upload model weights after 1200 training steps with frozen pretrained weights

checkpoint-1000/config.json

@@ -7,23 +7,21 @@
   "classifier_dropout": null,
   "dtype": "float32",
   "eos_token_id": 2,
-  "gradient_checkpointing": false,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
   "hidden_size": 768,
   "initializer_range": 0.02,
   "intermediate_size": 3072,
   "layer_norm_eps": 1e-05,
-  "max_position_embeddings": 1026,
+  "max_position_embeddings": 514,
   "model_type": "roberta",
   "num_attention_heads": 12,
   "num_hidden_layers": 12,
   "output_past": true,
   "pad_token_id": 1,
   "position_embedding_type": "absolute",
-  "problem_type": "single_label_classification",
   "transformers_version": "4.56.0",
-  "type_vocab_size": 10,
+  "type_vocab_size": 1,
   "use_cache": true,
-  "vocab_size": 51416
+  "vocab_size": 50265
 }

checkpoint-1000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9dba0da560adadcbd863063b6d8a4d4b56290c3be335cf41f14a557b6fee1c9f
-size 506505672
+oid sha256:665d1f6d7cabc187a2c05357e277b61c6a5caafa84e641cdaa9f74e1fc8f6b9e
+size 498612824

checkpoint-1000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6bdf390e9c9e5d555f636219141a77d3bb7ad8de7356aaf171bdfbbbf56a7182
-size 1007622411
+oid sha256:8df0daec4efa812ba208cd0dc90879540bae0bc13e657a311777d25929b1aea1
+size 4741859

checkpoint-1000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5183674b6e19178da7160a0d79f26425cdd2d73db6134e2a7310a7b772fca618
+oid sha256:0ea1b5dc44e07bb7fcc644910a14b04ebd8863c5a7cf8eb7438e00e0c30f16a6
 size 14645

checkpoint-1000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5c69cb9ab6a70db20a2d0ee7c18ddc9292442d3eec7b5ec05df69a67d327cf4b
-size 1977
+oid sha256:9f25fe7731b9278a60bff9cc7fd3ead4df1aef9e341ef501c56c76de421b4c24
+size 1465

checkpoint-1000/special_tokens_map.json

@@ -23,7 +23,7 @@
   "mask_token": {
     "content": "<mask>",
     "lstrip": true,
-    "normalized": true,
+    "normalized": false,
     "rstrip": false,
     "single_word": false
   },

checkpoint-1000/tokenizer_config.json

@@ -33,10 +33,10 @@
       "single_word": false,
       "special": true
     },
-    "4": {
+    "50264": {
       "content": "<mask>",
       "lstrip": true,
-      "normalized": true,
+      "normalized": false,
       "rstrip": false,
       "single_word": false,
       "special": true
@@ -49,7 +49,7 @@
   "errors": "replace",
   "extra_special_tokens": {},
   "mask_token": "<mask>",
-  "model_max_length": 1000000000000000019884624838656,
+  "model_max_length": 512,
   "pad_token": "<pad>",
   "sep_token": "</s>",
   "tokenizer_class": "RobertaTokenizer",

