diff --git "a/text_models/llama-7b/checkpoint-13200/trainer_state.json" "b/text_models/llama-7b/checkpoint-13200/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/text_models/llama-7b/checkpoint-13200/trainer_state.json"
@@ -0,0 +1,8467 @@
+{
+  "best_metric": 1.0564384460449219,
+  "best_model_checkpoint": "checkpoints/alpaca_text_52K/checkpoint-6000",
+  "epoch": 4.223662107031437,
+  "eval_steps": 200,
+  "global_step": 13200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 2.2432,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 2.184,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.5e-05,
+      "loss": 2.1422,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000102,
+      "loss": 1.7323,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 1.5829,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000162,
+      "loss": 1.5178,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019199999999999998,
+      "loss": 1.2123,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00022199999999999998,
+      "loss": 1.1177,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00025199999999999995,
+      "loss": 1.1426,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00028199999999999997,
+      "loss": 1.1881,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999227053140096,
+      "loss": 1.1073,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002997294685990338,
+      "loss": 1.1338,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029953623188405795,
+      "loss": 1.0979,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002993429951690821,
+      "loss": 1.1325,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029914975845410623,
+      "loss": 1.1361,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002989565217391304,
+      "loss": 1.1412,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029876328502415457,
+      "loss": 1.0977,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002985700483091787,
+      "loss": 1.1252,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029837681159420286,
+      "loss": 1.1185,
+      "step": 190
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000298183574879227,
+      "loss": 1.0904,
+      "step": 200
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 1.1103301048278809,
+      "eval_runtime": 299.3056,
+      "eval_samples_per_second": 6.682,
+      "eval_steps_per_second": 0.835,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002979903381642512,
+      "loss": 1.1097,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029779710144927534,
+      "loss": 1.1154,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002976038647342995,
+      "loss": 1.1053,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002974106280193236,
+      "loss": 1.1247,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029721739130434777,
+      "loss": 1.1455,
+      "step": 250
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029702415458937196,
+      "loss": 1.1066,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002968309178743961,
+      "loss": 1.0765,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029663768115942025,
+      "loss": 1.0976,
+      "step": 280
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002964444444444444,
+      "loss": 1.0866,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002962512077294686,
+      "loss": 1.0844,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029605797101449273,
+      "loss": 1.1352,
+      "step": 310
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002958647342995169,
+      "loss": 1.0878,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000295671497584541,
+      "loss": 1.0814,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029547826086956516,
+      "loss": 1.0676,
+      "step": 340
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029528502415458936,
+      "loss": 1.043,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002950917874396135,
+      "loss": 1.0888,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029489855072463764,
+      "loss": 1.0948,
+      "step": 370
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002947053140096618,
+      "loss": 1.0595,
+      "step": 380
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000294512077294686,
+      "loss": 1.1063,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002943188405797101,
+      "loss": 1.0437,
+      "step": 400
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 1.0938334465026855,
+      "eval_runtime": 298.9118,
+      "eval_samples_per_second": 6.691,
+      "eval_steps_per_second": 0.836,
+      "step": 400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029412560386473427,
+      "loss": 1.107,
+      "step": 410
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002939323671497584,
+      "loss": 1.1255,
+      "step": 420
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002937391304347826,
+      "loss": 1.1168,
+      "step": 430
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029354589371980675,
+      "loss": 1.0599,
+      "step": 440
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002933526570048309,
+      "loss": 1.0614,
+      "step": 450
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00029315942028985503,
+      "loss": 1.1127,
+      "step": 460
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002929661835748792,
+      "loss": 1.0629,
+      "step": 470
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00029277294685990337,
+      "loss": 1.0952,
+      "step": 480
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002925797101449275,
+      "loss": 1.1008,
+      "step": 490
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00029238647342995166,
+      "loss": 1.0932,
+      "step": 500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002921932367149758,
+      "loss": 1.0604,
+      "step": 510
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000292,
+      "loss": 1.0925,
+      "step": 520
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00029180676328502414,
+      "loss": 1.0489,
+      "step": 530
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002916135265700483,
+      "loss": 1.1185,
+      "step": 540
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002914202898550724,
+      "loss": 1.0939,
+      "step": 550
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00029122705314009657,
+      "loss": 1.1288,
+      "step": 560
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00029103381642512076,
+      "loss": 1.0368,
+      "step": 570
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002908405797101449,
+      "loss": 1.0762,
+      "step": 580
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00029064734299516905,
+      "loss": 1.0966,
+      "step": 590
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002904541062801932,
+      "loss": 1.0941,
+      "step": 600
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 1.0855311155319214,
+      "eval_runtime": 299.2543,
+      "eval_samples_per_second": 6.683,
+      "eval_steps_per_second": 0.835,
+      "step": 600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002902608695652174,
+      "loss": 1.0732,
+      "step": 610
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00029006763285024153,
+      "loss": 1.0497,
+      "step": 620
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00028987439613526567,
+      "loss": 1.0601,
+      "step": 630
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002896811594202898,
+      "loss": 1.0886,
+      "step": 640
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00028948792270531396,
+      "loss": 1.0681,
+      "step": 650
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00028929468599033815,
+      "loss": 1.0559,
+      "step": 660
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002891014492753623,
+      "loss": 1.0879,
+      "step": 670
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00028890821256038644,
+      "loss": 1.089,
+      "step": 680
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002887149758454106,
+      "loss": 1.1005,
+      "step": 690
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002885217391304348,
+      "loss": 1.1494,
+      "step": 700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002883285024154589,
+      "loss": 1.0847,
+      "step": 710
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00028813526570048306,
+      "loss": 1.0514,
+      "step": 720
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002879420289855072,
+      "loss": 1.1038,
+      "step": 730
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002877487922705314,
+      "loss": 1.0171,
+      "step": 740
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00028755555555555555,
+      "loss": 1.1281,
+      "step": 750
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002873623188405797,
+      "loss": 1.0677,
+      "step": 760
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00028716908212560383,
+      "loss": 1.0963,
+      "step": 770
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.000286975845410628,
+      "loss": 1.1041,
+      "step": 780
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00028678260869565217,
+      "loss": 1.0847,
+      "step": 790
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002865893719806763,
+      "loss": 1.0871,
+      "step": 800
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 1.0817850828170776,
+      "eval_runtime": 298.1366,
+      "eval_samples_per_second": 6.708,
+      "eval_steps_per_second": 0.839,
+      "step": 800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00028639613526570046,
+      "loss": 1.0593,
+      "step": 810
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002862028985507246,
+      "loss": 1.0254,
+      "step": 820
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002860096618357488,
+      "loss": 1.0503,
+      "step": 830
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00028581642512077294,
+      "loss": 1.1017,
+      "step": 840
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002856231884057971,
+      "loss": 1.0658,
+      "step": 850
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002854299516908212,
+      "loss": 1.1069,
+      "step": 860
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00028523671497584537,
+      "loss": 1.0603,
+      "step": 870
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00028504347826086956,
+      "loss": 1.082,
+      "step": 880
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002848502415458937,
+      "loss": 1.1248,
+      "step": 890
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00028465700483091785,
+      "loss": 1.1223,
+      "step": 900
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.000284463768115942,
+      "loss": 1.0672,
+      "step": 910
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002842705314009662,
+      "loss": 1.1095,
+      "step": 920
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00028407729468599033,
+      "loss": 1.0227,
+      "step": 930
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00028388405797101447,
+      "loss": 1.0969,
+      "step": 940
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002836908212560386,
+      "loss": 1.0638,
+      "step": 950
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00028349758454106276,
+      "loss": 1.0892,
+      "step": 960
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00028330434782608695,
+      "loss": 1.0565,
+      "step": 970
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002831111111111111,
+      "loss": 1.1078,
+      "step": 980
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00028291787439613524,
+      "loss": 1.1106,
+      "step": 990
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002827246376811594,
+      "loss": 1.066,
+      "step": 1000
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 1.0781841278076172,
+      "eval_runtime": 298.6342,
+      "eval_samples_per_second": 6.697,
+      "eval_steps_per_second": 0.837,
+      "step": 1000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002825314009661836,
+      "loss": 1.107,
+      "step": 1010
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002823381642512077,
+      "loss": 1.0654,
+      "step": 1020
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00028214492753623186,
+      "loss": 1.0342,
+      "step": 1030
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.000281951690821256,
+      "loss": 1.088,
+      "step": 1040
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002817584541062802,
+      "loss": 1.0871,
+      "step": 1050
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00028156521739130434,
+      "loss": 1.1038,
+      "step": 1060
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002813719806763285,
+      "loss": 1.0912,
+      "step": 1070
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00028117874396135263,
+      "loss": 1.0836,
+      "step": 1080
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00028098550724637677,
+      "loss": 1.0852,
+      "step": 1090
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00028079227053140097,
+      "loss": 1.0757,
+      "step": 1100
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002805990338164251,
+      "loss": 1.0274,
+      "step": 1110
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00028040579710144925,
+      "loss": 1.0815,
+      "step": 1120
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002802125603864734,
+      "loss": 1.0917,
+      "step": 1130
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002800193236714976,
+      "loss": 1.0986,
+      "step": 1140
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00027982608695652174,
+      "loss": 1.0712,
+      "step": 1150
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002796328502415459,
+      "loss": 1.0725,
+      "step": 1160
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00027943961352657,
+      "loss": 1.0358,
+      "step": 1170
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00027924637681159416,
+      "loss": 1.0775,
+      "step": 1180
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00027905314009661836,
+      "loss": 1.083,
+      "step": 1190
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002788599033816425,
+      "loss": 1.0618,
+      "step": 1200
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.0745840072631836,
+      "eval_runtime": 300.795,
+      "eval_samples_per_second": 6.649,
+      "eval_steps_per_second": 0.831,
+      "step": 1200
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00027866666666666665,
+      "loss": 1.0851,
+      "step": 1210
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002784734299516908,
+      "loss": 1.1221,
+      "step": 1220
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.000278280193236715,
+      "loss": 1.1222,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027808695652173913,
+      "loss": 1.0941,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027789371980676327,
+      "loss": 1.0994,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002777004830917874,
+      "loss": 1.0973,
+      "step": 1260
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00027750724637681156,
+      "loss": 0.9984,
+      "step": 1270
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00027731400966183575,
+      "loss": 1.0531,
+      "step": 1280
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002771207729468599,
+      "loss": 1.0872,
+      "step": 1290
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00027692753623188404,
+      "loss": 1.0575,
+      "step": 1300
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002767342995169082,
+      "loss": 1.0515,
+      "step": 1310
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002765410628019324,
+      "loss": 1.0888,
+      "step": 1320
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002763478260869565,
+      "loss": 1.058,
+      "step": 1330
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00027615458937198066,
+      "loss": 1.0355,
+      "step": 1340
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002759613526570048,
+      "loss": 1.0612,
+      "step": 1350
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000275768115942029,
+      "loss": 1.1059,
+      "step": 1360
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00027557487922705314,
+      "loss": 1.0836,
+      "step": 1370
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002753816425120773,
+      "loss": 1.0433,
+      "step": 1380
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00027518840579710143,
+      "loss": 1.1255,
+      "step": 1390
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00027499516908212557,
+      "loss": 1.0494,
+      "step": 1400
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 1.0709148645401,
+      "eval_runtime": 299.6045,
+      "eval_samples_per_second": 6.675,
+      "eval_steps_per_second": 0.834,
+      "step": 1400
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00027480193236714977,
+      "loss": 1.0646,
+      "step": 1410
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002746086956521739,
+      "loss": 1.0492,
+      "step": 1420
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00027441545893719805,
+      "loss": 1.1201,
+      "step": 1430
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002742222222222222,
+      "loss": 1.1151,
+      "step": 1440
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002740289855072464,
+      "loss": 1.0785,
+      "step": 1450
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00027383574879227054,
+      "loss": 1.1167,
+      "step": 1460
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002736425120772947,
+      "loss": 1.0623,
+      "step": 1470
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002734492753623188,
+      "loss": 1.084,
+      "step": 1480
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00027325603864734296,
+      "loss": 1.0931,
+      "step": 1490
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00027306280193236716,
+      "loss": 1.0783,
+      "step": 1500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002728695652173913,
+      "loss": 1.0801,
+      "step": 1510
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00027267632850241544,
+      "loss": 1.0362,
+      "step": 1520
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002724830917874396,
+      "loss": 1.1147,
+      "step": 1530
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002722898550724638,
+      "loss": 1.0659,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002720966183574879,
+      "loss": 1.0629,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00027190338164251207,
+      "loss": 1.0563,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002717101449275362,
+      "loss": 1.0661,
+      "step": 1570
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00027151690821256035,
+      "loss": 1.0722,
+      "step": 1580
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00027132367149758455,
+      "loss": 1.0175,
+      "step": 1590
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002711304347826087,
+      "loss": 1.0829,
+      "step": 1600
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 1.0700575113296509,
+      "eval_runtime": 301.3359,
+      "eval_samples_per_second": 6.637,
+      "eval_steps_per_second": 0.83,
+      "step": 1600
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00027093719806763284,
+      "loss": 1.0935,
+      "step": 1610
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.000270743961352657,
+      "loss": 1.094,
+      "step": 1620
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002705507246376812,
+      "loss": 1.0603,
+      "step": 1630
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002703574879227053,
+      "loss": 1.0813,
+      "step": 1640
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00027016425120772946,
+      "loss": 1.0836,
+      "step": 1650
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0002699710144927536,
+      "loss": 1.0538,
+      "step": 1660
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00026977777777777775,
+      "loss": 1.0892,
+      "step": 1670
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00026958454106280194,
+      "loss": 1.0814,
+      "step": 1680
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002693913043478261,
+      "loss": 1.0407,
+      "step": 1690
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00026919806763285023,
+      "loss": 1.0628,
+      "step": 1700
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00026900483091787437,
+      "loss": 1.0678,
+      "step": 1710
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002688115942028985,
+      "loss": 1.0832,
+      "step": 1720
