ccit_final

config.json

@@ -20,7 +20,7 @@
   "pad_token_id": 1,
   "position_embedding_type": "absolute",
   "torch_dtype": "float32",
-  "transformers_version": "4.41.1",
+  "transformers_version": "4.41.2",
   "type_vocab_size": 1,
   "use_cache": true,
   "vocab_size": 50265

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b9bcae3293b5d29ff9f520b875765460e825e12ec2ecccdf5fb2635ab3d058ec
-size 498692860
+oid sha256:b6aad7c2f42e665fe68a35f0abf79a2483077c39dd42fba313390c6515423a69
+size 498705152

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eeda6c76b7102eb22693937d10b880ae3af91f76df0a3a54359358853fc95e33
-size 997506746
+oid sha256:a7e0a9772004bf6dc7cb7d761cdf5209bcbc125cb4ff8281b6b4af487abc8608
+size 997531514

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cfc880f662532368769bf627f8612e2c4b90986f8b4c7b89b3f45762dc6d424d
+oid sha256:4297877051912922800eeda2dfabec7286e3f19a18bb48fac3d835eb91c6dbae
 size 14244

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f1b227ad2a10700168d2b014fe9a1d55c46d4ec5e2f18bd52971da8f64d4b44b
+oid sha256:db3e53efd18aae3a98eba7ce2c13a309a6e1555c5a2fe5b6c71d6270566d2fef
 size 1064

trainer_state.json

@@ -1,219 +1,556 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.685408299866131,
+  "epoch": 9.433962264150944,
   "eval_steps": 500,
-  "global_step": 3500,
+  "global_step": 9000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.6693440428380187,
-      "grad_norm": 14.38033676147461,
+      "epoch": 0.5241090146750524,
+      "grad_norm": 14.80826473236084,
       "learning_rate": 4.97e-05,
-      "loss": 1.6741,
+      "loss": 1.8113,
       "step": 500
     },
     {
-      "epoch": 0.6693440428380187,
-      "eval_accuracy": 0.6358458961474037,
-      "eval_f1_class_0": 0.9606299212598425,
-      "eval_f1_class_1": 0.4562647754137116,
-      "eval_f1_class_2": 0.5818673883626523,
-      "eval_f1_class_3": 0.6752,
-      "eval_f1_class_4": 0.4695259593679458,
-      "eval_f1_class_5": 0.6699186991869919,
-      "eval_f1_class_6": 0.6111908177905309,
-      "eval_f1_class_7": 0.6285714285714286,
-      "eval_f1_class_8": 0.7056603773584904,
-      "eval_f1_macro": 0.6398699297012882,
-      "eval_f1_micro": 0.6358458961474037,
-      "eval_f1_weighted": 0.6376816033602213,
-      "eval_loss": 1.0920162200927734,
-      "eval_recall_weighted": 0.6358458961474037,
-      "eval_runtime": 52.0874,
-      "eval_samples_per_second": 57.307,
-      "eval_steps_per_second": 1.805,
+      "epoch": 0.5241090146750524,
+      "eval_accuracy": 0.6428945988463556,
+      "eval_f1_class_0": 0.632183908045977,
+      "eval_f1_class_1": 0.9472222222222223,
+      "eval_f1_class_2": 0.3657331136738056,
+      "eval_f1_class_3": 0.5542168674698795,
+      "eval_f1_class_4": 0.6165254237288136,
+      "eval_f1_class_5": 0.6261808367071524,
+      "eval_f1_class_6": 0.6833541927409262,
+      "eval_f1_class_7": 0.5899053627760252,
+      "eval_f1_class_8": 0.662280701754386,
+      "eval_f1_class_9": 0.6956521739130435,
+      "eval_f1_macro": 0.6373254803032232,
+      "eval_f1_micro": 0.6428945988463556,
+      "eval_f1_weighted": 0.6376303124409798,
+      "eval_loss": 1.1190885305404663,
+      "eval_recall_weighted": 0.6428945988463556,
+      "eval_runtime": 18.5604,
+      "eval_samples_per_second": 205.491,
+      "eval_steps_per_second": 6.465,
       "step": 500
     },
     {
-      "epoch": 1.3386880856760375,
-      "grad_norm": 13.991486549377441,
-      "learning_rate": 4.643472022955524e-05,
-      "loss": 0.77,
+      "epoch": 1.0482180293501049,
+      "grad_norm": 6.733876705169678,
+      "learning_rate": 4.725110619469027e-05,
+      "loss": 0.885,
       "step": 1000
     },
     {
-      "epoch": 1.3386880856760375,
-      "eval_accuracy": 0.8395309882747068,
-      "eval_f1_class_0": 0.9654088050314467,
-      "eval_f1_class_1": 0.7083333333333334,
-      "eval_f1_class_2": 0.8022598870056498,
-      "eval_f1_class_3": 0.8431952662721893,
-      "eval_f1_class_4": 0.7962674961119751,
-      "eval_f1_class_5": 0.8715447154471545,
-      "eval_f1_class_6": 0.8243430152143845,
-      "eval_f1_class_7": 0.8963210702341138,
-      "eval_f1_class_8": 0.8755760368663594,
-      "eval_f1_macro": 0.8425832917240673,
-      "eval_f1_micro": 0.8395309882747068,
-      "eval_f1_weighted": 0.8406783887287406,
-      "eval_loss": 0.5749590992927551,
-      "eval_recall_weighted": 0.8395309882747068,
-      "eval_runtime": 52.1038,
-      "eval_samples_per_second": 57.289,
-      "eval_steps_per_second": 1.804,
+      "epoch": 1.0482180293501049,
+      "eval_accuracy": 0.8185631882538018,
+      "eval_f1_class_0": 0.7974522292993631,
+      "eval_f1_class_1": 0.9651324965132496,
+      "eval_f1_class_2": 0.6306569343065694,
+      "eval_f1_class_3": 0.7193229901269393,
+      "eval_f1_class_4": 0.8341584158415841,
+      "eval_f1_class_5": 0.807061790668348,
+      "eval_f1_class_6": 0.8753246753246754,
+      "eval_f1_class_7": 0.8368794326241134,
+      "eval_f1_class_8": 0.8787446504992867,
+      "eval_f1_class_9": 0.8255528255528255,
+      "eval_f1_macro": 0.8170286440756953,
+      "eval_f1_micro": 0.8185631882538018,
+      "eval_f1_weighted": 0.8175319894578669,
+      "eval_loss": 0.6289433836936951,
+      "eval_recall_weighted": 0.8185631882538018,
+      "eval_runtime": 18.5795,
+      "eval_samples_per_second": 205.28,
+      "eval_steps_per_second": 6.459,
       "step": 1000
     },
     {
-      "epoch": 2.0080321285140563,
-      "grad_norm": 2.8914544582366943,
-      "learning_rate": 4.284791965566714e-05,
-      "loss": 0.4007,
+      "epoch": 1.5723270440251573,
+      "grad_norm": 21.78145408630371,
+      "learning_rate": 4.449115044247788e-05,
+      "loss": 0.4651,
       "step": 1500
     },
     {
-      "epoch": 2.0080321285140563,
-      "eval_accuracy": 0.8931323283082077,
-      "eval_f1_class_0": 0.96875,
-      "eval_f1_class_1": 0.7915492957746477,
-      "eval_f1_class_2": 0.8837209302325582,
-      "eval_f1_class_3": 0.838235294117647,
-      "eval_f1_class_4": 0.844574780058651,
-      "eval_f1_class_5": 0.9504,
-      "eval_f1_class_6": 0.9249617151607963,
-      "eval_f1_class_7": 0.9345794392523364,
-      "eval_f1_class_8": 0.9199999999999999,
-      "eval_f1_macro": 0.895196828288515,
-      "eval_f1_micro": 0.8931323283082077,
-      "eval_f1_weighted": 0.8935187314222036,
-      "eval_loss": 0.3943232595920563,
-      "eval_recall_weighted": 0.8931323283082077,
-      "eval_runtime": 52.1055,
-      "eval_samples_per_second": 57.288,
-      "eval_steps_per_second": 1.804,
+      "epoch": 1.5723270440251573,
+      "eval_accuracy": 0.8615626638699528,
+      "eval_f1_class_0": 0.8236877523553162,
+      "eval_f1_class_1": 0.9653259361997226,
+      "eval_f1_class_2": 0.6986128625472887,
+      "eval_f1_class_3": 0.8308457711442785,
+      "eval_f1_class_4": 0.8667563930013459,
+      "eval_f1_class_5": 0.8346456692913385,
+      "eval_f1_class_6": 0.8898305084745762,
+      "eval_f1_class_7": 0.9136420525657072,
