Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Feb 19, 2025

• Update README.md (#239)

  ZeyuMi

  authored
  843195e

  Copy full SHA for 843195e

• Release code for OPT models (#245)

  Show description for 1113aef
  AliceRayLu

  authored
  1113aef

  Copy full SHA for 1113aef

Commits on Sep 6, 2024

• Fix segmentation fault for models exceeding 40B on AMD GPUs & optimize mul_mat_axpy operation (#217)

  Show description for 6ae7e06
  

  Tworan

  and

  tworan

  authored
  6ae7e06

  Copy full SHA for 6ae7e06

Commits on Jul 1, 2024

• add convert-hf-to-powerinfer-gguf.py to CMakeLists.txt (#205)

  MatthewCroughan

  authored
  61cac9b

  Copy full SHA for 61cac9b

Commits on Jun 27, 2024

• Update README (#204)

  YixinSong-e

  authored
  7639e72

  Copy full SHA for 7639e72

Commits on Jun 11, 2024

• minor in README(#193)

  YixinSong-e

  authored
  4896dfb

  Copy full SHA for 4896dfb

• add news (#192)

  Show description for 9bea6ce
  

  YixinSong-e

  and

  hodlen

  authored
  9bea6ce

  Copy full SHA for 9bea6ce

Commits on May 20, 2024

• Support AMD ROCm(#187)

  Show description for 3f638f7
  

  freelulul

  and

  hodlen

  authored
  3f638f7

  Copy full SHA for 3f638f7

Commits on Apr 7, 2024

• fix: offload ffn norm weights (#179)

  hodlen

  authored
  906830b

  Copy full SHA for 906830b

Commits on Apr 2, 2024

• Remove axpy dense op (#177)

  hodlen

  authored
  b478398

  Copy full SHA for b478398

• hotfix: failed to build ffn with LLM_GATE_SEQ (#178)

  hodlen

  authored
  c8e15df

  Copy full SHA for c8e15df

Commits on Mar 28, 2024

• Support Bamboo LM sparse and dense inference (#171)

  Show description for b25cb86
  hodlen

  authored
  b25cb86

  Copy full SHA for b25cb86

Commits on Mar 27, 2024

• Optimize mul_mat_sparse for INT4 quantized weights (#174)

  Show description for 1374ab0
  

  hodlen

  and

  syx

  authored
  1374ab0

  Copy full SHA for 1374ab0

• Use mul_mat_transpose at axpy op for large batch (#168)

  Show description for 94fd6d2
  

  Begunner

  and

  hodlen

  authored
  94fd6d2

  Copy full SHA for 94fd6d2

Commits on Mar 17, 2024

• Full-GPU computational graph and CUDA refactoring (#153)

  Show description for 47e9d7e
  hodlen

  authored
  47e9d7e

  Copy full SHA for 47e9d7e

Commits on Mar 14, 2024

• Add prosparse to README (#149)

  Show description for 8eea28c
  

  Raincleared-Song

  and

  hodlen

  authored
  8eea28c

  Copy full SHA for 8eea28c

Commits on Mar 7, 2024

• fix: unknown host compiler flags passed to nvcc (#161)

  hodlen

  authored
  c36f1df

  Copy full SHA for c36f1df

Commits on Feb 20, 2024

• Fix compiling issue under git worktrees (#146)

  hodlen

  authored
  7b09717

  Copy full SHA for 7b09717

Commits on Feb 18, 2024

• Fix CMake requirement in README (#145)

  hodlen

  authored
  c859c76

  Copy full SHA for c859c76

Commits on Feb 1, 2024

• fix(ggml): add macro definitions to allow compiling to rocm (#139)

  mack-w

  authored
  52ff38f

  Copy full SHA for 52ff38f

• Fix a bug of memory leak on Windows platform (#138)

  Show description for 4fd41d2
  Tan-YiFan

  authored
  4fd41d2

  Copy full SHA for 4fd41d2

Commits on Jan 30, 2024

• Docs: add detailed instruction on downloading HF models (#136)

  Show description for ca5acde
  hodlen

  authored
  ca5acde

  Copy full SHA for ca5acde

Commits on Jan 26, 2024

• Update README: add Windows-specific commands (#133)

  hodlen

  authored
  120b07d

  Copy full SHA for 120b07d

• fix activation file detection (#134)

  hodlen

  authored
  7f69428

  Copy full SHA for 7f69428

Commits on Jan 24, 2024

• remove unused toolchain files (#132)

  hodlen

  authored
  0128451

  Copy full SHA for 0128451

• Fix CUDA performance regression due to auto offloading (#127)

  Show description for 5c782ea
  hodlen

  authored
  5c782ea

  Copy full SHA for 5c782ea

Commits on Jan 11, 2024

• Support CPU/GPU inference on Windows (#114)

  Show description for 2217e7f
  

  hodlen

  and

  zwb

  authored
  2217e7f

  Copy full SHA for 2217e7f

Commits on Jan 4, 2024

• Convert HF models with sparse threshold specified (#76)

  Szy0127

  authored
  fee0adf

  Copy full SHA for fee0adf

Commits on Jan 3, 2024

• Fix argument parsing in examples/batched (#113)

  hodlen

  authored
  c2cfbf4

  Copy full SHA for c2cfbf4

• Support broader python and pip versions (#112)

  Show description for 0a98ca2
  hodlen

  authored
  0a98ca2

  Copy full SHA for 0a98ca2

Commits on Dec 29, 2023

• Support setting VRAM budget for examples/server (#106)

  Show description for 74c5c58
  hodlen

  authored
  74c5c58

  Copy full SHA for 74c5c58

Commits on Dec 27, 2023

• Fix generation error under INT4 quantization and batched prompting (#99)

  Show description for 79986ec
  hodlen

  authored
  79986ec

  Copy full SHA for 79986ec

• Add our Kansan to README.md (#89)

  hodlen

  authored
  9b830d6

  Copy full SHA for 9b830d6

• Update issue templates of PowerInfer (#90)

  Show description for 0dde047
  hodlen

  authored
  0dde047

  Copy full SHA for 0dde047

Commits on Dec 24, 2023

• Update README.md (#72)

  YixinSong-e

  authored
  c72c6da

  Copy full SHA for c72c6da

Pagination

Previous

Next