Upload 74 files

ANE/styletts2_ane_alignment.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79b2d334b136d0db3149ae324b9423c3c3985d12512d8bb977d6e157bf37c7d4
+size 243

ANE/styletts2_ane_alignment.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2578251e92caabd70dd2d68c453e0ed86d05c2e7db9b5511e25a9287c9fdeee9
+size 519

ANE/styletts2_ane_alignment.mlmodelc/metadata.json

@@ -0,0 +1,104 @@
+[
+  {
+    "metadataOutputVersion" : "3.0",
+    "shortDescription" : "StyleTTS2-ANE Stage 3: Alignment (cumsum+broadcast, T_a≤2000) — fp16, ANE.",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 1 × 640 × 2000)",
+        "shortDescription" : "",
+        "shape" : "[1, 640, 2000]",
+        "name" : "en",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 1 × 512 × 2000)",
+        "shortDescription" : "",
+        "shape" : "[1, 512, 2000]",
+        "name" : "asr",
+        "type" : "MultiArray"
+      }
+    ],
+    "storagePrecision" : "Float16",
+    "modelParameters" : [
+
+    ],
+    "specificationVersion" : 8,
+    "mlProgramOperationTypeHistogram" : {
+      "Ios16.cumsum" : 1,
+      "Ios17.logicalAnd" : 1,
+      "Ios17.cast" : 4,
+      "Ios17.less" : 1,
+      "Ios17.matmul" : 2,
+      "Ios17.greaterEqual" : 1,
+      "Ios17.sub" : 1,
+      "Ios17.expandDims" : 2
+    },
+    "computePrecision" : "Mixed (Float16, 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"
+    },
+    "inputSchema" : [
+      {
+        "dataType" : "Float32",
+        "hasShapeFlexibility" : "1",
+        "isOptional" : "0",
+        "shapeFlexibility" : "1 × 2...512",
+        "shapeRange" : "[[1, 1], [2, 512]]",
+        "formattedType" : "MultiArray (Float32 1 × 64)",
+        "type" : "MultiArray",
+        "shape" : "[1, 64]",
+        "name" : "pred_dur",
+        "shortDescription" : ""
+      },
+      {
+        "dataType" : "Float32",
+        "hasShapeFlexibility" : "1",
+        "isOptional" : "0",
+        "shapeFlexibility" : "1 × 2...512 × 640",
+        "shapeRange" : "[[1, 1], [2, 512], [640, 640]]",
+        "formattedType" : "MultiArray (Float32 1 × 64 × 640)",
+        "type" : "MultiArray",
+        "shape" : "[1, 64, 640]",
+        "name" : "d",
+        "shortDescription" : ""
+      },
+      {
+        "dataType" : "Float32",
+        "hasShapeFlexibility" : "1",
+        "isOptional" : "0",
+        "shapeFlexibility" : "1 × 512 × 2...512",
+        "shapeRange" : "[[1, 1], [512, 512], [2, 512]]",
+        "formattedType" : "MultiArray (Float32 1 × 512 × 64)",
+        "type" : "MultiArray",
+        "shape" : "[1, 512, 64]",
+        "name" : "t_en",
+        "shortDescription" : ""
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.source_dialect" : "TorchScript",
+      "com.github.apple.coremltools.source" : "torch==2.11.0",
+      "com.github.apple.coremltools.version" : "8.3.0"
+    },
+    "generatedClassName" : "styletts2_ane_alignment",
+    "method" : "predict"
+  }
+]

ANE/styletts2_ane_alignment.mlmodelc/model.mil

@@ -0,0 +1,33 @@
+program(1.0)
+[buildInfo = dict<tensor<string, []>, tensor<string, []>>({{"coremlc-component-MIL", "3520.4.1"}, {"coremlc-version", "3520.5.1"}})]
+{
+    func main<ios17>(tensor<fp32, [1, ?, 640]> d, tensor<fp32, [1, ?]> pred_dur, tensor<fp32, [1, 512, ?]> t_en) [FlexibleShapeInformation = tuple<tuple<tensor<string, []>, dict<tensor<string, []>, tensor<int32, [?]>>>, tuple<tensor<string, []>, dict<tensor<string, []>, list<tensor<int32, [2]>, ?>>>>((("DefaultShapes", {{"d", [1, 64, 640]}, {"pred_dur", [1, 64]}, {"t_en", [1, 512, 64]}}), ("RangeDims", {{"d", [[1, 1], [2, 512], [640, 640]]}, {"pred_dur", [[1, 1], [2, 512]]}, {"t_en", [[1, 1], [512, 512], [2, 512]]}})))] {
+            tensor<int32, []> var_9 = const()[name = tensor<string, []>("op_9"), val = tensor<int32, []>(-1)];
+            tensor<bool, []> cum_exclusive_0 = const()[name = tensor<string, []>("cum_exclusive_0"), val = tensor<bool, []>(false)];
+            tensor<bool, []> cum_reverse_0 = const()[name = tensor<string, []>("cum_reverse_0"), val = tensor<bool, []>(false)];
+            tensor<string, []> pred_dur_to_fp16_dtype_0 = const()[name = tensor<string, []>("pred_dur_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
+            tensor<fp16, [1, ?]> pred_dur_to_fp16 = cast(dtype = pred_dur_to_fp16_dtype_0, x = pred_dur)[name = tensor<string, []>("cast_3")];
+            tensor<fp16, [1, ?]> cum_cast_fp16 = cumsum(axis = var_9, exclusive = cum_exclusive_0, reverse = cum_reverse_0, x = pred_dur_to_fp16)[name = tensor<string, []>("cum_cast_fp16")];
+            tensor<fp16, [1, ?]> starts_cast_fp16 = sub(x = cum_cast_fp16, y = pred_dur_to_fp16)[name = tensor<string, []>("starts_cast_fp16")];
+            tensor<int32, [1]> var_25_axes_0 = const()[name = tensor<string, []>("op_25_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, [1, ?, 1]> var_25_cast_fp16 = expand_dims(axes = var_25_axes_0, x = starts_cast_fp16)[name = tensor<string, []>("op_25_cast_fp16")];
+            tensor<fp16, [1, 1, 2000]> frames_promoted_to_fp16 = const()[name = tensor<string, []>("frames_promoted_to_fp16"), val = tensor<fp16, [1, 1, 2000]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(64)))];
+            tensor<bool, [1, ?, 2000]> var_26_cast_fp16 = greater_equal(x = frames_promoted_to_fp16, y = var_25_cast_fp16)[name = tensor<string, []>("op_26_cast_fp16")];
+            tensor<int32, [1]> var_28_axes_0 = const()[name = tensor<string, []>("op_28_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, [1, ?, 1]> var_28_cast_fp16 = expand_dims(axes = var_28_axes_0, x = cum_cast_fp16)[name = tensor<string, []>("op_28_cast_fp16")];
+            tensor<bool, [1, ?, 2000]> var_29_cast_fp16 = less(x = frames_promoted_to_fp16, y = var_28_cast_fp16)[name = tensor<string, []>("op_29_cast_fp16")];
+            tensor<bool, [1, ?, 2000]> in_seg = logical_and(x = var_26_cast_fp16, y = var_29_cast_fp16)[name = tensor<string, []>("in_seg")];
+            tensor<string, []> d_to_fp16_dtype_0 = const()[name = tensor<string, []>("d_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
+            tensor<bool, []> var_44_transpose_x_1 = const()[name = tensor<string, []>("op_44_transpose_x_1"), val = tensor<bool, []>(true)];
+            tensor<bool, []> var_44_transpose_y_1 = const()[name = tensor<string, []>("op_44_transpose_y_1"), val = tensor<bool, []>(false)];
+            tensor<string, []> cast_3_to_fp16_dtype_0 = const()[name = tensor<string, []>("cast_3_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
+            tensor<fp16, [1, ?, 640]> d_to_fp16 = cast(dtype = d_to_fp16_dtype_0, x = d)[name = tensor<string, []>("cast_1")];
+            tensor<fp16, [1, ?, 2000]> in_seg_to_fp16 = cast(dtype = cast_3_to_fp16_dtype_0, x = in_seg)[name = tensor<string, []>("cast_2")];
+            tensor<fp16, [1, 640, 2000]> en = matmul(transpose_x = var_44_transpose_x_1, transpose_y = var_44_transpose_y_1, x = d_to_fp16, y = in_seg_to_fp16)[name = tensor<string, []>("op_44_cast_fp16")];
+            tensor<bool, []> var_50_transpose_x_0 = const()[name = tensor<string, []>("op_50_transpose_x_0"), val = tensor<bool, []>(false)];
+            tensor<bool, []> var_50_transpose_y_0 = const()[name = tensor<string, []>("op_50_transpose_y_0"), val = tensor<bool, []>(false)];
+            tensor<string, []> t_en_to_fp16_dtype_0 = const()[name = tensor<string, []>("t_en_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
+            tensor<fp16, [1, 512, ?]> t_en_to_fp16 = cast(dtype = t_en_to_fp16_dtype_0, x = t_en)[name = tensor<string, []>("cast_0")];
+            tensor<fp16, [1, 512, 2000]> asr = matmul(transpose_x = var_50_transpose_x_0, transpose_y = var_50_transpose_y_0, x = t_en_to_fp16, y = in_seg_to_fp16)[name = tensor<string, []>("op_50_cast_fp16")];
+        } -> (en, asr);
+}

ANE/styletts2_ane_alignment.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e7d69128b59d615fc3d3cf85637a687235fc086b1eb136359adb11a61615f6b
+size 4128

ANE/styletts2_ane_alignment.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a866791eb7d8e7fe4a54551d9c98d7e8a10bd442799862fa52dd81ac9c550e6
+size 4162

ANE/styletts2_ane_alignment.mlpackage/Data/com.apple.CoreML/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e7d69128b59d615fc3d3cf85637a687235fc086b1eb136359adb11a61615f6b
+size 4128

ANE/styletts2_ane_alignment.mlpackage/Manifest.json

@@ -0,0 +1,18 @@
+{
+    "fileFormatVersion": "1.0.0",
+    "itemInfoEntries": {
+        "B5C37E4A-EDAD-42AA-A450-1BF7CC840C90": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Weights",
+            "name": "weights",
+            "path": "com.apple.CoreML/weights"
+        },
+        "DBA3FF95-5E96-49D4-A75D-BD281A186EE8": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Specification",
+            "name": "model.mlmodel",
+            "path": "com.apple.CoreML/model.mlmodel"
+        }
+    },
+    "rootModelIdentifier": "DBA3FF95-5E96-49D4-A75D-BD281A186EE8"
+}

ANE/styletts2_ane_diffusion_step.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78d70f424ec17bfb8c861db787faf81ca410d2011942e5f50a7e2a251108d2f3
+size 243

ANE/styletts2_ane_diffusion_step.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:979bc3282748bfae788bd403e1bd2c91ff6e9f348e90545b14cb80c0aab0cf57
+size 472

ANE/styletts2_ane_diffusion_step.mlmodelc/metadata.json

@@ -0,0 +1,115 @@
+[
+  {
+    "metadataOutputVersion" : "3.0",
+    "shortDescription" : "StyleTTS2-ANE Stage 4: ADPM2 single denoise step (static shapes) — fp16, ANE.",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 1 × 1 × 256)",
+        "shortDescription" : "",
+        "shape" : "[1, 1, 256]",
+        "name" : "denoised",
+        "type" : "MultiArray"
+      }
+    ],
+    "storagePrecision" : "Mixed (Float16, Palettized (8 bits))",
+    "modelParameters" : [
+
+    ],
+    "specificationVersion" : 8,
+    "mlProgramOperationTypeHistogram" : {
+      "Ios17.cos" : 1,
+      "Ios17.reshape" : 20,
+      "Ios17.inverse" : 2,
+      "Ios16.softmax" : 3,
+      "Ios17.matmul" : 6,
+      "Ios17.sin" : 1,
+      "Ios17.transpose" : 27,
+      "Split" : 9,
+      "Ios16.reduceMean" : 1,
+      "Ios17.expandDims" : 2,
+      "Ios17.add" : 23,
+      "Tile" : 2,
+      "Ios16.reduceSum" : 1,
+      "Ios17.log" : 1,
+      "Ios17.layerNorm" : 3,
+      "Ios17.sqrt" : 1,
+      "Ios16.gelu" : 7,
+      "Ios16.constexprLutToDense" : 26,
+      "Ios17.cast" : 4,
+      "Ios17.realDiv" : 1,
+      "Ios17.linear" : 25,
+      "Stack" : 1,
+      "Ios17.concat" : 2,
+      "Ios17.conv" : 1,
+      "Ios17.mul" : 19
+    },
+    "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"
+    },
+    "inputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 1 × 256)",
+        "shortDescription" : "",
+        "shape" : "[1, 1, 256]",
+        "name" : "x_noisy",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1)",
+        "shortDescription" : "",
+        "shape" : "[1]",
+        "name" : "sigma",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 512 × 768)",
+        "shortDescription" : "",
+        "shape" : "[1, 512, 768]",
+        "name" : "embedding",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 256)",
+        "shortDescription" : "",
+        "shape" : "[1, 256]",
+        "name" : "features",
+        "type" : "MultiArray"
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.source_dialect" : "TorchScript",
+      "com.github.apple.coremltools.version" : "8.3.0",
+      "com.github.apple.coremltools.source" : "torch==2.11.0"
+    },
+    "generatedClassName" : "styletts2_ane_diffusion_step",
+    "method" : "predict"
+  }
+]

