tests : expose MoE weighted reduction in fusion baseline

Use 2 experts per token in the generated MoE test models so the Metal
MoE weighted reduction fusion (MUL + ADD) is exercised by test-fusion.

Assisted-by: pi:llama.cpp/DeepSeek-V4-Flash-Vision-Exp
This commit is contained in:
Georgi Gerganov
2026-09-15 15:17:50 +03:00
parent fcb9607ffe
commit 7c304a5114
2 changed files with 43 additions and 1 deletions
+42
View File
@@ -1,38 +1,49 @@
# test-fusion baseline for device MTL
# arch ,moe ,mode ,label , count
arcee ,0 ,any ,RMS_NORM+MUL , 5
arctic ,0 ,any ,MUL+ADD , 4
arctic ,0 ,any ,RMS_NORM+MUL , 7
arctic ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
baichuan ,0 ,any ,RMS_NORM+MUL , 5
bailingmoe ,1 ,any ,ADD+ADD , 2
bailingmoe ,1 ,any ,MUL+ADD , 4
bailingmoe ,1 ,any ,RMS_NORM+MUL , 5
bailingmoe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS , 2
bailingmoe2 ,1 ,any ,ADD+ADD , 1
bailingmoe2 ,1 ,any ,MUL+ADD , 2
bailingmoe2 ,1 ,any ,RMS_NORM+MUL , 9
bailingmoe3 ,1 ,any ,ADD+ADD , 1
bailingmoe3 ,1 ,any ,GATED_DELTA_NET+CPY , 1
bailingmoe3 ,1 ,any ,MUL+ADD , 2
bailingmoe3 ,1 ,any ,RMS_NORM+MUL , 8
bloom ,0 ,any ,NORM+MUL+ADD , 6
chatglm ,0 ,any ,RMS_NORM+MUL , 5
codeshell ,0 ,any ,NORM+MUL+ADD , 5
cogvlm ,0 ,any ,RMS_NORM+MUL , 5
command-r ,0 ,any ,NORM+MUL , 3
dbrx ,0 ,any ,MUL+ADD , 4
dbrx ,0 ,any ,NORM+MUL , 5
dbrx ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
deci ,0 ,any ,RMS_NORM+MUL , 5
deepseek ,0 ,any ,ADD+ADD , 1
deepseek ,0 ,any ,MUL+ADD , 2
deepseek ,0 ,any ,RMS_NORM+MUL , 5
deepseek ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS , 1
deepseek2 ,0 ,any ,ADD+ADD , 1
deepseek2 ,0 ,any ,MUL+ADD , 2
deepseek2 ,0 ,any ,RMS_NORM+MUL , 9
deepseek32 ,0 ,any ,ADD+ADD , 1
deepseek32 ,0 ,any ,MUL+ADD , 2
deepseek32 ,0 ,any ,NORM+MUL+ADD , 2
deepseek32 ,0 ,any ,RMS_NORM+MUL , 9
deepseek4 ,0 ,any ,MUL+ADD , 8
deepseek4 ,0 ,any ,RMS_NORM+MUL , 20
dots1 ,0 ,any ,ADD+ADD , 1
dots1 ,0 ,any ,MUL+ADD , 2
dots1 ,0 ,any ,RMS_NORM+MUL , 9
dream ,0 ,any ,RMS_NORM+MUL , 5
ernie4_5-moe ,1 ,any ,ADD+ADD , 1
ernie4_5-moe ,1 ,any ,MUL+ADD , 2
ernie4_5-moe ,1 ,any ,RMS_NORM+MUL , 5
ernie4_5 ,0 ,any ,RMS_NORM+MUL , 5
exaone ,0 ,any ,RMS_NORM+MUL , 5
@@ -46,34 +57,44 @@ gemma ,0 ,any ,RMS_NORM+MUL , 5
gemma2 ,0 ,any ,RMS_NORM+MUL , 5
gemma2 ,0 ,any ,RMS_NORM+MUL+ADD , 4
glm-dsa ,0 ,any ,ADD+ADD , 1
glm-dsa ,0 ,any ,MUL+ADD , 2
glm-dsa ,0 ,any ,NORM+MUL+ADD , 2
glm-dsa ,0 ,any ,RMS_NORM+MUL , 9
glm4 ,0 ,any ,RMS_NORM+MUL , 5
glm4 ,0 ,any ,RMS_NORM+MUL+ADD , 4
glm4moe ,1 ,any ,ADD+ADD , 1
glm4moe ,1 ,any ,MUL+ADD , 2
glm4moe ,1 ,any ,RMS_NORM+MUL , 9
gpt-oss ,0 ,any ,MUL+ADD , 4
gpt-oss ,0 ,any ,RMS_NORM+MUL , 5
gpt2 ,0 ,any ,NORM+MUL+ADD , 5
gptneox ,0 ,any ,NORM+MUL+ADD , 5
granite ,0 ,any ,RMS_NORM+MUL , 5
granite ,0 ,any ,MUL+ADD , 4
granite ,0 ,any ,RMS_NORM+MUL , 5
granite ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
granitehybrid ,0 ,any ,RMS_NORM+MUL , 6
granitemoe ,1 ,any ,RMS_NORM+MUL , 5
granitemoe ,1 ,any ,MUL+ADD , 4
granitemoe ,1 ,any ,RMS_NORM+MUL , 5
granitemoe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
grok ,0 ,any ,MUL+ADD , 4
grok ,0 ,any ,RMS_NORM+MUL , 5
grok ,0 ,any ,RMS_NORM+MUL+ADD , 4
grok ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
grovemoe ,1 ,any ,ADD+ADD , 2
grovemoe ,1 ,any ,MUL+ADD , 8
grovemoe ,1 ,any ,RMS_NORM+MUL , 9
hunyuan-dense ,0 ,any ,RMS_NORM+MUL , 9
hunyuan-moe ,1 ,any ,ADD+ADD , 2
hunyuan-moe ,1 ,any ,MUL+ADD , 4
hunyuan-moe ,1 ,any ,RMS_NORM+MUL , 9
hunyuan-moe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
hunyuan_vl ,0 ,any ,RMS_NORM+MUL , 9
hy_v3 ,0 ,any ,ADD+ADD , 2
hy_v3 ,0 ,any ,MUL+ADD , 4
hy_v3 ,0 ,any ,RMS_NORM+MUL , 9
hy_v4 ,0 ,any ,MUL+ADD , 2
hy_v4 ,0 ,any ,NORM+MUL+ADD , 1
hy_v4 ,0 ,any ,RMS_NORM+MUL , 9
internlm2 ,0 ,any ,RMS_NORM+MUL , 5
@@ -81,16 +102,21 @@ jais ,0 ,any ,NORM+MUL+ADD , 5
jais2 ,0 ,any ,NORM+MUL+ADD , 5
jamba ,0 ,any ,RMS_NORM+MUL , 8
kimi-k3 ,0 ,any ,GATED_DELTA_NET+CPY , 1
kimi-k3 ,0 ,any ,MUL+ADD , 2
kimi-k3 ,0 ,any ,RMS_NORM+MUL , 17
kimi-linear ,0 ,any ,ADD+ADD , 1
kimi-linear ,0 ,any ,GATED_DELTA_NET+CPY , 1
kimi-linear ,0 ,any ,MUL+ADD , 2
kimi-linear ,0 ,any ,RMS_NORM+MUL , 7
lfm2 ,0 ,any ,RMS_NORM+MUL , 7
lfm2moe ,1 ,any ,MUL+ADD , 2
lfm2moe ,1 ,any ,RMS_NORM+MUL , 7
llada ,0 ,any ,RMS_NORM+MUL , 5
llada-moe ,1 ,any ,MUL+ADD , 4
llada-moe ,1 ,any ,RMS_NORM+MUL , 9
llada-moe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS , 2
llama ,0 ,any ,RMS_NORM+MUL , 5
llama ,0 ,any ,MUL+ADD , 4
llama ,0 ,any ,RMS_NORM+MUL , 5
llama ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
llama4 ,0 ,any ,ADD+ADD , 2
@@ -99,24 +125,31 @@ maincoder ,0 ,any ,RMS_NORM+MUL , 9
mamba ,0 ,any ,RMS_NORM+MUL , 3
mamba2 ,0 ,any ,RMS_NORM+MUL , 5
minicpm ,0 ,any ,RMS_NORM+MUL , 5
minicpm ,0 ,any ,MUL+ADD , 4
minicpm ,0 ,any ,RMS_NORM+MUL , 5
minicpm ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
minicpm3 ,0 ,any ,RMS_NORM+MUL , 9
minimax-01 ,0 ,any ,MUL+ADD , 4
minimax-01 ,0 ,any ,RMS_NORM+MUL , 6
minimax-01 ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
minimax-m2 ,0 ,any ,MUL+ADD , 4
minimax-m2 ,0 ,any ,RMS_NORM+MUL , 9
minimax-m3 ,0 ,any ,ADD+ADD , 1
minimax-m3 ,0 ,any ,MUL+ADD , 2
minimax-m3 ,0 ,any ,RMS_NORM+MUL , 11
mistral3 ,0 ,any ,RMS_NORM+MUL , 5
mistral3 ,0 ,any ,MUL+ADD , 4
mistral3 ,0 ,any ,RMS_NORM+MUL , 5
mistral3 ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
mistral4 ,0 ,any ,ADD+ADD , 1
mistral4 ,0 ,any ,MUL+ADD , 2
mistral4 ,0 ,any ,RMS_NORM+MUL , 9
mpt ,0 ,any ,NORM+MUL+ADD , 5
nanbeige ,0 ,any ,RMS_NORM+MUL , 5
nemotron ,0 ,any ,NORM+MUL+ADD , 5
nemotron_h ,0 ,any ,RMS_NORM+MUL , 5
nemotron_h_moe ,1 ,any ,RMS_NORM+MUL , 5
olmoe ,1 ,any ,MUL+ADD , 4
olmoe ,1 ,any ,RMS_NORM+MUL , 9
olmoe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS , 2
openelm ,0 ,any ,RMS_NORM+MUL , 9
@@ -126,6 +159,7 @@ pangu-embedded ,0 ,any ,RMS_NORM+MUL , 5
phi2 ,0 ,any ,ADD+ADD , 2
phi2 ,0 ,any ,NORM+MUL+ADD , 3
phi3 ,0 ,any ,RMS_NORM+MUL , 5
phimoe ,1 ,any ,MUL+ADD , 4
phimoe ,1 ,any ,RMS_NORM+MUL+ADD , 5
phimoe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
plamo ,0 ,any ,ADD+ADD , 2
@@ -136,6 +170,7 @@ pockettts ,0 ,any ,NORM+MUL+ADD , 5
qwen ,0 ,any ,RMS_NORM+MUL , 5
qwen2 ,0 ,any ,RMS_NORM+MUL , 5
qwen2moe ,1 ,any ,ADD+ADD , 2
qwen2moe ,1 ,any ,MUL+ADD , 4
qwen2moe ,1 ,any ,RMS_NORM+MUL , 5
qwen2moe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS , 2
qwen2vl ,0 ,any ,RMS_NORM+MUL , 5
@@ -144,28 +179,35 @@ qwen35 ,0 ,any ,GATED_DELTA_NET+CPY , 1
qwen35 ,0 ,any ,RMS_NORM+MUL , 8
qwen35moe ,1 ,any ,ADD+ADD , 2
qwen35moe ,1 ,any ,GATED_DELTA_NET+CPY , 1
qwen35moe ,1 ,any ,MUL+ADD , 4
qwen35moe ,1 ,any ,RMS_NORM+MUL , 8
qwen35moe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
qwen3moe ,1 ,any ,MUL+ADD , 4
qwen3moe ,1 ,any ,RMS_NORM+MUL , 9
qwen3moe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
qwen3next ,0 ,any ,ADD+ADD , 2
qwen3next ,0 ,any ,GATED_DELTA_NET+CPY , 1
qwen3next ,0 ,any ,MUL+ADD , 4
qwen3next ,0 ,any ,RMS_NORM+MUL , 8
qwen3next ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
qwen3tts ,0 ,any ,RMS_NORM+MUL , 9
qwen3vl ,0 ,any ,RMS_NORM+MUL , 9
qwen3vlmoe ,1 ,any ,MUL+ADD , 4
qwen3vlmoe ,1 ,any ,RMS_NORM+MUL , 9
qwen3vlmoe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
qwen4exp ,0 ,any ,ADD+ADD+ADD , 5
qwen4exp ,0 ,any ,ADD+ADD+ADD+ADD+ADD+ADD+ADD , 9
qwen4exp ,0 ,any ,GATED_DELTA_NET+CPY , 1
qwen4exp ,0 ,any ,MUL+ADD , 4
qwen4exp ,0 ,any ,RMS_NORM+MUL , 13
qwen4exp ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
refact ,0 ,any ,RMS_NORM+MUL , 5
refact ,0 ,any ,RMS_NORM+MUL , 5
rnd1 ,0 ,any ,MUL+ADD , 4
rnd1 ,0 ,any ,RMS_NORM+MUL , 9
rnd1 ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
seed_oss ,0 ,any ,RMS_NORM+MUL , 5
smallthinker ,0 ,any ,MUL+ADD , 4
smallthinker ,0 ,any ,RMS_NORM+MUL , 5
smollm3 ,0 ,any ,RMS_NORM+MUL , 5
stablelm ,0 ,any ,NORM+MUL , 4
1 # test-fusion baseline for device MTL
2 # arch ,moe ,mode ,label , count
3 arcee ,0 ,any ,RMS_NORM+MUL , 5
4 arctic ,0 ,any ,MUL+ADD , 4
5 arctic ,0 ,any ,RMS_NORM+MUL , 7
6 arctic ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
7 baichuan ,0 ,any ,RMS_NORM+MUL , 5
8 bailingmoe ,1 ,any ,ADD+ADD , 2
9 bailingmoe ,1 ,any ,MUL+ADD , 4
10 bailingmoe ,1 ,any ,RMS_NORM+MUL , 5
11 bailingmoe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS , 2
12 bailingmoe2 ,1 ,any ,ADD+ADD , 1
13 bailingmoe2 ,1 ,any ,MUL+ADD , 2
14 bailingmoe2 ,1 ,any ,RMS_NORM+MUL , 9
15 bailingmoe3 ,1 ,any ,ADD+ADD , 1
16 bailingmoe3 ,1 ,any ,GATED_DELTA_NET+CPY , 1
17 bailingmoe3 ,1 ,any ,MUL+ADD , 2
18 bailingmoe3 ,1 ,any ,RMS_NORM+MUL , 8
19 bloom ,0 ,any ,NORM+MUL+ADD , 6
20 chatglm ,0 ,any ,RMS_NORM+MUL , 5
21 codeshell ,0 ,any ,NORM+MUL+ADD , 5
22 cogvlm ,0 ,any ,RMS_NORM+MUL , 5
23 command-r ,0 ,any ,NORM+MUL , 3
24 dbrx ,0 ,any ,MUL+ADD , 4
25 dbrx ,0 ,any ,NORM+MUL , 5
26 dbrx ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
27 deci ,0 ,any ,RMS_NORM+MUL , 5
28 deepseek ,0 ,any ,ADD+ADD , 1
29 deepseek ,0 ,any ,MUL+ADD , 2
30 deepseek ,0 ,any ,RMS_NORM+MUL , 5
31 deepseek ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS , 1
32 deepseek2 ,0 ,any ,ADD+ADD , 1
33 deepseek2 ,0 ,any ,MUL+ADD , 2
34 deepseek2 ,0 ,any ,RMS_NORM+MUL , 9
35 deepseek32 ,0 ,any ,ADD+ADD , 1
36 deepseek32 ,0 ,any ,MUL+ADD , 2
37 deepseek32 ,0 ,any ,NORM+MUL+ADD , 2
38 deepseek32 ,0 ,any ,RMS_NORM+MUL , 9
39 deepseek4 ,0 ,any ,MUL+ADD , 8
40 deepseek4 ,0 ,any ,RMS_NORM+MUL , 20
41 dots1 ,0 ,any ,ADD+ADD , 1
42 dots1 ,0 ,any ,MUL+ADD , 2
43 dots1 ,0 ,any ,RMS_NORM+MUL , 9
44 dream ,0 ,any ,RMS_NORM+MUL , 5
45 ernie4_5-moe ,1 ,any ,ADD+ADD , 1
46 ernie4_5-moe ,1 ,any ,MUL+ADD , 2
47 ernie4_5-moe ,1 ,any ,RMS_NORM+MUL , 5
48 ernie4_5 ,0 ,any ,RMS_NORM+MUL , 5
49 exaone ,0 ,any ,RMS_NORM+MUL , 5
57 gemma2 ,0 ,any ,RMS_NORM+MUL , 5
58 gemma2 ,0 ,any ,RMS_NORM+MUL+ADD , 4
59 glm-dsa ,0 ,any ,ADD+ADD , 1
60 glm-dsa ,0 ,any ,MUL+ADD , 2
61 glm-dsa ,0 ,any ,NORM+MUL+ADD , 2
62 glm-dsa ,0 ,any ,RMS_NORM+MUL , 9
63 glm4 ,0 ,any ,RMS_NORM+MUL , 5
64 glm4 ,0 ,any ,RMS_NORM+MUL+ADD , 4
65 glm4moe ,1 ,any ,ADD+ADD , 1
66 glm4moe ,1 ,any ,MUL+ADD , 2
67 glm4moe ,1 ,any ,RMS_NORM+MUL , 9
68 gpt-oss ,0 ,any ,MUL+ADD , 4
69 gpt-oss ,0 ,any ,RMS_NORM+MUL , 5
70 gpt2 ,0 ,any ,NORM+MUL+ADD , 5
71 gptneox ,0 ,any ,NORM+MUL+ADD , 5
72 granite ,0 ,any ,RMS_NORM+MUL , 5
73 granite ,0 ,any ,MUL+ADD , 4
74 granite ,0 ,any ,RMS_NORM+MUL , 5
75 granite ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
76 granitehybrid ,0 ,any ,RMS_NORM+MUL , 6
77 granitemoe ,1 ,any ,RMS_NORM+MUL , 5
78 granitemoe ,1 ,any ,MUL+ADD , 4
79 granitemoe ,1 ,any ,RMS_NORM+MUL , 5
80 granitemoe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
81 grok ,0 ,any ,MUL+ADD , 4
82 grok ,0 ,any ,RMS_NORM+MUL , 5
83 grok ,0 ,any ,RMS_NORM+MUL+ADD , 4
84 grok ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
85 grovemoe ,1 ,any ,ADD+ADD , 2
86 grovemoe ,1 ,any ,MUL+ADD , 8
87 grovemoe ,1 ,any ,RMS_NORM+MUL , 9
88 hunyuan-dense ,0 ,any ,RMS_NORM+MUL , 9
89 hunyuan-moe ,1 ,any ,ADD+ADD , 2
90 hunyuan-moe ,1 ,any ,MUL+ADD , 4
91 hunyuan-moe ,1 ,any ,RMS_NORM+MUL , 9
92 hunyuan-moe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
93 hunyuan_vl ,0 ,any ,RMS_NORM+MUL , 9
94 hy_v3 ,0 ,any ,ADD+ADD , 2
95 hy_v3 ,0 ,any ,MUL+ADD , 4
96 hy_v3 ,0 ,any ,RMS_NORM+MUL , 9
97 hy_v4 ,0 ,any ,MUL+ADD , 2
98 hy_v4 ,0 ,any ,NORM+MUL+ADD , 1
99 hy_v4 ,0 ,any ,RMS_NORM+MUL , 9
100 internlm2 ,0 ,any ,RMS_NORM+MUL , 5
102 jais2 ,0 ,any ,NORM+MUL+ADD , 5
103 jamba ,0 ,any ,RMS_NORM+MUL , 8
104 kimi-k3 ,0 ,any ,GATED_DELTA_NET+CPY , 1
105 kimi-k3 ,0 ,any ,MUL+ADD , 2
106 kimi-k3 ,0 ,any ,RMS_NORM+MUL , 17
107 kimi-linear ,0 ,any ,ADD+ADD , 1
108 kimi-linear ,0 ,any ,GATED_DELTA_NET+CPY , 1
109 kimi-linear ,0 ,any ,MUL+ADD , 2
110 kimi-linear ,0 ,any ,RMS_NORM+MUL , 7
111 lfm2 ,0 ,any ,RMS_NORM+MUL , 7
112 lfm2moe ,1 ,any ,MUL+ADD , 2
113 lfm2moe ,1 ,any ,RMS_NORM+MUL , 7
114 llada ,0 ,any ,RMS_NORM+MUL , 5
115 llada-moe ,1 ,any ,MUL+ADD , 4
116 llada-moe ,1 ,any ,RMS_NORM+MUL , 9
117 llada-moe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS , 2
118 llama ,0 ,any ,RMS_NORM+MUL , 5
119 llama ,0 ,any ,MUL+ADD , 4
120 llama ,0 ,any ,RMS_NORM+MUL , 5
121 llama ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
122 llama4 ,0 ,any ,ADD+ADD , 2
125 mamba ,0 ,any ,RMS_NORM+MUL , 3
126 mamba2 ,0 ,any ,RMS_NORM+MUL , 5
127 minicpm ,0 ,any ,RMS_NORM+MUL , 5
128 minicpm ,0 ,any ,MUL+ADD , 4
129 minicpm ,0 ,any ,RMS_NORM+MUL , 5
130 minicpm ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
131 minicpm3 ,0 ,any ,RMS_NORM+MUL , 9
132 minimax-01 ,0 ,any ,MUL+ADD , 4
133 minimax-01 ,0 ,any ,RMS_NORM+MUL , 6
134 minimax-01 ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
135 minimax-m2 ,0 ,any ,MUL+ADD , 4
136 minimax-m2 ,0 ,any ,RMS_NORM+MUL , 9
137 minimax-m3 ,0 ,any ,ADD+ADD , 1
138 minimax-m3 ,0 ,any ,MUL+ADD , 2
139 minimax-m3 ,0 ,any ,RMS_NORM+MUL , 11
140 mistral3 ,0 ,any ,RMS_NORM+MUL , 5
141 mistral3 ,0 ,any ,MUL+ADD , 4
142 mistral3 ,0 ,any ,RMS_NORM+MUL , 5
143 mistral3 ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
144 mistral4 ,0 ,any ,ADD+ADD , 1
145 mistral4 ,0 ,any ,MUL+ADD , 2
146 mistral4 ,0 ,any ,RMS_NORM+MUL , 9
147 mpt ,0 ,any ,NORM+MUL+ADD , 5
148 nanbeige ,0 ,any ,RMS_NORM+MUL , 5
149 nemotron ,0 ,any ,NORM+MUL+ADD , 5
150 nemotron_h ,0 ,any ,RMS_NORM+MUL , 5
151 nemotron_h_moe ,1 ,any ,RMS_NORM+MUL , 5
152 olmoe ,1 ,any ,MUL+ADD , 4
153 olmoe ,1 ,any ,RMS_NORM+MUL , 9
154 olmoe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS , 2
155 openelm ,0 ,any ,RMS_NORM+MUL , 9
159 phi2 ,0 ,any ,ADD+ADD , 2
160 phi2 ,0 ,any ,NORM+MUL+ADD , 3
161 phi3 ,0 ,any ,RMS_NORM+MUL , 5
162 phimoe ,1 ,any ,MUL+ADD , 4
163 phimoe ,1 ,any ,RMS_NORM+MUL+ADD , 5
164 phimoe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
165 plamo ,0 ,any ,ADD+ADD , 2
170 qwen ,0 ,any ,RMS_NORM+MUL , 5
171 qwen2 ,0 ,any ,RMS_NORM+MUL , 5
172 qwen2moe ,1 ,any ,ADD+ADD , 2
173 qwen2moe ,1 ,any ,MUL+ADD , 4
174 qwen2moe ,1 ,any ,RMS_NORM+MUL , 5
175 qwen2moe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS , 2
176 qwen2vl ,0 ,any ,RMS_NORM+MUL , 5
179 qwen35 ,0 ,any ,RMS_NORM+MUL , 8
180 qwen35moe ,1 ,any ,ADD+ADD , 2
181 qwen35moe ,1 ,any ,GATED_DELTA_NET+CPY , 1
182 qwen35moe ,1 ,any ,MUL+ADD , 4
183 qwen35moe ,1 ,any ,RMS_NORM+MUL , 8
184 qwen35moe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
185 qwen3moe ,1 ,any ,MUL+ADD , 4
186 qwen3moe ,1 ,any ,RMS_NORM+MUL , 9
187 qwen3moe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
188 qwen3next ,0 ,any ,ADD+ADD , 2
189 qwen3next ,0 ,any ,GATED_DELTA_NET+CPY , 1
190 qwen3next ,0 ,any ,MUL+ADD , 4
191 qwen3next ,0 ,any ,RMS_NORM+MUL , 8
192 qwen3next ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
193 qwen3tts ,0 ,any ,RMS_NORM+MUL , 9
194 qwen3vl ,0 ,any ,RMS_NORM+MUL , 9
195 qwen3vlmoe ,1 ,any ,MUL+ADD , 4
196 qwen3vlmoe ,1 ,any ,RMS_NORM+MUL , 9
197 qwen3vlmoe ,1 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
198 qwen4exp ,0 ,any ,ADD+ADD+ADD , 5
199 qwen4exp ,0 ,any ,ADD+ADD+ADD+ADD+ADD+ADD+ADD , 9
200 qwen4exp ,0 ,any ,GATED_DELTA_NET+CPY , 1
201 qwen4exp ,0 ,any ,MUL+ADD , 4
202 qwen4exp ,0 ,any ,RMS_NORM+MUL , 13
203 qwen4exp ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
204 refact ,0 ,any ,RMS_NORM+MUL , 5
205 refact ,0 ,any ,RMS_NORM+MUL , 5
206 rnd1 ,0 ,any ,MUL+ADD , 4
207 rnd1 ,0 ,any ,RMS_NORM+MUL , 9
208 rnd1 ,0 ,any ,SOFT_MAX+ARGSORT+GET_ROWS+SUM_ROWS+CLAMP+DIV, 2
209 seed_oss ,0 ,any ,RMS_NORM+MUL , 5
210 smallthinker ,0 ,any ,MUL+ADD , 4
211 smallthinker ,0 ,any ,RMS_NORM+MUL , 5
212 smollm3 ,0 ,any ,RMS_NORM+MUL , 5
213 stablelm ,0 ,any ,NORM+MUL , 4
+1 -1
View File
@@ -339,7 +339,7 @@ static gguf_context_ptr get_gguf_ctx(const llm_arch arch, const bool moe) {
ms.add_kv(LLM_KV_EXPERT_LATENT_LENGTH, n_ff);
ms.add_kv(LLM_KV_INTERLEAVE_MOE_LAYER_STEP, uint32_t(2));
ms.add_kv(LLM_KV_EXPERT_COUNT, uint32_t(2));
ms.add_kv(LLM_KV_EXPERT_USED_COUNT, uint32_t(1));
ms.add_kv(LLM_KV_EXPERT_USED_COUNT, uint32_t(2));
ms.add_kv(LLM_KV_EXPERT_SHARED_COUNT, uint32_t(1));
ms.add_kv(LLM_KV_EXPERT_GATING_FUNC, arch == LLM_ARCH_DEEPSEEK4 ? uint32_t(4) : uint32_t(2)); // sqrtsoftplus : sigmoid
ms.add_kv(LLM_KV_EXPERT_GROUP_SCALE, 1.0f);