Improve `setattr` performance of Pydantic models by caching setter functions #10868

MarkusSintonen · 2024-11-18T13:06:01Z

Change Summary

Attribute setting has been pretty slow for BaseModel due to the extensive checks it has been doing for every __setattr__ call. PR improves performance of __setattr__ by memoizing the attribute specific handlers to the model class. This makes the attribute assigning some 7x faster. Also add missing benchmarks for attribute usage.

from timeit import timeit
from pydantic import BaseModel

class Model(BaseModel):
    field: int

model = Model(field=1)

def run():
    model.field = 2

# Before 1.048
# After 0.147
print(timeit(run, number=1000000))

Related issue number

fix #10853

Checklist

The pull request title is a good summary of the changes - it will be used in the changelog
Unit tests for the changes exist
Tests pass on CI
Documentation reflects the changes where applicable
My PR is ready to review, please add a comment including the phrase "please review" to assign reviewers

Selected Reviewer: @sydney-runkle

codspeed-hq · 2024-11-18T13:11:44Z

CodSpeed Performance Report

Merging #10868 will not alter performance

_{Comparing MarkusSintonen:fast-setattr (404b8b7) with main (30ee4f4)}

Summary

✅ 44 untouched benchmarks

🆕 2 new benchmarks

Benchmarks breakdown

	Benchmark	`main`	`MarkusSintonen:fast-setattr`	Change
🆕	`test_getattr`	N/A	54 µs	N/A
🆕	`test_setattr`	N/A	87.7 µs	N/A

github-actions · 2024-11-18T13:13:24Z

Coverage report

Click to see where and how coverage changed

File	Statements	Missing	Coverage	Coverage (new stmts)	Lines missing
pydantic
main.py
pydantic/_internal
_model_construction.py
Project Total

_{This report was generated by python-coverage-comment-action}

MarkusSintonen · 2024-11-18T15:35:42Z

please review

sydney-runkle

Hi @MarkusSintonen,

Cool idea, thanks! I think memoization could be helpful here. Let me circle back with some colleagues to verify.

Specifically, @dmontagu, wdyt about this? I recall you've done a lot of work on these setattr branches.

pydantic/main.py

dmontagu

This is great!

Smart idea to memoize most of the checks that are really only dependent on the class and attribute name. I made a couple notes and generally defer to Sydney on specific style nits, but the overall approach seems sensible to me and a good improvement.

Consider the PR approved by me, at least conceptually; I'm just not explicitly approving due to the nit comments maybe meriting some minor changes before merging.

sydney-runkle · 2024-11-18T20:02:56Z

Consider the PR approved by me, at least conceptually

Fantastic, thanks for the prompt review.

I've give this a nit-picky review this evening, then we can move forward!

This makes me think more about what else we could potentially memoize in the schema gen department...

One other thing I want to make sure of - this doesn't leave us with any pickling issues? I don't think so, given passing tests, but we should check.

Viicos

Thanks, I think this is a smart idea. We might have to worry about the size of the cache for large models (with a lot of fields). If we encounter such issues, we could use proper functions defined once instead of creating lambdas everytime

pydantic/main.py

MarkusSintonen · 2024-11-19T06:20:53Z

Thanks, I think this is a smart idea. We might have to worry about the size of the cache for large models (with a lot of fields). If we encounter such issues, we could use proper functions defined once instead of creating lambdas everytime

I wouldnt worry about size of it as anyways all the fields are listed in various ClassVars. However if we want to remove the tiny overhead of field name strs we could push the handler fn into eg FieldInfo/ModelPrivateAttr.

functions defined once

I purposely didnt want to touch the model generation side to not make it anymore heavier than it already is. Because of the mentioned large models it could just do work for no good reason in case fields are not even used like this.

Viicos · 2024-11-19T09:49:42Z

Not sure exactly what you mean by field name strs / didnt want to touch the model generation side, but what I wanted to say is we could do something like this:

HANDLERS = {
    'descriptor': lambda m, val: attribute.__set__(m, val),
    'cached_property': lambda m, val: m.__dict__.__setitem__(name, val),
    ...
}

def _setattr_handler(name: str, value: Any):
    ...
    if hasattr(attribute, '__set__'):
        return HANDLERS['descriptor']
    ...
    elif isinstance(attr, cached_property):
        return HANDLERS['cached_property']

So that we don't create a new function every time.

MarkusSintonen · 2024-11-19T10:06:08Z

but what I wanted to say is we could do something like this

Ah yes I see! That would make sense yes

MarkusSintonen · 2024-11-19T10:54:22Z

@Viicos the suggestion now done here 7dfbe50

Viicos

Thanks for adding the simple dict. This looks good to me.

Overall this approach can look a bit weird as calling __setattr__ (i.e. at the instance level) mutates a class variable that will be valid for every instance. But functionally it makes sense.

pydantic/main.py

sydney-runkle

Looking good! A few more questions / comments.

pydantic/main.py

sydney-runkle

Great work, thank you!

fast setattr

3a285fd

github-actions bot added the relnotes-fix Used for bugfixes. label Nov 18, 2024

MarkusSintonen mentioned this pull request Nov 18, 2024

Slow setting an value of a basic Model #10853

Closed

1 task

pydantic-hooky bot added the ready for review label Nov 18, 2024

pydantic-hooky bot assigned sydney-runkle Nov 18, 2024

sydney-runkle reviewed Nov 18, 2024

View reviewed changes

pydantic/main.py Outdated Show resolved Hide resolved

pydantic/main.py Outdated Show resolved Hide resolved

sydney-runkle added relnotes-performance Used for performance improvements. and removed relnotes-fix Used for bugfixes. labels Nov 18, 2024

dmontagu reviewed Nov 18, 2024

View reviewed changes

pydantic/main.py Outdated Show resolved Hide resolved

dmontagu reviewed Nov 18, 2024

View reviewed changes

Code review fixes.

6f0ab2d

MarkusSintonen force-pushed the fast-setattr branch from d9d4272 to 6f0ab2d Compare November 18, 2024 20:37

Viicos reviewed Nov 18, 2024

View reviewed changes

pydantic/main.py Outdated Show resolved Hide resolved

pydantic/main.py Outdated Show resolved Hide resolved

Add more tests, fix priv field corner case, code review fixes

668d4dd

Viicos changed the title ~~Fix slow BaseModel.__setattr__~~ Improve __setattr__ performance of Pydantic models by caching setter functions Nov 19, 2024

Predefine simple attr setters without captured closure.

7dfbe50

MarkusSintonen force-pushed the fast-setattr branch from e8afcf3 to 7dfbe50 Compare November 19, 2024 10:46

Viicos approved these changes Nov 19, 2024

View reviewed changes

pydantic/main.py Outdated Show resolved Hide resolved

pydantic/main.py Outdated Show resolved Hide resolved

sydney-runkle reviewed Nov 19, 2024

View reviewed changes

Code review changes

404b8b7

sydney-runkle approved these changes Nov 19, 2024

View reviewed changes

sydney-runkle merged commit addf1f9 into pydantic:main Nov 19, 2024
53 checks passed

Emrys-Merlin mentioned this pull request Apr 10, 2025

Setting validate_assignment=False after model initialization and invoking a setter does not disable assignment validation #11729

Closed

1 task

Viicos mentioned this pull request Apr 15, 2025

Strange runtime performance on model instantiation, copying and attribute access #8711

Open

1 task

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

Improve `setattr` performance of Pydantic models by caching setter functions #10868

Improve `setattr` performance of Pydantic models by caching setter functions #10868

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Improve __setattr__ performance of Pydantic models by caching setter functions #10868

Improve __setattr__ performance of Pydantic models by caching setter functions #10868

Uh oh!

Conversation

Uh oh!

Change Summary

Related issue number

Checklist

Uh oh!

Uh oh!

CodSpeed Performance Report

Merging #10868 will not alter performance

Summary

Benchmarks breakdown

Uh oh!

Uh oh!

Coverage report

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Improve `setattr` performance of Pydantic models by caching setter functions #10868

Improve `setattr` performance of Pydantic models by caching setter functions #10868