Loading Q-TinySAM

[merveenoyan]

Hello! I came across two issues loading Q-TinySAM. First one is running demo_quant.py looking for normal checkpoints due to model type being vit_t, is this the proper type for loading quantized weights? Because due to this, during model loading it doesn't look for quantization layer and fails. Demo error:

Traceback (most recent call last):
  File "/content/TinySAM/./demo_quant.py", line 7, in <module>
    from demo import show_mask, show_points, show_box
  File "/content/TinySAM/demo.py", line 31, in <module>
    sam = sam_model_registry[model_type](checkpoint="./weights/tinysam.pth")
  File "/content/TinySAM/tinysam/build_sam.py", line 90, in build_sam_vit_t
    with open(checkpoint, "rb") as f:
FileNotFoundError: [Errno 2] No such file or directory: './weights/tinysam.pth'

When I try to infer myself with no demo, I get following error:

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
[<ipython-input-7-04f0ec6a92d3>](https://localhost:8080/#) in <cell line: 19>()
     17 
     18 cpt_path = "./tinysam/tinysam_w8a8.pth"
---> 19 quant_sam = torch.load(cpt_path)
     20 
     21 device = "cuda" if torch.cuda.is_available() else "cpu"

2 frames
[/usr/local/lib/python3.10/dist-packages/torch/serialization.py](https://localhost:8080/#) in find_class(self, mod_name, name)
   1413                     pass
   1414             mod_name = load_module_mapping.get(mod_name, mod_name)
-> 1415             return super().find_class(mod_name, name)
   1416 
   1417     # Load the data (which may in turn use `persistent_load` to load tensors)

ModuleNotFoundError: No module named 'quantization_layer'

[xinghaochen]

Hi,

The original codes in demo_quant.py try to import show_mask, show_points, show_box from demo.py, thus it tries to run demo.py. That why you encounter the above error, since no checkpoint tinysam.pth is found.

It is not elegant so we have just updated demo_quant.py to include the implementation of show_mask, show_points, show_box. Now the demo_quant.py should work well and you can have a try.

As for the error of ModuleNotFoundError: No module named 'quantization_layer', you could simply add sys.path.append("./tinysam") in your codes so that the tinysam.quantization_layer could be found.