Ludwig: Fine-Tune Mistral-7b missing LudwigModel import and/or definition
noahgift opened this issue · comments
Describe the bug
Ludwig: Fine-Tune Mistral-7b missing LudwigModel import and/or definition.
To Reproduce
Steps to reproduce the behavior:
- Go to 'https://colab.research.google.com/drive/1i_8A1n__b7ljRWHzIsAdhO7u7r49vUm4#scrollTo=k-dtCIj73498'
- Run Cells in cola
- Scroll down to 'qlora_fine_tuning_config = yaml.safe_load(' cell and notice LudwigModel not defined
- see error
Please provide code, yaml config file and a sample of data in order to entirely reproduce the issue.
Issues that are not reproducible will be ignored.
Expected behavior
A clear and concise description of what you expected to happen.
Screenshots
If applicable, add screenshots to help explain your problem.
Environment (please complete the following information):
- OS: [e.g. iOS]
- Version [e.g. 22]
- Python version
- Ludwig version
Additional context
Add any other context about the problem here.
Hey @noahgift - thanks for flagging this issue and sorry you ran into it. I'm taking a look right now and will get back to you once I know what's going on and find a fix.
@noahgift I was able to repro and I see the problem - just a few missing imports. Let me update the notebook so that it runs correctly.
@noahgift I was able to repro and I see the problem - just a few missing imports. Let me update the notebook so that it runs correctly.
Thank you so much! Putting this into a Coursera Duke course on LLMOps!
I only just saw this -- did I really not include imports into the notebook? Not sure how this could have happened if it ran fine. In fact, I just checked the original source of this notebook, and all the imports seem to be there (in addition, the references to the notebooks in the blog post also contain the imports). Apologies, @noahgift, for any confusion and lost productivity. Thank you!
@alexsherstinsky Thanks for checking! I believe this is a notebook I created a few months ago after our collaboration when Mistral first came out and I may have missed some imports. It was meant to a be a very lean/stripped down version of the original notebook adapted and blogpost we created together, but I adapted it for the code alpaca dataset in conjunction with fine-tuning on Llama-2-7b/13b as can be seen in the Ludwig README.
I'm also surprised that I missed some imports, but I've updated the notebook now to have the right imports! Just making sure that the training parameters are adjusted so that the notebook has good inference performance when it's run!
@arnavgarg1 Now I see what is going on! Thank you very much for clarifying! Whew! 😄
@noahgift Alright, things should be fixed up! Are you able to give it a try now?