isc/isc.py

# Copyright 2021 Zilliz. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import logging
import os
import warnings
from typing import Union, List
from pathlib import Path

import towhee
from towhee.operator.base import NNOperator, OperatorFlag
from towhee.types.arg import arg, to_image_color
from towhee import register
from towhee.models import isc
import sys
# from towhee.dc2 import accelerate

import torch
from torch import nn
from PIL import Image as PILImage
from timm.data import create_transform
from timm import create_model

try:
    from timm.models import get_pretrained_cfg
except ImportError:
    from timm.models.registry import _model_default_cfgs
    def get_pretrained_cfg(model_name):
        return _model_default_cfgs[model_name]

warnings.filterwarnings('ignore')
log = logging.getLogger('isc_op')
_ = sys.modules[__name__]


# @accelerate
class Model:
    def __init__(self, timm_backbone, checkpoint_path, device):
        self.device = device
        self.backbone = create_model(timm_backbone, features_only=True, pretrained=False)
        self.model = isc.create_model(pretrained=True, checkpoint_path=checkpoint_path, device=self.device,
                                      backbone=self.backbone, p=1.0, eval_p=1.0)
        self.model.eval()

    def __call__(self, x):
        x = x.to(self.device)
        return self.model(x)


@register(output_schema=['vec'])
class Isc(NNOperator):
    """
    The operator uses pretrained ISC model to extract features for an image input.

    Args:
        skip_preprocess (`bool = False`):
            Whether skip image transforms.
    """

    def __init__(self,
                 timm_backbone: str = 'tf_efficientnetv2_m_in21ft1k',
                 img_size: int = 512,
                 checkpoint_path: str = None,
                 skip_preprocess: bool = False,
                 device: str = None) -> None:
        super().__init__()
        if device is None:
            device = 'cuda' if torch.cuda.is_available() else 'cpu'
        self.device = device
        self.skip_tfms = skip_preprocess
        self.timm_backbone = timm_backbone

        if checkpoint_path is None:
            checkpoint_path = os.path.join(str(Path(__file__).parent), 'checkpoints', timm_backbone + '.pth')

        self.model = Model(self.timm_backbone, checkpoint_path, self.device)

        self.tfms = create_transform(
            input_size=img_size,
            interpolation=self.config['interpolation'],
            mean=self.config['mean'],
            std=self.config['std'],
            crop_pct=self.config['crop_pct']
        )

    def __call__(self, data: Union[List['towhee.types.Image'], 'towhee.types.Image']):
        if not isinstance(data, list):
            imgs = [data]
        else:
            imgs = data

        img_list = []
        for img in imgs:
            img = self.convert_img(img)
            img = img if self.skip_tfms else self.tfms(img)
            img_list.append(img)
        inputs = torch.stack(img_list)
        inputs = inputs.to(self.device)
        features = self.model(inputs)
        features = features.to('cpu')

        if isinstance(data, list):
            vecs = list(features.detach().numpy())
        else:
            vecs = features.squeeze(0).detach().numpy()
        return vecs

    @property
    def _model(self):
        return self.model.model

    @property
    def config(self):
        config = get_pretrained_cfg(self.timm_backbone)
        return config

    def save_model(self, format: str = 'pytorch', path: str = 'default'):
        if path == 'default':
            path = str(Path(__file__).parent)
            path = os.path.join(path, 'saved', format)
            os.makedirs(path, exist_ok=True)
            name = self.timm_backbone.replace('/', '-')
            path = os.path.join(path, name)
            if format in ['pytorch', 'torchscript']:
                path = path + '.pt'
            elif format == 'onnx':
                path = path + '.onnx'
            else:
                raise ValueError(f'Invalid format {format}.')
        dummy_input = torch.rand(1, 3, 224, 224).to(self.device)
        if format == 'pytorch':
            torch.save(self._model, path)
        elif format == 'torchscript':
            try:
                try:
                    jit_model = torch.jit.script(self._model)
                except Exception:
                    jit_model = torch.jit.trace(self._model, dummy_input, strict=False)
                torch.jit.save(jit_model, path)
            except Exception as e:
                log.error(f'Fail to save as torchscript: {e}.')
                raise RuntimeError(f'Fail to save as torchscript: {e}.')
        elif format == 'onnx':
            try:
                torch.onnx.export(self._model,
                                  dummy_input,
                                  path,
                                  input_names=['input_0'],
                                  output_names=['output_0'],
                                  opset_version=14,
                                  dynamic_axes={
                                      'input_0': {0: 'batch_size', 2: 'height', 3: 'width'},
                                      'output_0': {0: 'batch_size', 1: 'dim'}
                                  },
                                  do_constant_folding=True
                                  )
            except Exception as e:
                log.error(f'Fail to save as onnx: {e}.')
                raise RuntimeError(f'Fail to save as onnx: {e}.')
        # todo: elif format == 'tensorrt':
        else:
            log.error(f'Unsupported format "{format}".')
        return path

    @arg(1, to_image_color('RGB'))
    def convert_img(self, img: towhee._types.Image):
        img = PILImage.fromarray(img.astype('uint8'), 'RGB')
        return img

    @property
    def supported_formats(self):
        return ['onnx']

    def train(self, training_config=None,
              train_dataset=None,
              eval_dataset=None,
              resume_checkpoint_path=None, **kwargs):
        from .train_isc import train_isc
        training_args = kwargs.pop('training_args', None)
        train_isc(self._model, training_args)

# if __name__ == '__main__':
#     from towhee import ops
#
#     path = 'https://github.com/towhee-io/towhee/raw/main/towhee_logo.png'
#
#     decoder = ops.image_decode.cv2()
#     img = decoder(path)
#
#     op = Isc()
#     out = op(img)
#     assert out.shape == (256,)