Open ErfolgreichCharismatisch opened 2 years ago
Push
I think you need to have sentences on a new line and many more samples. I used https://www.statmt.org/wmt14/training-monolingual-europarl-v7/europarl-v7.fr.gz and modified run.sh to use this file. Run it and see what output .txt files look like.
I followed
I used the
europarl-v7.de-en.de
dataset and tookwith
data.dev.txt
being a long string on one line from kaldi, a speech-to-text engine. It's all lowercase, sometimes wrong words and no punctuation.<model_output_path>
is equal todata.dev.txt
Is the solution to train more lines or do I have to preprocess
data.dev.txt
? If the latter, how?