Skip to content

GitLab

Unverified Commit 4d735946 authored by Shaden Smith's avatar Shaden Smith Committed by GitHub
Browse files

Web edits (#146) 

* Add coming soon to posts

* Add what's new section to main page
parent 85cc16ae
Show whitespace changes
Inline Side-by-side
Showing with 102 additions and 256 deletions
docs/_config.yml
View file @ 4d735946
......@@ -37,10 +37,11 @@ defaults:
values:
layout: single
author_profile: false
read_time: true
read_time: false
comments: false
share: true
related: false
sneak_preview: false
# _tutorials
- scope:
path: ""
......
docs/_data/navigation.yml
View file @ 4d735946
main:
- title: "Getting Started"
url: /getting-started/
- title: "Blog"
url: /blog/
- title: "News"
url: /news/
- title: "Tutorials"
url: /tutorials/
- title: "Documentation"
url: https://ghpages-test.readthedocs.io/
url: https://deepspeed.readthedocs.io/
- title: "GitHub"
url: https://github.com/microsoft/DeepSpeed
lnav:
- title: "This is a floating nav bar."
- title: "DeepSpeed Documentation"
- title: "Getting Started"
url: /getting-started/
children:
......@@ -19,3 +19,5 @@ lnav:
url: /getting-started/#installation
- title: "Configuration"
url: /getting-started/#deepspeed-configuration
- title: "DeepSpeed Features"
url: /features/
docs/_layouts/news-home.html 0 → 100644
View file @ 4d735946
---
layout: archive
---
{{ content }}
{% if paginator %}
{% assign posts = paginator.posts %}
{% else %}
{% assign posts = site.posts %}
{% endif %}
<h2>Features Coming Soon</h2>
{% assign soon = posts | where: "sneak_preview", "true" %}
{% for post in soon %}
{% include archive-single.html %}
{% endfor %}
<h2>{{ site.data.ui-text[site.locale].recent_posts | default: "Recent Posts" }}</h2>
{% assign news = posts | where: "sneak_preview", "false" %}
{% for post in news %}
{% include archive-single.html %}
{% endfor %}
{% include paginator.html %}
docs/_posts/2020-03-17-reduce-scatter.md
View file @ 4d735946
---
title: "ZeRO stage 1 with reduced communication"
date: 2020-03-13
sneak_preview: true
excerpt: "Partition-aware ZeRO with up to 2x reduction in communication time!"
---
# ZeRO stage 1 with reduced communication
* Partition-aware approach instead of initial implementation that used a global collective (all-reduce)
* Total communication volume reduction 1.5x -> 1x of data parallelism
* Up to 2x reduction in communication time compared to all-reduce
# Further updates coming soon!
## Further updates coming soon!
docs/_posts/2020-03-17-zero-stage2.md
View file @ 4d735946
---
title: "ZeRO stage 2"
date: 2020-03-13
sneak_preview: true
excerpt: "Reduce memory footprint to enable training 10B models without model parallelism!"
---
# Zero Stage 2
* Reduce memory footprint of gradients
* Train larger models: e.g., 10B parameters on 32GPUs without model parallelism
* Train larger batch sizes
# Further updates coming soon!
## Further updates coming soon!
docs/_tutorials/getting-started.md
View file @ 4d735946
......@@ -213,7 +213,7 @@ deepspeed --include="worker-2:0,1" \
<client_entry.py> <client args> \
--deepspeed --deepspeed_config ds_config.json
```
This is a floating nav bar.
### MPI Compatibility
As described above, DeepSpeed provides its own parallel launcher to help launch
multi-node/multi-gpu training jobs. If you prefer to launch your training job
......
docs/assets/css/main.scss 0 → 100644
View file @ 4d735946
---
# Only the main Sass file needs front matter (the dashes are enough)
---
@charset "utf-8";
@import "minimal-mistakes/skins/{{ site.minimal_mistakes_skin | default: 'default' }}"; // skin
@import "minimal-mistakes"; // main partials
//
// DeepSpeed customizations
//
.site-title {
display: -webkit-box;
display: -ms-flexbox;
display: flex;
-ms-flex-item-align: center;
align-self: center;
font-weight: bold;
font-size: $type-size-2; // DeepSpeed: increase size
}
.toc {
font-family: $sans-serif-narrow;
color: $gray;
background-color: $background-color;
border: 1px solid $border-color;
border-radius: $border-radius;
-webkit-box-shadow: $box-shadow;
box-shadow: $box-shadow;
position: fixed;
.nav__title {
color: #fff;
font-size: $type-size-6;
background: $primary-color;
border-top-left-radius: $border-radius;
border-top-right-radius: $border-radius;
}
// Scrollspy marks toc items as .active when they are in focus
.active a {
@include yiq-contrasted($active-color);
}
}
docs/features.md
View file @ 4d735946
---
title: "Feature Overview"
layout: single
permalink: /features/
toc: true
toc_label: "Contents"
---
......
docs/index.md
View file @ 4d735946
......@@ -18,6 +18,16 @@ a language model (LM) with over 17B parameters called
[Turing-NLG](https://www.microsoft.com/en-us/research/blog/turing-nlg-a-17-billion-parameter-language-model-by-microsoft),
establishing a new SOTA in the LM category.
# What's New?
{% assign news = site.posts | where: "sneak_preview", "false" %}
{% for post in news limit:5 %}
{% if post.link %}
* [{{ post.title }}]({{ post.link }})
{% else %}
* [{{ post.title }}]({{ post.url }})
{% endif %}
{% endfor %}
