Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.
Python asraudiochineseemotion-recognitionfunasrmcp-servermultilingual-asropenai-compatible-apiparaformerpunctuationpytorchreal-time-asrspeaker-diarizationspeech-recognitionspeech-to-textstreaming-asrtranscriptionvllmvoice-activity-detectionwhisper-alternative
chore(release): prepare FunASR 1.4.19
zhifu gao7e6e16cfaee4
diff --git a/README.md b/README.mdindex c1a20b293..4300147fe 100644--- a/README.md+++ b/README.md@@ -364,8 +364,9 @@ recordings with the same evaluation scope.## What's new- **MOSS-Transcribe-Diarize** brings long-form ASR, timestamps, and anonymous speaker labels to FunASR services, Docker, Kubernetes, vLLM/SGLang workflows, and FunClip. [Deploy MOSS ->](./docs/moss_transcribe_diarize.md)-- **FunASR 1.4.18** fixes cross-script hotword misreplacement, filtered-token timestamp alignment, and Chinese VAD sentence/speaker drift. Install with `python -m pip install -U "funasr==1.4.18"`. [Release and verification scope ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.18)- **Previous release: FunASR 1.4.17** adds explicit PCM16 input and improves timestamp, hotword, speaker-sample-rate, and long-segment streaming handling. [Release and verification scope ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.17)+- **FunASR 1.4.19** handles empty-token Paraformer batches, enforces `max_speakers` for automatic clustering, and adds validated JSON-object `--hf_overrides` support to realtime WebSocket serving. Install with `python -m pip install -U "funasr==1.4.19"`. [Release and verification scope ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.19)+ Previous release **FunASR 1.4.18** fixes cross-script hotword misreplacement, filtered-token timestamp alignment, and Chinese VAD sentence/speaker drift. [Release and verification scope ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.18)+ **Earlier release: FunASR 1.4.17** adds explicit PCM16 input and improves timestamp, hotword, speaker-sample-rate, and long-segment streaming handling. [Release and verification scope ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.17)**Earlier release: FunASR 1.4.16** preserves explicit VAD, punctuation, and speaker-model device placement across inference calls, and adds tested native Transformers adoption guides. [Release and verification scope ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.16)- **Native Transformers:** Released **5.17.0** supports Fun-ASR-Nano with the official `-hf` checkpoint, CPU examples and a notebook. [Get started ->](./docs/transformers_native.md)diff --git a/README_ja.md b/README_ja.mdindex 6defef34f..32d37fd07 100644--- a/README_ja.md+++ b/README_ja.md@@ -112,8 +112,9 @@ CER/WER をそろえて比較してください。オフラインのスループ## 最新情報- **MOSS-Transcribe-Diarize** を FunASR service、Docker、Kubernetes、vLLM/SGLang workflow、FunClip に統合し、長時間 ASR、timestamp、匿名 speaker label を一度に処理できます。[MOSS をデプロイ ->](./docs/moss_transcribe_diarize.md)-- **FunASR 1.4.18** は中国語とラテン文字が混在する hotword の誤置換、特殊 token 除去後の timestamp 対応、中国語 VAD 分割での文と話者のずれを修正します。`python -m pip install -U "funasr==1.4.18"` でインストールできます。[リリースと検証範囲 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.18)- 前のリリース **FunASR 1.4.17** は明示的な PCM16 入力を追加し、timestamp、hotword、speaker sample rate、長い segment の streaming 処理を改善しました。[リリースと検証範囲 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.17)+- **FunASR 1.4.19** は Paraformer の空 token batch を処理し、自動クラスタリングの `max_speakers` 上限を適用します。リアルタイム WebSocket サービスには JSON object を検証する `--hf_overrides` を追加しました。`python -m pip install -U "funasr==1.4.19"` でインストールできます。