Dependencies: Update flash-attention

v2.6.1

Signed-off-by: kingbri <[email protected]>

pyproject.toml

@@ -66,14 +66,14 @@ cu121 = [
     "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.1.7/exllamav2-0.1.7+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/v2.5.9.post1/flash_attn-2.5.9.post1+cu122torch2.3.1cxx11abiFALSE-cp312-cp312-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.12'",
-    "flash_attn @ https://github.com/bdashore3/flash-attention/releases/download/v2.5.9.post1/flash_attn-2.5.9.post1+cu122torch2.3.1cxx11abiFALSE-cp311-cp311-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.11'",
-    "flash_attn @ https://github.com/bdashore3/flash-attention/releases/download/v2.5.9.post1/flash_attn-2.5.9.post1+cu122torch2.3.1cxx11abiFALSE-cp310-cp310-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.10'",
+    "flash_attn @ https://github.com/bdashore3/flash-attention/releases/download/v2.6.1/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'",
 
     # Linux FA2 from https://github.com/Dao-AILab/flash-attention/releases
-    "flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.5.9.post1/flash_attn-2.5.9.post1+cu122torch2.3cxx11abiFALSE-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/v2.5.9.post1/flash_attn-2.5.9.post1+cu122torch2.3cxx11abiFALSE-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/v2.5.9.post1/flash_attn-2.5.9.post1+cu122torch2.3cxx11abiFALSE-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/v2.6.1/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'",
 ]
 cu118 = [
     # Torch
@@ -93,9 +93,9 @@ cu118 = [
     "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.1.7/exllamav2-0.1.7+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/v2.5.9.post1/flash_attn-2.5.9.post1+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/v2.5.9.post1/flash_attn-2.5.9.post1+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/v2.5.9.post1/flash_attn-2.5.9.post1+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/v2.6.1/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'",
 ]
 amd = [
     # Torch triton for ROCm