Skipgram: Hang while processing small dataset

[facebookresearch]

Skipgram: Hang while processing small dataset

[pyk]

For small dataset minCount should be explicitly set to avoid this hang problem

% cat data.txt 
simple
statement
% ./fasttext skipgram -input data.txt -output model -minCount 1                                                                                                                      
Read 0M words
Number of words:  3
Number of labels: 0
Progress: 100.0%  words/sec/thread: 5  lr: 0.000001  loss: -nan  eta: 0h0m 
Train time: 5.000000 sec

I think it's better for fasttext(1) to throw an error if the case like this happen, not just hang

[EdouardGrave]

Yes, I agree. I will fix this tomorrow.

[EdouardGrave]

This should be fixed now.

[pyk]

Thanks! 👍

On Friday, 2 September 2016, Edouard Grave <[email protected]
javascript:_e(%7B%7D,'cvml','[email protected]');> wrote:

This should be fixed now.

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
#63 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/ACHHDst2eimC5jhhX00kUGLnPhl9EMLPks5qmCJggaJpZM4JtBTx
.

Bayu Aldi Yansyah,
Universitas Airlangga
https://github.com/pyk

[tianyunzqs]

some uncommon result showed as follow (but no error) when fasttext used.
The version of Python is 3.5
I get the result as follow with the introduction in https://github.com/miguelgfierro/sciblog_support/blob/master/Intro_to_NLP_with_fastText/Intro_to_NLP.ipynb
does anyone know how to fix it ?

-------uncommon result-------
3.5.3 (v3.5.3:1880cb95a742, Jan 16 2017, 16:02:32) [MSC v.1900 64 bit (AMD64)]
P@1: nan
R@1: nan
Number of examples: 0