Switch to cutlass 3.1, disable nvcc --threads

tridao · Aug 14, 2023 · 5e005c0 · 5e005c0
1 parent 922fcda
commit 5e005c0
Show file tree

Hide file tree

Showing 4 changed files with 9 additions and 5 deletions.
diff --git a/.github/workflows/publish.yml b/.github/workflows/publish.yml
@@ -91,8 +91,12 @@ jobs:
 
       - name: Free up disk space
         if: ${{ runner.os == 'Linux' }}
+        # https://github.com/easimon/maximize-build-space/blob/master/action.yml
+        # https://github.com/easimon/maximize-build-space/tree/test-report
         run: |
           sudo rm -rf /usr/share/dotnet
+          sudo rm -rf /opt/ghc
+          sudo rm -rf /opt/hostedtoolcache/CodeQL
 
       - name: Install CUDA ${{ matrix.cuda-version }}
         if: ${{ matrix.cuda-version != 'cpu' }}

diff --git a/csrc/cutlass b/csrc/cutlass
diff --git a/flash_attn/__init__.py b/flash_attn/__init__.py
@@ -1,4 +1,4 @@
-__version__ = "2.0.6.post3"
+__version__ = "2.0.6.post4"
 
 from flash_attn.flash_attn_interface import flash_attn_func
 from flash_attn.flash_attn_interface import flash_attn_kvpacked_func

diff --git a/setup.py b/setup.py
@@ -91,9 +91,9 @@ def raise_if_cuda_home_none(global_option: str) -> None:
 
 
 def append_nvcc_threads(nvcc_extra_args):
-    _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
-    if bare_metal_version >= Version("11.2"):
-        return nvcc_extra_args + ["--threads", "4"]
+    # _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+    # if bare_metal_version >= Version("11.2"):
+    #     return nvcc_extra_args + ["--threads", "4"]
     return nvcc_extra_args