checkpoint-1000/trainer_state.json

@@ -1,850 +1,780 @@
 {
   "best_global_step": 1000,
-  "best_metric": 0.9644889377060355,
-  "best_model_checkpoint": "taskA-unixcoder-focal/checkpoint-1000",
-  "epoch": 0.1223540927444023,
-  "eval_steps": 100,
+  "best_metric": 0.689588440856387,
+  "best_model_checkpoint": "./taskA-codebert-base/checkpoint-1000",
+  "epoch": 0.12799180852425446,
+  "eval_steps": 200,
   "global_step": 1000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.001223540927444023,
-      "grad_norm": 0.1898110806941986,
-      "learning_rate": 3.6734693877551025e-07,
-      "loss": 0.4422,
+      "epoch": 0.0012799180852425445,
+      "grad_norm": 3.6954128742218018,
+      "learning_rate": 3.840409643695328e-08,
+      "loss": 0.7763,
       "step": 10
     },
     {
-      "epoch": 0.002447081854888046,
-      "grad_norm": 0.3350900411605835,
-      "learning_rate": 7.346938775510205e-07,
-      "loss": 0.4323,
+      "epoch": 0.002559836170485089,
+      "grad_norm": 3.3660645484924316,
+      "learning_rate": 8.10753147002347e-08,
+      "loss": 0.7763,
       "step": 20
     },
     {
-      "epoch": 0.003670622782332069,
-      "grad_norm": 0.3723449110984802,
-      "learning_rate": 1.142857142857143e-06,
-      "loss": 0.4362,
+      "epoch": 0.0038397542557276334,
+      "grad_norm": 4.491067886352539,
+      "learning_rate": 1.2374653296351612e-07,
+      "loss": 0.7889,
       "step": 30
     },
     {
-      "epoch": 0.004894163709776092,
-      "grad_norm": 0.3299769461154938,
-      "learning_rate": 1.5510204081632654e-06,
-      "loss": 0.4191,
+      "epoch": 0.005119672340970178,
+      "grad_norm": 1.5920521020889282,
+      "learning_rate": 1.6641775122679754e-07,
+      "loss": 0.7888,
       "step": 40
     },
     {
-      "epoch": 0.006117704637220115,
-      "grad_norm": 0.26681044697761536,
-      "learning_rate": 1.9591836734693877e-06,
-      "loss": 0.4289,
+      "epoch": 0.006399590426212722,
+      "grad_norm": 4.410898208618164,
+      "learning_rate": 2.0908896949007894e-07,
+      "loss": 0.7692,
       "step": 50
     },
     {
-      "epoch": 0.007341245564664138,
-      "grad_norm": 0.30233338475227356,
-      "learning_rate": 2.3673469387755103e-06,
-      "loss": 0.4083,
+      "epoch": 0.007679508511455267,
+      "grad_norm": 2.5669264793395996,
+      "learning_rate": 2.517601877533604e-07,
+      "loss": 0.7735,
       "step": 60
     },
     {
-      "epoch": 0.008564786492108161,
-      "grad_norm": 0.31031334400177,
-      "learning_rate": 2.775510204081633e-06,
-      "loss": 0.404,
+      "epoch": 0.008959426596697812,
+      "grad_norm": 6.190986156463623,
+      "learning_rate": 2.944314060166418e-07,
+      "loss": 0.7679,
       "step": 70
     },
     {
-      "epoch": 0.009788327419552184,
-      "grad_norm": 0.2552475035190582,
-      "learning_rate": 3.183673469387755e-06,
-      "loss": 0.3889,
+      "epoch": 0.010239344681940356,
+      "grad_norm": 5.736413478851318,
+      "learning_rate": 3.371026242799232e-07,
+      "loss": 0.744,
       "step": 80
     },
     {
-      "epoch": 0.011011868346996207,
-      "grad_norm": 0.3131052553653717,
-      "learning_rate": 3.5918367346938777e-06,
-      "loss": 0.3813,
+      "epoch": 0.011519262767182901,
+      "grad_norm": 4.278866767883301,
+      "learning_rate": 3.7977384254320464e-07,
+      "loss": 0.7542,
       "step": 90
     },
     {
-      "epoch": 0.01223540927444023,
-      "grad_norm": 0.3377229571342468,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 0.3727,
+      "epoch": 0.012799180852425445,
+      "grad_norm": 5.565455436706543,
+      "learning_rate": 4.22445060806486e-07,
+      "loss": 0.752,
       "step": 100
     },
     {
-      "epoch": 0.01223540927444023,
-      "eval_accuracy": 0.6961,
-      "eval_f1_weighted": 0.6955947095331424,
-      "eval_loss": 0.15026023983955383,
-      "eval_macro_f1": 0.6959280508399847,
-      "eval_precision": 0.7019396279846728,
-      "eval_recall": 0.6961,
-      "eval_runtime": 789.7648,
-      "eval_samples_per_second": 126.62,
-      "eval_steps_per_second": 0.99,
-      "step": 100
-    },
-    {
-      "epoch": 0.013458950201884253,
-      "grad_norm": 0.3366774618625641,
-      "learning_rate": 4.408163265306123e-06,
-      "loss": 0.3671,
+      "epoch": 0.01407909893766799,
+      "grad_norm": 4.031781196594238,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 0.7614,
       "step": 110
     },
     {
-      "epoch": 0.014682491129328276,
-      "grad_norm": 0.33491575717926025,
-      "learning_rate": 4.816326530612245e-06,
-      "loss": 0.3498,
+      "epoch": 0.015359017022910534,
+      "grad_norm": 2.484126329421997,
+      "learning_rate": 5.077874973330489e-07,
+      "loss": 0.7664,
       "step": 120
     },
     {
-      "epoch": 0.0159060320567723,
-      "grad_norm": 0.4119006097316742,
-      "learning_rate": 5.2244897959183684e-06,
-      "loss": 0.3409,
+      "epoch": 0.016638935108153077,
+      "grad_norm": 2.7525155544281006,
+      "learning_rate": 5.504587155963304e-07,
+      "loss": 0.7521,
       "step": 130
     },
     {
-      "epoch": 0.017129572984216322,
-      "grad_norm": 0.2964725196361542,
-      "learning_rate": 5.63265306122449e-06,
-      "loss": 0.3433,
+      "epoch": 0.017918853193395624,
+      "grad_norm": 1.6374385356903076,
+      "learning_rate": 5.931299338596117e-07,
+      "loss": 0.7436,
       "step": 140
     },
     {
-      "epoch": 0.018353113911660345,
-      "grad_norm": 0.3075334429740906,
-      "learning_rate": 6.040816326530613e-06,
-      "loss": 0.3298,
+      "epoch": 0.019198771278638168,
+      "grad_norm": 2.370774269104004,
+      "learning_rate": 6.358011521228932e-07,
+      "loss": 0.7429,
       "step": 150
     },
     {
-      "epoch": 0.019576654839104368,
-      "grad_norm": 0.33718687295913696,
-      "learning_rate": 6.4489795918367345e-06,
-      "loss": 0.3297,
+      "epoch": 0.02047868936388071,
+      "grad_norm": 2.8948566913604736,
+      "learning_rate": 6.784723703861745e-07,
+      "loss": 0.7333,
       "step": 160
     },
     {
-      "epoch": 0.02080019576654839,
-      "grad_norm": 0.2809618413448334,
-      "learning_rate": 6.857142857142858e-06,
-      "loss": 0.3275,
+      "epoch": 0.021758607449123255,
+      "grad_norm": 2.7481186389923096,
+      "learning_rate": 7.21143588649456e-07,
+      "loss": 0.7294,
       "step": 170
     },
     {
-      "epoch": 0.022023736693992414,
-      "grad_norm": 0.37805014848709106,
-      "learning_rate": 7.2653061224489805e-06,
-      "loss": 0.3201,
+      "epoch": 0.023038525534365802,
+      "grad_norm": 2.0159404277801514,
+      "learning_rate": 7.638148069127374e-07,
+      "loss": 0.731,
       "step": 180
     },
     {
-      "epoch": 0.023247277621436437,
-      "grad_norm": 0.2940429151058197,
-      "learning_rate": 7.673469387755102e-06,
-      "loss": 0.3042,
+      "epoch": 0.024318443619608346,
+      "grad_norm": 3.6898131370544434,
+      "learning_rate": 8.064860251760189e-07,
+      "loss": 0.7153,
       "step": 190
     },
     {
-      "epoch": 0.02447081854888046,
-      "grad_norm": 0.3286060690879822,
-      "learning_rate": 8.081632653061225e-06,
-      "loss": 0.3071,
+      "epoch": 0.02559836170485089,
+      "grad_norm": 3.2842342853546143,
+      "learning_rate": 8.491572434393003e-07,
+      "loss": 0.7245,
       "step": 200
     },
     {
-      "epoch": 0.02447081854888046,
-      "eval_accuracy": 0.77807,
-      "eval_f1_weighted": 0.7781129395311969,
-      "eval_loss": 0.12153773754835129,
-      "eval_macro_f1": 0.7780651598452646,
-      "eval_precision": 0.7808306361972198,
-      "eval_recall": 0.77807,
-      "eval_runtime": 792.7157,
-      "eval_samples_per_second": 126.149,
-      "eval_steps_per_second": 0.986,
+      "epoch": 0.02559836170485089,
+      "eval_accuracy": 0.47695,
+      "eval_loss": 0.7148504853248596,
+      "eval_macro_f1": 0.3229290091066048,
+      "eval_precision": 0.238475,
+      "eval_recall": 0.5,
+      "eval_runtime": 1641.324,
+      "eval_samples_per_second": 60.926,
+      "eval_steps_per_second": 0.476,
       "step": 200
     },
     {
-      "epoch": 0.025694359476324483,
-      "grad_norm": 0.33857154846191406,
-      "learning_rate": 8.489795918367347e-06,
-      "loss": 0.2964,
+      "epoch": 0.026878279790093433,
+      "grad_norm": 2.294558048248291,
+      "learning_rate": 8.918284617025817e-07,
+      "loss": 0.718,
       "step": 210
     },
     {
-      "epoch": 0.026917900403768506,
-      "grad_norm": 0.2752092182636261,
-      "learning_rate": 8.89795918367347e-06,
-      "loss": 0.3,
+      "epoch": 0.02815819787533598,
+      "grad_norm": 2.9467175006866455,
+      "learning_rate": 9.344996799658632e-07,
+      "loss": 0.7076,
       "step": 220
     },
     {
-      "epoch": 0.02814144133121253,
-      "grad_norm": 0.4952544569969177,
-      "learning_rate": 9.306122448979593e-06,
-      "loss": 0.294,
+      "epoch": 0.029438115960578524,
+      "grad_norm": 2.6889193058013916,
+      "learning_rate": 9.771708982291445e-07,
+      "loss": 0.7019,
       "step": 230
     },
     {
-      "epoch": 0.029364982258656552,
-      "grad_norm": 0.37926042079925537,
-      "learning_rate": 9.714285714285715e-06,
-      "loss": 0.2964,
+      "epoch": 0.030718034045821067,
+      "grad_norm": 0.7491384744644165,
+      "learning_rate": 1.0198421164924258e-06,
+      "loss": 0.7046,
       "step": 240
     },
     {
-      "epoch": 0.030588523186100575,
-      "grad_norm": 0.3222455680370331,
-      "learning_rate": 1.0122448979591836e-05,
-      "loss": 0.283,
+      "epoch": 0.031997952131063614,
+      "grad_norm": 2.0363433361053467,
+      "learning_rate": 1.0625133347557074e-06,
+      "loss": 0.7085,
       "step": 250
     },
     {
-      "epoch": 0.0318120641135446,
-      "grad_norm": 0.2930658459663391,
-      "learning_rate": 1.053061224489796e-05,
-      "loss": 0.2793,
+      "epoch": 0.033277870216306155,
+      "grad_norm": 1.1431337594985962,
+      "learning_rate": 1.1051845530189888e-06,
+      "loss": 0.7008,
       "step": 260
     },
     {
-      "epoch": 0.03303560504098862,
-      "grad_norm": 0.4313414692878723,
-      "learning_rate": 1.0938775510204083e-05,
-      "loss": 0.2728,
+      "epoch": 0.0345577883015487,
+      "grad_norm": 3.7735233306884766,
+      "learning_rate": 1.1478557712822702e-06,
+      "loss": 0.7015,
       "step": 270
     },
     {
-      "epoch": 0.034259145968432644,
-      "grad_norm": 0.37351569533348083,
-      "learning_rate": 1.1346938775510206e-05,
-      "loss": 0.2752,
+      "epoch": 0.03583770638679125,
+      "grad_norm": 0.7889328002929688,
+      "learning_rate": 1.1905269895455517e-06,
+      "loss": 0.699,
       "step": 280
     },
     {
-      "epoch": 0.035482686895876664,
-      "grad_norm": 0.22772285342216492,
-      "learning_rate": 1.1755102040816328e-05,
-      "loss": 0.274,
+      "epoch": 0.03711762447203379,
+      "grad_norm": 0.8796226382255554,
+      "learning_rate": 1.233198207808833e-06,
+      "loss": 0.6922,
       "step": 290
     },
     {
-      "epoch": 0.03670622782332069,
-      "grad_norm": 0.15888753533363342,
-      "learning_rate": 1.2163265306122449e-05,
-      "loss": 0.264,
-      "step": 300
-    },
-    {
-      "epoch": 0.03670622782332069,
-      "eval_accuracy": 0.82326,
-      "eval_f1_weighted": 0.8227034206790359,
-      "eval_loss": 0.10234300792217255,
-      "eval_macro_f1": 0.8230100448030374,
-      "eval_precision": 0.8338333810964687,
-      "eval_recall": 0.82326,
-      "eval_runtime": 789.0168,
-      "eval_samples_per_second": 126.74,
-      "eval_steps_per_second": 0.991,
+      "epoch": 0.038397542557276336,
+      "grad_norm": 0.7966652512550354,
+      "learning_rate": 1.2758694260721145e-06,
+      "loss": 0.6949,
       "step": 300
     },
     {
-      "epoch": 0.03792976875076471,
-      "grad_norm": 0.23560188710689545,
-      "learning_rate": 1.2571428571428572e-05,
-      "loss": 0.2753,
+      "epoch": 0.039677460642518876,
+      "grad_norm": 0.7029069066047668,
+      "learning_rate": 1.318540644335396e-06,
+      "loss": 0.695,
       "step": 310
     },
     {
-      "epoch": 0.039153309678208736,
-      "grad_norm": 0.18699614703655243,
-      "learning_rate": 1.2979591836734696e-05,
-      "loss": 0.2565,
+      "epoch": 0.04095737872776142,
+      "grad_norm": 2.04327392578125,
+      "learning_rate": 1.3612118625986772e-06,
+      "loss": 0.6944,
       "step": 320
     },
     {
-      "epoch": 0.040376850605652756,
-      "grad_norm": 0.24035991728305817,
-      "learning_rate": 1.3387755102040818e-05,
-      "loss": 0.2716,
+      "epoch": 0.04223729681300397,
+      "grad_norm": 0.918025553226471,
+      "learning_rate": 1.4038830808619588e-06,
+      "loss": 0.6863,
       "step": 330
     },
     {
-      "epoch": 0.04160039153309678,
-      "grad_norm": 0.15309660136699677,
-      "learning_rate": 1.3795918367346941e-05,
-      "loss": 0.2515,
+      "epoch": 0.04351721489824651,
+      "grad_norm": 0.9528706073760986,
+      "learning_rate": 1.4465542991252401e-06,
+      "loss": 0.6907,
       "step": 340
     },
     {
-      "epoch": 0.0428239324605408,
