finnvoorhees
commited on
Upload folder using huggingface_hub
Browse files
Llama-3.2-3B-Instruct-4bit.mlmodelc/analytics/coremldata.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:35eb23d299aa3193ed8f221e095d0e2d71e0c8134fdfde43f52b97bd6fff6558
|
3 |
+
size 243
|
Llama-3.2-3B-Instruct-4bit.mlmodelc/coremldata.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4a3c257acfeb8acfe5b7c1253b2952c5f3e4f1dffe8b54191445879f1e887d74
|
3 |
+
size 447
|
Llama-3.2-3B-Instruct-4bit.mlmodelc/metadata.json
ADDED
@@ -0,0 +1,117 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"metadataOutputVersion" : "3.0",
|
4 |
+
"storagePrecision" : "Mixed (Float16, Int4)",
|
5 |
+
"outputSchema" : [
|
6 |
+
{
|
7 |
+
"hasShapeFlexibility" : "0",
|
8 |
+
"isOptional" : "0",
|
9 |
+
"dataType" : "Float16",
|
10 |
+
"formattedType" : "MultiArray (Float16)",
|
11 |
+
"shortDescription" : "",
|
12 |
+
"shape" : "[]",
|
13 |
+
"name" : "logits",
|
14 |
+
"type" : "MultiArray"
|
15 |
+
}
|
16 |
+
],
|
17 |
+
"modelParameters" : [
|
18 |
+
|
19 |
+
],
|
20 |
+
"specificationVersion" : 9,
|
21 |
+
"mlProgramOperationTypeHistogram" : {
|
22 |
+
"Ios18.expandDims" : 62,
|
23 |
+
"Range1d" : 1,
|
24 |
+
"Ios18.cos" : 1,
|
25 |
+
"Ios18.mul" : 310,
|
26 |
+
"Ios18.matmul" : 1,
|
27 |
+
"Ios18.rsqrt" : 57,
|
28 |
+
"Ios16.reduceMean" : 57,
|
29 |
+
"Ios18.realDiv" : 56,
|
30 |
+
"Ios18.scaledDotProductAttention" : 28,
|
31 |
+
"Ios18.sin" : 1,
|
32 |
+
"Ios18.readState" : 58,
|
33 |
+
"Tile" : 56,
|
34 |
+
"Ios18.gather" : 117,
|
35 |
+
"Ios18.add" : 171,
|
36 |
+
"Ios18.writeState" : 56,
|
37 |
+
"Ios18.sliceUpdate" : 56,
|
38 |
+
"Ios18.reshape" : 168,
|
39 |
+
"Shape" : 172,
|
40 |
+
"Ios18.linear" : 197,
|
41 |
+
"Ios18.concat" : 199,
|
42 |
+
"Ios18.transpose" : 113,
|
43 |
+
"Ios18.sub" : 1,
|
44 |
+
"Ios18.pow" : 57,
|
45 |
+
"Ios18.cast" : 233,
|
46 |
+
"Ios18.silu" : 28,
|
47 |
+
"Ios18.sliceByIndex" : 252,
|
48 |
+
"Ios18.constexprBlockwiseShiftScale" : 197
|
49 |
+
},
|
50 |
+
"computePrecision" : "Mixed (Float16, Float32, Int16, Int32, UInt16)",
|
51 |
+
"isUpdatable" : "0",
|
52 |
+
"stateSchema" : [
|
53 |
+
{
|
54 |
+
"dataType" : "Float16",
|
55 |
+
"isOptional" : "0",
|
56 |
+
"formattedType" : "State (Float16 28 × 1 × 8 × 2048 × 128)",
|
57 |
+
"shortDescription" : "",
|
58 |
+
"shape" : "[28, 1, 8, 2048, 128]",
|
59 |
+
"name" : "key_cache",
|
60 |
+
"type" : "State"
|
61 |
+
},
|
62 |
+
{
|
63 |
+
"dataType" : "Float16",
|
64 |
+
"isOptional" : "0",
|
65 |
+
"formattedType" : "State (Float16 28 × 1 × 8 × 2048 × 128)",
|
66 |
+
"shortDescription" : "",
|
67 |
+
"shape" : "[28, 1, 8, 2048, 128]",
|
68 |
+
"name" : "value_cache",
|
69 |
+
"type" : "State"
|
70 |
+
}
|
71 |
+
],
|
72 |
+
"availability" : {
|
73 |
+
"macOS" : "15.0",
|
74 |
+
"tvOS" : "18.0",
|
75 |
+
"visionOS" : "2.0",
|
76 |
+
"watchOS" : "11.0",
|
77 |
+
"iOS" : "18.0",
|
78 |
+
"macCatalyst" : "18.0"
|
79 |
+
},
|
80 |
+
"modelType" : {
|
81 |
+
"name" : "MLModelType_mlProgram"
|
82 |
+
},
|
83 |
+
"userDefinedMetadata" : {
|
84 |
+
"com.github.apple.coremltools.source_dialect" : "TorchScript",
|
85 |
+
"com.github.apple.coremltools.source" : "torch==2.5.1",
|
86 |
+
"com.github.apple.coremltools.version" : "8.1"
|
87 |
+
},
|
88 |
+
"inputSchema" : [
|
89 |
+
{
|
90 |
+
"dataType" : "Int32",
|
91 |
+
"hasShapeFlexibility" : "1",
|
92 |
+
"isOptional" : "0",
|
93 |
+
"shapeFlexibility" : "1 × 1...2048",
|
94 |
+
"shapeRange" : "[[1, 1], [1, 2048]]",
|
95 |
+
"formattedType" : "MultiArray (Int32 1 × 1)",
|
96 |
+
"type" : "MultiArray",
|
97 |
+
"shape" : "[1, 1]",
|
98 |
+
"name" : "input_ids",
|
99 |
+
"shortDescription" : ""
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"dataType" : "Float16",
|
103 |
+
"hasShapeFlexibility" : "1",
|
104 |
+
"isOptional" : "0",
|
105 |
+
"shapeFlexibility" : "1 × 1 × 1...2048 × 1...2048",
|
106 |
+
"shapeRange" : "[[1, 1], [1, 1], [1, 2048], [1, 2048]]",
|
107 |
+
"formattedType" : "MultiArray (Float16 1 × 1 × 1 × 1)",
|
108 |
+
"type" : "MultiArray",
|
109 |
+
"shape" : "[1, 1, 1, 1]",
|
110 |
+
"name" : "causal_mask",
|
111 |
+
"shortDescription" : ""
|
112 |
+
}
|
113 |
+
],
|
114 |
+
"generatedClassName" : "Llama_3_2_3B_Instruct_4bit",
|
115 |
+
"method" : "predict"
|
116 |
+
}
|
117 |
+
]
|
Llama-3.2-3B-Instruct-4bit.mlmodelc/model.mil
ADDED
The diff for this file is too large to render.
See raw diff
|
|
Llama-3.2-3B-Instruct-4bit.mlmodelc/weights/weight.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f880602f4726a897555457adf5c8f7195ca11df85769c122dd3fa79fe4960298
|
3 |
+
size 1807733952
|