ray-project / ray

👋

Welcome to the ray community!

A place to connect with other community members about ray

Browse

Answered Unanswered

🙏 Q&A

🙏

[Cluster] Exit signal (intentional) received

🙏 Q&A

🙏

[rllib] Using gpu for inference ( rollout )

🙏 Q&A

🙏

[rllib] Processing a batch of observations by stacking one of the observation dimensions along the batch dimension

💡 Ideas

💡

Adding Slurm to the autoscaler

🙏 Q&A

🙏

Django models are not loaded on Ray workers

🙏 Q&A

🙏

How can I set the x axis variable/time step for logging in functional API?

🙏 Q&A

🙏

What replaced ray.register_custom_serializer() in v1.0?

🙏 Q&A

🙏

[ray] no test for slurm allocation in python/ray/utils.py: get_num_cpus

🙏 Q&A

🙏

How to properly use the `custom_loss` override of a model?

🙏 Q&A

🙏

[rllib] Questions about extending DDPG with online expert; and possible DDPG bug with tf-eager

🙏 Q&A

🙏

Cannot use more than one node on gcp with simple python script

🙏 Q&A

🙏

Expected latency for ray.wait()/ray.get() for completed remotes

🙏 Q&A

🙏

Add another item to `train_batch`

🙏 Q&A

🙏

[tune] Shared File Systems + sync_on_checkpoint Compatibility

🙏 Q&A

🙏

[tune] what is the suggested workflow for ray?

💡 Ideas

💡

Check if tune is active

🙏 Q&A

🙏

[SGD+Serve] Batch train/serve millions of models, one for each example

🙏 Q&A

🙏

[tune] slurm job preemption + restart -> restarting a tune run comprehensively from saved state?

🙏 Q&A

🙏

[ray] GCSServer take a lot of memory

🙏 Q&A

🙏

[rllib] For DDPPO, is there a way to share GPU resources with RolloutWorkers for performing additional tasks?

💡 Ideas

💡

🙏 Q&A

🙏

[ray][tune] __cinit__() takes exactly 1 positional argument

#️⃣ General

#️⃣

Help Installing Ray on NVIDIA Jetson TX2

🙏 Q&A

🙏

[rllib][tune] Each trial of tune with multi-agent sets the same config for all agents?

🙏 Q&A

🙏

Failing to keep worker nodes alive on a PBS cluster since Ray 1.0.0

Beta

You can’t perform that action at this time.