cedricbonhomme commited on
Commit
6f06f3e
·
verified ·
1 Parent(s): 20bc801

CIRCL/cwe-parent-vulnerability-classification-roberta-base

Browse files
Files changed (5) hide show
  1. README.md +49 -52
  2. config.json +52 -52
  3. emissions.csv +1 -1
  4. metrics.json +6 -6
  5. model.safetensors +1 -1
README.md CHANGED
@@ -9,8 +9,6 @@ metrics:
9
  model-index:
10
  - name: cwe-parent-vulnerability-classification-roberta-base
11
  results: []
12
- datasets:
13
- - CIRCL/vulnerability-cwe-patch
14
  ---
15
 
16
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -18,12 +16,11 @@ should probably proofread and complete it, then remove this comment. -->
18
 
19
  # cwe-parent-vulnerability-classification-roberta-base
20
 
21
- This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base).
22
-
23
  It achieves the following results on the evaluation set:
24
- - Loss: 1.2078
25
- - Accuracy: 0.875
26
- - F1 Macro: 0.6248
27
 
28
  ## Model description
29
 
@@ -46,7 +43,7 @@ The following hyperparameters were used during training:
46
  - train_batch_size: 32
47
  - eval_batch_size: 32
48
  - seed: 42
49
- - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
50
  - lr_scheduler_type: linear
51
  - num_epochs: 40
52
 
@@ -54,51 +51,51 @@ The following hyperparameters were used during training:
54
 
55
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 Macro |
56
  |:-------------:|:-----:|:----:|:---------------:|:--------:|:--------:|
57
- | 3.2699 | 1.0 | 25 | 3.1492 | 0.0341 | 0.0055 |
58
- | 3.1972 | 2.0 | 50 | 2.9909 | 0.0114 | 0.0064 |
59
- | 3.1211 | 3.0 | 75 | 3.0017 | 0.0341 | 0.0140 |
60
- | 3.0888 | 4.0 | 100 | 3.0223 | 0.2841 | 0.0463 |
61
- | 2.9467 | 5.0 | 125 | 2.9608 | 0.0114 | 0.0018 |
62
- | 2.9851 | 6.0 | 150 | 2.8743 | 0.1932 | 0.0641 |
63
- | 2.9083 | 7.0 | 175 | 2.7687 | 0.375 | 0.0963 |
64
- | 2.7652 | 8.0 | 200 | 2.7049 | 0.4318 | 0.1953 |
65
- | 2.6893 | 9.0 | 225 | 2.5547 | 0.4886 | 0.1952 |
66
- | 2.5636 | 10.0 | 250 | 2.4970 | 0.5682 | 0.3314 |
67
- | 2.477 | 11.0 | 275 | 2.3499 | 0.6136 | 0.3790 |
68
- | 2.2936 | 12.0 | 300 | 2.2659 | 0.6364 | 0.3949 |
69
- | 2.1369 | 13.0 | 325 | 2.1758 | 0.625 | 0.4002 |
70
- | 2.0615 | 14.0 | 350 | 2.1015 | 0.6477 | 0.4169 |
71
- | 1.9548 | 15.0 | 375 | 1.9444 | 0.6932 | 0.3972 |
72
- | 1.7943 | 16.0 | 400 | 1.8892 | 0.6818 | 0.4210 |
73
- | 1.6619 | 17.0 | 425 | 1.8439 | 0.6818 | 0.4149 |
74
- | 1.5391 | 18.0 | 450 | 1.7247 | 0.7159 | 0.4848 |
75
- | 1.4415 | 19.0 | 475 | 1.6650 | 0.7273 | 0.4749 |
76
- | 1.2834 | 20.0 | 500 | 1.5743 | 0.7727 | 0.5574 |
77
- | 1.2245 | 21.0 | 525 | 1.5396 | 0.7614 | 0.5373 |
78
- | 1.1629 | 22.0 | 550 | 1.5005 | 0.7614 | 0.5350 |
79
- | 1.0894 | 23.0 | 575 | 1.4478 | 0.7614 | 0.5383 |
80
- | 0.9755 | 24.0 | 600 | 1.4335 | 0.7841 | 0.5599 |
81
- | 0.9271 | 25.0 | 625 | 1.4195 | 0.7841 | 0.5562 |
82
- | 0.8761 | 26.0 | 650 | 1.3740 | 0.8182 | 0.6015 |
83
- | 0.8312 | 27.0 | 675 | 1.3479 | 0.8295 | 0.6086 |
84
- | 0.7523 | 28.0 | 700 | 1.3379 | 0.8295 | 0.5948 |
85
- | 0.718 | 29.0 | 725 | 1.2991 | 0.8295 | 0.5948 |
86
- | 0.6819 | 30.0 | 750 | 1.3059 | 0.8409 | 0.6047 |
87
- | 0.6771 | 31.0 | 775 | 1.2650 | 0.8636 | 0.6167 |
88
- | 0.6267 | 32.0 | 800 | 1.2905 | 0.8523 | 0.6252 |
89
- | 0.6068 | 33.0 | 825 | 1.2559 | 0.875 | 0.6248 |
90
- | 0.5811 | 34.0 | 850 | 1.2371 | 0.875 | 0.6248 |
91
- | 0.5579 | 35.0 | 875 | 1.2231 | 0.875 | 0.6248 |
92
- | 0.5385 | 36.0 | 900 | 1.2342 | 0.875 | 0.6248 |
93
- | 0.5334 | 37.0 | 925 | 1.2255 | 0.875 | 0.6248 |
94
- | 0.4868 | 38.0 | 950 | 1.2223 | 0.875 | 0.6248 |
95
- | 0.5228 | 39.0 | 975 | 1.2078 | 0.875 | 0.6248 |
96
- | 0.5325 | 40.0 | 1000 | 1.2101 | 0.875 | 0.6248 |
97
 
98
 
99
  ### Framework versions
100
 
101
- - Transformers 4.55.4
102
- - Pytorch 2.7.1+cu126
103
- - Datasets 4.0.0
104
- - Tokenizers 0.21.2
 
9
  model-index:
10
  - name: cwe-parent-vulnerability-classification-roberta-base
11
  results: []
 
 
12
  ---
13
 
14
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
16
 
17
  # cwe-parent-vulnerability-classification-roberta-base
18
 
19
+ This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on an unknown dataset.
 
20
  It achieves the following results on the evaluation set:
21
+ - Loss: 1.8213
22
+ - Accuracy: 0.5673
23
+ - F1 Macro: 0.3582
24
 
25
  ## Model description
26
 
 
43
  - train_batch_size: 32
44
  - eval_batch_size: 32
45
  - seed: 42
46
+ - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
47
  - lr_scheduler_type: linear
48
  - num_epochs: 40
49
 
 
51
 
52
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 Macro |
53
  |:-------------:|:-----:|:----:|:---------------:|:--------:|:--------:|
54
+ | 3.2359 | 1.0 | 237 | 3.0937 | 0.0334 | 0.0114 |
55
+ | 2.3278 | 2.0 | 474 | 2.2992 | 0.4303 | 0.2323 |
56
+ | 1.8161 | 3.0 | 711 | 1.9653 | 0.5089 | 0.2917 |
57
+ | 1.6283 | 4.0 | 948 | 1.8788 | 0.5292 | 0.3007 |
58
+ | 1.3258 | 5.0 | 1185 | 1.8253 | 0.5864 | 0.3392 |
59
+ | 1.3328 | 6.0 | 1422 | 1.8213 | 0.5673 | 0.3582 |
60
+ | 1.2914 | 7.0 | 1659 | 1.8573 | 0.5900 | 0.3604 |
61
+ | 0.8614 | 8.0 | 1896 | 1.8345 | 0.6198 | 0.3810 |
62
+ | 0.8253 | 9.0 | 2133 | 2.0121 | 0.6067 | 0.3568 |
63
+ | 0.7363 | 10.0 | 2370 | 2.0059 | 0.6508 | 0.4071 |
64
+ | 0.6406 | 11.0 | 2607 | 2.0866 | 0.6472 | 0.3906 |
65
+ | 0.4574 | 12.0 | 2844 | 2.1904 | 0.6377 | 0.3902 |
66
+ | 0.4834 | 13.0 | 3081 | 2.2417 | 0.6281 | 0.3797 |
67
+ | 0.3078 | 14.0 | 3318 | 2.3233 | 0.6830 | 0.4132 |
68
+ | 0.3217 | 15.0 | 3555 | 2.4203 | 0.6639 | 0.4020 |
69
+ | 0.2823 | 16.0 | 3792 | 2.4227 | 0.6615 | 0.3918 |
70
+ | 0.2392 | 17.0 | 4029 | 2.6133 | 0.6865 | 0.3844 |
71
+ | 0.1956 | 18.0 | 4266 | 2.6611 | 0.6675 | 0.3789 |
72
+ | 0.1504 | 19.0 | 4503 | 2.7612 | 0.6746 | 0.4158 |
73
+ | 0.1109 | 20.0 | 4740 | 2.8752 | 0.6710 | 0.3776 |
74
+ | 0.1091 | 21.0 | 4977 | 3.0530 | 0.6949 | 0.3776 |
75
+ | 0.13 | 22.0 | 5214 | 3.0540 | 0.6889 | 0.3766 |
76
+ | 0.0797 | 23.0 | 5451 | 3.2854 | 0.6770 | 0.4038 |
77
+ | 0.145 | 24.0 | 5688 | 3.2146 | 0.6973 | 0.3877 |
78
+ | 0.1004 | 25.0 | 5925 | 3.4159 | 0.6937 | 0.3850 |
79
+ | 0.0486 | 26.0 | 6162 | 3.4003 | 0.6865 | 0.3767 |
80
+ | 0.0544 | 27.0 | 6399 | 3.3643 | 0.6889 | 0.3822 |
81
+ | 0.082 | 28.0 | 6636 | 3.4874 | 0.6913 | 0.3818 |
82
+ | 0.0468 | 29.0 | 6873 | 3.5810 | 0.6877 | 0.3815 |
83
+ | 0.0308 | 30.0 | 7110 | 3.7565 | 0.6949 | 0.3837 |
84
+ | 0.0366 | 31.0 | 7347 | 3.6714 | 0.6961 | 0.3814 |
85
+ | 0.0388 | 32.0 | 7584 | 3.8502 | 0.6973 | 0.3957 |
86
+ | 0.0341 | 33.0 | 7821 | 3.8415 | 0.6973 | 0.3973 |
87
+ | 0.0417 | 34.0 | 8058 | 3.9342 | 0.7056 | 0.3864 |
88
+ | 0.0218 | 35.0 | 8295 | 3.9367 | 0.6996 | 0.3843 |
89
+ | 0.0607 | 36.0 | 8532 | 3.9262 | 0.6996 | 0.3841 |
90
+ | 0.0148 | 37.0 | 8769 | 3.9609 | 0.7020 | 0.4189 |
91
+ | 0.0186 | 38.0 | 9006 | 3.9656 | 0.6985 | 0.3822 |
92
+ | 0.0387 | 39.0 | 9243 | 3.9738 | 0.7044 | 0.3850 |
93
+ | 0.008 | 40.0 | 9480 | 3.9906 | 0.7044 | 0.3861 |
94
 
