Set pad_token to eos_token and exclude user query from response another fix

app.py

@@ -43,7 +43,7 @@ def chat():
     inputs = tokenizer(user_input, return_tensors='pt', padding=True, truncation=True)
     input_ids = inputs['input_ids']
     attention_mask = inputs['attention_mask']
-    outputs = model.generate(input_ids, attention_mask=attention_mask)
+    outputs = model.generate(input_ids, attention_mask=attention_mask, max_new_tokens=100,pad_token_id=tokenizer.eos_token_id)
     response_text = tokenizer.decode(outputs[0], skip_special_tokens=True).replace(user_input, '').strip()
     
     return jsonify({"response": response_text})