From b3a417f48e1f300a4426f74fce089662e835333a Mon Sep 17 00:00:00 2001 From: Prajwal Date: Tue, 15 Sep 2020 12:07:04 +0530 Subject: [PATCH] Add loss info --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index c8db1d5..6593708 100644 --- a/README.md +++ b/README.md @@ -109,6 +109,7 @@ Training on other datasets might require modifications to the code. Please read - You must train the expert discriminator for your own dataset before training Wav2Lip. - If it is your own dataset downloaded from the web, in most cases, needs to be sync-corrected. - Be mindful of the FPS of the videos of your dataset. Changes to FPS would need significant code changes. +- The expert discriminator's eval loss should go down to ~0.25 and the Wav2Lip eval sync loss should go down to ~0.2 to get good results. When raising an issue on this topic, please let us know that you are aware of all these points.