TwoSix Review: FineTune-Mistral-LLM-OwnData
- Add description of what Mistral is and why it is important.
- Overall, a lot of terms relating to Mistral and fine-tuning LLMs are brought up without being explained, motivating why they are important, or overall teaching the subject.
- Remove “(There may be a better way to do this... if so please do let me know!)”
- Add docstrings and explanations for the functions made
- For Accelerator description either remove it if it seems too complicated or explain a through pros/cons to using it which makes it hard to determine whether to use it or not.
- Provide context and useful resources to the complicated topics like tokenization
- When running peft model there is an error that says “can’t find adapter_config.json”