Precompiled Windows AMD64 Wheels for GPTQ-for-LLaMa CUDA
See the Linux-x64 branch for Linux x86_64 wheels.
Wheels in root directory compiled from oobabooga's fork
- Supports Pascal+ (compute 6.0+)
832e220 wheels compiled from latest (as of writing) commit of GPTQ-for-LLaMa
610fdae wheels compiled from equivalent commit of GPTQ-for-LLaMa
0cc4m wheels compiled from 0cc4m's fork for KoboldAI
Deprecated quant_cuda wheel is included for those who want it.
- Supports late-Kepler+ (compute 3.5+)
Wheels are compiled using GitHub Actions.
Intended for use with:
text-generation-webui
KoboldAI