Sovit Ranjan Rath

@Sowgandh6 Yes, it is solved. You do not need to install `vision_transformers` to use Faster RCNN VitDet models. However, you will need to do `pip install vision_transformers` if you wish...

@edehino The checkpoint will have to have a `model_state_dict` key where the state dictionary should be saved. Else the script won't be able to load the pretrained weihts.

@edehino You can directly use COCO pretrained weights just by using `--model fasterrcnn_resnet50_fpn_v2`. It is one of the best models in the repository and directly uses the PyTorch pretrained model.

Hi. To train without pretrained weights, just go into `models/fasterrcnn_resnet50_fnp_v2.py` and make `weights=None`. Only a few hundred parameters are frozen when fine-tuning. Almost 99% of the model gets fine-tuned when...

Hi @DungHD-1997 The `output` is the key name that we provide when exporting the model. You will find this code in the `export.py` file. ``` dynamic_axes={ 'input' : {0 :...

By default, I have not added that functionality. I will need to take a look at whether that's possible or not.

Sure.

Hello. It seems that you do not have write access to the disk. Can you try running the code in a terminal as Administrator.

Got. I had not seen that you did not pass the epochs argument in the previous command. Glad that it was solved.

@tuanmanh1410 It is just a training and inference pipeline repository. Not an end-to-end one, i.e. does not contain deployment and conversion code yet. But it contains almost everything that you...