Hugging Face's logo

jondin
/
ExploraVistAppModels

Core ML
Safetensors
Model card Files
xet
Community
Jonathan Ipe commited on
Commit
bdd4eae
·
1 Parent(s): e1f5585

model files

Browse files
Files changed (48) hide show
  1. CoreMLModels/.DS_Store +0 -0
  2. CoreMLModels/ImageEncoder_float32_batch_16.mlmodelc/analytics/coremldata.bin +3 -0
  3. CoreMLModels/ImageEncoder_float32_batch_16.mlmodelc/coremldata.bin +3 -0
  4. CoreMLModels/ImageEncoder_float32_batch_16.mlmodelc/metadata.json +81 -0
  5. CoreMLModels/ImageEncoder_float32_batch_16.mlmodelc/model.mil +0 -0
  6. CoreMLModels/ImageEncoder_float32_batch_16.mlmodelc/weights/weight.bin +3 -0
  7. CoreMLModels/ImageEncoder_mobileCLIP_s2.mlmodelc/analytics/coremldata.bin +3 -0
  8. CoreMLModels/ImageEncoder_mobileCLIP_s2.mlmodelc/coremldata.bin +3 -0
  9. CoreMLModels/ImageEncoder_mobileCLIP_s2.mlmodelc/metadata.json +75 -0
  10. CoreMLModels/ImageEncoder_mobileCLIP_s2.mlmodelc/model.mil +0 -0
  11. CoreMLModels/ImageEncoder_mobileCLIP_s2.mlmodelc/weights/weight.bin +3 -0
  12. CoreMLModels/TextEncoder_mobileCLIP_s2.mlmodelc/analytics/coremldata.bin +3 -0
  13. CoreMLModels/TextEncoder_mobileCLIP_s2.mlmodelc/coremldata.bin +3 -0
  14. CoreMLModels/TextEncoder_mobileCLIP_s2.mlmodelc/metadata.json +70 -0
  15. CoreMLModels/TextEncoder_mobileCLIP_s2.mlmodelc/model.mil +0 -0
  16. CoreMLModels/TextEncoder_mobileCLIP_s2.mlmodelc/weights/weight.bin +3 -0
  17. CoreMLModels/clip_mci_image_s2.mlmodelc/analytics/coremldata.bin +3 -0
  18. CoreMLModels/clip_mci_image_s2.mlmodelc/coremldata.bin +3 -0
  19. CoreMLModels/clip_mci_image_s2.mlmodelc/metadata.json +82 -0
  20. CoreMLModels/clip_mci_image_s2.mlmodelc/model.mil +0 -0
  21. CoreMLModels/clip_mci_image_s2.mlmodelc/weights/weight.bin +3 -0
  22. CoreMLModels/clip_mci_image_s2_mult_batch.mlmodelc/analytics/coremldata.bin +3 -0
  23. CoreMLModels/clip_mci_image_s2_mult_batch.mlmodelc/coremldata.bin +3 -0
  24. CoreMLModels/clip_mci_image_s2_mult_batch.mlmodelc/metadata.json +85 -0
  25. CoreMLModels/clip_mci_image_s2_mult_batch.mlmodelc/model.mil +0 -0
  26. CoreMLModels/clip_mci_image_s2_mult_batch.mlmodelc/weights/weight.bin +3 -0
  27. CoreMLModels/clip_text_s2.mlmodelc/analytics/coremldata.bin +3 -0
  28. CoreMLModels/clip_text_s2.mlmodelc/coremldata.bin +3 -0
  29. CoreMLModels/clip_text_s2.mlmodelc/metadata.json +80 -0
  30. CoreMLModels/clip_text_s2.mlmodelc/model.mil +0 -0
  31. CoreMLModels/clip_text_s2.mlmodelc/weights/weight.bin +3 -0
  32. CoreMLModels/merges_clip.txt +0 -0
  33. CoreMLModels/vocab_clip.json +0 -0
  34. model/README.md +222 -0
  35. model/added_tokens.json +6 -0
  36. model/config.json +50 -0
  37. model/fastvithd.mlpackage/Data/com.apple.CoreML/model.mlmodel +3 -0
  38. model/fastvithd.mlpackage/Data/com.apple.CoreML/weights/weight.bin +3 -0
  39. model/fastvithd.mlpackage/Manifest.json +3 -0
  40. model/merges.txt +0 -0
  41. model/model.safetensors +3 -0
  42. model/model.safetensors.index.json +646 -0
  43. model/preprocessor_config.json +28 -0
  44. model/processor_config.json +7 -0
  45. model/special_tokens_map.json +20 -0
  46. model/tokenizer.json +3 -0
  47. model/tokenizer_config.json +52 -0
  48. model/vocab.json +0 -0
CoreMLModels/.DS_Store ADDED
Binary file (8.2 kB). View file
 
CoreMLModels/ImageEncoder_float32_batch_16.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9de20562aed22140187fec66d8963877160327490059811b41b047c288ac328a
3
+ size 243
CoreMLModels/ImageEncoder_float32_batch_16.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8e0b6fcb77f4b1245fa1df61b527e03e066586fe0dfb162399f8d52c6d57209
3
+ size 338
CoreMLModels/ImageEncoder_float32_batch_16.mlmodelc/metadata.json ADDED
@@ -0,0 +1,81 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "metadataOutputVersion" : "3.0",
4
+ "storagePrecision" : "Float16",
5
+ "outputSchema" : [
6
+ {
7
+ "hasShapeFlexibility" : "0",
8
+ "isOptional" : "0",
9
+ "dataType" : "Float32",
10
+ "formattedType" : "MultiArray (Float32)",
11
+ "shortDescription" : "",
12
+ "shape" : "[]",
13
+ "name" : "embOutput",
14
+ "type" : "MultiArray"
15
+ }
16
+ ],
17
+ "modelParameters" : [
18
+
19
+ ],
20
+ "specificationVersion" : 7,
21
+ "mlProgramOperationTypeHistogram" : {
22
+ "Concat" : 3,
23
+ "Ios16.cast" : 2,
24
+ "Ios16.mul" : 37,
25
+ "Ios16.layerNorm" : 26,
26
+ "SliceByIndex" : 37,
27
+ "Transpose" : 63,
28
+ "Shape" : 2,
29
+ "Ios16.sigmoid" : 12,
30
+ "Ios16.linear" : 49,
31
+ "Ios16.add" : 27,
32
+ "Fill" : 1,
33
+ "Ios16.conv" : 1,
34
+ "Squeeze" : 12,
35
+ "Ios16.matmul" : 24,
36
+ "ExpandDims" : 12,
37
+ "Ios16.gather" : 2,
38
+ "Ios16.softmax" : 12,
39
+ "Ios16.reshape" : 109
40
+ },
41
+ "computePrecision" : "Mixed (Float16, Float32, Int32)",
42
+ "isUpdatable" : "0",
43
+ "stateSchema" : [
44
+
45
+ ],
46
+ "availability" : {
47
+ "macOS" : "13.0",
48
+ "tvOS" : "16.0",
49
+ "visionOS" : "1.0",
50
+ "watchOS" : "9.0",
51
+ "iOS" : "16.0",
52
+ "macCatalyst" : "16.0"
53
+ },
54
+ "modelType" : {
55
+ "name" : "MLModelType_mlProgram"
56
+ },
57
+ "userDefinedMetadata" : {
58
+ "com.github.apple.coremltools.source_dialect" : "TorchScript",
59
+ "com.github.apple.coremltools.source" : "torch==2.6.0+cu124",
60
+ "com.github.apple.coremltools.version" : "8.3.0"
61
+ },
62
+ "inputSchema" : [
63
+ {
64
+ "formattedType" : "Image (Color 224 × 224)",
65
+ "hasSizeFlexibility" : "1",
66
+ "shortDescription" : "",
67
+ "sizeRange" : "[[224, 224], [224, 224]]",
68
+ "width" : "224",
69
+ "type" : "Image",
70
+ "isColor" : "1",
71
+ "height" : "224",
72
+ "sizeFlexibility" : "224 × 224",
73
+ "colorspace" : "RGB",
74
+ "name" : "colorImage",
75
+ "isOptional" : "0"
76
+ }
77
+ ],
78
+ "generatedClassName" : "ImageEncoder_float32_batch_16",
79
+ "method" : "predict"
80
+ }
81
+ ]
CoreMLModels/ImageEncoder_float32_batch_16.mlmodelc/model.mil ADDED
The diff for this file is too large to render. See raw diff
 
CoreMLModels/ImageEncoder_float32_batch_16.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41abc3b89a92dfa06f8b2950f8d67494bb3ef6c47ee1b14d0c39fc3aee2cfb9d
3
+ size 175709312
CoreMLModels/ImageEncoder_mobileCLIP_s2.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f0a9affc51514969939fde6631d71cdd98e8b8029e50f00dc84d219192f1b629
3
+ size 243
CoreMLModels/ImageEncoder_mobileCLIP_s2.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:df007948a786f05e9a969fafadab22352921898530176662785ad4869485b40b
3
+ size 317
CoreMLModels/ImageEncoder_mobileCLIP_s2.mlmodelc/metadata.json ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "metadataOutputVersion" : "3.0",
4
+ "storagePrecision" : "Float16",
5
+ "outputSchema" : [
6
+ {
7
+ "hasShapeFlexibility" : "0",
8
+ "isOptional" : "0",
9
+ "dataType" : "Float32",
10
+ "formattedType" : "MultiArray (Float32 1 × 512)",
11
+ "shortDescription" : "",
12
+ "shape" : "[1, 512]",
13
+ "name" : "embOutput",
14
+ "type" : "MultiArray"
15
+ }
16
+ ],
17
+ "modelParameters" : [
18
+
19
+ ],
20
+ "specificationVersion" : 7,
21
+ "mlProgramOperationTypeHistogram" : {
22
+ "Split" : 4,
23
+ "Squeeze" : 12,
24
+ "Ios16.cast" : 2,
25
+ "Ios16.mul" : 12,
26
+ "Transpose" : 20,
27
+ "Ios16.sigmoid" : 3,
28
+ "Ios16.conv" : 189,
29
+ "Ios16.add" : 48,
30
+ "Ios16.linear" : 9,
31
+ "Ios16.avgPool" : 1,
32
+ "Ios16.gelu" : 54,
33
+ "Ios16.matmul" : 8,
34
+ "Ios16.reduceMean" : 3,
35
+ "Ios16.batchNorm" : 4,
36
+ "Ios16.relu" : 3,
37
+ "Ios16.reshape" : 16,
38
+ "Ios16.softmax" : 4
39
+ },
40
+ "computePrecision" : "Mixed (Float16, Float32, Int32)",
41
+ "isUpdatable" : "0",
42
+ "availability" : {
43
+ "macOS" : "13.0",
44
+ "tvOS" : "16.0",
45
+ "visionOS" : "1.0",
46
+ "watchOS" : "9.0",
47
+ "iOS" : "16.0",
48
+ "macCatalyst" : "16.0"
49
+ },
50
+ "modelType" : {
51
+ "name" : "MLModelType_mlProgram"
52
+ },
53
+ "userDefinedMetadata" : {
54
+ "com.github.apple.coremltools.source_dialect" : "TorchScript",
55
+ "com.github.apple.coremltools.source" : "torch==1.13.1",
56
+ "com.github.apple.coremltools.version" : "7.1"
57
+ },
58
+ "inputSchema" : [
59
+ {
60
+ "height" : "256",
61
+ "colorspace" : "RGB",
62
+ "isOptional" : "0",
63
+ "width" : "256",
64
+ "isColor" : "1",
65
+ "formattedType" : "Image (Color 256 × 256)",
66
+ "hasSizeFlexibility" : "0",
67
+ "type" : "Image",
68
+ "shortDescription" : "",
69
+ "name" : "colorImage"
70
+ }
71
+ ],
72
+ "generatedClassName" : "ImageEncoder_mobileCLIP_s2",
73
+ "method" : "predict"
74
+ }
75
+ ]
CoreMLModels/ImageEncoder_mobileCLIP_s2.mlmodelc/model.mil ADDED
The diff for this file is too large to render. See raw diff
 
CoreMLModels/ImageEncoder_mobileCLIP_s2.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6409033333b398af4189057a9e3178590157f188f3ef8d429b4c7778aa3fb37d
3
+ size 71397632
CoreMLModels/TextEncoder_mobileCLIP_s2.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc311528521250462a48096fc99b4e95b4332c26304dd444fef92dcbf5a24ffc
3
+ size 243
CoreMLModels/TextEncoder_mobileCLIP_s2.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:252f75dc59d16f253fc611287794e37a7f9e7052420bbb0ff188866dab74729d
3
+ size 263
CoreMLModels/TextEncoder_mobileCLIP_s2.mlmodelc/metadata.json ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "metadataOutputVersion" : "3.0",
4
+ "storagePrecision" : "Float16",
5
+ "outputSchema" : [
6
+ {
7
+ "hasShapeFlexibility" : "0",
8
+ "isOptional" : "0",
9
+ "dataType" : "Float32",
10
+ "formattedType" : "MultiArray (Float32 1 × 512)",
11
+ "shortDescription" : "",
12
+ "shape" : "[1, 512]",
13
+ "name" : "text_embeddings",
14
+ "type" : "MultiArray"
15
+ }
16
+ ],
17
+ "modelParameters" : [
18
+
19
+ ],
20
+ "specificationVersion" : 7,
21
+ "mlProgramOperationTypeHistogram" : {
22
+ "Ios16.cast" : 1,
23
+ "Ios16.mul" : 12,
24
+ "Ios16.layerNorm" : 25,
25
+ "SliceByIndex" : 36,
26
+ "Stack" : 1,
27
+ "Transpose" : 36,
