Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- .DS_Store +0 -0
- DeepSeek_FFN_PF_lut6_chunk_01of08.mlmodelc/analytics/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_01of08.mlmodelc/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_01of08.mlmodelc/metadata.json +336 -0
- DeepSeek_FFN_PF_lut6_chunk_01of08.mlmodelc/model.mil +0 -0
- DeepSeek_FFN_PF_lut6_chunk_01of08.mlmodelc/weights/weight.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_02of08.mlmodelc/analytics/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_02of08.mlmodelc/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_02of08.mlmodelc/metadata.json +336 -0
- DeepSeek_FFN_PF_lut6_chunk_02of08.mlmodelc/model.mil +0 -0
- DeepSeek_FFN_PF_lut6_chunk_02of08.mlmodelc/weights/weight.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_03of08.mlmodelc/analytics/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_03of08.mlmodelc/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_03of08.mlmodelc/metadata.json +336 -0
- DeepSeek_FFN_PF_lut6_chunk_03of08.mlmodelc/model.mil +0 -0
- DeepSeek_FFN_PF_lut6_chunk_03of08.mlmodelc/weights/weight.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_04of08.mlmodelc/analytics/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_04of08.mlmodelc/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_04of08.mlmodelc/metadata.json +336 -0
- DeepSeek_FFN_PF_lut6_chunk_04of08.mlmodelc/model.mil +0 -0
- DeepSeek_FFN_PF_lut6_chunk_04of08.mlmodelc/weights/weight.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_05of08.mlmodelc/analytics/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_05of08.mlmodelc/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_05of08.mlmodelc/metadata.json +336 -0
- DeepSeek_FFN_PF_lut6_chunk_05of08.mlmodelc/model.mil +0 -0
- DeepSeek_FFN_PF_lut6_chunk_05of08.mlmodelc/weights/weight.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_06of08.mlmodelc/analytics/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_06of08.mlmodelc/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_06of08.mlmodelc/metadata.json +336 -0
- DeepSeek_FFN_PF_lut6_chunk_06of08.mlmodelc/model.mil +0 -0
- DeepSeek_FFN_PF_lut6_chunk_06of08.mlmodelc/weights/weight.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_07of08.mlmodelc/analytics/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_07of08.mlmodelc/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_07of08.mlmodelc/metadata.json +336 -0
- DeepSeek_FFN_PF_lut6_chunk_07of08.mlmodelc/model.mil +0 -0
- DeepSeek_FFN_PF_lut6_chunk_07of08.mlmodelc/weights/weight.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_08of08.mlmodelc/analytics/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_08of08.mlmodelc/coremldata.bin +3 -0
- DeepSeek_FFN_PF_lut6_chunk_08of08.mlmodelc/metadata.json +336 -0
- DeepSeek_FFN_PF_lut6_chunk_08of08.mlmodelc/model.mil +0 -0
- DeepSeek_FFN_PF_lut6_chunk_08of08.mlmodelc/weights/weight.bin +3 -0
- DeepSeek_embeddings.mlmodelc/analytics/coremldata.bin +3 -0
- DeepSeek_embeddings.mlmodelc/coremldata.bin +3 -0
- DeepSeek_embeddings.mlmodelc/metadata.json +67 -0
- DeepSeek_embeddings.mlmodelc/model.mil +11 -0
- DeepSeek_embeddings.mlmodelc/weights/weight.bin +3 -0
- DeepSeek_lm_head_lut6.mlmodelc/analytics/coremldata.bin +3 -0
- DeepSeek_lm_head_lut6.mlmodelc/coremldata.bin +3 -0
- DeepSeek_lm_head_lut6.mlmodelc/metadata.json +139 -0
- DeepSeek_lm_head_lut6.mlmodelc/model.mil +98 -0
.DS_Store
ADDED
|
Binary file (12.3 kB). View file
|
|
|
DeepSeek_FFN_PF_lut6_chunk_01of08.mlmodelc/analytics/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e6d0c648a227f89537451cc9d08494aecb42bfadcdbc5ed1239cfe12195ce6cf
|
| 3 |
+
size 243
|
DeepSeek_FFN_PF_lut6_chunk_01of08.mlmodelc/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d5f0d403305809f6171f0c2ae5b07f5398a78fd66ff5325f295bd658709f133
|
| 3 |
+
size 986
|
DeepSeek_FFN_PF_lut6_chunk_01of08.mlmodelc/metadata.json
ADDED
|
@@ -0,0 +1,336 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"metadataOutputVersion" : "3.0",
|
| 4 |
+
"userDefinedMetadata" : {
|
| 5 |
+
"com.github.apple.coremltools.source" : "torch==2.5.0",
|
| 6 |
+
"com.github.apple.coremltools.version" : "8.2",
|
| 7 |
+
"com.anemll.context_length" : "1024",
|
| 8 |
+
"com.anemll.chunk_no" : "1",
|
| 9 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
| 10 |
+
"com.anemll.num_chunks" : "8",
|
| 11 |
+
"com.anemll.info" : "Converted with Anemll v0.2.0",
|
| 12 |
+
"com.anemll.batch_size" : "256",
|
| 13 |
+
"com.anemll.lut_bits" : "6"
|
| 14 |
+
},
|
| 15 |
+
"availability" : {
|
| 16 |
+
"macOS" : "15.0",
|
| 17 |
+
"tvOS" : "18.0",
|
| 18 |
+
"visionOS" : "2.0",
|
| 19 |
+
"watchOS" : "11.0",
|
| 20 |
+
"iOS" : "18.0",
|
| 21 |
+
"macCatalyst" : "18.0"
|
| 22 |
+
},
|
| 23 |
+
"inputSchema" : [
|
| 24 |
+
{
|
| 25 |
+
"hasShapeFlexibility" : "0",
|
| 26 |
+
"isOptional" : "0",
|
| 27 |
+
"dataType" : "Float16",
|
| 28 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 29 |
+
"shortDescription" : "",
|
| 30 |
+
"shape" : "[1, 1, 4096]",
|
| 31 |
+
"name" : "hidden_states",
|
| 32 |
+
"type" : "MultiArray"
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"hasShapeFlexibility" : "0",
|
| 36 |
+
"isOptional" : "0",
|
| 37 |
+
"dataType" : "Int32",
|
| 38 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 39 |
+
"shortDescription" : "",
|
| 40 |
+
"shape" : "[1]",
|
| 41 |
+
"name" : "position_ids",
|
| 42 |
+
"type" : "MultiArray"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"hasShapeFlexibility" : "0",
|
| 46 |
+
"isOptional" : "0",
|
| 47 |
+
"dataType" : "Float16",
|
| 48 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 49 |
+
"shortDescription" : "",
|
| 50 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 51 |
+
"name" : "causal_mask",
|
| 52 |
+
"type" : "MultiArray"
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"hasShapeFlexibility" : "0",
|
| 56 |
+
"isOptional" : "0",
|
| 57 |
+
"dataType" : "Int32",
|
| 58 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 59 |
+
"shortDescription" : "",
|
| 60 |
+
"shape" : "[1]",
|
| 61 |
+
"name" : "current_pos",
|
| 62 |
+
"type" : "MultiArray"
|
| 63 |
+
}
|
| 64 |
+
],
|
| 65 |
+
"outputSchema" : [
|
| 66 |
+
{
|
| 67 |
+
"hasShapeFlexibility" : "0",
|
| 68 |
+
"isOptional" : "0",
|
| 69 |
+
"dataType" : "Float16",
|
| 70 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 71 |
+
"shortDescription" : "",
|
| 72 |
+
"shape" : "[1, 1, 4096]",
|
| 73 |
+
"name" : "output_hidden_states",
|
| 74 |
+
"type" : "MultiArray"
|
| 75 |
+
}
|
| 76 |
+
],
|
| 77 |
+
"modelParameters" : [
|
| 78 |
+
|
| 79 |
+
],
|
| 80 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 81 |
+
"method" : "predict",
|
| 82 |
+
"functions" : [
|
| 83 |
+
{
|
| 84 |
+
"inputSchema" : [
|
| 85 |
+
{
|
| 86 |
+
"hasShapeFlexibility" : "0",
|
| 87 |
+
"isOptional" : "0",
|
| 88 |
+
"dataType" : "Float16",
|
| 89 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 90 |
+
"shortDescription" : "",
|
| 91 |
+
"shape" : "[1, 1, 4096]",
|
| 92 |
+
"name" : "hidden_states",
|
| 93 |
+
"type" : "MultiArray"
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"hasShapeFlexibility" : "0",
|
| 97 |
+
"isOptional" : "0",
|
| 98 |
+
"dataType" : "Int32",
|
| 99 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 100 |
+
"shortDescription" : "",
|
| 101 |
+
"shape" : "[1]",
|
| 102 |
+
"name" : "position_ids",
|
| 103 |
+
"type" : "MultiArray"
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"hasShapeFlexibility" : "0",
|
| 107 |
+
"isOptional" : "0",
|
| 108 |
+
"dataType" : "Float16",
|
| 109 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 110 |
+
"shortDescription" : "",
|
| 111 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 112 |
+
"name" : "causal_mask",
|
| 113 |
+
"type" : "MultiArray"
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"hasShapeFlexibility" : "0",
|
| 117 |
+
"isOptional" : "0",
|
| 118 |
+
"dataType" : "Int32",
|
| 119 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 120 |
+
"shortDescription" : "",
|
| 121 |
+
"shape" : "[1]",
|
| 122 |
+
"name" : "current_pos",
|
| 123 |
+
"type" : "MultiArray"
|
| 124 |
+
}
|
| 125 |
+
],
|
| 126 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 127 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 128 |
+
"stateSchema" : [
|
| 129 |
+
{
|
| 130 |
+
"dataType" : "Float16",
|
| 131 |
+
"isOptional" : "0",
|
| 132 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 133 |
+
"shortDescription" : "",
|
| 134 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 135 |
+
"name" : "model_model_kv_cache_0",
|
| 136 |
+
"type" : "State"
|
| 137 |
+
}
|
| 138 |
+
],
|
| 139 |
+
"outputSchema" : [
|
| 140 |
+
{
|
| 141 |
+
"hasShapeFlexibility" : "0",
|
| 142 |
+
"isOptional" : "0",
|
| 143 |
+
"dataType" : "Float16",
|
| 144 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 145 |
+
"shortDescription" : "",
|
| 146 |
+
"shape" : "[1, 1, 4096]",
|
| 147 |
+
"name" : "output_hidden_states",
|
| 148 |
+
"type" : "MultiArray"
|
| 149 |
+
}
|
| 150 |
+
],
|
| 151 |
+
"name" : "infer",
|
| 152 |
+
"mlProgramOperationTypeHistogram" : {
|
| 153 |
+
"Ios18.expandDims" : 16,
|
| 154 |
+
"Ios18.mul" : 40,
|
| 155 |
+
"Ios18.matmul" : 8,
|
| 156 |
+
"Identity" : 1,
|
| 157 |
+
"Ios16.reduceMean" : 8,
|
| 158 |
+
"Ios18.exp" : 4,
|
| 159 |
+
"Ios18.realDiv" : 4,
|
| 160 |
+
"Ios18.greaterEqual" : 1,
|
| 161 |
+
"Select" : 1,
|
| 162 |
+
"Ios18.readState" : 9,
|
| 163 |
+
"Tile" : 8,
|
| 164 |
+
"Ios18.gather" : 2,
|
| 165 |
+
"Ios18.add" : 22,
|
| 166 |
+
"Ios18.layerNorm" : 8,
|
| 167 |
+
"Ios18.sliceUpdate" : 8,
|
| 168 |
+
"Ios18.writeState" : 8,
|
| 169 |
+
"Ios18.reshape" : 26,
|
| 170 |
+
"Ios16.reduceMax" : 4,
|
| 171 |
+
"Ios16.reduceSum" : 4,
|
| 172 |
+
"Ios18.constexprLutToDense" : 28,
|
| 173 |
+
"Ios18.conv" : 24,
|
| 174 |
+
"Ios18.concat" : 24,
|
| 175 |
+
"Ios18.transpose" : 16,
|
| 176 |
+
"Ios18.sub" : 20,
|
| 177 |
+
"Ios18.linear" : 4,
|
| 178 |
+
"Ios18.silu" : 4,
|
| 179 |
+
"Ios18.sliceByIndex" : 26,
|
| 180 |
+
"Ios18.squeeze" : 12
|
| 181 |
+
}
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"inputSchema" : [
|
| 185 |
+
{
|
| 186 |
+
"hasShapeFlexibility" : "0",
|
| 187 |
+
"isOptional" : "0",
|
| 188 |
+
"dataType" : "Float16",
|
| 189 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 190 |
+
"shortDescription" : "",
|
| 191 |
+
"shape" : "[1, 256, 4096]",
|
| 192 |
+
"name" : "hidden_states",
|
| 193 |
+
"type" : "MultiArray"
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"hasShapeFlexibility" : "0",
|
| 197 |
+
"isOptional" : "0",
|
| 198 |
+
"dataType" : "Int32",
|
| 199 |
+
"formattedType" : "MultiArray (Int32 256)",
|
| 200 |
+
"shortDescription" : "",
|
| 201 |
+
"shape" : "[256]",
|
| 202 |
+
"name" : "position_ids",
|
| 203 |
+
"type" : "MultiArray"
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"hasShapeFlexibility" : "0",
|
| 207 |
+
"isOptional" : "0",
|
| 208 |
+
"dataType" : "Float16",
|
| 209 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 256 × 1024)",
|
| 210 |
+
"shortDescription" : "",
|
| 211 |
+
"shape" : "[1, 1, 256, 1024]",
|
| 212 |
+
"name" : "causal_mask",
|
| 213 |
+
"type" : "MultiArray"
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"hasShapeFlexibility" : "0",
|
| 217 |
+
"isOptional" : "0",
|
| 218 |
+
"dataType" : "Int32",
|
| 219 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 220 |
+
"shortDescription" : "",
|
| 221 |
+
"shape" : "[1]",
|
| 222 |
+
"name" : "current_pos",
|
| 223 |
+
"type" : "MultiArray"
|
| 224 |
+
}
|
| 225 |
+
],
|
| 226 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 227 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 228 |
+
"stateSchema" : [
|
| 229 |
+
{
|
| 230 |
+
"dataType" : "Float16",
|
| 231 |
+
"isOptional" : "0",
|
| 232 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 233 |
+
"shortDescription" : "",
|
| 234 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 235 |
+
"name" : "model_model_kv_cache_0",
|
| 236 |
+
"type" : "State"
|
| 237 |
+
}
|
| 238 |
+
],
|
| 239 |
+
"outputSchema" : [
|
| 240 |
+
{
|
| 241 |
+
"hasShapeFlexibility" : "0",
|
| 242 |
+
"isOptional" : "0",
|
| 243 |
+
"dataType" : "Float16",
|
| 244 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 245 |
+
"shortDescription" : "",
|
| 246 |
+
"shape" : "[1, 256, 4096]",
|
| 247 |
+
"name" : "output_hidden_states",
|
| 248 |
+
"type" : "MultiArray"
|
| 249 |
+
}
|
| 250 |
+
],
|
| 251 |
+
"name" : "prefill",
|
| 252 |
+
"mlProgramOperationTypeHistogram" : {
|
| 253 |
+
"Ios18.expandDims" : 16,
|
| 254 |
+
"Ios18.mul" : 40,
|
| 255 |
+
"Ios18.matmul" : 8,
|
| 256 |
+
"Ios16.reduceMean" : 8,
|
| 257 |
+
"Ios18.exp" : 4,
|
| 258 |
+
"Ios18.realDiv" : 4,
|
| 259 |
+
"Ios18.greaterEqual" : 1,
|
| 260 |
+
"Select" : 1,
|
| 261 |
+
"Ios18.readState" : 9,
|
| 262 |
+
"Tile" : 8,
|
| 263 |
+
"Ios18.gather" : 2,
|
| 264 |
+
"Ios18.add" : 22,
|
| 265 |
+
"Ios18.layerNorm" : 8,
|
| 266 |
+
"Ios18.sliceUpdate" : 8,
|
| 267 |
+
"Ios18.writeState" : 8,
|
| 268 |
+
"Ios18.reshape" : 34,
|
| 269 |
+
"Ios16.reduceMax" : 4,
|
| 270 |
+
"Ios16.reduceSum" : 4,
|
| 271 |
+
"Ios18.constexprLutToDense" : 28,
|
| 272 |
+
"Ios18.conv" : 24,
|
| 273 |
+
"Ios18.concat" : 24,
|
| 274 |
+
"Ios18.transpose" : 30,
|
| 275 |
+
"Ios18.sub" : 20,
|
| 276 |
+
"Ios18.linear" : 4,
|
| 277 |
+
"Ios18.silu" : 4,
|
| 278 |
+
"Ios18.sliceByIndex" : 26,
|
| 279 |
+
"Ios18.squeeze" : 12
|
| 280 |
+
}
|
| 281 |
+
}
|
| 282 |
+
],
|
| 283 |
+
"version" : "0.2.0",
|
| 284 |
+
"isUpdatable" : "0",
|
| 285 |
+
"defaultFunctionName" : "infer",
|
| 286 |
+
"specificationVersion" : 9,
|
| 287 |
+
"stateSchema" : [
|
| 288 |
+
{
|
| 289 |
+
"dataType" : "Float16",
|
| 290 |
+
"isOptional" : "0",
|
| 291 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 292 |
+
"shortDescription" : "",
|
| 293 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 294 |
+
"name" : "model_model_kv_cache_0",
|
| 295 |
+
"type" : "State"
|
| 296 |
+
}
|
| 297 |
+
],
|
| 298 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 299 |
+
"mlProgramOperationTypeHistogram" : {
|
| 300 |
+
"Ios18.expandDims" : 16,
|
| 301 |
+
"Ios18.mul" : 40,
|
| 302 |
+
"Ios18.matmul" : 8,
|
| 303 |
+
"Identity" : 1,
|
| 304 |
+
"Ios16.reduceMean" : 8,
|
| 305 |
+
"Ios18.exp" : 4,
|
| 306 |
+
"Ios18.realDiv" : 4,
|
| 307 |
+
"Ios18.greaterEqual" : 1,
|
| 308 |
+
"Select" : 1,
|
| 309 |
+
"Ios18.readState" : 9,
|
| 310 |
+
"Tile" : 8,
|
| 311 |
+
"Ios18.gather" : 2,
|
| 312 |
+
"Ios18.add" : 22,
|
| 313 |
+
"Ios18.layerNorm" : 8,
|
| 314 |
+
"Ios18.sliceUpdate" : 8,
|
| 315 |
+
"Ios18.writeState" : 8,
|
| 316 |
+
"Ios18.reshape" : 26,
|
| 317 |
+
"Ios16.reduceMax" : 4,
|
| 318 |
+
"Ios16.reduceSum" : 4,
|
| 319 |
+
"Ios18.constexprLutToDense" : 28,
|
| 320 |
+
"Ios18.conv" : 24,
|
| 321 |
+
"Ios18.concat" : 24,
|
| 322 |
+
"Ios18.transpose" : 16,
|
| 323 |
+
"Ios18.sub" : 20,
|
| 324 |
+
"Ios18.linear" : 4,
|
| 325 |
+
"Ios18.silu" : 4,
|
| 326 |
+
"Ios18.sliceByIndex" : 26,
|
| 327 |
+
"Ios18.squeeze" : 12
|
| 328 |
+
},
|
| 329 |
+
"shortDescription" : "Anemll Model: Multifunction FFN+Prefill",
|
| 330 |
+
"generatedClassName" : "DeepSeek_FFN_PF_lut6_chunk_01of08",
|
| 331 |
+
"author" : "Converted with Anemll v0.2.0",
|
| 332 |
+
"modelType" : {
|
| 333 |
+
"name" : "MLModelType_mlProgram"
|
| 334 |
+
}
|
| 335 |
+
}
|
| 336 |
+
]
|
DeepSeek_FFN_PF_lut6_chunk_01of08.mlmodelc/model.mil
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
DeepSeek_FFN_PF_lut6_chunk_01of08.mlmodelc/weights/weight.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8e6008c782dfa4d52873aa42cd92c307664f9accefc1ac9a4b0c6ab11b0ab660
|
| 3 |
+
size 724250880
|
DeepSeek_FFN_PF_lut6_chunk_02of08.mlmodelc/analytics/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2431f39b018f69db743ec14adf09e391c5906fdd689962cee552e03916a82b97
|
| 3 |
+
size 243
|
DeepSeek_FFN_PF_lut6_chunk_02of08.mlmodelc/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3fef0e796e819264297adbdad9f45821d9a78ac1a090f0f265bdd6d58bf65548
|
| 3 |
+
size 986
|
DeepSeek_FFN_PF_lut6_chunk_02of08.mlmodelc/metadata.json
ADDED
|
@@ -0,0 +1,336 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"metadataOutputVersion" : "3.0",
|
| 4 |
+
"userDefinedMetadata" : {
|
| 5 |
+
"com.anemll.lut_bits" : "6",
|
| 6 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
| 7 |
+
"com.anemll.context_length" : "1024",
|
| 8 |
+
"com.github.apple.coremltools.source" : "torch==2.5.0",
|
| 9 |
+
"com.github.apple.coremltools.version" : "8.2",
|
| 10 |
+
"com.anemll.num_chunks" : "8",
|
| 11 |
+
"com.anemll.batch_size" : "256",
|
| 12 |
+
"com.anemll.info" : "Converted with Anemll v0.2.0",
|
| 13 |
+
"com.anemll.chunk_no" : "2"
|
| 14 |
+
},
|
| 15 |
+
"availability" : {
|
| 16 |
+
"macOS" : "15.0",
|
| 17 |
+
"tvOS" : "18.0",
|
| 18 |
+
"visionOS" : "2.0",
|
| 19 |
+
"watchOS" : "11.0",
|
| 20 |
+
"iOS" : "18.0",
|
| 21 |
+
"macCatalyst" : "18.0"
|
| 22 |
+
},
|
| 23 |
+
"inputSchema" : [
|
| 24 |
+
{
|
| 25 |
+
"hasShapeFlexibility" : "0",
|
| 26 |
+
"isOptional" : "0",
|
| 27 |
+
"dataType" : "Float16",
|
| 28 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 29 |
+
"shortDescription" : "",
|
| 30 |
+
"shape" : "[1, 1, 4096]",
|
| 31 |
+
"name" : "hidden_states",
|
| 32 |
+
"type" : "MultiArray"
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"hasShapeFlexibility" : "0",
|
| 36 |
+
"isOptional" : "0",
|
| 37 |
+
"dataType" : "Int32",
|
| 38 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 39 |
+
"shortDescription" : "",
|
| 40 |
+
"shape" : "[1]",
|
| 41 |
+
"name" : "position_ids",
|
| 42 |
+
"type" : "MultiArray"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"hasShapeFlexibility" : "0",
|
| 46 |
+
"isOptional" : "0",
|
| 47 |
+
"dataType" : "Float16",
|
| 48 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 49 |
+
"shortDescription" : "",
|
| 50 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 51 |
+
"name" : "causal_mask",
|
| 52 |
+
"type" : "MultiArray"
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"hasShapeFlexibility" : "0",
|
| 56 |
+
"isOptional" : "0",
|
| 57 |
+
"dataType" : "Int32",
|
| 58 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 59 |
+
"shortDescription" : "",
|
| 60 |
+
"shape" : "[1]",
|
| 61 |
+
"name" : "current_pos",
|
| 62 |
+
"type" : "MultiArray"
|
| 63 |
+
}
|
| 64 |
+
],
|
| 65 |
+
"outputSchema" : [
|
| 66 |
+
{
|
| 67 |
+
"hasShapeFlexibility" : "0",
|
| 68 |
+
"isOptional" : "0",
|
| 69 |
+
"dataType" : "Float16",
|
| 70 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 71 |
+
"shortDescription" : "",
|
| 72 |
+
"shape" : "[1, 1, 4096]",
|
| 73 |
+
"name" : "output_hidden_states",
|
| 74 |
+
"type" : "MultiArray"
|
| 75 |
+
}
|
| 76 |
+
],
|
| 77 |
+
"modelParameters" : [
|
| 78 |
+
|
| 79 |
+
],
|
| 80 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 81 |
+
"method" : "predict",
|
| 82 |
+
"functions" : [
|
| 83 |
+
{
|
| 84 |
+
"inputSchema" : [
|
| 85 |
+
{
|
| 86 |
+
"hasShapeFlexibility" : "0",
|
| 87 |
+
"isOptional" : "0",
|
| 88 |
+
"dataType" : "Float16",
|
| 89 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 90 |
+
"shortDescription" : "",
|
| 91 |
+
"shape" : "[1, 1, 4096]",
|
| 92 |
+
"name" : "hidden_states",
|
| 93 |
+
"type" : "MultiArray"
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"hasShapeFlexibility" : "0",
|
| 97 |
+
"isOptional" : "0",
|
| 98 |
+
"dataType" : "Int32",
|
| 99 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 100 |
+
"shortDescription" : "",
|
| 101 |
+
"shape" : "[1]",
|
| 102 |
+
"name" : "position_ids",
|
| 103 |
+
"type" : "MultiArray"
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"hasShapeFlexibility" : "0",
|
| 107 |
+
"isOptional" : "0",
|
| 108 |
+
"dataType" : "Float16",
|
| 109 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 110 |
+
"shortDescription" : "",
|
| 111 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 112 |
+
"name" : "causal_mask",
|
| 113 |
+
"type" : "MultiArray"
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"hasShapeFlexibility" : "0",
|
| 117 |
+
"isOptional" : "0",
|
| 118 |
+
"dataType" : "Int32",
|
| 119 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 120 |
+
"shortDescription" : "",
|
| 121 |
+
"shape" : "[1]",
|
| 122 |
+
"name" : "current_pos",
|
| 123 |
+
"type" : "MultiArray"
|
| 124 |
+
}
|
| 125 |
+
],
|
| 126 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 127 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 128 |
+
"stateSchema" : [
|
| 129 |
+
{
|
| 130 |
+
"dataType" : "Float16",
|
| 131 |
+
"isOptional" : "0",
|
| 132 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 133 |
+
"shortDescription" : "",
|
| 134 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 135 |
+
"name" : "model_model_kv_cache_0",
|
| 136 |
+
"type" : "State"
|
| 137 |
+
}
|
| 138 |
+
],
|
| 139 |
+
"outputSchema" : [
|
| 140 |
+
{
|
| 141 |
+
"hasShapeFlexibility" : "0",
|
| 142 |
+
"isOptional" : "0",
|
| 143 |
+
"dataType" : "Float16",
|
| 144 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 145 |
+
"shortDescription" : "",
|
| 146 |
+
"shape" : "[1, 1, 4096]",
|
| 147 |
+
"name" : "output_hidden_states",
|
| 148 |
+
"type" : "MultiArray"
|
| 149 |
+
}
|
| 150 |
+
],
|
| 151 |
+
"name" : "infer",
|
| 152 |
+
"mlProgramOperationTypeHistogram" : {
|
| 153 |
+
"Ios18.expandDims" : 16,
|
| 154 |
+
"Ios18.mul" : 40,
|
| 155 |
+
"Ios18.matmul" : 8,
|
| 156 |
+
"Identity" : 1,
|
| 157 |
+
"Ios16.reduceMean" : 8,
|
| 158 |
+
"Ios18.exp" : 4,
|
| 159 |
+
"Ios18.realDiv" : 4,
|
| 160 |
+
"Ios18.greaterEqual" : 1,
|
| 161 |
+
"Select" : 1,
|
| 162 |
+
"Ios18.readState" : 9,
|
| 163 |
+
"Tile" : 8,
|
| 164 |
+
"Ios18.gather" : 2,
|
| 165 |
+
"Ios18.add" : 22,
|
| 166 |
+
"Ios18.layerNorm" : 8,
|
| 167 |
+
"Ios18.sliceUpdate" : 8,
|
| 168 |
+
"Ios18.writeState" : 8,
|
| 169 |
+
"Ios18.reshape" : 26,
|
| 170 |
+
"Ios16.reduceMax" : 4,
|
| 171 |
+
"Ios16.reduceSum" : 4,
|
| 172 |
+
"Ios18.constexprLutToDense" : 28,
|
| 173 |
+
"Ios18.conv" : 24,
|
| 174 |
+
"Ios18.concat" : 24,
|
| 175 |
+
"Ios18.transpose" : 16,
|
| 176 |
+
"Ios18.sub" : 20,
|
| 177 |
+
"Ios18.linear" : 4,
|
| 178 |
+
"Ios18.silu" : 4,
|
| 179 |
+
"Ios18.sliceByIndex" : 26,
|
| 180 |
+
"Ios18.squeeze" : 12
|
| 181 |
+
}
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"inputSchema" : [
|
| 185 |
+
{
|
| 186 |
+
"hasShapeFlexibility" : "0",
|
| 187 |
+
"isOptional" : "0",
|
| 188 |
+
"dataType" : "Float16",
|
| 189 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 190 |
+
"shortDescription" : "",
|
| 191 |
+
"shape" : "[1, 256, 4096]",
|
| 192 |
+
"name" : "hidden_states",
|
| 193 |
+
"type" : "MultiArray"
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"hasShapeFlexibility" : "0",
|
| 197 |
+
"isOptional" : "0",
|
| 198 |
+
"dataType" : "Int32",
|
| 199 |
+
"formattedType" : "MultiArray (Int32 256)",
|
| 200 |
+
"shortDescription" : "",
|
| 201 |
+
"shape" : "[256]",
|
| 202 |
+
"name" : "position_ids",
|
| 203 |
+
"type" : "MultiArray"
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"hasShapeFlexibility" : "0",
|
| 207 |
+
"isOptional" : "0",
|
| 208 |
+
"dataType" : "Float16",
|
| 209 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 256 × 1024)",
|
| 210 |
+
"shortDescription" : "",
|
| 211 |
+
"shape" : "[1, 1, 256, 1024]",
|
| 212 |
+
"name" : "causal_mask",
|
| 213 |
+
"type" : "MultiArray"
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"hasShapeFlexibility" : "0",
|
| 217 |
+
"isOptional" : "0",
|
| 218 |
+
"dataType" : "Int32",
|
| 219 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 220 |
+
"shortDescription" : "",
|
| 221 |
+
"shape" : "[1]",
|
| 222 |
+
"name" : "current_pos",
|
| 223 |
+
"type" : "MultiArray"
|
| 224 |
+
}
|
| 225 |
+
],
|
| 226 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 227 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 228 |
+
"stateSchema" : [
|
| 229 |
+
{
|
| 230 |
+
"dataType" : "Float16",
|
| 231 |
+
"isOptional" : "0",
|
| 232 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 233 |
+
"shortDescription" : "",
|
| 234 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 235 |
+
"name" : "model_model_kv_cache_0",
|
| 236 |
+
"type" : "State"
|
| 237 |
+
}
|
| 238 |
+
],
|
| 239 |
+
"outputSchema" : [
|
| 240 |
+
{
|
| 241 |
+
"hasShapeFlexibility" : "0",
|
| 242 |
+
"isOptional" : "0",
|
| 243 |
+
"dataType" : "Float16",
|
| 244 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 245 |
+
"shortDescription" : "",
|
| 246 |
+
"shape" : "[1, 256, 4096]",
|
| 247 |
+
"name" : "output_hidden_states",
|
| 248 |
+
"type" : "MultiArray"
|
| 249 |
+
}
|
| 250 |
+
],
|
| 251 |
+
"name" : "prefill",
|
| 252 |
+
"mlProgramOperationTypeHistogram" : {
|
| 253 |
+
"Ios18.expandDims" : 16,
|
| 254 |
+
"Ios18.mul" : 40,
|
| 255 |
+
"Ios18.matmul" : 8,
|
| 256 |
+
"Ios16.reduceMean" : 8,
|
| 257 |
+
"Ios18.exp" : 4,
|
| 258 |
+
"Ios18.realDiv" : 4,
|
| 259 |
+
"Ios18.greaterEqual" : 1,
|
| 260 |
+
"Select" : 1,
|
| 261 |
+
"Ios18.readState" : 9,
|
| 262 |
+
"Tile" : 8,
|
| 263 |
+
"Ios18.gather" : 2,
|
| 264 |
+
"Ios18.add" : 22,
|
| 265 |
+
"Ios18.layerNorm" : 8,
|
| 266 |
+
"Ios18.sliceUpdate" : 8,
|
| 267 |
+
"Ios18.writeState" : 8,
|
| 268 |
+
"Ios18.reshape" : 34,
|
| 269 |
+
"Ios16.reduceMax" : 4,
|
| 270 |
+
"Ios16.reduceSum" : 4,
|
| 271 |
+
"Ios18.constexprLutToDense" : 28,
|
| 272 |
+
"Ios18.conv" : 24,
|
| 273 |
+
"Ios18.concat" : 24,
|
| 274 |
+
"Ios18.transpose" : 30,
|
| 275 |
+
"Ios18.sub" : 20,
|
| 276 |
+
"Ios18.linear" : 4,
|
| 277 |
+
"Ios18.silu" : 4,
|
| 278 |
+
"Ios18.sliceByIndex" : 26,
|
| 279 |
+
"Ios18.squeeze" : 12
|
| 280 |
+
}
|
| 281 |
+
}
|
| 282 |
+
],
|
| 283 |
+
"version" : "0.2.0",
|
| 284 |
+
"isUpdatable" : "0",
|
| 285 |
+
"defaultFunctionName" : "infer",
|
| 286 |
+
"specificationVersion" : 9,
|
| 287 |
+
"stateSchema" : [
|
| 288 |
+
{
|
| 289 |
+
"dataType" : "Float16",
|
| 290 |
+
"isOptional" : "0",
|
| 291 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 292 |
+
"shortDescription" : "",
|
| 293 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 294 |
+
"name" : "model_model_kv_cache_0",
|
| 295 |
+
"type" : "State"
|
| 296 |
+
}
|
| 297 |
+
],
|
| 298 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 299 |
+
"mlProgramOperationTypeHistogram" : {
|
| 300 |
+
"Ios18.expandDims" : 16,
|
| 301 |
+
"Ios18.mul" : 40,
|
| 302 |
+
"Ios18.matmul" : 8,
|
| 303 |
+
"Identity" : 1,
|
| 304 |
+
"Ios16.reduceMean" : 8,
|
| 305 |
+
"Ios18.exp" : 4,
|
| 306 |
+
"Ios18.realDiv" : 4,
|
| 307 |
+
"Ios18.greaterEqual" : 1,
|
| 308 |
+
"Select" : 1,
|
| 309 |
+
"Ios18.readState" : 9,
|
| 310 |
+
"Tile" : 8,
|
| 311 |
+
"Ios18.gather" : 2,
|
| 312 |
+
"Ios18.add" : 22,
|
| 313 |
+
"Ios18.layerNorm" : 8,
|
| 314 |
+
"Ios18.sliceUpdate" : 8,
|
| 315 |
+
"Ios18.writeState" : 8,
|
| 316 |
+
"Ios18.reshape" : 26,
|
| 317 |
+
"Ios16.reduceMax" : 4,
|
| 318 |
+
"Ios16.reduceSum" : 4,
|
| 319 |
+
"Ios18.constexprLutToDense" : 28,
|
| 320 |
+
"Ios18.conv" : 24,
|
| 321 |
+
"Ios18.concat" : 24,
|
| 322 |
+
"Ios18.transpose" : 16,
|
| 323 |
+
"Ios18.sub" : 20,
|
| 324 |
+
"Ios18.linear" : 4,
|
| 325 |
+
"Ios18.silu" : 4,
|
| 326 |
+
"Ios18.sliceByIndex" : 26,
|
| 327 |
+
"Ios18.squeeze" : 12
|
| 328 |
+
},
|
| 329 |
+
"shortDescription" : "Anemll Model: Multifunction FFN+Prefill",
|
| 330 |
+
"generatedClassName" : "DeepSeek_FFN_PF_lut6_chunk_02of08",
|
| 331 |
+
"author" : "Converted with Anemll v0.2.0",
|
| 332 |
+
"modelType" : {
|
| 333 |
+
"name" : "MLModelType_mlProgram"
|
| 334 |
+
}
|
| 335 |
+
}
|
| 336 |
+
]
|
DeepSeek_FFN_PF_lut6_chunk_02of08.mlmodelc/model.mil
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
DeepSeek_FFN_PF_lut6_chunk_02of08.mlmodelc/weights/weight.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f623547faa68eb35416429a83901fc79abed89abe406259cb36991d5531c60e3
|
| 3 |
+
size 724250880
|
DeepSeek_FFN_PF_lut6_chunk_03of08.mlmodelc/analytics/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d6eb2beed48b0143756c361ada0f9fa5d2727ed57b0ceed598afbb79b4b86df
|
| 3 |
+
size 243
|
DeepSeek_FFN_PF_lut6_chunk_03of08.mlmodelc/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:93467c294dcd986ac246c1a81a4decc6f9f31c3fdf6406c16c2844dcc04d153d
|
| 3 |
+
size 986
|
DeepSeek_FFN_PF_lut6_chunk_03of08.mlmodelc/metadata.json
ADDED
|
@@ -0,0 +1,336 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"metadataOutputVersion" : "3.0",
|
| 4 |
+
"userDefinedMetadata" : {
|
| 5 |
+
"com.anemll.lut_bits" : "6",
|
| 6 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
| 7 |
+
"com.anemll.context_length" : "1024",
|
| 8 |
+
"com.github.apple.coremltools.source" : "torch==2.5.0",
|
| 9 |
+
"com.github.apple.coremltools.version" : "8.2",
|
| 10 |
+
"com.anemll.num_chunks" : "8",
|
| 11 |
+
"com.anemll.batch_size" : "256",
|
| 12 |
+
"com.anemll.info" : "Converted with Anemll v0.2.0",
|
| 13 |
+
"com.anemll.chunk_no" : "3"
|
| 14 |
+
},
|
| 15 |
+
"availability" : {
|
| 16 |
+
"macOS" : "15.0",
|
| 17 |
+
"tvOS" : "18.0",
|
| 18 |
+
"visionOS" : "2.0",
|
| 19 |
+
"watchOS" : "11.0",
|
| 20 |
+
"iOS" : "18.0",
|
| 21 |
+
"macCatalyst" : "18.0"
|
| 22 |
+
},
|
| 23 |
+
"inputSchema" : [
|
| 24 |
+
{
|
| 25 |
+
"hasShapeFlexibility" : "0",
|
| 26 |
+
"isOptional" : "0",
|
| 27 |
+
"dataType" : "Float16",
|
| 28 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 29 |
+
"shortDescription" : "",
|
| 30 |
+
"shape" : "[1, 1, 4096]",
|
| 31 |
+
"name" : "hidden_states",
|
| 32 |
+
"type" : "MultiArray"
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"hasShapeFlexibility" : "0",
|
| 36 |
+
"isOptional" : "0",
|
| 37 |
+
"dataType" : "Int32",
|
| 38 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 39 |
+
"shortDescription" : "",
|
| 40 |
+
"shape" : "[1]",
|
| 41 |
+
"name" : "position_ids",
|
| 42 |
+
"type" : "MultiArray"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"hasShapeFlexibility" : "0",
|
| 46 |
+
"isOptional" : "0",
|
| 47 |
+
"dataType" : "Float16",
|
| 48 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 49 |
+
"shortDescription" : "",
|
| 50 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 51 |
+
"name" : "causal_mask",
|
| 52 |
+
"type" : "MultiArray"
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"hasShapeFlexibility" : "0",
|
| 56 |
+
"isOptional" : "0",
|
| 57 |
+
"dataType" : "Int32",
|
| 58 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 59 |
+
"shortDescription" : "",
|
| 60 |
+
"shape" : "[1]",
|
| 61 |
+
"name" : "current_pos",
|
| 62 |
+
"type" : "MultiArray"
|
| 63 |
+
}
|
| 64 |
+
],
|
| 65 |
+
"outputSchema" : [
|
| 66 |
+
{
|
| 67 |
+
"hasShapeFlexibility" : "0",
|
| 68 |
+
"isOptional" : "0",
|
| 69 |
+
"dataType" : "Float16",
|
| 70 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 71 |
+
"shortDescription" : "",
|
| 72 |
+
"shape" : "[1, 1, 4096]",
|
| 73 |
+
"name" : "output_hidden_states",
|
| 74 |
+
"type" : "MultiArray"
|
| 75 |
+
}
|
| 76 |
+
],
|
| 77 |
+
"modelParameters" : [
|
| 78 |
+
|
| 79 |
+
],
|
| 80 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 81 |
+
"method" : "predict",
|
| 82 |
+
"functions" : [
|
| 83 |
+
{
|
| 84 |
+
"inputSchema" : [
|
| 85 |
+
{
|
| 86 |
+
"hasShapeFlexibility" : "0",
|
