(args: Args)
| 53 | |
| 54 | |
| 55 | def run_path_integral(args: Args): |
| 56 | |
| 57 | rng = jax.random.PRNGKey(seed=args.seed) |
| 58 | |
| 59 | update_fn = { |
| 60 | "mppi": softmax_update, |
| 61 | "cma-es": cma_es_update, |
| 62 | "cem": cem_update, |
| 63 | }[args.update_method] |
| 64 | |
| 65 | ## setup env |
| 66 | |
| 67 | # recommended temperature for envs |
| 68 | temp_recommend = { |
| 69 | "ant": 0.1, |
| 70 | "halfcheetah": 0.4, |
| 71 | "hopper": 0.1, |
| 72 | "humanoidstandup": 0.1, |
| 73 | "humanoidrun": 0.1, |
| 74 | "walker2d": 0.1, |
| 75 | "pushT": 0.2, |
| 76 | } |
| 77 | Nrefine_recommend = { |
| 78 | "pushT": 200, |
| 79 | "humanoidrun": 300, |
| 80 | } |
| 81 | Nsample_recommend = { |
| 82 | "humanoidrun": 8192, |
| 83 | } |
| 84 | Hsample_recommend = { |
| 85 | "pushT": 40, |
| 86 | } |
| 87 | if not args.disable_recommended_params: |
| 88 | args.temp_sample = temp_recommend.get(args.env_name, args.temp_sample) |
| 89 | args.Nrefine = Nrefine_recommend.get(args.env_name, args.Nrefine) |
| 90 | args.Nsample = Nsample_recommend.get(args.env_name, args.Nsample) |
| 91 | args.Hsample = Hsample_recommend.get(args.env_name, args.Hsample) |
| 92 | print(f"override temp_sample to {args.temp_sample}") |
| 93 | env = mbd.envs.get_env(args.env_name) |
| 94 | Nx = env.observation_size |
| 95 | Nu = env.action_size |
| 96 | # env functions |
| 97 | step_env_jit = jax.jit(env.step) |
| 98 | reset_env_jit = jax.jit(env.reset) |
| 99 | eval_us = jax.jit(functools.partial(mbd.utils.eval_us, step_env_jit)) |
| 100 | render_us = functools.partial( |
| 101 | mbd.utils.render_us, step_env_jit, env.sys.replace(dt=env.dt) |
| 102 | ) |
| 103 | |
| 104 | rng, rng_reset = jax.random.split(rng) # NOTE: rng_reset should never be changed. |
| 105 | state_init = reset_env_jit(rng_reset) |
| 106 | |
| 107 | ## run path interal |
| 108 | |
| 109 | mu_0T = jnp.zeros([args.Hsample, Nu]) |
| 110 | |
| 111 | @jax.jit |
| 112 | def update_once(carry, unused): |
no test coverage detected