TalkMate commited on
Commit
764c546
·
verified ·
1 Parent(s): 6954465

Upload 25 files

Browse files
Decoder.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9660ecb2a69b7444cfb066510758cb2e8a6047b829c0300d26dea3bc3dcb4ba
3
+ size 243
Decoder.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc6009dec04d99448124a1fcbb6dd986b9ee463a20cb7bd8008524e46354e6f7
3
+ size 492
Decoder.mlmodelc/metadata.json ADDED
@@ -0,0 +1,121 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "metadataOutputVersion" : "3.0",
4
+ "storagePrecision" : "Float16",
5
+ "outputSchema" : [
6
+ {
7
+ "hasShapeFlexibility" : "0",
8
+ "isOptional" : "0",
9
+ "dataType" : "Float32",
10
+ "formattedType" : "MultiArray (Float32 1 × 640 × 1)",
11
+ "shortDescription" : "",
12
+ "shape" : "[1, 640, 1]",
13
+ "name" : "decoder",
14
+ "type" : "MultiArray"
15
+ },
16
+ {
17
+ "hasShapeFlexibility" : "0",
18
+ "isOptional" : "0",
19
+ "dataType" : "Float32",
20
+ "formattedType" : "MultiArray (Float32 2 × 1 × 640)",
21
+ "shortDescription" : "",
22
+ "shape" : "[2, 1, 640]",
23
+ "name" : "h_out",
24
+ "type" : "MultiArray"
25
+ },
26
+ {
27
+ "hasShapeFlexibility" : "0",
28
+ "isOptional" : "0",
29
+ "dataType" : "Float32",
30
+ "formattedType" : "MultiArray (Float32 2 × 1 × 640)",
31
+ "shortDescription" : "",
32
+ "shape" : "[2, 1, 640]",
33
+ "name" : "c_out",
34
+ "type" : "MultiArray"
35
+ }
36
+ ],
37
+ "modelParameters" : [
38
+
39
+ ],
40
+ "specificationVersion" : 9,
41
+ "mlProgramOperationTypeHistogram" : {
42
+ "Select" : 1,
43
+ "Ios18.lstm" : 2,
44
+ "Ios18.gather" : 1,
45
+ "Stack" : 2,
46
+ "Identity" : 1,
47
+ "Split" : 2,
48
+ "Ios18.transpose" : 2,
49
+ "Ios18.squeeze" : 4,
50
+ "Ios18.cast" : 8,
51
+ "Ios18.greaterEqual" : 1,
52
+ "Ios18.add" : 1
53
+ },
54
+ "computePrecision" : "Mixed (Float16, Float32, Int16, Int32)",
55
+ "isUpdatable" : "0",
56
+ "stateSchema" : [
57
+
58
+ ],
59
+ "availability" : {
60
+ "macOS" : "15.0",
61
+ "tvOS" : "18.0",
62
+ "visionOS" : "2.0",
63
+ "watchOS" : "11.0",
64
+ "iOS" : "18.0",
65
+ "macCatalyst" : "18.0"
66
+ },
67
+ "modelType" : {
68
+ "name" : "MLModelType_mlProgram"
69
+ },
70
+ "userDefinedMetadata" : {
71
+ "com.github.apple.coremltools.conversion_date" : "2026-11-01",
72
+ "com.github.apple.coremltools.source" : "torch==2.15.0",
73
+ "com.github.apple.coremltools.version" : "9.0",
74
+ "com.github.apple.coremltools.source_dialect" : "TorchScript"
75
+ },
76
+ "inputSchema" : [
77
+ {
78
+ "hasShapeFlexibility" : "0",
79
+ "isOptional" : "0",
80
+ "dataType" : "Int32",
81
+ "formattedType" : "MultiArray (Int32 1 × 1)",
82
+ "shortDescription" : "",
83
+ "shape" : "[1, 1]",
84
+ "name" : "targets",
85
+ "type" : "MultiArray"
86
+ },
87
+ {
88
+ "hasShapeFlexibility" : "0",
89
+ "isOptional" : "0",
90
+ "dataType" : "Int32",
91
+ "formattedType" : "MultiArray (Int32 1)",
92
+ "shortDescription" : "",
93
+ "shape" : "[1]",
94
+ "name" : "target_length",
95
+ "type" : "MultiArray"
96
+ },
97
+ {
98
+ "hasShapeFlexibility" : "0",
99
+ "isOptional" : "0",
100
+ "dataType" : "Float32",
101
+ "formattedType" : "MultiArray (Float32 2 × 1 × 640)",
102
+ "shortDescription" : "",
103
+ "shape" : "[2, 1, 640]",
104
+ "name" : "h_in",
105
+ "type" : "MultiArray"
106
+ },
107
+ {
108
+ "hasShapeFlexibility" : "0",
109
+ "isOptional" : "0",
110
+ "dataType" : "Float32",
111
+ "formattedType" : "MultiArray (Float32 2 × 1 × 640)",
112
+ "shortDescription" : "",
113
+ "shape" : "[2, 1, 640]",
114
+ "name" : "c_in",
115
+ "type" : "MultiArray"
116
+ }
117
+ ],
118
+ "generatedClassName" : "Decoder",
119
+ "method" : "predict"
120
+ }
121
+ ]
Decoder.mlmodelc/model.mil ADDED
@@ -0,0 +1,73 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ program(1.3)
2
+ [buildInfo = dict<string, string>({{"coremlc-component-MIL", "3600.16.1"}, {"coremlc-version", "3600.25.1"}, {"coremltools-component-torch", "2.15.0"}, {"coremltools-source-dialect", "TorchScript"}, {"coremltools-version", "9.0"}})]
3
+ {
4
+ func main<ios18>(tensor<fp32, [2, 1, 640]> c_in, tensor<fp32, [2, 1, 640]> h_in, tensor<int32, [1]> target_length, tensor<int32, [1, 1]> targets) {
5
+ int32 var_15_batch_dims_0 = const()[name = string("op_15_batch_dims_0"), val = int32(0)];
6
+ bool var_15_validate_indices_0 = const()[name = string("op_15_validate_indices_0"), val = bool(false)];
7
+ tensor<fp16, [8193, 640]> embed_weight_to_fp16 = const()[name = string("embed_weight_to_fp16"), val = tensor<fp16, [8193, 640]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(64)))];
8
+ string targets_to_int16_dtype_0 = const()[name = string("targets_to_int16_dtype_0"), val = string("int16")];
9
+ string cast_1_dtype_0 = const()[name = string("cast_1_dtype_0"), val = string("int32")];
10
+ int32 greater_equal_0_y_0 = const()[name = string("greater_equal_0_y_0"), val = int32(0)];
11
+ tensor<int16, [1, 1]> targets_to_int16 = cast(dtype = targets_to_int16_dtype_0, x = targets)[name = string("cast_9")];
12
+ tensor<int32, [1, 1]> cast_1 = cast(dtype = cast_1_dtype_0, x = targets_to_int16)[name = string("cast_8")];
13
+ tensor<bool, [1, 1]> greater_equal_0 = greater_equal(x = cast_1, y = greater_equal_0_y_0)[name = string("greater_equal_0")];
14
+ int32 slice_by_index_0 = const()[name = string("slice_by_index_0"), val = int32(8193)];
15
+ tensor<int32, [1, 1]> add_2 = add(x = cast_1, y = slice_by_index_0)[name = string("add_2")];
16
+ tensor<int32, [1, 1]> select_0 = select(a = cast_1, b = add_2, cond = greater_equal_0)[name = string("select_0")];
17
+ int32 var_15_cast_fp16_cast_uint16_axis_0 = const()[name = string("op_15_cast_fp16_cast_uint16_axis_0"), val = int32(0)];
18
+ string select_0_to_int16_dtype_0 = const()[name = string("select_0_to_int16_dtype_0"), val = string("int16")];
19
+ tensor<int16, [1, 1]> select_0_to_int16 = cast(dtype = select_0_to_int16_dtype_0, x = select_0)[name = string("cast_7")];
20
+ tensor<fp16, [1, 1, 640]> var_15_cast_fp16_cast_uint16_cast_uint16 = gather(axis = var_15_cast_fp16_cast_uint16_axis_0, batch_dims = var_15_batch_dims_0, indices = select_0_to_int16, validate_indices = var_15_validate_indices_0, x = embed_weight_to_fp16)[name = string("op_15_cast_fp16_cast_uint16_cast_uint16")];
21
+ tensor<int32, [3]> input_perm_0 = const()[name = string("input_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
22
+ int32 split_0_num_splits_0 = const()[name = string("split_0_num_splits_0"), val = int32(2)];
23
+ int32 split_0_axis_0 = const()[name = string("split_0_axis_0"), val = int32(0)];
24
+ string h_in_to_fp16_dtype_0 = const()[name = string("h_in_to_fp16_dtype_0"), val = string("fp16")];
25
+ tensor<fp16, [2, 1, 640]> h_in_to_fp16 = cast(dtype = h_in_to_fp16_dtype_0, x = h_in)[name = string("cast_6")];
26
+ tensor<fp16, [1, 1, 640]> split_0_cast_fp16_0, tensor<fp16, [1, 1, 640]> split_0_cast_fp16_1 = split(axis = split_0_axis_0, num_splits = split_0_num_splits_0, x = h_in_to_fp16)[name = string("split_0_cast_fp16")];
27
+ int32 split_1_num_splits_0 = const()[name = string("split_1_num_splits_0"), val = int32(2)];
28
+ int32 split_1_axis_0 = const()[name = string("split_1_axis_0"), val = int32(0)];
29
+ string c_in_to_fp16_dtype_0 = const()[name = string("c_in_to_fp16_dtype_0"), val = string("fp16")];
30
+ tensor<fp16, [2, 1, 640]> c_in_to_fp16 = cast(dtype = c_in_to_fp16_dtype_0, x = c_in)[name = string("cast_5")];
31
+ tensor<fp16, [1, 1, 640]> split_1_cast_fp16_0, tensor<fp16, [1, 1, 640]> split_1_cast_fp16_1 = split(axis = split_1_axis_0, num_splits = split_1_num_splits_0, x = c_in_to_fp16)[name = string("split_1_cast_fp16")];
32
+ tensor<int32, [1]> y_lstm_layer_0_lstm_h0_squeeze_axes_0 = const()[name = string("y_lstm_layer_0_lstm_h0_squeeze_axes_0"), val = tensor<int32, [1]>([0])];
33
+ tensor<fp16, [1, 640]> y_lstm_layer_0_lstm_h0_squeeze_cast_fp16 = squeeze(axes = y_lstm_layer_0_lstm_h0_squeeze_axes_0, x = split_0_cast_fp16_0)[name = string("y_lstm_layer_0_lstm_h0_squeeze_cast_fp16")];
34
+ tensor<int32, [1]> y_lstm_layer_0_lstm_c0_squeeze_axes_0 = const()[name = string("y_lstm_layer_0_lstm_c0_squeeze_axes_0"), val = tensor<int32, [1]>([0])];
35
+ tensor<fp16, [1, 640]> y_lstm_layer_0_lstm_c0_squeeze_cast_fp16 = squeeze(axes = y_lstm_layer_0_lstm_c0_squeeze_axes_0, x = split_1_cast_fp16_0)[name = string("y_lstm_layer_0_lstm_c0_squeeze_cast_fp16")];
36
+ string y_lstm_layer_0_direction_0 = const()[name = string("y_lstm_layer_0_direction_0"), val = string("forward")];
37
+ bool y_lstm_layer_0_output_sequence_0 = const()[name = string("y_lstm_layer_0_output_sequence_0"), val = bool(true)];
38
+ string y_lstm_layer_0_recurrent_activation_0 = const()[name = string("y_lstm_layer_0_recurrent_activation_0"), val = string("sigmoid")];
39
+ string y_lstm_layer_0_cell_activation_0 = const()[name = string("y_lstm_layer_0_cell_activation_0"), val = string("tanh")];
40
+ string y_lstm_layer_0_activation_0 = const()[name = string("y_lstm_layer_0_activation_0"), val = string("tanh")];
41
+ tensor<fp16, [2560, 640]> concat_1_to_fp16 = const()[name = string("concat_1_to_fp16"), val = tensor<fp16, [2560, 640]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10487168)))];
42
+ tensor<fp16, [2560, 640]> concat_2_to_fp16 = const()[name = string("concat_2_to_fp16"), val = tensor<fp16, [2560, 640]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(13764032)))];
43
