| .. |
|
__init__.py
|
9d81716bfd
[v0.5.3] Release Candidate (#388)
|
10 mesi fa |
|
abstract.py
|
1390915778
multi-step: add support for flashinfer attention backend (#1033)
|
2 mesi fa |
|
blocksparse_attn.py
|
1405051912
attention: add `AttentionState` abstraction (#863)
|
3 mesi fa |
|
flash_attn.py
|
1390915778
multi-step: add support for flashinfer attention backend (#1033)
|
2 mesi fa |
|
flashinfer.py
|
c951a54d21
fix: multi-step + flashinfer with cuda graphs (#1036)
|
2 mesi fa |
|
ipex_attn.py
|
6951928522
xpu: bump IPEX to 2.3, support GQA (#1042)
|
2 mesi fa |
|
openvino.py
|
1405051912
attention: add `AttentionState` abstraction (#863)
|
3 mesi fa |
|
pallas.py
|
032974a28a
tpu: fix TPU type api (#975)
|
2 mesi fa |
|
placeholder_attn.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
2 mesi fa |
|
rocm_flash_attn.py
|
4a7cb8f232
rocm: add custom paged attention kernels for ROCm (#1043)
|
2 mesi fa |
|
torch_sdpa.py
|
1405051912
attention: add `AttentionState` abstraction (#863)
|
3 mesi fa |
|
utils.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
2 mesi fa |
|
xformers.py
|
1405051912
attention: add `AttentionState` abstraction (#863)
|
3 mesi fa |
AlpinDale
4a7cb8f232
rocm: add custom paged attention kernels for ROCm (#1043)