Stable Audio 3.0
Generate short stereo sound effects from English descriptions locally on your desktop.
Desktop app
Run Stable Audio 3.0 on your computer
Download the desktop app to load this model, provide input and view results locally.
Public desktop downloads are coming soon.Device requirements
If your device falls short
MemoryAdvisory only: close other apps, reduce image/tile/token/batch settings or choose a smaller verified model/precision. CPU can need more RAM than GPU.
GPUIf the selected GPU runtime fails, existing CPU fallback may recover when enabled, but CPU has its own memory/runtime needs.
PlatformUse a compatible OS/architecture/runtime or another model. These estimates do not add hardware detection or block runs.
Benchmark evidence
Saved desktop measurements apply to the recorded hardware and runtime. Fallback cases report the device that actually ran; they do not prove GPU support on the requested device.
Stable Audio 3.0 Small Music
22 configurations testedQuality criteria
No explicit quality gate recorded. Review per-case evidence; full certification is pending.
| Precision | Device | Result | Platform | Inference / load | Task metric | |
|---|---|---|---|---|---|---|
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 733 ms / 3525 ms | 0.147 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 747 ms / 3575 ms | 0.149 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 749 ms / 3525 ms | 0.150 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 763 ms / 3575 ms | 0.153 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 763 ms / 3581 ms | 0.153 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 802 ms / 3581 ms | 0.160 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 865 ms / 4961 ms | 0.173 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 993 ms / 4961 ms | 0.199 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 1205 ms / 3525 ms | 0.241 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 1220 ms / 3575 ms | 0.244 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 3629 ms / 3581 ms | 0.726 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2053 ms / 3098 ms | 0.411 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2057 ms / 3157 ms | 0.411 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2069 ms / 3127 ms | 0.414 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2077 ms / 3157 ms | 0.415 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2079 ms / 3098 ms | 0.416 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2083 ms / 3127 ms | 0.417 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2097 ms / 3098 ms | 0.419 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2098 ms / 3127 ms | 0.420 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2109 ms / 3157 ms | 0.422 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2259 ms / 3145 ms | 0.452 RTF |
FP32 | cpu | Passed | Fast | macos · aarch64 | 2296 ms / 3145 ms | 0.459 RTF |
Test setup & sample details
- Quality gate
- No explicit quality gate recorded. Review the per-case quality evidence; full certification is pending.
- Tested
- 2026-11-07T14:47:12.310944+00:00
- Environment
- { "os": "macos", "arch": "aarch64", "hardware": "Apple M3 Max", "osVersion": "27.2", "runtime": "Stable Audio 3 pinned 790793b / Torch 2.7.1 / Transformers 5.8.0", "interpreter": "App-owned CPython 3.13.18 private probe venv", "concurrentActivity": "Pinned candidate source downloads and tracked background archival service active. CLAP scoring runs between generation cases." }
- Runtime
- stable-audio-3 0.1.0+790793b
- Artifact revision
- 8bdfb53c6b0fbc62113cd571fd37373efe15fa14
- Fixture
{ "file": "model-samples/desktop/stable-audio3/small-music/small-music-piano.txt", "size": 83, "sha1": "bdf829476f6d8979181472b69c8ff1e40d822a37", "sha256": "598e2c9b870380f0e7632de52d63bcaf3ec5343badf49ba10b38b507d057d783", "url": "https://assets.localai.tools/samples/v1/files/598e2c9b870380f0e7632de52d63bcaf3ec5343badf49ba10b38b507d057d783/small-music-piano.txt" }- Rerun when
- Checkpoint/runtime/adapter/config changes; Fixture or quality gate changes; Relevant device/backend behavior
Per-case quality & failure details
[
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.9920029640197754
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.007462271023541689
},
{
"label": "A person giving a speech.",
"score": 0.0003950182581320405
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.00013151219172868878
},
{
"label": "The sound of a dog barking.",
"score": 0.000008267261364380829
}
]
},
"expected": "A slow piano melody.",
"score": 0.9920029640197754,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 3629.268124990631,
"durationMs": 3637.992749994737,
"audioSeconds": 5,
"audioSecondsPerSecond": 1.3776882356998377,
"realTimeFactor": 0.7258536249981262,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.1201879233121872,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3394961408,
"loadMs": 3581.1152499954915,
"totalMs": 7219.107999990229,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.9997418522834778
},
{
"label": "The sound of a dog barking.",
"score": 0.000100907520391047
},
{
"label": "A person giving a speech.",
"score": 0.00008659282320877537
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.00004083924068254419
},
{
"label": "A slow piano melody.",
"score": 0.000029727703804383054
}
]
},
"expected": "Electronic dance music with synthesizers and drums.",
