models/community/cv/snn

• Contents
• SNN Description
• • Description
  • Paper
• Model Architecture
• Dataset
• Environment Requirements
• Quick Start
• Script Description
• • Script and Sample Code
  • Script Parameters
  • Training Process
  • • Usage
    • • Running on Ascend
  • Resume Process
  • • Usage
    • • Running on Ascend
    • Result
  • Evaluation Process
  • • Usage
    • • Running on Ascend
    • Result
• Model Description
• • Performance
  • • Evaluation Performance
    • • LeNet on CIFAR-10
• Description of Random Situation
• ModelZoo Homepage

Contents

• Contents
• SNN Description
  • Description
  • Paper
• Model Architecture
• Dataset
• Environment Requirements
• Quick Start
• Script Description
  • Script and Sample Code
  • Script Parameters
  • Training Process
    • Usage
      • Running on Ascend
      • Running parameter server mode training
      • Evaluation while training
    • Result
  • Evaluation Process
    • Usage
      • Running on Ascend
    • Result
• Model Description
  • Performance
    • Evaluation Performance
      • LeNet on CIFAR-10
• Description of Random Situation
• ModelZoo Homepage

SNN Description

Description

SNN (Spiking neural networks) was proposed by Kaushik Roy and other two authors.

These are examples of training LeNet/ResNet50 with CIFAR-10 dataset.

Paper

1.paper:Kaushik Roy, Akhilesh Jaiswal, Priyadarshini Panda. "Towards spike-based machine intelligence with neuromorphic computing"

Model Architecture

The overall network architecture of SNN is show below:
Link

Dataset

Dataset used: CIFAR-10

• Dataset size：60,000 32*32 colorful images in 10 classes
  • Train：50,000 images
  • Test： 10,000 images
• Data format：binary files
  • Note：Data will be processed in dataset.py
• Download the dataset, the directory structure is as follows:

├─cifar-10-batches-bin
│
└─cifar-10-verify-bin

Environment Requirements

• Hardware（Ascend）
  • Prepare hardware environment with Ascend processor.
• Framework
  • MindSpore
• For more information, please check the resources below：
  • MindSpore Tutorials
  • MindSpore Python API

Quick Start

After installing MindSpore via the official website, you can start training and evaluation as follows:

• During training, if CIFAR-10 dataset is used, DATASET_PATH={CIFAR-10 directory}/cifar-10-batches-bin;
• During evaluating and inferring, if CIFAR-10 dataset is used, DATASET_PATH={CIFAR-10 directory}/cifar-10-verify-bin;

• Running on Ascend

# distributed training
Usage: bash run_distribute_train_ascend.sh [RANK_TABLE_FILE] [DATASET_PATH] [CONFIG_PATH] [PRETRAINED_CKPT_PATH](optional)

# standalone training
Usage: bash run_standalone_train_ascend.sh [DATASET_PATH] [CONFIG_PATH] [PRETRAINED_CKPT_PATH](optional)

# run evaluation example
Usage: bash run_eval.sh [DATASET_PATH] [CHECKPOINT_PATH] [CONFIG_PATH]

# infer example
python eval.py --data_path=[DATASET_PATH] --ckpt_path=[CHECKPOINT_PATH] --config_path [CONFIG_PATH]

If you want to run in modelarts, please check the official documentation of modelarts, and you can start training and evaluation as follows:

# run distributed training on modelarts example
# (1) Add "config_path='/path_to_code/config/snn_lenet_cifar10_config.yaml'" on the website UI interface.
# (2) First, Perform a or b.
#       a. Set "enable_modelarts=True" on yaml file.
#          Set other parameters on yaml file you need.
#       b. Add "enable_modelarts=True" on the website UI interface.
#          Add other parameters on the website UI interface.
# (3) Set the code directory to "/path/snn" on the website UI interface.
# (4) Set the startup file to "train.py" on the website UI interface.
# (5) Set the "Dataset path" and "Output file path" and "Job log path" to your path on the website UI interface.
# (6) Create your job.

# run evaluation on modelarts example
# (1) Add "config_path='/path_to_code/config/snn_lenet_cifar10_config.yaml'" on the website UI interface.
# (2) Copy or upload your trained model to S3 bucket.
# (3) Perform a or b.
#       a. Set "enable_modelarts=True" on yaml file.
#          Set "checkpoint_file_path='/cache/checkpoint_path/model.ckpt'" on yaml file.
#          Set "checkpoint_url=/The path of checkpoint in S3/" on yaml file.
#       b. Add "enable_modelarts=True" on the website UI interface.
#          Add "checkpoint_file_path='/cache/checkpoint_path/model.ckpt'" on the website UI interface.
#          Add "checkpoint_url=/The path of checkpoint in S3/" on the website UI interface.
# (4) Set the code directory to "/path/snn" on the website UI interface.
# (5) Set the startup file to "eval.py" on the website UI interface.
# (6) Set the "Dataset path" and "Output file path" and "Job log path" to your path on the website UI interface.
# (7) Create your job.

Script Description

Script and Sample Code

.
└──snn
  ├── README.md
  ├── config                               # parameter configuration
    ├── snn_lenet_cifar10_config.yaml
    ├── snn_resnet50_cifar10_config.yaml
  ├── scripts
    ├── run_distribute_train_ascend.sh     # launch ascend distributed training(8 pcs)
    ├── run_eval.sh                        # launch ascend evaluation
    ├── run_standalone_train_ascend.sh     # launch ascend standalone training(1 pcs)
  ├── src
    ├── dataset.py                         # data preprocessing
    ├── ifnode.py                          # ifnode cell for snn
    ├── lr_generator.py                    # generate learning rate for each step
    ├── snn_lenet.py                       # lenet_snn for ascend benchmark
    ├── snn_resnet.py                      # resnet50_snn for ascend benchmark
    ├── model_utils
       ├──config.py                        # parameter configuration
       ├──device_adapter.py                # device adapter
       ├──local_adapter.py                 # local adapter
       ├──moxing_adapter.py                # moxing adapter
  ├── eval.py                              # eval net
  └── train.py                             # train net

Script Parameters

Parameters for both training and evaluation can be set in config file.

• Config for LeNet and ResNet50, CIFAR-10 dataset

"class_num": 10,                  # dataset class num
"batch_size": 32,                 # batch size of input tensor
"loss_scale": 1024,               # loss scale
"momentum": 0.9,                  # momentum
"weight_decay": 1e-4,             # weight decay
"epoch_size": 5,                  # only valid for taining, which is always 1 for inference
"save_checkpoint": True,          # whether save checkpoint or not
"save_checkpoint_epochs": 1,      # the epoch interval between two checkpoints. By default, the last checkpoint will be saved after the last step
"keep_checkpoint_max": 5,         # only keep the last keep_checkpoint_max checkpoint
"warmup_epochs": 5,               # number of warmup epoch
"lr_init": 0.001,                 # initial learning rate
"save_graphs": False,             # save graph results

Training Process

Usage

Running on Ascend

# distributed training
Usage: bash run_distribute_train_ascend.sh [RANK_TABLE_FILE] [DATASET_PATH] [CONFIG_PATH] [PRETRAINED_CKPT_PATH](optional)

# standalone training
Usage: bash run_standalone_train_ascend.sh [DATASET_PATH] [CONFIG_PATH] [PRETRAINED_CKPT_PATH](optional)

# run evaluation example
Usage: bash run_eval.sh [DATASET_PATH] [CHECKPOINT_PATH] [CONFIG_PATH]

For distributed training, a hccl configuration file with JSON format needs to be created in advance.

Please follow the instructions in the link hccn_tools.

Training result will be stored in the example path, whose folder name begins with "train" or "train_parallel". Under this, you can find checkpoint file together with result like the following in log.

If you want to change device_id for standalone training, you can set environment variable export DEVICE_ID=x or set device_id=x in context.

Resume Process

Usage

Running on Ascend

# distributed training
Usage: bash run_distribute_train_ascend.sh [RANK_TABLE_FILE] [DATASET_PATH] [CONFIG_PATH] [PRETRAINED_CKPT_PATH](optional)

# standalone training
Usage: bash run_standalone_train_ascend.sh [DATASET_PATH] [CONFIG_PATH] [PRETRAINED_CKPT_PATH](optional)

Result

• Training LeNet with CIFAR-10 dataset

# training result on GRAPH mode(1 pcs)
epoch: 2, step: 250, loss is loss:0.090041, epoch time: 60457.911ms, per step time: 241.832ms
epoch: 3, step: 250, loss is loss:0.088589, epoch time: 60414.800ms, per step time: 241.659ms
epoch: 4, step: 250, loss is loss:0.078889, epoch time: 60488.454ms, per step time: 241.954ms
epoch: 5, step: 250, loss is loss:0.072030, epoch time: 60465.275ms, per step time: 241.861ms

• Training ResNet50 with CIFAR-10 dataset

# training result on GRAPH mode(1 pcs)
epoch: 2, step: 1562, loss is loss:2.575632, epoch time: 106556.557ms, per step time: 68.218ms
epoch: 3, step: 1562, loss is loss:2.307327, epoch time: 106474.296ms, per step time: 68.165ms
epoch: 4, step: 1562, loss is loss:2.308245, epoch time: 106434.503ms, per step time: 68.140ms
epoch: 5, step: 1562, loss is loss:2.309555, epoch time: 108063.124ms, per step time: 69.183ms

Evaluation Process

Usage

Running on Ascend

# evaluation
Usage: bash run_eval.sh [DATASET_PATH] [CHECKPOINT_PATH] [CONFIG_PATH]

checkpoint can be produced in training process.

Result

Evaluation result will be stored in the example path, whose folder name is "eval". Under this, you can find result like the following in log.

• Evaluating LeNet with CIFAR-10 dataset

result: {'acc': 59.5400 %} ckpt=~/snn/train/output/checkpoint/lenet-5_250.ckpt

Model Description

Performance

Evaluation Performance

LeNet on CIFAR-10

Parameters	Ascend 910
Model Version	LeNet
Resource	Ascend 910; CPU 2.60GHz, 192cores; Memory 755G; OS Euler2.8
uploaded Date	06/30/2022 (month/day/year)
MindSpore Version	1.8.0
Dataset	CIFAR-10
Training Parameters	epoch=5, steps per epoch=250, batch_size = 200
Optimizer	Adam
Loss Function	MSE
outputs	probability
Loss	0.072030
Speed	241 ms/step（1pcs)
Total time	7 mins
Checkpoint for Fine tuning	1.3M (.ckpt file)
Accuracy	59.54%
config	Link

Description of Random Situation

In dataset.py, we set the seed inside “create_dataset" function. We also use random seed in train.py.

ModelZoo Homepage

Please check the official homepage.