Ray VS ara

22. Ray | Github | tutorial

Training times for GSM8k are mentioned here: https://github.com/ray-project/ray/tree/master/doc/source/te...

Here's a reference from RLlib that points to docs and an example, and here's one from one of my projects that includes all my own implementations

RLlib

⦿ AWS Glue Data Quality - Automatic data quality rule recommendations based on your data AWS Glue for Ray - Data integration with Ray (ray.io), a popular new open-source compute framework that helps you scale Python workloads

https://ray.io (just dropping the link)

I've wondered whether it's easier to add data analyst stuff to Elixir that Python seems to have, or add features to Python that Erlang (and by extension Elixir) provides out of the box.

By what I can see, if you want multiprocessing on Python in an easier way (let's say running async), you have to use something like ray core[0], then if you want multiple machines you need redis(?). Elixir/Erlang supports this out of the box.

Explorer[1] is an interesting approach, where it uses Rust via Rustler (Elixir library to call Rust code) and uses Polars as its dataframe library. I think Rustler needs to be reworked for this usecase, as it can be slow to return data. I made initial improvements which drastically improves encoding (https://github.com/elixir-nx/explorer/pull/282 and https://github.com/elixir-nx/explorer/pull/286, tldr 20+ seconds down to 3).

[0] https://github.com/ray-project/ray

Look into AWX as an alternative to Tower. If you just want better reporting on runs, check out ARA or callback plugins.

Also worth considering is ARA for playbook reporting, and then whatever you want for orchestration (Jenkins, Azure Devops, Rundeck, etc).

Why not use ara?

For general recording of playbook activity with a web dashboard, ARA works really well.

If you don’t want any UI / access control then you can also look at ARA - https://ara.recordsansible.org This works as a callback plugin to capture the job run data.

I'm not sure how to interpret running 100 debug messages (https://github.com/ansible-community/ara/blob/master/tests/integration/benchmark_tasks.yaml) into real life performance. Mitogen's Benchmark used either 100 times a "hostname" command on the target machine (https://github.com/mitogen-hq/mitogen/blob/master/tests/ansible/bench/loop-100-items.yml) or running the DebOps project (https://github.com/debops/debops-playbooks/blob/master/playbooks/common.yml) for some sorta real-world module usage.

Reach out on Mastodon or see this issue on GitHub.

There has been discussions about a prometheus exporter for monitoring, metrics and eventually fancy graphs in grafana but work on this has not started yet: https://github.com/ansible-community/ara/issues/177