"score": 0.9997418522834778,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 802.0211660041241,
"durationMs": 806.9512079964625,
"audioSeconds": 5,
"audioSecondsPerSecond": 6.234249433729146,
"realTimeFactor": 0.1604042332008248,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.14332476258277893,
"clippedFraction": 0.000029478458049886623,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3394961408,
"loadMs": 3581.1152499954915,
"totalMs": 4388.066457991954,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.9994122982025146
},
{
"label": "A slow piano melody.",
"score": 0.0002746071550063789
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0002629700175020844
},
{
"label": "The sound of a dog barking.",
"score": 0.00003894661131198518
},
{
"label": "A person giving a speech.",
"score": 0.000011159519090142567
}
]
},
"expected": "A solo acoustic guitar playing a melody.",
"score": 0.9994122982025146,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 762.7774169959594,
"durationMs": 766.8402090057498,
"audioSeconds": 5,
"audioSecondsPerSecond": 6.554992175425779,
"realTimeFactor": 0.15255548339919187,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.10740576684474945,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3394961408,
"loadMs": 3581.1152499954915,
"totalMs": 4347.955459001241,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.9920029640197754
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.007462271023541689
},
{
"label": "A person giving a speech.",
"score": 0.0003950182581320405
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.00013151219172868878
},
{
"label": "The sound of a dog barking.",
"score": 0.000008267261364380829
}
]
},
"expected": "A slow piano melody.",
"score": 0.9920029640197754,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 1204.62720800424,
"durationMs": 1210.1769580040127,
"audioSeconds": 5,
"audioSecondsPerSecond": 4.15066168751387,
"realTimeFactor": 0.240925441600848,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.1201879233121872,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3394961408,
"loadMs": 3524.91474999988,
"totalMs": 4735.091708003893,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.9997418522834778
},
{
"label": "The sound of a dog barking.",
"score": 0.000100907520391047
},
{
"label": "A person giving a speech.",
"score": 0.00008659282320877537
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.00004083924068254419
},
{
"label": "A slow piano melody.",
"score": 0.000029727703804383054
}
]
},
"expected": "Electronic dance music with synthesizers and drums.",
"score": 0.9997418522834778,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 732.9417500004638,
"durationMs": 736.5957500005607,
"audioSeconds": 5,
"audioSecondsPerSecond": 6.821824517428344,
"realTimeFactor": 0.14658835000009277,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.14332476258277893,
"clippedFraction": 0.000029478458049886623,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3394961408,
"loadMs": 3524.91474999988,
"totalMs": 4261.5105000004405,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.9994122982025146
},
{
"label": "A slow piano melody.",
"score": 0.0002746071550063789
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0002629700175020844
},
{
"label": "The sound of a dog barking.",
"score": 0.00003894661131198518
},
{
"label": "A person giving a speech.",
"score": 0.000011159519090142567
}
]
},
"expected": "A solo acoustic guitar playing a melody.",
"score": 0.9994122982025146,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 748.9287920034258,
"durationMs": 754.4481250079116,
"audioSeconds": 5,
"audioSecondsPerSecond": 6.676202134818084,
"realTimeFactor": 0.14978575840068514,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.10740576684474945,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3394961408,
"loadMs": 3524.91474999988,
"totalMs": 4279.362875007791,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.9920029640197754
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.007462271023541689
},
{
"label": "A person giving a speech.",
"score": 0.0003950182581320405
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.00013151219172868878
},
{
"label": "The sound of a dog barking.",
"score": 0.000008267261364380829
}
]
},
"expected": "A slow piano melody.",
"score": 0.9920029640197754,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 1220.4867499967804,
"durationMs": 1225.9417499881238,
"audioSeconds": 5,
"audioSecondsPerSecond": 4.096726162748747,
"realTimeFactor": 0.2440973499993561,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.1201879233121872,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3394961408,
"loadMs": 3575.173542005359,
"totalMs": 4801.115291993483,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.9997418522834778
},
{
"label": "The sound of a dog barking.",
"score": 0.000100907520391047
},
{
"label": "A person giving a speech.",
"score": 0.00008659282320877537
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.00004083924068254419
},
{
"label": "A slow piano melody.",
"score": 0.000029727703804383054
}
]
},
"expected": "Electronic dance music with synthesizers and drums.",
"score": 0.9997418522834778,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 762.7447089907946,
"durationMs": 766.9907499948749,
"audioSeconds": 5,
"audioSecondsPerSecond": 6.555273266484689,
