FastText bigram vs words words sentences
Given the input file, I want to get dictionary vectors for bitrams like "new york", "machine learning", etc. I am just wondering which is the correct approach to make from the following two approaches.
Approach 1:
#Change the word_ngrams parameter to two
./fasttext skipgram -input file.txt -output model -wordNgrams 2
Approach 2:
#Train with default parameters
./fasttext skipgram -input file.txt -output model
#Obtain word vectors for bigrams (where `queries.txt` includes bigrams such as new york",
"machine learning" etc.)
./fasttext print-word-vectors model.bin < queries.txt
Please let me know which approach is correct?
0
user8566323
source
to share
No one has answered this question yet
See similar questions:
or similar: