1.我们现在的进展到哪了 还差什么 ?
2."PowerShell 7 环境已加载 (版本: 7.5.2)
PS C:\Users\Administrator\Desktop> cd E:\PyTorch_Build\pytorch
PS E:\PyTorch_Build\pytorch> .\pytorch_env\Scripts\activate
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 列出所有可用的 Conda 环境
(pytorch_env) PS E:\PyTorch_Build\pytorch> conda env list
# conda environments:
#
base C:\Miniconda3
pytorch-env C:\Miniconda3\envs\pytorch-env
(pytorch_env) PS E:\PyTorch_Build\pytorch>
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 如果 pytorch_env 不存在,创建新环境
(pytorch_env) PS E:\PyTorch_Build\pytorch> conda create -n pytorch_env python=3.10 -y
3 channel Terms of Service accepted
Channels:
- defaults
- conda-forge
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main
Platform: win-64
Collecting package metadata (repodata.json): done
Solving environment: done
## Package Plan ##
environment location: C:\Miniconda3\envs\pytorch_env
added / updated specs:
- python=3.10
done
#
# To activate this environment, use
#
# $ conda activate pytorch_env
#
# To deactivate an active environment, use
#
# $ conda deactivate
(pytorch_env) PS E:\PyTorch_Build\pytorch> conda activate pytorch_env
(pytorch_env) PS E:\PyTorch_Build\pytorch>
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 安装必要的 Python 包
(pytorch_env) PS E:\PyTorch_Build\pytorch> conda install -c conda-forge -y numpy mkl mkl-include mkl-service intel-openmp
3 channel Terms of Service accepted
Channels:
- conda-forge
- defaults
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main
Platform: win-64
Collecting package metadata (repodata.json): done
Solving environment: done
## Package Plan ##
environment location: C:\Miniconda3
added / updated specs:
- intel-openmp
- mkl
- mkl-include
- mkl-service
- numpy
The following packages will be downloaded:
package | build
---------------------------|-----------------
ca-certificates-2025.8.3 | h4c7d964_0 151 KB conda-forge
intel-openmp-2025.2.0 | h57928b3_757 21.4 MB conda-forge
libblas-3.9.0 | 34_h5709861_mkl 69 KB conda-forge
libcblas-3.9.0 | 34_h2a3cdd5_mkl 69 KB conda-forge
liblapack-3.9.0 | 34_hf9ab0e9_mkl 80 KB conda-forge
llvm-openmp-20.1.8 | h29ce207_0 329 KB defaults
mkl-2024.2.2 | h57928b3_16 98.3 MB conda-forge
mkl-include-2025.2.0 | h57928b3_628 692 KB conda-forge
mkl-service-2.5.2 | py313haca3b5c_0 64 KB conda-forge
openssl-3.5.2 | h725018a_0 8.8 MB conda-forge
------------------------------------------------------------
Total: 130.0 MB
The following NEW packages will be INSTALLED:
llvm-openmp pkgs/main/win-64::llvm-openmp-20.1.8-h29ce207_0
mkl-service conda-forge/win-64::mkl-service-2.5.2-py313haca3b5c_0
The following packages will be UPDATED:
ca-certificates pkgs/main/win-64::ca-certificates-202~ --> conda-forge/noarch::ca-certificates-2025.8.3-h4c7d964_0
intel-openmp 2024.2.1-h57928b3_1083 --> 2025.2.0-h57928b3_757
libblas 3.9.0-24_win64_mkl --> 3.9.0-34_h5709861_mkl
libcblas 3.9.0-24_win64_mkl --> 3.9.0-34_h2a3cdd5_mkl
liblapack 3.9.0-24_win64_mkl --> 3.9.0-34_hf9ab0e9_mkl
mkl 2024.1.0-h66d3029_694 --> 2024.2.2-h57928b3_16
mkl-include 2024.1.0-h66d3029_694 --> 2025.2.0-h57928b3_628
openssl 3.1.8-ha4e3fda_0 --> 3.5.2-h725018a_0
Downloading and Extracting Packages:
Preparing transaction: done
Verifying transaction: done
Executing transaction: done
(pytorch_env) PS E:\PyTorch_Build\pytorch>
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 验证 MKL 安装
(pytorch_env) PS E:\PyTorch_Build\pytorch> python -c "import mkl; print(f'MKL version: {mkl.__version__}')"
Traceback (most recent call last):
File "<string>", line 1, in <module>
ModuleNotFoundError: No module named 'mkl'
(pytorch_env) PS E:\PyTorch_Build\pytorch> python -c "import numpy as np; np.show_config()"
Traceback (most recent call last):
File "<string>", line 1, in <module>
ModuleNotFoundError: No module named 'numpy'
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 确保在 pytorch 源码目录
(pytorch_env) PS E:\PyTorch_Build\pytorch> cd E:\PyTorch_Build\pytorch
(pytorch_env) PS E:\PyTorch_Build\pytorch>
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 设置环境变量(根据您的实际路径修改)
(pytorch_env) PS E:\PyTorch_Build\pytorch> $env:CUDA_PATH = "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0"
(pytorch_env) PS E:\PyTorch_Build\pytorch> $env:CUDNN_INCLUDE_DIR = "E:\Program Files\NVIDIA\CUNND\v9.12\include"
(pytorch_env) PS E:\PyTorch_Build\pytorch> $env:CUDNN_LIBRARY = "E:\Program Files\NVIDIA\CUNND\v9.12\lib\x64\cudnn.lib"
(pytorch_env) PS E:\PyTorch_Build\pytorch>
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 安装构建依赖
(pytorch_env) PS E:\PyTorch_Build\pytorch> pip install cmake ninja
Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple
Requirement already satisfied: cmake in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (4.1.0)
Requirement already satisfied: ninja in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (1.13.0)
(pytorch_env) PS E:\PyTorch_Build\pytorch>
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 开始构建
(pytorch_env) PS E:\PyTorch_Build\pytorch> python setup.py install --cmake -G "Ninja" `
>> -DCMAKE_CUDA_COMPILER="$env:CUDA_PATH\bin\nvcc.exe" `
>> -DCUDNN_INCLUDE_DIR="$env:CUDNN_INCLUDE_DIR" `
>> -DCUDNN_LIBRARY="$env:CUDNN_LIBRARY"
Building wheel torch-2.9.0a0+git2d31c3d
option -G not recognized
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 卸载现有版本
(pytorch_env) PS E:\PyTorch_Build\pytorch> pip uninstall -y torch torchvision torchaudio
Found existing installation: torch 2.9.0a0+git2d31c3d
Uninstalling torch-2.9.0a0+git2d31c3d:
Successfully uninstalled torch-2.9.0a0+git2d31c3d
WARNING: Skipping torchvision as it is not installed.
WARNING: Skipping torchaudio as it is not installed.
(pytorch_env) PS E:\PyTorch_Build\pytorch>
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 安装 CUDA 12.1 版本的 PyTorch(兼容 CUDA 13.0)
(pytorch_env) PS E:\PyTorch_Build\pytorch> pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
Looking in indexes: https://download.pytorch.org/whl/cu121
Collecting torch
Using cached https://download.pytorch.org/whl/cu121/torch-2.5.1%2Bcu121-cp310-cp310-win_amd64.whl (2449.4 MB)
Collecting torchvision
Using cached https://download.pytorch.org/whl/cu121/torchvision-0.20.1%2Bcu121-cp310-cp310-win_amd64.whl (6.1 MB)
Collecting torchaudio
Using cached https://download.pytorch.org/whl/cu121/torchaudio-2.5.1%2Bcu121-cp310-cp310-win_amd64.whl (4.1 MB)
Requirement already satisfied: filelock in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch) (3.19.1)
Requirement already satisfied: typing-extensions>=4.8.0 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch) (4.15.0)
Requirement already satisfied: networkx in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch) (3.4.2)
Requirement already satisfied: jinja2 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch) (3.1.6)
Requirement already satisfied: fsspec in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch) (2025.7.0)
Collecting sympy==1.13.1 (from torch)
Using cached https://download.pytorch.org/whl/sympy-1.13.1-py3-none-any.whl (6.2 MB)
Requirement already satisfied: mpmath<1.4,>=1.1.0 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from sympy==1.13.1->torch) (1.3.0)
Collecting numpy (from torchvision)
Downloading https://download.pytorch.org/whl/numpy-2.1.2-cp310-cp310-win_amd64.whl.metadata (59 kB)
Collecting pillow!=8.3.*,>=5.3.0 (from torchvision)
Downloading https://download.pytorch.org/whl/pillow-11.0.0-cp310-cp310-win_amd64.whl.metadata (9.3 kB)
Requirement already satisfied: MarkupSafe>=2.0 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from jinja2->torch) (3.0.2)
Downloading https://download.pytorch.org/whl/pillow-11.0.0-cp310-cp310-win_amd64.whl (2.6 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 2.6/2.6 MB 5.7 MB/s 0:00:00
Downloading https://download.pytorch.org/whl/numpy-2.1.2-cp310-cp310-win_amd64.whl (12.9 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 12.9/12.9 MB 27.8 MB/s 0:00:00
Installing collected packages: sympy, pillow, numpy, torch, torchvision, torchaudio
Attempting uninstall: sympy
Found existing installation: sympy 1.14.0
Uninstalling sympy-1.14.0:
Successfully uninstalled sympy-1.14.0
Successfully installed numpy-2.1.2 pillow-11.0.0 sympy-1.13.1 torch-2.5.1+cu121 torchaudio-2.5.1+cu121 torchvision-0.20.1+cu121
