https://github.com/hrishioa/tough-llm-tests

Some tough questions to test new models.
https://github.com/hrishioa/tough-llm-tests

Last synced: about 1 month ago
JSON representation

Some tough questions to test new models.

• Host: GitHub
• URL: https://github.com/hrishioa/tough-llm-tests
• Owner: hrishioa
• Created: 2024-04-20T07:33:22.000Z (9 months ago)
• Default Branch: master
• Last Pushed: 2024-04-20T07:33:38.000Z (9 months ago)
• Last Synced: 2024-10-15T03:55:27.037Z (3 months ago)
• Size: 82 KB
• Stars: 25
• Watchers: 4
• Forks: 3
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome_ai_agents - Tough-Llm-Tests - Some tough questions to test new models. (Building / LLM Models)
• awesome_ai_agents - Tough-Llm-Tests - Some tough questions to test new models. (Building / LLM Models)

README

        
Accompaniment to [this video](https://youtu.be/CvPEz2ngusw) where I test Llama 3 7b against a bunch of models. Here's a bunch of relatively tough questions I use to grade models and see how they perform!

## How to use

Each folder in `prompts` has a bunch of task prompts and at least one source to use. Copy them together and you're done!