ANE/styletts2_ane_diffusion_step.mlmodelc/model.mil

@@ -0,0 +1,351 @@
+program(1.0)
+[buildInfo = dict<tensor<string, []>, tensor<string, []>>({{"coremlc-component-MIL", "3520.4.1"}, {"coremlc-version", "3520.5.1"}})]
+{
+    func main<ios17>(tensor<fp32, [1, 512, 768]> embedding, tensor<fp32, [1, 256]> features, tensor<fp32, [1]> sigma, tensor<fp32, [1, 1, 256]> x_noisy) {
+            tensor<int32, [3]> var_25 = const()[name = tensor<string, []>("op_25"), val = tensor<int32, [3]>([-1, 1, 1])];
+            tensor<string, []> sigma_to_fp16_dtype_0 = const()[name = tensor<string, []>("sigma_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
+            tensor<fp16, [1]> sigma_to_fp16 = cast(dtype = sigma_to_fp16_dtype_0, x = sigma)[name = tensor<string, []>("cast_3")];
+            tensor<fp16, [1, 1, 1]> s_cast_fp16 = reshape(shape = var_25, x = sigma_to_fp16)[name = tensor<string, []>("s_cast_fp16")];
+            tensor<fp16, [1, 1, 1]> var_27_cast_fp16 = mul(x = s_cast_fp16, y = s_cast_fp16)[name = tensor<string, []>("op_27_cast_fp16")];
+            tensor<fp16, []> var_29_to_fp16 = const()[name = tensor<string, []>("op_29_to_fp16"), val = tensor<fp16, []>(0x1.47cp-5)];
+            tensor<fp16, [1, 1, 1]> var_30_cast_fp16 = add(x = var_27_cast_fp16, y = var_29_to_fp16)[name = tensor<string, []>("op_30_cast_fp16")];
+            tensor<fp32, []> var_31_epsilon_0 = const()[name = tensor<string, []>("op_31_epsilon_0"), val = tensor<fp32, []>(0x1.a36e2ep-14)];
+            tensor<fp16, [1, 1, 1]> var_31_cast_fp16 = inverse(epsilon = var_31_epsilon_0, x = var_30_cast_fp16)[name = tensor<string, []>("op_31_cast_fp16")];
+            tensor<fp16, []> var_32_to_fp16 = const()[name = tensor<string, []>("op_32_to_fp16"), val = tensor<fp16, []>(0x1.47cp-5)];
+            tensor<fp16, [1, 1, 1]> c_skip_cast_fp16 = mul(x = var_31_cast_fp16, y = var_32_to_fp16)[name = tensor<string, []>("c_skip_cast_fp16")];
+            tensor<fp16, []> var_34_to_fp16 = const()[name = tensor<string, []>("op_34_to_fp16"), val = tensor<fp16, []>(0x1.998p-3)];
+            tensor<fp16, [1, 1, 1]> var_35_cast_fp16 = mul(x = s_cast_fp16, y = var_34_to_fp16)[name = tensor<string, []>("op_35_cast_fp16")];
+            tensor<fp16, [1, 1, 1]> var_40_cast_fp16 = sqrt(x = var_30_cast_fp16)[name = tensor<string, []>("op_40_cast_fp16")];
+            tensor<fp16, [1, 1, 1]> c_out_cast_fp16 = real_div(x = var_35_cast_fp16, y = var_40_cast_fp16)[name = tensor<string, []>("c_out_cast_fp16")];
+            tensor<fp32, []> var_47_epsilon_0 = const()[name = tensor<string, []>("op_47_epsilon_0"), val = tensor<fp32, []>(0x1.a36e2ep-14)];
+            tensor<fp16, [1, 1, 1]> var_47_cast_fp16 = inverse(epsilon = var_47_epsilon_0, x = var_40_cast_fp16)[name = tensor<string, []>("op_47_cast_fp16")];
+            tensor<fp32, []> var_50_epsilon_0 = const()[name = tensor<string, []>("op_50_epsilon_0"), val = tensor<fp32, []>(0x1p-149)];
+            tensor<fp16, [1]> var_50_cast_fp16 = log(epsilon = var_50_epsilon_0, x = sigma_to_fp16)[name = tensor<string, []>("op_50_cast_fp16")];
+            tensor<fp16, []> var_51_to_fp16 = const()[name = tensor<string, []>("op_51_to_fp16"), val = tensor<fp16, []>(0x1p-2)];
+            tensor<fp16, [1]> x_1_cast_fp16 = mul(x = var_50_cast_fp16, y = var_51_to_fp16)[name = tensor<string, []>("x_1_cast_fp16")];
+            tensor<string, []> x_noisy_to_fp16_dtype_0 = const()[name = tensor<string, []>("x_noisy_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
+            tensor<fp16, [1, 1, 256]> x_noisy_to_fp16 = cast(dtype = x_noisy_to_fp16_dtype_0, x = x_noisy)[name = tensor<string, []>("cast_2")];
+            tensor<fp16, [1, 1, 256]> x_11_cast_fp16 = mul(x = var_47_cast_fp16, y = x_noisy_to_fp16)[name = tensor<string, []>("x_11_cast_fp16")];
+            tensor<int32, []> var_55 = const()[name = tensor<string, []>("op_55"), val = tensor<int32, []>(-1)];
+            tensor<int32, [2]> var_67 = const()[name = tensor<string, []>("op_67"), val = tensor<int32, [2]>([1, 1])];
+            tensor<fp16, [1, 1]> x_5_cast_fp16 = reshape(shape = var_67, x = x_1_cast_fp16)[name = tensor<string, []>("x_5_cast_fp16")];
+            tensor<fp16, [1, 128]> var_75_to_fp16 = const()[name = tensor<string, []>("op_75_to_fp16"), val = tensor<fp16, [1, 128]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(64)))];
+            tensor<fp16, [1, 128]> var_76_cast_fp16 = mul(x = x_5_cast_fp16, y = var_75_to_fp16)[name = tensor<string, []>("op_76_cast_fp16")];
+            tensor<fp16, []> var_77_promoted_to_fp16 = const()[name = tensor<string, []>("op_77_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+1)];
+            tensor<fp16, [1, 128]> var_78_cast_fp16 = mul(x = var_76_cast_fp16, y = var_77_promoted_to_fp16)[name = tensor<string, []>("op_78_cast_fp16")];
+            tensor<fp16, []> var_79_to_fp16 = const()[name = tensor<string, []>("op_79_to_fp16"), val = tensor<fp16, []>(0x1.92p+1)];
+            tensor<fp16, [1, 128]> freqs_cast_fp16 = mul(x = var_78_cast_fp16, y = var_79_to_fp16)[name = tensor<string, []>("freqs_cast_fp16")];
+            tensor<fp16, [1, 128]> var_81_cast_fp16 = sin(x = freqs_cast_fp16)[name = tensor<string, []>("op_81_cast_fp16")];
+            tensor<fp16, [1, 128]> var_82_cast_fp16 = cos(x = freqs_cast_fp16)[name = tensor<string, []>("op_82_cast_fp16")];
+            tensor<bool, []> input_1_interleave_0 = const()[name = tensor<string, []>("input_1_interleave_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, 257]> input_1_cast_fp16 = concat(axis = var_55, interleave = input_1_interleave_0, values = (x_5_cast_fp16, var_81_cast_fp16, var_82_cast_fp16))[name = tensor<string, []>("input_1_cast_fp16")];
+            tensor<fp16, [1024, 257]> transformer_to_time_0_1_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [263168]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(384))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(263616))), name = tensor<string, []>("transformer_to_time_0_1_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 257])];
+            tensor<fp16, [1024]> transformer_to_time_0_1_bias_to_fp16 = const()[name = tensor<string, []>("transformer_to_time_0_1_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(264192)))];
+            tensor<fp16, [1, 1024]> linear_0_cast_fp16 = linear(bias = transformer_to_time_0_1_bias_to_fp16, weight = transformer_to_time_0_1_weight_to_fp16_palettized, x = input_1_cast_fp16)[name = tensor<string, []>("linear_0_cast_fp16")];
+            tensor<string, []> var_88_mode_0 = const()[name = tensor<string, []>("op_88_mode_0"), val = tensor<string, []>("EXACT")];
+            tensor<fp16, [1, 1024]> var_88_cast_fp16 = gelu(mode = var_88_mode_0, x = linear_0_cast_fp16)[name = tensor<string, []>("op_88_cast_fp16")];
+            tensor<string, []> features_to_fp16_dtype_0 = const()[name = tensor<string, []>("features_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
+            tensor<fp16, [1024, 256]> transformer_to_features_0_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [262144]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(266304))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(528512))), name = tensor<string, []>("transformer_to_features_0_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 256])];
+            tensor<fp16, [1024]> transformer_to_features_0_bias_to_fp16 = const()[name = tensor<string, []>("transformer_to_features_0_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(529088)))];
+            tensor<fp16, [1, 256]> features_to_fp16 = cast(dtype = features_to_fp16_dtype_0, x = features)[name = tensor<string, []>("cast_1")];
+            tensor<fp16, [1, 1024]> linear_1_cast_fp16 = linear(bias = transformer_to_features_0_bias_to_fp16, weight = transformer_to_features_0_weight_to_fp16_palettized, x = features_to_fp16)[name = tensor<string, []>("linear_1_cast_fp16")];
+            tensor<string, []> var_94_mode_0 = const()[name = tensor<string, []>("op_94_mode_0"), val = tensor<string, []>("EXACT")];
+            tensor<fp16, [1, 1024]> var_94_cast_fp16 = gelu(mode = var_94_mode_0, x = linear_1_cast_fp16)[name = tensor<string, []>("op_94_cast_fp16")];
+            tensor<int32, []> x_7_axis_0 = const()[name = tensor<string, []>("x_7_axis_0"), val = tensor<int32, []>(0)];
+            tensor<fp16, [2, 1, 1024]> x_7_cast_fp16 = stack(axis = x_7_axis_0, values = (var_88_cast_fp16, var_94_cast_fp16))[name = tensor<string, []>("x_7_cast_fp16")];
+            tensor<int32, [3]> var_101 = const()[name = tensor<string, []>("op_101"), val = tensor<int32, [3]>([1, 2, 0])];
+            tensor<int32, [1]> input_7_axes_0 = const()[name = tensor<string, []>("input_7_axes_0"), val = tensor<int32, [1]>([2])];
+            tensor<bool, []> input_7_keep_dims_0 = const()[name = tensor<string, []>("input_7_keep_dims_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, 1024, 2]> x_9_cast_fp16 = transpose(perm = var_101, x = x_7_cast_fp16)[name = tensor<string, []>("transpose_44")];
+            tensor<fp16, [1, 1024]> input_7_cast_fp16 = reduce_sum(axes = input_7_axes_0, keep_dims = input_7_keep_dims_0, x = x_9_cast_fp16)[name = tensor<string, []>("input_7_cast_fp16")];
+            tensor<fp16, [1024, 1024]> transformer_to_mapping_0_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [1048576]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(531200))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1579840))), name = tensor<string, []>("transformer_to_mapping_0_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 1024])];
+            tensor<fp16, [1024]> transformer_to_mapping_0_bias_to_fp16 = const()[name = tensor<string, []>("transformer_to_mapping_0_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1580416)))];
+            tensor<fp16, [1, 1024]> linear_2_cast_fp16 = linear(bias = transformer_to_mapping_0_bias_to_fp16, weight = transformer_to_mapping_0_weight_to_fp16_palettized, x = input_7_cast_fp16)[name = tensor<string, []>("linear_2_cast_fp16")];
+            tensor<string, []> input_11_mode_0 = const()[name = tensor<string, []>("input_11_mode_0"), val = tensor<string, []>("EXACT")];
+            tensor<fp16, [1, 1024]> input_11_cast_fp16 = gelu(mode = input_11_mode_0, x = linear_2_cast_fp16)[name = tensor<string, []>("input_11_cast_fp16")];
+            tensor<fp16, [1024, 1024]> transformer_to_mapping_2_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [1048576]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1582528))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2631168))), name = tensor<string, []>("transformer_to_mapping_2_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 1024])];
+            tensor<fp16, [1024]> transformer_to_mapping_2_bias_to_fp16 = const()[name = tensor<string, []>("transformer_to_mapping_2_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2631744)))];
+            tensor<fp16, [1, 1024]> linear_3_cast_fp16 = linear(bias = transformer_to_mapping_2_bias_to_fp16, weight = transformer_to_mapping_2_weight_to_fp16_palettized, x = input_11_cast_fp16)[name = tensor<string, []>("linear_3_cast_fp16")];
+            tensor<string, []> mapping_1_mode_0 = const()[name = tensor<string, []>("mapping_1_mode_0"), val = tensor<string, []>("EXACT")];
+            tensor<fp16, [1, 1024]> mapping_1_cast_fp16 = gelu(mode = mapping_1_mode_0, x = linear_3_cast_fp16)[name = tensor<string, []>("mapping_1_cast_fp16")];
+            tensor<int32, [3]> var_127_reps_0 = const()[name = tensor<string, []>("op_127_reps_0"), val = tensor<int32, [3]>([1, 512, 1])];
+            tensor<fp16, [1, 512, 256]> var_127_cast_fp16 = tile(reps = var_127_reps_0, x = x_11_cast_fp16)[name = tensor<string, []>("op_127_cast_fp16")];
+            tensor<int32, []> var_129 = const()[name = tensor<string, []>("op_129"), val = tensor<int32, []>(-1)];
+            tensor<bool, []> x_13_interleave_0 = const()[name = tensor<string, []>("x_13_interleave_0"), val = tensor<bool, []>(false)];
+            tensor<string, []> embedding_to_fp16_dtype_0 = const()[name = tensor<string, []>("embedding_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
+            tensor<fp16, [1, 512, 768]> embedding_to_fp16 = cast(dtype = embedding_to_fp16_dtype_0, x = embedding)[name = tensor<string, []>("cast_0")];
+            tensor<fp16, [1, 512, 1024]> x_13_cast_fp16 = concat(axis = var_129, interleave = x_13_interleave_0, values = (var_127_cast_fp16, embedding_to_fp16))[name = tensor<string, []>("x_13_cast_fp16")];
+            tensor<int32, [1]> var_132_axes_0 = const()[name = tensor<string, []>("op_132_axes_0"), val = tensor<int32, [1]>([1])];
+            tensor<fp16, [1, 1, 1024]> var_132_cast_fp16 = expand_dims(axes = var_132_axes_0, x = mapping_1_cast_fp16)[name = tensor<string, []>("op_132_cast_fp16")];
+            tensor<int32, [3]> mapping_reps_0 = const()[name = tensor<string, []>("mapping_reps_0"), val = tensor<int32, [3]>([1, 512, 1])];
+            tensor<fp16, [1, 512, 1024]> mapping_cast_fp16 = tile(reps = mapping_reps_0, x = var_132_cast_fp16)[name = tensor<string, []>("mapping_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> x_15_cast_fp16 = add(x = x_13_cast_fp16, y = mapping_cast_fp16)[name = tensor<string, []>("x_15_cast_fp16")];
+            tensor<int32, []> var_153 = const()[name = tensor<string, []>("op_153"), val = tensor<int32, []>(-1)];
+            tensor<fp16, [2048, 256]> transformer_blocks_0_attention_norm_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2633856))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3158208))), name = tensor<string, []>("transformer_blocks_0_attention_norm_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([2048, 256])];
+            tensor<fp16, [2048]> transformer_blocks_0_attention_norm_fc_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_0_attention_norm_fc_bias_to_fp16"), val = tensor<fp16, [2048]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3158784)))];
+            tensor<fp16, [1, 2048]> linear_4_cast_fp16 = linear(bias = transformer_blocks_0_attention_norm_fc_bias_to_fp16, weight = transformer_blocks_0_attention_norm_fc_weight_to_fp16_palettized, x = features_to_fp16)[name = tensor<string, []>("linear_4_cast_fp16")];
+            tensor<int32, [3]> var_172 = const()[name = tensor<string, []>("op_172"), val = tensor<int32, [3]>([1, 2048, 1])];
+            tensor<fp16, [1, 2048, 1]> h_3_cast_fp16 = reshape(shape = var_172, x = linear_4_cast_fp16)[name = tensor<string, []>("h_3_cast_fp16")];
+            tensor<int32, [2]> var_174_split_sizes_0 = const()[name = tensor<string, []>("op_174_split_sizes_0"), val = tensor<int32, [2]>([1024, 1024])];
+            tensor<int32, []> var_174_axis_0 = const()[name = tensor<string, []>("op_174_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 1024, 1]> var_174_cast_fp16_0, tensor<fp16, [1, 1024, 1]> var_174_cast_fp16_1 = split(axis = var_174_axis_0, split_sizes = var_174_split_sizes_0, x = h_3_cast_fp16)[name = tensor<string, []>("op_174_cast_fp16")];
+            tensor<int32, [3]> gamma_3_perm_0 = const()[name = tensor<string, []>("gamma_3_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [3]> beta_3_perm_0 = const()[name = tensor<string, []>("beta_3_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [1]> x_19_axes_0 = const()[name = tensor<string, []>("x_19_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, []> var_146_to_fp16 = const()[name = tensor<string, []>("op_146_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<fp16, [1, 512, 1024]> x_19_cast_fp16 = layer_norm(axes = x_19_axes_0, epsilon = var_146_to_fp16, x = x_15_cast_fp16)[name = tensor<string, []>("x_19_cast_fp16")];
+            tensor<fp16, []> var_180_promoted_to_fp16 = const()[name = tensor<string, []>("op_180_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 1, 1024]> gamma_3_cast_fp16 = transpose(perm = gamma_3_perm_0, x = var_174_cast_fp16_0)[name = tensor<string, []>("transpose_43")];
+            tensor<fp16, [1, 1, 1024]> var_181_cast_fp16 = add(x = gamma_3_cast_fp16, y = var_180_promoted_to_fp16)[name = tensor<string, []>("op_181_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> var_182_cast_fp16 = mul(x = var_181_cast_fp16, y = x_19_cast_fp16)[name = tensor<string, []>("op_182_cast_fp16")];
+            tensor<fp16, [1, 1, 1024]> beta_3_cast_fp16 = transpose(perm = beta_3_perm_0, x = var_174_cast_fp16_1)[name = tensor<string, []>("transpose_42")];
+            tensor<fp16, [1, 512, 1024]> x_21_cast_fp16 = add(x = var_182_cast_fp16, y = beta_3_cast_fp16)[name = tensor<string, []>("x_21_cast_fp16")];
+            tensor<fp16, [2048, 256]> transformer_blocks_0_attention_norm_context_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3162944))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3687296))), name = tensor<string, []>("transformer_blocks_0_attention_norm_context_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([2048, 256])];
+            tensor<fp16, [2048]> transformer_blocks_0_attention_norm_context_fc_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_0_attention_norm_context_fc_bias_to_fp16"), val = tensor<fp16, [2048]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3687872)))];
+            tensor<fp16, [1, 2048]> linear_5_cast_fp16 = linear(bias = transformer_blocks_0_attention_norm_context_fc_bias_to_fp16, weight = transformer_blocks_0_attention_norm_context_fc_weight_to_fp16_palettized, x = features_to_fp16)[name = tensor<string, []>("linear_5_cast_fp16")];
+            tensor<int32, [3]> var_194 = const()[name = tensor<string, []>("op_194"), val = tensor<int32, [3]>([1, 2048, 1])];
+            tensor<fp16, [1, 2048, 1]> h_7_cast_fp16 = reshape(shape = var_194, x = linear_5_cast_fp16)[name = tensor<string, []>("h_7_cast_fp16")];
+            tensor<int32, [2]> var_196_split_sizes_0 = const()[name = tensor<string, []>("op_196_split_sizes_0"), val = tensor<int32, [2]>([1024, 1024])];
+            tensor<int32, []> var_196_axis_0 = const()[name = tensor<string, []>("op_196_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 1024, 1]> var_196_cast_fp16_0, tensor<fp16, [1, 1024, 1]> var_196_cast_fp16_1 = split(axis = var_196_axis_0, split_sizes = var_196_split_sizes_0, x = h_7_cast_fp16)[name = tensor<string, []>("op_196_cast_fp16")];
+            tensor<int32, [3]> gamma_7_perm_0 = const()[name = tensor<string, []>("gamma_7_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [3]> beta_7_perm_0 = const()[name = tensor<string, []>("beta_7_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<fp16, []> var_202_promoted_to_fp16 = const()[name = tensor<string, []>("op_202_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 1, 1024]> gamma_7_cast_fp16 = transpose(perm = gamma_7_perm_0, x = var_196_cast_fp16_0)[name = tensor<string, []>("transpose_41")];
+            tensor<fp16, [1, 1, 1024]> var_203_cast_fp16 = add(x = gamma_7_cast_fp16, y = var_202_promoted_to_fp16)[name = tensor<string, []>("op_203_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> var_204_cast_fp16 = mul(x = var_203_cast_fp16, y = x_19_cast_fp16)[name = tensor<string, []>("op_204_cast_fp16")];
+            tensor<fp16, [1, 1, 1024]> beta_7_cast_fp16 = transpose(perm = beta_7_perm_0, x = var_196_cast_fp16_1)[name = tensor<string, []>("transpose_40")];
+            tensor<fp16, [1, 512, 1024]> x_27_cast_fp16 = add(x = var_204_cast_fp16, y = beta_7_cast_fp16)[name = tensor<string, []>("x_27_cast_fp16")];
+            tensor<fp16, [512, 1024]> transformer_blocks_0_attention_to_q_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3692032))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4216384))), name = tensor<string, []>("transformer_blocks_0_attention_to_q_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([512, 1024])];
+            tensor<fp16, [512]> linear_6_bias_0_to_fp16 = const()[name = tensor<string, []>("linear_6_bias_0_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4216960)))];
+            tensor<fp16, [1, 512, 512]> linear_6_cast_fp16 = linear(bias = linear_6_bias_0_to_fp16, weight = transformer_blocks_0_attention_to_q_weight_to_fp16_palettized, x = x_21_cast_fp16)[name = tensor<string, []>("linear_6_cast_fp16")];
+            tensor<fp16, [1024, 1024]> transformer_blocks_0_attention_to_kv_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [1048576]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4218048))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5266688))), name = tensor<string, []>("transformer_blocks_0_attention_to_kv_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 1024])];
+            tensor<fp16, [1024]> linear_7_bias_0_to_fp16 = const()[name = tensor<string, []>("linear_7_bias_0_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5267264)))];
+            tensor<fp16, [1, 512, 1024]> linear_7_cast_fp16 = linear(bias = linear_7_bias_0_to_fp16, weight = transformer_blocks_0_attention_to_kv_weight_to_fp16_palettized, x = x_27_cast_fp16)[name = tensor<string, []>("linear_7_cast_fp16")];
+            tensor<int32, [2]> var_212_split_sizes_0 = const()[name = tensor<string, []>("op_212_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_212_axis_0 = const()[name = tensor<string, []>("op_212_axis_0"), val = tensor<int32, []>(-1)];
+            tensor<fp16, [1, 512, 512]> var_212_cast_fp16_0, tensor<fp16, [1, 512, 512]> var_212_cast_fp16_1 = split(axis = var_212_axis_0, split_sizes = var_212_split_sizes_0, x = linear_7_cast_fp16)[name = tensor<string, []>("op_212_cast_fp16")];
+            tensor<int32, [4]> var_221 = const()[name = tensor<string, []>("op_221"), val = tensor<int32, [4]>([1, 512, 8, 64])];
+            tensor<fp16, [1, 512, 8, 64]> x_31_cast_fp16 = reshape(shape = var_221, x = linear_6_cast_fp16)[name = tensor<string, []>("x_31_cast_fp16")];
+            tensor<int32, [4]> var_231 = const()[name = tensor<string, []>("op_231"), val = tensor<int32, [4]>([1, 512, 8, 64])];
+            tensor<fp16, [1, 512, 8, 64]> x_35_cast_fp16 = reshape(shape = var_231, x = var_212_cast_fp16_0)[name = tensor<string, []>("x_35_cast_fp16")];
+            tensor<int32, [4]> var_241 = const()[name = tensor<string, []>("op_241"), val = tensor<int32, [4]>([1, 512, 8, 64])];
+            tensor<fp16, [1, 512, 8, 64]> x_39_cast_fp16 = reshape(shape = var_241, x = var_212_cast_fp16_1)[name = tensor<string, []>("x_39_cast_fp16")];
+            tensor<int32, [4]> var_243 = const()[name = tensor<string, []>("op_243"), val = tensor<int32, [4]>([0, 2, -3, -1])];
+            tensor<bool, []> sim_1_transpose_x_0 = const()[name = tensor<string, []>("sim_1_transpose_x_0"), val = tensor<bool, []>(false)];
+            tensor<bool, []> sim_1_transpose_y_0 = const()[name = tensor<string, []>("sim_1_transpose_y_0"), val = tensor<bool, []>(false)];
+            tensor<int32, [4]> transpose_12_perm_0 = const()[name = tensor<string, []>("transpose_12_perm_0"), val = tensor<int32, [4]>([0, 2, -3, -1])];
+            tensor<int32, [4]> transpose_13_perm_0 = const()[name = tensor<string, []>("transpose_13_perm_0"), val = tensor<int32, [4]>([0, 2, -1, -3])];
+            tensor<fp16, [1, 8, 64, 512]> transpose_13 = transpose(perm = transpose_13_perm_0, x = x_35_cast_fp16)[name = tensor<string, []>("transpose_38")];
+            tensor<fp16, [1, 8, 512, 64]> transpose_12 = transpose(perm = transpose_12_perm_0, x = x_31_cast_fp16)[name = tensor<string, []>("transpose_39")];
+            tensor<fp16, [1, 8, 512, 512]> sim_1_cast_fp16 = matmul(transpose_x = sim_1_transpose_x_0, transpose_y = sim_1_transpose_y_0, x = transpose_12, y = transpose_13)[name = tensor<string, []>("sim_1_cast_fp16")];
+            tensor<fp16, []> var_247_to_fp16 = const()[name = tensor<string, []>("op_247_to_fp16"), val = tensor<fp16, []>(0x1p-3)];
+            tensor<fp16, [1, 8, 512, 512]> sim_3_cast_fp16 = mul(x = sim_1_cast_fp16, y = var_247_to_fp16)[name = tensor<string, []>("sim_3_cast_fp16")];
+            tensor<fp16, [1, 8, 512, 512]> attn_1_cast_fp16 = softmax(axis = var_153, x = sim_3_cast_fp16)[name = tensor<string, []>("attn_1_cast_fp16")];
+            tensor<bool, []> x_41_transpose_x_0 = const()[name = tensor<string, []>("x_41_transpose_x_0"), val = tensor<bool, []>(false)];
+            tensor<bool, []> x_41_transpose_y_0 = const()[name = tensor<string, []>("x_41_transpose_y_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, 8, 512, 64]> v_1_cast_fp16 = transpose(perm = var_243, x = x_39_cast_fp16)[name = tensor<string, []>("transpose_37")];
+            tensor<fp16, [1, 8, 512, 64]> x_41_cast_fp16 = matmul(transpose_x = x_41_transpose_x_0, transpose_y = x_41_transpose_y_0, x = attn_1_cast_fp16, y = v_1_cast_fp16)[name = tensor<string, []>("x_41_cast_fp16")];
+            tensor<int32, [4]> var_269 = const()[name = tensor<string, []>("op_269"), val = tensor<int32, [4]>([0, 2, 1, 3])];
+            tensor<int32, [3]> var_271 = const()[name = tensor<string, []>("op_271"), val = tensor<int32, [3]>([1, 512, 512])];
+            tensor<fp16, [1, 512, 8, 64]> x_43_cast_fp16 = transpose(perm = var_269, x = x_41_cast_fp16)[name = tensor<string, []>("transpose_36")];
+            tensor<fp16, [1, 512, 512]> input_23_cast_fp16 = reshape(shape = var_271, x = x_43_cast_fp16)[name = tensor<string, []>("input_23_cast_fp16")];
+            tensor<fp16, [1024, 512]> transformer_blocks_0_attention_attention_to_out_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5269376))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5793728))), name = tensor<string, []>("transformer_blocks_0_attention_attention_to_out_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 512])];
+            tensor<fp16, [1024]> transformer_blocks_0_attention_attention_to_out_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_0_attention_attention_to_out_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5794304)))];
+            tensor<fp16, [1, 512, 1024]> linear_8_cast_fp16 = linear(bias = transformer_blocks_0_attention_attention_to_out_bias_to_fp16, weight = transformer_blocks_0_attention_attention_to_out_weight_to_fp16_palettized, x = input_23_cast_fp16)[name = tensor<string, []>("linear_8_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> input_25_cast_fp16 = add(x = linear_8_cast_fp16, y = x_15_cast_fp16)[name = tensor<string, []>("input_25_cast_fp16")];
+            tensor<fp16, [2048, 1024]> transformer_blocks_0_feed_forward_0_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [2097152]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5796416))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7893632))), name = tensor<string, []>("transformer_blocks_0_feed_forward_0_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([2048, 1024])];
+            tensor<fp16, [2048]> transformer_blocks_0_feed_forward_0_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_0_feed_forward_0_bias_to_fp16"), val = tensor<fp16, [2048]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7894208)))];
+            tensor<fp16, [1, 512, 2048]> linear_9_cast_fp16 = linear(bias = transformer_blocks_0_feed_forward_0_bias_to_fp16, weight = transformer_blocks_0_feed_forward_0_weight_to_fp16_palettized, x = input_25_cast_fp16)[name = tensor<string, []>("linear_9_cast_fp16")];
+            tensor<string, []> input_29_mode_0 = const()[name = tensor<string, []>("input_29_mode_0"), val = tensor<string, []>("EXACT")];
+            tensor<fp16, [1, 512, 2048]> input_29_cast_fp16 = gelu(mode = input_29_mode_0, x = linear_9_cast_fp16)[name = tensor<string, []>("input_29_cast_fp16")];
+            tensor<fp16, [1024, 2048]> transformer_blocks_0_feed_forward_2_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [2097152]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7898368))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(9995584))), name = tensor<string, []>("transformer_blocks_0_feed_forward_2_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 2048])];
+            tensor<fp16, [1024]> transformer_blocks_0_feed_forward_2_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_0_feed_forward_2_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(9996160)))];
+            tensor<fp16, [1, 512, 1024]> linear_10_cast_fp16 = linear(bias = transformer_blocks_0_feed_forward_2_bias_to_fp16, weight = transformer_blocks_0_feed_forward_2_weight_to_fp16_palettized, x = input_29_cast_fp16)[name = tensor<string, []>("linear_10_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> x_45_cast_fp16 = add(x = linear_10_cast_fp16, y = input_25_cast_fp16)[name = tensor<string, []>("x_45_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> x_47_cast_fp16 = add(x = x_45_cast_fp16, y = mapping_cast_fp16)[name = tensor<string, []>("x_47_cast_fp16")];
+            tensor<int32, []> var_298 = const()[name = tensor<string, []>("op_298"), val = tensor<int32, []>(-1)];
+            tensor<fp16, [2048, 256]> transformer_blocks_1_attention_norm_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(9998272))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(10522624))), name = tensor<string, []>("transformer_blocks_1_attention_norm_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([2048, 256])];
+            tensor<fp16, [2048]> transformer_blocks_1_attention_norm_fc_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_1_attention_norm_fc_bias_to_fp16"), val = tensor<fp16, [2048]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(10523200)))];
+            tensor<fp16, [1, 2048]> linear_11_cast_fp16 = linear(bias = transformer_blocks_1_attention_norm_fc_bias_to_fp16, weight = transformer_blocks_1_attention_norm_fc_weight_to_fp16_palettized, x = features_to_fp16)[name = tensor<string, []>("linear_11_cast_fp16")];
+            tensor<int32, [3]> var_317 = const()[name = tensor<string, []>("op_317"), val = tensor<int32, [3]>([1, 2048, 1])];
+            tensor<fp16, [1, 2048, 1]> h_11_cast_fp16 = reshape(shape = var_317, x = linear_11_cast_fp16)[name = tensor<string, []>("h_11_cast_fp16")];
+            tensor<int32, [2]> var_319_split_sizes_0 = const()[name = tensor<string, []>("op_319_split_sizes_0"), val = tensor<int32, [2]>([1024, 1024])];
+            tensor<int32, []> var_319_axis_0 = const()[name = tensor<string, []>("op_319_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 1024, 1]> var_319_cast_fp16_0, tensor<fp16, [1, 1024, 1]> var_319_cast_fp16_1 = split(axis = var_319_axis_0, split_sizes = var_319_split_sizes_0, x = h_11_cast_fp16)[name = tensor<string, []>("op_319_cast_fp16")];
+            tensor<int32, [3]> gamma_11_perm_0 = const()[name = tensor<string, []>("gamma_11_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [3]> beta_11_perm_0 = const()[name = tensor<string, []>("beta_11_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [1]> x_51_axes_0 = const()[name = tensor<string, []>("x_51_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, []> var_291_to_fp16 = const()[name = tensor<string, []>("op_291_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<fp16, [1, 512, 1024]> x_51_cast_fp16 = layer_norm(axes = x_51_axes_0, epsilon = var_291_to_fp16, x = x_47_cast_fp16)[name = tensor<string, []>("x_51_cast_fp16")];
+            tensor<fp16, []> var_325_promoted_to_fp16 = const()[name = tensor<string, []>("op_325_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 1, 1024]> gamma_11_cast_fp16 = transpose(perm = gamma_11_perm_0, x = var_319_cast_fp16_0)[name = tensor<string, []>("transpose_35")];
+            tensor<fp16, [1, 1, 1024]> var_326_cast_fp16 = add(x = gamma_11_cast_fp16, y = var_325_promoted_to_fp16)[name = tensor<string, []>("op_326_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> var_327_cast_fp16 = mul(x = var_326_cast_fp16, y = x_51_cast_fp16)[name = tensor<string, []>("op_327_cast_fp16")];
+            tensor<fp16, [1, 1, 1024]> beta_11_cast_fp16 = transpose(perm = beta_11_perm_0, x = var_319_cast_fp16_1)[name = tensor<string, []>("transpose_34")];
+            tensor<fp16, [1, 512, 1024]> x_53_cast_fp16 = add(x = var_327_cast_fp16, y = beta_11_cast_fp16)[name = tensor<string, []>("x_53_cast_fp16")];
+            tensor<fp16, [2048, 256]> transformer_blocks_1_attention_norm_context_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(10527360))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(11051712))), name = tensor<string, []>("transformer_blocks_1_attention_norm_context_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([2048, 256])];
+            tensor<fp16, [2048]> transformer_blocks_1_attention_norm_context_fc_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_1_attention_norm_context_fc_bias_to_fp16"), val = tensor<fp16, [2048]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(11052288)))];
+            tensor<fp16, [1, 2048]> linear_12_cast_fp16 = linear(bias = transformer_blocks_1_attention_norm_context_fc_bias_to_fp16, weight = transformer_blocks_1_attention_norm_context_fc_weight_to_fp16_palettized, x = features_to_fp16)[name = tensor<string, []>("linear_12_cast_fp16")];
+            tensor<int32, [3]> var_339 = const()[name = tensor<string, []>("op_339"), val = tensor<int32, [3]>([1, 2048, 1])];
+            tensor<fp16, [1, 2048, 1]> h_15_cast_fp16 = reshape(shape = var_339, x = linear_12_cast_fp16)[name = tensor<string, []>("h_15_cast_fp16")];
+            tensor<int32, [2]> var_341_split_sizes_0 = const()[name = tensor<string, []>("op_341_split_sizes_0"), val = tensor<int32, [2]>([1024, 1024])];
+            tensor<int32, []> var_341_axis_0 = const()[name = tensor<string, []>("op_341_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 1024, 1]> var_341_cast_fp16_0, tensor<fp16, [1, 1024, 1]> var_341_cast_fp16_1 = split(axis = var_341_axis_0, split_sizes = var_341_split_sizes_0, x = h_15_cast_fp16)[name = tensor<string, []>("op_341_cast_fp16")];
+            tensor<int32, [3]> gamma_15_perm_0 = const()[name = tensor<string, []>("gamma_15_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [3]> beta_15_perm_0 = const()[name = tensor<string, []>("beta_15_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<fp16, []> var_347_promoted_to_fp16 = const()[name = tensor<string, []>("op_347_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 1, 1024]> gamma_15_cast_fp16 = transpose(perm = gamma_15_perm_0, x = var_341_cast_fp16_0)[name = tensor<string, []>("transpose_33")];
+            tensor<fp16, [1, 1, 1024]> var_348_cast_fp16 = add(x = gamma_15_cast_fp16, y = var_347_promoted_to_fp16)[name = tensor<string, []>("op_348_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> var_349_cast_fp16 = mul(x = var_348_cast_fp16, y = x_51_cast_fp16)[name = tensor<string, []>("op_349_cast_fp16")];
+            tensor<fp16, [1, 1, 1024]> beta_15_cast_fp16 = transpose(perm = beta_15_perm_0, x = var_341_cast_fp16_1)[name = tensor<string, []>("transpose_32")];
+            tensor<fp16, [1, 512, 1024]> x_59_cast_fp16 = add(x = var_349_cast_fp16, y = beta_15_cast_fp16)[name = tensor<string, []>("x_59_cast_fp16")];
+            tensor<fp16, [512, 1024]> transformer_blocks_1_attention_to_q_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(11056448))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(11580800))), name = tensor<string, []>("transformer_blocks_1_attention_to_q_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([512, 1024])];
+            tensor<fp16, [1, 512, 512]> linear_13_cast_fp16 = linear(bias = linear_6_bias_0_to_fp16, weight = transformer_blocks_1_attention_to_q_weight_to_fp16_palettized, x = x_53_cast_fp16)[name = tensor<string, []>("linear_13_cast_fp16")];
+            tensor<fp16, [1024, 1024]> transformer_blocks_1_attention_to_kv_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [1048576]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(11581376))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(12630016))), name = tensor<string, []>("transformer_blocks_1_attention_to_kv_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 1024])];
+            tensor<fp16, [1, 512, 1024]> linear_14_cast_fp16 = linear(bias = linear_7_bias_0_to_fp16, weight = transformer_blocks_1_attention_to_kv_weight_to_fp16_palettized, x = x_59_cast_fp16)[name = tensor<string, []>("linear_14_cast_fp16")];
+            tensor<int32, [2]> var_357_split_sizes_0 = const()[name = tensor<string, []>("op_357_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_357_axis_0 = const()[name = tensor<string, []>("op_357_axis_0"), val = tensor<int32, []>(-1)];
+            tensor<fp16, [1, 512, 512]> var_357_cast_fp16_0, tensor<fp16, [1, 512, 512]> var_357_cast_fp16_1 = split(axis = var_357_axis_0, split_sizes = var_357_split_sizes_0, x = linear_14_cast_fp16)[name = tensor<string, []>("op_357_cast_fp16")];
+            tensor<int32, [4]> var_366 = const()[name = tensor<string, []>("op_366"), val = tensor<int32, [4]>([1, 512, 8, 64])];
+            tensor<fp16, [1, 512, 8, 64]> x_63_cast_fp16 = reshape(shape = var_366, x = linear_13_cast_fp16)[name = tensor<string, []>("x_63_cast_fp16")];
+            tensor<int32, [4]> var_376 = const()[name = tensor<string, []>("op_376"), val = tensor<int32, [4]>([1, 512, 8, 64])];
+            tensor<fp16, [1, 512, 8, 64]> x_67_cast_fp16 = reshape(shape = var_376, x = var_357_cast_fp16_0)[name = tensor<string, []>("x_67_cast_fp16")];
+            tensor<int32, [4]> var_386 = const()[name = tensor<string, []>("op_386"), val = tensor<int32, [4]>([1, 512, 8, 64])];
+            tensor<fp16, [1, 512, 8, 64]> x_71_cast_fp16 = reshape(shape = var_386, x = var_357_cast_fp16_1)[name = tensor<string, []>("x_71_cast_fp16")];
+            tensor<int32, [4]> var_388 = const()[name = tensor<string, []>("op_388"), val = tensor<int32, [4]>([0, 2, -3, -1])];
+            tensor<bool, []> sim_5_transpose_x_0 = const()[name = tensor<string, []>("sim_5_transpose_x_0"), val = tensor<bool, []>(false)];
+            tensor<bool, []> sim_5_transpose_y_0 = const()[name = tensor<string, []>("sim_5_transpose_y_0"), val = tensor<bool, []>(false)];
+            tensor<int32, [4]> transpose_14_perm_0 = const()[name = tensor<string, []>("transpose_14_perm_0"), val = tensor<int32, [4]>([0, 2, -3, -1])];
+            tensor<int32, [4]> transpose_15_perm_0 = const()[name = tensor<string, []>("transpose_15_perm_0"), val = tensor<int32, [4]>([0, 2, -1, -3])];
+            tensor<fp16, [1, 8, 64, 512]> transpose_15 = transpose(perm = transpose_15_perm_0, x = x_67_cast_fp16)[name = tensor<string, []>("transpose_30")];
+            tensor<fp16, [1, 8, 512, 64]> transpose_14 = transpose(perm = transpose_14_perm_0, x = x_63_cast_fp16)[name = tensor<string, []>("transpose_31")];
+            tensor<fp16, [1, 8, 512, 512]> sim_5_cast_fp16 = matmul(transpose_x = sim_5_transpose_x_0, transpose_y = sim_5_transpose_y_0, x = transpose_14, y = transpose_15)[name = tensor<string, []>("sim_5_cast_fp16")];
+            tensor<fp16, []> var_392_to_fp16 = const()[name = tensor<string, []>("op_392_to_fp16"), val = tensor<fp16, []>(0x1p-3)];
+            tensor<fp16, [1, 8, 512, 512]> sim_7_cast_fp16 = mul(x = sim_5_cast_fp16, y = var_392_to_fp16)[name = tensor<string, []>("sim_7_cast_fp16")];
+            tensor<fp16, [1, 8, 512, 512]> attn_3_cast_fp16 = softmax(axis = var_298, x = sim_7_cast_fp16)[name = tensor<string, []>("attn_3_cast_fp16")];
+            tensor<bool, []> x_73_transpose_x_0 = const()[name = tensor<string, []>("x_73_transpose_x_0"), val = tensor<bool, []>(false)];
+            tensor<bool, []> x_73_transpose_y_0 = const()[name = tensor<string, []>("x_73_transpose_y_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, 8, 512, 64]> v_3_cast_fp16 = transpose(perm = var_388, x = x_71_cast_fp16)[name = tensor<string, []>("transpose_29")];
+            tensor<fp16, [1, 8, 512, 64]> x_73_cast_fp16 = matmul(transpose_x = x_73_transpose_x_0, transpose_y = x_73_transpose_y_0, x = attn_3_cast_fp16, y = v_3_cast_fp16)[name = tensor<string, []>("x_73_cast_fp16")];
+            tensor<int32, [4]> var_414 = const()[name = tensor<string, []>("op_414"), val = tensor<int32, [4]>([0, 2, 1, 3])];
+            tensor<int32, [3]> var_416 = const()[name = tensor<string, []>("op_416"), val = tensor<int32, [3]>([1, 512, 512])];
+            tensor<fp16, [1, 512, 8, 64]> x_75_cast_fp16 = transpose(perm = var_414, x = x_73_cast_fp16)[name = tensor<string, []>("transpose_28")];
+            tensor<fp16, [1, 512, 512]> input_39_cast_fp16 = reshape(shape = var_416, x = x_75_cast_fp16)[name = tensor<string, []>("input_39_cast_fp16")];
+            tensor<fp16, [1024, 512]> transformer_blocks_1_attention_attention_to_out_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(12630592))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(13154944))), name = tensor<string, []>("transformer_blocks_1_attention_attention_to_out_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 512])];
+            tensor<fp16, [1024]> transformer_blocks_1_attention_attention_to_out_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_1_attention_attention_to_out_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(13155520)))];
+            tensor<fp16, [1, 512, 1024]> linear_15_cast_fp16 = linear(bias = transformer_blocks_1_attention_attention_to_out_bias_to_fp16, weight = transformer_blocks_1_attention_attention_to_out_weight_to_fp16_palettized, x = input_39_cast_fp16)[name = tensor<string, []>("linear_15_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> input_41_cast_fp16 = add(x = linear_15_cast_fp16, y = x_47_cast_fp16)[name = tensor<string, []>("input_41_cast_fp16")];
+            tensor<fp16, [2048, 1024]> transformer_blocks_1_feed_forward_0_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [2097152]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(13157632))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(15254848))), name = tensor<string, []>("transformer_blocks_1_feed_forward_0_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([2048, 1024])];
+            tensor<fp16, [2048]> transformer_blocks_1_feed_forward_0_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_1_feed_forward_0_bias_to_fp16"), val = tensor<fp16, [2048]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(15255424)))];
+            tensor<fp16, [1, 512, 2048]> linear_16_cast_fp16 = linear(bias = transformer_blocks_1_feed_forward_0_bias_to_fp16, weight = transformer_blocks_1_feed_forward_0_weight_to_fp16_palettized, x = input_41_cast_fp16)[name = tensor<string, []>("linear_16_cast_fp16")];
+            tensor<string, []> input_45_mode_0 = const()[name = tensor<string, []>("input_45_mode_0"), val = tensor<string, []>("EXACT")];
+            tensor<fp16, [1, 512, 2048]> input_45_cast_fp16 = gelu(mode = input_45_mode_0, x = linear_16_cast_fp16)[name = tensor<string, []>("input_45_cast_fp16")];
+            tensor<fp16, [1024, 2048]> transformer_blocks_1_feed_forward_2_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [2097152]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(15259584))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(17356800))), name = tensor<string, []>("transformer_blocks_1_feed_forward_2_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 2048])];
+            tensor<fp16, [1024]> transformer_blocks_1_feed_forward_2_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_1_feed_forward_2_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(17357376)))];
+            tensor<fp16, [1, 512, 1024]> linear_17_cast_fp16 = linear(bias = transformer_blocks_1_feed_forward_2_bias_to_fp16, weight = transformer_blocks_1_feed_forward_2_weight_to_fp16_palettized, x = input_45_cast_fp16)[name = tensor<string, []>("linear_17_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> x_77_cast_fp16 = add(x = linear_17_cast_fp16, y = input_41_cast_fp16)[name = tensor<string, []>("x_77_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> x_79_cast_fp16 = add(x = x_77_cast_fp16, y = mapping_cast_fp16)[name = tensor<string, []>("x_79_cast_fp16")];
+            tensor<int32, []> var_443 = const()[name = tensor<string, []>("op_443"), val = tensor<int32, []>(-1)];
+            tensor<fp16, [2048, 256]> transformer_blocks_2_attention_norm_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(17359488))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(17883840))), name = tensor<string, []>("transformer_blocks_2_attention_norm_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([2048, 256])];
+            tensor<fp16, [2048]> transformer_blocks_2_attention_norm_fc_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_2_attention_norm_fc_bias_to_fp16"), val = tensor<fp16, [2048]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(17884416)))];
+            tensor<fp16, [1, 2048]> linear_18_cast_fp16 = linear(bias = transformer_blocks_2_attention_norm_fc_bias_to_fp16, weight = transformer_blocks_2_attention_norm_fc_weight_to_fp16_palettized, x = features_to_fp16)[name = tensor<string, []>("linear_18_cast_fp16")];
+            tensor<int32, [3]> var_462 = const()[name = tensor<string, []>("op_462"), val = tensor<int32, [3]>([1, 2048, 1])];
+            tensor<fp16, [1, 2048, 1]> h_19_cast_fp16 = reshape(shape = var_462, x = linear_18_cast_fp16)[name = tensor<string, []>("h_19_cast_fp16")];
+            tensor<int32, [2]> var_464_split_sizes_0 = const()[name = tensor<string, []>("op_464_split_sizes_0"), val = tensor<int32, [2]>([1024, 1024])];
+            tensor<int32, []> var_464_axis_0 = const()[name = tensor<string, []>("op_464_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 1024, 1]> var_464_cast_fp16_0, tensor<fp16, [1, 1024, 1]> var_464_cast_fp16_1 = split(axis = var_464_axis_0, split_sizes = var_464_split_sizes_0, x = h_19_cast_fp16)[name = tensor<string, []>("op_464_cast_fp16")];
+            tensor<int32, [3]> gamma_19_perm_0 = const()[name = tensor<string, []>("gamma_19_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [3]> beta_19_perm_0 = const()[name = tensor<string, []>("beta_19_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [1]> x_83_axes_0 = const()[name = tensor<string, []>("x_83_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, []> var_436_to_fp16 = const()[name = tensor<string, []>("op_436_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<fp16, [1, 512, 1024]> x_83_cast_fp16 = layer_norm(axes = x_83_axes_0, epsilon = var_436_to_fp16, x = x_79_cast_fp16)[name = tensor<string, []>("x_83_cast_fp16")];
+            tensor<fp16, []> var_470_promoted_to_fp16 = const()[name = tensor<string, []>("op_470_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 1, 1024]> gamma_19_cast_fp16 = transpose(perm = gamma_19_perm_0, x = var_464_cast_fp16_0)[name = tensor<string, []>("transpose_27")];
+            tensor<fp16, [1, 1, 1024]> var_471_cast_fp16 = add(x = gamma_19_cast_fp16, y = var_470_promoted_to_fp16)[name = tensor<string, []>("op_471_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> var_472_cast_fp16 = mul(x = var_471_cast_fp16, y = x_83_cast_fp16)[name = tensor<string, []>("op_472_cast_fp16")];
+            tensor<fp16, [1, 1, 1024]> beta_19_cast_fp16 = transpose(perm = beta_19_perm_0, x = var_464_cast_fp16_1)[name = tensor<string, []>("transpose_26")];
+            tensor<fp16, [1, 512, 1024]> x_85_cast_fp16 = add(x = var_472_cast_fp16, y = beta_19_cast_fp16)[name = tensor<string, []>("x_85_cast_fp16")];
+            tensor<fp16, [2048, 256]> transformer_blocks_2_attention_norm_context_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(17888576))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(18412928))), name = tensor<string, []>("transformer_blocks_2_attention_norm_context_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([2048, 256])];
+            tensor<fp16, [2048]> transformer_blocks_2_attention_norm_context_fc_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_2_attention_norm_context_fc_bias_to_fp16"), val = tensor<fp16, [2048]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(18413504)))];
+            tensor<fp16, [1, 2048]> linear_19_cast_fp16 = linear(bias = transformer_blocks_2_attention_norm_context_fc_bias_to_fp16, weight = transformer_blocks_2_attention_norm_context_fc_weight_to_fp16_palettized, x = features_to_fp16)[name = tensor<string, []>("linear_19_cast_fp16")];
+            tensor<int32, [3]> var_484 = const()[name = tensor<string, []>("op_484"), val = tensor<int32, [3]>([1, 2048, 1])];
+            tensor<fp16, [1, 2048, 1]> h_cast_fp16 = reshape(shape = var_484, x = linear_19_cast_fp16)[name = tensor<string, []>("h_cast_fp16")];
+            tensor<int32, [2]> var_486_split_sizes_0 = const()[name = tensor<string, []>("op_486_split_sizes_0"), val = tensor<int32, [2]>([1024, 1024])];
+            tensor<int32, []> var_486_axis_0 = const()[name = tensor<string, []>("op_486_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 1024, 1]> var_486_cast_fp16_0, tensor<fp16, [1, 1024, 1]> var_486_cast_fp16_1 = split(axis = var_486_axis_0, split_sizes = var_486_split_sizes_0, x = h_cast_fp16)[name = tensor<string, []>("op_486_cast_fp16")];
+            tensor<int32, [3]> gamma_perm_0 = const()[name = tensor<string, []>("gamma_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [3]> beta_perm_0 = const()[name = tensor<string, []>("beta_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<fp16, []> var_492_promoted_to_fp16 = const()[name = tensor<string, []>("op_492_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 1, 1024]> gamma_cast_fp16 = transpose(perm = gamma_perm_0, x = var_486_cast_fp16_0)[name = tensor<string, []>("transpose_25")];
+            tensor<fp16, [1, 1, 1024]> var_493_cast_fp16 = add(x = gamma_cast_fp16, y = var_492_promoted_to_fp16)[name = tensor<string, []>("op_493_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> var_494_cast_fp16 = mul(x = var_493_cast_fp16, y = x_83_cast_fp16)[name = tensor<string, []>("op_494_cast_fp16")];
+            tensor<fp16, [1, 1, 1024]> beta_cast_fp16 = transpose(perm = beta_perm_0, x = var_486_cast_fp16_1)[name = tensor<string, []>("transpose_24")];
+            tensor<fp16, [1, 512, 1024]> x_91_cast_fp16 = add(x = var_494_cast_fp16, y = beta_cast_fp16)[name = tensor<string, []>("x_91_cast_fp16")];
+            tensor<fp16, [512, 1024]> transformer_blocks_2_attention_to_q_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(18417664))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(18942016))), name = tensor<string, []>("transformer_blocks_2_attention_to_q_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([512, 1024])];
+            tensor<fp16, [1, 512, 512]> linear_20_cast_fp16 = linear(bias = linear_6_bias_0_to_fp16, weight = transformer_blocks_2_attention_to_q_weight_to_fp16_palettized, x = x_85_cast_fp16)[name = tensor<string, []>("linear_20_cast_fp16")];
+            tensor<fp16, [1024, 1024]> transformer_blocks_2_attention_to_kv_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [1048576]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(18942592))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(19991232))), name = tensor<string, []>("transformer_blocks_2_attention_to_kv_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 1024])];
+            tensor<fp16, [1, 512, 1024]> linear_21_cast_fp16 = linear(bias = linear_7_bias_0_to_fp16, weight = transformer_blocks_2_attention_to_kv_weight_to_fp16_palettized, x = x_91_cast_fp16)[name = tensor<string, []>("linear_21_cast_fp16")];
+            tensor<int32, [2]> var_502_split_sizes_0 = const()[name = tensor<string, []>("op_502_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_502_axis_0 = const()[name = tensor<string, []>("op_502_axis_0"), val = tensor<int32, []>(-1)];
+            tensor<fp16, [1, 512, 512]> var_502_cast_fp16_0, tensor<fp16, [1, 512, 512]> var_502_cast_fp16_1 = split(axis = var_502_axis_0, split_sizes = var_502_split_sizes_0, x = linear_21_cast_fp16)[name = tensor<string, []>("op_502_cast_fp16")];
+            tensor<int32, [4]> var_511 = const()[name = tensor<string, []>("op_511"), val = tensor<int32, [4]>([1, 512, 8, 64])];
+            tensor<fp16, [1, 512, 8, 64]> x_95_cast_fp16 = reshape(shape = var_511, x = linear_20_cast_fp16)[name = tensor<string, []>("x_95_cast_fp16")];
+            tensor<int32, [4]> var_521 = const()[name = tensor<string, []>("op_521"), val = tensor<int32, [4]>([1, 512, 8, 64])];
+            tensor<fp16, [1, 512, 8, 64]> x_99_cast_fp16 = reshape(shape = var_521, x = var_502_cast_fp16_0)[name = tensor<string, []>("x_99_cast_fp16")];
+            tensor<int32, [4]> var_531 = const()[name = tensor<string, []>("op_531"), val = tensor<int32, [4]>([1, 512, 8, 64])];
+            tensor<fp16, [1, 512, 8, 64]> x_103_cast_fp16 = reshape(shape = var_531, x = var_502_cast_fp16_1)[name = tensor<string, []>("x_103_cast_fp16")];
+            tensor<int32, [4]> var_533 = const()[name = tensor<string, []>("op_533"), val = tensor<int32, [4]>([0, 2, -3, -1])];
+            tensor<bool, []> sim_9_transpose_x_0 = const()[name = tensor<string, []>("sim_9_transpose_x_0"), val = tensor<bool, []>(false)];
+            tensor<bool, []> sim_9_transpose_y_0 = const()[name = tensor<string, []>("sim_9_transpose_y_0"), val = tensor<bool, []>(false)];
+            tensor<int32, [4]> transpose_16_perm_0 = const()[name = tensor<string, []>("transpose_16_perm_0"), val = tensor<int32, [4]>([0, 2, -3, -1])];
+            tensor<int32, [4]> transpose_17_perm_0 = const()[name = tensor<string, []>("transpose_17_perm_0"), val = tensor<int32, [4]>([0, 2, -1, -3])];
+            tensor<fp16, [1, 8, 64, 512]> transpose_17 = transpose(perm = transpose_17_perm_0, x = x_99_cast_fp16)[name = tensor<string, []>("transpose_22")];
+            tensor<fp16, [1, 8, 512, 64]> transpose_16 = transpose(perm = transpose_16_perm_0, x = x_95_cast_fp16)[name = tensor<string, []>("transpose_23")];
+            tensor<fp16, [1, 8, 512, 512]> sim_9_cast_fp16 = matmul(transpose_x = sim_9_transpose_x_0, transpose_y = sim_9_transpose_y_0, x = transpose_16, y = transpose_17)[name = tensor<string, []>("sim_9_cast_fp16")];
+            tensor<fp16, []> var_537_to_fp16 = const()[name = tensor<string, []>("op_537_to_fp16"), val = tensor<fp16, []>(0x1p-3)];
+            tensor<fp16, [1, 8, 512, 512]> sim_cast_fp16 = mul(x = sim_9_cast_fp16, y = var_537_to_fp16)[name = tensor<string, []>("sim_cast_fp16")];
+            tensor<fp16, [1, 8, 512, 512]> attn_cast_fp16 = softmax(axis = var_443, x = sim_cast_fp16)[name = tensor<string, []>("attn_cast_fp16")];
+            tensor<bool, []> x_105_transpose_x_0 = const()[name = tensor<string, []>("x_105_transpose_x_0"), val = tensor<bool, []>(false)];
+            tensor<bool, []> x_105_transpose_y_0 = const()[name = tensor<string, []>("x_105_transpose_y_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, 8, 512, 64]> v_cast_fp16 = transpose(perm = var_533, x = x_103_cast_fp16)[name = tensor<string, []>("transpose_21")];
+            tensor<fp16, [1, 8, 512, 64]> x_105_cast_fp16 = matmul(transpose_x = x_105_transpose_x_0, transpose_y = x_105_transpose_y_0, x = attn_cast_fp16, y = v_cast_fp16)[name = tensor<string, []>("x_105_cast_fp16")];
+            tensor<int32, [4]> var_559 = const()[name = tensor<string, []>("op_559"), val = tensor<int32, [4]>([0, 2, 1, 3])];
+            tensor<int32, [3]> var_561 = const()[name = tensor<string, []>("op_561"), val = tensor<int32, [3]>([1, 512, 512])];
+            tensor<fp16, [1, 512, 8, 64]> x_107_cast_fp16 = transpose(perm = var_559, x = x_105_cast_fp16)[name = tensor<string, []>("transpose_20")];
+            tensor<fp16, [1, 512, 512]> input_55_cast_fp16 = reshape(shape = var_561, x = x_107_cast_fp16)[name = tensor<string, []>("input_55_cast_fp16")];
+            tensor<fp16, [1024, 512]> transformer_blocks_2_attention_attention_to_out_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [524288]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(19991808))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(20516160))), name = tensor<string, []>("transformer_blocks_2_attention_attention_to_out_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 512])];
+            tensor<fp16, [1024]> transformer_blocks_2_attention_attention_to_out_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_2_attention_attention_to_out_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(20516736)))];
+            tensor<fp16, [1, 512, 1024]> linear_22_cast_fp16 = linear(bias = transformer_blocks_2_attention_attention_to_out_bias_to_fp16, weight = transformer_blocks_2_attention_attention_to_out_weight_to_fp16_palettized, x = input_55_cast_fp16)[name = tensor<string, []>("linear_22_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> input_57_cast_fp16 = add(x = linear_22_cast_fp16, y = x_79_cast_fp16)[name = tensor<string, []>("input_57_cast_fp16")];
+            tensor<fp16, [2048, 1024]> transformer_blocks_2_feed_forward_0_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [2097152]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(20518848))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(22616064))), name = tensor<string, []>("transformer_blocks_2_feed_forward_0_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([2048, 1024])];
+            tensor<fp16, [2048]> transformer_blocks_2_feed_forward_0_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_2_feed_forward_0_bias_to_fp16"), val = tensor<fp16, [2048]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(22616640)))];
+            tensor<fp16, [1, 512, 2048]> linear_23_cast_fp16 = linear(bias = transformer_blocks_2_feed_forward_0_bias_to_fp16, weight = transformer_blocks_2_feed_forward_0_weight_to_fp16_palettized, x = input_57_cast_fp16)[name = tensor<string, []>("linear_23_cast_fp16")];
+            tensor<string, []> input_61_mode_0 = const()[name = tensor<string, []>("input_61_mode_0"), val = tensor<string, []>("EXACT")];
+            tensor<fp16, [1, 512, 2048]> input_61_cast_fp16 = gelu(mode = input_61_mode_0, x = linear_23_cast_fp16)[name = tensor<string, []>("input_61_cast_fp16")];
+            tensor<fp16, [1024, 2048]> transformer_blocks_2_feed_forward_2_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [2097152]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(22620800))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(24718016))), name = tensor<string, []>("transformer_blocks_2_feed_forward_2_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 2048])];
+            tensor<fp16, [1024]> transformer_blocks_2_feed_forward_2_bias_to_fp16 = const()[name = tensor<string, []>("transformer_blocks_2_feed_forward_2_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(24718592)))];
+            tensor<fp16, [1, 512, 1024]> linear_24_cast_fp16 = linear(bias = transformer_blocks_2_feed_forward_2_bias_to_fp16, weight = transformer_blocks_2_feed_forward_2_weight_to_fp16_palettized, x = input_61_cast_fp16)[name = tensor<string, []>("linear_24_cast_fp16")];
+            tensor<fp16, [1, 512, 1024]> x_109_cast_fp16 = add(x = linear_24_cast_fp16, y = input_57_cast_fp16)[name = tensor<string, []>("x_109_cast_fp16")];
+            tensor<int32, [1]> var_581_axes_0 = const()[name = tensor<string, []>("op_581_axes_0"), val = tensor<int32, [1]>([1])];
+            tensor<bool, []> var_581_keep_dims_0 = const()[name = tensor<string, []>("op_581_keep_dims_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, 1024]> var_581_cast_fp16 = reduce_mean(axes = var_581_axes_0, keep_dims = var_581_keep_dims_0, x = x_109_cast_fp16)[name = tensor<string, []>("op_581_cast_fp16")];
+            tensor<int32, [1]> x_111_axes_0 = const()[name = tensor<string, []>("x_111_axes_0"), val = tensor<int32, [1]>([1])];
+            tensor<fp16, [1, 1, 1024]> x_111_cast_fp16 = expand_dims(axes = x_111_axes_0, x = var_581_cast_fp16)[name = tensor<string, []>("x_111_cast_fp16")];
+            tensor<int32, [3]> var_590 = const()[name = tensor<string, []>("op_590"), val = tensor<int32, [3]>([0, 2, 1])];
+            tensor<string, []> x_pad_type_0 = const()[name = tensor<string, []>("x_pad_type_0"), val = tensor<string, []>("valid")];
+            tensor<int32, [1]> x_strides_0 = const()[name = tensor<string, []>("x_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [2]> x_pad_0 = const()[name = tensor<string, []>("x_pad_0"), val = tensor<int32, [2]>([0, 0])];
+            tensor<int32, [1]> x_dilations_0 = const()[name = tensor<string, []>("x_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> x_groups_0 = const()[name = tensor<string, []>("x_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 1024, 1]> transformer_to_out_1_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [262144]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(24720704))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(24982912))), name = tensor<string, []>("transformer_to_out_1_weight_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 1024, 1])];
+            tensor<fp16, [256]> transformer_to_out_1_bias_to_fp16 = const()[name = tensor<string, []>("transformer_to_out_1_bias_to_fp16"), val = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(24983488)))];
+            tensor<fp16, [1, 1024, 1]> input_cast_fp16 = transpose(perm = var_590, x = x_111_cast_fp16)[name = tensor<string, []>("transpose_19")];
+            tensor<fp16, [1, 256, 1]> x_cast_fp16 = conv(bias = transformer_to_out_1_bias_to_fp16, dilations = x_dilations_0, groups = x_groups_0, pad = x_pad_0, pad_type = x_pad_type_0, strides = x_strides_0, weight = transformer_to_out_1_weight_to_fp16_palettized, x = input_cast_fp16)[name = tensor<string, []>("x_cast_fp16")];
+            tensor<int32, [3]> x_pred_perm_0 = const()[name = tensor<string, []>("x_pred_perm_0"), val = tensor<int32, [3]>([0, -1, -2])];
+            tensor<fp16, [1, 1, 256]> var_602_cast_fp16 = mul(x = c_skip_cast_fp16, y = x_noisy_to_fp16)[name = tensor<string, []>("op_602_cast_fp16")];
+            tensor<fp16, [1, 1, 256]> x_pred_cast_fp16 = transpose(perm = x_pred_perm_0, x = x_cast_fp16)[name = tensor<string, []>("transpose_18")];
+            tensor<fp16, [1, 1, 256]> var_603_cast_fp16 = mul(x = c_out_cast_fp16, y = x_pred_cast_fp16)[name = tensor<string, []>("op_603_cast_fp16")];
+            tensor<fp16, [1, 1, 256]> denoised = add(x = var_602_cast_fp16, y = var_603_cast_fp16)[name = tensor<string, []>("op_605_cast_fp16")];
+        } -> (denoised);
+}