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002686183574879227,
+      "loss": 1.1199,
+      "step": 1730
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00026842512077294685,
+      "loss": 1.0674,
+      "step": 1740
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.000268231884057971,
+      "loss": 1.0719,
+      "step": 1750
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00026803864734299514,
+      "loss": 1.0754,
+      "step": 1760
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00026784541062801933,
+      "loss": 1.0248,
+      "step": 1770
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002676521739130435,
+      "loss": 1.0659,
+      "step": 1780
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002674589371980676,
+      "loss": 1.0341,
+      "step": 1790
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00026726570048309176,
+      "loss": 1.1131,
+      "step": 1800
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 1.0687577724456787,
+      "eval_runtime": 298.631,
+      "eval_samples_per_second": 6.697,
+      "eval_steps_per_second": 0.837,
+      "step": 1800
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002670724637681159,
+      "loss": 1.0585,
+      "step": 1810
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002668792270531401,
+      "loss": 1.0657,
+      "step": 1820
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00026668599033816424,
+      "loss": 1.0435,
+      "step": 1830
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002664927536231884,
+      "loss": 1.0697,
+      "step": 1840
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00026629951690821253,
+      "loss": 1.086,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002661062801932367,
+      "loss": 1.0853,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00026591304347826087,
+      "loss": 1.0214,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.000265719806763285,
+      "loss": 1.069,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00026552657004830915,
+      "loss": 1.0793,
+      "step": 1890
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002653333333333333,
+      "loss": 1.0631,
+      "step": 1900
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002651400966183575,
+      "loss": 1.0605,
+      "step": 1910
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00026494685990338163,
+      "loss": 1.0322,
+      "step": 1920
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002647536231884058,
+      "loss": 1.0644,
+      "step": 1930
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002645603864734299,
+      "loss": 1.0665,
+      "step": 1940
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00026436714975845406,
+      "loss": 1.1168,
+      "step": 1950
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00026417391304347826,
+      "loss": 1.0518,
+      "step": 1960
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002639806763285024,
+      "loss": 1.0634,
+      "step": 1970
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00026378743961352654,
+      "loss": 1.0483,
+      "step": 1980
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002635942028985507,
+      "loss": 1.0808,
+      "step": 1990
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002634009661835749,
+      "loss": 1.0519,
+      "step": 2000
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 1.0664184093475342,
+      "eval_runtime": 297.6732,
+      "eval_samples_per_second": 6.719,
+      "eval_steps_per_second": 0.84,
+      "step": 2000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.000263207729468599,
+      "loss": 1.0351,
+      "step": 2010
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00026301449275362317,
+      "loss": 1.0342,
+      "step": 2020
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002628212560386473,
+      "loss": 1.029,
+      "step": 2030
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00026262801932367145,
+      "loss": 1.0726,
+      "step": 2040
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00026243478260869565,
+      "loss": 1.0187,
+      "step": 2050
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002622415458937198,
+      "loss": 1.0743,
+      "step": 2060
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00026204830917874394,
+      "loss": 1.0544,
+      "step": 2070
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002618550724637681,
+      "loss": 1.0619,
+      "step": 2080
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002616618357487922,
+      "loss": 1.0642,
+      "step": 2090
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002614685990338164,
+      "loss": 1.0353,
+      "step": 2100
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00026127536231884056,
+      "loss": 1.0435,
+      "step": 2110
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002610821256038647,
+      "loss": 1.0759,
+      "step": 2120
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00026088888888888885,
+      "loss": 1.1161,
+      "step": 2130
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00026069565217391304,
+      "loss": 1.0283,
+      "step": 2140
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002605024154589372,
+      "loss": 1.0889,
+      "step": 2150
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00026030917874396133,
+      "loss": 1.0649,
+      "step": 2160
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00026011594202898547,
+      "loss": 1.0541,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002599227053140096,
+      "loss": 1.086,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002597294685990338,
+      "loss": 1.0836,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00025953623188405795,
+      "loss": 1.0128,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 1.0654979944229126,
+      "eval_runtime": 298.984,
+      "eval_samples_per_second": 6.689,
+      "eval_steps_per_second": 0.836,
+      "step": 2200
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002593429951690821,
+      "loss": 1.0502,
+      "step": 2210
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00025914975845410624,
+      "loss": 1.0275,
+      "step": 2220
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00025895652173913043,
+      "loss": 1.0567,
+      "step": 2230
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002587632850241546,
+      "loss": 1.1105,
+      "step": 2240
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002585700483091787,
+      "loss": 1.0271,
+      "step": 2250
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00025837681159420286,
+      "loss": 1.0471,
+      "step": 2260
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.000258183574879227,
+      "loss": 1.0862,
+      "step": 2270
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002579903381642512,
+      "loss": 1.0829,
+      "step": 2280
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00025779710144927534,
+      "loss": 1.0542,
+      "step": 2290
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002576038647342995,
+      "loss": 1.0632,
+      "step": 2300
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00025741062801932363,
+      "loss": 1.0714,
+      "step": 2310
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00025721739130434777,
+      "loss": 0.9918,
+      "step": 2320
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00025702415458937197,
+      "loss": 1.0651,
+      "step": 2330
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002568309178743961,
+      "loss": 1.0521,
+      "step": 2340
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00025663768115942025,
+      "loss": 1.057,
+      "step": 2350
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002564444444444444,
+      "loss": 1.0675,
+      "step": 2360
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002562512077294686,
+      "loss": 1.1029,
+      "step": 2370
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00025605797101449273,
+      "loss": 1.0636,
+      "step": 2380
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002558647342995169,
+      "loss": 1.0586,
+      "step": 2390
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.000255671497584541,
+      "loss": 1.0517,
+      "step": 2400
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 1.0638346672058105,
+      "eval_runtime": 305.5987,
+      "eval_samples_per_second": 6.545,
+      "eval_steps_per_second": 0.818,
+      "step": 2400
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00025547826086956516,
+      "loss": 1.0955,
+      "step": 2410
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00025528502415458936,
+      "loss": 1.0724,
+      "step": 2420
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002550917874396135,
+      "loss": 1.0682,
+      "step": 2430
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00025489855072463764,
+      "loss": 1.0906,
+      "step": 2440
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002547053140096618,
+      "loss": 1.0637,
+      "step": 2450
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00025451207729468593,
+      "loss": 1.0794,
+      "step": 2460
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002543188405797101,
+      "loss": 1.0554,
+      "step": 2470
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00025412560386473427,
+      "loss": 1.1048,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002539323671497584,
+      "loss": 1.0526,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00025373913043478255,
+      "loss": 1.1208,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00025354589371980675,
+      "loss": 1.058,
+      "step": 2510
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002533526570048309,
+      "loss": 1.0374,
+      "step": 2520
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00025315942028985504,
+      "loss": 1.0768,
+      "step": 2530
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002529661835748792,
+      "loss": 1.0313,
+      "step": 2540
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002527729468599033,
+      "loss": 1.0322,
+      "step": 2550
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002525797101449275,
+      "loss": 1.0753,
+      "step": 2560
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00025238647342995166,
+      "loss": 1.0702,
+      "step": 2570
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002521932367149758,
+      "loss": 1.0889,
+      "step": 2580
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00025199999999999995,
+      "loss": 1.0907,
+      "step": 2590
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00025180676328502414,
+      "loss": 1.0813,
+      "step": 2600
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 1.06317138671875,
+      "eval_runtime": 301.103,
+      "eval_samples_per_second": 6.642,
+      "eval_steps_per_second": 0.83,
+      "step": 2600
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002516135265700483,
+      "loss": 1.0696,
+      "step": 2610
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00025142028985507243,
+      "loss": 1.05,
+      "step": 2620
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00025122705314009657,
+      "loss": 1.049,
+      "step": 2630
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002510338164251207,
+      "loss": 1.0756,
+      "step": 2640
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002508405797101449,
+      "loss": 1.0656,
+      "step": 2650
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00025064734299516905,
+      "loss": 1.0376,
+      "step": 2660
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002504541062801932,
+      "loss": 1.0677,
+      "step": 2670
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00025026086956521734,
+      "loss": 1.0488,
+      "step": 2680
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00025006763285024153,
+      "loss": 1.0751,
+      "step": 2690
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002498743961352657,
+      "loss": 1.054,
+      "step": 2700
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002496811594202898,
+      "loss": 1.0478,
+      "step": 2710
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00024948792270531396,
+      "loss": 1.0872,
+      "step": 2720
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002492946859903381,
+      "loss": 1.0308,
+      "step": 2730
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0002491014492753623,
+      "loss": 1.0349,
+      "step": 2740
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024890821256038644,
+      "loss": 1.0775,
+      "step": 2750
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0002487149758454106,
+      "loss": 1.1327,
+      "step": 2760
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00024852173913043473,
+      "loss": 1.0442,
+      "step": 2770
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0002483285024154589,
+      "loss": 1.0428,
+      "step": 2780
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00024813526570048307,
+      "loss": 1.0661,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002479420289855072,
+      "loss": 1.1036,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 1.0611621141433716,
+      "eval_runtime": 303.5211,
+      "eval_samples_per_second": 6.589,
+      "eval_steps_per_second": 0.824,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00024774879227053135,
+      "loss": 1.0631,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002475555555555555,
+      "loss": 1.0571,
+      "step": 2820
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002473623188405797,
+      "loss": 1.0389,
+      "step": 2830
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00024716908212560383,
+      "loss": 1.0354,
+      "step": 2840
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.000246975845410628,
+      "loss": 1.0385,
+      "step": 2850
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002467826086956521,
+      "loss": 1.0689,
+      "step": 2860
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002465893719806763,
+      "loss": 1.0414,
+      "step": 2870
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00024639613526570046,
+      "loss": 1.069,
+      "step": 2880
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002462028985507246,
+      "loss": 1.0096,
+      "step": 2890
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00024600966183574874,
+      "loss": 1.0859,
+      "step": 2900
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00024581642512077294,
+      "loss": 1.0518,
+      "step": 2910
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002456231884057971,
+      "loss": 1.0696,
+      "step": 2920
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002454299516908212,
+      "loss": 1.0559,
+      "step": 2930
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00024523671497584537,
+      "loss": 1.0453,
+      "step": 2940
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002450434782608695,
+      "loss": 1.0325,
+      "step": 2950
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002448502415458937,
+      "loss": 1.1118,
+      "step": 2960
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00024465700483091785,
+      "loss": 1.0666,
+      "step": 2970
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.000244463768115942,
+      "loss": 1.0467,
+      "step": 2980
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00024427053140096614,
+      "loss": 1.0605,
+      "step": 2990
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00024407729468599033,
+      "loss": 1.0787,
+      "step": 3000
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 1.0607348680496216,
+      "eval_runtime": 301.6471,
+      "eval_samples_per_second": 6.63,
+      "eval_steps_per_second": 0.829,
+      "step": 3000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00024388405797101448,
+      "loss": 1.0407,
+      "step": 3010
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00024369082125603862,
+      "loss": 1.0713,
+      "step": 3020
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00024349758454106276,
+      "loss": 1.0489,
+      "step": 3030
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00024330434782608693,
+      "loss": 1.0443,
+      "step": 3040
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0002431111111111111,
+      "loss": 1.0606,
+      "step": 3050
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00024291787439613524,
+      "loss": 1.0929,
+      "step": 3060
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00024272463768115938,
+      "loss": 1.0656,
+      "step": 3070
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00024253140096618355,
+      "loss": 1.0528,
+      "step": 3080
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00024233816425120772,
+      "loss": 1.1073,
+      "step": 3090
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00024214492753623187,
+      "loss": 1.0326,
+      "step": 3100
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.000241951690821256,
+      "loss": 1.0446,
+      "step": 3110
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00024175845410628015,
+      "loss": 1.0253,
+      "step": 3120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00024156521739130432,
+      "loss": 1.0186,
+      "step": 3130
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0002413719806763285,
+      "loss": 1.0226,
+      "step": 3140
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00024117874396135263,
+      "loss": 0.9921,
+      "step": 3150
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00024098550724637678,
+      "loss": 0.9769,
+      "step": 3160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00024079227053140095,
+      "loss": 1.0191,
+      "step": 3170
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00024059903381642512,
+      "loss": 1.0193,
+      "step": 3180
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00024040579710144926,
+      "loss": 1.0023,
+      "step": 3190
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0002402125603864734,
+      "loss": 1.0317,
+      "step": 3200
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 1.0621122121810913,
+      "eval_runtime": 301.2241,
+      "eval_samples_per_second": 6.64,
+      "eval_steps_per_second": 0.83,
+      "step": 3200
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00024001932367149754,
+      "loss": 1.0001,
+      "step": 3210
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00023982608695652174,
+      "loss": 0.9865,
+      "step": 3220
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00023963285024154588,
+      "loss": 0.9851,
+      "step": 3230
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00023943961352657002,
+      "loss": 1.017,
+      "step": 3240
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00023924637681159417,
+      "loss": 1.0149,
+      "step": 3250
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0002390531400966183,
+      "loss": 0.9958,
+      "step": 3260
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0002388599033816425,
+      "loss": 1.0346,
+      "step": 3270
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00023866666666666665,
+      "loss": 1.0612,
+      "step": 3280
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0002384734299516908,
+      "loss": 0.9862,
+      "step": 3290
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00023828019323671493,
+      "loss": 0.9982,
+      "step": 3300
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00023808695652173913,
+      "loss": 1.046,
+      "step": 3310
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00023789371980676327,
+      "loss": 1.0176,
+      "step": 3320
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00023770048309178742,
+      "loss": 0.9974,
+      "step": 3330
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00023750724637681156,
+      "loss": 1.0194,
+      "step": 3340
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0002373140096618357,
+      "loss": 1.0272,
+      "step": 3350
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0002371207729468599,
+      "loss": 1.0707,
+      "step": 3360
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00023692753623188404,
+      "loss": 0.9941,
+      "step": 3370
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00023673429951690818,
+      "loss": 1.0298,
+      "step": 3380
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00023654106280193233,
+      "loss": 1.0192,
+      "step": 3390
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00023634782608695652,
+      "loss": 0.99,
+      "step": 3400
+    },
+    {
+      "epoch": 1.09,
+      "eval_loss": 1.0633840560913086,
+      "eval_runtime": 301.1449,
+      "eval_samples_per_second": 6.641,
