[AMD][MI355X] update model for gpt-oss

[ukannika]

Switch moe backend to CK to leverage the benefits from this PR.
vllm-project/vllm#42098

---

Note

Low Risk
Single YAML model ID change for a benchmark config; no application or auth logic touched.

Overview
Updates the MI355X gptoss-fp4-mi355x-vllm benchmark entry so the model Hugging Face ID is openai/gpt-oss-120b instead of amd/gpt-oss-120b-w-mxfp4-a-fp8. Image (vllm/vllm-openai-rocm:v0.22.0), runner, precision label, and fixed-seq-len search spaces are unchanged.

This aligns the MI355X vLLM gpt-oss FP4 job with the other gpt-oss entries in the same config (e.g. MI325X/MI300X) and is intended to exercise vLLM’s CK MoE path (per upstream vLLM work referenced in the PR description).

^{Reviewed by Cursor Bugbot for commit decdda7. Bugbot is set up for automated code reviews on this repo. Configure here.}

[claude]

Claude Code Review

This pull request is from a fork — automated review is disabled. A repository maintainer can comment @claude review to run a one-time review.

[chunfangamd]

/sweep test-config --config-files .github/configs/amd-master.yaml --config-keys gptoss-fp4-mi355x-vllm

[github-actions]

@chunfangamd Kicking off a sweep.

Run: https://github.com/SemiAnalysisAI/InferenceX/actions/runs/26845310867
Command: test-config --config-files .github/configs/amd-master.yaml --config-keys gptoss-fp4-mi355x-vllm
Pinned ref: decdda7
Approval: not required (trusted collaborator).