diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7219 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.0473597903539947,
+  "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-06,
+      "loss": 31.3285,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 28.7725,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5e-05,
+      "loss": 25.4191,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 21.4758,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5e-05,
+      "loss": 16.1213,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-05,
+      "loss": 11.1718,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5e-05,
+      "loss": 8.9586,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4e-05,
+      "loss": 8.4075,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.5e-05,
+      "loss": 8.2418,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 8.0737,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 7.9055,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6e-05,
+      "loss": 7.7686,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 7.6632,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7e-05,
+      "loss": 7.4765,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 7.3471,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8e-05,
+      "loss": 7.1586,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.5e-05,
+      "loss": 6.8925,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9e-05,
+      "loss": 6.752,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.5e-05,
+      "loss": 6.4513,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 6.2755,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999999961506464e-05,
+      "loss": 6.1679,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999999846025851e-05,
+      "loss": 6.0127,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999999653558166e-05,
+      "loss": 5.936,
+      "step": 115
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999999384103412e-05,
+      "loss": 5.7234,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999999037661592e-05,
+      "loss": 5.4162,
+      "step": 125
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999998614232713e-05,
+      "loss": 5.4169,
+      "step": 130
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99999811381678e-05,
+      "loss": 5.3181,
+      "step": 135
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.9999975364138e-05,
+      "loss": 5.2402,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999996882023784e-05,
+      "loss": 4.9345,
+      "step": 145
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99999615064674e-05,
+      "loss": 4.924,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99999534228268e-05,
+      "loss": 4.7958,
+      "step": 155
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999994456931619e-05,
+      "loss": 4.7021,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999993494593566e-05,
+      "loss": 4.6688,
+      "step": 165
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999992455268539e-05,
+      "loss": 4.4738,
+      "step": 170
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999991338956554e-05,
+      "loss": 4.4177,
+      "step": 175
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999990145657626e-05,
+      "loss": 4.3629,
+      "step": 180
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999988875371776e-05,
+      "loss": 4.2112,
+      "step": 185
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999987528099021e-05,
+      "loss": 4.1168,
+      "step": 190
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999986103839384e-05,
+      "loss": 4.091,
+      "step": 195
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999984602592884e-05,
+      "loss": 4.014,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999983024359549e-05,
+      "loss": 4.0023,
+      "step": 205
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.9999813691394e-05,
+      "loss": 3.9321,
+      "step": 210
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999979636932461e-05,
+      "loss": 3.785,
+      "step": 215
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999977827738762e-05,
+      "loss": 3.7226,
+      "step": 220
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999975941558328e-05,
+      "loss": 3.6735,
+      "step": 225
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999973978391191e-05,
+      "loss": 3.7661,
+      "step": 230
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99997193823738e-05,
+      "loss": 3.6616,
+      "step": 235
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999969821096924e-05,
+      "loss": 3.507,
+      "step": 240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99996762696986e-05,
+      "loss": 3.6003,
+      "step": 245
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999965355856218e-05,
+      "loss": 3.5171,
+      "step": 250
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999963007756035e-05,
+      "loss": 3.459,
+      "step": 255
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999960582669346e-05,
+      "loss": 3.5772,
+      "step": 260
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99995808059619e-05,
+      "loss": 3.4475,
+      "step": 265
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999955501536603e-05,
+      "loss": 3.2552,
+      "step": 270
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999952845490626e-05,
+      "loss": 3.3913,
+      "step": 275
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999950112458302e-05,
+      "loss": 3.3408,
+      "step": 280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99994730243967e-05,
+      "loss": 3.322,
+      "step": 285
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999944415434774e-05,
+      "loss": 3.3111,
+      "step": 290
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99994145144366e-05,
+      "loss": 3.3261,
+      "step": 295
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999938410466371e-05,
+      "loss": 3.3336,
+      "step": 300
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999935292502958e-05,
+      "loss": 3.328,
+      "step": 305
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999932097553463e-05,
+      "loss": 3.1969,
+      "step": 310
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999928825617941e-05,
+      "loss": 3.1167,
+      "step": 315
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999925476696439e-05,
+      "loss": 3.0775,
+      "step": 320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99992205078901e-05,
+      "loss": 3.1141,
+      "step": 325
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999918547895705e-05,
+      "loss": 3.1053,
+      "step": 330
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999914968016579e-05,
+      "loss": 3.1544,
+      "step": 335
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999911311151689e-05,
+      "loss": 3.1204,
+      "step": 340
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999907577301089e-05,
+      "loss": 3.03,
+      "step": 345
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999903766464837e-05,
+      "loss": 3.0965,
+      "step": 350
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99989987864299e-05,
+      "loss": 3.0011,
+      "step": 355
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999895913835613e-05,
+      "loss": 3.0307,
+      "step": 360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999891872042762e-05,
+      "loss": 3.1558,
+      "step": 365
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.9998877532645e-05,
+      "loss": 3.0856,
+      "step": 370
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999883557500892e-05,
+      "loss": 3.1212,
+      "step": 375
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999879284752003e-05,
+      "loss": 3.0439,
+      "step": 380
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999874935017897e-05,
+      "loss": 3.0008,
+      "step": 385
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999870508298641e-05,
+      "loss": 2.8993,
+      "step": 390
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999866004594303e-05,
+      "loss": 2.9422,
+      "step": 395
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999861423904954e-05,
+      "loss": 2.9066,
+      "step": 400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999856766230665e-05,
+      "loss": 3.0064,
+      "step": 405
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999852031571506e-05,
+      "loss": 3.0118,
+      "step": 410
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999847219927549e-05,
+      "loss": 3.0655,
+      "step": 415
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999842331298872e-05,
+      "loss": 2.8881,
+      "step": 420
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999837365685544e-05,
+      "loss": 2.9323,
+      "step": 425
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999832323087649e-05,
+      "loss": 2.9559,
+      "step": 430
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999827203505259e-05,
+      "loss": 2.9535,
+      "step": 435
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999822006938455e-05,
+      "loss": 2.8567,
+      "step": 440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999816733387316e-05,
+      "loss": 3.0198,
+      "step": 445
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999811382851924e-05,
+      "loss": 2.8562,
+      "step": 450
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999805955332362e-05,
+      "loss": 2.9086,
+      "step": 455
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999800450828711e-05,
+      "loss": 2.8259,
+      "step": 460
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99979486934106e-05,
+      "loss": 2.8738,
+      "step": 465
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99978921086949e-05,
+      "loss": 2.7782,
+      "step": 470
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999783475414093e-05,
+      "loss": 3.011,
+      "step": 475
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999777662974952e-05,
+      "loss": 2.8474,
+      "step": 480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999771773552162e-05,
+      "loss": 2.7734,
+      "step": 485
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999765807145811e-05,
+      "loss": 2.9079,
+      "step": 490
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99975976375599e-05,
+      "loss": 2.872,
+      "step": 495
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999753643382793e-05,
+      "loss": 2.8606,
+      "step": 500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999747446026315e-05,
+      "loss": 2.8687,
+      "step": 505
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999741171686649e-05,
+      "loss": 2.7748,
+      "step": 510
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999734820363895e-05,
+      "loss": 2.9025,
+      "step": 515
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99972839205815e-05,
+      "loss": 2.828,
+      "step": 520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999721886769511e-05,
+      "loss": 2.8609,
+      "step": 525
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999715304498078e-05,
+      "loss": 2.8119,
+      "step": 530
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999708645243955e-05,
+      "loss": 2.8324,
+      "step": 535
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999701909007242e-05,
+      "loss": 2.8067,
+      "step": 540
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999695095788044e-05,
+      "loss": 2.8003,
+      "step": 545
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999688205586468e-05,
+      "loss": 2.7513,
+      "step": 550
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999681238402615e-05,
+      "loss": 2.7573,
+      "step": 555
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999674194236597e-05,
+      "loss": 2.7483,
+      "step": 560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999667073088519e-05,
+      "loss": 2.8177,
+      "step": 565
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999659874958493e-05,
+      "loss": 2.741,
+      "step": 570
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.99965259984663e-05,
+      "loss": 2.7904,
+      "step": 575
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999645247753039e-05,
+      "loss": 2.7096,
+      "step": 580
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999637818677836e-05,
+      "loss": 2.6989,
+      "step": 585
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999630312621136e-05,
+      "loss": 2.7366,
+      "step": 590
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999622729583051e-05,
+      "loss": 2.8125,
+      "step": 595
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999615069563701e-05,
+      "loss": 2.7226,
+      "step": 600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999607332563205e-05,
+      "loss": 2.716,
+      "step": 605
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999599518581676e-05,
+      "loss": 2.7116,
+      "step": 610
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999591627619242e-05,
+      "loss": 2.7992,
+      "step": 615
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999583659676019e-05,
+      "loss": 2.7818,
+      "step": 620
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999575614752133e-05,
+      "loss": 2.7484,
+      "step": 625
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999567492847706e-05,
+      "loss": 2.6899,
+      "step": 630
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999559293962864e-05,
+      "loss": 2.7821,
+      "step": 635
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99955101809773e-05,
+      "loss": 2.6708,
+      "step": 640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999542665252437e-05,
+      "loss": 2.764,
+      "step": 645
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999534235427109e-05,
+      "loss": 2.64,
+      "step": 650
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999525728621879e-05,
+      "loss": 2.7156,
+      "step": 655
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999517144836876e-05,
+      "loss": 2.7384,
+      "step": 660
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999508484072233e-05,
+      "loss": 2.7237,
+      "step": 665
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999499746328083e-05,
+      "loss": 2.732,
+      "step": 670
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99949093160456e-05,
+      "loss": 2.6818,
+      "step": 675
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.9994820399018e-05,
+      "loss": 2.693,
+      "step": 680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999473071219941e-05,
+      "loss": 2.6411,
+      "step": 685
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999464025559121e-05,
+      "loss": 2.708,
+      "step": 690
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999454902919478e-05,
+      "loss": 2.6651,
+      "step": 695
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999445703301154e-05,
+      "loss": 2.6026,
+      "step": 700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999436426704287e-05,
+      "loss": 2.6089,
+      "step": 705
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999427073129026e-05,
+      "loss": 2.6368,
+      "step": 710
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999417642575509e-05,
+      "loss": 2.6682,
+      "step": 715
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999408135043886e-05,
+      "loss": 2.6859,
+      "step": 720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999398550534298e-05,
+      "loss": 2.5584,
+      "step": 725
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999388889046899e-05,
+      "loss": 2.6617,
+      "step": 730
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999379150581833e-05,
+      "loss": 2.6157,
+      "step": 735
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999369335139251e-05,
+      "loss": 2.7022,
+      "step": 740
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999359442719306e-05,
+      "loss": 2.7542,
+      "step": 745
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999349473322149e-05,
+      "loss": 2.7249,
+      "step": 750
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999339426947932e-05,
+      "loss": 2.6759,
+      "step": 755
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999329303596812e-05,
+      "loss": 2.5962,
+      "step": 760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999319103268945e-05,
+      "loss": 2.6601,
+      "step": 765
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999308825964485e-05,
+      "loss": 2.5728,
+      "step": 770
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999298471683593e-05,
+      "loss": 2.7093,
+      "step": 775
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999288040426429e-05,
+      "loss": 2.6036,
+      "step": 780
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999277532193151e-05,
+      "loss": 2.643,
+      "step": 785
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999266946983923e-05,
