Pre-to-Post-2/lichess_pretrain_elo_cutoff
Lichess pretraining games, binned by Elo All 307,921,006 non-bullet Lichess games in our chess pretraining set, split into 12 bins of 200 Elo by the game's average rating, from 600 to 3000. Each bin is provided both as text (one game per line, SAN movetext) and as pre-tokenized uint16 shards ready for the LLM-Pretraining trainer. In total that is 69,605,643,984 tokens (69.61B). Bins A game goes in bin [lo, hi) when lo <= avg_elo < hi, where avg_elo = (white_elo +… See the full description on the dataset page: https://huggingface.co/datasets/Pre-to-Post-2/lichess_pretrain_elo_cutoff.
| Author / organization | Pre-to-Post-2 |
|---|---|
| License (as declared) | cc0-1.0 |
| Downloads | 455 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 9 October 2026 |
| Last updated at source | 15 October 2026 |
Popularity, last 13 days
455 now (+11 since 2026-10-26). Counts as reported by the source, snapshotted daily by Silytics.
More from Pre-to-Post-2
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 7 November 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.