-      "grad_norm": 0.3382055461406708,
-      "learning_rate": 1.4204081632653062e-05,
-      "loss": 0.2597,
+      "epoch": 0.04479713298348906,
+      "grad_norm": 1.071359634399414,
+      "learning_rate": 1.4892255173885215e-06,
+      "loss": 0.6873,
       "step": 350
     },
     {
-      "epoch": 0.04404747338798483,
-      "grad_norm": 0.24439455568790436,
-      "learning_rate": 1.4612244897959185e-05,
-      "loss": 0.247,
+      "epoch": 0.046077051068731605,
+      "grad_norm": 1.9875625371932983,
+      "learning_rate": 1.531896735651803e-06,
+      "loss": 0.6942,
       "step": 360
     },
     {
-      "epoch": 0.04527101431542885,
-      "grad_norm": 0.26233720779418945,
-      "learning_rate": 1.5020408163265307e-05,
-      "loss": 0.2431,
+      "epoch": 0.047356969153974145,
+      "grad_norm": 2.0222439765930176,
+      "learning_rate": 1.5745679539150842e-06,
+      "loss": 0.6898,
       "step": 370
     },
     {
-      "epoch": 0.046494555242872875,
-      "grad_norm": 0.20477451384067535,
-      "learning_rate": 1.5387755102040818e-05,
-      "loss": 0.2538,
+      "epoch": 0.04863688723921669,
+      "grad_norm": 1.7827857732772827,
+      "learning_rate": 1.6172391721783658e-06,
+      "loss": 0.6909,
       "step": 380
     },
     {
-      "epoch": 0.047718096170316894,
-      "grad_norm": 0.2765439748764038,
-      "learning_rate": 1.5795918367346942e-05,
-      "loss": 0.2404,
+      "epoch": 0.04991680532445923,
+      "grad_norm": 1.065674901008606,
+      "learning_rate": 1.6599103904416472e-06,
+      "loss": 0.6877,
       "step": 390
     },
     {
-      "epoch": 0.04894163709776092,
-      "grad_norm": 0.13196279108524323,
-      "learning_rate": 1.6204081632653063e-05,
-      "loss": 0.2393,
+      "epoch": 0.05119672340970178,
+      "grad_norm": 0.8346193432807922,
+      "learning_rate": 1.7025816087049288e-06,
+      "loss": 0.6889,
       "step": 400
     },
     {
-      "epoch": 0.04894163709776092,
-      "eval_accuracy": 0.84563,
-      "eval_f1_weighted": 0.8448128964509274,
-      "eval_loss": 0.09028046578168869,
-      "eval_macro_f1": 0.8451923512067675,
-      "eval_precision": 0.861024732271907,
-      "eval_recall": 0.84563,
-      "eval_runtime": 787.6067,
-      "eval_samples_per_second": 126.967,
-      "eval_steps_per_second": 0.993,
+      "epoch": 0.05119672340970178,
+      "eval_accuracy": 0.54043,
+      "eval_loss": 0.6827595233917236,
+      "eval_macro_f1": 0.41833840347598905,
+      "eval_precision": 0.5839205994464474,
+      "eval_recall": 0.5204152767503025,
+      "eval_runtime": 1634.6244,
+      "eval_samples_per_second": 61.176,
+      "eval_steps_per_second": 0.478,
       "step": 400
     },
     {
-      "epoch": 0.05016517802520494,
-      "grad_norm": 0.0924634337425232,
-      "learning_rate": 1.6612244897959184e-05,
-      "loss": 0.2331,
+      "epoch": 0.052476641494944326,
+      "grad_norm": 0.8074935674667358,
+      "learning_rate": 1.7452528269682101e-06,
+      "loss": 0.684,
       "step": 410
     },
     {
-      "epoch": 0.05138871895264897,
-      "grad_norm": 0.277784138917923,
-      "learning_rate": 1.7020408163265308e-05,
-      "loss": 0.2301,
+      "epoch": 0.053756559580186866,
+      "grad_norm": 0.8965069651603699,
+      "learning_rate": 1.7879240452314913e-06,
+      "loss": 0.6922,
       "step": 420
     },
     {
-      "epoch": 0.052612259880092986,
-      "grad_norm": 0.32299959659576416,
-      "learning_rate": 1.742857142857143e-05,
-      "loss": 0.2218,
+      "epoch": 0.05503647766542941,
+      "grad_norm": 2.597215414047241,
+      "learning_rate": 1.8305952634947729e-06,
+      "loss": 0.6909,
       "step": 430
     },
     {
-      "epoch": 0.05383580080753701,
-      "grad_norm": 0.18170586228370667,
-      "learning_rate": 1.7836734693877553e-05,
-      "loss": 0.2318,
+      "epoch": 0.05631639575067196,
+      "grad_norm": 0.8233373165130615,
+      "learning_rate": 1.8732664817580542e-06,
+      "loss": 0.6831,
       "step": 440
     },
     {
-      "epoch": 0.05505934173498103,
-      "grad_norm": 0.1520216166973114,
-      "learning_rate": 1.8244897959183674e-05,
-      "loss": 0.2136,
+      "epoch": 0.0575963138359145,
+      "grad_norm": 1.0624489784240723,
+      "learning_rate": 1.915937700021336e-06,
+      "loss": 0.6817,
       "step": 450
     },
     {
-      "epoch": 0.05628288266242506,
-      "grad_norm": 0.23452645540237427,
-      "learning_rate": 1.8653061224489795e-05,
-      "loss": 0.2219,
+      "epoch": 0.05887623192115705,
+      "grad_norm": 2.0623559951782227,
+      "learning_rate": 1.958608918284617e-06,
+      "loss": 0.6819,
       "step": 460
     },
     {
-      "epoch": 0.05750642358986908,
-      "grad_norm": 0.10036301612854004,
-      "learning_rate": 1.906122448979592e-05,
-      "loss": 0.2213,
+      "epoch": 0.06015615000639959,
+      "grad_norm": 0.912433922290802,
+      "learning_rate": 2.0012801365478988e-06,
+      "loss": 0.686,
       "step": 470
     },
     {
-      "epoch": 0.058729964517313105,
-      "grad_norm": 0.24100220203399658,
-      "learning_rate": 1.9469387755102044e-05,
-      "loss": 0.2079,
+      "epoch": 0.061436068091642135,
+      "grad_norm": 0.6926271319389343,
+      "learning_rate": 2.04395135481118e-06,
+      "loss": 0.6834,
       "step": 480
     },
     {
-      "epoch": 0.059953505444757124,
-      "grad_norm": 0.28893032670021057,
-      "learning_rate": 1.9877551020408165e-05,
-      "loss": 0.1978,
+      "epoch": 0.06271598617688467,
+      "grad_norm": 1.498634934425354,
+      "learning_rate": 2.0866225730744615e-06,
+      "loss": 0.6805,
       "step": 490
     },
     {
-      "epoch": 0.06117704637220115,
-      "grad_norm": 0.22918537259101868,
-      "learning_rate": 1.9999836143766126e-05,
-      "loss": 0.202,
+      "epoch": 0.06399590426212723,
+      "grad_norm": 1.1818156242370605,
+      "learning_rate": 2.129293791337743e-06,
+      "loss": 0.6805,
       "step": 500
     },
     {
-      "epoch": 0.06117704637220115,
-      "eval_accuracy": 0.91628,
-      "eval_f1_weighted": 0.9163231793856368,
-      "eval_loss": 0.055908624082803726,
-      "eval_macro_f1": 0.9162187423616152,
-      "eval_precision": 0.9170320825159874,
-      "eval_recall": 0.91628,
-      "eval_runtime": 760.1138,
-      "eval_samples_per_second": 131.559,
-      "eval_steps_per_second": 1.029,
-      "step": 500
-    },
-    {
-      "epoch": 0.06240058729964517,
-      "grad_norm": 0.2930404841899872,
-      "learning_rate": 1.999903359554741e-05,
-      "loss": 0.1979,
+      "epoch": 0.06527582234736977,
+      "grad_norm": 1.4476051330566406,
+      "learning_rate": 2.1719650096010242e-06,
+      "loss": 0.6773,
       "step": 510
     },
     {
-      "epoch": 0.0636241282270892,
-      "grad_norm": 0.0889664739370346,
-      "learning_rate": 1.9997562312908153e-05,
-      "loss": 0.1953,
+      "epoch": 0.06655574043261231,
+      "grad_norm": 1.1143743991851807,
+      "learning_rate": 2.214636227864306e-06,
+      "loss": 0.6801,
       "step": 520
     },
     {
-      "epoch": 0.06484766915453322,
-      "grad_norm": 0.1687588095664978,
-      "learning_rate": 1.9995422394247598e-05,
-      "loss": 0.1935,
+      "epoch": 0.06783565851785486,
+      "grad_norm": 1.7030247449874878,
+      "learning_rate": 2.257307446127587e-06,
+      "loss": 0.6794,
       "step": 530
     },
     {
-      "epoch": 0.06607121008197724,
-      "grad_norm": 0.2399725466966629,
-      "learning_rate": 1.9992613982683296e-05,
-      "loss": 0.1818,
+      "epoch": 0.0691155766030974,
+      "grad_norm": 0.8592290282249451,
+      "learning_rate": 2.2999786643908685e-06,
+      "loss": 0.6736,
       "step": 540
     },
     {
-      "epoch": 0.06729475100942127,
-      "grad_norm": 0.2016785889863968,
-      "learning_rate": 1.9989137266041552e-05,
-      "loss": 0.1893,
+      "epoch": 0.07039549468833994,
+      "grad_norm": 0.8231940865516663,
+      "learning_rate": 2.34264988265415e-06,
+      "loss": 0.67,
       "step": 550
     },
     {
-      "epoch": 0.06851829193686529,
-      "grad_norm": 0.17420713603496552,
-      "learning_rate": 1.9984992476844845e-05,
-      "loss": 0.1895,
+      "epoch": 0.0716754127735825,
+      "grad_norm": 0.88897705078125,
+      "learning_rate": 2.3853211009174317e-06,
+      "loss": 0.685,
       "step": 560
     },
     {
-      "epoch": 0.06974183286430931,
-      "grad_norm": 0.2180609554052353,
-      "learning_rate": 1.998017989229629e-05,
-      "loss": 0.1836,
+      "epoch": 0.07295533085882504,
+      "grad_norm": 1.6297541856765747,
+      "learning_rate": 2.427992319180713e-06,
+      "loss": 0.6797,
       "step": 570
     },
     {
-      "epoch": 0.07096537379175333,
-      "grad_norm": 0.15563039481639862,
-      "learning_rate": 1.9974699834261074e-05,
-      "loss": 0.182,
+      "epoch": 0.07423524894406758,
+      "grad_norm": 2.9297447204589844,
+      "learning_rate": 2.470663537443994e-06,
+      "loss": 0.6766,
       "step": 580
     },
     {
-      "epoch": 0.07218891471919736,
-      "grad_norm": 0.25410088896751404,
-      "learning_rate": 1.996855266924497e-05,
-      "loss": 0.1862,
+      "epoch": 0.07551516702931012,
+      "grad_norm": 0.9277703762054443,
+      "learning_rate": 2.5133347557072756e-06,
+      "loss": 0.6824,
       "step": 590
     },
     {
-      "epoch": 0.07341245564664138,
-      "grad_norm": 0.24206259846687317,
-      "learning_rate": 1.996173880836979e-05,
-      "loss": 0.1818,
+      "epoch": 0.07679508511455267,
+      "grad_norm": 0.5883333683013916,
+      "learning_rate": 2.556005973970557e-06,
+      "loss": 0.6818,
       "step": 600
     },
     {
-      "epoch": 0.07341245564664138,
-      "eval_accuracy": 0.91633,
-      "eval_f1_weighted": 0.9162437523349878,
-      "eval_loss": 0.05625538155436516,
-      "eval_macro_f1": 0.9163072974421806,
-      "eval_precision": 0.9233764457540103,
-      "eval_recall": 0.91633,
-      "eval_runtime": 759.6815,
-      "eval_samples_per_second": 131.634,
-      "eval_steps_per_second": 1.029,
+      "epoch": 0.07679508511455267,
+      "eval_accuracy": 0.63909,
+      "eval_loss": 0.6700075268745422,
+      "eval_macro_f1": 0.621848143053056,
+      "eval_precision": 0.6524313071088819,
+      "eval_recall": 0.6305863955136095,
+      "eval_runtime": 1633.9732,
+      "eval_samples_per_second": 61.201,
+      "eval_steps_per_second": 0.479,
       "step": 600
     },
     {
-      "epoch": 0.0746359965740854,
-      "grad_norm": 0.27251410484313965,
-      "learning_rate": 1.9954258707345903e-05,
-      "loss": 0.1845,
+      "epoch": 0.07807500319979521,
+      "grad_norm": 0.7149448394775391,
+      "learning_rate": 2.5986771922338383e-06,
+      "loss": 0.6746,
       "step": 610
     },
     {
-      "epoch": 0.07585953750152942,
-      "grad_norm": 0.1545839011669159,
-      "learning_rate": 1.9946112866441757e-05,
-      "loss": 0.1861,
+      "epoch": 0.07935492128503775,
+      "grad_norm": 2.0167856216430664,
+      "learning_rate": 2.64134841049712e-06,
+      "loss": 0.682,
       "step": 620
     },
     {
-      "epoch": 0.07708307842897345,
-      "grad_norm": 0.2535873055458069,
-      "learning_rate": 1.9937301830450423e-05,
-      "loss": 0.1772,
+      "epoch": 0.0806348393702803,
+      "grad_norm": 1.303001880645752,
+      "learning_rate": 2.6840196287604015e-06,
+      "loss": 0.6735,
       "step": 630
     },
     {
-      "epoch": 0.07830661935641747,
-      "grad_norm": 0.2617948651313782,
-      "learning_rate": 1.992782618865315e-05,
-      "loss": 0.1711,
+      "epoch": 0.08191475745552285,
+      "grad_norm": 2.0516321659088135,
+      "learning_rate": 2.7266908470236826e-06,
+      "loss": 0.6705,
       "step": 640
     },
     {
-      "epoch": 0.07953016028386149,
-      "grad_norm": 0.38676971197128296,
-      "learning_rate": 1.9917686574779972e-05,
-      "loss": 0.1748,
+      "epoch": 0.08319467554076539,
+      "grad_norm": 2.242105484008789,
+      "learning_rate": 2.7693620652869642e-06,
+      "loss": 0.6798,
       "step": 650
     },
     {
-      "epoch": 0.08075370121130551,
-      "grad_norm": 0.1511957049369812,
-      "learning_rate": 1.9906883666967296e-05,
-      "loss": 0.1825,
+      "epoch": 0.08447459362600794,
+      "grad_norm": 1.1836538314819336,
+      "learning_rate": 2.812033283550246e-06,
+      "loss": 0.6684,
       "step": 660
     },
     {
-      "epoch": 0.08197724213874955,
-      "grad_norm": 0.2525976002216339,
-      "learning_rate": 1.989541818771258e-05,
-      "loss": 0.1651,
+      "epoch": 0.08575451171125048,
+      "grad_norm": 0.8640874624252319,
+      "learning_rate": 2.8547045018135274e-06,
+      "loss": 0.6783,
       "step": 670
     },
     {
-      "epoch": 0.08320078306619356,
-      "grad_norm": 0.25827622413635254,
-      "learning_rate": 1.9883290903825994e-05,
-      "loss": 0.1763,
+      "epoch": 0.08703442979649302,
+      "grad_norm": 0.6901177763938904,
+      "learning_rate": 2.897375720076808e-06,
+      "loss": 0.6724,
       "step": 680
     },
     {
-      "epoch": 0.08442432399363758,
-      "grad_norm": 0.24877820909023285,
-      "learning_rate": 1.9870502626379127e-05,
