PTQ-Level Efficiency, QAT-Level Accuracy | 训练后量化的效率,量化感知训练的精度
Commits on main
enable llava & Qwen-VL multimodal model quantization (#165)
aeb9e40 WeiweiZhang1 committed 2 years ago
fix autoround format mixed precision issue and refine gptq format code (#183)
e281499 wenhuach21 committed 2 years ago
[Large impact]set the default nsamples to 128 and low_gpu_mem_usage to False (#174)
ab61482 wenhuach21 committed 2 years ago
1 fix a bug in autoround format with the latest transformers 2 rename n_samples n_blocks to nsamples nblocks (#163)
f9e7d79 wenhuach21 committed 2 years ago
fix bug at whole block is excluded from quantization (#156)
edcec56 wenhuach21 committed 2 years ago