def run_configuration(argv):
# params = ParameterServer(filename="examples/example_params/tfa_params.json")
params = ParameterServer()
# NOTE: Modify these paths in order to save the checkpoints and summaries
params["ML"]["BehaviorTFAAgents"][
"CheckpointPath"] = "/Users/hart/Development/bark-ml/checkpoints/"
params["ML"]["TFARunner"][
"SummaryPath"] = "/Users/hart/Development/bark-ml/summaries/"
params["World"]["remove_agents_out_of_map"] = True
# create environment
bp = ContinuousMergingBlueprint(params,
number_of_senarios=2500,
random_seed=0)
env = SingleAgentRuntime(blueprint=bp, render=False)
# PPO-agent
# ppo_agent = BehaviorPPOAgent(environment=env,
# params=params)
# env.ml_behavior = ppo_agent
# runner = PPORunner(params=params,
# environment=env,
# agent=ppo_agent)
# SAC-agent
sac_agent = BehaviorSACAgent(environment=env, params=params)
env.ml_behavior = sac_agent
runner = SACRunner(params=params, environment=env, agent=sac_agent)
if FLAGS.mode == "train":
runner.SetupSummaryWriter()
runner.Train()
elif FLAGS.mode == "visualize":
runner.Visualize(5)
def prepare_agent(agent, params, env): env.ml_behavior = agent runner = SACRunner(params=params, environment=env, agent=agent) iterator = iter(agent._dataset) runner._collection_driver.run() experience, _ = next(iterator) agent._agent.train(experience)
def run_rl_example(env, agent, params, mode="visualize"):
env.ml_behavior = agent
runner = SACRunner(params=params,
environment=env,
agent=agent)
if mode == "train":
runner.SetupSummaryWriter()
runner.Train()
elif mode == "visualize":
runner.Visualize(5)
elif mode == "evaluate":
runner.Evaluate()
def run_rl_example(env, agent, params, mode="visualize"):
env.ml_behavior = agent
runner = SACRunner(params=params,
environment=env,
agent=agent)
if mode == "train":
runner.SetupSummaryWriter()
runner.Train()
elif mode == "visualize":
runner.Run(num_episodes=10, render=True)
elif mode == "evaluate":
runner.Run(num_episodes=10, render=False)
def run_configuration(argv):
# Uncomment one of the following default parameter filename definitions,
# depending on which GNN library you'd like to use.
# File with standard parameters for tf2_gnn use:
# param_filename = "examples/example_params/tfa_sac_gnn_tf2_gnn_default.json"
# File with standard parameters for spektral use:
param_filename = "examples/example_params/tfa_sac_gnn_spektral_default.json"
params = ParameterServer(filename=param_filename)
# NOTE: Modify these paths to specify your preferred path for checkpoints and summaries
# params["ML"]["BehaviorTFAAgents"]["CheckpointPath"] = "YOUR_PATH"
# params["ML"]["TFARunner"]["SummaryPath"] = "YOUR_PATH"
#viewer = MPViewer(
# params=params,
# x_range=[-35, 35],
# y_range=[-35, 35],
# follow_agent_id=True)
#viewer = VideoRenderer(
# renderer=viewer,
# world_step_time=0.2,
# fig_path="/your_path_here/training/video/")
# create environment
bp = ContinuousHighwayBlueprint(params,
number_of_senarios=2500,
random_seed=0)
observer = GraphObserver(params=params)
env = SingleAgentRuntime(blueprint=bp, observer=observer, render=False)
sac_agent = BehaviorGraphSACAgent(environment=env,
observer=observer,
params=params)
env.ml_behavior = sac_agent
runner = SACRunner(params=params, environment=env, agent=sac_agent)
if FLAGS.mode == "train":
runner.SetupSummaryWriter()
runner.Train()
elif FLAGS.mode == "visualize":
runner.Visualize(5)
elif FLAGS.mode == "evaluate":
runner.Evaluate()
def run_configuration(argv):
params = ParameterServer(
filename="examples/example_params/tfa_params.json")
# params = ParameterServer()
# NOTE: Modify these paths in order to save the checkpoints and summaries
# params["ML"]["BehaviorTFAAgents"]["CheckpointPath"] = "/Users/hart/Development/bark-ml/checkpoints_merging_nn/"
# params["ML"]["TFARunner"]["SummaryPath"] = "/Users/hart/Development/bark-ml/checkpoints_merging_nn/"
params["Visualization"]["Agents"]["Alpha"]["Other"] = 0.2
params["Visualization"]["Agents"]["Alpha"]["Controlled"] = 0.2
params["Visualization"]["Agents"]["Alpha"]["Controlled"] = 0.2
params["ML"]["VisualizeCfWorlds"] = False
params["ML"]["VisualizeCfHeatmap"] = True
params["World"]["remove_agents_out_of_map"] = False
viewer = MPViewer(params=params,
x_range=[-35, 35],
y_range=[-35, 35],
follow_agent_id=True)
# create environment
bp = ContinuousMergingBlueprint(params, num_scenarios=10000, random_seed=0)
env = SingleAgentRuntime(blueprint=bp, render=False, viewer=viewer)
# PPO-agent
# ppo_agent = BehaviorPPOAgent(environment=env,
# params=params)
# env.ml_behavior = ppo_agent
# runner = PPORunner(params=params,
# environment=env,
# agent=ppo_agent)
# SAC-agent
sac_agent = BehaviorSACAgent(environment=env, params=params)
