Owner · IST-DASLab
IST-DASLab
IST-DASLab's tracked open-source repos, sorted by stars.
3 repos
- #1
Code for the ICLR 2023 paper "GPTQ: Accurate Post-training Quantization of Generative Pretrained Transformers".
★ 2,362+6Star change over the last 7 days - #2
FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.
★ 1,137+4Star change over the last 7 days - #3
Code for the ICML 2023 paper "SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shot".
★ 893+1Star change over the last 7 days
