Open sfvnielsen opened 7 years ago
...
We now have fastText trained on our own dataset as word-embeddings. Should we also try with a pretrained fastText-set?
...