Cion-lab/ounce100m-mix-v1

ounce100m-mix-v1 Tokenised, deduplicated English pre-training mix built for project ounce100m — a 90–110M-parameter decoder-only base model trained from scratch on ~1B tokens on 2x Tesla T4. This dataset is the exact input to that run, published so the result can be reproduced. Training tokens: 1,109,714,831 across 139 shards Held-out validation tokens: 22,934,043 in 12 shards (val/), sampled every 50th merged document — across sources, never whole documents, and excluded from… See the full description on the dataset page: https://huggingface.co/datasets/Cion-lab/ounce100m-mix-v1.

<iframe src="https://silytics.si/embed/item/9e72a740-01fa-4498-9eea-1eb6bb9e45e7" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.