[リリースと検証範囲 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.19)+ 前のリリース **FunASR 1.4.18** は中国語とラテン文字が混在する hotword の誤置換、特殊 token 除去後の timestamp 対応、中国語 VAD 分割での文と話者のずれを修正しました。[リリースと検証範囲 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.18)+ 以前のリリース **FunASR 1.4.17** は明示的な PCM16 入力を追加し、timestamp、hotword、speaker sample rate、長い segment の streaming 処理を改善しました。[リリースと検証範囲 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.17)以前のリリース **FunASR 1.4.16** は複数回の推論でも VAD、句読点、話者モデルの明示的な device 配置を保持し、テスト済みのネイティブ Transformers 導入ガイドを追加しました。[リリースと検証範囲 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.16)- **ネイティブ Transformers:** 正式版 **5.17.0** が Fun-ASR-Nano に対応。公式 `-hf` checkpoint、CPU サンプル、Notebook:[導入ガイド(英語) ->](./docs/transformers_native.md)diff --git a/README_ko.md b/README_ko.mdindex b3ebff301..c5b328865 100644--- a/README_ko.md+++ b/README_ko.md@@ -111,8 +111,9 @@ checkpoint/revision, 오디오 집합, 하드웨어, 배치, 워밍업, 측정## 최신 소식- **MOSS-Transcribe-Diarize**를 FunASR service, Docker, Kubernetes, vLLM/SGLang workflow, FunClip에 통합해 긴 오디오 ASR, timestamp, 익명 speaker label을 한 번에 처리합니다. [MOSS 배포 ->](./docs/moss_transcribe_diarize.md)-- **FunASR 1.4.18**은 중국어와 라틴 문자가 섞인 hotword의 잘못된 치환, 특수 token 제거 후 timestamp 정렬, 중국어 VAD 분할의 문장 및 화자 불일치를 수정합니다. `python -m pip install -U "funasr==1.4.18"`로 설치할 수 있습니다. [릴리스 및 검증 범위 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.18)- 이전 릴리스 **FunASR 1.4.17**은 명시적 PCM16 입력을 추가하고 timestamp, hotword, speaker sample rate, 긴 segment의 streaming 처리를 개선했습니다. [릴리스 및 검증 범위 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.17)+- **FunASR 1.4.19**는 Paraformer의 빈 token batch를 처리하고 자동 클러스터링에 `max_speakers` 상한을 적용합니다. 실시간 WebSocket 서비스에 JSON object를 검증하는 `--hf_overrides`도 추가했습니다. `python -m pip install -U "funasr==1.4.19"`로 설치할 수 있습니다. [릴리스 및 검증 범위 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.19)+ 이전 릴리스 **FunASR 1.4.18**은 중국어와 라틴 문자가 섞인 hotword의 잘못된 치환, 특수 token 제거 후 timestamp 정렬, 중국어 VAD 분할의 문장 및 화자 불일치를 수정했습니다. [릴리스 및 검증 범위 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.18)+ 그 이전 릴리스 **FunASR 1.4.17**은 명시적 PCM16 입력을 추가하고 timestamp, hotword, speaker sample rate, 긴 segment의 streaming 처리를 개선했습니다. [릴리스 및 검증 범위 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.17)그 이전 릴리스 **FunASR 1.4.16**은 여러 번의 추론에서도 VAD, 문장 부호, 화자 모델의 명시적 device 배치를 유지하고 검증된 네이티브 Transformers 도입 가이드를 추가했습니다. [릴리스 및 검증 범위 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.16)- **네이티브 Transformers:** 정식 버전 **5.17.0**이 Fun-ASR-Nano를 지원합니다. 공식 `-hf` checkpoint, CPU 예제, Notebook: [설치 가이드(영문) ->](./docs/transformers_native.md)diff --git a/README_zh.md b/README_zh.mdindex 53af3bed2..2c5ac1e3a 100644--- a/README_zh.md+++ b/README_zh.md@@ -157,8 +157,9 @@ checkpoint/revision、音频集、硬件、批量大小、预热、计时范围## 最新动态- **MOSS-Transcribe-Diarize** 已接入 FunASR 服务、Docker、Kubernetes、vLLM/SGLang 工作流和 FunClip,一次完成长音频转写、时间戳与匿名说话人标注。[部署 MOSS ->](./docs/moss_transcribe_diarize_zh.md)-- **FunASR 1.4.18** 修复跨中英文热词误替换、过滤特殊符号后的时间戳对齐,以及中文 VAD 分段的句子与说话人错位。安装:`python -m pip install -U "funasr==1.4.18"`。[发布说明与验证范围 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.18)- **上一版本:FunASR 1.4.17** 新增显式 PCM16 输入,改进时间戳、热词、说话人采样率与长片段流式处理。[发布说明与验证范围 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.17)+- **FunASR 1.4.19** 修复 Paraformer 空 token 批次处理,使自动聚类遵守 `max_speakers` 上限,并为实时 WebSocket 服务新增带 JSON 对象校验的 `--hf_overrides` 参数。安装:`python -m pip install -U "funasr==1.4.19"`。[发布说明与验证范围 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.19)+ 上一版本 **FunASR 1.4.18** 修复跨中英文热词误替换、过滤特殊符号后的时间戳对齐,以及中文 VAD 分段的句子与说话人错位。[发布说明与验证范围 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.18)+ **更早版本:FunASR 1.4.17** 新增显式 PCM16 输入,改进时间戳、热词、说话人采样率与长片段流式处理。