Hugging Face's logo

matthewdubois
/
GLM-OCR-4bit-mlx

Safetensors
glm_ocr
4-bit precision
Model card Files
xet
 Community
matthewdubois commited on
Commit
9b8cb30
·
verified ·
1 Parent(s): 7a71833

Upload folder using huggingface_hub

Browse files
Files changed (9) hide show
  1. chat_template.jinja +140 -0
  2. config.json +1855 -0
  3. generation_config.json +10 -0
  4. model.safetensors +3 -0
  5. model.safetensors.index.json +961 -0
  6. preprocessor_config.json +22 -0
  7. processor_config.json +30 -0
  8. tokenizer.json +0 -0
  9. tokenizer_config.json +49 -0
chat_template.jinja ADDED
@@ -0,0 +1,140 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ {% elif m.content is iterable and m.content is not mapping %}
104
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
105
+ {{- '<|observation|>' }}
106
+ {%- endif %}
107
+ {{- '\n<tool_response>\n' }}
108
+ {%- for tr in m.content -%}
109
+ {%- if tr is mapping and tr.type is defined -%}
110
+ {%- set t = tr.type | lower -%}
111
+ {%- if t == 'text' and tr.text is defined -%}
112
+ {{ tr.text }}
113
+ {%- elif t in ['image', 'image_url'] -%}
114
+ <|begin_of_image|><|image|><|end_of_image|>
115
+ {%- elif t in ['video', 'video_url'] -%}
116
+ <|begin_of_video|><|video|><|end_of_video|>
117
+ {%- else -%}
118
+ {{ tr | tojson(ensure_ascii=False) }}
119
+ {%- endif -%}
120
+ {%- else -%}
121
+ {{ tr.output if tr.output is defined else tr }}
122
+ {%- endif -%}
123
+ {%- endfor -%}
124
+ {{- '\n</tool_response>' }}
125
+ {%- else -%}
126
+ <|observation|>{% for tr in m.content %}
127
+
128
+ <tool_response>
129
+ {{ tr.output if tr.output is defined else tr }}
130
+ </tool_response>{% endfor -%}
131
+ {% endif -%}
132
+ {%- elif m.role == 'system' -%}
133
+ <|system|>
134
+ {{ visible_text(m.content) }}
135
+ {%- endif -%}
136
+ {%- endfor -%}
137
+ {%- if add_generation_prompt -%}
138
+ <|assistant|>
139
+ {{'<think></think>\n' if (enable_thinking is defined and not enable_thinking) else ''}}
140
+ {%- endif -%}
config.json ADDED
@@ -0,0 +1,1855 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "GlmOcrForConditionalGeneration"
4
+ ],
5
+ "eos_token_id": [
6
+ 59246,
7
+ 59253
8
+ ],
9
+ "image_end_token_id": 59257,
10
+ "image_start_token_id": 59256,
11
+ "image_token_id": 59280,
12
+ "model_type": "glm_ocr",
13
+ "quantization": {
14
+ "group_size": 64,
15
+ "bits": 4,
16
+ "mode": "affine",
17
+ "vision_tower.blocks.0.attn.qkv": {
18
+ "group_size": 64,
19
+ "bits": 4
20
+ },
21
+ "vision_tower.blocks.0.attn.proj": {
22
+ "group_size": 64,
23
+ "bits": 4
24
+ },
25
+ "vision_tower.blocks.0.mlp.gate_proj": {
26
+ "group_size": 64,
27
+ "bits": 4
28
+ },
29
+ "vision_tower.blocks.0.mlp.up_proj": {
30
+ "group_size": 64,
31
+ "bits": 4
32
+ },
33
+ "vision_tower.blocks.0.mlp.down_proj": {
34
+ "group_size": 64,
35
+ "bits": 4
36
+ },
37
+ "vision_tower.blocks.1.attn.qkv": {
38
+ "group_size": 64,
39
+ "bits": 4
40
+ },
41
+ "vision_tower.blocks.1.attn.proj": {
42
+ "group_size": 64,
43
+ "bits": 4
44
+ },
45
+ "vision_tower.blocks.1.mlp.gate_proj": {
46
+ "group_size": 64,
47
+ "bits": 4
48
+ },
49
+ "vision_tower.blocks.1.mlp.up_proj": {
50
+ "group_size": 64,
51
+ "bits": 4
52
+ },
53
+ "vision_tower.blocks.1.mlp.down_proj": {
54
+ "group_size": 64,
55
+ "bits": 4
56
+ },
57
+ "vision_tower.blocks.2.attn.qkv": {
58
+ "group_size": 64,
59
+ "bits": 4
60
+ },
61
+ "vision_tower.blocks.2.attn.proj": {
62
+ "group_size": 64,
63
+ "bits": 4
64
+ },
65
+ "vision_tower.blocks.2.mlp.gate_proj": {
66
+ "group_size": 64,
67
+ "bits": 4
68
+ },
69
+ "vision_tower.blocks.2.mlp.up_proj": {
70
+ "group_size": 64,
71
+ "bits": 4
72
+ },
73
+ "vision_tower.blocks.2.mlp.down_proj": {
74
+ "group_size": 64,
75
+ "bits": 4
76
+ },
77
+ "vision_tower.blocks.3.attn.qkv": {
78
+ "group_size": 64,
79
+ "bits": 4
80
+ },
81
+ "vision_tower.blocks.3.attn.proj": {
82
+ "group_size": 64,
83
+ "bits": 4
84
+ },
85
+ "vision_tower.blocks.3.mlp.gate_proj": {
86
+ "group_size": 64,
87
+ "bits": 4
88
+ },
89
+ "vision_tower.blocks.3.mlp.up_proj": {
90
+ "group_size": 64,
91
+ "bits": 4
92
+ },
93
+ "vision_tower.blocks.3.mlp.down_proj": {
94
+ "group_size": 64,
95
+ "bits": 4
96
+ },
97
+ "vision_tower.blocks.4.attn.qkv": {
98
+ "group_size": 64,
99
+ "bits": 4
100
+ },
101
+ "vision_tower.blocks.4.attn.proj": {
102
+ "group_size": 64,
103
+ "bits": 4
104
+ },
105
+ "vision_tower.blocks.4.mlp.gate_proj": {
106
+ "group_size": 64,
107
+ "bits": 4
108
+ },
109
+ "vision_tower.blocks.4.mlp.up_proj": {
110
+ "group_size": 64,
111
+ "bits": 4
112
+ },
113
+ "vision_tower.blocks.4.mlp.down_proj": {
114
+ "group_size": 64,
115
+ "bits": 4
116
+ },
117
+ "vision_tower.blocks.5.attn.qkv": {
118
+ "group_size": 64,
119
+ "bits": 4
120
+ },
121
+ "vision_tower.blocks.5.attn.proj": {
122
+ "group_size": 64,
123
+ "bits": 4
124
+ },
125
+ "vision_tower.blocks.5.mlp.gate_proj": {
126
+ "group_size": 64,
127
+ "bits": 4
128
+ },
129
+ "vision_tower.blocks.5.mlp.up_proj": {
130
+ "group_size": 64,
131
+ "bits": 4
132
+ },
133
+ "vision_tower.blocks.5.mlp.down_proj": {
134
+ "group_size": 64,
135
+ "bits": 4
136
+ },
137
+ "vision_tower.blocks.6.attn.qkv": {
138
+ "group_size": 64,
139
+ "bits": 4
140
+ },
141
+ "vision_tower.blocks.6.attn.proj": {
142
+ "group_size": 64,
143
+ "bits": 4
144
+ },
145
+ "vision_tower.blocks.6.mlp.gate_proj": {
146
+ "group_size": 64,
147
+ "bits": 4
148
+ },
149
+ "vision_tower.blocks.6.mlp.up_proj": {
150
+ "group_size": 64,
151
+ "bits": 4
152
+ },
153
+ "vision_tower.blocks.6.mlp.down_proj": {
154
+ "group_size": 64,
155
+ "bits": 4
156
+ },
157
+ "vision_tower.blocks.7.attn.qkv": {
158
+ "group_size": 64,
159
+ "bits": 4
160
+ },
161
+ "vision_tower.blocks.7.attn.proj": {
162
+ "group_size": 64,
163
+ "bits": 4
164
+ },
165
+ "vision_tower.blocks.7.mlp.gate_proj": {
166
+ "group_size": 64,
167
+ "bits": 4
168
+ },
169
+ "vision_tower.blocks.7.mlp.up_proj": {
170
+ "group_size": 64,
171
+ "bits": 4
172
+ },
173
+ "vision_tower.blocks.7.mlp.down_proj": {
174
+ "group_size": 64,
175
+ "bits": 4
176
+ },
177
+ "vision_tower.blocks.8.attn.qkv": {
178
+ "group_size": 64,
179
+ "bits": 4
180
+ },
181
+ "vision_tower.blocks.8.attn.proj": {
182
+ "group_size": 64,
183
+ "bits": 4
184
+ },
185
+ "vision_tower.blocks.8.mlp.gate_proj": {
186
+ "group_size": 64,
187
+ "bits": 4
188
+ },
189
+ "vision_tower.blocks.8.mlp.up_proj": {
190
+ "group_size": 64,
191
+ "bits": 4
192
+ },
193
+ "vision_tower.blocks.8.mlp.down_proj": {
194
+ "group_size": 64,
195
+ "bits": 4
196
+ },
197
+ "vision_tower.blocks.9.attn.qkv": {
198
+ "group_size": 64,
199
+ "bits": 4
200
+ },
201
+ "vision_tower.blocks.9.attn.proj": {
202
+ "group_size": 64,
203
+ "bits": 4
204
+ },
205
+ "vision_tower.blocks.9.mlp.gate_proj": {
206
+ "group_size": 64,
207
+ "bits": 4
208
+ },
209
+ "vision_tower.blocks.9.mlp.up_proj": {
210
+ "group_size": 64,
211
+ "bits": 4
212
+ },
213
+ "vision_tower.blocks.9.mlp.down_proj": {
214
+ "group_size": 64,
215
+ "bits": 4
216
+ },
217
+ "vision_tower.blocks.10.attn.qkv": {
218
+ "group_size": 64,
219
+ "bits": 4
220
+ },
221
+ "vision_tower.blocks.10.attn.proj": {
222
+ "group_size": 64,
223
+ "bits": 4
224
+ },
225
+ "vision_tower.blocks.10.mlp.gate_proj": {
226
+ "group_size": 64,
227
+ "bits": 4
228
+ },
229
+ "vision_tower.blocks.10.mlp.up_proj": {
230
+ "group_size": 64,
231
+ "bits": 4
232
+ },
233
+ "vision_tower.blocks.10.mlp.down_proj": {
234
+ "group_size": 64,
235
+ "bits": 4
236
+ },
237
+ "vision_tower.blocks.11.attn.qkv": {
238
+ "group_size": 64,
239
+ "bits": 4
240
+ },
241
+ "vision_tower.blocks.11.attn.proj": {
242
+ "group_size": 64,
243
+ "bits": 4
244
+ },
245
+ "vision_tower.blocks.11.mlp.gate_proj": {
246
+ "group_size": 64,
247
+ "bits": 4
248
+ },
249
+ "vision_tower.blocks.11.mlp.up_proj": {
250
+ "group_size": 64,
251
+ "bits": 4
252
+ },
253
+ "vision_tower.blocks.11.mlp.down_proj": {
254
+ "group_size": 64,
255
+ "bits": 4
256
+ },
257
+ "vision_tower.blocks.12.attn.qkv": {
258
+ "group_size": 64,
259
+ "bits": 4
260
+ },
261
+ "vision_tower.blocks.12.attn.proj": {
262
+ "group_size": 64,
263
+ "bits": 4
264
+ },
265
+ "vision_tower.blocks.12.mlp.gate_proj": {
266
+ "group_size": 64,
267
+ "bits": 4
268
+ },
269
+ "vision_tower.blocks.12.mlp.up_proj": {
270
+ "group_size": 64,
271
+ "bits": 4
272
+ },
273
+ "vision_tower.blocks.12.mlp.down_proj": {
274
+ "group_size": 64,
275
+ "bits": 4
276
+ },
277
+ "vision_tower.blocks.13.attn.qkv": {
278
+ "group_size": 64,
279
+ "bits": 4
280
+ },
281
+ "vision_tower.blocks.13.attn.proj": {
282
+ "group_size": 64,
283
+ "bits": 4
284
+ },
285
+ "vision_tower.blocks.13.mlp.gate_proj": {
286
+ "group_size": 64,
287
+ "bits": 4
288
+ },
289
+ "vision_tower.blocks.13.mlp.up_proj": {
290
+ "group_size": 64,
291
+ "bits": 4
292
+ },
293
+ "vision_tower.blocks.13.mlp.down_proj": {
294
+ "group_size": 64,
295
+ "bits": 4
296
+ },
297
+ "vision_tower.blocks.14.attn.qkv": {
298
+ "group_size": 64,
299
+ "bits": 4
300
+ },
301
+ "vision_tower.blocks.14.attn.proj": {
302
+ "group_size": 64,
303
+ "bits": 4
304
+ },
305
+ "vision_tower.blocks.14.mlp.gate_proj": {
306
+ "group_size": 64,
307
+ "bits": 4
308
+ },
309
+ "vision_tower.blocks.14.mlp.up_proj": {
310
+ "group_size": 64,
311
+ "bits": 4
312
+ },
313
+ "vision_tower.blocks.14.mlp.down_proj": {
314
+ "group_size": 64,
315
+ "bits": 4
316
+ },
317
+ "vision_tower.blocks.15.attn.qkv": {
318
+ "group_size": 64,
319
+ "bits": 4
320
+ },
321
+ "vision_tower.blocks.15.attn.proj": {
322
+ "group_size": 64,
323
+ "bits": 4
324
+ },
325
+ "vision_tower.blocks.15.mlp.gate_proj": {
326
+ "group_size": 64,
327
+ "bits": 4
328
+ },
329
+ "vision_tower.blocks.15.mlp.up_proj": {
330
+ "group_size": 64,
331
+ "bits": 4
332
+ },
333
+ "vision_tower.blocks.15.mlp.down_proj": {
334
+ "group_size": 64,
335
+ "bits": 4
336
+ },
337
+ "vision_tower.blocks.16.attn.qkv": {
338
+ "group_size": 64,
339
+ "bits": 4
340
+ },
341
+ "vision_tower.blocks.16.attn.proj": {
342
+ "group_size": 64,
343
+ "bits": 4
344
+ },
345
+ "vision_tower.blocks.16.mlp.gate_proj": {
346
+ "group_size": 64,
347
+ "bits": 4
348
+ },
349
+ "vision_tower.blocks.16.mlp.up_proj": {
350
+ "group_size": 64,
351
+ "bits": 4
352
+ },
353
+ "vision_tower.blocks.16.mlp.down_proj": {
354
+ "group_size": 64,
355
+ "bits": 4
356
+ },
357
+ "vision_tower.blocks.17.attn.qkv": {
358
+ "group_size": 64,
359
+ "bits": 4
360
+ },
361
+ "vision_tower.blocks.17.attn.proj": {
362
+ "group_size": 64,
363
+ "bits": 4
364
+ },
365
+ "vision_tower.blocks.17.mlp.gate_proj": {
366
+ "group_size": 64,
367
+ "bits": 4
368
+ },
369
+ "vision_tower.blocks.17.mlp.up_proj": {
370
+ "group_size": 64,
371
+ "bits": 4
372
+ },
373
+ "vision_tower.blocks.17.mlp.down_proj": {
374
+ "group_size": 64,
375
+ "bits": 4
376
+ },
377
+ "vision_tower.blocks.18.attn.qkv": {
378
+ "group_size": 64,
379
+ "bits": 4
380
+ },
381
+ "vision_tower.blocks.18.attn.proj": {
382
+ "group_size": 64,
383
+ "bits": 4
384
+ },
385
+ "vision_tower.blocks.18.mlp.gate_proj": {
386
+ "group_size": 64,
387
+ "bits": 4
388
+ },
389
+ "vision_tower.blocks.18.mlp.up_proj": {
390
+ "group_size": 64,
391
+ "bits": 4
392
+ },
393
+ "vision_tower.blocks.18.mlp.down_proj": {
394
+ "group_size": 64,
395
+ "bits": 4
396
+ },
397
+ "vision_tower.blocks.19.attn.qkv": {
398
+ "group_size": 64,
399
+ "bits": 4
400
+ },
401
+ "vision_tower.blocks.19.attn.proj": {
402
+ "group_size": 64,
403
+ "bits": 4
404
+ },
405
+ "vision_tower.blocks.19.mlp.gate_proj": {
406
+ "group_size": 64,
407
+ "bits": 4
408
+ },
409
+ "vision_tower.blocks.19.mlp.up_proj": {
410
+ "group_size": 64,
411
+ "bits": 4
412
+ },
413
+ "vision_tower.blocks.19.mlp.down_proj": {
414
+ "group_size": 64,
415
+ "bits": 4
416
+ },
417
+ "vision_tower.blocks.20.attn.qkv": {
418
+ "group_size": 64,
419
+ "bits": 4
420
+ },
421
+ "vision_tower.blocks.20.attn.proj": {
422
+ "group_size": 64,
423
+ "bits": 4
424
+ },
425
+ "vision_tower.blocks.20.mlp.gate_proj": {
426
+ "group_size": 64,
427
+ "bits": 4
428
+ },
429
+ "vision_tower.blocks.20.mlp.up_proj": {
430
+ "group_size": 64,
431
+ "bits": 4
432
+ },
433
+ "vision_tower.blocks.20.mlp.down_proj": {
434
+ "group_size": 64,
435
+ "bits": 4
436
+ },
437
+ "vision_tower.blocks.21.attn.qkv": {
438
+ "group_size": 64,
439
+ "bits": 4
440
+ },
441
+ "vision_tower.blocks.21.attn.proj": {
442
+ "group_size": 64,
443
+ "bits": 4
444
+ },
445
+ "vision_tower.blocks.21.mlp.gate_proj": {
446
+ "group_size": 64,
447
+ "bits": 4
448
+ },
449
+ "vision_tower.blocks.21.mlp.up_proj": {
450
+ "group_size": 64,
451
+ "bits": 4
452
+ },
453
+ "vision_tower.blocks.21.mlp.down_proj": {
454
+ "group_size": 64,
455
+ "bits": 4
456
+ },
457
+ "vision_tower.blocks.22.attn.qkv": {
458
+ "group_size": 64,
459
+ "bits": 4
460
+ },
461
+ "vision_tower.blocks.22.attn.proj": {
462
+ "group_size": 64,
463
+ "bits": 4
464
+ },
465
+ "vision_tower.blocks.22.mlp.gate_proj": {
466
+ "group_size": 64,
467
+ "bits": 4
468
+ },
469
+ "vision_tower.blocks.22.mlp.up_proj": {
470
+ "group_size": 64,
471
+ "bits": 4
472
+ },
473
+ "vision_tower.blocks.22.mlp.down_proj": {
474
+ "group_size": 64,
475
+ "bits": 4
476
+ },
477
+ "vision_tower.blocks.23.attn.qkv": {
478
+ "group_size": 64,
479
+ "bits": 4
480
+ },
481
+ "vision_tower.blocks.23.attn.proj": {
482
+ "group_size": 64,
483
+ "bits": 4
484
+ },
485
+ "vision_tower.blocks.23.mlp.gate_proj": {
486
+ "group_size": 64,
487
+ "bits": 4
488
+ },
489
+ "vision_tower.blocks.23.mlp.up_proj": {
490
+ "group_size": 64,
491
+ "bits": 4
492
+ },
493
+ "vision_tower.blocks.23.mlp.down_proj": {
494
+ "group_size": 64,
495
+ "bits": 4
496
+ },
497
+ "vision_tower.merger.proj": {
498
+ "group_size": 64,
499
+ "bits": 4
500
+ },
501
+ "vision_tower.merger.gate_proj": {
502
+ "group_size": 64,
503
+ "bits": 4
504
+ },
505
+ "vision_tower.merger.up_proj": {
506
+ "group_size": 64,
507
+ "bits": 4
508
+ },
509
+ "vision_tower.merger.down_proj": {
510
+ "group_size": 64,
511
+ "bits": 4
512
+ },
513
+ "language_model.model.embed_tokens": {
514
+ "group_size": 64,
515
+ "bits": 4
516
+ },
517
+ "language_model.model.layers.0.self_attn.q_proj": {
518
+ "group_size": 64,
519
+ "bits": 4
520
+ },
521
+ "language_model.model.layers.0.self_attn.k_proj": {
522
+ "group_size": 64,
523
+ "bits": 4
524
+ },
525
+ "language_model.model.layers.0.self_attn.v_proj": {
526
+ "group_size": 64,
527
+ "bits": 4
528
+ },
529
+ "language_model.model.layers.0.self_attn.o_proj": {
530
+ "group_size": 64,
531
+ "bits": 4
532
+ },
533
+ "language_model.model.layers.0.mlp.gate_up_proj": {
534
+ "group_size": 64,
535
+ "bits": 4
536
+ },
537
+ "language_model.model.layers.0.mlp.down_proj": {
538
+ "group_size": 64,
539
+ "bits": 4
540
+ },
541
+ "language_model.model.layers.1.self_attn.q_proj": {
542
+ "group_size": 64,
543
+ "bits": 4
544
+ },
545
+ "language_model.model.layers.1.self_attn.k_proj": {
546
+ "group_size": 64,
547
+ "bits": 4
548
+ },
549
+ "language_model.model.layers.1.self_attn.v_proj": {
550
+ "group_size": 64,
551
+ "bits": 4
552
+ },
553
+ "language_model.model.layers.1.self_attn.o_proj": {
554
+ "group_size": 64,
555
+ "bits": 4
556
+ },
557