"realTimeFactor": 0.15254894179815892,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.14332476258277893,
"clippedFraction": 0.000029478458049886623,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3394961408,
"loadMs": 3575.173542005359,
"totalMs": 4342.164292000234,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.9994122982025146
},
{
"label": "A slow piano melody.",
"score": 0.0002746071550063789
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0002629700175020844
},
{
"label": "The sound of a dog barking.",
"score": 0.00003894661131198518
},
{
"label": "A person giving a speech.",
"score": 0.000011159519090142567
}
]
},
"expected": "A solo acoustic guitar playing a melody.",
"score": 0.9994122982025146,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 746.8642919993727,
"durationMs": 750.9577090095263,
"audioSeconds": 5,
"audioSecondsPerSecond": 6.694656651230288,
"realTimeFactor": 0.14937285839987452,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.10740576684474945,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3394961408,
"loadMs": 3575.173542005359,
"totalMs": 4326.1312510148855,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.9519044756889343
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.04515129700303078
},
{
"label": "A person giving a speech.",
"score": 0.0022107241675257683
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0006932063843123615
},
{
"label": "The sound of a dog barking.",
"score": 0.00004028583862236701
}
]
},
"expected": "A slow piano melody.",
"score": 0.9519044756889343,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 2079.3046670005424,
"durationMs": 2083.5227920033503,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.4046500156288526,
"realTimeFactor": 0.4158609334001085,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.11466941237449646,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"loadMs": 3098.456082996563,
"totalMs": 5181.978874999913,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.9993126392364502
},
{
"label": "A person giving a speech.",
"score": 0.0003860131255351007
},
{
"label": "The sound of a dog barking.",
"score": 0.00019279711705166847
},
{
"label": "A slow piano melody.",
"score": 0.00006709693843731657
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.00004150768290855922
}
]
},
"expected": "Electronic dance music with synthesizers and drums.",
"score": 0.9993126392364502,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 2096.705459000077,
"durationMs": 2099.6870000089984,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.38469355747493,
"realTimeFactor": 0.41934109180001544,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.10930800437927246,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"loadMs": 3098.456082996563,
"totalMs": 5198.1430830055615,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.9953279495239258
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0024406753946095705
},
{
"label": "A slow piano melody.",
"score": 0.0021972625982016325
},
{
"label": "The sound of a dog barking.",
"score": 0.000027037332984036766
},
{
"label": "A person giving a speech.",
"score": 0.000007161980647651944
}
]
},
"expected": "A solo acoustic guitar playing a melody.",
"score": 0.9953279495239258,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 2052.633416999015,
"durationMs": 2055.975708004553,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.4358952546480923,
"realTimeFactor": 0.410526683399803,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.13945604860782623,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"loadMs": 3098.456082996563,
"totalMs": 5154.431791001116,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.9519044756889343
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.04515129700303078
},
{
"label": "A person giving a speech.",
"score": 0.0022107241675257683
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0006932063843123615
},
{
"label": "The sound of a dog barking.",
"score": 0.00004028583862236701
}
]
},
"expected": "A slow piano melody.",
"score": 0.9519044756889343,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 2083.0607079988113,
"durationMs": 2087.451665996923,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.4003141054892643,
"realTimeFactor": 0.41661214159976223,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.11466941237449646,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"loadMs": 3127.0569170010276,
"totalMs": 5214.508582997951,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.9993126392364502
},
{
"label": "A person giving a speech.",
"score": 0.0003860131255351007
},
{
"label": "The sound of a dog barking.",
"score": 0.00019279711705166847
},
{
"label": "A slow piano melody.",
"score": 0.00006709693843731657
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.00004150768290855922
}
]
},
"expected": "Electronic dance music with synthesizers and drums.",
"score": 0.9993126392364502,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 2097.923292007181,
"durationMs": 2101.054208003916,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.383309255895752,
"realTimeFactor": 0.41958465840143616,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.10930800437927246,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"loadMs": 3127.0569170010276,
