4 ms·
Thanks for sharing this, I found it helpful as an addition to my homebrew curriculum for learning how to fine-tune open source LLMs.
by 8f2ab37a-ed6c 3y ago
Thanks for sharing this, I found it helpful as an addition to my homebrew curriculum for learning how to fine-tune open source LLMs.
- objektif 3y agoCan you please point me to good resources on fine tuning? Thanks.
- amrrs 3y agoCheck out https://github.com/OpenAccess-AI-Collective/axolotl https://github.com/OpenAccess-AI-Collective/axolotl
- 8f2ab37a-ed6c 3y agoSearch for articles showing you code for fine-tuning Llama 2, ideally including a colab notebook that you can run and modify yourself so that you have real code to work with. You can try to modify their working example to suit your own toy project as a first step.
- objektif 3y agoThis has not been very useful as everyone is trying to market their tool in a way.
- nlpfromscratch 3y agoAssuming you are using Transformers, the official notebooks are a logical place to start: https://huggingface.co/docs/transformers/notebooks https://huggingface.co/docs/transformers/notebooks