ImageCaption Bot : ValueError: No gradients provided for any variable

when I train my model, it’s giving me this error : ValueError: No gradients provided for any variable: [‘dense/kernel:0’, ‘dense/bias:0’, ‘lstm/lstm_cell/kernel:0’, ‘lstm/lstm_cell/recurrent_kernel:0’, ‘lstm/lstm_cell/bias:0’, ‘dense_1/kernel:0’, ‘dense_1/bias:0’, ‘dense_2/kernel:0’, ‘dense_2/bias:0’].

link to project : https://drive.google.com/file/d/13qnjL9yojd94jWg8q9Zwti3l227mLrmn/view?usp=sharing

Unzip file
Run notebook named wordEmbedding_and_modelArchitecture and rest should be fine

1 Like

did u get any solution for above problem ?

SAME QUESTION, please help. Getting the same chunk of error.

Facing the same error. Did anyone found the fix ?!!!

Just found the fix !!!.
Tensorflow 2.0 + expects tuple from data_generator,
so make sure you’re yielding a tuple not array