"totalMs": 5228.1111250049435,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.9953279495239258
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0024406753946095705
},
{
"label": "A slow piano melody.",
"score": 0.0021972625982016325
},
{
"label": "The sound of a dog barking.",
"score": 0.000027037332984036766
},
{
"label": "A person giving a speech.",
"score": 0.000007161980647651944
}
]
},
"expected": "A solo acoustic guitar playing a melody.",
"score": 0.9953279495239258,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 2068.9605840016156,
"durationMs": 2071.973959013121,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.416672428978519,
"realTimeFactor": 0.41379211680032313,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.13945604860782623,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"loadMs": 3127.0569170010276,
"totalMs": 5199.030876014149,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.9519044756889343
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.04515129700303078
},
{
"label": "A person giving a speech.",
"score": 0.0022107241675257683
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0006932063843123615
},
{
"label": "The sound of a dog barking.",
"score": 0.00004028583862236701
}
]
},
"expected": "A slow piano melody.",
"score": 0.9519044756889343,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 2109.4080839975504,
"durationMs": 2113.374374996056,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.3703331934352283,
"realTimeFactor": 0.4218816167995101,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.11466941237449646,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"loadMs": 3156.891167003778,
"totalMs": 5270.265541999834,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.9993126392364502
},
{
"label": "A person giving a speech.",
"score": 0.0003860131255351007
},
{
"label": "The sound of a dog barking.",
"score": 0.00019279711705166847
},
{
"label": "A slow piano melody.",
"score": 0.00006709693843731657
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.00004150768290855922
}
]
},
"expected": "Electronic dance music with synthesizers and drums.",
"score": 0.9993126392364502,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 2057.4952500028303,
"durationMs": 2060.7480829930864,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.4301392676328764,
"realTimeFactor": 0.4114990500005661,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.10930800437927246,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"loadMs": 3156.891167003778,
"totalMs": 5217.639249996864,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.9953279495239258
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0024406753946095705
},
{
"label": "A slow piano melody.",
"score": 0.0021972625982016325
},
{
"label": "The sound of a dog barking.",
"score": 0.000027037332984036766
},
{
"label": "A person giving a speech.",
"score": 0.000007161980647651944
}
]
},
"expected": "A solo acoustic guitar playing a melody.",
"score": 0.9953279495239258,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 2077.179874991998,
"durationMs": 2080.3636249911506,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.4071097838935405,
"realTimeFactor": 0.4154359749983996,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.13945604860782623,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"loadMs": 3156.891167003778,
"totalMs": 5237.254791994928,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"changed": true,
"outsideEnvelopeCosine": 0.9793803095817566,
"preserved": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.7845216989517212
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.1999359130859375
},
{
"label": "A person giving a speech.",
"score": 0.008600927889347076
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0066994959488511086
},
{
"label": "The sound of a dog barking.",
"score": 0.00024192176351789385
}
]
},
"metric": "CLAP intended source class, valid signal, actual reference encoder execution and source dependence; editing also checks outside-interval amplitude envelope",
"score": 0.7845216989517212
},
"runtime": {
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"torchVersion": "2.7.1",
"loadMs": 4961.355249994085,
"provider": "Metal/MPS",
"acceleration": "partial",
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"accelerationDetails": "Neural text encoder, diffusion and audio decoder on selected device; tokenization and WAV encoding on CPU",
"inputAudioSha256": "1b07f92056c63d313e2e71313ae6cd972a7d3ca1de88faadf60d5137775bbc4f",
"inputAudioSeconds": 5,
"inputSampleRate": 44100,
"inputMode": "variation",
"initNoiseLevel": 0.2,
"peakNormalization": false,
"inferenceMs": 992.5909159937873,
"durationMs": 997.2508329956327,
"audioSeconds": 5,
"audioSecondsPerSecond": 5.037321941430396,
"realTimeFactor": 0.19851818319875747,
"activationPlacement": {
"audioEncoderOutput": "mps:0",
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.18469901382923126,
"clippedFraction": 0.0003401360544217687,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3395026944,
"device": "mps",
"precision": "fp32",
"cpuThreads": 4,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"totalMs": 5958.606082989718,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"changed": true,
"outsideEnvelopeCosine": 0.9679448008537292,
"preserved": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.7467095255851746
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.20012681186199188
},
{
"label": "A person giving a speech.",
"score": 0.033698104321956635
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0180469062179327
},
{
"label": "The sound of a dog barking.",
"score": 0.0014187140623107553
}
]
},
"metric": "CLAP intended source class, valid signal, actual reference encoder execution and source dependence; editing also checks outside-interval amplitude envelope",
"score": 0.7467095255851746
},
"runtime": {
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"torchVersion": "2.7.1",
"loadMs": 3144.75195799605,
"provider": "PyTorch CPU",
"acceleration": "cpu",
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"accelerationDetails": "Neural text encoder, diffusion and audio decoder on selected device; tokenization and WAV encoding on CPU",
"inputAudioSha256": "1b07f92056c63d313e2e71313ae6cd972a7d3ca1de88faadf60d5137775bbc4f",
"inputAudioSeconds": 5,
"inputSampleRate": 44100,
"inputMode": "variation",
"initNoiseLevel": 0.2,
"peakNormalization": false,
"inferenceMs": 2258.8283749937546,
"durationMs": 2261.8192079971777,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.213536918232588,
"realTimeFactor": 0.4517656749987509,
"activationPlacement": {
"audioEncoderOutput": "cpu",
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.18487906455993652,
"clippedFraction": 0.0007732426303854876,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"device": "cpu",
"precision": "fp32",
"cpuThreads": 4,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"totalMs": 5406.571165993228,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"changed": true,
"outsideEnvelopeCosine": 0.9998035430908203,
"preserved": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.9970663189888
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.0023353134747594595
},
{
"label": "A person giving a speech.",
"score": 0.000447980739409104
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.00014634009858127683
},
{
"label": "The sound of a dog barking.",
"score": 0.0000040311383600055706
}
]
},
"metric": "CLAP intended source class, valid signal, actual reference encoder execution and source dependence; editing also checks outside-interval amplitude envelope",
"score": 0.9970663189888
},
"runtime": {
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"torchVersion": "2.7.1",
"loadMs": 4961.355249994085,
"provider": "Metal/MPS",
"acceleration": "partial",
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"accelerationDetails": "Neural text encoder, diffusion and audio decoder on selected device; tokenization and WAV encoding on CPU",
"inputAudioSha256": "1b07f92056c63d313e2e71313ae6cd972a7d3ca1de88faadf60d5137775bbc4f",
"inputAudioSeconds": 5,
"inputSampleRate": 44100,
"inputMode": "inpaint",
"editStartSeconds": 2,
"editEndSeconds": 3,
"peakNormalization": false,
"inferenceMs": 864.7240410064114,
"durationMs": 868.921916000545,
"audioSeconds": 5,
"audioSecondsPerSecond": 5.7821915002857285,
"realTimeFactor": 0.17294480820128227,
"activationPlacement": {
"audioEncoderOutput": "mps:0",
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.12083068490028381,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": 3395043328,
"device": "mps",
"precision": "fp32",
"cpuThreads": 4,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"totalMs": 5830.27716599463,
"fallback": false,
"grade": "fast"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"changed": true,
"outsideEnvelopeCosine": 0.9997914433479309,
"preserved": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.9976342916488647
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.0020166579633951187
},
{
"label": "A person giving a speech.",
"score": 0.00023357960162684321
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.000112050591269508
},
{
"label": "The sound of a dog barking.",
"score": 0.0000034622867133293767
}
]
},
"metric": "CLAP intended source class, valid signal, actual reference encoder execution and source dependence; editing also checks outside-interval amplitude envelope",
"score": 0.9976342916488647
},
"runtime": {
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"torchVersion": "2.7.1",
"loadMs": 3144.75195799605,
"provider": "PyTorch CPU",
"acceleration": "cpu",
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"accelerationDetails": "Neural text encoder, diffusion and audio decoder on selected device; tokenization and WAV encoding on CPU",
"inputAudioSha256": "1b07f92056c63d313e2e71313ae6cd972a7d3ca1de88faadf60d5137775bbc4f",
"inputAudioSeconds": 5,
"inputSampleRate": 44100,
"inputMode": "inpaint",
"editStartSeconds": 2,
"editEndSeconds": 3,
"peakNormalization": false,
"inferenceMs": 2295.62991700368,
"durationMs": 2299.8287090013037,
"audioSeconds": 5,
"audioSecondsPerSecond": 2.178051419771589,
"realTimeFactor": 0.459125983400736,
"activationPlacement": {
"audioEncoderOutput": "cpu",
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.12046680599451065,
"clippedFraction": 0,
"seed": 42,
"steps": 8,
"cfgScale": 1,
"driverAllocatedMemoryBytes": null,
"device": "cpu",
"precision": "fp32",
"cpuThreads": 4,
"artifactRevision": "8bdfb53c6b0fbc62113cd571fd37373efe15fa14",
"totalMs": 5444.580666997354,
"fallback": false,
"grade": "fast"
}
}
]Stable Audio 3.0 Small Music Base
12 configurations testedQuality criteria
No explicit quality gate recorded. Review per-case evidence; full certification is pending.
