Loading fasttext binary model from s3 fails

Question

I am hosting a pretrained fasttext model on s3 (uncompressed) and I am trying to load it in a lambda function. I am using the gensim.models.fasttext module to load the model:

from gensim.models.fasttext import load_facebook_vectors

def load_model(obj):
    model = load_facebook_vectors(obj["path"])

with obj["path"] being the s3 path, but I keep getting the following error:

"errorMessage": "fileno"
"errorType": "UnsupportedOperation"
"stackTrace": [
...
"  File \"/var/task/gensim/models/fasttext.py\", line 784, in load_facebook_vectors\n    full_model = _load_fasttext_format(path, encoding=encoding, full_model=False)\n"
"  File \"/var/task/gensim/models/fasttext.py\", line 808, in _load_fasttext_format\n    m = gensim.models._fasttext_bin.load(fin, encoding=encoding, full_model=full_model)\n"
"  File \"/var/task/gensim/models/_fasttext_bin.py\", line 348, in load\n    vectors_ngrams = _load_matrix(fin, new_format=new_format)\n"
"  File \"/var/task/gensim/models/_fasttext_bin.py\", line 282, in _load_matrix\n    matrix = np.fromfile(fin, _FLOAT_DTYPE, count)\n"
]

Answer 1

The documentation for load_facebook_vectors says:

This function uses the smart_open library to open the path. The path may be on a remote host (e.g. HTTP, S3, etc).

There are examples of accessing S3 objects at smart_open. I have not personally tried this but I wanted to make sure you eliminated all options before deciding to forcibly download the object and access it locally.