{ "checkpoint": { "algorithm_id": "ppo", "filename": "model.zip", "kind": "checkpoint", "model_class": "stable_baselines3.ppo.ppo.PPO", "sha256": "df8d1f4bfa889fa788fbdd1c22439088cddfebea81bf40eb0ad9282e992eacf1", "size_bytes": 21054228, "step": 7500000 }, "document_type": "rlab.model", "format_version": 1, "policy": { "algorithm_id": "ppo", "model_class": "stable_baselines3.ppo.ppo.PPO", "training_backend_config_hash": "4a6293118b258140c5a2285db3a2d4e49bdf370706890b27bb462191edf67380", "training_backend_id": "sb3.ppo" }, "provenance": { "goal_slug": "Level4-1", "kind": "checkpoint", "metadata_version": 2, "queue_train_job_id": 14, "recipe_slug": "base", "repo_git_commit": "044dcc3d2a978fb1a77524822238c4c7814c002d", "run_description": "Seed 1 runs the Level4-1 transfer of the successful Level1-1 B55 low-KL late-decay recipe, created to test whether that policy recipe transfers without changing PPO or reward hyperparameters while overriding only Level4-1 identity and reporting metadata.", "run_name": "Level4-1_base_s1_20260704T182045Z", "runtime_image_ref": "docker:ghcr.io/tsilva/gradlab/gradlab-train@sha256:f358d35bce7a87c17cbf0e642523c2daa78ef4d904895f867208c1cc25e4e323", "seed": 1, "training_metadata": { "env_config": { "action_set": "simple", "clip_rewards": false, "completion_reward": 0.0, "death_penalty": 25.0, "done_on_events": [ "life_loss", "level_change" ], "env_provider": "supermariobrosnes-turbo", "env_threads": 4, "env_wrappers": [ { "id": "SuperMarioBrosNesProgressInfoWrapper", "kwargs": {} }, { "id": "SuperMarioBrosNesRewardEnvWrapper", "kwargs": { "completion_reward": 0.0, "death_penalty": 25.0, "progress_reward_scale": 1.0, "reward_mode": "score", "reward_scale": 10.0, "score_progress_clipped": false, "terminal_reward": 50.0 } } ], "frame_skip": 4, "game": "SuperMarioBros-Nes-v0", "hud_crop_top": -1, "info_events": {}, "max_episode_steps": 4500, "max_pool_frames": false, "no_progress_min_delta": 0, "no_progress_timeout_steps": 0, "obs_crop": [ 32, 0, 0, 0 ], "obs_resize_algorithm": "area", "observation_size": 84, "progress_reward_cap": 30.0, "progress_reward_scale": 1.0, "reward_mode": "score", "reward_scale": 10.0, "score_progress_clipped": false, "state": "Level4-1", "state_distribution": [], "state_probs": [], "state_sampling_mode": "single", "states": [], "sticky_action_prob": 0.0, "task_conditioning": false, "task_conditioning_info_values": [], "task_conditioning_info_vars": [], "terminal_reward": 50.0, "time_penalty": 0.0, "use_retro_reward": false }, "environment": { "env_id": "supermariobrosnes-turbo:SuperMarioBros-Nes-v0", "preprocessing": { "frame_skip": 4, "frame_stack": 4, "max_pool_frames": false, "obs_copy": "safe_view", "obs_crop": [ 32, 0, 0, 0 ], "obs_crop_fill": 0, "obs_crop_mode": "remove", "obs_grayscale": true, "obs_resize": [ 84, 84 ], "obs_resize_algorithm": "area", "pipeline": "supermariobrosnes_turbo_native_vec_env", "policy_observation_layout": "channel_first", "sticky_action_prob": 0.0 }, "provider_args": { "frame_stack": 4, "info": "data", "info_filter": "all", "inttype": "stable", "noop_reset_max": 0, "obs_copy": "safe_view", "obs_grayscale": true, "obs_layout": "chw", "obs_type": "image", "players": 1, "record": false, "render_mode": "rgb_array", "reward_clip": false, "rom_path": null, "scenario": "scenario", "use_restricted_actions": "filtered" }, "schema_version": 2, "state": "Level4-1", "task": { "action": { "set": "simple" }, "events": { "level_change": { "operation": "change", "signal": "level" }, "life_loss": { "operation": "decrease", "signal": "lives" } }, "id": "mario", "reward": { "clip_rewards": false, "completion_reward": 0.0, "death_penalty": 25.0, "progress_reward_cap": 30.0, "progress_reward_scale": 1.0, "reward_mode": "score", "reward_scale": 10.0, "score_progress_clipped": false, "terminal_reward": 50.0, "time_penalty": 0.0, "use_native_reward": false }, "signals": { "level": [ "levelHi", "levelLo" ], "lives": "lives", "score": "score", "x": [ "xscrollHi", "xscrollLo" ] }, "termination": { "failure": [ "life_loss" ], "max_episode_steps": 4500, "success": [ "level_change" ] } } }, "environment_hash": "sha256:fa0f3dd473fac528938a2cc2130dfe2d516ef164946b6545f2262ff6777c4bef", "preprocessing": { "frame_skip": 4, "frame_stack": 4, "max_pool_frames": false, "obs_copy": "safe_view", "obs_crop": [ 32, 0, 0, 0 ], "obs_crop_fill": 0, "obs_crop_mode": "remove", "obs_grayscale": true, "obs_resize": [ 84, 84 ], "obs_resize_algorithm": "area", "pipeline": "supermariobrosnes_turbo_native_vec_env", "policy_observation_layout": "channel_first", "sticky_action_prob": 0.0 }, "versions": { "stable_baselines3": "2.8.0", "stable_retro_turbo": "1.0.1.post5", "supermariobrosnes_turbo": "0.2.5" } }, "training_metadata_hash": "5cbb8afd5c4b1b6ef34cab6e6953076b8fcd5725c1b5bfe93de85d43a6f2ca7a", "wandb_project": "SuperMarioBros-Nes-v0", "wandb_run_id": "4xrsz2v1", "wandb_run_path": "tsilva/SuperMarioBros-Nes-v0/4xrsz2v1" }, "recipe": { "document_type": "rlab.recipe", "filename": "recipe.json", "format_version": 1, "sha256": "0eed87cd734f16a4e0b23165cab978a01a465563167b3d240e16b5c26c106139", "size_bytes": 10832 } }