According to a blog, i have install fasttext in my python, by use pip install fasttext. I used it to process my dataset which is a data of classification question. It seem to work very well, the acceptable accuracy is achieved and run fast.
But today, i look at this repository carefully. I found the example in python's document is used train_supervised, but it in my code is supervised. Finally, i found these are two different libraries: fasttext and fastText!
I want to know the source of fasttext library, and whether these two libraries have some potential difference? or it just another wrap of the c code of this repository?
Hi @nomadlx,
The recommended method to install our python bindings for fastText is the following:
$ git clone https://github.com/facebookresearch/fastText.git
$ cd fastText
$ pip install .
Then, from python, you should use
import fastText
You can find examples on how to use the bindings in the folder python/doc/examples of our repo.
Best,
Edouard
@EdouardGrave
Imagine a world where every library recommended this process. How would you ever keep up? Pip & wheels exists to solve a problem you seem to be ignoring. No one can create anything repeatable out of raw git clones that could change day to day.
Hi @ruckc ,
We are addressing this issue, we will unify everything. Thank you for your patience :)
Regards,
Onur
Most helpful comment
@EdouardGrave
Imagine a world where every library recommended this process. How would you ever keep up? Pip & wheels exists to solve a problem you seem to be ignoring. No one can create anything repeatable out of raw git clones that could change day to day.