Update to Turing v0.35 #559
Merged
+73
−18
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
penelopeysm
force-pushed
the
py/v0.35
branch
from
24862bb
to
ae31cce
Compare
penelopeysm
force-pushed
the
py/v0.35
branch
from
ae31cce
to
02c24ef
Compare
|
Preview the changes: https://turinglang.org/docs/pr-previews/559 |
penelopeysm
commented
Nov 26, 2024
This was referenced Nov 26, 2024
Merged
yebai
approved these changes
Nov 26, 2024
Merged
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Outdated suggestions cannot be applied.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.
I suspect this might yet crash and burn because of the Bijectors stuff, but have to try it.It seems to be fine although we are restricted to Julia 1.10 and Turing 0.35.1 (not 0.35.2), as the latter bumps Bijectors compat to 0.14 which causes a bunch of dependency constraints that I didn't dig into. My suspicion is that all of the upgrades (Julia -> 1.11 and Turing -> 0.35.2) are contingent on TuringLang/Bijectors.jl#341 being merged.
This PR also updates the Bayesian neural network tutorial to work with Mooncake. It didn't work out of the box, I think that this was because of mixed-precision floating point arithmetic. Wrapping the neural network parameters in
Lux.f32(see https://lux.csail.mit.edu/stable/api/Lux/utilities#Updating-Floating-Point-Precision) allowed AD and inference to run locally.I'm fairly confident that the docs should build now.
Closes #521.
Closes #541.