Spaces:
Running
on
L4
Running
on
L4
updated readme template
Browse files
app.py
CHANGED
@@ -104,6 +104,67 @@ SUPPORTED_MODEL_TYPES = ['llama',
|
|
104 |
'cohere',
|
105 |
'minicpm']
|
106 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
107 |
|
108 |
def button_click(hf_model_id, conv_template, quantization, oauth_token: gr.OAuthToken | None, progress=gr.Progress()):
|
109 |
if oauth_token.token is None:
|
@@ -180,59 +241,11 @@ def button_click(hf_model_id, conv_template, quantization, oauth_token: gr.OAuth
|
|
180 |
card.data.tags.append("MLC-Weight-Conversion")
|
181 |
card.data.base_model = hf_model_id
|
182 |
|
183 |
-
card.text =
|
184 |
-
f""
|
185 |
-
|
186 |
-
|
187 |
-
|
188 |
-
The conversion was done using the [MLC-Weight-Conversion](https://huggingface.co/spaces/mlc-ai/MLC-Weight-Conversion) space.
|
189 |
-
The model can be used for projects [MLC-LLM](https://github.com/mlc-ai/mlc-llm).
|
190 |
-
|
191 |
-
## Example Usage
|
192 |
-
|
193 |
-
Here are some examples of using this model in MLC LLM.
|
194 |
-
Before running the examples, please install MLC LLM by following the [installation documentation](https://llm.mlc.ai/docs/install/mlc_llm.html#install-mlc-packages).
|
195 |
-
|
196 |
-
### Chat
|
197 |
-
|
198 |
-
In command line, run
|
199 |
-
```bash
|
200 |
-
mlc_llm chat HF://{created_repo_id}
|
201 |
-
```
|
202 |
-
|
203 |
-
### REST Server
|
204 |
-
|
205 |
-
In command line, run
|
206 |
-
```bash
|
207 |
-
mlc_llm serve HF://{created_repo_id}
|
208 |
-
```
|
209 |
-
|
210 |
-
### Python API
|
211 |
-
|
212 |
-
```python
|
213 |
-
from mlc_llm import MLCEngine
|
214 |
-
|
215 |
-
# Create engine
|
216 |
-
model = "HF://{created_repo_id}"
|
217 |
-
engine = MLCEngine(model)
|
218 |
-
|
219 |
-
# Run chat completion in OpenAI API.
|
220 |
-
for response in engine.chat.completions.create(
|
221 |
-
messages=[{"role": "user", "content": "What is the meaning of life?"}],
|
222 |
-
model=model,
|
223 |
-
stream=True,
|
224 |
-
):
|
225 |
-
for choice in response.choices:
|
226 |
-
print(choice.delta.content, end="", flush=True)
|
227 |
-
print("\n")
|
228 |
-
|
229 |
-
engine.terminate()
|
230 |
-
```
|
231 |
-
|
232 |
-
## Documentation
|
233 |
-
|
234 |
-
For more information on MLC LLM project, please visit our [documentation](https://llm.mlc.ai/docs/) and [GitHub repo](http://github.com/mlc-ai/mlc-llm).
|
235 |
-
"""
|
236 |
)
|
237 |
card.save("./dist/README.md")
|
238 |
|
|
|
104 |
'cohere',
|
105 |
'minicpm']
|
106 |
|
107 |
+
readme_template = """
|
108 |
+
---
|
109 |
+
library_name: mlc-llm
|
110 |
+
base_model: {base_model}
|
111 |
+
tags:
|
112 |
+
- mlc-llm
|
113 |
+
- web-llm
|
114 |
+
---
|
115 |
+
|
116 |
+
# {model_name}
|
117 |
+
|
118 |
+
This is the [{base_model_name}](https://huggingface.co/{base_model}) model in MLC format `{quant_format}`.
|
119 |
+
The conversion was done using the [MLC-Weight-Conversion](https://huggingface.co/spaces/mlc-ai/MLC-Weight-Conversion) space.
|
120 |
+
The model can be used for projects [MLC-LLM](https://github.com/mlc-ai/mlc-llm) and [WebLLM](https://github.com/mlc-ai/web-llm).
|
121 |
+
|
122 |
+
## Example Usage
|
123 |
+
|
124 |
+
Here are some examples of using this model in MLC LLM.
|
125 |
+
Before running the examples, please install MLC LLM by following the [installation documentation](https://llm.mlc.ai/docs/install/mlc_llm.html#install-mlc-packages).
|
126 |
+
|
127 |
+
### Chat
|
128 |
+
|
129 |
+
In command line, run
|
130 |
+
```bash
|
131 |
+
mlc_llm chat HF://mlc-ai/{model_name}
|
132 |
+
```
|
133 |
+
|
134 |
+
### REST Server
|
135 |
+
|
136 |
+
In command line, run
|
137 |
+
```bash
|
138 |
+
mlc_llm serve HF://mlc-ai/{model_name}
|
139 |
+
```
|
140 |
+
|
141 |
+
### Python API
|
142 |
+
|
143 |
+
```python
|
144 |
+
from mlc_llm import MLCEngine
|
145 |
+
|
146 |
+
# Create engine
|
147 |
+
model = "HF://mlc-ai/{model_name}"
|
148 |
+
engine = MLCEngine(model)
|
149 |
+
|
150 |
+
# Run chat completion in OpenAI API.
|
151 |
+
for response in engine.chat.completions.create(
|
152 |
+
messages=[{{"role": "user", "content": "What is the meaning of life?"}}],
|
153 |
+
model=model,
|
154 |
+
stream=True,
|
155 |
+
):
|
156 |
+
for choice in response.choices:
|
157 |
+
print(choice.delta.content, end="", flush=True)
|
158 |
+
print("\\n")
|
159 |
+
|
160 |
+
engine.terminate()
|
161 |
+
```
|
162 |
+
|
163 |
+
## Documentation
|
164 |
+
|
165 |
+
For more information on MLC LLM project, please visit our [documentation](https://llm.mlc.ai/docs/) and [GitHub repo](http://github.com/mlc-ai/mlc-llm).
|
166 |
+
""".strip()
|
167 |
+
|
168 |
|
169 |
def button_click(hf_model_id, conv_template, quantization, oauth_token: gr.OAuthToken | None, progress=gr.Progress()):
|
170 |
if oauth_token.token is None:
|
|
|
241 |
card.data.tags.append("MLC-Weight-Conversion")
|
242 |
card.data.base_model = hf_model_id
|
243 |
|
244 |
+
card.text = readme_template.format(
|
245 |
+
model_name=f"{user_name}/{mlc_model_name}",
|
246 |
+
base_model=hf_model_id,
|
247 |
+
base_model_name=model_dir_name,
|
248 |
+
quant_format=quantization,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
249 |
)
|
250 |
card.save("./dist/README.md")
|
251 |
|