跳至内容

TFLite,ONNX,CoreML,TensorRT Export

📚 This guide explains how to export a trained YOLOv5 🚀 model from PyTorch to ONNX and TorchScript formats.

开始之前

克隆 repo 并将requirements.txt安装在 Python>=3.8.0环境中安装 requirements txt,包括 PyTorch>=1.8.模型数据集会自动从最新的YOLOv5 版本下载。

git clone https://github.com/ultralytics/yolov5  # clone
cd yolov5
pip install -r requirements.txt  # install

对于 TensorRT 导出示例(需要GPU )请参阅我们的 Colab 笔记本 附录部分。 在 Colab 中打开

格式

YOLOv5 有 11 种格式正式支持推理:

💡 专业提示:导出到ONNX 或OpenVINO ,可将CPU 的速度提高 3 倍。请参见CPU Benchmarks。💡 ProTip:导出到TensorRT ,速度可提高 5 倍GPU 。请参见GPU Benchmarks

格式 export.py --include 模型
PyTorch - yolov5s.pt
TorchScript torchscript yolov5s.torchscript
ONNX onnx yolov5s.onnx
OpenVINO openvino yolov5s_openvino_model/
TensorRT engine yolov5s.engine
CoreML coreml yolov5s.mlmodel
TensorFlow SavedModel saved_model yolov5s_saved_model/
TensorFlow GraphDef pb yolov5s.pb
TensorFlow 轻型 tflite yolov5s.tflite
TensorFlow 边缘TPU edgetpu yolov5s_edgetpu.tflite
TensorFlow.js tfjs yolov5s_web_model/
PaddlePaddle paddle yolov5s_paddle_model/

基准

以下基准测试在配备YOLOv5 教程的 Colab Pro 笔记本电脑上运行 在 Colab 中打开.重现:

python benchmarks.py --weights yolov5s.pt --imgsz 640 --device 0

Colab Pro V100GPU

benchmarks: weights=/content/yolov5/yolov5s.pt, imgsz=640, batch_size=1, data=/content/yolov5/data/coco128.yaml, device=0, half=False, test=False
Checking setup...
YOLOv5 🚀 v6.1-135-g7926afc torch 1.10.0+cu111 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)
Setup complete ✅ (8 CPUs, 51.0 GB RAM, 46.7/166.8 GB disk)

Benchmarks complete (458.07s)
                   Format  mAP@0.5:0.95  Inference time (ms)
0                 PyTorch        0.4623                10.19
1             TorchScript        0.4623                 6.85
2                    ONNX        0.4623                14.63
3                OpenVINO           NaN                  NaN
4                TensorRT        0.4617                 1.89
5                  CoreML           NaN                  NaN
6   TensorFlow SavedModel        0.4623                21.28
7     TensorFlow GraphDef        0.4623                21.22
8         TensorFlow Lite           NaN                  NaN
9     TensorFlow Edge TPU           NaN                  NaN
10          TensorFlow.js           NaN                  NaN

Colab ProCPU

benchmarks: weights=/content/yolov5/yolov5s.pt, imgsz=640, batch_size=1, data=/content/yolov5/data/coco128.yaml, device=cpu, half=False, test=False
Checking setup...
YOLOv5 🚀 v6.1-135-g7926afc torch 1.10.0+cu111 CPU
Setup complete ✅ (8 CPUs, 51.0 GB RAM, 41.5/166.8 GB disk)

Benchmarks complete (241.20s)
                   Format  mAP@0.5:0.95  Inference time (ms)
0                 PyTorch        0.4623               127.61
1             TorchScript        0.4623               131.23
2                    ONNX        0.4623                69.34
3                OpenVINO        0.4623                66.52
4                TensorRT           NaN                  NaN
5                  CoreML           NaN                  NaN
6   TensorFlow SavedModel        0.4623               123.79
7     TensorFlow GraphDef        0.4623               121.57
8         TensorFlow Lite        0.4623               316.61
9     TensorFlow Edge TPU           NaN                  NaN
10          TensorFlow.js           NaN                  NaN

导出训练有素的YOLOv5 模型

此命令将预训练的 YOLOv5s 模型导出为TorchScript 和ONNX 格式。 yolov5s.pt 是 "小型 "机型,是第二小的机型。其他选择包括 yolov5n.pt, yolov5m.pt, yolov5l.ptyolov5x.pt与 P6 对应的是,即 yolov5s6.pt 或您自己定制的培训检查点,即 runs/exp/weights/best.pt.有关所有可用型号的详细信息,请参阅我们的 README。 .

python export.py --weights yolov5s.pt --include torchscript onnx

💡 专业提示:添加 --half to export models at FP16 half precision for smaller file sizes

输出:

export: data=data/coco128.yaml, weights=['yolov5s.pt'], imgsz=[640, 640], batch_size=1, device=cpu, half=False, inplace=False, train=False, keras=False, optimize=False, int8=False, dynamic=False, simplify=False, opset=12, verbose=False, workspace=4, nms=False, agnostic_nms=False, topk_per_class=100, topk_all=100, iou_thres=0.45, conf_thres=0.25, include=['torchscript', 'onnx']
YOLOv5 🚀 v6.2-104-ge3e5122 Python-3.8.0 torch-1.12.1+cu113 CPU

