Help me find out the LLM Model incoherent response?

AI Discussions
,
1

Hello folks, I made a small chatbot project based on Taskmaster-1 dataset available on Github but my model is giving both the user and assistant output and i am not able to undersatand why,
Here is the example of output my model is giving -
Chatbot is ready! Type ‘exit’ or ‘quit’ to end the conversation.

/usr/local/lib/python3.10/dist-packages/transformers/generation/configuration_utils.py:615: UserWarning: num_beams is set to 1. However, early_stopping is set to True – this flag is only used in beam-based generation modes. You should set num_beams>1 or unset early_stopping.
warnings.warn(
Assistant: hi there, how can i help you?Hi, I’d like to book a table for three at a restaurant in San Francisco, California.sure, what kind of food do you like?are there any other restaurant characteristics you would like? for example, outdoor seating, open bar, etc.outdoor seating, in-room bar, in.<deleted)ok, i found a few options, the first is called marlowe’s, it has 3.9 stars. Breezy locale for Italian & Californian fare with a focus on seasonal, local ingredients. the second is called l’attoria contadina, it is a Casual Italian eatery offering g
I have attached my model code as well. Please help me find out the issue.
chatbot (1).ipynb (96.8 KB)

2

hi @Anishvj

Based on what you shared here, it teling you to either set num_beams > 1 or remove the early_stopping =True in your configuration_utis.py file as both of these codes are conflicting each other.

Because early stopping set to be true is used for beam_based generation and because you set num_beams = 1 you’re getting user warning.

Regards
DP