+      "loss": 2.6513,
+      "step": 790
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999256284798906e-05,
+      "loss": 2.7043,
+      "step": 795
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999245545638267e-05,
+      "loss": 2.6427,
+      "step": 800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999234729502167e-05,
+      "loss": 2.5795,
+      "step": 805
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999223836390777e-05,
+      "loss": 2.604,
+      "step": 810
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999212866304262e-05,
+      "loss": 2.6267,
+      "step": 815
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999201819242794e-05,
+      "loss": 2.6257,
+      "step": 820
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99919069520654e-05,
+      "loss": 2.5922,
+      "step": 825
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99917949419567e-05,
+      "loss": 2.6699,
+      "step": 830
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999168216210362e-05,
+      "loss": 2.617,
+      "step": 835
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999156861250784e-05,
+      "loss": 2.5489,
+      "step": 840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999145429317114e-05,
+      "loss": 2.5703,
+      "step": 845
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999133920409527e-05,
+      "loss": 2.637,
+      "step": 850
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999122334528199e-05,
+      "loss": 2.5256,
+      "step": 855
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999110671673313e-05,
+      "loss": 2.6027,
+      "step": 860
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999098931845044e-05,
+      "loss": 2.5248,
+      "step": 865
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999087115043573e-05,
+      "loss": 2.5238,
+      "step": 870
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999075221269083e-05,
+      "loss": 2.5997,
+      "step": 875
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999063250521759e-05,
+      "loss": 2.5637,
+      "step": 880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999051202801781e-05,
+      "loss": 2.6535,
+      "step": 885
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999039078109341e-05,
+      "loss": 2.5874,
+      "step": 890
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999026876444617e-05,
+      "loss": 2.608,
+      "step": 895
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999014597807805e-05,
+      "loss": 2.5475,
+      "step": 900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99900224219909e-05,
+      "loss": 2.5285,
+      "step": 905
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998989809618662e-05,
+      "loss": 2.597,
+      "step": 910
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998977300066715e-05,
+      "loss": 2.6007,
+      "step": 915
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998964713543437e-05,
+      "loss": 2.5219,
+      "step": 920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998952050049026e-05,
+      "loss": 2.448,
+      "step": 925
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998939309583677e-05,
+      "loss": 2.6175,
+      "step": 930
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998926492147585e-05,
+      "loss": 2.4657,
+      "step": 935
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998913597740945e-05,
+      "loss": 2.5963,
+      "step": 940
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99890062636396e-05,
+      "loss": 2.5266,
+      "step": 945
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998887578016827e-05,
+      "loss": 2.5464,
+      "step": 950
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998874452699747e-05,
+      "loss": 2.5791,
+      "step": 955
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998861250412922e-05,
+      "loss": 2.5929,
+      "step": 960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998847971156556e-05,
+      "loss": 2.6085,
+      "step": 965
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998834614930855e-05,
+      "loss": 2.606,
+      "step": 970
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99882118173602e-05,
+      "loss": 2.5438,
+      "step": 975
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998807671572263e-05,
+      "loss": 2.5037,
+      "step": 980
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998794084439789e-05,
+      "loss": 2.5326,
+      "step": 985
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998780420338809e-05,
+      "loss": 2.526,
+      "step": 990
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998766679269531e-05,
+      "loss": 2.5609,
+      "step": 995
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998752861232168e-05,
+      "loss": 2.5791,
+      "step": 1000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998738966226933e-05,
+      "loss": 2.5382,
+      "step": 1005
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998724994254041e-05,
+      "loss": 2.575,
+      "step": 1010
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998710945313705e-05,
+      "loss": 2.5069,
+      "step": 1015
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998696819406141e-05,
+      "loss": 2.5852,
+      "step": 1020
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99868261653157e-05,
+      "loss": 2.4902,
+      "step": 1025
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998668336690206e-05,
+      "loss": 2.5166,
+      "step": 1030
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998653979882271e-05,
+      "loss": 2.5311,
+      "step": 1035
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998639546107989e-05,
+      "loss": 2.5292,
+      "step": 1040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998625035367577e-05,
+      "loss": 2.5739,
+      "step": 1045
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998610447661261e-05,
+      "loss": 2.5822,
+      "step": 1050
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998595782989266e-05,
+      "loss": 2.5616,
+      "step": 1055
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998581041351818e-05,
+      "loss": 2.446,
+      "step": 1060
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998566222749142e-05,
+      "loss": 2.5766,
+      "step": 1065
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998551327181468e-05,
+      "loss": 2.5128,
+      "step": 1070
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998536354649026e-05,
+      "loss": 2.5724,
+      "step": 1075
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998521305152045e-05,
+      "loss": 2.527,
+      "step": 1080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998506178690755e-05,
+      "loss": 2.577,
+      "step": 1085
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998490975265393e-05,
+      "loss": 2.5156,
+      "step": 1090
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99847569487619e-05,
+      "loss": 2.5398,
+      "step": 1095
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998460337523383e-05,
+      "loss": 2.5207,
+      "step": 1100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998444903207207e-05,
+      "loss": 2.4793,
+      "step": 1105
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998429391927902e-05,
+      "loss": 2.5408,
+      "step": 1110
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998413803685705e-05,
+      "loss": 2.5424,
+      "step": 1115
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998398138480855e-05,
+      "loss": 2.4463,
+      "step": 1120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998382396313597e-05,
+      "loss": 2.5269,
+      "step": 1125
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998366577184168e-05,
+      "loss": 2.5562,
+      "step": 1130
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998350681092817e-05,
+      "loss": 2.5724,
+      "step": 1135
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998334708039785e-05,
+      "loss": 2.4432,
+      "step": 1140
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998318658025318e-05,
+      "loss": 2.5919,
+      "step": 1145
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998302531049667e-05,
+      "loss": 2.5866,
+      "step": 1150
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998286327113076e-05,
+      "loss": 2.5586,
+      "step": 1155
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998270046215797e-05,
+      "loss": 2.5257,
+      "step": 1160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998253688358078e-05,
+      "loss": 2.5186,
+      "step": 1165
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998237253540174e-05,
+      "loss": 2.4606,
+      "step": 1170
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998220741762338e-05,
+      "loss": 2.3913,
+      "step": 1175
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998204153024821e-05,
+      "loss": 2.5368,
+      "step": 1180
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998187487327882e-05,
+      "loss": 2.501,
+      "step": 1185
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998170744671775e-05,
+      "loss": 2.4491,
+      "step": 1190
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998153925056758e-05,
+      "loss": 2.552,
+      "step": 1195
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998137028483092e-05,
+      "loss": 2.4892,
+      "step": 1200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998120054951035e-05,
+      "loss": 2.5213,
+      "step": 1205
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99810300446085e-05,
+      "loss": 2.4682,
+      "step": 1210
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.9980858770128e-05,
+      "loss": 2.5004,
+      "step": 1215
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998068672607146e-05,
+      "loss": 2.495,
+      "step": 1220
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998051391244155e-05,
+      "loss": 2.4078,
+      "step": 1225
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998034032924092e-05,
+      "loss": 2.5424,
+      "step": 1230
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.998016597647225e-05,
+      "loss": 2.5211,
+      "step": 1235
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997999085413823e-05,
+      "loss": 2.5175,
+      "step": 1240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997981496224153e-05,
+      "loss": 2.45,
+      "step": 1245
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99796383007849e-05,
+      "loss": 2.5494,
+      "step": 1250
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997946086977103e-05,
+      "loss": 2.4066,
+      "step": 1255
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997928266920266e-05,
+      "loss": 2.544,
+      "step": 1260
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997910369908254e-05,
+      "loss": 2.4463,
+      "step": 1265
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99789239594134e-05,
+      "loss": 2.5495,
+      "step": 1270
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997874345019805e-05,
+      "loss": 2.4245,
+      "step": 1275
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997856217143921e-05,
+      "loss": 2.5031,
+      "step": 1280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997838012313974e-05,
+      "loss": 2.4639,
+      "step": 1285
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99781973053024e-05,
+      "loss": 2.503,
+      "step": 1290
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997801371793001e-05,
+      "loss": 2.4389,
+      "step": 1295
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997782936102542e-05,
+      "loss": 2.4613,
+      "step": 1300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997764423459144e-05,
+      "loss": 2.4497,
+      "step": 1305
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997745833863092e-05,
+      "loss": 2.4507,
+      "step": 1310
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997727167314676e-05,
+      "loss": 2.4956,
+      "step": 1315
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997708423814178e-05,
+      "loss": 2.4657,
+      "step": 1320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99768960336189e-05,
+      "loss": 2.473,
+      "step": 1325
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997670705958102e-05,
+      "loss": 2.3665,
+      "step": 1330
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997651731603104e-05,
+      "loss": 2.5718,
+      "step": 1335
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997632680297188e-05,
+      "loss": 2.487,
+      "step": 1340
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997613552040648e-05,
+      "loss": 2.394,
+      "step": 1345
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997594346833779e-05,
+      "loss": 2.4425,
+      "step": 1350
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997575064676875e-05,
+      "loss": 2.4625,
+      "step": 1355
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997555705570232e-05,
+      "loss": 2.523,
+      "step": 1360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997536269514152e-05,
+      "loss": 2.494,
+      "step": 1365
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99751675650893e-05,
+      "loss": 2.5851,
+      "step": 1370
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99749716655487e-05,
+      "loss": 2.462,
+      "step": 1375
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997477499652273e-05,
+      "loss": 2.5842,
+      "step": 1380
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99745775580144e-05,
+      "loss": 2.4579,
+      "step": 1385
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997437935002676e-05,
+      "loss": 2.3896,
+      "step": 1390
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997418037256285e-05,
+      "loss": 2.2939,
+      "step": 1395
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997398062562575e-05,
+      "loss": 2.4816,
+      "step": 1400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997378010921856e-05,
+      "loss": 2.4322,
+      "step": 1405
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99735788233443e-05,
+      "loss": 2.5676,
+      "step": 1410
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997337676800613e-05,
+      "loss": 2.4026,
+      "step": 1415
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997317394320713e-05,
+      "loss": 2.4415,
+      "step": 1420
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997297034895044e-05,
+      "loss": 2.4965,
+      "step": 1425
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99727659852392e-05,
+      "loss": 2.4705,
+      "step": 1430
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997256085207652e-05,
+      "loss": 2.5072,
+      "step": 1435
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997235494946557e-05,
+      "loss": 2.4418,
+      "step": 1440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997214827740955e-05,
+      "loss": 2.4522,
+      "step": 1445
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997194083591164e-05,
+      "loss": 2.4988,
+      "step": 1450
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.9971732624975e-05,
+      "loss": 2.435,
+      "step": 1455
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997152364460287e-05,
+      "loss": 2.4739,
+      "step": 1460
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997131389479844e-05,
+      "loss": 2.4496,
+      "step": 1465
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997110337556494e-05,
+      "loss": 2.3305,
+      "step": 1470
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997089208690564e-05,
+      "loss": 2.4344,
+      "step": 1475
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997068002882377e-05,
+      "loss": 2.3736,
+      "step": 1480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997046720132261e-05,
+      "loss": 2.5,
+      "step": 1485
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997025360440541e-05,
+      "loss": 2.3823,
+      "step": 1490
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.997003923807551e-05,
+      "loss": 2.5366,
+      "step": 1495
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996982410233615e-05,
+      "loss": 2.3723,
+      "step": 1500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996960819719068e-05,
+      "loss": 2.5048,
+      "step": 1505
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996939152264242e-05,
+      "loss": 2.4097,
+      "step": 1510
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99691740786947e-05,
+      "loss": 2.4082,
+      "step": 1515
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996895586535086e-05,
+      "loss": 2.4948,
+      "step": 1520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996873688261428e-05,
+      "loss": 2.4843,
+      "step": 1525
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996851713048831e-05,
+      "loss": 2.4844,
+      "step": 1530
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996829660897637e-05,
+      "loss": 2.4169,
+      "step": 1535
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996807531808181e-05,