| 87 |
+
"isOptional" : "0",
|
| 88 |
+
"dataType" : "Float16",
|
| 89 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 90 |
+
"shortDescription" : "",
|
| 91 |
+
"shape" : "[1, 1, 4096]",
|
| 92 |
+
"name" : "hidden_states",
|
| 93 |
+
"type" : "MultiArray"
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"hasShapeFlexibility" : "0",
|
| 97 |
+
"isOptional" : "0",
|
| 98 |
+
"dataType" : "Int32",
|
| 99 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 100 |
+
"shortDescription" : "",
|
| 101 |
+
"shape" : "[1]",
|
| 102 |
+
"name" : "position_ids",
|
| 103 |
+
"type" : "MultiArray"
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"hasShapeFlexibility" : "0",
|
| 107 |
+
"isOptional" : "0",
|
| 108 |
+
"dataType" : "Float16",
|
| 109 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 110 |
+
"shortDescription" : "",
|
| 111 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 112 |
+
"name" : "causal_mask",
|
| 113 |
+
"type" : "MultiArray"
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"hasShapeFlexibility" : "0",
|
| 117 |
+
"isOptional" : "0",
|
| 118 |
+
"dataType" : "Int32",
|
| 119 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 120 |
+
"shortDescription" : "",
|
| 121 |
+
"shape" : "[1]",
|
| 122 |
+
"name" : "current_pos",
|
| 123 |
+
"type" : "MultiArray"
|
| 124 |
+
}
|
| 125 |
+
],
|
| 126 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 127 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 128 |
+
"stateSchema" : [
|
| 129 |
+
{
|
| 130 |
+
"dataType" : "Float16",
|
| 131 |
+
"isOptional" : "0",
|
| 132 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 133 |
+
"shortDescription" : "",
|
| 134 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 135 |
+
"name" : "model_model_kv_cache_0",
|
| 136 |
+
"type" : "State"
|
| 137 |
+
}
|
| 138 |
+
],
|
| 139 |
+
"outputSchema" : [
|
| 140 |
+
{
|
| 141 |
+
"hasShapeFlexibility" : "0",
|
| 142 |
+
"isOptional" : "0",
|
| 143 |
+
"dataType" : "Float16",
|
| 144 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 145 |
+
"shortDescription" : "",
|
| 146 |
+
"shape" : "[1, 1, 4096]",
|
| 147 |
+
"name" : "output_hidden_states",
|
| 148 |
+
"type" : "MultiArray"
|
| 149 |
+
}
|
| 150 |
+
],
|
| 151 |
+
"name" : "infer",
|
| 152 |
+
"mlProgramOperationTypeHistogram" : {
|
| 153 |
+
"Ios18.expandDims" : 16,
|
| 154 |
+
"Ios18.mul" : 40,
|
| 155 |
+
"Ios18.matmul" : 8,
|
| 156 |
+
"Identity" : 1,
|
| 157 |
+
"Ios16.reduceMean" : 8,
|
| 158 |
+
"Ios18.exp" : 4,
|
| 159 |
+
"Ios18.realDiv" : 4,
|
| 160 |
+
"Ios18.greaterEqual" : 1,
|
| 161 |
+
"Select" : 1,
|
| 162 |
+
"Ios18.readState" : 9,
|
| 163 |
+
"Tile" : 8,
|
| 164 |
+
"Ios18.gather" : 2,
|
| 165 |
+
"Ios18.add" : 22,
|
| 166 |
+
"Ios18.layerNorm" : 8,
|
| 167 |
+
"Ios18.sliceUpdate" : 8,
|
| 168 |
+
"Ios18.writeState" : 8,
|
| 169 |
+
"Ios18.reshape" : 26,
|
| 170 |
+
"Ios16.reduceMax" : 4,
|
| 171 |
+
"Ios16.reduceSum" : 4,
|
| 172 |
+
"Ios18.constexprLutToDense" : 28,
|
| 173 |
+
"Ios18.conv" : 24,
|
| 174 |
+
"Ios18.concat" : 24,
|
| 175 |
+
"Ios18.transpose" : 16,
|
| 176 |
+
"Ios18.sub" : 20,
|
| 177 |
+
"Ios18.linear" : 4,
|
| 178 |
+
"Ios18.silu" : 4,
|
| 179 |
+
"Ios18.sliceByIndex" : 26,
|
| 180 |
+
"Ios18.squeeze" : 12
|
| 181 |
+
}
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"inputSchema" : [
|
| 185 |
+
{
|
| 186 |
+
"hasShapeFlexibility" : "0",
|
| 187 |
+
"isOptional" : "0",
|
| 188 |
+
"dataType" : "Float16",
|
| 189 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 190 |
+
"shortDescription" : "",
|
| 191 |
+
"shape" : "[1, 256, 4096]",
|
| 192 |
+
"name" : "hidden_states",
|
| 193 |
+
"type" : "MultiArray"
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"hasShapeFlexibility" : "0",
|
| 197 |
+
"isOptional" : "0",
|
| 198 |
+
"dataType" : "Int32",
|
| 199 |
+
"formattedType" : "MultiArray (Int32 256)",
|
| 200 |
+
"shortDescription" : "",
|
| 201 |
+
"shape" : "[256]",
|
| 202 |
+
"name" : "position_ids",
|
| 203 |
+
"type" : "MultiArray"
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"hasShapeFlexibility" : "0",
|
| 207 |
+
"isOptional" : "0",
|
| 208 |
+
"dataType" : "Float16",
|
| 209 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 256 × 1024)",
|
| 210 |
+
"shortDescription" : "",
|
| 211 |
+
"shape" : "[1, 1, 256, 1024]",
|
| 212 |
+
"name" : "causal_mask",
|
| 213 |
+
"type" : "MultiArray"
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"hasShapeFlexibility" : "0",
|
| 217 |
+
"isOptional" : "0",
|
| 218 |
+
"dataType" : "Int32",
|
| 219 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 220 |
+
"shortDescription" : "",
|
| 221 |
+
"shape" : "[1]",
|
| 222 |
+
"name" : "current_pos",
|
| 223 |
+
"type" : "MultiArray"
|
| 224 |
+
}
|
| 225 |
+
],
|
| 226 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 227 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 228 |
+
"stateSchema" : [
|
| 229 |
+
{
|
| 230 |
+
"dataType" : "Float16",
|
| 231 |
+
"isOptional" : "0",
|
| 232 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 233 |
+
"shortDescription" : "",
|
| 234 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 235 |
+
"name" : "model_model_kv_cache_0",
|
| 236 |
+
"type" : "State"
|
| 237 |
+
}
|
| 238 |
+
],
|
| 239 |
+
"outputSchema" : [
|
| 240 |
+
{
|
| 241 |
+
"hasShapeFlexibility" : "0",
|
| 242 |
+
"isOptional" : "0",
|
| 243 |
+
"dataType" : "Float16",
|
| 244 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 245 |
+
"shortDescription" : "",
|
| 246 |
+
"shape" : "[1, 256, 4096]",
|
| 247 |
+
"name" : "output_hidden_states",
|
| 248 |
+
"type" : "MultiArray"
|
| 249 |
+
}
|
| 250 |
+
],
|
| 251 |
+
"name" : "prefill",
|
| 252 |
+
"mlProgramOperationTypeHistogram" : {
|
| 253 |
+
"Ios18.expandDims" : 16,
|
| 254 |
+
"Ios18.mul" : 40,
|
| 255 |
+
"Ios18.matmul" : 8,
|
| 256 |
+
"Ios16.reduceMean" : 8,
|
| 257 |
+
"Ios18.exp" : 4,
|
| 258 |
+
"Ios18.realDiv" : 4,
|
| 259 |
+
"Ios18.greaterEqual" : 1,
|
| 260 |
+
"Select" : 1,
|
| 261 |
+
"Ios18.readState" : 9,
|
| 262 |
+
"Tile" : 8,
|
| 263 |
+
"Ios18.gather" : 2,
|
| 264 |
+
"Ios18.add" : 22,
|
| 265 |
+
"Ios18.layerNorm" : 8,
|
| 266 |
+
"Ios18.sliceUpdate" : 8,
|
| 267 |
+
"Ios18.writeState" : 8,
|
| 268 |
+
"Ios18.reshape" : 34,
|
| 269 |
+
"Ios16.reduceMax" : 4,
|
| 270 |
+
"Ios16.reduceSum" : 4,
|
| 271 |
+
"Ios18.constexprLutToDense" : 28,
|
| 272 |
+
"Ios18.conv" : 24,
|
| 273 |
+
"Ios18.concat" : 24,
|
| 274 |
+
"Ios18.transpose" : 30,
|
| 275 |
+
"Ios18.sub" : 20,
|
| 276 |
+
"Ios18.linear" : 4,
|
| 277 |
+
"Ios18.silu" : 4,
|
| 278 |
+
"Ios18.sliceByIndex" : 26,
|
| 279 |
+
"Ios18.squeeze" : 12
|
| 280 |
+
}
|
| 281 |
+
}
|
| 282 |
+
],
|
| 283 |
+
"version" : "0.2.0",
|
| 284 |
+
"isUpdatable" : "0",
|
| 285 |
+
"defaultFunctionName" : "infer",
|
| 286 |
+
"specificationVersion" : 9,
|
| 287 |
+
"stateSchema" : [
|
| 288 |
+
{
|
| 289 |
+
"dataType" : "Float16",
|
| 290 |
+
"isOptional" : "0",
|
| 291 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 292 |
+
"shortDescription" : "",
|
| 293 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 294 |
+
"name" : "model_model_kv_cache_0",
|
| 295 |
+
"type" : "State"
|
| 296 |
+
}
|
| 297 |
+
],
|
| 298 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 299 |
+
"mlProgramOperationTypeHistogram" : {
|
| 300 |
+
"Ios18.expandDims" : 16,
|
| 301 |
+
"Ios18.mul" : 40,
|
| 302 |
+
"Ios18.matmul" : 8,
|
| 303 |
+
"Identity" : 1,
|
| 304 |
+
"Ios16.reduceMean" : 8,
|
| 305 |
+
"Ios18.exp" : 4,
|
| 306 |
+
"Ios18.realDiv" : 4,
|
| 307 |
+
"Ios18.greaterEqual" : 1,
|
| 308 |
+
"Select" : 1,
|
| 309 |
+
"Ios18.readState" : 9,
|
| 310 |
+
"Tile" : 8,
|
| 311 |
+
"Ios18.gather" : 2,
|
| 312 |
+
"Ios18.add" : 22,
|
| 313 |
+
"Ios18.layerNorm" : 8,
|
| 314 |
+
"Ios18.sliceUpdate" : 8,
|
| 315 |
+
"Ios18.writeState" : 8,
|
| 316 |
+
"Ios18.reshape" : 26,
|
| 317 |
+
"Ios16.reduceMax" : 4,
|
| 318 |
+
"Ios16.reduceSum" : 4,
|
| 319 |
+
"Ios18.constexprLutToDense" : 28,
|
| 320 |
+
"Ios18.conv" : 24,
|
| 321 |
+
"Ios18.concat" : 24,
|
| 322 |
+
"Ios18.transpose" : 16,
|
| 323 |
+
"Ios18.sub" : 20,
|
| 324 |
+
"Ios18.linear" : 4,
|
| 325 |
+
"Ios18.silu" : 4,
|
| 326 |
+
"Ios18.sliceByIndex" : 26,
|
| 327 |
+
"Ios18.squeeze" : 12
|
| 328 |
+
},
|
| 329 |
+
"shortDescription" : "Anemll Model: Multifunction FFN+Prefill",
|
| 330 |
+
"generatedClassName" : "DeepSeek_FFN_PF_lut6_chunk_03of08",
|
| 331 |
+
"author" : "Converted with Anemll v0.2.0",
|
| 332 |
+
"modelType" : {
|
| 333 |
+
"name" : "MLModelType_mlProgram"
|
| 334 |
+
}
|
| 335 |
+
}
|
| 336 |
+
]
|
DeepSeek_FFN_PF_lut6_chunk_03of08.mlmodelc/model.mil
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
DeepSeek_FFN_PF_lut6_chunk_03of08.mlmodelc/weights/weight.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:707eaa237412a09f3d91f79e4c2cb43cc782e9b0b772a70feab0ed2f15b7f3f7
|
| 3 |
+
size 724250880
|
DeepSeek_FFN_PF_lut6_chunk_04of08.mlmodelc/analytics/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:682a1d0d7f4fd3af26d8d1ca0fffb14c495d2a4f31eaf8d382e9de5a06b90870
|
| 3 |
+
size 243
|
DeepSeek_FFN_PF_lut6_chunk_04of08.mlmodelc/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df981ddc4b93953f29c99b2b486d8abee43b03a4c785772772b7e5fee1bf473a
|
| 3 |
+
size 986
|
DeepSeek_FFN_PF_lut6_chunk_04of08.mlmodelc/metadata.json
ADDED
|
@@ -0,0 +1,336 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"metadataOutputVersion" : "3.0",
|
| 4 |
+
"userDefinedMetadata" : {
|
| 5 |
+
"com.github.apple.coremltools.source" : "torch==2.5.0",
|
| 6 |
+
"com.github.apple.coremltools.version" : "8.2",
|
| 7 |
+
"com.anemll.context_length" : "1024",
|
| 8 |
+
"com.anemll.chunk_no" : "4",
|
| 9 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
| 10 |
+
"com.anemll.num_chunks" : "8",
|
| 11 |
+
"com.anemll.info" : "Converted with Anemll v0.2.0",
|
| 12 |
+
"com.anemll.batch_size" : "256",
|
| 13 |
+
"com.anemll.lut_bits" : "6"
|
| 14 |
+
},
|
| 15 |
+
"availability" : {
|
| 16 |
+
"macOS" : "15.0",
|
| 17 |
+
"tvOS" : "18.0",
|
| 18 |
+
"visionOS" : "2.0",
|
| 19 |
+
"watchOS" : "11.0",
|
| 20 |
+
"iOS" : "18.0",
|
| 21 |
+
"macCatalyst" : "18.0"
|
| 22 |
+
},
|
| 23 |
+
"inputSchema" : [
|
| 24 |
+
{
|
| 25 |
+
"hasShapeFlexibility" : "0",
|
| 26 |
+
"isOptional" : "0",
|
| 27 |
+
"dataType" : "Float16",
|
| 28 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 29 |
+
"shortDescription" : "",
|
| 30 |
+
"shape" : "[1, 1, 4096]",
|
| 31 |
+
"name" : "hidden_states",
|
| 32 |
+
"type" : "MultiArray"
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"hasShapeFlexibility" : "0",
|
| 36 |
+
"isOptional" : "0",
|
| 37 |
+
"dataType" : "Int32",
|
| 38 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 39 |
+
"shortDescription" : "",
|
| 40 |
+
"shape" : "[1]",
|
| 41 |
+
"name" : "position_ids",
|
| 42 |
+
"type" : "MultiArray"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"hasShapeFlexibility" : "0",
|
| 46 |
+
"isOptional" : "0",
|
| 47 |
+
"dataType" : "Float16",
|
| 48 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 49 |
+
"shortDescription" : "",
|
| 50 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 51 |
+
"name" : "causal_mask",
|
| 52 |
+
"type" : "MultiArray"
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"hasShapeFlexibility" : "0",
|
| 56 |
+
"isOptional" : "0",
|
| 57 |
+
"dataType" : "Int32",
|
| 58 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 59 |
+
"shortDescription" : "",
|
| 60 |
+
"shape" : "[1]",
|
| 61 |
+
"name" : "current_pos",
|
| 62 |
+
"type" : "MultiArray"
|
| 63 |
+
}
|
| 64 |
+
],
|
| 65 |
+
"outputSchema" : [
|
| 66 |
+
{
|
| 67 |
+
"hasShapeFlexibility" : "0",
|
| 68 |
+
"isOptional" : "0",
|
| 69 |
+
"dataType" : "Float16",
|
| 70 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 71 |
+
"shortDescription" : "",
|
| 72 |
+
"shape" : "[1, 1, 4096]",
|
| 73 |
+
"name" : "output_hidden_states",
|
| 74 |
+
"type" : "MultiArray"
|
| 75 |
+
}
|
| 76 |
+
],
|
| 77 |
+
"modelParameters" : [
|
| 78 |
+
|
| 79 |
+
],
|
| 80 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 81 |
+
"method" : "predict",
|
| 82 |
+
"functions" : [
|
| 83 |
+
{
|
| 84 |
+
"inputSchema" : [
|
| 85 |
+
{
|
| 86 |
+
"hasShapeFlexibility" : "0",
|
| 87 |
+
"isOptional" : "0",
|
| 88 |
+
"dataType" : "Float16",
|
| 89 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 90 |
+
"shortDescription" : "",
|
| 91 |
+
"shape" : "[1, 1, 4096]",
|
| 92 |
+
"name" : "hidden_states",
|
| 93 |
+
"type" : "MultiArray"
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"hasShapeFlexibility" : "0",
|
| 97 |
+
"isOptional" : "0",
|
| 98 |
+
"dataType" : "Int32",
|
| 99 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 100 |
+
"shortDescription" : "",
|
| 101 |
+
"shape" : "[1]",
|
| 102 |
+
"name" : "position_ids",
|
| 103 |
+
"type" : "MultiArray"
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"hasShapeFlexibility" : "0",
|
| 107 |
+
"isOptional" : "0",
|
| 108 |
+
"dataType" : "Float16",
|
| 109 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 110 |
+
"shortDescription" : "",
|
| 111 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 112 |
+
"name" : "causal_mask",
|
| 113 |
+
"type" : "MultiArray"
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"hasShapeFlexibility" : "0",
|
| 117 |
+
"isOptional" : "0",
|
| 118 |
+
"dataType" : "Int32",
|
| 119 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 120 |
+
"shortDescription" : "",
|
| 121 |
+
"shape" : "[1]",
|
| 122 |
+
"name" : "current_pos",
|
| 123 |
+
"type" : "MultiArray"
|
| 124 |
+
}
|
| 125 |
+
],
|
| 126 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 127 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 128 |
+
"stateSchema" : [
|
| 129 |
+
{
|
| 130 |
+
"dataType" : "Float16",
|
| 131 |
+
"isOptional" : "0",
|
| 132 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 133 |
+
"shortDescription" : "",
|
| 134 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 135 |
+
"name" : "model_model_kv_cache_0",
|
| 136 |
+
"type" : "State"
|
| 137 |
+
}
|
| 138 |
+
],
|
| 139 |
+
"outputSchema" : [
|
| 140 |
+
{
|
| 141 |
+
"hasShapeFlexibility" : "0",
|
| 142 |
+
"isOptional" : "0",
|
| 143 |
+
"dataType" : "Float16",
|
| 144 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 145 |
+
"shortDescription" : "",
|
| 146 |
+
"shape" : "[1, 1, 4096]",
|
| 147 |
+
"name" : "output_hidden_states",
|
| 148 |
+
"type" : "MultiArray"
|
| 149 |
+
}
|
| 150 |
+
],
|
| 151 |
+
"name" : "infer",
|
| 152 |
+
"mlProgramOperationTypeHistogram" : {
|
| 153 |
+
"Ios18.expandDims" : 16,
|
| 154 |
+
"Ios18.mul" : 40,
|
| 155 |
+
"Ios18.matmul" : 8,
|
| 156 |
+
"Identity" : 1,
|
| 157 |
+
"Ios16.reduceMean" : 8,
|
| 158 |
+
"Ios18.exp" : 4,
|
| 159 |
+
"Ios18.realDiv" : 4,
|
| 160 |
+
"Ios18.greaterEqual" : 1,
|
| 161 |
+
"Select" : 1,
|
| 162 |
+
"Ios18.readState" : 9,
|
| 163 |
+
"Tile" : 8,
|
| 164 |
+
"Ios18.gather" : 2,
|
| 165 |
+
"Ios18.add" : 22,
|
| 166 |
+
"Ios18.layerNorm" : 8,
|
| 167 |
+
"Ios18.sliceUpdate" : 8,
|
| 168 |
+
"Ios18.writeState" : 8,
|
| 169 |
+
"Ios18.reshape" : 26,
|
| 170 |
+
"Ios16.reduceMax" : 4,
|
| 171 |
+
"Ios16.reduceSum" : 4,
|
| 172 |
+
"Ios18.constexprLutToDense" : 28,
|
| 173 |
+
"Ios18.conv" : 24,
|
| 174 |
+
"Ios18.concat" : 24,
|
| 175 |
+
"Ios18.transpose" : 16,
|
| 176 |
+
"Ios18.sub" : 20,
|
| 177 |
+
"Ios18.linear" : 4,
|
| 178 |
+
"Ios18.silu" : 4,
|
| 179 |
+
"Ios18.sliceByIndex" : 26,
|
| 180 |
+
"Ios18.squeeze" : 12
|
| 181 |
+
}
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"inputSchema" : [
|
| 185 |
+
{
|
| 186 |
+
"hasShapeFlexibility" : "0",
|
| 187 |
+
"isOptional" : "0",
|
| 188 |
