search

chesterli29 / jupyterlab_tensorboard

Tensorboard extension for jupyterlab.
MIT License
316 stars 36 forks source link

Tensorboard 2.0 error #16

Closed juliusvonkohout closed 2 years ago

juliusvonkohout commented 5 years ago

Hello,

There is an incompatibility with tensorboard 2.0 which was just relased with tensorflow 2.0

Event Multiplexer initializing. Event Multiplexer done initializing Launching reload in a daemon thread TensorBoard reload process beginning Starting AddRunsFromDirectory: /opt/app-root/src/logs/fit GetLogdirSubdirectories: Starting to list directories via walking. Adding run from directory /opt/app-root/src/logs/fit/20191001-153636/validation Constructing EventAccumulator for /opt/app-root/src/logs/fit/20191001-153636/validation [E 15:48:17.963 LabApp] Uncaught exception POST /api/tensorboard?1569944897920 (127.0.0.1) HTTPServerRequest(protocol='http', host='hostname', method='POST', uri='/api/tensorboard?1569944897920', version='HTTP/1.1', remote_ip='127.0.0.1') Traceback (most recent call last): File "/opt/app-root/lib/python3.6/site-packages/tornado/web.py", line 1697, in _execute result = method(*self.path_args, *self.path_kwargs) File "/opt/app-root/lib/python3.6/site-packages/tornado/web.py", line 3174, in wrapper return method(self, args, **kwargs) File "/opt/app-root/lib/python3.6/site-packages/jupyter_tensorboard/api_handlers.py", line 40, in post .new_instance(data["logdir"], reload_interval=reload_interval) File "/opt/app-root/lib/python3.6/site-packages/jupyter_tensorboard/tensorboard_manager.py", line 148, in new_instance purge_orphaned_data=purge_orphaned_data) File "/opt/app-root/lib/python3.6/site-packages/jupyter_tensorboard/tensorboard_manager.py", line 46, in create_tb_app tensorboard.assets_zip_provider) File "/opt/app-root/lib/python3.6/site-packages/tensorboard/backend/application.py", line 157, in standard_tensorboard_wsgi flags, plugin_loaders, data_provider, assets_zip_provider, multiplexer) File "/opt/app-root/lib/python3.6/site-packages/jupyter_tensorboard/tensorboard_manager.py", line 113, in TensorBoardWSGIApp path_to_run = application.parse_event_files_spec(logdir) File "/opt/app-root/lib/python3.6/site-packages/tensorboard/backend/application.py", line 455, in parse_event_files_spec for specification in logdir_spec.split(','): AttributeError: 'Namespace' object has no attribute 'split' Adding run from directory /opt/app-root/src/logs/fit/20191001-153636/train Constructing EventAccumulator for /opt/app-root/src/logs/fit/20191001-153636/train ... [E 15:48:17.968 LabApp] 500 POST /api/tensorboard?1569944897920 (127.0.0.1) 19.22ms referer=https://hostname/lab? POST /api/tensorboard?1569944897920 500 20.322 ms - 1529 Done with AddRunsFromDirectory: /opt/app-root/src/logs/fit

jaekookang commented 4 years ago

Hi. I have the same issue too. Uninstalling and re-installing didn't fix the problem.

metya commented 4 years ago

Hi, in my case the same problem. Tensorboard just doesn't work.

chesterli29 commented 4 years ago

lspvic/jupyter_tensorboard#55 lspvic/jupyter_tensorboard/pull/56

metya commented 4 years ago

Yes, but unfortunately repo of jupyter_tensorboard not much activity of this time((

edoardogiacomello commented 4 years ago

lspvic/jupyter_tensorboard#55 lspvic/jupyter_tensorboard/pull/56

As some users pointed out, the new official method of running tensorboard inside a notebook doesn't work if behind a firewall. Moreover the UI of this extension was more simple, since didn't require to manually type in logdirs for each tensorboard instance you wanted to run. Is there any chance we well see an update of this? :)