jalr/tabbyAPI-ollama
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Dependencies: Update Flash Attention 2

Browse source 
v2.6.3 with torch 2.3 wheels.

Signed-off-by: kingbri <bdashore3@proton.me>
This commit is contained in:
kingbri 2024-07-28 13:50:15 -04:00
parent c79e0832d5
commit d85414738d
1 changed files with 9 additions and 9 deletions
Download patch file Download diff file Expand all files Collapse all files

18
pyproject.toml
View file

@ -70,14 +70,14 @@ cu121 = [
"exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.1.8/exllamav2-0.1.8+cu121.torch2.3.1-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
# Windows FA2 from https://github.com/bdashore3/flash-attention/releases
"flash_attn @ https://github.com/bdashore3/flash-attention/releases/download/v2.6.1/flash_attn-2.6.1+cu123torch2.3.1cxx11abiFALSE-cp312-cp312-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.12'",
"flash_attn @ https://github.com/bdashore3/flash-attention/releases/download/v2.6.1/flash_attn-2.6.1+cu123torch2.3.1cxx11abiFALSE-cp311-cp311-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.11'",
"flash_attn @ https://github.com/bdashore3/flash-attention/releases/download/v2.6.1/flash_attn-2.6.1+cu123torch2.3.1cxx11abiFALSE-cp310-cp310-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.10'",
"flash_attn @ https://github.com/bdashore3/flash-attention/releases/download/v2.6.3/flash_attn-2.6.3+cu123torch2.3.1cxx11abiFALSE-cp312-cp312-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.12'",
"flash_attn @ https://github.com/bdashore3/flash-attention/releases/download/v2.6.3/flash_attn-2.6.3+cu123torch2.3.1cxx11abiFALSE-cp311-cp311-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.11'",
"flash_attn @ https://github.com/bdashore3/flash-attention/releases/download/v2.6.3/flash_attn-2.6.3+cu123torch2.3.1cxx11abiFALSE-cp310-cp310-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.10'",
# Linux FA2 from https://github.com/Dao-AILab/flash-attention/releases
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.1/flash_attn-2.6.1+cu123torch2.3cxx11abiFALSE-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.12'",
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.1/flash_attn-2.6.1+cu123torch2.3cxx11abiFALSE-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.1/flash_attn-2.6.1+cu123torch2.3cxx11abiFALSE-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.3/flash_attn-2.6.3+cu123torch2.3cxx11abiFALSE-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.12'",
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.3/flash_attn-2.6.3+cu123torch2.3cxx11abiFALSE-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.3/flash_attn-2.6.3+cu123torch2.3cxx11abiFALSE-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
]
cu118 = [
# Torch
@ -97,9 +97,9 @@ cu118 = [
"exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.1.8/exllamav2-0.1.8+cu118.torch2.3.1-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
# Linux FA2 from https://github.com/Dao-AILab/flash-attention/releases
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.1/flash_attn-2.6.1+cu118torch2.3cxx11abiFALSE-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.12'",
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.1/flash_attn-2.6.1+cu118torch2.3cxx11abiFALSE-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.1/flash_attn-2.6.1+cu118torch2.3cxx11abiFALSE-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.3/flash_attn-2.6.3+cu118torch2.3cxx11abiFALSE-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.12'",
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.3/flash_attn-2.6.3+cu118torch2.3cxx11abiFALSE-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
"flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.6.3/flash_attn-2.6.3+cu118torch2.3cxx11abiFALSE-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
]
amd = [
# Torch triton for ROCm