[Benchmarks] Metadata generator

[PatKamin]

Generate Compute Benchmarks scenarios' metadata automatically based on Benchmark.explicit_group.

[Copilot]

Pull Request Overview

This PR refactors the metadata generation for Compute Benchmarks by introducing an automated metadata generator. The changes centralize metadata creation logic and ensure consistency between benchmark group membership and metadata definitions.

• Introduces a new ComputeMetadataGenerator class to automatically generate metadata based on Benchmark.explicit_group
• Updates type annotations for optional fields in BenchmarkMetadata to use union syntax
• Replaces manual metadata creation with automated generation in the Compute suite

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
devops/scripts/benchmarks/utils/result.py	Updates type annotations for range fields to use explicit union syntax
devops/scripts/benchmarks/benches/compute_metadata.py	Adds new metadata generator class with centralized group metadata definitions
devops/scripts/benchmarks/benches/compute.py	Refactors to use automated metadata generation instead of manual creation

[lslusarczyk]

I would love to see some test that executes this code, while we do a change. To not rely on manual checks. But UTs, if any, we can add in other PR.

[Copilot]

Pull Request Overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

[lslusarczyk]

thank you, good job

[PatKamin]

@intel/llvm-gatekeepers, please merge. Failing e2e test is unrelated to this benchmarks framework change. This change should fix missing charts issue in the dashboard. Benchmark run CI job passed.

[lslusarczyk]

@intel/llvm-gatekeepers , could you please merge this? Changes are OK, CI passes and this is an important fix that makes benchmarks plots working again.