donghyunli/Llama-2-7b-KronQ-HG

Llama-2-7b — KronQ H_G (output-side gradient covariance)

Paper: arXiv:2607.07964 · Code: GitHub

Pre-computed H_G for Llama-2-7b, the output-side curvature factor used by KronQ under the K-FAC factorization H ≈ H_X ⊗ H_G. H_G is the per-sublayer sampled-Fisher gradient covariance (labels drawn from the model distribution) (E[g gᵀ] over the layer output), distinct from the standard input-side Hessian H_X (which GPTQ/GPTAQ build online during calibration).

Publishing this lets you reproduce KronQ quantization without the offline Fisher precompute step.

Contents (32 layers × 7 sublayers, ~39 GB)

layer_<i>/self_attn_{q,k,v,o}_proj_G.pt
layer_<i>/mlp_{gate,up,down}_proj_G.pt
metadata.pt

Each *_G.pt is the out_features × out_features gradient covariance for that sublayer.

Usage

Point KronQ's --grad_dir at the downloaded folder — it skips precompute_gradients.py:

python main.py --model meta-llama/Llama-2-7b-hf \
    --w_bits 4 --w_groupsize -1 --w_clip --w_asym --a_bits 16 --act_order \
    --bi_calibration --use_gptaq --incoh_rotate --incoh_kernel had --incoh_mode full \
    --alpha 0.25 --grad_dir <downloaded_HG_dir>

This is raw (unrotated) H_G, the form used by the weight-only recipe (H_G cancels in the OBS update, so raw and rotated give identical weights for per-channel weight-only).

License

Derived from Llama-2-7b — subject to the Llama 2 Community License.