mirror of
https://github.com/THU-MIG/yolov10.git
synced 2025-05-23 05:24:22 +08:00
ultralytics 8.0.150 new Apple *.mlpackage export format (#4043)
Co-authored-by: MLBoy_DaisukeMajima <rockyshikoku@gmail.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Louis Lac <lac.louis5@gmail.com>
This commit is contained in:
Glenn Jocher
GitHub
parent
f5fc807fa7
commit
7dfdb63cde
6
.github/workflows/ci.yaml
vendored
6
.github/workflows/ci.yaml
vendored
@ -111,7 +111,7 @@ jobs:
|
||||
else
|
||||
pip install -e ".[export]" --extra-index-url https://download.pytorch.org/whl/cpu
|
||||
fi
|
||||
yolo export format=tflite imgsz=32
|
||||
yolo export format=tflite imgsz=32 || true
|
||||
- name: Check environment
|
||||
run: |
|
||||
echo "RUNNER_OS is ${{ runner.os }}"
|
||||
@ -165,9 +165,9 @@ jobs:
|
||||
run: | # CoreML must be installed before export due to protobuf error from AutoInstall
|
||||
python -m pip install --upgrade pip wheel
|
||||
if [ "${{ matrix.torch }}" == "1.8.0" ]; then
|
||||
pip install -e . torch==1.8.0 torchvision==0.9.0 pytest "coremltools>=6.0,<=6.2" --extra-index-url https://download.pytorch.org/whl/cpu
|
||||
pip install -e . torch==1.8.0 torchvision==0.9.0 pytest "coremltools>=7.0.b1" --extra-index-url https://download.pytorch.org/whl/cpu
|
||||
else
|
||||
pip install -e . pytest "coremltools>=6.0,<=6.2" --extra-index-url https://download.pytorch.org/whl/cpu
|
||||
pip install -e . pytest "coremltools>=7.0.b1" --extra-index-url https://download.pytorch.org/whl/cpu
|
||||
fi
|
||||
- name: Check environment
|
||||
run: |
|
||||
|
||||
1
.gitignore
vendored
1
.gitignore
vendored
@ -147,6 +147,7 @@ weights/
|
||||
*.onnx
|
||||
*.engine
|
||||
*.mlmodel
|
||||
*.mlpackage
|
||||
*.torchscript
|
||||
*.tflite
|
||||
*.h5
|
||||
|
||||
@ -42,20 +42,20 @@ Welcome to the Ultralytics Integrations page! This page provides an overview of
|
||||
|
||||
We also support a variety of model export formats for deployment in different environments. Here are the available formats:
|
||||
|
||||
| Format | `format` Argument | Model | Metadata | Arguments |
|
||||
|-------------------------------------------------------------------|-------------------|---------------------------|----------|-----------------------------------------------------|
|
||||
| [PyTorch](https://pytorch.org/) | - | `yolov8n.pt` | ✅ | - |
|
||||
| [TorchScript](https://pytorch.org/docs/stable/jit.html) | `torchscript` | `yolov8n.torchscript` | ✅ | `imgsz`, `optimize` |
|
||||
| [ONNX](https://onnx.ai/) | `onnx` | `yolov8n.onnx` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `opset` |
|
||||
| [OpenVINO](openvino.md) | `openvino` | `yolov8n_openvino_model/` | ✅ | `imgsz`, `half` |
|
||||
| [TensorRT](https://developer.nvidia.com/tensorrt) | `engine` | `yolov8n.engine` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `workspace` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlmodel` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [TF SavedModel](https://www.tensorflow.org/guide/saved_model) | `saved_model` | `yolov8n_saved_model/` | ✅ | `imgsz`, `keras` |
|
||||
| Format | `format` Argument | Model | Metadata | Arguments |
|
||||
|--------------------------------------------------------------------|-------------------|---------------------------|----------|-----------------------------------------------------|
|
||||
| [PyTorch](https://pytorch.org/) | - | `yolov8n.pt` | ✅ | - |
|
||||
| [TorchScript](https://pytorch.org/docs/stable/jit.html) | `torchscript` | `yolov8n.torchscript` | ✅ | `imgsz`, `optimize` |
|
||||
| [ONNX](https://onnx.ai/) | `onnx` | `yolov8n.onnx` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `opset` |
|
||||
| [OpenVINO](openvino.md) | `openvino` | `yolov8n_openvino_model/` | ✅ | `imgsz`, `half` |
|
||||
| [TensorRT](https://developer.nvidia.com/tensorrt) | `engine` | `yolov8n.engine` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `workspace` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlpackage` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [TF SavedModel](https://www.tensorflow.org/guide/saved_model) | `saved_model` | `yolov8n_saved_model/` | ✅ | `imgsz`, `keras` |
|
||||
| [TF GraphDef](https://www.tensorflow.org/api_docs/python/tf/Graph) | `pb` | `yolov8n.pb` | ❌ | `imgsz` |
|
||||
| [TF Lite](https://www.tensorflow.org/lite) | `tflite` | `yolov8n.tflite` | ✅ | `imgsz`, `half`, `int8` |
|
||||
| [TF Edge TPU](https://coral.ai/docs/edgetpu/models-intro/) | `edgetpu` | `yolov8n_edgetpu.tflite` | ✅ | `imgsz` |
|
||||
| [TF.js](https://www.tensorflow.org/js) | `tfjs` | `yolov8n_web_model/` | ✅ | `imgsz` |
|
||||
| [PaddlePaddle](https://github.com/PaddlePaddle) | `paddle` | `yolov8n_paddle_model/` | ✅ | `imgsz` |
|
||||
| [NCNN](https://github.com/Tencent/ncnn) | `ncnn` | `yolov8n_ncnn_model/` | ✅ | `imgsz`, `half` |
|
||||
| [TF Lite](https://www.tensorflow.org/lite) | `tflite` | `yolov8n.tflite` | ✅ | `imgsz`, `half`, `int8` |
|
||||
| [TF Edge TPU](https://coral.ai/docs/edgetpu/models-intro/) | `edgetpu` | `yolov8n_edgetpu.tflite` | ✅ | `imgsz` |
|
||||
| [TF.js](https://www.tensorflow.org/js) | `tfjs` | `yolov8n_web_model/` | ✅ | `imgsz` |
|
||||
| [PaddlePaddle](https://github.com/PaddlePaddle) | `paddle` | `yolov8n_paddle_model/` | ✅ | `imgsz` |
|
||||
| [NCNN](https://github.com/Tencent/ncnn) | `ncnn` | `yolov8n_ncnn_model/` | ✅ | `imgsz`, `half` |
|
||||
|
||||
Explore the links to learn more about each integration and how to get the most out of them with Ultralytics.
|
||||
|
||||
@ -64,7 +64,7 @@ Benchmarks will attempt to run automatically on all possible export formats belo
|
||||
| [ONNX](https://onnx.ai/) | `onnx` | `yolov8n.onnx` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `opset` |
|
||||
| [OpenVINO](https://docs.openvino.ai/latest/index.html) | `openvino` | `yolov8n_openvino_model/` | ✅ | `imgsz`, `half` |
|
||||
| [TensorRT](https://developer.nvidia.com/tensorrt) | `engine` | `yolov8n.engine` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `workspace` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlmodel` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlpackage` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [TF SavedModel](https://www.tensorflow.org/guide/saved_model) | `saved_model` | `yolov8n_saved_model/` | ✅ | `imgsz`, `keras` |
|
||||
| [TF GraphDef](https://www.tensorflow.org/api_docs/python/tf/Graph) | `pb` | `yolov8n.pb` | ❌ | `imgsz` |
|
||||
| [TF Lite](https://www.tensorflow.org/lite) | `tflite` | `yolov8n.tflite` | ✅ | `imgsz`, `half`, `int8` |
|
||||
|
||||
@ -78,7 +78,7 @@ i.e. `format='onnx'` or `format='engine'`.
