Iris roadmap

These docs are outdated! Please check out https://docs.titanml.co for the latest information on the TitanML platform. If there's anything that's not covered there, please contact us on our discord.

We are always working on ways to improve the Iris platform. Here is a rough list of things that we are working on completing in the near future. This list is neither in chronological nor importance order.

If there is something you think we should be focused on send us a message by email at [e-mail] or join the Discord.

Takeoff

• Reduced conversion memory overhead

• 4bit support

• Accelerated cpu builds with math acceleration libraries (MKL, oneDNN, etc)

• Improved large batch performance with vLLM

• Improved decoding methods

• Chat interface front-end

Train / Optimise / Front-end

• Experiment cards for easier 'getting started'

• Streaming logs

• Improved dataset metadata visble in the front end

• Better ETAs for competition

• Post jobs without the CLI