Dockerfile.ci

# syntax=docker/dockerfile:1.3-labs

ARG FROM_IMAGE_NAME
FROM $FROM_IMAGE_NAME as build_causal_conv1d
WORKDIR /opt
RUN CAUSAL_CONV1D_FORCE_BUILD=TRUE pip3 wheel -v git+https://github.com/Dao-AILab/causal-conv1d.git@v1.2.2.post1

FROM $FROM_IMAGE_NAME as build_grouped_gemm
WORKDIR /opt
RUN pip3 wheel -v git+https://github.com/fanshiqing/grouped_gemm@v1.1.2

FROM $FROM_IMAGE_NAME as build_mamba_ssm
WORKDIR /opt
RUN MAMBA_FORCE_BUILD=TRUE pip3 wheel -v git+https://github.com/state-spaces/mamba.git@v2.0.3

FROM $FROM_IMAGE_NAME as main
ENV DEBIAN_FRONTEND=noninteractive

RUN apt-get update && \
    apt-get install -y --no-install-recommends gettext python3-venv && \
    apt-get clean && \
    python -m venv /opt/jet && \
    wget https://github.com/mikefarah/yq/releases/download/v4.44.1/yq_linux_amd64 -O /usr/local/bin/yq && \
    chmod a+x /usr/local/bin/yq

COPY --from=build_causal_conv1d /opt/causal_conv1d-1.2.2.post1-cp310-cp310-linux_x86_64.whl ./
COPY --from=build_grouped_gemm /opt/grouped_gemm-1.1.2-cp310-cp310-linux_x86_64.whl ./
COPY --from=build_mamba_ssm /opt/mamba_ssm-2.0.3-cp310-cp310-linux_x86_64.whl ./

RUN pip3 install --no-cache-dir --upgrade-strategy only-if-needed -v \
einops \
flask-restful \
nltk \
pytest \
pytest-cov \
pytest_mock \
pytest-random-order \
sentencepiece \
tiktoken \
wrapt \
zarr \
wandb \
triton==2.1.0 \
causal_conv1d-1.2.2.post1-cp310-cp310-linux_x86_64.whl \
mamba_ssm-2.0.3-cp310-cp310-linux_x86_64.whl \
grouped_gemm-1.1.2-cp310-cp310-linux_x86_64.whl \
tensorstore==0.1.45 && \
rm *.whl

# Since megatron does not have any dependencies (and isn't a dependency to any other package), we can install it separately to make everything a bit quicker
COPY . /opt/megatron-lm
RUN pip install /opt/megatron-lm
ENV PYTHONPATH="/opt/megatron-lm:$PYTHONPATH"

##### For NVIDIANS only #####
FROM main as jet
ARG CACHEBUST=0
RUN --mount=type=secret,id=JET_INDEX_URLS \
    JET_INDEX_URLS=$(cat /run/secrets/JET_INDEX_URLS) && \
    pip install jet-client --upgrade $JET_INDEX_URLS && \
    /opt/jet/bin/pip install jet-api --upgrade $JET_INDEX_URLS 
ENV PATH="$PATH:/opt/jet/bin"
###