+
"dataType" : "Float16",
|
| 189 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 190 |
+
"shortDescription" : "",
|
| 191 |
+
"shape" : "[1, 256, 4096]",
|
| 192 |
+
"name" : "hidden_states",
|
| 193 |
+
"type" : "MultiArray"
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"hasShapeFlexibility" : "0",
|
| 197 |
+
"isOptional" : "0",
|
| 198 |
+
"dataType" : "Int32",
|
| 199 |
+
"formattedType" : "MultiArray (Int32 256)",
|
| 200 |
+
"shortDescription" : "",
|
| 201 |
+
"shape" : "[256]",
|
| 202 |
+
"name" : "position_ids",
|
| 203 |
+
"type" : "MultiArray"
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"hasShapeFlexibility" : "0",
|
| 207 |
+
"isOptional" : "0",
|
| 208 |
+
"dataType" : "Float16",
|
| 209 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 256 × 1024)",
|
| 210 |
+
"shortDescription" : "",
|
| 211 |
+
"shape" : "[1, 1, 256, 1024]",
|
| 212 |
+
"name" : "causal_mask",
|
| 213 |
+
"type" : "MultiArray"
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"hasShapeFlexibility" : "0",
|
| 217 |
+
"isOptional" : "0",
|
| 218 |
+
"dataType" : "Int32",
|
| 219 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 220 |
+
"shortDescription" : "",
|
| 221 |
+
"shape" : "[1]",
|
| 222 |
+
"name" : "current_pos",
|
| 223 |
+
"type" : "MultiArray"
|
| 224 |
+
}
|
| 225 |
+
],
|
| 226 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 227 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 228 |
+
"stateSchema" : [
|
| 229 |
+
{
|
| 230 |
+
"dataType" : "Float16",
|
| 231 |
+
"isOptional" : "0",
|
| 232 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 233 |
+
"shortDescription" : "",
|
| 234 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 235 |
+
"name" : "model_model_kv_cache_0",
|
| 236 |
+
"type" : "State"
|
| 237 |
+
}
|
| 238 |
+
],
|
| 239 |
+
"outputSchema" : [
|
| 240 |
+
{
|
| 241 |
+
"hasShapeFlexibility" : "0",
|
| 242 |
+
"isOptional" : "0",
|
| 243 |
+
"dataType" : "Float16",
|
| 244 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 245 |
+
"shortDescription" : "",
|
| 246 |
+
"shape" : "[1, 256, 4096]",
|
| 247 |
+
"name" : "output_hidden_states",
|
| 248 |
+
"type" : "MultiArray"
|
| 249 |
+
}
|
| 250 |
+
],
|
| 251 |
+
"name" : "prefill",
|
| 252 |
+
"mlProgramOperationTypeHistogram" : {
|
| 253 |
+
"Ios18.expandDims" : 16,
|
| 254 |
+
"Ios18.mul" : 40,
|
| 255 |
+
"Ios18.matmul" : 8,
|
| 256 |
+
"Ios16.reduceMean" : 8,
|
| 257 |
+
"Ios18.exp" : 4,
|
| 258 |
+
"Ios18.realDiv" : 4,
|
| 259 |
+
"Ios18.greaterEqual" : 1,
|
| 260 |
+
"Select" : 1,
|
| 261 |
+
"Ios18.readState" : 9,
|
| 262 |
+
"Tile" : 8,
|
| 263 |
+
"Ios18.gather" : 2,
|
| 264 |
+
"Ios18.add" : 22,
|
| 265 |
+
"Ios18.layerNorm" : 8,
|
| 266 |
+
"Ios18.sliceUpdate" : 8,
|
| 267 |
+
"Ios18.writeState" : 8,
|
| 268 |
+
"Ios18.reshape" : 34,
|
| 269 |
+
"Ios16.reduceMax" : 4,
|
| 270 |
+
"Ios16.reduceSum" : 4,
|
| 271 |
+
"Ios18.constexprLutToDense" : 28,
|
| 272 |
+
"Ios18.conv" : 24,
|
| 273 |
+
"Ios18.concat" : 24,
|
| 274 |
+
"Ios18.transpose" : 30,
|
| 275 |
+
"Ios18.sub" : 20,
|
| 276 |
+
"Ios18.linear" : 4,
|
| 277 |
+
"Ios18.silu" : 4,
|
| 278 |
+
"Ios18.sliceByIndex" : 26,
|
| 279 |
+
"Ios18.squeeze" : 12
|
| 280 |
+
}
|
| 281 |
+
}
|
| 282 |
+
],
|
| 283 |
+
"version" : "0.2.0",
|
| 284 |
+
"isUpdatable" : "0",
|
| 285 |
+
"defaultFunctionName" : "infer",
|
| 286 |
+
"specificationVersion" : 9,
|
| 287 |
+
"stateSchema" : [
|
| 288 |
+
{
|
| 289 |
+
"dataType" : "Float16",
|
| 290 |
+
"isOptional" : "0",
|
| 291 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 292 |
+
"shortDescription" : "",
|
| 293 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 294 |
+
"name" : "model_model_kv_cache_0",
|
| 295 |
+
"type" : "State"
|
| 296 |
+
}
|
| 297 |
+
],
|
| 298 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 299 |
+
"mlProgramOperationTypeHistogram" : {
|
| 300 |
+
"Ios18.expandDims" : 16,
|
| 301 |
+
"Ios18.mul" : 40,
|
| 302 |
+
"Ios18.matmul" : 8,
|
| 303 |
+
"Identity" : 1,
|
| 304 |
+
"Ios16.reduceMean" : 8,
|
| 305 |
+
"Ios18.exp" : 4,
|
| 306 |
+
"Ios18.realDiv" : 4,
|
| 307 |
+
"Ios18.greaterEqual" : 1,
|
| 308 |
+
"Select" : 1,
|
| 309 |
+
"Ios18.readState" : 9,
|
| 310 |
+
"Tile" : 8,
|
| 311 |
+
"Ios18.gather" : 2,
|
| 312 |
+
"Ios18.add" : 22,
|
| 313 |
+
"Ios18.layerNorm" : 8,
|
| 314 |
+
"Ios18.sliceUpdate" : 8,
|
| 315 |
+
"Ios18.writeState" : 8,
|
| 316 |
+
"Ios18.reshape" : 26,
|
| 317 |
+
"Ios16.reduceMax" : 4,
|
| 318 |
+
"Ios16.reduceSum" : 4,
|
| 319 |
+
"Ios18.constexprLutToDense" : 28,
|
| 320 |
+
"Ios18.conv" : 24,
|
| 321 |
+
"Ios18.concat" : 24,
|
| 322 |
+
"Ios18.transpose" : 16,
|
| 323 |
+
"Ios18.sub" : 20,
|
| 324 |
+
"Ios18.linear" : 4,
|
| 325 |
+
"Ios18.silu" : 4,
|
| 326 |
+
"Ios18.sliceByIndex" : 26,
|
| 327 |
+
"Ios18.squeeze" : 12
|
| 328 |
+
},
|
| 329 |
+
"shortDescription" : "Anemll Model: Multifunction FFN+Prefill",
|
| 330 |
+
"generatedClassName" : "DeepSeek_FFN_PF_lut6_chunk_04of08",
|
| 331 |
+
"author" : "Converted with Anemll v0.2.0",
|
| 332 |
+
"modelType" : {
|
| 333 |
+
"name" : "MLModelType_mlProgram"
|
| 334 |
+
}
|
| 335 |
+
}
|
| 336 |
+
]
|
DeepSeek_FFN_PF_lut6_chunk_04of08.mlmodelc/model.mil
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
DeepSeek_FFN_PF_lut6_chunk_04of08.mlmodelc/weights/weight.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d2083debfd509db551c7d371baa871384d409eccd21dbe97a5d1b70d3d1a8635
|
| 3 |
+
size 724250880
|
DeepSeek_FFN_PF_lut6_chunk_05of08.mlmodelc/analytics/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:11fd55f3a88a19593a12f28f3e3df4959c61af126a538c4b9264d0318f87f92c
|
| 3 |
+
size 243
|
DeepSeek_FFN_PF_lut6_chunk_05of08.mlmodelc/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d2187177566f70918110e8806292cc44e72bd8a938f6933e297f0eaf210579e
|
| 3 |
+
size 986
|
DeepSeek_FFN_PF_lut6_chunk_05of08.mlmodelc/metadata.json
ADDED
|
@@ -0,0 +1,336 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"metadataOutputVersion" : "3.0",
|
| 4 |
+
"userDefinedMetadata" : {
|
| 5 |
+
"com.anemll.chunk_no" : "5",
|
| 6 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
| 7 |
+
"com.github.apple.coremltools.source" : "torch==2.5.0",
|
| 8 |
+
"com.anemll.context_length" : "1024",
|
| 9 |
+
"com.github.apple.coremltools.version" : "8.2",
|
| 10 |
+
"com.anemll.num_chunks" : "8",
|
| 11 |
+
"com.anemll.batch_size" : "256",
|
| 12 |
+
"com.anemll.info" : "Converted with Anemll v0.2.0",
|
| 13 |
+
"com.anemll.lut_bits" : "6"
|
| 14 |
+
},
|
| 15 |
+
"availability" : {
|
| 16 |
+
"macOS" : "15.0",
|
| 17 |
+
"tvOS" : "18.0",
|
| 18 |
+
"visionOS" : "2.0",
|
| 19 |
+
"watchOS" : "11.0",
|
| 20 |
+
"iOS" : "18.0",
|
| 21 |
+
"macCatalyst" : "18.0"
|
| 22 |
+
},
|
| 23 |
+
"inputSchema" : [
|
| 24 |
+
{
|
| 25 |
+
"hasShapeFlexibility" : "0",
|
| 26 |
+
"isOptional" : "0",
|
| 27 |
+
"dataType" : "Float16",
|
| 28 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 29 |
+
"shortDescription" : "",
|
| 30 |
+
"shape" : "[1, 1, 4096]",
|
| 31 |
+
"name" : "hidden_states",
|
| 32 |
+
"type" : "MultiArray"
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"hasShapeFlexibility" : "0",
|
| 36 |
+
"isOptional" : "0",
|
| 37 |
+
"dataType" : "Int32",
|
| 38 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 39 |
+
"shortDescription" : "",
|
| 40 |
+
"shape" : "[1]",
|
| 41 |
+
"name" : "position_ids",
|
| 42 |
+
"type" : "MultiArray"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"hasShapeFlexibility" : "0",
|
| 46 |
+
"isOptional" : "0",
|
| 47 |
+
"dataType" : "Float16",
|
| 48 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 49 |
+
"shortDescription" : "",
|
| 50 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 51 |
+
"name" : "causal_mask",
|
| 52 |
+
"type" : "MultiArray"
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"hasShapeFlexibility" : "0",
|
| 56 |
+
"isOptional" : "0",
|
| 57 |
+
"dataType" : "Int32",
|
| 58 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 59 |
+
"shortDescription" : "",
|
| 60 |
+
"shape" : "[1]",
|
| 61 |
+
"name" : "current_pos",
|
| 62 |
+
"type" : "MultiArray"
|
| 63 |
+
}
|
| 64 |
+
],
|
| 65 |
+
"outputSchema" : [
|
| 66 |
+
{
|
| 67 |
+
"hasShapeFlexibility" : "0",
|
| 68 |
+
"isOptional" : "0",
|
| 69 |
+
"dataType" : "Float16",
|
| 70 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 71 |
+
"shortDescription" : "",
|
| 72 |
+
"shape" : "[1, 1, 4096]",
|
| 73 |
+
"name" : "output_hidden_states",
|
| 74 |
+
"type" : "MultiArray"
|
| 75 |
+
}
|
| 76 |
+
],
|
| 77 |
+
"modelParameters" : [
|
| 78 |
+
|
| 79 |
+
],
|
| 80 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 81 |
+
"method" : "predict",
|
| 82 |
+
"functions" : [
|
| 83 |
+
{
|
| 84 |
+
"inputSchema" : [
|
| 85 |
+
{
|
| 86 |
+
"hasShapeFlexibility" : "0",
|
| 87 |
+
"isOptional" : "0",
|
| 88 |
+
"dataType" : "Float16",
|
| 89 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 90 |
+
"shortDescription" : "",
|
| 91 |
+
"shape" : "[1, 1, 4096]",
|
| 92 |
+
"name" : "hidden_states",
|
| 93 |
+
"type" : "MultiArray"
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"hasShapeFlexibility" : "0",
|
| 97 |
+
"isOptional" : "0",
|
| 98 |
+
"dataType" : "Int32",
|
| 99 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 100 |
+
"shortDescription" : "",
|
| 101 |
+
"shape" : "[1]",
|
| 102 |
+
"name" : "position_ids",
|
| 103 |
+
"type" : "MultiArray"
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"hasShapeFlexibility" : "0",
|
| 107 |
+
"isOptional" : "0",
|
| 108 |
+
"dataType" : "Float16",
|
| 109 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 110 |
+
"shortDescription" : "",
|
| 111 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 112 |
+
"name" : "causal_mask",
|
| 113 |
+
"type" : "MultiArray"
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"hasShapeFlexibility" : "0",
|
| 117 |
+
"isOptional" : "0",
|
| 118 |
+
"dataType" : "Int32",
|
| 119 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 120 |
+
"shortDescription" : "",
|
| 121 |
+
"shape" : "[1]",
|
| 122 |
+
"name" : "current_pos",
|
| 123 |
+
"type" : "MultiArray"
|
| 124 |
+
}
|
| 125 |
+
],
|
| 126 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 127 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 128 |
+
"stateSchema" : [
|
| 129 |
+
{
|
| 130 |
+
"dataType" : "Float16",
|
| 131 |
+
"isOptional" : "0",
|
| 132 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 133 |
+
"shortDescription" : "",
|
| 134 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 135 |
+
"name" : "model_model_kv_cache_0",
|
| 136 |
+
"type" : "State"
|
| 137 |
+
}
|
| 138 |
+
],
|
| 139 |
+
"outputSchema" : [
|
| 140 |
+
{
|
| 141 |
+
"hasShapeFlexibility" : "0",
|
| 142 |
+
"isOptional" : "0",
|
| 143 |
+
"dataType" : "Float16",
|
| 144 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 145 |
+
"shortDescription" : "",
|
| 146 |
+
"shape" : "[1, 1, 4096]",
|
| 147 |
+
"name" : "output_hidden_states",
|
| 148 |
+
"type" : "MultiArray"
|
| 149 |
+
}
|
| 150 |
+
],
|
| 151 |
+
"name" : "infer",
|
| 152 |
+
"mlProgramOperationTypeHistogram" : {
|
| 153 |
+
"Ios18.expandDims" : 16,
|
| 154 |
+
"Ios18.mul" : 40,
|
| 155 |
+
"Ios18.matmul" : 8,
|
| 156 |
+
"Identity" : 1,
|
| 157 |
+
"Ios16.reduceMean" : 8,
|
| 158 |
+
"Ios18.exp" : 4,
|
| 159 |
+
"Ios18.realDiv" : 4,
|
| 160 |
+
"Ios18.greaterEqual" : 1,
|
| 161 |
+
"Select" : 1,
|
| 162 |
+
"Ios18.readState" : 9,
|
| 163 |
+
"Tile" : 8,
|
| 164 |
+
"Ios18.gather" : 2,
|
| 165 |
+
"Ios18.add" : 22,
|
| 166 |
+
"Ios18.layerNorm" : 8,
|
| 167 |
+
"Ios18.sliceUpdate" : 8,
|
| 168 |
+
"Ios18.writeState" : 8,
|
| 169 |
+
"Ios18.reshape" : 26,
|
| 170 |
+
"Ios16.reduceMax" : 4,
|
| 171 |
+
"Ios16.reduceSum" : 4,
|
| 172 |
+
"Ios18.constexprLutToDense" : 28,
|
| 173 |
+
"Ios18.conv" : 24,
|
| 174 |
+
"Ios18.concat" : 24,
|
| 175 |
+
"Ios18.transpose" : 16,
|
| 176 |
+
"Ios18.sub" : 20,
|
| 177 |
+
"Ios18.linear" : 4,
|
| 178 |
+
"Ios18.silu" : 4,
|
| 179 |
+
"Ios18.sliceByIndex" : 26,
|
| 180 |
+
"Ios18.squeeze" : 12
|
| 181 |
+
}
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"inputSchema" : [
|
| 185 |
+
{
|
| 186 |
+
"hasShapeFlexibility" : "0",
|
| 187 |
+
"isOptional" : "0",
|
| 188 |
+
"dataType" : "Float16",
|
| 189 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 190 |
+
"shortDescription" : "",
|
| 191 |
+
"shape" : "[1, 256, 4096]",
|
| 192 |
+
"name" : "hidden_states",
|
| 193 |
+
"type" : "MultiArray"
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"hasShapeFlexibility" : "0",
|
| 197 |
+
"isOptional" : "0",
|
| 198 |
+
"dataType" : "Int32",
|
| 199 |
+
"formattedType" : "MultiArray (Int32 256)",
|
| 200 |
+
"shortDescription" : "",
|
| 201 |
+
"shape" : "[256]",
|
| 202 |
+
"name" : "position_ids",
|
| 203 |
+
"type" : "MultiArray"
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"hasShapeFlexibility" : "0",
|
| 207 |
+
"isOptional" : "0",
|
| 208 |
+
"dataType" : "Float16",
|
| 209 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 256 × 1024)",
|
| 210 |
+
"shortDescription" : "",
|
| 211 |
+
"shape" : "[1, 1, 256, 1024]",
|
| 212 |
+
"name" : "causal_mask",
|
| 213 |
+
"type" : "MultiArray"
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"hasShapeFlexibility" : "0",
|
| 217 |
+
"isOptional" : "0",
|
| 218 |
+
"dataType" : "Int32",
|
| 219 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 220 |
+
"shortDescription" : "",
|
| 221 |
+
"shape" : "[1]",
|
| 222 |
+
"name" : "current_pos",
|
| 223 |
+
"type" : "MultiArray"
|
| 224 |
+
}
|
| 225 |
+
],
|
| 226 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 227 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 228 |
+
"stateSchema" : [
|
| 229 |
+
{
|
| 230 |
+
"dataType" : "Float16",
|
| 231 |
+
"isOptional" : "0",
|
| 232 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 233 |
+
"shortDescription" : "",
|
| 234 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 235 |
+
"name" : "model_model_kv_cache_0",
|
| 236 |
+
"type" : "State"
|
| 237 |
+
}
|
| 238 |
+
],
|
| 239 |
+
"outputSchema" : [
|
| 240 |
+
{
|
| 241 |
+
"hasShapeFlexibility" : "0",
|
| 242 |
+
"isOptional" : "0",
|
| 243 |
+
"dataType" : "Float16",
|
| 244 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 245 |
+
"shortDescription" : "",
|
| 246 |
+
"shape" : "[1, 256, 4096]",
|
| 247 |
+
"name" : "output_hidden_states",
|
| 248 |
+
"type" : "MultiArray"
|
| 249 |
+
}
|
| 250 |
+
],
|
| 251 |
+
"name" : "prefill",
|
| 252 |
+
"mlProgramOperationTypeHistogram" : {
|
| 253 |
+
"Ios18.expandDims" : 16,
|
| 254 |
+
"Ios18.mul" : 40,
|
| 255 |
+
"Ios18.matmul" : 8,
|
| 256 |
+
"Ios16.reduceMean" : 8,
|
| 257 |
+
"Ios18.exp" : 4,
|
| 258 |
+
"Ios18.realDiv" : 4,
|
| 259 |
+
"Ios18.greaterEqual" : 1,
|
| 260 |
+
"Select" : 1,
|
| 261 |
+
"Ios18.readState" : 9,
|
| 262 |
+
"Tile" : 8,
|
| 263 |
+
"Ios18.gather" : 2,
|
| 264 |
+
"Ios18.add" : 22,
|
| 265 |
+
"Ios18.layerNorm" : 8,
|
| 266 |
+
"Ios18.sliceUpdate" : 8,
|
| 267 |
+
"Ios18.writeState" : 8,
|
| 268 |
+
"Ios18.reshape" : 34,
|
| 269 |
+
"Ios16.reduceMax" : 4,
|
| 270 |
+
"Ios16.reduceSum" : 4,
|
| 271 |
+
"Ios18.constexprLutToDense" : 28,
|
| 272 |
+
"Ios18.conv" : 24,
|
| 273 |
+
"Ios18.concat" : 24,
|
| 274 |
+
"Ios18.transpose" : 30,
|
| 275 |
+
"Ios18.sub" : 20,
|
| 276 |
+
"Ios18.linear" : 4,
|
| 277 |
+
"Ios18.silu" : 4,
|
| 278 |
+
"Ios18.sliceByIndex" : 26,
|
| 279 |
+
"Ios18.squeeze" : 12
|
| 280 |
+
}
|
| 281 |
+
}
|
| 282 |
+
],
|
| 283 |
+
"version" : "0.2.0",
|
| 284 |
+
"isUpdatable" : "0",
|
| 285 |
+
"defaultFunctionName" : "infer",
|
| 286 |
+
"specificationVersion" : 9,
|
| 287 |
+
"stateSchema" : [
|
