bitsandbytes-foundation bitsandbytes issues

bitsandbytes-foundation / bitsandbytes

Accessible large language models via k-bit quantization for PyTorch.

https://huggingface.co/docs/bitsandbytes/main/en/index

MIT License

6.31k stars 634 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

torch.chunks(Params4bit) not returning Params4bit subclass

#1424 mreso opened 16 hours ago
0
python -m bitsandbytes ·

#1423 gunnusravani opened 1 day ago
0
Add Ascend NPU support for nf4 quant

#1422 statelesshz opened 1 day ago
3
Installed with cuda support but doesn't detect gpu runtime

#1421 mahmoodn opened 2 days ago
2
Fix invalid escape sequence warning in Python 3.12

#1420 oshiteku closed 2 days ago
2
Fix issue that no valid semantic version tag found when installing from source via personal repo

#1419 statelesshz closed 2 days ago
2
Enable XPU and optimize cpu/xpu op

#1418 jiqing-feng opened 3 days ago
4
Source Code for `cadam_8bit_blockwise_grad_bf16`

#1417 ionutmodo opened 3 days ago
0
[WIP] Add npu support for nf4 quantization

#1416 statelesshz closed 4 days ago
0
[Question] How to manually quantize every linear of a given model?

#1415 AaronZLT opened 1 week ago
0
Support for quantization of convolutional layers

#1414 JohnnyRacer opened 1 week ago
0
Remove triton.ops, copy necessary bits here

#1413 bertmaher opened 1 week ago
0
Integration of BitNet into bitsandbytes

#1412 ParagEkbote closed 20 hours ago
1
Device_map='auto' not working along with bitsandbytes (transformers)

#1411 guillemram97 closed 2 days ago
3
Update wheel requirement from ~=0.44.0 to ~=0.45.0

#1410 dependabot[bot] closed 2 days ago
2
It seems that the current version of bitsandbytes is not compatible with my CUDA 12.4 library.

#1409 Yukiiceeee opened 1 week ago
0
Rewrite `assert`s as exceptions

#1408 akx opened 2 weeks ago
0
Support

#1407 Qarqor5555555 opened 3 weeks ago
0
bitsandbytes-0.44.1.dev0-py3-none-macosx_13_1_arm64.whl is not a supported wheel on this platform.

#1406 wuhongsheng opened 3 weeks ago
0
Error when saving FSDP weights with cpu_offload=True [rank1]: AttributeError: 'Params4bit' object has no attribute 'absmax'

#1405 PanagiotisFytas opened 3 weeks ago
1
CUDA Setup failed despite CUDA being available.

#1404 N1ana7mi opened 3 weeks ago
1
unable to run FSDP2 with low bit optimizers like adam 8 bit

#1403 nighting0le01 opened 3 weeks ago
0
Support running on CPU

#1402 fzyzcjy opened 4 weeks ago
2
LLM.int8() Refactoring: Part 1

#1401 matthewdouglas opened 4 weeks ago
3
Questions about the details of LLM.int8

#1400 bg51717 closed 3 weeks ago
5
[Windows] [Error] Compile from source

#1398 blap closed 3 days ago
1
Update pandas requirement from ~=2.2.2 to ~=2.2.3

#1397 dependabot[bot] closed 4 weeks ago
2
too large numeric difference with pytorch inference

#1396 weixsong opened 1 month ago
0
Remove depth option in installation steps

#1395 pnunna93 closed 1 month ago
0
CUDA Setup failed despite GPU being available

#1394 carlxc911 opened 1 month ago
3
[Feature]Cutlass (re) integration

#1393 abhilash1910 opened 1 month ago
0
I ran a NF4 72B model in 2xA6000 using llamafactory

#1392 charleswg opened 1 month ago
0
CUDA Architecture 80+ Causing Incorrect Model Behavior with BitsAndBytes Quantization

#1391 gunjunlee opened 1 month ago
0
ARM Runners December 2024

#1390 johnnynunez opened 1 month ago
0
Available version for 4-bit quantization supporting CUDA 11.1?

#1388 justin4ai closed 3 weeks ago
0
Failed to import transformers.integrations.bitsandbytes because of the following error (look up to see its traceback): CUDA Setup failed despite GPU being available. Please run the following command to get more information: python -m bitsandbytes Inspect the output of the command and see if you can locate CUDA libraries. You might need to add them to your LD_LIBRARY_PATH. If you suspect a bug, please take the information from python -m bitsandbytes and open an issue at: https://github.com/TimDettmers/bitsandbytes/issues

#1387 smillpine opened 1 month ago
0
Update CI tools & fix typos

#1386 akx closed 4 weeks ago
1
README: Replace special Unicode text symbols with regular characters

#1385 akx closed 1 month ago
2
An error occurred: CUDA is required but not available for bitsandbytes.

#1384 GaoDalie opened 1 month ago
0
[TESTS] Add global random seed for deterministic test

#1383 allrob23 opened 1 month ago
0
AdEMA NaN when loading from state_dict

#1382 darius-lam opened 1 month ago
1
Paged optimizer resuming from checkpoint - attributeError: 'int' object has no attribute 'cpu'

#1381 shivam15s opened 1 month ago
2
refine docs for multi-backend alpha release

#1380 Titus-von-Koeller closed 1 month ago
1
Fix optimizer support for Python <= 3.9

#1379 matthewdouglas closed 1 month ago
2
Release v44 not available for Mac

#1378 ACMCMC opened 1 month ago
5
Update pandas requirement from ~=2.2.2 to ~=2.2.3 in the major group

#1377 dependabot[bot] closed 1 month ago
2
Python 3.9 support broken in 0.44.0

#1376 Benzhaomin closed 1 month ago
0
chore: migrate config files to `pyproject.toml`

#1373 SauravMaheshkar opened 1 month ago
2
replace tl.libdevice.llrint with tl.extra.cuda.libdevice.rint

#1372 bjmsong opened 1 month ago
0
Model architecture is modified when I use BitsAndBytesConfig with default params

#1371 yunhao-tech opened 1 month ago
0