Spaces:
Running
Running
Update project documentation and dependencies to English, modify HuggingFace model initialization
Browse files- README.md +33 -33
- app.py +1 -1
- requirements.txt +3 -2
README.md
CHANGED
@@ -11,83 +11,83 @@ short_description: Backend of SmolLM2 chat
|
|
11 |
|
12 |
# SmolLM2 Backend
|
13 |
|
14 |
-
|
15 |
|
16 |
-
##
|
17 |
|
18 |
-
###
|
19 |
|
20 |
-
|
21 |
|
22 |
-
1.
|
23 |
-
2.
|
24 |
-
-
|
25 |
-
-
|
26 |
-
-
|
27 |
-
-
|
28 |
|
29 |
-
###
|
30 |
|
31 |
-
|
32 |
|
33 |
-
1.
|
34 |
-
2.
|
35 |
```
|
36 |
-
HUGGINGFACE_TOKEN=
|
37 |
```
|
38 |
-
3.
|
39 |
```
|
40 |
pip install -r requirements.txt
|
41 |
```
|
42 |
|
43 |
-
##
|
44 |
|
45 |
```bash
|
46 |
uvicorn app:app --reload
|
47 |
```
|
48 |
|
49 |
-
|
50 |
|
51 |
## Endpoints
|
52 |
|
53 |
-
### GET
|
54 |
|
55 |
-
|
56 |
|
57 |
-
### POST
|
58 |
|
59 |
-
Endpoint
|
60 |
|
61 |
-
**
|
62 |
```json
|
63 |
{
|
64 |
-
"query": "
|
65 |
-
"thread_id": "
|
66 |
}
|
67 |
```
|
68 |
|
69 |
-
**
|
70 |
```json
|
71 |
{
|
72 |
-
"generated_text": "
|
73 |
-
"thread_id": "
|
74 |
}
|
75 |
```
|
76 |
|
77 |
## Docker
|
78 |
|
79 |
-
|
80 |
|
81 |
```bash
|
82 |
-
#
|
83 |
docker build -t smollm2-backend .
|
84 |
|
85 |
-
#
|
86 |
docker run -p 8000:8000 --env-file .env smollm2-backend
|
87 |
```
|
88 |
|
89 |
-
##
|
90 |
|
91 |
-
|
92 |
- Swagger UI: `http://localhost:8000/docs`
|
93 |
- ReDoc: `http://localhost:8000/redoc`
|
|
|
11 |
|
12 |
# SmolLM2 Backend
|
13 |
|
14 |
+
This project implements a FastAPI API that uses LangChain and LangGraph to generate text with the Qwen2.5-72B-Instruct model from HuggingFace.
|
15 |
|
16 |
+
## Configuration
|
17 |
|
18 |
+
### In HuggingFace Spaces
|
19 |
|
20 |
+
This project is designed to run in HuggingFace Spaces. To configure it:
|
21 |
|
22 |
+
1. Create a new Space in HuggingFace with SDK Docker
|
23 |
+
2. Configure the `HUGGINGFACE_TOKEN` or `HF_TOKEN` environment variable in the Space configuration:
|
24 |
+
- Go to the "Settings" tab of your Space
|
25 |
+
- Scroll down to the "Repository secrets" section
|
26 |
+
- Add a new variable with the name `HUGGINGFACE_TOKEN` and your token as the value
|
27 |
+
- Save the changes
|
28 |
|
29 |
+
### Local development
|
30 |
|
31 |
+
For local development:
|
32 |
|
33 |
+
1. Clone this repository
|
34 |
+
2. Create a `.env` file in the project root with your HuggingFace token:
|
35 |
```
|
36 |
+
HUGGINGFACE_TOKEN=your_token_here
|
37 |
```
|
38 |
+
3. Install the dependencies:
|
39 |
```
|
40 |
pip install -r requirements.txt
|
41 |
```
|
42 |
|
43 |
+
## Local execution
|
44 |
|
45 |
```bash
|
46 |
uvicorn app:app --reload
|
47 |
```
|
48 |
|
49 |
+
The API will be available at `http://localhost:8000`.
|
50 |
|
51 |
## Endpoints
|
52 |
|
53 |
+
### GET `/`
|
54 |
|
55 |
+
Welcome endpoint that returns a greeting message.
|
56 |
|
57 |
+
### POST `/generate`
|
58 |
|
59 |
+
Endpoint to generate text using the language model.
|
60 |
|
61 |
+
**Request parameters:**
|
62 |
```json
|
63 |
{
|
64 |
+
"query": "Your question here",
|
65 |
+
"thread_id": "optional_thread_identifier"
|
66 |
}
|
67 |
```
|
68 |
|
69 |
+
**Response:**
|
70 |
```json
|
71 |
{
|
72 |
+
"generated_text": "Generated text by the model",
|
73 |
+
"thread_id": "thread identifier"
|
74 |
}
|
75 |
```
|
76 |
|
77 |
## Docker
|
78 |
|
79 |
+
To run the application in a Docker container:
|
80 |
|
81 |
```bash
|
82 |
+
# Build the image
|
83 |
docker build -t smollm2-backend .
|
84 |
|
85 |
+
# Run the container
|
86 |
docker run -p 8000:8000 --env-file .env smollm2-backend
|
87 |
```
|
88 |
|
89 |
+
## API documentation
|
90 |
|
91 |
+
The interactive API documentation is available at:
|
92 |
- Swagger UI: `http://localhost:8000/docs`
|
93 |
- ReDoc: `http://localhost:8000/redoc`
|
app.py
CHANGED
@@ -21,7 +21,7 @@ if not HUGGINGFACE_TOKEN:
|
|
21 |
|
22 |
# Inicializar el modelo
|
23 |
model = HuggingFaceEndpoint(
|
24 |
-
|
25 |
huggingfacehub_api_token=HUGGINGFACE_TOKEN,
|
26 |
max_new_tokens=64,
|
27 |
temperature=0.5,
|
|
|
21 |
|
22 |
# Inicializar el modelo
|
23 |
model = HuggingFaceEndpoint(
|
24 |
+
repo_id="Qwen/Qwen2.5-72B-Instruct",
|
25 |
huggingfacehub_api_token=HUGGINGFACE_TOKEN,
|
26 |
max_new_tokens=64,
|
27 |
temperature=0.5,
|
requirements.txt
CHANGED
@@ -1,10 +1,11 @@
|
|
1 |
-
fastapi
|
2 |
uvicorn
|
3 |
requests
|
4 |
pydantic>=2.0.0
|
5 |
langchain
|
6 |
langchain-huggingface
|
7 |
-
|
|
|
8 |
python-dotenv
|
9 |
clarifai
|
10 |
Pillow
|
|
|
1 |
+
fastapi
|
2 |
uvicorn
|
3 |
requests
|
4 |
pydantic>=2.0.0
|
5 |
langchain
|
6 |
langchain-huggingface
|
7 |
+
langchain-core
|
8 |
+
langgraph > 0.2.27
|
9 |
python-dotenv
|
10 |
clarifai
|
11 |
Pillow
|