[ { "metadataOutputVersion" : "3.0", "storagePrecision" : "Float16", "outputSchema" : [ { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 512 × 8)", "shortDescription" : "", "shape" : "[1, 512, 8]", "name" : "encoded_output", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Int32", "formattedType" : "MultiArray (Int32 1)", "shortDescription" : "", "shape" : "[1]", "name" : "encoded_length", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 128 × 9)", "shortDescription" : "", "shape" : "[1, 128, 9]", "name" : "new_pre_cache", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 17 × 1 × 70 × 512)", "shortDescription" : "", "shape" : "[17, 1, 70, 512]", "name" : "new_cache_last_channel", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 17 × 1 × 512 × 8)", "shortDescription" : "", "shape" : "[17, 1, 512, 8]", "name" : "new_cache_last_time", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Int32", "formattedType" : "MultiArray (Int32 1)", "shortDescription" : "", "shape" : "[1]", "name" : "new_cache_last_channel_len", "type" : "MultiArray" } ], "modelParameters" : [ ], "specificationVersion" : 8, "mlProgramOperationTypeHistogram" : { "Ios17.logicalAnd" : 3, "Ios17.reshape" : 103, "Ios16.softmax" : 17, "Ios17.matmul" : 51, "Ios17.transpose" : 157, "Split" : 17, "Ios17.expandDims" : 18, "Select" : 51, "Ios17.add" : 132, "Tile" : 8, "Ios17.sliceByIndex" : 140, "Ios16.sigmoid" : 17, "Pad" : 20, "Ios17.logicalNot" : 2, "Ios17.layerNorm" : 102, "Ios17.less" : 5, "Ios17.sub" : 4, "Ios17.conv" : 56, "Ios16.relu" : 3, "Ios17.clip" : 2, "Ios17.linear" : 137, "Ios17.concat" : 52, "Ios17.floorDiv" : 3, "Ios17.greaterEqual" : 1, "Ios17.cast" : 20, "Ios16.silu" : 51, "Stack" : 2, "Ios17.mul" : 78 }, "computePrecision" : "Mixed (Float16, Float32, Int32)", "isUpdatable" : "0", "stateSchema" : [ ], "availability" : { "macOS" : "14.0", "tvOS" : "17.0", "visionOS" : "1.0", "watchOS" : "10.0", "iOS" : "17.0", "macCatalyst" : "17.0" }, "modelType" : { "name" : "MLModelType_mlProgram" }, "userDefinedMetadata" : { "com.github.apple.coremltools.source_dialect" : "TorchScript", "com.github.apple.coremltools.source" : "torch==2.4.0", "com.github.apple.coremltools.version" : "8.3.0" }, "inputSchema" : [ { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 128 × 64)", "shortDescription" : "", "shape" : "[1, 128, 64]", "name" : "audio_signal", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Int32", "formattedType" : "MultiArray (Int32 1)", "shortDescription" : "", "shape" : "[1]", "name" : "audio_length", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 1 × 128 × 9)", "shortDescription" : "", "shape" : "[1, 128, 9]", "name" : "pre_cache", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 17 × 1 × 70 × 512)", "shortDescription" : "", "shape" : "[17, 1, 70, 512]", "name" : "cache_last_channel", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Float32", "formattedType" : "MultiArray (Float32 17 × 1 × 512 × 8)", "shortDescription" : "", "shape" : "[17, 1, 512, 8]", "name" : "cache_last_time", "type" : "MultiArray" }, { "hasShapeFlexibility" : "0", "isOptional" : "0", "dataType" : "Int32", "formattedType" : "MultiArray (Int32 1)", "shortDescription" : "", "shape" : "[1]", "name" : "cache_last_channel_len", "type" : "MultiArray" } ], "generatedClassName" : "streaming_encoder", "method" : "predict" } ]