diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,3750 @@
+{
+  "best_metric": 0.6239999995057423,
+  "best_model_checkpoint": "/shared/data/jatin2/sec-ner/multirun/wnut/ner-seq/run-42/checkpoints/checkpoint-6855",
+  "epoch": 253.88888888888889,
+  "global_step": 6855,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.56,
+      "eval_loss": 0.6107152104377747,
+      "eval_micro_f1": 0.0,
+      "eval_runtime": 2.717,
+      "eval_samples_per_second": 371.359,
+      "step": 15
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 0.5758534669876099,
+      "eval_micro_f1": 0.0,
+      "eval_runtime": 2.7914,
+      "eval_samples_per_second": 361.463,
+      "step": 30
+    },
+    {
+      "epoch": 1.67,
+      "eval_loss": 0.5382413864135742,
+      "eval_micro_f1": 0.0,
+      "eval_runtime": 2.6918,
+      "eval_samples_per_second": 374.835,
+      "step": 45
+    },
+    {
+      "epoch": 2.22,
+      "eval_loss": 0.47368818521499634,
+      "eval_micro_f1": 0.0,
+      "eval_runtime": 2.683,
+      "eval_samples_per_second": 376.071,
+      "step": 60
+    },
+    {
+      "epoch": 2.78,
+      "eval_loss": 0.41591086983680725,
+      "eval_micro_f1": 0.0,
+      "eval_runtime": 2.576,
+      "eval_samples_per_second": 391.688,
+      "step": 75
+    },
+    {
+      "epoch": 3.33,
+      "eval_loss": 0.3757789433002472,
+      "eval_micro_f1": 0.05340699780386233,
+      "eval_runtime": 2.7733,
+      "eval_samples_per_second": 363.824,
+      "step": 90
+    },
+    {
+      "epoch": 3.89,
+      "eval_loss": 0.37088578939437866,
+      "eval_micro_f1": 0.12999999957713887,
+      "eval_runtime": 2.6835,
+      "eval_samples_per_second": 376.002,
+      "step": 105
+    },
+    {
+      "epoch": 4.44,
+      "eval_loss": 0.3661174178123474,
+      "eval_micro_f1": 0.23357664185972188,
+      "eval_runtime": 2.6872,
+      "eval_samples_per_second": 375.486,
+      "step": 120
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 0.346435010433197,
+      "eval_micro_f1": 0.2872777012883089,
+      "eval_runtime": 2.6707,
+      "eval_samples_per_second": 377.8,
+      "step": 135
+    },
+    {
+      "epoch": 5.56,
+      "eval_loss": 0.3376510739326477,
+      "eval_micro_f1": 0.32138114160392167,
+      "eval_runtime": 2.7366,
+      "eval_samples_per_second": 368.7,
+      "step": 150
+    },
+    {
+      "epoch": 6.11,
+      "eval_loss": 0.3373923599720001,
+      "eval_micro_f1": 0.33155792227560416,
+      "eval_runtime": 2.7902,
+      "eval_samples_per_second": 361.621,
+      "step": 165
+    },
+    {
+      "epoch": 6.67,
+      "eval_loss": 0.3485959768295288,
+      "eval_micro_f1": 0.37016574536780783,
+      "eval_runtime": 2.8453,
+      "eval_samples_per_second": 354.622,
+      "step": 180
+    },
+    {
+      "epoch": 7.22,
+      "eval_loss": 0.3558724820613861,
+      "eval_micro_f1": 0.3803101815709932,
+      "eval_runtime": 3.229,
+      "eval_samples_per_second": 312.48,
+      "step": 195
+    },
+    {
+      "epoch": 7.78,
+      "eval_loss": 0.34389063715934753,
+      "eval_micro_f1": 0.4108843532503791,
+      "eval_runtime": 2.7049,
+      "eval_samples_per_second": 373.024,
+      "step": 210
+    },
+    {
+      "epoch": 8.33,
+      "eval_loss": 0.3563588261604309,
+      "eval_micro_f1": 0.4243261916331337,
+      "eval_runtime": 2.7986,
+      "eval_samples_per_second": 360.542,
+      "step": 225
+    },
+    {
+      "epoch": 8.89,
+      "eval_loss": 0.3286007344722748,
+      "eval_micro_f1": 0.4667571229819302,
+      "eval_runtime": 2.7294,
+      "eval_samples_per_second": 369.673,
+      "step": 240
+    },
+    {
+      "epoch": 9.44,
+      "eval_loss": 0.33356600999832153,
+      "eval_micro_f1": 0.4773662546541498,
+      "eval_runtime": 2.7428,
+      "eval_samples_per_second": 367.879,
+      "step": 255
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 0.32206007838249207,
+      "eval_micro_f1": 0.49345549688599377,
+      "eval_runtime": 3.0813,
+      "eval_samples_per_second": 327.456,
+      "step": 270
+    },
+    {
+      "epoch": 10.56,
+      "eval_loss": 0.33766070008277893,
+      "eval_micro_f1": 0.4926174491712283,
+      "eval_runtime": 2.9682,
+      "eval_samples_per_second": 339.935,
+      "step": 285
+    },
+    {
+      "epoch": 11.11,
+      "eval_loss": 0.33275020122528076,
+      "eval_micro_f1": 0.504383006924838,
+      "eval_runtime": 2.948,
+      "eval_samples_per_second": 342.264,
+      "step": 300
+    },
+    {
+      "epoch": 11.67,
+      "eval_loss": 0.32769575715065,
+      "eval_micro_f1": 0.5029392548922187,
+      "eval_runtime": 2.8324,
+      "eval_samples_per_second": 356.234,
+      "step": 315
+    },
+    {
+      "epoch": 12.22,
+      "eval_loss": 0.3339318037033081,
+      "eval_micro_f1": 0.5143620569543754,
+      "eval_runtime": 3.0089,
+      "eval_samples_per_second": 335.34,
+      "step": 330
+    },
+    {
+      "epoch": 12.78,
+      "eval_loss": 0.3339863121509552,
+      "eval_micro_f1": 0.5246548318513781,
+      "eval_runtime": 2.9602,
+      "eval_samples_per_second": 340.851,
+      "step": 345
+    },
+    {
+      "epoch": 13.33,
+      "eval_loss": 0.34505823254585266,
+      "eval_micro_f1": 0.5254691684073772,
+      "eval_runtime": 2.9263,
+      "eval_samples_per_second": 344.806,
+      "step": 360
+    },
+    {
+      "epoch": 13.89,
+      "eval_loss": 0.35517218708992004,
+      "eval_micro_f1": 0.532970767693519,
+      "eval_runtime": 2.9694,
+      "eval_samples_per_second": 339.798,
+      "step": 375
+    },
+    {
+      "epoch": 14.44,
+      "eval_loss": 0.35111522674560547,
+      "eval_micro_f1": 0.5282258059585483,
+      "eval_runtime": 2.947,
+      "eval_samples_per_second": 342.386,
+      "step": 390
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 0.33189719915390015,
+      "eval_micro_f1": 0.5354330703701545,
+      "eval_runtime": 2.8801,
+      "eval_samples_per_second": 350.339,
+      "step": 405
+    },
+    {
+      "epoch": 15.56,
+      "eval_loss": 0.3263826072216034,
+      "eval_micro_f1": 0.5364583328365545,
+      "eval_runtime": 2.9203,
+      "eval_samples_per_second": 345.512,
+      "step": 420
+    },
+    {
+      "epoch": 16.11,
+      "eval_loss": 0.3304557800292969,
+      "eval_micro_f1": 0.5435897430916239,
+      "eval_runtime": 2.8072,
+      "eval_samples_per_second": 359.437,
+      "step": 435
+    },
+    {
+      "epoch": 16.67,
+      "eval_loss": 0.33612725138664246,
+      "eval_micro_f1": 0.5604249663047992,
+      "eval_runtime": 2.7339,
+      "eval_samples_per_second": 369.076,
+      "step": 450
+    },
+    {
+      "epoch": 17.22,
+      "eval_loss": 0.3405936658382416,
+      "eval_micro_f1": 0.5579227691461416,
+      "eval_runtime": 3.0589,
+      "eval_samples_per_second": 329.853,
+      "step": 465
+    },
+    {
+      "epoch": 17.78,
+      "eval_loss": 0.3675728440284729,
+      "eval_micro_f1": 0.5583277136004227,
+      "eval_runtime": 2.8145,
+      "eval_samples_per_second": 358.503,
+      "step": 480
+    },
+    {
+      "epoch": 18.33,
+      "eval_loss": 0.34318840503692627,
+      "eval_micro_f1": 0.5621693116741674,
+      "eval_runtime": 2.8624,
+      "eval_samples_per_second": 352.498,
+      "step": 495
+    },
+    {
+      "epoch": 18.52,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 0.1305,
+      "step": 500
+    },
+    {
+      "epoch": 18.89,
+      "eval_loss": 0.3430013656616211,
+      "eval_micro_f1": 0.5587467357958742,
+      "eval_runtime": 2.7585,
+      "eval_samples_per_second": 365.778,
+      "step": 510
+    },
+    {
+      "epoch": 19.44,
+      "eval_loss": 0.35118457674980164,
+      "eval_micro_f1": 0.5649867369056948,
+      "eval_runtime": 2.6475,
+      "eval_samples_per_second": 381.11,
+      "step": 525
+    },
+    {
+      "epoch": 20.0,
+      "eval_loss": 0.3561754524707794,
+      "eval_micro_f1": 0.5742705565343297,
+      "eval_runtime": 2.7367,
+      "eval_samples_per_second": 368.692,
+      "step": 540
+    },
+    {
+      "epoch": 20.56,
+      "eval_loss": 0.3560926616191864,
+      "eval_micro_f1": 0.5754339113884877,
+      "eval_runtime": 2.8003,
+      "eval_samples_per_second": 360.319,
+      "step": 555
+    },
+    {
+      "epoch": 21.11,
+      "eval_loss": 0.33654725551605225,
+      "eval_micro_f1": 0.5714285709294672,
+      "eval_runtime": 2.8885,
+      "eval_samples_per_second": 349.322,
+      "step": 570
+    },
+    {
+      "epoch": 21.67,
+      "eval_loss": 0.356985867023468,
+      "eval_micro_f1": 0.5862758156282931,
+      "eval_runtime": 3.3601,
+      "eval_samples_per_second": 300.284,
+      "step": 585
+    },
+    {
+      "epoch": 22.22,
+      "eval_loss": 0.35574978590011597,
+      "eval_micro_f1": 0.5898437495031517,
+      "eval_runtime": 3.0683,
+      "eval_samples_per_second": 328.85,
+      "step": 600
+    },
+    {
+      "epoch": 22.78,
+      "eval_loss": 0.3793315589427948,
+      "eval_micro_f1": 0.5806887233427784,
+      "eval_runtime": 2.5646,
+      "eval_samples_per_second": 393.43,
+      "step": 615
+    },
+    {
+      "epoch": 23.33,
+      "eval_loss": 0.3607286214828491,
+      "eval_micro_f1": 0.5815789468726559,
+      "eval_runtime": 2.6735,
+      "eval_samples_per_second": 377.407,
+      "step": 630
+    },
+    {
+      "epoch": 23.89,
+      "eval_loss": 0.3553391695022583,
+      "eval_micro_f1": 0.5871559628065596,
+      "eval_runtime": 2.7444,
+      "eval_samples_per_second": 367.655,
+      "step": 645
+    },
+    {
+      "epoch": 24.44,
+      "eval_loss": 0.3693152666091919,
+      "eval_micro_f1": 0.5723641121429205,
+      "eval_runtime": 2.5481,
+      "eval_samples_per_second": 395.981,
+      "step": 660
+    },
+    {
+      "epoch": 25.0,
+      "eval_loss": 0.37512636184692383,
+      "eval_micro_f1": 0.5902731507382145,
+      "eval_runtime": 2.8652,
+      "eval_samples_per_second": 352.158,
+      "step": 675
+    },
+    {
+      "epoch": 25.56,
+      "eval_loss": 0.3612546920776367,
+      "eval_micro_f1": 0.5912031042891023,
+      "eval_runtime": 2.7393,
+      "eval_samples_per_second": 368.343,
+      "step": 690
+    },
+    {
+      "epoch": 26.11,
+      "eval_loss": 0.36965033411979675,
+      "eval_micro_f1": 0.5904887709709797,
+      "eval_runtime": 2.9837,
+      "eval_samples_per_second": 338.171,
+      "step": 705
+    },
+    {
+      "epoch": 26.67,
+      "eval_loss": 0.3751550316810608,
+      "eval_micro_f1": 0.5959129856613228,
+      "eval_runtime": 2.7434,
+      "eval_samples_per_second": 367.792,
+      "step": 720
+    },
+    {
+      "epoch": 27.22,
+      "eval_loss": 0.3682175874710083,
+      "eval_micro_f1": 0.5988409524963693,
+      "eval_runtime": 2.9516,
+      "eval_samples_per_second": 341.849,
+      "step": 735
+    },
+    {
+      "epoch": 27.78,
+      "eval_loss": 0.37221601605415344,
+      "eval_micro_f1": 0.5996055221865852,
+      "eval_runtime": 3.24,
+      "eval_samples_per_second": 311.423,
+      "step": 750
+    },
+    {
+      "epoch": 28.33,
+      "eval_loss": 0.37376993894577026,
+      "eval_micro_f1": 0.5964450291290404,
+      "eval_runtime": 2.8752,
+      "eval_samples_per_second": 350.932,
+      "step": 765
+    },
+    {
+      "epoch": 28.89,
+      "eval_loss": 0.39057737588882446,
+      "eval_micro_f1": 0.6032171576834126,
+      "eval_runtime": 2.663,
+      "eval_samples_per_second": 378.891,
+      "step": 780
+    },
+    {
+      "epoch": 29.44,
+      "eval_loss": 0.37028804421424866,
+      "eval_micro_f1": 0.5925449866485467,
+      "eval_runtime": 2.8947,
+      "eval_samples_per_second": 348.573,
+      "step": 795
+    },
+    {
+      "epoch": 30.0,
+      "eval_loss": 0.3838958144187927,
+      "eval_micro_f1": 0.5931307788970002,
+      "eval_runtime": 2.6937,
+      "eval_samples_per_second": 374.58,
+      "step": 810
+    },
+    {
+      "epoch": 30.56,
+      "eval_loss": 0.38781028985977173,
+      "eval_micro_f1": 0.6078301255834884,
+      "eval_runtime": 3.1486,
+      "eval_samples_per_second": 320.462,
+      "step": 825
+    },
+    {
+      "epoch": 31.11,
+      "eval_loss": 0.38530829548835754,
+      "eval_micro_f1": 0.6028833546807199,
+      "eval_runtime": 3.0719,
+      "eval_samples_per_second": 328.458,
+      "step": 840
+    },
+    {
+      "epoch": 31.67,
+      "eval_loss": 0.3929048478603363,
+      "eval_micro_f1": 0.6027944106831641,
+      "eval_runtime": 2.8053,
+      "eval_samples_per_second": 359.672,
+      "step": 855
+    },
+    {
+      "epoch": 32.22,
+      "eval_loss": 0.388367235660553,
+      "eval_micro_f1": 0.5975049239951369,
+      "eval_runtime": 2.8501,
+      "eval_samples_per_second": 354.026,
+      "step": 870
+    },
+    {
+      "epoch": 32.78,
+      "eval_loss": 0.3855569660663605,
+      "eval_micro_f1": 0.5972850673757953,
+      "eval_runtime": 2.8904,
+      "eval_samples_per_second": 349.092,
+      "step": 885
+    },
+    {
+      "epoch": 33.33,
+      "eval_loss": 0.3963012099266052,
+      "eval_micro_f1": 0.5907284763261647,
+      "eval_runtime": 3.0151,
+      "eval_samples_per_second": 334.644,
+      "step": 900
+    },
+    {
+      "epoch": 33.89,
+      "eval_loss": 0.3843013644218445,
+      "eval_micro_f1": 0.5935483865991059,
+      "eval_runtime": 2.7799,
+      "eval_samples_per_second": 362.957,
+      "step": 915
+    },
+    {
+      "epoch": 34.44,
+      "eval_loss": 0.37854063510894775,
+      "eval_micro_f1": 0.6034593204498652,
+      "eval_runtime": 2.7274,
+      "eval_samples_per_second": 369.946,
+      "step": 930
+    },
+    {
+      "epoch": 35.0,
+      "eval_loss": 0.39461299777030945,
+      "eval_micro_f1": 0.6010568026750609,
+      "eval_runtime": 2.9473,
+      "eval_samples_per_second": 342.353,
+      "step": 945
+    },
+    {
+      "epoch": 35.56,
+      "eval_loss": 0.3603731095790863,
+      "eval_micro_f1": 0.586633662866211,
+      "eval_runtime": 2.7317,
+      "eval_samples_per_second": 369.373,
+      "step": 960
+    },
+    {
+      "epoch": 36.11,
+      "eval_loss": 0.3944607675075531,
+      "eval_micro_f1": 0.609271522683756,
+      "eval_runtime": 2.7367,
+      "eval_samples_per_second": 368.694,
+      "step": 975
+    },
+    {
+      "epoch": 36.67,
+      "eval_loss": 0.3918357193470001,
+      "eval_micro_f1": 0.6022353709702767,
+      "eval_runtime": 2.7729,
+      "eval_samples_per_second": 363.878,
+      "step": 990
+    },
+    {
+      "epoch": 37.04,
+      "learning_rate": 8.765432098765432e-06,
+      "loss": 0.0121,
+      "step": 1000
+    },
+    {
+      "epoch": 37.22,
+      "eval_loss": 0.4166373610496521,
+      "eval_micro_f1": 0.6089872563811323,
+      "eval_runtime": 2.7289,
+      "eval_samples_per_second": 369.75,
+      "step": 1005
+    },
+    {
+      "epoch": 37.78,
+      "eval_loss": 0.4003227651119232,
+      "eval_micro_f1": 0.6101249173213512,
+      "eval_runtime": 2.7811,
+      "eval_samples_per_second": 362.807,
+      "step": 1020
+    },
+    {
+      "epoch": 38.33,
+      "eval_loss": 0.4043891727924347,
+      "eval_micro_f1": 0.6104152928465378,
+      "eval_runtime": 2.7081,
+      "eval_samples_per_second": 372.592,
+      "step": 1035
+    },
+    {
+      "epoch": 38.89,
+      "eval_loss": 0.3939799964427948,
+      "eval_micro_f1": 0.6044386418010366,
+      "eval_runtime": 2.837,
+      "eval_samples_per_second": 355.657,
+      "step": 1050
+    },
+    {
+      "epoch": 39.44,
+      "eval_loss": 0.42168179154396057,
+      "eval_micro_f1": 0.6025554803408225,
+      "eval_runtime": 2.878,
+      "eval_samples_per_second": 350.586,
+      "step": 1065
+    },
+    {
+      "epoch": 40.0,
+      "eval_loss": 0.4030210077762604,
+      "eval_micro_f1": 0.5996106419428092,
+      "eval_runtime": 2.809,
+      "eval_samples_per_second": 359.203,
+      "step": 1080
+    },
+    {
+      "epoch": 40.56,
+      "eval_loss": 0.42451536655426025,
+      "eval_micro_f1": 0.597175520690527,
+      "eval_runtime": 2.8788,
+      "eval_samples_per_second": 350.493,
+      "step": 1095
+    },
+    {
+      "epoch": 41.11,
+      "eval_loss": 0.39370641112327576,
+      "eval_micro_f1": 0.599744245025914,
+      "eval_runtime": 2.9772,
+      "eval_samples_per_second": 338.907,
+      "step": 1110
+    },
+    {
+      "epoch": 41.67,
+      "eval_loss": 0.4258524477481842,
+      "eval_micro_f1": 0.6032171576834126,
+      "eval_runtime": 2.9461,
+      "eval_samples_per_second": 342.492,
+      "step": 1125
+    },
+    {
+      "epoch": 42.22,
+      "eval_loss": 0.4032389223575592,
+      "eval_micro_f1": 0.6085825742754358,
+      "eval_runtime": 2.8051,
+      "eval_samples_per_second": 359.708,
+      "step": 1140
+    },
+    {
+      "epoch": 42.78,
+      "eval_loss": 0.41863372921943665,
+      "eval_micro_f1": 0.6116248343791421,
+      "eval_runtime": 2.899,
+      "eval_samples_per_second": 348.047,
+      "step": 1155
+    },
+    {
+      "epoch": 43.33,
+      "eval_loss": 0.39685216546058655,
+      "eval_micro_f1": 0.603846153347957,
+      "eval_runtime": 2.7322,
+      "eval_samples_per_second": 369.3,
+      "step": 1170
+    },
+    {
+      "epoch": 43.89,
+      "eval_loss": 0.3990769386291504,
+      "eval_micro_f1": 0.5938921372547491,
+      "eval_runtime": 2.7574,
+      "eval_samples_per_second": 365.922,
+      "step": 1185
+    },
+    {
+      "epoch": 44.44,
+      "eval_loss": 0.39750733971595764,
+      "eval_micro_f1": 0.6001287825028067,
+      "eval_runtime": 2.7581,
+      "eval_samples_per_second": 365.828,
+      "step": 1200
+    },
+    {
+      "epoch": 45.0,
+      "eval_loss": 0.41059523820877075,
+      "eval_micro_f1": 0.6062992121023449,
+      "eval_runtime": 3.0236,
+      "eval_samples_per_second": 333.713,
+      "step": 1215
+    },
+    {
+      "epoch": 45.56,
+      "eval_loss": 0.428274005651474,
+      "eval_micro_f1": 0.6052104203477077,
+      "eval_runtime": 2.8825,
+      "eval_samples_per_second": 350.041,
+      "step": 1230
+    },
+    {
+      "epoch": 46.11,
+      "eval_loss": 0.39209499955177307,
+      "eval_micro_f1": 0.5911949680540454,
+      "eval_runtime": 2.889,
+      "eval_samples_per_second": 349.261,
+      "step": 1245
+    },
+    {
+      "epoch": 46.67,
+      "eval_loss": 0.4244146943092346,
+      "eval_micro_f1": 0.6013422813858476,
+      "eval_runtime": 2.8974,
+      "eval_samples_per_second": 348.243,
+      "step": 1260
+    },
+    {
+      "epoch": 47.22,
+      "eval_loss": 0.4031459093093872,
+      "eval_micro_f1": 0.6010430242751011,
+      "eval_runtime": 2.6784,
+      "eval_samples_per_second": 376.722,
+      "step": 1275
+    },
+    {
+      "epoch": 47.78,
+      "eval_loss": 0.4066299498081207,
+      "eval_micro_f1": 0.6081258186387732,
+      "eval_runtime": 2.7745,
+      "eval_samples_per_second": 363.667,
+      "step": 1290
+    },
+    {
+      "epoch": 48.33,
+      "eval_loss": 0.4201914072036743,
+      "eval_micro_f1": 0.5976095612582693,
+      "eval_runtime": 2.7634,
+      "eval_samples_per_second": 365.126,
+      "step": 1305
+    },
+    {
+      "epoch": 48.89,
+      "eval_loss": 0.430967777967453,
+      "eval_micro_f1": 0.6076458747580622,
+      "eval_runtime": 2.9174,
+      "eval_samples_per_second": 345.859,
+      "step": 1320
+    },
+    {
+      "epoch": 49.44,
+      "eval_loss": 0.4001426100730896,
+      "eval_micro_f1": 0.6021780904690853,
+      "eval_runtime": 2.9498,
+      "eval_samples_per_second": 342.058,
+      "step": 1335
+    },
+    {
+      "epoch": 50.0,
+      "eval_loss": 0.4081067442893982,
+      "eval_micro_f1": 0.6018396841295583,
+      "eval_runtime": 2.8863,
+      "eval_samples_per_second": 349.588,
+      "step": 1350
+    },
+    {
+      "epoch": 50.56,
+      "eval_loss": 0.4145914912223816,
+      "eval_micro_f1": 0.6063408185271014,
+      "eval_runtime": 2.6576,
+      "eval_samples_per_second": 379.668,
+      "step": 1365
+    },
+    {
+      "epoch": 51.11,
+      "eval_loss": 0.4008004665374756,
+      "eval_micro_f1": 0.6031746026757336,
+      "eval_runtime": 3.4866,
+      "eval_samples_per_second": 289.394,
+      "step": 1380
+    },
+    {
+      "epoch": 51.67,
+      "eval_loss": 0.4332279562950134,
+      "eval_micro_f1": 0.5998655005156747,
+      "eval_runtime": 3.0258,
+      "eval_samples_per_second": 333.471,
+      "step": 1395
+    },
+    {
+      "epoch": 52.22,
+      "eval_loss": 0.415213942527771,
+      "eval_micro_f1": 0.5937090427541264,
+      "eval_runtime": 3.0336,
+      "eval_samples_per_second": 332.609,
+      "step": 1410
+    },
+    {
+      "epoch": 52.78,
+      "eval_loss": 0.42885804176330566,
+      "eval_micro_f1": 0.5978695068291776,
+      "eval_runtime": 2.9679,
+      "eval_samples_per_second": 339.97,
+      "step": 1425
+    },
+    {
+      "epoch": 53.33,
+      "eval_loss": 0.40946725010871887,
+      "eval_micro_f1": 0.6024405903818509,
+      "eval_runtime": 2.6405,
+      "eval_samples_per_second": 382.12,
+      "step": 1440
+    },
+    {
+      "epoch": 53.89,
+      "eval_loss": 0.42097437381744385,
+      "eval_micro_f1": 0.6076618224901117,
+      "eval_runtime": 2.7372,
+      "eval_samples_per_second": 368.631,
+      "step": 1455
+    },
+    {
+      "epoch": 54.44,
+      "eval_loss": 0.40631037950515747,
+      "eval_micro_f1": 0.6012903220829668,
+      "eval_runtime": 5.6745,
+      "eval_samples_per_second": 177.812,
+      "step": 1470
+    },
+    {
+      "epoch": 55.0,
+      "eval_loss": 0.41115182638168335,
+      "eval_micro_f1": 0.6002604161698049,
+      "eval_runtime": 2.8323,
+      "eval_samples_per_second": 356.252,
+      "step": 1485
+    },
+    {
+      "epoch": 55.56,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.0047,
+      "step": 1500
+    },
+    {
+      "epoch": 55.56,
+      "eval_loss": 0.42716607451438904,
+      "eval_micro_f1": 0.5974717227257528,
+      "eval_runtime": 2.9218,
+      "eval_samples_per_second": 345.331,
+      "step": 1500
+    },
+    {
+      "epoch": 56.11,
+      "eval_loss": 0.40432748198509216,
+      "eval_micro_f1": 0.595527156051097,
+      "eval_runtime": 2.7818,
+      "eval_samples_per_second": 362.719,
+      "step": 1515
+    },
+    {
+      "epoch": 56.67,
+      "eval_loss": 0.41262897849082947,
+      "eval_micro_f1": 0.6005256236827796,
+      "eval_runtime": 2.914,
+      "eval_samples_per_second": 346.26,
+      "step": 1530
+    },
+    {
+      "epoch": 57.22,
+      "eval_loss": 0.4249415695667267,
+      "eval_micro_f1": 0.6082337312449748,
+      "eval_runtime": 3.0242,
+      "eval_samples_per_second": 333.642,
+      "step": 1545
+    },
+    {
+      "epoch": 57.78,
+      "eval_loss": 0.42977356910705566,
+      "eval_micro_f1": 0.5931307788970002,
+      "eval_runtime": 2.7928,
+      "eval_samples_per_second": 361.283,
+      "step": 1560
+    },
+    {
+      "epoch": 58.33,
+      "eval_loss": 0.4262886643409729,
+      "eval_micro_f1": 0.5954397389168877,
+      "eval_runtime": 2.6782,
+      "eval_samples_per_second": 376.743,
+      "step": 1575
+    },
+    {
+      "epoch": 58.89,
+      "eval_loss": 0.4387041926383972,
+      "eval_micro_f1": 0.5960264895711908,
+      "eval_runtime": 2.7213,
+      "eval_samples_per_second": 370.776,
+      "step": 1590
+    },
+    {
+      "epoch": 59.44,
+      "eval_loss": 0.4377950131893158,
+      "eval_micro_f1": 0.593522801884116,
+      "eval_runtime": 3.3156,
+      "eval_samples_per_second": 304.322,
+      "step": 1605
+    },
+    {
+      "epoch": 60.0,
+      "eval_loss": 0.4203121066093445,
+      "eval_micro_f1": 0.5881595876617141,
+      "eval_runtime": 3.0481,
+      "eval_samples_per_second": 331.026,
+      "step": 1620
+    },
+    {
+      "epoch": 60.56,
+      "eval_loss": 0.4478697180747986,
+      "eval_micro_f1": 0.5954301070337267,
+      "eval_runtime": 2.9076,
+      "eval_samples_per_second": 347.02,
+      "step": 1635
+    },
+    {
+      "epoch": 61.11,
+      "eval_loss": 0.4146019518375397,
+      "eval_micro_f1": 0.5924984101814561,
+      "eval_runtime": 2.6629,
+      "eval_samples_per_second": 378.905,
+      "step": 1650
+    },
+    {
+      "epoch": 61.67,
+      "eval_loss": 0.43386971950531006,
+      "eval_micro_f1": 0.5980327863891075,
+      "eval_runtime": 2.821,
+      "eval_samples_per_second": 357.673,
+      "step": 1665
+    },
+    {
+      "epoch": 62.22,
+      "eval_loss": 0.44057419896125793,
+      "eval_micro_f1": 0.5917159758355107,
+      "eval_runtime": 2.7609,
+      "eval_samples_per_second": 365.466,
+      "step": 1680
+    },
+    {
+      "epoch": 62.78,
+      "eval_loss": 0.439024418592453,
+      "eval_micro_f1": 0.5890052351057639,
+      "eval_runtime": 2.6749,
+      "eval_samples_per_second": 377.207,
+      "step": 1695
+    },
+    {
+      "epoch": 63.33,
+      "eval_loss": 0.43084436655044556,
+      "eval_micro_f1": 0.5979247725248276,
+      "eval_runtime": 2.8108,
+      "eval_samples_per_second": 358.977,
+      "step": 1710
+    },
+    {
+      "epoch": 63.89,
+      "eval_loss": 0.44356903433799744,
+      "eval_micro_f1": 0.596281540009931,
+      "eval_runtime": 2.8271,
+      "eval_samples_per_second": 356.898,
+      "step": 1725
+    },
+    {
+      "epoch": 64.44,
+      "eval_loss": 0.4429439902305603,
+      "eval_micro_f1": 0.5951283734345062,
+      "eval_runtime": 2.831,
+      "eval_samples_per_second": 356.413,
+      "step": 1740
+    },
+    {
+      "epoch": 65.0,
+      "eval_loss": 0.4403350353240967,
+      "eval_micro_f1": 0.5892388446482992,
+      "eval_runtime": 2.8422,
+      "eval_samples_per_second": 355.008,
+      "step": 1755
+    },
+    {
+      "epoch": 65.56,
+      "eval_loss": 0.46583083271980286,
+      "eval_micro_f1": 0.5951416999122735,
+      "eval_runtime": 2.7669,
+      "eval_samples_per_second": 364.664,
+      "step": 1770
+    },
+    {
+      "epoch": 66.11,
+      "eval_loss": 0.4445974826812744,
+      "eval_micro_f1": 0.5980198014847782,
+      "eval_runtime": 3.0164,
+      "eval_samples_per_second": 334.505,
+      "step": 1785
+    },
+    {
+      "epoch": 66.67,
+      "eval_loss": 0.43158385157585144,
+      "eval_micro_f1": 0.6071197406029162,
+      "eval_runtime": 2.7661,
+      "eval_samples_per_second": 364.775,
+      "step": 1800
+    },
+    {
+      "epoch": 67.22,
+      "eval_loss": 0.4658968448638916,
+      "eval_micro_f1": 0.5960890082733319,
+      "eval_runtime": 2.7622,
+      "eval_samples_per_second": 365.283,
+      "step": 1815
+    },
+    {
+      "epoch": 67.78,
+      "eval_loss": 0.41817206144332886,
+      "eval_micro_f1": 0.6071887029678795,
+      "eval_runtime": 3.0896,
+      "eval_samples_per_second": 326.582,
+      "step": 1830
+    },
+    {
+      "epoch": 68.33,
+      "eval_loss": 0.4263642132282257,
+      "eval_micro_f1": 0.6036745401863378,
+      "eval_runtime": 2.951,
+      "eval_samples_per_second": 341.914,
+      "step": 1845
+    },
+    {
+      "epoch": 68.89,
+      "eval_loss": 0.462801992893219,
+      "eval_micro_f1": 0.5931506844412365,
+      "eval_runtime": 2.8805,
+      "eval_samples_per_second": 350.286,
+      "step": 1860
+    },
+    {
+      "epoch": 69.44,
+      "eval_loss": 0.43784838914871216,
+      "eval_micro_f1": 0.6040709121763497,
+      "eval_runtime": 5.0995,
+      "eval_samples_per_second": 197.862,
+      "step": 1875
+    },
+    {
+      "epoch": 70.0,
+      "eval_loss": 0.4683259129524231,
+      "eval_micro_f1": 0.589918255639642,
+      "eval_runtime": 2.688,
+      "eval_samples_per_second": 375.372,
+      "step": 1890
+    },
+    {
+      "epoch": 70.56,
+      "eval_loss": 0.4364282190799713,
+      "eval_micro_f1": 0.6083550908871933,
+      "eval_runtime": 2.9256,
+      "eval_samples_per_second": 344.881,
+      "step": 1905
+    },
+    {
+      "epoch": 71.11,
+      "eval_loss": 0.420198529958725,
+      "eval_micro_f1": 0.6038709672442539,
+      "eval_runtime": 2.7887,
+      "eval_samples_per_second": 361.813,
+      "step": 1920
+    },
+    {
+      "epoch": 71.67,
+      "eval_loss": 0.4335069954395294,
+      "eval_micro_f1": 0.6136514242896463,
+      "eval_runtime": 2.8957,
+      "eval_samples_per_second": 348.451,
+      "step": 1935
+    },
+    {
+      "epoch": 72.22,
+      "eval_loss": 0.43404510617256165,
+      "eval_micro_f1": 0.6122448974634501,
+      "eval_runtime": 2.8428,
+      "eval_samples_per_second": 354.931,
+      "step": 1950
+    },
+    {
+      "epoch": 72.78,
+      "eval_loss": 0.4198473393917084,
+      "eval_micro_f1": 0.6124279303153245,
+      "eval_runtime": 2.7227,
+      "eval_samples_per_second": 370.594,
+      "step": 1965
+    },
+    {
+      "epoch": 73.33,
+      "eval_loss": 0.4370868504047394,
+      "eval_micro_f1": 0.6040709121763497,
+      "eval_runtime": 2.8337,
+      "eval_samples_per_second": 356.07,
+      "step": 1980
+    },
+    {
+      "epoch": 73.89,
+      "eval_loss": 0.438154935836792,
+      "eval_micro_f1": 0.6039215681309684,
+      "eval_runtime": 2.872,
+      "eval_samples_per_second": 351.325,
+      "step": 1995
+    },
+    {
+      "epoch": 74.07,
+      "learning_rate": 7.530864197530865e-06,
+      "loss": 0.0027,
+      "step": 2000
+    },
+    {
+      "epoch": 74.44,
+      "eval_loss": 0.44406837224960327,
+      "eval_micro_f1": 0.610928241768916,
+      "eval_runtime": 2.8987,
+      "eval_samples_per_second": 348.09,
+      "step": 2010
+    },
+    {
+      "epoch": 75.0,
+      "eval_loss": 0.43494340777397156,
+      "eval_micro_f1": 0.6054687495031313,
+      "eval_runtime": 2.7285,
+      "eval_samples_per_second": 369.796,
+      "step": 2025
+    },
+    {
+      "epoch": 75.56,
+      "eval_loss": 0.42609041929244995,
+      "eval_micro_f1": 0.599613152306912,
+      "eval_runtime": 5.45,
+      "eval_samples_per_second": 185.136,
+      "step": 2040
+    },
+    {
+      "epoch": 76.11,
+      "eval_loss": 0.43071994185447693,
+      "eval_micro_f1": 0.6015523927755054,
+      "eval_runtime": 3.0024,
+      "eval_samples_per_second": 336.062,
+      "step": 2055
+    },
+    {
+      "epoch": 76.67,
+      "eval_loss": 0.42303502559661865,
+      "eval_micro_f1": 0.6058673464402023,
+      "eval_runtime": 2.9279,
+      "eval_samples_per_second": 344.615,
+      "step": 2070
+    },
+    {
+      "epoch": 77.22,
+      "eval_loss": 0.42637309432029724,
+      "eval_micro_f1": 0.6069142120497178,
+      "eval_runtime": 2.8897,
+      "eval_samples_per_second": 349.172,
+      "step": 2085
+    },
+    {
+      "epoch": 77.78,
+      "eval_loss": 0.4545460343360901,
+      "eval_micro_f1": 0.6010638292926734,
+      "eval_runtime": 2.7907,
+      "eval_samples_per_second": 361.556,
+      "step": 2100
+    },
+    {
+      "epoch": 78.33,
+      "eval_loss": 0.4475354254245758,
+      "eval_micro_f1": 0.6023622042283344,
+      "eval_runtime": 2.7193,
+      "eval_samples_per_second": 371.053,
+      "step": 2115
+    },
+    {
+      "epoch": 78.89,
+      "eval_loss": 0.4469280242919922,
+      "eval_micro_f1": 0.6076618224901117,
+      "eval_runtime": 2.782,
+      "eval_samples_per_second": 362.689,
+      "step": 2130
+    },
+    {
+      "epoch": 79.44,
+      "eval_loss": 0.428404837846756,
+      "eval_micro_f1": 0.6079182625921914,
+      "eval_runtime": 3.2034,
+      "eval_samples_per_second": 314.98,
+      "step": 2145
+    },
+    {
+      "epoch": 80.0,
+      "eval_loss": 0.44551852345466614,
+      "eval_micro_f1": 0.6031331587723175,
+      "eval_runtime": 2.7718,
+      "eval_samples_per_second": 364.029,
+      "step": 2160
+    },
+    {
+      "epoch": 80.56,
+      "eval_loss": 0.4296429455280304,
+      "eval_micro_f1": 0.6059456035489609,
+      "eval_runtime": 2.9659,
+      "eval_samples_per_second": 340.203,
+      "step": 2175
+    },
+    {
+      "epoch": 81.11,
+      "eval_loss": 0.43831172585487366,
+      "eval_micro_f1": 0.6083601281194115,
+      "eval_runtime": 2.8059,
+      "eval_samples_per_second": 359.602,
+      "step": 2190
+    },
+    {
+      "epoch": 81.67,
+      "eval_loss": 0.4473990797996521,
+      "eval_micro_f1": 0.6040443569723377,
+      "eval_runtime": 2.839,
+      "eval_samples_per_second": 355.409,
+      "step": 2205
+    },
+    {
+      "epoch": 82.22,
+      "eval_loss": 0.4587768316268921,
+      "eval_micro_f1": 0.604527296443017,
+      "eval_runtime": 3.2702,
+      "eval_samples_per_second": 308.547,
+      "step": 2220
+    },
+    {
+      "epoch": 82.78,
+      "eval_loss": 0.44519737362861633,
+      "eval_micro_f1": 0.6028833546807199,
+      "eval_runtime": 2.8894,
+      "eval_samples_per_second": 349.207,
+      "step": 2235
+    },
+    {
+      "epoch": 83.33,
+      "eval_loss": 0.4549151062965393,
+      "eval_micro_f1": 0.6127320949958172,
+      "eval_runtime": 2.8658,
+      "eval_samples_per_second": 352.078,
+      "step": 2250
+    },
+    {
+      "epoch": 83.89,
+      "eval_loss": 0.45440971851348877,
+      "eval_micro_f1": 0.609271522683756,
+      "eval_runtime": 2.701,
+      "eval_samples_per_second": 373.57,
+      "step": 2265
+    },
+    {
+      "epoch": 84.44,
+      "eval_loss": 0.4553206264972687,
+      "eval_micro_f1": 0.6033003295375763,
+      "eval_runtime": 2.9556,
+      "eval_samples_per_second": 341.383,
+      "step": 2280
+    },
+    {
+      "epoch": 85.0,
+      "eval_loss": 0.45224007964134216,
+      "eval_micro_f1": 0.6035502953621225,
+      "eval_runtime": 2.8217,
+      "eval_samples_per_second": 357.59,
+      "step": 2295
+    },
+    {
+      "epoch": 85.56,
+      "eval_loss": 0.45437538623809814,
+      "eval_micro_f1": 0.602220770241534,
+      "eval_runtime": 3.301,
+      "eval_samples_per_second": 305.665,
+      "step": 2310
+    },
+    {
+      "epoch": 86.11,
+      "eval_loss": 0.4574569761753082,
+      "eval_micro_f1": 0.60397350943873,
+      "eval_runtime": 2.8985,
+      "eval_samples_per_second": 348.115,
+      "step": 2325
+    },
+    {
+      "epoch": 86.67,
+      "eval_loss": 0.4623734951019287,
+      "eval_micro_f1": 0.6065464256917272,
+      "eval_runtime": 2.8932,
+      "eval_samples_per_second": 348.744,
+      "step": 2340
+    },
+    {
+      "epoch": 87.22,
+      "eval_loss": 0.4351001977920532,
+      "eval_micro_f1": 0.6079182625921914,
+      "eval_runtime": 2.7308,
+      "eval_samples_per_second": 369.494,
+      "step": 2355
+    },
+    {
+      "epoch": 87.78,
+      "eval_loss": 0.45341917872428894,
+      "eval_micro_f1": 0.6104152928465378,
+      "eval_runtime": 2.8101,
+      "eval_samples_per_second": 359.06,
+      "step": 2370
+    },
+    {
+      "epoch": 88.33,
+      "eval_loss": 0.4562045931816101,
+      "eval_micro_f1": 0.6096752811484984,
+      "eval_runtime": 2.8413,
+      "eval_samples_per_second": 355.119,
+      "step": 2385
+    },
+    {
+      "epoch": 88.89,
+      "eval_loss": 0.4691969156265259,
+      "eval_micro_f1": 0.6022803482657825,
+      "eval_runtime": 2.9074,
+      "eval_samples_per_second": 347.04,
+      "step": 2400
+    },
+    {
+      "epoch": 89.44,
+      "eval_loss": 0.45213738083839417,
+      "eval_micro_f1": 0.6061801441458139,
+      "eval_runtime": 2.7749,
+      "eval_samples_per_second": 363.623,
+      "step": 2415
+    },
+    {
+      "epoch": 90.0,
+      "eval_loss": 0.46215492486953735,
+      "eval_micro_f1": 0.6038487055436795,
+      "eval_runtime": 2.9253,
+      "eval_samples_per_second": 344.927,
+      "step": 2430
+    },
+    {
+      "epoch": 90.56,
+      "eval_loss": 0.46368929743766785,
+      "eval_micro_f1": 0.6040978184075654,
+      "eval_runtime": 2.7791,
+      "eval_samples_per_second": 363.069,
+      "step": 2445
+    },
+    {
+      "epoch": 91.11,
+      "eval_loss": 0.4574613869190216,
+      "eval_micro_f1": 0.604467805023116,
+      "eval_runtime": 2.9721,
+      "eval_samples_per_second": 339.485,
+      "step": 2460
+    },
+    {
+      "epoch": 91.67,
+      "eval_loss": 0.45631158351898193,
+      "eval_micro_f1": 0.5914754093399357,
+      "eval_runtime": 2.9451,
+      "eval_samples_per_second": 342.608,
+      "step": 2475
+    },
+    {
+      "epoch": 92.22,
+      "eval_loss": 0.4667295515537262,
+      "eval_micro_f1": 0.6009265382209632,
+      "eval_runtime": 2.7344,
+      "eval_samples_per_second": 369.006,
+      "step": 2490
+    },
+    {
+      "epoch": 92.59,
+      "learning_rate": 6.913580246913581e-06,
+      "loss": 0.0019,
+      "step": 2500
+    },
+    {
+      "epoch": 92.78,
+      "eval_loss": 0.47763481736183167,
+      "eval_micro_f1": 0.6078824310357469,
+      "eval_runtime": 2.5881,
+      "eval_samples_per_second": 389.854,
+      "step": 2505
+    },
+    {
+      "epoch": 93.33,
+      "eval_loss": 0.4719112515449524,
+      "eval_micro_f1": 0.6050870142318563,
+      "eval_runtime": 2.95,
+      "eval_samples_per_second": 342.028,
+      "step": 2520
+    },
+    {
+      "epoch": 93.89,
+      "eval_loss": 0.48878467082977295,
+      "eval_micro_f1": 0.6040816321616808,
+      "eval_runtime": 2.8063,
+      "eval_samples_per_second": 359.554,
+      "step": 2535
+    },
+    {
+      "epoch": 94.44,
+      "eval_loss": 0.49078139662742615,
+      "eval_micro_f1": 0.6062246273833015,
+      "eval_runtime": 2.9277,
+      "eval_samples_per_second": 344.635,
+      "step": 2550
+    },
+    {
+      "epoch": 95.0,
+      "eval_loss": 0.45815572142601013,
+      "eval_micro_f1": 0.6007751933008817,
+      "eval_runtime": 2.7383,
+      "eval_samples_per_second": 368.477,
+      "step": 2565
+    },
+    {
+      "epoch": 95.56,
+      "eval_loss": 0.4678950309753418,
+      "eval_micro_f1": 0.6070959259166737,
