facebookresearch / StarSpace

Learning embeddings for classification, retrieval and ranking.
MIT License
3.95k stars 531 forks source link

output of trainMode 5 #260

Open Fahad92 opened 5 years ago

Fahad92 commented 5 years ago

I have trained a model with trainMode 5 and it successfully got trained with 0 known labels. I am trying to predict the labels/corpus of the words/sentences from the trained model. Is there a way I can do that in StarSpace.

ledw commented 5 years ago

@Fahad92 trainMode 5 is for word only and label is not included in the training. Why do you want to predict label from the trained model?

Fahad92 commented 5 years ago

Thank you for your reply @ledw. I am trying to get the corpus value in which the words are stored. There is an implementation with a different algorithm: https://towardsdatascience.com/applying-machine-learning-to-classify-an-unsupervised-text-document-e7bb6265f52. I am trying to do a similar thing with StarSpace. I trained the model with trainMode 5 and it trained successfully, but I need to find out the corpus in which different words are stored. Is there any way to do it?