Skip to content
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.

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

fix(botocore): accept AWS ARN bedrock model IDs [backport 2.18] #11972

Merged
merged 2 commits into from
Jan 15, 2025

Conversation

github-actions[bot]
Copy link
Contributor

@github-actions github-actions bot commented Jan 15, 2025

Backport e8d68ae from #11944 to 2.18.

This PR fixes the model ID parsing in the bedrock integration to additionally accept model IDs in AWS ARN format.

Previously bedrock only worked with base/foundation models which were of the format {model_provider}.{model_name}. However now bedrock has expanded to allow custom/provisioned/imported/prompt/inference/sagemaker models, following the typical AWS ARN format:

  Base model: "{model_provider}.{model_name}"
  Cross-region model: "{region}.{model_provider}.{model_name}"
  AWS ARNs: Prefixed by "arn:aws{+region?}:bedrock:{region}:{account-id}:"
        a. Foundation model: ARN prefix + "foundation-model/{region?}.{model_provider}.{model_name}"
        b. Custom model: ARN prefix + "custom-model/{model_provider}.{model_name}"
        c. Provisioned model: ARN prefix + "provisioned-model/{model-id}"
        d. Imported model: ARN prefix + "imported-module/{model-id}"
        e. Prompt management: ARN prefix + "prompt/{prompt-id}"
        f. Sagemaker: ARN prefix + "endpoint/{model-id}"
        g. Inference profile: ARN prefix + "{application-?}inference-profile/{model-id}"
        h. Default prompt router: ARN prefix + "default-prompt-router/{prompt-id}"

Currently if an AWS ARN gets submitted as the model ID (for example, prompt management or a provisioned model) then because we attempt to split the string into a 2/3-part period-separated list, our code breaks when there is no period delimitter in the string.

This PR makes a best effort attempt to check each case and extract a model provider and name from the model ID. However, due to the nature of some formats, we default model provider to "custom" and just return the model/prompt/resource ID from the rest of the AWS ARN.

Checklist

  • PR author has checked that all the criteria below are met
  • The PR description includes an overview of the change
  • The PR description articulates the motivation for the change
  • The change includes tests OR the PR description describes a testing strategy
  • The PR description notes risks associated with the change, if any
  • Newly-added code is easy to change
  • The change follows the library release note guidelines
  • The change includes or references documentation updates if necessary
  • Backport labels are set (if applicable)

Reviewer Checklist

  • Reviewer has checked that all the criteria below are met
  • Title is accurate
  • All changes are related to the pull request's stated goal
  • Avoids breaking API changes
  • Testing strategy adequately addresses listed risks
  • Newly-added code is easy to change
  • Release note makes sense to a user of the library
  • If necessary, author has acknowledged and discussed the performance implications of this PR as reported in the benchmarks PR comment
  • Backport labels are set in a manner that is consistent with the release branch maintenance policy

This PR fixes the model ID parsing in the bedrock integration to
additionally accept model IDs in AWS ARN format.

Previously bedrock only worked with base/foundation models which were of
the format `{model_provider}.{model_name}`. However now
[bedrock](https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_InvokeModel.html#API_runtime_InvokeModel_RequestSyntax)
has expanded to allow
custom/provisioned/imported/prompt/inference/sagemaker models, following
the typical AWS ARN format:

```
  Base model: "{model_provider}.{model_name}"
  Cross-region model: "{region}.{model_provider}.{model_name}"
  AWS ARNs: Prefixed by "arn:aws{+region?}:bedrock:{region}:{account-id}:"
        a. Foundation model: ARN prefix + "foundation-model/{region?}.{model_provider}.{model_name}"
        b. Custom model: ARN prefix + "custom-model/{model_provider}.{model_name}"
        c. Provisioned model: ARN prefix + "provisioned-model/{model-id}"
        d. Imported model: ARN prefix + "imported-module/{model-id}"
        e. Prompt management: ARN prefix + "prompt/{prompt-id}"
        f. Sagemaker: ARN prefix + "endpoint/{model-id}"
        g. Inference profile: ARN prefix + "{application-?}inference-profile/{model-id}"
        h. Default prompt router: ARN prefix + "default-prompt-router/{prompt-id}"
```
Currently if an AWS ARN gets submitted as the model ID (for example,
prompt management or a provisioned model) then because we attempt to
split the string into a 2/3-part period-separated list, our code breaks
when there is no period delimitter in the string.

This PR makes a best effort attempt to check each case and extract a
model provider and name from the model ID. However, due to the nature of
some formats, we default model provider to "custom" and just return the
model/prompt/resource ID from the rest of the AWS ARN.

## Checklist
- [x] PR author has checked that all the criteria below are met
- The PR description includes an overview of the change
- The PR description articulates the motivation for the change
- The change includes tests OR the PR description describes a testing
strategy
- The PR description notes risks associated with the change, if any
- Newly-added code is easy to change
- The change follows the [library release note
guidelines](https://ddtrace.readthedocs.io/en/stable/releasenotes.html)
- The change includes or references documentation updates if necessary
- Backport labels are set (if
[applicable](https://ddtrace.readthedocs.io/en/latest/contributing.html#backporting))

## Reviewer Checklist
- [x] Reviewer has checked that all the criteria below are met
- Title is accurate
- All changes are related to the pull request's stated goal
- Avoids breaking
[API](https://ddtrace.readthedocs.io/en/stable/versioning.html#interfaces)
changes
- Testing strategy adequately addresses listed risks
- Newly-added code is easy to change
- Release note makes sense to a user of the library
- If necessary, author has acknowledged and discussed the performance
implications of this PR as reported in the benchmarks PR comment
- Backport labels are set in a manner that is consistent with the
[release branch maintenance
policy](https://ddtrace.readthedocs.io/en/latest/contributing.html#backporting)

(cherry picked from commit e8d68ae)
@github-actions github-actions bot requested review from a team as code owners January 15, 2025 19:50
@Yun-Kim Yun-Kim added changelog/no-changelog A changelog entry is not required for this PR. and removed changelog/no-changelog A changelog entry is not required for this PR. labels Jan 15, 2025
@Yun-Kim Yun-Kim requested a review from erikayasuda January 15, 2025 20:03
@Yun-Kim Yun-Kim closed this Jan 15, 2025
@Yun-Kim Yun-Kim reopened this Jan 15, 2025
Copy link
Contributor Author

CODEOWNERS have been resolved as:

releasenotes/notes/fix-bedrock-model-id-parsing-611aea2ca2e00656.yaml   @DataDog/apm-python
tests/contrib/botocore/bedrock_cassettes/amazon_invoke_model_arn.yaml   @DataDog/ml-observability
tests/snapshots/tests.contrib.botocore.test_bedrock.test_invoke_model_using_aws_arn_model_id.json  @DataDog/apm-python
ddtrace/contrib/internal/botocore/services/bedrock.py                   @DataDog/ml-observability
docs/spelling_wordlist.txt                                              @DataDog/python-guild
tests/contrib/botocore/test_bedrock.py                                  @DataDog/ml-observability

@pr-commenter
Copy link

pr-commenter bot commented Jan 15, 2025

Benchmarks

Benchmark execution time: 2025-01-15 21:25:48

Comparing candidate commit 4dad3e7 in PR branch backport-11944-to-2.18 with baseline commit 75a0205 in branch 2.18.

Found 1 performance improvements and 1 performance regressions! Performance is the same for 392 metrics, 2 unstable metrics.

scenario:iast_aspects-ospathdirname_aspect

  • 🟩 execution_time [-386.588ns; -317.001ns] or [-9.500%; -7.790%]

scenario:iast_aspects-ospathsplit_aspect

  • 🟥 execution_time [+612.081ns; +672.375ns] or [+15.870%; +17.434%]

@Yun-Kim Yun-Kim enabled auto-merge (squash) January 15, 2025 21:35
@Yun-Kim Yun-Kim merged commit 0e3612f into 2.18 Jan 15, 2025
564 checks passed
@Yun-Kim Yun-Kim deleted the backport-11944-to-2.18 branch January 15, 2025 21:39
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