search

jtkim-kaist / VAD

Voice activity detection (VAD) toolkit including DNN, bDNN, LSTM and ACAM based VAD. We also provide our directly recorded dataset.
834 stars 232 forks source link

After how many iterations did your proposed VAD model converge ? #22

Open eskandarhassan opened 5 years ago

jtkim-kaist commented 5 years ago

This QUESTION is quite difficult to answer, because the necessary training time is according to the training data. Therefore, you should inspect your training & validation curve to find appropriate iteration number.

eskandarhassan commented 5 years ago

Hello, Thank You for the fast reply..

  1. Yes I meant on your dataset, Approximately, How many iterations did it take you to abort the training process using the proposed VAD model.
  2. Could you please confirm if the checkpoint in 'saved_model/temp/temp_ACAM' corresponds to the end of the training (i.e. final and best/optimised weights) or not.
jtkim-kaist commented 5 years ago

  1. Approximately 2 epochs for whole of my dataset I used.

  2. No, that checkpoint is not the optimized one. Now, I only provide the optimized .pb file. If you need the optimized ckpt, I will share for you.

Thx.

eskandarhassan commented 5 years ago

This comment is to kindly share the optimised checkpoint(.meta, .data and .index) of the ACAM model.

Thanks,