Upload 12 files

models/lora_weights/chreckpoint-6000/adapter_config.json

@@ -0,0 +1,17 @@
+{
+  "base_model_name_or_path": "../alpaca_weights/chinese-alpaca",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

models/lora_weights/chreckpoint-6000/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5e1621f48d9ad8feb1d6d31050275f0aafd080c5c07153301fe2f48411f4406
+size 443

models/lora_weights/chreckpoint-6000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24c6c6c69e422a9866b7e63c2105fa82e54729b25a736f558a023442102d0a61
+size 52562821

models/lora_weights/chreckpoint-6000/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:299c3d73b7055b0575c71ab25f0f4654913c2f6756efefb3f2746c403834dc02
+size 19639

models/lora_weights/chreckpoint-6000/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a777a12417bf6e41c4dda371cfcb9199f23f25f5637806ac2caeaa43af71825
+size 19639

models/lora_weights/chreckpoint-6000/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21475b9e62fa35cc2d8376ab55c56b2a7e6ce65d6b6d92853b43491060cbe4c3
+size 19639

models/lora_weights/chreckpoint-6000/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc475b3aa37f01ed5a14ca69d49d559e4cfef1a4d013577aaefc788b04482440
+size 19639

models/lora_weights/chreckpoint-6000/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b47919b7327a49b2f503a4016fa7fd95bcf64d51ec91b8cfcf64458ab4a4839a
+size 19639

models/lora_weights/chreckpoint-6000/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a6b0d94556d648cd537b0ae5bc0078f84327bfe34f981e4340ddc55b00377d0
+size 19639

models/lora_weights/chreckpoint-6000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bea775cd3a033d0f762b2df7c1217ebdbe9e335687e204a10a0620aa276592b6
+size 627

models/lora_weights/chreckpoint-6000/trainer_state.json

@@ -0,0 +1,3859 @@
+{
+  "best_metric": 0.9106292724609375,
+  "best_model_checkpoint": "./results/chinese-llama-2-lora-2nov/checkpoint-6000",
+  "epoch": 9.868421052631579,
+  "eval_steps": 200,
+  "global_step": 6000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 2.2931,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 2.2482,
+      "step": 20
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 2.0171,
+      "step": 30
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 1.6749,
+      "step": 40
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00015,
+      "loss": 1.5164,
+      "step": 50
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 1.4101,
+      "step": 60
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 1.3538,
+      "step": 70
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 1.2861,
+      "step": 80
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00027,
+      "loss": 1.2444,
+      "step": 90
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003,
+      "loss": 1.2369,
+      "step": 100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002994983277591973,
+      "loss": 1.2258,
+      "step": 110
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002989966555183946,
+      "loss": 1.1953,
+      "step": 120
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00029849498327759195,
+      "loss": 1.2009,
+      "step": 130
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002979933110367893,
+      "loss": 1.1498,
+      "step": 140
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002974916387959866,
+      "loss": 1.1646,
+      "step": 150
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002969899665551839,
+      "loss": 1.1375,
+      "step": 160
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00029648829431438127,
+      "loss": 1.1429,
+      "step": 170
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00029598662207357856,
+      "loss": 1.1322,
+      "step": 180
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002954849498327759,
+      "loss": 1.1255,
+      "step": 190
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00029498327759197324,
+      "loss": 1.1026,
+      "step": 200
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 1.1218857765197754,
+      "eval_runtime": 33.0988,
+      "eval_samples_per_second": 60.425,
+      "eval_steps_per_second": 1.269,
+      "step": 200
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00029448160535117053,
+      "loss": 1.1085,
+      "step": 210
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002939799331103679,
+      "loss": 1.0916,
+      "step": 220
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002934782608695652,
+      "loss": 1.1211,
+      "step": 230
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002929765886287625,
+      "loss": 1.0773,
+      "step": 240
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00029247491638795985,
+      "loss": 1.085,
+      "step": 250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002919732441471572,
+      "loss": 1.0696,
+      "step": 260
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002914715719063545,
+      "loss": 1.0814,
+      "step": 270
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002909698996655518,
+      "loss": 1.0931,
+      "step": 280
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00029046822742474917,
+      "loss": 1.064,
+      "step": 290
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00028996655518394646,
+      "loss": 1.076,
+      "step": 300
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002894648829431438,
+      "loss": 1.0667,
+      "step": 310
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00028896321070234115,
+      "loss": 1.0682,
+      "step": 320
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00028846153846153843,
+      "loss": 1.0589,
+      "step": 330
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002879598662207358,
+      "loss": 1.0524,
+      "step": 340
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002874581939799331,
+      "loss": 1.061,
+      "step": 350
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002869565217391304,
+      "loss": 1.0575,
+      "step": 360
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00028645484949832775,
+      "loss": 1.0381,
+      "step": 370
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00028595317725752504,
+      "loss": 1.0463,
+      "step": 380
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002854515050167224,
+      "loss": 1.0616,
+      "step": 390
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00028494983277591973,
+      "loss": 1.0462,
+      "step": 400
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 1.0583561658859253,
+      "eval_runtime": 34.9211,
+      "eval_samples_per_second": 57.272,
+      "eval_steps_per_second": 1.203,
+      "step": 400
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.000284448160535117,
+      "loss": 1.0481,
+      "step": 410
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00028394648829431436,
+      "loss": 1.0362,
+      "step": 420
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002834448160535117,
+      "loss": 1.0517,
+      "step": 430
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.000282943143812709,
+      "loss": 1.0465,
+      "step": 440
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00028244147157190634,
+      "loss": 1.0313,
+      "step": 450
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002819397993311037,
+      "loss": 1.0134,
+      "step": 460
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00028143812709030097,
+      "loss": 1.0373,
+      "step": 470
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002809364548494983,
+      "loss": 1.0357,
+      "step": 480
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00028043478260869565,
+      "loss": 1.0113,
+      "step": 490
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00027993311036789294,
+      "loss": 1.0201,
+      "step": 500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002794314381270903,
+      "loss": 1.0359,
+      "step": 510
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00027892976588628763,
+      "loss": 1.0064,
+      "step": 520
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002784280936454849,
+      "loss": 1.0166,
+      "step": 530
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00027792642140468226,
+      "loss": 1.0025,
+      "step": 540
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002774247491638796,
+      "loss": 1.028,
+      "step": 550
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002769230769230769,
+      "loss": 1.0216,
+      "step": 560
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00027642140468227424,
+      "loss": 1.0037,
+      "step": 570
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002759197324414716,
+      "loss": 1.0028,
+      "step": 580
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00027541806020066887,
+      "loss": 1.0193,
+      "step": 590
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00027491638795986616,
+      "loss": 1.0128,
+      "step": 600
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 1.0269545316696167,
+      "eval_runtime": 33.0707,
+      "eval_samples_per_second": 60.476,
+      "eval_steps_per_second": 1.27,
+      "step": 600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0002744147157190635,
+      "loss": 0.9854,
+      "step": 610
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00027391304347826085,
+      "loss": 1.0161,
+      "step": 620
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00027341137123745813,
+      "loss": 0.9755,
+      "step": 630
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0002729096989966555,
+      "loss": 0.997,
+      "step": 640
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0002724080267558528,
+      "loss": 0.9834,
+      "step": 650
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0002719063545150501,
+      "loss": 1.0024,
+      "step": 660
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00027140468227424745,
+      "loss": 0.985,
+      "step": 670
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0002709030100334448,
+      "loss": 1.0037,
+      "step": 680
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0002704013377926421,
+      "loss": 1.0004,
+      "step": 690
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00026989966555183943,
+      "loss": 1.006,
+      "step": 700
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00026939799331103677,
+      "loss": 0.9887,
+      "step": 710
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00026889632107023406,
+      "loss": 0.9817,
+      "step": 720
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0002683946488294314,
+      "loss": 0.9718,
+      "step": 730
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00026789297658862875,
+      "loss": 0.9958,
+      "step": 740
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00026739130434782604,
+      "loss": 0.9654,
+      "step": 750
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0002668896321070234,
+      "loss": 0.9906,
+      "step": 760
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0002663879598662207,
+      "loss": 0.9922,
+      "step": 770
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.000265886287625418,
+      "loss": 0.9803,
+      "step": 780
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00026538461538461536,
+      "loss": 0.9709,
+      "step": 790
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0002648829431438127,
+      "loss": 0.9562,
+      "step": 800
+    },
+    {
+      "epoch": 1.32,
+      "eval_loss": 1.007567048072815,
+      "eval_runtime": 33.1229,
+      "eval_samples_per_second": 60.381,
+      "eval_steps_per_second": 1.268,
+      "step": 800
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00026438127090301,
+      "loss": 0.98,
+      "step": 810
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00026387959866220733,
+      "loss": 1.0006,
+      "step": 820
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0002633779264214047,
+      "loss": 0.9748,
+      "step": 830
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00026287625418060196,
+      "loss": 0.9878,
+      "step": 840
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0002623745819397993,
+      "loss": 0.9827,
+      "step": 850
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00026187290969899665,
+      "loss": 0.9752,
+      "step": 860
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00026137123745819394,
+      "loss": 0.9675,
+      "step": 870
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0002608695652173913,
+      "loss": 0.9563,
+      "step": 880
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0002603678929765886,
+      "loss": 0.99,
+      "step": 890
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0002598662207357859,
+      "loss": 1.0059,
+      "step": 900
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00025936454849498326,
+      "loss": 0.9916,
+      "step": 910
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0002588628762541806,
+      "loss": 0.9866,
+      "step": 920
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0002583612040133779,
+      "loss": 0.9522,
+      "step": 930
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00025785953177257523,
+      "loss": 0.9421,
+      "step": 940
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0002573578595317725,
+      "loss": 0.9678,
+      "step": 950
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00025685618729096986,
+      "loss": 0.9556,
+      "step": 960
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0002563545150501672,
+      "loss": 0.9794,
+      "step": 970
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0002558528428093645,
+      "loss": 0.9603,
+      "step": 980
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00025535117056856184,
+      "loss": 0.9457,
+      "step": 990
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0002548494983277592,
+      "loss": 0.9656,
+      "step": 1000
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 0.9902360439300537,
+      "eval_runtime": 33.1444,
+      "eval_samples_per_second": 60.342,
+      "eval_steps_per_second": 1.267,
+      "step": 1000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00025434782608695647,
+      "loss": 0.955,
+      "step": 1010
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0002538461538461538,
+      "loss": 0.9664,
+      "step": 1020
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00025334448160535116,
+      "loss": 0.9712,
+      "step": 1030
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00025284280936454845,
+      "loss": 0.947,
+      "step": 1040
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0002523411371237458,
+      "loss": 0.973,
+      "step": 1050
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00025183946488294313,
+      "loss": 0.9664,
+      "step": 1060
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0002513377926421404,
+      "loss": 0.9581,
+      "step": 1070
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00025083612040133777,
+      "loss": 0.9544,
+      "step": 1080
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0002503344481605351,
+      "loss": 0.9774,
+      "step": 1090
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0002498327759197324,
+      "loss": 0.9493,
+      "step": 1100
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00024933110367892974,
+      "loss": 0.9584,
+      "step": 1110
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.0002488294314381271,
+      "loss": 0.947,
+      "step": 1120
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0002483277591973244,
+      "loss": 0.9502,
+      "step": 1130
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0002478260869565217,
+      "loss": 0.944,
+      "step": 1140
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00024732441471571906,
+      "loss": 0.9556,
+      "step": 1150
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00024682274247491635,
+      "loss": 0.9627,
+      "step": 1160
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0002463210702341137,
+      "loss": 0.9376,
+      "step": 1170
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00024581939799331104,
+      "loss": 0.9399,
+      "step": 1180
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0002453177257525083,
+      "loss": 0.9449,
+      "step": 1190
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00024481605351170567,
+      "loss": 0.9514,
+      "step": 1200
+    },
+    {
+      "epoch": 1.97,
+      "eval_loss": 0.9769035577774048,
+      "eval_runtime": 33.1416,
+      "eval_samples_per_second": 60.347,
+      "eval_steps_per_second": 1.267,
+      "step": 1200
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.000244314381270903,
+      "loss": 0.9439,
+      "step": 1210
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00024381270903010033,
+      "loss": 0.9512,
+      "step": 1220
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00024331103678929764,
+      "loss": 0.948,
+      "step": 1230
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00024280936454849496,
+      "loss": 0.9421,
+      "step": 1240
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0002423076923076923,
+      "loss": 0.9323,
+      "step": 1250
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00024180602006688962,
+      "loss": 0.9278,
+      "step": 1260
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.00024130434782608694,
+      "loss": 0.9255,
+      "step": 1270
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.00024080267558528428,
+      "loss": 0.9214,
+      "step": 1280
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0002403010033444816,
+      "loss": 0.9456,
+      "step": 1290
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0002397993311036789,
+      "loss": 0.9435,
+      "step": 1300
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.00023929765886287623,
+      "loss": 0.9261,
+      "step": 1310
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.00023879598662207357,
+      "loss": 0.9225,
+      "step": 1320
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.00023829431438127089,
+      "loss": 0.9184,
+      "step": 1330
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.0002377926421404682,
+      "loss": 0.9301,
+      "step": 1340
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00023729096989966555,
+      "loss": 0.9599,
+      "step": 1350
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00023678929765886286,
+      "loss": 0.9316,
+      "step": 1360
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.00023628762541806018,
+      "loss": 0.9214,
+      "step": 1370
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00023578595317725752,
+      "loss": 0.9206,
+      "step": 1380
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.00023528428093645484,
+      "loss": 0.9413,
+      "step": 1390
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00023478260869565215,
+      "loss": 0.9136,
+      "step": 1400
+    },
+    {
+      "epoch": 2.3,
+      "eval_loss": 0.9681736826896667,
+      "eval_runtime": 33.0949,
+      "eval_samples_per_second": 60.432,
+      "eval_steps_per_second": 1.269,
+      "step": 1400
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.0002342809364548495,
+      "loss": 0.9128,
+      "step": 1410
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0002337792642140468,
+      "loss": 0.9229,
+      "step": 1420
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00023327759197324413,
+      "loss": 0.9127,
+      "step": 1430
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00023277591973244147,
+      "loss": 0.9305,
+      "step": 1440
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0002322742474916388,
+      "loss": 0.9238,
+      "step": 1450
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.0002317725752508361,
+      "loss": 0.9179,
+      "step": 1460
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.00023127090301003345,
+      "loss": 0.9392,
+      "step": 1470
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.00023076923076923076,
+      "loss": 0.9301,
+      "step": 1480
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.00023026755852842808,
+      "loss": 0.9298,
+      "step": 1490
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.00022976588628762542,
+      "loss": 0.9426,
+      "step": 1500
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.00022926421404682274,
+      "loss": 0.9228,
+      "step": 1510
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00022876254180602006,
+      "loss": 0.9242,
+      "step": 1520
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0002282608695652174,
+      "loss": 0.9317,
+      "step": 1530
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.00022775919732441471,
+      "loss": 0.9093,
+      "step": 1540
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.00022725752508361203,
+      "loss": 0.91,
+      "step": 1550
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.00022675585284280935,
+      "loss": 0.9177,
+      "step": 1560
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0002262541806020067,
+      "loss": 0.9222,
+      "step": 1570
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.000225752508361204,
+      "loss": 0.9165,
+      "step": 1580
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.00022525083612040132,
+      "loss": 0.9159,
+      "step": 1590
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00022474916387959864,
+      "loss": 0.887,
+      "step": 1600
+    },
+    {
+      "epoch": 2.63,
+      "eval_loss": 0.9578931927680969,
+      "eval_runtime": 33.1845,
+      "eval_samples_per_second": 60.269,
+      "eval_steps_per_second": 1.266,
+      "step": 1600
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.00022424749163879595,
+      "loss": 0.8999,
+      "step": 1610
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00022374581939799327,
+      "loss": 0.9044,
+      "step": 1620
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00022324414715719061,
+      "loss": 0.8885,
+      "step": 1630
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.00022274247491638793,
+      "loss": 0.9134,
+      "step": 1640
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00022224080267558525,
+      "loss": 0.9055,
+      "step": 1650
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.00022173913043478256,
+      "loss": 0.919,
+      "step": 1660
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.0002212374581939799,
+      "loss": 0.9358,
+      "step": 1670
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.00022073578595317722,
+      "loss": 0.918,
+      "step": 1680
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.00022023411371237454,
+      "loss": 0.9072,
+      "step": 1690
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00021973244147157188,
+      "loss": 0.9131,
+      "step": 1700
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.0002192307692307692,
+      "loss": 0.9092,
+      "step": 1710
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.0002187290969899665,
+      "loss": 0.8981,
+      "step": 1720
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.00021822742474916386,
+      "loss": 0.9273,
+      "step": 1730
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.00021772575250836117,
+      "loss": 0.9226,
+      "step": 1740
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.0002172240802675585,
+      "loss": 0.9169,
+      "step": 1750
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00021672240802675583,
+      "loss": 0.9278,
+      "step": 1760
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00021622073578595315,
+      "loss": 0.907,
+      "step": 1770
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.00021571906354515046,
+      "loss": 0.8985,
+      "step": 1780
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.0002152173913043478,
+      "loss": 0.9093,
+      "step": 1790
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.00021471571906354512,
+      "loss": 0.8969,
+      "step": 1800
+    },
+    {
+      "epoch": 2.96,
+      "eval_loss": 0.9505146145820618,
+      "eval_runtime": 33.1216,
+      "eval_samples_per_second": 60.383,
+      "eval_steps_per_second": 1.268,
+      "step": 1800
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00021421404682274244,
+      "loss": 0.8899,
+      "step": 1810
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00021371237458193978,
+      "loss": 0.8995,
+      "step": 1820
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.0002132107023411371,
+      "loss": 0.9034,
+      "step": 1830
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00021270903010033441,
+      "loss": 0.8901,
+      "step": 1840
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.00021220735785953176,
+      "loss": 0.8977,
+      "step": 1850
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00021170568561872907,
+      "loss": 0.891,
+      "step": 1860
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.0002112040133779264,
+      "loss": 0.8848,
+      "step": 1870
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00021070234113712373,
+      "loss": 0.8829,
+      "step": 1880
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00021020066889632105,
+      "loss": 0.8728,
+      "step": 1890
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00020969899665551837,
+      "loss": 0.8927,
+      "step": 1900
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00020919732441471568,
+      "loss": 0.8834,
+      "step": 1910
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.00020869565217391303,
+      "loss": 0.9003,
+      "step": 1920
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00020819397993311034,
+      "loss": 0.8958,
+      "step": 1930
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00020769230769230766,
+      "loss": 0.9081,
+      "step": 1940
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.000207190635451505,
+      "loss": 0.9027,
+      "step": 1950
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.00020668896321070232,
+      "loss": 0.8997,
+      "step": 1960
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00020618729096989963,
+      "loss": 0.8906,
+      "step": 1970
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00020568561872909698,
+      "loss": 0.8886,
+      "step": 1980
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.0002051839464882943,
+      "loss": 0.8886,
+      "step": 1990
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.0002046822742474916,
+      "loss": 0.8885,
+      "step": 2000
+    },
+    {
+      "epoch": 3.29,
+      "eval_loss": 0.9456945061683655,
+      "eval_runtime": 33.1317,
+      "eval_samples_per_second": 60.365,
+      "eval_steps_per_second": 1.268,
+      "step": 2000
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00020418060200668895,
+      "loss": 0.8992,
+      "step": 2010
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.00020367892976588627,
+      "loss": 0.892,
+      "step": 2020
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.00020317725752508358,
+      "loss": 0.8848,
+      "step": 2030
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.00020267558528428093,
+      "loss": 0.8889,
+      "step": 2040
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 0.00020217391304347824,
+      "loss": 0.8872,
+      "step": 2050
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.00020167224080267556,
+      "loss": 0.8933,
+      "step": 2060
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.0002011705685618729,
+      "loss": 0.8816,
+      "step": 2070
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 0.00020066889632107022,
+      "loss": 0.8836,
+      "step": 2080
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 0.00020016722408026753,
+      "loss": 0.8918,
+      "step": 2090
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.00019966555183946488,
+      "loss": 0.8944,
+      "step": 2100
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.0001991638795986622,
+      "loss": 0.9055,
+      "step": 2110
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 0.0001986622073578595,
+      "loss": 0.8809,
+      "step": 2120
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.00019816053511705685,
+      "loss": 0.8982,
+      "step": 2130
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 0.00019765886287625417,
+      "loss": 0.89,
+      "step": 2140
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.00019715719063545149,
+      "loss": 0.8798,
+      "step": 2150
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 0.0001966555183946488,
+      "loss": 0.9001,
+      "step": 2160
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.00019615384615384615,
+      "loss": 0.8746,
+      "step": 2170
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.00019565217391304346,
+      "loss": 0.8945,
+      "step": 2180
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 0.00019515050167224078,
+      "loss": 0.8745,
+      "step": 2190
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 0.00019464882943143812,
+      "loss": 0.8695,
+      "step": 2200
+    },
+    {
+      "epoch": 3.62,
+      "eval_loss": 0.9407016038894653,
+      "eval_runtime": 33.1522,
+      "eval_samples_per_second": 60.328,
+      "eval_steps_per_second": 1.267,
+      "step": 2200
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 0.00019414715719063544,
+      "loss": 0.8917,
+      "step": 2210
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 0.00019364548494983275,
+      "loss": 0.8822,
+      "step": 2220
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 0.0001931438127090301,
+      "loss": 0.8694,
+      "step": 2230
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.0001926421404682274,
+      "loss": 0.8904,
+      "step": 2240
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 0.00019214046822742473,
+      "loss": 0.8759,
+      "step": 2250
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 0.00019163879598662207,
+      "loss": 0.8784,
+      "step": 2260
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 0.0001911371237458194,
+      "loss": 0.8829,
+      "step": 2270
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.0001906354515050167,
+      "loss": 0.8788,
+      "step": 2280
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 0.00019013377926421405,
+      "loss": 0.8966,
+      "step": 2290
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 0.00018963210702341136,
+      "loss": 0.8803,
+      "step": 2300
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 0.00018913043478260868,
+      "loss": 0.8783,
+      "step": 2310
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 0.00018862876254180602,
+      "loss": 0.8874,
+      "step": 2320
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.00018812709030100334,
+      "loss": 0.8711,
+      "step": 2330
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.00018762541806020065,
+      "loss": 0.8676,
+      "step": 2340
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 0.000187123745819398,
+      "loss": 0.8638,
+      "step": 2350
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 0.00018662207357859531,
+      "loss": 0.8945,
+      "step": 2360
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 0.00018612040133779263,
+      "loss": 0.8772,
+      "step": 2370
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 0.00018561872909698997,
+      "loss": 0.8895,
+      "step": 2380
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 0.0001851170568561873,
+      "loss": 0.87,
+      "step": 2390
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.0001846153846153846,
+      "loss": 0.8608,
+      "step": 2400
+    },
+    {
+      "epoch": 3.95,
+      "eval_loss": 0.936105489730835,
+      "eval_runtime": 33.1676,
+      "eval_samples_per_second": 60.3,
+      "eval_steps_per_second": 1.266,
+      "step": 2400
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 0.00018411371237458192,
+      "loss": 0.8848,
+      "step": 2410
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 0.00018361204013377926,
+      "loss": 0.8823,
+      "step": 2420
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 0.00018311036789297658,
+      "loss": 0.8692,
+      "step": 2430
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 0.0001826086956521739,
+      "loss": 0.896,
+      "step": 2440
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 0.00018210702341137124,
+      "loss": 0.8704,
+      "step": 2450
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 0.00018160535117056856,
+      "loss": 0.8869,
+      "step": 2460
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 0.00018110367892976587,
+      "loss": 0.8492,
+      "step": 2470
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 0.00018060200668896322,
+      "loss": 0.8558,
+      "step": 2480
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 0.00018010033444816053,
+      "loss": 0.8607,
+      "step": 2490
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 0.00017959866220735785,
+      "loss": 0.8651,
+      "step": 2500
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 0.0001790969899665552,
+      "loss": 0.8648,
+      "step": 2510
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 0.0001785953177257525,
+      "loss": 0.8614,
+      "step": 2520
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 0.00017809364548494982,
+      "loss": 0.8759,
+      "step": 2530
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 0.00017759197324414717,
+      "loss": 0.8653,
+      "step": 2540
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 0.00017709030100334448,
+      "loss": 0.8478,
+      "step": 2550
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 0.0001765886287625418,
+      "loss": 0.8636,
+      "step": 2560
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 0.00017608695652173914,
+      "loss": 0.859,
+      "step": 2570
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 0.00017558528428093646,
+      "loss": 0.8685,
+      "step": 2580
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 0.00017508361204013377,
+      "loss": 0.8595,
+      "step": 2590
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 0.00017458193979933106,
+      "loss": 0.8698,
+      "step": 2600
+    },
+    {
+      "epoch": 4.28,
+      "eval_loss": 0.9336724877357483,
+      "eval_runtime": 33.1889,
+      "eval_samples_per_second": 60.261,
+      "eval_steps_per_second": 1.265,
+      "step": 2600
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 0.0001740802675585284,
+      "loss": 0.8696,
+      "step": 2610
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 0.00017357859531772572,
+      "loss": 0.8667,
+      "step": 2620
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 0.00017307692307692304,
+      "loss": 0.8648,
+      "step": 2630
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 0.00017257525083612038,
+      "loss": 0.8524,
+      "step": 2640
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 0.0001720735785953177,
+      "loss": 0.8506,
+      "step": 2650
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 0.00017157190635451501,
+      "loss": 0.8638,
+      "step": 2660
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 0.00017107023411371236,
+      "loss": 0.8769,
+      "step": 2670
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 0.00017056856187290967,
+      "loss": 0.8581,
+      "step": 2680
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 0.000170066889632107,
+      "loss": 0.8564,
+      "step": 2690
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 0.00016956521739130433,
+      "loss": 0.8369,
+      "step": 2700
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 0.00016906354515050165,
+      "loss": 0.879,
+      "step": 2710
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 0.00016856187290969897,
+      "loss": 0.8426,
+      "step": 2720
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 0.00016806020066889628,
+      "loss": 0.8603,
+      "step": 2730
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 0.00016755852842809362,
+      "loss": 0.8613,
+      "step": 2740
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 0.00016705685618729094,
+      "loss": 0.8695,
+      "step": 2750
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 0.00016655518394648826,
+      "loss": 0.8652,
+      "step": 2760
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 0.0001660535117056856,
+      "loss": 0.8537,
+      "step": 2770
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 0.00016555183946488292,
+      "loss": 0.8529,
+      "step": 2780
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.00016505016722408023,
+      "loss": 0.8426,
+      "step": 2790
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 0.00016454849498327758,
+      "loss": 0.8691,
+      "step": 2800
+    },
+    {
+      "epoch": 4.61,
+      "eval_loss": 0.9289954900741577,
+      "eval_runtime": 33.1507,
+      "eval_samples_per_second": 60.331,
+      "eval_steps_per_second": 1.267,
+      "step": 2800
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 0.0001640468227424749,
+      "loss": 0.8589,
+      "step": 2810
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 0.0001635451505016722,
+      "loss": 0.8583,
+      "step": 2820
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 0.00016304347826086955,
+      "loss": 0.8408,
+      "step": 2830
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 0.00016254180602006687,
+      "loss": 0.8634,
+      "step": 2840
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 0.00016204013377926418,
+      "loss": 0.8513,
+      "step": 2850
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 0.00016153846153846153,
+      "loss": 0.8568,
+      "step": 2860
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.00016103678929765884,
+      "loss": 0.8703,
+      "step": 2870
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 0.00016053511705685616,
+      "loss": 0.8545,
+      "step": 2880
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 0.0001600334448160535,
+      "loss": 0.8456,
+      "step": 2890
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 0.00015953177257525082,
+      "loss": 0.8548,
+      "step": 2900
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 0.00015903010033444813,
+      "loss": 0.8515,
+      "step": 2910
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 0.00015852842809364548,
+      "loss": 0.8562,
+      "step": 2920
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 0.0001580267558528428,
+      "loss": 0.8623,
+      "step": 2930
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 0.0001575250836120401,
+      "loss": 0.855,
+      "step": 2940
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 0.00015702341137123745,
+      "loss": 0.876,
+      "step": 2950
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 0.00015652173913043477,
+      "loss": 0.8582,
+      "step": 2960
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 0.00015602006688963209,
+      "loss": 0.8488,
+      "step": 2970
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 0.0001555183946488294,
+      "loss": 0.8466,
+      "step": 2980
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 0.00015501672240802674,
+      "loss": 0.8407,
+      "step": 2990
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 0.00015451505016722406,
+      "loss": 0.8732,
+      "step": 3000
+    },
+    {
+      "epoch": 4.93,
+      "eval_loss": 0.9256382584571838,
+      "eval_runtime": 33.1403,
+      "eval_samples_per_second": 60.349,
+      "eval_steps_per_second": 1.267,
+      "step": 3000
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 0.00015401337792642138,
+      "loss": 0.8422,
+      "step": 3010
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 0.00015351170568561872,
+      "loss": 0.8528,
+      "step": 3020
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 0.00015301003344481604,
+      "loss": 0.854,
+      "step": 3030
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.00015250836120401335,
+      "loss": 0.8598,
+      "step": 3040
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 0.0001520066889632107,
+      "loss": 0.843,
+      "step": 3050
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 0.000151505016722408,
+      "loss": 0.8394,
+      "step": 3060
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 0.00015100334448160533,
+      "loss": 0.8593,
+      "step": 3070
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 0.00015050167224080267,
+      "loss": 0.8566,
+      "step": 3080
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 0.00015,
+      "loss": 0.8469,
+      "step": 3090
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 0.0001494983277591973,
+      "loss": 0.8421,
+      "step": 3100
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 0.00014899665551839465,
+      "loss": 0.8425,
+      "step": 3110
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 0.00014849498327759196,
+      "loss": 0.8341,
+      "step": 3120
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 0.00014804347826086954,
+      "loss": 0.8527,
+      "step": 3130
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 0.00014754180602006688,
+      "loss": 0.8517,
+      "step": 3140
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 0.0001470401337792642,
+      "loss": 0.8389,
+      "step": 3150
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 0.00014653846153846151,
+      "loss": 0.8311,
+      "step": 3160
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 0.00014603678929765886,
+      "loss": 0.85,
+      "step": 3170
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 0.00014553511705685617,
+      "loss": 0.8469,
+      "step": 3180
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 0.0001450334448160535,
+      "loss": 0.8569,
+      "step": 3190
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 0.00014453177257525083,
+      "loss": 0.853,
+      "step": 3200
+    },
+    {
+      "epoch": 5.26,
+      "eval_loss": 0.9253825545310974,
+      "eval_runtime": 33.1231,
+      "eval_samples_per_second": 60.381,
+      "eval_steps_per_second": 1.268,
+      "step": 3200
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 0.00014403010033444815,
+      "loss": 0.8445,
+      "step": 3210
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 0.00014352842809364547,
+      "loss": 0.8462,
+      "step": 3220
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 0.0001430267558528428,
+      "loss": 0.8638,
+      "step": 3230
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 0.00014252508361204012,
+      "loss": 0.8405,
+      "step": 3240
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 0.00014202341137123744,
+      "loss": 0.8329,
+      "step": 3250
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 0.00014152173913043476,
+      "loss": 0.8502,
+      "step": 3260
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 0.0001410200668896321,
+      "loss": 0.8398,
+      "step": 3270
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 0.00014051839464882942,
+      "loss": 0.8667,
+      "step": 3280
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 0.00014001672240802673,
+      "loss": 0.8393,
+      "step": 3290
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 0.00013951505016722408,
+      "loss": 0.8493,
+      "step": 3300
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 0.0001390133779264214,
+      "loss": 0.8369,
+      "step": 3310
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 0.0001385117056856187,
+      "loss": 0.8305,
+      "step": 3320
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 0.00013801003344481605,
+      "loss": 0.8425,
+      "step": 3330
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 0.00013750836120401337,
+      "loss": 0.8392,
+      "step": 3340
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 0.00013700668896321068,
+      "loss": 0.8354,
+      "step": 3350
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 0.00013650501672240803,
+      "loss": 0.8296,
+      "step": 3360
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 0.00013600334448160534,
+      "loss": 0.8442,
+      "step": 3370
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 0.00013550167224080266,
+      "loss": 0.84,
+      "step": 3380
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 0.000135,
+      "loss": 0.8538,
+      "step": 3390
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 0.00013449832775919732,
+      "loss": 0.8356,
+      "step": 3400
+    },
+    {
+      "epoch": 5.59,
+      "eval_loss": 0.9222854971885681,
+      "eval_runtime": 33.1332,
+      "eval_samples_per_second": 60.362,
+      "eval_steps_per_second": 1.268,
+      "step": 3400
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 0.00013399665551839463,
+      "loss": 0.8417,
+      "step": 3410
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 0.00013349498327759198,
+      "loss": 0.8287,
+      "step": 3420
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 0.0001329933110367893,
+      "loss": 0.8429,
+      "step": 3430
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 0.0001324916387959866,
+      "loss": 0.8453,
+      "step": 3440
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 0.00013198996655518395,
+      "loss": 0.8436,
+      "step": 3450
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 0.00013148829431438127,
+      "loss": 0.8357,
+      "step": 3460
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 0.00013098662207357858,
+      "loss": 0.8334,
+      "step": 3470
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 0.00013048494983277593,
+      "loss": 0.8401,
+      "step": 3480
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 0.00012998327759197324,
+      "loss": 0.8513,
+      "step": 3490
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 0.00012948160535117056,
+      "loss": 0.8369,
+      "step": 3500
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 0.00012897993311036788,
+      "loss": 0.8437,
+      "step": 3510
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 0.00012847826086956522,
+      "loss": 0.8422,
+      "step": 3520
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 0.00012797658862876254,
+      "loss": 0.8437,
+      "step": 3530
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 0.00012747491638795985,
+      "loss": 0.8277,
+      "step": 3540
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 0.00012697324414715717,
+      "loss": 0.8317,
+      "step": 3550
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 0.00012647157190635448,
+      "loss": 0.8322,
+      "step": 3560
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 0.00012596989966555183,
+      "loss": 0.8221,
+      "step": 3570
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 0.00012546822742474914,
+      "loss": 0.8295,
+      "step": 3580
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 0.00012496655518394646,
+      "loss": 0.8411,
+      "step": 3590
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 0.0001244648829431438,
+      "loss": 0.8279,
+      "step": 3600
+    },
+    {
+      "epoch": 5.92,
+      "eval_loss": 0.9197421669960022,
+      "eval_runtime": 33.1031,
+      "eval_samples_per_second": 60.417,
+      "eval_steps_per_second": 1.269,
+      "step": 3600
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 0.00012396321070234112,
+      "loss": 0.8296,
+      "step": 3610
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 0.00012346153846153844,
+      "loss": 0.8278,
+      "step": 3620
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 0.00012295986622073578,
+      "loss": 0.8408,
+      "step": 3630
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 0.0001224581939799331,
+      "loss": 0.8232,
+      "step": 3640
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 0.00012195652173913042,
+      "loss": 0.826,
+      "step": 3650
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 0.00012145484949832774,
+      "loss": 0.811,
+      "step": 3660
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 0.00012095317725752507,
+      "loss": 0.85,
+      "step": 3670
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 0.0001204515050167224,
+      "loss": 0.8395,
+      "step": 3680
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 0.00011994983277591972,
+      "loss": 0.8179,
+      "step": 3690
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 0.00011944816053511705,
+      "loss": 0.8279,
+      "step": 3700
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 0.00011894648829431436,
+      "loss": 0.8227,
+      "step": 3710
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 0.00011844481605351169,
+      "loss": 0.8248,
+      "step": 3720
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 0.00011794314381270902,
+      "loss": 0.8332,
+      "step": 3730
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 0.00011744147157190634,
+      "loss": 0.8204,
+      "step": 3740
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 0.00011693979933110367,
+      "loss": 0.8336,
+      "step": 3750
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 0.000116438127090301,
+      "loss": 0.813,
+      "step": 3760
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 0.00011593645484949831,
+      "loss": 0.8211,
+      "step": 3770
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 0.00011543478260869564,
+      "loss": 0.8193,
+      "step": 3780
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 0.00011493311036789297,
+      "loss": 0.8425,
+      "step": 3790
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 0.00011443143812709029,
+      "loss": 0.8209,
+      "step": 3800
+    },
+    {
+      "epoch": 6.25,
+      "eval_loss": 0.9195341467857361,
+      "eval_runtime": 33.159,
+      "eval_samples_per_second": 60.315,
+      "eval_steps_per_second": 1.267,
+      "step": 3800
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 0.00011392976588628762,
+      "loss": 0.8276,
+      "step": 3810
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 0.00011342809364548495,
+      "loss": 0.845,
+      "step": 3820
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 0.00011292642140468226,
+      "loss": 0.83,
+      "step": 3830
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 0.00011242474916387959,
+      "loss": 0.8226,
+      "step": 3840
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 0.00011197324414715718,
+      "loss": 0.8329,
+      "step": 3850
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 0.0001114715719063545,
+      "loss": 0.8384,
+      "step": 3860
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 0.00011096989966555183,
+      "loss": 0.8186,
+      "step": 3870
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 0.00011046822742474914,
+      "loss": 0.8027,
+      "step": 3880
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 0.00010996655518394647,
+      "loss": 0.8214,
+      "step": 3890
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 0.0001094648829431438,
+      "loss": 0.8066,
+      "step": 3900
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 0.00010896321070234112,
+      "loss": 0.8303,
+      "step": 3910
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 0.00010846153846153845,
+      "loss": 0.8342,
+      "step": 3920
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 0.00010795986622073578,
+      "loss": 0.8304,
+      "step": 3930
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 0.0001074581939799331,
+      "loss": 0.8279,
+      "step": 3940
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 0.00010695652173913043,
+      "loss": 0.8104,
+      "step": 3950
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 0.00010645484949832776,
+      "loss": 0.8372,
+      "step": 3960
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 0.00010595317725752507,
+      "loss": 0.8239,
+      "step": 3970
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 0.0001054515050167224,
+      "loss": 0.8404,
+      "step": 3980
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 0.00010494983277591972,
+      "loss": 0.8206,
+      "step": 3990
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 0.00010444816053511705,
+      "loss": 0.8174,
+      "step": 4000
+    },
+    {
+      "epoch": 6.58,
+      "eval_loss": 0.9175052642822266,
+      "eval_runtime": 33.1942,
+      "eval_samples_per_second": 60.251,
+      "eval_steps_per_second": 1.265,
+      "step": 4000
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 0.00010394648829431438,
+      "loss": 0.8335,
+      "step": 4010
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 0.00010344481605351169,
+      "loss": 0.84,
+      "step": 4020
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 0.00010294314381270902,
+      "loss": 0.815,
+      "step": 4030
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 0.00010244147157190635,
+      "loss": 0.8112,
+      "step": 4040
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 0.00010193979933110367,
+      "loss": 0.8302,
+      "step": 4050
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 0.000101438127090301,
+      "loss": 0.8284,
+      "step": 4060
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 0.00010093645484949833,
+      "loss": 0.823,
+      "step": 4070
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 0.00010043478260869564,
+      "loss": 0.8192,
+      "step": 4080
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 9.993311036789297e-05,
+      "loss": 0.8285,
+      "step": 4090
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 9.94314381270903e-05,
+      "loss": 0.8222,
+      "step": 4100
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 9.892976588628762e-05,
+      "loss": 0.8273,
+      "step": 4110
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 9.842809364548495e-05,
+      "loss": 0.8145,
+      "step": 4120
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 9.792642140468226e-05,
+      "loss": 0.8279,
+      "step": 4130
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 9.74247491638796e-05,
+      "loss": 0.8193,
+      "step": 4140
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 9.692307692307692e-05,
+      "loss": 0.8243,
+      "step": 4150
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 9.642140468227424e-05,
+      "loss": 0.825,
+      "step": 4160
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 9.591973244147157e-05,
+      "loss": 0.8185,
+      "step": 4170
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 9.54180602006689e-05,
+      "loss": 0.8358,
+      "step": 4180
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 9.491638795986622e-05,
+      "loss": 0.8164,
+      "step": 4190
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 9.441471571906355e-05,
+      "loss": 0.833,
+      "step": 4200
+    },
+    {
+      "epoch": 6.91,
+      "eval_loss": 0.9155363440513611,
+      "eval_runtime": 33.1004,
+      "eval_samples_per_second": 60.422,
+      "eval_steps_per_second": 1.269,
+      "step": 4200
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 9.391304347826087e-05,
+      "loss": 0.8186,
+      "step": 4210
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 9.341137123745819e-05,
+      "loss": 0.8211,
+      "step": 4220
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 9.290969899665552e-05,
+      "loss": 0.8216,
+      "step": 4230
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 9.240802675585282e-05,
+      "loss": 0.7982,
+      "step": 4240
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 9.190635451505015e-05,
+      "loss": 0.8025,
+      "step": 4250
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 9.140468227424748e-05,
+      "loss": 0.814,
+      "step": 4260
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 9.09030100334448e-05,
+      "loss": 0.8048,
+      "step": 4270
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 9.040133779264213e-05,
+      "loss": 0.8279,
+      "step": 4280
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 8.989966555183944e-05,
+      "loss": 0.8208,
+      "step": 4290
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 8.939799331103677e-05,
+      "loss": 0.8009,
+      "step": 4300
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 8.88963210702341e-05,
+      "loss": 0.8151,
+      "step": 4310
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 8.839464882943142e-05,
+      "loss": 0.8066,
+      "step": 4320
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 8.789297658862875e-05,
+      "loss": 0.8239,
+      "step": 4330
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 8.739130434782608e-05,
+      "loss": 0.8055,
+      "step": 4340
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 8.68896321070234e-05,
+      "loss": 0.816,
+      "step": 4350
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 8.638795986622073e-05,
+      "loss": 0.815,
+      "step": 4360
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 8.588628762541805e-05,
+      "loss": 0.7989,
+      "step": 4370
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 8.538461538461537e-05,
+      "loss": 0.7924,
+      "step": 4380
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 8.48829431438127e-05,
+      "loss": 0.8151,
+      "step": 4390
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 8.438127090301002e-05,
+      "loss": 0.82,
+      "step": 4400
+    },
+    {
+      "epoch": 7.24,
+      "eval_loss": 0.9152849316596985,
+      "eval_runtime": 33.2483,
+      "eval_samples_per_second": 60.153,
+      "eval_steps_per_second": 1.263,
+      "step": 4400
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 8.387959866220735e-05,
+      "loss": 0.8091,
+      "step": 4410
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 8.337792642140468e-05,
+      "loss": 0.8195,
+      "step": 4420
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 8.287625418060199e-05,
+      "loss": 0.7982,
+      "step": 4430
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 8.237458193979932e-05,
+      "loss": 0.8106,
+      "step": 4440
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 8.187290969899665e-05,
+      "loss": 0.811,
+      "step": 4450
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 8.137123745819397e-05,
+      "loss": 0.8213,
+      "step": 4460
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 8.08695652173913e-05,
+      "loss": 0.8143,
+      "step": 4470
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 8.036789297658863e-05,
+      "loss": 0.8151,
+      "step": 4480
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 7.986622073578594e-05,
+      "loss": 0.8208,
+      "step": 4490
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 7.936454849498327e-05,
+      "loss": 0.8198,
+      "step": 4500
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 7.88628762541806e-05,
+      "loss": 0.8219,
