--- license: cc-by-nc-sa-4.0 --- You can try listening to the performance of this model [here](./example_audio) Description: The model is used to separate aspiration, which will be useful for mixing to some mixrs.
Instruments: aspiration, other
Dataset: My own datasets(117 songs for training and 17 songs for validation).
Metrics: Based on the SDR of 17 songs for validation.
[aspiration_mel_band_roformer_less_aggr_sdr_18.1201.ckpt](./aspiration_mel_band_roformer_less_aggr_sdr_18.1201.ckpt)
Epoch: 27
Instr SDR aspiration: 9.0704
Instr SDR other: 27.1699
SDR Avg: 18.1201
[aspiration_mel_band_roformer_sdr_18.9845.ckpt](./aspiration_mel_band_roformer_sdr_18.9845.ckpt)
Epoch: 123
Instr SDR aspiration: 9.8554
Instr SDR other: 28.1136
SDR Avg: 18.9845
training logs: ![image](./training_logs.png)