Files
Run-Skeleton-Run/trying_ddpg_with_implicit_dynamics.ipynb
T
2018-01-19 12:08:49 +08:00

15 KiB

In [1]:
import os
os.environ['CUDA_VISIBLE_DEVICES']=""
os.environ["PYTHONPATH"]='.'
In [2]:
%pylab --no-import-all inline
%reload_ext autoreload
%autoreload 2
Populating the interactive namespace from numpy and matplotlib
In [3]:
os.sys.argv="ddpg/train.py --logdir ./outputs/logs_ddpg2 \
--num-threads 3 --num-train-threads 3 \
--reward-scale 0.001 \
--actor-layers 64-64-64 --actor-layer-norm --actor-parameters-noise --actor-lr 0.001 --actor-lr-end 0.000001 \
--dynamics-lr 0.0001 --dynamics-lr-end 0.0000001 \
--critic-layers 64-32 --critic-layer-norm --critic-lr 0.002 --critic-lr-end 0.000001 \
--initial-epsilon 0.5 --final-epsilon 0.001 --tau 0.0001".split(" ")
os.sys.argv
Out [3]:
['ddpg/train.py',
 '--logdir',
 './outputs/logs_ddpg2',
 '--num-threads',
 '3',
 '--reward-scale',
 '0.001',
 '--actor-layers',
 '64-64-64',
 '--actor-layer-norm',
 '--actor-parameters-noise',
 '--actor-lr',
 '0.001',
 '--actor-lr-end',
 '0.000001',
 '--dynamics-lr',
 '0.0001',
 '--dynamics-lr-end',
 '0.0000001',
 '--critic-layers',
 '64-32',
 '--critic-layer-norm',
 '--critic-lr',
 '0.002',
 '--critic-lr-end',
 '0.000001',
 '--initial-epsilon',
 '0.5',
 '--final-epsilon',
 '0.001',
 '--tau',
 '0.0001']
In [ ]:
from ddpg.train import *
In [ ]:
os.environ['OMP_NUM_THREADS'] = '3'
torch.set_num_threads(1)
args = parse_args()
train(args,
      create_model,
      create_act_update_fns,
      train_multi_thread,
      train_single_thread,
      play_single_thread)
[2018-01-19 10:03:05,421] Making new env: Pendulum-v0
Base (
  (feature_net): LinearNet (
    (net): Sequential (
      (linear_0): NoisyLinear (3 -> 64)
      (layer_norm_0): LayerNorm (
      )
      (act_0): ReLU ()
      (linear_1): NoisyLinear (64 -> 64)
      (layer_norm_1): LayerNorm (
      )
      (act_1): ReLU ()
      (linear_2): NoisyLinear (64 -> 64)
      (layer_norm_2): LayerNorm (
      )
      (act_2): ReLU ()
    )
  )
)
ActorHead (
  (base): Base (
    (feature_net): LinearNet (
      (net): Sequential (
        (linear_0): NoisyLinear (3 -> 64)
        (layer_norm_0): LayerNorm (
        )
        (act_0): ReLU ()
        (linear_1): NoisyLinear (64 -> 64)
        (layer_norm_1): LayerNorm (
        )
        (act_1): ReLU ()
        (linear_2): NoisyLinear (64 -> 64)
        (layer_norm_2): LayerNorm (
        )
        (act_2): ReLU ()
      )
    )
  )
  (policy_net): LinearNet (
    (net): Sequential (
      (linear_0): Linear (64 -> 1)
      (act_0): ELU (alpha=1.0)
    )
  )
)
CriticHead (
  (base): Base (
    (feature_net): LinearNet (
      (net): Sequential (
        (linear_0): NoisyLinear (3 -> 64)
        (layer_norm_0): LayerNorm (
        )
        (act_0): ReLU ()
        (linear_1): NoisyLinear (64 -> 64)
        (layer_norm_1): LayerNorm (
        )
        (act_1): ReLU ()
        (linear_2): NoisyLinear (64 -> 64)
        (layer_norm_2): LayerNorm (
        )
        (act_2): ReLU ()
      )
    )
  )
  (value_net): Linear (64 -> 1)
)
DynamicsHead (
  (base): Base (
    (feature_net): LinearNet (
      (net): Sequential (
        (linear_0): NoisyLinear (3 -> 64)
        (layer_norm_0): LayerNorm (
        )
        (act_0): ReLU ()
        (linear_1): NoisyLinear (64 -> 64)
        (layer_norm_1): LayerNorm (
        )
        (act_1): ReLU ()
        (linear_2): NoisyLinear (64 -> 64)
        (layer_norm_2): LayerNorm (
        )
        (act_2): ReLU ()
      )
    )
  )
  (value_net): LinearNet (
    (net): Sequential (
      (linear_0): Linear (65 -> 64)
      (layer_norm_0): LayerNorm (
      )
      (act_0): ReLU ()
    )
  )
  (value_net2): Linear (64 -> 3)
)
[2018-01-19 10:03:05,570] Making new env: Pendulum-v0
[2018-01-19 10:03:05,589] Making new env: Pendulum-v0
Process Process-1:
Traceback (most recent call last):
  File "/home/isisilon/.pyenv/versions/3.6.0/lib/python3.6/multiprocessing/process.py", line 249, in _bootstrap
    self.run()
  File "/home/isisilon/.pyenv/versions/3.6.0/lib/python3.6/multiprocessing/process.py", line 93, in run
    self._target(*self._args, **self._kwargs)
  File "/media/isisilon/Data/My_Documents/Documents/eclipse-workspace/ddpg_implicit_state/Run-Skeleton-Run/ddpg/model.py", line 440, in train_single_thread
    logger.scalar_summary(key, value, update_step)
  File "/media/isisilon/Data/My_Documents/Documents/eclipse-workspace/ddpg_implicit_state/Run-Skeleton-Run/common/logger.py", line 17, in scalar_summary
    self.writer.add_scalar(tag, value, step)
  File "/home/isisilon/.pyenv/versions/3.6.0/envs/jupyter3/lib/python3.6/site-packages/tensorboardX/writer.py", line 262, in add_scalar
    self.file_writer.add_summary(scalar(tag, scalar_value), global_step)
  File "/home/isisilon/.pyenv/versions/3.6.0/envs/jupyter3/lib/python3.6/site-packages/tensorboardX/summary.py", line 88, in scalar
    assert(scalar.squeeze().ndim==0), 'scalar should be 0D'
AttributeError: 'NoneType' object has no attribute 'squeeze'
[2018-01-19 10:04:01,980] Making new env: Pendulum-v0
[2018-01-19 10:04:05,897] Making new env: Pendulum-v0
[2018-01-19 10:04:56,915] Making new env: Pendulum-v0
[2018-01-19 10:04:59,842] Making new env: Pendulum-v0
[2018-01-19 10:05:51,136] Making new env: Pendulum-v0
[2018-01-19 10:05:54,795] Making new env: Pendulum-v0
[2018-01-19 10:06:58,471] Making new env: Pendulum-v0
[2018-01-19 10:07:01,303] Making new env: Pendulum-v0
[2018-01-19 10:08:06,776] Making new env: Pendulum-v0
[2018-01-19 10:08:11,823] Making new env: Pendulum-v0
[2018-01-19 10:09:24,555] Making new env: Pendulum-v0
[2018-01-19 10:09:28,326] Making new env: Pendulum-v0
[2018-01-19 10:10:25,677] Making new env: Pendulum-v0
[2018-01-19 10:10:28,134] Making new env: Pendulum-v0
[2018-01-19 10:11:21,746] Making new env: Pendulum-v0
[2018-01-19 10:11:22,767] Making new env: Pendulum-v0
[2018-01-19 10:12:19,668] Making new env: Pendulum-v0
[2018-01-19 10:12:20,498] Making new env: Pendulum-v0
[2018-01-19 10:13:16,087] Making new env: Pendulum-v0
[2018-01-19 10:13:19,701] Making new env: Pendulum-v0
[2018-01-19 10:14:20,287] Making new env: Pendulum-v0
[2018-01-19 10:14:25,255] Making new env: Pendulum-v0
[2018-01-19 10:15:25,534] Making new env: Pendulum-v0
[2018-01-19 10:15:30,307] Making new env: Pendulum-v0
[2018-01-19 10:16:24,529] Making new env: Pendulum-v0
[2018-01-19 10:16:27,554] Making new env: Pendulum-v0
[2018-01-19 10:17:23,602] Making new env: Pendulum-v0
[2018-01-19 10:17:24,952] Making new env: Pendulum-v0
[2018-01-19 10:18:19,517] Making new env: Pendulum-v0
[2018-01-19 10:18:21,783] Making new env: Pendulum-v0
[2018-01-19 10:19:17,384] Making new env: Pendulum-v0
[2018-01-19 10:19:19,827] Making new env: Pendulum-v0
[2018-01-19 10:20:15,270] Making new env: Pendulum-v0
[2018-01-19 10:20:15,364] Making new env: Pendulum-v0
[2018-01-19 10:21:09,305] Making new env: Pendulum-v0
[2018-01-19 10:21:12,535] Making new env: Pendulum-v0
[2018-01-19 10:22:03,694] Making new env: Pendulum-v0
[2018-01-19 10:22:05,211] Making new env: Pendulum-v0
[2018-01-19 10:22:59,166] Making new env: Pendulum-v0
[2018-01-19 10:23:03,510] Making new env: Pendulum-v0
[2018-01-19 10:23:52,806] Making new env: Pendulum-v0
[2018-01-19 10:23:58,895] Making new env: Pendulum-v0
[2018-01-19 10:24:44,687] Making new env: Pendulum-v0
[2018-01-19 10:24:52,964] Making new env: Pendulum-v0
[2018-01-19 10:25:37,681] Making new env: Pendulum-v0
[2018-01-19 10:25:51,033] Making new env: Pendulum-v0
[2018-01-19 10:26:34,755] Making new env: Pendulum-v0
[2018-01-19 10:26:45,860] Making new env: Pendulum-v0
[2018-01-19 10:27:27,254] Making new env: Pendulum-v0
[2018-01-19 10:27:44,223] Making new env: Pendulum-v0
[2018-01-19 10:28:25,083] Making new env: Pendulum-v0
[2018-01-19 10:28:40,452] Making new env: Pendulum-v0
[2018-01-19 10:29:21,854] Making new env: Pendulum-v0
[2018-01-19 10:29:40,183] Making new env: Pendulum-v0
[2018-01-19 10:30:20,224] Making new env: Pendulum-v0
[2018-01-19 10:30:43,777] Making new env: Pendulum-v0
[2018-01-19 10:31:21,376] Making new env: Pendulum-v0
[2018-01-19 10:31:43,822] Making new env: Pendulum-v0
[2018-01-19 10:32:19,617] Making new env: Pendulum-v0
[2018-01-19 10:32:42,093] Making new env: Pendulum-v0
[2018-01-19 10:33:13,267] Making new env: Pendulum-v0
[2018-01-19 10:33:38,685] Making new env: Pendulum-v0
[2018-01-19 10:34:08,173] Making new env: Pendulum-v0
[2018-01-19 10:34:36,100] Making new env: Pendulum-v0
[2018-01-19 10:35:03,554] Making new env: Pendulum-v0
[2018-01-19 10:35:31,869] Making new env: Pendulum-v0
[2018-01-19 10:36:02,086] Making new env: Pendulum-v0
[2018-01-19 10:36:30,077] Making new env: Pendulum-v0
[2018-01-19 10:37:04,189] Making new env: Pendulum-v0
[2018-01-19 10:37:33,748] Making new env: Pendulum-v0
[2018-01-19 10:38:03,424] Making new env: Pendulum-v0
[2018-01-19 10:38:35,036] Making new env: Pendulum-v0
In [ ]:
In [ ]:
In [ ]: