voice_audio_multi/model_specs/mossformer2.json
Fabio c36510c21e
Some checks failed
Linux Build / Linux cpu (push) Has been cancelled
Linux Build / Linux cpu-clang (push) Has been cancelled
Linux Build / Linux vulkan (push) Has been cancelled
mac-build / macOS CPU (push) Has been cancelled
Nix Build / Nix (cpu) (push) Has been cancelled
Nix Build / Nix (vulkan) (push) Has been cancelled
windows-build / Windows CPU (push) Has been cancelled
Initial import: audio.cpp with Orange Pi documentation
2026-11-04 19:05:22 +01:00

118 lines
2.8 KiB
JSON

{
"schema_version": 1,
"family": "mossformer2",
"display_name": "MossFormer2",
"description": "MossFormer2 overlapping speech separation at 16 kHz.",
"category": "audio_tools",
"status": "supported",
"tasks": [
"sep"
],
"modes": [
"offline"
],
"languages": [
"language_agnostic"
],
"capabilities": {
"sep": [
"stems"
]
},
"runtime": {
"tags": [
"gguf"
]
},
"options": {
"request": [
{
"name": "audio_chunk_duration_sec",
"type": "float",
"description": "Separation window duration; zero selects whole-file inference.",
"min": 0,
"default": 2,
"required": false
},
{
"name": "audio_chunk_overlap_sec",
"type": "float",
"description": "Overlap between adjacent audio chunks, in seconds; must be smaller than audio_chunk_duration_sec when chunking is enabled. Recommended: 25% of audio_chunk_duration_sec, matching the official Python settings. Overlapping edges are trimmed rather than averaged. Ignored for whole-file inference.",
"min": 0,
"default": 0.5,
"required": false
},
{
"name": "normalize_output",
"type": "bool",
"description": "Match each trimmed track RMS to the input as in ClearVoice's NumPy interface.",
"default": true,
"required": false
}
],
"session": [],
"load": []
},
"package_defaults": {
"download": {
"kind": "huggingface_snapshot",
"repo": "audio-cpp/audio.cpp-gguf",
"revision": "main",
"gated": false
}
},
"packages": [
{
"id": "mossformer2_ss_16k_f32",
"display_name": "MossFormer2 SS 16K F32",
"default": true,
"format": "gguf",
"precision": "f32",
"target_directory": "MossFormer2-GGUF",
"files": [
"MossFormer2-GGUF/mossformer2-ss-16k-f32.gguf"
],
"strip_prefix": "MossFormer2-GGUF"
},
{
"id": "mossformer2_ss_16k_f16",
"display_name": "MossFormer2 SS 16K F16",
"default": false,
"format": "gguf",
"precision": "f16",
"target_directory": "MossFormer2-GGUF",
"files": [
"MossFormer2-GGUF/mossformer2-ss-16k-f16.gguf"
],
"strip_prefix": "MossFormer2-GGUF"
}
],
"dependencies": [],
"ui": {
"tags": [
"GGUF"
],
"docs": [
"docs/models/mossformer2.md"
],
"recommended_package": "mossformer2_ss_16k_f32"
},
"sources": [
{
"format": "gguf",
"roots": {
"model": ".",
"weights": "$gguf"
},
"files": {
"config": "model:config.json"
},
"tensors": {
"weights": {
"source": "weights:",
"prefix": "weights"
}
}
}
]
}