+      "eval_steps_per_second": 0.83,
+      "step": 3400
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00023615458937198067,
+      "loss": 1.0316,
+      "step": 3410
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0002359613526570048,
+      "loss": 0.9605,
+      "step": 3420
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00023576811594202895,
+      "loss": 0.9902,
+      "step": 3430
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002355748792270531,
+      "loss": 0.99,
+      "step": 3440
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002353816425120773,
+      "loss": 0.9923,
+      "step": 3450
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00023518840579710143,
+      "loss": 0.993,
+      "step": 3460
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00023499516908212557,
+      "loss": 1.0103,
+      "step": 3470
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00023480193236714972,
+      "loss": 1.0536,
+      "step": 3480
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00023460869565217391,
+      "loss": 0.9961,
+      "step": 3490
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00023441545893719806,
+      "loss": 1.0461,
+      "step": 3500
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0002342222222222222,
+      "loss": 0.9959,
+      "step": 3510
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00023402898550724634,
+      "loss": 0.9932,
+      "step": 3520
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00023383574879227054,
+      "loss": 1.0258,
+      "step": 3530
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00023364251207729468,
+      "loss": 1.0591,
+      "step": 3540
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00023344927536231882,
+      "loss": 1.0672,
+      "step": 3550
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00023325603864734297,
+      "loss": 0.9972,
+      "step": 3560
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0002330628019323671,
+      "loss": 1.0063,
+      "step": 3570
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002328695652173913,
+      "loss": 1.0153,
+      "step": 3580
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00023267632850241545,
+      "loss": 1.0611,
+      "step": 3590
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002324830917874396,
+      "loss": 1.0215,
+      "step": 3600
+    },
+    {
+      "epoch": 1.15,
+      "eval_loss": 1.0626596212387085,
+      "eval_runtime": 302.4484,
+      "eval_samples_per_second": 6.613,
+      "eval_steps_per_second": 0.827,
+      "step": 3600
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00023228985507246373,
+      "loss": 1.0118,
+      "step": 3610
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00023209661835748793,
+      "loss": 1.0626,
+      "step": 3620
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00023190338164251207,
+      "loss": 1.0051,
+      "step": 3630
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00023171014492753622,
+      "loss": 0.9824,
+      "step": 3640
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00023151690821256036,
+      "loss": 1.0567,
+      "step": 3650
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0002313236714975845,
+      "loss": 1.015,
+      "step": 3660
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0002311304347826087,
+      "loss": 1.0129,
+      "step": 3670
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00023093719806763284,
+      "loss": 0.9826,
+      "step": 3680
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00023074396135265698,
+      "loss": 1.0257,
+      "step": 3690
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00023055072463768112,
+      "loss": 1.0426,
+      "step": 3700
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00023035748792270532,
+      "loss": 1.0054,
+      "step": 3710
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00023016425120772946,
+      "loss": 1.0168,
+      "step": 3720
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0002299710144927536,
+      "loss": 1.0,
+      "step": 3730
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00022977777777777775,
+      "loss": 1.0138,
+      "step": 3740
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0002295845410628019,
+      "loss": 1.0339,
+      "step": 3750
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0002293913043478261,
+      "loss": 0.9986,
+      "step": 3760
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00022919806763285023,
+      "loss": 1.0171,
+      "step": 3770
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00022900483091787437,
+      "loss": 0.9846,
+      "step": 3780
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00022881159420289852,
+      "loss": 1.0338,
+      "step": 3790
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00022861835748792269,
+      "loss": 1.0161,
+      "step": 3800
+    },
+    {
+      "epoch": 1.22,
+      "eval_loss": 1.0625114440917969,
+      "eval_runtime": 301.9731,
+      "eval_samples_per_second": 6.623,
+      "eval_steps_per_second": 0.828,
+      "step": 3800
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00022842512077294686,
+      "loss": 1.0253,
+      "step": 3810
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.000228231884057971,
+      "loss": 1.0187,
+      "step": 3820
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00022803864734299514,
+      "loss": 0.9766,
+      "step": 3830
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0002278454106280193,
+      "loss": 1.0336,
+      "step": 3840
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00022765217391304348,
+      "loss": 1.0084,
+      "step": 3850
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00022745893719806762,
+      "loss": 1.0152,
+      "step": 3860
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00022726570048309177,
+      "loss": 0.9628,
+      "step": 3870
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0002270724637681159,
+      "loss": 0.9494,
+      "step": 3880
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00022687922705314008,
+      "loss": 1.0688,
+      "step": 3890
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00022668599033816425,
+      "loss": 1.0242,
+      "step": 3900
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0002264927536231884,
+      "loss": 1.0518,
+      "step": 3910
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00022629951690821253,
+      "loss": 0.9777,
+      "step": 3920
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0002261062801932367,
+      "loss": 0.9929,
+      "step": 3930
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00022591304347826084,
+      "loss": 1.0192,
+      "step": 3940
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00022571980676328501,
+      "loss": 0.9951,
+      "step": 3950
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00022552657004830916,
+      "loss": 1.0014,
+      "step": 3960
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0002253333333333333,
+      "loss": 1.035,
+      "step": 3970
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00022514009661835747,
+      "loss": 0.9909,
+      "step": 3980
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00022494685990338164,
+      "loss": 1.0372,
+      "step": 3990
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00022475362318840578,
+      "loss": 0.992,
+      "step": 4000
+    },
+    {
+      "epoch": 1.28,
+      "eval_loss": 1.0622905492782593,
+      "eval_runtime": 303.6073,
+      "eval_samples_per_second": 6.587,
+      "eval_steps_per_second": 0.823,
+      "step": 4000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00022456038647342992,
+      "loss": 1.0224,
+      "step": 4010
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0002243671497584541,
+      "loss": 1.0159,
+      "step": 4020
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00022417391304347824,
+      "loss": 1.0257,
+      "step": 4030
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0002239806763285024,
+      "loss": 1.0391,
+      "step": 4040
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00022378743961352655,
+      "loss": 0.9996,
+      "step": 4050
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0002235942028985507,
+      "loss": 1.0807,
+      "step": 4060
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00022340096618357486,
+      "loss": 0.975,
+      "step": 4070
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00022320772946859903,
+      "loss": 1.0069,
+      "step": 4080
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00022301449275362317,
+      "loss": 0.9844,
+      "step": 4090
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00022282125603864732,
+      "loss": 1.0336,
+      "step": 4100
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00022262801932367148,
+      "loss": 1.0124,
+      "step": 4110
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00022243478260869563,
+      "loss": 0.9849,
+      "step": 4120
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0002222415458937198,
+      "loss": 1.0065,
+      "step": 4130
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00022204830917874394,
+      "loss": 1.0319,
+      "step": 4140
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00022185507246376808,
+      "loss": 1.0118,
+      "step": 4150
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00022166183574879225,
+      "loss": 1.0255,
+      "step": 4160
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0002214685990338164,
+      "loss": 0.9649,
+      "step": 4170
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00022127536231884056,
+      "loss": 0.9756,
+      "step": 4180
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0002210821256038647,
+      "loss": 0.9827,
+      "step": 4190
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00022088888888888888,
+      "loss": 1.0245,
+      "step": 4200
+    },
+    {
+      "epoch": 1.34,
+      "eval_loss": 1.0623623132705688,
+      "eval_runtime": 303.2092,
+      "eval_samples_per_second": 6.596,
+      "eval_steps_per_second": 0.825,
+      "step": 4200
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00022069565217391302,
+      "loss": 1.0284,
+      "step": 4210
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0002205024154589372,
+      "loss": 1.0258,
+      "step": 4220
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00022030917874396133,
+      "loss": 1.0665,
+      "step": 4230
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0002201159420289855,
+      "loss": 0.9961,
+      "step": 4240
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00021992270531400964,
+      "loss": 1.0137,
+      "step": 4250
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00021972946859903379,
+      "loss": 0.9965,
+      "step": 4260
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00021953623188405796,
+      "loss": 1.0181,
+      "step": 4270
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0002193429951690821,
+      "loss": 1.0047,
+      "step": 4280
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00021914975845410627,
+      "loss": 1.0193,
+      "step": 4290
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0002189565217391304,
+      "loss": 1.011,
+      "step": 4300
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00021876328502415458,
+      "loss": 1.0327,
+      "step": 4310
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00021857004830917872,
+      "loss": 1.0434,
+      "step": 4320
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0002183768115942029,
+      "loss": 1.0138,
+      "step": 4330
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00021818357487922703,
+      "loss": 0.9737,
+      "step": 4340
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00021799033816425118,
+      "loss": 0.9977,
+      "step": 4350
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00021779710144927535,
+      "loss": 1.0181,
+      "step": 4360
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0002176038647342995,
+      "loss": 1.0434,
+      "step": 4370
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00021741062801932366,
+      "loss": 1.0143,
+      "step": 4380
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0002172173913043478,
+      "loss": 0.9786,
+      "step": 4390
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00021702415458937194,
+      "loss": 0.9861,
+      "step": 4400
+    },
+    {
+      "epoch": 1.41,
+      "eval_loss": 1.0632884502410889,
+      "eval_runtime": 302.3081,
+      "eval_samples_per_second": 6.616,
+      "eval_steps_per_second": 0.827,
+      "step": 4400
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00021683091787439611,
+      "loss": 1.0298,
+      "step": 4410
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00021663768115942028,
+      "loss": 1.0203,
+      "step": 4420
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00021644444444444443,
+      "loss": 1.0476,
+      "step": 4430
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00021625120772946857,
+      "loss": 1.0071,
+      "step": 4440
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00021605797101449274,
+      "loss": 1.0125,
+      "step": 4450
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00021586473429951688,
+      "loss": 1.0706,
+      "step": 4460
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00021567149758454105,
+      "loss": 0.9954,
+      "step": 4470
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0002154782608695652,
+      "loss": 1.0229,
+      "step": 4480
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00021528502415458934,
+      "loss": 0.9828,
+      "step": 4490
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0002150917874396135,
+      "loss": 1.0033,
+      "step": 4500
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00021489855072463767,
+      "loss": 1.0132,
+      "step": 4510
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00021470531400966182,
+      "loss": 1.0256,
+      "step": 4520
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00021451207729468596,
+      "loss": 1.0227,
+      "step": 4530
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0002143188405797101,
+      "loss": 1.0434,
+      "step": 4540
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0002141256038647343,
+      "loss": 1.0596,
+      "step": 4550
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00021393236714975844,
+      "loss": 0.9893,
+      "step": 4560
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00021373913043478258,
+      "loss": 1.0225,
+      "step": 4570
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00021354589371980673,
+      "loss": 0.981,
+      "step": 4580
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0002133526570048309,
+      "loss": 1.0348,
+      "step": 4590
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00021315942028985507,
+      "loss": 1.0208,
+      "step": 4600
+    },
+    {
+      "epoch": 1.47,
+      "eval_loss": 1.0610480308532715,
+      "eval_runtime": 302.534,
+      "eval_samples_per_second": 6.611,
+      "eval_steps_per_second": 0.826,
+      "step": 4600
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0002129661835748792,
+      "loss": 1.0074,
+      "step": 4610
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00021277294685990335,
+      "loss": 1.0173,
+      "step": 4620
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0002125797101449275,
+      "loss": 0.9835,
+      "step": 4630
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0002123864734299517,
+      "loss": 0.9819,
+      "step": 4640
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00021219323671497583,
+      "loss": 1.0399,
+      "step": 4650
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00021199999999999998,
+      "loss": 1.0076,
+      "step": 4660
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00021180676328502412,
+      "loss": 1.0235,
+      "step": 4670
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0002116135265700483,
+      "loss": 1.0269,
+      "step": 4680
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00021142028985507246,
+      "loss": 1.0428,
+      "step": 4690
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0002112270531400966,
+      "loss": 1.0019,
+      "step": 4700
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00021103381642512074,
+      "loss": 0.9701,
+      "step": 4710
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00021084057971014489,
+      "loss": 0.977,
+      "step": 4720
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00021064734299516908,
+      "loss": 0.9753,
+      "step": 4730
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00021045410628019322,
+      "loss": 1.0432,
+      "step": 4740
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00021026086956521737,
+      "loss": 1.051,
+      "step": 4750
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0002100676328502415,
+      "loss": 1.0301,
+      "step": 4760
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00020987439613526565,
+      "loss": 0.9793,
+      "step": 4770
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00020968115942028985,
+      "loss": 1.0151,
+      "step": 4780
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.000209487922705314,
+      "loss": 0.9979,
+      "step": 4790
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00020929468599033813,
+      "loss": 1.0126,
+      "step": 4800
+    },
+    {
+      "epoch": 1.54,
+      "eval_loss": 1.0597844123840332,
+      "eval_runtime": 303.6517,
+      "eval_samples_per_second": 6.586,
+      "eval_steps_per_second": 0.823,
+      "step": 4800
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00020910144927536228,
+      "loss": 1.0428,
+      "step": 4810
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00020890821256038647,
+      "loss": 1.0049,
+      "step": 4820
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00020871497584541062,
+      "loss": 1.0006,
+      "step": 4830
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00020852173913043476,
+      "loss": 1.022,
+      "step": 4840
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.0002083285024154589,
+      "loss": 1.0342,
+      "step": 4850
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0002081352657004831,
+      "loss": 1.0146,
+      "step": 4860
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00020794202898550724,
+      "loss": 1.029,
+      "step": 4870
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00020774879227053138,
+      "loss": 1.0381,
+      "step": 4880
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00020755555555555553,
+      "loss": 1.0372,
+      "step": 4890
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00020736231884057967,
+      "loss": 1.0119,
+      "step": 4900
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00020716908212560387,
+      "loss": 1.0797,
+      "step": 4910
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.000206975845410628,
+      "loss": 0.95,
+      "step": 4920
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00020678260869565215,
+      "loss": 1.0012,
+      "step": 4930
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0002065893719806763,
+      "loss": 1.0194,
+      "step": 4940
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0002063961352657005,
+      "loss": 1.0135,
+      "step": 4950
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00020620289855072463,
+      "loss": 1.0294,
+      "step": 4960
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00020600966183574877,
+      "loss": 0.9954,
+      "step": 4970
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00020581642512077292,
+      "loss": 1.0082,
+      "step": 4980
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00020562318840579706,
+      "loss": 1.0343,
+      "step": 4990
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00020542995169082126,
+      "loss": 0.9967,
+      "step": 5000
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 1.060504674911499,
+      "eval_runtime": 303.6571,
+      "eval_samples_per_second": 6.586,
+      "eval_steps_per_second": 0.823,
+      "step": 5000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0002052367149758454,
+      "loss": 1.0436,
+      "step": 5010
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00020504347826086954,
+      "loss": 1.037,
+      "step": 5020
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00020485024154589368,
+      "loss": 1.0613,
+      "step": 5030
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00020465700483091788,
+      "loss": 1.0305,
+      "step": 5040
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00020446376811594202,
+      "loss": 1.0585,
+      "step": 5050
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00020427053140096617,
+      "loss": 1.0071,
+      "step": 5060
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0002040772946859903,
+      "loss": 0.9929,
+      "step": 5070
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00020388405797101445,
+      "loss": 1.051,
+      "step": 5080
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00020369082125603865,
+      "loss": 0.9953,
+      "step": 5090
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0002034975845410628,
+      "loss": 1.0176,
+      "step": 5100
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00020330434782608693,
+      "loss": 1.0107,
+      "step": 5110
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00020311111111111108,
+      "loss": 1.0618,
+      "step": 5120
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00020291787439613527,
+      "loss": 1.0147,
+      "step": 5130
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00020272463768115942,
+      "loss": 0.9754,
+      "step": 5140
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00020253140096618356,
+      "loss": 1.0381,
+      "step": 5150
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0002023381642512077,
+      "loss": 0.9645,
+      "step": 5160
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00020214492753623187,
+      "loss": 0.9645,
+      "step": 5170
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00020195169082125604,
+      "loss": 1.0371,
+      "step": 5180
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00020175845410628018,
+      "loss": 1.0058,
+      "step": 5190
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00020156521739130432,
+      "loss": 1.02,
+      "step": 5200
+    },
+    {
+      "epoch": 1.66,
+      "eval_loss": 1.0592375993728638,
+      "eval_runtime": 303.4125,
+      "eval_samples_per_second": 6.592,
+      "eval_steps_per_second": 0.824,
+      "step": 5200
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00020137198067632847,
+      "loss": 1.0409,
+      "step": 5210
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00020117874396135266,
+      "loss": 1.0193,
+      "step": 5220
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0002009855072463768,
+      "loss": 1.0169,
+      "step": 5230
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00020079227053140095,
+      "loss": 1.0484,
+      "step": 5240
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0002005990338164251,
+      "loss": 0.9895,
+      "step": 5250
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00020040579710144926,
+      "loss": 0.9858,
+      "step": 5260
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00020021256038647343,
+      "loss": 0.9993,
+      "step": 5270
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00020001932367149757,
+      "loss": 0.9755,
+      "step": 5280
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00019982608695652172,
+      "loss": 1.0392,
+      "step": 5290
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00019963285024154586,
+      "loss": 1.0123,
+      "step": 5300
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00019943961352657003,
+      "loss": 1.0386,
+      "step": 5310
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0001992463768115942,
+      "loss": 1.0359,
+      "step": 5320
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00019905314009661834,
+      "loss": 1.0003,
+      "step": 5330
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00019885990338164248,
+      "loss": 0.9989,
+      "step": 5340
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00019866666666666665,
+      "loss": 0.9515,
+      "step": 5350
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00019847342995169082,
+      "loss": 1.0035,
+      "step": 5360
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00019828019323671497,
+      "loss": 1.0219,
+      "step": 5370
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.0001980869565217391,
+      "loss": 1.0096,
+      "step": 5380
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00019789371980676325,
+      "loss": 1.0077,
+      "step": 5390
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00019770048309178742,
+      "loss": 1.0298,
+      "step": 5400
+    },
+    {
+      "epoch": 1.73,
+      "eval_loss": 1.0586228370666504,
+      "eval_runtime": 302.8462,
+      "eval_samples_per_second": 6.604,
+      "eval_steps_per_second": 0.826,
+      "step": 5400
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0001975072463768116,
+      "loss": 1.0001,
+      "step": 5410
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00019731400966183573,
+      "loss": 1.0465,
+      "step": 5420
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00019712077294685987,
+      "loss": 1.0265,
+      "step": 5430
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00019692753623188404,
+      "loss": 1.0148,
+      "step": 5440
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0001967342995169082,
+      "loss": 0.9763,
+      "step": 5450
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00019654106280193236,
+      "loss": 0.9861,
+      "step": 5460
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.0001963478260869565,
+      "loss": 0.9877,
+      "step": 5470
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00019615458937198067,
+      "loss": 0.9489,
+      "step": 5480
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0001959613526570048,
+      "loss": 1.0694,
+      "step": 5490
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00019576811594202898,
+      "loss": 1.0311,
+      "step": 5500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00019557487922705312,
+      "loss": 1.0494,
+      "step": 5510
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00019538164251207727,
+      "loss": 0.9967,
+      "step": 5520
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00019518840579710144,
+      "loss": 1.0278,
+      "step": 5530
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00019499516908212558,
+      "loss": 1.0384,
+      "step": 5540
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00019480193236714975,
+      "loss": 1.0051,
+      "step": 5550
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0001946086956521739,
+      "loss": 1.0387,
+      "step": 5560
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00019441545893719806,
+      "loss": 1.0094,
+      "step": 5570
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0001942222222222222,
+      "loss": 1.0155,
+      "step": 5580
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00019402898550724637,
+      "loss": 1.0104,
+      "step": 5590
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00019383574879227052,
+      "loss": 1.0264,
+      "step": 5600
+    },
+    {
+      "epoch": 1.79,
+      "eval_loss": 1.058441162109375,
+      "eval_runtime": 301.1664,
+      "eval_samples_per_second": 6.641,
+      "eval_steps_per_second": 0.83,
+      "step": 5600
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00019364251207729466,
+      "loss": 0.9823,
+      "step": 5610
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00019344927536231883,
+      "loss": 0.9871,
+      "step": 5620
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00019325603864734297,
+      "loss": 1.022,
+      "step": 5630
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00019306280193236714,
+      "loss": 1.0128,
+      "step": 5640
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00019286956521739128,
+      "loss": 1.0502,
+      "step": 5650
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00019267632850241545,
+      "loss": 1.0427,
+      "step": 5660
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0001924830917874396,
+      "loss": 1.0088,
+      "step": 5670
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00019228985507246374,
+      "loss": 1.0319,
+      "step": 5680
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0001920966183574879,
+      "loss": 1.02,
+      "step": 5690
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00019190338164251205,
+      "loss": 1.0376,
+      "step": 5700
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00019171014492753622,
+      "loss": 1.0233,
+      "step": 5710
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00019151690821256036,
+      "loss": 1.0285,
+      "step": 5720
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00019132367149758453,
+      "loss": 1.0579,
+      "step": 5730
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00019113043478260867,
+      "loss": 1.0542,
+      "step": 5740
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00019093719806763284,
+      "loss": 1.0427,
+      "step": 5750
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00019074396135265699,
+      "loss": 0.9983,
+      "step": 5760
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00019055072463768113,
+      "loss": 1.0145,
+      "step": 5770
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.0001903574879227053,
+      "loss": 1.0471,
+      "step": 5780
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00019016425120772947,
+      "loss": 1.0436,
+      "step": 5790
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0001899710144927536,
+      "loss": 1.0312,
+      "step": 5800
+    },
+    {
+      "epoch": 1.86,
+      "eval_loss": 1.056977391242981,
+      "eval_runtime": 105.8988,
+      "eval_samples_per_second": 18.886,
+      "eval_steps_per_second": 2.361,
+      "step": 5800
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00018977777777777775,
+      "loss": 0.9989,
+      "step": 5810
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00018958454106280192,
+      "loss": 1.0022,
+      "step": 5820
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00018939130434782606,
+      "loss": 1.0295,
+      "step": 5830
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00018919806763285023,
+      "loss": 1.0644,
+      "step": 5840
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00018900483091787438,
+      "loss": 1.0238,
+      "step": 5850
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00018881159420289852,
+      "loss": 1.06,
+      "step": 5860
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0001886183574879227,
+      "loss": 0.9779,
+      "step": 5870
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00018842512077294686,
+      "loss": 1.0072,
+      "step": 5880
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.000188231884057971,
+      "loss": 1.0261,
+      "step": 5890
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00018803864734299514,
+      "loss": 1.0632,
+      "step": 5900
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.0001878454106280193,
+      "loss": 0.9929,
+      "step": 5910
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00018765217391304346,
+      "loss": 0.9593,
+      "step": 5920
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00018745893719806763,
+      "loss": 1.051,
+      "step": 5930
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00018726570048309177,
+      "loss": 1.0117,
+      "step": 5940
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0001870724637681159,
+      "loss": 1.041,
+      "step": 5950
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00018687922705314008,
+      "loss": 0.9918,
+      "step": 5960
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00018668599033816425,
+      "loss": 1.0187,
+      "step": 5970
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.0001864927536231884,
+      "loss": 1.0205,
+      "step": 5980
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00018629951690821254,
+      "loss": 1.0267,
+      "step": 5990
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00018610628019323668,
+      "loss": 1.0381,
+      "step": 6000
+    },
+    {
+      "epoch": 1.92,
+      "eval_loss": 1.0564384460449219,
+      "eval_runtime": 105.9283,
+      "eval_samples_per_second": 18.881,
+      "eval_steps_per_second": 2.36,
+      "step": 6000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00018591304347826085,
+      "loss": 1.0032,
+      "step": 6010
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00018571980676328502,
+      "loss": 1.0641,
+      "step": 6020
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00018552657004830916,
+      "loss": 1.0396,
+      "step": 6030
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.0001853333333333333,
+      "loss": 1.0209,
+      "step": 6040
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00018514009661835745,
+      "loss": 1.0351,
+      "step": 6050
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00018494685990338164,
+      "loss": 1.0288,
+      "step": 6060
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00018475362318840578,
+      "loss": 0.9853,
+      "step": 6070
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00018456038647342993,
+      "loss": 1.0462,
+      "step": 6080
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00018436714975845407,
+      "loss": 0.964,
+      "step": 6090
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00018417391304347827,
+      "loss": 0.9786,
+      "step": 6100
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0001839806763285024,
+      "loss": 1.0175,
+      "step": 6110
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00018378743961352655,
+      "loss": 0.9656,
+      "step": 6120
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0001835942028985507,
+      "loss": 1.0188,
+      "step": 6130
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00018340096618357484,
+      "loss": 1.0321,
+      "step": 6140
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00018320772946859903,
+      "loss": 1.0227,
+      "step": 6150
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00018301449275362318,
+      "loss": 1.0134,
+      "step": 6160
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00018282125603864732,
+      "loss": 0.9888,
+      "step": 6170
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00018262801932367146,
+      "loss": 1.0341,
+      "step": 6180
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00018243478260869566,
+      "loss": 1.0129,
+      "step": 6190
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0001822415458937198,
+      "loss": 0.9857,
+      "step": 6200
+    },
+    {
+      "epoch": 1.98,
+      "eval_loss": 1.057225227355957,
+      "eval_runtime": 105.9015,
+      "eval_samples_per_second": 18.885,
+      "eval_steps_per_second": 2.361,
+      "step": 6200
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00018204830917874394,
+      "loss": 0.9737,
+      "step": 6210
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00018185507246376809,
+      "loss": 1.0512,
+      "step": 6220
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00018166183574879223,
+      "loss": 1.0394,
+      "step": 6230
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.00018146859903381642,
+      "loss": 1.0026,
+      "step": 6240
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.00018127536231884057,
+      "loss": 1.0175,
+      "step": 6250
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0001810821256038647,
+      "loss": 0.9493,
+      "step": 6260
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00018088888888888885,
+      "loss": 0.9625,
+      "step": 6270
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00018069565217391305,
+      "loss": 0.9273,
+      "step": 6280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.0001805024154589372,
+      "loss": 0.9381,
+      "step": 6290
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00018030917874396133,
+      "loss": 0.9485,
+      "step": 6300
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00018011594202898548,
+      "loss": 0.9448,
+      "step": 6310
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00017992270531400962,
+      "loss": 0.9574,
+      "step": 6320
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.00017972946859903382,
+      "loss": 0.9652,
+      "step": 6330
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.00017953623188405796,
+      "loss": 0.9474,
+      "step": 6340
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0001793429951690821,
+      "loss": 0.9497,
+      "step": 6350
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00017914975845410624,
+      "loss": 0.9024,
+      "step": 6360
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00017895652173913044,
+      "loss": 0.9315,
+      "step": 6370
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00017876328502415458,
+      "loss": 0.9413,
+      "step": 6380
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00017857004830917873,
+      "loss": 1.0023,
+      "step": 6390
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.00017837681159420287,
+      "loss": 0.948,
+      "step": 6400
+    },
+    {
+      "epoch": 2.05,
+      "eval_loss": 1.0676686763763428,
+      "eval_runtime": 105.8497,
+      "eval_samples_per_second": 18.895,
+      "eval_steps_per_second": 2.362,
+      "step": 6400
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.00017818357487922707,
+      "loss": 0.9117,
+      "step": 6410
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.0001779903381642512,
+      "loss": 0.9872,
+      "step": 6420
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.00017779710144927535,
+      "loss": 0.9494,
+      "step": 6430
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0001776038647342995,
+      "loss": 0.9379,
+      "step": 6440
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.00017741062801932364,
+      "loss": 0.9193,
+      "step": 6450
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00017721739130434783,
+      "loss": 0.8931,
+      "step": 6460
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00017702415458937197,
+      "loss": 0.9422,
+      "step": 6470
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00017683091787439612,
+      "loss": 0.9522,
+      "step": 6480
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.00017663768115942026,
+      "loss": 0.9141,
+      "step": 6490
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.00017644444444444446,
+      "loss": 0.9531,
+      "step": 6500
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0001762512077294686,
+      "loss": 0.9344,
+      "step": 6510
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.00017605797101449274,
+      "loss": 0.9716,
+      "step": 6520
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.00017586473429951688,
+      "loss": 0.9833,
+      "step": 6530
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.00017567149758454103,
+      "loss": 0.9097,
+      "step": 6540
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.00017547826086956522,
+      "loss": 0.9237,
+      "step": 6550
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.00017528502415458937,
+      "loss": 0.9659,
+      "step": 6560
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.0001750917874396135,
+      "loss": 0.9806,
+      "step": 6570
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.00017489855072463765,
+      "loss": 0.9091,
+      "step": 6580
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.00017470531400966182,
+      "loss": 0.9439,
+      "step": 6590
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.000174512077294686,
+      "loss": 0.9215,
+      "step": 6600
+    },
+    {
+      "epoch": 2.11,
+      "eval_loss": 1.0705487728118896,
+      "eval_runtime": 105.9435,
+      "eval_samples_per_second": 18.878,
+      "eval_steps_per_second": 2.36,
+      "step": 6600
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.00017431884057971013,
+      "loss": 0.9445,
+      "step": 6610
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.00017412560386473428,
+      "loss": 0.9648,
+      "step": 6620
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.00017393236714975842,
+      "loss": 0.9412,
+      "step": 6630
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.00017373913043478262,
+      "loss": 0.9656,
+      "step": 6640
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.00017354589371980676,
+      "loss": 0.9525,
+      "step": 6650
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.0001733526570048309,
+      "loss": 0.9165,
+      "step": 6660
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.00017315942028985504,
+      "loss": 0.9437,
+      "step": 6670
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0001729661835748792,
+      "loss": 0.9133,
+      "step": 6680
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.00017277294685990338,
+      "loss": 0.9673,
+      "step": 6690
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.00017257971014492752,
+      "loss": 0.9383,
+      "step": 6700
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.00017238647342995167,
+      "loss": 0.9627,
+      "step": 6710
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.00017219323671497584,
+      "loss": 0.975,
+      "step": 6720
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.000172,
+      "loss": 0.9567,
+      "step": 6730
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.00017180676328502415,
+      "loss": 0.988,
+      "step": 6740
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.0001716135265700483,
+      "loss": 0.9782,
+      "step": 6750
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.00017142028985507243,
+      "loss": 0.9429,
+      "step": 6760
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.0001712270531400966,
+      "loss": 0.9769,
+      "step": 6770
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.00017103381642512077,
+      "loss": 0.8864,
+      "step": 6780
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.00017084057971014492,
+      "loss": 0.954,
+      "step": 6790
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00017064734299516906,
+      "loss": 0.9934,
+      "step": 6800
+    },
+    {
+      "epoch": 2.18,
+      "eval_loss": 1.0722672939300537,
+      "eval_runtime": 105.8772,
+      "eval_samples_per_second": 18.89,
+      "eval_steps_per_second": 2.361,
+      "step": 6800
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00017045410628019323,
+      "loss": 0.926,
+      "step": 6810
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00017026086956521737,
+      "loss": 0.9026,
+      "step": 6820
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.00017006763285024154,
+      "loss": 0.9445,
+      "step": 6830
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.00016987439613526568,
+      "loss": 0.9492,
+      "step": 6840
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.00016968115942028983,
+      "loss": 0.9434,
+      "step": 6850
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.000169487922705314,
+      "loss": 0.9664,
+      "step": 6860
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00016929468599033817,
+      "loss": 0.9702,
+      "step": 6870
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.0001691014492753623,
+      "loss": 0.9021,
+      "step": 6880
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00016890821256038645,
+      "loss": 0.9608,
+      "step": 6890
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.00016871497584541062,
+      "loss": 0.9275,
+      "step": 6900
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.00016852173913043476,
+      "loss": 0.9441,
+      "step": 6910
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.00016832850241545893,
+      "loss": 0.9755,
+      "step": 6920
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00016813526570048307,
+      "loss": 0.9349,
+      "step": 6930
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00016794202898550722,
+      "loss": 0.9295,
+      "step": 6940
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.0001677487922705314,
+      "loss": 0.9618,
+      "step": 6950
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.00016755555555555553,
+      "loss": 0.9377,
+      "step": 6960
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0001673623188405797,
+      "loss": 0.9333,
+      "step": 6970
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.00016716908212560384,
+      "loss": 0.9005,
+      "step": 6980
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.000166975845410628,
+      "loss": 0.9766,
+      "step": 6990
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00016678260869565215,
+      "loss": 0.9591,
+      "step": 7000
+    },
+    {
+      "epoch": 2.24,
+      "eval_loss": 1.0725497007369995,
+      "eval_runtime": 105.7895,
+      "eval_samples_per_second": 18.905,
+      "eval_steps_per_second": 2.363,
+      "step": 7000
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00016658937198067632,
+      "loss": 0.9478,
+      "step": 7010
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.00016639613526570047,
+      "loss": 0.925,
+      "step": 7020
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.00016620289855072464,
+      "loss": 0.9257,
+      "step": 7030
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.00016600966183574878,
+      "loss": 0.9935,
+      "step": 7040
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.00016581642512077292,
+      "loss": 0.9546,
+      "step": 7050
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0001656231884057971,
+      "loss": 0.9571,
+      "step": 7060
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.00016542995169082123,
+      "loss": 1.001,
+      "step": 7070
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.0001652367149758454,
+      "loss": 0.9939,
+      "step": 7080
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00016504347826086955,
+      "loss": 0.9993,
+      "step": 7090
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00016485024154589372,
+      "loss": 0.9141,
+      "step": 7100
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.00016465700483091786,
+      "loss": 0.9383,
+      "step": 7110
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.00016446376811594203,
+      "loss": 0.9555,
+      "step": 7120
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.00016427053140096617,
+      "loss": 0.963,
+      "step": 7130
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0001640772946859903,
+      "loss": 0.9495,
+      "step": 7140
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.00016388405797101448,
+      "loss": 0.925,
+      "step": 7150
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.00016369082125603862,
+      "loss": 0.9237,
+      "step": 7160
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.0001634975845410628,
+      "loss": 0.9585,
+      "step": 7170
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00016330434782608694,
+      "loss": 0.9858,
+      "step": 7180
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00016311111111111108,
+      "loss": 0.9253,
+      "step": 7190
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00016291787439613525,
+      "loss": 0.9361,
+      "step": 7200
+    },
+    {
+      "epoch": 2.3,
+      "eval_loss": 1.0764214992523193,
+      "eval_runtime": 105.7836,
+      "eval_samples_per_second": 18.907,
+      "eval_steps_per_second": 2.363,
+      "step": 7200
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.00016272463768115942,
+      "loss": 1.0001,
+      "step": 7210
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.00016253140096618356,
+      "loss": 0.9428,
+      "step": 7220
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0001623381642512077,
+      "loss": 0.9294,
+      "step": 7230
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00016214492753623187,
+      "loss": 0.9508,
+      "step": 7240
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00016195169082125602,
+      "loss": 0.9423,
+      "step": 7250
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00016175845410628019,
+      "loss": 0.9655,
+      "step": 7260
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.00016156521739130433,
+      "loss": 0.9826,
+      "step": 7270
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.00016137198067632847,
+      "loss": 0.9505,
+      "step": 7280
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.00016117874396135264,
+      "loss": 0.9119,
+      "step": 7290
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0001609855072463768,
+      "loss": 0.949,
+      "step": 7300
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.00016079227053140095,
+      "loss": 0.9056,
+      "step": 7310
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0001605990338164251,
+      "loss": 0.9203,
+      "step": 7320
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00016040579710144926,
+      "loss": 0.9134,
+      "step": 7330
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00016021256038647343,
+      "loss": 0.9175,
+      "step": 7340
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00016001932367149758,
+      "loss": 0.921,
+      "step": 7350
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00015982608695652172,
+      "loss": 0.9443,
+      "step": 7360
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00015963285024154586,
+      "loss": 0.9423,
+      "step": 7370
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00015943961352657003,
+      "loss": 0.9292,
+      "step": 7380
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.0001592463768115942,
+      "loss": 0.968,
+      "step": 7390
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00015905314009661834,
+      "loss": 0.9355,
+      "step": 7400
+    },
+    {
+      "epoch": 2.37,
+      "eval_loss": 1.0742346048355103,
+      "eval_runtime": 105.7786,
+      "eval_samples_per_second": 18.907,
+      "eval_steps_per_second": 2.363,
+      "step": 7400
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.0001588599033816425,
+      "loss": 0.9083,
+      "step": 7410
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00015866666666666663,
+      "loss": 0.9229,
+      "step": 7420
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.00015847342995169083,
+      "loss": 0.9415,
+      "step": 7430
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.00015828019323671497,
+      "loss": 0.9782,
+      "step": 7440
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0001580869565217391,
+      "loss": 0.9653,
+      "step": 7450
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.00015789371980676325,
+      "loss": 0.9629,
+      "step": 7460
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.00015770048309178742,
+      "loss": 0.9472,
+      "step": 7470
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.0001575072463768116,
+      "loss": 0.955,
+      "step": 7480
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00015731400966183574,
+      "loss": 0.9727,
+      "step": 7490
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00015712077294685988,
+      "loss": 0.9746,
+      "step": 7500
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00015692753623188402,
+      "loss": 0.9836,
+      "step": 7510
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.00015673429951690822,
+      "loss": 0.9561,
+      "step": 7520
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.00015654106280193236,
+      "loss": 0.9581,
+      "step": 7530
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0001563478260869565,
+      "loss": 0.961,
+      "step": 7540
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.00015615458937198065,
+      "loss": 0.9313,
+      "step": 7550
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0001559613526570048,
+      "loss": 0.9707,
+      "step": 7560
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.00015576811594202898,
+      "loss": 0.937,
+      "step": 7570
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.00015557487922705313,
+      "loss": 0.9776,
+      "step": 7580
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.00015538164251207727,
+      "loss": 0.9475,
+      "step": 7590
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.0001551884057971014,
+      "loss": 0.9576,
+      "step": 7600
+    },
+    {
+      "epoch": 2.43,
+      "eval_loss": 1.0753015279769897,
+      "eval_runtime": 105.8392,
+      "eval_samples_per_second": 18.897,
+      "eval_steps_per_second": 2.362,
+      "step": 7600
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.0001549951690821256,
+      "loss": 0.8983,
+      "step": 7610
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.00015480193236714975,
+      "loss": 0.9855,
+      "step": 7620
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.0001546086956521739,
+      "loss": 0.9291,
+      "step": 7630
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.00015441545893719804,
+      "loss": 0.9364,
+      "step": 7640
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.00015422222222222223,
+      "loss": 0.9144,
+      "step": 7650
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.00015402898550724638,
+      "loss": 0.9659,
+      "step": 7660
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.00015383574879227052,
+      "loss": 0.9632,
+      "step": 7670
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.00015364251207729466,
+      "loss": 0.9451,
+      "step": 7680
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0001534492753623188,
+      "loss": 0.9601,
+      "step": 7690
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.000153256038647343,
+      "loss": 0.9643,
+      "step": 7700
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.00015306280193236714,
+      "loss": 0.936,
+      "step": 7710
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.00015286956521739129,
+      "loss": 0.9286,
+      "step": 7720
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.00015267632850241543,
+      "loss": 0.9404,
+      "step": 7730
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.00015248309178743962,
+      "loss": 0.9496,
+      "step": 7740
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.00015228985507246377,
+      "loss": 0.9468,
+      "step": 7750
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0001520966183574879,
+      "loss": 0.9185,
+      "step": 7760
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.00015190338164251205,
+      "loss": 0.9377,
+      "step": 7770
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0001517101449275362,
+      "loss": 0.945,
+      "step": 7780
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0001515169082125604,
+      "loss": 0.9862,
+      "step": 7790
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00015132367149758453,
+      "loss": 0.9432,
+      "step": 7800
+    },
+    {
+      "epoch": 2.5,
+      "eval_loss": 1.073928713798523,
+      "eval_runtime": 105.8871,
+      "eval_samples_per_second": 18.888,
+      "eval_steps_per_second": 2.361,
+      "step": 7800
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00015113043478260868,
+      "loss": 0.9618,
+      "step": 7810
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00015093719806763282,
+      "loss": 0.9377,
+      "step": 7820
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.00015074396135265702,
+      "loss": 0.9477,
+      "step": 7830
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.00015055072463768116,
+      "loss": 0.9616,
+      "step": 7840
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0001503574879227053,
+      "loss": 0.9327,
+      "step": 7850
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.00015016425120772944,
+      "loss": 0.9718,
+      "step": 7860
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.00014997101449275361,
+      "loss": 0.9431,
+      "step": 7870
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.00014977777777777776,
+      "loss": 0.9504,
+      "step": 7880
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.00014958454106280193,
+      "loss": 0.9267,
+      "step": 7890
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.00014939130434782607,
+      "loss": 0.9409,
+      "step": 7900
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.00014919806763285024,
+      "loss": 0.9531,
+      "step": 7910
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.00014900483091787438,
+      "loss": 0.9397,
+      "step": 7920
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.00014881159420289855,
+      "loss": 0.969,
+      "step": 7930
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.0001486183574879227,
+      "loss": 0.9509,
+      "step": 7940
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.00014842512077294686,
+      "loss": 0.9136,
+      "step": 7950
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.000148231884057971,
+      "loss": 0.9522,
+      "step": 7960
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.00014803864734299515,
+      "loss": 0.9331,
+      "step": 7970
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.00014784541062801932,
+      "loss": 0.925,
+      "step": 7980
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00014765217391304346,
+      "loss": 0.9437,
+      "step": 7990
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00014745893719806763,
+      "loss": 0.9682,
+      "step": 8000
+    },
+    {
+      "epoch": 2.56,
+      "eval_loss": 1.073836326599121,
+      "eval_runtime": 105.8058,
+      "eval_samples_per_second": 18.903,
+      "eval_steps_per_second": 2.363,
+      "step": 8000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00014726570048309177,
+      "loss": 0.9431,
+      "step": 8010
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.00014707246376811594,
+      "loss": 0.9381,
+      "step": 8020
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.00014687922705314008,
+      "loss": 0.9446,
+      "step": 8030
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.00014668599033816425,
+      "loss": 0.9888,
+      "step": 8040
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0001464927536231884,
+      "loss": 0.9139,
+      "step": 8050
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.00014629951690821257,
+      "loss": 0.9272,
+      "step": 8060
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0001461062801932367,
+      "loss": 0.926,
+      "step": 8070
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.00014591304347826085,
+      "loss": 0.9274,
+      "step": 8080
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.00014571980676328502,
+      "loss": 0.988,
+      "step": 8090
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.00014552657004830916,
+      "loss": 0.9504,
+      "step": 8100
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.00014533333333333333,
+      "loss": 0.9279,
+      "step": 8110
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00014514009661835748,
+      "loss": 0.9821,
+      "step": 8120
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00014494685990338165,
+      "loss": 0.9624,
+      "step": 8130
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.0001447536231884058,
+      "loss": 0.9398,
+      "step": 8140
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00014456038647342996,
+      "loss": 0.953,
+      "step": 8150
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.0001443671497584541,
+      "loss": 0.9907,
+      "step": 8160
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00014417391304347824,
+      "loss": 0.9152,
+      "step": 8170
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.0001439806763285024,
+      "loss": 0.9368,
+      "step": 8180
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.00014378743961352656,
+      "loss": 0.9304,
+      "step": 8190
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.00014359420289855072,
+      "loss": 0.9873,
+      "step": 8200
+    },
+    {
+      "epoch": 2.62,
+      "eval_loss": 1.0735951662063599,
+      "eval_runtime": 105.698,
+      "eval_samples_per_second": 18.922,
+      "eval_steps_per_second": 2.365,
+      "step": 8200
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00014340096618357487,
+      "loss": 0.9527,
+      "step": 8210
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00014320772946859904,
+      "loss": 0.933,
+      "step": 8220
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00014301449275362318,
+      "loss": 0.9563,
+      "step": 8230
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.00014282125603864735,
+      "loss": 0.9734,
+      "step": 8240
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0001426280193236715,
+      "loss": 0.987,
+      "step": 8250
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.00014243478260869563,
+      "loss": 0.9276,
+      "step": 8260
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.0001422415458937198,
+      "loss": 0.9906,
+      "step": 8270
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.00014204830917874395,
+      "loss": 0.9927,
+      "step": 8280
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.00014185507246376812,
+      "loss": 0.916,
+      "step": 8290
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00014166183574879226,
+      "loss": 0.9458,
+      "step": 8300
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00014146859903381643,
+      "loss": 0.9491,
+      "step": 8310
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00014127536231884057,
+      "loss": 0.916,
+      "step": 8320
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.0001410821256038647,
+      "loss": 0.9742,
+      "step": 8330
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.00014088888888888888,
+      "loss": 0.9526,
+      "step": 8340
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.00014069565217391303,
+      "loss": 0.9452,
+      "step": 8350