+      "loss": 2.4401,
+      "step": 1540
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996785325780806e-05,
+      "loss": 2.4023,
+      "step": 1545
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996763042815853e-05,
+      "loss": 2.4986,
+      "step": 1550
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996740682913668e-05,
+      "loss": 2.4329,
+      "step": 1555
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996718246074591e-05,
+      "loss": 2.4427,
+      "step": 1560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99669573229897e-05,
+      "loss": 2.4527,
+      "step": 1565
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996673141587151e-05,
+      "loss": 2.3753,
+      "step": 1570
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996650473939482e-05,
+      "loss": 2.3921,
+      "step": 1575
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996627729356314e-05,
+      "loss": 2.4195,
+      "step": 1580
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996604907837993e-05,
+      "loss": 2.4184,
+      "step": 1585
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996582009384874e-05,
+      "loss": 2.4636,
+      "step": 1590
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996559033997307e-05,
+      "loss": 2.3548,
+      "step": 1595
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996535981675648e-05,
+      "loss": 2.4444,
+      "step": 1600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996512852420251e-05,
+      "loss": 2.4453,
+      "step": 1605
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996489646231471e-05,
+      "loss": 2.4925,
+      "step": 1610
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996466363109667e-05,
+      "loss": 2.4515,
+      "step": 1615
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996443003055196e-05,
+      "loss": 2.4808,
+      "step": 1620
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99641956606842e-05,
+      "loss": 2.5005,
+      "step": 1625
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996396052149698e-05,
+      "loss": 2.4451,
+      "step": 1630
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996372461299392e-05,
+      "loss": 2.481,
+      "step": 1635
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996348793517866e-05,
+      "loss": 2.4231,
+      "step": 1640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996325048805484e-05,
+      "loss": 2.4542,
+      "step": 1645
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996301227162612e-05,
+      "loss": 2.425,
+      "step": 1650
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996277328589618e-05,
+      "loss": 2.4129,
+      "step": 1655
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996253353086866e-05,
+      "loss": 2.4454,
+      "step": 1660
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996229300654729e-05,
+      "loss": 2.3946,
+      "step": 1665
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996205171293576e-05,
+      "loss": 2.3503,
+      "step": 1670
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996180965003778e-05,
+      "loss": 2.4151,
+      "step": 1675
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996156681785708e-05,
+      "loss": 2.5114,
+      "step": 1680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996132321639741e-05,
+      "loss": 2.3828,
+      "step": 1685
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996107884566252e-05,
+      "loss": 2.3487,
+      "step": 1690
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996083370565616e-05,
+      "loss": 2.4185,
+      "step": 1695
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99605877963821e-05,
+      "loss": 2.4196,
+      "step": 1700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996034111784413e-05,
+      "loss": 2.4839,
+      "step": 1705
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.996009367004607e-05,
+      "loss": 2.4537,
+      "step": 1710
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995984545299171e-05,
+      "loss": 2.4431,
+      "step": 1715
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995959646668488e-05,
+      "loss": 2.3891,
+      "step": 1720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995934671112941e-05,
+      "loss": 2.3896,
+      "step": 1725
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995909618632915e-05,
+      "loss": 2.3581,
+      "step": 1730
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995884489228794e-05,
+      "loss": 2.4947,
+      "step": 1735
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995859282900967e-05,
+      "loss": 2.4027,
+      "step": 1740
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995833999649821e-05,
+      "loss": 2.4233,
+      "step": 1745
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995808639475746e-05,
+      "loss": 2.4081,
+      "step": 1750
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995783202379131e-05,
+      "loss": 2.3955,
+      "step": 1755
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99575768836037e-05,
+      "loss": 2.4884,
+      "step": 1760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995732097419855e-05,
+      "loss": 2.409,
+      "step": 1765
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995706429557978e-05,
+      "loss": 2.448,
+      "step": 1770
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995680684775137e-05,
+      "loss": 2.41,
+      "step": 1775
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995654863071729e-05,
+      "loss": 2.3777,
+      "step": 1780
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995628964448147e-05,
+      "loss": 2.4343,
+      "step": 1785
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995602988904793e-05,
+      "loss": 2.3848,
+      "step": 1790
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995576936442068e-05,
+      "loss": 2.4433,
+      "step": 1795
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995550807060371e-05,
+      "loss": 2.4459,
+      "step": 1800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995524600760105e-05,
+      "loss": 2.4139,
+      "step": 1805
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995498317541674e-05,
+      "loss": 2.4005,
+      "step": 1810
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995471957405483e-05,
+      "loss": 2.4289,
+      "step": 1815
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995445520351935e-05,
+      "loss": 2.4081,
+      "step": 1820
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995419006381442e-05,
+      "loss": 2.4445,
+      "step": 1825
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995392415494407e-05,
+      "loss": 2.4924,
+      "step": 1830
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995365747691243e-05,
+      "loss": 2.4143,
+      "step": 1835
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995339002972359e-05,
+      "loss": 2.4053,
+      "step": 1840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995312181338168e-05,
+      "loss": 2.4104,
+      "step": 1845
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995285282789082e-05,
+      "loss": 2.3929,
+      "step": 1850
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995258307325515e-05,
+      "loss": 2.3994,
+      "step": 1855
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995231254947883e-05,
+      "loss": 2.4254,
+      "step": 1860
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995204125656603e-05,
+      "loss": 2.352,
+      "step": 1865
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995176919452092e-05,
+      "loss": 2.4527,
+      "step": 1870
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995149636334769e-05,
+      "loss": 2.3883,
+      "step": 1875
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995122276305054e-05,
+      "loss": 2.4859,
+      "step": 1880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995094839363368e-05,
+      "loss": 2.2985,
+      "step": 1885
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995067325510132e-05,
+      "loss": 2.417,
+      "step": 1890
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995039734745774e-05,
+      "loss": 2.3872,
+      "step": 1895
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.995012067070715e-05,
+      "loss": 2.3647,
+      "step": 1900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994984322485383e-05,
+      "loss": 2.4384,
+      "step": 1905
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994956500990205e-05,
+      "loss": 2.4623,
+      "step": 1910
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994928602585608e-05,
+      "loss": 2.4103,
+      "step": 1915
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994900627272023e-05,
+      "loss": 2.3676,
+      "step": 1920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99487257504988e-05,
+      "loss": 2.4409,
+      "step": 1925
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994844445919612e-05,
+      "loss": 2.373,
+      "step": 1930
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994816239881648e-05,
+      "loss": 2.5025,
+      "step": 1935
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99478795693643e-05,
+      "loss": 2.4449,
+      "step": 1940
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994759597084384e-05,
+      "loss": 2.3656,
+      "step": 1945
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994731160325955e-05,
+      "loss": 2.4325,
+      "step": 1950
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994702646661577e-05,
+      "loss": 2.301,
+      "step": 1955
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994674056091689e-05,
+      "loss": 2.4401,
+      "step": 1960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994645388616732e-05,
+      "loss": 2.377,
+      "step": 1965
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994616644237148e-05,
+      "loss": 2.3128,
+      "step": 1970
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994587822953378e-05,
+      "loss": 2.5291,
+      "step": 1975
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994558924765866e-05,
+      "loss": 2.3376,
+      "step": 1980
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99452994967506e-05,
+      "loss": 2.3627,
+      "step": 1985
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994500897681401e-05,
+      "loss": 2.3505,
+      "step": 1990
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994471768785339e-05,
+      "loss": 2.3763,
+      "step": 1995
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994442562987324e-05,
+      "loss": 2.4746,
+      "step": 2000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994413280287803e-05,
+      "loss": 2.3738,
+      "step": 2005
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994383920687229e-05,
+      "loss": 2.3992,
+      "step": 2010
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994354484186053e-05,
+      "loss": 2.3451,
+      "step": 2015
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994324970784727e-05,
+      "loss": 2.4516,
+      "step": 2020
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994295380483709e-05,
+      "loss": 2.3444,
+      "step": 2025
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99426571328345e-05,
+      "loss": 2.3135,
+      "step": 2030
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994235969184411e-05,
+      "loss": 2.4334,
+      "step": 2035
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994206148187046e-05,
+      "loss": 2.3167,
+      "step": 2040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994176250291818e-05,
+      "loss": 2.3463,
+      "step": 2045
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994146275499186e-05,
+      "loss": 2.3544,
+      "step": 2050
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994116223809611e-05,
+      "loss": 2.3879,
+      "step": 2055
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994086095223557e-05,
+      "loss": 2.3984,
+      "step": 2060
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994055889741485e-05,
+      "loss": 2.4373,
+      "step": 2065
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994025607363862e-05,
+      "loss": 2.376,
+      "step": 2070
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993995248091155e-05,
+      "loss": 2.4802,
+      "step": 2075
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993964811923831e-05,
+      "loss": 2.3893,
+      "step": 2080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993934298862358e-05,
+      "loss": 2.4356,
+      "step": 2085
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993903708907205e-05,
+      "loss": 2.4178,
+      "step": 2090
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993873042058845e-05,
+      "loss": 2.4467,
+      "step": 2095
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993842298317749e-05,
+      "loss": 2.3824,
+      "step": 2100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993811477684392e-05,
+      "loss": 2.3588,
+      "step": 2105
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993780580159246e-05,
+      "loss": 2.2839,
+      "step": 2110
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993749605742788e-05,
+      "loss": 2.4074,
+      "step": 2115
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993718554435495e-05,
+      "loss": 2.3843,
+      "step": 2120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993687426237846e-05,
+      "loss": 2.3552,
+      "step": 2125
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993656221150318e-05,
+      "loss": 2.3654,
+      "step": 2130
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993624939173394e-05,
+      "loss": 2.3168,
+      "step": 2135
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993593580307553e-05,
+      "loss": 2.3784,
+      "step": 2140
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99356214455328e-05,
+      "loss": 2.3739,
+      "step": 2145
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993530631911059e-05,
+      "loss": 2.3465,
+      "step": 2150
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993499042381373e-05,
+      "loss": 2.3981,
+      "step": 2155
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993467375964711e-05,
+      "loss": 2.318,
+      "step": 2160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99343563266156e-05,
+      "loss": 2.4133,
+      "step": 2165
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993403812472408e-05,
+      "loss": 2.4693,
+      "step": 2170
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993371915397745e-05,
+      "loss": 2.4214,
+      "step": 2175
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993339941438062e-05,
+      "loss": 2.3688,
+      "step": 2180
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993307890593853e-05,
+      "loss": 2.3853,
+      "step": 2185
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993275762865609e-05,
+      "loss": 2.3267,
+      "step": 2190
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993243558253826e-05,
+      "loss": 2.3306,
+      "step": 2195
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993211276759e-05,
+      "loss": 2.2614,
+      "step": 2200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993178918381628e-05,
+      "loss": 2.3915,
+      "step": 2205
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993146483122207e-05,
+      "loss": 2.403,
+      "step": 2210
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993113970981238e-05,
+      "loss": 2.3505,
+      "step": 2215
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993081381959221e-05,
+      "loss": 2.4745,
+      "step": 2220
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99304871605666e-05,
+      "loss": 2.3792,
+      "step": 2225
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.993015973274053e-05,
+      "loss": 2.3561,
+      "step": 2230
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992983153611908e-05,
+      "loss": 2.3619,
+      "step": 2235
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99295025707073e-05,
+      "loss": 2.3838,
+      "step": 2240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992917283651024e-05,
+      "loss": 2.3265,
+      "step": 2245
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992884233353298e-05,
+      "loss": 2.4172,
+      "step": 2250
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992851106178063e-05,
+      "loss": 2.2826,
+      "step": 2255
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992817902125826e-05,
+      "loss": 2.3871,
+      "step": 2260
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.9927846211971e-05,
+      "loss": 2.3853,
+      "step": 2265
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992751263392397e-05,
+      "loss": 2.233,
+      "step": 2270
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992717828712232e-05,
+      "loss": 2.4092,
+      "step": 2275
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992684317157117e-05,
+      "loss": 2.3586,
+      "step": 2280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992650728727571e-05,
