[tune] Storage: 🐙 🧠 Tune tests and examples {using RLlib} migration #38895

krfricke · 2023-08-25T15:02:41Z

Why are these changes needed?

Migrates failing tests of this suite: https://buildkite.com/ray-project/oss-ci-build-pr/builds/33691#018a2b92-326f-4336-9c75-4f358721c816

Related issue number

Checks

I've signed off every commit(by using the -s flag, i.e., git commit -s) in this PR.
I've run scripts/format.sh to lint the changes in this PR.
I've included any doc changes needed for https://docs.ray.io/en/master/.
- I've added any new APIs to the API Reference. For example, if I added a
  method in Tune, I've added it in doc/source/tune/api/ under the
  corresponding .rst file.
I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/
Testing Strategy
- Unit tests
- Release tests
- This PR is not tested :(

Signed-off-by: Kai Fricke <kai@anyscale.com>

krfricke · 2023-08-25T15:04:23Z

python/ray/train/_internal/storage.py

+        if not self.storage_filesystem or isinstance(
+            self.storage_filesystem, pyarrow.fs.LocalFileSystem
+        ):
+            self.storage_local_path = self.storage_path
+        else:
+            self.storage_local_path = default_results_dir
+


@justinvyu lmk what you think. Previously the storage_local_path was always the ray_results directory. But I think we should actually set it to the local path if a local path was set. This also matches previous behavior.

I'm ok with this. @ericl did not want ANY parsing of the storage path (i.e detecting if it's remote or not).

But, because this is AFTER pyrarrow.fs.from_uri, and we're just doing a typecheck of the pyarrow fs, I think this is ok.

The difference is that previously, we did some checks on the raw uri that was being passed by the user.

This will also remove the need for us to either re-introduce some kind of local_dir param or make the environment variable more visible.
storage_path is the 1 way to set the experiment path.

Signed-off-by: Kai Fricke <kai@anyscale.com>

justinvyu

Thanks! One fix, and some other comments

justinvyu · 2023-08-25T17:50:08Z

python/ray/train/_internal/storage.py

+        if not self.storage_filesystem or isinstance(
+            self.storage_filesystem, pyarrow.fs.LocalFileSystem
+        ):
+            self.storage_local_path = self.storage_path
+        else:
+            self.storage_local_path = default_results_dir
+


I'm ok with this. @ericl did not want ANY parsing of the storage path (i.e detecting if it's remote or not).

But, because this is AFTER pyrarrow.fs.from_uri, and we're just doing a typecheck of the pyarrow fs, I think this is ok.

The difference is that previously, we did some checks on the raw uri that was being passed by the user.

This will also remove the need for us to either re-introduce some kind of local_dir param or make the environment variable more visible.
storage_path is the 1 way to set the experiment path.

justinvyu · 2023-08-25T17:52:02Z

python/ray/tune/trainable/session.py

+        # Backwards compatibility
+        from ray import train
+
+        report_dict = kwargs
+        if _metric is not None:
+            report_dict["_metric"] = _metric
+
+        train.report(report_dict)


should we just force users to switch to train.report? otherwise we'll have to force users to do another migration post 2.7

Or, at minimum add a warning here.

I'll add a deprecation warning for now. Let's check if we can hard deprecate in a follow-up

justinvyu · 2023-08-25T17:54:24Z

python/ray/train/_internal/storage.py

+        if not self.storage_filesystem or isinstance(
+            self.storage_filesystem, pyarrow.fs.LocalFileSystem
+        ):
+            self.storage_local_path = self.storage_path
+        else:
+            self.storage_local_path = default_results_dir


Suggested change

if not self.storage_filesystem or isinstance(

self.storage_filesystem, pyarrow.fs.LocalFileSystem

):

self.storage_local_path = self.storage_path

else:

self.storage_local_path = default_results_dir

if isinstance(

self.storage_filesystem, pyarrow.fs.LocalFileSystem

):

self.storage_local_path = self.storage_path

else:

self.storage_local_path = default_results_dir

otherwise this will set local path to a s3 uri in the case of

storage_path="s3://...", storage_filesystem=None

In that case, wouldn't pyarrow resolve the filesystem to s3 filesystem anyway? this is checking self.storage_filesystem after parsing, not storage_filesystem

ericl

[flagging as a major API change]

It's super late in the release process, and we cannot make breaking API changes like this that go against choices previously made in the REP.

Let's revert the behavior changes from this PR and we can discuss them separately.

Signed-off-by: Kai Fricke <kai@anyscale.com>

# Conflicts: # python/ray/tune/BUILD # python/ray/tune/tests/test_api.py # python/ray/tune/tests/test_cluster.py

Signed-off-by: Kai Fricke <kai@anyscale.com>

krfricke · 2023-08-26T14:02:34Z

Ok, reverted the changes for now.

It also looks like test_cluster was fixed in another PR, so this leaves only two test suites in this PR.

Signed-off-by: Kai Fricke <kai@anyscale.com>

matthewdeng · 2023-08-27T15:23:05Z

python/ray/tune/BUILD

@@ -59,7 +59,7 @@ py_test(
    size = "large",
    srcs = ["tests/test_api.py"],
    deps = [":tune_lib"],
-    tags = ["team:ml", "exclusive", "rllib"],
+    tags = ["team:ml", "exclusive", "rllib", "new_storage"],


nit/for future reference - we don't need this if we're enabling the FF (or is this test run in multiple suites?)

These tests are disabled in the old path but enabled in the new path. When the FF is enabled for everything (and not overwritten) we don't need it anymore.

Let's clean those up in one go.

Ack, just to confirm my understanding - these tests are run in other suites in addition to ":octopus: :brain: Tune tests and examples {using RLlib}"?

matthewdeng · 2023-08-27T15:33:57Z

python/ray/tune/tests/test_api.py

+
+        register_trainable("f1", train_fn)
+        with unittest.mock.patch.dict(os.environ, {"TEST_TMPDIR": logdir}):
+            tune.run("f1")


Is this test this supposed to test run_experiments specifically? Or does it need fail_fast="raise" in order for the test failure to be propagated?

It's just testing the logdir location

Oh woops, for some reason I was under the impression that Tune would swallow the error and tune.run would finish "successfully" with an error file.

python/ray/tune/tests/test_api.py

python/ray/tune/tests/test_tune_restore.py

matthewdeng · 2023-08-27T15:43:32Z

python/ray/tune/tests/test_tune_restore.py

@@ -64,7 +67,7 @@ def testTuneRestore(self):
            name="TuneRestoreTest",
            stop={"training_iteration": 2},  # train one more iteration.
            checkpoint_config=CheckpointConfig(checkpoint_frequency=1),
-            restore=self.checkpoint_path,  # Restore the checkpoint
+            restore=self.checkpoint_parent,  # Restore the checkpoint


@justinvyu to confirm that this is the intended behavior - restore takes in the directory instead of the file? Seems logical, but are there backwards compatibility issues with this?

matthewdeng · 2023-08-28T00:05:15Z

@krfricke I'm going to merge in master to re-trigger the tests. (I took off the tests-ok label because I saw a different test that failed, but can add it back after this run)

Signed-off-by: Kai Fricke <kai@anyscale.com>

…ay-project#38895) Signed-off-by: Kai Fricke <kai@anyscale.com> Co-authored-by: matthewdeng <matt@anyscale.com>

