diff --git a/cvnet_small_v1_global/checkpoint-1000/model.safetensors b/cvnet_small_v1_global/checkpoint-1000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..2d6cd2a75496a5b2e33c799d7d410384596d0e69
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94b64e1aea865196e9b19a03a0b7e3ac2eec8b15357819313767513259f8a27d
+size 237860264
diff --git a/cvnet_small_v1_global/checkpoint-1000/optimizer.pt b/cvnet_small_v1_global/checkpoint-1000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..bd4d9a49fecf912b711c2c1032847680cf430373
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3087e6b510df04a6a96e86fb6a37ae576f3f26f7ce8e9068e3c726c8440dd3f
+size 475804282
diff --git a/cvnet_small_v1_global/checkpoint-1000/rng_state_0.pth b/cvnet_small_v1_global/checkpoint-1000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..4ab31eb89973990d9b8610104399e7e5aaec5986
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f68045b77ef059786b9d730ec7bfcab3532c91c560ed23338e0d4d518b2be5e
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-1000/rng_state_1.pth b/cvnet_small_v1_global/checkpoint-1000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c53f9ece2cbb84f4662850cb2a534a064415698c
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:164a39302aced8efcfa4d5b883ab7d4f208ce2d3aaf57fa0fff0dc5bc35e9b53
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-1000/rng_state_2.pth b/cvnet_small_v1_global/checkpoint-1000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5513d1ce6c2aed658bc3b07da4002acb5b6ce7dc
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74928fb7ef7944af9aedd43d3456cb7cc320ccd38e5d7989ea54cf81f2bdc4ac
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-1000/rng_state_3.pth b/cvnet_small_v1_global/checkpoint-1000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..873724ab5b9b6f006d53045897c53d31e33804d8
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7390fd7db05b7bd4f309fe3c16e3133e51c0b7bbf36881c88fd513b9ebb7b75a
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-1000/rng_state_4.pth b/cvnet_small_v1_global/checkpoint-1000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..81ff0496e02da863836522e59c26181e5043ad3f
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:315c200f1cd4bb6a4a593accb8ba546796343253845919cb5b2b52089686830d
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-1000/rng_state_5.pth b/cvnet_small_v1_global/checkpoint-1000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..dde498f5924607b3543e00cedc5f8d8e7ead0775
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:561c368392fbdb6e8a7e5e086de59545625d80b821b5c72a4024b1f47b1390bb
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-1000/rng_state_6.pth b/cvnet_small_v1_global/checkpoint-1000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..004d54378694fe0b6ec4833301aba9fe98fda662
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56a358b4a84de8c0fefda163466d1b792e3d23a409d2f7f54fdd9b3535793d39
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-1000/rng_state_7.pth b/cvnet_small_v1_global/checkpoint-1000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e54c9d0e00105e7c7921fd6d4d170f670ee0b5a7
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17ad99f8730fc68cc1e3438467270d0c73427583fe0bffeb5bca595169156c2f
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-1000/scheduler.pt b/cvnet_small_v1_global/checkpoint-1000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9ff2c05006e51f3f771730c9055c6c181da2abd0
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c34ba9e6f50c2313df84e91c7dfed2be6b77dec7394adba34d3bd12d40d5ba2d
+size 1064
diff --git a/cvnet_small_v1_global/checkpoint-1000/trainer_state.json b/cvnet_small_v1_global/checkpoint-1000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..9464ad2cb3743f8522ca37b90c705ff3c670ddde
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/trainer_state.json
@@ -0,0 +1,621 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3373534620899047,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7081,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6774,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6781,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6772,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6802,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6745,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6729,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.636,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6098,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5978,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.602,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5778,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5795,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5883,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5882,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5747,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5846,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5823,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5722,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5599,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5832,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5691,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5711,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5821,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5557,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5581,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5761,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5643,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5702,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5648,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5691,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.556,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5578,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.564,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5562,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.559,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.565,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5508,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5588,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5584,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5555,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5597,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5483,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5621,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5568,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5527,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5524,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5458,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5524,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5553,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5582,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5518,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5485,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5444,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5522,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5417,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5543,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.56,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5394,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5475,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5437,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5438,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5454,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5481,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5434,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5422,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5465,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5447,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5396,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.543,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5461,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5482,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5453,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5493,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5477,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5365,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5433,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5368,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5467,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5382,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5354,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5322,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_global/checkpoint-1000/training_args.bin b/cvnet_small_v1_global/checkpoint-1000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..89dfdb803f04836e7afc0c6a28f9dae664c6e171
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-1000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dddfb786197274aad152ee4c2cc9787279a04f3c148146f1b57e85d4fb19f6fc
+size 4856
diff --git a/cvnet_small_v1_global/checkpoint-2000/model.safetensors b/cvnet_small_v1_global/checkpoint-2000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..12c9675d79b5f5ba6adba9943753084a0c317647
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbdd0ac1006e9b3741dea6c7d1cbaed5ad86865589792b5cd08c72bb887c128e
+size 237860264
diff --git a/cvnet_small_v1_global/checkpoint-2000/optimizer.pt b/cvnet_small_v1_global/checkpoint-2000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..7b83476b9b88f57d0b18da882a705f7881554fae
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3dcf359b402ab155e141b07b029eddf98e5e63df26e4e808b51a68a0fa44c1bb
+size 475804282
diff --git a/cvnet_small_v1_global/checkpoint-2000/rng_state_0.pth b/cvnet_small_v1_global/checkpoint-2000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..33d35d46eb49c7694ddf02abc1f806fe21ce3c03
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24545b6bb7493bdf35c20f30780f8e0793c67065421d64d7180f5a320abd8417
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-2000/rng_state_1.pth b/cvnet_small_v1_global/checkpoint-2000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..6310340e83ce52050038f3421730ede96037a6ba
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0232b218d0ca1f5487f91d43e6fb49bb92e0051351056294e6e73f362f38dc4
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-2000/rng_state_2.pth b/cvnet_small_v1_global/checkpoint-2000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5f2d7f68fb7fd5fa3b96e6ab8e2a6ed8ae0cc685
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3dffc06c0c6d363c38ffe8a64d78f552b7bfccf1d412a5e6d8b05a20badc3ab4
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-2000/rng_state_3.pth b/cvnet_small_v1_global/checkpoint-2000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ee633cf4e658e3f4b63beab3a61ea2a072f17879
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c47dc8892b7412ec8da71c7704256b96b8469c5c7f3b6490b83502e4b2ee928
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-2000/rng_state_4.pth b/cvnet_small_v1_global/checkpoint-2000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..a68c4918b64151b1245b3a97d067f09d0feb75a0
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:905b45076269c4efc0bfdb46513ea882d1de5bce7ddd5bed4ce98092c7da2c0b
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-2000/rng_state_5.pth b/cvnet_small_v1_global/checkpoint-2000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..37b3a6d3dd2e51bee05f2894dcebb9a4fe902f47
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97ee8e3283737d5d9165ab7bbed91fab8f8054e727830b0ac207e19ea1b998ca
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-2000/rng_state_6.pth b/cvnet_small_v1_global/checkpoint-2000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..24f763bcd01fe0ca32d12a74bd2f9a71337dd568
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7205ffddd0d11808dcecbed7a8e9b2051147bffa0bcf02b1ee5ef28b9e65162
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-2000/rng_state_7.pth b/cvnet_small_v1_global/checkpoint-2000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1a55221cfbb5c000b877e8e6b14977b7163966dc
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07bde508a1939450d8ec45b79bab2568f6dd5d0ac0cf2add96faab85481697da
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-2000/scheduler.pt b/cvnet_small_v1_global/checkpoint-2000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4605de17b07049a57c10990ac0db26645b375276
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:115fc5d3c71fc8b112d5c1701359950aa4e48ea559451a73de1d1d8ca1116d89
+size 1064
diff --git a/cvnet_small_v1_global/checkpoint-2000/trainer_state.json b/cvnet_small_v1_global/checkpoint-2000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..97d26beb7095a7619de2410acc29982f83ce9d6b
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/trainer_state.json
@@ -0,0 +1,1221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6747069241798094,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7081,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6774,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6781,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6772,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6802,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6745,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6729,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.636,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6098,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5978,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.602,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5778,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5795,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5883,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5882,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5747,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5846,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5823,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5722,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5599,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5832,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5691,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5711,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5821,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5557,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5581,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5761,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5643,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5702,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5648,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5691,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.556,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5578,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.564,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5562,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.559,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.565,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5508,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5588,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5584,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5555,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5597,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5483,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5621,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5568,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5527,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5524,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5458,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5524,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5553,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5582,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5518,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5485,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5444,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5522,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5417,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5543,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.56,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5394,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5475,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5437,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5438,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5454,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5481,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5434,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5422,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5465,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5447,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5396,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.543,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5461,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5482,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5453,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5493,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5477,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5365,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5433,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5368,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5467,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5382,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5354,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5322,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5379,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5398,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5337,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5384,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5256,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5353,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5233,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5483,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5362,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5347,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.5377,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.5374,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.532,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5315,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5274,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.537,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5427,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5288,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5259,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.531,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5263,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5305,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5245,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5306,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.526,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5308,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5226,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5295,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5297,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5293,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5252,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.524,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5248,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5332,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5206,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5237,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5201,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5336,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5282,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5269,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5245,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5203,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5311,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5229,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5266,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5283,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5279,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5267,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5238,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5249,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5226,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.525,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5249,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5281,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5204,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5193,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5278,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5224,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5218,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5239,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5164,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5181,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5166,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5179,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5169,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5275,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5179,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5216,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5236,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5125,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.527,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5224,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5304,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5242,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5351,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5262,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5178,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5172,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.517,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5154,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5285,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5228,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5131,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_global/checkpoint-2000/training_args.bin b/cvnet_small_v1_global/checkpoint-2000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..89dfdb803f04836e7afc0c6a28f9dae664c6e171
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-2000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dddfb786197274aad152ee4c2cc9787279a04f3c148146f1b57e85d4fb19f6fc
+size 4856
diff --git a/cvnet_small_v1_global/checkpoint-3000/model.safetensors b/cvnet_small_v1_global/checkpoint-3000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..f73a5307890fffc413b81ed0da91c17f7ac5988a
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6f885f65a815e27442367ce964bc2c0559b43681759c08c7f00bb28cb1d07e1
+size 237860264
diff --git a/cvnet_small_v1_global/checkpoint-3000/optimizer.pt b/cvnet_small_v1_global/checkpoint-3000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..fda4745001dc9ab2673cd3d8e8068f6e46e4f801
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97471d9b30876e8c0dbc7c483bcba3c96168b58bf31900a8605c5dc276a793de
+size 475804282
diff --git a/cvnet_small_v1_global/checkpoint-3000/rng_state_0.pth b/cvnet_small_v1_global/checkpoint-3000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cb8e78a2252e7489e47adb734bb1f370f80fb558
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7431609fc9d10c784464e14fe45f5a89b22573e0769c9c21d7d0a2dead49ec58
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-3000/rng_state_1.pth b/cvnet_small_v1_global/checkpoint-3000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..14409c37cb8c37cf95a2a40f01e3bde6c0a86a8f
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5081bec13a1a7ac2ad3917351c18540a7bae0cfef6e0df05d537d5c0163fce12
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-3000/rng_state_2.pth b/cvnet_small_v1_global/checkpoint-3000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..2e0781a6af83c0376c3bc340f290368a3cdf5f3b
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:535456fe6d77127c3883ae9c325dccec963de01449e321e9c704f78c50992939
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-3000/rng_state_3.pth b/cvnet_small_v1_global/checkpoint-3000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c3a155aae9463722acad62167de281b7e03ad81a
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:880e7012bf446334aa49027d1791b52f192a6dc11d44fea74095b153fd34bc97
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-3000/rng_state_4.pth b/cvnet_small_v1_global/checkpoint-3000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..50505327adf88ab24ab9f90d8faf76553bc4d2a9
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40038cca0e1cc4c41ee9eb36cfd70f6e868697faa2d2d79afba994d00387b725
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-3000/rng_state_5.pth b/cvnet_small_v1_global/checkpoint-3000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1db65e964f07df377b31f7937f3fad02cb5361f7
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86ec016c5b320b9210f7edc5e4b4d309aa8c19437ccbdf2fd68012814e7b0be6
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-3000/rng_state_6.pth b/cvnet_small_v1_global/checkpoint-3000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9d27f215baa695e5cea378d01affc7631aa9463d
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:049da174472b1bfd728972b360f9b2b80d3b4842c4c00b3019ff285417d7b3ca
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-3000/rng_state_7.pth b/cvnet_small_v1_global/checkpoint-3000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ae156e93c6ee694319aecdfa93bbb81b80472851
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2dcd919cf3c5f19cfa3f8a17ee1d2a7803a15af43f3917c55ec05ed6ca5858d
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-3000/scheduler.pt b/cvnet_small_v1_global/checkpoint-3000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0a885c8b891f90d887f6c40db2b8ae74f564dd98
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4df43d46412429a4bda78060914e68357cbee658ac819784da79443fe1ce6982
+size 1064
diff --git a/cvnet_small_v1_global/checkpoint-3000/trainer_state.json b/cvnet_small_v1_global/checkpoint-3000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..0f461f61fc933d6244beb15f69843dd8ebe46d4d
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/trainer_state.json
@@ -0,0 +1,1821 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0120603862697142,
+  "eval_steps": 500,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7081,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6774,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6781,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6772,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6802,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6745,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6729,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.636,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6098,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5978,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.602,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5778,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5795,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5883,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5882,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5747,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5846,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5823,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5722,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5599,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5832,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5691,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5711,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5821,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5557,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5581,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5761,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5643,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5702,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5648,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5691,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.556,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5578,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.564,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5562,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.559,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.565,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5508,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5588,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5584,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5555,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5597,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5483,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5621,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5568,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5527,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5524,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5458,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5524,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5553,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5582,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5518,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5485,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5444,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5522,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5417,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5543,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.56,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5394,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5475,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5437,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5438,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5454,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5481,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5434,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5422,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5465,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5447,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5396,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.543,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5461,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5482,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5453,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5493,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5477,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5365,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5433,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5368,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5467,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5382,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5354,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5322,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5379,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5398,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5337,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5384,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5256,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5353,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5233,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5483,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5362,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5347,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.5377,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.5374,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.532,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5315,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5274,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.537,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5427,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5288,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5259,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.531,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5263,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5305,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5245,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5306,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.526,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5308,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5226,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5295,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5297,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5293,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5252,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.524,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5248,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5332,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5206,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5237,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5201,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5336,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5282,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5269,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5245,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5203,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5311,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5229,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5266,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5283,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5279,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5267,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5238,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5249,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5226,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.525,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5249,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5281,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5204,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5193,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5278,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5224,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5218,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5239,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5164,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5181,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5166,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5179,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5169,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5275,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5179,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5216,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5236,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5125,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.527,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5224,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5304,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5242,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5351,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5262,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5178,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5172,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.517,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5154,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5285,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5228,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5131,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5173,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5137,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5132,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5154,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5194,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.521,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5166,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5161,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5143,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5133,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5099,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5181,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5148,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5155,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5108,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5026,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5187,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5156,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5164,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5129,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5143,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5206,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5101,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5069,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5082,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5206,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5099,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5114,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5132,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5125,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5055,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.514,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5093,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.506,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5198,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5187,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5112,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5189,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5105,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5165,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5088,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.504,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5065,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5053,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5078,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5038,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5114,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5055,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5011,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5113,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.506,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5031,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.505,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4966,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.5132,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4975,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.5108,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.514,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5071,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5021,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5112,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5106,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5011,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5103,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5004,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5045,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5019,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4928,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5029,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5175,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5114,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5021,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5084,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5008,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5102,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5133,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5054,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5191,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5051,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4943,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5044,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4943,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.504,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4978,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.492,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4997,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4968,
+      "step": 3000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_global/checkpoint-3000/training_args.bin b/cvnet_small_v1_global/checkpoint-3000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..89dfdb803f04836e7afc0c6a28f9dae664c6e171
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-3000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dddfb786197274aad152ee4c2cc9787279a04f3c148146f1b57e85d4fb19f6fc
+size 4856
diff --git a/cvnet_small_v1_global/checkpoint-4000/model.safetensors b/cvnet_small_v1_global/checkpoint-4000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..281414c4d9590f4d407fda3a5c744694b08804a5
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d25bb4e965b6951e2ee3ab893a0d0ce4998c4876a8469be3e67a9e30cbf24767
+size 237860264
diff --git a/cvnet_small_v1_global/checkpoint-4000/optimizer.pt b/cvnet_small_v1_global/checkpoint-4000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..74191f5642b9dce176111b6f6884111f164673d1
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb4094205d4f0d797c40d85e82a83e95c93d5a364494deadaafd79984544eee9
+size 475804282
diff --git a/cvnet_small_v1_global/checkpoint-4000/rng_state_0.pth b/cvnet_small_v1_global/checkpoint-4000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..27157ed6df83ce9d6c109c9b7dea4f945557c6c2
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30e9d80f2c7b48e823660edbe971433389c5987502ff5b8904c6fa06a4b8fac4
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-4000/rng_state_1.pth b/cvnet_small_v1_global/checkpoint-4000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d07f16bf0d35084ca8183acbe7d1ec6d1d90a220
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6dc64ef283107cc20517b99d8a4260d5dc35fe0739df81e0627d71f753f071e
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-4000/rng_state_2.pth b/cvnet_small_v1_global/checkpoint-4000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..a5a4f71f56747c23a569fdab1e672a231cf7beb8
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3984ffaba3ef211236ed7149a8baca5ccfdc9f355d0c4a2f94ee586ef97610e1
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-4000/rng_state_3.pth b/cvnet_small_v1_global/checkpoint-4000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..8bc5aa77c0a5136f2602b61102d9fa19e74049d1
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aef672fc70b0c0d7f41d3aa0c8a5fc28c8b3a305a6b083477713cb302d8d72a0
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-4000/rng_state_4.pth b/cvnet_small_v1_global/checkpoint-4000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..23ce76a0046bef79ca5a99524908ab2f3b052e03
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a0341c1736e01084f618482fe049b36f4ff83d3c592a9632213c226953a9084
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-4000/rng_state_5.pth b/cvnet_small_v1_global/checkpoint-4000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..78895c9a779442c5c0a14d8426eb7659ed5f69f5
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89b1c55b3a76909bc8b26c01f698db08cb90488d67ec9d13d047aea4788fd653
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-4000/rng_state_6.pth b/cvnet_small_v1_global/checkpoint-4000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1009d8998b054def1787038b788e28e743447e38
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73db3b651070332fced69b13785993fdefdaac849cbcb23e06c08cde7fdc5d55
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-4000/rng_state_7.pth b/cvnet_small_v1_global/checkpoint-4000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..642ce42df3671964f6b34b6543afd6eb286a0714
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f21a5acb7274574d65d51c6e1a0d8b70345cf7b9c9dd99e4b85e293546a6e40
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-4000/scheduler.pt b/cvnet_small_v1_global/checkpoint-4000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..503d2b402ef8df237c8826d9429a15068a9f8e5d
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1f5b5ac4a53f35c9923ba2e114d395c8f0ca2ff8283a0adc3fb6db30ab7f070
+size 1064
diff --git a/cvnet_small_v1_global/checkpoint-4000/trainer_state.json b/cvnet_small_v1_global/checkpoint-4000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..281cf43b7355f13d85eea0684f37ac254ce06d9f
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/trainer_state.json
@@ -0,0 +1,2421 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.3494138483596188,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7081,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6774,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6781,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6772,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6802,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6745,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6729,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.636,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6098,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5978,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.602,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5778,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5795,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5883,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5882,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5747,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5846,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5823,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5722,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5599,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5832,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5691,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5711,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5821,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5557,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5581,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5761,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5643,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5702,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5648,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5691,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.556,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5578,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.564,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5562,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.559,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.565,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5508,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5588,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5584,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5555,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5597,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5483,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5621,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5568,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5527,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5524,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5458,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5524,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5553,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5582,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5518,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5485,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5444,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5522,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5417,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5543,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.56,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5394,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5475,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5437,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5438,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5454,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5481,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5434,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5422,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5465,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5447,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5396,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.543,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5461,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5482,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5453,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5493,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5477,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5365,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5433,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5368,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5467,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5382,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5354,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5322,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5379,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5398,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5337,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5384,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5256,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5353,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5233,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5483,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5362,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5347,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.5377,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.5374,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.532,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5315,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5274,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.537,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5427,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5288,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5259,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.531,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5263,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5305,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5245,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5306,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.526,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5308,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5226,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5295,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5297,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5293,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5252,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.524,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5248,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5332,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5206,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5237,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5201,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5336,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5282,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5269,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5245,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5203,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5311,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5229,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5266,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5283,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5279,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5267,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5238,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5249,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5226,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.525,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5249,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5281,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5204,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5193,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5278,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5224,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5218,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5239,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5164,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5181,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5166,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5179,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5169,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5275,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5179,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5216,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5236,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5125,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.527,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5224,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5304,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5242,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5351,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5262,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5178,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5172,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.517,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5154,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5285,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5228,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5131,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5173,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5137,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5132,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5154,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5194,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.521,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5166,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5161,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5143,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5133,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5099,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5181,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5148,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5155,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5108,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5026,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5187,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5156,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5164,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5129,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5143,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5206,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5101,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5069,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5082,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5206,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5099,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5114,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5132,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5125,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5055,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.514,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5093,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.506,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5198,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5187,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5112,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5189,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5105,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5165,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5088,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.504,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5065,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5053,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5078,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5038,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5114,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5055,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5011,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5113,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.506,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5031,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.505,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4966,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.5132,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4975,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.5108,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.514,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5071,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5021,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5112,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5106,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5011,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5103,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5004,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5045,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5019,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4928,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5029,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5175,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5114,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5021,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5084,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5008,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5102,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5133,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5054,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5191,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5051,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4943,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5044,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4943,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.504,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4978,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.492,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4997,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4968,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4963,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5034,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5049,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5021,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.512,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5033,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4975,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5091,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4988,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4971,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5038,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4997,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5012,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4974,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.503,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.501,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4961,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4941,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4883,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5048,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4951,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.493,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4831,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4989,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4923,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4943,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4928,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4967,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4873,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4954,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.502,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5034,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4933,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4932,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4954,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4863,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4959,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4893,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4831,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4878,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4862,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4887,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4916,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4947,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4935,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4935,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4843,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.493,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4921,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4813,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4873,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4897,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4902,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4904,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4834,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4917,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.49,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4849,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4799,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4888,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4926,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4955,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4795,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4894,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4818,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4862,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4807,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4857,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4981,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4911,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4773,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4755,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.481,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4815,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4854,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4811,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4809,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4726,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4906,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4862,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.48,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4757,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.481,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_global/checkpoint-4000/training_args.bin b/cvnet_small_v1_global/checkpoint-4000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..89dfdb803f04836e7afc0c6a28f9dae664c6e171
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-4000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dddfb786197274aad152ee4c2cc9787279a04f3c148146f1b57e85d4fb19f6fc
+size 4856
diff --git a/cvnet_small_v1_global/checkpoint-5000/model.safetensors b/cvnet_small_v1_global/checkpoint-5000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..59a9ba0820f60821ccf988669213be6adbcd3ca2
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d9495d7a3d49a9be241c1af32a062b96e69e2fd6501e7bff8a7e3249bd4785a
+size 237860264
diff --git a/cvnet_small_v1_global/checkpoint-5000/optimizer.pt b/cvnet_small_v1_global/checkpoint-5000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5771daa7bdf85cf7eb24db25122e99a4cc54350d
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85aa11f57b0e71cb3040f095948ffe07c37a9c111f1418d99850c3412b5d22d6
+size 475804282
diff --git a/cvnet_small_v1_global/checkpoint-5000/rng_state_0.pth b/cvnet_small_v1_global/checkpoint-5000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..11e0c340c3c865d1230d06fc8147457636c36c7e
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fdb01d742970be957c0ece5354beb317869a9c3ca2df86c171b56533ec8f43a
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-5000/rng_state_1.pth b/cvnet_small_v1_global/checkpoint-5000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9dd478a8c9ea4164341bf243db8034288705a41e
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f451b776653c44b20c2f0aa86ce758d2f6b160a2aed1623e73fb9ce65aab45dc
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-5000/rng_state_2.pth b/cvnet_small_v1_global/checkpoint-5000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..64b4a78013a2d2da348987b2be5e0f0f9768f97d
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ec45f9a6fa09681ce69490ec8693348a5cd79d4e73833ec2c3c3616ad0cfb45
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-5000/rng_state_3.pth b/cvnet_small_v1_global/checkpoint-5000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1c64fb5462b0001c919706fb5b95cbb3049c2daf
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88c41930d7b79073d68661afecfe85fb979ce0c66aa50931c41c30c84094e01f
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-5000/rng_state_4.pth b/cvnet_small_v1_global/checkpoint-5000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..fe1ec346d9089256afe9858e9a678cc5a0f9e725
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f195114778c83dcdf347276c88ba35590abcc726622eff124827e1184d650b85
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-5000/rng_state_5.pth b/cvnet_small_v1_global/checkpoint-5000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..551227058e212b8b1b4c9946a46abd488003f429
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e790930ffa2fcd8eaf8532e85b32332d84d65708e9d2dbf5498070faacf67367
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-5000/rng_state_6.pth b/cvnet_small_v1_global/checkpoint-5000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..dff0e3a4547f927a68daf8bdea3c57bf24a69713
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dcc050ca3e609d48f5075b0221433e64883cc24379379e0f87b9bfc6fc692cd
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-5000/rng_state_7.pth b/cvnet_small_v1_global/checkpoint-5000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c4f9ceecc118db03d3225ab3c794d989fceb08fc
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ee5311f27f1199ba1ac2117100e5fb29127ebdb52900e88cf9b6a1d8cd6cd3a
+size 15920
diff --git a/cvnet_small_v1_global/checkpoint-5000/scheduler.pt b/cvnet_small_v1_global/checkpoint-5000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1cb6e935fff7477b8e4bab40ab97b33a5268355e
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5efa72feef3a398c4152d1d39792b540dbc6e38d02d8ab41bc4a583b249405d
+size 1064
diff --git a/cvnet_small_v1_global/checkpoint-5000/trainer_state.json b/cvnet_small_v1_global/checkpoint-5000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..9cff7ded2e6525cc3c48bb515e59b378d9ef3438
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/trainer_state.json
@@ -0,0 +1,3021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.6867673104495235,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7081,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6774,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6781,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6772,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6802,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6745,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6729,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.636,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6098,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5978,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.602,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5778,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5795,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5883,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5882,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5747,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5846,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5823,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5722,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5599,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5832,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5691,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5711,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5821,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5557,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5581,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5761,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5643,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5702,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5648,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5691,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.556,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5578,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.564,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5562,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.559,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.565,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5508,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5588,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5584,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5555,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5597,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5483,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5621,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5568,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5527,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5524,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5458,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5524,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5553,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5582,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5518,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5485,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5444,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5522,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5417,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5543,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.56,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5394,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5475,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5437,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5438,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5454,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5481,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5434,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5422,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5465,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5447,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5396,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.543,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5461,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5482,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5453,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5493,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5477,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5365,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5433,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5368,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.542,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5467,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5382,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5354,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5322,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5379,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5398,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5337,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5384,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5256,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5353,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5233,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5483,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5362,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5347,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.5377,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.5374,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.532,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5315,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5274,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.537,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5427,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5288,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5259,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.531,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5263,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5305,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5245,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5306,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.526,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5308,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5226,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5295,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5297,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5293,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5252,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.524,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5248,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5332,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5206,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5237,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5201,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5366,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5336,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5282,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5269,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5245,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5203,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5311,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5229,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5266,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5283,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5279,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5267,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5238,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5249,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5226,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.525,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5249,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5225,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5281,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5204,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5193,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5278,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5224,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5218,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5239,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5231,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5164,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5181,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5166,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5179,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5169,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5275,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5179,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5216,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5236,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5125,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.527,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5224,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5304,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5242,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5351,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5262,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5178,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5172,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.517,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5154,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5285,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5228,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5131,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5173,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5137,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5132,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5154,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5194,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.521,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5166,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5161,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5143,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5133,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5099,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5181,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5148,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5155,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5108,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5026,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5187,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5156,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5164,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5129,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5143,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5206,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5101,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5069,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5082,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5206,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5099,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5114,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5132,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5125,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5055,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.514,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5093,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.506,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5198,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5187,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5112,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5189,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5104,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5105,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5165,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5088,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.504,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5065,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5053,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5078,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5038,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5114,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5055,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5011,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5113,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.506,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5031,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.505,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4966,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.5132,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4975,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.5108,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.514,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5071,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5021,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5112,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5106,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5011,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5103,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5004,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5045,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5019,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4928,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5029,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5175,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5114,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5021,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5084,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5008,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5102,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5133,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.508,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5054,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5191,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5051,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4943,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5044,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4943,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.504,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4978,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.492,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4997,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4968,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4963,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5034,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5049,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5021,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.512,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5033,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4975,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5091,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4988,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4971,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5038,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4997,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5012,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4974,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.503,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.501,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4961,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4941,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4883,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5048,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4951,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.493,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4831,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4989,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4923,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4943,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4928,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4967,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4873,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4954,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.502,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5034,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4933,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4932,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4954,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4863,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4959,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4893,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4831,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4878,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4862,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4887,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4916,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4947,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4935,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4935,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4843,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.493,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4921,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4813,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4873,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4897,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4902,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4904,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4834,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4917,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.49,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4849,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4799,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4888,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4926,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4955,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4795,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4894,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4818,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4862,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4807,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4857,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4981,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4911,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4773,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4755,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.481,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4815,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4854,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4811,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4809,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4726,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4906,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4862,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.48,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4757,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.481,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4806,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4686,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4822,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4695,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4749,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.477,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4819,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4759,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4734,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4695,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4767,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4692,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4782,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4748,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4733,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4755,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4748,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.474,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4707,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4731,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4756,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4772,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4781,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.475,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4662,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4705,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4726,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4785,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4696,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4759,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4698,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4773,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.47,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4808,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4681,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4712,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4761,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4725,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.473,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4749,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4767,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4662,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4789,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4737,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.483,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4721,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4698,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4672,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4743,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4668,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4668,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4679,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4688,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4717,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4715,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4726,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4717,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.473,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4742,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4753,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4697,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4726,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4748,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4657,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.474,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4624,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4725,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4736,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4671,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4752,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4689,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4595,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4679,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4665,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4747,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.469,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4668,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4713,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4687,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.455,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4613,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4757,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4631,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4631,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4595,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4693,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4593,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4725,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4646,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4624,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4677,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_global/checkpoint-5000/training_args.bin b/cvnet_small_v1_global/checkpoint-5000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..89dfdb803f04836e7afc0c6a28f9dae664c6e171
--- /dev/null
+++ b/cvnet_small_v1_global/checkpoint-5000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dddfb786197274aad152ee4c2cc9787279a04f3c148146f1b57e85d4fb19f6fc
+size 4856
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/model.safetensors b/cvnet_small_v1_noglobal/checkpoint-1000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..d1b8a957464af7ea593df71c21592f4abb505264
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ee09c633670537ad32071c6f48f72aedcb3cb4b22de6141e667c924f9bd3abf
+size 231565528
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/optimizer.pt b/cvnet_small_v1_noglobal/checkpoint-1000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..182beb17de0b88daf4ee7f28e3877259fa756fae
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a5d473c72a379045a2e0f1e56c9af753f440c0392a002b4a5c1806f18fd5f73
+size 463213562
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_0.pth b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..76199e8cd59370e56d78379e7f0857d098932fdd
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20b275069de3c999174e0675baa7922d66052a0c898935ff704e23e3cef878fe
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_1.pth b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..db236d0038dbd238bc789a4e77c23f1457370409
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d09eab743b5b749ebce1ebcc0acafce88f180d92e6c1574df2cd2ec6493f284
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_2.pth b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..612080d1b966edb9c548b38c39066628aed041b5
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f10f3b4810f8be1a94c6a3f8e84e6d8d72c53f7a363a52cc624ecfd9cc5eb83
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_3.pth b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..55551738ab573081f638dfdb48a0106c65248f31
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5240772816e916400c0a6d093b395cc427e69eb6b82ad16a4cecc2332561b8af
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_4.pth b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..7cd9a50c7c25b4903ef43a29d7252dbb96151f5f
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03cb1f9d9d54e20989e075d4d0a8459635eb37e623741f99efb7d2427f4993b4
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_5.pth b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..924b34104933c59c34ca55f6744bd89d9f634e41
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:607b7a381bc052b0a9b35286824f0b454ed165491e7b87a3e74ed4c3336494bc
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_6.pth b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c178a16483d317ef9f4c0b585ff4cfc137bdd296
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae6d53d1c065b1230e113ea0f5e5404c6c649c2fcca72b82bef06b2e3f1200e1
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_7.pth b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d0baf74f7428518103c26e94815df83a9cafe655
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c32bd7774791739fb7c8a3c2e4d976b9b5b1d261f4312586bdf00cd5306c5794
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/scheduler.pt b/cvnet_small_v1_noglobal/checkpoint-1000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9ff2c05006e51f3f771730c9055c6c181da2abd0
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c34ba9e6f50c2313df84e91c7dfed2be6b77dec7394adba34d3bd12d40d5ba2d
+size 1064
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/trainer_state.json b/cvnet_small_v1_noglobal/checkpoint-1000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..41b306bf7753ba937b18fa386ef83f6268365346
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/trainer_state.json
@@ -0,0 +1,621 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3373534620899047,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7302,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6767,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6744,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6699,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6799,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6763,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.674,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6733,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.668,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6679,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6663,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6793,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6748,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6671,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6688,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6701,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6828,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.667,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6661,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6624,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6657,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.664,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6605,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6664,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6696,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.658,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6611,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6575,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6703,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.665,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6557,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.663,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6627,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.659,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6594,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6572,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6546,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.656,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6544,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6517,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6562,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6549,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6535,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6489,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6459,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6444,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6353,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6469,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.649,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6551,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6486,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal/checkpoint-1000/training_args.bin b/cvnet_small_v1_noglobal/checkpoint-1000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ddc854e996b1e5937c51594c3f01138b37a11f27
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-1000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e33b021e61b4db84dbe9910e89543e5342c7e95bc5945ad2e0f9f1fee840a78
+size 4856
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/model.safetensors b/cvnet_small_v1_noglobal/checkpoint-2000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..fd99606fde888ac987b991a1a8ea18a104bd2d4d
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36633585b0e17168614ca9692b64c7f3fea9f7946ac150d06c96c18bbb7047dd
+size 231565528
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/optimizer.pt b/cvnet_small_v1_noglobal/checkpoint-2000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..152a06fe22fa993bc537189d9eeb3469d86c5fb4
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38ac31606569293367a9da6568e5fc983e47a2329e0fce8a4e57cc5272b371bc
+size 463213562
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_0.pth b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..4986a8aac36333c2e13582ab4cc0e13f51a0e3c1
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:399a88ae31a12e7e3c780b5188c2dd2baf3b17cf459e26aa3c46eb48ac6b98fa
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_1.pth b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b3e8e99f59e43dbfee6761a8b163596d53d37a2b
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d11c4e0a8d7af0b51cdf0c49070411aa85a8ec60ebc52f77a275788f28703d46
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_2.pth b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..128a6430adf59dc9ae51af972ee34823bf35d745
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:602981e9d45a5154a05538455b041260754c018d792f1e6e664edfc3e0572feb
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_3.pth b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..8a638077e62343a306f380ba3c92cb3eb2f6c05b
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0a6236572b94e610082d6971859606d884a94f9265f9ddd8545182ffc6cc636
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_4.pth b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d23874a128181fbca324694c9acc20fcb5e40998
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa496dbe24521e796da8d4b2a1507ea4c55cda06ae1c7336e5dd0396a4559190
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_5.pth b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9e98e7cfc462af0bc887d61d6706110ea3f76c11
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:004586c0da87a64f32d6edcb395860503348b22a799755ca3fb2314708a49c7e
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_6.pth b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..40a36a2eda59b410e2ef79790cee3180f801fee8
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abbafa128291a4aa7088b96a35651dcc22e18ab4b22698d7e9ee3f40e398147d
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_7.pth b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..467f22eb368091fd8bf28f2a71eb7a725fa85d41
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b18a8aafa7fd87ce9ec0c644422f207e743a3863e733fbfbb45374297107d5af
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/scheduler.pt b/cvnet_small_v1_noglobal/checkpoint-2000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4605de17b07049a57c10990ac0db26645b375276
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:115fc5d3c71fc8b112d5c1701359950aa4e48ea559451a73de1d1d8ca1116d89
+size 1064
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/trainer_state.json b/cvnet_small_v1_noglobal/checkpoint-2000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..fc9338585adfb20354421820e901b5d1a2479c5c
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/trainer_state.json
@@ -0,0 +1,1221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6747069241798094,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7302,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6767,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6744,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6699,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6799,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6763,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.674,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6733,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.668,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6679,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6663,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6793,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6748,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6671,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6688,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6701,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6828,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.667,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6661,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6624,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6657,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.664,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6605,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6664,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6696,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.658,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6611,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6575,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6703,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.665,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6557,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.663,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6627,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.659,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6594,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6572,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6546,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.656,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6544,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6517,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6562,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6549,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6535,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6489,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6459,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6444,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6353,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6469,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.649,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6551,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6486,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6395,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6378,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6614,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6449,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6443,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6411,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.642,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6364,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.65,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.645,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6401,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6452,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6334,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6395,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6422,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6421,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6434,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6417,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6404,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.646,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6752,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6394,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6308,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal/checkpoint-2000/training_args.bin b/cvnet_small_v1_noglobal/checkpoint-2000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ddc854e996b1e5937c51594c3f01138b37a11f27
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-2000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e33b021e61b4db84dbe9910e89543e5342c7e95bc5945ad2e0f9f1fee840a78
+size 4856
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/model.safetensors b/cvnet_small_v1_noglobal/checkpoint-3000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..ed6c1a40a819a6fd1bde4e580e0d9435d4d8d9c7
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d688504fbd162544f39051f7bf7f722ab0b172635e68f4bac8af1af0d6d58300
+size 231565528
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/optimizer.pt b/cvnet_small_v1_noglobal/checkpoint-3000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..01da1dab0299b3a3ddc09395f27ce82d417e13ea
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:daa82394884bf57dfca1b54d42fdcf921c3c1a028333ec1e680a0a04ba9f694a
+size 463213562
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_0.pth b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..6279c89246d3ac79f9433ce1aa3c10763c6ab1de
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f066bac4e29b58cbd92ad08aebed608a49ce7ec8b8a08d89b29cd19eaa6526e1
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_1.pth b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..0d7ac0de90e233d9281952b42edad13d015b1f0f
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb4d6e21ee9e308a02de7f522194a6d3b8b29550b32bccb7b8f66ac8c0f18ccb
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_2.pth b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..554425bcf5d398faa6ea7d5da15ab29d5b0e641a
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21f4fb5dab010c6c7d1a6d363b3da29775b73f0d66b80ff6beb870b648827933
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_3.pth b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e94745960de1137c6534584cb54cff62b9babf23
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c67fb1b5eb9abc7454d2105b449747cd44bdcb81a729fbe025d4b657fa734669
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_4.pth b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..0d2d0ccd4eda40390432c3f449afd1bfc8388f78
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32b7ebffea3f19a0a27c942a6bdac48a361468d1d42ff8818c4b30b3e8efa91f
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_5.pth b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ddbc33f84b15728f6c193dc691f895ccc9806728
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb4144ca487b6f9fcbc6ab7fd4b4d135a1c75f1ab342fd8e9be48da4e8ebe68b
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_6.pth b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..2178437bf7e585fe72d165f15e0b103f82981938
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:270a519b5b8f8590b94caed17210ea4d103747c2f9c69c6214d89fa23be4ac65
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_7.pth b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..deeedfae0fb52086053eeadab93961e81a68fdad
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d85449fcca7e93c9c3fbc9d9a9c4305741c156b214ad2e628d107917a899673
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/scheduler.pt b/cvnet_small_v1_noglobal/checkpoint-3000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0a885c8b891f90d887f6c40db2b8ae74f564dd98
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4df43d46412429a4bda78060914e68357cbee658ac819784da79443fe1ce6982
+size 1064
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/trainer_state.json b/cvnet_small_v1_noglobal/checkpoint-3000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a4cf76c1b24b1c1d3d014871cb7a41909373ec1
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/trainer_state.json
@@ -0,0 +1,1821 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0120603862697142,
+  "eval_steps": 500,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7302,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6767,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6744,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6699,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6799,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6763,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.674,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6733,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.668,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6679,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6663,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6793,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6748,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6671,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6688,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6701,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6828,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.667,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6661,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6624,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6657,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.664,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6605,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6664,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6696,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.658,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6611,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6575,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6703,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.665,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6557,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.663,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6627,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.659,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6594,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6572,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6546,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.656,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6544,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6517,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6562,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6549,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6535,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6489,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6459,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6444,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6353,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6469,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.649,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6551,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6486,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6395,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6378,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6614,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6449,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6443,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6411,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.642,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6364,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.65,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.645,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6401,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6452,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6334,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6395,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6422,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6421,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6434,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6417,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6404,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.646,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6752,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6394,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6308,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6417,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6363,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6294,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6286,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6282,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6296,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6646,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6334,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6287,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6286,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6397,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6281,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6397,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6317,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6272,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6269,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6184,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6244,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.62,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6424,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6368,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6177,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6245,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6539,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.623,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6254,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.626,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6309,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.62,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6271,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6172,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6169,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6165,
+      "step": 3000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal/checkpoint-3000/training_args.bin b/cvnet_small_v1_noglobal/checkpoint-3000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ddc854e996b1e5937c51594c3f01138b37a11f27
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-3000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e33b021e61b4db84dbe9910e89543e5342c7e95bc5945ad2e0f9f1fee840a78
+size 4856
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/model.safetensors b/cvnet_small_v1_noglobal/checkpoint-4000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..fae19eb9f643878678aa6b2cd33bfb6e90012872
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0a2818292ffcfb576e45a296de056ce7265d665ff9e40af04198f15926bcfae
+size 231565528
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/optimizer.pt b/cvnet_small_v1_noglobal/checkpoint-4000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b2457b4b5272963a868ab799ff9e9306764bcf46
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c901728a29230568247ae52bc6d870d69d35f0401b47e33e4c58df44f422d5a5
+size 463213562
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_0.pth b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..99493da2e5370a66dbed7827e1d7ba6410e8828d
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f26c3d1768c6ae4c474def0c137b69314ae8c72e465f6f95a800af794a2a09e
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_1.pth b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..fcc291e14a12baedccc1431ea7a9430e1b8c0e72
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea3ca7f799a3e1fd36dc83bd11dc3d9b99112d7064e93ea65eaa391986ef5cd6
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_2.pth b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9d99186aaa674d522b45d36579003eb151a96fa7
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ccfd3472e4e4cfcd011db16885b574284403045f147713f913c42e21104c294
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_3.pth b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cf177a48bdceb75cd47590d3dc4590deb2b704ae
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49ee0246ba535c4ea39226df08b91cfa3a9cb800772ae18a4e932595aafae3f3
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_4.pth b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..453a7e71ab9e0cdde170e3e2136d3597466543b3
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3c3a5050775f7ed2cb135b9675df1194aef481dc07c07e8786a990008c237d2
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_5.pth b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..6f1123cc148666196af551f7e7c508b3b7d5188a
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac45cf7b32e235682cc63de0934056563e30db7b9fb88500545c0fd6933f7ae4
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_6.pth b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..7a461608b4dcf872f8b375665b3b295abb61f341
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6550a2c59c5405e5668721566e21841986fa2fca1e14cac0ba7bd09a4c6c6d87
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_7.pth b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..dac85a2304e36e9a8808904608cf8511947a8270
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9397f0988a7295194f8742f890e6c203ae61671c0c74b2e5b3c0f3ab4fe9c03b
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/scheduler.pt b/cvnet_small_v1_noglobal/checkpoint-4000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..503d2b402ef8df237c8826d9429a15068a9f8e5d
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1f5b5ac4a53f35c9923ba2e114d395c8f0ca2ff8283a0adc3fb6db30ab7f070
+size 1064
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/trainer_state.json b/cvnet_small_v1_noglobal/checkpoint-4000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..46deeff68f10f6f46ecb976514bcc54121117bc7
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/trainer_state.json
@@ -0,0 +1,2421 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.3494138483596188,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7302,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6767,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6744,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6699,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6799,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6763,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.674,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6733,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.668,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6679,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6663,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6793,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6748,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6671,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6688,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6701,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6828,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.667,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6661,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6624,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6657,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.664,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6605,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6664,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6696,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.658,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6611,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6575,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6703,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.665,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6557,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.663,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6627,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.659,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6594,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6572,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6546,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.656,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6544,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6517,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6562,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6549,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6535,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6489,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6459,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6444,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6353,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6469,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.649,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6551,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6486,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6395,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6378,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6614,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6449,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6443,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6411,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.642,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6364,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.65,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.645,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6401,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6452,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6334,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6395,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6422,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6421,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6434,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6417,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6404,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.646,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6752,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6394,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6308,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6417,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6363,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6294,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6286,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6282,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6296,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6646,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6334,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6287,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6286,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6397,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6281,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6397,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6317,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6272,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6269,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6184,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6244,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.62,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6424,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6368,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6177,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6245,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6539,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.623,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6254,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.626,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6309,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.62,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6271,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6172,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6169,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6165,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6196,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6436,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6169,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6254,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6287,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6256,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6162,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6154,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6164,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6169,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6179,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6254,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6172,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6154,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6198,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6141,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6285,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6147,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6215,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6186,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6221,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6183,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6183,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6201,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6144,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6136,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6186,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6179,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6251,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6091,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6199,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6111,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6179,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.611,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.611,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6113,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6218,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6192,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6057,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.615,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6166,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6122,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.616,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal/checkpoint-4000/training_args.bin b/cvnet_small_v1_noglobal/checkpoint-4000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ddc854e996b1e5937c51594c3f01138b37a11f27
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-4000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e33b021e61b4db84dbe9910e89543e5342c7e95bc5945ad2e0f9f1fee840a78
+size 4856
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/model.safetensors b/cvnet_small_v1_noglobal/checkpoint-5000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..cc6d979a85f15f7a52c91535ff7abfae2093d9b5
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f787ac129d749961894bfafe697f66d13f573d5f38b432e59d9288f5d8e08b55
+size 231565528
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/optimizer.pt b/cvnet_small_v1_noglobal/checkpoint-5000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5f510502d840d9acfc9205a602082dcdc5c90cd3
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:707c1a4e5b6b8d2ccb9187696f9dbcf8bed3f25718bee672fa20cdf3d01ee54c
+size 463213562
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_0.pth b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..39677673ea6d33ece6c7e54d42dda570337ef321
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cafb0e9235f8047bc55b0b621a4e93d80fc892e63ee3da03af79fc7f2e23579c
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_1.pth b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c12058346ac872ccdb8f02d78f4563b6f47ae847
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d28fb3cde459b36c79b37d7ed796adba69b956c9e9e4bf18151be2743f515ce
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_2.pth b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d1e634c602498a40b047d3b1bb9640f53ddc31d6
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be30b240d749c0c071986c83b1160d241a87f1b44562a3e04f2c33d1b0deba77
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_3.pth b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..119a7adf2ae3effe42deccb69216298ffd0ae4c5
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d15ee6cb6f002f026de91076d83dd67712e8d59d6e35203594152ccd458c034
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_4.pth b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..abcf417a906294784464e6ff4d87caff7743d479
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3540bcc17b69b78275876e9c411218fe0e7d9e9c456a543c715fddaf72580b12
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_5.pth b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d6b847665195bf326e14d3f49401daec9be86fd4
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5577daf3bb02d1cdf478e7569ab5f72486e526bd3f0a7dfaafd5b876c20b233c
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_6.pth b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..de5876b2c74a579837ca196c64dadccca8f794aa
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5c6b0fa46b2cd29408ac0acf133813f28d691d2c969970edd02d477c516b9de
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_7.pth b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e39ef30cc80b584ca184856225d68fea94b5904f
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:119af508b7160eed58791eb97d94873e53469668209aaa0085013a4450bcaf5f
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/scheduler.pt b/cvnet_small_v1_noglobal/checkpoint-5000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1cb6e935fff7477b8e4bab40ab97b33a5268355e
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5efa72feef3a398c4152d1d39792b540dbc6e38d02d8ab41bc4a583b249405d
+size 1064
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/trainer_state.json b/cvnet_small_v1_noglobal/checkpoint-5000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..c0db2e7a7962ade00795af9d37c617f44073e1c6
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/trainer_state.json
@@ -0,0 +1,3021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.6867673104495235,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7302,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6767,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6744,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6699,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6799,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6763,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.674,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6733,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.668,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6679,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6663,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6793,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6748,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6671,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6688,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6701,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6828,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.667,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6661,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6624,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6657,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.664,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6605,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6664,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6696,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.658,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6611,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6575,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6703,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.665,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6557,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.663,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6627,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.659,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6594,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6572,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6546,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.656,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6544,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6517,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6562,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6549,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6535,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6489,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6459,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6444,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6353,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6469,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.649,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6551,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6486,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6395,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6378,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6614,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6449,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6443,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6411,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.642,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6364,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.65,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.645,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6401,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6452,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6334,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6395,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6422,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6421,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6434,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6417,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6404,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.646,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6752,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6394,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6308,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6417,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6363,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6294,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6286,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6282,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6296,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6646,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6334,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6287,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6286,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6397,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6281,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6397,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6317,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6272,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6269,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6184,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6244,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.62,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6424,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6368,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6177,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6245,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6539,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.623,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6254,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.626,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6309,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.62,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6271,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6172,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6169,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6165,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6196,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6436,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6169,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6254,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6287,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6256,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6162,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6154,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6164,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6169,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6179,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6254,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6172,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6154,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6198,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6141,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6285,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6147,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6215,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6186,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6221,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6183,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6183,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6201,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6144,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6136,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6186,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6179,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6251,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6091,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6199,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6111,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6179,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.611,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.611,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6113,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6218,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6192,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6057,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.615,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6166,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6122,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.616,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6044,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6113,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6134,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6099,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6168,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6127,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.611,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.601,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.611,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6081,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6101,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6031,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6109,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6045,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6099,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6165,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6076,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.617,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6109,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6132,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6126,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6063,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6052,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6044,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5983,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6113,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6079,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6128,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6085,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6161,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6099,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.613,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6106,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6166,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6099,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6216,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6115,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.608,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6052,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6163,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6088,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6118,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6044,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.614,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6215,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6018,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.615,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.605,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6069,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6127,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5996,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6091,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6133,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6057,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6117,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6083,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal/checkpoint-5000/training_args.bin b/cvnet_small_v1_noglobal/checkpoint-5000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ddc854e996b1e5937c51594c3f01138b37a11f27