Downloading https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s.pt to yolov5s.pt...
100% 14.1M/14.1M [00:00<00:00, 274MB/s]

Fusing layers...
YOLOv5s summary: 213 layers, 7225885 parameters, 0 gradients

PyTorch: starting from yolov5s.pt with output shape (1, 25200, 85) (14.1 MB)

TorchScript: starting export with torch 1.12.1+cu113...
TorchScript: export success  1.7s, saved as yolov5s.torchscript (28.1 MB)

ONNX: starting export with onnx 1.12.0...
ONNX: export success  2.3s, saved as yolov5s.onnx (28.0 MB)

Export complete (5.5s)
Results saved to /content/yolov5
Detect:          python detect.py --weights yolov5s.onnx
Validate:        python val.py --weights yolov5s.onnx
PyTorch Hub:     model = torch.hub.load('ultralytics/yolov5', 'custom', 'yolov5s.onnx')
Visualize:       https://netron.app/

导出的 3 个模型将与原始PyTorch 模型一起保存:

YOLO 出口地点

建议使用Netron Viewer对导出的模型进行可视化:

YOLO 模型可视化

导出模型使用示例

detect.py 对导出的模型进行推理:

python detect.py --weights yolov5s.pt                 # PyTorch
                           yolov5s.torchscript        # TorchScript
                           yolov5s.onnx               # ONNX Runtime or OpenCV DNN with dnn=True
                           yolov5s_openvino_model     # OpenVINO
                           yolov5s.engine             # TensorRT
                           yolov5s.mlmodel            # CoreML (macOS only)
                           yolov5s_saved_model        # TensorFlow SavedModel
                           yolov5s.pb                 # TensorFlow GraphDef
                           yolov5s.tflite             # TensorFlow Lite
                           yolov5s_edgetpu.tflite     # TensorFlow Edge TPU
                           yolov5s_paddle_model       # PaddlePaddle

val.py 对导出的模型进行验证:

python val.py --weights yolov5s.pt                 # PyTorch
                        yolov5s.torchscript        # TorchScript
                        yolov5s.onnx               # ONNX Runtime or OpenCV DNN with dnn=True
                        yolov5s_openvino_model     # OpenVINO
                        yolov5s.engine             # TensorRT
                        yolov5s.mlmodel            # CoreML (macOS Only)
                        yolov5s_saved_model        # TensorFlow SavedModel
                        yolov5s.pb                 # TensorFlow GraphDef
                        yolov5s.tflite             # TensorFlow Lite
                        yolov5s_edgetpu.tflite     # TensorFlow Edge TPU
                        yolov5s_paddle_model       # PaddlePaddle

将PyTorch Hub 与导出的YOLOv5 模型一起使用:

import torch

# Model
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s.pt")
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s.torchscript ")  # TorchScript
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s.onnx")  # ONNX Runtime
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s_openvino_model")  # OpenVINO
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s.engine")  # TensorRT
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s.mlmodel")  # CoreML (macOS Only)
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s_saved_model")  # TensorFlow SavedModel
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s.pb")  # TensorFlow GraphDef
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s.tflite")  # TensorFlow Lite
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s_edgetpu.tflite")  # TensorFlow Edge TPU
model = torch.hub.load("ultralytics/yolov5", "custom", "yolov5s_paddle_model")  # PaddlePaddle

# Images
img = "https://ultralytics.com/images/zidane.jpg"  # or file, Path, PIL, OpenCV, numpy, list

# Inference
results = model(img)

# Results
results.print()  # or .show(), .save(), .crop(), .pandas(), etc.

OpenCV DNN 推断

OpenCV inference with ONNX models:

python export.py --weights yolov5s.pt --include onnx

python detect.py --weights yolov5s.onnx --dnn  # detect
python val.py --weights yolov5s.onnx --dnn  # validate

C++ 推理

YOLOv5 OpenCV DNN C++ 对导出的ONNX 模型示例进行推理:

YOLOv5 OpenVINO C++ 推理示例

TensorFlow.js 网络浏览器推理

支持的环境

Ultralytics 提供了一系列随时可用的环境,每个环境都预装了基本的依赖项,如 CUDACUDNNPythonPyTorch等基本依赖项,以便启动您的项目。

项目现状

YOLOv5 CI

此徽章表示YOLOv5 GitHub Actions 的所有持续集成(CI)测试均已成功通过。这些 CI 测试严格检查了YOLOv5 在训练验证推理导出基准等多个关键方面的功能和性能。它们确保在 macOS、Windows 和 Ubuntu 上运行的一致性和可靠性,每 24 小时和每次新提交时都会进行一次测试。


📅 Created 11 months ago ✏️ Updated 17 days ago

评论