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.0001405024154589372,
+      "loss": 0.949,
+      "step": 8360
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00014030917874396134,
+      "loss": 0.9686,
+      "step": 8370
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.0001401159420289855,
+      "loss": 0.948,
+      "step": 8380
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00013992270531400965,
+      "loss": 0.925,
+      "step": 8390
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.0001397294685990338,
+      "loss": 0.9325,
+      "step": 8400
+    },
+    {
+      "epoch": 2.69,
+      "eval_loss": 1.0741145610809326,
+      "eval_runtime": 105.7009,
+      "eval_samples_per_second": 18.921,
+      "eval_steps_per_second": 2.365,
+      "step": 8400
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.00013953623188405796,
+      "loss": 0.9426,
+      "step": 8410
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.0001393429951690821,
+      "loss": 0.9794,
+      "step": 8420
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.00013914975845410627,
+      "loss": 0.9779,
+      "step": 8430
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.00013895652173913042,
+      "loss": 0.9258,
+      "step": 8440
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.0001387632850241546,
+      "loss": 0.9499,
+      "step": 8450
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00013857004830917873,
+      "loss": 0.9712,
+      "step": 8460
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00013837681159420287,
+      "loss": 0.9244,
+      "step": 8470
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00013818357487922704,
+      "loss": 0.9513,
+      "step": 8480
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00013799033816425118,
+      "loss": 0.9629,
+      "step": 8490
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00013779710144927535,
+      "loss": 0.9266,
+      "step": 8500
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.0001376038647342995,
+      "loss": 0.9795,
+      "step": 8510
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.00013741062801932367,
+      "loss": 0.9562,
+      "step": 8520
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.0001372173913043478,
+      "loss": 0.9301,
+      "step": 8530
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.00013702415458937198,
+      "loss": 0.9652,
+      "step": 8540
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.00013683091787439612,
+      "loss": 0.9698,
+      "step": 8550
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.00013663768115942026,
+      "loss": 0.9742,
+      "step": 8560
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.00013644444444444443,
+      "loss": 0.9189,
+      "step": 8570
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.00013625120772946858,
+      "loss": 0.9731,
+      "step": 8580
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.00013605797101449275,
+      "loss": 1.0023,
+      "step": 8590
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.0001358647342995169,
+      "loss": 0.9228,
+      "step": 8600
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 1.0715253353118896,
+      "eval_runtime": 105.7865,
+      "eval_samples_per_second": 18.906,
+      "eval_steps_per_second": 2.363,
+      "step": 8600
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.00013567149758454106,
+      "loss": 0.9774,
+      "step": 8610
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.0001354782608695652,
+      "loss": 0.9616,
+      "step": 8620
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.00013528502415458934,
+      "loss": 0.9512,
+      "step": 8630
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.0001350917874396135,
+      "loss": 0.9505,
+      "step": 8640
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.00013489855072463765,
+      "loss": 0.9213,
+      "step": 8650
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.00013470531400966182,
+      "loss": 0.9283,
+      "step": 8660
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.00013451207729468597,
+      "loss": 0.9408,
+      "step": 8670
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.00013431884057971014,
+      "loss": 0.9554,
+      "step": 8680
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.00013412560386473428,
+      "loss": 0.958,
+      "step": 8690
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.00013393236714975842,
+      "loss": 0.9351,
+      "step": 8700
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.0001337391304347826,
+      "loss": 0.951,
+      "step": 8710
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.00013354589371980673,
+      "loss": 0.9103,
+      "step": 8720
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.0001333526570048309,
+      "loss": 1.0058,
+      "step": 8730
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00013315942028985505,
+      "loss": 0.8872,
+      "step": 8740
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00013296618357487922,
+      "loss": 0.9944,
+      "step": 8750
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00013277294685990336,
+      "loss": 0.9418,
+      "step": 8760
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.00013257971014492753,
+      "loss": 0.9814,
+      "step": 8770
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.00013238647342995167,
+      "loss": 0.9732,
+      "step": 8780
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.00013219323671497584,
+      "loss": 0.9644,
+      "step": 8790
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 0.9546,
+      "step": 8800
+    },
+    {
+      "epoch": 2.82,
+      "eval_loss": 1.0739343166351318,
+      "eval_runtime": 105.7738,
+      "eval_samples_per_second": 18.908,
+      "eval_steps_per_second": 2.364,
+      "step": 8800
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.00013180676328502413,
+      "loss": 0.9472,
+      "step": 8810
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.0001316135265700483,
+      "loss": 0.9088,
+      "step": 8820
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.00013142028985507244,
+      "loss": 0.8972,
+      "step": 8830
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.0001312270531400966,
+      "loss": 0.9297,
+      "step": 8840
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.00013103381642512075,
+      "loss": 0.9581,
+      "step": 8850
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.00013084057971014492,
+      "loss": 0.9641,
+      "step": 8860
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.00013064734299516906,
+      "loss": 0.9598,
+      "step": 8870
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.00013045410628019323,
+      "loss": 0.9838,
+      "step": 8880
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.00013026086956521737,
+      "loss": 0.9136,
+      "step": 8890
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.00013006763285024152,
+      "loss": 0.9375,
+      "step": 8900
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.0001298743961352657,
+      "loss": 0.9772,
+      "step": 8910
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.00012968115942028983,
+      "loss": 0.9662,
+      "step": 8920
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.000129487922705314,
+      "loss": 0.9301,
+      "step": 8930
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.00012929468599033814,
+      "loss": 0.9543,
+      "step": 8940
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.0001291014492753623,
+      "loss": 0.9831,
+      "step": 8950
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.00012890821256038645,
+      "loss": 0.9418,
+      "step": 8960
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.00012871497584541062,
+      "loss": 0.9684,
+      "step": 8970
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.00012852173913043477,
+      "loss": 0.92,
+      "step": 8980
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00012832850241545894,
+      "loss": 0.9661,
+      "step": 8990
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00012813526570048308,
+      "loss": 0.9488,
+      "step": 9000
+    },
+    {
+      "epoch": 2.88,
+      "eval_loss": 1.0742032527923584,
+      "eval_runtime": 105.8222,
+      "eval_samples_per_second": 18.9,
+      "eval_steps_per_second": 2.362,
+      "step": 9000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00012794202898550722,
+      "loss": 0.9173,
+      "step": 9010
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.0001277487922705314,
+      "loss": 0.9515,
+      "step": 9020
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00012755555555555553,
+      "loss": 0.9241,
+      "step": 9030
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.0001273623188405797,
+      "loss": 0.9537,
+      "step": 9040
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.00012716908212560385,
+      "loss": 0.9495,
+      "step": 9050
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.00012697584541062801,
+      "loss": 0.9617,
+      "step": 9060
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.00012678260869565216,
+      "loss": 0.9857,
+      "step": 9070
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00012658937198067633,
+      "loss": 0.9406,
+      "step": 9080
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00012639613526570047,
+      "loss": 0.9825,
+      "step": 9090
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00012620289855072464,
+      "loss": 0.9971,
+      "step": 9100
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00012600966183574878,
+      "loss": 0.9819,
+      "step": 9110
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.00012581642512077292,
+      "loss": 0.9232,
+      "step": 9120
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.0001256231884057971,
+      "loss": 0.9198,
+      "step": 9130
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.00012542995169082124,
+      "loss": 0.9956,
+      "step": 9140
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.0001252367149758454,
+      "loss": 0.9817,
+      "step": 9150
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.00012504347826086955,
+      "loss": 0.948,
+      "step": 9160
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.00012485024154589372,
+      "loss": 0.972,
+      "step": 9170
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00012465700483091786,
+      "loss": 0.9885,
+      "step": 9180
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00012446376811594203,
+      "loss": 0.9966,
+      "step": 9190
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00012427053140096617,
+      "loss": 0.9444,
+      "step": 9200
+    },
+    {
+      "epoch": 2.94,
+      "eval_loss": 1.0732178688049316,
+      "eval_runtime": 105.7104,
+      "eval_samples_per_second": 18.92,
+      "eval_steps_per_second": 2.365,
+      "step": 9200
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00012407729468599032,
+      "loss": 0.9398,
+      "step": 9210
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00012388405797101449,
+      "loss": 0.9518,
+      "step": 9220
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00012369082125603863,
+      "loss": 0.8983,
+      "step": 9230
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.0001234975845410628,
+      "loss": 0.9363,
+      "step": 9240
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.00012330434782608694,
+      "loss": 0.9339,
+      "step": 9250
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.0001231111111111111,
+      "loss": 0.9962,
+      "step": 9260
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00012291787439613525,
+      "loss": 0.9377,
+      "step": 9270
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00012272463768115942,
+      "loss": 0.951,
+      "step": 9280
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00012253140096618356,
+      "loss": 0.9589,
+      "step": 9290
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00012233816425120773,
+      "loss": 0.9865,
+      "step": 9300
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00012214492753623188,
+      "loss": 0.9463,
+      "step": 9310
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00012195169082125602,
+      "loss": 0.8978,
+      "step": 9320
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00012175845410628019,
+      "loss": 0.9443,
+      "step": 9330
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00012156521739130433,
+      "loss": 0.904,
+      "step": 9340
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.0001213719806763285,
+      "loss": 0.9817,
+      "step": 9350
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00012117874396135264,
+      "loss": 0.9092,
+      "step": 9360
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0001209855072463768,
+      "loss": 0.9397,
+      "step": 9370
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.00012079227053140096,
+      "loss": 0.881,
+      "step": 9380
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.00012059903381642511,
+      "loss": 0.8538,
+      "step": 9390
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.00012040579710144927,
+      "loss": 0.8364,
+      "step": 9400
+    },
+    {
+      "epoch": 3.01,
+      "eval_loss": 1.0937556028366089,
+      "eval_runtime": 105.6105,
+      "eval_samples_per_second": 18.938,
+      "eval_steps_per_second": 2.367,
+      "step": 9400
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.00012021256038647342,
+      "loss": 0.8731,
+      "step": 9410
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.00012001932367149758,
+      "loss": 0.8553,
+      "step": 9420
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.00011982608695652172,
+      "loss": 0.8505,
+      "step": 9430
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.00011963285024154588,
+      "loss": 0.9045,
+      "step": 9440
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.00011943961352657004,
+      "loss": 0.8725,
+      "step": 9450
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00011924637681159419,
+      "loss": 0.8939,
+      "step": 9460
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00011905314009661835,
+      "loss": 0.9085,
+      "step": 9470
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.0001188599033816425,
+      "loss": 0.8813,
+      "step": 9480
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.00011866666666666666,
+      "loss": 0.9056,
+      "step": 9490
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.00011847342995169082,
+      "loss": 0.885,
+      "step": 9500
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.00011828019323671496,
+      "loss": 0.8836,
+      "step": 9510
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00011808695652173911,
+      "loss": 0.8932,
+      "step": 9520
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00011789371980676327,
+      "loss": 0.8395,
+      "step": 9530
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00011770048309178743,
+      "loss": 0.888,
+      "step": 9540
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00011750724637681158,
+      "loss": 0.8839,
+      "step": 9550
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00011731400966183574,
+      "loss": 0.8919,
+      "step": 9560
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.0001171207729468599,
+      "loss": 0.8569,
+      "step": 9570
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00011692753623188405,
+      "loss": 0.8518,
+      "step": 9580
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00011673429951690821,
+      "loss": 0.9049,
+      "step": 9590
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00011654106280193235,
+      "loss": 0.9091,
+      "step": 9600
+    },
+    {
+      "epoch": 3.07,
+      "eval_loss": 1.0996880531311035,
+      "eval_runtime": 105.5655,
+      "eval_samples_per_second": 18.946,
+      "eval_steps_per_second": 2.368,
+      "step": 9600
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00011634782608695652,
+      "loss": 0.8709,
+      "step": 9610
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00011615458937198066,
+      "loss": 0.9005,
+      "step": 9620
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00011596135265700482,
+      "loss": 0.9014,
+      "step": 9630
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00011576811594202897,
+      "loss": 0.8451,
+      "step": 9640
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00011557487922705313,
+      "loss": 0.8486,
+      "step": 9650
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00011538164251207729,
+      "loss": 0.8667,
+      "step": 9660
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00011518840579710143,
+      "loss": 0.8479,
+      "step": 9670
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.0001149951690821256,
+      "loss": 0.89,
+      "step": 9680
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.00011480193236714974,
+      "loss": 0.9254,
+      "step": 9690
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.00011460869565217391,
+      "loss": 0.8853,
+      "step": 9700
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00011441545893719805,
+      "loss": 0.8741,
+      "step": 9710
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00011422222222222222,
+      "loss": 0.8632,
+      "step": 9720
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00011402898550724637,
+      "loss": 0.8529,
+      "step": 9730
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00011383574879227051,
+      "loss": 0.8668,
+      "step": 9740
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00011364251207729468,
+      "loss": 0.8963,
+      "step": 9750
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00011344927536231882,
+      "loss": 0.8793,
+      "step": 9760
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.00011325603864734299,
+      "loss": 0.9182,
+      "step": 9770
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.00011306280193236713,
+      "loss": 0.8473,
+      "step": 9780
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.0001128695652173913,
+      "loss": 0.8931,
+      "step": 9790
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00011267632850241545,
+      "loss": 0.8351,
+      "step": 9800
+    },
+    {
+      "epoch": 3.14,
+      "eval_loss": 1.0953755378723145,
+      "eval_runtime": 105.6296,
+      "eval_samples_per_second": 18.934,
+      "eval_steps_per_second": 2.367,
+      "step": 9800
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00011248309178743961,
+      "loss": 0.8867,
+      "step": 9810
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00011228985507246376,
+      "loss": 0.8693,
+      "step": 9820
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.0001120966183574879,
+      "loss": 0.9383,
+      "step": 9830
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00011190338164251207,
+      "loss": 0.8753,
+      "step": 9840
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00011171014492753621,
+      "loss": 0.8897,
+      "step": 9850
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00011151690821256038,
+      "loss": 0.9105,
+      "step": 9860
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.00011132367149758452,
+      "loss": 0.8594,
+      "step": 9870
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.0001111304347826087,
+      "loss": 0.8909,
+      "step": 9880
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.00011093719806763284,
+      "loss": 0.8908,
+      "step": 9890
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.000110743961352657,
+      "loss": 0.9055,
+      "step": 9900
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00011055072463768115,
+      "loss": 0.8832,
+      "step": 9910
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00011035748792270532,
+      "loss": 0.8862,
+      "step": 9920
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.00011016425120772946,
+      "loss": 0.8566,
+      "step": 9930
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.0001099710144927536,
+      "loss": 0.8839,
+      "step": 9940
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.00010977777777777777,
+      "loss": 0.8942,
+      "step": 9950
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00010958454106280192,
+      "loss": 0.8885,
+      "step": 9960
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00010939130434782609,
+      "loss": 0.8567,
+      "step": 9970
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00010919806763285023,
+      "loss": 0.8847,
+      "step": 9980
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.0001090048309178744,
+      "loss": 0.9023,
+      "step": 9990
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.00010881159420289854,
+      "loss": 0.8417,
+      "step": 10000
