export_qat.py

import argparse
import contextlib
import json
import os
import platform
import re
import subprocess
import sys
import time
import warnings
from pathlib import Path

import pandas as pd
import torch
from torch.utils.mobile_optimizer import optimize_for_mobile
import onnx_graphsurgeon as gs

import models.quantize as quantize

from pytorch_quantization import nn as quant_nn

FILE = Path(__file__).resolve()
ROOT = FILE.parents[0]  # YOLO root directory
if str(ROOT) not in sys.path:
    sys.path.append(str(ROOT))  # add ROOT to PATH
if platform.system() != 'Windows':
    ROOT = Path(os.path.relpath(ROOT, Path.cwd()))  # relative

from models.experimental import attempt_load, End2End
from models.yolo import ClassificationModel, Detect, DDetect, DualDetect, DualDDetect, DetectionModel, SegmentationModel
from utils.dataloaders import LoadImages
from utils.general import (LOGGER, Profile, check_dataset, check_img_size, check_requirements, check_version,
                           check_yaml, colorstr, file_size, get_default_args, print_args, url2file, yaml_save)
from utils.torch_utils import select_device, smart_inference_mode
from models.quantize import remove_redundant_qdq_model

MACOS = platform.system() == 'Darwin'  # macOS environment



def export_formats():
    # YOLO export formats
    x = [
        ['PyTorch', '-', '.pt', True, True],
        ['TorchScript', 'torchscript', '.torchscript', True, True],
        ['ONNX', 'onnx', '.onnx', True, True],
        ['ONNX END2END', 'onnx_end2end', '_end2end.onnx', True, True],
        ['OpenVINO', 'openvino', '_openvino_model', True, False],
        ['TensorRT', 'engine', '.engine', False, True],
        ['CoreML', 'coreml', '.mlmodel', True, False],
        ['TensorFlow SavedModel', 'saved_model', '_saved_model', True, True],
        ['TensorFlow GraphDef', 'pb', '.pb', True, True],
        ['TensorFlow Lite', 'tflite', '.tflite', True, False],
        ['TensorFlow Edge TPU', 'edgetpu', '_edgetpu.tflite', False, False],
        ['TensorFlow.js', 'tfjs', '_web_model', False, False],
        ['PaddlePaddle', 'paddle', '_paddle_model', True, True],]
    return pd.DataFrame(x, columns=['Format', 'Argument', 'Suffix', 'CPU', 'GPU'])


def try_export(inner_func):
    # YOLO export decorator, i..e @try_export
    inner_args = get_default_args(inner_func)

    def outer_func(*args, **kwargs):
        prefix = inner_args['prefix']
        try:
            with Profile() as dt:
                f, model = inner_func(*args, **kwargs)
            LOGGER.info(f'{prefix} export success ✅ {dt.t:.1f}s, saved as {f} ({file_size(f):.1f} MB)')
            return f, model
        except Exception as e:
            LOGGER.info(f'{prefix} export failure ❌ {dt.t:.1f}s: {e}')
            return None, None

    return outer_func


@try_export
def export_torchscript(model, im, file, optimize, prefix=colorstr('TorchScript:')):
    # YOLO TorchScript model export
    LOGGER.info(f'\n{prefix} starting export with torch {torch.__version__}...')
    f = file.with_suffix('.torchscript')

    ts = torch.jit.trace(model, im, strict=False)
    d = {"shape": im.shape, "stride": int(max(model.stride)), "names": model.names}
    extra_files = {'config.txt': json.dumps(d)}  # torch._C.ExtraFilesMap()
    if optimize:  # https://pytorch.org/tutorials/recipes/mobile_interpreter.html
        optimize_for_mobile(ts)._save_for_lite_interpreter(str(f), _extra_files=extra_files)
    else:
        ts.save(str(f), _extra_files=extra_files)
    return f, None


@try_export
def export_onnx(model, im, file, opset, dynamic, simplify, prefix=colorstr('ONNX:')):
    # YOLO ONNX export
    check_requirements('onnx')
    import onnx

    is_model_qat=False
    for i in range(0, len(model.model)):
        layer = model.model[i]
        if quantize.have_quantizer(layer):
            is_model_qat=True
            break

