twe-1: in except-except branch, train_sentence_topic not defined about topical_word_embeddings HOT 4 OPEN

You may notice the 'word2vec_inner.pyx' file. This is the code of cython which have been modified by author. You need to compile it at first, then you can import 'train_sentence_topic' defined in it.
Give more hints you may feel useful when running code.

1. In all three models, author have modified the source code of gensim and included them, so you needn't to install the gensim by yourself.
2. I failed to compile the word2vec_inner.pyx.pyx in the original folder. Then I only copied the word2vec_inner.pyx and head file to new folder. It succeed this time.
3. If you have compiled the *.pyx, only import the word2_inner.pyx is enough. You can delete import 'pyximport' that can prevent some errors.
4. When compiling *.pyx, you may meet the the error of scipy ‘from scipy.... import fabls’, it may happen for the version of scipy. You can change the code, in word2vec_inner.pyx, 'from 'scipy.linalg.blas import fblas' to 'import scipy.linalg.blas as fblas'

from topical_word_embeddings.

Speaking of the cython issue:
When I try to compile word2vec_inner.pyx, I get the following issue:

Error compiling Cython file:
------------------------------------------------------------
...
                k = sentence_len
            for j in range(j, k):
                if j == i or codelens[j] == 0:
                    continue
                if hs:
                    fast_sentence_topic(points[i], codes[i], codelens[i], syn0, syn1, syn0_topic, size, topic[j], _alpha, work)
                                                                                                            ^
------------------------------------------------------------

word2vec_inner.pyx:890:109: Cannot assign type 'uint32_t *' to 'uint32_t'

has anyone seen this before or know how to resolve it?

I already tried the fixes above and it seems like this issue is stopping me from running the code.

from topical_word_embeddings.

I meet the same error.

from topical_word_embeddings.

I also met the same error.

from topical_word_embeddings.