+      "step": 4510
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 7.836120401337792e-05,
+      "loss": 0.7986,
+      "step": 4520
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 7.785953177257525e-05,
+      "loss": 0.8054,
+      "step": 4530
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 7.735785953177256e-05,
+      "loss": 0.8138,
+      "step": 4540
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 7.68561872909699e-05,
+      "loss": 0.8009,
+      "step": 4550
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 7.635451505016722e-05,
+      "loss": 0.8142,
+      "step": 4560
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 7.585284280936454e-05,
+      "loss": 0.8348,
+      "step": 4570
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 7.535117056856187e-05,
+      "loss": 0.8042,
+      "step": 4580
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 7.48494983277592e-05,
+      "loss": 0.819,
+      "step": 4590
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 7.434782608695652e-05,
+      "loss": 0.8027,
+      "step": 4600
+    },
+    {
+      "epoch": 7.57,
+      "eval_loss": 0.914962887763977,
+      "eval_runtime": 33.1103,
+      "eval_samples_per_second": 60.404,
+      "eval_steps_per_second": 1.268,
+      "step": 4600
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 7.384615384615384e-05,
+      "loss": 0.7976,
+      "step": 4610
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 7.334448160535116e-05,
+      "loss": 0.8168,
+      "step": 4620
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 7.284280936454849e-05,
+      "loss": 0.7952,
+      "step": 4630
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 7.234113712374581e-05,
+      "loss": 0.8238,
+      "step": 4640
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 7.183946488294314e-05,
+      "loss": 0.8196,
+      "step": 4650
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 7.133779264214047e-05,
+      "loss": 0.8032,
+      "step": 4660
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 7.083612040133778e-05,
+      "loss": 0.8067,
+      "step": 4670
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 7.033444816053511e-05,
+      "loss": 0.8167,
+      "step": 4680
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 6.983277591973243e-05,
+      "loss": 0.7856,
+      "step": 4690
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 6.933110367892976e-05,
+      "loss": 0.8234,
+      "step": 4700
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 6.882943143812709e-05,
+      "loss": 0.805,
+      "step": 4710
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 6.83277591973244e-05,
+      "loss": 0.8065,
+      "step": 4720
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 6.782608695652173e-05,
+      "loss": 0.8084,
+      "step": 4730
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 6.732441471571906e-05,
+      "loss": 0.8105,
+      "step": 4740
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 6.682274247491638e-05,
+      "loss": 0.8065,
+      "step": 4750
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 6.632107023411371e-05,
+      "loss": 0.8007,
+      "step": 4760
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 6.581939799331104e-05,
+      "loss": 0.8238,
+      "step": 4770
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 6.531772575250835e-05,
+      "loss": 0.8083,
+      "step": 4780
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 6.481605351170568e-05,
+      "loss": 0.81,
+      "step": 4790
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 6.4314381270903e-05,
+      "loss": 0.7963,
+      "step": 4800
+    },
+    {
+      "epoch": 7.89,
+      "eval_loss": 0.9136397838592529,
+      "eval_runtime": 33.1953,
+      "eval_samples_per_second": 60.249,
+      "eval_steps_per_second": 1.265,
+      "step": 4800
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 6.381270903010033e-05,
+      "loss": 0.7906,
+      "step": 4810
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 6.331103678929766e-05,
+      "loss": 0.8132,
+      "step": 4820
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 6.280936454849498e-05,
+      "loss": 0.8175,
+      "step": 4830
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 6.23076923076923e-05,
+      "loss": 0.801,
+      "step": 4840
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 6.180602006688964e-05,
+      "loss": 0.8117,
+      "step": 4850
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 6.130434782608695e-05,
+      "loss": 0.8019,
+      "step": 4860
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 6.0802675585284274e-05,
+      "loss": 0.802,
+      "step": 4870
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 6.03010033444816e-05,
+      "loss": 0.821,
+      "step": 4880
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 5.979933110367892e-05,
+      "loss": 0.7962,
+      "step": 4890
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 5.929765886287625e-05,
+      "loss": 0.7992,
+      "step": 4900
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 5.879598662207357e-05,
+      "loss": 0.7998,
+      "step": 4910
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 5.8294314381270895e-05,
+      "loss": 0.7934,
+      "step": 4920
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 5.779264214046822e-05,
+      "loss": 0.801,
+      "step": 4930
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 5.729096989966555e-05,
+      "loss": 0.8039,
+      "step": 4940
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 5.678929765886287e-05,
+      "loss": 0.802,
+      "step": 4950
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 5.6287625418060194e-05,
+      "loss": 0.781,
+      "step": 4960
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 5.5785953177257517e-05,
+      "loss": 0.7938,
+      "step": 4970
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 5.5284280936454846e-05,
+      "loss": 0.8031,
+      "step": 4980
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 5.478260869565217e-05,
+      "loss": 0.7999,
+      "step": 4990
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 5.428093645484949e-05,
+      "loss": 0.7985,
+      "step": 5000
+    },
+    {
+      "epoch": 8.22,
+      "eval_loss": 0.9126840829849243,
+      "eval_runtime": 33.2306,
+      "eval_samples_per_second": 60.185,
+      "eval_steps_per_second": 1.264,
+      "step": 5000
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 5.377926421404682e-05,
+      "loss": 0.8056,
+      "step": 5010
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 5.3277591973244145e-05,
+      "loss": 0.8099,
+      "step": 5020
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 5.277591973244147e-05,
+      "loss": 0.7982,
+      "step": 5030
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 5.227424749163879e-05,
+      "loss": 0.8021,
+      "step": 5040
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 5.177257525083612e-05,
+      "loss": 0.8044,
+      "step": 5050
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 5.127090301003344e-05,
+      "loss": 0.8099,
+      "step": 5060
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 5.0769230769230766e-05,
+      "loss": 0.818,
+      "step": 5070
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 5.0267558528428095e-05,
+      "loss": 0.796,
+      "step": 5080
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 4.976588628762542e-05,
+      "loss": 0.8102,
+      "step": 5090
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 4.926421404682274e-05,
+      "loss": 0.8092,
+      "step": 5100
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 4.8762541806020064e-05,
+      "loss": 0.8064,
+      "step": 5110
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 4.826086956521738e-05,
+      "loss": 0.8171,
+      "step": 5120
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 4.775919732441471e-05,
+      "loss": 0.8077,
+      "step": 5130
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 4.725752508361203e-05,
+      "loss": 0.8084,
+      "step": 5140
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 4.6755852842809356e-05,
+      "loss": 0.8081,
+      "step": 5150
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 4.6254180602006685e-05,
+      "loss": 0.7926,
+      "step": 5160
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 4.575250836120401e-05,
+      "loss": 0.7938,
+      "step": 5170
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 4.525083612040133e-05,
+      "loss": 0.8196,
+      "step": 5180
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 4.4749163879598654e-05,
+      "loss": 0.8059,
+      "step": 5190
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 4.4247491638795984e-05,
+      "loss": 0.7997,
+      "step": 5200
+    },
+    {
+      "epoch": 8.55,
+      "eval_loss": 0.9122925996780396,
+      "eval_runtime": 33.1494,
+      "eval_samples_per_second": 60.333,
+      "eval_steps_per_second": 1.267,
+      "step": 5200
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 4.374581939799331e-05,
+      "loss": 0.79,
+      "step": 5210
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 4.324414715719063e-05,
+      "loss": 0.8103,
+      "step": 5220
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 4.274247491638796e-05,
+      "loss": 0.7972,
+      "step": 5230
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 4.224080267558528e-05,
+      "loss": 0.8117,
+      "step": 5240
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 4.1739130434782605e-05,
+      "loss": 0.8058,
+      "step": 5250
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 4.123745819397993e-05,
+      "loss": 0.8043,
+      "step": 5260
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 4.073578595317726e-05,
+      "loss": 0.7986,
