How to save custom embedding matrix to .txt file format?

Question

I have made a dictionary which contains word and its corresponding word vector in the following format:

{'word1': array([ 4.530e-02, -1.170e-02, -1.201e-01,  2.439e-01,  4.670e-02d], type=float32),
'word2': array([ 4.530e-02, -1.170e-02, -1.201e-01,  2.439e-01,  4.670e-02d], type=float32)}

I would like to save this dictionary to custom_embeddings.txt file in the following format:

The format of your custom_embeddings.txt file needs to be the token followed by the values of each of the dimensions for the embedding, all separated by a single space, eg here's two tokens with 5 dimensional embeddings:

word1 4.530e-02 -1.170e-02 -1.201e-01  2.439e-01  4.670e-02d
word2 4.530e-02 -1.170e-02 -1.201e-01  2.439e-01  4.670e-02d

It will be really helpful if you could tell me how to achieve this result?

Thanks in advance

Answer 1

Python's .items() call is an elegant way to loop over all the words in your dictionary. This will save the output as lines of a text file:

txt_filename = 'output.txt'

with open(txt_filename, 'w') as f:
    for word, vec in my_wordvec_dict.items():
        f.write('{} {}\n'.format(word, ' '.join(['{:e}'.format(item) for item in vec])))