princeton-nlp/prolong-data-64K

princeton-nlp/prolong-data-64K [Paper] [HF Collection] [Code] ProLong (Princeton long-context language models) is a family of long-context models that are continued trained and supervised fine-tuned from Llama-3-8B, with a maximum context window of 512K tokens. Our main ProLong model is one of the best-performing long-context models at the 10B scale (evaluated by HELMET). To train this strong long-context model, we conduct thorough ablations on the long-context pre-training data… See the full description on the dataset page: https://huggingface.co/datasets/princeton-nlp/prolong-data-64K.

<iframe src="https://silytics.si/embed/item/48113587-d86e-4112-a0f4-6071c1e13463" 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 Information may be incomplete or out of date.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.