Add files using upload-large-folder tool

ood-vigoemotions-phobert-source-source-00-s42-20260826T130931Z-0c086fe0/trainer/checkpoint-3672/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

ood-vigoemotions-phobert-source-source-00-s42-20260826T130931Z-0c086fe0/trainer/checkpoint-3672/special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-00-s42-20260826T130931Z-0c086fe0/trainer/checkpoint-3672/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 256,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-00-s42-20260826T130931Z-0c086fe0/trainer/checkpoint-3672/trainer_state.json

@@ -0,0 +1,1281 @@
+{
+  "best_global_step": 3672,
+  "best_metric": 0.5437383183547814,
+  "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vigoemotions-phobert-source-source-00-s42-20260826T130931Z-0c086fe0/trainer/checkpoint-3672",
+  "epoch": 18.0,
+  "eval_steps": 500,
+  "global_step": 3672,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.12300123001230012,
+      "grad_norm": 5.347543716430664,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 5.0951,
+      "step": 25
+    },
+    {
+      "epoch": 0.24600246002460024,
+      "grad_norm": 5.039833068847656,
+      "learning_rate": 2.401960784313726e-06,
+      "loss": 5.1964,
+      "step": 50
+    },
+    {
+      "epoch": 0.36900369003690037,
+      "grad_norm": 4.420496463775635,
+      "learning_rate": 3.6274509803921573e-06,
+      "loss": 5.1619,
+      "step": 75
+    },
+    {
+      "epoch": 0.4920049200492005,
+      "grad_norm": 5.998644828796387,
+      "learning_rate": 4.852941176470589e-06,
+      "loss": 5.228,
+      "step": 100
+    },
+    {
+      "epoch": 0.6150061500615006,
+      "grad_norm": 5.091968536376953,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 5.1265,
+      "step": 125
+    },
+    {
+      "epoch": 0.7380073800738007,
+      "grad_norm": 6.671788215637207,
+      "learning_rate": 7.3039215686274515e-06,
+      "loss": 5.0759,
+      "step": 150
+    },
+    {
+      "epoch": 0.8610086100861009,
+      "grad_norm": 12.5631742477417,
+      "learning_rate": 8.529411764705883e-06,
+      "loss": 5.2254,
+      "step": 175
+    },
+    {
+      "epoch": 0.984009840098401,
+      "grad_norm": 4.14892053604126,
+      "learning_rate": 9.754901960784315e-06,
+      "loss": 4.8142,
+      "step": 200
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 1.1878608465194702,
+      "eval_macro_f1": 0.21320196736877045,
+      "eval_micro_f1": 0.23826145704312576,
+      "eval_runtime": 0.4282,
+      "eval_samples_per_second": 1688.602,
+      "eval_steps_per_second": 107.435,
+      "eval_weighted_f1": 0.2757431924455255,
+      "step": 204
+    },
+    {
+      "epoch": 1.103321033210332,
+      "grad_norm": 5.696617603302002,
+      "learning_rate": 1.0980392156862747e-05,
+      "loss": 4.578,
+      "step": 225
+    },
+    {
+      "epoch": 1.2263222632226323,
+      "grad_norm": 7.200258255004883,
+      "learning_rate": 1.2205882352941179e-05,
+      "loss": 4.6407,
+      "step": 250
+    },
+    {
+      "epoch": 1.3493234932349323,
+      "grad_norm": 7.213965892791748,
+      "learning_rate": 1.3431372549019607e-05,
+      "loss": 4.53,
+      "step": 275
+    },
+    {
+      "epoch": 1.4723247232472325,
+      "grad_norm": 18.122190475463867,
+      "learning_rate": 1.4656862745098039e-05,
+      "loss": 4.4205,
+      "step": 300
+    },
+    {
+      "epoch": 1.5953259532595325,
+      "grad_norm": 10.154242515563965,
+      "learning_rate": 1.5882352941176473e-05,
+      "loss": 4.2928,
+      "step": 325
+    },
+    {
+      "epoch": 1.7183271832718328,
+      "grad_norm": 14.085644721984863,
+      "learning_rate": 1.7107843137254903e-05,
+      "loss": 4.0917,
+      "step": 350
+    },
+    {
+      "epoch": 1.841328413284133,
+      "grad_norm": 7.562018871307373,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 4.1262,
+      "step": 375
+    },
+    {
+      "epoch": 1.964329643296433,
+      "grad_norm": 9.274996757507324,
+      "learning_rate": 1.9558823529411766e-05,
+      "loss": 4.0357,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 0.9473660588264465,
+      "eval_macro_f1": 0.2907416898851504,
+      "eval_micro_f1": 0.30697176502074136,
+      "eval_runtime": 0.4394,
+      "eval_samples_per_second": 1645.561,
+      "eval_steps_per_second": 104.697,
+      "eval_weighted_f1": 0.3731837751611428,
+      "step": 408
+    },
+    {
+      "epoch": 2.0836408364083643,
+      "grad_norm": 5.191472053527832,
+      "learning_rate": 1.9912854030501092e-05,
+      "loss": 3.8519,
+      "step": 425
+    },
+    {
+      "epoch": 2.206642066420664,
+      "grad_norm": 13.798820495605469,
+      "learning_rate": 1.9776688453159043e-05,
+      "loss": 3.618,
+      "step": 450
+    },
+    {
+      "epoch": 2.3296432964329643,
+      "grad_norm": 11.70262336730957,
+      "learning_rate": 1.9640522875816993e-05,
+      "loss": 3.7455,
+      "step": 475
+    },
+    {
+      "epoch": 2.4526445264452645,
+      "grad_norm": 7.336067199707031,
+      "learning_rate": 1.9504357298474947e-05,
+      "loss": 3.787,
+      "step": 500
+    },
+    {
+      "epoch": 2.5756457564575648,
+      "grad_norm": 13.359006881713867,
+      "learning_rate": 1.9368191721132898e-05,
+      "loss": 3.6748,
+      "step": 525
+    },
+    {
+      "epoch": 2.6986469864698646,
+      "grad_norm": 10.397160530090332,
+      "learning_rate": 1.9232026143790852e-05,
+      "loss": 3.6422,
+      "step": 550
+    },
+    {
+      "epoch": 2.821648216482165,
+      "grad_norm": 18.4261531829834,
+      "learning_rate": 1.9095860566448803e-05,
+      "loss": 3.5534,
+      "step": 575
+    },
+    {
+      "epoch": 2.944649446494465,
+      "grad_norm": 7.37200927734375,
+      "learning_rate": 1.8959694989106757e-05,
+      "loss": 3.5361,
+      "step": 600
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 0.8620179295539856,
+      "eval_macro_f1": 0.34796481421021197,
+      "eval_micro_f1": 0.3669754591256298,
+      "eval_runtime": 0.4439,
+      "eval_samples_per_second": 1628.883,
+      "eval_steps_per_second": 103.636,
+      "eval_weighted_f1": 0.4245846547213332,
+      "step": 612
+    },
+    {
+      "epoch": 3.063960639606396,
+      "grad_norm": 7.358486652374268,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 3.3656,
+      "step": 625
+    },
+    {
+      "epoch": 3.1869618696186963,
+      "grad_norm": 7.551677227020264,
+      "learning_rate": 1.8687363834422662e-05,
+      "loss": 3.2034,
+      "step": 650
+    },
+    {
+      "epoch": 3.3099630996309966,
+      "grad_norm": 10.687773704528809,
+      "learning_rate": 1.8551198257080612e-05,
+      "loss": 3.3985,
+      "step": 675
+    },
+    {
+      "epoch": 3.4329643296432963,
+      "grad_norm": 7.113092422485352,
+      "learning_rate": 1.8415032679738563e-05,
+      "loss": 3.24,
+      "step": 700
+    },
+    {
+      "epoch": 3.5559655596555966,
+      "grad_norm": 8.03344440460205,
+      "learning_rate": 1.8278867102396517e-05,
+      "loss": 3.2581,
+      "step": 725
+    },
+    {
+      "epoch": 3.678966789667897,
+      "grad_norm": 6.487556457519531,
+      "learning_rate": 1.8142701525054468e-05,
+      "loss": 3.2576,
+      "step": 750
+    },
+    {
+      "epoch": 3.8019680196801966,
+      "grad_norm": 6.6811699867248535,
+      "learning_rate": 1.800653594771242e-05,
+      "loss": 3.1186,
+      "step": 775
+    },
+    {
+      "epoch": 3.924969249692497,
+      "grad_norm": 16.565784454345703,
+      "learning_rate": 1.7870370370370373e-05,
+      "loss": 3.1422,
+      "step": 800
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.005532503457814661,
+      "eval_loss": 0.794185221195221,
+      "eval_macro_f1": 0.36775263884169107,
+      "eval_micro_f1": 0.3921229918811539,
+      "eval_runtime": 0.4322,
+      "eval_samples_per_second": 1672.868,
+      "eval_steps_per_second": 106.434,
+      "eval_weighted_f1": 0.4494910853177473,
+      "step": 816
+    },
+    {
+      "epoch": 4.044280442804428,
+      "grad_norm": 6.690476417541504,
+      "learning_rate": 1.7734204793028323e-05,
+      "loss": 3.0639,
+      "step": 825
+    },
+    {
+      "epoch": 4.167281672816729,
+      "grad_norm": 7.374212741851807,
+      "learning_rate": 1.7598039215686274e-05,
+      "loss": 2.979,
+      "step": 850
+    },
+    {
+      "epoch": 4.290282902829028,
+      "grad_norm": 6.836591720581055,
+      "learning_rate": 1.7461873638344228e-05,
+      "loss": 2.9576,
+      "step": 875
+    },
+    {
+      "epoch": 4.413284132841328,
+      "grad_norm": 17.42571258544922,
+      "learning_rate": 1.732570806100218e-05,
+      "loss": 2.9641,
+      "step": 900
+    },
+    {
+      "epoch": 4.536285362853628,
+      "grad_norm": 5.6428141593933105,
+      "learning_rate": 1.7189542483660133e-05,
+      "loss": 2.9495,
+      "step": 925
+    },
+    {
+      "epoch": 4.659286592865929,
+      "grad_norm": 5.34227180480957,
+      "learning_rate": 1.7053376906318084e-05,
+      "loss": 2.9363,
+      "step": 950
+    },
+    {
+      "epoch": 4.782287822878229,
+      "grad_norm": 6.289327144622803,
+      "learning_rate": 1.6917211328976034e-05,
+      "loss": 2.9018,
+      "step": 975
+    },
+    {
+      "epoch": 4.905289052890529,
+      "grad_norm": 16.238441467285156,
+      "learning_rate": 1.678104575163399e-05,
+      "loss": 2.8826,
+      "step": 1000
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.02074688796680498,
+      "eval_loss": 0.7590930461883545,
+      "eval_macro_f1": 0.3985394691102169,
+      "eval_micro_f1": 0.43172147001934236,
+      "eval_runtime": 0.4883,
+      "eval_samples_per_second": 1480.617,
+      "eval_steps_per_second": 94.202,
+      "eval_weighted_f1": 0.48269412632767783,
+      "step": 1020
+    },
+    {
+      "epoch": 5.02460024600246,
+      "grad_norm": 9.875899314880371,
+      "learning_rate": 1.664488017429194e-05,
+      "loss": 2.7175,
+      "step": 1025
+    },
+    {
+      "epoch": 5.14760147601476,
+      "grad_norm": 11.602205276489258,
+      "learning_rate": 1.6508714596949893e-05,
+      "loss": 2.6652,
+      "step": 1050
+    },
+    {
+      "epoch": 5.27060270602706,
+      "grad_norm": 8.778997421264648,
+      "learning_rate": 1.6372549019607844e-05,
+      "loss": 2.7818,
+      "step": 1075
+    },
+    {
+      "epoch": 5.39360393603936,
+      "grad_norm": 7.099167346954346,
+      "learning_rate": 1.6236383442265798e-05,
+      "loss": 2.7251,
+      "step": 1100
+    },
+    {
+      "epoch": 5.516605166051661,
+      "grad_norm": 5.344371318817139,
+      "learning_rate": 1.610021786492375e-05,
+      "loss": 2.6626,
+      "step": 1125
+    },
+    {
+      "epoch": 5.639606396063961,
+      "grad_norm": 9.371820449829102,
+      "learning_rate": 1.5964052287581703e-05,
+      "loss": 2.5991,
+      "step": 1150
+    },
+    {
+      "epoch": 5.762607626076261,
+      "grad_norm": 5.649630069732666,
+      "learning_rate": 1.5827886710239654e-05,
+      "loss": 2.6888,
+      "step": 1175
+    },
+    {
+      "epoch": 5.885608856088561,
+      "grad_norm": 6.022938251495361,
+      "learning_rate": 1.5691721132897604e-05,
+      "loss": 2.593,
+      "step": 1200
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.04426002766251729,
+      "eval_loss": 0.7480583786964417,
+      "eval_macro_f1": 0.4246178060686669,
+      "eval_micro_f1": 0.46047098402018505,
+      "eval_runtime": 0.415,
+      "eval_samples_per_second": 1742.003,
+      "eval_steps_per_second": 110.833,
+      "eval_weighted_f1": 0.5006830733841459,
+      "step": 1224
+    },
+    {
+      "epoch": 6.004920049200492,
+      "grad_norm": 13.547445297241211,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 2.5475,
+      "step": 1225
+    },
+    {
+      "epoch": 6.127921279212792,
+      "grad_norm": 10.183258056640625,
+      "learning_rate": 1.541938997821351e-05,
+      "loss": 2.5095,
+      "step": 1250
+    },
+    {
+      "epoch": 6.250922509225092,
+      "grad_norm": 7.828505516052246,
+      "learning_rate": 1.528322440087146e-05,
+      "loss": 2.4239,
+      "step": 1275
+    },
+    {
+      "epoch": 6.373923739237393,
+      "grad_norm": 8.002518653869629,
+      "learning_rate": 1.5147058823529412e-05,
+      "loss": 2.4879,
+      "step": 1300
+    },
+    {
+      "epoch": 6.496924969249693,
+      "grad_norm": 12.85448169708252,
+      "learning_rate": 1.5010893246187366e-05,
+      "loss": 2.4154,
+      "step": 1325
+    },
+    {
+      "epoch": 6.619926199261993,
+      "grad_norm": 5.428205966949463,
+      "learning_rate": 1.4874727668845317e-05,
+      "loss": 2.4019,
+      "step": 1350
+    },
+    {
+      "epoch": 6.7429274292742925,
+      "grad_norm": 10.2520112991333,
+      "learning_rate": 1.473856209150327e-05,
+      "loss": 2.4149,
+      "step": 1375
+    },
+    {
+      "epoch": 6.865928659286593,
+      "grad_norm": 8.956914901733398,
+      "learning_rate": 1.4602396514161222e-05,
+      "loss": 2.3976,
+      "step": 1400
+    },
+    {
+      "epoch": 6.988929889298893,
+      "grad_norm": 7.741453647613525,
+      "learning_rate": 1.4466230936819172e-05,
+      "loss": 2.4199,
+      "step": 1425
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.08298755186721991,
+      "eval_loss": 0.7325209975242615,
+      "eval_macro_f1": 0.43862133890959265,
+      "eval_micro_f1": 0.4823896567097637,
+      "eval_runtime": 0.4143,
+      "eval_samples_per_second": 1744.944,
+      "eval_steps_per_second": 111.02,
+      "eval_weighted_f1": 0.5131973346966094,
+      "step": 1428
+    },
+    {
+      "epoch": 7.108241082410824,
+      "grad_norm": 5.299953937530518,
+      "learning_rate": 1.4330065359477125e-05,
+      "loss": 2.2119,
+      "step": 1450
+    },
+    {
+      "epoch": 7.231242312423125,
+      "grad_norm": 8.74158763885498,
+      "learning_rate": 1.4193899782135077e-05,
+      "loss": 2.3311,
+      "step": 1475
+    },
+    {
+      "epoch": 7.354243542435424,
+      "grad_norm": 10.328490257263184,
+      "learning_rate": 1.405773420479303e-05,
+      "loss": 2.2658,
+      "step": 1500
+    },
+    {
+      "epoch": 7.477244772447724,
+      "grad_norm": 8.569588661193848,
+      "learning_rate": 1.392156862745098e-05,
+      "loss": 2.2812,
+      "step": 1525
+    },
+    {
+      "epoch": 7.6002460024600245,
+      "grad_norm": 7.103965759277344,
+      "learning_rate": 1.3785403050108934e-05,
+      "loss": 2.2201,
+      "step": 1550
+    },
+    {
+      "epoch": 7.723247232472325,
+      "grad_norm": 6.22614049911499,
+      "learning_rate": 1.3649237472766885e-05,
+      "loss": 2.2075,
+      "step": 1575
+    },
+    {
+      "epoch": 7.846248462484625,
+      "grad_norm": 7.187036991119385,
+      "learning_rate": 1.351307189542484e-05,
+      "loss": 2.1669,
+      "step": 1600
+    },
+    {
+      "epoch": 7.969249692496925,
+      "grad_norm": 6.168943405151367,
+      "learning_rate": 1.337690631808279e-05,
+      "loss": 2.2443,
+      "step": 1625
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.10926694329183956,
+      "eval_loss": 0.7492125630378723,
+      "eval_macro_f1": 0.47865805519264376,
+      "eval_micro_f1": 0.5181168350998274,
+      "eval_runtime": 0.4166,
+      "eval_samples_per_second": 1735.553,
+      "eval_steps_per_second": 110.422,
+      "eval_weighted_f1": 0.5433474238239877,
+      "step": 1632
+    },
+    {
+      "epoch": 8.088560885608857,
+      "grad_norm": 6.345441818237305,
+      "learning_rate": 1.3240740740740742e-05,
+      "loss": 2.1594,
+      "step": 1650
+    },
+    {
+      "epoch": 8.211562115621156,
+      "grad_norm": 6.153682708740234,
+      "learning_rate": 1.3104575163398695e-05,
+      "loss": 2.086,
+      "step": 1675
+    },
+    {
+      "epoch": 8.334563345633457,
+      "grad_norm": 6.458537578582764,
+      "learning_rate": 1.2968409586056645e-05,
+      "loss": 2.0523,
+      "step": 1700
+    },
+    {
+      "epoch": 8.457564575645756,
+      "grad_norm": 4.463102340698242,
+      "learning_rate": 1.2832244008714598e-05,
+      "loss": 2.0716,
+      "step": 1725
+    },
+    {
+      "epoch": 8.580565805658056,
+      "grad_norm": 5.9551897048950195,
+      "learning_rate": 1.269607843137255e-05,
+      "loss": 2.0998,
+      "step": 1750
+    },
+    {
+      "epoch": 8.703567035670357,
+      "grad_norm": 12.094724655151367,
+      "learning_rate": 1.2559912854030503e-05,
+      "loss": 2.1279,
+      "step": 1775
+    },
+    {
+      "epoch": 8.826568265682656,
+      "grad_norm": 5.6941142082214355,
+      "learning_rate": 1.2423747276688453e-05,
+      "loss": 2.022,
+      "step": 1800
+    },
+    {
+      "epoch": 8.949569495694957,
+      "grad_norm": 5.646745204925537,
+      "learning_rate": 1.2287581699346407e-05,
+      "loss": 2.0617,
+      "step": 1825
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.09820193637621023,
+      "eval_loss": 0.7172796130180359,
+      "eval_macro_f1": 0.4703039493047723,
+      "eval_micro_f1": 0.5005804504295334,
+      "eval_runtime": 0.4506,
+      "eval_samples_per_second": 1604.522,
+      "eval_steps_per_second": 102.086,
+      "eval_weighted_f1": 0.5364196628370508,
+      "step": 1836
+    },
+    {
+      "epoch": 9.068880688806889,
+      "grad_norm": 6.38439416885376,
+      "learning_rate": 1.2151416122004358e-05,
+      "loss": 2.0243,
+      "step": 1850
+    },
+    {
+      "epoch": 9.191881918819188,
+      "grad_norm": 15.08652400970459,
+      "learning_rate": 1.2015250544662312e-05,
+      "loss": 1.9855,
+      "step": 1875
+    },
+    {
+      "epoch": 9.314883148831488,
+      "grad_norm": 5.886865615844727,
+      "learning_rate": 1.1879084967320263e-05,
+      "loss": 1.9351,
+      "step": 1900
+    },
+    {
+      "epoch": 9.437884378843789,
+      "grad_norm": 5.136407375335693,
+      "learning_rate": 1.1742919389978214e-05,
+      "loss": 1.9242,
+      "step": 1925
+    },
+    {
+      "epoch": 9.560885608856088,
+      "grad_norm": 7.7304911613464355,
+      "learning_rate": 1.1606753812636166e-05,
+      "loss": 1.987,
+      "step": 1950
+    },
+    {
+      "epoch": 9.68388683886839,
+      "grad_norm": 9.386374473571777,
+      "learning_rate": 1.1470588235294118e-05,
+      "loss": 1.9295,
+      "step": 1975
+    },
+    {
+      "epoch": 9.806888068880689,
+      "grad_norm": 6.307365894317627,
+      "learning_rate": 1.133442265795207e-05,
+      "loss": 1.9393,
+      "step": 2000
+    },
+    {
+      "epoch": 9.92988929889299,
+      "grad_norm": 6.248044967651367,
+      "learning_rate": 1.1198257080610021e-05,
+      "loss": 1.9582,
+      "step": 2025
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.13416320885200553,
+      "eval_loss": 0.7009500861167908,
+      "eval_macro_f1": 0.49583988112082567,
+      "eval_micro_f1": 0.5355776587605203,
+      "eval_runtime": 0.4279,
+      "eval_samples_per_second": 1689.492,
+      "eval_steps_per_second": 107.492,
+      "eval_weighted_f1": 0.5556808037515301,
+      "step": 2040
+    },
+    {
+      "epoch": 10.04920049200492,
+      "grad_norm": 6.902281284332275,
+      "learning_rate": 1.1062091503267976e-05,
+      "loss": 1.7956,
+      "step": 2050
+    },
+    {
+      "epoch": 10.17220172201722,
+      "grad_norm": 5.536381244659424,
+      "learning_rate": 1.0925925925925926e-05,
+      "loss": 1.856,
+      "step": 2075
+    },
+    {
+      "epoch": 10.29520295202952,
+      "grad_norm": 3.8086092472076416,
+      "learning_rate": 1.078976034858388e-05,
+      "loss": 1.8511,
+      "step": 2100
+    },
+    {
+      "epoch": 10.418204182041821,
+      "grad_norm": 6.091910362243652,
+      "learning_rate": 1.0653594771241831e-05,
+      "loss": 1.789,
+      "step": 2125
+    },
+    {
+      "epoch": 10.54120541205412,
+      "grad_norm": 5.5678839683532715,
+      "learning_rate": 1.0517429193899782e-05,
+      "loss": 1.8157,
+      "step": 2150
+    },
+    {
+      "epoch": 10.664206642066421,
+      "grad_norm": 5.289937496185303,
+      "learning_rate": 1.0381263616557736e-05,
+      "loss": 1.8265,
+      "step": 2175
+    },
+    {
+      "epoch": 10.78720787207872,
+      "grad_norm": 7.392284393310547,
+      "learning_rate": 1.0245098039215687e-05,
+      "loss": 1.7552,
+      "step": 2200
+    },
+    {
+      "epoch": 10.91020910209102,
+      "grad_norm": 7.5922746658325195,
+      "learning_rate": 1.0108932461873639e-05,
+      "loss": 1.805,
+      "step": 2225
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.16044260027662519,
+      "eval_loss": 0.7213637828826904,
+      "eval_macro_f1": 0.514443422016327,
+      "eval_micro_f1": 0.5513196480938416,
+      "eval_runtime": 0.4384,
+      "eval_samples_per_second": 1649.083,
+      "eval_steps_per_second": 104.921,
+      "eval_weighted_f1": 0.5660369113766512,
+      "step": 2244
+    },
+    {
+      "epoch": 11.029520295202952,
+      "grad_norm": 5.942030906677246,
+      "learning_rate": 9.972766884531591e-06,
+      "loss": 1.7107,
+      "step": 2250
+    },
+    {
+      "epoch": 11.152521525215253,
+      "grad_norm": 6.128445625305176,
+      "learning_rate": 9.836601307189542e-06,
+      "loss": 1.7904,
+      "step": 2275
+    },
+    {
+      "epoch": 11.275522755227552,
+      "grad_norm": 6.908228874206543,
+      "learning_rate": 9.700435729847494e-06,
+      "loss": 1.6947,
+      "step": 2300
+    },
+    {
+      "epoch": 11.398523985239853,
+      "grad_norm": 12.505904197692871,
+      "learning_rate": 9.564270152505447e-06,
+      "loss": 1.7133,
+      "step": 2325
+    },
+    {
+      "epoch": 11.521525215252153,
+      "grad_norm": 4.8953986167907715,
+      "learning_rate": 9.4281045751634e-06,
+      "loss": 1.678,
+      "step": 2350
+    },
+    {
+      "epoch": 11.644526445264452,
+      "grad_norm": 6.927519798278809,
+      "learning_rate": 9.291938997821352e-06,
+      "loss": 1.6724,
+      "step": 2375
+    },
+    {
+      "epoch": 11.767527675276753,
+      "grad_norm": 11.44202709197998,
+      "learning_rate": 9.155773420479304e-06,
+      "loss": 1.7256,
+      "step": 2400
+    },
+    {
+      "epoch": 11.890528905289052,
+      "grad_norm": 6.743433952331543,
+      "learning_rate": 9.019607843137256e-06,
+      "loss": 1.6917,
+      "step": 2425
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.14937759336099585,
+      "eval_loss": 0.7130449414253235,
+      "eval_macro_f1": 0.510854907416043,
+      "eval_micro_f1": 0.5542811416377701,
+      "eval_runtime": 0.4717,
+      "eval_samples_per_second": 1532.598,
+      "eval_steps_per_second": 97.51,
+      "eval_weighted_f1": 0.5733406737839715,
+      "step": 2448
+    },
+    {
+      "epoch": 12.009840098400984,
+      "grad_norm": 9.664470672607422,
+      "learning_rate": 8.883442265795209e-06,
+      "loss": 1.6364,
+      "step": 2450
+    },
+    {
+      "epoch": 12.132841328413285,
+      "grad_norm": 12.487649917602539,
+      "learning_rate": 8.74727668845316e-06,
+      "loss": 1.6667,
+      "step": 2475
+    },
+    {
+      "epoch": 12.255842558425584,
+      "grad_norm": 6.491523265838623,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 1.5484,
+      "step": 2500
+    },
+    {
+      "epoch": 12.378843788437884,
+      "grad_norm": 8.131217956542969,
+      "learning_rate": 8.474945533769063e-06,
+      "loss": 1.6584,
+      "step": 2525
+    },
+    {
+      "epoch": 12.501845018450185,
+      "grad_norm": 6.362199306488037,
+      "learning_rate": 8.338779956427015e-06,
+      "loss": 1.6223,
+      "step": 2550
+    },
+    {
+      "epoch": 12.624846248462484,
+      "grad_norm": 9.135724067687988,
+      "learning_rate": 8.202614379084967e-06,
+      "loss": 1.6026,
+      "step": 2575
+    },
+    {
+      "epoch": 12.747847478474785,
+      "grad_norm": 7.459300518035889,
+      "learning_rate": 8.06644880174292e-06,
+      "loss": 1.5687,
+      "step": 2600
+    },
+    {
+      "epoch": 12.870848708487085,
+      "grad_norm": 5.186668395996094,
+      "learning_rate": 7.930283224400872e-06,
+      "loss": 1.5528,
+      "step": 2625
+    },
+    {
+      "epoch": 12.993849938499386,
+      "grad_norm": 13.283559799194336,
+      "learning_rate": 7.794117647058825e-06,
+      "loss": 1.654,
+      "step": 2650
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.16597510373443983,
+      "eval_loss": 0.6994755268096924,
+      "eval_macro_f1": 0.5197501449610717,
+      "eval_micro_f1": 0.5607123583378305,
+      "eval_runtime": 0.4299,
+      "eval_samples_per_second": 1681.739,
+      "eval_steps_per_second": 106.999,
+      "eval_weighted_f1": 0.5785792673842536,
+      "step": 2652
+    },
+    {
+      "epoch": 13.113161131611315,
+      "grad_norm": 7.693202972412109,
+      "learning_rate": 7.657952069716777e-06,
+      "loss": 1.4981,
+      "step": 2675
+    },
+    {
+      "epoch": 13.236162361623617,
+      "grad_norm": 4.944885730743408,
+      "learning_rate": 7.5217864923747285e-06,
+      "loss": 1.563,
+      "step": 2700
+    },
+    {
+      "epoch": 13.359163591635916,
+      "grad_norm": 7.11841344833374,
+      "learning_rate": 7.385620915032681e-06,
+      "loss": 1.5465,
+      "step": 2725
+    },
+    {
+      "epoch": 13.482164821648217,
+      "grad_norm": 7.9372968673706055,
+      "learning_rate": 7.249455337690633e-06,
+      "loss": 1.5258,
+      "step": 2750
+    },
+    {
+      "epoch": 13.605166051660516,
+      "grad_norm": 5.245449542999268,
+      "learning_rate": 7.113289760348584e-06,
+      "loss": 1.5063,
+      "step": 2775
+    },
+    {
+      "epoch": 13.728167281672818,
+      "grad_norm": 8.229990005493164,
+      "learning_rate": 6.977124183006536e-06,
+      "loss": 1.5031,
+      "step": 2800
+    },
+    {
+      "epoch": 13.851168511685117,
+      "grad_norm": 7.4326653480529785,
+      "learning_rate": 6.840958605664488e-06,
+      "loss": 1.5227,
+      "step": 2825
+    },
+    {
+      "epoch": 13.974169741697416,
+      "grad_norm": 17.769107818603516,
+      "learning_rate": 6.70479302832244e-06,
+      "loss": 1.5404,
+      "step": 2850
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.1839557399723375,
+      "eval_loss": 0.7303460240364075,
+      "eval_macro_f1": 0.5237214993143648,
+      "eval_micro_f1": 0.5703682557807593,
+      "eval_runtime": 0.4195,
+      "eval_samples_per_second": 1723.501,
+      "eval_steps_per_second": 109.656,
+      "eval_weighted_f1": 0.5845632043986984,
+      "step": 2856
+    },
+    {
+      "epoch": 14.093480934809348,
+      "grad_norm": 4.83207368850708,
+      "learning_rate": 6.568627450980393e-06,
+      "loss": 1.3969,
+      "step": 2875
+    },
+    {
+      "epoch": 14.216482164821649,
+      "grad_norm": 8.71630573272705,
+      "learning_rate": 6.432461873638345e-06,
+      "loss": 1.4554,
+      "step": 2900
+    },
+    {
+      "epoch": 14.339483394833948,
+      "grad_norm": 9.311717987060547,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 1.5311,
+      "step": 2925
+    },
+    {
+      "epoch": 14.46248462484625,
+      "grad_norm": 16.771753311157227,
+      "learning_rate": 6.160130718954249e-06,
+      "loss": 1.4538,
+      "step": 2950
+    },
+    {
+      "epoch": 14.585485854858549,
+      "grad_norm": 5.463100433349609,
+      "learning_rate": 6.0239651416122014e-06,
+      "loss": 1.4487,
+      "step": 2975
+    },
+    {
+      "epoch": 14.708487084870848,
+      "grad_norm": 5.630454063415527,
+      "learning_rate": 5.887799564270154e-06,
+      "loss": 1.4717,
+      "step": 3000
+    },
+    {
+      "epoch": 14.83148831488315,
+      "grad_norm": 11.105183601379395,
+      "learning_rate": 5.7516339869281045e-06,
+      "loss": 1.4308,
+      "step": 3025
+    },
+    {
+      "epoch": 14.954489544895448,
+      "grad_norm": 8.745418548583984,
+      "learning_rate": 5.615468409586057e-06,
+      "loss": 1.526,
+      "step": 3050
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.18948824343015214,
+      "eval_loss": 0.7252681255340576,
+      "eval_macro_f1": 0.5210228179882834,
+      "eval_micro_f1": 0.572573718866304,
+      "eval_runtime": 0.4056,
+      "eval_samples_per_second": 1782.557,
+      "eval_steps_per_second": 113.413,
+      "eval_weighted_f1": 0.5857314004710027,
+      "step": 3060
+    },
+    {
+      "epoch": 15.07380073800738,
+      "grad_norm": 6.491910934448242,
+      "learning_rate": 5.479302832244009e-06,
+      "loss": 1.3952,
+      "step": 3075
+    },
+    {
+      "epoch": 15.196801968019681,
+      "grad_norm": 6.425677299499512,
+      "learning_rate": 5.343137254901961e-06,
+      "loss": 1.4124,
+      "step": 3100
+    },
+    {
+      "epoch": 15.31980319803198,
+      "grad_norm": 8.645101547241211,
+      "learning_rate": 5.206971677559913e-06,
+      "loss": 1.4557,
+      "step": 3125
+    },
+    {
+      "epoch": 15.44280442804428,
+      "grad_norm": 7.3229570388793945,
+      "learning_rate": 5.070806100217866e-06,
+      "loss": 1.3918,
+      "step": 3150
+    },
+    {
+      "epoch": 15.565805658056581,
+      "grad_norm": 6.167361736297607,
+      "learning_rate": 4.934640522875817e-06,
+      "loss": 1.3895,
+      "step": 3175
+    },
+    {
+      "epoch": 15.68880688806888,
+      "grad_norm": 5.834020137786865,
+      "learning_rate": 4.79847494553377e-06,
+      "loss": 1.4155,
+      "step": 3200
+    },
+    {
+      "epoch": 15.811808118081181,
+      "grad_norm": 4.483017444610596,
+      "learning_rate": 4.662309368191721e-06,
+      "loss": 1.4209,
+      "step": 3225
+    },
+    {
+      "epoch": 15.93480934809348,
+      "grad_norm": 6.58746337890625,
+      "learning_rate": 4.5261437908496736e-06,
+      "loss": 1.3601,
+      "step": 3250
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.1825726141078838,
+      "eval_loss": 0.7127264142036438,
+      "eval_macro_f1": 0.5253492968273261,
+      "eval_micro_f1": 0.5687867750070048,
+      "eval_runtime": 0.4116,
+      "eval_samples_per_second": 1756.457,
+      "eval_steps_per_second": 111.752,
+      "eval_weighted_f1": 0.5827125157247528,
+      "step": 3264
+    },
+    {
+      "epoch": 16.05412054120541,
+      "grad_norm": 8.892355918884277,
+      "learning_rate": 4.389978213507626e-06,
+      "loss": 1.3576,
+      "step": 3275
+    },
+    {
+      "epoch": 16.177121771217713,
+      "grad_norm": 4.933442115783691,
+      "learning_rate": 4.253812636165578e-06,
+      "loss": 1.3908,
+      "step": 3300
+    },
+    {
+      "epoch": 16.300123001230013,
+      "grad_norm": 4.983952522277832,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 1.3372,
+      "step": 3325
+    },
+    {
+      "epoch": 16.423124231242312,
+      "grad_norm": 4.208073616027832,
+      "learning_rate": 3.9814814814814814e-06,
+      "loss": 1.3834,
+      "step": 3350
+    },
+    {
+      "epoch": 16.54612546125461,
+      "grad_norm": 9.286803245544434,
+      "learning_rate": 3.845315904139434e-06,
+      "loss": 1.4217,
+      "step": 3375
+    },
+    {
+      "epoch": 16.669126691266914,
+      "grad_norm": 11.953864097595215,
+      "learning_rate": 3.7091503267973862e-06,
+      "loss": 1.3355,
+      "step": 3400
+    },
+    {
+      "epoch": 16.792127921279214,
+      "grad_norm": 4.970104694366455,
+      "learning_rate": 3.572984749455338e-06,
+      "loss": 1.3761,
+      "step": 3425
+    },
+    {
+      "epoch": 16.915129151291513,
+      "grad_norm": 6.42612361907959,
+      "learning_rate": 3.4368191721132898e-06,
+      "loss": 1.3635,
+      "step": 3450
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.20193637621023514,
+      "eval_loss": 0.7259645462036133,
+      "eval_macro_f1": 0.5314443842165251,
+      "eval_micro_f1": 0.5772661870503597,
+      "eval_runtime": 0.4131,
+      "eval_samples_per_second": 1750.37,
+      "eval_steps_per_second": 111.365,
+      "eval_weighted_f1": 0.5906881193547726,
+      "step": 3468
+    },
+    {
+      "epoch": 17.034440344403443,
+      "grad_norm": 9.278027534484863,
+      "learning_rate": 3.300653594771242e-06,
+      "loss": 1.3039,
+      "step": 3475
+    },
+    {
+      "epoch": 17.157441574415746,
+      "grad_norm": 4.947128772735596,
+      "learning_rate": 3.164488017429194e-06,
+      "loss": 1.342,
+      "step": 3500
+    },
+    {
+      "epoch": 17.280442804428045,
+      "grad_norm": 5.7515082359313965,
+      "learning_rate": 3.0283224400871465e-06,
+      "loss": 1.3199,
+      "step": 3525
+    },
+    {
+      "epoch": 17.403444034440344,
+      "grad_norm": 7.482638359069824,
+      "learning_rate": 2.8921568627450985e-06,
+      "loss": 1.3333,
+      "step": 3550
+    },
+    {
+      "epoch": 17.526445264452644,
+      "grad_norm": 6.239217758178711,
+      "learning_rate": 2.75599128540305e-06,
+      "loss": 1.3035,
+      "step": 3575
+    },
+    {
+      "epoch": 17.649446494464943,
+      "grad_norm": 8.896585464477539,
+      "learning_rate": 2.6198257080610024e-06,
+      "loss": 1.4039,
+      "step": 3600
+    },
+    {
+      "epoch": 17.772447724477246,
+      "grad_norm": 6.399275779724121,
+      "learning_rate": 2.4836601307189544e-06,
+      "loss": 1.3413,
+      "step": 3625
+    },
+    {
+      "epoch": 17.895448954489545,
+      "grad_norm": 5.9048075675964355,
+      "learning_rate": 2.3474945533769064e-06,
+      "loss": 1.2671,
+      "step": 3650
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.21853388658367912,
+      "eval_loss": 0.7562486529350281,
+      "eval_macro_f1": 0.5437383183547814,
+      "eval_micro_f1": 0.5892334698654184,
+      "eval_runtime": 0.4025,
+      "eval_samples_per_second": 1796.272,
+      "eval_steps_per_second": 114.286,
+      "eval_weighted_f1": 0.5998390621765434,
+      "step": 3672
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 4080,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

ood-vigoemotions-phobert-source-source-00-s42-20260826T130931Z-0c086fe0/trainer/checkpoint-4080/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

ood-vigoemotions-phobert-source-source-00-s42-20260826T130931Z-0c086fe0/trainer/checkpoint-4080/special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-00-s42-20260826T130931Z-0c086fe0/trainer/checkpoint-4080/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 256,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-00-s42-20260826T130931Z-0c086fe0/trainer/checkpoint-4080/trainer_state.json

@@ -0,0 +1,1424 @@
+{
+  "best_global_step": 3672,
+  "best_metric": 0.5437383183547814,
+  "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vigoemotions-phobert-source-source-00-s42-20260826T130931Z-0c086fe0/trainer/checkpoint-3672",
+  "epoch": 20.0,
+  "eval_steps": 500,
+  "global_step": 4080,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.12300123001230012,
+      "grad_norm": 5.347543716430664,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 5.0951,
+      "step": 25
+    },
+    {
+      "epoch": 0.24600246002460024,
+      "grad_norm": 5.039833068847656,
+      "learning_rate": 2.401960784313726e-06,
+      "loss": 5.1964,
+      "step": 50
+    },
+    {
+      "epoch": 0.36900369003690037,
+      "grad_norm": 4.420496463775635,
+      "learning_rate": 3.6274509803921573e-06,
+      "loss": 5.1619,
+      "step": 75
+    },
+    {
+      "epoch": 0.4920049200492005,
+      "grad_norm": 5.998644828796387,
+      "learning_rate": 4.852941176470589e-06,
+      "loss": 5.228,
+      "step": 100
+    },
+    {
+      "epoch": 0.6150061500615006,
+      "grad_norm": 5.091968536376953,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 5.1265,
+      "step": 125
+    },
+    {
+      "epoch": 0.7380073800738007,
+      "grad_norm": 6.671788215637207,
+      "learning_rate": 7.3039215686274515e-06,
+      "loss": 5.0759,
+      "step": 150
+    },
+    {
+      "epoch": 0.8610086100861009,
+      "grad_norm": 12.5631742477417,
+      "learning_rate": 8.529411764705883e-06,
+      "loss": 5.2254,
+      "step": 175
+    },
+    {
+      "epoch": 0.984009840098401,
+      "grad_norm": 4.14892053604126,
+      "learning_rate": 9.754901960784315e-06,
+      "loss": 4.8142,
+      "step": 200
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 1.1878608465194702,
+      "eval_macro_f1": 0.21320196736877045,
+      "eval_micro_f1": 0.23826145704312576,
+      "eval_runtime": 0.4282,
+      "eval_samples_per_second": 1688.602,
+      "eval_steps_per_second": 107.435,
+      "eval_weighted_f1": 0.2757431924455255,
+      "step": 204
+    },
+    {
+      "epoch": 1.103321033210332,
+      "grad_norm": 5.696617603302002,
+      "learning_rate": 1.0980392156862747e-05,
+      "loss": 4.578,
+      "step": 225
+    },
+    {
+      "epoch": 1.2263222632226323,
+      "grad_norm": 7.200258255004883,
+      "learning_rate": 1.2205882352941179e-05,
+      "loss": 4.6407,
+      "step": 250
+    },
+    {
+      "epoch": 1.3493234932349323,
+      "grad_norm": 7.213965892791748,
+      "learning_rate": 1.3431372549019607e-05,
+      "loss": 4.53,
+      "step": 275
+    },
+    {
+      "epoch": 1.4723247232472325,
+      "grad_norm": 18.122190475463867,
+      "learning_rate": 1.4656862745098039e-05,
+      "loss": 4.4205,
+      "step": 300
+    },
+    {
+      "epoch": 1.5953259532595325,
+      "grad_norm": 10.154242515563965,
+      "learning_rate": 1.5882352941176473e-05,
+      "loss": 4.2928,
+      "step": 325
+    },
+    {
+      "epoch": 1.7183271832718328,
+      "grad_norm": 14.085644721984863,
+      "learning_rate": 1.7107843137254903e-05,
+      "loss": 4.0917,
+      "step": 350
+    },
+    {
+      "epoch": 1.841328413284133,
+      "grad_norm": 7.562018871307373,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 4.1262,
+      "step": 375
+    },
+    {
+      "epoch": 1.964329643296433,
+      "grad_norm": 9.274996757507324,
+      "learning_rate": 1.9558823529411766e-05,
+      "loss": 4.0357,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 0.9473660588264465,
+      "eval_macro_f1": 0.2907416898851504,
+      "eval_micro_f1": 0.30697176502074136,
+      "eval_runtime": 0.4394,
+      "eval_samples_per_second": 1645.561,
+      "eval_steps_per_second": 104.697,
+      "eval_weighted_f1": 0.3731837751611428,
+      "step": 408
+    },
+    {
+      "epoch": 2.0836408364083643,
+      "grad_norm": 5.191472053527832,
+      "learning_rate": 1.9912854030501092e-05,
+      "loss": 3.8519,
+      "step": 425
+    },
+    {
+      "epoch": 2.206642066420664,
+      "grad_norm": 13.798820495605469,
+      "learning_rate": 1.9776688453159043e-05,
+      "loss": 3.618,
+      "step": 450
+    },
+    {
+      "epoch": 2.3296432964329643,
+      "grad_norm": 11.70262336730957,
+      "learning_rate": 1.9640522875816993e-05,
+      "loss": 3.7455,
+      "step": 475
+    },
+    {
+      "epoch": 2.4526445264452645,
+      "grad_norm": 7.336067199707031,
+      "learning_rate": 1.9504357298474947e-05,
+      "loss": 3.787,
+      "step": 500
+    },
+    {
+      "epoch": 2.5756457564575648,
+      "grad_norm": 13.359006881713867,
+      "learning_rate": 1.9368191721132898e-05,
+      "loss": 3.6748,
+      "step": 525
+    },
+    {
+      "epoch": 2.6986469864698646,
+      "grad_norm": 10.397160530090332,
+      "learning_rate": 1.9232026143790852e-05,
+      "loss": 3.6422,
+      "step": 550
+    },
+    {
+      "epoch": 2.821648216482165,
+      "grad_norm": 18.4261531829834,
+      "learning_rate": 1.9095860566448803e-05,
+      "loss": 3.5534,
+      "step": 575
+    },
+    {
+      "epoch": 2.944649446494465,
+      "grad_norm": 7.37200927734375,
+      "learning_rate": 1.8959694989106757e-05,
+      "loss": 3.5361,
+      "step": 600
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 0.8620179295539856,
+      "eval_macro_f1": 0.34796481421021197,
+      "eval_micro_f1": 0.3669754591256298,
+      "eval_runtime": 0.4439,
+      "eval_samples_per_second": 1628.883,
+      "eval_steps_per_second": 103.636,
+      "eval_weighted_f1": 0.4245846547213332,
+      "step": 612
+    },
+    {
+      "epoch": 3.063960639606396,
+      "grad_norm": 7.358486652374268,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 3.3656,
+      "step": 625
+    },
+    {
+      "epoch": 3.1869618696186963,
+      "grad_norm": 7.551677227020264,
+      "learning_rate": 1.8687363834422662e-05,
+      "loss": 3.2034,
+      "step": 650
+    },
+    {
+      "epoch": 3.3099630996309966,
+      "grad_norm": 10.687773704528809,
+      "learning_rate": 1.8551198257080612e-05,
+      "loss": 3.3985,
+      "step": 675
+    },
+    {
+      "epoch": 3.4329643296432963,
+      "grad_norm": 7.113092422485352,
+      "learning_rate": 1.8415032679738563e-05,
+      "loss": 3.24,
+      "step": 700
+    },
+    {
+      "epoch": 3.5559655596555966,
+      "grad_norm": 8.03344440460205,
+      "learning_rate": 1.8278867102396517e-05,
+      "loss": 3.2581,
+      "step": 725
+    },
+    {
+      "epoch": 3.678966789667897,
+      "grad_norm": 6.487556457519531,
+      "learning_rate": 1.8142701525054468e-05,
+      "loss": 3.2576,
+      "step": 750
+    },
+    {
+      "epoch": 3.8019680196801966,
+      "grad_norm": 6.6811699867248535,
+      "learning_rate": 1.800653594771242e-05,
+      "loss": 3.1186,
+      "step": 775
+    },
+    {
+      "epoch": 3.924969249692497,
+      "grad_norm": 16.565784454345703,
+      "learning_rate": 1.7870370370370373e-05,
+      "loss": 3.1422,
+      "step": 800
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.005532503457814661,
+      "eval_loss": 0.794185221195221,
+      "eval_macro_f1": 0.36775263884169107,
+      "eval_micro_f1": 0.3921229918811539,
+      "eval_runtime": 0.4322,
+      "eval_samples_per_second": 1672.868,
+      "eval_steps_per_second": 106.434,
+      "eval_weighted_f1": 0.4494910853177473,
+      "step": 816
+    },
+    {
+      "epoch": 4.044280442804428,
+      "grad_norm": 6.690476417541504,
+      "learning_rate": 1.7734204793028323e-05,
+      "loss": 3.0639,
+      "step": 825
+    },
+    {
+      "epoch": 4.167281672816729,
+      "grad_norm": 7.374212741851807,
+      "learning_rate": 1.7598039215686274e-05,
+      "loss": 2.979,
+      "step": 850
+    },
+    {
+      "epoch": 4.290282902829028,
+      "grad_norm": 6.836591720581055,
+      "learning_rate": 1.7461873638344228e-05,
+      "loss": 2.9576,
+      "step": 875
+    },
+    {
+      "epoch": 4.413284132841328,
+      "grad_norm": 17.42571258544922,
+      "learning_rate": 1.732570806100218e-05,
+      "loss": 2.9641,
+      "step": 900
+    },
+    {
+      "epoch": 4.536285362853628,
+      "grad_norm": 5.6428141593933105,
+      "learning_rate": 1.7189542483660133e-05,
+      "loss": 2.9495,
+      "step": 925
+    },
+    {
+      "epoch": 4.659286592865929,
+      "grad_norm": 5.34227180480957,
+      "learning_rate": 1.7053376906318084e-05,
+      "loss": 2.9363,
+      "step": 950
+    },
+    {
+      "epoch": 4.782287822878229,
+      "grad_norm": 6.289327144622803,
+      "learning_rate": 1.6917211328976034e-05,
+      "loss": 2.9018,
+      "step": 975
+    },
+    {
+      "epoch": 4.905289052890529,
+      "grad_norm": 16.238441467285156,
+      "learning_rate": 1.678104575163399e-05,
+      "loss": 2.8826,
+      "step": 1000
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.02074688796680498,
+      "eval_loss": 0.7590930461883545,
+      "eval_macro_f1": 0.3985394691102169,
+      "eval_micro_f1": 0.43172147001934236,
+      "eval_runtime": 0.4883,
+      "eval_samples_per_second": 1480.617,
+      "eval_steps_per_second": 94.202,
+      "eval_weighted_f1": 0.48269412632767783,
+      "step": 1020
+    },
+    {
+      "epoch": 5.02460024600246,
+      "grad_norm": 9.875899314880371,
+      "learning_rate": 1.664488017429194e-05,
+      "loss": 2.7175,
+      "step": 1025
+    },
+    {
+      "epoch": 5.14760147601476,
+      "grad_norm": 11.602205276489258,
+      "learning_rate": 1.6508714596949893e-05,
+      "loss": 2.6652,
+      "step": 1050
+    },
+    {
+      "epoch": 5.27060270602706,
+      "grad_norm": 8.778997421264648,
+      "learning_rate": 1.6372549019607844e-05,
+      "loss": 2.7818,
+      "step": 1075
+    },
+    {
+      "epoch": 5.39360393603936,
+      "grad_norm": 7.099167346954346,
+      "learning_rate": 1.6236383442265798e-05,
+      "loss": 2.7251,
+      "step": 1100
+    },
+    {
+      "epoch": 5.516605166051661,
+      "grad_norm": 5.344371318817139,
+      "learning_rate": 1.610021786492375e-05,
+      "loss": 2.6626,
+      "step": 1125
+    },
+    {
+      "epoch": 5.639606396063961,
+      "grad_norm": 9.371820449829102,
+      "learning_rate": 1.5964052287581703e-05,
+      "loss": 2.5991,
+      "step": 1150
+    },
+    {
+      "epoch": 5.762607626076261,
+      "grad_norm": 5.649630069732666,
+      "learning_rate": 1.5827886710239654e-05,
+      "loss": 2.6888,
+      "step": 1175
+    },
+    {
+      "epoch": 5.885608856088561,
+      "grad_norm": 6.022938251495361,
+      "learning_rate": 1.5691721132897604e-05,
+      "loss": 2.593,
+      "step": 1200
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.04426002766251729,
+      "eval_loss": 0.7480583786964417,
+      "eval_macro_f1": 0.4246178060686669,
+      "eval_micro_f1": 0.46047098402018505,
+      "eval_runtime": 0.415,
+      "eval_samples_per_second": 1742.003,
+      "eval_steps_per_second": 110.833,
+      "eval_weighted_f1": 0.5006830733841459,
+      "step": 1224
+    },
+    {
+      "epoch": 6.004920049200492,
+      "grad_norm": 13.547445297241211,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 2.5475,
+      "step": 1225
+    },
+    {
+      "epoch": 6.127921279212792,
+      "grad_norm": 10.183258056640625,
+      "learning_rate": 1.541938997821351e-05,
+      "loss": 2.5095,
+      "step": 1250
+    },
+    {
+      "epoch": 6.250922509225092,
+      "grad_norm": 7.828505516052246,
+      "learning_rate": 1.528322440087146e-05,
+      "loss": 2.4239,
+      "step": 1275
+    },
+    {
+      "epoch": 6.373923739237393,
+      "grad_norm": 8.002518653869629,
+      "learning_rate": 1.5147058823529412e-05,
+      "loss": 2.4879,
+      "step": 1300
+    },
+    {
+      "epoch": 6.496924969249693,
+      "grad_norm": 12.85448169708252,
+      "learning_rate": 1.5010893246187366e-05,
+      "loss": 2.4154,
+      "step": 1325
+    },
+    {
+      "epoch": 6.619926199261993,
+      "grad_norm": 5.428205966949463,
+      "learning_rate": 1.4874727668845317e-05,
+      "loss": 2.4019,
+      "step": 1350
+    },
+    {
+      "epoch": 6.7429274292742925,
+      "grad_norm": 10.2520112991333,
+      "learning_rate": 1.473856209150327e-05,
+      "loss": 2.4149,
+      "step": 1375
+    },
+    {
+      "epoch": 6.865928659286593,
+      "grad_norm": 8.956914901733398,
+      "learning_rate": 1.4602396514161222e-05,
+      "loss": 2.3976,
+      "step": 1400
+    },
+    {
+      "epoch": 6.988929889298893,
+      "grad_norm": 7.741453647613525,
+      "learning_rate": 1.4466230936819172e-05,
+      "loss": 2.4199,
+      "step": 1425
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.08298755186721991,
+      "eval_loss": 0.7325209975242615,
+      "eval_macro_f1": 0.43862133890959265,
+      "eval_micro_f1": 0.4823896567097637,
+      "eval_runtime": 0.4143,
+      "eval_samples_per_second": 1744.944,
+      "eval_steps_per_second": 111.02,
+      "eval_weighted_f1": 0.5131973346966094,
+      "step": 1428
+    },
+    {
+      "epoch": 7.108241082410824,
+      "grad_norm": 5.299953937530518,
+      "learning_rate": 1.4330065359477125e-05,
+      "loss": 2.2119,
+      "step": 1450
+    },
+    {
+      "epoch": 7.231242312423125,
+      "grad_norm": 8.74158763885498,
+      "learning_rate": 1.4193899782135077e-05,
+      "loss": 2.3311,
+      "step": 1475
+    },
+    {
+      "epoch": 7.354243542435424,
+      "grad_norm": 10.328490257263184,
+      "learning_rate": 1.405773420479303e-05,
+      "loss": 2.2658,
+      "step": 1500
+    },
+    {
+      "epoch": 7.477244772447724,
+      "grad_norm": 8.569588661193848,
+      "learning_rate": 1.392156862745098e-05,
+      "loss": 2.2812,
+      "step": 1525
+    },
+    {
+      "epoch": 7.6002460024600245,
+      "grad_norm": 7.103965759277344,
+      "learning_rate": 1.3785403050108934e-05,
+      "loss": 2.2201,
+      "step": 1550
+    },
+    {
+      "epoch": 7.723247232472325,
+      "grad_norm": 6.22614049911499,
+      "learning_rate": 1.3649237472766885e-05,
+      "loss": 2.2075,
+      "step": 1575
+    },
+    {
+      "epoch": 7.846248462484625,
+      "grad_norm": 7.187036991119385,
+      "learning_rate": 1.351307189542484e-05,
+      "loss": 2.1669,
+      "step": 1600
+    },
+    {
+      "epoch": 7.969249692496925,
+      "grad_norm": 6.168943405151367,
+      "learning_rate": 1.337690631808279e-05,
+      "loss": 2.2443,
+      "step": 1625
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.10926694329183956,
+      "eval_loss": 0.7492125630378723,
+      "eval_macro_f1": 0.47865805519264376,
+      "eval_micro_f1": 0.5181168350998274,
+      "eval_runtime": 0.4166,
+      "eval_samples_per_second": 1735.553,
+      "eval_steps_per_second": 110.422,
+      "eval_weighted_f1": 0.5433474238239877,
+      "step": 1632
+    },
+    {
+      "epoch": 8.088560885608857,
+      "grad_norm": 6.345441818237305,
+      "learning_rate": 1.3240740740740742e-05,
+      "loss": 2.1594,
+      "step": 1650
+    },
+    {
+      "epoch": 8.211562115621156,
+      "grad_norm": 6.153682708740234,
+      "learning_rate": 1.3104575163398695e-05,
+      "loss": 2.086,
+      "step": 1675
+    },
+    {
+      "epoch": 8.334563345633457,
+      "grad_norm": 6.458537578582764,
+      "learning_rate": 1.2968409586056645e-05,
+      "loss": 2.0523,
+      "step": 1700
+    },
+    {
+      "epoch": 8.457564575645756,
+      "grad_norm": 4.463102340698242,
+      "learning_rate": 1.2832244008714598e-05,
+      "loss": 2.0716,
+      "step": 1725
+    },
+    {
+      "epoch": 8.580565805658056,
+      "grad_norm": 5.9551897048950195,
+      "learning_rate": 1.269607843137255e-05,
+      "loss": 2.0998,
+      "step": 1750
+    },
+    {
+      "epoch": 8.703567035670357,
+      "grad_norm": 12.094724655151367,
+      "learning_rate": 1.2559912854030503e-05,
+      "loss": 2.1279,
+      "step": 1775
+    },
+    {
+      "epoch": 8.826568265682656,
+      "grad_norm": 5.6941142082214355,
+      "learning_rate": 1.2423747276688453e-05,
+      "loss": 2.022,
+      "step": 1800
+    },
+    {
+      "epoch": 8.949569495694957,
+      "grad_norm": 5.646745204925537,
+      "learning_rate": 1.2287581699346407e-05,
+      "loss": 2.0617,
+      "step": 1825
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.09820193637621023,
+      "eval_loss": 0.7172796130180359,
+      "eval_macro_f1": 0.4703039493047723,
+      "eval_micro_f1": 0.5005804504295334,
+      "eval_runtime": 0.4506,
+      "eval_samples_per_second": 1604.522,
+      "eval_steps_per_second": 102.086,
+      "eval_weighted_f1": 0.5364196628370508,
+      "step": 1836
+    },
+    {
+      "epoch": 9.068880688806889,
+      "grad_norm": 6.38439416885376,
+      "learning_rate": 1.2151416122004358e-05,
+      "loss": 2.0243,
+      "step": 1850
+    },
+    {
+      "epoch": 9.191881918819188,
+      "grad_norm": 15.08652400970459,
+      "learning_rate": 1.2015250544662312e-05,
+      "loss": 1.9855,
+      "step": 1875
+    },
+    {
+      "epoch": 9.314883148831488,
+      "grad_norm": 5.886865615844727,
+      "learning_rate": 1.1879084967320263e-05,
+      "loss": 1.9351,
+      "step": 1900
+    },
+    {
+      "epoch": 9.437884378843789,
+      "grad_norm": 5.136407375335693,
+      "learning_rate": 1.1742919389978214e-05,
+      "loss": 1.9242,
+      "step": 1925
+    },
+    {
+      "epoch": 9.560885608856088,
+      "grad_norm": 7.7304911613464355,
+      "learning_rate": 1.1606753812636166e-05,
+      "loss": 1.987,
+      "step": 1950
+    },
+    {
+      "epoch": 9.68388683886839,
+      "grad_norm": 9.386374473571777,
+      "learning_rate": 1.1470588235294118e-05,
+      "loss": 1.9295,
+      "step": 1975
+    },
+    {
+      "epoch": 9.806888068880689,
+      "grad_norm": 6.307365894317627,
+      "learning_rate": 1.133442265795207e-05,
+      "loss": 1.9393,
+      "step": 2000
+    },
+    {
+      "epoch": 9.92988929889299,
+      "grad_norm": 6.248044967651367,
+      "learning_rate": 1.1198257080610021e-05,
+      "loss": 1.9582,
+      "step": 2025
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.13416320885200553,
+      "eval_loss": 0.7009500861167908,
+      "eval_macro_f1": 0.49583988112082567,
+      "eval_micro_f1": 0.5355776587605203,
+      "eval_runtime": 0.4279,
+      "eval_samples_per_second": 1689.492,
+      "eval_steps_per_second": 107.492,
+      "eval_weighted_f1": 0.5556808037515301,
+      "step": 2040
+    },
+    {
+      "epoch": 10.04920049200492,
+      "grad_norm": 6.902281284332275,
+      "learning_rate": 1.1062091503267976e-05,
+      "loss": 1.7956,
+      "step": 2050
+    },
+    {
+      "epoch": 10.17220172201722,
+      "grad_norm": 5.536381244659424,
+      "learning_rate": 1.0925925925925926e-05,
+      "loss": 1.856,
+      "step": 2075
+    },
+    {
+      "epoch": 10.29520295202952,
+      "grad_norm": 3.8086092472076416,
+      "learning_rate": 1.078976034858388e-05,
+      "loss": 1.8511,
+      "step": 2100
+    },
+    {
+      "epoch": 10.418204182041821,
+      "grad_norm": 6.091910362243652,
+      "learning_rate": 1.0653594771241831e-05,
+      "loss": 1.789,
+      "step": 2125
+    },
+    {
+      "epoch": 10.54120541205412,
+      "grad_norm": 5.5678839683532715,
+      "learning_rate": 1.0517429193899782e-05,
+      "loss": 1.8157,
+      "step": 2150
+    },
+    {
+      "epoch": 10.664206642066421,
+      "grad_norm": 5.289937496185303,
+      "learning_rate": 1.0381263616557736e-05,
+      "loss": 1.8265,
+      "step": 2175
+    },
+    {
+      "epoch": 10.78720787207872,
+      "grad_norm": 7.392284393310547,
+      "learning_rate": 1.0245098039215687e-05,
+      "loss": 1.7552,
+      "step": 2200
+    },
+    {
+      "epoch": 10.91020910209102,
+      "grad_norm": 7.5922746658325195,
+      "learning_rate": 1.0108932461873639e-05,
+      "loss": 1.805,
+      "step": 2225
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.16044260027662519,
+      "eval_loss": 0.7213637828826904,
+      "eval_macro_f1": 0.514443422016327,
+      "eval_micro_f1": 0.5513196480938416,
+      "eval_runtime": 0.4384,
+      "eval_samples_per_second": 1649.083,
+      "eval_steps_per_second": 104.921,
+      "eval_weighted_f1": 0.5660369113766512,
+      "step": 2244
+    },
+    {
+      "epoch": 11.029520295202952,
+      "grad_norm": 5.942030906677246,
+      "learning_rate": 9.972766884531591e-06,
+      "loss": 1.7107,
+      "step": 2250
+    },
+    {
+      "epoch": 11.152521525215253,
+      "grad_norm": 6.128445625305176,
+      "learning_rate": 9.836601307189542e-06,
+      "loss": 1.7904,
+      "step": 2275
+    },
+    {
+      "epoch": 11.275522755227552,
+      "grad_norm": 6.908228874206543,
+      "learning_rate": 9.700435729847494e-06,
+      "loss": 1.6947,
+      "step": 2300
+    },
+    {
+      "epoch": 11.398523985239853,
+      "grad_norm": 12.505904197692871,
+      "learning_rate": 9.564270152505447e-06,
+      "loss": 1.7133,
+      "step": 2325
+    },
+    {
+      "epoch": 11.521525215252153,
+      "grad_norm": 4.8953986167907715,
+      "learning_rate": 9.4281045751634e-06,
+      "loss": 1.678,
+      "step": 2350
+    },
+    {
+      "epoch": 11.644526445264452,
+      "grad_norm": 6.927519798278809,
+      "learning_rate": 9.291938997821352e-06,
+      "loss": 1.6724,
+      "step": 2375
+    },
+    {
+      "epoch": 11.767527675276753,
+      "grad_norm": 11.44202709197998,
+      "learning_rate": 9.155773420479304e-06,
+      "loss": 1.7256,
+      "step": 2400
+    },
+    {
+      "epoch": 11.890528905289052,
+      "grad_norm": 6.743433952331543,
+      "learning_rate": 9.019607843137256e-06,
+      "loss": 1.6917,
+      "step": 2425
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.14937759336099585,
+      "eval_loss": 0.7130449414253235,
+      "eval_macro_f1": 0.510854907416043,
+      "eval_micro_f1": 0.5542811416377701,
+      "eval_runtime": 0.4717,
+      "eval_samples_per_second": 1532.598,
+      "eval_steps_per_second": 97.51,
+      "eval_weighted_f1": 0.5733406737839715,
+      "step": 2448
+    },
+    {
+      "epoch": 12.009840098400984,
+      "grad_norm": 9.664470672607422,
+      "learning_rate": 8.883442265795209e-06,
+      "loss": 1.6364,
+      "step": 2450
+    },
+    {
+      "epoch": 12.132841328413285,
+      "grad_norm": 12.487649917602539,
+      "learning_rate": 8.74727668845316e-06,
+      "loss": 1.6667,
+      "step": 2475
+    },
+    {
+      "epoch": 12.255842558425584,
+      "grad_norm": 6.491523265838623,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 1.5484,
+      "step": 2500
+    },
+    {
+      "epoch": 12.378843788437884,
+      "grad_norm": 8.131217956542969,
+      "learning_rate": 8.474945533769063e-06,
+      "loss": 1.6584,
+      "step": 2525
+    },
+    {
+      "epoch": 12.501845018450185,
+      "grad_norm": 6.362199306488037,
+      "learning_rate": 8.338779956427015e-06,
+      "loss": 1.6223,
+      "step": 2550
+    },
+    {
+      "epoch": 12.624846248462484,
+      "grad_norm": 9.135724067687988,
+      "learning_rate": 8.202614379084967e-06,
+      "loss": 1.6026,
+      "step": 2575
+    },
+    {
+      "epoch": 12.747847478474785,
+      "grad_norm": 7.459300518035889,
+      "learning_rate": 8.06644880174292e-06,
+      "loss": 1.5687,
+      "step": 2600
+    },
+    {
+      "epoch": 12.870848708487085,
+      "grad_norm": 5.186668395996094,
+      "learning_rate": 7.930283224400872e-06,
+      "loss": 1.5528,
+      "step": 2625
+    },
+    {
+      "epoch": 12.993849938499386,
+      "grad_norm": 13.283559799194336,
+      "learning_rate": 7.794117647058825e-06,
+      "loss": 1.654,
+      "step": 2650
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.16597510373443983,
+      "eval_loss": 0.6994755268096924,
+      "eval_macro_f1": 0.5197501449610717,
+      "eval_micro_f1": 0.5607123583378305,
+      "eval_runtime": 0.4299,
+      "eval_samples_per_second": 1681.739,
+      "eval_steps_per_second": 106.999,
+      "eval_weighted_f1": 0.5785792673842536,
+      "step": 2652
+    },
+    {
+      "epoch": 13.113161131611315,
+      "grad_norm": 7.693202972412109,
+      "learning_rate": 7.657952069716777e-06,
+      "loss": 1.4981,
+      "step": 2675
+    },
+    {
+      "epoch": 13.236162361623617,
+      "grad_norm": 4.944885730743408,
+      "learning_rate": 7.5217864923747285e-06,
+      "loss": 1.563,
+      "step": 2700
+    },
+    {
+      "epoch": 13.359163591635916,
+      "grad_norm": 7.11841344833374,
+      "learning_rate": 7.385620915032681e-06,
+      "loss": 1.5465,
+      "step": 2725
+    },
+    {
+      "epoch": 13.482164821648217,
+      "grad_norm": 7.9372968673706055,
+      "learning_rate": 7.249455337690633e-06,
+      "loss": 1.5258,
+      "step": 2750
+    },
+    {
+      "epoch": 13.605166051660516,
+      "grad_norm": 5.245449542999268,
+      "learning_rate": 7.113289760348584e-06,
+      "loss": 1.5063,
+      "step": 2775
+    },
+    {
+      "epoch": 13.728167281672818,
+      "grad_norm": 8.229990005493164,
+      "learning_rate": 6.977124183006536e-06,
+      "loss": 1.5031,
+      "step": 2800
+    },
+    {
+      "epoch": 13.851168511685117,
+      "grad_norm": 7.4326653480529785,
+      "learning_rate": 6.840958605664488e-06,
+      "loss": 1.5227,
+      "step": 2825
+    },
+    {
+      "epoch": 13.974169741697416,
+      "grad_norm": 17.769107818603516,
+      "learning_rate": 6.70479302832244e-06,
+      "loss": 1.5404,
+      "step": 2850
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.1839557399723375,
+      "eval_loss": 0.7303460240364075,
+      "eval_macro_f1": 0.5237214993143648,
+      "eval_micro_f1": 0.5703682557807593,
+      "eval_runtime": 0.4195,
+      "eval_samples_per_second": 1723.501,
+      "eval_steps_per_second": 109.656,
+      "eval_weighted_f1": 0.5845632043986984,
+      "step": 2856
+    },
+    {
+      "epoch": 14.093480934809348,
+      "grad_norm": 4.83207368850708,
+      "learning_rate": 6.568627450980393e-06,
+      "loss": 1.3969,
+      "step": 2875
+    },
+    {
+      "epoch": 14.216482164821649,
+      "grad_norm": 8.71630573272705,
+      "learning_rate": 6.432461873638345e-06,
+      "loss": 1.4554,
+      "step": 2900
+    },
+    {
+      "epoch": 14.339483394833948,
+      "grad_norm": 9.311717987060547,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 1.5311,
+      "step": 2925
+    },
+    {
+      "epoch": 14.46248462484625,
+      "grad_norm": 16.771753311157227,
+      "learning_rate": 6.160130718954249e-06,
+      "loss": 1.4538,
+      "step": 2950
+    },
+    {
+      "epoch": 14.585485854858549,
+      "grad_norm": 5.463100433349609,
+      "learning_rate": 6.0239651416122014e-06,
+      "loss": 1.4487,
+      "step": 2975
+    },
+    {
+      "epoch": 14.708487084870848,
+      "grad_norm": 5.630454063415527,
+      "learning_rate": 5.887799564270154e-06,
+      "loss": 1.4717,
+      "step": 3000
+    },
+    {
+      "epoch": 14.83148831488315,
+      "grad_norm": 11.105183601379395,
+      "learning_rate": 5.7516339869281045e-06,
+      "loss": 1.4308,
+      "step": 3025
+    },
+    {
+      "epoch": 14.954489544895448,
+      "grad_norm": 8.745418548583984,
+      "learning_rate": 5.615468409586057e-06,
+      "loss": 1.526,
+      "step": 3050
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.18948824343015214,
+      "eval_loss": 0.7252681255340576,
+      "eval_macro_f1": 0.5210228179882834,
+      "eval_micro_f1": 0.572573718866304,
+      "eval_runtime": 0.4056,
+      "eval_samples_per_second": 1782.557,
+      "eval_steps_per_second": 113.413,
+      "eval_weighted_f1": 0.5857314004710027,
+      "step": 3060
+    },
+    {
+      "epoch": 15.07380073800738,
+      "grad_norm": 6.491910934448242,
+      "learning_rate": 5.479302832244009e-06,
+      "loss": 1.3952,
+      "step": 3075
+    },
+    {
+      "epoch": 15.196801968019681,
+      "grad_norm": 6.425677299499512,
+      "learning_rate": 5.343137254901961e-06,
+      "loss": 1.4124,
+      "step": 3100
+    },
+    {
+      "epoch": 15.31980319803198,
+      "grad_norm": 8.645101547241211,
+      "learning_rate": 5.206971677559913e-06,
+      "loss": 1.4557,
+      "step": 3125
+    },
+    {
+      "epoch": 15.44280442804428,
+      "grad_norm": 7.3229570388793945,
+      "learning_rate": 5.070806100217866e-06,
+      "loss": 1.3918,
+      "step": 3150
+    },
+    {
+      "epoch": 15.565805658056581,
+      "grad_norm": 6.167361736297607,
+      "learning_rate": 4.934640522875817e-06,
+      "loss": 1.3895,
+      "step": 3175
+    },
+    {
+      "epoch": 15.68880688806888,
+      "grad_norm": 5.834020137786865,
+      "learning_rate": 4.79847494553377e-06,
+      "loss": 1.4155,
+      "step": 3200
+    },
+    {
+      "epoch": 15.811808118081181,
+      "grad_norm": 4.483017444610596,
+      "learning_rate": 4.662309368191721e-06,
+      "loss": 1.4209,
+      "step": 3225
+    },
+    {
+      "epoch": 15.93480934809348,
+      "grad_norm": 6.58746337890625,
+      "learning_rate": 4.5261437908496736e-06,
+      "loss": 1.3601,
+      "step": 3250
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.1825726141078838,
+      "eval_loss": 0.7127264142036438,
+      "eval_macro_f1": 0.5253492968273261,
+      "eval_micro_f1": 0.5687867750070048,
+      "eval_runtime": 0.4116,
+      "eval_samples_per_second": 1756.457,
+      "eval_steps_per_second": 111.752,
+      "eval_weighted_f1": 0.5827125157247528,
+      "step": 3264
+    },
+    {
+      "epoch": 16.05412054120541,
+      "grad_norm": 8.892355918884277,
+      "learning_rate": 4.389978213507626e-06,
+      "loss": 1.3576,
+      "step": 3275
+    },
+    {
+      "epoch": 16.177121771217713,
+      "grad_norm": 4.933442115783691,
+      "learning_rate": 4.253812636165578e-06,
+      "loss": 1.3908,
+      "step": 3300
+    },
+    {
+      "epoch": 16.300123001230013,
+      "grad_norm": 4.983952522277832,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 1.3372,
+      "step": 3325
+    },
+    {
+      "epoch": 16.423124231242312,
+      "grad_norm": 4.208073616027832,
+      "learning_rate": 3.9814814814814814e-06,
+      "loss": 1.3834,
+      "step": 3350
+    },
+    {
+      "epoch": 16.54612546125461,
+      "grad_norm": 9.286803245544434,
+      "learning_rate": 3.845315904139434e-06,
+      "loss": 1.4217,
+      "step": 3375
+    },
+    {
+      "epoch": 16.669126691266914,
+      "grad_norm": 11.953864097595215,
+      "learning_rate": 3.7091503267973862e-06,
+      "loss": 1.3355,
+      "step": 3400
+    },
+    {
+      "epoch": 16.792127921279214,
+      "grad_norm": 4.970104694366455,
+      "learning_rate": 3.572984749455338e-06,
+      "loss": 1.3761,
+      "step": 3425
+    },
+    {
+      "epoch": 16.915129151291513,
+      "grad_norm": 6.42612361907959,
+      "learning_rate": 3.4368191721132898e-06,
+      "loss": 1.3635,
+      "step": 3450
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.20193637621023514,
+      "eval_loss": 0.7259645462036133,
+      "eval_macro_f1": 0.5314443842165251,
+      "eval_micro_f1": 0.5772661870503597,
+      "eval_runtime": 0.4131,
+      "eval_samples_per_second": 1750.37,
+      "eval_steps_per_second": 111.365,
+      "eval_weighted_f1": 0.5906881193547726,
+      "step": 3468
+    },
+    {
+      "epoch": 17.034440344403443,
+      "grad_norm": 9.278027534484863,
+      "learning_rate": 3.300653594771242e-06,
+      "loss": 1.3039,
+      "step": 3475
+    },
+    {
+      "epoch": 17.157441574415746,
+      "grad_norm": 4.947128772735596,
+      "learning_rate": 3.164488017429194e-06,
+      "loss": 1.342,
+      "step": 3500
+    },
+    {
+      "epoch": 17.280442804428045,
+      "grad_norm": 5.7515082359313965,
+      "learning_rate": 3.0283224400871465e-06,
+      "loss": 1.3199,
+      "step": 3525
+    },
+    {
+      "epoch": 17.403444034440344,
+      "grad_norm": 7.482638359069824,
+      "learning_rate": 2.8921568627450985e-06,
+      "loss": 1.3333,
+      "step": 3550
+    },
+    {
+      "epoch": 17.526445264452644,
+      "grad_norm": 6.239217758178711,
+      "learning_rate": 2.75599128540305e-06,
+      "loss": 1.3035,
+      "step": 3575
+    },
+    {
+      "epoch": 17.649446494464943,
+      "grad_norm": 8.896585464477539,
+      "learning_rate": 2.6198257080610024e-06,
+      "loss": 1.4039,
+      "step": 3600
+    },
+    {
+      "epoch": 17.772447724477246,
+      "grad_norm": 6.399275779724121,
+      "learning_rate": 2.4836601307189544e-06,
+      "loss": 1.3413,
+      "step": 3625
+    },
+    {
+      "epoch": 17.895448954489545,
+      "grad_norm": 5.9048075675964355,
+      "learning_rate": 2.3474945533769064e-06,
+      "loss": 1.2671,
+      "step": 3650
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.21853388658367912,
+      "eval_loss": 0.7562486529350281,
+      "eval_macro_f1": 0.5437383183547814,
+      "eval_micro_f1": 0.5892334698654184,
+      "eval_runtime": 0.4025,
+      "eval_samples_per_second": 1796.272,
+      "eval_steps_per_second": 114.286,
+      "eval_weighted_f1": 0.5998390621765434,
+      "step": 3672
+    },
+    {
+      "epoch": 18.014760147601475,
+      "grad_norm": 5.023840427398682,
+      "learning_rate": 2.2113289760348588e-06,
+      "loss": 1.2706,
+      "step": 3675
+    },
+    {
+      "epoch": 18.137761377613778,
+      "grad_norm": 4.5124831199646,
+      "learning_rate": 2.0751633986928107e-06,
+      "loss": 1.296,
+      "step": 3700
+    },
+    {
+      "epoch": 18.260762607626077,
+      "grad_norm": 9.173319816589355,
+      "learning_rate": 1.9389978213507627e-06,
+      "loss": 1.3194,
+      "step": 3725
+    },
+    {
+      "epoch": 18.383763837638377,
+      "grad_norm": 5.279863357543945,
+      "learning_rate": 1.802832244008715e-06,
+      "loss": 1.343,
+      "step": 3750
+    },
+    {
+      "epoch": 18.506765067650676,
+      "grad_norm": 4.328301906585693,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.2895,
+      "step": 3775
+    },
+    {
+      "epoch": 18.629766297662975,
+      "grad_norm": 4.910815238952637,
+      "learning_rate": 1.5305010893246188e-06,
+      "loss": 1.3101,
+      "step": 3800
+    },
+    {
+      "epoch": 18.752767527675278,
+      "grad_norm": 11.867794036865234,
+      "learning_rate": 1.394335511982571e-06,
+      "loss": 1.259,
+      "step": 3825
+    },
+    {
+      "epoch": 18.875768757687577,
+      "grad_norm": 5.709174633026123,
+      "learning_rate": 1.258169934640523e-06,
+      "loss": 1.3082,
+      "step": 3850
+    },
+    {
+      "epoch": 18.998769987699877,
+      "grad_norm": 4.429140567779541,
+      "learning_rate": 1.122004357298475e-06,
+      "loss": 1.3049,
+      "step": 3875
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.2033195020746888,
+      "eval_loss": 0.7472886443138123,
+      "eval_macro_f1": 0.5311168116175851,
+      "eval_micro_f1": 0.5780080598733448,
+      "eval_runtime": 0.4035,
+      "eval_samples_per_second": 1791.725,
+      "eval_steps_per_second": 113.996,
+      "eval_weighted_f1": 0.590838038205878,
+      "step": 3876
+    },
+    {
+      "epoch": 19.118081180811807,
+      "grad_norm": 5.378570079803467,
+      "learning_rate": 9.858387799564272e-07,
+      "loss": 1.2442,
+      "step": 3900
+    },
+    {
+      "epoch": 19.24108241082411,
+      "grad_norm": 5.290021896362305,
+      "learning_rate": 8.496732026143792e-07,
+      "loss": 1.344,
+      "step": 3925
+    },
+    {
+      "epoch": 19.36408364083641,
+      "grad_norm": 6.8393683433532715,
+      "learning_rate": 7.135076252723312e-07,
+      "loss": 1.3138,
+      "step": 3950
+    },
+    {
+      "epoch": 19.487084870848708,
+      "grad_norm": 6.9358601570129395,
+      "learning_rate": 5.773420479302833e-07,
+      "loss": 1.2876,
+      "step": 3975
+    },
+    {
+      "epoch": 19.610086100861007,
+      "grad_norm": 7.213063716888428,
+      "learning_rate": 4.4117647058823536e-07,
+      "loss": 1.2293,
+      "step": 4000
+    },
+    {
+      "epoch": 19.73308733087331,
+      "grad_norm": 8.128819465637207,
+      "learning_rate": 3.050108932461874e-07,
+      "loss": 1.3021,
+      "step": 4025
+    },
+    {
+      "epoch": 19.85608856088561,
+      "grad_norm": 7.749570369720459,
+      "learning_rate": 1.6884531590413946e-07,
+      "loss": 1.2288,
+      "step": 4050
+    },
+    {
+      "epoch": 19.97908979089791,
+      "grad_norm": 5.202642440795898,
+      "learning_rate": 3.267973856209151e-08,
+      "loss": 1.2972,
+      "step": 4075
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.2074688796680498,
+      "eval_loss": 0.7472759485244751,
+      "eval_macro_f1": 0.5309981980472833,
+      "eval_micro_f1": 0.5790697674418605,
+      "eval_runtime": 0.4263,
+      "eval_samples_per_second": 1695.817,
+      "eval_steps_per_second": 107.894,
+      "eval_weighted_f1": 0.5905288025272599,
+      "step": 4080
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 4080,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 2
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

