birgermoell/oellm-longctx-tokenized-superlong-512k-1m-2m-v2

OELLM Superlong 512K/1M/2M Tokenized V2 This is a Megatron-LM tokenized artifact for long-context continuation training toward 1M-2M context windows. It combines the original v1 512K/1M/2M superlong artifact with a second extension batch of non-duplicate real long-source slices and harder synthetic long-dependency examples. The dataset is intended for context-extension training, not benchmark evaluation. The main goal is to expose the model to full-span dependencies and… See the full description on the dataset page: https://huggingface.co/datasets/birgermoell/oellm-longctx-tokenized-superlong-

<iframe src="https://silytics.si/embed/item/eb52930c-7036-4385-8e2f-5ad89948675c" 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 Information may be incomplete or out of date.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.