diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10535 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.999956610404825,
+  "global_step": 17285,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0,
+      "loss": 8.2501,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.936416184971099e-07,
+      "loss": 7.9965,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.0057803468208094e-06,
+      "loss": 8.0268,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.317919075144509e-06,
+      "loss": 7.9878,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.398843930635839e-06,
+      "loss": 7.6933,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.710982658959537e-06,
+      "loss": 7.4624,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2023121387283238e-05,
+      "loss": 7.3907,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4335260115606938e-05,
+      "loss": 7.0312,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6647398843930635e-05,
+      "loss": 6.6125,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8959537572254336e-05,
+      "loss": 6.3013,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1271676300578036e-05,
+      "loss": 6.068,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3583815028901734e-05,
+      "loss": 5.8308,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5895953757225434e-05,
+      "loss": 5.7656,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8208092485549138e-05,
+      "loss": 5.5955,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0520231213872835e-05,
+      "loss": 5.3842,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.283236994219653e-05,
+      "loss": 5.2866,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.514450867052023e-05,
+      "loss": 5.0532,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.7456647398843934e-05,
+      "loss": 4.9624,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.976878612716764e-05,
+      "loss": 4.8342,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.2080924855491335e-05,
+      "loss": 4.6055,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.439306358381503e-05,
+      "loss": 4.5276,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.670520231213873e-05,
+      "loss": 4.3676,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9017341040462426e-05,
+      "loss": 4.2029,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.1329479768786124e-05,
+      "loss": 4.0336,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.364161849710983e-05,
+      "loss": 3.8993,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.595375722543353e-05,
+      "loss": 3.834,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.8265895953757235e-05,
+      "loss": 3.7466,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.057803468208093e-05,
+      "loss": 3.6144,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.289017341040462e-05,
+      "loss": 3.4977,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.520231213872833e-05,
+      "loss": 3.4428,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.751445086705203e-05,
+      "loss": 3.2823,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.982658959537573e-05,
+      "loss": 3.2296,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 7.213872832369943e-05,
+      "loss": 3.1029,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 7.445086705202312e-05,
+      "loss": 3.1717,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 7.676300578034682e-05,
+      "loss": 3.0968,
+      "step": 340
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 7.907514450867053e-05,
+      "loss": 2.9926,
+      "step": 350
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 8.138728323699423e-05,
+      "loss": 3.0021,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 8.369942196531792e-05,
+      "loss": 2.891,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 8.601156069364162e-05,
+      "loss": 2.8498,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 8.832369942196532e-05,
+      "loss": 2.8172,
+      "step": 390
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.063583815028902e-05,
+      "loss": 2.8302,
+      "step": 400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.294797687861271e-05,
+      "loss": 2.7333,
+      "step": 410
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.526011560693642e-05,
+      "loss": 2.7135,
+      "step": 420
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.757225433526012e-05,
+      "loss": 2.6811,
+      "step": 430
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.988439306358382e-05,
+      "loss": 2.6537,
+      "step": 440
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00010219653179190752,
+      "loss": 2.6031,
+      "step": 450
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00010450867052023121,
+      "loss": 2.6037,
+      "step": 460
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00010682080924855491,
+      "loss": 2.5387,
+      "step": 470
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00010913294797687861,
+      "loss": 2.5393,
+      "step": 480
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00011144508670520233,
+      "loss": 2.5387,
+      "step": 490
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00011375722543352603,
+      "loss": 2.4848,
+      "step": 500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00011606936416184973,
+      "loss": 2.4773,
+      "step": 510
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00011838150289017342,
+      "loss": 2.4453,
+      "step": 520
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00012069364161849712,
+      "loss": 2.3941,
+      "step": 530
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00012300578034682083,
+      "loss": 2.431,
+      "step": 540
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00012531791907514453,
+      "loss": 2.4208,
+      "step": 550
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00012763005780346823,
+      "loss": 2.4313,
+      "step": 560
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00012971098265895952,
+      "loss": 2.3427,
+      "step": 570
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00013202312138728322,
+      "loss": 2.3415,
+      "step": 580
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00013433526011560694,
+      "loss": 2.2621,
+      "step": 590
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00013664739884393064,
+      "loss": 2.3606,
+      "step": 600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00013895953757225434,
+      "loss": 2.3175,
+      "step": 610
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00014127167630057804,
+      "loss": 2.2297,
+      "step": 620
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00014358381502890176,
+      "loss": 2.1856,
+      "step": 630
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00014589595375722546,
+      "loss": 2.2633,
+      "step": 640
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00014820809248554915,
+      "loss": 2.2474,
+      "step": 650
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00015052023121387285,
+      "loss": 2.2024,
+      "step": 660
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00015260115606936415,
+      "loss": 2.1947,
+      "step": 670
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00015491329479768785,
+      "loss": 2.234,
+      "step": 680
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00015722543352601157,
+      "loss": 2.2061,
+      "step": 690
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001593063583815029,
+      "loss": 2.1553,
+      "step": 700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001616184971098266,
+      "loss": 2.2286,
+      "step": 710
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001639306358381503,
+      "loss": 2.161,
+      "step": 720
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000166242774566474,
+      "loss": 2.1628,
+      "step": 730
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00016855491329479768,
+      "loss": 2.1371,
+      "step": 740
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017063583815028904,
+      "loss": 2.2181,
+      "step": 750
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017271676300578033,
+      "loss": 2.154,
+      "step": 760
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017502890173410406,
+      "loss": 2.1695,
+      "step": 770
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017734104046242776,
+      "loss": 2.1685,
+      "step": 780
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017965317919075145,
+      "loss": 2.192,
+      "step": 790
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00018196531791907515,
+      "loss": 2.1567,
+      "step": 800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00018427745664739887,
+      "loss": 2.0987,
+      "step": 810
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00018658959537572257,
+      "loss": 2.1687,
+      "step": 820
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00018890173410404627,
+      "loss": 2.0736,
+      "step": 830
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019121387283236997,
+      "loss": 2.0683,
+      "step": 840
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019352601156069366,
+      "loss": 2.0872,
+      "step": 850
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019583815028901736,
+      "loss": 2.1029,
+      "step": 860
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019815028901734106,
+      "loss": 2.0301,
+      "step": 870
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019999999267878048,
+      "loss": 2.0957,
+      "step": 880
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001999997364362091,
+      "loss": 2.0484,
+      "step": 890
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019999911413373273,
+      "loss": 2.0489,
+      "step": 900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019999812577362934,
+      "loss": 2.0073,
+      "step": 910
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001999967713595169,
+      "loss": 2.0241,
+      "step": 920
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019999505089635347,
+      "loss": 2.0097,
+      "step": 930
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001999929643904369,
+      "loss": 2.0251,
+      "step": 940
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019999051184940516,
+      "loss": 1.9893,
+      "step": 950
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019998769328223598,
+      "loss": 1.9893,
+      "step": 960
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019998450869924703,
+      "loss": 1.9321,
+      "step": 970
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019998095811209587,
+      "loss": 2.0008,
+      "step": 980
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019997704153377978,
+      "loss": 2.0254,
+      "step": 990
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001999727589786358,
+      "loss": 1.873,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 1.0050371885299683,
+      "eval_runtime": 62.5449,
+      "eval_samples_per_second": 8.378,
+      "eval_steps_per_second": 0.528,
+      "step": 1000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019996811046234077,
+      "loss": 1.9664,
+      "step": 1010
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019996309600191098,
+      "loss": 1.9404,
+      "step": 1020
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019995771561570248,
+      "loss": 1.969,
+      "step": 1030
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019995196932341073,
+      "loss": 1.9545,
+      "step": 1040
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019994585714607066,
+      "loss": 1.9141,
+      "step": 1050
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019993937910605658,
+      "loss": 1.9299,
+      "step": 1060
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019993253522708205,
+      "loss": 1.9305,
+      "step": 1070
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001999253255341998,
+      "loss": 1.902,
+      "step": 1080
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019991775005380173,
+      "loss": 1.9416,
+      "step": 1090
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019991061939600934,
+      "loss": 1.9164,
+      "step": 1100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019990234899683635,
+      "loss": 1.947,
+      "step": 1110
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019989371289425568,
+      "loss": 1.9242,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019988471111988062,
+      "loss": 1.9037,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019987534370666328,
+      "loss": 1.915,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.000199865610688894,
+      "loss": 1.9268,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019985551210220158,
+      "loss": 1.9268,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019984611084327463,
+      "loss": 1.9629,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019983531777857815,
+      "loss": 1.854,
+      "step": 1180
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019982415925584902,
+      "loss": 1.9051,
+      "step": 1190
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019981263531593422,
+      "loss": 1.8801,
+      "step": 1200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019980074600101837,
+      "loss": 1.8322,
+      "step": 1210
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019978849135462366,
+      "loss": 1.8857,
+      "step": 1220
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019977587142160945,
+      "loss": 1.8805,
+      "step": 1230
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019976288624817248,
+      "loss": 1.8511,
+      "step": 1240
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019974953588184632,
+      "loss": 1.8872,
+      "step": 1250
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019973582037150148,
+      "loss": 1.8636,
+      "step": 1260
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019972173976734507,
+      "loss": 1.8701,
+      "step": 1270
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019970729412092063,
+      "loss": 1.8454,
+      "step": 1280
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019969248348510808,
+      "loss": 1.8941,
+      "step": 1290
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019967730791412328,
+      "loss": 1.8561,
+      "step": 1300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019966176746351818,
+      "loss": 1.8992,
+      "step": 1310
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019964586219018018,
+      "loss": 1.8372,
+      "step": 1320
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001996295921523323,
+      "loss": 1.8278,
+      "step": 1330
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019961295740953278,
+      "loss": 1.8311,
+      "step": 1340
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019959595802267492,
+      "loss": 1.8281,
+      "step": 1350
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001995785940539868,
+      "loss": 1.8188,
+      "step": 1360
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019956086556703113,
+      "loss": 1.8156,
+      "step": 1370
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019954277262670495,
+      "loss": 1.7751,
+      "step": 1380
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019952431529923949,
+      "loss": 1.832,
+      "step": 1390
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019950549365219968,
+      "loss": 1.8475,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019948630775448433,
+      "loss": 1.8329,
+      "step": 1410
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019946675767632544,
+      "loss": 1.8352,
+      "step": 1420
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019944684348928822,
+      "loss": 1.8325,
+      "step": 1430
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019942860946808643,
+      "loss": 1.8484,
+      "step": 1440
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019940800367611585,
+      "loss": 1.837,
+      "step": 1450
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019938703399034234,
+      "loss": 1.8295,
+      "step": 1460
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019936570048752775,
+      "loss": 1.8153,
+      "step": 1470
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019934400324576564,
+      "loss": 1.7925,
+      "step": 1480
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001993219423444811,
+      "loss": 1.8383,
+      "step": 1490
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001992995178644305,
+      "loss": 1.8135,
+      "step": 1500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019927672988770105,
+      "loss": 1.8036,
+      "step": 1510
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019925357849771066,
+      "loss": 1.8035,
+      "step": 1520
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019923243159839809,
+      "loss": 1.8135,
+      "step": 1530
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019920858995779232,
+      "loss": 1.7839,
+      "step": 1540
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019918438515335927,
+      "loss": 1.7759,
+      "step": 1550
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019915981727370316,
+      "loss": 1.7933,
+      "step": 1560
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019913488640875744,
+      "loss": 1.7977,
+      "step": 1570
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019910959264978422,
+      "loss": 1.7797,
+      "step": 1580
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019908393608937406,
+      "loss": 1.7656,
+      "step": 1590
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019905791682144557,
+      "loss": 1.798,
+      "step": 1600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019903153494124518,
+      "loss": 1.7618,
+      "step": 1610
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019900479054534652,
+      "loss": 1.7879,
+      "step": 1620
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019897768373165046,
+      "loss": 1.7972,
+      "step": 1630
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019895297781409127,
+      "loss": 1.7738,
+      "step": 1640
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019892518268104788,
+      "loss": 1.7901,
+      "step": 1650
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019889702542162026,
+      "loss": 1.7489,
+      "step": 1660
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019887137435523912,
+      "loss": 1.8051,
+      "step": 1670
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019884252934074216,
+      "loss": 1.7428,
+      "step": 1680
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001988133225024225,
+      "loss": 1.7961,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019878375394719502,
+      "loss": 1.7779,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019875382378329857,
+      "loss": 1.8037,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001987235321202958,
+      "loss": 1.7767,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019869287906907265,
+      "loss": 1.8044,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001986618647418379,
+      "loss": 1.7517,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001986304892521229,
+      "loss": 1.8253,
+      "step": 1750
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019859875271478102,
+      "loss": 1.7588,
+      "step": 1760
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019856665524598733,
+      "loss": 1.7948,
+      "step": 1770
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019853419696323806,
+      "loss": 1.8023,
+      "step": 1780
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019850467611100676,
+      "loss": 1.7663,
+      "step": 1790
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019847153261017426,
+      "loss": 1.7566,
+      "step": 1800
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019843802864359298,
+      "loss": 1.7882,
+      "step": 1810
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019840416433390782,
+      "loss": 1.782,
+      "step": 1820
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019836993980508268,
+      "loss": 1.7849,
+      "step": 1830
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019833535518240031,
+      "loss": 1.7793,
+      "step": 1840
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001983004105924614,
+      "loss": 1.7761,
+      "step": 1850
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019826865279520944,
+      "loss": 1.7676,
+      "step": 1860
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001982330246209872,
+      "loss": 1.7274,
+      "step": 1870
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019819703685410058,
+      "loss": 1.7513,
+      "step": 1880
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001981606896262867,
+      "loss": 1.7478,
+      "step": 1890
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019812398307059856,
+      "loss": 1.781,
+      "step": 1900
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019808691732140448,
+      "loss": 1.7504,
+      "step": 1910
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019804949251438767,
+      "loss": 1.7552,
+      "step": 1920
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001980117087865457,
+      "loss": 1.8154,
+      "step": 1930
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019797356627619,
+      "loss": 1.7762,
+      "step": 1940
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019793506512294542,
+      "loss": 1.7263,
+      "step": 1950
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019789620546774956,
+      "loss": 1.7446,
+      "step": 1960
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019786485971773587,
+      "loss": 1.694,
+      "step": 1970
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001978253551183793,
+      "loss": 1.7198,
+      "step": 1980
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019778549241867687,
+      "loss": 1.7423,
+      "step": 1990
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001977452717645503,
+      "loss": 1.7434,
+      "step": 2000
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.90117347240448,
+      "eval_runtime": 61.8399,
+      "eval_samples_per_second": 8.473,
+      "eval_steps_per_second": 0.534,
+      "step": 2000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019770469330323174,
+      "loss": 1.7791,
+      "step": 2010
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019766375718326297,
+      "loss": 1.7459,
+      "step": 2020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019762246355449516,
+      "loss": 1.7342,
+      "step": 2030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019758081256808816,
+      "loss": 1.7564,
+      "step": 2040
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019753880437650985,
+      "loss": 1.7394,
+      "step": 2050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019749643913353582,
+      "loss": 1.7663,
+      "step": 2060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019745371699424864,
+      "loss": 1.7222,
+      "step": 2070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019741063811503734,
+      "loss": 1.7046,
+      "step": 2080
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001973759182648501,
+      "loss": 1.6904,
+      "step": 2090
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019733219765204383,
+      "loss": 1.6956,
+      "step": 2100
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019728812074414819,
+      "loss": 1.7511,
+      "step": 2110
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019724368770251155,
+      "loss": 1.7262,
+      "step": 2120
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001971988986897858,
+      "loss": 1.7114,
+      "step": 2130
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019715375386992608,
+      "loss": 1.7182,
+      "step": 2140
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019710825340818987,
+      "loss": 1.7034,
+      "step": 2150
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019706239747113656,
+      "loss": 1.7282,
+      "step": 2160
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019701618622662678,
+      "loss": 1.74,
+      "step": 2170
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019696961984382182,
+      "loss": 1.6854,
+      "step": 2180
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019692269849318303,
+      "loss": 1.756,
+      "step": 2190
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019687542234647106,
+      "loss": 1.7159,
+      "step": 2200
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019682779157674537,
+      "loss": 1.7095,
+      "step": 2210
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019677980635836363,
+      "loss": 1.7071,
+      "step": 2220
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019673146686698093,
+      "loss": 1.7077,
+      "step": 2230
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019668277327954917,
+      "loss": 1.7144,
+      "step": 2240
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019663372577431663,
+      "loss": 1.6873,
+      "step": 2250
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001965892805682537,
+      "loss": 1.7081,
+      "step": 2260
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019653956111491275,
+      "loss": 1.6979,
+      "step": 2270
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019648948826801467,
+      "loss": 1.698,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019644412070578336,
+      "loss": 1.692,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019639337691717884,
+      "loss": 1.6938,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019634228027014033,
+      "loss": 1.7152,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019629083095171264,
+      "loss": 1.7155,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001962390291502316,
+      "loss": 1.6839,
+      "step": 2330
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019618687505532334,
+      "loss": 1.6888,
+      "step": 2340
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001961343688579036,
+      "loss": 1.6962,
+      "step": 2350
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000196081510750177,
+      "loss": 1.6784,
+      "step": 2360
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019602830092563643,
+      "loss": 1.672,
+      "step": 2370
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019597473957906224,
+      "loss": 1.6769,
+      "step": 2380
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019592082690652148,
+      "loss": 1.6975,
+      "step": 2390
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019586656310536743,
+      "loss": 1.7687,
+      "step": 2400
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019581194837423857,
+      "loss": 1.685,
+      "step": 2410
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019575698291305813,
+      "loss": 1.6858,
+      "step": 2420
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001957016669230331,
+      "loss": 1.6883,
+      "step": 2430
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019565158299718013,
+      "loss": 1.6471,
+      "step": 2440
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001955956015612708,
+      "loss": 1.6831,
+      "step": 2450
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001955392701872709,
+      "loss": 1.6686,
+      "step": 2460
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019548258908138753,
+      "loss": 1.7006,
+      "step": 2470
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019542555845110805,
+      "loss": 1.7317,
+      "step": 2480
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019536817850519927,
+      "loss": 1.6572,
+      "step": 2490
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001953104494537067,
+      "loss": 1.6916,
+      "step": 2500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001952523715079538,
+      "loss": 1.6533,
+      "step": 2510
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019519394488054127,
+      "loss": 1.6463,
+      "step": 2520
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019513516978534608,
+      "loss": 1.6984,
+      "step": 2530
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019508197443751353,
+      "loss": 1.6643,
+      "step": 2540
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001950284971627635,
+      "loss": 1.6853,
+      "step": 2550
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019496874750645754,
+      "loss": 1.6864,
+      "step": 2560
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019490865020672837,
+      "loss": 1.6562,
+      "step": 2570
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019484820548356873,
+      "loss": 1.6825,
+      "step": 2580
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019478741355824313,
+      "loss": 1.7296,
+      "step": 2590
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019472627465328692,
+      "loss": 1.7077,
+      "step": 2600
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001946647889925058,
+      "loss": 1.7098,
+      "step": 2610
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019460915560757066,
+      "loss": 1.6647,
+      "step": 2620
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.000194547011731852,
+      "loss": 1.6919,
+      "step": 2630
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00019448452175651983,
+      "loss": 1.6805,
+      "step": 2640
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001944216859103255,
+      "loss": 1.7276,
+      "step": 2650
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00019435850442328637,
+      "loss": 1.6987,
+      "step": 2660
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00019429497752668516,
+      "loss": 1.6923,
+      "step": 2670
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00019423110545306908,
+      "loss": 1.6908,
+      "step": 2680
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00019416688843624873,
+      "loss": 1.6799,
+      "step": 2690
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00019410232671129745,
+      "loss": 1.7065,
+      "step": 2700
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001940374205145505,
+      "loss": 1.682,
+      "step": 2710
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00019397217008360404,
+      "loss": 1.654,
+      "step": 2720
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001939065756573144,
+      "loss": 1.6809,
+      "step": 2730
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019384063747579706,
+      "loss": 1.6426,
+      "step": 2740
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019377435578042592,
+      "loss": 1.6453,
+      "step": 2750
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019370773081383235,
+      "loss": 1.6419,
+      "step": 2760
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019364076281990427,
+      "loss": 1.7025,
+      "step": 2770
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001935734520437853,
+      "loss": 1.6897,
+      "step": 2780
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019350579873187384,
+      "loss": 1.652,
+      "step": 2790
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001934378031318222,
+      "loss": 1.6461,
+      "step": 2800
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00019336946549253567,
+      "loss": 1.6482,
+      "step": 2810
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00019330078606417164,
+      "loss": 1.684,
+      "step": 2820
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00019323176509813855,
+      "loss": 1.7247,
+      "step": 2830
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001931624028470952,
+      "loss": 1.6417,
+      "step": 2840
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00019309269956494963,
+      "loss": 1.688,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00019302967524028727,
+      "loss": 1.6498,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00019295932470303454,
+      "loss": 1.6872,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00019289571826614754,
+      "loss": 1.6668,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00019283183638479643,
+      "loss": 1.652,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00019276053369488895,
+      "loss": 1.6777,
+      "step": 2900
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.000192688891444965,
+      "loss": 1.6466,
+      "step": 2910
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019261690989727875,
+      "loss": 1.6432,
+      "step": 2920
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019254458931532655,
+      "loss": 1.6499,
+      "step": 2930
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019247192996384572,
+      "loss": 1.6599,
+      "step": 2940
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019239893210881373,
+      "loss": 1.6458,
+      "step": 2950
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019232559601744712,
+      "loss": 1.69,
+      "step": 2960
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019225192195820067,
+      "loss": 1.6294,
+      "step": 2970
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019217791020076627,
+      "loss": 1.7088,
+      "step": 2980
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.000192103561016072,
+      "loss": 1.6531,
+      "step": 2990
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019202887467628115,
+      "loss": 1.6708,
+      "step": 3000
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 0.862983226776123,
+      "eval_runtime": 61.7517,
+      "eval_samples_per_second": 8.486,
+      "eval_steps_per_second": 0.534,
+      "step": 3000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019195385145479116,
+      "loss": 1.6732,
+      "step": 3010
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001918784916262327,
+      "loss": 1.6961,
+      "step": 3020
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001918027954664686,
+      "loss": 1.6361,
+      "step": 3030
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019172676325259288,
+      "loss": 1.708,
+      "step": 3040
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019165039526292975,
+      "loss": 1.6377,
+      "step": 3050
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001915736917770325,
+      "loss": 1.667,
+      "step": 3060
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019149665307568263,
+      "loss": 1.6649,
+      "step": 3070
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019141927944088863,
+      "loss": 1.6981,
+      "step": 3080
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001913415711558851,
+      "loss": 1.6095,
+      "step": 3090
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00019127916377084718,
+      "loss": 1.6629,
+      "step": 3100
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00019120085383312737,
+      "loss": 1.6908,
+      "step": 3110
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00019112221004476872,
+      "loss": 1.6602,
+      "step": 3120
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00019104323269365537,
+      "loss": 1.6378,
+      "step": 3130
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00019096392206889248,
+      "loss": 1.642,
+      "step": 3140
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00019088427846080527,
+      "loss": 1.6605,
+      "step": 3150
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00019080430216093778,
+      "loss": 1.6055,
+      "step": 3160
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00019072399346205197,
+      "loss": 1.6423,
+      "step": 3170
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00019064335265812652,
+      "loss": 1.6856,
+      "step": 3180
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00019056238004435593,
+      "loss": 1.6115,
+      "step": 3190
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001904810759171492,
+      "loss": 1.6145,
+      "step": 3200
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019042396593693816,
+      "loss": 1.6301,
+      "step": 3210
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019034209892058318,
+      "loss": 1.6615,
+      "step": 3220
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019025990119715506,
+      "loss": 1.6515,
+      "step": 3230
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019017737306754754,
+      "loss": 1.7024,
+      "step": 3240
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019009451483386375,
+      "loss": 1.6598,
+      "step": 3250
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001900113267994153,
+      "loss": 1.6995,
+      "step": 3260
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00018992780926872102,
+      "loss": 1.684,
+      "step": 3270
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00018984396254750593,
+      "loss": 1.6553,
+      "step": 3280
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00018975978694270003,
+      "loss": 1.6515,
+      "step": 3290
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00018967528276243734,
+      "loss": 1.6754,
+      "step": 3300
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00018959045031605453,
+      "loss": 1.6483,
+      "step": 3310
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00018950528991409,
+      "loss": 1.6569,
+      "step": 3320
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00018941980186828263,
+      "loss": 1.6626,
+      "step": 3330
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001893339864915708,
+      "loss": 1.6508,
+      "step": 3340
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00018924784409809093,
+      "loss": 1.6662,
+      "step": 3350
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001891613750031767,
+      "loss": 1.6426,
+      "step": 3360
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00018907457952335754,
+      "loss": 1.6468,
+      "step": 3370
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001889874579763578,
+      "loss": 1.6326,
+      "step": 3380
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00018890001068109534,
+      "loss": 1.6034,
+      "step": 3390
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00018881223795768053,
+      "loss": 1.6951,
+      "step": 3400
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00018872414012741494,
+      "loss": 1.5691,
+      "step": 3410
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001886445743803333,
+      "loss": 1.6343,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00018857362860914253,
+      "loss": 1.6143,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00018848465460459042,
+      "loss": 1.663,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00018839535669234195,
+      "loss": 1.617,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00018830573519928195,
+      "loss": 1.6374,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001882157904534795,
+      "loss": 1.6472,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00018812552278418726,
+      "loss": 1.6326,
+      "step": 3480
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00018803493252183976,
+      "loss": 1.6444,
+      "step": 3490
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00018794401999805248,
+      "loss": 1.6167,
+      "step": 3500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00018785278554562065,
+      "loss": 1.6498,
+      "step": 3510
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00018776122949851792,
+      "loss": 1.6605,
+      "step": 3520
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00018766935219189507,
+      "loss": 1.6455,
+      "step": 3530
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00018757715396207903,
+      "loss": 1.671,
+      "step": 3540
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00018748463514657146,
+      "loss": 1.6176,
+      "step": 3550
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00018739179608404747,
+      "loss": 1.6459,
+      "step": 3560
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00018729863711435457,
+      "loss": 1.6481,
+      "step": 3570
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00018720515857851132,
+      "loss": 1.6823,
+      "step": 3580
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00018711136081870605,
+      "loss": 1.6239,
+      "step": 3590
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00018701724417829565,
+      "loss": 1.6209,
+      "step": 3600
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001869228090018043,
+      "loss": 1.6381,
+      "step": 3610
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00018682805563492225,
+      "loss": 1.6498,
+      "step": 3620
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00018673298442450448,
+      "loss": 1.6377,
+      "step": 3630
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00018663759571856952,
+      "loss": 1.6513,
+      "step": 3640
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001865514747131347,
+      "loss": 1.6385,
+      "step": 3650
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00018646509707450926,
+      "loss": 1.6137,
+      "step": 3660
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00018636882124247248,
+      "loss": 1.6402,
+      "step": 3670
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001862819026646694,
+      "loss": 1.6949,
+      "step": 3680
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001861850264262445,
+      "loss": 1.6283,
+      "step": 3690
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00018608783469816221,
+      "loss": 1.633,
+      "step": 3700
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00018599032783620342,
+      "loss": 1.6442,
+      "step": 3710
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00018589250619730253,
+      "loss": 1.629,
+      "step": 3720
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00018580419788394125,
+      "loss": 1.599,
+      "step": 3730
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00018570577915633075,
+      "loss": 1.6526,
+      "step": 3740
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00018560704669339962,
+      "loss": 1.6345,
+      "step": 3750
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00018550800085656875,
+      "loss": 1.6609,
+      "step": 3760
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00018540864200840615,
+      "loss": 1.6241,
+      "step": 3770
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001853089705126257,
+      "loss": 1.6081,
+      "step": 3780
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00018520898673408576,
+      "loss": 1.6591,
+      "step": 3790
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00018510869103878781,
+      "loss": 1.6196,
+      "step": 3800
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00018500808379387515,
+      "loss": 1.6015,
+      "step": 3810
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00018490716536763153,
+      "loss": 1.6196,
+      "step": 3820
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00018480593612947978,
+      "loss": 1.6504,
+      "step": 3830
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00018470439644998062,
+      "loss": 1.6474,
+      "step": 3840
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00018460254670083103,
+      "loss": 1.6038,
+      "step": 3850
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00018450038725486306,
+      "loss": 1.712,
+      "step": 3860
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00018439791848604253,
+      "loss": 1.6311,
+      "step": 3870
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00018429514076946746,
+      "loss": 1.626,
+      "step": 3880
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00018419205448136686,
+      "loss": 1.6255,
+      "step": 3890
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00018408865999909932,
+      "loss": 1.6269,
+      "step": 3900
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00018398495770115153,
+      "loss": 1.5649,
+      "step": 3910
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001838809479671371,
+      "loss": 1.6243,
+      "step": 3920
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001837766311777949,
+      "loss": 1.6366,
+      "step": 3930
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00018367200771498787,
+      "loss": 1.6387,
+      "step": 3940
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00018356707796170161,
+      "loss": 1.6256,
+      "step": 3950
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00018346184230204292,
+      "loss": 1.6158,
+      "step": 3960
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001833563011212383,
+      "loss": 1.6103,
+      "step": 3970
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00018325045480563273,
+      "loss": 1.6038,
+      "step": 3980
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00018314430374268817,
+      "loss": 1.5909,
+      "step": 3990
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001830378483209821,
+      "loss": 1.612,
+      "step": 4000
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.8420035243034363,
+      "eval_runtime": 62.0042,
+      "eval_samples_per_second": 8.451,
+      "eval_steps_per_second": 0.532,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001829310889302062,
+      "loss": 1.6545,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00018284546282243836,
+      "loss": 1.6391,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00018273815727291054,
+      "loss": 1.6237,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00018263054885136454,
+      "loss": 1.6281,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00018252263795171263,
+      "loss": 1.6102,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00018241442496897444,
+      "loss": 1.6246,
+      "step": 4060
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00018230591029927537,
+      "loss": 1.5991,
+      "step": 4070
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00018219709433984512,
+      "loss": 1.6252,
+      "step": 4080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00018208797748901637,
+      "loss": 1.6047,
+      "step": 4090
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001819785601462232,
+      "loss": 1.6173,
+      "step": 4100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00018186884271199967,
+      "loss": 1.5678,
+      "step": 4110
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001817588255879784,
+      "loss": 1.6143,
+      "step": 4120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.000181648509176889,
+      "loss": 1.6248,
+      "step": 4130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00018153789388255677,
+      "loss": 1.6552,
+      "step": 4140
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001814269801099009,
+      "loss": 1.626,
+      "step": 4150
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00018131576826493337,
+      "loss": 1.6096,
+      "step": 4160
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00018120425875475723,
+      "loss": 1.6182,
+      "step": 4170
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00018109245198756518,
+      "loss": 1.6014,
+      "step": 4180
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00018099157208059183,
+      "loss": 1.5923,
+      "step": 4190
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001808792016535363,
+      "loss": 1.5841,
+      "step": 4200
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00018076653515937166,
+      "loss": 1.5837,
+      "step": 4210
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00018065357301052593,
+      "loss": 1.6354,
+      "step": 4220
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00018054031562050928,
+      "loss": 1.6433,
+      "step": 4230
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001804267634039127,
+      "loss": 1.6109,
+      "step": 4240
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001803243146757791,
+      "loss": 1.6059,
+      "step": 4250
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00018021020343474294,
+      "loss": 1.6127,
+      "step": 4260
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001800957985755384,
+      "loss": 1.6006,
+      "step": 4270
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00017998110051695688,
+      "loss": 1.5927,
+      "step": 4280
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001798661096788631,
+      "loss": 1.6142,
+      "step": 4290
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00017975082648219356,
+      "loss": 1.6272,
+      "step": 4300
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001796352513489549,
+      "loss": 1.6459,
+      "step": 4310
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00017951938470222247,
+      "loss": 1.6373,
+      "step": 4320
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001794032269661387,
+      "loss": 1.6331,
+      "step": 4330
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00017928677856591163,
+      "loss": 1.6007,
+      "step": 4340
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001791700399278133,
+      "loss": 1.6359,
+      "step": 4350
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00017905301147917816,
+      "loss": 1.5939,
+      "step": 4360
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00017893569364840154,
+      "loss": 1.5889,
+      "step": 4370
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001788180868649382,
+      "loss": 1.6206,
+      "step": 4380
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00017870019155930047,
+      "loss": 1.5902,
+      "step": 4390
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00017858200816305697,
+      "loss": 1.6394,
+      "step": 4400
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00017846353710883087,
+      "loss": 1.6193,
+      "step": 4410
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00017835666757086383,
+      "loss": 1.6162,
+      "step": 4420
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00017823765116211767,
+      "loss": 1.6329,
+      "step": 4430
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001781183483559451,
+      "loss": 1.6248,
+      "step": 4440
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00017799875958906703,
+      "loss": 1.6109,
+      "step": 4450
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001778788852992512,
+      "loss": 1.5499,
+      "step": 4460
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001777587259253104,
