clocksmith commited on
Commit
09caa4d
·
verified ·
1 Parent(s): c9597d5

Publish gemma-3-270m-it-wq4k-ef16-hf16 canonical artifact

Browse files
.gitattributes CHANGED
@@ -41,3 +41,4 @@ models/google-embeddinggemma-300m-q4k-ehf16-af32/tokenizer.json filter=lfs diff=
41
  models/gemma-3-1b-it-q4k-ehf16-af32/tokenizer.json filter=lfs diff=lfs merge=lfs -text
42
  models/gemma-3-1b-it-f16-af32/tokenizer.json filter=lfs diff=lfs merge=lfs -text
43
  models/translategemma-4b-it-q4k-ehf16-af32/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 
 
41
  models/gemma-3-1b-it-q4k-ehf16-af32/tokenizer.json filter=lfs diff=lfs merge=lfs -text
42
  models/gemma-3-1b-it-f16-af32/tokenizer.json filter=lfs diff=lfs merge=lfs -text
43
  models/translategemma-4b-it-q4k-ehf16-af32/tokenizer.json filter=lfs diff=lfs merge=lfs -text
44
+ models/gemma-3-270m-it-wq4k-ef16-hf16/tokenizer.json filter=lfs diff=lfs merge=lfs -text
models/gemma-3-270m-it-wq4k-ef16-hf16/manifest.json ADDED
@@ -0,0 +1,3731 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "version": 1,
3
+ "modelId": "gemma-3-270m-it-wq4k-ef16-hf16",
4
+ "modelType": "transformer",
5
+ "quantization": "Q4_K_M",
6
+ "quantizationInfo": {
7
+ "weights": "q4k",
8
+ "embeddings": "f16",
9
+ "compute": "f32",
10
+ "layout": "row",
11
+ "variantTag": "wq4k-ef16"
12
+ },
13
+ "architecture": {
14
+ "numLayers": 18,
15
+ "hiddenSize": 640,
16
+ "intermediateSize": 2048,
17
+ "numAttentionHeads": 4,
18
+ "numKeyValueHeads": 1,
19
+ "headDim": 256,
20
+ "vocabSize": 262144,
21
+ "maxSeqLen": 32768,
22
+ "ropeTheta": 1000000
23
+ },
24
+ "moeConfig": null,
25
+ "inference": {
26
+ "schema": "doppler.execution/v0",
27
+ "presetId": "gemma3",
28
+ "attention": {
29
+ "queryPreAttnScalar": 256,
30
+ "attnLogitSoftcapping": null,
31
+ "slidingWindow": 512,
32
+ "queryKeyNorm": true,
33
+ "attentionOutputGate": false,
34
+ "causal": true,
35
+ "attentionBias": false
36
+ },
37
+ "normalization": {
38
+ "rmsNormEps": 0.000001,
39
+ "rmsNormWeightOffset": true,
40
+ "postAttentionNorm": true,
41
+ "preFeedforwardNorm": true,
42
+ "postFeedforwardNorm": true
43
+ },
44
+ "ffn": {
45
+ "activation": "gelu",
46
+ "gatedActivation": true,
47
+ "swigluLimit": null
48
+ },
49
+ "rope": {
50
+ "ropeTheta": 1000000,
51
+ "ropeLocalTheta": 10000,
52
+ "mropeInterleaved": false,
53
+ "mropeSection": null,
54
+ "partialRotaryFactor": null,
55
+ "ropeScalingType": null,
56
+ "ropeScalingFactor": 1,
57
+ "yarnBetaFast": null,
58
+ "yarnBetaSlow": null,
59
+ "yarnOriginalMaxPos": null,
60
+ "ropeLocalScalingType": null,
61
+ "ropeLocalScalingFactor": 1,
62
+ "ropeLocalYarnBetaFast": null,
63
+ "ropeLocalYarnBetaSlow": null,
64
+ "ropeLocalYarnOriginalMaxPos": null
65
+ },
66
+ "output": {
67
+ "finalLogitSoftcapping": null,
68
+ "tieWordEmbeddings": true,
69
+ "scaleEmbeddings": true,
70
+ "embeddingTranspose": false,
71
+ "embeddingVocabSize": 262144
72
+ },
73
+ "layerPattern": {
74
+ "type": "every_n",
75
+ "globalPattern": null,
76
+ "period": 6,
77
+ "offset": 5,
78
+ "layerTypes": null
79
+ },
80
+ "chatTemplate": {
81
+ "type": "gemma",
82
+ "enabled": true
83
+ },
84
+ "pipeline": null,
85
+ "sessionDefaults": {
86
+ "compute": {
87
+ "defaults": {
88
+ "activationDtype": "f32",
89
+ "mathDtype": "f32",
90
+ "accumDtype": "f32",
91
+ "outputDtype": "f32"
92
+ },
93
+ "kernelProfiles": [
94
+ {
95
+ "kernelRef": {
96
+ "id": "gather.main",
97
+ "version": "1.0.0",
98
+ "digest": "sha256:777991fb6e4b3b506e4493b47ee998afe541924ddd7c04e1eadf4cb7fd719ef8"
99
+ }
100
+ },
101
+ {
102
+ "kernelRef": {
103
+ "id": "rmsnorm.main",
104
+ "version": "1.0.0",
105
+ "digest": "sha256:c529986befb29a04b94d89744585923a7cef82baf4b2b0a243aa2431618622cc"
106
+ }
107
+ },
108
+ {
109
+ "kernelRef": {
110
+ "id": "gather.f16.main",
111
+ "version": "1.0.0",
112
+ "digest": "sha256:a4829f4067091c98ad6ebbc9b0744cdd5bbcd4fbf6092b2f7cc7f1098695860f"
113
+ }
114
+ },
115
+ {
116
+ "kernelRef": {
117
+ "id": "matmul.gemv.subgroup.main.vec4",
118
+ "version": "1.0.0",
119
+ "digest": "sha256:3cee3bed453b40c5564a751d2a917649e10ad52f5268e77cbfecfcee34780457"
120
+ }
121
+ },
122
+ {
123
+ "kernelRef": {
124
+ "id": "rope.main",
125
+ "version": "1.0.0",
126
+ "digest": "sha256:b639fe8a54508115c82c13c923bfea89f59c6e15a5bef66bfc34e12f0ab4e32f"
127
+ }
128
+ },
129
+ {
130
+ "kernelRef": {
131
+ "id": "attention.decode.online.f16kv.main",
132
+ "version": "1.0.0",
133
+ "digest": "sha256:4c5d8c92a0a111af716d6b46b9559446807c086027445c7fefa150202f43dae4"
134
+ }
135
+ },
136
+ {
137
+ "kernelRef": {
138
+ "id": "residual.main",
139
+ "version": "1.0.0",
140
+ "digest": "sha256:1fc456b14e2fb2bc9627107b4e51e7a2098f723b5ba6ab5542cd9455af99f423"
141
+ }
142
+ },
143
+ {
144
+ "kernelRef": {
145
+ "id": "gelu.main",
146
+ "version": "1.0.0",
147
+ "digest": "sha256:a9007ea08aaff98f9be08f1e0490a6bcf252883eac5513de876ab9ce918865e6"
148
+ }
149
+ },
150
+ {
151
+ "kernelRef": {
152
+ "id": "matmul.f16w.f32a.tiled.main",
153
+ "version": "1.0.0",
154
+ "digest": "sha256:e94ae5374e8b43dd48b663eff59a45c822c3784d5702a1145266b6ffd15ba78c"
155
+ }
156
+ },
157
+ {
158
+ "kernelRef": {
159
+ "id": "matmul.f16w.f32a.main",
160
+ "version": "1.0.0",
161
+ "digest": "sha256:027a8f1cd9713cbe0b0ada160bd175e0542bb90896ad85441b023522d9a1befc"
162
+ }
163
+ },
164
+ {
165
+ "kernelRef": {
166
+ "id": "attention.small.f16kv.main",
167
+ "version": "1.0.0",
168
+ "digest": "sha256:7e3acc24b8b45294d18052c759319fe8a202d4ebfc4b653d4b904b245ae7e5c9"
169
+ }
170
+ },
171
+ {
172
+ "kernelRef": {
173
+ "id": "attention.streaming.f16kv.main",
174
+ "version": "1.0.0",
175
+ "digest": "sha256:b337a2dcf7b40431a733d1726eee8bf23504136fd5f915bec057fb59f3ea7480"
176
+ }
177
+ },
178
+ {
179
+ "kernelRef": {
180
+ "id": "matmul.gemv.subgroup.main.multicol",
181
+ "version": "1.0.0",
182
+ "digest": "sha256:96c38c15e6fed0d7efdc5cd094db5843a8e8ddfe01eee3bc7322fa555dacf3d0"
183
+ }
184
+ },
185
+ {
186
+ "kernelRef": {
187
+ "id": "sample.sample.single.pass",
188
+ "version": "1.0.0",
189
+ "digest": "sha256:4412357e84113ee2f1bc0dc8bf89e314c2ab482c89c14ca016ea9949d16a9d0c"
190
+ }
191
+ }
192
+ ]
193
+ },
194
+ "kvcache": {
195
+ "kvDtype": "f16"
196
+ },
197
+ "decodeLoop": null
198
+ },
199
+ "execution": {
200
+ "steps": [
201
+ {
202
+ "id": "preLayer_both_0_embed",
203
+ "phase": "both",
204
+ "section": "preLayer",
205
+ "op": "embed",
206
+ "kernel": "gather_f16.wgsl",
207
+ "entry": "main",
208
+ "weights": "embed_tokens",
209
+ "layers": "all",
210
+ "src": "state",
211
+ "dst": "state",
212
+ "kernelRef": {
213
+ "id": "gather.f16.main",
214
+ "version": "1.0.0",
215
+ "digest": "sha256:a4829f4067091c98ad6ebbc9b0744cdd5bbcd4fbf6092b2f7cc7f1098695860f"
216
+ }
217
+ },
218
+ {
219
+ "id": "layer_decode_1_input_norm",
220
+ "phase": "decode",
221
+ "section": "layer",
222
+ "op": "input_norm",
223
+ "kernel": "rmsnorm.wgsl",
224
+ "entry": "main",
225
+ "layers": "all",
226
+ "src": "state",
227
+ "dst": "state",
228
+ "kernelRef": {
229
+ "id": "rmsnorm.main",
230
+ "version": "1.0.0",
231
+ "digest": "sha256:c529986befb29a04b94d89744585923a7cef82baf4b2b0a243aa2431618622cc"
232
+ }
233
+ },
234
+ {
235
+ "id": "layer_decode_2_q_proj",
236
+ "phase": "decode",
237
+ "section": "layer",
238
+ "op": "q_proj",
239
+ "kernel": "matmul_gemv_subgroup.wgsl",
240
+ "entry": "main_vec4",
241
+ "weights": "layer.{L}.self_attn.q_proj",
242
+ "layers": "all",
243
+ "src": "state",
244
+ "dst": "state",
245
+ "kernelRef": {
246
+ "id": "matmul.gemv.subgroup.main.vec4",
247
+ "version": "1.0.0",
248
+ "digest": "sha256:3cee3bed453b40c5564a751d2a917649e10ad52f5268e77cbfecfcee34780457"
249
+ }
250
+ },
251
+ {
252
+ "id": "layer_decode_3_k_proj",
253
+ "phase": "decode",
254
+ "section": "layer",
255
+ "op": "k_proj",
256
+ "kernel": "matmul_gemv_subgroup.wgsl",
257
+ "entry": "main_vec4",
258
+ "weights": "layer.{L}.self_attn.k_proj",
259
+ "layers": "all",
260
+ "src": "state",
261
+ "dst": "state",
262
+ "kernelRef": {
263
+ "id": "matmul.gemv.subgroup.main.vec4",
264
+ "version": "1.0.0",
265
+ "digest": "sha256:3cee3bed453b40c5564a751d2a917649e10ad52f5268e77cbfecfcee34780457"
266
+ }
267
+ },
268
+ {
269
+ "id": "layer_decode_4_v_proj",
270
+ "phase": "decode",
271
+ "section": "layer",
272
+ "op": "v_proj",
273
+ "kernel": "matmul_gemv_subgroup.wgsl",
274
+ "entry": "main_vec4",
275
+ "weights": "layer.{L}.self_attn.v_proj",
276
+ "layers": "all",
277
+ "src": "state",
278
+ "dst": "state",
279
+ "kernelRef": {
280
+ "id": "matmul.gemv.subgroup.main.vec4",
281
+ "version": "1.0.0",
282
+ "digest": "sha256:3cee3bed453b40c5564a751d2a917649e10ad52f5268e77cbfecfcee34780457"
283
+ }
284
+ },
285
+ {
286
+ "id": "layer_decode_5_rope_q",
287
+ "phase": "decode",
288
+ "section": "layer",
289
+ "op": "rope_q",
290
+ "kernel": "rope.wgsl",
291
+ "entry": "main",
292
+ "layers": "all",
293
+ "src": "state",
294
+ "dst": "state",
295
+ "kernelRef": {
296
+ "id": "rope.main",
297
+ "version": "1.0.0",
298
+ "digest": "sha256:b639fe8a54508115c82c13c923bfea89f59c6e15a5bef66bfc34e12f0ab4e32f"
299
+ }
300
+ },
301
+ {
302
+ "id": "layer_decode_6_rope_k",
303
+ "phase": "decode",
304
+ "section": "layer",
305
+ "op": "rope_k",
306
+ "kernel": "rope.wgsl",
307
+ "entry": "main",
308
+ "layers": "all",
309
+ "src": "state",
310
+ "dst": "state",
311
+ "kernelRef": {
312
+ "id": "rope.main",
313
+ "version": "1.0.0",
314
+ "digest": "sha256:b639fe8a54508115c82c13c923bfea89f59c6e15a5bef66bfc34e12f0ab4e32f"
315
+ }
316
+ },
317
+ {
318
+ "id": "layer_decode_7_attention",
319
+ "phase": "decode",
320
+ "section": "layer",
321
+ "op": "attention",
322
+ "kernel": "attention_decode_online_f16kv.wgsl",
323
+ "entry": "main",
324
+ "layers": "all",
325
+ "src": "state",
326
+ "dst": "state",
327
+ "kernelRef": {
328
+ "id": "attention.decode.online.f16kv.main",
329
+ "version": "1.0.0",
330
+ "digest": "sha256:4c5d8c92a0a111af716d6b46b9559446807c086027445c7fefa150202f43dae4"
331
+ }
332
+ },
333
+ {
334
+ "id": "layer_decode_8_o_proj",
335
+ "phase": "decode",
336
+ "section": "layer",
337
+ "op": "o_proj",
338
+ "kernel": "matmul_gemv_subgroup.wgsl",
339
+ "entry": "main_vec4",
340
+ "weights": "layer.{L}.self_attn.o_proj",
341
+ "layers": "all",
342
+ "src": "state",
343
+ "dst": "state",
344
+ "kernelRef": {
345
+ "id": "matmul.gemv.subgroup.main.vec4",
346
+ "version": "1.0.0",
347
+ "digest": "sha256:3cee3bed453b40c5564a751d2a917649e10ad52f5268e77cbfecfcee34780457"
348
+ }
349
+ },
350
+ {
351
+ "id": "layer_decode_9_attn_residual",
352
+ "phase": "decode",
353
+ "section": "layer",
354
+ "op": "attn_residual",
355
+ "kernel": "residual.wgsl",
356
+ "entry": "main",
357
+ "layers": "all",
358
+ "src": "state",
359
+ "dst": "state",
360
+ "kernelRef": {
361
+ "id": "residual.main",
362
+ "version": "1.0.0",
363
+ "digest": "sha256:1fc456b14e2fb2bc9627107b4e51e7a2098f723b5ba6ab5542cd9455af99f423"
364
+ }
365
+ },
366
+ {
367
+ "id": "layer_decode_10_post_attn_norm",
368
+ "phase": "decode",
369
+ "section": "layer",
370
+ "op": "post_attn_norm",
371
+ "kernel": "rmsnorm.wgsl",
372
+ "entry": "main",
373
+ "layers": "all",
374
+ "src": "state",
375
+ "dst": "state",
376
+ "kernelRef": {
377
+ "id": "rmsnorm.main",
378
+ "version": "1.0.0",
379
+ "digest": "sha256:c529986befb29a04b94d89744585923a7cef82baf4b2b0a243aa2431618622cc"
380
+ }
381
+ },
382
+ {
383
+ "id": "layer_decode_11_gate_proj",
384
+ "phase": "decode",
385
+ "section": "layer",
386
+ "op": "gate_proj",
387
+ "kernel": "matmul_gemv_subgroup.wgsl",
388
+ "entry": "main_vec4",
389
+ "weights": "layer.{L}.mlp.gate_proj",
390
+ "layers": "all",
391
+ "src": "state",
392
+ "dst": "state",
393
+ "kernelRef": {
394
+ "id": "matmul.gemv.subgroup.main.vec4",
395
+ "version": "1.0.0",
396
+ "digest": "sha256:3cee3bed453b40c5564a751d2a917649e10ad52f5268e77cbfecfcee34780457"
397
+ }
398
+ },
399
+ {
400
+ "id": "layer_decode_12_up_proj",
401
+ "phase": "decode",
402
+ "section": "layer",
403
+ "op": "up_proj",
404
+ "kernel": "matmul_gemv_subgroup.wgsl",
405
+ "entry": "main_vec4",
406
+ "weights": "layer.{L}.mlp.up_proj",
407
+ "layers": "all",
408
+ "src": "state",
409
+ "dst": "state",
410
+ "kernelRef": {
411
+ "id": "matmul.gemv.subgroup.main.vec4",
412
+ "version": "1.0.0",
413
+ "digest": "sha256:3cee3bed453b40c5564a751d2a917649e10ad52f5268e77cbfecfcee34780457"
414
+ }
415
+ },
416
+ {
417
+ "id": "layer_decode_13_activation",
418
+ "phase": "decode",
419
+ "section": "layer",
420
+ "op": "activation",
421
+ "kernel": "gelu.wgsl",
422
+ "entry": "main",
423
+ "constants": {
424
+ "HAS_GATE": true
425
+ },
426
+ "layers": "all",
427
+ "src": "state",
428
+ "dst": "state",
429
+ "kernelRef": {
430
+ "id": "gelu.main",
431
+ "version": "1.0.0",
432
+ "digest": "sha256:a9007ea08aaff98f9be08f1e0490a6bcf252883eac5513de876ab9ce918865e6"
433
+ }
434
+ },
435
+ {
436
+ "id": "layer_decode_14_down_proj",
437
+ "phase": "decode",
438
+ "section": "layer",
439
+ "op": "down_proj",
440
+ "kernel": "matmul_gemv_subgroup.wgsl",
441
+ "entry": "main_vec4",
442
+ "weights": "layer.{L}.mlp.down_proj",
443
+ "layers": "all",
444
+ "src": "state",
445
+ "dst": "state",
446
+ "kernelRef": {
447
+ "id": "matmul.gemv.subgroup.main.vec4",
448
+ "version": "1.0.0",
449
+ "digest": "sha256:3cee3bed453b40c5564a751d2a917649e10ad52f5268e77cbfecfcee34780457"
450
+ }
451
+ },
452
+ {
453
+ "id": "layer_decode_15_ffn_residual",
454
+ "phase": "decode",
455
+ "section": "layer",
456
+ "op": "ffn_residual",
457
+ "kernel": "residual.wgsl",
458
+ "entry": "main",
459
+ "layers": "all",
460
+ "src": "state",
461
+ "dst": "state",
462
+ "kernelRef": {
463
+ "id": "residual.main",
464
+ "version": "1.0.0",
465
+ "digest": "sha256:1fc456b14e2fb2bc9627107b4e51e7a2098f723b5ba6ab5542cd9455af99f423"
466
+ }
467
+ },
468
+ {
469
+ "id": "layer_prefill_16_input_norm",
470
+ "phase": "prefill",
471
+ "section": "layer",
472
+ "op": "input_norm",
473
+ "kernel": "rmsnorm.wgsl",
474
+ "entry": "main",
475
+ "layers": "all",
476
+ "src": "state",
477
+ "dst": "state",
478
+ "kernelRef": {
479
+ "id": "rmsnorm.main",
480
+ "version": "1.0.0",
481
+ "digest": "sha256:c529986befb29a04b94d89744585923a7cef82baf4b2b0a243aa2431618622cc"
482
+ }
483
+ },
484
+ {
485
+ "id": "layer_prefill_17_q_proj",
486
+ "phase": "prefill",
487
+ "section": "layer",
488
+ "op": "q_proj",
489
+ "kernel": "matmul_f16w_f32a.wgsl",
490
+ "entry": "main",
491
+ "weights": "layer.{L}.self_attn.q_proj",
492
+ "layers": "all",
493
+ "src": "state",
494
+ "dst": "state",
495
+ "kernelRef": {
496
+ "id": "matmul.f16w.f32a.main",
497
+ "version": "1.0.0",
498
+ "digest": "sha256:027a8f1cd9713cbe0b0ada160bd175e0542bb90896ad85441b023522d9a1befc"
499
+ }
500
+ },
501
+ {
502
+ "id": "layer_prefill_18_k_proj",
503
+ "phase": "prefill",
504
+ "section": "layer",
505
+ "op": "k_proj",
506
+ "kernel": "matmul_f16w_f32a.wgsl",
507
+ "entry": "main",
508
+ "weights": "layer.{L}.self_attn.k_proj",
509
+ "layers": "all",
510
+ "src": "state",
511
+ "dst": "state",
512
+ "kernelRef": {
513
+ "id": "matmul.f16w.f32a.main",
514
+ "version": "1.0.0",
515
+ "digest": "sha256:027a8f1cd9713cbe0b0ada160bd175e0542bb90896ad85441b023522d9a1befc"
516
+ }
517
+ },
518
+ {
519
+ "id": "layer_prefill_19_v_proj",
520
+ "phase": "prefill",
521
+ "section": "layer",
522
+ "op": "v_proj",
523
+ "kernel": "matmul_f16w_f32a.wgsl",
524
+ "entry": "main",
525
+ "weights": "layer.{L}.self_attn.v_proj",
526
+ "layers": "all",
527
+ "src": "state",
528
+ "dst": "state",
529
+ "kernelRef": {
530
+ "id": "matmul.f16w.f32a.main",
531
+ "version": "1.0.0",
532
+ "digest": "sha256:027a8f1cd9713cbe0b0ada160bd175e0542bb90896ad85441b023522d9a1befc"
533
+ }
534
+ },
535
+ {
536
+ "id": "layer_prefill_20_rope_q",
537
+ "phase": "prefill",
538
+ "section": "layer",
539
+ "op": "rope_q",
540
+ "kernel": "rope.wgsl",
541
+ "entry": "main",
542
+ "layers": "all",
543
+ "src": "state",
544
+ "dst": "state",
545
+ "kernelRef": {
546
+ "id": "rope.main",
547
+ "version": "1.0.0",
548
+ "digest": "sha256:b639fe8a54508115c82c13c923bfea89f59c6e15a5bef66bfc34e12f0ab4e32f"
549
+ }
550
+ },
551
+ {
552
+ "id": "layer_prefill_21_rope_k",
553
+ "phase": "prefill",
554
+ "section": "layer",
555
+ "op": "rope_k",
556
+ "kernel": "rope.wgsl",
557
+ "entry": "main",
558
+ "layers": "all",
559
+ "src": "state",
560
+ "dst": "state",
561
+ "kernelRef": {
562
+ "id": "rope.main",
563
+ "version": "1.0.0",
564
+ "digest": "sha256:b639fe8a54508115c82c13c923bfea89f59c6e15a5bef66bfc34e12f0ab4e32f"
565
+ }
566
+ },
567
+ {
568
+ "id": "layer_prefill_22_attention",
569
+ "phase": "prefill",
570
+ "section": "layer",
571
+ "op": "attention",
572
+ "kernel": "attention_streaming_f16kv.wgsl",
573
+ "entry": "main",
574
+ "layers": "all",
575
+ "src": "state",
576
+ "dst": "state",
577
+ "kernelRef": {
578
+ "id": "attention.streaming.f16kv.main",
579
+ "version": "1.0.0",
580
+ "digest": "sha256:b337a2dcf7b40431a733d1726eee8bf23504136fd5f915bec057fb59f3ea7480"
581
+ }
582
+ },
583
+ {
584
+ "id": "layer_prefill_23_o_proj",
585
+ "phase": "prefill",
586
+ "section": "layer",
587
+ "op": "o_proj",
588
+ "kernel": "matmul_f16w_f32a.wgsl",
589
+ "entry": "main",
590
+ "weights": "layer.{L}.self_attn.o_proj",
591
+ "layers": "all",
592
+ "src": "state",
593
+ "dst": "state",
594
+ "kernelRef": {
595
+ "id": "matmul.f16w.f32a.main",
596
+ "version": "1.0.0",
597
+ "digest": "sha256:027a8f1cd9713cbe0b0ada160bd175e0542bb90896ad85441b023522d9a1befc"
598
+ }
599
+ },
600
+ {
601
+ "id": "layer_prefill_24_attn_residual",
602
+ "phase": "prefill",
603
+ "section": "layer",
604
+ "op": "attn_residual",
605
+ "kernel": "residual.wgsl",
606
+ "entry": "main",
607
+ "layers": "all",
608
+ "src": "state",
609
+ "dst": "state",
610
+ "kernelRef": {
611
+ "id": "residual.main",
612
+ "version": "1.0.0",
613
+ "digest": "sha256:1fc456b14e2fb2bc9627107b4e51e7a2098f723b5ba6ab5542cd9455af99f423"
614
+ }
615
+ },
616
+ {
617
+ "id": "layer_prefill_25_post_attn_norm",
618
+ "phase": "prefill",
619
+ "section": "layer",
620
+ "op": "post_attn_norm",
621
+ "kernel": "rmsnorm.wgsl",
622
+ "entry": "main",
623
+ "layers": "all",
624
+ "src": "state",
625
+ "dst": "state",
626
+ "kernelRef": {
627
+ "id": "rmsnorm.main",
628
+ "version": "1.0.0",
629
+ "digest": "sha256:c529986befb29a04b94d89744585923a7cef82baf4b2b0a243aa2431618622cc"
630
+ }
631
+ },
632
+ {
633
+ "id": "layer_prefill_26_gate_proj",
634
+ "phase": "prefill",
635
+ "section": "layer",
636
+ "op": "gate_proj",
637
+ "kernel": "matmul_f16w_f32a.wgsl",
638
+ "entry": "main",
639
+ "weights": "layer.{L}.mlp.gate_proj",
640
+ "layers": "all",
641
+ "src": "state",
642
+ "dst": "state",
643
+ "kernelRef": {
644
+ "id": "matmul.f16w.f32a.main",
645
+ "version": "1.0.0",
646
+ "digest": "sha256:027a8f1cd9713cbe0b0ada160bd175e0542bb90896ad85441b023522d9a1befc"
647
+ }
648
+ },
649
+ {
650
+ "id": "layer_prefill_27_up_proj",
651
+ "phase": "prefill",
652
+ "section": "layer",
653
+ "op": "up_proj",
654
+ "kernel": "matmul_f16w_f32a.wgsl",
655
+ "entry": "main",
656
+ "weights": "layer.{L}.mlp.up_proj",
657
+ "layers": "all",
658
+ "src": "state",
659
+ "dst": "state",
660
+ "kernelRef": {
661
+ "id": "matmul.f16w.f32a.main",
662
+ "version": "1.0.0",
663
+ "digest": "sha256:027a8f1cd9713cbe0b0ada160bd175e0542bb90896ad85441b023522d9a1befc"
664
+ }
665
+ },
666
+ {
667
+ "id": "layer_prefill_28_activation",
668
+ "phase": "prefill",
669
+ "section": "layer",
670
+ "op": "activation",
671
+ "kernel": "gelu.wgsl",
672
+ "entry": "main",
673
+ "constants": {
674
+ "HAS_GATE": true
675
+ },
676
+ "layers": "all",
677
+ "src": "state",
678
+ "dst": "state",
679
+ "kernelRef": {
680
+ "id": "gelu.main",
681
+ "version": "1.0.0",
682
+ "digest": "sha256:a9007ea08aaff98f9be08f1e0490a6bcf252883eac5513de876ab9ce918865e6"
683
+ }
684
+ },
685
+ {
686
+ "id": "layer_prefill_29_down_proj",
687
+ "phase": "prefill",
688
+ "section": "layer",
689
+ "op": "down_proj",
690
+ "kernel": "matmul_f16w_f32a.wgsl",
691
+ "entry": "main",
692
+ "weights": "layer.{L}.mlp.down_proj",
693
+ "layers": "all",
694
+ "src": "state",
695
+ "dst": "state",
696
+ "kernelRef": {
697
+ "id": "matmul.f16w.f32a.main",
698
+ "version": "1.0.0",
699
+ "digest": "sha256:027a8f1cd9713cbe0b0ada160bd175e0542bb90896ad85441b023522d9a1befc"
700
+ }
701
+ },
702
+ {
703
+ "id": "layer_prefill_30_ffn_residual",
704
+ "phase": "prefill",
705
+ "section": "layer",
706
+ "op": "ffn_residual",
707
+ "kernel": "residual.wgsl",
708
+ "entry": "main",
709
+ "layers": "all",
710
+ "src": "state",
711
+ "dst": "state",
712
+ "kernelRef": {
713
+ "id": "residual.main",
714
+ "version": "1.0.0",
715
+ "digest": "sha256:1fc456b14e2fb2bc9627107b4e51e7a2098f723b5ba6ab5542cd9455af99f423"
716
+ }
717
+ },
718
+ {
719
+ "id": "postLayer_both_31_final_norm",
720
+ "phase": "both",
721
+ "section": "postLayer",
722
+ "op": "final_norm",
723
+ "kernel": "rmsnorm.wgsl",
724
+ "entry": "main",
725
+ "layers": "all",
726
+ "src": "state",
727
+ "dst": "state",
728
+ "kernelRef": {
729
+ "id": "rmsnorm.main",
730
+ "version": "1.0.0",
731
+ "digest": "sha256:c529986befb29a04b94d89744585923a7cef82baf4b2b0a243aa2431618622cc"
732
+ }
733
+ },
734
+ {
735
+ "id": "postLayer_both_32_lm_head",
736
+ "phase": "both",
737
+ "section": "postLayer",
738
+ "op": "lm_head",
739
+ "kernel": "matmul_gemv_subgroup.wgsl",
740
+ "entry": "main_multicol",
741
+ "weights": "lm_head",
742
+ "constants": {
743
+ "MULTICOL_COLS_PER_WG": 64,
744
+ "MULTICOL_THREADS_PER_COL": 4
745
+ },
746
+ "layers": "all",
747
+ "src": "state",
748
+ "dst": "state",
749
+ "kernelRef": {
750
+ "id": "matmul.gemv.subgroup.main.multicol",
751
+ "version": "1.0.0",
752
+ "digest": "sha256:96c38c15e6fed0d7efdc5cd094db5843a8e8ddfe01eee3bc7322fa555dacf3d0"
753
+ }
754
+ },
755
+ {
756
+ "id": "postLayer_both_33_lm_head_prefill",
757
+ "phase": "both",
758
+ "section": "postLayer",
759
+ "op": "lm_head_prefill",
760
+ "kernel": "matmul_f16w_f32a.wgsl",
761
+ "entry": "main",
762
+ "weights": "lm_head",
763
+ "layers": "all",
764
+ "src": "state",
765
+ "dst": "state",
766
+ "kernelRef": {
767
+ "id": "matmul.f16w.f32a.main",
768
+ "version": "1.0.0",
769
+ "digest": "sha256:027a8f1cd9713cbe0b0ada160bd175e0542bb90896ad85441b023522d9a1befc"
770
+ }
771
+ },
772
+ {
773
+ "id": "sampling_decode_34_sample",
774
+ "phase": "decode",
775
+ "section": "sampling",
776
+ "op": "sample",
777
+ "kernel": "sample.wgsl",
778
+ "entry": "sample_single_pass",
779
+ "layers": "all",
780
+ "src": "state",
781
+ "dst": "state",
782
+ "kernelRef": {
783
+ "id": "sample.sample.single.pass",
784
+ "version": "1.0.0",
785
+ "digest": "sha256:4412357e84113ee2f1bc0dc8bf89e314c2ab482c89c14ca016ea9949d16a9d0c"
786
+ }
787
+ }
788
+ ],
789
+ "policies": {
790
+ "precisionPrecedence": "step_then_kernel_profile_then_session_default",
791
+ "unsupportedPrecision": "error",
792
+ "dtypeTransition": "require_cast_step",
793
+ "unresolvedKernel": "error"
794
+ }
795
+ },
796
+ "defaultKernelPath": "gemma3-q4k-dequant-f32a-online"
797
+ },
798
+ "shards": [
799
+ {
800
+ "index": 0,
801
+ "filename": "shard_00000.bin",
802
+ "size": 67108864,
803
+ "hash": "93d8ad2033e7ec42376dafb41c1e89753d0ca40896835c6f012efe47dbdda6d1",
804
+ "offset": 0
805
+ },
806
+ {
807
+ "index": 1,
808
+ "filename": "shard_00001.bin",
809
+ "size": 67108864,
810
+ "hash": "7896aa344a3da4539cdb76bdbfec3a5f665da5e23d15d15ef9a3b0536ebcf3c5",
811
+ "offset": 67108864
812
+ },
813
+ {
814
+ "index": 2,
815
+ "filename": "shard_00002.bin",
816
+ "size": 67108864,
817
+ "hash": "ff2fc36d9b0008d734f0a82923d47db2383e5a5bae055172630e8ad93ad23ca5",
818
+ "offset": 134217728
819
+ },
820
+ {
821
+ "index": 3,
822
+ "filename": "shard_00003.bin",
823
+ "size": 67108864,
824
+ "hash": "6170b5f1376dbfe33a58de2603987abde826c0eb7318d6c0a1baa78cdf42f77b",
825
+ "offset": 201326592
826
+ },
827
+ {
828
+ "index": 4,
829
+ "filename": "shard_00004.bin",
830
+ "size": 67108864,
831
+ "hash": "5c8767e71c920596247b7ea3d0d55b45d6af4255757758dda6aedd907cfa3e2c",
832
+ "offset": 268435456
833
+ },
834
+ {
835
+ "index": 5,
836
+ "filename": "shard_00005.bin",
837
+ "size": 63812864,
838
+ "hash": "81ff27711b246ec27e3396871547e77f5e4d5d3d029357a66a8f547b6f32c015",
839
+ "offset": 335544320
840
+ }
841
+ ],
842
+ "tensors": {
843
+ "model.embed_tokens.weight": {
844
+ "spans": [
845
+ {
846
+ "shardIndex": 0,
847
+ "offset": 0,
848
+ "size": 67108864
849
+ },
850
+ {
851
+ "shardIndex": 1,
852
+ "offset": 0,
853
+ "size": 67108864
854
+ },
855
+ {
856
+ "shardIndex": 2,
857
+ "offset": 0,
858
+ "size": 67108864
859
+ },
860
+ {
861
+ "shardIndex": 3,
862
+ "offset": 0,
863
+ "size": 67108864
864
+ },
865
+ {
866
+ "shardIndex": 4,
867
+ "offset": 0,
868
+ "size": 67108864
869
+ }
870
+ ],
871
+ "size": 335544320,
872
+ "shape": [
873
+ 262144,
874
+ 640
875
+ ],
876
+ "dtype": "F16",
877
+ "role": "embedding",
878
+ "group": "embed"
879
+ },
880
+ "model.layers.0.input_layernorm.weight": {
881
+ "shard": 5,
882
+ "offset": 0,
883
+ "size": 1280,
884
+ "shape": [
885
+ 640
886
+ ],
887
+ "dtype": "BF16",
888
+ "role": "norm",
889
+ "group": "layer.0"
890
+ },
891
+ "model.layers.0.mlp.down_proj.weight": {
892
+ "shard": 5,
893
+ "offset": 1280,
894
+ "size": 737280,
895
+ "shape": [
896
+ 640,
897
+ 2048
898
+ ],
899
+ "dtype": "Q4_K_M",
900
+ "role": "matmul",
901
+ "group": "layer.0",
902
+ "layout": "row"
903
+ },
904
+ "model.layers.0.mlp.gate_proj.weight": {
905
+ "shard": 5,
906
+ "offset": 738560,
907
+ "size": 884736,
908
+ "shape": [
909
+ 2048,
910
+ 640
911
+ ],
912
+ "dtype": "Q4_K_M",
913
+ "role": "matmul",
914
+ "group": "layer.0",
915
+ "layout": "row"
916
+ },
917
+ "model.layers.0.mlp.up_proj.weight": {
918
+ "shard": 5,
919
+ "offset": 1623296,
920
+ "size": 884736,
921
+ "shape": [
922
+ 2048,
923
+ 640
924
+ ],
925
+ "dtype": "Q4_K_M",
926
+ "role": "matmul",
927
+ "group": "layer.0",
928
+ "layout": "row"
929
+ },
930
+ "model.layers.0.post_attention_layernorm.weight": {
931
+ "shard": 5,
932
+ "offset": 2508032,
933
+ "size": 1280,
934
+ "shape": [
935
+ 640
936
+ ],
937
+ "dtype": "BF16",
938
+ "role": "norm",
939
+ "group": "layer.0"
940
+ },
941
+ "model.layers.0.post_feedforward_layernorm.weight": {
942
+ "shard": 5,
943
+ "offset": 2509312,
944
+ "size": 1280,
945
+ "shape": [
946
+ 640
947
+ ],
948
+ "dtype": "BF16",
949
+ "role": "norm",
950
+ "group": "layer.0"
951
+ },
952
+ "model.layers.0.pre_feedforward_layernorm.weight": {
953
+ "shard": 5,
954
+ "offset": 2510592,
955
+ "size": 1280,
956
+ "shape": [
957
+ 640
958
+ ],
959
+ "dtype": "BF16",
960
+ "role": "norm",
961
+ "group": "layer.0"
962
+ },
963
+ "model.layers.0.self_attn.k_norm.weight": {
964
+ "shard": 5,
965
+ "offset": 2511872,
966
+ "size": 512,
967
+ "shape": [
968
+ 256
969
+ ],
970
+ "dtype": "BF16",
971
+ "role": "norm",
972
+ "group": "layer.0"
973
+ },
974
+ "model.layers.0.self_attn.k_proj.weight": {
975
+ "shard": 5,
976
+ "offset": 2512384,
977
+ "size": 110592,
978
+ "shape": [
979
+ 256,
980
+ 640
981
+ ],
982
+ "dtype": "Q4_K_M",
983
+ "role": "matmul",
984
+ "group": "layer.0",
985
+ "layout": "row"
986
+ },
987
+ "model.layers.0.self_attn.o_proj.weight": {
988
+ "shard": 5,
989
+ "offset": 2622976,
990
+ "size": 368640,
991
+ "shape": [
992
+ 640,
993
+ 1024
994
+ ],
995
+ "dtype": "Q4_K_M",
996
+ "role": "matmul",
997
+ "group": "layer.0",
998
+ "layout": "row"
999
+ },
1000
+ "model.layers.0.self_attn.q_norm.weight": {
1001
+ "shard": 5,
1002
+ "offset": 2991616,
1003
+ "size": 512,
1004
+ "shape": [
1005
+ 256
1006
+ ],
1007
+ "dtype": "BF16",
1008
+ "role": "norm",
1009
+ "group": "layer.0"
1010
+ },
1011
+ "model.layers.0.self_attn.q_proj.weight": {
1012
+ "shard": 5,
1013
+ "offset": 2992128,
1014
+ "size": 442368,
1015
+ "shape": [
1016
+ 1024,
1017
+ 640
1018
+ ],
1019
+ "dtype": "Q4_K_M",
1020
+ "role": "matmul",
1021
+ "group": "layer.0",
1022
+ "layout": "row"
1023
+ },
1024
+ "model.layers.0.self_attn.v_proj.weight": {
1025
+ "shard": 5,
1026
+ "offset": 3434496,
1027
+ "size": 110592,
1028
+ "shape": [
1029
+ 256,
1030
+ 640
1031
+ ],
1032
+ "dtype": "Q4_K_M",
1033
+ "role": "matmul",
1034
+ "group": "layer.0",
1035
+ "layout": "row"
1036
+ },
1037
+ "model.layers.1.input_layernorm.weight": {
1038
+ "shard": 5,
1039
+ "offset": 3545088,
1040
+ "size": 1280,
1041
+ "shape": [
1042
+ 640
1043
+ ],
1044
+ "dtype": "BF16",
1045
+ "role": "norm",
1046
+ "group": "layer.1"
1047
+ },
1048
+ "model.layers.1.mlp.down_proj.weight": {
1049
+ "shard": 5,
1050
+ "offset": 3546368,
1051
+ "size": 737280,
1052
+ "shape": [
1053
+ 640,
1054
+ 2048
1055
+ ],
1056
+ "dtype": "Q4_K_M",
1057
+ "role": "matmul",
1058
+ "group": "layer.1",
1059
+ "layout": "row"
1060
+ },
1061
+ "model.layers.1.mlp.gate_proj.weight": {
1062
+ "shard": 5,
1063
+ "offset": 4283648,
1064
+ "size": 884736,
1065
+ "shape": [
1066
+ 2048,
1067
+ 640
1068
+ ],
1069
+ "dtype": "Q4_K_M",
1070
+ "role": "matmul",
1071
+ "group": "layer.1",
1072
+ "layout": "row"
1073
+ },
1074
+ "model.layers.1.mlp.up_proj.weight": {
1075
+ "shard": 5,
1076
+ "offset": 5168384,
1077
+ "size": 884736,
1078
+ "shape": [
1079
+ 2048,
1080
+ 640
1081
+ ],
1082
+ "dtype": "Q4_K_M",
1083
+ "role": "matmul",
1084
+ "group": "layer.1",
1085
+ "layout": "row"
1086
+ },
1087
+ "model.layers.1.post_attention_layernorm.weight": {
1088
+ "shard": 5,
1089
+ "offset": 6053120,
1090
+ "size": 1280,
1091
+ "shape": [
1092
+ 640
1093
+ ],
1094
+ "dtype": "BF16",
1095
+ "role": "norm",
1096
+ "group": "layer.1"
1097
+ },
1098
+ "model.layers.1.post_feedforward_layernorm.weight": {
1099
+ "shard": 5,
1100
+ "offset": 6054400,
1101
+ "size": 1280,
1102
+ "shape": [
1103
+ 640
1104
+ ],
1105
+ "dtype": "BF16",
1106
+ "role": "norm",
1107
+ "group": "layer.1"
1108
+ },
1109
+ "model.layers.1.pre_feedforward_layernorm.weight": {
1110
+ "shard": 5,
1111
+ "offset": 6055680,
1112
+ "size": 1280,
1113
+ "shape": [
1114
+ 640
1115
+ ],
1116
+ "dtype": "BF16",
1117
+ "role": "norm",
1118
+ "group": "layer.1"
1119
+ },
1120
+ "model.layers.1.self_attn.k_norm.weight": {
1121
+ "shard": 5,
1122
+ "offset": 6056960,
1123
+ "size": 512,
1124
+ "shape": [
1125
+ 256
1126
+ ],
1127
+ "dtype": "BF16",
1128
+ "role": "norm",
1129
+ "group": "layer.1"
1130
+ },
1131
+ "model.layers.1.self_attn.k_proj.weight": {
1132
+ "shard": 5,
1133
+ "offset": 6057472,
1134
+ "size": 110592,
1135
+ "shape": [
1136
+ 256,
1137
+ 640
1138
+ ],
1139
+ "dtype": "Q4_K_M",
1140
+ "role": "matmul",
1141
+ "group": "layer.1",
1142
+ "layout": "row"
1143
+ },
1144
+ "model.layers.1.self_attn.o_proj.weight": {
1145
+ "shard": 5,
1146
+ "offset": 6168064,
1147
+ "size": 368640,
1148
+ "shape": [
1149
+ 640,
1150
+ 1024
1151
+ ],
1152
+ "dtype": "Q4_K_M",
1153
+ "role": "matmul",
1154
+ "group": "layer.1",
1155
+ "layout": "row"
1156
+ },
1157
+ "model.layers.1.self_attn.q_norm.weight": {
1158
+ "shard": 5,
1159
+ "offset": 6536704,
1160
+ "size": 512,
1161
+ "shape": [
1162
+ 256
1163
+ ],
1164
+ "dtype": "BF16",
1165
+ "role": "norm",
1166
+ "group": "layer.1"
1167
+ },
1168
+ "model.layers.1.self_attn.q_proj.weight": {
1169
+ "shard": 5,
1170
+ "offset": 6537216,
1171
+ "size": 442368,
1172
+ "shape": [
1173
+ 1024,
1174
+ 640
1175
+ ],
1176
+ "dtype": "Q4_K_M",
1177
+ "role": "matmul",
1178
+ "group": "layer.1",
1179
+ "layout": "row"
1180
+ },
1181
+ "model.layers.1.self_attn.v_proj.weight": {
1182
+ "shard": 5,
1183
+ "offset": 6979584,
1184
+ "size": 110592,
1185
+ "shape": [
1186
+ 256,
1187
+ 640
1188
+ ],
1189
+ "dtype": "Q4_K_M",
1190
+ "role": "matmul",
1191
+ "group": "layer.1",
1192
+ "layout": "row"
1193
+ },
1194
+ "model.layers.10.input_layernorm.weight": {
1195
+ "shard": 5,
1196
+ "offset": 7090176,
1197
+ "size": 1280,
1198
+ "shape": [
1199
+ 640
1200
+ ],
1201
+ "dtype": "BF16",
1202
+ "role": "norm",
1203
+ "group": "layer.10"
1204
+ },
1205
+ "model.layers.10.mlp.down_proj.weight": {
1206
+ "shard": 5,
1207
+ "offset": 7091456,
1208
+ "size": 737280,
1209
+ "shape": [
1210
+ 640,
1211
+ 2048
1212
+ ],
1213
+ "dtype": "Q4_K_M",
1214
+ "role": "matmul",
1215
+ "group": "layer.10",
1216
+ "layout": "row"
1217
+ },
1218
+ "model.layers.10.mlp.gate_proj.weight": {
1219
+ "shard": 5,
1220
+ "offset": 7828736,
1221
+ "size": 884736,
1222
+ "shape": [
1223
+ 2048,
1224
+ 640
1225
+ ],
1226
+ "dtype": "Q4_K_M",
1227
+ "role": "matmul",
1228
+ "group": "layer.10",
1229
+ "layout": "row"
1230
+ },
1231
+ "model.layers.10.mlp.up_proj.weight": {
1232
+ "shard": 5,
1233
+ "offset": 8713472,
1234
+ "size": 884736,
1235
+ "shape": [
1236
+ 2048,
1237
+ 640
1238
+ ],
1239
+ "dtype": "Q4_K_M",
1240
+ "role": "matmul",
1241
+ "group": "layer.10",
1242
+ "layout": "row"
1243
+ },
1244
+ "model.layers.10.post_attention_layernorm.weight": {
1245
+ "shard": 5,
1246
+ "offset": 9598208,
1247
+ "size": 1280,
1248
+ "shape": [
1249
+ 640
1250
+ ],
1251
+ "dtype": "BF16",
1252
+ "role": "norm",
1253
+ "group": "layer.10"
1254
+ },
1255
+ "model.layers.10.post_feedforward_layernorm.weight": {
1256
+ "shard": 5,
1257
+ "offset": 9599488,
1258
+ "size": 1280,
1259
+ "shape": [
1260
+ 640
1261
+ ],
1262
+ "dtype": "BF16",
1263
+ "role": "norm",
1264
+ "group": "layer.10"
1265
+ },
1266
+ "model.layers.10.pre_feedforward_layernorm.weight": {
1267
+ "shard": 5,
1268
+ "offset": 9600768,
1269
+ "size": 1280,
1270
+ "shape": [
1271
+ 640
1272
+ ],
1273
+ "dtype": "BF16",
1274
+ "role": "norm",
1275
+ "group": "layer.10"
1276
+ },
1277
+ "model.layers.10.self_attn.k_norm.weight": {
1278
+ "shard": 5,
1279
+ "offset": 9602048,
1280
+ "size": 512,
1281
+ "shape": [
1282
+ 256
1283
+ ],
1284
+ "dtype": "BF16",
1285
+ "role": "norm",
1286
+ "group": "layer.10"
1287
+ },
1288
+ "model.layers.10.self_attn.k_proj.weight": {
1289
+ "shard": 5,
1290
+ "offset": 9602560,
1291
+ "size": 110592,
1292
+ "shape": [
1293
+ 256,
1294
+ 640
1295
+ ],
1296
+ "dtype": "Q4_K_M",
1297
+ "role": "matmul",
1298
+ "group": "layer.10",
1299
+ "layout": "row"
1300
+ },
1301
+ "model.layers.10.self_attn.o_proj.weight": {
1302
+ "shard": 5,
1303
+ "offset": 9713152,
1304
+ "size": 368640,
1305
+ "shape": [
1306
+ 640,
1307
+ 1024
1308
+ ],
1309
+ "dtype": "Q4_K_M",
1310
+ "role": "matmul",
1311
+ "group": "layer.10",
1312
+ "layout": "row"
1313
+ },
1314
+ "model.layers.10.self_attn.q_norm.weight": {
1315
+ "shard": 5,
1316
+ "offset": 10081792,
1317
+ "size": 512,
1318
+ "shape": [
1319
+ 256
1320
+ ],
1321
+ "dtype": "BF16",
1322
+ "role": "norm",
1323
+ "group": "layer.10"
1324
+ },
1325
+ "model.layers.10.self_attn.q_proj.weight": {
1326
+ "shard": 5,
1327
+ "offset": 10082304,
1328
+ "size": 442368,
1329
+ "shape": [
1330
+ 1024,
1331
+ 640
1332
+ ],
1333
+ "dtype": "Q4_K_M",
1334
+ "role": "matmul",
1335
+ "group": "layer.10",
1336
+ "layout": "row"
1337
+ },
1338
+ "model.layers.10.self_attn.v_proj.weight": {
1339
+ "shard": 5,
1340
+ "offset": 10524672,
1341
+ "size": 110592,
1342
+ "shape": [
1343
+ 256,
1344
+ 640
1345
+ ],
1346
+ "dtype": "Q4_K_M",
1347
+ "role": "matmul",
1348
+ "group": "layer.10",
1349
+ "layout": "row"
1350
+ },
1351
+ "model.layers.11.input_layernorm.weight": {
1352
+ "shard": 5,
1353
+ "offset": 10635264,
1354
+ "size": 1280,
1355
+ "shape": [
1356
+ 640
1357
+ ],
1358
+ "dtype": "BF16",
1359
+ "role": "norm",
1360
+ "group": "layer.11"
1361
+ },
1362
+ "model.layers.11.mlp.down_proj.weight": {
1363
+ "shard": 5,
1364
+ "offset": 10636544,
1365
+ "size": 737280,
1366
+ "shape": [
1367
+ 640,
1368
+ 2048
1369
+ ],
1370
+ "dtype": "Q4_K_M",
1371
+ "role": "matmul",
1372
+ "group": "layer.11",
1373
+ "layout": "row"
1374
+ },
1375
+ "model.layers.11.mlp.gate_proj.weight": {
1376
+ "shard": 5,
1377
+ "offset": 11373824,
1378
+ "size": 884736,
1379
+ "shape": [
1380
+ 2048,
1381
+ 640
1382
+ ],
1383
+ "dtype": "Q4_K_M",
1384
+ "role": "matmul",
1385
+ "group": "layer.11",
1386
+ "layout": "row"
1387
+ },
1388
+ "model.layers.11.mlp.up_proj.weight": {
1389
+ "shard": 5,
1390
+ "offset": 12258560,
1391
+ "size": 884736,
1392
+ "shape": [
1393
+ 2048,
1394
+ 640
1395
+ ],
1396
+ "dtype": "Q4_K_M",
1397
+ "role": "matmul",
1398
+ "group": "layer.11",
1399
+ "layout": "row"
1400
+ },
1401
+ "model.layers.11.post_attention_layernorm.weight": {
1402
+ "shard": 5,
1403
+ "offset": 13143296,
1404
+ "size": 1280,
1405
+ "shape": [
1406
+ 640
1407
+ ],
1408
+ "dtype": "BF16",
1409
+ "role": "norm",
1410
+ "group": "layer.11"
1411
+ },
1412
+ "model.layers.11.post_feedforward_layernorm.weight": {
1413
+ "shard": 5,
1414
+ "offset": 13144576,
1415
+ "size": 1280,
1416
+ "shape": [
1417
+ 640
1418
+ ],
1419
+ "dtype": "BF16",
1420
+ "role": "norm",
1421
+ "group": "layer.11"
1422
+ },
1423
+ "model.layers.11.pre_feedforward_layernorm.weight": {
1424
+ "shard": 5,
1425
+ "offset": 13145856,
1426
+ "size": 1280,
1427
+ "shape": [
1428
+ 640
1429
+ ],
1430
+ "dtype": "BF16",
1431
+ "role": "norm",
1432
+ "group": "layer.11"
1433
+ },
1434
+ "model.layers.11.self_attn.k_norm.weight": {
1435
+ "shard": 5,
1436
+ "offset": 13147136,
1437
+ "size": 512,
1438
+ "shape": [
1439
+ 256
1440
+ ],
1441
+ "dtype": "BF16",
1442
+ "role": "norm",
1443
+ "group": "layer.11"
1444
+ },
1445
+ "model.layers.11.self_attn.k_proj.weight": {
1446
+ "shard": 5,
1447
+ "offset": 13147648,
1448
+ "size": 110592,
1449
+ "shape": [
1450
+ 256,
1451
+ 640
1452
+ ],
1453
+ "dtype": "Q4_K_M",
1454
+ "role": "matmul",
1455
+ "group": "layer.11",
1456
+ "layout": "row"
1457
+ },
1458
+ "model.layers.11.self_attn.o_proj.weight": {
1459
+ "shard": 5,
1460
+ "offset": 13258240,
1461
+ "size": 368640,
1462
+ "shape": [
1463
+ 640,
1464
+ 1024
1465
+ ],
1466
+ "dtype": "Q4_K_M",
1467
+ "role": "matmul",
1468
+ "group": "layer.11",
1469
+ "layout": "row"
1470
+ },
1471
+ "model.layers.11.self_attn.q_norm.weight": {
1472
+ "shard": 5,
1473
+ "offset": 13626880,
1474
+ "size": 512,
1475
+ "shape": [
1476
+ 256
1477
+ ],
1478
+ "dtype": "BF16",
1479
+ "role": "norm",
1480
+ "group": "layer.11"
1481
+ },
1482
+ "model.layers.11.self_attn.q_proj.weight": {
1483
+ "shard": 5,
1484
+ "offset": 13627392,
1485
+ "size": 442368,
1486
+ "shape": [
1487
+ 1024,
1488
+ 640
1489
+ ],
1490
+ "dtype": "Q4_K_M",
1491
+ "role": "matmul",
1492
+ "group": "layer.11",
1493
+ "layout": "row"
1494
+ },
1495
+ "model.layers.11.self_attn.v_proj.weight": {
1496
+ "shard": 5,
1497
+ "offset": 14069760,
1498
+ "size": 110592,
1499
+ "shape": [
1500
+ 256,
1501
+ 640
1502
+ ],
1503
+ "dtype": "Q4_K_M",
1504
+ "role": "matmul",
1505
+ "group": "layer.11",
1506
+ "layout": "row"
1507
+ },
1508
+ "model.layers.12.input_layernorm.weight": {
1509
+ "shard": 5,
1510
+ "offset": 14180352,
1511
+ "size": 1280,
1512
+ "shape": [
1513
+ 640
1514
+ ],
1515
+ "dtype": "BF16",
1516
+ "role": "norm",
1517
+ "group": "layer.12"
1518
+ },
1519
+ "model.layers.12.mlp.down_proj.weight": {
1520
+ "shard": 5,
1521
+ "offset": 14181632,
1522
+ "size": 737280,
1523
+ "shape": [
1524
+ 640,
1525
+ 2048
1526
+ ],
1527
+ "dtype": "Q4_K_M",
1528
+ "role": "matmul",
1529
+ "group": "layer.12",
1530
+ "layout": "row"
1531
+ },
1532
+ "model.layers.12.mlp.gate_proj.weight": {
1533
+ "shard": 5,
1534
+ "offset": 14918912,
1535
+ "size": 884736,
1536
+ "shape": [
1537
+ 2048,
1538
+ 640
1539
+ ],
1540
+ "dtype": "Q4_K_M",
1541
+ "role": "matmul",
1542
+ "group": "layer.12",
1543
+ "layout": "row"
1544
+ },
1545
+ "model.layers.12.mlp.up_proj.weight": {
1546
+ "shard": 5,
1547
+ "offset": 15803648,
1548
+ "size": 884736,
1549
+ "shape": [
1550
+ 2048,
1551
+ 640
1552
+ ],
1553
+ "dtype": "Q4_K_M",
1554
+ "role": "matmul",
1555
+ "group": "layer.12",
1556
+ "layout": "row"
1557
+ },
1558
+ "model.layers.12.post_attention_layernorm.weight": {
1559
+ "shard": 5,
1560
+ "offset": 16688384,
1561
+ "size": 1280,
1562
+ "shape": [
1563
+ 640
1564
+ ],
1565
+ "dtype": "BF16",
1566
+ "role": "norm",
1567
+ "group": "layer.12"
1568
+ },
1569
+ "model.layers.12.post_feedforward_layernorm.weight": {
1570
+ "shard": 5,
1571
+ "offset": 16689664,
1572
+ "size": 1280,
1573
+ "shape": [
1574
+ 640
1575
+ ],
1576
+ "dtype": "BF16",
1577
+ "role": "norm",
1578
+ "group": "layer.12"
1579
+ },
1580
+ "model.layers.12.pre_feedforward_layernorm.weight": {
1581
+ "shard": 5,
1582
+ "offset": 16690944,
1583
+ "size": 1280,
1584
+ "shape": [
1585
+ 640
1586
+ ],
1587
+ "dtype": "BF16",
1588
+ "role": "norm",
1589
+ "group": "layer.12"
1590
+ },
1591
+ "model.layers.12.self_attn.k_norm.weight": {
1592
+ "shard": 5,
1593
+ "offset": 16692224,
1594
+ "size": 512,
1595
+ "shape": [
1596
+ 256
1597
+ ],
1598
+ "dtype": "BF16",
1599
+ "role": "norm",
1600
+ "group": "layer.12"
1601
+ },
1602
+ "model.layers.12.self_attn.k_proj.weight": {
1603
+ "shard": 5,
1604
+ "offset": 16692736,
1605
+ "size": 110592,
1606
+ "shape": [
1607
+ 256,
1608
+ 640
1609
+ ],
1610
+ "dtype": "Q4_K_M",
1611
+ "role": "matmul",
1612
+ "group": "layer.12",
1613
+ "layout": "row"
1614
+ },
1615
+ "model.layers.12.self_attn.o_proj.weight": {
1616
+ "shard": 5,
1617
+ "offset": 16803328,
1618
+ "size": 368640,
1619
+ "shape": [
1620
+ 640,
1621
+ 1024
1622
+ ],
1623
+ "dtype": "Q4_K_M",
1624
+ "role": "matmul",
1625
+ "group": "layer.12",
1626
+ "layout": "row"
1627
+ },
1628
+ "model.layers.12.self_attn.q_norm.weight": {
1629
+ "shard": 5,
1630
+ "offset": 17171968,
1631
+ "size": 512,
1632
+ "shape": [
1633
+ 256
1634
+ ],
1635
+ "dtype": "BF16",
1636
+ "role": "norm",
1637
+ "group": "layer.12"
1638
+ },
1639
+ "model.layers.12.self_attn.q_proj.weight": {
1640
+ "shard": 5,
1641
+ "offset": 17172480,
1642
+ "size": 442368,
1643
+ "shape": [
1644
+ 1024,
1645
+ 640
1646
+ ],
1647
+ "dtype": "Q4_K_M",
1648
+ "role": "matmul",
1649
+ "group": "layer.12",
1650
+ "layout": "row"
1651
+ },
1652
+ "model.layers.12.self_attn.v_proj.weight": {
1653
+ "shard": 5,
1654
+ "offset": 17614848,
1655
+ "size": 110592,
1656
+ "shape": [
1657
+ 256,
1658
+ 640
1659
+ ],
1660
+ "dtype": "Q4_K_M",
1661
+ "role": "matmul",
1662
+ "group": "layer.12",
1663
+ "layout": "row"
1664
+ },
1665
+ "model.layers.13.input_layernorm.weight": {
1666
+ "shard": 5,
1667
+ "offset": 17725440,
1668
+ "size": 1280,
1669
+ "shape": [
1670
+ 640
1671
+ ],
1672
+ "dtype": "BF16",
1673
+ "role": "norm",
1674
+ "group": "layer.13"
1675
+ },
1676
+ "model.layers.13.mlp.down_proj.weight": {
1677
+ "shard": 5,
1678
+ "offset": 17726720,
1679
+ "size": 737280,
1680
+ "shape": [
1681
+ 640,
1682
+ 2048
1683
+ ],
1684
+ "dtype": "Q4_K_M",
1685
+ "role": "matmul",
1686
+ "group": "layer.13",
1687
+ "layout": "row"
1688
+ },
1689
+ "model.layers.13.mlp.gate_proj.weight": {
1690
+ "shard": 5,
1691
+ "offset": 18464000,
1692
+ "size": 884736,
1693
+ "shape": [
1694
+ 2048,
1695
+ 640
1696
+ ],
1697
+ "dtype": "Q4_K_M",
1698
+ "role": "matmul",
1699
+ "group": "layer.13",
1700
+ "layout": "row"
1701
+ },
1702
+ "model.layers.13.mlp.up_proj.weight": {
1703
+ "shard": 5,
1704
+ "offset": 19348736,
1705
+ "size": 884736,
1706
+ "shape": [
1707
+ 2048,
1708
+ 640
1709
+ ],
1710
+ "dtype": "Q4_K_M",
1711
+ "role": "matmul",
1712
+ "group": "layer.13",
1713
+ "layout": "row"
1714
+ },
1715
+ "model.layers.13.post_attention_layernorm.weight": {
1716
+ "shard": 5,
1717
+ "offset": 20233472,
1718
+ "size": 1280,
1719
+ "shape": [
1720
+ 640
1721
+ ],
1722
+ "dtype": "BF16",
1723
+ "role": "norm",
1724
+ "group": "layer.13"
1725
+ },
1726
+ "model.layers.13.post_feedforward_layernorm.weight": {
1727
+ "shard": 5,
1728
+ "offset": 20234752,
1729
+ "size": 1280,
1730
+ "shape": [
1731
+ 640
1732
+ ],
1733
+ "dtype": "BF16",
1734
+ "role": "norm",
1735
+ "group": "layer.13"
1736
+ },
1737
+ "model.layers.13.pre_feedforward_layernorm.weight": {
1738
+ "shard": 5,
1739
+ "offset": 20236032,
1740
+ "size": 1280,
1741
+ "shape": [
1742
+ 640
1743
+ ],
1744
+ "dtype": "BF16",
1745
+ "role": "norm",
1746
+ "group": "layer.13"
1747
+ },
1748
+ "model.layers.13.self_attn.k_norm.weight": {
1749
+ "shard": 5,
1750
+ "offset": 20237312,
1751
+ "size": 512,
1752
+ "shape": [
1753
+ 256
1754
+ ],
1755
+ "dtype": "BF16",
1756
+ "role": "norm",
1757
+ "group": "layer.13"
1758
+ },
1759
+ "model.layers.13.self_attn.k_proj.weight": {
1760
+ "shard": 5,
1761
+ "offset": 20237824,
1762
+ "size": 110592,
1763
+ "shape": [
1764
+ 256,
1765
+ 640
1766
+ ],
1767
+ "dtype": "Q4_K_M",
1768
+ "role": "matmul",
1769
+ "group": "layer.13",
1770
+ "layout": "row"
1771
+ },
1772
+ "model.layers.13.self_attn.o_proj.weight": {
1773
+ "shard": 5,
1774
+ "offset": 20348416,
1775
+ "size": 368640,
1776
+ "shape": [
1777
+ 640,
1778
+ 1024
1779
+ ],
1780
+ "dtype": "Q4_K_M",
1781
+ "role": "matmul",
1782
+ "group": "layer.13",
1783
+ "layout": "row"
1784
+ },
1785
+ "model.layers.13.self_attn.q_norm.weight": {
1786
+ "shard": 5,
1787
+ "offset": 20717056,
1788
+ "size": 512,
1789
+ "shape": [
1790
+ 256
1791
+ ],
1792
+ "dtype": "BF16",
1793
+ "role": "norm",
1794
+ "group": "layer.13"
1795
+ },
1796
+ "model.layers.13.self_attn.q_proj.weight": {
1797
+ "shard": 5,
1798
+ "offset": 20717568,
1799
+ "size": 442368,
1800
+ "shape": [
1801
+ 1024,
1802
+ 640
1803
+ ],
1804
+ "dtype": "Q4_K_M",
1805
+ "role": "matmul",
1806
+ "group": "layer.13",
1807
+ "layout": "row"
1808
+ },
1809
+ "model.layers.13.self_attn.v_proj.weight": {
1810
+ "shard": 5,
1811
+ "offset": 21159936,
1812
+ "size": 110592,
1813
+ "shape": [
1814
+ 256,
1815
+ 640
1816
+ ],
1817
+ "dtype": "Q4_K_M",
1818
+ "role": "matmul",
1819
+ "group": "layer.13",
1820
+ "layout": "row"
1821
+ },
1822
+ "model.layers.14.input_layernorm.weight": {
1823
+ "shard": 5,
1824
+ "offset": 21270528,
1825
+ "size": 1280,
1826
+ "shape": [
1827
+ 640
1828
+ ],
1829
+ "dtype": "BF16",
1830
+ "role": "norm",
1831
+ "group": "layer.14"
1832
+ },
1833
+ "model.layers.14.mlp.down_proj.weight": {
1834
+ "shard": 5,
1835
+ "offset": 21271808,
1836
+ "size": 737280,
1837
+ "shape": [
1838
+ 640,
1839
+ 2048
1840
+ ],
1841
+ "dtype": "Q4_K_M",
1842
+ "role": "matmul",
1843
+ "group": "layer.14",
1844
+ "layout": "row"
1845
+ },
1846
+ "model.layers.14.mlp.gate_proj.weight": {
1847
+ "shard": 5,
1848
+ "offset": 22009088,
1849
+ "size": 884736,
1850
+ "shape": [
1851
+ 2048,
1852
+ 640
1853
+ ],
1854
+ "dtype": "Q4_K_M",
1855
+ "role": "matmul",
1856
+ "group": "layer.14",
1857
+ "layout": "row"
1858
+ },
1859
+ "model.layers.14.mlp.up_proj.weight": {
1860
+ "shard": 5,
1861
+ "offset": 22893824,
1862
+ "size": 884736,
1863
+ "shape": [
1864
+ 2048,
1865
+ 640
1866
+ ],
1867
+ "dtype": "Q4_K_M",
1868
+ "role": "matmul",
1869
+ "group": "layer.14",
1870
+ "layout": "row"
1871
+ },
1872
+ "model.layers.14.post_attention_layernorm.weight": {
1873
+ "shard": 5,
1874
+ "offset": 23778560,
1875
+ "size": 1280,
1876
+ "shape": [
1877
+ 640
1878
+ ],
1879
+ "dtype": "BF16",
1880
+ "role": "norm",
1881
+ "group": "layer.14"
1882
+ },
1883
+ "model.layers.14.post_feedforward_layernorm.weight": {
1884
+ "shard": 5,
1885
+ "offset": 23779840,
1886
+ "size": 1280,
1887
+ "shape": [
1888
+ 640
1889
+ ],
1890
+ "dtype": "BF16",
1891
+ "role": "norm",
1892
+ "group": "layer.14"
1893
+ },
1894
+ "model.layers.14.pre_feedforward_layernorm.weight": {
1895
+ "shard": 5,
1896
+ "offset": 23781120,
1897
+ "size": 1280,
1898
+ "shape": [
1899
+ 640
1900
+ ],
1901
+ "dtype": "BF16",
1902
+ "role": "norm",
1903
+ "group": "layer.14"
1904
+ },
1905
+ "model.layers.14.self_attn.k_norm.weight": {
1906
+ "shard": 5,
1907
+ "offset": 23782400,
1908
+ "size": 512,
1909
+ "shape": [
1910
+ 256
1911
+ ],
1912
+ "dtype": "BF16",
1913
+ "role": "norm",
1914
+ "group": "layer.14"
1915
+ },
1916
+ "model.layers.14.self_attn.k_proj.weight": {
1917
+ "shard": 5,
1918
+ "offset": 23782912,
1919
+ "size": 110592,
1920
+ "shape": [
1921
+ 256,
1922
+ 640
1923
+ ],
1924
+ "dtype": "Q4_K_M",
1925
+ "role": "matmul",
1926
+ "group": "layer.14",
1927
+ "layout": "row"
1928
+ },
1929
+ "model.layers.14.self_attn.o_proj.weight": {
1930
+ "shard": 5,
1931
+ "offset": 23893504,
1932
+ "size": 368640,
1933
+ "shape": [
1934
+ 640,
1935
+ 1024
1936
+ ],
1937
+ "dtype": "Q4_K_M",
1938
+ "role": "matmul",
1939
+ "group": "layer.14",
1940
+ "layout": "row"
1941
+ },
1942
+ "model.layers.14.self_attn.q_norm.weight": {
1943
+ "shard": 5,
1944
+ "offset": 24262144,
1945
+ "size": 512,
1946
+ "shape": [
1947
+ 256
1948
+ ],
1949
+ "dtype": "BF16",
1950
+ "role": "norm",
1951
+ "group": "layer.14"
1952
+ },
1953
+ "model.layers.14.self_attn.q_proj.weight": {
1954
+ "shard": 5,
1955
+ "offset": 24262656,
1956
+ "size": 442368,
1957
+ "shape": [
1958
+ 1024,
1959
+ 640
1960
+ ],
1961
+ "dtype": "Q4_K_M",
1962
+ "role": "matmul",
1963
+ "group": "layer.14",
1964
+ "layout": "row"
1965
+ },
1966
+ "model.layers.14.self_attn.v_proj.weight": {
1967
+ "shard": 5,
1968
+ "offset": 24705024,
1969
+ "size": 110592,
1970
+ "shape": [
1971
+ 256,
1972
+ 640
1973
+ ],
1974
+ "dtype": "Q4_K_M",
1975
+ "role": "matmul",
1976
+ "group": "layer.14",
1977
+ "layout": "row"
1978
+ },
1979
+ "model.layers.15.input_layernorm.weight": {
1980
+ "shard": 5,
1981
+ "offset": 24815616,
1982
+ "size": 1280,
1983
+ "shape": [
1984
+ 640
1985
+ ],
1986
+ "dtype": "BF16",
1987
+ "role": "norm",
1988
+ "group": "layer.15"
1989
+ },
1990
+ "model.layers.15.mlp.down_proj.weight": {
1991
+ "shard": 5,
1992
+ "offset": 24816896,
1993
+ "size": 737280,
1994
+ "shape": [
1995
+ 640,
1996
+ 2048
1997
+ ],
1998
+ "dtype": "Q4_K_M",
1999
+ "role": "matmul",
2000
+ "group": "layer.15",
2001
+ "layout": "row"
2002
+ },
2003
+ "model.layers.15.mlp.gate_proj.weight": {
2004
+ "shard": 5,
2005
+ "offset": 25554176,
2006
+ "size": 884736,
2007
+ "shape": [
2008
+ 2048,
2009
+ 640
2010
+ ],
2011
+ "dtype": "Q4_K_M",
2012
+ "role": "matmul",
2013
+ "group": "layer.15",
2014
+ "layout": "row"
2015
+ },
2016
+ "model.layers.15.mlp.up_proj.weight": {
2017
+ "shard": 5,
2018
+ "offset": 26438912,
2019
+ "size": 884736,
2020
+ "shape": [
2021
+ 2048,
2022
+ 640
2023
+ ],
2024
+ "dtype": "Q4_K_M",
2025
+ "role": "matmul",
2026
+ "group": "layer.15",
2027
+ "layout": "row"
2028
+ },
2029
+ "model.layers.15.post_attention_layernorm.weight": {
2030
+ "shard": 5,
2031
+ "offset": 27323648,
2032
+ "size": 1280,
2033
+ "shape": [
2034
+ 640
2035
+ ],
2036
+ "dtype": "BF16",
2037
+ "role": "norm",
2038
+ "group": "layer.15"
2039
+ },
2040
+ "model.layers.15.post_feedforward_layernorm.weight": {
2041
+ "shard": 5,
2042
+ "offset": 27324928,
2043
+ "size": 1280,
2044
+ "shape": [
2045
+ 640
2046
+ ],
2047
+ "dtype": "BF16",
2048
+ "role": "norm",
2049
+ "group": "layer.15"
2050
+ },
2051
+ "model.layers.15.pre_feedforward_layernorm.weight": {
2052
+ "shard": 5,
2053
+ "offset": 27326208,
2054
+ "size": 1280,
2055
+ "shape": [
2056
+ 640
2057
+ ],
2058
+ "dtype": "BF16",
2059
+ "role": "norm",
2060
+ "group": "layer.15"
2061
+ },
2062
+ "model.layers.15.self_attn.k_norm.weight": {
2063
+ "shard": 5,
2064
+ "offset": 27327488,
2065
+ "size": 512,
2066
+ "shape": [
2067
+ 256
2068
+ ],
2069
+ "dtype": "BF16",
2070
+ "role": "norm",
2071
+ "group": "layer.15"
2072
+ },
2073
+ "model.layers.15.self_attn.k_proj.weight": {
2074
+ "shard": 5,
2075
+ "offset": 27328000,
2076
+ "size": 110592,
2077
+ "shape": [
2078
+ 256,
2079
+ 640
2080
+ ],
2081
+ "dtype": "Q4_K_M",
2082
+ "role": "matmul",
2083
+ "group": "layer.15",
2084
+ "layout": "row"
2085
+ },
2086
+ "model.layers.15.self_attn.o_proj.weight": {
2087
+ "shard": 5,
2088
+ "offset": 27438592,
2089
+ "size": 368640,
2090
+ "shape": [
2091
+ 640,
2092
+ 1024
2093
+ ],
2094
+ "dtype": "Q4_K_M",
2095
+ "role": "matmul",
2096
+ "group": "layer.15",
2097
+ "layout": "row"
2098
+ },
2099
+ "model.layers.15.self_attn.q_norm.weight": {
2100
+ "shard": 5,
2101
+ "offset": 27807232,
2102
+ "size": 512,
2103
+ "shape": [
2104
+ 256
2105
+ ],
2106
+ "dtype": "BF16",
2107
+ "role": "norm",
2108
+ "group": "layer.15"
2109
+ },
2110
+ "model.layers.15.self_attn.q_proj.weight": {
2111
+ "shard": 5,
2112
+ "offset": 27807744,
2113
+ "size": 442368,
2114
+ "shape": [
2115
+ 1024,
2116
+ 640
2117
+ ],
2118
+ "dtype": "Q4_K_M",
2119
+ "role": "matmul",
2120
+ "group": "layer.15",
2121
+ "layout": "row"
2122
+ },
2123
+ "model.layers.15.self_attn.v_proj.weight": {
2124
+ "shard": 5,
2125
+ "offset": 28250112,
2126
+ "size": 110592,
2127
+ "shape": [
2128
+ 256,
2129
+ 640
2130
+ ],
2131
+ "dtype": "Q4_K_M",
2132
+ "role": "matmul",
2133
+ "group": "layer.15",
2134
+ "layout": "row"
2135
+ },
2136
+ "model.layers.16.input_layernorm.weight": {
2137
+ "shard": 5,
2138
+ "offset": 28360704,
2139
+ "size": 1280,
2140
+ "shape": [
2141
+ 640
2142
+ ],
2143
+ "dtype": "BF16",
2144
+ "role": "norm",
2145
+ "group": "layer.16"
2146
+ },
2147
+ "model.layers.16.mlp.down_proj.weight": {
2148
+ "shard": 5,
2149
+ "offset": 28361984,
2150
+ "size": 737280,
2151
+ "shape": [
2152
+ 640,
2153
+ 2048
2154
+ ],
2155
+ "dtype": "Q4_K_M",
2156
+ "role": "matmul",
2157
+ "group": "layer.16",
2158
+ "layout": "row"
2159
+ },
2160
+ "model.layers.16.mlp.gate_proj.weight": {
2161
+ "shard": 5,
2162
+ "offset": 29099264,
2163
+ "size": 884736,
2164
+ "shape": [
2165
+ 2048,
2166
+ 640
2167
+ ],
2168
+ "dtype": "Q4_K_M",
2169
+ "role": "matmul",
2170
+ "group": "layer.16",
2171
+ "layout": "row"
2172
+ },
2173
+ "model.layers.16.mlp.up_proj.weight": {
2174
+ "shard": 5,
2175
+ "offset": 29984000,
2176
+ "size": 884736,
2177
+ "shape": [
2178
+ 2048,
2179
+ 640
2180
+ ],
2181
+ "dtype": "Q4_K_M",
2182
+ "role": "matmul",
2183
+ "group": "layer.16",
2184
+ "layout": "row"
2185
+ },
2186
+ "model.layers.16.post_attention_layernorm.weight": {
2187
+ "shard": 5,
2188
+ "offset": 30868736,
2189
+ "size": 1280,
2190
+ "shape": [
2191
+ 640
2192
+ ],
2193
+ "dtype": "BF16",
2194
+ "role": "norm",
2195
+ "group": "layer.16"
2196
+ },
2197
+ "model.layers.16.post_feedforward_layernorm.weight": {
2198
+ "shard": 5,
2199
+ "offset": 30870016,
2200
+ "size": 1280,
2201
+ "shape": [
2202
+ 640
2203
+ ],
2204
+ "dtype": "BF16",
2205
+ "role": "norm",
2206
+ "group": "layer.16"
2207
+ },
2208
+ "model.layers.16.pre_feedforward_layernorm.weight": {
2209
+ "shard": 5,
2210
+ "offset": 30871296,
2211
+ "size": 1280,
2212
+ "shape": [
2213
+ 640
2214
+ ],
2215
+ "dtype": "BF16",
2216
+ "role": "norm",
2217
+ "group": "layer.16"
2218
+ },
2219
+ "model.layers.16.self_attn.k_norm.weight": {
2220
+ "shard": 5,
2221
+ "offset": 30872576,
2222
+ "size": 512,
2223
+ "shape": [
2224
+ 256
2225
+ ],
2226
+ "dtype": "BF16",
2227
+ "role": "norm",
2228
+ "group": "layer.16"
2229
+ },
2230
+ "model.layers.16.self_attn.k_proj.weight": {
2231
+ "shard": 5,
2232
+ "offset": 30873088,
2233
+ "size": 110592,
2234
+ "shape": [
2235
+ 256,
2236
+ 640
2237
+ ],
2238
+ "dtype": "Q4_K_M",
2239
+ "role": "matmul",
2240
+ "group": "layer.16",
2241
+ "layout": "row"
2242
+ },
2243
+ "model.layers.16.self_attn.o_proj.weight": {
2244
+ "shard": 5,
2245
+ "offset": 30983680,
2246
+ "size": 368640,
2247
+ "shape": [
2248
+ 640,
2249
+ 1024
2250
+ ],
2251
+ "dtype": "Q4_K_M",
2252
+ "role": "matmul",
2253
+ "group": "layer.16",
2254
+ "layout": "row"
2255
+ },
2256
+ "model.layers.16.self_attn.q_norm.weight": {
2257
+ "shard": 5,
2258
+ "offset": 31352320,
2259
+ "size": 512,
2260
+ "shape": [
2261
+ 256
2262
+ ],
2263
+ "dtype": "BF16",
2264
+ "role": "norm",
2265
+ "group": "layer.16"
2266
+ },
2267
+ "model.layers.16.self_attn.q_proj.weight": {
2268
+ "shard": 5,
2269
+ "offset": 31352832,
2270
+ "size": 442368,
2271
+ "shape": [
2272
+ 1024,
2273
+ 640
2274
+ ],
2275
+ "dtype": "Q4_K_M",
2276
+ "role": "matmul",
2277
+ "group": "layer.16",
2278
+ "layout": "row"
2279
+ },
2280
+ "model.layers.16.self_attn.v_proj.weight": {
2281
+ "shard": 5,
2282
+ "offset": 31795200,
2283
+ "size": 110592,
2284
+ "shape": [
2285
+ 256,
2286
+ 640
2287
+ ],
2288
+ "dtype": "Q4_K_M",
2289
+ "role": "matmul",
2290
+ "group": "layer.16",
2291
+ "layout": "row"
2292
+ },
2293
+ "model.layers.17.input_layernorm.weight": {
2294
+ "shard": 5,
2295
+ "offset": 31905792,
2296
+ "size": 1280,
2297
+ "shape": [
2298
+ 640
2299
+ ],
2300
+ "dtype": "BF16",
2301
+ "role": "norm",
2302
+ "group": "layer.17"
2303
+ },
2304
+ "model.layers.17.mlp.down_proj.weight": {
2305
+ "shard": 5,
2306
+ "offset": 31907072,
2307
+ "size": 737280,
2308
+ "shape": [
2309
+ 640,
2310
+ 2048
2311
+ ],
2312
+ "dtype": "Q4_K_M",
2313
+ "role": "matmul",
2314
+ "group": "layer.17",
2315
+ "layout": "row"
2316
+ },
2317
+ "model.layers.17.mlp.gate_proj.weight": {
2318
+ "shard": 5,
2319
+ "offset": 32644352,
2320
+ "size": 884736,
2321
+ "shape": [
2322
+ 2048,
2323
+ 640
2324
+ ],
2325
+ "dtype": "Q4_K_M",
2326
+ "role": "matmul",
2327
+ "group": "layer.17",
2328
+ "layout": "row"
2329
+ },
2330
+ "model.layers.17.mlp.up_proj.weight": {
2331
+ "shard": 5,
2332
+ "offset": 33529088,
2333
+ "size": 884736,
2334
+ "shape": [
2335
+ 2048,
2336
+ 640
2337
+ ],
2338
+ "dtype": "Q4_K_M",
2339
+ "role": "matmul",
2340
+ "group": "layer.17",
2341
+ "layout": "row"
2342
+ },
2343
+ "model.layers.17.post_attention_layernorm.weight": {
2344
+ "shard": 5,
2345
+ "offset": 34413824,
2346
+ "size": 1280,
2347
+ "shape": [
2348
+ 640
2349
+ ],
2350
+ "dtype": "BF16",
2351
+ "role": "norm",
2352
+ "group": "layer.17"
2353
+ },
2354
+ "model.layers.17.post_feedforward_layernorm.weight": {
2355
+ "shard": 5,
2356
+ "offset": 34415104,
2357
+ "size": 1280,
2358
+ "shape": [
2359
+ 640
2360
+ ],
2361
+ "dtype": "BF16",
2362
+ "role": "norm",
2363
+ "group": "layer.17"
2364
+ },
2365
+ "model.layers.17.pre_feedforward_layernorm.weight": {
2366
+ "shard": 5,
2367
+ "offset": 34416384,
2368
+ "size": 1280,
2369
+ "shape": [
2370
+ 640
2371
+ ],
2372
+ "dtype": "BF16",
2373
+ "role": "norm",
2374
+ "group": "layer.17"
2375
+ },
2376
+ "model.layers.17.self_attn.k_norm.weight": {
2377
+ "shard": 5,
2378
+ "offset": 34417664,
2379
+ "size": 512,
2380
+ "shape": [
2381
+ 256
2382
+ ],
2383
+ "dtype": "BF16",
2384
+ "role": "norm",
2385
+ "group": "layer.17"
2386
+ },
2387
+ "model.layers.17.self_attn.k_proj.weight": {
2388
+ "shard": 5,
2389
+ "offset": 34418176,
2390
+ "size": 110592,
2391
+ "shape": [
2392
+ 256,
2393
+ 640
2394
+ ],
2395
+ "dtype": "Q4_K_M",
2396
+ "role": "matmul",
2397
+ "group": "layer.17",
2398
+ "layout": "row"
2399
+ },
2400
+ "model.layers.17.self_attn.o_proj.weight": {
2401
+ "shard": 5,
2402
+ "offset": 34528768,
2403
+ "size": 368640,
2404
+ "shape": [
2405
+ 640,
2406
+ 1024
2407
+ ],
2408
+ "dtype": "Q4_K_M",
2409
+ "role": "matmul",
2410
+ "group": "layer.17",
2411
+ "layout": "row"
2412
+ },
2413
+ "model.layers.17.self_attn.q_norm.weight": {
2414
+ "shard": 5,
2415
+ "offset": 34897408,
2416
+ "size": 512,
2417
+ "shape": [
2418
+ 256
2419
+ ],
2420
+ "dtype": "BF16",
2421
+ "role": "norm",
2422
+ "group": "layer.17"
2423
+ },
2424
+ "model.layers.17.self_attn.q_proj.weight": {
2425
+ "shard": 5,
2426
+ "offset": 34897920,
2427
+ "size": 442368,
2428
+ "shape": [
2429
+ 1024,
2430
+ 640
2431
+ ],
2432
+ "dtype": "Q4_K_M",
2433
+ "role": "matmul",
2434
+ "group": "layer.17",
2435
+ "layout": "row"
2436
+ },
2437
+ "model.layers.17.self_attn.v_proj.weight": {
2438
+ "shard": 5,
2439
+ "offset": 35340288,
2440
+ "size": 110592,
2441
+ "shape": [
2442
+ 256,
2443
+ 640
2444
+ ],
2445
+ "dtype": "Q4_K_M",
2446
+ "role": "matmul",
2447
+ "group": "layer.17",
2448
+ "layout": "row"
2449
+ },
2450
+ "model.layers.2.input_layernorm.weight": {
2451
+ "shard": 5,
2452
+ "offset": 35450880,
2453
+ "size": 1280,
2454
+ "shape": [
2455
+ 640
2456
+ ],
2457
+ "dtype": "BF16",
2458
+ "role": "norm",
2459
+ "group": "layer.2"
2460
+ },
2461
+ "model.layers.2.mlp.down_proj.weight": {
2462
+ "shard": 5,
2463
+ "offset": 35452160,
2464
+ "size": 737280,
2465
+ "shape": [
2466
+ 640,
2467
+ 2048
2468
+ ],
2469
+ "dtype": "Q4_K_M",
2470
+ "role": "matmul",
2471
+ "group": "layer.2",
2472
+ "layout": "row"
2473
+ },
2474
+ "model.layers.2.mlp.gate_proj.weight": {
2475
+ "shard": 5,
2476
+ "offset": 36189440,
2477
+ "size": 884736,
2478
+ "shape": [
2479
+ 2048,
2480
+ 640
2481
+ ],
2482
+ "dtype": "Q4_K_M",
2483
+ "role": "matmul",
2484
+ "group": "layer.2",
2485
+ "layout": "row"
2486
+ },
2487
+ "model.layers.2.mlp.up_proj.weight": {
2488
+ "shard": 5,
2489
+ "offset": 37074176,
2490
+ "size": 884736,
2491
+ "shape": [
2492
+ 2048,
2493
+ 640
2494
+ ],
2495
+ "dtype": "Q4_K_M",
2496
+ "role": "matmul",
2497
+ "group": "layer.2",
2498
+ "layout": "row"
2499
+ },
2500
+ "model.layers.2.post_attention_layernorm.weight": {
2501
+ "shard": 5,
2502
+ "offset": 37958912,
2503
+ "size": 1280,
2504
+ "shape": [
2505
+ 640
2506
+ ],
2507
+ "dtype": "BF16",
2508
+ "role": "norm",
2509
+ "group": "layer.2"
2510
+ },
2511
+ "model.layers.2.post_feedforward_layernorm.weight": {
2512
+ "shard": 5,
2513
+ "offset": 37960192,
2514
+ "size": 1280,
2515
+ "shape": [
2516
+ 640
2517
+ ],
2518
+ "dtype": "BF16",
2519
+ "role": "norm",
2520
+ "group": "layer.2"
2521
+ },
2522
+ "model.layers.2.pre_feedforward_layernorm.weight": {
2523
+ "shard": 5,
2524
+ "offset": 37961472,
2525
+ "size": 1280,
2526
+ "shape": [
2527
+ 640
2528
+ ],
2529
+ "dtype": "BF16",
2530
+ "role": "norm",
2531
+ "group": "layer.2"
2532
+ },
2533
+ "model.layers.2.self_attn.k_norm.weight": {
2534
+ "shard": 5,
2535
+ "offset": 37962752,
2536
+ "size": 512,
2537
+ "shape": [
2538
+ 256
2539
+ ],
2540
+ "dtype": "BF16",
2541
+ "role": "norm",
2542
+ "group": "layer.2"
2543
+ },
2544
+ "model.layers.2.self_attn.k_proj.weight": {
2545
+ "shard": 5,
2546
+ "offset": 37963264,
2547
+ "size": 110592,
2548
+ "shape": [
2549
+ 256,
2550
+ 640
2551
+ ],
2552
+ "dtype": "Q4_K_M",
2553
+ "role": "matmul",
2554
+ "group": "layer.2",
2555
+ "layout": "row"
2556
+ },
2557
+ "model.layers.2.self_attn.o_proj.weight": {
2558
+ "shard": 5,
2559
+ "offset": 38073856,
2560
+ "size": 368640,
2561
+ "shape": [
2562
+ 640,
2563
+ 1024
2564
+ ],
2565
+ "dtype": "Q4_K_M",
2566
+ "role": "matmul",
2567
+ "group": "layer.2",
2568
+ "layout": "row"
2569
+ },
2570
+ "model.layers.2.self_attn.q_norm.weight": {
2571
+ "shard": 5,
2572
+ "offset": 38442496,
2573
+ "size": 512,
2574
+ "shape": [
2575
+ 256
2576
+ ],
2577
+ "dtype": "BF16",
2578
+ "role": "norm",
2579
+ "group": "layer.2"
2580
+ },
2581
+ "model.layers.2.self_attn.q_proj.weight": {
2582
+ "shard": 5,
2583
+ "offset": 38443008,
2584
+ "size": 442368,
2585
+ "shape": [
2586
+ 1024,
2587
+ 640
2588
+ ],
2589
+ "dtype": "Q4_K_M",
2590
+ "role": "matmul",
2591
+ "group": "layer.2",
2592
+ "layout": "row"
2593
+ },
2594
+ "model.layers.2.self_attn.v_proj.weight": {
2595
+ "shard": 5,
2596
+ "offset": 38885376,
2597
+ "size": 110592,
2598
+ "shape": [
2599
+ 256,
2600
+ 640
2601
+ ],
2602
+ "dtype": "Q4_K_M",
2603
+ "role": "matmul",
2604
+ "group": "layer.2",
2605
+ "layout": "row"
2606
+ },
2607
+ "model.layers.3.input_layernorm.weight": {
2608
+ "shard": 5,
2609
+ "offset": 38995968,
2610
+ "size": 1280,
2611
+ "shape": [
2612
+ 640
2613
+ ],
2614
+ "dtype": "BF16",
2615
+ "role": "norm",
2616
+ "group": "layer.3"
2617
+ },
2618
+ "model.layers.3.mlp.down_proj.weight": {
2619
+ "shard": 5,
2620
+ "offset": 38997248,
2621
+ "size": 737280,
2622
+ "shape": [
2623
+ 640,
2624
+ 2048
2625
+ ],
2626
+ "dtype": "Q4_K_M",
2627
+ "role": "matmul",
2628
+ "group": "layer.3",
2629
+ "layout": "row"
2630
+ },
2631
+ "model.layers.3.mlp.gate_proj.weight": {
2632
+ "shard": 5,
2633
+ "offset": 39734528,
2634
+ "size": 884736,
2635
+ "shape": [
2636
+ 2048,
2637
+ 640
2638
+ ],
2639
+ "dtype": "Q4_K_M",
2640
+ "role": "matmul",
2641
+ "group": "layer.3",
2642
+ "layout": "row"
2643
+ },
2644
+ "model.layers.3.mlp.up_proj.weight": {
2645
+ "shard": 5,
2646
+ "offset": 40619264,
2647
+ "size": 884736,
2648
+ "shape": [
2649
+ 2048,
2650
+ 640
2651
+ ],
2652
+ "dtype": "Q4_K_M",
2653
+ "role": "matmul",
2654
+ "group": "layer.3",
2655
+ "layout": "row"
2656
+ },
2657
+ "model.layers.3.post_attention_layernorm.weight": {
2658
+ "shard": 5,
2659
+ "offset": 41504000,
2660
+ "size": 1280,
2661
+ "shape": [
2662
+ 640
2663
+ ],
2664
+ "dtype": "BF16",
2665
+ "role": "norm",
2666
+ "group": "layer.3"
2667
+ },
2668
+ "model.layers.3.post_feedforward_layernorm.weight": {
2669
+ "shard": 5,
2670
+ "offset": 41505280,
2671
+ "size": 1280,
2672
+ "shape": [
2673
+ 640
2674
+ ],
2675
+ "dtype": "BF16",
2676
+ "role": "norm",
2677
+ "group": "layer.3"
2678
+ },
2679
+ "model.layers.3.pre_feedforward_layernorm.weight": {
2680
+ "shard": 5,
2681
+ "offset": 41506560,
2682
+ "size": 1280,
2683
+ "shape": [
2684
+ 640
2685
+ ],
2686
+ "dtype": "BF16",
2687
+ "role": "norm",
2688
+ "group": "layer.3"
2689
+ },
2690
+ "model.layers.3.self_attn.k_norm.weight": {
2691
+ "shard": 5,
2692
+ "offset": 41507840,
2693
+ "size": 512,
2694
+ "shape": [
2695
+ 256
2696
+ ],
2697
+ "dtype": "BF16",
2698
+ "role": "norm",
2699
+ "group": "layer.3"
2700
+ },
2701
+ "model.layers.3.self_attn.k_proj.weight": {
2702
+ "shard": 5,
2703
+ "offset": 41508352,
2704
+ "size": 110592,
2705
+ "shape": [
2706
+ 256,
2707
+ 640
2708
+ ],
2709
+ "dtype": "Q4_K_M",
2710
+ "role": "matmul",
2711
+ "group": "layer.3",
2712
+ "layout": "row"
2713
+ },
2714
+ "model.layers.3.self_attn.o_proj.weight": {
2715
+ "shard": 5,
2716
+ "offset": 41618944,
2717
+ "size": 368640,
2718
+ "shape": [
2719
+ 640,
2720
+ 1024
2721
+ ],
2722
+ "dtype": "Q4_K_M",
2723
+ "role": "matmul",
2724
+ "group": "layer.3",
2725
+ "layout": "row"
2726
+ },
2727
+ "model.layers.3.self_attn.q_norm.weight": {
2728
+ "shard": 5,
2729
+ "offset": 41987584,
2730
+ "size": 512,
2731
+ "shape": [
2732
+ 256
2733
+ ],
2734
+ "dtype": "BF16",
2735
+ "role": "norm",
2736
+ "group": "layer.3"
2737
+ },
2738
+ "model.layers.3.self_attn.q_proj.weight": {
2739
+ "shard": 5,
2740
+ "offset": 41988096,
2741
+ "size": 442368,
2742
+ "shape": [
2743
+ 1024,
2744
+ 640
2745
+ ],
2746
+ "dtype": "Q4_K_M",
2747
+ "role": "matmul",
2748
+ "group": "layer.3",
2749
+ "layout": "row"
2750
+ },
2751
+ "model.layers.3.self_attn.v_proj.weight": {
2752
+ "shard": 5,
2753
+ "offset": 42430464,
2754
+ "size": 110592,
2755
+ "shape": [
2756
+ 256,
2757
+ 640
2758
+ ],
2759
+ "dtype": "Q4_K_M",
2760
+ "role": "matmul",
2761
+ "group": "layer.3",
2762
+ "layout": "row"
2763
+ },
2764
+ "model.layers.4.input_layernorm.weight": {
2765
+ "shard": 5,
2766
+ "offset": 42541056,
2767
+ "size": 1280,
2768
+ "shape": [
2769
+ 640
2770
+ ],
2771
+ "dtype": "BF16",
2772
+ "role": "norm",
2773
+ "group": "layer.4"
2774
+ },
2775
+ "model.layers.4.mlp.down_proj.weight": {
2776
+ "shard": 5,
2777
+ "offset": 42542336,
2778
+ "size": 737280,
2779
+ "shape": [
2780
+ 640,
2781
+ 2048
2782
+ ],
2783
+ "dtype": "Q4_K_M",
2784
+ "role": "matmul",
2785
+ "group": "layer.4",
2786
+ "layout": "row"
2787
+ },
2788
+ "model.layers.4.mlp.gate_proj.weight": {
2789
+ "shard": 5,
2790
+ "offset": 43279616,
2791
+ "size": 884736,
2792
+ "shape": [
2793
+ 2048,
2794
+ 640
2795
+ ],
2796
+ "dtype": "Q4_K_M",
2797
+ "role": "matmul",
2798
+ "group": "layer.4",
2799
+ "layout": "row"
2800
+ },
2801
+ "model.layers.4.mlp.up_proj.weight": {
2802
+ "shard": 5,
2803
+ "offset": 44164352,
2804
+ "size": 884736,
2805
+ "shape": [
2806
+ 2048,
2807
+ 640
2808
+ ],
2809
+ "dtype": "Q4_K_M",
2810
+ "role": "matmul",
2811
+ "group": "layer.4",
2812
+ "layout": "row"
2813
+ },
2814
+ "model.layers.4.post_attention_layernorm.weight": {
2815
+ "shard": 5,
2816
+ "offset": 45049088,
2817
+ "size": 1280,
2818
+ "shape": [
2819
+ 640
2820
+ ],
2821
+ "dtype": "BF16",
2822
+ "role": "norm",
2823
+ "group": "layer.4"
2824
+ },
2825
+ "model.layers.4.post_feedforward_layernorm.weight": {
2826
+ "shard": 5,
2827
+ "offset": 45050368,
2828
+ "size": 1280,
2829
+ "shape": [
2830
+ 640
2831
+ ],
2832
+ "dtype": "BF16",
2833
+ "role": "norm",
2834
+ "group": "layer.4"
2835
+ },
2836
+ "model.layers.4.pre_feedforward_layernorm.weight": {
2837
+ "shard": 5,
2838
+ "offset": 45051648,
2839
+ "size": 1280,
2840
+ "shape": [
2841
+ 640
2842
+ ],
2843
+ "dtype": "BF16",
2844
+ "role": "norm",
2845
+ "group": "layer.4"
2846
+ },
2847
+ "model.layers.4.self_attn.k_norm.weight": {
2848
+ "shard": 5,
2849
+ "offset": 45052928,
2850
+ "size": 512,
2851
+ "shape": [
2852
+ 256
2853
+ ],
2854
+ "dtype": "BF16",
2855
+ "role": "norm",
2856
+ "group": "layer.4"
2857
+ },
2858
+ "model.layers.4.self_attn.k_proj.weight": {
2859
+ "shard": 5,
2860
+ "offset": 45053440,
2861
+ "size": 110592,
2862
+ "shape": [
2863
+ 256,
2864
+ 640
2865
+ ],
2866
+ "dtype": "Q4_K_M",
2867
+ "role": "matmul",
2868
+ "group": "layer.4",
2869
+ "layout": "row"
2870
+ },
2871
+ "model.layers.4.self_attn.o_proj.weight": {
2872
+ "shard": 5,
2873
+ "offset": 45164032,
2874
+ "size": 368640,
2875
+ "shape": [
2876
+ 640,
2877
+ 1024
2878
+ ],
2879
+ "dtype": "Q4_K_M",
2880
+ "role": "matmul",
2881
+ "group": "layer.4",
2882
+ "layout": "row"
2883
+ },
2884
+ "model.layers.4.self_attn.q_norm.weight": {
2885
+ "shard": 5,
2886
+ "offset": 45532672,
2887
+ "size": 512,
2888
+ "shape": [
2889
+ 256
2890
+ ],
2891
+ "dtype": "BF16",
2892
+ "role": "norm",
2893
+ "group": "layer.4"
2894
+ },
2895
+ "model.layers.4.self_attn.q_proj.weight": {
2896
+ "shard": 5,
2897
+ "offset": 45533184,
2898
+ "size": 442368,
2899
+ "shape": [
2900
+ 1024,
2901
+ 640
2902
+ ],
2903
+ "dtype": "Q4_K_M",
2904
+ "role": "matmul",
2905
+ "group": "layer.4",
2906
+ "layout": "row"
2907
+ },
2908
+ "model.layers.4.self_attn.v_proj.weight": {
2909
+ "shard": 5,
2910
+ "offset": 45975552,
2911
+ "size": 110592,
2912
+ "shape": [
2913
+ 256,
2914
+ 640
2915
+ ],
2916
+ "dtype": "Q4_K_M",
2917
+ "role": "matmul",
2918
+ "group": "layer.4",
2919
+ "layout": "row"
2920
+ },
2921
+ "model.layers.5.input_layernorm.weight": {
2922
+ "shard": 5,
2923
+ "offset": 46086144,
2924
+ "size": 1280,
2925
+ "shape": [
2926
+ 640
2927
+ ],
2928
+ "dtype": "BF16",
2929
+ "role": "norm",
2930
+ "group": "layer.5"
2931
+ },
2932
+ "model.layers.5.mlp.down_proj.weight": {
2933
+ "shard": 5,
2934
+ "offset": 46087424,
2935
+ "size": 737280,
2936
+ "shape": [
2937
+ 640,
2938
+ 2048
2939
+ ],
2940
+ "dtype": "Q4_K_M",
2941
+ "role": "matmul",
2942
+ "group": "layer.5",
2943
+ "layout": "row"
2944
+ },
2945
+ "model.layers.5.mlp.gate_proj.weight": {
2946
+ "shard": 5,
2947
+ "offset": 46824704,
2948
+ "size": 884736,
2949
+ "shape": [
2950
+ 2048,
2951
+ 640
2952
+ ],
2953
+ "dtype": "Q4_K_M",
2954
+ "role": "matmul",
2955
+ "group": "layer.5",
2956
+ "layout": "row"
2957
+ },
2958
+ "model.layers.5.mlp.up_proj.weight": {
2959
+ "shard": 5,
2960
+ "offset": 47709440,
2961
+ "size": 884736,
2962
+ "shape": [
2963
+ 2048,
2964
+ 640
2965
+ ],
2966
+ "dtype": "Q4_K_M",
2967
+ "role": "matmul",
2968
+ "group": "layer.5",
2969
+ "layout": "row"
2970
+ },
2971
+ "model.layers.5.post_attention_layernorm.weight": {
2972
+ "shard": 5,
2973
+ "offset": 48594176,
2974
+ "size": 1280,
2975
+ "shape": [
2976
+ 640
2977
+ ],
2978
+ "dtype": "BF16",
2979
+ "role": "norm",
2980
+ "group": "layer.5"
2981
+ },
2982
+ "model.layers.5.post_feedforward_layernorm.weight": {
2983
+ "shard": 5,
2984
+ "offset": 48595456,
2985
+ "size": 1280,
2986
+ "shape": [
2987
+ 640
2988
+ ],
2989
+ "dtype": "BF16",
2990
+ "role": "norm",
2991
+ "group": "layer.5"
2992
+ },
2993
+ "model.layers.5.pre_feedforward_layernorm.weight": {
2994
+ "shard": 5,
2995
+ "offset": 48596736,
2996
+ "size": 1280,
2997
+ "shape": [
2998
+ 640
2999
+ ],
3000
+ "dtype": "BF16",
3001
+ "role": "norm",
3002
+ "group": "layer.5"
3003
+ },
3004
+ "model.layers.5.self_attn.k_norm.weight": {
3005
+ "shard": 5,
3006
+ "offset": 48598016,
3007
+ "size": 512,
3008
+ "shape": [
3009
+ 256
3010
+ ],
3011
+ "dtype": "BF16",
3012
+ "role": "norm",
3013
+ "group": "layer.5"
3014
+ },
3015
+ "model.layers.5.self_attn.k_proj.weight": {
3016
+ "shard": 5,
3017
+ "offset": 48598528,
3018
+ "size": 110592,
3019
+ "shape": [
3020
+ 256,
3021
+ 640
3022
+ ],
3023
+ "dtype": "Q4_K_M",
3024
+ "role": "matmul",
3025
+ "group": "layer.5",
3026
+ "layout": "row"
3027
+ },
3028
+ "model.layers.5.self_attn.o_proj.weight": {
3029
+ "shard": 5,
3030
+ "offset": 48709120,
3031
+ "size": 368640,
3032
+ "shape": [
3033
+ 640,
3034
+ 1024
3035
+ ],
3036
+ "dtype": "Q4_K_M",
3037
+ "role": "matmul",
3038
+ "group": "layer.5",
3039
+ "layout": "row"
3040
+ },
3041
+ "model.layers.5.self_attn.q_norm.weight": {
3042
+ "shard": 5,
3043
+ "offset": 49077760,
3044
+ "size": 512,
3045
+ "shape": [
3046
+ 256
3047
+ ],
3048
+ "dtype": "BF16",
3049
+ "role": "norm",
3050
+ "group": "layer.5"
3051
+ },
3052
+ "model.layers.5.self_attn.q_proj.weight": {
3053
+ "shard": 5,
3054
+ "offset": 49078272,
3055
+ "size": 442368,
3056
+ "shape": [
3057
+ 1024,
3058
+ 640
3059
+ ],
3060
+ "dtype": "Q4_K_M",
3061
+ "role": "matmul",
3062
+ "group": "layer.5",
3063
+ "layout": "row"
3064
+ },
3065
+ "model.layers.5.self_attn.v_proj.weight": {
3066
+ "shard": 5,
3067
+ "offset": 49520640,
3068
+ "size": 110592,
3069
+ "shape": [
3070
+ 256,
3071
+ 640
3072
+ ],
3073
+ "dtype": "Q4_K_M",
3074
+ "role": "matmul",
3075
+ "group": "layer.5",
3076
+ "layout": "row"
3077
+ },
3078
+ "model.layers.6.input_layernorm.weight": {
3079
+ "shard": 5,
3080
+ "offset": 49631232,
3081
+ "size": 1280,
3082
+ "shape": [
3083
+ 640
3084
+ ],
3085
+ "dtype": "BF16",
3086
+ "role": "norm",
3087
+ "group": "layer.6"
3088
+ },
3089
+ "model.layers.6.mlp.down_proj.weight": {
3090
+ "shard": 5,
3091
+ "offset": 49632512,
3092
+ "size": 737280,
3093
+ "shape": [
3094
+ 640,
3095
+ 2048
3096
+ ],
3097
+ "dtype": "Q4_K_M",
3098
+ "role": "matmul",
3099
+ "group": "layer.6",
3100
+ "layout": "row"
3101
+ },
3102
+ "model.layers.6.mlp.gate_proj.weight": {
3103
+ "shard": 5,
3104
+ "offset": 50369792,
3105
+ "size": 884736,
3106
+ "shape": [
3107
+ 2048,
3108
+ 640
3109
+ ],
3110
+ "dtype": "Q4_K_M",
3111
+ "role": "matmul",
3112
+ "group": "layer.6",
3113
+ "layout": "row"
3114
+ },
3115
+ "model.layers.6.mlp.up_proj.weight": {
3116
+ "shard": 5,
3117
+ "offset": 51254528,
3118
+ "size": 884736,
3119
+ "shape": [
3120
+ 2048,
3121
+ 640
3122
+ ],
3123
+ "dtype": "Q4_K_M",
3124
+ "role": "matmul",
3125
+ "group": "layer.6",
3126
+ "layout": "row"
3127
+ },
3128
+ "model.layers.6.post_attention_layernorm.weight": {
3129
+ "shard": 5,
3130
+ "offset": 52139264,
3131
+ "size": 1280,
3132
+ "shape": [
3133
+ 640
3134
+ ],
3135
+ "dtype": "BF16",
3136
+ "role": "norm",
3137
+ "group": "layer.6"
3138
+ },
3139
+ "model.layers.6.post_feedforward_layernorm.weight": {
3140
+ "shard": 5,
3141
+ "offset": 52140544,
3142
+ "size": 1280,
3143
+ "shape": [
3144
+ 640
3145
+ ],
3146
+ "dtype": "BF16",
3147
+ "role": "norm",
3148
+ "group": "layer.6"
3149
+ },
3150
+ "model.layers.6.pre_feedforward_layernorm.weight": {
3151
+ "shard": 5,
3152
+ "offset": 52141824,
3153
+ "size": 1280,
3154
+ "shape": [
3155
+ 640
3156
+ ],
3157
+ "dtype": "BF16",
3158
+ "role": "norm",
3159
+ "group": "layer.6"
3160
+ },
3161
+ "model.layers.6.self_attn.k_norm.weight": {
3162
+ "shard": 5,
3163
+ "offset": 52143104,
3164
+ "size": 512,
3165
+ "shape": [
3166
+ 256
3167
+ ],
3168
+ "dtype": "BF16",
3169
+ "role": "norm",
3170
+ "group": "layer.6"
3171
+ },
3172
+ "model.layers.6.self_attn.k_proj.weight": {
3173
+ "shard": 5,
3174
+ "offset": 52143616,
3175
+ "size": 110592,
3176
+ "shape": [
3177
+ 256,
3178
+ 640
3179
+ ],
3180
+ "dtype": "Q4_K_M",
3181
+ "role": "matmul",
3182
+ "group": "layer.6",
3183
+ "layout": "row"
3184
+ },
3185
+ "model.layers.6.self_attn.o_proj.weight": {
3186
+ "shard": 5,
3187
+ "offset": 52254208,
3188
+ "size": 368640,
3189
+ "shape": [
3190
+ 640,
3191
+ 1024
3192
+ ],
3193
+ "dtype": "Q4_K_M",
3194
+ "role": "matmul",
3195
+ "group": "layer.6",
3196
+ "layout": "row"
3197
+ },
3198
+ "model.layers.6.self_attn.q_norm.weight": {
3199
+ "shard": 5,
3200
+ "offset": 52622848,
3201
+ "size": 512,
3202
+ "shape": [
3203
+ 256
3204
+ ],
3205
+ "dtype": "BF16",
3206
+ "role": "norm",
3207
+ "group": "layer.6"
3208
+ },
3209
+ "model.layers.6.self_attn.q_proj.weight": {
3210
+ "shard": 5,
3211
+ "offset": 52623360,
3212
+ "size": 442368,
3213
+ "shape": [
3214
+ 1024,
3215
+ 640
3216
+ ],
3217
+ "dtype": "Q4_K_M",
3218
+ "role": "matmul",
3219
+ "group": "layer.6",
3220
+ "layout": "row"
3221
+ },
3222
+ "model.layers.6.self_attn.v_proj.weight": {
3223
+ "shard": 5,
3224
+ "offset": 53065728,
3225
+ "size": 110592,
3226
+ "shape": [
3227
+ 256,
3228
+ 640
3229
+ ],
3230
+ "dtype": "Q4_K_M",
3231
+ "role": "matmul",
3232
+ "group": "layer.6",
3233
+ "layout": "row"
3234
+ },
3235
+ "model.layers.7.input_layernorm.weight": {
3236
+ "shard": 5,
3237
+ "offset": 53176320,
3238
+ "size": 1280,
3239
+ "shape": [
3240
+ 640
3241
+ ],
3242
+ "dtype": "BF16",
3243
+ "role": "norm",
3244
+ "group": "layer.7"
3245
+ },
3246
+ "model.layers.7.mlp.down_proj.weight": {
3247
+ "shard": 5,
3248
+ "offset": 53177600,
3249
+ "size": 737280,
3250
+ "shape": [
3251
+ 640,
3252
+ 2048
3253
+ ],
3254
+ "dtype": "Q4_K_M",
3255
+ "role": "matmul",
3256
+ "group": "layer.7",
3257
+ "layout": "row"
3258
+ },
3259
+ "model.layers.7.mlp.gate_proj.weight": {
3260
+ "shard": 5,
3261
+ "offset": 53914880,
3262
+ "size": 884736,
3263
+ "shape": [
3264
+ 2048,
3265
+ 640
3266
+ ],
3267
+ "dtype": "Q4_K_M",
3268
+ "role": "matmul",
3269
+ "group": "layer.7",
3270
+ "layout": "row"
3271
+ },
3272
+ "model.layers.7.mlp.up_proj.weight": {
3273
+ "shard": 5,
3274
+ "offset": 54799616,
3275
+ "size": 884736,
3276
+ "shape": [
3277
+ 2048,
3278
+ 640
3279
+ ],
3280
+ "dtype": "Q4_K_M",
3281
+ "role": "matmul",
3282
+ "group": "layer.7",
3283
+ "layout": "row"
3284
+ },
3285
+ "model.layers.7.post_attention_layernorm.weight": {
3286
+ "shard": 5,
3287
+ "offset": 55684352,
3288
+ "size": 1280,
3289
+ "shape": [
3290
+ 640
3291
+ ],
3292
+ "dtype": "BF16",
3293
+ "role": "norm",
3294
+ "group": "layer.7"
3295
+ },
3296
+ "model.layers.7.post_feedforward_layernorm.weight": {
3297
+ "shard": 5,
3298
+ "offset": 55685632,
3299
+ "size": 1280,
3300
+ "shape": [
3301
+ 640
3302
+ ],
3303
+ "dtype": "BF16",
3304
+ "role": "norm",
3305
+ "group": "layer.7"
3306
+ },
3307
+ "model.layers.7.pre_feedforward_layernorm.weight": {
3308
+ "shard": 5,
3309
+ "offset": 55686912,
3310
+ "size": 1280,
3311
+ "shape": [
3312
+ 640
3313
+ ],
3314
+ "dtype": "BF16",
3315
+ "role": "norm",
3316
+ "group": "layer.7"
3317
+ },
3318
+ "model.layers.7.self_attn.k_norm.weight": {
3319
+ "shard": 5,
3320
+ "offset": 55688192,
3321
+ "size": 512,
3322
+ "shape": [
3323
+ 256
3324
+ ],
3325
+ "dtype": "BF16",
3326
+ "role": "norm",
3327
+ "group": "layer.7"
3328
+ },
3329
+ "model.layers.7.self_attn.k_proj.weight": {
3330
+ "shard": 5,
3331
+ "offset": 55688704,
3332
+ "size": 110592,
3333
+ "shape": [
3334
+ 256,
3335
+ 640
3336
+ ],
3337
+ "dtype": "Q4_K_M",
3338
+ "role": "matmul",
3339
+ "group": "layer.7",
3340
+ "layout": "row"
3341
+ },
3342
+ "model.layers.7.self_attn.o_proj.weight": {
3343
+ "shard": 5,
3344
+ "offset": 55799296,
3345
+ "size": 368640,
3346
+ "shape": [
3347
+ 640,
3348
+ 1024
3349
+ ],
3350
+ "dtype": "Q4_K_M",
3351
+ "role": "matmul",
3352
+ "group": "layer.7",
3353
+ "layout": "row"
3354
+ },
3355
+ "model.layers.7.self_attn.q_norm.weight": {
3356
+ "shard": 5,
3357
+ "offset": 56167936,
3358
+ "size": 512,
3359
+ "shape": [
3360
+ 256
3361
+ ],
3362
+ "dtype": "BF16",
3363
+ "role": "norm",
3364
+ "group": "layer.7"
3365
+ },
3366
+ "model.layers.7.self_attn.q_proj.weight": {
3367
+ "shard": 5,
3368
+ "offset": 56168448,
3369
+ "size": 442368,
3370
+ "shape": [
3371
+ 1024,
3372
+ 640
3373
+ ],
3374
+ "dtype": "Q4_K_M",
3375
+ "role": "matmul",
3376
+ "group": "layer.7",
3377
+ "layout": "row"
3378
+ },
3379
+ "model.layers.7.self_attn.v_proj.weight": {
3380
+ "shard": 5,
3381
+ "offset": 56610816,
3382
+ "size": 110592,
3383
+ "shape": [
3384
+ 256,
3385
+ 640
3386
+ ],
3387
+ "dtype": "Q4_K_M",
3388
+ "role": "matmul",
3389
+ "group": "layer.7",
3390
+ "layout": "row"
3391
+ },
3392
+ "model.layers.8.input_layernorm.weight": {
3393
+ "shard": 5,
3394
+ "offset": 56721408,
3395
+ "size": 1280,
3396
+ "shape": [
3397
+ 640
3398
+ ],
3399
+ "dtype": "BF16",
3400
+ "role": "norm",
3401
+ "group": "layer.8"
3402
+ },
3403
+ "model.layers.8.mlp.down_proj.weight": {
3404
+ "shard": 5,
3405
+ "offset": 56722688,
3406
+ "size": 737280,
3407
+ "shape": [
3408
+ 640,
3409
+ 2048
3410
+ ],
3411
+ "dtype": "Q4_K_M",
3412
+ "role": "matmul",
3413
+ "group": "layer.8",
3414
+ "layout": "row"
3415
+ },
3416
+ "model.layers.8.mlp.gate_proj.weight": {
3417
+ "shard": 5,
3418
+ "offset": 57459968,
3419
+ "size": 884736,
3420
+ "shape": [
3421
+ 2048,
3422
+ 640
3423
+ ],
3424
+ "dtype": "Q4_K_M",
3425
+ "role": "matmul",
3426
+ "group": "layer.8",
3427
+ "layout": "row"
3428
+ },
3429
+ "model.layers.8.mlp.up_proj.weight": {
3430
+ "shard": 5,
3431
+ "offset": 58344704,
3432
+ "size": 884736,
3433
+ "shape": [
3434
+ 2048,
3435
+ 640
3436
+ ],
3437
+ "dtype": "Q4_K_M",
3438
+ "role": "matmul",
3439
+ "group": "layer.8",
3440
+ "layout": "row"
3441
+ },
3442
+ "model.layers.8.post_attention_layernorm.weight": {
3443
+ "shard": 5,
3444
+ "offset": 59229440,
3445
+ "size": 1280,
3446
+ "shape": [
3447
+ 640
3448
+ ],
3449
+ "dtype": "BF16",
3450
+ "role": "norm",
3451
+ "group": "layer.8"
3452
+ },
3453
+ "model.layers.8.post_feedforward_layernorm.weight": {
3454
+ "shard": 5,
3455
+ "offset": 59230720,
3456
+ "size": 1280,
3457
+ "shape": [
3458
+ 640
3459
+ ],
3460
+ "dtype": "BF16",
3461
+ "role": "norm",
3462
+ "group": "layer.8"
3463
+ },
3464
+ "model.layers.8.pre_feedforward_layernorm.weight": {
3465
+ "shard": 5,
3466
+ "offset": 59232000,
3467
+ "size": 1280,
3468
+ "shape": [
3469
+ 640
3470
+ ],
3471
+ "dtype": "BF16",
3472
+ "role": "norm",
3473
+ "group": "layer.8"
3474
+ },
3475
+ "model.layers.8.self_attn.k_norm.weight": {
3476
+ "shard": 5,
3477
+ "offset": 59233280,
3478
+ "size": 512,
3479
+ "shape": [
3480
+ 256
3481
+ ],
3482
+ "dtype": "BF16",
3483
+ "role": "norm",
3484
+ "group": "layer.8"
3485
+ },
3486
+ "model.layers.8.self_attn.k_proj.weight": {
3487
+ "shard": 5,
3488
+ "offset": 59233792,
3489
+ "size": 110592,
3490
+ "shape": [
3491
+ 256,
3492
+ 640
3493
+ ],
3494
+ "dtype": "Q4_K_M",
3495
+ "role": "matmul",
3496
+ "group": "layer.8",
3497
+ "layout": "row"
3498
+ },
3499
+ "model.layers.8.self_attn.o_proj.weight": {
3500
+ "shard": 5,
3501
+ "offset": 59344384,
3502
+ "size": 368640,
3503
+ "shape": [
3504
+ 640,
3505
+ 1024
3506
+ ],
3507
+ "dtype": "Q4_K_M",
3508
+ "role": "matmul",
3509
+ "group": "layer.8",
3510
+ "layout": "row"
3511
+ },
3512
+ "model.layers.8.self_attn.q_norm.weight": {
3513
+ "shard": 5,
3514
+ "offset": 59713024,
3515
+ "size": 512,
3516
+ "shape": [
3517
+ 256
3518
+ ],
3519
+ "dtype": "BF16",
3520
+ "role": "norm",
3521
+ "group": "layer.8"
3522
+ },
3523
+ "model.layers.8.self_attn.q_proj.weight": {
3524
+ "shard": 5,
3525
+ "offset": 59713536,
3526
+ "size": 442368,
3527
+ "shape": [
3528
+ 1024,
3529
+ 640
3530
+ ],
3531
+ "dtype": "Q4_K_M",
3532
+ "role": "matmul",
3533
+ "group": "layer.8",
3534
+ "layout": "row"
3535
+ },
3536
+ "model.layers.8.self_attn.v_proj.weight": {
3537
+ "shard": 5,
3538
+ "offset": 60155904,
3539
+ "size": 110592,
3540
+ "shape": [
3541
+ 256,
3542
+ 640
3543
+ ],
3544
+ "dtype": "Q4_K_M",
3545
+ "role": "matmul",
3546
+ "group": "layer.8",
3547
+ "layout": "row"
3548
+ },
3549
+ "model.layers.9.input_layernorm.weight": {
3550
+ "shard": 5,
3551
+ "offset": 60266496,
3552
+ "size": 1280,
3553
+ "shape": [
3554
+ 640
3555
+ ],
3556
+ "dtype": "BF16",
3557
+ "role": "norm",
3558
+ "group": "layer.9"
3559
+ },
3560
+ "model.layers.9.mlp.down_proj.weight": {
3561
+ "shard": 5,
3562
+ "offset": 60267776,
3563
+ "size": 737280,
3564
+ "shape": [
3565
+ 640,
3566
+ 2048
3567
+ ],
3568
+ "dtype": "Q4_K_M",
3569
+ "role": "matmul",
3570
+ "group": "layer.9",
3571
+ "layout": "row"
3572
+ },
3573
+ "model.layers.9.mlp.gate_proj.weight": {
3574
+ "shard": 5,
3575
+ "offset": 61005056,
3576
+ "size": 884736,
3577
+ "shape": [
3578
+ 2048,
3579
+ 640
3580
+ ],
3581
+ "dtype": "Q4_K_M",
3582
+ "role": "matmul",
3583
+ "group": "layer.9",
3584
+ "layout": "row"
3585
+ },
3586
+ "model.layers.9.mlp.up_proj.weight": {
3587
+ "shard": 5,
3588
+ "offset": 61889792,
3589
+ "size": 884736,
3590
+ "shape": [
3591
+ 2048,
3592
+ 640
3593
+ ],
3594
+ "dtype": "Q4_K_M",
3595
+ "role": "matmul",
3596
+ "group": "layer.9",
3597
+ "layout": "row"
3598
+ },
3599
+ "model.layers.9.post_attention_layernorm.weight": {
3600
+ "shard": 5,
3601
+ "offset": 62774528,
3602
+ "size": 1280,
3603
+ "shape": [
3604
+ 640
3605
+ ],
3606
+ "dtype": "BF16",
3607
+ "role": "norm",
3608
+ "group": "layer.9"
3609
+ },
3610
+ "model.layers.9.post_feedforward_layernorm.weight": {
3611
+ "shard": 5,
3612
+ "offset": 62775808,
3613
+ "size": 1280,
3614
+ "shape": [
3615
+ 640
3616
+ ],
3617
+ "dtype": "BF16",
3618
+ "role": "norm",
3619
+ "group": "layer.9"
3620
+ },
3621
+ "model.layers.9.pre_feedforward_layernorm.weight": {
3622
+ "shard": 5,
3623
+ "offset": 62777088,
3624
+ "size": 1280,
3625
+ "shape": [
3626
+ 640
3627
+ ],
3628
+ "dtype": "BF16",
3629
+ "role": "norm",
3630
+ "group": "layer.9"
3631
+ },
3632
+ "model.layers.9.self_attn.k_norm.weight": {
3633
+ "shard": 5,
3634
+ "offset": 62778368,
3635
+ "size": 512,
3636
+ "shape": [
3637
+ 256
3638
+ ],
3639
+ "dtype": "BF16",
3640
+ "role": "norm",
3641
+ "group": "layer.9"
3642
+ },
3643
+ "model.layers.9.self_attn.k_proj.weight": {
3644
+ "shard": 5,
3645
+ "offset": 62778880,
3646
+ "size": 110592,
3647
+ "shape": [
3648
+ 256,
3649
+ 640
3650
+ ],
3651
+ "dtype": "Q4_K_M",
3652
+ "role": "matmul",
3653
+ "group": "layer.9",
3654
+ "layout": "row"
3655
+ },
3656
+ "model.layers.9.self_attn.o_proj.weight": {
3657
+ "shard": 5,
3658
+ "offset": 62889472,
3659
+ "size": 368640,
3660
+ "shape": [
3661
+ 640,
3662
+ 1024
3663
+ ],
3664
+ "dtype": "Q4_K_M",
3665
+ "role": "matmul",
3666
+ "group": "layer.9",
3667
+ "layout": "row"
3668
+ },
3669
+ "model.layers.9.self_attn.q_norm.weight": {
3670
+ "shard": 5,
3671
+ "offset": 63258112,
3672
+ "size": 512,
3673
+ "shape": [
3674
+ 256
3675
+ ],
3676
+ "dtype": "BF16",
3677
+ "role": "norm",
3678
+ "group": "layer.9"
3679
+ },
3680
+ "model.layers.9.self_attn.q_proj.weight": {
3681
+ "shard": 5,
3682
+ "offset": 63258624,
3683
+ "size": 442368,
3684
+ "shape": [
3685
+ 1024,
3686
+ 640
3687
+ ],
3688
+ "dtype": "Q4_K_M",
3689
+ "role": "matmul",
3690
+ "group": "layer.9",
3691
+ "layout": "row"
3692
+ },
3693
+ "model.layers.9.self_attn.v_proj.weight": {
3694
+ "shard": 5,
3695
+ "offset": 63700992,
3696
+ "size": 110592,
3697
+ "shape": [
3698
+ 256,
3699
+ 640
3700
+ ],
3701
+ "dtype": "Q4_K_M",
3702
+ "role": "matmul",
3703
+ "group": "layer.9",
3704
+ "layout": "row"
3705
+ },
3706
+ "model.norm.weight": {
3707
+ "shard": 5,
3708
+ "offset": 63811584,
3709
+ "size": 1280,
3710
+ "shape": [
3711
+ 640
3712
+ ],
3713
+ "dtype": "BF16",
3714
+ "role": "norm",
3715
+ "group": "head"
3716
+ }
3717
+ },
3718
+ "totalSize": 399357184,
3719
+ "hashAlgorithm": "sha256",
3720
+ "eos_token_id": 1,
3721
+ "metadata": {
3722
+ "source": "convert-core",
3723
+ "convertedAt": "2026-03-07T01:40:23.980Z",
3724
+ "hasTokenizer": true
3725
+ },
3726
+ "tokenizer": {
3727
+ "type": "bundled",
3728
+ "vocabSize": 3119,
3729
+ "file": "tokenizer.json"
3730
+ }
3731
+ }
models/gemma-3-270m-it-wq4k-ef16-hf16/origin.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "sourceModel": "gemma-3-270m-it",
3
+ "sourceRepo": "google/gemma-3-270m-it",
4
+ "sourceFormat": "safetensors",
5
+ "sourceRevision": "ac82b4e820549b854eebf28ce6dedaf9fdfa17b3",
6
+ "variant": "wq4k-ef16-hf16",
7
+ "convertedAt": "2026-03-07T01:40:23.980Z"
8
+ }
models/gemma-3-270m-it-wq4k-ef16-hf16/shard_00000.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93d8ad2033e7ec42376dafb41c1e89753d0ca40896835c6f012efe47dbdda6d1
3
+ size 67108864
models/gemma-3-270m-it-wq4k-ef16-hf16/shard_00001.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7896aa344a3da4539cdb76bdbfec3a5f665da5e23d15d15ef9a3b0536ebcf3c5
3
+ size 67108864
models/gemma-3-270m-it-wq4k-ef16-hf16/shard_00002.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff2fc36d9b0008d734f0a82923d47db2383e5a5bae055172630e8ad93ad23ca5
3
+ size 67108864
models/gemma-3-270m-it-wq4k-ef16-hf16/shard_00003.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6170b5f1376dbfe33a58de2603987abde826c0eb7318d6c0a1baa78cdf42f77b
3
+ size 67108864
models/gemma-3-270m-it-wq4k-ef16-hf16/shard_00004.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c8767e71c920596247b7ea3d0d55b45d6af4255757758dda6aedd907cfa3e2c
3
+ size 67108864
models/gemma-3-270m-it-wq4k-ef16-hf16/shard_00005.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:81ff27711b246ec27e3396871547e77f5e4d5d3d029357a66a8f547b6f32c015
3
+ size 63812864
models/gemma-3-270m-it-wq4k-ef16-hf16/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be6955513a2509d69d18e82180c2a477ad99358d15212fb81d957566d8186075
3
+ size 14386247
models/gemma-3-270m-it-wq4k-ef16-hf16/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
3
+ size 4689074