| name: "MultiMaskMultiSourceBandSplitRNN" | |
| audio: | |
| chunk_size: 264600 | |
| num_channels: 2 | |
| sample_rate: 44100 | |
| min_mean_abs: 0.001 | |
| model: | |
| in_channel: 1 | |
| stems: ['speech', 'music', 'effects'] | |
| band_specs: "musical" | |
| n_bands: 64 | |
| fs: 44100 | |
| require_no_overlap: false | |
| require_no_gap: true | |
| normalize_channel_independently: false | |
| treat_channel_as_feature: true | |
| n_sqm_modules: 8 | |
| emb_dim: 128 | |
| rnn_dim: 256 | |
| bidirectional: true | |
| rnn_type: "GRU" | |
| mlp_dim: 512 | |
| hidden_activation: "Tanh" | |
| hidden_activation_kwargs: null | |
| complex_mask: true | |
| n_fft: 2048 | |
| win_length: 2048 | |
| hop_length: 512 | |
| window_fn: "hann_window" | |
| wkwargs: null | |
| power: null | |
| center: true | |
| normalized: true | |
| pad_mode: "constant" | |
| onesided: true | |
| training: | |
| batch_size: 4 | |
| gradient_accumulation_steps: 4 | |
| grad_clip: 0 | |
| instruments: | |
| - speech | |
| - music | |
| - effects | |
| lr: 9.0e-05 | |
| patience: 2 | |
| reduce_factor: 0.95 | |
| target_instrument: null | |
| num_epochs: 1000 | |
| num_steps: 1000 | |
| q: 0.95 | |
| coarse_loss_clip: true | |
| ema_momentum: 0.999 | |
| optimizer: adam | |
| other_fix: true # it's needed for checking on multisong dataset if other is actually instrumental | |
| use_amp: true # enable or disable usage of mixed precision (float16) - usually it must be true | |
| augmentations: | |
| enable: true # enable or disable all augmentations (to fast disable if needed) | |
| loudness: true # randomly change loudness of each stem on the range (loudness_min; loudness_max) | |
| loudness_min: 0.5 | |
| loudness_max: 1.5 | |
| mixup: true # mix several stems of same type with some probability (only works for dataset types: 1, 2, 3) | |
| mixup_probs: !!python/tuple # 2 additional stems of the same type (1st with prob 0.2, 2nd with prob 0.02) | |
| - 0.2 | |
| - 0.02 | |
| mixup_loudness_min: 0.5 | |
| mixup_loudness_max: 1.5 | |
| all: | |
| channel_shuffle: 0.5 # Set 0 or lower to disable | |
| random_inverse: 0.1 # inverse track (better lower probability) | |
| random_polarity: 0.5 # polarity change (multiply waveform to -1) | |
| inference: | |
| batch_size: 1 | |
| dim_t: 256 | |
| num_overlap: 4 |