Hugging Face's logo

zai-org
/
GLM-4.5V-FP8

Image-Text-to-Text
Transformers
Safetensors
Chinese
English
glm4v_moe
conversational
compressed-tensors
Model card Files
xet
Community
3
zR commited on
Commit
56e1d12
·
verified ·
1 Parent(s): 8419b5d

Add files using upload-large-folder tool

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +1 -0
  2. chat_template.jinja +118 -0
  3. config.json +627 -0
  4. generation_config.json +14 -0
  5. model-00001-of-00046.safetensors +3 -0
  6. model-00002-of-00046.safetensors +3 -0
  7. model-00003-of-00046.safetensors +3 -0
  8. model-00004-of-00046.safetensors +3 -0
  9. model-00005-of-00046.safetensors +3 -0
  10. model-00006-of-00046.safetensors +3 -0
  11. model-00007-of-00046.safetensors +3 -0
  12. model-00008-of-00046.safetensors +3 -0
  13. model-00009-of-00046.safetensors +3 -0
  14. model-00010-of-00046.safetensors +3 -0
  15. model-00011-of-00046.safetensors +3 -0
  16. model-00012-of-00046.safetensors +3 -0
  17. model-00013-of-00046.safetensors +3 -0
  18. model-00014-of-00046.safetensors +3 -0
  19. model-00015-of-00046.safetensors +3 -0
  20. model-00016-of-00046.safetensors +3 -0
  21. model-00017-of-00046.safetensors +3 -0
  22. model-00018-of-00046.safetensors +3 -0
  23. model-00019-of-00046.safetensors +3 -0
  24. model-00020-of-00046.safetensors +3 -0
  25. model-00021-of-00046.safetensors +3 -0
  26. model-00022-of-00046.safetensors +3 -0
  27. model-00023-of-00046.safetensors +3 -0
  28. model-00024-of-00046.safetensors +3 -0
  29. model-00025-of-00046.safetensors +3 -0
  30. model-00026-of-00046.safetensors +3 -0
  31. model-00027-of-00046.safetensors +3 -0
  32. model-00028-of-00046.safetensors +3 -0
  33. model-00029-of-00046.safetensors +3 -0
  34. model-00030-of-00046.safetensors +3 -0
  35. model-00031-of-00046.safetensors +3 -0
  36. model-00032-of-00046.safetensors +3 -0
  37. model-00033-of-00046.safetensors +3 -0
  38. model-00034-of-00046.safetensors +3 -0
  39. model-00036-of-00046.safetensors +3 -0
  40. model-00038-of-00046.safetensors +3 -0
  41. model-00039-of-00046.safetensors +3 -0
  42. model-00041-of-00046.safetensors +3 -0
  43. model-00042-of-00046.safetensors +3 -0
  44. model-00043-of-00046.safetensors +3 -0
  45. model-00045-of-00046.safetensors +3 -0
  46. model-00046-of-00046.safetensors +3 -0
  47. model.safetensors.index.json +0 -0
  48. preprocessor_config.json +11 -0
  49. tokenizer.json +3 -0
  50. tokenizer_config.json +325 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
chat_template.jinja ADDED
@@ -0,0 +1,118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [gMASK]<sop>
2
+ {%- if tools -%}
3
+ <|system|>
4
+ # Tools
5
+
6
+ You may call one or more functions to assist with the user query.
7
+
8
+ You are provided with function signatures within <tools></tools> XML tags:
9
+ <tools>
10
+ {% for tool in tools %}
11
+ {{ tool | tojson(ensure_ascii=False) }}
12
+ {% endfor %}
13
+ </tools>
14
+
15
+ For each function call, output the function name and arguments within the following XML format:
16
+ <tool_call>{function-name}
17
+ <arg_key>{arg-key-1}</arg_key>
18
+ <arg_value>{arg-value-1}</arg_value>
19
+ <arg_key>{arg-key-2}</arg_key>
20
+ <arg_value>{arg-value-2}</arg_value>
21
+ ...
22
+ </tool_call>{%- endif -%}
23
+ {%- macro visible_text(content) -%}
24
+ {%- if content is string -%}
25
+ {{- content }}
26
+ {%- elif content is iterable and content is not mapping -%}
27
+ {%- for item in content -%}
28
+ {%- if item is mapping and item.type == 'text' -%}
29
+ {{- item.text }}
30
+ {%- elif item is mapping and (item.type == 'image' or 'image' in item) -%}
31
+ <|begin_of_image|><|image|><|end_of_image|>
32
+ {%- elif item is mapping and (item.type == 'video' or 'video' in item) -%}
33
+ <|begin_of_video|><|video|><|end_of_video|>
34
+ {%- elif item is string -%}
35
+ {{- item }}
36
+ {%- endif -%}
37
+ {%- endfor -%}
38
+ {%- else -%}
39
+ {{- content }}
40
+ {%- endif -%}
41
+ {%- endmacro -%}
42
+ {%- set ns = namespace(last_user_index=-1) %}
43
+ {%- for m in messages %}
44
+ {%- if m.role == 'user' %}
45
+ {% set ns.last_user_index = loop.index0 -%}
46
+ {%- endif %}
47
+ {%- endfor %}
48
+ {% for m in messages %}
49
+ {%- if m.role == 'user' -%}<|user|>
50
+ {% if m.content is string %}
51
+ {{ m.content }}
52
+ {%- else %}
53
+ {%- for item in m.content %}
54
+ {% if item.type == 'video' or 'video' in item %}
55
+ <|begin_of_video|><|video|><|end_of_video|>{% elif item.type == 'image' or 'image' in item %}
56
+ <|begin_of_image|><|image|><|end_of_image|>{% elif item.type == 'text' %}
57
+ {{ item.text }}
58
+ {%- endif %}
59
+ {%- endfor %}
60
+ {%- endif %}
61
+ {{- '/nothink' if (enable_thinking is defined and not enable_thinking and not visible_text(m.content).endswith("/nothink")) else '' -}}
62
+ {%- elif m.role == 'assistant' -%}
63
+ <|assistant|>
64
+ {%- set reasoning_content = '' %}
65
+ {%- set content = visible_text(m.content) %}
66
+ {%- if m.reasoning_content is string %}
67
+ {%- set reasoning_content = m.reasoning_content %}
68
+ {%- else %}
69
+ {%- if '</think>' in content %}
70
+ {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
71
+ {%- set content = content.split('</think>')[-1].lstrip('\n') %}
72
+ {%- endif %}
73
+ {%- endif %}
74
+ {%- if loop.index0 > ns.last_user_index and reasoning_content -%}
75
+ {{ '\n<think>' + reasoning_content.strip() + '</think>'}}
76
+ {%- else -%}
77
+ {{ '\n<think></think>' }}
78
+ {%- endif -%}
79
+ {%- if content.strip() -%}
80
+ {{ '\n' + content.strip() }}
81
+ {%- endif -%}
82
+ {% if m.tool_calls %}
83
+ {% for tc in m.tool_calls %}
84
+ {%- if tc.function %}
85
+ {%- set tc = tc.function %}
86
+ {%- endif %}
87
+ {{ '\n<tool_call>' + tc.name }}
88
+ {% set _args = tc.arguments %}
89
+ {% for k, v in _args.items() %}
90
+ <arg_key>{{ k }}</arg_key>
91
+ <arg_value>{{ v | tojson(ensure_ascii=False) if v is not string else v }}</arg_value>
92
+ {% endfor %}
93
+ </tool_call>{% endfor %}
94
+ {% endif %}
95
+ {%- elif m.role == 'tool' -%}
96
+ {%- if m.content is string -%}
97
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
98
+ {{- '<|observation|>' }}
99
+ {%- endif %}
100
+ {{- '\n<tool_response>\n' }}
101
+ {{- m.content }}
102
+ {{- '\n</tool_response>' }}
103
+ {%- else -%}
104
+ <|observation|>{% for tr in m.content %}
105
+
106
+ <tool_response>
107
+ {{ tr.output if tr.output is defined else tr }}