ANE/styletts2_ane_diffusion_step.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7ec58d48cbfd5094ca3c0a9de778a789878e9d2cae35cc4ad277cff669d538d
+size 24984064

ANE/styletts2_ane_diffusion_step.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7012e3db65ab67a3ad9a85a421d83b9ceb2363b370c7fc3f27d23271af6ca9e
+size 58407

ANE/styletts2_ane_diffusion_step.mlpackage/Data/com.apple.CoreML/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7ec58d48cbfd5094ca3c0a9de778a789878e9d2cae35cc4ad277cff669d538d
+size 24984064

ANE/styletts2_ane_diffusion_step.mlpackage/Manifest.json

@@ -0,0 +1,18 @@
+{
+    "fileFormatVersion": "1.0.0",
+    "itemInfoEntries": {
+        "0D2868A2-D1E7-4D21-BDBB-2078B73600CD": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Specification",
+            "name": "model.mlmodel",
+            "path": "com.apple.CoreML/model.mlmodel"
+        },
+        "99F44601-8B02-4B68-8756-183CA2AB7B21": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Weights",
+            "name": "weights",
+            "path": "com.apple.CoreML/weights"
+        }
+    },
+    "rootModelIdentifier": "0D2868A2-D1E7-4D21-BDBB-2078B73600CD"
+}

ANE/styletts2_ane_noise.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f4c8e0ce15ecb4ac16b403361160287c7638ac0dfdd8327835be795979b6395
+size 243

ANE/styletts2_ane_noise.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:848949d5bf9dd6db5ef5cd49faa59a9ecfe47efcf442562eb568471c9787ff16
+size 407

ANE/styletts2_ane_noise.mlmodelc/metadata.json

@@ -0,0 +1,87 @@
+[
+  {
+    "metadataOutputVersion" : "3.0",
+    "shortDescription" : "StyleTTS2-ANE Stage 6: SineGen (static T_a=2000) — fp32, ALL.",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 1200000 × 9)",
+        "shortDescription" : "",
+        "shape" : "[1, 1200000, 9]",
+        "name" : "sine_waves",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 1200000 × 1)",
+        "shortDescription" : "",
+        "shape" : "[1, 1200000, 1]",
+        "name" : "uv",
+        "type" : "MultiArray"
+      }
+    ],
+    "storagePrecision" : "Palettized (8 bits)",
+    "modelParameters" : [
+
+    ],
+    "specificationVersion" : 8,
+    "mlProgramOperationTypeHistogram" : {
+      "Ios17.mul" : 14,
+      "Ios17.floor" : 1,
+      "Ios17.transpose" : 7,
+      "Ios17.sub" : 2,
+      "Ios16.constexprLutToDense" : 2,
+      "Ios17.sin" : 2,
+      "Ios16.cumsum" : 1,
+      "Ios17.greater" : 1,
+      "Ios17.concat" : 1,
+      "Ios17.sliceByIndex" : 3,
+      "UpsampleNearestNeighbor" : 1,
+      "Ios17.add" : 3,
+      "Tile" : 2,
+      "Ios17.expandDims" : 2,
+      "Ios17.squeeze" : 1,
+      "Ios17.reshape" : 4,
+      "Ios17.cast" : 1
+    },
+    "computePrecision" : "Mixed (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"
+    },
+    "inputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 4000)",
+        "shortDescription" : "",
+        "shape" : "[1, 4000]",
+        "name" : "F0_curve",
+        "type" : "MultiArray"
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.version" : "8.3.0",
+      "com.github.apple.coremltools.source_dialect" : "TorchScript",
+      "com.github.apple.coremltools.source" : "torch==2.11.0"
+    },
+    "generatedClassName" : "styletts2_ane_noise",
+    "method" : "predict"
+  }
+]

