str2vec

str2vec copied to clipboard

关于训练时长

2

谢谢你的开源代码，str2vec貌似要运行很长的时间 请问你当时的训练语料有多长，训练的时候花了多长时间!,语料很大的时候是不是可以把Max iterations设小一点？

hit-joseph

ZeroDivisionError: division by zero

1

Hi, when I run the command ,I meet the problem ./mpi-train.sh 10 Instances file: /users4/xyan/experiment/phrase/phrase_vec_embedding/str2vec-master/demo-data/str2vec-demo/input/sample-training-file.txt Model file: /users4/xyan/experiment/phrase/phrase_vec_embedding/str2vec-master/demo-data/str2vec-demo/output/sample-training-file.mpi-10.model.gz Word vector file: /users4/xyan/experiment/phrase/phrase_vec_embedding/str2vec-master/demo-data/str2vec-demo/input/sample-word-vectors-trained-by-word2vec.txt lambda_reg: 0.149999999999999994 Max iterations: 200 load word vectors......

hit-joseph

I'm getting `ValueError: invalid literal for int() with base 10:` for strings with non-ascii symbols, e.g Finnish äö. Any thoughts on how to fix? Thanks

Coolnesss

Hi, MAy i Know how do i get the phrasal count. for the phrases generated by giza++. It would be great if I am getting an earnest reply.

sanjanasri