+      "loss": 1.6107,
+      "step": 4470
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00017763828190710113,
+      "loss": 1.5865,
+      "step": 4480
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00017751755368552178,
+      "loss": 1.6013,
+      "step": 4490
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00017739654170251116,
+      "loss": 1.5829,
+      "step": 4500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00017727524640104674,
+      "loss": 1.6356,
+      "step": 4510
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00017715366822514318,
+      "loss": 1.6237,
+      "step": 4520
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00017703180761985063,
+      "loss": 1.5802,
+      "step": 4530
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00017690966503125307,
+      "loss": 1.5659,
+      "step": 4540
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001767872409064667,
+      "loss": 1.61,
+      "step": 4550
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00017666453569363836,
+      "loss": 1.6184,
+      "step": 4560
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00017654154984194382,
+      "loss": 1.5797,
+      "step": 4570
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00017641828380158612,
+      "loss": 1.6256,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00017629473802379403,
+      "loss": 1.5783,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00017617091296082032,
+      "loss": 1.5988,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001760468090659401,
+      "loss": 1.5904,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001759224267934491,
+      "loss": 1.611,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00017579776659866218,
+      "loss": 1.6066,
+      "step": 4630
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00017569783864540068,
+      "loss": 1.5947,
+      "step": 4640
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00017557267934112085,
+      "loss": 1.576,
+      "step": 4650
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00017544724339483368,
+      "loss": 1.6143,
+      "step": 4660
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00017532153126571107,
+      "loss": 1.5985,
+      "step": 4670
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00017519554341393593,
+      "loss": 1.5992,
+      "step": 4680
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00017506928030070054,
+      "loss": 1.5891,
+      "step": 4690
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00017494274238820468,
+      "loss": 1.5622,
+      "step": 4700
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00017482862369640954,
+      "loss": 1.5646,
+      "step": 4710
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00017470156494228082,
+      "loss": 1.6121,
+      "step": 4720
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001745742327349537,
+      "loss": 1.5766,
+      "step": 4730
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00017444662754054156,
+      "loss": 1.557,
+      "step": 4740
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00017431874982615708,
+      "loss": 1.5716,
+      "step": 4750
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00017419060005991054,
+      "loss": 1.5992,
+      "step": 4760
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001740621787109081,
+      "loss": 1.6036,
+      "step": 4770
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00017393348624925004,
+      "loss": 1.6121,
+      "step": 4780
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00017380452314602916,
+      "loss": 1.6076,
+      "step": 4790
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00017367528987332885,
+      "loss": 1.5798,
+      "step": 4800
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00017354578690422157,
+      "loss": 1.5597,
+      "step": 4810
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00017341601471276708,
+      "loss": 1.5834,
+      "step": 4820
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001732859737740105,
+      "loss": 1.6169,
+      "step": 4830
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00017315566456398086,
+      "loss": 1.5933,
+      "step": 4840
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00017303815729724509,
+      "loss": 1.6006,
+      "step": 4850
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00017292043346556449,
+      "loss": 1.6013,
+      "step": 4860
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001727893756367969,
+      "loss": 1.6042,
+      "step": 4870
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00017265805135460778,
+      "loss": 1.5738,
+      "step": 4880
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00017252646109972383,
+      "loss": 1.6376,
+      "step": 4890
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00017240780286177955,
+      "loss": 1.5904,
+      "step": 4900
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001722757085866635,
+      "loss": 1.5651,
+      "step": 4910
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00017214334973845988,
+      "loss": 1.5923,
+      "step": 4920
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001720107268016827,
+      "loss": 1.6032,
+      "step": 4930
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00017187784026181265,
+      "loss": 1.5859,
+      "step": 4940
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00017174469060529527,
+      "loss": 1.5376,
+      "step": 4950
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00017161127831953946,
+      "loss": 1.5445,
+      "step": 4960
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001714776038929153,
+      "loss": 1.5652,
+      "step": 4970
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00017134366781475262,
+      "loss": 1.5267,
+      "step": 4980
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00017120947057533897,
+      "loss": 1.5805,
+      "step": 4990
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00017108847017299018,
+      "loss": 1.5876,
+      "step": 5000
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.826651930809021,
+      "eval_runtime": 61.8972,
+      "eval_samples_per_second": 8.466,
+      "eval_steps_per_second": 0.533,
+      "step": 5000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00017095377808136445,
+      "loss": 1.6163,
+      "step": 5010
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00017083233311224484,
+      "loss": 1.629,
+      "step": 5020
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001706971479483343,
+      "loss": 1.5993,
+      "step": 5030
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00017056170398982906,
+      "loss": 1.6056,
+      "step": 5040
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00017042600173253645,
+      "loss": 1.5728,
+      "step": 5050
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00017029004167320926,
+      "loss": 1.6298,
+      "step": 5060
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00017015382430954413,
+      "loss": 1.5792,
+      "step": 5070
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00017001735014017955,
+      "loss": 1.5615,
+      "step": 5080
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.000169880619664694,
+      "loss": 1.6449,
+      "step": 5090
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00016974363338360425,
+      "loss": 1.5903,
+      "step": 5100
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001696063917983635,
+      "loss": 1.5682,
+      "step": 5110
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00016946889541135946,
+      "loss": 1.5754,
+      "step": 5120
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00016933114472591262,
+      "loss": 1.6168,
+      "step": 5130
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00016919314024627436,
+      "loss": 1.5977,
+      "step": 5140
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00016906871963807865,
+      "loss": 1.6037,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00016894409423469082,
+      "loss": 1.582,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00016880538182183466,
+      "loss": 1.5863,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00016866641753939926,
+      "loss": 1.5792,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00016852720189607857,
+      "loss": 1.5481,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00016838773540148655,
+      "loss": 1.594,
+      "step": 5200
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00016824801856615547,
+      "loss": 1.5484,
+      "step": 5210
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00016810805190153397,
+      "loss": 1.551,
+      "step": 5220
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001679678359199853,
+      "loss": 1.6115,
+      "step": 5230
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001678273711347852,
+      "loss": 1.5713,
+      "step": 5240
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00016770074052593968,
+      "loss": 1.532,
+      "step": 5250
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00016755980443113736,
+      "loss": 1.6103,
+      "step": 5260
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00016741862102632728,
+      "loss": 1.5881,
+      "step": 5270
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00016727719082832666,
+      "loss": 1.5909,
+      "step": 5280
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00016713551435485608,
+      "loss": 1.5632,
+      "step": 5290
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001669935921245377,
+      "loss": 1.5794,
+      "step": 5300
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00016685142465689326,
+      "loss": 1.5779,
+      "step": 5310
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00016670901247234224,
+      "loss": 1.615,
+      "step": 5320
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001665663560921999,
+      "loss": 1.6188,
+      "step": 5330
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00016642345603867545,
+      "loss": 1.5351,
+      "step": 5340
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00016628031283487006,
+      "loss": 1.6056,
+      "step": 5350
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00016613692700477494,
+      "loss": 1.6137,
+      "step": 5360
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001659932990732696,
+      "loss": 1.5894,
+      "step": 5370
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00016584942956611963,
+      "loss": 1.6144,
+      "step": 5380
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00016570531900997497,
+      "loss": 1.5606,
+      "step": 5390
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00016556096793236805,
+      "loss": 1.5671,
+      "step": 5400
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00016541637686171167,
+      "loss": 1.5839,
+      "step": 5410
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00016527154632729713,
+      "loss": 1.5991,
+      "step": 5420
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00016512647685929235,
+      "loss": 1.5912,
+      "step": 5430
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001649811689887399,
+      "loss": 1.578,
+      "step": 5440
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00016483562324755502,
+      "loss": 1.5861,
+      "step": 5450
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00016468984016852374,
+      "loss": 1.5587,
+      "step": 5460
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001645438202853008,
+      "loss": 1.5391,
+      "step": 5470
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00016439756413240793,
+      "loss": 1.5762,
+      "step": 5480
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00016425107224523168,
+      "loss": 1.6125,
+      "step": 5490
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001641190284371531,
+      "loss": 1.5776,
+      "step": 5500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00016397209013291726,
+      "loss": 1.5469,
+      "step": 5510
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00016382491765189186,
+      "loss": 1.5636,
+      "step": 5520
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00016367751153281774,
+      "loss": 1.5732,
+      "step": 5530
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00016352987231529103,
+      "loss": 1.547,
+      "step": 5540
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00016338200053976108,
+      "loss": 1.6132,
+      "step": 5550
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00016323389674752868,
+      "loss": 1.5797,
+      "step": 5560
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00016308556148074378,
+      "loss": 1.5448,
+      "step": 5570
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00016293699528240386,
+      "loss": 1.5573,
+      "step": 5580
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001627881986963515,
+      "loss": 1.5472,
+      "step": 5590
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00016263917226727286,
+      "loss": 1.5518,
+      "step": 5600
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001624899165406954,
+      "loss": 1.5813,
+      "step": 5610
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00016234043206298586,
+      "loss": 1.5408,
+      "step": 5620
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00016219071938134845,
+      "loss": 1.5346,
+      "step": 5630
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001620557833064539,
+      "loss": 1.5452,
+      "step": 5640
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00016192066323037722,
+      "loss": 1.5825,
+      "step": 5650
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00016177031449597098,
+      "loss": 1.5871,
+      "step": 5660
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001616197396446142,
+      "loss": 1.5518,
+      "step": 5670
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001614689392275025,
+      "loss": 1.5645,
+      "step": 5680
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00016131791379665717,
+      "loss": 1.549,
+      "step": 5690
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00016116666390492325,
+      "loss": 1.5868,
+      "step": 5700
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00016101519010596743,
+      "loss": 1.5491,
+      "step": 5710
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00016086349295427595,
+      "loss": 1.5768,
+      "step": 5720
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00016072677501010647,
+      "loss": 1.5492,
+      "step": 5730
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00016057465501875367,
+      "loss": 1.5695,
+      "step": 5740
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00016042231328729185,
+      "loss": 1.5693,
+      "step": 5750
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001602697503733844,
+      "loss": 1.5273,
+      "step": 5760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00016011696683550456,
+      "loss": 1.4587,
+      "step": 5770
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00015996396323293295,
+      "loss": 1.4827,
+      "step": 5780
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00015981074012575593,
+      "loss": 1.4734,
+      "step": 5790
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001596572980748634,
+      "loss": 1.4913,
+      "step": 5800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00015950363764194662,
+      "loss": 1.4804,
+      "step": 5810
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001593497593894963,
+      "loss": 1.4434,
+      "step": 5820
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00015919566388080048,
+      "loss": 1.4501,
+      "step": 5830
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00015904135167994264,
+      "loss": 1.4815,
+      "step": 5840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00015888682335179924,
+      "loss": 1.4588,
+      "step": 5850
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00015873207946203802,
+      "loss": 1.4268,
+      "step": 5860
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00015857712057711592,
+      "loss": 1.4502,
+      "step": 5870
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00015843747422863421,
+      "loss": 1.4486,
+      "step": 5880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00015828210841631188,
+      "loss": 1.4331,
+      "step": 5890
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0001581265292559965,
+      "loss": 1.4565,
+      "step": 5900
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00015797073731720253,
+      "loss": 1.453,
+      "step": 5910
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00015781473317022333,
+      "loss": 1.45,
+      "step": 5920
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00015765851738612895,
+      "loss": 1.4356,
+      "step": 5930
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00015750209053676432,
+      "loss": 1.4915,
+      "step": 5940
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00015734545319474693,
+      "loss": 1.4443,
+      "step": 5950
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00015718860593346473,
+      "loss": 1.4245,
+      "step": 5960
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001570315493270742,
+      "loss": 1.4927,
+      "step": 5970
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00015687428395049814,
+      "loss": 1.4698,
+      "step": 5980
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00015671681037942355,
+      "loss": 1.4568,
+      "step": 5990
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00015655912919029953,
+      "loss": 1.4498,
+      "step": 6000
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 0.8124380707740784,
+      "eval_runtime": 62.0134,
+      "eval_samples_per_second": 8.45,
+      "eval_steps_per_second": 0.532,
+      "step": 6000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00015640124096033526,
+      "loss": 1.4347,
+      "step": 6010
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001562431462674977,
+      "loss": 1.4754,
+      "step": 6020
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00015608484569050975,
+      "loss": 1.4716,
+      "step": 6030
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00015592633980884778,
+      "loss": 1.4523,
+      "step": 6040
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00015578350945939874,
+      "loss": 1.4578,
+      "step": 6050
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00015562461509800382,
+      "loss": 1.4311,
+      "step": 6060
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0001554814360610988,
+      "loss": 1.4419,
+      "step": 6070
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00015532215531972608,
+      "loss": 1.449,
+      "step": 6080
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001551786294874456,
+      "loss": 1.4604,
+      "step": 6090
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001550189644709561,
+      "loss": 1.4673,
+      "step": 6100
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00015485909805156665,
+      "loss": 1.4787,
+      "step": 6110
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001546990308144857,
+      "loss": 1.4571,
+      "step": 6120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001545387633456568,
+      "loss": 1.4773,
+      "step": 6130
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00015437829623175637,
+      "loss": 1.4816,
+      "step": 6140
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00015421763006019177,
+      "loss": 1.5024,
+      "step": 6150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00015405676541909897,
+      "loss": 1.4956,
+      "step": 6160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00015389570289734046,
+      "loss": 1.4515,
+      "step": 6170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00015373444308450313,
+      "loss": 1.4745,
+      "step": 6180
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00015357298657089606,
+      "loss": 1.4185,
+      "step": 6190
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00015341133394754838,
+      "loss": 1.4462,
+      "step": 6200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00015324948580620703,
+      "loss": 1.4664,
+      "step": 6210
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00015308744273933477,
+      "loss": 1.4747,
+      "step": 6220
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00015292520534010784,
+      "loss": 1.4091,
+      "step": 6230
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00015277902601747382,
+      "loss": 1.4634,
+      "step": 6240
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0001526164210235197,
+      "loss": 1.4462,
+      "step": 6250
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0001524536234214371,
+      "loss": 1.4771,
+      "step": 6260
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0001522906338071643,
+      "loss": 1.4639,
+      "step": 6270
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00015212745277734259,
+      "loss": 1.4302,
+      "step": 6280
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00015196408092931383,
+      "loss": 1.4412,
+      "step": 6290
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0001518005188611184,
+      "loss": 1.4518,
+      "step": 6300
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00015163676717149308,
+      "loss": 1.4316,
+      "step": 6310
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00015147282645986866,
+      "loss": 1.4317,
+      "step": 6320
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00015130869732636804,
+      "loss": 1.4911,
+      "step": 6330
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00015114438037180364,
+      "loss": 1.4359,
+      "step": 6340
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00015097987619767556,
+      "loss": 1.4517,
+      "step": 6350
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001508316628659255,
+      "loss": 1.4638,
+      "step": 6360
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.000150666804634212,
+      "loss": 1.4206,
+      "step": 6370
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0001505017609311527,
+      "loss": 1.4665,
+      "step": 6380
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00015033653236090806,
+      "loss": 1.4544,
+      "step": 6390
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00015018766908612838,
+      "loss": 1.4356,
+      "step": 6400
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00015002209093511546,
+      "loss": 1.4497,
+      "step": 6410
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00014985632967280134,
+      "loss": 1.4469,
+      "step": 6420
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014969038590597315,
+      "loss": 1.4334,
+      "step": 6430
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0001495242602420861,
+      "loss": 1.4714,
+      "step": 6440
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014935795328926125,
+      "loss": 1.4577,
+      "step": 6450
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014919146565628327,
+      "loss": 1.4247,
+      "step": 6460
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014902479795259822,
+      "loss": 1.4811,
+      "step": 6470
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014885795078831132,
+      "loss": 1.4309,
+      "step": 6480
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00014869092477418482,
+      "loss": 1.4853,
+      "step": 6490
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00014852372052163553,
+      "loss": 1.4507,
+      "step": 6500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00014835633864273287,
+      "loss": 1.4455,
+      "step": 6510
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001482055435875876,
+      "loss": 1.4421,
+      "step": 6520
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001480378259071914,
+      "loss": 1.4758,
+      "step": 6530
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00014786993237911187,
+      "loss": 1.4564,
+      "step": 6540
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00014771867836201847,
+      "loss": 1.4497,
+      "step": 6550
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00014755045241707308,
+      "loss": 1.4975,
+      "step": 6560
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00014738205240852806,
+      "loss": 1.4897,
+      "step": 6570
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00014721347895282978,
+      "loss": 1.4567,
+      "step": 6580
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0001470447326670598,
+      "loss": 1.4813,
+      "step": 6590
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014687581416893218,
+      "loss": 1.4861,
+      "step": 6600
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0001467067240767915,
+      "loss": 1.4759,
+      "step": 6610
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014653746300961038,
+      "loss": 1.4753,
+      "step": 6620
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014636803158698738,
+      "loss": 1.464,
+      "step": 6630
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014619843042914466,
+      "loss": 1.4849,
+      "step": 6640
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014602866015692563,