ANE/styletts2_ane_noise.mlmodelc/model.mil

@@ -0,0 +1,99 @@
+program(1.0)
+[buildInfo = dict<tensor<string, []>, tensor<string, []>>({{"coremlc-component-MIL", "3520.4.1"}, {"coremlc-version", "3520.5.1"}})]
+{
+    func main<ios17>(tensor<fp32, [1, 4000]> F0_curve) {
+            tensor<int32, [1]> input_axes_0 = const()[name = tensor<string, []>("input_axes_0"), val = tensor<int32, [1]>([1])];
+            tensor<fp32, [1, 1, 4000]> input = expand_dims(axes = input_axes_0, x = F0_curve)[name = tensor<string, []>("input")];
+            tensor<int32, [1]> expand_dims_0_axes_0 = const()[name = tensor<string, []>("expand_dims_0_axes_0"), val = tensor<int32, [1]>([3])];
+            tensor<fp32, [1, 1, 4000, 1]> expand_dims_0 = expand_dims(axes = expand_dims_0_axes_0, x = input)[name = tensor<string, []>("expand_dims_0")];
+            tensor<int32, []> upsample_nearest_neighbor_0_scale_factor_height_0 = const()[name = tensor<string, []>("upsample_nearest_neighbor_0_scale_factor_height_0"), val = tensor<int32, []>(300)];
+            tensor<int32, []> upsample_nearest_neighbor_0_scale_factor_width_0 = const()[name = tensor<string, []>("upsample_nearest_neighbor_0_scale_factor_width_0"), val = tensor<int32, []>(1)];
+            tensor<fp32, [1, 1, 1200000, 1]> upsample_nearest_neighbor_0 = upsample_nearest_neighbor(scale_factor_height = upsample_nearest_neighbor_0_scale_factor_height_0, scale_factor_width = upsample_nearest_neighbor_0_scale_factor_width_0, x = expand_dims_0)[name = tensor<string, []>("upsample_nearest_neighbor_0")];
+            tensor<int32, [1]> var_12_axes_0 = const()[name = tensor<string, []>("op_12_axes_0"), val = tensor<int32, [1]>([3])];
+            tensor<fp32, [1, 1, 1200000]> var_12 = squeeze(axes = var_12_axes_0, x = upsample_nearest_neighbor_0)[name = tensor<string, []>("op_12")];
+            tensor<int32, [3]> f0_perm_0 = const()[name = tensor<string, []>("f0_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
+            tensor<int32, []> var_23 = const()[name = tensor<string, []>("op_23"), val = tensor<int32, []>(1)];
+            tensor<fp32, [1, 1, 9]> const_1 = const()[name = tensor<string, []>("const_1"), val = tensor<fp32, [1, 1, 9]>([[[0x1p+0, 0x1p+1, 0x1.8p+1, 0x1p+2, 0x1.4p+2, 0x1.8p+2, 0x1.cp+2, 0x1p+3, 0x1.2p+3]]])];
+            tensor<fp32, [1, 1200000, 1]> f0 = transpose(perm = f0_perm_0, x = var_12)[name = tensor<string, []>("transpose_6")];
+            tensor<fp32, [1, 1200000, 9]> f0_values = mul(x = f0, y = const_1)[name = tensor<string, []>("f0_values")];
+            tensor<fp32, []> _inversed_rad_div_y_0 = const()[name = tensor<string, []>("_inversed_rad_div_y_0"), val = tensor<fp32, []>(0x1.5d867cp-15)];
+            tensor<fp32, [1, 1200000, 9]> _inversed_rad_div = mul(x = f0_values, y = _inversed_rad_div_y_0)[name = tensor<string, []>("_inversed_rad_div")];
+            tensor<fp32, [1, 1200000, 9]> var_35 = floor(x = _inversed_rad_div)[name = tensor<string, []>("op_35")];
+            tensor<fp32, [1, 1200000, 9]> rad_mod = sub(x = _inversed_rad_div, y = var_35)[name = tensor<string, []>("rad_mod")];
+            tensor<int32, [3]> var_38_begin_0 = const()[name = tensor<string, []>("op_38_begin_0"), val = tensor<int32, [3]>([0, 0, 0])];
+            tensor<int32, [3]> var_38_end_0 = const()[name = tensor<string, []>("op_38_end_0"), val = tensor<int32, [3]>([1, 1200000, 9])];
+            tensor<bool, [3]> var_38_end_mask_0 = const()[name = tensor<string, []>("op_38_end_mask_0"), val = tensor<bool, [3]>([true, true, true])];
+            tensor<int32, [3]> var_38_stride_0 = const()[name = tensor<string, []>("op_38_stride_0"), val = tensor<int32, [3]>([1, 300, 1])];
+            tensor<fp32, [1, 4000, 9]> var_38 = slice_by_index(begin = var_38_begin_0, end = var_38_end_0, end_mask = var_38_end_mask_0, stride = var_38_stride_0, x = rad_mod)[name = tensor<string, []>("op_38")];
+            tensor<bool, []> var_40_exclusive_0 = const()[name = tensor<string, []>("op_40_exclusive_0"), val = tensor<bool, []>(false)];
+            tensor<bool, []> var_40_reverse_0 = const()[name = tensor<string, []>("op_40_reverse_0"), val = tensor<bool, []>(false)];
+            tensor<fp32, [1, 4000, 9]> var_40 = cumsum(axis = var_23, exclusive = var_40_exclusive_0, reverse = var_40_reverse_0, x = var_38)[name = tensor<string, []>("op_40")];
+            tensor<fp32, []> var_41_promoted = const()[name = tensor<string, []>("op_41_promoted"), val = tensor<fp32, []>(0x1p+1)];
+            tensor<fp32, [1, 4000, 9]> var_42 = mul(x = var_40, y = var_41_promoted)[name = tensor<string, []>("op_42")];
+            tensor<fp32, []> var_43 = const()[name = tensor<string, []>("op_43"), val = tensor<fp32, []>(0x1.921fb6p+1)];
+            tensor<fp32, [1, 4000, 9]> phase = mul(x = var_42, y = var_43)[name = tensor<string, []>("phase")];
+            tensor<fp32, []> var_45_promoted = const()[name = tensor<string, []>("op_45_promoted"), val = tensor<fp32, []>(0x1.2cp+8)];
+            tensor<fp32, [1, 4000, 9]> phase_scaled = mul(x = phase, y = var_45_promoted)[name = tensor<string, []>("phase_scaled")];
+            tensor<int32, [3]> transpose_0_perm_0 = const()[name = tensor<string, []>("transpose_0_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
+            tensor<int32, [3]> tile_0_reps_0 = const()[name = tensor<string, []>("tile_0_reps_0"), val = tensor<int32, [3]>([300, 1, 1])];
+            tensor<fp32, [4000, 1, 9]> transpose_0 = transpose(perm = transpose_0_perm_0, x = phase_scaled)[name = tensor<string, []>("transpose_5")];
+            tensor<fp32, [1200000, 1, 9]> tile_0 = tile(reps = tile_0_reps_0, x = transpose_0)[name = tensor<string, []>("tile_0")];
+            tensor<int32, [4]> concat_0 = const()[name = tensor<string, []>("concat_0"), val = tensor<int32, [4]>([300, 4000, 1, 9])];
+            tensor<fp32, [300, 4000, 1, 9]> reshape_0 = reshape(shape = concat_0, x = tile_0)[name = tensor<string, []>("reshape_0")];
+            tensor<int32, [4]> transpose_1_perm_0 = const()[name = tensor<string, []>("transpose_1_perm_0"), val = tensor<int32, [4]>([1, 0, 2, 3])];
+            tensor<int32, [3]> concat_1 = const()[name = tensor<string, []>("concat_1"), val = tensor<int32, [3]>([-1, 1, 9])];
+            tensor<fp32, [4000, 300, 1, 9]> transpose_1 = transpose(perm = transpose_1_perm_0, x = reshape_0)[name = tensor<string, []>("transpose_4")];
+            tensor<fp32, [1200000, 1, 9]> reshape_1 = reshape(shape = concat_1, x = transpose_1)[name = tensor<string, []>("reshape_1")];
+            tensor<int32, [3]> left_perm_0 = const()[name = tensor<string, []>("left_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
+            tensor<int32, [3]> var_49_begin_0 = const()[name = tensor<string, []>("op_49_begin_0"), val = tensor<int32, [3]>([0, -1, 0])];
+            tensor<int32, [3]> var_49_end_0 = const()[name = tensor<string, []>("op_49_end_0"), val = tensor<int32, [3]>([1, 4000, 9])];
+            tensor<bool, [3]> var_49_end_mask_0 = const()[name = tensor<string, []>("op_49_end_mask_0"), val = tensor<bool, [3]>([true, true, true])];
+            tensor<fp32, [1, 1, 9]> var_49 = slice_by_index(begin = var_49_begin_0, end = var_49_end_0, end_mask = var_49_end_mask_0, x = phase_scaled)[name = tensor<string, []>("op_49")];
+            tensor<int32, [3]> var_52_begin_0 = const()[name = tensor<string, []>("op_52_begin_0"), val = tensor<int32, [3]>([0, 1, 0])];
+            tensor<int32, [3]> var_52_end_0 = const()[name = tensor<string, []>("op_52_end_0"), val = tensor<int32, [3]>([1, 4000, 9])];
+            tensor<bool, [3]> var_52_end_mask_0 = const()[name = tensor<string, []>("op_52_end_mask_0"), val = tensor<bool, [3]>([true, true, true])];
+            tensor<fp32, [1, 3999, 9]> var_52 = slice_by_index(begin = var_52_begin_0, end = var_52_end_0, end_mask = var_52_end_mask_0, x = phase_scaled)[name = tensor<string, []>("op_52")];
+            tensor<bool, []> shifted_interleave_0 = const()[name = tensor<string, []>("shifted_interleave_0"), val = tensor<bool, []>(false)];
+            tensor<fp32, [1, 4000, 9]> shifted = concat(axis = var_23, interleave = shifted_interleave_0, values = (var_52, var_49))[name = tensor<string, []>("shifted")];
+            tensor<int32, [3]> transpose_2_perm_0 = const()[name = tensor<string, []>("transpose_2_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
+            tensor<int32, [3]> tile_1_reps_0 = const()[name = tensor<string, []>("tile_1_reps_0"), val = tensor<int32, [3]>([300, 1, 1])];
+            tensor<fp32, [4000, 1, 9]> transpose_2 = transpose(perm = transpose_2_perm_0, x = shifted)[name = tensor<string, []>("transpose_3")];
+            tensor<fp32, [1200000, 1, 9]> tile_1 = tile(reps = tile_1_reps_0, x = transpose_2)[name = tensor<string, []>("tile_1")];
+            tensor<int32, [4]> concat_2 = const()[name = tensor<string, []>("concat_2"), val = tensor<int32, [4]>([300, 4000, 1, 9])];
+            tensor<fp32, [300, 4000, 1, 9]> reshape_2 = reshape(shape = concat_2, x = tile_1)[name = tensor<string, []>("reshape_2")];
+            tensor<int32, [4]> transpose_3_perm_0 = const()[name = tensor<string, []>("transpose_3_perm_0"), val = tensor<int32, [4]>([1, 0, 2, 3])];
+            tensor<int32, [3]> concat_3 = const()[name = tensor<string, []>("concat_3"), val = tensor<int32, [3]>([-1, 1, 9])];
+            tensor<fp32, [4000, 300, 1, 9]> transpose_3 = transpose(perm = transpose_3_perm_0, x = reshape_2)[name = tensor<string, []>("transpose_2")];
+            tensor<fp32, [1200000, 1, 9]> reshape_3 = reshape(shape = concat_3, x = transpose_3)[name = tensor<string, []>("reshape_3")];
+            tensor<int32, [3]> right_perm_0 = const()[name = tensor<string, []>("right_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
+            tensor<fp32, [1, 1200000, 1]> const_2_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [1200000]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(64))), lut = tensor<fp32, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1200128))), name = tensor<string, []>("const_2_palettized"), shape = tensor<uint32, [3]>([1, 1200000, 1])];
+            tensor<fp32, [1, 1200000, 1]> op_58_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [1200000]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1201216))), lut = tensor<fp32, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2401280))), name = tensor<string, []>("op_58_palettized"), shape = tensor<uint32, [3]>([1, 1200000, 1])];
+            tensor<fp32, [1, 1200000, 9]> left = transpose(perm = left_perm_0, x = reshape_1)[name = tensor<string, []>("transpose_1")];
+            tensor<fp32, [1, 1200000, 9]> var_59 = mul(x = left, y = op_58_palettized)[name = tensor<string, []>("op_59")];
+            tensor<fp32, [1, 1200000, 9]> right = transpose(perm = right_perm_0, x = reshape_3)[name = tensor<string, []>("transpose_0")];
+            tensor<fp32, [1, 1200000, 9]> var_60 = mul(x = right, y = const_2_palettized)[name = tensor<string, []>("op_60")];
+            tensor<fp32, [1, 1200000, 9]> phase_up = add(x = var_59, y = var_60)[name = tensor<string, []>("phase_up")];
+            tensor<fp32, [1, 1200000, 9]> sines = sin(x = phase_up)[name = tensor<string, []>("sines")];
+            tensor<fp32, []> var_63 = const()[name = tensor<string, []>("op_63"), val = tensor<fp32, []>(0x1.99999ap-4)];
+            tensor<fp32, [1, 1200000, 9]> sine_waves_1 = mul(x = sines, y = var_63)[name = tensor<string, []>("sine_waves")];
+            tensor<fp32, []> var_16_promoted = const()[name = tensor<string, []>("op_16_promoted"), val = tensor<fp32, []>(0x1.4p+3)];
+            tensor<bool, [1, 1200000, 1]> var_65 = greater(x = f0, y = var_16_promoted)[name = tensor<string, []>("op_65")];
+            tensor<string, []> cast_2_dtype_0 = const()[name = tensor<string, []>("cast_2_dtype_0"), val = tensor<string, []>("fp32")];
+            tensor<fp32, []> var_67 = const()[name = tensor<string, []>("op_67"), val = tensor<fp32, []>(0x1.89374cp-9)];
+            tensor<fp32, [1, 1200000, 1]> uv = cast(dtype = cast_2_dtype_0, x = var_65)[name = tensor<string, []>("cast_0")];
+            tensor<fp32, [1, 1200000, 1]> var_68 = mul(x = uv, y = var_67)[name = tensor<string, []>("op_68")];
+            tensor<fp32, []> var_23_promoted = const()[name = tensor<string, []>("op_23_promoted"), val = tensor<fp32, []>(0x1p+0)];
+            tensor<fp32, [1, 1200000, 1]> var_69 = sub(x = var_23_promoted, y = uv)[name = tensor<string, []>("op_69")];
+            tensor<fp32, []> var_70 = const()[name = tensor<string, []>("op_70"), val = tensor<fp32, []>(0x1.99999ap-4)];
+            tensor<fp32, [1, 1200000, 1]> var_71 = mul(x = var_69, y = var_70)[name = tensor<string, []>("op_71")];
+            tensor<fp32, []> _inversed_73_y_0 = const()[name = tensor<string, []>("_inversed_73_y_0"), val = tensor<fp32, []>(0x1.555556p-2)];
+            tensor<fp32, [1, 1200000, 1]> _inversed_73 = mul(x = var_71, y = _inversed_73_y_0)[name = tensor<string, []>("_inversed_73")];
+            tensor<fp32, [1, 1200000, 1]> noise_amp = add(x = var_68, y = _inversed_73)[name = tensor<string, []>("noise_amp")];
+            tensor<fp32, []> var_75 = const()[name = tensor<string, []>("op_75"), val = tensor<fp32, []>(0x1.9p+6)];
+            tensor<fp32, [1, 1200000, 9]> var_76 = mul(x = f0_values, y = var_75)[name = tensor<string, []>("op_76")];
+            tensor<fp32, [1, 1200000, 9]> var_77 = sin(x = var_76)[name = tensor<string, []>("op_77")];
+            tensor<fp32, [1, 1200000, 9]> noise = mul(x = noise_amp, y = var_77)[name = tensor<string, []>("noise")];
+            tensor<fp32, [1, 1200000, 9]> var_79 = mul(x = sine_waves_1, y = uv)[name = tensor<string, []>("op_79")];
+            tensor<fp32, [1, 1200000, 9]> sine_waves = add(x = var_79, y = noise)[name = tensor<string, []>("op_80")];
+        } -> (sine_waves, uv);
+}

ANE/styletts2_ane_noise.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b988942b88ad44925a6cf81bf3d1108b1da6c2f2c84edc19091e75cd29e53d6
+size 2402368

ANE/styletts2_ane_noise.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54d28925e686100531855b225d53c4c50f94bac6eac7a2d0193340d4dcecb0bc
+size 12114

ANE/styletts2_ane_noise.mlpackage/Data/com.apple.CoreML/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b988942b88ad44925a6cf81bf3d1108b1da6c2f2c84edc19091e75cd29e53d6
+size 2402368

ANE/styletts2_ane_noise.mlpackage/Manifest.json

@@ -0,0 +1,18 @@
+{
+    "fileFormatVersion": "1.0.0",
+    "itemInfoEntries": {
+        "47473672-28F9-487C-9ED6-63C6D82ABB57": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Weights",
+            "name": "weights",
+            "path": "com.apple.CoreML/weights"
+        },
+        "88C15F91-FFC7-4B76-B3DF-CFFD20CF2957": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Specification",
+            "name": "model.mlmodel",
+            "path": "com.apple.CoreML/model.mlmodel"
+        }
+    },
+    "rootModelIdentifier": "88C15F91-FFC7-4B76-B3DF-CFFD20CF2957"
+}

ANE/styletts2_ane_plbert.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a828a4caef0ea1661731eb81369d07f3c0bcbfa9bbb9d2aaebe0c75fd1136221
+size 243

ANE/styletts2_ane_plbert.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e968b81438916aee673e4daa1410ca1b6490946302de377e2a3c96389f8c9bed
+size 380

ANE/styletts2_ane_plbert.mlmodelc/metadata.json

@@ -0,0 +1,83 @@
+[
+  {
+    "metadataOutputVersion" : "3.0",
+    "shortDescription" : "StyleTTS2-ANE Stage 1: PLBERT (Albert) — fp16, ANE.",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16)",
+        "shortDescription" : "",
+        "shape" : "[]",
+        "name" : "bert_dur",
+        "type" : "MultiArray"
+      }
+    ],
+    "storagePrecision" : "Mixed (Float16, Palettized (8 bits))",
+    "modelParameters" : [
+
+    ],
+    "specificationVersion" : 8,
+    "mlProgramOperationTypeHistogram" : {
+      "Fill" : 1,
+      "Ios17.reshape" : 48,
+      "Ios16.softmax" : 12,
+      "Ios17.matmul" : 24,
+      "Ios17.transpose" : 48,
+      "Select" : 1,
+      "Ios17.expandDims" : 2,
+      "Ios17.add" : 38,
+      "Tile" : 2,
+      "Ios17.sliceByIndex" : 2,
+      "Shape" : 16,
+      "Ios17.gather" : 29,
+      "Ios17.layerNorm" : 25,
+      "Ios17.cast" : 44,
+      "Ios17.sub" : 1,
+      "Ios16.constexprLutToDense" : 9,
+      "Ios16.gelu" : 12,
+      "Ios17.realDiv" : 2,
+      "Ios17.linear" : 73,
+      "Ios17.concat" : 26,
+      "Ios17.mul" : 12
+    },
+    "computePrecision" : "Mixed (Float16, Int16, Int32, UInt16)",
+    "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"
+    },
+    "inputSchema" : [
+      {
+        "dataType" : "Int32",
+        "hasShapeFlexibility" : "1",
+        "isOptional" : "0",
+        "shapeFlexibility" : "1 × 2...512",
+        "shapeRange" : "[[1, 1], [2, 512]]",
+        "formattedType" : "MultiArray (Int32 1 × 64)",
+        "type" : "MultiArray",
+        "shape" : "[1, 64]",
+        "name" : "tokens",
+        "shortDescription" : ""
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.version" : "8.3.0",
+      "com.github.apple.coremltools.source_dialect" : "TorchScript",
+      "com.github.apple.coremltools.source" : "torch==2.11.0"
+    },
+    "generatedClassName" : "styletts2_ane_plbert",
+    "method" : "predict"
+  }
+]

ANE/styletts2_ane_plbert.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59c7da8562b26c0842a3941543ebf48d94629105022320c9d8cd590761b6fcab
+size 5718848

ANE/styletts2_ane_plbert.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a91ff072efb2aa10a26f76e280108f110a877429d74ec76769aa560c3ebbc7e
+size 148365

ANE/styletts2_ane_plbert.mlpackage/Data/com.apple.CoreML/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59c7da8562b26c0842a3941543ebf48d94629105022320c9d8cd590761b6fcab
+size 5718848

ANE/styletts2_ane_plbert.mlpackage/Manifest.json

@@ -0,0 +1,18 @@
+{
+    "fileFormatVersion": "1.0.0",
+    "itemInfoEntries": {
+        "3B613FCB-B965-4C2D-94A7-4D5C80EDE23F": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Weights",
+            "name": "weights",
+            "path": "com.apple.CoreML/weights"
+        },
+        "440AF525-7EC6-4E07-8019-7BF58173E928": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Specification",
+            "name": "model.mlmodel",
+            "path": "com.apple.CoreML/model.mlmodel"
+        }
+    },
+    "rootModelIdentifier": "440AF525-7EC6-4E07-8019-7BF58173E928"
+}

ANE/styletts2_ane_postbert.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:140544065220d26a02c8c8c6592a298c95dd14b80c0b22bc4005189cc86698d2
+size 243

ANE/styletts2_ane_postbert.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dde3c94cb2d81022ada31a29ca27224bdc7d3c75672565969188fe7be761c172
+size 542

ANE/styletts2_ane_postbert.mlmodelc/metadata.json

@@ -0,0 +1,133 @@
+[
+  {
+    "metadataOutputVersion" : "3.0",
+    "shortDescription" : "StyleTTS2-ANE Stage 2: TextEncoder + DurationEncoder + duration head — fp16, ANE.",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16)",
+        "shortDescription" : "",
+        "shape" : "[]",
+        "name" : "t_en",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16)",
+        "shortDescription" : "",
+        "shape" : "[]",
+        "name" : "d",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16)",
+        "shortDescription" : "",
+        "shape" : "[]",
+        "name" : "pred_dur_log",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16)",
+        "shortDescription" : "",
+        "shape" : "[]",
+        "name" : "fixed_embedding",
+        "type" : "MultiArray"
+      }
+    ],
+    "storagePrecision" : "Float16",
+    "modelParameters" : [
+
+    ],
+    "specificationVersion" : 8,
+    "mlProgramOperationTypeHistogram" : {
+      "Split" : 3,
+      "Ios17.mul" : 3,
+      "Ios17.linear" : 5,
+      "Ios17.transpose" : 27,
+      "Ios17.sub" : 2,
+      "Ios16.fillLike" : 1,
+      "Ios17.conv" : 3,
+      "Ios16.cumsum" : 1,
+      "Ios17.leakyRelu" : 3,
+      "Ios17.gather" : 2,
+      "Ios17.add" : 7,
+      "Ios17.sliceByIndex" : 3,
+      "Ios17.lstm" : 5,
+      "Ios17.concat" : 4,
+      "Ios17.expandDims" : 2,
+      "Ios17.squeeze" : 1,
+      "Ios17.cast" : 3,
+      "Ios17.layerNorm" : 6,
+      "Ios17.reshape" : 3
+    },
+    "computePrecision" : "Mixed (Float16, Float32, Int16, Int32, UInt16)",
+    "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"
+    },
+    "inputSchema" : [
+      {
+        "dataType" : "Float16",
+        "hasShapeFlexibility" : "1",
+        "isOptional" : "0",
+        "shapeFlexibility" : "1 × 2...512 × 768",
+        "shapeRange" : "[[1, 1], [2, 512], [768, 768]]",
+        "formattedType" : "MultiArray (Float16 1 × 64 × 768)",
+        "type" : "MultiArray",
+        "shape" : "[1, 64, 768]",
+        "name" : "bert_dur",
+        "shortDescription" : ""
+      },
+      {
+        "dataType" : "Int32",
+        "hasShapeFlexibility" : "1",
+        "isOptional" : "0",
+        "shapeFlexibility" : "1 × 2...512",
+        "shapeRange" : "[[1, 1], [2, 512]]",
+        "formattedType" : "MultiArray (Int32 1 × 64)",
+        "type" : "MultiArray",
+        "shape" : "[1, 64]",
+        "name" : "tokens",
+        "shortDescription" : ""
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 1 × 256)",
+        "shortDescription" : "",
+        "shape" : "[1, 256]",
+        "name" : "style",
+        "type" : "MultiArray"
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.version" : "8.3.0",
+      "com.github.apple.coremltools.source" : "torch==2.11.0",
+      "com.github.apple.coremltools.source_dialect" : "TorchScript"
+    },
+    "generatedClassName" : "styletts2_ane_postbert",
+    "method" : "predict"
+  }
+]

ANE/styletts2_ane_postbert.mlmodelc/model.mil

@@ -0,0 +1,265 @@
+program(1.0)
+[buildInfo = dict<tensor<string, []>, tensor<string, []>>({{"coremlc-component-MIL", "3520.4.1"}, {"coremlc-version", "3520.5.1"}, {"coremltools-component-torch", "2.11.0"}, {"coremltools-source-dialect", "TorchScript"}, {"coremltools-version", "8.3.0"}})]
+{
+    func main<ios17>(tensor<fp16, [1, ?, 768]> bert_dur, tensor<fp16, [1, 256]> style, tensor<int32, [1, ?]> tokens) [FlexibleShapeInformation = tuple<tuple<tensor<string, []>, dict<tensor<string, []>, tensor<int32, [?]>>>, tuple<tensor<string, []>, dict<tensor<string, []>, list<tensor<int32, [2]>, ?>>>>((("DefaultShapes", {{"bert_dur", [1, 64, 768]}, {"tokens", [1, 64]}}), ("RangeDims", {{"bert_dur", [[1, 1], [2, 512], [768, 768]]}, {"tokens", [[1, 1], [2, 512]]}})))] {
+            tensor<int32, [2]> style_begin_0 = const()[name = tensor<string, []>("style_begin_0"), val = tensor<int32, [2]>([0, 0])];
+            tensor<int32, [2]> style_end_0 = const()[name = tensor<string, []>("style_end_0"), val = tensor<int32, [2]>([1, 128])];
+            tensor<bool, [2]> style_end_mask_0 = const()[name = tensor<string, []>("style_end_mask_0"), val = tensor<bool, [2]>([true, false])];
+            tensor<fp16, [1, 128]> style_cast_fp16 = slice_by_index(begin = style_begin_0, end = style_end_0, end_mask = style_end_mask_0, x = style)[name = tensor<string, []>("style_cast_fp16")];
+            tensor<fp32, []> var_30 = const()[name = tensor<string, []>("op_30"), val = tensor<fp32, []>(0x1.99999ap-3)];
+            tensor<int32, []> x_1_axis_0 = const()[name = tensor<string, []>("x_1_axis_0"), val = tensor<int32, []>(0)];
+            tensor<int32, []> x_1_batch_dims_0 = const()[name = tensor<string, []>("x_1_batch_dims_0"), val = tensor<int32, []>(0)];
+            tensor<bool, []> x_1_validate_indices_0 = const()[name = tensor<string, []>("x_1_validate_indices_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [178, 512]> text_encoder_embedding_weight_to_fp16 = const()[name = tensor<string, []>("text_encoder_embedding_weight_to_fp16"), val = tensor<fp16, [178, 512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(64)))];
+            tensor<string, []> tokens_to_int16_dtype_0 = const()[name = tensor<string, []>("tokens_to_int16_dtype_0"), val = tensor<string, []>("int16")];
+            tensor<int16, [1, ?]> tokens_to_int16 = cast(dtype = tokens_to_int16_dtype_0, x = tokens)[name = tensor<string, []>("cast_14")];
+            tensor<fp16, [1, ?, 512]> x_1_cast_fp16_cast_uint16 = gather(axis = x_1_axis_0, batch_dims = x_1_batch_dims_0, indices = tokens_to_int16, validate_indices = x_1_validate_indices_0, x = text_encoder_embedding_weight_to_fp16)[name = tensor<string, []>("x_1_cast_fp16_cast_uint16")];
+            tensor<int32, [3]> input_1_perm_0 = const()[name = tensor<string, []>("input_1_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
+            tensor<string, []> x_3_pad_type_0 = const()[name = tensor<string, []>("x_3_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> x_3_pad_0 = const()[name = tensor<string, []>("x_3_pad_0"), val = tensor<int32, [2]>([2, 2])];
+            tensor<int32, [1]> x_3_strides_0 = const()[name = tensor<string, []>("x_3_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> x_3_dilations_0 = const()[name = tensor<string, []>("x_3_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> x_3_groups_0 = const()[name = tensor<string, []>("x_3_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [512, 512, 5]> weight_3_to_fp16 = const()[name = tensor<string, []>("weight_3_to_fp16"), val = tensor<fp16, [512, 512, 5]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(182400)))];
+            tensor<fp16, [512]> text_encoder_cnn_0_0_bias_to_fp16 = const()[name = tensor<string, []>("text_encoder_cnn_0_0_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2803904)))];
+            tensor<fp16, [1, 512, ?]> input_1_cast_fp16 = transpose(perm = input_1_perm_0, x = x_1_cast_fp16_cast_uint16)[name = tensor<string, []>("transpose_36")];
+            tensor<fp16, [1, 512, ?]> x_3_cast_fp16 = conv(bias = text_encoder_cnn_0_0_bias_to_fp16, dilations = x_3_dilations_0, groups = x_3_groups_0, pad = x_3_pad_0, pad_type = x_3_pad_type_0, strides = x_3_strides_0, weight = weight_3_to_fp16, x = input_1_cast_fp16)[name = tensor<string, []>("x_3_cast_fp16")];
+            tensor<int32, [3]> input_3_perm_0 = const()[name = tensor<string, []>("input_3_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [1]> x_5_axes_0 = const()[name = tensor<string, []>("x_5_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, [512]> text_encoder_cnn_0_1_gamma_to_fp16 = const()[name = tensor<string, []>("text_encoder_cnn_0_1_gamma_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2804992)))];
+            tensor<fp16, [512]> text_encoder_cnn_0_1_beta_to_fp16 = const()[name = tensor<string, []>("text_encoder_cnn_0_1_beta_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2806080)))];
+            tensor<fp16, []> var_28_to_fp16 = const()[name = tensor<string, []>("op_28_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<fp16, [1, ?, 512]> input_3_cast_fp16 = transpose(perm = input_3_perm_0, x = x_3_cast_fp16)[name = tensor<string, []>("transpose_35")];
+            tensor<fp16, [1, ?, 512]> x_5_cast_fp16 = layer_norm(axes = x_5_axes_0, beta = text_encoder_cnn_0_1_beta_to_fp16, epsilon = var_28_to_fp16, gamma = text_encoder_cnn_0_1_gamma_to_fp16, x = input_3_cast_fp16)[name = tensor<string, []>("x_5_cast_fp16")];
+            tensor<int32, [3]> input_5_perm_0 = const()[name = tensor<string, []>("input_5_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<fp16, [1, 512, ?]> input_5_cast_fp16 = transpose(perm = input_5_perm_0, x = x_5_cast_fp16)[name = tensor<string, []>("transpose_34")];
+            tensor<fp16, [1, 512, ?]> input_7_cast_fp16 = leaky_relu(alpha = var_30, x = input_5_cast_fp16)[name = tensor<string, []>("input_7_cast_fp16")];
+            tensor<string, []> x_7_pad_type_0 = const()[name = tensor<string, []>("x_7_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> x_7_pad_0 = const()[name = tensor<string, []>("x_7_pad_0"), val = tensor<int32, [2]>([2, 2])];
+            tensor<int32, [1]> x_7_strides_0 = const()[name = tensor<string, []>("x_7_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> x_7_dilations_0 = const()[name = tensor<string, []>("x_7_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> x_7_groups_0 = const()[name = tensor<string, []>("x_7_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [512, 512, 5]> weight_7_to_fp16 = const()[name = tensor<string, []>("weight_7_to_fp16"), val = tensor<fp16, [512, 512, 5]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2807168)))];
+            tensor<fp16, [512]> text_encoder_cnn_1_0_bias_to_fp16 = const()[name = tensor<string, []>("text_encoder_cnn_1_0_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5428672)))];
+            tensor<fp16, [1, 512, ?]> x_7_cast_fp16 = conv(bias = text_encoder_cnn_1_0_bias_to_fp16, dilations = x_7_dilations_0, groups = x_7_groups_0, pad = x_7_pad_0, pad_type = x_7_pad_type_0, strides = x_7_strides_0, weight = weight_7_to_fp16, x = input_7_cast_fp16)[name = tensor<string, []>("x_7_cast_fp16")];
+            tensor<int32, [3]> input_11_perm_0 = const()[name = tensor<string, []>("input_11_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [1]> x_9_axes_0 = const()[name = tensor<string, []>("x_9_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, [512]> text_encoder_cnn_1_1_gamma_to_fp16 = const()[name = tensor<string, []>("text_encoder_cnn_1_1_gamma_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5429760)))];
+            tensor<fp16, [512]> text_encoder_cnn_1_1_beta_to_fp16 = const()[name = tensor<string, []>("text_encoder_cnn_1_1_beta_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5430848)))];
+            tensor<fp16, [1, ?, 512]> input_11_cast_fp16 = transpose(perm = input_11_perm_0, x = x_7_cast_fp16)[name = tensor<string, []>("transpose_33")];
+            tensor<fp16, [1, ?, 512]> x_9_cast_fp16 = layer_norm(axes = x_9_axes_0, beta = text_encoder_cnn_1_1_beta_to_fp16, epsilon = var_28_to_fp16, gamma = text_encoder_cnn_1_1_gamma_to_fp16, x = input_11_cast_fp16)[name = tensor<string, []>("x_9_cast_fp16")];
+            tensor<int32, [3]> input_13_perm_0 = const()[name = tensor<string, []>("input_13_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<fp16, [1, 512, ?]> input_13_cast_fp16 = transpose(perm = input_13_perm_0, x = x_9_cast_fp16)[name = tensor<string, []>("transpose_32")];
+            tensor<fp16, [1, 512, ?]> input_15_cast_fp16 = leaky_relu(alpha = var_30, x = input_13_cast_fp16)[name = tensor<string, []>("input_15_cast_fp16")];
+            tensor<string, []> x_11_pad_type_0 = const()[name = tensor<string, []>("x_11_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> x_11_pad_0 = const()[name = tensor<string, []>("x_11_pad_0"), val = tensor<int32, [2]>([2, 2])];
+            tensor<int32, [1]> x_11_strides_0 = const()[name = tensor<string, []>("x_11_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> x_11_dilations_0 = const()[name = tensor<string, []>("x_11_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> x_11_groups_0 = const()[name = tensor<string, []>("x_11_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [512, 512, 5]> weight_11_to_fp16 = const()[name = tensor<string, []>("weight_11_to_fp16"), val = tensor<fp16, [512, 512, 5]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5431936)))];
+            tensor<fp16, [512]> text_encoder_cnn_2_0_bias_to_fp16 = const()[name = tensor<string, []>("text_encoder_cnn_2_0_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8053440)))];
+            tensor<fp16, [1, 512, ?]> x_11_cast_fp16 = conv(bias = text_encoder_cnn_2_0_bias_to_fp16, dilations = x_11_dilations_0, groups = x_11_groups_0, pad = x_11_pad_0, pad_type = x_11_pad_type_0, strides = x_11_strides_0, weight = weight_11_to_fp16, x = input_15_cast_fp16)[name = tensor<string, []>("x_11_cast_fp16")];
+            tensor<int32, [3]> input_19_perm_0 = const()[name = tensor<string, []>("input_19_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [1]> x_13_axes_0 = const()[name = tensor<string, []>("x_13_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, [512]> text_encoder_cnn_2_1_gamma_to_fp16 = const()[name = tensor<string, []>("text_encoder_cnn_2_1_gamma_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8054528)))];
+            tensor<fp16, [512]> text_encoder_cnn_2_1_beta_to_fp16 = const()[name = tensor<string, []>("text_encoder_cnn_2_1_beta_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8055616)))];
+            tensor<fp16, [1, ?, 512]> input_19_cast_fp16 = transpose(perm = input_19_perm_0, x = x_11_cast_fp16)[name = tensor<string, []>("transpose_31")];
+            tensor<fp16, [1, ?, 512]> x_13_cast_fp16 = layer_norm(axes = x_13_axes_0, beta = text_encoder_cnn_2_1_beta_to_fp16, epsilon = var_28_to_fp16, gamma = text_encoder_cnn_2_1_gamma_to_fp16, x = input_19_cast_fp16)[name = tensor<string, []>("x_13_cast_fp16")];
+            tensor<fp16, [1, ?, 512]> input_23_cast_fp16 = leaky_relu(alpha = var_30, x = x_13_cast_fp16)[name = tensor<string, []>("input_23_cast_fp16")];
+            tensor<int32, [3]> input_25_batch_first_transpose_perm_0 = const()[name = tensor<string, []>("input_25_batch_first_transpose_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
+            tensor<string, []> x_17_batch_first_direction_0 = const()[name = tensor<string, []>("x_17_batch_first_direction_0"), val = tensor<string, []>("bidirectional")];
+            tensor<bool, []> x_17_batch_first_output_sequence_0 = const()[name = tensor<string, []>("x_17_batch_first_output_sequence_0"), val = tensor<bool, []>(true)];
+            tensor<string, []> x_17_batch_first_recurrent_activation_0 = const()[name = tensor<string, []>("x_17_batch_first_recurrent_activation_0"), val = tensor<string, []>("sigmoid")];
+            tensor<string, []> x_17_batch_first_cell_activation_0 = const()[name = tensor<string, []>("x_17_batch_first_cell_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<string, []> x_17_batch_first_activation_0 = const()[name = tensor<string, []>("x_17_batch_first_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<fp16, [1, 512]> x_17_batch_first_lstm_h0_reshaped_to_fp16 = const()[name = tensor<string, []>("x_17_batch_first_lstm_h0_reshaped_to_fp16"), val = tensor<fp16, [1, 512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8056704)))];
+            tensor<fp16, [1024, 512]> concat_4_to_fp16 = const()[name = tensor<string, []>("concat_4_to_fp16"), val = tensor<fp16, [1024, 512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8057792)))];
+            tensor<fp16, [1024, 256]> concat_5_to_fp16 = const()[name = tensor<string, []>("concat_5_to_fp16"), val = tensor<fp16, [1024, 256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(9106432)))];
+            tensor<fp16, [1024]> add_1_to_fp16 = const()[name = tensor<string, []>("add_1_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(9630784)))];
+            tensor<fp16, [1024, 512]> concat_6_to_fp16 = const()[name = tensor<string, []>("concat_6_to_fp16"), val = tensor<fp16, [1024, 512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(9632896)))];
+            tensor<fp16, [1024, 256]> concat_7_to_fp16 = const()[name = tensor<string, []>("concat_7_to_fp16"), val = tensor<fp16, [1024, 256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(10681536)))];
+            tensor<fp16, [1024]> add_2_to_fp16 = const()[name = tensor<string, []>("add_2_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(11205888)))];
+            tensor<fp16, [?, 1, 512]> input_25_batch_first_transpose_cast_fp16 = transpose(perm = input_25_batch_first_transpose_perm_0, x = input_23_cast_fp16)[name = tensor<string, []>("transpose_30")];
+            tensor<fp16, [?, 1, 512]> x_17_batch_first_cast_fp16_0, tensor<fp16, [1, 512]> x_17_batch_first_cast_fp16_1, tensor<fp16, [1, 512]> x_17_batch_first_cast_fp16_2 = lstm(activation = x_17_batch_first_activation_0, bias = add_1_to_fp16, bias_back = add_2_to_fp16, cell_activation = x_17_batch_first_cell_activation_0, direction = x_17_batch_first_direction_0, initial_c = x_17_batch_first_lstm_h0_reshaped_to_fp16, initial_h = x_17_batch_first_lstm_h0_reshaped_to_fp16, output_sequence = x_17_batch_first_output_sequence_0, recurrent_activation = x_17_batch_first_recurrent_activation_0, weight_hh = concat_5_to_fp16, weight_hh_back = concat_7_to_fp16, weight_ih = concat_4_to_fp16, weight_ih_back = concat_6_to_fp16, x = input_25_batch_first_transpose_cast_fp16)[name = tensor<string, []>("x_17_batch_first_cast_fp16")];
+            tensor<int32, [3]> transpose_0_perm_0 = const()[name = tensor<string, []>("transpose_0_perm_0"), val = tensor<int32, [3]>([1, 2, 0])];
+            tensor<fp16, [512, 768]> bert_encoder_weight_to_fp16 = const()[name = tensor<string, []>("bert_encoder_weight_to_fp16"), val = tensor<fp16, [512, 768]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(11208000)))];
+            tensor<fp16, [512]> bert_encoder_bias_to_fp16 = const()[name = tensor<string, []>("bert_encoder_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(11994496)))];
+            tensor<fp16, [1, ?, 512]> linear_0_cast_fp16 = linear(bias = bert_encoder_bias_to_fp16, weight = bert_encoder_weight_to_fp16, x = bert_dur)[name = tensor<string, []>("linear_0_cast_fp16")];
+            tensor<int32, [3]> x_19_perm_0 = const()[name = tensor<string, []>("x_19_perm_0"), val = tensor<int32, [3]>([0, -1, -2])];
+            tensor<int32, []> var_147 = const()[name = tensor<string, []>("op_147"), val = tensor<int32, []>(1)];
+            tensor<int32, [3]> var_163_begin_0 = const()[name = tensor<string, []>("op_163_begin_0"), val = tensor<int32, [3]>([0, 0, 0])];
+            tensor<int32, [3]> var_163_end_0 = const()[name = tensor<string, []>("op_163_end_0"), val = tensor<int32, [3]>([1, 1, 0])];
+            tensor<bool, [3]> var_163_end_mask_0 = const()[name = tensor<string, []>("op_163_end_mask_0"), val = tensor<bool, [3]>([true, false, true])];
+            tensor<fp16, [1, 512, ?]> x_19_cast_fp16 = transpose(perm = x_19_perm_0, x = linear_0_cast_fp16)[name = tensor<string, []>("transpose_28")];
+            tensor<fp16, [1, 1, ?]> var_163_cast_fp16 = slice_by_index(begin = var_163_begin_0, end = var_163_end_0, end_mask = var_163_end_mask_0, x = x_19_cast_fp16)[name = tensor<string, []>("op_163_cast_fp16")];
+            tensor<fp16, [1, 1, ?]> sub_0_cast_fp16 = sub(x = var_163_cast_fp16, y = var_163_cast_fp16)[name = tensor<string, []>("sub_0_cast_fp16")];
+            tensor<int32, [1]> var_166_axes_0 = const()[name = tensor<string, []>("op_166_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, [1, 128, 1]> var_166_cast_fp16 = expand_dims(axes = var_166_axes_0, x = style_cast_fp16)[name = tensor<string, []>("op_166_cast_fp16")];
+            tensor<fp16, [1, 128, ?]> s_chan_cast_fp16 = add(x = var_166_cast_fp16, y = sub_0_cast_fp16)[name = tensor<string, []>("s_chan_cast_fp16")];
+            tensor<bool, []> x_21_interleave_0 = const()[name = tensor<string, []>("x_21_interleave_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, 640, ?]> x_21_cast_fp16 = concat(axis = var_147, interleave = x_21_interleave_0, values = (x_19_cast_fp16, s_chan_cast_fp16))[name = tensor<string, []>("x_21_cast_fp16")];
+            tensor<int32, [3]> transpose_1_perm_0 = const()[name = tensor<string, []>("transpose_1_perm_0"), val = tensor<int32, [3]>([-1, 0, -2])];
+            tensor<string, []> input_29_batch_first_direction_0 = const()[name = tensor<string, []>("input_29_batch_first_direction_0"), val = tensor<string, []>("bidirectional")];
+            tensor<bool, []> input_29_batch_first_output_sequence_0 = const()[name = tensor<string, []>("input_29_batch_first_output_sequence_0"), val = tensor<bool, []>(true)];
+            tensor<string, []> input_29_batch_first_recurrent_activation_0 = const()[name = tensor<string, []>("input_29_batch_first_recurrent_activation_0"), val = tensor<string, []>("sigmoid")];
+            tensor<string, []> input_29_batch_first_cell_activation_0 = const()[name = tensor<string, []>("input_29_batch_first_cell_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<string, []> input_29_batch_first_activation_0 = const()[name = tensor<string, []>("input_29_batch_first_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<fp16, [1024, 640]> concat_14_to_fp16 = const()[name = tensor<string, []>("concat_14_to_fp16"), val = tensor<fp16, [1024, 640]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(11995584)))];
+            tensor<fp16, [1024, 256]> concat_15_to_fp16 = const()[name = tensor<string, []>("concat_15_to_fp16"), val = tensor<fp16, [1024, 256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(13306368)))];
+            tensor<fp16, [1024]> add_4_to_fp16 = const()[name = tensor<string, []>("add_4_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(13830720)))];
+            tensor<fp16, [1024, 640]> concat_16_to_fp16 = const()[name = tensor<string, []>("concat_16_to_fp16"), val = tensor<fp16, [1024, 640]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(13832832)))];
+            tensor<fp16, [1024, 256]> concat_17_to_fp16 = const()[name = tensor<string, []>("concat_17_to_fp16"), val = tensor<fp16, [1024, 256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(15143616)))];
+            tensor<fp16, [1024]> add_5_to_fp16 = const()[name = tensor<string, []>("add_5_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(15667968)))];
+            tensor<fp16, [?, 1, 640]> transpose_1_cast_fp16 = transpose(perm = transpose_1_perm_0, x = x_21_cast_fp16)[name = tensor<string, []>("transpose_27")];
+            tensor<fp16, [?, 1, 512]> input_29_batch_first_cast_fp16_0, tensor<fp16, [1, 512]> input_29_batch_first_cast_fp16_1, tensor<fp16, [1, 512]> input_29_batch_first_cast_fp16_2 = lstm(activation = input_29_batch_first_activation_0, bias = add_4_to_fp16, bias_back = add_5_to_fp16, cell_activation = input_29_batch_first_cell_activation_0, direction = input_29_batch_first_direction_0, initial_c = x_17_batch_first_lstm_h0_reshaped_to_fp16, initial_h = x_17_batch_first_lstm_h0_reshaped_to_fp16, output_sequence = input_29_batch_first_output_sequence_0, recurrent_activation = input_29_batch_first_recurrent_activation_0, weight_hh = concat_15_to_fp16, weight_hh_back = concat_17_to_fp16, weight_ih = concat_14_to_fp16, weight_ih_back = concat_16_to_fp16, x = transpose_1_cast_fp16)[name = tensor<string, []>("input_29_batch_first_cast_fp16")];
+            tensor<int32, [3]> transpose_7_perm_0 = const()[name = tensor<string, []>("transpose_7_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
+            tensor<fp16, [1024, 128]> duration_encoder_norms_0_fc_weight_to_fp16 = const()[name = tensor<string, []>("duration_encoder_norms_0_fc_weight_to_fp16"), val = tensor<fp16, [1024, 128]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(15670080)))];
+            tensor<fp16, [1024]> duration_encoder_norms_0_fc_bias_to_fp16 = const()[name = tensor<string, []>("duration_encoder_norms_0_fc_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(15932288)))];
+            tensor<fp16, [1, 1024]> linear_1_cast_fp16 = linear(bias = duration_encoder_norms_0_fc_bias_to_fp16, weight = duration_encoder_norms_0_fc_weight_to_fp16, x = style_cast_fp16)[name = tensor<string, []>("linear_1_cast_fp16")];
+            tensor<int32, [3]> var_200 = const()[name = tensor<string, []>("op_200"), val = tensor<int32, [3]>([1, 1024, 1])];
+            tensor<fp16, [1, 1024, 1]> h_3_cast_fp16 = reshape(shape = var_200, x = linear_1_cast_fp16)[name = tensor<string, []>("h_3_cast_fp16")];
+            tensor<int32, [2]> var_202_split_sizes_0 = const()[name = tensor<string, []>("op_202_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_202_axis_0 = const()[name = tensor<string, []>("op_202_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 1]> var_202_cast_fp16_0, tensor<fp16, [1, 512, 1]> var_202_cast_fp16_1 = split(axis = var_202_axis_0, split_sizes = var_202_split_sizes_0, x = h_3_cast_fp16)[name = tensor<string, []>("op_202_cast_fp16")];
+            tensor<int32, [3]> gamma_3_perm_0 = const()[name = tensor<string, []>("gamma_3_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [3]> beta_3_perm_0 = const()[name = tensor<string, []>("beta_3_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [1]> x_31_axes_0 = const()[name = tensor<string, []>("x_31_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, []> var_136_to_fp16 = const()[name = tensor<string, []>("op_136_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<fp16, [1, ?, 512]> transpose_7_cast_fp16 = transpose(perm = transpose_7_perm_0, x = input_29_batch_first_cast_fp16_0)[name = tensor<string, []>("transpose_26")];
+            tensor<fp16, [1, ?, 512]> x_31_cast_fp16 = layer_norm(axes = x_31_axes_0, epsilon = var_136_to_fp16, x = transpose_7_cast_fp16)[name = tensor<string, []>("x_31_cast_fp16")];
+            tensor<fp16, []> var_208_promoted_to_fp16 = const()[name = tensor<string, []>("op_208_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 1, 512]> gamma_3_cast_fp16 = transpose(perm = gamma_3_perm_0, x = var_202_cast_fp16_0)[name = tensor<string, []>("transpose_25")];
+            tensor<fp16, [1, 1, 512]> var_209_cast_fp16 = add(x = gamma_3_cast_fp16, y = var_208_promoted_to_fp16)[name = tensor<string, []>("op_209_cast_fp16")];
+            tensor<fp16, [1, ?, 512]> var_210_cast_fp16 = mul(x = var_209_cast_fp16, y = x_31_cast_fp16)[name = tensor<string, []>("op_210_cast_fp16")];
+            tensor<fp16, [1, 1, 512]> beta_3_cast_fp16 = transpose(perm = beta_3_perm_0, x = var_202_cast_fp16_1)[name = tensor<string, []>("transpose_24")];
+            tensor<fp16, [1, ?, 512]> x_33_cast_fp16 = add(x = var_210_cast_fp16, y = beta_3_cast_fp16)[name = tensor<string, []>("x_33_cast_fp16")];
+            tensor<int32, [3]> x_35_perm_0 = const()[name = tensor<string, []>("x_35_perm_0"), val = tensor<int32, [3]>([0, -1, -2])];
+            tensor<bool, []> x_37_interleave_0 = const()[name = tensor<string, []>("x_37_interleave_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, 512, ?]> x_35_cast_fp16 = transpose(perm = x_35_perm_0, x = x_33_cast_fp16)[name = tensor<string, []>("transpose_23")];
+            tensor<fp16, [1, 640, ?]> x_37_cast_fp16 = concat(axis = var_147, interleave = x_37_interleave_0, values = (x_35_cast_fp16, s_chan_cast_fp16))[name = tensor<string, []>("x_37_cast_fp16")];
+            tensor<int32, [3]> transpose_3_perm_0 = const()[name = tensor<string, []>("transpose_3_perm_0"), val = tensor<int32, [3]>([-1, 0, -2])];
+            tensor<string, []> input_35_batch_first_direction_0 = const()[name = tensor<string, []>("input_35_batch_first_direction_0"), val = tensor<string, []>("bidirectional")];
+            tensor<bool, []> input_35_batch_first_output_sequence_0 = const()[name = tensor<string, []>("input_35_batch_first_output_sequence_0"), val = tensor<bool, []>(true)];
+            tensor<string, []> input_35_batch_first_recurrent_activation_0 = const()[name = tensor<string, []>("input_35_batch_first_recurrent_activation_0"), val = tensor<string, []>("sigmoid")];
+            tensor<string, []> input_35_batch_first_cell_activation_0 = const()[name = tensor<string, []>("input_35_batch_first_cell_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<string, []> input_35_batch_first_activation_0 = const()[name = tensor<string, []>("input_35_batch_first_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<fp16, [1024, 640]> concat_24_to_fp16 = const()[name = tensor<string, []>("concat_24_to_fp16"), val = tensor<fp16, [1024, 640]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(15934400)))];
+            tensor<fp16, [1024, 256]> concat_25_to_fp16 = const()[name = tensor<string, []>("concat_25_to_fp16"), val = tensor<fp16, [1024, 256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(17245184)))];
+            tensor<fp16, [1024]> add_7_to_fp16 = const()[name = tensor<string, []>("add_7_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(17769536)))];
+            tensor<fp16, [1024, 640]> concat_26_to_fp16 = const()[name = tensor<string, []>("concat_26_to_fp16"), val = tensor<fp16, [1024, 640]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(17771648)))];
+            tensor<fp16, [1024, 256]> concat_27_to_fp16 = const()[name = tensor<string, []>("concat_27_to_fp16"), val = tensor<fp16, [1024, 256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(19082432)))];
+            tensor<fp16, [1024]> add_8_to_fp16 = const()[name = tensor<string, []>("add_8_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(19606784)))];
+            tensor<fp16, [?, 1, 640]> transpose_3_cast_fp16 = transpose(perm = transpose_3_perm_0, x = x_37_cast_fp16)[name = tensor<string, []>("transpose_22")];
+            tensor<fp16, [?, 1, 512]> input_35_batch_first_cast_fp16_0, tensor<fp16, [1, 512]> input_35_batch_first_cast_fp16_1, tensor<fp16, [1, 512]> input_35_batch_first_cast_fp16_2 = lstm(activation = input_35_batch_first_activation_0, bias = add_7_to_fp16, bias_back = add_8_to_fp16, cell_activation = input_35_batch_first_cell_activation_0, direction = input_35_batch_first_direction_0, initial_c = x_17_batch_first_lstm_h0_reshaped_to_fp16, initial_h = x_17_batch_first_lstm_h0_reshaped_to_fp16, output_sequence = input_35_batch_first_output_sequence_0, recurrent_activation = input_35_batch_first_recurrent_activation_0, weight_hh = concat_25_to_fp16, weight_hh_back = concat_27_to_fp16, weight_ih = concat_24_to_fp16, weight_ih_back = concat_26_to_fp16, x = transpose_3_cast_fp16)[name = tensor<string, []>("input_35_batch_first_cast_fp16")];
+            tensor<int32, [3]> transpose_8_perm_0 = const()[name = tensor<string, []>("transpose_8_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
+            tensor<fp16, [1024, 128]> duration_encoder_norms_1_fc_weight_to_fp16 = const()[name = tensor<string, []>("duration_encoder_norms_1_fc_weight_to_fp16"), val = tensor<fp16, [1024, 128]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(19608896)))];
+            tensor<fp16, [1024]> duration_encoder_norms_1_fc_bias_to_fp16 = const()[name = tensor<string, []>("duration_encoder_norms_1_fc_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(19871104)))];
+            tensor<fp16, [1, 1024]> linear_2_cast_fp16 = linear(bias = duration_encoder_norms_1_fc_bias_to_fp16, weight = duration_encoder_norms_1_fc_weight_to_fp16, x = style_cast_fp16)[name = tensor<string, []>("linear_2_cast_fp16")];
+            tensor<int32, [3]> var_247 = const()[name = tensor<string, []>("op_247"), val = tensor<int32, [3]>([1, 1024, 1])];
+            tensor<fp16, [1, 1024, 1]> h_7_cast_fp16 = reshape(shape = var_247, x = linear_2_cast_fp16)[name = tensor<string, []>("h_7_cast_fp16")];
+            tensor<int32, [2]> var_249_split_sizes_0 = const()[name = tensor<string, []>("op_249_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_249_axis_0 = const()[name = tensor<string, []>("op_249_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 1]> var_249_cast_fp16_0, tensor<fp16, [1, 512, 1]> var_249_cast_fp16_1 = split(axis = var_249_axis_0, split_sizes = var_249_split_sizes_0, x = h_7_cast_fp16)[name = tensor<string, []>("op_249_cast_fp16")];
+            tensor<int32, [3]> gamma_7_perm_0 = const()[name = tensor<string, []>("gamma_7_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [3]> beta_7_perm_0 = const()[name = tensor<string, []>("beta_7_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [1]> x_47_axes_0 = const()[name = tensor<string, []>("x_47_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, [1, ?, 512]> transpose_8_cast_fp16 = transpose(perm = transpose_8_perm_0, x = input_35_batch_first_cast_fp16_0)[name = tensor<string, []>("transpose_21")];
+            tensor<fp16, [1, ?, 512]> x_47_cast_fp16 = layer_norm(axes = x_47_axes_0, epsilon = var_136_to_fp16, x = transpose_8_cast_fp16)[name = tensor<string, []>("x_47_cast_fp16")];
+            tensor<fp16, []> var_255_promoted_to_fp16 = const()[name = tensor<string, []>("op_255_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 1, 512]> gamma_7_cast_fp16 = transpose(perm = gamma_7_perm_0, x = var_249_cast_fp16_0)[name = tensor<string, []>("transpose_20")];
+            tensor<fp16, [1, 1, 512]> var_256_cast_fp16 = add(x = gamma_7_cast_fp16, y = var_255_promoted_to_fp16)[name = tensor<string, []>("op_256_cast_fp16")];
+            tensor<fp16, [1, ?, 512]> var_257_cast_fp16 = mul(x = var_256_cast_fp16, y = x_47_cast_fp16)[name = tensor<string, []>("op_257_cast_fp16")];
+            tensor<fp16, [1, 1, 512]> beta_7_cast_fp16 = transpose(perm = beta_7_perm_0, x = var_249_cast_fp16_1)[name = tensor<string, []>("transpose_19")];
+            tensor<fp16, [1, ?, 512]> x_49_cast_fp16 = add(x = var_257_cast_fp16, y = beta_7_cast_fp16)[name = tensor<string, []>("x_49_cast_fp16")];
+            tensor<int32, [3]> x_51_perm_0 = const()[name = tensor<string, []>("x_51_perm_0"), val = tensor<int32, [3]>([0, -1, -2])];
+            tensor<bool, []> x_53_interleave_0 = const()[name = tensor<string, []>("x_53_interleave_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, 512, ?]> x_51_cast_fp16 = transpose(perm = x_51_perm_0, x = x_49_cast_fp16)[name = tensor<string, []>("transpose_18")];
+            tensor<fp16, [1, 640, ?]> x_53_cast_fp16 = concat(axis = var_147, interleave = x_53_interleave_0, values = (x_51_cast_fp16, s_chan_cast_fp16))[name = tensor<string, []>("x_53_cast_fp16")];
+            tensor<int32, [3]> transpose_5_perm_0 = const()[name = tensor<string, []>("transpose_5_perm_0"), val = tensor<int32, [3]>([-1, 0, -2])];
+            tensor<string, []> input_41_batch_first_direction_0 = const()[name = tensor<string, []>("input_41_batch_first_direction_0"), val = tensor<string, []>("bidirectional")];
+            tensor<bool, []> input_41_batch_first_output_sequence_0 = const()[name = tensor<string, []>("input_41_batch_first_output_sequence_0"), val = tensor<bool, []>(true)];
+            tensor<string, []> input_41_batch_first_recurrent_activation_0 = const()[name = tensor<string, []>("input_41_batch_first_recurrent_activation_0"), val = tensor<string, []>("sigmoid")];
+            tensor<string, []> input_41_batch_first_cell_activation_0 = const()[name = tensor<string, []>("input_41_batch_first_cell_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<string, []> input_41_batch_first_activation_0 = const()[name = tensor<string, []>("input_41_batch_first_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<fp16, [1024, 640]> concat_34_to_fp16 = const()[name = tensor<string, []>("concat_34_to_fp16"), val = tensor<fp16, [1024, 640]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(19873216)))];
+            tensor<fp16, [1024, 256]> concat_35_to_fp16 = const()[name = tensor<string, []>("concat_35_to_fp16"), val = tensor<fp16, [1024, 256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(21184000)))];
+            tensor<fp16, [1024]> add_10_to_fp16 = const()[name = tensor<string, []>("add_10_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(21708352)))];
+            tensor<fp16, [1024, 640]> concat_36_to_fp16 = const()[name = tensor<string, []>("concat_36_to_fp16"), val = tensor<fp16, [1024, 640]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(21710464)))];
+            tensor<fp16, [1024, 256]> concat_37_to_fp16 = const()[name = tensor<string, []>("concat_37_to_fp16"), val = tensor<fp16, [1024, 256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(23021248)))];
+            tensor<fp16, [1024]> add_11_to_fp16 = const()[name = tensor<string, []>("add_11_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(23545600)))];
+            tensor<fp16, [?, 1, 640]> transpose_5_cast_fp16 = transpose(perm = transpose_5_perm_0, x = x_53_cast_fp16)[name = tensor<string, []>("transpose_17")];
+            tensor<fp16, [?, 1, 512]> input_41_batch_first_cast_fp16_0, tensor<fp16, [1, 512]> input_41_batch_first_cast_fp16_1, tensor<fp16, [1, 512]> input_41_batch_first_cast_fp16_2 = lstm(activation = input_41_batch_first_activation_0, bias = add_10_to_fp16, bias_back = add_11_to_fp16, cell_activation = input_41_batch_first_cell_activation_0, direction = input_41_batch_first_direction_0, initial_c = x_17_batch_first_lstm_h0_reshaped_to_fp16, initial_h = x_17_batch_first_lstm_h0_reshaped_to_fp16, output_sequence = input_41_batch_first_output_sequence_0, recurrent_activation = input_41_batch_first_recurrent_activation_0, weight_hh = concat_35_to_fp16, weight_hh_back = concat_37_to_fp16, weight_ih = concat_34_to_fp16, weight_ih_back = concat_36_to_fp16, x = transpose_5_cast_fp16)[name = tensor<string, []>("input_41_batch_first_cast_fp16")];
+            tensor<int32, [3]> transpose_9_perm_0 = const()[name = tensor<string, []>("transpose_9_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
+            tensor<fp16, [1024, 128]> duration_encoder_norms_2_fc_weight_to_fp16 = const()[name = tensor<string, []>("duration_encoder_norms_2_fc_weight_to_fp16"), val = tensor<fp16, [1024, 128]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(23547712)))];
+            tensor<fp16, [1024]> duration_encoder_norms_2_fc_bias_to_fp16 = const()[name = tensor<string, []>("duration_encoder_norms_2_fc_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(23809920)))];
+            tensor<fp16, [1, 1024]> linear_3_cast_fp16 = linear(bias = duration_encoder_norms_2_fc_bias_to_fp16, weight = duration_encoder_norms_2_fc_weight_to_fp16, x = style_cast_fp16)[name = tensor<string, []>("linear_3_cast_fp16")];
+            tensor<int32, [3]> var_294 = const()[name = tensor<string, []>("op_294"), val = tensor<int32, [3]>([1, 1024, 1])];
+            tensor<fp16, [1, 1024, 1]> h_cast_fp16 = reshape(shape = var_294, x = linear_3_cast_fp16)[name = tensor<string, []>("h_cast_fp16")];
+            tensor<int32, [2]> var_296_split_sizes_0 = const()[name = tensor<string, []>("op_296_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_296_axis_0 = const()[name = tensor<string, []>("op_296_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 1]> var_296_cast_fp16_0, tensor<fp16, [1, 512, 1]> var_296_cast_fp16_1 = split(axis = var_296_axis_0, split_sizes = var_296_split_sizes_0, x = h_cast_fp16)[name = tensor<string, []>("op_296_cast_fp16")];
+            tensor<int32, [3]> gamma_11_perm_0 = const()[name = tensor<string, []>("gamma_11_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [3]> beta_11_perm_0 = const()[name = tensor<string, []>("beta_11_perm_0"), val = tensor<int32, [3]>([0, -1, 1])];
+            tensor<int32, [1]> x_63_axes_0 = const()[name = tensor<string, []>("x_63_axes_0"), val = tensor<int32, [1]>([-1])];
+            tensor<fp16, [1, ?, 512]> transpose_9_cast_fp16 = transpose(perm = transpose_9_perm_0, x = input_41_batch_first_cast_fp16_0)[name = tensor<string, []>("transpose_16")];
+            tensor<fp16, [1, ?, 512]> x_63_cast_fp16 = layer_norm(axes = x_63_axes_0, epsilon = var_136_to_fp16, x = transpose_9_cast_fp16)[name = tensor<string, []>("x_63_cast_fp16")];
+            tensor<fp16, []> var_302_promoted_to_fp16 = const()[name = tensor<string, []>("op_302_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 1, 512]> gamma_11_cast_fp16 = transpose(perm = gamma_11_perm_0, x = var_296_cast_fp16_0)[name = tensor<string, []>("transpose_15")];
+            tensor<fp16, [1, 1, 512]> var_303_cast_fp16 = add(x = gamma_11_cast_fp16, y = var_302_promoted_to_fp16)[name = tensor<string, []>("op_303_cast_fp16")];
+            tensor<fp16, [1, ?, 512]> var_304_cast_fp16 = mul(x = var_303_cast_fp16, y = x_63_cast_fp16)[name = tensor<string, []>("op_304_cast_fp16")];
+            tensor<fp16, [1, 1, 512]> beta_11_cast_fp16 = transpose(perm = beta_11_perm_0, x = var_296_cast_fp16_1)[name = tensor<string, []>("transpose_14")];
+            tensor<fp16, [1, ?, 512]> x_65_cast_fp16 = add(x = var_304_cast_fp16, y = beta_11_cast_fp16)[name = tensor<string, []>("x_65_cast_fp16")];
+            tensor<int32, [3]> x_67_perm_0 = const()[name = tensor<string, []>("x_67_perm_0"), val = tensor<int32, [3]>([0, -1, -2])];
+            tensor<bool, []> x_interleave_0 = const()[name = tensor<string, []>("x_interleave_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, 512, ?]> x_67_cast_fp16 = transpose(perm = x_67_perm_0, x = x_65_cast_fp16)[name = tensor<string, []>("transpose_13")];
+            tensor<fp16, [1, 640, ?]> x_cast_fp16 = concat(axis = var_147, interleave = x_interleave_0, values = (x_67_cast_fp16, s_chan_cast_fp16))[name = tensor<string, []>("x_cast_fp16")];
+            tensor<int32, [3]> input_45_perm_0 = const()[name = tensor<string, []>("input_45_perm_0"), val = tensor<int32, [3]>([0, -1, -2])];
+            tensor<int32, [3]> input_45_batch_first_transpose_perm_0 = const()[name = tensor<string, []>("input_45_batch_first_transpose_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
+            tensor<string, []> input_47_batch_first_direction_0 = const()[name = tensor<string, []>("input_47_batch_first_direction_0"), val = tensor<string, []>("bidirectional")];
+            tensor<bool, []> input_47_batch_first_output_sequence_0 = const()[name = tensor<string, []>("input_47_batch_first_output_sequence_0"), val = tensor<bool, []>(true)];
+            tensor<string, []> input_47_batch_first_recurrent_activation_0 = const()[name = tensor<string, []>("input_47_batch_first_recurrent_activation_0"), val = tensor<string, []>("sigmoid")];
+            tensor<string, []> input_47_batch_first_cell_activation_0 = const()[name = tensor<string, []>("input_47_batch_first_cell_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<string, []> input_47_batch_first_activation_0 = const()[name = tensor<string, []>("input_47_batch_first_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<fp16, [1024, 640]> concat_44_to_fp16 = const()[name = tensor<string, []>("concat_44_to_fp16"), val = tensor<fp16, [1024, 640]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(23812032)))];
+            tensor<fp16, [1024, 256]> concat_45_to_fp16 = const()[name = tensor<string, []>("concat_45_to_fp16"), val = tensor<fp16, [1024, 256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(25122816)))];
+            tensor<fp16, [1024]> add_13_to_fp16 = const()[name = tensor<string, []>("add_13_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(25647168)))];
+            tensor<fp16, [1024, 640]> concat_46_to_fp16 = const()[name = tensor<string, []>("concat_46_to_fp16"), val = tensor<fp16, [1024, 640]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(25649280)))];
+            tensor<fp16, [1024, 256]> concat_47_to_fp16 = const()[name = tensor<string, []>("concat_47_to_fp16"), val = tensor<fp16, [1024, 256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(26960064)))];
+            tensor<fp16, [1024]> add_14_to_fp16 = const()[name = tensor<string, []>("add_14_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(27484416)))];
+            tensor<fp16, [1, ?, 640]> d = transpose(perm = input_45_perm_0, x = x_cast_fp16)[name = tensor<string, []>("transpose_12")];
+            tensor<fp16, [?, 1, 640]> input_45_batch_first_transpose_cast_fp16 = transpose(perm = input_45_batch_first_transpose_perm_0, x = d)[name = tensor<string, []>("transpose_11")];
+            tensor<fp16, [?, 1, 512]> input_47_batch_first_cast_fp16_0, tensor<fp16, [1, 512]> input_47_batch_first_cast_fp16_1, tensor<fp16, [1, 512]> input_47_batch_first_cast_fp16_2 = lstm(activation = input_47_batch_first_activation_0, bias = add_13_to_fp16, bias_back = add_14_to_fp16, cell_activation = input_47_batch_first_cell_activation_0, direction = input_47_batch_first_direction_0, initial_c = x_17_batch_first_lstm_h0_reshaped_to_fp16, initial_h = x_17_batch_first_lstm_h0_reshaped_to_fp16, output_sequence = input_47_batch_first_output_sequence_0, recurrent_activation = input_47_batch_first_recurrent_activation_0, weight_hh = concat_45_to_fp16, weight_hh_back = concat_47_to_fp16, weight_ih = concat_44_to_fp16, weight_ih_back = concat_46_to_fp16, x = input_45_batch_first_transpose_cast_fp16)[name = tensor<string, []>("input_47_batch_first_cast_fp16")];
+            tensor<int32, [3]> input_47_perm_0 = const()[name = tensor<string, []>("input_47_perm_0"), val = tensor<int32, [3]>([1, 0, 2])];
+            tensor<fp16, [50, 512]> duration_proj_linear_layer_weight_to_fp16 = const()[name = tensor<string, []>("duration_proj_linear_layer_weight_to_fp16"), val = tensor<fp16, [50, 512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(27486528)))];
+            tensor<fp16, [50]> duration_proj_linear_layer_bias_to_fp16 = const()[name = tensor<string, []>("duration_proj_linear_layer_bias_to_fp16"), val = tensor<fp16, [50]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(27537792)))];
+            tensor<fp16, [1, ?, 512]> input_47_cast_fp16 = transpose(perm = input_47_perm_0, x = input_47_batch_first_cast_fp16_0)[name = tensor<string, []>("transpose_10")];
+            tensor<fp16, [1, ?, 50]> pred_dur_log = linear(bias = duration_proj_linear_layer_bias_to_fp16, weight = duration_proj_linear_layer_weight_to_fp16, x = input_47_cast_fp16)[name = tensor<string, []>("linear_4_cast_fp16")];
+            tensor<int32, []> var_345 = const()[name = tensor<string, []>("op_345"), val = tensor<int32, []>(-1)];
+            tensor<int32, [3]> var_355_begin_0 = const()[name = tensor<string, []>("op_355_begin_0"), val = tensor<int32, [3]>([0, 0, 0])];
+            tensor<int32, [3]> var_355_end_0 = const()[name = tensor<string, []>("op_355_end_0"), val = tensor<int32, [3]>([1, 0, 1])];
+            tensor<bool, [3]> var_355_end_mask_0 = const()[name = tensor<string, []>("op_355_end_mask_0"), val = tensor<bool, [3]>([true, true, false])];
+            tensor<bool, [3]> var_355_squeeze_mask_0 = const()[name = tensor<string, []>("op_355_squeeze_mask_0"), val = tensor<bool, [3]>([false, false, true])];
+            tensor<fp16, [1, ?]> var_355_cast_fp16 = slice_by_index(begin = var_355_begin_0, end = var_355_end_0, end_mask = var_355_end_mask_0, squeeze_mask = var_355_squeeze_mask_0, x = bert_dur)[name = tensor<string, []>("op_355_cast_fp16")];
+            tensor<fp16, []> fill_like_0_value_0_to_fp16 = const()[name = tensor<string, []>("fill_like_0_value_0_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, ?]> fill_like_0_cast_fp16 = fill_like(ref_tensor = var_355_cast_fp16, value = fill_like_0_value_0_to_fp16)[name = tensor<string, []>("fill_like_0_cast_fp16")];
+            tensor<bool, []> var_357_exclusive_0 = const()[name = tensor<string, []>("op_357_exclusive_0"), val = tensor<bool, []>(false)];
+            tensor<bool, []> var_357_reverse_0 = const()[name = tensor<string, []>("op_357_reverse_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [1, ?]> var_357_cast_fp16 = cumsum(axis = var_345, exclusive = var_357_exclusive_0, reverse = var_357_reverse_0, x = fill_like_0_cast_fp16)[name = tensor<string, []>("op_357_cast_fp16")];
+            tensor<string, []> cast_6_dtype_0 = const()[name = tensor<string, []>("cast_6_dtype_0"), val = tensor<string, []>("int32")];
+            tensor<int32, []> var_359 = const()[name = tensor<string, []>("op_359"), val = tensor<int32, []>(1)];
+            tensor<int32, [1, ?]> var_357_cast_fp16_to_int32 = cast(dtype = cast_6_dtype_0, x = var_357_cast_fp16)[name = tensor<string, []>("cast_13")];
+            tensor<int32, [1, ?]> indices = sub(x = var_357_cast_fp16_to_int32, y = var_359)[name = tensor<string, []>("indices")];
+            tensor<int32, [1]> input_axes_0 = const()[name = tensor<string, []>("input_axes_0"), val = tensor<int32, [1]>([0])];
+            tensor<string, []> indices_to_uint16_dtype_0 = const()[name = tensor<string, []>("indices_to_uint16_dtype_0"), val = tensor<string, []>("uint16")];
+            tensor<uint16, [1, ?]> indices_to_uint16 = cast(dtype = indices_to_uint16_dtype_0, x = indices)[name = tensor<string, []>("cast_12")];
+            tensor<uint16, [?]> input_cast_uint16 = squeeze(axes = input_axes_0, x = indices_to_uint16)[name = tensor<string, []>("input_cast_uint16")];
+            tensor<int32, []> fe_axis_0 = const()[name = tensor<string, []>("fe_axis_0"), val = tensor<int32, []>(0)];
+            tensor<int32, []> fe_batch_dims_0 = const()[name = tensor<string, []>("fe_batch_dims_0"), val = tensor<int32, []>(0)];
+            tensor<bool, []> fe_validate_indices_0 = const()[name = tensor<string, []>("fe_validate_indices_0"), val = tensor<bool, []>(false)];
+            tensor<fp16, [512, 768]> fixed_embedding_embedding_weight_to_fp16 = const()[name = tensor<string, []>("fixed_embedding_embedding_weight_to_fp16"), val = tensor<fp16, [512, 768]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(27537984)))];
+            tensor<fp16, [?, 768]> fe_cast_fp16_cast_uint16 = gather(axis = fe_axis_0, batch_dims = fe_batch_dims_0, indices = input_cast_uint16, validate_indices = fe_validate_indices_0, x = fixed_embedding_embedding_weight_to_fp16)[name = tensor<string, []>("fe_cast_fp16_cast_uint16")];
+            tensor<int32, [1]> var_364_axes_0 = const()[name = tensor<string, []>("op_364_axes_0"), val = tensor<int32, [1]>([0])];
+            tensor<fp16, [1, ?, 768]> fixed_embedding = expand_dims(axes = var_364_axes_0, x = fe_cast_fp16_cast_uint16)[name = tensor<string, []>("op_364_cast_fp16")];
+            tensor<fp16, [1, 512, ?]> t_en = transpose(perm = transpose_0_perm_0, x = x_17_batch_first_cast_fp16_0)[name = tensor<string, []>("transpose_29")];
+        } -> (t_en, d, pred_dur_log, fixed_embedding);
+}

