Update app.py

app.py

@@ -1,36 +1,51 @@
 import gradio as gr
 import torch
 from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
-import traceback
 
-# 👈 استخدم موديلك انت
+# =========================
+# CONFIG
+# =========================
 MODEL_NAME = "kawkabelaloom/astramindx"
-
 SYSTEM_PROMPT = "أنت مساعد عربي ذكي، تجيب بوضوح وبأسلوب بسيط ومفيد."
 
+# =========================
+# LOAD TOKENIZER
+# =========================
 print("🔄 Loading tokenizer...")
 tokenizer = AutoTokenizer.from_pretrained(
     MODEL_NAME,
     trust_remote_code=True
 )
 
-print("🔄 Loading model (CPU)...")
+# =========================
+# LOAD MODEL (FORCE CPU – NO QUANTIZATION)
+# =========================
+print("🔄 Loading model (CPU, no quantization)...")
+
 model = AutoModelForCausalLM.from_pretrained(
     MODEL_NAME,
-    device_map="cpu",
-    torch_dtype=torch.float32,
-    low_cpu_mem_usage=True,
-    trust_remote_code=True
+    device_map=None,                # ❌ لا device_map auto
+    torch_dtype=torch.float32,       # CPU safe
+    low_cpu_mem_usage=False,         # مهم
+    trust_remote_code=True,
+    quantization_config=None         # 🔥 الحل الأساسي
 )
 
-print("✅ Model loaded")
+model.eval()
+print("✅ Model loaded successfully")
 
+# =========================
+# PIPELINE
+# =========================
 generator = pipeline(
     "text-generation",
     model=model,
     tokenizer=tokenizer
 )
 
+# =========================
+# CHAT LOGIC
+# =========================
 def build_prompt(history, user_message):
     prompt = SYSTEM_PROMPT + "\n\n"
     for user, bot in history:
@@ -38,13 +53,14 @@ def build_prompt(history, user_message):
     prompt += f"المستخدم: {user_message}\nالمساعد:"
     return prompt
 
+
 def chat(user_message, history):
     try:
         prompt = build_prompt(history, user_message)
 
         output = generator(
             prompt,
-            max_new_tokens=128,  # 👈 قللها للسرعة
+            max_new_tokens=128,
             temperature=0.7,
             top_p=0.9,
             do_sample=True
@@ -55,13 +71,16 @@ def chat(user_message, history):
         return history, ""
 
     except Exception as e:
-        history.append((user_message, str(e)))
+        history.append((user_message, f"❌ Error: {str(e)}"))
         return history, ""
 
+# =========================
+# GRADIO UI
+# =========================
 with gr.Blocks() as demo:
     gr.Markdown("# 🤖 Astramindx Chatbot")
     chatbot = gr.Chatbot(height=450)
-    msg = gr.Textbox()
+    msg = gr.Textbox(placeholder="اكتب سؤالك هنا...")
     msg.submit(chat, [msg, chatbot], [chatbot, msg])
 
 demo.launch()