| Precision | Device | Result | Platform | Inference / load | Task metric | |
|---|---|---|---|---|---|---|
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 5922 ms / 5287 ms | 1.184 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 5945 ms / 5287 ms | 1.189 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 6415 ms / 5287 ms | 1.283 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 6660 ms / 3544 ms | 1.332 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 6868 ms / 3544 ms | 1.374 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 6976 ms / 3544 ms | 1.395 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 8894 ms / 3622 ms | 1.779 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 9646 ms / 3622 ms | 1.929 RTF |
FP32 | mpsAccelerated | Passed | Fast | macos · aarch64 | 10332 ms / 3622 ms | 2.066 RTF |
FP32 | cpu | Failed | — | macos · aarch64 | 17755 ms / 3135 ms | 3.551 RTF |
FP32 | cpu | Failed | — | macos · aarch64 | 17243 ms / 3135 ms | 3.449 RTF |
FP32 | cpu | Failed | — | macos · aarch64 | 17440 ms / 3135 ms | 3.488 RTF |
Test setup & sample details
- Quality gate
- No explicit quality gate recorded. Review the per-case quality evidence; full certification is pending.
- Tested
- 2026-11-07T14:11:17.974084+00:00
- Environment
- { "os": "macos", "arch": "aarch64", "hardware": "Apple M3 Max", "osVersion": "27.2", "runtime": "Stable Audio 3 pinned 790793b / Torch 2.7.1 / Transformers 5.8.0", "interpreter": "App-owned CPython 3.13.18 private probe venv", "concurrentActivity": "Pinned candidate source downloads and tracked background archival service active. CLAP scoring runs between generation cases." }
- Runtime
- stable-audio-3 0.1.0+790793b
- Artifact revision
- 4249e9179cd64a03fc016919c88d0e6f2c5617f3
- Fixture
{ "file": "model-samples/desktop/stable-audio3/small-music-base/small-music-base-piano.txt", "size": 83, "sha1": "bdf829476f6d8979181472b69c8ff1e40d822a37", "sha256": "598e2c9b870380f0e7632de52d63bcaf3ec5343badf49ba10b38b507d057d783", "url": "https://assets.localai.tools/samples/v1/files/598e2c9b870380f0e7632de52d63bcaf3ec5343badf49ba10b38b507d057d783/small-music-base-piano.txt" }- Rerun when
- Checkpoint/runtime/adapter/config changes; Fixture or quality gate changes; Relevant device/backend behavior
Per-case quality & failure details
[
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.6031748652458191
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.3888615071773529
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0036690866108983755
},
{
"label": "A person giving a speech.",
"score": 0.0031573809683322906
},
{
"label": "The sound of a dog barking.",
"score": 0.001137159182690084
}
]
},
"expected": "A slow piano melody.",
"score": 0.6031748652458191,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 6414.789334012312,
"durationMs": 6420.259375008754,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.7794488235941193,
"realTimeFactor": 1.2829578668024624,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.16253401339054108,
"clippedFraction": 0.0019750566893424037,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": 3405365248,
"loadMs": 5287.417957995785,
"totalMs": 11707.67733300454,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "good"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.9987886548042297
},
{
"label": "A slow piano melody.",
"score": 0.0004895455203950405
},
{
"label": "A person giving a speech.",
"score": 0.0004496182082220912
},
{
"label": "The sound of a dog barking.",
"score": 0.00021117953292559832
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.00006104465137468651
}
]
},
"expected": "Electronic dance music with synthesizers and drums.",
"score": 0.9987886548042297,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 5921.839875009027,
"durationMs": 5925.791750007193,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.8443321848503,
"realTimeFactor": 1.1843679750018055,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.20005479454994202,
"clippedFraction": 0,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": 3405365248,
"loadMs": 5287.417957995785,
"totalMs": 11213.209708002978,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "good"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.9939483404159546
},
{
"label": "A slow piano melody.",
"score": 0.003949727863073349
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0009806999005377293
},
{
"label": "A person giving a speech.",
"score": 0.0007986784912645817
},
{
"label": "The sound of a dog barking.",
"score": 0.0003225805121473968
}
]
},
"expected": "A solo acoustic guitar playing a melody.",
"score": 0.9939483404159546,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 5945.4735420004,
"durationMs": 5948.920209004427,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.8409759062383636,
"realTimeFactor": 1.18909470840008,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.14377078413963318,
"clippedFraction": 0.0003151927437641723,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": 3405365248,
"loadMs": 5287.417957995785,
"totalMs": 11236.338167000213,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "good"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.6031748652458191
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.3888615071773529
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0036690866108983755
},
{
"label": "A person giving a speech.",
"score": 0.0031573809683322906
},
{
"label": "The sound of a dog barking.",
"score": 0.001137159182690084
}
]
},
"expected": "A slow piano melody.",
"score": 0.6031748652458191,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 6975.604999999632,
"durationMs": 6980.919291992905,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.7167837054994174,
"realTimeFactor": 1.3951209999999263,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.16253401339054108,
"clippedFraction": 0.0019750566893424037,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": 3405365248,
"loadMs": 3544.179874996189,
"totalMs": 10525.099166989094,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "good"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.9987886548042297
},
{
"label": "A slow piano melody.",
"score": 0.0004895455203950405
},
{
"label": "A person giving a speech.",
"score": 0.0004496182082220912
},
{
"label": "The sound of a dog barking.",
"score": 0.00021117953292559832
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.00006104465137468651
}
]
},
"expected": "Electronic dance music with synthesizers and drums.",
"score": 0.9987886548042297,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 6659.557457998744,
"durationMs": 6663.861458000611,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.7508006397624121,
"realTimeFactor": 1.3319114915997488,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.20005479454994202,
"clippedFraction": 0,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": 3405365248,
"loadMs": 3544.179874996189,
"totalMs": 10208.0413329968,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "good"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.9939483404159546
},
{
"label": "A slow piano melody.",
"score": 0.003949727863073349
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0009806999005377293
},
{
"label": "A person giving a speech.",
"score": 0.0007986784912645817
},
{
"label": "The sound of a dog barking.",
"score": 0.0003225805121473968
}
]
},
"expected": "A solo acoustic guitar playing a melody.",
"score": 0.9939483404159546,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 6867.9025000019465,
"durationMs": 6871.931667003082,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.7280243130997539,
"realTimeFactor": 1.3735805000003893,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.14377078413963318,
"clippedFraction": 0.0003151927437641723,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": 3405365248,
"loadMs": 3544.179874996189,
"totalMs": 10416.111541999271,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "good"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A slow piano melody.",
"score": 0.6031748652458191
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.3888615071773529
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0036690866108983755
},
{
"label": "A person giving a speech.",
"score": 0.0031573809683322906
},
{
"label": "The sound of a dog barking.",
"score": 0.001137159182690084
}
]
},
"expected": "A slow piano melody.",
"score": 0.6031748652458191,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 8894.497874993249,
"durationMs": 8899.48112500133,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.5621452801801693,
"realTimeFactor": 1.7788995749986498,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.16253401339054108,
"clippedFraction": 0.0019750566893424037,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": 3405365248,
"loadMs": 3622.359041997697,
"totalMs": 12521.840166999027,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "good"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.9987886548042297
},
{
"label": "A slow piano melody.",
"score": 0.0004895455203950405
},
{
"label": "A person giving a speech.",
"score": 0.0004496182082220912
},
{
"label": "The sound of a dog barking.",
"score": 0.00021117953292559832
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.00006104465137468651
}
]
},
"expected": "Electronic dance music with synthesizers and drums.",
"score": 0.9987886548042297,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 9645.906709003611,
"durationMs": 9656.999708997319,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.518354588204024,
"realTimeFactor": 1.9291813418007222,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.20005479454994202,
"clippedFraction": 0,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": 3405365248,
"loadMs": 3622.359041997697,