+      "loss": 2.3862,
+      "step": 2285
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99261706342411e-05,
+      "loss": 2.3466,
+      "step": 2290
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99258332124725e-05,
+      "loss": 2.3949,
+      "step": 2295
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992549502197514e-05,
+      "loss": 2.3853,
+      "step": 2300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992515606275423e-05,
+      "loss": 2.3902,
+      "step": 2305
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992481633481495e-05,
+      "loss": 2.3866,
+      "step": 2310
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992447583816257e-05,
+      "loss": 2.3882,
+      "step": 2315
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992413457280231e-05,
+      "loss": 2.3806,
+      "step": 2320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992379253873943e-05,
+      "loss": 2.466,
+      "step": 2325
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99234497359792e-05,
+      "loss": 2.3619,
+      "step": 2330
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992310616452689e-05,
+      "loss": 2.3696,
+      "step": 2335
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99227618243878e-05,
+      "loss": 2.368,
+      "step": 2340
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992241671556724e-05,
+      "loss": 2.3657,
+      "step": 2345
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99220708380705e-05,
+      "loss": 2.3522,
+      "step": 2350
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99217241919029e-05,
+      "loss": 2.2997,
+      "step": 2355
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992137677706982e-05,
+      "loss": 2.3536,
+      "step": 2360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99210285935766e-05,
+      "loss": 2.3626,
+      "step": 2365
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992067964142856e-05,
+      "loss": 2.313,
+      "step": 2370
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.992032992063109e-05,
+      "loss": 2.3206,
+      "step": 2375
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99199794311896e-05,
+      "loss": 2.3265,
+      "step": 2380
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991962817310947e-05,
+      "loss": 2.2704,
+      "step": 2385
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991927614639612e-05,
+      "loss": 2.2784,
+      "step": 2390
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991892335105493e-05,
+      "loss": 2.3154,
+      "step": 2395
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991856978709137e-05,
+      "loss": 2.3381,
+      "step": 2400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99182154545109e-05,
+      "loss": 2.3826,
+      "step": 2405
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991786035331893e-05,
+      "loss": 2.2945,
+      "step": 2410
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991750448352096e-05,
+      "loss": 2.3473,
+      "step": 2415
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991714784512243e-05,
+      "loss": 2.4255,
+      "step": 2420
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991679043812887e-05,
+      "loss": 2.3745,
+      "step": 2425
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99164322625458e-05,
+      "loss": 2.3992,
+      "step": 2430
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991607331837867e-05,
+      "loss": 2.3323,
+      "step": 2435
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991571360563306e-05,
+      "loss": 2.4292,
+      "step": 2440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99153531243145e-05,
+      "loss": 2.3548,
+      "step": 2445
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991499187442852e-05,
+      "loss": 2.3719,
+      "step": 2450
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99146298559807e-05,
+      "loss": 2.3623,
+      "step": 2455
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991426706897661e-05,
+      "loss": 2.4133,
+      "step": 2460
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991390351342185e-05,
+      "loss": 2.4138,
+      "step": 2465
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991353918932198e-05,
+      "loss": 2.3205,
+      "step": 2470
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991317409668264e-05,
+      "loss": 2.3684,
+      "step": 2475
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991280823550945e-05,
+      "loss": 2.3346,
+      "step": 2480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991244160580803e-05,
+      "loss": 2.3879,
+      "step": 2485
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991207420758404e-05,
+      "loss": 2.3802,
+      "step": 2490
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991170604084314e-05,
+      "loss": 2.3434,
+      "step": 2495
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991133710559097e-05,
+      "loss": 2.3505,
+      "step": 2500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991096740183324e-05,
+      "loss": 2.3724,
+      "step": 2505
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991059692957563e-05,
+      "loss": 2.3889,
+      "step": 2510
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.991022568882384e-05,
+      "loss": 2.3868,
+      "step": 2515
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99098536795836e-05,
+      "loss": 2.2963,
+      "step": 2520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990948090186063e-05,
+      "loss": 2.3465,
+      "step": 2525
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990910735566065e-05,
+      "loss": 2.3814,
+      "step": 2530
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990873304098946e-05,
+      "loss": 2.3325,
+      "step": 2535
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990835795785278e-05,
+      "loss": 2.3165,
+      "step": 2540
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990798210625642e-05,
+      "loss": 2.2947,
+      "step": 2545
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990760548620613e-05,
+      "loss": 2.4551,
+      "step": 2550
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990722809770773e-05,
+      "loss": 2.3211,
+      "step": 2555
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990684994076702e-05,
+      "loss": 2.4179,
+      "step": 2560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990647101538983e-05,
+      "loss": 2.3238,
+      "step": 2565
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990609132158201e-05,
+      "loss": 2.3992,
+      "step": 2570
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990571085934938e-05,
+      "loss": 2.2842,
+      "step": 2575
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990532962869783e-05,
+      "loss": 2.4038,
+      "step": 2580
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990494762963318e-05,
+      "loss": 2.3693,
+      "step": 2585
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990456486216136e-05,
+      "loss": 2.358,
+      "step": 2590
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990418132628824e-05,
+      "loss": 2.4003,
+      "step": 2595
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990379702201974e-05,
+      "loss": 2.4024,
+      "step": 2600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990341194936173e-05,
+      "loss": 2.3517,
+      "step": 2605
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990302610832021e-05,
+      "loss": 2.4183,
+      "step": 2610
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990263949890108e-05,
+      "loss": 2.4618,
+      "step": 2615
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990225212111029e-05,
+      "loss": 2.2794,
+      "step": 2620
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990186397495383e-05,
+      "loss": 2.2619,
+      "step": 2625
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990147506043766e-05,
+      "loss": 2.2734,
+      "step": 2630
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990108537756775e-05,
+      "loss": 2.3587,
+      "step": 2635
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990069492635015e-05,
+      "loss": 2.3266,
+      "step": 2640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.990030370679082e-05,
+      "loss": 2.3427,
+      "step": 2645
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989991171889581e-05,
+      "loss": 2.3216,
+      "step": 2650
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989951896267115e-05,
+      "loss": 2.3416,
+      "step": 2655
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989912543812289e-05,
+      "loss": 2.3477,
+      "step": 2660
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989873114525709e-05,
+      "loss": 2.3372,
+      "step": 2665
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989833608407981e-05,
+      "loss": 2.3167,
+      "step": 2670
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989794025459715e-05,
+      "loss": 2.3239,
+      "step": 2675
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989754365681519e-05,
+      "loss": 2.3563,
+      "step": 2680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989714629074006e-05,
+      "loss": 2.3345,
+      "step": 2685
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989674815637784e-05,
+      "loss": 2.4123,
+      "step": 2690
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989634925373468e-05,
+      "loss": 2.3507,
+      "step": 2695
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989594958281673e-05,
+      "loss": 2.292,
+      "step": 2700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989554914363015e-05,
+      "loss": 2.3743,
+      "step": 2705
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989514793618109e-05,
+      "loss": 2.3874,
+      "step": 2710
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989474596047572e-05,
+      "loss": 2.3209,
+      "step": 2715
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989434321652024e-05,
+      "loss": 2.3604,
+      "step": 2720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989393970432085e-05,
+      "loss": 2.3424,
+      "step": 2725
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989353542388378e-05,
+      "loss": 2.3024,
+      "step": 2730
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989313037521524e-05,
+      "loss": 2.4235,
+      "step": 2735
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989272455832145e-05,
+      "loss": 2.3034,
+      "step": 2740
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989231797320869e-05,
+      "loss": 2.3289,
+      "step": 2745
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989191061988319e-05,
+      "loss": 2.3564,
+      "step": 2750
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989150249835125e-05,
+      "loss": 2.3878,
+      "step": 2755
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989109360861913e-05,
+      "loss": 2.3046,
+      "step": 2760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989068395069313e-05,
+      "loss": 2.3763,
+      "step": 2765
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.989027352457958e-05,
+      "loss": 2.3645,
+      "step": 2770
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988986233028478e-05,
+      "loss": 2.2728,
+      "step": 2775
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988945036781507e-05,
+      "loss": 2.2989,
+      "step": 2780
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988903763717679e-05,
+      "loss": 2.2334,
+      "step": 2785
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988862413837628e-05,
+      "loss": 2.2327,
+      "step": 2790
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988820987141993e-05,
+      "loss": 2.3926,
+      "step": 2795
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988779483631409e-05,
+      "loss": 2.3845,
+      "step": 2800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98873790330652e-05,
+      "loss": 2.2483,
+      "step": 2805
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988696246167961e-05,
+      "loss": 2.3089,
+      "step": 2810
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988654512216378e-05,
+      "loss": 2.2047,
+      "step": 2815
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988612701452409e-05,
+      "loss": 2.3429,
+      "step": 2820
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988570813876701e-05,
+      "loss": 2.3429,
+      "step": 2825
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988528849489897e-05,
+      "loss": 2.338,
+      "step": 2830
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988486808292646e-05,
+      "loss": 2.2374,
+      "step": 2835
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988444690285592e-05,
+      "loss": 2.2463,
+      "step": 2840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988402495469385e-05,
+      "loss": 2.3205,
+      "step": 2845
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988360223844673e-05,
+      "loss": 2.341,
+      "step": 2850
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98831787541211e-05,
+      "loss": 2.3593,
+      "step": 2855
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988275450172346e-05,
+      "loss": 2.3226,
+      "step": 2860
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988232948126037e-05,
+      "loss": 2.2739,
+      "step": 2865
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988190369273833e-05,
+      "loss": 2.2765,
+      "step": 2870
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988147713616391e-05,
+      "loss": 2.3146,
+      "step": 2875
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988104981154371e-05,
+      "loss": 2.2633,
+      "step": 2880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988062171888429e-05,
+      "loss": 2.2912,
+      "step": 2885
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.988019285819221e-05,
+      "loss": 2.388,
+      "step": 2890
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987976322947411e-05,
+      "loss": 2.3713,
+      "step": 2895
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98793328327366e-05,
+      "loss": 2.2527,
+      "step": 2900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98789016679863e-05,
+      "loss": 2.3604,
+      "step": 2905
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987846973522985e-05,
+      "loss": 2.1892,
+      "step": 2910
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98780370344739e-05,
+      "loss": 2.2705,
+      "step": 2915
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987760356572512e-05,
+      "loss": 2.3203,
+      "step": 2920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987716932899018e-05,
+      "loss": 2.2626,
+      "step": 2925
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987673432427576e-05,
+      "loss": 2.339,
+      "step": 2930
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987629855158855e-05,
+      "loss": 2.2512,
+      "step": 2935
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98758620109353e-05,
+      "loss": 2.319,
+      "step": 2940
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98754247023227e-05,
+      "loss": 2.2824,
+      "step": 2945
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987498662575747e-05,
+      "loss": 2.4025,
+      "step": 2950
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987454778124639e-05,
+      "loss": 2.3919,
+      "step": 2955
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987410816879619e-05,
+      "loss": 2.3347,
+      "step": 2960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987366778841365e-05,
+      "loss": 2.3593,
+      "step": 2965
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987322664010554e-05,
+      "loss": 2.2969,
+      "step": 2970
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987278472387867e-05,
+      "loss": 2.2978,
+      "step": 2975
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987234203973986e-05,
+      "loss": 2.3982,
+      "step": 2980
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987189858769586e-05,
+      "loss": 2.4113,
+      "step": 2985
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987145436775357e-05,
+      "loss": 2.434,
+      "step": 2990
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98710093799198e-05,
+      "loss": 2.2792,
+      "step": 2995
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.987056362420138e-05,
+      "loss": 2.3523,
+      "step": 3000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98701171006052e-05,
+      "loss": 2.2841,
+      "step": 3005
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986966980913817e-05,
+      "loss": 2.3459,
+      "step": 3010
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986922174980709e-05,
+      "loss": 2.2959,
+      "step": 3015
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986877292261893e-05,
+      "loss": 2.3091,
+      "step": 3020
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986832332758058e-05,
+      "loss": 2.2999,
+      "step": 3025
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986787296469895e-05,
+      "loss": 2.2881,
