Update README.md
Browse files
README.md
CHANGED
@@ -1,14 +1,6 @@
|
|
1 |
# Model Card for GPT_2_CODE
|
2 |
-
|
3 |
-
<!-- Provide a quick summary of what the model is/does. [Optional] -->
|
4 |
-
WIP,
|
5 |
-
Goal is to create a small GPT2 python coder
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
# Table of Contents
|
11 |
-
|
12 |
- [Model Card for GPT_2_CODE](#model-card-for--model_id-)
|
13 |
- [Table of Contents](#table-of-contents)
|
14 |
- [Table of Contents](#table-of-contents-1)
|
@@ -44,16 +36,9 @@ Goal is to create a small GPT2 python coder
|
|
44 |
- [Model Card Authors [optional]](#model-card-authors-optional)
|
45 |
- [Model Card Contact](#model-card-contact)
|
46 |
- [How to Get Started with the Model](#how-to-get-started-with-the-model)
|
47 |
-
|
48 |
-
|
49 |
# Model Details
|
50 |
-
|
51 |
## Model Description
|
52 |
-
|
53 |
-
<!-- Provide a longer summary of what this model is/does. -->
|
54 |
-
WIP,
|
55 |
-
Goal is to create a small GPT2 python coder
|
56 |
-
|
57 |
- **Developed by:** C, o, d, e, M, o, n, k, e, y
|
58 |
- **Shared by [Optional]:** More information needed
|
59 |
- **Model type:** Language model
|
@@ -63,186 +48,65 @@ Goal is to create a small GPT2 python coder
|
|
63 |
- **Resources for more information:** More information needed
|
64 |
- [GitHub Repo](None)
|
65 |
- [Associated Paper](None)
|
66 |
-
|
67 |
# Uses
|
68 |
-
|
69 |
-
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
70 |
-
|
71 |
## Direct Use
|
72 |
-
|
73 |
-
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
74 |
-
<!-- If the user enters content, print that. If not, but they enter a task in the list, use that. If neither, say "more info needed." -->
|
75 |
-
|
76 |
generate python code snippets
|
77 |
-
|
78 |
-
|
79 |
## Downstream Use [Optional]
|
80 |
-
|
81 |
-
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
82 |
-
<!-- If the user enters content, print that. If not, but they enter a task in the list, use that. If neither, say "more info needed." -->
|
83 |
-
|
84 |
semi auto coder
|
85 |
-
|
86 |
-
|
87 |
## Out-of-Scope Use
|
88 |
-
|
89 |
-
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
90 |
-
<!-- If the user enters content, print that. If not, but they enter a task in the list, use that. If neither, say "more info needed." -->
|
91 |
-
|
92 |
describe code
|
93 |
-
|
94 |
-
|
95 |
-
# Bias, Risks, and Limitations
|
96 |
-
|
97 |
-
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
98 |
-
|
99 |
-
Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). Predictions generated by the model may include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups.