    LOGGER.info(f'\n{prefix} starting export with onnx {onnx.__version__}...')
    f = file.with_suffix('.onnx')

    output_names = ['output0', 'output1'] if isinstance(model, SegmentationModel) else ['output0']
    if dynamic:
        dynamic = {'images': {0: 'batch', 2: 'height', 3: 'width'}}  # shape(1,3,640,640)
        if isinstance(model, SegmentationModel):
            dynamic['output0'] = {0: 'batch', 1: 'anchors'}  # shape(1,25200,85)
            dynamic['output1'] = {0: 'batch', 2: 'mask_height', 3: 'mask_width'}  # shape(1,32,160,160)
        elif isinstance(model, DetectionModel):
            dynamic['output0'] = {0: 'batch', 1: 'anchors'}  # shape(1,25200,85)
    
    if is_model_qat:
        warnings.filterwarnings("ignore")
        LOGGER.info(f'{prefix} Model QAT Detected ...')
        quant_nn.TensorQuantizer.use_fb_fake_quant = True
        model.eval()
        quantize.initialize()
        quantize.replace_custom_module_forward(model)
        with torch.no_grad():
            torch.onnx.export(
                model, 
                im, 
                f,
                verbose=False,
                opset_version=13,
                do_constant_folding=True,
                input_names=['images'],
                output_names=output_names,
                dynamic_axes=dynamic)

    else: 
        torch.onnx.export(
            model.cpu() if dynamic else model,  # --dynamic only compatible with cpu
            im.cpu() if dynamic else im,
            f,
            verbose=False,
            opset_version=opset,
            do_constant_folding=True,
            input_names=['images'],
            output_names=output_names,
            dynamic_axes=dynamic or None)

    # Checks
    model_onnx = onnx.load(f)  # load onnx model
    onnx.checker.check_model(model_onnx)  # check onnx model

    # Metadata
    d = {'stride': int(max(model.stride)), 'names': model.names}
    for k, v in d.items():
        meta = model_onnx.metadata_props.add()
        meta.key, meta.value = k, str(v)

    # Simplify
    if simplify:
        try:
            cuda = torch.cuda.is_available()
            check_requirements(('onnxruntime-gpu' if cuda else 'onnxruntime', 'onnx-simplifier>=0.4.1'))
            import onnxsim

            LOGGER.info(f'{prefix} simplifying with onnx-simplifier {onnxsim.__version__}...')
            model_onnx, check = onnxsim.simplify(model_onnx)
            assert check, 'assert check failed'
            onnx.save(model_onnx, f)
        except Exception as e:
            LOGGER.info(f'{prefix} simplifier failure: {e}')

    LOGGER.info(f'{prefix} Removing redundant Q/DQ layer with onnx_graphsurgeon {gs.__version__}...')
    remove_redundant_qdq_model(model_onnx, f) 
    model_onnx = onnx.load(f)
    return f, model_onnx
    

@try_export
def export_onnx_end2end(model, im, file, simplify, topk_all, iou_thres, conf_thres, device, labels, prefix=colorstr('ONNX END2END:')):
    if not isinstance(model, DetectionModel) or isinstance(model, SegmentationModel):
        raise RuntimeError("Model not supported. Only Detection Models can be exported with End2End functionality.")
    # YOLO ONNX export
    check_requirements('onnx')
    import onnx
    
    is_model_qat=False
    for i in range(0, len(model.model)):
        layer = model.model[i]
        if quantize.have_quantizer(layer):
            is_model_qat=True
            break
    
    LOGGER.info(f'\n{prefix} starting export with onnx {onnx.__version__}...')
    f = os.path.splitext(file)[0] + "-end2end.onnx"
    batch_size = 'batch'

    dynamic_axes = {'images': {0 : 'batch', 2: 'height', 3:'width'}, } # variable length axes

    output_axes = {
                    'num_dets': {0: 'batch'},
                    'det_boxes': {0: 'batch'},
                    'det_scores': {0: 'batch'},
                    'det_classes': {0: 'batch'},
                }
    dynamic_axes.update(output_axes)
    model = End2End(model, topk_all, iou_thres, conf_thres, None ,device, labels)

    output_names = ['num_dets', 'det_boxes', 'det_scores', 'det_classes']
    shapes = [ batch_size, 1,  batch_size,  topk_all, 4,
               batch_size,  topk_all,  batch_size,  topk_all]
    

    
    if is_model_qat:
        warnings.filterwarnings("ignore")
        LOGGER.info(f'{prefix} Model QAT Detected ...')
        quant_nn.TensorQuantizer.use_fb_fake_quant = True
        model.eval()
        quantize.initialize()
        quantize.replace_custom_module_forward(model)
        
        with torch.no_grad():
            torch.onnx.export(model, 
                            im, 
                            f, 
                            verbose=False, 
                            export_params=True,       # store the trained parameter weights inside the model file
                            opset_version=13, 
                            do_constant_folding=True, # whether to execute constant folding for optimization
                            input_names=['images'],
                            output_names=output_names,
                            dynamic_axes=dynamic_axes)
        quant_nn.TensorQuantizer.use_fb_fake_quant = False
    else:
        torch.onnx.export(model, 
                    im, 
                    f, 
                    verbose=False, 
                    export_params=True,       # store the trained parameter weights inside the model file
                    opset_version=12, 
                    do_constant_folding=True, # whether to execute constant folding for optimization
                    input_names=['images'],
                    output_names=output_names,
                    dynamic_axes=dynamic_axes)

    # Checks
    model_onnx = onnx.load(f)  # load onnx model
    onnx.checker.check_model(model_onnx)  # check onnx model
    for i in model_onnx.graph.output:
        for j in i.type.tensor_type.shape.dim:
            j.dim_param = str(shapes.pop(0))

    if simplify:
        try:
            import onnxsim

            print('\nStarting to simplify ONNX...')
            model_onnx, check = onnxsim.simplify(model_onnx)
            assert check, 'assert check failed'
        except Exception as e:
            print(f'Simplifier failure: {e}')

        # print(onnx.helper.printable_graph(onnx_model.graph))  # print a human readable model
        onnx.save(model_onnx,f)
        print('ONNX export success, saved as %s' % f)

    LOGGER.info(f'{prefix} Removing redundant Q/DQ layer with onnx_graphsurgeon {gs.__version__}...')
    remove_redundant_qdq_model(model_onnx, f) 
    model_onnx = onnx.load(f)

    return f, model_onnx


@try_export
def export_openvino(file, metadata, half, prefix=colorstr('OpenVINO:')):
    # YOLO OpenVINO export
    check_requirements('openvino-dev')  # requires openvino-dev: https://pypi.org/project/openvino-dev/
    import openvino.inference_engine as ie

    LOGGER.info(f'\n{prefix} starting export with openvino {ie.__version__}...')
    f = str(file).replace('.pt', f'_openvino_model{os.sep}')

    #cmd = f"mo --input_model {file.with_suffix('.onnx')} --output_dir {f} --data_type {'FP16' if half else 'FP32'}"
    #cmd = f"mo --input_model {file.with_suffix('.onnx')} --output_dir {f} {"--compress_to_fp16" if half else ""}"
    half_arg = "--compress_to_fp16" if half else ""
    cmd = f"mo --input_model {file.with_suffix('.onnx')} --output_dir {f} {half_arg}"
    subprocess.run(cmd.split(), check=True, env=os.environ)  # export
    yaml_save(Path(f) / file.with_suffix('.yaml').name, metadata)  # add metadata.yaml
    return f, None


@try_export
def export_paddle(model, im, file, metadata, prefix=colorstr('PaddlePaddle:')):
    # YOLO Paddle export
    check_requirements(('paddlepaddle', 'x2paddle'))
    import x2paddle
    from x2paddle.convert import pytorch2paddle

    LOGGER.info(f'\n{prefix} starting export with X2Paddle {x2paddle.__version__}...')
    f = str(file).replace('.pt', f'_paddle_model{os.sep}')

    pytorch2paddle(module=model, save_dir=f, jit_type='trace', input_examples=[im])  # export
    yaml_save(Path(f) / file.with_suffix('.yaml').name, metadata)  # add metadata.yaml
    return f, None


@try_export
def export_coreml(model, im, file, int8, half, prefix=colorstr('CoreML:')):
    # YOLO CoreML export
    check_requirements('coremltools')
    import coremltools as ct

    LOGGER.info(f'\n{prefix} starting export with coremltools {ct.__version__}...')
    f = file.with_suffix('.mlmodel')

    ts = torch.jit.trace(model, im, strict=False)  # TorchScript model
    ct_model = ct.convert(ts, inputs=[ct.ImageType('image', shape=im.shape, scale=1 / 255, bias=[0, 0, 0])])
    bits, mode = (8, 'kmeans_lut') if int8 else (16, 'linear') if half else (32, None)
    if bits < 32:
        if MACOS:  # quantization only supported on macOS
            with warnings.catch_warnings():
                warnings.filterwarnings("ignore", category=DeprecationWarning)  # suppress numpy==1.20 float warning
                ct_model = ct.models.neural_network.quantization_utils.quantize_weights(ct_model, bits, mode)
        else:
            print(f'{prefix} quantization only supported on macOS, skipping...')
    ct_model.save(f)
    return f, ct_model