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-5000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e33b021e61b4db84dbe9910e89543e5342c7e95bc5945ad2e0f9f1fee840a78
+size 4856
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/model.safetensors b/cvnet_small_v1_noglobal/checkpoint-6000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..68cebcf1d677a76c95c3f66c82899fc7725601f2
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c13daa9b98407416c717a5bfe1e85c9a851537397f95f7ef180a5edf56505aa
+size 231565528
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/optimizer.pt b/cvnet_small_v1_noglobal/checkpoint-6000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c9601183f85dc288cd5621621adc06c2739f0197
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f89cb3d65ea08823e01a4e530a9c6c4f3c7cdd4dd460191eafae7900256851e8
+size 463213562
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_0.pth b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ddd59c5c15b5a55e7a3987f420003a1e30c4da49
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e2bce1b939178bdb91cb1f874a7a000b36cf3be74fa8948eb1676bc6975d384
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_1.pth b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..120cf5fd288918e83403c3cf97aaa6c39448c409
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fd052c2e662b89e3963ab8505c968beabf71be9e15da0b0f6052abee3c9f435
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_2.pth b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1492ebad21f46552c2ff152143ba94df35d21ece
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d35d0d9bd0209acb9ce919c49f55de09c5cbf806f05e2fb1298640387aef6d5
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_3.pth b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..f83b7ba8355058c7be166f6cbf59d4539450907a
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4deb81c597430038ccd03b227752c64d8d38d92686235a61f2e148d6258ee9c
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_4.pth b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5453ac3bb6ed37aa8c478debc89795d2db1f0dae
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bfd68fceaf4d06e82547852230652ef63243b906e535881a62b0ffcb13ab0eb4
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_5.pth b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..3a65bc2ad9e7b7165847a88c1f97572ab2321e1b
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c84cd4836653a84a020c85b2ebfa25193e9127ef7a20cbec7ac46fab556269db
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_6.pth b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..88ef8e5ae0ae6144d3bb7be2cab5bce5f887b80f
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc4e4217f50cef231a5e408ea5f4ef762ced599fe76d8ce39bc4fc3531204ae3
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_7.pth b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..82bffb3c13d50406bfe64ab0b3555186d63ae810
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4b914c17166db505f1b1a01880e439a87ca31848337c59df04dafea9f7a9bea
+size 15920
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/scheduler.pt b/cvnet_small_v1_noglobal/checkpoint-6000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2a19f49862de9ff057d5dbf1579e3849486e1ddb
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c891c6464db0fd4c307d9cb1087a0312b5de2d7c04375cf97f0f59aa2882e71c
+size 1064
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/trainer_state.json b/cvnet_small_v1_noglobal/checkpoint-6000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..fe3b7d98f3d882871e998d8ff752b585a3c5f4d0
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/trainer_state.json
@@ -0,0 +1,3621 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0241207725394283,
+  "eval_steps": 500,
+  "global_step": 6000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7302,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6767,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6744,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6699,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6799,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6763,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.674,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6733,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.668,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6679,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6663,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6793,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6749,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6748,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6722,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6671,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6688,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6701,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6828,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6716,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6706,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.667,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6725,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6661,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6624,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6657,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.664,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6605,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6635,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6664,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6696,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.658,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6611,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6575,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6703,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.665,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6557,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.663,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6627,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6586,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.659,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6594,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6572,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6546,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.656,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6544,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6528,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6574,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6517,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6562,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6549,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6535,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6489,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6459,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6444,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6353,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6566,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6469,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.649,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6551,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6512,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6486,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6617,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6395,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6378,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.648,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6614,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6449,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6443,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6411,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.642,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6364,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.65,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.645,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6401,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6452,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6334,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6395,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6422,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6421,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6434,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6417,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6404,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.646,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6752,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6394,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6308,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6396,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6417,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6363,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6294,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6286,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6282,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6296,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6646,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6307,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6334,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6287,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6286,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6397,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6281,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6397,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6317,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6272,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6269,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6184,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6244,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.62,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6424,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6368,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6346,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6177,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6245,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6539,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.623,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.632,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6254,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.626,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6309,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.62,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6271,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6336,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6172,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6169,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6165,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6196,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6436,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6441,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6169,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6254,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6287,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6256,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6162,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6154,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6164,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6169,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6179,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6254,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6172,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6154,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6198,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6141,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6285,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6147,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6215,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6186,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6221,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6183,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6183,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6201,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6144,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6136,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6186,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.633,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6179,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6251,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6091,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6199,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6111,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6179,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.611,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.611,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6113,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6218,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6192,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6057,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.615,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6166,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6122,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.616,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6044,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6113,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6134,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6099,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6168,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6127,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.611,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.601,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.611,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6081,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6101,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6031,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6109,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6045,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6099,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6165,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6076,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.617,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6109,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6132,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6126,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6063,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6052,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6044,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5983,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6113,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6079,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6128,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6085,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6161,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6099,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.613,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6106,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6166,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6099,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6216,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6115,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.608,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6052,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6163,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6088,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6118,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6044,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.614,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6215,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6018,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.615,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.605,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6069,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6127,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5996,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6091,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6133,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6057,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6117,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6083,
+      "step": 5000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5989,
+      "step": 5010
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6041,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.617,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5998,
+      "step": 5050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6172,
+      "step": 5060
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6036,
+      "step": 5070
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6039,
+      "step": 5080
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5963,
+      "step": 5090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5983,
+      "step": 5100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6009,
+      "step": 5110
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 5120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.598,
+      "step": 5130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6172,
+      "step": 5140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6118,
+      "step": 5150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6066,
+      "step": 5160
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 5170
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.614,
+      "step": 5180
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6066,
+      "step": 5190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6128,
+      "step": 5200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 5210
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5964,
+      "step": 5220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6077,
+      "step": 5230
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6115,
+      "step": 5240
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 5250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5997,
+      "step": 5260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6037,
+      "step": 5270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5995,
+      "step": 5280
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6005,
+      "step": 5290
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6009,
+      "step": 5300
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6007,
+      "step": 5310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6097,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6013,
+      "step": 5350
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6004,
+      "step": 5360
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5915,
+      "step": 5370
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.603,
+      "step": 5380
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6016,
+      "step": 5390
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6006,
+      "step": 5400
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6079,
+      "step": 5410
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5957,
+      "step": 5420
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6034,
+      "step": 5430
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6019,
+      "step": 5440
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5969,
+      "step": 5450
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6236,
+      "step": 5460
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6127,
+      "step": 5470
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5989,
+      "step": 5480
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6068,
+      "step": 5490
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6005,
+      "step": 5500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5998,
+      "step": 5510
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 5520
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5973,
+      "step": 5530
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6003,
+      "step": 5540
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.602,
+      "step": 5550
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6076,
+      "step": 5560
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 5570
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6116,
+      "step": 5580
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6015,
+      "step": 5590
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.599,
+      "step": 5600
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6077,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6082,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6118,
+      "step": 5640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6106,
+      "step": 5650
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6216,
+      "step": 5660
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5945,
+      "step": 5670
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5996,
+      "step": 5680
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6057,
+      "step": 5690
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6102,
+      "step": 5700
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5976,
+      "step": 5710
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5969,
+      "step": 5720
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6012,
+      "step": 5730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 5740
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5999,
+      "step": 5750
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5989,
+      "step": 5760
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6155,
+      "step": 5770
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6164,
+      "step": 5780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5993,
+      "step": 5790
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5874,
+      "step": 5800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5907,
+      "step": 5810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5989,
+      "step": 5820
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6055,
+      "step": 5830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5989,
+      "step": 5840
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5932,
+      "step": 5850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5994,
+      "step": 5860
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 5870
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6036,
+      "step": 5880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6149,
+      "step": 5890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6261,
+      "step": 5900
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 5910
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5974,
+      "step": 5920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6069,
+      "step": 5930
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5981,
+      "step": 5940
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6081,
+      "step": 5950
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6016,
+      "step": 5960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6003,
+      "step": 5970
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6016,
+      "step": 5980
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6022,
+      "step": 5990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5956,
+      "step": 6000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal/checkpoint-6000/training_args.bin b/cvnet_small_v1_noglobal/checkpoint-6000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ddc854e996b1e5937c51594c3f01138b37a11f27
--- /dev/null
+++ b/cvnet_small_v1_noglobal/checkpoint-6000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e33b021e61b4db84dbe9910e89543e5342c7e95bc5945ad2e0f9f1fee840a78
+size 4856
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/model.safetensors b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..368eeaefc888c407e3fbffa50f9d0a5772bf21a8
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:936bbc48893a3c243c74e36640e5bbafa8749420a2404b899e8b4186c930eee4
+size 231565528
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/optimizer.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..7790424cacc9034ed3e464bb39dfa56898d47c74
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73b765d50a2fdb6ce5b4d3b395bcc2e9007df0363d89b3e7d6df4ca3d2258f59
+size 463213562
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_0.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..4ab31eb89973990d9b8610104399e7e5aaec5986
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f68045b77ef059786b9d730ec7bfcab3532c91c560ed23338e0d4d518b2be5e
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_1.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c53f9ece2cbb84f4662850cb2a534a064415698c
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:164a39302aced8efcfa4d5b883ab7d4f208ce2d3aaf57fa0fff0dc5bc35e9b53
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_2.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5513d1ce6c2aed658bc3b07da4002acb5b6ce7dc
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74928fb7ef7944af9aedd43d3456cb7cc320ccd38e5d7989ea54cf81f2bdc4ac
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_3.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..873724ab5b9b6f006d53045897c53d31e33804d8
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7390fd7db05b7bd4f309fe3c16e3133e51c0b7bbf36881c88fd513b9ebb7b75a
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_4.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..81ff0496e02da863836522e59c26181e5043ad3f
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:315c200f1cd4bb6a4a593accb8ba546796343253845919cb5b2b52089686830d
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_5.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..dde498f5924607b3543e00cedc5f8d8e7ead0775
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:561c368392fbdb6e8a7e5e086de59545625d80b821b5c72a4024b1f47b1390bb
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_6.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..004d54378694fe0b6ec4833301aba9fe98fda662
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56a358b4a84de8c0fefda163466d1b792e3d23a409d2f7f54fdd9b3535793d39
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_7.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e54c9d0e00105e7c7921fd6d4d170f670ee0b5a7
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17ad99f8730fc68cc1e3438467270d0c73427583fe0bffeb5bca595169156c2f
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/scheduler.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9ff2c05006e51f3f771730c9055c6c181da2abd0
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c34ba9e6f50c2313df84e91c7dfed2be6b77dec7394adba34d3bd12d40d5ba2d
+size 1064
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/trainer_state.json b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..47414831e09a6ed28a726fb93ce021890ba59c3c
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/trainer_state.json
@@ -0,0 +1,621 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3373534620899047,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7047,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6771,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.676,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6731,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6726,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6734,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6683,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6684,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6675,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6806,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6743,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6761,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6714,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6651,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6759,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6698,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6687,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6604,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6655,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6666,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6647,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6612,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6607,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6596,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6631,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6728,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6484,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6628,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.655,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6496,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6445,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6482,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6473,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.651,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6508,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6541,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6405,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6533,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6416,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6507,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6481,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6442,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6418,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.637,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/training_args.bin b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..76c6e8e44c071d2b5c47ee143f33b3b0c3f9fe92
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-1000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee26f62127373b4e313bacaf1f87bc66c9583ad97e93c127ff53a983dbadb481
+size 4856
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/model.safetensors b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..1d35c30d20e50f06177edb17f013b47f1eb5dfb1
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48a72f5940f8e43d49a57b55ad43a818ac37cd2211cd93ce0d848cc170bf3773
+size 231565528
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/optimizer.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..984155bc7b1807045b6c9bd81d29088056de3770
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cb5526677174a85abc57b556ba718a00d467a978752bf5471b14ce804eba771
+size 463213562
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_0.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..33d35d46eb49c7694ddf02abc1f806fe21ce3c03
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24545b6bb7493bdf35c20f30780f8e0793c67065421d64d7180f5a320abd8417
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_1.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..6310340e83ce52050038f3421730ede96037a6ba
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0232b218d0ca1f5487f91d43e6fb49bb92e0051351056294e6e73f362f38dc4
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_2.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5f2d7f68fb7fd5fa3b96e6ab8e2a6ed8ae0cc685
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3dffc06c0c6d363c38ffe8a64d78f552b7bfccf1d412a5e6d8b05a20badc3ab4
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_3.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ee633cf4e658e3f4b63beab3a61ea2a072f17879
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c47dc8892b7412ec8da71c7704256b96b8469c5c7f3b6490b83502e4b2ee928
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_4.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..a68c4918b64151b1245b3a97d067f09d0feb75a0
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:905b45076269c4efc0bfdb46513ea882d1de5bce7ddd5bed4ce98092c7da2c0b
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_5.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..37b3a6d3dd2e51bee05f2894dcebb9a4fe902f47
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97ee8e3283737d5d9165ab7bbed91fab8f8054e727830b0ac207e19ea1b998ca
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_6.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..24f763bcd01fe0ca32d12a74bd2f9a71337dd568
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7205ffddd0d11808dcecbed7a8e9b2051147bffa0bcf02b1ee5ef28b9e65162
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_7.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1a55221cfbb5c000b877e8e6b14977b7163966dc
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07bde508a1939450d8ec45b79bab2568f6dd5d0ac0cf2add96faab85481697da
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/scheduler.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4605de17b07049a57c10990ac0db26645b375276
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:115fc5d3c71fc8b112d5c1701359950aa4e48ea559451a73de1d1d8ca1116d89
+size 1064
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/trainer_state.json b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..983cf3c63d861fff9a53997510e4eefcd5d62ff5
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/trainer_state.json
@@ -0,0 +1,1221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6747069241798094,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7047,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6771,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.676,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6731,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6726,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6734,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6683,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6684,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6675,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6806,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6743,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6761,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6714,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6651,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6759,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6698,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6687,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6604,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6655,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6666,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6647,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6612,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6607,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6596,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6631,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6728,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6484,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6628,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.655,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6496,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6445,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6482,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6473,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.651,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6508,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6541,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6405,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6533,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6416,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6507,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6481,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6442,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6418,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.637,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6427,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6413,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6374,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6365,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6386,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6304,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6303,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6466,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6323,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.638,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6653,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6297,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6532,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6344,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6398,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6354,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6313,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/training_args.bin b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..76c6e8e44c071d2b5c47ee143f33b3b0c3f9fe92
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-2000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee26f62127373b4e313bacaf1f87bc66c9583ad97e93c127ff53a983dbadb481
+size 4856
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/model.safetensors b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..6fd944fe8cf63721714c351ea86044086138ded0
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c4f3f7bc7cfe4d62fba38a407fcadae0e37b7fa1b2a681226dbfb3a3888a34c
+size 231565528
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/optimizer.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..51932223f52edcdc969e29f2ab99cda182dde289
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:51f34a7b9911f5a43a65b50192839b4bda6d94d2973a75104786e8e37bc85721
+size 463213562
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_0.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cb8e78a2252e7489e47adb734bb1f370f80fb558
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7431609fc9d10c784464e14fe45f5a89b22573e0769c9c21d7d0a2dead49ec58
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_1.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..14409c37cb8c37cf95a2a40f01e3bde6c0a86a8f
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5081bec13a1a7ac2ad3917351c18540a7bae0cfef6e0df05d537d5c0163fce12
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_2.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..2e0781a6af83c0376c3bc340f290368a3cdf5f3b
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:535456fe6d77127c3883ae9c325dccec963de01449e321e9c704f78c50992939
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_3.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c3a155aae9463722acad62167de281b7e03ad81a
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:880e7012bf446334aa49027d1791b52f192a6dc11d44fea74095b153fd34bc97
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_4.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..50505327adf88ab24ab9f90d8faf76553bc4d2a9
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40038cca0e1cc4c41ee9eb36cfd70f6e868697faa2d2d79afba994d00387b725
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_5.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1db65e964f07df377b31f7937f3fad02cb5361f7
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86ec016c5b320b9210f7edc5e4b4d309aa8c19437ccbdf2fd68012814e7b0be6
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_6.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9d27f215baa695e5cea378d01affc7631aa9463d
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:049da174472b1bfd728972b360f9b2b80d3b4842c4c00b3019ff285417d7b3ca
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_7.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ae156e93c6ee694319aecdfa93bbb81b80472851
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2dcd919cf3c5f19cfa3f8a17ee1d2a7803a15af43f3917c55ec05ed6ca5858d
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/scheduler.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0a885c8b891f90d887f6c40db2b8ae74f564dd98
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4df43d46412429a4bda78060914e68357cbee658ac819784da79443fe1ce6982
+size 1064
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/trainer_state.json b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..2eabfca3fc8876c65cb7b8564d6949d43550114f
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/trainer_state.json
@@ -0,0 +1,1821 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0120603862697142,
+  "eval_steps": 500,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7047,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6771,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.676,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6731,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6726,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6734,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6683,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6684,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6675,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6806,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6743,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6761,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6714,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6651,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6759,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6698,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6687,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6604,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6655,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6666,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6647,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6612,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6607,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6596,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6631,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6728,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6484,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6628,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.655,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6496,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6445,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6482,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6473,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.651,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6508,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6541,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6405,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6533,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6416,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6507,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6481,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6442,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6418,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.637,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6427,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6413,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6374,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6365,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6386,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6304,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6303,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6466,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6323,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.638,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6653,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6297,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6532,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6344,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6398,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6354,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6313,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6337,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6292,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.634,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6277,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6324,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6306,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6382,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6358,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6199,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6446,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6209,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6141,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6249,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6259,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.622,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6366,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6414,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6218,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6362,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6221,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/training_args.bin b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..76c6e8e44c071d2b5c47ee143f33b3b0c3f9fe92
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-3000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee26f62127373b4e313bacaf1f87bc66c9583ad97e93c127ff53a983dbadb481
+size 4856
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/model.safetensors b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..9239f70972b9b1aee3747b256f4acf8301b2d9c5
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9060a0b79baf213a69af7c32b37dc43593dcdce701c552cd3e079c08bd20b190
+size 231565528
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/optimizer.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a4fc1582c82d2755d01be2e645a86d5460185ae6
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd21c43f0fe0f026c757b0425a0b4eaa3f1249e9d06c6bc88194c7dca92928d0
+size 463213562
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_0.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..27157ed6df83ce9d6c109c9b7dea4f945557c6c2
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30e9d80f2c7b48e823660edbe971433389c5987502ff5b8904c6fa06a4b8fac4
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_1.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d07f16bf0d35084ca8183acbe7d1ec6d1d90a220
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6dc64ef283107cc20517b99d8a4260d5dc35fe0739df81e0627d71f753f071e
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_2.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..a5a4f71f56747c23a569fdab1e672a231cf7beb8
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3984ffaba3ef211236ed7149a8baca5ccfdc9f355d0c4a2f94ee586ef97610e1
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_3.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..8bc5aa77c0a5136f2602b61102d9fa19e74049d1
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aef672fc70b0c0d7f41d3aa0c8a5fc28c8b3a305a6b083477713cb302d8d72a0
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_4.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..23ce76a0046bef79ca5a99524908ab2f3b052e03
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a0341c1736e01084f618482fe049b36f4ff83d3c592a9632213c226953a9084
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_5.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..78895c9a779442c5c0a14d8426eb7659ed5f69f5
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89b1c55b3a76909bc8b26c01f698db08cb90488d67ec9d13d047aea4788fd653
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_6.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1009d8998b054def1787038b788e28e743447e38
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73db3b651070332fced69b13785993fdefdaac849cbcb23e06c08cde7fdc5d55
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_7.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..642ce42df3671964f6b34b6543afd6eb286a0714
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f21a5acb7274574d65d51c6e1a0d8b70345cf7b9c9dd99e4b85e293546a6e40
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/scheduler.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..503d2b402ef8df237c8826d9429a15068a9f8e5d
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1f5b5ac4a53f35c9923ba2e114d395c8f0ca2ff8283a0adc3fb6db30ab7f070
+size 1064
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/trainer_state.json b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..6d8026acf4abf1c04906e50fd084e1fa7123437e
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/trainer_state.json
@@ -0,0 +1,2421 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.3494138483596188,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7047,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6771,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.676,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6731,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6726,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6734,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6683,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6684,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6675,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6806,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6743,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6761,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6714,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6651,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6759,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6698,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6687,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6604,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6655,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6666,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6647,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6612,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6607,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6596,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6631,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6728,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6484,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6628,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.655,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6496,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6445,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6482,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6473,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.651,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6508,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6541,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6405,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6533,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6416,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6507,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6481,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6442,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6418,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.637,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6427,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6413,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6374,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6365,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6386,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6304,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6303,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6466,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6323,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.638,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6653,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6297,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6532,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6344,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6398,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6354,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6313,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6337,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6292,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.634,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6277,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6324,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6306,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6382,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6358,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6199,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6446,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6209,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6141,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6249,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6259,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.622,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6366,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6414,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6218,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6362,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6221,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6205,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.612,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6117,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6171,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6238,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6143,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6243,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6131,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6202,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6081,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6082,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6251,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6201,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6088,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6144,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6155,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6111,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6164,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6109,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6216,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6049,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6193,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6135,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6037,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/training_args.bin b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..76c6e8e44c071d2b5c47ee143f33b3b0c3f9fe92
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-4000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee26f62127373b4e313bacaf1f87bc66c9583ad97e93c127ff53a983dbadb481
+size 4856
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/model.safetensors b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..05a9e0f529fe0265c1899a5ac7e8d9d3b7e97164
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:534383927a64da5f7d6d7c482ca72d50d0a6151259b79422d21686d1e27f15d4
+size 231565528
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/optimizer.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..bf9163571635c1f5e158ec5cc80dc97251f8f611
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50a79fb3b066c830961f4271527eb5029a4bd988b5562dc93ade9cbfbd569e83
+size 463213562
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_0.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..11e0c340c3c865d1230d06fc8147457636c36c7e
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fdb01d742970be957c0ece5354beb317869a9c3ca2df86c171b56533ec8f43a
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_1.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9dd478a8c9ea4164341bf243db8034288705a41e
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f451b776653c44b20c2f0aa86ce758d2f6b160a2aed1623e73fb9ce65aab45dc
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_2.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..64b4a78013a2d2da348987b2be5e0f0f9768f97d
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ec45f9a6fa09681ce69490ec8693348a5cd79d4e73833ec2c3c3616ad0cfb45
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_3.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1c64fb5462b0001c919706fb5b95cbb3049c2daf
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88c41930d7b79073d68661afecfe85fb979ce0c66aa50931c41c30c84094e01f
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_4.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..fe1ec346d9089256afe9858e9a678cc5a0f9e725
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f195114778c83dcdf347276c88ba35590abcc726622eff124827e1184d650b85
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_5.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..551227058e212b8b1b4c9946a46abd488003f429
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e790930ffa2fcd8eaf8532e85b32332d84d65708e9d2dbf5498070faacf67367
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_6.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..dff0e3a4547f927a68daf8bdea3c57bf24a69713
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dcc050ca3e609d48f5075b0221433e64883cc24379379e0f87b9bfc6fc692cd
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_7.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c4f9ceecc118db03d3225ab3c794d989fceb08fc
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ee5311f27f1199ba1ac2117100e5fb29127ebdb52900e88cf9b6a1d8cd6cd3a
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/scheduler.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1cb6e935fff7477b8e4bab40ab97b33a5268355e
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5efa72feef3a398c4152d1d39792b540dbc6e38d02d8ab41bc4a583b249405d
+size 1064
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/trainer_state.json b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..e30c74581a1a035afce4978c3eb859bbc14f88fc
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/trainer_state.json
@@ -0,0 +1,3021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.6867673104495235,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7047,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6771,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.676,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6731,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6726,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6734,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6683,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6684,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6675,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6806,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6743,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6761,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6714,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6651,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6759,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6698,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6687,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6604,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6655,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6666,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6647,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6612,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6607,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6596,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6631,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6728,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6484,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6628,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.655,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6496,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6445,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6482,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6473,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.651,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6508,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6541,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6405,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6533,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6416,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6507,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6481,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6442,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6418,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.637,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6427,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6413,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6374,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6365,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6386,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6304,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6303,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6466,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6323,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.638,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6653,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6297,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6532,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6344,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6398,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6354,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6313,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6337,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6292,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.634,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6277,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6324,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6306,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6382,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6358,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6199,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6446,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6209,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6141,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6249,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6259,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.622,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6366,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6414,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6218,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6362,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6221,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6205,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.612,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6117,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6171,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6238,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6143,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6243,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6131,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6202,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6081,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6082,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6251,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6201,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6088,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6144,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6155,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6111,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6164,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6109,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6216,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6049,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6193,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6135,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6037,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6162,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6253,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6086,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6123,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6095,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.623,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6115,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6085,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6012,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6042,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6091,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6116,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6108,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6175,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6086,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.617,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6146,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.613,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5997,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6056,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6211,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6097,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6177,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6126,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.603,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6016,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6083,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6128,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6101,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5993,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6063,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5992,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6025,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6022,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6065,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6032,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.61,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.605,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/training_args.bin b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..76c6e8e44c071d2b5c47ee143f33b3b0c3f9fe92
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-5000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee26f62127373b4e313bacaf1f87bc66c9583ad97e93c127ff53a983dbadb481
+size 4856
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/model.safetensors b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..fdf578d811eee8f047f7d6690e96940268f42d2b
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de9e2b537175fba00dba9d9f77023284a722877e58b0614c9902334258c2a531
+size 231565528
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/optimizer.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6d977b5bcca5fd385fc1c4292d44185fd97f33a7
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e1e3956d7c24864bf9055686cfc0c8ab1f4b6209440413f437201740b299957
+size 463213562
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_0.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..573734be361b33074bad7976bc01de749f5d6479
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab4e74bbbd65d4c41a2a3bb1ea3fab4232793342026d4ffddf16cc632eae7a62
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_1.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..00dad64f42e12d0afa1e60caca6210040dc49131
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a697538c03af5ff8d6452dc15809a2894197ffa5681aa24edbcffc60e254a44f
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_2.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c9ae0f8d5f1ebe6f385347e0bbf38ca560f444b7
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fbf72c1542dcec16af36c0222d3bbd5999db9343d956cbbc8773f3f66517526
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_3.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cb0bf81d6134a09c791172c3ca07fa2b29f739b5
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75a68317944e9f214474713c9c1db897f06cab2c5a4c170c0c5134a6091190b5
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_4.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..dc85f134ea3fbc6bc5d8d8a23706f5d0fe08b3fd
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cca38c6e7daab0237ed6da434a7df41d9f170d92b83d3a756bf26c5e8ec11fd
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_5.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ef8acbd251c3c21ed20e9316184f5ddb5b7c29e2
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4b8da61059b60a1e04b5439ccfded49d90ee8196f887e88b6a19571fb41b8ce
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_6.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5e9fed676d4b294296d8083ffa689e570dc39075
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6285187171ba9b778016af44ec47ceb282740a4dd23535d002b80bf5b2bdb97
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_7.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..4afd09a3df8f4d942ac7d014c7a0b8fecf1a0423
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6473ee5bdf6c7a45259c4f355f40e16b480317db5b2b9d24efa31f9bf9d22202
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/scheduler.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2a19f49862de9ff057d5dbf1579e3849486e1ddb
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c891c6464db0fd4c307d9cb1087a0312b5de2d7c04375cf97f0f59aa2882e71c
+size 1064
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/trainer_state.json b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f0bcb3a292c748e2c12134be5b564457f31056e
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/trainer_state.json
@@ -0,0 +1,3621 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0241207725394283,
+  "eval_steps": 500,
+  "global_step": 6000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7047,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6771,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.676,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6731,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6726,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6734,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6683,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6684,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6675,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6806,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6743,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6761,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6714,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6651,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6759,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6698,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6687,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6604,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6655,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6666,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6647,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6612,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6607,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6596,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6631,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6728,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6484,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6628,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.655,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6496,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6445,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6482,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6473,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.651,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6508,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6541,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6405,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6533,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6416,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6507,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6481,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6442,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6418,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.637,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6427,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6413,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6374,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6365,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6386,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6304,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6303,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6466,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6323,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.638,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6653,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6297,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6532,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6344,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6398,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6354,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6313,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6337,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6292,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.634,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6277,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6324,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6306,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6382,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6358,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6199,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6446,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6209,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6141,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6249,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6259,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.622,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6366,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6414,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6218,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6362,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6221,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6205,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.612,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6117,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6171,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6238,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6143,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6243,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6131,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6202,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6081,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6082,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6251,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6201,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6088,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6144,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6155,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6111,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6164,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6109,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6216,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6049,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6193,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6135,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6037,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6162,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6253,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6086,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6123,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6095,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.623,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6115,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6085,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6012,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6042,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6091,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6116,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6108,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6175,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6086,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.617,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6146,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.613,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5997,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6056,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6211,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6097,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6177,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6126,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.603,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6016,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6083,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6128,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6101,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5993,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6063,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5992,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6025,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6022,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6065,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6032,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.61,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.605,
+      "step": 5000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5961,
+      "step": 5010
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5957,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6069,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5924,
+      "step": 5050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6059,
+      "step": 5060
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5999,
+      "step": 5070
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5981,
+      "step": 5080
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5953,
+      "step": 5090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5974,
+      "step": 5100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5941,
+      "step": 5110
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 5120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5919,
+      "step": 5130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 5140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5972,
+      "step": 5150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 5160
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6066,
+      "step": 5170
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6047,
+      "step": 5180
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 5190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6077,
+      "step": 5200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5961,
+      "step": 5210
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.593,
+      "step": 5220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6012,
+      "step": 5230
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 5240
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5976,
+      "step": 5250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 5260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5945,
+      "step": 5270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6092,
+      "step": 5280
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6062,
+      "step": 5290
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6021,
+      "step": 5300
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5988,
+      "step": 5310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6035,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6163,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6015,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5948,
+      "step": 5350
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 5360
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5871,
+      "step": 5370
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5958,
+      "step": 5380
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5939,
+      "step": 5390
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5985,
+      "step": 5400
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6004,
+      "step": 5410
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5908,
+      "step": 5420
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6019,
+      "step": 5430
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 5440
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.602,
+      "step": 5450
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6098,
+      "step": 5460
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6045,
+      "step": 5470
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5921,
+      "step": 5480
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5962,
+      "step": 5490
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5969,
+      "step": 5500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5975,
+      "step": 5510
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5998,
+      "step": 5520
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6049,
+      "step": 5530
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5540
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.599,
+      "step": 5550
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 5560
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6019,
+      "step": 5570
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.601,
+      "step": 5580
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.592,
+      "step": 5590
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.5925,
+      "step": 5600
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5952,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 5640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5987,
+      "step": 5650
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6125,
+      "step": 5660
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5889,
+      "step": 5670
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5944,
+      "step": 5680
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5690
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5973,
+      "step": 5700
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6004,
+      "step": 5710
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5989,
+      "step": 5720
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5983,
+      "step": 5730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6079,
+      "step": 5740
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6028,
+      "step": 5750
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5953,
+      "step": 5760
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6108,
+      "step": 5770
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 5780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5943,
+      "step": 5790
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5845,
+      "step": 5800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5952,
+      "step": 5810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 5820
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5987,
+      "step": 5830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5895,
+      "step": 5840
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6007,
+      "step": 5850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 5860
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6054,
+      "step": 5870
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5896,
+      "step": 5880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.598,
+      "step": 5890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5963,
+      "step": 5900
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5910
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5944,
+      "step": 5920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 5930
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6001,
+      "step": 5940
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5975,
+      "step": 5950
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5864,
+      "step": 5960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5899,
+      "step": 5970
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 5980
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5957,
+      "step": 5990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5888,
+      "step": 6000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/training_args.bin b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..76c6e8e44c071d2b5c47ee143f33b3b0c3f9fe92
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-6000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee26f62127373b4e313bacaf1f87bc66c9583ad97e93c127ff53a983dbadb481
+size 4856
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/model.safetensors b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..c8c00dd25bb175e9ffa589c17429fcd10ece3327
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:862730b4ea6b5f5205c5f7949cb2cae6b0e1bad04fe2ceeefdb066bf683f282b
+size 231565528
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/optimizer.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..85afc1d33265d6500c56a1593146ee563d4db6d4
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da7f4dee1f0577b85b72fa874a88c8fbcb1db4236741544bbe0613ce4cb09a55
+size 463213562
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_0.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..45df2fc5a65b8ed8addaca82738706bcbc3d354d
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79070614809b14d7d9ec529b08959bdf21acd157981c8790f18646e727ecb7a2
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_1.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c948306a0faaf3e0501f52348767111b70b59cd7
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1271e97eea72dbeacaa4e28a8ba68c9001897356682fed9d29d54f495006853d
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_2.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..57db45639fee1b370933ee6c97f112bfaac04f90
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:271fc7f1ca0ec9742ec537789e746b0401ed98834a2288108d94cb14a694f803
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_3.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..22fa7388ac0e625e289ce37e6b79dd608755fcf3
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ef2b67b03c9810b08b56e4947a089f8d8ab7b89b86348255ecd2cd8c779631b
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_4.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..8e99c104b1f483d5eda22b12f56398f6de67e26d
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:936416ff61c500377528f245d1c7100ba09e42993fe5edd62b64fc08b0ae180a
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_5.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..2ef2b61a9a61c20ff816eb7d53d5a100b8a34071
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56af862adbec6e2b62670042d928d0eefc403b58310fa35f2528aa926c375ec7
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_6.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..6d351974ca9fca1dda7126f22f4ee7f42fab3fb7
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6cbc2fc39c7d03cec9648447b91c124a625c578e8c93e9870d9bdb8892f049a6
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_7.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..6d475c0fc82bccf2e2063e998260e0a996e01701
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63b8eff2583ba40500598c1be7bd1e1875b644238c72d057fe1345d8c310b5ac
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/scheduler.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f80910e440c97c6830904084432a446f7a721cd5
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5445e128a7bcf4177cc50c7224415006e388ca1019002cdd1a4b59301d72556c
+size 1064
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/trainer_state.json b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..0a25e6d75451865595727c4fe69bbdb12fa1bc26
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/trainer_state.json
@@ -0,0 +1,4221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.3614742346293327,
+  "eval_steps": 500,
+  "global_step": 7000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7047,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6771,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.676,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6731,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6726,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6734,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6683,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6684,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6675,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6806,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6743,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6761,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6714,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6651,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6759,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6698,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6687,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6604,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6655,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6666,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6647,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6612,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6607,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6596,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6631,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6728,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6484,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6628,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.655,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6496,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6445,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6482,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6473,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.651,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6508,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6541,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6405,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6533,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6416,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6507,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6481,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6442,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6418,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.637,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6427,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6413,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6374,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6365,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6386,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6304,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6303,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6466,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6323,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.638,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6653,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6297,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6532,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6344,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6398,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6354,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6313,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6337,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6292,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.634,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6277,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6324,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6306,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6382,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6358,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6199,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6446,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6209,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6141,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6249,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6259,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.622,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6366,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6414,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6218,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6362,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6221,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6205,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.612,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6117,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6171,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6238,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6143,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6243,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6131,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6202,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6081,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6082,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6251,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6201,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6088,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6144,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6155,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6111,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6164,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6109,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6216,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6049,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6193,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6135,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6037,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6162,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6253,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6086,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6123,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6095,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.623,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6115,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6085,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6012,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6042,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6091,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6116,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6108,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6175,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6086,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.617,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6146,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.613,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5997,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6056,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6211,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6097,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6177,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6126,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.603,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6016,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6083,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6128,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6101,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5993,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6063,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5992,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6025,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6022,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6065,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6032,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.61,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.605,
+      "step": 5000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5961,
+      "step": 5010
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5957,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6069,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5924,
+      "step": 5050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6059,
+      "step": 5060
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5999,
+      "step": 5070
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5981,
+      "step": 5080
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5953,
+      "step": 5090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5974,
+      "step": 5100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5941,
+      "step": 5110
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 5120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5919,
+      "step": 5130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 5140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5972,
+      "step": 5150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 5160
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6066,
+      "step": 5170
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6047,
+      "step": 5180
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 5190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6077,
+      "step": 5200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5961,
+      "step": 5210
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.593,
+      "step": 5220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6012,
+      "step": 5230
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 5240
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5976,
+      "step": 5250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 5260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5945,
+      "step": 5270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6092,
+      "step": 5280
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6062,
+      "step": 5290
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6021,
+      "step": 5300
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5988,
+      "step": 5310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6035,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6163,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6015,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5948,
+      "step": 5350
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 5360
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5871,
+      "step": 5370
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5958,
+      "step": 5380
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5939,
+      "step": 5390
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5985,
+      "step": 5400
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6004,
+      "step": 5410
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5908,
+      "step": 5420
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6019,
+      "step": 5430
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 5440
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.602,
+      "step": 5450
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6098,
+      "step": 5460
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6045,
+      "step": 5470
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5921,
+      "step": 5480
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5962,
+      "step": 5490
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5969,
+      "step": 5500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5975,
+      "step": 5510
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5998,
+      "step": 5520
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6049,
+      "step": 5530
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5540
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.599,
+      "step": 5550
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 5560
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6019,
+      "step": 5570
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.601,
+      "step": 5580
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.592,
+      "step": 5590
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.5925,
+      "step": 5600
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5952,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 5640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5987,
+      "step": 5650
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6125,
+      "step": 5660
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5889,
+      "step": 5670
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5944,
+      "step": 5680
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5690
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5973,
+      "step": 5700
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6004,
+      "step": 5710
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5989,
+      "step": 5720
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5983,
+      "step": 5730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6079,
+      "step": 5740
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6028,
+      "step": 5750
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5953,
+      "step": 5760
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6108,
+      "step": 5770
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 5780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5943,
+      "step": 5790
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5845,
+      "step": 5800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5952,
+      "step": 5810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 5820
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5987,
+      "step": 5830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5895,
+      "step": 5840
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6007,
+      "step": 5850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 5860
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6054,
+      "step": 5870
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5896,
+      "step": 5880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.598,
+      "step": 5890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5963,
+      "step": 5900
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5910
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5944,
+      "step": 5920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 5930
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6001,
+      "step": 5940
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5975,
+      "step": 5950
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5864,
+      "step": 5960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5899,
+      "step": 5970
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 5980
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5957,
+      "step": 5990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5888,
+      "step": 6000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5994,
+      "step": 6010
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5914,
+      "step": 6020
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5981,
+      "step": 6030
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5807,
+      "step": 6040
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.594,
+      "step": 6050
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5983,
+      "step": 6060
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5895,
+      "step": 6070
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5896,
+      "step": 6080
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5949,
+      "step": 6090
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.579,
+      "step": 6100
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.591,
+      "step": 6110
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5881,
+      "step": 6120
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6055,
+      "step": 6130
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6037,
+      "step": 6140
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5881,
+      "step": 6150
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5937,
+      "step": 6160
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5962,
+      "step": 6170
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5838,
+      "step": 6180
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6035,
+      "step": 6190
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5993,
+      "step": 6200
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6014,
+      "step": 6210
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5968,
+      "step": 6220
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 6230
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6003,
+      "step": 6240
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5939,
+      "step": 6250
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5855,
+      "step": 6260
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5876,
+      "step": 6270
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 6280
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5845,
+      "step": 6290
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5929,
+      "step": 6300
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 6310
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5938,
+      "step": 6320
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5935,
+      "step": 6330
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5904,
+      "step": 6340
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6015,
+      "step": 6350
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.59,
+      "step": 6360
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5894,
+      "step": 6370
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6066,
+      "step": 6380
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5921,
+      "step": 6390
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 6400
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5919,
+      "step": 6410
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 6420
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5948,
+      "step": 6430
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5858,
+      "step": 6440
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5927,
+      "step": 6450
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5877,
+      "step": 6460
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5874,
+      "step": 6470
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5934,
+      "step": 6480
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 6490
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5994,
+      "step": 6500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5986,
+      "step": 6510
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5863,
+      "step": 6520
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5844,
+      "step": 6530
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5891,
+      "step": 6540
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5848,
+      "step": 6550
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6097,
+      "step": 6560
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 6570
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5913,
+      "step": 6580
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5933,
+      "step": 6590
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 6600
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5811,
+      "step": 6610
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5866,
+      "step": 6620
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 6630
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5871,
+      "step": 6640
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5849,
+      "step": 6650
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5923,
+      "step": 6660
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.59,
+      "step": 6670
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5842,
+      "step": 6680
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5902,
+      "step": 6690
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5971,
+      "step": 6700
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 6710
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5885,
+      "step": 6720
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5924,
+      "step": 6730
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5837,
+      "step": 6740
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.588,
+      "step": 6750
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5903,
+      "step": 6760
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5896,
+      "step": 6770
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5885,
+      "step": 6780
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.588,
+      "step": 6790
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 6800
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 6810
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5811,
+      "step": 6820
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5904,
+      "step": 6830
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5817,
+      "step": 6840
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 6850
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5847,
+      "step": 6860
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 6870
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5886,
+      "step": 6880
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5828,
+      "step": 6890
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 6900
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5843,
+      "step": 6910
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5907,
+      "step": 6920
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5802,
+      "step": 6930
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5883,
+      "step": 6940
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5842,
+      "step": 6950
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5908,
+      "step": 6960
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.593,
+      "step": 6970
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5913,
+      "step": 6980
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5846,
+      "step": 6990
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5924,
+      "step": 7000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/training_args.bin b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..76c6e8e44c071d2b5c47ee143f33b3b0c3f9fe92
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-7000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee26f62127373b4e313bacaf1f87bc66c9583ad97e93c127ff53a983dbadb481
+size 4856
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/model.safetensors b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..0e0850bbb425b37c4f9346a502c70413b2bb7409
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05a0f07a7b809d891043275cfe2c44f080df708c5c52942fa8a6bab8fa8cd2e5
+size 231565528
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/optimizer.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..86154cd7cd88e4b843a431b83bade155f3dfeb86
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d76f5d5f819c3e9fe6615fe5e9f51ac9a949720c85e58d4c85886bd28b14f595
+size 463213562
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_0.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e1a7000b2a2a9b7e66995194c7def83701672681
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6041fbe6d64344a7314c873a27b77bc02d1482287c212202188d36f4518a660d
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_1.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cca238fc6d5fb18fc6b45ee6a12e0b6439381ca2
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d98371ac0922b4275c5af92449fd0f7232b9afa2afae6edaeff681f991c72d33
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_2.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..3ea9d02214ccd6e3e63d6f099324d649a3685139
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:780c0966ec7854b6c6e39f3e8fa4117f78c3e4680730e9fa98df60d0fe7eab92
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_3.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e3324706fc72c3c4eb41b160a595cc8c3920f031
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c663a5160d7288107012a869bab934436414514919f633809e6413d97c7ad945
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_4.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..474dbcb768ff8961068af90d14d5a6e7761b2bda
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ecb80974b256cc4c9e5cf15046168299cd65b34a55bc197240e42ed1c885ae5
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_5.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c8a6ddebeed1dfa08a5e502a3e3253075dcc5eca
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d98f2e800c82fc9025786118eb439c72f5f75939349d70c7f996ee05d9efbc0d
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_6.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ba103507ac4db9853bb68e7c0c4c9864ee383c1b
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb1e4bb44c061e0bde43201c6a526de0aa0db51d6e379cb474dc0df9403bef4a
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_7.pth b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b78fa1dc7fa2a0a928683c007abd9004fb520aba
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f30a1cb01c4953e01f8a0855c963b5ad50f1bb2cc19a235a31b0d397d966d70c
+size 15920
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/scheduler.pt b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0bc1a1fc7418b56600e0a35fa744b78dec1110ac
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33657ad014a83329319798b77914fe9b3b1f8d598a6a6cf79c7c92f3a50382cc
+size 1064
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/trainer_state.json b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..43a9feab656ad652111b03f8e1cbe38e0d9bf80f
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/trainer_state.json
@@ -0,0 +1,4821 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.6988276967192375,
+  "eval_steps": 500,
+  "global_step": 8000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7047,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6771,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.676,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6731,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6726,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6734,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6683,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6684,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6675,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6806,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6743,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6761,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6714,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6651,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6759,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6698,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6687,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6604,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6655,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6666,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6647,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6612,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6607,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6596,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6631,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6728,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6484,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6628,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.655,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6496,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6445,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6482,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6473,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.651,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6508,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6541,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6405,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6533,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6416,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6507,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6481,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6442,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6418,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.637,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6427,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6413,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6374,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6365,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6386,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6304,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6303,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6466,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6323,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.638,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6653,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6297,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6532,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6344,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6398,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6354,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6313,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6337,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6292,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.634,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6277,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6324,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6306,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6382,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6358,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6199,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6446,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6209,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6141,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6249,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6259,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.622,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6366,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6414,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6218,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6362,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6221,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6205,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.612,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6117,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6171,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6238,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6143,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6243,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6131,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6202,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6081,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6082,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6251,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6201,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6088,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6144,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6155,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6111,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6164,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6109,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6216,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6049,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6193,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6135,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6037,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6162,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6253,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6086,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6123,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6095,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.623,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6115,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6085,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6012,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6042,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6091,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6116,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6108,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6175,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6086,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.617,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6146,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.613,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5997,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6056,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6211,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6097,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6177,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6126,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.603,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6016,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6083,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6128,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6101,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5993,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6063,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5992,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6025,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6022,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6065,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6032,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.61,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.605,
+      "step": 5000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5961,
+      "step": 5010
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5957,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6069,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5924,
+      "step": 5050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6059,
+      "step": 5060
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5999,
+      "step": 5070
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5981,
+      "step": 5080
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5953,
+      "step": 5090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5974,
+      "step": 5100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5941,
+      "step": 5110
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 5120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5919,
+      "step": 5130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 5140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5972,
+      "step": 5150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 5160
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6066,
+      "step": 5170
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6047,
+      "step": 5180
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 5190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6077,
+      "step": 5200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5961,
+      "step": 5210
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.593,
+      "step": 5220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6012,
+      "step": 5230
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 5240
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5976,
+      "step": 5250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 5260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5945,
+      "step": 5270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6092,
+      "step": 5280
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6062,
+      "step": 5290
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6021,
+      "step": 5300
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5988,
+      "step": 5310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6035,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6163,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6015,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5948,
+      "step": 5350
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 5360
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5871,
+      "step": 5370
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5958,
+      "step": 5380
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5939,
+      "step": 5390
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5985,
+      "step": 5400
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6004,
+      "step": 5410
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5908,
+      "step": 5420
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6019,
+      "step": 5430
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 5440
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.602,
+      "step": 5450
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6098,
+      "step": 5460
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6045,
+      "step": 5470
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5921,
+      "step": 5480
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5962,
+      "step": 5490
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5969,
+      "step": 5500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5975,
+      "step": 5510
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5998,
+      "step": 5520
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6049,
+      "step": 5530
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5540
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.599,
+      "step": 5550
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 5560
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6019,
+      "step": 5570
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.601,
+      "step": 5580
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.592,
+      "step": 5590
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.5925,
+      "step": 5600
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5952,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 5640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5987,
+      "step": 5650
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6125,
+      "step": 5660
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5889,
+      "step": 5670
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5944,
+      "step": 5680
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5690
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5973,
+      "step": 5700
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6004,
+      "step": 5710
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5989,
+      "step": 5720
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5983,
+      "step": 5730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6079,
+      "step": 5740
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6028,
+      "step": 5750
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5953,
+      "step": 5760
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6108,
+      "step": 5770
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 5780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5943,
+      "step": 5790
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5845,
+      "step": 5800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5952,
+      "step": 5810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 5820
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5987,
+      "step": 5830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5895,
+      "step": 5840
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6007,
+      "step": 5850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 5860
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6054,
+      "step": 5870
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5896,
+      "step": 5880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.598,
+      "step": 5890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5963,
+      "step": 5900
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5910
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5944,
+      "step": 5920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 5930
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6001,
+      "step": 5940
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5975,
+      "step": 5950
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5864,
+      "step": 5960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5899,
+      "step": 5970
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 5980
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5957,
+      "step": 5990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5888,
+      "step": 6000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5994,
+      "step": 6010
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5914,
+      "step": 6020
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5981,
+      "step": 6030
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5807,
+      "step": 6040
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.594,
+      "step": 6050
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5983,
+      "step": 6060
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5895,
+      "step": 6070
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5896,
+      "step": 6080
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5949,
+      "step": 6090
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.579,
+      "step": 6100
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.591,
+      "step": 6110
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5881,
+      "step": 6120
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6055,
+      "step": 6130
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6037,
+      "step": 6140
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5881,
+      "step": 6150
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5937,
+      "step": 6160
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5962,
+      "step": 6170
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5838,
+      "step": 6180
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6035,
+      "step": 6190
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5993,
+      "step": 6200
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6014,
+      "step": 6210
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5968,
+      "step": 6220
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 6230
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6003,
+      "step": 6240
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5939,
+      "step": 6250
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5855,
+      "step": 6260
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5876,
+      "step": 6270
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 6280
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5845,
+      "step": 6290
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5929,
+      "step": 6300
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 6310
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5938,
+      "step": 6320
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5935,
+      "step": 6330
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5904,
+      "step": 6340
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6015,
+      "step": 6350
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.59,
+      "step": 6360
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5894,
+      "step": 6370
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6066,
+      "step": 6380
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5921,
+      "step": 6390
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 6400
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5919,
+      "step": 6410
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 6420
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5948,
+      "step": 6430
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5858,
+      "step": 6440
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5927,
+      "step": 6450
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5877,
+      "step": 6460
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5874,
+      "step": 6470
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5934,
+      "step": 6480
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 6490
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5994,
+      "step": 6500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5986,
+      "step": 6510
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5863,
+      "step": 6520
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5844,
+      "step": 6530
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5891,
+      "step": 6540
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5848,
+      "step": 6550
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6097,
+      "step": 6560
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 6570
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5913,
+      "step": 6580
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5933,
+      "step": 6590
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 6600
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5811,
+      "step": 6610
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5866,
+      "step": 6620
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 6630
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5871,
+      "step": 6640
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5849,
+      "step": 6650
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5923,
+      "step": 6660
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.59,
+      "step": 6670
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5842,
+      "step": 6680
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5902,
+      "step": 6690
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5971,
+      "step": 6700
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 6710
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5885,
+      "step": 6720
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5924,
+      "step": 6730
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5837,
+      "step": 6740
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.588,
+      "step": 6750
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5903,
+      "step": 6760
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5896,
+      "step": 6770
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5885,
+      "step": 6780
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.588,
+      "step": 6790
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 6800
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 6810
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5811,
+      "step": 6820
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5904,
+      "step": 6830
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5817,
+      "step": 6840
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 6850
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5847,
+      "step": 6860
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 6870
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5886,
+      "step": 6880
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5828,
+      "step": 6890
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 6900
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5843,
+      "step": 6910
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5907,
+      "step": 6920
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5802,
+      "step": 6930
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5883,
+      "step": 6940
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5842,
+      "step": 6950
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5908,
+      "step": 6960
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.593,
+      "step": 6970
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5913,
+      "step": 6980
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5846,
+      "step": 6990
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5924,
+      "step": 7000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 7010
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 7020
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.581,
+      "step": 7030
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5893,
+      "step": 7040
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.5792,
+      "step": 7050
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6,
+      "step": 7060
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 7070
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.579,
+      "step": 7080
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5834,
+      "step": 7090
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5823,
+      "step": 7100
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5852,
+      "step": 7110
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 7120
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5831,
+      "step": 7130
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.581,
+      "step": 7140
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5791,
+      "step": 7150
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5779,
+      "step": 7160
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5843,
+      "step": 7170
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5849,
+      "step": 7180
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.588,
+      "step": 7190
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 7200
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5785,
+      "step": 7210
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.59,
+      "step": 7220
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5897,
+      "step": 7230
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5827,
+      "step": 7240
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5874,
+      "step": 7250
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5853,
+      "step": 7260
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5849,
+      "step": 7270
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5926,
+      "step": 7280
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.586,
+      "step": 7290
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5897,
+      "step": 7300
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 7310
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5805,
+      "step": 7320
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5942,
+      "step": 7330
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5872,
+      "step": 7340
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5785,
+      "step": 7350
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5852,
+      "step": 7360
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5964,
+      "step": 7370
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.584,
+      "step": 7380
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5775,
+      "step": 7390
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5909,
+      "step": 7400
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5876,
+      "step": 7410
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.585,
+      "step": 7420
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.585,
+      "step": 7430
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5876,
+      "step": 7440
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5876,
+      "step": 7450
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5846,
+      "step": 7460
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5909,
+      "step": 7470
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5862,
+      "step": 7480
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.575,
+      "step": 7490
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5819,
+      "step": 7500
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5867,
+      "step": 7510
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5963,
+      "step": 7520
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 7530
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5914,
+      "step": 7540
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5847,
+      "step": 7550
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5905,
+      "step": 7560
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5744,
+      "step": 7570
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5771,
+      "step": 7580
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5827,
+      "step": 7590
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5844,
+      "step": 7600
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5807,
+      "step": 7610
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5872,
+      "step": 7620
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5827,
+      "step": 7630
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5827,
+      "step": 7640
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5894,
+      "step": 7650
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5827,
+      "step": 7660
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5815,
+      "step": 7670
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5828,
+      "step": 7680
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5835,
+      "step": 7690
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5798,
+      "step": 7700
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5886,
+      "step": 7710
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5966,
+      "step": 7720
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5878,
+      "step": 7730
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5855,
+      "step": 7740
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5816,
+      "step": 7750
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 7760
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5893,
+      "step": 7770
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5839,
+      "step": 7780
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5791,
+      "step": 7790
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5974,
+      "step": 7800
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5769,
+      "step": 7810
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5816,
+      "step": 7820
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5831,
+      "step": 7830
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5885,
+      "step": 7840
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5866,
+      "step": 7850
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5805,
+      "step": 7860
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5888,
+      "step": 7870
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5713,
+      "step": 7880
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5797,
+      "step": 7890
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5757,
+      "step": 7900
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 7910
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5921,
+      "step": 7920
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5825,
+      "step": 7930
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5749,
+      "step": 7940
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5858,
+      "step": 7950
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5773,
+      "step": 7960
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5913,
+      "step": 7970
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5758,
+      "step": 7980
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 7990
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5885,
+      "step": 8000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/training_args.bin b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..76c6e8e44c071d2b5c47ee143f33b3b0c3f9fe92
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/checkpoint-8000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee26f62127373b4e313bacaf1f87bc66c9583ad97e93c127ff53a983dbadb481
+size 4856
diff --git a/cvnet_small_v1_noglobal_noshuf/model.safetensors b/cvnet_small_v1_noglobal_noshuf/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..76ca21d3991b619e97fc0284f2dbf1af7d0b7598
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bdfe92fcbb79a012510882d98b5bda0b2937bc0cc8648dfa5ab2e2302abb3fd
+size 231565528
diff --git a/cvnet_small_v1_noglobal_noshuf/trainer_state.json b/cvnet_small_v1_noglobal_noshuf/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..07748219517da4290718f41ef4e4c261738846fa
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/trainer_state.json
@@ -0,0 +1,5364 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9997469849034326,
+  "eval_steps": 500,
+  "global_step": 8892,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.7047,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6771,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.676,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6731,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6726,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6734,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6718,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6683,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6719,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6684,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6675,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6806,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6743,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6735,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6739,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6761,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6714,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6707,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6651,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6757,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6759,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6698,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6687,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6697,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6678,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6604,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6669,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6655,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6666,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6647,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6612,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6616,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6607,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6596,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6606,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.654,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6631,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6558,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6728,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6649,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6569,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6582,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6484,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6476,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6628,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6523,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.655,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6496,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6498,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6542,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6445,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6511,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6482,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6623,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6473,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.651,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6453,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6508,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6541,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6467,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6405,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6533,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6416,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6461,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6439,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6507,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6481,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6442,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6418,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6505,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.637,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6737,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.64,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6583,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6326,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6384,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6427,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6379,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6352,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6413,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6374,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.641,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6332,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6351,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6365,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6386,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6433,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6304,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6361,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6303,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6299,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6376,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6447,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6325,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6343,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6389,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6302,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6315,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6293,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6466,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6371,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6323,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6239,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.638,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6391,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6494,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6448,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6454,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6521,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6653,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6432,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6347,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6373,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.635,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6322,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6297,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6532,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6428,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6344,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6273,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6372,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6398,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6491,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6318,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6339,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6355,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6381,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6327,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6354,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6383,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.63,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6313,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6393,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.644,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6387,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6388,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6337,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.6284,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6292,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6312,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.634,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6267,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6227,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.6319,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6425,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6301,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6316,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6277,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6399,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6324,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6268,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6306,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6203,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6305,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6385,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6288,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6283,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6349,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6223,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6382,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6345,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6462,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6367,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6358,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.6338,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6255,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6199,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.6252,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6258,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6328,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6359,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6329,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.624,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6463,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6446,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6264,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6209,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6141,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6249,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6259,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6237,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.622,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6206,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.6366,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6341,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6414,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6218,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6194,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6278,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6248,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6213,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6362,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6377,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6333,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6526,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6221,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6226,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6246,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6311,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6182,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6145,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6314,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6357,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.6295,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6205,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6242,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.6233,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6235,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6207,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.6224,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.612,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6335,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6117,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6171,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.6112,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6238,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6266,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6204,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6241,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.625,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6217,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6234,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.6185,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6189,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6143,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6243,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.6156,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6131,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.621,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6202,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6229,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6081,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6082,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6251,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.639,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6265,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6247,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.6201,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6212,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6197,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.6094,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6228,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6274,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.627,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6088,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6144,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6195,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6155,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6111,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.6153,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6164,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6109,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.6216,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6049,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6193,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6157,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.6135,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6178,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.6037,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6222,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6119,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.6129,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6162,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.6253,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6176,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6086,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6214,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.6138,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6121,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6123,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.6159,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6187,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.6071,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6095,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6188,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.623,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6115,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.6085,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6012,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.6042,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.607,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6104,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.6091,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6116,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6108,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.6175,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6114,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.6086,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.6152,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6043,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.617,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.6151,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6146,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.6107,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.613,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5997,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.6056,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6211,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.6124,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6097,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.6177,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.6126,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.603,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6219,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.6016,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6083,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.6128,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6101,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5993,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6173,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.6072,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.628,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6024,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.6063,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5992,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.6025,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6022,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.6103,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6065,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.6061,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6208,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6032,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.6084,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.61,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.605,
+      "step": 5000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5961,
+      "step": 5010
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5957,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6073,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6069,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5924,
+      "step": 5050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.6059,
+      "step": 5060
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5999,
+      "step": 5070
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5981,
+      "step": 5080
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5953,
+      "step": 5090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5974,
+      "step": 5100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5941,
+      "step": 5110
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.6051,
+      "step": 5120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5919,
+      "step": 5130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.618,
+      "step": 5140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5972,
+      "step": 5150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 5160
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.6066,
+      "step": 5170
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6047,
+      "step": 5180
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6096,
+      "step": 5190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.6077,
+      "step": 5200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5961,
+      "step": 5210
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.593,
+      "step": 5220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.6012,
+      "step": 5230
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 5240
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5976,
+      "step": 5250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.6027,
+      "step": 5260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5945,
+      "step": 5270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6092,
+      "step": 5280
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.6062,
+      "step": 5290
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6021,
+      "step": 5300
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5988,
+      "step": 5310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.6035,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6163,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.6015,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5948,
+      "step": 5350
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 5360
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5871,
+      "step": 5370
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5958,
+      "step": 5380
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5939,
+      "step": 5390
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5985,
+      "step": 5400
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6004,
+      "step": 5410
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5908,
+      "step": 5420
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.6019,
+      "step": 5430
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6048,
+      "step": 5440
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.602,
+      "step": 5450
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.6098,
+      "step": 5460
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.6045,
+      "step": 5470
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5921,
+      "step": 5480
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5962,
+      "step": 5490
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5969,
+      "step": 5500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5975,
+      "step": 5510
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5998,
+      "step": 5520
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.6049,
+      "step": 5530
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5540
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.599,
+      "step": 5550
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6075,
+      "step": 5560
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.6019,
+      "step": 5570
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.601,
+      "step": 5580
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.592,
+      "step": 5590
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.5925,
+      "step": 5600
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.6064,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.6089,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5952,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 5640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5987,
+      "step": 5650
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.6125,
+      "step": 5660
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5889,
+      "step": 5670
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5944,
+      "step": 5680
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5690
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5973,
+      "step": 5700
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.6004,
+      "step": 5710
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5989,
+      "step": 5720
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5983,
+      "step": 5730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6079,
+      "step": 5740
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.6028,
+      "step": 5750
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5953,
+      "step": 5760
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6108,
+      "step": 5770
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 5780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5943,
+      "step": 5790
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5845,
+      "step": 5800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5952,
+      "step": 5810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 5820
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5987,
+      "step": 5830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5895,
+      "step": 5840
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.6007,
+      "step": 5850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 5860
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.6054,
+      "step": 5870
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5896,
+      "step": 5880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.598,
+      "step": 5890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5963,
+      "step": 5900
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5984,
+      "step": 5910
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5944,
+      "step": 5920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.606,
+      "step": 5930
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.6001,
+      "step": 5940
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5975,
+      "step": 5950
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5864,
+      "step": 5960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5899,
+      "step": 5970
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.604,
+      "step": 5980
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5957,
+      "step": 5990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5888,
+      "step": 6000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5994,
+      "step": 6010
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5914,
+      "step": 6020
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5981,
+      "step": 6030
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5807,
+      "step": 6040
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.594,
+      "step": 6050
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5983,
+      "step": 6060
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5895,
+      "step": 6070
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5896,
+      "step": 6080
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5949,
+      "step": 6090
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.579,
+      "step": 6100
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.591,
+      "step": 6110
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5881,
+      "step": 6120
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6055,
+      "step": 6130
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.6037,
+      "step": 6140
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5881,
+      "step": 6150
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5937,
+      "step": 6160
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5962,
+      "step": 6170
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5838,
+      "step": 6180
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6035,
+      "step": 6190
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5993,
+      "step": 6200
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.6014,
+      "step": 6210
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5968,
+      "step": 6220
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 6230
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.6003,
+      "step": 6240
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5939,
+      "step": 6250
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.5855,
+      "step": 6260
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5876,
+      "step": 6270
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 6280
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.5845,
+      "step": 6290
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5929,
+      "step": 6300
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 6310
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.5938,
+      "step": 6320
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5935,
+      "step": 6330
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5904,
+      "step": 6340
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.6015,
+      "step": 6350
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.59,
+      "step": 6360
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5894,
+      "step": 6370
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.6066,
+      "step": 6380
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5921,
+      "step": 6390
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5967,
+      "step": 6400
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5919,
+      "step": 6410
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 6420
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5948,
+      "step": 6430
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5858,
+      "step": 6440
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5927,
+      "step": 6450
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5877,
+      "step": 6460
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.5874,
+      "step": 6470
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5934,
+      "step": 6480
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6038,
+      "step": 6490
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5994,
+      "step": 6500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5986,
+      "step": 6510
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5863,
+      "step": 6520
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.5844,
+      "step": 6530
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5891,
+      "step": 6540
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.5848,
+      "step": 6550
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.6097,
+      "step": 6560
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.6017,
+      "step": 6570
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5913,
+      "step": 6580
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.5933,
+      "step": 6590
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 6600
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5811,
+      "step": 6610
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.5866,
+      "step": 6620
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5979,
+      "step": 6630
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5871,
+      "step": 6640
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.5849,
+      "step": 6650
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5923,
+      "step": 6660
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.59,
+      "step": 6670
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.5842,
+      "step": 6680
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5902,
+      "step": 6690
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5971,
+      "step": 6700
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 6710
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5885,
+      "step": 6720
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5924,
+      "step": 6730
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.5837,
+      "step": 6740
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.588,
+      "step": 6750
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5903,
+      "step": 6760
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.5896,
+      "step": 6770
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5885,
+      "step": 6780
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.588,
+      "step": 6790
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 6800
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 6810
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5811,
+      "step": 6820
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.5904,
+      "step": 6830
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5817,
+      "step": 6840
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 6850
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.5847,
+      "step": 6860
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 6870
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5886,
+      "step": 6880
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.5828,
+      "step": 6890
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5906,
+      "step": 6900
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5843,
+      "step": 6910
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.5907,
+      "step": 6920
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5802,
+      "step": 6930
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5883,
+      "step": 6940
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.5842,
+      "step": 6950
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5908,
+      "step": 6960
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.593,
+      "step": 6970
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.5913,
+      "step": 6980
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5846,
+      "step": 6990
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5924,
+      "step": 7000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.5959,
+      "step": 7010
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 7020
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.581,
+      "step": 7030
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.5893,
+      "step": 7040
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.5792,
+      "step": 7050
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.6,
+      "step": 7060
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 7070
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.579,
+      "step": 7080
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.5834,
+      "step": 7090
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5823,
+      "step": 7100
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5852,
+      "step": 7110
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 7120
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5831,
+      "step": 7130
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.581,
+      "step": 7140
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.5791,
+      "step": 7150
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5779,
+      "step": 7160
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5843,
+      "step": 7170
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.5849,
+      "step": 7180
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.588,
+      "step": 7190
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 7200
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.5785,
+      "step": 7210
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.59,
+      "step": 7220
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5897,
+      "step": 7230
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.5827,
+      "step": 7240
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5874,
+      "step": 7250
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5853,
+      "step": 7260
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.5849,
+      "step": 7270
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5926,
+      "step": 7280
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.586,
+      "step": 7290
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5897,
+      "step": 7300
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 7310
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5805,
+      "step": 7320
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5942,
+      "step": 7330
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5872,
+      "step": 7340
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5785,
+      "step": 7350
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.5852,
+      "step": 7360
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5964,
+      "step": 7370
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.584,
+      "step": 7380
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.5775,
+      "step": 7390
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5909,
+      "step": 7400
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5876,
+      "step": 7410
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.585,
+      "step": 7420
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.585,
+      "step": 7430
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5876,
+      "step": 7440
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.5876,
+      "step": 7450
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5846,
+      "step": 7460
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5909,
+      "step": 7470
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.5862,
+      "step": 7480
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.575,
+      "step": 7490
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5819,
+      "step": 7500
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.5867,
+      "step": 7510
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5963,
+      "step": 7520
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5869,
+      "step": 7530
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.5914,
+      "step": 7540
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5847,
+      "step": 7550
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5905,
+      "step": 7560
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.5744,
+      "step": 7570
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5771,
+      "step": 7580
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5827,
+      "step": 7590
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.5844,
+      "step": 7600
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5807,
+      "step": 7610
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5872,
+      "step": 7620
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.5827,
+      "step": 7630
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5827,
+      "step": 7640
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5894,
+      "step": 7650
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5827,
+      "step": 7660
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5815,
+      "step": 7670
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5828,
+      "step": 7680
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.5835,
+      "step": 7690
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5798,
+      "step": 7700
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5886,
+      "step": 7710
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.5966,
+      "step": 7720
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5878,
+      "step": 7730
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5855,
+      "step": 7740
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.5816,
+      "step": 7750
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5856,
+      "step": 7760
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5893,
+      "step": 7770
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.5839,
+      "step": 7780
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5791,
+      "step": 7790
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5974,
+      "step": 7800
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.5769,
+      "step": 7810
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5816,
+      "step": 7820
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5831,
+      "step": 7830
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.5885,
+      "step": 7840
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5866,
+      "step": 7850
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5805,
+      "step": 7860
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.5888,
+      "step": 7870
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5713,
+      "step": 7880
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.5797,
+      "step": 7890
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5757,
+      "step": 7900
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5954,
+      "step": 7910
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.5921,
+      "step": 7920
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5825,
+      "step": 7930
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5749,
+      "step": 7940
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.5858,
+      "step": 7950
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5773,
+      "step": 7960
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5913,
+      "step": 7970
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.5758,
+      "step": 7980
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5898,
+      "step": 7990
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5885,
+      "step": 8000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.5803,
+      "step": 8010
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.576,
+      "step": 8020
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5821,
+      "step": 8030
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.5784,
+      "step": 8040
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5787,
+      "step": 8050
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5862,
+      "step": 8060
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.5778,
+      "step": 8070
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5894,
+      "step": 8080
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5868,
+      "step": 8090
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.578,
+      "step": 8100
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5721,
+      "step": 8110
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5771,
+      "step": 8120
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.5795,
+      "step": 8130
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5847,
+      "step": 8140
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5801,
+      "step": 8150
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.5759,
+      "step": 8160
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.581,
+      "step": 8170
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5797,
+      "step": 8180
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.5798,
+      "step": 8190
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5811,
+      "step": 8200
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.5724,
+      "step": 8210
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.577,
+      "step": 8220
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5823,
+      "step": 8230
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5812,
+      "step": 8240
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.5763,
+      "step": 8250
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5789,
+      "step": 8260
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.578,
+      "step": 8270
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.5826,
+      "step": 8280
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.587,
+      "step": 8290
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5859,
+      "step": 8300
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.5904,
+      "step": 8310
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5706,
+      "step": 8320
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5772,
+      "step": 8330
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.5859,
+      "step": 8340
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5824,
+      "step": 8350
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5798,
+      "step": 8360
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.5788,
+      "step": 8370
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5742,
+      "step": 8380
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5829,
+      "step": 8390
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.5823,
+      "step": 8400
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5741,
+      "step": 8410
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5789,
+      "step": 8420
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.5873,
+      "step": 8430
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5781,
+      "step": 8440
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5789,
+      "step": 8450
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.5738,
+      "step": 8460
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.5822,
+      "step": 8470
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.593,
+      "step": 8480
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.58,
+      "step": 8490
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5768,
+      "step": 8500
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5818,
+      "step": 8510
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5802,
+      "step": 8520
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.5753,
+      "step": 8530
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.5794,
+      "step": 8540
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.574,
+      "step": 8550
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.5712,
+      "step": 8560
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.5756,
+      "step": 8570
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.5687,
+      "step": 8580
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5673,
+      "step": 8590
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5771,
+      "step": 8600
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.5781,
+      "step": 8610
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.573,
+      "step": 8620
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5821,
+      "step": 8630
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.5797,
+      "step": 8640
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5812,
+      "step": 8650
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5777,
+      "step": 8660
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5755,
+      "step": 8670
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5988,
+      "step": 8680
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.5892,
+      "step": 8690
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.576,
+      "step": 8700
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5888,
+      "step": 8710
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 0.5845,
+      "step": 8720
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5799,
+      "step": 8730
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5784,
+      "step": 8740
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.5774,
+      "step": 8750
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5776,
+      "step": 8760
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5793,
+      "step": 8770
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.5814,
+      "step": 8780
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5645,
+      "step": 8790
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5851,
+      "step": 8800
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 8810
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5776,
+      "step": 8820
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.5755,
+      "step": 8830
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.581,
+      "step": 8840
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.577,
+      "step": 8850
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5703,
+      "step": 8860
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.5934,
+      "step": 8870
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5754,
+      "step": 8880
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5745,
+      "step": 8890
+    },
+    {
+      "epoch": 3.0,
+      "step": 8892,
+      "total_flos": 0.0,
+      "train_loss": 0.6122167715069629,
+      "train_runtime": 62943.9838,
+      "train_samples_per_second": 36.166,
+      "train_steps_per_second": 0.141
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/cvnet_small_v1_noglobal_noshuf/training_args.bin b/cvnet_small_v1_noglobal_noshuf/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..76c6e8e44c071d2b5c47ee143f33b3b0c3f9fe92
--- /dev/null
+++ b/cvnet_small_v1_noglobal_noshuf/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee26f62127373b4e313bacaf1f87bc66c9583ad97e93c127ff53a983dbadb481
+size 4856
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/model.safetensors b/dinov2_small_v1_global_nominmax/checkpoint-1000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..c001701709c6078cacadc4a004e88ad9db527f71
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:715056d99912e2e190ad33a1daa3583a0f5d55c0dac206b44d830e802fb8a79e
+size 228416552
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/optimizer.pt b/dinov2_small_v1_global_nominmax/checkpoint-1000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..94a01a45a7883c5540092ada819f2d409fb0bc2d
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:124ae6e79ebf74bb762aac0c09532b19612907250e4e1b519426e39c737eaa04
+size 456914234
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_0.pth b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..4ab31eb89973990d9b8610104399e7e5aaec5986
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f68045b77ef059786b9d730ec7bfcab3532c91c560ed23338e0d4d518b2be5e
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_1.pth b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c53f9ece2cbb84f4662850cb2a534a064415698c
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:164a39302aced8efcfa4d5b883ab7d4f208ce2d3aaf57fa0fff0dc5bc35e9b53
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_2.pth b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5513d1ce6c2aed658bc3b07da4002acb5b6ce7dc
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74928fb7ef7944af9aedd43d3456cb7cc320ccd38e5d7989ea54cf81f2bdc4ac
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_3.pth b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..873724ab5b9b6f006d53045897c53d31e33804d8
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7390fd7db05b7bd4f309fe3c16e3133e51c0b7bbf36881c88fd513b9ebb7b75a
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_4.pth b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..81ff0496e02da863836522e59c26181e5043ad3f
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:315c200f1cd4bb6a4a593accb8ba546796343253845919cb5b2b52089686830d
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_5.pth b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..dde498f5924607b3543e00cedc5f8d8e7ead0775
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:561c368392fbdb6e8a7e5e086de59545625d80b821b5c72a4024b1f47b1390bb
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_6.pth b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..004d54378694fe0b6ec4833301aba9fe98fda662
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56a358b4a84de8c0fefda163466d1b792e3d23a409d2f7f54fdd9b3535793d39
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_7.pth b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e54c9d0e00105e7c7921fd6d4d170f670ee0b5a7
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17ad99f8730fc68cc1e3438467270d0c73427583fe0bffeb5bca595169156c2f
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/scheduler.pt b/dinov2_small_v1_global_nominmax/checkpoint-1000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9ff2c05006e51f3f771730c9055c6c181da2abd0
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c34ba9e6f50c2313df84e91c7dfed2be6b77dec7394adba34d3bd12d40d5ba2d
+size 1064
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/trainer_state.json b/dinov2_small_v1_global_nominmax/checkpoint-1000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..cbe217494c7cf0d89b1ffc4ab4749ebe3d540935
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/trainer_state.json
@@ -0,0 +1,621 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3373534620899047,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5854,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5466,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.55,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5355,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5183,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5221,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5063,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5058,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5059,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5077,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5089,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5075,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5017,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5022,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5046,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5174,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4867,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4998,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4982,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4931,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4859,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5014,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5076,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4979,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4999,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4908,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.478,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4838,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4879,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4761,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4853,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.477,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.491,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4678,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4814,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4822,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4792,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4791,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4839,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4821,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4783,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4771,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4747,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4692,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4684,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4775,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4714,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4744,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4752,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4727,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4693,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4699,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4694,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4745,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4754,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4796,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4708,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.466,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4706,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4756,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4687,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-1000/training_args.bin b/dinov2_small_v1_global_nominmax/checkpoint-1000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ca4990ad6e2d6dc5c39192293644d1530e9bbcd5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-1000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:898baa792356a9e4502a27fce84c053686ecb4d412bc621b4bb97d064e9b460c
+size 4856
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/model.safetensors b/dinov2_small_v1_global_nominmax/checkpoint-2000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..bae9b8c1e7bc411250451fe91ac78bf373622eaf
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffe63ae8223f8f25d3c8d974bda9bfe7efe0437b404ca376b82fad1765a6027c
+size 228416552
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/optimizer.pt b/dinov2_small_v1_global_nominmax/checkpoint-2000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e5f7e1af3406454cdacaf5b54d5b500dc2988be1
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28de628253cf7185c40364c91e74d4d8977fe1f386e6ce82b267c4edd48d0d71
+size 456914234
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_0.pth b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..33d35d46eb49c7694ddf02abc1f806fe21ce3c03
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24545b6bb7493bdf35c20f30780f8e0793c67065421d64d7180f5a320abd8417
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_1.pth b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..6310340e83ce52050038f3421730ede96037a6ba
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0232b218d0ca1f5487f91d43e6fb49bb92e0051351056294e6e73f362f38dc4
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_2.pth b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5f2d7f68fb7fd5fa3b96e6ab8e2a6ed8ae0cc685
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3dffc06c0c6d363c38ffe8a64d78f552b7bfccf1d412a5e6d8b05a20badc3ab4
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_3.pth b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ee633cf4e658e3f4b63beab3a61ea2a072f17879
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c47dc8892b7412ec8da71c7704256b96b8469c5c7f3b6490b83502e4b2ee928
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_4.pth b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..a68c4918b64151b1245b3a97d067f09d0feb75a0
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:905b45076269c4efc0bfdb46513ea882d1de5bce7ddd5bed4ce98092c7da2c0b
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_5.pth b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..37b3a6d3dd2e51bee05f2894dcebb9a4fe902f47
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97ee8e3283737d5d9165ab7bbed91fab8f8054e727830b0ac207e19ea1b998ca
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_6.pth b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..24f763bcd01fe0ca32d12a74bd2f9a71337dd568
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7205ffddd0d11808dcecbed7a8e9b2051147bffa0bcf02b1ee5ef28b9e65162
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_7.pth b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1a55221cfbb5c000b877e8e6b14977b7163966dc
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07bde508a1939450d8ec45b79bab2568f6dd5d0ac0cf2add96faab85481697da
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/scheduler.pt b/dinov2_small_v1_global_nominmax/checkpoint-2000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4605de17b07049a57c10990ac0db26645b375276
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:115fc5d3c71fc8b112d5c1701359950aa4e48ea559451a73de1d1d8ca1116d89
+size 1064
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/trainer_state.json b/dinov2_small_v1_global_nominmax/checkpoint-2000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..2d454e2938361e4c5bc3e3ef1064eef03e069b6e
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/trainer_state.json
@@ -0,0 +1,1221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6747069241798094,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5854,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5466,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.55,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5355,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5183,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5221,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5063,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5058,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5059,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5077,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5089,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5075,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5017,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5022,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5046,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5174,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4867,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4998,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4982,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4931,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4859,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5014,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5076,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4979,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4999,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4908,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.478,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4838,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4879,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4761,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4853,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.477,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.491,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4678,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4814,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4822,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4792,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4791,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4839,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4821,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4783,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4771,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4747,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4692,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4684,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4775,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4714,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4744,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4752,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4727,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4693,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4699,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4694,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4745,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4754,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4796,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4708,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.466,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4706,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4756,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4687,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4702,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4739,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.472,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4632,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.458,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.463,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4655,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4682,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4627,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4608,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4718,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4612,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4615,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4621,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4575,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4602,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4672,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4613,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4685,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4643,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4637,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4595,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4593,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4585,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4538,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4559,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4623,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4605,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4569,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4649,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4572,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4578,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4617,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4646,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.457,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4564,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.455,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4591,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4543,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4536,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4547,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.451,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4534,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4537,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4485,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4518,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4601,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-2000/training_args.bin b/dinov2_small_v1_global_nominmax/checkpoint-2000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ca4990ad6e2d6dc5c39192293644d1530e9bbcd5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-2000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:898baa792356a9e4502a27fce84c053686ecb4d412bc621b4bb97d064e9b460c
+size 4856
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/model.safetensors b/dinov2_small_v1_global_nominmax/checkpoint-3000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..a5f4f5bf0e9203ec5083aacebbc30981019305b9
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c266272c4e71a36614fdd3904297971925f5e1c08abf7ad1d449815c569e8e7
+size 228416552
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/optimizer.pt b/dinov2_small_v1_global_nominmax/checkpoint-3000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d6a7366f12e739816ab11b63f9787793c6522737
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d131f48c4ec990cfb534a7c6235ebf4ef0fd190eb8238e084de7f830d48bbce5
+size 456914234
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_0.pth b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cb8e78a2252e7489e47adb734bb1f370f80fb558
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7431609fc9d10c784464e14fe45f5a89b22573e0769c9c21d7d0a2dead49ec58
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_1.pth b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..14409c37cb8c37cf95a2a40f01e3bde6c0a86a8f
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5081bec13a1a7ac2ad3917351c18540a7bae0cfef6e0df05d537d5c0163fce12
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_2.pth b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..2e0781a6af83c0376c3bc340f290368a3cdf5f3b
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:535456fe6d77127c3883ae9c325dccec963de01449e321e9c704f78c50992939
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_3.pth b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c3a155aae9463722acad62167de281b7e03ad81a
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:880e7012bf446334aa49027d1791b52f192a6dc11d44fea74095b153fd34bc97
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_4.pth b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..50505327adf88ab24ab9f90d8faf76553bc4d2a9
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40038cca0e1cc4c41ee9eb36cfd70f6e868697faa2d2d79afba994d00387b725
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_5.pth b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1db65e964f07df377b31f7937f3fad02cb5361f7
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86ec016c5b320b9210f7edc5e4b4d309aa8c19437ccbdf2fd68012814e7b0be6
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_6.pth b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9d27f215baa695e5cea378d01affc7631aa9463d
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:049da174472b1bfd728972b360f9b2b80d3b4842c4c00b3019ff285417d7b3ca
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_7.pth b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ae156e93c6ee694319aecdfa93bbb81b80472851
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2dcd919cf3c5f19cfa3f8a17ee1d2a7803a15af43f3917c55ec05ed6ca5858d
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/scheduler.pt b/dinov2_small_v1_global_nominmax/checkpoint-3000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0a885c8b891f90d887f6c40db2b8ae74f564dd98
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4df43d46412429a4bda78060914e68357cbee658ac819784da79443fe1ce6982
+size 1064
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/trainer_state.json b/dinov2_small_v1_global_nominmax/checkpoint-3000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..43d11899a95fdd698986089a18286f3486cd49b1
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/trainer_state.json
@@ -0,0 +1,1821 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0120603862697142,
+  "eval_steps": 500,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5854,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5466,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.55,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5355,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5183,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5221,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5063,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5058,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5059,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5077,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5089,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5075,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5017,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5022,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5046,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5174,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4867,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4998,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4982,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4931,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4859,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5014,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5076,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4979,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4999,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4908,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.478,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4838,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4879,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4761,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4853,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.477,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.491,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4678,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4814,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4822,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4792,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4791,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4839,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4821,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4783,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4771,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4747,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4692,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4684,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4775,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4714,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4744,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4752,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4727,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4693,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4699,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4694,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4745,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4754,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4796,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4708,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.466,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4706,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4756,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4687,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4702,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4739,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.472,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4632,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.458,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.463,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4655,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4682,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4627,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4608,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4718,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4612,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4615,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4621,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4575,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4602,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4672,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4613,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4685,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4643,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4637,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4595,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4593,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4585,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4538,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4559,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4623,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4605,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4569,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4649,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4572,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4578,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4617,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4646,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.457,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4564,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.455,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4591,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4543,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4536,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4547,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.451,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4534,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4537,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4485,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4518,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4601,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4609,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4606,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.456,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.4526,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4517,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4546,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4565,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4498,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4529,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4513,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4482,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4555,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4474,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4563,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4488,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4468,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4515,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4502,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4472,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4428,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.443,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4549,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4554,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4481,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4492,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4505,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.438,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4464,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4504,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4461,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4432,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4458,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4507,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4506,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4393,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4483,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4368,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4442,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4427,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4489,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4371,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4452,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4407,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4473,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4337,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4409,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4332,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4416,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-3000/training_args.bin b/dinov2_small_v1_global_nominmax/checkpoint-3000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ca4990ad6e2d6dc5c39192293644d1530e9bbcd5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-3000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:898baa792356a9e4502a27fce84c053686ecb4d412bc621b4bb97d064e9b460c
+size 4856
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/model.safetensors b/dinov2_small_v1_global_nominmax/checkpoint-4000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..27c66488b94c61c4700d3db227c7d26ed1da4038
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1e7ed581ced44beb63e53b4f97b2c055d9fec9189ab1eaa63eeb5fe161a132c
+size 228416552
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/optimizer.pt b/dinov2_small_v1_global_nominmax/checkpoint-4000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0c62e85e9c8dac2f9c75a70da314b1b8fe801278
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ae9f1d52c25734c652ac506fefebfd259b1370b7de9d99a7ca8b21931d2cf17
+size 456914234
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_0.pth b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..27157ed6df83ce9d6c109c9b7dea4f945557c6c2
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30e9d80f2c7b48e823660edbe971433389c5987502ff5b8904c6fa06a4b8fac4
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_1.pth b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d07f16bf0d35084ca8183acbe7d1ec6d1d90a220
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6dc64ef283107cc20517b99d8a4260d5dc35fe0739df81e0627d71f753f071e
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_2.pth b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..a5a4f71f56747c23a569fdab1e672a231cf7beb8
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3984ffaba3ef211236ed7149a8baca5ccfdc9f355d0c4a2f94ee586ef97610e1
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_3.pth b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..8bc5aa77c0a5136f2602b61102d9fa19e74049d1
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aef672fc70b0c0d7f41d3aa0c8a5fc28c8b3a305a6b083477713cb302d8d72a0
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_4.pth b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..23ce76a0046bef79ca5a99524908ab2f3b052e03
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a0341c1736e01084f618482fe049b36f4ff83d3c592a9632213c226953a9084
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_5.pth b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..78895c9a779442c5c0a14d8426eb7659ed5f69f5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89b1c55b3a76909bc8b26c01f698db08cb90488d67ec9d13d047aea4788fd653
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_6.pth b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1009d8998b054def1787038b788e28e743447e38
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73db3b651070332fced69b13785993fdefdaac849cbcb23e06c08cde7fdc5d55
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_7.pth b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..642ce42df3671964f6b34b6543afd6eb286a0714
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f21a5acb7274574d65d51c6e1a0d8b70345cf7b9c9dd99e4b85e293546a6e40
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/scheduler.pt b/dinov2_small_v1_global_nominmax/checkpoint-4000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..503d2b402ef8df237c8826d9429a15068a9f8e5d
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1f5b5ac4a53f35c9923ba2e114d395c8f0ca2ff8283a0adc3fb6db30ab7f070
+size 1064
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/trainer_state.json b/dinov2_small_v1_global_nominmax/checkpoint-4000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..f8e5123fd23c9c7fdda76be5cd9e29596b98bdda
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/trainer_state.json
@@ -0,0 +1,2421 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.3494138483596188,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5854,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5466,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.55,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5355,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5183,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5221,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5063,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5058,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5059,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5077,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5089,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5075,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5017,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5022,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5046,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5174,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4867,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4998,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4982,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4931,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4859,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5014,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5076,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4979,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4999,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4908,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.478,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4838,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4879,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4761,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4853,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.477,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.491,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4678,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4814,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4822,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4792,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4791,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4839,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4821,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4783,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4771,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4747,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4692,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4684,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4775,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4714,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4744,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4752,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4727,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4693,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4699,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4694,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4745,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4754,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4796,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4708,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.466,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4706,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4756,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4687,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4702,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4739,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.472,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4632,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.458,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.463,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4655,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4682,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4627,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4608,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4718,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4612,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4615,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4621,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4575,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4602,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4672,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4613,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4685,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4643,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4637,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4595,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4593,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4585,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4538,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4559,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4623,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4605,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4569,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4649,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4572,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4578,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4617,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4646,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.457,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4564,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.455,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4591,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4543,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4536,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4547,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.451,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4534,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4537,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4485,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4518,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4601,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4609,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4606,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.456,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.4526,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4517,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4546,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4565,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4498,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4529,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4513,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4482,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4555,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4474,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4563,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4488,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4468,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4515,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4502,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4472,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4428,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.443,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4549,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4554,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4481,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4492,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4505,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.438,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4464,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4504,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4461,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4432,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4458,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4507,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4506,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4393,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4483,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4368,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4442,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4427,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4489,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4371,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4452,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4407,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4473,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4337,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4409,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4332,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4416,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.445,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4453,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4402,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4457,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4385,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4414,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4396,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4355,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4382,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4408,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4398,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4343,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4352,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4345,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4342,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4308,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4361,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4388,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4424,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.437,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4449,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4366,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4324,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4318,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4374,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4359,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4351,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4397,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4354,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4291,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4304,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4277,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4288,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4253,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4295,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4298,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4278,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4306,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4347,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.425,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4357,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4322,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4307,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4275,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4367,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4258,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4325,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4299,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4268,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4314,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.419,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4254,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-4000/training_args.bin b/dinov2_small_v1_global_nominmax/checkpoint-4000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ca4990ad6e2d6dc5c39192293644d1530e9bbcd5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-4000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:898baa792356a9e4502a27fce84c053686ecb4d412bc621b4bb97d064e9b460c
+size 4856
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/model.safetensors b/dinov2_small_v1_global_nominmax/checkpoint-5000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..8e4d431b56f098473ba1d9956afabe99f002f100
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8510009479b79e6b36a6db87f3d23d450b37bb861528b08dc9eaf81ec54a016
+size 228416552
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/optimizer.pt b/dinov2_small_v1_global_nominmax/checkpoint-5000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..58b2aaebfc1cd0dd851137b732872ec951ad1bd6
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7b4b078e570f91d4e0b87e60ace89d2ba1af907976914ad1fb2d9a66327dd64
+size 456914234
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_0.pth b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..11e0c340c3c865d1230d06fc8147457636c36c7e
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fdb01d742970be957c0ece5354beb317869a9c3ca2df86c171b56533ec8f43a
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_1.pth b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9dd478a8c9ea4164341bf243db8034288705a41e
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f451b776653c44b20c2f0aa86ce758d2f6b160a2aed1623e73fb9ce65aab45dc
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_2.pth b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..64b4a78013a2d2da348987b2be5e0f0f9768f97d
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ec45f9a6fa09681ce69490ec8693348a5cd79d4e73833ec2c3c3616ad0cfb45
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_3.pth b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1c64fb5462b0001c919706fb5b95cbb3049c2daf
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88c41930d7b79073d68661afecfe85fb979ce0c66aa50931c41c30c84094e01f
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_4.pth b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..fe1ec346d9089256afe9858e9a678cc5a0f9e725
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f195114778c83dcdf347276c88ba35590abcc726622eff124827e1184d650b85
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_5.pth b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..551227058e212b8b1b4c9946a46abd488003f429
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e790930ffa2fcd8eaf8532e85b32332d84d65708e9d2dbf5498070faacf67367
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_6.pth b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..dff0e3a4547f927a68daf8bdea3c57bf24a69713
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dcc050ca3e609d48f5075b0221433e64883cc24379379e0f87b9bfc6fc692cd
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_7.pth b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c4f9ceecc118db03d3225ab3c794d989fceb08fc
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ee5311f27f1199ba1ac2117100e5fb29127ebdb52900e88cf9b6a1d8cd6cd3a
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/scheduler.pt b/dinov2_small_v1_global_nominmax/checkpoint-5000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1cb6e935fff7477b8e4bab40ab97b33a5268355e
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5efa72feef3a398c4152d1d39792b540dbc6e38d02d8ab41bc4a583b249405d
+size 1064
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/trainer_state.json b/dinov2_small_v1_global_nominmax/checkpoint-5000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..d055cff65acb79830a92cfc1dbe8ea188c4e7f17
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/trainer_state.json
@@ -0,0 +1,3021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.6867673104495235,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5854,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5466,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.55,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5355,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5183,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5221,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5063,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5058,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5059,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5077,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5089,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5075,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5017,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5022,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5046,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5174,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4867,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4998,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4982,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4931,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4859,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5014,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5076,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4979,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4999,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4908,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.478,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4838,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4879,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4761,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4853,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.477,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.491,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4678,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4814,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4822,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4792,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4791,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4839,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4821,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4783,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4771,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4747,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4692,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4684,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4775,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4714,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4744,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4752,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4727,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4693,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4699,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4694,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4745,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4754,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4796,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4708,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.466,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4706,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4756,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4687,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4702,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4739,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.472,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4632,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.458,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.463,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4655,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4682,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4627,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4608,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4718,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4612,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4615,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4621,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4575,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4602,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4672,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4613,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4685,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4643,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4637,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4595,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4593,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4585,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4538,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4559,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4623,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4605,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4569,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4649,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4572,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4578,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4617,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4646,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.457,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4564,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.455,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4591,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4543,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4536,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4547,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.451,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4534,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4537,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4485,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4518,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4601,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4609,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4606,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.456,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.4526,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4517,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4546,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4565,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4498,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4529,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4513,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4482,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4555,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4474,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4563,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4488,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4468,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4515,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4502,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4472,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4428,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.443,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4549,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4554,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4481,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4492,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4505,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.438,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4464,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4504,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4461,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4432,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4458,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4507,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4506,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4393,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4483,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4368,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4442,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4427,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4489,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4371,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4452,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4407,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4473,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4337,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4409,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4332,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4416,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.445,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4453,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4402,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4457,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4385,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4414,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4396,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4355,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4382,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4408,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4398,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4343,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4352,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4345,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4342,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4308,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4361,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4388,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4424,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.437,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4449,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4366,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4324,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4318,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4374,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4359,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4351,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4397,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4354,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4291,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4304,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4277,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4288,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4253,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4295,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4298,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4278,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4306,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4347,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.425,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4357,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4322,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4307,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4275,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4367,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4258,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4325,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4299,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4268,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4314,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.419,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4254,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4218,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4204,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4219,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4174,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4195,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4217,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4196,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4195,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4189,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4179,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4187,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4217,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4186,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4155,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4111,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4119,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4075,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4121,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4016,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4039,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4036,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4014,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4002,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4041,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.396,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4055,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3981,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3986,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3912,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3863,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3913,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3858,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3886,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3874,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3859,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3846,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3785,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3858,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3882,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3813,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3835,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3775,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3794,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3788,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3724,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3812,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3694,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3775,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3654,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.377,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3701,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3682,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3643,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3737,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3724,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.369,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.3634,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3674,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3569,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3675,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3634,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.362,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3589,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3591,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3549,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3611,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.362,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3554,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3564,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3604,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3574,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.36,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3511,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3529,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3565,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3501,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3475,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.353,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3524,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.35,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3497,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3439,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3552,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3406,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3531,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3459,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3476,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3421,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3451,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.342,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.3357,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.345,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.3437,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3381,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3365,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3412,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3413,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-5000/training_args.bin b/dinov2_small_v1_global_nominmax/checkpoint-5000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ca4990ad6e2d6dc5c39192293644d1530e9bbcd5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-5000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:898baa792356a9e4502a27fce84c053686ecb4d412bc621b4bb97d064e9b460c
+size 4856
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/model.safetensors b/dinov2_small_v1_global_nominmax/checkpoint-6000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..350853f9d61610045e9daaf87fd7ad0da3ef76f0
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6c32ba7e1a3137f7b4fd8a18b2be6de84aa447b1968ba6041ef264ff42d0d27
+size 228416552
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/optimizer.pt b/dinov2_small_v1_global_nominmax/checkpoint-6000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a6455d38da6c5faf35692bc8bf5a587bdec781fd
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06dd1abe9f459f0ebeaf8a798e05077ee310e675fa0c08de1e5a893212723461
+size 456914234
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_0.pth b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..573734be361b33074bad7976bc01de749f5d6479
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab4e74bbbd65d4c41a2a3bb1ea3fab4232793342026d4ffddf16cc632eae7a62
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_1.pth b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..00dad64f42e12d0afa1e60caca6210040dc49131
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a697538c03af5ff8d6452dc15809a2894197ffa5681aa24edbcffc60e254a44f
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_2.pth b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c9ae0f8d5f1ebe6f385347e0bbf38ca560f444b7
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fbf72c1542dcec16af36c0222d3bbd5999db9343d956cbbc8773f3f66517526
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_3.pth b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cb0bf81d6134a09c791172c3ca07fa2b29f739b5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75a68317944e9f214474713c9c1db897f06cab2c5a4c170c0c5134a6091190b5
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_4.pth b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..dc85f134ea3fbc6bc5d8d8a23706f5d0fe08b3fd
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cca38c6e7daab0237ed6da434a7df41d9f170d92b83d3a756bf26c5e8ec11fd
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_5.pth b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ef8acbd251c3c21ed20e9316184f5ddb5b7c29e2
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4b8da61059b60a1e04b5439ccfded49d90ee8196f887e88b6a19571fb41b8ce
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_6.pth b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5e9fed676d4b294296d8083ffa689e570dc39075
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6285187171ba9b778016af44ec47ceb282740a4dd23535d002b80bf5b2bdb97
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_7.pth b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..4afd09a3df8f4d942ac7d014c7a0b8fecf1a0423
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6473ee5bdf6c7a45259c4f355f40e16b480317db5b2b9d24efa31f9bf9d22202
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/scheduler.pt b/dinov2_small_v1_global_nominmax/checkpoint-6000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2a19f49862de9ff057d5dbf1579e3849486e1ddb
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c891c6464db0fd4c307d9cb1087a0312b5de2d7c04375cf97f0f59aa2882e71c
+size 1064
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/trainer_state.json b/dinov2_small_v1_global_nominmax/checkpoint-6000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..772dc69343f211b9af0770bca95a608779beafd2
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/trainer_state.json
@@ -0,0 +1,3621 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0241207725394283,
+  "eval_steps": 500,
+  "global_step": 6000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5854,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5466,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.55,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5355,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5183,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5221,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5063,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5058,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5059,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5077,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5089,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5075,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5017,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5022,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5046,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5174,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4867,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4998,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4982,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4931,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4859,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5014,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5076,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4979,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4999,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4908,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.478,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4838,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4879,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4761,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4853,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.477,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.491,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4678,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4814,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4822,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4792,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4791,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4839,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4821,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4783,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4771,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4747,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4692,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4684,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4775,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4714,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4744,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4752,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4727,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4693,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4699,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4694,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4745,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4754,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4796,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4708,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.466,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4706,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4756,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4687,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4702,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4739,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.472,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4632,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.458,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.463,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4655,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4682,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4627,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4608,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4718,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4612,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4615,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4621,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4575,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4602,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4672,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4613,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4685,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4643,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4637,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4595,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4593,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4585,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4538,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4559,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4623,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4605,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4569,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4649,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4572,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4578,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4617,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4646,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.457,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4564,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.455,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4591,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4543,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4536,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4547,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.451,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4534,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4537,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4485,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4518,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4601,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4609,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4606,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.456,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.4526,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4517,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4546,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4565,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4498,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4529,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4513,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4482,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4555,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4474,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4563,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4488,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4468,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4515,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4502,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4472,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4428,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.443,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4549,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4554,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4481,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4492,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4505,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.438,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4464,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4504,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4461,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4432,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4458,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4507,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4506,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4393,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4483,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4368,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4442,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4427,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4489,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4371,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4452,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4407,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4473,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4337,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4409,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4332,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4416,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.445,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4453,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4402,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4457,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4385,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4414,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4396,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4355,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4382,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4408,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4398,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4343,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4352,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4345,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4342,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4308,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4361,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4388,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4424,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.437,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4449,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4366,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4324,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4318,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4374,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4359,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4351,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4397,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4354,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4291,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4304,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4277,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4288,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4253,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4295,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4298,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4278,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4306,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4347,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.425,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4357,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4322,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4307,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4275,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4367,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4258,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4325,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4299,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4268,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4314,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.419,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4254,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4218,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4204,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4219,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4174,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4195,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4217,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4196,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4195,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4189,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4179,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4187,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4217,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4186,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4155,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4111,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4119,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4075,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4121,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4016,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4039,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4036,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4014,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4002,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4041,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.396,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4055,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3981,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3986,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3912,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3863,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3913,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3858,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3886,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3874,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3859,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3846,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3785,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3858,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3882,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3813,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3835,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3775,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3794,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3788,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3724,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3812,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3694,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3775,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3654,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.377,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3701,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3682,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3643,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3737,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3724,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.369,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.3634,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3674,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3569,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3675,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3634,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.362,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3589,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3591,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3549,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3611,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.362,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3554,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3564,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3604,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3574,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.36,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3511,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3529,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3565,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3501,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3475,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.353,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3524,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.35,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3497,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3439,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3552,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3406,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3531,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3459,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3476,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3421,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3451,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.342,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.3357,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.345,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.3437,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3381,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3365,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3412,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3413,
+      "step": 5000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3327,
+      "step": 5010
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3368,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3289,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3353,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3372,
+      "step": 5050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3321,
+      "step": 5060
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3324,
+      "step": 5070
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3292,
+      "step": 5080
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3231,
+      "step": 5090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3289,
+      "step": 5100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3249,
+      "step": 5110
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3259,
+      "step": 5120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3316,
+      "step": 5130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3353,
+      "step": 5140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3313,
+      "step": 5150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3332,
+      "step": 5160
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3239,
+      "step": 5170
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3232,
+      "step": 5180
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3295,
+      "step": 5190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3244,
+      "step": 5200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.3208,
+      "step": 5210
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.328,
+      "step": 5230
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3221,
+      "step": 5240
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3203,
+      "step": 5250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3146,
+      "step": 5260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.3184,
+      "step": 5270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.316,
+      "step": 5280
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.3189,
+      "step": 5290
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3119,
+      "step": 5300
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3159,
+      "step": 5310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3115,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3161,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3122,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3124,
+      "step": 5350
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3182,
+      "step": 5360
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3108,
+      "step": 5370
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3129,
+      "step": 5380
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.3149,
+      "step": 5390
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.314,
+      "step": 5400
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3141,
+      "step": 5410
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3063,
+      "step": 5420
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3087,
+      "step": 5430
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.3092,
+      "step": 5440
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.3059,
+      "step": 5450
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.32,
+      "step": 5460
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3131,
+      "step": 5470
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.304,
+      "step": 5480
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3136,
+      "step": 5490
