Problems in using pre-trained models
liverishabh opened this issue · comments
Hi @hpanwar08 ,
I was trying to use your pre-trained models for prediction by following your instructions and using the below command:
python demo/demo.py --config-file configs/DLA_mask_rcnn_X_101_32x8d_FPN_3x.yaml --input "<path to image.jpg>" --output <path to save the predicted image> --confidence-threshold 0.5 --opts MODEL.WEIGHTS <path to model_final_trimmed.pth> MODEL.DEVICE cpu
I am using Google Colab for the work and I am getting this error:
Traceback (most recent call last):
File "demo/demo.py", line 73, in <module>
demo = VisualizationDemo(cfg)
File "/content/drive/My Drive/PublayNet/detectron2/demo/predictor.py", line 35, in __init__
self.predictor = DefaultPredictor(cfg)
File "/usr/local/lib/python3.6/dist-packages/detectron2/engine/defaults.py", line 187, in __init__
checkpointer.load(cfg.MODEL.WEIGHTS)
File "/usr/local/lib/python3.6/dist-packages/fvcore/common/checkpoint.py", line 117, in load
checkpoint = self._load_file(path)
File "/usr/local/lib/python3.6/dist-packages/detectron2/checkpoint/detection_checkpoint.py", line 42, in _load_file
loaded = super()._load_file(filename) # load native pth checkpoint
File "/usr/local/lib/python3.6/dist-packages/fvcore/common/checkpoint.py", line 213, in _load_file
return torch.load(f, map_location=torch.device("cpu"))
File "/usr/local/lib/python3.6/dist-packages/torch/serialization.py", line 586, in load
with _open_zipfile_reader(f) as opened_zipfile:
File "/usr/local/lib/python3.6/dist-packages/torch/serialization.py", line 246, in __init__
super(_open_zipfile_reader, self).__init__(torch._C.PyTorchFileReader(name_or_buffer))
RuntimeError: [enforce fail at inline_container.cc:208] . file not found: /version
Found the issue.
Thanks for this great repository
Hey, sometimes I miss the notifications. Was the problem related to wrong model file?
Yeah, sort of.
I was using wget
to download the models to colab and it was getting downloaded in zipped format (although the extension was still '.pth').
got caught by this as well. thanks!