| 288 |
+
{
|
| 289 |
+
"dataType" : "Float16",
|
| 290 |
+
"isOptional" : "0",
|
| 291 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 292 |
+
"shortDescription" : "",
|
| 293 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 294 |
+
"name" : "model_model_kv_cache_0",
|
| 295 |
+
"type" : "State"
|
| 296 |
+
}
|
| 297 |
+
],
|
| 298 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 299 |
+
"mlProgramOperationTypeHistogram" : {
|
| 300 |
+
"Ios18.expandDims" : 16,
|
| 301 |
+
"Ios18.mul" : 40,
|
| 302 |
+
"Ios18.matmul" : 8,
|
| 303 |
+
"Identity" : 1,
|
| 304 |
+
"Ios16.reduceMean" : 8,
|
| 305 |
+
"Ios18.exp" : 4,
|
| 306 |
+
"Ios18.realDiv" : 4,
|
| 307 |
+
"Ios18.greaterEqual" : 1,
|
| 308 |
+
"Select" : 1,
|
| 309 |
+
"Ios18.readState" : 9,
|
| 310 |
+
"Tile" : 8,
|
| 311 |
+
"Ios18.gather" : 2,
|
| 312 |
+
"Ios18.add" : 22,
|
| 313 |
+
"Ios18.layerNorm" : 8,
|
| 314 |
+
"Ios18.sliceUpdate" : 8,
|
| 315 |
+
"Ios18.writeState" : 8,
|
| 316 |
+
"Ios18.reshape" : 26,
|
| 317 |
+
"Ios16.reduceMax" : 4,
|
| 318 |
+
"Ios16.reduceSum" : 4,
|
| 319 |
+
"Ios18.constexprLutToDense" : 28,
|
| 320 |
+
"Ios18.conv" : 24,
|
| 321 |
+
"Ios18.concat" : 24,
|
| 322 |
+
"Ios18.transpose" : 16,
|
| 323 |
+
"Ios18.sub" : 20,
|
| 324 |
+
"Ios18.linear" : 4,
|
| 325 |
+
"Ios18.silu" : 4,
|
| 326 |
+
"Ios18.sliceByIndex" : 26,
|
| 327 |
+
"Ios18.squeeze" : 12
|
| 328 |
+
},
|
| 329 |
+
"shortDescription" : "Anemll Model: Multifunction FFN+Prefill",
|
| 330 |
+
"generatedClassName" : "DeepSeek_FFN_PF_lut6_chunk_05of08",
|
| 331 |
+
"author" : "Converted with Anemll v0.2.0",
|
| 332 |
+
"modelType" : {
|
| 333 |
+
"name" : "MLModelType_mlProgram"
|
| 334 |
+
}
|
| 335 |
+
}
|
| 336 |
+
]
|
DeepSeek_FFN_PF_lut6_chunk_05of08.mlmodelc/model.mil
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
DeepSeek_FFN_PF_lut6_chunk_05of08.mlmodelc/weights/weight.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3371621923371d80917a6d5853d246fb83728c358f67a72655458b02b029010b
|
| 3 |
+
size 724250880
|
DeepSeek_FFN_PF_lut6_chunk_06of08.mlmodelc/analytics/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:334e8f4c83cded8bd2f833e0cea828228d3356b904eb3a1be5b9daf31016b9d8
|
| 3 |
+
size 243
|
DeepSeek_FFN_PF_lut6_chunk_06of08.mlmodelc/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:78e1339d57779b788f7eedd2d452dc5628f955130dfccc6e0ceaeaccbfd12880
|
| 3 |
+
size 986
|
DeepSeek_FFN_PF_lut6_chunk_06of08.mlmodelc/metadata.json
ADDED
|
@@ -0,0 +1,336 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"metadataOutputVersion" : "3.0",
|
| 4 |
+
"userDefinedMetadata" : {
|
| 5 |
+
"com.anemll.lut_bits" : "6",
|
| 6 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
| 7 |
+
"com.github.apple.coremltools.source" : "torch==2.5.0",
|
| 8 |
+
"com.anemll.context_length" : "1024",
|
| 9 |
+
"com.github.apple.coremltools.version" : "8.2",
|
| 10 |
+
"com.anemll.num_chunks" : "8",
|
| 11 |
+
"com.anemll.info" : "Converted with Anemll v0.2.0",
|
| 12 |
+
"com.anemll.batch_size" : "256",
|
| 13 |
+
"com.anemll.chunk_no" : "6"
|
| 14 |
+
},
|
| 15 |
+
"availability" : {
|
| 16 |
+
"macOS" : "15.0",
|
| 17 |
+
"tvOS" : "18.0",
|
| 18 |
+
"visionOS" : "2.0",
|
| 19 |
+
"watchOS" : "11.0",
|
| 20 |
+
"iOS" : "18.0",
|
| 21 |
+
"macCatalyst" : "18.0"
|
| 22 |
+
},
|
| 23 |
+
"inputSchema" : [
|
| 24 |
+
{
|
| 25 |
+
"hasShapeFlexibility" : "0",
|
| 26 |
+
"isOptional" : "0",
|
| 27 |
+
"dataType" : "Float16",
|
| 28 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 29 |
+
"shortDescription" : "",
|
| 30 |
+
"shape" : "[1, 1, 4096]",
|
| 31 |
+
"name" : "hidden_states",
|
| 32 |
+
"type" : "MultiArray"
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"hasShapeFlexibility" : "0",
|
| 36 |
+
"isOptional" : "0",
|
| 37 |
+
"dataType" : "Int32",
|
| 38 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 39 |
+
"shortDescription" : "",
|
| 40 |
+
"shape" : "[1]",
|
| 41 |
+
"name" : "position_ids",
|
| 42 |
+
"type" : "MultiArray"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"hasShapeFlexibility" : "0",
|
| 46 |
+
"isOptional" : "0",
|
| 47 |
+
"dataType" : "Float16",
|
| 48 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 49 |
+
"shortDescription" : "",
|
| 50 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 51 |
+
"name" : "causal_mask",
|
| 52 |
+
"type" : "MultiArray"
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"hasShapeFlexibility" : "0",
|
| 56 |
+
"isOptional" : "0",
|
| 57 |
+
"dataType" : "Int32",
|
| 58 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 59 |
+
"shortDescription" : "",
|
| 60 |
+
"shape" : "[1]",
|
| 61 |
+
"name" : "current_pos",
|
| 62 |
+
"type" : "MultiArray"
|
| 63 |
+
}
|
| 64 |
+
],
|
| 65 |
+
"outputSchema" : [
|
| 66 |
+
{
|
| 67 |
+
"hasShapeFlexibility" : "0",
|
| 68 |
+
"isOptional" : "0",
|
| 69 |
+
"dataType" : "Float16",
|
| 70 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 71 |
+
"shortDescription" : "",
|
| 72 |
+
"shape" : "[1, 1, 4096]",
|
| 73 |
+
"name" : "output_hidden_states",
|
| 74 |
+
"type" : "MultiArray"
|
| 75 |
+
}
|
| 76 |
+
],
|
| 77 |
+
"modelParameters" : [
|
| 78 |
+
|
| 79 |
+
],
|
| 80 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 81 |
+
"method" : "predict",
|
| 82 |
+
"functions" : [
|
| 83 |
+
{
|
| 84 |
+
"inputSchema" : [
|
| 85 |
+
{
|
| 86 |
+
"hasShapeFlexibility" : "0",
|
| 87 |
+
"isOptional" : "0",
|
| 88 |
+
"dataType" : "Float16",
|
| 89 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 90 |
+
"shortDescription" : "",
|
| 91 |
+
"shape" : "[1, 1, 4096]",
|
| 92 |
+
"name" : "hidden_states",
|
| 93 |
+
"type" : "MultiArray"
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"hasShapeFlexibility" : "0",
|
| 97 |
+
"isOptional" : "0",
|
| 98 |
+
"dataType" : "Int32",
|
| 99 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 100 |
+
"shortDescription" : "",
|
| 101 |
+
"shape" : "[1]",
|
| 102 |
+
"name" : "position_ids",
|
| 103 |
+
"type" : "MultiArray"
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"hasShapeFlexibility" : "0",
|
| 107 |
+
"isOptional" : "0",
|
| 108 |
+
"dataType" : "Float16",
|
| 109 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 110 |
+
"shortDescription" : "",
|
| 111 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 112 |
+
"name" : "causal_mask",
|
| 113 |
+
"type" : "MultiArray"
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"hasShapeFlexibility" : "0",
|
| 117 |
+
"isOptional" : "0",
|
| 118 |
+
"dataType" : "Int32",
|
| 119 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 120 |
+
"shortDescription" : "",
|
| 121 |
+
"shape" : "[1]",
|
| 122 |
+
"name" : "current_pos",
|
| 123 |
+
"type" : "MultiArray"
|
| 124 |
+
}
|
| 125 |
+
],
|
| 126 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 127 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 128 |
+
"stateSchema" : [
|
| 129 |
+
{
|
| 130 |
+
"dataType" : "Float16",
|
| 131 |
+
"isOptional" : "0",
|
| 132 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 133 |
+
"shortDescription" : "",
|
| 134 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 135 |
+
"name" : "model_model_kv_cache_0",
|
| 136 |
+
"type" : "State"
|
| 137 |
+
}
|
| 138 |
+
],
|
| 139 |
+
"outputSchema" : [
|
| 140 |
+
{
|
| 141 |
+
"hasShapeFlexibility" : "0",
|
| 142 |
+
"isOptional" : "0",
|
| 143 |
+
"dataType" : "Float16",
|
| 144 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 145 |
+
"shortDescription" : "",
|
| 146 |
+
"shape" : "[1, 1, 4096]",
|
| 147 |
+
"name" : "output_hidden_states",
|
| 148 |
+
"type" : "MultiArray"
|
| 149 |
+
}
|
| 150 |
+
],
|
| 151 |
+
"name" : "infer",
|
| 152 |
+
"mlProgramOperationTypeHistogram" : {
|
| 153 |
+
"Ios18.expandDims" : 16,
|
| 154 |
+
"Ios18.mul" : 40,
|
| 155 |
+
"Ios18.matmul" : 8,
|
| 156 |
+
"Identity" : 1,
|
| 157 |
+
"Ios16.reduceMean" : 8,
|
| 158 |
+
"Ios18.exp" : 4,
|
| 159 |
+
"Ios18.realDiv" : 4,
|
| 160 |
+
"Ios18.greaterEqual" : 1,
|
| 161 |
+
"Select" : 1,
|
| 162 |
+
"Ios18.readState" : 9,
|
| 163 |
+
"Tile" : 8,
|
| 164 |
+
"Ios18.gather" : 2,
|
| 165 |
+
"Ios18.add" : 22,
|
| 166 |
+
"Ios18.layerNorm" : 8,
|
| 167 |
+
"Ios18.sliceUpdate" : 8,
|
| 168 |
+
"Ios18.writeState" : 8,
|
| 169 |
+
"Ios18.reshape" : 26,
|
| 170 |
+
"Ios16.reduceMax" : 4,
|
| 171 |
+
"Ios16.reduceSum" : 4,
|
| 172 |
+
"Ios18.constexprLutToDense" : 28,
|
| 173 |
+
"Ios18.conv" : 24,
|
| 174 |
+
"Ios18.concat" : 24,
|
| 175 |
+
"Ios18.transpose" : 16,
|
| 176 |
+
"Ios18.sub" : 20,
|
| 177 |
+
"Ios18.linear" : 4,
|
| 178 |
+
"Ios18.silu" : 4,
|
| 179 |
+
"Ios18.sliceByIndex" : 26,
|
| 180 |
+
"Ios18.squeeze" : 12
|
| 181 |
+
}
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"inputSchema" : [
|
| 185 |
+
{
|
| 186 |
+
"hasShapeFlexibility" : "0",
|
| 187 |
+
"isOptional" : "0",
|
| 188 |
+
"dataType" : "Float16",
|
| 189 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 190 |
+
"shortDescription" : "",
|
| 191 |
+
"shape" : "[1, 256, 4096]",
|
| 192 |
+
"name" : "hidden_states",
|
| 193 |
+
"type" : "MultiArray"
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"hasShapeFlexibility" : "0",
|
| 197 |
+
"isOptional" : "0",
|
| 198 |
+
"dataType" : "Int32",
|
| 199 |
+
"formattedType" : "MultiArray (Int32 256)",
|
| 200 |
+
"shortDescription" : "",
|
| 201 |
+
"shape" : "[256]",
|
| 202 |
+
"name" : "position_ids",
|
| 203 |
+
"type" : "MultiArray"
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"hasShapeFlexibility" : "0",
|
| 207 |
+
"isOptional" : "0",
|
| 208 |
+
"dataType" : "Float16",
|
| 209 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 256 × 1024)",
|
| 210 |
+
"shortDescription" : "",
|
| 211 |
+
"shape" : "[1, 1, 256, 1024]",
|
| 212 |
+
"name" : "causal_mask",
|
| 213 |
+
"type" : "MultiArray"
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"hasShapeFlexibility" : "0",
|
| 217 |
+
"isOptional" : "0",
|
| 218 |
+
"dataType" : "Int32",
|
| 219 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 220 |
+
"shortDescription" : "",
|
| 221 |
+
"shape" : "[1]",
|
| 222 |
+
"name" : "current_pos",
|
| 223 |
+
"type" : "MultiArray"
|
| 224 |
+
}
|
| 225 |
+
],
|
| 226 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 227 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 228 |
+
"stateSchema" : [
|
| 229 |
+
{
|
| 230 |
+
"dataType" : "Float16",
|
| 231 |
+
"isOptional" : "0",
|
| 232 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 233 |
+
"shortDescription" : "",
|
| 234 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 235 |
+
"name" : "model_model_kv_cache_0",
|
| 236 |
+
"type" : "State"
|
| 237 |
+
}
|
| 238 |
+
],
|
| 239 |
+
"outputSchema" : [
|
| 240 |
+
{
|
| 241 |
+
"hasShapeFlexibility" : "0",
|
| 242 |
+
"isOptional" : "0",
|
| 243 |
+
"dataType" : "Float16",
|
| 244 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 245 |
+
"shortDescription" : "",
|
| 246 |
+
"shape" : "[1, 256, 4096]",
|
| 247 |
+
"name" : "output_hidden_states",
|
| 248 |
+
"type" : "MultiArray"
|
| 249 |
+
}
|
| 250 |
+
],
|
| 251 |
+
"name" : "prefill",
|
| 252 |
+
"mlProgramOperationTypeHistogram" : {
|
| 253 |
+
"Ios18.expandDims" : 16,
|
| 254 |
+
"Ios18.mul" : 40,
|
| 255 |
+
"Ios18.matmul" : 8,
|
| 256 |
+
"Ios16.reduceMean" : 8,
|
| 257 |
+
"Ios18.exp" : 4,
|
| 258 |
+
"Ios18.realDiv" : 4,
|
| 259 |
+
"Ios18.greaterEqual" : 1,
|
| 260 |
+
"Select" : 1,
|
| 261 |
+
"Ios18.readState" : 9,
|
| 262 |
+
"Tile" : 8,
|
| 263 |
+
"Ios18.gather" : 2,
|
| 264 |
+
"Ios18.add" : 22,
|
| 265 |
+
"Ios18.layerNorm" : 8,
|
| 266 |
+
"Ios18.sliceUpdate" : 8,
|
| 267 |
+
"Ios18.writeState" : 8,
|
| 268 |
+
"Ios18.reshape" : 34,
|
| 269 |
+
"Ios16.reduceMax" : 4,
|
| 270 |
+
"Ios16.reduceSum" : 4,
|
| 271 |
+
"Ios18.constexprLutToDense" : 28,
|
| 272 |
+
"Ios18.conv" : 24,
|
| 273 |
+
"Ios18.concat" : 24,
|
| 274 |
+
"Ios18.transpose" : 30,
|
| 275 |
+
"Ios18.sub" : 20,
|
| 276 |
+
"Ios18.linear" : 4,
|
| 277 |
+
"Ios18.silu" : 4,
|
| 278 |
+
"Ios18.sliceByIndex" : 26,
|
| 279 |
+
"Ios18.squeeze" : 12
|
| 280 |
+
}
|
| 281 |
+
}
|
| 282 |
+
],
|
| 283 |
+
"version" : "0.2.0",
|
| 284 |
+
"isUpdatable" : "0",
|
| 285 |
+
"defaultFunctionName" : "infer",
|
| 286 |
+
"specificationVersion" : 9,
|
| 287 |
+
"stateSchema" : [
|
| 288 |
+
{
|
| 289 |
+
"dataType" : "Float16",
|
| 290 |
+
"isOptional" : "0",
|
| 291 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 292 |
+
"shortDescription" : "",
|
| 293 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 294 |
+
"name" : "model_model_kv_cache_0",
|
| 295 |
+
"type" : "State"
|
| 296 |
+
}
|
| 297 |
+
],
|
| 298 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 299 |
+
"mlProgramOperationTypeHistogram" : {
|
| 300 |
+
"Ios18.expandDims" : 16,
|
| 301 |
+
"Ios18.mul" : 40,
|
| 302 |
+
"Ios18.matmul" : 8,
|
| 303 |
+
"Identity" : 1,
|
| 304 |
+
"Ios16.reduceMean" : 8,
|
| 305 |
+
"Ios18.exp" : 4,
|
| 306 |
+
"Ios18.realDiv" : 4,
|
| 307 |
+
"Ios18.greaterEqual" : 1,
|
| 308 |
+
"Select" : 1,
|
| 309 |
+
"Ios18.readState" : 9,
|
| 310 |
+
"Tile" : 8,
|
| 311 |
+
"Ios18.gather" : 2,
|
| 312 |
+
"Ios18.add" : 22,
|
| 313 |
+
"Ios18.layerNorm" : 8,
|
| 314 |
+
"Ios18.sliceUpdate" : 8,
|
| 315 |
+
"Ios18.writeState" : 8,
|
| 316 |
+
"Ios18.reshape" : 26,
|
| 317 |
+
"Ios16.reduceMax" : 4,
|
| 318 |
+
"Ios16.reduceSum" : 4,
|
| 319 |
+
"Ios18.constexprLutToDense" : 28,
|
| 320 |
+
"Ios18.conv" : 24,
|
| 321 |
+
"Ios18.concat" : 24,
|
| 322 |
+
"Ios18.transpose" : 16,
|
| 323 |
+
"Ios18.sub" : 20,
|
| 324 |
+
"Ios18.linear" : 4,
|
| 325 |
+
"Ios18.silu" : 4,
|
| 326 |
+
"Ios18.sliceByIndex" : 26,
|
| 327 |
+
"Ios18.squeeze" : 12
|
| 328 |
+
},
|
| 329 |
+
"shortDescription" : "Anemll Model: Multifunction FFN+Prefill",
|
| 330 |
+
"generatedClassName" : "DeepSeek_FFN_PF_lut6_chunk_06of08",
|
| 331 |
+
"author" : "Converted with Anemll v0.2.0",
|
| 332 |
+
"modelType" : {
|
| 333 |
+
"name" : "MLModelType_mlProgram"
|
| 334 |
+
}
|
| 335 |
+
}
|
| 336 |
+
]
|
DeepSeek_FFN_PF_lut6_chunk_06of08.mlmodelc/model.mil
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
DeepSeek_FFN_PF_lut6_chunk_06of08.mlmodelc/weights/weight.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:75be29ca890f52ef61838424278b105843a7ea4c82d439362e017fbb69fa3c39
|
| 3 |
+
size 724250880
|
DeepSeek_FFN_PF_lut6_chunk_07of08.mlmodelc/analytics/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:30d1f3be7ac5c7737c829ea215284af0ce09c6719d2e3799dd7e70e5bc2e2dc3
|
| 3 |
+
size 243
|
DeepSeek_FFN_PF_lut6_chunk_07of08.mlmodelc/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:407fe96eea51e844e91398b5758b47c63915a1ed34502c4bec576da7ce648c86
|
| 3 |
+
size 986
|
DeepSeek_FFN_PF_lut6_chunk_07of08.mlmodelc/metadata.json
ADDED
|
@@ -0,0 +1,336 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"metadataOutputVersion" : "3.0",
|
| 4 |
+
"userDefinedMetadata" : {
|
| 5 |
+
"com.anemll.lut_bits" : "6",
|
| 6 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
| 7 |
+
"com.anemll.context_length" : "1024",
|
| 8 |
+
"com.github.apple.coremltools.source" : "torch==2.5.0",
|
| 9 |
+
"com.github.apple.coremltools.version" : "8.2",
|
| 10 |
+
"com.anemll.num_chunks" : "8",
|
| 11 |
+
"com.anemll.batch_size" : "256",
|
| 12 |
+
"com.anemll.info" : "Converted with Anemll v0.2.0",
|
| 13 |
+
"com.anemll.chunk_no" : "7"
|
| 14 |
+
},
|
| 15 |
+
"availability" : {
|
| 16 |
+
"macOS" : "15.0",
|
| 17 |
+
"tvOS" : "18.0",
|
| 18 |
+
"visionOS" : "2.0",
|
| 19 |
+
"watchOS" : "11.0",
|
| 20 |
+
"iOS" : "18.0",
|
| 21 |
+
"macCatalyst" : "18.0"
|
| 22 |
+
},
|
| 23 |
+
"inputSchema" : [
|
| 24 |
+
{
|
| 25 |
+
"hasShapeFlexibility" : "0",
|
| 26 |
+
"isOptional" : "0",
|
| 27 |
