echo/pytorch-handbook
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Update 5_data_parallel_tutorial.ipynb

Browse Source
This commit is contained in:
Muli Yang 2019-02-19 17:32:23 +08:00 committed by GitHub
parent 4de06da2a1
commit e2b5e871ff
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 7 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
chapter1/5_data_parallel_tutorial.ipynb
View File

@ -14,13 +14,13 @@
"metadata": {}, "metadata": {},
"source": [ "source": [
"\n", "\n",
"数据并行(选读)\n", "数据并行（选读）\n",
"==========================\n", "==========================\n",
"**Authors**: `Sung Kim <https://github.com/hunkim>`_ and `Jenny Kang <https://github.com/jennykang>`_\n", "**Authors**: [Sung Kim](https://github.com/hunkim) and [Jenny Kang](https://github.com/jennykang)\n",
"\n", "\n",
"在这个教程里,我们将学习如何使用``DataParallel``来使用多GPU. \n", "在这个教程里，我们将学习如何使用 ``DataParallel`` 来使用多GPU。 \n",
"\n", "\n",
"PyTorch非常容易的就可以使用多GPU,用如下方式把一个模型放到GPU上\n", "PyTorch非常容易就可以使用多GPU，用如下方式把一个模型放到GPU上：\n",
"\n", "\n",
"```python\n", "```python\n",
"\n", "\n",
@ -28,12 +28,12 @@
" model.to(device)\n", " model.to(device)\n",
"```\n", "```\n",
" GPU:\n", " GPU:\n",
"然后复制所有的张量到GPU上\n", "然后复制所有的张量到GPU上：\n",
"```python\n", "```python\n",
"\n", "\n",
" mytensor = my_tensor.to(device)\n", " mytensor = my_tensor.to(device)\n",
"```\n", "```\n",
"请注意,只调用``my_tensor.to(device)``并没有复制张量到GPU上,而是返回了一个copy。所以你需要把它赋值给一个新的张量并在GPU上使用这个张量。\n", "请注意，只调用``my_tensor.to(device)``并没有复制张量到GPU上，而是返回了一个copy。所以你需要把它赋值给一个新的张量并在GPU上使用这个张量。\n",
"\n", "\n",
"在多GPU上执行前向和反向传播是自然而然的事。\n", "在多GPU上执行前向和反向传播是自然而然的事。\n",
"但是PyTorch默认将只是用一个GPU。\n", "但是PyTorch默认将只是用一个GPU。\n",
@ -104,7 +104,7 @@
"虚拟数据集\n", "虚拟数据集\n",
"-------------\n", "-------------\n",
"\n", "\n",
"制作一个虚拟（随机）数据集\n", "制作一个虚拟（随机）数据集，\n",
"你只需实现 `__getitem__`\n", "你只需实现 `__getitem__`\n",
"\n", "\n",
"\n" "\n"