+ "language_model.model.layers.1.mlp.gate_up_proj": {
558
+ "group_size": 64,
559
+ "bits": 4
560
+ },
561
+ "language_model.model.layers.1.mlp.down_proj": {
562
+ "group_size": 64,
563
+ "bits": 4
564
+ },
565
+ "language_model.model.layers.2.self_attn.q_proj": {
566
+ "group_size": 64,
567
+ "bits": 4
568
+ },
569
+ "language_model.model.layers.2.self_attn.k_proj": {
570
+ "group_size": 64,
571
+ "bits": 4
572
+ },
573
+ "language_model.model.layers.2.self_attn.v_proj": {
574
+ "group_size": 64,
575
+ "bits": 4
576
+ },
577
+ "language_model.model.layers.2.self_attn.o_proj": {
578
+ "group_size": 64,
579
+ "bits": 4
580
+ },
581
+ "language_model.model.layers.2.mlp.gate_up_proj": {
582
+ "group_size": 64,
583
+ "bits": 4
584
+ },
585
+ "language_model.model.layers.2.mlp.down_proj": {
586
+ "group_size": 64,
587
+ "bits": 4
588
+ },
589
+ "language_model.model.layers.3.self_attn.q_proj": {
590
+ "group_size": 64,
591
+ "bits": 4
592
+ },
593
+ "language_model.model.layers.3.self_attn.k_proj": {
594
+ "group_size": 64,
595
+ "bits": 4
596
+ },
597
+ "language_model.model.layers.3.self_attn.v_proj": {
598
+ "group_size": 64,
599
+ "bits": 4
600
+ },
601
+ "language_model.model.layers.3.self_attn.o_proj": {
602
+ "group_size": 64,
603
+ "bits": 4
604
+ },
605
+ "language_model.model.layers.3.mlp.gate_up_proj": {
606
+ "group_size": 64,
607
+ "bits": 4
608
+ },
609
+ "language_model.model.layers.3.mlp.down_proj": {
610
+ "group_size": 64,
611
+ "bits": 4
612
+ },
613
+ "language_model.model.layers.4.self_attn.q_proj": {
614
+ "group_size": 64,
615
+ "bits": 4
616
+ },
617
+ "language_model.model.layers.4.self_attn.k_proj": {
618
+ "group_size": 64,
619
+ "bits": 4
620
+ },
621
+ "language_model.model.layers.4.self_attn.v_proj": {
622
+ "group_size": 64,
623
+ "bits": 4
624
+ },
625
+ "language_model.model.layers.4.self_attn.o_proj": {
626
+ "group_size": 64,
627
+ "bits": 4
628
+ },
629
+ "language_model.model.layers.4.mlp.gate_up_proj": {
630
+ "group_size": 64,
631
+ "bits": 4
632
+ },
633
+ "language_model.model.layers.4.mlp.down_proj": {
634
+ "group_size": 64,
635
+ "bits": 4
636
+ },
637
+ "language_model.model.layers.5.self_attn.q_proj": {
638
+ "group_size": 64,
639
+ "bits": 4
640
+ },
641
+ "language_model.model.layers.5.self_attn.k_proj": {
642
+ "group_size": 64,
643
+ "bits": 4
644
+ },
645
+ "language_model.model.layers.5.self_attn.v_proj": {
646
+ "group_size": 64,
647
+ "bits": 4
648
+ },
649
+ "language_model.model.layers.5.self_attn.o_proj": {
650
+ "group_size": 64,
651
+ "bits": 4
652
+ },
653
+ "language_model.model.layers.5.mlp.gate_up_proj": {
654
+ "group_size": 64,
655
+ "bits": 4
656
+ },
657
+ "language_model.model.layers.5.mlp.down_proj": {
658
+ "group_size": 64,
659
+ "bits": 4
660
+ },
661
+ "language_model.model.layers.6.self_attn.q_proj": {
662
+ "group_size": 64,
663
+ "bits": 4
664
+ },
665
+ "language_model.model.layers.6.self_attn.k_proj": {
666
+ "group_size": 64,
667
+ "bits": 4
668
+ },
669
+ "language_model.model.layers.6.self_attn.v_proj": {
670
+ "group_size": 64,
671
+ "bits": 4
672
+ },
673
+ "language_model.model.layers.6.self_attn.o_proj": {
674
+ "group_size": 64,
675
+ "bits": 4
676
+ },
677
+ "language_model.model.layers.6.mlp.gate_up_proj": {
678
+ "group_size": 64,
679
+ "bits": 4
680
+ },
681
+ "language_model.model.layers.6.mlp.down_proj": {
682
+ "group_size": 64,
683
+ "bits": 4
684
+ },
685
+ "language_model.model.layers.7.self_attn.q_proj": {
686
+ "group_size": 64,
687
+ "bits": 4
688
+ },
689
+ "language_model.model.layers.7.self_attn.k_proj": {
690
+ "group_size": 64,
691
+ "bits": 4
692
+ },
693
+ "language_model.model.layers.7.self_attn.v_proj": {
694
+ "group_size": 64,
695
+ "bits": 4
696
+ },
697
+ "language_model.model.layers.7.self_attn.o_proj": {
698
+ "group_size": 64,
699
+ "bits": 4
700
+ },
701
+ "language_model.model.layers.7.mlp.gate_up_proj": {
702
+ "group_size": 64,
703
+ "bits": 4
704
+ },
705
+ "language_model.model.layers.7.mlp.down_proj": {
706
+ "group_size": 64,
707
+ "bits": 4
708
+ },
709
+ "language_model.model.layers.8.self_attn.q_proj": {
710
+ "group_size": 64,
711
+ "bits": 4
712
+ },
713
+ "language_model.model.layers.8.self_attn.k_proj": {
714
+ "group_size": 64,
715
+ "bits": 4
716
+ },
717
+ "language_model.model.layers.8.self_attn.v_proj": {
718
+ "group_size": 64,
719
+ "bits": 4
720
+ },
721
+ "language_model.model.layers.8.self_attn.o_proj": {
722
+ "group_size": 64,
723
+ "bits": 4
724
+ },
725
+ "language_model.model.layers.8.mlp.gate_up_proj": {
726
+ "group_size": 64,
727
+ "bits": 4
728
+ },
729
+ "language_model.model.layers.8.mlp.down_proj": {
730
+ "group_size": 64,
731
+ "bits": 4
732
+ },
733
+ "language_model.model.layers.9.self_attn.q_proj": {
734
+ "group_size": 64,
735
+ "bits": 4
736
+ },
737
+ "language_model.model.layers.9.self_attn.k_proj": {
738
+ "group_size": 64,
739
+ "bits": 4
740
+ },
741
+ "language_model.model.layers.9.self_attn.v_proj": {
742
+ "group_size": 64,
743
+ "bits": 4
744
+ },
745
+ "language_model.model.layers.9.self_attn.o_proj": {
746
+ "group_size": 64,
747
+ "bits": 4
748
+ },
749
+ "language_model.model.layers.9.mlp.gate_up_proj": {
750
+ "group_size": 64,
751
+ "bits": 4
752
+ },
753
+ "language_model.model.layers.9.mlp.down_proj": {
754
+ "group_size": 64,
755
+ "bits": 4
756
+ },
757
+ "language_model.model.layers.10.self_attn.q_proj": {
758
+ "group_size": 64,
759
+ "bits": 4
760
+ },
761
+ "language_model.model.layers.10.self_attn.k_proj": {
762
+ "group_size": 64,
763
+ "bits": 4
764
+ },
765
+ "language_model.model.layers.10.self_attn.v_proj": {
766
+ "group_size": 64,
767
+ "bits": 4
768
+ },
769
+ "language_model.model.layers.10.self_attn.o_proj": {
770
+ "group_size": 64,
771
+ "bits": 4
772
+ },
773
+ "language_model.model.layers.10.mlp.gate_up_proj": {
774
+ "group_size": 64,
775
+ "bits": 4
776
+ },
777
+ "language_model.model.layers.10.mlp.down_proj": {
778
+ "group_size": 64,
779
+ "bits": 4
780
+ },
781
+ "language_model.model.layers.11.self_attn.q_proj": {
782
+ "group_size": 64,
783
+ "bits": 4
784
+ },
785
+ "language_model.model.layers.11.self_attn.k_proj": {
786
+ "group_size": 64,
787
+ "bits": 4
788
+ },
789
+ "language_model.model.layers.11.self_attn.v_proj": {
790
+ "group_size": 64,
791
+ "bits": 4
792
+ },
793
+ "language_model.model.layers.11.self_attn.o_proj": {
794
+ "group_size": 64,
795
+ "bits": 4
796
+ },
797
+ "language_model.model.layers.11.mlp.gate_up_proj": {
798
+ "group_size": 64,
799
+ "bits": 4
800
+ },
801
+ "language_model.model.layers.11.mlp.down_proj": {
802
+ "group_size": 64,
803
+ "bits": 4
804
+ },
805
+ "language_model.model.layers.12.self_attn.q_proj": {
806
+ "group_size": 64,
807
+ "bits": 4
808
+ },
809
+ "language_model.model.layers.12.self_attn.k_proj": {
810
+ "group_size": 64,
811
+ "bits": 4
812
+ },
813
+ "language_model.model.layers.12.self_attn.v_proj": {
814
+ "group_size": 64,
815
+ "bits": 4
816
+ },
817
+ "language_model.model.layers.12.self_attn.o_proj": {
818
+ "group_size": 64,
819
+ "bits": 4
820
+ },
821
+ "language_model.model.layers.12.mlp.gate_up_proj": {
822
+ "group_size": 64,
823
+ "bits": 4
824
+ },
825
+ "language_model.model.layers.12.mlp.down_proj": {
826
+ "group_size": 64,
827
+ "bits": 4
828
+ },
829
+ "language_model.model.layers.13.self_attn.q_proj": {
830
+ "group_size": 64,
831
+ "bits": 4
832
+ },
833
+ "language_model.model.layers.13.self_attn.k_proj": {
834
+ "group_size": 64,
835
+ "bits": 4
836
+ },
837
+ "language_model.model.layers.13.self_attn.v_proj": {
838
+ "group_size": 64,
839
+ "bits": 4
840
+ },
841
+ "language_model.model.layers.13.self_attn.o_proj": {
842
+ "group_size": 64,
843
+ "bits": 4
844
+ },
845
+ "language_model.model.layers.13.mlp.gate_up_proj": {
846
+ "group_size": 64,
847
+ "bits": 4
848
+ },
849
+ "language_model.model.layers.13.mlp.down_proj": {
850
+ "group_size": 64,
851
+ "bits": 4
852
+ },
853
+ "language_model.model.layers.14.self_attn.q_proj": {
854
+ "group_size": 64,
855
+ "bits": 4
856
+ },
857
+ "language_model.model.layers.14.self_attn.k_proj": {
858
+ "group_size": 64,
859
+ "bits": 4
860
+ },
861
+ "language_model.model.layers.14.self_attn.v_proj": {
862
+ "group_size": 64,
863
+ "bits": 4
864
+ },
865
+ "language_model.model.layers.14.self_attn.o_proj": {
866
+ "group_size": 64,
867
+ "bits": 4
868
+ },
869
+ "language_model.model.layers.14.mlp.gate_up_proj": {
870
+ "group_size": 64,
871
+ "bits": 4
872
+ },
873
+ "language_model.model.layers.14.mlp.down_proj": {
874
+ "group_size": 64,
875
+ "bits": 4
876
+ },
877
+ "language_model.model.layers.15.self_attn.q_proj": {
878
+ "group_size": 64,
879
+ "bits": 4
880
+ },
881
+ "language_model.model.layers.15.self_attn.k_proj": {
882
+ "group_size": 64,
883
+ "bits": 4
884
+ },
885
+ "language_model.model.layers.15.self_attn.v_proj": {
886
+ "group_size": 64,
887
+ "bits": 4
888
+ },
889
+ "language_model.model.layers.15.self_attn.o_proj": {
890
+ "group_size": 64,
891
+ "bits": 4
892
+ },
893
+ "language_model.model.layers.15.mlp.gate_up_proj": {
894
+ "group_size": 64,
895
+ "bits": 4
896
+ },
897
+ "language_model.model.layers.15.mlp.down_proj": {
898
+ "group_size": 64,
899
+ "bits": 4
900
+ },
901
+ "language_model.lm_head": {
902
+ "group_size": 64,
903
+ "bits": 4
904
+ }
905
+ },
906
+ "quantization_config": {
907
+ "group_size": 64,
908
+ "bits": 4,
909
+ "mode": "affine",
910
+ "vision_tower.blocks.0.attn.qkv": {
911
+ "group_size": 64,
912
+ "bits": 4
913
+ },
914
+ "vision_tower.blocks.0.attn.proj": {
915
+ "group_size": 64,
916
+ "bits": 4
917
+ },
918
+ "vision_tower.blocks.0.mlp.gate_proj": {
919
+ "group_size": 64,
920
+ "bits": 4
921
+ },
922
+ "vision_tower.blocks.0.mlp.up_proj": {
923
+ "group_size": 64,
924
+ "bits": 4
925
+ },
926
+ "vision_tower.blocks.0.mlp.down_proj": {
927
+ "group_size": 64,
928
+ "bits": 4
929
+ },
930
+ "vision_tower.blocks.1.attn.qkv": {
931
+ "group_size": 64,
932
+ "bits": 4
933
+ },
934
+ "vision_tower.blocks.1.attn.proj": {
935
+ "group_size": 64,
936
+ "bits": 4
937
+ },
938
+ "vision_tower.blocks.1.mlp.gate_proj": {
939
+ "group_size": 64,
940
+ "bits": 4
941
+ },
942
+ "vision_tower.blocks.1.mlp.up_proj": {
943
+ "group_size": 64,
944
+ "bits": 4
945
+ },
946
+ "vision_tower.blocks.1.mlp.down_proj": {
947
+ "group_size": 64,
948
+ "bits": 4
949
+ },
950
+ "vision_tower.blocks.2.attn.qkv": {
951
+ "group_size": 64,
952
+ "bits": 4
953
+ },
954
+ "vision_tower.blocks.2.attn.proj": {
955
+ "group_size": 64,
956
+ "bits": 4
957
+ },
958
+ "vision_tower.blocks.2.mlp.gate_proj": {
959
+ "group_size": 64,
960
+ "bits": 4
961
+ },
962
+ "vision_tower.blocks.2.mlp.up_proj": {
963
+ "group_size": 64,
964
+ "bits": 4
965
+ },
966
+ "vision_tower.blocks.2.mlp.down_proj": {
967
+ "group_size": 64,
968
+ "bits": 4
969
+ },
970
+ "vision_tower.blocks.3.attn.qkv": {
971
+ "group_size": 64,
972
+ "bits": 4
973
+ },
974
+ "vision_tower.blocks.3.attn.proj": {
975
+ "group_size": 64,
976
+ "bits": 4
977
+ },
978
+ "vision_tower.blocks.3.mlp.gate_proj": {
979
+ "group_size": 64,
980
+ "bits": 4
981
+ },
982
+ "vision_tower.blocks.3.mlp.up_proj": {
983
+ "group_size": 64,
984
+ "bits": 4
985
+ },
986
+ "vision_tower.blocks.3.mlp.down_proj": {
987
+ "group_size": 64,
988
+ "bits": 4
989
+ },
990
+ "vision_tower.blocks.4.attn.qkv": {
991
+ "group_size": 64,
992
+ "bits": 4
993
+ },
994
+ "vision_tower.blocks.4.attn.proj": {
995
+ "group_size": 64,
996
+ "bits": 4
997
+ },
998
+ "vision_tower.blocks.4.mlp.gate_proj": {
999
+ "group_size": 64,
1000
+ "bits": 4
1001
+ },
1002
+ "vision_tower.blocks.4.mlp.up_proj": {
1003
+ "group_size": 64,
1004
+ "bits": 4
1005
+ },
1006
+ "vision_tower.blocks.4.mlp.down_proj": {
1007
+ "group_size": 64,
1008
+ "bits": 4
1009
+ },
1010
+ "vision_tower.blocks.5.attn.qkv": {
1011
+ "group_size": 64,
1012
+ "bits": 4
1013
+ },
1014
+ "vision_tower.blocks.5.attn.proj": {
1015
+ "group_size": 64,
1016
+ "bits": 4
1017
+ },
1018
+ "vision_tower.blocks.5.mlp.gate_proj": {
1019
+ "group_size": 64,
1020
+ "bits": 4
1021
+ },
1022
+ "vision_tower.blocks.5.mlp.up_proj": {
1023
+ "group_size": 64,
1024
+ "bits": 4
1025
+ },
1026
+ "vision_tower.blocks.5.mlp.down_proj": {
1027
+ "group_size": 64,
1028
+ "bits": 4
1029
+ },
1030
+ "vision_tower.blocks.6.attn.qkv": {
1031
+ "group_size": 64,
1032
+ "bits": 4
1033
+ },
1034
+ "vision_tower.blocks.6.attn.proj": {
1035
+ "group_size": 64,
1036
+ "bits": 4
1037
+ },
1038
+ "vision_tower.blocks.6.mlp.gate_proj": {
1039
+ "group_size": 64,
1040
+ "bits": 4
1041
+ },
1042
+ "vision_tower.blocks.6.mlp.up_proj": {
1043
+ "group_size": 64,
1044
+ "bits": 4
1045
+ },
1046
+ "vision_tower.blocks.6.mlp.down_proj": {
1047
+ "group_size": 64,
1048
+ "bits": 4
1049
+ },
1050
+ "vision_tower.blocks.7.attn.qkv": {
1051
+ "group_size": 64,
1052
+ "bits": 4
1053
+ },
1054
+ "vision_tower.blocks.7.attn.proj": {
1055
+ "group_size": 64,
1056
+ "bits": 4
1057
+ },
1058
+ "vision_tower.blocks.7.mlp.gate_proj": {
1059
+ "group_size": 64,
1060
+ "bits": 4
1061
+ },
1062
+ "vision_tower.blocks.7.mlp.up_proj": {
1063
+ "group_size": 64,
1064
+ "bits": 4
1065
+ },
1066
+ "vision_tower.blocks.7.mlp.down_proj": {
1067
+ "group_size": 64,
1068
+ "bits": 4
1069
+ },
1070
+ "vision_tower.blocks.8.attn.qkv": {
1071
+ "group_size": 64,
1072
+ "bits": 4
1073
+ },
1074
+ "vision_tower.blocks.8.attn.proj": {
1075
+ "group_size": 64,
1076
+ "bits": 4
1077
+ },
1078
+ "vision_tower.blocks.8.mlp.gate_proj": {
1079
+ "group_size": 64,
1080
+ "bits": 4
1081
+ },
1082
+ "vision_tower.blocks.8.mlp.up_proj": {
1083
+ "group_size": 64,
1084
+ "bits": 4
1085
+ },
1086
+ "vision_tower.blocks.8.mlp.down_proj": {
1087
+ "group_size": 64,
1088
+ "bits": 4
1089
+ },
1090
+ "vision_tower.blocks.9.attn.qkv": {
1091
+ "group_size": 64,
1092
+ "bits": 4
1093
+ },
1094
+ "vision_tower.blocks.9.attn.proj": {
1095
+ "group_size": 64,
1096
+ "bits": 4
1097
+ },
1098
+ "vision_tower.blocks.9.mlp.gate_proj": {
1099
+ "group_size": 64,
1100
+ "bits": 4
1101
+ },
1102
+ "vision_tower.blocks.9.mlp.up_proj": {
1103
+ "group_size": 64,
1104
+ "bits": 4
1105
+ },
1106
+ "vision_tower.blocks.9.mlp.down_proj": {
1107
+ "group_size": 64,
1108
+ "bits": 4
1109
+ },
1110
+ "vision_tower.blocks.10.attn.qkv": {
1111
+ "group_size": 64,
1112
+ "bits": 4
1113
+ },
1114
+ "vision_tower.blocks.10.attn.proj": {
1115
+ "group_size": 64,
1116
+ "bits": 4
1117
+ },
1118
+ "vision_tower.blocks.10.mlp.gate_proj": {
1119
+ "group_size": 64,
1120
+ "bits": 4
1121
+ },
1122
+ "vision_tower.blocks.10.mlp.up_proj": {
1123
+ "group_size": 64,
1124
+ "bits": 4
1125
+ },
1126
+ "vision_tower.blocks.10.mlp.down_proj": {
1127
+ "group_size": 64,
1128
+ "bits": 4
1129
+ },
1130
+ "vision_tower.blocks.11.attn.qkv": {
1131
+ "group_size": 64,
1132
+ "bits": 4
1133
+ },
1134
+ "vision_tower.blocks.11.attn.proj": {
1135
+ "group_size": 64,
1136
+ "bits": 4
1137
+ },
1138
+ "vision_tower.blocks.11.mlp.gate_proj": {
1139
+ "group_size": 64,
1140
+ "bits": 4
1141
+ },
1142
+ "vision_tower.blocks.11.mlp.up_proj": {
1143
+ "group_size": 64,
1144
+ "bits": 4
1145
+ },
1146
+ "vision_tower.blocks.11.mlp.down_proj": {
1147
+ "group_size": 64,
1148
+ "bits": 4
1149
+ },
1150
+ "vision_tower.blocks.12.attn.qkv": {
