File size: 1,136 Bytes
fec2d9d 961ab91 f433466 961ab91 f433466 961ab91 7b02262 fec2d9d 7b02262 fec2d9d 961ab91 f433466 961ab91 fec2d9d f433466 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 |
from fastapi import FastAPI
from pydantic import BaseModel
from huggingface_hub import InferenceClient
app = FastAPI()
# Use Hugging Face Inference API (Replace model name if needed)
# Charger le modèle et le tokenizer
#model_name = "mistralai/Mistral-7B-Instruct-v0.1" # Modèle Mistral 7B
#model_name = "HuggingFaceH4/zephyr-3b"
#model_name = "serkanarslan/mistral-7b-mini-ft"
# Choose a smaller model for free-tier
#model_name = "microsoft/phi-2"
#model_name = "TinyLlama/TinyLlama-1.1B-Chat-v1.0" # You can switch to Phi-2, OpenChat, etc.
# ✅ Use the full Hugging Face Inference API URL
HF_API_URL = "https://api-inference.huggingface.co/models/TinyLlama/TinyLlama-1.1B-Chat-v1.0"
client = InferenceClient(HF_API_URL)
# Define request format
class ChatRequest(BaseModel):
message: str
@app.post("/chat")
async def chat(request: ChatRequest):
# ✅ Corrected function call with `model` argument
response = client.text_generation(
request.message,
model="TinyLlama/TinyLlama-1.1B-Chat-v1.0",
max_new_tokens=100
)
return {"response": response} # ✅ Removed extra quote
|