Skip to content

Conversation

@wangrunji0408
Copy link
Collaborator

This PR introduces a streaming option to the map_batches function. When streaming=true, the user function will receive an iterator and return an iterator. This enables users to customize the streaming logic for data processing, such as parallelizing CPU and GPU operations to improve utilization of both resources. However, this approach is still not straightforward enough for users. In the future, we plan to allow users to pass multiple functions, which will automatically be executed in parallel across different threads in a pipelined manner.

Signed-off-by: Runji Wang <runji@deepseek.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants