Hugging Face's logo

Cadododoom
/
Agents-A1-TextOnly-FP4

Safetensors
qwen3_5_moe_text
modelopt
fp4
vllm
mamba
text-only
tensor-parallel
vocabulary-pruned
8-bit precision
Model card Files
xet
 Community
Cadododoom commited on
Commit
1ca9f2f
·
verified ·
1 Parent(s): e5beb05

Upload hf_quant_config.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. hf_quant_config.json +186 -0
hf_quant_config.json ADDED
@@ -0,0 +1,186 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "producer": {
3
+ "name": "modelopt",
4
+ "version": "0.46.0.dev135+gb0ee95363"
5
+ },
6
+ "quantization": {
7
+ "quant_algo": "W4A16_NVFP4",
8
+ "kv_cache_quant_algo": null,
9
+ "group_size": 16,
10
+ "exclude_modules": [
11
+ "lm_head",
12
+ "model.language_model.embed_tokens",
13
+ "model.language_model.layers.0.linear_attn.conv1d",
14
+ "model.language_model.layers.0.linear_attn.in_proj_a",
15
+ "model.language_model.layers.0.linear_attn.in_proj_b",
16
+ "model.language_model.layers.0.mlp.gate",
17
+ "model.language_model.layers.0.mlp.shared_expert_gate",
18
+ "model.language_model.layers.1.linear_attn.conv1d",
19
+ "model.language_model.layers.1.linear_attn.in_proj_a",
20
+ "model.language_model.layers.1.linear_attn.in_proj_b",
21
+ "model.language_model.layers.1.mlp.gate",
22
+ "model.language_model.layers.1.mlp.shared_expert_gate",
23
+ "model.language_model.layers.10.linear_attn.conv1d",
24
+ "model.language_model.layers.10.linear_attn.in_proj_a",
25
+ "model.language_model.layers.10.linear_attn.in_proj_b",
26
+ "model.language_model.layers.10.mlp.gate",
27
+ "model.language_model.layers.10.mlp.shared_expert_gate",
28
+ "model.language_model.layers.11.mlp.gate",
29
+ "model.language_model.layers.11.mlp.shared_expert_gate",
30
+ "model.language_model.layers.12.linear_attn.conv1d",
31
+ "model.language_model.layers.12.linear_attn.in_proj_a",
32
+ "model.language_model.layers.12.linear_attn.in_proj_b",
33
+ "model.language_model.layers.12.mlp.gate",
34
+ "model.language_model.layers.12.mlp.shared_expert_gate",
35
+ "model.language_model.layers.13.linear_attn.conv1d",
36
+ "model.language_model.layers.13.linear_attn.in_proj_a",
37
+ "model.language_model.layers.13.linear_attn.in_proj_b",
38
+ "model.language_model.layers.13.mlp.gate",
39
+ "model.language_model.layers.13.mlp.shared_expert_gate",
40
+ "model.language_model.layers.14.linear_attn.conv1d",
41
+ "model.language_model.layers.14.linear_attn.in_proj_a",
42
+ "model.language_model.layers.14.linear_attn.in_proj_b",
43
+ "model.language_model.layers.14.mlp.gate",
44
+ "model.language_model.layers.14.mlp.shared_expert_gate",
45
+ "model.language_model.layers.15.mlp.gate",
46
+ "model.language_model.layers.15.mlp.shared_expert_gate",
47
+ "model.language_model.layers.16.linear_attn.conv1d",
48
+ "model.language_model.layers.16.linear_attn.in_proj_a",
49
+ "model.language_model.layers.16.linear_attn.in_proj_b",
50
+ "model.language_model.layers.16.mlp.gate",
51
+ "model.language_model.layers.16.mlp.shared_expert_gate",
52
+ "model.language_model.layers.17.linear_attn.conv1d",
53
+ "model.language_model.layers.17.linear_attn.in_proj_a",
54
+ "model.language_model.layers.17.linear_attn.in_proj_b",
55
+ "model.language_model.layers.17.mlp.gate",
56
+ "model.language_model.layers.17.mlp.shared_expert_gate",
57
+ "model.language_model.layers.18.linear_attn.conv1d",
58
+ "model.language_model.layers.18.linear_attn.in_proj_a",
59
+ "model.language_model.layers.18.linear_attn.in_proj_b",
60
+ "model.language_model.layers.18.mlp.gate",
61
+ "model.language_model.layers.18.mlp.shared_expert_gate",
62
+ "model.language_model.layers.19.mlp.gate",
63
+ "model.language_model.layers.19.mlp.shared_expert_gate",
64
+ "model.language_model.layers.2.linear_attn.conv1d",
65
+ "model.language_model.layers.2.linear_attn.in_proj_a",
66
+ "model.language_model.layers.2.linear_attn.in_proj_b",
67
+ "model.language_model.layers.2.mlp.gate",
68
+ "model.language_model.layers.2.mlp.shared_expert_gate",
69
+ "model.language_model.layers.20.linear_attn.conv1d",
70
+ "model.language_model.layers.20.linear_attn.in_proj_a",
71
+ "model.language_model.layers.20.linear_attn.in_proj_b",
72
+ "model.language_model.layers.20.mlp.gate",
73
+ "model.language_model.layers.20.mlp.shared_expert_gate",
74
+ "model.language_model.layers.21.linear_attn.conv1d",
75
+ "model.language_model.layers.21.linear_attn.in_proj_a",
76
+ "model.language_model.layers.21.linear_attn.in_proj_b",
77
+ "model.language_model.layers.21.mlp.gate",
78
+ "model.language_model.layers.21.mlp.shared_expert_gate",
79
+ "model.language_model.layers.22.linear_attn.conv1d",
80
+ "model.language_model.layers.22.linear_attn.in_proj_a",
81
+ "model.language_model.layers.22.linear_attn.in_proj_b",
82
+ "model.language_model.layers.22.mlp.gate",
83
+ "model.language_model.layers.22.mlp.shared_expert_gate",
84
+ "model.language_model.layers.23.mlp.gate",
85
+ "model.language_model.layers.23.mlp.shared_expert_gate",
86
+ "model.language_model.layers.24.linear_attn.conv1d",
87
+ "model.language_model.layers.24.linear_attn.in_proj_a",
88
+ "model.language_model.layers.24.linear_attn.in_proj_b",
89
+ "model.language_model.layers.24.mlp.gate",
90
+ "model.language_model.layers.24.mlp.shared_expert_gate",
91
+ "model.language_model.layers.25.linear_attn.conv1d",
92
+ "model.language_model.layers.25.linear_attn.in_proj_a",
93
+ "model.language_model.layers.25.linear_attn.in_proj_b",
94
+ "model.language_model.layers.25.mlp.gate",
95
+ "model.language_model.layers.25.mlp.shared_expert_gate",
96
+ "model.language_model.layers.26.linear_attn.conv1d",
97
+ "model.language_model.layers.26.linear_attn.in_proj_a",
98
+ "model.language_model.layers.26.linear_attn.in_proj_b",
99
+ "model.language_model.layers.26.mlp.gate",
100
+ "model.language_model.layers.26.mlp.shared_expert_gate",
101
+ "model.language_model.layers.27.mlp.gate",
102
+ "model.language_model.layers.27.mlp.shared_expert_gate",
103
+ "model.language_model.layers.28.linear_attn.conv1d",
104
+ "model.language_model.layers.28.linear_attn.in_proj_a",
105
+ "model.language_model.layers.28.linear_attn.in_proj_b",
106
+ "model.language_model.layers.28.mlp.gate",
107
+ "model.language_model.layers.28.mlp.shared_expert_gate",
108
+ "model.language_model.layers.29.linear_attn.conv1d",
109
+ "model.language_model.layers.29.linear_attn.in_proj_a",
110
+ "model.language_model.layers.29.linear_attn.in_proj_b",
111
+ "model.language_model.layers.29.mlp.gate",
112
+ "model.language_model.layers.29.mlp.shared_expert_gate",
113
+ "model.language_model.layers.3.mlp.gate",
114
+ "model.language_model.layers.3.mlp.shared_expert_gate",
115
+ "model.language_model.layers.30.linear_attn.conv1d",
116
+ "model.language_model.layers.30.linear_attn.in_proj_a",
117
+ "model.language_model.layers.30.linear_attn.in_proj_b",
118
+ "model.language_model.layers.30.mlp.gate",
119
+ "model.language_model.layers.30.mlp.shared_expert_gate",
120
+ "model.language_model.layers.31.mlp.gate",
121
+ "model.language_model.layers.31.mlp.shared_expert_gate",
122
+ "model.language_model.layers.32.linear_attn.conv1d",
123
+ "model.language_model.layers.32.linear_attn.in_proj_a",
124
+ "model.language_model.layers.32.linear_attn.in_proj_b",
125
+ "model.language_model.layers.32.mlp.gate",
126
+ "model.language_model.layers.32.mlp.shared_expert_gate",
127
+ "model.language_model.layers.33.linear_attn.conv1d",
128
+ "model.language_model.layers.33.linear_attn.in_proj_a",
129
+ "model.language_model.layers.33.linear_attn.in_proj_b",
130
+ "model.language_model.layers.33.mlp.gate",
131
+ "model.language_model.layers.33.mlp.shared_expert_gate",
132
+ "model.language_model.layers.34.linear_attn.conv1d",
133
+ "model.language_model.layers.34.linear_attn.in_proj_a",
134
+ "model.language_model.layers.34.linear_attn.in_proj_b",
135
+ "model.language_model.layers.34.mlp.gate",
136
+ "model.language_model.layers.34.mlp.shared_expert_gate",
137
+ "model.language_model.layers.35.mlp.gate",
138
+ "model.language_model.layers.35.mlp.shared_expert_gate",
139
+ "model.language_model.layers.36.linear_attn.conv1d",
140
+ "model.language_model.layers.36.linear_attn.in_proj_a",
141
+ "model.language_model.layers.36.linear_attn.in_proj_b",
142
+ "model.language_model.layers.36.mlp.gate",
143
+ "model.language_model.layers.36.mlp.shared_expert_gate",
144
+ "model.language_model.layers.37.linear_attn.conv1d",
145
+ "model.language_model.layers.37.linear_attn.in_proj_a",
146
+ "model.language_model.layers.37.linear_attn.in_proj_b",
147
+ "model.language_model.layers.37.mlp.gate",
148
+ "model.language_model.layers.37.mlp.shared_expert_gate",
149
+ "model.language_model.layers.38.linear_attn.conv1d",
150
+ "model.language_model.layers.38.linear_attn.in_proj_a",
151
+ "model.language_model.layers.38.linear_attn.in_proj_b",
152
+ "model.language_model.layers.38.mlp.gate",
153
+ "model.language_model.layers.38.mlp.shared_expert_gate",
154
+ "model.language_model.layers.39.mlp.gate",
155
+ "model.language_model.layers.39.mlp.shared_expert_gate",
156
+ "model.language_model.layers.4.linear_attn.conv1d",
157
+ "model.language_model.layers.4.linear_attn.in_proj_a",
158
+ "model.language_model.layers.4.linear_attn.in_proj_b",
159
+ "model.language_model.layers.4.mlp.gate",
160
+ "model.language_model.layers.4.mlp.shared_expert_gate",
161
+ "model.language_model.layers.5.linear_attn.conv1d",
162
+ "model.language_model.layers.5.linear_attn.in_proj_a",
163
+ "model.language_model.layers.5.linear_attn.in_proj_b",
164
+ "model.language_model.layers.5.mlp.gate",
165
+ "model.language_model.layers.5.mlp.shared_expert_gate",
166
+ "model.language_model.layers.6.linear_attn.conv1d",
167
+ "model.language_model.layers.6.linear_attn.in_proj_a",
168
+ "model.language_model.layers.6.linear_attn.in_proj_b",
169
+ "model.language_model.layers.6.mlp.gate",
170
+ "model.language_model.layers.6.mlp.shared_expert_gate",
171
+ "model.language_model.layers.7.mlp.gate",
172
+ "model.language_model.layers.7.mlp.shared_expert_gate",
173
+ "model.language_model.layers.8.linear_attn.conv1d",
174
+ "model.language_model.layers.8.linear_attn.in_proj_a",
175
+ "model.language_model.layers.8.linear_attn.in_proj_b",
176
+ "model.language_model.layers.8.mlp.gate",
177
+ "model.language_model.layers.8.mlp.shared_expert_gate",
178
+ "model.language_model.layers.9.linear_attn.conv1d",
179
+ "model.language_model.layers.9.linear_attn.in_proj_a",
180
+ "model.language_model.layers.9.linear_attn.in_proj_b",
181
+ "model.language_model.layers.9.mlp.gate",
182
+ "model.language_model.layers.9.mlp.shared_expert_gate",
183
+ "model.visual*"
184
+ ]
185
+ }
186
+ }