# Why DeepSpeed?
Training advanced deep learning models is challenging. Beyond model design,
......@@ -56,8 +66,6 @@ optimizations on advanced hyperparameter tuning and optimizers. For example:
| 256 V100 GPUs | NVIDIA | 3.9 |
| 256 V100 GPUs | DeepSpeed | **3.7** |
<!---*Read more*: [BERT tutorial](../../Tutorials/bert_pretraining/deepspeed_bert_training.md)-->
*BERT Tutorial*: Coming Soon
* DeepSpeed trains GPT2 (1.5 billion parameters) 3.75x faster than state-of-art, NVIDIA
......@@ -157,247 +165,6 @@ overview](features) for descriptions and usage.
* [Performance Analysis and Debugging](features.md#performance-analysis-and-debugging)
# Getting Started
## Installation
* Please see our [Azure tutorial](docs/azure.md) to get started with DeepSpeed on Azure!
* If you're not on Azure, we recommend using our docker image via `docker pull deepspeed/deepspeed:latest` which contains a pre-installed version of DeepSpeed and all the necessary dependencies.
* If you want to install DeepSpeed manually, we provide an install script [install.sh](install.sh) to help install on a local machine or across an entire cluster.
## Writing DeepSpeed Models
DeepSpeed model training is accomplished using the DeepSpeed engine. The engine
can wrap any arbitrary model of type `torch.nn.module` and has a minimal set of APIs
for training and checkpointing the model. Please see the tutorials for detailed
examples.
To initialize the DeepSpeed engine:
```python
model_engine, optimizer, _, _ = deepspeed.initialize(args=cmd_args,
model=model,
model_parameters=params)
```
`deepspeed.inialize` ensures that all of the necessary setup required for
distributed data parallel or mixed precision training are done
appropriately under the hood. In addition to wrapping the model, DeepSpeed can
construct and manage the training optimizer, data loader, and the learning rate
scheduler based on the parameters passed to `deepspeed.initialze` and the
DeepSpeed [configuration file](#deepspeed-configuration).
### Training
Once the DeepSpeed engine has been initialized, it can be used to train the
model using three simple APIs for forward propagation (`()`), backward
propagation (`backward`), and weight updates (`step`).
```python
for step, batch in enumerate(data_loader):
#forward() method
loss = model_engine(batch)
#runs backpropagation
model_engine.backward(loss)
#weight update
model_engine.step()
```
Under the hood, DeepSpeed automatically performs the necessary operations
required for distributed data parallel training, in mixed precision, with a
pre-defined learning rate schedule:
* **Gradient Averaging**: in distributed data parallel training, `backward`
ensures that gradients are averaged across data parallel processes after
training on an `train_batch_size`.
* **Loss Scaling**: in FP16/mixed precision training, the DeepSpeed
engine automatically handles scaling the loss to avoid precision loss in the
gradients.
* **Learning Rate Schedule**: if using DeepSpeed's learning rate
schedule, then DeepSpeed automatically handles any updates to the learning
rate when `step` is executed.
### Model Checkpointing
Saving and loading the training state is handled via the `save_checkpoint` and
`load_checkpoint` API in DeepSpeed which takes two arguments to uniquely
identify a checkpoint:
* `ckpt_dir`: the directory where checkpoints will be saved.
* `ckpt_id`: an identifier that uniquely identifies a checkpoint in the directory.
In the following code snippet, we use the loss value as the checkpoint identifier.
```python
#load checkpoint
_, client_sd = model_engine.load_checkpoint(args.load_dir, args.ckpt_id)
step = client_sd['step']
#advance data loader to ckpt step
dataloader_to_step(data_loader, step + 1)
for step, batch in enumerate(data_loader):
#forward() method
loss = model_engine(batch)
#runs backpropagation
model_engine.backward(loss)
#weight update
model_engine.step()
#save checkpoint
if step % args.save_interval:
client_sd['step'] = step
ckpt_id = loss.item()
model_engine.save_checkpoint(args.save_dir, ckpt_id, client_sd = client_sd)
```
DeepSpeed can automatically save and restore the model, optimizer, and the
learning rate scheduler states while hiding away these details from the user.
However, the user may want to save other data in addition to these that are
unique to a given model training. To support these items, `save_checkpoint`
accepts a client state dictionary `client_sd` for saving. These items can be
retrieved from `load_checkpoint` as a return argument. In the example above,
the `step` value is stored as part of the `client_sd`.
## DeepSpeed Configuration
DeepSpeed features can be enabled, disabled, or configured using a config JSON
file that should be specified as `args.deepspeed_config`. A sample config file
is shown below. For a full set of features see [core API
doc](https://microsoft.github.io/DeepSpeed/docs/htmlfiles/api/full/index.html).
```json
{
"train_batch_size": 8,
"gradient_accumulation_steps": 1,
"optimizer": {
"type": "Adam",
"params": {
"lr": 0.00015
