动手学深度学习d2l包M4芯片 gpu加速

conda创建环境

CONDA_SUBDIR=osx-arm64 conda create -n ml python=3.9 -c conda-forge
conda env config vars set CONDA_SUBDIR=osx-arm64
conda activate ml

pip安装包

pip install --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/cpu
pip install transformers datasets
pip install matplotlib

下载del源文件

del链接
放到本地项目内

修改del下的torch.py的两个函数内容

# 修改try gpu函数
def try_gpu(i=0):
    """Return gpu(i) if exists, otherwise return cpu().

    Defined in :numref:`sec_use_gpu`"""
    if torch.cuda.device_count() >= i + 1:
        return torch.device(f'cuda:{i}')
    try:
        return torch.device('mps')
    except:
        return torch.device('cpu')

# 修改try gpu函数
def try_all_gpus():
    """Return all available GPUs, or [cpu(),] if no GPU exists.

    Defined in :numref:`sec_use_gpu`"""
    devices = [torch.device(f'cuda:{i}')
               for i in range(torch.cuda.device_count())]
    try:
        device_macos = torch.device('mps')
    except:
        device_macos = torch.device('cpu')
    return devices if devices else [device_macos]

测试

运行lenet.ipynb测试效果

速度还可以。

还不懂的可以看M1版本的教程