Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
You can not select more than 25 topics
Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
Wh1isper
cbddb2a0e3
|
1 year ago | |
|---|---|---|
| .. | ||
| README.md | 1 year ago | |
| resnet.py | 1 year ago | |
| resnet50_distributed_training.py | 1 year ago | |
| runner.py | 1 year ago | |
| train_cluster.ipynb | 1 year ago | |
| train_cluster_by_sdk.ipynb | 1 year ago | |
README.md
本文件夹下是 Ascend TensorFlow的改造、训练教程
基于MindSpore1.5分布式训练教程 ，参考其中的代码给出在靶场运行的案例
注意，MindSpore的代码遵循Apache License, Version 2.0协议，相关修改已经注明
项目结构
├── mindspore
│ ├── README.md
│ ├── resnet50_distributed_training.py # 分布式训练文件（从MindSpore官方引入）
│ ├── resnet.py # resnet模型（从MindSpore官方引入）
│ ├── runner.py # davincirunsdk简易实现
│ ├── train_cluster_by_sdk.ipynb # 使用davincirunsdk的运行样例
│ └── train_cluster.ipynb # 使用runner简易实现样例
数据集下载
在此 仓库 下载cifar10数据集
如何上传数据集、使用数据集及运行实验参考本仓库根目录下 教程
本仓库提供在AI靶场（https://datai.pcl.ac.cn/）进行分布式训练样例，也可以作为其他使用Jupyter或类似交互式工具在Ascend芯片上进行分布式训练的参考
Python Jupyter Notebook
Contributors (1)