env.ml_behavior = sac_agent
runner = SACRunner(params=params, environment=env, agent=sac_agent)
if FLAGS.mode == "train":
runner.SetupSummaryWriter()
runner.Train()
elif FLAGS.mode == "visualize":
runner.Run(num_episodes=50, render=True)
elif FLAGS.mode == "evaluate":
runner.Run(num_episodes=100, render=False)
def run_configuration(argv):
params = ParameterServer()
# NOTE: Modify these paths to specify your preferred path for checkpoints and summaries
# params["ML"]["BehaviorTFAAgents"]["CheckpointPath"] = "/Users/hart/Development/bark-ml/checkpoints_merge_spektral_att2/"
# params["ML"]["TFARunner"]["SummaryPath"] = "/Users/hart/Development/bark-ml/checkpoints_merge_spektral_att2/"
#viewer = MPViewer(
# params=params,
# x_range=[-35, 35],
# y_range=[-35, 35],
# follow_agent_id=True)
#viewer = VideoRenderer(
# renderer=viewer,
# world_step_time=0.2,
# fig_path="/your_path_here/training/video/")
# create environment
bp = ContinuousMergingBlueprint(params, num_scenarios=2500, random_seed=0)
observer = GraphObserver(params=params)
env = SingleAgentRuntime(blueprint=bp, observer=observer, render=False)
sac_agent = BehaviorGraphSACAgent(environment=env,
observer=observer,
params=params,
init_gnn='init_interaction_network')
env.ml_behavior = sac_agent
runner = SACRunner(params=params, environment=env, agent=sac_agent)
if FLAGS.mode == "train":
runner.SetupSummaryWriter()
runner.Train()
elif FLAGS.mode == "visualize":
runner.Run(num_episodes=10, render=True)
elif FLAGS.mode == "evaluate":
runner.Run(num_episodes=250, render=False)
from barkscape.server.base_server import BaseServer
from barkscape.server.base_handler import BaseHandler
from barkscape.server.runners.bark_ml_runner_runner import BARKMLRunnerRunner
def load_exp_runner(file_name):
return ExperimentRunner(
json_file=file_name,
mode=None,
random_seed=0)
if __name__ == "__main__":
# load experiment
exp_runner_gnn = load_exp_runner(
"/Users/hart/Development/bark-ml/experiments/configs/phd/01_hyperparams/gnns/merging_large_embedding.json")
runtime = exp_runner_gnn._experiment._runtime
# set buffered viewer
viewer = BufferedViewer()
runtime._viewer = viewer
runner = SACRunner(params=exp_runner_gnn._params,
environment=runtime,
agent=exp_runner_gnn._experiment._agent)
# run-stuff
logger = logging.getLogger()
bark_server = BaseServer(
runner=BARKMLRunnerRunner, runnable_object=runner, logger=logger)
bark_server.Start()
def run_configuration(argv):
params = ParameterServer()
# NOTE: Modify these paths to specify your preferred path for checkpoints and summaries
# params["ML"]["BehaviorTFAAgents"]["CheckpointPath"] = "/Users/hart/Development/bark-ml/checkpoints/"
# params["ML"]["TFARunner"]["SummaryPath"] = "/Users/hart/Development/bark-ml/checkpoints/"
params["Visualization"]["Agents"]["Alpha"]["Other"] = 0.2
params["Visualization"]["Agents"]["Alpha"]["Controlled"] = 0.2
params["Visualization"]["Agents"]["Alpha"]["Controlled"] = 0.2
params["ML"]["VisualizeCfWorlds"] = False
params["ML"]["VisualizeCfHeatmap"] = False
# params["ML"]["ResultsFolder"] = "/Users/hart/Development/bark-ml/results/data/"
# viewer = MPViewer(
# params=params,
# x_range=[-35, 35],
# y_range=[-35, 35],
# follow_agent_id=True)
# create environment
bp = ContinuousMergingBlueprint(params,
num_scenarios=2500,
random_seed=0)
observer = GraphObserver(params=params)
behavior_model_pool = []
for count, a in enumerate([-5., 0., 5.]):
local_params = params.AddChild("local_"+str(count))
local_params["BehaviorConstantAcceleration"]["ConstAcceleration"] = a
behavior = BehaviorConstantAcceleration(local_params)
behavior_model_pool.append(behavior)
env = CounterfactualRuntime(
blueprint=bp,
observer=observer,
render=False,
params=params,
behavior_model_pool=behavior_model_pool)
sac_agent = BehaviorGraphSACAgent(environment=env,
observer=observer,
params=params)
env.ml_behavior = sac_agent
runner = SACRunner(params=params,
environment=env,
agent=sac_agent)
if FLAGS.mode == "train":
runner.SetupSummaryWriter()
runner.Train()
elif FLAGS.mode == "visualize":
runner._environment._max_col_rate = 0.
runner.Run(num_episodes=1, render=True)
elif FLAGS.mode == "evaluate":
for cr in np.arange(0, 1, 0.1):
runner._environment._max_col_rate = cr
runner.Run(num_episodes=250, render=False, max_col_rate=cr)
runner._environment._tracer.Save(
params["ML"]["ResultsFolder"] + "evaluation_results_runtime.pckl")
goal_reached = runner._tracer.success_rate
runner._tracer.Save(
params["ML"]["ResultsFolder"] + "evaluation_results_runner.pckl")