Run-Skeleton-Run/trying_ddpg_with_implicit_dynamics_debug_1thread.ipynb

In [6]:

import os
os.environ['CUDA_VISIBLE_DEVICES']=""
os.environ["PYTHONPATH"]='.'

In [7]:

%pylab --no-import-all inline
%reload_ext autoreload
%autoreload 2

Populating the interactive namespace from numpy and matplotlib

In [8]:

os.sys.argv="ddpg/train.py --logdir ./outputs/logs_ddpg \
--num-threads 1 \
--reward-scale 1 \
--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 [8]:

['ddpg/train.py',
 '--logdir',
 './outputs/logs_ddpg',
 '--num-threads',
 '1',
 '--reward-scale',
 '1',
 '--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 [9]:

from ddpg.train import *

In [10]:

os.environ['OMP_NUM_THREADS'] = '1'
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 10:39:09,999] 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)
)

In [ ]:

In [ ]:

In [ ]: