language-evaluation
language-evaluation copied to clipboard
bckim92
:clipboard: Collection of evaluation code for natural language generation.
According to the [original ROUGE paper](http://www.aclweb.org/anthology/W04-1013), current ROUGE implementation (from google-research) only support sentence-level LCS. Thus, summary-level LCS needs to be implemented.
Maybe you should change `sudo add-apt-repository ppa:webupd8team/java` to `sudo add-apt-repository ppa:ts.sch.gr/ppa`. I have tested this source, it worked.
Each assessment will print these times, how to turn off these printing prompts? SPICE evaluation took: 1.309 s PTBTokenizer tokenized 1885 tokens at 25046.90 tokens per second. PTBTokenizer tokenized 1894...
