Skip to content

shouraykumra/LLM

Repository files navigation

LLM

Use Case 1: Utilizing the Mistral Model without Unslothing and Unsmashed Versions (Original 7B Parameters Model)

  • CPU Usage: 13GB
  • Data Performance: Unable to load on the free version of Google Colab; had to interrupt the kernel as it took a significant amount of time.

⠀Use Case 2: Utilizing the Mistral Model without Unslothing and Smashed Versions

  • CPU Usage: 5290MB/s
  • Data Performance:
    • Task: Running the base model with prompt engineering to generate replies for YouTube comments.
    • CPU Time: 4.57 seconds to generate a reply to a prompt. Significant improvement observed with the introduction of the smashed version of the model.
    • To assess the model's performance, it needs to be trained for a specific use case. For this base case, it performs as expected.

Use Case 3: Utilizing Mistral Model with Lightening Module Use Case 4: Utilizing Mistral Model with Unsloth

About

Swimming across ocean of LLMs

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published