| .. |
|
__init__.py
|
9d81716bfd
[v0.5.3] Release Candidate (#388)
|
há 8 meses atrás |
|
abstract.py
|
7a313483f1
chore: move update_flash_attn_metadata to attn backend (#731)
|
há 4 meses atrás |
|
blocksparse_attn.py
|
f1d0b77c92
[0.6.0] Release Candidate (#481)
|
há 4 meses atrás |
|
flash_attn.py
|
7a313483f1
chore: move update_flash_attn_metadata to attn backend (#731)
|
há 4 meses atrás |
|
flashinfer.py
|
60b702a827
chore: register custom torch ops for flash-attn and flashinfer (#724)
|
há 4 meses atrás |
|
ipex_attn.py
|
f1d0b77c92
[0.6.0] Release Candidate (#481)
|
há 4 meses atrás |
|
openvino.py
|
f1d0b77c92
[0.6.0] Release Candidate (#481)
|
há 4 meses atrás |
|
pallas.py
|
f1d0b77c92
[0.6.0] Release Candidate (#481)
|
há 4 meses atrás |
|
placeholder_attn.py
|
bf88c8567e
feat: mamba model support (#674)
|
há 4 meses atrás |
|
rocm_flash_attn.py
|
e200775863
feat: enable using fp8 kv and prefix caching with chunked prefill (#668)
|
há 4 meses atrás |
|
torch_sdpa.py
|
f1d0b77c92
[0.6.0] Release Candidate (#481)
|
há 4 meses atrás |
|
utils.py
|
3bbb3f2086
feat: add numpy implementation of `compute_slot_mapping` (#678)
|
há 4 meses atrás |
|
xformers.py
|
e200775863
feat: enable using fp8 kv and prefix caching with chunked prefill (#668)
|
há 4 meses atrás |
AlpinDale
7a313483f1
chore: move update_flash_attn_metadata to attn backend (#731)