+ tensor<fp16, [2560]> concat_0_to_fp16 = const()[name = string("concat_0_to_fp16"), val = tensor<fp16, [2560]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(17040896)))];
44
+ tensor<fp16, [1, 1, 640]> input_cast_fp16 = transpose(perm = input_perm_0, x = var_15_cast_fp16_cast_uint16_cast_uint16)[name = string("transpose_1")];
45
+ tensor<fp16, [1, 1, 640]> y_lstm_layer_0_cast_fp16_0, tensor<fp16, [1, 640]> y_lstm_layer_0_cast_fp16_1, tensor<fp16, [1, 640]> y_lstm_layer_0_cast_fp16_2 = lstm(activation = y_lstm_layer_0_activation_0, bias = concat_0_to_fp16, cell_activation = y_lstm_layer_0_cell_activation_0, direction = y_lstm_layer_0_direction_0, initial_c = y_lstm_layer_0_lstm_c0_squeeze_cast_fp16, initial_h = y_lstm_layer_0_lstm_h0_squeeze_cast_fp16, output_sequence = y_lstm_layer_0_output_sequence_0, recurrent_activation = y_lstm_layer_0_recurrent_activation_0, weight_hh = concat_2_to_fp16, weight_ih = concat_1_to_fp16, x = input_cast_fp16)[name = string("y_lstm_layer_0_cast_fp16")];
46
+ tensor<int32, [1]> y_lstm_h0_squeeze_axes_0 = const()[name = string("y_lstm_h0_squeeze_axes_0"), val = tensor<int32, [1]>([0])];
47
+ tensor<fp16, [1, 640]> y_lstm_h0_squeeze_cast_fp16 = squeeze(axes = y_lstm_h0_squeeze_axes_0, x = split_0_cast_fp16_1)[name = string("y_lstm_h0_squeeze_cast_fp16")];
48
+ tensor<int32, [1]> y_lstm_c0_squeeze_axes_0 = const()[name = string("y_lstm_c0_squeeze_axes_0"), val = tensor<int32, [1]>([0])];
49
+ tensor<fp16, [1, 640]> y_lstm_c0_squeeze_cast_fp16 = squeeze(axes = y_lstm_c0_squeeze_axes_0, x = split_1_cast_fp16_1)[name = string("y_lstm_c0_squeeze_cast_fp16")];
50
+ string y_direction_0 = const()[name = string("y_direction_0"), val = string("forward")];
51
+ bool y_output_sequence_0 = const()[name = string("y_output_sequence_0"), val = bool(true)];
52
+ string y_recurrent_activation_0 = const()[name = string("y_recurrent_activation_0"), val = string("sigmoid")];
53
+ string y_cell_activation_0 = const()[name = string("y_cell_activation_0"), val = string("tanh")];
54
+ string y_activation_0 = const()[name = string("y_activation_0"), val = string("tanh")];
55
+ tensor<fp16, [2560, 640]> concat_4_to_fp16 = const()[name = string("concat_4_to_fp16"), val = tensor<fp16, [2560, 640]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(17046080)))];
56
+ tensor<fp16, [2560, 640]> concat_5_to_fp16 = const()[name = string("concat_5_to_fp16"), val = tensor<fp16, [2560, 640]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(20322944)))];
57
+ tensor<fp16, [2560]> concat_3_to_fp16 = const()[name = string("concat_3_to_fp16"), val = tensor<fp16, [2560]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(23599808)))];
58
+ tensor<fp16, [1, 1, 640]> y_cast_fp16_0, tensor<fp16, [1, 640]> y_cast_fp16_1, tensor<fp16, [1, 640]> y_cast_fp16_2 = lstm(activation = y_activation_0, bias = concat_3_to_fp16, cell_activation = y_cell_activation_0, direction = y_direction_0, initial_c = y_lstm_c0_squeeze_cast_fp16, initial_h = y_lstm_h0_squeeze_cast_fp16, output_sequence = y_output_sequence_0, recurrent_activation = y_recurrent_activation_0, weight_hh = concat_5_to_fp16, weight_ih = concat_4_to_fp16, x = y_lstm_layer_0_cast_fp16_0)[name = string("y_cast_fp16")];
59
+ int32 var_34_axis_0 = const()[name = string("op_34_axis_0"), val = int32(0)];
60
+ tensor<fp16, [2, 1, 640]> var_34_cast_fp16 = stack(axis = var_34_axis_0, values = (y_lstm_layer_0_cast_fp16_1, y_cast_fp16_1))[name = string("op_34_cast_fp16")];
61
+ string var_34_cast_fp16_to_fp32_dtype_0 = const()[name = string("op_34_cast_fp16_to_fp32_dtype_0"), val = string("fp32")];
62
+ int32 var_35_axis_0 = const()[name = string("op_35_axis_0"), val = int32(0)];
63
+ tensor<fp16, [2, 1, 640]> var_35_cast_fp16 = stack(axis = var_35_axis_0, values = (y_lstm_layer_0_cast_fp16_2, y_cast_fp16_2))[name = string("op_35_cast_fp16")];
64
+ string var_35_cast_fp16_to_fp32_dtype_0 = const()[name = string("op_35_cast_fp16_to_fp32_dtype_0"), val = string("fp32")];
65
+ tensor<int32, [3]> var_43 = const()[name = string("op_43"), val = tensor<int32, [3]>([1, 2, 0])];
66
+ string var_44_cast_fp16_to_fp32_dtype_0 = const()[name = string("op_44_cast_fp16_to_fp32_dtype_0"), val = string("fp32")];
67
+ tensor<fp16, [1, 640, 1]> var_44_cast_fp16 = transpose(perm = var_43, x = y_cast_fp16_0)[name = string("transpose_0")];
68
+ tensor<fp32, [1, 640, 1]> decoder = cast(dtype = var_44_cast_fp16_to_fp32_dtype_0, x = var_44_cast_fp16)[name = string("cast_2")];
69
+ tensor<fp32, [2, 1, 640]> c_out = cast(dtype = var_35_cast_fp16_to_fp32_dtype_0, x = var_35_cast_fp16)[name = string("cast_3")];
70
+ tensor<fp32, [2, 1, 640]> h_out = cast(dtype = var_34_cast_fp16_to_fp32_dtype_0, x = var_34_cast_fp16)[name = string("cast_4")];
71
+ tensor<int32, [1]> target_length_tmp = identity(x = target_length)[name = string("target_length_tmp")];
72
+ } -> (decoder, h_out, c_out);
73
+ }
Decoder.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca377d5f7432a3019b7a9d9fb2cc4c489b87934108213adac9aba2384945b6e3
3
+ size 23604992
Encoder.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01839b132224cdaff9d5e052b8b3fa7e10caae2c67afc84196b5fa38bb3dd10b
3
+ size 243
Encoder.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f18c98c393d56ba4689915f49316768528a6264a3cf1a49230df116cc688311
3
+ size 428
Encoder.mlmodelc/metadata.json ADDED
@@ -0,0 +1,106 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "metadataOutputVersion" : "3.0",
4
+ "storagePrecision" : "Mixed (Float16, Int32, Int8)",
5
+ "outputSchema" : [
6
+ {
7
+ "hasShapeFlexibility" : "0",
8
+ "isOptional" : "0",
9
+ "dataType" : "Float32",
10
+ "formattedType" : "MultiArray (Float32 1 × 1024 × 188)",
11
+ "shortDescription" : "",
12
+ "shape" : "[1, 1024, 188]",
13
+ "name" : "encoder",
14
+ "type" : "MultiArray"
15
+ },
16
+ {
17
+ "hasShapeFlexibility" : "0",
18
+ "isOptional" : "0",
19
+ "dataType" : "Int32",
20
+ "formattedType" : "MultiArray (Int32 1)",
21
+ "shortDescription" : "",
22
+ "shape" : "[1]",
23
+ "name" : "encoder_length",
24
+ "type" : "MultiArray"
25
+ }
26
+ ],
27
+ "modelParameters" : [
28
+
29
+ ],
30
+ "specificationVersion" : 9,
31
+ "mlProgramOperationTypeHistogram" : {
32
+ "Ios18.expandDims" : 17,
33
+ "Ios18.mul" : 105,
34
+ "Ios18.softmax" : 24,
35
+ "Ios18.floorDiv" : 3,
36
+ "Ios18.matmul" : 72,
37
+ "Ios18.logicalNot" : 2,
38
+ "Ios18.sigmoid" : 24,
39
+ "Split" : 24,
40
+ "Select" : 72,
41
+ "Tile" : 8,
42
+ "Ios18.add" : 178,
43
+ "Ios18.layerNorm" : 121,
44
+ "Ios18.reshape" : 145,
45
+ "Pad" : 48,
46
+ "Ios18.linear" : 195,
47
+ "Ios18.conv" : 78,
48
+ "Ios18.relu" : 3,
49
+ "Ios18.sub" : 3,
50
+ "Ios18.gelu" : 1,
51
+ "Ios18.logicalAnd" : 2,
52
+ "Ios18.silu" : 72,
53
+ "Ios18.transpose" : 174,
54
+ "Ios18.cast" : 11,
55
+ "Ios18.less" : 5,
56
+ "Ios18.sliceByIndex" : 48,
57
+ "Ios18.constexprBlockwiseShiftScale" : 297
58
+ },
59
+ "computePrecision" : "Mixed (Float16, Float32, Int32)",
60
+ "isUpdatable" : "0",
61
+ "stateSchema" : [
62
+
63
+ ],
64
+ "availability" : {
65
+ "macOS" : "15.0",
66
+ "tvOS" : "18.0",
67
+ "visionOS" : "2.0",
68
+ "watchOS" : "11.0",
69
+ "iOS" : "18.0",
70
+ "macCatalyst" : "18.0"
71
+ },
72
+ "modelType" : {
73
+ "name" : "MLModelType_mlProgram"
74
+ },
75
+ "userDefinedMetadata" : {
76
+ "com.github.apple.coremltools.conversion_date" : "2026-11-01",
77
+ "com.github.apple.coremltools.source" : "torch==2.15.0",
78
+ "com.github.apple.coremltools.version" : "9.0",
79
+ "com.github.apple.coremltools.source_dialect" : "TorchScript"
80
+ },
81
+ "inputSchema" : [
82
+ {
83
+ "hasShapeFlexibility" : "0",
84
+ "isOptional" : "0",
85
+ "dataType" : "Float32",
86
+ "formattedType" : "MultiArray (Float32 1 × 128 × 1501)",
87
+ "shortDescription" : "",
88
+ "shape" : "[1, 128, 1501]",
89
+ "name" : "mel",
90
+ "type" : "MultiArray"
91
+ },
92
+ {
93
+ "hasShapeFlexibility" : "0",
94
+ "isOptional" : "0",
95
+ "dataType" : "Int32",
96
+ "formattedType" : "MultiArray (Int32 1)",
97
+ "shortDescription" : "",
98
+ "shape" : "[1]",
99
+ "name" : "mel_length",
100
+ "type" : "MultiArray"
101
+ }
102
+ ],
103
+ "generatedClassName" : "Encoder",
104
+ "method" : "predict"
105
+ }
106
+ ]
Encoder.mlmodelc/model.mil ADDED
The diff for this file is too large to render. See raw diff
 
Encoder.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9987e515d3467b9ad6fccc9ca3b15b9f9379b5d3351064aed101b30b3e0314d
3
+ size 596344384
JointDecision.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8714d5844e15858042bb80c8f740d2d9be793d2360a509a07df08568de038135
3
+ size 243
JointDecision.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a86b62650d03a4125daf9f5de0570d6cbf552b3194cab545e5781e86a70774ed
3
+ size 463
JointDecision.mlmodelc/metadata.json ADDED
@@ -0,0 +1,99 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "metadataOutputVersion" : "3.0",
4
+ "storagePrecision" : "Float16",
5
+ "outputSchema" : [
6
+ {
7
+ "hasShapeFlexibility" : "0",
8
+ "isOptional" : "0",
9
+ "dataType" : "Int32",
10
+ "formattedType" : "MultiArray (Int32 1 × 1 × 1)",
11
+ "shortDescription" : "",
12
+ "shape" : "[1, 1, 1]",
13
+ "name" : "token_id",
14
+ "type" : "MultiArray"
15
+ },
16
+ {
17
+ "hasShapeFlexibility" : "0",
18
+ "isOptional" : "0",
19
+ "dataType" : "Float32",
20
+ "formattedType" : "MultiArray (Float32 1 × 1 × 1)",
21
+ "shortDescription" : "",
22
+ "shape" : "[1, 1, 1]",
23
+ "name" : "token_prob",
