This is the model I personally use, I use it at a low res but you can reuse this encoder and decoder file to make a model of your desired res, these are just the encoder and decoder files not the entire model.
https://mega.nz/file/fJEFUDra#uF0Kch7nunHxM_mp3jbNwm0lz6NjtGZ81I6JkSzKNxI
This model has 2Million iteration pretraining on RTM dataset using mve fork of deepfacelab and then further more 2-3 million training, It works the best with mve fork of deepfacelab, though it would work with almost any fork, This is the fastest converging model I have used and The best I have
https://mega.nz/file/fJEFUDra#uF0Kch7nunHxM_mp3jbNwm0lz6NjtGZ81I6JkSzKNxI
This model has 2Million iteration pretraining on RTM dataset using mve fork of deepfacelab and then further more 2-3 million training, It works the best with mve fork of deepfacelab, though it would work with almost any fork, This is the fastest converging model I have used and The best I have
