Update the list of models for evaluation

yourbench_space/leaderboard_space/env.py

@@ -2,16 +2,14 @@ import os
 
 
 INIT_MODELS = [
-    # 70B
-    ("Qwen/Qwen2.5-72B-Instruct", "novita"),
-    ("meta-llama/Llama-3.3-70B-Instruct", "novita"),
-    ("deepseek-ai/DeepSeek-R1-Distill-Llama-70B", "novita"),
-    # 20 to 30B
-    ("Qwen/QwQ-32B", "novita"),
-    ("mistralai/Mistral-Small-24B-Instruct-2501", "together"),
+    # ("meta-llama/Llama-4-Scout-17B-16E-Instruct", "together"),
+    ("Qwen/Qwen3-32B", "novita"),
+    ("Qwen/Qwen2.5-72B-Instruct", "together"),
+    ("deepseek-ai/DeepSeek-R1", "together"),
+    ("google/gemma-3-27b-it", "nebius"),
 ]
 MODELS = [m[0] for m in INIT_MODELS]
 TASK = os.getenv("TASK")
 # With storage
 HF_TOKEN = os.getenv("HF_TOKEN")
-ORG_NAME = os.getenv("ORG_NAME")
+ORG_NAME = os.getenv("ORG_NAME")