Update app.py

app.py

@@ -1,8 +1,46 @@
 import os
+import gradio as gr
+import torch
 from transformers import AutoTokenizer, AutoModelForCausalLM
 
-model_id = "your-username/my-wizardmath-finetuned"
-hf_token = os.getenv("HF_TOKEN")
+# Replace this with your actual model repo ID on Hugging Face
+MODEL_ID = "your-username/wizardmath-finetuned"
+
+# If your model is private, provide your Hugging Face token via environment variable
+HF_TOKEN = os.getenv("HF_TOKEN", None)
+
+# Load tokenizer and model (use token if private)
+tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, use_auth_token=HF_TOKEN)
+model = AutoModelForCausalLM.from_pretrained(MODEL_ID, use_auth_token=HF_TOKEN,
+                                             torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32,
+                                             device_map="auto")
+
+def generate_answer(prompt, max_new_tokens=256, temperature=0.7, top_p=0.9):
+    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+    with torch.no_grad():
+        outputs = model.generate(
+            **inputs,
+            max_new_tokens=max_new_tokens,
+            temperature=temperature,
+            top_p=top_p,
+            do_sample=True,
+            pad_token_id=tokenizer.eos_token_id
+        )
+    return tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+iface = gr.Interface(
+    fn=generate_answer,
+    inputs=[
+        gr.Textbox(lines=4, label="Prompt"),
+        gr.Slider(50, 1024, value=256, step=1, label="Max new tokens"),
+        gr.Slider(0.1, 1.0, value=0.7, step=0.05, label="Temperature"),
+        gr.Slider(0.1, 1.0, value=0.9, step=0.05, label="Top-p"),
+    ],
+    outputs=gr.Textbox(label="Response"),
+    title="WizardMath Fine-Tuned Model",
+    description="Ask math questions to your fine-tuned WizardMath model!"
+)
+
+if __name__ == "__main__":
+    iface.launch()
 
-tokenizer = AutoTokenizer.from_pretrained(model_id, token=hf_token)
-model = AutoModelForCausalLM.from_pretrained(model_id, token=hf_token)