1151
+ "group_size": 64,
1152
+ "bits": 4
1153
+ },
1154
+ "vision_tower.blocks.12.attn.proj": {
1155
+ "group_size": 64,
1156
+ "bits": 4
1157
+ },
1158
+ "vision_tower.blocks.12.mlp.gate_proj": {
1159
+ "group_size": 64,
1160
+ "bits": 4
1161
+ },
1162
+ "vision_tower.blocks.12.mlp.up_proj": {
1163
+ "group_size": 64,
1164
+ "bits": 4
1165
+ },
1166
+ "vision_tower.blocks.12.mlp.down_proj": {
1167
+ "group_size": 64,
1168
+ "bits": 4
1169
+ },
1170
+ "vision_tower.blocks.13.attn.qkv": {
1171
+ "group_size": 64,
1172
+ "bits": 4
1173
+ },
1174
+ "vision_tower.blocks.13.attn.proj": {
1175
+ "group_size": 64,
1176
+ "bits": 4
1177
+ },
1178
+ "vision_tower.blocks.13.mlp.gate_proj": {
1179
+ "group_size": 64,
1180
+ "bits": 4
1181
+ },
1182
+ "vision_tower.blocks.13.mlp.up_proj": {
1183
+ "group_size": 64,
1184
+ "bits": 4
1185
+ },
1186
+ "vision_tower.blocks.13.mlp.down_proj": {
1187
+ "group_size": 64,
1188
+ "bits": 4
1189
+ },
1190
+ "vision_tower.blocks.14.attn.qkv": {
1191
+ "group_size": 64,
1192
+ "bits": 4
1193
+ },
1194
+ "vision_tower.blocks.14.attn.proj": {
1195
+ "group_size": 64,
1196
+ "bits": 4
1197
+ },
1198
+ "vision_tower.blocks.14.mlp.gate_proj": {
1199
+ "group_size": 64,
1200
+ "bits": 4
1201
+ },
1202
+ "vision_tower.blocks.14.mlp.up_proj": {
1203
+ "group_size": 64,
1204
+ "bits": 4
1205
+ },
1206
+ "vision_tower.blocks.14.mlp.down_proj": {
1207
+ "group_size": 64,
1208
+ "bits": 4
1209
+ },
1210
+ "vision_tower.blocks.15.attn.qkv": {
1211
+ "group_size": 64,
1212
+ "bits": 4
1213
+ },
1214
+ "vision_tower.blocks.15.attn.proj": {
1215
+ "group_size": 64,
1216
+ "bits": 4
1217
+ },
1218
+ "vision_tower.blocks.15.mlp.gate_proj": {
1219
+ "group_size": 64,
1220
+ "bits": 4
1221
+ },
1222
+ "vision_tower.blocks.15.mlp.up_proj": {
1223
+ "group_size": 64,
1224
+ "bits": 4
1225
+ },
1226
+ "vision_tower.blocks.15.mlp.down_proj": {
1227
+ "group_size": 64,
1228
+ "bits": 4
1229
+ },
1230
+ "vision_tower.blocks.16.attn.qkv": {
1231
+ "group_size": 64,
1232
+ "bits": 4
1233
+ },
1234
+ "vision_tower.blocks.16.attn.proj": {
1235
+ "group_size": 64,
1236
+ "bits": 4
1237
+ },
1238
+ "vision_tower.blocks.16.mlp.gate_proj": {
1239
+ "group_size": 64,
1240
+ "bits": 4
1241
+ },
1242
+ "vision_tower.blocks.16.mlp.up_proj": {
1243
+ "group_size": 64,
1244
+ "bits": 4
1245
+ },
1246
+ "vision_tower.blocks.16.mlp.down_proj": {
1247
+ "group_size": 64,
1248
+ "bits": 4
1249
+ },
1250
+ "vision_tower.blocks.17.attn.qkv": {
1251
+ "group_size": 64,
1252
+ "bits": 4
1253
+ },
1254
+ "vision_tower.blocks.17.attn.proj": {
1255
+ "group_size": 64,
1256
+ "bits": 4
1257
+ },
1258
+ "vision_tower.blocks.17.mlp.gate_proj": {
1259
+ "group_size": 64,
1260
+ "bits": 4
1261
+ },
1262
+ "vision_tower.blocks.17.mlp.up_proj": {
1263
+ "group_size": 64,
1264
+ "bits": 4
1265
+ },
1266
+ "vision_tower.blocks.17.mlp.down_proj": {
1267
+ "group_size": 64,
1268
+ "bits": 4
1269
+ },
1270
+ "vision_tower.blocks.18.attn.qkv": {
1271
+ "group_size": 64,
1272
+ "bits": 4
1273
+ },
1274
+ "vision_tower.blocks.18.attn.proj": {
1275
+ "group_size": 64,
1276
+ "bits": 4
1277
+ },
1278
+ "vision_tower.blocks.18.mlp.gate_proj": {
1279
+ "group_size": 64,
1280
+ "bits": 4
1281
+ },
1282
+ "vision_tower.blocks.18.mlp.up_proj": {
1283
+ "group_size": 64,
1284
+ "bits": 4
1285
+ },
1286
+ "vision_tower.blocks.18.mlp.down_proj": {
1287
+ "group_size": 64,
1288
+ "bits": 4
1289
+ },
1290
+ "vision_tower.blocks.19.attn.qkv": {
1291
+ "group_size": 64,
1292
+ "bits": 4
1293
+ },
1294
+ "vision_tower.blocks.19.attn.proj": {
1295
+ "group_size": 64,
1296
+ "bits": 4
1297
+ },
1298
+ "vision_tower.blocks.19.mlp.gate_proj": {
1299
+ "group_size": 64,
1300
+ "bits": 4
1301
+ },
1302
+ "vision_tower.blocks.19.mlp.up_proj": {
1303
+ "group_size": 64,
1304
+ "bits": 4
1305
+ },
1306
+ "vision_tower.blocks.19.mlp.down_proj": {
1307
+ "group_size": 64,
1308
+ "bits": 4
1309
+ },
1310
+ "vision_tower.blocks.20.attn.qkv": {
1311
+ "group_size": 64,
1312
+ "bits": 4
1313
+ },
1314
+ "vision_tower.blocks.20.attn.proj": {
1315
+ "group_size": 64,
1316
+ "bits": 4
1317
+ },
1318
+ "vision_tower.blocks.20.mlp.gate_proj": {
1319
+ "group_size": 64,
1320
+ "bits": 4
1321
+ },
1322
+ "vision_tower.blocks.20.mlp.up_proj": {
1323
+ "group_size": 64,
1324
+ "bits": 4
1325
+ },
1326
+ "vision_tower.blocks.20.mlp.down_proj": {
1327
+ "group_size": 64,
1328
+ "bits": 4
1329
+ },
1330
+ "vision_tower.blocks.21.attn.qkv": {
1331
+ "group_size": 64,
1332
+ "bits": 4
1333
+ },
1334
+ "vision_tower.blocks.21.attn.proj": {
1335
+ "group_size": 64,
1336
+ "bits": 4
1337
+ },
1338
+ "vision_tower.blocks.21.mlp.gate_proj": {
1339
+ "group_size": 64,
1340
+ "bits": 4
1341
+ },
1342
+ "vision_tower.blocks.21.mlp.up_proj": {
1343
+ "group_size": 64,
1344
+ "bits": 4
1345
+ },
1346
+ "vision_tower.blocks.21.mlp.down_proj": {
1347
+ "group_size": 64,
1348
+ "bits": 4
1349
+ },
1350
+ "vision_tower.blocks.22.attn.qkv": {
1351
+ "group_size": 64,
1352
+ "bits": 4
1353
+ },
1354
+ "vision_tower.blocks.22.attn.proj": {
1355
+ "group_size": 64,
1356
+ "bits": 4
1357
+ },
1358
+ "vision_tower.blocks.22.mlp.gate_proj": {
1359
+ "group_size": 64,
1360
+ "bits": 4
1361
+ },
1362
+ "vision_tower.blocks.22.mlp.up_proj": {
1363
+ "group_size": 64,
1364
+ "bits": 4
1365
+ },
1366
+ "vision_tower.blocks.22.mlp.down_proj": {
1367
+ "group_size": 64,
1368
+ "bits": 4
1369
+ },
1370
+ "vision_tower.blocks.23.attn.qkv": {
1371
+ "group_size": 64,
1372
+ "bits": 4
1373
+ },
1374
+ "vision_tower.blocks.23.attn.proj": {
1375
+ "group_size": 64,
1376
+ "bits": 4
1377
+ },
1378
+ "vision_tower.blocks.23.mlp.gate_proj": {
1379
+ "group_size": 64,
1380
+ "bits": 4
1381
+ },
1382
+ "vision_tower.blocks.23.mlp.up_proj": {
1383
+ "group_size": 64,
1384
+ "bits": 4
1385
+ },
1386
+ "vision_tower.blocks.23.mlp.down_proj": {
1387
+ "group_size": 64,
1388
+ "bits": 4
1389
+ },
1390
+ "vision_tower.merger.proj": {
1391
+ "group_size": 64,
1392
+ "bits": 4
1393
+ },
1394
+ "vision_tower.merger.gate_proj": {
1395
+ "group_size": 64,
1396
+ "bits": 4
1397
+ },
1398
+ "vision_tower.merger.up_proj": {
1399
+ "group_size": 64,
1400
+ "bits": 4
1401
+ },
1402
+ "vision_tower.merger.down_proj": {
1403
+ "group_size": 64,
1404
+ "bits": 4
1405
+ },
1406
+ "language_model.model.embed_tokens": {
1407
+ "group_size": 64,
1408
+ "bits": 4
1409
+ },
1410
+ "language_model.model.layers.0.self_attn.q_proj": {
1411
+ "group_size": 64,
1412
+ "bits": 4
1413
+ },
1414
+ "language_model.model.layers.0.self_attn.k_proj": {
1415
+ "group_size": 64,
1416
+ "bits": 4
1417
+ },
1418
+ "language_model.model.layers.0.self_attn.v_proj": {
1419
+ "group_size": 64,
1420
+ "bits": 4
1421
+ },
1422
+ "language_model.model.layers.0.self_attn.o_proj": {
1423
+ "group_size": 64,
1424
+ "bits": 4
1425
+ },
1426
+ "language_model.model.layers.0.mlp.gate_up_proj": {
1427
+ "group_size": 64,
1428
+ "bits": 4
1429
+ },
1430
+ "language_model.model.layers.0.mlp.down_proj": {
1431
+ "group_size": 64,
1432
+ "bits": 4
1433
+ },
1434
+ "language_model.model.layers.1.self_attn.q_proj": {
1435
+ "group_size": 64,
1436
+ "bits": 4
1437
+ },
1438
+ "language_model.model.layers.1.self_attn.k_proj": {
1439
+ "group_size": 64,
1440
+ "bits": 4
1441
+ },
1442
+ "language_model.model.layers.1.self_attn.v_proj": {
1443
+ "group_size": 64,
1444
+ "bits": 4
1445
+ },
1446
+ "language_model.model.layers.1.self_attn.o_proj": {
1447
+ "group_size": 64,
1448
+ "bits": 4
1449
+ },
1450
+ "language_model.model.layers.1.mlp.gate_up_proj": {
1451
+ "group_size": 64,
1452
+ "bits": 4
1453
+ },
1454
+ "language_model.model.layers.1.mlp.down_proj": {
1455
+ "group_size": 64,
1456
+ "bits": 4
1457
+ },
1458
+ "language_model.model.layers.2.self_attn.q_proj": {
1459
+ "group_size": 64,
1460
+ "bits": 4
1461
+ },
1462
+ "language_model.model.layers.2.self_attn.k_proj": {
1463
+ "group_size": 64,
1464
+ "bits": 4
1465
+ },
1466
+ "language_model.model.layers.2.self_attn.v_proj": {
1467
+ "group_size": 64,
1468
+ "bits": 4
1469
+ },
1470
+ "language_model.model.layers.2.self_attn.o_proj": {
1471
+ "group_size": 64,
1472
+ "bits": 4
1473
+ },
1474
+ "language_model.model.layers.2.mlp.gate_up_proj": {
1475
+ "group_size": 64,
1476
+ "bits": 4
1477
+ },
1478
+ "language_model.model.layers.2.mlp.down_proj": {
1479
+ "group_size": 64,
1480
+ "bits": 4
1481
+ },
1482
+ "language_model.model.layers.3.self_attn.q_proj": {
1483
+ "group_size": 64,
1484
+ "bits": 4
1485
+ },
1486
+ "language_model.model.layers.3.self_attn.k_proj": {
1487
+ "group_size": 64,
1488
+ "bits": 4
1489
+ },
1490
+ "language_model.model.layers.3.self_attn.v_proj": {
1491
+ "group_size": 64,
1492
+ "bits": 4
1493
+ },
1494
+ "language_model.model.layers.3.self_attn.o_proj": {
1495
+ "group_size": 64,
1496
+ "bits": 4
1497
+ },
1498
+ "language_model.model.layers.3.mlp.gate_up_proj": {
1499
+ "group_size": 64,
1500
+ "bits": 4
1501
+ },
1502
+ "language_model.model.layers.3.mlp.down_proj": {
1503
+ "group_size": 64,
1504
+ "bits": 4
1505
+ },
1506
+ "language_model.model.layers.4.self_attn.q_proj": {
1507
+ "group_size": 64,
1508
+ "bits": 4
1509
+ },
1510
+ "language_model.model.layers.4.self_attn.k_proj": {
1511
+ "group_size": 64,
1512
+ "bits": 4
1513
+ },
1514
+ "language_model.model.layers.4.self_attn.v_proj": {
1515
+ "group_size": 64,
1516
+ "bits": 4
1517
+ },
1518
+ "language_model.model.layers.4.self_attn.o_proj": {
1519
+ "group_size": 64,
1520
+ "bits": 4
1521
+ },
1522
+ "language_model.model.layers.4.mlp.gate_up_proj": {
1523
+ "group_size": 64,
1524
+ "bits": 4
1525
+ },
1526
+ "language_model.model.layers.4.mlp.down_proj": {
1527
+ "group_size": 64,
1528
+ "bits": 4
1529
+ },
1530
+ "language_model.model.layers.5.self_attn.q_proj": {
1531
+ "group_size": 64,
1532
+ "bits": 4
1533
+ },
1534
+ "language_model.model.layers.5.self_attn.k_proj": {
1535
+ "group_size": 64,
1536
+ "bits": 4
1537
+ },
1538
+ "language_model.model.layers.5.self_attn.v_proj": {
1539
+ "group_size": 64,
1540
+ "bits": 4
1541
+ },
1542
+ "language_model.model.layers.5.self_attn.o_proj": {
1543
+ "group_size": 64,
1544
+ "bits": 4
1545
+ },
1546
+ "language_model.model.layers.5.mlp.gate_up_proj": {
1547
+ "group_size": 64,
1548
+ "bits": 4
1549
+ },
1550
+ "language_model.model.layers.5.mlp.down_proj": {
1551
+ "group_size": 64,
1552
+ "bits": 4
1553
+ },
1554
+ "language_model.model.layers.6.self_attn.q_proj": {
1555
+ "group_size": 64,
1556
+ "bits": 4
1557
+ },
1558
+ "language_model.model.layers.6.self_attn.k_proj": {
1559
+ "group_size": 64,
1560
+ "bits": 4
1561
+ },
1562
+ "language_model.model.layers.6.self_attn.v_proj": {
1563
+ "group_size": 64,
1564
+ "bits": 4
1565
+ },
1566
+ "language_model.model.layers.6.self_attn.o_proj": {
1567
+ "group_size": 64,
1568
+ "bits": 4
1569
+ },
1570
+ "language_model.model.layers.6.mlp.gate_up_proj": {
1571
+ "group_size": 64,
1572
+ "bits": 4
1573
+ },
1574
+ "language_model.model.layers.6.mlp.down_proj": {
1575
+ "group_size": 64,
1576
+ "bits": 4
1577
+ },
1578
+ "language_model.model.layers.7.self_attn.q_proj": {
1579
+ "group_size": 64,
1580
+ "bits": 4
1581
+ },
1582
+ "language_model.model.layers.7.self_attn.k_proj": {
1583
+ "group_size": 64,
1584
+ "bits": 4
1585
+ },
1586
+ "language_model.model.layers.7.self_attn.v_proj": {
1587
+ "group_size": 64,
1588
+ "bits": 4
1589
+ },
1590
+ "language_model.model.layers.7.self_attn.o_proj": {
1591
+ "group_size": 64,
1592
+ "bits": 4
1593
+ },
1594
+ "language_model.model.layers.7.mlp.gate_up_proj": {
1595
+ "group_size": 64,
1596
+ "bits": 4
1597
+ },
1598
+ "language_model.model.layers.7.mlp.down_proj": {
1599
+ "group_size": 64,
1600
+ "bits": 4
1601
+ },
1602
+ "language_model.model.layers.8.self_attn.q_proj": {
1603
+ "group_size": 64,
1604
+ "bits": 4
1605
+ },
1606
+ "language_model.model.layers.8.self_attn.k_proj": {
1607
+ "group_size": 64,
1608
+ "bits": 4
1609
+ },
1610
+ "language_model.model.layers.8.self_attn.v_proj": {
1611
+ "group_size": 64,
1612
+ "bits": 4
1613
+ },
1614
+ "language_model.model.layers.8.self_attn.o_proj": {
1615
+ "group_size": 64,
1616
+ "bits": 4
1617
+ },
1618
+ "language_model.model.layers.8.mlp.gate_up_proj": {
1619
+ "group_size": 64,
1620
+ "bits": 4
1621
+ },
1622
+ "language_model.model.layers.8.mlp.down_proj": {
1623
+ "group_size": 64,
1624
+ "bits": 4
1625
+ },
1626
+ "language_model.model.layers.9.self_attn.q_proj": {
1627
+ "group_size": 64,
1628
+ "bits": 4
1629
+ },
1630
+ "language_model.model.layers.9.self_attn.k_proj": {
1631
+ "group_size": 64,
1632
+ "bits": 4
1633
+ },
1634
+ "language_model.model.layers.9.self_attn.v_proj": {
1635
+ "group_size": 64,
1636
+ "bits": 4
1637
+ },
1638
+ "language_model.model.layers.9.self_attn.o_proj": {
1639
+ "group_size": 64,
1640
+ "bits": 4
1641
+ },
1642
+ "language_model.model.layers.9.mlp.gate_up_proj": {
1643
+ "group_size": 64,
1644
+ "bits": 4
1645
+ },
1646
+ "language_model.model.layers.9.mlp.down_proj": {
1647
+ "group_size": 64,
1648
+ "bits": 4
1649
+ },
1650
+ "language_model.model.layers.10.self_attn.q_proj": {
1651
+ "group_size": 64,
1652
+ "bits": 4
1653
+ },
1654
+ "language_model.model.layers.10.self_attn.k_proj": {
1655
+ "group_size": 64,
1656
+ "bits": 4
1657
+ },
1658
+ "language_model.model.layers.10.self_attn.v_proj": {
1659
+ "group_size": 64,
1660
+ "bits": 4
1661
+ },
1662
+ "language_model.model.layers.10.self_attn.o_proj": {
1663
+ "group_size": 64,
1664
+ "bits": 4
1665
+ },
1666
+ "language_model.model.layers.10.mlp.gate_up_proj": {
1667
+ "group_size": 64,
1668
+ "bits": 4
1669
+ },
1670
+ "language_model.model.layers.10.mlp.down_proj": {
1671
+ "group_size": 64,
1672
+ "bits": 4
1673
+ },
1674
+ "language_model.model.layers.11.self_attn.q_proj": {
1675
+ "group_size": 64,
1676
+ "bits": 4
1677
+ },
1678
+ "language_model.model.layers.11.self_attn.k_proj": {
1679
+ "group_size": 64,
1680
+ "bits": 4
1681
+ },
1682
+ "language_model.model.layers.11.self_attn.v_proj": {
1683
+ "group_size": 64,
1684
+ "bits": 4
1685
+ },
1686
+ "language_model.model.layers.11.self_attn.o_proj": {
1687
+ "group_size": 64,
1688
+ "bits": 4
1689
+ },
1690
+ "language_model.model.layers.11.mlp.gate_up_proj": {
1691
+ "group_size": 64,
1692
+ "bits": 4
1693
+ },
1694
+ "language_model.model.layers.11.mlp.down_proj": {
1695
+ "group_size": 64,
1696
+ "bits": 4
1697
+ },
1698
+ "language_model.model.layers.12.self_attn.q_proj": {
1699
+ "group_size": 64,
1700
+ "bits": 4
1701
+ },
1702
+ "language_model.model.layers.12.self_attn.k_proj": {
1703
+ "group_size": 64,
1704
+ "bits": 4
1705
+ },
1706
+ "language_model.model.layers.12.self_attn.v_proj": {
1707
+ "group_size": 64,
1708
+ "bits": 4
1709
+ },
1710
+ "language_model.model.layers.12.self_attn.o_proj": {
1711
+ "group_size": 64,
1712
+ "bits": 4
1713
+ },
1714
+ "language_model.model.layers.12.mlp.gate_up_proj": {
1715
+ "group_size": 64,
1716
+ "bits": 4
1717
+ },
1718
+ "language_model.model.layers.12.mlp.down_proj": {
1719
+ "group_size": 64,
1720
+ "bits": 4
1721
+ },
1722
+ "language_model.model.layers.13.self_attn.q_proj": {
1723
+ "group_size": 64,
1724
+ "bits": 4
1725
+ },
1726
+ "language_model.model.layers.13.self_attn.k_proj": {
1727
+ "group_size": 64,
1728
+ "bits": 4
1729
+ },
1730
+ "language_model.model.layers.13.self_attn.v_proj": {
1731
+ "group_size": 64,
1732
+ "bits": 4
1733
+ },
1734
+ "language_model.model.layers.13.self_attn.o_proj": {
1735
+ "group_size": 64,
1736
+ "bits": 4
1737
+ },
1738
+ "language_model.model.layers.13.mlp.gate_up_proj": {
1739
+ "group_size": 64,
1740
+ "bits": 4
1741
+ },
1742
+ "language_model.model.layers.13.mlp.down_proj": {
1743
+ "group_size": 64,
1744
+ "bits": 4
1745
+ },
1746
+ "language_model.model.layers.14.self_attn.q_proj": {
1747
+ "group_size": 64,
1748
+ "bits": 4
1749
+ },
1750
+ "language_model.model.layers.14.self_attn.k_proj": {
1751
+ "group_size": 64,
1752
+ "bits": 4
1753
+ },
1754
+ "language_model.model.layers.14.self_attn.v_proj": {
1755
+ "group_size": 64,
1756
+ "bits": 4
1757
+ },
1758
+ "language_model.model.layers.14.self_attn.o_proj": {
1759
+ "group_size": 64,
1760
+ "bits": 4
1761
+ },
1762
+ "language_model.model.layers.14.mlp.gate_up_proj": {
1763
+ "group_size": 64,
1764
+ "bits": 4
1765
+ },
1766
+ "language_model.model.layers.14.mlp.down_proj": {
1767
+ "group_size": 64,
1768
+ "bits": 4
1769
+ },
1770
+ "language_model.model.layers.15.self_attn.q_proj": {
1771
+ "group_size": 64,
1772
+ "bits": 4
1773
+ },
1774
+ "language_model.model.layers.15.self_attn.k_proj": {
1775
+ "group_size": 64,
1776
+ "bits": 4
1777
+ },
1778
+ "language_model.model.layers.15.self_attn.v_proj": {
1779
+ "group_size": 64,
1780
+ "bits": 4
1781
+ },
1782
+ "language_model.model.layers.15.self_attn.o_proj": {
1783
+ "group_size": 64,
1784
+ "bits": 4
1785
+ },
1786
+ "language_model.model.layers.15.mlp.gate_up_proj": {
1787
+ "group_size": 64,
1788
+ "bits": 4
1789
+ },
1790
+ "language_model.model.layers.15.mlp.down_proj": {
1791
+ "group_size": 64,
1792
+ "bits": 4
1793
+ },
1794
+ "language_model.lm_head": {
1795
+ "group_size": 64,
1796
+ "bits": 4
1797
+ }
1798
+ },
1799
+ "text_config": {
1800
+ "model_type": "glm_ocr_text",
1801
+ "pad_token_id": 59246,
1802
+ "vocab_size": 59392,
1803
+ "eos_token_id": [
1804
+ 59246,
1805
+ 59253
1806
+ ],
1807
+ "attention_bias": false,
1808
+ "attention_dropout": 0.0,
1809
+ "head_dim": 128,
1810
+ "hidden_act": "silu",
1811
+ "hidden_size": 1536,
1812
+ "initializer_range": 0.02,
1813
+ "intermediate_size": 4608,
1814
+ "max_position_embeddings": 131072,
1815
+ "num_attention_heads": 16,
1816
+ "num_hidden_layers": 16,
1817
+ "num_nextn_predict_layers": 1,
1818
+ "num_key_value_heads": 8,
1819
+ "rms_norm_eps": 1e-05,
1820
+ "dtype": "bfloat16",
1821
+ "rope_parameters": {
1822
+ "rope_type": "default",
1823
+ "mrope_section": [
1824
+ 16,
1825
+ 24,
1826
+ 24
1827
+ ],
1828
+ "partial_rotary_factor": 1.0,
1829
+ "rope_theta": 10000
1830
+ },
1831
+ "tie_word_embeddings": false,
1832
+ "use_cache": true
1833
+ },
1834
+ "transformers_version": "5.0.1dev0",
1835
+ "video_end_token_id": 59259,
1836
+ "video_start_token_id": 59258,
1837
+ "video_token_id": 59281,
1838
+ "vision_config": {
1839
+ "model_type": "glm_ocr_vision",
1840
+ "hidden_size": 1024,
1841
+ "depth": 24,
1842
+ "num_heads": 16,
1843
+ "attention_bias": true,
1844
+ "intermediate_size": 4096,
1845
+ "hidden_act": "silu",
1846
+ "hidden_dropout_prob": 0.0,
1847
+ "initializer_range": 0.02,
1848
+ "image_size": 336,
1849
+ "patch_size": 14,
1850
+ "out_hidden_size": 1536,
1851
+ "rms_norm_eps": 1e-05,
1852
+ "spatial_merge_size": 2,
1853
+ "temporal_patch_size": 2
1854
+ }
1855
+ }
generation_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "do_sample": false,
4
+ "eos_token_id": [
5
+ 59246,
6
+ 59253
7
+ ],
8
+ "pad_token_id": 59246,
9
+ "transformers_version": "5.0.1dev0"
10
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fb6401925acb40ad8ad91e9b73805d65dbe10c809e681d48f454f0c870310c8
3
+ size 634491294
model.safetensors.index.json ADDED
@@ -0,0 +1,961 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 634378240
4
+ },
5
+ "weight_map": {
6
+ "language_model.lm_head.biases": "model.safetensors",
7
+ "language_model.lm_head.scales": "model.safetensors",
8
+ "language_model.lm_head.weight": "model.safetensors",
9
+ "language_model.model.embed_tokens.biases": "model.safetensors",
10
+ "language_model.model.embed_tokens.scales": "model.safetensors",
11
+ "language_model.model.embed_tokens.weight": "model.safetensors",
12
+ "language_model.model.layers.0.input_layernorm.weight": "model.safetensors",
13
+ "language_model.model.layers.0.mlp.down_proj.biases": "model.safetensors",
14
+ "language_model.model.layers.0.mlp.down_proj.scales": "model.safetensors",
15
+ "language_model.model.layers.0.mlp.down_proj.weight": "model.safetensors",
16
+ "language_model.model.layers.0.mlp.gate_up_proj.biases": "model.safetensors",
17
+ "language_model.model.layers.0.mlp.gate_up_proj.scales": "model.safetensors",
18
+ "language_model.model.layers.0.mlp.gate_up_proj.weight": "model.safetensors",
19
+ "language_model.model.layers.0.post_attention_layernorm.weight": "model.safetensors",
20
+ "language_model.model.layers.0.post_mlp_layernorm.weight": "model.safetensors",
21
+ "language_model.model.layers.0.post_self_attn_layernorm.weight": "model.safetensors",
22
+ "language_model.model.layers.0.self_attn.k_proj.biases": "model.safetensors",
23
+ "language_model.model.layers.0.self_attn.k_proj.scales": "model.safetensors",
24
+ "language_model.model.layers.0.self_attn.k_proj.weight": "model.safetensors",
25
+ "language_model.model.layers.0.self_attn.o_proj.biases": "model.safetensors",
26
+ "language_model.model.layers.0.self_attn.o_proj.scales": "model.safetensors",
27
+ "language_model.model.layers.0.self_attn.o_proj.weight": "model.safetensors",
28
+ "language_model.model.layers.0.self_attn.q_proj.biases": "model.safetensors",
29
+ "language_model.model.layers.0.self_attn.q_proj.scales": "model.safetensors",
30
+ "language_model.model.layers.0.self_attn.q_proj.weight": "model.safetensors",
31
+ "language_model.model.layers.0.self_attn.v_proj.biases": "model.safetensors",
32
+ "language_model.model.layers.0.self_attn.v_proj.scales": "model.safetensors",
33
+ "language_model.model.layers.0.self_attn.v_proj.weight": "model.safetensors",
34
+ "language_model.model.layers.1.input_layernorm.weight": "model.safetensors",
35
+ "language_model.model.layers.1.mlp.down_proj.biases": "model.safetensors",
36
+ "language_model.model.layers.1.mlp.down_proj.scales": "model.safetensors",
37
+ "language_model.model.layers.1.mlp.down_proj.weight": "model.safetensors",
38
+ "language_model.model.layers.1.mlp.gate_up_proj.biases": "model.safetensors",
39
+ "language_model.model.layers.1.mlp.gate_up_proj.scales": "model.safetensors",
40
+ "language_model.model.layers.1.mlp.gate_up_proj.weight": "model.safetensors",
41
+ "language_model.model.layers.1.post_attention_layernorm.weight": "model.safetensors",
42
+ "language_model.model.layers.1.post_mlp_layernorm.weight": "model.safetensors",
43
+ "language_model.model.layers.1.post_self_attn_layernorm.weight": "model.safetensors",
44
+ "language_model.model.layers.1.self_attn.k_proj.biases": "model.safetensors",
45
+ "language_model.model.layers.1.self_attn.k_proj.scales": "model.safetensors",
46
+ "language_model.model.layers.1.self_attn.k_proj.weight": "model.safetensors",
47
+ "language_model.model.layers.1.self_attn.o_proj.biases": "model.safetensors",
48
+ "language_model.model.layers.1.self_attn.o_proj.scales": "model.safetensors",
49
+ "language_model.model.layers.1.self_attn.o_proj.weight": "model.safetensors",
50
+ "language_model.model.layers.1.self_attn.q_proj.biases": "model.safetensors",
51
+ "language_model.model.layers.1.self_attn.q_proj.scales": "model.safetensors",
52
+ "language_model.model.layers.1.self_attn.q_proj.weight": "model.safetensors",
53
+ "language_model.model.layers.1.self_attn.v_proj.biases": "model.safetensors",
54
+ "language_model.model.layers.1.self_attn.v_proj.scales": "model.safetensors",
55
+ "language_model.model.layers.1.self_attn.v_proj.weight": "model.safetensors",
56
+ "language_model.model.layers.10.input_layernorm.weight": "model.safetensors",
57
+ "language_model.model.layers.10.mlp.down_proj.biases": "model.safetensors",
58
+ "language_model.model.layers.10.mlp.down_proj.scales": "model.safetensors",
59
+ "language_model.model.layers.10.mlp.down_proj.weight": "model.safetensors",
60
+ "language_model.model.layers.10.mlp.gate_up_proj.biases": "model.safetensors",
61
+ "language_model.model.layers.10.mlp.gate_up_proj.scales": "model.safetensors",
62
+ "language_model.model.layers.10.mlp.gate_up_proj.weight": "model.safetensors",
63
+ "language_model.model.layers.10.post_attention_layernorm.weight": "model.safetensors",
64
+ "language_model.model.layers.10.post_mlp_layernorm.weight": "model.safetensors",
65
+ "language_model.model.layers.10.post_self_attn_layernorm.weight": "model.safetensors",
66
+ "language_model.model.layers.10.self_attn.k_proj.biases": "model.safetensors",
67
+ "language_model.model.layers.10.self_attn.k_proj.scales": "model.safetensors",
68
+ "language_model.model.layers.10.self_attn.k_proj.weight": "model.safetensors",
69
+ "language_model.model.layers.10.self_attn.o_proj.biases": "model.safetensors",
70
+ "language_model.model.layers.10.self_attn.o_proj.scales": "model.safetensors",
71
+ "language_model.model.layers.10.self_attn.o_proj.weight": "model.safetensors",
72
+ "language_model.model.layers.10.self_attn.q_proj.biases": "model.safetensors",
73
+ "language_model.model.layers.10.self_attn.q_proj.scales": "model.safetensors",
74
+ "language_model.model.layers.10.self_attn.q_proj.weight": "model.safetensors",
75
+ "language_model.model.layers.10.self_attn.v_proj.biases": "model.safetensors",
76
+ "language_model.model.layers.10.self_attn.v_proj.scales": "model.safetensors",
77
+ "language_model.model.layers.10.self_attn.v_proj.weight": "model.safetensors",
78
+ "language_model.model.layers.11.input_layernorm.weight": "model.safetensors",
79
+ "language_model.model.layers.11.mlp.down_proj.biases": "model.safetensors",
80
+ "language_model.model.layers.11.mlp.down_proj.scales": "model.safetensors",
81
+ "language_model.model.layers.11.mlp.down_proj.weight": "model.safetensors",
82
+ "language_model.model.layers.11.mlp.gate_up_proj.biases": "model.safetensors",
83
+ "language_model.model.layers.11.mlp.gate_up_proj.scales": "model.safetensors",
84
+ "language_model.model.layers.11.mlp.gate_up_proj.weight": "model.safetensors",
85
+ "language_model.model.layers.11.post_attention_layernorm.weight": "model.safetensors",
86
+ "language_model.model.layers.11.post_mlp_layernorm.weight": "model.safetensors",
87
+ "language_model.model.layers.11.post_self_attn_layernorm.weight": "model.safetensors",
88
+ "language_model.model.layers.11.self_attn.k_proj.biases": "model.safetensors",
89
+ "language_model.model.layers.11.self_attn.k_proj.scales": "model.safetensors",
90
+ "language_model.model.layers.11.self_attn.k_proj.weight": "model.safetensors",
91
+ "language_model.model.layers.11.self_attn.o_proj.biases": "model.safetensors",
92
+ "language_model.model.layers.11.self_attn.o_proj.scales": "model.safetensors",
93
+ "language_model.model.layers.11.self_attn.o_proj.weight": "model.safetensors",
94
+ "language_model.model.layers.11.self_attn.q_proj.biases": "model.safetensors",
95
+ "language_model.model.layers.11.self_attn.q_proj.scales": "model.safetensors",
96
+ "language_model.model.layers.11.self_attn.q_proj.weight": "model.safetensors",
97
+ "language_model.model.layers.11.self_attn.v_proj.biases": "model.safetensors",
98
+ "language_model.model.layers.11.self_attn.v_proj.scales": "model.safetensors",
99
+ "language_model.model.layers.11.self_attn.v_proj.weight": "model.safetensors",
100
+ "language_model.model.layers.12.input_layernorm.weight": "model.safetensors",
101
+ "language_model.model.layers.12.mlp.down_proj.biases": "model.safetensors",
102
+ "language_model.model.layers.12.mlp.down_proj.scales": "model.safetensors",
103
+ "language_model.model.layers.12.mlp.down_proj.weight": "model.safetensors",
104
+ "language_model.model.layers.12.mlp.gate_up_proj.biases": "model.safetensors",
105
+ "language_model.model.layers.12.mlp.gate_up_proj.scales": "model.safetensors",
106
+ "language_model.model.layers.12.mlp.gate_up_proj.weight": "model.safetensors",
107
+ "language_model.model.layers.12.post_attention_layernorm.weight": "model.safetensors",
108
+ "language_model.model.layers.12.post_mlp_layernorm.weight": "model.safetensors",
109
+ "language_model.model.layers.12.post_self_attn_layernorm.weight": "model.safetensors",
110
+ "language_model.model.layers.12.self_attn.k_proj.biases": "model.safetensors",
111
+ "language_model.model.layers.12.self_attn.k_proj.scales": "model.safetensors",
112
+ "language_model.model.layers.12.self_attn.k_proj.weight": "model.safetensors",
113
+ "language_model.model.layers.12.self_attn.o_proj.biases": "model.safetensors",
114
+ "language_model.model.layers.12.self_attn.o_proj.scales": "model.safetensors",
115
+ "language_model.model.layers.12.self_attn.o_proj.weight": "model.safetensors",
116
+ "language_model.model.layers.12.self_attn.q_proj.biases": "model.safetensors",
117
+ "language_model.model.layers.12.self_attn.q_proj.scales": "model.safetensors",
118
+ "language_model.model.layers.12.self_attn.q_proj.weight": "model.safetensors",
119
+ "language_model.model.layers.12.self_attn.v_proj.biases": "model.safetensors",
120
+ "language_model.model.layers.12.self_attn.v_proj.scales": "model.safetensors",
121
+ "language_model.model.layers.12.self_attn.v_proj.weight": "model.safetensors",
122
+ "language_model.model.layers.13.input_layernorm.weight": "model.safetensors",
123
+ "language_model.model.layers.13.mlp.down_proj.biases": "model.safetensors",
124
+ "language_model.model.layers.13.mlp.down_proj.scales": "model.safetensors",
125
+ "language_model.model.layers.13.mlp.down_proj.weight": "model.safetensors",
126
+ "language_model.model.layers.13.mlp.gate_up_proj.biases": "model.safetensors",
127
+ "language_model.model.layers.13.mlp.gate_up_proj.scales": "model.safetensors",
128
+ "language_model.model.layers.13.mlp.gate_up_proj.weight": "model.safetensors",
129
+ "language_model.model.layers.13.post_attention_layernorm.weight": "model.safetensors",
130
+ "language_model.model.layers.13.post_mlp_layernorm.weight": "model.safetensors",
131
+ "language_model.model.layers.13.post_self_attn_layernorm.weight": "model.safetensors",
132
+ "language_model.model.layers.13.self_attn.k_proj.biases": "model.safetensors",
133
+ "language_model.model.layers.13.self_attn.k_proj.scales": "model.safetensors",
134
+ "language_model.model.layers.13.self_attn.k_proj.weight": "model.safetensors",
135
+ "language_model.model.layers.13.self_attn.o_proj.biases": "model.safetensors",
136
+ "language_model.model.layers.13.self_attn.o_proj.scales": "model.safetensors",
137
+ "language_model.model.layers.13.self_attn.o_proj.weight": "model.safetensors",
138
+ "language_model.model.layers.13.self_attn.q_proj.biases": "model.safetensors",
139
+ "language_model.model.layers.13.self_attn.q_proj.scales": "model.safetensors",
140
