From c61aca407415e0c35020e78011b646bf274f5964 Mon Sep 17 00:00:00 2001 From: Debian Date: Sat, 3 Jan 2026 20:17:10 +0000 Subject: [PATCH] Reduce build parallelism to avoid OOM during SageAttention compile MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 --- Dockerfile | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/Dockerfile b/Dockerfile index 14685f6..76f7080 100644 --- a/Dockerfile +++ b/Dockerfile @@ -62,9 +62,9 @@ RUN pip install -r /tmp/requirements.txt && rm -rf /root/.cache/pip # Compile SageAttention 2.2 from source with no build isolation WORKDIR /tmp -ENV EXT_PARALLEL=2 -ENV NVCC_APPEND_FLAGS="--threads 2" -ENV MAX_JOBS=4 +ENV EXT_PARALLEL=1 +ENV NVCC_APPEND_FLAGS="--threads 1" +ENV MAX_JOBS=2 # Target RunPod GPU architectures: # 8.0 = A100, 8.6 = A10/RTX 3090, 8.9 = RTX 4090/L40, 9.0 = H100/H200 # Note: Blackwell (10.0) not yet supported by SageAttention