Closed zdx3578 closed 6 years ago
3 gpus and train in 4 thread
(coach_env) sdc@frankwang-Tri01:~/Projects/coach$ python3 coach.py -p Carla_A3C -n 4
/home/sdc/Projects/coach/coach_env/lib/python3.5/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float
to np.floating
is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type
.
from ._conv import register_converters as _register_converters
Warning: failed to import the following packages - RoboSchool, GymExtensions, PyBullet, ViZDoom, Neon
Please enter an experiment name: c
Distributed Training
/home/sdc/Projects/coach/coach_env/lib/python3.5/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float
to np.floating
is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type
.
from ._conv import register_converters as _register_converters
/home/sdc/Projects/coach/coach_env/lib/python3.5/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float
to np.floating
is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type
.
from ._conv import register_converters as _register_converters
E0319 23:24:41.996985406 12448 ev_epoll1_linux.c:1051] grpc epoll fd: 3
/home/sdc/Projects/coach/coach_env/lib/python3.5/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float
to np.floating
is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type
.
from ._conv import register_converters as _register_converters
/home/sdc/Projects/coach/coach_env/lib/python3.5/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float
to np.floating
is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type
.
from ._conv import register_converters as _register_converters
/home/sdc/Projects/coach/coach_env/lib/python3.5/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float
to np.floating
is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type
.
from ._conv import register_converters as _register_converters
/home/sdc/Projects/coach/coach_env/lib/python3.5/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float
to np.floating
is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type
.
from ._conv import register_converters as _register_converters
E0319 23:24:42.861840739 12456 ev_epoll1_linux.c:1051] grpc epoll fd: 3
E0319 23:24:42.964169270 12454 ev_epoll1_linux.c:1051] grpc epoll fd: 3
E0319 23:24:42.964269364 12455 ev_epoll1_linux.c:1051] grpc epoll fd: 3
E0319 23:24:42.964458426 12453 ev_epoll1_linux.c:1051] grpc epoll fd: 3
E0319 23:24:43.027589433 12452 ev_epoll1_linux.c:1051] grpc epoll fd: 3
Traceback (most recent call last):
File "./parallel_actor.py", line 101, in
Results stored at: ./experiments/c/19_3_2018-23_24_0 Total runtime: 0:00:17.242453
Do you want to discard the experiment results (Warning: this cannot be undone)? (y/N)
Using tensorflow framework
Traceback (most recent call last):
File "coach.py", line 271, in
fix by: and open server then add time sleep;
change: @property def measurements(self): assert True
@measurements.setter
def measurements(self, value):
assert True
@property
def observation(self):
assert True
@observation.setter
def observation(self, value):
assert True
to Ture
error:
Using tensorflow framework
Traceback (most recent call last):
File "coach.py", line 271, in
Hi,
Checkout commit 125c7ee38d413e4709271292c473f5930196b137 ! The latest commit seems to have broken quite a number of things..
# @property
# def observation(self):
# assert False
#
# @observation.setter
# def observation(self, value):
# assert False
#observation = self.step([1.0, 0]).state['observation']
observation = self.preprocess_observation(self.env.observation)
@ciumonk thanks!
We pushed some fixes to Carla in the following commit: f31159aad6e84d738732d22fc8aa1509c6cda275 This should solve the measurements / observation issues
carla version 0.7 and 0.71 both same error and tensorflow=cpu 14.1 and tensorflow-gpu=1.6 is samw error
(coach_env) sdc@frankwang-Tri01:~/Projects/coach$ python3 coach.py -p Carla_A3C -n 1 /home/sdc/Projects/coach/coach_env/lib/python3.5/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from
float
tonp.floating
is deprecated. In future, it will be treated asnp.float64 == np.dtype(float).type
. from ._conv import register_converters as _register_converters Warning: failed to import the following packages - RoboSchool, Neon, ViZDoom, PyBullet, GymExtensions Please enter an experiment name: cUsing tensorflow framework Traceback (most recent call last): File "coach.py", line 271, in
env_instance = create_environment(tuning_parameters)
File "/home/sdc/Projects/coach/environments/init.py", line 32, in create_environment
env = eval(env_type)(tuning_parameters)
File "/home/sdc/Projects/coach/environments/carla_environment_wrapper.py", line 136, in init
self.reset(True)
File "/home/sdc/Projects/coach/environments/environment_wrapper.py", line 165, in reset
self._restart_environment_episode(force_environment_reset)
File "/home/sdc/Projects/coach/environments/carla_environment_wrapper.py", line 229, in _restart_environment_episode
observation = self.step([1.0, 0])['observation']
File "/home/sdc/Projects/coach/environments/environment_wrapper.py", line 140, in step
self._update_state()
File "/home/sdc/Projects/coach/environments/carla_environment_wrapper.py", line 185, in _update_state
'measurements': [measurements.player_measurements.forward_speed],
File "/home/sdc/Projects/coach/environments/environment_wrapper.py", line 81, in observation
assert False
AssertionError
Results stored at: ./experiments/c/19_03_2018-23_30 Total runtime: 0:00:29.358656
Do you want to discard the experiment results (Warning: this cannot be undone)? (y/N)