|
100 |
-
|
101 |
-
|
102 |
-
## Recommendations
|
103 |
-
|
104 |
-
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
105 |
-
|
106 |
-
|
107 |
Keep Finetuning on question/python datasets
|
108 |
-
|
109 |
-
|
110 |
# Training Details
|
111 |
|
112 |
## Training Data
|
113 |
-
|
114 |
-
<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
115 |
-
|
116 |
flytech/python-codes-25k
|
117 |
espejelomar/code_search_net_python_10000_examples
|
118 |
-
|
119 |
-
|
120 |
## Training Procedure
|
121 |
-
|
122 |
-
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
123 |
-
|
124 |
### Preprocessing
|
125 |
-
|
126 |
More information needed
|
127 |
-
|
128 |
### Speeds, Sizes, Times
|
129 |
-
|
130 |
-
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
131 |
-
|
132 |
Epochs 3
|
133 |
flytech/python-codes-25k (4600)
|
134 |
Training Loss: 0.4007
|
135 |
Validation Loss: 0.5526
|
136 |
-
|
137 |
Epochs 3
|
138 |
espejelomar/code_search_net_python_10000_examples (4800)
|
139 |
Training Loss: 1.5355
|
140 |
Validation Loss: 1.1723
|
141 |
-
|
142 |
# Evaluation
|
143 |
-
|
144 |
-
<!-- This section describes the evaluation protocols and provides the results. -->
|
145 |
-
|
146 |
-
## Testing Data, Factors & Metrics
|
147 |
-
|
148 |
### Testing Data
|
149 |
-
|
150 |
-
<!-- This should link to a Data Card if possible. -->
|
151 |
-
|
152 |
flytech/python-codes-25k
|
153 |
espejelomar/code_search_net_python_10000_examples
|
154 |
-
|
155 |
-
|
156 |
### Factors
|
157 |
-
|
158 |
-
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
159 |
-
|
160 |
80/20 train/val
|
161 |
-
|
162 |
### Metrics
|
163 |
-
|
164 |
-
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
165 |
-
|
166 |
train/validate
|
167 |
lr scheduling
|
168 |
-
|
169 |
## Results
|
170 |
-
|
171 |
Better in python code generation as base gpt2-medium model
|
172 |
-
|
173 |
# Model Examination
|
174 |
-
|
175 |
More information needed
|
176 |
-
|
177 |
# Environmental Impact
|
178 |
-
|
179 |
-
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
180 |
-
|
181 |
-
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
182 |
-
|
183 |
- **Hardware Type:** CPU and Colab T4
|
184 |
- **Hours used:** 4
|
185 |
- **Cloud Provider:** Google Colab
|
186 |
- **Compute Region:** NL
|
187 |
-
- **Carbon Emitted:** ???
|
188 |
-
|
189 |
-
# Technical Specifications [optional]
|
190 |
-
|
191 |
## Model Architecture and Objective
|
192 |
-
|
193 |
gpt2
|
194 |
-
|
195 |
## Compute Infrastructure
|
196 |
-
|
197 |
More information needed
|
198 |
-
|
199 |
### Hardware
|
200 |
-
|
201 |
CPU and Colab T4
|
202 |
-
|
203 |
### Software
|
204 |
-
|
205 |
pytorch, custom python
|
206 |
-
|
207 |
-
# Citation
|
208 |
-
|
209 |
-
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
210 |
-
|
211 |
-
**BibTeX:**
|
212 |
-
|
213 |
-
More information needed
|
214 |
-
|
215 |
-
**APA:**
|
216 |
-
|
217 |
-
More information needed
|
218 |
-
|
219 |
-
# Glossary [optional]
|
220 |
-
|
221 |
-
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
222 |
-
|
223 |
-
More information needed
|
224 |
-
|
225 |
# More Information [optional]
|
226 |
-
|
227 |
Experimental
|
228 |
-
|
229 |
# Model Card Authors [optional]
|
230 |
-
|
231 |
-
<!-- This section provides another layer of transparency and accountability. Whose views is this model card representing? How many voices were included in its construction? Etc. -->
|
232 |
-
|
233 |
CodeMonkeyXL
|
234 |
-
|
235 |
# Model Card Contact
|
236 |
-
|
237 |
K00B404 huggingface
|
238 |
-
|
239 |
# How to Get Started with the Model
|
240 |
-
|
241 |
Use the code below to get started with the model.
|
242 |
-
|
243 |
-
<details>
|
244 |
-
<summary> Click to expand </summary>
|
245 |
-
|
246 |
-
More information needed
|
247 |
-
|
248 |
-
</details>
|
|
|
1 |
# Model Card for GPT_2_CODE
|
2 |
+
-Goal is to create a small GPT2 python coder
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
# Table of Contents
|
|
|
4 |
- [Model Card for GPT_2_CODE](#model-card-for--model_id-)
|
5 |
- [Table of Contents](#table-of-contents)
|
6 |
- [Table of Contents](#table-of-contents-1)
|
|
|
36 |
- [Model Card Authors [optional]](#model-card-authors-optional)
|
37 |
- [Model Card Contact](#model-card-contact)
|
38 |
- [How to Get Started with the Model](#how-to-get-started-with-the-model)
|
|
|
|
|
39 |
# Model Details
|
|
|
40 |
## Model Description
|
41 |
+
WIP,Goal is to create a small GPT2 python coder
|
|
|
|
|
|
|
|
|
42 |
- **Developed by:** C, o, d, e, M, o, n, k, e, y
|
43 |
- **Shared by [Optional]:** More information needed
|
44 |
- **Model type:** Language model
|
|
|
48 |
- **Resources for more information:** More information needed
|
49 |
- [GitHub Repo](None)
|
50 |
- [Associated Paper](None)
|
|
|
51 |
# Uses
|
52 |
+
coding assistant
|
|
|
|
|
53 |
## Direct Use
|
|
|
|
|
|
|
|
|
54 |
generate python code snippets
|
|
|
|
|
55 |
## Downstream Use [Optional]
|
|
|
|
|
|
|
|
|
56 |
semi auto coder
|
|
|
|
|
57 |
## Out-of-Scope Use
|
|
|
|
|
|
|
|
|
58 |
describe code
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
59 |
Keep Finetuning on question/python datasets
|
|
|
|
|
60 |
# Training Details
|
61 |
|
62 |
## Training Data
|
|
|
|
|
|
|
63 |
flytech/python-codes-25k
|
64 |
espejelomar/code_search_net_python_10000_examples
|
|
|
|
|
65 |
## Training Procedure
|
66 |
+
Train/Val/Scheduler
|
|
|
|
|
67 |
### Preprocessing
|
|
|
68 |
More information needed
|
|
|
69 |
### Speeds, Sizes, Times
|
|
|
|
|
|
|
70 |
Epochs 3
|
71 |
flytech/python-codes-25k (4600)
|
72 |
Training Loss: 0.4007
|
73 |
Validation Loss: 0.5526
|
|
|
74 |
Epochs 3
|
75 |
espejelomar/code_search_net_python_10000_examples (4800)
|
76 |
Training Loss: 1.5355
|
77 |
Validation Loss: 1.1723
|
|
|
78 |
# Evaluation
|
79 |
+
Manual comparison with base model
|
|
|
|
|
|
|
|
|
80 |
### Testing Data
|
|
|
|
|
|
|
81 |
flytech/python-codes-25k
|
82 |
espejelomar/code_search_net_python_10000_examples
|
|
|
|
|
83 |
### Factors
|
|
|
|
|
|
|
84 |
80/20 train/val
|
|
|
85 |
### Metrics
|
|
|
|
|
|
|
86 |
train/validate
|
87 |
lr scheduling
|
|
|
88 |
## Results
|
|
|
89 |
Better in python code generation as base gpt2-medium model
|
|
|
90 |
# Model Examination
|
|
|
91 |
More information needed
|
|
|
92 |
# Environmental Impact
|
|
|
|
|
|
|
|
|
|
|
93 |
- **Hardware Type:** CPU and Colab T4
|
94 |
- **Hours used:** 4
|
95 |
- **Cloud Provider:** Google Colab
|
96 |
- **Compute Region:** NL
|
|
|
|
|
|
|
|
|
97 |
## Model Architecture and Objective
|
|
|
98 |
gpt2
|
|
|
99 |
## Compute Infrastructure
|
|
|
100 |
More information needed
|
|
|
101 |
### Hardware
|
|
|
102 |
CPU and Colab T4
|
|
|
103 |
### Software
|
|
|
104 |
pytorch, custom python
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
105 |
# More Information [optional]
|
|
|
106 |
Experimental
|
|
|
107 |
# Model Card Authors [optional]
|
|
|
|
|
|
|
108 |
CodeMonkeyXL
|
|
|
109 |
# Model Card Contact
|
|
|
110 |
K00B404 huggingface
|
|
|
111 |
# How to Get Started with the Model
|
|
|
112 |
Use the code below to get started with the model.
|
|
|
|
|
|
|
|
|
|
|
|
|
|