@try_export
def export_engine(model, im, file, half, dynamic, simplify, workspace=4, verbose=False, prefix=colorstr('TensorRT:')):
    # YOLO TensorRT export https://developer.nvidia.com/tensorrt
    assert im.device.type != 'cpu', 'export running on CPU but must be on GPU, i.e. `python export.py --device 0`'
    try:
        import tensorrt as trt
    except Exception:
        if platform.system() == 'Linux':
            check_requirements('nvidia-tensorrt', cmds='-U --index-url https://pypi.ngc.nvidia.com')
        import tensorrt as trt

    if trt.__version__[0] == '7':  # TensorRT 7 handling https://github.com/ultralytics/yolov5/issues/6012
        grid = model.model[-1].anchor_grid
        model.model[-1].anchor_grid = [a[..., :1, :1, :] for a in grid]
        export_onnx(model, im, file, 12, dynamic, simplify)  # opset 12
        model.model[-1].anchor_grid = grid
    else:  # TensorRT >= 8
        check_version(trt.__version__, '8.0.0', hard=True)  # require tensorrt>=8.0.0
        export_onnx(model, im, file, 12, dynamic, simplify)  # opset 12
    onnx = file.with_suffix('.onnx')

    LOGGER.info(f'\n{prefix} starting export with TensorRT {trt.__version__}...')
    assert onnx.exists(), f'failed to export ONNX file: {onnx}'
    f = file.with_suffix('.engine')  # TensorRT engine file
    logger = trt.Logger(trt.Logger.INFO)
    if verbose:
        logger.min_severity = trt.Logger.Severity.VERBOSE

    builder = trt.Builder(logger)
    config = builder.create_builder_config()
    config.max_workspace_size = workspace * 1 << 30
    # config.set_memory_pool_limit(trt.MemoryPoolType.WORKSPACE, workspace << 30)  # fix TRT 8.4 deprecation notice

    flag = (1 << int(trt.NetworkDefinitionCreationFlag.EXPLICIT_BATCH))
    network = builder.create_network(flag)
    parser = trt.OnnxParser(network, logger)
    if not parser.parse_from_file(str(onnx)):
        raise RuntimeError(f'failed to load ONNX file: {onnx}')

    inputs = [network.get_input(i) for i in range(network.num_inputs)]
    outputs = [network.get_output(i) for i in range(network.num_outputs)]
    for inp in inputs:
        LOGGER.info(f'{prefix} input "{inp.name}" with shape{inp.shape} {inp.dtype}')
    for out in outputs:
        LOGGER.info(f'{prefix} output "{out.name}" with shape{out.shape} {out.dtype}')

    if dynamic:
        if im.shape[0] <= 1:
            LOGGER.warning(f"{prefix} WARNING ⚠️ --dynamic model requires maximum --batch-size argument")
        profile = builder.create_optimization_profile()
        for inp in inputs:
            profile.set_shape(inp.name, (1, *im.shape[1:]), (max(1, im.shape[0] // 2), *im.shape[1:]), im.shape)
        config.add_optimization_profile(profile)

    LOGGER.info(f'{prefix} building FP{16 if builder.platform_has_fast_fp16 and half else 32} engine as {f}')
    if builder.platform_has_fast_fp16 and half:
        config.set_flag(trt.BuilderFlag.FP16)
    with builder.build_engine(network, config) as engine, open(f, 'wb') as t:
        t.write(engine.serialize())
    return f, None


@try_export
def export_saved_model(model,
                       im,
                       file,
                       dynamic,
                       tf_nms=False,
                       agnostic_nms=False,
                       topk_per_class=100,
                       topk_all=100,
                       iou_thres=0.45,
                       conf_thres=0.25,
                       keras=False,
                       prefix=colorstr('TensorFlow SavedModel:')):
    # YOLO TensorFlow SavedModel export
    try:
        import tensorflow as tf
    except Exception:
        check_requirements(f"tensorflow{'' if torch.cuda.is_available() else '-macos' if MACOS else '-cpu'}")
        import tensorflow as tf
    from tensorflow.python.framework.convert_to_constants import convert_variables_to_constants_v2

    from models.tf import TFModel

    LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...')
    f = str(file).replace('.pt', '_saved_model')
    batch_size, ch, *imgsz = list(im.shape)  # BCHW

