I got error when I run server.py in iris-GCE. I am not sure but issue may due to different version , which version of fastai do you use ?
fastai 2.1.8
torch 1.7.1
torchvision 0.8.2

(myenv) E:\user\Notebooks\AP\iris-ai\iris-GCE\app>python server.py serve
Traceback (most recent call last):
File "server.py", line 107, in
learn = loop.run_until_complete(asyncio.gather(*tasks))[0]
File "E:\user\anaconda\envs\myenv\lib\asyncio\base_events.py", line 573, in run_until_complete
return future.result()
File "server.py", line 76, in setup_learner
learn = load_learner(path/model_file_name)
File "E:\user\anaconda\envs\myenv\lib\site-packages\fastai\learner.py", line 553, in load_learner
res = torch.load(fname, map_location='cpu' if cpu else None)
File "E:\user\anaconda\envs\myenv\lib\site-packages\torch\serialization.py", line 595, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File "E:\user\anaconda\envs\myenv\lib\site-packages\torch\serialization.py", line 774, in _legacy_load
result = unpickler.load()
AttributeError: Can't get attribute 'FlattenedLoss' on <module 'fastai.layers' from 'E:\user\anaconda\envs\myenv\lib\site-packages\fastai\layers.py'>