ANE/styletts2_ane_postbert.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bffc0f40bfeee94d93537de68e3057f8e18e847c2197bef2045118d25e950a0d
+size 28324480

ANE/styletts2_ane_postbert.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:982078ab6840241c1b6717dab503af3f123373606fd6b31967f33845468aa0fd
+size 43623

ANE/styletts2_ane_postbert.mlpackage/Data/com.apple.CoreML/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bffc0f40bfeee94d93537de68e3057f8e18e847c2197bef2045118d25e950a0d
+size 28324480

ANE/styletts2_ane_postbert.mlpackage/Manifest.json

@@ -0,0 +1,18 @@
+{
+    "fileFormatVersion": "1.0.0",
+    "itemInfoEntries": {
+        "4577D915-6FB9-4738-8BE7-3DAA463EE8BD": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Specification",
+            "name": "model.mlmodel",
+            "path": "com.apple.CoreML/model.mlmodel"
+        },
+        "8CBB3422-373A-42D0-8E4F-AF8A2A241439": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Weights",
+            "name": "weights",
+            "path": "com.apple.CoreML/weights"
+        }
+    },
+    "rootModelIdentifier": "4577D915-6FB9-4738-8BE7-3DAA463EE8BD"
+}

ANE/styletts2_ane_prosody.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a569de5bd6b2a4cb48a7ec8c1544f80e7c6484933bb9160dc7c88b462c23c9cc
+size 243

ANE/styletts2_ane_prosody.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97801e9d7ade59ab0c78110290f93222ad0430e0c479125ef09755433bff3ac7
+size 417

ANE/styletts2_ane_prosody.mlmodelc/metadata.json

@@ -0,0 +1,97 @@
+[
+  {
+    "metadataOutputVersion" : "3.0",
+    "shortDescription" : "StyleTTS2-ANE Stage 5: Prosody F0Ntrain (static T_a=2000) — fp16, ANE.",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 1 × 4000)",
+        "shortDescription" : "",
+        "shape" : "[1, 4000]",
+        "name" : "F0",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 1 × 4000)",
+        "shortDescription" : "",
+        "shape" : "[1, 4000]",
+        "name" : "N",
+        "type" : "MultiArray"
+      }
+    ],
+    "storagePrecision" : "Mixed (Float16, Palettized (8 bits))",
+    "modelParameters" : [
+
+    ],
+    "specificationVersion" : 8,
+    "mlProgramOperationTypeHistogram" : {
+      "Ios17.squeeze" : 4,
+      "Ios17.mul" : 18,
+      "Ios17.linear" : 12,
+      "Ios17.transpose" : 2,
+      "Ios16.constexprLutToDense" : 30,
+      "Ios17.conv" : 16,
+      "Ios17.leakyRelu" : 12,
+      "Ios17.add" : 30,
+      "Ios17.lstm" : 1,
+      "Ios17.convTranspose" : 2,
+      "Ios17.sliceByIndex" : 2,
+      "UpsampleNearestNeighbor" : 2,
+      "Ios17.expandDims" : 2,
+      "Ios17.instanceNorm" : 11,
+      "Ios17.reshape" : 12,
+      "Ios17.cast" : 2,
+      "Split" : 12
+    },
+    "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"
+    },
+    "inputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 640 × 2000)",
+        "shortDescription" : "",
+        "shape" : "[1, 640, 2000]",
+        "name" : "en",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 128)",
+        "shortDescription" : "",
+        "shape" : "[1, 128]",
+        "name" : "s",
+        "type" : "MultiArray"
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.version" : "8.3.0",
+      "com.github.apple.coremltools.source" : "torch==2.11.0",
+      "com.github.apple.coremltools.source_dialect" : "TorchScript"
+    },
+    "generatedClassName" : "styletts2_ane_prosody",
+    "method" : "predict"
+  }
+]

ANE/styletts2_ane_prosody.mlmodelc/model.mil

@@ -0,0 +1,396 @@
+program(1.0)
+[buildInfo = dict<tensor<string, []>, tensor<string, []>>({{"coremlc-component-MIL", "3520.4.1"}, {"coremlc-version", "3520.5.1"}})]
+{
+    func main<ios17>(tensor<fp32, [1, 640, 2000]> en, tensor<fp32, [1, 128]> s) {
+            tensor<int32, [3]> transpose_0_perm_0 = const()[name = tensor<string, []>("transpose_0_perm_0"), val = tensor<int32, [3]>([-1, 0, -2])];
+            tensor<string, []> en_to_fp16_dtype_0 = const()[name = tensor<string, []>("en_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
+            tensor<string, []> x_batch_first_direction_0 = const()[name = tensor<string, []>("x_batch_first_direction_0"), val = tensor<string, []>("bidirectional")];
+            tensor<bool, []> x_batch_first_output_sequence_0 = const()[name = tensor<string, []>("x_batch_first_output_sequence_0"), val = tensor<bool, []>(true)];
+            tensor<string, []> x_batch_first_recurrent_activation_0 = const()[name = tensor<string, []>("x_batch_first_recurrent_activation_0"), val = tensor<string, []>("sigmoid")];
+            tensor<string, []> x_batch_first_cell_activation_0 = const()[name = tensor<string, []>("x_batch_first_cell_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<string, []> x_batch_first_activation_0 = const()[name = tensor<string, []>("x_batch_first_activation_0"), val = tensor<string, []>("tanh")];
+            tensor<fp16, [1, 512]> x_batch_first_lstm_h0_reshaped_to_fp16 = const()[name = tensor<string, []>("x_batch_first_lstm_h0_reshaped_to_fp16"), val = tensor<fp16, [1, 512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(64)))];
+            tensor<fp16, [1024, 640]> concat_4_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [655360]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1152))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(656576))), name = tensor<string, []>("concat_4_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 640])];
+            tensor<fp16, [1024, 256]> concat_5_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [262144]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(657152))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(919360))), name = tensor<string, []>("concat_5_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 256])];
+            tensor<fp16, [1024]> add_0_to_fp16 = const()[name = tensor<string, []>("add_0_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(919936)))];
+            tensor<fp16, [1024, 640]> concat_6_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [655360]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(922048))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1577472))), name = tensor<string, []>("concat_6_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 640])];
+            tensor<fp16, [1024, 256]> concat_7_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [262144]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1578048))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1840256))), name = tensor<string, []>("concat_7_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 256])];
+            tensor<fp16, [1024]> add_1_to_fp16 = const()[name = tensor<string, []>("add_1_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1840832)))];
+            tensor<fp16, [1, 640, 2000]> en_to_fp16 = cast(dtype = en_to_fp16_dtype_0, x = en)[name = tensor<string, []>("cast_1")];
+            tensor<fp16, [2000, 1, 640]> transpose_0_cast_fp16 = transpose(perm = transpose_0_perm_0, x = en_to_fp16)[name = tensor<string, []>("transpose_2")];
+            tensor<fp16, [2000, 1, 512]> x_batch_first_cast_fp16_0, tensor<fp16, [1, 512]> x_batch_first_cast_fp16_1, tensor<fp16, [1, 512]> x_batch_first_cast_fp16_2 = lstm(activation = x_batch_first_activation_0, bias = add_0_to_fp16, bias_back = add_1_to_fp16, cell_activation = x_batch_first_cell_activation_0, direction = x_batch_first_direction_0, initial_c = x_batch_first_lstm_h0_reshaped_to_fp16, initial_h = x_batch_first_lstm_h0_reshaped_to_fp16, output_sequence = x_batch_first_output_sequence_0, recurrent_activation = x_batch_first_recurrent_activation_0, weight_hh = concat_5_to_fp16_palettized, weight_hh_back = concat_7_to_fp16_palettized, weight_ih = concat_4_to_fp16_palettized, weight_ih_back = concat_6_to_fp16_palettized, x = transpose_0_cast_fp16)[name = tensor<string, []>("x_batch_first_cast_fp16")];
+            tensor<fp32, []> var_53 = const()[name = tensor<string, []>("op_53"), val = tensor<fp32, []>(0x1.99999ap-3)];
+            tensor<string, []> s_to_fp16_dtype_0 = const()[name = tensor<string, []>("s_to_fp16_dtype_0"), val = tensor<string, []>("fp16")];
+            tensor<fp16, [1024, 128]> F0_0_norm1_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [131072]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1842944))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1974080))), name = tensor<string, []>("F0_0_norm1_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 128])];
+            tensor<fp16, [1024]> F0_0_norm1_fc_bias_to_fp16 = const()[name = tensor<string, []>("F0_0_norm1_fc_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1974656)))];
+            tensor<fp16, [1, 128]> s_to_fp16 = cast(dtype = s_to_fp16_dtype_0, x = s)[name = tensor<string, []>("cast_0")];
+            tensor<fp16, [1, 1024]> linear_0_cast_fp16 = linear(bias = F0_0_norm1_fc_bias_to_fp16, weight = F0_0_norm1_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_0_cast_fp16")];
+            tensor<int32, [3]> var_78 = const()[name = tensor<string, []>("op_78"), val = tensor<int32, [3]>([1, 1024, 1])];
+            tensor<fp16, [1, 1024, 1]> h_3_cast_fp16 = reshape(shape = var_78, x = linear_0_cast_fp16)[name = tensor<string, []>("h_3_cast_fp16")];
+            tensor<int32, [2]> var_80_split_sizes_0 = const()[name = tensor<string, []>("op_80_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_80_axis_0 = const()[name = tensor<string, []>("op_80_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 1]> var_80_cast_fp16_0, tensor<fp16, [1, 512, 1]> var_80_cast_fp16_1 = split(axis = var_80_axis_0, split_sizes = var_80_split_sizes_0, x = h_3_cast_fp16)[name = tensor<string, []>("op_80_cast_fp16")];
+            tensor<fp16, []> var_82_promoted_to_fp16 = const()[name = tensor<string, []>("op_82_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 512, 1]> var_83_cast_fp16 = add(x = var_80_cast_fp16_0, y = var_82_promoted_to_fp16)[name = tensor<string, []>("op_83_cast_fp16")];
+            tensor<fp16, []> var_56_to_fp16 = const()[name = tensor<string, []>("op_56_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<int32, [3]> transpose_0_perm_0_1 = const()[name = tensor<string, []>("transpose_0_perm_0_1"), val = tensor<int32, [3]>([1, 2, 0])];
+            tensor<fp16, [1, 512, 2000]> transpose_0 = transpose(perm = transpose_0_perm_0_1, x = x_batch_first_cast_fp16_0)[name = tensor<string, []>("transpose_1")];
+            tensor<fp16, [1, 512, 2000]> var_84_cast_fp16 = instance_norm(epsilon = var_56_to_fp16, x = transpose_0)[name = tensor<string, []>("op_84_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> var_85_cast_fp16 = mul(x = var_83_cast_fp16, y = var_84_cast_fp16)[name = tensor<string, []>("op_85_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_5_cast_fp16 = add(x = var_85_cast_fp16, y = var_80_cast_fp16_1)[name = tensor<string, []>("input_5_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_7_cast_fp16 = leaky_relu(alpha = var_53, x = input_5_cast_fp16)[name = tensor<string, []>("input_7_cast_fp16")];
+            tensor<string, []> input_11_pad_type_0 = const()[name = tensor<string, []>("input_11_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> input_11_pad_0 = const()[name = tensor<string, []>("input_11_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> input_11_strides_0 = const()[name = tensor<string, []>("input_11_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> input_11_dilations_0 = const()[name = tensor<string, []>("input_11_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> input_11_groups_0 = const()[name = tensor<string, []>("input_11_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [512, 512, 3]> weight_1_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [786432]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(1976768))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2763264))), name = tensor<string, []>("weight_1_to_fp16_palettized"), shape = tensor<uint32, [3]>([512, 512, 3])];
+            tensor<fp16, [512]> F0_0_conv1_bias_to_fp16 = const()[name = tensor<string, []>("F0_0_conv1_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2763840)))];
+            tensor<fp16, [1, 512, 2000]> input_11_cast_fp16 = conv(bias = F0_0_conv1_bias_to_fp16, dilations = input_11_dilations_0, groups = input_11_groups_0, pad = input_11_pad_0, pad_type = input_11_pad_type_0, strides = input_11_strides_0, weight = weight_1_to_fp16_palettized, x = input_7_cast_fp16)[name = tensor<string, []>("input_11_cast_fp16")];
+            tensor<fp16, [1024, 128]> F0_0_norm2_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [131072]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2764928))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2896064))), name = tensor<string, []>("F0_0_norm2_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 128])];
+            tensor<fp16, [1024]> F0_0_norm2_fc_bias_to_fp16 = const()[name = tensor<string, []>("F0_0_norm2_fc_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2896640)))];
+            tensor<fp16, [1, 1024]> linear_1_cast_fp16 = linear(bias = F0_0_norm2_fc_bias_to_fp16, weight = F0_0_norm2_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_1_cast_fp16")];
+            tensor<int32, [3]> var_102 = const()[name = tensor<string, []>("op_102"), val = tensor<int32, [3]>([1, 1024, 1])];
+            tensor<fp16, [1, 1024, 1]> h_7_cast_fp16 = reshape(shape = var_102, x = linear_1_cast_fp16)[name = tensor<string, []>("h_7_cast_fp16")];
+            tensor<int32, [2]> var_104_split_sizes_0 = const()[name = tensor<string, []>("op_104_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_104_axis_0 = const()[name = tensor<string, []>("op_104_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 1]> var_104_cast_fp16_0, tensor<fp16, [1, 512, 1]> var_104_cast_fp16_1 = split(axis = var_104_axis_0, split_sizes = var_104_split_sizes_0, x = h_7_cast_fp16)[name = tensor<string, []>("op_104_cast_fp16")];
+            tensor<fp16, []> var_106_promoted_to_fp16 = const()[name = tensor<string, []>("op_106_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 512, 1]> var_107_cast_fp16 = add(x = var_104_cast_fp16_0, y = var_106_promoted_to_fp16)[name = tensor<string, []>("op_107_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> var_108_cast_fp16 = instance_norm(epsilon = var_56_to_fp16, x = input_11_cast_fp16)[name = tensor<string, []>("op_108_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> var_109_cast_fp16 = mul(x = var_107_cast_fp16, y = var_108_cast_fp16)[name = tensor<string, []>("op_109_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_13_cast_fp16 = add(x = var_109_cast_fp16, y = var_104_cast_fp16_1)[name = tensor<string, []>("input_13_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_15_cast_fp16 = leaky_relu(alpha = var_53, x = input_13_cast_fp16)[name = tensor<string, []>("input_15_cast_fp16")];
+            tensor<string, []> out_1_pad_type_0 = const()[name = tensor<string, []>("out_1_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> out_1_pad_0 = const()[name = tensor<string, []>("out_1_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> out_1_strides_0 = const()[name = tensor<string, []>("out_1_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> out_1_dilations_0 = const()[name = tensor<string, []>("out_1_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> out_1_groups_0 = const()[name = tensor<string, []>("out_1_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [512, 512, 3]> weight_3_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [786432]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(2898752))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3685248))), name = tensor<string, []>("weight_3_to_fp16_palettized"), shape = tensor<uint32, [3]>([512, 512, 3])];
+            tensor<fp16, [512]> F0_0_conv2_bias_to_fp16 = const()[name = tensor<string, []>("F0_0_conv2_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3685824)))];
+            tensor<fp16, [1, 512, 2000]> out_1_cast_fp16 = conv(bias = F0_0_conv2_bias_to_fp16, dilations = out_1_dilations_0, groups = out_1_groups_0, pad = out_1_pad_0, pad_type = out_1_pad_type_0, strides = out_1_strides_0, weight = weight_3_to_fp16_palettized, x = input_15_cast_fp16)[name = tensor<string, []>("out_1_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> var_120_cast_fp16 = add(x = out_1_cast_fp16, y = transpose_0)[name = tensor<string, []>("op_120_cast_fp16")];
+            tensor<fp16, []> _inversed_input_19_y_0_to_fp16 = const()[name = tensor<string, []>("_inversed_input_19_y_0_to_fp16"), val = tensor<fp16, []>(0x1.6ap-1)];
+            tensor<fp16, [1, 512, 2000]> _inversed_input_19_cast_fp16 = mul(x = var_120_cast_fp16, y = _inversed_input_19_y_0_to_fp16)[name = tensor<string, []>("_inversed_input_19_cast_fp16")];
+            tensor<fp32, []> var_126 = const()[name = tensor<string, []>("op_126"), val = tensor<fp32, []>(0x1.99999ap-3)];
+            tensor<fp16, [1024, 128]> F0_1_norm1_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [131072]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3686912))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3818048))), name = tensor<string, []>("F0_1_norm1_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 128])];
+            tensor<fp16, [1024]> F0_1_norm1_fc_bias_to_fp16 = const()[name = tensor<string, []>("F0_1_norm1_fc_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3818624)))];
+            tensor<fp16, [1, 1024]> linear_2_cast_fp16 = linear(bias = F0_1_norm1_fc_bias_to_fp16, weight = F0_1_norm1_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_2_cast_fp16")];
+            tensor<int32, [3]> var_161 = const()[name = tensor<string, []>("op_161"), val = tensor<int32, [3]>([1, 1024, 1])];
+            tensor<fp16, [1, 1024, 1]> h_11_cast_fp16 = reshape(shape = var_161, x = linear_2_cast_fp16)[name = tensor<string, []>("h_11_cast_fp16")];
+            tensor<int32, [2]> var_163_split_sizes_0 = const()[name = tensor<string, []>("op_163_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_163_axis_0 = const()[name = tensor<string, []>("op_163_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 1]> var_163_cast_fp16_0, tensor<fp16, [1, 512, 1]> var_163_cast_fp16_1 = split(axis = var_163_axis_0, split_sizes = var_163_split_sizes_0, x = h_11_cast_fp16)[name = tensor<string, []>("op_163_cast_fp16")];
+            tensor<fp16, []> var_165_promoted_to_fp16 = const()[name = tensor<string, []>("op_165_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 512, 1]> var_166_cast_fp16 = add(x = var_163_cast_fp16_0, y = var_165_promoted_to_fp16)[name = tensor<string, []>("op_166_cast_fp16")];
+            tensor<fp16, []> var_130_to_fp16 = const()[name = tensor<string, []>("op_130_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<fp16, [1, 512, 2000]> var_167_cast_fp16 = instance_norm(epsilon = var_130_to_fp16, x = _inversed_input_19_cast_fp16)[name = tensor<string, []>("op_167_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> var_168_cast_fp16 = mul(x = var_166_cast_fp16, y = var_167_cast_fp16)[name = tensor<string, []>("op_168_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_21_cast_fp16 = add(x = var_168_cast_fp16, y = var_163_cast_fp16_1)[name = tensor<string, []>("input_21_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_23_cast_fp16 = leaky_relu(alpha = var_126, x = input_21_cast_fp16)[name = tensor<string, []>("input_23_cast_fp16")];
+            tensor<string, []> conv_transpose_0_pad_type_0 = const()[name = tensor<string, []>("conv_transpose_0_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> conv_transpose_0_pad_0 = const()[name = tensor<string, []>("conv_transpose_0_pad_0"), val = tensor<int32, [2]>([0, 0])];
+            tensor<int32, [1]> conv_transpose_0_strides_0 = const()[name = tensor<string, []>("conv_transpose_0_strides_0"), val = tensor<int32, [1]>([2])];
+            tensor<int32, []> conv_transpose_0_groups_0 = const()[name = tensor<string, []>("conv_transpose_0_groups_0"), val = tensor<int32, []>(512)];
+            tensor<int32, [1]> conv_transpose_0_dilations_0 = const()[name = tensor<string, []>("conv_transpose_0_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [3]> conv_transpose_0_has_output_shape_output_shape_0 = const()[name = tensor<string, []>("conv_transpose_0_has_output_shape_output_shape_0"), val = tensor<int32, [3]>([1, 512, 4001])];
+            tensor<fp16, [512, 1, 3]> var_171_to_fp16 = const()[name = tensor<string, []>("op_171_to_fp16"), val = tensor<fp16, [512, 1, 3]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3820736)))];
+            tensor<fp16, [512]> F0_1_pool_bias_to_fp16 = const()[name = tensor<string, []>("F0_1_pool_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3823872)))];
+            tensor<fp16, [1, 512, 4001]> conv_transpose_0_has_output_shape_cast_fp16 = conv_transpose(bias = F0_1_pool_bias_to_fp16, dilations = conv_transpose_0_dilations_0, groups = conv_transpose_0_groups_0, output_shape = conv_transpose_0_has_output_shape_output_shape_0, pad = conv_transpose_0_pad_0, pad_type = conv_transpose_0_pad_type_0, strides = conv_transpose_0_strides_0, weight = var_171_to_fp16, x = input_23_cast_fp16)[name = tensor<string, []>("conv_transpose_0_has_output_shape_cast_fp16")];
+            tensor<int32, [3]> input_25_begin_0 = const()[name = tensor<string, []>("input_25_begin_0"), val = tensor<int32, [3]>([0, 0, 1])];
+            tensor<int32, [3]> input_25_end_0 = const()[name = tensor<string, []>("input_25_end_0"), val = tensor<int32, [3]>([0, 0, 0])];
+            tensor<bool, [3]> input_25_begin_mask_0 = const()[name = tensor<string, []>("input_25_begin_mask_0"), val = tensor<bool, [3]>([true, true, false])];
+            tensor<bool, [3]> input_25_end_mask_0 = const()[name = tensor<string, []>("input_25_end_mask_0"), val = tensor<bool, [3]>([true, true, true])];
+            tensor<fp16, [1, 512, 4000]> input_25_cast_fp16 = slice_by_index(begin = input_25_begin_0, begin_mask = input_25_begin_mask_0, end = input_25_end_0, end_mask = input_25_end_mask_0, x = conv_transpose_0_has_output_shape_cast_fp16)[name = tensor<string, []>("input_25_cast_fp16")];
+            tensor<string, []> input_29_pad_type_0 = const()[name = tensor<string, []>("input_29_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> input_29_pad_0 = const()[name = tensor<string, []>("input_29_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> input_29_strides_0 = const()[name = tensor<string, []>("input_29_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> input_29_dilations_0 = const()[name = tensor<string, []>("input_29_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> input_29_groups_0 = const()[name = tensor<string, []>("input_29_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 512, 3]> weight_5_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [393216]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(3824960))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4218240))), name = tensor<string, []>("weight_5_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 512, 3])];
+            tensor<fp16, [256]> F0_1_conv1_bias_to_fp16 = const()[name = tensor<string, []>("F0_1_conv1_bias_to_fp16"), val = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4218816)))];
+            tensor<fp16, [1, 256, 4000]> input_29_cast_fp16 = conv(bias = F0_1_conv1_bias_to_fp16, dilations = input_29_dilations_0, groups = input_29_groups_0, pad = input_29_pad_0, pad_type = input_29_pad_type_0, strides = input_29_strides_0, weight = weight_5_to_fp16_palettized, x = input_25_cast_fp16)[name = tensor<string, []>("input_29_cast_fp16")];
+            tensor<fp16, [512, 128]> F0_1_norm2_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [65536]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4219392))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4284992))), name = tensor<string, []>("F0_1_norm2_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([512, 128])];
+            tensor<fp16, [512]> F0_1_norm2_fc_bias_to_fp16 = const()[name = tensor<string, []>("F0_1_norm2_fc_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4285568)))];
+            tensor<fp16, [1, 512]> linear_3_cast_fp16 = linear(bias = F0_1_norm2_fc_bias_to_fp16, weight = F0_1_norm2_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_3_cast_fp16")];
+            tensor<int32, [3]> var_192 = const()[name = tensor<string, []>("op_192"), val = tensor<int32, [3]>([1, 512, 1])];
+            tensor<fp16, [1, 512, 1]> h_15_cast_fp16 = reshape(shape = var_192, x = linear_3_cast_fp16)[name = tensor<string, []>("h_15_cast_fp16")];
+            tensor<int32, [2]> var_194_split_sizes_0 = const()[name = tensor<string, []>("op_194_split_sizes_0"), val = tensor<int32, [2]>([256, 256])];
+            tensor<int32, []> var_194_axis_0 = const()[name = tensor<string, []>("op_194_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 256, 1]> var_194_cast_fp16_0, tensor<fp16, [1, 256, 1]> var_194_cast_fp16_1 = split(axis = var_194_axis_0, split_sizes = var_194_split_sizes_0, x = h_15_cast_fp16)[name = tensor<string, []>("op_194_cast_fp16")];
+            tensor<fp16, []> var_196_promoted_to_fp16 = const()[name = tensor<string, []>("op_196_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 256, 1]> var_197_cast_fp16 = add(x = var_194_cast_fp16_0, y = var_196_promoted_to_fp16)[name = tensor<string, []>("op_197_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_198_cast_fp16 = instance_norm(epsilon = var_130_to_fp16, x = input_29_cast_fp16)[name = tensor<string, []>("op_198_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_199_cast_fp16 = mul(x = var_197_cast_fp16, y = var_198_cast_fp16)[name = tensor<string, []>("op_199_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_31_cast_fp16 = add(x = var_199_cast_fp16, y = var_194_cast_fp16_1)[name = tensor<string, []>("input_31_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_33_cast_fp16 = leaky_relu(alpha = var_126, x = input_31_cast_fp16)[name = tensor<string, []>("input_33_cast_fp16")];
+            tensor<string, []> out_3_pad_type_0 = const()[name = tensor<string, []>("out_3_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> out_3_pad_0 = const()[name = tensor<string, []>("out_3_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> out_3_strides_0 = const()[name = tensor<string, []>("out_3_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> out_3_dilations_0 = const()[name = tensor<string, []>("out_3_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> out_3_groups_0 = const()[name = tensor<string, []>("out_3_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 256, 3]> weight_7_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [196608]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4286656))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4483328))), name = tensor<string, []>("weight_7_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 256, 3])];
+            tensor<fp16, [256]> F0_1_conv2_bias_to_fp16 = const()[name = tensor<string, []>("F0_1_conv2_bias_to_fp16"), val = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4483904)))];
+            tensor<fp16, [1, 256, 4000]> out_3_cast_fp16 = conv(bias = F0_1_conv2_bias_to_fp16, dilations = out_3_dilations_0, groups = out_3_groups_0, pad = out_3_pad_0, pad_type = out_3_pad_type_0, strides = out_3_strides_0, weight = weight_7_to_fp16_palettized, x = input_33_cast_fp16)[name = tensor<string, []>("out_3_cast_fp16")];
+            tensor<int32, [1]> expand_dims_0_axes_0 = const()[name = tensor<string, []>("expand_dims_0_axes_0"), val = tensor<int32, [1]>([3])];
+            tensor<fp16, [1, 512, 2000, 1]> expand_dims_0_cast_fp16 = expand_dims(axes = expand_dims_0_axes_0, x = _inversed_input_19_cast_fp16)[name = tensor<string, []>("expand_dims_0_cast_fp16")];
+            tensor<int32, []> upsample_nearest_neighbor_0_scale_factor_height_0 = const()[name = tensor<string, []>("upsample_nearest_neighbor_0_scale_factor_height_0"), val = tensor<int32, []>(2)];
+            tensor<int32, []> upsample_nearest_neighbor_0_scale_factor_width_0 = const()[name = tensor<string, []>("upsample_nearest_neighbor_0_scale_factor_width_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 4000, 1]> upsample_nearest_neighbor_0_cast_fp16 = upsample_nearest_neighbor(scale_factor_height = upsample_nearest_neighbor_0_scale_factor_height_0, scale_factor_width = upsample_nearest_neighbor_0_scale_factor_width_0, x = expand_dims_0_cast_fp16)[name = tensor<string, []>("upsample_nearest_neighbor_0_cast_fp16")];
+            tensor<int32, [1]> input_37_axes_0 = const()[name = tensor<string, []>("input_37_axes_0"), val = tensor<int32, [1]>([3])];
+            tensor<fp16, [1, 512, 4000]> input_37_cast_fp16 = squeeze(axes = input_37_axes_0, x = upsample_nearest_neighbor_0_cast_fp16)[name = tensor<string, []>("input_37_cast_fp16")];
+            tensor<string, []> var_217_pad_type_0 = const()[name = tensor<string, []>("op_217_pad_type_0"), val = tensor<string, []>("valid")];
+            tensor<int32, [1]> var_217_strides_0 = const()[name = tensor<string, []>("op_217_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [2]> var_217_pad_0 = const()[name = tensor<string, []>("op_217_pad_0"), val = tensor<int32, [2]>([0, 0])];
+            tensor<int32, [1]> var_217_dilations_0 = const()[name = tensor<string, []>("op_217_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> var_217_groups_0 = const()[name = tensor<string, []>("op_217_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 512, 1]> weight_9_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [131072]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4484480))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4615616))), name = tensor<string, []>("weight_9_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 512, 1])];
+            tensor<fp16, [1, 256, 4000]> var_217_cast_fp16 = conv(dilations = var_217_dilations_0, groups = var_217_groups_0, pad = var_217_pad_0, pad_type = var_217_pad_type_0, strides = var_217_strides_0, weight = weight_9_to_fp16_palettized, x = input_37_cast_fp16)[name = tensor<string, []>("op_217_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_218_cast_fp16 = add(x = out_3_cast_fp16, y = var_217_cast_fp16)[name = tensor<string, []>("op_218_cast_fp16")];
+            tensor<fp16, []> _inversed_input_39_y_0_to_fp16 = const()[name = tensor<string, []>("_inversed_input_39_y_0_to_fp16"), val = tensor<fp16, []>(0x1.6ap-1)];
+            tensor<fp16, [1, 256, 4000]> _inversed_input_39_cast_fp16 = mul(x = var_218_cast_fp16, y = _inversed_input_39_y_0_to_fp16)[name = tensor<string, []>("_inversed_input_39_cast_fp16")];
+            tensor<fp32, []> var_222 = const()[name = tensor<string, []>("op_222"), val = tensor<fp32, []>(0x1.99999ap-3)];
+            tensor<fp16, [512, 128]> F0_2_norm1_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [65536]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4616192))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4681792))), name = tensor<string, []>("F0_2_norm1_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([512, 128])];
+            tensor<fp16, [512]> F0_2_norm1_fc_bias_to_fp16 = const()[name = tensor<string, []>("F0_2_norm1_fc_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4682368)))];
+            tensor<fp16, [1, 512]> linear_4_cast_fp16 = linear(bias = F0_2_norm1_fc_bias_to_fp16, weight = F0_2_norm1_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_4_cast_fp16")];
+            tensor<int32, [3]> var_247 = const()[name = tensor<string, []>("op_247"), val = tensor<int32, [3]>([1, 512, 1])];
+            tensor<fp16, [1, 512, 1]> h_19_cast_fp16 = reshape(shape = var_247, x = linear_4_cast_fp16)[name = tensor<string, []>("h_19_cast_fp16")];
+            tensor<int32, [2]> var_249_split_sizes_0 = const()[name = tensor<string, []>("op_249_split_sizes_0"), val = tensor<int32, [2]>([256, 256])];
+            tensor<int32, []> var_249_axis_0 = const()[name = tensor<string, []>("op_249_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 256, 1]> var_249_cast_fp16_0, tensor<fp16, [1, 256, 1]> var_249_cast_fp16_1 = split(axis = var_249_axis_0, split_sizes = var_249_split_sizes_0, x = h_19_cast_fp16)[name = tensor<string, []>("op_249_cast_fp16")];
+            tensor<fp16, []> var_251_promoted_to_fp16 = const()[name = tensor<string, []>("op_251_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 256, 1]> var_252_cast_fp16 = add(x = var_249_cast_fp16_0, y = var_251_promoted_to_fp16)[name = tensor<string, []>("op_252_cast_fp16")];
+            tensor<fp16, []> var_225_to_fp16 = const()[name = tensor<string, []>("op_225_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<fp16, [1, 256, 4000]> var_253_cast_fp16 = instance_norm(epsilon = var_225_to_fp16, x = _inversed_input_39_cast_fp16)[name = tensor<string, []>("op_253_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_254_cast_fp16 = mul(x = var_252_cast_fp16, y = var_253_cast_fp16)[name = tensor<string, []>("op_254_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_41_cast_fp16 = add(x = var_254_cast_fp16, y = var_249_cast_fp16_1)[name = tensor<string, []>("input_41_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_43_cast_fp16 = leaky_relu(alpha = var_222, x = input_41_cast_fp16)[name = tensor<string, []>("input_43_cast_fp16")];
+            tensor<string, []> input_47_pad_type_0 = const()[name = tensor<string, []>("input_47_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> input_47_pad_0 = const()[name = tensor<string, []>("input_47_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> input_47_strides_0 = const()[name = tensor<string, []>("input_47_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> input_47_dilations_0 = const()[name = tensor<string, []>("input_47_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> input_47_groups_0 = const()[name = tensor<string, []>("input_47_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 256, 3]> weight_11_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [196608]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4683456))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4880128))), name = tensor<string, []>("weight_11_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 256, 3])];
+            tensor<fp16, [256]> F0_2_conv1_bias_to_fp16 = const()[name = tensor<string, []>("F0_2_conv1_bias_to_fp16"), val = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4880704)))];
+            tensor<fp16, [1, 256, 4000]> input_47_cast_fp16 = conv(bias = F0_2_conv1_bias_to_fp16, dilations = input_47_dilations_0, groups = input_47_groups_0, pad = input_47_pad_0, pad_type = input_47_pad_type_0, strides = input_47_strides_0, weight = weight_11_to_fp16_palettized, x = input_43_cast_fp16)[name = tensor<string, []>("input_47_cast_fp16")];
+            tensor<fp16, [512, 128]> F0_2_norm2_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [65536]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4881280))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4946880))), name = tensor<string, []>("F0_2_norm2_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([512, 128])];
+            tensor<fp16, [512]> F0_2_norm2_fc_bias_to_fp16 = const()[name = tensor<string, []>("F0_2_norm2_fc_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4947456)))];
+            tensor<fp16, [1, 512]> linear_5_cast_fp16 = linear(bias = F0_2_norm2_fc_bias_to_fp16, weight = F0_2_norm2_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_5_cast_fp16")];
+            tensor<int32, [3]> var_271 = const()[name = tensor<string, []>("op_271"), val = tensor<int32, [3]>([1, 512, 1])];
+            tensor<fp16, [1, 512, 1]> h_23_cast_fp16 = reshape(shape = var_271, x = linear_5_cast_fp16)[name = tensor<string, []>("h_23_cast_fp16")];
+            tensor<int32, [2]> var_273_split_sizes_0 = const()[name = tensor<string, []>("op_273_split_sizes_0"), val = tensor<int32, [2]>([256, 256])];
+            tensor<int32, []> var_273_axis_0 = const()[name = tensor<string, []>("op_273_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 256, 1]> var_273_cast_fp16_0, tensor<fp16, [1, 256, 1]> var_273_cast_fp16_1 = split(axis = var_273_axis_0, split_sizes = var_273_split_sizes_0, x = h_23_cast_fp16)[name = tensor<string, []>("op_273_cast_fp16")];
+            tensor<fp16, []> var_275_promoted_to_fp16 = const()[name = tensor<string, []>("op_275_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 256, 1]> var_276_cast_fp16 = add(x = var_273_cast_fp16_0, y = var_275_promoted_to_fp16)[name = tensor<string, []>("op_276_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_277_cast_fp16 = instance_norm(epsilon = var_225_to_fp16, x = input_47_cast_fp16)[name = tensor<string, []>("op_277_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_278_cast_fp16 = mul(x = var_276_cast_fp16, y = var_277_cast_fp16)[name = tensor<string, []>("op_278_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_49_cast_fp16 = add(x = var_278_cast_fp16, y = var_273_cast_fp16_1)[name = tensor<string, []>("input_49_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_51_cast_fp16 = leaky_relu(alpha = var_222, x = input_49_cast_fp16)[name = tensor<string, []>("input_51_cast_fp16")];
+            tensor<string, []> out_5_pad_type_0 = const()[name = tensor<string, []>("out_5_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> out_5_pad_0 = const()[name = tensor<string, []>("out_5_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> out_5_strides_0 = const()[name = tensor<string, []>("out_5_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> out_5_dilations_0 = const()[name = tensor<string, []>("out_5_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> out_5_groups_0 = const()[name = tensor<string, []>("out_5_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 256, 3]> weight_13_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [196608]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(4948544))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5145216))), name = tensor<string, []>("weight_13_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 256, 3])];
+            tensor<fp16, [256]> F0_2_conv2_bias_to_fp16 = const()[name = tensor<string, []>("F0_2_conv2_bias_to_fp16"), val = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5145792)))];
+            tensor<fp16, [1, 256, 4000]> out_5_cast_fp16 = conv(bias = F0_2_conv2_bias_to_fp16, dilations = out_5_dilations_0, groups = out_5_groups_0, pad = out_5_pad_0, pad_type = out_5_pad_type_0, strides = out_5_strides_0, weight = weight_13_to_fp16_palettized, x = input_51_cast_fp16)[name = tensor<string, []>("out_5_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_289_cast_fp16 = add(x = out_5_cast_fp16, y = _inversed_input_39_cast_fp16)[name = tensor<string, []>("op_289_cast_fp16")];
+            tensor<fp16, []> _inversed_input_55_y_0_to_fp16 = const()[name = tensor<string, []>("_inversed_input_55_y_0_to_fp16"), val = tensor<fp16, []>(0x1.6ap-1)];
+            tensor<fp16, [1, 256, 4000]> _inversed_input_55_cast_fp16 = mul(x = var_289_cast_fp16, y = _inversed_input_55_y_0_to_fp16)[name = tensor<string, []>("_inversed_input_55_cast_fp16")];
+            tensor<string, []> var_302_pad_type_0 = const()[name = tensor<string, []>("op_302_pad_type_0"), val = tensor<string, []>("valid")];
+            tensor<int32, [1]> var_302_strides_0 = const()[name = tensor<string, []>("op_302_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [2]> var_302_pad_0 = const()[name = tensor<string, []>("op_302_pad_0"), val = tensor<int32, [2]>([0, 0])];
+            tensor<int32, [1]> var_302_dilations_0 = const()[name = tensor<string, []>("op_302_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> var_302_groups_0 = const()[name = tensor<string, []>("op_302_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 256, 1]> F0_proj_weight_to_fp16 = const()[name = tensor<string, []>("F0_proj_weight_to_fp16"), val = tensor<fp16, [1, 256, 1]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5146368)))];
+            tensor<fp16, [1]> F0_proj_bias_to_fp16 = const()[name = tensor<string, []>("F0_proj_bias_to_fp16"), val = tensor<fp16, [1]>([0x1.adp-4])];
+            tensor<fp16, [1, 1, 4000]> var_302_cast_fp16 = conv(bias = F0_proj_bias_to_fp16, dilations = var_302_dilations_0, groups = var_302_groups_0, pad = var_302_pad_0, pad_type = var_302_pad_type_0, strides = var_302_strides_0, weight = F0_proj_weight_to_fp16, x = _inversed_input_55_cast_fp16)[name = tensor<string, []>("op_302_cast_fp16")];
+            tensor<int32, [1]> var_304_axes_0 = const()[name = tensor<string, []>("op_304_axes_0"), val = tensor<int32, [1]>([1])];
+            tensor<fp16, [1, 4000]> F0 = squeeze(axes = var_304_axes_0, x = var_302_cast_fp16)[name = tensor<string, []>("op_304_cast_fp16")];
+            tensor<fp32, []> var_309 = const()[name = tensor<string, []>("op_309"), val = tensor<fp32, []>(0x1.99999ap-3)];
+            tensor<fp16, [1024, 128]> N_0_norm1_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [131072]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5146944))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5278080))), name = tensor<string, []>("N_0_norm1_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 128])];
+            tensor<fp16, [1024]> N_0_norm1_fc_bias_to_fp16 = const()[name = tensor<string, []>("N_0_norm1_fc_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5278656)))];
+            tensor<fp16, [1, 1024]> linear_6_cast_fp16 = linear(bias = N_0_norm1_fc_bias_to_fp16, weight = N_0_norm1_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_6_cast_fp16")];
+            tensor<int32, [3]> var_334 = const()[name = tensor<string, []>("op_334"), val = tensor<int32, [3]>([1, 1024, 1])];
+            tensor<fp16, [1, 1024, 1]> h_27_cast_fp16 = reshape(shape = var_334, x = linear_6_cast_fp16)[name = tensor<string, []>("h_27_cast_fp16")];
+            tensor<int32, [2]> var_336_split_sizes_0 = const()[name = tensor<string, []>("op_336_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_336_axis_0 = const()[name = tensor<string, []>("op_336_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 1]> var_336_cast_fp16_0, tensor<fp16, [1, 512, 1]> var_336_cast_fp16_1 = split(axis = var_336_axis_0, split_sizes = var_336_split_sizes_0, x = h_27_cast_fp16)[name = tensor<string, []>("op_336_cast_fp16")];
+            tensor<fp16, []> var_338_promoted_to_fp16 = const()[name = tensor<string, []>("op_338_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 512, 1]> var_339_cast_fp16 = add(x = var_336_cast_fp16_0, y = var_338_promoted_to_fp16)[name = tensor<string, []>("op_339_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> var_341_cast_fp16 = mul(x = var_339_cast_fp16, y = var_84_cast_fp16)[name = tensor<string, []>("op_341_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_59_cast_fp16 = add(x = var_341_cast_fp16, y = var_336_cast_fp16_1)[name = tensor<string, []>("input_59_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_61_cast_fp16 = leaky_relu(alpha = var_309, x = input_59_cast_fp16)[name = tensor<string, []>("input_61_cast_fp16")];
+            tensor<string, []> input_65_pad_type_0 = const()[name = tensor<string, []>("input_65_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> input_65_pad_0 = const()[name = tensor<string, []>("input_65_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> input_65_strides_0 = const()[name = tensor<string, []>("input_65_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> input_65_dilations_0 = const()[name = tensor<string, []>("input_65_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> input_65_groups_0 = const()[name = tensor<string, []>("input_65_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [512, 512, 3]> weight_17_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [786432]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(5280768))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(6067264))), name = tensor<string, []>("weight_17_to_fp16_palettized"), shape = tensor<uint32, [3]>([512, 512, 3])];
+            tensor<fp16, [512]> N_0_conv1_bias_to_fp16 = const()[name = tensor<string, []>("N_0_conv1_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(6067840)))];
+            tensor<fp16, [1, 512, 2000]> input_65_cast_fp16 = conv(bias = N_0_conv1_bias_to_fp16, dilations = input_65_dilations_0, groups = input_65_groups_0, pad = input_65_pad_0, pad_type = input_65_pad_type_0, strides = input_65_strides_0, weight = weight_17_to_fp16_palettized, x = input_61_cast_fp16)[name = tensor<string, []>("input_65_cast_fp16")];
+            tensor<fp16, [1024, 128]> N_0_norm2_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [131072]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(6068928))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(6200064))), name = tensor<string, []>("N_0_norm2_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 128])];
+            tensor<fp16, [1024]> N_0_norm2_fc_bias_to_fp16 = const()[name = tensor<string, []>("N_0_norm2_fc_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(6200640)))];
+            tensor<fp16, [1, 1024]> linear_7_cast_fp16 = linear(bias = N_0_norm2_fc_bias_to_fp16, weight = N_0_norm2_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_7_cast_fp16")];
+            tensor<int32, [3]> var_358 = const()[name = tensor<string, []>("op_358"), val = tensor<int32, [3]>([1, 1024, 1])];
+            tensor<fp16, [1, 1024, 1]> h_31_cast_fp16 = reshape(shape = var_358, x = linear_7_cast_fp16)[name = tensor<string, []>("h_31_cast_fp16")];
+            tensor<int32, [2]> var_360_split_sizes_0 = const()[name = tensor<string, []>("op_360_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_360_axis_0 = const()[name = tensor<string, []>("op_360_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 1]> var_360_cast_fp16_0, tensor<fp16, [1, 512, 1]> var_360_cast_fp16_1 = split(axis = var_360_axis_0, split_sizes = var_360_split_sizes_0, x = h_31_cast_fp16)[name = tensor<string, []>("op_360_cast_fp16")];
+            tensor<fp16, []> var_362_promoted_to_fp16 = const()[name = tensor<string, []>("op_362_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 512, 1]> var_363_cast_fp16 = add(x = var_360_cast_fp16_0, y = var_362_promoted_to_fp16)[name = tensor<string, []>("op_363_cast_fp16")];
+            tensor<fp16, []> var_312_to_fp16 = const()[name = tensor<string, []>("op_312_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<fp16, [1, 512, 2000]> var_364_cast_fp16 = instance_norm(epsilon = var_312_to_fp16, x = input_65_cast_fp16)[name = tensor<string, []>("op_364_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> var_365_cast_fp16 = mul(x = var_363_cast_fp16, y = var_364_cast_fp16)[name = tensor<string, []>("op_365_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_67_cast_fp16 = add(x = var_365_cast_fp16, y = var_360_cast_fp16_1)[name = tensor<string, []>("input_67_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_69_cast_fp16 = leaky_relu(alpha = var_309, x = input_67_cast_fp16)[name = tensor<string, []>("input_69_cast_fp16")];
+            tensor<string, []> out_7_pad_type_0 = const()[name = tensor<string, []>("out_7_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> out_7_pad_0 = const()[name = tensor<string, []>("out_7_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> out_7_strides_0 = const()[name = tensor<string, []>("out_7_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> out_7_dilations_0 = const()[name = tensor<string, []>("out_7_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> out_7_groups_0 = const()[name = tensor<string, []>("out_7_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [512, 512, 3]> weight_19_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [786432]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(6202752))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(6989248))), name = tensor<string, []>("weight_19_to_fp16_palettized"), shape = tensor<uint32, [3]>([512, 512, 3])];
+            tensor<fp16, [512]> N_0_conv2_bias_to_fp16 = const()[name = tensor<string, []>("N_0_conv2_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(6989824)))];
+            tensor<fp16, [1, 512, 2000]> out_7_cast_fp16 = conv(bias = N_0_conv2_bias_to_fp16, dilations = out_7_dilations_0, groups = out_7_groups_0, pad = out_7_pad_0, pad_type = out_7_pad_type_0, strides = out_7_strides_0, weight = weight_19_to_fp16_palettized, x = input_69_cast_fp16)[name = tensor<string, []>("out_7_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> var_376_cast_fp16 = add(x = out_7_cast_fp16, y = transpose_0)[name = tensor<string, []>("op_376_cast_fp16")];
+            tensor<fp16, []> _inversed_input_73_y_0_to_fp16 = const()[name = tensor<string, []>("_inversed_input_73_y_0_to_fp16"), val = tensor<fp16, []>(0x1.6ap-1)];
+            tensor<fp16, [1, 512, 2000]> _inversed_input_73_cast_fp16 = mul(x = var_376_cast_fp16, y = _inversed_input_73_y_0_to_fp16)[name = tensor<string, []>("_inversed_input_73_cast_fp16")];
+            tensor<fp32, []> var_382 = const()[name = tensor<string, []>("op_382"), val = tensor<fp32, []>(0x1.99999ap-3)];
+            tensor<fp16, [1024, 128]> N_1_norm1_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [131072]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(6990912))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7122048))), name = tensor<string, []>("N_1_norm1_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([1024, 128])];
+            tensor<fp16, [1024]> N_1_norm1_fc_bias_to_fp16 = const()[name = tensor<string, []>("N_1_norm1_fc_bias_to_fp16"), val = tensor<fp16, [1024]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7122624)))];
+            tensor<fp16, [1, 1024]> linear_8_cast_fp16 = linear(bias = N_1_norm1_fc_bias_to_fp16, weight = N_1_norm1_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_8_cast_fp16")];
+            tensor<int32, [3]> var_417 = const()[name = tensor<string, []>("op_417"), val = tensor<int32, [3]>([1, 1024, 1])];
+            tensor<fp16, [1, 1024, 1]> h_35_cast_fp16 = reshape(shape = var_417, x = linear_8_cast_fp16)[name = tensor<string, []>("h_35_cast_fp16")];
+            tensor<int32, [2]> var_419_split_sizes_0 = const()[name = tensor<string, []>("op_419_split_sizes_0"), val = tensor<int32, [2]>([512, 512])];
+            tensor<int32, []> var_419_axis_0 = const()[name = tensor<string, []>("op_419_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 1]> var_419_cast_fp16_0, tensor<fp16, [1, 512, 1]> var_419_cast_fp16_1 = split(axis = var_419_axis_0, split_sizes = var_419_split_sizes_0, x = h_35_cast_fp16)[name = tensor<string, []>("op_419_cast_fp16")];
+            tensor<fp16, []> var_421_promoted_to_fp16 = const()[name = tensor<string, []>("op_421_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 512, 1]> var_422_cast_fp16 = add(x = var_419_cast_fp16_0, y = var_421_promoted_to_fp16)[name = tensor<string, []>("op_422_cast_fp16")];
+            tensor<fp16, []> var_386_to_fp16 = const()[name = tensor<string, []>("op_386_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<fp16, [1, 512, 2000]> var_423_cast_fp16 = instance_norm(epsilon = var_386_to_fp16, x = _inversed_input_73_cast_fp16)[name = tensor<string, []>("op_423_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> var_424_cast_fp16 = mul(x = var_422_cast_fp16, y = var_423_cast_fp16)[name = tensor<string, []>("op_424_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_75_cast_fp16 = add(x = var_424_cast_fp16, y = var_419_cast_fp16_1)[name = tensor<string, []>("input_75_cast_fp16")];
+            tensor<fp16, [1, 512, 2000]> input_77_cast_fp16 = leaky_relu(alpha = var_382, x = input_75_cast_fp16)[name = tensor<string, []>("input_77_cast_fp16")];
+            tensor<string, []> conv_transpose_1_pad_type_0 = const()[name = tensor<string, []>("conv_transpose_1_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> conv_transpose_1_pad_0 = const()[name = tensor<string, []>("conv_transpose_1_pad_0"), val = tensor<int32, [2]>([0, 0])];
+            tensor<int32, [1]> conv_transpose_1_strides_0 = const()[name = tensor<string, []>("conv_transpose_1_strides_0"), val = tensor<int32, [1]>([2])];
+            tensor<int32, []> conv_transpose_1_groups_0 = const()[name = tensor<string, []>("conv_transpose_1_groups_0"), val = tensor<int32, []>(512)];
+            tensor<int32, [1]> conv_transpose_1_dilations_0 = const()[name = tensor<string, []>("conv_transpose_1_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [3]> conv_transpose_1_has_output_shape_output_shape_0 = const()[name = tensor<string, []>("conv_transpose_1_has_output_shape_output_shape_0"), val = tensor<int32, [3]>([1, 512, 4001])];
+            tensor<fp16, [512, 1, 3]> var_427_to_fp16 = const()[name = tensor<string, []>("op_427_to_fp16"), val = tensor<fp16, [512, 1, 3]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7124736)))];
+            tensor<fp16, [512]> N_1_pool_bias_to_fp16 = const()[name = tensor<string, []>("N_1_pool_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7127872)))];
+            tensor<fp16, [1, 512, 4001]> conv_transpose_1_has_output_shape_cast_fp16 = conv_transpose(bias = N_1_pool_bias_to_fp16, dilations = conv_transpose_1_dilations_0, groups = conv_transpose_1_groups_0, output_shape = conv_transpose_1_has_output_shape_output_shape_0, pad = conv_transpose_1_pad_0, pad_type = conv_transpose_1_pad_type_0, strides = conv_transpose_1_strides_0, weight = var_427_to_fp16, x = input_77_cast_fp16)[name = tensor<string, []>("conv_transpose_1_has_output_shape_cast_fp16")];
+            tensor<int32, [3]> input_79_begin_0 = const()[name = tensor<string, []>("input_79_begin_0"), val = tensor<int32, [3]>([0, 0, 1])];
+            tensor<int32, [3]> input_79_end_0 = const()[name = tensor<string, []>("input_79_end_0"), val = tensor<int32, [3]>([0, 0, 0])];
+            tensor<bool, [3]> input_79_begin_mask_0 = const()[name = tensor<string, []>("input_79_begin_mask_0"), val = tensor<bool, [3]>([true, true, false])];
+            tensor<bool, [3]> input_79_end_mask_0 = const()[name = tensor<string, []>("input_79_end_mask_0"), val = tensor<bool, [3]>([true, true, true])];
+            tensor<fp16, [1, 512, 4000]> input_79_cast_fp16 = slice_by_index(begin = input_79_begin_0, begin_mask = input_79_begin_mask_0, end = input_79_end_0, end_mask = input_79_end_mask_0, x = conv_transpose_1_has_output_shape_cast_fp16)[name = tensor<string, []>("input_79_cast_fp16")];
+            tensor<string, []> input_83_pad_type_0 = const()[name = tensor<string, []>("input_83_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> input_83_pad_0 = const()[name = tensor<string, []>("input_83_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> input_83_strides_0 = const()[name = tensor<string, []>("input_83_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> input_83_dilations_0 = const()[name = tensor<string, []>("input_83_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> input_83_groups_0 = const()[name = tensor<string, []>("input_83_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 512, 3]> weight_21_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [393216]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7128960))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7522240))), name = tensor<string, []>("weight_21_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 512, 3])];
+            tensor<fp16, [256]> N_1_conv1_bias_to_fp16 = const()[name = tensor<string, []>("N_1_conv1_bias_to_fp16"), val = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7522816)))];
+            tensor<fp16, [1, 256, 4000]> input_83_cast_fp16 = conv(bias = N_1_conv1_bias_to_fp16, dilations = input_83_dilations_0, groups = input_83_groups_0, pad = input_83_pad_0, pad_type = input_83_pad_type_0, strides = input_83_strides_0, weight = weight_21_to_fp16_palettized, x = input_79_cast_fp16)[name = tensor<string, []>("input_83_cast_fp16")];
+            tensor<fp16, [512, 128]> N_1_norm2_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [65536]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7523392))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7588992))), name = tensor<string, []>("N_1_norm2_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([512, 128])];
+            tensor<fp16, [512]> N_1_norm2_fc_bias_to_fp16 = const()[name = tensor<string, []>("N_1_norm2_fc_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7589568)))];
+            tensor<fp16, [1, 512]> linear_9_cast_fp16 = linear(bias = N_1_norm2_fc_bias_to_fp16, weight = N_1_norm2_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_9_cast_fp16")];
+            tensor<int32, [3]> var_448 = const()[name = tensor<string, []>("op_448"), val = tensor<int32, [3]>([1, 512, 1])];
+            tensor<fp16, [1, 512, 1]> h_39_cast_fp16 = reshape(shape = var_448, x = linear_9_cast_fp16)[name = tensor<string, []>("h_39_cast_fp16")];
+            tensor<int32, [2]> var_450_split_sizes_0 = const()[name = tensor<string, []>("op_450_split_sizes_0"), val = tensor<int32, [2]>([256, 256])];
+            tensor<int32, []> var_450_axis_0 = const()[name = tensor<string, []>("op_450_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 256, 1]> var_450_cast_fp16_0, tensor<fp16, [1, 256, 1]> var_450_cast_fp16_1 = split(axis = var_450_axis_0, split_sizes = var_450_split_sizes_0, x = h_39_cast_fp16)[name = tensor<string, []>("op_450_cast_fp16")];
+            tensor<fp16, []> var_452_promoted_to_fp16 = const()[name = tensor<string, []>("op_452_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 256, 1]> var_453_cast_fp16 = add(x = var_450_cast_fp16_0, y = var_452_promoted_to_fp16)[name = tensor<string, []>("op_453_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_454_cast_fp16 = instance_norm(epsilon = var_386_to_fp16, x = input_83_cast_fp16)[name = tensor<string, []>("op_454_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_455_cast_fp16 = mul(x = var_453_cast_fp16, y = var_454_cast_fp16)[name = tensor<string, []>("op_455_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_85_cast_fp16 = add(x = var_455_cast_fp16, y = var_450_cast_fp16_1)[name = tensor<string, []>("input_85_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_87_cast_fp16 = leaky_relu(alpha = var_382, x = input_85_cast_fp16)[name = tensor<string, []>("input_87_cast_fp16")];
+            tensor<string, []> out_9_pad_type_0 = const()[name = tensor<string, []>("out_9_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> out_9_pad_0 = const()[name = tensor<string, []>("out_9_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> out_9_strides_0 = const()[name = tensor<string, []>("out_9_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> out_9_dilations_0 = const()[name = tensor<string, []>("out_9_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> out_9_groups_0 = const()[name = tensor<string, []>("out_9_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 256, 3]> weight_23_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [196608]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7590656))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7787328))), name = tensor<string, []>("weight_23_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 256, 3])];
+            tensor<fp16, [256]> N_1_conv2_bias_to_fp16 = const()[name = tensor<string, []>("N_1_conv2_bias_to_fp16"), val = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7787904)))];
+            tensor<fp16, [1, 256, 4000]> out_9_cast_fp16 = conv(bias = N_1_conv2_bias_to_fp16, dilations = out_9_dilations_0, groups = out_9_groups_0, pad = out_9_pad_0, pad_type = out_9_pad_type_0, strides = out_9_strides_0, weight = weight_23_to_fp16_palettized, x = input_87_cast_fp16)[name = tensor<string, []>("out_9_cast_fp16")];
+            tensor<int32, [1]> expand_dims_1_axes_0 = const()[name = tensor<string, []>("expand_dims_1_axes_0"), val = tensor<int32, [1]>([3])];
+            tensor<fp16, [1, 512, 2000, 1]> expand_dims_1_cast_fp16 = expand_dims(axes = expand_dims_1_axes_0, x = _inversed_input_73_cast_fp16)[name = tensor<string, []>("expand_dims_1_cast_fp16")];
+            tensor<int32, []> upsample_nearest_neighbor_1_scale_factor_height_0 = const()[name = tensor<string, []>("upsample_nearest_neighbor_1_scale_factor_height_0"), val = tensor<int32, []>(2)];
+            tensor<int32, []> upsample_nearest_neighbor_1_scale_factor_width_0 = const()[name = tensor<string, []>("upsample_nearest_neighbor_1_scale_factor_width_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 512, 4000, 1]> upsample_nearest_neighbor_1_cast_fp16 = upsample_nearest_neighbor(scale_factor_height = upsample_nearest_neighbor_1_scale_factor_height_0, scale_factor_width = upsample_nearest_neighbor_1_scale_factor_width_0, x = expand_dims_1_cast_fp16)[name = tensor<string, []>("upsample_nearest_neighbor_1_cast_fp16")];
+            tensor<int32, [1]> input_91_axes_0 = const()[name = tensor<string, []>("input_91_axes_0"), val = tensor<int32, [1]>([3])];
+            tensor<fp16, [1, 512, 4000]> input_91_cast_fp16 = squeeze(axes = input_91_axes_0, x = upsample_nearest_neighbor_1_cast_fp16)[name = tensor<string, []>("input_91_cast_fp16")];
+            tensor<string, []> var_473_pad_type_0 = const()[name = tensor<string, []>("op_473_pad_type_0"), val = tensor<string, []>("valid")];
+            tensor<int32, [1]> var_473_strides_0 = const()[name = tensor<string, []>("op_473_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [2]> var_473_pad_0 = const()[name = tensor<string, []>("op_473_pad_0"), val = tensor<int32, [2]>([0, 0])];
+            tensor<int32, [1]> var_473_dilations_0 = const()[name = tensor<string, []>("op_473_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> var_473_groups_0 = const()[name = tensor<string, []>("op_473_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 512, 1]> weight_25_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [131072]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7788480))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7919616))), name = tensor<string, []>("weight_25_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 512, 1])];
+            tensor<fp16, [1, 256, 4000]> var_473_cast_fp16 = conv(dilations = var_473_dilations_0, groups = var_473_groups_0, pad = var_473_pad_0, pad_type = var_473_pad_type_0, strides = var_473_strides_0, weight = weight_25_to_fp16_palettized, x = input_91_cast_fp16)[name = tensor<string, []>("op_473_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_474_cast_fp16 = add(x = out_9_cast_fp16, y = var_473_cast_fp16)[name = tensor<string, []>("op_474_cast_fp16")];
+            tensor<fp16, []> _inversed_input_93_y_0_to_fp16 = const()[name = tensor<string, []>("_inversed_input_93_y_0_to_fp16"), val = tensor<fp16, []>(0x1.6ap-1)];
+            tensor<fp16, [1, 256, 4000]> _inversed_input_93_cast_fp16 = mul(x = var_474_cast_fp16, y = _inversed_input_93_y_0_to_fp16)[name = tensor<string, []>("_inversed_input_93_cast_fp16")];
+            tensor<fp32, []> var_478 = const()[name = tensor<string, []>("op_478"), val = tensor<fp32, []>(0x1.99999ap-3)];
+            tensor<fp16, [512, 128]> N_2_norm1_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [65536]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7920192))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7985792))), name = tensor<string, []>("N_2_norm1_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([512, 128])];
+            tensor<fp16, [512]> N_2_norm1_fc_bias_to_fp16 = const()[name = tensor<string, []>("N_2_norm1_fc_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7986368)))];
+            tensor<fp16, [1, 512]> linear_10_cast_fp16 = linear(bias = N_2_norm1_fc_bias_to_fp16, weight = N_2_norm1_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_10_cast_fp16")];
+            tensor<int32, [3]> var_503 = const()[name = tensor<string, []>("op_503"), val = tensor<int32, [3]>([1, 512, 1])];
+            tensor<fp16, [1, 512, 1]> h_43_cast_fp16 = reshape(shape = var_503, x = linear_10_cast_fp16)[name = tensor<string, []>("h_43_cast_fp16")];
+            tensor<int32, [2]> var_505_split_sizes_0 = const()[name = tensor<string, []>("op_505_split_sizes_0"), val = tensor<int32, [2]>([256, 256])];
+            tensor<int32, []> var_505_axis_0 = const()[name = tensor<string, []>("op_505_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 256, 1]> var_505_cast_fp16_0, tensor<fp16, [1, 256, 1]> var_505_cast_fp16_1 = split(axis = var_505_axis_0, split_sizes = var_505_split_sizes_0, x = h_43_cast_fp16)[name = tensor<string, []>("op_505_cast_fp16")];
+            tensor<fp16, []> var_507_promoted_to_fp16 = const()[name = tensor<string, []>("op_507_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 256, 1]> var_508_cast_fp16 = add(x = var_505_cast_fp16_0, y = var_507_promoted_to_fp16)[name = tensor<string, []>("op_508_cast_fp16")];
+            tensor<fp16, []> var_481_to_fp16 = const()[name = tensor<string, []>("op_481_to_fp16"), val = tensor<fp16, []>(0x1.5p-17)];
+            tensor<fp16, [1, 256, 4000]> var_509_cast_fp16 = instance_norm(epsilon = var_481_to_fp16, x = _inversed_input_93_cast_fp16)[name = tensor<string, []>("op_509_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_510_cast_fp16 = mul(x = var_508_cast_fp16, y = var_509_cast_fp16)[name = tensor<string, []>("op_510_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_95_cast_fp16 = add(x = var_510_cast_fp16, y = var_505_cast_fp16_1)[name = tensor<string, []>("input_95_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_97_cast_fp16 = leaky_relu(alpha = var_478, x = input_95_cast_fp16)[name = tensor<string, []>("input_97_cast_fp16")];
+            tensor<string, []> input_101_pad_type_0 = const()[name = tensor<string, []>("input_101_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> input_101_pad_0 = const()[name = tensor<string, []>("input_101_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> input_101_strides_0 = const()[name = tensor<string, []>("input_101_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> input_101_dilations_0 = const()[name = tensor<string, []>("input_101_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> input_101_groups_0 = const()[name = tensor<string, []>("input_101_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 256, 3]> weight_27_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [196608]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(7987456))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8184128))), name = tensor<string, []>("weight_27_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 256, 3])];
+            tensor<fp16, [256]> N_2_conv1_bias_to_fp16 = const()[name = tensor<string, []>("N_2_conv1_bias_to_fp16"), val = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8184704)))];
+            tensor<fp16, [1, 256, 4000]> input_101_cast_fp16 = conv(bias = N_2_conv1_bias_to_fp16, dilations = input_101_dilations_0, groups = input_101_groups_0, pad = input_101_pad_0, pad_type = input_101_pad_type_0, strides = input_101_strides_0, weight = weight_27_to_fp16_palettized, x = input_97_cast_fp16)[name = tensor<string, []>("input_101_cast_fp16")];
+            tensor<fp16, [512, 128]> N_2_norm2_fc_weight_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [65536]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8185280))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8250880))), name = tensor<string, []>("N_2_norm2_fc_weight_to_fp16_palettized"), shape = tensor<uint32, [2]>([512, 128])];
+            tensor<fp16, [512]> N_2_norm2_fc_bias_to_fp16 = const()[name = tensor<string, []>("N_2_norm2_fc_bias_to_fp16"), val = tensor<fp16, [512]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8251456)))];
+            tensor<fp16, [1, 512]> linear_11_cast_fp16 = linear(bias = N_2_norm2_fc_bias_to_fp16, weight = N_2_norm2_fc_weight_to_fp16_palettized, x = s_to_fp16)[name = tensor<string, []>("linear_11_cast_fp16")];
+            tensor<int32, [3]> var_527 = const()[name = tensor<string, []>("op_527"), val = tensor<int32, [3]>([1, 512, 1])];
+            tensor<fp16, [1, 512, 1]> h_cast_fp16 = reshape(shape = var_527, x = linear_11_cast_fp16)[name = tensor<string, []>("h_cast_fp16")];
+            tensor<int32, [2]> var_529_split_sizes_0 = const()[name = tensor<string, []>("op_529_split_sizes_0"), val = tensor<int32, [2]>([256, 256])];
+            tensor<int32, []> var_529_axis_0 = const()[name = tensor<string, []>("op_529_axis_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 256, 1]> var_529_cast_fp16_0, tensor<fp16, [1, 256, 1]> var_529_cast_fp16_1 = split(axis = var_529_axis_0, split_sizes = var_529_split_sizes_0, x = h_cast_fp16)[name = tensor<string, []>("op_529_cast_fp16")];
+            tensor<fp16, []> var_531_promoted_to_fp16 = const()[name = tensor<string, []>("op_531_promoted_to_fp16"), val = tensor<fp16, []>(0x1p+0)];
+            tensor<fp16, [1, 256, 1]> var_532_cast_fp16 = add(x = var_529_cast_fp16_0, y = var_531_promoted_to_fp16)[name = tensor<string, []>("op_532_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_533_cast_fp16 = instance_norm(epsilon = var_481_to_fp16, x = input_101_cast_fp16)[name = tensor<string, []>("op_533_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_534_cast_fp16 = mul(x = var_532_cast_fp16, y = var_533_cast_fp16)[name = tensor<string, []>("op_534_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_103_cast_fp16 = add(x = var_534_cast_fp16, y = var_529_cast_fp16_1)[name = tensor<string, []>("input_103_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> input_105_cast_fp16 = leaky_relu(alpha = var_478, x = input_103_cast_fp16)[name = tensor<string, []>("input_105_cast_fp16")];
+            tensor<string, []> out_pad_type_0 = const()[name = tensor<string, []>("out_pad_type_0"), val = tensor<string, []>("custom")];
+            tensor<int32, [2]> out_pad_0 = const()[name = tensor<string, []>("out_pad_0"), val = tensor<int32, [2]>([1, 1])];
+            tensor<int32, [1]> out_strides_0 = const()[name = tensor<string, []>("out_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [1]> out_dilations_0 = const()[name = tensor<string, []>("out_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> out_groups_0 = const()[name = tensor<string, []>("out_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [256, 256, 3]> weight_29_to_fp16_palettized = constexpr_lut_to_dense()[indices = tensor<uint8, [196608]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8252544))), lut = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8449216))), name = tensor<string, []>("weight_29_to_fp16_palettized"), shape = tensor<uint32, [3]>([256, 256, 3])];
+            tensor<fp16, [256]> N_2_conv2_bias_to_fp16 = const()[name = tensor<string, []>("N_2_conv2_bias_to_fp16"), val = tensor<fp16, [256]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8449792)))];
+            tensor<fp16, [1, 256, 4000]> out_cast_fp16 = conv(bias = N_2_conv2_bias_to_fp16, dilations = out_dilations_0, groups = out_groups_0, pad = out_pad_0, pad_type = out_pad_type_0, strides = out_strides_0, weight = weight_29_to_fp16_palettized, x = input_105_cast_fp16)[name = tensor<string, []>("out_cast_fp16")];
+            tensor<fp16, [1, 256, 4000]> var_545_cast_fp16 = add(x = out_cast_fp16, y = _inversed_input_93_cast_fp16)[name = tensor<string, []>("op_545_cast_fp16")];
+            tensor<fp16, []> _inversed_input_y_0_to_fp16 = const()[name = tensor<string, []>("_inversed_input_y_0_to_fp16"), val = tensor<fp16, []>(0x1.6ap-1)];
+            tensor<fp16, [1, 256, 4000]> _inversed_input_cast_fp16 = mul(x = var_545_cast_fp16, y = _inversed_input_y_0_to_fp16)[name = tensor<string, []>("_inversed_input_cast_fp16")];
+            tensor<string, []> var_558_pad_type_0 = const()[name = tensor<string, []>("op_558_pad_type_0"), val = tensor<string, []>("valid")];
+            tensor<int32, [1]> var_558_strides_0 = const()[name = tensor<string, []>("op_558_strides_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, [2]> var_558_pad_0 = const()[name = tensor<string, []>("op_558_pad_0"), val = tensor<int32, [2]>([0, 0])];
+            tensor<int32, [1]> var_558_dilations_0 = const()[name = tensor<string, []>("op_558_dilations_0"), val = tensor<int32, [1]>([1])];
+            tensor<int32, []> var_558_groups_0 = const()[name = tensor<string, []>("op_558_groups_0"), val = tensor<int32, []>(1)];
+            tensor<fp16, [1, 256, 1]> N_proj_weight_to_fp16 = const()[name = tensor<string, []>("N_proj_weight_to_fp16"), val = tensor<fp16, [1, 256, 1]>(BLOBFILE(path = tensor<string, []>("@model_path/weights/weight.bin"), offset = tensor<uint64, []>(8450368)))];
+            tensor<fp16, [1]> N_proj_bias_to_fp16 = const()[name = tensor<string, []>("N_proj_bias_to_fp16"), val = tensor<fp16, [1]>([0x1.6ep-4])];
+            tensor<fp16, [1, 1, 4000]> var_558_cast_fp16 = conv(bias = N_proj_bias_to_fp16, dilations = var_558_dilations_0, groups = var_558_groups_0, pad = var_558_pad_0, pad_type = var_558_pad_type_0, strides = var_558_strides_0, weight = N_proj_weight_to_fp16, x = _inversed_input_cast_fp16)[name = tensor<string, []>("op_558_cast_fp16")];
+            tensor<int32, [1]> var_560_axes_0 = const()[name = tensor<string, []>("op_560_axes_0"), val = tensor<int32, [1]>([1])];
+            tensor<fp16, [1, 4000]> N = squeeze(axes = var_560_axes_0, x = var_558_cast_fp16)[name = tensor<string, []>("op_560_cast_fp16")];
+        } -> (F0, N);
+}