+      "eval_runtime": 2.7006,
+      "eval_samples_per_second": 373.624,
+      "step": 2580
+    },
+    {
+      "epoch": 96.11,
+      "eval_loss": 0.45970863103866577,
+      "eval_micro_f1": 0.6091503263008963,
+      "eval_runtime": 2.7672,
+      "eval_samples_per_second": 364.624,
+      "step": 2595
+    },
+    {
+      "epoch": 96.67,
+      "eval_loss": 0.4653020203113556,
+      "eval_micro_f1": 0.6136514242896463,
+      "eval_runtime": 2.8119,
+      "eval_samples_per_second": 358.828,
+      "step": 2610
+    },
+    {
+      "epoch": 97.22,
+      "eval_loss": 0.4667150378227234,
+      "eval_micro_f1": 0.6098535281340884,
+      "eval_runtime": 2.8393,
+      "eval_samples_per_second": 355.37,
+      "step": 2625
+    },
+    {
+      "epoch": 97.78,
+      "eval_loss": 0.4920254349708557,
+      "eval_micro_f1": 0.5974202303297673,
+      "eval_runtime": 2.6949,
+      "eval_samples_per_second": 374.413,
+      "step": 2640
+    },
+    {
+      "epoch": 98.33,
+      "eval_loss": 0.47361138463020325,
+      "eval_micro_f1": 0.6040268451442332,
+      "eval_runtime": 2.681,
+      "eval_samples_per_second": 376.351,
+      "step": 2655
+    },
+    {
+      "epoch": 98.89,
+      "eval_loss": 0.47226226329803467,
+      "eval_micro_f1": 0.6090425526969181,
+      "eval_runtime": 2.7279,
+      "eval_samples_per_second": 369.88,
+      "step": 2670
+    },
+    {
+      "epoch": 99.44,
+      "eval_loss": 0.46143531799316406,
+      "eval_micro_f1": 0.6098360650776167,
+      "eval_runtime": 2.961,
+      "eval_samples_per_second": 340.769,
+      "step": 2685
+    },
+    {
+      "epoch": 100.0,
+      "eval_loss": 0.47409743070602417,
+      "eval_micro_f1": 0.6027944106831641,
+      "eval_runtime": 2.7887,
+      "eval_samples_per_second": 361.812,
+      "step": 2700
+    },
+    {
+      "epoch": 100.56,
+      "eval_loss": 0.47864341735839844,
+      "eval_micro_f1": 0.5961152037930597,
+      "eval_runtime": 2.9125,
+      "eval_samples_per_second": 346.443,
+      "step": 2715
+    },
+    {
+      "epoch": 101.11,
+      "eval_loss": 0.4809187650680542,
+      "eval_micro_f1": 0.5916055957748548,
+      "eval_runtime": 2.8567,
+      "eval_samples_per_second": 353.205,
+      "step": 2730
+    },
+    {
+      "epoch": 101.67,
+      "eval_loss": 0.48038962483406067,
+      "eval_micro_f1": 0.5961665560149785,
+      "eval_runtime": 3.113,
+      "eval_samples_per_second": 324.128,
+      "step": 2745
+    },
+    {
+      "epoch": 102.22,
+      "eval_loss": 0.47263458371162415,
+      "eval_micro_f1": 0.5973856204185587,
+      "eval_runtime": 2.8272,
+      "eval_samples_per_second": 356.892,
+      "step": 2760
+    },
+    {
+      "epoch": 102.78,
+      "eval_loss": 0.49390164017677307,
+      "eval_micro_f1": 0.5964912275775887,
+      "eval_runtime": 2.839,
+      "eval_samples_per_second": 355.406,
+      "step": 2775
+    },
+    {
+      "epoch": 103.33,
+      "eval_loss": 0.4963178038597107,
+      "eval_micro_f1": 0.5960890082733319,
+      "eval_runtime": 2.94,
+      "eval_samples_per_second": 343.198,
+      "step": 2790
+    },
+    {
+      "epoch": 103.89,
+      "eval_loss": 0.48177576065063477,
+      "eval_micro_f1": 0.598540145490601,
+      "eval_runtime": 2.8107,
+      "eval_samples_per_second": 358.986,
+      "step": 2805
+    },
+    {
+      "epoch": 104.44,
+      "eval_loss": 0.4600963592529297,
+      "eval_micro_f1": 0.6080729161697948,
+      "eval_runtime": 3.0704,
+      "eval_samples_per_second": 328.617,
+      "step": 2820
+    },
+    {
+      "epoch": 105.0,
+      "eval_loss": 0.4570736587047577,
+      "eval_micro_f1": 0.6127989652426145,
+      "eval_runtime": 2.704,
+      "eval_samples_per_second": 373.15,
+      "step": 2835
+    },
+    {
+      "epoch": 105.56,
+      "eval_loss": 0.461261123418808,
+      "eval_micro_f1": 0.6085825742754358,
+      "eval_runtime": 2.8777,
+      "eval_samples_per_second": 350.626,
+      "step": 2850
+    },
+    {
+      "epoch": 106.11,
+      "eval_loss": 0.4839690923690796,
+      "eval_micro_f1": 0.5975935823938219,
+      "eval_runtime": 2.8026,
+      "eval_samples_per_second": 360.024,
+      "step": 2865
+    },
+    {
+      "epoch": 106.67,
+      "eval_loss": 0.4647506773471832,
+      "eval_micro_f1": 0.5983017630566925,
+      "eval_runtime": 2.9877,
+      "eval_samples_per_second": 337.713,
+      "step": 2880
+    },
+    {
+      "epoch": 107.22,
+      "eval_loss": 0.4726808965206146,
+      "eval_micro_f1": 0.602757714540107,
+      "eval_runtime": 2.9112,
+      "eval_samples_per_second": 346.596,
+      "step": 2895
+    },
+    {
+      "epoch": 107.78,
+      "eval_loss": 0.4857654869556427,
+      "eval_micro_f1": 0.60357379169548,
+      "eval_runtime": 2.7549,
+      "eval_samples_per_second": 366.255,
+      "step": 2910
+    },
+    {
+      "epoch": 108.33,
+      "eval_loss": 0.4960106909275055,
+      "eval_micro_f1": 0.6008064511197411,
+      "eval_runtime": 2.8094,
+      "eval_samples_per_second": 359.154,
+      "step": 2925
+    },
+    {
+      "epoch": 108.89,
+      "eval_loss": 0.48388785123825073,
+      "eval_micro_f1": 0.605540896602121,
+      "eval_runtime": 2.7583,
+      "eval_samples_per_second": 365.811,
+      "step": 2940
+    },
+    {
+      "epoch": 109.44,
+      "eval_loss": 0.4782085716724396,
+      "eval_micro_f1": 0.600920446578431,
+      "eval_runtime": 2.7759,
+      "eval_samples_per_second": 363.492,
+      "step": 2955
+    },
+    {
+      "epoch": 110.0,
+      "eval_loss": 0.46463891863822937,
+      "eval_micro_f1": 0.6036269425078358,
+      "eval_runtime": 2.9231,
+      "eval_samples_per_second": 345.179,
+      "step": 2970
+    },
+    {
+      "epoch": 110.56,
+      "eval_loss": 0.4623663127422333,
+      "eval_micro_f1": 0.6022063590096456,
+      "eval_runtime": 2.916,
+      "eval_samples_per_second": 346.024,
+      "step": 2985
+    },
+    {
+      "epoch": 111.11,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 0.0015,
+      "step": 3000
+    },
+    {
+      "epoch": 111.11,
+      "eval_loss": 0.48743540048599243,
+      "eval_micro_f1": 0.6105121288872556,
+      "eval_runtime": 2.8491,
+      "eval_samples_per_second": 354.151,
+      "step": 3000
+    },
+    {
+      "epoch": 111.67,
+      "eval_loss": 0.481542706489563,
+      "eval_micro_f1": 0.6048064080506649,
+      "eval_runtime": 2.7482,
+      "eval_samples_per_second": 367.151,
+      "step": 3015
+    },
+    {
+      "epoch": 112.22,
+      "eval_loss": 0.48628905415534973,
+      "eval_micro_f1": 0.6056149727681427,
+      "eval_runtime": 2.7068,
+      "eval_samples_per_second": 372.77,
+      "step": 3030
+    },
+    {
+      "epoch": 112.78,
+      "eval_loss": 0.4836028814315796,
+      "eval_micro_f1": 0.6103038304161319,
+      "eval_runtime": 2.8829,
+      "eval_samples_per_second": 349.99,
+      "step": 3045
+    },
+    {
+      "epoch": 113.33,
+      "eval_loss": 0.4918506145477295,
+      "eval_micro_f1": 0.609271522683756,
+      "eval_runtime": 2.6465,
+      "eval_samples_per_second": 381.254,
+      "step": 3060
+    },
+    {
+      "epoch": 113.89,
+      "eval_loss": 0.4956188499927521,
+      "eval_micro_f1": 0.6121902205378472,
+      "eval_runtime": 2.7907,
+      "eval_samples_per_second": 361.56,
+      "step": 3075
+    },
+    {
+      "epoch": 114.44,
+      "eval_loss": 0.45582953095436096,
+      "eval_micro_f1": 0.6027044425156818,
+      "eval_runtime": 3.1167,
+      "eval_samples_per_second": 323.738,
+      "step": 3090
+    },
+    {
+      "epoch": 115.0,
+      "eval_loss": 0.457035094499588,
+      "eval_micro_f1": 0.6025974021002934,
+      "eval_runtime": 2.7508,
+      "eval_samples_per_second": 366.798,
+      "step": 3105
+    },
+    {
+      "epoch": 115.56,
+      "eval_loss": 0.4658013880252838,
+      "eval_micro_f1": 0.5986928099610406,
+      "eval_runtime": 2.8868,
+      "eval_samples_per_second": 349.524,
+      "step": 3120
+    },
+    {
+      "epoch": 116.11,
+      "eval_loss": 0.5294644236564636,
+      "eval_micro_f1": 0.5947441212264672,
+      "eval_runtime": 2.7786,
+      "eval_samples_per_second": 363.131,
+      "step": 3135
+    },
+    {
+      "epoch": 116.67,
+      "eval_loss": 0.4735589921474457,
+      "eval_micro_f1": 0.6054196954729965,
+      "eval_runtime": 2.8461,
+      "eval_samples_per_second": 354.522,
+      "step": 3150
+    },
+    {
+      "epoch": 117.22,
+      "eval_loss": 0.49260324239730835,
+      "eval_micro_f1": 0.6074576266267394,
+      "eval_runtime": 2.8108,
+      "eval_samples_per_second": 358.972,
+      "step": 3165
+    },
+    {
+      "epoch": 117.78,
+      "eval_loss": 0.47801533341407776,
+      "eval_micro_f1": 0.6075446719072551,
+      "eval_runtime": 3.207,
+      "eval_samples_per_second": 314.621,
+      "step": 3180
+    },
+    {
+      "epoch": 118.33,
+      "eval_loss": 0.46694859862327576,
+      "eval_micro_f1": 0.6062992121023449,
+      "eval_runtime": 2.9404,
+      "eval_samples_per_second": 343.154,
+      "step": 3195
+    },
+    {
+      "epoch": 118.89,
+      "eval_loss": 0.4873449504375458,
+      "eval_micro_f1": 0.6041807142746697,
+      "eval_runtime": 2.9009,
+      "eval_samples_per_second": 347.819,
+      "step": 3210
+    },
+    {
+      "epoch": 119.44,
+      "eval_loss": 0.5059316754341125,
+      "eval_micro_f1": 0.6101231185245294,
+      "eval_runtime": 2.7072,
+      "eval_samples_per_second": 372.717,
+      "step": 3225
+    },
+    {
+      "epoch": 120.0,
+      "eval_loss": 0.484662264585495,
+      "eval_micro_f1": 0.6057371576112552,
+      "eval_runtime": 2.6285,
+      "eval_samples_per_second": 383.873,
+      "step": 3240
+    },
+    {
+      "epoch": 120.56,
+      "eval_loss": 0.4770389795303345,
+      "eval_micro_f1": 0.5992115632360008,
+      "eval_runtime": 2.8218,
+      "eval_samples_per_second": 357.571,
+      "step": 3255
+    },
+    {
+      "epoch": 121.11,
+      "eval_loss": 0.4697888195514679,
+      "eval_micro_f1": 0.6039215681309684,
+      "eval_runtime": 2.7896,
+      "eval_samples_per_second": 361.707,
+      "step": 3270
+    },
+    {
+      "epoch": 121.67,
+      "eval_loss": 0.4786725640296936,
+      "eval_micro_f1": 0.6033222586415673,
+      "eval_runtime": 2.7713,
+      "eval_samples_per_second": 364.087,
+      "step": 3285
+    },
+    {
+      "epoch": 122.22,
+      "eval_loss": 0.4789953827857971,
+      "eval_micro_f1": 0.6050531909947958,
+      "eval_runtime": 2.8107,
+      "eval_samples_per_second": 358.989,
+      "step": 3300
+    },
+    {
+      "epoch": 122.78,
+      "eval_loss": 0.48166513442993164,
+      "eval_micro_f1": 0.6030687119808384,
+      "eval_runtime": 2.8521,
+      "eval_samples_per_second": 353.778,
+      "step": 3315
+    },
+    {
+      "epoch": 123.33,
+      "eval_loss": 0.4727168679237366,
+      "eval_micro_f1": 0.6074950685376597,
+      "eval_runtime": 2.8478,
+      "eval_samples_per_second": 354.309,
+      "step": 3330
+    },
+    {
+      "epoch": 123.89,
+      "eval_loss": 0.4770917296409607,
+      "eval_micro_f1": 0.6074270552080258,
+      "eval_runtime": 2.732,
+      "eval_samples_per_second": 369.331,
+      "step": 3345
+    },
+    {
+      "epoch": 124.44,
+      "eval_loss": 0.4960639178752899,
+      "eval_micro_f1": 0.6017579440648337,
+      "eval_runtime": 2.8519,
+      "eval_samples_per_second": 353.801,
+      "step": 3360
+    },
+    {
+      "epoch": 125.0,
+      "eval_loss": 0.48805487155914307,
+      "eval_micro_f1": 0.6034712945660202,
+      "eval_runtime": 3.0062,
+      "eval_samples_per_second": 335.635,
+      "step": 3375
+    },
+    {
+      "epoch": 125.56,
+      "eval_loss": 0.4866376519203186,
+      "eval_micro_f1": 0.603874415003688,
+      "eval_runtime": 2.8917,
+      "eval_samples_per_second": 348.924,
+      "step": 3390
+    },
+    {
+      "epoch": 126.11,
+      "eval_loss": 0.4823364019393921,
+      "eval_micro_f1": 0.5997340420586326,
+      "eval_runtime": 2.9129,
+      "eval_samples_per_second": 346.385,
+      "step": 3405
+    },
+    {
+      "epoch": 126.67,
+      "eval_loss": 0.4744039475917816,
+      "eval_micro_f1": 0.605384109812592,
+      "eval_runtime": 3.075,
+      "eval_samples_per_second": 328.126,
+      "step": 3420
+    },
+    {
+      "epoch": 127.22,
+      "eval_loss": 0.5212647914886475,
+      "eval_micro_f1": 0.6041379305453698,
+      "eval_runtime": 2.9232,
+      "eval_samples_per_second": 345.173,
+      "step": 3435
+    },
+    {
+      "epoch": 127.78,
+      "eval_loss": 0.4715367555618286,
+      "eval_micro_f1": 0.6022063590096456,
+      "eval_runtime": 2.8539,
+      "eval_samples_per_second": 353.545,
+      "step": 3450
+    },
+    {
+      "epoch": 128.33,
+      "eval_loss": 0.47289079427719116,
+      "eval_micro_f1": 0.6022063590096456,
+      "eval_runtime": 2.7893,
+      "eval_samples_per_second": 361.741,
+      "step": 3465
+    },
+    {
+      "epoch": 128.89,
+      "eval_loss": 0.4845651090145111,
+      "eval_micro_f1": 0.6059405935639757,
+      "eval_runtime": 2.7207,
+      "eval_samples_per_second": 370.854,
+      "step": 3480
+    },
+    {
+      "epoch": 129.44,
+      "eval_loss": 0.4824657142162323,
+      "eval_micro_f1": 0.6064601181596609,
+      "eval_runtime": 2.9028,
+      "eval_samples_per_second": 347.591,
+      "step": 3495
+    },
+    {
+      "epoch": 129.63,
+      "learning_rate": 5.6790123456790125e-06,
+      "loss": 0.0014,
+      "step": 3500
+    },
+    {
+      "epoch": 130.0,
+      "eval_loss": 0.48599931597709656,
+      "eval_micro_f1": 0.6022501649582217,
+      "eval_runtime": 5.6624,
+      "eval_samples_per_second": 178.194,
+      "step": 3510
+    },
+    {
+      "epoch": 130.56,
+      "eval_loss": 0.49249857664108276,
+      "eval_micro_f1": 0.6046822737536912,
+      "eval_runtime": 2.944,
+      "eval_samples_per_second": 342.729,
+      "step": 3525
+    },
+    {
+      "epoch": 131.11,
+      "eval_loss": 0.49113982915878296,
+      "eval_micro_f1": 0.6041527121654535,
+      "eval_runtime": 2.7747,
+      "eval_samples_per_second": 363.645,
+      "step": 3540
+    },
+    {
+      "epoch": 131.67,
+      "eval_loss": 0.5029997229576111,
+      "eval_micro_f1": 0.6005434777692837,
+      "eval_runtime": 2.8234,
+      "eval_samples_per_second": 357.371,
+      "step": 3555
+    },
+    {
+      "epoch": 132.22,
+      "eval_loss": 0.5053459405899048,
+      "eval_micro_f1": 0.6035182674374324,
+      "eval_runtime": 2.9126,
+      "eval_samples_per_second": 346.429,
+      "step": 3570
+    },
+    {
+      "epoch": 132.78,
+      "eval_loss": 0.5123698711395264,
+      "eval_micro_f1": 0.6036709716363189,
+      "eval_runtime": 2.8359,
+      "eval_samples_per_second": 355.802,
+      "step": 3585
+    },
+    {
+      "epoch": 133.33,
+      "eval_loss": 0.5037609338760376,
+      "eval_micro_f1": 0.5991957099622383,
+      "eval_runtime": 2.9017,
+      "eval_samples_per_second": 347.727,
+      "step": 3600
+    },
+    {
+      "epoch": 133.89,
+      "eval_loss": 0.5023306012153625,
+      "eval_micro_f1": 0.5998655005156747,
+      "eval_runtime": 2.8797,
+      "eval_samples_per_second": 350.383,
+      "step": 3615
+    },
+    {
+      "epoch": 134.44,
+      "eval_loss": 0.5197464227676392,
+      "eval_micro_f1": 0.6027210879439955,
+      "eval_runtime": 2.9897,
+      "eval_samples_per_second": 337.49,
+      "step": 3630
+    },
+    {
+      "epoch": 135.0,
+      "eval_loss": 0.4896288514137268,
+      "eval_micro_f1": 0.6081258186387732,
+      "eval_runtime": 2.712,
+      "eval_samples_per_second": 372.047,
+      "step": 3645
+    },
+    {
+      "epoch": 135.56,
+      "eval_loss": 0.4946237802505493,
+      "eval_micro_f1": 0.6007957554732866,
+      "eval_runtime": 2.6958,
+      "eval_samples_per_second": 374.283,
+      "step": 3660
+    },
+    {
+      "epoch": 136.11,
+      "eval_loss": 0.49949172139167786,
+      "eval_micro_f1": 0.6076458747580622,
+      "eval_runtime": 2.8029,
+      "eval_samples_per_second": 359.985,
+      "step": 3675
+    },
+    {
+      "epoch": 136.67,
+      "eval_loss": 0.5209092497825623,
+      "eval_micro_f1": 0.6053169729240714,
+      "eval_runtime": 2.725,
+      "eval_samples_per_second": 370.279,
+      "step": 3690
+    },
+    {
+      "epoch": 137.22,
+      "eval_loss": 0.4862090051174164,
+      "eval_micro_f1": 0.6060209419120245,
+      "eval_runtime": 3.0966,
+      "eval_samples_per_second": 325.842,
+      "step": 3705
+    },
+    {
+      "epoch": 137.78,
+      "eval_loss": 0.4811948537826538,
+      "eval_micro_f1": 0.6026143785884865,
+      "eval_runtime": 2.8315,
+      "eval_samples_per_second": 356.347,
+      "step": 3720
+    },
+    {
+      "epoch": 138.33,
+      "eval_loss": 0.4642077386379242,
+      "eval_micro_f1": 0.5969387750116423,
+      "eval_runtime": 2.8594,
+      "eval_samples_per_second": 352.869,
+      "step": 3735
+    },
+    {
+      "epoch": 138.89,
+      "eval_loss": 0.47143450379371643,
+      "eval_micro_f1": 0.5992268036259283,
+      "eval_runtime": 5.669,
+      "eval_samples_per_second": 177.986,
+      "step": 3750
+    },
+    {
+      "epoch": 139.44,
+      "eval_loss": 0.4970646798610687,
+      "eval_micro_f1": 0.6022651560679786,
+      "eval_runtime": 2.6901,
+      "eval_samples_per_second": 375.072,
+      "step": 3765
+    },
+    {
+      "epoch": 140.0,
+      "eval_loss": 0.4777345657348633,
+      "eval_micro_f1": 0.6035042175430636,
+      "eval_runtime": 2.8423,
+      "eval_samples_per_second": 354.996,
+      "step": 3780
+    },
+    {
+      "epoch": 140.56,
+      "eval_loss": 0.4680575132369995,
+      "eval_micro_f1": 0.603072982856381,
+      "eval_runtime": 2.8397,
+      "eval_samples_per_second": 355.323,
+      "step": 3795
+    },
+    {
+      "epoch": 141.11,
+      "eval_loss": 0.5216084122657776,
+      "eval_micro_f1": 0.6095751849422281,
+      "eval_runtime": 2.7935,
+      "eval_samples_per_second": 361.199,
+      "step": 3810
+    },
+    {
+      "epoch": 141.67,
+      "eval_loss": 0.5302681922912598,
+      "eval_micro_f1": 0.6048387091842518,
+      "eval_runtime": 2.7321,
+      "eval_samples_per_second": 369.309,
+      "step": 3825
+    },
+    {
+      "epoch": 142.22,
+      "eval_loss": 0.479887992143631,
+      "eval_micro_f1": 0.5989650706539047,
+      "eval_runtime": 2.7677,
+      "eval_samples_per_second": 364.562,
+      "step": 3840
+    },
+    {
+      "epoch": 142.78,
+      "eval_loss": 0.4849691390991211,
+      "eval_micro_f1": 0.5993485337051564,
+      "eval_runtime": 3.1584,
+      "eval_samples_per_second": 319.461,
+      "step": 3855
+    },
+    {
+      "epoch": 143.33,
+      "eval_loss": 0.4907560348510742,
+      "eval_micro_f1": 0.5968379441683912,
+      "eval_runtime": 2.6758,
+      "eval_samples_per_second": 377.087,
+      "step": 3870
+    },
+    {
+      "epoch": 143.89,
+      "eval_loss": 0.48843497037887573,
+      "eval_micro_f1": 0.5934210521357981,
+      "eval_runtime": 2.8527,
+      "eval_samples_per_second": 353.704,
+      "step": 3885
+    },
+    {
+      "epoch": 144.44,
+      "eval_loss": 0.46706297993659973,
+      "eval_micro_f1": 0.6025641020659073,
+      "eval_runtime": 2.6934,
+      "eval_samples_per_second": 374.624,
+      "step": 3900
+    },
+    {
+      "epoch": 145.0,
+      "eval_loss": 0.46969130635261536,
+      "eval_micro_f1": 0.5964010277796472,
+      "eval_runtime": 2.8551,
+      "eval_samples_per_second": 353.405,
+      "step": 3915
+    },
+    {
+      "epoch": 145.56,
+      "eval_loss": 0.5060657262802124,
+      "eval_micro_f1": 0.597840755242904,
+      "eval_runtime": 2.8721,
+      "eval_samples_per_second": 351.312,
+      "step": 3930
+    },
+    {
+      "epoch": 146.11,
+      "eval_loss": 0.502547025680542,
+      "eval_micro_f1": 0.6022803482657825,
+      "eval_runtime": 2.7564,
+      "eval_samples_per_second": 366.055,
+      "step": 3945
+    },
+    {
+      "epoch": 146.67,
+      "eval_loss": 0.4989144504070282,
+      "eval_micro_f1": 0.6033444811115527,
+      "eval_runtime": 2.6862,
+      "eval_samples_per_second": 375.618,
+      "step": 3960
+    },
+    {
+      "epoch": 147.22,
+      "eval_loss": 0.5036029815673828,
+      "eval_micro_f1": 0.6078167110975018,
+      "eval_runtime": 2.6805,
+      "eval_samples_per_second": 376.427,
+      "step": 3975
+    },
+    {
+      "epoch": 147.78,
+      "eval_loss": 0.48515963554382324,
+      "eval_micro_f1": 0.6076115480603484,
+      "eval_runtime": 2.7528,
+      "eval_samples_per_second": 366.539,
+      "step": 3990
+    },
+    {
+      "epoch": 148.15,
+      "learning_rate": 5.061728395061729e-06,
+      "loss": 0.0012,
+      "step": 4000
+    },
+    {
+      "epoch": 148.33,
+      "eval_loss": 0.4917222559452057,
+      "eval_micro_f1": 0.6105960259950125,
+      "eval_runtime": 2.8887,
+      "eval_samples_per_second": 349.293,
+      "step": 4005
+    },
+    {
+      "epoch": 148.89,
+      "eval_loss": 0.4824866056442261,
+      "eval_micro_f1": 0.6094364346282867,
+      "eval_runtime": 2.7544,
+      "eval_samples_per_second": 366.324,
+      "step": 4020
+    },
+    {
+      "epoch": 149.44,
+      "eval_loss": 0.48811236023902893,
+      "eval_micro_f1": 0.5998672855038706,
+      "eval_runtime": 2.8464,
+      "eval_samples_per_second": 354.481,
+      "step": 4035
+    },
+    {
+      "epoch": 150.0,
+      "eval_loss": 0.4892898201942444,
+      "eval_micro_f1": 0.6002656037549456,
+      "eval_runtime": 2.8905,
+      "eval_samples_per_second": 349.08,
+      "step": 4050
+    },
+    {
+      "epoch": 150.56,
+      "eval_loss": 0.48368731141090393,
+      "eval_micro_f1": 0.6113306977915605,
+      "eval_runtime": 2.8043,
+      "eval_samples_per_second": 359.805,
+      "step": 4065
+    },
+    {
+      "epoch": 151.11,
+      "eval_loss": 0.47161540389060974,
+      "eval_micro_f1": 0.6069587623888049,
+      "eval_runtime": 5.4193,
+      "eval_samples_per_second": 186.187,
+      "step": 4080
+    },
+    {
+      "epoch": 151.67,
+      "eval_loss": 0.51511150598526,
+      "eval_micro_f1": 0.6008119074915631,
+      "eval_runtime": 2.7237,
+      "eval_samples_per_second": 370.456,
+      "step": 4095
+    },
+    {
+      "epoch": 152.22,
+      "eval_loss": 0.48475515842437744,
+      "eval_micro_f1": 0.5958279004159163,
+      "eval_runtime": 2.9922,
+      "eval_samples_per_second": 337.209,
+      "step": 4110
+    },
+    {
+      "epoch": 152.78,
+      "eval_loss": 0.49488502740859985,
+      "eval_micro_f1": 0.5973509928824472,
+      "eval_runtime": 2.9176,
+      "eval_samples_per_second": 345.833,
+      "step": 4125
+    },
+    {
+      "epoch": 153.33,
+      "eval_loss": 0.5112221240997314,
+      "eval_micro_f1": 0.6006779656097994,
+      "eval_runtime": 2.748,
+      "eval_samples_per_second": 367.169,
+      "step": 4140
+    },
+    {
+      "epoch": 153.89,
+      "eval_loss": 0.5146142244338989,
+      "eval_micro_f1": 0.5979661012030235,
+      "eval_runtime": 2.7242,
+      "eval_samples_per_second": 370.38,
+      "step": 4155
+    },
+    {
+      "epoch": 154.44,
+      "eval_loss": 0.5149275660514832,
+      "eval_micro_f1": 0.5983717769844653,
+      "eval_runtime": 2.7671,
+      "eval_samples_per_second": 364.646,
+      "step": 4170
+    },
+    {
+      "epoch": 155.0,
+      "eval_loss": 0.5111104846000671,
+      "eval_micro_f1": 0.6025727821754687,
+      "eval_runtime": 3.0733,
+      "eval_samples_per_second": 328.311,
+      "step": 4185
+    },
+    {
+      "epoch": 155.56,
+      "eval_loss": 0.5137137174606323,
+      "eval_micro_f1": 0.6043360428684316,
+      "eval_runtime": 2.923,
+      "eval_samples_per_second": 345.194,
+      "step": 4200
+    },
+    {
+      "epoch": 156.11,
+      "eval_loss": 0.5015988945960999,
+      "eval_micro_f1": 0.597074467590551,
+      "eval_runtime": 2.7706,
+      "eval_samples_per_second": 364.177,
+      "step": 4215
+    },
+    {
+      "epoch": 156.67,
+      "eval_loss": 0.4961719512939453,
+      "eval_micro_f1": 0.5931307788970002,
+      "eval_runtime": 2.9019,
+      "eval_samples_per_second": 347.704,
+      "step": 4230
+    },
+    {
+      "epoch": 157.22,
+      "eval_loss": 0.5027093887329102,
+      "eval_micro_f1": 0.597333332839111,
+      "eval_runtime": 2.6238,
+      "eval_samples_per_second": 384.55,
+      "step": 4245
+    },
+    {
+      "epoch": 157.78,
+      "eval_loss": 0.5096471905708313,
+      "eval_micro_f1": 0.6002673791852624,
+      "eval_runtime": 2.7421,
+      "eval_samples_per_second": 367.961,
+      "step": 4260
+    },
+    {
+      "epoch": 158.33,
+      "eval_loss": 0.492143452167511,
+      "eval_micro_f1": 0.6025049434727842,
+      "eval_runtime": 2.693,
+      "eval_samples_per_second": 374.682,
+      "step": 4275
+    },
+    {
+      "epoch": 158.89,
+      "eval_loss": 0.4941489100456238,
+      "eval_micro_f1": 0.6069057099966367,
+      "eval_runtime": 2.7257,
+      "eval_samples_per_second": 370.186,
+      "step": 4290
+    },
+    {
