fix some wav2lip docs errors (#177)

* fix wav2lip bugs (#175)

* fix wav2lip docs error (#176)

docs/en_US/tutorials/wav2lip.md

@@ -11,7 +11,7 @@ Runing the following command to complete the lip-syning task. The output is the
 
 ```
 cd applications
-python tools/wav2lip.py --face ../../imgs/mona7s.mp4 --audio ../../imgs/guangquan.m4a --outfile pp_guangquan_mona7s.mp4
+python tools/wav2lip.py --face ../docs/imgs/mona7s.mp4 --audio ../docs/imgs/guangquan.m4a --outfile pp_guangquan_mona7s.mp4
 ```
 
 **params:**
@@ -36,7 +36,7 @@ Place the LRS2 filelists(train, val, test) `.txt` files in the `filelists/` fold
 - For single GPU:
 ```
 export CUDA_VISIBLE_DEVICES=0
-python tools/main.py --confit-file configs/wav2lip.yaml
+python tools/main.py --config-file configs/wav2lip.yaml
 ```
 
 - For multiple GPUs:

docs/zh_CN/tutorials/wav2lip.md

@@ -13,10 +13,10 @@ Wav2Lip实现的是视频人物根据输入音频生成与语音同步的人物
 
 ```
 cd applications
-python tools/wav2lip.py --face ../../imgs/mona7s.mp4 --audio ../../imgs/guangquan.m4a --outfile pp_guangquan_mona7s.mp4
+python tools/wav2lip.py --face ../docs/imgs/mona7s.mp4 --audio ../docs/imgs/guangquan.m4a --outfile pp_guangquan_mona7s.mp4
 ```
 **参数说明:**
-- face: 原始视频，视频中都人物都唇形将根据音频进行唇形合成，以和音频同步
+- face: 视频或图片，视频或图片中的人物唇形将根据音频进行唇形合成，以和音频同步
 - audio: 驱动唇形合成的音频，视频中的人物将根据此音频进行唇形合成
 
 ### 2.2 训练
@@ -38,7 +38,7 @@ preprocessed_root (lrs2_preprocessed)
 - GPU单卡训练:
 ```
 export CUDA_VISIBLE_DEVICES=0
-python tools/main.py --confit-file configs/wav2lip.yaml
+python tools/main.py --config-file configs/wav2lip.yaml
 ```
 
 - GPU多卡训练:

ppgan/datasets/wav2lip_dataset.py

@@ -34,7 +34,10 @@ def get_image_list(data_root, split):
         for line in f:
             line = line.strip()
             if ' ' in line: line = line.split()[0]
-            filelist.append(os.path.join(data_root, line))
+            video_path = os.path.join(data_root, line)
+            assert os.path.exists(video_path), '{} is not found'.format(
+                video_path)
+            filelist.append(video_path)
 
     return filelist
 
@@ -179,14 +182,6 @@ class Wav2LipDataset(paddle.io.Dataset):
             mel = np.transpose(mel)
             mel = np.expand_dims(mel, 0)
             indiv_mels = np.expand_dims(indiv_mels, 1)
-            #np.random.seed(200)
-            #x = np.random.rand(*x.shape).astype('float32')
-            #np.random.seed(200)
-            #mel = np.random.rand(*mel.shape)
-            #np.random.seed(200)
-            #indiv_mels = np.random.rand(*indiv_mels.shape)
-            #np.random.seed(200)
-            #y = np.random.rand(*y.shape)
 
             return {
                 'x': x,