search

ml5js / training-charRNN

Training charRNN model for ml5js
Other
96 stars 46 forks source link

Fix hyperparameter naming in readme, add learning resources #9

Closed rafajak closed 4 years ago

rafajak commented 5 years ago

Hello there,

I made some changes in the hyperparameters section to reflect the argument names the model accepts, namely:

I also changed the suggested value of the latter from 0.25 to 0.75, as I'm assuming the author suggests dropout rate of 0.25, which corresponds to output_keep_prob of 0.75 in TensorFlow's DropoutWrapper.

Additionally, I added the learning resources Daniel lists in issue #7.

Thanks so much for your work on ml5 and tutorials, it's an absolute joy to play with. :)

Cheers, Raf

shiffman commented 4 years ago

Thank you for this, sorry for the long delay is taking a look!