Repository navigation
fix(runtime): release ONNX non-WAV audio loading as 0.4.10 - #3876
Merged
Merged
Conversation
Publish the already accepted #3865 standalone wrapper fixes with hosted regression coverage. Signed-off-by: LauraGPT <zhifu.gzf@alibaba-inc.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Scope
Publish the already-merged #3865 fix as standalone
funasr-onnx0.4.10.Offline/streaming Paraformer, SenseVoice and offline VAD decode supported audio
directly with librosa rather than treating all non-WAV input as MP3 and creating
a WAV sidecar. Codec availability still depends on the installed backend.
This PR changes version/docs/contracts only; no new inference implementation.
It does not include the open SenseVoice #3861 or VAD #3862 work.
Evidence
old undeclared pydub path (2 failures). The original results are retained.
Four loaders cover lossless FLAC, WAV/FLAC lists and order, ndarray identity,
visible invalid-audio errors and no sidecar writes. Existing wrapper errors,
streaming overlap/empty-final/VAD and punctuation cases remain selected.
Paraformer FLAC path versus ndarray,48000 samples,10 CUDA nodes/0 CPU nodes.
Four changed source modules are byte-identical to that accepted PR snapshot.
This is not new GPU execution, all-model inference, pretrained ASR quality,
BF16, or full-root-suite validation. No new weights were downloaded.
environment. The new hosted Python3.11/3.12 matrix must separately prove
clean torch-free installation and the installed wheel/sdist contracts.
Publication Gates
Normal signed+DCO PR merge only after exact-head CI and fresh full protection,
identity, permission, head/main and discussion checks. New-main CI and actual
checkout trees are accepted separately. The runtime tag is accepted against
the verified main artifacts and actual workflow triggers, not nonexistent CI.
GitHub receives wheel,sdist,SHA256SUMS as a non-latest runtime release; root
1.4.25 stays latest. PyPI is last, followed by actual official download hashes.
Nano and production remain unchanged.