+ "language_model.model.layers.13.self_attn.q_proj.weight": "model.safetensors",
141
+ "language_model.model.layers.13.self_attn.v_proj.biases": "model.safetensors",
142
+ "language_model.model.layers.13.self_attn.v_proj.scales": "model.safetensors",
143
+ "language_model.model.layers.13.self_attn.v_proj.weight": "model.safetensors",
144
+ "language_model.model.layers.14.input_layernorm.weight": "model.safetensors",
145
+ "language_model.model.layers.14.mlp.down_proj.biases": "model.safetensors",
146
+ "language_model.model.layers.14.mlp.down_proj.scales": "model.safetensors",
147
+ "language_model.model.layers.14.mlp.down_proj.weight": "model.safetensors",
148
+ "language_model.model.layers.14.mlp.gate_up_proj.biases": "model.safetensors",
149
+ "language_model.model.layers.14.mlp.gate_up_proj.scales": "model.safetensors",
150
+ "language_model.model.layers.14.mlp.gate_up_proj.weight": "model.safetensors",
151
+ "language_model.model.layers.14.post_attention_layernorm.weight": "model.safetensors",
152
+ "language_model.model.layers.14.post_mlp_layernorm.weight": "model.safetensors",
153
+ "language_model.model.layers.14.post_self_attn_layernorm.weight": "model.safetensors",
154
+ "language_model.model.layers.14.self_attn.k_proj.biases": "model.safetensors",
155
+ "language_model.model.layers.14.self_attn.k_proj.scales": "model.safetensors",
156
+ "language_model.model.layers.14.self_attn.k_proj.weight": "model.safetensors",
157
+ "language_model.model.layers.14.self_attn.o_proj.biases": "model.safetensors",
158
+ "language_model.model.layers.14.self_attn.o_proj.scales": "model.safetensors",
159
+ "language_model.model.layers.14.self_attn.o_proj.weight": "model.safetensors",
160
+ "language_model.model.layers.14.self_attn.q_proj.biases": "model.safetensors",
161
+ "language_model.model.layers.14.self_attn.q_proj.scales": "model.safetensors",
162
+ "language_model.model.layers.14.self_attn.q_proj.weight": "model.safetensors",
163
+ "language_model.model.layers.14.self_attn.v_proj.biases": "model.safetensors",
164
+ "language_model.model.layers.14.self_attn.v_proj.scales": "model.safetensors",
165
+ "language_model.model.layers.14.self_attn.v_proj.weight": "model.safetensors",
166
+ "language_model.model.layers.15.input_layernorm.weight": "model.safetensors",
167
+ "language_model.model.layers.15.mlp.down_proj.biases": "model.safetensors",
168
+ "language_model.model.layers.15.mlp.down_proj.scales": "model.safetensors",
169
+ "language_model.model.layers.15.mlp.down_proj.weight": "model.safetensors",
170
+ "language_model.model.layers.15.mlp.gate_up_proj.biases": "model.safetensors",
171
+ "language_model.model.layers.15.mlp.gate_up_proj.scales": "model.safetensors",
172
+ "language_model.model.layers.15.mlp.gate_up_proj.weight": "model.safetensors",
173
+ "language_model.model.layers.15.post_attention_layernorm.weight": "model.safetensors",
174
+ "language_model.model.layers.15.post_mlp_layernorm.weight": "model.safetensors",
175
+ "language_model.model.layers.15.post_self_attn_layernorm.weight": "model.safetensors",
176
+ "language_model.model.layers.15.self_attn.k_proj.biases": "model.safetensors",
177
+ "language_model.model.layers.15.self_attn.k_proj.scales": "model.safetensors",
178
+ "language_model.model.layers.15.self_attn.k_proj.weight": "model.safetensors",
179
+ "language_model.model.layers.15.self_attn.o_proj.biases": "model.safetensors",
180
+ "language_model.model.layers.15.self_attn.o_proj.scales": "model.safetensors",
181
+ "language_model.model.layers.15.self_attn.o_proj.weight": "model.safetensors",
182
+ "language_model.model.layers.15.self_attn.q_proj.biases": "model.safetensors",
183
+ "language_model.model.layers.15.self_attn.q_proj.scales": "model.safetensors",
184
+ "language_model.model.layers.15.self_attn.q_proj.weight": "model.safetensors",
185
+ "language_model.model.layers.15.self_attn.v_proj.biases": "model.safetensors",
186
+ "language_model.model.layers.15.self_attn.v_proj.scales": "model.safetensors",
187
+ "language_model.model.layers.15.self_attn.v_proj.weight": "model.safetensors",
188
+ "language_model.model.layers.2.input_layernorm.weight": "model.safetensors",
189
+ "language_model.model.layers.2.mlp.down_proj.biases": "model.safetensors",
190
+ "language_model.model.layers.2.mlp.down_proj.scales": "model.safetensors",
191
+ "language_model.model.layers.2.mlp.down_proj.weight": "model.safetensors",
192
+ "language_model.model.layers.2.mlp.gate_up_proj.biases": "model.safetensors",
193
+ "language_model.model.layers.2.mlp.gate_up_proj.scales": "model.safetensors",
194
+ "language_model.model.layers.2.mlp.gate_up_proj.weight": "model.safetensors",
195
+ "language_model.model.layers.2.post_attention_layernorm.weight": "model.safetensors",
196
+ "language_model.model.layers.2.post_mlp_layernorm.weight": "model.safetensors",
197
+ "language_model.model.layers.2.post_self_attn_layernorm.weight": "model.safetensors",
198
+ "language_model.model.layers.2.self_attn.k_proj.biases": "model.safetensors",
199
+ "language_model.model.layers.2.self_attn.k_proj.scales": "model.safetensors",
200
+ "language_model.model.layers.2.self_attn.k_proj.weight": "model.safetensors",
201
+ "language_model.model.layers.2.self_attn.o_proj.biases": "model.safetensors",
202
+ "language_model.model.layers.2.self_attn.o_proj.scales": "model.safetensors",
203
+ "language_model.model.layers.2.self_attn.o_proj.weight": "model.safetensors",
204
+ "language_model.model.layers.2.self_attn.q_proj.biases": "model.safetensors",
205
+ "language_model.model.layers.2.self_attn.q_proj.scales": "model.safetensors",
206
+ "language_model.model.layers.2.self_attn.q_proj.weight": "model.safetensors",
207
+ "language_model.model.layers.2.self_attn.v_proj.biases": "model.safetensors",
208
+ "language_model.model.layers.2.self_attn.v_proj.scales": "model.safetensors",
209
+ "language_model.model.layers.2.self_attn.v_proj.weight": "model.safetensors",
210
+ "language_model.model.layers.3.input_layernorm.weight": "model.safetensors",
211
+ "language_model.model.layers.3.mlp.down_proj.biases": "model.safetensors",
212
+ "language_model.model.layers.3.mlp.down_proj.scales": "model.safetensors",
213
+ "language_model.model.layers.3.mlp.down_proj.weight": "model.safetensors",
214
+ "language_model.model.layers.3.mlp.gate_up_proj.biases": "model.safetensors",
215
+ "language_model.model.layers.3.mlp.gate_up_proj.scales": "model.safetensors",
216
+ "language_model.model.layers.3.mlp.gate_up_proj.weight": "model.safetensors",
217
+ "language_model.model.layers.3.post_attention_layernorm.weight": "model.safetensors",
218
+ "language_model.model.layers.3.post_mlp_layernorm.weight": "model.safetensors",
219
+ "language_model.model.layers.3.post_self_attn_layernorm.weight": "model.safetensors",
220
+ "language_model.model.layers.3.self_attn.k_proj.biases": "model.safetensors",
221
+ "language_model.model.layers.3.self_attn.k_proj.scales": "model.safetensors",
222
+ "language_model.model.layers.3.self_attn.k_proj.weight": "model.safetensors",
223
+ "language_model.model.layers.3.self_attn.o_proj.biases": "model.safetensors",
224
+ "language_model.model.layers.3.self_attn.o_proj.scales": "model.safetensors",
225
+ "language_model.model.layers.3.self_attn.o_proj.weight": "model.safetensors",
226
+ "language_model.model.layers.3.self_attn.q_proj.biases": "model.safetensors",
227
+ "language_model.model.layers.3.self_attn.q_proj.scales": "model.safetensors",
228
+ "language_model.model.layers.3.self_attn.q_proj.weight": "model.safetensors",
229
+ "language_model.model.layers.3.self_attn.v_proj.biases": "model.safetensors",
230
+ "language_model.model.layers.3.self_attn.v_proj.scales": "model.safetensors",
231
+ "language_model.model.layers.3.self_attn.v_proj.weight": "model.safetensors",
232
+ "language_model.model.layers.4.input_layernorm.weight": "model.safetensors",
233
+ "language_model.model.layers.4.mlp.down_proj.biases": "model.safetensors",
234
+ "language_model.model.layers.4.mlp.down_proj.scales": "model.safetensors",
235
+ "language_model.model.layers.4.mlp.down_proj.weight": "model.safetensors",
236
+ "language_model.model.layers.4.mlp.gate_up_proj.biases": "model.safetensors",
237
+ "language_model.model.layers.4.mlp.gate_up_proj.scales": "model.safetensors",
238
+ "language_model.model.layers.4.mlp.gate_up_proj.weight": "model.safetensors",
239
+ "language_model.model.layers.4.post_attention_layernorm.weight": "model.safetensors",
240
+ "language_model.model.layers.4.post_mlp_layernorm.weight": "model.safetensors",
241
+ "language_model.model.layers.4.post_self_attn_layernorm.weight": "model.safetensors",
242
+ "language_model.model.layers.4.self_attn.k_proj.biases": "model.safetensors",
243
+ "language_model.model.layers.4.self_attn.k_proj.scales": "model.safetensors",
244
+ "language_model.model.layers.4.self_attn.k_proj.weight": "model.safetensors",
245
+ "language_model.model.layers.4.self_attn.o_proj.biases": "model.safetensors",
246
+ "language_model.model.layers.4.self_attn.o_proj.scales": "model.safetensors",
247
+ "language_model.model.layers.4.self_attn.o_proj.weight": "model.safetensors",
248
+ "language_model.model.layers.4.self_attn.q_proj.biases": "model.safetensors",
249
+ "language_model.model.layers.4.self_attn.q_proj.scales": "model.safetensors",
250
+ "language_model.model.layers.4.self_attn.q_proj.weight": "model.safetensors",
251
+ "language_model.model.layers.4.self_attn.v_proj.biases": "model.safetensors",
252
+ "language_model.model.layers.4.self_attn.v_proj.scales": "model.safetensors",
253
+ "language_model.model.layers.4.self_attn.v_proj.weight": "model.safetensors",
254
+ "language_model.model.layers.5.input_layernorm.weight": "model.safetensors",
255
+ "language_model.model.layers.5.mlp.down_proj.biases": "model.safetensors",
256
+ "language_model.model.layers.5.mlp.down_proj.scales": "model.safetensors",
257
+ "language_model.model.layers.5.mlp.down_proj.weight": "model.safetensors",
258
+ "language_model.model.layers.5.mlp.gate_up_proj.biases": "model.safetensors",
259
+ "language_model.model.layers.5.mlp.gate_up_proj.scales": "model.safetensors",
260
+ "language_model.model.layers.5.mlp.gate_up_proj.weight": "model.safetensors",
261
+ "language_model.model.layers.5.post_attention_layernorm.weight": "model.safetensors",
262
+ "language_model.model.layers.5.post_mlp_layernorm.weight": "model.safetensors",
263
+ "language_model.model.layers.5.post_self_attn_layernorm.weight": "model.safetensors",
264
+ "language_model.model.layers.5.self_attn.k_proj.biases": "model.safetensors",
265
+ "language_model.model.layers.5.self_attn.k_proj.scales": "model.safetensors",
266
+ "language_model.model.layers.5.self_attn.k_proj.weight": "model.safetensors",
267
+ "language_model.model.layers.5.self_attn.o_proj.biases": "model.safetensors",
268
+ "language_model.model.layers.5.self_attn.o_proj.scales": "model.safetensors",
269
+ "language_model.model.layers.5.self_attn.o_proj.weight": "model.safetensors",
270
+ "language_model.model.layers.5.self_attn.q_proj.biases": "model.safetensors",
271
+ "language_model.model.layers.5.self_attn.q_proj.scales": "model.safetensors",
272
+ "language_model.model.layers.5.self_attn.q_proj.weight": "model.safetensors",
273
+ "language_model.model.layers.5.self_attn.v_proj.biases": "model.safetensors",
274
+ "language_model.model.layers.5.self_attn.v_proj.scales": "model.safetensors",
275
+ "language_model.model.layers.5.self_attn.v_proj.weight": "model.safetensors",
276
+ "language_model.model.layers.6.input_layernorm.weight": "model.safetensors",
277
+ "language_model.model.layers.6.mlp.down_proj.biases": "model.safetensors",
278
+ "language_model.model.layers.6.mlp.down_proj.scales": "model.safetensors",
279
+ "language_model.model.layers.6.mlp.down_proj.weight": "model.safetensors",
280
+ "language_model.model.layers.6.mlp.gate_up_proj.biases": "model.safetensors",
281
+ "language_model.model.layers.6.mlp.gate_up_proj.scales": "model.safetensors",
282
+ "language_model.model.layers.6.mlp.gate_up_proj.weight": "model.safetensors",
283
+ "language_model.model.layers.6.post_attention_layernorm.weight": "model.safetensors",
284
+ "language_model.model.layers.6.post_mlp_layernorm.weight": "model.safetensors",
285
+ "language_model.model.layers.6.post_self_attn_layernorm.weight": "model.safetensors",
286
+ "language_model.model.layers.6.self_attn.k_proj.biases": "model.safetensors",
287
+ "language_model.model.layers.6.self_attn.k_proj.scales": "model.safetensors",
288
+ "language_model.model.layers.6.self_attn.k_proj.weight": "model.safetensors",
289
+ "language_model.model.layers.6.self_attn.o_proj.biases": "model.safetensors",
290
+ "language_model.model.layers.6.self_attn.o_proj.scales": "model.safetensors",
291
+ "language_model.model.layers.6.self_attn.o_proj.weight": "model.safetensors",
292
+ "language_model.model.layers.6.self_attn.q_proj.biases": "model.safetensors",
293
+ "language_model.model.layers.6.self_attn.q_proj.scales": "model.safetensors",
294
+ "language_model.model.layers.6.self_attn.q_proj.weight": "model.safetensors",
295
+ "language_model.model.layers.6.self_attn.v_proj.biases": "model.safetensors",
296
+ "language_model.model.layers.6.self_attn.v_proj.scales": "model.safetensors",
297
+ "language_model.model.layers.6.self_attn.v_proj.weight": "model.safetensors",
298
+ "language_model.model.layers.7.input_layernorm.weight": "model.safetensors",
299
+ "language_model.model.layers.7.mlp.down_proj.biases": "model.safetensors",
300
+ "language_model.model.layers.7.mlp.down_proj.scales": "model.safetensors",
301
+ "language_model.model.layers.7.mlp.down_proj.weight": "model.safetensors",
302
+ "language_model.model.layers.7.mlp.gate_up_proj.biases": "model.safetensors",
303
+ "language_model.model.layers.7.mlp.gate_up_proj.scales": "model.safetensors",
304
+ "language_model.model.layers.7.mlp.gate_up_proj.weight": "model.safetensors",
305
+ "language_model.model.layers.7.post_attention_layernorm.weight": "model.safetensors",
306
+ "language_model.model.layers.7.post_mlp_layernorm.weight": "model.safetensors",
307
+ "language_model.model.layers.7.post_self_attn_layernorm.weight": "model.safetensors",
308
+ "language_model.model.layers.7.self_attn.k_proj.biases": "model.safetensors",
309
+ "language_model.model.layers.7.self_attn.k_proj.scales": "model.safetensors",
310
+ "language_model.model.layers.7.self_attn.k_proj.weight": "model.safetensors",
311
+ "language_model.model.layers.7.self_attn.o_proj.biases": "model.safetensors",
312
+ "language_model.model.layers.7.self_attn.o_proj.scales": "model.safetensors",
313
+ "language_model.model.layers.7.self_attn.o_proj.weight": "model.safetensors",
314
+ "language_model.model.layers.7.self_attn.q_proj.biases": "model.safetensors",
315
+ "language_model.model.layers.7.self_attn.q_proj.scales": "model.safetensors",
316
+ "language_model.model.layers.7.self_attn.q_proj.weight": "model.safetensors",
317
+ "language_model.model.layers.7.self_attn.v_proj.biases": "model.safetensors",
318
+ "language_model.model.layers.7.self_attn.v_proj.scales": "model.safetensors",
319
+ "language_model.model.layers.7.self_attn.v_proj.weight": "model.safetensors",
320
+ "language_model.model.layers.8.input_layernorm.weight": "model.safetensors",
321
+ "language_model.model.layers.8.mlp.down_proj.biases": "model.safetensors",
322
+ "language_model.model.layers.8.mlp.down_proj.scales": "model.safetensors",
323
+ "language_model.model.layers.8.mlp.down_proj.weight": "model.safetensors",
324
+ "language_model.model.layers.8.mlp.gate_up_proj.biases": "model.safetensors",
325
+ "language_model.model.layers.8.mlp.gate_up_proj.scales": "model.safetensors",
326
+ "language_model.model.layers.8.mlp.gate_up_proj.weight": "model.safetensors",
327
+ "language_model.model.layers.8.post_attention_layernorm.weight": "model.safetensors",
328
+ "language_model.model.layers.8.post_mlp_layernorm.weight": "model.safetensors",
329
+ "language_model.model.layers.8.post_self_attn_layernorm.weight": "model.safetensors",
330
+ "language_model.model.layers.8.self_attn.k_proj.biases": "model.safetensors",
331
+ "language_model.model.layers.8.self_attn.k_proj.scales": "model.safetensors",
332
+ "language_model.model.layers.8.self_attn.k_proj.weight": "model.safetensors",
333
+ "language_model.model.layers.8.self_attn.o_proj.biases": "model.safetensors",
334
+ "language_model.model.layers.8.self_attn.o_proj.scales": "model.safetensors",
335
+ "language_model.model.layers.8.self_attn.o_proj.weight": "model.safetensors",
336
+ "language_model.model.layers.8.self_attn.q_proj.biases": "model.safetensors",
337
+ "language_model.model.layers.8.self_attn.q_proj.scales": "model.safetensors",
338
+ "language_model.model.layers.8.self_attn.q_proj.weight": "model.safetensors",
339
+ "language_model.model.layers.8.self_attn.v_proj.biases": "model.safetensors",
340
+ "language_model.model.layers.8.self_attn.v_proj.scales": "model.safetensors",
341
+ "language_model.model.layers.8.self_attn.v_proj.weight": "model.safetensors",
342
+ "language_model.model.layers.9.input_layernorm.weight": "model.safetensors",
343
+ "language_model.model.layers.9.mlp.down_proj.biases": "model.safetensors",
344
+ "language_model.model.layers.9.mlp.down_proj.scales": "model.safetensors",
345
+ "language_model.model.layers.9.mlp.down_proj.weight": "model.safetensors",
346
+ "language_model.model.layers.9.mlp.gate_up_proj.biases": "model.safetensors",
347
+ "language_model.model.layers.9.mlp.gate_up_proj.scales": "model.safetensors",
348
+ "language_model.model.layers.9.mlp.gate_up_proj.weight": "model.safetensors",
349
+ "language_model.model.layers.9.post_attention_layernorm.weight": "model.safetensors",
350
+ "language_model.model.layers.9.post_mlp_layernorm.weight": "model.safetensors",
351
+ "language_model.model.layers.9.post_self_attn_layernorm.weight": "model.safetensors",
352
+ "language_model.model.layers.9.self_attn.k_proj.biases": "model.safetensors",
353
+ "language_model.model.layers.9.self_attn.k_proj.scales": "model.safetensors",
354
+ "language_model.model.layers.9.self_attn.k_proj.weight": "model.safetensors",
355
+ "language_model.model.layers.9.self_attn.o_proj.biases": "model.safetensors",
356
+ "language_model.model.layers.9.self_attn.o_proj.scales": "model.safetensors",
357
+ "language_model.model.layers.9.self_attn.o_proj.weight": "model.safetensors",
358
+ "language_model.model.layers.9.self_attn.q_proj.biases": "model.safetensors",
359
+ "language_model.model.layers.9.self_attn.q_proj.scales": "model.safetensors",
360
+ "language_model.model.layers.9.self_attn.q_proj.weight": "model.safetensors",
361
+ "language_model.model.layers.9.self_attn.v_proj.biases": "model.safetensors",
362
+ "language_model.model.layers.9.self_attn.v_proj.scales": "model.safetensors",
363
+ "language_model.model.layers.9.self_attn.v_proj.weight": "model.safetensors",
364
+ "language_model.model.norm.weight": "model.safetensors",
365
+ "vision_tower.blocks.0.attn.k_norm.weight": "model.safetensors",
366
+ "vision_tower.blocks.0.attn.proj.bias": "model.safetensors",
367
+ "vision_tower.blocks.0.attn.proj.biases": "model.safetensors",
368
+ "vision_tower.blocks.0.attn.proj.scales": "model.safetensors",
369
+ "vision_tower.blocks.0.attn.proj.weight": "model.safetensors",
370
+ "vision_tower.blocks.0.attn.q_norm.weight": "model.safetensors",
371
+ "vision_tower.blocks.0.attn.qkv.bias": "model.safetensors",
372
+ "vision_tower.blocks.0.attn.qkv.biases": "model.safetensors",
373
+ "vision_tower.blocks.0.attn.qkv.scales": "model.safetensors",
374
+ "vision_tower.blocks.0.attn.qkv.weight": "model.safetensors",
375
+ "vision_tower.blocks.0.mlp.down_proj.bias": "model.safetensors",
376
+ "vision_tower.blocks.0.mlp.down_proj.biases": "model.safetensors",
377
+ "vision_tower.blocks.0.mlp.down_proj.scales": "model.safetensors",
378
+ "vision_tower.blocks.0.mlp.down_proj.weight": "model.safetensors",
379
+ "vision_tower.blocks.0.mlp.gate_proj.bias": "model.safetensors",
380
+ "vision_tower.blocks.0.mlp.gate_proj.biases": "model.safetensors",
381
+ "vision_tower.blocks.0.mlp.gate_proj.scales": "model.safetensors",
382
+ "vision_tower.blocks.0.mlp.gate_proj.weight": "model.safetensors",
383
+ "vision_tower.blocks.0.mlp.up_proj.bias": "model.safetensors",
384
+ "vision_tower.blocks.0.mlp.up_proj.biases": "model.safetensors",
385
+ "vision_tower.blocks.0.mlp.up_proj.scales": "model.safetensors",
386
+ "vision_tower.blocks.0.mlp.up_proj.weight": "model.safetensors",
387
+ "vision_tower.blocks.0.norm1.weight": "model.safetensors",
388
+ "vision_tower.blocks.0.norm2.weight": "model.safetensors",
389
+ "vision_tower.blocks.1.attn.k_norm.weight": "model.safetensors",
390
+ "vision_tower.blocks.1.attn.proj.bias": "model.safetensors",
391
+ "vision_tower.blocks.1.attn.proj.biases": "model.safetensors",
392
+ "vision_tower.blocks.1.attn.proj.scales": "model.safetensors",
393
+ "vision_tower.blocks.1.attn.proj.weight": "model.safetensors",
394
+ "vision_tower.blocks.1.attn.q_norm.weight": "model.safetensors",
395
+ "vision_tower.blocks.1.attn.qkv.bias": "model.safetensors",
396
+ "vision_tower.blocks.1.attn.qkv.biases": "model.safetensors",
397
+ "vision_tower.blocks.1.attn.qkv.scales": "model.safetensors",
398
+ "vision_tower.blocks.1.attn.qkv.weight": "model.safetensors",
399
+ "vision_tower.blocks.1.mlp.down_proj.bias": "model.safetensors",
400
+ "vision_tower.blocks.1.mlp.down_proj.biases": "model.safetensors",
401
+ "vision_tower.blocks.1.mlp.down_proj.scales": "model.safetensors",
402
+ "vision_tower.blocks.1.mlp.down_proj.weight": "model.safetensors",
403
+ "vision_tower.blocks.1.mlp.gate_proj.bias": "model.safetensors",
404
+ "vision_tower.blocks.1.mlp.gate_proj.biases": "model.safetensors",
405
+ "vision_tower.blocks.1.mlp.gate_proj.scales": "model.safetensors",
406
+ "vision_tower.blocks.1.mlp.gate_proj.weight": "model.safetensors",
407
+ "vision_tower.blocks.1.mlp.up_proj.bias": "model.safetensors",
408
+ "vision_tower.blocks.1.mlp.up_proj.biases": "model.safetensors",
409
+ "vision_tower.blocks.1.mlp.up_proj.scales": "model.safetensors",
410
+ "vision_tower.blocks.1.mlp.up_proj.weight": "model.safetensors",
411
+ "vision_tower.blocks.1.norm1.weight": "model.safetensors",
412
+ "vision_tower.blocks.1.norm2.weight": "model.safetensors",
413
+ "vision_tower.blocks.10.attn.k_norm.weight": "model.safetensors",
414
+ "vision_tower.blocks.10.attn.proj.bias": "model.safetensors",
415
+ "vision_tower.blocks.10.attn.proj.biases": "model.safetensors",
416
+ "vision_tower.blocks.10.attn.proj.scales": "model.safetensors",
417
+ "vision_tower.blocks.10.attn.proj.weight": "model.safetensors",
418
+ "vision_tower.blocks.10.attn.q_norm.weight": "model.safetensors",
419
+ "vision_tower.blocks.10.attn.qkv.bias": "model.safetensors",
420
+ "vision_tower.blocks.10.attn.qkv.biases": "model.safetensors",
421
+ "vision_tower.blocks.10.attn.qkv.scales": "model.safetensors",
422
+ "vision_tower.blocks.10.attn.qkv.weight": "model.safetensors",
423
+ "vision_tower.blocks.10.mlp.down_proj.bias": "model.safetensors",
424
+ "vision_tower.blocks.10.mlp.down_proj.biases": "model.safetensors",
425
+ "vision_tower.blocks.10.mlp.down_proj.scales": "model.safetensors",
426
+ "vision_tower.blocks.10.mlp.down_proj.weight": "model.safetensors",
427
+ "vision_tower.blocks.10.mlp.gate_proj.bias": "model.safetensors",
428
+ "vision_tower.blocks.10.mlp.gate_proj.biases": "model.safetensors",
429
+ "vision_tower.blocks.10.mlp.gate_proj.scales": "model.safetensors",
430
+ "vision_tower.blocks.10.mlp.gate_proj.weight": "model.safetensors",
431
+ "vision_tower.blocks.10.mlp.up_proj.bias": "model.safetensors",
432
+ "vision_tower.blocks.10.mlp.up_proj.biases": "model.safetensors",
433
+ "vision_tower.blocks.10.mlp.up_proj.scales": "model.safetensors",
434
+ "vision_tower.blocks.10.mlp.up_proj.weight": "model.safetensors",
435
+ "vision_tower.blocks.10.norm1.weight": "model.safetensors",
436
+ "vision_tower.blocks.10.norm2.weight": "model.safetensors",
437
+ "vision_tower.blocks.11.attn.k_norm.weight": "model.safetensors",
438
+ "vision_tower.blocks.11.attn.proj.bias": "model.safetensors",
439
+ "vision_tower.blocks.11.attn.proj.biases": "model.safetensors",
440
+ "vision_tower.blocks.11.attn.proj.scales": "model.safetensors",
441
+ "vision_tower.blocks.11.attn.proj.weight": "model.safetensors",
442
+ "vision_tower.blocks.11.attn.q_norm.weight": "model.safetensors",
443
+ "vision_tower.blocks.11.attn.qkv.bias": "model.safetensors",
444
+ "vision_tower.blocks.11.attn.qkv.biases": "model.safetensors",
445
+ "vision_tower.blocks.11.attn.qkv.scales": "model.safetensors",
446
+ "vision_tower.blocks.11.attn.qkv.weight": "model.safetensors",
447
+ "vision_tower.blocks.11.mlp.down_proj.bias": "model.safetensors",
448
+ "vision_tower.blocks.11.mlp.down_proj.biases": "model.safetensors",
449
+ "vision_tower.blocks.11.mlp.down_proj.scales": "model.safetensors",
450
+ "vision_tower.blocks.11.mlp.down_proj.weight": "model.safetensors",
451
+ "vision_tower.blocks.11.mlp.gate_proj.bias": "model.safetensors",
452
+ "vision_tower.blocks.11.mlp.gate_proj.biases": "model.safetensors",
453
+ "vision_tower.blocks.11.mlp.gate_proj.scales": "model.safetensors",
454
+ "vision_tower.blocks.11.mlp.gate_proj.weight": "model.safetensors",
455
+ "vision_tower.blocks.11.mlp.up_proj.bias": "model.safetensors",
456
+ "vision_tower.blocks.11.mlp.up_proj.biases": "model.safetensors",
457
+ "vision_tower.blocks.11.mlp.up_proj.scales": "model.safetensors",
458
+ "vision_tower.blocks.11.mlp.up_proj.weight": "model.safetensors",
459
+ "vision_tower.blocks.11.norm1.weight": "model.safetensors",
460
+ "vision_tower.blocks.11.norm2.weight": "model.safetensors",
461
+ "vision_tower.blocks.12.attn.k_norm.weight": "model.safetensors",
462
+ "vision_tower.blocks.12.attn.proj.bias": "model.safetensors",
463
+ "vision_tower.blocks.12.attn.proj.biases": "model.safetensors",
464
+ "vision_tower.blocks.12.attn.proj.scales": "model.safetensors",
465
+ "vision_tower.blocks.12.attn.proj.weight": "model.safetensors",
466
+ "vision_tower.blocks.12.attn.q_norm.weight": "model.safetensors",
467
+ "vision_tower.blocks.12.attn.qkv.bias": "model.safetensors",
468
+ "vision_tower.blocks.12.attn.qkv.biases": "model.safetensors",
469
+ "vision_tower.blocks.12.attn.qkv.scales": "model.safetensors",
470
+ "vision_tower.blocks.12.attn.qkv.weight": "model.safetensors",
471
+ "vision_tower.blocks.12.mlp.down_proj.bias": "model.safetensors",
472
+ "vision_tower.blocks.12.mlp.down_proj.biases": "model.safetensors",
473
+ "vision_tower.blocks.12.mlp.down_proj.scales": "model.safetensors",
474
+ "vision_tower.blocks.12.mlp.down_proj.weight": "model.safetensors",
475
+ "vision_tower.blocks.12.mlp.gate_proj.bias": "model.safetensors",
476
+ "vision_tower.blocks.12.mlp.gate_proj.biases": "model.safetensors",
477
+ "vision_tower.blocks.12.mlp.gate_proj.scales": "model.safetensors",
478
+ "vision_tower.blocks.12.mlp.gate_proj.weight": "model.safetensors",
479
+ "vision_tower.blocks.12.mlp.up_proj.bias": "model.safetensors",
480
+ "vision_tower.blocks.12.mlp.up_proj.biases": "model.safetensors",
481
+ "vision_tower.blocks.12.mlp.up_proj.scales": "model.safetensors",
482
+ "vision_tower.blocks.12.mlp.up_proj.weight": "model.safetensors",
483
+ "vision_tower.blocks.12.norm1.weight": "model.safetensors",
484
+ "vision_tower.blocks.12.norm2.weight": "model.safetensors",
485
+ "vision_tower.blocks.13.attn.k_norm.weight": "model.safetensors",
486
+ "vision_tower.blocks.13.attn.proj.bias": "model.safetensors",
487
+ "vision_tower.blocks.13.attn.proj.biases": "model.safetensors",
488
+ "vision_tower.blocks.13.attn.proj.scales": "model.safetensors",
489
+ "vision_tower.blocks.13.attn.proj.weight": "model.safetensors",
490
+ "vision_tower.blocks.13.attn.q_norm.weight": "model.safetensors",
491
+ "vision_tower.blocks.13.attn.qkv.bias": "model.safetensors",
492
+ "vision_tower.blocks.13.attn.qkv.biases": "model.safetensors",
493
+ "vision_tower.blocks.13.attn.qkv.scales": "model.safetensors",
494
+ "vision_tower.blocks.13.attn.qkv.weight": "model.safetensors",
495
+ "vision_tower.blocks.13.mlp.down_proj.bias": "model.safetensors",
496
+ "vision_tower.blocks.13.mlp.down_proj.biases": "model.safetensors",
497
+ "vision_tower.blocks.13.mlp.down_proj.scales": "model.safetensors",
498
+ "vision_tower.blocks.13.mlp.down_proj.weight": "model.safetensors",
499
+ "vision_tower.blocks.13.mlp.gate_proj.bias": "model.safetensors",
500
+ "vision_tower.blocks.13.mlp.gate_proj.biases": "model.safetensors",
501
+ "vision_tower.blocks.13.mlp.gate_proj.scales": "model.safetensors",
502
+ "vision_tower.blocks.13.mlp.gate_proj.weight": "model.safetensors",
503
+ "vision_tower.blocks.13.mlp.up_proj.bias": "model.safetensors",
504
+ "vision_tower.blocks.13.mlp.up_proj.biases": "model.safetensors",
505
+ "vision_tower.blocks.13.mlp.up_proj.scales": "model.safetensors",
506
+ "vision_tower.blocks.13.mlp.up_proj.weight": "model.safetensors",
507
+ "vision_tower.blocks.13.norm1.weight": "model.safetensors",
508
+ "vision_tower.blocks.13.norm2.weight": "model.safetensors",
509
+ "vision_tower.blocks.14.attn.k_norm.weight": "model.safetensors",
510
+ "vision_tower.blocks.14.attn.proj.bias": "model.safetensors",
511
+ "vision_tower.blocks.14.attn.proj.biases": "model.safetensors",
512
+ "vision_tower.blocks.14.attn.proj.scales": "model.safetensors",
513
+ "vision_tower.blocks.14.attn.proj.weight": "model.safetensors",
514
+ "vision_tower.blocks.14.attn.q_norm.weight": "model.safetensors",
515
+ "vision_tower.blocks.14.attn.qkv.bias": "model.safetensors",
516
+ "vision_tower.blocks.14.attn.qkv.biases": "model.safetensors",
517
+ "vision_tower.blocks.14.attn.qkv.scales": "model.safetensors",
518
+ "vision_tower.blocks.14.attn.qkv.weight": "model.safetensors",
519
+ "vision_tower.blocks.14.mlp.down_proj.bias": "model.safetensors",
520
+ "vision_tower.blocks.14.mlp.down_proj.biases": "model.safetensors",
521
+ "vision_tower.blocks.14.mlp.down_proj.scales": "model.safetensors",
522
+ "vision_tower.blocks.14.mlp.down_proj.weight": "model.safetensors",
523
+ "vision_tower.blocks.14.mlp.gate_proj.bias": "model.safetensors",
524
+ "vision_tower.blocks.14.mlp.gate_proj.biases": "model.safetensors",
525
+ "vision_tower.blocks.14.mlp.gate_proj.scales": "model.safetensors",
