Hugging Face's logo

amd
/
GLM-5-NVFP4

Safetensors
glm_moe_dsa
8-bit precision
quark
Model card Files
xet
 Community
6

mtp-glm5 support 0812

#6
by huidongji - opened
base: refs/heads/main
←
from: refs/pr/6
Discussion Files changed
+987
-6
Files changed (13) hide show
  1. config.json +954 -3
  2. generation_config.json +1 -1
  3. model-00001-of-00010.safetensors +3 -0
  4. model-00002-of-00010.safetensors +3 -0
  5. model-00003-of-00010.safetensors +3 -0
  6. model-00004-of-00010.safetensors +3 -0
  7. model-00005-of-00010.safetensors +3 -0
  8. model-00006-of-00010.safetensors +3 -0
  9. model-00007-of-00010.safetensors +3 -0
  10. model-00008-of-00010.safetensors +3 -0
  11. model-00009-of-00010.safetensors +3 -0
  12. model-00010-of-00010.safetensors +3 -0
  13. model.safetensors.index.json +2 -2
config.json CHANGED
@@ -20,10 +20,171 @@
20
  "index_n_heads": 32,
21
  "index_topk": 2048,
22
  "indexer_rope_interleave": true,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
23
  "initializer_range": 0.02,
24
  "intermediate_size": 12288,
25
  "kv_lora_rank": 512,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
26
  "max_position_embeddings": 202752,
 
27
  "mlp_layer_types": [
28
  "dense",
29
  "dense",
@@ -758,7 +919,797 @@
758
  "model.layers.77.self_attn.indexer.wq_b",
759
  "model.layers.77.self_attn.indexer.wk",
760
  "model.layers.77.self_attn.indexer.weights_proj",
761
- "lm_head"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
762
  ],
763
  "export": {
764
  "kv_cache_group": [],
@@ -855,7 +1806,7 @@
855
  "quant_method": "quark",
856
  "quant_mode": "eager_mode",
857
  "softmax_quant_spec": null,
858
- "version": "0.13+488baccced9"
859
  },
860
  "rms_norm_eps": 1e-05,
861
  "rope_interleave": true,
@@ -868,7 +1819,7 @@
868
  "tie_word_embeddings": false,
869
  "topk_group": 1,
870
  "topk_method": "noaux_tc",
871
- "transformers_version": "5.2.0",
872
  "use_cache": true,
873
  "v_head_dim": 256,
874
  "vocab_size": 154880
 
20
  "index_n_heads": 32,
21
  "index_topk": 2048,
22
  "indexer_rope_interleave": true,
23
+ "indexer_types": [
24
+ "full",
25
+ "full",
26
+ "full",
27
+ "full",
28
+ "full",
29
+ "full",
30
+ "full",
31
+ "full",
32
+ "full",
33
+ "full",
34
+ "full",
35
+ "full",
36
+ "full",
37
+ "full",
38
+ "full",
39
+ "full",
40
+ "full",
41
+ "full",
42
+ "full",
43
+ "full",
44
+ "full",
45
+ "full",
46
+ "full",
47
+ "full",
48
+ "full",
49
+ "full",
50
+ "full",
51
+ "full",
52
+ "full",
53
+ "full",
54
+ "full",
55
+ "full",
56
+ "full",
57
+ "full",
58
+ "full",
59
+ "full",
60
+ "full",
61
+ "full",
62
+ "full",
63
+ "full",
64
+ "full",
65
+ "full",
66
+ "full",
67
+ "full",
68
+ "full",
69
+ "full",
70
+ "full",
71
+ "full",
72
+ "full",
73
+ "full",
74
+ "full",
75
+ "full",
76
+ "full",
77
+ "full",
78
+ "full",
79
+ "full",
80
+ "full",
81
+ "full",
82
+ "full",
83
+ "full",
84
+ "full",
85
+ "full",
86
+ "full",
87
+ "full",
88
+ "full",
89
+ "full",
90
+ "full",
91
+ "full",
92
+ "full",
93
+ "full",
94
+ "full",
95
+ "full",
96
+ "full",
97
+ "full",
98
+ "full",
99
+ "full",
100
+ "full",
101
+ "full"
102
+ ],
103
  "initializer_range": 0.02,
104
  "intermediate_size": 12288,
105
  "kv_lora_rank": 512,
106
+ "layer_types": [
107
+ "deepseek_sparse_attention",
108
+ "deepseek_sparse_attention",
109
+ "deepseek_sparse_attention",
110
+ "deepseek_sparse_attention",
111
+ "deepseek_sparse_attention",
112
+ "deepseek_sparse_attention",
113
+ "deepseek_sparse_attention",
114
+ "deepseek_sparse_attention",
115
+ "deepseek_sparse_attention",
116
+ "deepseek_sparse_attention",
117
+ "deepseek_sparse_attention",
118
+ "deepseek_sparse_attention",
119
+ "deepseek_sparse_attention",
120
+ "deepseek_sparse_attention",
121
+ "deepseek_sparse_attention",
122
+ "deepseek_sparse_attention",
123
+ "deepseek_sparse_attention",
124
+ "deepseek_sparse_attention",
125
+ "deepseek_sparse_attention",
126
+ "deepseek_sparse_attention",
127
+ "deepseek_sparse_attention",
128
+ "deepseek_sparse_attention",
129
+ "deepseek_sparse_attention",
130
+ "deepseek_sparse_attention",
131
+ "deepseek_sparse_attention",
132
+ "deepseek_sparse_attention",
133
+ "deepseek_sparse_attention",
134
+ "deepseek_sparse_attention",
135
+ "deepseek_sparse_attention",
136
+ "deepseek_sparse_attention",
137
+ "deepseek_sparse_attention",
138
+ "deepseek_sparse_attention",
139
+ "deepseek_sparse_attention",
140
+ "deepseek_sparse_attention",
141
+ "deepseek_sparse_attention",
142
+ "deepseek_sparse_attention",
143
+ "deepseek_sparse_attention",
144
+ "deepseek_sparse_attention",
145
+ "deepseek_sparse_attention",
146
+ "deepseek_sparse_attention",
147
+ "deepseek_sparse_attention",
148
+ "deepseek_sparse_attention",
149
+ "deepseek_sparse_attention",
150
+ "deepseek_sparse_attention",
151
+ "deepseek_sparse_attention",
152
+ "deepseek_sparse_attention",
153
+ "deepseek_sparse_attention",
154
+ "deepseek_sparse_attention",
155
+ "deepseek_sparse_attention",
156
+ "deepseek_sparse_attention",
157
+ "deepseek_sparse_attention",
158
+ "deepseek_sparse_attention",
159
+ "deepseek_sparse_attention",
160
+ "deepseek_sparse_attention",
161
+ "deepseek_sparse_attention",
162
+ "deepseek_sparse_attention",
163
+ "deepseek_sparse_attention",
164
+ "deepseek_sparse_attention",
165
+ "deepseek_sparse_attention",
166
+ "deepseek_sparse_attention",
167
+ "deepseek_sparse_attention",
168
+ "deepseek_sparse_attention",
169
+ "deepseek_sparse_attention",
170
+ "deepseek_sparse_attention",
171
+ "deepseek_sparse_attention",
172
+ "deepseek_sparse_attention",
173
+ "deepseek_sparse_attention",
174
+ "deepseek_sparse_attention",
175
+ "deepseek_sparse_attention",
176
+ "deepseek_sparse_attention",
177
+ "deepseek_sparse_attention",
178
+ "deepseek_sparse_attention",
179
+ "deepseek_sparse_attention",
180
+ "deepseek_sparse_attention",
181
+ "deepseek_sparse_attention",
182
+ "deepseek_sparse_attention",
183
+ "deepseek_sparse_attention",
184
+ "deepseek_sparse_attention"
185
+ ],
186
  "max_position_embeddings": 202752,
187
+ "mlp_bias": false,
188
  "mlp_layer_types": [
189
  "dense",
190
  "dense",
 
919
  "model.layers.77.self_attn.indexer.wq_b",
920
  "model.layers.77.self_attn.indexer.wk",
921
  "model.layers.77.self_attn.indexer.weights_proj",
922
+ "lm_head",
923
+ "model.layers.78.eh_proj",
924
+ "model.layers.78.enorm",
925
+ "model.layers.78.hnorm",
926
+ "model.layers.78.input_layernorm",
927
+ "model.layers.78.mlp.experts.0.down_proj",
928
+ "model.layers.78.mlp.experts.0.gate_proj",
929
+ "model.layers.78.mlp.experts.0.up_proj",
930
+ "model.layers.78.mlp.experts.1.down_proj",
931
+ "model.layers.78.mlp.experts.1.gate_proj",
932
+ "model.layers.78.mlp.experts.1.up_proj",
933
+ "model.layers.78.mlp.experts.10.down_proj",
934
+ "model.layers.78.mlp.experts.10.gate_proj",
935
+ "model.layers.78.mlp.experts.10.up_proj",
936
+ "model.layers.78.mlp.experts.100.down_proj",
937
+ "model.layers.78.mlp.experts.100.gate_proj",
938
+ "model.layers.78.mlp.experts.100.up_proj",
939
+ "model.layers.78.mlp.experts.101.down_proj",
940
+ "model.layers.78.mlp.experts.101.gate_proj",
941
+ "model.layers.78.mlp.experts.101.up_proj",
942
+ "model.layers.78.mlp.experts.102.down_proj",
943
+ "model.layers.78.mlp.experts.102.gate_proj",
944
+ "model.layers.78.mlp.experts.102.up_proj",
945
+ "model.layers.78.mlp.experts.103.down_proj",
946
+ "model.layers.78.mlp.experts.103.gate_proj",
947
+ "model.layers.78.mlp.experts.103.up_proj",
948
+ "model.layers.78.mlp.experts.104.down_proj",
949
+ "model.layers.78.mlp.experts.104.gate_proj",
950
+ "model.layers.78.mlp.experts.104.up_proj",
951
+ "model.layers.78.mlp.experts.105.down_proj",
952
+ "model.layers.78.mlp.experts.105.gate_proj",
953
+ "model.layers.78.mlp.experts.105.up_proj",
954
+ "model.layers.78.mlp.experts.106.down_proj",
955
+ "model.layers.78.mlp.experts.106.gate_proj",
956
+ "model.layers.78.mlp.experts.106.up_proj",
957
+ "model.layers.78.mlp.experts.107.down_proj",
958
+ "model.layers.78.mlp.experts.107.gate_proj",
959
+ "model.layers.78.mlp.experts.107.up_proj",
960
+ "model.layers.78.mlp.experts.108.down_proj",
961
+ "model.layers.78.mlp.experts.108.gate_proj",
962
+ "model.layers.78.mlp.experts.108.up_proj",
963
+ "model.layers.78.mlp.experts.109.down_proj",
964
+ "model.layers.78.mlp.experts.109.gate_proj",
965
+ "model.layers.78.mlp.experts.109.up_proj",
966
+ "model.layers.78.mlp.experts.11.down_proj",
967
+ "model.layers.78.mlp.experts.11.gate_proj",
968
+ "model.layers.78.mlp.experts.11.up_proj",
969
+ "model.layers.78.mlp.experts.110.down_proj",
970
+ "model.layers.78.mlp.experts.110.gate_proj",
971
+ "model.layers.78.mlp.experts.110.up_proj",
972
+ "model.layers.78.mlp.experts.111.down_proj",
973
+ "model.layers.78.mlp.experts.111.gate_proj",
974
+ "model.layers.78.mlp.experts.111.up_proj",
975
+ "model.layers.78.mlp.experts.112.down_proj",
976
+ "model.layers.78.mlp.experts.112.gate_proj",
977
+ "model.layers.78.mlp.experts.112.up_proj",
978
+ "model.layers.78.mlp.experts.113.down_proj",
979
+ "model.layers.78.mlp.experts.113.gate_proj",
980
+ "model.layers.78.mlp.experts.113.up_proj",
981
+ "model.layers.78.mlp.experts.114.down_proj",
982
+ "model.layers.78.mlp.experts.114.gate_proj",
983
+ "model.layers.78.mlp.experts.114.up_proj",
984
+ "model.layers.78.mlp.experts.115.down_proj",
985
+ "model.layers.78.mlp.experts.115.gate_proj",
986
+ "model.layers.78.mlp.experts.115.up_proj",
987
+ "model.layers.78.mlp.experts.116.down_proj",
988
+ "model.layers.78.mlp.experts.116.gate_proj",
989
+ "model.layers.78.mlp.experts.116.up_proj",
990
+ "model.layers.78.mlp.experts.117.down_proj",
991
+ "model.layers.78.mlp.experts.117.gate_proj",
992
+ "model.layers.78.mlp.experts.117.up_proj",
993
+ "model.layers.78.mlp.experts.118.down_proj",
994
+ "model.layers.78.mlp.experts.118.gate_proj",
995
+ "model.layers.78.mlp.experts.118.up_proj",
996
+ "model.layers.78.mlp.experts.119.down_proj",
997
+ "model.layers.78.mlp.experts.119.gate_proj",
998
+ "model.layers.78.mlp.experts.119.up_proj",
999
+ "model.layers.78.mlp.experts.12.down_proj",
1000
+ "model.layers.78.mlp.experts.12.gate_proj",
1001
+ "model.layers.78.mlp.experts.12.up_proj",
1002
+ "model.layers.78.mlp.experts.120.down_proj",
1003
+ "model.layers.78.mlp.experts.120.gate_proj",
1004
+ "model.layers.78.mlp.experts.120.up_proj",
1005
+ "model.layers.78.mlp.experts.121.down_proj",
1006
+ "model.layers.78.mlp.experts.121.gate_proj",
1007
+ "model.layers.78.mlp.experts.121.up_proj",
1008
+ "model.layers.78.mlp.experts.122.down_proj",
1009
+ "model.layers.78.mlp.experts.122.gate_proj",
1010
+ "model.layers.78.mlp.experts.122.up_proj",
1011
+ "model.layers.78.mlp.experts.123.down_proj",
1012
+ "model.layers.78.mlp.experts.123.gate_proj",
1013
+ "model.layers.78.mlp.experts.123.up_proj",
1014
+ "model.layers.78.mlp.experts.124.down_proj",
1015
+ "model.layers.78.mlp.experts.124.gate_proj",
1016
+ "model.layers.78.mlp.experts.124.up_proj",
1017
+ "model.layers.78.mlp.experts.125.down_proj",
1018
+ "model.layers.78.mlp.experts.125.gate_proj",
1019
+ "model.layers.78.mlp.experts.125.up_proj",
1020
+ "model.layers.78.mlp.experts.126.down_proj",
1021
+ "model.layers.78.mlp.experts.126.gate_proj",
1022
+ "model.layers.78.mlp.experts.126.up_proj",
1023
+ "model.layers.78.mlp.experts.127.down_proj",
1024
+ "model.layers.78.mlp.experts.127.gate_proj",
1025
+ "model.layers.78.mlp.experts.127.up_proj",
1026
+ "model.layers.78.mlp.experts.128.down_proj",
1027
+ "model.layers.78.mlp.experts.128.gate_proj",
1028
+ "model.layers.78.mlp.experts.128.up_proj",
1029
+ "model.layers.78.mlp.experts.129.down_proj",
1030
+ "model.layers.78.mlp.experts.129.gate_proj",
1031
+ "model.layers.78.mlp.experts.129.up_proj",
1032
+ "model.layers.78.mlp.experts.13.down_proj",
1033
+ "model.layers.78.mlp.experts.13.gate_proj",
1034
+ "model.layers.78.mlp.experts.13.up_proj",
1035
+ "model.layers.78.mlp.experts.130.down_proj",
1036
+ "model.layers.78.mlp.experts.130.gate_proj",
1037
+ "model.layers.78.mlp.experts.130.up_proj",
1038
+ "model.layers.78.mlp.experts.131.down_proj",
1039
+ "model.layers.78.mlp.experts.131.gate_proj",
1040
+ "model.layers.78.mlp.experts.131.up_proj",
1041
+ "model.layers.78.mlp.experts.132.down_proj",
1042
+ "model.layers.78.mlp.experts.132.gate_proj",
1043
+ "model.layers.78.mlp.experts.132.up_proj",
1044
+ "model.layers.78.mlp.experts.133.down_proj",
1045
+ "model.layers.78.mlp.experts.133.gate_proj",
1046
+ "model.layers.78.mlp.experts.133.up_proj",
1047
+ "model.layers.78.mlp.experts.134.down_proj",
1048
+ "model.layers.78.mlp.experts.134.gate_proj",
1049
+ "model.layers.78.mlp.experts.134.up_proj",
1050
+ "model.layers.78.mlp.experts.135.down_proj",
1051
+ "model.layers.78.mlp.experts.135.gate_proj",
1052
+ "model.layers.78.mlp.experts.135.up_proj",
1053
+ "model.layers.78.mlp.experts.136.down_proj",
1054
+ "model.layers.78.mlp.experts.136.gate_proj",
1055
+ "model.layers.78.mlp.experts.136.up_proj",
1056
+ "model.layers.78.mlp.experts.137.down_proj",
1057
+ "model.layers.78.mlp.experts.137.gate_proj",
1058
+ "model.layers.78.mlp.experts.137.up_proj",
1059
+ "model.layers.78.mlp.experts.138.down_proj",
1060
+ "model.layers.78.mlp.experts.138.gate_proj",
1061
+ "model.layers.78.mlp.experts.138.up_proj",
1062
+ "model.layers.78.mlp.experts.139.down_proj",
1063
+ "model.layers.78.mlp.experts.139.gate_proj",
1064
+ "model.layers.78.mlp.experts.139.up_proj",
1065
+ "model.layers.78.mlp.experts.14.down_proj",
1066
+ "model.layers.78.mlp.experts.14.gate_proj",
1067
+ "model.layers.78.mlp.experts.14.up_proj",
1068
+ "model.layers.78.mlp.experts.140.down_proj",
1069
+ "model.layers.78.mlp.experts.140.gate_proj",
1070
+ "model.layers.78.mlp.experts.140.up_proj",
1071
+ "model.layers.78.mlp.experts.141.down_proj",
1072
+ "model.layers.78.mlp.experts.141.gate_proj",
1073
+ "model.layers.78.mlp.experts.141.up_proj",
1074
+ "model.layers.78.mlp.experts.142.down_proj",
1075
+ "model.layers.78.mlp.experts.142.gate_proj",
1076
+ "model.layers.78.mlp.experts.142.up_proj",
1077
+ "model.layers.78.mlp.experts.143.down_proj",
1078
+ "model.layers.78.mlp.experts.143.gate_proj",
1079
+ "model.layers.78.mlp.experts.143.up_proj",
1080
+ "model.layers.78.mlp.experts.144.down_proj",
1081
+ "model.layers.78.mlp.experts.144.gate_proj",
1082
+ "model.layers.78.mlp.experts.144.up_proj",
1083
+ "model.layers.78.mlp.experts.145.down_proj",
1084
+ "model.layers.78.mlp.experts.145.gate_proj",
1085
+ "model.layers.78.mlp.experts.145.up_proj",
1086
+ "model.layers.78.mlp.experts.146.down_proj",
1087
+ "model.layers.78.mlp.experts.146.gate_proj",
1088
+ "model.layers.78.mlp.experts.146.up_proj",
1089
+ "model.layers.78.mlp.experts.147.down_proj",
1090
+ "model.layers.78.mlp.experts.147.gate_proj",
1091
+ "model.layers.78.mlp.experts.147.up_proj",
1092
+ "model.layers.78.mlp.experts.148.down_proj",
1093
+ "model.layers.78.mlp.experts.148.gate_proj",
1094
+ "model.layers.78.mlp.experts.148.up_proj",
1095
+ "model.layers.78.mlp.experts.149.down_proj",
1096
+ "model.layers.78.mlp.experts.149.gate_proj",
1097
+ "model.layers.78.mlp.experts.149.up_proj",
1098
+ "model.layers.78.mlp.experts.15.down_proj",
1099
+ "model.layers.78.mlp.experts.15.gate_proj",
1100
+ "model.layers.78.mlp.experts.15.up_proj",
1101
+ "model.layers.78.mlp.experts.150.down_proj",
1102
+ "model.layers.78.mlp.experts.150.gate_proj",
1103
+ "model.layers.78.mlp.experts.150.up_proj",
1104
+ "model.layers.78.mlp.experts.151.down_proj",
1105
+ "model.layers.78.mlp.experts.151.gate_proj",
1106
+ "model.layers.78.mlp.experts.151.up_proj",
1107
+ "model.layers.78.mlp.experts.152.down_proj",
1108
+ "model.layers.78.mlp.experts.152.gate_proj",
1109
+ "model.layers.78.mlp.experts.152.up_proj",
1110
+ "model.layers.78.mlp.experts.153.down_proj",
1111
+ "model.layers.78.mlp.experts.153.gate_proj",
1112
+ "model.layers.78.mlp.experts.153.up_proj",
1113
+ "model.layers.78.mlp.experts.154.down_proj",
1114
+ "model.layers.78.mlp.experts.154.gate_proj",
1115
+ "model.layers.78.mlp.experts.154.up_proj",
1116
+ "model.layers.78.mlp.experts.155.down_proj",
1117
+ "model.layers.78.mlp.experts.155.gate_proj",
1118
+ "model.layers.78.mlp.experts.155.up_proj",
1119
+ "model.layers.78.mlp.experts.156.down_proj",
1120
+ "model.layers.78.mlp.experts.156.gate_proj",
1121
+ "model.layers.78.mlp.experts.156.up_proj",
1122
+ "model.layers.78.mlp.experts.157.down_proj",
1123
+ "model.layers.78.mlp.experts.157.gate_proj",
1124
+ "model.layers.78.mlp.experts.157.up_proj",
1125
+ "model.layers.78.mlp.experts.158.down_proj",
1126
+ "model.layers.78.mlp.experts.158.gate_proj",
1127
+ "model.layers.78.mlp.experts.158.up_proj",
1128
+ "model.layers.78.mlp.experts.159.down_proj",
1129
+ "model.layers.78.mlp.experts.159.gate_proj",
1130
+ "model.layers.78.mlp.experts.159.up_proj",
1131
+ "model.layers.78.mlp.experts.16.down_proj",
1132
+ "model.layers.78.mlp.experts.16.gate_proj",
1133
+ "model.layers.78.mlp.experts.16.up_proj",
1134
+ "model.layers.78.mlp.experts.160.down_proj",
1135
+ "model.layers.78.mlp.experts.160.gate_proj",
1136
+ "model.layers.78.mlp.experts.160.up_proj",
1137
+ "model.layers.78.mlp.experts.161.down_proj",
1138
+ "model.layers.78.mlp.experts.161.gate_proj",
1139
+ "model.layers.78.mlp.experts.161.up_proj",
1140
+ "model.layers.78.mlp.experts.162.down_proj",
1141
+ "model.layers.78.mlp.experts.162.gate_proj",
1142
+ "model.layers.78.mlp.experts.162.up_proj",
1143
+ "model.layers.78.mlp.experts.163.down_proj",
1144
+ "model.layers.78.mlp.experts.163.gate_proj",
1145
+ "model.layers.78.mlp.experts.163.up_proj",
1146
+ "model.layers.78.mlp.experts.164.down_proj",
1147
+ "model.layers.78.mlp.experts.164.gate_proj",
1148
+ "model.layers.78.mlp.experts.164.up_proj",
1149
+ "model.layers.78.mlp.experts.165.down_proj",
1150
+ "model.layers.78.mlp.experts.165.gate_proj",
1151
+ "model.layers.78.mlp.experts.165.up_proj",
1152
+ "model.layers.78.mlp.experts.166.down_proj",
1153
+ "model.layers.78.mlp.experts.166.gate_proj",
1154
+ "model.layers.78.mlp.experts.166.up_proj",
1155
+ "model.layers.78.mlp.experts.167.down_proj",
1156
+ "model.layers.78.mlp.experts.167.gate_proj",
1157
+ "model.layers.78.mlp.experts.167.up_proj",
1158
+ "model.layers.78.mlp.experts.168.down_proj",
1159
+ "model.layers.78.mlp.experts.168.gate_proj",
1160
+ "model.layers.78.mlp.experts.168.up_proj",
1161
+ "model.layers.78.mlp.experts.169.down_proj",
1162
+ "model.layers.78.mlp.experts.169.gate_proj",
1163
+ "model.layers.78.mlp.experts.169.up_proj",
1164
+ "model.layers.78.mlp.experts.17.down_proj",
1165
+ "model.layers.78.mlp.experts.17.gate_proj",
1166
+ "model.layers.78.mlp.experts.17.up_proj",
1167
+ "model.layers.78.mlp.experts.170.down_proj",
1168
+ "model.layers.78.mlp.experts.170.gate_proj",
1169
+ "model.layers.78.mlp.experts.170.up_proj",
1170
+ "model.layers.78.mlp.experts.171.down_proj",
1171
+ "model.layers.78.mlp.experts.171.gate_proj",
1172
+ "model.layers.78.mlp.experts.171.up_proj",
1173
+ "model.layers.78.mlp.experts.172.down_proj",
1174
+ "model.layers.78.mlp.experts.172.gate_proj",
1175
+ "model.layers.78.mlp.experts.172.up_proj",
1176
+ "model.layers.78.mlp.experts.173.down_proj",
1177
+ "model.layers.78.mlp.experts.173.gate_proj",
1178
+ "model.layers.78.mlp.experts.173.up_proj",
1179
+ "model.layers.78.mlp.experts.174.down_proj",
1180
+ "model.layers.78.mlp.experts.174.gate_proj",
1181
+ "model.layers.78.mlp.experts.174.up_proj",
1182
+ "model.layers.78.mlp.experts.175.down_proj",
1183
+ "model.layers.78.mlp.experts.175.gate_proj",
1184
+ "model.layers.78.mlp.experts.175.up_proj",
1185
+ "model.layers.78.mlp.experts.176.down_proj",
1186
+ "model.layers.78.mlp.experts.176.gate_proj",
1187
+ "model.layers.78.mlp.experts.176.up_proj",
1188
+ "model.layers.78.mlp.experts.177.down_proj",
1189
+ "model.layers.78.mlp.experts.177.gate_proj",
1190
+ "model.layers.78.mlp.experts.177.up_proj",
1191
+ "model.layers.78.mlp.experts.178.down_proj",
1192
+ "model.layers.78.mlp.experts.178.gate_proj",
1193
+ "model.layers.78.mlp.experts.178.up_proj",
1194
+ "model.layers.78.mlp.experts.179.down_proj",
1195
+ "model.layers.78.mlp.experts.179.gate_proj",
1196
+ "model.layers.78.mlp.experts.179.up_proj",
1197
+ "model.layers.78.mlp.experts.18.down_proj",
1198
+ "model.layers.78.mlp.experts.18.gate_proj",
1199
+ "model.layers.78.mlp.experts.18.up_proj",
1200
+ "model.layers.78.mlp.experts.180.down_proj",
1201
+ "model.layers.78.mlp.experts.180.gate_proj",
1202
+ "model.layers.78.mlp.experts.180.up_proj",
1203
+ "model.layers.78.mlp.experts.181.down_proj",
1204
+ "model.layers.78.mlp.experts.181.gate_proj",
1205
+ "model.layers.78.mlp.experts.181.up_proj",
1206
+ "model.layers.78.mlp.experts.182.down_proj",
1207
+ "model.layers.78.mlp.experts.182.gate_proj",
1208
+ "model.layers.78.mlp.experts.182.up_proj",
1209
+ "model.layers.78.mlp.experts.183.down_proj",
1210
+ "model.layers.78.mlp.experts.183.gate_proj",
1211
+ "model.layers.78.mlp.experts.183.up_proj",
1212
+ "model.layers.78.mlp.experts.184.down_proj",
1213
+ "model.layers.78.mlp.experts.184.gate_proj",
1214
+ "model.layers.78.mlp.experts.184.up_proj",
1215
+ "model.layers.78.mlp.experts.185.down_proj",
1216
+ "model.layers.78.mlp.experts.185.gate_proj",
1217
+ "model.layers.78.mlp.experts.185.up_proj",
1218
+ "model.layers.78.mlp.experts.186.down_proj",
1219
+ "model.layers.78.mlp.experts.186.gate_proj",
1220
+ "model.layers.78.mlp.experts.186.up_proj",
1221
+ "model.layers.78.mlp.experts.187.down_proj",
1222
+ "model.layers.78.mlp.experts.187.gate_proj",
1223
+ "model.layers.78.mlp.experts.187.up_proj",
1224
+ "model.layers.78.mlp.experts.188.down_proj",
1225
+ "model.layers.78.mlp.experts.188.gate_proj",
1226
+ "model.layers.78.mlp.experts.188.up_proj",
1227
+ "model.layers.78.mlp.experts.189.down_proj",
1228
+ "model.layers.78.mlp.experts.189.gate_proj",
1229
+ "model.layers.78.mlp.experts.189.up_proj",
1230
+ "model.layers.78.mlp.experts.19.down_proj",
1231
+ "model.layers.78.mlp.experts.19.gate_proj",
1232
+ "model.layers.78.mlp.experts.19.up_proj",
1233
+ "model.layers.78.mlp.experts.190.down_proj",
1234
+ "model.layers.78.mlp.experts.190.gate_proj",
1235
+ "model.layers.78.mlp.experts.190.up_proj",
1236
+ "model.layers.78.mlp.experts.191.down_proj",
1237
+ "model.layers.78.mlp.experts.191.gate_proj",
1238
+ "model.layers.78.mlp.experts.191.up_proj",
1239
+ "model.layers.78.mlp.experts.192.down_proj",
1240
+ "model.layers.78.mlp.experts.192.gate_proj",
1241
+ "model.layers.78.mlp.experts.192.up_proj",
1242
+ "model.layers.78.mlp.experts.193.down_proj",
1243
+ "model.layers.78.mlp.experts.193.gate_proj",
1244
+ "model.layers.78.mlp.experts.193.up_proj",
1245
+ "model.layers.78.mlp.experts.194.down_proj",
1246
+ "model.layers.78.mlp.experts.194.gate_proj",
1247
+ "model.layers.78.mlp.experts.194.up_proj",
1248
+ "model.layers.78.mlp.experts.195.down_proj",
1249
+ "model.layers.78.mlp.experts.195.gate_proj",
1250
+ "model.layers.78.mlp.experts.195.up_proj",
1251
+ "model.layers.78.mlp.experts.196.down_proj",
1252
+ "model.layers.78.mlp.experts.196.gate_proj",
1253
+ "model.layers.78.mlp.experts.196.up_proj",
1254
+ "model.layers.78.mlp.experts.197.down_proj",
1255
+ "model.layers.78.mlp.experts.197.gate_proj",
1256
+ "model.layers.78.mlp.experts.197.up_proj",
1257
+ "model.layers.78.mlp.experts.198.down_proj",
1258
+ "model.layers.78.mlp.experts.198.gate_proj",
1259
+ "model.layers.78.mlp.experts.198.up_proj",
1260
+ "model.layers.78.mlp.experts.199.down_proj",
1261
+ "model.layers.78.mlp.experts.199.gate_proj",
1262
+ "model.layers.78.mlp.experts.199.up_proj",
1263
+ "model.layers.78.mlp.experts.2.down_proj",
1264
+ "model.layers.78.mlp.experts.2.gate_proj",
1265
+ "model.layers.78.mlp.experts.2.up_proj",
1266
+ "model.layers.78.mlp.experts.20.down_proj",
1267
+ "model.layers.78.mlp.experts.20.gate_proj",
1268
+ "model.layers.78.mlp.experts.20.up_proj",
1269
+ "model.layers.78.mlp.experts.200.down_proj",
1270
+ "model.layers.78.mlp.experts.200.gate_proj",
1271
+ "model.layers.78.mlp.experts.200.up_proj",
1272
+ "model.layers.78.mlp.experts.201.down_proj",
1273
+ "model.layers.78.mlp.experts.201.gate_proj",
1274
+ "model.layers.78.mlp.experts.201.up_proj",
1275
+ "model.layers.78.mlp.experts.202.down_proj",
1276
+ "model.layers.78.mlp.experts.202.gate_proj",
1277
+ "model.layers.78.mlp.experts.202.up_proj",
1278
+ "model.layers.78.mlp.experts.203.down_proj",
1279
+ "model.layers.78.mlp.experts.203.gate_proj",
1280
+ "model.layers.78.mlp.experts.203.up_proj",
1281
+ "model.layers.78.mlp.experts.204.down_proj",
1282
+ "model.layers.78.mlp.experts.204.gate_proj",
1283
+ "model.layers.78.mlp.experts.204.up_proj",
1284
+ "model.layers.78.mlp.experts.205.down_proj",
1285
+ "model.layers.78.mlp.experts.205.gate_proj",
1286
+ "model.layers.78.mlp.experts.205.up_proj",
1287
+ "model.layers.78.mlp.experts.206.down_proj",
1288
+ "model.layers.78.mlp.experts.206.gate_proj",
1289
+ "model.layers.78.mlp.experts.206.up_proj",
1290
+ "model.layers.78.mlp.experts.207.down_proj",
1291
+ "model.layers.78.mlp.experts.207.gate_proj",
1292
+ "model.layers.78.mlp.experts.207.up_proj",
1293
+ "model.layers.78.mlp.experts.208.down_proj",
1294
+ "model.layers.78.mlp.experts.208.gate_proj",
1295
+ "model.layers.78.mlp.experts.208.up_proj",
1296
+ "model.layers.78.mlp.experts.209.down_proj",
1297
+ "model.layers.78.mlp.experts.209.gate_proj",
1298
+ "model.layers.78.mlp.experts.209.up_proj",
1299
+ "model.layers.78.mlp.experts.21.down_proj",
1300
+ "model.layers.78.mlp.experts.21.gate_proj",
1301
+ "model.layers.78.mlp.experts.21.up_proj",
1302
+ "model.layers.78.mlp.experts.210.down_proj",
1303
+ "model.layers.78.mlp.experts.210.gate_proj",
1304
+ "model.layers.78.mlp.experts.210.up_proj",
1305
+ "model.layers.78.mlp.experts.211.down_proj",
1306
+ "model.layers.78.mlp.experts.211.gate_proj",
1307
+ "model.layers.78.mlp.experts.211.up_proj",
1308
+ "model.layers.78.mlp.experts.212.down_proj",
1309
+ "model.layers.78.mlp.experts.212.gate_proj",
1310
+ "model.layers.78.mlp.experts.212.up_proj",
1311
+ "model.layers.78.mlp.experts.213.down_proj",
1312
+ "model.layers.78.mlp.experts.213.gate_proj",
1313
+ "model.layers.78.mlp.experts.213.up_proj",
1314
+ "model.layers.78.mlp.experts.214.down_proj",
1315
+ "model.layers.78.mlp.experts.214.gate_proj",
1316
+ "model.layers.78.mlp.experts.214.up_proj",
1317
+ "model.layers.78.mlp.experts.215.down_proj",
1318
+ "model.layers.78.mlp.experts.215.gate_proj",
1319
+ "model.layers.78.mlp.experts.215.up_proj",
1320
+ "model.layers.78.mlp.experts.216.down_proj",
1321
+ "model.layers.78.mlp.experts.216.gate_proj",
1322
+ "model.layers.78.mlp.experts.216.up_proj",
1323
+ "model.layers.78.mlp.experts.217.down_proj",
1324
+ "model.layers.78.mlp.experts.217.gate_proj",
1325
+ "model.layers.78.mlp.experts.217.up_proj",
1326
+ "model.layers.78.mlp.experts.218.down_proj",
1327
+ "model.layers.78.mlp.experts.218.gate_proj",
1328
+ "model.layers.78.mlp.experts.218.up_proj",
1329
+ "model.layers.78.mlp.experts.219.down_proj",
1330
+ "model.layers.78.mlp.experts.219.gate_proj",
1331
+ "model.layers.78.mlp.experts.219.up_proj",
1332
+ "model.layers.78.mlp.experts.22.down_proj",
1333
+ "model.layers.78.mlp.experts.22.gate_proj",
1334
+ "model.layers.78.mlp.experts.22.up_proj",
1335
+ "model.layers.78.mlp.experts.220.down_proj",
1336
+ "model.layers.78.mlp.experts.220.gate_proj",
1337
+ "model.layers.78.mlp.experts.220.up_proj",
1338
+ "model.layers.78.mlp.experts.221.down_proj",
1339
+ "model.layers.78.mlp.experts.221.gate_proj",
1340
+ "model.layers.78.mlp.experts.221.up_proj",
1341
+ "model.layers.78.mlp.experts.222.down_proj",
1342
+ "model.layers.78.mlp.experts.222.gate_proj",
1343
+ "model.layers.78.mlp.experts.222.up_proj",
1344
+ "model.layers.78.mlp.experts.223.down_proj",
1345
+ "model.layers.78.mlp.experts.223.gate_proj",
1346
+ "model.layers.78.mlp.experts.223.up_proj",
1347
+ "model.layers.78.mlp.experts.224.down_proj",
1348
+ "model.layers.78.mlp.experts.224.gate_proj",
1349
+ "model.layers.78.mlp.experts.224.up_proj",
1350
+ "model.layers.78.mlp.experts.225.down_proj",
1351
+ "model.layers.78.mlp.experts.225.gate_proj",
1352
+ "model.layers.78.mlp.experts.225.up_proj",
1353
+ "model.layers.78.mlp.experts.226.down_proj",
1354
+ "model.layers.78.mlp.experts.226.gate_proj",
1355
+ "model.layers.78.mlp.experts.226.up_proj",
1356
+ "model.layers.78.mlp.experts.227.down_proj",
1357
+ "model.layers.78.mlp.experts.227.gate_proj",
1358
+ "model.layers.78.mlp.experts.227.up_proj",
1359
+ "model.layers.78.mlp.experts.228.down_proj",
1360
+ "model.layers.78.mlp.experts.228.gate_proj",
1361
+ "model.layers.78.mlp.experts.228.up_proj",
1362
+ "model.layers.78.mlp.experts.229.down_proj",
1363
+ "model.layers.78.mlp.experts.229.gate_proj",
1364
+ "model.layers.78.mlp.experts.229.up_proj",
1365
+ "model.layers.78.mlp.experts.23.down_proj",
1366
+ "model.layers.78.mlp.experts.23.gate_proj",
1367
+ "model.layers.78.mlp.experts.23.up_proj",
1368
+ "model.layers.78.mlp.experts.230.down_proj",
1369
+ "model.layers.78.mlp.experts.230.gate_proj",
1370
+ "model.layers.78.mlp.experts.230.up_proj",
1371
+ "model.layers.78.mlp.experts.231.down_proj",
1372
+ "model.layers.78.mlp.experts.231.gate_proj",
1373
+ "model.layers.78.mlp.experts.231.up_proj",
1374
+ "model.layers.78.mlp.experts.232.down_proj",
1375
+ "model.layers.78.mlp.experts.232.gate_proj",
1376
+ "model.layers.78.mlp.experts.232.up_proj",
1377
+ "model.layers.78.mlp.experts.233.down_proj",
1378
+ "model.layers.78.mlp.experts.233.gate_proj",
1379
+ "model.layers.78.mlp.experts.233.up_proj",
1380
+ "model.layers.78.mlp.experts.234.down_proj",
1381
+ "model.layers.78.mlp.experts.234.gate_proj",
1382
+ "model.layers.78.mlp.experts.234.up_proj",
1383
+ "model.layers.78.mlp.experts.235.down_proj",
1384
+ "model.layers.78.mlp.experts.235.gate_proj",
1385
+ "model.layers.78.mlp.experts.235.up_proj",
1386
+ "model.layers.78.mlp.experts.236.down_proj",
1387
+ "model.layers.78.mlp.experts.236.gate_proj",
1388
+ "model.layers.78.mlp.experts.236.up_proj",
1389
+ "model.layers.78.mlp.experts.237.down_proj",
1390
+ "model.layers.78.mlp.experts.237.gate_proj",
1391
+ "model.layers.78.mlp.experts.237.up_proj",
1392
+ "model.layers.78.mlp.experts.238.down_proj",
1393
+ "model.layers.78.mlp.experts.238.gate_proj",
1394
+ "model.layers.78.mlp.experts.238.up_proj",
1395
+ "model.layers.78.mlp.experts.239.down_proj",
1396
+ "model.layers.78.mlp.experts.239.gate_proj",
1397
+ "model.layers.78.mlp.experts.239.up_proj",
1398
+ "model.layers.78.mlp.experts.24.down_proj",
1399
+ "model.layers.78.mlp.experts.24.gate_proj",
1400
+ "model.layers.78.mlp.experts.24.up_proj",
1401
+ "model.layers.78.mlp.experts.240.down_proj",
1402
+ "model.layers.78.mlp.experts.240.gate_proj",
1403
+ "model.layers.78.mlp.experts.240.up_proj",
1404
+ "model.layers.78.mlp.experts.241.down_proj",
1405
+ "model.layers.78.mlp.experts.241.gate_proj",
1406
+ "model.layers.78.mlp.experts.241.up_proj",
1407
+ "model.layers.78.mlp.experts.242.down_proj",
1408
+ "model.layers.78.mlp.experts.242.gate_proj",
1409
+ "model.layers.78.mlp.experts.242.up_proj",
1410
+ "model.layers.78.mlp.experts.243.down_proj",
1411
+ "model.layers.78.mlp.experts.243.gate_proj",
1412
+ "model.layers.78.mlp.experts.243.up_proj",
1413
+ "model.layers.78.mlp.experts.244.down_proj",
1414
+ "model.layers.78.mlp.experts.244.gate_proj",
1415
+ "model.layers.78.mlp.experts.244.up_proj",
1416
+ "model.layers.78.mlp.experts.245.down_proj",
1417
+ "model.layers.78.mlp.experts.245.gate_proj",
1418
+ "model.layers.78.mlp.experts.245.up_proj",
1419
+ "model.layers.78.mlp.experts.246.down_proj",
1420
+ "model.layers.78.mlp.experts.246.gate_proj",
1421
+ "model.layers.78.mlp.experts.246.up_proj",
1422
+ "model.layers.78.mlp.experts.247.down_proj",
1423
+ "model.layers.78.mlp.experts.247.gate_proj",
1424
+ "model.layers.78.mlp.experts.247.up_proj",
1425
+ "model.layers.78.mlp.experts.248.down_proj",
1426
+ "model.layers.78.mlp.experts.248.gate_proj",
1427
+ "model.layers.78.mlp.experts.248.up_proj",
1428
+ "model.layers.78.mlp.experts.249.down_proj",
1429
+ "model.layers.78.mlp.experts.249.gate_proj",
1430
+ "model.layers.78.mlp.experts.249.up_proj",
1431
+ "model.layers.78.mlp.experts.25.down_proj",
1432
+ "model.layers.78.mlp.experts.25.gate_proj",
1433
+ "model.layers.78.mlp.experts.25.up_proj",
1434
+ "model.layers.78.mlp.experts.250.down_proj",
1435
+ "model.layers.78.mlp.experts.250.gate_proj",
1436
+ "model.layers.78.mlp.experts.250.up_proj",
1437
+ "model.layers.78.mlp.experts.251.down_proj",
1438
+ "model.layers.78.mlp.experts.251.gate_proj",
1439
+ "model.layers.78.mlp.experts.251.up_proj",
1440
+ "model.layers.78.mlp.experts.252.down_proj",
1441
+ "model.layers.78.mlp.experts.252.gate_proj",
1442
+ "model.layers.78.mlp.experts.252.up_proj",
1443
+ "model.layers.78.mlp.experts.253.down_proj",
1444
+ "model.layers.78.mlp.experts.253.gate_proj",
1445
+ "model.layers.78.mlp.experts.253.up_proj",
1446
+ "model.layers.78.mlp.experts.254.down_proj",
1447
+ "model.layers.78.mlp.experts.254.gate_proj",
1448
+ "model.layers.78.mlp.experts.254.up_proj",
1449
+ "model.layers.78.mlp.experts.255.down_proj",
1450
+ "model.layers.78.mlp.experts.255.gate_proj",
1451
+ "model.layers.78.mlp.experts.255.up_proj",
1452
+ "model.layers.78.mlp.experts.26.down_proj",
1453
+ "model.layers.78.mlp.experts.26.gate_proj",
1454
+ "model.layers.78.mlp.experts.26.up_proj",
1455
+ "model.layers.78.mlp.experts.27.down_proj",
1456
+ "model.layers.78.mlp.experts.27.gate_proj",
1457
+ "model.layers.78.mlp.experts.27.up_proj",
1458
+ "model.layers.78.mlp.experts.28.down_proj",
1459
+ "model.layers.78.mlp.experts.28.gate_proj",
1460
+ "model.layers.78.mlp.experts.28.up_proj",
1461
+ "model.layers.78.mlp.experts.29.down_proj",
1462
+ "model.layers.78.mlp.experts.29.gate_proj",
1463
+ "model.layers.78.mlp.experts.29.up_proj",
1464
+ "model.layers.78.mlp.experts.3.down_proj",
1465
+ "model.layers.78.mlp.experts.3.gate_proj",
1466
+ "model.layers.78.mlp.experts.3.up_proj",
1467
+ "model.layers.78.mlp.experts.30.down_proj",
1468
+ "model.layers.78.mlp.experts.30.gate_proj",
1469
+ "model.layers.78.mlp.experts.30.up_proj",
1470
+ "model.layers.78.mlp.experts.31.down_proj",
1471
+ "model.layers.78.mlp.experts.31.gate_proj",
1472
+ "model.layers.78.mlp.experts.31.up_proj",
1473
+ "model.layers.78.mlp.experts.32.down_proj",
1474
+ "model.layers.78.mlp.experts.32.gate_proj",
1475
+ "model.layers.78.mlp.experts.32.up_proj",
1476
+ "model.layers.78.mlp.experts.33.down_proj",
1477
+ "model.layers.78.mlp.experts.33.gate_proj",
1478
+ "model.layers.78.mlp.experts.33.up_proj",
1479
+ "model.layers.78.mlp.experts.34.down_proj",
1480
+ "model.layers.78.mlp.experts.34.gate_proj",
1481
+ "model.layers.78.mlp.experts.34.up_proj",
1482
+ "model.layers.78.mlp.experts.35.down_proj",
1483
+ "model.layers.78.mlp.experts.35.gate_proj",
1484
+ "model.layers.78.mlp.experts.35.up_proj",
1485
+ "model.layers.78.mlp.experts.36.down_proj",
1486
+ "model.layers.78.mlp.experts.36.gate_proj",
1487
+ "model.layers.78.mlp.experts.36.up_proj",
1488
+ "model.layers.78.mlp.experts.37.down_proj",
1489
+ "model.layers.78.mlp.experts.37.gate_proj",
1490
+ "model.layers.78.mlp.experts.37.up_proj",
1491
+ "model.layers.78.mlp.experts.38.down_proj",
1492
+ "model.layers.78.mlp.experts.38.gate_proj",
1493
+ "model.layers.78.mlp.experts.38.up_proj",
1494
+ "model.layers.78.mlp.experts.39.down_proj",
1495
+ "model.layers.78.mlp.experts.39.gate_proj",
1496
+ "model.layers.78.mlp.experts.39.up_proj",
1497
+ "model.layers.78.mlp.experts.4.down_proj",
1498
+ "model.layers.78.mlp.experts.4.gate_proj",
1499
+ "model.layers.78.mlp.experts.4.up_proj",
1500
+ "model.layers.78.mlp.experts.40.down_proj",
1501
+ "model.layers.78.mlp.experts.40.gate_proj",
1502
+ "model.layers.78.mlp.experts.40.up_proj",
1503
+ "model.layers.78.mlp.experts.41.down_proj",
1504
+ "model.layers.78.mlp.experts.41.gate_proj",
1505
+ "model.layers.78.mlp.experts.41.up_proj",
1506
+ "model.layers.78.mlp.experts.42.down_proj",
1507
+ "model.layers.78.mlp.experts.42.gate_proj",
1508
+ "model.layers.78.mlp.experts.42.up_proj",
1509
+ "model.layers.78.mlp.experts.43.down_proj",
1510
+ "model.layers.78.mlp.experts.43.gate_proj",
1511
+ "model.layers.78.mlp.experts.43.up_proj",
1512
+ "model.layers.78.mlp.experts.44.down_proj",
1513
+ "model.layers.78.mlp.experts.44.gate_proj",
1514
+ "model.layers.78.mlp.experts.44.up_proj",
1515
+ "model.layers.78.mlp.experts.45.down_proj",
1516
+ "model.layers.78.mlp.experts.45.gate_proj",
1517
+ "model.layers.78.mlp.experts.45.up_proj",
1518
+ "model.layers.78.mlp.experts.46.down_proj",
1519
+ "model.layers.78.mlp.experts.46.gate_proj",
1520
+ "model.layers.78.mlp.experts.46.up_proj",
1521
+ "model.layers.78.mlp.experts.47.down_proj",
1522
+ "model.layers.78.mlp.experts.47.gate_proj",
1523
+ "model.layers.78.mlp.experts.47.up_proj",
1524
+ "model.layers.78.mlp.experts.48.down_proj",
1525
+ "model.layers.78.mlp.experts.48.gate_proj",
1526
+ "model.layers.78.mlp.experts.48.up_proj",
1527
+ "model.layers.78.mlp.experts.49.down_proj",
1528
+ "model.layers.78.mlp.experts.49.gate_proj",
1529
+ "model.layers.78.mlp.experts.49.up_proj",
1530
+ "model.layers.78.mlp.experts.5.down_proj",
1531
+ "model.layers.78.mlp.experts.5.gate_proj",
1532
+ "model.layers.78.mlp.experts.5.up_proj",
1533
+ "model.layers.78.mlp.experts.50.down_proj",
1534
+ "model.layers.78.mlp.experts.50.gate_proj",
1535
+ "model.layers.78.mlp.experts.50.up_proj",
1536
+ "model.layers.78.mlp.experts.51.down_proj",
1537
+ "model.layers.78.mlp.experts.51.gate_proj",
1538
+ "model.layers.78.mlp.experts.51.up_proj",
1539
+ "model.layers.78.mlp.experts.52.down_proj",
1540
+ "model.layers.78.mlp.experts.52.gate_proj",
1541
+ "model.layers.78.mlp.experts.52.up_proj",
1542
+ "model.layers.78.mlp.experts.53.down_proj",
1543
+ "model.layers.78.mlp.experts.53.gate_proj",
1544
+ "model.layers.78.mlp.experts.53.up_proj",
1545
+ "model.layers.78.mlp.experts.54.down_proj",
1546
+ "model.layers.78.mlp.experts.54.gate_proj",
1547
+ "model.layers.78.mlp.experts.54.up_proj",
1548
+ "model.layers.78.mlp.experts.55.down_proj",
1549
+ "model.layers.78.mlp.experts.55.gate_proj",
1550
+ "model.layers.78.mlp.experts.55.up_proj",
1551
+ "model.layers.78.mlp.experts.56.down_proj",
1552
+ "model.layers.78.mlp.experts.56.gate_proj",
1553
+ "model.layers.78.mlp.experts.56.up_proj",
1554
+ "model.layers.78.mlp.experts.57.down_proj",
1555
+ "model.layers.78.mlp.experts.57.gate_proj",
1556
+ "model.layers.78.mlp.experts.57.up_proj",
1557
+ "model.layers.78.mlp.experts.58.down_proj",
1558
+ "model.layers.78.mlp.experts.58.gate_proj",
1559
+ "model.layers.78.mlp.experts.58.up_proj",
1560
+ "model.layers.78.mlp.experts.59.down_proj",
1561
+ "model.layers.78.mlp.experts.59.gate_proj",
1562
+ "model.layers.78.mlp.experts.59.up_proj",
1563
+ "model.layers.78.mlp.experts.6.down_proj",
1564
+ "model.layers.78.mlp.experts.6.gate_proj",
1565
+ "model.layers.78.mlp.experts.6.up_proj",
1566
+ "model.layers.78.mlp.experts.60.down_proj",
1567
+ "model.layers.78.mlp.experts.60.gate_proj",
1568
+ "model.layers.78.mlp.experts.60.up_proj",
1569
+ "model.layers.78.mlp.experts.61.down_proj",
1570
+ "model.layers.78.mlp.experts.61.gate_proj",
1571
+ "model.layers.78.mlp.experts.61.up_proj",
1572
+ "model.layers.78.mlp.experts.62.down_proj",
1573
+ "model.layers.78.mlp.experts.62.gate_proj",
1574
+ "model.layers.78.mlp.experts.62.up_proj",
1575
+ "model.layers.78.mlp.experts.63.down_proj",
1576
+ "model.layers.78.mlp.experts.63.gate_proj",
1577
+ "model.layers.78.mlp.experts.63.up_proj",
1578
+ "model.layers.78.mlp.experts.64.down_proj",
1579
+ "model.layers.78.mlp.experts.64.gate_proj",
1580
+ "model.layers.78.mlp.experts.64.up_proj",
1581
+ "model.layers.78.mlp.experts.65.down_proj",
1582
+ "model.layers.78.mlp.experts.65.gate_proj",
1583
+ "model.layers.78.mlp.experts.65.up_proj",
1584
+ "model.layers.78.mlp.experts.66.down_proj",
1585
+ "model.layers.78.mlp.experts.66.gate_proj",
1586
+ "model.layers.78.mlp.experts.66.up_proj",
1587
+ "model.layers.78.mlp.experts.67.down_proj",
1588
+ "model.layers.78.mlp.experts.67.gate_proj",
1589
+ "model.layers.78.mlp.experts.67.up_proj",
1590
+ "model.layers.78.mlp.experts.68.down_proj",
1591
+ "model.layers.78.mlp.experts.68.gate_proj",
1592
+ "model.layers.78.mlp.experts.68.up_proj",
1593
+ "model.layers.78.mlp.experts.69.down_proj",
1594
+ "model.layers.78.mlp.experts.69.gate_proj",
1595
+ "model.layers.78.mlp.experts.69.up_proj",
1596
+ "model.layers.78.mlp.experts.7.down_proj",
1597
+ "model.layers.78.mlp.experts.7.gate_proj",
1598
+ "model.layers.78.mlp.experts.7.up_proj",
1599
+ "model.layers.78.mlp.experts.70.down_proj",
1600
+ "model.layers.78.mlp.experts.70.gate_proj",
1601
+ "model.layers.78.mlp.experts.70.up_proj",
1602
+ "model.layers.78.mlp.experts.71.down_proj",
1603
+ "model.layers.78.mlp.experts.71.gate_proj",
1604
+ "model.layers.78.mlp.experts.71.up_proj",
1605
+ "model.layers.78.mlp.experts.72.down_proj",
1606
+ "model.layers.78.mlp.experts.72.gate_proj",
1607
+ "model.layers.78.mlp.experts.72.up_proj",
1608
+ "model.layers.78.mlp.experts.73.down_proj",
1609
+ "model.layers.78.mlp.experts.73.gate_proj",
1610
+ "model.layers.78.mlp.experts.73.up_proj",
1611
+ "model.layers.78.mlp.experts.74.down_proj",
1612
+ "model.layers.78.mlp.experts.74.gate_proj",
1613
+ "model.layers.78.mlp.experts.74.up_proj",
1614
+ "model.layers.78.mlp.experts.75.down_proj",
1615
+ "model.layers.78.mlp.experts.75.gate_proj",
1616
+ "model.layers.78.mlp.experts.75.up_proj",
1617
+ "model.layers.78.mlp.experts.76.down_proj",
1618
+ "model.layers.78.mlp.experts.76.gate_proj",
1619
+ "model.layers.78.mlp.experts.76.up_proj",
1620
+ "model.layers.78.mlp.experts.77.down_proj",
1621
+ "model.layers.78.mlp.experts.77.gate_proj",
1622
+ "model.layers.78.mlp.experts.77.up_proj",
1623
+ "model.layers.78.mlp.experts.78.down_proj",
1624
+ "model.layers.78.mlp.experts.78.gate_proj",
1625
+ "model.layers.78.mlp.experts.78.up_proj",
1626
+ "model.layers.78.mlp.experts.79.down_proj",
1627
+ "model.layers.78.mlp.experts.79.gate_proj",
1628
+ "model.layers.78.mlp.experts.79.up_proj",
1629
+ "model.layers.78.mlp.experts.8.down_proj",
1630
+ "model.layers.78.mlp.experts.8.gate_proj",
1631
+ "model.layers.78.mlp.experts.8.up_proj",
1632
+ "model.layers.78.mlp.experts.80.down_proj",
1633
+ "model.layers.78.mlp.experts.80.gate_proj",
1634
+ "model.layers.78.mlp.experts.80.up_proj",
1635
+ "model.layers.78.mlp.experts.81.down_proj",
1636
+ "model.layers.78.mlp.experts.81.gate_proj",
1637
+ "model.layers.78.mlp.experts.81.up_proj",
1638
+ "model.layers.78.mlp.experts.82.down_proj",
1639
+ "model.layers.78.mlp.experts.82.gate_proj",
1640
+ "model.layers.78.mlp.experts.82.up_proj",
1641
+ "model.layers.78.mlp.experts.83.down_proj",
1642
+ "model.layers.78.mlp.experts.83.gate_proj",
1643
+ "model.layers.78.mlp.experts.83.up_proj",
1644
+ "model.layers.78.mlp.experts.84.down_proj",
1645
+ "model.layers.78.mlp.experts.84.gate_proj",
1646
+ "model.layers.78.mlp.experts.84.up_proj",
1647
+ "model.layers.78.mlp.experts.85.down_proj",
1648
+ "model.layers.78.mlp.experts.85.gate_proj",
1649
+ "model.layers.78.mlp.experts.85.up_proj",
1650
+ "model.layers.78.mlp.experts.86.down_proj",
1651
+ "model.layers.78.mlp.experts.86.gate_proj",
1652
+ "model.layers.78.mlp.experts.86.up_proj",
1653
+ "model.layers.78.mlp.experts.87.down_proj",
1654
+ "model.layers.78.mlp.experts.87.gate_proj",
1655
+ "model.layers.78.mlp.experts.87.up_proj",
1656
+ "model.layers.78.mlp.experts.88.down_proj",
1657
+ "model.layers.78.mlp.experts.88.gate_proj",
1658
+ "model.layers.78.mlp.experts.88.up_proj",
1659
+ "model.layers.78.mlp.experts.89.down_proj",
1660
+ "model.layers.78.mlp.experts.89.gate_proj",
1661
+ "model.layers.78.mlp.experts.89.up_proj",
1662
+ "model.layers.78.mlp.experts.9.down_proj",
1663
+ "model.layers.78.mlp.experts.9.gate_proj",
1664
+ "model.layers.78.mlp.experts.9.up_proj",
1665
+ "model.layers.78.mlp.experts.90.down_proj",
1666
+ "model.layers.78.mlp.experts.90.gate_proj",
1667
+ "model.layers.78.mlp.experts.90.up_proj",
1668
+ "model.layers.78.mlp.experts.91.down_proj",
1669
+ "model.layers.78.mlp.experts.91.gate_proj",
1670
+ "model.layers.78.mlp.experts.91.up_proj",
1671
+ "model.layers.78.mlp.experts.92.down_proj",
1672
+ "model.layers.78.mlp.experts.92.gate_proj",
1673
+ "model.layers.78.mlp.experts.92.up_proj",
1674
+ "model.layers.78.mlp.experts.93.down_proj",
1675
+ "model.layers.78.mlp.experts.93.gate_proj",
1676
+ "model.layers.78.mlp.experts.93.up_proj",
1677
+ "model.layers.78.mlp.experts.94.down_proj",
1678
+ "model.layers.78.mlp.experts.94.gate_proj",
1679
+ "model.layers.78.mlp.experts.94.up_proj",
1680
+ "model.layers.78.mlp.experts.95.down_proj",
1681
+ "model.layers.78.mlp.experts.95.gate_proj",
1682
+ "model.layers.78.mlp.experts.95.up_proj",
1683
+ "model.layers.78.mlp.experts.96.down_proj",
1684
+ "model.layers.78.mlp.experts.96.gate_proj",
1685
+ "model.layers.78.mlp.experts.96.up_proj",
1686
+ "model.layers.78.mlp.experts.97.down_proj",
1687
+ "model.layers.78.mlp.experts.97.gate_proj",
1688
+ "model.layers.78.mlp.experts.97.up_proj",
1689
+ "model.layers.78.mlp.experts.98.down_proj",
1690
+ "model.layers.78.mlp.experts.98.gate_proj",
1691
+ "model.layers.78.mlp.experts.98.up_proj",
1692
+ "model.layers.78.mlp.experts.99.down_proj",
1693
+ "model.layers.78.mlp.experts.99.gate_proj",
1694
+ "model.layers.78.mlp.experts.99.up_proj",
1695
+ "model.layers.78.mlp.gate",
1696
+ "model.layers.78.mlp.gate.e_score_correction_bias",
1697
+ "model.layers.78.mlp.shared_experts.down_proj",
1698
+ "model.layers.78.mlp.shared_experts.gate_proj",
1699
+ "model.layers.78.mlp.shared_experts.up_proj",
1700
+ "model.layers.78.post_attention_layernorm",
1701
+ "model.layers.78.self_attn.indexer.k_norm",
1702
+ "model.layers.78.self_attn.indexer.weights_proj",
1703
+ "model.layers.78.self_attn.indexer.wk",
1704
+ "model.layers.78.self_attn.indexer.wq_b",
1705
+ "model.layers.78.self_attn.kv_a_layernorm",
1706
+ "model.layers.78.self_attn.kv_a_proj_with_mqa",
1707
+ "model.layers.78.self_attn.kv_b_proj",
1708
+ "model.layers.78.self_attn.o_proj",
1709
+ "model.layers.78.self_attn.q_a_layernorm",
1710
+ "model.layers.78.self_attn.q_a_proj",
1711
+ "model.layers.78.self_attn.q_b_proj",
1712
+ "model.layers.78.shared_head.norm"
1713
  ],
1714
  "export": {
1715
  "kv_cache_group": [],
 
1806
  "quant_method": "quark",
1807
  "quant_mode": "eager_mode",
1808
  "softmax_quant_spec": null,
1809
+ "version": "0.13+43a6fd31a8b"
1810
  },
1811
  "rms_norm_eps": 1e-05,
1812
  "rope_interleave": true,
 
1819
  "tie_word_embeddings": false,
1820
  "topk_group": 1,
1821
  "topk_method": "noaux_tc",
1822
+ "transformers_version": "5.14.1",
1823
  "use_cache": true,
1824
  "v_head_dim": 256,
1825
  "vocab_size": 154880
generation_config.json CHANGED
@@ -9,5 +9,5 @@
9
  "pad_token_id": 154820,
10
  "temperature": 1.0,
11
  "top_p": 0.95,
12
- "transformers_version": "5.2.0"
13
  }
 
9
  "pad_token_id": 154820,
10
  "temperature": 1.0,
11
  "top_p": 0.95,
12
+ "transformers_version": "5.14.1"
13
  }
model-00001-of-00010.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a45e81d75b805837fc03630ed117c9540ece023a078d3b88a5ea6e9683d5eac
3
+ size 49999051524
model-00002-of-00010.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:109d737b0859e6c9681146ac8fe4b602a007e53b01df4afd70b256ba94d44d9a
3
+ size 49998227584
model-00003-of-00010.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d061a391fc8663cf7469e88844b62c6e0d5bf5664e91110ed1d7cbfd9f04bd8c
3
+ size 50000184968
model-00004-of-00010.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2cc0a2fc3866520927f4d02a26d02444a6bf839831566dced758d2ee65cef2ae
3
+ size 49998227296
model-00005-of-00010.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93b24ee3751b9be14b6d63cb389871c5e1264e1326933c0c611b174ff39c36fd
3
+ size 49916961116
model-00006-of-00010.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:874b3be12440770807602fe1dd75d9500481e5d29474354d92378ea9b66331fa
3
+ size 50002802316
model-00007-of-00010.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b24ff29036acc0d4418d3fd23aaece4132789d93464d7758d072494070f5faa1
3
+ size 49999014384
model-00008-of-00010.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:315c426f17a0a948c16847eae99cf364094fc3512021305646d0abcca002a762
3
+ size 50000185032
model-00009-of-00010.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11e978447cd26dab09f5d3cdb6e3e57e6d0d33b09a1ce19abc62b79559310204
3
+ size 49979254916
model-00010-of-00010.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7deb9740073b7d7afdb0cb37af1b3316275a3a6b04abb3f540c239ac23abd1b
3
+ size 11928311112
model.safetensors.index.json CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0f23d65b82e78a3c64d765704f0f473a4792760186fbb191ea8bec8c05c65ee3
3
- size 22403819
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eb3fa16d9f6b25fdd4fab62466a8894aa45c63b86e27d9a728520e909739d304
3
+ size 22403820