+      "loss": 1.4503,
+      "step": 6650
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00014587572283276284,
+      "loss": 1.4429,
+      "step": 6660
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00014572264948280539,
+      "loss": 1.481,
+      "step": 6670
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00014555240899848083,
+      "loss": 1.4358,
+      "step": 6680
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00014538200176461162,
+      "loss": 1.4236,
+      "step": 6690
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00014521142840499203,
+      "loss": 1.4508,
+      "step": 6700
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0001450406895440244,
+      "loss": 1.4858,
+      "step": 6710
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001448697858067168,
+      "loss": 1.4396,
+      "step": 6720
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00014469871781868098,
+      "loss": 1.4119,
+      "step": 6730
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00014452748620612992,
+      "loss": 1.4862,
+      "step": 6740
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00014435609159587555,
+      "loss": 1.4585,
+      "step": 6750
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001442016976021512,
+      "loss": 1.4725,
+      "step": 6760
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00014402999502526254,
+      "loss": 1.4497,
+      "step": 6770
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014385813127179106,
+      "loss": 1.4455,
+      "step": 6780
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014368610697086277,
+      "loss": 1.4252,
+      "step": 6790
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014351392275219134,
+      "loss": 1.4629,
+      "step": 6800
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014334157924607578,
+      "loss": 1.4628,
+      "step": 6810
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014316907708339822,
+      "loss": 1.4321,
+      "step": 6820
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014299641689562156,
+      "loss": 1.4558,
+      "step": 6830
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0001428235993147873,
+      "loss": 1.4762,
+      "step": 6840
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014265062497351285,
+      "loss": 1.4601,
+      "step": 6850
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014247749450498962,
+      "loss": 1.4782,
+      "step": 6860
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014230420854298054,
+      "loss": 1.4407,
+      "step": 6870
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014213076772181767,
+      "loss": 1.4164,
+      "step": 6880
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00014195717267640004,
+      "loss": 1.4223,
+      "step": 6890
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00014178342404219118,
+      "loss": 1.3949,
+      "step": 6900
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00014160952245521682,
+      "loss": 1.4718,
+      "step": 6910
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00014145288077845185,
+      "loss": 1.4722,
+      "step": 6920
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00014127869033547745,
+      "loss": 1.4446,
+      "step": 6930
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00014112178972372757,
+      "loss": 1.4475,
+      "step": 6940
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00014094731272664267,
+      "loss": 1.4202,
+      "step": 6950
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00014077268583746858,
+      "loss": 1.4854,
+      "step": 6960
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001405979096954461,
+      "loss": 1.4541,
+      "step": 6970
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00014042298494036228,
+      "loss": 1.4756,
+      "step": 6980
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00014024791221254815,
+      "loss": 1.435,
+      "step": 6990
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001400726921528765,
+      "loss": 1.3868,
+      "step": 7000
+    },
+    {
+      "epoch": 1.21,
+      "eval_loss": 0.8074263334274292,
+      "eval_runtime": 61.8502,
+      "eval_samples_per_second": 8.472,
+      "eval_steps_per_second": 0.534,
+      "step": 7000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001398973254027594,
+      "loss": 1.4526,
+      "step": 7010
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013972181260414585,
+      "loss": 1.4434,
+      "step": 7020
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001395461543995196,
+      "loss": 1.4733,
+      "step": 7030
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013937035143189657,
+      "loss": 1.4456,
+      "step": 7040
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013919440434482266,
+      "loss": 1.4451,
+      "step": 7050
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00013901831378237124,
+      "loss": 1.4572,
+      "step": 7060
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.000138842080389141,
+      "loss": 1.4455,
+      "step": 7070
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00013866570481025346,
+      "loss": 1.438,
+      "step": 7080
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00013848918769135055,
+      "loss": 1.4261,
+      "step": 7090
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00013831252967859238,
+      "loss": 1.436,
+      "step": 7100
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00013813573141865484,
+      "loss": 1.4295,
+      "step": 7110
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00013797649360826399,
+      "loss": 1.4416,
+      "step": 7120
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00013779943066211437,
+      "loss": 1.4494,
+      "step": 7130
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001376222293470401,
+      "loss": 1.4558,
+      "step": 7140
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00013744489031170578,
+      "loss": 1.4371,
+      "step": 7150
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001372851679656103,
+      "loss": 1.4876,
+      "step": 7160
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00013710756905065686,
+      "loss": 1.4465,
+      "step": 7170
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00013692983429941337,
+      "loss": 1.5011,
+      "step": 7180
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00013675196436249725,
+      "loss": 1.4399,
+      "step": 7190
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00013657395989102067,
+      "loss": 1.4586,
+      "step": 7200
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00013639582153658842,
+      "loss": 1.431,
+      "step": 7210
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00013621754995129522,
+      "loss": 1.4681,
+      "step": 7220
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0001360391457877237,
+      "loss": 1.47,
+      "step": 7230
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001358606096989416,
+      "loss": 1.4658,
+      "step": 7240
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001356819423384997,
+      "loss": 1.4789,
+      "step": 7250
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00013550314436042932,
+      "loss": 1.4218,
+      "step": 7260
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001353242164192399,
+      "loss": 1.4351,
+      "step": 7270
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00013514515916991657,
+      "loss": 1.4711,
+      "step": 7280
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00013496597326791786,
+      "loss": 1.4263,
+      "step": 7290
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00013478665936917332,
+      "loss": 1.4692,
+      "step": 7300
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00013460721813008086,
+      "loss": 1.457,
+      "step": 7310
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001344276502075047,
+      "loss": 1.4114,
+      "step": 7320
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00013424795625877276,
+      "loss": 1.395,
+      "step": 7330
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001340681369416742,
+      "loss": 1.4456,
+      "step": 7340
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00013388819291445723,
+      "loss": 1.4459,
+      "step": 7350
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001337081248358265,
+      "loss": 1.4642,
+      "step": 7360
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001335279333649408,
+      "loss": 1.4538,
+      "step": 7370
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00013334761916141064,
+      "loss": 1.4443,
+      "step": 7380
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00013316718288529567,
+      "loss": 1.4769,
+      "step": 7390
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00013300468641063172,
+      "loss": 1.4329,
+      "step": 7400
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013282402001666874,
+      "loss": 1.4563,
+      "step": 7410
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013264323346681258,
+      "loss": 1.487,
+      "step": 7420
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013246232742285206,
+      "loss": 1.4135,
+      "step": 7430
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013228130254701342,
+      "loss": 1.485,
+      "step": 7440
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0001321182791051834,
+      "loss": 1.4554,
+      "step": 7450
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013193703027476557,
+      "loss": 1.4543,
+      "step": 7460
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00013175566453537692,
+      "loss": 1.4419,
+      "step": 7470
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001315741825509265,
+      "loss": 1.4353,
+      "step": 7480
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00013139258498574873,
+      "loss": 1.4382,
+      "step": 7490
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00013121087250460132,
+      "loss": 1.4579,
+      "step": 7500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00013102904577266255,
+      "loss": 1.4475,
+      "step": 7510
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00013084710545552893,
+      "loss": 1.442,
+      "step": 7520
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00013066505221921273,
+      "loss": 1.4578,
+      "step": 7530
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00013048288673013966,
+      "loss": 1.4778,
+      "step": 7540
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00013030060965514632,
+      "loss": 1.4279,
+      "step": 7550
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00013011822166147767,
+      "loss": 1.4175,
+      "step": 7560
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00012993572341678483,
+      "loss": 1.4537,
+      "step": 7570
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00012975311558912248,
+      "loss": 1.473,
+      "step": 7580
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00012957039884694638,
+      "loss": 1.4041,
+      "step": 7590
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00012938757385911104,
+      "loss": 1.4453,
+      "step": 7600
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00012920464129486723,
+      "loss": 1.4795,
+      "step": 7610
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00012903991056267166,
+      "loss": 1.4592,
+      "step": 7620
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001288567754484459,
+      "loss": 1.4767,
+      "step": 7630
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00012867353470085696,
+      "loss": 1.4769,
+      "step": 7640
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00012849018899067748,
+      "loss": 1.4212,
+      "step": 7650
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00012830673898906435,
+      "loss": 1.4932,
+      "step": 7660
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00012812318536755622,
+      "loss": 1.4644,
+      "step": 7670
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00012795789906852118,
+      "loss": 1.4359,
+      "step": 7680
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001277741504206582,
+      "loss": 1.4672,
+      "step": 7690
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00012759030010249867,
+      "loss": 1.4161,
+      "step": 7700
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00012740634878704655,
+      "loss": 1.4479,
+      "step": 7710
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00012722229714767566,
+      "loss": 1.5016,
+      "step": 7720
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00012703814585812706,
+      "loss": 1.4459,
+      "step": 7730
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00012685389559250655,
+      "loss": 1.4491,
+      "step": 7740
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00012666954702528224,
+      "loss": 1.4229,
+      "step": 7750
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00012648510083128212,
+      "loss": 1.4286,
+      "step": 7760
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00012631901634382203,
+      "loss": 1.4043,
+      "step": 7770
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00012613438651937683,
+      "loss": 1.44,
+      "step": 7780
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00012594966102716905,
+      "loss": 1.4456,
+      "step": 7790
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00012576484054340636,
+      "loss": 1.4206,
+      "step": 7800
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00012557992574464428,
+      "loss": 1.4273,
+      "step": 7810
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00012539491730778355,
+      "loss": 1.4658,
+      "step": 7820
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001252098159100676,
+      "loss": 1.423,
+      "step": 7830
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00012502462222908025,
+      "loss": 1.4591,
+      "step": 7840
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001248393369427431,
+      "loss": 1.3987,
+      "step": 7850
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00012465396072931307,
+      "loss": 1.4278,
+      "step": 7860
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00012446849426737996,
+      "loss": 1.4273,
+      "step": 7870
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00012428293823586387,
+      "loss": 1.4464,
+      "step": 7880
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00012409729331401288,
+      "loss": 1.4407,
+      "step": 7890
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0001239115601814004,
+      "loss": 1.4192,
+      "step": 7900
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00012372573951792271,
+      "loss": 1.4327,
+      "step": 7910
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00012355842664409558,
+      "loss": 1.4286,
+      "step": 7920
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00012337244154623397,
+      "loss": 1.4381,
+      "step": 7930
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0001231863708910095,
+      "loss": 1.4084,
+      "step": 7940
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00012300021535955412,
+      "loss": 1.4431,
+      "step": 7950
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0001228139756333103,
+      "loss": 1.4226,
+      "step": 7960
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00012262765239402884,
+      "loss": 1.3949,
+      "step": 7970
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0001224412463237662,
+      "loss": 1.4388,
+      "step": 7980
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00012225475810488206,
+      "loss": 1.4102,
+      "step": 7990
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00012208684903502762,
+      "loss": 1.4059,
+      "step": 8000
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 0.8009569048881531,
+      "eval_runtime": 62.0166,
+      "eval_samples_per_second": 8.449,
+      "eval_steps_per_second": 0.532,
+      "step": 8000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00012190020661473858,
+      "loss": 1.4513,
+      "step": 8010
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00012171348402636268,
+      "loss": 1.4368,
+      "step": 8020
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00012152668195341832,
+      "loss": 1.4503,
+      "step": 8030
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00012133980107971474,
+      "loss": 1.4039,
+      "step": 8040
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00012115284208934969,
+      "loss": 1.4467,
+      "step": 8050
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00012096580566670692,
+      "loss": 1.4028,
+      "step": 8060
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00012077869249645357,
+      "loss": 1.4299,
+      "step": 8070
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00012059150326353772,
+      "loss": 1.4264,
+      "step": 8080
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00012040423865318591,
+      "loss": 1.453,
+      "step": 8090
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001202168993509006,
+      "loss": 1.4452,
+      "step": 8100
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00012002948604245768,
+      "loss": 1.4251,
+      "step": 8110
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00011984199941390392,
+      "loss": 1.4419,
+      "step": 8120
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00011965444015155452,
+      "loss": 1.4453,
+      "step": 8130
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00011946680894199054,
+      "loss": 1.4178,
+      "step": 8140
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00011927910647205644,
+      "loss": 1.4655,
+      "step": 8150
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00011909133342885747,
+      "loss": 1.4289,
+      "step": 8160
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00011890349049975729,
+      "loss": 1.4673,
+      "step": 8170
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00011871557837237537,
+      "loss": 1.4313,
+      "step": 8180
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00011852759773458446,
+      "loss": 1.4631,
+      "step": 8190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00011835835715290196,
+      "loss": 1.4695,
+      "step": 8200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00011817024824131962,
+      "loss": 1.4487,
+      "step": 8210
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00011798207281556853,
+      "loss": 1.4025,
+      "step": 8220
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00011779383156448527,
+      "loss": 1.4665,
+      "step": 8230
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00011760552517714743,
+      "loss": 1.4005,
+      "step": 8240
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00011741715434287097,
+      "loss": 1.4488,
+      "step": 8250
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00011722871975120782,
+      "loss": 1.4511,
+      "step": 8260
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00011705907467624817,
+      "loss": 1.4192,
+      "step": 8270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00011687052084609971,
+      "loss": 1.4309,
+      "step": 8280
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0001166819052595759,
+      "loss": 1.4495,
+      "step": 8290
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00011649322860712455,
+      "loss": 1.4375,
+      "step": 8300
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00011630449157941714,
+      "loss": 1.4502,
+      "step": 8310
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00011611569486734603,
+      "loss": 1.4179,
+      "step": 8320
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00011592683916202211,
+      "loss": 1.4581,
+      "step": 8330
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00011573792515477222,
+      "loss": 1.4211,
+      "step": 8340
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00011554895353713662,
+      "loss": 1.4118,
+      "step": 8350
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00011535992500086643,
+      "loss": 1.4308,
+      "step": 8360
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0001151708402379212,
+      "loss": 1.429,
+      "step": 8370
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00011498169994046621,
+      "loss": 1.4262,
+      "step": 8380
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00011479250480087011,
+      "loss": 1.4375,
+      "step": 8390
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00011462218285760746,
+      "loss": 1.4508,
+      "step": 8400
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00011443288542613578,
+      "loss": 1.4318,
+      "step": 8410
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00011424353516151814,
+      "loss": 1.4416,
+      "step": 8420
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00011405413275689179,
+      "loss": 1.4296,
+      "step": 8430
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0001138646789055848,
+      "loss": 1.4192,
+      "step": 8440
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00011367517430111365,
+      "loss": 1.4411,
+      "step": 8450
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001134856196371805,
+      "loss": 1.407,
+      "step": 8460
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00011329601560767078,
+      "loss": 1.447,
+      "step": 8470
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001131063629066507,
+      "loss": 1.4344,
+      "step": 8480
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00011291666222836454,
+      "loss": 1.4774,
+      "step": 8490
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00011274589117127904,
+      "loss": 1.4209,
+      "step": 8500
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00011255610124945745,
+      "loss": 1.3907,
+      "step": 8510
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00011236626536466241,
+      "loss": 1.4373,
+      "step": 8520
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00011217638421180883,
+      "loss": 1.4305,
+      "step": 8530
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00011198645848597729,
+      "loss": 1.4338,
+      "step": 8540
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00011179648888241155,
+      "loss": 1.4363,
+      "step": 8550
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00011160647609651597,
+      "loss": 1.4053,
+      "step": 8560
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00011141642082385304,
+      "loss": 1.4481,
+      "step": 8570
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00011122632376014078,
+      "loss": 1.3928,
+      "step": 8580
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00011103618560125007,
+      "loss": 1.3817,
+      "step": 8590
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00011084600704320238,
+      "loss": 1.4496,
+      "step": 8600
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00011065578878216696,
+      "loss": 1.4582,
+      "step": 8610
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00011046553151445844,
+      "loss": 1.451,
+      "step": 8620
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001102752359365342,
+      "loss": 1.469,
+      "step": 8630
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00011008490274499193,
+      "loss": 1.4299,
+      "step": 8640
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010989453263656697,
+      "loss": 1.4298,
+      "step": 8650
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010972316855101048,
+      "loss": 1.4192,
+      "step": 8660
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010953273022049615,
+      "loss": 1.4405,
+      "step": 8670
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00010934225699438665,
+      "loss": 1.4636,
+      "step": 8680
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00010917080183142705,
+      "loss": 1.4231,
+      "step": 8690
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00010898026422470837,
+      "loss": 1.4357,
+      "step": 8700
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00010878969374475633,
+      "loss": 1.4173,
+      "step": 8710
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00010859909108917496,
+      "loss": 1.4286,
+      "step": 8720
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00010840845695568593,
+      "loss": 1.4365,
+      "step": 8730
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00010821779204212623,
+      "loss": 1.4292,
+      "step": 8740
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00010804616787981517,
+      "loss": 1.4429,
+      "step": 8750
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00010785544640706349,
+      "loss": 1.43,
+      "step": 8760
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0001076646961785964,
+      "loss": 1.4428,
+      "step": 8770
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0001074739178926758,
+      "loss": 1.4511,
+      "step": 8780
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00010728311224766634,
+      "loss": 1.4283,
+      "step": 8790
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00010709227994203286,
+      "loss": 1.4041,
+      "step": 8800
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00010690142167433773,