+      "eval_f1_class_8": 0.8668280871670703,
+      "eval_f1_class_9": 0.9410150891632374,
+      "eval_f1_macro": 0.8631190121909882,
+      "eval_f1_micro": 0.8615626638699528,
+      "eval_f1_weighted": 0.8631956720576096,
+      "eval_loss": 0.5186640024185181,
+      "eval_recall_weighted": 0.8615626638699528,
+      "eval_runtime": 18.5317,
+      "eval_samples_per_second": 205.809,
+      "eval_steps_per_second": 6.475,
       "step": 1500
     },
     {
-      "epoch": 2.677376171352075,
-      "grad_norm": 2.264118194580078,
-      "learning_rate": 3.926111908177906e-05,
-      "loss": 0.2025,
+      "epoch": 2.0964360587002098,
+      "grad_norm": 6.887778282165527,
+      "learning_rate": 4.172566371681416e-05,
+      "loss": 0.3292,
       "step": 2000
     },
     {
-      "epoch": 2.677376171352075,
-      "eval_accuracy": 0.9165829145728643,
-      "eval_f1_class_0": 0.9734789391575663,
-      "eval_f1_class_1": 0.8170563961485556,
-      "eval_f1_class_2": 0.9121338912133892,
-      "eval_f1_class_3": 0.910384068278805,
-      "eval_f1_class_4": 0.8715447154471545,
-      "eval_f1_class_5": 0.9562499999999999,
-      "eval_f1_class_6": 0.9147058823529411,
-      "eval_f1_class_7": 0.9559748427672956,
-      "eval_f1_class_8": 0.9525368248772503,
-      "eval_f1_macro": 0.918229506693662,
-      "eval_f1_micro": 0.9165829145728643,
-      "eval_f1_weighted": 0.9169458307013622,
-      "eval_loss": 0.392281711101532,
-      "eval_recall_weighted": 0.9165829145728643,
-      "eval_runtime": 52.1029,
-      "eval_samples_per_second": 57.29,
-      "eval_steps_per_second": 1.804,
+      "epoch": 2.0964360587002098,
+      "eval_accuracy": 0.9003670686942842,
+      "eval_f1_class_0": 0.8623188405797102,
+      "eval_f1_class_1": 0.9615384615384616,
+      "eval_f1_class_2": 0.7830045523520485,
+      "eval_f1_class_3": 0.9002624671916011,
+      "eval_f1_class_4": 0.8952618453865336,
+      "eval_f1_class_5": 0.8938271604938272,
+      "eval_f1_class_6": 0.9402173913043479,
+      "eval_f1_class_7": 0.9341935483870967,
+      "eval_f1_class_8": 0.9295039164490861,
+      "eval_f1_class_9": 0.8950131233595802,
+      "eval_f1_macro": 0.8995141307042294,
+      "eval_f1_micro": 0.9003670686942842,
+      "eval_f1_weighted": 0.899989630913346,
+      "eval_loss": 0.37930119037628174,
+      "eval_recall_weighted": 0.9003670686942842,
+      "eval_runtime": 18.5937,
+      "eval_samples_per_second": 205.123,
+      "eval_steps_per_second": 6.454,
       "step": 2000
     },
     {
-      "epoch": 3.3467202141900936,
-      "grad_norm": 9.838397979736328,
-      "learning_rate": 3.5674318507890965e-05,
-      "loss": 0.1351,
+      "epoch": 2.620545073375262,
+      "grad_norm": 18.899120330810547,
+      "learning_rate": 3.896017699115044e-05,
+      "loss": 0.2064,
       "step": 2500
     },
     {
-      "epoch": 3.3467202141900936,
-      "eval_accuracy": 0.9252931323283082,
-      "eval_f1_class_0": 0.9750778816199377,
-      "eval_f1_class_1": 0.8636363636363636,
-      "eval_f1_class_2": 0.909814323607427,
-      "eval_f1_class_3": 0.8973607038123167,
-      "eval_f1_class_4": 0.8852459016393442,
-      "eval_f1_class_5": 0.9549839228295819,
-      "eval_f1_class_6": 0.9319526627218935,
-      "eval_f1_class_7": 0.9635499207606973,
-      "eval_f1_class_8": 0.9556962025316457,
-      "eval_f1_macro": 0.9263686536843565,
-      "eval_f1_micro": 0.9252931323283082,
-      "eval_f1_weighted": 0.9251223151765453,
-      "eval_loss": 0.38727396726608276,
-      "eval_recall_weighted": 0.9252931323283082,
-      "eval_runtime": 52.0796,
-      "eval_samples_per_second": 57.316,
-      "eval_steps_per_second": 1.805,
+      "epoch": 2.620545073375262,
+      "eval_accuracy": 0.916885159937074,
+      "eval_f1_class_0": 0.8660049627791564,
+      "eval_f1_class_1": 0.9670329670329672,
+      "eval_f1_class_2": 0.8037889039242219,
+      "eval_f1_class_3": 0.9132653061224489,
+      "eval_f1_class_4": 0.9252217997465145,
+      "eval_f1_class_5": 0.896719319562576,
+      "eval_f1_class_6": 0.9539295392953929,
+      "eval_f1_class_7": 0.9450000000000001,
+      "eval_f1_class_8": 0.9398601398601398,
+      "eval_f1_class_9": 0.9660056657223797,
+      "eval_f1_macro": 0.9176828604045797,
+      "eval_f1_micro": 0.916885159937074,
+      "eval_f1_weighted": 0.9176253934359663,
+      "eval_loss": 0.34075525403022766,
+      "eval_recall_weighted": 0.916885159937074,
+      "eval_runtime": 18.4847,
+      "eval_samples_per_second": 206.333,
+      "eval_steps_per_second": 6.492,
       "step": 2500
     },
     {
-      "epoch": 4.016064257028113,
-      "grad_norm": 3.1932365894317627,
-      "learning_rate": 3.208751793400287e-05,
-      "loss": 0.1193,
+      "epoch": 3.1446540880503147,
+      "grad_norm": 1.2899887561798096,
+      "learning_rate": 3.619469026548672e-05,
+      "loss": 0.1562,
       "step": 3000
     },
     {
-      "epoch": 4.016064257028113,
-      "eval_accuracy": 0.9366834170854271,
-      "eval_f1_class_0": 0.9753086419753085,
-      "eval_f1_class_1": 0.8781954887218045,
-      "eval_f1_class_2": 0.9465020576131686,
-      "eval_f1_class_3": 0.8873626373626373,
-      "eval_f1_class_4": 0.9245283018867925,
-      "eval_f1_class_5": 0.9478672985781991,
-      "eval_f1_class_6": 0.948301329394387,
-      "eval_f1_class_7": 0.9695999999999999,
-      "eval_f1_class_8": 0.9602543720190778,
-      "eval_f1_macro": 0.9375466808390418,
-      "eval_f1_micro": 0.9366834170854271,
-      "eval_f1_weighted": 0.9367077203287543,
-      "eval_loss": 0.36861082911491394,
-      "eval_recall_weighted": 0.9366834170854271,
-      "eval_runtime": 52.0777,
-      "eval_samples_per_second": 57.318,
-      "eval_steps_per_second": 1.805,
+      "epoch": 3.1446540880503147,
+      "eval_accuracy": 0.9326166754063975,
+      "eval_f1_class_0": 0.8903061224489796,
+      "eval_f1_class_1": 0.9594594594594594,
+      "eval_f1_class_2": 0.8427128427128427,
+      "eval_f1_class_3": 0.9363867684478372,
+      "eval_f1_class_4": 0.9300000000000002,
+      "eval_f1_class_5": 0.9108433734939759,
+      "eval_f1_class_6": 0.9784366576819407,
+      "eval_f1_class_7": 0.9619289340101522,
+      "eval_f1_class_8": 0.9572192513368984,
+      "eval_f1_class_9": 0.9567642956764295,
+      "eval_f1_macro": 0.9324057705268516,
+      "eval_f1_micro": 0.9326166754063975,
+      "eval_f1_weighted": 0.9323951969454738,
+      "eval_loss": 0.29545533657073975,
+      "eval_recall_weighted": 0.9326166754063975,
+      "eval_runtime": 18.4597,
+      "eval_samples_per_second": 206.612,
+      "eval_steps_per_second": 6.501,
       "step": 3000
     },
     {
-      "epoch": 4.685408299866131,
-      "grad_norm": 16.49643325805664,
-      "learning_rate": 2.850071736011478e-05,
-      "loss": 0.063,
+      "epoch": 3.668763102725367,
+      "grad_norm": 7.331942558288574,
+      "learning_rate": 3.342920353982301e-05,
+      "loss": 0.1097,
       "step": 3500
     },
     {
-      "epoch": 4.685408299866131,
-      "eval_accuracy": 0.9396984924623115,
-      "eval_f1_class_0": 0.9751552795031054,
-      "eval_f1_class_1": 0.8555240793201134,
-      "eval_f1_class_2": 0.9222222222222223,
-      "eval_f1_class_3": 0.9184549356223176,
-      "eval_f1_class_4": 0.9200603318250377,
-      "eval_f1_class_5": 0.9730586370839936,
-      "eval_f1_class_6": 0.9650986342943855,
-      "eval_f1_class_7": 0.964968152866242,
-      "eval_f1_class_8": 0.9774193548387097,
-      "eval_f1_macro": 0.9413290697306808,
-      "eval_f1_micro": 0.9396984924623115,
-      "eval_f1_weighted": 0.9399337568761867,
-      "eval_loss": 0.3695576786994934,
-      "eval_recall_weighted": 0.9396984924623115,
-      "eval_runtime": 52.0778,
-      "eval_samples_per_second": 57.318,
-      "eval_steps_per_second": 1.805,
+      "epoch": 3.668763102725367,
+      "eval_accuracy": 0.936811746198217,
+      "eval_f1_class_0": 0.8847497089639115,
+      "eval_f1_class_1": 0.967391304347826,
+      "eval_f1_class_2": 0.8646153846153847,
+      "eval_f1_class_3": 0.9375830013280213,
+      "eval_f1_class_4": 0.9284818067754077,
+      "eval_f1_class_5": 0.9315068493150684,
+      "eval_f1_class_6": 0.9733333333333333,
+      "eval_f1_class_7": 0.9563046192259677,
+      "eval_f1_class_8": 0.9731903485254693,
+      "eval_f1_class_9": 0.9495225102319237,
+      "eval_f1_macro": 0.9366678866662314,
+      "eval_f1_micro": 0.936811746198217,
+      "eval_f1_weighted": 0.9365912559521494,
+      "eval_loss": 0.3290887176990509,
+      "eval_recall_weighted": 0.936811746198217,
+      "eval_runtime": 18.5309,
+      "eval_samples_per_second": 205.818,
+      "eval_steps_per_second": 6.476,
       "step": 3500
+    },
+    {
+      "epoch": 4.1928721174004195,
+      "grad_norm": 0.010240758769214153,
+      "learning_rate": 3.066924778761062e-05,
+      "loss": 0.1123,
+      "step": 4000
+    },
+    {
+      "epoch": 4.1928721174004195,
+      "eval_accuracy": 0.9431043523859465,
+      "eval_f1_class_0": 0.9055118110236221,
+      "eval_f1_class_1": 0.9780821917808219,
+      "eval_f1_class_2": 0.8661870503597122,
+      "eval_f1_class_3": 0.9431524547803619,
+      "eval_f1_class_4": 0.9369592088998764,
+      "eval_f1_class_5": 0.9245742092457422,
+      "eval_f1_class_6": 0.9734042553191489,
+      "eval_f1_class_7": 0.958904109589041,
+      "eval_f1_class_8": 0.9736842105263158,
+      "eval_f1_class_9": 0.9680998613037449,
+      "eval_f1_macro": 0.9428559362828388,
+      "eval_f1_micro": 0.9431043523859465,
+      "eval_f1_weighted": 0.9427303581875908,
+      "eval_loss": 0.3067891001701355,
+      "eval_recall_weighted": 0.9431043523859465,
+      "eval_runtime": 18.589,
+      "eval_samples_per_second": 205.175,
+      "eval_steps_per_second": 6.455,
+      "step": 4000
+    },
+    {
+      "epoch": 4.716981132075472,
+      "grad_norm": 0.4483562707901001,
+      "learning_rate": 2.7903761061946903e-05,
+      "loss": 0.0732,
+      "step": 4500
+    },
+    {
+      "epoch": 4.716981132075472,
+      "eval_accuracy": 0.940744625065548,
+      "eval_f1_class_0": 0.8982826948480845,
+      "eval_f1_class_1": 0.9741496598639455,
+      "eval_f1_class_2": 0.8587570621468926,
+      "eval_f1_class_3": 0.940127388535032,
+      "eval_f1_class_4": 0.9477707006369427,
+      "eval_f1_class_5": 0.9290012033694345,
+      "eval_f1_class_6": 0.972972972972973,
+      "eval_f1_class_7": 0.9485294117647058,
+      "eval_f1_class_8": 0.9693741677762983,
+      "eval_f1_class_9": 0.9666666666666666,
+      "eval_f1_macro": 0.9405631928580975,
+      "eval_f1_micro": 0.940744625065548,
+      "eval_f1_weighted": 0.9405207858122142,
+      "eval_loss": 0.3208290636539459,
+      "eval_recall_weighted": 0.940744625065548,
+      "eval_runtime": 18.5774,
+      "eval_samples_per_second": 205.303,
+      "eval_steps_per_second": 6.459,
+      "step": 4500
+    },
+    {
+      "epoch": 5.241090146750524,
+      "grad_norm": 0.00789484940469265,
+      "learning_rate": 2.5138274336283185e-05,
+      "loss": 0.0582,
+      "step": 5000
+    },
+    {
+      "epoch": 5.241090146750524,
+      "eval_accuracy": 0.9444153120083901,
+      "eval_f1_class_0": 0.8981132075471698,
+      "eval_f1_class_1": 0.9780821917808219,
+      "eval_f1_class_2": 0.8748241912798874,
+      "eval_f1_class_3": 0.9479166666666665,
+      "eval_f1_class_4": 0.9436795994993743,
+      "eval_f1_class_5": 0.942189421894219,
+      "eval_f1_class_6": 0.9613259668508287,
+      "eval_f1_class_7": 0.9551122194513716,
+      "eval_f1_class_8": 0.9713541666666666,
+      "eval_f1_class_9": 0.9721448467966574,
+      "eval_f1_macro": 0.9444742478433662,
+      "eval_f1_micro": 0.9444153120083901,
+      "eval_f1_weighted": 0.9443976005879655,
+      "eval_loss": 0.3178301155567169,
+      "eval_recall_weighted": 0.9444153120083901,
+      "eval_runtime": 18.6189,
+      "eval_samples_per_second": 204.845,
+      "eval_steps_per_second": 6.445,
+      "step": 5000
+    },
+    {
+      "epoch": 5.765199161425577,
+      "grad_norm": 0.2585061490535736,
+      "learning_rate": 2.237278761061947e-05,
+      "loss": 0.0533,
+      "step": 5500
+    },
+    {
+      "epoch": 5.765199161425577,
+      "eval_accuracy": 0.951232302045097,
+      "eval_f1_class_0": 0.9144316730523627,
+      "eval_f1_class_1": 0.9754768392370572,
+      "eval_f1_class_2": 0.8732782369146006,
+      "eval_f1_class_3": 0.9662337662337663,
+      "eval_f1_class_4": 0.958974358974359,
+      "eval_f1_class_5": 0.9410288582183187,
+      "eval_f1_class_6": 0.9840848806366048,
+      "eval_f1_class_7": 0.9597989949748743,
+      "eval_f1_class_8": 0.9712793733681462,
+      "eval_f1_class_9": 0.9667590027700832,
+      "eval_f1_macro": 0.9511345984380173,
+      "eval_f1_micro": 0.9512323020450971,
+      "eval_f1_weighted": 0.9512610025073973,
+      "eval_loss": 0.30544909834861755,
+      "eval_recall_weighted": 0.951232302045097,
+      "eval_runtime": 18.5675,
+      "eval_samples_per_second": 205.412,
+      "eval_steps_per_second": 6.463,
+      "step": 5500
+    },
+    {
+      "epoch": 6.289308176100629,
+      "grad_norm": 0.020399658009409904,
+      "learning_rate": 1.9607300884955755e-05,
+      "loss": 0.0424,
+      "step": 6000
+    },
+    {
+      "epoch": 6.289308176100629,
+      "eval_accuracy": 0.9520188778185632,
+      "eval_f1_class_0": 0.9041450777202072,
+      "eval_f1_class_1": 0.9782016348773842,
+      "eval_f1_class_2": 0.8834019204389574,
+      "eval_f1_class_3": 0.9584415584415584,
+      "eval_f1_class_4": 0.9491094147582698,
+      "eval_f1_class_5": 0.9521472392638036,
+      "eval_f1_class_6": 0.9824086603518267,
+      "eval_f1_class_7": 0.9638854296388544,
+      "eval_f1_class_8": 0.9814323607427056,
+      "eval_f1_class_9": 0.9669421487603306,
+      "eval_f1_macro": 0.9520115444993896,
+      "eval_f1_micro": 0.9520188778185632,
+      "eval_f1_weighted": 0.9520110205225185,
+      "eval_loss": 0.30924680829048157,
+      "eval_recall_weighted": 0.9520188778185632,
