upload graphcodebert robust, best f1 score at 0.54 at robust checkpoint 200

inference/graphcodebert-robust/inference.log

@@ -67,3 +67,117 @@ RobertaForSequenceClassification(
 [[422 355]
  [142  81]]
 2026-04-16 09:16:55,921 - INFO - ✅ Predictions saved to test/inference/graphcodebert-robust/submission.csv
+2026-04-16 10:06:49,138 - INFO - Loading model and tokenizer from: ./output_checkpoints/graphcodebert-robust/checkpoint-1000
+2026-04-16 10:06:49,327 - INFO - ===== Model Architecture =====
+2026-04-16 10:06:49,331 - INFO - 
+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=0.2, 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=0.2, 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=0.2, 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=0.2, inplace=False)
+          )
+        )
+      )
+    )
+  )
+  (classifier): RobertaClassificationHead(
+    (dense): Linear(in_features=768, out_features=768, bias=True)
+    (dropout): Dropout(p=0.2, inplace=False)
+    (out_proj): Linear(in_features=768, out_features=2, bias=True)
+  )
+)
+2026-04-16 10:06:49,337 - INFO - ===== Parameter Summary =====
+2026-04-16 10:06:49,340 - INFO - Total Parameters:         124,647,170
+2026-04-16 10:06:49,343 - INFO - Trainable Parameters:     124,647,170
+2026-04-16 10:06:49,346 - INFO - Non-trainable Parameters: 0
+2026-04-16 10:06:49,349 - INFO - ===== Tokenizer Summary =====
+2026-04-16 10:06:49,366 - INFO - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
+2026-04-16 10:06:49,369 - INFO - ===== End of Architecture Log =====
+2026-04-16 10:06:49,539 - INFO - Loading dataset: dzungpham/SemEval-2026-TaskA-dataset (default)
+2026-04-16 10:08:44,659 - INFO - Loading model and tokenizer from: ./output_checkpoints/graphcodebert-robust/checkpoint-1000
+2026-04-16 10:08:44,856 - INFO - ===== Model Architecture =====
+2026-04-16 10:08:44,861 - INFO - 
+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=0.2, 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=0.2, 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=0.2, 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=0.2, inplace=False)
+          )
+        )
+      )
+    )
+  )
+  (classifier): RobertaClassificationHead(
+    (dense): Linear(in_features=768, out_features=768, bias=True)
+    (dropout): Dropout(p=0.2, inplace=False)
+    (out_proj): Linear(in_features=768, out_features=2, bias=True)
+  )
+)
+2026-04-16 10:08:44,865 - INFO - ===== Parameter Summary =====
+2026-04-16 10:08:44,867 - INFO - Total Parameters:         124,647,170
+2026-04-16 10:08:44,870 - INFO - Trainable Parameters:     124,647,170
+2026-04-16 10:08:44,874 - INFO - Non-trainable Parameters: 0
+2026-04-16 10:08:44,876 - INFO - ===== Tokenizer Summary =====
+2026-04-16 10:08:44,893 - INFO - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
+2026-04-16 10:08:44,896 - INFO - ===== End of Architecture Log =====
+2026-04-16 10:08:45,082 - INFO - Loading dataset: dzungpham/SemEval-2026-TaskA-dataset (default)
+2026-04-16 10:08:51,304 - WARNING - Default loading failed due to schema mismatch: An error occurred while generating the dataset
+2026-04-16 10:08:51,307 - INFO - Attempting to load split 'test' using data_files...
+2026-04-16 10:08:55,114 - INFO - Tokenizing dataset...
+2026-04-16 10:14:03,634 - INFO - Running inference on 500000 examples...