search

NextAudioGen / ultimatevocalremover_api

API for a Vocal Remover that uses Deep Neural Networks.
MIT License
91 stars 10 forks source link

Should the UVR5 HP5_only_main_vocal model work? #9

Open ybwai opened 7 months ago

ybwai commented 7 months ago

models.json

  "vr_network": {
    "1_HP-UVR": {
      "model_path": [
        "https://github.com/TRvlvr/model_repo/releases/download/all_public_uvr_models/1_HP-UVR.pth"
      ]
    },
    "HP5_only_main_vocal": {
      "model_path": [
        "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/uvr5_weights/HP5_only_main_vocal.pth"
      ]
    }
  },

code:


vrNetwork = models.VrNetwork(
     name="HP5_only_main_vocal",
     other_metadata={},
     device=device,
     logger=None,
 )

res= vrNetwork(name)

gives error:


  File "/Users/dan/Workspaces/myproject2/worker/uvr/models.py", line 175, in __init__
    model_run, mp, is_vr_51_model, stems  = vr_api.load_model(self.model_path, device)
  File "/Users/dan/Workspaces/myproject2/worker/uvr/models_dir/vr_network/vr_interface.py", line 205, in load_model
    model_data = MODELS_DATA[model_hash]
KeyError: '0b633752cb273e9758813724ee8afa13'

I can see it downloaded the weights into the relevant folder.