hbfreed/dolci-distill-packed
Dolci Distill Packed Pre-packed training data for knowledge distillation from OLMo-3-7B-Instruct to a pruned student model. Description This dataset contains tensorized and sequence-packed batches ready for distillation training. The data was preprocessed from hbfreed/Dolci-Instruct-RL-Completions to avoid preprocessing overhead during training. Format 35 .pt files: 34 training chunks + 1 validation chunk Pack length: 6144 tokens ~158GB total Each… See the full description on the dataset page: https://huggingface.co/datasets/hbfreed/dolci-distill-packed.
| Author / organization | hbfreed |
|---|---|
| License (as declared) | apache-2.0 |
| Downloads | 749 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 19 January 2026 |
| Last updated at source | 19 January 2026 |
Popularity, last 3 days
749 now (+135 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Recent changes
- 16 Oct 2026 Description: Dolci Distill Packed Pre-packed training data for knowledge distillation from OLMo-3-7B-Instruct to a pruned student model. Description This dataset contains tensorized and sequence-packed batches ready for distillation training. The data w → Dolci Distill Packed Pre-packed training data for knowledge distillation from OLMo-3-7B-Instruct to a pruned student model. Description This dataset contains tensorized and sequence-packed batches ready for distillation training. The data w
Changes detected by Silytics between refreshes of the source.
More from hbfreed
- hbfreed/flexolmo-math-logprobsDataset
- hbfreed/Picklebot-50KDataset
- hbfreed/Picklebot-2MDataset
- hbfreed/Picklebot-130KDataset
- hbfreed/subliminal-number-datasetsDataset
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 Information may be incomplete or out of date.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.