Update app.py

app.py

@@ -1,11 +1,7 @@
-from fastapi import FastAPI, HTTPException
-from pydantic import BaseModel
-import torch
+import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
 
-app = FastAPI(title="Qwen API Server")
-
-# استخدام نسخة 1.5B لتجنب مشاكل الذاكرة والانهيار على الـ CPU
 MODEL_NAME = "Qwen/Qwen2.5-1.5B-Instruct"
 
 print(f"Loading Model: {MODEL_NAME} ...")
@@ -17,59 +13,43 @@ model = AutoModelForCausalLM.from_pretrained(
 )
 print("Model successfully loaded on CPU!")
 
-class ChatMessage(BaseModel):
-    role: str
-    content: str
-
-class ChatRequest(BaseModel):
-    messages: list[ChatMessage]
-    max_tokens: int = 512
-    temperature: float = 0.7
-
-@app.get("/")
-def home():
-    return {"status": "Running", "model": MODEL_NAME}
-
-@app.post("/v1/chat/completions")
-async def chat_completion(request: ChatRequest):
-    try:
-        formatted_messages = [{"role": m.role, "content": m.content} for m in request.messages]
-        
-        text = tokenizer.apply_chat_template(
-            formatted_messages,
-            tokenize=False,
-            add_generation_prompt=True
+def generate_response(message, history):
+    # تجهيز الرسائل بالصيغة الصحيحة
+    messages = []
+    for human, assistant in history:
+        messages.append({"role": "user", "content": human})
+        messages.append({"role": "assistant", "content": assistant})
+    messages.append({"role": "user", "content": message})
+    
+    text = tokenizer.apply_chat_template(
+        messages,
+        tokenize=False,
+        add_generation_prompt=True
+    )
+    
+    inputs = tokenizer([text], return_tensors="pt")
+    
+    with torch.no_grad():
+        outputs = model.generate(
+            **inputs,
+            max_new_tokens=512,
+            temperature=0.7,
+            do_sample=True,
+            top_p=0.9
         )
-        
-        inputs = tokenizer([text], return_tensors="pt")
-        
-        with torch.no_grad():
-            outputs = model.generate(
-                **inputs,
-                max_new_tokens=request.max_tokens,
-                temperature=request.temperature,
-                do_sample=True,
-                top_p=0.9
-            )
-        
-        generated_ids = [
-            output[len(input_ids):] for input_ids, output in zip(inputs.input_ids, outputs)
-        ]
-        response_text = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
-        
-        return {
-            "choices": [
-                {
-                    "message": {
-                        "role": "assistant",
-                        "content": response_text
-                    }
-                }
-            ]
-        }
-    except Exception as e:
-        raise HTTPException(status_code=500, detail=str(e))
+    
+    generated_ids = [
+        output[len(input_ids):] for input_ids, output in zip(inputs.input_ids, outputs)
+    ]
+    response_text = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
+    return response_text
+
+# إنشاء واجهة Gradio التي تدعم الـ UI وفي نفس الوقت تتيح API جاهز لأي منصة خارجية
+demo = gr.ChatInterface(
+    fn=generate_response,
+    title="Qwen 2.5 1.5B API & Chat",
+    description="هذا السيرفر يعمل بشكل دائم على Hugging Face ويمكنك التحدث معه هنا أو ربطه بأي تطبيق خارجي عبر الـ API."
+)
 
 if __name__ == "__main__":
-    import uvicorn
-    uvicorn.run(app, host="0.0.0.0", port=7860)
+    demo.launch(server_name="0.0.0.0", server_port=7860, share=False)