facebookresearch/xformers

Issues

• Can't install xformers

  #1148 opened 7 days ago by GUEST-1001
  0

• Are there any version for Torch 2.5 including dev? like 0.0.29.dev921?

  #1121 opened 8 days ago by FurkanGozukara
  6

• Customization of BlockDiagonalMask or LowerTriangularMask

  #1147 opened 8 days ago by kc334
  0

• Can I use xformers with torch-2.4.1?

  #1146 opened 9 days ago by LukeLIN-web
  1

• scaled_dot_product_attention output is different from memory_efficient_attention

  #1114 opened 2 months ago by aenoca
  1

• [Bug] Unexpected behavior of `memory_efficient_attention` with `BlockDiagonalMask`

  #1122 opened a month ago by xiangxu-google
  1

• incorrect causal mask in global attention

  #1135 opened a month ago by davidqqq
  1

• Is there an efficient way to use memory_efficient_attention with a causal mask that has a small rectangle of zeros?

  #1131 opened a month ago by arilato
  1

• RuntimeError: CUDA error: CUBLAS_STATUS_NOT_INITIALIZED when calling `cublasCreate(handle)`

  #1144 opened 13 days ago by antony-frolov
  1

• build without flash attn?

  #1145 opened 14 days ago by sipie800
  1

• Cant use powershell command to build for Nightly PyTorch on Win - filename too long

  #1132 opened a month ago by Mescalamba
  5

• Generalization of dual_gemm_silu_identity_mul to use custom activation function, not only SiLU

  #1140 opened 16 days ago by warpuv
  0

• # ❓ Questions and Help NotImplementedError: No operator found for `memory_efficient_attention_forward` with inputs:

  #1143 opened 16 days ago by Was-11-wendy
  1

• How to install to aarch64

  #1142 opened 17 days ago by gongchangsui
  1

• Hey, you! The people that made this thing. You know that Torch v 2.4.1 is out, right??

  #1096 opened 2 months ago by Xephier102
  11

• torch2.5.1 support

  #1137 opened 21 days ago by luohao123
  4

• Cann't install xformers==0.0.28.post2

  #1139 opened 21 days ago by Wiselnn570
  1

• vllm 0.6.3 createLLM error TypeError: autotune() got an unexpected keyword argument 'use_cuda_graph' on windows

  #1138 opened 22 days ago by xiezhipeng-git
  4

• Activation checkpointing is not working on SwiGLU

  #1126 opened 23 days ago by warpuv
  0

• timm.models.layers is deprecated/resume_download is deprecated

  #1134 opened a month ago by zd391
  1

• AMD64 version of Xformers

  #1130 opened a month ago by pandayummy
  2

• AttributeError: module 'xformers.ops' has no attribute 'AttentionOpDispatch'

  #1129 opened a month ago by LarsDoorenbos
  1

• I tried upgrade visual studio,cuda 12.4 and 12.6,but it still can't build the wheel,and return this

  #1117 opened 2 months ago by neutronslime
  5

• `ScaledDotProduct` with attention mask returns different result as standard attention

  #1089 opened a month ago by amyxlu
  3

• xformers 0.0.20 memoery efficient attent cutalss backward non deterministic

  #1128 opened a month ago by Bavesh-B
  0

• xformers.sparse.utils._coo_to_csr is incorrect when n > m

  #1125 opened a month ago by francois-rozet
  0

• Incorrect attention output with SparseCS mask

  #1124 opened a month ago by francois-rozet
  1

• Add support for CUDNN attention via CUDNN_FRONTEND Python API?

  #1123 opened a month ago by Skylion007
  0

• CUTLASS Fused multi head attention

  #1112 opened 2 months ago by yoon5862
  2

• 🚀 Precompiled xFormers for CUDA 12.4 and PyTorch 2.4 Compatibility

  #1079 opened 3 months ago by sashaok123
  14

• Why xformers 0.0.28.post1 doesn't have pre-compiled wheel for Windows?

  #1118 opened 2 months ago by FurkanGozukara
  7

• Does memory efficient attention cutlass kernel support various seq len inputs for q/k/v + tensor bias?

  #1120 opened 2 months ago by ShijunK
  0

• I cannot run xformers on an AMD Radeon RX 6600 with torch 2.4.1

  #1116 opened 2 months ago by renich
  2

• no module 'xformers'. Processing without...

  #1102 opened 2 months ago by ZeroCool22
  2

• Enabling softcap option

  #1113 opened 2 months ago by SpyrosMouselinos
  1

• what version about torch in FLUX bf16？

  #1111 opened 2 months ago by Zhuangvictor0
  1

• compile for rocm w/ gfx1032 card

  #1110 opened 2 months ago by brcisna
  4

• unable to pip install 0.0.28.post1 on windows

  #1104 opened 2 months ago by ajkessel
  5

• No operator found for `memory_efficient_attention_forward` with inputs:

  #1109 opened 2 months ago by brcisna
  1

• `memory_efficient_attention` is slower than `scaled_dot_product_attention` of PyTorch?

  #1107 opened 2 months ago by QinlongHuang
  2

• Prebuilt wheel for Windows

  #1098 opened 2 months ago by KohakuBlueleaf
  3

• xformers Installation failed

  #1095 opened 2 months ago by wzgrx
  1

• why conflicting release notes?

  #1097 opened 2 months ago by BBC-Esq
  1

• BlockDiagonalGappyKeysMask backward support

  #1093 opened 3 months ago by cumulo-autumn
  2

• flash attn bug

  #1090 opened 3 months ago by zhw-zhang
  0

• Sparse attention will not reduce peak memory usage

  #1086 opened 3 months ago by ThisisBillhe
  2

• Error in [conda install xformers -c xformers/label/dev]

  #1084 opened 3 months ago by 11whitewater
  6

• Build with CUDA 12.6 and VS 2022 17.11.1 broken on Windows

  #1083 opened 3 months ago by levicki
  14

• How to get Q @ K^T similarity?

  #1080 opened 3 months ago by volcverse
  0

• Local attention mask size mismatch

  #1076 opened 4 months ago by samuelwheeler
  1