hexhowells / hexhowells/tinyLLM
Train on more data
Open
@hexhowells is already working on this.
Since Sep 6, 2026.
enhancement
- Dominant language
- Python
- Stars
- 0
- Forks
- 0
- Avg merge
- 1m
- Merged PRs (30d)
- 1
Description
Currently planned to train on fineweb-edu-10BT, however for a small model it's better to overfit on more tokens and ignoring chinchilla scaling laws. Thus it might make sense to download & pre-process fineweb-edu-30BT instead, and we can just stop early if needed (will need to take into account cosine annealing)
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Assessment
This issue has not been assessed yet.