Hi,

I get the (truncated) error message below when using the run_tile.sh script currently in github (and attached). It runs the pannuke_type model ... but with nr_type = 0.

I suspect the nr_type = 0 is a typo: When I set nr_type = 6, the script runs to successful completion.

Is nr_type = 6 correct for this model? (I copied it from run_wsi.sh, which uses the same model). In general, would we set this parameter by going back to the description of the corresponding data set to see the number of cell types annotated within it? I understand nr_type must be 0 for segmentation only.

Thank you for making your code available!

Best,
Brian

|2021-02-03|17:33:41.748| [INFO] .... Detect #GPUS: 1
Traceback (most recent call last):
File "run_infer.py", line 181, in
infer = InferManager(**method_args)
File "/pod/2/chuang-lab/USERS/whitebr/hover_net/infer/base.py", line 27, in init
self.__load_model()
File "/pod/2/chuang-lab/USERS/whitebr/hover_net/infer/base.py", line 68, in __load_model
net.load_state_dict(saved_state_dict, strict=True)
File "/home/whitebr/anaconda3/envs/hovernet-gpu/lib/python3.6/site-packages/torch/nn/modules/module.py", line 1045, in load_state_dict
self.class.name, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for HoVerNet:
Unexpected key(s) in state_dict: "decoder.tp.u3.dense.units.7.preact_bna/bn.weight", "decoder.tp.u3.dense.units.7.conv2.weight",
...
run_tile.sh.txt