Simplify for Gradio 6.x compatibility

app.py

@@ -1,7 +1,6 @@
 """
 Stack 2.9 - HuggingFace Space
-Minimal Gradio UI for code generation
-Pinned to Gradio 4.12.0 for compatibility
+Compatible with Gradio 6.x
 """
 import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer
@@ -9,26 +8,19 @@ import torch
 
 print("Loading model...")
 
-# Load model on CPU to fit free tier
 MODEL_NAME = "Qwen/Qwen2.5-Coder-1.5B-Instruct"
 tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME, trust_remote_code=True)
 model = AutoModelForCausalLM.from_pretrained(
     MODEL_NAME,
-    torch_dtype=torch.float32,  # CPU-friendly
+    torch_dtype=torch.float32,
     device_map="cpu",
     trust_remote_code=True,
     low_cpu_mem_usage=True
 )
 print("Model loaded!")
 
-def generate_code(prompt, system_prompt="You are a helpful coding assistant.", max_tokens=256, temperature=0.7):
-    """Generate code from prompt"""
-    messages = [
-        {"role": "user", "content": prompt}
-    ]
-    if system_prompt:
-        messages.insert(0, {"role": "system", "content": system_prompt})
-    
+def generate(prompt, max_tokens=256, temperature=0.7):
+    messages = [{"role": "user", "content": prompt}]
     text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
     inputs = tokenizer([text], return_tensors="pt")
     
@@ -44,45 +36,22 @@ def generate_code(prompt, system_prompt="You are a helpful coding assistant.", m
     response = tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True)
     return response.strip()
 
-# Simple UI
-with gr.Blocks(title="Stack 2.9", theme=gr.themes.Soft()) as demo:
-    gr.Markdown("# 💻 Stack 2.9 Code Assistant")
-    gr.Markdown(f"Running: `{MODEL_NAME}` · 1.5B parameters · CPU")
-    
-    with gr.Row():
-        with gr.Column():
-            prompt = gr.Textbox(
-                label="Your prompt",
-                placeholder="Write a Python function to calculate fibonacci...",
-                lines=6
-            )
-            system = gr.Textbox(
-                label="System prompt",
-                value="You are a helpful coding assistant.",
-                lines=2
-            )
-            with gr.Row():
-                max_tokens = gr.Slider(64, 512, value=256, label="Max tokens")
-                temp = gr.Slider(0.1, 1.0, value=0.7, label="Temperature")
-            btn = gr.Button("Generate", variant="primary")
-        
-        with gr.Column():
-            output = gr.Code(label="Generated code", language="python", lines=20)
-            clear = gr.Button("Clear")
-    
-    gr.Examples(
-        examples=[
-            ["Write a Python function to calculate fibonacci numbers"],
-            ["Explain what this code does: def foo(x): return x * 2"],
-            ["Write a SQL query to find duplicate emails"],
-            ["How do I handle exceptions in Python?"],
-        ],
-        inputs=prompt
-    )
-    
-    btn.click(generate_code, [prompt, system, max_tokens, temp], output)
-    prompt.submit(generate_code, [prompt, system, max_tokens, temp], output)
-    clear.click(lambda: "", None, prompt)
+demo = gr.Interface(
+    fn=generate,
+    inputs=[
+        gr.Textbox(label="Prompt", placeholder="Write a Python function to calculate fibonacci...", lines=6),
+        gr.Slider(64, 512, value=256, label="Max tokens"),
+        gr.Slider(0.1, 1.0, value=0.7, label="Temperature"),
+    ],
+    outputs=gr.Textbox(label="Response", lines=10),
+    title="Stack 2.9 Code Assistant",
+    description="Powered by Qwen2.5-Coder-1.5B",
+    examples=[
+        ["Write a Python function to calculate fibonacci numbers"],
+        ["Explain what this code does: def foo(x): return x * 2"],
+        ["Write a SQL query to find duplicate emails"],
+        ["How do I handle exceptions in Python?"],
+    ]
+)
 
-if __name__ == "__main__":
-    demo.queue().launch(server_name="0.0.0.0", server_port=7860)
+demo.launch(server_name="0.0.0.0", server_port=7860)

requirements.txt

@@ -1,5 +1,5 @@
-gradio==4.12.0
-transformers==4.40.0
-torch==2.1.2
-accelerate==0.25.0
-sentencepiece==0.2.0
+gradio>=4.0.0
+transformers>=4.40.0
+torch>=2.0.0
+accelerate>=0.25.0
+sentencepiece>=0.2.0