UI: Enhancements on the layout, wordings, etc

ファイル、テスト、またはエントリーポイントは指定されていません。まず、この issue の未チェック項目に照らして既存の評価画面と設定画面を確認し、変更を加える前に、対象を絞ったサブセットと完了基準について合意してください。

Describe the current behavior
Currently the frontend has been built with keeping only one thing in mind that it is clear enough to run evaluations, config management etc. UI-wise our frontend may not be the best, we connected with noopur during sprint and these are the things she pointed out that we can take care of to make the UI better, More things can be added to this list also

Describe the enhancement you'd like

• Naming can be clearer, that users can understand
• Back button on Configuration bucket, also browse library may not be the best idea
• Laundry list of old evals that have been run can be a good addition
• using specific colour codes when showing scores to show bad or good score
• evaluation page layout can be - the first thing we see is a list of old evals, and then we have a button to create a new eval, and we click on it and we select which evaluation to run
• eval metric overview can be smaller, does not have to take much space, the detailed results part deserves more space to take
• lots of texts in score reasoning, maybe turn them to points/takeaways
• markdown should not show up in texts in answers/ground truth
• question and ground truth column should be kept together, and then answer column comes, also these columns can be shown in a better way noopur will come back to us for this
• evaluation score comparision, with particular filters
• for evaluation, dont show the timestamp it took but if you have to show the amount of time it took
• change configuration library name?
• "datasets" can come under evaluation
• the word configuration by itself does not mean much
• when we compress the menu, there still should be smth we can see is there
• under configurations, maybe it should not be called "prompt editor", it should be called "config editor"