Histórico de Commits

Autor SHA1 Mensagem Data
  AlpinDale 2b85ffb1a5 chore: minor cleanups há 4 meses atrás
  AlpinDale 4d4e767838 ci: take one of fixing lint issues há 4 meses atrás
  AlpinDale 0e6c400b13 feat: re-add GGUF (#600) há 4 meses atrás
  AlpinDale edffcecc67 chore: add proper logging for spec decoding verification há 4 meses atrás
  AlpinDale d357341203 chore: add pipeline parallel support for Qwen há 4 meses atrás
  AlpinDale 98f9dbd734 feat: Triton Kernels for Punica (#613) há 4 meses atrás
  AlpinDale 07cc8a56bb fix: add nemotron to PP_SUPPORTED_MODELS há 4 meses atrás
  AlpinDale ea838abb6b fix: disable enforce_eager for bnb há 4 meses atrás
  AlpinDale fce2c2e304 fix: support ignore patterns in model loader há 4 meses atrás
  AlpinDale cb44c8daa8 feat: support FP8 KV Cache scales from compressed-tensors há 4 meses atrás
  AlpinDale ba371fbbbd feat: AWQ marlin kernels (#603) há 4 meses atrás
  AlpinDale a4cbcfe59f feat: disable logprob serialization to CPU for spec decode há 4 meses atrás
  AlpinDale 9be43994fe feat: fbgemm quantization support (#601) há 4 meses atrás
  AlpinDale 45a004874c chore: allow specifying custom Executor há 4 meses atrás
  AlpinDale b7a2d52e47 fix: allow using mp executor for pipeline parallel há 4 meses atrás
  AlpinDale 6671e3a162 feat: add CPU offloading support (#598) há 4 meses atrás
  AlpinDale ee2c5d34da feat: add fp8 channel-wise weight quantization support há 4 meses atrás
  AlpinDale 6c4c20652b feat: pipeline parallel support for mixtral há 4 meses atrás
  AlpinDale 5289c14b24 feat: Asymmetric Tensor Parallel (#594) há 4 meses atrás
  AlpinDale ddb28a80a3 fix: bump torch for rocm, unify CUDA_VISIBLE_DEVICES for cuda and rocm há 4 meses atrás
  AlpinDale 99680b2d23 feat: soft prompts (#589) há 4 meses atrás
  AlpinDale 5761ef8c35 feat: gemma-2 support há 4 meses atrás
  AlpinDale 1ff6d4c3d7 feat: support pipeline parallel on indivisible GPU count (#587) há 4 meses atrás
  AlpinDale 4f7d212b70 feat: remove vision language config há 4 meses atrás
  AlpinDale bdf1cc1aec fix: allow using custom all reduce when pp_size > 1 há 4 meses atrás
  AlpinDale 5240c0da23 fix: avoid unnecessary ray import warnings há 4 meses atrás
  AlpinDale 5be90c3859 Mamba infrastrucuture support (#586) há 4 meses atrás
  AlpinDale ae04f57ec1 feat: Pipeline Parallel support (#581) há 4 meses atrás
  AlpinDale dd378ea063 feat: MLPSpeculator with tensor parallel há 4 meses atrás
  AlpinDale 3a0fdf7b9b chore: remove `image_input_type` from VLM config há 4 meses atrás