feat: 完整中文翻译 maths-cs-ai-compendium（数学·计算机科学·AI 知识大全）
翻译自英文原版 maths-cs-ai-compendium，共 20 章全部完成。 第01章 向量 | 第02章 矩阵 | 第03章 微积分 第04章 统计学 | 第05章 概率论 | 第06章 机器学习 第07章 计算语言学 | 第08章 计算机视觉 | 第09章 音频与语音 第10章 多模态学习 | 第11章 自主系统 | 第12章 图神经网络 第13章 计算与操作系统 | 第14章 数据结构与算法 第15章 生产级软件工程 | 第16章 SIMD与GPU编程 第17章 AI推理 | 第18章 ML系统设计 第19章 应用人工智能 | 第20章 前沿人工智能 翻译说明： - 所有数学公式 $...$ / $$...$$、代码块、图片引用完整保留 - mkdocs.yml 配置中文导航 + language: zh - README.md 已翻译为中文（兼 docs/index.md） - docs/ 目录包含指向各章文件的 symlink - 约 29,000 行中文内容，排除 .cache/ 构建缓存
This commit is contained in:
@@ -0,0 +1,127 @@
|
||||
<svg xmlns="http://www.w3.org/2000/svg" width="700" height="290">
|
||||
<defs>
|
||||
<marker id="arr3" viewBox="0 0 10 6" refX="10" refY="3" markerWidth="8" markerHeight="6" orient="auto">
|
||||
<polygon points="0,0 10,3 0,6" fill="#666"/>
|
||||
</marker>
|
||||
</defs>
|
||||
|
||||
<!-- Title -->
|
||||
<text x="350" y="22" text-anchor="middle" font-family="Arial, sans-serif" font-size="14" font-weight="bold" fill="#333">HiFi-GAN Generator Structure</text>
|
||||
|
||||
<!-- ===== INPUT: Mel Spectrogram (narrow, tall) ===== -->
|
||||
<rect x="20" y="55" width="50" height="100" rx="6" fill="#9b59b6" fill-opacity="0.15" stroke="#9b59b6" stroke-width="1.5"/>
|
||||
<text x="45" y="98" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" font-weight="bold" fill="#9b59b6">Mel</text>
|
||||
<text x="45" y="110" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" fill="#9b59b6">Spec</text>
|
||||
<text x="45" y="122" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" fill="#666">80 ch</text>
|
||||
|
||||
<!-- Arrow -->
|
||||
<line x1="70" y1="105" x2="90" y2="105" stroke="#666" stroke-width="1.2" marker-end="url(#arr3)"/>
|
||||
|
||||
<!-- ===== Upsample Block 1 ===== -->
|
||||
<rect x="92" y="60" width="62" height="90" rx="6" fill="#3498db" fill-opacity="0.12" stroke="#3498db" stroke-width="1.5"/>
|
||||
<text x="123" y="82" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" font-weight="bold" fill="#3498db">TransConv</text>
|
||||
<text x="123" y="93" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" fill="#3498db">Upsample</text>
|
||||
<text x="123" y="104" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" fill="#666">8x</text>
|
||||
|
||||
<!-- MRF block 1 -->
|
||||
<rect x="97" y="109" width="52" height="34" rx="4" fill="#e74c3c" fill-opacity="0.12" stroke="#e74c3c" stroke-width="1"/>
|
||||
<text x="123" y="121" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" font-weight="bold" fill="#e74c3c">MRF</text>
|
||||
<text x="123" y="131" text-anchor="middle" font-family="Arial, sans-serif" font-size="6" fill="#666">k=3,7,11</text>
|
||||
|
||||
<!-- Arrow -->
|
||||
<line x1="154" y1="105" x2="176" y2="105" stroke="#666" stroke-width="1.2" marker-end="url(#arr3)"/>
|
||||
|
||||
<!-- ===== Upsample Block 2 ===== -->
|
||||
<rect x="178" y="55" width="72" height="100" rx="6" fill="#3498db" fill-opacity="0.12" stroke="#3498db" stroke-width="1.5"/>
|
||||
<text x="214" y="78" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" font-weight="bold" fill="#3498db">TransConv</text>
|
||||
<text x="214" y="89" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" fill="#3498db">Upsample</text>
|
||||
<text x="214" y="100" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" fill="#666">8x</text>
|
||||
|
||||
<!-- MRF block 2 -->
|
||||
<rect x="185" y="106" width="58" height="40" rx="4" fill="#e74c3c" fill-opacity="0.12" stroke="#e74c3c" stroke-width="1"/>
|
||||
<text x="214" y="120" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" font-weight="bold" fill="#e74c3c">MRF</text>
|
||||
<text x="214" y="130" text-anchor="middle" font-family="Arial, sans-serif" font-size="6" fill="#666">k=3,7,11</text>
|
||||
<text x="214" y="140" text-anchor="middle" font-family="Arial, sans-serif" font-size="6" fill="#666">d=1,3,5</text>
