doc updates (microsoft#3415)

davidsolomon21cn · May 1, 2023 · b4b63f5 · b4b63f5
1 parent 7ddc3b0
commit b4b63f5
Showing 1 changed file with 15 additions and 15 deletions.
diff --git a/docs/_tutorials/cifar-10.md b/docs/_tutorials/cifar-10.md
@@ -8,21 +8,21 @@ If you haven't already, we advise you to first read through the
 [Getting Started](/getting-started/) guide before stepping through this
 tutorial.
 
-In this tutorial we will be adding DeepSpeed to CIFAR-10 model, which is small image classification model.
+In this tutorial we will be adding DeepSpeed to the CIFAR-10 model, which is a small image classification model.
 
-First we will go over how to run original CIFAR-10. Then we will proceed step-by-step in enabling this model to run with DeepSpeed.
+First we will go over how to run the original CIFAR-10 model. Then we will proceed step-by-step in enabling this model to run with DeepSpeed.
 
 
 
 ## Running Original CIFAR-10
 
-Original model code from [CIFAR-10 Tutorial](https://github.com/pytorch/tutorials/blob/main/beginner_source/blitz/cifar10_tutorial.py), We've copied this repo under [DeepSpeedExamples/training/cifar/](https://github.com/microsoft/DeepSpeedExamples/tree/master/training/cifar) and made it available as a submodule. To download, execute:
+Original model code from the [CIFAR-10 Tutorial](https://github.com/pytorch/tutorials/blob/main/beginner_source/blitz/cifar10_tutorial.py), We've copied this repo under [DeepSpeedExamples/training/cifar/](https://github.com/microsoft/DeepSpeedExamples/tree/master/training/cifar) and made it available as a submodule. To download, execute:
 
 ```bash
 git submodule update --init --recursive
 ```
 
-To install requirements for CIFAR-10:
+To install the requirements for the CIFAR-10 model:
 ```bash
 cd DeepSpeedExamples/cifar
 pip install -r requirements.txt
@@ -82,22 +82,22 @@ The first step to apply DeepSpeed is adding DeepSpeed arguments to CIFAR-10 mode
 
      parser=argparse.ArgumentParser(description='CIFAR')
 
-     #data
-     # cuda
+     # Data.
+     # Cuda.
      parser.add_argument('--with_cuda', default=False, action='store_true',
                          help='use CPU in case there\'s no GPU support')
      parser.add_argument('--use_ema', default=False, action='store_true',
                          help='whether use exponential moving average')
 
-     # train
+     # Train.
      parser.add_argument('-b', '--batch_size', default=32, type=int,
                          help='mini-batch size (default: 32)')
      parser.add_argument('-e', '--epochs', default=30, type=int,
                          help='number of total epochs (default: 30)')
      parser.add_argument('--local_rank', type=int, default=-1,
                         help='local rank passed from distributed launcher')
 
-     # Include DeepSpeed configuration arguments
+     # Include DeepSpeed configuration arguments.
      parser = deepspeed.add_config_arguments(parser)
 
      args=parser.parse_args()
@@ -123,16 +123,16 @@ def initialize(args,
                collate_fn=None):
 ```
 
-Here we initialize DeepSpeed with CIFAR-10 model (`net`), `args`, `parameters` and `trainset`:
+Here we initialize DeepSpeed with the CIFAR-10 model (`net`), `args`, `parameters` and `trainset`:
 
 ```python
  parameters = filter(lambda p: p.requires_grad, net.parameters())
  args=add_argument()
 
  # Initialize DeepSpeed to use the following features
- # 1) Distributed model
- # 2) Distributed data loader
- # 3) DeepSpeed optimizer
+ # 1) Distributed model.
+ # 2) Distributed data loader.
+ # 3) DeepSpeed optimizer.
  model_engine, optimizer, trainloader, _ = deepspeed.initialize(args=args, model=net, model_parameters=parameters, training_data=trainset)
 
 ```
@@ -155,7 +155,7 @@ The `model` returned by `deepspeed.initialize` is the _DeepSpeed Model Engine_ t
 
 ```python
      for i, data in enumerate(trainloader):
-         # get the inputs; data is a list of [inputs, labels]
+         # Get the inputs; data is a list of [inputs, labels].
          inputs = data[0].to(model_engine.device)
          labels = data[1].to(model_engine.device)
 
@@ -206,13 +206,13 @@ The next step to use DeepSpeed is to create a configuration JSON file (ds_config
 
 ### Run CIFAR-10 Model with DeepSpeed Enabled
 
-To start training CIFAR-10 model with DeepSpeed applied, execute the following command, it will use all detected GPUs by default.
+To start training the CIFAR-10 model with DeepSpeed applied, execute the following command, it will use all detected GPUs by default.
 
 ```bash
 deepspeed cifar10_deepspeed.py --deepspeed_config ds_config.json
 ```
 
-DeepSpeed usually prints more training details for user to monitor, including training settings, performance statistics and loss trends.
+DeepSpeed usually prints more training details for the user to monitor, including training settings, performance statistics and loss trends.
 ```
 deepspeed.pt cifar10_deepspeed.py --deepspeed_config ds_config.json
 Warning: Permanently added '[192.168.0.22]:42227' (ECDSA) to the list of known hosts.