arkanathp/1M-stratified-pile-uncopyrighted
1M stratified Pile-uncopyrighted Training corpus of the paper Repurposing Pre-trained LLMs as High Fidelity Continuous Text Autoencoders (paper, code): 999,990 documents drawn from Pile-uncopyrighted, general web and technical text. One JSONL file with a text field (999,990 documents). Documents were collected stratified by length into 41 buckets of 100 characters so that the training set is balanced from short through full-length inputs. The LLMAE recipes (configs/*.py in the… See the full description on the dataset page: https://huggingface.co/datasets/arkanathp/1M-stratified-pile-uncopyrigh
| Author / organization | arkanathp |
|---|---|
| License (as declared) | other |
| Downloads | 137 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 7 February 2026 |
| Last updated at source | 12 October 2026 |
Popularity, last 3 days
137 now (+3 since 2026-10-17). Counts as reported by the source, snapshotted daily by Silytics.
Connected in the knowledge graph
Related papers
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 19 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.