DevOps & Infrastructure
Gptq
Post-training 4-bit quantization for LLMs with minimal accuracy loss. Use for deploying large models (70B, 405B) on consumer GPUs, when you need 4× memory reduction with <2% perplexity degradation, or for faster inference (3-4× speedup) vs FP16. Integrates with transformers and PEFT for QLoRA fine-tuning.
- Publisher / source
- Orchestra-Research/AI-Research-SKILLs (found on skills.sh, an awesome list)
- Upstream record
- Orchestra-Research/AI-Research-SKILLs
- Version
- Version 3
- Published
- Publication review
- Approved for publication
- Content fingerprint
ce19de416afa03a0e6a764c9cb8b8ba9a0b94812b9de1ad5a0c2cb2932009f28