ood-vigoemotions-phobert-source-source-00-s43-20260826T134203Z-1f9ffdf3/trainer/checkpoint-3876/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

ood-vigoemotions-phobert-source-source-00-s43-20260826T134203Z-1f9ffdf3/trainer/checkpoint-3876/special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-00-s43-20260826T134203Z-1f9ffdf3/trainer/checkpoint-3876/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 256,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-00-s43-20260826T134203Z-1f9ffdf3/trainer/checkpoint-3876/trainer_state.json

@@ -0,0 +1,1356 @@
+{
+  "best_global_step": 3468,
+  "best_metric": 0.5334202111120329,
+  "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vigoemotions-phobert-source-source-00-s43-20260826T134203Z-1f9ffdf3/trainer/checkpoint-3468",
+  "epoch": 19.0,
+  "eval_steps": 500,
+  "global_step": 3876,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.12300123001230012,
+      "grad_norm": 8.09683895111084,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 5.2456,
+      "step": 25
+    },
+    {
+      "epoch": 0.24600246002460024,
+      "grad_norm": 5.68751859664917,
+      "learning_rate": 2.401960784313726e-06,
+      "loss": 5.2497,
+      "step": 50
+    },
+    {
+      "epoch": 0.36900369003690037,
+      "grad_norm": 306.9134216308594,
+      "learning_rate": 3.6274509803921573e-06,
+      "loss": 5.1922,
+      "step": 75
+    },
+    {
+      "epoch": 0.4920049200492005,
+      "grad_norm": 5.2206339836120605,
+      "learning_rate": 4.852941176470589e-06,
+      "loss": 5.2109,
+      "step": 100
+    },
+    {
+      "epoch": 0.6150061500615006,
+      "grad_norm": 6.440758228302002,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 5.1985,
+      "step": 125
+    },
+    {
+      "epoch": 0.7380073800738007,
+      "grad_norm": 5.165765762329102,
+      "learning_rate": 7.3039215686274515e-06,
+      "loss": 5.0125,
+      "step": 150
+    },
+    {
+      "epoch": 0.8610086100861009,
+      "grad_norm": 7.758053302764893,
+      "learning_rate": 8.529411764705883e-06,
+      "loss": 5.0253,
+      "step": 175
+    },
+    {
+      "epoch": 0.984009840098401,
+      "grad_norm": 10.424230575561523,
+      "learning_rate": 9.754901960784315e-06,
+      "loss": 4.9213,
+      "step": 200
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 1.194104790687561,
+      "eval_macro_f1": 0.20881396339551378,
+      "eval_micro_f1": 0.2245830234781685,
+      "eval_runtime": 0.4229,
+      "eval_samples_per_second": 1709.77,
+      "eval_steps_per_second": 108.782,
+      "eval_weighted_f1": 0.2739720167784619,
+      "step": 204
+    },
+    {
+      "epoch": 1.103321033210332,
+      "grad_norm": 5.252875804901123,
+      "learning_rate": 1.0980392156862747e-05,
+      "loss": 4.7508,
+      "step": 225
+    },
+    {
+      "epoch": 1.2263222632226323,
+      "grad_norm": 5.531340599060059,
+      "learning_rate": 1.2205882352941179e-05,
+      "loss": 4.4645,
+      "step": 250
+    },
+    {
+      "epoch": 1.3493234932349323,
+      "grad_norm": 7.683099746704102,
+      "learning_rate": 1.3431372549019607e-05,
+      "loss": 4.5754,
+      "step": 275
+    },
+    {
+      "epoch": 1.4723247232472325,
+      "grad_norm": 8.739664077758789,
+      "learning_rate": 1.4656862745098039e-05,
+      "loss": 4.5256,
+      "step": 300
+    },
+    {
+      "epoch": 1.5953259532595325,
+      "grad_norm": 8.408071517944336,
+      "learning_rate": 1.5882352941176473e-05,
+      "loss": 4.3144,
+      "step": 325
+    },
+    {
+      "epoch": 1.7183271832718328,
+      "grad_norm": 5.407787322998047,
+      "learning_rate": 1.7107843137254903e-05,
+      "loss": 4.1639,
+      "step": 350
+    },
+    {
+      "epoch": 1.841328413284133,
+      "grad_norm": 6.228161811828613,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 4.0869,
+      "step": 375
+    },
+    {
+      "epoch": 1.964329643296433,
+      "grad_norm": 5.647112846374512,
+      "learning_rate": 1.9558823529411766e-05,
+      "loss": 3.9973,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.0013831258644536654,
+      "eval_loss": 0.9395498037338257,
+      "eval_macro_f1": 0.3101064655101556,
+      "eval_micro_f1": 0.3253012048192771,
+      "eval_runtime": 0.4162,
+      "eval_samples_per_second": 1736.959,
+      "eval_steps_per_second": 110.512,
+      "eval_weighted_f1": 0.38850470823525307,
+      "step": 408
+    },
+    {
+      "epoch": 2.0836408364083643,
+      "grad_norm": 5.767376899719238,
+      "learning_rate": 1.9912854030501092e-05,
+      "loss": 3.7444,
+      "step": 425
+    },
+    {
+      "epoch": 2.206642066420664,
+      "grad_norm": 8.322330474853516,
+      "learning_rate": 1.9776688453159043e-05,
+      "loss": 3.6566,
+      "step": 450
+    },
+    {
+      "epoch": 2.3296432964329643,
+      "grad_norm": 12.773788452148438,
+      "learning_rate": 1.9640522875816993e-05,
+      "loss": 3.8016,
+      "step": 475
+    },
+    {
+      "epoch": 2.4526445264452645,
+      "grad_norm": 12.092753410339355,
+      "learning_rate": 1.9504357298474947e-05,
+      "loss": 3.6083,
+      "step": 500
+    },
+    {
+      "epoch": 2.5756457564575648,
+      "grad_norm": 13.042563438415527,
+      "learning_rate": 1.9368191721132898e-05,
+      "loss": 3.6308,
+      "step": 525
+    },
+    {
+      "epoch": 2.6986469864698646,
+      "grad_norm": 6.5783562660217285,
+      "learning_rate": 1.9232026143790852e-05,
+      "loss": 3.5555,
+      "step": 550
+    },
+    {
+      "epoch": 2.821648216482165,
+      "grad_norm": 14.115478515625,
+      "learning_rate": 1.9095860566448803e-05,
+      "loss": 3.5335,
+      "step": 575
+    },
+    {
+      "epoch": 2.944649446494465,
+      "grad_norm": 16.671173095703125,
+      "learning_rate": 1.8959694989106757e-05,
+      "loss": 3.4805,
+      "step": 600
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.011065006915629323,
+      "eval_loss": 0.8417011499404907,
+      "eval_macro_f1": 0.3377949019335999,
+      "eval_micro_f1": 0.3597916995423702,
+      "eval_runtime": 0.4516,
+      "eval_samples_per_second": 1601.102,
+      "eval_steps_per_second": 101.868,
+      "eval_weighted_f1": 0.4160828254989279,
+      "step": 612
+    },
+    {
+      "epoch": 3.063960639606396,
+      "grad_norm": 5.159065246582031,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 3.3403,
+      "step": 625
+    },
+    {
+      "epoch": 3.1869618696186963,
+      "grad_norm": 5.292785167694092,
+      "learning_rate": 1.8687363834422662e-05,
+      "loss": 3.2876,
+      "step": 650
+    },
+    {
+      "epoch": 3.3099630996309966,
+      "grad_norm": 10.624845504760742,
+      "learning_rate": 1.8551198257080612e-05,
+      "loss": 3.2157,
+      "step": 675
+    },
+    {
+      "epoch": 3.4329643296432963,
+      "grad_norm": 7.054944038391113,
+      "learning_rate": 1.8415032679738563e-05,
+      "loss": 3.2664,
+      "step": 700
+    },
+    {
+      "epoch": 3.5559655596555966,
+      "grad_norm": 4.8366522789001465,
+      "learning_rate": 1.8278867102396517e-05,
+      "loss": 3.2077,
+      "step": 725
+    },
+    {
+      "epoch": 3.678966789667897,
+      "grad_norm": 10.388249397277832,
+      "learning_rate": 1.8142701525054468e-05,
+      "loss": 3.2117,
+      "step": 750
+    },
+    {
+      "epoch": 3.8019680196801966,
+      "grad_norm": 7.776172161102295,
+      "learning_rate": 1.800653594771242e-05,
+      "loss": 3.1885,
+      "step": 775
+    },
+    {
+      "epoch": 3.924969249692497,
+      "grad_norm": 7.569451332092285,
+      "learning_rate": 1.7870370370370373e-05,
+      "loss": 3.1334,
+      "step": 800
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.02627939142461964,
+      "eval_loss": 0.7910987734794617,
+      "eval_macro_f1": 0.3759416595017794,
+      "eval_micro_f1": 0.40867328188166113,
+      "eval_runtime": 0.4274,
+      "eval_samples_per_second": 1691.472,
+      "eval_steps_per_second": 107.618,
+      "eval_weighted_f1": 0.4538096853436305,
+      "step": 816
+    },
+    {
+      "epoch": 4.044280442804428,
+      "grad_norm": 6.28005313873291,
+      "learning_rate": 1.7734204793028323e-05,
+      "loss": 2.9181,
+      "step": 825
+    },
+    {
+      "epoch": 4.167281672816729,
+      "grad_norm": 8.796655654907227,
+      "learning_rate": 1.7598039215686274e-05,
+      "loss": 2.9481,
+      "step": 850
+    },
+    {
+      "epoch": 4.290282902829028,
+      "grad_norm": 7.713093280792236,
+      "learning_rate": 1.7461873638344228e-05,
+      "loss": 3.0248,
+      "step": 875
+    },
+    {
+      "epoch": 4.413284132841328,
+      "grad_norm": 6.511098861694336,
+      "learning_rate": 1.732570806100218e-05,
+      "loss": 2.9415,
+      "step": 900
+    },
+    {
+      "epoch": 4.536285362853628,
+      "grad_norm": 4.849259853363037,
+      "learning_rate": 1.7189542483660133e-05,
+      "loss": 2.8507,
+      "step": 925
+    },
+    {
+      "epoch": 4.659286592865929,
+      "grad_norm": 7.472237586975098,
+      "learning_rate": 1.7053376906318084e-05,
+      "loss": 2.8718,
+      "step": 950
+    },
+    {
+      "epoch": 4.782287822878229,
+      "grad_norm": 6.3157148361206055,
+      "learning_rate": 1.6917211328976034e-05,
+      "loss": 2.8972,
+      "step": 975
+    },
+    {
+      "epoch": 4.905289052890529,
+      "grad_norm": 7.70225191116333,
+      "learning_rate": 1.678104575163399e-05,
+      "loss": 2.871,
+      "step": 1000
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.034578146611341634,
+      "eval_loss": 0.7648706436157227,
+      "eval_macro_f1": 0.4042134872250745,
+      "eval_micro_f1": 0.43454367026496565,
+      "eval_runtime": 0.4066,
+      "eval_samples_per_second": 1778.019,
+      "eval_steps_per_second": 113.124,
+      "eval_weighted_f1": 0.4766167643046378,
+      "step": 1020
+    },
+    {
+      "epoch": 5.02460024600246,
+      "grad_norm": 7.741610050201416,
+      "learning_rate": 1.664488017429194e-05,
+      "loss": 2.7584,
+      "step": 1025
+    },
+    {
+      "epoch": 5.14760147601476,
+      "grad_norm": 7.474596977233887,
+      "learning_rate": 1.6508714596949893e-05,
+      "loss": 2.6958,
+      "step": 1050
+    },
+    {
+      "epoch": 5.27060270602706,
+      "grad_norm": 7.814574718475342,
+      "learning_rate": 1.6372549019607844e-05,
+      "loss": 2.6284,
+      "step": 1075
+    },
+    {
+      "epoch": 5.39360393603936,
+      "grad_norm": 6.433797359466553,
+      "learning_rate": 1.6236383442265798e-05,
+      "loss": 2.6836,
+      "step": 1100
+    },
+    {
+      "epoch": 5.516605166051661,
+      "grad_norm": 4.551105499267578,
+      "learning_rate": 1.610021786492375e-05,
+      "loss": 2.6436,
+      "step": 1125
+    },
+    {
+      "epoch": 5.639606396063961,
+      "grad_norm": 7.557528018951416,
+      "learning_rate": 1.5964052287581703e-05,
+      "loss": 2.6489,
+      "step": 1150
+    },
+    {
+      "epoch": 5.762607626076261,
+      "grad_norm": 9.051876068115234,
+      "learning_rate": 1.5827886710239654e-05,
+      "loss": 2.6136,
+      "step": 1175
+    },
+    {
+      "epoch": 5.885608856088561,
+      "grad_norm": 7.060078144073486,
+      "learning_rate": 1.5691721132897604e-05,
+      "loss": 2.5916,
+      "step": 1200
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.06500691562932227,
+      "eval_loss": 0.7515914440155029,
+      "eval_macro_f1": 0.4233426125745359,
+      "eval_micro_f1": 0.46203208556149733,
+      "eval_runtime": 0.443,
+      "eval_samples_per_second": 1632.187,
+      "eval_steps_per_second": 103.846,
+      "eval_weighted_f1": 0.4926703504033719,
+      "step": 1224
+    },
+    {
+      "epoch": 6.004920049200492,
+      "grad_norm": 3.625218629837036,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 2.5225,
+      "step": 1225
+    },
+    {
+      "epoch": 6.127921279212792,
+      "grad_norm": 5.750714302062988,
+      "learning_rate": 1.541938997821351e-05,
+      "loss": 2.4854,
+      "step": 1250
+    },
+    {
+      "epoch": 6.250922509225092,
+      "grad_norm": 7.495558261871338,
+      "learning_rate": 1.528322440087146e-05,
+      "loss": 2.4865,
+      "step": 1275
+    },
+    {
+      "epoch": 6.373923739237393,
+      "grad_norm": 10.852934837341309,
+      "learning_rate": 1.5147058823529412e-05,
+      "loss": 2.432,
+      "step": 1300
+    },
+    {
+      "epoch": 6.496924969249693,
+      "grad_norm": 7.2275261878967285,
+      "learning_rate": 1.5010893246187366e-05,
+      "loss": 2.4735,
+      "step": 1325
+    },
+    {
+      "epoch": 6.619926199261993,
+      "grad_norm": 12.832423210144043,
+      "learning_rate": 1.4874727668845317e-05,
+      "loss": 2.3393,
+      "step": 1350
+    },
+    {
+      "epoch": 6.7429274292742925,
+      "grad_norm": 5.984768390655518,
+      "learning_rate": 1.473856209150327e-05,
+      "loss": 2.4116,
+      "step": 1375
+    },
+    {
+      "epoch": 6.865928659286593,
+      "grad_norm": 5.80078125,
+      "learning_rate": 1.4602396514161222e-05,
+      "loss": 2.3488,
+      "step": 1400
+    },
+    {
+      "epoch": 6.988929889298893,
+      "grad_norm": 6.97410249710083,
+      "learning_rate": 1.4466230936819172e-05,
+      "loss": 2.3563,
+      "step": 1425
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.08852005532503458,
+      "eval_loss": 0.7310532927513123,
+      "eval_macro_f1": 0.4418108265923872,
+      "eval_micro_f1": 0.48302226219923544,
+      "eval_runtime": 0.422,
+      "eval_samples_per_second": 1713.437,
+      "eval_steps_per_second": 109.015,
+      "eval_weighted_f1": 0.5135287722671217,
+      "step": 1428
+    },
+    {
+      "epoch": 7.108241082410824,
+      "grad_norm": 4.801399230957031,
+      "learning_rate": 1.4330065359477125e-05,
+      "loss": 2.2417,
+      "step": 1450
+    },
+    {
+      "epoch": 7.231242312423125,
+      "grad_norm": 6.208895683288574,
+      "learning_rate": 1.4193899782135077e-05,
+      "loss": 2.187,
+      "step": 1475
+    },
+    {
+      "epoch": 7.354243542435424,
+      "grad_norm": 6.7103190422058105,
+      "learning_rate": 1.405773420479303e-05,
+      "loss": 2.2413,
+      "step": 1500
+    },
+    {
+      "epoch": 7.477244772447724,
+      "grad_norm": 8.126374244689941,
+      "learning_rate": 1.392156862745098e-05,
+      "loss": 2.2087,
+      "step": 1525
+    },
+    {
+      "epoch": 7.6002460024600245,
+      "grad_norm": 15.926607131958008,
+      "learning_rate": 1.3785403050108934e-05,
+      "loss": 2.2598,
+      "step": 1550
+    },
+    {
+      "epoch": 7.723247232472325,
+      "grad_norm": 12.520524978637695,
+      "learning_rate": 1.3649237472766885e-05,
+      "loss": 2.2066,
+      "step": 1575
+    },
+    {
+      "epoch": 7.846248462484625,
+      "grad_norm": 7.5365681648254395,
+      "learning_rate": 1.351307189542484e-05,
+      "loss": 2.1874,
+      "step": 1600
+    },
+    {
+      "epoch": 7.969249692496925,
+      "grad_norm": 9.78396987915039,
+      "learning_rate": 1.337690631808279e-05,
+      "loss": 2.1823,
+      "step": 1625
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.10235131396957123,
+      "eval_loss": 0.7341471314430237,
+      "eval_macro_f1": 0.4605108588723891,
+      "eval_micro_f1": 0.49837587006960554,
+      "eval_runtime": 0.412,
+      "eval_samples_per_second": 1755.046,
+      "eval_steps_per_second": 111.663,
+      "eval_weighted_f1": 0.5262486497475692,
+      "step": 1632
+    },
+    {
+      "epoch": 8.088560885608857,
+      "grad_norm": 8.340449333190918,
+      "learning_rate": 1.3240740740740742e-05,
+      "loss": 2.0857,
+      "step": 1650
+    },
+    {
+      "epoch": 8.211562115621156,
+      "grad_norm": 7.824148178100586,
+      "learning_rate": 1.3104575163398695e-05,
+      "loss": 2.1127,
+      "step": 1675
+    },
+    {
+      "epoch": 8.334563345633457,
+      "grad_norm": 7.566495895385742,
+      "learning_rate": 1.2968409586056645e-05,
+      "loss": 2.0538,
+      "step": 1700
+    },
+    {
+      "epoch": 8.457564575645756,
+      "grad_norm": 6.453400611877441,
+      "learning_rate": 1.2832244008714598e-05,
+      "loss": 2.0553,
+      "step": 1725
+    },
+    {
+      "epoch": 8.580565805658056,
+      "grad_norm": 6.763232231140137,
+      "learning_rate": 1.269607843137255e-05,
+      "loss": 2.0757,
+      "step": 1750
+    },
+    {
+      "epoch": 8.703567035670357,
+      "grad_norm": 7.584934234619141,
+      "learning_rate": 1.2559912854030503e-05,
+      "loss": 2.0295,
+      "step": 1775
+    },
+    {
+      "epoch": 8.826568265682656,
+      "grad_norm": 5.602021217346191,
+      "learning_rate": 1.2423747276688453e-05,
+      "loss": 2.0571,
+      "step": 1800
+    },
+    {
+      "epoch": 8.949569495694957,
+      "grad_norm": 10.217915534973145,
+      "learning_rate": 1.2287581699346407e-05,
+      "loss": 2.0523,
+      "step": 1825
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.1355463347164592,
+      "eval_loss": 0.7010074853897095,
+      "eval_macro_f1": 0.48181601637870647,
+      "eval_micro_f1": 0.5239146431199412,
+      "eval_runtime": 0.4331,
+      "eval_samples_per_second": 1669.531,
+      "eval_steps_per_second": 106.222,
+      "eval_weighted_f1": 0.5459274955079247,
+      "step": 1836
+    },
+    {
+      "epoch": 9.068880688806889,
+      "grad_norm": 10.751814842224121,
+      "learning_rate": 1.2151416122004358e-05,
+      "loss": 1.9293,
+      "step": 1850
+    },
+    {
+      "epoch": 9.191881918819188,
+      "grad_norm": 10.432910919189453,
+      "learning_rate": 1.2015250544662312e-05,
+      "loss": 1.9795,
+      "step": 1875
+    },
+    {
+      "epoch": 9.314883148831488,
+      "grad_norm": 7.785835266113281,
+      "learning_rate": 1.1879084967320263e-05,
+      "loss": 1.9476,
+      "step": 1900
+    },
+    {
+      "epoch": 9.437884378843789,
+      "grad_norm": 5.261160373687744,
+      "learning_rate": 1.1742919389978214e-05,
+      "loss": 1.8623,
+      "step": 1925
+    },
+    {
+      "epoch": 9.560885608856088,
+      "grad_norm": 5.456881523132324,
+      "learning_rate": 1.1606753812636166e-05,
+      "loss": 1.9236,
+      "step": 1950
+    },
+    {
+      "epoch": 9.68388683886839,
+      "grad_norm": 8.351594924926758,
+      "learning_rate": 1.1470588235294118e-05,
+      "loss": 1.9185,
+      "step": 1975
+    },
+    {
+      "epoch": 9.806888068880689,
+      "grad_norm": 6.483122825622559,
+      "learning_rate": 1.133442265795207e-05,
+      "loss": 1.8525,
+      "step": 2000
+    },
+    {
+      "epoch": 9.92988929889299,
+      "grad_norm": 8.558549880981445,
+      "learning_rate": 1.1198257080610021e-05,
+      "loss": 1.9423,
+      "step": 2025
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.1396957123098202,
+      "eval_loss": 0.7128008008003235,
+      "eval_macro_f1": 0.4875816054036432,
+      "eval_micro_f1": 0.5319255907491202,
+      "eval_runtime": 0.4125,
+      "eval_samples_per_second": 1752.67,
+      "eval_steps_per_second": 111.512,
+      "eval_weighted_f1": 0.5514922942549983,
+      "step": 2040
+    },
+    {
+      "epoch": 10.04920049200492,
+      "grad_norm": 6.171204566955566,
+      "learning_rate": 1.1062091503267976e-05,
+      "loss": 1.7874,
+      "step": 2050
+    },
+    {
+      "epoch": 10.17220172201722,
+      "grad_norm": 11.2750883102417,
+      "learning_rate": 1.0925925925925926e-05,
+      "loss": 1.8845,
+      "step": 2075
+    },
+    {
+      "epoch": 10.29520295202952,
+      "grad_norm": 7.489241600036621,
+      "learning_rate": 1.078976034858388e-05,
+      "loss": 1.7853,
+      "step": 2100
+    },
+    {
+      "epoch": 10.418204182041821,
+      "grad_norm": 8.48294448852539,
+      "learning_rate": 1.0653594771241831e-05,
+      "loss": 1.8354,
+      "step": 2125
+    },
+    {
+      "epoch": 10.54120541205412,
+      "grad_norm": 3.9956068992614746,
+      "learning_rate": 1.0517429193899782e-05,
+      "loss": 1.7633,
+      "step": 2150
+    },
+    {
+      "epoch": 10.664206642066421,
+      "grad_norm": 5.930418491363525,
+      "learning_rate": 1.0381263616557736e-05,
+      "loss": 1.7956,
+      "step": 2175
+    },
+    {
+      "epoch": 10.78720787207872,
+      "grad_norm": 6.483989238739014,
+      "learning_rate": 1.0245098039215687e-05,
+      "loss": 1.8058,
+      "step": 2200
+    },
+    {
+      "epoch": 10.91020910209102,
+      "grad_norm": 8.547865867614746,
+      "learning_rate": 1.0108932461873639e-05,
+      "loss": 1.8056,
+      "step": 2225
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.17289073305670816,
+      "eval_loss": 0.7329915165901184,
+      "eval_macro_f1": 0.5138081792281348,
+      "eval_micro_f1": 0.5559449986519277,
+      "eval_runtime": 0.4211,
+      "eval_samples_per_second": 1716.839,
+      "eval_steps_per_second": 109.232,
+      "eval_weighted_f1": 0.5711498727391536,
+      "step": 2244
+    },
+    {
+      "epoch": 11.029520295202952,
+      "grad_norm": 7.899692535400391,
+      "learning_rate": 9.972766884531591e-06,
+      "loss": 1.7049,
+      "step": 2250
+    },
+    {
+      "epoch": 11.152521525215253,
+      "grad_norm": 7.067365646362305,
+      "learning_rate": 9.836601307189542e-06,
+      "loss": 1.7172,
+      "step": 2275
+    },
+    {
+      "epoch": 11.275522755227552,
+      "grad_norm": 6.566714763641357,
+      "learning_rate": 9.700435729847494e-06,
+      "loss": 1.6484,
+      "step": 2300
+    },
+    {
+      "epoch": 11.398523985239853,
+      "grad_norm": 10.03989315032959,
+      "learning_rate": 9.564270152505447e-06,
+      "loss": 1.7287,
+      "step": 2325
+    },
+    {
+      "epoch": 11.521525215252153,
+      "grad_norm": 5.717295169830322,
+      "learning_rate": 9.4281045751634e-06,
+      "loss": 1.6768,
+      "step": 2350
+    },
+    {
+      "epoch": 11.644526445264452,
+      "grad_norm": 5.902117729187012,
+      "learning_rate": 9.291938997821352e-06,
+      "loss": 1.6792,
+      "step": 2375
+    },
+    {
+      "epoch": 11.767527675276753,
+      "grad_norm": 6.71498966217041,
+      "learning_rate": 9.155773420479304e-06,
+      "loss": 1.6388,
+      "step": 2400
+    },
+    {
+      "epoch": 11.890528905289052,
+      "grad_norm": 13.973197937011719,
+      "learning_rate": 9.019607843137256e-06,
+      "loss": 1.6361,
+      "step": 2425
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.17704011065006917,
+      "eval_loss": 0.7418239712715149,
+      "eval_macro_f1": 0.514102228550647,
+      "eval_micro_f1": 0.5543418896777718,
+      "eval_runtime": 0.4568,
+      "eval_samples_per_second": 1582.871,
+      "eval_steps_per_second": 100.708,
+      "eval_weighted_f1": 0.5681533053605753,
+      "step": 2448
+    },
+    {
+      "epoch": 12.009840098400984,
+      "grad_norm": 10.411066055297852,
+      "learning_rate": 8.883442265795209e-06,
+      "loss": 1.6661,
+      "step": 2450
+    },
+    {
+      "epoch": 12.132841328413285,
+      "grad_norm": 5.354274749755859,
+      "learning_rate": 8.74727668845316e-06,
+      "loss": 1.647,
+      "step": 2475
+    },
+    {
+      "epoch": 12.255842558425584,
+      "grad_norm": 7.938675880432129,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 1.5896,
+      "step": 2500
+    },
+    {
+      "epoch": 12.378843788437884,
+      "grad_norm": 8.076502799987793,
+      "learning_rate": 8.474945533769063e-06,
+      "loss": 1.6035,
+      "step": 2525
+    },
+    {
+      "epoch": 12.501845018450185,
+      "grad_norm": 4.712522983551025,
+      "learning_rate": 8.338779956427015e-06,
+      "loss": 1.584,
+      "step": 2550
+    },
+    {
+      "epoch": 12.624846248462484,
+      "grad_norm": 9.102602005004883,
+      "learning_rate": 8.202614379084967e-06,
+      "loss": 1.5626,
+      "step": 2575
+    },
+    {
+      "epoch": 12.747847478474785,
+      "grad_norm": 8.958296775817871,
+      "learning_rate": 8.06644880174292e-06,
+      "loss": 1.5575,
+      "step": 2600
+    },
+    {
+      "epoch": 12.870848708487085,
+      "grad_norm": 3.7900683879852295,
+      "learning_rate": 7.930283224400872e-06,
+      "loss": 1.5755,
+      "step": 2625
+    },
+    {
+      "epoch": 12.993849938499386,
+      "grad_norm": 4.654300689697266,
+      "learning_rate": 7.794117647058825e-06,
+      "loss": 1.6144,
+      "step": 2650
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.18672199170124482,
+      "eval_loss": 0.7420814633369446,
+      "eval_macro_f1": 0.5182252425366539,
+      "eval_micro_f1": 0.565118912797282,
+      "eval_runtime": 0.4279,
+      "eval_samples_per_second": 1689.604,
+      "eval_steps_per_second": 107.499,
+      "eval_weighted_f1": 0.5758384113601707,
+      "step": 2652
+    },
+    {
+      "epoch": 13.113161131611315,
+      "grad_norm": 8.709327697753906,
+      "learning_rate": 7.657952069716777e-06,
+      "loss": 1.4696,
+      "step": 2675
+    },
+    {
+      "epoch": 13.236162361623617,
+      "grad_norm": 4.688850402832031,
+      "learning_rate": 7.5217864923747285e-06,
+      "loss": 1.4891,
+      "step": 2700
+    },
+    {
+      "epoch": 13.359163591635916,
+      "grad_norm": 4.056413173675537,
+      "learning_rate": 7.385620915032681e-06,
+      "loss": 1.5349,
+      "step": 2725
+    },
+    {
+      "epoch": 13.482164821648217,
+      "grad_norm": 10.238755226135254,
+      "learning_rate": 7.249455337690633e-06,
+      "loss": 1.5372,
+      "step": 2750
+    },
+    {
+      "epoch": 13.605166051660516,
+      "grad_norm": 5.842587471008301,
+      "learning_rate": 7.113289760348584e-06,
+      "loss": 1.4805,
+      "step": 2775
+    },
+    {
+      "epoch": 13.728167281672818,
+      "grad_norm": 6.376543998718262,
+      "learning_rate": 6.977124183006536e-06,
+      "loss": 1.5407,
+      "step": 2800
+    },
+    {
+      "epoch": 13.851168511685117,
+      "grad_norm": 7.761598110198975,
+      "learning_rate": 6.840958605664488e-06,
+      "loss": 1.4828,
+      "step": 2825
+    },
+    {
+      "epoch": 13.974169741697416,
+      "grad_norm": 4.524773120880127,
+      "learning_rate": 6.70479302832244e-06,
+      "loss": 1.5614,
+      "step": 2850
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.19225449515905949,
+      "eval_loss": 0.7390348315238953,
+      "eval_macro_f1": 0.5248361915567423,
+      "eval_micro_f1": 0.5727298605180757,
+      "eval_runtime": 0.4194,
+      "eval_samples_per_second": 1723.824,
+      "eval_steps_per_second": 109.676,
+      "eval_weighted_f1": 0.5843580985419131,
+      "step": 2856
+    },
+    {
+      "epoch": 14.093480934809348,
+      "grad_norm": 5.846768856048584,
+      "learning_rate": 6.568627450980393e-06,
+      "loss": 1.4189,
+      "step": 2875
+    },
+    {
+      "epoch": 14.216482164821649,
+      "grad_norm": 6.395859718322754,
+      "learning_rate": 6.432461873638345e-06,
+      "loss": 1.4489,
+      "step": 2900
+    },
+    {
+      "epoch": 14.339483394833948,
+      "grad_norm": 7.327369213104248,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 1.4864,
+      "step": 2925
+    },
+    {
+      "epoch": 14.46248462484625,
+      "grad_norm": 4.31693696975708,
+      "learning_rate": 6.160130718954249e-06,
+      "loss": 1.4329,
+      "step": 2950
+    },
+    {
+      "epoch": 14.585485854858549,
+      "grad_norm": 7.08527946472168,
+      "learning_rate": 6.0239651416122014e-06,
+      "loss": 1.425,
+      "step": 2975
+    },
+    {
+      "epoch": 14.708487084870848,
+      "grad_norm": 4.74684476852417,
+      "learning_rate": 5.887799564270154e-06,
+      "loss": 1.4275,
+      "step": 3000
+    },
+    {
+      "epoch": 14.83148831488315,
+      "grad_norm": 13.72156047821045,
+      "learning_rate": 5.7516339869281045e-06,
+      "loss": 1.4445,
+      "step": 3025
+    },
+    {
+      "epoch": 14.954489544895448,
+      "grad_norm": 9.880606651306152,
+      "learning_rate": 5.615468409586057e-06,
+      "loss": 1.4172,
+      "step": 3050
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.18118948824343015,
+      "eval_loss": 0.7427704334259033,
+      "eval_macro_f1": 0.5143420941342366,
+      "eval_micro_f1": 0.5656338028169015,
+      "eval_runtime": 0.4638,
+      "eval_samples_per_second": 1558.864,
+      "eval_steps_per_second": 99.181,
+      "eval_weighted_f1": 0.5768156402908098,
+      "step": 3060
+    },
+    {
+      "epoch": 15.07380073800738,
+      "grad_norm": 16.806764602661133,
+      "learning_rate": 5.479302832244009e-06,
+      "loss": 1.3593,
+      "step": 3075
+    },
+    {
+      "epoch": 15.196801968019681,
+      "grad_norm": 6.311575889587402,
+      "learning_rate": 5.343137254901961e-06,
+      "loss": 1.4274,
+      "step": 3100
+    },
+    {
+      "epoch": 15.31980319803198,
+      "grad_norm": 13.33009147644043,
+      "learning_rate": 5.206971677559913e-06,
+      "loss": 1.3595,
+      "step": 3125
+    },
+    {
+      "epoch": 15.44280442804428,
+      "grad_norm": 4.078766822814941,
+      "learning_rate": 5.070806100217866e-06,
+      "loss": 1.4074,
+      "step": 3150
+    },
+    {
+      "epoch": 15.565805658056581,
+      "grad_norm": 10.644097328186035,
+      "learning_rate": 4.934640522875817e-06,
+      "loss": 1.4424,
+      "step": 3175
+    },
+    {
+      "epoch": 15.68880688806888,
+      "grad_norm": 5.109310626983643,
+      "learning_rate": 4.79847494553377e-06,
+      "loss": 1.3553,
+      "step": 3200
+    },
+    {
+      "epoch": 15.811808118081181,
+      "grad_norm": 7.613893985748291,
+      "learning_rate": 4.662309368191721e-06,
+      "loss": 1.3812,
+      "step": 3225
+    },
+    {
+      "epoch": 15.93480934809348,
+      "grad_norm": 6.8762946128845215,
+      "learning_rate": 4.5261437908496736e-06,
+      "loss": 1.4065,
+      "step": 3250
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.1825726141078838,
+      "eval_loss": 0.7540271282196045,
+      "eval_macro_f1": 0.5191546798139748,
+      "eval_micro_f1": 0.5683185333715268,
+      "eval_runtime": 0.4403,
+      "eval_samples_per_second": 1642.168,
+      "eval_steps_per_second": 104.481,
+      "eval_weighted_f1": 0.5804307137894656,
+      "step": 3264
+    },
+    {
+      "epoch": 16.05412054120541,
+      "grad_norm": 13.340099334716797,
+      "learning_rate": 4.389978213507626e-06,
+      "loss": 1.3355,
+      "step": 3275
+    },
+    {
+      "epoch": 16.177121771217713,
+      "grad_norm": 5.658074855804443,
+      "learning_rate": 4.253812636165578e-06,
+      "loss": 1.3459,
+      "step": 3300
+    },
+    {
+      "epoch": 16.300123001230013,
+      "grad_norm": 7.94587516784668,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 1.3617,
+      "step": 3325
+    },
+    {
+      "epoch": 16.423124231242312,
+      "grad_norm": 4.613985061645508,
+      "learning_rate": 3.9814814814814814e-06,
+      "loss": 1.3341,
+      "step": 3350
+    },
+    {
+      "epoch": 16.54612546125461,
+      "grad_norm": 7.527010917663574,
+      "learning_rate": 3.845315904139434e-06,
+      "loss": 1.3143,
+      "step": 3375
+    },
+    {
+      "epoch": 16.669126691266914,
+      "grad_norm": 9.891593933105469,
+      "learning_rate": 3.7091503267973862e-06,
+      "loss": 1.4211,
+      "step": 3400
+    },
+    {
+      "epoch": 16.792127921279214,
+      "grad_norm": 6.552762031555176,
+      "learning_rate": 3.572984749455338e-06,
+      "loss": 1.3691,
+      "step": 3425
+    },
+    {
+      "epoch": 16.915129151291513,
+      "grad_norm": 5.981565475463867,
+      "learning_rate": 3.4368191721132898e-06,
+      "loss": 1.2688,
+      "step": 3450
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.21023513139695713,
+      "eval_loss": 0.7745928764343262,
+      "eval_macro_f1": 0.5334202111120329,
+      "eval_micro_f1": 0.581764705882353,
+      "eval_runtime": 0.4488,
+      "eval_samples_per_second": 1610.932,
+      "eval_steps_per_second": 102.494,
+      "eval_weighted_f1": 0.5901403459769167,
+      "step": 3468
+    },
+    {
+      "epoch": 17.034440344403443,
+      "grad_norm": 6.895166397094727,
+      "learning_rate": 3.300653594771242e-06,
+      "loss": 1.2851,
+      "step": 3475
+    },
+    {
+      "epoch": 17.157441574415746,
+      "grad_norm": 4.59776496887207,
+      "learning_rate": 3.164488017429194e-06,
+      "loss": 1.3046,
+      "step": 3500
+    },
+    {
+      "epoch": 17.280442804428045,
+      "grad_norm": 4.178014278411865,
+      "learning_rate": 3.0283224400871465e-06,
+      "loss": 1.3453,
+      "step": 3525
+    },
+    {
+      "epoch": 17.403444034440344,
+      "grad_norm": 3.9644834995269775,
+      "learning_rate": 2.8921568627450985e-06,
+      "loss": 1.3331,
+      "step": 3550
+    },
+    {
+      "epoch": 17.526445264452644,
+      "grad_norm": 6.1651201248168945,
+      "learning_rate": 2.75599128540305e-06,
+      "loss": 1.3013,
+      "step": 3575
+    },
+    {
+      "epoch": 17.649446494464943,
+      "grad_norm": 6.653693199157715,
+      "learning_rate": 2.6198257080610024e-06,
+      "loss": 1.3299,
+      "step": 3600
+    },
+    {
+      "epoch": 17.772447724477246,
+      "grad_norm": 4.887754917144775,
+      "learning_rate": 2.4836601307189544e-06,
+      "loss": 1.2617,
+      "step": 3625
+    },
+    {
+      "epoch": 17.895448954489545,
+      "grad_norm": 9.397178649902344,
+      "learning_rate": 2.3474945533769064e-06,
+      "loss": 1.3397,
+      "step": 3650
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.19778699861687413,
+      "eval_loss": 0.7525197267532349,
+      "eval_macro_f1": 0.5230016464764999,
+      "eval_micro_f1": 0.574739281575898,
+      "eval_runtime": 0.4704,
+      "eval_samples_per_second": 1536.984,
+      "eval_steps_per_second": 97.789,
+      "eval_weighted_f1": 0.5859785736459122,
+      "step": 3672
+    },
+    {
+      "epoch": 18.014760147601475,
+      "grad_norm": 6.480631351470947,
+      "learning_rate": 2.2113289760348588e-06,
+      "loss": 1.2651,
+      "step": 3675
+    },
+    {
+      "epoch": 18.137761377613778,
+      "grad_norm": 4.3607707023620605,
+      "learning_rate": 2.0751633986928107e-06,
+      "loss": 1.2688,
+      "step": 3700
+    },
+    {
+      "epoch": 18.260762607626077,
+      "grad_norm": 6.317516803741455,
+      "learning_rate": 1.9389978213507627e-06,
+      "loss": 1.3253,
+      "step": 3725
+    },
+    {
+      "epoch": 18.383763837638377,
+      "grad_norm": 3.8014440536499023,
+      "learning_rate": 1.802832244008715e-06,
+      "loss": 1.3243,
+      "step": 3750
+    },
+    {
+      "epoch": 18.506765067650676,
+      "grad_norm": 22.600187301635742,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.2883,
+      "step": 3775
+    },
+    {
+      "epoch": 18.629766297662975,
+      "grad_norm": 9.801239013671875,
+      "learning_rate": 1.5305010893246188e-06,
+      "loss": 1.2581,
+      "step": 3800
+    },
+    {
+      "epoch": 18.752767527675278,
+      "grad_norm": 13.6274995803833,
+      "learning_rate": 1.394335511982571e-06,
+      "loss": 1.2963,
+      "step": 3825
+    },
+    {
+      "epoch": 18.875768757687577,
+      "grad_norm": 6.531538963317871,
+      "learning_rate": 1.258169934640523e-06,
+      "loss": 1.2437,
+      "step": 3850
+    },
+    {
+      "epoch": 18.998769987699877,
+      "grad_norm": 7.813934326171875,
+      "learning_rate": 1.122004357298475e-06,
+      "loss": 1.3025,
+      "step": 3875
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.20193637621023514,
+      "eval_loss": 0.766295313835144,
+      "eval_macro_f1": 0.5272206155524268,
+      "eval_micro_f1": 0.5776997366110623,
+      "eval_runtime": 0.4575,
+      "eval_samples_per_second": 1580.481,
+      "eval_steps_per_second": 100.556,
+      "eval_weighted_f1": 0.5870572148716042,
+      "step": 3876
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 4080,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 2
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

ood-vigoemotions-phobert-source-source-00-s43-20260826T134203Z-1f9ffdf3/trainer/checkpoint-4080/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

ood-vigoemotions-phobert-source-source-00-s43-20260826T134203Z-1f9ffdf3/trainer/checkpoint-4080/special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-00-s43-20260826T134203Z-1f9ffdf3/trainer/checkpoint-4080/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 256,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-00-s43-20260826T134203Z-1f9ffdf3/trainer/checkpoint-4080/trainer_state.json