-      "loss": 0.1652,
+      "epoch": 0.08831434788173557,
+      "grad_norm": 1.3443949222564697,
+      "learning_rate": 2.9400469383400897e-06,
+      "loss": 0.67,
       "step": 690
     },
     {
-      "epoch": 0.0856478649210816,
-      "grad_norm": 0.278772234916687,
-      "learning_rate": 1.9857054210650776e-05,
-      "loss": 0.1651,
-      "step": 700
-    },
-    {
-      "epoch": 0.0856478649210816,
-      "eval_accuracy": 0.94905,
-      "eval_f1_weighted": 0.9490731118481598,
-      "eval_loss": 0.03884471580386162,
-      "eval_macro_f1": 0.9489981325237288,
-      "eval_precision": 0.9494795831024372,
-      "eval_recall": 0.94905,
-      "eval_runtime": 759.9337,
-      "eval_samples_per_second": 131.59,
-      "eval_steps_per_second": 1.029,
+      "epoch": 0.08959426596697811,
+      "grad_norm": 0.7775189280509949,
+      "learning_rate": 2.9827181566033713e-06,
+      "loss": 0.6733,
       "step": 700
     },
     {
-      "epoch": 0.08687140584852564,
-      "grad_norm": 0.1895805150270462,
-      "learning_rate": 1.9842946556069705e-05,
-      "loss": 0.1628,
+      "epoch": 0.09087418405222066,
+      "grad_norm": 1.16916823387146,
+      "learning_rate": 3.025389374866653e-06,
+      "loss": 0.6713,
       "step": 710
     },
     {
-      "epoch": 0.08809494677596966,
-      "grad_norm": 0.1171497106552124,
-      "learning_rate": 1.982818060615452e-05,
-      "loss": 0.1715,
+      "epoch": 0.09215410213746321,
+      "grad_norm": 1.707152009010315,
+      "learning_rate": 3.068060593129934e-06,
+      "loss": 0.6659,
       "step": 720
     },
     {
-      "epoch": 0.08931848770341368,
-      "grad_norm": 0.22032728791236877,
-      "learning_rate": 1.9812757348450547e-05,
-      "loss": 0.1712,
+      "epoch": 0.09343402022270575,
+      "grad_norm": 0.9780622124671936,
+      "learning_rate": 3.1107318113932156e-06,
+      "loss": 0.6639,
       "step": 730
     },
     {
-      "epoch": 0.0905420286308577,
-      "grad_norm": 0.20700488984584808,
-      "learning_rate": 1.9796677814463812e-05,
-      "loss": 0.1638,
+      "epoch": 0.09471393830794829,
+      "grad_norm": 2.3687374591827393,
+      "learning_rate": 3.153403029656497e-06,
+      "loss": 0.6666,
       "step": 740
     },
     {
-      "epoch": 0.09176556955830173,
-      "grad_norm": 0.24688613414764404,
-      "learning_rate": 1.977994307959202e-05,
-      "loss": 0.1647,
+      "epoch": 0.09599385639319083,
+      "grad_norm": 0.8506460189819336,
+      "learning_rate": 3.1960742479197783e-06,
+      "loss": 0.6778,
       "step": 750
     },
     {
-      "epoch": 0.09298911048574575,
-      "grad_norm": 0.1886432021856308,
-      "learning_rate": 1.9762554263052655e-05,
-      "loss": 0.1734,
+      "epoch": 0.09727377447843338,
+      "grad_norm": 1.0102107524871826,
+      "learning_rate": 3.23874546618306e-06,
+      "loss": 0.6635,
       "step": 760
     },
     {
-      "epoch": 0.09421265141318977,
-      "grad_norm": 0.08407370001077652,
-      "learning_rate": 1.9744512527808127e-05,
-      "loss": 0.1755,
+      "epoch": 0.09855369256367592,
+      "grad_norm": 1.1480194330215454,
+      "learning_rate": 3.2814166844463415e-06,
+      "loss": 0.6709,
       "step": 770
     },
     {
-      "epoch": 0.09543619234063379,
-      "grad_norm": 0.14160771667957306,
-      "learning_rate": 1.9725819080487976e-05,
-      "loss": 0.1749,
+      "epoch": 0.09983361064891846,
+      "grad_norm": 1.3119179010391235,
+      "learning_rate": 3.324087902709623e-06,
+      "loss": 0.6564,
       "step": 780
     },
     {
-      "epoch": 0.09665973326807782,
-      "grad_norm": 0.17038081586360931,
-      "learning_rate": 1.970647517130819e-05,
-      "loss": 0.1795,
+      "epoch": 0.10111352873416102,
+      "grad_norm": 0.681377649307251,
+      "learning_rate": 3.3667591209729038e-06,
+      "loss": 0.6656,
       "step": 790
     },
     {
-      "epoch": 0.09788327419552184,
-      "grad_norm": 0.24339532852172852,
-      "learning_rate": 1.968648209398759e-05,
-      "loss": 0.161,
+      "epoch": 0.10239344681940356,
+      "grad_norm": 1.022257685661316,
+      "learning_rate": 3.4094303392361854e-06,
+      "loss": 0.6709,
       "step": 800
     },
     {
-      "epoch": 0.09788327419552184,
-      "eval_accuracy": 0.95721,
-      "eval_f1_weighted": 0.9572295598206603,
-      "eval_loss": 0.03337462991476059,
-      "eval_macro_f1": 0.9571669846830866,
-      "eval_precision": 0.9576467364395861,
-      "eval_recall": 0.95721,
-      "eval_runtime": 759.9975,
-      "eval_samples_per_second": 131.579,
-      "eval_steps_per_second": 1.029,
+      "epoch": 0.10239344681940356,
+      "eval_accuracy": 0.66247,
+      "eval_loss": 0.6552017331123352,
+      "eval_macro_f1": 0.6514270557181998,
+      "eval_precision": 0.6718270655389242,
+      "eval_recall": 0.6556580720913894,
+      "eval_runtime": 1633.5478,
+      "eval_samples_per_second": 61.216,
+      "eval_steps_per_second": 0.479,
       "step": 800
     },
     {
-      "epoch": 0.09910681512296586,
-      "grad_norm": 0.16346058249473572,
-      "learning_rate": 1.9665841185661283e-05,
-      "loss": 0.1596,
+      "epoch": 0.1036733649046461,
+      "grad_norm": 1.0268975496292114,
+      "learning_rate": 3.452101557499467e-06,
+      "loss": 0.6602,
       "step": 810
     },
     {
-      "epoch": 0.10033035605040988,
-      "grad_norm": 0.25031423568725586,
-      "learning_rate": 1.964455382679128e-05,
-      "loss": 0.1662,
+      "epoch": 0.10495328298988865,
+      "grad_norm": 2.6206603050231934,
+      "learning_rate": 3.494772775762748e-06,
+      "loss": 0.6705,
       "step": 820
     },
     {
-      "epoch": 0.10155389697785391,
-      "grad_norm": 0.36678382754325867,
-      "learning_rate": 1.9622621441074113e-05,
-      "loss": 0.1552,
+      "epoch": 0.10623320107513119,
+      "grad_norm": 1.035157561302185,
+      "learning_rate": 3.5374439940260297e-06,
+      "loss": 0.6643,
       "step": 830
     },
     {
-      "epoch": 0.10277743790529793,
-      "grad_norm": 0.34810611605644226,
-      "learning_rate": 1.960004549534567e-05,
-      "loss": 0.1663,
+      "epoch": 0.10751311916037373,
+      "grad_norm": 1.736412763595581,
+      "learning_rate": 3.5801152122893113e-06,
+      "loss": 0.6612,
       "step": 840
     },
     {
-      "epoch": 0.10400097883274195,
-      "grad_norm": 0.15511688590049744,
-      "learning_rate": 1.957682749948307e-05,
-      "loss": 0.1571,
+      "epoch": 0.10879303724561629,
+      "grad_norm": 0.6906964778900146,
+      "learning_rate": 3.622786430552593e-06,
+      "loss": 0.6576,
       "step": 850
     },
     {
-      "epoch": 0.10522451976018597,
-      "grad_norm": 0.2082190364599228,
-      "learning_rate": 1.9552969006303666e-05,
-      "loss": 0.1497,
+      "epoch": 0.11007295533085883,
+      "grad_norm": 0.8148729801177979,
+      "learning_rate": 3.665457648815874e-06,
+      "loss": 0.6561,
       "step": 860
     },
     {
-      "epoch": 0.10644806068763,
-      "grad_norm": 0.14936988055706024,
-      "learning_rate": 1.9528471611461235e-05,
-      "loss": 0.1435,
+      "epoch": 0.11135287341610137,
+      "grad_norm": 1.0217723846435547,
+      "learning_rate": 3.7081288670791556e-06,
+      "loss": 0.6581,
       "step": 870
     },
     {
-      "epoch": 0.10767160161507403,
-      "grad_norm": 0.252530574798584,
-      "learning_rate": 1.9503336953339223e-05,
-      "loss": 0.1536,
+      "epoch": 0.11263279150134392,
+      "grad_norm": 1.2743115425109863,
+      "learning_rate": 3.750800085342437e-06,
+      "loss": 0.6606,
       "step": 880
     },
     {
-      "epoch": 0.10889514254251804,
-      "grad_norm": 0.07253506034612656,
-      "learning_rate": 1.947756671294118e-05,
-      "loss": 0.1605,
+      "epoch": 0.11391270958658646,
+      "grad_norm": 0.6405361294746399,
+      "learning_rate": 3.793471303605718e-06,
+      "loss": 0.652,
       "step": 890
     },
     {
-      "epoch": 0.11011868346996206,
-      "grad_norm": 0.09060605615377426,
-      "learning_rate": 1.9451162613778353e-05,
-      "loss": 0.1581,
-      "step": 900
-    },
-    {
-      "epoch": 0.11011868346996206,
-      "eval_accuracy": 0.95459,
-      "eval_f1_weighted": 0.9546135918699761,
-      "eval_loss": 0.03526490181684494,
-      "eval_macro_f1": 0.9545725664001194,
-      "eval_precision": 0.9559789658562307,
-      "eval_recall": 0.95459,
-      "eval_runtime": 759.7202,
-      "eval_samples_per_second": 131.627,
-      "eval_steps_per_second": 1.029,
+      "epoch": 0.115192627671829,
+      "grad_norm": 0.8357161283493042,
+      "learning_rate": 3.836142521869e-06,
+      "loss": 0.6558,
       "step": 900
     },
     {
-      "epoch": 0.1113422243974061,
-      "grad_norm": 0.21080073714256287,
-      "learning_rate": 1.942412642175439e-05,
-      "loss": 0.164,
+      "epoch": 0.11647254575707154,
+      "grad_norm": 1.3505173921585083,
+      "learning_rate": 3.8788137401322815e-06,
+      "loss": 0.6576,
       "step": 910
     },
     {
-      "epoch": 0.11256576532485012,
-      "grad_norm": 0.22832269966602325,
-      "learning_rate": 1.939645994504726e-05,
-      "loss": 0.1595,
+      "epoch": 0.1177524638423141,
+      "grad_norm": 1.2253319025039673,
+      "learning_rate": 3.921484958395563e-06,
+      "loss": 0.6548,
       "step": 920
     },
     {
-      "epoch": 0.11378930625229414,
-      "grad_norm": 0.19044892489910126,
-      "learning_rate": 1.93681650339883e-05,
-      "loss": 0.1648,
+      "epoch": 0.11903238192755664,
+      "grad_norm": 0.6269568800926208,
+      "learning_rate": 3.964156176658844e-06,
+      "loss": 0.6562,
       "step": 930
     },
     {
-      "epoch": 0.11501284717973816,
-      "grad_norm": 0.2413518726825714,
-      "learning_rate": 1.9339243580938493e-05,
-      "loss": 0.155,
+      "epoch": 0.12031230001279918,
+      "grad_norm": 0.9413424134254456,
+      "learning_rate": 4.006827394922125e-06,
+      "loss": 0.6461,
       "step": 940
     },
     {
-      "epoch": 0.11623638810718219,
-      "grad_norm": 0.13218896090984344,
-      "learning_rate": 1.930969752016187e-05,
-      "loss": 0.1499,
+      "epoch": 0.12159221809804173,
+      "grad_norm": 1.0903513431549072,
+      "learning_rate": 4.049498613185407e-06,
+      "loss": 0.6457,
       "step": 950
     },
     {
-      "epoch": 0.11745992903462621,
-      "grad_norm": 0.23770441114902496,
-      "learning_rate": 1.927952882769619e-05,
-      "loss": 0.1555,
+      "epoch": 0.12287213618328427,
+      "grad_norm": 2.0023152828216553,
+      "learning_rate": 4.092169831448688e-06,
+      "loss": 0.6458,
       "step": 960
     },
     {
-      "epoch": 0.11868346996207023,
-      "grad_norm": 0.2193710207939148,
-      "learning_rate": 1.924873952122074e-05,
-      "loss": 0.1505,
+      "epoch": 0.12415205426852681,
+      "grad_norm": 0.9638109803199768,
+      "learning_rate": 4.134841049711969e-06,
+      "loss": 0.6486,
       "step": 970
     },
     {
-      "epoch": 0.11990701088951425,
-      "grad_norm": 0.1577174812555313,
-      "learning_rate": 1.9217331659921432e-05,
-      "loss": 0.1551,
+      "epoch": 0.12543197235376935,
+      "grad_norm": 0.9894546270370483,
+      "learning_rate": 4.177512267975251e-06,
+      "loss": 0.6453,
       "step": 980
     },
     {
-      "epoch": 0.12113055181695828,
-      "grad_norm": 0.1814967393875122,
-      "learning_rate": 1.9185307344353052e-05,
-      "loss": 0.1607,
+      "epoch": 0.1267118904390119,
+      "grad_norm": 1.2289057970046997,
+      "learning_rate": 4.220183486238532e-06,
+      "loss": 0.6605,
       "step": 990
     },
     {
-      "epoch": 0.1223540927444023,
-      "grad_norm": 0.27239707112312317,
-      "learning_rate": 1.91526687162988e-05,
-      "loss": 0.1506,
+      "epoch": 0.12799180852425446,
+      "grad_norm": 1.119125247001648,
+      "learning_rate": 4.262854704501814e-06,
+      "loss": 0.6564,
       "step": 1000
     },
     {
-      "epoch": 0.1223540927444023,
-      "eval_accuracy": 0.96454,
-      "eval_f1_weighted": 0.9645510150201642,
-      "eval_loss": 0.028770776465535164,
-      "eval_macro_f1": 0.9644889377060355,
-      "eval_precision": 0.9646912114769961,
-      "eval_recall": 0.96454,
-      "eval_runtime": 759.5962,
-      "eval_samples_per_second": 131.649,
-      "eval_steps_per_second": 1.029,
+      "epoch": 0.12799180852425446,
+      "eval_accuracy": 0.69068,
+      "eval_loss": 0.6396247148513794,
+      "eval_macro_f1": 0.689588440856387,
+      "eval_precision": 0.6900076024604327,
+      "eval_recall": 0.6894114140312133,
+      "eval_runtime": 1634.4321,
+      "eval_samples_per_second": 61.183,
+      "eval_steps_per_second": 0.478,
       "step": 1000
     }
   ],
   "logging_steps": 10,
-  "max_steps": 8173,
+  "max_steps": 23439,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 1,
-  "save_steps": 100,
+  "num_train_epochs": 3,
+  "save_steps": 200,
   "stateful_callbacks": {
     "EarlyStoppingCallback": {
       "args": {
@@ -866,7 +796,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.6974575173632e+16,
+  "total_flos": 1.683910754304e+16,
   "train_batch_size": 64,
   "trial_name": null,
   "trial_params": null

