ValueError: Layer #314 (named "time_distributed_2"), weight <tf.Variable 'time_distributed_2/kernel:0' shape=(2048, 4096) dtype=float32_ref> has shape (2048, 4096), but the saved weight has shape (4096, 4096).

[Qcatbot]

I trained using densenet121_weights_tf_dim_ordering_tf_kernels_notop.h5. But when I test, I get the following error. Any suggestions?

WARNING:tensorflow:From /usr/local/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:3445: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.
Instructions for updating:
Please use rate instead of keep_prob. Rate should be set to rate = 1 - keep_prob.
Loading weights from models/densenet/voc.hdf5
Traceback (most recent call last):
File "test_frcnn.py", line 185, in
model_classifier.load_weights(C.model_path, by_name=True)
File "/usr/local/lib/python3.6/site-packages/keras/engine/network.py", line 1163, in load_weights
reshape=reshape)
File "/usr/local/lib/python3.6/site-packages/keras/engine/saving.py", line 1149, in load_weights_from_hdf5_group_by_name
str(weight_values[i].shape) + '.')
ValueError: Layer #314 (named "time_distributed_2"), weight <tf.Variable 'time_distributed_2/kernel:0' shape=(2048, 4096) dtype=float32_ref> has shape (2048, 4096), but the saved weight has shape (4096, 4096).

[kentaroy47]

Sorry I haven't tried out densenet for this repo.
Resnets should work fine and achieve similar accuracy.

[ravi-lourem]

@kentaroy47 Hi..Need for help
all_imgs[filename]['width'] = cols
NameError: name 'cols' is not defined ..

this happen simple parser... class mapping check but can solve..