|
||||
| [ONNX](https://onnx.ai/) | `onnx` | `yolov8n.onnx` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `opset` |
|
||||
| [OpenVINO](https://docs.openvino.ai/latest/index.html) | `openvino` | `yolov8n_openvino_model/` | ✅ | `imgsz`, `half` |
|
||||
| [TensorRT](https://developer.nvidia.com/tensorrt) | `engine` | `yolov8n.engine` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `workspace` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlmodel` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlpackage` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [TF SavedModel](https://www.tensorflow.org/guide/saved_model) | `saved_model` | `yolov8n_saved_model/` | ✅ | `imgsz`, `keras` |
|
||||
| [TF GraphDef](https://www.tensorflow.org/api_docs/python/tf/Graph) | `pb` | `yolov8n.pb` | ❌ | `imgsz` |
|
||||
| [TF Lite](https://www.tensorflow.org/lite) | `tflite` | `yolov8n.tflite` | ✅ | `imgsz`, `half`, `int8` |
|
||||
|
||||
@ -169,7 +169,7 @@ i.e. `yolo predict model=yolov8n-cls.onnx`. Usage examples are shown for your mo
|
||||
| [ONNX](https://onnx.ai/) | `onnx` | `yolov8n-cls.onnx` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `opset` |
|
||||
| [OpenVINO](https://docs.openvino.ai/latest/index.html) | `openvino` | `yolov8n-cls_openvino_model/` | ✅ | `imgsz`, `half` |
|
||||
| [TensorRT](https://developer.nvidia.com/tensorrt) | `engine` | `yolov8n-cls.engine` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `workspace` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n-cls.mlmodel` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n-cls.mlpackage` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [TF SavedModel](https://www.tensorflow.org/guide/saved_model) | `saved_model` | `yolov8n-cls_saved_model/` | ✅ | `imgsz`, `keras` |
|
||||
| [TF GraphDef](https://www.tensorflow.org/api_docs/python/tf/Graph) | `pb` | `yolov8n-cls.pb` | ❌ | `imgsz` |
|
||||
| [TF Lite](https://www.tensorflow.org/lite) | `tflite` | `yolov8n-cls.tflite` | ✅ | `imgsz`, `half`, `int8` |
|
||||
|
||||
@ -160,7 +160,7 @@ Available YOLOv8 export formats are in the table below. You can predict or valid
|
||||
| [ONNX](https://onnx.ai/) | `onnx` | `yolov8n.onnx` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `opset` |
|
||||
| [OpenVINO](https://docs.openvino.ai/latest/index.html) | `openvino` | `yolov8n_openvino_model/` | ✅ | `imgsz`, `half` |
|
||||
| [TensorRT](https://developer.nvidia.com/tensorrt) | `engine` | `yolov8n.engine` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `workspace` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlmodel` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlpackage` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [TF SavedModel](https://www.tensorflow.org/guide/saved_model) | `saved_model` | `yolov8n_saved_model/` | ✅ | `imgsz`, `keras` |
|
||||
| [TF GraphDef](https://www.tensorflow.org/api_docs/python/tf/Graph) | `pb` | `yolov8n.pb` | ❌ | `imgsz` |
|
||||
| [TF Lite](https://www.tensorflow.org/lite) | `tflite` | `yolov8n.tflite` | ✅ | `imgsz`, `half`, `int8` |
|
||||
|
||||
@ -174,7 +174,7 @@ i.e. `yolo predict model=yolov8n-pose.onnx`. Usage examples are shown for your m
|
||||
| [ONNX](https://onnx.ai/) | `onnx` | `yolov8n-pose.onnx` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `opset` |
|
||||
| [OpenVINO](https://docs.openvino.ai/latest/index.html) | `openvino` | `yolov8n-pose_openvino_model/` | ✅ | `imgsz`, `half` |
|
||||
| [TensorRT](https://developer.nvidia.com/tensorrt) | `engine` | `yolov8n-pose.engine` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `workspace` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n-pose.mlmodel` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n-pose.mlpackage` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [TF SavedModel](https://www.tensorflow.org/guide/saved_model) | `saved_model` | `yolov8n-pose_saved_model/` | ✅ | `imgsz`, `keras` |
|
||||
| [TF GraphDef](https://www.tensorflow.org/api_docs/python/tf/Graph) | `pb` | `yolov8n-pose.pb` | ❌ | `imgsz` |
|
||||
| [TF Lite](https://www.tensorflow.org/lite) | `tflite` | `yolov8n-pose.tflite` | ✅ | `imgsz`, `half`, `int8` |
|
||||
|
||||
@ -174,7 +174,7 @@ i.e. `yolo predict model=yolov8n-seg.onnx`. Usage examples are shown for your mo
|
||||
| [ONNX](https://onnx.ai/) | `onnx` | `yolov8n-seg.onnx` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `opset` |
|
||||
| [OpenVINO](https://docs.openvino.ai/latest/index.html) | `openvino` | `yolov8n-seg_openvino_model/` | ✅ | `imgsz`, `half` |
|
||||
| [TensorRT](https://developer.nvidia.com/tensorrt) | `engine` | `yolov8n-seg.engine` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `workspace` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n-seg.mlmodel` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n-seg.mlpackage` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [TF SavedModel](https://www.tensorflow.org/guide/saved_model) | `saved_model` | `yolov8n-seg_saved_model/` | ✅ | `imgsz`, `keras` |
|
||||
| [TF GraphDef](https://www.tensorflow.org/api_docs/python/tf/Graph) | `pb` | `yolov8n-seg.pb` | ❌ | `imgsz` |
|
||||
| [TF Lite](https://www.tensorflow.org/lite) | `tflite` | `yolov8n-seg.tflite` | ✅ | `imgsz`, `half`, `int8` |
|
||||
|
||||
@ -172,7 +172,7 @@ i.e. `format='onnx'` or `format='engine'`.
|
||||
| [ONNX](https://onnx.ai/) | `onnx` | `yolov8n.onnx` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `opset` |
|
||||
| [OpenVINO](https://docs.openvino.ai/latest/index.html) | `openvino` | `yolov8n_openvino_model/` | ✅ | `imgsz`, `half` |
|
||||
| [TensorRT](https://developer.nvidia.com/tensorrt) | `engine` | `yolov8n.engine` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `workspace` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlmodel` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlpackage` | ✅ | `imgsz`, `half`, `int8`, `nms` |
|
||||
| [TF SavedModel](https://www.tensorflow.org/guide/saved_model) | `saved_model` | `yolov8n_saved_model/` | ✅ | `imgsz`, `keras` |
|
||||
| [TF GraphDef](https://www.tensorflow.org/api_docs/python/tf/Graph) | `pb` | `yolov8n.pb` | ❌ | `imgsz` |
|
||||
| [TF Lite](https://www.tensorflow.org/lite) | `tflite` | `yolov8n.tflite` | ✅ | `imgsz`, `half`, `int8` |
|
||||
|
||||
@ -334,7 +334,7 @@
|
||||
"| [ONNX](https://onnx.ai/) | `onnx` | `yolov8n.onnx` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `opset` |\n",
|
||||
"| [OpenVINO](https://docs.openvino.ai/latest/index.html) | `openvino` | `yolov8n_openvino_model/` | ✅ | `imgsz`, `half` |\n",
|
||||
"| [TensorRT](https://developer.nvidia.com/tensorrt) | `engine` | `yolov8n.engine` | ✅ | `imgsz`, `half`, `dynamic`, `simplify`, `workspace` |\n",
|
||||
"| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlmodel` | ✅ | `imgsz`, `half`, `int8`, `nms` |\n",
|
||||
"| [CoreML](https://github.com/apple/coremltools) | `coreml` | `yolov8n.mlpackage` | ✅ | `imgsz`, `half`, `int8`, `nms` |\n",
|
||||
"| [TF SavedModel](https://www.tensorflow.org/guide/saved_model) | `saved_model` | `yolov8n_saved_model/` | ✅ | `imgsz`, `keras` |\n",
|
||||
"| [TF GraphDef](https://www.tensorflow.org/api_docs/python/tf/Graph) | `pb` | `yolov8n.pb` | ❌ | `imgsz` |\n",
|
||||
"| [TF Lite](https://www.tensorflow.org/lite) | `tflite` | `yolov8n.tflite` | ✅ | `imgsz`, `half`, `int8` |\n",
|
||||
|
||||
@ -24,7 +24,7 @@ pandas>=1.1.4
|
||||
seaborn>=0.11.0
|
||||
|
||||
# Export --------------------------------------
|
||||
# coremltools>=6.0,<=6.2 # CoreML export
|
||||
# coremltools>=7.0.b1 # CoreML export
|
||||
# onnx>=1.12.0 # ONNX export
|
||||
# onnxsim>=0.4.1 # ONNX simplifier
|
||||
# nvidia-pyindex # TensorRT export
|
||||
|
||||
2
setup.py
2
setup.py
@ -50,7 +50,7 @@ setup(
|
||||
'mkdocs-ultralytics-plugin>=0.0.25', # for meta descriptions and images, dates and authors
|
||||
],
|
||||
'export': [
|
||||
'coremltools>=6.0,<=6.2',
|
||||
'coremltools>=7.0.b1',
|
||||
'openvino-dev>=2023.0',
|
||||
'tensorflowjs', # automatically installs tensorflow
|
||||
], },
|
||||
|
||||
@ -1,6 +1,6 @@
|
||||
# Ultralytics YOLO 🚀, AGPL-3.0 license
|
||||
|
||||
__version__ = '8.0.149'
|
||||
__version__ = '8.0.150'
|
||||
|
||||
from ultralytics.hub import start
|
||||
from ultralytics.models import RTDETR, SAM, YOLO
|
||||
|
||||
@ -72,7 +72,7 @@ download: |
