From 9092b481af04272f5cd942083c4cf29ea2c22782 Mon Sep 17 00:00:00 2001
From: Muli Yang <31494617+mattmoevil@users.noreply.github.com>
Date: Tue, 19 Feb 2019 17:39:44 +0800
Subject: [PATCH] Update 5_data_parallel_tutorial.ipynb

---
 chapter1/5_data_parallel_tutorial.ipynb | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/chapter1/5_data_parallel_tutorial.ipynb b/chapter1/5_data_parallel_tutorial.ipynb
index 0be5795f..da0ebe31 100644
--- a/chapter1/5_data_parallel_tutorial.ipynb
+++ b/chapter1/5_data_parallel_tutorial.ipynb
@@ -36,7 +36,7 @@
     "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",
     "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默认将只使用一个GPU。\n",
     "\n",
     "使用``DataParallel``可以轻易的让模型并行运行在多个GPU上。\n",
     "\n",
@@ -139,7 +139,7 @@
     "简单模型\n",
     "------------\n",
     "作为演示，我们的模型只接受一个输入，执行一个线性操作，然后得到结果。\n",
-    "说明，``DataParallel``能在任何模型（CNN，RNN，Capsule Net等）上使用。\n",
+    "说明：``DataParallel``能在任何模型（CNN，RNN，Capsule Net等）上使用。\n",
     "\n",
     "\n",
     "我们在模型内部放置了一条打印语句来打印输入和输出向量的大小。\n",
@@ -286,7 +286,7 @@
     "    Outside: input size torch.Size([10, 5]) output_size torch.Size([10, 2])\n",
     "\n",
     "3 GPUs\n",
-    "~~~~~~\n",
+    "~\n",
     "\n",
     "If you have 3 GPUs, you will see:\n",
     "\n",
@@ -311,7 +311,7 @@
     "    Outside: input size torch.Size([10, 5]) output_size torch.Size([10, 2])\n",
     "\n",
     "8 GPUs\n",
-    "~~~~~~~~~~~~~~\n",
+    "~~\n",
     "\n",
     "If you have 8, you will see:\n",
     "\n",