How can I deploy a pretrained fasttext model?

TheDoctor2035 · 03-13-2022 01:31 AM

Hi, I have this code :

"

import fasttext

ft = fasttext.load_model('pretrained model location')

"

Is there any way I could deploy this model using the google cloud platform? I've been looking around and there doesn't seem to be a way to do it. When I want to create a model it wants me to choose scikit or tensorflow but nothing allows me to deploy this model. Can anyone help? Thanks!

Horace

Hello,

Seeing that there is already an existing post [1] on Stack Overflow, kindly follow up on that post to avoid any work duplication.

[1] https://stackoverflow.com/questions/71455187/how-can-i-deploy-a-fasttext-model-on-google-cloud

TheDoctor2035

Hi! Thanks for answering, that question actually is mine, and there hasn't been an answer yet. Anyways I have a question, all I want to do, is deploy a couple pretrained fasttext models :

import fasttext

ft = fasttext.load_model('pretrainedmodelpath')

Just that, a couple times and be able to have an api get word vectors for words, which google cloud service would best be suited for this task? Most of the google cloud services want me to train my own model and don't let me just load a pretrained one. And in for example Dataflow, I can't just send text via the API and get an answer because it wants a file input. Any ideas? Thanks!