ANE/styletts2_ane_prosody.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4242f6c193f8d1c995e9da98e63cda3f9559daaa9154be4d7a325c976d347ad8
+size 8450944

ANE/styletts2_ane_prosody.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c8bb395a1d0b34fceb6da4e42c4cd201fe6c4121f788475940abaeb7ccd9e6f
+size 65948

ANE/styletts2_ane_prosody.mlpackage/Data/com.apple.CoreML/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4242f6c193f8d1c995e9da98e63cda3f9559daaa9154be4d7a325c976d347ad8
+size 8450944

ANE/styletts2_ane_prosody.mlpackage/Manifest.json

@@ -0,0 +1,18 @@
+{
+    "fileFormatVersion": "1.0.0",
+    "itemInfoEntries": {
+        "5239C997-4433-48DE-B7F7-8F278231936E": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Weights",
+            "name": "weights",
+            "path": "com.apple.CoreML/weights"
+        },
+        "EAD2ABD5-31E9-4A23-A7AB-80FE05510307": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Specification",
+            "name": "model.mlmodel",
+            "path": "com.apple.CoreML/model.mlmodel"
+        }
+    },
+    "rootModelIdentifier": "EAD2ABD5-31E9-4A23-A7AB-80FE05510307"
+}

ANE/styletts2_ane_vocoder.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e00e88bd1cdc41d123c27fe2f02f15f06711c8f6a271a0e6f458a482f066a233
+size 243