|
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189 |
- # Contents
-
- - [MobileNetV2 Description](#mobilenetv2-description)
- - [Model Architecture](#model-architecture)
- - [Dataset](#dataset)
- - [Features](#features)
- - [Mixed Precision](#mixed-precision)
- - [Environment Requirements](#environment-requirements)
- - [Script Description](#script-description)
- - [Script and Sample Code](#script-and-sample-code)
- - [Training Process](#training-process)
- - [Evaluation Process](#evaluation-process)
- - [Model Description](#model-description)
- - [Performance](#performance)
- - [Training Performance](#training-performance)
- - [Evaluation Performance](#evaluation-performance)
- - [Description of Random Situation](#description-of-random-situation)
- - [ModelZoo Homepage](#modelzoo-homepage)
-
- # [MobileNetV2 Description](#contents)
-
-
- MobileNetV2 is tuned to mobile phone CPUs through a combination of hardware- aware network architecture search (NAS) complemented by the NetAdapt algorithm and then subsequently improved through novel architecture advances.Nov 20, 2019.
-
- [Paper](https://arxiv.org/pdf/1905.02244) Howard, Andrew, Mark Sandler, Grace Chu, Liang-Chieh Chen, Bo Chen, Mingxing Tan, Weijun Wang et al. "Searching for MobileNetV2." In Proceedings of the IEEE International Conference on Computer Vision, pp. 1314-1324. 2019.
-
- This is the quantitative network of MobileNetV2.
-
- # [Model architecture](#contents)
-
- The overall network architecture of MobileNetV2 is show below:
-
- [Link](https://arxiv.org/pdf/1905.02244)
-
- # [Dataset](#contents)
-
- Dataset used: [imagenet](http://www.image-net.org/)
-
- - Dataset size: ~125G, 1.2W colorful images in 1000 classes
- - Train: 120G, 1.2W images
- - Test: 5G, 50000 images
- - Data format: RGB images.
- - Note: Data will be processed in src/dataset.py
-
-
- # [Features](#contents)
-
- ## [Mixed Precision](#contents)
-
- The [mixed precision](https://www.mindspore.cn/tutorial/zh-CN/master/advanced_use/mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware.
- For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’.
-
- # [Environment Requirements](#contents)
-
- - Hardware:Ascend
- - Prepare hardware environment with Ascend. If you want to try Ascend , please send the [application form](https://obs-9be7.obs.cn-east-2.myhuaweicloud.com/file/other/Ascend%20Model%20Zoo%E4%BD%93%E9%AA%8C%E8%B5%84%E6%BA%90%E7%94%B3%E8%AF%B7%E8%A1%A8.docx) to ascend@huawei.com. Once approved, you can get the resources.
- - Framework
- - [MindSpore](https://www.mindspore.cn/install/en)
- - For more information, please check the resources below
- - [MindSpore tutorials](https://www.mindspore.cn/tutorial/zh-CN/master/index.html)
- - [MindSpore API](https://www.mindspore.cn/api/zh-CN/master/index.html)
-
-
- # [Script description](#contents)
-
- ## [Script and sample code](#contents)
-
- ```python
- ├── mobileNetv2_quant
- ├── Readme.md # descriptions about MobileNetV2-Quant
- ├── scripts
- │ ├──run_train.sh # shell script for train on Ascend
- │ ├──run_infer.sh # shell script for evaluation on Ascend
- ├── src
- │ ├──config.py # parameter configuration
- │ ├──dataset.py # creating dataset
- │ ├──launch.py # start python script
- │ ├──lr_generator.py # learning rate config
- │ ├──mobilenetV2.py # MobileNetV2 architecture
- │ ├──utils.py # supply the monitor module
- ├── train.py # training script
- ├── eval.py # evaluation script
- ├── export.py # export checkpoint files into air/onnx
- ```
-
- ## [Training process](#contents)
-
- ### Usage
-
-
- You can start training using python or shell scripts. The usage of shell scripts as follows:
-
- - bash run_train.sh [Ascend] [RANK_TABLE_FILE] [DATASET_PATH] [PRETRAINED_CKPT_PATH]\(optional)
- - bash run_train.sh [GPU] [DEVICE_ID_LIST] [DATASET_PATH] [PRETRAINED_CKPT_PATH]\(optional)
-
-
- ### Launch
-
- ``` bash
- # training example
- >>> bash run_train.sh Ascend ~/hccl_4p_0123_x.x.x.x.json ~/imagenet/train/ ~/mobilenet.ckpt
- >>> bash run_train.sh GPU 1,2 ~/imagenet/train/ ~/mobilenet.ckpt
- ```
-
- ### Result
-
- Training result will be stored in the example path. Checkpoints trained by `Ascend` will be stored at `./train/device$i/checkpoint` by default, and training log will be redirected to `./train/device$i/train.log`. Checkpoints trained by `GPU` will be stored in `./train/checkpointckpt_$i` by default, and training log will be redirected to `./train/train.log`.
- `train.log` is as follows:
-
- ```
- epoch: [ 0/200], step:[ 624/ 625], loss:[5.258/5.258], time:[140412.236], lr:[0.100]
- epoch time: 140522.500, per step time: 224.836, avg loss: 5.258
- epoch: [ 1/200], step:[ 624/ 625], loss:[3.917/3.917], time:[138221.250], lr:[0.200]
- epoch time: 138331.250, per step time: 221.330, avg loss: 3.917
- ```
-
- ## [Evaluation process](#contents)
-
- ### Usage
-
- You can start training using python or shell scripts. The usage of shell scripts as follows:
-
- - Ascend: sh run_infer_quant.sh Ascend [DATASET_PATH] [CHECKPOINT_PATH]
-
- ### Launch
-
- ```
- # infer example
- shell:
- Ascend: sh run_infer_quant.sh Ascend ~/imagenet/val/ ~/train/mobilenet-60_1601.ckpt
- ```
-
- > checkpoint can be produced in training process.
-
- ### Result
-
- Inference result will be stored in the example path, you can find result like the followings in `./val/infer.log`.
-
- ```
- result: {'acc': 0.71976314102564111}
- ```
-
- # [Model description](#contents)
-
- ## [Performance](#contents)
-
- ### Training Performance
-
- | Parameters | MobilenetV2 |
- | -------------------------- | ---------------------------------------------------------- |
- | Model Version | V2 |
- | Resource | Ascend 910, cpu:2.60GHz 56cores, memory:314G |
- | uploaded Date | 06/06/2020 |
- | MindSpore Version | 0.3.0 |
- | Dataset | ImageNet |
- | Training Parameters | src/config.py |
- | Optimizer | Momentum |
- | Loss Function | SoftmaxCrossEntropy |
- | outputs | ckpt file |
- | Loss | 1.913 |
- | Accuracy | |
- | Total time | 16h |
- | Params (M) | batch_size=192, epoch=60 |
- | Checkpoint for Fine tuning | |
- | Model for inference | |
-
- #### Evaluation Performance
-
- | Parameters | |
- | -------------------------- | ----------------------------- |
- | Model Version | V2 |
- | Resource | Ascend 910 |
- | uploaded Date | 06/06/2020 |
- | MindSpore Version | 0.3.0 |
- | Dataset | ImageNet, 1.2W |
- | batch_size | 130(8P) |
- | outputs | probability |
- | Accuracy | ACC1[71.78%] ACC5[90.90%] |
- | Speed | 200ms/step |
- | Total time | 5min |
- | Model for inference | |
-
- # [Description of Random Situation](#contents)
-
- In dataset.py, we set the seed inside “create_dataset" function. We also use random seed in train.py.
-
- # [ModelZoo Homepage](#contents)
-
- Please check the official [homepage](https://gitee.com/mindspore/mindspore/tree/master/model_zoo).
|
