I just added a note in my previous post here:
In fact, it seems that, in the second part of my test, the embeddings are not really changing. I certainly have inverted 2 files.
Are the fix_word_vecs options stored in the model, thus restored when no more defined on the command line ?
You are right, embeddings remain fixed after a retraining. This does not seem like a good behavior (unless you are using
-continue of course). I will look into it.
Perhaps these 2 fix_word_vecs options may be defined with 0/1 explicit values on the command line, rather than set/omitted with implicit values when omitted.
-continue is not used,
-fix_word_vecs_dec are correctly applied.
FYI, this approach is now on