(pytorch_env) PS E:\PyTorch_Build\pytorch>
(pytorch_env) PS E:\PyTorch_Build\pytorch> # 或者安装特定版本
(pytorch_env) PS E:\PyTorch_Build\pytorch> pip install torch==2.3.0+cu121 torchvision==0.18.0+cu121 torchaudio==2.3.0 -f https://download.pytorch.org/whl/torch_stable.html
Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple
Looking in links: https://download.pytorch.org/whl/torch_stable.html
Collecting torch==2.3.0+cu121
Downloading https://download.pytorch.org/whl/cu121/torch-2.3.0%2Bcu121-cp310-cp310-win_amd64.whl (2413.3 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 2.4/2.4 GB 27.9 MB/s 0:01:13
Collecting torchvision==0.18.0+cu121
Downloading https://download.pytorch.org/whl/cu121/torchvision-0.18.0%2Bcu121-cp310-cp310-win_amd64.whl (5.7 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 5.7/5.7 MB 115.0 MB/s 0:00:00
Collecting torchaudio==2.3.0
Downloading https://download.pytorch.org/whl/cu121/torchaudio-2.3.0%2Bcu121-cp310-cp310-win_amd64.whl (4.1 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 4.1/4.1 MB 80.9 MB/s 0:00:00
Requirement already satisfied: filelock in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch==2.3.0+cu121) (3.19.1)
Requirement already satisfied: typing-extensions>=4.8.0 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch==2.3.0+cu121) (4.15.0)
Requirement already satisfied: sympy in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch==2.3.0+cu121) (1.13.1)
Requirement already satisfied: networkx in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch==2.3.0+cu121) (3.4.2)
Requirement already satisfied: jinja2 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch==2.3.0+cu121) (3.1.6)
Requirement already satisfied: fsspec in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torch==2.3.0+cu121) (2025.7.0)
Collecting mkl<=2021.4.0,>=2021.1.1 (from torch==2.3.0+cu121)
Downloading https://pypi.tuna.tsinghua.edu.cn/packages/fe/1c/5f6dbf18e8b73e0a5472466f0ea8d48ce9efae39bd2ff38cebf8dce61259/mkl-2021.4.0-py2.py3-none-win_amd64.whl (228.5 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 228.5/228.5 MB 46.8 MB/s 0:00:05
Requirement already satisfied: numpy in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torchvision==0.18.0+cu121) (2.1.2)
Requirement already satisfied: pillow!=8.3.*,>=5.3.0 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from torchvision==0.18.0+cu121) (11.0.0)
Collecting intel-openmp==2021.* (from mkl<=2021.4.0,>=2021.1.1->torch==2.3.0+cu121)
Downloading https://pypi.tuna.tsinghua.edu.cn/packages/6f/21/b590c0cc3888b24f2ac9898c41d852d7454a1695fbad34bee85dba6dc408/intel_openmp-2021.4.0-py2.py3-none-win_amd64.whl (3.5 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 3.5/3.5 MB 51.7 MB/s 0:00:00
Requirement already satisfied: tbb==2021.* in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from mkl<=2021.4.0,>=2021.1.1->torch==2.3.0+cu121) (2021.13.1)
Requirement already satisfied: MarkupSafe>=2.0 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from jinja2->torch==2.3.0+cu121) (3.0.2)
Requirement already satisfied: mpmath<1.4,>=1.1.0 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from sympy->torch==2.3.0+cu121) (1.3.0)
Installing collected packages: intel-openmp, mkl, torch, torchvision, torchaudio
Attempting uninstall: intel-openmp
Found existing installation: intel-openmp 2024.2.1
Uninstalling intel-openmp-2024.2.1:
Successfully uninstalled intel-openmp-2024.2.1
Attempting uninstall: torch
Found existing installation: torch 2.5.1+cu121
Uninstalling torch-2.5.1+cu121:
Successfully uninstalled torch-2.5.1+cu121
Attempting uninstall: torchvision
Found existing installation: torchvision 0.20.1+cu121
Uninstalling torchvision-0.20.1+cu121:
Successfully uninstalled torchvision-0.20.1+cu121
Attempting uninstall: torchaudio
Found existing installation: torchaudio 2.5.1+cu121
Uninstalling torchaudio-2.5.1+cu121:
Successfully uninstalled torchaudio-2.5.1+cu121
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
mkl-static 2024.1.0 requires intel-openmp==2024.*, but you have intel-openmp 2021.4.0 which is incompatible.
