Tensorboard Info

Example specifying tensorboard callback for Classification training

Example specifying tensorboard callback for Object Detection API training

NOTE: Object detection API uses scripts that have the Tensorboard callback specified inside of them as default. It will be stored in the train subdirectory of where training results are

stored

Example Output Tensorboard Files From Training

Example code to run Tensorboard from Local Machine(assuming it is installed)

ASSUMES: that you have installed TensorBoard in your environment AND when you trained you have specified a TensorBoard callback to save necessary files

%load_ext tensorboard
print(mylog_dir)
%tensorboard --logdir=mylog_dir --host=local

Example code to run Tensorboard from Colab (or notebook).

# first install it 
!pip install tensorboard      
logs_dir = base_dir + "/efficientdetd0training/train"  
%ls {logs_dir}      
%reload_ext tensorboard  
%tensorboard --logdir {logs_dir}

Example results of running Tensorboard from local machine for an ObjectDetection API model --NON

NOTE: when running on command line it will launch a web page on localhost. When running from within a notebook it will runn in the notebook itself (but, will less control)