add new categories

utils.py

@@ -6,9 +6,13 @@ from typing import Literal, List
 import pandas as pd
 from huggingface_hub import HfFileSystem, hf_hub_download
 
+# from: https://github.com/lm-sys/FastChat/blob/main/fastchat/serve/monitor/monitor.py#L389
 KEY_TO_CATEGORY_NAME = {
     "full": "Overall",
+    "dedup": "De-duplicate Top Redundant Queries (soon to be default)",
     "coding": "Coding",
+    "hard_6": "Hard Prompts (Overall)",
+    "hard_english_6": "Hard Prompts (English)",
     "long_user": "Longer Query",
     "english": "English",
     "chinese": "Chinese",
@@ -16,10 +20,14 @@ KEY_TO_CATEGORY_NAME = {
     "no_tie": "Exclude Ties",
     "no_short": "Exclude Short Query (< 5 tokens)",
     "no_refusal": "Exclude Refusal",
+    "overall_limit_5_user_vote": "overall_limit_5_user_vote",
 }
 CAT_NAME_TO_EXPLANATION = {
     "Overall": "Overall Questions",
+    "De-duplicate Top Redundant Queries (soon to be default)": "De-duplicate top redundant queries (top 0.1%). See details in [blog post](https://lmsys.org/blog/2024-05-17-category-hard/#note-enhancing-quality-through-de-duplication).",
     "Coding": "Coding: whether conversation contains code snippets",
+    "Hard Prompts (Overall)": "Hard Prompts (Overall): details in [blog post](https://lmsys.org/blog/2024-05-17-category-hard/)",
+    "Hard Prompts (English)": "Hard Prompts (English), note: the delta is to English Category. details in [blog post](https://lmsys.org/blog/2024-05-17-category-hard/)",
     "Longer Query": "Longer Query (>= 500 tokens)",
     "English": "English Prompts",
     "Chinese": "Chinese Prompts",
@@ -27,6 +35,7 @@ CAT_NAME_TO_EXPLANATION = {
     "Exclude Ties": "Exclude Ties and Bothbad",
     "Exclude Short Query (< 5 tokens)": "Exclude Short User Query (< 5 tokens)",
     "Exclude Refusal": 'Exclude model responses with refusal (e.g., "I cannot answer")',
+    "overall_limit_5_user_vote": "overall_limit_5_user_vote",
 }
 
 PROPRIETARY_LICENSES = ["Proprietary", "Proprietory"]