File size: 1,396 Bytes
f3eaffd | 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 | import numpy as np
from rlbench.action_modes.action_mode import MoveArmThenGripper
from rlbench.action_modes.arm_action_modes import JointVelocity
from rlbench.action_modes.gripper_action_modes import Discrete
from rlbench.environment import Environment
from rlbench.observation_config import ObservationConfig
from rlbench.tasks import ReachTarget
class Agent(object):
def __init__(self, action_shape):
self.action_shape = action_shape
def act(self, obs):
arm = np.random.normal(0.0, 0.1, size=(self.action_shape[0] - 1,))
gripper = [1.0] # Always open
return np.concatenate([arm, gripper], axis=-1)
obs_config = ObservationConfig()
obs_config.set_all(True)
obs_config.gripper_touch_forces = False
action_mode = MoveArmThenGripper(
arm_action_mode=JointVelocity(), gripper_action_mode=Discrete())
env = Environment(
action_mode, obs_config=obs_config, headless=False,
robot_setup='sawyer')
env.launch()
task = env.get_task(ReachTarget)
agent = Agent(env.action_shape) # 6DoF + 1 for gripper
training_steps = 120
episode_length = 40
obs = None
for i in range(training_steps):
if i % episode_length == 0:
print('Reset Episode')
descriptions, obs = task.reset()
print(descriptions)
action = agent.act(obs)
print(action)
obs, reward, terminate = task.step(action)
print('Done')
env.shutdown()
|