cross_embodiment_transfer/train_bc.py at main · ExistentialRobotics/cross_embodiment_transfer · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
import pathlib
import argparse
import time
from ruamel.yaml import YAML

import numpy as np
import torch
from torch.utils.tensorboard import SummaryWriter

import utils
import replay_buffer
from bc import BCObsActAgent as Agent


def parse_args():
    parser = argparse.ArgumentParser()
    # environment
    parser.add_argument('--config', help='train config file path')
    args = parser.parse_args()
    return args

def main():
    args = parse_args()
    yaml = YAML(typ='safe')
    params = yaml.load(open(args.config))

    ##################################
    ### CREATE DIRECTORY FOR LOGGING
    ##################################
    if params['expert_folder'] is not None:
        demo_dir = (pathlib.Path(params['expert_folder']) /
            params['env_name'] / params['robots'] / params['controller_type']).resolve()

    if params['logdir_prefix'] is None:
        logdir_prefix = pathlib.Path(__file__).parent
    else:
        logdir_prefix = pathlib.Path(params['logdir_prefix'])
    data_path = logdir_prefix / 'logs' / time.strftime("%m.%d.%Y")
    logdir = '_'.join([
        time.strftime("%H-%M-%S"),
        params['env_name'],
        params['robots'],
        params['controller_type'],
        params['suffix']
    ])
    logdir = data_path / logdir
    params['logdir'] = str(logdir)
    print(params)

    # dump params
    logdir.mkdir(parents=True, exist_ok=True)
    import yaml
    with open(logdir / 'params.yml', 'w') as fp:
        yaml.safe_dump(params, fp, sort_keys=False)

    model_dir = logdir / 'models'
    pathlib.Path(model_dir).mkdir(parents=True, exist_ok=True)

    ##################################
    ### SETUP ENV, AGENT
    ##################################
    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

    env = utils.make_robosuite_env(params['env_name'], robots=params['robots'],
        controller_type=params['controller_type'], **params['env_kwargs'])

    obs = env.reset()
    robot_obs_shape = np.concatenate([obs[k] for k in params['robot_obs_keys']]).shape
    obj_obs_shape = np.concatenate([obs[k] for k in params['obj_obs_keys']]).shape

    params['obs_keys'] = params['robot_obs_keys'] + params['obj_obs_keys']

    env = utils.make(
        params['env_name'],
        robots=params['robots'],
        controller_type=params['controller_type'],
        obs_keys=params['obs_keys'],
        seed=params['seed'],
        **params['env_kwargs'],
    )
    obs_shape = env.observation_space.shape
    act_shape = env.action_space.shape

    print(f"Environment observation space shape {obs_shape}")
    print(f"Environment action space shape {act_shape}")

    eval_env = utils.make(
        params['env_name'],
        robots=params['robots'],
        controller_type=params['controller_type'],
        obs_keys=params['obs_keys'],
        seed=params['seed']+100,
        **params['env_kwargs'],
    )

    logger = SummaryWriter(log_dir=params['logdir'])

    obs_dims = {
        'obs_dim': obs_shape[0],
        'robot_obs_dim': robot_obs_shape[0],
        'obj_obs_dim': obj_obs_shape[0],
        'lat_obs_dim': params['lat_obs_dim']
    }
    act_dims = {
        'act_dim': act_shape[0],
        'lat_act_dim': params['lat_act_dim']
    }
    agent = Agent(
        obs_dims,
        act_dims,
        device,
    )

    agent_replay_buffer = replay_buffer.ReplayBuffer(
        obs_shape=obs_shape,
        action_shape=act_shape,
        capacity=2000000,
        batch_size=params['batch_size'],
        device=device
    )
    demo_paths = utils.load_episodes(demo_dir, params['obs_keys'])
    agent_replay_buffer.add_rollouts(demo_paths)

    for step in range(params['total_timesteps']):
        if step % params['evaluation']['interval'] == 0:
            print(f"Evaluating at step {step}")
            agent.eval_mode()
            utils.evaluate(eval_env, agent, 4, logger, step)
            agent.train_mode()
        if step % params['evaluation']['save_interval'] == 0:
            print(f"Saving model at step {step}")
            step_dir = model_dir / f"step_{step:07d}"
            pathlib.Path(step_dir).mkdir(parents=True, exist_ok=True)
            agent.save(step_dir)

        agent.update(agent_replay_buffer, logger, step)

    logger.close()

if __name__ == '__main__':
    main()