8000 Add support for other ways of training than using tf.Estimator · Issue #4 · merantix/imitation-learning · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

Add support for other ways of training than using tf.Estimator #4

New issue

Have a question about this project? Sig 8F8B n 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

Jump to bottom
Open
roberttorfason opened this issue Dec 12, 2018 · 0 comments
Open

Add support for other ways of training than using tf.Estimator #4

roberttorfason opened this issue Dec 12, 2018 · 0 comments
Labels
enhancement New feature or request

Comments

@roberttorfason
Copy link

Currently training only works using the tf.Estimator framework. Some users might prefer using the standard sess.run way of calling the training operation for a more low level way of doing training. A starting point for that version might look like this

def main():
    features, labels = input_fn.train_input_fn(tfrecord_path, batch_size=bs, shuffle_buffer_size=sbs)()
    model = trainer.model_fn(features, labels, tf.estimator.ModeKeys.TRAIN)
    train_op = model.train_op

    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for i in range(num_epochs):
            sess.run(train_op)
@markus-hinsche markus-hinsche added the enhancement New feature or request label Jan 26, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
enhancement New feature or request
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@markus-hinsche @roberttorfason
0