95
 
96
  ### Framework versions
97
 
98
+ - Transformers 4.57.1
99
+ - Pytorch 2.9.1+cu128
100
+ - Datasets 4.4.1
101
+ - Tokenizers 0.22.1
config.json CHANGED
@@ -11,62 +11,62 @@
11
  "hidden_dropout_prob": 0.1,
12
  "hidden_size": 768,
13
  "id2label": {
14
- "0": "LABEL_0",
15
- "1": "LABEL_1",
16
- "2": "LABEL_2",
17
- "3": "LABEL_3",
18
- "4": "LABEL_4",
19
- "5": "LABEL_5",
20
- "6": "LABEL_6",
21
- "7": "LABEL_7",
22
- "8": "LABEL_8",
23
- "9": "LABEL_9",
24
- "10": "LABEL_10",
25
- "11": "LABEL_11",
26
- "12": "LABEL_12",
27
- "13": "LABEL_13",
28
- "14": "LABEL_14",
29
- "15": "LABEL_15",
30
- "16": "LABEL_16",
31
- "17": "LABEL_17",
32
- "18": "LABEL_18",
33
- "19": "LABEL_19",
34
- "20": "LABEL_20",
35
- "21": "LABEL_21",
36
- "22": "LABEL_22",
37
- "23": "LABEL_23",
38
- "24": "LABEL_24",
39
- "25": "LABEL_25"
40
  },
41
  "initializer_range": 0.02,
42
  "intermediate_size": 3072,
43
  "label2id": {
44
- "LABEL_0": 0,
45
- "LABEL_1": 1,
46
- "LABEL_10": 10,
47
- "LABEL_11": 11,
48
- "LABEL_12": 12,
49
- "LABEL_13": 13,
50
- "LABEL_14": 14,
51
- "LABEL_15": 15,
52
- "LABEL_16": 16,
53
- "LABEL_17": 17,
54
- "LABEL_18": 18,
55
- "LABEL_19": 19,
56
- "LABEL_2": 2,
57
- "LABEL_20": 20,
58
- "LABEL_21": 21,
59
- "LABEL_22": 22,
60
- "LABEL_23": 23,
61
- "LABEL_24": 24,
62
- "LABEL_25": 25,
63
- "LABEL_3": 3,
64
- "LABEL_4": 4,
65
- "LABEL_5": 5,
66
- "LABEL_6": 6,
67
- "LABEL_7": 7,
68
- "LABEL_8": 8,
69
- "LABEL_9": 9
70
  },
71
  "layer_norm_eps": 1e-05,
72
  "max_position_embeddings": 514,
 
