Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jan 21, 2025

• fix g_idx may be None (AutoGPTQ#771)

  CSY-ModelCloud

  authored
  c0f8edb

  Copy full SHA for c0f8edb

• fix the issue of qlinear packing being too slow. (AutoGPTQ#770)

  Show description for 191bd77
  Marxlp

  authored
  191bd77

  Copy full SHA for 191bd77

Commits on Jan 16, 2025

• Fix import of mapping from PEFT (AutoGPTQ#764)

  Show description for c039255
  BenjaminBossan

  authored
  c039255

  Copy full SHA for c039255

• Mod add warning (AutoGPTQ#766)

  Show description for d3793f5
  CL-ModelCloud

  authored
  d3793f5

  Copy full SHA for d3793f5

Commits on Dec 15, 2024

• spell (AutoGPTQ#757)

  Qubitium

  authored
  6689349

  Copy full SHA for 6689349

• Note about autogptq dev status (AutoGPTQ#756)

  Qubitium

  authored
  f90a77f

  Copy full SHA for f90a77f

Commits on Nov 27, 2024

• Updated readme for HPU (AutoGPTQ#749)

  Show description for 7e291e9
  

  4 peopleauthored
  7e291e9

  Copy full SHA for 7e291e9

Commits on Sep 28, 2024

• Quickfix for failure after merged PR 731 (AutoGPTQ#734)

  Show description for bd18bb8
  avoroshilov

  authored
  bd18bb8

  Copy full SHA for bd18bb8

• support for minicpm3.0 (AutoGPTQ#731)

  Show description for e8e9219
  LDLINGLINGLING

  authored
  e8e9219

  Copy full SHA for e8e9219

Commits on Aug 14, 2024

• Fix error when loading quantized model with low_cpu_mem_usage is true (AutoGPTQ#719)

  Show description for 097dd04
  ryanmiao

  authored
  097dd04

  Copy full SHA for 097dd04

Commits on Jul 16, 2024

• Add Gemma 2 support (AutoGPTQ#710)

  maxmet91

  authored
  caf343b

  Copy full SHA for caf343b

Commits on Jul 9, 2024

• Update README.md with Intel Gaudi HPU information (AutoGPTQ#706)

  Show description for 65c2e15
  

  4 peopleauthored
  65c2e15

  Copy full SHA for 65c2e15

Commits on Jun 29, 2024

• Fix upstream regression when there's no HPU device (AutoGPTQ#701)

  Show description for 52eaddd
  HolyFalafel

  authored
  52eaddd

  Copy full SHA for 52eaddd

Commits on Jun 27, 2024

• [BUG] Fix H100 crash/compat with Marlin (AutoGPTQ#654)

  Qubitium

  authored
  8d8e0ea

  Copy full SHA for 8d8e0ea

• [PERFORMANCE] Fix Packing thread regression in code (AutoGPTQ#642)

  Qubitium

  authored
  ecd0c1c

  Copy full SHA for ecd0c1c

Commits on Jun 24, 2024

• Supporting uint4 inference of pre-quantized models in HPU (AutoGPTQ#689)

  Show description for b57bea0
  

  HolyFalafel

  and

  Yantom1

  authored
  b57bea0

  Copy full SHA for b57bea0

Commits on Apr 27, 2024

• now that h100 is validated, remove hopper check

  Qubitium

  committed
  5b456a6

  Copy full SHA for 5b456a6

• revert debug code

  Qubitium

  committed
  5b307da

  Copy full SHA for 5b307da

• fix h100

  Qubitium

  committed
  e474efd

  Copy full SHA for e474efd

• revert

  Qubitium

  committed
  aa2e385

  Copy full SHA for aa2e385

• Revert "backport h100 fixed marlin kernel from vllm"

  Show description for 25decb0
  Qubitium

  committed
  25decb0

  Copy full SHA for 25decb0

• backport h100 fixed marlin kernel from vllm

  Qubitium

  committed
  8ac1b87

  Copy full SHA for 8ac1b87

Commits on Apr 16, 2024

• stream pytest output

  Qubitium

  committed
  991508a

  Copy full SHA for 991508a

• fix pack() thread regression via code

  Qubitium

  committed
  35cfceb

  Copy full SHA for 35cfceb

Commits on Apr 12, 2024

• Force warn users to reduce threads or packing may regress ~100x (AutoGPTQ#628)

  Show description for ea829c7
  Qubitium

  authored
  ea829c7

  Copy full SHA for ea829c7

• Cohere Command-R Added (AutoGPTQ#631)

  Show description for ce39d61
  

  egortolmachev

  and

  t3ga

  authored
  ce39d61

  Copy full SHA for ce39d61

Commits on Apr 5, 2024

• Fix padding not used correctly in exllama v2 layer (AutoGPTQ#626)

  Show description for b4b801c
  Qubitium

  authored
  b4b801c

  Copy full SHA for b4b801c

Commits on Mar 28, 2024

• Add checkpoint_format to unify all current/future saved weight formats (AutoGPTQ#603)

  Show description for 866b4c8
  

  Qubitium

  and

  fxmarty

  authored
  866b4c8

  Copy full SHA for 866b4c8

• Show meaningful error on Maxwell build (AutoGPTQ#622)

  Show description for ff3dcc4
  fxmarty

  authored
  ff3dcc4

  Copy full SHA for ff3dcc4

• Add --no-build-isolation to install instructions (AutoGPTQ#620)

  Show description for a78daec
  fxmarty

  authored
  a78daec

  Copy full SHA for a78daec

• Add support for MPT (#73)

  Show description for a9f3d46
  LaaZa

  authored
  a9f3d46

  Copy full SHA for a9f3d46

• fix tritonv2 in make_sure_no_tensor_in_meta_device (AutoGPTQ#617)

  LaaZa

  authored
  ca187fb

  Copy full SHA for ca187fb

Commits on Mar 26, 2024

• Patch load_checkpoint_in_model till accelerate 0.29 (AutoGPTQ#613)

  Show description for 9fd7a6c
  fxmarty

  authored
  9fd7a6c

  Copy full SHA for 9fd7a6c

• fix quantization samples

  fxmarty

  committed
  24efa31

  Copy full SHA for 24efa31

Commits on Mar 25, 2024

• Fix: bias option for qlinear should depend on origin layer (AutoGPTQ#602)

  Show description for 82070a1
  Coco58323

  authored
  82070a1

  Copy full SHA for 82070a1

Pagination

Previous

Next