+      "step": 3030
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986742183398099e-05,
+      "loss": 2.2579,
+      "step": 3035
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986696993543366e-05,
+      "loss": 2.3496,
+      "step": 3040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986651726906387e-05,
+      "loss": 2.2874,
+      "step": 3045
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986606383487862e-05,
+      "loss": 2.3023,
+      "step": 3050
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98656096328849e-05,
+      "loss": 2.3129,
+      "step": 3055
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98651546630897e-05,
+      "loss": 2.4075,
+      "step": 3060
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986469892550001e-05,
+      "loss": 2.2214,
+      "step": 3065
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986424242012287e-05,
+      "loss": 2.2589,
+      "step": 3070
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986378514696528e-05,
+      "loss": 2.3294,
+      "step": 3075
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98633271060343e-05,
+      "loss": 2.2725,
+      "step": 3080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.9862868297337e-05,
+      "loss": 2.3247,
+      "step": 3085
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98624087208804e-05,
+      "loss": 2.3375,
+      "step": 3090
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98619483766716e-05,
+      "loss": 2.3534,
+      "step": 3095
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98614872647177e-05,
+      "loss": 2.2878,
+      "step": 3100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986102538502579e-05,
+      "loss": 2.3655,
+      "step": 3105
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986056273760297e-05,
+      "loss": 2.2853,
+      "step": 3110
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.986009932245637e-05,
+      "loss": 2.3147,
+      "step": 3115
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.985963513959313e-05,
+      "loss": 2.2755,
+      "step": 3120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.985917018902041e-05,
+      "loss": 2.2856,
+      "step": 3125
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.985870447074535e-05,
+      "loss": 2.3196,
+      "step": 3130
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.985823798477512e-05,
+      "loss": 2.2685,
+      "step": 3135
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.985777073111691e-05,
+      "loss": 2.2899,
+      "step": 3140
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.985730270977791e-05,
+      "loss": 2.2988,
+      "step": 3145
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.985683392076534e-05,
+      "loss": 2.397,
+      "step": 3150
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.98563643640864e-05,
+      "loss": 2.2802,
+      "step": 3155
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.985589403974833e-05,
+      "loss": 2.2991,
+      "step": 3160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.985542294775838e-05,
+      "loss": 2.2802,
+      "step": 3165
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.985495108812377e-05,
+      "loss": 2.3144,
+      "step": 3170
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.985447846085181e-05,
+      "loss": 2.3108,
+      "step": 3175
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.985400506594974e-05,
+      "loss": 2.3358,
+      "step": 3180
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.985353090342487e-05,
+      "loss": 2.3165,
+      "step": 3185
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.985305597328449e-05,
+      "loss": 2.3595,
+      "step": 3190
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.985258027553593e-05,
+      "loss": 2.1832,
+      "step": 3195
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98521038101865e-05,
+      "loss": 2.2893,
+      "step": 3200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.985162657724354e-05,
+      "loss": 2.3715,
+      "step": 3205
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98511485767144e-05,
+      "loss": 2.3812,
+      "step": 3210
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.985066980860642e-05,
+      "loss": 2.2409,
+      "step": 3215
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.9850190272927e-05,
+      "loss": 2.3263,
+      "step": 3220
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984970996968352e-05,
+      "loss": 2.3001,
+      "step": 3225
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984922889888335e-05,
+      "loss": 2.2641,
+      "step": 3230
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984874706053393e-05,
+      "loss": 2.2901,
+      "step": 3235
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984826445464266e-05,
+      "loss": 2.3808,
+      "step": 3240
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984778108121698e-05,
+      "loss": 2.2285,
+      "step": 3245
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984729694026431e-05,
+      "loss": 2.2539,
+      "step": 3250
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984681203179214e-05,
+      "loss": 2.3509,
+      "step": 3255
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984632635580792e-05,
+      "loss": 2.3404,
+      "step": 3260
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984583991231913e-05,
+      "loss": 2.3602,
+      "step": 3265
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984535270133324e-05,
+      "loss": 2.3682,
+      "step": 3270
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984486472285777e-05,
+      "loss": 2.3913,
+      "step": 3275
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984437597690024e-05,
+      "loss": 2.3063,
+      "step": 3280
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984388646346817e-05,
+      "loss": 2.3705,
+      "step": 3285
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984339618256909e-05,
+      "loss": 2.2668,
+      "step": 3290
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984290513421057e-05,
+      "loss": 2.35,
+      "step": 3295
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984241331840015e-05,
+      "loss": 2.3137,
+      "step": 3300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984192073514538e-05,
+      "loss": 2.3296,
+      "step": 3305
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98414273844539e-05,
+      "loss": 2.199,
+      "step": 3310
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984093326633325e-05,
+      "loss": 2.2811,
+      "step": 3315
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98404383807911e-05,
+      "loss": 2.2835,
+      "step": 3320
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983994272783502e-05,
+      "loss": 2.3144,
+      "step": 3325
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983944630747266e-05,
+      "loss": 2.3136,
+      "step": 3330
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983894911971167e-05,
+      "loss": 2.3045,
+      "step": 3335
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983845116455969e-05,
+      "loss": 2.2329,
+      "step": 3340
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983795244202439e-05,
+      "loss": 2.3072,
+      "step": 3345
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983745295211347e-05,
+      "loss": 2.3152,
+      "step": 3350
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983695269483459e-05,
+      "loss": 2.3246,
+      "step": 3355
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983645167019546e-05,
+      "loss": 2.3248,
+      "step": 3360
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983594987820381e-05,
+      "loss": 2.3302,
+      "step": 3365
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983544731886735e-05,
+      "loss": 2.334,
+      "step": 3370
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983494399219386e-05,
+      "loss": 2.314,
+      "step": 3375
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983443989819103e-05,
+      "loss": 2.3419,
+      "step": 3380
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983393503686665e-05,
+      "loss": 2.2459,
+      "step": 3385
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983342940822849e-05,
+      "loss": 2.3459,
+      "step": 3390
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983292301228432e-05,
+      "loss": 2.3459,
+      "step": 3395
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983241584904197e-05,
+      "loss": 2.3051,
+      "step": 3400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983190791850925e-05,
+      "loss": 2.2793,
+      "step": 3405
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983139922069394e-05,
+      "loss": 2.2498,
+      "step": 3410
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983088975560391e-05,
+      "loss": 2.3175,
+      "step": 3415
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983037952324699e-05,
+      "loss": 2.2861,
+      "step": 3420
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982986852363102e-05,
+      "loss": 2.2955,
+      "step": 3425
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98293567567639e-05,
+      "loss": 2.2792,
+      "step": 3430
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982884422265349e-05,
+      "loss": 2.2862,
+      "step": 3435
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982833092130768e-05,
+      "loss": 2.2474,
+      "step": 3440
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982781685273438e-05,
+      "loss": 2.348,
+      "step": 3445
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982730201694153e-05,
+      "loss": 2.3897,
+      "step": 3450
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982678641393701e-05,
+      "loss": 2.3274,
+      "step": 3455
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98262700437288e-05,
+      "loss": 2.3125,
+      "step": 3460
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982575290632481e-05,
+      "loss": 2.2552,
+      "step": 3465
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982523500173304e-05,
+      "loss": 2.2147,
+      "step": 3470
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982471632996144e-05,
+      "loss": 2.2603,
+      "step": 3475
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.9824196891018e-05,
+      "loss": 2.2767,
+      "step": 3480
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982367668491075e-05,
+      "loss": 2.2521,
+      "step": 3485
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982315571164765e-05,
+      "loss": 2.3282,
+      "step": 3490
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982263397123675e-05,
+      "loss": 2.3706,
+      "step": 3495
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982211146368608e-05,
+      "loss": 2.3938,
+      "step": 3500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982158818900369e-05,
+      "loss": 2.2767,
+      "step": 3505
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982106414719762e-05,
+      "loss": 2.2298,
+      "step": 3510
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982053933827595e-05,
+      "loss": 2.3042,
+      "step": 3515
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.982001376224677e-05,
+      "loss": 2.333,
+      "step": 3520
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981948741911816e-05,
+      "loss": 2.3507,
+      "step": 3525
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981896030889821e-05,
+      "loss": 2.295,
+      "step": 3530
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981843243159507e-05,
+      "loss": 2.3,
+      "step": 3535
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981790378721684e-05,
+      "loss": 2.2971,
+      "step": 3540
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98173743757717e-05,
+      "loss": 2.3063,
+      "step": 3545
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981684419726775e-05,
+      "loss": 2.272,
+      "step": 3550
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981631325171318e-05,
+      "loss": 2.2542,
+      "step": 3555
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981578153911616e-05,
+      "loss": 2.3389,
+      "step": 3560
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981524905948487e-05,
+      "loss": 2.3276,
+      "step": 3565
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981471581282753e-05,
+      "loss": 2.3268,
+      "step": 3570
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981418179915234e-05,
+      "loss": 2.247,
+      "step": 3575
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981364701846752e-05,
+      "loss": 2.3676,
+      "step": 3580
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981311147078128e-05,
+      "loss": 2.2852,
+      "step": 3585
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981257515610192e-05,
+      "loss": 2.2176,
+      "step": 3590
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981203807443764e-05,
+      "loss": 2.2895,
+      "step": 3595
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981150022579676e-05,
+      "loss": 2.3178,
+      "step": 3600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981096161018754e-05,
+      "loss": 2.2514,
+      "step": 3605
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.981042222761826e-05,
+      "loss": 2.3644,
+      "step": 3610
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980988207809726e-05,
+      "loss": 2.3326,
+      "step": 3615
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98093411616328e-05,
+      "loss": 2.2961,
+      "step": 3620
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980879947823327e-05,
+      "loss": 2.2616,
+      "step": 3625
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980825702790696e-05,
+      "loss": 2.2439,
+      "step": 3630
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980771381066227e-05,
+      "loss": 2.2797,
+      "step": 3635
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980716982650753e-05,
+      "loss": 2.25,
+      "step": 3640
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980662507545113e-05,
+      "loss": 2.3018,
+      "step": 3645
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980607955750146e-05,
+      "loss": 2.2686,
+      "step": 3650
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98055332726669e-05,
+      "loss": 2.3805,
+      "step": 3655
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980498622095588e-05,
+      "loss": 2.3218,
+      "step": 3660
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980443840237681e-05,
+      "loss": 2.3353,
+      "step": 3665
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980388981693816e-05,
+      "loss": 2.3342,
+      "step": 3670
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980334046464832e-05,
+      "loss": 2.2184,
+      "step": 3675
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980279034551579e-05,
+      "loss": 2.341,
+      "step": 3680
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980223945954902e-05,
+      "loss": 2.2364,
+      "step": 3685
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980168780675651e-05,
+      "loss": 2.2659,
+      "step": 3690
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980113538714675e-05,
+      "loss": 2.3344,
+      "step": 3695
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.980058220072825e-05,
+      "loss": 2.285,
+      "step": 3700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98000282475095e-05,
+      "loss": 2.2328,
+      "step": 3705
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979947352749906e-05,
+      "loss": 2.2904,
+      "step": 3710
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979891804070545e-05,
+      "loss": 2.2441,
+      "step": 3715
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979836178713724e-05,
+      "loss": 2.2306,
+      "step": 3720
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979780476680298e-05,
+      "loss": 2.2437,
+      "step": 3725
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979724697971125e-05,
+      "loss": 2.2094,
+      "step": 3730
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979668842587065e-05,
+      "loss": 2.2705,
+      "step": 3735
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979612910528976e-05,
+      "loss": 2.2357,
+      "step": 3740
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979556901797723e-05,
+      "loss": 2.1606,
+      "step": 3745
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979500816394165e-05,
+      "loss": 2.2645,
+      "step": 3750
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979444654319165e-05,
+      "loss": 2.2358,
+      "step": 3755
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979388415573591e-05,
+      "loss": 2.2744,
+      "step": 3760
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979332100158307e-05,
+      "loss": 2.2737,
+      "step": 3765
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979275708074179e-05,
+      "loss": 2.2459,
+      "step": 3770
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979219239322078e-05,
+      "loss": 2.2298,
+      "step": 3775