24
+ "type" : "MultiArray"
25
+ },
26
+ {
27
+ "hasShapeFlexibility" : "0",
28
+ "isOptional" : "0",
29
+ "dataType" : "Int32",
30
+ "formattedType" : "MultiArray (Int32 1 × 1 × 1)",
31
+ "shortDescription" : "",
32
+ "shape" : "[1, 1, 1]",
33
+ "name" : "duration",
34
+ "type" : "MultiArray"
35
+ }
36
+ ],
37
+ "modelParameters" : [
38
+
39
+ ],
40
+ "specificationVersion" : 9,
41
+ "mlProgramOperationTypeHistogram" : {
42
+ "Ios16.reduceMax" : 1,
43
+ "Ios18.relu" : 1,
44
+ "Ios18.reduceArgmax" : 2,
45
+ "Ios18.linear" : 3,
46
+ "Ios18.transpose" : 2,
47
+ "Ios18.cast" : 3,
48
+ "Ios18.sliceByIndex" : 2,
49
+ "Ios18.add" : 1,
50
+ "Ios18.softmax" : 1
51
+ },
52
+ "computePrecision" : "Mixed (Float16, Float32, Int32)",
53
+ "isUpdatable" : "0",
54
+ "stateSchema" : [
55
+
56
+ ],
57
+ "availability" : {
58
+ "macOS" : "15.0",
59
+ "tvOS" : "18.0",
60
+ "visionOS" : "2.0",
61
+ "watchOS" : "11.0",
62
+ "iOS" : "18.0",
63
+ "macCatalyst" : "18.0"
64
+ },
65
+ "modelType" : {
66
+ "name" : "MLModelType_mlProgram"
67
+ },
68
+ "userDefinedMetadata" : {
69
+ "com.github.apple.coremltools.conversion_date" : "2026-11-01",
70
+ "com.github.apple.coremltools.source" : "torch==2.15.0",
71
+ "com.github.apple.coremltools.version" : "9.0",
72
+ "com.github.apple.coremltools.source_dialect" : "TorchScript"
73
+ },
74
+ "inputSchema" : [
75
+ {
76
+ "hasShapeFlexibility" : "0",
77
+ "isOptional" : "0",
78
+ "dataType" : "Float32",
79
+ "formattedType" : "MultiArray (Float32 1 × 1024 × 1)",
80
+ "shortDescription" : "",
81
+ "shape" : "[1, 1024, 1]",
82
+ "name" : "encoder_step",
83
+ "type" : "MultiArray"
84
+ },
85
+ {
86
+ "hasShapeFlexibility" : "0",
87
+ "isOptional" : "0",
88
+ "dataType" : "Float32",
89
+ "formattedType" : "MultiArray (Float32 1 × 640 × 1)",
90
+ "shortDescription" : "",
91
+ "shape" : "[1, 640, 1]",
92
+ "name" : "decoder_step",
93
+ "type" : "MultiArray"
94
+ }
95
+ ],
96
+ "generatedClassName" : "JointDecision",
97
+ "method" : "predict"
98
+ }
99
+ ]
JointDecision.mlmodelc/model.mil ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ program(1.3)
2
+ [buildInfo = dict<string, string>({{"coremlc-component-MIL", "3600.16.1"}, {"coremlc-version", "3600.25.1"}, {"coremltools-component-torch", "2.15.0"}, {"coremltools-source-dialect", "TorchScript"}, {"coremltools-version", "9.0"}})]
3
+ {
4
+ func main<ios18>(tensor<fp32, [1, 640, 1]> decoder_step, tensor<fp32, [1, 1024, 1]> encoder_step) {
5
+ tensor<int32, [3]> input_1_perm_0 = const()[name = string("input_1_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
6
+ string encoder_step_to_fp16_dtype_0 = const()[name = string("encoder_step_to_fp16_dtype_0"), val = string("fp16")];
7
+ tensor<fp16, [640, 1024]> enc_weight_to_fp16 = const()[name = string("enc_weight_to_fp16"), val = tensor<fp16, [640, 1024]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(64)))];
8
+ tensor<fp16, [640]> enc_bias_to_fp16 = const()[name = string("enc_bias_to_fp16"), val = tensor<fp16, [640]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(1310848)))];
9
+ tensor<fp16, [1, 1024, 1]> encoder_step_to_fp16 = cast(dtype = encoder_step_to_fp16_dtype_0, x = encoder_step)[name = string("cast_2")];
10
+ tensor<fp16, [1, 1, 1024]> input_1_cast_fp16 = transpose(perm = input_1_perm_0, x = encoder_step_to_fp16)[name = string("transpose_1")];
11
+ tensor<fp16, [1, 1, 640]> linear_0_cast_fp16 = linear(bias = enc_bias_to_fp16, weight = enc_weight_to_fp16, x = input_1_cast_fp16)[name = string("linear_0_cast_fp16")];
12
+ tensor<int32, [3]> input_3_perm_0 = const()[name = string("input_3_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
13
+ string decoder_step_to_fp16_dtype_0 = const()[name = string("decoder_step_to_fp16_dtype_0"), val = string("fp16")];
14
+ tensor<fp16, [640, 640]> pred_weight_to_fp16 = const()[name = string("pred_weight_to_fp16"), val = tensor<fp16, [640, 640]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(1312192)))];
15
+ tensor<fp16, [640]> pred_bias_to_fp16 = const()[name = string("pred_bias_to_fp16"), val = tensor<fp16, [640]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(2131456)))];
16
+ tensor<fp16, [1, 640, 1]> decoder_step_to_fp16 = cast(dtype = decoder_step_to_fp16_dtype_0, x = decoder_step)[name = string("cast_1")];
17
+ tensor<fp16, [1, 1, 640]> input_3_cast_fp16 = transpose(perm = input_3_perm_0, x = decoder_step_to_fp16)[name = string("transpose_0")];
18
+ tensor<fp16, [1, 1, 640]> linear_1_cast_fp16 = linear(bias = pred_bias_to_fp16, weight = pred_weight_to_fp16, x = input_3_cast_fp16)[name = string("linear_1_cast_fp16")];
19
+ tensor<fp16, [1, 1, 640]> var_19_cast_fp16 = add(x = linear_0_cast_fp16, y = linear_1_cast_fp16)[name = string("op_19_cast_fp16")];
20
+ tensor<fp16, [1, 1, 640]> input_cast_fp16 = relu(x = var_19_cast_fp16)[name = string("input_cast_fp16")];
21
+ tensor<fp16, [8198, 640]> head_weight_to_fp16 = const()[name = string("head_weight_to_fp16"), val = tensor<fp16, [8198, 640]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(2132800)))];
22
+ tensor<fp16, [8198]> head_bias_to_fp16 = const()[name = string("head_bias_to_fp16"), val = tensor<fp16, [8198]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(12626304)))];
23
+ tensor<fp16, [1, 1, 8198]> linear_2_cast_fp16 = linear(bias = head_bias_to_fp16, weight = head_weight_to_fp16, x = input_cast_fp16)[name = string("linear_2_cast_fp16")];
24
+ tensor<int32, [3]> tokens_begin_0 = const()[name = string("tokens_begin_0"), val = tensor<int32, [3]>([0, 0, 0])];
25
+ tensor<int32, [3]> tokens_end_0 = const()[name = string("tokens_end_0"), val = tensor<int32, [3]>([1, 1, 8193])];
26
+ tensor<bool, [3]> tokens_end_mask_0 = const()[name = string("tokens_end_mask_0"), val = tensor<bool, [3]>([true, true, false])];
27
+ tensor<fp16, [1, 1, 8193]> tokens_cast_fp16 = slice_by_index(begin = tokens_begin_0, end = tokens_end_0, end_mask = tokens_end_mask_0, x = linear_2_cast_fp16)[name = string("tokens_cast_fp16")];
28
+ tensor<int32, [3]> durations_begin_0 = const()[name = string("durations_begin_0"), val = tensor<int32, [3]>([0, 0, 8193])];
29
+ tensor<int32, [3]> durations_end_0 = const()[name = string("durations_end_0"), val = tensor<int32, [3]>([1, 1, 8198])];
30
+ tensor<bool, [3]> durations_end_mask_0 = const()[name = string("durations_end_mask_0"), val = tensor<bool, [3]>([true, true, true])];
31
+ tensor<fp16, [1, 1, 5]> durations_cast_fp16 = slice_by_index(begin = durations_begin_0, end = durations_end_0, end_mask = durations_end_mask_0, x = linear_2_cast_fp16)[name = string("durations_cast_fp16")];
32
+ int32 var_36_axis_0 = const()[name = string("op_36_axis_0"), val = int32(-1)];
33
+ bool var_36_keep_dims_0 = const()[name = string("op_36_keep_dims_0"), val = bool(true)];
34
+ string var_36_output_dtype_0 = const()[name = string("op_36_output_dtype_0"), val = string("int32")];
35
+ tensor<int32, [1, 1, 1]> token_id = reduce_argmax(axis = var_36_axis_0, keep_dims = var_36_keep_dims_0, output_dtype = var_36_output_dtype_0, x = tokens_cast_fp16)[name = string("op_36_cast_fp16")];
36
+ int32 var_42 = const()[name = string("op_42"), val = int32(-1)];
37
+ tensor<fp16, [1, 1, 8193]> var_44_cast_fp16 = softmax(axis = var_42, x = tokens_cast_fp16)[name = string("op_44_cast_fp16")];
38
+ bool var_47 = const()[name = string("op_47"), val = bool(true)];
39
+ tensor<int32, [1]> var_48_axes_0 = const()[name = string("op_48_axes_0"), val = tensor<int32, [1]>([-1])];
40
+ tensor<fp16, [1, 1, 1]> var_48_cast_fp16 = reduce_max(axes = var_48_axes_0, keep_dims = var_47, x = var_44_cast_fp16)[name = string("op_48_cast_fp16")];
41
+ string var_48_cast_fp16_to_fp32_dtype_0 = const()[name = string("op_48_cast_fp16_to_fp32_dtype_0"), val = string("fp32")];
42
+ int32 var_51_axis_0 = const()[name = string("op_51_axis_0"), val = int32(-1)];
43
+ bool var_51_keep_dims_0 = const()[name = string("op_51_keep_dims_0"), val = bool(true)];
44
+ string var_51_output_dtype_0 = const()[name = string("op_51_output_dtype_0"), val = string("int32")];
45
+ tensor<int32, [1, 1, 1]> duration = reduce_argmax(axis = var_51_axis_0, keep_dims = var_51_keep_dims_0, output_dtype = var_51_output_dtype_0, x = durations_cast_fp16)[name = string("op_51_cast_fp16")];
46
+ tensor<fp32, [1, 1, 1]> token_prob = cast(dtype = var_48_cast_fp16_to_fp32_dtype_0, x = var_48_cast_fp16)[name = string("cast_0")];
47
+ } -> (token_id, token_prob, duration);
48
+ }
JointDecision.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f310b85b82341c53ec383025ab094a4e462ee1c592e1ad7c6bfe39cff66ca25
3
+ size 12642764
LICENSE ADDED
@@ -0,0 +1,396 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Attribution 4.0 International
2
+
3
+ =======================================================================
4
+
5
+ Creative Commons Corporation ("Creative Commons") is not a law firm and
6
+ does not provide legal services or legal advice. Distribution of
7
+ Creative Commons public licenses does not create a lawyer-client or
8
+ other relationship. Creative Commons makes its licenses and related
9
+ information available on an "as-is" basis. Creative Commons gives no
10
+ warranties regarding its licenses, any material licensed under their
11
+ terms and conditions, or any related information. Creative Commons
12
+ disclaims all liability for damages resulting from their use to the
13
+ fullest extent possible.
