Create your own GitHub profile
Sign up for your own profile on GitHub, the best place to host code, manage projects, and build software alongside 50 million developers.
Sign up
Popular repositories
-
-
-
-
Forked from ray-project/ray
A fast and simple framework for building and running distributed applications. Ray is packaged with RLlib, a scalable reinforcement learning library, and Tune, a scalable hyperparameter tuning libr…
Python
-
153 contributions in the last year
Contribution activity
August 2020
- krfricke/transformers Python
- krfricke/tune-sklearn Python
Created a pull request in ray-project/ray that received 4 comments
[tune] PBT replay utility scheduler
Why are these changes needed? This adds a replay utility for Population Based Training parameter schedules. The implementation uses a Scheduler wh…
+390
−26
•
4
comments
- [tune/rllib] made wandb compatible with rllib trainables
- [tune] added FAQ to docs
- [tune] Added PyTorch Lightning callbacks to integrations
- [tune] Added external PyTorch tutorial test
- [tune] Added Kubernetes syncer and sync client
- [tune] fix flaky PBT replay test
- [tune] Add OptunaSearcher wrapper around Optuna samplers
- [tune] Close logfile contexts
- [tune] Ray Tune + Serve end-to-end integration example
- [tune] Prevent leak of magic keys in trial config
- [Tune] Update PBT Transformers Example
- [tune] custom trial directory name
- [Tune] Synchronous Mode for PBT
- [tune] cleanup error messaging/diagnose_serialization helper
- [tune] support rerunning failed trials
- [tune] Update node syncing documentation
- [Tune] Remove `checkpoint_at_end` from tune+serve docs
- [tune] option to raise on error
- [tune] Add OptunaSearcher wrapper around Optuna samplers
- [tune] support resume for search algorithms
- [tune] Add doc builds for Tune
- [tune] Fix restoration for function API PBT
- [tune] Search alg checkpointing during training