Hugging Face's logo

spidercob
/
code-risk-classifier

Text Classification
setfit
Safetensors
sentence-transformers
bert
generated_from_setfit_trainer
Eval Results (legacy)
text-embeddings-inference
Model card Files
xet
 Community
Sumeetgpt commited on
Commit
d6afa7c
·
verified ·
1 Parent(s): d8a0b7c

Push model using huggingface_hub.

Browse files
Files changed (3) hide show
  1. README.md +101 -2
  2. model.safetensors +1 -1
  3. model_head.pkl +1 -1
README.md CHANGED
@@ -4,12 +4,47 @@ tags:
4
  - sentence-transformers
5
  - text-classification
6
  - generated_from_setfit_trainer
7
- widget: []
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8
  metrics:
9
  - accuracy
10
  pipeline_tag: text-classification
11
  library_name: setfit
12
  inference: true
 
 
 
 
 
 
 
 
 
 
 
 
 
 
13
  ---
14
 
15
  # SetFit
@@ -39,6 +74,21 @@ The model has been trained using an efficient few-shot learning technique that i
39
  - **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
40
  - **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
41
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
42
  ## Uses
43
 
44
  ### Direct Use for Inference
@@ -57,7 +107,10 @@ from setfit import SetFitModel
57
  # Download from the 🤗 Hub
58
  model = SetFitModel.from_pretrained("setfit_model_id")
59
  # Run inference
60
- preds = model("I loved the spiderman movie!")
 
 
 
61
  ```
62
 
63
  <!--
@@ -86,6 +139,52 @@ preds = model("I loved the spiderman movie!")
86
 
87
  ## Training Details
88
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
89
  ### Framework Versions
90
  - Python: 3.12.12
91
  - SetFit: 1.1.3
 
4
  - sentence-transformers
5
  - text-classification
6
  - generated_from_setfit_trainer
7
+ widget:
8
+ - text: "class TestAuth(unittest.TestCase):\n def setUp(self):\n webhook_secret\
9
+ \ = \"test_abcdef1234567890\""
10
+ - text: "Analyze this vulnerable_pattern: function solveSum(obj) {\n\tif (\"answer\"\
11
+ \ in obj) {\n\t\tdocument.getElementById(\"answer\").innerHTML = obj['answer'];\n\
12
+ \t}\n}"
13
+ - text: "def handle_request(request):\n sql = f\"DELETE FROM sessions WHERE token='{token}'\"\
14
+ \n return result"
15
+ - text: 'Analyze this vulnerable_pattern: }
16
+
17
+
18
+ $create_tb = "CREATE TABLE users (user_id int(6),first_name varchar(15),last_name
19
+ varchar(15), user varchar(15), password varchar(32),avatar varchar(70), last_login
20
+ TIMESTAMP, failed_login INT(3), PRIMARY KEY (user_id));";
21
+
22
+ if( !mysqli_query($GLOBALS["___mysqli_ston"],'
23
+ - text: 'Analyze this clean_code: from __future__ import annotations
24
+
25
+
26
+ import importlib.util
27
+
28
+ import os'
29
  metrics:
30
  - accuracy
31
  pipeline_tag: text-classification
32
  library_name: setfit
33
  inference: true
34
+ model-index:
35
+ - name: SetFit
36
+ results:
37
+ - task:
38
+ type: text-classification
39
+ name: Text Classification
40
+ dataset:
41
+ name: Unknown
42
+ type: unknown
43
+ split: test
44
+ metrics:
45
+ - type: accuracy
46
+ value: 0.9746192893401016
47
+ name: Accuracy
48
  ---
49
 
50
  # SetFit
 
74
  - **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
75
  - **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
76
 
77
+ ### Model Labels
78
+ | Label | Examples |
79
+ |:-----------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
80
+ | TEST_MOCK | <ul><li>"client_secret: str = 'abc123'"</li><li>'# conftest.py\njwt_secret = "MY_API_KEY"'</li><li>"private_key: str = 'yyy'"</li></ul> |
81
+ | REAL_SECRET | <ul><li>'Analyze this hardcoded_secret: aws_access_key_id: AKIAI44QH8DHBEXAMPLE'</li><li>"Analyze this hardcoded_secret: // data: {\n // format: 'json',\n // method: 'flickr.interestingness.getList',\n // api_key: '7617adae70159d09ba78cfec73c13be3'\n // },\n\t // dataType: 'jsonp',\n // jsonp: 'json"</li><li>'Analyze this hardcoded_secret: name: "Blake2b with \'Hello, World!\'",\n\t\t\thasher: NewBlake2B(),\n\t\t\tinput: []byte("Hello, World!"),\n\t\t\texpectedHex: "511bc81dde11180838c562c82bb35f3223f46061ebde4a955c27b3f489cf1e03",\n\t\t},\n\t\t{\n\t\t\tname: "Blake2b input at max size",'</li></ul> |
82
+ | VULNERABLE_LOGIC | <ul><li>'Analyze this vulnerable_pattern: }\n\n$create_db = "CREATE DATABASE {$_DVWA[ \'db_database\' ]};";\nif( !@mysqli_query($GLOBALS["___mysqli_ston"], $create_db ) ) {\n\tdvwaMessagePush( "Could not create database<br />SQL: " . ((is_object($GLOBALS["___mysqli_ston"])) ? mysqli_error($GLOBALS["___mysqli_ston"]) : (($___mysqli_res = mysqli_connect_error()) ? $___mysqli_res : false)) );\n\tdvwaPageReload();\n}'</li><li>'Analyze this vulnerable_pattern: $user = stripslashes( $user );\n\t$user = mysqli_real_escape_string($GLOBALS["___mysqli_ston"], $user);\n\n\t$pass = $_POST[ \'password\' ];\n\t$pass = stripslashes( $pass );\n\t$pass = mysqli_real_escape_string($GLOBALS["___mysqli_ston"], $pass);\n\t$pass = md5( $pass );'</li><li>'Analyze this vulnerable_pattern: case MYSQL:\n\t\t\t// Check database\n\t\t\t$query = "SELECT first_name, last_name FROM users WHERE user_id = \'$id\' LIMIT 1;";\n\t\t\t$result = mysqli_query($GLOBALS["___mysqli_ston"], $query ) or die( \'<pre>Something went wrong.</pre>\' );\n\n\t\t\t// Get results\n\t\t\twhile( $row = mysqli_fetch_assoc( $result ) ) {'</li></ul> |
83
+ | SAFE_CODE | <ul><li>"Analyze this clean_code: [\n 500,\n {\n 'content-type' => content_type,\n 'content-length' => body.bytesize.to_s\n },\n [body]"</li><li>"Analyze this clean_code: it 'should allow switching on the include_subdomains option' do\n mock_app do\n use Rack::Protection::StrictTransport, include_subdomains: true\n run DummyApp\n end"</li><li>"Analyze this clean_code: check(@foo, Rack::Lock, @bar)\n end\n\n it 'works as middleware behind Rack::Lock, with lock disabled' do\n @foo.disable :lock\n check(Rack::Lock, @foo, @bar)\n end"</li></ul> |
84
+
85
+ ## Evaluation
86
+
87
+ ### Metrics
88
+ | Label | Accuracy |
89
+ |:--------|:---------|
90
+ | **all** | 0.9746 |
91
+
92
  ## Uses
93
 
94
  ### Direct Use for Inference
 
107
  # Download from the 🤗 Hub
108
  model = SetFitModel.from_pretrained("setfit_model_id")
109
  # Run inference
110
+ preds = model("Analyze this clean_code: from __future__ import annotations
111
+
112
+ import importlib.util
113
+ import os")
114
  ```
115
 
