| { | |
| "model_type": "autoencoder", | |
| "sample_rate": 44100, | |
| "model": { | |
| "encoder": { | |
| "type": "oobleck", | |
| "config": { | |
| "in_channels": 4, | |
| "channels": 128, | |
| "c_mults": [ | |
| 1, | |
| 2, | |
| 4, | |
| 8, | |
| 16 | |
| ], | |
| "strides": [ | |
| 2, | |
| 2, | |
| 4, | |
| 8, | |
| 8 | |
| ], | |
| "latent_dim": 128, | |
| "use_snake": true | |
| } | |
| }, | |
| "decoder": { | |
| "type": "oobleck", | |
| "config": { | |
| "out_channels": 4, | |
| "channels": 128, | |
| "c_mults": [ | |
| 1, | |
| 2, | |
| 4, | |
| 8, | |
| 16 | |
| ], | |
| "strides": [ | |
| 2, | |
| 2, | |
| 4, | |
| 8, | |
| 8 | |
| ], | |
| "latent_dim": 64, | |
| "use_snake": true, | |
| "final_tanh": false | |
| } | |
| }, | |
| "bottleneck": { | |
| "type": "grouped_vae", | |
| "config": { | |
| "n_w": 40 | |
| } | |
| }, | |
| "latent_dim": 64, | |
| "downsampling_ratio": 1024, | |
| "io_channels": 4 | |
| } | |
| } | |