|
||||
|
||||
<!-- Arrow -->
|
||||
<line x1="250" y1="105" x2="274" y2="105" stroke="#666" stroke-width="1.2" marker-end="url(#arr3)"/>
|
||||
|
||||
<!-- ===== Upsample Block 3 ===== -->
|
||||
<rect x="276" y="50" width="82" height="110" rx="6" fill="#3498db" fill-opacity="0.12" stroke="#3498db" stroke-width="1.5"/>
|
||||
<text x="317" y="73" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" font-weight="bold" fill="#3498db">TransConv</text>
|
||||
<text x="317" y="84" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" fill="#3498db">Upsample</text>
|
||||
<text x="317" y="95" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" fill="#666">2x</text>
|
||||
|
||||
<!-- MRF block 3 -->
|
||||
<rect x="284" y="101" width="66" height="50" rx="4" fill="#e74c3c" fill-opacity="0.12" stroke="#e74c3c" stroke-width="1"/>
|
||||
<text x="317" y="116" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" font-weight="bold" fill="#e74c3c">MRF</text>
|
||||
<!-- Parallel kernel paths -->
|
||||
<rect x="289" y="121" width="16" height="12" rx="2" fill="#e74c3c" fill-opacity="0.25" stroke="#e74c3c" stroke-width="0.5"/>
|
||||
<text x="297" y="130" text-anchor="middle" font-family="Arial, sans-serif" font-size="5" fill="#e74c3c">k3</text>
|
||||
<rect x="308" y="121" width="16" height="12" rx="2" fill="#e74c3c" fill-opacity="0.25" stroke="#e74c3c" stroke-width="0.5"/>
|
||||
<text x="316" y="130" text-anchor="middle" font-family="Arial, sans-serif" font-size="5" fill="#e74c3c">k7</text>
|
||||
<rect x="327" y="121" width="18" height="12" rx="2" fill="#e74c3c" fill-opacity="0.25" stroke="#e74c3c" stroke-width="0.5"/>
|
||||
<text x="336" y="130" text-anchor="middle" font-family="Arial, sans-serif" font-size="5" fill="#e74c3c">k11</text>
|
||||
<text x="317" y="146" text-anchor="middle" font-family="Arial, sans-serif" font-size="6" fill="#666">sum outputs</text>
|
||||
|
||||
<!-- Arrow -->
|
||||
<line x1="358" y1="105" x2="382" y2="105" stroke="#666" stroke-width="1.2" marker-end="url(#arr3)"/>
|
||||
|
||||
<!-- ===== Upsample Block 4 ===== -->
|
||||
<rect x="384" y="47" width="90" height="116" rx="6" fill="#3498db" fill-opacity="0.12" stroke="#3498db" stroke-width="1.5"/>
|
||||
<text x="429" y="68" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" font-weight="bold" fill="#3498db">TransConv</text>
|
||||
<text x="429" y="79" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" fill="#3498db">Upsample</text>
|
||||
<text x="429" y="90" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" fill="#666">2x</text>
|
||||
|
||||
<!-- MRF block 4 -->
|
||||
<rect x="392" y="97" width="74" height="56" rx="4" fill="#e74c3c" fill-opacity="0.12" stroke="#e74c3c" stroke-width="1"/>
|
||||
<text x="429" y="112" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" font-weight="bold" fill="#e74c3c">MRF</text>
|
||||
<rect x="397" y="117" width="18" height="12" rx="2" fill="#e74c3c" fill-opacity="0.25" stroke="#e74c3c" stroke-width="0.5"/>
|
||||
<text x="406" y="126" text-anchor="middle" font-family="Arial, sans-serif" font-size="5" fill="#e74c3c">k3</text>
|
||||
<rect x="419" y="117" width="18" height="12" rx="2" fill="#e74c3c" fill-opacity="0.25" stroke="#e74c3c" stroke-width="0.5"/>
|
||||
<text x="428" y="126" text-anchor="middle" font-family="Arial, sans-serif" font-size="5" fill="#e74c3c">k7</text>
|
||||
<rect x="441" y="117" width="18" height="12" rx="2" fill="#e74c3c" fill-opacity="0.25" stroke="#e74c3c" stroke-width="0.5"/>
|
||||
<text x="450" y="126" text-anchor="middle" font-family="Arial, sans-serif" font-size="5" fill="#e74c3c">k11</text>
|
||||
<text x="429" y="146" text-anchor="middle" font-family="Arial, sans-serif" font-size="6" fill="#666">sum outputs</text>
|
||||
|
||||
<!-- Arrow -->
|
||||
<line x1="474" y1="105" x2="498" y2="105" stroke="#666" stroke-width="1.2" marker-end="url(#arr3)"/>
|
||||
|
||||
<!-- ===== Final Conv ===== -->
|
||||
<rect x="500" y="72" width="65" height="66" rx="6" fill="#27ae60" fill-opacity="0.12" stroke="#27ae60" stroke-width="1.5"/>
|
||||
<text x="532" y="96" text-anchor="middle" font-family="Arial, sans-serif" font-size="8" font-weight="bold" fill="#27ae60">Conv1d</text>
|
||||
<text x="532" y="108" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" fill="#666">k=7</text>
|
||||
<text x="532" y="119" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" fill="#27ae60">+ tanh</text>
|
||||
|
||||
<!-- Arrow -->
|
||||
<line x1="565" y1="105" x2="589" y2="105" stroke="#666" stroke-width="1.2" marker-end="url(#arr3)"/>
|
||||
|
||||
<!-- ===== OUTPUT: Waveform (wide, shorter) ===== -->
|
||||
<rect x="591" y="75" width="90" height="60" rx="6" fill="#27ae60" fill-opacity="0.18" stroke="#27ae60" stroke-width="1.5"/>
|
||||
<text x="636" y="100" text-anchor="middle" font-family="Arial, sans-serif" font-size="9" font-weight="bold" fill="#27ae60">Waveform</text>
|
||||
<text x="636" y="113" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" fill="#666">24 kHz audio</text>
|
||||
<text x="636" y="125" text-anchor="middle" font-family="Arial, sans-serif" font-size="7" fill="#666">1-channel</text>
|
||||
|
||||
<!-- Waveform squiggle in output box -->
|
||||
<polyline points="600,95 606,90 612,100 618,88 624,102 630,89 636,98 642,90 648,100 654,92 660,97 666,91 672,96" fill="none" stroke="#27ae60" stroke-width="0.8" stroke-opacity="0.4"/>
|
||||
|
||||
<!-- ===== MRF Explanation ===== -->
|
||||
<rect x="20" y="175" width="310" height="55" rx="6" fill="#e74c3c" fill-opacity="0.06" stroke="#e74c3c" stroke-width="1" stroke-dasharray="3,2"/>
|
||||
<text x="30" y="192" font-family="Arial, sans-serif" font-size="8" font-weight="bold" fill="#e74c3c">Multi-Receptive Field Fusion (MRF):</text>
|
||||
<text x="30" y="204" font-family="Arial, sans-serif" font-size="8" fill="#666">Parallel ResBlocks with different kernel sizes (3, 7, 11) and</text>
|
||||
<text x="30" y="216" font-family="Arial, sans-serif" font-size="8" fill="#666">dilation rates (1, 3, 5). Outputs summed to capture patterns at</text>
|
||||
<text x="30" y="228" font-family="Arial, sans-serif" font-size="8" fill="#666">multiple temporal scales simultaneously.</text>
|
||||
|
||||
<!-- Upsample factor label -->
|
||||
<rect x="345" y="175" width="170" height="55" rx="6" fill="#3498db" fill-opacity="0.06" stroke="#3498db" stroke-width="1" stroke-dasharray="3,2"/>
|
||||
<text x="355" y="192" font-family="Arial, sans-serif" font-size="8" font-weight="bold" fill="#3498db">Transposed Convolution Upsampling:</text>
|
||||
<text x="355" y="204" font-family="Arial, sans-serif" font-size="8" fill="#666">Total upsample factor = 8 x 8 x 2 x 2 = 256</text>
|
||||
<text x="355" y="216" font-family="Arial, sans-serif" font-size="8" fill="#666">Matches hop_size=256 of mel transform.</text>
|
||||
<text x="355" y="228" font-family="Arial, sans-serif" font-size="8" fill="#666">Each stage: LeakyReLU activation.</text>
|
||||
|
||||
<!-- Note box -->
|
||||
<rect x="20" y="245" width="660" height="35" rx="6" fill="#f5f5f5" stroke="#333" stroke-width="1"/>
|
||||
<text x="35" y="260" font-family="Arial, sans-serif" font-size="9" font-weight="bold" fill="#333">Note:</text>
|
||||
<text x="70" y="260" font-family="Arial, sans-serif" font-size="9" fill="#666">Multi-period + multi-scale discriminators train the generator adversarially. MPD captures periodic structures</text>
|
||||
<text x="35" y="273" font-family="Arial, sans-serif" font-size="9" fill="#666">(pitch harmonics) while MSD evaluates audio quality at different time resolutions. Combined with mel loss for stability.</text>
|
||||
</svg>
|
||||
|
After Width: | Height: | Size: 10 KiB |
Reference in New Issue
Block a user