feat: setup project #1
Conversation
| from .datasource import LanceDatasource | ||
|
|
||
|
|
||
| def read_lance( |
There was a problem hiding this comment.
Do we still need this if we have upstream ray.data.read_lance?
There was a problem hiding this comment.
I think we can have both? the ray.data.read_lance will just call this.
There was a problem hiding this comment.
What I'm quite curious about is what features will be mainly developed here in the future based on Ray?
There was a problem hiding this comment.
What I'm quite curious about is what features will be mainly developed here in the future based on Ray?
The PR submission process in the official Ray repo is too long, for example, @Jay-ju several bug fix PRs have not been merged for nearly two months 😁. We need faster iterations. This repo allows for faster iterations. Additionally, we have implemented distributed index building, distributed vector queries, and caching capabilities, which can also be gradually opened up in this repo.
There was a problem hiding this comment.
Do we still need this if we have upstream
ray.data.read_lance?
We may be able to use the official ray.data.read_lance as an alias for lance_ray.read_lance.
|
I added a few tasks in issues, let's get this in and iterate on it! |
4 participants
No description provided.