+    },
+    {
+      "epoch": 3.2,
+      "eval_loss": 1.1017428636550903,
+      "eval_runtime": 105.7728,
+      "eval_samples_per_second": 18.908,
+      "eval_steps_per_second": 2.364,
+      "step": 10000
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.0001086183574879227,
+      "loss": 0.8819,
+      "step": 10010
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.00010842512077294685,
+      "loss": 0.861,
+      "step": 10020
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.000108231884057971,
+      "loss": 0.9112,
+      "step": 10030
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.00010803864734299516,
+      "loss": 0.8673,
+      "step": 10040
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.00010784541062801931,
+      "loss": 0.88,
+      "step": 10050
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.00010765217391304348,
+      "loss": 0.8634,
+      "step": 10060
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.00010745893719806762,
+      "loss": 0.9328,
+      "step": 10070
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00010726570048309178,
+      "loss": 0.8646,
+      "step": 10080
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00010707246376811593,
+      "loss": 0.8252,
+      "step": 10090
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00010687922705314009,
+      "loss": 0.9282,
+      "step": 10100
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00010668599033816424,
+      "loss": 0.8955,
+      "step": 10110
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.0001064927536231884,
+      "loss": 0.8344,
+      "step": 10120
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00010629951690821256,
+      "loss": 0.8901,
+      "step": 10130
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.0001061062801932367,
+      "loss": 0.8484,
+      "step": 10140
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.00010591304347826087,
+      "loss": 0.8562,
+      "step": 10150
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.00010571980676328501,
+      "loss": 0.9222,
+      "step": 10160
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.00010552657004830917,
+      "loss": 0.8509,
+      "step": 10170
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00010533333333333332,
+      "loss": 0.8725,
+      "step": 10180
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00010514009661835748,
+      "loss": 0.8866,
+      "step": 10190
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00010494685990338164,
+      "loss": 0.8822,
+      "step": 10200
+    },
+    {
+      "epoch": 3.26,
+      "eval_loss": 1.1016156673431396,
+      "eval_runtime": 105.8053,
+      "eval_samples_per_second": 18.903,
+      "eval_steps_per_second": 2.363,
+      "step": 10200
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00010475362318840579,
+      "loss": 0.9291,
+      "step": 10210
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00010456038647342995,
+      "loss": 0.889,
+      "step": 10220
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.0001043671497584541,
+      "loss": 0.926,
+      "step": 10230
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.00010417391304347825,
+      "loss": 0.8719,
+      "step": 10240
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.0001039806763285024,
+      "loss": 0.902,
+      "step": 10250
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.00010378743961352656,
+      "loss": 0.8727,
+      "step": 10260
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00010359420289855071,
+      "loss": 0.8451,
+      "step": 10270
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00010340096618357487,
+      "loss": 0.8685,
+      "step": 10280
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00010320772946859903,
+      "loss": 0.8748,
+      "step": 10290
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00010301449275362318,
+      "loss": 0.8771,
+      "step": 10300
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00010282125603864733,
+      "loss": 0.846,
+      "step": 10310
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.0001026280193236715,
+      "loss": 0.8616,
+      "step": 10320
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00010243478260869564,
+      "loss": 0.8989,
+      "step": 10330
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.0001022415458937198,
+      "loss": 0.9234,
+      "step": 10340
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00010204830917874395,
+      "loss": 0.9422,
+      "step": 10350
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.0001018550724637681,
+      "loss": 0.9013,
+      "step": 10360
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.00010166183574879226,
+      "loss": 0.8427,
+      "step": 10370
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.0001014685990338164,
+      "loss": 0.8651,
+      "step": 10380
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.00010127536231884057,
+      "loss": 0.8754,
+      "step": 10390
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00010108212560386472,
+      "loss": 0.9252,
+      "step": 10400
+    },
+    {
+      "epoch": 3.33,
+      "eval_loss": 1.1023533344268799,
+      "eval_runtime": 105.7857,
+      "eval_samples_per_second": 18.906,
+      "eval_steps_per_second": 2.363,
+      "step": 10400
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00010088888888888889,
+      "loss": 0.8828,
+      "step": 10410
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00010069565217391303,
+      "loss": 0.8745,
+      "step": 10420
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.0001005024154589372,
+      "loss": 0.8903,
+      "step": 10430
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.00010030917874396134,
+      "loss": 0.881,
+      "step": 10440
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.00010011594202898548,
+      "loss": 0.8799,
+      "step": 10450
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 9.992270531400965e-05,
+      "loss": 0.8923,
+      "step": 10460
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 9.97294685990338e-05,
+      "loss": 0.8649,
+      "step": 10470
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 9.953623188405797e-05,
+      "loss": 0.8552,
+      "step": 10480
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 9.934299516908211e-05,
+      "loss": 0.8901,
+      "step": 10490
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 9.914975845410628e-05,
+      "loss": 0.8467,
+      "step": 10500
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 9.895652173913042e-05,
+      "loss": 0.8572,
+      "step": 10510
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 9.876328502415459e-05,
+      "loss": 0.8769,
+      "step": 10520
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 9.857004830917873e-05,
+      "loss": 0.8712,
+      "step": 10530
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 9.83768115942029e-05,
+      "loss": 0.9274,
+      "step": 10540
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 9.818357487922705e-05,
+      "loss": 0.8769,
+      "step": 10550
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 9.799033816425119e-05,
+      "loss": 0.9091,
+      "step": 10560
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 9.779710144927536e-05,
+      "loss": 0.8762,
+      "step": 10570
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 9.76038647342995e-05,
+      "loss": 0.8546,
+      "step": 10580
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 9.741062801932367e-05,
+      "loss": 0.8882,
+      "step": 10590
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 9.721739130434781e-05,
+      "loss": 0.9302,
+      "step": 10600
+    },
+    {
+      "epoch": 3.39,
+      "eval_loss": 1.099436640739441,
+      "eval_runtime": 105.6452,
+      "eval_samples_per_second": 18.931,
+      "eval_steps_per_second": 2.366,
+      "step": 10600
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 9.702415458937198e-05,
+      "loss": 0.8952,
+      "step": 10610
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 9.683091787439612e-05,
+      "loss": 0.8826,
+      "step": 10620
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 9.66376811594203e-05,
+      "loss": 0.8515,
+      "step": 10630
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 9.644444444444444e-05,
+      "loss": 0.9101,
+      "step": 10640
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 9.625120772946858e-05,
+      "loss": 0.9107,
+      "step": 10650
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 9.605797101449275e-05,
+      "loss": 0.8618,
+      "step": 10660
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 9.586473429951689e-05,
+      "loss": 0.829,
+      "step": 10670
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 9.567149758454106e-05,
+      "loss": 0.9084,
+      "step": 10680
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 9.54782608695652e-05,
+      "loss": 0.8791,
+      "step": 10690
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 9.528502415458937e-05,
+      "loss": 0.8995,
+      "step": 10700
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 9.509178743961352e-05,
+      "loss": 0.8697,
+      "step": 10710
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 9.489855072463767e-05,
+      "loss": 0.8598,
+      "step": 10720
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 9.470531400966183e-05,
+      "loss": 0.8892,
+      "step": 10730
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 9.451207729468598e-05,
+      "loss": 0.8876,
+      "step": 10740
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 9.431884057971014e-05,
+      "loss": 0.9099,
+      "step": 10750
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 9.412560386473428e-05,
+      "loss": 0.8841,
+      "step": 10760
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 9.393236714975845e-05,
+      "loss": 0.9175,
+      "step": 10770
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 9.37391304347826e-05,
+      "loss": 0.8894,
+      "step": 10780
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 9.354589371980676e-05,
+      "loss": 0.8984,
+      "step": 10790
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 9.335265700483091e-05,
+      "loss": 0.899,
+      "step": 10800
+    },
+    {
+      "epoch": 3.46,
+      "eval_loss": 1.1025915145874023,
+      "eval_runtime": 105.6092,
+      "eval_samples_per_second": 18.938,
+      "eval_steps_per_second": 2.367,
+      "step": 10800
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 9.315942028985506e-05,
+      "loss": 0.9025,
+      "step": 10810
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 9.296618357487922e-05,
+      "loss": 0.8491,
+      "step": 10820
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.277294685990338e-05,
+      "loss": 0.864,
+      "step": 10830
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.257971014492753e-05,
+      "loss": 0.8667,
+      "step": 10840
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.238647342995169e-05,
+      "loss": 0.8511,
+      "step": 10850
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.219323671497584e-05,
+      "loss": 0.8823,
+      "step": 10860
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.199999999999999e-05,
+      "loss": 0.9321,
+      "step": 10870
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.180676328502414e-05,
+      "loss": 0.9039,
+      "step": 10880
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.16135265700483e-05,
+      "loss": 0.8711,
+      "step": 10890
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 9.142028985507245e-05,
+      "loss": 0.9234,
+      "step": 10900
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 9.122705314009661e-05,
+      "loss": 0.8885,
+      "step": 10910
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 9.103381642512077e-05,
+      "loss": 0.8579,
+      "step": 10920
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.084057971014492e-05,
+      "loss": 0.8505,
+      "step": 10930
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.064734299516908e-05,
+      "loss": 0.874,
+      "step": 10940
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.045410628019322e-05,
+      "loss": 0.891,
+      "step": 10950
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.026086956521738e-05,
+      "loss": 0.8843,
+      "step": 10960
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.006763285024153e-05,
+      "loss": 0.8923,
+      "step": 10970
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 8.987439613526569e-05,
+      "loss": 0.8668,
+      "step": 10980
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 8.968115942028985e-05,
+      "loss": 0.8462,
+      "step": 10990
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 8.9487922705314e-05,
+      "loss": 0.8824,
+      "step": 11000
+    },
+    {
+      "epoch": 3.52,
+      "eval_loss": 1.103293538093567,
+      "eval_runtime": 105.6938,
+      "eval_samples_per_second": 18.923,
+      "eval_steps_per_second": 2.365,
+      "step": 11000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 8.929468599033816e-05,
+      "loss": 0.9079,
+      "step": 11010
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 8.91014492753623e-05,
+      "loss": 0.8729,
+      "step": 11020
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 8.890821256038647e-05,
+      "loss": 0.8724,
+      "step": 11030
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 8.871497584541061e-05,
+      "loss": 0.8586,
+      "step": 11040
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 8.852173913043478e-05,
+      "loss": 0.8635,
+      "step": 11050
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 8.832850241545893e-05,
+      "loss": 0.9202,
+      "step": 11060
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 8.813526570048308e-05,
+      "loss": 0.8761,
+      "step": 11070
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 8.794202898550724e-05,
+      "loss": 0.8373,
+      "step": 11080
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 8.77487922705314e-05,
+      "loss": 0.8886,
+      "step": 11090
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 8.755555555555555e-05,
+      "loss": 0.8835,
+      "step": 11100
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 8.736231884057969e-05,
+      "loss": 0.829,
+      "step": 11110
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 8.716908212560386e-05,
+      "loss": 0.8461,
+      "step": 11120
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 8.6975845410628e-05,
+      "loss": 0.9023,
+      "step": 11130
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 8.678260869565217e-05,
+      "loss": 0.9455,
+      "step": 11140
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 8.658937198067632e-05,
+      "loss": 0.8482,
+      "step": 11150
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 8.639613526570049e-05,
+      "loss": 0.8522,
+      "step": 11160
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 8.620289855072463e-05,
+      "loss": 0.8928,
+      "step": 11170
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 8.600966183574877e-05,
+      "loss": 0.9059,
+      "step": 11180
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 8.581642512077294e-05,
+      "loss": 0.9116,
+      "step": 11190
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 8.562318840579708e-05,
+      "loss": 0.8806,
+      "step": 11200
+    },
+    {
+      "epoch": 3.58,
+      "eval_loss": 1.1037750244140625,
+      "eval_runtime": 105.6165,
+      "eval_samples_per_second": 18.936,
+      "eval_steps_per_second": 2.367,
+      "step": 11200
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 8.542995169082125e-05,
+      "loss": 0.8387,
+      "step": 11210
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 8.52367149758454e-05,
+      "loss": 0.8371,
+      "step": 11220
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 8.504347826086957e-05,
+      "loss": 0.9119,
+      "step": 11230
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 8.485024154589371e-05,
+      "loss": 0.8651,
+      "step": 11240
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 8.465700483091788e-05,
+      "loss": 0.8814,
+      "step": 11250
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 8.446376811594202e-05,
+      "loss": 0.8982,
+      "step": 11260
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 8.427053140096616e-05,
+      "loss": 0.9436,
+      "step": 11270
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 8.407729468599033e-05,
+      "loss": 0.8837,
+      "step": 11280
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 8.388405797101448e-05,
+      "loss": 0.9212,
+      "step": 11290
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 8.369082125603865e-05,
+      "loss": 0.899,
+      "step": 11300
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 8.349758454106279e-05,
+      "loss": 0.8822,
+      "step": 11310
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 8.330434782608696e-05,
+      "loss": 0.8565,
+      "step": 11320
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 8.31111111111111e-05,
+      "loss": 0.8496,
+      "step": 11330
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 8.291787439613527e-05,
+      "loss": 0.9226,
+      "step": 11340
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 8.272463768115941e-05,
+      "loss": 0.9008,
+      "step": 11350
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 8.253140096618358e-05,
+      "loss": 0.8852,
+      "step": 11360
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 8.233816425120772e-05,
+      "loss": 0.8944,
+      "step": 11370
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 8.214492753623187e-05,
+      "loss": 0.8915,
+      "step": 11380
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 8.195169082125604e-05,
+      "loss": 0.8543,
+      "step": 11390
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 8.175845410628018e-05,
+      "loss": 0.8812,
+      "step": 11400
+    },
+    {
+      "epoch": 3.65,
+      "eval_loss": 1.1034456491470337,
+      "eval_runtime": 105.6832,
+      "eval_samples_per_second": 18.924,
+      "eval_steps_per_second": 2.366,
+      "step": 11400
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 8.156521739130435e-05,
+      "loss": 0.9243,
+      "step": 11410
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 8.137198067632849e-05,
+      "loss": 0.8573,
+      "step": 11420
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.117874396135266e-05,
+      "loss": 0.9056,
+      "step": 11430
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.09855072463768e-05,
+      "loss": 0.9333,
+      "step": 11440
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.079227053140096e-05,
+      "loss": 0.9058,
+      "step": 11450
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 8.059903381642512e-05,
+      "loss": 0.8713,
+      "step": 11460
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 8.040579710144927e-05,
+      "loss": 0.878,
+      "step": 11470
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 8.021256038647343e-05,
+      "loss": 0.8883,
+      "step": 11480
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 8.001932367149757e-05,
+      "loss": 0.888,
+      "step": 11490
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 7.982608695652174e-05,
+      "loss": 0.9062,
+      "step": 11500
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 7.963285024154588e-05,
+      "loss": 0.9023,
+      "step": 11510
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 7.943961352657004e-05,
+      "loss": 0.912,
+      "step": 11520
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 7.92463768115942e-05,
+      "loss": 0.8954,
+      "step": 11530
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 7.905314009661835e-05,
+      "loss": 0.8716,
+      "step": 11540
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 7.885990338164251e-05,
+      "loss": 0.8667,
+      "step": 11550
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 7.866666666666666e-05,
+      "loss": 0.8903,
+      "step": 11560
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 7.847342995169082e-05,
+      "loss": 0.8682,
+      "step": 11570
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 7.828019323671496e-05,
+      "loss": 0.969,
+      "step": 11580
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 7.808695652173912e-05,
+      "loss": 0.8721,
+      "step": 11590
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 7.789371980676327e-05,
+      "loss": 0.8605,
+      "step": 11600
+    },
+    {
+      "epoch": 3.71,
+      "eval_loss": 1.1016223430633545,
+      "eval_runtime": 105.5836,
+      "eval_samples_per_second": 18.942,
+      "eval_steps_per_second": 2.368,
+      "step": 11600
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 7.770048309178743e-05,
+      "loss": 0.8777,
+      "step": 11610
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 7.750724637681159e-05,
+      "loss": 0.8876,
+      "step": 11620
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 7.731400966183574e-05,
+      "loss": 0.8854,
+      "step": 11630
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 7.71207729468599e-05,
+      "loss": 0.8464,
+      "step": 11640
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 7.692753623188405e-05,
+      "loss": 0.8346,
+      "step": 11650
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 7.673429951690821e-05,
+      "loss": 0.9311,
+      "step": 11660
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 7.654106280193237e-05,
+      "loss": 0.8562,
+      "step": 11670
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 7.634782608695651e-05,
+      "loss": 0.8871,
+      "step": 11680
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 7.615458937198067e-05,
+      "loss": 0.8344,
+      "step": 11690
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 7.596135265700482e-05,
+      "loss": 0.9221,
+      "step": 11700
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 7.576811594202898e-05,
+      "loss": 0.8681,
+      "step": 11710
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 7.557487922705313e-05,
+      "loss": 0.8834,
+      "step": 11720
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 7.538164251207729e-05,
+      "loss": 0.8816,
+      "step": 11730
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 7.518840579710145e-05,
+      "loss": 0.9228,
+      "step": 11740
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 7.499516908212559e-05,
+      "loss": 0.9043,
+      "step": 11750
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 7.480193236714974e-05,
+      "loss": 0.86,
+      "step": 11760
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 7.46086956521739e-05,
+      "loss": 0.8998,
+      "step": 11770
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 7.441545893719806e-05,
+      "loss": 0.9084,
+      "step": 11780
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 7.422222222222221e-05,
+      "loss": 0.8533,
+      "step": 11790
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 7.402898550724637e-05,
+      "loss": 0.915,
+      "step": 11800
+    },
+    {
+      "epoch": 3.78,
+      "eval_loss": 1.1024399995803833,
+      "eval_runtime": 105.6515,
+      "eval_samples_per_second": 18.93,
+      "eval_steps_per_second": 2.366,
+      "step": 11800
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 7.383574879227053e-05,
+      "loss": 0.9233,
+      "step": 11810
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 7.364251207729468e-05,
+      "loss": 0.8587,
+      "step": 11820
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 7.344927536231884e-05,
+      "loss": 0.8828,
+      "step": 11830
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 7.325603864734298e-05,
+      "loss": 0.8618,
+      "step": 11840
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 7.306280193236714e-05,
+      "loss": 0.8875,
+      "step": 11850
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 7.286956521739129e-05,
+      "loss": 0.8818,
+      "step": 11860
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 7.267632850241545e-05,
+      "loss": 0.9362,
+      "step": 11870
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 7.24830917874396e-05,
+      "loss": 0.9056,
+      "step": 11880
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 7.228985507246376e-05,
+      "loss": 0.8877,
+      "step": 11890
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 7.209661835748792e-05,
+      "loss": 0.8903,
+      "step": 11900
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 7.190338164251207e-05,
+      "loss": 0.9009,
+      "step": 11910
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 7.171014492753623e-05,
+      "loss": 0.8953,
+      "step": 11920
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 7.151690821256039e-05,
+      "loss": 0.8667,
+      "step": 11930
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 7.132367149758453e-05,
+      "loss": 0.8662,
+      "step": 11940
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 7.113043478260868e-05,
+      "loss": 0.8844,
+      "step": 11950
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 7.093719806763284e-05,
+      "loss": 0.8586,
+      "step": 11960
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 7.0743961352657e-05,
+      "loss": 0.8625,
+      "step": 11970
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 7.055072463768115e-05,
+      "loss": 0.9116,
+      "step": 11980
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 7.035748792270531e-05,
+      "loss": 0.9215,
+      "step": 11990
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 7.016425120772946e-05,
+      "loss": 0.9084,
+      "step": 12000
+    },
+    {
+      "epoch": 3.84,
+      "eval_loss": 1.102295994758606,
+      "eval_runtime": 105.6367,
+      "eval_samples_per_second": 18.933,
+      "eval_steps_per_second": 2.367,
+      "step": 12000
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 6.997101449275362e-05,
+      "loss": 0.8855,
+      "step": 12010
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 6.977777777777778e-05,
+      "loss": 0.8989,
+      "step": 12020
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 6.958454106280193e-05,
+      "loss": 0.8573,
+      "step": 12030
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 6.939130434782609e-05,
+      "loss": 0.8793,
+      "step": 12040
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 6.919806763285023e-05,
+      "loss": 0.9006,
+      "step": 12050
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 6.900483091787439e-05,
+      "loss": 0.9057,
+      "step": 12060
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 6.881159420289854e-05,
+      "loss": 0.8872,
+      "step": 12070
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 6.86183574879227e-05,
+      "loss": 0.8857,
+      "step": 12080
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 6.842512077294686e-05,
+      "loss": 0.8748,
+      "step": 12090
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 6.823188405797101e-05,
+      "loss": 0.8398,
+      "step": 12100
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 6.803864734299517e-05,
+      "loss": 0.9099,
+      "step": 12110
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 6.784541062801932e-05,
+      "loss": 0.9245,
+      "step": 12120
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 6.765217391304348e-05,
+      "loss": 0.9001,
+      "step": 12130
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 6.745893719806764e-05,
+      "loss": 0.892,
+      "step": 12140
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 6.726570048309178e-05,
+      "loss": 0.8583,
+      "step": 12150
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 6.707246376811594e-05,
+      "loss": 0.8508,
+      "step": 12160
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 6.687922705314009e-05,
+      "loss": 0.893,
+      "step": 12170
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 6.668599033816425e-05,
+      "loss": 0.9071,
+      "step": 12180
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 6.64927536231884e-05,
+      "loss": 0.8679,
+      "step": 12190
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 6.629951690821256e-05,
+      "loss": 0.8924,
+      "step": 12200
+    },
+    {
+      "epoch": 3.9,
+      "eval_loss": 1.1033905744552612,
+      "eval_runtime": 105.7703,
+      "eval_samples_per_second": 18.909,
+      "eval_steps_per_second": 2.364,
+      "step": 12200
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 6.610628019323672e-05,
+      "loss": 0.9134,
+      "step": 12210
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 6.591304347826087e-05,
+      "loss": 0.8522,
+      "step": 12220
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 6.571980676328501e-05,
+      "loss": 0.8552,
+      "step": 12230
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 6.552657004830917e-05,
+      "loss": 0.8762,
+      "step": 12240
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 6.533333333333333e-05,
+      "loss": 0.8451,
+      "step": 12250
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 6.514009661835748e-05,
+      "loss": 0.8675,
+      "step": 12260
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 6.494685990338164e-05,
+      "loss": 0.8734,
+      "step": 12270
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 6.47536231884058e-05,
+      "loss": 0.9088,
+      "step": 12280
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 6.456038647342995e-05,
+      "loss": 0.8844,
+      "step": 12290
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 6.436714975845411e-05,
+      "loss": 0.859,
+      "step": 12300
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 6.417391304347825e-05,
+      "loss": 0.8829,
+      "step": 12310
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 6.39806763285024e-05,
+      "loss": 0.8861,
+      "step": 12320
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 6.378743961352656e-05,
+      "loss": 0.8807,
+      "step": 12330
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 6.359420289855072e-05,
+      "loss": 0.8704,
+      "step": 12340
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 6.340096618357487e-05,
+      "loss": 0.8682,
+      "step": 12350
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 6.320772946859903e-05,
+      "loss": 0.907,
+      "step": 12360
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 6.301449275362319e-05,
+      "loss": 0.8967,
+      "step": 12370
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 6.282125603864733e-05,
+      "loss": 0.9041,
+      "step": 12380
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 6.262801932367149e-05,
+      "loss": 0.8572,
+      "step": 12390
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 6.243478260869564e-05,
+      "loss": 0.8666,
+      "step": 12400
+    },
+    {
+      "epoch": 3.97,
+      "eval_loss": 1.1044241189956665,
+      "eval_runtime": 105.816,
+      "eval_samples_per_second": 18.901,
+      "eval_steps_per_second": 2.363,
+      "step": 12400
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 6.22415458937198e-05,
+      "loss": 0.867,
+      "step": 12410
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 6.204830917874395e-05,
+      "loss": 0.8644,
+      "step": 12420
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 6.185507246376811e-05,
+      "loss": 0.8764,
+      "step": 12430
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 6.166183574879227e-05,
+      "loss": 0.9048,
+      "step": 12440
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 6.146859903381642e-05,
+      "loss": 0.9189,
+      "step": 12450
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 6.127536231884056e-05,
+      "loss": 0.8635,
+      "step": 12460
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 6.108212560386472e-05,
+      "loss": 0.8915,
+      "step": 12470
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 6.0888888888888883e-05,
+      "loss": 0.8982,
+      "step": 12480
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.069565217391304e-05,
+      "loss": 0.8921,
+      "step": 12490
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.0502415458937196e-05,
+      "loss": 0.8547,
+      "step": 12500
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.0309178743961345e-05,
+      "loss": 0.835,
+      "step": 12510
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.01159420289855e-05,
+      "loss": 0.8325,
+      "step": 12520
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 5.992270531400966e-05,
+      "loss": 0.8735,
+      "step": 12530
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 5.972946859903381e-05,
+      "loss": 0.8042,
+      "step": 12540
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.953623188405797e-05,
+      "loss": 0.8002,
+      "step": 12550
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.934299516908212e-05,
+      "loss": 0.817,
+      "step": 12560
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.9149758454106275e-05,
+      "loss": 0.7889,
+      "step": 12570
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.8956521739130424e-05,
+      "loss": 0.8434,
+      "step": 12580
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.876328502415458e-05,
+      "loss": 0.8364,
+      "step": 12590
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.8570048309178736e-05,
+      "loss": 0.8324,
+      "step": 12600
+    },
+    {
+      "epoch": 4.03,
+      "eval_loss": 1.128263235092163,
+      "eval_runtime": 144.6511,
+      "eval_samples_per_second": 13.826,
+      "eval_steps_per_second": 1.728,
+      "step": 12600
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.837681159420289e-05,
+      "loss": 0.812,
+      "step": 12610
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.818357487922705e-05,
+      "loss": 0.8342,
+      "step": 12620
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.7990338164251205e-05,
+      "loss": 0.8237,
+      "step": 12630
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.779710144927536e-05,
+      "loss": 0.8634,
+      "step": 12640
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.760386473429952e-05,
+      "loss": 0.8526,
+      "step": 12650
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.741062801932367e-05,
+      "loss": 0.8189,
+      "step": 12660
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.7217391304347816e-05,
+      "loss": 0.8458,
+      "step": 12670
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 5.702415458937197e-05,
+      "loss": 0.8197,
+      "step": 12680
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 5.683091787439613e-05,
+      "loss": 0.8548,
+      "step": 12690
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 5.6637681159420284e-05,
+      "loss": 0.8136,
+      "step": 12700
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 5.644444444444444e-05,
+      "loss": 0.7845,
+      "step": 12710
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 5.6251207729468596e-05,
+      "loss": 0.8379,
+      "step": 12720
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 5.605797101449275e-05,
+      "loss": 0.835,
+      "step": 12730
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 5.586473429951691e-05,
+      "loss": 0.7949,
+      "step": 12740
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 5.567149758454106e-05,
+      "loss": 0.8672,
+      "step": 12750
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 5.5478260869565214e-05,
+      "loss": 0.8598,
+      "step": 12760
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 5.528502415458936e-05,
+      "loss": 0.851,
+      "step": 12770
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 5.509178743961352e-05,
+      "loss": 0.8648,
+      "step": 12780
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 5.4898550724637675e-05,
+      "loss": 0.8189,
+      "step": 12790
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.470531400966183e-05,
+      "loss": 0.803,
+      "step": 12800
+    },
+    {
+      "epoch": 4.1,
+      "eval_loss": 1.1351523399353027,
+      "eval_runtime": 144.5863,
+      "eval_samples_per_second": 13.833,
+      "eval_steps_per_second": 1.729,
+      "step": 12800
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.451207729468599e-05,
+      "loss": 0.8366,
+      "step": 12810
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.4318840579710144e-05,
+      "loss": 0.7907,
+      "step": 12820
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.412560386473429e-05,
+      "loss": 0.8722,
+      "step": 12830
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.393236714975845e-05,
+      "loss": 0.7995,
+      "step": 12840
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.3739130434782605e-05,
+      "loss": 0.8069,
+      "step": 12850
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.354589371980676e-05,
+      "loss": 0.8203,
+      "step": 12860
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 5.335265700483091e-05,
+      "loss": 0.8192,
+      "step": 12870
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 5.315942028985507e-05,
+      "loss": 0.8393,
+      "step": 12880
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 5.296618357487922e-05,
+      "loss": 0.8543,
+      "step": 12890
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 5.277294685990337e-05,
+      "loss": 0.863,
+      "step": 12900
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 5.257971014492753e-05,
+      "loss": 0.7979,
+      "step": 12910
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 5.2386473429951685e-05,
+      "loss": 0.8627,
+      "step": 12920
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 5.219323671497584e-05,
+      "loss": 0.8104,
+      "step": 12930
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 5.2e-05,
+      "loss": 0.8331,
+      "step": 12940
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 5.180676328502415e-05,
+      "loss": 0.79,
+      "step": 12950
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 5.161352657004831e-05,
+      "loss": 0.8166,
+      "step": 12960
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 5.142028985507246e-05,
+      "loss": 0.8166,
+      "step": 12970
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 5.122705314009661e-05,
+      "loss": 0.8314,
+      "step": 12980
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 5.1033816425120764e-05,
+      "loss": 0.807,
+      "step": 12990
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 5.084057971014492e-05,
+      "loss": 0.8223,
+      "step": 13000
+    },
+    {
+      "epoch": 4.16,
+      "eval_loss": 1.1353418827056885,
+      "eval_runtime": 141.9713,
+      "eval_samples_per_second": 14.087,
+      "eval_steps_per_second": 1.761,
+      "step": 13000
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 5.0647342995169076e-05,
+      "loss": 0.8345,
+      "step": 13010
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 5.045410628019323e-05,
+      "loss": 0.8237,
+      "step": 13020
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 5.026086956521739e-05,
+      "loss": 0.8273,
+      "step": 13030
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 5.0067632850241544e-05,
+      "loss": 0.8312,
+      "step": 13040
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 4.98743961352657e-05,
+      "loss": 0.8002,
+      "step": 13050
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 4.9681159420289857e-05,
+      "loss": 0.8421,
+      "step": 13060
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 4.9487922705314006e-05,
+      "loss": 0.8292,
+      "step": 13070
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.9294685990338155e-05,
+      "loss": 0.8,
+      "step": 13080
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.910144927536231e-05,
+      "loss": 0.8483,
+      "step": 13090
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.890821256038647e-05,
+      "loss": 0.8095,
+      "step": 13100
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.8714975845410624e-05,
+      "loss": 0.847,
+      "step": 13110
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 4.852173913043478e-05,
+      "loss": 0.849,
+      "step": 13120
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 4.8328502415458936e-05,
+      "loss": 0.7999,
+      "step": 13130
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 4.813526570048309e-05,
+      "loss": 0.8007,
+      "step": 13140
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 4.794202898550724e-05,
+      "loss": 0.8408,
+      "step": 13150
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 4.77487922705314e-05,
+      "loss": 0.8149,
+      "step": 13160
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 4.7555555555555553e-05,
+      "loss": 0.8488,
+      "step": 13170
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 4.73623188405797e-05,
+      "loss": 0.8104,
+      "step": 13180
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 4.716908212560386e-05,
+      "loss": 0.7863,
+      "step": 13190
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 4.6975845410628015e-05,
+      "loss": 0.8401,
+      "step": 13200
+    },
+    {
+      "epoch": 4.22,
+      "eval_loss": 1.139245867729187,
+      "eval_runtime": 145.0946,
+      "eval_samples_per_second": 13.784,
+      "eval_steps_per_second": 1.723,
+      "step": 13200
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 15625,
+  "num_train_epochs": 5,
+  "save_steps": 200,
+  "total_flos": 1.4568502076111585e+18,
+  "trial_name": null,
+  "trial_params": null
+}