epsilon: 0.2
lambd: 0.95
num_epoch: 3
learning_rate_schedule: linear
learning_rate_schedule: constant
network_settings:
normalize: false
hidden_units: 256
# encoding_size: 256
# learning_rate: 0.0003
keep_checkpoints: 5
max_steps: 20000000
max_steps: 50000000
summary_freq: 5000
summary_freq: 20000
threaded: true