    tf_model = TFModel(cfg=model.yaml, model=model, nc=model.nc, imgsz=imgsz)
    im = tf.zeros((batch_size, *imgsz, ch))  # BHWC order for TensorFlow
    _ = tf_model.predict(im, tf_nms, agnostic_nms, topk_per_class, topk_all, iou_thres, conf_thres)
    inputs = tf.keras.Input(shape=(*imgsz, ch), batch_size=None if dynamic else batch_size)
    outputs = tf_model.predict(inputs, tf_nms, agnostic_nms, topk_per_class, topk_all, iou_thres, conf_thres)
    keras_model = tf.keras.Model(inputs=inputs, outputs=outputs)
    keras_model.trainable = False
    keras_model.summary()
    if keras:
        keras_model.save(f, save_format='tf')
    else:
        spec = tf.TensorSpec(keras_model.inputs[0].shape, keras_model.inputs[0].dtype)
        m = tf.function(lambda x: keras_model(x))  # full model
        m = m.get_concrete_function(spec)
        frozen_func = convert_variables_to_constants_v2(m)
        tfm = tf.Module()
        tfm.__call__ = tf.function(lambda x: frozen_func(x)[:4] if tf_nms else frozen_func(x), [spec])
        tfm.__call__(im)
        tf.saved_model.save(tfm,
                            f,
                            options=tf.saved_model.SaveOptions(experimental_custom_gradients=False) if check_version(
                                tf.__version__, '2.6') else tf.saved_model.SaveOptions())
    return f, keras_model


@try_export
def export_pb(keras_model, file, prefix=colorstr('TensorFlow GraphDef:')):
    # YOLO TensorFlow GraphDef *.pb export https://github.com/leimao/Frozen_Graph_TensorFlow
    import tensorflow as tf
    from tensorflow.python.framework.convert_to_constants import convert_variables_to_constants_v2

    LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...')
    f = file.with_suffix('.pb')

    m = tf.function(lambda x: keras_model(x))  # full model
    m = m.get_concrete_function(tf.TensorSpec(keras_model.inputs[0].shape, keras_model.inputs[0].dtype))
    frozen_func = convert_variables_to_constants_v2(m)
    frozen_func.graph.as_graph_def()
    tf.io.write_graph(graph_or_graph_def=frozen_func.graph, logdir=str(f.parent), name=f.name, as_text=False)
    return f, None


@try_export
def export_tflite(keras_model, im, file, int8, data, nms, agnostic_nms, prefix=colorstr('TensorFlow Lite:')):
    # YOLOv5 TensorFlow Lite export
    import tensorflow as tf

    LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...')
    batch_size, ch, *imgsz = list(im.shape)  # BCHW
    f = str(file).replace('.pt', '-fp16.tflite')

    converter = tf.lite.TFLiteConverter.from_keras_model(keras_model)
    converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS]
    converter.target_spec.supported_types = [tf.float16]
    converter.optimizations = [tf.lite.Optimize.DEFAULT]
    if int8:
        from models.tf import representative_dataset_gen
        dataset = LoadImages(check_dataset(check_yaml(data))['train'], img_size=imgsz, auto=False)
        converter.representative_dataset = lambda: representative_dataset_gen(dataset, ncalib=100)
        converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8]
        converter.target_spec.supported_types = []
        converter.inference_input_type = tf.uint8  # or tf.int8
        converter.inference_output_type = tf.uint8  # or tf.int8
        converter.experimental_new_quantizer = True
        f = str(file).replace('.pt', '-int8.tflite')
    if nms or agnostic_nms:
        converter.target_spec.supported_ops.append(tf.lite.OpsSet.SELECT_TF_OPS)

    tflite_model = converter.convert()
    open(f, "wb").write(tflite_model)
    return f, None


@try_export
def export_edgetpu(file, prefix=colorstr('Edge TPU:')):
    # YOLO Edge TPU export https://coral.ai/docs/edgetpu/models-intro/
    cmd = 'edgetpu_compiler --version'
    help_url = 'https://coral.ai/docs/edgetpu/compiler/'
    assert platform.system() == 'Linux', f'export only supported on Linux. See {help_url}'
    if subprocess.run(f'{cmd} >/dev/null', shell=True).returncode != 0:
        LOGGER.info(f'\n{prefix} export requires Edge TPU compiler. Attempting install from {help_url}')
        sudo = subprocess.run('sudo --version >/dev/null', shell=True).returncode == 0  # sudo installed on system
        for c in (
                'curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add -',
                'echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" | sudo tee /etc/apt/sources.list.d/coral-edgetpu.list',
                'sudo apt-get update', 'sudo apt-get install edgetpu-compiler'):
            subprocess.run(c if sudo else c.replace('sudo ', ''), shell=True, check=True)
    ver = subprocess.run(cmd, shell=True, capture_output=True, check=True).stdout.decode().split()[-1]

