Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
lizhenyu 2e35802a5c | 10 months ago | |
---|---|---|
.. | ||
wide_deep | 10 months ago | |
README.md | 1 year ago |
CPU:Inter(R) Xeon(R) Platinum 8160@2.10GHz
GPU: Tesla V100-SXM2-16GB * 8
MindSpore v1.8
CUDA Version: 10.1
Ubuntu 16.04
vocab_size: 5860000
deep_embedding_vec_size: 16
hidden_units_num: 7,output dim: 1024
criteo(one day)
1. 克隆代码
git clone https://gitee.com/mindspore/recommender.git
cd benchmarks/wide_deep
2. 下载数据集并进行格式转换
详见datasets目录,使用对应脚本预处理数据,生成的MindRecord数据并存放在./data/mindrecord下。
3. 开始训练
python train_and_eval.py --data_path=./data/mindrecord --device_target="GPU"
或
bash scripts/run_standalone_train_for_gpu.sh 5 ./data/mindrecord
日志中,per step time: xxx ms,表示单step平均性能为xxx ms
bash scripts/run_distribute_train_for_gpu.sh 4 5 ./data/mindrecord 127.0.0.1 2898
训练日志保存在./worker_*/内,日志中,per step time: xxx ms,表示单step平均性能为xxx ms
bash scripts/run_distribute_train_for_gpu.sh 8 5 ./data/mindrecord 127.0.0.1 2898
训练日志保存在./worker_*/内,日志中,per step time: xxx ms,表示单step平均性能为xxx ms
GPU Num | Throughput |
---|---|
1p | 267,558 |
4p | 767,663 |
8p | 1,163,636 |
MindSpore large-scale recommender system library.
Python Shell C++ Markdown Text other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》