Update app.py

app.py

@@ -1,35 +1,4 @@
 import gradio as gr
-import os
-import torch
-from transformers import AutoTokenizer, AutoModelForCausalLM
-
-
-MODEL_ID = os.environ.get("HF_MODEL_ID", "teamaMohamed115/smollm-360m-code-lora")
-DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
-
-
-# Safe loader: try with device_map for HF inference if possible
-print(f"Loading tokenizer and model from {MODEL_ID} on {DEVICE}")
-
-
-tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, use_fast=True)
-# If the model was pushed with custom config (like trusting remote code), we handle gracefully
-try:
-model = AutoModelForCausalLM.from_pretrained(MODEL_ID, trust_remote_code=True)
-except Exception:
-model = AutoModelForCausalLM.from_pretrained(MODEL_ID)
-
-
-model.to(DEVICE)
-model.eval()
-
-
-# Generation helper
-GEN_KWARGS = dict(
-max_new_tokens=256,
-do_sample=True,
-temperature=0.2,
-top_p=0.95,
 top_k=50,
 num_return_sequences=1,
 )
@@ -82,4 +51,35 @@ return decoded.strip()
 
 with gr.Blocks(title="SmolLM Python Code Assistant") as demo:
 gr.Markdown("# SmolLM — Python Code Generation\nEnter an instruction and get a multi-function Python module.")
+
+
+with gr.Row():
+instr = gr.Textbox(lines=6, placeholder="Describe the Python module you want...", label="Instruction")
+with gr.Column(scale=1):
+max_t = gr.Slider(minimum=32, maximum=1024, value=256, step=32, label="Max new tokens")
+temp = gr.Slider(minimum=0.0, maximum=1.0, value=0.2, step=0.05, label="Temperature")
+top_p = gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.01, label="Top-p")
+run_btn = gr.Button("Generate")
+
+
+output = gr.Code(label="Generated Python module", language="python")
+
+
+def run(instruction, max_tokens, temperature, top_p):
+try:
+return generate_code(instruction, max_tokens, temperature, top_p)
+except Exception as e:
+return f"Error during generation: {e}"
+
+
+run_btn.click(run, inputs=[instr, max_t, temp, top_p], outputs=[output])
+
+
+gr.Examples(examples=[
+"Implement a Python module that includes: a function to compute Fibonacci sequence, a function to check primality, and a function to compute factorial, all with type hints and docstrings.",
+"Create a Python module for basic matrix operations (add, multiply, transpose) with appropriate error handling and tests.",
+], inputs=instr)
+
+
+if __name__ == "__main__":
 demo.launch()