Training in progress, epoch 2, checkpoint

last-checkpoint/global_step624/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8faa3c1a56d2e4b6a2127fa947980cca888393c36509b9151211d17f2f9d6509
+size 16060527788

last-checkpoint/global_step624/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50aad7dfe7d4605443882a0c887da8e0d2c5dfba98b06a3c81e13fd7f8eddc16
+size 16060527788

last-checkpoint/global_step624/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f7c1b3c32ca42fcff708a9eea9c0ae8cf0564a876450edb7c6bfad3bba8499f
+size 16060527788

last-checkpoint/global_step624/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd4528261c2a2f9faa4b1d743038e7735468feca6c64efe97b10fd2f8f39d51e
+size 16060527788

last-checkpoint/global_step624/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b5c5d8a21ab06d56887e86b68b03881ecd6b57c2cb4e5a8a5990099c47ea14b
+size 150693

last-checkpoint/global_step624/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a1fdff8e9100b2b4c12402d7b0dec2b2e5f0825a70159357ee69e460e446e58
+size 150693

last-checkpoint/global_step624/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1aae3f8006170d79cd2384ff9600a67776cdd1627ec59f247567c161efcb0f1b
+size 150693

last-checkpoint/global_step624/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6352ff4375f786bbb4deb191d8efeebb3bfebc1e09751decef3b1b92ddffd155
+size 150693

last-checkpoint/latest

@@ -1 +1 @@
-global_step312
+global_step624

last-checkpoint/model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c8f466e2478b0406023b68ea02d6d3997d7bc9762430b2cbeeac9c01a5e57a46
+oid sha256:dbbd032fa9662e528e5b422ba0c466f38e864fd3738692339c79e0e94af61c91
 size 4976698672

last-checkpoint/model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a82ddd433c542f3145189754603e53591801184672a703f3d6d770fc67c4ceba
+oid sha256:5130f7d7be6def99210631ab039ac4d1dd8e2bfb3b341bbea09774358a48a87d
 size 4999802720

last-checkpoint/model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:226ddfc6b5f35efca888be4141dc8e65de7479e8be595f4e71ef5c25dc6352c2
+oid sha256:102195ce8a84f213e428d6a9cc1a8f78a76b23711a1c6ebbb34f7ca8d53eead5
 size 4915916176

last-checkpoint/model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9832c1be84570095d5a050aa8680f5be583723c1859fe50e6676f4e4e2fc932f
+oid sha256:b1cdebb604c781670825469e1a2fcdc35d6afdf2695801f942744266e9ad0495
 size 1168138808

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f1ce770862b76ceecb3453a5d0b9ef65311eab0f6a60f52997726a89d718de08
+oid sha256:2307c03867cef25b5028feb9a23f80e784b9af9a615de13ddca560a6a90fb593
 size 15024

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2bcae6573442578b752be3e988d4ccb38056b45c31b5c02eb579ec6cebcfa62e
+oid sha256:50218cfaecdd818354e567b7167c13899e3b42297e7d8f58bd7e732cfa547800
 size 15024

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:398ff45072fee2975e88b9e078f915103d75d94bc08753303cf855b915973623
+oid sha256:9814a66b49861f5495b06dae3be12ddf7185b88e2cae1fb808ca9efd99d5807f
 size 15024

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7f4b210bc832d4c648e3bee8c21dca26e5a8b365d6ec90c638062005a052e57b
+oid sha256:a7be93040a65e0a29975f6c70b94418e1fdf88423a50c58aa572141d3c92fbfc
 size 15024

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ff41da86439a5cd2aa8fd5d3c55e2154d1b88c51f3be19137724b98e60eefb4d
+oid sha256:eba46a04fca9d4584d0e7af07b2d004fba441dbf422bdb9977bfac0706fb41fe
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.0,
+  "epoch": 2.0,
   "eval_steps": 500,
-  "global_step": 312,
+  "global_step": 624,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -503,6 +503,487 @@
       "eval_samples_per_second": 13.4,
       "eval_steps_per_second": 0.469,
       "step": 312
+    },
+    {
+      "epoch": 1.0256410256410255,
+      "grad_norm": 16.449029965031666,
+      "learning_rate": 3.6579572446555817e-07,
+      "logits/chosen": -1.59375,
+      "logits/rejected": -1.578125,
+      "logps/chosen": -117.0,
+      "logps/rejected": -195.0,
+      "loss": 0.1531,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": -0.34765625,
+      "rewards/margins": 3.1875,
+      "rewards/rejected": -3.546875,
+      "step": 320
+    },
+    {
+      "epoch": 1.0576923076923077,
+      "grad_norm": 6.562166239087373,
+      "learning_rate": 3.598574821852731e-07,
+      "logits/chosen": -1.59375,
+      "logits/rejected": -1.5703125,
+      "logps/chosen": -111.0,
+      "logps/rejected": -187.0,
+      "loss": 0.1452,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": -0.1923828125,
+      "rewards/margins": 3.40625,
+      "rewards/rejected": -3.609375,
+      "step": 330
+    },
+    {
+      "epoch": 1.0897435897435896,
+      "grad_norm": 25.028970432454248,
+      "learning_rate": 3.5391923990498813e-07,
+      "logits/chosen": -1.6484375,
+      "logits/rejected": -1.609375,
+      "logps/chosen": -134.0,
+      "logps/rejected": -219.0,
+      "loss": 0.1345,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": -0.69921875,
+      "rewards/margins": 3.609375,
+      "rewards/rejected": -4.3125,
+      "step": 340
+    },
+    {
+      "epoch": 1.1217948717948718,
+      "grad_norm": 15.596444813277682,
+      "learning_rate": 3.479809976247031e-07,
+      "logits/chosen": -1.65625,
+      "logits/rejected": -1.6015625,
+      "logps/chosen": -130.0,
+      "logps/rejected": -207.0,
+      "loss": 0.1092,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": -0.00677490234375,
+      "rewards/margins": 3.828125,
+      "rewards/rejected": -3.828125,
+      "step": 350
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 22.37742199079234,
+      "learning_rate": 3.42042755344418e-07,
+      "logits/chosen": -1.5859375,
+      "logits/rejected": -1.5703125,
+      "logps/chosen": -136.0,
+      "logps/rejected": -242.0,
+      "loss": 0.0918,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.640625,
+      "rewards/margins": 4.4375,
+      "rewards/rejected": -5.09375,
+      "step": 360
+    },
+    {
+      "epoch": 1.185897435897436,
+      "grad_norm": 27.25128799333949,
+      "learning_rate": 3.36104513064133e-07,
+      "logits/chosen": -1.609375,
+      "logits/rejected": -1.59375,
+      "logps/chosen": -154.0,
+      "logps/rejected": -221.0,
+      "loss": 0.1215,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.333984375,
+      "rewards/margins": 4.46875,
+      "rewards/rejected": -4.78125,
+      "step": 370
+    },
+    {
+      "epoch": 1.217948717948718,
+      "grad_norm": 16.802650370434502,
+      "learning_rate": 3.3016627078384796e-07,
+      "logits/chosen": -1.625,
+      "logits/rejected": -1.515625,
+      "logps/chosen": -121.0,
+      "logps/rejected": -184.0,
+      "loss": 0.1051,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": -0.75390625,
+      "rewards/margins": 3.75,
+      "rewards/rejected": -4.5,
+      "step": 380
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.14599606798184,
