TensorRT-Model-Connect/Dockerfile.community-cpu at main · NVIDIA/TensorRT-Model-Connect · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0

# Public, GPU-free source validation image. It contains the CUDA/TensorRT SDK
# needed to compile the CLI, but runtime containers receive no GPU devices.
ARG CUDA_IMAGE=nvidia/cuda:13.4.3-base-ubuntu24.04@sha256:07adb6792c59025f2c45bfb1a3f1d1148bf6d48bf2326d604a6b2e24c4f04d6f
ARG TENSORRT_VERSION=11.4.0.106
ARG TENSORRT_APT_VERSION=11.4.0.106-1+cuda13.4
FROM ${CUDA_IMAGE}

ARG TENSORRT_VERSION
ARG TENSORRT_APT_VERSION
ARG PYTORCH_INDEX=https://download.pytorch.org/whl/cu130
ARG TORCH_VERSION=2.12.0+cu130

ENV DEBIAN_FRONTEND=noninteractive
ENV PIP_NO_CACHE_DIR=1

RUN apt-get update && \
    apt-get install -y --no-install-recommends \
      build-essential \
      cmake \
      cuda-cudart-dev-13-4 \
      cuda-nvcc-13-4 \
      cuda-nvrtc-dev-13-4 \
      git \
      jq \
      ninja-build \
      nlohmann-json3-dev \
      libicu-dev \
      pkg-config \
      python3.12 \
      python3.12-dev \
      python3.12-venv \
      libcublas-dev-13-4 \
      libcurand-dev-13-4 \
      "libnvinfer-dev=${TENSORRT_APT_VERSION}" \
      "libnvinfer-headers-dev=${TENSORRT_APT_VERSION}" \
      "libnvinfer-headers-plugin-dev=${TENSORRT_APT_VERSION}" \
      "libnvinfer-safe-headers-dev=${TENSORRT_APT_VERSION}" \
      "libnvinfer-plugin11=${TENSORRT_APT_VERSION}" \
      "libnvinfer11=${TENSORRT_APT_VERSION}" \
      "libnvonnxparsers-dev=${TENSORRT_APT_VERSION}" \
      "libnvonnxparsers11=${TENSORRT_APT_VERSION}" && \
    rm -rf /var/lib/apt/lists/*

ENV VIRTUAL_ENV=/opt/venv
RUN python3.12 -m venv ${VIRTUAL_ENV}
ENV PATH="${VIRTUAL_ENV}/bin:${PATH}"

RUN pip install --upgrade pip && \
    pip install \
      "cuda-python>=13.0.3,<14" \
      "huggingface_hub>=0.23" \
      "ml_dtypes>=0.4" \
      "numpy>=1.24,<2.5" \
      "onnx>=1.16" \
      "safetensors>=0.4" \
      "sentencepiece>=0.1.99" \
      "transformers==5.2.0" && \
    pip install --no-deps \
      "tensorrt_cu13==${TENSORRT_VERSION}" \
      "tensorrt_cu13_bindings==${TENSORRT_VERSION}" && \
    pip install "torch==${TORCH_VERSION}" --index-url "${PYTORCH_INDEX}" && \
    pip install "setuptools>=80,<82"

COPY community-ci.txt /tmp/trtmc-community-ci.txt
RUN pip install --requirement /tmp/trtmc-community-ci.txt && \
    rm -f /tmp/trtmc-community-ci.txt

RUN multiarch="$(gcc -dumpmachine)" && \
    test -d "/usr/lib/${multiarch}" && \
    test -d "/usr/include/${multiarch}" && \
    ln -s "/usr/lib/${multiarch}" /opt/trtmc-tensorrt-lib && \
    ln -s "/usr/include/${multiarch}" /opt/trtmc-tensorrt-include

ENV TRT_LIB_DIR=/opt/trtmc-tensorrt-lib
ENV TRT_INC_DIR=/opt/trtmc-tensorrt-include
ENV TVM_FFI_LIB_DIR=/opt/venv/lib/python3.12/site-packages/tvm_ffi/lib
ENV LD_LIBRARY_PATH="${TRT_LIB_DIR}:${TVM_FFI_LIB_DIR}:/usr/local/cuda/lib64"
ENV TORCH_CUDA_ARCH_LIST=10.0

RUN python3.12 -c \
      "import tensorrt, torch, transformers, tvm_ffi; assert tensorrt.__version__ == '${TENSORRT_VERSION}'; assert torch.__version__ == '${TORCH_VERSION}'; assert torch.version.cuda == '13.0'; assert transformers.__version__ == '5.2.0'" && \
    test -f "${TRT_INC_DIR}/NvInferVersion.h" && \
    test -f "${TRT_LIB_DIR}/libnvinfer.so.11" && \
    test -f "${TVM_FFI_LIB_DIR}/libtvm_ffi.so"

WORKDIR /workspace/tensorrt-model-connect
CMD ["bash"]