+      "loss": 1.4574,
+      "step": 8810
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00010671053814323834,
+      "loss": 1.4421,
+      "step": 8820
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00010651963004748471,
+      "loss": 1.453,
+      "step": 8830
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00010632869808591662,
+      "loss": 1.4239,
+      "step": 8840
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00010613774295746124,
+      "loss": 1.4069,
+      "step": 8850
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001059467653611306,
+      "loss": 1.4335,
+      "step": 8860
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00010575576599601895,
+      "loss": 1.4341,
+      "step": 8870
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00010556474556130025,
+      "loss": 1.3864,
+      "step": 8880
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00010537370475622554,
+      "loss": 1.4389,
+      "step": 8890
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00010518264428012043,
+      "loss": 1.4895,
+      "step": 8900
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00010499156483238262,
+      "loss": 1.4252,
+      "step": 8910
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00010480046711247918,
+      "loss": 1.4495,
+      "step": 8920
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00010460935181994404,
+      "loss": 1.4292,
+      "step": 8930
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00010441821965437556,
+      "loss": 1.405,
+      "step": 8940
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00010422707131543377,
+      "loss": 1.4666,
+      "step": 8950
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00010405502456046876,
+      "loss": 1.4412,
+      "step": 8960
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0001038638474198912,
+      "loss": 1.3975,
+      "step": 8970
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00010367265613528012,
+      "loss": 1.4423,
+      "step": 8980
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00010348145140651204,
+      "loss": 1.4614,
+      "step": 8990
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00010329023393351272,
+      "loss": 1.4521,
+      "step": 9000
+    },
+    {
+      "epoch": 1.56,
+      "eval_loss": 0.7902594208717346,
+      "eval_runtime": 61.869,
+      "eval_samples_per_second": 8.47,
+      "eval_steps_per_second": 0.533,
+      "step": 9000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00010309900441625435,
+      "loss": 1.4428,
+      "step": 9010
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001029077635547535,
+      "loss": 1.4417,
+      "step": 9020
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00010271651204906811,
+      "loss": 1.4228,
+      "step": 9030
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001025252505992951,
+      "loss": 1.4302,
+      "step": 9040
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00010233397990556775,
+      "loss": 1.4544,
+      "step": 9050
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001021618289563197,
+      "loss": 1.4378,
+      "step": 9060
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00010198967159704729,
+      "loss": 1.4494,
+      "step": 9070
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00010179837874523537,
+      "loss": 1.4467,
+      "step": 9080
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00010160707931026259,
+      "loss": 1.4443,
+      "step": 9090
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00010143490476895921,
+      "loss": 1.4272,
+      "step": 9100
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.000101243594755249,
+      "loss": 1.4492,
+      "step": 9110
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00010105228018922502,
+      "loss": 1.4289,
+      "step": 9120
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00010086096177121504,
+      "loss": 1.3822,
+      "step": 9130
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00010066964020156091,
+      "loss": 1.413,
+      "step": 9140
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.000100478316180616,
+      "loss": 1.4419,
+      "step": 9150
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00010028699040874277,
+      "loss": 1.3911,
+      "step": 9160
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00010009566358630991,
+      "loss": 1.4321,
+      "step": 9170
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.990433641369012e-05,
+      "loss": 1.4358,
+      "step": 9180
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.971300959125727e-05,
+      "loss": 1.4307,
+      "step": 9190
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.952168381938401e-05,
+      "loss": 1.4235,
+      "step": 9200
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.933035979843912e-05,
+      "loss": 1.4123,
+      "step": 9210
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.913903822878499e-05,
+      "loss": 1.4267,
+      "step": 9220
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.8947719810775e-05,
+      "loss": 1.4151,
+      "step": 9230
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.875640524475103e-05,
+      "loss": 1.4215,
+      "step": 9240
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.856509523104083e-05,
+      "loss": 1.3905,
+      "step": 9250
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.83737904699555e-05,
+      "loss": 1.4147,
+      "step": 9260
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.820162125476466e-05,
+      "loss": 1.4225,
+      "step": 9270
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.802945737193441e-05,
+      "loss": 1.4273,
+      "step": 9280
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.783817104368033e-05,
+      "loss": 1.4644,
+      "step": 9290
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.764689262903611e-05,
+      "loss": 1.4026,
+      "step": 9300
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.74556228281972e-05,
+      "loss": 1.4154,
+      "step": 9310
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.726436234132755e-05,
+      "loss": 1.438,
+      "step": 9320
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.707311186855684e-05,
+      "loss": 1.4191,
+      "step": 9330
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.68818721099783e-05,
+      "loss": 1.4027,
+      "step": 9340
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.669064376564584e-05,
+      "loss": 1.4297,
+      "step": 9350
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.64994275355716e-05,
+      "loss": 1.4164,
+      "step": 9360
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.630822411972336e-05,
+      "loss": 1.3995,
+      "step": 9370
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.611703421802204e-05,
+      "loss": 1.4077,
+      "step": 9380
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.592585853033905e-05,
+      "loss": 1.4506,
+      "step": 9390
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.573469775649374e-05,
+      "loss": 1.4236,
+      "step": 9400
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.554355259625092e-05,
+      "loss": 1.3765,
+      "step": 9410
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.535242374931823e-05,
+      "loss": 1.4404,
+      "step": 9420
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.516131191534359e-05,
+      "loss": 1.4332,
+      "step": 9430
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.497021779391265e-05,
+      "loss": 1.4082,
+      "step": 9440
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.477914208454618e-05,
+      "loss": 1.4542,
+      "step": 9450
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.458808548669762e-05,
+      "loss": 1.4153,
+      "step": 9460
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.439704869975043e-05,
+      "loss": 1.4286,
+      "step": 9470
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.426333511085766e-05,
+      "loss": 1.3663,
+      "step": 9480
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.407233360732119e-05,
+      "loss": 1.389,
+      "step": 9490
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.388135380265187e-05,
+      "loss": 1.4096,
+      "step": 9500
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.369039639595209e-05,
+      "loss": 1.4111,
+      "step": 9510
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.349946208624212e-05,
+      "loss": 1.3776,
+      "step": 9520
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.330855157245775e-05,
+      "loss": 1.4375,
+      "step": 9530
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.31176655534477e-05,
+      "loss": 1.4131,
+      "step": 9540
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.292680472797101e-05,
+      "loss": 1.3917,
+      "step": 9550
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.273596979469446e-05,
+      "loss": 1.4346,
+      "step": 9560
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.254516145219005e-05,
+      "loss": 1.4056,
+      "step": 9570
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.235438039893248e-05,
+      "loss": 1.4045,
+      "step": 9580
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.216362733329655e-05,
+      "loss": 1.4248,
+      "step": 9590
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.197290295355454e-05,
+      "loss": 1.4291,
+      "step": 9600
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.17822079578738e-05,
+      "loss": 1.4162,
+      "step": 9610
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.159154304431409e-05,
+      "loss": 1.4189,
+      "step": 9620
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.140090891082506e-05,
+      "loss": 1.4231,
+      "step": 9630
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.121030625524365e-05,
+      "loss": 1.4008,
+      "step": 9640
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.103879135550087e-05,
+      "loss": 1.4321,
+      "step": 9650
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.084825043007008e-05,
+      "loss": 1.4719,
+      "step": 9660
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.065774300561337e-05,
+      "loss": 1.4252,
+      "step": 9670
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.04672697795039e-05,
+      "loss": 1.4116,
+      "step": 9680
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.029587369187029e-05,
+      "loss": 1.421,
+      "step": 9690
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.010546736343308e-05,
+      "loss": 1.4574,
+      "step": 9700
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.991509725500809e-05,
+      "loss": 1.3972,
+      "step": 9710
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.972476406346583e-05,
+      "loss": 1.3886,
+      "step": 9720
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.953446848554158e-05,
+      "loss": 1.4333,
+      "step": 9730
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.934421121783305e-05,
+      "loss": 1.3839,
+      "step": 9740
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.915399295679763e-05,
+      "loss": 1.4411,
+      "step": 9750
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.896381439874992e-05,
+      "loss": 1.4303,
+      "step": 9760
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.877367623985927e-05,
+      "loss": 1.453,
+      "step": 9770
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.858357917614699e-05,
+      "loss": 1.4066,
+      "step": 9780
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.839352390348404e-05,
+      "loss": 1.3924,
+      "step": 9790
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.820351111758849e-05,
+      "loss": 1.3878,
+      "step": 9800
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.801354151402274e-05,
+      "loss": 1.3849,
+      "step": 9810
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.782361578819118e-05,
+      "loss": 1.4135,
+      "step": 9820
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.763373463533758e-05,
+      "loss": 1.4187,
+      "step": 9830
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.748186227269857e-05,
+      "loss": 1.4153,
+      "step": 9840
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.729206310269713e-05,
+      "loss": 1.4083,
+      "step": 9850
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.710231045148006e-05,
+      "loss": 1.4055,
+      "step": 9860
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.691260501365754e-05,
+      "loss": 1.42,
+      "step": 9870
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.672294748366692e-05,
+      "loss": 1.438,
+      "step": 9880
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.653333855577024e-05,
+      "loss": 1.4077,
+      "step": 9890
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.634377892405157e-05,
+      "loss": 1.4448,
+      "step": 9900
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.615426928241457e-05,
+      "loss": 1.3921,
+      "step": 9910
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.596481032457986e-05,
+      "loss": 1.4282,
+      "step": 9920
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.577540274408256e-05,
+      "loss": 1.4084,
+      "step": 9930
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.558604723426972e-05,
+      "loss": 1.4007,
+      "step": 9940
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.539674448829775e-05,
+      "loss": 1.4371,
+      "step": 9950
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.520749519912991e-05,
+      "loss": 1.4304,
+      "step": 9960
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.501830005953381e-05,
+      "loss": 1.3866,
+      "step": 9970
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.482915976207883e-05,
+      "loss": 1.4156,
+      "step": 9980
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.464007499913359e-05,
+      "loss": 1.4735,
+      "step": 9990
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.445104646286339e-05,
+      "loss": 1.3907,
+      "step": 10000
+    },
+    {
+      "epoch": 1.74,
+      "eval_loss": 0.7795238494873047,
+      "eval_runtime": 61.8635,
+      "eval_samples_per_second": 8.47,
+      "eval_steps_per_second": 0.533,
+      "step": 10000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.428096942593624e-05,
+      "loss": 1.3785,
+      "step": 10010
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.409204962652496e-05,
+      "loss": 1.4134,
+      "step": 10020
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.392207157637791e-05,
+      "loss": 1.3956,
+      "step": 10030
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.373326301036039e-05,
+      "loss": 1.3801,
+      "step": 10040
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.354451399050185e-05,
+      "loss": 1.3976,
+      "step": 10050
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.335582520773848e-05,
+      "loss": 1.4065,
+      "step": 10060
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.316719735278616e-05,
+      "loss": 1.4058,
+      "step": 10070
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.29786311161376e-05,
+      "loss": 1.3986,
+      "step": 10080
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.279012718806004e-05,
+      "loss": 1.3801,
+      "step": 10090
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.260168625859259e-05,
+      "loss": 1.4051,
+      "step": 10100
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.241330901754376e-05,
+      "loss": 1.4178,
+      "step": 10110
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.222499615448894e-05,
+      "loss": 1.3994,
+      "step": 10120
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.203674835876778e-05,
+      "loss": 1.3948,
+      "step": 10130
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.184856631948184e-05,
+      "loss": 1.412,
+      "step": 10140
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.16604507254919e-05,
+      "loss": 1.4018,
+      "step": 10150
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.147240226541555e-05,
+      "loss": 1.395,
+      "step": 10160
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.128442162762465e-05,
+      "loss": 1.4177,
+      "step": 10170
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.109650950024272e-05,
+      "loss": 1.4684,
+      "step": 10180
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.090866657114254e-05,
+      "loss": 1.3745,
+      "step": 10190
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.072089352794361e-05,
+      "loss": 1.4243,
+      "step": 10200
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.055195810962145e-05,
+      "loss": 1.3984,
+      "step": 10210
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.036431974310813e-05,
+      "loss": 1.4003,
+      "step": 10220
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.017675325513676e-05,
+      "loss": 1.4197,
+      "step": 10230
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.000800543960246e-05,
+      "loss": 1.389,
+      "step": 10240
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.9820577412245e-05,
+      "loss": 1.3928,
+      "step": 10250
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.96332232538574e-05,
+      "loss": 1.4402,
+      "step": 10260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.94459436502699e-05,
+      "loss": 1.4059,
+      "step": 10270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.925873928703986e-05,
+      "loss": 1.4186,
+      "step": 10280
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.90716108494492e-05,
+      "loss": 1.4009,
+      "step": 10290
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.888455902250194e-05,
+      "loss": 1.4537,
+      "step": 10300
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.869758449092155e-05,
+      "loss": 1.4237,
+      "step": 10310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.851068793914867e-05,
+      "loss": 1.4158,
+      "step": 10320
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.832387005133845e-05,
+      "loss": 1.3977,
+      "step": 10330
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.813713151135805e-05,
+      "loss": 1.4253,
+      "step": 10340
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.795047300278422e-05,
+      "loss": 1.3966,
+      "step": 10350
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.776389520890071e-05,
+      "loss": 1.4211,
+      "step": 10360
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.757739881269582e-05,
+      "loss": 1.3777,
+      "step": 10370
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.739098449685987e-05,
+      "loss": 1.3922,
+      "step": 10380
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.720465294378272e-05,
+      "loss": 1.3884,
+      "step": 10390
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.70184048355513e-05,
+      "loss": 1.4122,
+      "step": 10400
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.683224085394702e-05,
+      "loss": 1.403,
+      "step": 10410
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.664616168044339e-05,
+      "loss": 1.433,
+      "step": 10420
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.646016799620345e-05,
+      "loss": 1.3929,
+      "step": 10430
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.631143505862324e-05,
+      "loss": 1.4136,
+      "step": 10440
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.614417685784577e-05,
+      "loss": 1.3835,
+      "step": 10450
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.595841739456996e-05,
+      "loss": 1.3911,
+      "step": 10460
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.577274593812058e-05,
+      "loss": 1.3981,
+      "step": 10470
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.558716316816814e-05,
+      "loss": 1.4022,
+      "step": 10480
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.54016697640586e-05,
+      "loss": 1.4368,
+      "step": 10490
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.521626640481061e-05,
+      "loss": 1.4221,
+      "step": 10500
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.503095376911342e-05,
+      "loss": 1.4189,
+      "step": 10510
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.484573253532406e-05,
+      "loss": 1.4389,
+      "step": 10520
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.466060338146498e-05,
+      "loss": 1.3616,
+      "step": 10530
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.447556698522156e-05,
+      "loss": 1.4295,
+      "step": 10540
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.429062402393965e-05,
+      "loss": 1.3832,
+      "step": 10550
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.410577517462307e-05,
+      "loss": 1.3762,
+      "step": 10560
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.392102111393116e-05,
+      "loss": 1.4565,
+      "step": 10570
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.373636251817615e-05,
+      "loss": 1.4553,
+      "step": 10580
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.355180006332097e-05,
+      "loss": 1.4171,
+      "step": 10590
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.336733442497654e-05,
+      "loss": 1.3761,
+      "step": 10600
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.318296627839935e-05,
+      "loss": 1.3928,
+      "step": 10610
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.299869629848908e-05,
+      "loss": 1.3833,
+      "step": 10620
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.281452515978599e-05,
+      "loss": 1.3991,
+      "step": 10630
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.263045353646861e-05,
+      "loss": 1.4247,
+      "step": 10640
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.248326834170777e-05,
+      "loss": 1.473,
+      "step": 10650
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.229937754384992e-05,
+      "loss": 1.3753,
+      "step": 10660
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.211558814713165e-05,
+      "loss": 1.41,
+      "step": 10670
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.195026494412065e-05,
+      "loss": 1.4259,
+      "step": 10680
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.176667006277049e-05,
+      "loss": 1.4327,
+      "step": 10690
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.158317853259342e-05,
+      "loss": 1.397,
+      "step": 10700
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.13997910252802e-05,
+      "loss": 1.3827,
+      "step": 10710
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.121650821214074e-05,
+      "loss": 1.417,
+      "step": 10720
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.103333076410166e-05,
+      "loss": 1.4142,
+      "step": 10730
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.085025935170397e-05,
+      "loss": 1.4047,
+      "step": 10740
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.066729464510045e-05,
+      "loss": 1.4153,
+      "step": 10750
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.04844373140533e-05,
+      "loss": 1.3896,
+      "step": 10760
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.030168802793164e-05,
+      "loss": 1.4255,
+      "step": 10770
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.011904745570912e-05,
+      "loss": 1.4286,
+      "step": 10780
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.993651626596138e-05,
+      "loss": 1.4337,
+      "step": 10790
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.97540951268637e-05,
+      "loss": 1.3943,
+      "step": 10800
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.95717847061885e-05,
+      "loss": 1.4138,
+      "step": 10810
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.938958567130285e-05,
+      "loss": 1.3604,
+      "step": 10820
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.920749868916618e-05,
+      "loss": 1.4096,
+      "step": 10830
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.902552442632765e-05,
+      "loss": 1.3915,
+      "step": 10840
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.88436635489238e-05,
+      "loss": 1.3946,
+      "step": 10850
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.868008625403449e-05,
+      "loss": 1.3984,
+      "step": 10860
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.851660182560898e-05,
+      "loss": 1.4222,
+      "step": 10870
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.833506196772657e-05,
+      "loss": 1.43,
+      "step": 10880
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.815363802279173e-05,
+      "loss": 1.3911,
+      "step": 10890
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.797233065492654e-05,
+      "loss": 1.4225,
+      "step": 10900
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.779114052782636e-05,
+      "loss": 1.4097,
+      "step": 10910
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.761006830475733e-05,
+      "loss": 1.3827,
+      "step": 10920
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.742911464855399e-05,
+      "loss": 1.4585,
+      "step": 10930
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.724828022161692e-05,
+      "loss": 1.4062,
+      "step": 10940
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.706756568591013e-05,
+      "loss": 1.4324,
+      "step": 10950
