Auto-deploy: 2025-10-15 16:07:02

app.py

@@ -2,6 +2,8 @@ from fastapi import FastAPI, HTTPException
 from fastapi.responses import HTMLResponse, JSONResponse
 from pydantic import BaseModel
 import os
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
 
 app = FastAPI(
     title="Sheikh LLM API",
@@ -9,6 +11,10 @@ app = FastAPI(
     version="1.0.0"
 )
 
+# Load model and tokenizer
+tokenizer = AutoTokenizer.from_pretrained("microsoft/DialoGPT-medium")
+model = AutoModelForCausalLM.from_pretrained("microsoft/DialoGPT-medium")
+
 class ChatRequest(BaseModel):
     message: str
     max_tokens: int = 100
@@ -107,11 +113,18 @@ async def api_status():
 
 @app.post("/api/chat", response_model=ChatResponse)
 async def chat_endpoint(request: ChatRequest):
-    """Simple chat endpoint that echoes the message"""
+    """Chat endpoint that uses a Hugging Face model"""
     if not request.message.strip():
         raise HTTPException(status_code=400, detail="Message cannot be empty")
-    
-    response_text = f"Received your message: '{request.message}'. This is from Sheikh LLM API!"
+
+    # Encode the new user input, add the eos_token and return a tensor in Pytorch
+    new_user_input_ids = tokenizer.encode(request.message + tokenizer.eos_token, return_tensors='pt')
+
+    # Generate a response
+    chat_history_ids = model.generate(new_user_input_ids, max_length=request.max_tokens, pad_token_id=tokenizer.eos_token_id)
+
+    # Decode the response
+    response_text = tokenizer.decode(chat_history_ids[:, new_user_input_ids.shape[-1]:][0], skip_special_tokens=True)
     
     return ChatResponse(
         response=response_text,
@@ -120,4 +133,4 @@ async def chat_endpoint(request: ChatRequest):
 
 if __name__ == "__main__":
     import uvicorn
-    uvicorn.run(app, host="0.0.0.0", port=7860)
+    uvicorn.run(app, host="0.0.0.0", port=7860)

requirements.txt

@@ -2,3 +2,5 @@ fastapi==0.104.1
 uvicorn[standard]==0.24.0
 pydantic==2.5.0
 python-multipart==0.0.6
+transformers
+torch