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979162693902871e-05,
+      "loss": 2.3438,
+      "step": 3780
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97910607181743e-05,
+      "loss": 2.2774,
+      "step": 3785
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.979049373066626e-05,
+      "loss": 2.3293,
+      "step": 3790
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978992597651334e-05,
+      "loss": 2.3051,
+      "step": 3795
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978935745572424e-05,
+      "loss": 2.2934,
+      "step": 3800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978878816830776e-05,
+      "loss": 2.1663,
+      "step": 3805
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978821811427263e-05,
+      "loss": 2.2536,
+      "step": 3810
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978764729362766e-05,
+      "loss": 2.3146,
+      "step": 3815
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978707570638162e-05,
+      "loss": 2.2096,
+      "step": 3820
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978650335254332e-05,
+      "loss": 2.3275,
+      "step": 3825
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978593023212155e-05,
+      "loss": 2.3102,
+      "step": 3830
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978535634512515e-05,
+      "loss": 2.3296,
+      "step": 3835
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978478169156298e-05,
+      "loss": 2.2232,
+      "step": 3840
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978420627144385e-05,
+      "loss": 2.3118,
+      "step": 3845
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978363008477665e-05,
+      "loss": 2.1779,
+      "step": 3850
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978305313157022e-05,
+      "loss": 2.2689,
+      "step": 3855
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978247541183348e-05,
+      "loss": 2.1777,
+      "step": 3860
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978189692557529e-05,
+      "loss": 2.3502,
+      "step": 3865
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978131767280456e-05,
+      "loss": 2.2188,
+      "step": 3870
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978073765353026e-05,
+      "loss": 2.2625,
+      "step": 3875
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.978015686776125e-05,
+      "loss": 2.288,
+      "step": 3880
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977957531550652e-05,
+      "loss": 2.3494,
+      "step": 3885
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977899299677502e-05,
+      "loss": 2.333,
+      "step": 3890
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97784099115757e-05,
+      "loss": 2.2879,
+      "step": 3895
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977782605991754e-05,
+      "loss": 2.273,
+      "step": 3900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977724144180953e-05,
+      "loss": 2.2467,
+      "step": 3905
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977665605726067e-05,
+      "loss": 2.3313,
+      "step": 3910
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977606990628001e-05,
+      "loss": 2.2959,
+      "step": 3915
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977548298887651e-05,
+      "loss": 2.3654,
+      "step": 3920
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977489530505927e-05,
+      "loss": 2.2376,
+      "step": 3925
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97743068548373e-05,
+      "loss": 2.3269,
+      "step": 3930
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977371763821966e-05,
+      "loss": 2.2641,
+      "step": 3935
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977312765521545e-05,
+      "loss": 2.2596,
+      "step": 3940
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977253690583372e-05,
+      "loss": 2.2769,
+      "step": 3945
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97719453900836e-05,
+      "loss": 2.2082,
+      "step": 3950
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977135310797418e-05,
+      "loss": 2.2451,
+      "step": 3955
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977076005951458e-05,
+      "loss": 2.3196,
+      "step": 3960
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.977016624471392e-05,
+      "loss": 2.2339,
+      "step": 3965
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976957166358137e-05,
+      "loss": 2.2546,
+      "step": 3970
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976897631612607e-05,
+      "loss": 2.3357,
+      "step": 3975
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97683802023572e-05,
+      "loss": 2.2903,
+      "step": 3980
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976778332228389e-05,
+      "loss": 2.2212,
+      "step": 3985
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97671856759154e-05,
+      "loss": 2.2799,
+      "step": 3990
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976658726326088e-05,
+      "loss": 2.3605,
+      "step": 3995
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976598808432957e-05,
+      "loss": 2.2614,
+      "step": 4000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976538813913068e-05,
+      "loss": 2.3054,
+      "step": 4005
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976478742767347e-05,
+      "loss": 2.2845,
+      "step": 4010
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976418594996716e-05,
+      "loss": 2.2929,
+      "step": 4015
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976358370602103e-05,
+      "loss": 2.2686,
+      "step": 4020
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976298069584438e-05,
+      "loss": 2.2737,
+      "step": 4025
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976237691944643e-05,
+      "loss": 2.2341,
+      "step": 4030
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976177237683653e-05,
+      "loss": 2.2159,
+      "step": 4035
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976116706802395e-05,
+      "loss": 2.2229,
+      "step": 4040
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.976056099301805e-05,
+      "loss": 2.1922,
+      "step": 4045
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975995415182812e-05,
+      "loss": 2.2847,
+      "step": 4050
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975934654446355e-05,
+      "loss": 2.3277,
+      "step": 4055
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975873817093367e-05,
+      "loss": 2.2143,
+      "step": 4060
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975812903124782e-05,
+      "loss": 2.2929,
+      "step": 4065
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975751912541544e-05,
+      "loss": 2.2916,
+      "step": 4070
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975690845344588e-05,
+      "loss": 2.3112,
+      "step": 4075
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975629701534854e-05,
+      "loss": 2.2847,
+      "step": 4080
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975568481113285e-05,
+      "loss": 2.3033,
+      "step": 4085
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975507184080823e-05,
+      "loss": 2.2035,
+      "step": 4090
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975445810438413e-05,
+      "loss": 2.3418,
+      "step": 4095
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975384360186999e-05,
+      "loss": 2.1919,
+      "step": 4100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975322833327527e-05,
+      "loss": 2.244,
+      "step": 4105
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975261229860944e-05,
+      "loss": 2.2319,
+      "step": 4110
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975199549788199e-05,
+      "loss": 2.2881,
+      "step": 4115
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975137793110243e-05,
+      "loss": 2.2925,
+      "step": 4120
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975075959828026e-05,
+      "loss": 2.3477,
+      "step": 4125
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.975014049942499e-05,
+      "loss": 2.2205,
+      "step": 4130
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974952063454615e-05,
+      "loss": 2.222,
+      "step": 4135
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974890000365332e-05,
+      "loss": 2.2326,
+      "step": 4140
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.9748278606756e-05,
+      "loss": 2.2545,
+      "step": 4145
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974765644386382e-05,
+      "loss": 2.251,
+      "step": 4150
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974703351498631e-05,
+      "loss": 2.225,
+      "step": 4155
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974640982013307e-05,
+      "loss": 2.2383,
+      "step": 4160
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974578535931374e-05,
+      "loss": 2.2304,
+      "step": 4165
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974516013253789e-05,
+      "loss": 2.1174,
+      "step": 4170
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974453413981517e-05,
+      "loss": 2.2287,
+      "step": 4175
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974390738115521e-05,
+      "loss": 2.1452,
+      "step": 4180
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974327985656767e-05,
+      "loss": 2.2313,
+      "step": 4185
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974265156606221e-05,
+      "loss": 2.244,
+      "step": 4190
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974202250964849e-05,
+      "loss": 2.2096,
+      "step": 4195
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974139268733623e-05,
+      "loss": 2.2665,
+      "step": 4200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974076209913508e-05,
+      "loss": 2.2563,
+      "step": 4205
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.974013074505477e-05,
+      "loss": 2.2404,
+      "step": 4210
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.973949862510503e-05,
+      "loss": 2.2702,
+      "step": 4215
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97388657392956e-05,
+      "loss": 2.2505,
+      "step": 4220
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97382320876362e-05,
+      "loss": 2.3272,
+      "step": 4225
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.973759767013662e-05,
+      "loss": 2.3273,
+      "step": 4230
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.973696248680659e-05,
+      "loss": 2.2721,
+      "step": 4235
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97363265376559e-05,
+      "loss": 2.2646,
+      "step": 4240
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.973568982269436e-05,
+      "loss": 2.3456,
+      "step": 4245
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.973505234193177e-05,
+      "loss": 2.242,
+      "step": 4250
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.973441409537793e-05,
+      "loss": 2.2289,
+      "step": 4255
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.973377508304269e-05,
+      "loss": 2.1981,
+      "step": 4260
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.973313530493586e-05,
+      "loss": 2.2683,
+      "step": 4265
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.973249476106732e-05,
+      "loss": 2.231,
+      "step": 4270
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97318534514469e-05,
+      "loss": 2.2083,
+      "step": 4275
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97312113760845e-05,
+      "loss": 2.1693,
+      "step": 4280
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.973056853499001e-05,
+      "loss": 2.2361,
+      "step": 4285
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972992492817331e-05,
+      "loss": 2.2527,
+      "step": 4290
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972928055564432e-05,
+      "loss": 2.27,
+      "step": 4295
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972863541741297e-05,
+      "loss": 2.2584,
+      "step": 4300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972798951348918e-05,
+      "loss": 2.3309,
+      "step": 4305
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972734284388288e-05,
+      "loss": 2.2751,
+      "step": 4310
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972669540860406e-05,
+      "loss": 2.3204,
+      "step": 4315
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972604720766268e-05,
+      "loss": 2.2387,
+      "step": 4320
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97253982410687e-05,
+      "loss": 2.2123,
+      "step": 4325
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972474850883213e-05,
+      "loss": 2.352,
+      "step": 4330
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972409801096299e-05,
+      "loss": 2.2467,
+      "step": 4335
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972344674747125e-05,
+      "loss": 2.2255,
+      "step": 4340
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972279471836698e-05,
+      "loss": 2.2355,
+      "step": 4345
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972214192366021e-05,
+      "loss": 2.2793,
+      "step": 4350
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972148836336098e-05,
+      "loss": 2.2235,
+      "step": 4355
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972083403747937e-05,
+      "loss": 2.2527,
+      "step": 4360
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.972017894602542e-05,
+      "loss": 2.223,
+      "step": 4365
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971952308900925e-05,
+      "loss": 2.2136,
+      "step": 4370
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971886646644096e-05,
+      "loss": 2.2413,
+      "step": 4375
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971820907833065e-05,
+      "loss": 2.3761,
+      "step": 4380
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971755092468843e-05,
+      "loss": 2.2545,
+      "step": 4385
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971689200552445e-05,
+      "loss": 2.28,
+      "step": 4390
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971623232084886e-05,
+      "loss": 2.3012,
+      "step": 4395
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971557187067179e-05,
+      "loss": 2.1371,
+      "step": 4400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971491065500345e-05,
+      "loss": 2.2286,
+      "step": 4405
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.9714248673854e-05,
+      "loss": 2.2123,
+      "step": 4410
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971358592723361e-05,
+      "loss": 2.2225,
+      "step": 4415
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.97129224151525e-05,
+      "loss": 2.2789,
+      "step": 4420
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971225813762094e-05,
+      "loss": 2.2073,
+      "step": 4425
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.971159309464906e-05,
+      "loss": 2.2576,
+      "step": 4430
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.971092728624718e-05,
+      "loss": 2.296,
+      "step": 4435
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.97102607124255e-05,
+      "loss": 2.3219,
+      "step": 4440
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970959337319433e-05,
+      "loss": 2.3401,
+      "step": 4445
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970892526856392e-05,
+      "loss": 2.3417,
+      "step": 4450
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970825639854454e-05,
+      "loss": 2.2046,
+      "step": 4455
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970758676314652e-05,
+      "loss": 2.2525,
+      "step": 4460
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970691636238017e-05,
+      "loss": 2.2026,
+      "step": 4465
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970624519625578e-05,
+      "loss": 2.1958,
+      "step": 4470
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970557326478371e-05,
+      "loss": 2.2265,
+      "step": 4475
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.97049005679743e-05,
+      "loss": 2.3012,
+      "step": 4480
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970422710583791e-05,
+      "loss": 2.33,
+      "step": 4485
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970355287838491e-05,
+      "loss": 2.221,
+      "step": 4490
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970287788562567e-05,
+      "loss": 2.1757,
+      "step": 4495
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.97022021275706e-05,
+      "loss": 2.2082,
+      "step": 4500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.97015256042301e-05,
+      "loss": 2.2574,
+      "step": 4505
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.97008483156146e-05,
+      "loss": 2.2297,
+      "step": 4510
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970017026173448e-05,
+      "loss": 2.2699,
+      "step": 4515
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969949144260022e-05,
+      "loss": 2.2965,
+      "step": 4520
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969881185822227e-05,
+      "loss": 2.2645,
+      "step": 4525
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969813150861111e-05,
+      "loss": 2.2684,
+      "step": 4530
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969745039377717e-05,
+      "loss": 2.285,
+      "step": 4535
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969676851373098e-05,
+      "loss": 2.1331,
+      "step": 4540
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969608586848301e-05,
+      "loss": 2.2608,
+      "step": 4545
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969540245804377e-05,
+      "loss": 2.1923,
+      "step": 4550
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96947182824238e-05,
+      "loss": 2.3028,
+      "step": 4555
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969403334163364e-05,
+      "loss": 2.2328,
+      "step": 4560
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969334763568383e-05,
+      "loss": 2.1814,
+      "step": 4565
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969266116458491e-05,
+      "loss": 2.2635,
+      "step": 4570
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969197392834747e-05,
+      "loss": 2.2506,
+      "step": 4575
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969128592698207e-05,
+      "loss": 2.2357,
+      "step": 4580
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969059716049934e-05,
+      "loss": 2.2448,
+      "step": 4585
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968990762890986e-05,
+      "loss": 2.2805,
+      "step": 4590
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968921733222425e-05,
+      "loss": 2.2564,
+      "step": 4595
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968852627045313e-05,
+      "loss": 2.1926,
+      "step": 4600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968783444360716e-05,
+      "loss": 2.2539,
+      "step": 4605
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968714185169698e-05,
+      "loss": 2.3144,
+      "step": 4610
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968644849473326e-05,
+      "loss": 2.3377,
+      "step": 4615
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968575437272667e-05,
+      "loss": 2.1945,
+      "step": 4620
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968505948568792e-05,
+      "loss": 2.2898,
+      "step": 4625
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968436383362766e-05,
+      "loss": 2.3409,
+      "step": 4630
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968366741655663e-05,
+      "loss": 2.3223,
+      "step": 4635
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968297023448558e-05,
+      "loss": 2.221,
+      "step": 4640
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96822722874252e-05,
+      "loss": 2.2661,
+      "step": 4645
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968157357538625e-05,
+      "loss": 2.2018,
+      "step": 4650
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968087409837951e-05,
+      "loss": 2.2,
+      "step": 4655
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968017385641574e-05,
+      "loss": 2.2231,
+      "step": 4660
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967947284950569e-05,
+      "loss": 2.2945,
+      "step": 4665
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967877107766021e-05,
+      "loss": 2.2125,
+      "step": 4670
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967806854089006e-05,
+      "loss": 2.2498,
+      "step": 4675
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967736523920607e-05,
+      "loss": 2.3166,
+      "step": 4680
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967666117261907e-05,
+      "loss": 2.1852,
+      "step": 4685
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96759563411399e-05,
+      "loss": 2.1743,
+      "step": 4690
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967525074477943e-05,
+      "loss": 2.1962,
+      "step": 4695
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96745443835485e-05,
+      "loss": 2.2867,
+      "step": 4700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967383725745799e-05,
+      "loss": 2.291,
+      "step": 4705
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96731293665188e-05,
+      "loss": 2.2361,
+      "step": 4710
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967242071074181e-05,
+      "loss": 2.2383,
+      "step": 4715
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967171129013797e-05,
+      "loss": 2.2743,
+      "step": 4720
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967100110471817e-05,
+      "loss": 2.2753,
+      "step": 4725
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.967029015449334e-05,
+      "loss": 2.2499,
+      "step": 4730
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.966957843947444e-05,
+      "loss": 2.2933,
+      "step": 4735
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.966886595967243e-05,
+      "loss": 2.2203,
+      "step": 4740
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.966815271509828e-05,
+      "loss": 2.246,
+      "step": 4745
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.966743870576297e-05,
+      "loss": 2.2648,
+      "step": 4750
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96667239316775e-05,
+      "loss": 2.2499,
+      "step": 4755
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.966600839285286e-05,
+      "loss": 2.1508,
+      "step": 4760
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.966529208930008e-05,
+      "loss": 2.2601,
+      "step": 4765
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96645750210302e-05,
+      "loss": 2.2032,
+      "step": 4770
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.966385718805423e-05,
+      "loss": 2.2731,
+      "step": 4775
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.966313859038325e-05,
+      "loss": 2.2752,
+      "step": 4780
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.966241922802832e-05,
+      "loss": 2.3035,
+      "step": 4785
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96616991010005e-05,
+      "loss": 2.2656,
+      "step": 4790
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96609782093109e-05,
+      "loss": 2.2331,
+      "step": 4795
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96602565529706e-05,
+      "loss": 2.2404,
+      "step": 4800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965953413199071e-05,
+      "loss": 2.2402,
+      "step": 4805
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965881094638238e-05,
+      "loss": 2.3075,
+      "step": 4810
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965808699615674e-05,
+      "loss": 2.269,
+      "step": 4815
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965736228132492e-05,
+      "loss": 2.3009,
+      "step": 4820
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965663680189807e-05,
+      "loss": 2.2563,
+      "step": 4825
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965591055788738e-05,
+      "loss": 2.1944,
+      "step": 4830
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965518354930405e-05,
+      "loss": 2.3139,
+      "step": 4835
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965445577615922e-05,
+      "loss": 2.0964,
+      "step": 4840
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965372723846415e-05,
+      "loss": 2.2211,
+      "step": 4845
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965299793623003e-05,
+      "loss": 2.1963,
+      "step": 4850
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96522678694681e-05,
+      "loss": 2.2473,
+      "step": 4855
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965153703818959e-05,
+      "loss": 2.2405,
+      "step": 4860
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965080544240575e-05,
+      "loss": 2.2519,
+      "step": 4865
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.965007308212786e-05,
+      "loss": 2.2718,
+      "step": 4870
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964933995736721e-05,
+      "loss": 2.116,
+      "step": 4875
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964860606813504e-05,
+      "loss": 2.2864,
+      "step": 4880
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96478714144427e-05,
+      "loss": 2.2851,
+      "step": 4885
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964713599630146e-05,
+      "loss": 2.279,
+      "step": 4890
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964639981372267e-05,
+      "loss": 2.2644,
+      "step": 4895
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964566286671766e-05,
+      "loss": 2.1937,
+      "step": 4900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964492515529777e-05,
+      "loss": 2.2716,
+      "step": 4905
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964418667947438e-05,
+      "loss": 2.2255,
+      "step": 4910
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964344743925884e-05,
+      "loss": 2.2354,
+      "step": 4915
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964270743466253e-05,
+      "loss": 2.2318,
+      "step": 4920
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964196666569686e-05,
+      "loss": 2.2621,
+      "step": 4925
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964122513237322e-05,
+      "loss": 2.1787,
+      "step": 4930
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.964048283470305e-05,
+      "loss": 2.2432,
+      "step": 4935
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963973977269776e-05,
+      "loss": 2.2742,
+      "step": 4940
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963899594636879e-05,
+      "loss": 2.2599,
+      "step": 4945
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96382513557276e-05,
+      "loss": 2.1767,
+      "step": 4950
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963750600078565e-05,
+      "loss": 2.3009,
+      "step": 4955
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963675988155443e-05,
+      "loss": 2.2184,
+      "step": 4960
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963601299804541e-05,
+      "loss": 2.2126,
+      "step": 4965
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963526535027012e-05,
+      "loss": 2.2596,
+      "step": 4970
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963451693824003e-05,
+      "loss": 2.2653,
+      "step": 4975
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963376776196669e-05,
+      "loss": 2.2172,
+      "step": 4980
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963301782146163e-05,
+      "loss": 2.2542,
+      "step": 4985
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96322671167364e-05,
+      "loss": 2.1849,
+      "step": 4990
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963151564780256e-05,
+      "loss": 2.201,
+      "step": 4995
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963076341467165e-05,
+      "loss": 2.2669,
+      "step": 5000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.963001041735531e-05,
+      "loss": 2.2083,
+      "step": 5005
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962925665586509e-05,
+      "loss": 2.2689,
+      "step": 5010
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962850213021262e-05,
+      "loss": 2.2726,
+      "step": 5015
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96277468404095e-05,
+      "loss": 2.2922,
+      "step": 5020
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962699078646735e-05,
+      "loss": 2.2322,
+      "step": 5025
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962623396839785e-05,
+      "loss": 2.28,
+      "step": 5030
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962547638621262e-05,
+      "loss": 2.2594,
+      "step": 5035
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962471803992334e-05,
+      "loss": 2.2896,
+      "step": 5040
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962395892954168e-05,
+      "loss": 2.2414,
+      "step": 5045
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962319905507934e-05,
+      "loss": 2.1559,
+      "step": 5050
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.9622438416548e-05,
+      "loss": 2.1629,
+      "step": 5055
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962167701395937e-05,
+      "loss": 2.224,
+      "step": 5060
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962091484732521e-05,
+      "loss": 2.2712,
+      "step": 5065
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.962015191665722e-05,
+      "loss": 2.3655,
+      "step": 5070
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961938822196716e-05,
+      "loss": 2.3186,
+      "step": 5075
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961862376326678e-05,
+      "loss": 2.218,
+      "step": 5080
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961785854056788e-05,
+      "loss": 2.1533,
+      "step": 5085
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961709255388221e-05,
+      "loss": 2.2316,
+      "step": 5090
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961632580322156e-05,
+      "loss": 2.2182,
+      "step": 5095
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961555828859777e-05,
+      "loss": 2.2901,
+      "step": 5100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961479001002264e-05,
+      "loss": 2.2504,
+      "step": 5105
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961402096750799e-05,
+      "loss": 2.184,
+      "step": 5110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961325116106566e-05,
+      "loss": 2.2117,
+      "step": 5115
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961248059070751e-05,
+      "loss": 2.1772,
+      "step": 5120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961170925644542e-05,
+      "loss": 2.2296,
+      "step": 5125
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961093715829127e-05,
+      "loss": 2.2211,
+      "step": 5130
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.961016429625691e-05,
+      "loss": 2.3033,
+      "step": 5135
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960939067035426e-05,
+      "loss": 2.1957,
+      "step": 5140
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960861628059524e-05,
+      "loss": 2.21,
+      "step": 5145
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960784112699177e-05,
+      "loss": 2.2984,
+      "step": 5150
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960706520955578e-05,
+      "loss": 2.2722,
+      "step": 5155
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960628852829923e-05,
+      "loss": 2.1871,
+      "step": 5160
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960551108323406e-05,
+      "loss": 2.2718,
+      "step": 5165
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960473287437224e-05,
+      "loss": 2.2753,
+      "step": 5170
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960395390172579e-05,
+      "loss": 2.3343,
+      "step": 5175
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960317416530665e-05,
+      "loss": 2.2347,
+      "step": 5180
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960239366512686e-05,
+      "loss": 2.2149,
+      "step": 5185
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960161240119843e-05,
+      "loss": 2.225,
+      "step": 5190
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.96008303735334e-05,
+      "loss": 2.3047,
+      "step": 5195
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.960004758214379e-05,
+      "loss": 2.229,
+      "step": 5200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.959926402704165e-05,
+      "loss": 2.2022,
+      "step": 5205
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.959847970823908e-05,
+      "loss": 2.2636,
+      "step": 5210
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.959769462574813e-05,
+      "loss": 2.1253,
+      "step": 5215
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.959690877958088e-05,
+      "loss": 2.1823,
+      "step": 5220
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.959612216974946e-05,
+      "loss": 2.1677,
+      "step": 5225
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.959533479626596e-05,
+      "loss": 2.2464,
+      "step": 5230
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.95945466591425e-05,
+      "loss": 2.3058,
+      "step": 5235
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.959375775839124e-05,
+      "loss": 2.2855,
+      "step": 5240
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.95929680940243e-05,
+      "loss": 2.2259,
+      "step": 5245
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.959217766605384e-05,
+      "loss": 2.2208,
+      "step": 5250
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.959138647449205e-05,
+      "loss": 2.2245,
+      "step": 5255
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.95905945193511e-05,
+      "loss": 2.1998,
+      "step": 5260
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958980180064318e-05,
+      "loss": 2.2566,
