donghyunli/Llama-2-70b-KronQ-HG
Llama-2-70b-hf — KronQ H_G (output-side gradient covariance) Paper: arXiv:2607.07964 · Code: GitHub Pre-computed H_G for Llama-2-70b-hf, the output-side curvature factor used by KronQ under the K-FAC factorization H ≈ H_X ⊗ H_G. Per-sublayer empirical-Fisher gradient covariance (E[g gᵀ] over the layer output), distinct from the input-side Hessian H_X. Publishing this lets you reproduce KronQ quantization without the offline Fisher precompute step. Contents (80 layers… See the full description on the dataset page: https://huggingface.co/datasets/donghyunli/Llama-2-70b-KronQ-HG.
| Author / organization | donghyunli |
|---|---|
| License (as declared) | llama2 |
| Downloads | 2.9K (as reported by Hugging Face) |
| Likes | 0 |
| Created | 23 June 2026 |
| Last updated at source | 13 July 2026 |
Popularity, last 3 days
2,877 now (+105 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Connected in the knowledge graph
Related papers
More from donghyunli
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 Information may be incomplete or out of date.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.