+      "learning_rate": 3.2422802850356297e-07,
+      "logits/chosen": -1.65625,
+      "logits/rejected": -1.5625,
+      "logps/chosen": -118.0,
+      "logps/rejected": -203.0,
+      "loss": 0.1691,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": -0.4609375,
+      "rewards/margins": 4.4375,
+      "rewards/rejected": -4.875,
+      "step": 390
+    },
+    {
+      "epoch": 1.282051282051282,
+      "grad_norm": 14.341477798166297,
+      "learning_rate": 3.182897862232779e-07,
+      "logits/chosen": -1.609375,
+      "logits/rejected": -1.5546875,
+      "logps/chosen": -112.0,
+      "logps/rejected": -202.0,
+      "loss": 0.1084,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": -0.2578125,
+      "rewards/margins": 4.65625,
+      "rewards/rejected": -4.90625,
+      "step": 400
+    },
+    {
+      "epoch": 1.314102564102564,
+      "grad_norm": 22.538352683910524,
+      "learning_rate": 3.1235154394299283e-07,
+      "logits/chosen": -1.671875,
+      "logits/rejected": -1.625,
+      "logps/chosen": -168.0,
+      "logps/rejected": -270.0,
+      "loss": 0.136,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -1.1796875,
+      "rewards/margins": 4.71875,
+      "rewards/rejected": -5.90625,
+      "step": 410
+    },
+    {
+      "epoch": 1.3461538461538463,
+      "grad_norm": 7.69063972210861,
+      "learning_rate": 3.0641330166270784e-07,
+      "logits/chosen": -1.609375,
+      "logits/rejected": -1.59375,
+      "logps/chosen": -128.0,
+      "logps/rejected": -232.0,
+      "loss": 0.1123,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -1.0859375,
+      "rewards/margins": 4.59375,
+      "rewards/rejected": -5.6875,
+      "step": 420
+    },
+    {
+      "epoch": 1.3782051282051282,
+      "grad_norm": 22.81215314331503,
+      "learning_rate": 3.004750593824228e-07,
+      "logits/chosen": -1.6328125,
+      "logits/rejected": -1.5390625,
+      "logps/chosen": -122.5,
+      "logps/rejected": -209.0,
+      "loss": 0.0971,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.33984375,
+      "rewards/margins": 4.65625,
+      "rewards/rejected": -5.0,
+      "step": 430
+    },
+    {
+      "epoch": 1.4102564102564101,
+      "grad_norm": 16.88060293661355,
+      "learning_rate": 2.9453681710213776e-07,
+      "logits/chosen": -1.5859375,
+      "logits/rejected": -1.4765625,
+      "logps/chosen": -123.0,
+      "logps/rejected": -205.0,
+      "loss": 0.11,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.4296875,
+      "rewards/margins": 4.5625,
+      "rewards/rejected": -5.0,
+      "step": 440
+    },
+    {
+      "epoch": 1.4423076923076923,
+      "grad_norm": 14.494261083829612,
+      "learning_rate": 2.885985748218527e-07,
+      "logits/chosen": -1.546875,
+      "logits/rejected": -1.5546875,
+      "logps/chosen": -118.5,
+      "logps/rejected": -221.0,
+      "loss": 0.0896,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -1.0859375,
+      "rewards/margins": 4.59375,
+      "rewards/rejected": -5.6875,
+      "step": 450
+    },
+    {
+      "epoch": 1.4743589743589745,
+      "grad_norm": 16.655718879118293,
+      "learning_rate": 2.8266033254156767e-07,
+      "logits/chosen": -1.59375,
+      "logits/rejected": -1.5859375,
+      "logps/chosen": -142.0,
+      "logps/rejected": -256.0,
+      "loss": 0.1071,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -1.25,
+      "rewards/margins": 4.84375,
+      "rewards/rejected": -6.09375,
+      "step": 460
+    },
+    {
+      "epoch": 1.5064102564102564,
+      "grad_norm": 20.51874312804957,
+      "learning_rate": 2.7672209026128263e-07,
+      "logits/chosen": -1.5625,
+      "logits/rejected": -1.53125,
+      "logps/chosen": -132.0,