+
"dataType" : "Float16",
|
| 28 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 29 |
+
"shortDescription" : "",
|
| 30 |
+
"shape" : "[1, 1, 4096]",
|
| 31 |
+
"name" : "hidden_states",
|
| 32 |
+
"type" : "MultiArray"
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"hasShapeFlexibility" : "0",
|
| 36 |
+
"isOptional" : "0",
|
| 37 |
+
"dataType" : "Int32",
|
| 38 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 39 |
+
"shortDescription" : "",
|
| 40 |
+
"shape" : "[1]",
|
| 41 |
+
"name" : "position_ids",
|
| 42 |
+
"type" : "MultiArray"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"hasShapeFlexibility" : "0",
|
| 46 |
+
"isOptional" : "0",
|
| 47 |
+
"dataType" : "Float16",
|
| 48 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 49 |
+
"shortDescription" : "",
|
| 50 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 51 |
+
"name" : "causal_mask",
|
| 52 |
+
"type" : "MultiArray"
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"hasShapeFlexibility" : "0",
|
| 56 |
+
"isOptional" : "0",
|
| 57 |
+
"dataType" : "Int32",
|
| 58 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 59 |
+
"shortDescription" : "",
|
| 60 |
+
"shape" : "[1]",
|
| 61 |
+
"name" : "current_pos",
|
| 62 |
+
"type" : "MultiArray"
|
| 63 |
+
}
|
| 64 |
+
],
|
| 65 |
+
"outputSchema" : [
|
| 66 |
+
{
|
| 67 |
+
"hasShapeFlexibility" : "0",
|
| 68 |
+
"isOptional" : "0",
|
| 69 |
+
"dataType" : "Float16",
|
| 70 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 71 |
+
"shortDescription" : "",
|
| 72 |
+
"shape" : "[1, 1, 4096]",
|
| 73 |
+
"name" : "output_hidden_states",
|
| 74 |
+
"type" : "MultiArray"
|
| 75 |
+
}
|
| 76 |
+
],
|
| 77 |
+
"modelParameters" : [
|
| 78 |
+
|
| 79 |
+
],
|
| 80 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 81 |
+
"method" : "predict",
|
| 82 |
+
"functions" : [
|
| 83 |
+
{
|
| 84 |
+
"inputSchema" : [
|
| 85 |
+
{
|
| 86 |
+
"hasShapeFlexibility" : "0",
|
| 87 |
+
"isOptional" : "0",
|
| 88 |
+
"dataType" : "Float16",
|
| 89 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 90 |
+
"shortDescription" : "",
|
| 91 |
+
"shape" : "[1, 1, 4096]",
|
| 92 |
+
"name" : "hidden_states",
|
| 93 |
+
"type" : "MultiArray"
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"hasShapeFlexibility" : "0",
|
| 97 |
+
"isOptional" : "0",
|
| 98 |
+
"dataType" : "Int32",
|
| 99 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 100 |
+
"shortDescription" : "",
|
| 101 |
+
"shape" : "[1]",
|
| 102 |
+
"name" : "position_ids",
|
| 103 |
+
"type" : "MultiArray"
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"hasShapeFlexibility" : "0",
|
| 107 |
+
"isOptional" : "0",
|
| 108 |
+
"dataType" : "Float16",
|
| 109 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 110 |
+
"shortDescription" : "",
|
| 111 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 112 |
+
"name" : "causal_mask",
|
| 113 |
+
"type" : "MultiArray"
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"hasShapeFlexibility" : "0",
|
| 117 |
+
"isOptional" : "0",
|
| 118 |
+
"dataType" : "Int32",
|
| 119 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 120 |
+
"shortDescription" : "",
|
| 121 |
+
"shape" : "[1]",
|
| 122 |
+
"name" : "current_pos",
|
| 123 |
+
"type" : "MultiArray"
|
| 124 |
+
}
|
| 125 |
+
],
|
| 126 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 127 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 128 |
+
"stateSchema" : [
|
| 129 |
+
{
|
| 130 |
+
"dataType" : "Float16",
|
| 131 |
+
"isOptional" : "0",
|
| 132 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 133 |
+
"shortDescription" : "",
|
| 134 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 135 |
+
"name" : "model_model_kv_cache_0",
|
| 136 |
+
"type" : "State"
|
| 137 |
+
}
|
| 138 |
+
],
|
| 139 |
+
"outputSchema" : [
|
| 140 |
+
{
|
| 141 |
+
"hasShapeFlexibility" : "0",
|
| 142 |
+
"isOptional" : "0",
|
| 143 |
+
"dataType" : "Float16",
|
| 144 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 145 |
+
"shortDescription" : "",
|
| 146 |
+
"shape" : "[1, 1, 4096]",
|
| 147 |
+
"name" : "output_hidden_states",
|
| 148 |
+
"type" : "MultiArray"
|
| 149 |
+
}
|
| 150 |
+
],
|
| 151 |
+
"name" : "infer",
|
| 152 |
+
"mlProgramOperationTypeHistogram" : {
|
| 153 |
+
"Ios18.expandDims" : 16,
|
| 154 |
+
"Ios18.mul" : 40,
|
| 155 |
+
"Ios18.matmul" : 8,
|
| 156 |
+
"Identity" : 1,
|
| 157 |
+
"Ios16.reduceMean" : 8,
|
| 158 |
+
"Ios18.exp" : 4,
|
| 159 |
+
"Ios18.realDiv" : 4,
|
| 160 |
+
"Ios18.greaterEqual" : 1,
|
| 161 |
+
"Select" : 1,
|
| 162 |
+
"Ios18.readState" : 9,
|
| 163 |
+
"Tile" : 8,
|
| 164 |
+
"Ios18.gather" : 2,
|
| 165 |
+
"Ios18.add" : 22,
|
| 166 |
+
"Ios18.layerNorm" : 8,
|
| 167 |
+
"Ios18.sliceUpdate" : 8,
|
| 168 |
+
"Ios18.writeState" : 8,
|
| 169 |
+
"Ios18.reshape" : 26,
|
| 170 |
+
"Ios16.reduceMax" : 4,
|
| 171 |
+
"Ios16.reduceSum" : 4,
|
| 172 |
+
"Ios18.constexprLutToDense" : 28,
|
| 173 |
+
"Ios18.conv" : 24,
|
| 174 |
+
"Ios18.concat" : 24,
|
| 175 |
+
"Ios18.transpose" : 16,
|
| 176 |
+
"Ios18.sub" : 20,
|
| 177 |
+
"Ios18.linear" : 4,
|
| 178 |
+
"Ios18.silu" : 4,
|
| 179 |
+
"Ios18.sliceByIndex" : 26,
|
| 180 |
+
"Ios18.squeeze" : 12
|
| 181 |
+
}
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"inputSchema" : [
|
| 185 |
+
{
|
| 186 |
+
"hasShapeFlexibility" : "0",
|
| 187 |
+
"isOptional" : "0",
|
| 188 |
+
"dataType" : "Float16",
|
| 189 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 190 |
+
"shortDescription" : "",
|
| 191 |
+
"shape" : "[1, 256, 4096]",
|
| 192 |
+
"name" : "hidden_states",
|
| 193 |
+
"type" : "MultiArray"
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"hasShapeFlexibility" : "0",
|
| 197 |
+
"isOptional" : "0",
|
| 198 |
+
"dataType" : "Int32",
|
| 199 |
+
"formattedType" : "MultiArray (Int32 256)",
|
| 200 |
+
"shortDescription" : "",
|
| 201 |
+
"shape" : "[256]",
|
| 202 |
+
"name" : "position_ids",
|
| 203 |
+
"type" : "MultiArray"
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"hasShapeFlexibility" : "0",
|
| 207 |
+
"isOptional" : "0",
|
| 208 |
+
"dataType" : "Float16",
|
| 209 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 256 × 1024)",
|
| 210 |
+
"shortDescription" : "",
|
| 211 |
+
"shape" : "[1, 1, 256, 1024]",
|
| 212 |
+
"name" : "causal_mask",
|
| 213 |
+
"type" : "MultiArray"
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"hasShapeFlexibility" : "0",
|
| 217 |
+
"isOptional" : "0",
|
| 218 |
+
"dataType" : "Int32",
|
| 219 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 220 |
+
"shortDescription" : "",
|
| 221 |
+
"shape" : "[1]",
|
| 222 |
+
"name" : "current_pos",
|
| 223 |
+
"type" : "MultiArray"
|
| 224 |
+
}
|
| 225 |
+
],
|
| 226 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 227 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 228 |
+
"stateSchema" : [
|
| 229 |
+
{
|
| 230 |
+
"dataType" : "Float16",
|
| 231 |
+
"isOptional" : "0",
|
| 232 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 233 |
+
"shortDescription" : "",
|
| 234 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 235 |
+
"name" : "model_model_kv_cache_0",
|
| 236 |
+
"type" : "State"
|
| 237 |
+
}
|
| 238 |
+
],
|
| 239 |
+
"outputSchema" : [
|
| 240 |
+
{
|
| 241 |
+
"hasShapeFlexibility" : "0",
|
| 242 |
+
"isOptional" : "0",
|
| 243 |
+
"dataType" : "Float16",
|
| 244 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 245 |
+
"shortDescription" : "",
|
| 246 |
+
"shape" : "[1, 256, 4096]",
|
| 247 |
+
"name" : "output_hidden_states",
|
| 248 |
+
"type" : "MultiArray"
|
| 249 |
+
}
|
| 250 |
+
],
|
| 251 |
+
"name" : "prefill",
|
| 252 |
+
"mlProgramOperationTypeHistogram" : {
|
| 253 |
+
"Ios18.expandDims" : 16,
|
| 254 |
+
"Ios18.mul" : 40,
|
| 255 |
+
"Ios18.matmul" : 8,
|
| 256 |
+
"Ios16.reduceMean" : 8,
|
| 257 |
+
"Ios18.exp" : 4,
|
| 258 |
+
"Ios18.realDiv" : 4,
|
| 259 |
+
"Ios18.greaterEqual" : 1,
|
| 260 |
+
"Select" : 1,
|
| 261 |
+
"Ios18.readState" : 9,
|
| 262 |
+
"Tile" : 8,
|
| 263 |
+
"Ios18.gather" : 2,
|
| 264 |
+
"Ios18.add" : 22,
|
| 265 |
+
"Ios18.layerNorm" : 8,
|
| 266 |
+
"Ios18.sliceUpdate" : 8,
|
| 267 |
+
"Ios18.writeState" : 8,
|
| 268 |
+
"Ios18.reshape" : 34,
|
| 269 |
+
"Ios16.reduceMax" : 4,
|
| 270 |
+
"Ios16.reduceSum" : 4,
|
| 271 |
+
"Ios18.constexprLutToDense" : 28,
|
| 272 |
+
"Ios18.conv" : 24,
|
| 273 |
+
"Ios18.concat" : 24,
|
| 274 |
+
"Ios18.transpose" : 30,
|
| 275 |
+
"Ios18.sub" : 20,
|
| 276 |
+
"Ios18.linear" : 4,
|
| 277 |
+
"Ios18.silu" : 4,
|
| 278 |
+
"Ios18.sliceByIndex" : 26,
|
| 279 |
+
"Ios18.squeeze" : 12
|
| 280 |
+
}
|
| 281 |
+
}
|
| 282 |
+
],
|
| 283 |
+
"version" : "0.2.0",
|
| 284 |
+
"isUpdatable" : "0",
|
| 285 |
+
"defaultFunctionName" : "infer",
|
| 286 |
+
"specificationVersion" : 9,
|
| 287 |
+
"stateSchema" : [
|
| 288 |
+
{
|
| 289 |
+
"dataType" : "Float16",
|
| 290 |
+
"isOptional" : "0",
|
| 291 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 292 |
+
"shortDescription" : "",
|
| 293 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 294 |
+
"name" : "model_model_kv_cache_0",
|
| 295 |
+
"type" : "State"
|
| 296 |
+
}
|
| 297 |
+
],
|
| 298 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 299 |
+
"mlProgramOperationTypeHistogram" : {
|
| 300 |
+
"Ios18.expandDims" : 16,
|
| 301 |
+
"Ios18.mul" : 40,
|
| 302 |
+
"Ios18.matmul" : 8,
|
| 303 |
+
"Identity" : 1,
|
| 304 |
+
"Ios16.reduceMean" : 8,
|
| 305 |
+
"Ios18.exp" : 4,
|
| 306 |
+
"Ios18.realDiv" : 4,
|
| 307 |
+
"Ios18.greaterEqual" : 1,
|
| 308 |
+
"Select" : 1,
|
| 309 |
+
"Ios18.readState" : 9,
|
| 310 |
+
"Tile" : 8,
|
| 311 |
+
"Ios18.gather" : 2,
|
| 312 |
+
"Ios18.add" : 22,
|
| 313 |
+
"Ios18.layerNorm" : 8,
|
| 314 |
+
"Ios18.sliceUpdate" : 8,
|
| 315 |
+
"Ios18.writeState" : 8,
|
| 316 |
+
"Ios18.reshape" : 26,
|
| 317 |
+
"Ios16.reduceMax" : 4,
|
| 318 |
+
"Ios16.reduceSum" : 4,
|
| 319 |
+
"Ios18.constexprLutToDense" : 28,
|
| 320 |
+
"Ios18.conv" : 24,
|
| 321 |
+
"Ios18.concat" : 24,
|
| 322 |
+
"Ios18.transpose" : 16,
|
| 323 |
+
"Ios18.sub" : 20,
|
| 324 |
+
"Ios18.linear" : 4,
|
| 325 |
+
"Ios18.silu" : 4,
|
| 326 |
+
"Ios18.sliceByIndex" : 26,
|
| 327 |
+
"Ios18.squeeze" : 12
|
| 328 |
+
},
|
| 329 |
+
"shortDescription" : "Anemll Model: Multifunction FFN+Prefill",
|
| 330 |
+
"generatedClassName" : "DeepSeek_FFN_PF_lut6_chunk_07of08",
|
| 331 |
+
"author" : "Converted with Anemll v0.2.0",
|
| 332 |
+
"modelType" : {
|
| 333 |
+
"name" : "MLModelType_mlProgram"
|
| 334 |
+
}
|
| 335 |
+
}
|
| 336 |
+
]
|
DeepSeek_FFN_PF_lut6_chunk_07of08.mlmodelc/model.mil
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
DeepSeek_FFN_PF_lut6_chunk_07of08.mlmodelc/weights/weight.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fb4da9a7c833ac64910d0e584a36816b35e3c4729bcc73fcdd4ea42ca5bfb997
|
| 3 |
+
size 724250880
|
DeepSeek_FFN_PF_lut6_chunk_08of08.mlmodelc/analytics/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:36d47258a982906bcdbc6614d35edf78270dae850548645167296cd251ee93ab
|
| 3 |
+
size 243
|
DeepSeek_FFN_PF_lut6_chunk_08of08.mlmodelc/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a6881f709589ef69ecb988f10e8061e0654e8c3c59d787b92951f9372e77d3e
|
| 3 |
+
size 985
|
DeepSeek_FFN_PF_lut6_chunk_08of08.mlmodelc/metadata.json
ADDED
|
@@ -0,0 +1,336 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"metadataOutputVersion" : "3.0",
|
| 4 |
+
"userDefinedMetadata" : {
|
| 5 |
+
"com.github.apple.coremltools.source" : "torch==2.5.0",
|
| 6 |
+
"com.github.apple.coremltools.version" : "8.2",
|
| 7 |
+
"com.anemll.context_length" : "1024",
|
| 8 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
| 9 |
+
"com.anemll.chunk_no" : "8",
|
| 10 |
+
"com.anemll.num_chunks" : "8",
|
| 11 |
+
"com.anemll.info" : "Converted with Anemll v0.2.0",
|
| 12 |
+
"com.anemll.batch_size" : "256",
|
| 13 |
+
"com.anemll.lut_bits" : "6"
|
| 14 |
+
},
|
| 15 |
+
"availability" : {
|
| 16 |
+
"macOS" : "15.0",
|
| 17 |
+
"tvOS" : "18.0",
|
| 18 |
+
"visionOS" : "2.0",
|
| 19 |
+
"watchOS" : "11.0",
|
| 20 |
+
"iOS" : "18.0",
|
| 21 |
+
"macCatalyst" : "18.0"
|
| 22 |
+
},
|
| 23 |
+
"inputSchema" : [
|
| 24 |
+
{
|
| 25 |
+
"hasShapeFlexibility" : "0",
|
| 26 |
+
"isOptional" : "0",
|
| 27 |
+
"dataType" : "Float16",
|
| 28 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 29 |
+
"shortDescription" : "",
|
| 30 |
+
"shape" : "[1, 1, 4096]",
|
| 31 |
+
"name" : "hidden_states",
|
| 32 |
+
"type" : "MultiArray"
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"hasShapeFlexibility" : "0",
|
| 36 |
+
"isOptional" : "0",
|
| 37 |
+
"dataType" : "Int32",
|
| 38 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 39 |
+
"shortDescription" : "",
|
| 40 |
+
"shape" : "[1]",
|
| 41 |
+
"name" : "position_ids",
|
| 42 |
+
"type" : "MultiArray"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"hasShapeFlexibility" : "0",
|
| 46 |
+
"isOptional" : "0",
|
| 47 |
+
"dataType" : "Float16",
|
| 48 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 49 |
+
"shortDescription" : "",
|
| 50 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 51 |
+
"name" : "causal_mask",
|
| 52 |
+
"type" : "MultiArray"
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"hasShapeFlexibility" : "0",
|
| 56 |
+
"isOptional" : "0",
|
| 57 |
+
"dataType" : "Int32",
|
| 58 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 59 |
+
"shortDescription" : "",
|
| 60 |
+
"shape" : "[1]",
|
| 61 |
+
"name" : "current_pos",
|
| 62 |
+
"type" : "MultiArray"
|
| 63 |
+
}
|
| 64 |
+
],
|
| 65 |
+
"outputSchema" : [
|
| 66 |
+
{
|
| 67 |
+
"hasShapeFlexibility" : "0",
|
| 68 |
+
"isOptional" : "0",
|
| 69 |
+
"dataType" : "Float16",
|
| 70 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 71 |
+
"shortDescription" : "",
|
| 72 |
+
"shape" : "[1, 1, 4096]",
|
| 73 |
+
"name" : "output_hidden_states",
|
| 74 |
+
"type" : "MultiArray"
|
| 75 |
+
}
|
| 76 |
+
],
|
| 77 |
+
"modelParameters" : [
|
| 78 |
+
|
| 79 |
+
],
|
| 80 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 81 |
+
"method" : "predict",
|
| 82 |
+
"functions" : [
|
| 83 |
+
{
|
| 84 |
+
"inputSchema" : [
|
| 85 |
+
{
|
| 86 |
+
"hasShapeFlexibility" : "0",
|
| 87 |
+
"isOptional" : "0",
|
| 88 |
+
"dataType" : "Float16",
|
| 89 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 90 |
+
"shortDescription" : "",
|
| 91 |
+
"shape" : "[1, 1, 4096]",
|
| 92 |
+
"name" : "hidden_states",
|
| 93 |
+
"type" : "MultiArray"
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"hasShapeFlexibility" : "0",
|
| 97 |
+
"isOptional" : "0",
|
| 98 |
+
"dataType" : "Int32",
|
| 99 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 100 |
+
"shortDescription" : "",
|
| 101 |
+
"shape" : "[1]",
|
| 102 |
+
"name" : "position_ids",
|
| 103 |
+
"type" : "MultiArray"
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"hasShapeFlexibility" : "0",
|
| 107 |
+
"isOptional" : "0",
|
| 108 |
+
"dataType" : "Float16",
|
| 109 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1024)",
|
| 110 |
+
"shortDescription" : "",
|
| 111 |
+
"shape" : "[1, 1, 1, 1024]",
|
| 112 |
+
"name" : "causal_mask",
|
| 113 |
+
"type" : "MultiArray"
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"hasShapeFlexibility" : "0",
|
| 117 |
+
"isOptional" : "0",
|
| 118 |
+
"dataType" : "Int32",
|
| 119 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 120 |
+
"shortDescription" : "",
|
| 121 |
+
"shape" : "[1]",
|
| 122 |
+
"name" : "current_pos",
|
| 123 |
+
"type" : "MultiArray"
|
| 124 |
+
}
|
| 125 |
+
],
|
| 126 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 127 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 128 |
+