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.68869717029588e-05,
+      "loss": 1.3791,
+      "step": 10960
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.670649893384692e-05,
+      "loss": 1.4147,
+      "step": 10970
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.652614803921461e-05,
+      "loss": 1.4273,
+      "step": 10980
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.634591967925598e-05,
+      "loss": 1.3976,
+      "step": 10990
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.616581451371651e-05,
+      "loss": 1.3826,
+      "step": 11000
+    },
+    {
+      "epoch": 1.91,
+      "eval_loss": 0.7706289887428284,
+      "eval_runtime": 63.1592,
+      "eval_samples_per_second": 8.296,
+      "eval_steps_per_second": 0.522,
+      "step": 11000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.598583320189075e-05,
+      "loss": 1.4273,
+      "step": 11010
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.580597640261978e-05,
+      "loss": 1.3904,
+      "step": 11020
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.562624477428905e-05,
+      "loss": 1.3824,
+      "step": 11030
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.544663897482568e-05,
+      "loss": 1.3709,
+      "step": 11040
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.528510188239592e-05,
+      "loss": 1.4232,
+      "step": 11050
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.510573696871829e-05,
+      "loss": 1.3842,
+      "step": 11060
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.492649978928341e-05,
+      "loss": 1.3655,
+      "step": 11070
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.47473910002085e-05,
+      "loss": 1.3651,
+      "step": 11080
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.456841125714071e-05,
+      "loss": 1.3925,
+      "step": 11090
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.440744036422758e-05,
+      "loss": 1.3946,
+      "step": 11100
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.422870761318759e-05,
+      "loss": 1.4216,
+      "step": 11110
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.405010580685171e-05,
+      "loss": 1.3654,
+      "step": 11120
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.387163559901117e-05,
+      "loss": 1.3438,
+      "step": 11130
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.36932976429756e-05,
+      "loss": 1.4255,
+      "step": 11140
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.35150925915705e-05,
+      "loss": 1.4268,
+      "step": 11150
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.333702109713477e-05,
+      "loss": 1.3947,
+      "step": 11160
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.315908381151857e-05,
+      "loss": 1.4452,
+      "step": 11170
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.298128138608059e-05,
+      "loss": 1.4187,
+      "step": 11180
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.280361447168603e-05,
+      "loss": 1.3878,
+      "step": 11190
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.264383064821323e-05,
+      "loss": 1.3753,
+      "step": 11200
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.246642299615586e-05,
+      "loss": 1.372,
+      "step": 11210
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.230687356416249e-05,
+      "loss": 1.361,
+      "step": 11220
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.212972751884663e-05,
+      "loss": 1.4421,
+      "step": 11230
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.195272010177959e-05,
+      "loss": 1.4402,
+      "step": 11240
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.177585196091631e-05,
+      "loss": 1.3958,
+      "step": 11250
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.159912374370183e-05,
+      "loss": 1.4365,
+      "step": 11260
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.142253609706898e-05,
+      "loss": 1.4041,
+      "step": 11270
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.124608966743606e-05,
+      "loss": 1.4321,
+      "step": 11280
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.106978510070443e-05,
+      "loss": 1.4154,
+      "step": 11290
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.089362304225603e-05,
+      "loss": 1.4208,
+      "step": 11300
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.071760413695131e-05,
+      "loss": 1.3757,
+      "step": 11310
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.054172902912656e-05,
+      "loss": 1.4266,
+      "step": 11320
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.0365998362591744e-05,
+      "loss": 1.4081,
+      "step": 11330
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.019041278062807e-05,
+      "loss": 1.4108,
+      "step": 11340
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.0032510335413086e-05,
+      "loss": 1.3667,
+      "step": 11350
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.985720218447026e-05,
+      "loss": 1.3987,
+      "step": 11360
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.9682040980602316e-05,
+      "loss": 1.3553,
+      "step": 11370
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.9524522066830346e-05,
+      "loss": 1.3722,
+      "step": 11380
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.934964182845485e-05,
+      "loss": 1.4445,
+      "step": 11390
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.917491039513411e-05,
+      "loss": 1.3968,
+      "step": 11400
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.9000328406491425e-05,
+      "loss": 1.3855,
+      "step": 11410
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.882589650160322e-05,
+      "loss": 1.3988,
+      "step": 11420
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.865161531899642e-05,
+      "loss": 1.3642,
+      "step": 11430
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.8477485496646245e-05,
+      "loss": 1.4189,
+      "step": 11440
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.8303507671973864e-05,
+      "loss": 1.4004,
+      "step": 11450
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.812968248184392e-05,
+      "loss": 1.3928,
+      "step": 11460
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.795601056256257e-05,
+      "loss": 1.4273,
+      "step": 11470
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.778249254987461e-05,
+      "loss": 1.3824,
+      "step": 11480
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.7609129078961655e-05,
+      "loss": 1.4004,
+      "step": 11490
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.7435920784439514e-05,
+      "loss": 1.3801,
+      "step": 11500
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.7262868300355975e-05,
+      "loss": 1.373,
+      "step": 11510
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.7089972260188485e-05,
+      "loss": 1.414,
+      "step": 11520
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.6917233296841776e-05,
+      "loss": 1.3001,
+      "step": 11530
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.6744652042645616e-05,
+      "loss": 1.3162,
+      "step": 11540
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.6572229129352474e-05,
+      "loss": 1.2765,
+      "step": 11550
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.6399965188135084e-05,
+      "loss": 1.2453,
+      "step": 11560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.622786084958437e-05,
+      "loss": 1.2787,
+      "step": 11570
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.6090292716047934e-05,
+      "loss": 1.2726,
+      "step": 11580
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.591847724951989e-05,
+      "loss": 1.2786,
+      "step": 11590
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.574682314819745e-05,
+      "loss": 1.2788,
+      "step": 11600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.557533104043913e-05,
+      "loss": 1.2654,
+      "step": 11610
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.54040015540104e-05,
+      "loss": 1.2507,
+      "step": 11620
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5249944575829906e-05,
+      "loss": 1.3315,
+      "step": 11630
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.507892579728751e-05,
+      "loss": 1.2968,
+      "step": 11640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.490807145722008e-05,
+      "loss": 1.3051,
+      "step": 11650
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.47373821810585e-05,
+      "loss": 1.306,
+      "step": 11660
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.4566858593629454e-05,
+      "loss": 1.3139,
+      "step": 11670
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.439650131915299e-05,
+      "loss": 1.277,
+      "step": 11680
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.4226310981240466e-05,
+      "loss": 1.2737,
+      "step": 11690
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.4056288202892126e-05,
+      "loss": 1.311,
+      "step": 11700
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.3886433606494804e-05,
+      "loss": 1.2775,
+      "step": 11710
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.37167478138197e-05,
+      "loss": 1.2633,
+      "step": 11720
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.354723144602016e-05,
+      "loss": 1.3022,
+      "step": 11730
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.337788512362931e-05,
+      "loss": 1.2979,
+      "step": 11740
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.320870946655765e-05,
+      "loss": 1.2726,
+      "step": 11750
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.303970509409113e-05,
+      "loss": 1.2303,
+      "step": 11760
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.2870872624888615e-05,
+      "loss": 1.2648,
+      "step": 11770
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.2702212676979704e-05,
+      "loss": 1.2865,
+      "step": 11780
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.253372586776248e-05,
+      "loss": 1.2944,
+      "step": 11790
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.236541281400122e-05,
+      "loss": 1.3188,
+      "step": 11800
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.219727413182419e-05,
+      "loss": 1.3009,
+      "step": 11810
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.202931043672124e-05,
+      "loss": 1.2727,
+      "step": 11820
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.186152234354172e-05,
+      "loss": 1.2515,
+      "step": 11830
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.172741871515152e-05,
+      "loss": 1.3243,
+      "step": 11840
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.1559948252801414e-05,
+      "loss": 1.3009,
+      "step": 11850
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.139265511052607e-05,
+      "loss": 1.3033,
+      "step": 11860
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.122553990072023e-05,
+      "loss": 1.2961,
+      "step": 11870
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.10586032351273e-05,
+      "loss": 1.2361,
+      "step": 11880
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.090851339647496e-05,
+      "loss": 1.2856,
+      "step": 11890
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.074191764789694e-05,
+      "loss": 1.2688,
+      "step": 11900
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.0575502213883655e-05,
+      "loss": 1.2338,
+      "step": 11910
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.040926770361687e-05,
+      "loss": 1.3065,
+      "step": 11920
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.0243214725616126e-05,
+      "loss": 1.2683,
+      "step": 11930
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.00773438877363e-05,
+      "loss": 1.3036,
+      "step": 11940
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.99116557971657e-05,
+      "loss": 1.25,
+      "step": 11950
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.9746151060423564e-05,
+      "loss": 1.2719,
+      "step": 11960
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.958083028335794e-05,
+      "loss": 1.2411,
+      "step": 11970
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.9415694071143584e-05,
+      "loss": 1.286,
+      "step": 11980
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.9250743028279486e-05,
+      "loss": 1.2786,
+      "step": 11990
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.9085977758586906e-05,
+      "loss": 1.2634,
+      "step": 12000
+    },
+    {
+      "epoch": 2.08,
+      "eval_loss": 0.7820777893066406,
+      "eval_runtime": 62.0421,
+      "eval_samples_per_second": 8.446,
+      "eval_steps_per_second": 0.532,
+      "step": 12000
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.8921398865207045e-05,
+      "loss": 1.3052,
+      "step": 12010
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.875700695059875e-05,
+      "loss": 1.3028,
+      "step": 12020
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.859280261653654e-05,
+      "loss": 1.3132,
+      "step": 12030
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.8428786464108225e-05,
+      "loss": 1.3153,
+      "step": 12040
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.826495909371276e-05,
+      "loss": 1.3391,
+      "step": 12050
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.810132110505804e-05,
+      "loss": 1.2821,
+      "step": 12060
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.793787309715871e-05,
+      "loss": 1.2542,
+      "step": 12070
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.779093281812042e-05,
+      "loss": 1.2344,
+      "step": 12080
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.7627847421449165e-05,
+      "loss": 1.2916,
+      "step": 12090
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.746495373873521e-05,
+      "loss": 1.2703,
+      "step": 12100
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.730225236626855e-05,
+      "loss": 1.3033,
+      "step": 12110
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.713974389963527e-05,
+      "loss": 1.2804,
+      "step": 12120
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.697742893371525e-05,
+      "loss": 1.317,
+      "step": 12130
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.6815308062680086e-05,
+      "loss": 1.3142,
+      "step": 12140
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.665338187999084e-05,
+      "loss": 1.2592,
+      "step": 12150
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.649165097839591e-05,
+      "loss": 1.2689,
+      "step": 12160
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.6330115949928876e-05,
+      "loss": 1.2734,
+      "step": 12170
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.618490238457079e-05,
+      "loss": 1.2713,
+      "step": 12180
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.602374114352934e-05,
+      "loss": 1.3216,
+      "step": 12190
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.586277748845055e-05,
+      "loss": 1.2775,
+      "step": 12200
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.570201200855939e-05,
+      "loss": 1.2749,
+      "step": 12210
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.554144529235537e-05,
+      "loss": 1.2809,
+      "step": 12220
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.538107792761041e-05,
+      "loss": 1.2817,
+      "step": 12230
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.522091050136663e-05,
+      "loss": 1.2324,
+      "step": 12240
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.50609435999344e-05,
+      "loss": 1.2649,
+      "step": 12250
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.4901177808889936e-05,
+      "loss": 1.2493,
+      "step": 12260
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.474161371307322e-05,
+      "loss": 1.2946,
+      "step": 12270
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.458225189658598e-05,
+      "loss": 1.2523,
+      "step": 12280
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.44230929427895e-05,
+      "loss": 1.2769,
+      "step": 12290
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.426413743430241e-05,
+      "loss": 1.2823,
+      "step": 12300
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.410538595299864e-05,
+      "loss": 1.2536,
+      "step": 12310
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.3962684543383956e-05,
+      "loss": 1.2686,
+      "step": 12320
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.380432231411452e-05,
+      "loss": 1.2826,
+      "step": 12330
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.364616579523162e-05,
+      "loss": 1.2906,
+      "step": 12340
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.348821556568439e-05,
+      "loss": 1.2777,
+      "step": 12350
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.3346237214366844e-05,
+      "loss": 1.2865,
+      "step": 12360
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.3188680526855985e-05,
+      "loss": 1.2784,
+      "step": 12370
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.303133180335535e-05,
+      "loss": 1.2866,
+      "step": 12380
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.287419161985704e-05,
+      "loss": 1.2954,
+      "step": 12390
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.2717260551589775e-05,
+      "loss": 1.2677,
+      "step": 12400
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.2560539173016813e-05,
+      "loss": 1.2825,
+      "step": 12410
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.240402805783377e-05,
+      "loss": 1.2749,
+      "step": 12420
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.224772777896659e-05,
+      "loss": 1.2797,
+      "step": 12430
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.209163890856951e-05,
+      "loss": 1.2819,
+      "step": 12440
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.193576201802268e-05,
+      "loss": 1.2607,
+      "step": 12450
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.1780097677930485e-05,
+      "loss": 1.2767,
+      "step": 12460
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.162464645811913e-05,
+      "loss": 1.2333,
+      "step": 12470
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.146940892763472e-05,
+      "loss": 1.2667,
+      "step": 12480
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.131438565474112e-05,
+      "loss": 1.3182,
+      "step": 12490
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.11595772069178e-05,
+      "loss": 1.2889,
+      "step": 12500
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.100498415085804e-05,
+      "loss": 1.3046,
+      "step": 12510
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.085060705246642e-05,
+      "loss": 1.2576,
+      "step": 12520
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.069644647685712e-05,
+      "loss": 1.2588,
+      "step": 12530
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.0542502988351686e-05,
+      "loss": 1.2901,
+      "step": 12540
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.038877715047699e-05,
+      "loss": 1.3228,
+      "step": 12550
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.0250610452792004e-05,
+      "loss": 1.2813,
+      "step": 12560
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.011262091761672e-05,
+      "loss": 1.3074,
+      "step": 12570
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.997480895410295e-05,
+      "loss": 1.2753,
+      "step": 12580
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.9821893310242744e-05,
+      "loss": 1.2519,
+      "step": 12590
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.966919795488333e-05,
+      "loss": 1.331,
+      "step": 12600
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.9516723446982664e-05,
+      "loss": 1.3126,
+      "step": 12610
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.936447034469024e-05,
+      "loss": 1.2616,
+      "step": 12620
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.92124392053451e-05,
+      "loss": 1.2957,
+      "step": 12630
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.9060630585473746e-05,
+      "loss": 1.309,
+      "step": 12640
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.890904504078814e-05,
+      "loss": 1.2873,
+      "step": 12650
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.8757683126183654e-05,
+      "loss": 1.283,
+      "step": 12660
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.8606545395737005e-05,
+      "loss": 1.3069,
+      "step": 12670
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.84556324027043e-05,
+      "loss": 1.2527,
+      "step": 12680
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.8304944699518954e-05,
+      "loss": 1.2922,
+      "step": 12690
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.816951884539331e-05,
+      "loss": 1.2795,
+      "step": 12700
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.801926071191671e-05,
+      "loss": 1.285,
+      "step": 12710
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.786922946567352e-05,
+      "loss": 1.2804,
+      "step": 12720
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.771942565586933e-05,
+      "loss": 1.318,
+      "step": 12730
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.7569849830877333e-05,
+      "loss": 1.3102,
+      "step": 12740
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.742050253823604e-05,
+      "loss": 1.3083,
+      "step": 12750
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.7286285821885306e-05,
+      "loss": 1.2854,
+      "step": 12760
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.713737424618142e-05,
+      "loss": 1.305,
+      "step": 12770
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.6988692785952173e-05,
+      "loss": 1.2948,
+      "step": 12780
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.68402419854622e-05,
+      "loss": 1.2779,
+      "step": 12790
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.6692022388131795e-05,
+      "loss": 1.3145,
+      "step": 12800
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.654403453653494e-05,
+      "loss": 1.2673,
+      "step": 12810
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.639627897239718e-05,
+      "loss": 1.2883,
+      "step": 12820
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.6248756236593863e-05,
+      "loss": 1.2678,
+      "step": 12830
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.6101466869147995e-05,
+      "loss": 1.2771,
+      "step": 12840
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.5954411409228294e-05,
+      "loss": 1.3139,
+      "step": 12850
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.580759039514729e-05,
+      "loss": 1.2914,
+      "step": 12860
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.566100436435924e-05,
+      "loss": 1.2685,
+      "step": 12870
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.551465385345826e-05,
+      "loss": 1.2932,
+      "step": 12880
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.5383140205951094e-05,
+      "loss": 1.2751,
+      "step": 12890
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.5237238658062945e-05,
+      "loss": 1.2775,
+      "step": 12900
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.5091574181302256e-05,
+      "loss": 1.2826,
+      "step": 12910
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.494614730888971e-05,
+      "loss": 1.2661,
+      "step": 12920
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.480095857317618e-05,
+      "loss": 1.27,
+      "step": 12930
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.4656008505640814e-05,
+      "loss": 1.3088,
+      "step": 12940
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.4511297636889095e-05,
+      "loss": 1.263,
+      "step": 12950
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.4366826496650886e-05,
+      "loss": 1.2896,
+      "step": 12960
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.422259561377853e-05,
+      "loss": 1.2919,
+      "step": 12970
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.4078605516244785e-05,
+      "loss": 1.2451,
+      "step": 12980
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.396358715789669e-05,
+      "loss": 1.3136,
+      "step": 12990
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.383437644428432e-05,
+      "loss": 1.2834,
+      "step": 13000
+    },
+    {
+      "epoch": 2.26,
+      "eval_loss": 0.78007972240448,
+      "eval_runtime": 61.984,
+      "eval_samples_per_second": 8.454,
+      "eval_steps_per_second": 0.532,
+      "step": 13000
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.369103909748521e-05,
+      "loss": 1.2585,
+      "step": 13010
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.354794448184514e-05,
+      "loss": 1.27,
+      "step": 13020
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.340509312117752e-05,
+      "loss": 1.2923,
+      "step": 13030
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.32624855384053e-05,
+      "loss": 1.2711,
+      "step": 13040
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.3120122255559e-05,
+      "loss": 1.2891,
+      "step": 13050
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.2978003793774914e-05,
+      "loss": 1.2635,
+      "step": 13060
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.283613067329311e-05,