+      "step": 5265
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958900831838051e-05,
+      "loss": 2.1666,
+      "step": 5270
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958821407257531e-05,
+      "loss": 2.258,
+      "step": 5275
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958741906323978e-05,
+      "loss": 2.2267,
+      "step": 5280
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.95866232903862e-05,
+      "loss": 2.2954,
+      "step": 5285
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958582675402677e-05,
+      "loss": 2.1452,
+      "step": 5290
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958502945417382e-05,
+      "loss": 2.19,
+      "step": 5295
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958423139083958e-05,
+      "loss": 2.2357,
+      "step": 5300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958343256403635e-05,
+      "loss": 2.2354,
+      "step": 5305
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958263297377643e-05,
+      "loss": 2.2369,
+      "step": 5310
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958183262007215e-05,
+      "loss": 2.1878,
+      "step": 5315
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.95810315029358e-05,
+      "loss": 2.181,
+      "step": 5320
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.958022962237974e-05,
+      "loss": 2.2528,
+      "step": 5325
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957942697841632e-05,
+      "loss": 2.2367,
+      "step": 5330
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957862357105785e-05,
+      "loss": 2.1905,
+      "step": 5335
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957781940031676e-05,
+      "loss": 2.2062,
+      "step": 5340
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957701446620543e-05,
+      "loss": 2.2256,
+      "step": 5345
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957620876873621e-05,
+      "loss": 2.234,
+      "step": 5350
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957540230792152e-05,
+      "loss": 2.2726,
+      "step": 5355
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957459508377381e-05,
+      "loss": 2.2848,
+      "step": 5360
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957378709630548e-05,
+      "loss": 2.2896,
+      "step": 5365
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957297834552898e-05,
+      "loss": 2.237,
+      "step": 5370
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957216883145675e-05,
+      "loss": 2.2086,
+      "step": 5375
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957135855410127e-05,
+      "loss": 2.2448,
+      "step": 5380
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.957054751347501e-05,
+      "loss": 2.1829,
+      "step": 5385
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956973570959045e-05,
+      "loss": 2.1896,
+      "step": 5390
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956892314246012e-05,
+      "loss": 2.1687,
+      "step": 5395
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956810981209647e-05,
+      "loss": 2.2002,
+      "step": 5400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956729571851211e-05,
+      "loss": 2.3063,
+      "step": 5405
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956648086171949e-05,
+      "loss": 2.228,
+      "step": 5410
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956566524173121e-05,
+      "loss": 2.2152,
+      "step": 5415
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956484885855982e-05,
+      "loss": 2.1367,
+      "step": 5420
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956403171221786e-05,
+      "loss": 2.1953,
+      "step": 5425
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956321380271796e-05,
+      "loss": 2.2763,
+      "step": 5430
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956239513007268e-05,
+      "loss": 2.244,
+      "step": 5435
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956157569429464e-05,
+      "loss": 2.2184,
+      "step": 5440
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.956075549539643e-05,
+      "loss": 2.2014,
+      "step": 5445
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955993453339072e-05,
+      "loss": 2.2225,
+      "step": 5450
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955911280829011e-05,
+      "loss": 2.2369,
+      "step": 5455
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.95582903201073e-05,
+      "loss": 2.2416,
+      "step": 5460
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955746706885492e-05,
+      "loss": 2.1798,
+      "step": 5465
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955664305454564e-05,
+      "loss": 2.1911,
+      "step": 5470
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955581827719216e-05,
+      "loss": 2.1335,
+      "step": 5475
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955499273680719e-05,
+      "loss": 2.2778,
+      "step": 5480
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955416643340343e-05,
+      "loss": 2.2643,
+      "step": 5485
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.95533393669936e-05,
+      "loss": 2.3167,
+      "step": 5490
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955251153759044e-05,
+      "loss": 2.2957,
+      "step": 5495
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955168294520672e-05,
+      "loss": 2.3075,
+      "step": 5500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955085358985514e-05,
+      "loss": 2.2352,
+      "step": 5505
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.955002347154851e-05,
+      "loss": 2.1775,
+      "step": 5510
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954919259029962e-05,
+      "loss": 2.1675,
+      "step": 5515
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954836094612123e-05,
+      "loss": 2.1308,
+      "step": 5520
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954752853902619e-05,
+      "loss": 2.2289,
+      "step": 5525
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954669536902728e-05,
+      "loss": 2.2759,
+      "step": 5530
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954586143613733e-05,
+      "loss": 2.2623,
+      "step": 5535
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954502674036919e-05,
+      "loss": 2.1587,
+      "step": 5540
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954419128173572e-05,
+      "loss": 2.1853,
+      "step": 5545
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954335506024978e-05,
+      "loss": 2.205,
+      "step": 5550
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954251807592423e-05,
+      "loss": 2.1612,
+      "step": 5555
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954168032877198e-05,
+      "loss": 2.2281,
+      "step": 5560
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.95408418188059e-05,
+      "loss": 2.1966,
+      "step": 5565
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.954000254603894e-05,
+      "loss": 2.2299,
+      "step": 5570
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.9539162510484e-05,
+      "loss": 2.26,
+      "step": 5575
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.953832171215401e-05,
+      "loss": 2.2781,
+      "step": 5580
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.953748015106193e-05,
+      "loss": 2.2335,
+      "step": 5585
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.953663782722069e-05,
+      "loss": 2.2139,
+      "step": 5590
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.953579474064329e-05,
+      "loss": 2.2681,
+      "step": 5595
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.95349508913427e-05,
+      "loss": 2.2618,
+      "step": 5600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.953410627933193e-05,
+      "loss": 2.2938,
+      "step": 5605
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.953326090462397e-05,
+      "loss": 2.2778,
+      "step": 5610
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.953241476723181e-05,
+      "loss": 2.1932,
+      "step": 5615
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.953156786716851e-05,
+      "loss": 2.1645,
+      "step": 5620
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.953072020444711e-05,
+      "loss": 2.2234,
+      "step": 5625
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952987177908066e-05,
+      "loss": 2.2933,
+      "step": 5630
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.95290225910822e-05,
+      "loss": 2.2171,
+      "step": 5635
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952817264046486e-05,
+      "loss": 2.3139,
+      "step": 5640
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952732192724167e-05,
+      "loss": 2.2854,
+      "step": 5645
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952647045142576e-05,
+      "loss": 2.1567,
+      "step": 5650
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952561821303021e-05,
+      "loss": 2.2923,
+      "step": 5655
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952476521206818e-05,
+      "loss": 2.2824,
+      "step": 5660
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952391144855278e-05,
+      "loss": 2.1538,
+      "step": 5665
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952305692249717e-05,
+      "loss": 2.3051,
+      "step": 5670
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952220163391451e-05,
+      "loss": 2.2092,
+      "step": 5675
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952134558281793e-05,
+      "loss": 2.2457,
+      "step": 5680
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.952048876922066e-05,
+      "loss": 2.2462,
+      "step": 5685
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.951963119313588e-05,
+      "loss": 2.2181,
+      "step": 5690
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.951877285457678e-05,
+      "loss": 2.1723,
+      "step": 5695
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.951791375355659e-05,
+      "loss": 2.2092,
+      "step": 5700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.951705389008852e-05,
+      "loss": 2.261,
+      "step": 5705
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.951619326418582e-05,
+      "loss": 2.221,
+      "step": 5710
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.951533187586178e-05,
+      "loss": 2.1836,
+      "step": 5715
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.951446972512958e-05,
+      "loss": 2.2236,
+      "step": 5720
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.951360681200256e-05,
+      "loss": 2.1634,
+      "step": 5725
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.9512743136494e-05,
+      "loss": 2.2121,
+      "step": 5730
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.951187869861715e-05,
+      "loss": 2.2864,
+      "step": 5735
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.951101349838538e-05,
+      "loss": 2.2263,
+      "step": 5740
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.9510147535812e-05,
+      "loss": 2.1629,
+      "step": 5745
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.950928081091032e-05,
+      "loss": 2.1985,
+      "step": 5750
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.95084133236937e-05,
+      "loss": 2.2684,
+      "step": 5755
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.950754507417547e-05,
+      "loss": 2.2916,
+      "step": 5760
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.950667606236906e-05,
+      "loss": 2.2178,
+      "step": 5765
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.95058062882878e-05,
+      "loss": 2.2671,
+      "step": 5770
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.950493575194509e-05,
+      "loss": 2.281,
+      "step": 5775
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.950406445335434e-05,
+      "loss": 2.2875,
+      "step": 5780
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.950319239252896e-05,
+      "loss": 2.2335,
+      "step": 5785
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.95023195694824e-05,
+      "loss": 2.2066,
+      "step": 5790
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.950144598422809e-05,
+      "loss": 2.2195,
+      "step": 5795
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.950057163677944e-05,
+      "loss": 2.2383,
+      "step": 5800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949969652714997e-05,
+      "loss": 2.1299,
+      "step": 5805
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949882065535312e-05,
+      "loss": 2.2448,
+      "step": 5810
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949794402140239e-05,
+      "loss": 2.2574,
+      "step": 5815
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949706662531127e-05,
+      "loss": 2.2185,
+      "step": 5820
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949618846709327e-05,
+      "loss": 2.1732,
+      "step": 5825
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949530954676194e-05,
+      "loss": 2.167,
+      "step": 5830
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949442986433076e-05,
+      "loss": 2.152,
+      "step": 5835
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949354941981332e-05,
+      "loss": 2.1885,
+      "step": 5840
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949266821322315e-05,
+      "loss": 2.1959,
+      "step": 5845
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949178624457384e-05,
+      "loss": 2.1348,
+      "step": 5850
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949090351387895e-05,
+      "loss": 2.1167,
+      "step": 5855
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.949002002115208e-05,
+      "loss": 2.1807,
+      "step": 5860
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948913576640683e-05,
+      "loss": 2.1584,
+      "step": 5865
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948825074965682e-05,
+      "loss": 2.1837,
+      "step": 5870
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948736497091569e-05,
+      "loss": 2.1224,
+      "step": 5875
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948647843019705e-05,
+      "loss": 2.1738,
+      "step": 5880
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948559112751456e-05,
+      "loss": 2.2349,
+      "step": 5885
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948470306288189e-05,
+      "loss": 2.283,
+      "step": 5890
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948381423631273e-05,
+      "loss": 2.2771,
+      "step": 5895
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948292464782072e-05,
+      "loss": 2.2011,
+      "step": 5900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948203429741961e-05,
+      "loss": 2.2963,
+      "step": 5905
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948114318512307e-05,
+      "loss": 2.1404,
+      "step": 5910
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948025131094482e-05,
+      "loss": 2.2242,
+      "step": 5915
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947935867489862e-05,
+      "loss": 2.2917,
+      "step": 5920
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.94784652769982e-05,
+      "loss": 2.2651,
+      "step": 5925
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947757111725732e-05,
+      "loss": 2.2225,
+      "step": 5930
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947667619568974e-05,
+      "loss": 2.2055,
+      "step": 5935
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947578051230925e-05,
+      "loss": 2.2388,
+      "step": 5940
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947488406712962e-05,
+      "loss": 2.2449,
+      "step": 5945
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947398686016469e-05,
+      "loss": 2.1951,
+      "step": 5950
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947308889142824e-05,
+      "loss": 2.1418,
+      "step": 5955
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.94721901609341e-05,
+      "loss": 2.2238,
+      "step": 5960
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947129066869614e-05,
+      "loss": 2.1567,
+      "step": 5965
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947039041472817e-05,
+      "loss": 2.1627,
+      "step": 5970
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.946948939904407e-05,
+      "loss": 2.1972,
+      "step": 5975
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.946858762165772e-05,
+      "loss": 2.299,
+      "step": 5980
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.946768508258299e-05,
+      "loss": 2.243,
+      "step": 5985
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.946678178183377e-05,
+      "loss": 2.1656,
+      "step": 5990
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.946587771942401e-05,
+      "loss": 2.2966,
+      "step": 5995
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.946497289536758e-05,
+      "loss": 2.2317,
+      "step": 6000
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 126689,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 1.7487862362537984e+19,
+  "trial_name": null,
+  "trial_params": null
+}