108
+ </tool_response>{% endfor -%}
109
+ {% endif -%}
110
+ {%- elif m.role == 'system' -%}
111
+ <|system|>
112
+ {{ visible_text(m.content) }}
113
+ {%- endif -%}
114
+ {%- endfor -%}
115
+ {%- if add_generation_prompt -%}
116
+ <|assistant|>
117
+ {{'<think></think>\n' if (enable_thinking is defined and not enable_thinking) else ''}}
118
+ {%- endif -%}
config.json ADDED
@@ -0,0 +1,627 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Glm4vMoeForConditionalGeneration"
4
+ ],
5
+ "model_type": "glm4v_moe",
6
+ "rope_scaling": {
7
+ "rope_type": "default",
8
+ "mrope_section": [
9
+ 8,
10
+ 12,
11
+ 12
12
+ ]
13
+ },
14
+ "text_config": {
15
+ "pad_token_id": 151329,
16
+ "vocab_size": 151552,
17
+ "eos_token_id": [
18
+ 151329,
19
+ 151336,
20
+ 151338
21
+ ],
22
+ "image_end_token_id": 151340,
23
+ "image_start_token_id": 151339,
24
+ "image_token_id": 151363,
25
+ "head_dim": 128,
26
+ "attention_bias": true,
27
+ "attention_dropout": 0.0,
28
+ "first_k_dense_replace": 1,
29
+ "hidden_act": "silu",
30
+ "hidden_size": 4096,
31
+ "initializer_range": 0.02,
32
+ "intermediate_size": 10944,
33
+ "max_position_embeddings": 65536,
34
+ "model_type": "glm4v_moe_text",
35
+ "moe_intermediate_size": 1408,
36
+ "n_group": 1,
37
+ "n_routed_experts": 128,
38
+ "n_shared_experts": 1,
39
+ "norm_topk_prob": true,
40
+ "num_attention_heads": 96,
41
+ "num_experts_per_tok": 8,
42
+ "num_hidden_layers": 46,
43
+ "num_key_value_heads": 8,
44
+ "partial_rotary_factor": 0.5,
45
+ "rms_norm_eps": 1e-05,
46
+ "torch_dtype": "bfloat16",
47
+ "rope_scaling": {
48
+ "rope_type": "default",
49
+ "mrope_section": [
50
+ 8,
51
+ 12,
52
+ 12
53
+ ]
54
+ },
55
+ "rope_theta": 10000.0,
56
+ "routed_scaling_factor": 1.0,
57
+ "topk_group": 1,
58
+ "use_cache": true,
59
+ "use_qk_norm": false
60
+ },
61
+ "torch_dtype": "bfloat16",
62
+ "transformers_version": "4.55.0.dev0",
63
+ "video_end_token_id": 151342,
64
+ "video_start_token_id": 151341,
65
+ "video_token_id": 151364,
66
+ "vision_config": {
67
+ "attention_bias": false,
68
+ "attention_dropout": 0.0,
69
+ "depth": 24,
70
+ "hidden_act": "silu",
71
+ "hidden_size": 1536,
72
+ "image_size": 336,
73
+ "in_channels": 3,
74
+ "initializer_range": 0.02,
75
+ "intermediate_size": 10944,
76
+ "model_type": "glm4v_moe",
77
+ "num_heads": 12,
78
+ "out_hidden_size": 4096,
79
+ "patch_size": 14,
80
+ "rms_norm_eps": 1e-05,
81
+ "spatial_merge_size": 2,
82
+ "temporal_patch_size": 2
83
+ },
84
+ "quantization_config": {
85
+ "config_groups": {
86
+ "group_0": {
87
+ "input_activations": {
88
+ "actorder": null,
89
+ "block_structure": null,
90
+ "dynamic": true,
91
+ "group_size": null,
92
+ "num_bits": 8,
93
+ "observer": null,
94
+ "observer_kwargs": {},
95
+ "strategy": "token",
96
+ "symmetric": true,
97
+ "type": "float"
98
+ },
99
+ "output_activations": null,
100
+ "targets": [
101
+ "Linear"
102
+ ],
103
+ "weights": {
104
+ "actorder": null,
105
+ "block_structure": null,
106
+ "dynamic": false,
107
+ "group_size": null,
108
+ "num_bits": 8,
109
+ "observer": "minmax",
110
+ "observer_kwargs": {},
111
+ "strategy": "channel",
112
+ "symmetric": true,
113
+ "type": "float"
114
+ }
115
+ }
116
+ },
117
+ "format": "float-quantized",
118
+ "ignore": [
119
+ "model.language_model.layers.4.input_layernorm",
120
+ "model.language_model.layers.3.input_layernorm",
121
+ "model.language_model.layers.2.input_layernorm",
122
+ "model.language_model.layers.0.input_layernorm",
123
+ "model.language_model.layers.1.input_layernorm",
124
+ "model.language_model.layers.8.input_layernorm",
125
+ "model.language_model.layers.6.input_layernorm",
126
+ "model.language_model.layers.7.input_layernorm",
127
+ "model.language_model.layers.5.input_layernorm",
128
+ "model.language_model.layers.9.input_layernorm",
129
+ "model.visual.blocks.6.norm1",
130
+ "model.visual.blocks.7.norm1",
131
+ "model.visual.blocks.8.norm1",
132
+ "model.visual.blocks.9.norm1",
133
+ "model.visual.patch_embed.proj.bias",
134
+ "model.visual.downsample.bias",
135
+ "model.language_model.layers.38.mlp.gate",
136
+ "model.language_model.layers.39.mlp.gate",
137
+ "model.language_model.layers.40.mlp.gate",
138
+ "model.language_model.layers.41.mlp.gate",
139
+ "model.language_model.layers.42.mlp.gate",
140
+ "model.language_model.layers.28.mlp.gate",
141
+ "model.language_model.layers.29.mlp.gate",
142
+ "model.language_model.layers.30.mlp.gate",
143
+ "model.language_model.layers.31.mlp.gate",
144
+ "model.language_model.layers.32.mlp.gate",
145
+ "model.language_model.layers.31.self_attn.q_proj.bias",
146
+ "model.language_model.layers.32.self_attn.q_proj.bias",
147
+ "model.language_model.layers.33.self_attn.q_proj.bias",
148
+ "model.language_model.layers.43.self_attn.q_proj.bias",
149
+ "model.language_model.layers.44.self_attn.q_proj.bias",
150
+ "model.language_model.layers.45.self_attn.q_proj.bias",
151
+ "model.visual.blocks.0.mlp.up_proj",
152
+ "model.visual.blocks.1.mlp.up_proj",
153
+ "model.visual.blocks.2.mlp.up_proj",
154
+ "model.visual.blocks.3.mlp.up_proj",
155
+ "model.visual.blocks.4.mlp.up_proj",
156
+ "model.visual.blocks.5.mlp.up_proj",
157
+ "model.visual.blocks.6.mlp.up_proj",
158
+ "model.visual.blocks.7.mlp.up_proj",
159
+ "model.visual.blocks.8.mlp.up_proj",
160
+ "model.visual.blocks.9.mlp.up_proj",
161
+ "model.language_model.layers.20.self_attn.q_proj.bias",
162
+ "model.language_model.layers.22.self_attn.q_proj.bias",
163
+ "model.language_model.layers.21.self_attn.q_proj.bias",
164
+ "model.language_model.layers.23.self_attn.q_proj.bias",
165
+ "model.language_model.layers.10.self_attn.q_proj.bias",
166
+ "model.language_model.layers.11.self_attn.q_proj.bias",
167
+ "model.language_model.layers.12.self_attn.q_proj.bias",
168
+ "model.language_model.layers.14.self_attn.q_proj.bias",
169
+ "model.language_model.layers.38.self_attn.q_proj.bias",
170
+ "model.language_model.layers.39.self_attn.q_proj.bias",
171
+ "model.language_model.layers.37.post_attention_layernorm",
172
+ "model.language_model.layers.38.post_attention_layernorm",
173
+ "model.language_model.layers.39.post_attention_layernorm",
174
+ "model.language_model.layers.20.post_attention_layernorm",
175
+ "model.language_model.layers.22.post_attention_layernorm",
176
+ "model.language_model.layers.21.post_attention_layernorm",
177
+ "model.language_model.layers.23.post_attention_layernorm",
178
+ "model.language_model.layers.24.post_attention_layernorm",
179
+ "model.language_model.layers.25.post_attention_layernorm",
180
+ "model.language_model.layers.26.post_attention_layernorm",
181
+ "model.language_model.layers.43.self_attn.k_proj.bias",
182
+ "model.language_model.layers.44.self_attn.k_proj.bias",
183
+ "model.language_model.layers.45.self_attn.k_proj.bias",
184
+ "model.language_model.layers.5.self_attn.v_proj.bias",
185
+ "model.language_model.layers.9.self_attn.v_proj.bias",
186
+ "model.language_model.layers.33.mlp.gate",
187
+ "model.language_model.layers.34.mlp.gate",
188
+ "model.language_model.layers.35.mlp.gate",
189
+ "model.language_model.layers.36.mlp.gate",
190
+ "model.language_model.layers.37.mlp.gate",
191
+ "model.language_model.layers.10.mlp.gate",
192
+ "model.language_model.layers.11.mlp.gate",
193
+ "model.language_model.layers.12.mlp.gate",
194
+ "model.language_model.layers.14.mlp.gate",
195
+ "model.language_model.layers.13.mlp.gate",
196
+ "model.language_model.layers.6.mlp.gate",
197
+ "model.language_model.layers.7.mlp.gate",
198
+ "model.language_model.layers.5.mlp.gate",
199
+ "model.language_model.layers.3.mlp.gate",
200
+ "model.language_model.layers.2.mlp.gate",
201
+ "model.language_model.layers.4.mlp.gate",
202
+ "model.language_model.layers.1.mlp.gate",
203
+ "model.language_model.layers.8.mlp.gate",
204
+ "model.language_model.layers.9.mlp.gate",
205
+ "model.visual.blocks.15.mlp.gate_proj",
206
+ "model.visual.blocks.16.mlp.down_proj",
207
+ "model.visual.blocks.16.mlp.gate_proj",
208
+ "model.visual.blocks.17.mlp.down_proj",
209
+ "model.visual.blocks.17.mlp.gate_proj",
210
+ "model.visual.blocks.18.mlp.down_proj",
211
+ "model.visual.blocks.18.mlp.gate_proj",
212
+ "model.visual.blocks.19.mlp.down_proj",
213
+ "model.visual.blocks.19.mlp.gate_proj",
214
+ "model.visual.blocks.20.mlp.down_proj",
215
+ "model.visual.post_layernorm",
216
+ "model.language_model.layers.40.mlp.gate.e_score_correction_bias",
217
+ "model.language_model.layers.41.mlp.gate.e_score_correction_bias",
218
+ "model.language_model.layers.42.mlp.gate.e_score_correction_bias",
219
+ "model.language_model.layers.29.mlp.gate.e_score_correction_bias",
220
+ "model.language_model.layers.30.mlp.gate.e_score_correction_bias",
221
+ "model.language_model.layers.31.mlp.gate.e_score_correction_bias",
222
+ "model.language_model.layers.32.mlp.gate.e_score_correction_bias",
223
+ "model.language_model.layers.33.mlp.gate.e_score_correction_bias",
224
+ "model.language_model.layers.34.mlp.gate.e_score_correction_bias",
225
+ "model.language_model.layers.35.mlp.gate.e_score_correction_bias",
226
+ "model.visual.blocks.12.norm1",
227
+ "model.visual.blocks.12.norm2",
228
+ "model.visual.blocks.13.norm1",
229
+ "model.visual.blocks.13.norm2",
230
+ "model.visual.blocks.14.norm1",
231
+ "model.visual.blocks.14.norm2",
232
+ "model.visual.blocks.15.norm1",
233
+ "model.visual.blocks.15.norm2",
234
+ "model.visual.blocks.16.norm2",
235
+ "model.visual.blocks.17.norm2",
236
+ "model.language_model.layers.45.mlp.gate.e_score_correction_bias",
237
+ "model.language_model.layers.17.post_attention_layernorm",
238
+ "model.language_model.layers.18.post_attention_layernorm",
239
+ "model.language_model.layers.19.post_attention_layernorm",
240
+ "model.language_model.layers.43.post_attention_layernorm",
241
+ "model.language_model.layers.44.post_attention_layernorm",
242
+ "model.language_model.layers.45.post_attention_layernorm",
243
+ "model.language_model.layers.20.self_attn.v_proj.bias",
244
+ "model.language_model.layers.22.self_attn.v_proj.bias",
245
+ "model.language_model.layers.21.self_attn.v_proj.bias",
246
+ "model.language_model.layers.23.self_attn.v_proj.bias",
247
+ "model.language_model.layers.24.self_attn.v_proj.bias",
248
+ "model.language_model.layers.25.self_attn.v_proj.bias",
249
+ "model.language_model.layers.26.self_attn.v_proj.bias",
250
+ "model.language_model.layers.27.self_attn.v_proj.bias",
251
+ "model.language_model.layers.28.self_attn.v_proj.bias",
252
+ "model.language_model.layers.29.self_attn.v_proj.bias",
253
+ "model.visual.blocks.11.mlp.gate_proj",
254
+ "model.visual.blocks.5.mlp.gate_proj",
255
+ "model.visual.blocks.5.norm1",
256
+ "model.visual.blocks.6.mlp.gate_proj",
257
+ "model.visual.blocks.7.mlp.gate_proj",
258
+ "model.visual.blocks.8.mlp.gate_proj",
259
+ "model.visual.blocks.9.mlp.gate_proj",
260
+ "model.language_model.layers.22.mlp.gate.e_score_correction_bias",
261
+ "model.language_model.layers.10.mlp.gate.e_score_correction_bias",
262
+ "model.language_model.layers.11.mlp.gate.e_score_correction_bias",
263
+ "model.language_model.layers.12.mlp.gate.e_score_correction_bias",
264
+ "model.visual.blocks.12.mlp.down_proj",
265
+ "model.visual.blocks.12.mlp.gate_proj",
266
+ "model.language_model.layers.14.mlp.gate.e_score_correction_bias",
267
+ "model.language_model.layers.13.mlp.gate.e_score_correction_bias",
268
+ "model.language_model.layers.15.mlp.gate.e_score_correction_bias",
269
+ "model.visual.blocks.13.mlp.down_proj",
270
+ "model.visual.blocks.13.mlp.gate_proj",
271
+ "model.visual.blocks.14.mlp.down_proj",
272
+ "model.visual.blocks.14.mlp.gate_proj",
273
+ "model.visual.blocks.15.mlp.down_proj",
274
+ "model.language_model.embed_tokens",
275
+ "model.language_model.layers.10.input_layernorm",
276
+ "model.language_model.layers.11.input_layernorm",
277
+ "model.language_model.layers.12.input_layernorm",
278
+ "model.language_model.layers.14.input_layernorm",
279
+ "model.language_model.layers.39.input_layernorm",
280
+ "model.language_model.layers.13.input_layernorm",
281
+ "model.language_model.layers.15.input_layernorm",
282
+ "model.language_model.layers.16.input_layernorm",
283
+ "model.language_model.layers.17.input_layernorm",
284
+ "model.language_model.layers.18.input_layernorm",
285
+ "model.language_model.layers.27.post_attention_layernorm",
286
+ "model.language_model.layers.28.post_attention_layernorm",
287
+ "model.language_model.layers.29.post_attention_layernorm",
288
+ "model.language_model.layers.10.post_attention_layernorm",
289
+ "model.language_model.layers.11.post_attention_layernorm",
290
+ "model.language_model.layers.12.post_attention_layernorm",
291
+ "model.language_model.layers.14.post_attention_layernorm",
292
+ "model.language_model.layers.13.post_attention_layernorm",
293
+ "model.language_model.layers.15.post_attention_layernorm",
294
+ "model.language_model.layers.16.post_attention_layernorm",
295
+ "model.visual.merger.up_proj",
296
+ "model.visual.blocks.19.norm2",
297
+ "model.visual.blocks.20.norm2",
298
+ "model.visual.blocks.21.norm2",
299
+ "model.visual.blocks.22.norm2",
300
+ "model.visual.blocks.23.norm2",
301
+ "model.visual.blocks.10.norm2",
302
+ "model.visual.blocks.11.norm2",
303
+ "model.language_model.layers.2.mlp.gate.e_score_correction_bias",
304
+ "model.language_model.layers.4.mlp.gate.e_score_correction_bias",
305
+ "model.language_model.layers.3.mlp.gate.e_score_correction_bias",
306
+ "model.language_model.layers.1.mlp.gate.e_score_correction_bias",
307
+ "model.language_model.layers.8.mlp.gate.e_score_correction_bias",
308
+ "model.language_model.layers.6.mlp.gate.e_score_correction_bias",
309
+ "model.language_model.layers.7.mlp.gate.e_score_correction_bias",
310
+ "model.language_model.layers.5.mlp.gate.e_score_correction_bias",
311
+ "model.language_model.layers.9.mlp.gate.e_score_correction_bias",
312
+ "model.language_model.layers.2.self_attn.k_proj.bias",
313
+ "model.language_model.layers.4.self_attn.k_proj.bias",
314
+ "model.language_model.layers.3.self_attn.k_proj.bias",
315
+ "model.language_model.layers.0.self_attn.k_proj.bias",
316
+ "model.language_model.layers.1.self_attn.k_proj.bias",
317
+ "model.language_model.layers.8.self_attn.k_proj.bias",
318
+ "model.language_model.layers.6.self_attn.k_proj.bias",
319
+ "model.language_model.layers.7.self_attn.k_proj.bias",
320
+ "model.language_model.layers.5.self_attn.k_proj.bias",
321
+ "model.language_model.layers.9.self_attn.k_proj.bias",
322
+ "model.language_model.layers.34.self_attn.q_proj.bias",
323
+ "model.language_model.layers.35.self_attn.q_proj.bias",
324
+ "model.language_model.layers.36.self_attn.q_proj.bias",
325
+ "model.language_model.layers.37.self_attn.q_proj.bias",
326
+ "model.language_model.layers.13.self_attn.q_proj.bias",
327
+ "model.language_model.layers.15.self_attn.q_proj.bias",
328
+ "model.language_model.layers.16.self_attn.q_proj.bias",
329
+ "model.language_model.layers.17.self_attn.q_proj.bias",
330
+ "model.language_model.layers.18.self_attn.q_proj.bias",
331
+ "model.language_model.layers.19.self_attn.q_proj.bias",
332
+ "model.language_model.layers.40.self_attn.q_proj.bias",
333
+ "model.language_model.layers.41.self_attn.q_proj.bias",
334
+ "model.language_model.layers.42.self_attn.q_proj.bias",
335
+ "model.language_model.layers.24.self_attn.q_proj.bias",
336
+ "model.language_model.layers.25.self_attn.q_proj.bias",
337
+ "model.language_model.layers.26.self_attn.q_proj.bias",
338
+ "model.language_model.layers.27.self_attn.q_proj.bias",
339
+ "model.language_model.layers.28.self_attn.q_proj.bias",
340
+ "model.language_model.layers.29.self_attn.q_proj.bias",
341
+ "model.language_model.layers.30.self_attn.q_proj.bias",
342
+ "model.language_model.layers.25.mlp.gate.e_score_correction_bias",
343
+ "model.language_model.layers.26.mlp.gate.e_score_correction_bias",
344
+ "model.language_model.layers.27.mlp.gate.e_score_correction_bias",
345
+ "model.language_model.layers.28.mlp.gate.e_score_correction_bias",
346
+ "model.language_model.layers.39.mlp.gate.e_score_correction_bias",
347
+ "model.language_model.layers.16.mlp.gate.e_score_correction_bias",
348
+ "model.language_model.layers.17.mlp.gate.e_score_correction_bias",
349
+ "model.language_model.layers.18.mlp.gate.e_score_correction_bias",
350
+ "model.language_model.layers.19.mlp.gate.e_score_correction_bias",
351
+ "model.language_model.layers.20.mlp.gate.e_score_correction_bias",
352
+ "model.language_model.layers.36.mlp.gate.e_score_correction_bias",
353
+ "model.language_model.layers.37.mlp.gate.e_score_correction_bias",
354
+ "model.language_model.layers.38.mlp.gate.e_score_correction_bias",
355
+ "model.language_model.layers.21.mlp.gate.e_score_correction_bias",
356
+ "model.language_model.layers.23.mlp.gate.e_score_correction_bias",
357
+ "model.language_model.layers.24.mlp.gate.e_score_correction_bias",
358
+ "model.language_model.layers.22.self_attn.k_proj.bias",
359
+ "model.language_model.layers.21.self_attn.k_proj.bias",
360
+ "model.language_model.layers.23.self_attn.k_proj.bias",
361
+ "model.language_model.layers.24.self_attn.k_proj.bias",
362
+ "model.language_model.layers.25.self_attn.k_proj.bias",
363
+ "model.language_model.layers.26.self_attn.k_proj.bias",
364
+ "model.language_model.layers.27.self_attn.k_proj.bias",
365
+ "model.language_model.layers.28.self_attn.k_proj.bias",
366
+ "model.language_model.layers.38.self_attn.k_proj.bias",
367
+ "model.language_model.layers.39.self_attn.k_proj.bias",
368
+ "model.visual.blocks.0.norm1",
369
+ "model.visual.blocks.1.norm1",
370
+ "model.visual.blocks.2.norm1",
371
+ "model.visual.blocks.0.norm2",
372
+ "model.visual.blocks.1.norm2",
373
+ "model.visual.blocks.2.norm2",
374
+ "model.visual.blocks.3.norm1",
375
+ "model.visual.blocks.3.norm2",
376
+ "model.visual.blocks.4.norm1",
377
+ "model.visual.blocks.19.attn.qkv",
378
+ "model.visual.blocks.20.attn.qkv",
379
+ "model.visual.blocks.21.attn.qkv",
380
+ "model.visual.blocks.22.attn.qkv",
381
+ "model.visual.blocks.23.attn.qkv",
382
+ "model.visual.blocks.10.attn.qkv",
383
+ "model.visual.blocks.11.attn.qkv",
384
+ "model.visual.blocks.4.norm2",
385
+ "model.visual.blocks.5.norm2",
386
+ "model.visual.blocks.6.norm2",
387
+ "model.visual.blocks.7.norm2",
388
+ "model.visual.blocks.8.norm2",
389
+ "model.visual.blocks.9.norm2",
390
+ "model.visual.blocks.22.mlp.up_proj",
391
+ "model.visual.blocks.23.mlp.up_proj",
392
+ "model.visual.blocks.10.mlp.up_proj",
393
+ "model.visual.blocks.11.mlp.up_proj",
394
+ "model.visual.blocks.20.mlp.gate_proj",
395
+ "model.visual.blocks.21.mlp.down_proj",
396
+ "model.visual.blocks.21.mlp.gate_proj",
397
+ "model.visual.blocks.22.mlp.down_proj",
398
+ "model.visual.blocks.22.mlp.gate_proj",
399
+ "model.visual.blocks.23.mlp.down_proj",
400
+ "model.visual.blocks.23.mlp.gate_proj",
401
+ "model.visual.blocks.10.mlp.down_proj",
402
+ "model.visual.blocks.10.mlp.gate_proj",
403
+ "model.visual.blocks.11.mlp.down_proj",
404
+ "model.visual.embeddings.position_embedding",
405
+ "model.language_model.layers.15.mlp.gate",
406
+ "model.language_model.layers.16.mlp.gate",
407
+ "model.language_model.layers.17.mlp.gate",
408
+ "model.visual.merger.proj",
409
+ "model.language_model.layers.43.mlp.gate",
410
+ "model.language_model.layers.44.mlp.gate",
411
+ "model.language_model.layers.45.mlp.gate",
412
+ "model.language_model.layers.32.input_layernorm",
413
+ "model.language_model.layers.33.input_layernorm",
414
+ "model.language_model.layers.34.input_layernorm",
415
+ "model.language_model.layers.35.input_layernorm",
416
+ "model.language_model.layers.36.input_layernorm",
417
+ "model.language_model.layers.37.input_layernorm",
418
+ "model.language_model.layers.38.input_layernorm",
419
+ "model.language_model.layers.40.input_layernorm",
420
+ "model.language_model.layers.41.input_layernorm",
421
+ "model.language_model.layers.42.input_layernorm",
422
+ "model.language_model.layers.10.self_attn.k_proj.bias",
423
+ "model.language_model.layers.11.self_attn.k_proj.bias",
424
+ "model.language_model.layers.12.self_attn.k_proj.bias",
425
+ "model.language_model.layers.14.self_attn.k_proj.bias",
426
+ "model.language_model.layers.13.self_attn.k_proj.bias",
427
+ "model.language_model.layers.15.self_attn.k_proj.bias",
428
+ "model.language_model.layers.16.self_attn.k_proj.bias",
429
+ "model.language_model.layers.17.self_attn.k_proj.bias",
430
+ "model.language_model.layers.18.self_attn.k_proj.bias",
431
+ "model.language_model.layers.19.self_attn.k_proj.bias",
432
+ "model.visual.patch_embed.proj",
433
+ "model.language_model.layers.32.self_attn.k_proj.bias",
434
+ "model.language_model.layers.33.self_attn.k_proj.bias",
435
+ "model.language_model.layers.34.self_attn.k_proj.bias",
436
+ "model.language_model.layers.35.self_attn.k_proj.bias",
437
+ "model.language_model.layers.36.self_attn.k_proj.bias",
438
+ "model.language_model.layers.37.self_attn.k_proj.bias",
439
+ "model.language_model.layers.40.self_attn.k_proj.bias",
440
+ "model.language_model.layers.41.self_attn.k_proj.bias",
441
+ "model.language_model.layers.42.self_attn.k_proj.bias",
442
+ "model.language_model.layers.29.self_attn.k_proj.bias",
443
+ "lm_head",
444
+ "model.language_model.layers.40.post_attention_layernorm",
445
+ "model.language_model.layers.41.post_attention_layernorm",
446
+ "model.language_model.layers.42.post_attention_layernorm",
447
+ "model.language_model.layers.30.post_attention_layernorm",
448
+ "model.language_model.layers.31.post_attention_layernorm",
449
+ "model.language_model.layers.32.post_attention_layernorm",
450
+ "model.language_model.layers.33.post_attention_layernorm",
451
+ "model.language_model.layers.34.post_attention_layernorm",
452
+ "model.language_model.layers.35.post_attention_layernorm",
453
+ "model.language_model.layers.36.post_attention_layernorm",
454
+ "model.language_model.layers.30.self_attn.k_proj.bias",
455
+ "model.language_model.layers.31.self_attn.k_proj.bias",
456
+ "model.language_model.layers.20.self_attn.k_proj.bias",
457
+ "model.language_model.layers.21.input_layernorm",
458
+ "model.language_model.layers.23.input_layernorm",
459
+ "model.language_model.layers.24.input_layernorm",
460
+ "model.language_model.layers.25.input_layernorm",
461
+ "model.language_model.layers.26.input_layernorm",
462
+ "model.language_model.layers.27.input_layernorm",
463
+ "model.language_model.layers.28.input_layernorm",
464
+ "model.language_model.layers.29.input_layernorm",
465
+ "model.language_model.layers.30.input_layernorm",
466
+ "model.language_model.layers.31.input_layernorm",
467
+ "model.visual.blocks.22.attn.proj",
468
+ "model.visual.blocks.23.attn.proj",
469
+ "model.visual.blocks.10.attn.proj",
470
+ "model.visual.blocks.11.attn.proj",
471
+ "model.visual.merger.down_proj",
472
+ "model.visual.merger.gate_proj",
473
+ "model.visual.merger.post_projection_norm.bias",
474
+ "model.language_model.norm",
475
+ "model.visual.blocks.16.norm1",
476
+ "model.visual.blocks.17.norm1",
477
+ "model.visual.blocks.18.norm1",
478
+ "model.visual.blocks.19.norm1",
479
+ "model.visual.blocks.20.norm1",
480
+ "model.visual.blocks.21.norm1",
481
+ "model.visual.blocks.22.norm1",
482
+ "model.visual.blocks.23.norm1",
483
+ "model.visual.blocks.10.norm1",
484
+ "model.visual.blocks.11.norm1",
485
+ "model.visual.blocks.18.norm2",
486
+ "model.language_model.layers.10.self_attn.v_proj.bias",
487
+ "model.language_model.layers.11.self_attn.v_proj.bias",
488
+ "model.language_model.layers.12.self_attn.v_proj.bias",
489
+ "model.language_model.layers.14.self_attn.v_proj.bias",
490
+ "model.language_model.layers.13.self_attn.v_proj.bias",
491
+ "model.language_model.layers.15.self_attn.v_proj.bias",
492
+ "model.language_model.layers.16.self_attn.v_proj.bias",
493
+ "model.language_model.layers.17.self_attn.v_proj.bias",
494
+ "model.language_model.layers.18.self_attn.v_proj.bias",
495
+ "model.language_model.layers.19.self_attn.v_proj.bias",
496
+ "model.visual.blocks.0.attn.qkv",
497
+ "model.visual.blocks.1.attn.qkv",
498
+ "model.visual.blocks.2.attn.qkv",
499
+ "model.visual.blocks.3.attn.qkv",
500
+ "model.visual.blocks.4.attn.qkv",
501
+ "model.visual.blocks.5.attn.qkv",
502
+ "model.visual.blocks.6.attn.qkv",
503
+ "model.visual.blocks.7.attn.qkv",
504
+ "model.visual.blocks.8.attn.qkv",
505
+ "model.visual.blocks.9.attn.qkv",
506
+ "model.language_model.layers.31.self_attn.v_proj.bias",
507
+ "model.language_model.layers.32.self_attn.v_proj.bias",
508
+ "model.language_model.layers.33.self_attn.v_proj.bias",
509
+ "model.language_model.layers.43.self_attn.v_proj.bias",
510
+ "model.language_model.layers.44.self_attn.v_proj.bias",
511
+ "model.language_model.layers.45.self_attn.v_proj.bias",
512
+ "model.visual.post_conv_layernorm",
513
+ "model.visual.blocks.0.mlp.down_proj",
514
+ "model.visual.blocks.0.mlp.gate_proj",
515
+ "model.visual.blocks.1.mlp.down_proj",
516
+ "model.visual.blocks.1.mlp.gate_proj",
517
+ "model.visual.blocks.2.mlp.down_proj",
518
+ "model.visual.blocks.2.mlp.gate_proj",
519
+ "model.visual.blocks.3.mlp.down_proj",
520
+ "model.visual.blocks.3.mlp.gate_proj",
521
+ "model.visual.blocks.4.mlp.down_proj",
522
+ "model.visual.blocks.4.mlp.gate_proj",
523
+ "model.visual.merger.post_projection_norm",
524
+ "model.visual.blocks.12.attn.proj",
525
+ "model.visual.blocks.13.attn.proj",
526
+ "model.visual.blocks.14.attn.proj",
527
+ "model.visual.blocks.15.attn.proj",
528
+ "model.visual.blocks.16.attn.proj",
529
+ "model.visual.blocks.17.attn.proj",
530
+ "model.visual.blocks.18.attn.proj",
531
+ "model.visual.blocks.19.attn.proj",
532
+ "model.visual.blocks.20.attn.proj",
533
+ "model.visual.blocks.21.attn.proj",
534
+ "model.visual.blocks.5.mlp.down_proj",
535
+ "model.visual.blocks.6.mlp.down_proj",
536
+ "model.visual.blocks.7.mlp.down_proj",
537
+ "model.visual.blocks.8.mlp.down_proj",
538
+ "model.visual.blocks.9.mlp.down_proj",
539
+ "model.visual.blocks.0.attn.proj",
540
+ "model.visual.blocks.1.attn.proj",
541
+ "model.visual.blocks.2.attn.proj",
542
+ "model.visual.blocks.3.attn.proj",
543
+ "model.visual.blocks.4.attn.proj",
544
+ "model.visual.blocks.5.attn.proj",
545
+ "model.visual.blocks.6.attn.proj",
546
+ "model.visual.blocks.7.attn.proj",
547
+ "model.visual.blocks.8.attn.proj",
548
+ "model.visual.blocks.9.attn.proj",
549
+ "model.visual.blocks.12.attn.qkv",
550
+ "model.visual.blocks.13.attn.qkv",
551
+ "model.visual.blocks.14.attn.qkv",
552
+ "model.visual.blocks.15.attn.qkv",
553
+ "model.visual.blocks.16.attn.qkv",
554
+ "model.visual.blocks.17.attn.qkv",
555
+ "model.visual.blocks.18.attn.qkv",
556
+ "model.language_model.layers.34.self_attn.v_proj.bias",
557
+ "model.language_model.layers.35.self_attn.v_proj.bias",
558
+ "model.language_model.layers.36.self_attn.v_proj.bias",
559
+ "model.language_model.layers.37.self_attn.v_proj.bias",
560
+ "model.language_model.layers.38.self_attn.v_proj.bias",
561
+ "model.language_model.layers.39.self_attn.v_proj.bias",
562
+ "model.language_model.layers.40.self_attn.v_proj.bias",
563
+ "model.language_model.layers.41.self_attn.v_proj.bias",
564
+ "model.language_model.layers.42.self_attn.v_proj.bias",
565
+ "model.language_model.layers.30.self_attn.v_proj.bias",
566
+ "model.visual.blocks.12.mlp.up_proj",
567
+ "model.visual.blocks.13.mlp.up_proj",
568
+ "model.visual.blocks.14.mlp.up_proj",
569
+ "model.visual.blocks.15.mlp.up_proj",
570
+ "model.visual.blocks.16.mlp.up_proj",
571
+ "model.visual.blocks.17.mlp.up_proj",
572
+ "model.visual.blocks.18.mlp.up_proj",
573
+ "model.visual.blocks.19.mlp.up_proj",
574
+ "model.visual.blocks.20.mlp.up_proj",
575
+ "model.visual.blocks.21.mlp.up_proj",
576
+ "model.language_model.layers.2.self_attn.q_proj.bias",
577
+ "model.language_model.layers.4.self_attn.q_proj.bias",
578
+ "model.language_model.layers.3.self_attn.q_proj.bias",
579
+ "model.language_model.layers.0.self_attn.q_proj.bias",
580
+ "model.language_model.layers.1.self_attn.q_proj.bias",
581
+ "model.language_model.layers.8.self_attn.q_proj.bias",
582
+ "model.language_model.layers.6.self_attn.q_proj.bias",
583
+ "model.language_model.layers.7.self_attn.q_proj.bias",
584
+ "model.language_model.layers.5.self_attn.q_proj.bias",
585
+ "model.language_model.layers.9.self_attn.q_proj.bias",
586
+ "model.language_model.layers.2.self_attn.v_proj.bias",
587
+ "model.language_model.layers.4.self_attn.v_proj.bias",
588
+ "model.language_model.layers.3.self_attn.v_proj.bias",
589
+ "model.language_model.layers.0.self_attn.v_proj.bias",
590
+ "model.language_model.layers.1.self_attn.v_proj.bias",
591
+ "model.language_model.layers.8.self_attn.v_proj.bias",
592
+ "model.language_model.layers.6.self_attn.v_proj.bias",
593
+ "model.language_model.layers.7.self_attn.v_proj.bias",
594
+ "model.language_model.layers.18.mlp.gate",
595
+ "model.language_model.layers.19.mlp.gate",
596
+ "model.language_model.layers.20.mlp.gate",
597
+ "model.language_model.layers.22.mlp.gate",
598
+ "model.language_model.layers.21.mlp.gate",
599
+ "model.language_model.layers.23.mlp.gate",
600
+ "model.language_model.layers.24.mlp.gate",
601
+ "model.language_model.layers.25.mlp.gate",
602
+ "model.language_model.layers.26.mlp.gate",
603
+ "model.language_model.layers.27.mlp.gate",
604
+ "model.language_model.layers.43.mlp.gate.e_score_correction_bias",
605
+ "model.language_model.layers.44.mlp.gate.e_score_correction_bias",
606
+ "model.language_model.layers.19.input_layernorm",
607
+ "model.language_model.layers.20.input_layernorm",
608
+ "model.language_model.layers.22.input_layernorm",
609
+ "model.language_model.layers.43.input_layernorm",
610
+ "model.language_model.layers.44.input_layernorm",
611
+ "model.language_model.layers.45.input_layernorm",
612
+ "model.language_model.layers.2.post_attention_layernorm",
613
+ "model.language_model.layers.4.post_attention_layernorm",
614
+ "model.language_model.layers.3.post_attention_layernorm",
615
+ "model.language_model.layers.0.post_attention_layernorm",
616
+ "model.language_model.layers.1.post_attention_layernorm",
617
+ "model.language_model.layers.8.post_attention_layernorm",
618
+ "model.language_model.layers.6.post_attention_layernorm",
619
+ "model.language_model.layers.7.post_attention_layernorm",
620
+ "model.language_model.layers.5.post_attention_layernorm",
621
+ "model.language_model.layers.9.post_attention_layernorm",
622
+ "model.visual.downsample"
623
+ ],
624
+ "quant_method": "compressed-tensors",
625
+ "quantization_status": "compressed"
626
+ }
627
+ }
generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 151329,
6
+ 151336,
7
+ 151338
8
+ ],
9
+ "pad_token_id": 151329,
10
+ "temperature": 1.0,
11
+ "top_k": 1,
12
+ "top_p": 0.0001,
13
+ "transformers_version": "4.55.0.dev"
14
+ }
model-00001-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c0026c1097650dede9bbc7893416a6c56fd86dcf32ef566ef195a1ffbb51bb3
3
+ size 2589546376
model-00002-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b499d123943a4fd52eb5e4891eb818b07532e29ce365a9a2f58508ccf0c81db
3
+ size 2345789304
model-00003-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:82727d15e28b67d590b476147797a034af5df1bda60afdaadd1d6b15717e0c10
3
+ size 2345789304
model-00004-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af67eecd864912aca8247d8ca4f9e6ae6229671f8addb2d30c091328d9548255
3
+ size 2345789304
model-00005-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a498348dd97f4787defe9451a331cce25ebccd6b955b729e9363924b456ced22
3
+ size 2345789304
model-00006-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9743eeba15d76650e49da82d6c0aa8cc899609bd7e7c8cdecf3b4bed9864c43d
3
+ size 2345789304
model-00007-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:22a4ea664ad038757fddee9a5ea51c94e935bc3913a9e8f061ed76a16660e600
3
+ size 2345789304
model-00008-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d997cb2a4e73c7933622f7d054ad356a177296bd166a88a8385e1d4f16ef3add
3
+ size 2345789304
model-00009-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d0417629496b9816d8b804c94961f270da7eba106432308c7f70e04fc3f6d16f
3
+ size 2345789304
model-00010-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1fc5d0e0c6be9c991fb30838fb916f65b59a3001a6d45910ea2f2b5b5884467e
3
+ size 2345790088
model-00011-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:42ea51bd9237e8daf79ab29df7b3d4a761fdba43bf00721fd859538651dcb6ef
3
+ size 2345790088
model-00012-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f3c04ccf3e26d0eda5a67449c2ad29f01ac8e6472b02f239ae9246d4f89e443
3
+ size 2345790088
model-00013-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7e6b8fc35578a270a5c9eb87faadbe9cc22464f47418a4c1837f083d06680a8
3
+ size 2345790088
model-00014-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:36e86fd429e5f83a612be7ac9b1ab610666f9567cd8899d08a258ea6cdf0a691
3
+ size 2345790088
model-00015-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4f0c2706171c49ffe01e95bd890fd15cf52552f6ddf5926fa9be1385fe0fc184
3
+ size 2345790088
model-00016-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05fa041f2402fdf6f1d2dbb202513d1b17ca5de5f16a0d05d9b5b4dcbf3bfb45
3
+ size 2345790088
model-00017-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:081941f1d321b7bc08b543a96fe6c4f8733ff76c23b588329071c897a9947cf2
3
+ size 2345790088
model-00018-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77d4a16e9f6b6228cfa9182af0485d18af15fa4909c66531e69c8f7b355a1258
3
+ size 2345790088
model-00019-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d43582ed4756714587fd0a40b998bef7aa7730bb0c039888acfc2b14ead150d9
3
+ size 2345790088
model-00020-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:03428d53dd36f6794c46ee25c99a6c75bda1be39ab932a76880c836378e1eb6c
3
+ size 2345790088
model-00021-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2e1ddb79657265803c09bb0f9055e6dc7bb63bca50955fc69afb8c6a50eb613
3
+ size 2345790088
model-00022-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69498b48c34dad39580154988e6b7e6c2eca094137cbe7073d4e4adf94aca460
3
+ size 2345790088
model-00023-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:22507d8b255ea3ed470a92d91c4ce80cc31ae4f2415b2330f06b3a94ffdc4442
3
+ size 2345790088
model-00024-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a041264a8f85653758b9d86415af6b0a69452635978f3992e2ec62b1d5ee587
3
+ size 2345790088
model-00025-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:818895f30d480a87af7e8e366391144b82efe08261a0416d79ecbc93f81cb193
3
+ size 2345790088
model-00026-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:90b82c161129374c39d8a9af9292b0c0c5267a0adabded559b5514a9acc2b48b
3
+ size 2345790088
model-00027-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6db358d11a9ca48bceb512a7ed3112e64a10063467f7df7c5b3e9c1ae0e0b66d
3
+ size 2345790088
model-00028-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69bb06e0eddba31f9b7158c4ab3bbd4027d9599ddccdcb3b2fa031e81977baa4
3
+ size 2345790088
model-00029-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a18b70bbca580315a7c6cc02b9ad46f2d90e0edb69867b64fd2355f106aa0eab
3
+ size 2345790088
model-00030-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b32f26dddfe6602af964a3b75399b13acd31d1a20acd723dc5fb2c608c2fb820
3
+ size 2345790088
model-00031-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:037e5337e1cfd3785be716b81577abbc448113d8acb4bd5663236cb0ed373d79
3
+ size 2345790088
model-00032-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51d097c3d9ae3f5e1ee58b29ffe8406fed9e27bbbbf94d643c2d6187135383eb
3
+ size 2345790088
model-00033-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eac2ec893d4c0865c9e0e010e37066b04b75428108532f9653435de681ff7eaa
3
+ size 2345790088
model-00034-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71ad484ce5341cfbcc34e1c4f7561924f4fddc19a792af791a38fd0d7bae45f1
3
+ size 2345790088
model-00036-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:391e4d870d3cac92d92af66b29afdec74fd14920ec7e96187ac6509346879bfd
3
+ size 2345790088
model-00038-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59a38f7f5663a3af97fee68b5c78305f110a2cbeb2dc3a5f1c4eedc584df1d75
3
+ size 2345790088
model-00039-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd378b918247789e9770ce4f56152d8f8e3775897dd644ab005de548633e76ab
3
+ size 2345790088
model-00041-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e05919e962b44df34c66477a74cd01fc7dc780f68217e2584a25f80ef98a84ec
3
+ size 2345790088
model-00042-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:42f5879666cca387c0d593fdf522e5820880f6cd74ca90d244f418567cd0498c
3
+ size 2345790088
model-00043-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7a2d9af66d73ba137252b42c56e3123b0389c1e237461449e9e98ddabe11c5f
3
+ size 2345790088
model-00045-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:413bd3fe7a33d0df18382d67d8fc1cc7bf2dd8fbe4e6dd29bf2a90b703496702
3
+ size 3025350880
model-00046-of-00046.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e42f8c540b7eaa349cc6dac69d9c066bc863328867c3291e1f47af20781b3f96
3
+ size 3520860544
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
preprocessor_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "size": {"shortest_edge": 12544, "longest_edge": 9633792},
3
+ "do_rescale": true,
4
+ "patch_size": 14,
5
+ "temporal_patch_size": 2,
6
+ "merge_size": 2,
7
+ "image_mean": [0.48145466, 0.4578275, 0.40821073],
8
+ "image_std": [0.26862954, 0.26130258, 0.27577711],
9
+ "image_processor_type": "Glm4vImageProcessor",
10
+ "processor_class": "Glm4vProcessor"
11
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9340665016419c825c4bdabbcc9acc43b7ca2c68ce142724afa829abb1be5efd
3
+ size 19970699
tokenizer_config.json ADDED
@@ -0,0 +1,325 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "151329": {
4
+ "content": "<|endoftext|>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "151330": {
12
+ "content": "[MASK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "151331": {
20
+ "content": "[gMASK]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "151332": {
28
+ "content": "[sMASK]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "151333": {
36
+ "content": "<sop>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "151334": {
44
+ "content": "<eop>",
45
+ "lstrip": false,
46
+ "normalized": false,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": true
50
+ },
