tocmo0nlord/axolotl
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

add runpod envs to .bashrc, fix bnb env (#316)

Browse Source 
* hopper support for base dockerfile, add runpod envs to .bashrc

* set BNB_CUDA_VERSION env for latest bnb

* don't support hopper yet w 118
This commit is contained in:
Wing Lian
2023-07-22 10:09:38 -04:00
committed by GitHub
parent c5df969262
commit cf62cfd661
3 changed files with 24 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
.github/workflows/main.yml vendored
View File

@@ -69,22 +69,22 @@ jobs:
strategy: strategy:
matrix: matrix:
include: include:
- cuda: cu118 - cuda: 118
cuda_version: 11.8.0 cuda_version: 11.8.0
python_version: "3.9" python_version: "3.9"
pytorch: 2.0.1 pytorch: 2.0.1
axolotl_extras: axolotl_extras:
- cuda: cu118 - cuda: 118
cuda_version: 11.8.0 cuda_version: 11.8.0
python_version: "3.10" python_version: "3.10"
pytorch: 2.0.1 pytorch: 2.0.1
axolotl_extras: axolotl_extras:
- cuda: cu118 - cuda: 118
cuda_version: 11.8.0 cuda_version: 11.8.0
python_version: "3.9" python_version: "3.9"
pytorch: 2.0.1 pytorch: 2.0.1
axolotl_extras: gptq axolotl_extras: gptq
- cuda: cu117 - cuda: 117
cuda_version: 11.7.1 cuda_version: 11.7.1
python_version: "3.9" python_version: "3.9"
pytorch: 1.13.1 pytorch: 1.13.1
@@ -110,8 +110,9 @@ jobs:
with: with:
context: . context: .
build-args: | build-args: |
BASE_TAG=${{ github.ref_name }}-py${{ matrix.python_version }}-${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }} BASE_TAG=${{ github.ref_name }}-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
CUDA=${{ matrix.cuda }}
file: ./docker/Dockerfile-runpod file: ./docker/Dockerfile-runpod
push: ${{ github.event_name != 'pull_request' }} push: ${{ github.event_name != 'pull_request' }}
tags: ${{ steps.metadata.outputs.tags }}-py${{ matrix.python_version }}-${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }} tags: ${{ steps.metadata.outputs.tags }}-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
labels: ${{ steps.metadata.outputs.labels }} labels: ${{ steps.metadata.outputs.labels }}

2
docker/Dockerfile
View File

@@ -3,6 +3,8 @@ FROM winglian/axolotl-base:$BASE_TAG
ARG TORCH_CUDA_ARCH_LIST="7.0 7.5 8.0 8.6+PTX" ARG TORCH_CUDA_ARCH_LIST="7.0 7.5 8.0 8.6+PTX"
ARG AXOLOTL_EXTRAS="" ARG AXOLOTL_EXTRAS=""
ARG CUDA="118"
ENV BNB_CUDA_VERSION=$CUDA
RUN apt-get update && \ RUN apt-get update && \
apt-get install -y vim curl apt-get install -y vim curl

19
scripts/runpod-entrypoint.sh Normal file → Executable file
View File

@@ -1,10 +1,21 @@
#!/bin/bash #!/bin/bash
echo $PUBLIC_KEY >> ~/.ssh/authorized_keys # Export specific ENV variables to /etc/rp_environment
chmod 700 -R ~/.ssh echo "Exporting environment variables..."
printenv | grep -E '^RUNPOD_|^PATH=|^_=' | sed 's/^\(.*\)=\(.*\)$/export \1="\2"/' >> /etc/rp_environment
echo 'source /etc/rp_environment' >> ~/.bashrc
# Start the SSH service in the background if [[ $PUBLIC_KEY ]]
service ssh start then
mkdir -p ~/.ssh
chmod 700 ~/.ssh
echo $PUBLIC_KEY >> ~/.ssh/authorized_keys
chmod 700 -R ~/.ssh
# Start the SSH service in the background
service ssh start
else
echo "No PUBLIC_KEY ENV variable provided, not starting openSSH daemon"
fi
# Execute the passed arguments (CMD) # Execute the passed arguments (CMD)
exec "$@" exec "$@"