Hugging Face's logo

dzungpham
/
graphcodebert-code-classification

Transformers
Safetensors
detection
AI-generated
bert
Model card Files
xet
 Community
graphcodebert-code-classification
File size: 5,106 Bytes
cd51d8b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ff9dd86
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cd51d8b
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
2026-04-24 17:43:47,257 - INFO - ===== Training Configuration =====
2026-04-24 17:43:47,259 - INFO - model_name           : microsoft/graphcodebert-base
2026-04-24 17:43:47,260 - INFO - output_dir           : output_checkpoints/graphcodebert-base-lowLR-highBatchSize/
2026-04-24 17:43:47,261 - INFO - num_epochs           : 1
2026-04-24 17:43:47,261 - INFO - max_steps            : -1
2026-04-24 17:43:47,262 - INFO - batch_size           : 256
2026-04-24 17:43:47,264 - INFO - learning_rate        : 1e-06
2026-04-24 17:43:47,265 - INFO - max_length           : 512
2026-04-24 17:43:47,266 - INFO - num_labels           : 2
2026-04-24 17:43:47,267 - INFO - use_wandb            : True
2026-04-24 17:43:47,268 - INFO - freeze_base          : True
2026-04-24 17:43:47,269 - INFO - loss_type            : r-drop
2026-04-24 17:43:47,270 - INFO - focal_alpha          : 1.0
2026-04-24 17:43:47,271 - INFO - focal_gamma          : 2.0
2026-04-24 17:43:47,272 - INFO - r_drop_alpha         : 10.0
2026-04-24 17:43:47,273 - INFO - infonce_temperature  : 0.07
2026-04-24 17:43:47,274 - INFO - infonce_weight       : 0.5
2026-04-24 17:43:47,275 - INFO - seed                 : 42
2026-04-24 17:43:47,276 - INFO - resume_from_checkpoint : checkpoints/graphcodebert-base-lowLR-highBatchSize/checkpoint-500
2026-04-24 17:43:47,276 - INFO - label_smoothing      : 0.5
2026-04-24 17:43:47,277 - INFO - adversarial_epsilon  : 0.5
2026-04-24 17:43:47,279 - INFO - use_swa              : False
2026-04-24 17:43:47,279 - INFO - swa_start_epoch      : 0
2026-04-24 17:43:47,280 - INFO - swa_lr               : 1e-05
2026-04-24 17:43:47,281 - INFO - data_augmentation    : True
2026-04-24 17:43:47,282 - INFO - aug_rename_prob      : 0.8
2026-04-24 17:43:47,283 - INFO - aug_format_prob      : 0.8
2026-04-24 17:43:47,285 - INFO - mixup_alpha          : 1.0
2026-04-24 17:43:47,285 - INFO - low_pass_keep_ratio  : 0.5
2026-04-24 17:43:47,286 - INFO - freq_consistency_weight : 0.5
2026-04-24 17:43:47,288 - INFO - hidden_dropout_prob  : 0.3
2026-04-24 17:43:47,289 - INFO - attention_probs_dropout_prob : 0.3
2026-04-24 17:43:47,290 - INFO - classifier_dropout   : 0.3
2026-04-24 17:43:47,290 - INFO - =================================
2026-04-24 17:43:53,414 - INFO - Model placed on cuda
2026-04-24 17:43:53,416 - INFO - ===== Model Architecture =====
2026-04-24 17:43:53,419 - 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=0.3, 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=0.3, 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=0.3, 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=0.3, inplace=False)
          )
        )
      )
    )
  )
  (classifier): RobertaClassificationHead(
    (dense): Linear(in_features=768, out_features=768, bias=True)
    (dropout): Dropout(p=0.3, inplace=False)
    (out_proj): Linear(in_features=768, out_features=2, bias=True)
  )
)
2026-04-24 17:43:53,420 - INFO - ===== Parameter Summary =====
2026-04-24 17:43:53,421 - INFO - Total Parameters:         124,647,170
2026-04-24 17:43:53,422 - INFO - Trainable Parameters:     592,130
2026-04-24 17:43:53,423 - INFO - Non-trainable Parameters: 124,055,040
2026-04-24 17:43:53,423 - INFO - ===== Tokenizer Summary =====
2026-04-24 17:43:53,438 - INFO - Vocab size: 50265 | Special tokens: ['<s>', '</s>', '<unk>', '<pad>', '<mask>']
2026-04-24 17:43:53,438 - INFO - ===== End of Architecture Log =====
2026-04-24 17:43:53,439 - INFO - Data augmentation enabled (rename=0.8, format=0.8)
2026-04-24 17:47:51,059 - INFO - === Starting training with MixCode + FFT low-pass consistency ===
2026-04-24 20:25:51,228 - INFO - Training completed successfully.
2026-04-24 20:25:52,599 - INFO - Final model saved to output_checkpoints/graphcodebert-base-lowLR-highBatchSize/final_model