Update main_ppo.py

main_ppo.py

@@ -17,13 +17,15 @@ Note that we don't combine the main with ray_trainer as ray_trainer is used by o
 
 from verl import DataProto
 import torch
-from verl.utils.reward_score import gsm8k, math, multiply, countdown, kk
+from verl.utils.reward_score import gsm8k, math, multiply, countdown, kk, amazonmix
 from verl.trainer.ppo.ray_trainer import RayPPOTrainer
 
 
 def _select_rm_score_fn(data_source):
     if data_source == 'openai/gsm8k':
         return gsm8k.compute_score
+    elif data_source == "IND_Diverse_Instruction":
+        return amazonmix.compute_score
     elif data_source == 'lighteval/MATH':
         return math.compute_score
     elif "multiply" in data_source or "arithmetic" in data_source: