Hello, I’m following the EfficientViT Tutorial ( EfficientViT - NVIDIA Jetson AI Lab ), but when I run python3 ./benchmark.py, I get the following error:
root@app:/opt/efficientvit# python3 ./benchmark.py
/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py:123: FutureWarning: Using TRANSFORMERS_CACHE is deprecated and will be removed in v5 of Transformers. Use HF_HOME instead.
warnings.warn(
Namespace(model=‘l2’, weight_url=‘/data/models/efficientvit/sam/l2.pt’, multimask=False, image_path=‘assets/fig/cat.jpg’, output_path=‘/data/benchmarks/efficientvit_sam_demo.png’, images=[‘/data/images/hoover.jpg’, ‘/data/images/lake.jpg’, ‘/opt/efficientvit/assets/fig/cat.jpg’], mode=‘box’, point=None, box=‘[150,70,630,400]’, runs=2, warmup=1, save=‘/data/benchmarks/Efficient_ViT.txt’)
Traceback (most recent call last):
File “/opt/efficientvit/./benchmark.py”, line 289, in
main()
File “/opt/efficientvit/./benchmark.py”, line 162, in main
efficientvit_sam = create_sam_model(args.model, True, args.weight_url).cuda().eval()
File “/opt/efficientvit/efficientvit/sam_model_zoo.py”, line 47, in create_sam_model
weight = load_state_dict_from_file(weight_url)
File “/opt/efficientvit/efficientvit/models/utils/network.py”, line 72, in load_state_dict_from_file
checkpoint = torch.load(file, map_location=“cpu”)
File “/usr/local/lib/python3.10/dist-packages/torch/serialization.py”, line 1028, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File “/usr/local/lib/python3.10/dist-packages/torch/serialization.py”, line 1246, in _legacy_load
magic_number = pickle_module.load(f, **pickle_load_args)
_pickle.UnpicklingError: invalid load key, ‘<’.
I’m working with a Jetson AGX Orin 64Gb,
Could someone guide me on how to proceed?