@@ -0,0 +1,1424 @@
+{
+  "best_global_step": 4080,
+  "best_metric": 0.533909530236441,
+  "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vigoemotions-phobert-source-source-00-s43-20260826T134203Z-1f9ffdf3/trainer/checkpoint-4080",
+  "epoch": 20.0,
+  "eval_steps": 500,
+  "global_step": 4080,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.12300123001230012,
+      "grad_norm": 8.09683895111084,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 5.2456,
+      "step": 25
+    },
+    {
+      "epoch": 0.24600246002460024,
+      "grad_norm": 5.68751859664917,
+      "learning_rate": 2.401960784313726e-06,
+      "loss": 5.2497,
+      "step": 50
+    },
+    {
+      "epoch": 0.36900369003690037,
+      "grad_norm": 306.9134216308594,
+      "learning_rate": 3.6274509803921573e-06,
+      "loss": 5.1922,
+      "step": 75
+    },
+    {
+      "epoch": 0.4920049200492005,
+      "grad_norm": 5.2206339836120605,
+      "learning_rate": 4.852941176470589e-06,
+      "loss": 5.2109,
+      "step": 100
+    },
+    {
+      "epoch": 0.6150061500615006,
+      "grad_norm": 6.440758228302002,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 5.1985,
+      "step": 125
+    },
+    {
+      "epoch": 0.7380073800738007,
+      "grad_norm": 5.165765762329102,
+      "learning_rate": 7.3039215686274515e-06,
+      "loss": 5.0125,
+      "step": 150
+    },
+    {
+      "epoch": 0.8610086100861009,
+      "grad_norm": 7.758053302764893,
+      "learning_rate": 8.529411764705883e-06,
+      "loss": 5.0253,
+      "step": 175
+    },
+    {
+      "epoch": 0.984009840098401,
+      "grad_norm": 10.424230575561523,
+      "learning_rate": 9.754901960784315e-06,
+      "loss": 4.9213,
+      "step": 200
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 1.194104790687561,
+      "eval_macro_f1": 0.20881396339551378,
+      "eval_micro_f1": 0.2245830234781685,
+      "eval_runtime": 0.4229,
+      "eval_samples_per_second": 1709.77,
+      "eval_steps_per_second": 108.782,
+      "eval_weighted_f1": 0.2739720167784619,
+      "step": 204
+    },
+    {
+      "epoch": 1.103321033210332,
+      "grad_norm": 5.252875804901123,
+      "learning_rate": 1.0980392156862747e-05,
+      "loss": 4.7508,
+      "step": 225
+    },
+    {
+      "epoch": 1.2263222632226323,
+      "grad_norm": 5.531340599060059,
+      "learning_rate": 1.2205882352941179e-05,
+      "loss": 4.4645,
+      "step": 250
+    },
+    {
+      "epoch": 1.3493234932349323,
+      "grad_norm": 7.683099746704102,
+      "learning_rate": 1.3431372549019607e-05,
+      "loss": 4.5754,
+      "step": 275
+    },
+    {
+      "epoch": 1.4723247232472325,
+      "grad_norm": 8.739664077758789,
+      "learning_rate": 1.4656862745098039e-05,
+      "loss": 4.5256,
+      "step": 300
+    },
+    {
+      "epoch": 1.5953259532595325,
+      "grad_norm": 8.408071517944336,
+      "learning_rate": 1.5882352941176473e-05,
+      "loss": 4.3144,
+      "step": 325
+    },
+    {
+      "epoch": 1.7183271832718328,
+      "grad_norm": 5.407787322998047,
+      "learning_rate": 1.7107843137254903e-05,
+      "loss": 4.1639,
+      "step": 350
+    },
+    {
+      "epoch": 1.841328413284133,
+      "grad_norm": 6.228161811828613,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 4.0869,
+      "step": 375
+    },
+    {
+      "epoch": 1.964329643296433,
+      "grad_norm": 5.647112846374512,
+      "learning_rate": 1.9558823529411766e-05,
+      "loss": 3.9973,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.0013831258644536654,
+      "eval_loss": 0.9395498037338257,
+      "eval_macro_f1": 0.3101064655101556,
+      "eval_micro_f1": 0.3253012048192771,
+      "eval_runtime": 0.4162,
+      "eval_samples_per_second": 1736.959,
+      "eval_steps_per_second": 110.512,
+      "eval_weighted_f1": 0.38850470823525307,
+      "step": 408
+    },
+    {
+      "epoch": 2.0836408364083643,
+      "grad_norm": 5.767376899719238,
+      "learning_rate": 1.9912854030501092e-05,
+      "loss": 3.7444,
+      "step": 425
+    },
+    {
+      "epoch": 2.206642066420664,
+      "grad_norm": 8.322330474853516,
+      "learning_rate": 1.9776688453159043e-05,
+      "loss": 3.6566,
+      "step": 450
+    },
+    {
+      "epoch": 2.3296432964329643,
+      "grad_norm": 12.773788452148438,
+      "learning_rate": 1.9640522875816993e-05,
+      "loss": 3.8016,
+      "step": 475
+    },
+    {
+      "epoch": 2.4526445264452645,
+      "grad_norm": 12.092753410339355,
+      "learning_rate": 1.9504357298474947e-05,
+      "loss": 3.6083,
+      "step": 500
+    },
+    {
+      "epoch": 2.5756457564575648,
+      "grad_norm": 13.042563438415527,
+      "learning_rate": 1.9368191721132898e-05,
+      "loss": 3.6308,
+      "step": 525
+    },
+    {
+      "epoch": 2.6986469864698646,
+      "grad_norm": 6.5783562660217285,
+      "learning_rate": 1.9232026143790852e-05,
+      "loss": 3.5555,
+      "step": 550
+    },
+    {
+      "epoch": 2.821648216482165,
+      "grad_norm": 14.115478515625,
+      "learning_rate": 1.9095860566448803e-05,
+      "loss": 3.5335,
+      "step": 575
+    },
+    {
+      "epoch": 2.944649446494465,
+      "grad_norm": 16.671173095703125,
+      "learning_rate": 1.8959694989106757e-05,
+      "loss": 3.4805,
+      "step": 600
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.011065006915629323,
+      "eval_loss": 0.8417011499404907,
+      "eval_macro_f1": 0.3377949019335999,
+      "eval_micro_f1": 0.3597916995423702,
+      "eval_runtime": 0.4516,
+      "eval_samples_per_second": 1601.102,
+      "eval_steps_per_second": 101.868,
+      "eval_weighted_f1": 0.4160828254989279,
+      "step": 612
+    },
+    {
+      "epoch": 3.063960639606396,
+      "grad_norm": 5.159065246582031,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 3.3403,
+      "step": 625
+    },
+    {
+      "epoch": 3.1869618696186963,
+      "grad_norm": 5.292785167694092,
+      "learning_rate": 1.8687363834422662e-05,
+      "loss": 3.2876,
+      "step": 650
+    },
+    {
+      "epoch": 3.3099630996309966,
+      "grad_norm": 10.624845504760742,
+      "learning_rate": 1.8551198257080612e-05,
+      "loss": 3.2157,
+      "step": 675
+    },
+    {
+      "epoch": 3.4329643296432963,
+      "grad_norm": 7.054944038391113,
+      "learning_rate": 1.8415032679738563e-05,
+      "loss": 3.2664,
+      "step": 700
+    },
+    {
+      "epoch": 3.5559655596555966,
+      "grad_norm": 4.8366522789001465,
+      "learning_rate": 1.8278867102396517e-05,
+      "loss": 3.2077,
+      "step": 725
+    },
+    {
+      "epoch": 3.678966789667897,
+      "grad_norm": 10.388249397277832,
+      "learning_rate": 1.8142701525054468e-05,
+      "loss": 3.2117,
+      "step": 750
+    },
+    {
+      "epoch": 3.8019680196801966,
+      "grad_norm": 7.776172161102295,
+      "learning_rate": 1.800653594771242e-05,
+      "loss": 3.1885,
+      "step": 775
+    },
+    {
+      "epoch": 3.924969249692497,
+      "grad_norm": 7.569451332092285,
+      "learning_rate": 1.7870370370370373e-05,
+      "loss": 3.1334,
+      "step": 800
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.02627939142461964,
+      "eval_loss": 0.7910987734794617,
+      "eval_macro_f1": 0.3759416595017794,
+      "eval_micro_f1": 0.40867328188166113,
+      "eval_runtime": 0.4274,
+      "eval_samples_per_second": 1691.472,
+      "eval_steps_per_second": 107.618,
+      "eval_weighted_f1": 0.4538096853436305,
+      "step": 816
+    },
+    {
+      "epoch": 4.044280442804428,
+      "grad_norm": 6.28005313873291,
+      "learning_rate": 1.7734204793028323e-05,
+      "loss": 2.9181,
+      "step": 825
+    },
+    {
+      "epoch": 4.167281672816729,
+      "grad_norm": 8.796655654907227,
+      "learning_rate": 1.7598039215686274e-05,
+      "loss": 2.9481,
+      "step": 850
+    },
+    {
+      "epoch": 4.290282902829028,
+      "grad_norm": 7.713093280792236,
+      "learning_rate": 1.7461873638344228e-05,
+      "loss": 3.0248,
+      "step": 875
+    },
+    {
+      "epoch": 4.413284132841328,
+      "grad_norm": 6.511098861694336,
+      "learning_rate": 1.732570806100218e-05,
+      "loss": 2.9415,
+      "step": 900
+    },
+    {
+      "epoch": 4.536285362853628,
+      "grad_norm": 4.849259853363037,
+      "learning_rate": 1.7189542483660133e-05,
+      "loss": 2.8507,
+      "step": 925
+    },
+    {
+      "epoch": 4.659286592865929,
+      "grad_norm": 7.472237586975098,
+      "learning_rate": 1.7053376906318084e-05,
+      "loss": 2.8718,
+      "step": 950
+    },
+    {
+      "epoch": 4.782287822878229,
+      "grad_norm": 6.3157148361206055,
+      "learning_rate": 1.6917211328976034e-05,
+      "loss": 2.8972,
+      "step": 975
+    },
+    {
+      "epoch": 4.905289052890529,
+      "grad_norm": 7.70225191116333,
+      "learning_rate": 1.678104575163399e-05,
+      "loss": 2.871,
+      "step": 1000
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.034578146611341634,
+      "eval_loss": 0.7648706436157227,
+      "eval_macro_f1": 0.4042134872250745,
+      "eval_micro_f1": 0.43454367026496565,
+      "eval_runtime": 0.4066,
+      "eval_samples_per_second": 1778.019,
+      "eval_steps_per_second": 113.124,
+      "eval_weighted_f1": 0.4766167643046378,
+      "step": 1020
+    },
+    {
+      "epoch": 5.02460024600246,
+      "grad_norm": 7.741610050201416,
+      "learning_rate": 1.664488017429194e-05,
+      "loss": 2.7584,
+      "step": 1025
+    },
+    {
+      "epoch": 5.14760147601476,
+      "grad_norm": 7.474596977233887,
+      "learning_rate": 1.6508714596949893e-05,
+      "loss": 2.6958,
+      "step": 1050
+    },
+    {
+      "epoch": 5.27060270602706,
+      "grad_norm": 7.814574718475342,
+      "learning_rate": 1.6372549019607844e-05,
+      "loss": 2.6284,
+      "step": 1075
+    },
+    {
+      "epoch": 5.39360393603936,
+      "grad_norm": 6.433797359466553,
+      "learning_rate": 1.6236383442265798e-05,
+      "loss": 2.6836,
+      "step": 1100
+    },
+    {
+      "epoch": 5.516605166051661,
+      "grad_norm": 4.551105499267578,
+      "learning_rate": 1.610021786492375e-05,
+      "loss": 2.6436,
+      "step": 1125
+    },
+    {
+      "epoch": 5.639606396063961,
+      "grad_norm": 7.557528018951416,
+      "learning_rate": 1.5964052287581703e-05,
+      "loss": 2.6489,
+      "step": 1150
+    },
+    {
+      "epoch": 5.762607626076261,
+      "grad_norm": 9.051876068115234,
+      "learning_rate": 1.5827886710239654e-05,
+      "loss": 2.6136,
+      "step": 1175
+    },
+    {
+      "epoch": 5.885608856088561,
+      "grad_norm": 7.060078144073486,
+      "learning_rate": 1.5691721132897604e-05,
+      "loss": 2.5916,
+      "step": 1200
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.06500691562932227,
+      "eval_loss": 0.7515914440155029,
+      "eval_macro_f1": 0.4233426125745359,
+      "eval_micro_f1": 0.46203208556149733,
+      "eval_runtime": 0.443,
+      "eval_samples_per_second": 1632.187,
+      "eval_steps_per_second": 103.846,
+      "eval_weighted_f1": 0.4926703504033719,
+      "step": 1224
+    },
+    {
+      "epoch": 6.004920049200492,
+      "grad_norm": 3.625218629837036,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 2.5225,
+      "step": 1225
+    },
+    {
+      "epoch": 6.127921279212792,
+      "grad_norm": 5.750714302062988,
+      "learning_rate": 1.541938997821351e-05,
+      "loss": 2.4854,
+      "step": 1250
+    },
+    {
+      "epoch": 6.250922509225092,
+      "grad_norm": 7.495558261871338,
+      "learning_rate": 1.528322440087146e-05,
+      "loss": 2.4865,
+      "step": 1275
+    },
+    {
+      "epoch": 6.373923739237393,
+      "grad_norm": 10.852934837341309,
+      "learning_rate": 1.5147058823529412e-05,
+      "loss": 2.432,
+      "step": 1300
+    },
+    {
+      "epoch": 6.496924969249693,
+      "grad_norm": 7.2275261878967285,
+      "learning_rate": 1.5010893246187366e-05,
+      "loss": 2.4735,
+      "step": 1325
+    },
+    {
+      "epoch": 6.619926199261993,
+      "grad_norm": 12.832423210144043,
+      "learning_rate": 1.4874727668845317e-05,
+      "loss": 2.3393,
+      "step": 1350
+    },
+    {
+      "epoch": 6.7429274292742925,
+      "grad_norm": 5.984768390655518,
+      "learning_rate": 1.473856209150327e-05,
+      "loss": 2.4116,
+      "step": 1375
+    },
+    {
+      "epoch": 6.865928659286593,
+      "grad_norm": 5.80078125,
+      "learning_rate": 1.4602396514161222e-05,
+      "loss": 2.3488,
+      "step": 1400
+    },
+    {
+      "epoch": 6.988929889298893,
+      "grad_norm": 6.97410249710083,
+      "learning_rate": 1.4466230936819172e-05,
+      "loss": 2.3563,
+      "step": 1425
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.08852005532503458,
+      "eval_loss": 0.7310532927513123,
+      "eval_macro_f1": 0.4418108265923872,
+      "eval_micro_f1": 0.48302226219923544,
+      "eval_runtime": 0.422,
+      "eval_samples_per_second": 1713.437,
+      "eval_steps_per_second": 109.015,
+      "eval_weighted_f1": 0.5135287722671217,
+      "step": 1428
+    },
+    {
+      "epoch": 7.108241082410824,
+      "grad_norm": 4.801399230957031,
+      "learning_rate": 1.4330065359477125e-05,
+      "loss": 2.2417,
+      "step": 1450
+    },
+    {
+      "epoch": 7.231242312423125,
+      "grad_norm": 6.208895683288574,
+      "learning_rate": 1.4193899782135077e-05,
+      "loss": 2.187,
+      "step": 1475
+    },
+    {
+      "epoch": 7.354243542435424,
+      "grad_norm": 6.7103190422058105,
+      "learning_rate": 1.405773420479303e-05,
+      "loss": 2.2413,
+      "step": 1500
+    },
+    {
+      "epoch": 7.477244772447724,
+      "grad_norm": 8.126374244689941,
+      "learning_rate": 1.392156862745098e-05,
+      "loss": 2.2087,
+      "step": 1525
+    },
+    {
+      "epoch": 7.6002460024600245,
+      "grad_norm": 15.926607131958008,
+      "learning_rate": 1.3785403050108934e-05,
+      "loss": 2.2598,
+      "step": 1550
+    },
+    {
+      "epoch": 7.723247232472325,
+      "grad_norm": 12.520524978637695,
+      "learning_rate": 1.3649237472766885e-05,
+      "loss": 2.2066,
+      "step": 1575
+    },
+    {
+      "epoch": 7.846248462484625,
+      "grad_norm": 7.5365681648254395,
+      "learning_rate": 1.351307189542484e-05,
+      "loss": 2.1874,
+      "step": 1600
+    },
+    {
+      "epoch": 7.969249692496925,
+      "grad_norm": 9.78396987915039,
+      "learning_rate": 1.337690631808279e-05,
+      "loss": 2.1823,
+      "step": 1625
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.10235131396957123,
+      "eval_loss": 0.7341471314430237,
+      "eval_macro_f1": 0.4605108588723891,
+      "eval_micro_f1": 0.49837587006960554,
+      "eval_runtime": 0.412,
+      "eval_samples_per_second": 1755.046,
+      "eval_steps_per_second": 111.663,
+      "eval_weighted_f1": 0.5262486497475692,
+      "step": 1632
+    },
+    {
+      "epoch": 8.088560885608857,
+      "grad_norm": 8.340449333190918,
+      "learning_rate": 1.3240740740740742e-05,
+      "loss": 2.0857,
+      "step": 1650
+    },
+    {
+      "epoch": 8.211562115621156,
+      "grad_norm": 7.824148178100586,
+      "learning_rate": 1.3104575163398695e-05,
+      "loss": 2.1127,
+      "step": 1675
+    },
+    {
+      "epoch": 8.334563345633457,
+      "grad_norm": 7.566495895385742,
+      "learning_rate": 1.2968409586056645e-05,
+      "loss": 2.0538,
+      "step": 1700
+    },
+    {
+      "epoch": 8.457564575645756,
+      "grad_norm": 6.453400611877441,
+      "learning_rate": 1.2832244008714598e-05,
+      "loss": 2.0553,
+      "step": 1725
+    },
+    {
+      "epoch": 8.580565805658056,
+      "grad_norm": 6.763232231140137,
+      "learning_rate": 1.269607843137255e-05,
+      "loss": 2.0757,
+      "step": 1750
+    },
+    {
+      "epoch": 8.703567035670357,
+      "grad_norm": 7.584934234619141,
+      "learning_rate": 1.2559912854030503e-05,
+      "loss": 2.0295,
+      "step": 1775
+    },
+    {
+      "epoch": 8.826568265682656,
+      "grad_norm": 5.602021217346191,
+      "learning_rate": 1.2423747276688453e-05,
+      "loss": 2.0571,
+      "step": 1800
+    },
+    {
+      "epoch": 8.949569495694957,
+      "grad_norm": 10.217915534973145,
+      "learning_rate": 1.2287581699346407e-05,
+      "loss": 2.0523,
+      "step": 1825
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.1355463347164592,
+      "eval_loss": 0.7010074853897095,
+      "eval_macro_f1": 0.48181601637870647,
+      "eval_micro_f1": 0.5239146431199412,
+      "eval_runtime": 0.4331,
+      "eval_samples_per_second": 1669.531,
+      "eval_steps_per_second": 106.222,
+      "eval_weighted_f1": 0.5459274955079247,
+      "step": 1836
+    },
+    {
+      "epoch": 9.068880688806889,
+      "grad_norm": 10.751814842224121,
+      "learning_rate": 1.2151416122004358e-05,
+      "loss": 1.9293,
+      "step": 1850
+    },
+    {
+      "epoch": 9.191881918819188,
+      "grad_norm": 10.432910919189453,
+      "learning_rate": 1.2015250544662312e-05,
+      "loss": 1.9795,
+      "step": 1875
+    },
+    {
+      "epoch": 9.314883148831488,
+      "grad_norm": 7.785835266113281,
+      "learning_rate": 1.1879084967320263e-05,
+      "loss": 1.9476,
+      "step": 1900
+    },
+    {
+      "epoch": 9.437884378843789,
+      "grad_norm": 5.261160373687744,
+      "learning_rate": 1.1742919389978214e-05,
+      "loss": 1.8623,
+      "step": 1925
+    },
+    {
+      "epoch": 9.560885608856088,
+      "grad_norm": 5.456881523132324,
+      "learning_rate": 1.1606753812636166e-05,
+      "loss": 1.9236,
+      "step": 1950
+    },
+    {
+      "epoch": 9.68388683886839,
+      "grad_norm": 8.351594924926758,
+      "learning_rate": 1.1470588235294118e-05,
+      "loss": 1.9185,
+      "step": 1975
+    },
+    {
+      "epoch": 9.806888068880689,
+      "grad_norm": 6.483122825622559,
+      "learning_rate": 1.133442265795207e-05,
+      "loss": 1.8525,
+      "step": 2000
+    },
+    {
+      "epoch": 9.92988929889299,
+      "grad_norm": 8.558549880981445,
+      "learning_rate": 1.1198257080610021e-05,
+      "loss": 1.9423,
+      "step": 2025
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.1396957123098202,
+      "eval_loss": 0.7128008008003235,
+      "eval_macro_f1": 0.4875816054036432,
+      "eval_micro_f1": 0.5319255907491202,
+      "eval_runtime": 0.4125,
+      "eval_samples_per_second": 1752.67,
+      "eval_steps_per_second": 111.512,
+      "eval_weighted_f1": 0.5514922942549983,
+      "step": 2040
+    },
+    {
+      "epoch": 10.04920049200492,
+      "grad_norm": 6.171204566955566,
+      "learning_rate": 1.1062091503267976e-05,
+      "loss": 1.7874,
+      "step": 2050
+    },
+    {
+      "epoch": 10.17220172201722,
+      "grad_norm": 11.2750883102417,
+      "learning_rate": 1.0925925925925926e-05,
+      "loss": 1.8845,
+      "step": 2075
+    },
+    {
+      "epoch": 10.29520295202952,
+      "grad_norm": 7.489241600036621,
+      "learning_rate": 1.078976034858388e-05,
+      "loss": 1.7853,
+      "step": 2100
+    },
+    {
+      "epoch": 10.418204182041821,
+      "grad_norm": 8.48294448852539,
+      "learning_rate": 1.0653594771241831e-05,
+      "loss": 1.8354,
+      "step": 2125
+    },
+    {
+      "epoch": 10.54120541205412,
+      "grad_norm": 3.9956068992614746,
+      "learning_rate": 1.0517429193899782e-05,
+      "loss": 1.7633,
+      "step": 2150
+    },
+    {
+      "epoch": 10.664206642066421,
+      "grad_norm": 5.930418491363525,
+      "learning_rate": 1.0381263616557736e-05,
+      "loss": 1.7956,
+      "step": 2175
+    },
+    {
+      "epoch": 10.78720787207872,
+      "grad_norm": 6.483989238739014,
+      "learning_rate": 1.0245098039215687e-05,
+      "loss": 1.8058,
+      "step": 2200
+    },
+    {
+      "epoch": 10.91020910209102,
+      "grad_norm": 8.547865867614746,
+      "learning_rate": 1.0108932461873639e-05,
+      "loss": 1.8056,
+      "step": 2225
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.17289073305670816,
+      "eval_loss": 0.7329915165901184,
+      "eval_macro_f1": 0.5138081792281348,
+      "eval_micro_f1": 0.5559449986519277,
+      "eval_runtime": 0.4211,
+      "eval_samples_per_second": 1716.839,
+      "eval_steps_per_second": 109.232,
+      "eval_weighted_f1": 0.5711498727391536,
+      "step": 2244
+    },
+    {
+      "epoch": 11.029520295202952,
+      "grad_norm": 7.899692535400391,
+      "learning_rate": 9.972766884531591e-06,
+      "loss": 1.7049,
+      "step": 2250
+    },
+    {
+      "epoch": 11.152521525215253,
+      "grad_norm": 7.067365646362305,
+      "learning_rate": 9.836601307189542e-06,
+      "loss": 1.7172,
+      "step": 2275
+    },
+    {
+      "epoch": 11.275522755227552,
+      "grad_norm": 6.566714763641357,
+      "learning_rate": 9.700435729847494e-06,
+      "loss": 1.6484,
+      "step": 2300
+    },
+    {
+      "epoch": 11.398523985239853,
+      "grad_norm": 10.03989315032959,
+      "learning_rate": 9.564270152505447e-06,
+      "loss": 1.7287,
+      "step": 2325
+    },
+    {
+      "epoch": 11.521525215252153,
+      "grad_norm": 5.717295169830322,
+      "learning_rate": 9.4281045751634e-06,
+      "loss": 1.6768,
+      "step": 2350
+    },
+    {
+      "epoch": 11.644526445264452,
+      "grad_norm": 5.902117729187012,
+      "learning_rate": 9.291938997821352e-06,
+      "loss": 1.6792,
+      "step": 2375
+    },
+    {
+      "epoch": 11.767527675276753,
+      "grad_norm": 6.71498966217041,
+      "learning_rate": 9.155773420479304e-06,
+      "loss": 1.6388,
+      "step": 2400
+    },
+    {
+      "epoch": 11.890528905289052,
+      "grad_norm": 13.973197937011719,
+      "learning_rate": 9.019607843137256e-06,
+      "loss": 1.6361,
+      "step": 2425
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.17704011065006917,
+      "eval_loss": 0.7418239712715149,
+      "eval_macro_f1": 0.514102228550647,
+      "eval_micro_f1": 0.5543418896777718,
+      "eval_runtime": 0.4568,
+      "eval_samples_per_second": 1582.871,
+      "eval_steps_per_second": 100.708,
+      "eval_weighted_f1": 0.5681533053605753,
+      "step": 2448
+    },
+    {
+      "epoch": 12.009840098400984,
+      "grad_norm": 10.411066055297852,
+      "learning_rate": 8.883442265795209e-06,
+      "loss": 1.6661,
+      "step": 2450
+    },
+    {
+      "epoch": 12.132841328413285,
+      "grad_norm": 5.354274749755859,
+      "learning_rate": 8.74727668845316e-06,
+      "loss": 1.647,
+      "step": 2475
+    },
+    {
+      "epoch": 12.255842558425584,
+      "grad_norm": 7.938675880432129,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 1.5896,
+      "step": 2500
+    },
+    {
+      "epoch": 12.378843788437884,
+      "grad_norm": 8.076502799987793,
+      "learning_rate": 8.474945533769063e-06,
+      "loss": 1.6035,
+      "step": 2525
+    },
+    {
+      "epoch": 12.501845018450185,
+      "grad_norm": 4.712522983551025,
+      "learning_rate": 8.338779956427015e-06,
+      "loss": 1.584,
+      "step": 2550
+    },
+    {
+      "epoch": 12.624846248462484,
+      "grad_norm": 9.102602005004883,
+      "learning_rate": 8.202614379084967e-06,
+      "loss": 1.5626,
+      "step": 2575
+    },
+    {
+      "epoch": 12.747847478474785,
+      "grad_norm": 8.958296775817871,
+      "learning_rate": 8.06644880174292e-06,
+      "loss": 1.5575,
+      "step": 2600
+    },
+    {
+      "epoch": 12.870848708487085,
+      "grad_norm": 3.7900683879852295,
+      "learning_rate": 7.930283224400872e-06,
+      "loss": 1.5755,
+      "step": 2625
+    },
+    {
+      "epoch": 12.993849938499386,
+      "grad_norm": 4.654300689697266,
+      "learning_rate": 7.794117647058825e-06,
+      "loss": 1.6144,
+      "step": 2650
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.18672199170124482,
+      "eval_loss": 0.7420814633369446,
+      "eval_macro_f1": 0.5182252425366539,
+      "eval_micro_f1": 0.565118912797282,
+      "eval_runtime": 0.4279,
+      "eval_samples_per_second": 1689.604,
+      "eval_steps_per_second": 107.499,
+      "eval_weighted_f1": 0.5758384113601707,
+      "step": 2652
+    },
+    {
+      "epoch": 13.113161131611315,
+      "grad_norm": 8.709327697753906,
+      "learning_rate": 7.657952069716777e-06,
+      "loss": 1.4696,
+      "step": 2675
+    },
+    {
+      "epoch": 13.236162361623617,
+      "grad_norm": 4.688850402832031,
+      "learning_rate": 7.5217864923747285e-06,
+      "loss": 1.4891,
+      "step": 2700
+    },
+    {
+      "epoch": 13.359163591635916,
+      "grad_norm": 4.056413173675537,
+      "learning_rate": 7.385620915032681e-06,
+      "loss": 1.5349,
+      "step": 2725
+    },
+    {
+      "epoch": 13.482164821648217,
+      "grad_norm": 10.238755226135254,
+      "learning_rate": 7.249455337690633e-06,
+      "loss": 1.5372,
+      "step": 2750
+    },
+    {
+      "epoch": 13.605166051660516,
+      "grad_norm": 5.842587471008301,
+      "learning_rate": 7.113289760348584e-06,
+      "loss": 1.4805,
+      "step": 2775
+    },
+    {
+      "epoch": 13.728167281672818,
+      "grad_norm": 6.376543998718262,
+      "learning_rate": 6.977124183006536e-06,
+      "loss": 1.5407,
+      "step": 2800
+    },
+    {
+      "epoch": 13.851168511685117,
+      "grad_norm": 7.761598110198975,
+      "learning_rate": 6.840958605664488e-06,
+      "loss": 1.4828,
+      "step": 2825
+    },
+    {
+      "epoch": 13.974169741697416,
+      "grad_norm": 4.524773120880127,
+      "learning_rate": 6.70479302832244e-06,
+      "loss": 1.5614,
+      "step": 2850
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.19225449515905949,
+      "eval_loss": 0.7390348315238953,
+      "eval_macro_f1": 0.5248361915567423,
+      "eval_micro_f1": 0.5727298605180757,
+      "eval_runtime": 0.4194,
+      "eval_samples_per_second": 1723.824,
+      "eval_steps_per_second": 109.676,
+      "eval_weighted_f1": 0.5843580985419131,
+      "step": 2856
+    },
+    {
+      "epoch": 14.093480934809348,
+      "grad_norm": 5.846768856048584,
+      "learning_rate": 6.568627450980393e-06,
+      "loss": 1.4189,
+      "step": 2875
+    },
+    {
+      "epoch": 14.216482164821649,
+      "grad_norm": 6.395859718322754,
+      "learning_rate": 6.432461873638345e-06,
+      "loss": 1.4489,
+      "step": 2900
+    },
+    {
+      "epoch": 14.339483394833948,
+      "grad_norm": 7.327369213104248,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 1.4864,
+      "step": 2925
+    },
+    {
+      "epoch": 14.46248462484625,
+      "grad_norm": 4.31693696975708,
+      "learning_rate": 6.160130718954249e-06,
+      "loss": 1.4329,
+      "step": 2950
+    },
+    {
+      "epoch": 14.585485854858549,
+      "grad_norm": 7.08527946472168,
+      "learning_rate": 6.0239651416122014e-06,
+      "loss": 1.425,
+      "step": 2975
+    },
+    {
+      "epoch": 14.708487084870848,
+      "grad_norm": 4.74684476852417,
+      "learning_rate": 5.887799564270154e-06,
+      "loss": 1.4275,
+      "step": 3000
+    },
+    {
+      "epoch": 14.83148831488315,
+      "grad_norm": 13.72156047821045,
+      "learning_rate": 5.7516339869281045e-06,
+      "loss": 1.4445,
+      "step": 3025
+    },
+    {
+      "epoch": 14.954489544895448,
+      "grad_norm": 9.880606651306152,
+      "learning_rate": 5.615468409586057e-06,
+      "loss": 1.4172,
+      "step": 3050
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.18118948824343015,
+      "eval_loss": 0.7427704334259033,
+      "eval_macro_f1": 0.5143420941342366,
+      "eval_micro_f1": 0.5656338028169015,
+      "eval_runtime": 0.4638,
+      "eval_samples_per_second": 1558.864,
+      "eval_steps_per_second": 99.181,
+      "eval_weighted_f1": 0.5768156402908098,
+      "step": 3060
+    },
+    {
+      "epoch": 15.07380073800738,
+      "grad_norm": 16.806764602661133,
+      "learning_rate": 5.479302832244009e-06,
+      "loss": 1.3593,
+      "step": 3075
+    },
+    {
+      "epoch": 15.196801968019681,
+      "grad_norm": 6.311575889587402,
+      "learning_rate": 5.343137254901961e-06,
+      "loss": 1.4274,
+      "step": 3100
+    },
+    {
+      "epoch": 15.31980319803198,
+      "grad_norm": 13.33009147644043,
+      "learning_rate": 5.206971677559913e-06,
+      "loss": 1.3595,
+      "step": 3125
+    },
+    {
+      "epoch": 15.44280442804428,
+      "grad_norm": 4.078766822814941,
+      "learning_rate": 5.070806100217866e-06,
+      "loss": 1.4074,
+      "step": 3150
+    },
+    {
+      "epoch": 15.565805658056581,
+      "grad_norm": 10.644097328186035,
+      "learning_rate": 4.934640522875817e-06,
+      "loss": 1.4424,
+      "step": 3175
+    },
+    {
+      "epoch": 15.68880688806888,
+      "grad_norm": 5.109310626983643,
+      "learning_rate": 4.79847494553377e-06,
+      "loss": 1.3553,
+      "step": 3200
+    },
+    {
+      "epoch": 15.811808118081181,
+      "grad_norm": 7.613893985748291,
+      "learning_rate": 4.662309368191721e-06,
+      "loss": 1.3812,
+      "step": 3225
+    },
+    {
+      "epoch": 15.93480934809348,
+      "grad_norm": 6.8762946128845215,
+      "learning_rate": 4.5261437908496736e-06,
+      "loss": 1.4065,
+      "step": 3250
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.1825726141078838,
+      "eval_loss": 0.7540271282196045,
+      "eval_macro_f1": 0.5191546798139748,
+      "eval_micro_f1": 0.5683185333715268,
+      "eval_runtime": 0.4403,
+      "eval_samples_per_second": 1642.168,
+      "eval_steps_per_second": 104.481,
+      "eval_weighted_f1": 0.5804307137894656,
+      "step": 3264
+    },
+    {
+      "epoch": 16.05412054120541,
+      "grad_norm": 13.340099334716797,
+      "learning_rate": 4.389978213507626e-06,
+      "loss": 1.3355,
+      "step": 3275
+    },
+    {
+      "epoch": 16.177121771217713,
+      "grad_norm": 5.658074855804443,
+      "learning_rate": 4.253812636165578e-06,
+      "loss": 1.3459,
+      "step": 3300
+    },
+    {
+      "epoch": 16.300123001230013,
+      "grad_norm": 7.94587516784668,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 1.3617,
+      "step": 3325
+    },
+    {
+      "epoch": 16.423124231242312,
+      "grad_norm": 4.613985061645508,
+      "learning_rate": 3.9814814814814814e-06,
+      "loss": 1.3341,
+      "step": 3350
+    },
+    {
+      "epoch": 16.54612546125461,
+      "grad_norm": 7.527010917663574,
+      "learning_rate": 3.845315904139434e-06,
+      "loss": 1.3143,
+      "step": 3375
+    },
+    {
+      "epoch": 16.669126691266914,
+      "grad_norm": 9.891593933105469,
+      "learning_rate": 3.7091503267973862e-06,
+      "loss": 1.4211,
+      "step": 3400
+    },
+    {
+      "epoch": 16.792127921279214,
+      "grad_norm": 6.552762031555176,
+      "learning_rate": 3.572984749455338e-06,
+      "loss": 1.3691,
+      "step": 3425
+    },
+    {
+      "epoch": 16.915129151291513,
+      "grad_norm": 5.981565475463867,
+      "learning_rate": 3.4368191721132898e-06,
+      "loss": 1.2688,
+      "step": 3450
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.21023513139695713,
+      "eval_loss": 0.7745928764343262,
+      "eval_macro_f1": 0.5334202111120329,
+      "eval_micro_f1": 0.581764705882353,
+      "eval_runtime": 0.4488,
+      "eval_samples_per_second": 1610.932,
+      "eval_steps_per_second": 102.494,
+      "eval_weighted_f1": 0.5901403459769167,
+      "step": 3468
+    },
+    {
+      "epoch": 17.034440344403443,
+      "grad_norm": 6.895166397094727,
+      "learning_rate": 3.300653594771242e-06,
+      "loss": 1.2851,
+      "step": 3475
+    },
+    {
+      "epoch": 17.157441574415746,
+      "grad_norm": 4.59776496887207,
+      "learning_rate": 3.164488017429194e-06,
+      "loss": 1.3046,
+      "step": 3500
+    },
+    {
+      "epoch": 17.280442804428045,
+      "grad_norm": 4.178014278411865,
+      "learning_rate": 3.0283224400871465e-06,
+      "loss": 1.3453,
+      "step": 3525
+    },
+    {
+      "epoch": 17.403444034440344,
+      "grad_norm": 3.9644834995269775,
+      "learning_rate": 2.8921568627450985e-06,
+      "loss": 1.3331,
+      "step": 3550
+    },
+    {
+      "epoch": 17.526445264452644,
+      "grad_norm": 6.1651201248168945,
+      "learning_rate": 2.75599128540305e-06,
+      "loss": 1.3013,
+      "step": 3575
+    },
+    {
+      "epoch": 17.649446494464943,
+      "grad_norm": 6.653693199157715,
+      "learning_rate": 2.6198257080610024e-06,
+      "loss": 1.3299,
+      "step": 3600
+    },
+    {
+      "epoch": 17.772447724477246,
+      "grad_norm": 4.887754917144775,
+      "learning_rate": 2.4836601307189544e-06,
+      "loss": 1.2617,
+      "step": 3625
+    },
+    {
+      "epoch": 17.895448954489545,
+      "grad_norm": 9.397178649902344,
+      "learning_rate": 2.3474945533769064e-06,
+      "loss": 1.3397,
+      "step": 3650
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.19778699861687413,
+      "eval_loss": 0.7525197267532349,
+      "eval_macro_f1": 0.5230016464764999,
+      "eval_micro_f1": 0.574739281575898,
+      "eval_runtime": 0.4704,
+      "eval_samples_per_second": 1536.984,
+      "eval_steps_per_second": 97.789,
+      "eval_weighted_f1": 0.5859785736459122,
+      "step": 3672
+    },
+    {
+      "epoch": 18.014760147601475,
+      "grad_norm": 6.480631351470947,
+      "learning_rate": 2.2113289760348588e-06,
+      "loss": 1.2651,
+      "step": 3675
+    },
+    {
+      "epoch": 18.137761377613778,
+      "grad_norm": 4.3607707023620605,
+      "learning_rate": 2.0751633986928107e-06,
+      "loss": 1.2688,
+      "step": 3700
+    },
+    {
+      "epoch": 18.260762607626077,
+      "grad_norm": 6.317516803741455,
+      "learning_rate": 1.9389978213507627e-06,
+      "loss": 1.3253,
+      "step": 3725
+    },
+    {
+      "epoch": 18.383763837638377,
+      "grad_norm": 3.8014440536499023,
+      "learning_rate": 1.802832244008715e-06,
+      "loss": 1.3243,
+      "step": 3750
+    },
+    {
+      "epoch": 18.506765067650676,
+      "grad_norm": 22.600187301635742,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.2883,
+      "step": 3775
+    },
+    {
+      "epoch": 18.629766297662975,
+      "grad_norm": 9.801239013671875,
+      "learning_rate": 1.5305010893246188e-06,
+      "loss": 1.2581,
+      "step": 3800
+    },
+    {
+      "epoch": 18.752767527675278,
+      "grad_norm": 13.6274995803833,
+      "learning_rate": 1.394335511982571e-06,
+      "loss": 1.2963,
+      "step": 3825
+    },
+    {
+      "epoch": 18.875768757687577,
+      "grad_norm": 6.531538963317871,
+      "learning_rate": 1.258169934640523e-06,
+      "loss": 1.2437,
+      "step": 3850
+    },
+    {
+      "epoch": 18.998769987699877,
+      "grad_norm": 7.813934326171875,
+      "learning_rate": 1.122004357298475e-06,
+      "loss": 1.3025,
+      "step": 3875
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.20193637621023514,
+      "eval_loss": 0.766295313835144,
+      "eval_macro_f1": 0.5272206155524268,
+      "eval_micro_f1": 0.5776997366110623,
+      "eval_runtime": 0.4575,
+      "eval_samples_per_second": 1580.481,
+      "eval_steps_per_second": 100.556,
+      "eval_weighted_f1": 0.5870572148716042,
+      "step": 3876
+    },
+    {
+      "epoch": 19.118081180811807,
+      "grad_norm": 5.070478916168213,
+      "learning_rate": 9.858387799564272e-07,
+      "loss": 1.222,
+      "step": 3900
+    },
+    {
+      "epoch": 19.24108241082411,
+      "grad_norm": 5.113666534423828,
+      "learning_rate": 8.496732026143792e-07,
+      "loss": 1.2611,
+      "step": 3925
+    },
+    {
+      "epoch": 19.36408364083641,
+      "grad_norm": 4.808750152587891,
+      "learning_rate": 7.135076252723312e-07,
+      "loss": 1.3198,
+      "step": 3950
+    },
+    {
+      "epoch": 19.487084870848708,
+      "grad_norm": 5.08709716796875,
+      "learning_rate": 5.773420479302833e-07,
+      "loss": 1.2533,
+      "step": 3975
+    },
+    {
+      "epoch": 19.610086100861007,
+      "grad_norm": 5.573187828063965,
+      "learning_rate": 4.4117647058823536e-07,
+      "loss": 1.2522,
+      "step": 4000
+    },
+    {
+      "epoch": 19.73308733087331,
+      "grad_norm": 3.478645086288452,
+      "learning_rate": 3.050108932461874e-07,
+      "loss": 1.2694,
+      "step": 4025
+    },
+    {
+      "epoch": 19.85608856088561,
+      "grad_norm": 30.846281051635742,
+      "learning_rate": 1.6884531590413946e-07,
+      "loss": 1.2605,
+      "step": 4050
+    },
+    {
+      "epoch": 19.97908979089791,
+      "grad_norm": 5.579220294952393,
+      "learning_rate": 3.267973856209151e-08,
+      "loss": 1.2804,
+      "step": 4075
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.20885200553250347,
+      "eval_loss": 0.7704569697380066,
+      "eval_macro_f1": 0.533909530236441,
+      "eval_micro_f1": 0.5841438255231358,
+      "eval_runtime": 0.4522,
+      "eval_samples_per_second": 1598.807,
+      "eval_steps_per_second": 101.722,
+      "eval_weighted_f1": 0.5925984184147985,
+      "step": 4080
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 4080,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

ood-vigoemotions-phobert-source-source-01-s42-20260826T131538Z-0c086fe0/trainer/checkpoint-8626/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

ood-vigoemotions-phobert-source-source-01-s42-20260826T131538Z-0c086fe0/trainer/checkpoint-8626/special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-01-s42-20260826T131538Z-0c086fe0/trainer/checkpoint-8626/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 256,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-01-s42-20260826T131538Z-0c086fe0/trainer/checkpoint-8626/trainer_state.json

