chore: remove the standalone llm-d launcher

[cquil11]

Remove the llmd-vllm launch path that #3613 added: the LLMD driver, LlmdRequest, and its policy route, plus benchmarks/multi_node/llm-d, benchmarks/multi_node/llm-d-recipes, benchmarks/llm-d, the llmd-vllm runtime settings, and the gb200-nv model override for it. No master config uses llmd-vllm. llm-d runs through srt-slurm's frontend.type: llm-d (NVIDIA/srt-slurm#588), which the submodule includes as of v2.49.0 (#3882).

[github-actions]

Thanks for the contribution!

• Review: If this PR changes files owned by someone other than a repository admin or @SemiAnalysisAI/core, ask one eligible CODEOWNER to complete the latest PR_REVIEW_CHECKLIST.md before contacting a core maintainer on Slack. Follow the template exactly, including As a PR reviewer and CODEOWNER, I have reviewed this and have, so sign-off verification triggers.
• PR verification: Sweeps run only when the PR appends an inferencex-e2e/perf-changelog.yaml entry and carries exactly one primary label: full-sweep-fail-fast (strongly recommended; canary plus per-matrix fail-fast), full-sweep-enabled (canary; matrix jobs continue after a failure), or non-canary-full-sweep-enabled (no canary or fail-fast). The modifiers all-evals, evals-only, and agentx-fast require a primary label. On fork PRs, a maintainer applies the label. See sweep labels and reuse.
• After merging: PR authors must ensure all GitHub Actions jobs pass. Transient failures often pass on rerun; see how to rerun failed jobs.

中文

感谢你的贡献！

• **审阅：**如果 PR 修改的文件归属于仓库管理员及 @SemiAnalysisAI/core 之外的 CODEOWNER，请先联系一位有资格的 CODEOWNER 填写最新的 PR_REVIEW_CHECKLIST.md，再通过 Slack 联系核心维护者。必须严格遵循模板，并保留 As a PR reviewer and CODEOWNER, I have reviewed this and have，才能触发签核验证。
• **PR 验证：**只有当 PR 在 inferencex-e2e/perf-changelog.yaml 末尾追加了条目，并且恰好带有一个主标签时，才会运行扫描：full-sweep-fail-fast（强烈推荐；canary 加逐矩阵 fail-fast）、full-sweep-enabled（有 canary；矩阵任务在失败后继续运行）或 non-canary-full-sweep-enabled（无 canary，也无 fail-fast）。修饰标签 all-evals、evals-only 和 agentx-fast 必须与主标签一起使用。fork PR 的标签由维护者添加。参见扫描标签与复用。
• **合并后：**PR 作者必须确保所有 GitHub Actions 任务通过。临时性失败通常可以通过重新运行恢复；参见重新运行失败任务的说明。

[github-actions]

inferencex-e2e/infx diff: +2 -321

Generally most InferenceX submissions don't need to add changes to the infx package. On the off chance they do, generally the diff is small. Please prompt your agent to clean your PR to keep your diff to only necessary & minimal changes too. We have an increased code quality bar for changes to the infx package.

中文

大多数 InferenceX 提交通常不需要修改 infx 包。即使确实需要，改动通常也很小。请让你的 agent 清理 PR，使改动只包含必要且最小的内容。infx 包的改动有更高的代码质量要求。

[claude]

I reviewed this PR and found no bugs. Beyond the bug-hunter's findings, I checked whether a perf-changelog.yaml entry was required for the removed llm-d recipes and ruled it out — AGENTS.md states deletions are tracked via git history and existing changelog entries, with the mandatory-entry rule applying to additions/modifications, not removals. I also confirmed no residual llmd/LlmdRequest references remain anywhere in infx/.

Extended reasoning...

This is a pure deletion PR (2 insertions / 2158 deletions across 21 files) removing the standalone llm-d launch path (driver, policy route, request model, benchmark scripts, recipes, tests) per the author's stated rationale that llm-d now runs through srt-slurm's own frontend.type. No security-sensitive surface is touched. I verified no dangling references to the removed code remain and that the one plausible gap (missing perf-changelog entry) is explicitly not required by AGENTS.md for deletions.