chore: consolidate gptoss example + fixes

[arekay-nv]

What does this PR do?

Consolidate the gpt example plus pipe clean to make sure it works end to end after recent refactors.

Type of change

• Bug fix
• New feature
• Documentation update
• Refactor/cleanup

Related issues

Testing

• Tests added/updated
• All tests pass locally
• Manual testing completed

Checklist

• Code follows project style
• Pre-commit hooks pass
• Documentation updated (if needed)

[github-actions]

MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅

[gemini-code-assist]

Code Review

This pull request focuses on extensive documentation updates, example consolidation, and internal API refactoring. Key changes include restructuring the GPT-OSS-120B benchmarking examples, updating CLI argument references in READMEs, and transitioning the CPU affinity configuration to a boolean flag. Additionally, the internal issue_query method was renamed to issue. Feedback is provided regarding the removal of conditional checks in the endpoint client configuration, which inadvertently prevents the programmatic override of adapters and accumulators.

[Copilot]

Pull request overview

This PR updates documentation and examples to align with recent endpoint-client refactors and consolidates the GPT-OSS-120B end-to-end example (configs + accuracy scripts) so it runs cleanly again.

Changes:

• Update docs/READMEs to reflect refactored endpoint client API (issue/poll/drain, updated worker internals, CLI flag updates).
• Consolidate GPT-OSS-120B example guidance (vLLM + SGLang) and remove the separate SGLang-only example README.
• Add standalone evaluation scripts (GPQA/AIME25/LiveCodeBench) and plumb --force-regenerate through dataset generation in the example runner.

Reviewed changes

Copilot reviewed 13 out of 16 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
src/inference_endpoint/profiling/README.md	Updates profiling callouts to match refactored client/worker method names.
src/inference_endpoint/endpoint_client/config.py	Adjusts default resolution so adapter/accumulator are always derived from api_type.
src/inference_endpoint/endpoint_client/README.md	Updates client API usage from issue_query to issue.
src/inference_endpoint/dataset_manager/README.md	Fixes import path for DatasetFormat.
examples/README.md	Updates GPT-OSS-120B example description and removes separate SGLang example entry.
examples/07_GPT-OSS-120B_SGLang_Example/README.md	Removes redundant SGLang-only README (content consolidated into 04 example).
examples/04_GPTOSS120B_Example/run.py	Adds force_regenerate passthrough for dataset generation.
examples/04_GPTOSS120B_Example/gptoss_120b_example.yaml	Updates default endpoint port to 30000.
examples/04_GPTOSS120B_Example/eval_livecodebench.py	Adds standalone LiveCodeBench re-scoring script from an existing report dir.
examples/04_GPTOSS120B_Example/eval_gpqa.py	Adds standalone GPQA re-scoring script from an existing report dir.
examples/04_GPTOSS120B_Example/eval_aime.py	Adds standalone AIME25 re-scoring script from an existing report dir.
examples/04_GPTOSS120B_Example/Readme.md	Consolidates end-to-end instructions for vLLM/SGLang + accuracy suite + troubleshooting.
examples/02_ServerBenchmarking/README.md	Updates CLI example flags to current --endpoints/--dataset/--model usage.
docs/CLI_QUICK_REFERENCE.md	Updates init command guidance to include concurrency and removes redundant line.
docs/CLIENT_PERFORMANCE_TUNING.md	Updates CPU affinity docs to match enable_cpu_affinity + --no-cpu-affinity.
AGENTS.md	Updates repository structure notes for core/ layout and record location.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Consolidates and repairs the gpt-oss-120b examples and related client configuration behavior so the selected endpoint_config.api_type reliably drives adapter/accumulator selection end-to-end after recent refactors.

Changes:

• Propagate endpoint_config.api_type into settings.client.api_type at config construction time, and make HTTPClientConfig.with_updates() clear stale auto-resolved fields when api_type changes.
• Update benchmark execution/docs to rely on the propagated api_type (no runtime patching in execute.py) and refresh client/profiling documentation for renamed methods.
• Consolidate GPT-OSS examples (remove the separate SGLang example README, enhance the unified example, add accuracy eval scripts, and update example docs/imports).

Reviewed changes

Copilot reviewed 16 out of 19 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
tests/unit/config/test_schema.py	Adds regression tests ensuring api_type propagation and adapter/accumulator resolution behave correctly across with_updates().
src/inference_endpoint/profiling/README.md	Updates profiling guide to match renamed client/worker methods.
src/inference_endpoint/endpoint_client/config.py	Adds HTTPClientConfig.with_updates() logic to clear adapter/accumulator when api_type changes.
src/inference_endpoint/endpoint_client/README.md	Updates docs to reflect HTTPEndpointClient.issue() API naming.
src/inference_endpoint/dataset_manager/README.md	Fixes import guidance for DatasetFormat (not exported from package root).
src/inference_endpoint/config/schema.py	Adds validator to propagate endpoint_config.api_type into the internal HTTP client config at construction.
src/inference_endpoint/commands/benchmark/execute.py	Removes runtime api_type override, relying on schema propagation.
examples/README.md	Updates GPT-OSS example description and removes separate SGLang example entry.
examples/07_GPT-OSS-120B_SGLang_Example/README.md	Deletes now-redundant standalone SGLang example README.
examples/04_GPTOSS120B_Example/run.py	Adds force_regenerate passthrough to dataset generators.
examples/04_GPTOSS120B_Example/gptoss_120b_example.yaml	Updates endpoint default port in the example config.
examples/04_GPTOSS120B_Example/eval_livecodebench.py	Adds standalone scoring script for LiveCodeBench from a saved dataset/report.
examples/04_GPTOSS120B_Example/eval_gpqa.py	Adds standalone scoring script for GPQA from a saved dataset/report.
examples/04_GPTOSS120B_Example/eval_aime.py	Adds standalone scoring script for AIME25 from a saved dataset/report.
examples/04_GPTOSS120B_Example/Readme.md	Expands and consolidates end-to-end instructions for vLLM + SGLang and accuracy workflows.
examples/02_ServerBenchmarking/README.md	Updates CLI usage example to current long-form flags.
docs/CLI_QUICK_REFERENCE.md	Updates init command quick reference to include concurrency template option.
docs/CLIENT_PERFORMANCE_TUNING.md	Updates CPU-affinity documentation to match enable_cpu_affinity config/CLI.
AGENTS.md	Updates repo structure documentation (core/types, record location).

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.