Datasets:
Update README_EN.md
Browse files- README_EN.md +5 -4
README_EN.md
CHANGED
@@ -109,7 +109,7 @@ Below are preview audios. **Scroll up, down, left, and right** to see them all.
|
|
109 |
</table>
|
110 |
</div>
|
111 |
|
112 |
-
Key Parameters
|
113 |
|
114 |
audio duration: total duration of the training dataset </br>
|
115 |
|
@@ -119,10 +119,11 @@ Others: </br>
|
|
119 |
|
120 |
batch_size = number of audio segments trained in one step </br>
|
121 |
|
122 |
-
segments = the number of segments that the audio is split into ,step = segments * epoch / batch_size, which is where the numbers in the model file name come from
|
123 |
|
124 |
-
Using the example of "Wanderer" (a character name)
|
125 |
-
|
|
|
126 |
shows the result of training on a 10-minute pure voice audio. The model achieved good performance at around
|
127 |
2800 epochs (10,000 steps), and the actual model used was trained for 5571 epochs (19,500 steps), with
|
128 |
slight differences between the trained voice and the original voice. Please refer to the preview audio above.
|
|
|
109 |
</table>
|
110 |
</div>
|
111 |
|
112 |
+
Key Parameters:</br>
|
113 |
|
114 |
audio duration: total duration of the training dataset </br>
|
115 |
|
|
|
119 |
|
120 |
batch_size = number of audio segments trained in one step </br>
|
121 |
|
122 |
+
segments = the number of segments that the audio is split into ,step = segments * epoch / batch_size, which is where the numbers in the model file name come from</br>
|
123 |
|
124 |
+
Using the example of "Wanderer" (a character name):</br>
|
125 |
+
Loss Function Graph: pay attention to step and loss5, for example:<br>
|
126 |
+
As a difficult test, all the original audios are high-pitched female voices, and this graph
|
127 |
shows the result of training on a 10-minute pure voice audio. The model achieved good performance at around
|
128 |
2800 epochs (10,000 steps), and the actual model used was trained for 5571 epochs (19,500 steps), with
|
129 |
slight differences between the trained voice and the original voice. Please refer to the preview audio above.
|