Repository Issue Activity (beta)

#1405 Unsloth на Windows для fine-tuning.

#1395 Blackwell Support

#1378 xFormers does not support running on device with `torch.distributed.is_available() = False`

#1401 [feat] ops: add XFORMERS_VERBOSE_DISPATCH env-var to log which attention kernel is selected at runtime

#1400 [BUG] checkpoint: scipy not listed in requirements.txt - get_optimal_checkpoint_policy() fails at runtime with no install-time warning

#1399 module 'xformers.ops' has no attribute 'memory_efficient_attention'

#1394 chore(tests): linux error handling to be behind flags as in pytorch

#1261 Cuda 12.9 and Pytorch 2.8 support

#1369 Add a canonical long-context attention boundary benchmark (one command + metric)

#1342 Issue with 50-Series Graphics Card Compatibility for xFormers

#1345 [BUG] Xformers not support NO-GIL

#1393 我的电脑需要pytorch2.9.0+cu128 Python3.10版本对应的xformers，有哪位大佬可以教一下我要下载哪个版本吗

#1392 NaNs from CUTLASS GQA op

#1388 xformers support matrix needs update

#1386 我的电脑需要这个版本xFormers PyTorch 2.11.0+cu130 Python 3.12.12的whl文件有哪位大佬编译成功的吗

#1380 [Feature Request] Support for Ascend NPU

#1251 memory_efficient_attention fails with half precision tensors, with pre-compiled xformers 0.0.30 on RTX 5090

#1356 requires device with capability < (8, 0) but your GPU has capability (12, 0) (too new)

#1314 xformers-0.0.32.post2 ROCm 6.4 with PyTorch 2.8.0 Aborted (core dumped) In WSL2

#1373 Status of CPU-only builds?

#1374 0.0.34 claims to support PyTorch>=2.10 but it doesn't?

#990 Customization of BlockDiagonalMask or BlockDiagonalCausalMask

#1371 Geometric sparse attention model

#1372 Provide official prebuilt CUDA-compatible wheels and fix common build/install failures for xFormers

#1367 Is there any expert who has successfully compiled the .whl file of xFormers for PyTorch 2.8.0+cu129 and Python 3.12.9, which is what my computer needs?