Repository navigation
feat(v2.0.1): align audio with current speech models and protocols (2.0.1) - #214
Merged
Merged
Conversation
added 2 commits
October 20, 2026 14:39
Merge the audio protocol updates and their ride-along fixes: - rework audio resources (asr/tts/realtime/voices/voice_design/ phrases/customization) for the current qwen-audio and cosyvoice model families; drop audio.translation, superseded by realtime sessions (realtime_translation) - harden the shared ws transport for multi-task connections: cross-task event guard, reader routing/error propagation and lifecycle cleanup; submit explicitly empty job parameters - agentstudio proxy rewrite: original error-code fidelity, wider stream-error translation, stale-cache and iterator-leak fixes - errors/transport: multi-frame SSE error parsing, stream close on error, SDK auth header precedence, tool_choice=None passthrough - examples: rename 01_atomic_capabilities/audio_speech to audio and merge in the speech sample suite (12 CLIs plus direct_call variants); drop the superseded audio_asr/audio_tts snippets - unmark the 2.0 audio preview: the audio API is final
lzsweb
changed the title
added 4 commits
October 20, 2026 14:53
The pre-commit CI job runs hooks over all files; wider inference context than changed-file runs surfaces three findings: - _agentstudio: restore the _STATUS_BY_CLASS Dict[type, int] annotation so __mro__ indexing type-checks - test_audio_management: pinpoint-disable unsubscriptable-object on the mocked Optional[Dict] preview_audio subscript - test_chat: pinpoint-disable not-an-iterable where pylint cannot resolve the stream=True overload
- the Unit Tests workflow only triggered on pull requests to main, so PRs into dev/v2 (the v2 development line) got no CI test coverage - add dev/v2 to the pull_request branch filter; the push trigger and the 3.9/3.12/3.13/3.14 matrix stay unchanged
This reverts commit dddffaa.
- the fixed 0.3s sleep left the duration assertion only 0.1s of slack; a shared CI runner starved the daemon flusher thread for ~0.17s and the 3.14 UT leg failed (0.13 > first + 0.2) - add _await_usage(): poll the usage file every 0.05s with a 5s deadline and use it in all three sleep-then-assert tests; the behavior contracts they pin are unchanged
lzsweb
pushed a commit
that referenced
this pull request
Oct 20, 2026
- re-apply cbed9ae's preview marking, which the 2.0.1 audio alignment (#214) dropped: client.audio mounts lazily and emits a one-time FutureWarning on first access, Audio/AsyncAudio docstrings carry the preview note, and the examples READMEs flag the audio row - the audio surface is still iterating, so the warning stays until it stabilizes
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
audio.translationremoved (use realtime sessions).audio_speech/renamed toaudio/and merged with the 12-CLI speech sample suite (+direct_call/variants).Test plan