Swin-T-使用TorchServe部署SwinTransformer目标检测算法-附项目源码-优质项目实战.zip

# Swin-Transformer-Serve Deploy Pre-trained [Swin-Transformer](https://github.com/microsoft/Swin-Transformer/) classifier trained on ImageNet 1K using [TorchServe](https://github.com/pytorch/serve) ### Create and activate virtual env ```bash virtualenv env --python=python3 source env/bin/activate ``` ### Install dependencies ```bash cd project # clone TorchServe git clone https://github.com/pytorch/serve.git cd serve # Refer to readme in serve repo for CUDA enabled local installtion # The instruction below is for CPU python ./ts_scripts/install_dependencies.py pip install torchserve torch-model-archiver ``` ### Pretrained Weights and Config #### Download weights. | name | pretrain | resolution | acc@1 | acc@5 | 1K model | |--------|--------------|------------|-------|-------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | Swin-T | ImageNet-1K | 224x224 | 81.2 | 95.5 | [github](https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_tiny_patch4_window7_224.pth)/[baidu](https://pan.baidu.com/s/156nWJy4Q28rDlrX-rRbI3w) | | Swin-S | ImageNet-1K | 224x224 | 83.2 | 96.2 | [github](https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_small_patch4_window7_224.pth)/[baidu](https://pan.baidu.com/s/1KFjpj3Efey3LmtE1QqPeQg) | | Swin-B | ImageNet-1K | 224x224 | 83.5 | 96.5 | [github](https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_base_patch4_window7_224.pth)/[baidu](https://pan.baidu.com/s/16bqCTEc70nC_isSsgBSaqQ) | | Swin-B | ImageNet-1K | 384x384 | 84.5 | 97.0 | [github](https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_base_patch4_window12_384.pth)/[baidu](https://pan.baidu.com/s/1xT1cu740-ejW7htUdVLnmw) | | Swin-B | ImageNet-22K | 224x224 | 85.2 | 97.5 | [github](https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_base_patch4_window7_224_22kto1k.pth)/[baidu](https://pan.baidu.com/s/1n_wNkcbRxVXit8r_KrfAVg) | | Swin-B | ImageNet-22K | 384x384 | 86.4 | 98.0 | [github](https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_base_patch4_window12_384_22kto1k.pth)/[baidu](https://pan.baidu.com/s/1caKTSdoLJYoi4WBcnmWuWg) | | Swin-L | ImageNet-22K | 224x224 | 86.3 | 97.9 | [github](https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_large_patch4_window7_224_22kto1k.pth)/[baidu](https://pan.baidu.com/s/1NkQApMWUhxBGjk1ne6VqBQ) | | Swin-L | ImageNet-22K | 384x384 | 87.3 | 98.2 | [github](https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_large_patch4_window12_384_22kto1k.pth)/[baidu](https://pan.baidu.com/s/1X0FLHQyPOC6Kmv2CmgxJvA) | For demo we will use Tiny model. ```bash wget https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_tiny_patch4_window7_224.pth -O weights/swin_tiny_patch4_window7_224.pth ``` Copy the corresonding model config file from to `swin_config.yaml` ```bash cp configs/swin_tiny_patch4_window7_224.yaml swin_config.yaml ``` ### TorchServe Create MAR file using [torch-model-archiver](https://github.com/pytorch/serve/tree/master/model-archiver). ```bash torch-model-archiver --model-name swin -v 1.0 --model-file swin_transformer.py --serialized-file weights/swin_tiny_patch4_window7_224.pth --handler swin_handler.py --extra-files index_to_name.json,swin_config.yaml --requirements-file requirements.txt # move the swin.mar to model_store folder mv swin.mar model_store ``` Start the `torchserve` using below cmd in the config.properties we have set the configurations.