Indus 3.0 Demo

app.py

@@ -42,13 +42,21 @@ def chat_fn(prompt, history):
     history = history or []
     history.append({"role": "user", "content": prompt})
 
+    # Tokenize and send to device
     inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+    input_length = inputs["input_ids"].shape[1]
+
+    # Generate full-sequence tokens (prompt + continuation)
     outputs = model.generate(**inputs, max_new_tokens=100)
-    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+    # Remove the prompt tokens from the start
+    generated_ids = outputs[0][input_length:]
+    response = tokenizer.decode(generated_ids, skip_special_tokens=True)
 
     history.append({"role": "assistant", "content": response})
     return history, ""
 
+
 # 6. Build Gradio interface
 with gr.Blocks() as demo:
     gr.Markdown("## 🌐 Indus 3.0 Hindi LLM Demo")