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.307,
+      "step": 5500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.313,
+      "step": 5510
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5520
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3069,
+      "step": 5530
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3065,
+      "step": 5540
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3017,
+      "step": 5550
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3067,
+      "step": 5560
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5570
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3023,
+      "step": 5580
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.3007,
+      "step": 5590
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2929,
+      "step": 5600
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2994,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.301,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2986,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2997,
+      "step": 5640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.302,
+      "step": 5650
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2951,
+      "step": 5660
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2956,
+      "step": 5670
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2912,
+      "step": 5680
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2929,
+      "step": 5690
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2951,
+      "step": 5700
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2939,
+      "step": 5710
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2978,
+      "step": 5720
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2914,
+      "step": 5730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2915,
+      "step": 5740
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2919,
+      "step": 5750
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2883,
+      "step": 5760
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2928,
+      "step": 5770
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.29,
+      "step": 5780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5790
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2983,
+      "step": 5800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2883,
+      "step": 5810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2848,
+      "step": 5820
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2855,
+      "step": 5830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2859,
+      "step": 5840
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2875,
+      "step": 5860
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2848,
+      "step": 5870
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2858,
+      "step": 5880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2838,
+      "step": 5890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5900
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2831,
+      "step": 5910
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2828,
+      "step": 5920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2819,
+      "step": 5930
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2736,
+      "step": 5940
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2767,
+      "step": 5950
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2724,
+      "step": 5960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2785,
+      "step": 5970
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2725,
+      "step": 5980
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2783,
+      "step": 5990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2706,
+      "step": 6000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-6000/training_args.bin b/dinov2_small_v1_global_nominmax/checkpoint-6000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ca4990ad6e2d6dc5c39192293644d1530e9bbcd5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-6000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:898baa792356a9e4502a27fce84c053686ecb4d412bc621b4bb97d064e9b460c
+size 4856
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/model.safetensors b/dinov2_small_v1_global_nominmax/checkpoint-7000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..e7e3e634e93da6afa978b4d839a88edbc9ff9cab
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:00fd64f51a7758482a99f6dd284b3abf14ad907110c13d332b4ca856cde59bda
+size 228416552
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/optimizer.pt b/dinov2_small_v1_global_nominmax/checkpoint-7000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d9703baf58f82245ead47a5b9c62817c2f4b0ffc
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bfaf55844819163fb48978293901c4fb71e2eec1349d77ade59040cbcf927942
+size 456914234
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_0.pth b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..45df2fc5a65b8ed8addaca82738706bcbc3d354d
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79070614809b14d7d9ec529b08959bdf21acd157981c8790f18646e727ecb7a2
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_1.pth b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c948306a0faaf3e0501f52348767111b70b59cd7
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1271e97eea72dbeacaa4e28a8ba68c9001897356682fed9d29d54f495006853d
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_2.pth b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..57db45639fee1b370933ee6c97f112bfaac04f90
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:271fc7f1ca0ec9742ec537789e746b0401ed98834a2288108d94cb14a694f803
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_3.pth b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..22fa7388ac0e625e289ce37e6b79dd608755fcf3
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ef2b67b03c9810b08b56e4947a089f8d8ab7b89b86348255ecd2cd8c779631b
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_4.pth b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..8e99c104b1f483d5eda22b12f56398f6de67e26d
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:936416ff61c500377528f245d1c7100ba09e42993fe5edd62b64fc08b0ae180a
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_5.pth b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..2ef2b61a9a61c20ff816eb7d53d5a100b8a34071
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56af862adbec6e2b62670042d928d0eefc403b58310fa35f2528aa926c375ec7
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_6.pth b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..6d351974ca9fca1dda7126f22f4ee7f42fab3fb7
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6cbc2fc39c7d03cec9648447b91c124a625c578e8c93e9870d9bdb8892f049a6
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_7.pth b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..6d475c0fc82bccf2e2063e998260e0a996e01701
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63b8eff2583ba40500598c1be7bd1e1875b644238c72d057fe1345d8c310b5ac
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/scheduler.pt b/dinov2_small_v1_global_nominmax/checkpoint-7000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f80910e440c97c6830904084432a446f7a721cd5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5445e128a7bcf4177cc50c7224415006e388ca1019002cdd1a4b59301d72556c
+size 1064
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/trainer_state.json b/dinov2_small_v1_global_nominmax/checkpoint-7000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a1ea2647ad2df94fc8031d2e5b454166a65a4fc
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/trainer_state.json
@@ -0,0 +1,4221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.3614742346293327,
+  "eval_steps": 500,
+  "global_step": 7000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5854,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5466,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.55,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5355,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5183,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5221,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5063,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5058,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5059,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5077,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5089,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5075,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5017,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5022,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5046,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5174,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4867,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4998,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4982,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4931,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4859,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5014,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5076,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4979,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4999,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4908,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.478,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4838,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4879,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4761,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4853,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.477,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.491,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4678,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4814,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4822,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4792,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4791,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4839,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4821,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4783,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4771,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4747,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4692,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4684,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4775,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4714,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4744,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4752,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4727,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4693,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4699,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4694,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4745,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4754,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4796,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4708,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.466,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4706,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4756,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4687,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4702,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4739,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.472,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4632,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.458,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.463,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4655,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4682,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4627,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4608,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4718,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4612,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4615,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4621,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4575,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4602,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4672,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4613,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4685,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4643,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4637,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4595,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4593,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4585,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4538,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4559,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4623,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4605,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4569,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4649,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4572,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4578,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4617,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4646,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.457,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4564,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.455,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4591,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4543,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4536,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4547,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.451,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4534,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4537,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4485,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4518,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4601,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4609,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4606,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.456,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.4526,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4517,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4546,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4565,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4498,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4529,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4513,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4482,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4555,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4474,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4563,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4488,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4468,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4515,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4502,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4472,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4428,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.443,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4549,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4554,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4481,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4492,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4505,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.438,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4464,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4504,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4461,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4432,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4458,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4507,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4506,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4393,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4483,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4368,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4442,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4427,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4489,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4371,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4452,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4407,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4473,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4337,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4409,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4332,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4416,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.445,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4453,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4402,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4457,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4385,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4414,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4396,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4355,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4382,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4408,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4398,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4343,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4352,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4345,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4342,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4308,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4361,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4388,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4424,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.437,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4449,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4366,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4324,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4318,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4374,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4359,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4351,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4397,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4354,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4291,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4304,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4277,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4288,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4253,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4295,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4298,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4278,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4306,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4347,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.425,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4357,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4322,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4307,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4275,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4367,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4258,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4325,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4299,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4268,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4314,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.419,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4254,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4218,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4204,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4219,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4174,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4195,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4217,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4196,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4195,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4189,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4179,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4187,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4217,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4186,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4155,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4111,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4119,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4075,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4121,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4016,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4039,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4036,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4014,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4002,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4041,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.396,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4055,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3981,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3986,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3912,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3863,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3913,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3858,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3886,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3874,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3859,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3846,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3785,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3858,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3882,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3813,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3835,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3775,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3794,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3788,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3724,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3812,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3694,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3775,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3654,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.377,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3701,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3682,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3643,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3737,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3724,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.369,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.3634,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3674,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3569,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3675,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3634,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.362,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3589,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3591,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3549,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3611,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.362,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3554,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3564,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3604,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3574,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.36,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3511,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3529,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3565,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3501,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3475,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.353,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3524,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.35,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3497,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3439,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3552,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3406,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3531,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3459,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3476,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3421,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3451,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.342,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.3357,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.345,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.3437,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3381,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3365,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3412,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3413,
+      "step": 5000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3327,
+      "step": 5010
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3368,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3289,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3353,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3372,
+      "step": 5050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3321,
+      "step": 5060
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3324,
+      "step": 5070
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3292,
+      "step": 5080
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3231,
+      "step": 5090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3289,
+      "step": 5100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3249,
+      "step": 5110
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3259,
+      "step": 5120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3316,
+      "step": 5130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3353,
+      "step": 5140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3313,
+      "step": 5150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3332,
+      "step": 5160
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3239,
+      "step": 5170
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3232,
+      "step": 5180
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3295,
+      "step": 5190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3244,
+      "step": 5200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.3208,
+      "step": 5210
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.328,
+      "step": 5230
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3221,
+      "step": 5240
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3203,
+      "step": 5250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3146,
+      "step": 5260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.3184,
+      "step": 5270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.316,
+      "step": 5280
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.3189,
+      "step": 5290
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3119,
+      "step": 5300
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3159,
+      "step": 5310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3115,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3161,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3122,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3124,
+      "step": 5350
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3182,
+      "step": 5360
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3108,
+      "step": 5370
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3129,
+      "step": 5380
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.3149,
+      "step": 5390
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.314,
+      "step": 5400
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3141,
+      "step": 5410
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3063,
+      "step": 5420
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3087,
+      "step": 5430
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.3092,
+      "step": 5440
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.3059,
+      "step": 5450
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.32,
+      "step": 5460
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3131,
+      "step": 5470
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.304,
+      "step": 5480
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3136,
+      "step": 5490
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.307,
+      "step": 5500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.313,
+      "step": 5510
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5520
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3069,
+      "step": 5530
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3065,
+      "step": 5540
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3017,
+      "step": 5550
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3067,
+      "step": 5560
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5570
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3023,
+      "step": 5580
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.3007,
+      "step": 5590
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2929,
+      "step": 5600
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2994,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.301,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2986,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2997,
+      "step": 5640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.302,
+      "step": 5650
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2951,
+      "step": 5660
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2956,
+      "step": 5670
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2912,
+      "step": 5680
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2929,
+      "step": 5690
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2951,
+      "step": 5700
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2939,
+      "step": 5710
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2978,
+      "step": 5720
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2914,
+      "step": 5730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2915,
+      "step": 5740
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2919,
+      "step": 5750
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2883,
+      "step": 5760
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2928,
+      "step": 5770
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.29,
+      "step": 5780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5790
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2983,
+      "step": 5800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2883,
+      "step": 5810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2848,
+      "step": 5820
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2855,
+      "step": 5830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2859,
+      "step": 5840
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2875,
+      "step": 5860
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2848,
+      "step": 5870
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2858,
+      "step": 5880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2838,
+      "step": 5890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5900
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2831,
+      "step": 5910
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2828,
+      "step": 5920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2819,
+      "step": 5930
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2736,
+      "step": 5940
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2767,
+      "step": 5950
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2724,
+      "step": 5960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2785,
+      "step": 5970
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2725,
+      "step": 5980
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2783,
+      "step": 5990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2706,
+      "step": 6000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2692,
+      "step": 6010
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2758,
+      "step": 6020
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2716,
+      "step": 6030
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2659,
+      "step": 6040
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2652,
+      "step": 6050
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2702,
+      "step": 6060
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2686,
+      "step": 6070
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2719,
+      "step": 6080
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2671,
+      "step": 6090
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2675,
+      "step": 6100
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2691,
+      "step": 6110
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2617,
+      "step": 6120
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2733,
+      "step": 6130
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2673,
+      "step": 6140
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2606,
+      "step": 6150
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2599,
+      "step": 6160
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.262,
+      "step": 6170
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2579,
+      "step": 6180
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2578,
+      "step": 6190
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2636,
+      "step": 6200
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2636,
+      "step": 6210
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2589,
+      "step": 6220
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2593,
+      "step": 6230
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2565,
+      "step": 6240
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2504,
+      "step": 6250
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2555,
+      "step": 6260
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2613,
+      "step": 6270
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2618,
+      "step": 6280
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2562,
+      "step": 6290
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2542,
+      "step": 6300
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2555,
+      "step": 6310
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.257,
+      "step": 6320
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2576,
+      "step": 6330
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2581,
+      "step": 6340
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2601,
+      "step": 6350
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2545,
+      "step": 6360
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2526,
+      "step": 6370
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.247,
+      "step": 6380
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2504,
+      "step": 6390
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2479,
+      "step": 6400
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2494,
+      "step": 6410
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2482,
+      "step": 6420
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2465,
+      "step": 6430
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2475,
+      "step": 6440
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2547,
+      "step": 6450
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2504,
+      "step": 6460
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2536,
+      "step": 6470
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.249,
+      "step": 6480
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2499,
+      "step": 6490
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2536,
+      "step": 6500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.2481,
+      "step": 6510
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.2433,
+      "step": 6520
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.244,
+      "step": 6530
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2446,
+      "step": 6540
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2463,
+      "step": 6550
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2418,
+      "step": 6560
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2406,
+      "step": 6570
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2483,
+      "step": 6580
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2457,
+      "step": 6590
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2411,
+      "step": 6600
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2371,
+      "step": 6610
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2381,
+      "step": 6620
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2418,
+      "step": 6630
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2386,
+      "step": 6640
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2401,
+      "step": 6650
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2431,
+      "step": 6660
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2425,
+      "step": 6670
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2391,
+      "step": 6680
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2329,
+      "step": 6690
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2397,
+      "step": 6700
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2306,
+      "step": 6710
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2403,
+      "step": 6720
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2468,
+      "step": 6730
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2374,
+      "step": 6740
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2364,
+      "step": 6750
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2396,
+      "step": 6760
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2299,
+      "step": 6770
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2406,
+      "step": 6780
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2334,
+      "step": 6790
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2367,
+      "step": 6800
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2385,
+      "step": 6810
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2324,
+      "step": 6820
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2312,
+      "step": 6830
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2257,
+      "step": 6840
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.233,
+      "step": 6850
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2337,
+      "step": 6860
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2348,
+      "step": 6870
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2382,
+      "step": 6880
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2309,
+      "step": 6890
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.236,
+      "step": 6900
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2344,
+      "step": 6910
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2301,
+      "step": 6920
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2304,
+      "step": 6930
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2324,
+      "step": 6940
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2252,
+      "step": 6950
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2334,
+      "step": 6960
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2294,
+      "step": 6970
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 6980
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.223,
+      "step": 6990
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2272,
+      "step": 7000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-7000/training_args.bin b/dinov2_small_v1_global_nominmax/checkpoint-7000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ca4990ad6e2d6dc5c39192293644d1530e9bbcd5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-7000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:898baa792356a9e4502a27fce84c053686ecb4d412bc621b4bb97d064e9b460c
+size 4856
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/model.safetensors b/dinov2_small_v1_global_nominmax/checkpoint-8000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..092b1495115ea12fb9de2a6ff58475e1fcab2303
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22acf5d7a956fe7665e960974a50966ea3b5b48a5bda10bbc0aadaa5e7344b7f
+size 228416552
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/optimizer.pt b/dinov2_small_v1_global_nominmax/checkpoint-8000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e401c10cf5e6d57efefd92769edd521457e2f416
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:241c9885449bfbef370df8dd9c78240397fee981ffdbe46c7f3fe0ab03c7cc3b
+size 456914234
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_0.pth b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e1a7000b2a2a9b7e66995194c7def83701672681
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6041fbe6d64344a7314c873a27b77bc02d1482287c212202188d36f4518a660d
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_1.pth b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cca238fc6d5fb18fc6b45ee6a12e0b6439381ca2
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d98371ac0922b4275c5af92449fd0f7232b9afa2afae6edaeff681f991c72d33
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_2.pth b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..3ea9d02214ccd6e3e63d6f099324d649a3685139
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:780c0966ec7854b6c6e39f3e8fa4117f78c3e4680730e9fa98df60d0fe7eab92
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_3.pth b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e3324706fc72c3c4eb41b160a595cc8c3920f031
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c663a5160d7288107012a869bab934436414514919f633809e6413d97c7ad945
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_4.pth b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_4.pth
new file mode 100644
index 0000000000000000000000000000000000000000..474dbcb768ff8961068af90d14d5a6e7761b2bda
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_4.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ecb80974b256cc4c9e5cf15046168299cd65b34a55bc197240e42ed1c885ae5
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_5.pth b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_5.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c8a6ddebeed1dfa08a5e502a3e3253075dcc5eca
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_5.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d98f2e800c82fc9025786118eb439c72f5f75939349d70c7f996ee05d9efbc0d
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_6.pth b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_6.pth
new file mode 100644
index 0000000000000000000000000000000000000000..ba103507ac4db9853bb68e7c0c4c9864ee383c1b
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_6.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb1e4bb44c061e0bde43201c6a526de0aa0db51d6e379cb474dc0df9403bef4a
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_7.pth b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_7.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b78fa1dc7fa2a0a928683c007abd9004fb520aba
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/rng_state_7.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f30a1cb01c4953e01f8a0855c963b5ad50f1bb2cc19a235a31b0d397d966d70c
+size 15920
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/scheduler.pt b/dinov2_small_v1_global_nominmax/checkpoint-8000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0bc1a1fc7418b56600e0a35fa744b78dec1110ac
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33657ad014a83329319798b77914fe9b3b1f8d598a6a6cf79c7c92f3a50382cc
+size 1064
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/trainer_state.json b/dinov2_small_v1_global_nominmax/checkpoint-8000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..1793ff988fa3990f1d23ed7a5e7da66fbfce1751
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/trainer_state.json
@@ -0,0 +1,4821 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.6988276967192375,
+  "eval_steps": 500,
+  "global_step": 8000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5854,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5466,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.55,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5355,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5183,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5221,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5063,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5058,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5059,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5077,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5089,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5075,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5017,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5022,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5046,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5174,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4867,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4998,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4982,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4931,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4859,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5014,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5076,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4979,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4999,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4908,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.478,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4838,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4879,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4761,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4853,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.477,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.491,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4678,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4814,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4822,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4792,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4791,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4839,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4821,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4783,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4771,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4747,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4692,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4684,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4775,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4714,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4744,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4752,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4727,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4693,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4699,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4694,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4745,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4754,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4796,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4708,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.466,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4706,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4756,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4687,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4702,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4739,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.472,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4632,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.458,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.463,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4655,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4682,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4627,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4608,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4718,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4612,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4615,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4621,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4575,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4602,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4672,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4613,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4685,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4643,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4637,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4595,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4593,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4585,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4538,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4559,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4623,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4605,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4569,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4649,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4572,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4578,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4617,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4646,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.457,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4564,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.455,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4591,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4543,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4536,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4547,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.451,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4534,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4537,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4485,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4518,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4601,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4609,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4606,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.456,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.4526,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4517,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4546,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4565,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4498,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4529,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4513,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4482,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4555,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4474,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4563,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4488,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4468,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4515,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4502,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4472,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4428,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.443,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4549,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4554,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4481,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4492,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4505,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.438,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4464,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4504,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4461,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4432,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4458,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4507,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4506,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4393,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4483,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4368,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4442,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4427,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4489,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4371,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4452,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4407,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4473,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4337,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4409,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4332,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4416,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.445,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4453,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4402,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4457,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4385,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4414,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4396,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4355,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4382,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4408,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4398,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4343,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4352,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4345,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4342,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4308,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4361,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4388,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4424,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.437,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4449,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4366,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4324,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4318,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4374,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4359,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4351,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4397,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4354,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4291,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4304,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4277,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4288,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4253,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4295,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4298,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4278,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4306,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4347,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.425,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4357,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4322,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4307,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4275,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4367,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4258,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4325,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4299,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4268,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4314,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.419,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4254,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4218,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4204,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4219,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4174,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4195,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4217,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4196,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4195,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4189,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4179,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4187,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4217,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4186,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4155,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4111,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4119,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4075,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4121,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4016,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4039,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4036,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4014,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4002,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4041,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.396,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4055,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3981,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3986,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3912,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3863,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3913,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3858,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3886,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3874,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3859,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3846,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3785,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3858,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3882,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3813,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3835,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3775,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3794,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3788,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3724,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3812,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3694,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3775,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3654,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.377,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3701,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3682,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3643,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3737,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3724,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.369,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.3634,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3674,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3569,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3675,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3634,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.362,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3589,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3591,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3549,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3611,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.362,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3554,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3564,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3604,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3574,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.36,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3511,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3529,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3565,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3501,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3475,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.353,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3524,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.35,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3497,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3439,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3552,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3406,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3531,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3459,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3476,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3421,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3451,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.342,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.3357,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.345,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.3437,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3381,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3365,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3412,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3413,
+      "step": 5000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3327,
+      "step": 5010
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3368,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3289,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3353,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3372,
+      "step": 5050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3321,
+      "step": 5060
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3324,
+      "step": 5070
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3292,
+      "step": 5080
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3231,
+      "step": 5090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3289,
+      "step": 5100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3249,
+      "step": 5110
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3259,
+      "step": 5120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3316,
+      "step": 5130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3353,
+      "step": 5140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3313,
+      "step": 5150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3332,
+      "step": 5160
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3239,
+      "step": 5170
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3232,
+      "step": 5180
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3295,
+      "step": 5190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3244,
+      "step": 5200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.3208,
+      "step": 5210
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.328,
+      "step": 5230
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3221,
+      "step": 5240
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3203,
+      "step": 5250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3146,
+      "step": 5260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.3184,
+      "step": 5270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.316,
+      "step": 5280
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.3189,
+      "step": 5290
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3119,
+      "step": 5300
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3159,
+      "step": 5310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3115,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3161,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3122,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3124,
+      "step": 5350
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3182,
+      "step": 5360
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3108,
+      "step": 5370
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3129,
+      "step": 5380
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.3149,
+      "step": 5390
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.314,
+      "step": 5400
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3141,
+      "step": 5410
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3063,
+      "step": 5420
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3087,
+      "step": 5430
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.3092,
+      "step": 5440
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.3059,
+      "step": 5450
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.32,
+      "step": 5460
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3131,
+      "step": 5470
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.304,
+      "step": 5480
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3136,
+      "step": 5490
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.307,
+      "step": 5500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.313,
+      "step": 5510
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5520
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3069,
+      "step": 5530
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3065,
+      "step": 5540
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3017,
+      "step": 5550
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3067,
+      "step": 5560
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5570
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3023,
+      "step": 5580
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.3007,
+      "step": 5590
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2929,
+      "step": 5600
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2994,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.301,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2986,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2997,
+      "step": 5640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.302,
+      "step": 5650
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2951,
+      "step": 5660
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2956,
+      "step": 5670
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2912,
+      "step": 5680
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2929,
+      "step": 5690
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2951,
+      "step": 5700
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2939,
+      "step": 5710
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2978,
+      "step": 5720
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2914,
+      "step": 5730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2915,
+      "step": 5740
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2919,
+      "step": 5750
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2883,
+      "step": 5760
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2928,
+      "step": 5770
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.29,
+      "step": 5780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5790
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2983,
+      "step": 5800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2883,
+      "step": 5810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2848,
+      "step": 5820
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2855,
+      "step": 5830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2859,
+      "step": 5840
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2875,
+      "step": 5860
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2848,
+      "step": 5870
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2858,
+      "step": 5880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2838,
+      "step": 5890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5900
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2831,
+      "step": 5910
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2828,
+      "step": 5920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2819,
+      "step": 5930
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2736,
+      "step": 5940
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2767,
+      "step": 5950
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2724,
+      "step": 5960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2785,
+      "step": 5970
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2725,
+      "step": 5980
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2783,
+      "step": 5990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2706,
+      "step": 6000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2692,
+      "step": 6010
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2758,
+      "step": 6020
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2716,
+      "step": 6030
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2659,
+      "step": 6040
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2652,
+      "step": 6050
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2702,
+      "step": 6060
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2686,
+      "step": 6070
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2719,
+      "step": 6080
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2671,
+      "step": 6090
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2675,
+      "step": 6100
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2691,
+      "step": 6110
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2617,
+      "step": 6120
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2733,
+      "step": 6130
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2673,
+      "step": 6140
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2606,
+      "step": 6150
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2599,
+      "step": 6160
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.262,
+      "step": 6170
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2579,
+      "step": 6180
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2578,
+      "step": 6190
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2636,
+      "step": 6200
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2636,
+      "step": 6210
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2589,
+      "step": 6220
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2593,
+      "step": 6230
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2565,
+      "step": 6240
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2504,
+      "step": 6250
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2555,
+      "step": 6260
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2613,
+      "step": 6270
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2618,
+      "step": 6280
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2562,
+      "step": 6290
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2542,
+      "step": 6300
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2555,
+      "step": 6310
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.257,
+      "step": 6320
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2576,
+      "step": 6330
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2581,
+      "step": 6340
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2601,
+      "step": 6350
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2545,
+      "step": 6360
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2526,
+      "step": 6370
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.247,
+      "step": 6380
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2504,
+      "step": 6390
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2479,