"stateSchema" : [
|
| 129 |
+
{
|
| 130 |
+
"dataType" : "Float16",
|
| 131 |
+
"isOptional" : "0",
|
| 132 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 133 |
+
"shortDescription" : "",
|
| 134 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 135 |
+
"name" : "model_model_kv_cache_0",
|
| 136 |
+
"type" : "State"
|
| 137 |
+
}
|
| 138 |
+
],
|
| 139 |
+
"outputSchema" : [
|
| 140 |
+
{
|
| 141 |
+
"hasShapeFlexibility" : "0",
|
| 142 |
+
"isOptional" : "0",
|
| 143 |
+
"dataType" : "Float16",
|
| 144 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 145 |
+
"shortDescription" : "",
|
| 146 |
+
"shape" : "[1, 1, 4096]",
|
| 147 |
+
"name" : "output_hidden_states",
|
| 148 |
+
"type" : "MultiArray"
|
| 149 |
+
}
|
| 150 |
+
],
|
| 151 |
+
"name" : "infer",
|
| 152 |
+
"mlProgramOperationTypeHistogram" : {
|
| 153 |
+
"Ios18.expandDims" : 16,
|
| 154 |
+
"Ios18.mul" : 40,
|
| 155 |
+
"Ios18.matmul" : 8,
|
| 156 |
+
"Identity" : 1,
|
| 157 |
+
"Ios16.reduceMean" : 9,
|
| 158 |
+
"Ios18.exp" : 4,
|
| 159 |
+
"Ios18.realDiv" : 4,
|
| 160 |
+
"Ios18.greaterEqual" : 1,
|
| 161 |
+
"Select" : 1,
|
| 162 |
+
"Ios18.readState" : 9,
|
| 163 |
+
"Tile" : 8,
|
| 164 |
+
"Ios18.gather" : 2,
|
| 165 |
+
"Ios18.add" : 22,
|
| 166 |
+
"Ios18.layerNorm" : 9,
|
| 167 |
+
"Ios18.sliceUpdate" : 8,
|
| 168 |
+
"Ios18.writeState" : 8,
|
| 169 |
+
"Ios18.reshape" : 26,
|
| 170 |
+
"Ios16.reduceMax" : 4,
|
| 171 |
+
"Ios16.reduceSum" : 4,
|
| 172 |
+
"Ios18.constexprLutToDense" : 28,
|
| 173 |
+
"Ios18.conv" : 24,
|
| 174 |
+
"Ios18.concat" : 24,
|
| 175 |
+
"Ios18.transpose" : 16,
|
| 176 |
+
"Ios18.sub" : 21,
|
| 177 |
+
"Ios18.linear" : 4,
|
| 178 |
+
"Ios18.silu" : 4,
|
| 179 |
+
"Ios18.sliceByIndex" : 26,
|
| 180 |
+
"Ios18.squeeze" : 12
|
| 181 |
+
}
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"inputSchema" : [
|
| 185 |
+
{
|
| 186 |
+
"hasShapeFlexibility" : "0",
|
| 187 |
+
"isOptional" : "0",
|
| 188 |
+
"dataType" : "Float16",
|
| 189 |
+
"formattedType" : "MultiArray (Float16 1 × 256 × 4096)",
|
| 190 |
+
"shortDescription" : "",
|
| 191 |
+
"shape" : "[1, 256, 4096]",
|
| 192 |
+
"name" : "hidden_states",
|
| 193 |
+
"type" : "MultiArray"
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"hasShapeFlexibility" : "0",
|
| 197 |
+
"isOptional" : "0",
|
| 198 |
+
"dataType" : "Int32",
|
| 199 |
+
"formattedType" : "MultiArray (Int32 256)",
|
| 200 |
+
"shortDescription" : "",
|
| 201 |
+
"shape" : "[256]",
|
| 202 |
+
"name" : "position_ids",
|
| 203 |
+
"type" : "MultiArray"
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"hasShapeFlexibility" : "0",
|
| 207 |
+
"isOptional" : "0",
|
| 208 |
+
"dataType" : "Float16",
|
| 209 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 256 × 1024)",
|
| 210 |
+
"shortDescription" : "",
|
| 211 |
+
"shape" : "[1, 1, 256, 1024]",
|
| 212 |
+
"name" : "causal_mask",
|
| 213 |
+
"type" : "MultiArray"
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"hasShapeFlexibility" : "0",
|
| 217 |
+
"isOptional" : "0",
|
| 218 |
+
"dataType" : "Int32",
|
| 219 |
+
"formattedType" : "MultiArray (Int32 1)",
|
| 220 |
+
"shortDescription" : "",
|
| 221 |
+
"shape" : "[1]",
|
| 222 |
+
"name" : "current_pos",
|
| 223 |
+
"type" : "MultiArray"
|
| 224 |
+
}
|
| 225 |
+
],
|
| 226 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 227 |
+
"storagePrecision" : "Mixed (Float16, Palettized (13 bits), Palettized (15 bits), Palettized (17 bits))",
|
| 228 |
+
"stateSchema" : [
|
| 229 |
+
{
|
| 230 |
+
"dataType" : "Float16",
|
| 231 |
+
"isOptional" : "0",
|
| 232 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 233 |
+
"shortDescription" : "",
|
| 234 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 235 |
+
"name" : "model_model_kv_cache_0",
|
| 236 |
+
"type" : "State"
|
| 237 |
+
}
|
| 238 |
+
],
|
| 239 |
+
"outputSchema" : [
|
| 240 |
+
{
|
| 241 |
+
"hasShapeFlexibility" : "0",
|
| 242 |
+
"isOptional" : "0",
|
| 243 |
+
"dataType" : "Float16",
|
| 244 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 245 |
+
"shortDescription" : "",
|
| 246 |
+
"shape" : "[1, 1, 4096]",
|
| 247 |
+
"name" : "output_hidden_states",
|
| 248 |
+
"type" : "MultiArray"
|
| 249 |
+
}
|
| 250 |
+
],
|
| 251 |
+
"name" : "prefill",
|
| 252 |
+
"mlProgramOperationTypeHistogram" : {
|
| 253 |
+
"Ios18.expandDims" : 15,
|
| 254 |
+
"Ios18.mul" : 39,
|
| 255 |
+
"Ios18.matmul" : 8,
|
| 256 |
+
"Ios16.reduceMean" : 7,
|
| 257 |
+
"Ios18.exp" : 4,
|
| 258 |
+
"Ios18.realDiv" : 4,
|
| 259 |
+
"Ios18.greaterEqual" : 1,
|
| 260 |
+
"Select" : 1,
|
| 261 |
+
"Ios18.readState" : 9,
|
| 262 |
+
"Tile" : 8,
|
| 263 |
+
"Ios18.gather" : 2,
|
| 264 |
+
"Ios18.add" : 21,
|
| 265 |
+
"Ios18.layerNorm" : 7,
|
| 266 |
+
"Ios18.sliceUpdate" : 8,
|
| 267 |
+
"Ios18.writeState" : 8,
|
| 268 |
+
"Ios18.reshape" : 34,
|
| 269 |
+
"Ios16.reduceMax" : 4,
|
| 270 |
+
"Ios16.reduceSum" : 4,
|
| 271 |
+
"Ios18.constexprLutToDense" : 25,
|
| 272 |
+
"Ios18.conv" : 21,
|
| 273 |
+
"Ios18.concat" : 24,
|
| 274 |
+
"Ios18.transpose" : 28,
|
| 275 |
+
"Ios18.sub" : 19,
|
| 276 |
+
"Ios18.linear" : 4,
|
| 277 |
+
"Ios18.silu" : 3,
|
| 278 |
+
"Ios18.sliceByIndex" : 27,
|
| 279 |
+
"Ios18.squeeze" : 11
|
| 280 |
+
}
|
| 281 |
+
}
|
| 282 |
+
],
|
| 283 |
+
"version" : "0.2.0",
|
| 284 |
+
"isUpdatable" : "0",
|
| 285 |
+
"defaultFunctionName" : "infer",
|
| 286 |
+
"specificationVersion" : 9,
|
| 287 |
+
"stateSchema" : [
|
| 288 |
+
{
|
| 289 |
+
"dataType" : "Float16",
|
| 290 |
+
"isOptional" : "0",
|
| 291 |
+
"formattedType" : "State (Float16 64 × 8 × 1024 × 128)",
|
| 292 |
+
"shortDescription" : "",
|
| 293 |
+
"shape" : "[64, 8, 1024, 128]",
|
| 294 |
+
"name" : "model_model_kv_cache_0",
|
| 295 |
+
"type" : "State"
|
| 296 |
+
}
|
| 297 |
+
],
|
| 298 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 299 |
+
"mlProgramOperationTypeHistogram" : {
|
| 300 |
+
"Ios18.expandDims" : 16,
|
| 301 |
+
"Ios18.mul" : 40,
|
| 302 |
+
"Ios18.matmul" : 8,
|
| 303 |
+
"Identity" : 1,
|
| 304 |
+
"Ios16.reduceMean" : 9,
|
| 305 |
+
"Ios18.exp" : 4,
|
| 306 |
+
"Ios18.realDiv" : 4,
|
| 307 |
+
"Ios18.greaterEqual" : 1,
|
| 308 |
+
"Select" : 1,
|
| 309 |
+
"Ios18.readState" : 9,
|
| 310 |
+
"Tile" : 8,
|
| 311 |
+
"Ios18.gather" : 2,
|
| 312 |
+
"Ios18.add" : 22,
|
| 313 |
+
"Ios18.layerNorm" : 9,
|
| 314 |
+
"Ios18.sliceUpdate" : 8,
|
| 315 |
+
"Ios18.writeState" : 8,
|
| 316 |
+
"Ios18.reshape" : 26,
|
| 317 |
+
"Ios16.reduceMax" : 4,
|
| 318 |
+
"Ios16.reduceSum" : 4,
|
| 319 |
+
"Ios18.constexprLutToDense" : 28,
|
| 320 |
+
"Ios18.conv" : 24,
|
| 321 |
+
"Ios18.concat" : 24,
|
| 322 |
+
"Ios18.transpose" : 16,
|
| 323 |
+
"Ios18.sub" : 21,
|
| 324 |
+
"Ios18.linear" : 4,
|
| 325 |
+
"Ios18.silu" : 4,
|
| 326 |
+
"Ios18.sliceByIndex" : 26,
|
| 327 |
+
"Ios18.squeeze" : 12
|
| 328 |
+
},
|
| 329 |
+
"shortDescription" : "Anemll Model: Multifunction FFN+Prefill",
|
| 330 |
+
"generatedClassName" : "DeepSeek_FFN_PF_lut6_chunk_08of08",
|
| 331 |
+
"author" : "Converted with Anemll v0.2.0",
|
| 332 |
+
"modelType" : {
|
| 333 |
+
"name" : "MLModelType_mlProgram"
|
| 334 |
+
}
|
| 335 |
+
}
|
| 336 |
+
]
|
DeepSeek_FFN_PF_lut6_chunk_08of08.mlmodelc/model.mil
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
DeepSeek_FFN_PF_lut6_chunk_08of08.mlmodelc/weights/weight.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cfb71b9c45f059d3d74f7dbf23c07946dac0a871df197ce3851b4d6bb81e6bcc
|
| 3 |
+
size 724259136
|
DeepSeek_embeddings.mlmodelc/analytics/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0d8c55ca24d1412f6bcc309abf969a35376272863bcc1fa10c74538151094b3
|
| 3 |
+
size 243
|
DeepSeek_embeddings.mlmodelc/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f34b2aaac71e780225a382e1d98e22992f969db563f071013a60a1ce5dcee4c5
|
| 3 |
+
size 502
|
DeepSeek_embeddings.mlmodelc/metadata.json
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"shortDescription" : "Anemll Model (Embeddings) converted to CoreML",
|
| 4 |
+
"metadataOutputVersion" : "3.0",
|
| 5 |
+
"outputSchema" : [
|
| 6 |
+
{
|
| 7 |
+
"hasShapeFlexibility" : "0",
|
| 8 |
+
"isOptional" : "0",
|
| 9 |
+
"dataType" : "Float16",
|
| 10 |
+
"formattedType" : "MultiArray (Float16)",
|
| 11 |
+
"shortDescription" : "",
|
| 12 |
+
"shape" : "[]",
|
| 13 |
+
"name" : "hidden_states",
|
| 14 |
+
"type" : "MultiArray"
|
| 15 |
+
}
|
| 16 |
+
],
|
| 17 |
+
"version" : "0.2.0",
|
| 18 |
+
"modelParameters" : [
|
| 19 |
+
|
| 20 |
+
],
|
| 21 |
+
"author" : "Converted with Anemll v0.2.0",
|
| 22 |
+
"specificationVersion" : 9,
|
| 23 |
+
"storagePrecision" : "Float16",
|
| 24 |
+
"mlProgramOperationTypeHistogram" : {
|
| 25 |
+
"Ios18.gather" : 1
|
| 26 |
+
},
|
| 27 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 28 |
+
"stateSchema" : [
|
| 29 |
+
|
| 30 |
+
],
|
| 31 |
+
"isUpdatable" : "0",
|
| 32 |
+
"availability" : {
|
| 33 |
+
"macOS" : "15.0",
|
| 34 |
+
"tvOS" : "18.0",
|
| 35 |
+
"visionOS" : "2.0",
|
| 36 |
+
"watchOS" : "11.0",
|
| 37 |
+
"iOS" : "18.0",
|
| 38 |
+
"macCatalyst" : "18.0"
|
| 39 |
+
},
|
| 40 |
+
"modelType" : {
|
| 41 |
+
"name" : "MLModelType_mlProgram"
|
| 42 |
+
},
|
| 43 |
+
"inputSchema" : [
|
| 44 |
+
{
|
| 45 |
+
"shortDescription" : "",
|
| 46 |
+
"dataType" : "Int32",
|
| 47 |
+
"hasShapeFlexibility" : "1",
|
| 48 |
+
"isOptional" : "0",
|
| 49 |
+
"shapeFlexibility" : "1 × 1 | 1 × 256",
|
| 50 |
+
"formattedType" : "MultiArray (Int32 1 × 1)",
|
| 51 |
+
"type" : "MultiArray",
|
| 52 |
+
"shape" : "[1, 1]",
|
| 53 |
+
"name" : "input_ids",
|
| 54 |
+
"enumeratedShapes" : "[[1, 1], [1, 256]]"
|
| 55 |
+
}
|
| 56 |
+
],
|
| 57 |
+
"userDefinedMetadata" : {
|
| 58 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
| 59 |
+
"com.github.apple.coremltools.version" : "8.0b2",
|
| 60 |
+
"com.github.apple.coremltools.source" : "torch==2.2.0",
|
| 61 |
+
"com.anemll.info" : "Converted with Anemll v0.2.0",
|
| 62 |
+
"com.anemll.context_length" : "1024"
|
| 63 |
+
},
|
| 64 |
+
"generatedClassName" : "DeepSeek_embeddings",
|
| 65 |
+
"method" : "predict"
|
| 66 |
+
}
|
| 67 |
+
]
|
DeepSeek_embeddings.mlmodelc/model.mil
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
program(1.3)
|
| 2 |
+
[buildInfo = dict<string, string>({{"coremlc-component-MIL", "3404.16.1"}, {"coremlc-version", "3404.23.1"}, {"coremltools-component-torch", "2.2.0"}, {"coremltools-source-dialect", "TorchScript"}, {"coremltools-version", "8.0b2"}})]
|
| 3 |
+
{
|
| 4 |
+
func main<ios18>(tensor<int32, [1, ?]> input_ids) [FlexibleShapeInformation = tuple<tuple<string, dict<string, tensor<int32, [?]>>>, tuple<string, dict<string, dict<string, tensor<int32, [?]>>>>>((("DefaultShapes", {{"input_ids", [1, 1]}}), ("EnumeratedShapes", {{"79ae981e", {{"input_ids", [1, 1]}}}, {"c09fdef5", {{"input_ids", [1, 256]}}}})))] {
|
| 5 |
+
int32 hidden_states_axis_0 = const()[name = string("hidden_states_axis_0"), val = int32(0)];
|
| 6 |
+
int32 hidden_states_batch_dims_0 = const()[name = string("hidden_states_batch_dims_0"), val = int32(0)];
|
| 7 |
+
bool hidden_states_validate_indices_0 = const()[name = string("hidden_states_validate_indices_0"), val = bool(false)];
|
| 8 |
+
tensor<fp16, [128256, 4096]> embed_tokens_weight_to_fp16 = const()[name = string("embed_tokens_weight_to_fp16"), val = tensor<fp16, [128256, 4096]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(64)))];
|
| 9 |
+
tensor<fp16, [1, ?, 4096]> hidden_states = gather(axis = hidden_states_axis_0, batch_dims = hidden_states_batch_dims_0, indices = input_ids, validate_indices = hidden_states_validate_indices_0, x = embed_tokens_weight_to_fp16)[name = string("hidden_states_cast_fp16")];
|
| 10 |
+
} -> (hidden_states);
|
| 11 |
+
}
|
DeepSeek_embeddings.mlmodelc/weights/weight.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:358fab4c0be122bdeccd0b39884fe0b6adc561d8bf56a676937e8e96b0a30c13
|
| 3 |
+
size 1050673280
|
DeepSeek_lm_head_lut6.mlmodelc/analytics/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b28b8755a8a5458892a0b062a469917905cc4c76101c8ff25f8541f8d9acbafc
|
| 3 |
+
size 243
|
DeepSeek_lm_head_lut6.mlmodelc/coremldata.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:543e78c6b96253d30bc6c16dbf88b8d9b9fbb7eff7e3e1f3c6b5f2bd3925f93a
|
| 3 |
+
size 689
|
DeepSeek_lm_head_lut6.mlmodelc/metadata.json
ADDED
|
@@ -0,0 +1,139 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"shortDescription" : "Anemll Model (LM Head) converted to CoreML",
|
| 4 |
+
"metadataOutputVersion" : "3.0",
|
| 5 |
+
"outputSchema" : [
|
| 6 |
+
{
|
| 7 |
+
"hasShapeFlexibility" : "0",
|
| 8 |
+
"isOptional" : "0",
|
| 9 |
+
"dataType" : "Float16",
|
| 10 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 16032)",
|
| 11 |
+
"shortDescription" : "",
|
| 12 |
+
"shape" : "[1, 1, 16032]",
|
| 13 |
+
"name" : "logits1",
|
| 14 |
+
"type" : "MultiArray"
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"hasShapeFlexibility" : "0",
|
| 18 |
+
"isOptional" : "0",
|
| 19 |
+
"dataType" : "Float16",
|
| 20 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 16032)",
|
| 21 |
+
"shortDescription" : "",
|
| 22 |
+
"shape" : "[1, 1, 16032]",
|
| 23 |
+
"name" : "logits2",
|
| 24 |
+
"type" : "MultiArray"
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"hasShapeFlexibility" : "0",
|
| 28 |
+
"isOptional" : "0",
|
| 29 |
+
"dataType" : "Float16",
|
| 30 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 16032)",
|
| 31 |
+
"shortDescription" : "",
|
| 32 |
+
"shape" : "[1, 1, 16032]",
|
| 33 |
+
"name" : "logits3",
|
| 34 |
+
"type" : "MultiArray"
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"hasShapeFlexibility" : "0",
|
| 38 |
+
"isOptional" : "0",
|
| 39 |
+
"dataType" : "Float16",
|
| 40 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 16032)",
|
| 41 |
+
"shortDescription" : "",
|
| 42 |
+
"shape" : "[1, 1, 16032]",
|
| 43 |
+
"name" : "logits4",
|
| 44 |
+
"type" : "MultiArray"
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"hasShapeFlexibility" : "0",
|
| 48 |
+
"isOptional" : "0",
|
| 49 |
+
"dataType" : "Float16",
|
| 50 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 16032)",
|
| 51 |
+
"shortDescription" : "",
|
| 52 |
+
"shape" : "[1, 1, 16032]",
|
| 53 |
+
"name" : "logits5",
|
| 54 |
+
"type" : "MultiArray"
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"hasShapeFlexibility" : "0",
|
| 58 |
+
"isOptional" : "0",
|
| 59 |
+
"dataType" : "Float16",
|
| 60 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 16032)",
|
| 61 |
+
"shortDescription" : "",
|
| 62 |
+
"shape" : "[1, 1, 16032]",
|
| 63 |
+
"name" : "logits6",
|
| 64 |
+
"type" : "MultiArray"
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"hasShapeFlexibility" : "0",
|
| 68 |
+
"isOptional" : "0",
|
| 69 |
+
"dataType" : "Float16",
|
| 70 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 16032)",
|
| 71 |
+
"shortDescription" : "",
|
| 72 |
+
"shape" : "[1, 1, 16032]",
|
| 73 |
+
"name" : "logits7",
|
| 74 |
+
"type" : "MultiArray"
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"hasShapeFlexibility" : "0",
|
| 78 |
+
"isOptional" : "0",
|
| 79 |
+
"dataType" : "Float16",
|
| 80 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 16032)",
|
| 81 |
+
"shortDescription" : "",
|
| 82 |
+
"shape" : "[1, 1, 16032]",
|
| 83 |
+
"name" : "logits8",
|
| 84 |
+
"type" : "MultiArray"
|
| 85 |
+
}
|
| 86 |
+
],
|
| 87 |
+
"version" : "0.2.0",
|
| 88 |
+
"modelParameters" : [
|
| 89 |
+
|
| 90 |
+
],
|
| 91 |
+
"author" : "Converted with Anemll v0.2.0",
|
| 92 |
+
"specificationVersion" : 9,
|
| 93 |
+
"storagePrecision" : "Float16",
|
| 94 |
+
"mlProgramOperationTypeHistogram" : {
|
| 95 |
+
"Ios18.transpose" : 9,
|
| 96 |
+
"Ios18.expandDims" : 1,
|
| 97 |
+
"Ios18.conv" : 8,
|
| 98 |
+
"Ios18.squeeze" : 8
|
| 99 |
+
},
|
| 100 |
+
"computePrecision" : "Mixed (Float16, Int32)",
|
| 101 |
+
"stateSchema" : [
|
| 102 |
+
|
| 103 |
+
],
|
| 104 |
+
"isUpdatable" : "0",
|
| 105 |
+
"availability" : {
|
| 106 |
+
"macOS" : "15.0",
|
| 107 |
+
"tvOS" : "18.0",
|
| 108 |
+
"visionOS" : "2.0",
|
| 109 |
+
"watchOS" : "11.0",
|
| 110 |
+
"iOS" : "18.0",
|
| 111 |
+
"macCatalyst" : "18.0"
|
| 112 |
+
},
|
| 113 |
+
"modelType" : {
|
| 114 |
+
"name" : "MLModelType_mlProgram"
|
| 115 |
+
},
|
| 116 |
+
"inputSchema" : [
|
| 117 |
+
{
|
| 118 |
+
"hasShapeFlexibility" : "0",
|
| 119 |
+
"isOptional" : "0",
|
| 120 |
+
"dataType" : "Float16",
|
| 121 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 4096)",
|
| 122 |
+
"shortDescription" : "",
|
| 123 |
+
"shape" : "[1, 1, 4096]",
|
| 124 |
+
"name" : "hidden_states",
|
| 125 |
+
"type" : "MultiArray"
|
| 126 |
+
}
|
| 127 |
+
],
|
| 128 |
+
"userDefinedMetadata" : {
|
| 129 |
+
"com.anemll.info" : "Converted with Anemll v0.2.0",
|
| 130 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
| 131 |
+
"com.anemll.lut_bits" : "6",
|
| 132 |
+
"com.github.apple.coremltools.source" : "torch==2.5.0",
|
| 133 |
+
"com.github.apple.coremltools.version" : "8.2",
|
| 134 |
+
"com.anemll.context_length" : "1024"
|
| 135 |
+
},
|
| 136 |
+
"generatedClassName" : "DeepSeek_lm_head_lut6",
|
| 137 |
+
"method" : "predict"
|
| 138 |
+
}
|
| 139 |
+
]
|
DeepSeek_lm_head_lut6.mlmodelc/model.mil
ADDED
|
@@ -0,0 +1,98 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
program(1.3)
|
| 2 |
+
[buildInfo = dict<string, string>({{"coremlc-component-MIL", "3404.16.1"}, {"coremlc-version", "3404.23.1"}, {"coremltools-component-torch", "2.5.0"}, {"coremltools-source-dialect", "TorchScript"}, {"coremltools-version", "8.2"}})]
|
| 3 |
+
{
|
| 4 |
+
func main<ios18>(tensor<fp16, [1, 1, 4096]> hidden_states) {
|
| 5 |
+
tensor<int32, [3]> var_5 = const()[name = string("op_5"), val = tensor<int32, [3]>([0, 2, 1])];
|
| 6 |
+
tensor<int32, [1]> input_axes_0 = const()[name = string("input_axes_0"), val = tensor<int32, [1]>([2])];
|
| 7 |
+
tensor<fp16, [1, 4096, 1]> var_6_cast_fp16 = transpose(perm = var_5, x = hidden_states)[name = string("transpose_8")];
|
| 8 |
+
tensor<fp16, [1, 4096, 1, 1]> input_cast_fp16 = expand_dims(axes = input_axes_0, x = var_6_cast_fp16)[name = string("input_cast_fp16")];
|
| 9 |
+
string var_29_pad_type_0 = const()[name = string("op_29_pad_type_0"), val = string("valid")];
|
| 10 |
+
tensor<int32, [2]> var_29_strides_0 = const()[name = string("op_29_strides_0"), val = tensor<int32, [2]>([1, 1])];
|
| 11 |
+
tensor<int32, [4]> var_29_pad_0 = const()[name = string("op_29_pad_0"), val = tensor<int32, [4]>([0, 0, 0, 0])];
|
| 12 |
+
tensor<int32, [2]> var_29_dilations_0 = const()[name = string("op_29_dilations_0"), val = tensor<int32, [2]>([1, 1])];
|
| 13 |
+
int32 var_29_groups_0 = const()[name = string("op_29_groups_0"), val = int32(1)];
|
| 14 |
+
tensor<fp16, [16032, 4096, 1, 1]> var_9_promoted_to_fp16 = const()[name = string("op_9_promoted_to_fp16"), val = tensor<fp16, [16032, 4096, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(64)))];
|
| 15 |
+
tensor<fp16, [1, 16032, 1, 1]> var_29_cast_fp16 = conv(dilations = var_29_dilations_0, groups = var_29_groups_0, pad = var_29_pad_0, pad_type = var_29_pad_type_0, strides = var_29_strides_0, weight = var_9_promoted_to_fp16, x = input_cast_fp16)[name = string("op_29_cast_fp16")];
|
| 16 |
+
tensor<int32, [1]> var_31_axes_0 = const()[name = string("op_31_axes_0"), val = tensor<int32, [1]>([2])];
|
| 17 |
+
tensor<fp16, [1, 16032, 1]> var_31_cast_fp16 = squeeze(axes = var_31_axes_0, x = var_29_cast_fp16)[name = string("op_31_cast_fp16")];
|
| 18 |
+
tensor<int32, [3]> var_34_perm_0 = const()[name = string("op_34_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
|
| 19 |
+
string var_55_pad_type_0 = const()[name = string("op_55_pad_type_0"), val = string("valid")];
|
| 20 |
+
tensor<int32, [2]> var_55_strides_0 = const()[name = string("op_55_strides_0"), val = tensor<int32, [2]>([1, 1])];
|
| 21 |
+
tensor<int32, [4]> var_55_pad_0 = const()[name = string("op_55_pad_0"), val = tensor<int32, [4]>([0, 0, 0, 0])];
|
| 22 |
+
tensor<int32, [2]> var_55_dilations_0 = const()[name = string("op_55_dilations_0"), val = tensor<int32, [2]>([1, 1])];
|
| 23 |
+
int32 var_55_groups_0 = const()[name = string("op_55_groups_0"), val = int32(1)];
|
| 24 |
+
tensor<fp16, [16032, 4096, 1, 1]> var_35_promoted_to_fp16 = const()[name = string("op_35_promoted_to_fp16"), val = tensor<fp16, [16032, 4096, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(131334272)))];
|
| 25 |
+
tensor<fp16, [1, 16032, 1, 1]> var_55_cast_fp16 = conv(dilations = var_55_dilations_0, groups = var_55_groups_0, pad = var_55_pad_0, pad_type = var_55_pad_type_0, strides = var_55_strides_0, weight = var_35_promoted_to_fp16, x = input_cast_fp16)[name = string("op_55_cast_fp16")];
|
| 26 |
+
tensor<int32, [1]> var_57_axes_0 = const()[name = string("op_57_axes_0"), val = tensor<int32, [1]>([2])];
|
| 27 |
+
tensor<fp16, [1, 16032, 1]> var_57_cast_fp16 = squeeze(axes = var_57_axes_0, x = var_55_cast_fp16)[name = string("op_57_cast_fp16")];
|
| 28 |
+
tensor<int32, [3]> var_60_perm_0 = const()[name = string("op_60_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
|
| 29 |
+
string var_81_pad_type_0 = const()[name = string("op_81_pad_type_0"), val = string("valid")];
|
| 30 |
+
tensor<int32, [2]> var_81_strides_0 = const()[name = string("op_81_strides_0"), val = tensor<int32, [2]>([1, 1])];
|
| 31 |
+
tensor<int32, [4]> var_81_pad_0 = const()[name = string("op_81_pad_0"), val = tensor<int32, [4]>([0, 0, 0, 0])];
|
| 32 |
+
tensor<int32, [2]> var_81_dilations_0 = const()[name = string("op_81_dilations_0"), val = tensor<int32, [2]>([1, 1])];
|
| 33 |
+
int32 var_81_groups_0 = const()[name = string("op_81_groups_0"), val = int32(1)];
|
| 34 |
+
tensor<fp16, [16032, 4096, 1, 1]> var_61_promoted_to_fp16 = const()[name = string("op_61_promoted_to_fp16"), val = tensor<fp16, [16032, 4096, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(262668480)))];
|
| 35 |
+
tensor<fp16, [1, 16032, 1, 1]> var_81_cast_fp16 = conv(dilations = var_81_dilations_0, groups = var_81_groups_0, pad = var_81_pad_0, pad_type = var_81_pad_type_0, strides = var_81_strides_0, weight = var_61_promoted_to_fp16, x = input_cast_fp16)[name = string("op_81_cast_fp16")];
|
| 36 |
+
tensor<int32, [1]> var_83_axes_0 = const()[name = string("op_83_axes_0"), val = tensor<int32, [1]>([2])];
|
| 37 |
+
tensor<fp16, [1, 16032, 1]> var_83_cast_fp16 = squeeze(axes = var_83_axes_0, x = var_81_cast_fp16)[name = string("op_83_cast_fp16")];
|
| 38 |
+
tensor<int32, [3]> var_86_perm_0 = const()[name = string("op_86_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
|
| 39 |
+
string var_107_pad_type_0 = const()[name = string("op_107_pad_type_0"), val = string("valid")];
|
| 40 |
+
tensor<int32, [2]> var_107_strides_0 = const()[name = string("op_107_strides_0"), val = tensor<int32, [2]>([1, 1])];
|
| 41 |
+
tensor<int32, [4]> var_107_pad_0 = const()[name = string("op_107_pad_0"), val = tensor<int32, [4]>([0, 0, 0, 0])];
|
| 42 |
+
tensor<int32, [2]> var_107_dilations_0 = const()[name = string("op_107_dilations_0"), val = tensor<int32, [2]>([1, 1])];
|
| 43 |
+
int32 var_107_groups_0 = const()[name = string("op_107_groups_0"), val = int32(1)];
|
| 44 |
+
tensor<fp16, [16032, 4096, 1, 1]> var_87_promoted_to_fp16 = const()[name = string("op_87_promoted_to_fp16"), val = tensor<fp16, [16032, 4096, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(394002688)))];
|
| 45 |
+
tensor<fp16, [1, 16032, 1, 1]> var_107_cast_fp16 = conv(dilations = var_107_dilations_0, groups = var_107_groups_0, pad = var_107_pad_0, pad_type = var_107_pad_type_0, strides = var_107_strides_0, weight = var_87_promoted_to_fp16, x = input_cast_fp16)[name = string("op_107_cast_fp16")];
|
| 46 |
+
tensor<int32, [1]> var_109_axes_0 = const()[name = string("op_109_axes_0"), val = tensor<int32, [1]>([2])];
|
| 47 |
+
tensor<fp16, [1, 16032, 1]> var_109_cast_fp16 = squeeze(axes = var_109_axes_0, x = var_107_cast_fp16)[name = string("op_109_cast_fp16")];
|
| 48 |
+
tensor<int32, [3]> var_112_perm_0 = const()[name = string("op_112_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
|
| 49 |
+
string var_133_pad_type_0 = const()[name = string("op_133_pad_type_0"), val = string("valid")];
|
| 50 |
+
tensor<int32, [2]> var_133_strides_0 = const()[name = string("op_133_strides_0"), val = tensor<int32, [2]>([1, 1])];
|
| 51 |
+
tensor<int32, [4]> var_133_pad_0 = const()[name = string("op_133_pad_0"), val = tensor<int32, [4]>([0, 0, 0, 0])];
|
| 52 |
+
tensor<int32, [2]> var_133_dilations_0 = const()[name = string("op_133_dilations_0"), val = tensor<int32, [2]>([1, 1])];
|
| 53 |
+
int32 var_133_groups_0 = const()[name = string("op_133_groups_0"), val = int32(1)];
|
| 54 |
+
tensor<fp16, [16032, 4096, 1, 1]> var_113_promoted_to_fp16 = const()[name = string("op_113_promoted_to_fp16"), val = tensor<fp16, [16032, 4096, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(525336896)))];
|
| 55 |
+
tensor<fp16, [1, 16032, 1, 1]> var_133_cast_fp16 = conv(dilations = var_133_dilations_0, groups = var_133_groups_0, pad = var_133_pad_0, pad_type = var_133_pad_type_0, strides = var_133_strides_0, weight = var_113_promoted_to_fp16, x = input_cast_fp16)[name = string("op_133_cast_fp16")];
|
| 56 |
+
tensor<int32, [1]> var_135_axes_0 = const()[name = string("op_135_axes_0"), val = tensor<int32, [1]>([2])];
|
| 57 |
+
tensor<fp16, [1, 16032, 1]> var_135_cast_fp16 = squeeze(axes = var_135_axes_0, x = var_133_cast_fp16)[name = string("op_135_cast_fp16")];
|
| 58 |
+
tensor<int32, [3]> var_138_perm_0 = const()[name = string("op_138_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
|
| 59 |
+
string var_159_pad_type_0 = const()[name = string("op_159_pad_type_0"), val = string("valid")];
|
| 60 |
+
tensor<int32, [2]> var_159_strides_0 = const()[name = string("op_159_strides_0"), val = tensor<int32, [2]>([1, 1])];
|
| 61 |
+
tensor<int32, [4]> var_159_pad_0 = const()[name = string("op_159_pad_0"), val = tensor<int32, [4]>([0, 0, 0, 0])];
|
| 62 |
+
tensor<int32, [2]> var_159_dilations_0 = const()[name = string("op_159_dilations_0"), val = tensor<int32, [2]>([1, 1])];
|
| 63 |
+
int32 var_159_groups_0 = const()[name = string("op_159_groups_0"), val = int32(1)];
|
| 64 |
+
tensor<fp16, [16032, 4096, 1, 1]> var_139_promoted_to_fp16 = const()[name = string("op_139_promoted_to_fp16"), val = tensor<fp16, [16032, 4096, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(656671104)))];
|
| 65 |
+
tensor<fp16, [1, 16032, 1, 1]> var_159_cast_fp16 = conv(dilations = var_159_dilations_0, groups = var_159_groups_0, pad = var_159_pad_0, pad_type = var_159_pad_type_0, strides = var_159_strides_0, weight = var_139_promoted_to_fp16, x = input_cast_fp16)[name = string("op_159_cast_fp16")];
|
| 66 |
+
tensor<int32, [1]> var_161_axes_0 = const()[name = string("op_161_axes_0"), val = tensor<int32, [1]>([2])];
|
| 67 |
+
tensor<fp16, [1, 16032, 1]> var_161_cast_fp16 = squeeze(axes = var_161_axes_0, x = var_159_cast_fp16)[name = string("op_161_cast_fp16")];
|
| 68 |
+
tensor<int32, [3]> var_164_perm_0 = const()[name = string("op_164_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
|
| 69 |
+
string var_185_pad_type_0 = const()[name = string("op_185_pad_type_0"), val = string("valid")];
|
| 70 |
+
tensor<int32, [2]> var_185_strides_0 = const()[name = string("op_185_strides_0"), val = tensor<int32, [2]>([1, 1])];
|
| 71 |
+
tensor<int32, [4]> var_185_pad_0 = const()[name = string("op_185_pad_0"), val = tensor<int32, [4]>([0, 0, 0, 0])];
|
| 72 |
+
tensor<int32, [2]> var_185_dilations_0 = const()[name = string("op_185_dilations_0"), val = tensor<int32, [2]>([1, 1])];
|
| 73 |
+
int32 var_185_groups_0 = const()[name = string("op_185_groups_0"), val = int32(1)];
|
| 74 |
+
tensor<fp16, [16032, 4096, 1, 1]> var_165_promoted_to_fp16 = const()[name = string("op_165_promoted_to_fp16"), val = tensor<fp16, [16032, 4096, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(788005312)))];
|
| 75 |
+
tensor<fp16, [1, 16032, 1, 1]> var_185_cast_fp16 = conv(dilations = var_185_dilations_0, groups = var_185_groups_0, pad = var_185_pad_0, pad_type = var_185_pad_type_0, strides = var_185_strides_0, weight = var_165_promoted_to_fp16, x = input_cast_fp16)[name = string("op_185_cast_fp16")];
|
| 76 |
+
tensor<int32, [1]> var_187_axes_0 = const()[name = string("op_187_axes_0"), val = tensor<int32, [1]>([2])];
|
| 77 |
+
tensor<fp16, [1, 16032, 1]> var_187_cast_fp16 = squeeze(axes = var_187_axes_0, x = var_185_cast_fp16)[name = string("op_187_cast_fp16")];
|
| 78 |
+
tensor<int32, [3]> var_190_perm_0 = const()[name = string("op_190_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
|
| 79 |
+
string var_211_pad_type_0 = const()[name = string("op_211_pad_type_0"), val = string("valid")];
|
| 80 |
+
tensor<int32, [2]> var_211_strides_0 = const()[name = string("op_211_strides_0"), val = tensor<int32, [2]>([1, 1])];
|
| 81 |
+
tensor<int32, [4]> var_211_pad_0 = const()[name = string("op_211_pad_0"), val = tensor<int32, [4]>([0, 0, 0, 0])];
|
| 82 |
+
tensor<int32, [2]> var_211_dilations_0 = const()[name = string("op_211_dilations_0"), val = tensor<int32, [2]>([1, 1])];
|
| 83 |
+
int32 var_211_groups_0 = const()[name = string("op_211_groups_0"), val = int32(1)];
|
| 84 |
+
tensor<fp16, [16032, 4096, 1, 1]> var_191_promoted_to_fp16 = const()[name = string("op_191_promoted_to_fp16"), val = tensor<fp16, [16032, 4096, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(919339520)))];
|
| 85 |
+
tensor<fp16, [1, 16032, 1, 1]> var_211_cast_fp16 = conv(dilations = var_211_dilations_0, groups = var_211_groups_0, pad = var_211_pad_0, pad_type = var_211_pad_type_0, strides = var_211_strides_0, weight = var_191_promoted_to_fp16, x = input_cast_fp16)[name = string("op_211_cast_fp16")];
|
| 86 |
+
tensor<int32, [1]> var_213_axes_0 = const()[name = string("op_213_axes_0"), val = tensor<int32, [1]>([2])];
|
| 87 |
+
tensor<fp16, [1, 16032, 1]> var_213_cast_fp16 = squeeze(axes = var_213_axes_0, x = var_211_cast_fp16)[name = string("op_213_cast_fp16")];
|
| 88 |
+
tensor<int32, [3]> var_216_perm_0 = const()[name = string("op_216_perm_0"), val = tensor<int32, [3]>([0, 2, 1])];
|
| 89 |
+
tensor<fp16, [1, 1, 16032]> logits8 = transpose(perm = var_216_perm_0, x = var_213_cast_fp16)[name = string("transpose_0")];
|
| 90 |
+
tensor<fp16, [1, 1, 16032]> logits7 = transpose(perm = var_190_perm_0, x = var_187_cast_fp16)[name = string("transpose_1")];
|
| 91 |
+
tensor<fp16, [1, 1, 16032]> logits6 = transpose(perm = var_164_perm_0, x = var_161_cast_fp16)[name = string("transpose_2")];
|
| 92 |
+
tensor<fp16, [1, 1, 16032]> logits5 = transpose(perm = var_138_perm_0, x = var_135_cast_fp16)[name = string("transpose_3")];
|
| 93 |
+
tensor<fp16, [1, 1, 16032]> logits4 = transpose(perm = var_112_perm_0, x = var_109_cast_fp16)[name = string("transpose_4")];
|
| 94 |
+
tensor<fp16, [1, 1, 16032]> logits3 = transpose(perm = var_86_perm_0, x = var_83_cast_fp16)[name = string("transpose_5")];
|
| 95 |
+
tensor<fp16, [1, 1, 16032]> logits2 = transpose(perm = var_60_perm_0, x = var_57_cast_fp16)[name = string("transpose_6")];
|
| 96 |
+
tensor<fp16, [1, 1, 16032]> logits1 = transpose(perm = var_34_perm_0, x = var_31_cast_fp16)[name = string("transpose_7")];
|
| 97 |
+
} -> (logits1, logits2, logits3, logits4, logits5, logits6, logits7, logits8);
|
| 98 |
+
}
|