11
  "hidden_dropout_prob": 0.1,
12
  "hidden_size": 768,
13
  "id2label": {
14
+ "0": "1025",
15
+ "1": "1071",
16
+ "2": "131",
17
+ "3": "138",
18
+ "4": "284",
19
+ "5": "285",
20
+ "6": "435",
21
+ "7": "436",
22
+ "8": "595",
23
+ "9": "657",
24
+ "10": "664",
25
+ "11": "682",
26
+ "12": "684",
27
+ "13": "691",
28
+ "14": "693",
29
+ "15": "697",
30
+ "16": "703",
31
+ "17": "706",
32
+ "18": "707",
33
+ "19": "710",
34
+ "20": "74",
35
+ "21": "754",
36
+ "22": "829",
37
+ "23": "862",
38
+ "24": "913",
39
+ "25": "94"
40
  },
41
  "initializer_range": 0.02,
42
  "intermediate_size": 3072,
43
  "label2id": {
44
+ "1025": 0,
45
+ "1071": 1,
46
+ "131": 2,
47
+ "138": 3,
48
+ "284": 4,
49
+ "285": 5,
50
+ "435": 6,
51
+ "436": 7,
52
+ "595": 8,
53
+ "657": 9,
54
+ "664": 10,
55
+ "682": 11,
56
+ "684": 12,
57
+ "691": 13,
58
+ "693": 14,
59
+ "697": 15,
60
+ "703": 16,
61
+ "706": 17,
62
+ "707": 18,
63
+ "710": 19,
64
+ "74": 20,
65
+ "754": 21,
66
+ "829": 22,
67
+ "862": 23,
68
+ "913": 24,
69
+ "94": 25
70
  },
71
  "layer_norm_eps": 1e-05,
72
  "max_position_embeddings": 514,
emissions.csv CHANGED
@@ -1,2 +1,2 @@
1
  timestamp,project_name,run_id,experiment_id,duration,emissions,emissions_rate,cpu_power,gpu_power,ram_power,cpu_energy,gpu_energy,ram_energy,energy_consumed,country_name,country_iso_code,region,cloud_provider,cloud_region,os,python_version,codecarbon_version,cpu_count,cpu_model,gpu_count,gpu_model,longitude,latitude,ram_total_size,tracking_mode,on_cloud,pue
2
- 2025-09-03T13:07:51,codecarbon,a4931482-bbba-40f5-bb09-0c95cf8cc22d,5b0fa12a-3dd7-45bb-9766-cc326314d9f1,381.010146320099,0.006199580425386055,1.6271431312954028e-05,42.5,399.8939178594448,94.34468507766725,0.004493555260882164,0.04442788498671746,0.009974662137721463,0.05889610238532109,Luxembourg,LUX,luxembourg,,,Linux-6.8.0-71-generic-x86_64-with-glibc2.39,3.12.3,2.8.4,64,AMD EPYC 9124 16-Core Processor,2,2 x NVIDIA L40S,6.1294,49.6113,251.5858268737793,machine,N,1.0
 
1
  timestamp,project_name,run_id,experiment_id,duration,emissions,emissions_rate,cpu_power,gpu_power,ram_power,cpu_energy,gpu_energy,ram_energy,energy_consumed,country_name,country_iso_code,region,cloud_provider,cloud_region,os,python_version,codecarbon_version,cpu_count,cpu_model,gpu_count,gpu_model,longitude,latitude,ram_total_size,tracking_mode,on_cloud,pue
2
+ 2025-11-21T08:46:27,codecarbon,bc463c8d-dd69-42da-9801-e35dabee8711,5b0fa12a-3dd7-45bb-9766-cc326314d9f1,2632.649387937039,0.050220601954459286,1.9076069219309326e-05,42.5,302.47241372038087,94.34468507766725,0.031054272330529063,0.3771095669652027,0.06893259491628256,0.4770964342120145,Luxembourg,LUX,luxembourg,,,Linux-6.8.0-71-generic-x86_64-with-glibc2.39,3.12.3,2.8.4,64,AMD EPYC 9124 16-Core Processor,2,2 x NVIDIA L40S,6.1294,49.6113,251.5858268737793,machine,N,1.0
metrics.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
- "eval_loss": 1.2077823877334595,
3
- "eval_accuracy": 0.875,
4
- "eval_f1_macro": 0.6247628726287263,
5
- "eval_runtime": 0.2845,
6
- "eval_samples_per_second": 309.345,
7
- "eval_steps_per_second": 10.546,
8
  "epoch": 40.0
9
  }
 
1
  {
2
+ "eval_loss": 1.821271300315857,
3
+ "eval_accuracy": 0.567342073897497,
4
+ "eval_f1_macro": 0.35818372664233605,
5
+ "eval_runtime": 2.5279,
6
+ "eval_samples_per_second": 331.898,
7
+ "eval_steps_per_second": 10.681,
8
  "epoch": 40.0
9
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b99289c0c2f3de76637a8067b418554499b4fdeecc09e90b4084705f13542c37
3
  size 498686648
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:097c2c599cc7c651891edb18006e591e48223f014edd119eff30a1bd53d18dfc
3
  size 498686648