checkpoint-1000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:84799314bcc123b16dc7e721b7c0a39a7148ba4c572d589c671664f2e203e207
-size 5841
+oid sha256:88587329aed582b4509ae035cda4f2b8b5747893ac474b5bc3b9cd7edddf9ba2
+size 5777

checkpoint-1200/config.json

@@ -0,0 +1,27 @@
+{
+  "architectures": [
+    "RobertaForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "roberta",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "output_past": true,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.56.0",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 50265
+}

checkpoint-1200/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61d5c9de3cde05f670095e14f42c9e4589698ab7e7544947e9899cb0fe38685d
+size 498612824

checkpoint-1200/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91d4f9219105610d445d1323ff323318ea75bcd29155e0cbd304baef2009c888
+size 4741859

checkpoint-1200/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a61d9b89bfb67c8331499ea02a3f044c3de94ae9858090c63f03cffdd302305e
+size 14645

checkpoint-1200/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd0aac41bfa547f4a1790a56d1fdcb6598cb130255e5d0420b6b94b14c070642
+size 1465

checkpoint-1200/special_tokens_map.json

@@ -0,0 +1,51 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "cls_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "<mask>",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-1200/tokenizer_config.json

@@ -0,0 +1,58 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "50264": {
+      "content": "<mask>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "RobertaTokenizer",
+  "trim_offsets": true,
+  "unk_token": "<unk>"
+}

checkpoint-1200/trainer_state.json

@@ -0,0 +1,955 @@
+{
+  "best_global_step": 1200,
+  "best_metric": 0.6993136399748257,
+  "best_model_checkpoint": "./taskA-codebert-base/checkpoint-1200",
+  "epoch": 0.15359017022910534,
+  "eval_steps": 200,
+  "global_step": 1200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0012799180852425445,
+      "grad_norm": 3.6954128742218018,
+      "learning_rate": 3.840409643695328e-08,
+      "loss": 0.7763,
+      "step": 10
+    },
+    {
+      "epoch": 0.002559836170485089,
+      "grad_norm": 3.3660645484924316,
+      "learning_rate": 8.10753147002347e-08,
+      "loss": 0.7763,
+      "step": 20
+    },
+    {
+      "epoch": 0.0038397542557276334,
+      "grad_norm": 4.491067886352539,
+      "learning_rate": 1.2374653296351612e-07,
+      "loss": 0.7889,
+      "step": 30
+    },
+    {
+      "epoch": 0.005119672340970178,
+      "grad_norm": 1.5920521020889282,
+      "learning_rate": 1.6641775122679754e-07,
+      "loss": 0.7888,
+      "step": 40
+    },
+    {
+      "epoch": 0.006399590426212722,
+      "grad_norm": 4.410898208618164,
+      "learning_rate": 2.0908896949007894e-07,
+      "loss": 0.7692,
+      "step": 50
+    },
+    {
+      "epoch": 0.007679508511455267,
+      "grad_norm": 2.5669264793395996,
+      "learning_rate": 2.517601877533604e-07,
+      "loss": 0.7735,
+      "step": 60
+    },
+    {
+      "epoch": 0.008959426596697812,
+      "grad_norm": 6.190986156463623,
+      "learning_rate": 2.944314060166418e-07,
+      "loss": 0.7679,
+      "step": 70
+    },
+    {
+      "epoch": 0.010239344681940356,
+      "grad_norm": 5.736413478851318,
+      "learning_rate": 3.371026242799232e-07,
+      "loss": 0.744,
+      "step": 80
+    },
+    {
+      "epoch": 0.011519262767182901,
+      "grad_norm": 4.278866767883301,
+      "learning_rate": 3.7977384254320464e-07,
+      "loss": 0.7542,
+      "step": 90
+    },
+    {
+      "epoch": 0.012799180852425445,
+      "grad_norm": 5.565455436706543,
+      "learning_rate": 4.22445060806486e-07,
+      "loss": 0.752,
+      "step": 100
+    },
+    {
+      "epoch": 0.01407909893766799,
+      "grad_norm": 4.031781196594238,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 0.7614,
+      "step": 110
+    },
+    {
+      "epoch": 0.015359017022910534,
+      "grad_norm": 2.484126329421997,
+      "learning_rate": 5.077874973330489e-07,
+      "loss": 0.7664,
+      "step": 120
+    },
+    {
+      "epoch": 0.016638935108153077,
+      "grad_norm": 2.7525155544281006,
+      "learning_rate": 5.504587155963304e-07,
+      "loss": 0.7521,
+      "step": 130
+    },
+    {
+      "epoch": 0.017918853193395624,
+      "grad_norm": 1.6374385356903076,
+      "learning_rate": 5.931299338596117e-07,
+      "loss": 0.7436,
+      "step": 140
+    },
+    {
+      "epoch": 0.019198771278638168,
+      "grad_norm": 2.370774269104004,
+      "learning_rate": 6.358011521228932e-07,
+      "loss": 0.7429,
+      "step": 150
+    },
+    {
+      "epoch": 0.02047868936388071,
+      "grad_norm": 2.8948566913604736,
+      "learning_rate": 6.784723703861745e-07,
+      "loss": 0.7333,
+      "step": 160
+    },
+    {
+      "epoch": 0.021758607449123255,
+      "grad_norm": 2.7481186389923096,
+      "learning_rate": 7.21143588649456e-07,
+      "loss": 0.7294,
+      "step": 170
+    },
+    {
+      "epoch": 0.023038525534365802,
+      "grad_norm": 2.0159404277801514,
+      "learning_rate": 7.638148069127374e-07,
+      "loss": 0.731,
+      "step": 180
+    },
+    {
+      "epoch": 0.024318443619608346,
+      "grad_norm": 3.6898131370544434,
+      "learning_rate": 8.064860251760189e-07,
+      "loss": 0.7153,
+      "step": 190
+    },
+    {
+      "epoch": 0.02559836170485089,
+      "grad_norm": 3.2842342853546143,
+      "learning_rate": 8.491572434393003e-07,
+      "loss": 0.7245,
+      "step": 200
+    },
+    {
+      "epoch": 0.02559836170485089,
+      "eval_accuracy": 0.47695,
+      "eval_loss": 0.7148504853248596,
+      "eval_macro_f1": 0.3229290091066048,
+      "eval_precision": 0.238475,
+      "eval_recall": 0.5,
+      "eval_runtime": 1641.324,
+      "eval_samples_per_second": 60.926,
+      "eval_steps_per_second": 0.476,
+      "step": 200
+    },
+    {
+      "epoch": 0.026878279790093433,
+      "grad_norm": 2.294558048248291,
+      "learning_rate": 8.918284617025817e-07,
+      "loss": 0.718,
+      "step": 210
+    },
+    {
+      "epoch": 0.02815819787533598,
+      "grad_norm": 2.9467175006866455,
+      "learning_rate": 9.344996799658632e-07,
+      "loss": 0.7076,
+      "step": 220
+    },
+    {
+      "epoch": 0.029438115960578524,
+      "grad_norm": 2.6889193058013916,
+      "learning_rate": 9.771708982291445e-07,
+      "loss": 0.7019,
+      "step": 230
+    },
+    {
+      "epoch": 0.030718034045821067,
+      "grad_norm": 0.7491384744644165,
+      "learning_rate": 1.0198421164924258e-06,
+      "loss": 0.7046,
+      "step": 240
+    },
+    {
+      "epoch": 0.031997952131063614,
+      "grad_norm": 2.0363433361053467,
+      "learning_rate": 1.0625133347557074e-06,
+      "loss": 0.7085,
+      "step": 250
+    },
+    {
+      "epoch": 0.033277870216306155,
+      "grad_norm": 1.1431337594985962,
+      "learning_rate": 1.1051845530189888e-06,
+      "loss": 0.7008,
+      "step": 260
+    },
+    {
+      "epoch": 0.0345577883015487,
+      "grad_norm": 3.7735233306884766,
+      "learning_rate": 1.1478557712822702e-06,
+      "loss": 0.7015,
+      "step": 270
+    },
+    {
+      "epoch": 0.03583770638679125,
+      "grad_norm": 0.7889328002929688,
+      "learning_rate": 1.1905269895455517e-06,
+      "loss": 0.699,
+      "step": 280
+    },
+    {
+      "epoch": 0.03711762447203379,
+      "grad_norm": 0.8796226382255554,
+      "learning_rate": 1.233198207808833e-06,
+      "loss": 0.6922,
+      "step": 290
+    },
+    {
+      "epoch": 0.038397542557276336,
+      "grad_norm": 0.7966652512550354,
+      "learning_rate": 1.2758694260721145e-06,
+      "loss": 0.6949,
+      "step": 300
+    },
+    {
+      "epoch": 0.039677460642518876,
+      "grad_norm": 0.7029069066047668,
+      "learning_rate": 1.318540644335396e-06,
+      "loss": 0.695,
+      "step": 310
+    },
+    {
+      "epoch": 0.04095737872776142,
+      "grad_norm": 2.04327392578125,
+      "learning_rate": 1.3612118625986772e-06,
+      "loss": 0.6944,
+      "step": 320
+    },
+    {
+      "epoch": 0.04223729681300397,
+      "grad_norm": 0.918025553226471,
+      "learning_rate": 1.4038830808619588e-06,
+      "loss": 0.6863,
+      "step": 330
+    },
+    {
+      "epoch": 0.04351721489824651,
+      "grad_norm": 0.9528706073760986,
+      "learning_rate": 1.4465542991252401e-06,
+      "loss": 0.6907,
+      "step": 340
+    },
+    {
+      "epoch": 0.04479713298348906,
+      "grad_norm": 1.071359634399414,
+      "learning_rate": 1.4892255173885215e-06,
+      "loss": 0.6873,
+      "step": 350
+    },
+    {
+      "epoch": 0.046077051068731605,
+      "grad_norm": 1.9875625371932983,
+      "learning_rate": 1.531896735651803e-06,
+      "loss": 0.6942,
+      "step": 360
+    },
+    {
+      "epoch": 0.047356969153974145,
+      "grad_norm": 2.0222439765930176,
+      "learning_rate": 1.5745679539150842e-06,
+      "loss": 0.6898,
+      "step": 370
+    },
+    {
+      "epoch": 0.04863688723921669,
+      "grad_norm": 1.7827857732772827,
+      "learning_rate": 1.6172391721783658e-06,
+      "loss": 0.6909,
+      "step": 380
+    },
+    {
+      "epoch": 0.04991680532445923,
+      "grad_norm": 1.065674901008606,
+      "learning_rate": 1.6599103904416472e-06,
+      "loss": 0.6877,
+      "step": 390
+    },
+    {
+      "epoch": 0.05119672340970178,
+      "grad_norm": 0.8346193432807922,
+      "learning_rate": 1.7025816087049288e-06,
+      "loss": 0.6889,
+      "step": 400
+    },
+    {
+      "epoch": 0.05119672340970178,
+      "eval_accuracy": 0.54043,
+      "eval_loss": 0.6827595233917236,
+      "eval_macro_f1": 0.41833840347598905,
+      "eval_precision": 0.5839205994464474,
+      "eval_recall": 0.5204152767503025,
+      "eval_runtime": 1634.6244,
+      "eval_samples_per_second": 61.176,
+      "eval_steps_per_second": 0.478,
+      "step": 400
+    },
+    {
+      "epoch": 0.052476641494944326,
+      "grad_norm": 0.8074935674667358,
+      "learning_rate": 1.7452528269682101e-06,
+      "loss": 0.684,
+      "step": 410
+    },
+    {
+      "epoch": 0.053756559580186866,
+      "grad_norm": 0.8965069651603699,
+      "learning_rate": 1.7879240452314913e-06,
+      "loss": 0.6922,
+      "step": 420
+    },
+    {
+      "epoch": 0.05503647766542941,
+      "grad_norm": 2.597215414047241,
+      "learning_rate": 1.8305952634947729e-06,
+      "loss": 0.6909,
+      "step": 430
+    },
+    {
+      "epoch": 0.05631639575067196,
+      "grad_norm": 0.8233373165130615,
+      "learning_rate": 1.8732664817580542e-06,
+      "loss": 0.6831,
+      "step": 440
+    },
+    {
+      "epoch": 0.0575963138359145,
+      "grad_norm": 1.0624489784240723,
+      "learning_rate": 1.915937700021336e-06,
+      "loss": 0.6817,
+      "step": 450
+    },
+    {
+      "epoch": 0.05887623192115705,
+      "grad_norm": 2.0623559951782227,
+      "learning_rate": 1.958608918284617e-06,
+      "loss": 0.6819,
+      "step": 460
+    },
+    {
+      "epoch": 0.06015615000639959,
+      "grad_norm": 0.912433922290802,
+      "learning_rate": 2.0012801365478988e-06,
+      "loss": 0.686,
+      "step": 470
+    },
+    {
+      "epoch": 0.061436068091642135,
+      "grad_norm": 0.6926271319389343,
+      "learning_rate": 2.04395135481118e-06,
+      "loss": 0.6834,
+      "step": 480
+    },
+    {
+      "epoch": 0.06271598617688467,
+      "grad_norm": 1.498634934425354,
+      "learning_rate": 2.0866225730744615e-06,
+      "loss": 0.6805,
+      "step": 490
+    },
+    {
+      "epoch": 0.06399590426212723,
+      "grad_norm": 1.1818156242370605,
+      "learning_rate": 2.129293791337743e-06,
+      "loss": 0.6805,
+      "step": 500
+    },
+    {
+      "epoch": 0.06527582234736977,
+      "grad_norm": 1.4476051330566406,
+      "learning_rate": 2.1719650096010242e-06,
+      "loss": 0.6773,
+      "step": 510
+    },
+    {
+      "epoch": 0.06655574043261231,
+      "grad_norm": 1.1143743991851807,
+      "learning_rate": 2.214636227864306e-06,
+      "loss": 0.6801,
+      "step": 520
+    },
+    {
+      "epoch": 0.06783565851785486,
+      "grad_norm": 1.7030247449874878,
+      "learning_rate": 2.257307446127587e-06,
+      "loss": 0.6794,
+      "step": 530
+    },
+    {
+      "epoch": 0.0691155766030974,
+      "grad_norm": 0.8592290282249451,
+      "learning_rate": 2.2999786643908685e-06,
+      "loss": 0.6736,
+      "step": 540
+    },
+    {
+      "epoch": 0.07039549468833994,
+      "grad_norm": 0.8231940865516663,
+      "learning_rate": 2.34264988265415e-06,
+      "loss": 0.67,
+      "step": 550
+    },
+    {
+      "epoch": 0.0716754127735825,
+      "grad_norm": 0.88897705078125,
+      "learning_rate": 2.3853211009174317e-06,
+      "loss": 0.685,
+      "step": 560
+    },
+    {
+      "epoch": 0.07295533085882504,
+      "grad_norm": 1.6297541856765747,
+      "learning_rate": 2.427992319180713e-06,
+      "loss": 0.6797,
+      "step": 570
+    },
+    {
+      "epoch": 0.07423524894406758,
+      "grad_norm": 2.9297447204589844,
+      "learning_rate": 2.470663537443994e-06,
+      "loss": 0.6766,
+      "step": 580
+    },
+    {
+      "epoch": 0.07551516702931012,
+      "grad_norm": 0.9277703762054443,
+      "learning_rate": 2.5133347557072756e-06,
+      "loss": 0.6824,
+      "step": 590
+    },
+    {
+      "epoch": 0.07679508511455267,
+      "grad_norm": 0.5883333683013916,
+      "learning_rate": 2.556005973970557e-06,
+      "loss": 0.6818,
+      "step": 600
+    },
+    {
+      "epoch": 0.07679508511455267,
+      "eval_accuracy": 0.63909,
+      "eval_loss": 0.6700075268745422,
+      "eval_macro_f1": 0.621848143053056,
+      "eval_precision": 0.6524313071088819,
+      "eval_recall": 0.6305863955136095,
+      "eval_runtime": 1633.9732,
+      "eval_samples_per_second": 61.201,
+      "eval_steps_per_second": 0.479,
+      "step": 600
+    },
+    {
+      "epoch": 0.07807500319979521,
