chore(deps-dev): Bump guidellm from 0.7.3 to 0.7.4 in the python-minor-patch group across 1 directory

[dependabot]

Bumps the python-minor-patch group with 1 update in the / directory: guidellm.

Updates guidellm from 0.7.3 to 0.7.4

Release notes

Sourced from guidellm's releases.

GuideLLM v0.7.4

Overview

GuideLLM v0.7.4 is a minor release which backports many bugfixes including a change to how token metrics interact with warmup/cooldown. This change will cause differences in which requests are sampled for latency metrics and will bound token event to only the tokens that occur outside warmup/cooldown.

To get started, install with:

pip install guidellm[recommended]==0.7.4

Or from source with:

pip install 'guidellm[recommended] @ git+https://github.com/vllm-project/guidellm.git'@v0.7.4

What's Fixed

• Change to the methodology around tracking token latency and throughput events. See #1078 for full details of the problem.

What's Changed

Bug fixes

• Bound token events by when they occur in vllm-project/guidellm#1079
• fix(openai): validate non-streaming tool calls before yielding in vllm-project/guidellm#1099
• fix(openai): surface nested Responses streaming errors in vllm-project/guidellm#1107
• fix(openai): accept SSE DONE markers without a space in vllm-project/guidellm#1106
• fix: serialize audio translation filename in vllm-project/guidellm#1096
• Pull pending requests into a worker only when a consumer is waiting in vllm-project/guidellm#1042
• fix: use spawn multiprocessing context on macOS in vllm-project/guidellm#1026
• Fix encoded audio sample frame metrics in vllm-project/guidellm#987
• Fix non-streaming audio response parsing in vllm-project/guidellm#981
• Fix multipart filenames for audio uploads in vllm-project/guidellm#990

CI environment

• [GitHub Actions]: Bump docker/setup-qemu-action from 4.2.0 to 4.3.0 in vllm-project/guidellm#1129
• [GitHub Actions]: Bump redhat-actions/buildah-build from 3.0.2 to 3.1.0 in vllm-project/guidellm#1062
• [GitHub Actions]: Bump redhat-actions/buildah-build from 2.13 to 3 in vllm-project/guidellm#1036
• [GitHub Actions]: Bump redhat-actions/push-to-registry from 2.8 to 3 in vllm-project/guidellm#1037
• fix: keep GHCR latest/stable as multi-arch manifests in vllm-project/guidellm#1018

Dependency updates

• build(deps): Bump transformers from 5.5.0 to 5.10.1 in vllm-project/guidellm#1076
• Bump h2 from 4.3.0 to 4.4.1 in vllm-project/guidellm#1011
• Bump aiohttp from 3.14.1 to 3.14.3 in vllm-project/guidellm#995

... (truncated)

Commits

• 291a6e6 Bound token events by when they occur (#1079)
• 67c2993 [GitHub Actions]: Bump docker/setup-qemu-action from 4.2.0 to 4.3.0 (#1129)
• 0e6e551 fix(openai): validate non-streaming tool calls before yielding (#1099)
• f9e9be2 fix(openai): surface nested Responses streaming errors (#1107)
• 6d9c1ea fix(openai): accept SSE DONE markers without a space (#1106)
• 3fe60ae fix: serialize audio translation filename (#1096)
• 1f964e5 build(deps): Bump transformers from 5.5.0 to 5.10.1 (#1076)
• be37e84 Pull pending requests into a worker only when a consumer is waiting (#1042)
• f083085 [GitHub Actions]: Bump redhat-actions/buildah-build from 3.0.2 to 3.1.0 (#1062)
• d26fec5 [GitHub Actions]: Bump redhat-actions/buildah-build from 2.13 to 3 (#1036)
• Additional commits viewable in compare view