chovanecm / sacredboard

Dashboard for sacred. Monitor and access your past machine learning experiments.
MIT License
183 stars 39 forks source link

Error using Tensorboard with 0.40 #83

Closed emited closed 6 years ago

emited commented 6 years ago

Hello,

First off, thanks a lot for your tool. I seem to be encountering errors with the master branch version of sacredboard. When I go to Tensorflow Logs and click on Run Tensorboard, I get the following error:

Internal Server Error

The server encountered an internal error and was unable to complete your request. Either the server is overloaded or there is an error in the application.

This error does not occur when installing with pip.

here is the trace:

  File "/home/debezenac/anaconda3/lib/python3.6/site-packages/flask/app.py", line 1982, in wsgi_app
    response = self.full_dispatch_request()
  File "/home/debezenac/anaconda3/lib/python3.6/site-packages/flask/app.py", line 1614, in full_dispatch_request
    rv = self.handle_user_exception(e)
  File "/home/debezenac/anaconda3/lib/python3.6/site-packages/flask/app.py", line 1517, in handle_user_exception
    reraise(exc_type, exc_value, tb)
  File "/home/debezenac/anaconda3/lib/python3.6/site-packages/flask/_compat.py", line 33, in reraise
    raise value
  File "/home/debezenac/anaconda3/lib/python3.6/site-packages/flask/app.py", line 1612, in full_dispatch_request
    rv = self.dispatch_request()
  File "/home/debezenac/anaconda3/lib/python3.6/site-packages/flask/app.py", line 1598, in dispatch_request
    return self.view_functions[rule.endpoint](**req.view_args)
  File "/home/debezenac/anaconda3/lib/python3.6/site-packages/sacredboard/app/webapi/routes.py", line 42, in run_tensorboard
    run = data.get(run_id)
AttributeError: 'PyMongoDataAccess' object has no attribute 'get'
chovanecm commented 6 years ago

I believe it has been fixed by you :-)

Sadly, there is still the problem that Sacredboard does not work with newer Tensorboard versions :-(

emited commented 6 years ago

Why is that?

chovanecm commented 6 years ago

I have not tried it yet, but I guess it has something to do with the way Tensorboard announces its port on the command line: Maybe it is enough to modify this line https://github.com/chovanecm/sacredboard/blob/develop/sacredboard/app/process/tensorboard.py#L52