+      "grad_norm": 0.7149448394775391,
+      "learning_rate": 2.5986771922338383e-06,
+      "loss": 0.6746,
+      "step": 610
+    },
+    {
+      "epoch": 0.07935492128503775,
+      "grad_norm": 2.0167856216430664,
+      "learning_rate": 2.64134841049712e-06,
+      "loss": 0.682,
+      "step": 620
+    },
+    {
+      "epoch": 0.0806348393702803,
+      "grad_norm": 1.303001880645752,
+      "learning_rate": 2.6840196287604015e-06,
+      "loss": 0.6735,
+      "step": 630
+    },
+    {
+      "epoch": 0.08191475745552285,
+      "grad_norm": 2.0516321659088135,
+      "learning_rate": 2.7266908470236826e-06,
+      "loss": 0.6705,
+      "step": 640
+    },
+    {
+      "epoch": 0.08319467554076539,
+      "grad_norm": 2.242105484008789,
+      "learning_rate": 2.7693620652869642e-06,
+      "loss": 0.6798,
+      "step": 650
+    },
+    {
+      "epoch": 0.08447459362600794,
+      "grad_norm": 1.1836538314819336,
+      "learning_rate": 2.812033283550246e-06,
+      "loss": 0.6684,
+      "step": 660
+    },
+    {
+      "epoch": 0.08575451171125048,
+      "grad_norm": 0.8640874624252319,
+      "learning_rate": 2.8547045018135274e-06,
+      "loss": 0.6783,
+      "step": 670
+    },
+    {
+      "epoch": 0.08703442979649302,
+      "grad_norm": 0.6901177763938904,
+      "learning_rate": 2.897375720076808e-06,
+      "loss": 0.6724,
+      "step": 680
+    },
+    {
+      "epoch": 0.08831434788173557,
+      "grad_norm": 1.3443949222564697,
+      "learning_rate": 2.9400469383400897e-06,
+      "loss": 0.67,
+      "step": 690
+    },
+    {
+      "epoch": 0.08959426596697811,
+      "grad_norm": 0.7775189280509949,
+      "learning_rate": 2.9827181566033713e-06,
+      "loss": 0.6733,
+      "step": 700
+    },
+    {
+      "epoch": 0.09087418405222066,
+      "grad_norm": 1.16916823387146,
+      "learning_rate": 3.025389374866653e-06,
+      "loss": 0.6713,
+      "step": 710
+    },
+    {
+      "epoch": 0.09215410213746321,
+      "grad_norm": 1.707152009010315,
+      "learning_rate": 3.068060593129934e-06,
+      "loss": 0.6659,
+      "step": 720
+    },
+    {
+      "epoch": 0.09343402022270575,
+      "grad_norm": 0.9780622124671936,
+      "learning_rate": 3.1107318113932156e-06,
+      "loss": 0.6639,
+      "step": 730
+    },
+    {
+      "epoch": 0.09471393830794829,
+      "grad_norm": 2.3687374591827393,
+      "learning_rate": 3.153403029656497e-06,
+      "loss": 0.6666,
+      "step": 740
+    },
+    {
+      "epoch": 0.09599385639319083,
+      "grad_norm": 0.8506460189819336,
+      "learning_rate": 3.1960742479197783e-06,
+      "loss": 0.6778,
+      "step": 750
+    },
+    {
+      "epoch": 0.09727377447843338,
+      "grad_norm": 1.0102107524871826,
+      "learning_rate": 3.23874546618306e-06,
+      "loss": 0.6635,
+      "step": 760
+    },
+    {
+      "epoch": 0.09855369256367592,
+      "grad_norm": 1.1480194330215454,
+      "learning_rate": 3.2814166844463415e-06,
+      "loss": 0.6709,
+      "step": 770
+    },
+    {
+      "epoch": 0.09983361064891846,
+      "grad_norm": 1.3119179010391235,
+      "learning_rate": 3.324087902709623e-06,
+      "loss": 0.6564,
+      "step": 780
+    },
+    {
+      "epoch": 0.10111352873416102,
+      "grad_norm": 0.681377649307251,
+      "learning_rate": 3.3667591209729038e-06,
+      "loss": 0.6656,
+      "step": 790
+    },
+    {
+      "epoch": 0.10239344681940356,
+      "grad_norm": 1.022257685661316,
+      "learning_rate": 3.4094303392361854e-06,
+      "loss": 0.6709,
+      "step": 800
+    },
+    {
+      "epoch": 0.10239344681940356,
+      "eval_accuracy": 0.66247,
+      "eval_loss": 0.6552017331123352,
+      "eval_macro_f1": 0.6514270557181998,
+      "eval_precision": 0.6718270655389242,
+      "eval_recall": 0.6556580720913894,
+      "eval_runtime": 1633.5478,
+      "eval_samples_per_second": 61.216,
+      "eval_steps_per_second": 0.479,
+      "step": 800
+    },
+    {
+      "epoch": 0.1036733649046461,
+      "grad_norm": 1.0268975496292114,
+      "learning_rate": 3.452101557499467e-06,
+      "loss": 0.6602,
+      "step": 810
+    },
+    {
+      "epoch": 0.10495328298988865,
+      "grad_norm": 2.6206603050231934,
+      "learning_rate": 3.494772775762748e-06,
+      "loss": 0.6705,
+      "step": 820
+    },
+    {
+      "epoch": 0.10623320107513119,
+      "grad_norm": 1.035157561302185,
+      "learning_rate": 3.5374439940260297e-06,
+      "loss": 0.6643,
+      "step": 830
+    },
+    {
+      "epoch": 0.10751311916037373,
+      "grad_norm": 1.736412763595581,
+      "learning_rate": 3.5801152122893113e-06,
+      "loss": 0.6612,
+      "step": 840
+    },
+    {
+      "epoch": 0.10879303724561629,
+      "grad_norm": 0.6906964778900146,
+      "learning_rate": 3.622786430552593e-06,
+      "loss": 0.6576,
+      "step": 850
+    },
+    {
+      "epoch": 0.11007295533085883,
+      "grad_norm": 0.8148729801177979,
+      "learning_rate": 3.665457648815874e-06,
+      "loss": 0.6561,
+      "step": 860
+    },
+    {
+      "epoch": 0.11135287341610137,
+      "grad_norm": 1.0217723846435547,
+      "learning_rate": 3.7081288670791556e-06,
+      "loss": 0.6581,
+      "step": 870
+    },
+    {
+      "epoch": 0.11263279150134392,
+      "grad_norm": 1.2743115425109863,
+      "learning_rate": 3.750800085342437e-06,
+      "loss": 0.6606,
+      "step": 880
+    },
+    {
+      "epoch": 0.11391270958658646,
+      "grad_norm": 0.6405361294746399,
+      "learning_rate": 3.793471303605718e-06,
+      "loss": 0.652,
+      "step": 890
+    },
+    {
+      "epoch": 0.115192627671829,
+      "grad_norm": 0.8357161283493042,
+      "learning_rate": 3.836142521869e-06,
+      "loss": 0.6558,
+      "step": 900
+    },
+    {
+      "epoch": 0.11647254575707154,
+      "grad_norm": 1.3505173921585083,
+      "learning_rate": 3.8788137401322815e-06,
+      "loss": 0.6576,
+      "step": 910
+    },
+    {
+      "epoch": 0.1177524638423141,
+      "grad_norm": 1.2253319025039673,
+      "learning_rate": 3.921484958395563e-06,
+      "loss": 0.6548,
+      "step": 920
+    },
+    {
+      "epoch": 0.11903238192755664,
+      "grad_norm": 0.6269568800926208,
+      "learning_rate": 3.964156176658844e-06,
+      "loss": 0.6562,
+      "step": 930
+    },
+    {
+      "epoch": 0.12031230001279918,
+      "grad_norm": 0.9413424134254456,
+      "learning_rate": 4.006827394922125e-06,
+      "loss": 0.6461,
+      "step": 940
+    },
+    {
+      "epoch": 0.12159221809804173,
+      "grad_norm": 1.0903513431549072,
+      "learning_rate": 4.049498613185407e-06,
+      "loss": 0.6457,
+      "step": 950
+    },
+    {
+      "epoch": 0.12287213618328427,
+      "grad_norm": 2.0023152828216553,
+      "learning_rate": 4.092169831448688e-06,
+      "loss": 0.6458,
+      "step": 960
+    },
+    {
+      "epoch": 0.12415205426852681,
+      "grad_norm": 0.9638109803199768,
+      "learning_rate": 4.134841049711969e-06,
+      "loss": 0.6486,
+      "step": 970
+    },
+    {
+      "epoch": 0.12543197235376935,
+      "grad_norm": 0.9894546270370483,
+      "learning_rate": 4.177512267975251e-06,
+      "loss": 0.6453,
+      "step": 980
+    },
+    {
+      "epoch": 0.1267118904390119,
+      "grad_norm": 1.2289057970046997,
+      "learning_rate": 4.220183486238532e-06,
+      "loss": 0.6605,
+      "step": 990
+    },
+    {
+      "epoch": 0.12799180852425446,
+      "grad_norm": 1.119125247001648,
+      "learning_rate": 4.262854704501814e-06,
+      "loss": 0.6564,
+      "step": 1000
+    },
+    {
+      "epoch": 0.12799180852425446,
+      "eval_accuracy": 0.69068,
+      "eval_loss": 0.6396247148513794,
+      "eval_macro_f1": 0.689588440856387,
+      "eval_precision": 0.6900076024604327,
+      "eval_recall": 0.6894114140312133,
+      "eval_runtime": 1634.4321,
+      "eval_samples_per_second": 61.183,
+      "eval_steps_per_second": 0.478,
+      "step": 1000
+    },
+    {
+      "epoch": 0.12927172660949698,
+      "grad_norm": 2.4629580974578857,
+      "learning_rate": 4.3055259227650956e-06,
+      "loss": 0.6485,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13055164469473954,
+      "grad_norm": 0.7082359790802002,
+      "learning_rate": 4.348197141028377e-06,
+      "loss": 0.6491,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1318315627799821,
+      "grad_norm": 1.4061448574066162,
+      "learning_rate": 4.390868359291658e-06,
+      "loss": 0.6439,
+      "step": 1030
+    },
+    {
+      "epoch": 0.13311148086522462,
+      "grad_norm": 0.8420212864875793,
+      "learning_rate": 4.4335395775549394e-06,
+      "loss": 0.6422,
+      "step": 1040
+    },
+    {
+      "epoch": 0.13439139895046717,
+      "grad_norm": 1.0872454643249512,
+      "learning_rate": 4.476210795818221e-06,
+      "loss": 0.6364,
+      "step": 1050
+    },
+    {
+      "epoch": 0.13567131703570973,
+      "grad_norm": 1.1760668754577637,
+      "learning_rate": 4.518882014081503e-06,
+      "loss": 0.6487,
+      "step": 1060
+    },
+    {
+      "epoch": 0.13695123512095225,
+      "grad_norm": 1.1484756469726562,
+      "learning_rate": 4.561553232344783e-06,
+      "loss": 0.6451,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1382311532061948,
+      "grad_norm": 0.7577368021011353,
+      "learning_rate": 4.604224450608065e-06,
+      "loss": 0.6419,
+      "step": 1080
+    },
+    {
+      "epoch": 0.13951107129143736,
+      "grad_norm": 0.6963039636611938,
+      "learning_rate": 4.6468956688713465e-06,
+      "loss": 0.634,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1407909893766799,
+      "grad_norm": 0.7238577008247375,
+      "learning_rate": 4.689566887134628e-06,
+      "loss": 0.6466,
+      "step": 1100
+    },
+    {
+      "epoch": 0.14207090746192244,
+      "grad_norm": 0.6751101016998291,
+      "learning_rate": 4.73223810539791e-06,
+      "loss": 0.6377,
+      "step": 1110
+    },
+    {
+      "epoch": 0.143350825547165,
+      "grad_norm": 0.7622250318527222,
+      "learning_rate": 4.774909323661191e-06,
+      "loss": 0.6318,
+      "step": 1120
+    },
+    {
+      "epoch": 0.14463074363240752,
+      "grad_norm": 1.3731235265731812,
+      "learning_rate": 4.817580541924473e-06,
+      "loss": 0.6464,
+      "step": 1130
+    },
+    {
+      "epoch": 0.14591066171765008,
+      "grad_norm": 0.8598525524139404,
+      "learning_rate": 4.8602517601877535e-06,
+      "loss": 0.6268,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1471905798028926,
+      "grad_norm": 1.1462016105651855,
+      "learning_rate": 4.902922978451035e-06,
+      "loss": 0.6357,
+      "step": 1150
+    },
+    {
+      "epoch": 0.14847049788813516,
+      "grad_norm": 1.2314528226852417,
+      "learning_rate": 4.945594196714317e-06,
+      "loss": 0.6407,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1497504159733777,
+      "grad_norm": 1.0393744707107544,
+      "learning_rate": 4.9882654149775974e-06,
+      "loss": 0.6425,
+      "step": 1170
+    },
+    {
+      "epoch": 0.15103033405862024,
+      "grad_norm": 0.9097604751586914,
+      "learning_rate": 5.030936633240879e-06,
+      "loss": 0.6275,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1523102521438628,
+      "grad_norm": 0.8274818062782288,
+      "learning_rate": 5.073607851504161e-06,
+      "loss": 0.6369,
+      "step": 1190
+    },
+    {
+      "epoch": 0.15359017022910534,
+      "grad_norm": 0.8929425477981567,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 0.6382,
+      "step": 1200
+    },
+    {
+      "epoch": 0.15359017022910534,
+      "eval_accuracy": 0.70003,
+      "eval_loss": 0.6226751804351807,
+      "eval_macro_f1": 0.6993136399748257,
+      "eval_precision": 0.6993716582660007,
+      "eval_recall": 0.6992659459810584,
+      "eval_runtime": 1635.7462,
+      "eval_samples_per_second": 61.134,
+      "eval_steps_per_second": 0.478,
+      "step": 1200
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 23439,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 200,
+  "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": 2.0206929051648e+16,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1200/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e72353f4c9aec8aa78e85916ea1e167db5c282ccfd48294f1986907e6699455a
+size 5777

