| [ | |
| { | |
| "shortDescription" : "SortFormer head (conformer + transformer + prediction) for speaker diarization", | |
| "metadataOutputVersion" : "3.0", | |
| "outputSchema" : [ | |
| { | |
| "hasShapeFlexibility" : "0", | |
| "isOptional" : "0", | |
| "dataType" : "Float32", | |
| "formattedType" : "MultiArray (Float32 1 × 400 × 4)", | |
| "shortDescription" : "Speaker probabilities [1, T, 4]", | |
| "shape" : "[1, 400, 4]", | |
| "name" : "speaker_preds", | |
| "type" : "MultiArray" | |
| } | |
| ], | |
| "version" : "1.0", | |
| "modelParameters" : [ | |
| ], | |
| "author" : "sortformer-ggml", | |
| "specificationVersion" : 7, | |
| "storagePrecision" : "Float16", | |
| "license" : "Apache-2.0", | |
| "mlProgramOperationTypeHistogram" : { | |
| "Transpose" : 192, | |
| "Ios16.softmax" : 35, | |
| "Split" : 17, | |
| "Ios16.linear" : 247, | |
| "Select" : 51, | |
| "Ios16.add" : 173, | |
| "Tile" : 1, | |
| "Ios16.sigmoid" : 18, | |
| "Ios16.logicalAnd" : 2, | |
| "Pad" : 34, | |
| "ExpandDims" : 8, | |
| "Ios16.sub" : 1, | |
| "Ios16.cast" : 4, | |
| "Ios16.less" : 2, | |
| "Ios16.conv" : 51, | |
| "Ios16.matmul" : 87, | |
| "Ios16.layerNorm" : 121, | |
| "SliceByIndex" : 34, | |
| "Ios16.reshape" : 174, | |
| "Ios16.silu" : 51, | |
| "Ios16.relu" : 20, | |
| "Ios16.mul" : 107, | |
| "Ios16.logicalNot" : 2 | |
| }, | |
| "computePrecision" : "Mixed (Float16, Float32, Int32)", | |
| "stateSchema" : [ | |
| ], | |
| "isUpdatable" : "0", | |
| "availability" : { | |
| "macOS" : "13.0", | |
| "tvOS" : "16.0", | |
| "visionOS" : "1.0", | |
| "watchOS" : "9.0", | |
| "iOS" : "16.0", | |
| "macCatalyst" : "16.0" | |
| }, | |
| "modelType" : { | |
| "name" : "MLModelType_mlProgram" | |
| }, | |
| "inputSchema" : [ | |
| { | |
| "hasShapeFlexibility" : "0", | |
| "isOptional" : "0", | |
| "dataType" : "Float32", | |
| "formattedType" : "MultiArray (Float32 1 × 400 × 512)", | |
| "shortDescription" : "Pre-encoder embeddings [1, T, 512], T <= 400", | |
| "shape" : "[1, 400, 512]", | |
| "name" : "pre_encoder_embs", | |
| "type" : "MultiArray" | |
| }, | |
| { | |
| "hasShapeFlexibility" : "0", | |
| "isOptional" : "0", | |
| "dataType" : "Int32", | |
| "formattedType" : "MultiArray (Int32 1)", | |
| "shortDescription" : "Valid sequence length", | |
| "shape" : "[1]", | |
| "name" : "pre_encoder_lengths", | |
| "type" : "MultiArray" | |
| } | |
| ], | |
| "userDefinedMetadata" : { | |
| "com.github.apple.coremltools.source" : "torch==2.10.0", | |
| "com.github.apple.coremltools.version" : "9.0", | |
| "com.github.apple.coremltools.conversion_date" : "2026-02-03", | |
| "com.github.apple.coremltools.source_dialect" : "TorchScript", | |
| "n_speakers" : "4", | |
| "precision" : "fp16", | |
| "max_seq_len" : "400", | |
| "fc_d_model" : "512" | |
| }, | |
| "generatedClassName" : "model_coreml_head", | |
| "method" : "predict" | |
| } | |
| ] |