from upstream #7
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This PR adds a `Tuner.restore(param_space=...)` argument. This allows object refs to be updated if used in the original run. This is a follow-up to #31927 Signed-off-by: Justin Yu <[email protected]>
…#32518) * RLTrainer -> Learner * TrainerRunner -> LearnerGroup Signed-off-by: Kourosh Hakhamaneshi <[email protected]>
Previously, `get_preprocessor` would always serialize the Checkpoint into a dictionary first. This is incredibly wasteful and causes huge memory usage and runtime with large directory-based Checkpoints. This PR changes the logic to first see if a directory Checkpoint should be loaded into a dictionary or not in order to obtain the preprocessor. Context: I had ran into it when trying to do predictions with 25 GB Hugging Face model. `HuggingFacePredictor` calls `get_preprocessor` internally, and that takes ages to complete and almost caused an OOM for me - and all of that is unnecessary as the preprocessor has to be loaded from a file anyway. Signed-off-by: Antoni Baum <[email protected]>
This PR improves Train lazy checkpointing with NFS setups. Previously, the logic to determine whether lazy checkpointing should be used was dependent on whether the Train worker-actor was on the same node as the Trainable actor. The new logic instead has the Trainable actor drop a marker file in the Trial's directory. If a worker-actor can detect that file, it means it can access the same directory as the Trainable actor. This PR also fixes lazy checkpointing env var propagation. Signed-off-by: Antoni Baum <[email protected]>
…_dir` w/ endpoint and params (#32479) Currently, URI handling with parameters is done in multiple places in different ways (using `urllib.parse` or splitting by `'?'` directly). In some places, it's not done at all, which **causes errors when performing cloud checkpointing.** In particular, `Trial.remote_checkpoint_dir` and `Trainable._storage_path` do not handle URI path appends correctly when URL params are present. Signed-off-by: Justin Yu <[email protected]>
Signed-off-by: SangBin Cho <[email protected]> Remove the tool sub-directory and put all the tools to the top level. Move ray dashboard to the top item from monitoring & debugging. Add a link to the overview page for all getting started guide. Add a observability section to the top-level getting started guide. Remove verbose dashboard overview and add a picture instead. Note: I will make an another PR to improve the overview page of the dashboard.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Why are these changes needed?
Related issue number
Checks
git commit -s) in this PR.scripts/format.shto lint the changes in this PR.