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
docs: align v1.4.14 Blackwell release links
zhifu gao1586c13efbd3
diff --git a/web-pages/product-site/legacy/blog/funasr-v1-4-14-portable-source-release.html b/web-pages/product-site/legacy/blog/funasr-v1-4-14-portable-source-release.htmlindex b32b8229f..c08d65f93 100644--- a/web-pages/product-site/legacy/blog/funasr-v1-4-14-portable-source-release.html+++ b/web-pages/product-site/legacy/blog/funasr-v1-4-14-portable-source-release.html@@ -44,20 +44,19 @@<h2>发布验证</h2><p>签名标签 <code>v1.4.14</code> 精确指向 <code>a23946a8b3e9146f9bf80e18f768298f0cd000b7</code>。审计重新执行 210 项发布、文档与站点回归,并通过 compileall、PEP 517 构建、Twine、隔离 wheel 安装和 public PyPI 回源校验;重建 wheel/sdist 的解包内容与公网制品零差异。</p>- <h2>12 个 v1.4.14 资产与 Blackwell 独立运行时</h2>- <p><a href="https://github.com/modelscope/FunASR/releases/tag/v1.4.14" target="_blank" rel="noopener">v1.4.14 Release</a> 固定复用 <a href="https://github.com/modelscope/FunASR/releases/tag/runtime-llamacpp-v0.2.6" target="_blank" rel="noopener">runtime-llamacpp-v0.2.6</a>。v1.4.14 页面共 12 个资产:2 个 Python 制品、9 个镜像运行时和 1 份校验清单;第 10 个 Blackwell 运行时保留在独立运行时发布页:</p>+ <h2>13 个 v1.4.14 资产,含 Blackwell 运行时</h2>+ <p><a href="https://github.com/modelscope/FunASR/releases/tag/v1.4.14" target="_blank" rel="noopener">v1.4.14 Release</a> 固定复用 <a href="https://github.com/modelscope/FunASR/releases/tag/runtime-llamacpp-v0.2.6" target="_blank" rel="noopener">runtime-llamacpp-v0.2.6</a>。v1.4.14 页面现有 13 个资产:2 个 Python 制品、10 个镜像运行时和 1 份校验清单;Windows CUDA Blackwell(sm_120)包已与其余运行时一同附在此发布页。</p><ul><li>Python:<code>funasr-1.4.14-py3-none-any.whl</code>、<code>funasr-1.4.14.tar.gz</code></li><li>Linux:<code>funasr-llamacpp-linux-arm64.tar.gz</code>、<code>funasr-llamacpp-linux-x64.tar.gz</code>、<code>funasr-llamacpp-linux-x64-avx2.tar.gz</code>、<code>funasr-llamacpp-linux-x64-vulkan.tar.gz</code></li><li>macOS:<code>funasr-llamacpp-macos-arm64.tar.gz</code></li>- <li>Windows 镜像:<code>funasr-llamacpp-windows-x64.zip</code>、<code>funasr-llamacpp-windows-x64-avx2.zip</code>、<code>funasr-llamacpp-windows-x64-cuda.zip</code>、<code>funasr-llamacpp-windows-x64-vulkan.zip</code></li>- <li>Blackwell 独立运行时:<a href="https://github.com/modelscope/FunASR/releases/download/runtime-llamacpp-v0.2.6/funasr-llamacpp-windows-x64-cuda-blackwell.zip" target="_blank" rel="noopener"><code>funasr-llamacpp-windows-x64-cuda-blackwell.zip</code></a></li>+ <li>Windows 镜像:<code>funasr-llamacpp-windows-x64.zip</code>、<code>funasr-llamacpp-windows-x64-avx2.zip</code>、<code>funasr-llamacpp-windows-x64-cuda.zip</code>、<code>funasr-llamacpp-windows-x64-vulkan.zip</code>,以及 Blackwell(sm_120)<a href="https://github.com/modelscope/FunASR/releases/download/v1.4.14/funasr-llamacpp-windows-x64-cuda-blackwell.zip" target="_blank" rel="noopener"><code>funasr-llamacpp-windows-x64-cuda-blackwell.zip</code></a></li><li>校验:<code>SHA256SUMS</code></li></ul><pre><code>sha256sum -c SHA256SUMS</code></pre><p>按硬件选择、启动命令、烟测和已知边界见<a href="/deploy/llama-cpp.html"> llama.cpp 工业部署页</a>;GPU 服务可继续参考 <a href="/deploy/vllm.html">vLLM 部署页</a>。</p>- <div class="cta"><p>从固定版本和可校验资产开始,在目标硬件与真实业务音频上复测后再上线。</p><a class="btn" href="https://github.com/modelscope/FunASR/releases/tag/v1.4.14" target="_blank" rel="noopener">查看 v1.4.14 的 12 个资产</a></div>+ <div class="cta"><p>从固定版本和可校验资产开始,在目标硬件与真实业务音频上复测后再上线。</p><a class="btn" href="https://github.com/modelscope/FunASR/releases/tag/v1.4.14" target="_blank" rel="noopener">查看 v1.4.14 的 13 个资产</a></div></div></article><footer><div class="container">© 2024-2026 Tongyi Lab, Alibaba Group · <a href="https://github.com/modelscope/FunASR/blob/main/LICENSE">MIT License</a></div></footer></body>diff --git a/web-pages/product-site/legacy/en/blog/funasr-v1-4-14-portable-source-release.html b/web-pages/product-site/legacy/en/blog/funasr-v1-4-14-portable-source-release.htmlindex 69fa4f19a..48ab7375b 100644--- a/web-pages/product-site/legacy/en/blog/funasr-v1-4-14-portable-source-release.html+++ b/web-pages/product-site/legacy/en/blog/funasr-v1-4-14-portable-source-release.html@@ -44,20 +44,19 @@<h2>Release verification</h2><p>The signed <code>v1.4.14</code> tag resolves exactly to <code>a23946a8b3e9146f9bf80e18f768298f0cd000b7</code>. The audit reran 210 release, documentation, and site regressions plus compileall, PEP 517 builds, Twine, an isolated wheel install, and public PyPI round trips. Extracted contents from the rebuilt wheel and sdist had zero differences from the public artifacts.</p>- <h2>12 v1.4.14 assets plus a separate Blackwell runtime</h2>- <p>The <a href="https://github.com/modelscope/FunASR/releases/tag/v1.4.14" target="_blank" rel="noopener">v1.4.14 Release</a> pins <a href="https://github.com/modelscope/FunASR/releases/tag/runtime-llamacpp-v0.2.6" target="_blank" rel="noopener">runtime-llamacpp-v0.2.6</a>. Its 12 assets are two Python distributions, 9 mirrored runtimes, and one checksum manifest; the tenth, Blackwell runtime remains on the separate runtime release:</p>+ <h2>13 v1.4.14 assets, including the Blackwell runtime</h2>+ <p>The <a href="https://github.com/modelscope/FunASR/releases/tag/v1.4.14" target="_blank" rel="noopener">v1.4.14 Release</a> pins <a href="https://github.com/modelscope/FunASR/releases/tag/runtime-llamacpp-v0.2.6" target="_blank" rel="noopener">runtime-llamacpp-v0.2.6</a>. Its 13 assets are two Python distributions, 10 mirrored runtimes, and one checksum manifest; the Windows CUDA Blackwell (sm_120) archive is now attached beside the other runtime packages.</p><ul><li>Python: <code>funasr-1.4.14-py3-none-any.whl</code> and <code>funasr-1.4.14.tar.gz</code></li><li>Linux: <code>funasr-llamacpp-linux-arm64.tar.gz</code>, <code>funasr-llamacpp-linux-x64.tar.gz</code>, <code>funasr-llamacpp-linux-x64-avx2.tar.gz</code>, <code>funasr-llamacpp-linux-x64-vulkan.tar.gz</code></li><li>macOS: <code>funasr-llamacpp-macos-arm64.tar.gz</code></li>- <li>Mirrored Windows runtimes: <code>funasr-llamacpp-windows-x64.zip</code>, <code>funasr-llamacpp-windows-x64-avx2.zip</code>, <code>funasr-llamacpp-windows-x64-cuda.zip</code>, <code>funasr-llamacpp-windows-x64-vulkan.zip</code></li>- <li>Separate Blackwell runtime: <a href="https://github.com/modelscope/FunASR/releases/download/runtime-llamacpp-v0.2.6/funasr-llamacpp-windows-x64-cuda-blackwell.zip" target="_blank" rel="noopener"><code>funasr-llamacpp-windows-x64-cuda-blackwell.zip</code></a></li>+ <li>Mirrored Windows runtimes: <code>funasr-llamacpp-windows-x64.zip</code>, <code>funasr-llamacpp-windows-x64-avx2.zip</code>, <code>funasr-llamacpp-windows-x64-cuda.zip</code>, <code>funasr-llamacpp-windows-x64-vulkan.zip</code>, and the Blackwell (sm_120) <a href="https://github.com/modelscope/FunASR/releases/download/v1.4.14/funasr-llamacpp-windows-x64-cuda-blackwell.zip" target="_blank" rel="noopener"><code>funasr-llamacpp-windows-x64-cuda-blackwell.zip</code></a></li><li>Verification: <code>SHA256SUMS</code></li></ul><pre><code>sha256sum -c SHA256SUMS</code></pre><p>Use the <a href="/en/deploy/llama-cpp.html">llama.cpp production deployment page</a> for hardware selection, launch commands, smoke tests, and known boundaries. GPU API deployments can continue with the <a href="/en/deploy/vllm.html">vLLM guide</a>.