Calvin Giles

I think what is required is to reinitialise the model state at the start of each line so that it doesn't learn to predict the start of the line based...

I did try that but I use other plugins (pandoc in particular) which are required during development. Sent from my Phone On 11 Jul 2014, at 13:12, Ben Smith [email protected]...

I have recently started to embed ipython notebooks into the site. These take a _very_ long time to index, so an ability to disable this for development is even more...

That sounds like exactly what I need! Thanks, I will try that in the morning. Sent from my Phone On 4 Aug 2014, at 15:28, Ben Smith [email protected] wrote: A...

This does works as intended. A flag remains preferable as it doesn't result in the script appearing as renamed in my git status, but it is fine as a workaround.

What do the extra `--` do? Is this going to be in a release any time soon? Otherwise I will look at integrating your changes.

Sure - here https://colab.research.google.com/drive/1gJSfUc9YwOxBAzviSGqmLQgS2q-uehj1#scrollTo=oTsyVxGFqGVt

Looks like the correct way to do this is: ``` import keras_nlp from tensorflow import keras inputs = keras.Input(shape=(), dtype="string") preprocessor = keras_nlp.models.BertPreprocessor.from_preset("bert_tiny_en_uncased") preprocessor(inputs) ``` for single sentences. It is...