def compute_gradient(self, params):
self.policy.set_weights(params)
rollout = self.pull_batch_from_queue()
batch = process_rollout(rollout, gamma=0.99, lambda_=1.0)
gradient = self.policy.get_gradients(batch)
info = {"id": self.id, "size": len(batch.a)}
return gradient, info
def compute_gradient(self, params):
self.policy.set_weights(params)
rollout = self.pull_batch_from_queue()
batch = process_rollout(rollout, gamma=0.99, lambda_=1.0)
gradient, info = self.policy.get_gradients(batch)
if "summary" in info:
self.summary_writer.add_summary(
tf.Summary.FromString(info['summary']),
self.policy.local_steps)
self.summary_writer.flush()
info = {"id": self.id, "size": len(batch.a)}
return gradient, info