</p>- <div class="cta"><p>Start from a pinned version and verified assets, then revalidate on the target hardware with representative audio before rollout.</p><a class="btn" href="https://github.com/modelscope/FunASR/releases/tag/v1.4.14" target="_blank" rel="noopener">Open all 12 v1.4.14 assets</a></div>+ <div class="cta"><p>Start from a pinned version and verified assets, then revalidate on the target hardware with representative audio before rollout.</p><a class="btn" href="https://github.com/modelscope/FunASR/releases/tag/v1.4.14" target="_blank" rel="noopener">Open all 13 v1.4.14 assets</a></div></div></article><footer><div class="container">© 2024-2026 Tongyi Lab, Alibaba Group · <a href="https://github.com/modelscope/FunASR/blob/main/LICENSE">MIT License</a></div></footer></body>diff --git a/web-pages/product-site/tests/browser/product-site.spec.ts b/web-pages/product-site/tests/browser/product-site.spec.tsindex 8f8f5ed85..e907f0ce5 100644--- a/web-pages/product-site/tests/browser/product-site.spec.ts+++ b/web-pages/product-site/tests/browser/product-site.spec.ts@@ -519,15 +519,15 @@ test('FunASR v1.4.14 release hashes wrap on mobile', async ({ page }, testInfo)await expect(page.locator('h1')).toContainText('v1.4.14');const layout = await page.evaluate(() => ({overflow: document.documentElement.scrollWidth - document.documentElement.clientWidth,- assetCount: document.body.innerText.includes('12'),+ assetCount: document.body.innerText.includes('13'),blackwell: document.querySelector<HTMLAnchorElement>(- 'a[href*="runtime-llamacpp-v0.2.6/funasr-llamacpp-windows-x64-cuda-blackwell.zip"]',+ 'a[href*="v1.4.14/funasr-llamacpp-windows-x64-cuda-blackwell.zip"]',)?.href,}));expect(layout.overflow).toBeLessThanOrEqual(1);expect(layout.assetCount).toBe(true);- expect(layout.blackwell).toContain('runtime-llamacpp-v0.2.6');+ expect(layout.blackwell).toContain('v1.4.14');}await page.screenshot({diff --git a/web-pages/product-site/tests/test_output.py b/web-pages/product-site/tests/test_output.pyindex 8da5bf736..e4e1e5a82 100644--- a/web-pages/product-site/tests/test_output.py+++ b/web-pages/product-site/tests/test_output.py@@ -679,9 +679,9 @@ def test_v1_4_3_release_blog_is_bilingual_and_verifiable('FunASR v1.4.14','可移植源码包','MOSS-Transcribe-Diarize',- '12 个 v1.4.14 资产',- '9 个镜像运行时',- 'Blackwell 独立运行时',+ '13 个 v1.4.14 资产',+ '10 个镜像运行时',+ '含 Blackwell 运行时','SHA256SUMS',),),@@ -692,9 +692,9 @@ def test_v1_4_3_release_blog_is_bilingual_and_verifiable('FunASR v1.4.14','Portable Source Archives','MOSS-Transcribe-Diarize',- '12 v1.4.14 assets',- '9 mirrored runtimes',- 'separate Blackwell runtime',+ '13 v1.4.14 assets',+ '10 mirrored runtimes',+ 'including the Blackwell runtime','SHA256SUMS',),),@@ -716,7 +716,7 @@ def test_v1_4_14_release_blog_is_bilingual_and_verifiable()assert soup.select_one('a[href="https://github.com/modelscope/FunASR/releases/download/'- 'runtime-llamacpp-v0.2.6/funasr-llamacpp-windows-x64-cuda-blackwell.zip"]'+ 'v1.4.14/funasr-llamacpp-windows-x64-cuda-blackwell.zip"]')for marker in markers:assert marker in text