fastchat-demo

Getting started

• log into http://jupyterhub.sdsu.edu/ using your sdsu credentials, select your resources (minimal is fine) and select "Stack PRP" image. Here is a guide for reference
• make sure you have open-ai 0.28.1 installed
  • go to the terminal and enter pip list | grep openai
  • if nothing if returned, stop the server and make sure you have selected the "Stack PRP" Image from the list of container images
• git clone this repo into your workspace
• create a env.yaml file
• copy and paste template from env-template.yaml
• paste your API key into API key and the file name under file_name if not using paper.txt

Make Changes

• click into the fastchat-demo.ipynb notebook
• to change the initial prompt for LLM, change initial_prompt
• to change the prompt, make changes to the "process paper.txt" section either by changing the code for pre-processing and/or by using a new txt file
• to change the LLM, change model = models.data[X].id, replacing X with the index of the model you desire to use as shown in print(models)