Hugging Face's logo

mlx-community
/
SoulX-Singer-8bit

Text-to-Audio
MLX
Safetensors
apple-silicon
singing-voice-synthesis
singing-voice-conversion
soulx-singer
Model card Files
xet
 Community
SoulX-Singer-8bit
File size: 1,894 Bytes
37a006f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
{
  "format": "soulx-singer-mlx",
  "source": "Soul-AILab/SoulX-Singer",
  "official_code": "https://github.com/Soul-AILab/SoulX-Singer",
  "precision": "8bit",
  "dtype": "bfloat16",
  "components": [
    {
      "component": "svs",
      "source_file": "model.pt",
      "tensor_count": 587,
      "skipped": [],
      "numel": 704471426,
      "weight_files": [
        "model-00001-of-00006.safetensors",
        "model-00002-of-00006.safetensors",
        "model-00003-of-00006.safetensors",
        "model-00004-of-00006.safetensors",
        "model-00005-of-00006.safetensors",
        "model-00006-of-00006.safetensors"
      ]
    },
    {
      "component": "svc",
      "source_file": "model-svc.pt",
      "tensor_count": 544,
      "skipped": [],
      "numel": 698444162,
      "weight_files": [
        "model-00001-of-00006.safetensors",
        "model-00002-of-00006.safetensors",
        "model-00003-of-00006.safetensors",
        "model-00004-of-00006.safetensors",
        "model-00005-of-00006.safetensors",
        "model-00006-of-00006.safetensors"
      ]
    }
  ],
  "runtime": {
    "svs": "SoulXSinger checkpoint, saved as MLX-compatible safetensors",
    "svc": "SoulXSingerSVC checkpoint, saved as MLX-compatible safetensors",
    "decoder": "full audio generation currently uses the official PyTorch bridge"
  },
  "quantization": {
    "bits": 8,
    "group_size": 64,
    "mode": "affine",
    "layout": "each quantized tensor is stored as .qweight/.scales/.biases"
  },
  "quantization_summary": [
    {
      "component": "svs",
      "quantized_tensors": 289,
      "kept_tensors": 298
    },
    {
      "component": "svc",
      "quantized_tensors": 270,
      "kept_tensors": 274
    }
  ],
  "target_repo": "mlx-community/SoulX-Singer-8bit",
  "collection": "https://huggingface.co/collections/mlx-community/soulx-singer-mlx-6a1c0525d0911ea400102840"
}