28
+ "Ios16.linear" : 49,
29
+ "Ios16.add" : 25,
30
+ "Ios16.matmul" : 24,
31
+ "Ios16.gelu" : 12,
32
+ "Ios16.softmax" : 12,
33
+ "Ios16.gatherNd" : 1,
34
+ "Ios16.gather" : 1,
35
+ "Ios16.reshape" : 24,
36
+ "Ios16.reduceArgmax" : 1
37
+ },
38
+ "computePrecision" : "Mixed (Float16, Float32, Int32)",
39
+ "isUpdatable" : "0",
40
+ "availability" : {
41
+ "macOS" : "13.0",
42
+ "tvOS" : "16.0",
43
+ "visionOS" : "1.0",
44
+ "watchOS" : "9.0",
45
+ "iOS" : "16.0",
46
+ "macCatalyst" : "16.0"
47
+ },
48
+ "modelType" : {
49
+ "name" : "MLModelType_mlProgram"
50
+ },
51
+ "userDefinedMetadata" : {
52
+ "com.github.apple.coremltools.version" : "7.0",
53
+ "com.github.apple.coremltools.source" : "torch==1.13.1"
54
+ },
55
+ "inputSchema" : [
56
+ {
57
+ "hasShapeFlexibility" : "0",
58
+ "isOptional" : "0",
59
+ "dataType" : "Int32",
60
+ "formattedType" : "MultiArray (Int32 1 × 77)",
61
+ "shortDescription" : "",
62
+ "shape" : "[1, 77]",
63
+ "name" : "input_tokens",
64
+ "type" : "MultiArray"
65
+ }
66
+ ],
67
+ "generatedClassName" : "TextEncoder_mobileCLIP_s2",
68
+ "method" : "predict"
69
+ }
70
+ ]
CoreMLModels/TextEncoder_mobileCLIP_s2.mlmodelc/model.mil ADDED
The diff for this file is too large to render. See raw diff
 
CoreMLModels/TextEncoder_mobileCLIP_s2.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d7924ab28ec16126b5e8d03236ddf921d801bac7af9aceab7fca2b299cb7bde
3
+ size 126866880
CoreMLModels/clip_mci_image_s2.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b04c1f64b14765adb9e691d4a7a140cb6309b09ae99b96bfc02c67a3a1142bf
3
+ size 243
CoreMLModels/clip_mci_image_s2.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:089f3ccbe7063bf8d835cf6b1538a4fa491ba92939ac01ab3f9b70e7e9a4a262
3
+ size 333
CoreMLModels/clip_mci_image_s2.mlmodelc/metadata.json ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "metadataOutputVersion" : "3.0",
4
+ "storagePrecision" : "Float16",
5
+ "outputSchema" : [
6
+ {
7
+ "hasShapeFlexibility" : "0",
8
+ "isOptional" : "0",
9
+ "dataType" : "Float16",
10
+ "formattedType" : "MultiArray (Float16 1 × 512)",
11
+ "shortDescription" : "",
12
+ "shape" : "[1, 512]",
13
+ "name" : "output_embeddings",
14
+ "type" : "MultiArray"
15
+ }
16
+ ],
17
+ "modelParameters" : [
18
+
19
+ ],
20
+ "specificationVersion" : 8,
21
+ "mlProgramOperationTypeHistogram" : {
22
+ "Ios16.reduceL2Norm" : 1,
23
+ "Ios17.reshape" : 16,
24
+ "Ios16.reduceMean" : 3,
25
+ "Ios16.softmax" : 4,
26
+ "Ios17.matmul" : 8,
27
+ "Ios17.maximum" : 1,
28
+ "Ios17.transpose" : 16,
29
+ "Split" : 4,
30
+ "Ios16.avgPool" : 1,
31
+ "Ios17.add" : 48,
32
+ "Tile" : 1,
33
+ "Ios16.sigmoid" : 3,
34
+ "Ios17.squeeze" : 12,
35
+ "Ios17.batchNorm" : 4,
36
+ "Ios17.conv" : 189,
37
+ "Ios16.relu" : 3,
38
+ "Ios16.gelu" : 54,
39
+ "Ios17.cast" : 1,
40
+ "Ios17.realDiv" : 1,
41
+ "Ios17.linear" : 9,
42
+ "Ios17.mul" : 12
43
+ },
44
+ "computePrecision" : "Mixed (Float16, Float32, Int32)",
45
+ "isUpdatable" : "0",
46
+ "stateSchema" : [
47
+
48
+ ],
49
+ "availability" : {
50
+ "macOS" : "14.0",
51
+ "tvOS" : "17.0",
52
+ "visionOS" : "1.0",
53
+ "watchOS" : "10.0",
54
+ "iOS" : "17.0",
55
+ "macCatalyst" : "17.0"
56
+ },
57
+ "modelType" : {
58
+ "name" : "MLModelType_mlProgram"
59
+ },
60
+ "userDefinedMetadata" : {
61
+ "com.github.apple.coremltools.source_dialect" : "TorchScript",
62
+ "com.github.apple.coremltools.source" : "torch==2.6.0+cu124",
63
+ "com.github.apple.coremltools.version" : "8.3.0"
64
+ },
65
+ "inputSchema" : [
66
+ {
67
+ "height" : "256",
68
+ "colorspace" : "RGB",
69
+ "isOptional" : "0",
70
+ "width" : "256",
71
+ "isColor" : "1",
72
+ "formattedType" : "Image (Color 256 × 256)",
73
+ "hasSizeFlexibility" : "0",
74
+ "type" : "Image",
75
+ "shortDescription" : "",
76
+ "name" : "input_image"
77
+ }
78
+ ],
79
+ "generatedClassName" : "clip_mci_image_s2",
80
+ "method" : "predict"
81
+ }
82
+ ]
CoreMLModels/clip_mci_image_s2.mlmodelc/model.mil ADDED
The diff for this file is too large to render. See raw diff
 
CoreMLModels/clip_mci_image_s2.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:97483acff55708b7bd0a6e6a35f91a942ef95c86d1bff3baf591cd2fa7535df0
3
+ size 71397632
CoreMLModels/clip_mci_image_s2_mult_batch.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:86adbc296697f11ccf9ee6c7367b4ac64cbf628a0a94bc540fd74e79ff873f0b
3
+ size 243
CoreMLModels/clip_mci_image_s2_mult_batch.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c378f20e96d2176e2cd8593c9fd10ed70c7cd36269b11a87667b1623995e6d7
3
+ size 341
CoreMLModels/clip_mci_image_s2_mult_batch.mlmodelc/metadata.json ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "metadataOutputVersion" : "3.0",
4
+ "storagePrecision" : "Float16",
5
+ "outputSchema" : [
6
+ {
7
+ "hasShapeFlexibility" : "0",
8
+ "isOptional" : "0",
9
+ "dataType" : "Float16",
10
+ "formattedType" : "MultiArray (Float16)",
11
+ "shortDescription" : "",
12
+ "shape" : "[]",
13
+ "name" : "output_embeddings",
14
+ "type" : "MultiArray"
15
+ }
16
+ ],
17
+ "modelParameters" : [
18
+
19
+ ],
20
+ "specificationVersion" : 8,
21
+ "mlProgramOperationTypeHistogram" : {
22
+ "Ios16.reduceL2Norm" : 1,
23
+ "Ios17.reshape" : 16,
24
+ "Ios16.reduceMean" : 3,
25
+ "Ios16.softmax" : 4,
26
+ "Ios17.matmul" : 8,
27
+ "Ios17.maximum" : 1,
28
+ "Ios17.transpose" : 16,
29
+ "Split" : 4,
30
+ "Ios16.avgPool" : 1,
31
+ "Ios17.add" : 48,
32
+ "Ios16.sigmoid" : 3,
33
+ "Tile" : 1,
34
+ "Ios17.squeeze" : 12,
35
+ "Shape" : 2,
36
+ "Ios17.batchNorm" : 4,
37
+ "Ios17.conv" : 189,
38
+ "Ios16.relu" : 3,
39
+ "Ios16.gelu" : 54,
40
+ "Ios17.cast" : 1,
41
+ "Ios17.realDiv" : 2,
42
+ "Ios17.linear" : 9,
43
+ "Ios17.mul" : 12
44
+ },
45
+ "computePrecision" : "Mixed (Float16, Float32, Int32)",
46
+ "isUpdatable" : "0",
47
+ "stateSchema" : [
48
+
49
+ ],
50
+ "availability" : {
51
+ "macOS" : "14.0",
52
+ "tvOS" : "17.0",
53
+ "visionOS" : "1.0",
54
+ "watchOS" : "10.0",
55
+ "iOS" : "17.0",
56
+ "macCatalyst" : "17.0"
57
+ },
58
+ "modelType" : {
59
+ "name" : "MLModelType_mlProgram"
60
+ },
61
+ "userDefinedMetadata" : {
62
+ "com.github.apple.coremltools.source_dialect" : "TorchScript",
63
+ "com.github.apple.coremltools.version" : "8.3.0",
64
+ "com.github.apple.coremltools.source" : "torch==2.7.1"
65
+ },
66
+ "inputSchema" : [
67
+ {
68
+ "formattedType" : "Image (Color 256 × 256)",
69
+ "hasSizeFlexibility" : "1",
70
+ "shortDescription" : "",
71
+ "sizeRange" : "[[256, 256], [256, 256]]",
72
+ "width" : "256",
73
+ "type" : "Image",
74
+ "isColor" : "1",
75
+ "height" : "256",
76
+ "sizeFlexibility" : "256 × 256",
77
+ "colorspace" : "RGB",
78
+ "name" : "input_image",
79
+ "isOptional" : "0"
80
+ }
81
+ ],
82
+ "generatedClassName" : "clip_mci_image_s2",
83
+ "method" : "predict"
84
+ }
85
+ ]
CoreMLModels/clip_mci_image_s2_mult_batch.mlmodelc/model.mil ADDED
The diff for this file is too large to render. See raw diff
 
CoreMLModels/clip_mci_image_s2_mult_batch.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:97483acff55708b7bd0a6e6a35f91a942ef95c86d1bff3baf591cd2fa7535df0
3
+ size 71397632
CoreMLModels/clip_text_s2.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e73ee341d708af9891780c19072f8648b3322fda7f53ef75f5b895e54fb33fa8
3
+ size 243
CoreMLModels/clip_text_s2.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:86019767fd5cd2fcfe9c113aef7ab610faef3a59c0b096c0b96f1cd18e6b8ecf
3
+ size 332
CoreMLModels/clip_text_s2.mlmodelc/metadata.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "metadataOutputVersion" : "3.0",
4
+ "storagePrecision" : "Float16",
5
+ "outputSchema" : [
6
+ {
7
+ "hasShapeFlexibility" : "0",
8
+ "isOptional" : "0",
9
+ "dataType" : "Float16",
10
+ "formattedType" : "MultiArray (Float16 1 × 512)",
11
+ "shortDescription" : "",
12
+ "shape" : "[1, 512]",
13
+ "name" : "output_embeddings",
14
+ "type" : "MultiArray"
15
+ }
16
+ ],
17
+ "modelParameters" : [
18
+
19
+ ],
20
+ "specificationVersion" : 8,
21
+ "mlProgramOperationTypeHistogram" : {
22
+ "Ios17.reduceArgmax" : 1,
23
+ "Ios16.reduceL2Norm" : 1,
24
+ "Ios17.reshape" : 24,
25
+ "Ios16.softmax" : 12,
26
+ "Ios17.matmul" : 24,
27
+ "Ios17.transpose" : 24,
28
+ "Ios17.maximum" : 1,
29
+ "Select" : 1,
30
+ "Ios17.add" : 26,
31
+ "Tile" : 1,
32
+ "Ios17.sliceByIndex" : 36,
33
+ "Ios17.gatherNd" : 1,
34
+ "Ios17.gather" : 1,
35
+ "Ios17.layerNorm" : 25,
36
+ "Ios16.gelu" : 12,
37
+ "Ios17.cast" : 1,
38
+ "Ios17.realDiv" : 1,
39
+ "Ios17.linear" : 49,
40
+ "Stack" : 1,
41
+ "Ios17.greaterEqual" : 1,
42
+ "Ios17.mul" : 12
43
+ },
44
+ "computePrecision" : "Mixed (Float16, Int32, UInt16)",
45
+ "isUpdatable" : "0",
46
+ "stateSchema" : [
47
+
48
+ ],
49
+ "availability" : {
50
+ "macOS" : "14.0",
51
+ "tvOS" : "17.0",
52
+ "visionOS" : "1.0",
53
+ "watchOS" : "10.0",
54
+ "iOS" : "17.0",
55
+ "macCatalyst" : "17.0"
56
+ },
57
+ "modelType" : {
58
+ "name" : "MLModelType_mlProgram"
59
+ },
60
+ "userDefinedMetadata" : {
61
+ "com.github.apple.coremltools.source_dialect" : "TorchScript",
62
+ "com.github.apple.coremltools.version" : "8.3.0",
63
+ "com.github.apple.coremltools.source" : "torch==2.6.0+cu124"
64
+ },
65
+ "inputSchema" : [
66
+ {
67
+ "hasShapeFlexibility" : "0",
68
+ "isOptional" : "0",
69
+ "dataType" : "Int32",
70
+ "formattedType" : "MultiArray (Int32 1 × 77)",
71
+ "shortDescription" : "",
72
+ "shape" : "[1, 77]",
73
+ "name" : "input_text",
74
+ "type" : "MultiArray"
75
+ }
76
+ ],
77
+ "generatedClassName" : "clip_text_s2",
78
+ "method" : "predict"
79
+ }
80
+ ]
CoreMLModels/clip_text_s2.mlmodelc/model.mil ADDED
The diff for this file is too large to render. See raw diff
 
CoreMLModels/clip_text_s2.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf5d0197667379dada155d1b18dc0d138e061570654f625ec79f9ba0443b15a6
3
+ size 126866880
CoreMLModels/merges_clip.txt ADDED
The diff for this file is too large to render. See raw diff
 
CoreMLModels/vocab_clip.json ADDED
The diff for this file is too large to render. See raw diff
 
model/README.md ADDED
@@ -0,0 +1,222 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apple-amlr
3
+ datasets:
4
+ - riddhimanrana/coco-fastvlm-2k-val2017
5
+ language:
6
+ - en
7
+ base_model:
8
+ - apple/FastVLM-0.5B
9
+ base_model_relation: finetune
10
+ pipeline_tag: image-text-to-text
11
+ library_name: transformers
12
+ tags:
13
+ - mlx
14
+ - finetuned
15
+ - 4bit
16
+ - llava_qwen2
17
+ - multimodal
18
+ ---
19
+
20
+ # fastvlm-0.5b-captions
21
+
22
+ ## Model Details
23
+
24
+ `fastvlm-0.5b-captions` is a finetuned version of **FastVLM-0.5B Stage 3** from the [FastVLM official repository](https://github.com/apple/ml-fastvlm), built for **efficient structured image captioning on mobile devices**. This model incorporates **LoRA fine-tuning**, **4-bit quantization**, and **MobileCLIP-S0** as its vision tower, achieving substantial RAM reductions for embedded inference. This is part of a larger research project that I'm conducting. Find out more at [orionlive.ai/research](https://orionlive.ai/research) or visit my git repo [riddhimanrana/orion](https://github.com/riddhimanrana/orion)
25
+
26
+ ### Model Description
27
+
28
+ - **Developed by:** Riddhiman Rana (fine-tuning and optimizations)
29
+ - **Model type:** VLM (Vision-Language Model)
30
+ - **Original model authors:** Pavan Kumar Anasosalu Vasu, Fartash Faghri, Chun-Liang Li, Cem Koc, Nate True, Albert Antony, Gokul Santhanam, James Gabriel, Peter Grasch, Oncel Tuzel, Hadi Pouransari
31
+ - **Language(s) (NLP):** English
32
+ - **License (base model):** apple-amlr
33
+ - **Finetuned from model:** [`apple/ml-fastvlm`](https://github.com/apple/ml-fastvlm), specifically `FastVLM-0.5B Stage 3`
34
+
35
+ ### Model Sources
36
+
37
+ <!-- Provide the basic links for the model. -->
38
+
39
+ - **Base Model Repository:** https://github.com/apple/ml-fastvlm
40
+ - **Fine-tuning Training Dataset:** https://huggingface.co/datasets/riddhimanrana/coco-fastvlm-2k-val2017
41
+ - **FastVLM Paper (CVPR 2025):** https://www.arxiv.org/abs/2412.13303
42
+
43
+ ## Uses
44
+
45
+ <table>
46
+ <tr>
47
+ <td><img src="https://huggingface.co/riddhimanrana/fastvlm-0.5b-captions/resolve/main/demo/demo.gif" alt="FastVLM - iOS App Demo"></td>
48
+ </tr>
49
+ </table>
50
+
51
+ *Demo on iPhone 13 Pro Max*
52
+
53
+ ### Direct Use
54
+
55
+ - Generating **highly detailed, structured captions** for images on mobile and embedded devices.
56
+ - Ideal for **low-resource environments** such as iPhones, MacBooks, and potentially other Apple Silicon devices via MLX and CoreML.
57
+ - Tested on iPhone 12/13 Pro Max/14 – reaching RAM usage **below 1 GB** and TTFT as low as **600ms** on higher-end iPhones.
58
+
59
+ ### Out-of-Scope Use
60
+
61
+ - This is not designed for general-purpose multimodal reasoning beyond descriptive image captioning.
62
+ - Not suitable for text-only language tasks.
63
+
64
+ ## Bias, Risks, and Limitations
65
+
66
+ - Dataset was limited to **2,000 images from COCO 2017 Validation** – captions may reflect biases in that dataset.
67
+ - The model’s structured captions might occasionally be verbose or repetitive depending on input complexity.
68
+ - Accuracy for extremely abstract or unfamiliar visual scenes may degrade.
69
+
70
+
71
+ ### Recommendations
72
+
73
+
74
+ ## How to Get Started with the Model
75
+
76
+ To run inference of PyTorch checkpoint, follow the instruction below. I recommend you go through [apple/ml-fastvlm](https://github.com/apple/ml-fastvlm) for further instructions on inference on Apple Silicon and other devices.
77
+
78
+ ```python
79
+ python predict.py --model-path /path/to/checkpoint-dir \
80
+ --image-file /path/to/image.png \
81
+ --prompt "Describe the image."
82
+ ```
83
+
84
+ The prompt I used for the dataset, in training, and in practice is:
85
+ ```
86
+ You are a vision-language model that analyzes images for context-aware reasoning.
87
+ Given a visual scene, generate a rich, structured, and detailed description that includes:\n\n
88
+ 1. Main Focus – What is the primary object, person, or action in the scene?\n
89
+ 2. Surrounding Objects & Context – List and describe notable secondary objects, people, or environment details.\n
90
+ 3. Spatial Relationships – Describe where the objects are relative to one another.\n
91
+ 4. Activities & Interactions – What are people or objects doing? Are there interactions or implied motions?\n
92
+ 5. Scene Type & Time – Describe the overall type of scene (e.g. urban street, kitchen, park) and visible time of day.\n
93
+ 6. Inferences & Intent – Based on visual cues, infer what might have just happened or what might happen next.\n
94
+ 7. Style & Aesthetic – Describe the scene’s mood, lighting, or style (e.g. bright, moody, colorful).\n\n
95
+ Your goal: make your description so complete and detailed that an image generator could reconstruct the scene with full visual accuracy from your output alone.
96
+ ```
97
+
98
+ ## Training Details
99
+
100
+ ### Training Data
101
+
102
+ * **Training data:** [`riddhimanrana/coco-fastvlm-2k-val2017`](https://huggingface.co/datasets/riddhimanrana/coco-fastvlm-2k-val2017)
103
+ * **Device:** MacBook Pro 16" (M2 Pro, 16GB RAM, Apple Silicon)
104
+ * **Vision tower:** [`MobileCLIP-S0`](https://github.com/apple/ml-mobileclip)
105
+ * **Lora parameters:**
106
+ * `r=128`
107
+ * `alpha=256`
108
+ * `Dropout = 0.1`
109
+ * Applied to the language model using PEFT
110
+ * **Epochs:** `1`
111
+ * **Model max tokens:** `512`
112
+ * **Quantization:** 4-bit (post-training, MLX conversion)
113
+
114
+ ### Training Procedure
115
+
116
+ #### Preprocessing
117
+
118
+ - Image aspect ratio padded to 256×256.
119
+ - Object detection tags from YOLOv11n were added at the start of each prompt.
120
+ - All prompts followed a structured, 7-point captioning rubric.
121
+ - Inputs were clipped at 512 tokens.
122
+
123
+ #### Training Hyperparameters
124
+ | Hyperparameter | Value |
125
+ | ---------------------- | ------------------------------------ |
126
+ | Precision | `fp32` (Apple Silicon, no bf16/fp16) |
127
+ | Learning rate | `2e-4` |
128
+ | Weight decay | `0.0` |
129
+ | Warmup ratio | `0.03` |
130
+ | Scheduler | `cosine` |
131
+ | Batch size (train) | `8` |
132
+ | Batch size (eval) | `4` |
133
+ | Gradient accumulation | `1` |
134
+ | Max token length | `512` |
135
+ | Logging steps | `1` |
136
+ | Evaluation strategy | `no` |
137
+ | Save strategy | `steps` (default step interval) |
138
+ | Gradient checkpointing | `True` |
139
+ | Lazy preprocessing | `True` |
140
+ | DataLoader workers | `4` |
141
+ #### Speeds, Sizes, Times
142
+
143
+ Training duration: ~1.2 hours on M2 Pro (1 epoch over 2k samples)
144
+
145
+ Peak RAM usage: ~11.5 GB
146
+
147
+ Merged model size: 3.0 GB (pre-quantization)
148
+
149
+ Post-quantization size: ~864 MB (MLX-quantized, 4-bit)
150
+
151
+ Inference memory on iPhone (MLX): ~980MB-1.2GB RAM with 256 token generation
152
+
153
+ All devices were fed the same image. However, this model is only compatible with iPhone 12 and newer models. It has been tested on iPhone 11, but it doesn’t work due to incompatibility issues with Apple MLX support and smaller neural engines.
154
+ | Device | Chip | RAM | TTFT | Generation |
155
+ |-------------------|--------|------|--------|------------|
156
+ | iPhone 12 | A14 | 4GB | 2392ms | 73.5 tok/s |
157
+ | iPhone 13 Pro Max | A15 | 6GB | 1138ms | 74.1 tok/s |
158
+ | iPhone 14 | A15 | 6GB | 1069ms | 71.3 tok/s |
159
+ | MacBook Air 2020 | M1 | 8GB | 673ms | 131 tok/s |
160
+ ## Evaluation
161
+
162
+ ### Testing Data, Factors & Metrics
163
+
164
+ #### Testing Data
165
+
166
+ - A subset of COCO val2017 images was manually evaluated.
167
+ - Dataset includes both common and edge cases: animals, street scenes, closeups, occlusion, and indoor scenes.
168
+
169
+ #### Factors
170
+
171
+ - Image complexity (single vs multi-object)
172
+ - Scene type (indoor vs outdoor)
173
+ - Visual density
174
+ - Prompt diversity (7-point rubric compliance)
175
+
176
+ #### Metrics
177
+
178
+ *Due to the direction of my current project, evaluation metrics weren’t particularly important so I didn't spend much time on it. However, I am open to community contributions for model evaluation.*
179
+
180
+ - **Human Evaluation** (1–5 scale):
181
+ - Completeness: How well the description matches the visible scene
182
+ - Structure: Coherence of the response relative to the 7-part prompt
183
+ - Detail & Accuracy: Visual correctness of relationships and entities
184
+ - **Quantitative** (for future release):
185
+ - CIDEr / METEOR / BLEU-4 (planned via COCO eval pipeline)
186
+
187
+ ### Results
188
+ | Metric | Avg Score |
189
+ | --------------- | --------- |
190
+ | Completeness | `4.6 / 5` |
191
+ | Structure | `4.8 / 5` |
192
+ | Visual Accuracy | `4.5 / 5` |
193
+ #### Summary
194
+
195
+ The model produces rich, well-structured, and highly relevant captions optimized for real-time mobile inference. With ~930 MB size and <1 GB RAM usage, it is deployable on older iPhones w/o Apple Intelligence(e.g., iPhone 12 or newer). Despite fine-tuning on just 2,000 examples, its reasoning capability generalizes well due to the high-quality distilled prompts.
196
+
197
+ ## Environmental Impact
198
+
199
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
200
+
201
+ - **Hardware Type:** MacBook Air M1 (dataset generation), MacBook Pro M2 Pro (training, quantization)
202
+ - **Hours used:** ~3 hours for dataset, ~1h for training
203
+ - **Compute Region:** Local / personal hardware
204
+ - **Carbon Emitted:** Minimal, due to small dataset size and single-device compute.