+      "step": 5270
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 4.023411371237458e-05,
+      "loss": 0.7967,
+      "step": 5280
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 3.9732441471571903e-05,
+      "loss": 0.8072,
+      "step": 5290
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 3.9230769230769226e-05,
+      "loss": 0.8061,
+      "step": 5300
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 3.8729096989966556e-05,
+      "loss": 0.8101,
+      "step": 5310
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 3.822742474916388e-05,
+      "loss": 0.7987,
+      "step": 5320
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 3.77257525083612e-05,
+      "loss": 0.7971,
+      "step": 5330
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 3.7224080267558525e-05,
+      "loss": 0.7826,
+      "step": 5340
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 3.672240802675585e-05,
+      "loss": 0.7956,
+      "step": 5350
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 3.622073578595318e-05,
+      "loss": 0.8029,
+      "step": 5360
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 3.57190635451505e-05,
+      "loss": 0.8041,
+      "step": 5370
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 3.521739130434782e-05,
+      "loss": 0.7844,
+      "step": 5380
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 3.4715719063545146e-05,
+      "loss": 0.7914,
+      "step": 5390
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 3.4214046822742476e-05,
+      "loss": 0.8197,
+      "step": 5400
+    },
+    {
+      "epoch": 8.88,
+      "eval_loss": 0.9114879369735718,
+      "eval_runtime": 33.1591,
+      "eval_samples_per_second": 60.315,
+      "eval_steps_per_second": 1.267,
+      "step": 5400
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 3.37123745819398e-05,
+      "loss": 0.8012,
+      "step": 5410
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 3.321070234113712e-05,
+      "loss": 0.8069,
+      "step": 5420
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 3.2709030100334444e-05,
+      "loss": 0.7984,
+      "step": 5430
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 3.220735785953177e-05,
+      "loss": 0.7877,
+      "step": 5440
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 3.170568561872909e-05,
+      "loss": 0.7964,
+      "step": 5450
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 3.120401337792642e-05,
+      "loss": 0.8064,
+      "step": 5460
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 3.070234113712374e-05,
+      "loss": 0.8088,
+      "step": 5470
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 3.020066889632107e-05,
+      "loss": 0.7983,
+      "step": 5480
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 2.9698996655518392e-05,
+      "loss": 0.7924,
+      "step": 5490
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 2.9197324414715718e-05,
+      "loss": 0.7847,
+      "step": 5500
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 2.869565217391304e-05,
+      "loss": 0.8061,
+      "step": 5510
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 2.8193979933110367e-05,
+      "loss": 0.8038,
+      "step": 5520
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 2.769230769230769e-05,
+      "loss": 0.7913,
+      "step": 5530
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 2.7190635451505016e-05,
+      "loss": 0.7976,
+      "step": 5540
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 2.6688963210702336e-05,
+      "loss": 0.7996,
+      "step": 5550
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 2.6187290969899662e-05,
+      "loss": 0.7963,
+      "step": 5560
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 2.5685618729096985e-05,
+      "loss": 0.7987,
+      "step": 5570
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 2.518394648829431e-05,
+      "loss": 0.7858,
+      "step": 5580
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 2.4682274247491638e-05,
+      "loss": 0.8028,
+      "step": 5590
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 2.418060200668896e-05,
+      "loss": 0.7899,
+      "step": 5600
+    },
+    {
+      "epoch": 9.21,
+      "eval_loss": 0.9114179611206055,
+      "eval_runtime": 33.1597,
+      "eval_samples_per_second": 60.314,
+      "eval_steps_per_second": 1.267,
+      "step": 5600
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 2.3678929765886287e-05,
+      "loss": 0.7896,
+      "step": 5610
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 2.317725752508361e-05,
+      "loss": 0.7928,
+      "step": 5620
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 2.2675585284280936e-05,
+      "loss": 0.8108,
+      "step": 5630
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 2.217391304347826e-05,
+      "loss": 0.8011,
+      "step": 5640
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 2.1672240802675585e-05,
+      "loss": 0.7965,
+      "step": 5650
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 2.1170568561872908e-05,
+      "loss": 0.7967,
+      "step": 5660
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 2.0668896321070234e-05,
+      "loss": 0.8006,
+      "step": 5670
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 2.0167224080267554e-05,
+      "loss": 0.7938,
+      "step": 5680
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 1.966555183946488e-05,
+      "loss": 0.8137,
+      "step": 5690
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 1.9163879598662203e-05,
+      "loss": 0.7994,
+      "step": 5700
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 1.866220735785953e-05,
+      "loss": 0.7994,
+      "step": 5710
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 1.8160535117056856e-05,
+      "loss": 0.7926,
+      "step": 5720
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 1.765886287625418e-05,
+      "loss": 0.7918,
+      "step": 5730
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 1.7157190635451505e-05,
+      "loss": 0.7811,
+      "step": 5740
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 1.6655518394648828e-05,
+      "loss": 0.8001,
+      "step": 5750
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 0.8106,
+      "step": 5760
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 1.5652173913043477e-05,
+      "loss": 0.7882,
+      "step": 5770
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 1.5150501672240801e-05,
+      "loss": 0.7873,
+      "step": 5780
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 1.4648829431438126e-05,
+      "loss": 0.7976,
+      "step": 5790
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 1.414715719063545e-05,
+      "loss": 0.7951,
+      "step": 5800
+    },
+    {
+      "epoch": 9.54,
+      "eval_loss": 0.9110632538795471,
+      "eval_runtime": 33.1804,
+      "eval_samples_per_second": 60.276,
+      "eval_steps_per_second": 1.266,
+      "step": 5800
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 1.3645484949832775e-05,
+      "loss": 0.7897,
+      "step": 5810
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 1.31438127090301e-05,
+      "loss": 0.8037,
+      "step": 5820
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 1.2642140468227423e-05,
+      "loss": 0.786,
+      "step": 5830
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 1.2140468227424747e-05,
+      "loss": 0.8044,
+      "step": 5840
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 1.1638795986622072e-05,
+      "loss": 0.804,
+      "step": 5850
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 1.1137123745819396e-05,
+      "loss": 0.8,
+      "step": 5860
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 1.0635451505016721e-05,
+      "loss": 0.7957,
+      "step": 5870
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 1.0133779264214047e-05,
+      "loss": 0.7875,
+      "step": 5880
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 9.632107023411369e-06,
+      "loss": 0.7969,
+      "step": 5890
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 9.130434782608695e-06,
+      "loss": 0.805,
+      "step": 5900
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 8.62876254180602e-06,
+      "loss": 0.7755,
+      "step": 5910
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 8.127090301003344e-06,
+      "loss": 0.7957,
+      "step": 5920
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 7.625418060200669e-06,
+      "loss": 0.8089,
+      "step": 5930
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 7.123745819397992e-06,
+      "loss": 0.8008,
+      "step": 5940
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 6.622073578595317e-06,
+      "loss": 0.7941,
+      "step": 5950
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 6.1204013377926415e-06,
+      "loss": 0.7932,
+      "step": 5960
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 5.618729096989965e-06,
+      "loss": 0.7993,
+      "step": 5970
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 5.117056856187291e-06,
+      "loss": 0.7872,
+      "step": 5980
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 4.615384615384615e-06,
+      "loss": 0.8088,
+      "step": 5990
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 4.11371237458194e-06,
+      "loss": 0.7947,
+      "step": 6000
+    },
+    {
+      "epoch": 9.87,
+      "eval_loss": 0.9106292724609375,
+      "eval_runtime": 33.2219,
+      "eval_samples_per_second": 60.201,
+      "eval_steps_per_second": 1.264,
+      "step": 6000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 6080,
+  "num_train_epochs": 10,
+  "save_steps": 200,
+  "total_flos": 1.3874338630022988e+19,
+  "trial_name": null,
+  "trial_params": null
+}

models/lora_weights/chreckpoint-6000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2683d2cc3b2cf2ed5efb5a43c93d20249d9a309d91a7775d0bbfbbf7ac3e5a62
+size 4027