+      "logps/rejected": -231.0,
+      "loss": 0.1012,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -1.015625,
+      "rewards/margins": 4.65625,
+      "rewards/rejected": -5.6875,
+      "step": 470
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 28.03019085733964,
+      "learning_rate": 2.7078384798099764e-07,
+      "logits/chosen": -1.4453125,
+      "logits/rejected": -1.4609375,
+      "logps/chosen": -104.5,
+      "logps/rejected": -203.0,
+      "loss": 0.1074,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": -1.6796875,
+      "rewards/margins": 4.59375,
+      "rewards/rejected": -6.3125,
+      "step": 480
+    },
+    {
+      "epoch": 1.5705128205128205,
+      "grad_norm": 14.674518317046845,
+      "learning_rate": 2.648456057007126e-07,
+      "logits/chosen": -1.5625,
+      "logits/rejected": -1.5390625,
+      "logps/chosen": -139.0,
+      "logps/rejected": -233.0,
+      "loss": 0.1003,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": -1.3125,
+      "rewards/margins": 4.9375,
+      "rewards/rejected": -6.28125,
+      "step": 490
+    },
+    {
+      "epoch": 1.6025641025641026,
+      "grad_norm": 18.731133870495857,
+      "learning_rate": 2.589073634204275e-07,
+      "logits/chosen": -1.5703125,
+      "logits/rejected": -1.5625,
+      "logps/chosen": -144.0,
+      "logps/rejected": -251.0,
+      "loss": 0.1228,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": -1.109375,
+      "rewards/margins": 5.25,
+      "rewards/rejected": -6.375,
+      "step": 500
+    },
+    {
+      "epoch": 1.6346153846153846,
+      "grad_norm": 16.81631331394118,
+      "learning_rate": 2.529691211401425e-07,
+      "logits/chosen": -1.5625,
+      "logits/rejected": -1.5,
+      "logps/chosen": -131.0,
+      "logps/rejected": -230.0,
+      "loss": 0.0819,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": -0.8359375,
+      "rewards/margins": 5.375,
+      "rewards/rejected": -6.21875,
+      "step": 510
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 15.111001669085608,
+      "learning_rate": 2.4703087885985747e-07,
+      "logits/chosen": -1.546875,
+      "logits/rejected": -1.4921875,
+      "logps/chosen": -114.0,
+      "logps/rejected": -202.0,
+      "loss": 0.1091,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.87109375,
+      "rewards/margins": 4.78125,
+      "rewards/rejected": -5.625,
+      "step": 520
+    },
+    {
+      "epoch": 1.6987179487179487,
+      "grad_norm": 26.75975769912583,
+      "learning_rate": 2.410926365795724e-07,
+      "logits/chosen": -1.5625,
+      "logits/rejected": -1.515625,
+      "logps/chosen": -140.0,
+      "logps/rejected": -221.0,
+      "loss": 0.1139,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -1.0625,
+      "rewards/margins": 4.8125,
+      "rewards/rejected": -5.875,
+      "step": 530
+    },
+    {
+      "epoch": 1.7307692307692308,
+      "grad_norm": 13.759559493772942,
+      "learning_rate": 2.351543942992874e-07,
+      "logits/chosen": -1.578125,
+      "logits/rejected": -1.5,
+      "logps/chosen": -143.0,
+      "logps/rejected": -229.0,
+      "loss": 0.0978,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.9765625,
+      "rewards/margins": 5.09375,
+      "rewards/rejected": -6.0625,
+      "step": 540
+    },
+    {
+      "epoch": 1.7628205128205128,
+      "grad_norm": 18.98900148314797,
+      "learning_rate": 2.2921615201900234e-07,
+      "logits/chosen": -1.4921875,
+      "logits/rejected": -1.5,
+      "logps/chosen": -114.0,
+      "logps/rejected": -232.0,
+      "loss": 0.1103,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": -0.6953125,
+      "rewards/margins": 5.03125,
+      "rewards/rejected": -5.71875,