+      "epoch": 159.44,
+      "eval_loss": 0.49916547536849976,
+      "eval_micro_f1": 0.6028131274367211,
+      "eval_runtime": 2.6556,
+      "eval_samples_per_second": 379.956,
+      "step": 4305
+    },
+    {
+      "epoch": 160.0,
+      "eval_loss": 0.5051697492599487,
+      "eval_micro_f1": 0.6039004702533963,
+      "eval_runtime": 2.8467,
+      "eval_samples_per_second": 354.447,
+      "step": 4320
+    },
+    {
+      "epoch": 160.56,
+      "eval_loss": 0.5021268725395203,
+      "eval_micro_f1": 0.6036217298888524,
+      "eval_runtime": 2.9021,
+      "eval_samples_per_second": 347.679,
+      "step": 4335
+    },
+    {
+      "epoch": 161.11,
+      "eval_loss": 0.49290674924850464,
+      "eval_micro_f1": 0.6039999995057689,
+      "eval_runtime": 2.7123,
+      "eval_samples_per_second": 372.004,
+      "step": 4350
+    },
+    {
+      "epoch": 161.67,
+      "eval_loss": 0.4955635666847229,
+      "eval_micro_f1": 0.6041250826725171,
+      "eval_runtime": 2.7055,
+      "eval_samples_per_second": 372.949,
+      "step": 4365
+    },
+    {
+      "epoch": 162.22,
+      "eval_loss": 0.4938003122806549,
+      "eval_micro_f1": 0.6070245190543998,
+      "eval_runtime": 2.729,
+      "eval_samples_per_second": 369.736,
+      "step": 4380
+    },
+    {
+      "epoch": 162.78,
+      "eval_loss": 0.49344369769096375,
+      "eval_micro_f1": 0.6088682986445134,
+      "eval_runtime": 2.7264,
+      "eval_samples_per_second": 370.086,
+      "step": 4395
+    },
+    {
+      "epoch": 163.33,
+      "eval_loss": 0.49135950207710266,
+      "eval_micro_f1": 0.6076618224901117,
+      "eval_runtime": 2.745,
+      "eval_samples_per_second": 367.579,
+      "step": 4410
+    },
+    {
+      "epoch": 163.89,
+      "eval_loss": 0.49814996123313904,
+      "eval_micro_f1": 0.6058588543657848,
+      "eval_runtime": 2.8875,
+      "eval_samples_per_second": 349.434,
+      "step": 4425
+    },
+    {
+      "epoch": 164.44,
+      "eval_loss": 0.4815930426120758,
+      "eval_micro_f1": 0.6020671829649626,
+      "eval_runtime": 2.6074,
+      "eval_samples_per_second": 386.973,
+      "step": 4440
+    },
+    {
+      "epoch": 165.0,
+      "eval_loss": 0.49755266308784485,
+      "eval_micro_f1": 0.6036988106010811,
+      "eval_runtime": 2.8028,
+      "eval_samples_per_second": 359.999,
+      "step": 4455
+    },
+    {
+      "epoch": 165.56,
+      "eval_loss": 0.5143761038780212,
+      "eval_micro_f1": 0.6035182674374324,
+      "eval_runtime": 2.912,
+      "eval_samples_per_second": 346.498,
+      "step": 4470
+    },
+    {
+      "epoch": 166.11,
+      "eval_loss": 0.5137454271316528,
+      "eval_micro_f1": 0.6040540535616525,
+      "eval_runtime": 5.4478,
+      "eval_samples_per_second": 185.211,
+      "step": 4485
+    },
+    {
+      "epoch": 166.67,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.0011,
+      "step": 4500
+    },
+    {
+      "epoch": 166.67,
+      "eval_loss": 0.5077248215675354,
+      "eval_micro_f1": 0.6040268451442332,
+      "eval_runtime": 2.8383,
+      "eval_samples_per_second": 355.499,
+      "step": 4500
+    },
+    {
+      "epoch": 167.22,
+      "eval_loss": 0.5226935744285583,
+      "eval_micro_f1": 0.6085772629532396,
+      "eval_runtime": 2.8242,
+      "eval_samples_per_second": 357.271,
+      "step": 4515
+    },
+    {
+      "epoch": 167.78,
+      "eval_loss": 0.5281339287757874,
+      "eval_micro_f1": 0.6057298767259631,
+      "eval_runtime": 2.9543,
+      "eval_samples_per_second": 341.537,
+      "step": 4530
+    },
+    {
+      "epoch": 168.33,
+      "eval_loss": 0.5277001857757568,
+      "eval_micro_f1": 0.6030013637655303,
+      "eval_runtime": 3.0817,
+      "eval_samples_per_second": 327.421,
+      "step": 4545
+    },
+    {
+      "epoch": 168.89,
+      "eval_loss": 0.5145075917243958,
+      "eval_micro_f1": 0.6005398105735348,
+      "eval_runtime": 2.7633,
+      "eval_samples_per_second": 365.14,
+      "step": 4560
+    },
+    {
+      "epoch": 169.44,
+      "eval_loss": 0.5113720297813416,
+      "eval_micro_f1": 0.6078167110975018,
+      "eval_runtime": 2.6603,
+      "eval_samples_per_second": 379.279,
+      "step": 4575
+    },
+    {
+      "epoch": 170.0,
+      "eval_loss": 0.5028752088546753,
+      "eval_micro_f1": 0.6122177949899896,
+      "eval_runtime": 2.6409,
+      "eval_samples_per_second": 382.07,
+      "step": 4590
+    },
+    {
+      "epoch": 170.56,
+      "eval_loss": 0.49598032236099243,
+      "eval_micro_f1": 0.6081258186387732,
+      "eval_runtime": 2.8411,
+      "eval_samples_per_second": 355.148,
+      "step": 4605
+    },
+    {
+      "epoch": 171.11,
+      "eval_loss": 0.48839080333709717,
+      "eval_micro_f1": 0.6007702177304038,
+      "eval_runtime": 3.0639,
+      "eval_samples_per_second": 329.319,
+      "step": 4620
+    },
+    {
+      "epoch": 171.67,
+      "eval_loss": 0.4986254572868347,
+      "eval_micro_f1": 0.5999999995035225,
+      "eval_runtime": 2.7467,
+      "eval_samples_per_second": 367.355,
+      "step": 4635
+    },
+    {
+      "epoch": 172.22,
+      "eval_loss": 0.5127492547035217,
+      "eval_micro_f1": 0.6006644513326007,
+      "eval_runtime": 2.816,
+      "eval_samples_per_second": 358.308,
+      "step": 4650
+    },
+    {
+      "epoch": 172.78,
+      "eval_loss": 0.5176624655723572,
+      "eval_micro_f1": 0.6008010675967316,
+      "eval_runtime": 2.8367,
+      "eval_samples_per_second": 355.7,
+      "step": 4665
+    },
+    {
+      "epoch": 173.33,
+      "eval_loss": 0.5219458937644958,
+      "eval_micro_f1": 0.6021505371412447,
+      "eval_runtime": 2.6289,
+      "eval_samples_per_second": 383.805,
+      "step": 4680
+    },
+    {
+      "epoch": 173.89,
+      "eval_loss": 0.513482391834259,
+      "eval_micro_f1": 0.6048064080506649,
+      "eval_runtime": 2.7042,
+      "eval_samples_per_second": 373.121,
+      "step": 4695
+    },
+    {
+      "epoch": 174.44,
+      "eval_loss": 0.5164716839790344,
+      "eval_micro_f1": 0.6071188712327275,
+      "eval_runtime": 2.718,
+      "eval_samples_per_second": 371.224,
+      "step": 4710
+    },
+    {
+      "epoch": 175.0,
+      "eval_loss": 0.5149679780006409,
+      "eval_micro_f1": 0.606951871163863,
+      "eval_runtime": 2.7247,
+      "eval_samples_per_second": 370.316,
+      "step": 4725
+    },
+    {
+      "epoch": 175.56,
+      "eval_loss": 0.5000478029251099,
+      "eval_micro_f1": 0.606258148134286,
+      "eval_runtime": 2.8806,
+      "eval_samples_per_second": 350.27,
+      "step": 4740
+    },
+    {
+      "epoch": 176.11,
+      "eval_loss": 0.5007814168930054,
+      "eval_micro_f1": 0.6069326352131936,
+      "eval_runtime": 2.89,
+      "eval_samples_per_second": 349.137,
+      "step": 4755
+    },
+    {
+      "epoch": 176.67,
+      "eval_loss": 0.49977752566337585,
+      "eval_micro_f1": 0.6075619290990822,
+      "eval_runtime": 2.7758,
+      "eval_samples_per_second": 363.5,
+      "step": 4770
+    },
+    {
+      "epoch": 177.22,
+      "eval_loss": 0.5320616364479065,
+      "eval_micro_f1": 0.6054922968941857,
+      "eval_runtime": 2.6602,
+      "eval_samples_per_second": 379.295,
+      "step": 4785
+    },
+    {
+      "epoch": 177.78,
+      "eval_loss": 0.536164402961731,
+      "eval_micro_f1": 0.6148648643724489,
+      "eval_runtime": 2.7976,
+      "eval_samples_per_second": 360.672,
+      "step": 4800
+    },
+    {
+      "epoch": 178.33,
+      "eval_loss": 0.531025230884552,
+      "eval_micro_f1": 0.6159029644667632,
+      "eval_runtime": 2.8465,
+      "eval_samples_per_second": 354.474,
+      "step": 4815
+    },
+    {
+      "epoch": 178.89,
+      "eval_loss": 0.5245648622512817,
+      "eval_micro_f1": 0.6206434311418341,
+      "eval_runtime": 2.8904,
+      "eval_samples_per_second": 349.086,
+      "step": 4830
+    },
+    {
+      "epoch": 179.44,
+      "eval_loss": 0.5113655924797058,
+      "eval_micro_f1": 0.6131386856365669,
+      "eval_runtime": 2.7606,
+      "eval_samples_per_second": 365.495,
+      "step": 4845
+    },
+    {
+      "epoch": 180.0,
+      "eval_loss": 0.525195300579071,
+      "eval_micro_f1": 0.6148648643724489,
+      "eval_runtime": 2.8542,
+      "eval_samples_per_second": 353.515,
+      "step": 4860
+    },
+    {
+      "epoch": 180.56,
+      "eval_loss": 0.505492091178894,
+      "eval_micro_f1": 0.6058201053249034,
+      "eval_runtime": 2.7054,
+      "eval_samples_per_second": 372.964,
+      "step": 4875
+    },
+    {
+      "epoch": 181.11,
+      "eval_loss": 0.5221575498580933,
+      "eval_micro_f1": 0.6072386054045867,
+      "eval_runtime": 2.9725,
+      "eval_samples_per_second": 339.444,
+      "step": 4890
+    },
+    {
+      "epoch": 181.67,
+      "eval_loss": 0.5590493679046631,
+      "eval_micro_f1": 0.6088154265078849,
+      "eval_runtime": 2.7217,
+      "eval_samples_per_second": 370.728,
+      "step": 4905
+    },
+    {
+      "epoch": 182.22,
+      "eval_loss": 0.5311306118965149,
+      "eval_micro_f1": 0.607407406914518,
+      "eval_runtime": 2.9279,
+      "eval_samples_per_second": 344.62,
+      "step": 4920
+    },
+    {
+      "epoch": 182.78,
+      "eval_loss": 0.534024715423584,
+      "eval_micro_f1": 0.6082265672753385,
+      "eval_runtime": 2.8479,
+      "eval_samples_per_second": 354.291,
+      "step": 4935
+    },
+    {
+      "epoch": 183.33,
+      "eval_loss": 0.5438510179519653,
+      "eval_micro_f1": 0.6151761512695008,
+      "eval_runtime": 2.7432,
+      "eval_samples_per_second": 367.825,
+      "step": 4950
+    },
+    {
+      "epoch": 183.89,
+      "eval_loss": 0.5505983233451843,
+      "eval_micro_f1": 0.615803814222713,
+      "eval_runtime": 2.8057,
+      "eval_samples_per_second": 359.624,
+      "step": 4965
+    },
+    {
+      "epoch": 184.44,
+      "eval_loss": 0.5643740296363831,
+      "eval_micro_f1": 0.6116838483075315,
+      "eval_runtime": 2.6873,
+      "eval_samples_per_second": 375.47,
+      "step": 4980
+    },
+    {
+      "epoch": 185.0,
+      "eval_loss": 0.5376504063606262,
+      "eval_micro_f1": 0.6112237993724612,
+      "eval_runtime": 3.3142,
+      "eval_samples_per_second": 304.448,
+      "step": 4995
+    },
+    {
+      "epoch": 185.19,
+      "learning_rate": 3.827160493827161e-06,
+      "loss": 0.0011,
+      "step": 5000
+    },
+    {
+      "epoch": 185.56,
+      "eval_loss": 0.5281098484992981,
+      "eval_micro_f1": 0.6072874489001112,
+      "eval_runtime": 2.9,
+      "eval_samples_per_second": 347.931,
+      "step": 5010
+    },
+    {
+      "epoch": 186.11,
+      "eval_loss": 0.5261973142623901,
+      "eval_micro_f1": 0.604432504543678,
+      "eval_runtime": 2.966,
+      "eval_samples_per_second": 340.194,
+      "step": 5025
+    },
+    {
+      "epoch": 186.67,
+      "eval_loss": 0.5327093005180359,
+      "eval_micro_f1": 0.6091644199923787,
+      "eval_runtime": 2.917,
+      "eval_samples_per_second": 345.908,
+      "step": 5040
+    },
+    {
+      "epoch": 187.22,
+      "eval_loss": 0.5286650061607361,
+      "eval_micro_f1": 0.6099195705520363,
+      "eval_runtime": 5.3656,
+      "eval_samples_per_second": 188.051,
+      "step": 5055
+    },
+    {
+      "epoch": 187.78,
+      "eval_loss": 0.5314157605171204,
+      "eval_micro_f1": 0.6091644199923787,
+      "eval_runtime": 3.2366,
+      "eval_samples_per_second": 311.744,
+      "step": 5070
+    },
+    {
+      "epoch": 188.33,
+      "eval_loss": 0.5322765707969666,
+      "eval_micro_f1": 0.6098715343285145,
+      "eval_runtime": 2.8334,
+      "eval_samples_per_second": 356.112,
+      "step": 5085
+    },
+    {
+      "epoch": 188.89,
+      "eval_loss": 0.5415607690811157,
+      "eval_micro_f1": 0.6137135093521486,
+      "eval_runtime": 2.6852,
+      "eval_samples_per_second": 375.758,
+      "step": 5100
+    },
+    {
+      "epoch": 189.44,
+      "eval_loss": 0.5384184122085571,
+      "eval_micro_f1": 0.6120514551612378,
+      "eval_runtime": 2.7928,
+      "eval_samples_per_second": 361.291,
+      "step": 5115
+    },
+    {
+      "epoch": 190.0,
+      "eval_loss": 0.5397367477416992,
+      "eval_micro_f1": 0.6106973590204137,
+      "eval_runtime": 2.7785,
+      "eval_samples_per_second": 363.142,
+      "step": 5130
+    },
+    {
+      "epoch": 190.56,
+      "eval_loss": 0.5299333930015564,
+      "eval_micro_f1": 0.6110363386725561,
+      "eval_runtime": 2.7555,
+      "eval_samples_per_second": 366.177,
+      "step": 5145
+    },
+    {
+      "epoch": 191.11,
+      "eval_loss": 0.5294369459152222,
+      "eval_micro_f1": 0.6082265672753385,
+      "eval_runtime": 2.7952,
+      "eval_samples_per_second": 360.976,
+      "step": 5160
+    },
+    {
+      "epoch": 191.67,
+      "eval_loss": 0.5222564935684204,
+      "eval_micro_f1": 0.6082887695595831,
+      "eval_runtime": 2.8356,
+      "eval_samples_per_second": 355.833,
+      "step": 5175
+    },
+    {
+      "epoch": 192.22,
+      "eval_loss": 0.5004547834396362,
+      "eval_micro_f1": 0.6116129027281149,
+      "eval_runtime": 2.8929,
+      "eval_samples_per_second": 348.781,
+      "step": 5190
+    },
+    {
+      "epoch": 192.78,
+      "eval_loss": 0.5099601149559021,
+      "eval_micro_f1": 0.6065359472159323,
+      "eval_runtime": 2.7484,
+      "eval_samples_per_second": 367.117,
+      "step": 5205
+    },
+    {
+      "epoch": 193.33,
+      "eval_loss": 0.5260539650917053,
+      "eval_micro_f1": 0.6132264524118253,
+      "eval_runtime": 5.6121,
+      "eval_samples_per_second": 179.791,
+      "step": 5220
+    },
+    {
+      "epoch": 193.89,
+      "eval_loss": 0.5357444882392883,
+      "eval_micro_f1": 0.6129032253132731,
+      "eval_runtime": 2.8606,
+      "eval_samples_per_second": 352.717,
+      "step": 5235
+    },
+    {
+      "epoch": 194.44,
+      "eval_loss": 0.5295007228851318,
+      "eval_micro_f1": 0.6124916046108264,
+      "eval_runtime": 2.6872,
+      "eval_samples_per_second": 375.479,
+      "step": 5250
+    },
+    {
+      "epoch": 195.0,
+      "eval_loss": 0.5287444591522217,
+      "eval_micro_f1": 0.6133153996414132,
+      "eval_runtime": 2.9186,
+      "eval_samples_per_second": 345.711,
+      "step": 5265
+    },
+    {
+      "epoch": 195.56,
+      "eval_loss": 0.5251892805099487,
+      "eval_micro_f1": 0.6106254198162655,
+      "eval_runtime": 2.9346,
+      "eval_samples_per_second": 343.833,
+      "step": 5280
+    },
+    {
+      "epoch": 196.11,
+      "eval_loss": 0.5428449511528015,
+      "eval_micro_f1": 0.6106973590204137,
+      "eval_runtime": 2.9111,
+      "eval_samples_per_second": 346.607,
+      "step": 5295
+    },
+    {
+      "epoch": 196.67,
+      "eval_loss": 0.5459561347961426,
+      "eval_micro_f1": 0.6151761512695008,
+      "eval_runtime": 2.7987,
+      "eval_samples_per_second": 360.529,
+      "step": 5310
+    },
+    {
+      "epoch": 197.22,
+      "eval_loss": 0.5328717827796936,
+      "eval_micro_f1": 0.610215053270266,
+      "eval_runtime": 2.7191,
+      "eval_samples_per_second": 371.085,
+      "step": 5325
+    },
+    {
+      "epoch": 197.78,
+      "eval_loss": 0.5316534042358398,
+      "eval_micro_f1": 0.6056149727681427,
+      "eval_runtime": 2.7493,
+      "eval_samples_per_second": 367.006,
+      "step": 5340
+    },
+    {
+      "epoch": 198.33,
+      "eval_loss": 0.5266748666763306,
+      "eval_micro_f1": 0.6061415215353092,
+      "eval_runtime": 2.8671,
+      "eval_samples_per_second": 351.925,
+      "step": 5355
+    },
+    {
+      "epoch": 198.89,
+      "eval_loss": 0.5218961834907532,
+      "eval_micro_f1": 0.6084056032416719,
+      "eval_runtime": 2.8139,
+      "eval_samples_per_second": 358.578,
+      "step": 5370
+    },
+    {
+      "epoch": 199.44,
+      "eval_loss": 0.5269371271133423,
+      "eval_micro_f1": 0.6061415215353092,
+      "eval_runtime": 2.7467,
+      "eval_samples_per_second": 367.355,
+      "step": 5385
+    },
+    {
+      "epoch": 200.0,
+      "eval_loss": 0.5311424136161804,
+      "eval_micro_f1": 0.6091030784888791,
+      "eval_runtime": 2.8756,
+      "eval_samples_per_second": 350.88,
+      "step": 5400
+    },
+    {
+      "epoch": 200.56,
+      "eval_loss": 0.534724235534668,
+      "eval_micro_f1": 0.6101010096081171,
+      "eval_runtime": 2.7988,
+      "eval_samples_per_second": 360.517,
+      "step": 5415
+    },
+    {
+      "epoch": 201.11,
+      "eval_loss": 0.5386430025100708,
+      "eval_micro_f1": 0.6072874489001112,
+      "eval_runtime": 2.8706,
+      "eval_samples_per_second": 351.499,
+      "step": 5430
+    },
+    {
+      "epoch": 201.67,
+      "eval_loss": 0.5445970296859741,
+      "eval_micro_f1": 0.6086956516823162,
+      "eval_runtime": 3.0976,
+      "eval_samples_per_second": 325.733,
+      "step": 5445
+    },
+    {
+      "epoch": 202.22,
+      "eval_loss": 0.5676397085189819,
+      "eval_micro_f1": 0.6073153894350898,
+      "eval_runtime": 2.8475,
+      "eval_samples_per_second": 354.352,
+      "step": 5460
+    },
+    {
+      "epoch": 202.78,
+      "eval_loss": 0.5405770540237427,
+      "eval_micro_f1": 0.6113360318960572,
+      "eval_runtime": 2.7437,
+      "eval_samples_per_second": 367.747,
+      "step": 5475
+    },
+    {
+      "epoch": 203.33,
+      "eval_loss": 0.5431000590324402,
+      "eval_micro_f1": 0.6117488178734589,
+      "eval_runtime": 2.9178,
+      "eval_samples_per_second": 345.806,
+      "step": 5490
+    },
+    {
+      "epoch": 203.7,
+      "learning_rate": 3.2098765432098767e-06,
+      "loss": 0.001,
+      "step": 5500
+    },
+    {
+      "epoch": 203.89,
+      "eval_loss": 0.5510136485099792,
+      "eval_micro_f1": 0.608282416344688,
+      "eval_runtime": 2.7988,
+      "eval_samples_per_second": 360.507,
+      "step": 5505
+    },
+    {
+      "epoch": 204.44,
+      "eval_loss": 0.5503948330879211,
+      "eval_micro_f1": 0.608282416344688,
+      "eval_runtime": 2.8343,
+      "eval_samples_per_second": 355.998,
+      "step": 5520
+    },
+    {
+      "epoch": 205.0,
+      "eval_loss": 0.5440623164176941,
+      "eval_micro_f1": 0.6121621616697498,
+      "eval_runtime": 2.8685,
+      "eval_samples_per_second": 351.754,
+      "step": 5535
+    },
+    {
+      "epoch": 205.56,
+      "eval_loss": 0.5460735559463501,
+      "eval_micro_f1": 0.6103983789808207,
+      "eval_runtime": 2.7631,
+      "eval_samples_per_second": 365.169,
+      "step": 5550
+    },
+    {
+      "epoch": 206.11,
+      "eval_loss": 0.49798014760017395,
+      "eval_micro_f1": 0.6099744240540339,
+      "eval_runtime": 2.9419,
+      "eval_samples_per_second": 342.977,
+      "step": 5565
+    },
+    {
+      "epoch": 206.67,
+      "eval_loss": 0.5046055316925049,
+      "eval_micro_f1": 0.614583332836453,
+      "eval_runtime": 2.7211,
+      "eval_samples_per_second": 370.812,
+      "step": 5580
+    },
+    {
+      "epoch": 207.22,
+      "eval_loss": 0.5200738906860352,
+      "eval_micro_f1": 0.6174318025660456,
+      "eval_runtime": 2.7568,
+      "eval_samples_per_second": 366.005,
+      "step": 5595
+    },
+    {
+      "epoch": 207.78,
+      "eval_loss": 0.5331651568412781,
+      "eval_micro_f1": 0.6153846148920032,
+      "eval_runtime": 2.9518,
+      "eval_samples_per_second": 341.825,
+      "step": 5610
+    },
+    {
+      "epoch": 208.33,
+      "eval_loss": 0.5340650081634521,
+      "eval_micro_f1": 0.6128813554402914,
+      "eval_runtime": 3.5266,
+      "eval_samples_per_second": 286.107,
+      "step": 5625
+    },
+    {
+      "epoch": 208.89,
+      "eval_loss": 0.5031038522720337,
+      "eval_micro_f1": 0.6209150321832339,
+      "eval_runtime": 2.9438,
+      "eval_samples_per_second": 342.757,
+      "step": 5640
+    },
+    {
+      "epoch": 209.44,
+      "eval_loss": 0.5017216801643372,
+      "eval_micro_f1": 0.6220472435983874,
+      "eval_runtime": 2.8041,
+      "eval_samples_per_second": 359.834,
+      "step": 5655
+    },
+    {
+      "epoch": 210.0,
+      "eval_loss": 0.5109838843345642,
+      "eval_micro_f1": 0.6148738374866659,
+      "eval_runtime": 2.7124,
+      "eval_samples_per_second": 371.994,
+      "step": 5670
+    },
+    {
+      "epoch": 210.56,
+      "eval_loss": 0.5132467746734619,
+      "eval_micro_f1": 0.6147704585873398,
+      "eval_runtime": 2.6306,
+      "eval_samples_per_second": 383.565,
+      "step": 5685
+    },
+    {
+      "epoch": 211.11,
+      "eval_loss": 0.5432448983192444,
+      "eval_micro_f1": 0.6152796720874781,
+      "eval_runtime": 2.8488,
+      "eval_samples_per_second": 354.183,
+      "step": 5700
+    },
+    {
+      "epoch": 211.67,
+      "eval_loss": 0.5429700613021851,
+      "eval_micro_f1": 0.6130790185823898,
+      "eval_runtime": 3.3425,
+      "eval_samples_per_second": 301.87,
+      "step": 5715
+    },
+    {
+      "epoch": 212.22,
+      "eval_loss": 0.49836409091949463,
+      "eval_micro_f1": 0.613071894928342,
+      "eval_runtime": 2.824,
+      "eval_samples_per_second": 357.295,
+      "step": 5730
+    },
+    {
+      "epoch": 212.78,
+      "eval_loss": 0.4905017018318176,
+      "eval_micro_f1": 0.6117192525607442,
+      "eval_runtime": 2.6924,
+      "eval_samples_per_second": 374.756,
+      "step": 5745
+    },
+    {
+      "epoch": 213.33,
+      "eval_loss": 0.5017233490943909,
+      "eval_micro_f1": 0.6152841275243385,
+      "eval_runtime": 2.7135,
+      "eval_samples_per_second": 371.85,
+      "step": 5760
+    },
+    {
+      "epoch": 213.89,
+      "eval_loss": 0.5155999660491943,
+      "eval_micro_f1": 0.6169220514710237,
+      "eval_runtime": 2.8442,
+      "eval_samples_per_second": 354.757,
+      "step": 5775
+    },
+    {
+      "epoch": 214.44,
+      "eval_loss": 0.5244033932685852,
+      "eval_micro_f1": 0.6196236554207911,
+      "eval_runtime": 6.0995,
+      "eval_samples_per_second": 165.424,
+      "step": 5790
+    },
+    {
+      "epoch": 215.0,
+      "eval_loss": 0.5281329154968262,
+      "eval_micro_f1": 0.6147651001777756,
+      "eval_runtime": 2.8456,
+      "eval_samples_per_second": 354.576,
+      "step": 5805
+    },
+    {
+      "epoch": 215.56,
+      "eval_loss": 0.5226531028747559,
+      "eval_micro_f1": 0.6147704585873398,
+      "eval_runtime": 2.8238,
+      "eval_samples_per_second": 357.314,
+      "step": 5820
+    },
+    {
+      "epoch": 216.11,
+      "eval_loss": 0.5318886637687683,
+      "eval_micro_f1": 0.6137281287129295,
+      "eval_runtime": 2.8198,
+      "eval_samples_per_second": 357.829,
+      "step": 5835
+    },
+    {
+      "epoch": 216.67,
+      "eval_loss": 0.5341318845748901,
+      "eval_micro_f1": 0.6099865042307419,
+      "eval_runtime": 3.2913,
+      "eval_samples_per_second": 306.565,
+      "step": 5850
+    },
+    {
+      "epoch": 217.22,
+      "eval_loss": 0.5242801308631897,
+      "eval_micro_f1": 0.6139410182732105,
+      "eval_runtime": 2.8593,
+      "eval_samples_per_second": 352.884,
+      "step": 5865
+    },
+    {
+      "epoch": 217.78,
+      "eval_loss": 0.5099356174468994,
+      "eval_micro_f1": 0.6134564638844324,
+      "eval_runtime": 2.6655,
+      "eval_samples_per_second": 378.547,
+      "step": 5880
+    },
+    {
+      "epoch": 218.33,
+      "eval_loss": 0.5128466486930847,
+      "eval_micro_f1": 0.6128391788562884,
+      "eval_runtime": 2.9234,
+      "eval_samples_per_second": 345.143,
+      "step": 5895
+    },
+    {
+      "epoch": 218.89,
+      "eval_loss": 0.5132299065589905,
+      "eval_micro_f1": 0.617218542551295,
+      "eval_runtime": 2.717,
+      "eval_samples_per_second": 371.366,
+      "step": 5910
+    },
+    {
+      "epoch": 219.44,
+      "eval_loss": 0.5213440656661987,
+      "eval_micro_f1": 0.6144578308315761,
+      "eval_runtime": 3.0885,
+      "eval_samples_per_second": 326.701,
+      "step": 5925
+    },
+    {
+      "epoch": 220.0,
+      "eval_loss": 0.5245106816291809,
+      "eval_micro_f1": 0.6115591392917696,
+      "eval_runtime": 2.7893,
+      "eval_samples_per_second": 361.738,
+      "step": 5940
+    },
+    {
+      "epoch": 220.56,
+      "eval_loss": 0.5291798710823059,
+      "eval_micro_f1": 0.6132075466770094,
+      "eval_runtime": 2.9637,
+      "eval_samples_per_second": 340.457,
+      "step": 5955
+    },
+    {
+      "epoch": 221.11,
+      "eval_loss": 0.5263174772262573,
+      "eval_micro_f1": 0.6080536907818116,
+      "eval_runtime": 2.7229,
+      "eval_samples_per_second": 370.555,
+      "step": 5970
+    },
+    {
+      "epoch": 221.67,
+      "eval_loss": 0.54996258020401,
+      "eval_micro_f1": 0.6146475012207442,
+      "eval_runtime": 3.0924,
+      "eval_samples_per_second": 326.279,
+      "step": 5985
+    },
+    {
+      "epoch": 222.22,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.001,
+      "step": 6000
+    },
+    {
+      "epoch": 222.22,
+      "eval_loss": 0.5206257700920105,
+      "eval_micro_f1": 0.6056149727681427,
+      "eval_runtime": 2.8278,
+      "eval_samples_per_second": 356.813,
+      "step": 6000
+    },
+    {