|
||||
xmlbox = obj.find('bndbox')
|
||||
bb = convert_box((w, h), [float(xmlbox.find(x).text) for x in ('xmin', 'xmax', 'ymin', 'ymax')])
|
||||
cls_id = names.index(cls) # class id
|
||||
out_file.write(" ".join([str(a) for a in (cls_id, *bb)]) + '\n')
|
||||
out_file.write(" ".join(str(a) for a in (cls_id, *bb)) + '\n')
|
||||
|
||||
|
||||
# Download
|
||||
|
||||
@ -9,7 +9,7 @@ TorchScript | `torchscript` | yolov8n.torchscript
|
||||
ONNX | `onnx` | yolov8n.onnx
|
||||
OpenVINO | `openvino` | yolov8n_openvino_model/
|
||||
TensorRT | `engine` | yolov8n.engine
|
||||
CoreML | `coreml` | yolov8n.mlmodel
|
||||
CoreML | `coreml` | yolov8n.mlpackage
|
||||
TensorFlow SavedModel | `saved_model` | yolov8n_saved_model/
|
||||
TensorFlow GraphDef | `pb` | yolov8n.pb
|
||||
TensorFlow Lite | `tflite` | yolov8n.tflite
|
||||
@ -35,7 +35,7 @@ Inference:
|
||||
yolov8n.onnx # ONNX Runtime or OpenCV DNN with dnn=True
|
||||
yolov8n_openvino_model # OpenVINO
|
||||
yolov8n.engine # TensorRT
|
||||
yolov8n.mlmodel # CoreML (macOS-only)
|
||||
yolov8n.mlpackage # CoreML (macOS-only)
|
||||
yolov8n_saved_model # TensorFlow SavedModel
|
||||
yolov8n.pb # TensorFlow GraphDef
|
||||
yolov8n.tflite # TensorFlow Lite
|
||||
@ -82,7 +82,7 @@ def export_formats():
|
||||
['ONNX', 'onnx', '.onnx', True, True],
|
||||
['OpenVINO', 'openvino', '_openvino_model', True, False],
|
||||
['TensorRT', 'engine', '.engine', False, True],
|
||||
['CoreML', 'coreml', '.mlmodel', True, False],
|
||||
['CoreML', 'coreml', '.mlpackage', True, False],
|
||||
['TensorFlow SavedModel', 'saved_model', '_saved_model', True, True],
|
||||
['TensorFlow GraphDef', 'pb', '.pb', True, True],
|
||||
['TensorFlow Lite', 'tflite', '.tflite', True, False],
|
||||
@ -149,8 +149,10 @@ class Exporter:
|
||||
self.run_callbacks('on_export_start')
|
||||
t = time.time()
|
||||
format = self.args.format.lower() # to lowercase
|
||||
if format in ('tensorrt', 'trt'): # engine aliases
|
||||
if format in ('tensorrt', 'trt'): # 'engine' aliases
|
||||
format = 'engine'
|
||||
if format in ('mlmodel', 'mlpackage', 'mlprogram', 'apple', 'ios'): # 'coreml' aliases
|
||||
format = 'coreml'
|
||||
fmts = tuple(export_formats()['Argument'][1:]) # available export formats
|
||||
flags = [x == format for x in fmts]
|
||||
if sum(flags) != 1:
|
||||
@ -319,7 +321,7 @@ class Exporter:
|
||||
dynamic['output0'] = {0: 'batch', 2: 'anchors'} # shape(1, 84, 8400)
|
||||
|
||||
torch.onnx.export(
|
||||
self.model.cpu() if dynamic else self.model, # --dynamic only compatible with cpu
|
||||
self.model.cpu() if dynamic else self.model, # dynamic=True only compatible with cpu
|
||||
self.im.cpu() if dynamic else self.im,
|
||||
f,
|
||||
verbose=False,
|
||||
@ -461,14 +463,16 @@ class Exporter:
|
||||
yaml_save(f / 'metadata.yaml', self.metadata) # add metadata.yaml
|
||||
return str(f), None
|
||||
|
||||
@try_export
|
||||
def export_coreml(self, prefix=colorstr('CoreML:')):
|
||||
"""YOLOv8 CoreML export."""
