https://github.com/tcapelle/apple_m1_pro_python

A collection of ML scripts to test the M1 Pro MacBook Pro
https://github.com/tcapelle/apple_m1_pro_python

Last synced: 14 days ago
JSON representation

A collection of ML scripts to test the M1 Pro MacBook Pro

• Host: GitHub
• URL: https://github.com/tcapelle/apple_m1_pro_python
• Owner: tcapelle
• Created: 2021-11-19T09:02:35.000Z (almost 3 years ago)
• Default Branch: main
• Last Pushed: 2023-06-21T13:57:56.000Z (over 1 year ago)
• Last Synced: 2024-10-15T22:52:49.110Z (about 1 month ago)
• Language: Jupyter Notebook
• Size: 571 KB
• Stars: 167
• Watchers: 5
• Forks: 42
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Apple Silicon DL benchmarks

Currently we have PyTorch and Tensorflow that have Metal backend.

## Results 

Varied results across frameworks:

- [Apple M1Pro Pytorch Training Results](https://wandb.me/pytorch_m1)

- [Apple M1Pro Tensorflow Training Results](https://wandb.me/m1pro)

### Tensorflow Resnet50:

![tf_resnet_50results.png](images/tf_resnet50_results.png)

### PyTorch Resnet50:

- Difference between CPU and GPU

![gpu_vs_cpu.png](images/pt_gpu_vs_cpu.png)

- Comparing with Nvidia

![samples_sec.png](images/pt_samples_sec.png)

### PyTorch Bert

- Running a Bert from Huggingface

![pt_bert.png](images/pt_bert.png)

## Pytorch

We have official PyTorch support! check [pytorch](pytorch) folder to start running your benchmarks

## Tensorflow

You can run tensorflow benchmarks by going to the [tensorflow](tensorflow) folder.