-

app.py

@@ -24,23 +24,26 @@ allowed_keywords = [
 def is_relevant_question(user_input):
     return any(keyword in user_input for keyword in allowed_keywords)
 
-# ===== 2) تحميل النموذج =====
 model_name = "ALLaM-AI/ALLaM-7B-Instruct-preview"
+
+device = "cuda" if torch.cuda.is_available() else "cpu"
+dtype = torch.float16 if device == "cuda" else torch.float32
+
 tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=False, trust_remote_code=True)
 model = AutoModelForCausalLM.from_pretrained(
     model_name,
     trust_remote_code=True,
-    torch_dtype=torch.float16,
+    torch_dtype=dtype,
     device_map="auto"
 )
 
-# ===== 3) توليد الرد =====
 def chatbot_response(user_input):
     if not is_relevant_question(user_input):
         return "أنا متخصص فقط في النخيل والتمور. يرجى طرح أسئلة ذات صلة."
 
     prompt = f"سؤال: {user_input}\nجواب:"
-    inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
+    inputs = tokenizer(prompt, return_tensors="pt")
+    inputs = {k: v.to(model.device) for k, v in inputs.items()}
 
     outputs = model.generate(
         **inputs,
@@ -53,7 +56,6 @@ def chatbot_response(user_input):
     response = tokenizer.decode(outputs[0], skip_special_tokens=True)
     return response.replace(prompt, "").strip()
 
-# ===== 4) واجهة Gradio =====
 iface = gr.Interface(
     fn=chatbot_response,
     inputs=gr.Textbox(lines=3, placeholder="اكتب سؤالك هنا..."),