|
||||
check_requirements('coremltools>=6.0,<=6.2')
|
||||
mlmodel = self.args.format.lower() == 'mlmodel' # legacy *.mlmodel export format requested
|
||||
check_requirements('coremltools>=6.0,<=6.2' if mlmodel else 'coremltools>=7.0.b1')
|
||||
import coremltools as ct # noqa
|
||||
|
||||
LOGGER.info(f'\n{prefix} starting export with coremltools {ct.__version__}...')
|
||||
f = self.file.with_suffix('.mlmodel')
|
||||
f = self.file.with_suffix('.mlmodel' if mlmodel else '.mlpackage')
|
||||
if f.is_dir():
|
||||
shutil.rmtree(f)
|
||||
|
||||
bias = [0.0, 0.0, 0.0]
|
||||
scale = 1 / 255
|
||||
@ -479,20 +483,38 @@ class Exporter:
|
||||
elif self.model.task == 'detect':
|
||||
model = iOSDetectModel(self.model, self.im) if self.args.nms else self.model
|
||||
else:
|
||||
# TODO CoreML Segment and Pose model pipelining
|
||||
if self.args.nms:
|
||||
LOGGER.warning(f"{prefix} WARNING ⚠️ 'nms=True' is only available for Detect models like 'yolov8n.pt'.")
|
||||
# TODO CoreML Segment and Pose model pipelining
|
||||
model = self.model
|
||||
|
||||
ts = torch.jit.trace(model.eval(), self.im, strict=False) # TorchScript model
|
||||
ct_model = ct.convert(ts,
|
||||
inputs=[ct.ImageType('image', shape=self.im.shape, scale=scale, bias=bias)],
|
||||
classifier_config=classifier_config)
|
||||
bits, mode = (8, 'kmeans_lut') if self.args.int8 else (16, 'linear') if self.args.half else (32, None)
|
||||
classifier_config=classifier_config,
|
||||
convert_to='neuralnetwork' if mlmodel else 'mlprogram')
|
||||
bits, mode = (8, 'kmeans') if self.args.int8 else (16, 'linear') if self.args.half else (32, None)
|
||||
if bits < 32:
|
||||
if 'kmeans' in mode:
|
||||
check_requirements('scikit-learn') # scikit-learn package required for k-means quantization
|
||||
ct_model = ct.models.neural_network.quantization_utils.quantize_weights(ct_model, bits, mode)
|
||||
if mlmodel:
|
||||
ct_model = ct.models.neural_network.quantization_utils.quantize_weights(ct_model, bits, mode)
|
||||
else:
|
||||
import coremltools.optimize.coreml as cto
|
||||
op_config = cto.OpPalettizerConfig(mode=mode, nbits=bits, weight_threshold=512)
|
||||
config = cto.OptimizationConfig(global_config=op_config)
|
||||
ct_model = cto.palettize_weights(ct_model, config=config)
|
||||
if self.args.nms and self.model.task == 'detect':
|
||||
ct_model = self._pipeline_coreml(ct_model)
|
||||
if mlmodel:
|
||||
import platform
|
||||
|
||||
# coremltools<=6.2 NMS export requires Python<3.11
|
||||
check_version(platform.python_version(), '<3.11', name='Python ', hard=True)
|
||||
weights_dir = None
|
||||
else:
|
||||
ct_model.save(str(f)) # save otherwise weights_dir does not exist
|
||||
weights_dir = str(f / 'Data/com.apple.CoreML/weights')
|
||||
ct_model = self._pipeline_coreml(ct_model, weights_dir=weights_dir)
|
||||
|
||||
m = self.metadata # metadata dict
|
||||
ct_model.short_description = m.pop('description')
|
||||
@ -500,7 +522,14 @@ class Exporter:
|
||||
ct_model.license = m.pop('license')
|
||||
ct_model.version = m.pop('version')
|
||||
ct_model.user_defined_metadata.update({k: str(v) for k, v in m.items()})
|
||||
ct_model.save(str(f))
|
||||
try:
|
||||
ct_model.save(str(f)) # save *.mlpackage
|
||||
except Exception as e:
|
||||
LOGGER.warning(
|
||||
f'{prefix} WARNING ⚠️ CoreML export to *.mlpackage failed ({e}), reverting to *.mlmodel export. '
|
||||
f'Known coremltools Python 3.11 and Windows bugs https://github.com/apple/coremltools/issues/1928.')
|
||||
f = f.with_suffix('.mlmodel')
|
||||
ct_model.save(str(f))
|
||||
return f, ct_model
|
||||
|
||||
@try_export
|
||||
@ -546,7 +575,7 @@ class Exporter:
|
||||
if self.args.dynamic:
|
||||
shape = self.im.shape
|
||||
if shape[0] <= 1:
|
||||
LOGGER.warning(f'{prefix} WARNING ⚠️ --dynamic model requires maximum --batch-size argument')
|
||||
LOGGER.warning(f"{prefix} WARNING ⚠️ 'dynamic=True' model requires max batch size, i.e. 'batch=16'")
|
||||
profile = builder.create_optimization_profile()
|
||||
for inp in inputs:
|
||||
profile.set_shape(inp.name, (1, *shape[1:]), (max(1, shape[0] // 2), *shape[1:]), shape)
|
||||
@ -805,7 +834,7 @@ class Exporter:
|
||||
populator.populate()
|
||||
tmp_file.unlink()
|
||||
|
||||
def _pipeline_coreml(self, model, prefix=colorstr('CoreML Pipeline:')):
|
||||
def _pipeline_coreml(self, model, weights_dir=None, prefix=colorstr('CoreML Pipeline:')):
|
||||
"""YOLOv8 CoreML pipeline."""
|
||||
import coremltools as ct # noqa
|
||||
|
||||
@ -853,7 +882,7 @@ class Exporter:
|
||||
# print(spec.description)
|
||||
|
||||
# Model from spec
|
||||
model = ct.models.MLModel(spec)
|
||||
model = ct.models.MLModel(spec, weights_dir=weights_dir)
|
||||
|
||||
# 3. Create NMS protobuf
|
||||
nms_spec = ct.proto.Model_pb2.Model()
|
||||
@ -912,7 +941,7 @@ class Exporter:
|
||||
'Confidence threshold': str(nms.confidenceThreshold)})
|
||||
|
||||
# Save the model
|
||||
model = ct.models.MLModel(pipeline.spec)
|
||||
model = ct.models.MLModel(pipeline.spec, weights_dir=weights_dir)
|
||||
model.input_description['image'] = 'Input image'
|
||||
model.input_description['iouThreshold'] = f'(optional) IOU threshold override (default: {nms.iouThreshold})'
|
||||
model.input_description['confidenceThreshold'] = \
|
||||
|
||||
@ -320,7 +320,7 @@ class Model:
|
||||
half=overrides['half'],
|
||||
int8=overrides['int8'],
|
||||
device=overrides['device'],
|
||||
verbose=overrides['verbose'])
|
||||
verbose=kwargs.get('verbose'))
|
||||
|
||||
def export(self, **kwargs):
|
||||
"""
|
||||
|
||||
@ -20,7 +20,7 @@ Usage - formats:
|
||||
yolov8n.onnx # ONNX Runtime or OpenCV DNN with dnn=True
|
||||
yolov8n_openvino_model # OpenVINO
|
||||
yolov8n.engine # TensorRT
|
||||
yolov8n.mlmodel # CoreML (macOS-only)
|
||||
yolov8n.mlpackage # CoreML (macOS-only)
|
||||
yolov8n_saved_model # TensorFlow SavedModel
|
||||
yolov8n.pb # TensorFlow GraphDef
|
||||
yolov8n.tflite # TensorFlow Lite
|
||||
|
||||
@ -11,7 +11,7 @@ Usage - formats:
|
||||
yolov8n.onnx # ONNX Runtime or OpenCV DNN with dnn=True
|
||||
yolov8n_openvino_model # OpenVINO
|
||||
yolov8n.engine # TensorRT
|
||||
yolov8n.mlmodel # CoreML (macOS-only)
|
||||
yolov8n.mlpackage # CoreML (macOS-only)
|
||||
yolov8n_saved_model # TensorFlow SavedModel
|
||||
yolov8n.pb # TensorFlow GraphDef
|
||||
yolov8n.tflite # TensorFlow Lite
|
||||
|
||||