+      "step": 6400
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2494,
+      "step": 6410
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2482,
+      "step": 6420
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2465,
+      "step": 6430
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2475,
+      "step": 6440
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2547,
+      "step": 6450
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2504,
+      "step": 6460
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2536,
+      "step": 6470
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.249,
+      "step": 6480
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2499,
+      "step": 6490
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2536,
+      "step": 6500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.2481,
+      "step": 6510
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.2433,
+      "step": 6520
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.244,
+      "step": 6530
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2446,
+      "step": 6540
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2463,
+      "step": 6550
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2418,
+      "step": 6560
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2406,
+      "step": 6570
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2483,
+      "step": 6580
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2457,
+      "step": 6590
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2411,
+      "step": 6600
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2371,
+      "step": 6610
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2381,
+      "step": 6620
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2418,
+      "step": 6630
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2386,
+      "step": 6640
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2401,
+      "step": 6650
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2431,
+      "step": 6660
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2425,
+      "step": 6670
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2391,
+      "step": 6680
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2329,
+      "step": 6690
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2397,
+      "step": 6700
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2306,
+      "step": 6710
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2403,
+      "step": 6720
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2468,
+      "step": 6730
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2374,
+      "step": 6740
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2364,
+      "step": 6750
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2396,
+      "step": 6760
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2299,
+      "step": 6770
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2406,
+      "step": 6780
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2334,
+      "step": 6790
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2367,
+      "step": 6800
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2385,
+      "step": 6810
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2324,
+      "step": 6820
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2312,
+      "step": 6830
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2257,
+      "step": 6840
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.233,
+      "step": 6850
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2337,
+      "step": 6860
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2348,
+      "step": 6870
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2382,
+      "step": 6880
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2309,
+      "step": 6890
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.236,
+      "step": 6900
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2344,
+      "step": 6910
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2301,
+      "step": 6920
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2304,
+      "step": 6930
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2324,
+      "step": 6940
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2252,
+      "step": 6950
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2334,
+      "step": 6960
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2294,
+      "step": 6970
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 6980
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.223,
+      "step": 6990
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2272,
+      "step": 7000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2267,
+      "step": 7010
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2299,
+      "step": 7020
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2228,
+      "step": 7030
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2315,
+      "step": 7040
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2262,
+      "step": 7050
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2291,
+      "step": 7060
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2242,
+      "step": 7070
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2215,
+      "step": 7080
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2272,
+      "step": 7090
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2275,
+      "step": 7100
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.225,
+      "step": 7110
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2259,
+      "step": 7120
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.226,
+      "step": 7130
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 7140
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2206,
+      "step": 7150
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2229,
+      "step": 7160
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2269,
+      "step": 7170
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2237,
+      "step": 7180
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2153,
+      "step": 7190
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2271,
+      "step": 7200
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2294,
+      "step": 7210
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.225,
+      "step": 7220
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2245,
+      "step": 7230
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.226,
+      "step": 7240
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2226,
+      "step": 7250
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2276,
+      "step": 7260
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2223,
+      "step": 7270
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2191,
+      "step": 7280
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2263,
+      "step": 7290
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2208,
+      "step": 7300
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2259,
+      "step": 7310
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2173,
+      "step": 7320
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.218,
+      "step": 7330
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2148,
+      "step": 7340
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2087,
+      "step": 7350
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2257,
+      "step": 7360
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2258,
+      "step": 7370
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2198,
+      "step": 7380
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2185,
+      "step": 7390
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2173,
+      "step": 7400
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.217,
+      "step": 7410
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2164,
+      "step": 7420
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2161,
+      "step": 7430
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2213,
+      "step": 7440
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.219,
+      "step": 7450
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2142,
+      "step": 7460
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2242,
+      "step": 7470
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2233,
+      "step": 7480
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.212,
+      "step": 7490
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.214,
+      "step": 7500
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2171,
+      "step": 7510
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2191,
+      "step": 7520
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2199,
+      "step": 7530
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2148,
+      "step": 7540
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2125,
+      "step": 7550
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2209,
+      "step": 7560
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2135,
+      "step": 7570
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2147,
+      "step": 7580
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2108,
+      "step": 7590
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.216,
+      "step": 7600
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.218,
+      "step": 7610
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2163,
+      "step": 7620
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2093,
+      "step": 7630
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2146,
+      "step": 7640
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2181,
+      "step": 7650
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2119,
+      "step": 7660
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2105,
+      "step": 7670
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.212,
+      "step": 7680
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2129,
+      "step": 7690
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.206,
+      "step": 7700
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2146,
+      "step": 7710
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2158,
+      "step": 7720
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2044,
+      "step": 7730
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2062,
+      "step": 7740
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.208,
+      "step": 7750
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2103,
+      "step": 7760
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2103,
+      "step": 7770
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2071,
+      "step": 7780
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2099,
+      "step": 7790
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2207,
+      "step": 7800
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2087,
+      "step": 7810
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2093,
+      "step": 7820
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2104,
+      "step": 7830
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2057,
+      "step": 7840
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2089,
+      "step": 7850
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2128,
+      "step": 7860
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2088,
+      "step": 7870
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2062,
+      "step": 7880
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2095,
+      "step": 7890
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2091,
+      "step": 7900
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2142,
+      "step": 7910
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2134,
+      "step": 7920
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.205,
+      "step": 7930
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2071,
+      "step": 7940
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2059,
+      "step": 7950
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2037,
+      "step": 7960
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2082,
+      "step": 7970
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2078,
+      "step": 7980
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2037,
+      "step": 7990
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2057,
+      "step": 8000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/dinov2_small_v1_global_nominmax/checkpoint-8000/training_args.bin b/dinov2_small_v1_global_nominmax/checkpoint-8000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ca4990ad6e2d6dc5c39192293644d1530e9bbcd5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/checkpoint-8000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:898baa792356a9e4502a27fce84c053686ecb4d412bc621b4bb97d064e9b460c
+size 4856
diff --git a/dinov2_small_v1_global_nominmax/model.safetensors b/dinov2_small_v1_global_nominmax/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..70961870ba7ecbb6648eb4fd7308ff5bc73b6be8
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:977e19446a37f20a4cf8d9e9c07fdb9663123a6cfdb71f20690d4bbc8637b418
+size 228416552
diff --git a/dinov2_small_v1_global_nominmax/trainer_state.json b/dinov2_small_v1_global_nominmax/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..6758e180ed2f9c656ff50f748cbebebf39ed3ba8
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/trainer_state.json
@@ -0,0 +1,5364 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9997469849034326,
+  "eval_steps": 500,
+  "global_step": 8892,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.5854,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5466,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5426,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 0.5393,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.55,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5355,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.5387,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5299,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5183,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.5188,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5186,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5135,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.5221,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5163,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5063,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.5134,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5058,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5059,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 0.5077,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5089,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5075,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 0.5017,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5171,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5022,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.5005,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5025,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5046,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 0.5174,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.5037,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4952,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4914,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4867,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4998,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 0.482,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.498,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4982,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4931,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4859,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4852,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4942,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 0.5014,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.5076,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4962,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5002,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4979,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4999,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4908,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4884,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4801,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.478,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4838,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4879,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4761,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4853,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4938,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.477,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.491,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4678,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4814,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4822,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4792,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4791,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4839,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4846,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4821,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4783,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4771,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4747,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4802,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4692,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4684,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4775,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4714,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4744,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4752,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4727,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4693,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4699,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4694,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4745,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4754,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4796,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4708,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4768,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4709,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.466,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4706,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4756,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4687,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4702,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4739,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.472,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4632,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.458,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.463,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4655,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4778,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4746,
+      "step": 1110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4738,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4682,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4627,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4608,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4718,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4612,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4645,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4615,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4621,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4575,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4602,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4672,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4613,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.4674,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.4685,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4643,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.4637,
+      "step": 1400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4595,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.4584,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4644,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4667,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.4593,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4585,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4634,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.4538,
+      "step": 1490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4559,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.4535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 1530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4623,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4605,
+      "step": 1550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 1560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4569,
+      "step": 1580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4611,
+      "step": 1590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4649,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4572,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1640
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.4578,
+      "step": 1670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4617,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.4646,
+      "step": 1700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.457,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.4629,
+      "step": 1730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.46,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4598,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.4564,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.455,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.4591,
+      "step": 1790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.4543,
+      "step": 1820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4536,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 1850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4547,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4669,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 1880
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.451,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.4534,
+      "step": 1910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4537,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4485,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4576,
+      "step": 1940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4518,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4601,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 1970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4607,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4609,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.4574,
+      "step": 2030
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4606,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.456,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.4526,
+      "step": 2080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4517,
+      "step": 2090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.4546,
+      "step": 2110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4565,
+      "step": 2120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4589,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.449,
+      "step": 2140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2150
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4562,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.4498,
+      "step": 2170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2180
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4529,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.4539,
+      "step": 2200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4531,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.4513,
+      "step": 2230
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4519,
+      "step": 2240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4482,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4555,
+      "step": 2260
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4514,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4474,
+      "step": 2290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4561,
+      "step": 2300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4563,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.4509,
+      "step": 2320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4488,
+      "step": 2330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4476,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.4524,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4468,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4515,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 2380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4502,
+      "step": 2390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4472,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4428,
+      "step": 2410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.443,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4549,
+      "step": 2440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4554,
+      "step": 2450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4481,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4492,
+      "step": 2480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4505,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.438,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4503,
+      "step": 2530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.4464,
+      "step": 2560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4504,
+      "step": 2570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.4461,
+      "step": 2590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.4479,
+      "step": 2620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4432,
+      "step": 2630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.4458,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4478,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4507,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.4471,
+      "step": 2680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4463,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4506,
+      "step": 2720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.4511,
+      "step": 2740
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4393,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 2770
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4451,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4483,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4469,
+      "step": 2800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.446,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.4443,
+      "step": 2830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4368,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4442,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.4427,
+      "step": 2860
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4489,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4371,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.4452,
+      "step": 2890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4407,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.4473,
+      "step": 2910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4337,
+      "step": 2920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4409,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.4332,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4437,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.4416,
+      "step": 2970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 2990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3010
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3020
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.4413,
+      "step": 3030
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3040
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.445,
+      "step": 3050
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.4453,
+      "step": 3060
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4402,
+      "step": 3070
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3080
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4434,
+      "step": 3090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4457,
+      "step": 3100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4475,
+      "step": 3110
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.4401,
+      "step": 3120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4385,
+      "step": 3130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4414,
+      "step": 3140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.4396,
+      "step": 3150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4355,
+      "step": 3160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4389,
+      "step": 3170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.4384,
+      "step": 3180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4382,
+      "step": 3190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4408,
+      "step": 3200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.4398,
+      "step": 3210
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3220
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4343,
+      "step": 3230
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.4352,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4345,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4363,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.4342,
+      "step": 3270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4308,
+      "step": 3280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.4361,
+      "step": 3290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3300
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3310
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3320
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3330
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4388,
+      "step": 3340
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4387,
+      "step": 3350
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3360
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3370
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4424,
+      "step": 3380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.4372,
+      "step": 3390
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.437,
+      "step": 3410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.4449,
+      "step": 3420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4426,
+      "step": 3440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.4415,
+      "step": 3450
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4366,
+      "step": 3460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4324,
+      "step": 3470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4318,
+      "step": 3490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4405,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4374,
+      "step": 3510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4359,
+      "step": 3520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4351,
+      "step": 3530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4397,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4354,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4327,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4358,
+      "step": 3570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4291,
+      "step": 3580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4304,
+      "step": 3590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.4277,
+      "step": 3600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4319,
+      "step": 3610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4392,
+      "step": 3630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4288,
+      "step": 3640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4305,
+      "step": 3650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4253,
+      "step": 3660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4295,
+      "step": 3670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4298,
+      "step": 3680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4278,
+      "step": 3700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4306,
+      "step": 3710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.4347,
+      "step": 3740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.435,
+      "step": 3750
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.425,
+      "step": 3760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4357,
+      "step": 3780
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4349,
+      "step": 3790
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.4322,
+      "step": 3800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4338,
+      "step": 3810
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4307,
+      "step": 3820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4283,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4275,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.4367,
+      "step": 3860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.4258,
+      "step": 3880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.431,
+      "step": 3890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4325,
+      "step": 3900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 3910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.4299,
+      "step": 3920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4268,
+      "step": 3930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4313,
+      "step": 3940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.4263,
+      "step": 3950
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4335,
+      "step": 3960
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.4314,
+      "step": 3970
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.419,
+      "step": 3980
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4254,
+      "step": 3990
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4323,
+      "step": 4000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.4218,
+      "step": 4010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4204,
+      "step": 4020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4219,
+      "step": 4030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.4174,
+      "step": 4040
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4195,
+      "step": 4050
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4217,
+      "step": 4060
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.4196,
+      "step": 4070
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4195,
+      "step": 4080
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4189,
+      "step": 4090
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4179,
+      "step": 4100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4187,
+      "step": 4110
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4217,
+      "step": 4120
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.4186,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4155,
+      "step": 4140
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4111,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.4119,
+      "step": 4160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4075,
+      "step": 4170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4121,
+      "step": 4180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.4016,
+      "step": 4190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4039,
+      "step": 4200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4036,
+      "step": 4210
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.4014,
+      "step": 4220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4002,
+      "step": 4230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.4041,
+      "step": 4240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.396,
+      "step": 4250
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.4055,
+      "step": 4260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3981,
+      "step": 4270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3986,
+      "step": 4280
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3912,
+      "step": 4290
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3863,
+      "step": 4300
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3913,
+      "step": 4310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3858,
+      "step": 4320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3886,
+      "step": 4330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3874,
+      "step": 4340
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3859,
+      "step": 4350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3846,
+      "step": 4360
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3785,
+      "step": 4370
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3858,
+      "step": 4380
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3882,
+      "step": 4390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3813,
+      "step": 4400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3835,
+      "step": 4410
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3775,
+      "step": 4420
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3794,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3788,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3724,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3812,
+      "step": 4460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3694,
+      "step": 4470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3775,
+      "step": 4480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3654,
+      "step": 4490
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.377,
+      "step": 4500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3701,
+      "step": 4510
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3682,
+      "step": 4520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3643,
+      "step": 4530
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3737,
+      "step": 4540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3724,
+      "step": 4550
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.369,
+      "step": 4560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.3634,
+      "step": 4570
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3674,
+      "step": 4580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3569,
+      "step": 4590
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3675,
+      "step": 4600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3634,
+      "step": 4610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.362,
+      "step": 4620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3589,
+      "step": 4630
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3591,
+      "step": 4640
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3549,
+      "step": 4650
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3611,
+      "step": 4660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.362,
+      "step": 4670
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3554,
+      "step": 4680
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3564,
+      "step": 4690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3604,
+      "step": 4700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3574,
+      "step": 4710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.36,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3511,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3529,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3565,
+      "step": 4750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3501,
+      "step": 4760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3475,
+      "step": 4770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.353,
+      "step": 4780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3524,
+      "step": 4790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.35,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3497,
+      "step": 4810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3439,
+      "step": 4820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3552,
+      "step": 4830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.3406,
+      "step": 4840
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3531,
+      "step": 4850
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3459,
+      "step": 4860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3476,
+      "step": 4870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3421,
+      "step": 4880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3451,
+      "step": 4890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.342,
+      "step": 4900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.3357,
+      "step": 4910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4930
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.345,
+      "step": 4940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.3437,
+      "step": 4950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.339,
+      "step": 4960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3381,
+      "step": 4970
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3365,
+      "step": 4980
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3412,
+      "step": 4990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3413,
+      "step": 5000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3327,
+      "step": 5010
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3368,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3289,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3353,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3372,
+      "step": 5050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3321,
+      "step": 5060
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3324,
+      "step": 5070
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3292,
+      "step": 5080
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3231,
+      "step": 5090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3289,
+      "step": 5100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3249,
+      "step": 5110
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3259,
+      "step": 5120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3316,
+      "step": 5130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3353,
+      "step": 5140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3313,
+      "step": 5150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3332,
+      "step": 5160
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3239,
+      "step": 5170
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3232,
+      "step": 5180
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3295,
+      "step": 5190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3244,
+      "step": 5200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.3208,
+      "step": 5210
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.328,
+      "step": 5230
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3221,
+      "step": 5240
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3203,
+      "step": 5250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3146,
+      "step": 5260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.3184,
+      "step": 5270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.316,
+      "step": 5280
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.3189,
+      "step": 5290
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3119,
+      "step": 5300
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3159,
+      "step": 5310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3115,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3161,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3122,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3124,
+      "step": 5350
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3182,
+      "step": 5360
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3108,
+      "step": 5370
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3129,
+      "step": 5380
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.3149,
+      "step": 5390
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.314,
+      "step": 5400
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3141,
+      "step": 5410
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3063,
+      "step": 5420
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3087,
+      "step": 5430
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.3092,
+      "step": 5440
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.3059,
+      "step": 5450
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.32,
+      "step": 5460
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3131,
+      "step": 5470
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.304,
+      "step": 5480
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3136,
+      "step": 5490
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.307,
+      "step": 5500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.313,
+      "step": 5510
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5520
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3069,
+      "step": 5530
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3065,
+      "step": 5540
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3017,
+      "step": 5550
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3067,
+      "step": 5560
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3079,
+      "step": 5570
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3023,
+      "step": 5580
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.3007,
+      "step": 5590
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2929,
+      "step": 5600
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2994,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.301,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2986,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2997,
+      "step": 5640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.302,
+      "step": 5650
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2951,
+      "step": 5660
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2956,
+      "step": 5670
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2912,
+      "step": 5680
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2929,
+      "step": 5690
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2951,
+      "step": 5700
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2939,
+      "step": 5710
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2978,
+      "step": 5720
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2914,
+      "step": 5730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2915,
+      "step": 5740
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2919,
+      "step": 5750
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2883,
+      "step": 5760
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2928,
+      "step": 5770
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.29,
+      "step": 5780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5790
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2983,
+      "step": 5800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2883,
+      "step": 5810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2848,
+      "step": 5820
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2855,
+      "step": 5830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2859,
+      "step": 5840
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2875,
+      "step": 5860
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2848,
+      "step": 5870
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2858,
+      "step": 5880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2838,
+      "step": 5890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 5900
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2831,
+      "step": 5910
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2828,
+      "step": 5920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2819,
+      "step": 5930
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2736,
+      "step": 5940
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2767,
+      "step": 5950
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2724,
+      "step": 5960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2785,
+      "step": 5970
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2725,
+      "step": 5980
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2783,
+      "step": 5990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2706,
+      "step": 6000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2692,
+      "step": 6010
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2758,
+      "step": 6020
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2716,
+      "step": 6030
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2659,
+      "step": 6040
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2652,
+      "step": 6050
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2702,
+      "step": 6060
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2686,
+      "step": 6070
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2719,
+      "step": 6080
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2671,
+      "step": 6090
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2675,
+      "step": 6100
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2691,
+      "step": 6110
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2617,
+      "step": 6120
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2733,
+      "step": 6130
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2673,
+      "step": 6140
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2606,
+      "step": 6150
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2599,
+      "step": 6160
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.262,
+      "step": 6170
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2579,
+      "step": 6180
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2578,
+      "step": 6190
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2636,
+      "step": 6200
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2636,
+      "step": 6210
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2589,
+      "step": 6220
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2593,
+      "step": 6230
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2565,
+      "step": 6240
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2504,
+      "step": 6250
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2555,
+      "step": 6260
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2613,
+      "step": 6270
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2618,
+      "step": 6280
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2562,
+      "step": 6290
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2542,
+      "step": 6300
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2555,
+      "step": 6310
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.257,
+      "step": 6320
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2576,
+      "step": 6330
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2581,
+      "step": 6340
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2601,
+      "step": 6350
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2545,
+      "step": 6360
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2526,
+      "step": 6370
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.247,
+      "step": 6380
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2504,
+      "step": 6390
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2479,
+      "step": 6400
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2494,
+      "step": 6410
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2482,
+      "step": 6420
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2465,
+      "step": 6430
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2475,
+      "step": 6440
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2547,
+      "step": 6450
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2504,
+      "step": 6460
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2536,
+      "step": 6470
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.249,
+      "step": 6480
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2499,
+      "step": 6490
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2536,
+      "step": 6500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.2481,
+      "step": 6510
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.2433,
+      "step": 6520
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.244,
+      "step": 6530
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2446,
+      "step": 6540
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2463,
+      "step": 6550
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2418,
+      "step": 6560
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2406,
+      "step": 6570
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2483,
+      "step": 6580
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2457,
+      "step": 6590
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2411,
+      "step": 6600
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2371,
+      "step": 6610
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2381,
+      "step": 6620
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2418,
+      "step": 6630
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2386,
+      "step": 6640
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2401,
+      "step": 6650
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2431,
+      "step": 6660
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2425,
+      "step": 6670
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2391,
+      "step": 6680
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2329,
+      "step": 6690
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2397,
+      "step": 6700
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2306,
+      "step": 6710
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2403,
+      "step": 6720
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2468,
+      "step": 6730
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2374,
+      "step": 6740
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2364,
+      "step": 6750
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2396,
+      "step": 6760
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2299,
+      "step": 6770
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2406,
+      "step": 6780
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2334,
+      "step": 6790
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2367,
+      "step": 6800
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2385,
+      "step": 6810
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2324,
+      "step": 6820
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2312,
+      "step": 6830
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2257,
+      "step": 6840
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.233,
+      "step": 6850
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2337,
+      "step": 6860
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2348,
+      "step": 6870
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2382,
+      "step": 6880
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2309,
+      "step": 6890
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.236,
+      "step": 6900
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2344,
+      "step": 6910
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2301,
+      "step": 6920
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2304,
+      "step": 6930
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2324,
+      "step": 6940
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2252,
+      "step": 6950
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2334,
+      "step": 6960
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2294,
+      "step": 6970
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 6980
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.223,
+      "step": 6990
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2272,
+      "step": 7000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2267,
+      "step": 7010
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2299,
+      "step": 7020
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2228,
+      "step": 7030
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2315,
+      "step": 7040
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2262,
+      "step": 7050
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2291,
+      "step": 7060
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2242,
+      "step": 7070
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2215,
+      "step": 7080
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2272,
+      "step": 7090
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2275,
+      "step": 7100
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.225,
+      "step": 7110
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2259,
+      "step": 7120
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.226,
+      "step": 7130
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 7140
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2206,
+      "step": 7150
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2229,
+      "step": 7160
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2269,
+      "step": 7170
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2237,
+      "step": 7180
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2153,
+      "step": 7190
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2271,
+      "step": 7200
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2294,
+      "step": 7210
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.225,
+      "step": 7220
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2245,
+      "step": 7230
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.226,
+      "step": 7240
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2226,
+      "step": 7250
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2276,
+      "step": 7260
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2223,
+      "step": 7270
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2191,
+      "step": 7280
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2263,
+      "step": 7290
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2208,
+      "step": 7300
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2259,
+      "step": 7310
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2173,
+      "step": 7320
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.218,
+      "step": 7330
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2148,
+      "step": 7340
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2087,
+      "step": 7350
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2257,
+      "step": 7360
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2258,
+      "step": 7370
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2198,
+      "step": 7380
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2185,
+      "step": 7390
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2173,
+      "step": 7400
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.217,
+      "step": 7410
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2164,
+      "step": 7420
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2161,
+      "step": 7430
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2213,
+      "step": 7440
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.219,
+      "step": 7450
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2142,
+      "step": 7460
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2242,
+      "step": 7470
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2233,
+      "step": 7480
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.212,
+      "step": 7490
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.214,
+      "step": 7500
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2171,
+      "step": 7510
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2191,
+      "step": 7520
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2199,
+      "step": 7530
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2148,
+      "step": 7540
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2125,
+      "step": 7550
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2209,
+      "step": 7560
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2135,
+      "step": 7570
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2147,
+      "step": 7580
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2108,
+      "step": 7590
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.216,
+      "step": 7600
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.218,
+      "step": 7610
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2163,
+      "step": 7620
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2093,
+      "step": 7630
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2146,
+      "step": 7640
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2181,
+      "step": 7650
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2119,
+      "step": 7660
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2105,
+      "step": 7670
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.212,
+      "step": 7680
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2129,
+      "step": 7690
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.206,
+      "step": 7700
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2146,
+      "step": 7710
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2158,
+      "step": 7720
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2044,
+      "step": 7730
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2062,
+      "step": 7740
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.208,
+      "step": 7750
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2103,
+      "step": 7760
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2103,
+      "step": 7770
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2071,
+      "step": 7780
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2099,
+      "step": 7790
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2207,
+      "step": 7800
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2087,
+      "step": 7810
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2093,
+      "step": 7820
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2104,
+      "step": 7830
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2057,
+      "step": 7840
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2089,
+      "step": 7850
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2128,
+      "step": 7860
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2088,
+      "step": 7870
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2062,
+      "step": 7880
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2095,
+      "step": 7890
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2091,
+      "step": 7900
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2142,
+      "step": 7910
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2134,
+      "step": 7920
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.205,
+      "step": 7930
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2071,
+      "step": 7940
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2059,
+      "step": 7950
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2037,
+      "step": 7960
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2082,
+      "step": 7970
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2078,
+      "step": 7980
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2037,
+      "step": 7990
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2057,
+      "step": 8000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2052,
+      "step": 8010
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2039,
+      "step": 8020
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2052,
+      "step": 8030
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2029,
+      "step": 8040
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2056,
+      "step": 8050
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2059,
+      "step": 8060
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2013,
+      "step": 8070
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2026,
+      "step": 8080
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2042,
+      "step": 8090
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2093,
+      "step": 8100
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2028,
+      "step": 8110
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.1982,
+      "step": 8120
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2014,
+      "step": 8130
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.211,
+      "step": 8140
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2018,
+      "step": 8150
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2021,
+      "step": 8160
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2102,
+      "step": 8170
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.1996,
+      "step": 8180
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2055,
+      "step": 8190
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2025,
+      "step": 8200
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.1975,
+      "step": 8210
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2049,
+      "step": 8220
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2004,
+      "step": 8230
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.1995,
+      "step": 8240
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.1995,
+      "step": 8250
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.1942,
+      "step": 8260
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2027,
+      "step": 8270
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.199,
+      "step": 8280
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2005,
+      "step": 8290
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.1987,
+      "step": 8300
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2051,
+      "step": 8310
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.1996,
+      "step": 8320
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.1976,
+      "step": 8330
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.1999,
+      "step": 8340
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.205,
+      "step": 8350
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.1972,
+      "step": 8360
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.1985,
+      "step": 8370
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.1995,
+      "step": 8380
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2038,
+      "step": 8390
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.1942,
+      "step": 8400
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.1976,
+      "step": 8410
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.1942,
+      "step": 8420
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.1911,
+      "step": 8430
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.1996,
+      "step": 8440
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.1974,
+      "step": 8450
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.1984,
+      "step": 8460
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.196,
+      "step": 8470
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.1954,
+      "step": 8480
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.1941,
+      "step": 8490
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.1949,
+      "step": 8500
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.1975,
+      "step": 8510
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2007,
+      "step": 8520
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2015,
+      "step": 8530
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.1943,
+      "step": 8540
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.203,
+      "step": 8550
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.1982,
+      "step": 8560
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.1948,
+      "step": 8570
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.1997,
+      "step": 8580
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.1941,
+      "step": 8590
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.1965,
+      "step": 8600
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.194,
+      "step": 8610
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.1951,
+      "step": 8620
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.1984,
+      "step": 8630
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.1924,
+      "step": 8640
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.19,
+      "step": 8650
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.1898,
+      "step": 8660
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.1942,
+      "step": 8670
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.189,
+      "step": 8680
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.1978,
+      "step": 8690
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.1945,
+      "step": 8700
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 0.1995,
+      "step": 8710
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 0.1984,
+      "step": 8720
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.1943,
+      "step": 8730
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.1954,
+      "step": 8740
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.1882,
+      "step": 8750
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.1931,
+      "step": 8760
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.1929,
+      "step": 8770
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.1903,
+      "step": 8780
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.193,
+      "step": 8790
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.192,
+      "step": 8800
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.19,
+      "step": 8810
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.194,
+      "step": 8820
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.1953,
+      "step": 8830
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.1914,
+      "step": 8840
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.1873,
+      "step": 8850
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.1902,
+      "step": 8860
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.1924,
+      "step": 8870
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 0.1854,
+      "step": 8880
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 0.1891,
+      "step": 8890
+    },
+    {
+      "epoch": 3.0,
+      "step": 8892,
+      "total_flos": 0.0,
+      "train_loss": 0.35511165388763033,
+      "train_runtime": 61372.9616,
+      "train_samples_per_second": 37.092,
+      "train_steps_per_second": 0.145
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8892,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/dinov2_small_v1_global_nominmax/training_args.bin b/dinov2_small_v1_global_nominmax/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ca4990ad6e2d6dc5c39192293644d1530e9bbcd5
--- /dev/null
+++ b/dinov2_small_v1_global_nominmax/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:898baa792356a9e4502a27fce84c053686ecb4d412bc621b4bb97d064e9b460c
+size 4856