Commit 4b45dab9 authored by jameskrw's avatar jameskrw
Browse files

updated truncation

parent 3893c056

scripts/debug/debug_alfworld_text_grpo_service/run.sh

+1 −1
Original line number Diff line number Diff line
@@ -24,7 +24,7 @@ python3 -m vagen.trainer.main_ppo \ 
    data.max_response_length=128 \

    data.max_trajectory_length=3600 \

    data.image_key=images \

    data.truncation=error \

    data.truncation=left \

    actor_rollout_ref.model.path=Qwen/Qwen2.5-0.5B-Instruct \

    actor_rollout_ref.actor.optim.lr=1e-6 \

    actor_rollout_ref.model.use_remove_padding=True \

scripts/debug/debug_alfworld_vision_grpo_service/run.sh

+1 −1
Original line number Diff line number Diff line
@@ -24,7 +24,7 @@ python3 -m vagen.trainer.main_ppo \ 
    data.max_response_length=128 \

    data.max_trajectory_length=3600 \

    data.image_key=images \

    data.truncation=error \

    data.truncation=left \

    actor_rollout_ref.model.path=Qwen/Qwen2.5-VL-3B-Instruct \

    actor_rollout_ref.actor.optim.lr=1e-6 \

    actor_rollout_ref.model.use_remove_padding=True \

scripts/debug/debug_frozen_lake_text_grpo_service/run.sh

+1 −1
Original line number Diff line number Diff line
@@ -23,7 +23,7 @@ python3 -m vagen.trainer.main_ppo \ 
    data.max_response_length=128 \

    data.max_trajectory_length=3600 \

    data.image_key=images \

    data.truncation=error \

    data.truncation=left \

    actor_rollout_ref.model.path=Qwen/Qwen2.5-0.5B-Instruct \

    actor_rollout_ref.actor.optim.lr=1e-6 \

    actor_rollout_ref.model.use_remove_padding=True \

scripts/debug/debug_navigation_vision_grpo_service/run.sh

+1 −1
Original line number Diff line number Diff line
@@ -23,7 +23,7 @@ python -m vagen.trainer.main_ppo \ 
    data.max_response_length=128 \

    data.max_trajectory_length=1800 \

    data.image_key=images \

    data.truncation=error \

    data.truncation=left \

    actor_rollout_ref.model.path=Qwen/Qwen2.5-VL-3B-Instruct \

    actor_rollout_ref.actor.optim.lr=1e-6 \

    actor_rollout_ref.model.use_remove_padding=True \

scripts/debug/debug_primitive_skill_text_grpo_service/run.sh

+1 −1
Original line number Diff line number Diff line
@@ -24,7 +24,7 @@ python -m vagen.trainer.main_ppo \ 
    data.max_response_length=128 \

    data.max_trajectory_length=1800 \

    data.image_key=images \

    data.truncation=error \

    data.truncation=left \

    actor_rollout_ref.model.path=Qwen/Qwen2.5-0.5B-Instruct \

    actor_rollout_ref.actor.optim.lr=1e-6 \

    actor_rollout_ref.model.use_remove_padding=True \

CRA Git | Maintained and supported by SUSTech CRA and CCSE