14
+
15
+ Using Creative Commons Public Licenses
16
+
17
+ Creative Commons public licenses provide a standard set of terms and
18
+ conditions that creators and other rights holders may use to share
19
+ original works of authorship and other material subject to copyright
20
+ and certain other rights specified in the public license below. The
21
+ following considerations are for informational purposes only, are not
22
+ exhaustive, and do not form part of our licenses.
23
+
24
+ Considerations for licensors: Our public licenses are
25
+ intended for use by those authorized to give the public
26
+ permission to use material in ways otherwise restricted by
27
+ copyright and certain other rights. Our licenses are
28
+ irrevocable. Licensors should read and understand the terms
29
+ and conditions of the license they choose before applying it.
30
+ Licensors should also secure all rights necessary before
31
+ applying our licenses so that the public can reuse the
32
+ material as expected. Licensors should clearly mark any
33
+ material not subject to the license. This includes other CC-
34
+ licensed material, or material used under an exception or
35
+ limitation to copyright. More considerations for licensors:
36
+ wiki.creativecommons.org/Considerations_for_licensors
37
+
38
+ Considerations for the public: By using one of our public
39
+ licenses, a licensor grants the public permission to use the
40
+ licensed material under specified terms and conditions. If
41
+ the licensor's permission is not necessary for any reason--for
42
+ example, because of any applicable exception or limitation to
43
+ copyright--then that use is not regulated by the license. Our
44
+ licenses grant only permissions under copyright and certain
45
+ other rights that a licensor has authority to grant. Use of
46
+ the licensed material may still be restricted for other
47
+ reasons, including because others have copyright or other
48
+ rights in the material. A licensor may make special requests,
49
+ such as asking that all changes be marked or described.
50
+ Although not required by our licenses, you are encouraged to
51
+ respect those requests where reasonable. More considerations
52
+ for the public:
53
+ wiki.creativecommons.org/Considerations_for_licensees
54
+
55
+ =======================================================================
56
+
57
+ Creative Commons Attribution 4.0 International Public License
58
+
59
+ By exercising the Licensed Rights (defined below), You accept and agree
60
+ to be bound by the terms and conditions of this Creative Commons
61
+ Attribution 4.0 International Public License ("Public License"). To the
62
+ extent this Public License may be interpreted as a contract, You are
63
+ granted the Licensed Rights in consideration of Your acceptance of
64
+ these terms and conditions, and the Licensor grants You such rights in
65
+ consideration of benefits the Licensor receives from making the
66
+ Licensed Material available under these terms and conditions.
67
+
68
+
69
+ Section 1 -- Definitions.
70
+
71
+ a. Adapted Material means material subject to Copyright and Similar
72
+ Rights that is derived from or based upon the Licensed Material
73
+ and in which the Licensed Material is translated, altered,
74
+ arranged, transformed, or otherwise modified in a manner requiring
75
+ permission under the Copyright and Similar Rights held by the
76
+ Licensor. For purposes of this Public License, where the Licensed
77
+ Material is a musical work, performance, or sound recording,
78
+ Adapted Material is always produced where the Licensed Material is
79
+ synched in timed relation with a moving image.
80
+
81
+ b. Adapter's License means the license You apply to Your Copyright
82
+ and Similar Rights in Your contributions to Adapted Material in
83
+ accordance with the terms and conditions of this Public License.
84
+
85
+ c. Copyright and Similar Rights means copyright and/or similar rights
86
+ closely related to copyright including, without limitation,
87
+ performance, broadcast, sound recording, and Sui Generis Database
88
+ Rights, without regard to how the rights are labeled or
89
+ categorized. For purposes of this Public License, the rights
90
+ specified in Section 2(b)(1)-(2) are not Copyright and Similar
91
+ Rights.
92
+
93
+ d. Effective Technological Measures means those measures that, in the
94
+ absence of proper authority, may not be circumvented under laws
95
+ fulfilling obligations under Article 11 of the WIPO Copyright
96
+ Treaty adopted on December 20, 1996, and/or similar international
97
+ agreements.
98
+
99
+ e. Exceptions and Limitations means fair use, fair dealing, and/or
100
+ any other exception or limitation to Copyright and Similar Rights
101
+ that applies to Your use of the Licensed Material.
102
+
103
+ f. Licensed Material means the artistic or literary work, database,
104
+ or other material to which the Licensor applied this Public
105
+ License.
106
+
107
+ g. Licensed Rights means the rights granted to You subject to the
108
+ terms and conditions of this Public License, which are limited to
109
+ all Copyright and Similar Rights that apply to Your use of the
110
+ Licensed Material and that the Licensor has authority to license.
111
+
112
+ h. Licensor means the individual(s) or entity(ies) granting rights
113
+ under this Public License.
114
+
115
+ i. Share means to provide material to the public by any means or
116
+ process that requires permission under the Licensed Rights, such
117
+ as reproduction, public display, public performance, distribution,
118
+ dissemination, communication, or importation, and to make material
119
+ available to the public including in ways that members of the
120
+ public may access the material from a place and at a time
121
+ individually chosen by them.
122
+
123
+ j. Sui Generis Database Rights means rights other than copyright
124
+ resulting from Directive 96/9/EC of the European Parliament and of
125
+ the Council of 11 March 1996 on the legal protection of databases,
126
+ as amended and/or succeeded, as well as other essentially
127
+ equivalent rights anywhere in the world.
128
+
129
+ k. You means the individual or entity exercising the Licensed Rights
130
+ under this Public License. Your has a corresponding meaning.
131
+
132
+
133
+ Section 2 -- Scope.
134
+
135
+ a. License grant.
136
+
137
+ 1. Subject to the terms and conditions of this Public License,
138
+ the Licensor hereby grants You a worldwide, royalty-free,
139
+ non-sublicensable, non-exclusive, irrevocable license to
140
+ exercise the Licensed Rights in the Licensed Material to:
141
+
142
+ a. reproduce and Share the Licensed Material, in whole or
143
+ in part; and
144
+
145
+ b. produce, reproduce, and Share Adapted Material.
146
+
147
+ 2. Exceptions and Limitations. For the avoidance of doubt, where
148
+ Exceptions and Limitations apply to Your use, this Public
149
+ License does not apply, and You do not need to comply with
150
+ its terms and conditions.
151
+
152
+ 3. Term. The term of this Public License is specified in Section
153
+ 6(a).
154
+
155
+ 4. Media and formats; technical modifications allowed. The
156
+ Licensor authorizes You to exercise the Licensed Rights in
157
+ all media and formats whether now known or hereafter created,
158
+ and to make technical modifications necessary to do so. The
159
+ Licensor waives and/or agrees not to assert any right or
160
+ authority to forbid You from making technical modifications
161
+ necessary to exercise the Licensed Rights, including
162
+ technical modifications necessary to circumvent Effective
163
+ Technological Measures. For purposes of this Public License,
164
+ simply making modifications authorized by this Section 2(a)
165
+ (4) never produces Adapted Material.
166
+
167
+ 5. Downstream recipients.
168
+
169
+ a. Offer from the Licensor -- Licensed Material. Every
170
+ recipient of the Licensed Material automatically
171
+ receives an offer from the Licensor to exercise the
172
+ Licensed Rights under the terms and conditions of this
173
+ Public License.
174
+
175
+ b. No downstream restrictions. You may not offer or impose
176
+ any additional or different terms or conditions on, or
177
+ apply any Effective Technological Measures to, the
178
+ Licensed Material if doing so restricts exercise of the
179
+ Licensed Rights by any recipient of the Licensed
180
+ Material.
181
+
182
+ 6. No endorsement. Nothing in this Public License constitutes or
183
+ may be construed as permission to assert or imply that You
184
+ are, or that Your use of the Licensed Material is, connected
185
+ with, or sponsored, endorsed, or granted official status by,
186
+ the Licensor or others designated to receive attribution as
187
+ provided in Section 3(a)(1)(A)(i).
188
+
189
+ b. Other rights.
190
+
191
+ 1. Moral rights, such as the right of integrity, are not
192
+ licensed under this Public License, nor are publicity,
193
+ privacy, and/or other similar personality rights; however, to
194
+ the extent possible, the Licensor waives and/or agrees not to
195
+ assert any such rights held by the Licensor to the limited
196
+ extent necessary to allow You to exercise the Licensed
197
+ Rights, but not otherwise.
198
+
199
+ 2. Patent and trademark rights are not licensed under this
200
+ Public License.
201
+
202
+ 3. To the extent possible, the Licensor waives any right to
203
+ collect royalties from You for the exercise of the Licensed
204
+ Rights, whether directly or through a collecting society
205
+ under any voluntary or waivable statutory or compulsory
206
+ licensing scheme. In all other cases the Licensor expressly
207
+ reserves any right to collect such royalties.
208
+
209
+
210
+ Section 3 -- License Conditions.
211
+
212
+ Your exercise of the Licensed Rights is expressly made subject to the
213
+ following conditions.
214
+
215
+ a. Attribution.
216
+
217
+ 1. If You Share the Licensed Material (including in modified
218
+ form), You must:
219
+
220
+ a. retain the following if it is supplied by the Licensor
221
+ with the Licensed Material:
222
+
223
+ i. identification of the creator(s) of the Licensed
224
+ Material and any others designated to receive
225
+ attribution, in any reasonable manner requested by
226
+ the Licensor (including by pseudonym if
227
+ designated);
228
+
229
+ ii. a copyright notice;
230
+
231
+ iii. a notice that refers to this Public License;
232
+
233
+ iv. a notice that refers to the disclaimer of
234
+ warranties;
235
+
236
+ v. a URI or hyperlink to the Licensed Material to the
237
+ extent reasonably practicable;
238
+
239
+ b. indicate if You modified the Licensed Material and
240
+ retain an indication of any previous modifications; and
241
+
242
+ c. indicate the Licensed Material is licensed under this
243
+ Public License, and include the text of, or the URI or
244
+ hyperlink to, this Public License.
245
+
246
+ 2. You may satisfy the conditions in Section 3(a)(1) in any
247
+ reasonable manner based on the medium, means, and context in
248
+ which You Share the Licensed Material. For example, it may be
249
+ reasonable to satisfy the conditions by providing a URI or
250
+ hyperlink to a resource that includes the required
251
+ information.
252
+
253
+ 3. If requested by the Licensor, You must remove any of the
254
+ information required by Section 3(a)(1)(A) to the extent
255
+ reasonably practicable.
256
+
257
+ 4. If You Share Adapted Material You produce, the Adapter's
258
+ License You apply must not prevent recipients of the Adapted
259
+ Material from complying with this Public License.
260
+
261
+
262
+ Section 4 -- Sui Generis Database Rights.
263
+
264
+ Where the Licensed Rights include Sui Generis Database Rights that
265
+ apply to Your use of the Licensed Material:
266
+
267
+ a. for the avoidance of doubt, Section 2(a)(1) grants You the right
268
+ to extract, reuse, reproduce, and Share all or a substantial
269
+ portion of the contents of the database;
270
+
271
+ b. if You include all or a substantial portion of the database
272
+ contents in a database in which You have Sui Generis Database
273
+ Rights, then the database in which You have Sui Generis Database
274
+ Rights (but not its individual contents) is Adapted Material; and
275
+
276
+ c. You must comply with the conditions in Section 3(a) if You Share
277
+ all or a substantial portion of the contents of the database.
278
+
279
+ For the avoidance of doubt, this Section 4 supplements and does not
280
+ replace Your obligations under this Public License where the Licensed
281
+ Rights include other Copyright and Similar Rights.
282
+
283
+
284
+ Section 5 -- Disclaimer of Warranties and Limitation of Liability.