Successfully installed intel-openmp-2021.4.0 mkl-2021.4.0 torch-2.3.0+cu121 torchaudio-2.3.0+cu121 torchvision-0.18.0+cu121
(pytorch_env) PS E:\PyTorch_Build\pytorch> python -c "
>> import torch
>> from torch.utils import cpp_extension
>>
>> print('='*50)
>> print(f'PyTorch 版本: {torch.__version__}')
>> print(f'CUDA 可用: {torch.cuda.is_available()}')
>> if torch.cuda.is_available():
>> print(f'CUDA 版本: {torch.version.cuda}')
>> print(f'cuDNN 版本: {torch.backends.cudnn.version()}')
>> print(f'GPU 名称: {torch.cuda.get_device_name(0)}')
>>
>> print('='*20 + ' 配置信息 ' + '='*20)
>> print(torch.__config__.show())
>>
>> print('='*20 + ' 简单计算测试 ' + '='*20)
>> x = torch.randn(1000, 1000, device='cuda')
>> y = torch.randn(1000, 1000, device='cuda')
>> z = x @ y
>> print(f'矩阵乘法完成: {z.size()}')
>>
>> print('='*20 + ' cuDNN 卷积测试 ' + '='*20)
>> conv = torch.nn.Conv2d(3, 64, kernel_size=3).cuda()
>> input = torch.randn(1, 3, 256, 256).cuda()
>> output = conv(input)
>> print(f'卷积输出尺寸: {output.size()}')
>>
>> print('='*20 + ' MKL 性能测试 ' + '='*20)
>> a = torch.randn(5000, 5000)
>> b = torch.randn(5000, 5000)
>> c = a @ b
>> print(f'CPU 矩阵乘法完成: {c.size()}')
>>
>> print('='*50)
>> "
==================================================
PyTorch 版本: 2.3.0+cu121
CUDA 可用: True
CUDA 版本: 12.1
cuDNN 版本: 8801
E:\PyTorch_Build\pytorch\pytorch_env\lib\site-packages\torch\cuda\__init__.py:209: UserWarning:
NVIDIA GeForce RTX 5070 with CUDA capability sm_120 is not compatible with the current PyTorch installation.
The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_61 sm_70 sm_75 sm_80 sm_86 sm_90.
If you want to use the NVIDIA GeForce RTX 5070 GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
warnings.warn(
GPU 名称: NVIDIA GeForce RTX 5070
==================== 配置信息 ====================
PyTorch built with:
- C++ Version: 201703
- MSVC 192930151
- Intel(R) oneAPI Math Kernel Library Version 2021.4-Product Build 20210904 for Intel(R) 64 architecture applications
- Intel(R) MKL-DNN v3.3.6 (Git Hash 86e6af5974177e513fd3fee58425e1063e7f1361)
- OpenMP 2019
- LAPACK is enabled (usually provided by MKL)
- CPU capability usage: AVX2
- CUDA Runtime 12.1
- NVCC architecture flags: -gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_90,code=sm_90
- CuDNN 8.8.1 (built against CUDA 12.0)
- Magma 2.5.4
- Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=12.1, CUDNN_VERSION=8.8.1, CXX_COMPILER=C:/actions-runner/_work/pytorch/pytorch/builder/windows/tmp_bin/sccache-cl.exe, CXX_FLAGS=/DWIN32 /D_WINDOWS /GR /EHsc /Zc:__cplusplus /bigobj /FS /utf-8 -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOCUPTI -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE /wd4624 /wd4068 /wd4067 /wd4267 /wd4661 /wd4717 /wd4244 /wd4804 /wd4273, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=2.3.0, USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=OFF, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=OFF, USE_NNPACK=OFF, USE_OPENMP=ON, USE_ROCM=OFF, USE_ROCM_KERNEL_ASSERT=OFF,
==================== 简单计算测试 ====================
A module that was compiled using NumPy 1.x cannot be run in
NumPy 2.1.2 as it may crash. To support both 1.x and 2.x
versions of NumPy, modules must be compiled with NumPy 2.0.
Some module may need to rebuild instead e.g. with 'pybind11>=2.12'.
If you are a user of the module, the easiest solution will be to
downgrade to 'numpy<2' or try to upgrade the affected module.
We expect that some modules will need time to support NumPy 2.
Traceback (most recent call last): File "<string>", line 17, in <module>
<string>:17: UserWarning: Failed to initialize NumPy: _ARRAY_API not found (Triggered internally at ..\torch\csrc\utils\tensor_numpy.cpp:84.)
Traceback (most recent call last):
File "<string>", line 17, in <module>
RuntimeError: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
“
最新发布