Skip to content

Fix: ModelBuilder deployment & optimization of JumpStart llama-3.1 models #4937

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 11 commits into from
Nov 22, 2024
Merged

Fix: ModelBuilder deployment & optimization of JumpStart llama-3.1 models #4937

merged 11 commits into from
Nov 22, 2024

Conversation

cj-zhang
Copy link
Collaborator

@cj-zhang cj-zhang commented Nov 21, 2024 •
edited
Loading

Description of changes:

  • Allow deploying JumpStart models w/gated draft models through ModelBuilder
  • Fixing regression on Fast Model Loading on GoldFinch
  • Fix compilation and quantization of Llama 3.1 models

Testing done: E2E notebook testing, UTs and integ tests.

Merge Checklist

Put an x in the boxes that apply. You can also fill these out after creating the PR. If you're unsure about any of them, don't hesitate to ask. We're here to help! This is simply a reminder of what we are going to look for before merging your pull request.

General

  • I have read the CONTRIBUTING doc
  • I certify that the changes I am introducing will be backward compatible, and I have discussed concerns about this, if any, with the Python SDK team
  • I used the commit message format described in CONTRIBUTING
  • I have passed the region in to all S3 and STS clients that I've initialized as part of this change.
  • I have updated any necessary documentation, including READMEs and API docs (if appropriate)

Tests

  • I have added tests that prove my fix is effective or that my feature works (if appropriate)
  • I have added unit and/or integration tests as appropriate to ensure backward compatibility of the changes
  • I have checked that my tests are not configured for a specific region or account (if appropriate)
  • I have used unique_name_from_base to create resource names in integ tests (if appropriate)
  • If adding any dependency in requirements.txt files, I have spell checked and ensured they exist in PyPi

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.

@cj-zhang cj-zhang requested a review from a team as a code owner November 21, 2024 00:20
@cj-zhang cj-zhang requested a review from nargokul November 21, 2024 00:20
gwang111
gwang111 reviewed Nov 21, 2024
View reviewed changes
gwang111
gwang111 previously approved these changes Nov 21, 2024
View reviewed changes
@cj-zhang cj-zhang changed the title Emit warning when cpu cores are requested w/sharded model deployment. Fix: ModelBuilder deployment & optimization of JumpStart llama-3.1 models Nov 22, 2024
Lokiiiiii
Lokiiiiii suggested changes Nov 22, 2024
View reviewed changes
src/sagemaker/jumpstart/utils.py
mutable_model_data_source.pop(
"HostingEulaKey"
) # pop when model access config is not applicable
if "HostingEulaKey" in mutable_model_data_source:
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we add a test for this in test_model_builder as well in addition to what you already have ? This is a very critical path in ModelBuilder

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

tests/integ/sagemaker/serve/test_serve_js_deep_unit_tests.py under test_js_model_with_optimize_speculative_decoding_config_gated_requests_are_expected

@gwang111
Copy link
Collaborator

gwang111 commented Nov 22, 2024 •
edited
Loading 

flake8 installed: flake8==4.0.1,flake8-future-import==0.4.6,mccabe==0.6.1,pycodestyle==2.8.0,pyflakes==2.4.0
flake8 run-test-pre: PYTHONHASHSEED='42'
flake8 run-test: commands[0] | flake8
_____________________________________________________________________________________________________________________ summary _____________________________________________________________________________________________________________________
  flake8: commands succeeded
  congratulations :)

Flake8 is passing on my end with the same versions etc.

@mufaddal-rohawala
Copy link
Member

RTD build has passed, not sure why it's not reflecting baack in PR check, overriding for now.

@mufaddal-rohawala mufaddal-rohawala merged commit 801db44 into aws:master Nov 22, 2024
13 of 14 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@gwang111 gwang111 gwang111 left review comments

@mufaddal-rohawala mufaddal-rohawala mufaddal-rohawala approved these changes

@nargokul nargokul Awaiting requested review from nargokul nargokul is a code owner automatically assigned from aws/sagemaker-ml-frameworks

+1 more reviewer

@Lokiiiiii Lokiiiiii Lokiiiiii requested changes

Reviewers whose approvals may not affect merge requirements
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
4 participants
@cj-zhang @gwang111 @mufaddal-rohawala @Lokiiiiii