training over new dataset

[agoel-cs]

I am training it over new dataset. I am getting this error in save checkpoint 36/1850 batch done in 2.356s. at epoch 0.97. loss cost = 9.295156, reg cost = 0.000000, ppl2 = 4.59 (smooth 14.32) evaluating val performance in batches of 100 Traceback (most recent call last): File "driver.py", line 315, in main(params) File "driver.py", line 232, in main val_ppl2 = eval_split('val', dp, model, params, misc) # perform the evaluation on VAL set File "/root/neuraltalk/imagernn/imagernn_utils.py", line 38, in eval_split ppl2 = 2 ** (logppl / logppln) ZeroDivisionError: integer division or modulo by zero

[lebronze]

Well, i had the same problem. You can make the para 'logppl' and 'logppln' to a very little value, such as 0.0001 in the line 72,73 driver.py. That works for me.

[lebronze]

Dear @goelankur7 , did you finally train your multimodal RNN model successfully? And could you test you own images on your multimodal RNN model?

[infinity0a]

hey @lebronze, I had this problem as well, and I use the "0.0001". It still doesn't work. Do you have any other fixes?

[nyj-ocean]

@agoel-cs @lebronze @infinity0a how did you create the json file of your own dataset like dataset_coco.json file?