+      "loss": 1.2926,
+      "step": 13070
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.269450341345558e-05,
+      "loss": 1.2621,
+      "step": 13080
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.2553122532704325e-05,
+      "loss": 1.291,
+      "step": 13090
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.241198854857938e-05,
+      "loss": 1.2658,
+      "step": 13100
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.227110197771703e-05,
+      "loss": 1.2671,
+      "step": 13110
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.213046333584792e-05,
+      "loss": 1.2686,
+      "step": 13120
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.1990073137795066e-05,
+      "loss": 1.2723,
+      "step": 13130
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.186393480377876e-05,
+      "loss": 1.3037,
+      "step": 13140
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.172401806405554e-05,
+      "loss": 1.284,
+      "step": 13150
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.1584351255985664e-05,
+      "loss": 1.2582,
+      "step": 13160
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.144493489083469e-05,
+      "loss": 1.3076,
+      "step": 13170
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.130576947895139e-05,
+      "loss": 1.3193,
+      "step": 13180
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.1166855529765825e-05,
+      "loss": 1.2829,
+      "step": 13190
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.102819355178763e-05,
+      "loss": 1.3018,
+      "step": 13200
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.0889784052604066e-05,
+      "loss": 1.2842,
+      "step": 13210
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.075162753887814e-05,
+      "loss": 1.2816,
+      "step": 13220
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.061372451634678e-05,
+      "loss": 1.2915,
+      "step": 13230
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.0503584951244668e-05,
+      "loss": 1.2235,
+      "step": 13240
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.0366139484357482e-05,
+      "loss": 1.2766,
+      "step": 13250
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.0228948919785782e-05,
+      "loss": 1.2729,
+      "step": 13260
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.0092013759730564e-05,
+      "loss": 1.2784,
+      "step": 13270
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.9955334505457845e-05,
+      "loss": 1.2827,
+      "step": 13280
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.981891165729691e-05,
+      "loss": 1.279,
+      "step": 13290
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.9682745714638417e-05,
+      "loss": 1.2917,
+      "step": 13300
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.9546837175932596e-05,
+      "loss": 1.2764,
+      "step": 13310
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.941118653868744e-05,
+      "loss": 1.3066,
+      "step": 13320
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 1.2945,
+      "step": 13330
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.9154162624127146e-05,
+      "loss": 1.2654,
+      "step": 13340
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.9019262705797567e-05,
+      "loss": 1.2715,
+      "step": 13350
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.888462262017233e-05,
+      "loss": 1.3311,
+      "step": 13360
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.875024286011615e-05,
+      "loss": 1.306,
+      "step": 13370
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.8616123917540673e-05,
+      "loss": 1.2865,
+      "step": 13380
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.848226628340287e-05,
+      "loss": 1.2676,
+      "step": 13390
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.8348670447703218e-05,
+      "loss": 1.2997,
+      "step": 13400
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.8215336899483768e-05,
+      "loss": 1.2863,
+      "step": 13410
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.808226612682646e-05,
+      "loss": 1.2532,
+      "step": 13420
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.7949458616851343e-05,
+      "loss": 1.2524,
+      "step": 13430
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.781691485571475e-05,
+      "loss": 1.263,
+      "step": 13440
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.7684635328607477e-05,
+      "loss": 1.2607,
+      "step": 13450
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.7552620519753137e-05,
+      "loss": 1.2505,
+      "step": 13460
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.742087091240628e-05,
+      "loss": 1.2687,
+      "step": 13470
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.7302523411710645e-05,
+      "loss": 1.2855,
+      "step": 13480
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.7171279015116002e-05,
+      "loss": 1.2412,
+      "step": 13490
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.7040301215970876e-05,
+      "loss": 1.2575,
+      "step": 13500
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.6922649533852228e-05,
+      "loss": 1.3001,
+      "step": 13510
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.6792179589961273e-05,
+      "loss": 1.2726,
+      "step": 13520
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.66619776312545e-05,
+      "loss": 1.2727,
+      "step": 13530
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.6532044134350288e-05,
+      "loss": 1.2466,
+      "step": 13540
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.6402379574884418e-05,
+      "loss": 1.2975,
+      "step": 13550
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.627298442750803e-05,
+      "loss": 1.2745,
+      "step": 13560
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.614385916588613e-05,
+      "loss": 1.2988,
+      "step": 13570
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.6015004262695798e-05,
+      "loss": 1.2541,
+      "step": 13580
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.5886420189624407e-05,
+      "loss": 1.2596,
+      "step": 13590
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.5758107417367915e-05,
+      "loss": 1.267,
+      "step": 13600
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.5630066415629195e-05,
+      "loss": 1.2716,
+      "step": 13610
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.550229765311628e-05,
+      "loss": 1.2824,
+      "step": 13620
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.5400278969684065e-05,
+      "loss": 1.3018,
+      "step": 13630
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.5273001415739562e-05,
+      "loss": 1.2786,
+      "step": 13640
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.5145997408096057e-05,
+      "loss": 1.2634,
+      "step": 13650
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.503192806757474e-05,
+      "loss": 1.3144,
+      "step": 13660
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.4905445077906675e-05,
+      "loss": 1.2597,
+      "step": 13670
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.477923698001955e-05,
+      "loss": 1.2646,
+      "step": 13680
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.4653304235911823e-05,
+      "loss": 1.2471,
+      "step": 13690
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.4527647306573998e-05,
+      "loss": 1.2835,
+      "step": 13700
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.4402266651986927e-05,
+      "loss": 1.2674,
+      "step": 13710
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.4277162731120108e-05,
+      "loss": 1.2984,
+      "step": 13720
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.4152336001930054e-05,
+      "loss": 1.2879,
+      "step": 13730
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.4027786921358607e-05,
+      "loss": 1.2361,
+      "step": 13740
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.3903515945331155e-05,
+      "loss": 1.3072,
+      "step": 13750
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.3779523528755145e-05,
+      "loss": 1.2665,
+      "step": 13760
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.3655810125518284e-05,
+      "loss": 1.3312,
+      "step": 13770
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.3532376188486948e-05,
+      "loss": 1.3107,
+      "step": 13780
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.340922216950443e-05,
+      "loss": 1.2497,
+      "step": 13790
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.328634851938949e-05,
+      "loss": 1.3204,
+      "step": 13800
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.316375568793443e-05,
+      "loss": 1.2692,
+      "step": 13810
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.3041444123903668e-05,
+      "loss": 1.26,
+      "step": 13820
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.2919414275031914e-05,
+      "loss": 1.2608,
+      "step": 13830
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.2797666588022748e-05,
+      "loss": 1.2862,
+      "step": 13840
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.2676201508546792e-05,
+      "loss": 1.2762,
+      "step": 13850
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.2567124933972495e-05,
+      "loss": 1.3093,
+      "step": 13860
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.2470360715755768e-05,
+      "loss": 1.2785,
+      "step": 13870
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.2349660894643332e-05,
+      "loss": 1.2841,
+      "step": 13880
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.222924532103765e-05,
+      "loss": 1.2676,
+      "step": 13890
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.2109114435733026e-05,
+      "loss": 1.2522,
+      "step": 13900
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.19892686784816e-05,
+      "loss": 1.2573,
+      "step": 13910
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.1869708487991812e-05,
+      "loss": 1.2572,
+      "step": 13920
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.1750434301926704e-05,
+      "loss": 1.23,
+      "step": 13930
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.163144655690249e-05,
+      "loss": 1.2547,
+      "step": 13940
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.1512745688486646e-05,
+      "loss": 1.29,
+      "step": 13950
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.139433213119664e-05,
+      "loss": 1.2863,
+      "step": 13960
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.127620631849816e-05,
+      "loss": 1.2471,
+      "step": 13970
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.11583686828036e-05,
+      "loss": 1.2756,
+      "step": 13980
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.104081965547041e-05,
+      "loss": 1.272,
+      "step": 13990
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.092355966679961e-05,
+      "loss": 1.2714,
+      "step": 14000
+    },
+    {
+      "epoch": 2.43,
+      "eval_loss": 0.7770761251449585,
+      "eval_runtime": 62.0435,
+      "eval_samples_per_second": 8.446,
+      "eval_steps_per_second": 0.532,
+      "step": 14000
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.080658914603415e-05,
+      "loss": 1.3074,
+      "step": 14010
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.068990852135728e-05,
+      "loss": 1.2862,
+      "step": 14020
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.057351821989113e-05,
+      "loss": 1.285,
+      "step": 14030
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.045741866769507e-05,
+      "loss": 1.2885,
+      "step": 14040
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.034161028976408e-05,
+      "loss": 1.2729,
+      "step": 14050
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.0226093510027388e-05,
+      "loss": 1.2783,
+      "step": 14060
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.0110868751346678e-05,
+      "loss": 1.2502,
+      "step": 14070
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.999593643551475e-05,
+      "loss": 1.2787,
+      "step": 14080
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9881296983253773e-05,
+      "loss": 1.2514,
+      "step": 14090
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9766950814213946e-05,
+      "loss": 1.3199,
+      "step": 14100
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.966429036520796e-05,
+      "loss": 1.321,
+      "step": 14110
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9550502586578255e-05,
+      "loss": 1.2912,
+      "step": 14120
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9437009302078558e-05,
+      "loss": 1.2402,
+      "step": 14130
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9323810927163365e-05,
+      "loss": 1.2623,
+      "step": 14140
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.921090787620764e-05,
+      "loss": 1.2941,
+      "step": 14150
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 1.2472,
+      "step": 14160
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.8985989398267557e-05,
+      "loss": 1.2568,
+      "step": 14170
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.887397479462174e-05,
+      "loss": 1.2569,
+      "step": 14180
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.8762257161609442e-05,
+      "loss": 1.2837,
+      "step": 14190
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.865083690818521e-05,
+      "loss": 1.255,
+      "step": 14200
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.8550813276774915e-05,
+      "loss": 1.264,
+      "step": 14210
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.845103114979575e-05,
+      "loss": 1.2722,
+      "step": 14220
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.8340445725584443e-05,
+      "loss": 1.2828,
+      "step": 14230
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.8230159225047806e-05,
+      "loss": 1.2776,
+      "step": 14240
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.8120172051901564e-05,
+      "loss": 1.2505,
+      "step": 14250
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.801048460876572e-05,
+      "loss": 1.2663,
+      "step": 14260
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.7901097297163094e-05,
+      "loss": 1.2305,
+      "step": 14270
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.779201051751783e-05,
+      "loss": 1.2955,
+      "step": 14280
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.768322466915392e-05,
+      "loss": 1.2682,
+      "step": 14290
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.7574740150293778e-05,
+      "loss": 1.2796,
+      "step": 14300
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.746655735805681e-05,
+      "loss": 1.2872,
+      "step": 14310
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.7380228633595075e-05,
+      "loss": 1.2768,
+      "step": 14320
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.7272589946494132e-05,
+      "loss": 1.2804,
+      "step": 14330
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.7165254092070015e-05,
+      "loss": 1.2582,
+      "step": 14340
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.7058221463237277e-05,
+      "loss": 1.3296,
+      "step": 14350
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.695149245180051e-05,
+      "loss": 1.2726,
+      "step": 14360
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.685569625731185e-05,
+      "loss": 1.273,
+      "step": 14370
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6749545194367288e-05,
+      "loss": 1.3058,
+      "step": 14380
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6643698878761716e-05,
+      "loss": 1.295,
+      "step": 14390
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6538157697957113e-05,
+      "loss": 1.3008,
+      "step": 14400
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.643292203829839e-05,
+      "loss": 1.2531,
+      "step": 14410
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.632799228501215e-05,
+      "loss": 1.2844,
+      "step": 14420
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.622336882220514e-05,
+      "loss": 1.2638,
+      "step": 14430
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.6119052032862915e-05,
+      "loss": 1.2261,
+      "step": 14440
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.601504229884846e-05,
+      "loss": 1.2561,
+      "step": 14450
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.5911340000900688e-05,
+      "loss": 1.2693,
+      "step": 14460
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.580794551863316e-05,
+      "loss": 1.267,
+      "step": 14470
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.5704859230532563e-05,
+      "loss": 1.3048,
+      "step": 14480
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.560208151395749e-05,
+      "loss": 1.2803,
+      "step": 14490
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.549961274513695e-05,
+      "loss": 1.2607,
+      "step": 14500
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.5407655313570525e-05,
+      "loss": 1.275,
+      "step": 14510
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.5315948706191573e-05,
+      "loss": 1.2627,
+      "step": 14520
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.5214346982990213e-05,
+      "loss": 1.2514,
+      "step": 14530
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.5113055626887762e-05,
+      "loss": 1.2496,
+      "step": 14540
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.5012075008672267e-05,
+      "loss": 1.3028,
+      "step": 14550
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.4911405497994235e-05,
+      "loss": 1.2599,
+      "step": 14560
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.4811047463365357e-05,
+      "loss": 1.2633,
+      "step": 14570
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.4711001272157132e-05,
+      "loss": 1.2443,
+      "step": 14580
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.4611267290599528e-05,
+      "loss": 1.3036,
+      "step": 14590
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.4511845883779607e-05,
+      "loss": 1.2934,
+      "step": 14600
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.4412737415640232e-05,
+      "loss": 1.2485,
+      "step": 14610
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.4313942248978752e-05,
+      "loss": 1.2625,
+      "step": 14620
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.4235131935781309e-05,
+      "loss": 1.272,
+      "step": 14630
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.4136901622367581e-05,
+      "loss": 1.2825,
+      "step": 14640
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.403898562015863e-05,
+      "loss": 1.2842,
+      "step": 14650
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.3941384287586633e-05,
+      "loss": 1.2833,
+      "step": 14660
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.384409798193188e-05,
+      "loss": 1.2957,
+      "step": 14670
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.3747127059321474e-05,
+      "loss": 1.2412,
+      "step": 14680
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.3650471874727967e-05,
+      "loss": 1.2911,
+      "step": 14690
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.3554132781968232e-05,
+      "loss": 1.3062,
+      "step": 14700
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.3458110133701962e-05,
+      "loss": 1.2822,
+      "step": 14710
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.3362404281430497e-05,
+      "loss": 1.2376,
+      "step": 14720
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.3267015575495512e-05,
+      "loss": 1.2577,
+      "step": 14730
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.3171944365077748e-05,
+      "loss": 1.2595,
+      "step": 14740
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.307719099819571e-05,
+      "loss": 1.2946,
+      "step": 14750
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.2982755821704372e-05,
+      "loss": 1.2915,
+      "step": 14760
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.288863918129396e-05,
+      "loss": 1.2599,
+      "step": 14770
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.2794841421488679e-05,
+      "loss": 1.2552,
+      "step": 14780
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.2710696364389941e-05,
+      "loss": 1.2647,
+      "step": 14790
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.2626810128213363e-05,
+      "loss": 1.2425,
+      "step": 14800
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.2533907057030315e-05,
+      "loss": 1.2571,
+      "step": 14810
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.244132416498789e-05,
+      "loss": 1.297,
+      "step": 14820
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.2349061790995841e-05,
+      "loss": 1.2686,
+      "step": 14830
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.225712027279059e-05,
+      "loss": 1.2944,
+      "step": 14840
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.21654999469341e-05,
+      "loss": 1.2394,
+      "step": 14850
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.2074201148812537e-05,
+      "loss": 1.2908,
+      "step": 14860
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1983224212635024e-05,
+      "loss": 1.2721,
+      "step": 14870
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1892569471432557e-05,
+      "loss": 1.2818,
+      "step": 14880
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1802237257056659e-05,
+      "loss": 1.2811,
+      "step": 14890
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.171222790017823e-05,
+      "loss": 1.2835,
+      "step": 14900
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.1622541730286296e-05,
+      "loss": 1.2731,
+      "step": 14910
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.153317907568684e-05,
+      "loss": 1.2946,
+      "step": 14920
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.1444140263501591e-05,
+      "loss": 1.2726,
+      "step": 14930
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.135542561966675e-05,
+      "loss": 1.2807,
+      "step": 14940
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.1275859872585081e-05,
+      "loss": 1.2817,
+      "step": 14950
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.1187762042319471e-05,
+      "loss": 1.2802,
+      "step": 14960
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.1108751952271423e-05,
+      "loss": 1.2721,
+      "step": 14970
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.1021272099769108e-05,
+      "loss": 1.2398,
+      "step": 14980
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.093411796357211e-05,
+      "loss": 1.2574,
+      "step": 14990
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.0847289862717614e-05,
+      "loss": 1.2228,
+      "step": 15000
+    },
+    {
+      "epoch": 2.6,
+      "eval_loss": 0.7751675248146057,
+      "eval_runtime": 62.0277,
+      "eval_samples_per_second": 8.448,
+      "eval_steps_per_second": 0.532,
+      "step": 15000
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.0760788115049313e-05,
+      "loss": 1.3108,
+      "step": 15010
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.0674613037216263e-05,
+      "loss": 1.2385,
+      "step": 15020
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.0588764944671713e-05,
+      "loss": 1.2627,
+      "step": 15030
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.0503244151671942e-05,
+      "loss": 1.2532,
+      "step": 15040
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.0426555537850258e-05,
+      "loss": 1.2731,
+      "step": 15050
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.034165747546959e-05,
+      "loss": 1.2618,
+      "step": 15060
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.0257087617197447e-05,
+      "loss": 1.2941,
+      "step": 15070
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.017284627261097e-05,
+      "loss": 1.229,
+      "step": 15080
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.008893375008475e-05,
+      "loss": 1.3288,
+      "step": 15090
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.0005350356789733e-05,
+      "loss": 1.2818,
+      "step": 15100
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 9.922096398692005e-06,
+      "loss": 1.2817,
+      "step": 15110
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 9.839172180551736e-06,
+      "loss": 1.2831,
+      "step": 15120
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.756578005922001e-06,
+      "loss": 1.2657,
+      "step": 15130
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.674314177147791e-06,
+      "loss": 1.2788,
+      "step": 15140
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.592380995364781e-06,
+      "loss": 1.2736,
+      "step": 15150
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.510778760498273e-06,
+      "loss": 1.262,
+      "step": 15160
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.429507771262148e-06,
+      "loss": 1.2497,
+      "step": 15170
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.348568325157681e-06,
+      "loss": 1.2698,
+      "step": 15180
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 9.267960718472513e-06,
+      "loss": 1.2894,
+      "step": 15190
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 9.187685246279565e-06,