526
+ "vision_tower.blocks.14.mlp.gate_proj.weight": "model.safetensors",
527
+ "vision_tower.blocks.14.mlp.up_proj.bias": "model.safetensors",
528
+ "vision_tower.blocks.14.mlp.up_proj.biases": "model.safetensors",
529
+ "vision_tower.blocks.14.mlp.up_proj.scales": "model.safetensors",
530
+ "vision_tower.blocks.14.mlp.up_proj.weight": "model.safetensors",
531
+ "vision_tower.blocks.14.norm1.weight": "model.safetensors",
532
+ "vision_tower.blocks.14.norm2.weight": "model.safetensors",
533
+ "vision_tower.blocks.15.attn.k_norm.weight": "model.safetensors",
534
+ "vision_tower.blocks.15.attn.proj.bias": "model.safetensors",
535
+ "vision_tower.blocks.15.attn.proj.biases": "model.safetensors",
536
+ "vision_tower.blocks.15.attn.proj.scales": "model.safetensors",
537
+ "vision_tower.blocks.15.attn.proj.weight": "model.safetensors",
538
+ "vision_tower.blocks.15.attn.q_norm.weight": "model.safetensors",
539
+ "vision_tower.blocks.15.attn.qkv.bias": "model.safetensors",
540
+ "vision_tower.blocks.15.attn.qkv.biases": "model.safetensors",
541
+ "vision_tower.blocks.15.attn.qkv.scales": "model.safetensors",
542
+ "vision_tower.blocks.15.attn.qkv.weight": "model.safetensors",
543
+ "vision_tower.blocks.15.mlp.down_proj.bias": "model.safetensors",
544
+ "vision_tower.blocks.15.mlp.down_proj.biases": "model.safetensors",
545
+ "vision_tower.blocks.15.mlp.down_proj.scales": "model.safetensors",
546
+ "vision_tower.blocks.15.mlp.down_proj.weight": "model.safetensors",
547
+ "vision_tower.blocks.15.mlp.gate_proj.bias": "model.safetensors",
548
+ "vision_tower.blocks.15.mlp.gate_proj.biases": "model.safetensors",
549
+ "vision_tower.blocks.15.mlp.gate_proj.scales": "model.safetensors",
550
+ "vision_tower.blocks.15.mlp.gate_proj.weight": "model.safetensors",
551
+ "vision_tower.blocks.15.mlp.up_proj.bias": "model.safetensors",
552
+ "vision_tower.blocks.15.mlp.up_proj.biases": "model.safetensors",
553
+ "vision_tower.blocks.15.mlp.up_proj.scales": "model.safetensors",
554
+ "vision_tower.blocks.15.mlp.up_proj.weight": "model.safetensors",
555
+ "vision_tower.blocks.15.norm1.weight": "model.safetensors",
556
+ "vision_tower.blocks.15.norm2.weight": "model.safetensors",
557
+ "vision_tower.blocks.16.attn.k_norm.weight": "model.safetensors",
558
+ "vision_tower.blocks.16.attn.proj.bias": "model.safetensors",
559
+ "vision_tower.blocks.16.attn.proj.biases": "model.safetensors",
560
+ "vision_tower.blocks.16.attn.proj.scales": "model.safetensors",
561
+ "vision_tower.blocks.16.attn.proj.weight": "model.safetensors",
562
+ "vision_tower.blocks.16.attn.q_norm.weight": "model.safetensors",
563
+ "vision_tower.blocks.16.attn.qkv.bias": "model.safetensors",
564
+ "vision_tower.blocks.16.attn.qkv.biases": "model.safetensors",
565
+ "vision_tower.blocks.16.attn.qkv.scales": "model.safetensors",
566
+ "vision_tower.blocks.16.attn.qkv.weight": "model.safetensors",
567
+ "vision_tower.blocks.16.mlp.down_proj.bias": "model.safetensors",
568
+ "vision_tower.blocks.16.mlp.down_proj.biases": "model.safetensors",
569
+ "vision_tower.blocks.16.mlp.down_proj.scales": "model.safetensors",
570
+ "vision_tower.blocks.16.mlp.down_proj.weight": "model.safetensors",
571
+ "vision_tower.blocks.16.mlp.gate_proj.bias": "model.safetensors",
572
+ "vision_tower.blocks.16.mlp.gate_proj.biases": "model.safetensors",
573
+ "vision_tower.blocks.16.mlp.gate_proj.scales": "model.safetensors",
574
+ "vision_tower.blocks.16.mlp.gate_proj.weight": "model.safetensors",
575
+ "vision_tower.blocks.16.mlp.up_proj.bias": "model.safetensors",
576
+ "vision_tower.blocks.16.mlp.up_proj.biases": "model.safetensors",
577
+ "vision_tower.blocks.16.mlp.up_proj.scales": "model.safetensors",
578
+ "vision_tower.blocks.16.mlp.up_proj.weight": "model.safetensors",
579
+ "vision_tower.blocks.16.norm1.weight": "model.safetensors",
580
+ "vision_tower.blocks.16.norm2.weight": "model.safetensors",
581
+ "vision_tower.blocks.17.attn.k_norm.weight": "model.safetensors",
582
+ "vision_tower.blocks.17.attn.proj.bias": "model.safetensors",
583
+ "vision_tower.blocks.17.attn.proj.biases": "model.safetensors",
584
+ "vision_tower.blocks.17.attn.proj.scales": "model.safetensors",
585
+ "vision_tower.blocks.17.attn.proj.weight": "model.safetensors",
586
+ "vision_tower.blocks.17.attn.q_norm.weight": "model.safetensors",
587
+ "vision_tower.blocks.17.attn.qkv.bias": "model.safetensors",
588
+ "vision_tower.blocks.17.attn.qkv.biases": "model.safetensors",
589
+ "vision_tower.blocks.17.attn.qkv.scales": "model.safetensors",
590
+ "vision_tower.blocks.17.attn.qkv.weight": "model.safetensors",
591
+ "vision_tower.blocks.17.mlp.down_proj.bias": "model.safetensors",
592
+ "vision_tower.blocks.17.mlp.down_proj.biases": "model.safetensors",
593
+ "vision_tower.blocks.17.mlp.down_proj.scales": "model.safetensors",
594
+ "vision_tower.blocks.17.mlp.down_proj.weight": "model.safetensors",
595
+ "vision_tower.blocks.17.mlp.gate_proj.bias": "model.safetensors",
596
+ "vision_tower.blocks.17.mlp.gate_proj.biases": "model.safetensors",
597
+ "vision_tower.blocks.17.mlp.gate_proj.scales": "model.safetensors",
598
+ "vision_tower.blocks.17.mlp.gate_proj.weight": "model.safetensors",
599
+ "vision_tower.blocks.17.mlp.up_proj.bias": "model.safetensors",
600
+ "vision_tower.blocks.17.mlp.up_proj.biases": "model.safetensors",
601
+ "vision_tower.blocks.17.mlp.up_proj.scales": "model.safetensors",
602
+ "vision_tower.blocks.17.mlp.up_proj.weight": "model.safetensors",
603
+ "vision_tower.blocks.17.norm1.weight": "model.safetensors",
604
+ "vision_tower.blocks.17.norm2.weight": "model.safetensors",
605
+ "vision_tower.blocks.18.attn.k_norm.weight": "model.safetensors",
606
+ "vision_tower.blocks.18.attn.proj.bias": "model.safetensors",
607
+ "vision_tower.blocks.18.attn.proj.biases": "model.safetensors",
608
+ "vision_tower.blocks.18.attn.proj.scales": "model.safetensors",
609
+ "vision_tower.blocks.18.attn.proj.weight": "model.safetensors",
610
+ "vision_tower.blocks.18.attn.q_norm.weight": "model.safetensors",
611
+ "vision_tower.blocks.18.attn.qkv.bias": "model.safetensors",
612
+ "vision_tower.blocks.18.attn.qkv.biases": "model.safetensors",
613
+ "vision_tower.blocks.18.attn.qkv.scales": "model.safetensors",
614
+ "vision_tower.blocks.18.attn.qkv.weight": "model.safetensors",
615
+ "vision_tower.blocks.18.mlp.down_proj.bias": "model.safetensors",
616
+ "vision_tower.blocks.18.mlp.down_proj.biases": "model.safetensors",
617
+ "vision_tower.blocks.18.mlp.down_proj.scales": "model.safetensors",
618
+ "vision_tower.blocks.18.mlp.down_proj.weight": "model.safetensors",
619
+ "vision_tower.blocks.18.mlp.gate_proj.bias": "model.safetensors",
620
+ "vision_tower.blocks.18.mlp.gate_proj.biases": "model.safetensors",
621
+ "vision_tower.blocks.18.mlp.gate_proj.scales": "model.safetensors",
622
+ "vision_tower.blocks.18.mlp.gate_proj.weight": "model.safetensors",
623
+ "vision_tower.blocks.18.mlp.up_proj.bias": "model.safetensors",
624
+ "vision_tower.blocks.18.mlp.up_proj.biases": "model.safetensors",
625
+ "vision_tower.blocks.18.mlp.up_proj.scales": "model.safetensors",
626
+ "vision_tower.blocks.18.mlp.up_proj.weight": "model.safetensors",
627
+ "vision_tower.blocks.18.norm1.weight": "model.safetensors",
628
+ "vision_tower.blocks.18.norm2.weight": "model.safetensors",
629
+ "vision_tower.blocks.19.attn.k_norm.weight": "model.safetensors",
630
+ "vision_tower.blocks.19.attn.proj.bias": "model.safetensors",
631
+ "vision_tower.blocks.19.attn.proj.biases": "model.safetensors",
632
+ "vision_tower.blocks.19.attn.proj.scales": "model.safetensors",
633
+ "vision_tower.blocks.19.attn.proj.weight": "model.safetensors",
634
+ "vision_tower.blocks.19.attn.q_norm.weight": "model.safetensors",
635
+ "vision_tower.blocks.19.attn.qkv.bias": "model.safetensors",
636
+ "vision_tower.blocks.19.attn.qkv.biases": "model.safetensors",
637
+ "vision_tower.blocks.19.attn.qkv.scales": "model.safetensors",
638
+ "vision_tower.blocks.19.attn.qkv.weight": "model.safetensors",
639
+ "vision_tower.blocks.19.mlp.down_proj.bias": "model.safetensors",
640
+ "vision_tower.blocks.19.mlp.down_proj.biases": "model.safetensors",
641
+ "vision_tower.blocks.19.mlp.down_proj.scales": "model.safetensors",
642
+ "vision_tower.blocks.19.mlp.down_proj.weight": "model.safetensors",
643
+ "vision_tower.blocks.19.mlp.gate_proj.bias": "model.safetensors",
644
+ "vision_tower.blocks.19.mlp.gate_proj.biases": "model.safetensors",
645
+ "vision_tower.blocks.19.mlp.gate_proj.scales": "model.safetensors",
646
+ "vision_tower.blocks.19.mlp.gate_proj.weight": "model.safetensors",
647
+ "vision_tower.blocks.19.mlp.up_proj.bias": "model.safetensors",
648
+ "vision_tower.blocks.19.mlp.up_proj.biases": "model.safetensors",
649
+ "vision_tower.blocks.19.mlp.up_proj.scales": "model.safetensors",
650
+ "vision_tower.blocks.19.mlp.up_proj.weight": "model.safetensors",
651
+ "vision_tower.blocks.19.norm1.weight": "model.safetensors",
652
+ "vision_tower.blocks.19.norm2.weight": "model.safetensors",
653
+ "vision_tower.blocks.2.attn.k_norm.weight": "model.safetensors",
654
+ "vision_tower.blocks.2.attn.proj.bias": "model.safetensors",
655
+ "vision_tower.blocks.2.attn.proj.biases": "model.safetensors",
656
+ "vision_tower.blocks.2.attn.proj.scales": "model.safetensors",
657
+ "vision_tower.blocks.2.attn.proj.weight": "model.safetensors",
658
+ "vision_tower.blocks.2.attn.q_norm.weight": "model.safetensors",
659
+ "vision_tower.blocks.2.attn.qkv.bias": "model.safetensors",
660
+ "vision_tower.blocks.2.attn.qkv.biases": "model.safetensors",
661
+ "vision_tower.blocks.2.attn.qkv.scales": "model.safetensors",
662
+ "vision_tower.blocks.2.attn.qkv.weight": "model.safetensors",
663
+ "vision_tower.blocks.2.mlp.down_proj.bias": "model.safetensors",
664
+ "vision_tower.blocks.2.mlp.down_proj.biases": "model.safetensors",
665
+ "vision_tower.blocks.2.mlp.down_proj.scales": "model.safetensors",
666
+ "vision_tower.blocks.2.mlp.down_proj.weight": "model.safetensors",
667
+ "vision_tower.blocks.2.mlp.gate_proj.bias": "model.safetensors",
668
+ "vision_tower.blocks.2.mlp.gate_proj.biases": "model.safetensors",
669
+ "vision_tower.blocks.2.mlp.gate_proj.scales": "model.safetensors",
670
+ "vision_tower.blocks.2.mlp.gate_proj.weight": "model.safetensors",
671
+ "vision_tower.blocks.2.mlp.up_proj.bias": "model.safetensors",
672
+ "vision_tower.blocks.2.mlp.up_proj.biases": "model.safetensors",
673
+ "vision_tower.blocks.2.mlp.up_proj.scales": "model.safetensors",
674
+ "vision_tower.blocks.2.mlp.up_proj.weight": "model.safetensors",
675
+ "vision_tower.blocks.2.norm1.weight": "model.safetensors",
676
+ "vision_tower.blocks.2.norm2.weight": "model.safetensors",
677
+ "vision_tower.blocks.20.attn.k_norm.weight": "model.safetensors",
678
+ "vision_tower.blocks.20.attn.proj.bias": "model.safetensors",
679
+ "vision_tower.blocks.20.attn.proj.biases": "model.safetensors",
680
+ "vision_tower.blocks.20.attn.proj.scales": "model.safetensors",
681
+ "vision_tower.blocks.20.attn.proj.weight": "model.safetensors",
682
+ "vision_tower.blocks.20.attn.q_norm.weight": "model.safetensors",
683
+ "vision_tower.blocks.20.attn.qkv.bias": "model.safetensors",
684
+ "vision_tower.blocks.20.attn.qkv.biases": "model.safetensors",
685
+ "vision_tower.blocks.20.attn.qkv.scales": "model.safetensors",
686
+ "vision_tower.blocks.20.attn.qkv.weight": "model.safetensors",
687
+ "vision_tower.blocks.20.mlp.down_proj.bias": "model.safetensors",
688
+ "vision_tower.blocks.20.mlp.down_proj.biases": "model.safetensors",
689
+ "vision_tower.blocks.20.mlp.down_proj.scales": "model.safetensors",
690
+ "vision_tower.blocks.20.mlp.down_proj.weight": "model.safetensors",
691
+ "vision_tower.blocks.20.mlp.gate_proj.bias": "model.safetensors",
692
+ "vision_tower.blocks.20.mlp.gate_proj.biases": "model.safetensors",
693
+ "vision_tower.blocks.20.mlp.gate_proj.scales": "model.safetensors",
694
+ "vision_tower.blocks.20.mlp.gate_proj.weight": "model.safetensors",
695
+ "vision_tower.blocks.20.mlp.up_proj.bias": "model.safetensors",
696
+ "vision_tower.blocks.20.mlp.up_proj.biases": "model.safetensors",
697
+ "vision_tower.blocks.20.mlp.up_proj.scales": "model.safetensors",
698
+ "vision_tower.blocks.20.mlp.up_proj.weight": "model.safetensors",
699
+ "vision_tower.blocks.20.norm1.weight": "model.safetensors",
700
+ "vision_tower.blocks.20.norm2.weight": "model.safetensors",
701
+ "vision_tower.blocks.21.attn.k_norm.weight": "model.safetensors",
702
+ "vision_tower.blocks.21.attn.proj.bias": "model.safetensors",
703
+ "vision_tower.blocks.21.attn.proj.biases": "model.safetensors",
704
+ "vision_tower.blocks.21.attn.proj.scales": "model.safetensors",
705
+ "vision_tower.blocks.21.attn.proj.weight": "model.safetensors",
706
+ "vision_tower.blocks.21.attn.q_norm.weight": "model.safetensors",
707
+ "vision_tower.blocks.21.attn.qkv.bias": "model.safetensors",
708
+ "vision_tower.blocks.21.attn.qkv.biases": "model.safetensors",
709
+ "vision_tower.blocks.21.attn.qkv.scales": "model.safetensors",
710
+ "vision_tower.blocks.21.attn.qkv.weight": "model.safetensors",
711
+ "vision_tower.blocks.21.mlp.down_proj.bias": "model.safetensors",
712
+ "vision_tower.blocks.21.mlp.down_proj.biases": "model.safetensors",
713
+ "vision_tower.blocks.21.mlp.down_proj.scales": "model.safetensors",
714
+ "vision_tower.blocks.21.mlp.down_proj.weight": "model.safetensors",
715
+ "vision_tower.blocks.21.mlp.gate_proj.bias": "model.safetensors",
716
+ "vision_tower.blocks.21.mlp.gate_proj.biases": "model.safetensors",
717
+ "vision_tower.blocks.21.mlp.gate_proj.scales": "model.safetensors",
718
+ "vision_tower.blocks.21.mlp.gate_proj.weight": "model.safetensors",
719
+ "vision_tower.blocks.21.mlp.up_proj.bias": "model.safetensors",
720
+ "vision_tower.blocks.21.mlp.up_proj.biases": "model.safetensors",
721
+ "vision_tower.blocks.21.mlp.up_proj.scales": "model.safetensors",
722
+ "vision_tower.blocks.21.mlp.up_proj.weight": "model.safetensors",
723
+ "vision_tower.blocks.21.norm1.weight": "model.safetensors",
724
+ "vision_tower.blocks.21.norm2.weight": "model.safetensors",
725
+ "vision_tower.blocks.22.attn.k_norm.weight": "model.safetensors",
726
+ "vision_tower.blocks.22.attn.proj.bias": "model.safetensors",
727
+ "vision_tower.blocks.22.attn.proj.biases": "model.safetensors",
728
+ "vision_tower.blocks.22.attn.proj.scales": "model.safetensors",
729
+ "vision_tower.blocks.22.attn.proj.weight": "model.safetensors",
730
+ "vision_tower.blocks.22.attn.q_norm.weight": "model.safetensors",
731
+ "vision_tower.blocks.22.attn.qkv.bias": "model.safetensors",
732
+ "vision_tower.blocks.22.attn.qkv.biases": "model.safetensors",
733
+ "vision_tower.blocks.22.attn.qkv.scales": "model.safetensors",
734
+ "vision_tower.blocks.22.attn.qkv.weight": "model.safetensors",
735
+ "vision_tower.blocks.22.mlp.down_proj.bias": "model.safetensors",
736
+ "vision_tower.blocks.22.mlp.down_proj.biases": "model.safetensors",
737
+ "vision_tower.blocks.22.mlp.down_proj.scales": "model.safetensors",
738
+ "vision_tower.blocks.22.mlp.down_proj.weight": "model.safetensors",
739
+ "vision_tower.blocks.22.mlp.gate_proj.bias": "model.safetensors",
740