@@ -0,0 +1,2686 @@
+{
+  "best_global_step": 8626,
+  "best_metric": 0.5389595030296993,
+  "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vigoemotions-phobert-source-source-01-s42-20260826T131538Z-0c086fe0/trainer/checkpoint-8626",
+  "epoch": 19.0,
+  "eval_steps": 500,
+  "global_step": 8626,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.05509641873278237,
+      "grad_norm": 4.227837562561035,
+      "learning_rate": 5.286343612334802e-07,
+      "loss": 5.2039,
+      "step": 25
+    },
+    {
+      "epoch": 0.11019283746556474,
+      "grad_norm": 5.331326484680176,
+      "learning_rate": 1.0792951541850223e-06,
+      "loss": 5.0579,
+      "step": 50
+    },
+    {
+      "epoch": 0.1652892561983471,
+      "grad_norm": 5.1418375968933105,
+      "learning_rate": 1.629955947136564e-06,
+      "loss": 5.1842,
+      "step": 75
+    },
+    {
+      "epoch": 0.22038567493112948,
+      "grad_norm": 4.12420654296875,
+      "learning_rate": 2.180616740088106e-06,
+      "loss": 5.1411,
+      "step": 100
+    },
+    {
+      "epoch": 0.27548209366391185,
+      "grad_norm": 3.9862260818481445,
+      "learning_rate": 2.731277533039648e-06,
+      "loss": 5.2115,
+      "step": 125
+    },
+    {
+      "epoch": 0.3305785123966942,
+      "grad_norm": 4.397335052490234,
+      "learning_rate": 3.2819383259911898e-06,
+      "loss": 5.1033,
+      "step": 150
+    },
+    {
+      "epoch": 0.3856749311294766,
+      "grad_norm": 6.729996681213379,
+      "learning_rate": 3.8325991189427315e-06,
+      "loss": 5.1591,
+      "step": 175
+    },
+    {
+      "epoch": 0.44077134986225897,
+      "grad_norm": 5.972109317779541,
+      "learning_rate": 4.383259911894274e-06,
+      "loss": 5.1763,
+      "step": 200
+    },
+    {
+      "epoch": 0.49586776859504134,
+      "grad_norm": 5.35798454284668,
+      "learning_rate": 4.933920704845816e-06,
+      "loss": 5.0333,
+      "step": 225
+    },
+    {
+      "epoch": 0.5509641873278237,
+      "grad_norm": 6.948966979980469,
+      "learning_rate": 5.484581497797358e-06,
+      "loss": 4.9953,
+      "step": 250
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 4.730196475982666,
+      "learning_rate": 6.035242290748899e-06,
+      "loss": 4.948,
+      "step": 275
+    },
+    {
+      "epoch": 0.6611570247933884,
+      "grad_norm": 5.168266773223877,
+      "learning_rate": 6.585903083700441e-06,
+      "loss": 4.9737,
+      "step": 300
+    },
+    {
+      "epoch": 0.7162534435261708,
+      "grad_norm": 6.052734851837158,
+      "learning_rate": 7.136563876651983e-06,
+      "loss": 4.8535,
+      "step": 325
+    },
+    {
+      "epoch": 0.7713498622589532,
+      "grad_norm": 6.672458171844482,
+      "learning_rate": 7.687224669603525e-06,
+      "loss": 4.8036,
+      "step": 350
+    },
+    {
+      "epoch": 0.8264462809917356,
+      "grad_norm": 5.597661018371582,
+      "learning_rate": 8.237885462555067e-06,
+      "loss": 4.7429,
+      "step": 375
+    },
+    {
+      "epoch": 0.8815426997245179,
+      "grad_norm": 12.606609344482422,
+      "learning_rate": 8.788546255506607e-06,
+      "loss": 4.5717,
+      "step": 400
+    },
+    {
+      "epoch": 0.9366391184573003,
+      "grad_norm": 6.42344856262207,
+      "learning_rate": 9.339207048458151e-06,
+      "loss": 4.5116,
+      "step": 425
+    },
+    {
+      "epoch": 0.9917355371900827,
+      "grad_norm": 6.29040002822876,
+      "learning_rate": 9.889867841409693e-06,
+      "loss": 4.5005,
+      "step": 450
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 1.075113296508789,
+      "eval_macro_f1": 0.23296379461720287,
+      "eval_micro_f1": 0.2440189231808966,
+      "eval_runtime": 0.8467,
+      "eval_samples_per_second": 1940.521,
+      "eval_steps_per_second": 121.652,
+      "eval_weighted_f1": 0.30895690386581653,
+      "step": 454
+    },
+    {
+      "epoch": 1.0462809917355371,
+      "grad_norm": 9.197685241699219,
+      "learning_rate": 1.0440528634361234e-05,
+      "loss": 4.3713,
+      "step": 475
+    },
+    {
+      "epoch": 1.1013774104683196,
+      "grad_norm": 10.728469848632812,
+      "learning_rate": 1.0991189427312776e-05,
+      "loss": 4.3226,
+      "step": 500
+    },
+    {
+      "epoch": 1.1564738292011019,
+      "grad_norm": 10.500930786132812,
+      "learning_rate": 1.154185022026432e-05,
+      "loss": 4.2326,
+      "step": 525
+    },
+    {
+      "epoch": 1.2115702479338843,
+      "grad_norm": 10.107128143310547,
+      "learning_rate": 1.209251101321586e-05,
+      "loss": 4.1633,
+      "step": 550
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 6.363664150238037,
+      "learning_rate": 1.2643171806167402e-05,
+      "loss": 4.1342,
+      "step": 575
+    },
+    {
+      "epoch": 1.321763085399449,
+      "grad_norm": 6.806942939758301,
+      "learning_rate": 1.3193832599118943e-05,
+      "loss": 4.1034,
+      "step": 600
+    },
+    {
+      "epoch": 1.3768595041322313,
+      "grad_norm": 5.18177604675293,
+      "learning_rate": 1.3744493392070487e-05,
+      "loss": 4.1358,
+      "step": 625
+    },
+    {
+      "epoch": 1.4319559228650138,
+      "grad_norm": 13.11739444732666,
+      "learning_rate": 1.4295154185022027e-05,
+      "loss": 4.0104,
+      "step": 650
+    },
+    {
+      "epoch": 1.487052341597796,
+      "grad_norm": 13.619627952575684,
+      "learning_rate": 1.484581497797357e-05,
+      "loss": 3.9651,
+      "step": 675
+    },
+    {
+      "epoch": 1.5421487603305786,
+      "grad_norm": 9.929632186889648,
+      "learning_rate": 1.539647577092511e-05,
+      "loss": 3.946,
+      "step": 700
+    },
+    {
+      "epoch": 1.5972451790633608,
+      "grad_norm": 11.750797271728516,
+      "learning_rate": 1.5947136563876654e-05,
+      "loss": 3.8697,
+      "step": 725
+    },
+    {
+      "epoch": 1.652341597796143,
+      "grad_norm": 18.73433494567871,
+      "learning_rate": 1.6497797356828196e-05,
+      "loss": 3.8729,
+      "step": 750
+    },
+    {
+      "epoch": 1.7074380165289256,
+      "grad_norm": 9.994765281677246,
+      "learning_rate": 1.7048458149779738e-05,
+      "loss": 3.8399,
+      "step": 775
+    },
+    {
+      "epoch": 1.762534435261708,
+      "grad_norm": 13.98775577545166,
+      "learning_rate": 1.759911894273128e-05,
+      "loss": 3.7893,
+      "step": 800
+    },
+    {
+      "epoch": 1.8176308539944903,
+      "grad_norm": 11.74811840057373,
+      "learning_rate": 1.814977973568282e-05,
+      "loss": 3.8683,
+      "step": 825
+    },
+    {
+      "epoch": 1.8727272727272726,
+      "grad_norm": 10.832721710205078,
+      "learning_rate": 1.8700440528634364e-05,
+      "loss": 3.6864,
+      "step": 850
+    },
+    {
+      "epoch": 1.927823691460055,
+      "grad_norm": 9.301114082336426,
+      "learning_rate": 1.9251101321585906e-05,
+      "loss": 3.8125,
+      "step": 875
+    },
+    {
+      "epoch": 1.9829201101928375,
+      "grad_norm": 9.613875389099121,
+      "learning_rate": 1.9801762114537445e-05,
+      "loss": 3.6356,
+      "step": 900
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 0.8913275003433228,
+      "eval_macro_f1": 0.31342985112925853,
+      "eval_micro_f1": 0.3350537901871774,
+      "eval_runtime": 0.7417,
+      "eval_samples_per_second": 2215.172,
+      "eval_steps_per_second": 138.87,
+      "eval_weighted_f1": 0.390772984117027,
+      "step": 908
+    },
+    {
+      "epoch": 2.037465564738292,
+      "grad_norm": 8.475796699523926,
+      "learning_rate": 1.9960841899167892e-05,
+      "loss": 3.6071,
+      "step": 925
+    },
+    {
+      "epoch": 2.0925619834710742,
+      "grad_norm": 8.122872352600098,
+      "learning_rate": 1.989965736661772e-05,
+      "loss": 3.4062,
+      "step": 950
+    },
+    {
+      "epoch": 2.1476584022038567,
+      "grad_norm": 10.734721183776855,
+      "learning_rate": 1.983847283406755e-05,
+      "loss": 3.5119,
+      "step": 975
+    },
+    {
+      "epoch": 2.202754820936639,
+      "grad_norm": 7.998258113861084,
+      "learning_rate": 1.9777288301517378e-05,
+      "loss": 3.5601,
+      "step": 1000
+    },
+    {
+      "epoch": 2.2578512396694217,
+      "grad_norm": 11.745463371276855,
+      "learning_rate": 1.9716103768967207e-05,
+      "loss": 3.4671,
+      "step": 1025
+    },
+    {
+      "epoch": 2.3129476584022037,
+      "grad_norm": 10.620390892028809,
+      "learning_rate": 1.9654919236417035e-05,
+      "loss": 3.3976,
+      "step": 1050
+    },
+    {
+      "epoch": 2.368044077134986,
+      "grad_norm": 18.419321060180664,
+      "learning_rate": 1.9593734703866864e-05,
+      "loss": 3.5408,
+      "step": 1075
+    },
+    {
+      "epoch": 2.4231404958677687,
+      "grad_norm": 10.657835960388184,
+      "learning_rate": 1.9532550171316693e-05,
+      "loss": 3.3784,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4782369146005507,
+      "grad_norm": 11.976256370544434,
+      "learning_rate": 1.947136563876652e-05,
+      "loss": 3.4036,
+      "step": 1125
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 15.081591606140137,
+      "learning_rate": 1.941018110621635e-05,
+      "loss": 3.3793,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5884297520661157,
+      "grad_norm": 9.404304504394531,
+      "learning_rate": 1.934899657366618e-05,
+      "loss": 3.3192,
+      "step": 1175
+    },
+    {
+      "epoch": 2.643526170798898,
+      "grad_norm": 7.783041477203369,
+      "learning_rate": 1.9287812041116007e-05,
+      "loss": 3.2732,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6986225895316807,
+      "grad_norm": 10.965785026550293,
+      "learning_rate": 1.9226627508565836e-05,
+      "loss": 3.2884,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7537190082644627,
+      "grad_norm": 8.72093677520752,
+      "learning_rate": 1.9165442976015665e-05,
+      "loss": 3.2833,
+      "step": 1250
+    },
+    {
+      "epoch": 2.808815426997245,
+      "grad_norm": 7.197648048400879,
+      "learning_rate": 1.9104258443465493e-05,
+      "loss": 3.2771,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8639118457300277,
+      "grad_norm": 6.953661918640137,
+      "learning_rate": 1.9043073910915322e-05,
+      "loss": 3.1917,
+      "step": 1300
+    },
+    {
+      "epoch": 2.9190082644628097,
+      "grad_norm": 11.918855667114258,
+      "learning_rate": 1.898188937836515e-05,
+      "loss": 3.2673,
+      "step": 1325
+    },
+    {
+      "epoch": 2.974104683195592,
+      "grad_norm": 9.502423286437988,
+      "learning_rate": 1.892070484581498e-05,
+      "loss": 3.2804,
+      "step": 1350
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.0012172854534388314,
+      "eval_loss": 0.8076587319374084,
+      "eval_macro_f1": 0.3553418429761633,
+      "eval_micro_f1": 0.38251753205355243,
+      "eval_runtime": 0.7432,
+      "eval_samples_per_second": 2210.799,
+      "eval_steps_per_second": 138.595,
+      "eval_weighted_f1": 0.4286911321436805,
+      "step": 1362
+    },
+    {
+      "epoch": 3.028650137741047,
+      "grad_norm": 14.517410278320312,
+      "learning_rate": 1.8859520313264808e-05,
+      "loss": 3.0605,
+      "step": 1375
+    },
+    {
+      "epoch": 3.0837465564738293,
+      "grad_norm": 6.636230945587158,
+      "learning_rate": 1.8798335780714636e-05,
+      "loss": 2.9334,
+      "step": 1400
+    },
+    {
+      "epoch": 3.1388429752066114,
+      "grad_norm": 7.712543964385986,
+      "learning_rate": 1.8737151248164465e-05,
+      "loss": 3.0362,
+      "step": 1425
+    },
+    {
+      "epoch": 3.193939393939394,
+      "grad_norm": 6.594825267791748,
+      "learning_rate": 1.8675966715614294e-05,
+      "loss": 2.9552,
+      "step": 1450
+    },
+    {
+      "epoch": 3.2490358126721763,
+      "grad_norm": 10.446346282958984,
+      "learning_rate": 1.8614782183064122e-05,
+      "loss": 3.0088,
+      "step": 1475
+    },
+    {
+      "epoch": 3.304132231404959,
+      "grad_norm": 11.816834449768066,
+      "learning_rate": 1.855359765051395e-05,
+      "loss": 2.9343,
+      "step": 1500
+    },
+    {
+      "epoch": 3.359228650137741,
+      "grad_norm": 11.104568481445312,
+      "learning_rate": 1.849241311796378e-05,
+      "loss": 3.0041,
+      "step": 1525
+    },
+    {
+      "epoch": 3.4143250688705233,
+      "grad_norm": 18.386274337768555,
+      "learning_rate": 1.843122858541361e-05,
+      "loss": 2.9729,
+      "step": 1550
+    },
+    {
+      "epoch": 3.469421487603306,
+      "grad_norm": 8.527523040771484,
+      "learning_rate": 1.8370044052863437e-05,
+      "loss": 2.8784,
+      "step": 1575
+    },
+    {
+      "epoch": 3.5245179063360883,
+      "grad_norm": 6.6398024559021,
+      "learning_rate": 1.8308859520313266e-05,
+      "loss": 2.9162,
+      "step": 1600
+    },
+    {
+      "epoch": 3.579614325068871,
+      "grad_norm": 8.599790573120117,
+      "learning_rate": 1.8247674987763094e-05,
+      "loss": 3.002,
+      "step": 1625
+    },
+    {
+      "epoch": 3.634710743801653,
+      "grad_norm": 13.401683807373047,
+      "learning_rate": 1.8186490455212923e-05,
+      "loss": 2.8535,
+      "step": 1650
+    },
+    {
+      "epoch": 3.6898071625344353,
+      "grad_norm": 10.858172416687012,
+      "learning_rate": 1.812530592266275e-05,
+      "loss": 2.9071,
+      "step": 1675
+    },
+    {
+      "epoch": 3.744903581267218,
+      "grad_norm": 14.278366088867188,
+      "learning_rate": 1.806412139011258e-05,
+      "loss": 2.8518,
+      "step": 1700
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 10.848072052001953,
+      "learning_rate": 1.800293685756241e-05,
+      "loss": 2.8515,
+      "step": 1725
+    },
+    {
+      "epoch": 3.8550964187327823,
+      "grad_norm": 10.81524658203125,
+      "learning_rate": 1.7941752325012238e-05,
+      "loss": 2.9081,
+      "step": 1750
+    },
+    {
+      "epoch": 3.910192837465565,
+      "grad_norm": 8.548844337463379,
+      "learning_rate": 1.7880567792462066e-05,
+      "loss": 2.8372,
+      "step": 1775
+    },
+    {
+      "epoch": 3.9652892561983473,
+      "grad_norm": 13.30102252960205,
+      "learning_rate": 1.7819383259911895e-05,
+      "loss": 2.7932,
+      "step": 1800
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.002434570906877663,
+      "eval_loss": 0.756788969039917,
+      "eval_macro_f1": 0.39802980850917896,
+      "eval_micro_f1": 0.42926200451176283,
+      "eval_runtime": 0.8119,
+      "eval_samples_per_second": 2023.73,
+      "eval_steps_per_second": 126.868,
+      "eval_weighted_f1": 0.46001421514565244,
+      "step": 1816
+    },
+    {
+      "epoch": 4.019834710743802,
+      "grad_norm": 16.3684024810791,
+      "learning_rate": 1.7758198727361724e-05,
+      "loss": 2.7434,
+      "step": 1825
+    },
+    {
+      "epoch": 4.074931129476584,
+      "grad_norm": 9.8795804977417,
+      "learning_rate": 1.7697014194811552e-05,
+      "loss": 2.6399,
+      "step": 1850
+    },
+    {
+      "epoch": 4.130027548209366,
+      "grad_norm": 9.345877647399902,
+      "learning_rate": 1.763582966226138e-05,
+      "loss": 2.678,
+      "step": 1875
+    },
+    {
+      "epoch": 4.1851239669421485,
+      "grad_norm": 9.299138069152832,
+      "learning_rate": 1.757464512971121e-05,
+      "loss": 2.6086,
+      "step": 1900
+    },
+    {
+      "epoch": 4.240220385674931,
+      "grad_norm": 5.941190719604492,
+      "learning_rate": 1.7513460597161038e-05,
+      "loss": 2.6337,
+      "step": 1925
+    },
+    {
+      "epoch": 4.2953168044077135,
+      "grad_norm": 8.25993537902832,
+      "learning_rate": 1.7452276064610867e-05,
+      "loss": 2.6188,
+      "step": 1950
+    },
+    {
+      "epoch": 4.350413223140496,
+      "grad_norm": 12.652754783630371,
+      "learning_rate": 1.7391091532060695e-05,
+      "loss": 2.5735,
+      "step": 1975
+    },
+    {
+      "epoch": 4.405509641873278,
+      "grad_norm": 7.743082523345947,
+      "learning_rate": 1.7329906999510527e-05,
+      "loss": 2.6304,
+      "step": 2000
+    },
+    {
+      "epoch": 4.460606060606061,
+      "grad_norm": 17.932268142700195,
+      "learning_rate": 1.7268722466960356e-05,
+      "loss": 2.5889,
+      "step": 2025
+    },
+    {
+      "epoch": 4.515702479338843,
+      "grad_norm": 5.641079425811768,
+      "learning_rate": 1.7207537934410185e-05,
+      "loss": 2.6445,
+      "step": 2050
+    },
+    {
+      "epoch": 4.570798898071625,
+      "grad_norm": 11.554326057434082,
+      "learning_rate": 1.7146353401860013e-05,
+      "loss": 2.5493,
+      "step": 2075
+    },
+    {
+      "epoch": 4.6258953168044075,
+      "grad_norm": 11.849327087402344,
+      "learning_rate": 1.7085168869309842e-05,
+      "loss": 2.6203,
+      "step": 2100
+    },
+    {
+      "epoch": 4.68099173553719,
+      "grad_norm": 8.938584327697754,
+      "learning_rate": 1.702398433675967e-05,
+      "loss": 2.5594,
+      "step": 2125
+    },
+    {
+      "epoch": 4.736088154269972,
+      "grad_norm": 8.740391731262207,
+      "learning_rate": 1.69627998042095e-05,
+      "loss": 2.5733,
+      "step": 2150
+    },
+    {
+      "epoch": 4.791184573002755,
+      "grad_norm": 9.344244003295898,
+      "learning_rate": 1.6901615271659325e-05,
+      "loss": 2.5557,
+      "step": 2175
+    },
+    {
+      "epoch": 4.846280991735537,
+      "grad_norm": 16.789382934570312,
+      "learning_rate": 1.6840430739109153e-05,
+      "loss": 2.5097,
+      "step": 2200
+    },
+    {
+      "epoch": 4.90137741046832,
+      "grad_norm": 10.325007438659668,
+      "learning_rate": 1.6779246206558982e-05,
+      "loss": 2.4844,
+      "step": 2225
+    },
+    {
+      "epoch": 4.9564738292011015,
+      "grad_norm": 12.274792671203613,
+      "learning_rate": 1.671806167400881e-05,
+      "loss": 2.4952,
+      "step": 2250
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.02556299452221546,
+      "eval_loss": 0.7364040613174438,
+      "eval_macro_f1": 0.4238555784117118,
+      "eval_micro_f1": 0.4553359683794466,
+      "eval_runtime": 0.8163,
+      "eval_samples_per_second": 2012.859,
+      "eval_steps_per_second": 126.187,
+      "eval_weighted_f1": 0.4764858778657165,
+      "step": 2270
+    },
+    {
+      "epoch": 5.011019283746556,
+      "grad_norm": 9.081082344055176,
+      "learning_rate": 1.665687714145864e-05,
+      "loss": 2.4144,
+      "step": 2275
+    },
+    {
+      "epoch": 5.066115702479339,
+      "grad_norm": 7.6597723960876465,
+      "learning_rate": 1.6595692608908468e-05,
+      "loss": 2.2944,
+      "step": 2300
+    },
+    {
+      "epoch": 5.121212121212121,
+      "grad_norm": 60.42306137084961,
+      "learning_rate": 1.6534508076358297e-05,
+      "loss": 2.2364,
+      "step": 2325
+    },
+    {
+      "epoch": 5.176308539944904,
+      "grad_norm": 10.49393081665039,
+      "learning_rate": 1.6473323543808125e-05,
+      "loss": 2.2605,
+      "step": 2350
+    },
+    {
+      "epoch": 5.231404958677686,
+      "grad_norm": 6.1078782081604,
+      "learning_rate": 1.6412139011257954e-05,
+      "loss": 2.2864,
+      "step": 2375
+    },
+    {
+      "epoch": 5.2865013774104685,
+      "grad_norm": 9.789051055908203,
+      "learning_rate": 1.6350954478707782e-05,
+      "loss": 2.3701,
+      "step": 2400
+    },
+    {
+      "epoch": 5.341597796143251,
+      "grad_norm": 12.479533195495605,
+      "learning_rate": 1.628976994615761e-05,
+      "loss": 2.3592,
+      "step": 2425
+    },
+    {
+      "epoch": 5.3966942148760335,
+      "grad_norm": 7.823705673217773,
+      "learning_rate": 1.622858541360744e-05,
+      "loss": 2.3578,
+      "step": 2450
+    },
+    {
+      "epoch": 5.451790633608815,
+      "grad_norm": 6.067243576049805,
+      "learning_rate": 1.616740088105727e-05,
+      "loss": 2.2772,
+      "step": 2475
+    },
+    {
+      "epoch": 5.506887052341598,
+      "grad_norm": 11.943389892578125,
+      "learning_rate": 1.6106216348507097e-05,
+      "loss": 2.2705,
+      "step": 2500
+    },
+    {
+      "epoch": 5.56198347107438,
+      "grad_norm": 10.336061477661133,
+      "learning_rate": 1.604503181595693e-05,
+      "loss": 2.2845,
+      "step": 2525
+    },
+    {
+      "epoch": 5.6170798898071626,
+      "grad_norm": 12.933616638183594,
+      "learning_rate": 1.5983847283406758e-05,
+      "loss": 2.2261,
+      "step": 2550
+    },
+    {
+      "epoch": 5.672176308539945,
+      "grad_norm": 10.702818870544434,
+      "learning_rate": 1.5922662750856586e-05,
+      "loss": 2.3344,
+      "step": 2575
+    },
+    {
+      "epoch": 5.7272727272727275,
+      "grad_norm": 10.82820987701416,
+      "learning_rate": 1.5861478218306415e-05,
+      "loss": 2.3006,
+      "step": 2600
+    },
+    {
+      "epoch": 5.78236914600551,
+      "grad_norm": 12.483839988708496,
+      "learning_rate": 1.5800293685756244e-05,
+      "loss": 2.25,
+      "step": 2625
+    },
+    {
+      "epoch": 5.837465564738292,
+      "grad_norm": 12.435521125793457,
+      "learning_rate": 1.5739109153206072e-05,
+      "loss": 2.2539,
+      "step": 2650
+    },
+    {
+      "epoch": 5.892561983471074,
+      "grad_norm": 10.098462104797363,
+      "learning_rate": 1.56779246206559e-05,
+      "loss": 2.1928,
+      "step": 2675
+    },
+    {
+      "epoch": 5.9476584022038566,
+      "grad_norm": 14.790334701538086,
+      "learning_rate": 1.561674008810573e-05,
+      "loss": 2.1906,
+      "step": 2700
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.02982349360925137,
+      "eval_loss": 0.7161994576454163,
+      "eval_macro_f1": 0.4152780695248448,
+      "eval_micro_f1": 0.44851338390443807,
+      "eval_runtime": 0.8477,
+      "eval_samples_per_second": 1938.149,
+      "eval_steps_per_second": 121.503,
+      "eval_weighted_f1": 0.473624177164651,
+      "step": 2724
+    },
+    {
+      "epoch": 6.002203856749311,
+      "grad_norm": 11.331801414489746,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 2.1592,
+      "step": 2725
+    },
+    {
+      "epoch": 6.057300275482094,
+      "grad_norm": 7.465226650238037,
+      "learning_rate": 1.5494371023005387e-05,
+      "loss": 2.1362,
+      "step": 2750
+    },
+    {
+      "epoch": 6.112396694214876,
+      "grad_norm": 7.91953706741333,
+      "learning_rate": 1.5433186490455216e-05,
+      "loss": 2.133,
+      "step": 2775
+    },
+    {
+      "epoch": 6.167493112947659,
+      "grad_norm": 43.329559326171875,
+      "learning_rate": 1.5372001957905044e-05,
+      "loss": 2.0267,
+      "step": 2800
+    },
+    {
+      "epoch": 6.222589531680441,
+      "grad_norm": 9.198493003845215,
+      "learning_rate": 1.5310817425354873e-05,
+      "loss": 1.9903,
+      "step": 2825
+    },
+    {
+      "epoch": 6.277685950413223,
+      "grad_norm": 14.81072998046875,
+      "learning_rate": 1.52496328928047e-05,
+      "loss": 2.0183,
+      "step": 2850
+    },
+    {
+      "epoch": 6.332782369146005,
+      "grad_norm": 7.481613636016846,
+      "learning_rate": 1.5188448360254529e-05,
+      "loss": 2.0888,
+      "step": 2875
+    },
+    {
+      "epoch": 6.387878787878788,
+      "grad_norm": 28.831653594970703,
+      "learning_rate": 1.5127263827704357e-05,
+      "loss": 2.0233,
+      "step": 2900
+    },
+    {
+      "epoch": 6.44297520661157,
+      "grad_norm": 10.08276653289795,
+      "learning_rate": 1.5066079295154186e-05,
+      "loss": 2.0979,
+      "step": 2925
+    },
+    {
+      "epoch": 6.498071625344353,
+      "grad_norm": 8.211263656616211,
+      "learning_rate": 1.5004894762604014e-05,
+      "loss": 2.0263,
+      "step": 2950
+    },
+    {
+      "epoch": 6.553168044077135,
+      "grad_norm": 13.02656364440918,
+      "learning_rate": 1.4943710230053843e-05,
+      "loss": 1.9823,
+      "step": 2975
+    },
+    {
+      "epoch": 6.608264462809918,
+      "grad_norm": 10.631696701049805,
+      "learning_rate": 1.4882525697503672e-05,
+      "loss": 1.987,
+      "step": 3000
+    },
+    {
+      "epoch": 6.663360881542699,
+      "grad_norm": 8.777915954589844,
+      "learning_rate": 1.48213411649535e-05,
+      "loss": 2.0683,
+      "step": 3025
+    },
+    {
+      "epoch": 6.718457300275482,
+      "grad_norm": 12.436280250549316,
+      "learning_rate": 1.476015663240333e-05,
+      "loss": 2.0,
+      "step": 3050
+    },
+    {
+      "epoch": 6.773553719008264,
+      "grad_norm": 11.935210227966309,
+      "learning_rate": 1.469897209985316e-05,
+      "loss": 1.9862,
+      "step": 3075
+    },
+    {
+      "epoch": 6.828650137741047,
+      "grad_norm": 11.0009765625,
+      "learning_rate": 1.4637787567302988e-05,
+      "loss": 2.0002,
+      "step": 3100
+    },
+    {
+      "epoch": 6.883746556473829,
+      "grad_norm": 8.148244857788086,
+      "learning_rate": 1.4576603034752817e-05,
+      "loss": 1.9712,
+      "step": 3125
+    },
+    {
+      "epoch": 6.938842975206612,
+      "grad_norm": 15.212584495544434,
+      "learning_rate": 1.4515418502202645e-05,
+      "loss": 1.9824,
+      "step": 3150
+    },
+    {
+      "epoch": 6.993939393939394,
+      "grad_norm": 6.5839457511901855,
+      "learning_rate": 1.4454233969652474e-05,
+      "loss": 1.9517,
+      "step": 3175
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.0426049908703591,
+      "eval_loss": 0.7072303891181946,
+      "eval_macro_f1": 0.4624544764817252,
+      "eval_micro_f1": 0.4939302709527047,
+      "eval_runtime": 0.7605,
+      "eval_samples_per_second": 2160.426,
+      "eval_steps_per_second": 135.438,
+      "eval_weighted_f1": 0.5083354781806794,
+      "step": 3178
+    },
+    {
+      "epoch": 7.048484848484849,
+      "grad_norm": 9.203914642333984,
+      "learning_rate": 1.4393049437102303e-05,
+      "loss": 1.8504,
+      "step": 3200
+    },
+    {
+      "epoch": 7.103581267217631,
+      "grad_norm": 10.75075912475586,
+      "learning_rate": 1.4331864904552131e-05,
+      "loss": 1.8917,
+      "step": 3225
+    },
+    {
+      "epoch": 7.158677685950413,
+      "grad_norm": 8.260217666625977,
+      "learning_rate": 1.427068037200196e-05,
+      "loss": 1.9492,
+      "step": 3250
+    },
+    {
+      "epoch": 7.213774104683195,
+      "grad_norm": 12.863761901855469,
+      "learning_rate": 1.4209495839451789e-05,
+      "loss": 1.8323,
+      "step": 3275
+    },
+    {
+      "epoch": 7.268870523415978,
+      "grad_norm": 7.896994113922119,
+      "learning_rate": 1.4148311306901617e-05,
+      "loss": 1.8989,
+      "step": 3300
+    },
+    {
+      "epoch": 7.32396694214876,
+      "grad_norm": 13.598098754882812,
+      "learning_rate": 1.4087126774351444e-05,
+      "loss": 1.8014,
+      "step": 3325
+    },
+    {
+      "epoch": 7.379063360881543,
+      "grad_norm": 10.687313079833984,
+      "learning_rate": 1.4025942241801273e-05,
+      "loss": 1.7836,
+      "step": 3350
+    },
+    {
+      "epoch": 7.434159779614325,
+      "grad_norm": 9.242302894592285,
+      "learning_rate": 1.3964757709251102e-05,
+      "loss": 1.8942,
+      "step": 3375
+    },
+    {
+      "epoch": 7.489256198347108,
+      "grad_norm": 8.878076553344727,
+      "learning_rate": 1.390357317670093e-05,
+      "loss": 1.7399,
+      "step": 3400
+    },
+    {
+      "epoch": 7.544352617079889,
+      "grad_norm": 14.274787902832031,
+      "learning_rate": 1.3842388644150759e-05,
+      "loss": 1.7946,
+      "step": 3425
+    },
+    {
+      "epoch": 7.599449035812672,
+      "grad_norm": 7.229008197784424,
+      "learning_rate": 1.3781204111600588e-05,
+      "loss": 1.8592,
+      "step": 3450
+    },
+    {
+      "epoch": 7.654545454545454,
+      "grad_norm": 7.422128200531006,
+      "learning_rate": 1.3720019579050416e-05,
+      "loss": 1.8041,
+      "step": 3475
+    },
+    {
+      "epoch": 7.709641873278237,
+      "grad_norm": 5.205471038818359,
+      "learning_rate": 1.3658835046500245e-05,
+      "loss": 1.8001,
+      "step": 3500
+    },
+    {
+      "epoch": 7.764738292011019,
+      "grad_norm": 14.030226707458496,
+      "learning_rate": 1.3597650513950073e-05,
+      "loss": 1.7651,
+      "step": 3525
+    },
+    {
+      "epoch": 7.819834710743802,
+      "grad_norm": 9.975546836853027,
+      "learning_rate": 1.3536465981399902e-05,
+      "loss": 1.8567,
+      "step": 3550
+    },
+    {
+      "epoch": 7.874931129476584,
+      "grad_norm": 7.343936920166016,
+      "learning_rate": 1.3475281448849732e-05,
+      "loss": 1.7966,
+      "step": 3575
+    },
+    {
+      "epoch": 7.930027548209367,
+      "grad_norm": 9.779473304748535,
+      "learning_rate": 1.3414096916299561e-05,
+      "loss": 1.7319,
+      "step": 3600
+    },
+    {
+      "epoch": 7.985123966942149,
+      "grad_norm": 9.262539863586426,
+      "learning_rate": 1.335291238374939e-05,
+      "loss": 1.7536,
+      "step": 3625
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.06086427267194157,
+      "eval_loss": 0.7092871069908142,
+      "eval_macro_f1": 0.4828944671231688,
+      "eval_micro_f1": 0.5134519867549668,
+      "eval_runtime": 0.8505,
+      "eval_samples_per_second": 1931.851,
+      "eval_steps_per_second": 121.108,
+      "eval_weighted_f1": 0.5272358823424601,
+      "step": 3632
+    },
+    {
+      "epoch": 8.039669421487604,
+      "grad_norm": 9.195419311523438,
+      "learning_rate": 1.3291727851199218e-05,
+      "loss": 1.7551,
+      "step": 3650
+    },
+    {
+      "epoch": 8.094765840220386,
+      "grad_norm": 8.99988842010498,
+      "learning_rate": 1.3230543318649047e-05,
+      "loss": 1.6994,
+      "step": 3675
+    },
+    {
+      "epoch": 8.149862258953169,
+      "grad_norm": 13.685859680175781,
+      "learning_rate": 1.3169358786098876e-05,
+      "loss": 1.7063,
+      "step": 3700
+    },
+    {
+      "epoch": 8.20495867768595,
+      "grad_norm": 7.462896347045898,
+      "learning_rate": 1.3108174253548704e-05,
+      "loss": 1.6769,
+      "step": 3725
+    },
+    {
+      "epoch": 8.260055096418732,
+      "grad_norm": 9.458789825439453,
+      "learning_rate": 1.3046989720998533e-05,
+      "loss": 1.6593,
+      "step": 3750
+    },
+    {
+      "epoch": 8.315151515151515,
+      "grad_norm": 9.853849411010742,
+      "learning_rate": 1.2985805188448362e-05,
+      "loss": 1.6467,
+      "step": 3775
+    },
+    {
+      "epoch": 8.370247933884297,
+      "grad_norm": 7.881902694702148,
+      "learning_rate": 1.292462065589819e-05,
+      "loss": 1.6405,
+      "step": 3800
+    },
+    {
+      "epoch": 8.42534435261708,
+      "grad_norm": 13.257023811340332,
+      "learning_rate": 1.2863436123348019e-05,
+      "loss": 1.6338,
+      "step": 3825
+    },
+    {
+      "epoch": 8.480440771349862,
+      "grad_norm": 8.842818260192871,
+      "learning_rate": 1.2802251590797848e-05,
+      "loss": 1.6426,
+      "step": 3850
+    },
+    {
+      "epoch": 8.535537190082644,
+      "grad_norm": 9.2915620803833,
+      "learning_rate": 1.2741067058247676e-05,
+      "loss": 1.6603,
+      "step": 3875
+    },
+    {
+      "epoch": 8.590633608815427,
+      "grad_norm": 6.625736713409424,
+      "learning_rate": 1.2679882525697505e-05,
+      "loss": 1.6491,
+      "step": 3900
+    },
+    {
+      "epoch": 8.64573002754821,
+      "grad_norm": 11.400477409362793,
+      "learning_rate": 1.2618697993147334e-05,
+      "loss": 1.6193,
+      "step": 3925
+    },
+    {
+      "epoch": 8.700826446280992,
+      "grad_norm": 8.049001693725586,
+      "learning_rate": 1.2557513460597162e-05,
+      "loss": 1.6068,
+      "step": 3950
+    },
+    {
+      "epoch": 8.755922865013774,
+      "grad_norm": 8.45368766784668,
+      "learning_rate": 1.2496328928046991e-05,
+      "loss": 1.629,
+      "step": 3975
+    },
+    {
+      "epoch": 8.811019283746557,
+      "grad_norm": 11.155718803405762,
+      "learning_rate": 1.243514439549682e-05,
+      "loss": 1.6357,
+      "step": 4000
+    },
+    {
+      "epoch": 8.86611570247934,
+      "grad_norm": 9.854330062866211,
+      "learning_rate": 1.2373959862946646e-05,
+      "loss": 1.6631,
+      "step": 4025
+    },
+    {
+      "epoch": 8.921212121212122,
+      "grad_norm": 11.173809051513672,
+      "learning_rate": 1.2312775330396475e-05,
+      "loss": 1.6442,
+      "step": 4050
+    },
+    {
+      "epoch": 8.976308539944904,
+      "grad_norm": 6.974915027618408,
+      "learning_rate": 1.2251590797846304e-05,
+      "loss": 1.6068,
+      "step": 4075
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.05660377358490566,
+      "eval_loss": 0.7177210450172424,
+      "eval_macro_f1": 0.4876323052992565,
+      "eval_micro_f1": 0.5183353934899052,
+      "eval_runtime": 0.7616,
+      "eval_samples_per_second": 2157.235,
+      "eval_steps_per_second": 135.237,
+      "eval_weighted_f1": 0.5293421961526377,
+      "step": 4086
+    },
+    {
+      "epoch": 9.030853994490359,
+      "grad_norm": 8.590648651123047,
+      "learning_rate": 1.2190406265296136e-05,
+      "loss": 1.604,
+      "step": 4100
+    },
+    {
+      "epoch": 9.08595041322314,
+      "grad_norm": 11.171589851379395,
+      "learning_rate": 1.2129221732745964e-05,
+      "loss": 1.5905,
+      "step": 4125
+    },
+    {
+      "epoch": 9.141046831955922,
+      "grad_norm": 17.820585250854492,
+      "learning_rate": 1.2068037200195793e-05,
+      "loss": 1.4744,
+      "step": 4150
+    },
+    {
+      "epoch": 9.196143250688705,
+      "grad_norm": 8.431110382080078,
+      "learning_rate": 1.2006852667645622e-05,
+      "loss": 1.5549,
+      "step": 4175
+    },
+    {
+      "epoch": 9.251239669421487,
+      "grad_norm": 7.0245680809021,
+      "learning_rate": 1.1945668135095449e-05,
+      "loss": 1.5031,
+      "step": 4200
+    },
+    {
+      "epoch": 9.30633608815427,
+      "grad_norm": 11.74380111694336,
+      "learning_rate": 1.1884483602545277e-05,
+      "loss": 1.456,
+      "step": 4225
+    },
+    {
+      "epoch": 9.361432506887052,
+      "grad_norm": 12.57530403137207,
+      "learning_rate": 1.1823299069995106e-05,
+      "loss": 1.5037,
+      "step": 4250
+    },
+    {
+      "epoch": 9.416528925619835,
+      "grad_norm": 7.103407859802246,
+      "learning_rate": 1.1762114537444935e-05,
+      "loss": 1.5581,
+      "step": 4275
+    },
+    {
+      "epoch": 9.471625344352617,
+      "grad_norm": 14.686859130859375,
+      "learning_rate": 1.1700930004894763e-05,
+      "loss": 1.5657,
+      "step": 4300
+    },
+    {
+      "epoch": 9.5267217630854,
+      "grad_norm": 9.481619834899902,
+      "learning_rate": 1.1639745472344592e-05,
+      "loss": 1.5319,
+      "step": 4325
+    },
+    {
+      "epoch": 9.581818181818182,
+      "grad_norm": 9.552848815917969,
+      "learning_rate": 1.157856093979442e-05,
+      "loss": 1.5062,
+      "step": 4350
+    },
+    {
+      "epoch": 9.636914600550964,
+      "grad_norm": 10.749216079711914,
+      "learning_rate": 1.151737640724425e-05,
+      "loss": 1.4887,
+      "step": 4375
+    },
+    {
+      "epoch": 9.692011019283747,
+      "grad_norm": 12.477849960327148,
+      "learning_rate": 1.1456191874694078e-05,
+      "loss": 1.5119,
+      "step": 4400
+    },
+    {
+      "epoch": 9.74710743801653,
+      "grad_norm": 9.663217544555664,
+      "learning_rate": 1.1395007342143907e-05,
+      "loss": 1.4636,
+      "step": 4425
+    },
+    {
+      "epoch": 9.802203856749312,
+      "grad_norm": 5.807657241821289,
+      "learning_rate": 1.1333822809593735e-05,
+      "loss": 1.537,
+      "step": 4450
+    },
+    {
+      "epoch": 9.857300275482094,
+      "grad_norm": 14.375182151794434,
+      "learning_rate": 1.1272638277043564e-05,
+      "loss": 1.4866,
+      "step": 4475
+    },
+    {
+      "epoch": 9.912396694214877,
+      "grad_norm": 6.251608848571777,
+      "learning_rate": 1.1211453744493393e-05,
+      "loss": 1.5255,
+      "step": 4500
+    },
+    {
+      "epoch": 9.967493112947658,
+      "grad_norm": 15.098796844482422,
+      "learning_rate": 1.1150269211943221e-05,
+      "loss": 1.434,
+      "step": 4525
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.0620815581253804,
+      "eval_loss": 0.7124394774436951,
+      "eval_macro_f1": 0.49507577431721617,
+      "eval_micro_f1": 0.5261606484893147,
+      "eval_runtime": 0.8034,
+      "eval_samples_per_second": 2044.959,
+      "eval_steps_per_second": 128.199,
+      "eval_weighted_f1": 0.5337704983302356,
+      "step": 4540
+    },
+    {
+      "epoch": 10.022038567493112,
+      "grad_norm": 7.172276496887207,
+      "learning_rate": 1.108908467939305e-05,
+      "loss": 1.4615,
+      "step": 4550
+    },
+    {
+      "epoch": 10.077134986225895,
+      "grad_norm": 6.198989391326904,
+      "learning_rate": 1.1027900146842878e-05,
+      "loss": 1.4007,
+      "step": 4575
+    },
+    {
+      "epoch": 10.132231404958677,
+      "grad_norm": 6.623414993286133,
+      "learning_rate": 1.0966715614292707e-05,
+      "loss": 1.4138,
+      "step": 4600
+    },
+    {
+      "epoch": 10.18732782369146,
+      "grad_norm": 8.518909454345703,
+      "learning_rate": 1.0905531081742536e-05,
+      "loss": 1.37,
+      "step": 4625
+    },
+    {
+      "epoch": 10.242424242424242,
+      "grad_norm": 9.576210021972656,
+      "learning_rate": 1.0844346549192366e-05,
+      "loss": 1.3478,
+      "step": 4650
+    },
+    {
+      "epoch": 10.297520661157025,
+      "grad_norm": 20.642356872558594,
+      "learning_rate": 1.0783162016642195e-05,
+      "loss": 1.4265,
+      "step": 4675
+    },
+    {
+      "epoch": 10.352617079889807,
+      "grad_norm": 11.112329483032227,
+      "learning_rate": 1.0721977484092023e-05,
+      "loss": 1.4461,
+      "step": 4700
+    },
+    {
+      "epoch": 10.40771349862259,
+      "grad_norm": 10.513445854187012,
+      "learning_rate": 1.0660792951541852e-05,
+      "loss": 1.3629,
+      "step": 4725
+    },
+    {
+      "epoch": 10.462809917355372,
+      "grad_norm": 8.798458099365234,
+      "learning_rate": 1.059960841899168e-05,
+      "loss": 1.3599,
+      "step": 4750
+    },
+    {
+      "epoch": 10.517906336088155,
+      "grad_norm": 9.605145454406738,
+      "learning_rate": 1.053842388644151e-05,
+      "loss": 1.4262,
+      "step": 4775
+    },
+    {
+      "epoch": 10.573002754820937,
+      "grad_norm": 12.231898307800293,
+      "learning_rate": 1.0477239353891338e-05,
+      "loss": 1.3747,
+      "step": 4800
+    },
+    {
+      "epoch": 10.62809917355372,
+      "grad_norm": 9.49361515045166,
+      "learning_rate": 1.0416054821341167e-05,
+      "loss": 1.3357,
+      "step": 4825
+    },
+    {
+      "epoch": 10.683195592286502,
+      "grad_norm": 14.282561302185059,
+      "learning_rate": 1.0354870288790995e-05,
+      "loss": 1.3724,
+      "step": 4850
+    },
+    {
+      "epoch": 10.738292011019285,
+      "grad_norm": 33.559486389160156,
+      "learning_rate": 1.0293685756240824e-05,
+      "loss": 1.4198,
+      "step": 4875
+    },
+    {
+      "epoch": 10.793388429752067,
+      "grad_norm": 10.780166625976562,
+      "learning_rate": 1.0232501223690653e-05,
+      "loss": 1.3685,
+      "step": 4900
+    },
+    {
+      "epoch": 10.848484848484848,
+      "grad_norm": 6.057034492492676,
+      "learning_rate": 1.017131669114048e-05,
+      "loss": 1.3914,
+      "step": 4925
+    },
+    {
+      "epoch": 10.90358126721763,
+      "grad_norm": 9.05207347869873,
+      "learning_rate": 1.0110132158590308e-05,
+      "loss": 1.3619,
+      "step": 4950
+    },
+    {
+      "epoch": 10.958677685950413,
+      "grad_norm": 6.542205810546875,
+      "learning_rate": 1.0048947626040137e-05,
+      "loss": 1.3213,
+      "step": 4975
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.08581862446743761,
+      "eval_loss": 0.7311891317367554,
+      "eval_macro_f1": 0.50394695762473,
+      "eval_micro_f1": 0.5346751257930431,
+      "eval_runtime": 0.7471,
+      "eval_samples_per_second": 2199.246,
+      "eval_steps_per_second": 137.871,
+      "eval_weighted_f1": 0.5450425663519038,
+      "step": 4994
+    },
+    {
+      "epoch": 11.013223140495867,
+      "grad_norm": 9.045429229736328,
+      "learning_rate": 9.987763093489966e-06,
+      "loss": 1.3719,
+      "step": 5000
+    },
+    {
+      "epoch": 11.06831955922865,
+      "grad_norm": 16.91636848449707,
+      "learning_rate": 9.926578560939796e-06,
+      "loss": 1.304,
+      "step": 5025
+    },
+    {
+      "epoch": 11.123415977961432,
+      "grad_norm": 12.180585861206055,
+      "learning_rate": 9.865394028389625e-06,
+      "loss": 1.282,
+      "step": 5050
+    },
+    {
+      "epoch": 11.178512396694215,
+      "grad_norm": 6.918673515319824,
+      "learning_rate": 9.804209495839453e-06,
+      "loss": 1.3003,
+      "step": 5075
+    },
+    {
+      "epoch": 11.233608815426997,
+      "grad_norm": 9.964341163635254,
+      "learning_rate": 9.743024963289282e-06,
+      "loss": 1.2704,
+      "step": 5100
+    },
+    {
+      "epoch": 11.28870523415978,
+      "grad_norm": 13.528480529785156,
+      "learning_rate": 9.68184043073911e-06,
+      "loss": 1.3075,
+      "step": 5125
+    },
+    {
+      "epoch": 11.343801652892562,
+      "grad_norm": 9.736034393310547,
+      "learning_rate": 9.620655898188939e-06,
+      "loss": 1.2758,
+      "step": 5150
+    },
+    {
+      "epoch": 11.398898071625345,
+      "grad_norm": 9.628288269042969,
+      "learning_rate": 9.559471365638768e-06,
+      "loss": 1.2939,
+      "step": 5175
+    },
+    {
+      "epoch": 11.453994490358127,
+      "grad_norm": 9.606924057006836,
+      "learning_rate": 9.498286833088596e-06,
+      "loss": 1.291,
+      "step": 5200
+    },
+    {
+      "epoch": 11.50909090909091,
+      "grad_norm": 9.936850547790527,
+      "learning_rate": 9.437102300538425e-06,
+      "loss": 1.2681,
+      "step": 5225
+    },
+    {
+      "epoch": 11.564187327823692,
+      "grad_norm": 7.868658542633057,
+      "learning_rate": 9.375917767988254e-06,
+      "loss": 1.2479,
+      "step": 5250
+    },
+    {
+      "epoch": 11.619283746556473,
+      "grad_norm": 8.432409286499023,
+      "learning_rate": 9.31473323543808e-06,
+      "loss": 1.2672,
+      "step": 5275
+    },
+    {
+      "epoch": 11.674380165289255,
+      "grad_norm": 8.701844215393066,
+      "learning_rate": 9.253548702887911e-06,
+      "loss": 1.3027,
+      "step": 5300
+    },
+    {
+      "epoch": 11.729476584022038,
+      "grad_norm": 16.532873153686523,
+      "learning_rate": 9.19236417033774e-06,
+      "loss": 1.244,
+      "step": 5325
+    },
+    {
+      "epoch": 11.78457300275482,
+      "grad_norm": 8.731383323669434,
+      "learning_rate": 9.131179637787568e-06,
+      "loss": 1.2405,
+      "step": 5350
+    },
+    {
+      "epoch": 11.839669421487603,
+      "grad_norm": 10.397244453430176,
+      "learning_rate": 9.069995105237397e-06,
+      "loss": 1.2831,
+      "step": 5375
+    },
+    {
+      "epoch": 11.894765840220385,
+      "grad_norm": 10.608807563781738,
+      "learning_rate": 9.008810572687226e-06,
+      "loss": 1.1933,
+      "step": 5400
+    },
+    {
+      "epoch": 11.949862258953168,
+      "grad_norm": 7.912414073944092,
+      "learning_rate": 8.947626040137054e-06,
+      "loss": 1.2603,
+      "step": 5425
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.08642726719415703,
+      "eval_loss": 0.7480849027633667,
+      "eval_macro_f1": 0.5113890808616641,
+      "eval_micro_f1": 0.5431139718122295,
+      "eval_runtime": 0.774,
+      "eval_samples_per_second": 2122.643,
+      "eval_steps_per_second": 133.069,
+      "eval_weighted_f1": 0.5492742671911292,
+      "step": 5448
+    },
+    {
+      "epoch": 12.004407713498622,
+      "grad_norm": 9.934858322143555,
+      "learning_rate": 8.886441507586883e-06,
+      "loss": 1.2531,
+      "step": 5450
+    },
+    {
+      "epoch": 12.059504132231405,
+      "grad_norm": 6.843869209289551,
+      "learning_rate": 8.825256975036712e-06,
+      "loss": 1.1773,
+      "step": 5475
+    },
+    {
+      "epoch": 12.114600550964187,
+      "grad_norm": 7.1369853019714355,
+      "learning_rate": 8.76407244248654e-06,
+      "loss": 1.226,
+      "step": 5500
+    },
+    {
+      "epoch": 12.16969696969697,
+      "grad_norm": 8.611149787902832,
+      "learning_rate": 8.702887909936369e-06,
+      "loss": 1.1832,
+      "step": 5525
+    },
+    {
+      "epoch": 12.224793388429752,
+      "grad_norm": 12.588682174682617,
+      "learning_rate": 8.641703377386198e-06,
+      "loss": 1.2135,
+      "step": 5550
+    },
+    {
+      "epoch": 12.279889807162535,
+      "grad_norm": 16.866628646850586,
+      "learning_rate": 8.580518844836026e-06,
+      "loss": 1.1874,
+      "step": 5575
+    },
+    {
+      "epoch": 12.334986225895317,
+      "grad_norm": 21.434120178222656,
+      "learning_rate": 8.519334312285855e-06,
+      "loss": 1.2108,
+      "step": 5600
+    },
+    {
+      "epoch": 12.3900826446281,
+      "grad_norm": 14.522948265075684,
+      "learning_rate": 8.458149779735683e-06,
+      "loss": 1.1988,
+      "step": 5625
+    },
+    {
+      "epoch": 12.445179063360882,
+      "grad_norm": 13.422613143920898,
+      "learning_rate": 8.396965247185512e-06,
+      "loss": 1.2145,
+      "step": 5650
+    },
+    {
+      "epoch": 12.500275482093663,
+      "grad_norm": 9.499993324279785,
+      "learning_rate": 8.33578071463534e-06,
+      "loss": 1.1397,
+      "step": 5675
+    },
+    {
+      "epoch": 12.555371900826445,
+      "grad_norm": 10.744636535644531,
+      "learning_rate": 8.27459618208517e-06,
+      "loss": 1.1531,
+      "step": 5700
+    },
+    {
+      "epoch": 12.610468319559228,
+      "grad_norm": 9.612398147583008,
+      "learning_rate": 8.213411649534998e-06,
+      "loss": 1.1706,
+      "step": 5725
+    },
+    {
+      "epoch": 12.66556473829201,
+      "grad_norm": 12.579903602600098,
+      "learning_rate": 8.152227116984827e-06,
+      "loss": 1.1717,
+      "step": 5750
+    },
+    {
+      "epoch": 12.720661157024793,
+      "grad_norm": 9.093660354614258,
+      "learning_rate": 8.091042584434655e-06,
+      "loss": 1.1956,
+      "step": 5775
+    },
+    {
+      "epoch": 12.775757575757575,
+      "grad_norm": 18.263643264770508,
+      "learning_rate": 8.029858051884484e-06,
+      "loss": 1.2099,
+      "step": 5800
+    },
+    {
+      "epoch": 12.830853994490358,
+      "grad_norm": 17.743852615356445,
+      "learning_rate": 7.968673519334313e-06,
+      "loss": 1.1798,
+      "step": 5825
+    },
+    {
+      "epoch": 12.88595041322314,
+      "grad_norm": 7.139863967895508,
+      "learning_rate": 7.907488986784141e-06,
+      "loss": 1.2019,
+      "step": 5850
+    },
+    {
+      "epoch": 12.941046831955923,
+      "grad_norm": 15.34650707244873,
+      "learning_rate": 7.84630445423397e-06,
+      "loss": 1.2193,
+      "step": 5875
+    },
+    {
+      "epoch": 12.996143250688705,
+      "grad_norm": 10.276274681091309,
+      "learning_rate": 7.785119921683799e-06,
+      "loss": 1.1547,
+      "step": 5900
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.09494826536822885,
+      "eval_loss": 0.7594033479690552,
+      "eval_macro_f1": 0.5202326488837883,
+      "eval_micro_f1": 0.5480618418104414,
+      "eval_runtime": 0.7284,
+      "eval_samples_per_second": 2255.513,
+      "eval_steps_per_second": 141.399,
+      "eval_weighted_f1": 0.5550034950523685,
+      "step": 5902
+    },
+    {
+      "epoch": 13.05068870523416,
+      "grad_norm": 12.496541023254395,
+      "learning_rate": 7.723935389133627e-06,
+      "loss": 1.0906,
+      "step": 5925
+    },
+    {
+      "epoch": 13.105785123966943,
+      "grad_norm": 7.705901145935059,
+      "learning_rate": 7.662750856583456e-06,
+      "loss": 1.1254,
+      "step": 5950
+    },
+    {
+      "epoch": 13.160881542699725,
+      "grad_norm": 7.635675430297852,
+      "learning_rate": 7.601566324033285e-06,
+      "loss": 1.1406,
+      "step": 5975
+    },
+    {
+      "epoch": 13.215977961432507,
+      "grad_norm": 8.631380081176758,
+      "learning_rate": 7.540381791483113e-06,
+      "loss": 1.1202,
+      "step": 6000
+    },
+    {
+      "epoch": 13.27107438016529,
+      "grad_norm": 22.504575729370117,
+      "learning_rate": 7.479197258932942e-06,
+      "loss": 1.139,
+      "step": 6025
+    },
+    {
+      "epoch": 13.326170798898072,
+      "grad_norm": 9.432455062866211,
+      "learning_rate": 7.4180127263827705e-06,
+      "loss": 1.1305,
+      "step": 6050
+    },
+    {
+      "epoch": 13.381267217630853,
+      "grad_norm": 9.018024444580078,
+      "learning_rate": 7.3568281938326e-06,
+      "loss": 1.1186,
+      "step": 6075
+    },
+    {
+      "epoch": 13.436363636363636,
+      "grad_norm": 9.342816352844238,
+      "learning_rate": 7.295643661282429e-06,
+      "loss": 1.0854,
+      "step": 6100
+    },
+    {
+      "epoch": 13.491460055096418,
+      "grad_norm": 4.410456657409668,
+      "learning_rate": 7.234459128732257e-06,
+      "loss": 1.1191,
+      "step": 6125
+    },
+    {
+      "epoch": 13.5465564738292,
+      "grad_norm": 6.352032661437988,
+      "learning_rate": 7.173274596182086e-06,
+      "loss": 1.1093,
+      "step": 6150
+    },
+    {
+      "epoch": 13.601652892561983,
+      "grad_norm": 6.334588527679443,
+      "learning_rate": 7.112090063631915e-06,
+      "loss": 1.1429,
+      "step": 6175
+    },
+    {
+      "epoch": 13.656749311294766,
+      "grad_norm": 7.5459418296813965,
+      "learning_rate": 7.050905531081743e-06,
+      "loss": 1.069,
+      "step": 6200
+    },
+    {
+      "epoch": 13.711845730027548,
+      "grad_norm": 9.283977508544922,
+      "learning_rate": 6.989720998531572e-06,
+      "loss": 1.0905,
+      "step": 6225
+    },
+    {
+      "epoch": 13.76694214876033,
+      "grad_norm": 9.334644317626953,
+      "learning_rate": 6.928536465981401e-06,
+      "loss": 1.107,
+      "step": 6250
+    },
+    {
+      "epoch": 13.822038567493113,
+      "grad_norm": 6.586839199066162,
+      "learning_rate": 6.867351933431229e-06,
+      "loss": 1.091,
+      "step": 6275
+    },
+    {
+      "epoch": 13.877134986225895,
+      "grad_norm": 11.92888069152832,
+      "learning_rate": 6.806167400881057e-06,
+      "loss": 1.1088,
+      "step": 6300
+    },
+    {
+      "epoch": 13.932231404958678,
+      "grad_norm": 9.216996192932129,
+      "learning_rate": 6.744982868330886e-06,
+      "loss": 1.0462,
+      "step": 6325
+    },
+    {
+      "epoch": 13.98732782369146,
+      "grad_norm": 6.684733867645264,
+      "learning_rate": 6.683798335780716e-06,
+      "loss": 1.0711,
+      "step": 6350
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.10590383444917834,
+      "eval_loss": 0.7497881054878235,
+      "eval_macro_f1": 0.5163370092093984,
+      "eval_micro_f1": 0.5468609865470851,
+      "eval_runtime": 0.727,
+      "eval_samples_per_second": 2260.083,
+      "eval_steps_per_second": 141.685,
+      "eval_weighted_f1": 0.5529208492347623,
+      "step": 6356
+    },
+    {
+      "epoch": 14.041873278236915,
+      "grad_norm": 7.138929843902588,
+      "learning_rate": 6.622613803230544e-06,
+      "loss": 1.0558,
+      "step": 6375
+    },
+    {
+      "epoch": 14.096969696969698,
+      "grad_norm": 10.733924865722656,
+      "learning_rate": 6.5614292706803725e-06,
+      "loss": 1.0532,
+      "step": 6400
+    },
+    {
+      "epoch": 14.15206611570248,
+      "grad_norm": 12.12220573425293,
+      "learning_rate": 6.500244738130201e-06,
+      "loss": 1.0375,
+      "step": 6425
+    },
+    {
+      "epoch": 14.207162534435263,
+      "grad_norm": 17.76565170288086,
+      "learning_rate": 6.43906020558003e-06,
+      "loss": 1.0618,
+      "step": 6450
+    },
+    {
+      "epoch": 14.262258953168043,
+      "grad_norm": 7.568203449249268,
+      "learning_rate": 6.3778756730298585e-06,
+      "loss": 1.0354,
+      "step": 6475
+    },
+    {
+      "epoch": 14.317355371900826,
+      "grad_norm": 10.111392974853516,
+      "learning_rate": 6.316691140479687e-06,
+      "loss": 1.0496,
+      "step": 6500
+    },
+    {
+      "epoch": 14.372451790633608,
+      "grad_norm": 6.254069805145264,
+      "learning_rate": 6.255506607929516e-06,
+      "loss": 1.0712,
+      "step": 6525
+    },
+    {
+      "epoch": 14.42754820936639,
+      "grad_norm": 7.912994861602783,
+      "learning_rate": 6.194322075379344e-06,
+      "loss": 1.0293,
+      "step": 6550
+    },
+    {
+      "epoch": 14.482644628099173,
+      "grad_norm": 12.278776168823242,
+      "learning_rate": 6.133137542829173e-06,
+      "loss": 1.0933,
+      "step": 6575
+    },
+    {
+      "epoch": 14.537741046831956,
+      "grad_norm": 12.193331718444824,
+      "learning_rate": 6.071953010279002e-06,
+      "loss": 1.0321,
+      "step": 6600
+    },
+    {
+      "epoch": 14.592837465564738,
+      "grad_norm": 6.962792873382568,
+      "learning_rate": 6.010768477728831e-06,
+      "loss": 1.0203,
+      "step": 6625
+    },
+    {
+      "epoch": 14.64793388429752,
+      "grad_norm": 14.487834930419922,
+      "learning_rate": 5.94958394517866e-06,
+      "loss": 1.0275,
+      "step": 6650
+    },
+    {
+      "epoch": 14.703030303030303,
+      "grad_norm": 26.301523208618164,
+      "learning_rate": 5.8883994126284885e-06,
+      "loss": 1.0524,
+      "step": 6675
+    },
+    {
+      "epoch": 14.758126721763086,
+      "grad_norm": 8.678800582885742,
+      "learning_rate": 5.827214880078317e-06,
+      "loss": 1.0461,
+      "step": 6700
+    },
+    {
+      "epoch": 14.813223140495868,
+      "grad_norm": 7.605831146240234,
+      "learning_rate": 5.766030347528146e-06,
+      "loss": 1.0164,
+      "step": 6725
+    },
+    {
+      "epoch": 14.86831955922865,
+      "grad_norm": 16.34834861755371,
+      "learning_rate": 5.704845814977974e-06,
+      "loss": 1.0022,
+      "step": 6750
+    },
+    {
+      "epoch": 14.923415977961433,
+      "grad_norm": 10.478272438049316,
+      "learning_rate": 5.643661282427802e-06,
+      "loss": 1.0283,
+      "step": 6775
+    },
+    {
+      "epoch": 14.978512396694216,
+      "grad_norm": 6.517714977264404,
+      "learning_rate": 5.582476749877631e-06,
+      "loss": 0.9951,
+      "step": 6800
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.1199026171637249,
+      "eval_loss": 0.786878764629364,
+      "eval_macro_f1": 0.5303045519693917,
+      "eval_micro_f1": 0.5607280364018201,
+      "eval_runtime": 0.7664,
+      "eval_samples_per_second": 2143.802,
+      "eval_steps_per_second": 134.395,
+      "eval_weighted_f1": 0.5648358632177137,
+      "step": 6810
+    },
+    {
+      "epoch": 15.03305785123967,
+      "grad_norm": 12.449422836303711,
+      "learning_rate": 5.5212922173274596e-06,
+      "loss": 1.0273,
+      "step": 6825
+    },
+    {
+      "epoch": 15.088154269972453,
+      "grad_norm": 16.88009262084961,
+      "learning_rate": 5.460107684777288e-06,
+      "loss": 0.9533,
+      "step": 6850
+    },
+    {
+      "epoch": 15.143250688705233,
+      "grad_norm": 13.051454544067383,
+      "learning_rate": 5.398923152227118e-06,
+      "loss": 1.0137,
+      "step": 6875
+    },
+    {
+      "epoch": 15.198347107438016,
+      "grad_norm": 9.161343574523926,
+      "learning_rate": 5.337738619676946e-06,
+      "loss": 0.963,
+      "step": 6900
+    },
+    {
+      "epoch": 15.253443526170798,
+      "grad_norm": 16.16594886779785,
+      "learning_rate": 5.276554087126775e-06,
+      "loss": 0.9616,
+      "step": 6925
+    },
+    {
+      "epoch": 15.30853994490358,
+      "grad_norm": 8.046670913696289,
+      "learning_rate": 5.215369554576604e-06,
+      "loss": 0.983,
+      "step": 6950
+    },
+    {
+      "epoch": 15.363636363636363,
+      "grad_norm": 10.402591705322266,
+      "learning_rate": 5.154185022026432e-06,
+      "loss": 1.0406,
+      "step": 6975
+    },
+    {
+      "epoch": 15.418732782369146,
+      "grad_norm": 6.809611797332764,
+      "learning_rate": 5.093000489476261e-06,
+      "loss": 1.0326,
+      "step": 7000
+    },
+    {
+      "epoch": 15.473829201101928,
+      "grad_norm": 8.409727096557617,
+      "learning_rate": 5.03181595692609e-06,
+      "loss": 0.9884,
+      "step": 7025
+    },
+    {
+      "epoch": 15.52892561983471,
+      "grad_norm": 5.694775581359863,
+      "learning_rate": 4.970631424375918e-06,
+      "loss": 0.978,
+      "step": 7050
+    },
+    {
+      "epoch": 15.584022038567493,
+      "grad_norm": 20.40342140197754,
+      "learning_rate": 4.909446891825747e-06,
+      "loss": 1.0045,
+      "step": 7075
+    },
+    {
+      "epoch": 15.639118457300276,
+      "grad_norm": 10.138327598571777,
+      "learning_rate": 4.8482623592755756e-06,
+      "loss": 0.9975,
+      "step": 7100
+    },
+    {
+      "epoch": 15.694214876033058,
+      "grad_norm": 8.172189712524414,
+      "learning_rate": 4.787077826725404e-06,
+      "loss": 0.9591,
+      "step": 7125
+    },
+    {
+      "epoch": 15.74931129476584,
+      "grad_norm": 9.392911911010742,
+      "learning_rate": 4.725893294175233e-06,
+      "loss": 0.992,
+      "step": 7150
+    },
+    {
+      "epoch": 15.804407713498623,
+      "grad_norm": 11.717973709106445,
+      "learning_rate": 4.6647087616250615e-06,
+      "loss": 0.9686,
+      "step": 7175
+    },
+    {
+      "epoch": 15.859504132231406,
+      "grad_norm": 5.921899318695068,
+      "learning_rate": 4.60352422907489e-06,
+      "loss": 0.9505,
+      "step": 7200
+    },
+    {
+      "epoch": 15.914600550964188,
+      "grad_norm": 6.638045310974121,
+      "learning_rate": 4.542339696524719e-06,
+      "loss": 0.9778,
+      "step": 7225
+    },
+    {
+      "epoch": 15.969696969696969,
+      "grad_norm": 11.249752044677734,
+      "learning_rate": 4.4811551639745475e-06,
+      "loss": 0.9656,
+      "step": 7250
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.12903225806451613,
+      "eval_loss": 0.7909542322158813,
+      "eval_macro_f1": 0.5332530090007124,
+      "eval_micro_f1": 0.5646370023419204,
+      "eval_runtime": 0.9191,
+      "eval_samples_per_second": 1787.597,
+      "eval_steps_per_second": 112.065,
+      "eval_weighted_f1": 0.5695274753104705,
+      "step": 7264
+    },
+    {
+      "epoch": 16.024242424242424,
+      "grad_norm": 22.084747314453125,
+      "learning_rate": 4.419970631424376e-06,
+      "loss": 0.9781,
+      "step": 7275
+    },
+    {
+      "epoch": 16.079338842975208,
+      "grad_norm": 7.531091690063477,
+      "learning_rate": 4.358786098874205e-06,
+      "loss": 0.9447,
+      "step": 7300
+    },
+    {
+      "epoch": 16.13443526170799,
+      "grad_norm": 9.669242858886719,
+      "learning_rate": 4.297601566324034e-06,
+      "loss": 0.9171,
+      "step": 7325
+    },
+    {
+      "epoch": 16.189531680440773,
+      "grad_norm": 17.005956649780273,
+      "learning_rate": 4.236417033773862e-06,
+      "loss": 0.9611,
+      "step": 7350
+    },
+    {
+      "epoch": 16.244628099173553,
+      "grad_norm": 10.098047256469727,
+      "learning_rate": 4.175232501223691e-06,
+      "loss": 0.9458,
+      "step": 7375
+    },
+    {
+      "epoch": 16.299724517906338,
+      "grad_norm": 8.807305335998535,
+      "learning_rate": 4.114047968673519e-06,
+      "loss": 0.9424,
+      "step": 7400
+    },
+    {
+      "epoch": 16.35482093663912,
+      "grad_norm": 10.43460464477539,
+      "learning_rate": 4.052863436123348e-06,
+      "loss": 0.9741,
+      "step": 7425
+    },
+    {
+      "epoch": 16.4099173553719,
+      "grad_norm": 9.416433334350586,
+      "learning_rate": 3.991678903573177e-06,
+      "loss": 0.9422,
+      "step": 7450
+    },
+    {
+      "epoch": 16.465013774104683,
+      "grad_norm": 26.025875091552734,
+      "learning_rate": 3.930494371023006e-06,
+      "loss": 0.9808,
+      "step": 7475
+    },
+    {
+      "epoch": 16.520110192837464,
+      "grad_norm": 6.642582893371582,
+      "learning_rate": 3.869309838472835e-06,
+      "loss": 0.932,
+      "step": 7500
+    },
+    {
+      "epoch": 16.57520661157025,
+      "grad_norm": 13.423791885375977,
+      "learning_rate": 3.808125305922663e-06,
+      "loss": 0.9568,
+      "step": 7525
+    },
+    {
+      "epoch": 16.63030303030303,
+      "grad_norm": 7.4708757400512695,
+      "learning_rate": 3.7469407733724917e-06,
+      "loss": 0.9453,
+      "step": 7550
+    },
+    {
+      "epoch": 16.685399449035813,
+      "grad_norm": 5.990961074829102,
+      "learning_rate": 3.6857562408223203e-06,
+      "loss": 0.928,
+      "step": 7575
+    },
+    {
+      "epoch": 16.740495867768594,
+      "grad_norm": 7.558459758758545,
+      "learning_rate": 3.6245717082721494e-06,
+      "loss": 0.9763,
+      "step": 7600
+    },
+    {
+      "epoch": 16.79559228650138,
+      "grad_norm": 5.2827935218811035,
+      "learning_rate": 3.563387175721978e-06,
+      "loss": 0.9528,
+      "step": 7625
+    },
+    {
+      "epoch": 16.85068870523416,
+      "grad_norm": 15.335264205932617,
+      "learning_rate": 3.5022026431718063e-06,
+      "loss": 0.933,
+      "step": 7650
+    },
+    {
+      "epoch": 16.905785123966943,
+      "grad_norm": 8.040648460388184,
+      "learning_rate": 3.441018110621635e-06,
+      "loss": 0.9064,
+      "step": 7675
+    },
+    {
+      "epoch": 16.960881542699724,
+      "grad_norm": 14.046770095825195,
+      "learning_rate": 3.3798335780714636e-06,
+      "loss": 0.9466,
+      "step": 7700
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.13268411442483263,
+      "eval_loss": 0.7997220754623413,
+      "eval_macro_f1": 0.5386078032263989,
+      "eval_micro_f1": 0.566721972031287,
+      "eval_runtime": 0.708,
+      "eval_samples_per_second": 2320.756,
+      "eval_steps_per_second": 145.489,
+      "eval_weighted_f1": 0.5716301004766311,
+      "step": 7718
+    },
+    {
+      "epoch": 17.01542699724518,
+      "grad_norm": 12.992655754089355,
+      "learning_rate": 3.3186490455212927e-06,
+      "loss": 0.9179,
+      "step": 7725
+    },
+    {
+      "epoch": 17.070523415977963,
+      "grad_norm": 16.123558044433594,
+      "learning_rate": 3.2574645129711213e-06,
+      "loss": 0.9245,
+      "step": 7750
+    },
+    {
+      "epoch": 17.125619834710744,
+      "grad_norm": 6.627078533172607,
+      "learning_rate": 3.19627998042095e-06,
+      "loss": 0.9164,
+      "step": 7775
+    },
+    {
+      "epoch": 17.180716253443528,
+      "grad_norm": 10.419363021850586,
+      "learning_rate": 3.1350954478707786e-06,
+      "loss": 0.8796,
+      "step": 7800
+    },
+    {
+      "epoch": 17.23581267217631,
+      "grad_norm": 17.208833694458008,
+      "learning_rate": 3.073910915320607e-06,
+      "loss": 0.9119,
+      "step": 7825
+    },
+    {
+      "epoch": 17.29090909090909,
+      "grad_norm": 10.589327812194824,
+      "learning_rate": 3.0127263827704355e-06,
+      "loss": 0.9052,
+      "step": 7850
+    },
+    {
+      "epoch": 17.346005509641873,
+      "grad_norm": 11.223331451416016,
+      "learning_rate": 2.9515418502202646e-06,
+      "loss": 0.9467,
+      "step": 7875
+    },
+    {
+      "epoch": 17.401101928374654,
+      "grad_norm": 8.815620422363281,
+      "learning_rate": 2.8903573176700932e-06,
+      "loss": 0.9018,
+      "step": 7900
+    },
+    {
+      "epoch": 17.45619834710744,
+      "grad_norm": 7.268686294555664,
+      "learning_rate": 2.829172785119922e-06,
+      "loss": 0.9054,
+      "step": 7925
+    },
+    {
+      "epoch": 17.51129476584022,
+      "grad_norm": 14.526104927062988,
+      "learning_rate": 2.7679882525697505e-06,
+      "loss": 0.9042,
+      "step": 7950
+    },
+    {
+      "epoch": 17.566391184573003,
+      "grad_norm": 8.908108711242676,
+      "learning_rate": 2.706803720019579e-06,
+      "loss": 0.8553,
+      "step": 7975
+    },
+    {
+      "epoch": 17.621487603305784,
+      "grad_norm": 5.285924434661865,
+      "learning_rate": 2.6456191874694083e-06,
+      "loss": 0.9036,
+      "step": 8000
+    },
+    {
+      "epoch": 17.67658402203857,
+      "grad_norm": 9.32984733581543,
+      "learning_rate": 2.584434654919237e-06,
+      "loss": 0.9144,
+      "step": 8025
+    },
+    {
+      "epoch": 17.73168044077135,
+      "grad_norm": 6.950353145599365,
+      "learning_rate": 2.523250122369065e-06,
+      "loss": 0.9188,
+      "step": 8050
+    },
+    {
+      "epoch": 17.786776859504133,
+      "grad_norm": 7.641226291656494,
+      "learning_rate": 2.4620655898188938e-06,
+      "loss": 0.9088,
+      "step": 8075
+    },
+    {
+      "epoch": 17.841873278236914,
+      "grad_norm": 10.661919593811035,
+      "learning_rate": 2.400881057268723e-06,
+      "loss": 0.8949,
+      "step": 8100
+    },
+    {
+      "epoch": 17.8969696969697,
+      "grad_norm": 5.028895378112793,
+      "learning_rate": 2.3396965247185515e-06,
+      "loss": 0.9146,
+      "step": 8125
+    },
+    {
+      "epoch": 17.95206611570248,
+      "grad_norm": 10.540121078491211,
+      "learning_rate": 2.27851199216838e-06,
+      "loss": 0.9099,
+      "step": 8150
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.13451004260499086,
+      "eval_loss": 0.8080140352249146,
+      "eval_macro_f1": 0.535982460938898,
+      "eval_micro_f1": 0.566127497621313,
+      "eval_runtime": 0.7341,
+      "eval_samples_per_second": 2238.178,
+      "eval_steps_per_second": 140.312,
+      "eval_weighted_f1": 0.570448162053066,
+      "step": 8172
+    },
+    {
+      "epoch": 18.006611570247934,
+      "grad_norm": 7.576343536376953,
+      "learning_rate": 2.217327459618209e-06,
+      "loss": 0.8805,
+      "step": 8175
+    },
+    {
+      "epoch": 18.061707988980718,
+      "grad_norm": 8.76292610168457,
+      "learning_rate": 2.1561429270680375e-06,
+      "loss": 0.8528,
+      "step": 8200
+    },
+    {
+      "epoch": 18.1168044077135,
+      "grad_norm": 7.125899791717529,
+      "learning_rate": 2.094958394517866e-06,
+      "loss": 0.872,
+      "step": 8225
+    },
+    {
+      "epoch": 18.17190082644628,
+      "grad_norm": 7.739444732666016,
+      "learning_rate": 2.0337738619676948e-06,
+      "loss": 0.8481,
+      "step": 8250
+    },
+    {
+      "epoch": 18.226997245179064,
+      "grad_norm": 9.100935935974121,
+      "learning_rate": 1.9725893294175234e-06,
+      "loss": 0.8929,
+      "step": 8275
+    },
+    {
+      "epoch": 18.282093663911844,
+      "grad_norm": 6.645345687866211,
+      "learning_rate": 1.911404796867352e-06,
+      "loss": 0.8944,
+      "step": 8300
+    },
+    {
+      "epoch": 18.33719008264463,
+      "grad_norm": 10.697574615478516,
+      "learning_rate": 1.8502202643171807e-06,
+      "loss": 0.8416,
+      "step": 8325
+    },
+    {
+      "epoch": 18.39228650137741,
+      "grad_norm": 10.242743492126465,
+      "learning_rate": 1.7890357317670096e-06,
+      "loss": 0.9235,
+      "step": 8350
+    },
+    {
+      "epoch": 18.447382920110194,
+      "grad_norm": 8.19653606414795,
+      "learning_rate": 1.7278511992168382e-06,
+      "loss": 0.8798,
+      "step": 8375
+    },
+    {
+      "epoch": 18.502479338842974,
+      "grad_norm": 7.815878868103027,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.8979,
+      "step": 8400
+    },
+    {
+      "epoch": 18.55757575757576,
+      "grad_norm": 10.144920349121094,
+      "learning_rate": 1.6054821341164955e-06,
+      "loss": 0.9083,
+      "step": 8425
+    },
+    {
+      "epoch": 18.61267217630854,
+      "grad_norm": 12.199047088623047,
+      "learning_rate": 1.5442976015663242e-06,
+      "loss": 0.8923,
+      "step": 8450
+    },
+    {
+      "epoch": 18.667768595041323,
+      "grad_norm": 14.669466972351074,
+      "learning_rate": 1.4831130690161526e-06,
+      "loss": 0.9132,
+      "step": 8475
+    },
+    {
+      "epoch": 18.722865013774104,
+      "grad_norm": 6.225057125091553,
+      "learning_rate": 1.4219285364659815e-06,
+      "loss": 0.8451,
+      "step": 8500
+    },
+    {
+      "epoch": 18.77796143250689,
+      "grad_norm": 5.0352301597595215,
+      "learning_rate": 1.3607440039158101e-06,
+      "loss": 0.8885,
+      "step": 8525
+    },
+    {
+      "epoch": 18.83305785123967,
+      "grad_norm": 11.503164291381836,
+      "learning_rate": 1.299559471365639e-06,
+      "loss": 0.8928,
+      "step": 8550
+    },
+    {
+      "epoch": 18.888154269972453,
+      "grad_norm": 7.827541828155518,
+      "learning_rate": 1.2383749388154676e-06,
+      "loss": 0.8906,
+      "step": 8575
+    },
+    {
+      "epoch": 18.943250688705234,
+      "grad_norm": 7.901440143585205,
+      "learning_rate": 1.1771904062652963e-06,
+      "loss": 0.8679,
+      "step": 8600
+    },
+    {
+      "epoch": 18.998347107438015,
+      "grad_norm": 6.281285762786865,
+      "learning_rate": 1.116005873715125e-06,
+      "loss": 0.8768,
+      "step": 8625
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.13694461351186854,
+      "eval_loss": 0.8208908438682556,
+      "eval_macro_f1": 0.5389595030296993,
+      "eval_micro_f1": 0.5692087018885967,
+      "eval_runtime": 0.7096,
+      "eval_samples_per_second": 2315.51,
+      "eval_steps_per_second": 145.16,
+      "eval_weighted_f1": 0.5728033344104219,
+      "step": 8626
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 9080,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

ood-vigoemotions-phobert-source-source-01-s42-20260826T131538Z-0c086fe0/trainer/checkpoint-9080/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

ood-vigoemotions-phobert-source-source-01-s42-20260826T131538Z-0c086fe0/trainer/checkpoint-9080/special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-01-s42-20260826T131538Z-0c086fe0/trainer/checkpoint-9080/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 256,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-01-s42-20260826T131538Z-0c086fe0/trainer/checkpoint-9080/trainer_state.json

