Commits · 21dc0d8415863835a7de0aeb0ee09ab685143c4b · OpenDAS / AutoAWQ

02 Jul, 2024 1 commit
- 增加DCU version信息 · 21dc0d84
  gaoqiong authored Jul 02, 2024
  
  21dc0d84
02 May, 2024 2 commits
- Bump to 0.2.5 (#468) · 5f3785dc
  Casper authored May 02, 2024
  
  5f3785dc
- support max_memory to specify mem usage for each GPU (#460) · 12581fbd
  laoda513 authored May 03, 2024
  
  12581fbd
01 May, 2024 2 commits
- FIX: Fix multiple generations for new HF cache format (#444) · 33af761a
  Younes Belkada authored May 01, 2024
  
  33af761a
- FIX: 'awq_ext' is not defined error (#465) · 76edff62
  Younes Belkada authored May 01, 2024
  
  76edff62
19 Apr, 2024 1 commit
- Fix path (#451) · 4fc6cc03
  Casper authored Apr 19, 2024
  
  4fc6cc03
18 Apr, 2024 2 commits
- [BUG] Fix github action documentation build (#449) · 24b98c25
  Shaun Prince authored Apr 18, 2024
  
  24b98c25
- Update generate example to llama 3 (#448) · edcf780f
  Casper authored Apr 18, 2024
  
  edcf780f
12 Apr, 2024 1 commit
- Fix starcoder2 fused norm (#442) · 0e52a5c2
  TechxGenus authored Apr 13, 2024
  
  0e52a5c2
06 Apr, 2024 8 commits
- Add StableLM support (#410) · e9f62694
  Isotr0py authored Apr 06, 2024
```
Co-authored-by: Casper <casperbh.96@gmail.com>
```
  e9f62694
- add starcoder2 support (#406) · 33dfb048
  少年 authored Apr 06, 2024
```
Co-authored-by: charrli <charrli@tencent.com>
```
  33dfb048
- Add download_kwargs for load model (#302) (#399) · eb85f67d
  Roshiago authored Apr 06, 2024
  
  eb85f67d
- Workaround: illegal memory access (#421) · f8353793
  Casper authored Apr 06, 2024
  
  f8353793
- Implement `apply_clip` argument to `quantize()` (#427) · b5db7fcd
  Casper authored Apr 06, 2024
  
  b5db7fcd
- Pin: lm_eval==0.4.1 (#426) · c780d650
  Casper authored Apr 06, 2024
  
  c780d650
- FIX: Add safe guards for static cache + llama on transformers latest (#401) · 1f07200a
  Younes Belkada authored Apr 06, 2024
  
  1f07200a
- Fix fused models for tf >= 4.39 (#418) · 5d7b0502
  TechxGenus authored Apr 06, 2024
  
  5d7b0502
24 Mar, 2024 2 commits
- Bump to v0.2.4 (#409) · 0fa9a2c1
  Casper authored Mar 24, 2024
  
  0fa9a2c1
- Pin transformers>=4.35.0,<=4.38.2 (#408) · 0f942181
  Casper authored Mar 24, 2024
  
  0f942181
11 Mar, 2024 1 commit
- Add Gemma Support (#393) · 94e73f0b
  TechxGenus authored Mar 11, 2024
  
  94e73f0b
02 Mar, 2024 3 commits
- Bump to 0.2.3 · d8ca1e2f
  Casper authored Mar 02, 2024
  
  d8ca1e2f
- x_max -> x_mean and w_max -> w_mean name changes and some comments (#378) · f713b888
  Oscar Savolainen authored Mar 02, 2024
  
  f713b888
- Fix double bias (#383) · d9dc8e56
  Casper authored Mar 02, 2024
  
  d9dc8e56
24 Feb, 2024 1 commit
- New optimized kernels (#365) · 68c727a1
  Casper authored Feb 25, 2024
  
  68c727a1
17 Feb, 2024 2 commits
- Bump to 0.2.2 (#356) · 6b7992aa
  Casper authored Feb 17, 2024
  
  6b7992aa
- Remove MoE Triton kernels (#355) · a40515f4
  Casper authored Feb 17, 2024
  
  a40515f4
16 Feb, 2024 5 commits
- Add multi-GPU benchmark of Mixtral (#353) · d54bf0ef
  Casper authored Feb 16, 2024
  
  d54bf0ef
- Support Fused Mixtral on multi-GPU (#352) · 79b6fbd8
  Casper authored Feb 16, 2024
  
  79b6fbd8
- Bump to 0.2.1 (#351) · 74053104
  Casper authored Feb 16, 2024
  
  74053104
- Fix triton dependency (#350) · 8849043a
  Casper authored Feb 16, 2024
  
  8849043a
- ENH / FIX: Few enhancements and fix for mixed-precision training (#348) · 969b290e
  Younes Belkada authored Feb 16, 2024
  
  969b290e
15 Feb, 2024 7 commits
- Avoid downloading ROCm (#347) · 2de6092a
  Casper authored Feb 15, 2024
  
  2de6092a
- Fix typo in setup.py (#346) · d8311027
  Casper authored Feb 15, 2024
  
  d8311027
- Fix workflow (#345) · be0f9e01
  Casper authored Feb 15, 2024
  
  be0f9e01
- Fix importlib (#344) · 171c7aff
  Casper authored Feb 15, 2024
  
  171c7aff
- Fix dependency (#343) · 890127fa
  Casper authored Feb 15, 2024
  
  890127fa
- Fix ROCm build (#342) · 9477ea5c
  Casper authored Feb 15, 2024
  
  9477ea5c
- v0.2.0 (#330) · bcaa8a36
  Casper authored Feb 15, 2024
```
Co-authored-by: jinz2014 <7799920+jinz2014@users.noreply.github.com>
Co-authored-by: Jin Z <5zj@cousteau.ftpn.ornl.gov>
```
  bcaa8a36
14 Feb, 2024 1 commit
- [`PEFT`] Fix PEFT batch size > 1 (#338) · c69d3b65
  Younes Belkada authored Feb 14, 2024
  
  c69d3b65
03 Feb, 2024 1 commit
- PEFT compatible GEMM (#324) · 29ee66d9
  Casper authored Feb 03, 2024
  
  29ee66d9