piddnad
question about training
Hi @lecanhduc02 , currently we do not have the code for colab training. Here is a colab example of inference that might be helpful：https://github.com/camenduru/DDColor-colab
I found this code on https://modelscope.cn/models/iic/cv_ddcolor_image-colorization/summary. Can it be used for training? import os import tempfile
from modelscope.hub.snapshot_download import snapshot_download
from modelscope.msdatasets import MsDataset
from modelscope.msdatasets.dataset_cls.custom_datasets.image_colorization import
ImageColorizationDataset
from modelscope.trainers import build_trainer
from modelscope.utils.config import Config
from modelscope.utils.constant import DownloadMode, ModelFile
tmp_dir = tempfile.TemporaryDirectory().name if not os.path.exists(tmp_dir): os.makedirs(tmp_dir) model_id = 'damo/cv_ddcolor_image-colorization' cache_path = snapshot_download(model_id) config = Config.from_file( os.path.join(cache_path, ModelFile.CONFIGURATION))
dataset_train = MsDataset.load( 'imagenet-val5k-image', namespace='damo', subset_name='default', split='validation', download_mode=DownloadMode.REUSE_DATASET_IF_EXISTS)._hf_ds dataset_val = MsDataset.load( 'imagenet-val5k-image', namespace='damo', subset_name='default', split='validation', download_mode=DownloadMode.REUSE_DATASET_IF_EXISTS)._hf_ds
dataset_train = ImageColorizationDataset( dataset_train, config.dataset, is_train=True) dataset_val = ImageColorizationDataset( dataset_val, config.dataset, is_train=False)
kwargs = dict( model=model_id, train_dataset=dataset_train, eval_dataset=dataset_val, work_dir=tmp_dir) trainer = build_trainer(default_args=kwargs) trainer.train()
Yes, it is possible to use the modelscope library for training. However, the reliability of this method (i.e., the alignment of the final results with our model results) is not guaranteed, as our model training was completed through the training code provided in this repository.