training.log

@@ -1,13 +1,372 @@
-2026-04-15 07:19:05,558 - INFO - __main__ - Logging to file: ./taskA-codebert-base/training.log
-2026-04-15 07:19:05,563 - INFO - __main__ - Loading tokenizer from: microsoft/codebert-base
-2026-04-15 07:19:05,823 - INFO - __main__ - Loading model from: microsoft/codebert-base
-2026-04-15 07:19:06,547 - INFO - __main__ - Model loaded successfully
-2026-04-15 07:19:06,551 - INFO - __main__ - Model config - num_labels: 2, hidden_size: 768
-2026-04-15 07:19:06,698 - INFO - __main__ - Model on device: cuda
-2026-04-15 07:19:06,702 - INFO - __main__ - Freezing base model weights, training classifier head only...
-2026-04-15 07:19:06,707 - INFO - __main__ - Base model frozen - only classifier head is trainable
-2026-04-15 07:19:06,710 - INFO - __main__ - Loading datasets from Hugging Face Hub...
-2026-04-15 07:19:07,548 - INFO - __main__ - Train samples: 500000, Val samples: 100000
-2026-04-15 07:19:07,552 - INFO - __main__ - Tokenizing datasets...
-2026-04-15 07:22:59,746 - INFO - __main__ - W&B set to offline mode for notebook stability
-2026-04-15 07:22:59,804 - INFO - __main__ - Starting training...
+2026-04-15 10:25:16,900 - INFO - train_pipeline - Logging to ./taskA-codebert-base/training.log
+2026-04-15 10:25:16,911 - INFO - train_pipeline - Loading model & tokenizer for 'microsoft/codebert-base'
+2026-04-15 10:25:18,028 - INFO - train_pipeline - Model placed on cuda
+2026-04-15 10:25:18,039 - INFO - train_pipeline - ===== Model Architecture =====
+2026-04-15 10:25:18,050 - INFO - train_pipeline - 
+RobertaForSequenceClassification(
+  (roberta): RobertaModel(
+    (embeddings): RobertaEmbeddings(
+      (word_embeddings): Embedding(50265, 768, padding_idx=1)
+      (position_embeddings): Embedding(514, 768, padding_idx=1)
+      (token_type_embeddings): Embedding(1, 768)
+      (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+      (dropout): Dropout(p=0.1, inplace=False)
+    )
+    (encoder): RobertaEncoder(
+      (layer): ModuleList(
+        (0-11): 12 x RobertaLayer(
+          (attention): RobertaAttention(
+            (self): RobertaSdpaSelfAttention(
+              (query): Linear(in_features=768, out_features=768, bias=True)
+              (key): Linear(in_features=768, out_features=768, bias=True)
+              (value): Linear(in_features=768, out_features=768, bias=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+            (output): RobertaSelfOutput(
+              (dense): Linear(in_features=768, out_features=768, bias=True)
+              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+          )
+          (intermediate): RobertaIntermediate(
+            (dense): Linear(in_features=768, out_features=3072, bias=True)
+            (intermediate_act_fn): GELUActivation()
+          )
+          (output): RobertaOutput(
+            (dense): Linear(in_features=3072, out_features=768, bias=True)
+            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+            (dropout): Dropout(p=0.1, inplace=False)
+          )
+        )
+      )
+    )
+  )
+  (classifier): RobertaClassificationHead(
+    (dense): Linear(in_features=768, out_features=768, bias=True)
+    (dropout): Dropout(p=0.1, inplace=False)
+    (out_proj): Linear(in_features=768, out_features=2, bias=True)
+  )
+)
+2026-04-15 10:25:18,063 - INFO - train_pipeline - ===== Tokenizer Summary =====
+2026-04-15 10:25:18,087 - INFO - train_pipeline - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
+2026-04-15 10:25:18,097 - INFO - train_pipeline - ===== End of Architecture Log =====
+2026-04-15 10:25:18,108 - INFO - train_pipeline - Base model weights frozen – only classifier head will be trained.
+2026-04-15 10:25:19,118 - INFO - train_pipeline - === Starting training ===
+2026-04-15 10:25:16,900 - INFO - train_pipeline - Logging to ./taskA-codebert-base/training.log
+2026-04-15 10:25:16,911 - INFO - train_pipeline - Loading model & tokenizer for 'microsoft/codebert-base'
+2026-04-15 10:25:18,028 - INFO - train_pipeline - Model placed on cuda
+2026-04-15 10:25:18,039 - INFO - train_pipeline - ===== Model Architecture =====
+2026-04-15 10:25:18,050 - INFO - train_pipeline - 
+RobertaForSequenceClassification(
+  (roberta): RobertaModel(
+    (embeddings): RobertaEmbeddings(
+      (word_embeddings): Embedding(50265, 768, padding_idx=1)
+      (position_embeddings): Embedding(514, 768, padding_idx=1)
+      (token_type_embeddings): Embedding(1, 768)
+      (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+      (dropout): Dropout(p=0.1, inplace=False)
+    )
+    (encoder): RobertaEncoder(
+      (layer): ModuleList(
+        (0-11): 12 x RobertaLayer(
+          (attention): RobertaAttention(
+            (self): RobertaSdpaSelfAttention(
+              (query): Linear(in_features=768, out_features=768, bias=True)
+              (key): Linear(in_features=768, out_features=768, bias=True)
+              (value): Linear(in_features=768, out_features=768, bias=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+            (output): RobertaSelfOutput(
+              (dense): Linear(in_features=768, out_features=768, bias=True)
+              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+          )
+          (intermediate): RobertaIntermediate(
+            (dense): Linear(in_features=768, out_features=3072, bias=True)
+            (intermediate_act_fn): GELUActivation()
+          )
+          (output): RobertaOutput(
+            (dense): Linear(in_features=3072, out_f2026-04-15 10:25:18,121 - INFO - __main__ - Loading datasets from Hugging Face Hub...
+2026-04-15 10:25:18,988 - INFO - __main__ - Train samples: 500000, Val samples: 100000
+2026-04-15 10:25:18,992 - INFO - __main__ - Tokenizing datasets...
+   (dense): Linear(in_features=768, out_features=768, bias=True)
+    (dropout): Dropout(p=0.1, inplace=False)
+    (out_proj): Linear(in_features=768, out_features=2, bias=True)
+  )
+)
+2026-04-15 10:25:18,063 - INFO - train_pipeline - ===== Tokenizer Summary =====
+2026-04-15 10:25:18,087 - INFO - train_pipeline - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
+2026-04-15 10:25:18,097 - INFO - train_pipeline - ===== End of Architecture Log =====
+2026-04-15 10:25:18,108 - INFO - train_pipeline - Base model weights frozen – only classifier head will be trained.
+2026-04-15 10:25:19,118 - INFO - train_pipeline - === Starting training ===
+2026-04-15 10:25:16,900 - INFO - train_pipeline - Logging to ./taskA-codebert-base/training.log
+2026-04-15 10:25:16,911 - INFO - train_pipeline - Loading model & tokenizer for 'microsoft/codebert-base'
+2026-04-15 10:25:18,028 - INFO - train_pipeline - Model placed on cuda
+2026-04-15 10:25:18,039 - INFO - train_pipeline - ===== Model Architecture =====
+2026-04-15 10:25:18,050 - INFO - train_pipeline - 
+RobertaForSequenceClassification(
+  (roberta): RobertaModel(
+    (embeddings): RobertaEmbeddings(
+      (word_embeddings): Embedding(50265, 768, padding_idx=1)
+      (position_embeddings): Embedding(514, 768, padding_idx=1)
+      (token_type_embeddings): Embedding(1, 768)
+      (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+      (dropout): Dropout(p=0.1, inplace=False)
+    )
+    (encoder): RobertaEncoder(
+      (layer): ModuleList(
+        (0-11): 12 x RobertaLayer(
+          (attention): RobertaAttention(
+            (self): RobertaSdpaSelfAttention(
+              (query): Linear(in_features=768, out_features=768, bias=True)
+              (key): Linear(in_features=768, out_features=768, bias=True)
+              (value): Linear(in_features=768, out_features=768, bias=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+            (output): RobertaSelfOutput(
+              (dense): Linear(in_features=768, out_features=768, bias=True)
+              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+          )
+          (intermediate): RobertaIntermediate(
+            (dense): Linear(in_features=768, out_features=3072, bias=True)
+            (intermediate_act_fn): GELUActivation()
+          )
+          (output): RobertaOutput(
+            (dense): Linear(in_features=3072, out_features=768, bias=True)
+            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+            (dropout): Dropout(p=0.1, inplace=False)
+          )
+        )
+      )
+    )
+  )
+  (classifier): RobertaClassificationHead(
+    (dense): Linear(in_features=768, out_features=768, bias=True)
+    (dropout): Dropout(p=0.1, inplace=False)
+    (out_proj): Linear(in_features=768, out_features=2, bias=True)
+  )
+)
+2026-04-15 10:25:18,063 - INFO - train_pipeline - ===== Tokenizer Summary =====
+2026-04-15 10:25:18,087 - INFO - train_pipeline - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
+2026-04-15 10:25:18,097 - INFO - train_pipeline - ===== End of Architecture Log =====
+2026-04-15 10:25:18,108 - INFO - train_pipeline - Base model weights frozen – only classifier head will be trained.
+2026-04-15 10:25:19,118 - INFO - train_pipeline - === Starting training ===
+2026-04-15 10:25:16,900 - INFO - train_pipeline - Logging to ./taskA-codebert-base/training.log
+2026-04-15 10:25:16,911 - INFO - train_pipeline - Loading model & tokenizer for 'microsoft/codebert-base'
+2026-04-15 10:25:18,028 - INFO - train_pipeline - Model placed on cuda
+2026-04-15 10:25:18,039 - INFO - train_pipeline - ===== Model Architecture =====
+2026-04-15 10:25:18,050 - INFO - train_pipeline - 
+RobertaForSequenceClassification(
+  (roberta): RobertaModel(
+    (embeddings): RobertaEmbeddings(
+      (word_embeddings): Embedding(50265, 768, padding_idx=1)
+      (position_embeddings): Embedding(514, 768, padding_idx=1)
+      (token_type_embeddings): Embedding(1, 768)
+      (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+      (dropout): Dropout(p=0.1, inplace=False)
+    )
+    (encoder): RobertaEncoder(
+      (layer): ModuleList(
+        (0-11): 12 x RobertaLayer(
+          (attention): RobertaAttention(
+            (self): RobertaSdpaSelfAttention(
+              (query): Linear(in_features=768, out_features=768, bias=True)
+              (key): Linear(in_features=768, out_features=768, bias=True)
+              (value): Linear(in_features=768, out_features=768, bias=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+            (output): RobertaSelfOutput(
+              (dense): Linear(in_features=768, out_features=768, bias=True)
+              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+          )
+          (intermediate): RobertaIntermediate(
+            (dense): Linear(in_features=768, out_features=3072, bias=True)
+            (intermediate_act_fn): GELUActivation()
+          )
+          (output): RobertaOutput(
+            (dense): Linear(in_features=3072, out_features=768, bias=True)
+            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+            (dropout): Dropout(p=0.1, inplace=False)
+          )
+        )
+      )
+    )
+  )
+  (classifier): RobertaClassificationHead(
+    (dense): Linear(in_features=768, out_features=768, bias=True)
+    (dropout): Dropout(p=0.1, inplace=False)
+    (out_proj): Linear(in_features=768, out_features=2, bias=True)
+  )
+)
+2026-04-15 10:25:18,063 - INFO - train_pipeline - ===== Tokenizer Summary =====
+2026-04-15 10:25:18,087 - INFO - train_pipeline - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
+2026-04-15 10:25:18,097 - INFO - train_pipeline - ===== End of Architecture Log =====
+2026-04-15 10:25:18,108 - INFO - train_pipeline - Base model weights frozen – only classifier head will be trained.
+2026-04-15 10:25:19,118 - INFO - train_pipeline - === Starting training ===
+2026-04-15 10:25:16,900 - INFO - train_pipeline - Logging to ./taskA-codebert-base/training.log
+2026-04-15 10:25:16,911 - INFO - train_pipeline - Loading model & tokenizer for 'microsoft/codebert-base'
+2026-04-15 10:25:18,028 - INFO - train_pipeline - Model placed on cuda
+2026-04-15 10:25:18,039 - INFO - train_pipeline - ===== Model Architecture =====
+2026-04-15 10:25:18,050 - INFO - train_pipeline - 
+RobertaForSequenceClassification(
+  (roberta): RobertaModel(
+    (embeddings): RobertaEmbeddings(
+      (word_embeddings): Embedding(50265, 768, padding_idx=1)
+      (position_embeddings): Embedding(514, 768, padding_idx=1)
+      (token_type_embeddings): Embedding(1, 768)
+      (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+      (dropout): Dropout(p=0.1, inplace=False)
+    )
+    (encoder): RobertaEncoder(
+      (layer): ModuleList(
+        (0-11): 12 x RobertaLayer(
+          (attention): RobertaAttention(
+            (self): RobertaSdpaSelfAttention(
+              (query): Linear(in_features=768, out_features=768, bias=True)
+              (key): Linear(in_features=768, out_features=768, bias=True)
+              (value): Linear(in_features=768, out_features=768, bias=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+            (output): RobertaSelfOutput(
+              (dense): Linear(in_features=768, out_features=768, bias=True)
+              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+          )
+          (intermediate): RobertaIntermediate(
+            (dense): Linear(in_features=768, out_features=3072, bias=True)
+            (intermediate_act_fn): GELUActivation()
+          )
+          (output): RobertaOutput(
+            (dense): Linear(in_features=3072, out_features=768, bias=True)
+            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+            (dropout): Dropout(p=0.1, inplace=False)
+          )
+        )
+      )
+    )
+  )
+  (classifier): RobertaClassificationHead(
+    (dense): Linear(in_features=768, out_features=768, bias=True)
+    (dropout): Dropout(p=0.1, inplace=False)
+    (out_proj): Linear(in_features=768, out_features=2, bias=True)
+  )
+)
+2026-04-15 10:25:18,063 - INFO - train_pipeline - ===== Tokenizer Summary =====
+2026-04-15 10:25:18,087 - INFO - train_pipeline - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
+2026-04-15 10:25:18,097 - INFO - train_pipeline - ===== End of Architecture Log =====
+2026-04-15 10:25:18,108 - INFO - train_pipeline - Base model weights frozen – only classifier head will be trained.
+2026-04-15 10:25:19,118 - INFO - train_pipeline - === Starting training ===
+2026-04-15 10:25:16,900 - INFO - train_pipeline - Logging to ./taskA-codebert-base/training.log
+2026-04-15 10:25:16,911 - INFO - train_pipeline - Loading model & tokenizer for 'microsoft/codebert-base'
+2026-04-15 10:25:18,028 - INFO - train_pipeline - Model placed on cuda
+2026-04-15 10:25:18,039 - INFO - train_pipeline - ===== Model Architecture =====
+2026-04-15 10:25:18,050 - INFO - train_pipeline - 
+RobertaForSequenceClassification(
+  (roberta): RobertaModel(
+    (embeddings): RobertaEmbeddings(
+      (word_embeddings): Embedding(50265, 768, padding_idx=1)
+      (position_embeddings): Embedding(514, 768, padding_idx=1)
+      (token_type_embeddings): Embedding(1, 768)
+      (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+      (dropout): Dropout(p=0.1, inplace=False)
+    )
+    (encoder): RobertaEncoder(
+      (layer): ModuleList(
+        (0-11): 12 x RobertaLayer(
+          (attention): RobertaAttention(
+            (self): RobertaSdpaSelfAttention(
+              (query): Linear(in_features=768, out_features=768, bias=True)
+              (key): Linear(in_features=768, out_features=768, bias=True)
+              (value): Linear(in_features=768, out_features=768, bias=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+            (output): RobertaSelfOutput(
+              (dense): Linear(in_features=768, out_features=768, bias=True)
+              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+          )
+          (intermediate): RobertaIntermediate(
+            (dense): Linear(in_features=768, out_features=3072, bias=True)
+            (intermediate_act_fn): GELUActivation()
+          )
+          (output): RobertaOutput(
+            (dense): Linear(in_features=3072, out_features=768, bias=True)
+            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+            (dropout): Dropout(p=0.1, inplace=False)
+          )
+        )
+      )
+    )
+  )
+  (classifier): RobertaClassificationHead(
+    (dense): Linear(in_features=768, out_features=768, bias=True)
+    (dropout): Dropout(p=0.1, inplace=False)
+    (out_proj): Linear(in_features=768, out_features=2, bias=True)
+  )
+)
+2026-04-15 10:25:18,063 - INFO - train_pipeline - ===== Tokenizer Summary =====
+2026-04-15 10:25:18,087 - INFO - train_pipeline - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
+2026-04-15 10:25:18,097 - INFO - train_pipeline - ===== End of Architecture Log =====
+2026-04-15 10:25:18,108 - INFO - train_pipeline - Base model weights frozen – only classifier head will be trained.
+2026-04-15 10:25:19,118 - INFO - train_pipeline - === Starting training ===
+2026-04-15 10:25:16,900 - INFO - train_pipeline - Logging to ./taskA-codebert-base/training.log
+2026-04-15 10:25:16,911 - INFO - train_pipeline - Loading model & tokenizer for 'microsoft/codebert-base'
+2026-04-15 10:25:18,028 - INFO - train_pipeline - Model placed on cuda
+2026-04-15 10:25:18,039 - INFO - train_pipeline - ===== Model Architecture =====
+2026-04-15 10:25:18,050 - INFO - train_pipeline - 
+RobertaForSequenceClassification(
+  (roberta): RobertaModel(
+    (embeddings): RobertaEmbeddings(
+      (word_embeddings): Embedding(50265, 768, padding_idx=1)
+      (position_embeddings): Embedding(514, 768, padding_idx=1)
+      (token_type_embeddings): Embedding(1, 768)
+      (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+      (dropout): Dropout(p=0.1, inplace=False)
+    )
+    (encoder): RobertaEncoder(
+      (layer): ModuleList(
+        (0-11): 12 x RobertaLayer(
+          (attention): RobertaAttention(
+            (self): RobertaSdpaSelfAttention(
+              (query): Linear(in_features=768, out_features=768, bias=True)
+              (key): Linear(in_features=768, out_features=768, bias=True)
+              (value): Linear(in_features=768, out_features=768, bias=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+            (output): RobertaSelfOutput(
+              (dense): Linear(in_features=768, out_features=768, bias=True)
+              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+              (dropout): Dropout(p=0.1, inplace=False)
+            )
+          )
+          (intermediate): RobertaIntermediate(
+            (dense): Linear(in_features=768, out_features=3072, bias=True)
+            (intermediate_act_fn): GELUActivation()
+          )
+          (output): RobertaOutput(
+            (dense): Linear(in_features=3072, out_features=768, bias=True)
+            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+            (dropout): Dropout(p=0.1, inplace=False)
+          )
+        )
+      )
+    )
+  )
+  (classifier): RobertaClassificationHead(
+    (dense): Linear(in_features=768, out_features=768, bias=True)
+    (dropout): Dropout(p=0.1, inplace=False)
+    (out_proj): Linear(in_features=768, out_features=2, bias=True)
+  )
+)
+2026-04-15 10:25:18,063 - INFO - train_pipeline - ===== Tokenizer Summary =====
+2026-04-15 10:25:18,087 - INFO - train_pipeline - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
+2026-04-15 10:25:18,097 - INFO - train_pipeline - ===== End of Architecture Log =====
+2026-04-15 10:25:18,108 - INFO - train_pipeline - Base model weights frozen – only classifier head will be trained.
+2026-04-15 10:25:19,118 - INFO - train_pipeline - === Starting training ===