upload code

2-应用/2.10-语音处理/2.10.3-DeepSpeech2端到端中文语音识别模型实现/DeepSpeech.ipynb

-{
+{
@@ -2,6 +2,9 @@
   "cells": [
     {
       "cell_type": "markdown",
+      "metadata": {
+        "id": "N9mSZSK7fpK_"
+      },
       "source": [
         "# 开始之前\n",
         "\n",
@@ -14,10 +17,7 @@
         "export CUDA_HOME=\"/usr/local/cuda\"\n",
         "cd ../pytorch_binding && python setup.py install\n",
         "```"
-      ],
-      "metadata": {
-        "id": "N9mSZSK7fpK_"
-      }
+      ]
     },
     {
       "cell_type": "code",
@@ -1566,7 +1566,7 @@
       "source": [
         "### Prepare datasets\n",
         "- 使用了 an4 音频数据集，该数据集相对较小且易于进行实验。\n",
-        "- 地址：http://www.speech.cs.cmu.edu/databases/an4/\n",
+        "- 地址：https://github.com/cmusphinx/an4\n",
         "\n",
         "- 输入是 wav 文件，\n",
         "- 输出是包含文本内容的文本文件。"
@@ -2811,6 +2811,12 @@
     }
   ],
   "metadata": {
+    "accelerator": "GPU",
+    "colab": {
+      "gpuType": "L4",
+      "machine_shape": "hm",
+      "provenance": []
+    },
     "kernelspec": {
       "display_name": "Python 3",
       "name": "python3"
@@ -2826,13 +2832,7 @@
       "nbconvert_exporter": "python",
       "pygments_lexer": "ipython3",
       "version": "3.7.7"
-    },
-    "colab": {
-      "provenance": [],
-      "machine_shape": "hm",
-      "gpuType": "L4"
-    },
-    "accelerator": "GPU"
+    }
   },
   "nbformat": 4,
   "nbformat_minor": 0