    LOGGER.info(f'\n{prefix} starting export with Edge TPU compiler {ver}...')
    f = str(file).replace('.pt', '-int8_edgetpu.tflite')  # Edge TPU model
    f_tfl = str(file).replace('.pt', '-int8.tflite')  # TFLite model

    cmd = f"edgetpu_compiler -s -d -k 10 --out_dir {file.parent} {f_tfl}"
    subprocess.run(cmd.split(), check=True)
    return f, None


@try_export
def export_tfjs(file, prefix=colorstr('TensorFlow.js:')):
    # YOLO TensorFlow.js export
    check_requirements('tensorflowjs')
    import tensorflowjs as tfjs

    LOGGER.info(f'\n{prefix} starting export with tensorflowjs {tfjs.__version__}...')
    f = str(file).replace('.pt', '_web_model')  # js dir
    f_pb = file.with_suffix('.pb')  # *.pb path
    f_json = f'{f}/model.json'  # *.json path

    cmd = f'tensorflowjs_converter --input_format=tf_frozen_model ' \
          f'--output_node_names=Identity,Identity_1,Identity_2,Identity_3 {f_pb} {f}'
    subprocess.run(cmd.split())

    json = Path(f_json).read_text()
    with open(f_json, 'w') as j:  # sort JSON Identity_* in ascending order
        subst = re.sub(
            r'{"outputs": {"Identity.?.?": {"name": "Identity.?.?"}, '
            r'"Identity.?.?": {"name": "Identity.?.?"}, '
            r'"Identity.?.?": {"name": "Identity.?.?"}, '
            r'"Identity.?.?": {"name": "Identity.?.?"}}}', r'{"outputs": {"Identity": {"name": "Identity"}, '
            r'"Identity_1": {"name": "Identity_1"}, '
            r'"Identity_2": {"name": "Identity_2"}, '
            r'"Identity_3": {"name": "Identity_3"}}}', json)
        j.write(subst)
    return f, None


def add_tflite_metadata(file, metadata, num_outputs):
    # Add metadata to *.tflite models per https://www.tensorflow.org/lite/models/convert/metadata
    with contextlib.suppress(ImportError):
        # check_requirements('tflite_support')
        from tflite_support import flatbuffers
        from tflite_support import metadata as _metadata
        from tflite_support import metadata_schema_py_generated as _metadata_fb

        tmp_file = Path('/tmp/meta.txt')
        with open(tmp_file, 'w') as meta_f:
            meta_f.write(str(metadata))

        model_meta = _metadata_fb.ModelMetadataT()
        label_file = _metadata_fb.AssociatedFileT()
        label_file.name = tmp_file.name
        model_meta.associatedFiles = [label_file]

        subgraph = _metadata_fb.SubGraphMetadataT()
        subgraph.inputTensorMetadata = [_metadata_fb.TensorMetadataT()]
        subgraph.outputTensorMetadata = [_metadata_fb.TensorMetadataT()] * num_outputs
        model_meta.subgraphMetadata = [subgraph]

        b = flatbuffers.Builder(0)
        b.Finish(model_meta.Pack(b), _metadata.MetadataPopulator.METADATA_FILE_IDENTIFIER)
        metadata_buf = b.Output()

        populator = _metadata.MetadataPopulator.with_model_file(file)
        populator.load_metadata_buffer(metadata_buf)
        populator.load_associated_files([str(tmp_file)])
        populator.populate()
        tmp_file.unlink()


@smart_inference_mode()
def run(
        data=ROOT / 'data/coco.yaml',  # 'dataset.yaml path'
        weights=ROOT / 'yolo.pt',  # weights path
        imgsz=(640, 640),  # image (height, width)
        batch_size=1,  # batch size
        device='cpu',  # cuda device, i.e. 0 or 0,1,2,3 or cpu
        include=('torchscript', 'onnx'),  # include formats
        half=False,  # FP16 half-precision export
        inplace=False,  # set YOLO Detect() inplace=True
        keras=False,  # use Keras
        optimize=False,  # TorchScript: optimize for mobile
        int8=False,  # CoreML/TF INT8 quantization
        dynamic=False,  # ONNX/TF/TensorRT: dynamic axes
        simplify=False,  # ONNX: simplify model
        opset=12,  # ONNX: opset version
        verbose=False,  # TensorRT: verbose log
        workspace=4,  # TensorRT: workspace size (GB)
        nms=False,  # TF: add NMS to model
        agnostic_nms=False,  # TF: add agnostic NMS to model
        topk_per_class=100,  # TF.js NMS: topk per class to keep
        topk_all=100,  # TF.js NMS: topk for all classes to keep
        iou_thres=0.45,  # TF.js NMS: IoU threshold
        conf_thres=0.25,  # TF.js NMS: confidence threshold
):
    t = time.time()
    include = [x.lower() for x in include]  # to lowercase
    fmts = tuple(export_formats()['Argument'][1:])  # --include arguments
    flags = [x in include for x in fmts]
    assert sum(flags) == len(include), f'ERROR: Invalid --include {include}, valid --include arguments are {fmts}'
    jit, onnx, onnx_end2end, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle = flags  # export booleans
    file = Path(url2file(weights) if str(weights).startswith(('http:/', 'https:/')) else weights)  # PyTorch weights

    # Load PyTorch model
    device = select_device(device)
    if half:
        assert device.type != 'cpu' or coreml, '--half only compatible with GPU export, i.e. use --device 0'
        assert not dynamic, '--half not compatible with --dynamic, i.e. use either --half or --dynamic but not both'
    model = attempt_load(weights, device=device, inplace=True, fuse=True)  # load FP32 model

    # Checks
    imgsz *= 2 if len(imgsz) == 1 else 1  # expand
    if optimize:
        assert device.type == 'cpu', '--optimize not compatible with cuda devices, i.e. use --device cpu'

    # Input
    gs = int(max(model.stride))  # grid size (max stride)
    imgsz = [check_img_size(x, gs) for x in imgsz]  # verify img_size are gs-multiples
    im = torch.zeros(batch_size, 3, *imgsz).to(device)  # image size(1,3,320,192) BCHW iDetection

    # Update model
    model.eval()
    for k, m in model.named_modules():
        if isinstance(m, (Detect, DDetect, DualDetect, DualDDetect)):
            m.inplace = inplace
            m.dynamic = dynamic
            m.export = True

    for _ in range(2):
        y = model(im)  # dry runs
    if half and not coreml:
        im, model = im.half(), model.half()  # to FP16
    shape = tuple((y[0] if isinstance(y, (tuple, list)) else y).shape)  # model output shape
    metadata = {'stride': int(max(model.stride)), 'names': model.names}  # model metadata
    LOGGER.info(f"\n{colorstr('PyTorch:')} starting from {file} with output shape {shape} ({file_size(file):.1f} MB)")

    # Exports
    f = [''] * len(fmts)  # exported filenames
    warnings.filterwarnings(action='ignore', category=torch.jit.TracerWarning)  # suppress TracerWarning
    if jit:  # TorchScript
        f[0], _ = export_torchscript(model, im, file, optimize)
    if engine:  # TensorRT required before ONNX
        f[1], _ = export_engine(model, im, file, half, dynamic, simplify, workspace, verbose)
    if onnx or xml:  # OpenVINO requires ONNX
        f[2], _ = export_onnx(model, im, file, opset, dynamic, simplify)
    if onnx_end2end:
        labels = model.names
        f[2], _ = export_onnx_end2end(model, im, file, simplify, topk_all, iou_thres, conf_thres, device, len(labels))
    if xml:  # OpenVINO
        f[3], _ = export_openvino(file, metadata, half)
    if coreml:  # CoreML
        f[4], _ = export_coreml(model, im, file, int8, half)
    if any((saved_model, pb, tflite, edgetpu, tfjs)):  # TensorFlow formats
        assert not tflite or not tfjs, 'TFLite and TF.js models must be exported separately, please pass only one type.'
        assert not isinstance(model, ClassificationModel), 'ClassificationModel export to TF formats not yet supported.'
        f[5], s_model = export_saved_model(model.cpu(),
                                           im,
                                           file,
                                           dynamic,
                                           tf_nms=nms or agnostic_nms or tfjs,
                                           agnostic_nms=agnostic_nms or tfjs,
                                           topk_per_class=topk_per_class,
                                           topk_all=topk_all,
                                           iou_thres=iou_thres,
                                           conf_thres=conf_thres,
                                           keras=keras)
        if pb or tfjs:  # pb prerequisite to tfjs
            f[6], _ = export_pb(s_model, file)
        if tflite or edgetpu:
            f[7], _ = export_tflite(s_model, im, file, int8 or edgetpu, data=data, nms=nms, agnostic_nms=agnostic_nms)
            if edgetpu:
                f[8], _ = export_edgetpu(file)
            add_tflite_metadata(f[8] or f[7], metadata, num_outputs=len(s_model.outputs))
        if tfjs:
            f[9], _ = export_tfjs(file)
    if paddle:  # PaddlePaddle
        f[10], _ = export_paddle(model, im, file, metadata)

    # Finish
    f = [str(x) for x in f if x]  # filter out '' and None
    if any(f):
        cls, det, seg = (isinstance(model, x) for x in (ClassificationModel, DetectionModel, SegmentationModel))  # type
        dir = Path('segment' if seg else 'classify' if cls else '')
        h = '--half' if half else ''  # --half FP16 inference arg
        s = "# WARNING ⚠️ ClassificationModel not yet supported for PyTorch Hub AutoShape inference" if cls else \
            "# WARNING ⚠️ SegmentationModel not yet supported for PyTorch Hub AutoShape inference" if seg else ''
        if onnx_end2end:
            LOGGER.info(f'\nExport complete ({time.time() - t:.1f}s)'
                        f"\nResults saved to {colorstr('bold', file.parent.resolve())}"
                        f"\nVisualize:       https://netron.app")
        else:
            LOGGER.info(f'\nExport complete ({time.time() - t:.1f}s)'
                        f"\nResults saved to {colorstr('bold', file.parent.resolve())}"
                        f"\nDetect:          python {dir / ('detect.py' if det else 'predict.py')} --weights {f[-1]} {h}"
                        f"\nValidate:        python {dir / 'val.py'} --weights {f[-1]} {h}"
                        f"\nPyTorch Hub:     model = torch.hub.load('ultralytics/yolov5', 'custom', '{f[-1]}')  {s}"
                        f"\nVisualize:       https://netron.app")
    return f  # return list of exported files/dirs


def parse_opt():
    parser = argparse.ArgumentParser()
    parser.add_argument('--data', type=str, default=ROOT / 'data/coco.yaml', help='dataset.yaml path')
    parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolo.pt', help='model.pt path(s)')
    parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640, 640], help='image (h, w)')
    parser.add_argument('--batch-size', type=int, default=1, help='batch size')
    parser.add_argument('--device', default='cpu', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
    parser.add_argument('--half', action='store_true', help='FP16 half-precision export')
    parser.add_argument('--inplace', action='store_true', help='set YOLO Detect() inplace=True')
    parser.add_argument('--keras', action='store_true', help='TF: use Keras')
    parser.add_argument('--optimize', action='store_true', help='TorchScript: optimize for mobile')
    parser.add_argument('--int8', action='store_true', help='CoreML/TF INT8 quantization')
    parser.add_argument('--dynamic', action='store_true', help='ONNX/TF/TensorRT: dynamic axes')
    parser.add_argument('--simplify', action='store_true', help='ONNX: simplify model')
    parser.add_argument('--opset', type=int, default=12, help='ONNX: opset version')
    parser.add_argument('--verbose', action='store_true', help='TensorRT: verbose log')
    parser.add_argument('--workspace', type=int, default=4, help='TensorRT: workspace size (GB)')
    parser.add_argument('--nms', action='store_true', help='TF: add NMS to model')
    parser.add_argument('--agnostic-nms', action='store_true', help='TF: add agnostic NMS to model')
    parser.add_argument('--topk-per-class', type=int, default=100, help='TF.js NMS: topk per class to keep')
    parser.add_argument('--topk-all', type=int, default=100, help='ONNX END2END/TF.js NMS: topk for all classes to keep')
    parser.add_argument('--iou-thres', type=float, default=0.45, help='ONNX END2END/TF.js NMS: IoU threshold')
    parser.add_argument('--conf-thres', type=float, default=0.25, help='ONNX END2END/TF.js NMS: confidence threshold')
    parser.add_argument(
        '--include',
        nargs='+',
        default=['torchscript'],
        help='torchscript, onnx, onnx_end2end, openvino, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle')
    opt = parser.parse_args()

    if 'onnx_end2end' in opt.include:  
        opt.simplify = True
        opt.dynamic = True
        opt.inplace = True
        opt.half = False

    print_args(vars(opt))
    return opt


def main(opt):
    for opt.weights in (opt.weights if isinstance(opt.weights, list) else [opt.weights]):
        run(**vars(opt))


if __name__ == "__main__":
    opt = parse_opt()
    main(opt)