IBM / unitxt / 15953443032

81%

push

github

Add to_markdown() to InstanceScores to pretty print output (#1846)

* Add wrapped to_markdown() method to InstanceScores to pretty print the results

Signed-off-by: Yoav Katz <katz@il.ibm.com>

* Updated examples

Signed-off-by: Yoav Katz <katz@il.ibm.com>

* Updated example to run with existing models

Signed-off-by: Yoav Katz <katz@il.ibm.com>

* Fixed max token renaming in WatsonSDK Chat enginer

Signed-off-by: Yoav Katz <katz@il.ibm.com>

* Improved summary printout of InstanceScores

Signed-off-by: Yoav Katz <katz@il.ibm.com>

* USed new summary feature

Signed-off-by: Yoav Katz <katz@il.ibm.com>

* Change lite lllm test to use large model for better and more consistent results

Signed-off-by: Yoav Katz <katz@il.ibm.com>

* Update InstanceScore summary to ignore columsn that do not exist.

Signed-off-by: Yoav Katz <katz@il.ibm.com>

---------

Signed-off-by: Yoav Katz <katz@il.ibm.com>

1538 of 1910 branches covered (80.52%)

Branch coverage included in aggregate %.

10466 of 12897 relevant lines covered (81.15%)

0.81 hits per line