@@ -0,0 +1,2824 @@
+{
+  "best_global_step": 9080,
+  "best_metric": 0.5395569926876561,
+  "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vigoemotions-phobert-source-source-01-s42-20260826T131538Z-0c086fe0/trainer/checkpoint-9080",
+  "epoch": 20.0,
+  "eval_steps": 500,
+  "global_step": 9080,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.05509641873278237,
+      "grad_norm": 4.227837562561035,
+      "learning_rate": 5.286343612334802e-07,
+      "loss": 5.2039,
+      "step": 25
+    },
+    {
+      "epoch": 0.11019283746556474,
+      "grad_norm": 5.331326484680176,
+      "learning_rate": 1.0792951541850223e-06,
+      "loss": 5.0579,
+      "step": 50
+    },
+    {
+      "epoch": 0.1652892561983471,
+      "grad_norm": 5.1418375968933105,
+      "learning_rate": 1.629955947136564e-06,
+      "loss": 5.1842,
+      "step": 75
+    },
+    {
+      "epoch": 0.22038567493112948,
+      "grad_norm": 4.12420654296875,
+      "learning_rate": 2.180616740088106e-06,
+      "loss": 5.1411,
+      "step": 100
+    },
+    {
+      "epoch": 0.27548209366391185,
+      "grad_norm": 3.9862260818481445,
+      "learning_rate": 2.731277533039648e-06,
+      "loss": 5.2115,
+      "step": 125
+    },
+    {
+      "epoch": 0.3305785123966942,
+      "grad_norm": 4.397335052490234,
+      "learning_rate": 3.2819383259911898e-06,
+      "loss": 5.1033,
+      "step": 150
+    },
+    {
+      "epoch": 0.3856749311294766,
+      "grad_norm": 6.729996681213379,
+      "learning_rate": 3.8325991189427315e-06,
+      "loss": 5.1591,
+      "step": 175
+    },
+    {
+      "epoch": 0.44077134986225897,
+      "grad_norm": 5.972109317779541,
+      "learning_rate": 4.383259911894274e-06,
+      "loss": 5.1763,
+      "step": 200
+    },
+    {
+      "epoch": 0.49586776859504134,
+      "grad_norm": 5.35798454284668,
+      "learning_rate": 4.933920704845816e-06,
+      "loss": 5.0333,
+      "step": 225
+    },
+    {
+      "epoch": 0.5509641873278237,
+      "grad_norm": 6.948966979980469,
+      "learning_rate": 5.484581497797358e-06,
+      "loss": 4.9953,
+      "step": 250
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 4.730196475982666,
+      "learning_rate": 6.035242290748899e-06,
+      "loss": 4.948,
+      "step": 275
+    },
+    {
+      "epoch": 0.6611570247933884,
+      "grad_norm": 5.168266773223877,
+      "learning_rate": 6.585903083700441e-06,
+      "loss": 4.9737,
+      "step": 300
+    },
+    {
+      "epoch": 0.7162534435261708,
+      "grad_norm": 6.052734851837158,
+      "learning_rate": 7.136563876651983e-06,
+      "loss": 4.8535,
+      "step": 325
+    },
+    {
+      "epoch": 0.7713498622589532,
+      "grad_norm": 6.672458171844482,
+      "learning_rate": 7.687224669603525e-06,
+      "loss": 4.8036,
+      "step": 350
+    },
+    {
+      "epoch": 0.8264462809917356,
+      "grad_norm": 5.597661018371582,
+      "learning_rate": 8.237885462555067e-06,
+      "loss": 4.7429,
+      "step": 375
+    },
+    {
+      "epoch": 0.8815426997245179,
+      "grad_norm": 12.606609344482422,
+      "learning_rate": 8.788546255506607e-06,
+      "loss": 4.5717,
+      "step": 400
+    },
+    {
+      "epoch": 0.9366391184573003,
+      "grad_norm": 6.42344856262207,
+      "learning_rate": 9.339207048458151e-06,
+      "loss": 4.5116,
+      "step": 425
+    },
+    {
+      "epoch": 0.9917355371900827,
+      "grad_norm": 6.29040002822876,
+      "learning_rate": 9.889867841409693e-06,
+      "loss": 4.5005,
+      "step": 450
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 1.075113296508789,
+      "eval_macro_f1": 0.23296379461720287,
+      "eval_micro_f1": 0.2440189231808966,
+      "eval_runtime": 0.8467,
+      "eval_samples_per_second": 1940.521,
+      "eval_steps_per_second": 121.652,
+      "eval_weighted_f1": 0.30895690386581653,
+      "step": 454
+    },
+    {
+      "epoch": 1.0462809917355371,
+      "grad_norm": 9.197685241699219,
+      "learning_rate": 1.0440528634361234e-05,
+      "loss": 4.3713,
+      "step": 475
+    },
+    {
+      "epoch": 1.1013774104683196,
+      "grad_norm": 10.728469848632812,
+      "learning_rate": 1.0991189427312776e-05,
+      "loss": 4.3226,
+      "step": 500
+    },
+    {
+      "epoch": 1.1564738292011019,
+      "grad_norm": 10.500930786132812,
+      "learning_rate": 1.154185022026432e-05,
+      "loss": 4.2326,
+      "step": 525
+    },
+    {
+      "epoch": 1.2115702479338843,
+      "grad_norm": 10.107128143310547,
+      "learning_rate": 1.209251101321586e-05,
+      "loss": 4.1633,
+      "step": 550
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 6.363664150238037,
+      "learning_rate": 1.2643171806167402e-05,
+      "loss": 4.1342,
+      "step": 575
+    },
+    {
+      "epoch": 1.321763085399449,
+      "grad_norm": 6.806942939758301,
+      "learning_rate": 1.3193832599118943e-05,
+      "loss": 4.1034,
+      "step": 600
+    },
+    {
+      "epoch": 1.3768595041322313,
+      "grad_norm": 5.18177604675293,
+      "learning_rate": 1.3744493392070487e-05,
+      "loss": 4.1358,
+      "step": 625
+    },
+    {
+      "epoch": 1.4319559228650138,
+      "grad_norm": 13.11739444732666,
+      "learning_rate": 1.4295154185022027e-05,
+      "loss": 4.0104,
+      "step": 650
+    },
+    {
+      "epoch": 1.487052341597796,
+      "grad_norm": 13.619627952575684,
+      "learning_rate": 1.484581497797357e-05,
+      "loss": 3.9651,
+      "step": 675
+    },
+    {
+      "epoch": 1.5421487603305786,
+      "grad_norm": 9.929632186889648,
+      "learning_rate": 1.539647577092511e-05,
+      "loss": 3.946,
+      "step": 700
+    },
+    {
+      "epoch": 1.5972451790633608,
+      "grad_norm": 11.750797271728516,
+      "learning_rate": 1.5947136563876654e-05,
+      "loss": 3.8697,
+      "step": 725
+    },
+    {
+      "epoch": 1.652341597796143,
+      "grad_norm": 18.73433494567871,
+      "learning_rate": 1.6497797356828196e-05,
+      "loss": 3.8729,
+      "step": 750
+    },
+    {
+      "epoch": 1.7074380165289256,
+      "grad_norm": 9.994765281677246,
+      "learning_rate": 1.7048458149779738e-05,
+      "loss": 3.8399,
+      "step": 775
+    },
+    {
+      "epoch": 1.762534435261708,
+      "grad_norm": 13.98775577545166,
+      "learning_rate": 1.759911894273128e-05,
+      "loss": 3.7893,
+      "step": 800
+    },
+    {
+      "epoch": 1.8176308539944903,
+      "grad_norm": 11.74811840057373,
+      "learning_rate": 1.814977973568282e-05,
+      "loss": 3.8683,
+      "step": 825
+    },
+    {
+      "epoch": 1.8727272727272726,
+      "grad_norm": 10.832721710205078,
+      "learning_rate": 1.8700440528634364e-05,
+      "loss": 3.6864,
+      "step": 850
+    },
+    {
+      "epoch": 1.927823691460055,
+      "grad_norm": 9.301114082336426,
+      "learning_rate": 1.9251101321585906e-05,
+      "loss": 3.8125,
+      "step": 875
+    },
+    {
+      "epoch": 1.9829201101928375,
+      "grad_norm": 9.613875389099121,
+      "learning_rate": 1.9801762114537445e-05,
+      "loss": 3.6356,
+      "step": 900
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 0.8913275003433228,
+      "eval_macro_f1": 0.31342985112925853,
+      "eval_micro_f1": 0.3350537901871774,
+      "eval_runtime": 0.7417,
+      "eval_samples_per_second": 2215.172,
+      "eval_steps_per_second": 138.87,
+      "eval_weighted_f1": 0.390772984117027,
+      "step": 908
+    },
+    {
+      "epoch": 2.037465564738292,
+      "grad_norm": 8.475796699523926,
+      "learning_rate": 1.9960841899167892e-05,
+      "loss": 3.6071,
+      "step": 925
+    },
+    {
+      "epoch": 2.0925619834710742,
+      "grad_norm": 8.122872352600098,
+      "learning_rate": 1.989965736661772e-05,
+      "loss": 3.4062,
+      "step": 950
+    },
+    {
+      "epoch": 2.1476584022038567,
+      "grad_norm": 10.734721183776855,
+      "learning_rate": 1.983847283406755e-05,
+      "loss": 3.5119,
+      "step": 975
+    },
+    {
+      "epoch": 2.202754820936639,
+      "grad_norm": 7.998258113861084,
+      "learning_rate": 1.9777288301517378e-05,
+      "loss": 3.5601,
+      "step": 1000
+    },
+    {
+      "epoch": 2.2578512396694217,
+      "grad_norm": 11.745463371276855,
+      "learning_rate": 1.9716103768967207e-05,
+      "loss": 3.4671,
+      "step": 1025
+    },
+    {
+      "epoch": 2.3129476584022037,
+      "grad_norm": 10.620390892028809,
+      "learning_rate": 1.9654919236417035e-05,
+      "loss": 3.3976,
+      "step": 1050
+    },
+    {
+      "epoch": 2.368044077134986,
+      "grad_norm": 18.419321060180664,
+      "learning_rate": 1.9593734703866864e-05,
+      "loss": 3.5408,
+      "step": 1075
+    },
+    {
+      "epoch": 2.4231404958677687,
+      "grad_norm": 10.657835960388184,
+      "learning_rate": 1.9532550171316693e-05,
+      "loss": 3.3784,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4782369146005507,
+      "grad_norm": 11.976256370544434,
+      "learning_rate": 1.947136563876652e-05,
+      "loss": 3.4036,
+      "step": 1125
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 15.081591606140137,
+      "learning_rate": 1.941018110621635e-05,
+      "loss": 3.3793,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5884297520661157,
+      "grad_norm": 9.404304504394531,
+      "learning_rate": 1.934899657366618e-05,
+      "loss": 3.3192,
+      "step": 1175
+    },
+    {
+      "epoch": 2.643526170798898,
+      "grad_norm": 7.783041477203369,
+      "learning_rate": 1.9287812041116007e-05,
+      "loss": 3.2732,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6986225895316807,
+      "grad_norm": 10.965785026550293,
+      "learning_rate": 1.9226627508565836e-05,
+      "loss": 3.2884,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7537190082644627,
+      "grad_norm": 8.72093677520752,
+      "learning_rate": 1.9165442976015665e-05,
+      "loss": 3.2833,
+      "step": 1250
+    },
+    {
+      "epoch": 2.808815426997245,
+      "grad_norm": 7.197648048400879,
+      "learning_rate": 1.9104258443465493e-05,
+      "loss": 3.2771,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8639118457300277,
+      "grad_norm": 6.953661918640137,
+      "learning_rate": 1.9043073910915322e-05,
+      "loss": 3.1917,
+      "step": 1300
+    },
+    {
+      "epoch": 2.9190082644628097,
+      "grad_norm": 11.918855667114258,
+      "learning_rate": 1.898188937836515e-05,
+      "loss": 3.2673,
+      "step": 1325
+    },
+    {
+      "epoch": 2.974104683195592,
+      "grad_norm": 9.502423286437988,
+      "learning_rate": 1.892070484581498e-05,
+      "loss": 3.2804,
+      "step": 1350
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.0012172854534388314,
+      "eval_loss": 0.8076587319374084,
+      "eval_macro_f1": 0.3553418429761633,
+      "eval_micro_f1": 0.38251753205355243,
+      "eval_runtime": 0.7432,
+      "eval_samples_per_second": 2210.799,
+      "eval_steps_per_second": 138.595,
+      "eval_weighted_f1": 0.4286911321436805,
+      "step": 1362
+    },
+    {
+      "epoch": 3.028650137741047,
+      "grad_norm": 14.517410278320312,
+      "learning_rate": 1.8859520313264808e-05,
+      "loss": 3.0605,
+      "step": 1375
+    },
+    {
+      "epoch": 3.0837465564738293,
+      "grad_norm": 6.636230945587158,
+      "learning_rate": 1.8798335780714636e-05,
+      "loss": 2.9334,
+      "step": 1400
+    },
+    {
+      "epoch": 3.1388429752066114,
+      "grad_norm": 7.712543964385986,
+      "learning_rate": 1.8737151248164465e-05,
+      "loss": 3.0362,
+      "step": 1425
+    },
+    {
+      "epoch": 3.193939393939394,
+      "grad_norm": 6.594825267791748,
+      "learning_rate": 1.8675966715614294e-05,
+      "loss": 2.9552,
+      "step": 1450
+    },
+    {
+      "epoch": 3.2490358126721763,
+      "grad_norm": 10.446346282958984,
+      "learning_rate": 1.8614782183064122e-05,
+      "loss": 3.0088,
+      "step": 1475
+    },
+    {
+      "epoch": 3.304132231404959,
+      "grad_norm": 11.816834449768066,
+      "learning_rate": 1.855359765051395e-05,
+      "loss": 2.9343,
+      "step": 1500
+    },
+    {
+      "epoch": 3.359228650137741,
+      "grad_norm": 11.104568481445312,
+      "learning_rate": 1.849241311796378e-05,
+      "loss": 3.0041,
+      "step": 1525
+    },
+    {
+      "epoch": 3.4143250688705233,
+      "grad_norm": 18.386274337768555,
+      "learning_rate": 1.843122858541361e-05,
+      "loss": 2.9729,
+      "step": 1550
+    },
+    {
+      "epoch": 3.469421487603306,
+      "grad_norm": 8.527523040771484,
+      "learning_rate": 1.8370044052863437e-05,
+      "loss": 2.8784,
+      "step": 1575
+    },
+    {
+      "epoch": 3.5245179063360883,
+      "grad_norm": 6.6398024559021,
+      "learning_rate": 1.8308859520313266e-05,
+      "loss": 2.9162,
+      "step": 1600
+    },
+    {
+      "epoch": 3.579614325068871,
+      "grad_norm": 8.599790573120117,
+      "learning_rate": 1.8247674987763094e-05,
+      "loss": 3.002,
+      "step": 1625
+    },
+    {
+      "epoch": 3.634710743801653,
+      "grad_norm": 13.401683807373047,
+      "learning_rate": 1.8186490455212923e-05,
+      "loss": 2.8535,
+      "step": 1650
+    },
+    {
+      "epoch": 3.6898071625344353,
+      "grad_norm": 10.858172416687012,
+      "learning_rate": 1.812530592266275e-05,
+      "loss": 2.9071,
+      "step": 1675
+    },
+    {
+      "epoch": 3.744903581267218,
+      "grad_norm": 14.278366088867188,
+      "learning_rate": 1.806412139011258e-05,
+      "loss": 2.8518,
+      "step": 1700
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 10.848072052001953,
+      "learning_rate": 1.800293685756241e-05,
+      "loss": 2.8515,
+      "step": 1725
+    },
+    {
+      "epoch": 3.8550964187327823,
+      "grad_norm": 10.81524658203125,
+      "learning_rate": 1.7941752325012238e-05,
+      "loss": 2.9081,
+      "step": 1750
+    },
+    {
+      "epoch": 3.910192837465565,
+      "grad_norm": 8.548844337463379,
+      "learning_rate": 1.7880567792462066e-05,
+      "loss": 2.8372,
+      "step": 1775
+    },
+    {
+      "epoch": 3.9652892561983473,
+      "grad_norm": 13.30102252960205,
+      "learning_rate": 1.7819383259911895e-05,
+      "loss": 2.7932,
+      "step": 1800
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.002434570906877663,
+      "eval_loss": 0.756788969039917,
+      "eval_macro_f1": 0.39802980850917896,
+      "eval_micro_f1": 0.42926200451176283,
+      "eval_runtime": 0.8119,
+      "eval_samples_per_second": 2023.73,
+      "eval_steps_per_second": 126.868,
+      "eval_weighted_f1": 0.46001421514565244,
+      "step": 1816
+    },
+    {
+      "epoch": 4.019834710743802,
+      "grad_norm": 16.3684024810791,
+      "learning_rate": 1.7758198727361724e-05,
+      "loss": 2.7434,
+      "step": 1825
+    },
+    {
+      "epoch": 4.074931129476584,
+      "grad_norm": 9.8795804977417,
+      "learning_rate": 1.7697014194811552e-05,
+      "loss": 2.6399,
+      "step": 1850
+    },
+    {
+      "epoch": 4.130027548209366,
+      "grad_norm": 9.345877647399902,
+      "learning_rate": 1.763582966226138e-05,
+      "loss": 2.678,
+      "step": 1875
+    },
+    {
+      "epoch": 4.1851239669421485,
+      "grad_norm": 9.299138069152832,
+      "learning_rate": 1.757464512971121e-05,
+      "loss": 2.6086,
+      "step": 1900
+    },
+    {
+      "epoch": 4.240220385674931,
+      "grad_norm": 5.941190719604492,
+      "learning_rate": 1.7513460597161038e-05,
+      "loss": 2.6337,
+      "step": 1925
+    },
+    {
+      "epoch": 4.2953168044077135,
+      "grad_norm": 8.25993537902832,
+      "learning_rate": 1.7452276064610867e-05,
+      "loss": 2.6188,
+      "step": 1950
+    },
+    {
+      "epoch": 4.350413223140496,
+      "grad_norm": 12.652754783630371,
+      "learning_rate": 1.7391091532060695e-05,
+      "loss": 2.5735,
+      "step": 1975
+    },
+    {
+      "epoch": 4.405509641873278,
+      "grad_norm": 7.743082523345947,
+      "learning_rate": 1.7329906999510527e-05,
+      "loss": 2.6304,
+      "step": 2000
+    },
+    {
+      "epoch": 4.460606060606061,
+      "grad_norm": 17.932268142700195,
+      "learning_rate": 1.7268722466960356e-05,
+      "loss": 2.5889,
+      "step": 2025
+    },
+    {
+      "epoch": 4.515702479338843,
+      "grad_norm": 5.641079425811768,
+      "learning_rate": 1.7207537934410185e-05,
+      "loss": 2.6445,
+      "step": 2050
+    },
+    {
+      "epoch": 4.570798898071625,
+      "grad_norm": 11.554326057434082,
+      "learning_rate": 1.7146353401860013e-05,
+      "loss": 2.5493,
+      "step": 2075
+    },
+    {
+      "epoch": 4.6258953168044075,
+      "grad_norm": 11.849327087402344,
+      "learning_rate": 1.7085168869309842e-05,
+      "loss": 2.6203,
+      "step": 2100
+    },
+    {
+      "epoch": 4.68099173553719,
+      "grad_norm": 8.938584327697754,
+      "learning_rate": 1.702398433675967e-05,
+      "loss": 2.5594,
+      "step": 2125
+    },
+    {
+      "epoch": 4.736088154269972,
+      "grad_norm": 8.740391731262207,
+      "learning_rate": 1.69627998042095e-05,
+      "loss": 2.5733,
+      "step": 2150
+    },
+    {
+      "epoch": 4.791184573002755,
+      "grad_norm": 9.344244003295898,
+      "learning_rate": 1.6901615271659325e-05,
+      "loss": 2.5557,
+      "step": 2175
+    },
+    {
+      "epoch": 4.846280991735537,
+      "grad_norm": 16.789382934570312,
+      "learning_rate": 1.6840430739109153e-05,
+      "loss": 2.5097,
+      "step": 2200
+    },
+    {
+      "epoch": 4.90137741046832,
+      "grad_norm": 10.325007438659668,
+      "learning_rate": 1.6779246206558982e-05,
+      "loss": 2.4844,
+      "step": 2225
+    },
+    {
+      "epoch": 4.9564738292011015,
+      "grad_norm": 12.274792671203613,
+      "learning_rate": 1.671806167400881e-05,
+      "loss": 2.4952,
+      "step": 2250
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.02556299452221546,
+      "eval_loss": 0.7364040613174438,
+      "eval_macro_f1": 0.4238555784117118,
+      "eval_micro_f1": 0.4553359683794466,
+      "eval_runtime": 0.8163,
+      "eval_samples_per_second": 2012.859,
+      "eval_steps_per_second": 126.187,
+      "eval_weighted_f1": 0.4764858778657165,
+      "step": 2270
+    },
+    {
+      "epoch": 5.011019283746556,
+      "grad_norm": 9.081082344055176,
+      "learning_rate": 1.665687714145864e-05,
+      "loss": 2.4144,
+      "step": 2275
+    },
+    {
+      "epoch": 5.066115702479339,
+      "grad_norm": 7.6597723960876465,
+      "learning_rate": 1.6595692608908468e-05,
+      "loss": 2.2944,
+      "step": 2300
+    },
+    {
+      "epoch": 5.121212121212121,
+      "grad_norm": 60.42306137084961,
+      "learning_rate": 1.6534508076358297e-05,
+      "loss": 2.2364,
+      "step": 2325
+    },
+    {
+      "epoch": 5.176308539944904,
+      "grad_norm": 10.49393081665039,
+      "learning_rate": 1.6473323543808125e-05,
+      "loss": 2.2605,
+      "step": 2350
+    },
+    {
+      "epoch": 5.231404958677686,
+      "grad_norm": 6.1078782081604,
+      "learning_rate": 1.6412139011257954e-05,
+      "loss": 2.2864,
+      "step": 2375
+    },
+    {
+      "epoch": 5.2865013774104685,
+      "grad_norm": 9.789051055908203,
+      "learning_rate": 1.6350954478707782e-05,
+      "loss": 2.3701,
+      "step": 2400
+    },
+    {
+      "epoch": 5.341597796143251,
+      "grad_norm": 12.479533195495605,
+      "learning_rate": 1.628976994615761e-05,
+      "loss": 2.3592,
+      "step": 2425
+    },
+    {
+      "epoch": 5.3966942148760335,
+      "grad_norm": 7.823705673217773,
+      "learning_rate": 1.622858541360744e-05,
+      "loss": 2.3578,
+      "step": 2450
+    },
+    {
+      "epoch": 5.451790633608815,
+      "grad_norm": 6.067243576049805,
+      "learning_rate": 1.616740088105727e-05,
+      "loss": 2.2772,
+      "step": 2475
+    },
+    {
+      "epoch": 5.506887052341598,
+      "grad_norm": 11.943389892578125,
+      "learning_rate": 1.6106216348507097e-05,
+      "loss": 2.2705,
+      "step": 2500
+    },
+    {
+      "epoch": 5.56198347107438,
+      "grad_norm": 10.336061477661133,
+      "learning_rate": 1.604503181595693e-05,
+      "loss": 2.2845,
+      "step": 2525
+    },
+    {
+      "epoch": 5.6170798898071626,
+      "grad_norm": 12.933616638183594,
+      "learning_rate": 1.5983847283406758e-05,
+      "loss": 2.2261,
+      "step": 2550
+    },
+    {
+      "epoch": 5.672176308539945,
+      "grad_norm": 10.702818870544434,
+      "learning_rate": 1.5922662750856586e-05,
+      "loss": 2.3344,
+      "step": 2575
+    },
+    {
+      "epoch": 5.7272727272727275,
+      "grad_norm": 10.82820987701416,
+      "learning_rate": 1.5861478218306415e-05,
+      "loss": 2.3006,
+      "step": 2600
+    },
+    {
+      "epoch": 5.78236914600551,
+      "grad_norm": 12.483839988708496,
+      "learning_rate": 1.5800293685756244e-05,
+      "loss": 2.25,
+      "step": 2625
+    },
+    {
+      "epoch": 5.837465564738292,
+      "grad_norm": 12.435521125793457,
+      "learning_rate": 1.5739109153206072e-05,
+      "loss": 2.2539,
+      "step": 2650
+    },
+    {
+      "epoch": 5.892561983471074,
+      "grad_norm": 10.098462104797363,
+      "learning_rate": 1.56779246206559e-05,
+      "loss": 2.1928,
+      "step": 2675
+    },
+    {
+      "epoch": 5.9476584022038566,
+      "grad_norm": 14.790334701538086,
+      "learning_rate": 1.561674008810573e-05,
+      "loss": 2.1906,
+      "step": 2700
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.02982349360925137,
+      "eval_loss": 0.7161994576454163,
+      "eval_macro_f1": 0.4152780695248448,
+      "eval_micro_f1": 0.44851338390443807,
+      "eval_runtime": 0.8477,
+      "eval_samples_per_second": 1938.149,
+      "eval_steps_per_second": 121.503,
+      "eval_weighted_f1": 0.473624177164651,
+      "step": 2724
+    },
+    {
+      "epoch": 6.002203856749311,
+      "grad_norm": 11.331801414489746,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 2.1592,
+      "step": 2725
+    },
+    {
+      "epoch": 6.057300275482094,
+      "grad_norm": 7.465226650238037,
+      "learning_rate": 1.5494371023005387e-05,
+      "loss": 2.1362,
+      "step": 2750
+    },
+    {
+      "epoch": 6.112396694214876,
+      "grad_norm": 7.91953706741333,
+      "learning_rate": 1.5433186490455216e-05,
+      "loss": 2.133,
+      "step": 2775
+    },
+    {
+      "epoch": 6.167493112947659,
+      "grad_norm": 43.329559326171875,
+      "learning_rate": 1.5372001957905044e-05,
+      "loss": 2.0267,
+      "step": 2800
+    },
+    {
+      "epoch": 6.222589531680441,
+      "grad_norm": 9.198493003845215,
+      "learning_rate": 1.5310817425354873e-05,
+      "loss": 1.9903,
+      "step": 2825
+    },
+    {
+      "epoch": 6.277685950413223,
+      "grad_norm": 14.81072998046875,
+      "learning_rate": 1.52496328928047e-05,
+      "loss": 2.0183,
+      "step": 2850
+    },
+    {
+      "epoch": 6.332782369146005,
+      "grad_norm": 7.481613636016846,
+      "learning_rate": 1.5188448360254529e-05,
+      "loss": 2.0888,
+      "step": 2875
+    },
+    {
+      "epoch": 6.387878787878788,
+      "grad_norm": 28.831653594970703,
+      "learning_rate": 1.5127263827704357e-05,
+      "loss": 2.0233,
+      "step": 2900
+    },
+    {
+      "epoch": 6.44297520661157,
+      "grad_norm": 10.08276653289795,
+      "learning_rate": 1.5066079295154186e-05,
+      "loss": 2.0979,
+      "step": 2925
+    },
+    {
+      "epoch": 6.498071625344353,
+      "grad_norm": 8.211263656616211,
+      "learning_rate": 1.5004894762604014e-05,
+      "loss": 2.0263,
+      "step": 2950
+    },
+    {
+      "epoch": 6.553168044077135,
+      "grad_norm": 13.02656364440918,
+      "learning_rate": 1.4943710230053843e-05,
+      "loss": 1.9823,
+      "step": 2975
+    },
+    {
+      "epoch": 6.608264462809918,
+      "grad_norm": 10.631696701049805,
+      "learning_rate": 1.4882525697503672e-05,
+      "loss": 1.987,
+      "step": 3000
+    },
+    {
+      "epoch": 6.663360881542699,
+      "grad_norm": 8.777915954589844,
+      "learning_rate": 1.48213411649535e-05,
+      "loss": 2.0683,
+      "step": 3025
+    },
+    {
+      "epoch": 6.718457300275482,
+      "grad_norm": 12.436280250549316,
+      "learning_rate": 1.476015663240333e-05,
+      "loss": 2.0,
+      "step": 3050
+    },
+    {
+      "epoch": 6.773553719008264,
+      "grad_norm": 11.935210227966309,
+      "learning_rate": 1.469897209985316e-05,
+      "loss": 1.9862,
+      "step": 3075
+    },
+    {
+      "epoch": 6.828650137741047,
+      "grad_norm": 11.0009765625,
+      "learning_rate": 1.4637787567302988e-05,
+      "loss": 2.0002,
+      "step": 3100
+    },
+    {
+      "epoch": 6.883746556473829,
+      "grad_norm": 8.148244857788086,
+      "learning_rate": 1.4576603034752817e-05,
+      "loss": 1.9712,
+      "step": 3125
+    },
+    {
+      "epoch": 6.938842975206612,
+      "grad_norm": 15.212584495544434,
+      "learning_rate": 1.4515418502202645e-05,
+      "loss": 1.9824,
+      "step": 3150
+    },
+    {
+      "epoch": 6.993939393939394,
+      "grad_norm": 6.5839457511901855,
+      "learning_rate": 1.4454233969652474e-05,
+      "loss": 1.9517,
+      "step": 3175
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.0426049908703591,
+      "eval_loss": 0.7072303891181946,
+      "eval_macro_f1": 0.4624544764817252,
+      "eval_micro_f1": 0.4939302709527047,
+      "eval_runtime": 0.7605,
+      "eval_samples_per_second": 2160.426,
+      "eval_steps_per_second": 135.438,
+      "eval_weighted_f1": 0.5083354781806794,
+      "step": 3178
+    },
+    {
+      "epoch": 7.048484848484849,
+      "grad_norm": 9.203914642333984,
+      "learning_rate": 1.4393049437102303e-05,
+      "loss": 1.8504,
+      "step": 3200
+    },
+    {
+      "epoch": 7.103581267217631,
+      "grad_norm": 10.75075912475586,
+      "learning_rate": 1.4331864904552131e-05,
+      "loss": 1.8917,
+      "step": 3225
+    },
+    {
+      "epoch": 7.158677685950413,
+      "grad_norm": 8.260217666625977,
+      "learning_rate": 1.427068037200196e-05,
+      "loss": 1.9492,
+      "step": 3250
+    },
+    {
+      "epoch": 7.213774104683195,
+      "grad_norm": 12.863761901855469,
+      "learning_rate": 1.4209495839451789e-05,
+      "loss": 1.8323,
+      "step": 3275
+    },
+    {
+      "epoch": 7.268870523415978,
+      "grad_norm": 7.896994113922119,
+      "learning_rate": 1.4148311306901617e-05,
+      "loss": 1.8989,
+      "step": 3300
+    },
+    {
+      "epoch": 7.32396694214876,
+      "grad_norm": 13.598098754882812,
+      "learning_rate": 1.4087126774351444e-05,
+      "loss": 1.8014,
+      "step": 3325
+    },
+    {
+      "epoch": 7.379063360881543,
+      "grad_norm": 10.687313079833984,
+      "learning_rate": 1.4025942241801273e-05,
+      "loss": 1.7836,
+      "step": 3350
+    },
+    {
+      "epoch": 7.434159779614325,
+      "grad_norm": 9.242302894592285,
+      "learning_rate": 1.3964757709251102e-05,
+      "loss": 1.8942,
+      "step": 3375
+    },
+    {
+      "epoch": 7.489256198347108,
+      "grad_norm": 8.878076553344727,
+      "learning_rate": 1.390357317670093e-05,
+      "loss": 1.7399,
+      "step": 3400
+    },
+    {
+      "epoch": 7.544352617079889,
+      "grad_norm": 14.274787902832031,
+      "learning_rate": 1.3842388644150759e-05,
+      "loss": 1.7946,
+      "step": 3425
+    },
+    {
+      "epoch": 7.599449035812672,
+      "grad_norm": 7.229008197784424,
+      "learning_rate": 1.3781204111600588e-05,
+      "loss": 1.8592,
+      "step": 3450
+    },
+    {
+      "epoch": 7.654545454545454,
+      "grad_norm": 7.422128200531006,
+      "learning_rate": 1.3720019579050416e-05,
+      "loss": 1.8041,
+      "step": 3475
+    },
+    {
+      "epoch": 7.709641873278237,
+      "grad_norm": 5.205471038818359,
+      "learning_rate": 1.3658835046500245e-05,
+      "loss": 1.8001,
+      "step": 3500
+    },
+    {
+      "epoch": 7.764738292011019,
+      "grad_norm": 14.030226707458496,
+      "learning_rate": 1.3597650513950073e-05,
+      "loss": 1.7651,
+      "step": 3525
+    },
+    {
+      "epoch": 7.819834710743802,
+      "grad_norm": 9.975546836853027,
+      "learning_rate": 1.3536465981399902e-05,
+      "loss": 1.8567,
+      "step": 3550
+    },
+    {
+      "epoch": 7.874931129476584,
+      "grad_norm": 7.343936920166016,
+      "learning_rate": 1.3475281448849732e-05,
+      "loss": 1.7966,
+      "step": 3575
+    },
+    {
+      "epoch": 7.930027548209367,
+      "grad_norm": 9.779473304748535,
+      "learning_rate": 1.3414096916299561e-05,
+      "loss": 1.7319,
+      "step": 3600
+    },
+    {
+      "epoch": 7.985123966942149,
+      "grad_norm": 9.262539863586426,
+      "learning_rate": 1.335291238374939e-05,
+      "loss": 1.7536,
+      "step": 3625
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.06086427267194157,
+      "eval_loss": 0.7092871069908142,
+      "eval_macro_f1": 0.4828944671231688,
+      "eval_micro_f1": 0.5134519867549668,
+      "eval_runtime": 0.8505,
+      "eval_samples_per_second": 1931.851,
+      "eval_steps_per_second": 121.108,
+      "eval_weighted_f1": 0.5272358823424601,
+      "step": 3632
+    },
+    {
+      "epoch": 8.039669421487604,
+      "grad_norm": 9.195419311523438,
+      "learning_rate": 1.3291727851199218e-05,
+      "loss": 1.7551,
+      "step": 3650
+    },
+    {
+      "epoch": 8.094765840220386,
+      "grad_norm": 8.99988842010498,
+      "learning_rate": 1.3230543318649047e-05,
+      "loss": 1.6994,
+      "step": 3675
+    },
+    {
+      "epoch": 8.149862258953169,
+      "grad_norm": 13.685859680175781,
+      "learning_rate": 1.3169358786098876e-05,
+      "loss": 1.7063,
+      "step": 3700
+    },
+    {
+      "epoch": 8.20495867768595,
+      "grad_norm": 7.462896347045898,
+      "learning_rate": 1.3108174253548704e-05,
+      "loss": 1.6769,
+      "step": 3725
+    },
+    {
+      "epoch": 8.260055096418732,
+      "grad_norm": 9.458789825439453,
+      "learning_rate": 1.3046989720998533e-05,
+      "loss": 1.6593,
+      "step": 3750
+    },
+    {
+      "epoch": 8.315151515151515,
+      "grad_norm": 9.853849411010742,
+      "learning_rate": 1.2985805188448362e-05,
+      "loss": 1.6467,
+      "step": 3775
+    },
+    {
+      "epoch": 8.370247933884297,
+      "grad_norm": 7.881902694702148,
+      "learning_rate": 1.292462065589819e-05,
+      "loss": 1.6405,
+      "step": 3800
+    },
+    {
+      "epoch": 8.42534435261708,
+      "grad_norm": 13.257023811340332,
+      "learning_rate": 1.2863436123348019e-05,
+      "loss": 1.6338,
+      "step": 3825
+    },
+    {
+      "epoch": 8.480440771349862,
+      "grad_norm": 8.842818260192871,
+      "learning_rate": 1.2802251590797848e-05,
+      "loss": 1.6426,
+      "step": 3850
+    },
+    {
+      "epoch": 8.535537190082644,
+      "grad_norm": 9.2915620803833,
+      "learning_rate": 1.2741067058247676e-05,
+      "loss": 1.6603,
+      "step": 3875
+    },
+    {
+      "epoch": 8.590633608815427,
+      "grad_norm": 6.625736713409424,
+      "learning_rate": 1.2679882525697505e-05,
+      "loss": 1.6491,
+      "step": 3900
+    },
+    {
+      "epoch": 8.64573002754821,
+      "grad_norm": 11.400477409362793,
+      "learning_rate": 1.2618697993147334e-05,
+      "loss": 1.6193,
+      "step": 3925
+    },
+    {
+      "epoch": 8.700826446280992,
+      "grad_norm": 8.049001693725586,
+      "learning_rate": 1.2557513460597162e-05,
+      "loss": 1.6068,
+      "step": 3950
+    },
+    {
+      "epoch": 8.755922865013774,
+      "grad_norm": 8.45368766784668,
+      "learning_rate": 1.2496328928046991e-05,
+      "loss": 1.629,
+      "step": 3975
+    },
+    {
+      "epoch": 8.811019283746557,
+      "grad_norm": 11.155718803405762,
+      "learning_rate": 1.243514439549682e-05,
+      "loss": 1.6357,
+      "step": 4000
+    },
+    {
+      "epoch": 8.86611570247934,
+      "grad_norm": 9.854330062866211,
+      "learning_rate": 1.2373959862946646e-05,
+      "loss": 1.6631,
+      "step": 4025
+    },
+    {
+      "epoch": 8.921212121212122,
+      "grad_norm": 11.173809051513672,
+      "learning_rate": 1.2312775330396475e-05,
+      "loss": 1.6442,
+      "step": 4050
+    },
+    {
+      "epoch": 8.976308539944904,
+      "grad_norm": 6.974915027618408,
+      "learning_rate": 1.2251590797846304e-05,
+      "loss": 1.6068,
+      "step": 4075
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.05660377358490566,
+      "eval_loss": 0.7177210450172424,
+      "eval_macro_f1": 0.4876323052992565,
+      "eval_micro_f1": 0.5183353934899052,
+      "eval_runtime": 0.7616,
+      "eval_samples_per_second": 2157.235,
+      "eval_steps_per_second": 135.237,
+      "eval_weighted_f1": 0.5293421961526377,
+      "step": 4086
+    },
+    {
+      "epoch": 9.030853994490359,
+      "grad_norm": 8.590648651123047,
+      "learning_rate": 1.2190406265296136e-05,
+      "loss": 1.604,
+      "step": 4100
+    },
+    {
+      "epoch": 9.08595041322314,
+      "grad_norm": 11.171589851379395,
+      "learning_rate": 1.2129221732745964e-05,
+      "loss": 1.5905,
+      "step": 4125
+    },
+    {
+      "epoch": 9.141046831955922,
+      "grad_norm": 17.820585250854492,
+      "learning_rate": 1.2068037200195793e-05,
+      "loss": 1.4744,
+      "step": 4150
+    },
+    {
+      "epoch": 9.196143250688705,
+      "grad_norm": 8.431110382080078,
+      "learning_rate": 1.2006852667645622e-05,
+      "loss": 1.5549,
+      "step": 4175
+    },
+    {
+      "epoch": 9.251239669421487,
+      "grad_norm": 7.0245680809021,
+      "learning_rate": 1.1945668135095449e-05,
+      "loss": 1.5031,
+      "step": 4200
+    },
+    {
+      "epoch": 9.30633608815427,
+      "grad_norm": 11.74380111694336,
+      "learning_rate": 1.1884483602545277e-05,
+      "loss": 1.456,
+      "step": 4225
+    },
+    {
+      "epoch": 9.361432506887052,
+      "grad_norm": 12.57530403137207,
+      "learning_rate": 1.1823299069995106e-05,
+      "loss": 1.5037,
+      "step": 4250
+    },
+    {
+      "epoch": 9.416528925619835,
+      "grad_norm": 7.103407859802246,
+      "learning_rate": 1.1762114537444935e-05,
+      "loss": 1.5581,
+      "step": 4275
+    },
+    {
+      "epoch": 9.471625344352617,
+      "grad_norm": 14.686859130859375,
+      "learning_rate": 1.1700930004894763e-05,
+      "loss": 1.5657,
+      "step": 4300
+    },
+    {
+      "epoch": 9.5267217630854,
+      "grad_norm": 9.481619834899902,
+      "learning_rate": 1.1639745472344592e-05,
+      "loss": 1.5319,
+      "step": 4325
+    },
+    {
+      "epoch": 9.581818181818182,
+      "grad_norm": 9.552848815917969,
+      "learning_rate": 1.157856093979442e-05,
+      "loss": 1.5062,
+      "step": 4350
+    },
+    {
+      "epoch": 9.636914600550964,
+      "grad_norm": 10.749216079711914,
+      "learning_rate": 1.151737640724425e-05,
+      "loss": 1.4887,
+      "step": 4375
+    },
+    {
+      "epoch": 9.692011019283747,
+      "grad_norm": 12.477849960327148,
+      "learning_rate": 1.1456191874694078e-05,
+      "loss": 1.5119,
+      "step": 4400
+    },
+    {
+      "epoch": 9.74710743801653,
+      "grad_norm": 9.663217544555664,
+      "learning_rate": 1.1395007342143907e-05,
+      "loss": 1.4636,
+      "step": 4425
+    },
+    {
+      "epoch": 9.802203856749312,
+      "grad_norm": 5.807657241821289,
+      "learning_rate": 1.1333822809593735e-05,
+      "loss": 1.537,
+      "step": 4450
+    },
+    {
+      "epoch": 9.857300275482094,
+      "grad_norm": 14.375182151794434,
+      "learning_rate": 1.1272638277043564e-05,
+      "loss": 1.4866,
+      "step": 4475
+    },
+    {
+      "epoch": 9.912396694214877,
+      "grad_norm": 6.251608848571777,
+      "learning_rate": 1.1211453744493393e-05,
+      "loss": 1.5255,
+      "step": 4500
+    },
+    {
+      "epoch": 9.967493112947658,
+      "grad_norm": 15.098796844482422,
+      "learning_rate": 1.1150269211943221e-05,
+      "loss": 1.434,
+      "step": 4525
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.0620815581253804,
+      "eval_loss": 0.7124394774436951,
+      "eval_macro_f1": 0.49507577431721617,
+      "eval_micro_f1": 0.5261606484893147,
+      "eval_runtime": 0.8034,
+      "eval_samples_per_second": 2044.959,
+      "eval_steps_per_second": 128.199,
+      "eval_weighted_f1": 0.5337704983302356,
+      "step": 4540
+    },
+    {
+      "epoch": 10.022038567493112,
+      "grad_norm": 7.172276496887207,
+      "learning_rate": 1.108908467939305e-05,
+      "loss": 1.4615,
+      "step": 4550
+    },
+    {
+      "epoch": 10.077134986225895,
+      "grad_norm": 6.198989391326904,
+      "learning_rate": 1.1027900146842878e-05,
+      "loss": 1.4007,
+      "step": 4575
+    },
+    {
+      "epoch": 10.132231404958677,
+      "grad_norm": 6.623414993286133,
+      "learning_rate": 1.0966715614292707e-05,
+      "loss": 1.4138,
+      "step": 4600
+    },
+    {
+      "epoch": 10.18732782369146,
+      "grad_norm": 8.518909454345703,
+      "learning_rate": 1.0905531081742536e-05,
+      "loss": 1.37,
+      "step": 4625
+    },
+    {
+      "epoch": 10.242424242424242,
+      "grad_norm": 9.576210021972656,
+      "learning_rate": 1.0844346549192366e-05,
+      "loss": 1.3478,
+      "step": 4650
+    },
+    {
+      "epoch": 10.297520661157025,
+      "grad_norm": 20.642356872558594,
+      "learning_rate": 1.0783162016642195e-05,
+      "loss": 1.4265,
+      "step": 4675
+    },
+    {
+      "epoch": 10.352617079889807,
+      "grad_norm": 11.112329483032227,
+      "learning_rate": 1.0721977484092023e-05,
+      "loss": 1.4461,
+      "step": 4700
+    },
+    {
+      "epoch": 10.40771349862259,
+      "grad_norm": 10.513445854187012,
+      "learning_rate": 1.0660792951541852e-05,
+      "loss": 1.3629,
+      "step": 4725
+    },
+    {
+      "epoch": 10.462809917355372,
+      "grad_norm": 8.798458099365234,
+      "learning_rate": 1.059960841899168e-05,
+      "loss": 1.3599,
+      "step": 4750
+    },
+    {
+      "epoch": 10.517906336088155,
+      "grad_norm": 9.605145454406738,
+      "learning_rate": 1.053842388644151e-05,
+      "loss": 1.4262,
+      "step": 4775
+    },
+    {
+      "epoch": 10.573002754820937,
+      "grad_norm": 12.231898307800293,
+      "learning_rate": 1.0477239353891338e-05,
+      "loss": 1.3747,
+      "step": 4800
+    },
+    {
+      "epoch": 10.62809917355372,
+      "grad_norm": 9.49361515045166,
+      "learning_rate": 1.0416054821341167e-05,
+      "loss": 1.3357,
+      "step": 4825
+    },
+    {
+      "epoch": 10.683195592286502,
+      "grad_norm": 14.282561302185059,
+      "learning_rate": 1.0354870288790995e-05,
+      "loss": 1.3724,
+      "step": 4850
+    },
+    {
+      "epoch": 10.738292011019285,
+      "grad_norm": 33.559486389160156,
+      "learning_rate": 1.0293685756240824e-05,
+      "loss": 1.4198,
+      "step": 4875
+    },
+    {
+      "epoch": 10.793388429752067,
+      "grad_norm": 10.780166625976562,
+      "learning_rate": 1.0232501223690653e-05,
+      "loss": 1.3685,
+      "step": 4900
+    },
+    {
+      "epoch": 10.848484848484848,
+      "grad_norm": 6.057034492492676,
+      "learning_rate": 1.017131669114048e-05,
+      "loss": 1.3914,
+      "step": 4925
+    },
+    {
+      "epoch": 10.90358126721763,
+      "grad_norm": 9.05207347869873,
+      "learning_rate": 1.0110132158590308e-05,
+      "loss": 1.3619,
+      "step": 4950
+    },
+    {
+      "epoch": 10.958677685950413,
+      "grad_norm": 6.542205810546875,
+      "learning_rate": 1.0048947626040137e-05,
+      "loss": 1.3213,
+      "step": 4975
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.08581862446743761,
+      "eval_loss": 0.7311891317367554,
+      "eval_macro_f1": 0.50394695762473,
+      "eval_micro_f1": 0.5346751257930431,
+      "eval_runtime": 0.7471,
+      "eval_samples_per_second": 2199.246,
+      "eval_steps_per_second": 137.871,
+      "eval_weighted_f1": 0.5450425663519038,
+      "step": 4994
+    },
+    {
+      "epoch": 11.013223140495867,
+      "grad_norm": 9.045429229736328,
+      "learning_rate": 9.987763093489966e-06,
+      "loss": 1.3719,
+      "step": 5000
+    },
+    {
+      "epoch": 11.06831955922865,
+      "grad_norm": 16.91636848449707,
+      "learning_rate": 9.926578560939796e-06,
+      "loss": 1.304,
+      "step": 5025
+    },
+    {
+      "epoch": 11.123415977961432,
+      "grad_norm": 12.180585861206055,
+      "learning_rate": 9.865394028389625e-06,
+      "loss": 1.282,
+      "step": 5050
+    },
+    {
+      "epoch": 11.178512396694215,
+      "grad_norm": 6.918673515319824,
+      "learning_rate": 9.804209495839453e-06,
+      "loss": 1.3003,
+      "step": 5075
+    },
+    {
+      "epoch": 11.233608815426997,
+      "grad_norm": 9.964341163635254,
+      "learning_rate": 9.743024963289282e-06,
+      "loss": 1.2704,
+      "step": 5100
+    },
+    {
+      "epoch": 11.28870523415978,
+      "grad_norm": 13.528480529785156,
+      "learning_rate": 9.68184043073911e-06,
+      "loss": 1.3075,
+      "step": 5125
+    },
+    {
+      "epoch": 11.343801652892562,
+      "grad_norm": 9.736034393310547,
+      "learning_rate": 9.620655898188939e-06,
+      "loss": 1.2758,
+      "step": 5150
+    },
+    {
+      "epoch": 11.398898071625345,
+      "grad_norm": 9.628288269042969,
+      "learning_rate": 9.559471365638768e-06,
+      "loss": 1.2939,
+      "step": 5175
+    },
+    {
+      "epoch": 11.453994490358127,
+      "grad_norm": 9.606924057006836,
+      "learning_rate": 9.498286833088596e-06,
+      "loss": 1.291,
+      "step": 5200
+    },
+    {
+      "epoch": 11.50909090909091,
+      "grad_norm": 9.936850547790527,
+      "learning_rate": 9.437102300538425e-06,
+      "loss": 1.2681,
+      "step": 5225
+    },
+    {
+      "epoch": 11.564187327823692,
+      "grad_norm": 7.868658542633057,
+      "learning_rate": 9.375917767988254e-06,
+      "loss": 1.2479,
+      "step": 5250
+    },
+    {
+      "epoch": 11.619283746556473,
+      "grad_norm": 8.432409286499023,
+      "learning_rate": 9.31473323543808e-06,
+      "loss": 1.2672,
+      "step": 5275
+    },
+    {
+      "epoch": 11.674380165289255,
+      "grad_norm": 8.701844215393066,
+      "learning_rate": 9.253548702887911e-06,
+      "loss": 1.3027,
+      "step": 5300
+    },
+    {
+      "epoch": 11.729476584022038,
+      "grad_norm": 16.532873153686523,
+      "learning_rate": 9.19236417033774e-06,
+      "loss": 1.244,
+      "step": 5325
+    },
+    {
+      "epoch": 11.78457300275482,
+      "grad_norm": 8.731383323669434,
+      "learning_rate": 9.131179637787568e-06,
+      "loss": 1.2405,
+      "step": 5350
+    },
+    {
+      "epoch": 11.839669421487603,
+      "grad_norm": 10.397244453430176,
+      "learning_rate": 9.069995105237397e-06,
+      "loss": 1.2831,
+      "step": 5375
+    },
+    {
+      "epoch": 11.894765840220385,
+      "grad_norm": 10.608807563781738,
+      "learning_rate": 9.008810572687226e-06,
+      "loss": 1.1933,
+      "step": 5400
+    },
+    {
+      "epoch": 11.949862258953168,
+      "grad_norm": 7.912414073944092,
+      "learning_rate": 8.947626040137054e-06,
+      "loss": 1.2603,
+      "step": 5425
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.08642726719415703,
+      "eval_loss": 0.7480849027633667,
+      "eval_macro_f1": 0.5113890808616641,
+      "eval_micro_f1": 0.5431139718122295,
+      "eval_runtime": 0.774,
+      "eval_samples_per_second": 2122.643,
+      "eval_steps_per_second": 133.069,
+      "eval_weighted_f1": 0.5492742671911292,
+      "step": 5448
+    },
+    {
+      "epoch": 12.004407713498622,
+      "grad_norm": 9.934858322143555,
+      "learning_rate": 8.886441507586883e-06,
+      "loss": 1.2531,
+      "step": 5450
+    },
+    {
+      "epoch": 12.059504132231405,
+      "grad_norm": 6.843869209289551,
+      "learning_rate": 8.825256975036712e-06,
+      "loss": 1.1773,
+      "step": 5475
+    },
+    {
+      "epoch": 12.114600550964187,
+      "grad_norm": 7.1369853019714355,
+      "learning_rate": 8.76407244248654e-06,
+      "loss": 1.226,
+      "step": 5500
+    },
+    {
+      "epoch": 12.16969696969697,
+      "grad_norm": 8.611149787902832,
+      "learning_rate": 8.702887909936369e-06,
+      "loss": 1.1832,
+      "step": 5525
+    },
+    {
+      "epoch": 12.224793388429752,
+      "grad_norm": 12.588682174682617,
+      "learning_rate": 8.641703377386198e-06,
+      "loss": 1.2135,
+      "step": 5550
+    },
+    {
+      "epoch": 12.279889807162535,
+      "grad_norm": 16.866628646850586,
+      "learning_rate": 8.580518844836026e-06,
+      "loss": 1.1874,
+      "step": 5575
+    },
+    {
+      "epoch": 12.334986225895317,
+      "grad_norm": 21.434120178222656,
+      "learning_rate": 8.519334312285855e-06,
+      "loss": 1.2108,
+      "step": 5600
+    },
+    {
+      "epoch": 12.3900826446281,
+      "grad_norm": 14.522948265075684,
+      "learning_rate": 8.458149779735683e-06,
+      "loss": 1.1988,
+      "step": 5625
+    },
+    {
+      "epoch": 12.445179063360882,
+      "grad_norm": 13.422613143920898,
+      "learning_rate": 8.396965247185512e-06,
+      "loss": 1.2145,
+      "step": 5650
+    },
+    {
+      "epoch": 12.500275482093663,
+      "grad_norm": 9.499993324279785,
+      "learning_rate": 8.33578071463534e-06,
+      "loss": 1.1397,
+      "step": 5675
+    },
+    {
+      "epoch": 12.555371900826445,
+      "grad_norm": 10.744636535644531,
+      "learning_rate": 8.27459618208517e-06,
+      "loss": 1.1531,
+      "step": 5700
+    },
+    {
+      "epoch": 12.610468319559228,
+      "grad_norm": 9.612398147583008,
+      "learning_rate": 8.213411649534998e-06,
+      "loss": 1.1706,
+      "step": 5725
+    },
+    {
+      "epoch": 12.66556473829201,
+      "grad_norm": 12.579903602600098,
+      "learning_rate": 8.152227116984827e-06,
+      "loss": 1.1717,
+      "step": 5750
+    },
+    {
+      "epoch": 12.720661157024793,
+      "grad_norm": 9.093660354614258,
+      "learning_rate": 8.091042584434655e-06,
+      "loss": 1.1956,
+      "step": 5775
+    },
+    {
+      "epoch": 12.775757575757575,
+      "grad_norm": 18.263643264770508,
+      "learning_rate": 8.029858051884484e-06,
+      "loss": 1.2099,
+      "step": 5800
+    },
+    {
+      "epoch": 12.830853994490358,
+      "grad_norm": 17.743852615356445,
+      "learning_rate": 7.968673519334313e-06,
+      "loss": 1.1798,
+      "step": 5825
+    },
+    {
+      "epoch": 12.88595041322314,
+      "grad_norm": 7.139863967895508,
+      "learning_rate": 7.907488986784141e-06,
+      "loss": 1.2019,
+      "step": 5850
+    },
+    {
+      "epoch": 12.941046831955923,
+      "grad_norm": 15.34650707244873,
+      "learning_rate": 7.84630445423397e-06,
+      "loss": 1.2193,
+      "step": 5875
+    },
+    {
+      "epoch": 12.996143250688705,
+      "grad_norm": 10.276274681091309,
+      "learning_rate": 7.785119921683799e-06,
+      "loss": 1.1547,
+      "step": 5900
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.09494826536822885,
+      "eval_loss": 0.7594033479690552,
+      "eval_macro_f1": 0.5202326488837883,
+      "eval_micro_f1": 0.5480618418104414,
+      "eval_runtime": 0.7284,
+      "eval_samples_per_second": 2255.513,
+      "eval_steps_per_second": 141.399,
+      "eval_weighted_f1": 0.5550034950523685,
+      "step": 5902
+    },
+    {
+      "epoch": 13.05068870523416,
+      "grad_norm": 12.496541023254395,
+      "learning_rate": 7.723935389133627e-06,
+      "loss": 1.0906,
+      "step": 5925
+    },
+    {
+      "epoch": 13.105785123966943,
+      "grad_norm": 7.705901145935059,
+      "learning_rate": 7.662750856583456e-06,
+      "loss": 1.1254,
+      "step": 5950
+    },
+    {
+      "epoch": 13.160881542699725,
+      "grad_norm": 7.635675430297852,
+      "learning_rate": 7.601566324033285e-06,
+      "loss": 1.1406,
+      "step": 5975
+    },
+    {
+      "epoch": 13.215977961432507,
+      "grad_norm": 8.631380081176758,
+      "learning_rate": 7.540381791483113e-06,
+      "loss": 1.1202,
+      "step": 6000
+    },
+    {
+      "epoch": 13.27107438016529,
+      "grad_norm": 22.504575729370117,
+      "learning_rate": 7.479197258932942e-06,
+      "loss": 1.139,
+      "step": 6025
+    },
+    {
+      "epoch": 13.326170798898072,
+      "grad_norm": 9.432455062866211,
+      "learning_rate": 7.4180127263827705e-06,
+      "loss": 1.1305,
+      "step": 6050
+    },
+    {
+      "epoch": 13.381267217630853,
+      "grad_norm": 9.018024444580078,
+      "learning_rate": 7.3568281938326e-06,
+      "loss": 1.1186,
+      "step": 6075
+    },
+    {
+      "epoch": 13.436363636363636,
+      "grad_norm": 9.342816352844238,
+      "learning_rate": 7.295643661282429e-06,
+      "loss": 1.0854,
+      "step": 6100
+    },
+    {
+      "epoch": 13.491460055096418,
+      "grad_norm": 4.410456657409668,
+      "learning_rate": 7.234459128732257e-06,
+      "loss": 1.1191,
+      "step": 6125
+    },
+    {
+      "epoch": 13.5465564738292,
+      "grad_norm": 6.352032661437988,
+      "learning_rate": 7.173274596182086e-06,
+      "loss": 1.1093,
+      "step": 6150
+    },
+    {
+      "epoch": 13.601652892561983,
+      "grad_norm": 6.334588527679443,
+      "learning_rate": 7.112090063631915e-06,
+      "loss": 1.1429,
+      "step": 6175
+    },
+    {
+      "epoch": 13.656749311294766,
+      "grad_norm": 7.5459418296813965,
+      "learning_rate": 7.050905531081743e-06,
+      "loss": 1.069,
+      "step": 6200
+    },
+    {
+      "epoch": 13.711845730027548,
+      "grad_norm": 9.283977508544922,
+      "learning_rate": 6.989720998531572e-06,
+      "loss": 1.0905,
+      "step": 6225
+    },
+    {
+      "epoch": 13.76694214876033,
+      "grad_norm": 9.334644317626953,
+      "learning_rate": 6.928536465981401e-06,
+      "loss": 1.107,
+      "step": 6250
+    },
+    {
+      "epoch": 13.822038567493113,
+      "grad_norm": 6.586839199066162,
+      "learning_rate": 6.867351933431229e-06,
+      "loss": 1.091,
+      "step": 6275
+    },
+    {
+      "epoch": 13.877134986225895,
+      "grad_norm": 11.92888069152832,
+      "learning_rate": 6.806167400881057e-06,
+      "loss": 1.1088,
+      "step": 6300
+    },
+    {
+      "epoch": 13.932231404958678,
+      "grad_norm": 9.216996192932129,
+      "learning_rate": 6.744982868330886e-06,
+      "loss": 1.0462,
+      "step": 6325
+    },
+    {
+      "epoch": 13.98732782369146,
+      "grad_norm": 6.684733867645264,
+      "learning_rate": 6.683798335780716e-06,
+      "loss": 1.0711,
+      "step": 6350
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.10590383444917834,
+      "eval_loss": 0.7497881054878235,
+      "eval_macro_f1": 0.5163370092093984,
+      "eval_micro_f1": 0.5468609865470851,
+      "eval_runtime": 0.727,
+      "eval_samples_per_second": 2260.083,
+      "eval_steps_per_second": 141.685,
+      "eval_weighted_f1": 0.5529208492347623,
+      "step": 6356
+    },
+    {
+      "epoch": 14.041873278236915,
+      "grad_norm": 7.138929843902588,
+      "learning_rate": 6.622613803230544e-06,
+      "loss": 1.0558,
+      "step": 6375
+    },
+    {
+      "epoch": 14.096969696969698,
+      "grad_norm": 10.733924865722656,
+      "learning_rate": 6.5614292706803725e-06,
+      "loss": 1.0532,
+      "step": 6400
+    },
+    {
+      "epoch": 14.15206611570248,
+      "grad_norm": 12.12220573425293,
+      "learning_rate": 6.500244738130201e-06,
+      "loss": 1.0375,
+      "step": 6425
+    },
+    {
+      "epoch": 14.207162534435263,
+      "grad_norm": 17.76565170288086,
+      "learning_rate": 6.43906020558003e-06,
+      "loss": 1.0618,
+      "step": 6450
+    },
+    {
+      "epoch": 14.262258953168043,
+      "grad_norm": 7.568203449249268,
+      "learning_rate": 6.3778756730298585e-06,
+      "loss": 1.0354,
+      "step": 6475
+    },
+    {
+      "epoch": 14.317355371900826,
+      "grad_norm": 10.111392974853516,
+      "learning_rate": 6.316691140479687e-06,
+      "loss": 1.0496,
+      "step": 6500
+    },
+    {
+      "epoch": 14.372451790633608,
+      "grad_norm": 6.254069805145264,
+      "learning_rate": 6.255506607929516e-06,
+      "loss": 1.0712,
+      "step": 6525
+    },
+    {
+      "epoch": 14.42754820936639,
+      "grad_norm": 7.912994861602783,
+      "learning_rate": 6.194322075379344e-06,
+      "loss": 1.0293,
+      "step": 6550
+    },
+    {
+      "epoch": 14.482644628099173,
+      "grad_norm": 12.278776168823242,
+      "learning_rate": 6.133137542829173e-06,
+      "loss": 1.0933,
+      "step": 6575
+    },
+    {
+      "epoch": 14.537741046831956,
+      "grad_norm": 12.193331718444824,
+      "learning_rate": 6.071953010279002e-06,
+      "loss": 1.0321,
+      "step": 6600
+    },
+    {
+      "epoch": 14.592837465564738,
+      "grad_norm": 6.962792873382568,
+      "learning_rate": 6.010768477728831e-06,
+      "loss": 1.0203,
+      "step": 6625
+    },
+    {
+      "epoch": 14.64793388429752,
+      "grad_norm": 14.487834930419922,
+      "learning_rate": 5.94958394517866e-06,
+      "loss": 1.0275,
+      "step": 6650
+    },
+    {
+      "epoch": 14.703030303030303,
+      "grad_norm": 26.301523208618164,
+      "learning_rate": 5.8883994126284885e-06,
+      "loss": 1.0524,
+      "step": 6675
+    },
+    {
+      "epoch": 14.758126721763086,
+      "grad_norm": 8.678800582885742,
+      "learning_rate": 5.827214880078317e-06,
+      "loss": 1.0461,
+      "step": 6700
+    },
+    {
+      "epoch": 14.813223140495868,
+      "grad_norm": 7.605831146240234,
+      "learning_rate": 5.766030347528146e-06,
+      "loss": 1.0164,
+      "step": 6725
+    },
+    {
+      "epoch": 14.86831955922865,
+      "grad_norm": 16.34834861755371,
+      "learning_rate": 5.704845814977974e-06,
+      "loss": 1.0022,
+      "step": 6750
+    },
+    {
+      "epoch": 14.923415977961433,
+      "grad_norm": 10.478272438049316,
+      "learning_rate": 5.643661282427802e-06,
+      "loss": 1.0283,
+      "step": 6775
+    },
+    {
+      "epoch": 14.978512396694216,
+      "grad_norm": 6.517714977264404,
+      "learning_rate": 5.582476749877631e-06,
+      "loss": 0.9951,
+      "step": 6800
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.1199026171637249,
+      "eval_loss": 0.786878764629364,
+      "eval_macro_f1": 0.5303045519693917,
+      "eval_micro_f1": 0.5607280364018201,
+      "eval_runtime": 0.7664,
+      "eval_samples_per_second": 2143.802,
+      "eval_steps_per_second": 134.395,
+      "eval_weighted_f1": 0.5648358632177137,
+      "step": 6810
+    },
+    {
+      "epoch": 15.03305785123967,
+      "grad_norm": 12.449422836303711,
+      "learning_rate": 5.5212922173274596e-06,
+      "loss": 1.0273,
+      "step": 6825
+    },
+    {
+      "epoch": 15.088154269972453,
+      "grad_norm": 16.88009262084961,
+      "learning_rate": 5.460107684777288e-06,
+      "loss": 0.9533,
+      "step": 6850
+    },
+    {
+      "epoch": 15.143250688705233,
+      "grad_norm": 13.051454544067383,
+      "learning_rate": 5.398923152227118e-06,
+      "loss": 1.0137,
+      "step": 6875
+    },
+    {
+      "epoch": 15.198347107438016,
+      "grad_norm": 9.161343574523926,
+      "learning_rate": 5.337738619676946e-06,
+      "loss": 0.963,
+      "step": 6900
+    },
+    {
+      "epoch": 15.253443526170798,
+      "grad_norm": 16.16594886779785,
+      "learning_rate": 5.276554087126775e-06,
+      "loss": 0.9616,
+      "step": 6925
+    },
+    {
+      "epoch": 15.30853994490358,
+      "grad_norm": 8.046670913696289,
+      "learning_rate": 5.215369554576604e-06,
+      "loss": 0.983,
+      "step": 6950
+    },
+    {
+      "epoch": 15.363636363636363,
+      "grad_norm": 10.402591705322266,
+      "learning_rate": 5.154185022026432e-06,
+      "loss": 1.0406,
+      "step": 6975
+    },
+    {
+      "epoch": 15.418732782369146,
+      "grad_norm": 6.809611797332764,
+      "learning_rate": 5.093000489476261e-06,
+      "loss": 1.0326,
+      "step": 7000
+    },
+    {
+      "epoch": 15.473829201101928,
+      "grad_norm": 8.409727096557617,
+      "learning_rate": 5.03181595692609e-06,
+      "loss": 0.9884,
+      "step": 7025
+    },
+    {
+      "epoch": 15.52892561983471,
+      "grad_norm": 5.694775581359863,
+      "learning_rate": 4.970631424375918e-06,
+      "loss": 0.978,
+      "step": 7050
+    },
+    {
+      "epoch": 15.584022038567493,
+      "grad_norm": 20.40342140197754,
+      "learning_rate": 4.909446891825747e-06,
+      "loss": 1.0045,
+      "step": 7075
+    },
+    {
+      "epoch": 15.639118457300276,
+      "grad_norm": 10.138327598571777,
+      "learning_rate": 4.8482623592755756e-06,
+      "loss": 0.9975,
+      "step": 7100
+    },
+    {
+      "epoch": 15.694214876033058,
+      "grad_norm": 8.172189712524414,
+      "learning_rate": 4.787077826725404e-06,
+      "loss": 0.9591,
+      "step": 7125
+    },
+    {
+      "epoch": 15.74931129476584,
+      "grad_norm": 9.392911911010742,
+      "learning_rate": 4.725893294175233e-06,
+      "loss": 0.992,
+      "step": 7150
+    },
+    {
+      "epoch": 15.804407713498623,
+      "grad_norm": 11.717973709106445,
+      "learning_rate": 4.6647087616250615e-06,
+      "loss": 0.9686,
+      "step": 7175
+    },
+    {
+      "epoch": 15.859504132231406,
+      "grad_norm": 5.921899318695068,
+      "learning_rate": 4.60352422907489e-06,
+      "loss": 0.9505,
+      "step": 7200
+    },
+    {
+      "epoch": 15.914600550964188,
+      "grad_norm": 6.638045310974121,
+      "learning_rate": 4.542339696524719e-06,
+      "loss": 0.9778,
+      "step": 7225
+    },
+    {
+      "epoch": 15.969696969696969,
+      "grad_norm": 11.249752044677734,
+      "learning_rate": 4.4811551639745475e-06,
+      "loss": 0.9656,
+      "step": 7250
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.12903225806451613,
+      "eval_loss": 0.7909542322158813,
+      "eval_macro_f1": 0.5332530090007124,
+      "eval_micro_f1": 0.5646370023419204,
+      "eval_runtime": 0.9191,
+      "eval_samples_per_second": 1787.597,
+      "eval_steps_per_second": 112.065,
+      "eval_weighted_f1": 0.5695274753104705,
+      "step": 7264
+    },
+    {
+      "epoch": 16.024242424242424,
+      "grad_norm": 22.084747314453125,
+      "learning_rate": 4.419970631424376e-06,
+      "loss": 0.9781,
+      "step": 7275
+    },
+    {
+      "epoch": 16.079338842975208,
+      "grad_norm": 7.531091690063477,
+      "learning_rate": 4.358786098874205e-06,
+      "loss": 0.9447,
+      "step": 7300
+    },
+    {
+      "epoch": 16.13443526170799,
+      "grad_norm": 9.669242858886719,
+      "learning_rate": 4.297601566324034e-06,
+      "loss": 0.9171,
+      "step": 7325
+    },
+    {
+      "epoch": 16.189531680440773,
+      "grad_norm": 17.005956649780273,
+      "learning_rate": 4.236417033773862e-06,
+      "loss": 0.9611,
+      "step": 7350
+    },
+    {
+      "epoch": 16.244628099173553,
+      "grad_norm": 10.098047256469727,
+      "learning_rate": 4.175232501223691e-06,
+      "loss": 0.9458,
+      "step": 7375
+    },
+    {
+      "epoch": 16.299724517906338,
+      "grad_norm": 8.807305335998535,
+      "learning_rate": 4.114047968673519e-06,
+      "loss": 0.9424,
+      "step": 7400
+    },
+    {
+      "epoch": 16.35482093663912,
+      "grad_norm": 10.43460464477539,
+      "learning_rate": 4.052863436123348e-06,
+      "loss": 0.9741,
+      "step": 7425
+    },
+    {
+      "epoch": 16.4099173553719,
+      "grad_norm": 9.416433334350586,
+      "learning_rate": 3.991678903573177e-06,
+      "loss": 0.9422,
+      "step": 7450
+    },
+    {
+      "epoch": 16.465013774104683,
+      "grad_norm": 26.025875091552734,
+      "learning_rate": 3.930494371023006e-06,
+      "loss": 0.9808,
+      "step": 7475
+    },
+    {
+      "epoch": 16.520110192837464,
+      "grad_norm": 6.642582893371582,
+      "learning_rate": 3.869309838472835e-06,
+      "loss": 0.932,
+      "step": 7500
+    },
+    {
+      "epoch": 16.57520661157025,
+      "grad_norm": 13.423791885375977,
+      "learning_rate": 3.808125305922663e-06,
+      "loss": 0.9568,
+      "step": 7525
+    },
+    {
+      "epoch": 16.63030303030303,
+      "grad_norm": 7.4708757400512695,
+      "learning_rate": 3.7469407733724917e-06,
+      "loss": 0.9453,
+      "step": 7550
+    },
+    {
+      "epoch": 16.685399449035813,
+      "grad_norm": 5.990961074829102,
+      "learning_rate": 3.6857562408223203e-06,
+      "loss": 0.928,
+      "step": 7575
+    },
+    {
+      "epoch": 16.740495867768594,
+      "grad_norm": 7.558459758758545,
+      "learning_rate": 3.6245717082721494e-06,
+      "loss": 0.9763,
+      "step": 7600
+    },
+    {
+      "epoch": 16.79559228650138,
+      "grad_norm": 5.2827935218811035,
+      "learning_rate": 3.563387175721978e-06,
+      "loss": 0.9528,
+      "step": 7625
+    },
+    {
+      "epoch": 16.85068870523416,
+      "grad_norm": 15.335264205932617,
+      "learning_rate": 3.5022026431718063e-06,
+      "loss": 0.933,
+      "step": 7650
+    },
+    {
+      "epoch": 16.905785123966943,
+      "grad_norm": 8.040648460388184,
+      "learning_rate": 3.441018110621635e-06,
+      "loss": 0.9064,
+      "step": 7675
+    },
+    {
+      "epoch": 16.960881542699724,
+      "grad_norm": 14.046770095825195,
+      "learning_rate": 3.3798335780714636e-06,
+      "loss": 0.9466,
+      "step": 7700
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.13268411442483263,
+      "eval_loss": 0.7997220754623413,
+      "eval_macro_f1": 0.5386078032263989,
+      "eval_micro_f1": 0.566721972031287,
+      "eval_runtime": 0.708,
+      "eval_samples_per_second": 2320.756,
+      "eval_steps_per_second": 145.489,
+      "eval_weighted_f1": 0.5716301004766311,
+      "step": 7718
+    },
+    {
+      "epoch": 17.01542699724518,
+      "grad_norm": 12.992655754089355,
+      "learning_rate": 3.3186490455212927e-06,
+      "loss": 0.9179,
+      "step": 7725
+    },
+    {
+      "epoch": 17.070523415977963,
+      "grad_norm": 16.123558044433594,
+      "learning_rate": 3.2574645129711213e-06,
+      "loss": 0.9245,
+      "step": 7750
+    },
+    {
+      "epoch": 17.125619834710744,
+      "grad_norm": 6.627078533172607,
+      "learning_rate": 3.19627998042095e-06,
+      "loss": 0.9164,
+      "step": 7775
+    },
+    {
+      "epoch": 17.180716253443528,
+      "grad_norm": 10.419363021850586,
+      "learning_rate": 3.1350954478707786e-06,
+      "loss": 0.8796,
+      "step": 7800
+    },
+    {
+      "epoch": 17.23581267217631,
+      "grad_norm": 17.208833694458008,
+      "learning_rate": 3.073910915320607e-06,
+      "loss": 0.9119,
+      "step": 7825
+    },
+    {
+      "epoch": 17.29090909090909,
+      "grad_norm": 10.589327812194824,
+      "learning_rate": 3.0127263827704355e-06,
+      "loss": 0.9052,
+      "step": 7850
+    },
+    {
+      "epoch": 17.346005509641873,
+      "grad_norm": 11.223331451416016,
+      "learning_rate": 2.9515418502202646e-06,
+      "loss": 0.9467,
+      "step": 7875
+    },
+    {
+      "epoch": 17.401101928374654,
+      "grad_norm": 8.815620422363281,
+      "learning_rate": 2.8903573176700932e-06,
+      "loss": 0.9018,
+      "step": 7900
+    },
+    {
+      "epoch": 17.45619834710744,
+      "grad_norm": 7.268686294555664,
+      "learning_rate": 2.829172785119922e-06,
+      "loss": 0.9054,
+      "step": 7925
+    },
+    {
+      "epoch": 17.51129476584022,
+      "grad_norm": 14.526104927062988,
+      "learning_rate": 2.7679882525697505e-06,
+      "loss": 0.9042,
+      "step": 7950
+    },
+    {
+      "epoch": 17.566391184573003,
+      "grad_norm": 8.908108711242676,
+      "learning_rate": 2.706803720019579e-06,
+      "loss": 0.8553,
+      "step": 7975
+    },
+    {
+      "epoch": 17.621487603305784,
+      "grad_norm": 5.285924434661865,
+      "learning_rate": 2.6456191874694083e-06,
+      "loss": 0.9036,
+      "step": 8000
+    },
+    {
+      "epoch": 17.67658402203857,
+      "grad_norm": 9.32984733581543,
+      "learning_rate": 2.584434654919237e-06,
+      "loss": 0.9144,
+      "step": 8025
+    },
+    {
+      "epoch": 17.73168044077135,
+      "grad_norm": 6.950353145599365,
+      "learning_rate": 2.523250122369065e-06,
+      "loss": 0.9188,
+      "step": 8050
+    },
+    {
+      "epoch": 17.786776859504133,
+      "grad_norm": 7.641226291656494,
+      "learning_rate": 2.4620655898188938e-06,
+      "loss": 0.9088,
+      "step": 8075
+    },
+    {
+      "epoch": 17.841873278236914,
+      "grad_norm": 10.661919593811035,
+      "learning_rate": 2.400881057268723e-06,
+      "loss": 0.8949,
+      "step": 8100
+    },
+    {
+      "epoch": 17.8969696969697,
+      "grad_norm": 5.028895378112793,
+      "learning_rate": 2.3396965247185515e-06,
+      "loss": 0.9146,
+      "step": 8125
+    },
+    {
+      "epoch": 17.95206611570248,
+      "grad_norm": 10.540121078491211,
+      "learning_rate": 2.27851199216838e-06,
+      "loss": 0.9099,
+      "step": 8150
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.13451004260499086,
+      "eval_loss": 0.8080140352249146,
+      "eval_macro_f1": 0.535982460938898,
+      "eval_micro_f1": 0.566127497621313,
+      "eval_runtime": 0.7341,
+      "eval_samples_per_second": 2238.178,
+      "eval_steps_per_second": 140.312,
+      "eval_weighted_f1": 0.570448162053066,
+      "step": 8172
+    },
+    {
+      "epoch": 18.006611570247934,
+      "grad_norm": 7.576343536376953,
+      "learning_rate": 2.217327459618209e-06,
+      "loss": 0.8805,
+      "step": 8175
+    },
+    {
+      "epoch": 18.061707988980718,
+      "grad_norm": 8.76292610168457,
+      "learning_rate": 2.1561429270680375e-06,
+      "loss": 0.8528,
+      "step": 8200
+    },
+    {
+      "epoch": 18.1168044077135,
+      "grad_norm": 7.125899791717529,
+      "learning_rate": 2.094958394517866e-06,
+      "loss": 0.872,
+      "step": 8225
+    },
+    {
+      "epoch": 18.17190082644628,
+      "grad_norm": 7.739444732666016,
+      "learning_rate": 2.0337738619676948e-06,
+      "loss": 0.8481,
+      "step": 8250
+    },
+    {
+      "epoch": 18.226997245179064,
+      "grad_norm": 9.100935935974121,
+      "learning_rate": 1.9725893294175234e-06,
+      "loss": 0.8929,
+      "step": 8275
+    },
+    {
+      "epoch": 18.282093663911844,
+      "grad_norm": 6.645345687866211,
+      "learning_rate": 1.911404796867352e-06,
+      "loss": 0.8944,
+      "step": 8300
+    },
+    {
+      "epoch": 18.33719008264463,
+      "grad_norm": 10.697574615478516,
+      "learning_rate": 1.8502202643171807e-06,
+      "loss": 0.8416,
+      "step": 8325
+    },
+    {
+      "epoch": 18.39228650137741,
+      "grad_norm": 10.242743492126465,
+      "learning_rate": 1.7890357317670096e-06,
+      "loss": 0.9235,
+      "step": 8350
+    },
+    {
+      "epoch": 18.447382920110194,
+      "grad_norm": 8.19653606414795,
+      "learning_rate": 1.7278511992168382e-06,
+      "loss": 0.8798,
+      "step": 8375
+    },
+    {
+      "epoch": 18.502479338842974,
+      "grad_norm": 7.815878868103027,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.8979,
+      "step": 8400
+    },
+    {
+      "epoch": 18.55757575757576,
+      "grad_norm": 10.144920349121094,
+      "learning_rate": 1.6054821341164955e-06,
+      "loss": 0.9083,
+      "step": 8425
+    },
+    {
+      "epoch": 18.61267217630854,
+      "grad_norm": 12.199047088623047,
+      "learning_rate": 1.5442976015663242e-06,
+      "loss": 0.8923,
+      "step": 8450
+    },
+    {
+      "epoch": 18.667768595041323,
+      "grad_norm": 14.669466972351074,
+      "learning_rate": 1.4831130690161526e-06,
+      "loss": 0.9132,
+      "step": 8475
+    },
+    {
+      "epoch": 18.722865013774104,
+      "grad_norm": 6.225057125091553,
+      "learning_rate": 1.4219285364659815e-06,
+      "loss": 0.8451,
+      "step": 8500
+    },
+    {
+      "epoch": 18.77796143250689,
+      "grad_norm": 5.0352301597595215,
+      "learning_rate": 1.3607440039158101e-06,
+      "loss": 0.8885,
+      "step": 8525
+    },
+    {
+      "epoch": 18.83305785123967,
+      "grad_norm": 11.503164291381836,
+      "learning_rate": 1.299559471365639e-06,
+      "loss": 0.8928,
+      "step": 8550
+    },
+    {
+      "epoch": 18.888154269972453,
+      "grad_norm": 7.827541828155518,
+      "learning_rate": 1.2383749388154676e-06,
+      "loss": 0.8906,
+      "step": 8575
+    },
+    {
+      "epoch": 18.943250688705234,
+      "grad_norm": 7.901440143585205,
+      "learning_rate": 1.1771904062652963e-06,
+      "loss": 0.8679,
+      "step": 8600
+    },
+    {
+      "epoch": 18.998347107438015,
+      "grad_norm": 6.281285762786865,
+      "learning_rate": 1.116005873715125e-06,
+      "loss": 0.8768,
+      "step": 8625
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.13694461351186854,
+      "eval_loss": 0.8208908438682556,
+      "eval_macro_f1": 0.5389595030296993,
+      "eval_micro_f1": 0.5692087018885967,
+      "eval_runtime": 0.7096,
+      "eval_samples_per_second": 2315.51,
+      "eval_steps_per_second": 145.16,
+      "eval_weighted_f1": 0.5728033344104219,
+      "step": 8626
+    },
+    {
+      "epoch": 19.05289256198347,
+      "grad_norm": 8.007085800170898,
+      "learning_rate": 1.0548213411649536e-06,
+      "loss": 0.8682,
+      "step": 8650
+    },
+    {
+      "epoch": 19.107988980716254,
+      "grad_norm": 14.798598289489746,
+      "learning_rate": 9.936368086147822e-07,
+      "loss": 0.869,
+      "step": 8675
+    },
+    {
+      "epoch": 19.163085399449034,
+      "grad_norm": 7.352877616882324,
+      "learning_rate": 9.32452276064611e-07,
+      "loss": 0.8763,
+      "step": 8700
+    },
+    {
+      "epoch": 19.21818181818182,
+      "grad_norm": 7.396852016448975,
+      "learning_rate": 8.712677435144396e-07,
+      "loss": 0.8786,
+      "step": 8725
+    },
+    {
+      "epoch": 19.2732782369146,
+      "grad_norm": 6.227280616760254,
+      "learning_rate": 8.100832109642683e-07,
+      "loss": 0.8422,
+      "step": 8750
+    },
+    {
+      "epoch": 19.328374655647384,
+      "grad_norm": 19.285799026489258,
+      "learning_rate": 7.48898678414097e-07,
+      "loss": 0.8577,
+      "step": 8775
+    },
+    {
+      "epoch": 19.383471074380164,
+      "grad_norm": 16.623361587524414,
+      "learning_rate": 6.877141458639257e-07,
+      "loss": 0.8771,
+      "step": 8800
+    },
+    {
+      "epoch": 19.43856749311295,
+      "grad_norm": 5.518881320953369,
+      "learning_rate": 6.265296133137544e-07,
+      "loss": 0.8508,
+      "step": 8825
+    },
+    {
+      "epoch": 19.49366391184573,
+      "grad_norm": 8.208395957946777,
+      "learning_rate": 5.65345080763583e-07,
+      "loss": 0.8586,
+      "step": 8850
+    },
+    {
+      "epoch": 19.548760330578514,
+      "grad_norm": 6.739044666290283,
+      "learning_rate": 5.041605482134117e-07,
+      "loss": 0.8657,
+      "step": 8875
+    },
+    {
+      "epoch": 19.603856749311294,
+      "grad_norm": 4.547929286956787,
+      "learning_rate": 4.429760156632404e-07,
+      "loss": 0.8717,
+      "step": 8900
+    },
+    {
+      "epoch": 19.65895316804408,
+      "grad_norm": 6.962479114532471,
+      "learning_rate": 3.81791483113069e-07,
+      "loss": 0.8457,
+      "step": 8925
+    },
+    {
+      "epoch": 19.71404958677686,
+      "grad_norm": 8.265941619873047,
+      "learning_rate": 3.2060695056289777e-07,
+      "loss": 0.886,
+      "step": 8950
+    },
+    {
+      "epoch": 19.769146005509644,
+      "grad_norm": 5.207814693450928,
+      "learning_rate": 2.594224180127264e-07,
+      "loss": 0.8617,
+      "step": 8975
+    },
+    {
+      "epoch": 19.824242424242424,
+      "grad_norm": 8.848176002502441,
+      "learning_rate": 1.982378854625551e-07,
+      "loss": 0.8424,
+      "step": 9000
+    },
+    {
+      "epoch": 19.879338842975205,
+      "grad_norm": 11.36594009399414,
+      "learning_rate": 1.3705335291238375e-07,
+      "loss": 0.8665,
+      "step": 9025
+    },
+    {
+      "epoch": 19.93443526170799,
+      "grad_norm": 8.019472122192383,
+      "learning_rate": 7.586882036221244e-08,
+      "loss": 0.8632,
+      "step": 9050
+    },
+    {
+      "epoch": 19.98953168044077,
+      "grad_norm": 6.396020412445068,
+      "learning_rate": 1.4684287812041116e-08,
+      "loss": 0.8703,
+      "step": 9075
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.1351186853317103,
+      "eval_loss": 0.8197515606880188,
+      "eval_macro_f1": 0.5395569926876561,
+      "eval_micro_f1": 0.5694461059935398,
+      "eval_runtime": 0.7121,
+      "eval_samples_per_second": 2307.191,
+      "eval_steps_per_second": 144.638,
+      "eval_weighted_f1": 0.5731101639165003,
+      "step": 9080
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 9080,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

ood-vigoemotions-phobert-source-source-01-s43-20260826T134812Z-1f9ffdf3/trainer/checkpoint-7264/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

ood-vigoemotions-phobert-source-source-01-s43-20260826T134812Z-1f9ffdf3/trainer/checkpoint-7264/special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-01-s43-20260826T134812Z-1f9ffdf3/trainer/checkpoint-7264/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 256,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-01-s43-20260826T134812Z-1f9ffdf3/trainer/checkpoint-7264/trainer_state.json

@@ -0,0 +1,2265 @@
+{
+  "best_global_step": 7264,
+  "best_metric": 0.5387736977274954,
+  "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vigoemotions-phobert-source-source-01-s43-20260826T134812Z-1f9ffdf3/trainer/checkpoint-7264",
+  "epoch": 16.0,
+  "eval_steps": 500,
+  "global_step": 7264,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.05509641873278237,
+      "grad_norm": 4.721506595611572,
+      "learning_rate": 5.286343612334802e-07,
+      "loss": 5.1824,
+      "step": 25
+    },
+    {
+      "epoch": 0.11019283746556474,
+      "grad_norm": 4.242921352386475,
+      "learning_rate": 1.0792951541850223e-06,
+      "loss": 5.243,
+      "step": 50
+    },
+    {
+      "epoch": 0.1652892561983471,
+      "grad_norm": 5.161743640899658,
+      "learning_rate": 1.629955947136564e-06,
+      "loss": 5.2076,
+      "step": 75
+    },
+    {
+      "epoch": 0.22038567493112948,
+      "grad_norm": 5.828020095825195,
+      "learning_rate": 2.180616740088106e-06,
+      "loss": 5.0939,
+      "step": 100
+    },
+    {
+      "epoch": 0.27548209366391185,
+      "grad_norm": 5.841954231262207,
+      "learning_rate": 2.731277533039648e-06,
+      "loss": 5.0982,
+      "step": 125
+    },
+    {
+      "epoch": 0.3305785123966942,
+      "grad_norm": 3.9589149951934814,
+      "learning_rate": 3.2819383259911898e-06,
+      "loss": 5.1753,
+      "step": 150
+    },
+    {
+      "epoch": 0.3856749311294766,
+      "grad_norm": 4.0094428062438965,
+      "learning_rate": 3.8325991189427315e-06,
+      "loss": 5.1995,
+      "step": 175
+    },
+    {
+      "epoch": 0.44077134986225897,
+      "grad_norm": 4.752462387084961,
+      "learning_rate": 4.383259911894274e-06,
+      "loss": 5.2052,
+      "step": 200
+    },
+    {
+      "epoch": 0.49586776859504134,
+      "grad_norm": 7.853937149047852,
+      "learning_rate": 4.933920704845816e-06,
+      "loss": 5.0651,
+      "step": 225
+    },
+    {
+      "epoch": 0.5509641873278237,
+      "grad_norm": 6.576372146606445,
+      "learning_rate": 5.484581497797358e-06,
+      "loss": 5.0573,
+      "step": 250
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 5.6159844398498535,
+      "learning_rate": 6.035242290748899e-06,
+      "loss": 5.0438,
+      "step": 275
+    },
+    {
+      "epoch": 0.6611570247933884,
+      "grad_norm": 4.743258476257324,
+      "learning_rate": 6.585903083700441e-06,
+      "loss": 4.9259,
+      "step": 300
+    },
+    {
+      "epoch": 0.7162534435261708,
+      "grad_norm": 5.2631425857543945,
+      "learning_rate": 7.136563876651983e-06,
+      "loss": 4.8988,
+      "step": 325
+    },
+    {
+      "epoch": 0.7713498622589532,
+      "grad_norm": 5.224791049957275,
+      "learning_rate": 7.687224669603525e-06,
+      "loss": 4.8268,
+      "step": 350
+    },
+    {
+      "epoch": 0.8264462809917356,
+      "grad_norm": 5.436279773712158,
+      "learning_rate": 8.237885462555067e-06,
+      "loss": 4.6179,
+      "step": 375
+    },
+    {
+      "epoch": 0.8815426997245179,
+      "grad_norm": 6.9627909660339355,
+      "learning_rate": 8.788546255506607e-06,
+      "loss": 4.5771,
+      "step": 400
+    },
+    {
+      "epoch": 0.9366391184573003,
+      "grad_norm": 6.738986968994141,
+      "learning_rate": 9.339207048458151e-06,
+      "loss": 4.5826,
+      "step": 425
+    },
+    {
+      "epoch": 0.9917355371900827,
+      "grad_norm": 7.674351215362549,
+      "learning_rate": 9.889867841409693e-06,
+      "loss": 4.3028,
+      "step": 450
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 1.064934253692627,
+      "eval_macro_f1": 0.25079695642070166,
+      "eval_micro_f1": 0.26526126486162455,
+      "eval_runtime": 0.8081,
+      "eval_samples_per_second": 2033.215,
+      "eval_steps_per_second": 127.463,
+      "eval_weighted_f1": 0.32978528051651723,
+      "step": 454
+    },
+    {
+      "epoch": 1.0462809917355371,
+      "grad_norm": 5.639154434204102,
+      "learning_rate": 1.0440528634361234e-05,
+      "loss": 4.2781,
+      "step": 475
+    },
+    {
+      "epoch": 1.1013774104683196,
+      "grad_norm": 6.368907928466797,
+      "learning_rate": 1.0991189427312776e-05,
+      "loss": 4.1193,
+      "step": 500
+    },
+    {
+      "epoch": 1.1564738292011019,
+      "grad_norm": 11.263838768005371,
+      "learning_rate": 1.154185022026432e-05,
+      "loss": 4.2441,
+      "step": 525
+    },
+    {
+      "epoch": 1.2115702479338843,
+      "grad_norm": 5.8739824295043945,
+      "learning_rate": 1.209251101321586e-05,
+      "loss": 4.2407,
+      "step": 550
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 8.794628143310547,
+      "learning_rate": 1.2643171806167402e-05,
+      "loss": 4.1258,
+      "step": 575
+    },
+    {
+      "epoch": 1.321763085399449,
+      "grad_norm": 9.760412216186523,
+      "learning_rate": 1.3193832599118943e-05,
+      "loss": 4.105,
+      "step": 600
+    },
+    {
+      "epoch": 1.3768595041322313,
+      "grad_norm": 6.199836730957031,
+      "learning_rate": 1.3744493392070487e-05,
+      "loss": 4.0819,
+      "step": 625
+    },
+    {
+      "epoch": 1.4319559228650138,
+      "grad_norm": 11.421781539916992,
+      "learning_rate": 1.4295154185022027e-05,
+      "loss": 3.9663,
+      "step": 650
+    },
+    {
+      "epoch": 1.487052341597796,
+      "grad_norm": 11.619317054748535,
+      "learning_rate": 1.484581497797357e-05,
+      "loss": 3.9551,
+      "step": 675
+    },
+    {
+      "epoch": 1.5421487603305786,
+      "grad_norm": 8.50078010559082,
+      "learning_rate": 1.539647577092511e-05,
+      "loss": 3.9145,
+      "step": 700
+    },
+    {
+      "epoch": 1.5972451790633608,
+      "grad_norm": 9.922685623168945,
+      "learning_rate": 1.5947136563876654e-05,
+      "loss": 3.8852,
+      "step": 725
+    },
+    {
+      "epoch": 1.652341597796143,
+      "grad_norm": 8.09848690032959,
+      "learning_rate": 1.6497797356828196e-05,
+      "loss": 3.7545,
+      "step": 750
+    },
+    {
+      "epoch": 1.7074380165289256,
+      "grad_norm": 13.60912799835205,
+      "learning_rate": 1.7048458149779738e-05,
+      "loss": 3.7962,
+      "step": 775
+    },
+    {
+      "epoch": 1.762534435261708,
+      "grad_norm": 9.702221870422363,
+      "learning_rate": 1.759911894273128e-05,
+      "loss": 3.8082,
+      "step": 800
+    },
+    {
+      "epoch": 1.8176308539944903,
+      "grad_norm": 5.398869037628174,
+      "learning_rate": 1.814977973568282e-05,
+      "loss": 3.7219,
+      "step": 825
+    },
+    {
+      "epoch": 1.8727272727272726,
+      "grad_norm": 9.586057662963867,
+      "learning_rate": 1.8700440528634364e-05,
+      "loss": 3.6499,
+      "step": 850
+    },
+    {
+      "epoch": 1.927823691460055,
+      "grad_norm": 11.937478065490723,
+      "learning_rate": 1.9251101321585906e-05,
+      "loss": 3.7809,
+      "step": 875
+    },
+    {
+      "epoch": 1.9829201101928375,
+      "grad_norm": 9.817349433898926,
+      "learning_rate": 1.9801762114537445e-05,
+      "loss": 3.7224,
+      "step": 900
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 0.8863740563392639,
+      "eval_macro_f1": 0.31581366295673935,
+      "eval_micro_f1": 0.3377152379189039,
+      "eval_runtime": 0.7837,
+      "eval_samples_per_second": 2096.4,
+      "eval_steps_per_second": 131.424,
+      "eval_weighted_f1": 0.3906050874186073,
+      "step": 908
+    },
+    {
+      "epoch": 2.037465564738292,
+      "grad_norm": 9.083897590637207,
+      "learning_rate": 1.9960841899167892e-05,
+      "loss": 3.5236,
+      "step": 925
+    },
+    {
+      "epoch": 2.0925619834710742,
+      "grad_norm": 7.97184944152832,
+      "learning_rate": 1.989965736661772e-05,
+      "loss": 3.3794,
+      "step": 950
+    },
+    {
+      "epoch": 2.1476584022038567,
+      "grad_norm": 9.009512901306152,
+      "learning_rate": 1.983847283406755e-05,
+      "loss": 3.5319,
+      "step": 975
+    },
+    {
+      "epoch": 2.202754820936639,
+      "grad_norm": 11.871821403503418,
+      "learning_rate": 1.9777288301517378e-05,
+      "loss": 3.4795,
+      "step": 1000
+    },
+    {
+      "epoch": 2.2578512396694217,
+      "grad_norm": 11.774299621582031,
+      "learning_rate": 1.9716103768967207e-05,
+      "loss": 3.4565,
+      "step": 1025
+    },
+    {
+      "epoch": 2.3129476584022037,
+      "grad_norm": 9.386924743652344,
+      "learning_rate": 1.9654919236417035e-05,
+      "loss": 3.3905,
+      "step": 1050
+    },
+    {
+      "epoch": 2.368044077134986,
+      "grad_norm": 7.399035930633545,
+      "learning_rate": 1.9593734703866864e-05,
+      "loss": 3.4593,
+      "step": 1075
+    },
+    {
+      "epoch": 2.4231404958677687,
+      "grad_norm": 9.72281265258789,
+      "learning_rate": 1.9532550171316693e-05,
+      "loss": 3.399,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4782369146005507,
+      "grad_norm": 11.944668769836426,
+      "learning_rate": 1.947136563876652e-05,
+      "loss": 3.309,
+      "step": 1125
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 10.8898286819458,
+      "learning_rate": 1.941018110621635e-05,
+      "loss": 3.3269,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5884297520661157,
+      "grad_norm": 12.417951583862305,
+      "learning_rate": 1.934899657366618e-05,
+      "loss": 3.4443,
+      "step": 1175
+    },
+    {
+      "epoch": 2.643526170798898,
+      "grad_norm": 7.977325916290283,
+      "learning_rate": 1.9287812041116007e-05,
+      "loss": 3.3148,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6986225895316807,
+      "grad_norm": 13.754631996154785,
+      "learning_rate": 1.9226627508565836e-05,
+      "loss": 3.2603,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7537190082644627,
+      "grad_norm": 7.910646915435791,
+      "learning_rate": 1.9165442976015665e-05,
+      "loss": 3.221,
+      "step": 1250
+    },
+    {
+      "epoch": 2.808815426997245,
+      "grad_norm": 8.343128204345703,
+      "learning_rate": 1.9104258443465493e-05,
+      "loss": 3.2729,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8639118457300277,
+      "grad_norm": 6.801973819732666,
+      "learning_rate": 1.9043073910915322e-05,
+      "loss": 3.2611,
+      "step": 1300
+    },
+    {
+      "epoch": 2.9190082644628097,
+      "grad_norm": 11.878301620483398,
+      "learning_rate": 1.898188937836515e-05,
+      "loss": 3.2498,
+      "step": 1325
+    },
+    {
+      "epoch": 2.974104683195592,
+      "grad_norm": 8.057741165161133,
+      "learning_rate": 1.892070484581498e-05,
+      "loss": 3.1971,
+      "step": 1350
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.0018259281801582471,
+      "eval_loss": 0.7955052852630615,
+      "eval_macro_f1": 0.371248723286375,
+      "eval_micro_f1": 0.3993803481853054,
+      "eval_runtime": 0.7772,
+      "eval_samples_per_second": 2114.059,
+      "eval_steps_per_second": 132.531,
+      "eval_weighted_f1": 0.4370260452300996,
+      "step": 1362
+    },
+    {
+      "epoch": 3.028650137741047,
+      "grad_norm": 9.659102439880371,
+      "learning_rate": 1.8859520313264808e-05,
+      "loss": 3.0296,
+      "step": 1375
+    },
+    {
+      "epoch": 3.0837465564738293,
+      "grad_norm": 6.875763416290283,
+      "learning_rate": 1.8798335780714636e-05,
+      "loss": 2.974,
+      "step": 1400
+    },
+    {
+      "epoch": 3.1388429752066114,
+      "grad_norm": 15.656290054321289,
+      "learning_rate": 1.8737151248164465e-05,
+      "loss": 3.071,
+      "step": 1425
+    },
+    {
+      "epoch": 3.193939393939394,
+      "grad_norm": 13.179049491882324,
+      "learning_rate": 1.8675966715614294e-05,
+      "loss": 2.9726,
+      "step": 1450
+    },
+    {
+      "epoch": 3.2490358126721763,
+      "grad_norm": 7.199787139892578,
+      "learning_rate": 1.8614782183064122e-05,
+      "loss": 2.9661,
+      "step": 1475
+    },
+    {
+      "epoch": 3.304132231404959,
+      "grad_norm": 6.702415466308594,
+      "learning_rate": 1.855359765051395e-05,
+      "loss": 2.9899,
+      "step": 1500
+    },
+    {
+      "epoch": 3.359228650137741,
+      "grad_norm": 7.977599620819092,
+      "learning_rate": 1.849241311796378e-05,
+      "loss": 2.9054,
+      "step": 1525
+    },
+    {
+      "epoch": 3.4143250688705233,
+      "grad_norm": 8.159268379211426,
+      "learning_rate": 1.843122858541361e-05,
+      "loss": 2.9468,
+      "step": 1550
+    },
+    {
+      "epoch": 3.469421487603306,
+      "grad_norm": 12.21645450592041,
+      "learning_rate": 1.8370044052863437e-05,
+      "loss": 2.9158,
+      "step": 1575
+    },
+    {
+      "epoch": 3.5245179063360883,
+      "grad_norm": 9.832043647766113,
+      "learning_rate": 1.8308859520313266e-05,
+      "loss": 2.9598,
+      "step": 1600
+    },
+    {
+      "epoch": 3.579614325068871,
+      "grad_norm": 9.701088905334473,
+      "learning_rate": 1.8247674987763094e-05,
+      "loss": 2.8472,
+      "step": 1625
+    },
+    {
+      "epoch": 3.634710743801653,
+      "grad_norm": 13.512517929077148,
+      "learning_rate": 1.8186490455212923e-05,
+      "loss": 2.9392,
+      "step": 1650
+    },
+    {
+      "epoch": 3.6898071625344353,
+      "grad_norm": 14.005834579467773,
+      "learning_rate": 1.812530592266275e-05,
+      "loss": 2.8569,
+      "step": 1675
+    },
+    {
+      "epoch": 3.744903581267218,
+      "grad_norm": 11.314921379089355,
+      "learning_rate": 1.806412139011258e-05,
+      "loss": 2.8256,
+      "step": 1700
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 9.094903945922852,
+      "learning_rate": 1.800293685756241e-05,
+      "loss": 2.8047,
+      "step": 1725
+    },
+    {
+      "epoch": 3.8550964187327823,
+      "grad_norm": 7.849555969238281,
+      "learning_rate": 1.7941752325012238e-05,
+      "loss": 2.8377,
+      "step": 1750
+    },
+    {
+      "epoch": 3.910192837465565,
+      "grad_norm": 13.487672805786133,
+      "learning_rate": 1.7880567792462066e-05,
+      "loss": 2.8002,
+      "step": 1775
+    },
+    {
+      "epoch": 3.9652892561983473,
+      "grad_norm": 21.90669822692871,
+      "learning_rate": 1.7819383259911895e-05,
+      "loss": 2.7817,
+      "step": 1800
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.010346926354230066,
+      "eval_loss": 0.750116765499115,
+      "eval_macro_f1": 0.39218802795663554,
+      "eval_micro_f1": 0.42694649150913166,
+      "eval_runtime": 0.833,
+      "eval_samples_per_second": 1972.37,
+      "eval_steps_per_second": 123.648,
+      "eval_weighted_f1": 0.45943726636348214,
+      "step": 1816
+    },
+    {
+      "epoch": 4.019834710743802,
+      "grad_norm": 10.072190284729004,
+      "learning_rate": 1.7758198727361724e-05,
+      "loss": 2.6486,
+      "step": 1825
+    },
+    {
+      "epoch": 4.074931129476584,
+      "grad_norm": 6.724724292755127,
+      "learning_rate": 1.7697014194811552e-05,
+      "loss": 2.5753,
+      "step": 1850
+    },
+    {
+      "epoch": 4.130027548209366,
+      "grad_norm": 14.258201599121094,
+      "learning_rate": 1.763582966226138e-05,
+      "loss": 2.5727,
+      "step": 1875
+    },
+    {
+      "epoch": 4.1851239669421485,
+      "grad_norm": 8.002503395080566,
+      "learning_rate": 1.757464512971121e-05,
+      "loss": 2.5374,
+      "step": 1900
+    },
+    {
+      "epoch": 4.240220385674931,
+      "grad_norm": 7.770499229431152,
+      "learning_rate": 1.7513460597161038e-05,
+      "loss": 2.586,
+      "step": 1925
+    },
+    {
+      "epoch": 4.2953168044077135,
+      "grad_norm": 8.485304832458496,
+      "learning_rate": 1.7452276064610867e-05,
+      "loss": 2.6794,
+      "step": 1950
+    },
+    {
+      "epoch": 4.350413223140496,
+      "grad_norm": 12.671653747558594,
+      "learning_rate": 1.7391091532060695e-05,
+      "loss": 2.6331,
+      "step": 1975
+    },
+    {
+      "epoch": 4.405509641873278,
+      "grad_norm": 10.223567008972168,
+      "learning_rate": 1.7329906999510527e-05,
+      "loss": 2.5868,
+      "step": 2000
+    },
+    {
+      "epoch": 4.460606060606061,
+      "grad_norm": 14.223559379577637,
+      "learning_rate": 1.7268722466960356e-05,
+      "loss": 2.5251,
+      "step": 2025
+    },
+    {
+      "epoch": 4.515702479338843,
+      "grad_norm": 8.538398742675781,
+      "learning_rate": 1.7207537934410185e-05,
+      "loss": 2.5315,
+      "step": 2050
+    },
+    {
+      "epoch": 4.570798898071625,
+      "grad_norm": 5.16135311126709,
+      "learning_rate": 1.7146353401860013e-05,
+      "loss": 2.5609,
+      "step": 2075
+    },
+    {
+      "epoch": 4.6258953168044075,
+      "grad_norm": 11.051167488098145,
+      "learning_rate": 1.7085168869309842e-05,
+      "loss": 2.4787,
+      "step": 2100
+    },
+    {
+      "epoch": 4.68099173553719,
+      "grad_norm": 9.579821586608887,
+      "learning_rate": 1.702398433675967e-05,
+      "loss": 2.6092,
+      "step": 2125
+    },
+    {
+      "epoch": 4.736088154269972,
+      "grad_norm": 7.40781831741333,
+      "learning_rate": 1.69627998042095e-05,
+      "loss": 2.5947,
+      "step": 2150
+    },
+    {
+      "epoch": 4.791184573002755,
+      "grad_norm": 6.301372528076172,
+      "learning_rate": 1.6901615271659325e-05,
+      "loss": 2.4604,
+      "step": 2175
+    },
+    {
+      "epoch": 4.846280991735537,
+      "grad_norm": 7.67243766784668,
+      "learning_rate": 1.6840430739109153e-05,
+      "loss": 2.5283,
+      "step": 2200
+    },
+    {
+      "epoch": 4.90137741046832,
+      "grad_norm": 6.105996608734131,
+      "learning_rate": 1.6779246206558982e-05,
+      "loss": 2.4024,
+      "step": 2225
+    },
+    {
+      "epoch": 4.9564738292011015,
+      "grad_norm": 13.376340866088867,
+      "learning_rate": 1.671806167400881e-05,
+      "loss": 2.481,
+      "step": 2250
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.021911138161898967,
+      "eval_loss": 0.7181954979896545,
+      "eval_macro_f1": 0.40061912618913414,
+      "eval_micro_f1": 0.4357821089455272,
+      "eval_runtime": 0.7813,
+      "eval_samples_per_second": 2102.788,
+      "eval_steps_per_second": 131.824,
+      "eval_weighted_f1": 0.4673903225695518,
+      "step": 2270
+    },
+    {
+      "epoch": 5.011019283746556,
+      "grad_norm": 8.279224395751953,
+      "learning_rate": 1.665687714145864e-05,
+      "loss": 2.4698,
+      "step": 2275
+    },
+    {
+      "epoch": 5.066115702479339,
+      "grad_norm": 7.251194477081299,
+      "learning_rate": 1.6595692608908468e-05,
+      "loss": 2.3897,
+      "step": 2300
+    },
+    {
+      "epoch": 5.121212121212121,
+      "grad_norm": 11.917134284973145,
+      "learning_rate": 1.6534508076358297e-05,
+      "loss": 2.3596,
+      "step": 2325
+    },
+    {
+      "epoch": 5.176308539944904,
+      "grad_norm": 11.278261184692383,
+      "learning_rate": 1.6473323543808125e-05,
+      "loss": 2.2631,
+      "step": 2350
+    },
+    {
+      "epoch": 5.231404958677686,
+      "grad_norm": 8.443638801574707,
+      "learning_rate": 1.6412139011257954e-05,
+      "loss": 2.1933,
+      "step": 2375
+    },
+    {
+      "epoch": 5.2865013774104685,
+      "grad_norm": 9.477250099182129,
+      "learning_rate": 1.6350954478707782e-05,
+      "loss": 2.3299,
+      "step": 2400
+    },
+    {
+      "epoch": 5.341597796143251,
+      "grad_norm": 5.243191242218018,
+      "learning_rate": 1.628976994615761e-05,
+      "loss": 2.29,
+      "step": 2425
+    },
+    {
+      "epoch": 5.3966942148760335,
+      "grad_norm": 10.38256549835205,
+      "learning_rate": 1.622858541360744e-05,
+      "loss": 2.3002,
+      "step": 2450
+    },
+    {
+      "epoch": 5.451790633608815,
+      "grad_norm": 10.228897094726562,
+      "learning_rate": 1.616740088105727e-05,
+      "loss": 2.3167,
+      "step": 2475
+    },
+    {
+      "epoch": 5.506887052341598,
+      "grad_norm": 13.564765930175781,
+      "learning_rate": 1.6106216348507097e-05,
+      "loss": 2.2538,
+      "step": 2500
+    },
+    {
+      "epoch": 5.56198347107438,
+      "grad_norm": 13.028368949890137,
+      "learning_rate": 1.604503181595693e-05,
+      "loss": 2.2295,
+      "step": 2525
+    },
+    {
+      "epoch": 5.6170798898071626,
+      "grad_norm": 11.13936710357666,
+      "learning_rate": 1.5983847283406758e-05,
+      "loss": 2.2291,
+      "step": 2550
+    },
+    {
+      "epoch": 5.672176308539945,
+      "grad_norm": 12.807024002075195,
+      "learning_rate": 1.5922662750856586e-05,
+      "loss": 2.2607,
+      "step": 2575
+    },
+    {
+      "epoch": 5.7272727272727275,
+      "grad_norm": 12.083641052246094,
+      "learning_rate": 1.5861478218306415e-05,
+      "loss": 2.2267,
+      "step": 2600
+    },
+    {
+      "epoch": 5.78236914600551,
+      "grad_norm": 6.688780784606934,
+      "learning_rate": 1.5800293685756244e-05,
+      "loss": 2.197,
+      "step": 2625
+    },
+    {
+      "epoch": 5.837465564738292,
+      "grad_norm": 10.367467880249023,
+      "learning_rate": 1.5739109153206072e-05,
+      "loss": 2.2441,
+      "step": 2650
+    },
+    {
+      "epoch": 5.892561983471074,
+      "grad_norm": 8.155801773071289,
+      "learning_rate": 1.56779246206559e-05,
+      "loss": 2.18,
+      "step": 2675
+    },
+    {
+      "epoch": 5.9476584022038566,
+      "grad_norm": 15.422148704528809,
+      "learning_rate": 1.561674008810573e-05,
+      "loss": 2.2327,
+      "step": 2700
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.03408399269628728,
+      "eval_loss": 0.7054561376571655,
+      "eval_macro_f1": 0.4520903933181108,
+      "eval_micro_f1": 0.4842243235652092,
+      "eval_runtime": 0.7411,
+      "eval_samples_per_second": 2216.967,
+      "eval_steps_per_second": 138.982,
+      "eval_weighted_f1": 0.503275021503218,
+      "step": 2724
+    },
+    {
+      "epoch": 6.002203856749311,
+      "grad_norm": 8.507722854614258,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 2.1608,
+      "step": 2725
+    },
+    {
+      "epoch": 6.057300275482094,
+      "grad_norm": 11.379617691040039,
+      "learning_rate": 1.5494371023005387e-05,
+      "loss": 2.0387,
+      "step": 2750
+    },
+    {
+      "epoch": 6.112396694214876,
+      "grad_norm": 8.296934127807617,
+      "learning_rate": 1.5433186490455216e-05,
+      "loss": 2.1263,
+      "step": 2775
+    },
+    {
+      "epoch": 6.167493112947659,
+      "grad_norm": 9.881864547729492,
+      "learning_rate": 1.5372001957905044e-05,
+      "loss": 2.0996,
+      "step": 2800
+    },
+    {
+      "epoch": 6.222589531680441,
+      "grad_norm": 15.106012344360352,
+      "learning_rate": 1.5310817425354873e-05,
+      "loss": 2.0111,
+      "step": 2825
+    },
+    {
+      "epoch": 6.277685950413223,
+      "grad_norm": 10.103724479675293,
+      "learning_rate": 1.52496328928047e-05,
+      "loss": 2.0883,
+      "step": 2850
+    },
+    {
+      "epoch": 6.332782369146005,
+      "grad_norm": 9.56401538848877,
+      "learning_rate": 1.5188448360254529e-05,
+      "loss": 2.0111,
+      "step": 2875
+    },
+    {
+      "epoch": 6.387878787878788,
+      "grad_norm": 7.425657749176025,
+      "learning_rate": 1.5127263827704357e-05,
+      "loss": 2.0236,
+      "step": 2900
+    },
+    {
+      "epoch": 6.44297520661157,
+      "grad_norm": 10.291604042053223,
+      "learning_rate": 1.5066079295154186e-05,
+      "loss": 2.0612,
+      "step": 2925
+    },
+    {
+      "epoch": 6.498071625344353,
+      "grad_norm": 9.244171142578125,
+      "learning_rate": 1.5004894762604014e-05,
+      "loss": 1.9911,
+      "step": 2950
+    },
+    {
+      "epoch": 6.553168044077135,
+      "grad_norm": 10.03786563873291,
+      "learning_rate": 1.4943710230053843e-05,
+      "loss": 2.0577,
+      "step": 2975
+    },
+    {
+      "epoch": 6.608264462809918,
+      "grad_norm": 7.586009502410889,
+      "learning_rate": 1.4882525697503672e-05,
+      "loss": 2.0063,
+      "step": 3000
+    },
+    {
+      "epoch": 6.663360881542699,
+      "grad_norm": 11.507920265197754,
+      "learning_rate": 1.48213411649535e-05,
+      "loss": 1.9918,
+      "step": 3025
+    },
+    {
+      "epoch": 6.718457300275482,
+      "grad_norm": 10.177037239074707,
+      "learning_rate": 1.476015663240333e-05,
+      "loss": 1.989,
+      "step": 3050
+    },
+    {
+      "epoch": 6.773553719008264,
+      "grad_norm": 11.146783828735352,
+      "learning_rate": 1.469897209985316e-05,
+      "loss": 1.9675,
+      "step": 3075
+    },
+    {
+      "epoch": 6.828650137741047,
+      "grad_norm": 23.576797485351562,
+      "learning_rate": 1.4637787567302988e-05,
+      "loss": 2.066,
+      "step": 3100
+    },
+    {
+      "epoch": 6.883746556473829,
+      "grad_norm": 6.994001865386963,
+      "learning_rate": 1.4576603034752817e-05,
+      "loss": 1.9676,
+      "step": 3125
+    },
+    {
+      "epoch": 6.938842975206612,
+      "grad_norm": 12.981430053710938,
+      "learning_rate": 1.4515418502202645e-05,
+      "loss": 1.951,
+      "step": 3150
+    },
+    {
+      "epoch": 6.993939393939394,
+      "grad_norm": 6.651545524597168,
+      "learning_rate": 1.4454233969652474e-05,
+      "loss": 1.9298,
+      "step": 3175
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.054777845404747415,
+      "eval_loss": 0.7065615653991699,
+      "eval_macro_f1": 0.47539287952540293,
+      "eval_micro_f1": 0.5047904191616767,
+      "eval_runtime": 0.7799,
+      "eval_samples_per_second": 2106.647,
+      "eval_steps_per_second": 132.066,
+      "eval_weighted_f1": 0.5200095920230331,
+      "step": 3178
+    },
+    {
+      "epoch": 7.048484848484849,
+      "grad_norm": 8.36910629272461,
+      "learning_rate": 1.4393049437102303e-05,
+      "loss": 1.8839,
+      "step": 3200
+    },
+    {
+      "epoch": 7.103581267217631,
+      "grad_norm": 11.923949241638184,
+      "learning_rate": 1.4331864904552131e-05,
+      "loss": 1.8423,
+      "step": 3225
+    },
+    {
+      "epoch": 7.158677685950413,
+      "grad_norm": 9.617166519165039,
+      "learning_rate": 1.427068037200196e-05,
+      "loss": 1.8561,
+      "step": 3250
+    },
+    {
+      "epoch": 7.213774104683195,
+      "grad_norm": 16.226926803588867,
+      "learning_rate": 1.4209495839451789e-05,
+      "loss": 1.8183,
+      "step": 3275
+    },
+    {
+      "epoch": 7.268870523415978,
+      "grad_norm": 12.089911460876465,
+      "learning_rate": 1.4148311306901617e-05,
+      "loss": 1.8012,
+      "step": 3300
+    },
+    {
+      "epoch": 7.32396694214876,
+      "grad_norm": 8.376047134399414,
+      "learning_rate": 1.4087126774351444e-05,
+      "loss": 1.8334,
+      "step": 3325
+    },
+    {
+      "epoch": 7.379063360881543,
+      "grad_norm": 13.260794639587402,
+      "learning_rate": 1.4025942241801273e-05,
+      "loss": 1.8614,
+      "step": 3350
+    },
+    {
+      "epoch": 7.434159779614325,
+      "grad_norm": 9.980430603027344,
+      "learning_rate": 1.3964757709251102e-05,
+      "loss": 1.8474,
+      "step": 3375
+    },
+    {
+      "epoch": 7.489256198347108,
+      "grad_norm": 7.031350135803223,
+      "learning_rate": 1.390357317670093e-05,
+      "loss": 1.838,
+      "step": 3400
+    },
+    {
+      "epoch": 7.544352617079889,
+      "grad_norm": 12.172435760498047,
+      "learning_rate": 1.3842388644150759e-05,
+      "loss": 1.8082,
+      "step": 3425
+    },
+    {
+      "epoch": 7.599449035812672,
+      "grad_norm": 10.703603744506836,
+      "learning_rate": 1.3781204111600588e-05,
+      "loss": 1.8367,
+      "step": 3450
+    },
+    {
+      "epoch": 7.654545454545454,
+      "grad_norm": 11.197491645812988,
+      "learning_rate": 1.3720019579050416e-05,
+      "loss": 1.7869,
+      "step": 3475
+    },
+    {
+      "epoch": 7.709641873278237,
+      "grad_norm": 23.729402542114258,
+      "learning_rate": 1.3658835046500245e-05,
+      "loss": 1.7855,
+      "step": 3500
+    },
+    {
+      "epoch": 7.764738292011019,
+      "grad_norm": 8.929984092712402,
+      "learning_rate": 1.3597650513950073e-05,
+      "loss": 1.8359,
+      "step": 3525
+    },
+    {
+      "epoch": 7.819834710743802,
+      "grad_norm": 17.336320877075195,
+      "learning_rate": 1.3536465981399902e-05,
+      "loss": 1.7919,
+      "step": 3550
+    },
+    {
+      "epoch": 7.874931129476584,
+      "grad_norm": 9.479592323303223,
+      "learning_rate": 1.3475281448849732e-05,
+      "loss": 1.7993,
+      "step": 3575
+    },
+    {
+      "epoch": 7.930027548209367,
+      "grad_norm": 8.46316909790039,
+      "learning_rate": 1.3414096916299561e-05,
+      "loss": 1.7886,
+      "step": 3600
+    },
+    {
+      "epoch": 7.985123966942149,
+      "grad_norm": 6.84007453918457,
+      "learning_rate": 1.335291238374939e-05,
+      "loss": 1.8102,
+      "step": 3625
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.06451612903225806,
+      "eval_loss": 0.6984469294548035,
+      "eval_macro_f1": 0.482549048079171,
+      "eval_micro_f1": 0.5128887070376432,
+      "eval_runtime": 0.7756,
+      "eval_samples_per_second": 2118.326,
+      "eval_steps_per_second": 132.798,
+      "eval_weighted_f1": 0.5248092674046481,
+      "step": 3632
+    },
+    {
+      "epoch": 8.039669421487604,
+      "grad_norm": 16.53033447265625,
+      "learning_rate": 1.3291727851199218e-05,
+      "loss": 1.7326,
+      "step": 3650
+    },
+    {
+      "epoch": 8.094765840220386,
+      "grad_norm": 7.538041114807129,
+      "learning_rate": 1.3230543318649047e-05,
+      "loss": 1.7154,
+      "step": 3675
+    },
+    {
+      "epoch": 8.149862258953169,
+      "grad_norm": 7.187880992889404,
+      "learning_rate": 1.3169358786098876e-05,
+      "loss": 1.6261,
+      "step": 3700
+    },
+    {
+      "epoch": 8.20495867768595,
+      "grad_norm": 6.9718403816223145,
+      "learning_rate": 1.3108174253548704e-05,
+      "loss": 1.698,
+      "step": 3725
+    },
+    {
+      "epoch": 8.260055096418732,
+      "grad_norm": 11.223376274108887,
+      "learning_rate": 1.3046989720998533e-05,
+      "loss": 1.6221,
+      "step": 3750
+    },
+    {
+      "epoch": 8.315151515151515,
+      "grad_norm": 9.946511268615723,
+      "learning_rate": 1.2985805188448362e-05,
+      "loss": 1.5966,
+      "step": 3775
+    },
+    {
+      "epoch": 8.370247933884297,
+      "grad_norm": 10.008794784545898,
+      "learning_rate": 1.292462065589819e-05,
+      "loss": 1.6861,
+      "step": 3800
+    },
+    {
+      "epoch": 8.42534435261708,
+      "grad_norm": 13.9457368850708,
+      "learning_rate": 1.2863436123348019e-05,
+      "loss": 1.6792,
+      "step": 3825
+    },
+    {
+      "epoch": 8.480440771349862,
+      "grad_norm": 7.465098857879639,
+      "learning_rate": 1.2802251590797848e-05,
+      "loss": 1.6933,
+      "step": 3850
+    },
+    {
+      "epoch": 8.535537190082644,
+      "grad_norm": 9.797990798950195,
+      "learning_rate": 1.2741067058247676e-05,
+      "loss": 1.673,
+      "step": 3875
+    },
+    {
+      "epoch": 8.590633608815427,
+      "grad_norm": 11.942303657531738,
+      "learning_rate": 1.2679882525697505e-05,
+      "loss": 1.643,
+      "step": 3900
+    },
+    {
+      "epoch": 8.64573002754821,
+      "grad_norm": 6.5281662940979,
+      "learning_rate": 1.2618697993147334e-05,
+      "loss": 1.6283,
+      "step": 3925
+    },
+    {
+      "epoch": 8.700826446280992,
+      "grad_norm": 7.67686653137207,
+      "learning_rate": 1.2557513460597162e-05,
+      "loss": 1.6683,
+      "step": 3950
+    },
+    {
+      "epoch": 8.755922865013774,
+      "grad_norm": 10.89754867553711,
+      "learning_rate": 1.2496328928046991e-05,
+      "loss": 1.6098,
+      "step": 3975
+    },
+    {
+      "epoch": 8.811019283746557,
+      "grad_norm": 7.0877685546875,
+      "learning_rate": 1.243514439549682e-05,
+      "loss": 1.6264,
+      "step": 4000
+    },
+    {
+      "epoch": 8.86611570247934,
+      "grad_norm": 6.110809803009033,
+      "learning_rate": 1.2373959862946646e-05,
+      "loss": 1.6501,
+      "step": 4025
+    },
+    {
+      "epoch": 8.921212121212122,
+      "grad_norm": 15.987994194030762,
+      "learning_rate": 1.2312775330396475e-05,
+      "loss": 1.6452,
+      "step": 4050
+    },
+    {
+      "epoch": 8.976308539944904,
+      "grad_norm": 9.162452697753906,
+      "learning_rate": 1.2251590797846304e-05,
+      "loss": 1.5748,
+      "step": 4075
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.054777845404747415,
+      "eval_loss": 0.6941969990730286,
+      "eval_macro_f1": 0.47838987201665556,
+      "eval_micro_f1": 0.5094188376753507,
+      "eval_runtime": 0.7496,
+      "eval_samples_per_second": 2191.758,
+      "eval_steps_per_second": 137.402,
+      "eval_weighted_f1": 0.5218786373886054,
+      "step": 4086
+    },
+    {
+      "epoch": 9.030853994490359,
+      "grad_norm": 7.348278999328613,
+      "learning_rate": 1.2190406265296136e-05,
+      "loss": 1.5912,
+      "step": 4100
+    },
+    {
+      "epoch": 9.08595041322314,
+      "grad_norm": 13.46335506439209,
+      "learning_rate": 1.2129221732745964e-05,
+      "loss": 1.5393,
+      "step": 4125
+    },
+    {
+      "epoch": 9.141046831955922,
+      "grad_norm": 12.748778343200684,
+      "learning_rate": 1.2068037200195793e-05,
+      "loss": 1.5056,
+      "step": 4150
+    },
+    {
+      "epoch": 9.196143250688705,
+      "grad_norm": 6.756908893585205,
+      "learning_rate": 1.2006852667645622e-05,
+      "loss": 1.4559,
+      "step": 4175
+    },
+    {
+      "epoch": 9.251239669421487,
+      "grad_norm": 14.214062690734863,
+      "learning_rate": 1.1945668135095449e-05,
+      "loss": 1.4885,
+      "step": 4200
+    },
+    {
+      "epoch": 9.30633608815427,
+      "grad_norm": 13.800779342651367,
+      "learning_rate": 1.1884483602545277e-05,
+      "loss": 1.5808,
+      "step": 4225
+    },
+    {
+      "epoch": 9.361432506887052,
+      "grad_norm": 7.899569034576416,
+      "learning_rate": 1.1823299069995106e-05,
+      "loss": 1.5607,
+      "step": 4250
+    },
+    {
+      "epoch": 9.416528925619835,
+      "grad_norm": 8.042000770568848,
+      "learning_rate": 1.1762114537444935e-05,
+      "loss": 1.4975,
+      "step": 4275
+    },
+    {
+      "epoch": 9.471625344352617,
+      "grad_norm": 12.677453994750977,
+      "learning_rate": 1.1700930004894763e-05,
+      "loss": 1.4892,
+      "step": 4300
+    },
+    {
+      "epoch": 9.5267217630854,
+      "grad_norm": 9.090930938720703,
+      "learning_rate": 1.1639745472344592e-05,
+      "loss": 1.5671,
+      "step": 4325
+    },
+    {
+      "epoch": 9.581818181818182,
+      "grad_norm": 18.41088104248047,
+      "learning_rate": 1.157856093979442e-05,
+      "loss": 1.4874,
+      "step": 4350
+    },
+    {
+      "epoch": 9.636914600550964,
+      "grad_norm": 9.162910461425781,
+      "learning_rate": 1.151737640724425e-05,
+      "loss": 1.4635,
+      "step": 4375
+    },
+    {
+      "epoch": 9.692011019283747,
+      "grad_norm": 10.90570068359375,
+      "learning_rate": 1.1456191874694078e-05,
+      "loss": 1.513,
+      "step": 4400
+    },
+    {
+      "epoch": 9.74710743801653,
+      "grad_norm": 10.234724044799805,
+      "learning_rate": 1.1395007342143907e-05,
+      "loss": 1.533,
+      "step": 4425
+    },
+    {
+      "epoch": 9.802203856749312,
+      "grad_norm": 6.85866641998291,
+      "learning_rate": 1.1333822809593735e-05,
+      "loss": 1.4384,
+      "step": 4450
+    },
+    {
+      "epoch": 9.857300275482094,
+      "grad_norm": 8.677523612976074,
+      "learning_rate": 1.1272638277043564e-05,
+      "loss": 1.5038,
+      "step": 4475
+    },
+    {
+      "epoch": 9.912396694214877,
+      "grad_norm": 7.9851460456848145,
+      "learning_rate": 1.1211453744493393e-05,
+      "loss": 1.4422,
+      "step": 4500
+    },
+    {
+      "epoch": 9.967493112947658,
+      "grad_norm": 7.877720832824707,
+      "learning_rate": 1.1150269211943221e-05,
+      "loss": 1.4556,
+      "step": 4525
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.08094948265368229,
+      "eval_loss": 0.72707200050354,
+      "eval_macro_f1": 0.5029402835068527,
+      "eval_micro_f1": 0.5317832764505119,
+      "eval_runtime": 0.7956,
+      "eval_samples_per_second": 2065.152,
+      "eval_steps_per_second": 129.465,
+      "eval_weighted_f1": 0.5432343330455452,
+      "step": 4540
+    },
+    {
+      "epoch": 10.022038567493112,
+      "grad_norm": 6.184263229370117,
+      "learning_rate": 1.108908467939305e-05,
+      "loss": 1.4583,
+      "step": 4550
+    },
+    {
+      "epoch": 10.077134986225895,
+      "grad_norm": 10.197240829467773,
+      "learning_rate": 1.1027900146842878e-05,
+      "loss": 1.4161,
+      "step": 4575
+    },
+    {
+      "epoch": 10.132231404958677,
+      "grad_norm": 12.444828033447266,
+      "learning_rate": 1.0966715614292707e-05,
+      "loss": 1.4309,
+      "step": 4600
+    },
+    {
+      "epoch": 10.18732782369146,
+      "grad_norm": 16.441001892089844,
+      "learning_rate": 1.0905531081742536e-05,
+      "loss": 1.388,
+      "step": 4625
+    },
+    {
+      "epoch": 10.242424242424242,
+      "grad_norm": 8.201318740844727,
+      "learning_rate": 1.0844346549192366e-05,
+      "loss": 1.4199,
+      "step": 4650
+    },
+    {
+      "epoch": 10.297520661157025,
+      "grad_norm": 13.441341400146484,
+      "learning_rate": 1.0783162016642195e-05,
+      "loss": 1.4134,
+      "step": 4675
+    },
+    {
+      "epoch": 10.352617079889807,
+      "grad_norm": 10.3322172164917,
+      "learning_rate": 1.0721977484092023e-05,
+      "loss": 1.3602,
+      "step": 4700
+    },
+    {
+      "epoch": 10.40771349862259,
+      "grad_norm": 5.628726482391357,
+      "learning_rate": 1.0660792951541852e-05,
+      "loss": 1.3711,
+      "step": 4725
+    },
+    {
+      "epoch": 10.462809917355372,
+      "grad_norm": 6.589449882507324,
+      "learning_rate": 1.059960841899168e-05,
+      "loss": 1.4131,
+      "step": 4750
+    },
+    {
+      "epoch": 10.517906336088155,
+      "grad_norm": 10.09038257598877,
+      "learning_rate": 1.053842388644151e-05,
+      "loss": 1.407,
+      "step": 4775
+    },
+    {
+      "epoch": 10.573002754820937,
+      "grad_norm": 16.832630157470703,
+      "learning_rate": 1.0477239353891338e-05,
+      "loss": 1.3365,
+      "step": 4800
+    },
+    {
+      "epoch": 10.62809917355372,
+      "grad_norm": 10.801997184753418,
+      "learning_rate": 1.0416054821341167e-05,
+      "loss": 1.3448,
+      "step": 4825
+    },
+    {
+      "epoch": 10.683195592286502,
+      "grad_norm": 10.650815963745117,
+      "learning_rate": 1.0354870288790995e-05,
+      "loss": 1.405,
+      "step": 4850
+    },
+    {
+      "epoch": 10.738292011019285,
+      "grad_norm": 9.310195922851562,
+      "learning_rate": 1.0293685756240824e-05,
+      "loss": 1.3109,
+      "step": 4875
+    },
+    {
+      "epoch": 10.793388429752067,
+      "grad_norm": 8.726702690124512,
+      "learning_rate": 1.0232501223690653e-05,
+      "loss": 1.3609,
+      "step": 4900
+    },
+    {
+      "epoch": 10.848484848484848,
+      "grad_norm": 8.540894508361816,
+      "learning_rate": 1.017131669114048e-05,
+      "loss": 1.3933,
+      "step": 4925
+    },
+    {
+      "epoch": 10.90358126721763,
+      "grad_norm": 9.035423278808594,
+      "learning_rate": 1.0110132158590308e-05,
+      "loss": 1.3052,
+      "step": 4950
+    },
+    {
+      "epoch": 10.958677685950413,
+      "grad_norm": 9.07214641571045,
+      "learning_rate": 1.0048947626040137e-05,
+      "loss": 1.4001,
+      "step": 4975
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.09007912355447352,
+      "eval_loss": 0.7351568341255188,
+      "eval_macro_f1": 0.5172892371402843,
+      "eval_micro_f1": 0.5423877327491785,
+      "eval_runtime": 0.786,
+      "eval_samples_per_second": 2090.312,
+      "eval_steps_per_second": 131.042,
+      "eval_weighted_f1": 0.5519542884665859,
+      "step": 4994
+    },
+    {
+      "epoch": 11.013223140495867,
+      "grad_norm": 5.78129768371582,
+      "learning_rate": 9.987763093489966e-06,
+      "loss": 1.32,
+      "step": 5000
+    },
+    {
+      "epoch": 11.06831955922865,
+      "grad_norm": 8.128321647644043,
+      "learning_rate": 9.926578560939796e-06,
+      "loss": 1.2663,
+      "step": 5025
+    },
+    {
+      "epoch": 11.123415977961432,
+      "grad_norm": 8.939966201782227,
+      "learning_rate": 9.865394028389625e-06,
+      "loss": 1.2988,
+      "step": 5050
+    },
+    {
+      "epoch": 11.178512396694215,
+      "grad_norm": 6.803643703460693,
+      "learning_rate": 9.804209495839453e-06,
+      "loss": 1.2649,
+      "step": 5075
+    },
+    {
+      "epoch": 11.233608815426997,
+      "grad_norm": 16.629180908203125,
+      "learning_rate": 9.743024963289282e-06,
+      "loss": 1.3324,
+      "step": 5100
+    },
+    {
+      "epoch": 11.28870523415978,
+      "grad_norm": 7.846168041229248,
+      "learning_rate": 9.68184043073911e-06,
+      "loss": 1.2661,
+      "step": 5125
+    },
+    {
+      "epoch": 11.343801652892562,
+      "grad_norm": 10.154409408569336,
+      "learning_rate": 9.620655898188939e-06,
+      "loss": 1.3156,
+      "step": 5150
+    },
+    {
+      "epoch": 11.398898071625345,
+      "grad_norm": 6.023170471191406,
+      "learning_rate": 9.559471365638768e-06,
+      "loss": 1.2855,
+      "step": 5175
+    },
+    {
+      "epoch": 11.453994490358127,
+      "grad_norm": 10.903104782104492,
+      "learning_rate": 9.498286833088596e-06,
+      "loss": 1.3085,
+      "step": 5200
+    },
+    {
+      "epoch": 11.50909090909091,
+      "grad_norm": 6.417316913604736,
+      "learning_rate": 9.437102300538425e-06,
+      "loss": 1.2589,
+      "step": 5225
+    },
+    {
+      "epoch": 11.564187327823692,
+      "grad_norm": 14.446656227111816,
+      "learning_rate": 9.375917767988254e-06,
+      "loss": 1.2417,
+      "step": 5250
+    },
+    {
+      "epoch": 11.619283746556473,
+      "grad_norm": 9.680697441101074,
+      "learning_rate": 9.31473323543808e-06,
+      "loss": 1.258,
+      "step": 5275
+    },
+    {
+      "epoch": 11.674380165289255,
+      "grad_norm": 7.470788478851318,
+      "learning_rate": 9.253548702887911e-06,
+      "loss": 1.2603,
+      "step": 5300
+    },
+    {
+      "epoch": 11.729476584022038,
+      "grad_norm": 6.185989856719971,
+      "learning_rate": 9.19236417033774e-06,
+      "loss": 1.2775,
+      "step": 5325
+    },
+    {
+      "epoch": 11.78457300275482,
+      "grad_norm": 6.857886791229248,
+      "learning_rate": 9.131179637787568e-06,
+      "loss": 1.2927,
+      "step": 5350
+    },
+    {
+      "epoch": 11.839669421487603,
+      "grad_norm": 8.878190040588379,
+      "learning_rate": 9.069995105237397e-06,
+      "loss": 1.2458,
+      "step": 5375
+    },
+    {
+      "epoch": 11.894765840220385,
+      "grad_norm": 12.2567720413208,
+      "learning_rate": 9.008810572687226e-06,
+      "loss": 1.2774,
+      "step": 5400
+    },
+    {
+      "epoch": 11.949862258953168,
+      "grad_norm": 7.592448711395264,
+      "learning_rate": 8.947626040137054e-06,
+      "loss": 1.274,
+      "step": 5425
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.10164333536214243,
+      "eval_loss": 0.7427167892456055,
+      "eval_macro_f1": 0.5173683757576815,
+      "eval_micro_f1": 0.5469133051036327,
+      "eval_runtime": 0.7544,
+      "eval_samples_per_second": 2177.917,
+      "eval_steps_per_second": 136.534,
+      "eval_weighted_f1": 0.5552655400400709,
+      "step": 5448
+    },
+    {
+      "epoch": 12.004407713498622,
+      "grad_norm": 8.367660522460938,
+      "learning_rate": 8.886441507586883e-06,
+      "loss": 1.2109,
+      "step": 5450
+    },
+    {
+      "epoch": 12.059504132231405,
+      "grad_norm": 146.01919555664062,
+      "learning_rate": 8.825256975036712e-06,
+      "loss": 1.1759,
+      "step": 5475
+    },
+    {
+      "epoch": 12.114600550964187,
+      "grad_norm": 5.678831100463867,
+      "learning_rate": 8.76407244248654e-06,
+      "loss": 1.2324,
+      "step": 5500
+    },
+    {
+      "epoch": 12.16969696969697,
+      "grad_norm": 6.2751851081848145,
+      "learning_rate": 8.702887909936369e-06,
+      "loss": 1.2089,
+      "step": 5525
+    },
+    {
+      "epoch": 12.224793388429752,
+      "grad_norm": 6.340836048126221,
+      "learning_rate": 8.641703377386198e-06,
+      "loss": 1.2303,
+      "step": 5550
+    },
+    {
+      "epoch": 12.279889807162535,
+      "grad_norm": 8.033232688903809,
+      "learning_rate": 8.580518844836026e-06,
+      "loss": 1.1842,
+      "step": 5575
+    },
+    {
+      "epoch": 12.334986225895317,
+      "grad_norm": 5.0165228843688965,
+      "learning_rate": 8.519334312285855e-06,
+      "loss": 1.1871,
+      "step": 5600
+    },
+    {
+      "epoch": 12.3900826446281,
+      "grad_norm": 6.247511863708496,
+      "learning_rate": 8.458149779735683e-06,
+      "loss": 1.1719,
+      "step": 5625
+    },
+    {
+      "epoch": 12.445179063360882,
+      "grad_norm": 6.809659957885742,
+      "learning_rate": 8.396965247185512e-06,
+      "loss": 1.174,
+      "step": 5650
+    },
+    {
+      "epoch": 12.500275482093663,
+      "grad_norm": 10.726896286010742,
+      "learning_rate": 8.33578071463534e-06,
+      "loss": 1.2047,
+      "step": 5675
+    },
+    {
+      "epoch": 12.555371900826445,
+      "grad_norm": 7.948911666870117,
+      "learning_rate": 8.27459618208517e-06,
+      "loss": 1.1879,
+      "step": 5700
+    },
+    {
+      "epoch": 12.610468319559228,
+      "grad_norm": 27.659639358520508,
+      "learning_rate": 8.213411649534998e-06,
+      "loss": 1.2368,
+      "step": 5725
+    },
+    {
+      "epoch": 12.66556473829201,
+      "grad_norm": 10.702486038208008,
+      "learning_rate": 8.152227116984827e-06,
+      "loss": 1.1614,
+      "step": 5750
+    },
+    {
+      "epoch": 12.720661157024793,
+      "grad_norm": 12.045475006103516,
+      "learning_rate": 8.091042584434655e-06,
+      "loss": 1.1367,
+      "step": 5775
+    },
+    {
+      "epoch": 12.775757575757575,
+      "grad_norm": 9.43398666381836,
+      "learning_rate": 8.029858051884484e-06,
+      "loss": 1.1774,
+      "step": 5800
+    },
+    {
+      "epoch": 12.830853994490358,
+      "grad_norm": 9.546597480773926,
+      "learning_rate": 7.968673519334313e-06,
+      "loss": 1.2087,
+      "step": 5825
+    },
+    {
+      "epoch": 12.88595041322314,
+      "grad_norm": 11.342320442199707,
+      "learning_rate": 7.907488986784141e-06,
+      "loss": 1.1969,
+      "step": 5850
+    },
+    {
+      "epoch": 12.941046831955923,
+      "grad_norm": 6.419647216796875,
+      "learning_rate": 7.84630445423397e-06,
+      "loss": 1.1185,
+      "step": 5875
+    },
+    {
+      "epoch": 12.996143250688705,
+      "grad_norm": 10.889259338378906,
+      "learning_rate": 7.785119921683799e-06,
+      "loss": 1.1474,
+      "step": 5900
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.09981740718198417,
+      "eval_loss": 0.7411310076713562,
+      "eval_macro_f1": 0.5188983413806106,
+      "eval_micro_f1": 0.5482582649212336,
+      "eval_runtime": 0.7318,
+      "eval_samples_per_second": 2245.254,
+      "eval_steps_per_second": 140.755,
+      "eval_weighted_f1": 0.5550465213529865,
+      "step": 5902
+    },
+    {
+      "epoch": 13.05068870523416,
+      "grad_norm": 10.158915519714355,
+      "learning_rate": 7.723935389133627e-06,
+      "loss": 1.1359,
+      "step": 5925
+    },
+    {
+      "epoch": 13.105785123966943,
+      "grad_norm": 9.29521369934082,
+      "learning_rate": 7.662750856583456e-06,
+      "loss": 1.1321,
+      "step": 5950
+    },
+    {
+      "epoch": 13.160881542699725,
+      "grad_norm": 10.788544654846191,
+      "learning_rate": 7.601566324033285e-06,
+      "loss": 1.1328,
+      "step": 5975
+    },
+    {
+      "epoch": 13.215977961432507,
+      "grad_norm": 9.468472480773926,
+      "learning_rate": 7.540381791483113e-06,
+      "loss": 1.119,
+      "step": 6000
+    },
+    {
+      "epoch": 13.27107438016529,
+      "grad_norm": 12.014350891113281,
+      "learning_rate": 7.479197258932942e-06,
+      "loss": 1.1089,
+      "step": 6025
+    },
+    {
+      "epoch": 13.326170798898072,
+      "grad_norm": 6.275262832641602,
+      "learning_rate": 7.4180127263827705e-06,
+      "loss": 1.1273,
+      "step": 6050
+    },
+    {
+      "epoch": 13.381267217630853,
+      "grad_norm": 6.748322010040283,
+      "learning_rate": 7.3568281938326e-06,
+      "loss": 1.1488,
+      "step": 6075
+    },
+    {
+      "epoch": 13.436363636363636,
+      "grad_norm": 9.268157958984375,
+      "learning_rate": 7.295643661282429e-06,
+      "loss": 1.0934,
+      "step": 6100
+    },
+    {
+      "epoch": 13.491460055096418,
+      "grad_norm": 9.821578025817871,
+      "learning_rate": 7.234459128732257e-06,
+      "loss": 1.1626,
+      "step": 6125
+    },
+    {
+      "epoch": 13.5465564738292,
+      "grad_norm": 16.561037063598633,
+      "learning_rate": 7.173274596182086e-06,
+      "loss": 1.124,
+      "step": 6150
+    },
+    {
+      "epoch": 13.601652892561983,
+      "grad_norm": 8.936065673828125,
+      "learning_rate": 7.112090063631915e-06,
+      "loss": 1.0531,
+      "step": 6175
+    },
+    {
+      "epoch": 13.656749311294766,
+      "grad_norm": 7.354989528656006,
+      "learning_rate": 7.050905531081743e-06,
+      "loss": 1.0947,
+      "step": 6200
+    },
+    {
+      "epoch": 13.711845730027548,
+      "grad_norm": 7.0242600440979,
+      "learning_rate": 6.989720998531572e-06,
+      "loss": 1.1123,
+      "step": 6225
+    },
+    {
+      "epoch": 13.76694214876033,
+      "grad_norm": 8.146479606628418,
+      "learning_rate": 6.928536465981401e-06,
+      "loss": 1.1017,
+      "step": 6250
+    },
+    {
+      "epoch": 13.822038567493113,
+      "grad_norm": 7.871755599975586,
+      "learning_rate": 6.867351933431229e-06,
+      "loss": 1.0755,
+      "step": 6275
+    },
+    {
+      "epoch": 13.877134986225895,
+      "grad_norm": 8.830574989318848,
+      "learning_rate": 6.806167400881057e-06,
+      "loss": 1.0948,
+      "step": 6300
+    },
+    {
+      "epoch": 13.932231404958678,
+      "grad_norm": 6.276432514190674,
+      "learning_rate": 6.744982868330886e-06,
+      "loss": 1.0978,
+      "step": 6325
+    },
+    {
+      "epoch": 13.98732782369146,
+      "grad_norm": 31.289966583251953,
+      "learning_rate": 6.683798335780716e-06,
+      "loss": 1.0909,
+      "step": 6350
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.11381618989653074,
+      "eval_loss": 0.7726384997367859,
+      "eval_macro_f1": 0.5301473406912293,
+      "eval_micro_f1": 0.5577033055015441,
+      "eval_runtime": 0.7668,
+      "eval_samples_per_second": 2142.769,
+      "eval_steps_per_second": 134.331,
+      "eval_weighted_f1": 0.5628237647563301,
+      "step": 6356
+    },
+    {
+      "epoch": 14.041873278236915,
+      "grad_norm": 9.511872291564941,
+      "learning_rate": 6.622613803230544e-06,
+      "loss": 1.0502,
+      "step": 6375
+    },
+    {
+      "epoch": 14.096969696969698,
+      "grad_norm": 8.232625007629395,
+      "learning_rate": 6.5614292706803725e-06,
+      "loss": 1.0125,
+      "step": 6400
+    },
+    {
+      "epoch": 14.15206611570248,
+      "grad_norm": 13.766243934631348,
+      "learning_rate": 6.500244738130201e-06,
+      "loss": 1.0749,
+      "step": 6425
+    },
+    {
+      "epoch": 14.207162534435263,
+      "grad_norm": 8.681984901428223,
+      "learning_rate": 6.43906020558003e-06,
+      "loss": 1.0245,
+      "step": 6450
+    },
+    {
+      "epoch": 14.262258953168043,
+      "grad_norm": 12.72724437713623,
+      "learning_rate": 6.3778756730298585e-06,
+      "loss": 1.0497,
+      "step": 6475
+    },
+    {
+      "epoch": 14.317355371900826,
+      "grad_norm": 10.159751892089844,
+      "learning_rate": 6.316691140479687e-06,
+      "loss": 1.0501,
+      "step": 6500
+    },
+    {
+      "epoch": 14.372451790633608,
+      "grad_norm": 8.054067611694336,
+      "learning_rate": 6.255506607929516e-06,
+      "loss": 1.0853,
+      "step": 6525
+    },
+    {
+      "epoch": 14.42754820936639,
+      "grad_norm": 9.144680976867676,
+      "learning_rate": 6.194322075379344e-06,
+      "loss": 1.0849,
+      "step": 6550
+    },
+    {
+      "epoch": 14.482644628099173,
+      "grad_norm": 7.8340020179748535,
+      "learning_rate": 6.133137542829173e-06,
+      "loss": 1.0418,
+      "step": 6575
+    },
+    {
+      "epoch": 14.537741046831956,
+      "grad_norm": 9.663851737976074,
+      "learning_rate": 6.071953010279002e-06,
+      "loss": 1.0477,
+      "step": 6600
+    },
+    {
+      "epoch": 14.592837465564738,
+      "grad_norm": 6.405959129333496,
+      "learning_rate": 6.010768477728831e-06,
+      "loss": 1.0559,
+      "step": 6625
+    },
+    {
+      "epoch": 14.64793388429752,
+      "grad_norm": 9.999762535095215,
+      "learning_rate": 5.94958394517866e-06,
+      "loss": 1.0371,
+      "step": 6650
+    },
+    {
+      "epoch": 14.703030303030303,
+      "grad_norm": 6.774183750152588,
+      "learning_rate": 5.8883994126284885e-06,
+      "loss": 1.0193,
+      "step": 6675
+    },
+    {
+      "epoch": 14.758126721763086,
+      "grad_norm": 13.159234046936035,
+      "learning_rate": 5.827214880078317e-06,
+      "loss": 1.0433,
+      "step": 6700
+    },
+    {
+      "epoch": 14.813223140495868,
+      "grad_norm": 12.729023933410645,
+      "learning_rate": 5.766030347528146e-06,
+      "loss": 1.0167,
+      "step": 6725
+    },
+    {
+      "epoch": 14.86831955922865,
+      "grad_norm": 15.757935523986816,
+      "learning_rate": 5.704845814977974e-06,
+      "loss": 1.0048,
+      "step": 6750
+    },
+    {
+      "epoch": 14.923415977961433,
+      "grad_norm": 25.768495559692383,
+      "learning_rate": 5.643661282427802e-06,
+      "loss": 1.0126,
+      "step": 6775
+    },
+    {
+      "epoch": 14.978512396694216,
+      "grad_norm": 8.229415893554688,
+      "learning_rate": 5.582476749877631e-06,
+      "loss": 1.0675,
+      "step": 6800
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.11929397443700548,
+      "eval_loss": 0.7604466080665588,
+      "eval_macro_f1": 0.5233293354606164,
+      "eval_micro_f1": 0.55444864307798,
+      "eval_runtime": 0.7401,
+      "eval_samples_per_second": 2220.088,
+      "eval_steps_per_second": 139.178,
+      "eval_weighted_f1": 0.5605802174257807,
+      "step": 6810
+    },
+    {
+      "epoch": 15.03305785123967,
+      "grad_norm": 7.657383441925049,
+      "learning_rate": 5.5212922173274596e-06,
+      "loss": 1.0134,
+      "step": 6825
+    },
+    {
+      "epoch": 15.088154269972453,
+      "grad_norm": 6.9480204582214355,
+      "learning_rate": 5.460107684777288e-06,
+      "loss": 0.9708,
+      "step": 6850
+    },
+    {
+      "epoch": 15.143250688705233,
+      "grad_norm": 4.3695387840271,
+      "learning_rate": 5.398923152227118e-06,
+      "loss": 0.9865,
+      "step": 6875
+    },
+    {
+      "epoch": 15.198347107438016,
+      "grad_norm": 6.225238800048828,
+      "learning_rate": 5.337738619676946e-06,
+      "loss": 1.0123,
+      "step": 6900
+    },
+    {
+      "epoch": 15.253443526170798,
+      "grad_norm": 11.188499450683594,
+      "learning_rate": 5.276554087126775e-06,
+      "loss": 0.9725,
+      "step": 6925
+    },
+    {
+      "epoch": 15.30853994490358,
+      "grad_norm": 36.433753967285156,
+      "learning_rate": 5.215369554576604e-06,
+      "loss": 0.9925,
+      "step": 6950
+    },
+    {
+      "epoch": 15.363636363636363,
+      "grad_norm": 7.442889213562012,
+      "learning_rate": 5.154185022026432e-06,
+      "loss": 1.0082,
+      "step": 6975
+    },
+    {
+      "epoch": 15.418732782369146,
+      "grad_norm": 16.42450714111328,
+      "learning_rate": 5.093000489476261e-06,
+      "loss": 1.0109,
+      "step": 7000
+    },
+    {
+      "epoch": 15.473829201101928,
+      "grad_norm": 7.358915328979492,
+      "learning_rate": 5.03181595692609e-06,
+      "loss": 0.9753,
+      "step": 7025
+    },
+    {
+      "epoch": 15.52892561983471,
+      "grad_norm": 17.838838577270508,
+      "learning_rate": 4.970631424375918e-06,
+      "loss": 0.979,
+      "step": 7050
+    },
+    {
+      "epoch": 15.584022038567493,
+      "grad_norm": 10.839521408081055,
+      "learning_rate": 4.909446891825747e-06,
+      "loss": 1.0254,
+      "step": 7075
+    },
+    {
+      "epoch": 15.639118457300276,
+      "grad_norm": 10.101546287536621,
+      "learning_rate": 4.8482623592755756e-06,
+      "loss": 1.0127,
+      "step": 7100
+    },
+    {
+      "epoch": 15.694214876033058,
+      "grad_norm": 10.862388610839844,
+      "learning_rate": 4.787077826725404e-06,
+      "loss": 0.998,
+      "step": 7125
+    },
+    {
+      "epoch": 15.74931129476584,
+      "grad_norm": 9.09751033782959,
+      "learning_rate": 4.725893294175233e-06,
+      "loss": 1.008,
+      "step": 7150
+    },
+    {
+      "epoch": 15.804407713498623,
+      "grad_norm": 6.354736328125,
+      "learning_rate": 4.6647087616250615e-06,
+      "loss": 0.9695,
+      "step": 7175
+    },
+    {
+      "epoch": 15.859504132231406,
+      "grad_norm": 8.806958198547363,
+      "learning_rate": 4.60352422907489e-06,
+      "loss": 0.9777,
+      "step": 7200
+    },
+    {
+      "epoch": 15.914600550964188,
+      "grad_norm": 7.194450378417969,
+      "learning_rate": 4.542339696524719e-06,
+      "loss": 0.9588,
+      "step": 7225
+    },
+    {
+      "epoch": 15.969696969696969,
+      "grad_norm": 5.09818172454834,
+      "learning_rate": 4.4811551639745475e-06,
+      "loss": 0.9835,
+      "step": 7250
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.13451004260499086,
+      "eval_loss": 0.7917519807815552,
+      "eval_macro_f1": 0.5387736977274954,
+      "eval_micro_f1": 0.5704451038575667,
+      "eval_runtime": 0.732,
+      "eval_samples_per_second": 2244.679,
+      "eval_steps_per_second": 140.719,
+      "eval_weighted_f1": 0.575498510225612,
+      "step": 7264
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 9080,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

ood-vigoemotions-phobert-source-source-01-s43-20260826T134812Z-1f9ffdf3/trainer/checkpoint-8626/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

ood-vigoemotions-phobert-source-source-01-s43-20260826T134812Z-1f9ffdf3/trainer/checkpoint-8626/special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-01-s43-20260826T134812Z-1f9ffdf3/trainer/checkpoint-8626/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 256,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}

ood-vigoemotions-phobert-source-source-01-s43-20260826T134812Z-1f9ffdf3/trainer/checkpoint-8626/trainer_state.json