@ -68,7 +68,7 @@ class AutoBackend(nn.Module):
|
||||
| ONNX Runtime | *.onnx |
|
||||
| ONNX OpenCV DNN | *.onnx dnn=True |
|
||||
| OpenVINO | *.xml |
|
||||
| CoreML | *.mlmodel |
|
||||
| CoreML | *.mlpackage |
|
||||
| TensorRT | *.engine |
|
||||
| TensorFlow SavedModel | *_saved_model |
|
||||
| TensorFlow GraphDef | *.pb |
|
||||
@ -485,8 +485,13 @@ class AutoBackend(nn.Module):
|
||||
sf = list(export_formats().Suffix) # export suffixes
|
||||
if not is_url(p, check=False) and not isinstance(p, str):
|
||||
check_suffix(p, sf) # checks
|
||||
url = urlparse(p) # if url may be Triton inference server
|
||||
types = [s in Path(p).name for s in sf]
|
||||
name = Path(p).name
|
||||
types = [s in name for s in sf]
|
||||
types[5] |= name.endswith('.mlmodel') # retain support for older Apple CoreML *.mlmodel formats
|
||||
types[8] &= not types[9] # tflite &= not edgetpu
|
||||
triton = not any(types) and all([any(s in url.scheme for s in ['http', 'grpc']), url.netloc])
|
||||
if any(types):
|
||||
triton = False
|
||||
else:
|
||||
url = urlparse(p) # if url may be Triton inference server
|
||||
triton = all([any(s in url.scheme for s in ['http', 'grpc']), url.netloc])
|
||||
return types + [triton]
|
||||
|
||||
@ -14,7 +14,7 @@ TorchScript | `torchscript` | yolov8n.torchscript
|
||||
ONNX | `onnx` | yolov8n.onnx
|
||||
OpenVINO | `openvino` | yolov8n_openvino_model/
|
||||
TensorRT | `engine` | yolov8n.engine
|
||||
CoreML | `coreml` | yolov8n.mlmodel
|
||||
CoreML | `coreml` | yolov8n.mlpackage
|
||||
TensorFlow SavedModel | `saved_model` | yolov8n_saved_model/
|
||||
TensorFlow GraphDef | `pb` | yolov8n.pb
|
||||
TensorFlow Lite | `tflite` | yolov8n.tflite
|
||||
|
||||
@ -265,20 +265,21 @@ def check_requirements(requirements=ROOT.parent / 'requirements.txt', exclude=()
|
||||
elif isinstance(requirements, str):
|
||||
requirements = [requirements]
|
||||
|
||||
s = '' # console string
|
||||
pkgs = []
|
||||
for r in requirements:
|
||||
r_stripped = r.split('/')[-1].replace('.git', '') # replace git+https://org/repo.git -> 'repo'
|
||||
try:
|
||||
pkg.require(r_stripped)
|
||||
except (pkg.VersionConflict, pkg.DistributionNotFound): # exception if requirements not met
|
||||
pkg.require(r_stripped) # exception if requirements not met
|
||||
except pkg.DistributionNotFound:
|
||||
try: # attempt to import (slower but more accurate)
|
||||
import importlib
|
||||
importlib.import_module(next(pkg.parse_requirements(r_stripped)).name)
|
||||
except ImportError:
|
||||
s += f'"{r}" '
|
||||
pkgs.append(r)
|
||||
except pkg.VersionConflict:
|
||||
pkgs.append(r)
|
||||
|
||||
s = ' '.join(f'"{x}"' for x in pkgs) # console string
|
||||
if s:
|
||||
if install and AUTOINSTALL: # check environment variable
|
||||
n = len(pkgs) # number of packages updates
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user