Hugging Face's logo

Datasets:
AnnyNguyen
/
VulnBridge

Tasks:
Text Classification
Modalities:
Tabular
Text
Languages:
code
Size:
100K - 1M
Tags:
software-security
vulnerability-detection
cwe-classification
source-code
benchmark
c
License:
Dataset card Data Studio Files
xet
 Community
1
AnnyNguyen commited on
Commit
e47cb59
·
verified ·
1 Parent(s): 463b354

Upload 2 files

Browse files
Files changed (1) hide show
  1. README.md +24 -13
README.md CHANGED
@@ -19,35 +19,35 @@ configs:
19
  - config_name: top10
20
  data_files:
21
  - split: train
22
- path: top10/train.jsonl
23
  - split: validation
24
- path: top10/validation.jsonl
25
  - split: test
26
- path: top10/test.jsonl
27
  - config_name: top20
28
  data_files:
29
  - split: train
30
- path: top20/train.jsonl
31
  - split: validation
32
- path: top20/validation.jsonl
33
  - split: test
34
- path: top20/test.jsonl
35
  - config_name: top50
36
  data_files:
37
  - split: train
38
- path: top50/train.jsonl
39
  - split: validation
40
- path: top50/validation.jsonl
41
  - split: test
42
- path: top50/test.jsonl
43
  - config_name: full
44
  data_files:
45
  - split: train
46
- path: full/train.jsonl
47
  - split: validation
48
- path: full/validation.jsonl
49
  - split: test
50
- path: full/test.jsonl
51
  ---
52
 
53
  # VulnBridge
@@ -77,6 +77,17 @@ variants are:
77
  | Top-50 | 43,740 | 5,467 | 5,468 | 54,675 | 50 | 50.2 | 93.9 | 3.0 |
78
  | Full | 45,100 | 5,638 | 5,638 | 56,376 | 174 | 51.7 | N/A | 0.0 |
79
 
 
 
 
 
 
 
 
 
 
 
 
80
  ## Source Datasets
81
 
82
  VulnBridge is constructed from the following public function-level vulnerability
@@ -95,7 +106,7 @@ sources:
95
 
96
  ## Data Format
97
 
98
- Each split is distributed as JSON Lines. A typical record has the following
99
  fields:
100
 
101
  ```json
 
19
  - config_name: top10
20
  data_files:
21
  - split: train
22
+ path: top10_train.jsonl
23
  - split: validation
24
+ path: top10_val.jsonl
25
  - split: test
26
+ path: top10_test.jsonl
27
  - config_name: top20
28
  data_files:
29
  - split: train
30
+ path: top20_train.jsonl
31
  - split: validation
32
+ path: top20_val.jsonl
33
  - split: test
34
+ path: top20_test.jsonl
35
  - config_name: top50
36
  data_files:
37
  - split: train
38
+ path: top50_train.jsonl
39
  - split: validation
40
+ path: top50_val.jsonl
41
  - split: test
42
+ path: top50_test.jsonl
43
  - config_name: full
44
  data_files:
45
  - split: train
46
+ path: full_train.jsonl
47
  - split: validation
48
+ path: full_val.jsonl
49
  - split: test
50
+ path: full_test.jsonl
51
  ---
52
 
53
  # VulnBridge
 
77
  | Top-50 | 43,740 | 5,467 | 5,468 | 54,675 | 50 | 50.2 | 93.9 | 3.0 |
78
  | Full | 45,100 | 5,638 | 5,638 | 56,376 | 174 | 51.7 | N/A | 0.0 |
79
 
80
+ ## Repository Files
81
+
82
+ The Hugging Face repository uses a flat file layout:
83
+
84
+ | Variant | Train | Validation | Test | CWE mapping |
85
+ |---|---|---|---|---|
86
+ | Top-10 | `top10_train.jsonl` | `top10_val.jsonl` | `top10_test.jsonl` | `top10_mapping` |
87
+ | Top-20 | `top20_train.jsonl` | `top20_val.jsonl` | `top20_test.jsonl` | `top20_mapping` |
88
+ | Top-50 | `top50_train.jsonl` | `top50_val.jsonl` | `top50_test.jsonl` | `top50_mapping` |
89
+ | Full | `full_train.jsonl` | `full_val.jsonl` | `full_test.jsonl` | `full_mapping` |
90
+
91
  ## Source Datasets
92
 
93
  VulnBridge is constructed from the following public function-level vulnerability
 
106
 
107
  ## Data Format
108
 
109
+ Each split is distributed as JSON Lines using the flat file names listed above. A typical record has the following
110
  fields:
111
 
112
  ```json