[发布说明与验证范围 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.17)**更早版本:FunASR 1.4.16** 修复 VAD、标点和说话人子模型在多次推理时的显式设备放置,并新增经过测试的原生 Transformers 使用指南。[发布说明与验证范围 ->](https://github.com/modelscope/FunASR/releases/tag/v1.4.16)- **原生 Transformers:** 正式版 **5.17.0** 已支持 Fun-ASR-Nano。官方 `-hf` 权重、CPU 示例与 Notebook:[安装与推理 ->](./docs/transformers_native_zh.md)diff --git a/docs/repository_roles.md b/docs/repository_roles.mdindex 75c3b88e6..5e19bda2e 100644--- a/docs/repository_roles.md+++ b/docs/repository_roles.md@@ -5,7 +5,7 @@ This document explains the responsibility boundaries, user entry points, and iss> **Directional roadmap, not a release promise.**> This document records shipped capabilities and active work, but does not commit to> future version numbers or dates. The current Python release is-> [`funasr==1.4.18`](https://github.com/modelscope/FunASR/releases/tag/v1.4.18).+> [`funasr==1.4.19`](https://github.com/modelscope/FunASR/releases/tag/v1.4.19).> Any future breaking release still requires a maintainer-approved milestone and> migration plan.diff --git a/docs/repository_roles_zh.md b/docs/repository_roles_zh.mdindex 4f6de0827..5f51abd75 100644--- a/docs/repository_roles_zh.md+++ b/docs/repository_roles_zh.md@@ -4,7 +4,7 @@> **方向性路线图,不是版本承诺。**> 本文档记录已交付能力与正在推进的工作,但不承诺未来版本号或日期。当前 Python-> 版本是 [`funasr==1.4.18`](https://github.com/modelscope/FunASR/releases/tag/v1.4.18)。+> 版本是 [`funasr==1.4.19`](https://github.com/modelscope/FunASR/releases/tag/v1.4.19)。> 任何未来的 breaking release 仍需 maintainer 确认 milestone 与迁移方案。---diff --git a/funasr/version.txt b/funasr/version.txtindex f689e8c1a..fd4ca57b8 100644--- a/funasr/version.txt+++ b/funasr/version.txt@@ -1 +1 @@-1.4.18+1.4.19diff --git a/tests/test_release_version_contract.py b/tests/test_release_version_contract.pyindex 91c5f2739..a5311cae5 100644--- a/tests/test_release_version_contract.py+++ b/tests/test_release_version_contract.py@@ -4,7 +4,7 @@ import pytestROOT = Path(__file__).resolve().parents[1]-RELEASE_VERSION = "1.4.18"+RELEASE_VERSION = "1.4.19"README_FILES = ("README.md", "README_zh.md", "README_ja.md", "README_ko.md")REPOSITORY_ROLE_FILES = ("docs/repository_roles.md", "docs/repository_roles_zh.md")@@ -17,7 +17,7 @@ REPOSITORY_ROLE_FILES = ("docs/repository_roles.md", "docs/repository_roles_zh.m])def test_release_hotword_summary_matches_cross_script_fix(name, expected, incorrect):lines = (ROOT / name).read_text(encoding="utf-8").splitlines()- summaries = [line for line in lines if f"**FunASR {RELEASE_VERSION}**" in line]+ summaries = [line for line in lines if "**FunASR 1.4.18**" in line]assert len(summaries) == 1assert expected in summaries[0]assert incorrect not in summaries[0]@@ -44,6 +44,8 @@ def test_release_is_recorded_in_all_top_level_readmes():def test_repository_role_guides_link_the_current_release():for name in REPOSITORY_ROLE_FILES:text = (ROOT / name).read_text(encoding="utf-8")+ assert f"funasr=={RELEASE_VERSION}" in text+ assert f"releases/tag/v{RELEASE_VERSION}" in text@@ -56,10 +58,10 @@ def test_release_documents_do_not_describe_a_pending_candidate():assert version_lines, nameassert not any(label in line.lower() for line in version_linesfor label in candidate_labels), name- assert 'pip install -U "funasr==1.4.17"' not in text, name+ assert 'pip install -U "funasr==1.4.18"' not in text, namefor name in REPOSITORY_ROLE_FILES:text = (ROOT / name).read_text(encoding="utf-8")- assert "funasr==1.4.17" not in text, name+ assert "funasr==1.4.18" not in text, nameassert "not published yet" not in text, nameassert "\u5c1a\u672a\u6b63\u5f0f\u53d1\u5e03" not in text, nameassert not any(label in text.lower() for label in candidate_labels), name