Reduces latency of rendering Jax arrays. #40
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
![copybara-service[bot]](https://avatars.githubusercontent.com/in/44061?s=60&v=4){kind=small}
copybara-service
bot
force-pushed
the
test_680276759
branch
6 times, most recently
from
7757819
to
f47b55b
Compare
Before: Cold rendering array of any size would take ~15-25 seconds (in colab). Now: Cold rendering of small-ish (less than 10M bytes) arrays is instantaneous (<< 1s) Larger arrays (e.g. > 2K x 2K) take 3-4 seconds, which is much more manageable To do this we do two things: a) We convert to numpy for arrays smaller than 10M for computing stats and doing slicing. (We still maintain jax visualization of sharding and types) b) We use a single jitted function for arrays larger than 10m to compute summaries rather than individual jax invocations for each stat, which end up jitted separately. PiperOrigin-RevId: 681055675
copybara-service
bot
force-pushed
the
test_680276759
branch
from
f47b55b
to
438b9a5
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Reduces latency of rendering Jax arrays.
Before:
Cold rendering array of any size would take ~15-25 seconds (in colab).
Now:
Cold rendering of small-ish (less than 10M bytes) arrays is instantaneous (<< 1s)
Larger arrays (e.g. > 2K x 2K) take 3-4 seconds, which is much more manageable
To do this we do two things:
a) We convert to numpy for arrays smaller than 10M for computing stats and doing slicing. (We still maintain jax visualization of sharding and types)
b) We use a single jitted function for arrays larger than 10m to compute summaries rather than individual jax invocations for each stat, which end up jitted separately.