dafnapension
dafnapension
A naïve question.. : Since the card is a json file, a simpler, and perhaps better understood way (addressing @yoavkatz concern) would be to first load that json file into...
reproducible via prepare.card.cola.py
For the case in question, where predictions = references = ['acceptable', 'acceptable', 'acceptable'], by the book:  we only have TP here (or only TN), all three other components are...
scikit's implementation faithfully follows the definition (as there is only TN or only TP, and all other **three** components are 0, hence the result, by definition of matthew_coef, is 0)....
Gladly, I think that in all of your cases, there is only a single input term that is 2, and the tree three others == 0, so the nominator is...
total loss for Matthews is -1, not 0:  I think that since Matthews returns 0, by definition, for any case that the nominator in the formula is 0, (namely:...
My 2 cents, having dug some: Rouge employs its default tokenizer as first step to computing the score. When lang = nepali, for example, no token identified, nor in prediction...
Hi Yoav, very nice and important catch! I managed to reproduce, and generated a test for it. And solved with deep copies. Please see [PR](https://github.com/IBM/unitxt/pull/379)
where? Copy has 'use_query' as a deprecated field. ``` (virtual39) dafna@DESKTOP-GM8R3J7:~/workspaces/unitxt$ cat tests/mine/copyfields.py from unitxt.operators import Copy from unitxt.stream import MultiStream instance = {"predictions" : { "a" : 3, "b"...
The warning is printed out (for no obvious reason) only in the preparation step, when the classes are built up, but not in a civilian use of the classes: ```...