LLM-QAT: Data-Free Quantization Aware Training for Large Language Models (Frantar et al., NVIDIA)
paper
Fields
id
sources/qlab-llm-qat-paperPublisher
arXiv
Kind
paper
Published
2023-05-29
Accessed
2026-10-03
notes
Single version. Distillation target is the pretrained model's own generations, hence 'data-free'.
Notes
Single version. Distillation target is the pretrained model's own generations, hence 'data-free'.
Cross-references
This record references no other records.