@justinvyu

* [train] enable new persistence mode for core and serve tests (#38938) Signed-off-by: Matthew Deng <matt@anyscale.com> * [train] New persistence mode: Update 🐠 `ML Libraries w/ Ray Client Examples (Python 3.7)` (#38923) Signed-off-by: Justin Yu <justinvyu@anyscale.com> * [train] remove non-URI assertion (#38944) Signed-off-by: Matthew Deng <matt@anyscale.com> * [train] New persistence mode: Update 📖 `Doc tests and examples (excluding Ray AIR examples)` (#38940) Signed-off-by: Justin Yu <justinvyu@anyscale.com> Signed-off-by: Matthew Deng <matt@anyscale.com> Co-authored-by: Matthew Deng <matt@anyscale.com> * disable legacy sync config logic in trainable (#38952) Signed-off-by: Justin Yu <justinvyu@anyscale.com> * [2.7 CI][New Persistent Mode][6/n] 📖 ✈️ Ray AIR examples (#38918) Signed-off-by: woshiyyya <xiaoyunxuan1998@gmail.com> * [2.7 CI][New Persistent Mode][2/n] 📺 📖 Doc GPU tests and examples (#38905) Signed-off-by: woshiyyya <xiaoyunxuan1998@gmail.com> * [2.7 CI][New Persistent Mode][4/n] 📺 🚂 Train GPU tests & 🚂 Datasets Train Integration GPU Tests and Examples (#38910) Signed-off-by: woshiyyya <xiaoyunxuan1998@gmail.com> Signed-off-by: Justin Yu <justinvyu@anyscale.com> Co-authored-by: Justin Yu <justinvyu@anyscale.com> * [2.7 CI][New Persistent Mode][1/n] 📺 ✈️ AIR GPU tests (ray/air) & ⚡ :python: Lightning 2.0 Train GPU tests (#38903) Signed-off-by: woshiyyya <xiaoyunxuan1998@gmail.com> Signed-off-by: Yunxuan Xiao <xiaoyunxuan1998@gmail.com> * [train] Fix broken tune tests and support ray storage (#38950) This PR re-introduces support for ray storage ray.init(storage="s3://...") and fixes a broken tune controller test. Signed-off-by: Justin Yu <justinvyu@anyscale.com> * [train] New persistence mode: Finish migrating `xgb`, `lgbm` and `sklearn` trainers, checkpoints + tests (#38959) Signed-off-by: Justin Yu <justinvyu@anyscale.com> * [2.7 CI][New Persistent Mode][5/n] 📖 Doc examples for external code (#38915) Signed-off-by: woshiyyya <xiaoyunxuan1998@gmail.com> * [train][rllib] temporarily disable new persistence mode for rllib tests (#38965) Signed-off-by: Matthew Deng <matt@anyscale.com> * [2.7 CI][New Persistent Mode][8/n] ✈️ AIR tests (ray/air) (#38932) Signed-off-by: woshiyyya <xiaoyunxuan1998@gmail.com> * [tune] Storage: 🐙 🧠 Tune tests and examples {using RLlib} migration (#38895) Signed-off-by: Kai Fricke <kai@anyscale.com> Co-authored-by: matthewdeng <matt@anyscale.com> * [train] Fix MosaicTrainer example and unit test (#38970) Signed-off-by: Justin Yu <justinvyu@anyscale.com> * [air/release] Fix dreambooth example image preprocessing logic (#39020) Signed-off-by: Justin Yu <justinvyu@anyscale.com> * [train] clean up ray.train._checkpoint imports (#38951) Signed-off-by: Matthew Deng <matt@anyscale.com> * [train] high level cleanup of Ray Train docs (#38971) Signed-off-by: Matthew Deng <matt@anyscale.com> * [wip][docs] update FrameworkPredictor examples (#38634) Signed-off-by: Matthew Deng <matt@anyscale.com> Signed-off-by: matthewdeng <matt@anyscale.com> * [train] Add documentation for using metadata argument to save preprocessors (#38701) * [Train] Restructure Ray Train Example Page (#38814) Signed-off-by: woshiyyya <xiaoyunxuan1998@gmail.com> * [air] Deprecate some fields/classes that are supposed to be gone in 2.6. (#38794) Signed-off-by: xwjiang2010 <xwjiang2010@gmail.com> * [tune/storage] Fix Tune multinode tests (#39050) Fixes multinode tests by using the new train.report() API. Signed-off-by: Kai Fricke <kai@anyscale.com> * [tune] Fix BOHB example for new storage (#38983) The new storage path does not create "empty" checkpoints per default anymore. Previously, when no checkpoint is saved, PAUSEing a trial would create a dummy checkpoint that only contains trial metadata (such as the iteration number). This is not the case anymore. Examples now have to implement checkpointing to properly restore previous state. This was also true previously - but some of our simple examples (e.g. the one in this PR) didn't implement it and still "worked". I think it's fine to keep the functionality as is and require our examples to show checkpointing implementations. This will ensure that users don't shoot their feet trying to use e.g. BOHB. Separately, BOHB was malfunctioning as trials were repeatedly PAUSED and restarted as they've never been removed from `bracket.trials_to_unpause`. @justinvyu mentioned this in the review where it was introduced and I believed at the time it wasn't necessary - turns out it is, as we can end up in a situation where a bracket is never finished because trials are constantly running. This was not caught by any tests. We should add one in a follow-up - for now we can proceed with this PR to pick onto Ray 2.7. Signed-off-by: Kai Fricke <kai@anyscale.com> * [Release Test] Fix `long_running_horovod_tune_test`. (#39012) Signed-off-by: Yunxuan Xiao <yunxuanx@anyscale.com> Signed-off-by: Yunxuan Xiao <xiaoyunxuan1998@gmail.com> * [train] New persistence mode: `StorageContext` unit tests (#39023) Signed-off-by: Justin Yu <justinvyu@anyscale.com> * [train] enable train + tune tests and examples (#39021) Signed-off-by: Matthew Deng <matt@anyscale.com> * [rllib] Fix storage-path related tests (#38947) This PR fixes rllib-related tests that didn't pass changes related to the new storage context. Signed-off-by: Kai Fricke <kai@anyscale.com> Signed-off-by: matthewdeng <matt@anyscale.com> Co-authored-by: matthewdeng <matt@anyscale.com> * [train] New persistence mode: Migrate 🐙 `Tune tests and examples (medium)` (#39081) Signed-off-by: Justin Yu <justinvyu@anyscale.com> --------- Signed-off-by: Matthew Deng <matt@anyscale.com> Signed-off-by: Justin Yu <justinvyu@anyscale.com> Signed-off-by: woshiyyya <xiaoyunxuan1998@gmail.com> Signed-off-by: Yunxuan Xiao <xiaoyunxuan1998@gmail.com> Signed-off-by: Kai Fricke <kai@anyscale.com> Signed-off-by: matthewdeng <matt@anyscale.com> Signed-off-by: xwjiang2010 <xwjiang2010@gmail.com> Signed-off-by: Yunxuan Xiao <yunxuanx@anyscale.com> Co-authored-by: Justin Yu <justinvyu@anyscale.com> Co-authored-by: Yunxuan Xiao <yunxuanx@anyscale.com> Co-authored-by: Kai Fricke <krfricke@users.noreply.github.com> Co-authored-by: Eric Liang <ekhliang@gmail.com> Co-authored-by: xwjiang2010 <87673679+xwjiang2010@users.noreply.github.com>

…ay-project#38895) Signed-off-by: Kai Fricke <kai@anyscale.com> Co-authored-by: matthewdeng <matt@anyscale.com> Signed-off-by: e428265 <arvind.chandramouli@lmco.com>

…ay-project#38895) Signed-off-by: Kai Fricke <kai@anyscale.com> Co-authored-by: matthewdeng <matt@anyscale.com>

…ay-project#38895) Signed-off-by: Kai Fricke <kai@anyscale.com> Co-authored-by: matthewdeng <matt@anyscale.com> Signed-off-by: Jim Thompson <jimthompson5802@gmail.com>

…ay-project#38895) Signed-off-by: Kai Fricke <kai@anyscale.com> Co-authored-by: matthewdeng <matt@anyscale.com> Signed-off-by: Victor <vctr.y.m@example.com>

Kai Fricke added 8 commits August 25, 2023 13:59

test_api, Durable

c12b5ef

Signed-off-by: Kai Fricke <kai@anyscale.com>

test_api

c050d27

Signed-off-by: Kai Fricke <kai@anyscale.com>

test_tune_restore

c19a072

Signed-off-by: Kai Fricke <kai@anyscale.com>

test_tune_restore

ffef986

Signed-off-by: Kai Fricke <kai@anyscale.com>

test_api

5cbb904

Signed-off-by: Kai Fricke <kai@anyscale.com>

test_tune_restore

b95b507

Signed-off-by: Kai Fricke <kai@anyscale.com>

test_cluster

f02926d

Signed-off-by: Kai Fricke <kai@anyscale.com>

Revert flag

6a0eb0d

Signed-off-by: Kai Fricke <kai@anyscale.com>

krfricke requested a review from justinvyu August 25, 2023 15:03

krfricke assigned justinvyu Aug 25, 2023

krfricke commented Aug 25, 2023

View reviewed changes

tests

ceef228

Signed-off-by: Kai Fricke <kai@anyscale.com>

justinvyu reviewed Aug 25, 2023

View reviewed changes

justinvyu requested a review from ericl August 25, 2023 17:56

ericl requested changes Aug 25, 2023

View reviewed changes

ericl added the @author-action-required The PR author is responsible for the next step. Remove tag to send back to the reviewer. label Aug 25, 2023

Kai Fricke added 5 commits August 25, 2023 20:14

deprecate tune.report

c9a7859

Signed-off-by: Kai Fricke <kai@anyscale.com>

Merge remote-tracking branch 'upstream/master' into train/test-api

f2211dc

# Conflicts: # python/ray/tune/BUILD # python/ray/tune/tests/test_api.py # python/ray/tune/tests/test_cluster.py

Revert storage changes

f4942e5

Signed-off-by: Kai Fricke <kai@anyscale.com>

revert session

d279f0f

Signed-off-by: Kai Fricke <kai@anyscale.com>

revert build

769b8b6

Signed-off-by: Kai Fricke <kai@anyscale.com>

krfricke removed the @author-action-required The PR author is responsible for the next step. Remove tag to send back to the reviewer. label Aug 26, 2023

krfricke requested review from ericl and justinvyu August 26, 2023 14:16

Kai Fricke added 5 commits August 26, 2023 20:55

Merge remote-tracking branch 'upstream/master' into train/test-api

874f8d5

flag to 1

57f61ec

Signed-off-by: Kai Fricke <kai@anyscale.com>

fix

d1dad8a

Signed-off-by: Kai Fricke <kai@anyscale.com>

Merge remote-tracking branch 'upstream/master' into train/test-api

89edda2

no local mode

f639439

Signed-off-by: Kai Fricke <kai@anyscale.com>

Remove syncer test

4cc6f1b

Signed-off-by: Kai Fricke <kai@anyscale.com>

krfricke changed the title ~~[tune] Storage: Tune tests + examples (incl Rllib) migration~~ [tune] Storage: 🐙 🧠 Tune tests and examples {using RLlib} migration Aug 27, 2023

krfricke requested a review from matthewdeng August 27, 2023 15:10

krfricke assigned ericl, matthewdeng and zhe-thoughts Aug 27, 2023

krfricke added tests-ok The tagger certifies test failures are unrelated and assumes personal liability. Ray-2.7 labels Aug 27, 2023

matthewdeng reviewed Aug 27, 2023

View reviewed changes

matthewdeng added v2.7.0-pick and removed tests-ok The tagger certifies test failures are unrelated and assumes personal liability. labels Aug 27, 2023

Merge branch 'master' into train/test-api

1686397

matthewdeng mentioned this pull request Aug 28, 2023

[train] clean up ray.train._checkpoint imports #38951

Merged

8 tasks

matthewdeng added the tests-ok The tagger certifies test failures are unrelated and assumes personal liability. label Aug 28, 2023

zhe-thoughts removed their assignment Aug 28, 2023

Kai Fricke added 2 commits August 28, 2023 17:26

review

101998a

Signed-off-by: Kai Fricke <kai@anyscale.com>

Merge remote-tracking branch 'origin/train/test-api' into train/test-api

2cd7292

krfricke added tests-ok The tagger certifies test failures are unrelated and assumes personal liability. and removed tests-ok The tagger certifies test failures are unrelated and assumes personal liability. labels Aug 28, 2023

matthewdeng approved these changes Aug 28, 2023

View reviewed changes

ericl approved these changes Aug 28, 2023

View reviewed changes

matthewdeng merged commit ed7186a into ray-project:master Aug 28, 2023

krfricke deleted the train/test-api branch August 29, 2023 07:59

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[tune] Storage: 🐙 🧠 Tune tests and examples {using RLlib} migration #38895

[tune] Storage: 🐙 🧠 Tune tests and examples {using RLlib} migration #38895

krfricke commented Aug 25, 2023 •

edited

Loading

krfricke Aug 25, 2023

justinvyu Aug 25, 2023 •

edited

Loading

justinvyu left a comment

justinvyu Aug 25, 2023 •

edited

Loading

justinvyu Aug 25, 2023

justinvyu Aug 25, 2023

krfricke Aug 25, 2023

justinvyu Aug 25, 2023

justinvyu Aug 25, 2023

krfricke Aug 25, 2023

ericl left a comment

krfricke commented Aug 26, 2023

matthewdeng Aug 27, 2023

krfricke Aug 27, 2023

matthewdeng Aug 27, 2023

matthewdeng Aug 27, 2023

krfricke Aug 27, 2023

matthewdeng Aug 27, 2023

matthewdeng Aug 27, 2023

matthewdeng commented Aug 28, 2023 •

edited

Loading

[tune] Storage: 🐙 🧠 Tune tests and examples {using RLlib} migration #38895

[tune] Storage: 🐙 🧠 Tune tests and examples {using RLlib} migration #38895

Conversation

krfricke commented Aug 25, 2023 • edited Loading

Why are these changes needed?

Related issue number

Checks

Choose a reason for hiding this comment

justinvyu Aug 25, 2023 • edited Loading

Choose a reason for hiding this comment

justinvyu left a comment

Choose a reason for hiding this comment

justinvyu Aug 25, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ericl left a comment

Choose a reason for hiding this comment

krfricke commented Aug 26, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

matthewdeng commented Aug 28, 2023 • edited Loading

krfricke commented Aug 25, 2023 •

edited

Loading

justinvyu Aug 25, 2023 •

edited

Loading

justinvyu Aug 25, 2023 •

edited

Loading

matthewdeng commented Aug 28, 2023 •

edited

Loading