small fixes

configs_backup/single_stem_models/config_apollo.yaml

@@ -29,5 +29,5 @@ augmentations:
   enable: false # enable or disable all augmentations (to fast disable if needed)
 
 inference:
-  batch_size: 4
-  num_overlap: 4
+  batch_size: 1
+  num_overlap: 2

data_backup/msst_model_map.json

@@ -136,19 +136,19 @@
     },
     {
       "name": "Apollo_LQ_MP3_restoration.ckpt",
-      "config_path": "configs_backup/single_stem_models/config_apollo.yaml",
+      "config_path": "configs/single_stem_models/config_apollo.yaml",
       "model_type": "apollo",
       "link": "https://huggingface.co/Sucial/Music_Source_Sepetration_Models/resolve/main/Apollo_LQ_MP3_restoration.ckpt"
     },
     {
       "name": "aspiration_mel_band_roformer_sdr_18.9845.ckpt",
-      "config_path": "configs_backup/single_stem_models/config_aspiration_mel_band_roformer.yaml",
+      "config_path": "configs/single_stem_models/config_aspiration_mel_band_roformer.yaml",
       "model_type": "mel_band_roformer",
       "link": "https://huggingface.co/Sucial/Aspiration_Mel_Band_Roformer/resolve/main/aspiration_mel_band_roformer_sdr_18.9845.ckpt"
     },
     {
       "name": "aspiration_mel_band_roformer_less_aggr_sdr_18.1201.ckpt",
-      "config_path": "configs_backup/single_stem_models/config_aspiration_mel_band_roformer.yaml",
+      "config_path": "configs/single_stem_models/config_aspiration_mel_band_roformer.yaml",
       "model_type": "mel_band_roformer",
       "link": "https://huggingface.co/Sucial/Aspiration_Mel_Band_Roformer/resolve/main/aspiration_mel_band_roformer_less_aggr_sdr_18.1201.ckpt"
     }

models/look2hear/apollo.py

@@ -313,6 +313,7 @@ def forward(self, input):
             this_RI = self.output[i](feature[:, i]).view(B * nch, 2, self.band_width[i], -1)
             est_spec.append(torch.complex(this_RI[:, 0], this_RI[:, 1]))
         est_spec = torch.cat(est_spec, 1)
+        est_spec = est_spec.to(dtype=torch.complex64)
         output = torch.istft(est_spec, n_fft=self.win, hop_length=self.stride,
                              window=torch.hann_window(self.win).to(input.device), length=nsample).view(B, nch, -1)
 

webUI.py

@@ -468,11 +468,11 @@ def save_config(selected_model, batch_size, dim_t, num_overlap, normalize):
     _, config_path, _, _ = get_msst_model(selected_model)
     config = load_configs(config_path)
     if config.inference.get('batch_size'):
-        config.inference['batch_size'] = int(batch_size) if batch_size.isdigit() else None
+        config.inference['batch_size'] = int(batch_size)
     if config.inference.get('dim_t'):
-        config.inference['dim_t'] = int(dim_t) if dim_t.isdigit() else None
+        config.inference['dim_t'] = int(dim_t)
     if config.inference.get('num_overlap'):
-        config.inference['num_overlap'] = int(num_overlap) if num_overlap.isdigit() else None
+        config.inference['num_overlap'] = int(num_overlap)
     if config.inference.get('normalize'):
         config.inference['normalize'] = normalize
     save_configs(config, config_path)