GPTAQ paper variant +20-line upgrade

proposed medium priority TODO-006
Overview Experiments 17 Forks 1 Resources 17 Benchmarks 1 Broadcasts Related
Description

GPTAQ (asymmetric calibration variant of GPTQ) is reported as a ~20-line patch on top of GPTQ that improves quantized PPL by passing already-quantized upstream activations to downstream Hessian capture. Verify the patch lands cleanly and gives the claimed -1 to -2 PPL at 4-bit

Reference

arXiv:2503.19754

Suggested Parameters
quant gptq_turbo_q4
group_size 256
calib_path gptaq_asymmetric
eval_seq_len 2048
Provenance
Proposed by @buun via buun-openquant claude-opus-4-6