Hi there, excited to be using Language Reactor! Thanks for what you’ve built.
I watched my first movie in Spanish using LR, and noticed that the English machine translations are still getting the genders confused. Partly I think this might be because the machine translation is doing each sentence piecemeal rather than looking at previous sentences for context.
(I should note that this movie (Volver) is only available with Spanish subtitles, no English subtitles. However, this seems to be the case for multiple films and TV shows I’m watching on Netflix (I am located in Spain).)
Here’s an example. Right before this image, Penelope Cruz’ character is talking about Paco (a man) but this subtitle screen it refers to an “it”. Of course in Spanish it is common for the pronoun to be omitted since it is partially embedded in the conjugation of the verb.
However in this case the conjugation is 3rd person singular, which could refer to it, him, or her. Not having this clarification makes it tough to understand the english translation and left me a little lost.
My suggestion was to consider using the context of the previous sentences to help establish the gender/pronouns for English machine translations. Another option was to somehow retrieve the ChatGPT typically does this although I know this costs money vs the Google Translate API. But I’d for one be willing to pay more/use my API key to get better English translations.
Thanks so much for your time!