Skip to content

Adding automatic mixed precision training support #113

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: master
Choose a base branch
from
Open

Adding automatic mixed precision training support #113

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
2edb759
add auto mixed precision training support for dqn
vinhngx Jul 24, 2019
1ff6ee2
add auto mixed precision training support for rainbow agent
vinhngx Jul 24, 2019
2ca68bb
add auto mixed precision training support for dqn agent
vinhngx Jul 24, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions dopamine/agents/dqn/dqn_agent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -181,6 +181,8 @@ def __init__(self,
self.eval_mode = eval_mode
self.training_steps = 0
self.optimizer = optimizer
if os.environ.get('TF_ENABLE_AUTO_MIXED_PRECISION', default='0') == '1':
self.optimizer = tf.train.experimental.enable_mixed_precision_graph_rewrite(optimizer)
self.summary_writer = summary_writer
self.summary_writing_frequency = summary_writing_frequency
self.allow_partial_reload = allow_partial_reload
Expand Down
4 changes: 3 additions & 1 deletion dopamine/agents/rainbow/rainbow_agent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,7 +38,7 @@
from __future__ import print_function

import collections

import os


from dopamine.agents.dqn import dqn_agent
Expand Down Expand Up @@ -127,6 +127,8 @@ def __init__(self,
self._replay_scheme = replay_scheme
# TODO(b/110897128): Make agent optimizer attribute private.
self.optimizer = optimizer
if os.environ.get('TF_ENABLE_AUTO_MIXED_PRECISION', default='0') == '1':
self.optimizer = tf.train.experimental.enable_mixed_precision_graph_rewrite(optimizer)

dqn_agent.DQNAgent.__init__(
self,
Expand Down