hiro/ray
1
0
Fork 0
mirror of https://github.com/vale981/ray synced 2025-03-06 02:21:39 -05:00
Code Issues Projects Releases Packages Wiki Activity Actions

[air doc] checkpoint_freq --> checkpoint_frequency (#27325)

Browse source 
Signed-off-by: xwjiang2010 <xwjiang2010@gmail.com>
This commit is contained in:
xwjiang2010 2022-08-02 03:34:10 -07:00 • committed by GitHub
parent 149c031c4b
commit 36cf1baa82
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 6 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
doc/source/tune/api_docs/trainable.rst
View file

@ -99,7 +99,7 @@ You can save and load checkpoint in Ray Tune in the following manner:
tuner = tune.Tuner(train_func)
results = tuner.fit()
.. note:: ``checkpoint_freq`` and ``checkpoint_at_end`` will not work with Function API checkpointing.
.. note:: ``checkpoint_frequency`` and ``checkpoint_at_end`` will not work with Function API checkpointing.
In this example, checkpoints will be saved by training iteration to ``local_dir/exp_name/trial_name/checkpoint_<step>``.
@ -177,7 +177,7 @@ You can also implement checkpoint/restore using the Trainable Class API:
checkpoint_path = os.path.join(tmp_checkpoint_dir, "model.pth")
self.model.load_state_dict(torch.load(checkpoint_path))
tuner = tune.Tuner(MyTrainableClass, run_config=air.RunConfig(checkpoint_config=air.CheckpointConfig(checkpoint_freq=2)))
tuner = tune.Tuner(MyTrainableClass, run_config=air.RunConfig(checkpoint_config=air.CheckpointConfig(checkpoint_frequency=2)))
results = tuner.fit()
You can checkpoint with three different mechanisms: manually, periodically, and at termination.
@ -197,7 +197,7 @@ This can be especially helpful in spot instances:
**Periodic Checkpointing**: periodic checkpointing can be used to provide fault-tolerance for experiments.
This can be enabled by setting ``checkpoint_freq=<int>`` and ``max_failures=<int>`` to checkpoint trials
This can be enabled by setting ``checkpoint_frequency=<int>`` and ``max_failures=<int>`` to checkpoint trials
every *N* iterations and recover from up to *M* crashes per trial, e.g.:
.. code-block:: python
@ -205,12 +205,12 @@ every *N* iterations and recover from up to *M* crashes per trial, e.g.:
tuner = tune.Tuner(
my_trainable,
run_config=air.RunConfig(
checkpoint_config=air.CheckpointConfig(checkpoint_freq=10),
checkpoint_config=air.CheckpointConfig(checkpoint_frequency=10),
failure_config=air.FailureConfig(max_failures=5))
)
results = tuner.fit()
**Checkpointing at Termination**: The checkpoint_freq may not coincide with the exact end of an experiment.
**Checkpointing at Termination**: The checkpoint_frequency may not coincide with the exact end of an experiment.
If you want a checkpoint to be created at the end of a trial, you can additionally set the ``checkpoint_at_end=True``:
.. code-block:: python
@ -219,7 +219,7 @@ If you want a checkpoint to be created at the end of a trial, you can additional
tuner = tune.Tuner(
my_trainable,
run_config=air.RunConfig(
checkpoint_config=air.CheckpointConfig(checkpoint_freq=10, checkpoint_at_end=True),
checkpoint_config=air.CheckpointConfig(checkpoint_frequency=10, checkpoint_at_end=True),
failure_config=air.FailureConfig(max_failures=5))
)
results = tuner.fit()