https://59d713fc.rickylin.pages.dev/tags/video/Recent content in Video on RickyHugo -- gohugo.ioenFri, 31 Jul 2026 10:46:55 +0800https://59d713fc.rickylin.pages.dev/posts/2026/20260731-local-llm-audio-video/Fri, 31 Jul 2026 10:46:55 +0800https://59d713fc.rickylin.pages.dev/posts/2026/20260731-local-llm-audio-video/<h3 id="video">Video</h3> <ul> <li><a href="https://huggingface.co/Wan-AI" target="_blank" rel="noopener">https://huggingface.co/Wan-AI</a></li> <li><a href="https://github.com/Wan-Video/Wan2.2" target="_blank" rel="noopener">https://github.com/Wan-Video/Wan2.2</a></li> <li><a href="https://github.com/Blaizzy/mlx-video" target="_blank" rel="noopener">https://github.com/Blaizzy/mlx-video</a></li> <li><a href="https://huggingface.co/fishaudio/s2-pro" target="_blank" rel="noopener">https://huggingface.co/fishaudio/s2-pro</a></li> </ul> <h4 id="install">Install</h4> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><code class="language-shell" data-lang="shell"><span style="display:flex;"><span>uv tool install git+https://github.com/Blaizzy/mlx-video.git </span></span></code></pre></div><h4 id="download-model">Download Model</h4> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><code class="language-shell" data-lang="shell"><span style="display:flex;"><span>HF_HUB_DISABLE_XET<span style="color:#f92672">=</span><span style="color:#ae81ff">1</span> hf download Wan-AI/Wan2.2-TI2V-5B --local-dir Wan2.2-TI2V-5B --max-workers <span style="color:#ae81ff">1</span> </span></span><span style="display:flex;"><span>HF_HUB_DISABLE_XET<span style="color:#f92672">=</span><span style="color:#ae81ff">1</span> hf download Wan-AI/Wan2.2-T2V-A14B --local-dir Wan2.2-T2V-A14B --max-workers <span style="color:#ae81ff">1</span> </span></span></code></pre></div><h4 id="convert">Convert</h4> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><code class="language-shell" data-lang="shell"><span style="display:flex;"><span>~/.local/share/uv/tools/mlx-video/bin/python -m mlx_video.models.wan_2.convert <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --checkpoint-dir models/Wan-AI/Wan2.2-TI2V-5B <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --output-dir models/Wan-AI/Wan2.2-TI2V-5B-mlx </span></span><span style="display:flex;"><span>~/.local/share/uv/tools/mlx-video/bin/python -m mlx_video.models.wan_2.convert <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --checkpoint-dir models/Wan-AI/Wan2.2-T2V-A14B <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --output-dir models/Wan-AI/Wan2.2-T2V-A14B-mlx <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --dtype bfloat16 </span></span></code></pre></div><h4 id="run">Run</h4> <h5 id="wan22-ti2v-5b">Wan2.2-TI2V-5B</h5> <h6 id="t2v">t2v</h6> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><code class="language-shell" data-lang="shell"><span style="display:flex;"><span>mlx_video.wan_2.generate <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --model-dir models/Wan-AI/Wan2.2-TI2V-5B-mlx <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --prompt <span style="color:#e6db74">&#39;A cinematic aerial shot of the Great Wall of China at sunrise&#39;</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --width <span style="color:#ae81ff">1280</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --height <span style="color:#ae81ff">704</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --num-frames <span style="color:#ae81ff">81</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --steps <span style="color:#ae81ff">40</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --guide-scale 5.0 <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --shift 5.0 <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --scheduler unipc <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --seed <span style="color:#ae81ff">3478</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --output-path video/great_wall1.mp4 </span></span></code></pre></div><h6 id="i2v">i2v</h6> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><code class="language-shell" data-lang="shell"><span style="display:flex;"><span>mlx_video.wan_2.generate <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --model-dir models/Wan-AI/Wan2.2-TI2V-5B-mlx <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --prompt <span style="color:#e6db74">&#39;A cat is boating on the river&#39;</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --image models/Wan-AI/Wan2.2-TI2V-5B/examples/i2v_input.JPG <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --width <span style="color:#ae81ff">1280</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --height <span style="color:#ae81ff">704</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --num-frames <span style="color:#ae81ff">81</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --steps <span style="color:#ae81ff">40</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --guide-scale 5.0 <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --shift 5.0 <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --scheduler unipc <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --seed <span style="color:#ae81ff">13633</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --output-path video/cat.mp4 </span></span></code></pre></div><h5 id="wan22-t2v-a14b">Wan2.2-T2V-A14B</h5> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><code class="language-shell" data-lang="shell"><span style="display:flex;"><span>mlx_video.wan_2.generate <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --model-dir models/Wan-AI/Wan2.2-T2V-A14B-mlx <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --prompt <span style="color:#e6db74">&#39;A cinematic aerial shot of the Great Wall of China at sunrise&#39;</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --width <span style="color:#ae81ff">1280</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --height <span style="color:#ae81ff">720</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --num-frames <span style="color:#ae81ff">81</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --steps <span style="color:#ae81ff">40</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --guide-scale 3.0,4.0 <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --shift 12.0 <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --scheduler unipc <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --seed <span style="color:#ae81ff">11846</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --output-path video/great_wall2.mp4 </span></span></code></pre></div><h4 id="upload">Upload</h4> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><code class="language-shell" data-lang="shell"><span style="display:flex;"><span>HF_HUB_DOWNLOAD_TIMEOUT<span style="color:#f92672">=</span><span style="color:#ae81ff">120</span> HF_HUB_DISABLE_XET<span style="color:#f92672">=</span><span style="color:#ae81ff">1</span> hf upload-large-folder rickylin20260522/Wan2.2-TI2V-5B-mlx models/Wan-AI/Wan2.2-TI2V-5B-mlx --repo-type model --num-workers <span style="color:#ae81ff">2</span> </span></span></code></pre></div><hr> <h3 id="audio">Audio</h3> <ul> <li><a href="https://huggingface.co/MediaTek-Research" target="_blank" rel="noopener">https://huggingface.co/MediaTek-Research</a></li> </ul> <h4 id="install-1">Install</h4> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><code class="language-shell" data-lang="shell"><span style="display:flex;"><span>brew install ffmpeg </span></span><span style="display:flex;"><span>uv pip install mlx mlx-audio </span></span></code></pre></div><h4 id="convert-1">Convert</h4> <h5 id="model">model</h5> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><code class="language-shell" data-lang="shell"><span style="display:flex;"><span>python -m mlx_audio.convert <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --hf-path models/MediaTek-Research/Breeze-ASR-25 <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --mlx-path models/MediaTek-Research/Breeze-ASR-25-mlx <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --dtype bfloat16 <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> --model-domain stt </span></span></code></pre></div><h5 id="audio-1">audio</h5> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><code class="language-shell" data-lang="shell"><span style="display:flex;"><span>ffmpeg <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> -i ~/Downloads/talk.m4a <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> -ar <span style="color:#ae81ff">16000</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> -ac <span style="color:#ae81ff">1</span> <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> -c:a pcm_s16le <span style="color:#ae81ff">\ </span></span></span><span style="display:flex;"><span><span style="color:#ae81ff"></span> ~/Downloads/talk.wav </span></span></code></pre></div><h4 id="run-1">Run</h4> <p><strong>STT</strong></p>https://59d713fc.rickylin.pages.dev/posts/2023/20230927-chromeos/Wed, 27 Sep 2023 09:36:00 +0800https://59d713fc.rickylin.pages.dev/posts/2023/20230927-chromeos/<ul> <li><a href="https://www.youtube.com/watch?v=Nc3bTBA4meo" target="_blank" rel="noopener">老電腦別丟掉！安裝 Google 免費作業系統安裝教學（Chrome OS Flex /CloudReady）</a></li> </ul> <div style="position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden;"> <iframe allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" loading="eager" referrerpolicy="strict-origin-when-cross-origin" src="https://www.youtube.com/embed/Nc3bTBA4meo?autoplay=0&amp;controls=1&amp;end=0&amp;loop=0&amp;mute=0&amp;start=0" style="position: absolute; top: 0; left: 0; width: 100%; height: 100%; border:0;" title="YouTube video"></iframe> </div> <div style="margin-top: rem;"></div>https://59d713fc.rickylin.pages.dev/posts/2023/20230927-ai/Wed, 27 Sep 2023 09:36:00 +0800https://59d713fc.rickylin.pages.dev/posts/2023/20230927-ai/<ul> <li><a href="https://www.youtube.com/watch?v=M3z6gI1AEns" target="_blank" rel="noopener">https://www.youtube.com/watch?v=M3z6gI1AEns</a></li> </ul> <div style="position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden;"> <iframe allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" loading="eager" referrerpolicy="strict-origin-when-cross-origin" src="https://www.youtube.com/embed/M3z6gI1AEns?autoplay=0&amp;controls=1&amp;end=0&amp;loop=0&amp;mute=0&amp;start=0" style="position: absolute; top: 0; left: 0; width: 100%; height: 100%; border:0;" title="YouTube video"></iframe> </div> <div style="margin-top: rem;"></div>