+      "loss": 1.277,
+      "step": 15200
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 9.107742202435876e-06,
+      "loss": 1.2803,
+      "step": 15210
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 9.028131879581714e-06,
+      "loss": 1.2451,
+      "step": 15220
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 8.948854569139287e-06,
+      "loss": 1.241,
+      "step": 15230
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.8699105613118e-06,
+      "loss": 1.2558,
+      "step": 15240
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.79914616687264e-06,
+      "loss": 1.2357,
+      "step": 15250
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.720836229152817e-06,
+      "loss": 1.2819,
+      "step": 15260
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.642860428733857e-06,
+      "loss": 1.288,
+      "step": 15270
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.565219051054663e-06,
+      "loss": 1.283,
+      "step": 15280
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.495627977514654e-06,
+      "loss": 1.2858,
+      "step": 15290
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 8.41862278503991e-06,
+      "loss": 1.2931,
+      "step": 15300
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 8.341952836151169e-06,
+      "loss": 1.2803,
+      "step": 15310
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 8.265618411507148e-06,
+      "loss": 1.2773,
+      "step": 15320
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 8.189619790538295e-06,
+      "loss": 1.2717,
+      "step": 15330
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 8.113957251445836e-06,
+      "loss": 1.2474,
+      "step": 15340
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 8.038631071200698e-06,
+      "loss": 1.2828,
+      "step": 15350
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 7.963641525542564e-06,
+      "loss": 1.2829,
+      "step": 15360
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 7.888988888978833e-06,
+      "loss": 1.2845,
+      "step": 15370
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 7.814673434783604e-06,
+      "loss": 1.2726,
+      "step": 15380
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 7.740695434996626e-06,
+      "loss": 1.2498,
+      "step": 15390
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 7.667055160422431e-06,
+      "loss": 1.2746,
+      "step": 15400
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 7.593752880629257e-06,
+      "loss": 1.271,
+      "step": 15410
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 7.52078886394807e-06,
+      "loss": 1.256,
+      "step": 15420
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 7.448163377471562e-06,
+      "loss": 1.2778,
+      "step": 15430
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 7.375876687053251e-06,
+      "loss": 1.2898,
+      "step": 15440
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 7.303929057306414e-06,
+      "loss": 1.2512,
+      "step": 15450
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 7.23232075160315e-06,
+      "loss": 1.2936,
+      "step": 15460
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 7.161052032073445e-06,
+      "loss": 1.2946,
+      "step": 15470
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 7.097200746323862e-06,
+      "loss": 1.2764,
+      "step": 15480
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 7.026577958239167e-06,
+      "loss": 1.2406,
+      "step": 15490
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 6.956295509471921e-06,
+      "loss": 1.2662,
+      "step": 15500
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 6.88635365729865e-06,
+      "loss": 1.244,
+      "step": 15510
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 6.8167526577491034e-06,
+      "loss": 1.2808,
+      "step": 15520
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.747492765605312e-06,
+      "loss": 1.3011,
+      "step": 15530
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.678574234400659e-06,
+      "loss": 1.2447,
+      "step": 15540
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.60999731641887e-06,
+      "loss": 1.2038,
+      "step": 15550
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.548570377045693e-06,
+      "loss": 1.3007,
+      "step": 15560
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.480643214749759e-06,
+      "loss": 1.2823,
+      "step": 15570
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.413058390224724e-06,
+      "loss": 1.2388,
+      "step": 15580
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 6.345816150872197e-06,
+      "loss": 1.2874,
+      "step": 15590
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 6.278916742839691e-06,
+      "loss": 1.2493,
+      "step": 15600
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 6.2123604110197686e-06,
+      "loss": 1.282,
+      "step": 15610
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 6.146147399049107e-06,
+      "loss": 1.2575,
+      "step": 15620
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 6.0802779493076665e-06,
+      "loss": 1.286,
+      "step": 15630
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 6.014752302917681e-06,
+      "loss": 1.281,
+      "step": 15640
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.949570699742935e-06,
+      "loss": 1.2855,
+      "step": 15650
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.8847333783877635e-06,
+      "loss": 1.2316,
+      "step": 15660
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.820240576196223e-06,
+      "loss": 1.2645,
+      "step": 15670
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.7560925292512335e-06,
+      "loss": 1.2897,
+      "step": 15680
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.69228947237368e-06,
+      "loss": 1.2823,
+      "step": 15690
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.635161880753381e-06,
+      "loss": 1.2788,
+      "step": 15700
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5.572014947411885e-06,
+      "loss": 1.2814,
+      "step": 15710
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5.515478243480177e-06,
+      "loss": 1.2719,
+      "step": 15720
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5.452988268147996e-06,
+      "loss": 1.2618,
+      "step": 15730
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5.390844392429362e-06,
+      "loss": 1.3436,
+      "step": 15740
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5.329046843808683e-06,
+      "loss": 1.2658,
+      "step": 15750
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5.267595848502604e-06,
+      "loss": 1.2742,
+      "step": 15760
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5.2064916314591646e-06,
+      "loss": 1.2553,
+      "step": 15770
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5.145734416356996e-06,
+      "loss": 1.2679,
+      "step": 15780
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5.085324425604499e-06,
+      "loss": 1.2254,
+      "step": 15790
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5.025261880338994e-06,
+      "loss": 1.2656,
+      "step": 15800
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.965547000425985e-06,
+      "loss": 1.2524,
+      "step": 15810
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.9061800044582385e-06,
+      "loss": 1.2899,
+      "step": 15820
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.853047328501259e-06,
+      "loss": 1.2837,
+      "step": 15830
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.794341909691191e-06,
+      "loss": 1.2689,
+      "step": 15840
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.735985001541243e-06,
+      "loss": 1.2794,
+      "step": 15850
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.677976817673235e-06,
+      "loss": 1.2599,
+      "step": 15860
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.62031757043242e-06,
+      "loss": 1.2905,
+      "step": 15870
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.563007470886749e-06,
+      "loss": 1.2726,
+      "step": 15880
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.506046728826075e-06,
+      "loss": 1.2318,
+      "step": 15890
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.449435552761372e-06,
+      "loss": 1.2712,
+      "step": 15900
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.398784544532874e-06,
+      "loss": 1.3048,
+      "step": 15910
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.342838113724712e-06,
+      "loss": 1.2803,
+      "step": 15920
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.2872418463554055e-06,
+      "loss": 1.3073,
+      "step": 15930
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.231995945941125e-06,
+      "loss": 1.2495,
+      "step": 15940
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.1771006147155015e-06,
+      "loss": 1.2985,
+      "step": 15950
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.122556053628868e-06,
+      "loss": 1.2603,
+      "step": 15960
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.068362462347508e-06,
+      "loss": 1.2751,
+      "step": 15970
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.014520039252956e-06,
+      "loss": 1.2502,
+      "step": 15980
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.961028981441251e-06,
+      "loss": 1.2988,
+      "step": 15990
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.907889484722238e-06,
+      "loss": 1.2901,
+      "step": 16000
+    },
+    {
+      "epoch": 2.78,
+      "eval_loss": 0.7744143605232239,
+      "eval_runtime": 61.9699,
+      "eval_samples_per_second": 8.456,
+      "eval_steps_per_second": 0.533,
+      "step": 16000
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.855101743618806e-06,
+      "loss": 1.278,
+      "step": 16010
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.8026659513662353e-06,
+      "loss": 1.2782,
+      "step": 16020
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.7505822999114206e-06,
+      "loss": 1.266,
+      "step": 16030
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.6988509799122494e-06,
+      "loss": 1.2606,
+      "step": 16040
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.647472180736833e-06,
+      "loss": 1.2544,
+      "step": 16050
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.5964460904628685e-06,
+      "loss": 1.2632,
+      "step": 16060
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.5457728958768642e-06,
+      "loss": 1.2793,
+      "step": 16070
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.495452782473596e-06,
+      "loss": 1.2691,
+      "step": 16080
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.4454859344552835e-06,
+      "loss": 1.2889,
+      "step": 16090
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.4008179643440496e-06,
+      "loss": 1.2798,
+      "step": 16100
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.3515228234023422e-06,
+      "loss": 1.2751,
+      "step": 16110
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.307459683817815e-06,
+      "loss": 1.2461,
+      "step": 16120
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.2588369013774933e-06,
+      "loss": 1.2488,
+      "step": 16130
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.210568250480306e-06,
+      "loss": 1.2522,
+      "step": 16140
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.1626539078188687e-06,
+      "loss": 1.2958,
+      "step": 16150
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.1150940487888804e-06,
+      "loss": 1.2353,
+      "step": 16160
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.0678888474883316e-06,
+      "loss": 1.2498,
+      "step": 16170
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.0210384767169975e-06,
+      "loss": 1.2708,
+      "step": 16180
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.97454310797578e-06,
+      "loss": 1.2368,
+      "step": 16190
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.9284029114660107e-06,
+      "loss": 1.2822,
+      "step": 16200
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.8826180560888927e-06,
+      "loss": 1.2863,
+      "step": 16210
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.837188709444882e-06,
+      "loss": 1.2477,
+      "step": 16220
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.792115037833032e-06,
+      "loss": 1.2577,
+      "step": 16230
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.7473972062503905e-06,
+      "loss": 1.2445,
+      "step": 16240
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.707455536371439e-06,
+      "loss": 1.2237,
+      "step": 16250
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.6634142507455885e-06,
+      "loss": 1.2587,
+      "step": 16260
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.624081735149897e-06,
+      "loss": 1.2759,
+      "step": 16270
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.580717577477021e-06,
+      "loss": 1.2786,
+      "step": 16280
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.5377100336767545e-06,
+      "loss": 1.272,
+      "step": 16290
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.495059261182886e-06,
+      "loss": 1.2404,
+      "step": 16300
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.452765416123215e-06,
+      "loss": 1.2751,
+      "step": 16310
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.4108286533189527e-06,
+      "loss": 1.2624,
+      "step": 16320
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.3692491262841785e-06,
+      "loss": 1.2965,
+      "step": 16330
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.3280269872252847e-06,
+      "loss": 1.2947,
+      "step": 16340
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.287162387040365e-06,
+      "loss": 1.2839,
+      "step": 16350
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.2506900662738086e-06,
+      "loss": 1.2637,
+      "step": 16360
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.210505200985846e-06,
+      "loss": 1.2615,
+      "step": 16370
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.1706783047731326e-06,
+      "loss": 1.2743,
+      "step": 16380
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.1312095234263807e-06,
+      "loss": 1.2368,
+      "step": 16390
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.0920990014253185e-06,
+      "loss": 1.2542,
+      "step": 16400
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.0533468819382893e-06,
+      "loss": 1.2367,
+      "step": 16410
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.014953306821632e-06,
+      "loss": 1.2418,
+      "step": 16420
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.976918416619211e-06,
+      "loss": 1.2555,
+      "step": 16430
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.939242350561854e-06,
+      "loss": 1.2583,
+      "step": 16440
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.9019252465669046e-06,
+      "loss": 1.2734,
+      "step": 16450
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.8649672412376916e-06,
+      "loss": 1.2125,
+      "step": 16460
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.8283684698629843e-06,
+      "loss": 1.2853,
+      "step": 16470
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.7921290664165923e-06,
+      "loss": 1.2621,
+      "step": 16480
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.756249163556778e-06,
+      "loss": 1.2599,
+      "step": 16490
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.7207288926258225e-06,
+      "loss": 1.2865,
+      "step": 16500
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.6855683836495383e-06,
+      "loss": 1.238,
+      "step": 16510
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.6507677653367915e-06,
+      "loss": 1.2989,
+      "step": 16520
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.6163271650790456e-06,
+      "loss": 1.2784,
+      "step": 16530
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.5822467089498304e-06,
+      "loss": 1.2912,
+      "step": 16540
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.5485265217043854e-06,
+      "loss": 1.2584,
+      "step": 16550
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.5184864851265469e-06,
+      "loss": 1.2535,
+      "step": 16560
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.4854511477372047e-06,
+      "loss": 1.3007,
+      "step": 16570
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.456027673515925e-06,
+      "loss": 1.2791,
+      "step": 16580
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.4236776225376336e-06,
+      "loss": 1.2686,
+      "step": 16590
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.3916884209024705e-06,
+      "loss": 1.2315,
+      "step": 16600
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.3600601857104101e-06,
+      "loss": 1.2747,
+      "step": 16610
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.3287930327400167e-06,
+      "loss": 1.2595,
+      "step": 16620
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.2978870764481232e-06,
+      "loss": 1.2505,
+      "step": 16630
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.2673424299693204e-06,
+      "loss": 1.2814,
+      "step": 16640
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.2371592051156345e-06,
+      "loss": 1.2427,
+      "step": 16650
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.2073375123760168e-06,
+      "loss": 1.2477,
+      "step": 16660
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.1778774609160436e-06,
+      "loss": 1.2516,
+      "step": 16670
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.1487791585774176e-06,
+      "loss": 1.2804,
+      "step": 16680
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.1200427118776224e-06,
+      "loss": 1.2826,
+      "step": 16690
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0916682260095789e-06,
+      "loss": 1.2703,
+      "step": 16700
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.063655804841146e-06,
+      "loss": 1.2501,
+      "step": 16710
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0360055509148535e-06,
+      "loss": 1.2323,
+      "step": 16720
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.008717565447448e-06,
+      "loss": 1.2773,
+      "step": 16730
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 9.871480775350161e-07,
+      "loss": 1.3079,
+      "step": 16740
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 9.605124261266474e-07,
+      "loss": 1.2767,
+      "step": 16750
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 9.34239319527963e-07,
+      "loss": 1.2475,
+      "step": 16760
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 9.083288539145196e-07,
+      "loss": 1.2586,
+      "step": 16770
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 8.827811241344131e-07,
+      "loss": 1.2465,
+      "step": 16780
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 8.575962237078572e-07,
+      "loss": 1.2841,
+      "step": 16790
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 8.327742448269394e-07,
+      "loss": 1.2984,
+      "step": 16800
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 8.083152783552095e-07,
+      "loss": 1.2587,
+      "step": 16810
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 7.842194138273584e-07,
+      "loss": 1.2659,
+      "step": 16820
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 7.628436608436595e-07,
+      "loss": 1.2654,
+      "step": 16830
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 7.3943793191662e-07,
+      "loss": 1.2415,
+      "step": 16840
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 7.163955570664738e-07,
+      "loss": 1.2256,
+      "step": 16850
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 6.937166206423485e-07,
+      "loss": 1.2541,
+      "step": 16860
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 6.714012056629693e-07,
+      "loss": 1.2869,
+      "step": 16870
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 6.494493938163038e-07,
+      "loss": 1.2867,
+      "step": 16880
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 6.278612654593729e-07,
+      "loss": 1.242,
+      "step": 16890
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 6.066368996178517e-07,
+      "loss": 1.2183,
+      "step": 16900
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5.85776373985858e-07,
+      "loss": 1.2517,
+      "step": 16910
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.652797649255969e-07,
+      "loss": 1.2363,
+      "step": 16920
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.4514714746714e-07,
+      "loss": 1.2635,
+      "step": 16930
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.253785953081125e-07,
+      "loss": 1.2782,
+      "step": 16940
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.059741808134621e-07,
+      "loss": 1.3006,
+      "step": 16950
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 4.869339750151469e-07,
+      "loss": 1.2425,
+      "step": 16960
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 4.682580476119247e-07,
+      "loss": 1.276,
+      "step": 16970
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 4.499464669690423e-07,
+      "loss": 1.2827,
+      "step": 16980
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 4.3199930011802446e-07,
+      "loss": 1.3223,
+      "step": 16990
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 4.1441661275645195e-07,
+      "loss": 1.2453,
+      "step": 17000
+    },
+    {
+      "epoch": 2.95,
+      "eval_loss": 0.7738975882530212,
+      "eval_runtime": 61.8545,
+      "eval_samples_per_second": 8.471,
+      "eval_steps_per_second": 0.534,
+      "step": 17000
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.971984692476394e-07,
+      "loss": 1.2683,
+      "step": 17010
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.820138772047788e-07,
+      "loss": 1.2824,
+      "step": 17020
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.6548853955771235e-07,
+      "loss": 1.271,
+      "step": 17030
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.493279248699355e-07,
+      "loss": 1.298,
+      "step": 17040
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.3353209229913806e-07,
+      "loss": 1.2695,
+      "step": 17050
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.181010996677003e-07,
+      "loss": 1.2698,
+      "step": 17060
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.030350034624374e-07,
+      "loss": 1.266,
+      "step": 17070
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.88333858834422e-07,
+      "loss": 1.3003,
+      "step": 17080
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.7399771959880637e-07,
+      "loss": 1.2441,
+      "step": 17090
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.600266382345895e-07,
+      "loss": 1.2973,
+      "step": 17100
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.4642066588441705e-07,
+      "loss": 1.2442,
+      "step": 17110
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.3317985235443707e-07,
+      "loss": 1.2901,
+      "step": 17120
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.215753710563373e-07,
+      "loss": 1.2668,
+      "step": 17130
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.0902849171310356e-07,
+      "loss": 1.2799,
+      "step": 17140
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.968469080681823e-07,
+      "loss": 1.2652,
+      "step": 17150
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.8619584749273167e-07,
+      "loss": 1.2909,
+      "step": 17160
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.747084474202576e-07,
+      "loss": 1.2791,
+      "step": 17170
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.6358646867835615e-07,
+      "loss": 1.269,
+      "step": 17180
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.5282995198021565e-07,
+      "loss": 1.2189,
+      "step": 17190
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.424389367012613e-07,
+      "loss": 1.2991,
+      "step": 17200
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.3241346087892182e-07,
+      "loss": 1.2893,
+      "step": 17210
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.2275356121254077e-07,
+      "loss": 1.2596,
+      "step": 17220
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.1345927306323224e-07,
+      "loss": 1.2393,
+      "step": 17230
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.0453063045375855e-07,
+      "loss": 1.2814,
+      "step": 17240
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.596766606836393e-08,
+      "loss": 1.2632,
+      "step": 17250
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.777041125273e-08,
+      "loss": 1.2705,
+      "step": 17260
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 7.993889601378701e-08,
+      "loss": 1.2207,
+      "step": 17270
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 7.24731490196584e-08,
+      "loss": 1.2858,
+      "step": 17280
+    },
+    {
+      "epoch": 3.0,
+      "step": 17285,
+      "total_flos": 9.82692376847319e+16,
+      "train_loss": 1.5514631569575188,
+      "train_runtime": 557290.3071,
+      "train_samples_per_second": 3.97,
+      "train_steps_per_second": 0.031
+    }
+  ],
+  "max_steps": 17285,
+  "num_train_epochs": 4,
+  "total_flos": 9.82692376847319e+16,
+  "trial_name": null,
+  "trial_params": null
+}