2026-04-16 09:16:18,911 - INFO - Loading model and tokenizer from: ./checkpoints/graphcodebert-robust/checkpoint-200
2026-04-16 09:16:20,848 - INFO - ===== Model Architecture =====
2026-04-16 09:16:20,851 - 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 09:16:20,853 - INFO - ===== Parameter Summary =====
2026-04-16 09:16:20,855 - INFO - Total Parameters:         124,647,170
2026-04-16 09:16:20,857 - INFO - Trainable Parameters:     124,647,170
2026-04-16 09:16:20,858 - INFO - Non-trainable Parameters: 0
2026-04-16 09:16:20,859 - INFO - ===== Tokenizer Summary =====
2026-04-16 09:16:20,874 - INFO - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
2026-04-16 09:16:20,876 - INFO - ===== End of Architecture Log =====
2026-04-16 09:16:21,287 - INFO - Loading dataset: DaniilOr/SemEval-2026-Task13 (A)
2026-04-16 09:16:28,538 - INFO - Tokenizing dataset...
2026-04-16 09:16:29,324 - INFO - Running inference on 1000 examples...
2026-04-16 09:16:55,877 - INFO - Calculating classification metrics...
2026-04-16 09:16:55,902 - INFO - ------------------------------
2026-04-16 09:16:55,904 - INFO - METRICS FOR SPLIT: test
2026-04-16 09:16:55,905 - INFO - Accuracy:  0.5030
2026-04-16 09:16:55,907 - INFO - Precision: 0.6228
2026-04-16 09:16:55,909 - INFO - Recall:    0.5030
2026-04-16 09:16:55,910 - INFO - F1-Score:  0.5438
2026-04-16 09:16:55,912 - INFO - ------------------------------
2026-04-16 09:16:55,918 - INFO - Confusion Matrix:
[[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...