"totalMs": 13279.358750995016,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "good"
}
},
{
"precision": "fp32",
"device": "mps",
"status": "pass",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.9939483404159546
},
{
"label": "A slow piano melody.",
"score": 0.003949727863073349
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0009806999005377293
},
{
"label": "A person giving a speech.",
"score": 0.0007986784912645817
},
{
"label": "The sound of a dog barking.",
"score": 0.0003225805121473968
}
]
},
"expected": "A solo acoustic guitar playing a melody.",
"score": 0.9939483404159546,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 10331.937499999185,
"durationMs": 10335.49533301266,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.4839363381747513,
"realTimeFactor": 2.066387499999837,
"placement": {
"mps:0/torch.float32": 672,
"mps:0/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "mps:0"
},
"waveformDevice": "mps:0",
"rms": 0.14377078413963318,
"clippedFraction": 0.0003151927437641723,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": 3405365248,
"loadMs": 3622.359041997697,
"totalMs": 13957.854375010356,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "mps",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "good"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "fail",
"quality": {
"passed": false,
"signal": false,
"semantic": false,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.93120938539505
},
{
"label": "A slow piano melody.",
"score": 0.05310773849487305
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.01023319736123085
},
{
"label": "The sound of a dog barking.",
"score": 0.003698357380926609
},
{
"label": "A person giving a speech.",
"score": 0.0017512980848550797
}
]
},
"expected": "A slow piano melody.",
"score": 0.05310773849487305,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 17754.78962500347,
"durationMs": 17760.751417008578,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.2816141506379028,
"realTimeFactor": 3.5509579250006937,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.2110019028186798,
"clippedFraction": 0.012476190476190476,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": null,
"loadMs": 3135.378249993664,
"totalMs": 20896.129667002242,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "slow"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "fail",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.9995396137237549
},
{
"label": "A person giving a speech.",
"score": 0.00016365383635275066
},
{
"label": "The sound of a dog barking.",
"score": 0.0001441351923858747
},
{
"label": "A slow piano melody.",
"score": 0.00012586727098096162
},
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.000026832189178094268
}
]
},
"expected": "Electronic dance music with synthesizers and drums.",
"score": 0.9995396137237549,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 17243.008791003376,
"durationMs": 17247.273540997412,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.2899725947253924,
"realTimeFactor": 3.448601758200675,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.15724915266036987,
"clippedFraction": 0.000004535147392290249,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": null,
"loadMs": 3135.378249993664,
"totalMs": 20382.651790991076,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "slow"
}
},
{
"precision": "fp32",
"device": "cpu",
"status": "fail",
"quality": {
"passed": true,
"signal": true,
"semantic": true,
"clap": {
"predictions": [
{
"label": "A solo acoustic guitar playing a melody.",
"score": 0.9951090216636658
},
{
"label": "A slow piano melody.",
"score": 0.003878642339259386
},
{
"label": "Electronic dance music with synthesizers and drums.",
"score": 0.0006580938934348524
},
{
"label": "A person giving a speech.",
"score": 0.00027963961474597454
},
{
"label": "The sound of a dog barking.",
"score": 0.0000746085133869201
}
]
},
"expected": "A solo acoustic guitar playing a melody.",
"score": 0.9951090216636658,
"metric": "CLAP top-1 intended class among five frozen labels plus signal validity"
},
"runtime": {
"inferenceMs": 17439.933874993585,
"durationMs": 17444.529833999695,
"audioSeconds": 5,
"audioSecondsPerSecond": 0.28669833474364814,
"realTimeFactor": 3.487986774998717,
"placement": {
"cpu/torch.float32": 672,
"cpu/torch.bfloat16": 134
},
"activationPlacement": {
"diffusionOutput": "cpu"
},
"waveformDevice": "cpu",
"rms": 0.11063375324010849,
"clippedFraction": 0,
"seed": 42,
"steps": 50,
"cfgScale": 7,
"driverAllocatedMemoryBytes": null,
"loadMs": 3135.378249993664,
"totalMs": 20579.90808399336,
"artifactRevision": "4249e9179cd64a03fc016919c88d0e6f2c5617f3",
"precision": "fp32",
"device": "cpu",
"engine": "stable-audio-3",
"version": "0.1.0+790793b",
"cpuThreads": 4,
"fallback": false,
"grade": "slow"
}
}
]Release validation still required
- Complete public dependency archive and forced model/runtime-source recovery remain pending.