+      "eval_runtime": 18.5678,
+      "eval_samples_per_second": 205.41,
+      "eval_steps_per_second": 6.463,
+      "step": 6000
+    },
+    {
+      "epoch": 6.813417190775681,
+      "grad_norm": 3.3116018772125244,
+      "learning_rate": 1.6841814159292034e-05,
+      "loss": 0.0378,
+      "step": 6500
+    },
+    {
+      "epoch": 6.813417190775681,
+      "eval_accuracy": 0.9530676455165181,
+      "eval_f1_class_0": 0.9155107187894074,
+      "eval_f1_class_1": 0.9768707482993196,
+      "eval_f1_class_2": 0.8838526912181304,
+      "eval_f1_class_3": 0.9488491048593349,
+      "eval_f1_class_4": 0.951885565669701,
+      "eval_f1_class_5": 0.9431680773881501,
+      "eval_f1_class_6": 0.9813829787234043,
+      "eval_f1_class_7": 0.9669211195928754,
+      "eval_f1_class_8": 0.9868073878627968,
+      "eval_f1_class_9": 0.9750000000000001,
+      "eval_f1_macro": 0.953024839240312,
+      "eval_f1_micro": 0.9530676455165181,
+      "eval_f1_weighted": 0.952958033459228,
+      "eval_loss": 0.30203908681869507,
+      "eval_recall_weighted": 0.9530676455165181,
+      "eval_runtime": 18.5121,
+      "eval_samples_per_second": 206.028,
+      "eval_steps_per_second": 6.482,
+      "step": 6500
+    },
+    {
+      "epoch": 7.337526205450734,
+      "grad_norm": 0.005160727072507143,
+      "learning_rate": 1.4081858407079645e-05,
+      "loss": 0.0305,
+      "step": 7000
+    },
+    {
+      "epoch": 7.337526205450734,
+      "eval_accuracy": 0.954116413214473,
+      "eval_f1_class_0": 0.9171817058096415,
+      "eval_f1_class_1": 0.9822646657571624,
+      "eval_f1_class_2": 0.8885672937771346,
+      "eval_f1_class_3": 0.95822454308094,
+      "eval_f1_class_4": 0.952020202020202,
+      "eval_f1_class_5": 0.9428918590522478,
+      "eval_f1_class_6": 0.9761273209549071,
+      "eval_f1_class_7": 0.9647355163727961,
+      "eval_f1_class_8": 0.9866310160427807,
+      "eval_f1_class_9": 0.9721448467966574,
+      "eval_f1_macro": 0.9540788969664471,
+      "eval_f1_micro": 0.954116413214473,
+      "eval_f1_weighted": 0.9540130488123523,
+      "eval_loss": 0.28602519631385803,
+      "eval_recall_weighted": 0.954116413214473,
+      "eval_runtime": 18.523,
+      "eval_samples_per_second": 205.906,
+      "eval_steps_per_second": 6.478,
+      "step": 7000
+    },
+    {
+      "epoch": 7.861635220125786,
+      "grad_norm": 0.02748439833521843,
+      "learning_rate": 1.131637168141593e-05,
+      "loss": 0.0216,
+      "step": 7500
+    },
+    {
+      "epoch": 7.861635220125786,
+      "eval_accuracy": 0.9562139486103828,
+      "eval_f1_class_0": 0.9250317662007626,
+      "eval_f1_class_1": 0.9794801641586868,
+      "eval_f1_class_2": 0.8888888888888887,
+      "eval_f1_class_3": 0.9527458492975734,
+      "eval_f1_class_4": 0.9604086845466155,
+      "eval_f1_class_5": 0.9476248477466506,
+      "eval_f1_class_6": 0.9853137516688919,
+      "eval_f1_class_7": 0.9682337992376113,
+      "eval_f1_class_8": 0.9790575916230366,
+      "eval_f1_class_9": 0.9736477115117891,
+      "eval_f1_macro": 0.9560433054880507,
+      "eval_f1_micro": 0.9562139486103828,
+      "eval_f1_weighted": 0.956078385428201,
+      "eval_loss": 0.303732693195343,
+      "eval_recall_weighted": 0.9562139486103828,
+      "eval_runtime": 18.6679,
+      "eval_samples_per_second": 204.308,
+      "eval_steps_per_second": 6.428,
+      "step": 7500
+    },
+    {
+      "epoch": 8.385744234800839,
+      "grad_norm": 0.017909903079271317,
