From 53e193e2c2d5f5a4e89f630e92150028b6e4b669 Mon Sep 17 00:00:00 2001
From: leasunhy <leasunhy@gmail.com>
Date: Sun, 23 Jun 2024 16:35:15 +0000
Subject: [PATCH] fix docker build

---
 .github/workflows/docker_rdma.yml        | 23 +++++++++++++++++++++-
 .github/workflows/docker_rdma_latest.yml | 25 ++++++++++++++++++++++--
 docker/rdma/Dockerfile                   |  4 ++--
 3 files changed, 47 insertions(+), 5 deletions(-)

diff --git a/.github/workflows/docker_rdma.yml b/.github/workflows/docker_rdma.yml
index d08db5c..6c03cad 100644
--- a/.github/workflows/docker_rdma.yml
+++ b/.github/workflows/docker_rdma.yml
@@ -25,7 +25,28 @@ jobs:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_TOKEN }}
       -
-        name: Build and push cu117 with rdma
+        name: Free Disk Space (Ubuntu)
+        uses: jlumbroso/free-disk-space@main
+        with:
+          # this might remove tools that are actually needed,
+          # if set to "true" but frees about 6 GB
+          tool-cache: false
+
+          # all of these default to true, but feel free to set to
+          # "false" if necessary for your workflow
+          android: true
+          dotnet: true
+          haskell: true
+          large-packages: true
+          docker-images: false
+          swap-storage: false
+      -
+        name: Set up swap space
+        uses: pierotofy/set-swap-space@v1.0
+        with:
+          swap-size-gb: 10
+      -
+        name: Build and push with rdma (nightly)
         uses: docker/build-push-action@v3
         with:
           context: ./docker/rdma/
diff --git a/.github/workflows/docker_rdma_latest.yml b/.github/workflows/docker_rdma_latest.yml
index 16481f2..ff01a49 100644
--- a/.github/workflows/docker_rdma_latest.yml
+++ b/.github/workflows/docker_rdma_latest.yml
@@ -8,7 +8,7 @@ on:
 
 jobs:
   docker:
-    runs-on: self-hosted
+    runs-on: ubuntu-latest
     steps:
       -
         name: Checkout
@@ -26,7 +26,28 @@ jobs:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_TOKEN }}
       -
-        name: Build and push cu12.4 with rdma
+        name: Free Disk Space (Ubuntu)
+        uses: jlumbroso/free-disk-space@main
+        with:
+          # this might remove tools that are actually needed,
+          # if set to "true" but frees about 6 GB
+          tool-cache: false
+
+          # all of these default to true, but feel free to set to
+          # "false" if necessary for your workflow
+          android: true
+          dotnet: true
+          haskell: true
+          large-packages: true
+          docker-images: false
+          swap-storage: false
+      -
+        name: Set up swap space
+        uses: pierotofy/set-swap-space@v1.0
+        with:
+          swap-size-gb: 10
+      -
+        name: Build and push with rdma (nightly)
         uses: docker/build-push-action@v3
         with:
           context: ./docker/rdma/
diff --git a/docker/rdma/Dockerfile b/docker/rdma/Dockerfile
index 32e0896..0cf248b 100644
--- a/docker/rdma/Dockerfile
+++ b/docker/rdma/Dockerfile
@@ -56,7 +56,7 @@ RUN APT_INSTALL="apt-get install -y --no-install-recommends" && \
 # # ------------------------------------------------------------------
 ENV TORCH_CUDA_ARCH_LIST "7.0;7.5;8.0;9.0"
 
-RUN pip3 install --upgrade sentry-sdk requests
+RUN pip3 install --no-cache-dir --upgrade sentry-sdk requests ninja typing packaging && rm -rf ~/.cache/pip
 
 RUN cd /tmp && \
     git clone https://github.com/dptech-corp/Uni-Core && \
@@ -66,7 +66,7 @@ RUN cd /tmp && \
 
 RUN pip3 install --no-cache-dir tokenizers lmdb biopython ml-collections timeout-decorator urllib3 tree dm-tree && rm -rf ~/.cache/pip 
 
-RUN MAX_JOBS=4 pip3 install -U 'flash-attn<2.5.0' --no-build-isolation --no-cache-dir
+RUN MAX_JOBS=1 pip3 install -U 'flash-attn<2.5.0' --no-build-isolation --no-cache-dir
 
 RUN ldconfig && \
     apt-get clean && \