Skip to content

Add int4 implementation on top of the bitstandbytes

active 2023-12-132024-07-10 (UTC)

Partial coverage18,243 / 23,362 hourly files (78%) · 2 absent upstream · 5,117 failed, retryable2023-12-132026-08-12 (UTC)— sampled evenly across the window, so rankings and trends hold; absolute counts scale up.
Events
2.6K
Pushes
51
Pull requests
156
Issues
570
Stars
735
Forks
89

Activity over time

Daily event counts in the loaded window

Line chart, 211 days from 2023-12-13 to 2024-07-10. Pushes: 51 total, peak 11 in a day. Pull requests: 156 total, peak 40 in a day. Issues: 570 total, peak 212 in a day. Comments: 863 total, peak 55 in a day. Stars: 735 total, peak 12 in a day.

  • Pushes
  • Pull requests
  • Issues
  • Comments
  • Stars

Top contributors

Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity

ContributorContributionsPushesPRsComments
github-actions[bot]51003245
TimDettmers153184384
Titus-von-Koeller142242667
akx1310685
younesbelkada11691467
rickardp730541
matthewdouglas560438
wkpark410327
jiqing-feng250117
Xia-Weiwen180213
jgong5150011
abhilash1910140010
arlo-phoenix110110
josemerinom110011
jianan-gu9006
dailingx9006
pnunna938035
stas007006
ehartford6006
MuhammadShifa6006

Recent activity

Latest issues, pull requests and releases

  • Issue comment#1258ClaireCJS2024-07-09 00:38
    error message not meaningful enough — why is it looking in folders that are not in my path and are unrelated environment variables?
  • Issue comment#306xiaocai20112024-07-08 01:07
    RuntimeError: Unrecognized CachingAllocator option: max_split_size_mb=8000
  • Issue comment#107katanazero862024-07-07 12:24
    Feature Request: ROCm support (AMD GPU)
  • Issue comment#1202etemiz2024-07-06 17:51
    AttributeError: 'NoneType' object has no attribute 'split' CUDA Setup failed despite CUDA being available.
  • Issue comment#490Oxi842024-07-05 23:27
    load_in_8bit very slow and slower than load_in_4bit now?
  • Pull request#1270Ther-nullptr2024-07-05 11:28
  • Issue comment#1093spydaz2024-07-05 06:35
    RuntimeError: Failed to import transformers.integrations.bitsandbytes because of the following error (look up to see its traceback):
  • Issue comment#614spydaz2024-07-04 10:51
    What is the latest version of bitsandbytes supported on Windows Python 3.10.x with cuda support?
  • Issue comment#1085spydaz2024-07-04 10:33
    ERROR Installing bitsandbytes on windows
  • Issue comment#1264dami10182024-07-04 07:01
    Regarding the issue of bitsandbytes package supporting cuda12.4 version
  • Issue comment#1251Xia-Weiwen2024-07-04 03:30
    CPU/XPU: disable torch.compile if g++ is not available
  • Issue comment#1258matthewdouglas2024-07-02 23:42
    error message not meaningful enough — why is it looking in folders that are not in my path and are unrelated environment variables?
  • Issue comment#1264matthewdouglas2024-07-02 23:35
    Regarding the issue of bitsandbytes package supporting cuda12.4 version
  • Issue#1268wizardforcel2024-07-02 04:03
    AttributeError: 'Parameter' object has no attribute 'SCB'
  • Issue comment#216RudraChopra2024-07-02 02:04
    BUG REPORT
  • Issue comment#12560wwafa2024-07-01 23:40
    [QUESTION] Quantizing in a different way...
  • Issue comment#591saramirabi2024-07-01 04:43
    ImportError: cannot import name 'translate_llvmir_to_hsaco' from 'triton._C.libtriton.triton'
  • Issue#1262Worromots2024-06-30 16:25
    quantization of T5 faild. int8 model cost more inference time and memory.
  • Issue#1261gau-nernst2024-06-30 06:19
    Wrong doc and function signature for 8-bit optim
  • Issue#1260artsera2024-06-29 18:35
    bitsandbytes broken with regards to cuda with name error 'str2optimizer32bit' and name 'str2optimizer8bit_blockwise' is not defined
  • Issue comment#1251Xia-Weiwen2024-06-26 00:42
    CPU/XPU: disable torch.compile if g++ is not available
  • Issue comment#653Eikwang2024-06-25 04:37
    CUDA Setup failed despite GPU being available. Please run the following command to get more information:
  • Issue#12560wwafa2024-06-23 18:06
    [QUESTION] Quantizing in a different way...
  • Issue comment#1194weilingfeng982024-06-23 13:39
    error on VectorstoreIndexCreator
  • Issue comment#1254alexgit2562024-06-22 09:03
    AssertionError: Torch not compiled with CUDA enabled

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 735 stars here means stars gained during the window, not the repo's star count.