285
+
286
+ a. UNLESS OTHERWISE SEPARATELY UNDERTAKEN BY THE LICENSOR, TO THE
287
+ EXTENT POSSIBLE, THE LICENSOR OFFERS THE LICENSED MATERIAL AS-IS
288
+ AND AS-AVAILABLE, AND MAKES NO REPRESENTATIONS OR WARRANTIES OF
289
+ ANY KIND CONCERNING THE LICENSED MATERIAL, WHETHER EXPRESS,
290
+ IMPLIED, STATUTORY, OR OTHER. THIS INCLUDES, WITHOUT LIMITATION,
291
+ WARRANTIES OF TITLE, MERCHANTABILITY, FITNESS FOR A PARTICULAR
292
+ PURPOSE, NON-INFRINGEMENT, ABSENCE OF LATENT OR OTHER DEFECTS,
293
+ ACCURACY, OR THE PRESENCE OR ABSENCE OF ERRORS, WHETHER OR NOT
294
+ KNOWN OR DISCOVERABLE. WHERE DISCLAIMERS OF WARRANTIES ARE NOT
295
+ ALLOWED IN FULL OR IN PART, THIS DISCLAIMER MAY NOT APPLY TO YOU.
296
+
297
+ b. TO THE EXTENT POSSIBLE, IN NO EVENT WILL THE LICENSOR BE LIABLE
298
+ TO YOU ON ANY LEGAL THEORY (INCLUDING, WITHOUT LIMITATION,
299
+ NEGLIGENCE) OR OTHERWISE FOR ANY DIRECT, SPECIAL, INDIRECT,
300
+ INCIDENTAL, CONSEQUENTIAL, PUNITIVE, EXEMPLARY, OR OTHER LOSSES,
301
+ COSTS, EXPENSES, OR DAMAGES ARISING OUT OF THIS PUBLIC LICENSE OR
302
+ USE OF THE LICENSED MATERIAL, EVEN IF THE LICENSOR HAS BEEN
303
+ ADVISED OF THE POSSIBILITY OF SUCH LOSSES, COSTS, EXPENSES, OR
304
+ DAMAGES. WHERE A LIMITATION OF LIABILITY IS NOT ALLOWED IN FULL OR
305
+ IN PART, THIS LIMITATION MAY NOT APPLY TO YOU.
306
+
307
+ c. The disclaimer of warranties and limitation of liability provided
308
+ above shall be interpreted in a manner that, to the extent
309
+ possible, most closely approximates an absolute disclaimer and
310
+ waiver of all liability.
311
+
312
+
313
+ Section 6 -- Term and Termination.
314
+
315
+ a. This Public License applies for the term of the Copyright and
316
+ Similar Rights licensed here. However, if You fail to comply with
317
+ this Public License, then Your rights under this Public License
318
+ terminate automatically.
319
+
320
+ b. Where Your right to use the Licensed Material has terminated under
321
+ Section 6(a), it reinstates:
322
+
323
+ 1. automatically as of the date the violation is cured, provided
324
+ it is cured within 30 days of Your discovery of the
325
+ violation; or
326
+
327
+ 2. upon express reinstatement by the Licensor.
328
+
329
+ For the avoidance of doubt, this Section 6(b) does not affect any
330
+ right the Licensor may have to seek remedies for Your violations
331
+ of this Public License.
332
+
333
+ c. For the avoidance of doubt, the Licensor may also offer the
334
+ Licensed Material under separate terms or conditions or stop
335
+ distributing the Licensed Material at any time; however, doing so
336
+ will not terminate this Public License.
337
+
338
+ d. Sections 1, 5, 6, 7, and 8 survive termination of this Public
339
+ License.
340
+
341
+
342
+ Section 7 -- Other Terms and Conditions.
343
+
344
+ a. The Licensor shall not be bound by any additional or different
345
+ terms or conditions communicated by You unless expressly agreed.
346
+
347
+ b. Any arrangements, understandings, or agreements regarding the
348
+ Licensed Material not stated herein are separate from and
349
+ independent of the terms and conditions of this Public License.
350
+
351
+
352
+ Section 8 -- Interpretation.
353
+
354
+ a. For the avoidance of doubt, this Public License does not, and
355
+ shall not be interpreted to, reduce, limit, restrict, or impose
356
+ conditions on any use of the Licensed Material that could lawfully
357
+ be made without permission under this Public License.
358
+
359
+ b. To the extent possible, if any provision of this Public License is
360
+ deemed unenforceable, it shall be automatically reformed to the
361
+ minimum extent necessary to make it enforceable. If the provision
362
+ cannot be reformed, it shall be severed from this Public License
363
+ without affecting the enforceability of the remaining terms and
364
+ conditions.
365
+
366
+ c. No term or condition of this Public License will be waived and no
367
+ failure to comply consented to unless expressly agreed to by the
368
+ Licensor.
369
+
370
+ d. Nothing in this Public License constitutes or may be interpreted
371
+ as a limitation upon, or waiver of, any privileges and immunities
372
+ that apply to the Licensor or You, including from the legal
373
+ processes of any jurisdiction or authority.
374
+
375
+
376
+ =======================================================================
377
+
378
+ Creative Commons is not a party to its public
379
+ licenses. Notwithstanding, Creative Commons may elect to apply one of
380
+ its public licenses to material it publishes and in those instances
381
+ will be considered the “Licensor.” The text of the Creative Commons
382
+ public licenses is dedicated to the public domain under the CC0 Public
383
+ Domain Dedication. Except for the limited purpose of indicating that
384
+ material is shared under a Creative Commons public license or as
385
+ otherwise permitted by the Creative Commons policies published at
386
+ creativecommons.org/policies, Creative Commons does not authorize the
387
+ use of the trademark "Creative Commons" or any other trademark or logo
388
+ of Creative Commons without its prior written consent including,
389
+ without limitation, in connection with any unauthorized modifications
390
+ to any of its public licenses or any other arrangements,
391
+ understandings, or agreements concerning use of licensed material. For
392
+ the avoidance of doubt, this paragraph does not form part of the
393
+ public licenses.
394
+
395
+ Creative Commons may be contacted at creativecommons.org.
396
+
NOTICE ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ Fluid Speech Pro by FluidVoice (https://huggingface.co/FluidVoice-Labs/fluid-speech-pro), CC BY 4.0.
2
+ Built on NVIDIA Parakeet TDT 0.6B v3 and Moondream Parakeet Ultra, both CC BY 4.0.
3
+ Converted to Core ML by TalkMate (8-bit encoder weights). The preprocessor is FluidInference's
4
+ Parakeet TDT 0.6B v3 Core ML preprocessor (CC BY 4.0).
Preprocessor.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c9beeb989c8d66f8be11df59bc6df277ec76cee404f6865b46243835ef562f6d
3
+ size 243
Preprocessor.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbde3f2300842c1fd51ef3ff948a0bcffe65ffd2dca10707f2509f32c1d65b1d
3
+ size 486
Preprocessor.mlmodelc/metadata.json ADDED
@@ -0,0 +1,105 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "metadataOutputVersion" : "3.0",
4
+ "shortDescription" : "int8-linear quantized - preprocessor",
5
+ "outputSchema" : [
6
+ {
7
+ "hasShapeFlexibility" : "0",
8
+ "isOptional" : "0",
9
+ "dataType" : "Float32",
10
+ "formattedType" : "MultiArray (Float32 1 × 128 × 1501)",
11
+ "shortDescription" : "",
12
+ "shape" : "[1, 128, 1501]",
13
+ "name" : "mel",
14
+ "type" : "MultiArray"
15
+ },
16
+ {
17
+ "hasShapeFlexibility" : "0",
18
+ "isOptional" : "0",
19
+ "dataType" : "Int32",
20
+ "formattedType" : "MultiArray (Int32 1)",
21
+ "shortDescription" : "",
22
+ "shape" : "[1]",
23
+ "name" : "mel_length",
24
+ "type" : "MultiArray"
25
+ }
26
+ ],
27
+ "storagePrecision" : "Int8",
28
+ "modelParameters" : [
29
+
30
+ ],
31
+ "author" : "Fluid Inference",
32
+ "specificationVersion" : 8,
33
+ "mlProgramOperationTypeHistogram" : {
34
+ "Ios17.reshape" : 2,
35
+ "Ios17.matmul" : 1,
36
+ "Select" : 3,
37
+ "Ios17.expandDims" : 8,
38
+ "Ios17.add" : 4,
39
+ "Tile" : 1,
40
+ "Ios17.sliceByIndex" : 3,
41
+ "Ios16.reduceSum" : 4,
42
+ "Pad" : 1,
43
+ "Ios17.log" : 1,
44
+ "Ios16.constexprAffineDequantize" : 4,
45
+ "Ios17.less" : 1,
46
+ "Ios17.sub" : 4,
47
+ "Ios17.conv" : 2,
48
+ "Ios17.pow" : 2,
49
+ "Ios17.cast" : 5,
50
+ "Stack" : 1,
51
+ "Ios17.concat" : 1,
52
+ "Ios17.floorDiv" : 1,
53
+ "Ios17.realDiv" : 3,
54
+ "Ios17.sqrt" : 1,
55
+ "Ios17.greaterEqual" : 1,
56
+ "Ios17.mul" : 1
57
+ },
58
+ "computePrecision" : "Mixed (Float16, Float32, Int32)",
59
+ "isUpdatable" : "0",
60
+ "stateSchema" : [
61
+
62
+ ],
63
+ "availability" : {
64
+ "macOS" : "14.0",
65
+ "tvOS" : "17.0",
66
+ "visionOS" : "1.0",
67
+ "watchOS" : "10.0",
68
+ "iOS" : "17.0",
69
+ "macCatalyst" : "17.0"
70
+ },
71
+ "modelType" : {
72
+ "name" : "MLModelType_mlProgram"
73
+ },
74
+ "inputSchema" : [
75
+ {
76
+ "hasShapeFlexibility" : "0",
77
+ "isOptional" : "0",
78
+ "dataType" : "Float32",
79
+ "formattedType" : "MultiArray (Float32 1 × 240000)",
80
+ "shortDescription" : "",
81
+ "shape" : "[1, 240000]",
82
+ "name" : "audio_signal",
83
+ "type" : "MultiArray"
84
+ },
85
+ {
86
+ "hasShapeFlexibility" : "0",
87
+ "isOptional" : "0",
88
+ "dataType" : "Int32",
89
+ "formattedType" : "MultiArray (Int32 1)",
90
+ "shortDescription" : "",
91
+ "shape" : "[1]",
92
+ "name" : "audio_length",
93
+ "type" : "MultiArray"
94
+ }
95
+ ],
96
+ "userDefinedMetadata" : {
97
+ "com.github.apple.coremltools.conversion_date" : "2025-09-19",
98
+ "com.github.apple.coremltools.source" : "torch==2.7.0",
99
+ "com.github.apple.coremltools.version" : "9.0b1",
100
+ "com.github.apple.coremltools.source_dialect" : "TorchScript"
101
+ },
102
+ "generatedClassName" : "parakeet_preprocessor",
103
+ "method" : "predict"
104
+ }
105
+ ]
Preprocessor.mlmodelc/model.mil ADDED
@@ -0,0 +1,125 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ program(1.0)
2
+ [buildInfo = dict<tensor<string, []>, tensor<string, []>>({{"coremlc-component-MIL", "3500.14.1"}, {"coremlc-version", "3500.32.1"}})]
3
+ {
4
+ func main<ios17>(tensor<int32, [1]> audio_length, tensor<fp32, [1, 240000]> audio_signal) {
5
+ tensor<int32, []> var_9 = const()[name = tensor<string, []>("op_9"), val = tensor<int32, []>(1)];
6
+ tensor<int32, []> var_10 = const()[name = tensor<string, []>("op_10"), val = tensor<int32, []>(160)];
7
+ tensor<int32, []> var_34 = const()[name = tensor<string, []>("op_34"), val = tensor<int32, []>(512)];
8
+ tensor<int32, [1]> var_35 = add(x = audio_length, y = var_34)[name = tensor<string, []>("op_35")];
9
+ tensor<int32, []> var_36 = const()[name = tensor<string, []>("op_36"), val = tensor<int32, []>(512)];
10
+ tensor<int32, [1]> var_37 = sub(x = var_35, y = var_36)[name = tensor<string, []>("op_37")];
11
+ tensor<int32, [1]> floor_div_0 = floor_div(x = var_37, y = var_10)[name = tensor<string, []>("floor_div_0")];
12
+ tensor<string, []> var_38_to_fp16_dtype_0 = const()[name = tensor<string, []>("op_38_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
13
+ tensor<fp16, []> var_39_promoted_to_fp16 = const()[name = tensor<string, []>("op_39_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
14
+ tensor<fp16, [1]> floor_div_0_to_fp16 = cast(dtype = var_38_to_fp16_dtype_0, x = floor_div_0)[name = tensor<string, []>("cast_4")];
15
+ tensor<fp16, [1]> seq_len_1_cast_fp16 = add(x = floor_div_0_to_fp16, y = var_39_promoted_to_fp16)[name = tensor<string, []>("seq_len_1_cast_fp16")];
16
+ tensor<string, []> seq_len_dtype_0 = const()[name = tensor<string, []>("seq_len_dtype_0"), val = tensor<string, []>("int32")];
17
+ tensor<int32, [2]> var_43_begin_0 = const()[name = tensor<string, []>("op_43_begin_0"), val = tensor<int32, [2]>([0, 0])];
18
+ tensor<int32, [2]> var_43_end_0 = const()[name = tensor<string, []>("op_43_end_0"), val = tensor<int32, [2]>([1, 1])];
19
+ tensor<bool, [2]> var_43_end_mask_0 = const()[name = tensor<string, []>("op_43_end_mask_0"), val = tensor<bool, [2]>([true, false])];
20
+ tensor<bool, [2]> var_43_squeeze_mask_0 = const()[name = tensor<string, []>("op_43_squeeze_mask_0"), val = tensor<bool, [2]>([false, true])];
21
+ tensor<string, []> audio_signal_to_fp16_dtype_0 = const()[name = tensor<string, []>("audio_signal_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
22
+ tensor<fp16, [1, 240000]> audio_signal_to_fp16 = cast(dtype = audio_signal_to_fp16_dtype_0, x = audio_signal)[name = tensor<string, []>("cast_3")];
23
+ tensor<fp16, [1]> var_43_cast_fp16 = slice_by_index(begin = var_43_begin_0, end = var_43_end_0, end_mask = var_43_end_mask_0, squeeze_mask = var_43_squeeze_mask_0, x = audio_signal_to_fp16)[name = tensor<string, []>("op_43_cast_fp16")];
24
+ tensor<int32, [1]> var_44_axes_0 = const()[name = tensor<string, []>("op_44_axes_0"), val = tensor<int32, [1]>([1])];
25
+ tensor<fp16, [1, 1]> var_44_cast_fp16 = expand_dims(axes = var_44_axes_0, x = var_43_cast_fp16)[name = tensor<string, []>("op_44_cast_fp16")];
26
+ tensor<int32, [2]> var_46_begin_0 = const()[name = tensor<string, []>("op_46_begin_0"), val = tensor<int32, [2]>([0, 1])];
27
+ tensor<int32, [2]> var_46_end_0 = const()[name = tensor<string, []>("op_46_end_0"), val = tensor<int32, [2]>([1, 240000])];
28
+ tensor<bool, [2]> var_46_end_mask_0 = const()[name = tensor<string, []>("op_46_end_mask_0"), val = tensor<bool, [2]>([true, true])];
29
+ tensor<fp16, [1, 239999]> var_46_cast_fp16 = slice_by_index(begin = var_46_begin_0, end = var_46_end_0, end_mask = var_46_end_mask_0, x = audio_signal_to_fp16)[name = tensor<string, []>("op_46_cast_fp16")];
30
+ tensor<int32, [2]> var_48_begin_0 = const()[name = tensor<string, []>("op_48_begin_0"), val = tensor<int32, [2]>([0, 0])];
31
+ tensor<int32, [2]> var_48_end_0 = const()[name = tensor<string, []>("op_48_end_0"), val = tensor<int32, [2]>([1, 239999])];
32
+ tensor<bool, [2]> var_48_end_mask_0 = const()[name = tensor<string, []>("op_48_end_mask_0"), val = tensor<bool, [2]>([true, false])];
33
+ tensor<fp16, [1, 239999]> var_48_cast_fp16 = slice_by_index(begin = var_48_begin_0, end = var_48_end_0, end_mask = var_48_end_mask_0, x = audio_signal_to_fp16)[name = tensor<string, []>("op_48_cast_fp16")];
34
+ tensor<fp16, []> var_49_to_fp16 = const()[name = tensor<string, []>("op_49_to_fp16"), val = tensor<fp16, []>(0x1.f0cp-1)];
35
+ tensor<fp16, [1, 239999]> var_50_cast_fp16 = mul(x = var_48_cast_fp16, y = var_49_to_fp16)[name = tensor<string, []>("op_50_cast_fp16")];
36
+ tensor<fp16, [1, 239999]> var_51_cast_fp16 = sub(x = var_46_cast_fp16, y = var_50_cast_fp16)[name = tensor<string, []>("op_51_cast_fp16")];
37
+ tensor<bool, []> input_1_interleave_0 = const()[name = tensor<string, []>("input_1_interleave_0"), val = tensor<bool, []>(false)];
38
+ tensor<fp16, [1, 240000]> input_1_cast_fp16 = concat(axis = var_9, interleave = input_1_interleave_0, values = (var_44_cast_fp16, var_51_cast_fp16))[name = tensor<string, []>("input_1_cast_fp16")];
39
+ tensor<int32, [3]> var_57 = const()[name = tensor<string, []>("op_57"), val = tensor<int32, [3]>([1, 1, 240000])];
40
+ tensor<fp16, [1, 1, 240000]> input_3_cast_fp16 = reshape(shape = var_57, x = input_1_cast_fp16)[name = tensor<string, []>("input_3_cast_fp16")];
41
+ tensor<int32, [6]> input_5_pad_0 = const()[name = tensor<string, []>("input_5_pad_0"), val = tensor<int32, [6]>([0, 0, 0, 0, 256, 256])];
42
+ tensor<string, []> input_5_mode_0 = const()[name = tensor<string, []>("input_5_mode_0"), val = tensor<string, []>("reflect")];
43
+ tensor<fp16, []> const_3_to_fp16 = const()[name = tensor<string, []>("const_3_to_fp16"), val = tensor<fp16, []>(0x0p+0)];
44
+ tensor<fp16, [1, 1, 240512]> input_5_cast_fp16 = pad(constant_val = const_3_to_fp16, mode = input_5_mode_0, pad = input_5_pad_0, x = input_3_cast_fp16)[name = tensor<string, []>("input_5_cast_fp16")];
45
+ tensor<int32, [2]> var_63 = const()[name = tensor<string, []>("op_63"), val = tensor<int32, [2]>([1, 240512])];
46
+ tensor<fp16, [1, 240512]> input_cast_fp16 = reshape(shape = var_63, x = input_5_cast_fp16)[name = tensor<string, []>("input_cast_fp16")];
47
+ tensor<int32, [1]> expand_dims_5 = const()[name = tensor<string, []>("expand_dims_5"), val = tensor<int32, [1]>([160])];
48
+ tensor<int32, [1]> expand_dims_6_axes_0 = const()[name = tensor<string, []>("expand_dims_6_axes_0"), val = tensor<int32, [1]>([1])];
49
+ tensor<fp16, [1, 1, 240512]> expand_dims_6_cast_fp16 = expand_dims(axes = expand_dims_6_axes_0, x = input_cast_fp16)[name = tensor<string, []>("expand_dims_6_cast_fp16")];
50
+ tensor<string, []> conv_0_pad_type_0 = const()[name = tensor<string, []>("conv_0_pad_type_0"), val = tensor<string, []>("valid")];
51
+ tensor<int32, [2]> conv_0_pad_0 = const()[name = tensor<string, []>("conv_0_pad_0"), val = tensor<int32, [2]>([0, 0])];
52
+ tensor<int32, [1]> conv_0_dilations_0 = const()[name = tensor<string, []>("conv_0_dilations_0"), val = tensor<int32, [1]>([1])];
53
+ tensor<int32, []> conv_0_groups_0 = const()[name = tensor<string, []>("conv_0_groups_0"), val = tensor<int32, []>(1)];
54
+ tensor<fp16, [257, 1, 512]> expand_dims_3_to_fp16_quantized = constexpr_affine_dequantize()[axis = tensor<int32, []>(0), name = tensor<string, []>("expand_dims_3_to_fp16_quantized"), quantized_data = tensor<int8, [257, 1, 512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(64))), scale = tensor<fp16, [257]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(132096))), zero_point = tensor<int8, [257]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(131712)))];
55
+ tensor<fp16, [1, 257, 1501]> conv_0_cast_fp16 = conv(dilations = conv_0_dilations_0, groups = conv_0_groups_0, pad = conv_0_pad_0, pad_type = conv_0_pad_type_0, strides = expand_dims_5, weight = expand_dims_3_to_fp16_quantized, x = expand_dims_6_cast_fp16)[name = tensor<string, []>("conv_0_cast_fp16")];
56
+ tensor<string, []> conv_1_pad_type_0 = const()[name = tensor<string, []>("conv_1_pad_type_0"), val = tensor<string, []>("valid")];
57
+ tensor<int32, [2]> conv_1_pad_0 = const()[name = tensor<string, []>("conv_1_pad_0"), val = tensor<int32, [2]>([0, 0])];
58
+ tensor<int32, [1]> conv_1_dilations_0 = const()[name = tensor<string, []>("conv_1_dilations_0"), val = tensor<int32, [1]>([1])];
59
+ tensor<int32, []> conv_1_groups_0 = const()[name = tensor<string, []>("conv_1_groups_0"), val = tensor<int32, []>(1)];
60
+ tensor<fp16, [257, 1, 512]> expand_dims_4_to_fp16_quantized = constexpr_affine_dequantize()[axis = tensor<int32, []>(0), name = tensor<string, []>("expand_dims_4_to_fp16_quantized"), quantized_data = tensor<int8, [257, 1, 512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(132736))), scale = tensor<fp16, [257]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(264768))), zero_point = tensor<int8, [257]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(264384)))];
61
+ tensor<fp16, [1, 257, 1501]> conv_1_cast_fp16 = conv(dilations = conv_1_dilations_0, groups = conv_1_groups_0, pad = conv_1_pad_0, pad_type = conv_1_pad_type_0, strides = expand_dims_5, weight = expand_dims_4_to_fp16_quantized, x = expand_dims_6_cast_fp16)[name = tensor<string, []>("conv_1_cast_fp16")];
62
+ tensor<int32, []> stack_0_axis_0 = const()[name = tensor<string, []>("stack_0_axis_0"), val = tensor<int32, []>(-1)];
63
+ tensor<fp16, [1, 257, 1501, 2]> stack_0_cast_fp16 = stack(axis = stack_0_axis_0, values = (conv_0_cast_fp16, conv_1_cast_fp16))[name = tensor<string, []>("stack_0_cast_fp16")];
64
+ tensor<fp16, []> var_17_promoted_to_fp16 = const()[name = tensor<string, []>("op_17_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+1)];
65
+ tensor<fp16, [1, 257, 1501, 2]> var_67_cast_fp16 = pow(x = stack_0_cast_fp16, y = var_17_promoted_to_fp16)[name = tensor<string, []>("op_67_cast_fp16")];
66
+ tensor<int32, [1]> var_69_axes_0 = const()[name = tensor<string, []>("op_69_axes_0"), val = tensor<int32, [1]>([-1])];
67
+ tensor<bool, []> var_69_keep_dims_0 = const()[name = tensor<string, []>("op_69_keep_dims_0"), val = tensor<bool, []>(false)];
68
+ tensor<fp16, [1, 257, 1501]> var_69_cast_fp16 = reduce_sum(axes = var_69_axes_0, keep_dims = var_69_keep_dims_0, x = var_67_cast_fp16)[name = tensor<string, []>("op_69_cast_fp16")];
69
+ tensor<bool, []> x_11_transpose_x_0 = const()[name = tensor<string, []>("x_11_transpose_x_0"), val = tensor<bool, []>(false)];
70
+ tensor<bool, []> x_11_transpose_y_0 = const()[name = tensor<string, []>("x_11_transpose_y_0"), val = tensor<bool, []>(false)];
71
+ tensor<fp16, [1, 128, 257]> const_6_to_fp16_quantized = constexpr_affine_dequantize()[axis = tensor<int32, []>(1), name = tensor<string, []>("const_6_to_fp16_quantized"), quantized_data = tensor<int8, [1, 128, 257]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(265408))), scale = tensor<fp16, [128]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(298560))), zero_point = tensor<int8, [128]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(298368)))];
72
+ tensor<fp16, [1, 128, 1501]> x_11_cast_fp16 = matmul(transpose_x = x_11_transpose_x_0, transpose_y = x_11_transpose_y_0, x = const_6_to_fp16_quantized, y = var_69_cast_fp16)[name = tensor<string, []>("x_11_cast_fp16")];
73
+ tensor<fp16, []> var_76_to_fp16 = const()[name = tensor<string, []>("op_76_to_fp16"), val = tensor<fp16, []>(0x1p-24)];
74
+ tensor<fp16, [1, 128, 1501]> var_77_cast_fp16 = add(x = x_11_cast_fp16, y = var_76_to_fp16)[name = tensor<string, []>("op_77_cast_fp16")];
75
+ tensor<fp32, []> x_13_epsilon_0 = const()[name = tensor<string, []>("x_13_epsilon_0"), val = tensor<fp32, []>(0x1p-149)];
76
+ tensor<fp16, [1, 128, 1501]> x_13_cast_fp16 = log(epsilon = x_13_epsilon_0, x = var_77_cast_fp16)[name = tensor<string, []>("x_13_cast_fp16")];
77
+ tensor<int32, [1, 1501]> var_82 = const()[name = tensor<string, []>("op_82"), val = tensor<int32, [1, 1501]>([[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 262, 263, 264, 265, 266, 267, 268, 269, 270, 271, 272, 273, 274, 275, 276, 277, 278, 279, 280, 281, 282, 283, 284, 285, 286, 287, 288, 289, 290, 291, 292, 293, 294, 295, 296, 297, 298, 299, 300, 301, 302, 303, 304, 305, 306, 307, 308, 309, 310, 311, 312, 313, 314, 315, 316, 317, 318, 319, 320, 321, 322, 323, 324, 325, 326, 327, 328, 329, 330, 331, 332, 333, 334, 335, 336, 337, 338, 339, 340, 341, 342, 343, 344, 345, 346, 347, 348, 349, 350, 351, 352, 353, 354, 355, 356, 357, 358, 359, 360, 361, 362, 363, 364, 365, 366, 367, 368, 369, 370, 371, 372, 373, 374, 375, 376, 377, 378, 379, 380, 381, 382, 383, 384, 385, 386, 387, 388, 389, 390, 391, 392, 393, 394, 395, 396, 397, 398, 399, 400, 401, 402, 403, 404, 405, 406, 407, 408, 409, 410, 411, 412, 413, 414, 415, 416, 417, 418, 419, 420, 421, 422, 423, 424, 425, 426, 427, 428, 429, 430, 431, 432, 433, 434, 435, 436, 437, 438, 439, 440, 441, 442, 443, 444, 445, 446, 447, 448, 449, 450, 451, 452, 453, 454, 455, 456, 457, 458, 459, 460, 461, 462, 463, 464, 465, 466, 467, 468, 469, 470, 471, 472, 473, 474, 475, 476, 477, 478, 479, 480, 481, 482, 483, 484, 485, 486, 487, 488, 489, 490, 491, 492, 493, 494, 495, 496, 497, 498, 499, 500, 501, 502, 503, 504, 505, 506, 507, 508, 509, 510, 511, 512, 513, 514, 515, 516, 517, 518, 519, 520, 521, 522, 523, 524, 525, 526, 527, 528, 529, 530, 531, 532, 533, 534, 535, 536, 537, 538, 539, 540, 541, 542, 543, 544, 545, 546, 547, 548, 549, 550, 551, 552, 553, 554, 555, 556, 557, 558, 559, 560, 561, 562, 563, 564, 565, 566, 567, 568, 569, 570, 571, 572, 573, 574, 575, 576, 577, 578, 579, 580, 581, 582, 583, 584, 585, 586, 587, 588, 589, 590, 591, 592, 593, 594, 595, 596, 597, 598, 599, 600, 601, 602, 603, 604, 605, 606, 607, 608, 609, 610, 611, 612, 613, 614, 615, 616, 617, 618, 619, 620, 621, 622, 623, 624, 625, 626, 627, 628, 629, 630, 631, 632, 633, 634, 635, 636, 637, 638, 639, 640, 641, 642, 643, 644, 645, 646, 647, 648, 649, 650, 651, 652, 653, 654, 655, 656, 657, 658, 659, 660, 661, 662, 663, 664, 665, 666, 667, 668, 669, 670, 671, 672, 673, 674, 675, 676, 677, 678, 679, 680, 681, 682, 683, 684, 685, 686, 687, 688, 689, 690, 691, 692, 693, 694, 695, 696, 697, 698, 699, 700, 701, 702, 703, 704, 705, 706, 707, 708, 709, 710, 711, 712, 713, 714, 715, 716, 717, 718, 719, 720, 721, 722, 723, 724, 725, 726, 727, 728, 729, 730, 731, 732, 733, 734, 735, 736, 737, 738, 739, 740, 741, 742, 743, 744, 745, 746, 747, 748, 749, 750, 751, 752, 753, 754, 755, 756, 757, 758, 759, 760, 761, 762, 763, 764, 765, 766, 767, 768, 769, 770, 771, 772, 773, 774, 775, 776, 777, 778, 779, 780, 781, 782, 783, 784, 785, 786, 787, 788, 789, 790, 791, 792, 793, 794, 795, 796, 797, 798, 799, 800, 801, 802, 803, 804, 805, 806, 807, 808, 809, 810, 811, 812, 813, 814, 815, 816, 817, 818, 819, 820, 821, 822, 823, 824, 825, 826, 827, 828, 829, 830, 831, 832, 833, 834, 835, 836, 837, 838, 839, 840, 841, 842, 843, 844, 845, 846, 847, 848, 849, 850, 851, 852, 853, 854, 855, 856, 857, 858, 859, 860, 861, 862, 863, 864, 865, 866, 867, 868, 869, 870, 871, 872, 873, 874, 875, 876, 877, 878, 879, 880, 881, 882, 883, 884, 885, 886, 887, 888, 889, 890, 891, 892, 893, 894, 895, 896, 897, 898, 899, 900, 901, 902, 903, 904, 905, 906, 907, 908, 909, 910, 911, 912, 913, 914, 915, 916, 917, 918, 919, 920, 921, 922, 923, 924, 925, 926, 927, 928, 929, 930, 931, 932, 933, 934, 935, 936, 937, 938, 939, 940, 941, 942, 943, 944, 945, 946, 947, 948, 949, 950, 951, 952, 953, 954, 955, 956, 957, 958, 959, 960, 961, 962, 963, 964, 965, 966, 967, 968, 969, 970, 971, 972, 973, 974, 975, 976, 977, 978, 979, 980, 981, 982, 983, 984, 985, 986, 987, 988, 989, 990, 991, 992, 993, 994, 995, 996, 997, 998, 999, 1000, 1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008, 1009, 1010, 1011, 1012, 1013, 1014, 1015, 1016, 1017, 1018, 1019, 1020, 1021, 1022, 1023, 1024, 1025, 1026, 1027, 1028, 1029, 1030, 1031, 1032, 1033, 1034, 1035, 1036, 1037, 1038, 1039, 1040, 1041, 1042, 1043, 1044, 1045, 1046, 1047, 1048, 1049, 1050, 1051, 1052, 1053, 1054, 1055, 1056, 1057, 1058, 1059, 1060, 1061, 1062, 1063, 1064, 1065, 1066, 1067, 1068, 1069, 1070, 1071, 1072, 1073, 1074, 1075, 1076, 1077, 1078, 1079, 1080, 1081, 1082, 1083, 1084, 1085, 1086, 1087, 1088, 1089, 1090, 1091, 1092, 1093, 1094, 1095, 1096, 1097, 1098, 1099, 1100, 1101, 1102, 1103, 1104, 1105, 1106, 1107, 1108, 1109, 1110, 1111, 1112, 1113, 1114, 1115, 1116, 1117, 1118, 1119, 1120, 1121, 1122, 1123, 1124, 1125, 1126, 1127, 1128, 1129, 1130, 1131, 1132, 1133, 1134, 1135, 1136, 1137, 1138, 1139, 1140, 1141, 1142, 1143, 1144, 1145, 1146, 1147, 1148, 1149, 1150, 1151, 1152, 1153, 1154, 1155, 1156, 1157, 1158, 1159, 1160, 1161, 1162, 1163, 1164, 1165, 1166, 1167, 1168, 1169, 1170, 1171, 1172, 1173, 1174, 1175, 1176, 1177, 1178, 1179, 1180, 1181, 1182, 1183, 1184, 1185, 1186, 1187, 1188, 1189, 1190, 1191, 1192, 1193, 1194, 1195, 1196, 1197, 1198, 1199, 1200, 1201, 1202, 1203, 1204, 1205, 1206, 1207, 1208, 1209, 1210, 1211, 1212, 1213, 1214, 1215, 1216, 1217, 1218, 1219, 1220, 1221, 1222, 1223, 1224, 1225, 1226, 1227, 1228, 1229, 1230, 1231, 1232, 1233, 1234, 1235, 1236, 1237, 1238, 1239, 1240, 1241, 1242, 1243, 1244, 1245, 1246, 1247, 1248, 1249, 1250, 1251, 1252, 1253, 1254, 1255, 1256, 1257, 1258, 1259, 1260, 1261, 1262, 1263, 1264, 1265, 1266, 1267, 1268, 1269, 1270, 1271, 1272, 1273, 1274, 1275, 1276, 1277, 1278, 1279, 1280, 1281, 1282, 1283, 1284, 1285, 1286, 1287, 1288, 1289, 1290, 1291, 1292, 1293, 1294, 1295, 1296, 1297, 1298, 1299, 1300, 1301, 1302, 1303, 1304, 1305, 1306, 1307, 1308, 1309, 1310, 1311, 1312, 1313, 1314, 1315, 1316, 1317, 1318, 1319, 1320, 1321, 1322, 1323, 1324, 1325, 1326, 1327, 1328, 1329, 1330, 1331, 1332, 1333, 1334, 1335, 1336, 1337, 1338, 1339, 1340, 1341, 1342, 1343, 1344, 1345, 1346, 1347, 1348, 1349, 1350, 1351, 1352, 1353, 1354, 1355, 1356, 1357, 1358, 1359, 1360, 1361, 1362, 1363, 1364, 1365, 1366, 1367, 1368, 1369, 1370, 1371, 1372, 1373, 1374, 1375, 1376, 1377, 1378, 1379, 1380, 1381, 1382, 1383, 1384, 1385, 1386, 1387, 1388, 1389, 1390, 1391, 1392, 1393, 1394, 1395, 1396, 1397, 1398, 1399, 1400, 1401, 1402, 1403, 1404, 1405, 1406, 1407, 1408, 1409, 1410, 1411, 1412, 1413, 1414, 1415, 1416, 1417, 1418, 1419, 1420, 1421, 1422, 1423, 1424, 1425, 1426, 1427, 1428, 1429, 1430, 1431, 1432, 1433, 1434, 1435, 1436, 1437, 1438, 1439, 1440, 1441, 1442, 1443, 1444, 1445, 1446, 1447, 1448, 1449, 1450, 1451, 1452, 1453, 1454, 1455, 1456, 1457, 1458, 1459, 1460, 1461, 1462, 1463, 1464, 1465, 1466, 1467, 1468, 1469, 1470, 1471, 1472, 1473, 1474, 1475, 1476, 1477, 1478, 1479, 1480, 1481, 1482, 1483, 1484, 1485, 1486, 1487, 1488, 1489, 1490, 1491, 1492, 1493, 1494, 1495, 1496, 1497, 1498, 1499, 1500]])];
78
+ tensor<int32, [1]> var_85_axes_0 = const()[name = tensor<string, []>("op_85_axes_0"), val = tensor<int32, [1]>([1])];
79
+ tensor<int32, [1]> mel_length = cast(dtype = seq_len_dtype_0, x = seq_len_1_cast_fp16)[name = tensor<string, []>("cast_2")];
80
+ tensor<int32, [1, 1]> var_85 = expand_dims(axes = var_85_axes_0, x = mel_length)[name = tensor<string, []>("op_85")];
81
+ tensor<bool, [1, 1501]> valid_mask = less(x = var_82, y = var_85)[name = tensor<string, []>("valid_mask")];
82
+ tensor<int32, [1]> var_87_axes_0 = const()[name = tensor<string, []>("op_87_axes_0"), val = tensor<int32, [1]>([1])];
83
+ tensor<bool, [1, 1, 1501]> var_87 = expand_dims(axes = var_87_axes_0, x = valid_mask)[name = tensor<string, []>("op_87")];
84
+ tensor<int32, [3]> var_87_after_broadcast_reps_0 = const()[name = tensor<string, []>("op_87_after_broadcast_reps_0"), val = tensor<int32, [3]>([1, 128, 1])];
85
+ tensor<bool, [1, 128, 1501]> var_87_after_broadcast = tile(reps = var_87_after_broadcast_reps_0, x = var_87)[name = tensor<string, []>("op_87_after_broadcast")];
86
+ tensor<fp16, [1, 128, 1501]> op_24_after_broadcast_to_fp16_quantized = constexpr_affine_dequantize()[axis = tensor<int32, []>(0), name = tensor<string, []>("op_24_after_broadcast_to_fp16_quantized"), quantized_data = tensor<int8, [1, 128, 1501]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(298880))), scale = tensor<fp16, []>(0x0p+0), zero_point = tensor<int8, []>(0)];
87
+ tensor<fp16, [1, 128, 1501]> var_88_cast_fp16 = select(a = x_13_cast_fp16, b = op_24_after_broadcast_to_fp16_quantized, cond = var_87_after_broadcast)[name = tensor<string, []>("op_88_cast_fp16")];
88
+ tensor<int32, [1]> x_mean_numerator_axes_0 = const()[name = tensor<string, []>("x_mean_numerator_axes_0"), val = tensor<int32, [1]>([2])];
89
+ tensor<bool, []> x_mean_numerator_keep_dims_0 = const()[name = tensor<string, []>("x_mean_numerator_keep_dims_0"), val = tensor<bool, []>(false)];
90
+ tensor<fp16, [1, 128]> x_mean_numerator_cast_fp16 = reduce_sum(axes = x_mean_numerator_axes_0, keep_dims = x_mean_numerator_keep_dims_0, x = var_88_cast_fp16)[name = tensor<string, []>("x_mean_numerator_cast_fp16")];
91
+ tensor<int32, [1]> x_mean_denominator_axes_0 = const()[name = tensor<string, []>("x_mean_denominator_axes_0"), val = tensor<int32, [1]>([1])];
92
+ tensor<bool, []> x_mean_denominator_keep_dims_0 = const()[name = tensor<string, []>("x_mean_denominator_keep_dims_0"), val = tensor<bool, []>(false)];
93
+ tensor<string, []> cast_2_to_fp16_dtype_0 = const()[name = tensor<string, []>("cast_2_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
94
+ tensor<fp16, [1, 1501]> valid_mask_to_fp16 = cast(dtype = cast_2_to_fp16_dtype_0, x = valid_mask)[name = tensor<string, []>("cast_1")];
95
+ tensor<fp16, [1]> x_mean_denominator_cast_fp16 = reduce_sum(axes = x_mean_denominator_axes_0, keep_dims = x_mean_denominator_keep_dims_0, x = valid_mask_to_fp16)[name = tensor<string, []>("x_mean_denominator_cast_fp16")];
96
+ tensor<int32, [1]> var_93_axes_0 = const()[name = tensor<string, []>("op_93_axes_0"), val = tensor<int32, [1]>([1])];
97
+ tensor<fp16, [1, 1]> var_93_cast_fp16 = expand_dims(axes = var_93_axes_0, x = x_mean_denominator_cast_fp16)[name = tensor<string, []>("op_93_cast_fp16")];
98
+ tensor<fp16, [1, 128]> x_mean_cast_fp16 = real_div(x = x_mean_numerator_cast_fp16, y = var_93_cast_fp16)[name = tensor<string, []>("x_mean_cast_fp16")];
99
+ tensor<int32, [1]> var_96_axes_0 = const()[name = tensor<string, []>("op_96_axes_0"), val = tensor<int32, [1]>([2])];
100
+ tensor<fp16, [1, 128, 1]> var_96_cast_fp16 = expand_dims(axes = var_96_axes_0, x = x_mean_cast_fp16)[name = tensor<string, []>("op_96_cast_fp16")];
101
+ tensor<fp16, [1, 128, 1501]> var_97_cast_fp16 = sub(x = x_13_cast_fp16, y = var_96_cast_fp16)[name = tensor<string, []>("op_97_cast_fp16")];
102
+ tensor<fp16, [1, 128, 1501]> var_98_cast_fp16 = select(a = var_97_cast_fp16, b = op_24_after_broadcast_to_fp16_quantized, cond = var_87_after_broadcast)[name = tensor<string, []>("op_98_cast_fp16")];
103
+ tensor<fp16, []> var_17_promoted_1_to_fp16 = const()[name = tensor<string, []>("op_17_promoted_1_to_fp16"), val = tensor<fp16, []>(0x1p+1)];
104
+ tensor<fp16, [1, 128, 1501]> var_99_cast_fp16 = pow(x = var_98_cast_fp16, y = var_17_promoted_1_to_fp16)[name = tensor<string, []>("op_99_cast_fp16")];
105
+ tensor<int32, [1]> var_101_axes_0 = const()[name = tensor<string, []>("op_101_axes_0"), val = tensor<int32, [1]>([2])];
106
+ tensor<bool, []> var_101_keep_dims_0 = const()[name = tensor<string, []>("op_101_keep_dims_0"), val = tensor<bool, []>(false)];
107
+ tensor<fp16, [1, 128]> var_101_cast_fp16 = reduce_sum(axes = var_101_axes_0, keep_dims = var_101_keep_dims_0, x = var_99_cast_fp16)[name = tensor<string, []>("op_101_cast_fp16")];
108
+ tensor<fp16, []> var_103_to_fp16 = const()[name = tensor<string, []>("op_103_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
109
+ tensor<fp16, [1, 1]> var_104_cast_fp16 = sub(x = var_93_cast_fp16, y = var_103_to_fp16)[name = tensor<string, []>("op_104_cast_fp16")];
110
+ tensor<fp16, [1, 128]> var_105_cast_fp16 = real_div(x = var_101_cast_fp16, y = var_104_cast_fp16)[name = tensor<string, []>("op_105_cast_fp16")];
111
+ tensor<fp16, [1, 128]> x_std_1_cast_fp16 = sqrt(x = var_105_cast_fp16)[name = tensor<string, []>("x_std_1_cast_fp16")];
112
+ tensor<fp16, []> var_25_to_fp16 = const()[name = tensor<string, []>("op_25_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
113
+ tensor<fp16, [1, 128]> x_std_cast_fp16 = add(x = x_std_1_cast_fp16, y = var_25_to_fp16)[name = tensor<string, []>("x_std_cast_fp16")];
114
+ tensor<int32, [1]> var_110_axes_0 = const()[name = tensor<string, []>("op_110_axes_0"), val = tensor<int32, [1]>([2])];
115
+ tensor<fp16, [1, 128, 1]> var_110_cast_fp16 = expand_dims(axes = var_110_axes_0, x = x_std_cast_fp16)[name = tensor<string, []>("op_110_cast_fp16")];
116
+ tensor<fp16, [1, 128, 1501]> x_cast_fp16 = real_div(x = var_97_cast_fp16, y = var_110_cast_fp16)[name = tensor<string, []>("x_cast_fp16")];
117
+ tensor<bool, [1, 1501]> mask = greater_equal(x = var_82, y = var_85)[name = tensor<string, []>("mask")];
118
+ tensor<int32, [1]> var_119_axes_0 = const()[name = tensor<string, []>("op_119_axes_0"), val = tensor<int32, [1]>([1])];
119
+ tensor<bool, [1, 1, 1501]> var_119 = expand_dims(axes = var_119_axes_0, x = mask)[name = tensor<string, []>("op_119")];
120
+ tensor<fp16, []> var_24_to_fp16 = const()[name = tensor<string, []>("op_24_to_fp16"), val = tensor<fp16, []>(0x0p+0)];
121
+ tensor<fp16, [1, 128, 1501]> processed_signal_cast_fp16 = select(a = var_24_to_fp16, b = x_cast_fp16, cond = var_119)[name = tensor<string, []>("processed_signal_cast_fp16")];
122
+ tensor<string, []> processed_signal_cast_fp16_to_fp32_dtype_0 = const()[name = tensor<string, []>("processed_signal_cast_fp16_to_fp32_dtype_0"), val = tensor<string, []>("fp32")];
123
+ tensor<fp32, [1, 128, 1501]> mel = cast(dtype = processed_signal_cast_fp16_to_fp32_dtype_0, x = processed_signal_cast_fp16)[name = tensor<string, []>("cast_0")];
124
+ } -> (mel, mel_length);
125
+ }
Preprocessor.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:129b76e3aeafa8afa3ea76d995b964b145fe83700d579f6ff42c4c38fa0968ea
3
+ size 491072
README.md CHANGED
@@ -1,3 +1,26 @@
1
  ---
2
  license: cc-by-4.0
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: cc-by-4.0
3
+ language: [en]
4
+ library_name: coreml
5
+ pipeline_tag: automatic-speech-recognition
6
+ base_model: FluidVoice-Labs/fluid-speech-pro
7
+ tags: [speech-recognition, parakeet, core-ml, apple-neural-engine]
8
  ---
9
+
10
+ # Fluid Speech Pro (Core ML, 8-bit)
11
+
12
+ **Fluid Speech Pro by FluidVoice**, converted to Core ML by TalkMate for the Apple Neural Engine.
13
+
14
+ Same inputs, outputs and tokenizer as `parakeet-tdt-0.6b-v3-coreml` (Preprocessor, Encoder, Decoder,
15
+ JointDecision), so FluidAudio loads it as a Parakeet Ultra-contract package.
16
+
17
+ | Part | Weights |
18
+ |---|---|
19
+ | Encoder | int8 per channel, Pro's extra pre-Conformer block included |
20
+ | Decoder, JointDecision | fp16 |
21
+ | Preprocessor | FluidInference's Parakeet TDT 0.6B v3 preprocessor |
22
+
23
+ On accented English (EdAcc, 600 clips, M4): 10.5% WER, against 10.1% for the original .nemo. About 0.6 GB.
24
+
25
+ Licensed CC BY 4.0. Built on NVIDIA Parakeet TDT 0.6B v3 and Moondream Parakeet Ultra, both CC BY 4.0;
26
+ the preprocessor is FluidInference's (CC BY 4.0).
config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_id": "fluid-speech-pro",
3
+ "name": "Fluid Speech Pro",
4
+ "source": "https://huggingface.co/FluidVoice-Labs/fluid-speech-pro",
5
+ "license": "CC-BY-4.0",
6
+ "minimum_os": {
7
+ "macOS": "15.0"
8
+ },
9
+ "audio": {
10
+ "sample_rate": 16000,
11
+ "window_seconds": 15
12
+ },
13
+ "vocabulary": "parakeet_vocab.json",
14
+ "vocab_size": 8192,
15
+ "contract": "Same inputs, outputs and tokenizer as parakeet-tdt-0.6b-v3-coreml (Preprocessor, Encoder, Decoder, JointDecision).",
16
+ "components": {
17
+ "Encoder.mlmodelc": "int8 per-channel weights, extra pre-conformer block included",
18
+ "Decoder.mlmodelc": "fp16",
19
+ "JointDecision.mlmodelc": "fp16",
20
+ "Preprocessor.mlmodelc": "FluidInference/parakeet-tdt-0.6b-v3-coreml@7dd20fe6b1797d35f5e3307e8b1732d9a178edfe"
21
+ }
22
+ }
parakeet_vocab.json ADDED
The diff for this file is too large to render. See raw diff