search

mrdbourke / tensorflow-deep-learning

All course materials for the Zero to Mastery Deep Learning with TensorFlow course.
https://dbourke.link/ZTMTFcourse
MIT License
5.14k stars 2.53k forks source link

Changes made on the default parameters of Text Vectorizer #435

Open arghanath007 opened 2 years ago

arghanath007 commented 2 years ago

The parameter in Text Vectorizer, which is pad_to_max_tokens has been changes from True to False.

From The Video:

image

From the Docs

image I don't know if it will make my model perform worse or not. I haven't trained my model yet. But I did notice something which is, even if I set pad_to_max_tokens= False which is by default then also it is padding the integer tensor with zeroes at the end

image So I am a bit confused now. Let me train my model and check the accuracy score then I think the confusion will go away.

My Text Vectorizer Config:

image

arghanath007 commented 2 years ago

I don't think it had any or little effect on the outcome of the project. All worked out perfectly for me. I don't know I am still a bit confused about it. If anyone could explain what is happening that would be great. Thanks in advance for it.