Seq2Seq Translation (NLP video 12)

[@osunkoyaopeoluwa8102]

There is a mistake in the notebook. The ft object is actually an instance of fasttext.FastText. Change `import fasttext as ft` to `from fasttext import FastText as ft`

[@AardraKA]

Rachel is an amazing teacher, and so is Jeremy. Can we please stop tearing down teachers? Honestly it is just upsetting to see these mean comments. If you are so upset about it, dont watch

[@alexandershepard3534]

A trick: you can watch series at Flixzone. Me and my gf have been using it for watching all kinds of movies during the lockdown.

[@aaronazariah3069]

@Alexander Shepard yea, I've been using flixzone for months myself :D

[@stefansavev2519]

Apologies for pointing this out, but the German translation looks wrong. It looks like it was translated via Google Translate which gets it wrong. The idiomatic translation is "He loves food" - "Er mag Essen". I also checked that "er liebte zu essen" gives 10 results in Google while "Er mag Essen" gives 18.500 results. Given that both are meant to mean the same, the more common phrase should be preferred.

[@anthonymalkoun6188]

At 4:04 the second question "Who are we?" is incorrectly translated into "Where are we?". The french question should be "Qui sommes-nous?". Just pointing out how knowing the languages you're working with can make it easier for you, mistakes in the database can be dangerous.

[@SupeHero00]

21:16 Why do we need another linear layer in the encoder?

[@bobli3467]

I was trying to run src = Seq2SeqTextList.from_df(df, cols='fr').split_by_rand_pct(0.2).label_from_df(cols='en', label_cls=TextList) from Lesson 7 in a terminal instead of a Jupyter notebook . But I keep getting . I really appreciate it if somebody could help me with that~

[@TheAnubhav27]

There is too much abstraction. Difficult to understand

[@dtm-555]

BLEU: 27:30

[@gopesh97]

I really like the work you guys are doing at fast.ai. I wish that Jeremy Howard should've taught this course. The way you teach is really hard to follow, it feels like that you yourself are not sure/confident about what you're teaching, so how can a student get confidence in these topics. Honest Opinion. I wish if there is a 2020 lecture series on NLP by fast ai, it should be taught by Jeremy.

[@zardarkhan9913]

Thanks for that assessment Freud. Helpful and specific feedback is better warranted rather than a sweeping generalization. Honest Opinion.

[@vladimirgetselevich4704]

Not very clear explanation about Attention

[@vladimirgetselevich4704]

Will she stop to say so much "kind of" expression? Is not she sure what she is talking about?

[@zardarkhan9913]

You suck