}
},
"fp16": {
"enabled": true
},
"zero_optimization": true
}
```
## Multi-Node Environment Variables
When training across multiple nodes we have found it useful to support
propagating user-defined environment variables. By default DeepSpeed will
propagate all NCCL and PYTHON related environment variables that are set. If
you would like to propagate additional variables you can specify them in a
dot-file named `.deepspeed_env` that contains a new-line separated list of
`VAR=VAL` entries. The DeepSpeed launcher will look in the local path you are
executing from and also in your home directory (`~/`).
As a concrete example, some clusters require special NCCL variables to set
prior to training. The user can simply add these variables to a
`.deepspeed_env` file in their home directory that looks like this:
```
NCCL_IB_DISABLE=1
NCCL_SOCKET_IFNAME=eth0
```
DeepSpeed will then make sure that these environment variables are set when
launching each process on every node across their training job.
# Launching DeepSpeed Training
DeepSpeed installs the entry point `deepspeed` to launch distributed training.
We illustrate an example usage of DeepSpeed with the following assumptions:
1. You have already integrated DeepSpeed into your model
2. `client_entry.py` is the entry script for your model
3. `client args` is the `argparse` command line arguments
4. `ds_config.json` is the configuration file for DeepSpeed
## Resource Configuration (multi-node)
DeepSpeed configures multi-node compute resources with hostfiles that are compatible with
[OpenMPI](https://www.open-mpi.org/) and [Horovod](https://github.com/horovod/horovod).
A hostfile is a list of *hostnames* (or SSH aliases), which are machines accessible via passwordless
SSH, and *slot counts*, which specify the number of GPUs available on the system. For
example,
```
worker-1 slots=4
worker-2 slots=4
```
specifies that two machines named *worker-1* and *worker-2* each have four GPUs to use
for training.
Hostfiles are specified with the `--hostfile` command line option. If no hostfile is
specified, DeepSpeed searches for `/job/hostfile`. If no hostfile is specified or found,
DeepSpeed queries the number of GPUs on the local machine to discover the number of local
slots available.
The following command launches a PyTorch training job across all available nodes and GPUs
specified in `myhostfile`:
```bash
deepspeed <client_entry.py> <client args> \
--deepspeed --deepspeed_config ds_config.json --hostfile=myhostfile
```
Alternatively, DeepSpeed allows you to restrict distributed training of your model to a
subset of the available nodes and GPUs. This feature is enabled through two command line
arguments: `--num_nodes` and `--num_gpus`. For example, distributed training can be
restricted to use only two nodes with the following command:
```bash
deepspeed --num_nodes=2 \
<client_entry.py> <client args> \
--deepspeed --deepspeed_config ds_config.json
```
You can instead include or exclude specific resources using the `--include` and
`--exclude` flags. For example, to use all available resources **except** GPU 0 on node
*worker-2* and GPUs 0 and 1 on *worker-3*:
```bash
deepspeed --exclude="worker-2:0@worker-3:0,1" \
<client_entry.py> <client args> \
--deepspeed --deepspeed_config ds_config.json
```
Similarly, you can use **only** GPUs 0 and 1 on *worker-2*:
```bash
deepspeed --include="worker-2:0,1" \
<client_entry.py> <client args> \
--deepspeed --deepspeed_config ds_config.json
```
### MPI Compatibility
As described above, DeepSpeed provides its own parallel launcher to help launch
multi-node/multi-gpu training jobs. If you prefer to launch your training job
using MPI (e.g., mpirun), we provide support for this. It should be noted that
DeepSpeed will still use the torch distributed NCCL backend and *not* the MPI
backend. To launch your training job with mpirun + DeepSpeed you simply pass us
an additional flag `--deepspeed_mpi`. DeepSpeed will then use
[mpi4py](https://pypi.org/project/mpi4py/) to discover the MPI environment (e.g.,
rank, world size) and properly initialize torch distributed for training. In this
case you will explicitly invoke `python` to launch your model script instead of using
the `deepspeed` launcher, here is an example:
```bash
mpirun <mpi-args> python \
<client_entry.py> <client args> \
--deepspeed_mpi --deepspeed --deepspeed_config ds_config.json
```
If you want to use this feature of DeepSpeed, please ensure that mpi4py is
installed via `pip install mpi4py`.
## Resource Configuration (single-node)
In the case that we are only running on a single node (with one or more GPUs)
DeepSpeed *does not* require a hostfile as described above. If a hostfile is
not detected or passed in then DeepSpeed will query the number of GPUs on the
local machine to discover the number of slots available. The `--include` and
`--exclude` arguments work as normal, but the user should specify 'localhost'
as the hostname.
# Further Reading
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment