minor

python3 -m vagen.trainer.main_ppo \

    algorithm.adv_estimator=bi_level_gae \

    algorithm.high_level_gamma=0.95 \

    algorithm.high_level_gamma=1.0 \

    data.train_files=data/frozenlake-vision-debug/train.parquet \

    data.val_files=data/frozenlake-vision-debug/test.parquet \

    data.train_batch_size=16 \

    data.train_batch_size=64 \

    data.max_prompt_length=1024 \

    data.max_response_length=128 \

    data.max_trajectory_length=2048 \

    rollout_manager.use_gae_mask=True \

    trainer.val_before_train=True \

    trainer.val_generations_to_log_to_wandb=8 \

    rollout_manager.n_trajectory=8 \

    rollout_manager.n_trajectory=2 \

    2>&1 | tee trico_frozenlake_vision.log