search

facebookresearch / gtn_applications

Applications using the GTN library and code to reproduce experiments in "Differentiable Weighted Finite-State Transducers"
MIT License
80 stars 7 forks source link

RNN model correction #14

Closed ronitd closed 3 years ago

ronitd commented 3 years ago

RNN model's default parameters contain stride. There are 2 convolutional layers each with a stride of 2. So for num_feature = 80, get reduced to 80 ->40 ->20. Updated the code to integrate it.

ronitd commented 3 years ago

Made the changes.

awni commented 3 years ago

Thank you!