diff --git a/README.md b/README.md index 3eaf7a2..4c1e069 100644 --- a/README.md +++ b/README.md @@ -115,21 +115,31 @@ All models released here will have the values they were trained with appended to ## Models Included +All of the models included in the release were trained on large datasets containing diverse sets of music genres. + **PLEASE NOTE:** Do not change the name of the models provided! The required parameters are specified and appended to the end of the filenames. Here's a list of the models included within the package - -- **v2 AI Engine** - - **Main Models** - - *(list pending)* - - **Stacked Models** - - *(list pending)* - - **v4 AI Engine** - **Main Models** - - *(list pending)* + - **MGM_MAIN_v4_sr44100_hl512_w512_nf2048.pth** - This is the main model that does an excellent job removing vocals from most tracks. + - **MGM_LOWEND_A_v4_sr32000_hl512_w512_nf2048.pth** - This model focuses a bit more on removing vocals from lower frequencies. + - **MGM_LOWEND_B_v4_sr33075_hl384_w512_nf2048.pth** - This is also a model that focuses on lower end frequencies, but trained with different parameters. + - **MGM_HIGHEND_v4_sr44100_hl1024_nf2048.pth** - This model slightly focuses a bit more on higher end frequencies. + - **MODEL_BVKARAOKE_by_aufr33_v4_sr33075_hl384_w512_nf1536.pth** - This is a beta model that removes main vocals while leaving background vocals intact. - **Stacked Models** - - *(list pending)* + - **StackedMGM_MM_v4_sr44100_hl512_w512_nf2048.pth** - This is a strong vocal artifact removal model. This model was made to run with *'MGM_MAIN_v4_sr44100_hl512_w512_nf2048.pth'*. However, any combination may yield the desired results. + - **StackedMGM_MLA_v4_sr32000_hl512_w512_nf2048.pth** - This is a strong vocal artifact removal model. This model was made to run with *'MGM_MAIN_v4_sr44100_hl512_w512_nf2048.pth'*. However, any combination may yield a desired results. + - **StackedMGM_LL_v4_sr32000_hl512_w512_nf2048.pth** - This is a strong vocal artifact removal model. This model was made to run with *'MGM_LOWEND_A_v4_sr32000_hl512_w512_nf2048.pth'*. However, any combination may yield a desired results. + +- **v2 AI Engine** + - **Main Models** + - **Multi_Genre_Model_v2_sr44100_hl1024_w512.pth** - This model yields excellent results for most tracks processed through it. + - **Stacked Models** + - **StackedRegA_v2_sr44100_hl1024_w512.pth** - This is a standard vocal artifact removal model. + - **StackedRegB_v2_sr44100_hl1024_w512.pth** - This is a standard vocal artifact removal model. + - **StackedArg_v2_sr44100_hl1024_w512.pth** - This model removes vocal artifacts a bit more aggressively, but may greatly degrade the audio quality of the output audio. A special thank you to aufr33 for helping me expand the dataset used to train these models and for the helpful training tips.