CV-CUDA/versions.env at main · CVCUDA/CV-CUDA · 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
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

# versions.env — single source of truth for managed dependency versions and source pins in CV-CUDA.
# Edit this file to update any pinned version, then regenerate requirements files:
#   bash generate_requirements.sh

# CUDA base versions
cuda_cu12=12.5.0
cuda_cu13=13.3.0
cuda_cu13_arm=13.0.1   # arm CI constraint: hosts lack drivers for 13.3

# PyTorch
torch_cu12=2.13.0
torch_cu12_index=cu126   # wheel index for CUDA 12.x (backward compatible with CUDA 12.5+)
torch_cu13=2.13.0
torch_cu13_index=cu130   # wheel index for CUDA 13.x (forward compatible with CUDA 13.3+, cu133 index not available yet)

# Minimum CUDA version the NVIDIA driver must report for the torch wheels above.
# The wheels bundle their own CUDA runtime, so this tracks the wheel index, not
# the installed toolkit. Consumed by samples/install_samples_dependencies.sh.
torch_cu12_driver_min=12.6
torch_cu13_driver_min=13.0

# torchvision (versions must be paired with torch)
torchvision_cu12=0.28.0
torchvision_cu13=0.28.0

# NumPy
numpy_v1=1.26.4          # Python 3.10-3.12 only (no 3.13+ support for NumPy 1.x)
numpy_v2_py310=2.2.6     # Python 3.10
numpy_v2_py311plus=2.4.4  # Python 3.11+
numpy_floor=1.23.5

# CuPy
cupy_cu12=14.0.1          # NumPy 2.x; CuPy 14 requires NumPy 2+
cupy_cu12_numpy1=13.6.0   # Last CuPy 13 release; supports NumPy 1.22-2.5
cupy_cu13=14.0.1          # NumPy 2.x; CuPy 14 requires NumPy 2+

# NVIDIA packages
nvimgcodec=0.7.0.11      # nvidia-nvimgcodec-cu{12,13} — same version for both
nvjpeg_cu12=12.4.0.76    # nvidia-nvjpeg-cu12 (CUDA 12 only)
nvjpeg_cu13=13.0.4.44    # nvidia-nvjpeg (no CUDA suffix) for CUDA 13

tensorrt_cu12=10.13.3.9.post1
tensorrt_cu13=11.1.0.106

cuda_python_cu12=12.9.4
cuda_python_cu13=13.0.3

pynvvideocodec=2.1.0

# Benchmarking dependencies
pycuda=2025.1.2
pandas=2.3.3
cuda_pathfinder=1.3.3
nvtx=0.2.13

# Samples dependencies
onnx=1.22.0
onnxscript=0.6.2
onnxslim=0.1.78

# Build dependencies
dlpack=1.3
dlpack_commit=84d107bf416c6bab9ae68ad285876600d230490d
pybind11=3.0.3
pybind11_stubgen=2.5.5
setuptools=83.0.0
wheel=0.46.2
build=1.3.0
patchelf=0.17.2.1
auditwheel=6.4.2
flake8=7.1.0

# Tests common dependencies
pytest=9.1.1
typing_extensions=4.13.1

# Docs dependencies
sphinx=8.1.3
exhale=0.3.7
breathe=4.36.0
sphinx_rtd_theme=3.0.2
sphinx_tabs=3.4.7
recommonmark=0.7.1
graphviz=0.20.3
sphinxcontrib_jquery=4.1
sphinxcontrib_jsmath=1.0.1
sphinxcontrib_applehelp=2.0.0
sphinxcontrib_devhelp=2.0.0
sphinxcontrib_htmlhelp=2.1.0
sphinxcontrib_qthelp=2.0.0
sphinxcontrib_serializinghtml=2.0.0