@@ -0,0 +1,2686 @@
+{
+  "best_global_step": 7264,
+  "best_metric": 0.5387736977274954,
+  "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vigoemotions-phobert-source-source-01-s43-20260826T134812Z-1f9ffdf3/trainer/checkpoint-7264",
+  "epoch": 19.0,
+  "eval_steps": 500,
+  "global_step": 8626,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.05509641873278237,
+      "grad_norm": 4.721506595611572,
+      "learning_rate": 5.286343612334802e-07,
+      "loss": 5.1824,
+      "step": 25
+    },
+    {
+      "epoch": 0.11019283746556474,
+      "grad_norm": 4.242921352386475,
+      "learning_rate": 1.0792951541850223e-06,
+      "loss": 5.243,
+      "step": 50
+    },
+    {
+      "epoch": 0.1652892561983471,
+      "grad_norm": 5.161743640899658,
+      "learning_rate": 1.629955947136564e-06,
+      "loss": 5.2076,
+      "step": 75
+    },
+    {
+      "epoch": 0.22038567493112948,
+      "grad_norm": 5.828020095825195,
+      "learning_rate": 2.180616740088106e-06,
+      "loss": 5.0939,
+      "step": 100
+    },
+    {
+      "epoch": 0.27548209366391185,
+      "grad_norm": 5.841954231262207,
+      "learning_rate": 2.731277533039648e-06,
+      "loss": 5.0982,
+      "step": 125
+    },
+    {
+      "epoch": 0.3305785123966942,
+      "grad_norm": 3.9589149951934814,
+      "learning_rate": 3.2819383259911898e-06,
+      "loss": 5.1753,
+      "step": 150
+    },
+    {
+      "epoch": 0.3856749311294766,
+      "grad_norm": 4.0094428062438965,
+      "learning_rate": 3.8325991189427315e-06,
+      "loss": 5.1995,
+      "step": 175
+    },
+    {
+      "epoch": 0.44077134986225897,
+      "grad_norm": 4.752462387084961,
+      "learning_rate": 4.383259911894274e-06,
+      "loss": 5.2052,
+      "step": 200
+    },
+    {
+      "epoch": 0.49586776859504134,
+      "grad_norm": 7.853937149047852,
+      "learning_rate": 4.933920704845816e-06,
+      "loss": 5.0651,
+      "step": 225
+    },
+    {
+      "epoch": 0.5509641873278237,
+      "grad_norm": 6.576372146606445,
+      "learning_rate": 5.484581497797358e-06,
+      "loss": 5.0573,
+      "step": 250
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 5.6159844398498535,
+      "learning_rate": 6.035242290748899e-06,
+      "loss": 5.0438,
+      "step": 275
+    },
+    {
+      "epoch": 0.6611570247933884,
+      "grad_norm": 4.743258476257324,
+      "learning_rate": 6.585903083700441e-06,
+      "loss": 4.9259,
+      "step": 300
+    },
+    {
+      "epoch": 0.7162534435261708,
+      "grad_norm": 5.2631425857543945,
+      "learning_rate": 7.136563876651983e-06,
+      "loss": 4.8988,
+      "step": 325
+    },
+    {
+      "epoch": 0.7713498622589532,
+      "grad_norm": 5.224791049957275,
+      "learning_rate": 7.687224669603525e-06,
+      "loss": 4.8268,
+      "step": 350
+    },
+    {
+      "epoch": 0.8264462809917356,
+      "grad_norm": 5.436279773712158,
+      "learning_rate": 8.237885462555067e-06,
+      "loss": 4.6179,
+      "step": 375
+    },
+    {
+      "epoch": 0.8815426997245179,
+      "grad_norm": 6.9627909660339355,
+      "learning_rate": 8.788546255506607e-06,
+      "loss": 4.5771,
+      "step": 400
+    },
+    {
+      "epoch": 0.9366391184573003,
+      "grad_norm": 6.738986968994141,
+      "learning_rate": 9.339207048458151e-06,
+      "loss": 4.5826,
+      "step": 425
+    },
+    {
+      "epoch": 0.9917355371900827,
+      "grad_norm": 7.674351215362549,
+      "learning_rate": 9.889867841409693e-06,
+      "loss": 4.3028,
+      "step": 450
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 1.064934253692627,
+      "eval_macro_f1": 0.25079695642070166,
+      "eval_micro_f1": 0.26526126486162455,
+      "eval_runtime": 0.8081,
+      "eval_samples_per_second": 2033.215,
+      "eval_steps_per_second": 127.463,
+      "eval_weighted_f1": 0.32978528051651723,
+      "step": 454
+    },
+    {
+      "epoch": 1.0462809917355371,
+      "grad_norm": 5.639154434204102,
+      "learning_rate": 1.0440528634361234e-05,
+      "loss": 4.2781,
+      "step": 475
+    },
+    {
+      "epoch": 1.1013774104683196,
+      "grad_norm": 6.368907928466797,
+      "learning_rate": 1.0991189427312776e-05,
+      "loss": 4.1193,
+      "step": 500
+    },
+    {
+      "epoch": 1.1564738292011019,
+      "grad_norm": 11.263838768005371,
+      "learning_rate": 1.154185022026432e-05,
+      "loss": 4.2441,
+      "step": 525
+    },
+    {
+      "epoch": 1.2115702479338843,
+      "grad_norm": 5.8739824295043945,
+      "learning_rate": 1.209251101321586e-05,
+      "loss": 4.2407,
+      "step": 550
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 8.794628143310547,
+      "learning_rate": 1.2643171806167402e-05,
+      "loss": 4.1258,
+      "step": 575
+    },
+    {
+      "epoch": 1.321763085399449,
+      "grad_norm": 9.760412216186523,
+      "learning_rate": 1.3193832599118943e-05,
+      "loss": 4.105,
+      "step": 600
+    },
+    {
+      "epoch": 1.3768595041322313,
+      "grad_norm": 6.199836730957031,
+      "learning_rate": 1.3744493392070487e-05,
+      "loss": 4.0819,
+      "step": 625
+    },
+    {
+      "epoch": 1.4319559228650138,
+      "grad_norm": 11.421781539916992,
+      "learning_rate": 1.4295154185022027e-05,
+      "loss": 3.9663,
+      "step": 650
+    },
+    {
+      "epoch": 1.487052341597796,
+      "grad_norm": 11.619317054748535,
+      "learning_rate": 1.484581497797357e-05,
+      "loss": 3.9551,
+      "step": 675
+    },
+    {
+      "epoch": 1.5421487603305786,
+      "grad_norm": 8.50078010559082,
+      "learning_rate": 1.539647577092511e-05,
+      "loss": 3.9145,
+      "step": 700
+    },
+    {
+      "epoch": 1.5972451790633608,
+      "grad_norm": 9.922685623168945,
+      "learning_rate": 1.5947136563876654e-05,
+      "loss": 3.8852,
+      "step": 725
+    },
+    {
+      "epoch": 1.652341597796143,
+      "grad_norm": 8.09848690032959,
+      "learning_rate": 1.6497797356828196e-05,
+      "loss": 3.7545,
+      "step": 750
+    },
+    {
+      "epoch": 1.7074380165289256,
+      "grad_norm": 13.60912799835205,
+      "learning_rate": 1.7048458149779738e-05,
+      "loss": 3.7962,
+      "step": 775
+    },
+    {
+      "epoch": 1.762534435261708,
+      "grad_norm": 9.702221870422363,
+      "learning_rate": 1.759911894273128e-05,
+      "loss": 3.8082,
+      "step": 800
+    },
+    {
+      "epoch": 1.8176308539944903,
+      "grad_norm": 5.398869037628174,
+      "learning_rate": 1.814977973568282e-05,
+      "loss": 3.7219,
+      "step": 825
+    },
+    {
+      "epoch": 1.8727272727272726,
+      "grad_norm": 9.586057662963867,
+      "learning_rate": 1.8700440528634364e-05,
+      "loss": 3.6499,
+      "step": 850
+    },
+    {
+      "epoch": 1.927823691460055,
+      "grad_norm": 11.937478065490723,
+      "learning_rate": 1.9251101321585906e-05,
+      "loss": 3.7809,
+      "step": 875
+    },
+    {
+      "epoch": 1.9829201101928375,
+      "grad_norm": 9.817349433898926,
+      "learning_rate": 1.9801762114537445e-05,
+      "loss": 3.7224,
+      "step": 900
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.0,
+      "eval_loss": 0.8863740563392639,
+      "eval_macro_f1": 0.31581366295673935,
+      "eval_micro_f1": 0.3377152379189039,
+      "eval_runtime": 0.7837,
+      "eval_samples_per_second": 2096.4,
+      "eval_steps_per_second": 131.424,
+      "eval_weighted_f1": 0.3906050874186073,
+      "step": 908
+    },
+    {
+      "epoch": 2.037465564738292,
+      "grad_norm": 9.083897590637207,
+      "learning_rate": 1.9960841899167892e-05,
+      "loss": 3.5236,
+      "step": 925
+    },
+    {
+      "epoch": 2.0925619834710742,
+      "grad_norm": 7.97184944152832,
+      "learning_rate": 1.989965736661772e-05,
+      "loss": 3.3794,
+      "step": 950
+    },
+    {
+      "epoch": 2.1476584022038567,
+      "grad_norm": 9.009512901306152,
+      "learning_rate": 1.983847283406755e-05,
+      "loss": 3.5319,
+      "step": 975
+    },
+    {
+      "epoch": 2.202754820936639,
+      "grad_norm": 11.871821403503418,
+      "learning_rate": 1.9777288301517378e-05,
+      "loss": 3.4795,
+      "step": 1000
+    },
+    {
+      "epoch": 2.2578512396694217,
+      "grad_norm": 11.774299621582031,
+      "learning_rate": 1.9716103768967207e-05,
+      "loss": 3.4565,
+      "step": 1025
+    },
+    {
+      "epoch": 2.3129476584022037,
+      "grad_norm": 9.386924743652344,
+      "learning_rate": 1.9654919236417035e-05,
+      "loss": 3.3905,
+      "step": 1050
+    },
+    {
+      "epoch": 2.368044077134986,
+      "grad_norm": 7.399035930633545,
+      "learning_rate": 1.9593734703866864e-05,
+      "loss": 3.4593,
+      "step": 1075
+    },
+    {
+      "epoch": 2.4231404958677687,
+      "grad_norm": 9.72281265258789,
+      "learning_rate": 1.9532550171316693e-05,
+      "loss": 3.399,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4782369146005507,
+      "grad_norm": 11.944668769836426,
+      "learning_rate": 1.947136563876652e-05,
+      "loss": 3.309,
+      "step": 1125
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 10.8898286819458,
+      "learning_rate": 1.941018110621635e-05,
+      "loss": 3.3269,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5884297520661157,
+      "grad_norm": 12.417951583862305,
+      "learning_rate": 1.934899657366618e-05,
+      "loss": 3.4443,
+      "step": 1175
+    },
+    {
+      "epoch": 2.643526170798898,
+      "grad_norm": 7.977325916290283,
+      "learning_rate": 1.9287812041116007e-05,
+      "loss": 3.3148,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6986225895316807,
+      "grad_norm": 13.754631996154785,
+      "learning_rate": 1.9226627508565836e-05,
+      "loss": 3.2603,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7537190082644627,
+      "grad_norm": 7.910646915435791,
+      "learning_rate": 1.9165442976015665e-05,
+      "loss": 3.221,
+      "step": 1250
+    },
+    {
+      "epoch": 2.808815426997245,
+      "grad_norm": 8.343128204345703,
+      "learning_rate": 1.9104258443465493e-05,
+      "loss": 3.2729,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8639118457300277,
+      "grad_norm": 6.801973819732666,
+      "learning_rate": 1.9043073910915322e-05,
+      "loss": 3.2611,
+      "step": 1300
+    },
+    {
+      "epoch": 2.9190082644628097,
+      "grad_norm": 11.878301620483398,
+      "learning_rate": 1.898188937836515e-05,
+      "loss": 3.2498,
+      "step": 1325
+    },
+    {
+      "epoch": 2.974104683195592,
+      "grad_norm": 8.057741165161133,
+      "learning_rate": 1.892070484581498e-05,
+      "loss": 3.1971,
+      "step": 1350
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.0018259281801582471,
+      "eval_loss": 0.7955052852630615,
+      "eval_macro_f1": 0.371248723286375,
+      "eval_micro_f1": 0.3993803481853054,
+      "eval_runtime": 0.7772,
+      "eval_samples_per_second": 2114.059,
+      "eval_steps_per_second": 132.531,
+      "eval_weighted_f1": 0.4370260452300996,
+      "step": 1362
+    },
+    {
+      "epoch": 3.028650137741047,
+      "grad_norm": 9.659102439880371,
+      "learning_rate": 1.8859520313264808e-05,
+      "loss": 3.0296,
+      "step": 1375
+    },
+    {
+      "epoch": 3.0837465564738293,
+      "grad_norm": 6.875763416290283,
+      "learning_rate": 1.8798335780714636e-05,
+      "loss": 2.974,
+      "step": 1400
+    },
+    {
+      "epoch": 3.1388429752066114,
+      "grad_norm": 15.656290054321289,
+      "learning_rate": 1.8737151248164465e-05,
+      "loss": 3.071,
+      "step": 1425
+    },
+    {
+      "epoch": 3.193939393939394,
+      "grad_norm": 13.179049491882324,
+      "learning_rate": 1.8675966715614294e-05,
+      "loss": 2.9726,
+      "step": 1450
+    },
+    {
+      "epoch": 3.2490358126721763,
+      "grad_norm": 7.199787139892578,
+      "learning_rate": 1.8614782183064122e-05,
+      "loss": 2.9661,
+      "step": 1475
+    },
+    {
+      "epoch": 3.304132231404959,
+      "grad_norm": 6.702415466308594,
+      "learning_rate": 1.855359765051395e-05,
+      "loss": 2.9899,
+      "step": 1500
+    },
+    {
+      "epoch": 3.359228650137741,
+      "grad_norm": 7.977599620819092,
+      "learning_rate": 1.849241311796378e-05,
+      "loss": 2.9054,
+      "step": 1525
+    },
+    {
+      "epoch": 3.4143250688705233,
+      "grad_norm": 8.159268379211426,
+      "learning_rate": 1.843122858541361e-05,
+      "loss": 2.9468,
+      "step": 1550
+    },
+    {
+      "epoch": 3.469421487603306,
+      "grad_norm": 12.21645450592041,
+      "learning_rate": 1.8370044052863437e-05,
+      "loss": 2.9158,
+      "step": 1575
+    },
+    {
+      "epoch": 3.5245179063360883,
+      "grad_norm": 9.832043647766113,
+      "learning_rate": 1.8308859520313266e-05,
+      "loss": 2.9598,
+      "step": 1600
+    },
+    {
+      "epoch": 3.579614325068871,
+      "grad_norm": 9.701088905334473,
+      "learning_rate": 1.8247674987763094e-05,
+      "loss": 2.8472,
+      "step": 1625
+    },
+    {
+      "epoch": 3.634710743801653,
+      "grad_norm": 13.512517929077148,
+      "learning_rate": 1.8186490455212923e-05,
+      "loss": 2.9392,
+      "step": 1650
+    },
+    {
+      "epoch": 3.6898071625344353,
+      "grad_norm": 14.005834579467773,
+      "learning_rate": 1.812530592266275e-05,
+      "loss": 2.8569,
+      "step": 1675
+    },
+    {
+      "epoch": 3.744903581267218,
+      "grad_norm": 11.314921379089355,
+      "learning_rate": 1.806412139011258e-05,
+      "loss": 2.8256,
+      "step": 1700
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 9.094903945922852,
+      "learning_rate": 1.800293685756241e-05,
+      "loss": 2.8047,
+      "step": 1725
+    },
+    {
+      "epoch": 3.8550964187327823,
+      "grad_norm": 7.849555969238281,
+      "learning_rate": 1.7941752325012238e-05,
+      "loss": 2.8377,
+      "step": 1750
+    },
+    {
+      "epoch": 3.910192837465565,
+      "grad_norm": 13.487672805786133,
+      "learning_rate": 1.7880567792462066e-05,
+      "loss": 2.8002,
+      "step": 1775
+    },
+    {
+      "epoch": 3.9652892561983473,
+      "grad_norm": 21.90669822692871,
+      "learning_rate": 1.7819383259911895e-05,
+      "loss": 2.7817,
+      "step": 1800
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.010346926354230066,
+      "eval_loss": 0.750116765499115,
+      "eval_macro_f1": 0.39218802795663554,
+      "eval_micro_f1": 0.42694649150913166,
+      "eval_runtime": 0.833,
+      "eval_samples_per_second": 1972.37,
+      "eval_steps_per_second": 123.648,
+      "eval_weighted_f1": 0.45943726636348214,
+      "step": 1816
+    },
+    {
+      "epoch": 4.019834710743802,
+      "grad_norm": 10.072190284729004,
+      "learning_rate": 1.7758198727361724e-05,
+      "loss": 2.6486,
+      "step": 1825
+    },
+    {
+      "epoch": 4.074931129476584,
+      "grad_norm": 6.724724292755127,
+      "learning_rate": 1.7697014194811552e-05,
+      "loss": 2.5753,
+      "step": 1850
+    },
+    {
+      "epoch": 4.130027548209366,
+      "grad_norm": 14.258201599121094,
+      "learning_rate": 1.763582966226138e-05,
+      "loss": 2.5727,
+      "step": 1875
+    },
+    {
+      "epoch": 4.1851239669421485,
+      "grad_norm": 8.002503395080566,
+      "learning_rate": 1.757464512971121e-05,
+      "loss": 2.5374,
+      "step": 1900
+    },
+    {
+      "epoch": 4.240220385674931,
+      "grad_norm": 7.770499229431152,
+      "learning_rate": 1.7513460597161038e-05,
+      "loss": 2.586,
+      "step": 1925
+    },
+    {
+      "epoch": 4.2953168044077135,
+      "grad_norm": 8.485304832458496,
+      "learning_rate": 1.7452276064610867e-05,
+      "loss": 2.6794,
+      "step": 1950
+    },
+    {
+      "epoch": 4.350413223140496,
+      "grad_norm": 12.671653747558594,
+      "learning_rate": 1.7391091532060695e-05,
+      "loss": 2.6331,
+      "step": 1975
+    },
+    {
+      "epoch": 4.405509641873278,
+      "grad_norm": 10.223567008972168,
+      "learning_rate": 1.7329906999510527e-05,
+      "loss": 2.5868,
+      "step": 2000
+    },
+    {
+      "epoch": 4.460606060606061,
+      "grad_norm": 14.223559379577637,
+      "learning_rate": 1.7268722466960356e-05,
+      "loss": 2.5251,
+      "step": 2025
+    },
+    {
+      "epoch": 4.515702479338843,
+      "grad_norm": 8.538398742675781,
+      "learning_rate": 1.7207537934410185e-05,
+      "loss": 2.5315,
+      "step": 2050
+    },
+    {
+      "epoch": 4.570798898071625,
+      "grad_norm": 5.16135311126709,
+      "learning_rate": 1.7146353401860013e-05,
+      "loss": 2.5609,
+      "step": 2075
+    },
+    {
+      "epoch": 4.6258953168044075,
+      "grad_norm": 11.051167488098145,
+      "learning_rate": 1.7085168869309842e-05,
+      "loss": 2.4787,
+      "step": 2100
+    },
+    {
+      "epoch": 4.68099173553719,
+      "grad_norm": 9.579821586608887,
+      "learning_rate": 1.702398433675967e-05,
+      "loss": 2.6092,
+      "step": 2125
+    },
+    {
+      "epoch": 4.736088154269972,
+      "grad_norm": 7.40781831741333,
+      "learning_rate": 1.69627998042095e-05,
+      "loss": 2.5947,
+      "step": 2150
+    },
+    {
+      "epoch": 4.791184573002755,
+      "grad_norm": 6.301372528076172,
+      "learning_rate": 1.6901615271659325e-05,
+      "loss": 2.4604,
+      "step": 2175
+    },
+    {
+      "epoch": 4.846280991735537,
+      "grad_norm": 7.67243766784668,
+      "learning_rate": 1.6840430739109153e-05,
+      "loss": 2.5283,
+      "step": 2200
+    },
+    {
+      "epoch": 4.90137741046832,
+      "grad_norm": 6.105996608734131,
+      "learning_rate": 1.6779246206558982e-05,
+      "loss": 2.4024,
+      "step": 2225
+    },
+    {
+      "epoch": 4.9564738292011015,
+      "grad_norm": 13.376340866088867,
+      "learning_rate": 1.671806167400881e-05,
+      "loss": 2.481,
+      "step": 2250
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.021911138161898967,
+      "eval_loss": 0.7181954979896545,
+      "eval_macro_f1": 0.40061912618913414,
+      "eval_micro_f1": 0.4357821089455272,
+      "eval_runtime": 0.7813,
+      "eval_samples_per_second": 2102.788,
+      "eval_steps_per_second": 131.824,
+      "eval_weighted_f1": 0.4673903225695518,
+      "step": 2270
+    },
+    {
+      "epoch": 5.011019283746556,
+      "grad_norm": 8.279224395751953,
+      "learning_rate": 1.665687714145864e-05,
+      "loss": 2.4698,
+      "step": 2275
+    },
+    {
+      "epoch": 5.066115702479339,
+      "grad_norm": 7.251194477081299,
+      "learning_rate": 1.6595692608908468e-05,
+      "loss": 2.3897,
+      "step": 2300
+    },
+    {
+      "epoch": 5.121212121212121,
+      "grad_norm": 11.917134284973145,
+      "learning_rate": 1.6534508076358297e-05,
+      "loss": 2.3596,
+      "step": 2325
+    },
+    {
+      "epoch": 5.176308539944904,
+      "grad_norm": 11.278261184692383,
+      "learning_rate": 1.6473323543808125e-05,
+      "loss": 2.2631,
+      "step": 2350
+    },
+    {
+      "epoch": 5.231404958677686,
+      "grad_norm": 8.443638801574707,
+      "learning_rate": 1.6412139011257954e-05,
+      "loss": 2.1933,
+      "step": 2375
+    },
+    {
+      "epoch": 5.2865013774104685,
+      "grad_norm": 9.477250099182129,
+      "learning_rate": 1.6350954478707782e-05,
+      "loss": 2.3299,
+      "step": 2400
+    },
+    {
+      "epoch": 5.341597796143251,
+      "grad_norm": 5.243191242218018,
+      "learning_rate": 1.628976994615761e-05,
+      "loss": 2.29,
+      "step": 2425
+    },
+    {
+      "epoch": 5.3966942148760335,
+      "grad_norm": 10.38256549835205,
+      "learning_rate": 1.622858541360744e-05,
+      "loss": 2.3002,
+      "step": 2450
+    },
+    {
+      "epoch": 5.451790633608815,
+      "grad_norm": 10.228897094726562,
+      "learning_rate": 1.616740088105727e-05,
+      "loss": 2.3167,
+      "step": 2475
+    },
+    {
+      "epoch": 5.506887052341598,
+      "grad_norm": 13.564765930175781,
+      "learning_rate": 1.6106216348507097e-05,
+      "loss": 2.2538,
+      "step": 2500
+    },
+    {
+      "epoch": 5.56198347107438,
+      "grad_norm": 13.028368949890137,
+      "learning_rate": 1.604503181595693e-05,
+      "loss": 2.2295,
+      "step": 2525
+    },
+    {
+      "epoch": 5.6170798898071626,
+      "grad_norm": 11.13936710357666,
+      "learning_rate": 1.5983847283406758e-05,
+      "loss": 2.2291,
+      "step": 2550
+    },
+    {
+      "epoch": 5.672176308539945,
+      "grad_norm": 12.807024002075195,
+      "learning_rate": 1.5922662750856586e-05,
+      "loss": 2.2607,
+      "step": 2575
+    },
+    {
+      "epoch": 5.7272727272727275,
+      "grad_norm": 12.083641052246094,
+      "learning_rate": 1.5861478218306415e-05,
+      "loss": 2.2267,
+      "step": 2600
+    },
+    {
+      "epoch": 5.78236914600551,
+      "grad_norm": 6.688780784606934,
+      "learning_rate": 1.5800293685756244e-05,
+      "loss": 2.197,
+      "step": 2625
+    },
+    {
+      "epoch": 5.837465564738292,
+      "grad_norm": 10.367467880249023,
+      "learning_rate": 1.5739109153206072e-05,
+      "loss": 2.2441,
+      "step": 2650
+    },
+    {
+      "epoch": 5.892561983471074,
+      "grad_norm": 8.155801773071289,
+      "learning_rate": 1.56779246206559e-05,
+      "loss": 2.18,
+      "step": 2675
+    },
+    {
+      "epoch": 5.9476584022038566,
+      "grad_norm": 15.422148704528809,
+      "learning_rate": 1.561674008810573e-05,
+      "loss": 2.2327,
+      "step": 2700
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.03408399269628728,
+      "eval_loss": 0.7054561376571655,
+      "eval_macro_f1": 0.4520903933181108,
+      "eval_micro_f1": 0.4842243235652092,
+      "eval_runtime": 0.7411,
+      "eval_samples_per_second": 2216.967,
+      "eval_steps_per_second": 138.982,
+      "eval_weighted_f1": 0.503275021503218,
+      "step": 2724
+    },
+    {
+      "epoch": 6.002203856749311,
+      "grad_norm": 8.507722854614258,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 2.1608,
+      "step": 2725
+    },
+    {
+      "epoch": 6.057300275482094,
+      "grad_norm": 11.379617691040039,
+      "learning_rate": 1.5494371023005387e-05,
+      "loss": 2.0387,
+      "step": 2750
+    },
+    {
+      "epoch": 6.112396694214876,
+      "grad_norm": 8.296934127807617,
+      "learning_rate": 1.5433186490455216e-05,
+      "loss": 2.1263,
+      "step": 2775
+    },
+    {
+      "epoch": 6.167493112947659,
+      "grad_norm": 9.881864547729492,
+      "learning_rate": 1.5372001957905044e-05,
+      "loss": 2.0996,
+      "step": 2800
+    },
+    {
+      "epoch": 6.222589531680441,
+      "grad_norm": 15.106012344360352,
+      "learning_rate": 1.5310817425354873e-05,
+      "loss": 2.0111,
+      "step": 2825
+    },
+    {
+      "epoch": 6.277685950413223,
+      "grad_norm": 10.103724479675293,
+      "learning_rate": 1.52496328928047e-05,
+      "loss": 2.0883,
+      "step": 2850
+    },
+    {
+      "epoch": 6.332782369146005,
+      "grad_norm": 9.56401538848877,
+      "learning_rate": 1.5188448360254529e-05,
+      "loss": 2.0111,
+      "step": 2875
+    },
+    {
+      "epoch": 6.387878787878788,
+      "grad_norm": 7.425657749176025,
+      "learning_rate": 1.5127263827704357e-05,
+      "loss": 2.0236,
+      "step": 2900
+    },
+    {
+      "epoch": 6.44297520661157,
+      "grad_norm": 10.291604042053223,
+      "learning_rate": 1.5066079295154186e-05,
+      "loss": 2.0612,
+      "step": 2925
+    },
+    {
+      "epoch": 6.498071625344353,
+      "grad_norm": 9.244171142578125,
+      "learning_rate": 1.5004894762604014e-05,
+      "loss": 1.9911,
+      "step": 2950
+    },
+    {
+      "epoch": 6.553168044077135,
+      "grad_norm": 10.03786563873291,
+      "learning_rate": 1.4943710230053843e-05,
+      "loss": 2.0577,
+      "step": 2975
+    },
+    {
+      "epoch": 6.608264462809918,
+      "grad_norm": 7.586009502410889,
+      "learning_rate": 1.4882525697503672e-05,
+      "loss": 2.0063,
+      "step": 3000
+    },
+    {
+      "epoch": 6.663360881542699,
+      "grad_norm": 11.507920265197754,
+      "learning_rate": 1.48213411649535e-05,
+      "loss": 1.9918,
+      "step": 3025
+    },
+    {
+      "epoch": 6.718457300275482,
+      "grad_norm": 10.177037239074707,
+      "learning_rate": 1.476015663240333e-05,
+      "loss": 1.989,
+      "step": 3050
+    },
+    {
+      "epoch": 6.773553719008264,
+      "grad_norm": 11.146783828735352,
+      "learning_rate": 1.469897209985316e-05,
+      "loss": 1.9675,
+      "step": 3075
+    },
+    {
+      "epoch": 6.828650137741047,
+      "grad_norm": 23.576797485351562,
+      "learning_rate": 1.4637787567302988e-05,
+      "loss": 2.066,
+      "step": 3100
+    },
+    {
+      "epoch": 6.883746556473829,
+      "grad_norm": 6.994001865386963,
+      "learning_rate": 1.4576603034752817e-05,
+      "loss": 1.9676,
+      "step": 3125
+    },
+    {
+      "epoch": 6.938842975206612,
+      "grad_norm": 12.981430053710938,
+      "learning_rate": 1.4515418502202645e-05,
+      "loss": 1.951,
+      "step": 3150
+    },
+    {
+      "epoch": 6.993939393939394,
+      "grad_norm": 6.651545524597168,
+      "learning_rate": 1.4454233969652474e-05,
+      "loss": 1.9298,
+      "step": 3175
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.054777845404747415,
+      "eval_loss": 0.7065615653991699,
+      "eval_macro_f1": 0.47539287952540293,
+      "eval_micro_f1": 0.5047904191616767,
+      "eval_runtime": 0.7799,
+      "eval_samples_per_second": 2106.647,
+      "eval_steps_per_second": 132.066,
+      "eval_weighted_f1": 0.5200095920230331,
+      "step": 3178
+    },
+    {
+      "epoch": 7.048484848484849,
+      "grad_norm": 8.36910629272461,
+      "learning_rate": 1.4393049437102303e-05,
+      "loss": 1.8839,
+      "step": 3200
+    },
+    {
+      "epoch": 7.103581267217631,
+      "grad_norm": 11.923949241638184,
+      "learning_rate": 1.4331864904552131e-05,
+      "loss": 1.8423,
+      "step": 3225
+    },
+    {
+      "epoch": 7.158677685950413,
+      "grad_norm": 9.617166519165039,
+      "learning_rate": 1.427068037200196e-05,
+      "loss": 1.8561,
+      "step": 3250
+    },
+    {
+      "epoch": 7.213774104683195,
+      "grad_norm": 16.226926803588867,
+      "learning_rate": 1.4209495839451789e-05,
+      "loss": 1.8183,
+      "step": 3275
+    },
+    {
+      "epoch": 7.268870523415978,
+      "grad_norm": 12.089911460876465,
+      "learning_rate": 1.4148311306901617e-05,
+      "loss": 1.8012,
+      "step": 3300
+    },
+    {
+      "epoch": 7.32396694214876,
+      "grad_norm": 8.376047134399414,
+      "learning_rate": 1.4087126774351444e-05,
+      "loss": 1.8334,
+      "step": 3325
+    },
+    {
+      "epoch": 7.379063360881543,
+      "grad_norm": 13.260794639587402,
+      "learning_rate": 1.4025942241801273e-05,
+      "loss": 1.8614,
+      "step": 3350
+    },
+    {
+      "epoch": 7.434159779614325,
+      "grad_norm": 9.980430603027344,
+      "learning_rate": 1.3964757709251102e-05,
+      "loss": 1.8474,
+      "step": 3375
+    },
+    {
+      "epoch": 7.489256198347108,
+      "grad_norm": 7.031350135803223,
+      "learning_rate": 1.390357317670093e-05,
+      "loss": 1.838,
+      "step": 3400
+    },
+    {
+      "epoch": 7.544352617079889,
+      "grad_norm": 12.172435760498047,
+      "learning_rate": 1.3842388644150759e-05,
+      "loss": 1.8082,
+      "step": 3425
+    },
+    {
+      "epoch": 7.599449035812672,
+      "grad_norm": 10.703603744506836,
+      "learning_rate": 1.3781204111600588e-05,
+      "loss": 1.8367,
+      "step": 3450
+    },
+    {
+      "epoch": 7.654545454545454,
+      "grad_norm": 11.197491645812988,
+      "learning_rate": 1.3720019579050416e-05,
+      "loss": 1.7869,
+      "step": 3475
+    },
+    {
+      "epoch": 7.709641873278237,
+      "grad_norm": 23.729402542114258,
+      "learning_rate": 1.3658835046500245e-05,
+      "loss": 1.7855,
+      "step": 3500
+    },
+    {
+      "epoch": 7.764738292011019,
+      "grad_norm": 8.929984092712402,
+      "learning_rate": 1.3597650513950073e-05,
+      "loss": 1.8359,
+      "step": 3525
+    },
+    {
+      "epoch": 7.819834710743802,
+      "grad_norm": 17.336320877075195,
+      "learning_rate": 1.3536465981399902e-05,
+      "loss": 1.7919,
+      "step": 3550
+    },
+    {
+      "epoch": 7.874931129476584,
+      "grad_norm": 9.479592323303223,
+      "learning_rate": 1.3475281448849732e-05,
+      "loss": 1.7993,
+      "step": 3575
+    },
+    {
+      "epoch": 7.930027548209367,
+      "grad_norm": 8.46316909790039,
+      "learning_rate": 1.3414096916299561e-05,
+      "loss": 1.7886,
+      "step": 3600
+    },
+    {
+      "epoch": 7.985123966942149,
+      "grad_norm": 6.84007453918457,
+      "learning_rate": 1.335291238374939e-05,
+      "loss": 1.8102,
+      "step": 3625
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.06451612903225806,
+      "eval_loss": 0.6984469294548035,
+      "eval_macro_f1": 0.482549048079171,
+      "eval_micro_f1": 0.5128887070376432,
+      "eval_runtime": 0.7756,
+      "eval_samples_per_second": 2118.326,
+      "eval_steps_per_second": 132.798,
+      "eval_weighted_f1": 0.5248092674046481,
+      "step": 3632
+    },
+    {
+      "epoch": 8.039669421487604,
+      "grad_norm": 16.53033447265625,
+      "learning_rate": 1.3291727851199218e-05,
+      "loss": 1.7326,
+      "step": 3650
+    },
+    {
+      "epoch": 8.094765840220386,
+      "grad_norm": 7.538041114807129,
+      "learning_rate": 1.3230543318649047e-05,
+      "loss": 1.7154,
+      "step": 3675
+    },
+    {
+      "epoch": 8.149862258953169,
+      "grad_norm": 7.187880992889404,
+      "learning_rate": 1.3169358786098876e-05,
+      "loss": 1.6261,
+      "step": 3700
+    },
+    {
+      "epoch": 8.20495867768595,
+      "grad_norm": 6.9718403816223145,
+      "learning_rate": 1.3108174253548704e-05,
+      "loss": 1.698,
+      "step": 3725
+    },
+    {
+      "epoch": 8.260055096418732,
+      "grad_norm": 11.223376274108887,
+      "learning_rate": 1.3046989720998533e-05,
+      "loss": 1.6221,
+      "step": 3750
+    },
+    {
+      "epoch": 8.315151515151515,
+      "grad_norm": 9.946511268615723,
+      "learning_rate": 1.2985805188448362e-05,
+      "loss": 1.5966,
+      "step": 3775
+    },
+    {
+      "epoch": 8.370247933884297,
+      "grad_norm": 10.008794784545898,
+      "learning_rate": 1.292462065589819e-05,
+      "loss": 1.6861,
+      "step": 3800
+    },
+    {
+      "epoch": 8.42534435261708,
+      "grad_norm": 13.9457368850708,
+      "learning_rate": 1.2863436123348019e-05,
+      "loss": 1.6792,
+      "step": 3825
+    },
+    {
+      "epoch": 8.480440771349862,
+      "grad_norm": 7.465098857879639,
+      "learning_rate": 1.2802251590797848e-05,
+      "loss": 1.6933,
+      "step": 3850
+    },
+    {
+      "epoch": 8.535537190082644,
+      "grad_norm": 9.797990798950195,
+      "learning_rate": 1.2741067058247676e-05,
+      "loss": 1.673,
+      "step": 3875
+    },
+    {
+      "epoch": 8.590633608815427,
+      "grad_norm": 11.942303657531738,
+      "learning_rate": 1.2679882525697505e-05,
+      "loss": 1.643,
+      "step": 3900
+    },
+    {
+      "epoch": 8.64573002754821,
+      "grad_norm": 6.5281662940979,
+      "learning_rate": 1.2618697993147334e-05,
+      "loss": 1.6283,
+      "step": 3925
+    },
+    {
+      "epoch": 8.700826446280992,
+      "grad_norm": 7.67686653137207,
+      "learning_rate": 1.2557513460597162e-05,
+      "loss": 1.6683,
+      "step": 3950
+    },
+    {
+      "epoch": 8.755922865013774,
+      "grad_norm": 10.89754867553711,
+      "learning_rate": 1.2496328928046991e-05,
+      "loss": 1.6098,
+      "step": 3975
+    },
+    {
+      "epoch": 8.811019283746557,
+      "grad_norm": 7.0877685546875,
+      "learning_rate": 1.243514439549682e-05,
+      "loss": 1.6264,
+      "step": 4000
+    },
+    {
+      "epoch": 8.86611570247934,
+      "grad_norm": 6.110809803009033,
+      "learning_rate": 1.2373959862946646e-05,
+      "loss": 1.6501,
+      "step": 4025
+    },
+    {
+      "epoch": 8.921212121212122,
+      "grad_norm": 15.987994194030762,
+      "learning_rate": 1.2312775330396475e-05,
+      "loss": 1.6452,
+      "step": 4050
+    },
+    {
+      "epoch": 8.976308539944904,
+      "grad_norm": 9.162452697753906,
+      "learning_rate": 1.2251590797846304e-05,
+      "loss": 1.5748,
+      "step": 4075
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.054777845404747415,
+      "eval_loss": 0.6941969990730286,
+      "eval_macro_f1": 0.47838987201665556,
+      "eval_micro_f1": 0.5094188376753507,
+      "eval_runtime": 0.7496,
+      "eval_samples_per_second": 2191.758,
+      "eval_steps_per_second": 137.402,
+      "eval_weighted_f1": 0.5218786373886054,
+      "step": 4086
+    },
+    {
+      "epoch": 9.030853994490359,
+      "grad_norm": 7.348278999328613,
+      "learning_rate": 1.2190406265296136e-05,
+      "loss": 1.5912,
+      "step": 4100
+    },
+    {
+      "epoch": 9.08595041322314,
+      "grad_norm": 13.46335506439209,
+      "learning_rate": 1.2129221732745964e-05,
+      "loss": 1.5393,
+      "step": 4125
+    },
+    {
+      "epoch": 9.141046831955922,
+      "grad_norm": 12.748778343200684,
+      "learning_rate": 1.2068037200195793e-05,
+      "loss": 1.5056,
+      "step": 4150
+    },
+    {
+      "epoch": 9.196143250688705,
+      "grad_norm": 6.756908893585205,
+      "learning_rate": 1.2006852667645622e-05,
+      "loss": 1.4559,
+      "step": 4175
+    },
+    {
+      "epoch": 9.251239669421487,
+      "grad_norm": 14.214062690734863,
+      "learning_rate": 1.1945668135095449e-05,
+      "loss": 1.4885,
+      "step": 4200
+    },
+    {
+      "epoch": 9.30633608815427,
+      "grad_norm": 13.800779342651367,
+      "learning_rate": 1.1884483602545277e-05,
+      "loss": 1.5808,
+      "step": 4225
+    },
+    {
+      "epoch": 9.361432506887052,
+      "grad_norm": 7.899569034576416,
+      "learning_rate": 1.1823299069995106e-05,
+      "loss": 1.5607,
+      "step": 4250
+    },
+    {
+      "epoch": 9.416528925619835,
+      "grad_norm": 8.042000770568848,
+      "learning_rate": 1.1762114537444935e-05,
+      "loss": 1.4975,
+      "step": 4275
+    },
+    {
+      "epoch": 9.471625344352617,
+      "grad_norm": 12.677453994750977,
+      "learning_rate": 1.1700930004894763e-05,
+      "loss": 1.4892,
+      "step": 4300
+    },
+    {
+      "epoch": 9.5267217630854,
+      "grad_norm": 9.090930938720703,
+      "learning_rate": 1.1639745472344592e-05,
+      "loss": 1.5671,
+      "step": 4325
+    },
+    {
+      "epoch": 9.581818181818182,
+      "grad_norm": 18.41088104248047,
+      "learning_rate": 1.157856093979442e-05,
+      "loss": 1.4874,
+      "step": 4350
+    },
+    {
+      "epoch": 9.636914600550964,
+      "grad_norm": 9.162910461425781,
+      "learning_rate": 1.151737640724425e-05,
+      "loss": 1.4635,
+      "step": 4375
+    },
+    {
+      "epoch": 9.692011019283747,
+      "grad_norm": 10.90570068359375,
+      "learning_rate": 1.1456191874694078e-05,
+      "loss": 1.513,
+      "step": 4400
+    },
+    {
+      "epoch": 9.74710743801653,
+      "grad_norm": 10.234724044799805,
+      "learning_rate": 1.1395007342143907e-05,
+      "loss": 1.533,
+      "step": 4425
+    },
+    {
+      "epoch": 9.802203856749312,
+      "grad_norm": 6.85866641998291,
+      "learning_rate": 1.1333822809593735e-05,
+      "loss": 1.4384,
+      "step": 4450
+    },
+    {
+      "epoch": 9.857300275482094,
+      "grad_norm": 8.677523612976074,
+      "learning_rate": 1.1272638277043564e-05,
+      "loss": 1.5038,
+      "step": 4475
+    },
+    {
+      "epoch": 9.912396694214877,
+      "grad_norm": 7.9851460456848145,
+      "learning_rate": 1.1211453744493393e-05,
+      "loss": 1.4422,
+      "step": 4500
+    },
+    {
+      "epoch": 9.967493112947658,
+      "grad_norm": 7.877720832824707,
+      "learning_rate": 1.1150269211943221e-05,
+      "loss": 1.4556,
+      "step": 4525
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.08094948265368229,
+      "eval_loss": 0.72707200050354,
+      "eval_macro_f1": 0.5029402835068527,
+      "eval_micro_f1": 0.5317832764505119,
+      "eval_runtime": 0.7956,
+      "eval_samples_per_second": 2065.152,
+      "eval_steps_per_second": 129.465,
+      "eval_weighted_f1": 0.5432343330455452,
+      "step": 4540
+    },
+    {
+      "epoch": 10.022038567493112,
+      "grad_norm": 6.184263229370117,
+      "learning_rate": 1.108908467939305e-05,
+      "loss": 1.4583,
+      "step": 4550
+    },
+    {
+      "epoch": 10.077134986225895,
+      "grad_norm": 10.197240829467773,
+      "learning_rate": 1.1027900146842878e-05,
+      "loss": 1.4161,
+      "step": 4575
+    },
+    {
+      "epoch": 10.132231404958677,
+      "grad_norm": 12.444828033447266,
+      "learning_rate": 1.0966715614292707e-05,
+      "loss": 1.4309,
+      "step": 4600
+    },
+    {
+      "epoch": 10.18732782369146,
+      "grad_norm": 16.441001892089844,
+      "learning_rate": 1.0905531081742536e-05,
+      "loss": 1.388,
+      "step": 4625
+    },
+    {
+      "epoch": 10.242424242424242,
+      "grad_norm": 8.201318740844727,
+      "learning_rate": 1.0844346549192366e-05,
+      "loss": 1.4199,
+      "step": 4650
+    },
+    {
+      "epoch": 10.297520661157025,
+      "grad_norm": 13.441341400146484,
+      "learning_rate": 1.0783162016642195e-05,
+      "loss": 1.4134,
+      "step": 4675
+    },
+    {
+      "epoch": 10.352617079889807,
+      "grad_norm": 10.3322172164917,
+      "learning_rate": 1.0721977484092023e-05,
+      "loss": 1.3602,
+      "step": 4700
+    },
+    {
+      "epoch": 10.40771349862259,
+      "grad_norm": 5.628726482391357,
+      "learning_rate": 1.0660792951541852e-05,
+      "loss": 1.3711,
+      "step": 4725
+    },
+    {
+      "epoch": 10.462809917355372,
+      "grad_norm": 6.589449882507324,
+      "learning_rate": 1.059960841899168e-05,
+      "loss": 1.4131,
+      "step": 4750
+    },
+    {
+      "epoch": 10.517906336088155,
+      "grad_norm": 10.09038257598877,
+      "learning_rate": 1.053842388644151e-05,
+      "loss": 1.407,
+      "step": 4775
+    },
+    {
+      "epoch": 10.573002754820937,
+      "grad_norm": 16.832630157470703,
+      "learning_rate": 1.0477239353891338e-05,
+      "loss": 1.3365,
+      "step": 4800
+    },
+    {
+      "epoch": 10.62809917355372,
+      "grad_norm": 10.801997184753418,
+      "learning_rate": 1.0416054821341167e-05,
+      "loss": 1.3448,
+      "step": 4825
+    },
+    {
+      "epoch": 10.683195592286502,
+      "grad_norm": 10.650815963745117,
+      "learning_rate": 1.0354870288790995e-05,
+      "loss": 1.405,
+      "step": 4850
+    },
+    {
+      "epoch": 10.738292011019285,
+      "grad_norm": 9.310195922851562,
+      "learning_rate": 1.0293685756240824e-05,
+      "loss": 1.3109,
+      "step": 4875
+    },
+    {
+      "epoch": 10.793388429752067,
+      "grad_norm": 8.726702690124512,
+      "learning_rate": 1.0232501223690653e-05,
+      "loss": 1.3609,
+      "step": 4900
+    },
+    {
+      "epoch": 10.848484848484848,
+      "grad_norm": 8.540894508361816,
+      "learning_rate": 1.017131669114048e-05,
+      "loss": 1.3933,
+      "step": 4925
+    },
+    {
+      "epoch": 10.90358126721763,
+      "grad_norm": 9.035423278808594,
+      "learning_rate": 1.0110132158590308e-05,
+      "loss": 1.3052,
+      "step": 4950
+    },
+    {
+      "epoch": 10.958677685950413,
+      "grad_norm": 9.07214641571045,
+      "learning_rate": 1.0048947626040137e-05,
+      "loss": 1.4001,
+      "step": 4975
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.09007912355447352,
+      "eval_loss": 0.7351568341255188,
+      "eval_macro_f1": 0.5172892371402843,
+      "eval_micro_f1": 0.5423877327491785,
+      "eval_runtime": 0.786,
+      "eval_samples_per_second": 2090.312,
+      "eval_steps_per_second": 131.042,
+      "eval_weighted_f1": 0.5519542884665859,
+      "step": 4994
+    },
+    {
+      "epoch": 11.013223140495867,
+      "grad_norm": 5.78129768371582,
+      "learning_rate": 9.987763093489966e-06,
+      "loss": 1.32,
+      "step": 5000
+    },
+    {
+      "epoch": 11.06831955922865,
+      "grad_norm": 8.128321647644043,
+      "learning_rate": 9.926578560939796e-06,
+      "loss": 1.2663,
+      "step": 5025
+    },
+    {
+      "epoch": 11.123415977961432,
+      "grad_norm": 8.939966201782227,
+      "learning_rate": 9.865394028389625e-06,
+      "loss": 1.2988,
+      "step": 5050
+    },
+    {
+      "epoch": 11.178512396694215,
+      "grad_norm": 6.803643703460693,
+      "learning_rate": 9.804209495839453e-06,
+      "loss": 1.2649,
+      "step": 5075
+    },
+    {
+      "epoch": 11.233608815426997,
+      "grad_norm": 16.629180908203125,
+      "learning_rate": 9.743024963289282e-06,
+      "loss": 1.3324,
+      "step": 5100
+    },
+    {
+      "epoch": 11.28870523415978,
+      "grad_norm": 7.846168041229248,
+      "learning_rate": 9.68184043073911e-06,
+      "loss": 1.2661,
+      "step": 5125
+    },
+    {
+      "epoch": 11.343801652892562,
+      "grad_norm": 10.154409408569336,
+      "learning_rate": 9.620655898188939e-06,
+      "loss": 1.3156,
+      "step": 5150
+    },
+    {
+      "epoch": 11.398898071625345,
+      "grad_norm": 6.023170471191406,
+      "learning_rate": 9.559471365638768e-06,
+      "loss": 1.2855,
+      "step": 5175
+    },
+    {
+      "epoch": 11.453994490358127,
+      "grad_norm": 10.903104782104492,
+      "learning_rate": 9.498286833088596e-06,
+      "loss": 1.3085,
+      "step": 5200
+    },
+    {
+      "epoch": 11.50909090909091,
+      "grad_norm": 6.417316913604736,
+      "learning_rate": 9.437102300538425e-06,
+      "loss": 1.2589,
+      "step": 5225
+    },
+    {
+      "epoch": 11.564187327823692,
+      "grad_norm": 14.446656227111816,
+      "learning_rate": 9.375917767988254e-06,
+      "loss": 1.2417,
+      "step": 5250
+    },
+    {
+      "epoch": 11.619283746556473,
+      "grad_norm": 9.680697441101074,
+      "learning_rate": 9.31473323543808e-06,
+      "loss": 1.258,
+      "step": 5275
+    },
+    {
+      "epoch": 11.674380165289255,
+      "grad_norm": 7.470788478851318,
+      "learning_rate": 9.253548702887911e-06,
+      "loss": 1.2603,
+      "step": 5300
+    },
+    {
+      "epoch": 11.729476584022038,
+      "grad_norm": 6.185989856719971,
+      "learning_rate": 9.19236417033774e-06,
+      "loss": 1.2775,
+      "step": 5325
+    },
+    {
+      "epoch": 11.78457300275482,
+      "grad_norm": 6.857886791229248,
+      "learning_rate": 9.131179637787568e-06,
+      "loss": 1.2927,
+      "step": 5350
+    },
+    {
+      "epoch": 11.839669421487603,
+      "grad_norm": 8.878190040588379,
+      "learning_rate": 9.069995105237397e-06,
+      "loss": 1.2458,
+      "step": 5375
+    },
+    {
+      "epoch": 11.894765840220385,
+      "grad_norm": 12.2567720413208,
+      "learning_rate": 9.008810572687226e-06,
+      "loss": 1.2774,
+      "step": 5400
+    },
+    {
+      "epoch": 11.949862258953168,
+      "grad_norm": 7.592448711395264,
+      "learning_rate": 8.947626040137054e-06,
+      "loss": 1.274,
+      "step": 5425
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.10164333536214243,
+      "eval_loss": 0.7427167892456055,
+      "eval_macro_f1": 0.5173683757576815,
+      "eval_micro_f1": 0.5469133051036327,
+      "eval_runtime": 0.7544,
+      "eval_samples_per_second": 2177.917,
+      "eval_steps_per_second": 136.534,
+      "eval_weighted_f1": 0.5552655400400709,
+      "step": 5448
+    },
+    {
+      "epoch": 12.004407713498622,
+      "grad_norm": 8.367660522460938,
+      "learning_rate": 8.886441507586883e-06,
+      "loss": 1.2109,
+      "step": 5450
+    },
+    {
+      "epoch": 12.059504132231405,
+      "grad_norm": 146.01919555664062,
+      "learning_rate": 8.825256975036712e-06,
+      "loss": 1.1759,
+      "step": 5475
+    },
+    {
+      "epoch": 12.114600550964187,
+      "grad_norm": 5.678831100463867,
+      "learning_rate": 8.76407244248654e-06,
+      "loss": 1.2324,
+      "step": 5500
+    },
+    {
+      "epoch": 12.16969696969697,
+      "grad_norm": 6.2751851081848145,
+      "learning_rate": 8.702887909936369e-06,
+      "loss": 1.2089,
+      "step": 5525
+    },
+    {
+      "epoch": 12.224793388429752,
+      "grad_norm": 6.340836048126221,
+      "learning_rate": 8.641703377386198e-06,
+      "loss": 1.2303,
+      "step": 5550
+    },
+    {
+      "epoch": 12.279889807162535,
+      "grad_norm": 8.033232688903809,
+      "learning_rate": 8.580518844836026e-06,
+      "loss": 1.1842,
+      "step": 5575
+    },
+    {
+      "epoch": 12.334986225895317,
+      "grad_norm": 5.0165228843688965,
+      "learning_rate": 8.519334312285855e-06,
+      "loss": 1.1871,
+      "step": 5600
+    },
+    {
+      "epoch": 12.3900826446281,
+      "grad_norm": 6.247511863708496,
+      "learning_rate": 8.458149779735683e-06,
+      "loss": 1.1719,
+      "step": 5625
+    },
+    {
+      "epoch": 12.445179063360882,
+      "grad_norm": 6.809659957885742,
+      "learning_rate": 8.396965247185512e-06,
+      "loss": 1.174,
+      "step": 5650
+    },
+    {
+      "epoch": 12.500275482093663,
+      "grad_norm": 10.726896286010742,
+      "learning_rate": 8.33578071463534e-06,
+      "loss": 1.2047,
+      "step": 5675
+    },
+    {
+      "epoch": 12.555371900826445,
+      "grad_norm": 7.948911666870117,
+      "learning_rate": 8.27459618208517e-06,
+      "loss": 1.1879,
+      "step": 5700
+    },
+    {
+      "epoch": 12.610468319559228,
+      "grad_norm": 27.659639358520508,
+      "learning_rate": 8.213411649534998e-06,
+      "loss": 1.2368,
+      "step": 5725
+    },
+    {
+      "epoch": 12.66556473829201,
+      "grad_norm": 10.702486038208008,
+      "learning_rate": 8.152227116984827e-06,
+      "loss": 1.1614,
+      "step": 5750
+    },
+    {
+      "epoch": 12.720661157024793,
+      "grad_norm": 12.045475006103516,
+      "learning_rate": 8.091042584434655e-06,
+      "loss": 1.1367,
+      "step": 5775
+    },
+    {
+      "epoch": 12.775757575757575,
+      "grad_norm": 9.43398666381836,
+      "learning_rate": 8.029858051884484e-06,
+      "loss": 1.1774,
+      "step": 5800
+    },
+    {
+      "epoch": 12.830853994490358,
+      "grad_norm": 9.546597480773926,
+      "learning_rate": 7.968673519334313e-06,
+      "loss": 1.2087,
+      "step": 5825
+    },
+    {
+      "epoch": 12.88595041322314,
+      "grad_norm": 11.342320442199707,
+      "learning_rate": 7.907488986784141e-06,
+      "loss": 1.1969,
+      "step": 5850
+    },
+    {
+      "epoch": 12.941046831955923,
+      "grad_norm": 6.419647216796875,
+      "learning_rate": 7.84630445423397e-06,
+      "loss": 1.1185,
+      "step": 5875
+    },
+    {
+      "epoch": 12.996143250688705,
+      "grad_norm": 10.889259338378906,
+      "learning_rate": 7.785119921683799e-06,
+      "loss": 1.1474,
+      "step": 5900
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.09981740718198417,
+      "eval_loss": 0.7411310076713562,
+      "eval_macro_f1": 0.5188983413806106,
+      "eval_micro_f1": 0.5482582649212336,
+      "eval_runtime": 0.7318,
+      "eval_samples_per_second": 2245.254,
+      "eval_steps_per_second": 140.755,
+      "eval_weighted_f1": 0.5550465213529865,
+      "step": 5902
+    },
+    {
+      "epoch": 13.05068870523416,
+      "grad_norm": 10.158915519714355,
+      "learning_rate": 7.723935389133627e-06,
+      "loss": 1.1359,
+      "step": 5925
+    },
+    {
+      "epoch": 13.105785123966943,
+      "grad_norm": 9.29521369934082,
+      "learning_rate": 7.662750856583456e-06,
+      "loss": 1.1321,
+      "step": 5950
+    },
+    {
+      "epoch": 13.160881542699725,
+      "grad_norm": 10.788544654846191,
+      "learning_rate": 7.601566324033285e-06,
+      "loss": 1.1328,
+      "step": 5975
+    },
+    {
+      "epoch": 13.215977961432507,
+      "grad_norm": 9.468472480773926,
+      "learning_rate": 7.540381791483113e-06,
+      "loss": 1.119,
+      "step": 6000
+    },
+    {
+      "epoch": 13.27107438016529,
+      "grad_norm": 12.014350891113281,
+      "learning_rate": 7.479197258932942e-06,
+      "loss": 1.1089,
+      "step": 6025
+    },
+    {
+      "epoch": 13.326170798898072,
+      "grad_norm": 6.275262832641602,
+      "learning_rate": 7.4180127263827705e-06,
+      "loss": 1.1273,
+      "step": 6050
+    },
+    {
+      "epoch": 13.381267217630853,
+      "grad_norm": 6.748322010040283,
+      "learning_rate": 7.3568281938326e-06,
+      "loss": 1.1488,
+      "step": 6075
+    },
+    {
+      "epoch": 13.436363636363636,
+      "grad_norm": 9.268157958984375,
+      "learning_rate": 7.295643661282429e-06,
+      "loss": 1.0934,
+      "step": 6100
+    },
+    {
+      "epoch": 13.491460055096418,
+      "grad_norm": 9.821578025817871,
+      "learning_rate": 7.234459128732257e-06,
+      "loss": 1.1626,
+      "step": 6125
+    },
+    {
+      "epoch": 13.5465564738292,
+      "grad_norm": 16.561037063598633,
+      "learning_rate": 7.173274596182086e-06,
+      "loss": 1.124,
+      "step": 6150
+    },
+    {
+      "epoch": 13.601652892561983,
+      "grad_norm": 8.936065673828125,
+      "learning_rate": 7.112090063631915e-06,
+      "loss": 1.0531,
+      "step": 6175
+    },
+    {
+      "epoch": 13.656749311294766,
+      "grad_norm": 7.354989528656006,
+      "learning_rate": 7.050905531081743e-06,
+      "loss": 1.0947,
+      "step": 6200
+    },
+    {
+      "epoch": 13.711845730027548,
+      "grad_norm": 7.0242600440979,
+      "learning_rate": 6.989720998531572e-06,
+      "loss": 1.1123,
+      "step": 6225
+    },
+    {
+      "epoch": 13.76694214876033,
+      "grad_norm": 8.146479606628418,
+      "learning_rate": 6.928536465981401e-06,
+      "loss": 1.1017,
+      "step": 6250
+    },
+    {
+      "epoch": 13.822038567493113,
+      "grad_norm": 7.871755599975586,
+      "learning_rate": 6.867351933431229e-06,
+      "loss": 1.0755,
+      "step": 6275
+    },
+    {
+      "epoch": 13.877134986225895,
+      "grad_norm": 8.830574989318848,
+      "learning_rate": 6.806167400881057e-06,
+      "loss": 1.0948,
+      "step": 6300
+    },
+    {
+      "epoch": 13.932231404958678,
+      "grad_norm": 6.276432514190674,
+      "learning_rate": 6.744982868330886e-06,
+      "loss": 1.0978,
+      "step": 6325
+    },
+    {
+      "epoch": 13.98732782369146,
+      "grad_norm": 31.289966583251953,
+      "learning_rate": 6.683798335780716e-06,
+      "loss": 1.0909,
+      "step": 6350
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.11381618989653074,
+      "eval_loss": 0.7726384997367859,
+      "eval_macro_f1": 0.5301473406912293,
+      "eval_micro_f1": 0.5577033055015441,
+      "eval_runtime": 0.7668,
+      "eval_samples_per_second": 2142.769,
+      "eval_steps_per_second": 134.331,
+      "eval_weighted_f1": 0.5628237647563301,
+      "step": 6356
+    },
+    {
+      "epoch": 14.041873278236915,
+      "grad_norm": 9.511872291564941,
+      "learning_rate": 6.622613803230544e-06,
+      "loss": 1.0502,
+      "step": 6375
+    },
+    {
+      "epoch": 14.096969696969698,
+      "grad_norm": 8.232625007629395,
+      "learning_rate": 6.5614292706803725e-06,
+      "loss": 1.0125,
+      "step": 6400
+    },
+    {
+      "epoch": 14.15206611570248,
+      "grad_norm": 13.766243934631348,
+      "learning_rate": 6.500244738130201e-06,
+      "loss": 1.0749,
+      "step": 6425
+    },
+    {
+      "epoch": 14.207162534435263,
+      "grad_norm": 8.681984901428223,
+      "learning_rate": 6.43906020558003e-06,
+      "loss": 1.0245,
+      "step": 6450
+    },
+    {
+      "epoch": 14.262258953168043,
+      "grad_norm": 12.72724437713623,
+      "learning_rate": 6.3778756730298585e-06,
+      "loss": 1.0497,
+      "step": 6475
+    },
+    {
+      "epoch": 14.317355371900826,
+      "grad_norm": 10.159751892089844,
+      "learning_rate": 6.316691140479687e-06,
+      "loss": 1.0501,
+      "step": 6500
+    },
+    {
+      "epoch": 14.372451790633608,
+      "grad_norm": 8.054067611694336,
+      "learning_rate": 6.255506607929516e-06,
+      "loss": 1.0853,
+      "step": 6525
+    },
+    {
+      "epoch": 14.42754820936639,
+      "grad_norm": 9.144680976867676,
+      "learning_rate": 6.194322075379344e-06,
+      "loss": 1.0849,
+      "step": 6550
+    },
+    {
+      "epoch": 14.482644628099173,
+      "grad_norm": 7.8340020179748535,
+      "learning_rate": 6.133137542829173e-06,
+      "loss": 1.0418,
+      "step": 6575
+    },
+    {
+      "epoch": 14.537741046831956,
+      "grad_norm": 9.663851737976074,
+      "learning_rate": 6.071953010279002e-06,
+      "loss": 1.0477,
+      "step": 6600
+    },
+    {
+      "epoch": 14.592837465564738,
+      "grad_norm": 6.405959129333496,
+      "learning_rate": 6.010768477728831e-06,
+      "loss": 1.0559,
+      "step": 6625
+    },
+    {
+      "epoch": 14.64793388429752,
+      "grad_norm": 9.999762535095215,
+      "learning_rate": 5.94958394517866e-06,
+      "loss": 1.0371,
+      "step": 6650
+    },
+    {
+      "epoch": 14.703030303030303,
+      "grad_norm": 6.774183750152588,
+      "learning_rate": 5.8883994126284885e-06,
+      "loss": 1.0193,
+      "step": 6675
+    },
+    {
+      "epoch": 14.758126721763086,
+      "grad_norm": 13.159234046936035,
+      "learning_rate": 5.827214880078317e-06,
+      "loss": 1.0433,
+      "step": 6700
+    },
+    {
+      "epoch": 14.813223140495868,
+      "grad_norm": 12.729023933410645,
+      "learning_rate": 5.766030347528146e-06,
+      "loss": 1.0167,
+      "step": 6725
+    },
+    {
+      "epoch": 14.86831955922865,
+      "grad_norm": 15.757935523986816,
+      "learning_rate": 5.704845814977974e-06,
+      "loss": 1.0048,
+      "step": 6750
+    },
+    {
+      "epoch": 14.923415977961433,
+      "grad_norm": 25.768495559692383,
+      "learning_rate": 5.643661282427802e-06,
+      "loss": 1.0126,
+      "step": 6775
+    },
+    {
+      "epoch": 14.978512396694216,
+      "grad_norm": 8.229415893554688,
+      "learning_rate": 5.582476749877631e-06,
+      "loss": 1.0675,
+      "step": 6800
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.11929397443700548,
+      "eval_loss": 0.7604466080665588,
+      "eval_macro_f1": 0.5233293354606164,
+      "eval_micro_f1": 0.55444864307798,
+      "eval_runtime": 0.7401,
+      "eval_samples_per_second": 2220.088,
+      "eval_steps_per_second": 139.178,
+      "eval_weighted_f1": 0.5605802174257807,
+      "step": 6810
+    },
+    {
+      "epoch": 15.03305785123967,
+      "grad_norm": 7.657383441925049,
+      "learning_rate": 5.5212922173274596e-06,
+      "loss": 1.0134,
+      "step": 6825
+    },
+    {
+      "epoch": 15.088154269972453,
+      "grad_norm": 6.9480204582214355,
+      "learning_rate": 5.460107684777288e-06,
+      "loss": 0.9708,
+      "step": 6850
+    },
+    {
+      "epoch": 15.143250688705233,
+      "grad_norm": 4.3695387840271,
+      "learning_rate": 5.398923152227118e-06,
+      "loss": 0.9865,
+      "step": 6875
+    },
+    {
+      "epoch": 15.198347107438016,
+      "grad_norm": 6.225238800048828,
+      "learning_rate": 5.337738619676946e-06,
+      "loss": 1.0123,
+      "step": 6900
+    },
+    {
+      "epoch": 15.253443526170798,
+      "grad_norm": 11.188499450683594,
+      "learning_rate": 5.276554087126775e-06,
+      "loss": 0.9725,
+      "step": 6925
+    },
+    {
+      "epoch": 15.30853994490358,
+      "grad_norm": 36.433753967285156,
+      "learning_rate": 5.215369554576604e-06,
+      "loss": 0.9925,
+      "step": 6950
+    },
+    {
+      "epoch": 15.363636363636363,
+      "grad_norm": 7.442889213562012,
+      "learning_rate": 5.154185022026432e-06,
+      "loss": 1.0082,
+      "step": 6975
+    },
+    {
+      "epoch": 15.418732782369146,
+      "grad_norm": 16.42450714111328,
+      "learning_rate": 5.093000489476261e-06,
+      "loss": 1.0109,
+      "step": 7000
+    },
+    {
+      "epoch": 15.473829201101928,
+      "grad_norm": 7.358915328979492,
+      "learning_rate": 5.03181595692609e-06,
+      "loss": 0.9753,
+      "step": 7025
+    },
+    {
+      "epoch": 15.52892561983471,
+      "grad_norm": 17.838838577270508,
+      "learning_rate": 4.970631424375918e-06,
+      "loss": 0.979,
+      "step": 7050
+    },
+    {
+      "epoch": 15.584022038567493,
+      "grad_norm": 10.839521408081055,
+      "learning_rate": 4.909446891825747e-06,
+      "loss": 1.0254,
+      "step": 7075
+    },
+    {
+      "epoch": 15.639118457300276,
+      "grad_norm": 10.101546287536621,
+      "learning_rate": 4.8482623592755756e-06,
+      "loss": 1.0127,
+      "step": 7100
+    },
+    {
+      "epoch": 15.694214876033058,
+      "grad_norm": 10.862388610839844,
+      "learning_rate": 4.787077826725404e-06,
+      "loss": 0.998,
+      "step": 7125
+    },
+    {
+      "epoch": 15.74931129476584,
+      "grad_norm": 9.09751033782959,
+      "learning_rate": 4.725893294175233e-06,
+      "loss": 1.008,
+      "step": 7150
+    },
+    {
+      "epoch": 15.804407713498623,
+      "grad_norm": 6.354736328125,
+      "learning_rate": 4.6647087616250615e-06,
+      "loss": 0.9695,
+      "step": 7175
+    },
+    {
+      "epoch": 15.859504132231406,
+      "grad_norm": 8.806958198547363,
+      "learning_rate": 4.60352422907489e-06,
+      "loss": 0.9777,
+      "step": 7200
+    },
+    {
+      "epoch": 15.914600550964188,
+      "grad_norm": 7.194450378417969,
+      "learning_rate": 4.542339696524719e-06,
+      "loss": 0.9588,
+      "step": 7225
+    },
+    {
+      "epoch": 15.969696969696969,
+      "grad_norm": 5.09818172454834,
+      "learning_rate": 4.4811551639745475e-06,
+      "loss": 0.9835,
+      "step": 7250
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.13451004260499086,
+      "eval_loss": 0.7917519807815552,
+      "eval_macro_f1": 0.5387736977274954,
+      "eval_micro_f1": 0.5704451038575667,
+      "eval_runtime": 0.732,
+      "eval_samples_per_second": 2244.679,
+      "eval_steps_per_second": 140.719,
+      "eval_weighted_f1": 0.575498510225612,
+      "step": 7264
+    },
+    {
+      "epoch": 16.024242424242424,
+      "grad_norm": 6.4887776374816895,
+      "learning_rate": 4.419970631424376e-06,
+      "loss": 0.9353,
+      "step": 7275
+    },
+    {
+      "epoch": 16.079338842975208,
+      "grad_norm": 11.537016868591309,
+      "learning_rate": 4.358786098874205e-06,
+      "loss": 0.9686,
+      "step": 7300
+    },
+    {
+      "epoch": 16.13443526170799,
+      "grad_norm": 9.97562026977539,
+      "learning_rate": 4.297601566324034e-06,
+      "loss": 0.9619,
+      "step": 7325
+    },
+    {
+      "epoch": 16.189531680440773,
+      "grad_norm": 16.37826919555664,
+      "learning_rate": 4.236417033773862e-06,
+      "loss": 0.9198,
+      "step": 7350
+    },
+    {
+      "epoch": 16.244628099173553,
+      "grad_norm": 15.019182205200195,
+      "learning_rate": 4.175232501223691e-06,
+      "loss": 0.9585,
+      "step": 7375
+    },
+    {
+      "epoch": 16.299724517906338,
+      "grad_norm": 9.785524368286133,
+      "learning_rate": 4.114047968673519e-06,
+      "loss": 0.9148,
+      "step": 7400
+    },
+    {
+      "epoch": 16.35482093663912,
+      "grad_norm": 8.5923433303833,
+      "learning_rate": 4.052863436123348e-06,
+      "loss": 0.9847,
+      "step": 7425
+    },
+    {
+      "epoch": 16.4099173553719,
+      "grad_norm": 8.138625144958496,
+      "learning_rate": 3.991678903573177e-06,
+      "loss": 0.9391,
+      "step": 7450
+    },
+    {
+      "epoch": 16.465013774104683,
+      "grad_norm": 8.172958374023438,
+      "learning_rate": 3.930494371023006e-06,
+      "loss": 0.9434,
+      "step": 7475
+    },
+    {
+      "epoch": 16.520110192837464,
+      "grad_norm": 4.830559253692627,
+      "learning_rate": 3.869309838472835e-06,
+      "loss": 0.9133,
+      "step": 7500
+    },
+    {
+      "epoch": 16.57520661157025,
+      "grad_norm": 9.164722442626953,
+      "learning_rate": 3.808125305922663e-06,
+      "loss": 0.9265,
+      "step": 7525
+    },
+    {
+      "epoch": 16.63030303030303,
+      "grad_norm": 9.447441101074219,
+      "learning_rate": 3.7469407733724917e-06,
+      "loss": 0.9427,
+      "step": 7550
+    },
+    {
+      "epoch": 16.685399449035813,
+      "grad_norm": 11.204426765441895,
+      "learning_rate": 3.6857562408223203e-06,
+      "loss": 0.9283,
+      "step": 7575
+    },
+    {
+      "epoch": 16.740495867768594,
+      "grad_norm": 10.056817054748535,
+      "learning_rate": 3.6245717082721494e-06,
+      "loss": 0.9617,
+      "step": 7600
+    },
+    {
+      "epoch": 16.79559228650138,
+      "grad_norm": 8.630284309387207,
+      "learning_rate": 3.563387175721978e-06,
+      "loss": 0.9385,
+      "step": 7625
+    },
+    {
+      "epoch": 16.85068870523416,
+      "grad_norm": 9.643190383911133,
+      "learning_rate": 3.5022026431718063e-06,
+      "loss": 0.9208,
+      "step": 7650
+    },
+    {
+      "epoch": 16.905785123966943,
+      "grad_norm": 9.304166793823242,
+      "learning_rate": 3.441018110621635e-06,
+      "loss": 0.9371,
+      "step": 7675
+    },
+    {
+      "epoch": 16.960881542699724,
+      "grad_norm": 6.273467540740967,
+      "learning_rate": 3.3798335780714636e-06,
+      "loss": 0.9175,
+      "step": 7700
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.12598904443091904,
+      "eval_loss": 0.7984939217567444,
+      "eval_macro_f1": 0.5367619871597079,
+      "eval_micro_f1": 0.5672975196896674,
+      "eval_runtime": 0.7179,
+      "eval_samples_per_second": 2288.569,
+      "eval_steps_per_second": 143.471,
+      "eval_weighted_f1": 0.5714722158764484,
+      "step": 7718
+    },
+    {
+      "epoch": 17.01542699724518,
+      "grad_norm": 5.473207950592041,
+      "learning_rate": 3.3186490455212927e-06,
+      "loss": 0.9151,
+      "step": 7725
+    },
+    {
+      "epoch": 17.070523415977963,
+      "grad_norm": 13.94350528717041,
+      "learning_rate": 3.2574645129711213e-06,
+      "loss": 0.8743,
+      "step": 7750
+    },
+    {
+      "epoch": 17.125619834710744,
+      "grad_norm": 5.383657455444336,
+      "learning_rate": 3.19627998042095e-06,
+      "loss": 0.8991,
+      "step": 7775
+    },
+    {
+      "epoch": 17.180716253443528,
+      "grad_norm": 9.38705062866211,
+      "learning_rate": 3.1350954478707786e-06,
+      "loss": 0.8857,
+      "step": 7800
+    },
+    {
+      "epoch": 17.23581267217631,
+      "grad_norm": 9.738102912902832,
+      "learning_rate": 3.073910915320607e-06,
+      "loss": 0.9093,
+      "step": 7825
+    },
+    {
+      "epoch": 17.29090909090909,
+      "grad_norm": 12.653076171875,
+      "learning_rate": 3.0127263827704355e-06,
+      "loss": 0.9191,
+      "step": 7850
+    },
+    {
+      "epoch": 17.346005509641873,
+      "grad_norm": 9.38412857055664,
+      "learning_rate": 2.9515418502202646e-06,
+      "loss": 0.8707,
+      "step": 7875
+    },
+    {
+      "epoch": 17.401101928374654,
+      "grad_norm": 10.618575096130371,
+      "learning_rate": 2.8903573176700932e-06,
+      "loss": 0.9536,
+      "step": 7900
+    },
+    {
+      "epoch": 17.45619834710744,
+      "grad_norm": 8.758997917175293,
+      "learning_rate": 2.829172785119922e-06,
+      "loss": 0.9043,
+      "step": 7925
+    },
+    {
+      "epoch": 17.51129476584022,
+      "grad_norm": 12.952402114868164,
+      "learning_rate": 2.7679882525697505e-06,
+      "loss": 0.9041,
+      "step": 7950
+    },
+    {
+      "epoch": 17.566391184573003,
+      "grad_norm": 8.233007431030273,
+      "learning_rate": 2.706803720019579e-06,
+      "loss": 0.9614,
+      "step": 7975
+    },
+    {
+      "epoch": 17.621487603305784,
+      "grad_norm": 18.04732322692871,
+      "learning_rate": 2.6456191874694083e-06,
+      "loss": 0.9338,
+      "step": 8000
+    },
+    {
+      "epoch": 17.67658402203857,
+      "grad_norm": 6.762555122375488,
+      "learning_rate": 2.584434654919237e-06,
+      "loss": 0.9321,
+      "step": 8025
+    },
+    {
+      "epoch": 17.73168044077135,
+      "grad_norm": 5.75241756439209,
+      "learning_rate": 2.523250122369065e-06,
+      "loss": 0.8526,
+      "step": 8050
+    },
+    {
+      "epoch": 17.786776859504133,
+      "grad_norm": 7.084556579589844,
+      "learning_rate": 2.4620655898188938e-06,
+      "loss": 0.8949,
+      "step": 8075
+    },
+    {
+      "epoch": 17.841873278236914,
+      "grad_norm": 9.39500904083252,
+      "learning_rate": 2.400881057268723e-06,
+      "loss": 0.914,
+      "step": 8100
+    },
+    {
+      "epoch": 17.8969696969697,
+      "grad_norm": 5.70341682434082,
+      "learning_rate": 2.3396965247185515e-06,
+      "loss": 0.8862,
+      "step": 8125
+    },
+    {
+      "epoch": 17.95206611570248,
+      "grad_norm": 11.6044921875,
+      "learning_rate": 2.27851199216838e-06,
+      "loss": 0.9043,
+      "step": 8150
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.13085818624467438,
+      "eval_loss": 0.8047019839286804,
+      "eval_macro_f1": 0.5374664109875137,
+      "eval_micro_f1": 0.5688203912270302,
+      "eval_runtime": 0.7594,
+      "eval_samples_per_second": 2163.66,
+      "eval_steps_per_second": 135.64,
+      "eval_weighted_f1": 0.5729212099169474,
+      "step": 8172
+    },
+    {
+      "epoch": 18.006611570247934,
+      "grad_norm": 9.203124046325684,
+      "learning_rate": 2.217327459618209e-06,
+      "loss": 0.9186,
+      "step": 8175
+    },
+    {
+      "epoch": 18.061707988980718,
+      "grad_norm": 11.816119194030762,
+      "learning_rate": 2.1561429270680375e-06,
+      "loss": 0.8877,
+      "step": 8200
+    },
+    {
+      "epoch": 18.1168044077135,
+      "grad_norm": 9.449479103088379,
+      "learning_rate": 2.094958394517866e-06,
+      "loss": 0.8621,
+      "step": 8225
+    },
+    {
+      "epoch": 18.17190082644628,
+      "grad_norm": 20.197847366333008,
+      "learning_rate": 2.0337738619676948e-06,
+      "loss": 0.899,
+      "step": 8250
+    },
+    {
+      "epoch": 18.226997245179064,
+      "grad_norm": 11.110992431640625,
+      "learning_rate": 1.9725893294175234e-06,
+      "loss": 0.9089,
+      "step": 8275
+    },
+    {
+      "epoch": 18.282093663911844,
+      "grad_norm": 7.784780025482178,
+      "learning_rate": 1.911404796867352e-06,
+      "loss": 0.8785,
+      "step": 8300
+    },
+    {
+      "epoch": 18.33719008264463,
+      "grad_norm": 12.796671867370605,
+      "learning_rate": 1.8502202643171807e-06,
+      "loss": 0.8987,
+      "step": 8325
+    },
+    {
+      "epoch": 18.39228650137741,
+      "grad_norm": 8.227873802185059,
+      "learning_rate": 1.7890357317670096e-06,
+      "loss": 0.8942,
+      "step": 8350
+    },
+    {
+      "epoch": 18.447382920110194,
+      "grad_norm": 10.541505813598633,
+      "learning_rate": 1.7278511992168382e-06,
+      "loss": 0.8513,
+      "step": 8375
+    },
+    {
+      "epoch": 18.502479338842974,
+      "grad_norm": 11.191814422607422,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.8549,
+      "step": 8400
+    },
+    {
+      "epoch": 18.55757575757576,
+      "grad_norm": 21.217105865478516,
+      "learning_rate": 1.6054821341164955e-06,
+      "loss": 0.9218,
+      "step": 8425
+    },
+    {
+      "epoch": 18.61267217630854,
+      "grad_norm": 5.7040557861328125,
+      "learning_rate": 1.5442976015663242e-06,
+      "loss": 0.8749,
+      "step": 8450
+    },
+    {
+      "epoch": 18.667768595041323,
+      "grad_norm": 6.811832904815674,
+      "learning_rate": 1.4831130690161526e-06,
+      "loss": 0.8652,
+      "step": 8475
+    },
+    {
+      "epoch": 18.722865013774104,
+      "grad_norm": 7.055276870727539,
+      "learning_rate": 1.4219285364659815e-06,
+      "loss": 0.8869,
+      "step": 8500
+    },
+    {
+      "epoch": 18.77796143250689,
+      "grad_norm": 10.736101150512695,
+      "learning_rate": 1.3607440039158101e-06,
+      "loss": 0.8619,
+      "step": 8525
+    },
+    {
+      "epoch": 18.83305785123967,
+      "grad_norm": 15.553990364074707,
+      "learning_rate": 1.299559471365639e-06,
+      "loss": 0.8802,
+      "step": 8550
+    },
+    {
+      "epoch": 18.888154269972453,
+      "grad_norm": 18.04989242553711,
+      "learning_rate": 1.2383749388154676e-06,
+      "loss": 0.8842,
+      "step": 8575
+    },
+    {
+      "epoch": 18.943250688705234,
+      "grad_norm": 10.656030654907227,
+      "learning_rate": 1.1771904062652963e-06,
+      "loss": 0.8953,
+      "step": 8600
+    },
+    {
+      "epoch": 18.998347107438015,
+      "grad_norm": 7.111612796783447,
+      "learning_rate": 1.116005873715125e-06,
+      "loss": 0.8614,
+      "step": 8625
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.13694461351186854,
+      "eval_loss": 0.8140851855278015,
+      "eval_macro_f1": 0.5383380316480916,
+      "eval_micro_f1": 0.5684865900383141,
+      "eval_runtime": 0.7732,
+      "eval_samples_per_second": 2124.851,
+      "eval_steps_per_second": 133.207,
+      "eval_weighted_f1": 0.57270769626429,
+      "step": 8626
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 9080,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 3
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

ood-vigoemotions-phobert-source-source-02-s42-20260826T132817Z-0c086fe0/trainer/checkpoint-10013/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}