Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

doc: bagua-net add benchmark pic #50

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

doc: bagua-net add benchmark pic #50

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
30 changes: 15 additions & 15 deletions src/more-optimizations/bagua-net.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,21 +38,7 @@ Img/sec per GPU: 126.5 +-6.4
Total img/sec on 32 GPU(s): 4046.6 +-205.2
```

<!--
## Some test results

### 1. Performance comparison of Bagua-Net and NCCL-TCP under 100G TCP network

![](source/img/nccl-test_Bagua-Net_vs_NCCL-TCP.png)

> Thanks to the tensor fusion of the communication library. The actual communication packets will be larger than 10MB. In this range, Bagua-Net has better performance than NCCL-TCP. I have also done some experiments. When training a small network, Bagua-Net is no worse than NCCL-TCP.

### 2. Bagua-Net's acceleration effect on Bagua's different algorithms

![](source/img/bagua-net_accelerate_bagua_algorithms.png)

> The data comes from the real 128 V100 ImageNet training. The throughput increase brought by Bagua-Net is 11% to 68%.
-->
## Quick Start

To enable Bagua-Net, you only need to pass the `--enable-bagua-net` argument in `bagua.distributed.launch` or `bagua.distributed.run`. No code change in your training script.

Expand All @@ -65,6 +51,20 @@ python3 -m bagua.distributed.launch --enable-bagua-net \

> It worth noting that you can even use `bagua.distributed.launch` or `bagua.distributed.run` with `--enable-bagua-net` argument to launch PyTorch-DDP jobs to improve the training throughput without migrating your code to Bagua.

## Benchmark

### 100G TCP network performance comparison with NCCL native implementation

![](bagua-net/source/img/nccl-test_Bagua-Net_vs_NCCL-TCP.png)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Fix both figures' style, conforming https://32808a3b.tutorials-8ro.pages.dev/benchmark/

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

OK, still under revision.


> Thanks to the tensor fusion of the communication library. The actual communication packets will be larger than 10MB. In this range, Bagua-Net has better performance than NCCL-TCP. I have also done some experiments. And when training a small module, Bagua-Net is no obvious worse than NCCL-TCP.

### Effect on bagua algorithm

![](bagua-net/source/img/bagua-net_accelerate_bagua_algorithms.png)

> The data comes from the real 128 V100 ImageNet training. The throughput increase brought by Bagua-Net is 11% to 68%.

<!--
## Enable Bagua-Net

Expand Down