205
+
206
+ ## Citation
207
+
208
+ **BibTeX:**
209
+
210
+ ```bibtex
211
+ @InProceedings{fastvlm2025,
212
+ author = {Pavan Kumar Anasosalu Vasu, Fartash Faghri, Chun-Liang Li, Cem Koc, Nate True, Albert Antony, Gokul Santhanam, James Gabriel, Peter Grasch, Oncel Tuzel, Hadi Pouransari},
213
+ title = {FastVLM: Efficient Vision Encoding for Vision Language Models},
214
+ booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
215
+ month = {June},
216
+ year = {2025}
217
+ }
218
+ ```
219
+
220
+ ## Model Card Contact
221
+
222
+ Contact: @riddhimanrana on Hugging Face or GitHub
model/added_tokens.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "<image>": 151646,
3
+ "<|endoftext|>": 151643,
4
+ "<|im_end|>": 151645,
5
+ "<|im_start|>": 151644
6
+ }
model/config.json ADDED
@@ -0,0 +1,50 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlavaQwen2ForCausalLM"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "bos_token_id": 151643,
7
+ "eos_token_id": 151645,
8
+ "freeze_mm_mlp_adapter": false,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 896,
11
+ "image_aspect_ratio": "pad",
12
+ "image_grid_pinpoints": null,
13
+ "image_token_index": 151646,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 4864,
16
+ "max_position_embeddings": 32768,
17
+ "max_window_layers": 24,
18
+ "mm_hidden_size": 3072,
19
+ "mm_patch_merge_type": "flat",
20
+ "mm_projector_lr": null,
21
+ "mm_projector_type": "mlp2x_gelu",
22
+ "mm_use_im_patch_token": false,
23
+ "mm_use_im_start_end": false,
24
+ "mm_vision_select_feature": "patch",
25
+ "mm_vision_select_layer": -2,
26
+ "mm_vision_tower": "mobileclip_l_1024",
27
+ "model_type": "llava_qwen2",
28
+ "num_attention_heads": 14,
29
+ "num_hidden_layers": 24,
30
+ "num_key_value_heads": 2,
31
+ "quantization": {
32
+ "group_size": 64,
33
+ "bits": 4
34
+ },
35
+ "rms_norm_eps": 1e-06,
36
+ "rope_theta": 1000000.0,
37
+ "sliding_window": 32768,
38
+ "tie_word_embeddings": false,
39
+ "tokenizer_model_max_length": 8192,
40
+ "tokenizer_padding_side": "right",
41
+ "torch_dtype": "bfloat16",
42
+ "transformers_version": "4.39.3",
43
+ "tune_mm_mlp_adapter": false,
44
+ "unfreeze_mm_vision_tower": true,
45
+ "use_cache": true,
46
+ "use_mm_proj": true,
47
+ "use_sliding_window": false,
48
+ "vision_config": {},
49
+ "vocab_size": 151936
50
+ }
model/fastvithd.mlpackage/Data/com.apple.CoreML/model.mlmodel ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c83bdaeb1b378c90a0af03a7b736d5acc13c5544432ea2df8fbe735037e037c
3
+ size 323557
model/fastvithd.mlpackage/Data/com.apple.CoreML/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93776b5c75c327a0d6a114f49f58c9dad17b1202aa3aee52b088799ab2080339
3
+ size 490786624
model/fastvithd.mlpackage/Manifest.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8d00bf654d43569b6dd2221b4dc8ead254460442638d59b781a5921d6a2a507
3
+ size 617
model/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2094ccb332cf37122cdd3583a12a3a193d3ce2b58dcc1cf45c95e5147559147
3
+ size 356655116
model/model.safetensors.index.json ADDED
@@ -0,0 +1,646 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 356575488
4
+ },
5
+ "weight_map": {
6
+ "language_model.lm_head.biases": "model.safetensors",
7
+ "language_model.lm_head.scales": "model.safetensors",
8
+ "language_model.lm_head.weight": "model.safetensors",
9
+ "language_model.model.embed_tokens.biases": "model.safetensors",
10
+ "language_model.model.embed_tokens.scales": "model.safetensors",
11
+ "language_model.model.embed_tokens.weight": "model.safetensors",
12
+ "language_model.model.layers.0.input_layernorm.weight": "model.safetensors",
13
+ "language_model.model.layers.0.mlp.down_proj.biases": "model.safetensors",
14
+ "language_model.model.layers.0.mlp.down_proj.scales": "model.safetensors",
15
+ "language_model.model.layers.0.mlp.down_proj.weight": "model.safetensors",
16
+ "language_model.model.layers.0.mlp.gate_proj.biases": "model.safetensors",
17
+ "language_model.model.layers.0.mlp.gate_proj.scales": "model.safetensors",
18
+ "language_model.model.layers.0.mlp.gate_proj.weight": "model.safetensors",
19
+ "language_model.model.layers.0.mlp.up_proj.biases": "model.safetensors",
20
+ "language_model.model.layers.0.mlp.up_proj.scales": "model.safetensors",
21
+ "language_model.model.layers.0.mlp.up_proj.weight": "model.safetensors",
22
+ "language_model.model.layers.0.post_attention_layernorm.weight": "model.safetensors",
23
+ "language_model.model.layers.0.self_attn.k_proj.bias": "model.safetensors",
24
+ "language_model.model.layers.0.self_attn.k_proj.biases": "model.safetensors",
25
+ "language_model.model.layers.0.self_attn.k_proj.scales": "model.safetensors",
26
+ "language_model.model.layers.0.self_attn.k_proj.weight": "model.safetensors",
27
+ "language_model.model.layers.0.self_attn.o_proj.biases": "model.safetensors",
28
+ "language_model.model.layers.0.self_attn.o_proj.scales": "model.safetensors",
29
+ "language_model.model.layers.0.self_attn.o_proj.weight": "model.safetensors",
30
+ "language_model.model.layers.0.self_attn.q_proj.bias": "model.safetensors",
31
+ "language_model.model.layers.0.self_attn.q_proj.biases": "model.safetensors",
32
+ "language_model.model.layers.0.self_attn.q_proj.scales": "model.safetensors",
33
+ "language_model.model.layers.0.self_attn.q_proj.weight": "model.safetensors",
34
+ "language_model.model.layers.0.self_attn.v_proj.bias": "model.safetensors",
35
+ "language_model.model.layers.0.self_attn.v_proj.biases": "model.safetensors",
36
+ "language_model.model.layers.0.self_attn.v_proj.scales": "model.safetensors",
37
+ "language_model.model.layers.0.self_attn.v_proj.weight": "model.safetensors",
38
+ "language_model.model.layers.1.input_layernorm.weight": "model.safetensors",
39
+ "language_model.model.layers.1.mlp.down_proj.biases": "model.safetensors",
40
+ "language_model.model.layers.1.mlp.down_proj.scales": "model.safetensors",
41
+ "language_model.model.layers.1.mlp.down_proj.weight": "model.safetensors",
42
+ "language_model.model.layers.1.mlp.gate_proj.biases": "model.safetensors",
43
+ "language_model.model.layers.1.mlp.gate_proj.scales": "model.safetensors",
44
+ "language_model.model.layers.1.mlp.gate_proj.weight": "model.safetensors",
45
+ "language_model.model.layers.1.mlp.up_proj.biases": "model.safetensors",
46
+ "language_model.model.layers.1.mlp.up_proj.scales": "model.safetensors",
47
+ "language_model.model.layers.1.mlp.up_proj.weight": "model.safetensors",
48
+ "language_model.model.layers.1.post_attention_layernorm.weight": "model.safetensors",
49
+ "language_model.model.layers.1.self_attn.k_proj.bias": "model.safetensors",
50
+ "language_model.model.layers.1.self_attn.k_proj.biases": "model.safetensors",
51
+ "language_model.model.layers.1.self_attn.k_proj.scales": "model.safetensors",
52
+ "language_model.model.layers.1.self_attn.k_proj.weight": "model.safetensors",
53
+ "language_model.model.layers.1.self_attn.o_proj.biases": "model.safetensors",
54
+ "language_model.model.layers.1.self_attn.o_proj.scales": "model.safetensors",
55
+ "language_model.model.layers.1.self_attn.o_proj.weight": "model.safetensors",
56
+ "language_model.model.layers.1.self_attn.q_proj.bias": "model.safetensors",
57
+ "language_model.model.layers.1.self_attn.q_proj.biases": "model.safetensors",
58
+ "language_model.model.layers.1.self_attn.q_proj.scales": "model.safetensors",
59
+ "language_model.model.layers.1.self_attn.q_proj.weight": "model.safetensors",
60
+ "language_model.model.layers.1.self_attn.v_proj.bias": "model.safetensors",
61
+ "language_model.model.layers.1.self_attn.v_proj.biases": "model.safetensors",
62
+ "language_model.model.layers.1.self_attn.v_proj.scales": "model.safetensors",
63
+ "language_model.model.layers.1.self_attn.v_proj.weight": "model.safetensors",
64
+ "language_model.model.layers.10.input_layernorm.weight": "model.safetensors",
65
+ "language_model.model.layers.10.mlp.down_proj.biases": "model.safetensors",
66
+ "language_model.model.layers.10.mlp.down_proj.scales": "model.safetensors",
67
+ "language_model.model.layers.10.mlp.down_proj.weight": "model.safetensors",
68
+ "language_model.model.layers.10.mlp.gate_proj.biases": "model.safetensors",
69
+ "language_model.model.layers.10.mlp.gate_proj.scales": "model.safetensors",
70
+ "language_model.model.layers.10.mlp.gate_proj.weight": "model.safetensors",
71
+ "language_model.model.layers.10.mlp.up_proj.biases": "model.safetensors",
72
+ "language_model.model.layers.10.mlp.up_proj.scales": "model.safetensors",
73
+ "language_model.model.layers.10.mlp.up_proj.weight": "model.safetensors",
74
+ "language_model.model.layers.10.post_attention_layernorm.weight": "model.safetensors",
75
+ "language_model.model.layers.10.self_attn.k_proj.bias": "model.safetensors",
76
+ "language_model.model.layers.10.self_attn.k_proj.biases": "model.safetensors",
77
+ "language_model.model.layers.10.self_attn.k_proj.scales": "model.safetensors",
78
+ "language_model.model.layers.10.self_attn.k_proj.weight": "model.safetensors",
79
+ "language_model.model.layers.10.self_attn.o_proj.biases": "model.safetensors",
80
+ "language_model.model.layers.10.self_attn.o_proj.scales": "model.safetensors",
81
+ "language_model.model.layers.10.self_attn.o_proj.weight": "model.safetensors",
82
+ "language_model.model.layers.10.self_attn.q_proj.bias": "model.safetensors",
83
+ "language_model.model.layers.10.self_attn.q_proj.biases": "model.safetensors",
84
+ "language_model.model.layers.10.self_attn.q_proj.scales": "model.safetensors",
85
+ "language_model.model.layers.10.self_attn.q_proj.weight": "model.safetensors",
86
+ "language_model.model.layers.10.self_attn.v_proj.bias": "model.safetensors",
87
+ "language_model.model.layers.10.self_attn.v_proj.biases": "model.safetensors",
88
+ "language_model.model.layers.10.self_attn.v_proj.scales": "model.safetensors",
89
+ "language_model.model.layers.10.self_attn.v_proj.weight": "model.safetensors",
90
+ "language_model.model.layers.11.input_layernorm.weight": "model.safetensors",
91
+ "language_model.model.layers.11.mlp.down_proj.biases": "model.safetensors",
92
+ "language_model.model.layers.11.mlp.down_proj.scales": "model.safetensors",
93
+ "language_model.model.layers.11.mlp.down_proj.weight": "model.safetensors",
94
+ "language_model.model.layers.11.mlp.gate_proj.biases": "model.safetensors",
95
+ "language_model.model.layers.11.mlp.gate_proj.scales": "model.safetensors",
96
+ "language_model.model.layers.11.mlp.gate_proj.weight": "model.safetensors",
97
+ "language_model.model.layers.11.mlp.up_proj.biases": "model.safetensors",
98
+ "language_model.model.layers.11.mlp.up_proj.scales": "model.safetensors",
99
+ "language_model.model.layers.11.mlp.up_proj.weight": "model.safetensors",
100
+ "language_model.model.layers.11.post_attention_layernorm.weight": "model.safetensors",
101
+ "language_model.model.layers.11.self_attn.k_proj.bias": "model.safetensors",
102
+ "language_model.model.layers.11.self_attn.k_proj.biases": "model.safetensors",
103
+ "language_model.model.layers.11.self_attn.k_proj.scales": "model.safetensors",
104
+ "language_model.model.layers.11.self_attn.k_proj.weight": "model.safetensors",
105
+ "language_model.model.layers.11.self_attn.o_proj.biases": "model.safetensors",
106
+ "language_model.model.layers.11.self_attn.o_proj.scales": "model.safetensors",
107
+ "language_model.model.layers.11.self_attn.o_proj.weight": "model.safetensors",
108
+ "language_model.model.layers.11.self_attn.q_proj.bias": "model.safetensors",
109
+ "language_model.model.layers.11.self_attn.q_proj.biases": "model.safetensors",
110
+ "language_model.model.layers.11.self_attn.q_proj.scales": "model.safetensors",
111
+ "language_model.model.layers.11.self_attn.q_proj.weight": "model.safetensors",
112
+ "language_model.model.layers.11.self_attn.v_proj.bias": "model.safetensors",
113
+ "language_model.model.layers.11.self_attn.v_proj.biases": "model.safetensors",
114
+ "language_model.model.layers.11.self_attn.v_proj.scales": "model.safetensors",
115
+ "language_model.model.layers.11.self_attn.v_proj.weight": "model.safetensors",
116
+ "language_model.model.layers.12.input_layernorm.weight": "model.safetensors",
117
+ "language_model.model.layers.12.mlp.down_proj.biases": "model.safetensors",
118
+ "language_model.model.layers.12.mlp.down_proj.scales": "model.safetensors",
119
+ "language_model.model.layers.12.mlp.down_proj.weight": "model.safetensors",
120
+ "language_model.model.layers.12.mlp.gate_proj.biases": "model.safetensors",
121
+ "language_model.model.layers.12.mlp.gate_proj.scales": "model.safetensors",
122
+ "language_model.model.layers.12.mlp.gate_proj.weight": "model.safetensors",
123
+ "language_model.model.layers.12.mlp.up_proj.biases": "model.safetensors",
124
+ "language_model.model.layers.12.mlp.up_proj.scales": "model.safetensors",
125
+ "language_model.model.layers.12.mlp.up_proj.weight": "model.safetensors",
126
+ "language_model.model.layers.12.post_attention_layernorm.weight": "model.safetensors",
127
+ "language_model.model.layers.12.self_attn.k_proj.bias": "model.safetensors",
128
+ "language_model.model.layers.12.self_attn.k_proj.biases": "model.safetensors",
129
+ "language_model.model.layers.12.self_attn.k_proj.scales": "model.safetensors",
130
+ "language_model.model.layers.12.self_attn.k_proj.weight": "model.safetensors",
131
+ "language_model.model.layers.12.self_attn.o_proj.biases": "model.safetensors",
132
+ "language_model.model.layers.12.self_attn.o_proj.scales": "model.safetensors",
133
+ "language_model.model.layers.12.self_attn.o_proj.weight": "model.safetensors",
134
+ "language_model.model.layers.12.self_attn.q_proj.bias": "model.safetensors",
135
+ "language_model.model.layers.12.self_attn.q_proj.biases": "model.safetensors",
136
+ "language_model.model.layers.12.self_attn.q_proj.scales": "model.safetensors",
137
+ "language_model.model.layers.12.self_attn.q_proj.weight": "model.safetensors",
138
+ "language_model.model.layers.12.self_attn.v_proj.bias": "model.safetensors",
139
+ "language_model.model.layers.12.self_attn.v_proj.biases": "model.safetensors",
140
+ "language_model.model.layers.12.self_attn.v_proj.scales": "model.safetensors",
141
+ "language_model.model.layers.12.self_attn.v_proj.weight": "model.safetensors",
142
+ "language_model.model.layers.13.input_layernorm.weight": "model.safetensors",
143
+ "language_model.model.layers.13.mlp.down_proj.biases": "model.safetensors",
144
+ "language_model.model.layers.13.mlp.down_proj.scales": "model.safetensors",
145
+ "language_model.model.layers.13.mlp.down_proj.weight": "model.safetensors",
146
+ "language_model.model.layers.13.mlp.gate_proj.biases": "model.safetensors",
147
+ "language_model.model.layers.13.mlp.gate_proj.scales": "model.safetensors",
148
+ "language_model.model.layers.13.mlp.gate_proj.weight": "model.safetensors",
149
+ "language_model.model.layers.13.mlp.up_proj.biases": "model.safetensors",
150
+ "language_model.model.layers.13.mlp.up_proj.scales": "model.safetensors",
151
+ "language_model.model.layers.13.mlp.up_proj.weight": "model.safetensors",
152
+ "language_model.model.layers.13.post_attention_layernorm.weight": "model.safetensors",
153
+ "language_model.model.layers.13.self_attn.k_proj.bias": "model.safetensors",
154
+ "language_model.model.layers.13.self_attn.k_proj.biases": "model.safetensors",
155
+ "language_model.model.layers.13.self_attn.k_proj.scales": "model.safetensors",
156
+ "language_model.model.layers.13.self_attn.k_proj.weight": "model.safetensors",
157
+ "language_model.model.layers.13.self_attn.o_proj.biases": "model.safetensors",
158
+ "language_model.model.layers.13.self_attn.o_proj.scales": "model.safetensors",
159
+ "language_model.model.layers.13.self_attn.o_proj.weight": "model.safetensors",
160
+ "language_model.model.layers.13.self_attn.q_proj.bias": "model.safetensors",
161
+ "language_model.model.layers.13.self_attn.q_proj.biases": "model.safetensors",
162
+ "language_model.model.layers.13.self_attn.q_proj.scales": "model.safetensors",
163
+ "language_model.model.layers.13.self_attn.q_proj.weight": "model.safetensors",
164
+ "language_model.model.layers.13.self_attn.v_proj.bias": "model.safetensors",
165
+ "language_model.model.layers.13.self_attn.v_proj.biases": "model.safetensors",
166
+ "language_model.model.layers.13.self_attn.v_proj.scales": "model.safetensors",
167
+ "language_model.model.layers.13.self_attn.v_proj.weight": "model.safetensors",
168
+ "language_model.model.layers.14.input_layernorm.weight": "model.safetensors",
169
+ "language_model.model.layers.14.mlp.down_proj.biases": "model.safetensors",
170
+ "language_model.model.layers.14.mlp.down_proj.scales": "model.safetensors",
171
+ "language_model.model.layers.14.mlp.down_proj.weight": "model.safetensors",
172
+ "language_model.model.layers.14.mlp.gate_proj.biases": "model.safetensors",
173
+ "language_model.model.layers.14.mlp.gate_proj.scales": "model.safetensors",
174
+ "language_model.model.layers.14.mlp.gate_proj.weight": "model.safetensors",
175
+ "language_model.model.layers.14.mlp.up_proj.biases": "model.safetensors",
176
+ "language_model.model.layers.14.mlp.up_proj.scales": "model.safetensors",
177
+ "language_model.model.layers.14.mlp.up_proj.weight": "model.safetensors",
178
+ "language_model.model.layers.14.post_attention_layernorm.weight": "model.safetensors",
179
+ "language_model.model.layers.14.self_attn.k_proj.bias": "model.safetensors",
180
+ "language_model.model.layers.14.self_attn.k_proj.biases": "model.safetensors",
181
+ "language_model.model.layers.14.self_attn.k_proj.scales": "model.safetensors",
182
+ "language_model.model.layers.14.self_attn.k_proj.weight": "model.safetensors",
183
+ "language_model.model.layers.14.self_attn.o_proj.biases": "model.safetensors",
184
+ "language_model.model.layers.14.self_attn.o_proj.scales": "model.safetensors",
185
+ "language_model.model.layers.14.self_attn.o_proj.weight": "model.safetensors",
186
+ "language_model.model.layers.14.self_attn.q_proj.bias": "model.safetensors",
187
+ "language_model.model.layers.14.self_attn.q_proj.biases": "model.safetensors",
188
+ "language_model.model.layers.14.self_attn.q_proj.scales": "model.safetensors",
189
+ "language_model.model.layers.14.self_attn.q_proj.weight": "model.safetensors",
190
+ "language_model.model.layers.14.self_attn.v_proj.bias": "model.safetensors",
191
+ "language_model.model.layers.14.self_attn.v_proj.biases": "model.safetensors",
192
+ "language_model.model.layers.14.self_attn.v_proj.scales": "model.safetensors",
193
+ "language_model.model.layers.14.self_attn.v_proj.weight": "model.safetensors",
194
+ "language_model.model.layers.15.input_layernorm.weight": "model.safetensors",
195
+ "language_model.model.layers.15.mlp.down_proj.biases": "model.safetensors",
196
+ "language_model.model.layers.15.mlp.down_proj.scales": "model.safetensors",
197
+ "language_model.model.layers.15.mlp.down_proj.weight": "model.safetensors",
198
+ "language_model.model.layers.15.mlp.gate_proj.biases": "model.safetensors",
199
+ "language_model.model.layers.15.mlp.gate_proj.scales": "model.safetensors",
200
+ "language_model.model.layers.15.mlp.gate_proj.weight": "model.safetensors",
201
+ "language_model.model.layers.15.mlp.up_proj.biases": "model.safetensors",
202
+ "language_model.model.layers.15.mlp.up_proj.scales": "model.safetensors",
203
+ "language_model.model.layers.15.mlp.up_proj.weight": "model.safetensors",
204
+ "language_model.model.layers.15.post_attention_layernorm.weight": "model.safetensors",
205
+ "language_model.model.layers.15.self_attn.k_proj.bias": "model.safetensors",
206
+ "language_model.model.layers.15.self_attn.k_proj.biases": "model.safetensors",
207
+ "language_model.model.layers.15.self_attn.k_proj.scales": "model.safetensors",
208
+ "language_model.model.layers.15.self_attn.k_proj.weight": "model.safetensors",
209
+ "language_model.model.layers.15.self_attn.o_proj.biases": "model.safetensors",
210
+ "language_model.model.layers.15.self_attn.o_proj.scales": "model.safetensors",
211
+ "language_model.model.layers.15.self_attn.o_proj.weight": "model.safetensors",
212
+ "language_model.model.layers.15.self_attn.q_proj.bias": "model.safetensors",
213
+ "language_model.model.layers.15.self_attn.q_proj.biases": "model.safetensors",
214
+ "language_model.model.layers.15.self_attn.q_proj.scales": "model.safetensors",
215
+ "language_model.model.layers.15.self_attn.q_proj.weight": "model.safetensors",
216
+ "language_model.model.layers.15.self_attn.v_proj.bias": "model.safetensors",
217
+ "language_model.model.layers.15.self_attn.v_proj.biases": "model.safetensors",
218
+ "language_model.model.layers.15.self_attn.v_proj.scales": "model.safetensors",
219
+ "language_model.model.layers.15.self_attn.v_proj.weight": "model.safetensors",
220
+ "language_model.model.layers.16.input_layernorm.weight": "model.safetensors",
221
+ "language_model.model.layers.16.mlp.down_proj.biases": "model.safetensors",
222
+ "language_model.model.layers.16.mlp.down_proj.scales": "model.safetensors",
223
+ "language_model.model.layers.16.mlp.down_proj.weight": "model.safetensors",
224
+ "language_model.model.layers.16.mlp.gate_proj.biases": "model.safetensors",
225
+ "language_model.model.layers.16.mlp.gate_proj.scales": "model.safetensors",
226
+ "language_model.model.layers.16.mlp.gate_proj.weight": "model.safetensors",
227
+ "language_model.model.layers.16.mlp.up_proj.biases": "model.safetensors",
228
+ "language_model.model.layers.16.mlp.up_proj.scales": "model.safetensors",
229
+ "language_model.model.layers.16.mlp.up_proj.weight": "model.safetensors",
230
+ "language_model.model.layers.16.post_attention_layernorm.weight": "model.safetensors",
231
+ "language_model.model.layers.16.self_attn.k_proj.bias": "model.safetensors",
232
+ "language_model.model.layers.16.self_attn.k_proj.biases": "model.safetensors",
233
+ "language_model.model.layers.16.self_attn.k_proj.scales": "model.safetensors",
234
+ "language_model.model.layers.16.self_attn.k_proj.weight": "model.safetensors",
235
+ "language_model.model.layers.16.self_attn.o_proj.biases": "model.safetensors",
236
+ "language_model.model.layers.16.self_attn.o_proj.scales": "model.safetensors",
237
+ "language_model.model.layers.16.self_attn.o_proj.weight": "model.safetensors",
238
+ "language_model.model.layers.16.self_attn.q_proj.bias": "model.safetensors",
239
+ "language_model.model.layers.16.self_attn.q_proj.biases": "model.safetensors",
240
+ "language_model.model.layers.16.self_attn.q_proj.scales": "model.safetensors",
241
+ "language_model.model.layers.16.self_attn.q_proj.weight": "model.safetensors",
242
+ "language_model.model.layers.16.self_attn.v_proj.bias": "model.safetensors",
243
+ "language_model.model.layers.16.self_attn.v_proj.biases": "model.safetensors",
244
+ "language_model.model.layers.16.self_attn.v_proj.scales": "model.safetensors",
245
+ "language_model.model.layers.16.self_attn.v_proj.weight": "model.safetensors",
246
+ "language_model.model.layers.17.input_layernorm.weight": "model.safetensors",
247
+ "language_model.model.layers.17.mlp.down_proj.biases": "model.safetensors",
248
+ "language_model.model.layers.17.mlp.down_proj.scales": "model.safetensors",
249
+ "language_model.model.layers.17.mlp.down_proj.weight": "model.safetensors",
250
+ "language_model.model.layers.17.mlp.gate_proj.biases": "model.safetensors",
251
+ "language_model.model.layers.17.mlp.gate_proj.scales": "model.safetensors",
252
+ "language_model.model.layers.17.mlp.gate_proj.weight": "model.safetensors",
253
+ "language_model.model.layers.17.mlp.up_proj.biases": "model.safetensors",
254
+ "language_model.model.layers.17.mlp.up_proj.scales": "model.safetensors",
255
+ "language_model.model.layers.17.mlp.up_proj.weight": "model.safetensors",
256
+ "language_model.model.layers.17.post_attention_layernorm.weight": "model.safetensors",
257
+ "language_model.model.layers.17.self_attn.k_proj.bias": "model.safetensors",
258
+ "language_model.model.layers.17.self_attn.k_proj.biases": "model.safetensors",
259
+ "language_model.model.layers.17.self_attn.k_proj.scales": "model.safetensors",
260
+ "language_model.model.layers.17.self_attn.k_proj.weight": "model.safetensors",
261
+ "language_model.model.layers.17.self_attn.o_proj.biases": "model.safetensors",
262
+ "language_model.model.layers.17.self_attn.o_proj.scales": "model.safetensors",
263
+ "language_model.model.layers.17.self_attn.o_proj.weight": "model.safetensors",
264
+ "language_model.model.layers.17.self_attn.q_proj.bias": "model.safetensors",
265
+ "language_model.model.layers.17.self_attn.q_proj.biases": "model.safetensors",
266
+ "language_model.model.layers.17.self_attn.q_proj.scales": "model.safetensors",
267
+ "language_model.model.layers.17.self_attn.q_proj.weight": "model.safetensors",
268
+ "language_model.model.layers.17.self_attn.v_proj.bias": "model.safetensors",
269
+ "language_model.model.layers.17.self_attn.v_proj.biases": "model.safetensors",
270
+ "language_model.model.layers.17.self_attn.v_proj.scales": "model.safetensors",
271
+ "language_model.model.layers.17.self_attn.v_proj.weight": "model.safetensors",
272
+ "language_model.model.layers.18.input_layernorm.weight": "model.safetensors",
273
+ "language_model.model.layers.18.mlp.down_proj.biases": "model.safetensors",
274
+ "language_model.model.layers.18.mlp.down_proj.scales": "model.safetensors",
275
+ "language_model.model.layers.18.mlp.down_proj.weight": "model.safetensors",
276
+ "language_model.model.layers.18.mlp.gate_proj.biases": "model.safetensors",
277
+ "language_model.model.layers.18.mlp.gate_proj.scales": "model.safetensors",
278
+ "language_model.model.layers.18.mlp.gate_proj.weight": "model.safetensors",
279
+ "language_model.model.layers.18.mlp.up_proj.biases": "model.safetensors",
280
+ "language_model.model.layers.18.mlp.up_proj.scales": "model.safetensors",
281
+ "language_model.model.layers.18.mlp.up_proj.weight": "model.safetensors",
282
+ "language_model.model.layers.18.post_attention_layernorm.weight": "model.safetensors",
283
+ "language_model.model.layers.18.self_attn.k_proj.bias": "model.safetensors",
284
+ "language_model.model.layers.18.self_attn.k_proj.biases": "model.safetensors",
285
+ "language_model.model.layers.18.self_attn.k_proj.scales": "model.safetensors",
286
+ "language_model.model.layers.18.self_attn.k_proj.weight": "model.safetensors",
287
+ "language_model.model.layers.18.self_attn.o_proj.biases": "model.safetensors",
288
+ "language_model.model.layers.18.self_attn.o_proj.scales": "model.safetensors",
289
+ "language_model.model.layers.18.self_attn.o_proj.weight": "model.safetensors",
290
+ "language_model.model.layers.18.self_attn.q_proj.bias": "model.safetensors",
291
+ "language_model.model.layers.18.self_attn.q_proj.biases": "model.safetensors",
292
+ "language_model.model.layers.18.self_attn.q_proj.scales": "model.safetensors",
293
+ "language_model.model.layers.18.self_attn.q_proj.weight": "model.safetensors",
294
+ "language_model.model.layers.18.self_attn.v_proj.bias": "model.safetensors",
295
+ "language_model.model.layers.18.self_attn.v_proj.biases": "model.safetensors",
296
+ "language_model.model.layers.18.self_attn.v_proj.scales": "model.safetensors",
297
+ "language_model.model.layers.18.self_attn.v_proj.weight": "model.safetensors",
298
+ "language_model.model.layers.19.input_layernorm.weight": "model.safetensors",
299
+ "language_model.model.layers.19.mlp.down_proj.biases": "model.safetensors",
300
+ "language_model.model.layers.19.mlp.down_proj.scales": "model.safetensors",
301
+ "language_model.model.layers.19.mlp.down_proj.weight": "model.safetensors",
302
+ "language_model.model.layers.19.mlp.gate_proj.biases": "model.safetensors",
303
+ "language_model.model.layers.19.mlp.gate_proj.scales": "model.safetensors",
304
+ "language_model.model.layers.19.mlp.gate_proj.weight": "model.safetensors",
305
+ "language_model.model.layers.19.mlp.up_proj.biases": "model.safetensors",
306
+ "language_model.model.layers.19.mlp.up_proj.scales": "model.safetensors",
307
+ "language_model.model.layers.19.mlp.up_proj.weight": "model.safetensors",
308
+ "language_model.model.layers.19.post_attention_layernorm.weight": "model.safetensors",
309
+ "language_model.model.layers.19.self_attn.k_proj.bias": "model.safetensors",
310
+ "language_model.model.layers.19.self_attn.k_proj.biases": "model.safetensors",
311
+ "language_model.model.layers.19.self_attn.k_proj.scales": "model.safetensors",
312
+ "language_model.model.layers.19.self_attn.k_proj.weight": "model.safetensors",
313
+ "language_model.model.layers.19.self_attn.o_proj.biases": "model.safetensors",
314
+ "language_model.model.layers.19.self_attn.o_proj.scales": "model.safetensors",
315
+ "language_model.model.layers.19.self_attn.o_proj.weight": "model.safetensors",
316
+ "language_model.model.layers.19.self_attn.q_proj.bias": "model.safetensors",
317
+ "language_model.model.layers.19.self_attn.q_proj.biases": "model.safetensors",
318
+ "language_model.model.layers.19.self_attn.q_proj.scales": "model.safetensors",
319
+ "language_model.model.layers.19.self_attn.q_proj.weight": "model.safetensors",
320
+ "language_model.model.layers.19.self_attn.v_proj.bias": "model.safetensors",
321
+ "language_model.model.layers.19.self_attn.v_proj.biases": "model.safetensors",
322
+ "language_model.model.layers.19.self_attn.v_proj.scales": "model.safetensors",
323
+ "language_model.model.layers.19.self_attn.v_proj.weight": "model.safetensors",
324
+ "language_model.model.layers.2.input_layernorm.weight": "model.safetensors",
325
+ "language_model.model.layers.2.mlp.down_proj.biases": "model.safetensors",
326
+ "language_model.model.layers.2.mlp.down_proj.scales": "model.safetensors",
327
+ "language_model.model.layers.2.mlp.down_proj.weight": "model.safetensors",
328
+ "language_model.model.layers.2.mlp.gate_proj.biases": "model.safetensors",
329
+ "language_model.model.layers.2.mlp.gate_proj.scales": "model.safetensors",
330
+ "language_model.model.layers.2.mlp.gate_proj.weight": "model.safetensors",
331
+ "language_model.model.layers.2.mlp.up_proj.biases": "model.safetensors",
332
+ "language_model.model.layers.2.mlp.up_proj.scales": "model.safetensors",
333
+ "language_model.model.layers.2.mlp.up_proj.weight": "model.safetensors",
334
+ "language_model.model.layers.2.post_attention_layernorm.weight": "model.safetensors",
335
+ "language_model.model.layers.2.self_attn.k_proj.bias": "model.safetensors",
336
+ "language_model.model.layers.2.self_attn.k_proj.biases": "model.safetensors",
337
+ "language_model.model.layers.2.self_attn.k_proj.scales": "model.safetensors",
338
+ "language_model.model.layers.2.self_attn.k_proj.weight": "model.safetensors",
339
+ "language_model.model.layers.2.self_attn.o_proj.biases": "model.safetensors",
340
+ "language_model.model.layers.2.self_attn.o_proj.scales": "model.safetensors",
341
+ "language_model.model.layers.2.self_attn.o_proj.weight": "model.safetensors",
342
+ "language_model.model.layers.2.self_attn.q_proj.bias": "model.safetensors",
343
+ "language_model.model.layers.2.self_attn.q_proj.biases": "model.safetensors",
344
+ "language_model.model.layers.2.self_attn.q_proj.scales": "model.safetensors",
345
+ "language_model.model.layers.2.self_attn.q_proj.weight": "model.safetensors",
346
+ "language_model.model.layers.2.self_attn.v_proj.bias": "model.safetensors",
347
+ "language_model.model.layers.2.self_attn.v_proj.biases": "model.safetensors",
348
+ "language_model.model.layers.2.self_attn.v_proj.scales": "model.safetensors",
349
+ "language_model.model.layers.2.self_attn.v_proj.weight": "model.safetensors",
350
+ "language_model.model.layers.20.input_layernorm.weight": "model.safetensors",
351
+ "language_model.model.layers.20.mlp.down_proj.biases": "model.safetensors",
352
+ "language_model.model.layers.20.mlp.down_proj.scales": "model.safetensors",
353
+ "language_model.model.layers.20.mlp.down_proj.weight": "model.safetensors",
354
+ "language_model.model.layers.20.mlp.gate_proj.biases": "model.safetensors",
355
+ "language_model.model.layers.20.mlp.gate_proj.scales": "model.safetensors",
356
+ "language_model.model.layers.20.mlp.gate_proj.weight": "model.safetensors",
357
+ "language_model.model.layers.20.mlp.up_proj.biases": "model.safetensors",
358
+ "language_model.model.layers.20.mlp.up_proj.scales": "model.safetensors",
359
+ "language_model.model.layers.20.mlp.up_proj.weight": "model.safetensors",
360
+ "language_model.model.layers.20.post_attention_layernorm.weight": "model.safetensors",
361
+ "language_model.model.layers.20.self_attn.k_proj.bias": "model.safetensors",
362
+ "language_model.model.layers.20.self_attn.k_proj.biases": "model.safetensors",
363
+ "language_model.model.layers.20.self_attn.k_proj.scales": "model.safetensors",
364
+ "language_model.model.layers.20.self_attn.k_proj.weight": "model.safetensors",
365
+ "language_model.model.layers.20.self_attn.o_proj.biases": "model.safetensors",
366
+ "language_model.model.layers.20.self_attn.o_proj.scales": "model.safetensors",
367
+ "language_model.model.layers.20.self_attn.o_proj.weight": "model.safetensors",
368
+ "language_model.model.layers.20.self_attn.q_proj.bias": "model.safetensors",
369
+ "language_model.model.layers.20.self_attn.q_proj.biases": "model.safetensors",
370
+ "language_model.model.layers.20.self_attn.q_proj.scales": "model.safetensors",
371
+ "language_model.model.layers.20.self_attn.q_proj.weight": "model.safetensors",
372
+ "language_model.model.layers.20.self_attn.v_proj.bias": "model.safetensors",
373
+ "language_model.model.layers.20.self_attn.v_proj.biases": "model.safetensors",
374
+ "language_model.model.layers.20.self_attn.v_proj.scales": "model.safetensors",
375
+ "language_model.model.layers.20.self_attn.v_proj.weight": "model.safetensors",
376
+ "language_model.model.layers.21.input_layernorm.weight": "model.safetensors",
377
+ "language_model.model.layers.21.mlp.down_proj.biases": "model.safetensors",
378
+ "language_model.model.layers.21.mlp.down_proj.scales": "model.safetensors",
379
+ "language_model.model.layers.21.mlp.down_proj.weight": "model.safetensors",
380
+ "language_model.model.layers.21.mlp.gate_proj.biases": "model.safetensors",
381
+ "language_model.model.layers.21.mlp.gate_proj.scales": "model.safetensors",
382
+ "language_model.model.layers.21.mlp.gate_proj.weight": "model.safetensors",
383
+ "language_model.model.layers.21.mlp.up_proj.biases": "model.safetensors",
384
+ "language_model.model.layers.21.mlp.up_proj.scales": "model.safetensors",
385
+ "language_model.model.layers.21.mlp.up_proj.weight": "model.safetensors",
386
+ "language_model.model.layers.21.post_attention_layernorm.weight": "model.safetensors",
387
+ "language_model.model.layers.21.self_attn.k_proj.bias": "model.safetensors",
388
+ "language_model.model.layers.21.self_attn.k_proj.biases": "model.safetensors",
389
+ "language_model.model.layers.21.self_attn.k_proj.scales": "model.safetensors",
390
+ "language_model.model.layers.21.self_attn.k_proj.weight": "model.safetensors",
391
+ "language_model.model.layers.21.self_attn.o_proj.biases": "model.safetensors",
392
+ "language_model.model.layers.21.self_attn.o_proj.scales": "model.safetensors",
393
+ "language_model.model.layers.21.self_attn.o_proj.weight": "model.safetensors",
394
+ "language_model.model.layers.21.self_attn.q_proj.bias": "model.safetensors",
395
+ "language_model.model.layers.21.self_attn.q_proj.biases": "model.safetensors",
396
+ "language_model.model.layers.21.self_attn.q_proj.scales": "model.safetensors",
397
+ "language_model.model.layers.21.self_attn.q_proj.weight": "model.safetensors",
398
+ "language_model.model.layers.21.self_attn.v_proj.bias": "model.safetensors",
399
+ "language_model.model.layers.21.self_attn.v_proj.biases": "model.safetensors",
400
+ "language_model.model.layers.21.self_attn.v_proj.scales": "model.safetensors",
401
+ "language_model.model.layers.21.self_attn.v_proj.weight": "model.safetensors",
402
+ "language_model.model.layers.22.input_layernorm.weight": "model.safetensors",
403
+ "language_model.model.layers.22.mlp.down_proj.biases": "model.safetensors",
404
+ "language_model.model.layers.22.mlp.down_proj.scales": "model.safetensors",
405
+ "language_model.model.layers.22.mlp.down_proj.weight": "model.safetensors",
406
+ "language_model.model.layers.22.mlp.gate_proj.biases": "model.safetensors",
407
+ "language_model.model.layers.22.mlp.gate_proj.scales": "model.safetensors",
408
+ "language_model.model.layers.22.mlp.gate_proj.weight": "model.safetensors",
409
+ "language_model.model.layers.22.mlp.up_proj.biases": "model.safetensors",
410
+ "language_model.model.layers.22.mlp.up_proj.scales": "model.safetensors",
411
+ "language_model.model.layers.22.mlp.up_proj.weight": "model.safetensors",
412
+ "language_model.model.layers.22.post_attention_layernorm.weight": "model.safetensors",
413
+ "language_model.model.layers.22.self_attn.k_proj.bias": "model.safetensors",
414
+ "language_model.model.layers.22.self_attn.k_proj.biases": "model.safetensors",
415
+ "language_model.model.layers.22.self_attn.k_proj.scales": "model.safetensors",
416
+ "language_model.model.layers.22.self_attn.k_proj.weight": "model.safetensors",
417
+ "language_model.model.layers.22.self_attn.o_proj.biases": "model.safetensors",
418
+ "language_model.model.layers.22.self_attn.o_proj.scales": "model.safetensors",
419
+ "language_model.model.layers.22.self_attn.o_proj.weight": "model.safetensors",
420
+ "language_model.model.layers.22.self_attn.q_proj.bias": "model.safetensors",
421
+ "language_model.model.layers.22.self_attn.q_proj.biases": "model.safetensors",
422
+ "language_model.model.layers.22.self_attn.q_proj.scales": "model.safetensors",
423
+ "language_model.model.layers.22.self_attn.q_proj.weight": "model.safetensors",
424
+ "language_model.model.layers.22.self_attn.v_proj.bias": "model.safetensors",
425
+ "language_model.model.layers.22.self_attn.v_proj.biases": "model.safetensors",
426
+ "language_model.model.layers.22.self_attn.v_proj.scales": "model.safetensors",
427
+ "language_model.model.layers.22.self_attn.v_proj.weight": "model.safetensors",
428
+ "language_model.model.layers.23.input_layernorm.weight": "model.safetensors",
429
+ "language_model.model.layers.23.mlp.down_proj.biases": "model.safetensors",
430
+ "language_model.model.layers.23.mlp.down_proj.scales": "model.safetensors",
431
+ "language_model.model.layers.23.mlp.down_proj.weight": "model.safetensors",
432
+ "language_model.model.layers.23.mlp.gate_proj.biases": "model.safetensors",
433
+ "language_model.model.layers.23.mlp.gate_proj.scales": "model.safetensors",
434
+ "language_model.model.layers.23.mlp.gate_proj.weight": "model.safetensors",
435
+ "language_model.model.layers.23.mlp.up_proj.biases": "model.safetensors",
436
+ "language_model.model.layers.23.mlp.up_proj.scales": "model.safetensors",
437
+ "language_model.model.layers.23.mlp.up_proj.weight": "model.safetensors",
438
+ "language_model.model.layers.23.post_attention_layernorm.weight": "model.safetensors",
439
+ "language_model.model.layers.23.self_attn.k_proj.bias": "model.safetensors",
440
+ "language_model.model.layers.23.self_attn.k_proj.biases": "model.safetensors",
441
+ "language_model.model.layers.23.self_attn.k_proj.scales": "model.safetensors",
442
+ "language_model.model.layers.23.self_attn.k_proj.weight": "model.safetensors",
443
+ "language_model.model.layers.23.self_attn.o_proj.biases": "model.safetensors",
444
+ "language_model.model.layers.23.self_attn.o_proj.scales": "model.safetensors",
445
+ "language_model.model.layers.23.self_attn.o_proj.weight": "model.safetensors",
446
+ "language_model.model.layers.23.self_attn.q_proj.bias": "model.safetensors",
447
+ "language_model.model.layers.23.self_attn.q_proj.biases": "model.safetensors",
448
+ "language_model.model.layers.23.self_attn.q_proj.scales": "model.safetensors",
449
+ "language_model.model.layers.23.self_attn.q_proj.weight": "model.safetensors",
450
+ "language_model.model.layers.23.self_attn.v_proj.bias": "model.safetensors",
451
+ "language_model.model.layers.23.self_attn.v_proj.biases": "model.safetensors",
452
+ "language_model.model.layers.23.self_attn.v_proj.scales": "model.safetensors",
453
+ "language_model.model.layers.23.self_attn.v_proj.weight": "model.safetensors",
454
+ "language_model.model.layers.3.input_layernorm.weight": "model.safetensors",
455
+ "language_model.model.layers.3.mlp.down_proj.biases": "model.safetensors",
456
+ "language_model.model.layers.3.mlp.down_proj.scales": "model.safetensors",
457
+ "language_model.model.layers.3.mlp.down_proj.weight": "model.safetensors",
458
+ "language_model.model.layers.3.mlp.gate_proj.biases": "model.safetensors",
459
+ "language_model.model.layers.3.mlp.gate_proj.scales": "model.safetensors",
460
+ "language_model.model.layers.3.mlp.gate_proj.weight": "model.safetensors",
461
+ "language_model.model.layers.3.mlp.up_proj.biases": "model.safetensors",
462
+ "language_model.model.layers.3.mlp.up_proj.scales": "model.safetensors",
463
+ "language_model.model.layers.3.mlp.up_proj.weight": "model.safetensors",
464
+ "language_model.model.layers.3.post_attention_layernorm.weight": "model.safetensors",
465
+ "language_model.model.layers.3.self_attn.k_proj.bias": "model.safetensors",
466
+ "language_model.model.layers.3.self_attn.k_proj.biases": "model.safetensors",
467
+ "language_model.model.layers.3.self_attn.k_proj.scales": "model.safetensors",
468
+ "language_model.model.layers.3.self_attn.k_proj.weight": "model.safetensors",
469
+ "language_model.model.layers.3.self_attn.o_proj.biases": "model.safetensors",
470
+ "language_model.model.layers.3.self_attn.o_proj.scales": "model.safetensors",
471
+ "language_model.model.layers.3.self_attn.o_proj.weight": "model.safetensors",
472
+ "language_model.model.layers.3.self_attn.q_proj.bias": "model.safetensors",
473
+ "language_model.model.layers.3.self_attn.q_proj.biases": "model.safetensors",
474
+ "language_model.model.layers.3.self_attn.q_proj.scales": "model.safetensors",
475
+ "language_model.model.layers.3.self_attn.q_proj.weight": "model.safetensors",
476
+ "language_model.model.layers.3.self_attn.v_proj.bias": "model.safetensors",
477
+ "language_model.model.layers.3.self_attn.v_proj.biases": "model.safetensors",
478
+ "language_model.model.layers.3.self_attn.v_proj.scales": "model.safetensors",
479
+ "language_model.model.layers.3.self_attn.v_proj.weight": "model.safetensors",
480
+ "language_model.model.layers.4.input_layernorm.weight": "model.safetensors",
481
+ "language_model.model.layers.4.mlp.down_proj.biases": "model.safetensors",
482
+ "language_model.model.layers.4.mlp.down_proj.scales": "model.safetensors",
483
+ "language_model.model.layers.4.mlp.down_proj.weight": "model.safetensors",
484
+ "language_model.model.layers.4.mlp.gate_proj.biases": "model.safetensors",
485
+ "language_model.model.layers.4.mlp.gate_proj.scales": "model.safetensors",
486
+ "language_model.model.layers.4.mlp.gate_proj.weight": "model.safetensors",
487
+ "language_model.model.layers.4.mlp.up_proj.biases": "model.safetensors",
488
+ "language_model.model.layers.4.mlp.up_proj.scales": "model.safetensors",
489
+ "language_model.model.layers.4.mlp.up_proj.weight": "model.safetensors",
490
+ "language_model.model.layers.4.post_attention_layernorm.weight": "model.safetensors",
491
+ "language_model.model.layers.4.self_attn.k_proj.bias": "model.safetensors",
492
+ "language_model.model.layers.4.self_attn.k_proj.biases": "model.safetensors",
493
+ "language_model.model.layers.4.self_attn.k_proj.scales": "model.safetensors",
494
+ "language_model.model.layers.4.self_attn.k_proj.weight": "model.safetensors",
495
+ "language_model.model.layers.4.self_attn.o_proj.biases": "model.safetensors",
496
+ "language_model.model.layers.4.self_attn.o_proj.scales": "model.safetensors",
497
+ "language_model.model.layers.4.self_attn.o_proj.weight": "model.safetensors",
498
+ "language_model.model.layers.4.self_attn.q_proj.bias": "model.safetensors",
499
+ "language_model.model.layers.4.self_attn.q_proj.biases": "model.safetensors",
500
+ "language_model.model.layers.4.self_attn.q_proj.scales": "model.safetensors",
501
+ "language_model.model.layers.4.self_attn.q_proj.weight": "model.safetensors",
502
+ "language_model.model.layers.4.self_attn.v_proj.bias": "model.safetensors",
503
+ "language_model.model.layers.4.self_attn.v_proj.biases": "model.safetensors",
504
+ "language_model.model.layers.4.self_attn.v_proj.scales": "model.safetensors",
505
+ "language_model.model.layers.4.self_attn.v_proj.weight": "model.safetensors",
506
+ "language_model.model.layers.5.input_layernorm.weight": "model.safetensors",
507
+ "language_model.model.layers.5.mlp.down_proj.biases": "model.safetensors",
508
+ "language_model.model.layers.5.mlp.down_proj.scales": "model.safetensors",
509
+ "language_model.model.layers.5.mlp.down_proj.weight": "model.safetensors",
510
+ "language_model.model.layers.5.mlp.gate_proj.biases": "model.safetensors",
511
+ "language_model.model.layers.5.mlp.gate_proj.scales": "model.safetensors",
512
+ "language_model.model.layers.5.mlp.gate_proj.weight": "model.safetensors",
513
+ "language_model.model.layers.5.mlp.up_proj.biases": "model.safetensors",
514
+ "language_model.model.layers.5.mlp.up_proj.scales": "model.safetensors",
515
+ "language_model.model.layers.5.mlp.up_proj.weight": "model.safetensors",
516
+ "language_model.model.layers.5.post_attention_layernorm.weight": "model.safetensors",
517
+ "language_model.model.layers.5.self_attn.k_proj.bias": "model.safetensors",
518
+ "language_model.model.layers.5.self_attn.k_proj.biases": "model.safetensors",
519
+ "language_model.model.layers.5.self_attn.k_proj.scales": "model.safetensors",
520
+ "language_model.model.layers.5.self_attn.k_proj.weight": "model.safetensors",
521
+ "language_model.model.layers.5.self_attn.o_proj.biases": "model.safetensors",
522
+ "language_model.model.layers.5.self_attn.o_proj.scales": "model.safetensors",
523
+ "language_model.model.layers.5.self_attn.o_proj.weight": "model.safetensors",
524
+ "language_model.model.layers.5.self_attn.q_proj.bias": "model.safetensors",
525
+ "language_model.model.layers.5.self_attn.q_proj.biases": "model.safetensors",
526
+ "language_model.model.layers.5.self_attn.q_proj.scales": "model.safetensors",
527
+ "language_model.model.layers.5.self_attn.q_proj.weight": "model.safetensors",
528
+ "language_model.model.layers.5.self_attn.v_proj.bias": "model.safetensors",
529
+ "language_model.model.layers.5.self_attn.v_proj.biases": "model.safetensors",
530
+ "language_model.model.layers.5.self_attn.v_proj.scales": "model.safetensors",
531
+ "language_model.model.layers.5.self_attn.v_proj.weight": "model.safetensors",
532
+ "language_model.model.layers.6.input_layernorm.weight": "model.safetensors",
533
+ "language_model.model.layers.6.mlp.down_proj.biases": "model.safetensors",
534
+ "language_model.model.layers.6.mlp.down_proj.scales": "model.safetensors",
535
+ "language_model.model.layers.6.mlp.down_proj.weight": "model.safetensors",
536
+ "language_model.model.layers.6.mlp.gate_proj.biases": "model.safetensors",
537
+ "language_model.model.layers.6.mlp.gate_proj.scales": "model.safetensors",
538
+ "language_model.model.layers.6.mlp.gate_proj.weight": "model.safetensors",
539
+ "language_model.model.layers.6.mlp.up_proj.biases": "model.safetensors",
540
+ "language_model.model.layers.6.mlp.up_proj.scales": "model.safetensors",
541
+ "language_model.model.layers.6.mlp.up_proj.weight": "model.safetensors",
542
+ "language_model.model.layers.6.post_attention_layernorm.weight": "model.safetensors",
543
+ "language_model.model.layers.6.self_attn.k_proj.bias": "model.safetensors",
544
+ "language_model.model.layers.6.self_attn.k_proj.biases": "model.safetensors",
545
+ "language_model.model.layers.6.self_attn.k_proj.scales": "model.safetensors",
546
+ "language_model.model.layers.6.self_attn.k_proj.weight": "model.safetensors",
547
+ "language_model.model.layers.6.self_attn.o_proj.biases": "model.safetensors",
548
+ "language_model.model.layers.6.self_attn.o_proj.scales": "model.safetensors",
549
+ "language_model.model.layers.6.self_attn.o_proj.weight": "model.safetensors",
550
+ "language_model.model.layers.6.self_attn.q_proj.bias": "model.safetensors",
551
+ "language_model.model.layers.6.self_attn.q_proj.biases": "model.safetensors",
552
+ "language_model.model.layers.6.self_attn.q_proj.scales": "model.safetensors",
553
+ "language_model.model.layers.6.self_attn.q_proj.weight": "model.safetensors",
554
+ "language_model.model.layers.6.self_attn.v_proj.bias": "model.safetensors",
555
+ "language_model.model.layers.6.self_attn.v_proj.biases": "model.safetensors",
556
+ "language_model.model.layers.6.self_attn.v_proj.scales": "model.safetensors",
557
+ "language_model.model.layers.6.self_attn.v_proj.weight": "model.safetensors",
558
+ "language_model.model.layers.7.input_layernorm.weight": "model.safetensors",
559
+ "language_model.model.layers.7.mlp.down_proj.biases": "model.safetensors",
560
+ "language_model.model.layers.7.mlp.down_proj.scales": "model.safetensors",
561
+ "language_model.model.layers.7.mlp.down_proj.weight": "model.safetensors",
562
+ "language_model.model.layers.7.mlp.gate_proj.biases": "model.safetensors",
563
+ "language_model.model.layers.7.mlp.gate_proj.scales": "model.safetensors",
564
+ "language_model.model.layers.7.mlp.gate_proj.weight": "model.safetensors",
565
+ "language_model.model.layers.7.mlp.up_proj.biases": "model.safetensors",
566
+ "language_model.model.layers.7.mlp.up_proj.scales": "model.safetensors",
567
+ "language_model.model.layers.7.mlp.up_proj.weight": "model.safetensors",
568
+ "language_model.model.layers.7.post_attention_layernorm.weight": "model.safetensors",
569
+ "language_model.model.layers.7.self_attn.k_proj.bias": "model.safetensors",
570
+ "language_model.model.layers.7.self_attn.k_proj.biases": "model.safetensors",
571
+ "language_model.model.layers.7.self_attn.k_proj.scales": "model.safetensors",
572
+ "language_model.model.layers.7.self_attn.k_proj.weight": "model.safetensors",
573
+ "language_model.model.layers.7.self_attn.o_proj.biases": "model.safetensors",
574
+ "language_model.model.layers.7.self_attn.o_proj.scales": "model.safetensors",
575
+ "language_model.model.layers.7.self_attn.o_proj.weight": "model.safetensors",
576
+ "language_model.model.layers.7.self_attn.q_proj.bias": "model.safetensors",
577
+ "language_model.model.layers.7.self_attn.q_proj.biases": "model.safetensors",
578
+ "language_model.model.layers.7.self_attn.q_proj.scales": "model.safetensors",
579
+ "language_model.model.layers.7.self_attn.q_proj.weight": "model.safetensors",
580
+ "language_model.model.layers.7.self_attn.v_proj.bias": "model.safetensors",
581
+ "language_model.model.layers.7.self_attn.v_proj.biases": "model.safetensors",
582
+ "language_model.model.layers.7.self_attn.v_proj.scales": "model.safetensors",
583
+ "language_model.model.layers.7.self_attn.v_proj.weight": "model.safetensors",
584
+ "language_model.model.layers.8.input_layernorm.weight": "model.safetensors",
585
+ "language_model.model.layers.8.mlp.down_proj.biases": "model.safetensors",
586
+ "language_model.model.layers.8.mlp.down_proj.scales": "model.safetensors",
587
+ "language_model.model.layers.8.mlp.down_proj.weight": "model.safetensors",
588
+ "language_model.model.layers.8.mlp.gate_proj.biases": "model.safetensors",
589
+ "language_model.model.layers.8.mlp.gate_proj.scales": "model.safetensors",
590
+ "language_model.model.layers.8.mlp.gate_proj.weight": "model.safetensors",
591
+ "language_model.model.layers.8.mlp.up_proj.biases": "model.safetensors",
592
+ "language_model.model.layers.8.mlp.up_proj.scales": "model.safetensors",
593
+ "language_model.model.layers.8.mlp.up_proj.weight": "model.safetensors",
594
+ "language_model.model.layers.8.post_attention_layernorm.weight": "model.safetensors",
595
+ "language_model.model.layers.8.self_attn.k_proj.bias": "model.safetensors",
596
+ "language_model.model.layers.8.self_attn.k_proj.biases": "model.safetensors",
597
+ "language_model.model.layers.8.self_attn.k_proj.scales": "model.safetensors",
598
+ "language_model.model.layers.8.self_attn.k_proj.weight": "model.safetensors",
599
+ "language_model.model.layers.8.self_attn.o_proj.biases": "model.safetensors",
600
+ "language_model.model.layers.8.self_attn.o_proj.scales": "model.safetensors",
601
+ "language_model.model.layers.8.self_attn.o_proj.weight": "model.safetensors",
602
+ "language_model.model.layers.8.self_attn.q_proj.bias": "model.safetensors",
603
+ "language_model.model.layers.8.self_attn.q_proj.biases": "model.safetensors",
604
+ "language_model.model.layers.8.self_attn.q_proj.scales": "model.safetensors",
605
+ "language_model.model.layers.8.self_attn.q_proj.weight": "model.safetensors",
606
+ "language_model.model.layers.8.self_attn.v_proj.bias": "model.safetensors",
607
+ "language_model.model.layers.8.self_attn.v_proj.biases": "model.safetensors",
608
+ "language_model.model.layers.8.self_attn.v_proj.scales": "model.safetensors",
609
+ "language_model.model.layers.8.self_attn.v_proj.weight": "model.safetensors",
610
+ "language_model.model.layers.9.input_layernorm.weight": "model.safetensors",
611
+ "language_model.model.layers.9.mlp.down_proj.biases": "model.safetensors",
612
+ "language_model.model.layers.9.mlp.down_proj.scales": "model.safetensors",
613
+ "language_model.model.layers.9.mlp.down_proj.weight": "model.safetensors",
614
+ "language_model.model.layers.9.mlp.gate_proj.biases": "model.safetensors",
615
+ "language_model.model.layers.9.mlp.gate_proj.scales": "model.safetensors",
616
+ "language_model.model.layers.9.mlp.gate_proj.weight": "model.safetensors",
617
+ "language_model.model.layers.9.mlp.up_proj.biases": "model.safetensors",
618
+ "language_model.model.layers.9.mlp.up_proj.scales": "model.safetensors",
619
+ "language_model.model.layers.9.mlp.up_proj.weight": "model.safetensors",
620
+ "language_model.model.layers.9.post_attention_layernorm.weight": "model.safetensors",
621
+ "language_model.model.layers.9.self_attn.k_proj.bias": "model.safetensors",
622
+ "language_model.model.layers.9.self_attn.k_proj.biases": "model.safetensors",
623
+ "language_model.model.layers.9.self_attn.k_proj.scales": "model.safetensors",
624
+ "language_model.model.layers.9.self_attn.k_proj.weight": "model.safetensors",
625
+ "language_model.model.layers.9.self_attn.o_proj.biases": "model.safetensors",
626
+ "language_model.model.layers.9.self_attn.o_proj.scales": "model.safetensors",
627
+ "language_model.model.layers.9.self_attn.o_proj.weight": "model.safetensors",
628
+ "language_model.model.layers.9.self_attn.q_proj.bias": "model.safetensors",
629
+ "language_model.model.layers.9.self_attn.q_proj.biases": "model.safetensors",
630
+ "language_model.model.layers.9.self_attn.q_proj.scales": "model.safetensors",
631
+ "language_model.model.layers.9.self_attn.q_proj.weight": "model.safetensors",
632
+ "language_model.model.layers.9.self_attn.v_proj.bias": "model.safetensors",
633
+ "language_model.model.layers.9.self_attn.v_proj.biases": "model.safetensors",
634
+ "language_model.model.layers.9.self_attn.v_proj.scales": "model.safetensors",
635
+ "language_model.model.layers.9.self_attn.v_proj.weight": "model.safetensors",
636
+ "language_model.model.norm.weight": "model.safetensors",
637
+ "multi_modal_projector.linear_0.bias": "model.safetensors",
638
+ "multi_modal_projector.linear_0.biases": "model.safetensors",
639
+ "multi_modal_projector.linear_0.scales": "model.safetensors",
640
+ "multi_modal_projector.linear_0.weight": "model.safetensors",
641
+ "multi_modal_projector.linear_2.bias": "model.safetensors",
642
+ "multi_modal_projector.linear_2.biases": "model.safetensors",
643
+ "multi_modal_projector.linear_2.scales": "model.safetensors",
644
+ "multi_modal_projector.linear_2.weight": "model.safetensors"
645
+ }
646
+ }
model/preprocessor_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": {
3
+ "height": 1024,
4
+ "width": 1024
5
+ },
6
+ "do_center_crop": true,
7
+ "do_convert_rgb": true,
8
+ "do_normalize": true,
9
+ "do_rescale": true,
10
+ "do_resize": true,
11
+ "image_mean": [
12
+ 0.0,
13
+ 0.0,
14
+ 0.0
15
+ ],
16
+ "image_processor_type": "CLIPImageProcessor",
17
+ "image_std": [
18
+ 1.0,
19
+ 1.0,
20
+ 1.0
21
+ ],
22
+ "processor_class": "LlavaProcessor",
23
+ "resample": 3,
24
+ "rescale_factor": 0.00392156862745098,
25
+ "size": {
26
+ "shortest_edge": 1024
27
+ }
28
+ }
model/processor_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_token": "<image>",
3
+ "num_additional_image_tokens": 0,
4
+ "patch_size": 64,
5
+ "processor_class": "LlavaProcessor",
6
+ "vision_feature_select_strategy": null
7
+ }
model/special_tokens_map.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>"
5
+ ],
6
+ "eos_token": {
7
+ "content": "<|im_end|>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false
12
+ },
13
+ "pad_token": {
14
+ "content": "<|endoftext|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false
19
+ }
20
+ }
model/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8bb55926dbf36523cd143a5805102a9a516df89f7010313d182e1e710d94fb15
3
+ size 11413284
model/tokenizer_config.json ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "151643": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "151644": {
13
+ "content": "<|im_start|>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "151645": {
21
+ "content": "<|im_end|>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "151646": {
29
+ "content": "<image>",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ }
36
+ },
37
+ "additional_special_tokens": ["<|im_start|>", "<|im_end|>"],
38
+ "bos_token": null,
39
+ "chat_template": "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful vision-language model that analyzes images for context-aware reasoning.<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
40
+ "clean_up_tokenization_spaces": false,
41
+ "eos_token": "<|im_end|>",
42
+ "errors": "replace",
43
+ "extra_special_tokens": {},
44
+ "model_max_length": 8192,
45
+ "only_llm": true,
46
+ "pad_token": "<|endoftext|>",
47
+ "padding_side": "right",
48
+ "processor_class": "LlavaProcessor",
49
+ "split_special_tokens": false,
50
+ "tokenizer_class": "Qwen2Tokenizer",
51
+ "unk_token": null
52
+ }
model/vocab.json ADDED
The diff for this file is too large to render. See raw diff