A model is still running

A model is still running or downloading. Leaving will stop the current operation. Stay on this page to let it finish.

Skip to content
LocalAI.toolsBeta

Join the waitlist

Get an email when the new apps and model runs launch.

No account needed.

Basic Pitch

Convert instrument recordings into note events and MIDI locally using the verified CPU path.

Desktop inferencePrivate · Runs on your computer
About this model

Convert instrument recordings into note events and MIDI locally using the verified CPU path.

Created by
Spotify
License
Apache-2.0
Model released
Jun 1, 2022
Source repository created
May 3, 2022
Model weights · FP32
0.3 MB

Best for

  • Editable note and MIDI drafts from isolated instruments

Keep in mind

  • CPU-only exception approved by the user; CoreML GPU failed chord and silence quality tests.
  • Noise, percussion, mixtures, overlapping instruments and expressive playing can produce extra or missing notes; review MIDI.
  • Only sampled piano melody, chords and silence locally evaluated; no broad instrument-quality claim.
  • Process recordings you have permission to use.
  • Up to60seconds normalized mono PCM; current Apple Silicon Mac only.

Processed on your deviceYour input is processed locally in the desktop app. Model and runtime files download from external hosts.

Target platforms · verification required

  • macos · aarch64, x86_64
  • windows · x86_64

Recommended configuration

FP32 · cpu (verified platform only)

Sources & technical details
Original model
https://github.com/spotify/basic-pitch
Release verification
Verified
Release evidence
Official Spotify announcement of original ICASSP2022 model. Explicit user-approved older-model exception;2024 CoreML/package serialization does not constitute a new checkpoint.
Source updated
Aug 16, 2024
Native model files
https://github.com/spotify/basic-pitch
Files updated
Aug 16, 2024
Artifact revision
9991303bba609a3b93089d13ec80d1d495083596

Desktop app

Sample result

Basic Pitch

Run Basic Pitch 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

macOS Apple SiliconCPU / FP32Tested4.0 GiBRAM estimate
Windows Intel / AMDCPU / FP32Unverified4.0 GiBRAM estimate
macOS Intel / AMDCPU / FP32Unverified4.0 GiBRAM estimate
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.

Basic Pitch

18 configurations tested
Passed18
Failed0
TestedNov 4, 2026
Runtimecoremltools 9.0

Quality criteria

No explicit quality gate recorded. Review per-case evidence; full certification is pending.

