IndexError: index 0 is out of bounds for axis 0 with size 0

[MrKyaw]

Hi Scotthawley,

I tried to run the train_network.py few days already but still got errors message "IndexError: index 0 is out of bounds for axis 0 with size 0"
It was happened in panotti\ datautils.py of line 237 and 238 and couldn't manage to solve until now even I debugged many times.
X[load_count,:,:] = melgram
Y[load_count,:] = this_Y
I used the data sets of drums and guitars not sure that affect to the training process and hopefully kindly advise me from someone.

Best regards

[MrKyaw]

Dear Dr Scotthawley,

Below are debugged result related to the above issue
X - float64 - (0,96,30571,2)
Y - float64 - (0,2)
mel_dims - tuple - (1,96,30571,2)
melgram - float16 - (1,96,30571,2)

Thanks and best regards

[funatsufumiya]

@MrKyaw
I got same error, but I found the reason is sample count is smaller than batch size.

Please try python train_network.py --batch_size=N (where N is file count)