v2 64

README.md

@@ -13,7 +13,7 @@ The framework of grad-svc-v1
 
 ![grad_svc_v2](./assets/grad_svc_v2.jpg)
 
-The framework of grad-svc-v2, encoder:768->512, diffusion:64->96
+The framework of grad-svc-v2-64, encoder:768->512, diffusion:64
 
 https://github.com/PlayVoice/Grad-SVC/assets/16432329/f9b66af7-b5b5-4efb-b73d-adb0dc84a0ae
 
@@ -34,6 +34,10 @@ https://github.com/PlayVoice/Grad-SVC/assets/16432329/f9b66af7-b5b5-4efb-b73d-ad
 
 6. Integrated Fast Maximum Likelihood Sampling Scheme 
 
+7. Low GPU memery required for train
+
+    `6.2GB for batch_size=8`
+
 ## Setup Environment
 1. Install project dependencies
 
@@ -47,13 +51,17 @@ https://github.com/PlayVoice/Grad-SVC/assets/16432329/f9b66af7-b5b5-4efb-b73d-ad
 
 4. Download pretrained [nsf_bigvgan_pretrain_32K.pth](https://github.com/PlayVoice/NSF-BigVGAN/releases/augment), and put it into `bigvgan_pretrain/`.
 
-5. Download pretrain model [gvc.pretrain.pth](https://github.com/PlayVoice/Grad-SVC/releases/tag/20230920), and put it into `grad_pretrain/`.
+	**Performance Bottleneck: Generator and Discriminator are 116Mb, but Generator is only 22Mb**
+
+	**系统性能瓶颈：生成器和判别器一共116M，而生成器只有22M**
+
+5. Download pretrain model [gvc.pretrain.pth](https://github.com/PlayVoice/Grad-SVC/releases/tag/20230922), and put it into `grad_pretrain/`.
     ```
     python gvc_inference.py --model ./grad_pretrain/gvc.pretrain.pth --spk ./configs/singers/singer0001.npy --wave test.wav
     ```
 
     For this pretrain model, `temperature` is set `temperature=1.015` in `gvc_inference.py` to get good result.
-   
+
 ## Dataset preparation
 Put the dataset into the `data_raw` directory following the structure below.
 ```