GensimBackend

[XiepengLi]

AttributeError: The vocab attribute was removed from KeyedVector in Gensim 4.0.0. Use KeyedVector's .key_to_index dict, .index_to_key list, and methods .get_vecattr(key, attr) and .set_vecattr(key, attr, new_val) instead. See https://github.com/RaRe-Technologies/gensim/wiki/Migrating-from-Gensim-3.x-to-4

[XiepengLi]

AttributeError: The vocab attribute was removed from KeyedVector in Gensim 4.0.0. Use KeyedVector's .key_to_index dict, .index_to_key list, and methods .get_vecattr(key, attr) and .set_vecattr(key, attr, new_val) instead. See https://github.com/RaRe-Technologies/gensim/wiki/Migrating-from-Gensim-3.x-to-4

vector_size = self.embedding_model.vector_size

[MaartenGr]

Could you share how you have initialized the model? Also, which version of KeyBERT and Gensim are you using?

[MaartenGr]

The last response has been a while ago, so I'll be closing this issue for now. Having said that, let me know if you are still experiencing this issue and I'll make sure to reopen it.

[dopc]

I have faced with the same problem,

how I have initialized the model:

import gensim.downloader as api
from keybert import KeyBERT

ft = api.load('fasttext-wiki-news-subwords-300')
kw_model = KeyBERT(model=ft)

the versions I am using:

gensim==4.2.0
keybert==0.6.0

and the error message:

File ~/miniconda3/envs/myenv/lib/python3.10/site-packages/keybert/backend/_gensim.py:53, in GensimBackend.embed(self, documents, verbose)
     40 def embed(self, documents: List[str], verbose: bool = False) -> np.ndarray:
     41     """Embed a list of n documents/words into an n-dimensional
     42     matrix of embeddings
     43 
   (...)
     50         that each have an embeddings size of `m`
     51     """
     52     vector_shape = self.embedding_model.word_vec(
---> 53         list(self.embedding_model.vocab.keys())[0]
     54     ).shape
     55     empty_vector = np.zeros(vector_shape[0])
     57     embeddings = []

File ~/miniconda3/envs/myenv/lib/python3.10/site-packages/gensim/models/keyedvectors.py:735, in KeyedVectors.vocab(self)
    733 @property
    734 def vocab(self):
--> 735     raise AttributeError(
    736         "The vocab attribute was removed from KeyedVector in Gensim 4.0.0.\n"
    737         "Use KeyedVector's .key_to_index dict, .index_to_key list, and methods "
    738         ".get_vecattr(key, attr) and .set_vecattr(key, attr, new_val) instead.\n"
    739         "See [https://github.com/RaRe-Technologies/gensim/wiki/Migrating-from-Gensim-3.x-to-4](https://github.com/RaRe-Technologies/gensim/wiki/Migrating-from-Gensim-3.x-to-4%3C/span%3E%3Cspan) style="color:rgb(175,0,0)">"
    740     )

AttributeError: The vocab attribute was removed from KeyedVector in Gensim 4.0.0.
Use KeyedVector's .key_to_index dict, .index_to_key list, and methods .get_vecattr(key, attr) and .set_vecattr(key, attr, new_val) instead.
See https://github.com/RaRe-Technologies/gensim/wiki/Migrating-from-Gensim-3.x-to-4

BR.

[MaartenGr]

@dopc The way Gensim accesses the embeddings in 4.0 was chanced and not yet updated in KeyBERT. For now, using Gensim < 4.0 should work without any issues. I'll make sure it gets updated in the upcoming released.

[dopc]

Thanks so much!