Hugging Face's logo

AngelSlim
/
Qwen2.5-VL-72B-Instruct-FP8-Dynamic

Safetensors
qwen2_5_vl
fp8
Model card Files
xet
Community
woodchen7 commited on
Commit
8afcdfe
·
verified ·
1 Parent(s): 6f33f50

Upload config.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. config.json +198 -0
config.json ADDED
@@ -0,0 +1,198 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2_5_VLForConditionalGeneration"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "bos_token_id": 151643,
7
+ "eos_token_id": 151645,
8
+ "hidden_act": "silu",
9
+ "hidden_size": 8192,
10
+ "image_token_id": 151655,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 29568,
13
+ "max_position_embeddings": 128000,
14
+ "max_window_layers": 80,
15
+ "model_type": "qwen2_5_vl",
16
+ "num_attention_heads": 64,
17
+ "num_hidden_layers": 80,
18
+ "num_key_value_heads": 8,
19
+ "quantization_config": {
20
+ "activation_scheme": "dynamic",
21
+ "ignored_layers": [
22
+ "model.visual.patch_embed.proj",
23
+ "model.lm_head",
24
+ "model.language_model.embed_tokens",
25
+ "model.visual.merger.mlp.0",
26
+ "model.visual.merger.mlp.2",
27
+ "lm_head"
28
+ ],
29
+ "quant_method": "fp8"
30
+ },
31
+ "rms_norm_eps": 1e-06,
32
+ "rope_scaling": {
33
+ "mrope_section": [
34
+ 16,
35
+ 24,
36
+ 24
37
+ ],
38
+ "rope_type": "default",
39
+ "type": "default"
40
+ },
41
+ "rope_theta": 1000000.0,
42
+ "sliding_window": 32768,
43
+ "text_config": {
44
+ "architectures": [
45
+ "Qwen2_5_VLForConditionalGeneration"
46
+ ],
47
+ "attention_dropout": 0.0,
48
+ "bos_token_id": 151643,
49
+ "eos_token_id": 151645,
50
+ "hidden_act": "silu",
51
+ "hidden_size": 8192,
52
+ "image_token_id": null,
53
+ "initializer_range": 0.02,
54
+ "intermediate_size": 29568,
55
+ "layer_types": [
56
+ "full_attention",
57
+ "full_attention",
58
+ "full_attention",
59
+ "full_attention",
60
+ "full_attention",
61
+ "full_attention",
62
+ "full_attention",
63
+ "full_attention",
64
+ "full_attention",
65
+ "full_attention",
66
+ "full_attention",
67
+ "full_attention",
68
+ "full_attention",
69
+ "full_attention",
70
+ "full_attention",
71
+ "full_attention",
72
+ "full_attention",
73
+ "full_attention",
74
+ "full_attention",
75
+ "full_attention",
76
+ "full_attention",
77
+ "full_attention",
78
+ "full_attention",
79
+ "full_attention",
80
+ "full_attention",
81
+ "full_attention",
82
+ "full_attention",
83
+ "full_attention",
84
+ "full_attention",
85
+ "full_attention",
86
+ "full_attention",
87
+ "full_attention",
88
+ "full_attention",
89
+ "full_attention",
90
+ "full_attention",
91
+ "full_attention",
92
+ "full_attention",
93
+ "full_attention",
94
+ "full_attention",
95
+ "full_attention",
96
+ "full_attention",
97
+ "full_attention",
98
+ "full_attention",
99
+ "full_attention",
100
+ "full_attention",
101
+ "full_attention",
102
+ "full_attention",
103
+ "full_attention",
104
+ "full_attention",
105
+ "full_attention",
106
+ "full_attention",
107
+ "full_attention",
108
+ "full_attention",
109
+ "full_attention",
110
+ "full_attention",
111
+ "full_attention",
112
+ "full_attention",
113
+ "full_attention",
114
+ "full_attention",
115
+ "full_attention",
116
+ "full_attention",
117
+ "full_attention",
118
+ "full_attention",
119
+ "full_attention",
120
+ "full_attention",
121
+ "full_attention",
122
+ "full_attention",
123
+ "full_attention",
124
+ "full_attention",
125
+ "full_attention",
126
+ "full_attention",
127
+ "full_attention",
128
+ "full_attention",
129
+ "full_attention",
130
+ "full_attention",
131
+ "full_attention",
132
+ "full_attention",
133
+ "full_attention",
134
+ "full_attention",
135
+ "full_attention"
136
+ ],
137
+ "max_position_embeddings": 128000,
138
+ "max_window_layers": 80,
139
+ "model_type": "qwen2_5_vl_text",
140
+ "num_attention_heads": 64,
141
+ "num_hidden_layers": 80,
142
+ "num_key_value_heads": 8,
143
+ "rms_norm_eps": 1e-06,
144
+ "rope_scaling": {
145
+ "mrope_section": [
146
+ 16,
147
+ 24,
148
+ 24
149
+ ],
150
+ "rope_type": "default",
151
+ "type": "default"
152
+ },
153
+ "rope_theta": 1000000.0,
154
+ "sliding_window": null,
155
+ "torch_dtype": "bfloat16",
156
+ "use_cache": true,
157
+ "use_sliding_window": false,
158
+ "video_token_id": null,
159
+ "vision_end_token_id": 151653,
160
+ "vision_start_token_id": 151652,
161
+ "vision_token_id": 151654,
162
+ "vocab_size": 152064
163
+ },
164
+ "tie_word_embeddings": false,
165
+ "torch_dtype": "bfloat16",
166
+ "transformers_version": "4.53.3",
167
+ "use_cache": false,
168
+ "use_sliding_window": false,
169
+ "video_token_id": 151656,
170
+ "vision_config": {
171
+ "depth": 32,
172
+ "fullatt_block_indexes": [
173
+ 7,
174
+ 15,
175
+ 23,
176
+ 31
177
+ ],
178
+ "hidden_act": "silu",
179
+ "hidden_size": 1280,
180
+ "in_channels": 3,
181
+ "in_chans": 3,
182
+ "initializer_range": 0.02,
183
+ "intermediate_size": 3456,
184
+ "model_type": "qwen2_5_vl",
185
+ "num_heads": 16,
186
+ "out_hidden_size": 8192,
187
+ "patch_size": 14,
188
+ "spatial_merge_size": 2,
189
+ "spatial_patch_size": 14,
190
+ "temporal_patch_size": 2,
191
+ "tokens_per_second": 2,
192
+ "window_size": 112
193
+ },
194
+ "vision_end_token_id": 151653,
195
+ "vision_start_token_id": 151652,
196
+ "vision_token_id": 151654,
197
+ "vocab_size": 152064
198
+ }