[Refer](https://github.com/pytorch/serve/blob/master/docs/configuration.md) ```bash torchserve --start --model-store model_store --models swin=swin.mar ``` test prediction using cURL ```bash curl http://127.0.0.1:8080/predictions/swin -T kitten_small.jpg ``` ```bash { "tabby": 0.44951513409614563, "tiger_cat": 0.1962115466594696, "lynx": 0.16013166308403015, "Egyptian_cat": 0.08244507014751434, "tiger": 0.015334611758589745 } ``` As we set in the configuration file only one worker will be created per model. To increase workers and batch inference can done using [Management API](https://github.com/pytorch/serve/blob/master/docs/management_api.md) #### Increase worker ```bash curl -v -X PUT "http://localhost:8081/models/swin?min_worker=3" ``` This will increase minimum workers to 3. #### Batch Inference. [Reference](https://github.com/pytorch/serve/blob/master/docs/batch_inference_with_ts.md) ```bash #if the model is already running dergister it first curl -X DELETE "http://localhost:8081/models/swin/1.0" # The following command will register a model "swin.mar" and configure TorchServe to use a batch_size of 8 and a max batch delay of 50 milli seconds. curl -X POST "localhost:8081/models?url=swin.mar&batch_size=8&max_batch_delay=50" ``` Stop torchserve ```bash torchserve --stop ``` ## Model deploy using docker Update `config.properties` file with following lines ``` install_py_dep_per_model=true default_workers_per_model=1 ``` Execute the below cmds to build a `cpu` docker image ```bash cd serve/docker sudo ./build_image.sh ``` `gpu` docker image ```bash sudo ./build_image.sh -g -cv cu102 ``` On successful docker build ##### CPU ```bash cd Swin-Transformer-Serve sudo docker run --rm -it -p 8080:8080 -p 8081:8081 -p 8082:8082 -p 7070:7070 -p 7071:7071 -v $(pwd)/model_store:/home/model-server/model-store pytorch/torchserve:latest-cpu ``` ##### GPU ```bash cd Swin-Transformer-Serve sudo docker run --rm -it --gpus all -p 8080:8080 -p 8081:8081 -p 8082:8082 -p 7070:7070 -p 7071:7071 -v $(pwd)/model_store:/home/model-server/model-store pytorch/torchserve:latest-gpu ``` #### Register model ```bash curl -X POST "localhost:8081/models?url=swin.mar&batch_size=1&max_batch_delay=50" ``` ### Creating mar file for torchscript mode model #### Generate serialized-file using TorchScript ```python import torch from swin_handler import get_config from swin_transformer import SwinTransformer config = get_config("swin_config.yaml") model = SwinTransformer(img_size=config.DATA.IMG_SIZE, patch_size=config.MODEL.SWIN.PATCH_SIZE, in_chans=config.MODEL.SWIN.IN_CHANS, num_classes=config.MODEL.NUM_CLASSES, embed_dim=config.MODEL.SWIN.EMBED_DIM, depths=config.MODEL.SWIN.DEPTHS, num_heads=config.MODEL.SWIN.NUM_HEADS, window_size=config.MODEL.SWIN.WINDOW_SIZE, mlp_ratio=config.MODEL.SWIN.MLP_RATIO, qkv_bias=config.MODEL.SWIN.QKV_BIAS, qk_scale=config.MODEL.SWIN.QK_SCALE, drop_rate=config.MODEL.DROP_RATE, drop_path_rate=config.MODEL.DROP_PATH_RATE, ape=config.MODEL.SWIN.APE, patch_norm=config.MODEL.SWIN.PATCH_NORM, use_checkpoint=config.TRAIN.USE_CHECKPOINT) model.load_state_dict(torch.load("weights/swin_tiny_patch4_window7_224.pth",map_location="cpu")["model"]) model.eval() example_input = torch.rand(1, 3, config.DATA.IMG_SIZE, config.DATA.IMG_SIZE) traced_script_module = torch.jit.trace(model, example_input) traced_script_module.save("swin.pt") ``` #### TorchServe ```bash torch-model-archiver --model-name swin -v 1.0 --serialized-file weights/swin_tiny_patch4_window7_224.pth --handler swin_handler.py --extra-files index_to_name.json,swin_conf