51
+ "151335": {
52
+ "content": "<|system|>",
53
+ "lstrip": false,
54
+ "normalized": false,
55
+ "rstrip": false,
56
+ "single_word": false,
57
+ "special": true
58
+ },
59
+ "151336": {
60
+ "content": "<|user|>",
61
+ "lstrip": false,
62
+ "normalized": false,
63
+ "rstrip": false,
64
+ "single_word": false,
65
+ "special": true
66
+ },
67
+ "151337": {
68
+ "content": "<|assistant|>",
69
+ "lstrip": false,
70
+ "normalized": false,
71
+ "rstrip": false,
72
+ "single_word": false,
73
+ "special": true
74
+ },
75
+ "151338": {
76
+ "content": "<|observation|>",
77
+ "lstrip": false,
78
+ "normalized": false,
79
+ "rstrip": false,
80
+ "single_word": false,
81
+ "special": true
82
+ },
83
+ "151339": {
84
+ "content": "<|begin_of_image|>",
85
+ "lstrip": false,
86
+ "normalized": false,
87
+ "rstrip": false,
88
+ "single_word": false,
89
+ "special": true
90
+ },
91
+ "151340": {
92
+ "content": "<|end_of_image|>",
93
+ "lstrip": false,
94
+ "normalized": false,
95
+ "rstrip": false,
96
+ "single_word": false,
97
+ "special": true
98
+ },
99
+ "151341": {
100
+ "content": "<|begin_of_video|>",
101
+ "lstrip": false,
102
+ "normalized": false,
103
+ "rstrip": false,
104
+ "single_word": false,
105
+ "special": true
106
+ },
107
+ "151342": {
108
+ "content": "<|end_of_video|>",
109
+ "lstrip": false,
110
+ "normalized": false,
111
+ "rstrip": false,
112
+ "single_word": false,
113
+ "special": true
114
+ },
115
+ "151343": {
116
+ "content": "<|begin_of_audio|>",
117
+ "lstrip": false,
118
+ "normalized": false,
119
+ "rstrip": false,
120
+ "single_word": false,
121
+ "special": true
122
+ },
123
+ "151344": {
124
+ "content": "<|end_of_audio|>",
125
+ "lstrip": false,
126
+ "normalized": false,
127
+ "rstrip": false,
128
+ "single_word": false,
129
+ "special": true
130
+ },
131
+ "151345": {
132
+ "content": "<|begin_of_transcription|>",
133
+ "lstrip": false,
134
+ "normalized": false,
135
+ "rstrip": false,
136
+ "single_word": false,
137
+ "special": true
138
+ },
139
+ "151346": {
140
+ "content": "<|end_of_transcription|>",
141
+ "lstrip": false,
142
+ "normalized": false,
143
+ "rstrip": false,
144
+ "single_word": false,
145
+ "special": true
146
+ },
147
+ "151347": {
148
+ "content": "<|code_prefix|>",
149
+ "lstrip": false,
150
+ "normalized": false,
151
+ "rstrip": false,
152
+ "single_word": false,
153
+ "special": true
154
+ },
155
+ "151348": {
156
+ "content": "<|code_middle|>",
157
+ "lstrip": false,
158
+ "normalized": false,
159
+ "rstrip": false,
160
+ "single_word": false,
161
+ "special": true
162
+ },
163
+ "151349": {
164
+ "content": "<|code_suffix|>",
165
+ "lstrip": false,
166
+ "normalized": false,
167
+ "rstrip": false,
168
+ "single_word": false,
169
+ "special": true
170
+ },
171
+ "151350": {
172
+ "content": "<think>",
173
+ "lstrip": false,
174
+ "normalized": false,
175
+ "rstrip": false,
176
+ "single_word": false,
177
+ "special": false
178
+ },
179
+ "151351": {
180
+ "content": "</think>",
181
+ "lstrip": false,
182
+ "normalized": false,
183
+ "rstrip": false,
184
+ "single_word": false,
185
+ "special": false
186
+ },
187
+ "151352": {
188
+ "content": "<tool_call>",
189
+ "lstrip": false,
190
+ "normalized": false,
191
+ "rstrip": false,
192
+ "single_word": false,
193
+ "special": false
194
+ },
195
+ "151353": {
196
+ "content": "</tool_call>",
197
+ "lstrip": false,
198
+ "normalized": false,
199
+ "rstrip": false,
200
+ "single_word": false,
201
+ "special": false
202
+ },
203
+ "151354": {
204
+ "content": "<tool_response>",
205
+ "lstrip": false,
206
+ "normalized": false,
207
+ "rstrip": false,
208
+ "single_word": false,
209
+ "special": false
210
+ },
211
+ "151355": {
212
+ "content": "</tool_response>",
213
+ "lstrip": false,
214
+ "normalized": false,
215
+ "rstrip": false,
216
+ "single_word": false,
217
+ "special": false
218
+ },
219
+ "151356": {
220
+ "content": "<arg_key>",
221
+ "lstrip": false,
222
+ "normalized": false,
223
+ "rstrip": false,
224
+ "single_word": false,
225
+ "special": false
226
+ },
227
+ "151357": {
228
+ "content": "</arg_key>",
229
+ "lstrip": false,
230
+ "normalized": false,
231
+ "rstrip": false,
232
+ "single_word": false,
233
+ "special": false
234
+ },
235
+ "151358": {
236
+ "content": "<arg_value>",
237
+ "lstrip": false,
238
+ "normalized": false,
239
+ "rstrip": false,
240
+ "single_word": false,
241
+ "special": false
242
+ },
243
+ "151359": {
244
+ "content": "</arg_value>",
245
+ "lstrip": false,
246
+ "normalized": false,
247
+ "rstrip": false,
248
+ "single_word": false,
249
+ "special": false
250
+ },
251
+ "151360": {
252
+ "content": "/nothink",
253
+ "lstrip": false,
254
+ "normalized": false,
255
+ "rstrip": false,
256
+ "single_word": false,
257
+ "special": true
258
+ },
259
+ "151361": {
260
+ "content": "<|begin_of_box|>",
261
+ "lstrip": false,
262
+ "normalized": false,
263
+ "rstrip": false,
264
+ "single_word": false,
265
+ "special": false
266
+ },
267
+ "151362": {
268
+ "content": "<|end_of_box|>",
269
+ "lstrip": false,
270
+ "normalized": false,
271
+ "rstrip": false,
272
+ "single_word": false,
273
+ "special": false
274
+ },
275
+ "151363": {
276
+ "content": "<|image|>",
277
+ "lstrip": false,
278
+ "normalized": false,
279
+ "rstrip": false,
280
+ "single_word": false,
281
+ "special": false
282
+ },
283
+ "151364": {
284
+ "content": "<|video|>",
285
+ "lstrip": false,
286
+ "normalized": false,
287
+ "rstrip": false,
288
+ "single_word": false,
289
+ "special": false
290
+ }
291
+ },
292
+ "additional_special_tokens": [
293
+ "<|endoftext|>",
294
+ "[MASK]",
295
+ "[gMASK]",
296
+ "[sMASK]",
297
+ "<sop>",
298
+ "<eop>",
299
+ "<|system|>",
300
+ "<|user|>",
301
+ "<|assistant|>",
302
+ "<|observation|>",
303
+ "<|begin_of_image|>",
304
+ "<|end_of_image|>",
305
+ "<|begin_of_video|>",
306
+ "<|end_of_video|>",
307
+ "<|begin_of_audio|>",
308
+ "<|end_of_audio|>",
309
+ "<|begin_of_transcription|>",
310
+ "<|end_of_transcription|>",
311
+ "<|code_prefix|>",
312
+ "<|code_middle|>",
313
+ "<|code_suffix|>",
314
+ "/nothink"
315
+ ],
316
+ "clean_up_tokenization_spaces": false,
317
+ "do_lower_case": false,
318
+ "eos_token": "<|endoftext|>",
319
+ "extra_special_tokens": {},
320
+ "model_max_length": 128000,
321
+ "pad_token": "<|endoftext|>",
322
+ "padding_side": "left",
323
+ "remove_space": false,
324
+ "tokenizer_class": "PreTrainedTokenizer"
325
+ }