Historique des commits

Auteur SHA1 Message Date
  AlpinDale e31c6f0b45 feat: refactor modeling logic and support more models (#274) il y a 10 mois
  AlpinDale 224b87b484 feat: add fused mixtral moe support (#238) il y a 10 mois
  AlpinDale d9b65e6c5f feat: DeepSeek MoE support (#237) il y a 11 mois
  AlpinDale c3a221eb02 feat: GGUF, QuIP#, and Marlin support (#228) il y a 11 mois
  AlpinDale 31c95011a6 feat: FP8 E5M2 KV Cache (#226) il y a 11 mois
  AlpinDale 641bb0f6e9 feat: add custom allreduce kernels (#224) il y a 11 mois
  AlpinDale 5053743c1c feat: speedup AWQ (#223) il y a 11 mois
  AlpinDale 8fa608aeb7 feat: replace Ray with NCCL for control plane comms (#221) il y a 11 mois
  AlpinDale 7e72ce0a73 feat: mixtral tensor parallelism (#193) il y a 1 an
  AlpinDale 15a0454172 feat: FP8 KV Cache (#185) il y a 1 an
  AlpinDale 801eda0b7a feat: support GPTQ 2, 3, and 8bit quants (#181) il y a 1 an
  AlpinDale 7c6fdea535 fix: GPTQ warnings and exllama states (#171) il y a 1 an
  AlpinDale 02f3ab3501 fix: replace head_mapping with num_kv_heads (#161) il y a 1 an
  AlpinDale 62b2c4119d feat: re-write GPTQ and refactor exllama kernels (#152) il y a 1 an
  AlpinDale 1334a833a4 feat: AMD ROCm support (#95) il y a 1 an
  AlpinDale 2b1ba581f9 feat: re-implement GPTQ (#141) il y a 1 an
  AlpinDale 8223f85c1b feat: SqueezeLLM support (#140) il y a 1 an
  AlpinDale 1aab8a7d6f feat: speedup compilation times by 3x (#130) il y a 1 an