+      "step": 550
+    },
+    {
+      "epoch": 1.7948717948717947,
+      "grad_norm": 7.478935072745169,
+      "learning_rate": 2.2327790973871732e-07,
+      "logits/chosen": -1.515625,
+      "logits/rejected": -1.4375,
+      "logps/chosen": -112.0,
+      "logps/rejected": -177.0,
+      "loss": 0.1139,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": -0.6640625,
+      "rewards/margins": 4.5625,
+      "rewards/rejected": -5.21875,
+      "step": 560
+    },
+    {
+      "epoch": 1.8269230769230769,
+      "grad_norm": 13.728806584543696,
+      "learning_rate": 2.173396674584323e-07,
+      "logits/chosen": -1.546875,
+      "logits/rejected": -1.5546875,
+      "logps/chosen": -130.0,
+      "logps/rejected": -239.0,
+      "loss": 0.1355,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": -0.466796875,
+      "rewards/margins": 4.875,
+      "rewards/rejected": -5.34375,
+      "step": 570
+    },
+    {
+      "epoch": 1.858974358974359,
+      "grad_norm": 9.926742756220793,
+      "learning_rate": 2.1140142517814726e-07,
+      "logits/chosen": -1.609375,
+      "logits/rejected": -1.5078125,
+      "logps/chosen": -158.0,
+      "logps/rejected": -202.0,
+      "loss": 0.1118,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": -1.09375,
+      "rewards/margins": 4.21875,
+      "rewards/rejected": -5.3125,
+      "step": 580
+    },
+    {
+      "epoch": 1.891025641025641,
+      "grad_norm": 12.431864869622972,
+      "learning_rate": 2.0546318289786222e-07,
+      "logits/chosen": -1.5546875,
+      "logits/rejected": -1.53125,
+      "logps/chosen": -122.0,
+      "logps/rejected": -202.0,
+      "loss": 0.1325,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": -0.61328125,
+      "rewards/margins": 4.65625,
+      "rewards/rejected": -5.28125,
+      "step": 590
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 12.291656293105886,
+      "learning_rate": 1.9952494061757718e-07,
+      "logits/chosen": -1.5703125,
+      "logits/rejected": -1.5625,
+      "logps/chosen": -123.0,
+      "logps/rejected": -227.0,
+      "loss": 0.1085,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": -0.67578125,
+      "rewards/margins": 4.8125,
+      "rewards/rejected": -5.5,
+      "step": 600
+    },
+    {
+      "epoch": 1.9551282051282053,
+      "grad_norm": 6.534649119157792,
+      "learning_rate": 1.9358669833729216e-07,
+      "logits/chosen": -1.5546875,
+      "logits/rejected": -1.484375,
+      "logps/chosen": -122.0,
+      "logps/rejected": -202.0,
+      "loss": 0.0935,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": -0.56640625,
+      "rewards/margins": 5.0,
+      "rewards/rejected": -5.59375,
+      "step": 610
+    },
+    {
+      "epoch": 1.9871794871794872,
+      "grad_norm": 10.91488724265466,
+      "learning_rate": 1.876484560570071e-07,
+      "logits/chosen": -1.5703125,
+      "logits/rejected": -1.515625,
+      "logps/chosen": -139.0,
+      "logps/rejected": -236.0,
+      "loss": 0.1071,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": -1.1953125,
+      "rewards/margins": 5.34375,
+      "rewards/rejected": -6.53125,
+      "step": 620
+    },
+    {
+      "epoch": 2.0,
+      "eval_logits/chosen": -1.5,
+      "eval_logits/rejected": -1.46875,
+      "eval_logps/chosen": -188.0,
+      "eval_logps/rejected": -206.0,
+      "eval_loss": 0.700976550579071,
+      "eval_rewards/accuracies": 0.6428571343421936,
+      "eval_rewards/chosen": -3.296875,
+      "eval_rewards/margins": 1.3046875,
+      "eval_rewards/rejected": -4.59375,
+      "eval_runtime": 15.0135,
+      "eval_samples_per_second": 13.321,
+      "eval_steps_per_second": 0.466,
+      "step": 624
     }
   ],
   "logging_steps": 10,