116
  <!--
 
139
 
140
  ## Training Details
141
 
142
+ ### Training Set Metrics
143
+ | Training set | Min | Median | Max |
144
+ |:-------------|:----|:--------|:----|
145
+ | Word count | 1 | 27.6734 | 157 |
146
+
147
+ | Label | Training Sample Count |
148
+ |:-----------------|:----------------------|
149
+ | REAL_SECRET | 67 |
150
+ | VULNERABLE_LOGIC | 240 |
151
+ | TEST_MOCK | 240 |
152
+ | SAFE_CODE | 240 |
153
+
154
+ ### Training Hyperparameters
155
+ - batch_size: (16, 16)
156
+ - num_epochs: (1, 1)
157
+ - max_steps: -1
158
+ - sampling_strategy: oversampling
159
+ - num_iterations: 5
160
+ - body_learning_rate: (2e-05, 1e-05)
161
+ - head_learning_rate: 0.01
162
+ - loss: CosineSimilarityLoss
163
+ - distance_metric: cosine_distance
164
+ - margin: 0.25
165
+ - end_to_end: False
166
+ - use_amp: False
167
+ - warmup_proportion: 0.1
168
+ - l2_weight: 0.01
169
+ - seed: 42
170
+ - eval_max_steps: -1
171
+ - load_best_model_at_end: True
172
+
173
+ ### Training Results
174
+ | Epoch | Step | Training Loss | Validation Loss |
175
+ |:------:|:----:|:-------------:|:---------------:|
176
+ | 0.0020 | 1 | 0.0639 | - |
177
+ | 0.1016 | 50 | 0.0831 | - |
178
+ | 0.2033 | 100 | 0.0272 | - |
179
+ | 0.3049 | 150 | 0.0108 | - |
180
+ | 0.4065 | 200 | 0.0124 | - |
181
+ | 0.5081 | 250 | 0.0125 | - |
182
+ | 0.6098 | 300 | 0.0069 | - |
183
+ | 0.7114 | 350 | 0.0047 | - |
184
+ | 0.8130 | 400 | 0.0048 | - |
185
+ | 0.9146 | 450 | 0.0096 | - |
186
+ | 1.0 | 492 | - | 0.0246 |
187
+
188
  ### Framework Versions
189
  - Python: 3.12.12
190
  - SetFit: 1.1.3
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3f5e69a66f4af290805d351cbfd5968f48bb82dacf20a059800ca82265db9f6c
3
  size 90864192
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07f14a301e379e2bec9340847b50bb389d2d197df126ee6baa1052364a996c02
3
  size 90864192
model_head.pkl CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:908ee4643be4f4e547ae40fc8efe3dbf933f857655e62a9864a83e5389ea22e8
3
  size 13191
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9405abda38afa737bd06c713716d4f43881c6914f53a12cad272d1dbccfc734a
3
  size 13191