+      "learning_rate": 8.550884955752212e-06,
+      "loss": 0.0198,
+      "step": 8000
+    },
+    {
+      "epoch": 8.385744234800839,
+      "eval_accuracy": 0.9562139486103828,
+      "eval_f1_class_0": 0.9207547169811321,
+      "eval_f1_class_1": 0.9795361527967258,
+      "eval_f1_class_2": 0.8829337094499294,
+      "eval_f1_class_3": 0.9623865110246432,
+      "eval_f1_class_4": 0.9592875318066159,
+      "eval_f1_class_5": 0.9451887941534713,
+      "eval_f1_class_6": 0.9826897470039946,
+      "eval_f1_class_7": 0.9707006369426752,
+      "eval_f1_class_8": 0.9855072463768115,
+      "eval_f1_class_9": 0.9721448467966574,
+      "eval_f1_macro": 0.9561129893332657,
+      "eval_f1_micro": 0.9562139486103828,
+      "eval_f1_weighted": 0.9561806063383737,
+      "eval_loss": 0.2868480682373047,
+      "eval_recall_weighted": 0.9562139486103828,
+      "eval_runtime": 18.5749,
+      "eval_samples_per_second": 205.331,
+      "eval_steps_per_second": 6.46,
+      "step": 8000
+    },
+    {
+      "epoch": 8.90985324947589,
+      "grad_norm": 0.019022395834326744,
+      "learning_rate": 5.785398230088496e-06,
+      "loss": 0.0145,
+      "step": 8500
+    },
+    {
+      "epoch": 8.90985324947589,
+      "eval_accuracy": 0.9580492920818039,
+      "eval_f1_class_0": 0.9221260815822002,
+      "eval_f1_class_1": 0.9754768392370572,
+      "eval_f1_class_2": 0.8932584269662921,
+      "eval_f1_class_3": 0.9621903520208606,
+      "eval_f1_class_4": 0.9541984732824428,
+      "eval_f1_class_5": 0.957920792079208,
+      "eval_f1_class_6": 0.9879518072289156,
+      "eval_f1_class_7": 0.9671717171717172,
+      "eval_f1_class_8": 0.9867724867724869,
+      "eval_f1_class_9": 0.9735006973500697,
+      "eval_f1_macro": 0.9580567673691249,
+      "eval_f1_micro": 0.9580492920818039,
+      "eval_f1_weighted": 0.9581079328217166,
+      "eval_loss": 0.29160091280937195,
+      "eval_recall_weighted": 0.9580492920818039,
+      "eval_runtime": 18.5713,
+      "eval_samples_per_second": 205.371,
+      "eval_steps_per_second": 6.462,
+      "step": 8500
+    },
+    {
+      "epoch": 9.433962264150944,
+      "grad_norm": 0.0038110397290438414,
+      "learning_rate": 3.019911504424779e-06,
+      "loss": 0.014,
+      "step": 9000
+    },
+    {
+      "epoch": 9.433962264150944,
+      "eval_accuracy": 0.959884635553225,
+      "eval_f1_class_0": 0.9228855721393034,
+      "eval_f1_class_1": 0.9809264305177112,
+      "eval_f1_class_2": 0.9080459770114943,
+      "eval_f1_class_3": 0.9571984435797665,
+      "eval_f1_class_4": 0.9566326530612245,
+      "eval_f1_class_5": 0.9585365853658537,
+      "eval_f1_class_6": 0.9840848806366048,
+      "eval_f1_class_7": 0.9745547073791349,
+      "eval_f1_class_8": 0.9868073878627968,
+      "eval_f1_class_9": 0.9680998613037449,
+      "eval_f1_macro": 0.9597772498857635,
+      "eval_f1_micro": 0.959884635553225,
+      "eval_f1_weighted": 0.9597604965241316,
+      "eval_loss": 0.28916341066360474,
+      "eval_recall_weighted": 0.959884635553225,
+      "eval_runtime": 18.5578,
+      "eval_samples_per_second": 205.52,
+      "eval_steps_per_second": 6.466,
+      "step": 9000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 7470,
+  "max_steps": 9540,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
@@ -229,7 +566,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.47218650942464e+16,
+  "total_flos": 3.787923857566925e+16,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null