+      "epoch": 222.78,
+      "eval_loss": 0.513648509979248,
+      "eval_micro_f1": 0.6073089696050171,
+      "eval_runtime": 3.1364,
+      "eval_samples_per_second": 321.706,
+      "step": 6015
+    },
+    {
+      "epoch": 223.33,
+      "eval_loss": 0.5244899392127991,
+      "eval_micro_f1": 0.6072386054045867,
+      "eval_runtime": 2.7902,
+      "eval_samples_per_second": 361.618,
+      "step": 6030
+    },
+    {
+      "epoch": 223.89,
+      "eval_loss": 0.5400967001914978,
+      "eval_micro_f1": 0.6100543473344884,
+      "eval_runtime": 2.8982,
+      "eval_samples_per_second": 348.148,
+      "step": 6045
+    },
+    {
+      "epoch": 224.44,
+      "eval_loss": 0.536340594291687,
+      "eval_micro_f1": 0.6067567562643518,
+      "eval_runtime": 2.895,
+      "eval_samples_per_second": 348.536,
+      "step": 6060
+    },
+    {
+      "epoch": 225.0,
+      "eval_loss": 0.5387406349182129,
+      "eval_micro_f1": 0.6081081076157012,
+      "eval_runtime": 2.9474,
+      "eval_samples_per_second": 342.333,
+      "step": 6075
+    },
+    {
+      "epoch": 225.56,
+      "eval_loss": 0.5473647117614746,
+      "eval_micro_f1": 0.6050305909429112,
+      "eval_runtime": 2.8255,
+      "eval_samples_per_second": 357.1,
+      "step": 6090
+    },
+    {
+      "epoch": 226.11,
+      "eval_loss": 0.539093554019928,
+      "eval_micro_f1": 0.6071670042406209,
+      "eval_runtime": 2.737,
+      "eval_samples_per_second": 368.651,
+      "step": 6105
+    },
+    {
+      "epoch": 226.67,
+      "eval_loss": 0.5333505868911743,
+      "eval_micro_f1": 0.6069986536119963,
+      "eval_runtime": 2.9349,
+      "eval_samples_per_second": 343.793,
+      "step": 6120
+    },
+    {
+      "epoch": 227.22,
+      "eval_loss": 0.5280427932739258,
+      "eval_micro_f1": 0.6068318816229181,
+      "eval_runtime": 2.8239,
+      "eval_samples_per_second": 357.309,
+      "step": 6135
+    },
+    {
+      "epoch": 227.78,
+      "eval_loss": 0.5226674675941467,
+      "eval_micro_f1": 0.6054557546618698,
+      "eval_runtime": 2.9059,
+      "eval_samples_per_second": 347.23,
+      "step": 6150
+    },
+    {
+      "epoch": 228.33,
+      "eval_loss": 0.5279316902160645,
+      "eval_micro_f1": 0.6048387091842518,
+      "eval_runtime": 2.8044,
+      "eval_samples_per_second": 359.789,
+      "step": 6165
+    },
+    {
+      "epoch": 228.89,
+      "eval_loss": 0.5393856763839722,
+      "eval_micro_f1": 0.6094594589670507,
+      "eval_runtime": 2.8813,
+      "eval_samples_per_second": 350.191,
+      "step": 6180
+    },
+    {
+      "epoch": 229.44,
+      "eval_loss": 0.5337560772895813,
+      "eval_micro_f1": 0.606469002202625,
+      "eval_runtime": 2.8228,
+      "eval_samples_per_second": 357.447,
+      "step": 6195
+    },
+    {
+      "epoch": 230.0,
+      "eval_loss": 0.5304192304611206,
+      "eval_micro_f1": 0.6092804299036916,
+      "eval_runtime": 2.8402,
+      "eval_samples_per_second": 355.259,
+      "step": 6210
+    },
+    {
+      "epoch": 230.56,
+      "eval_loss": 0.533469021320343,
+      "eval_micro_f1": 0.6082265672753385,
+      "eval_runtime": 2.8616,
+      "eval_samples_per_second": 352.599,
+      "step": 6225
+    },
+    {
+      "epoch": 231.11,
+      "eval_loss": 0.534256637096405,
+      "eval_micro_f1": 0.6086369765654265,
+      "eval_runtime": 2.6352,
+      "eval_samples_per_second": 382.897,
+      "step": 6240
+    },
+    {
+      "epoch": 231.67,
+      "eval_loss": 0.535795271396637,
+      "eval_micro_f1": 0.6081081076157012,
+      "eval_runtime": 2.7937,
+      "eval_samples_per_second": 361.168,
+      "step": 6255
+    },
+    {
+      "epoch": 232.22,
+      "eval_loss": 0.5363573431968689,
+      "eval_micro_f1": 0.6059379212347961,
+      "eval_runtime": 2.7109,
+      "eval_samples_per_second": 372.202,
+      "step": 6270
+    },
+    {
+      "epoch": 232.78,
+      "eval_loss": 0.5361554026603699,
+      "eval_micro_f1": 0.606469002202625,
+      "eval_runtime": 2.7052,
+      "eval_samples_per_second": 372.983,
+      "step": 6285
+    },
+    {
+      "epoch": 233.33,
+      "eval_loss": 0.538044810295105,
+      "eval_micro_f1": 0.6090479400881825,
+      "eval_runtime": 2.9592,
+      "eval_samples_per_second": 340.969,
+      "step": 6300
+    },
+    {
+      "epoch": 233.89,
+      "eval_loss": 0.5425122976303101,
+      "eval_micro_f1": 0.6115254232369034,
+      "eval_runtime": 2.6517,
+      "eval_samples_per_second": 380.517,
+      "step": 6315
+    },
+    {
+      "epoch": 234.44,
+      "eval_loss": 0.5380368828773499,
+      "eval_micro_f1": 0.6094594589670507,
+      "eval_runtime": 2.8201,
+      "eval_samples_per_second": 357.793,
+      "step": 6330
+    },
+    {
+      "epoch": 235.0,
+      "eval_loss": 0.5456990599632263,
+      "eval_micro_f1": 0.6139283294603548,
+      "eval_runtime": 2.8872,
+      "eval_samples_per_second": 349.469,
+      "step": 6345
+    },
+    {
+      "epoch": 235.56,
+      "eval_loss": 0.5452710390090942,
+      "eval_micro_f1": 0.6120514551612378,
+      "eval_runtime": 2.8711,
+      "eval_samples_per_second": 351.43,
+      "step": 6360
+    },
+    {
+      "epoch": 236.11,
+      "eval_loss": 0.5367301106452942,
+      "eval_micro_f1": 0.6109238026091178,
+      "eval_runtime": 2.6767,
+      "eval_samples_per_second": 376.958,
+      "step": 6375
+    },
+    {
+      "epoch": 236.67,
+      "eval_loss": 0.5360122919082642,
+      "eval_micro_f1": 0.6101010096081171,
+      "eval_runtime": 2.6853,
+      "eval_samples_per_second": 375.756,
+      "step": 6390
+    },
+    {
+      "epoch": 237.22,
+      "eval_loss": 0.5445055365562439,
+      "eval_micro_f1": 0.6129905272479743,
+      "eval_runtime": 2.8106,
+      "eval_samples_per_second": 359.001,
+      "step": 6405
+    },
+    {
+      "epoch": 237.78,
+      "eval_loss": 0.5479749441146851,
+      "eval_micro_f1": 0.6124661241692334,
+      "eval_runtime": 2.7604,
+      "eval_samples_per_second": 365.531,
+      "step": 6420
+    },
+    {
+      "epoch": 238.33,
+      "eval_loss": 0.5343855023384094,
+      "eval_micro_f1": 0.613011401250342,
+      "eval_runtime": 2.8294,
+      "eval_samples_per_second": 356.607,
+      "step": 6435
+    },
+    {
+      "epoch": 238.89,
+      "eval_loss": 0.5311406254768372,
+      "eval_micro_f1": 0.6153846148908009,
+      "eval_runtime": 2.8819,
+      "eval_samples_per_second": 350.114,
+      "step": 6450
+    },
+    {
+      "epoch": 239.44,
+      "eval_loss": 0.5145896077156067,
+      "eval_micro_f1": 0.614780902054256,
+      "eval_runtime": 2.6855,
+      "eval_samples_per_second": 375.715,
+      "step": 6465
+    },
+    {
+      "epoch": 240.0,
+      "eval_loss": 0.5146639943122864,
+      "eval_micro_f1": 0.61467889858634,
+      "eval_runtime": 3.1305,
+      "eval_samples_per_second": 322.309,
+      "step": 6480
+    },
+    {
+      "epoch": 240.56,
+      "eval_loss": 0.5215977430343628,
+      "eval_micro_f1": 0.6120290808001523,
+      "eval_runtime": 2.8584,
+      "eval_samples_per_second": 352.992,
+      "step": 6495
+    },
+    {
+      "epoch": 240.74,
+      "learning_rate": 1.9753086419753087e-06,
+      "loss": 0.0009,
+      "step": 6500
+    },
+    {
+      "epoch": 241.11,
+      "eval_loss": 0.5249710083007812,
+      "eval_micro_f1": 0.613848201902392,
+      "eval_runtime": 2.7447,
+      "eval_samples_per_second": 367.615,
+      "step": 6510
+    },
+    {
+      "epoch": 241.67,
+      "eval_loss": 0.5258700251579285,
+      "eval_micro_f1": 0.6146666661724214,
+      "eval_runtime": 2.8187,
+      "eval_samples_per_second": 357.963,
+      "step": 6525
+    },
+    {
+      "epoch": 242.22,
+      "eval_loss": 0.5263156294822693,
+      "eval_micro_f1": 0.6132264524118253,
+      "eval_runtime": 2.7948,
+      "eval_samples_per_second": 361.026,
+      "step": 6540
+    },
+    {
+      "epoch": 242.78,
+      "eval_loss": 0.5248748660087585,
+      "eval_micro_f1": 0.6137424945025052,
+      "eval_runtime": 3.1257,
+      "eval_samples_per_second": 322.805,
+      "step": 6555
+    },
+    {
+      "epoch": 243.33,
+      "eval_loss": 0.5191696286201477,
+      "eval_micro_f1": 0.6107072037347211,
+      "eval_runtime": 2.6808,
+      "eval_samples_per_second": 376.383,
+      "step": 6570
+    },
+    {
+      "epoch": 243.89,
+      "eval_loss": 0.5193530917167664,
+      "eval_micro_f1": 0.6124338619386512,
+      "eval_runtime": 2.7311,
+      "eval_samples_per_second": 369.45,
+      "step": 6585
+    },
+    {
+      "epoch": 244.44,
+      "eval_loss": 0.5190628170967102,
+      "eval_micro_f1": 0.6165016496695722,
+      "eval_runtime": 2.6973,
+      "eval_samples_per_second": 374.078,
+      "step": 6600
+    },
+    {
+      "epoch": 245.0,
+      "eval_loss": 0.5182846188545227,
+      "eval_micro_f1": 0.614775725098151,
+      "eval_runtime": 3.1251,
+      "eval_samples_per_second": 322.87,
+      "step": 6615
+    },
+    {
+      "epoch": 245.56,
+      "eval_loss": 0.5207036137580872,
+      "eval_micro_f1": 0.6182298541941929,
+      "eval_runtime": 2.764,
+      "eval_samples_per_second": 365.057,
+      "step": 6630
+    },
+    {
+      "epoch": 246.11,
+      "eval_loss": 0.5208626389503479,
+      "eval_micro_f1": 0.6187335087393067,
+      "eval_runtime": 2.767,
+      "eval_samples_per_second": 364.653,
+      "step": 6645
+    },
+    {
+      "epoch": 246.67,
+      "eval_loss": 0.5178468227386475,
+      "eval_micro_f1": 0.612352167703789,
+      "eval_runtime": 2.7864,
+      "eval_samples_per_second": 362.111,
+      "step": 6660
+    },
+    {
+      "epoch": 247.22,
+      "eval_loss": 0.5082351565361023,
+      "eval_micro_f1": 0.6095979242757852,
+      "eval_runtime": 2.8302,
+      "eval_samples_per_second": 356.51,
+      "step": 6675
+    },
+    {
+      "epoch": 247.78,
+      "eval_loss": 0.5035471320152283,
+      "eval_micro_f1": 0.6081342796831293,
+      "eval_runtime": 2.8265,
+      "eval_samples_per_second": 356.972,
+      "step": 6690
+    },
+    {
+      "epoch": 248.33,
+      "eval_loss": 0.5141299366950989,
+      "eval_micro_f1": 0.6166007900181675,
+      "eval_runtime": 2.6406,
+      "eval_samples_per_second": 382.111,
+      "step": 6705
+    },
+    {
+      "epoch": 248.89,
+      "eval_loss": 0.5190057754516602,
+      "eval_micro_f1": 0.6181336858053217,
+      "eval_runtime": 2.763,
+      "eval_samples_per_second": 365.183,
+      "step": 6720
+    },
+    {
+      "epoch": 249.44,
+      "eval_loss": 0.5207487940788269,
+      "eval_micro_f1": 0.6197743857029152,
+      "eval_runtime": 2.915,
+      "eval_samples_per_second": 346.138,
+      "step": 6735
+    },
+    {
+      "epoch": 250.0,
+      "eval_loss": 0.5243783593177795,
+      "eval_micro_f1": 0.6190793857633389,
+      "eval_runtime": 2.9211,
+      "eval_samples_per_second": 345.415,
+      "step": 6750
+    },
+    {
+      "epoch": 250.56,
+      "eval_loss": 0.5288082957267761,
+      "eval_micro_f1": 0.6175485594527764,
+      "eval_runtime": 2.9863,
+      "eval_samples_per_second": 337.876,
+      "step": 6765
+    },
+    {
+      "epoch": 251.11,
+      "eval_loss": 0.5287660360336304,
+      "eval_micro_f1": 0.6157965189172504,
+      "eval_runtime": 3.2231,
+      "eval_samples_per_second": 313.055,
+      "step": 6780
+    },
+    {
+      "epoch": 251.67,
+      "eval_loss": 0.5403092503547668,
+      "eval_micro_f1": 0.6189189184264975,
+      "eval_runtime": 2.7036,
+      "eval_samples_per_second": 373.204,
+      "step": 6795
+    },
+    {
+      "epoch": 252.22,
+      "eval_loss": 0.5443710684776306,
+      "eval_micro_f1": 0.6151761512695008,
+      "eval_runtime": 2.9158,
+      "eval_samples_per_second": 346.05,
+      "step": 6810
+    },
+    {
+      "epoch": 252.78,
+      "eval_loss": 0.531115710735321,
+      "eval_micro_f1": 0.6203208551210643,
+      "eval_runtime": 2.7013,
+      "eval_samples_per_second": 373.522,
+      "step": 6825
+    },
+    {
+      "epoch": 253.33,
+      "eval_loss": 0.5341444611549377,
+      "eval_micro_f1": 0.6211512712599473,
+      "eval_runtime": 2.7801,
+      "eval_samples_per_second": 362.943,
+      "step": 6840
+    },
+    {
+      "epoch": 253.89,
+      "eval_loss": 0.5271020531654358,
+      "eval_micro_f1": 0.6239999995057423,
+      "eval_runtime": 2.6391,
+      "eval_samples_per_second": 382.324,
+      "step": 6855
+    }
+  ],
+  "max_steps": 8100,
+  "num_train_epochs": 300,
+  "total_flos": 37983401679245100,
+  "trial_name": null,
+  "trial_params": null
+}