remove unused epsilon-clip

LovelyBuggies · LovelyBuggies · commit f6b8e0801475 · 2025-10-05T10:43:45.000-04:00
diff --git a/configs/grpo_che_config.yaml b/configs/grpo_che_config.yaml
@@ -47,7 +47,6 @@ grpo:
   discount: 0.9
   termination_threshold: -0.1
   reward_shift: -2.1
-  epsilon_clip: null
 
 # wandb
 wandb:
diff --git a/configs/grpo_he_config.yaml b/configs/grpo_he_config.yaml
@@ -47,7 +47,6 @@ grpo:
   discount: 0.9
   termination_threshold: -0.1
   reward_shift: -2.1
-  epsilon_clip: null
 
 # wandb
 wandb:
diff --git a/configs/grpo_mbpp_config.yaml b/configs/grpo_mbpp_config.yaml
@@ -47,7 +47,6 @@ grpo:
   discount: 0.9
   termination_threshold: -0.1
   reward_shift: -2.1
-  epsilon_clip: null
 
 # wandb
 wandb:
diff --git a/configs/magrpo_che_config.yaml b/configs/magrpo_che_config.yaml
@@ -48,7 +48,6 @@ magrpo:
   discount: 0.9
   termination_threshold: -0.2
   reward_shift: -4
-  epsilon_clip: null
 
 # wandb
 wandb:
diff --git a/configs/magrpo_he_config.yaml b/configs/magrpo_he_config.yaml
@@ -46,7 +46,6 @@ magrpo:
   discount: 0.9
   termination_threshold: -0.2
   reward_shift: -4
-  epsilon_clip: null
 
 # wandb
 wandb:
diff --git a/configs/magrpo_mbpp_config.yaml b/configs/magrpo_mbpp_config.yaml
@@ -48,7 +48,6 @@ magrpo:
   discount: 0.9
   termination_threshold: -0.2
   reward_shift: -4
-  epsilon_clip: null
 
 # wandb
 wandb:
diff --git a/train_grpo.py b/train_grpo.py
@@ -366,7 +366,6 @@ def _resolver(prompt: str):
         discount=grpo_config.get("discount", 0.9),
         joint_mode=grpo_config.get("joint_mode", "aligned"),
         termination_threshold=grpo_config.get("termination_threshold", None),
-        epsilon_clip=grpo_config.get("epsilon_clip", None),
     )
 
     # ------------------------------------------------------------------
diff --git a/train_magrpo.py b/train_magrpo.py
@@ -419,7 +419,6 @@ def _resolver(prompt: str):
         discount=magrpo_config.get("discount", 0.9),
         joint_mode=magrpo_config.get("joint_mode", "aligned"),
         termination_threshold=magrpo_config.get("termination_threshold", None),
-        epsilon_clip=magrpo_config.get("epsilon_clip", None),
     )
 
     # ------------------------------------------------------------------

Original file line number	Diff line number	Diff line change
`@@ -366,7 +366,6 @@ def _resolver(prompt: str):`
`366`	`366`	`discount=grpo_config.get("discount", 0.9),`
`367`	`367`	`joint_mode=grpo_config.get("joint_mode", "aligned"),`
`368`	`368`	`termination_threshold=grpo_config.get("termination_threshold", None),`
`369`		`- epsilon_clip=grpo_config.get("epsilon_clip", None),`
`370`	`369`	`)`
`371`	`370`
`372`	`371`	`# ------------------------------------------------------------------`
Original file line number	Diff line number	Diff line change
`@@ -419,7 +419,6 @@ def _resolver(prompt: str):`
`419`	`419`	`discount=magrpo_config.get("discount", 0.9),`
`420`	`420`	`joint_mode=magrpo_config.get("joint_mode", "aligned"),`
`421`	`421`	`termination_threshold=magrpo_config.get("termination_threshold", None),`
`422`		`- epsilon_clip=magrpo_config.get("epsilon_clip", None),`
`423`	`422`	`)`
`424`	`423`
`425`	`424`	`# ------------------------------------------------------------------`