Commit History

Autor SHA1 Mensaxe Data
  AlpinDale ba371fbbbd feat: AWQ marlin kernels (#603) hai 5 meses
  AlpinDale fa15bad2ea chore: minor AMD fixes hai 5 meses
  AlpinDale dd18c5042c move prepare_inputs to the GPU (#596) hai 5 meses
  AlpinDale f5d52320da Port mamba kernels to Aphrodite (#595) hai 5 meses
  AlpinDale 321a0892e0 let's not build these for now hai 5 meses
  AlpinDale b1e61268a8 bump torch to 2.3.1 hai 5 meses
  AlpinDale ddb28a80a3 fix: bump torch for rocm, unify CUDA_VISIBLE_DEVICES for cuda and rocm hai 5 meses
  AlpinDale ad24e74a99 feat: FP8 weight-only quantization support for Ampere GPUs hai 5 meses
  AlpinDale 5be90c3859 Mamba infrastrucuture support (#586) hai 5 meses
  AlpinDale 5b464d36ea feat: bias epilogue support for cutlass kernels hai 5 meses
  AlpinDale 2c321ce1f2 chore: upgrade to rocm 6.1, update docker hai 5 meses
  AlpinDale 7e54c3916d chore: factor out epilogues from cutlass kernels hai 5 meses
  AlpinDale 156f577f79 feat: switch from `PYBIND11_MODULE` to `TORCH_LIBRARY` (#569) hai 5 meses
  AlpinDale 00acf371f9 rocm: fused topk softmax hai 5 meses
  AlpinDale 0307da9e15 refactor: bitsandbytes -> autoquant hai 5 meses
  AlpinDale 4ab4c5c87c oops hai 5 meses
  AlpinDale d45c846c8c do not build sm_90a for cuda 11 hai 5 meses
  AlpinDale f4ea11b982 feat: initial support for activation quantization hai 5 meses
  AlpinDale 2313c97e3d add cutlass w8a8 kernels (#556) hai 6 meses
  AlpinDale c66b1b57b1 Marlin 2:4 sparsity (#555) hai 6 meses
  AlpinDale 251568470e initial nvidia fp8 e4m3 for kv cache hai 6 meses
  AlpinDale e3f2ea4850 make punica kernels work with rocm hai 6 meses
  AlpinDale f22b700ee4 feat: marlin kernels for GPTQ (#547) hai 6 meses
  AlpinDale fca911ee0a vLLM Upstream Sync (#526) hai 6 meses
  AlpinDale 9d81716bfd [v0.5.3] Release Candidate (#388) hai 8 meses
  AlpinDale e120404436 Revert "feat: CMake Build System Generator (#332)" hai 10 meses
  AlpinDale ad6802690f feat: CMake Build System Generator (#332) hai 10 meses