PrecisionDeviceResultPlatformInference / loadTask metric
FP32cpuPassedVery fastmacos · aarch6439 ms / 1081 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6439 ms / 1081 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6439 ms / 1081 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6439 ms / 50360 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6440 ms / 1120 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6440 ms / 1120 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6440 ms / 50360 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6440 ms / 1081 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6440 ms / 50360 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6440 ms / 1081 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6441 ms / 50360 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6441 ms / 1120 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6441 ms / 1120 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6441 ms / 50360 ms0.008 RTF
FP32cpuPassedVery fastmacos · aarch6444 ms / 1120 ms0.009 RTF
FP32cpuPassedVery fastmacos · aarch64327 ms / 1081 ms0.065 RTF
FP32cpuPassedVery fastmacos · aarch64386 ms / 1120 ms0.077 RTF
FP32cpuPassedNot measuredmacos · aarch645912 ms / 50360 ms1.182 RTF
Test setup & sample details
Quality gate
No explicit quality gate recorded. Review the per-case quality evidence; full certification is pending.
Tested
2026-11-04T19:40:23.832Z
Environment
{ "os": "macos", "arch": "aarch64", "hardware": "Apple M3 Max", "osVersion": "27.2" }
Runtime
coremltools 9.0
Artifact revision
9991303bba609a3b93089d13ec80d1d495083596
Fixture
{
  "file": "model-samples/desktop/transkun-v2/melody.wav",
  "size": 441044,
  "sha1": "3134ef8497155ef317c1c36d43178c331f53492a",
  "sha256": "f28de3b2a0d500b424e944b3060c0eaaa7ca540d4d313a568f84ca3dff827f6a",
  "url": "https://assets.localai.tools/samples/v1/files/f28de3b2a0d500b424e944b3060c0eaaa7ca540d4d313a568f84ca3dff827f6a/melody.wav"
}
Rerun when
Checkpoint/runtime/adapter/config changes; Fixture/quality-gate changes; Relevant backend behavior
Per-case quality & failure details
[
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 0.8888888888888888,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 0.8457744642558349,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 6031.661833,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 5911.74150002189,
      "loadMs": 50359.721542,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 1.182348300004378,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 56271.46304202189,
      "grade": "usable"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 0.8888888888888888,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 121.0291103870384,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 159.71975,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 41.31237504770979,
      "loadMs": 50359.721542,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.00826247500954196,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 50401.03391704771,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 122.3448625704934,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 128.326584,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 40.86808301508427,
      "loadMs": 50359.721542,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.008173616603016853,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 50400.58962501508,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 123.8563904468086,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 127.58037499999999,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 40.369334048591554,
      "loadMs": 50359.721542,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.00807386680971831,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 50400.09087604859,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 124.90138102269356,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 85.17675,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 40.0315829901956,
      "loadMs": 50359.721542,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.008006316598039121,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 50399.753124990195,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 128.80069266691032,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 83.10466699999999,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 38.81966701010242,
      "loadMs": 50359.721542,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.007763933402020484,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 50398.5412090101,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 0.8888888888888888,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 15.2807513196403,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 441.924583,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 327.2090419777669,
      "loadMs": 1081.298791,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.06544180839555339,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1408.5078329777668,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 0.8888888888888888,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 129.81224717304477,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 152.812708,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 38.51716697681695,
      "loadMs": 1081.298791,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.0077034333953633904,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1119.815957976817,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 124.22668900319071,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 125.47533299999999,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 40.24899995420128,
      "loadMs": 1081.298791,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.008049799990840257,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1121.5477909542012,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 123.57413377634532,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 125.94429200000002,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 40.46154196839779,
      "loadMs": 1081.298791,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.008092308393679559,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1121.7603329683977,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 129.66146582567026,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 82.75833300000001,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 38.56195800472051,
      "loadMs": 1081.298791,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.007712391600944102,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1119.8607490047204,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 128.8441158460156,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 82.134666,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 38.806583965197206,
      "loadMs": 1081.298791,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.007761316793039441,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1120.1053749651971,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 0.8888888888888888,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 12.944221407038643,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 505.111042,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 386.2727500381879,
      "loadMs": 1120.465625,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.07725455000763759,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1506.738375038188,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 0.8888888888888888,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 126.00806455700187,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 157.095334,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 39.67999998712912,
      "loadMs": 1120.465625,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.007935999997425825,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1160.1456249871292,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 126.30422369833464,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 128.99949999999998,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 39.58695800974965,
      "loadMs": 1120.465625,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.00791739160194993,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1160.0525830097497,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 113.24088028752011,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 129.913792,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 44.15366594912484,
      "loadMs": 1120.465625,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.008830733189824969,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1164.619290949125,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 121.8056469329381,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 84.855584,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 41.04899999219924,
      "loadMs": 1120.465625,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.008209799998439848,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1161.5146249921993,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "noteOnsetF1": 1,
      "threshold": 0.85,
      "passed": true
    },
    "runtime": {
      "acceleration": "cpu",
      "accelerationDetails": "Explicit CPU_ONLY compute units. fp32 labels the audio input/runtime selection, not the mixed-precision stored CoreML weights. GPU excluded after chord and silence failures.",
      "artifactRevision": "9991303bba609a3b93089d13ec80d1d495083596",
      "audioSeconds": 5,
      "audioSecondsPerSecond": 121.03264958077963,
      "cpuThreads": 4,
      "device": "cpu",
      "durationMs": 85.76391699999999,
      "engine": "coremltools",
      "fallback": false,
      "fallbackReason": null,
      "family": {
        "catalogId": "basic-pitch",
        "defaultModelId": "basic-pitch-native",
        "displayName": "Basic Pitch",
        "id": "basic-pitch",
        "variantId": "icassp-2022",
        "variantName": "ICASSP2022"
      },
      "inferenceMs": 41.311167005915195,
      "loadMs": 1120.465625,
      "measurementScope": "Resampling, neural inference and note decoding; excludes IPC and visualization",
      "modelId": "basic-pitch-native",
      "parameterDtype": "upstream mixed CoreML precision",
      "placement": {
        "CPU_ONLY": true
      },
      "precision": "fp32",
      "provider": "CoreML CPU_ONLY",
      "realTimeFactor": 0.008262233401183039,
      "requestedDevice": "cpu",
      "requestedPrecision": "fp32",
      "reused": false,
      "selectionReason": "Explicit device selection; actual placement is reported after inference",
      "torchVersion": null,
      "version": "9.0",
      "totalMs": 1161.7767920059152,
      "grade": "good"
    }
  }
]

Something not quite right?

Tell us what happened with Basic Pitch. Loading, results, or the page itself—we’re here for all of it.

Attachments Optional

Up to 3 files · 50 MB each

Browser, OS, CPU/GPU, memory, screen, storage, and network details are included automatically. Model inputs, results, and browsing history are not collected.