[ { "metadataOutputVersion" : "3.0", "storagePrecision" : "Mixed (Float16, Int8)", "outputSchema" : [ { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float16", "formattedType" : "MultiArray (Float16 390 × 1024)", "shortDescription" : "", "shape" : "[390, 1024]", "name" : "hidden", "type" : "MultiArray" } ], "modelParameters" : [ ], "specificationVersion" : 9, "functions" : [ { "inputSchema" : [ { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 128 × 500)", "shortDescription" : "", "shape" : "[1, 128, 500]", "name" : "audios", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 1 × 65 × 65)", "shortDescription" : "", "shape" : "[1, 1, 65, 65]", "name" : "attn_mask", "type" : "MultiArray" } ], "computePrecision" : "Mixed (Float16, Int32)", "storagePrecision" : "Mixed (Float16, Int8)", "stateSchema" : [ ], "outputSchema" : [ { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float16", "formattedType" : "MultiArray (Float16 65 × 1024)", "shortDescription" : "", "shape" : "[65, 1024]", "name" : "hidden", "type" : "MultiArray" } ], "name" : "tower_5s", "mlProgramOperationTypeHistogram" : { "Ios18.linear" : 119, "Ios18.conv" : 3, "Ios18.expandDims" : 55, "Ios18.matmul" : 36, "Ios18.gelu" : 26, "Ios16.constexprAffineDequantize" : 123, "Ios18.add" : 59, "Ios18.softmax" : 18, "Ios18.sliceByIndex" : 1, "Ios18.layerNorm" : 42, "Ios18.transpose" : 75, "Ios18.cast" : 2, "Ios18.squeeze" : 18, "Ios18.reshape" : 75, "Ios18.mul" : 18 } }, { "inputSchema" : [ { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 128 × 1000)", "shortDescription" : "", "shape" : "[1, 128, 1000]", "name" : "audios", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 1 × 130 × 130)", "shortDescription" : "", "shape" : "[1, 1, 130, 130]", "name" : "attn_mask", "type" : "MultiArray" } ], "computePrecision" : "Mixed (Float16, Int32)", "storagePrecision" : "Mixed (Float16, Int8)", "stateSchema" : [ ], "outputSchema" : [ { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float16", "formattedType" : "MultiArray (Float16 130 × 1024)", "shortDescription" : "", "shape" : "[130, 1024]", "name" : "hidden", "type" : "MultiArray" } ], "name" : "tower_10s", "mlProgramOperationTypeHistogram" : { "Ios18.linear" : 119, "Ios18.conv" : 3, "Ios18.expandDims" : 55, "Ios18.matmul" : 36, "Ios18.gelu" : 26, "Ios16.constexprAffineDequantize" : 123, "Ios18.add" : 59, "Ios18.softmax" : 18, "Ios18.sliceByIndex" : 1, "Ios18.layerNorm" : 42, "Ios18.transpose" : 75, "Ios18.cast" : 2, "Ios18.squeeze" : 18, "Ios18.reshape" : 75, "Ios18.mul" : 18 } }, { "inputSchema" : [ { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 128 × 3000)", "shortDescription" : "", "shape" : "[1, 128, 3000]", "name" : "audios", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 1 × 390 × 390)", "shortDescription" : "", "shape" : "[1, 1, 390, 390]", "name" : "attn_mask", "type" : "MultiArray" } ], "computePrecision" : "Mixed (Float16, Int32)", "storagePrecision" : "Mixed (Float16, Int8)", "stateSchema" : [ ], "outputSchema" : [ { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float16", "formattedType" : "MultiArray (Float16 390 × 1024)", "shortDescription" : "", "shape" : "[390, 1024]", "name" : "hidden", "type" : "MultiArray" } ], "name" : "tower_30s", "mlProgramOperationTypeHistogram" : { "Ios18.linear" : 119, "Ios18.conv" : 3, "Ios18.expandDims" : 55, "Ios18.matmul" : 36, "Ios18.gelu" : 26, "Ios16.constexprAffineDequantize" : 123, "Ios18.add" : 59, "Ios18.softmax" : 18, "Ios18.sliceByIndex" : 1, "Ios18.layerNorm" : 42, "Ios18.transpose" : 75, "Ios18.cast" : 2, "Ios18.squeeze" : 18, "Ios18.reshape" : 75, "Ios18.mul" : 18 } } ], "mlProgramOperationTypeHistogram" : { "Ios18.linear" : 119, "Ios18.conv" : 3, "Ios18.expandDims" : 55, "Ios18.matmul" : 36, "Ios18.gelu" : 26, "Ios16.constexprAffineDequantize" : 123, "Ios18.add" : 59, "Ios18.softmax" : 18, "Ios18.sliceByIndex" : 1, "Ios18.layerNorm" : 42, "Ios18.transpose" : 75, "Ios18.cast" : 2, "Ios18.squeeze" : 18, "Ios18.reshape" : 75, "Ios18.mul" : 18 }, "isUpdatable" : "0", "stateSchema" : [ ], "availability" : { "macOS" : "15.0", "tvOS" : "18.0", "visionOS" : "2.0", "watchOS" : "11.0", "iOS" : "18.0", "macCatalyst" : "18.0" }, "computePrecision" : "Mixed (Float16, Int32)", "modelType" : { "name" : "MLModelType_mlProgram" }, "inputSchema" : [ { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 128 × 3000)", "shortDescription" : "", "shape" : "[1, 128, 3000]", "name" : "audios", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 1 × 390 × 390)", "shortDescription" : "", "shape" : "[1, 1, 390, 390]", "name" : "attn_mask", "type" : "MultiArray" } ], "defaultFunctionName" : "tower_30s", "generatedClassName" : "audio_tower_multifunction_w8", "userDefinedMetadata" : { }, "method" : "predict" } ]