+ "vision_tower.blocks.22.mlp.gate_proj.biases": "model.safetensors",
741
+ "vision_tower.blocks.22.mlp.gate_proj.scales": "model.safetensors",
742
+ "vision_tower.blocks.22.mlp.gate_proj.weight": "model.safetensors",
743
+ "vision_tower.blocks.22.mlp.up_proj.bias": "model.safetensors",
744
+ "vision_tower.blocks.22.mlp.up_proj.biases": "model.safetensors",
745
+ "vision_tower.blocks.22.mlp.up_proj.scales": "model.safetensors",
746
+ "vision_tower.blocks.22.mlp.up_proj.weight": "model.safetensors",
747
+ "vision_tower.blocks.22.norm1.weight": "model.safetensors",
748
+ "vision_tower.blocks.22.norm2.weight": "model.safetensors",
749
+ "vision_tower.blocks.23.attn.k_norm.weight": "model.safetensors",
750
+ "vision_tower.blocks.23.attn.proj.bias": "model.safetensors",
751
+ "vision_tower.blocks.23.attn.proj.biases": "model.safetensors",
752
+ "vision_tower.blocks.23.attn.proj.scales": "model.safetensors",
753
+ "vision_tower.blocks.23.attn.proj.weight": "model.safetensors",
754
+ "vision_tower.blocks.23.attn.q_norm.weight": "model.safetensors",
755
+ "vision_tower.blocks.23.attn.qkv.bias": "model.safetensors",
756
+ "vision_tower.blocks.23.attn.qkv.biases": "model.safetensors",
757
+ "vision_tower.blocks.23.attn.qkv.scales": "model.safetensors",
758
+ "vision_tower.blocks.23.attn.qkv.weight": "model.safetensors",
759
+ "vision_tower.blocks.23.mlp.down_proj.bias": "model.safetensors",
760
+ "vision_tower.blocks.23.mlp.down_proj.biases": "model.safetensors",
761
+ "vision_tower.blocks.23.mlp.down_proj.scales": "model.safetensors",
762
+ "vision_tower.blocks.23.mlp.down_proj.weight": "model.safetensors",
763
+ "vision_tower.blocks.23.mlp.gate_proj.bias": "model.safetensors",
764
+ "vision_tower.blocks.23.mlp.gate_proj.biases": "model.safetensors",
765
+ "vision_tower.blocks.23.mlp.gate_proj.scales": "model.safetensors",
766
+ "vision_tower.blocks.23.mlp.gate_proj.weight": "model.safetensors",
767
+ "vision_tower.blocks.23.mlp.up_proj.bias": "model.safetensors",
768
+ "vision_tower.blocks.23.mlp.up_proj.biases": "model.safetensors",
769
+ "vision_tower.blocks.23.mlp.up_proj.scales": "model.safetensors",
770
+ "vision_tower.blocks.23.mlp.up_proj.weight": "model.safetensors",
771
+ "vision_tower.blocks.23.norm1.weight": "model.safetensors",
772
+ "vision_tower.blocks.23.norm2.weight": "model.safetensors",
773
+ "vision_tower.blocks.3.attn.k_norm.weight": "model.safetensors",
774
+ "vision_tower.blocks.3.attn.proj.bias": "model.safetensors",
775
+ "vision_tower.blocks.3.attn.proj.biases": "model.safetensors",
776
+ "vision_tower.blocks.3.attn.proj.scales": "model.safetensors",
777
+ "vision_tower.blocks.3.attn.proj.weight": "model.safetensors",
778
+ "vision_tower.blocks.3.attn.q_norm.weight": "model.safetensors",
779
+ "vision_tower.blocks.3.attn.qkv.bias": "model.safetensors",
780
+ "vision_tower.blocks.3.attn.qkv.biases": "model.safetensors",
781
+ "vision_tower.blocks.3.attn.qkv.scales": "model.safetensors",
782
+ "vision_tower.blocks.3.attn.qkv.weight": "model.safetensors",
783
+ "vision_tower.blocks.3.mlp.down_proj.bias": "model.safetensors",
784
+ "vision_tower.blocks.3.mlp.down_proj.biases": "model.safetensors",
785
+ "vision_tower.blocks.3.mlp.down_proj.scales": "model.safetensors",
786
+ "vision_tower.blocks.3.mlp.down_proj.weight": "model.safetensors",
787
+ "vision_tower.blocks.3.mlp.gate_proj.bias": "model.safetensors",
788
+ "vision_tower.blocks.3.mlp.gate_proj.biases": "model.safetensors",
789
+ "vision_tower.blocks.3.mlp.gate_proj.scales": "model.safetensors",
790
+ "vision_tower.blocks.3.mlp.gate_proj.weight": "model.safetensors",
791
+ "vision_tower.blocks.3.mlp.up_proj.bias": "model.safetensors",
792
+ "vision_tower.blocks.3.mlp.up_proj.biases": "model.safetensors",
793
+ "vision_tower.blocks.3.mlp.up_proj.scales": "model.safetensors",
794
+ "vision_tower.blocks.3.mlp.up_proj.weight": "model.safetensors",
795
+ "vision_tower.blocks.3.norm1.weight": "model.safetensors",
796
+ "vision_tower.blocks.3.norm2.weight": "model.safetensors",
797
+ "vision_tower.blocks.4.attn.k_norm.weight": "model.safetensors",
798
+ "vision_tower.blocks.4.attn.proj.bias": "model.safetensors",
799
+ "vision_tower.blocks.4.attn.proj.biases": "model.safetensors",
800
+ "vision_tower.blocks.4.attn.proj.scales": "model.safetensors",
801
+ "vision_tower.blocks.4.attn.proj.weight": "model.safetensors",
802
+ "vision_tower.blocks.4.attn.q_norm.weight": "model.safetensors",
803
+ "vision_tower.blocks.4.attn.qkv.bias": "model.safetensors",
804
+ "vision_tower.blocks.4.attn.qkv.biases": "model.safetensors",
805
+ "vision_tower.blocks.4.attn.qkv.scales": "model.safetensors",
806
+ "vision_tower.blocks.4.attn.qkv.weight": "model.safetensors",
807
+ "vision_tower.blocks.4.mlp.down_proj.bias": "model.safetensors",
808
+ "vision_tower.blocks.4.mlp.down_proj.biases": "model.safetensors",
809
+ "vision_tower.blocks.4.mlp.down_proj.scales": "model.safetensors",
810
+ "vision_tower.blocks.4.mlp.down_proj.weight": "model.safetensors",
811
+ "vision_tower.blocks.4.mlp.gate_proj.bias": "model.safetensors",
812
+ "vision_tower.blocks.4.mlp.gate_proj.biases": "model.safetensors",
813
+ "vision_tower.blocks.4.mlp.gate_proj.scales": "model.safetensors",
814
+ "vision_tower.blocks.4.mlp.gate_proj.weight": "model.safetensors",
815
+ "vision_tower.blocks.4.mlp.up_proj.bias": "model.safetensors",
816
+ "vision_tower.blocks.4.mlp.up_proj.biases": "model.safetensors",
817
+ "vision_tower.blocks.4.mlp.up_proj.scales": "model.safetensors",
818
+ "vision_tower.blocks.4.mlp.up_proj.weight": "model.safetensors",
819
+ "vision_tower.blocks.4.norm1.weight": "model.safetensors",
820
+ "vision_tower.blocks.4.norm2.weight": "model.safetensors",
821
+ "vision_tower.blocks.5.attn.k_norm.weight": "model.safetensors",
822
+ "vision_tower.blocks.5.attn.proj.bias": "model.safetensors",
823
+ "vision_tower.blocks.5.attn.proj.biases": "model.safetensors",
824
+ "vision_tower.blocks.5.attn.proj.scales": "model.safetensors",
825
+ "vision_tower.blocks.5.attn.proj.weight": "model.safetensors",
826
+ "vision_tower.blocks.5.attn.q_norm.weight": "model.safetensors",
827
+ "vision_tower.blocks.5.attn.qkv.bias": "model.safetensors",
828
+ "vision_tower.blocks.5.attn.qkv.biases": "model.safetensors",
829
+ "vision_tower.blocks.5.attn.qkv.scales": "model.safetensors",
830
+ "vision_tower.blocks.5.attn.qkv.weight": "model.safetensors",
831
+ "vision_tower.blocks.5.mlp.down_proj.bias": "model.safetensors",
832
+ "vision_tower.blocks.5.mlp.down_proj.biases": "model.safetensors",
833
+ "vision_tower.blocks.5.mlp.down_proj.scales": "model.safetensors",
834
+ "vision_tower.blocks.5.mlp.down_proj.weight": "model.safetensors",
835
+ "vision_tower.blocks.5.mlp.gate_proj.bias": "model.safetensors",
836
+ "vision_tower.blocks.5.mlp.gate_proj.biases": "model.safetensors",
837
+ "vision_tower.blocks.5.mlp.gate_proj.scales": "model.safetensors",
838
+ "vision_tower.blocks.5.mlp.gate_proj.weight": "model.safetensors",
839
+ "vision_tower.blocks.5.mlp.up_proj.bias": "model.safetensors",
840
+ "vision_tower.blocks.5.mlp.up_proj.biases": "model.safetensors",
841
+ "vision_tower.blocks.5.mlp.up_proj.scales": "model.safetensors",
842
+ "vision_tower.blocks.5.mlp.up_proj.weight": "model.safetensors",
843
+ "vision_tower.blocks.5.norm1.weight": "model.safetensors",
844
+ "vision_tower.blocks.5.norm2.weight": "model.safetensors",
845
+ "vision_tower.blocks.6.attn.k_norm.weight": "model.safetensors",
846
+ "vision_tower.blocks.6.attn.proj.bias": "model.safetensors",
847
+ "vision_tower.blocks.6.attn.proj.biases": "model.safetensors",
848
+ "vision_tower.blocks.6.attn.proj.scales": "model.safetensors",
849
+ "vision_tower.blocks.6.attn.proj.weight": "model.safetensors",
850
+ "vision_tower.blocks.6.attn.q_norm.weight": "model.safetensors",
851
+ "vision_tower.blocks.6.attn.qkv.bias": "model.safetensors",
852
+ "vision_tower.blocks.6.attn.qkv.biases": "model.safetensors",
853
+ "vision_tower.blocks.6.attn.qkv.scales": "model.safetensors",
854
+ "vision_tower.blocks.6.attn.qkv.weight": "model.safetensors",
855
+ "vision_tower.blocks.6.mlp.down_proj.bias": "model.safetensors",
856
+ "vision_tower.blocks.6.mlp.down_proj.biases": "model.safetensors",
857
+ "vision_tower.blocks.6.mlp.down_proj.scales": "model.safetensors",
858
+ "vision_tower.blocks.6.mlp.down_proj.weight": "model.safetensors",
859
+ "vision_tower.blocks.6.mlp.gate_proj.bias": "model.safetensors",
860
+ "vision_tower.blocks.6.mlp.gate_proj.biases": "model.safetensors",
861
+ "vision_tower.blocks.6.mlp.gate_proj.scales": "model.safetensors",
862
+ "vision_tower.blocks.6.mlp.gate_proj.weight": "model.safetensors",
863
+ "vision_tower.blocks.6.mlp.up_proj.bias": "model.safetensors",
864
+ "vision_tower.blocks.6.mlp.up_proj.biases": "model.safetensors",
865
+ "vision_tower.blocks.6.mlp.up_proj.scales": "model.safetensors",
866
+ "vision_tower.blocks.6.mlp.up_proj.weight": "model.safetensors",
867
+ "vision_tower.blocks.6.norm1.weight": "model.safetensors",
868
+ "vision_tower.blocks.6.norm2.weight": "model.safetensors",
869
+ "vision_tower.blocks.7.attn.k_norm.weight": "model.safetensors",
870
+ "vision_tower.blocks.7.attn.proj.bias": "model.safetensors",
871
+ "vision_tower.blocks.7.attn.proj.biases": "model.safetensors",
872
+ "vision_tower.blocks.7.attn.proj.scales": "model.safetensors",
873
+ "vision_tower.blocks.7.attn.proj.weight": "model.safetensors",
874
+ "vision_tower.blocks.7.attn.q_norm.weight": "model.safetensors",
875
+ "vision_tower.blocks.7.attn.qkv.bias": "model.safetensors",
876
+ "vision_tower.blocks.7.attn.qkv.biases": "model.safetensors",
877
+ "vision_tower.blocks.7.attn.qkv.scales": "model.safetensors",
878
+ "vision_tower.blocks.7.attn.qkv.weight": "model.safetensors",
879
+ "vision_tower.blocks.7.mlp.down_proj.bias": "model.safetensors",
880
+ "vision_tower.blocks.7.mlp.down_proj.biases": "model.safetensors",
881
+ "vision_tower.blocks.7.mlp.down_proj.scales": "model.safetensors",
882
+ "vision_tower.blocks.7.mlp.down_proj.weight": "model.safetensors",
883
+ "vision_tower.blocks.7.mlp.gate_proj.bias": "model.safetensors",
884
+ "vision_tower.blocks.7.mlp.gate_proj.biases": "model.safetensors",
885
+ "vision_tower.blocks.7.mlp.gate_proj.scales": "model.safetensors",
886
+ "vision_tower.blocks.7.mlp.gate_proj.weight": "model.safetensors",
887
+ "vision_tower.blocks.7.mlp.up_proj.bias": "model.safetensors",
888
+ "vision_tower.blocks.7.mlp.up_proj.biases": "model.safetensors",
889
+ "vision_tower.blocks.7.mlp.up_proj.scales": "model.safetensors",
890
+ "vision_tower.blocks.7.mlp.up_proj.weight": "model.safetensors",
891
+ "vision_tower.blocks.7.norm1.weight": "model.safetensors",
892
+ "vision_tower.blocks.7.norm2.weight": "model.safetensors",
893
+ "vision_tower.blocks.8.attn.k_norm.weight": "model.safetensors",
894
+ "vision_tower.blocks.8.attn.proj.bias": "model.safetensors",
895
+ "vision_tower.blocks.8.attn.proj.biases": "model.safetensors",
896
+ "vision_tower.blocks.8.attn.proj.scales": "model.safetensors",
897
+ "vision_tower.blocks.8.attn.proj.weight": "model.safetensors",
898
+ "vision_tower.blocks.8.attn.q_norm.weight": "model.safetensors",
899
+ "vision_tower.blocks.8.attn.qkv.bias": "model.safetensors",
900
+ "vision_tower.blocks.8.attn.qkv.biases": "model.safetensors",
901
+ "vision_tower.blocks.8.attn.qkv.scales": "model.safetensors",
902
+ "vision_tower.blocks.8.attn.qkv.weight": "model.safetensors",
903
+ "vision_tower.blocks.8.mlp.down_proj.bias": "model.safetensors",
904
+ "vision_tower.blocks.8.mlp.down_proj.biases": "model.safetensors",
905
+ "vision_tower.blocks.8.mlp.down_proj.scales": "model.safetensors",
906
+ "vision_tower.blocks.8.mlp.down_proj.weight": "model.safetensors",
907
+ "vision_tower.blocks.8.mlp.gate_proj.bias": "model.safetensors",
908
+ "vision_tower.blocks.8.mlp.gate_proj.biases": "model.safetensors",
909
+ "vision_tower.blocks.8.mlp.gate_proj.scales": "model.safetensors",
910
+ "vision_tower.blocks.8.mlp.gate_proj.weight": "model.safetensors",
911
+ "vision_tower.blocks.8.mlp.up_proj.bias": "model.safetensors",
912
+ "vision_tower.blocks.8.mlp.up_proj.biases": "model.safetensors",
913
+ "vision_tower.blocks.8.mlp.up_proj.scales": "model.safetensors",
914
+ "vision_tower.blocks.8.mlp.up_proj.weight": "model.safetensors",
915
+ "vision_tower.blocks.8.norm1.weight": "model.safetensors",
916
+ "vision_tower.blocks.8.norm2.weight": "model.safetensors",
917
+ "vision_tower.blocks.9.attn.k_norm.weight": "model.safetensors",
918
+ "vision_tower.blocks.9.attn.proj.bias": "model.safetensors",
919
+ "vision_tower.blocks.9.attn.proj.biases": "model.safetensors",
920
+ "vision_tower.blocks.9.attn.proj.scales": "model.safetensors",
921
+ "vision_tower.blocks.9.attn.proj.weight": "model.safetensors",
922
+ "vision_tower.blocks.9.attn.q_norm.weight": "model.safetensors",
923
+ "vision_tower.blocks.9.attn.qkv.bias": "model.safetensors",
924
+ "vision_tower.blocks.9.attn.qkv.biases": "model.safetensors",
925
+ "vision_tower.blocks.9.attn.qkv.scales": "model.safetensors",
926
+ "vision_tower.blocks.9.attn.qkv.weight": "model.safetensors",
927
+ "vision_tower.blocks.9.mlp.down_proj.bias": "model.safetensors",
928
+ "vision_tower.blocks.9.mlp.down_proj.biases": "model.safetensors",
929
+ "vision_tower.blocks.9.mlp.down_proj.scales": "model.safetensors",
930
+ "vision_tower.blocks.9.mlp.down_proj.weight": "model.safetensors",
931
+ "vision_tower.blocks.9.mlp.gate_proj.bias": "model.safetensors",
932
+ "vision_tower.blocks.9.mlp.gate_proj.biases": "model.safetensors",
933
+ "vision_tower.blocks.9.mlp.gate_proj.scales": "model.safetensors",
934
+ "vision_tower.blocks.9.mlp.gate_proj.weight": "model.safetensors",
935
+ "vision_tower.blocks.9.mlp.up_proj.bias": "model.safetensors",
936
+ "vision_tower.blocks.9.mlp.up_proj.biases": "model.safetensors",
937
+ "vision_tower.blocks.9.mlp.up_proj.scales": "model.safetensors",
938
+ "vision_tower.blocks.9.mlp.up_proj.weight": "model.safetensors",
939
+ "vision_tower.blocks.9.norm1.weight": "model.safetensors",
940
+ "vision_tower.blocks.9.norm2.weight": "model.safetensors",
941
+ "vision_tower.downsample.bias": "model.safetensors",
942
+ "vision_tower.downsample.weight": "model.safetensors",
943
+ "vision_tower.merger.down_proj.biases": "model.safetensors",
944
+ "vision_tower.merger.down_proj.scales": "model.safetensors",
945
+ "vision_tower.merger.down_proj.weight": "model.safetensors",
946
+ "vision_tower.merger.gate_proj.biases": "model.safetensors",
947
+ "vision_tower.merger.gate_proj.scales": "model.safetensors",
948
+ "vision_tower.merger.gate_proj.weight": "model.safetensors",
949
+ "vision_tower.merger.post_projection_norm.bias": "model.safetensors",
950
+ "vision_tower.merger.post_projection_norm.weight": "model.safetensors",
951
+ "vision_tower.merger.proj.biases": "model.safetensors",
952
+ "vision_tower.merger.proj.scales": "model.safetensors",
953
+ "vision_tower.merger.proj.weight": "model.safetensors",
954
+ "vision_tower.merger.up_proj.biases": "model.safetensors",
955
+ "vision_tower.merger.up_proj.scales": "model.safetensors",
956
+ "vision_tower.merger.up_proj.weight": "model.safetensors",
957
+ "vision_tower.patch_embed.proj.bias": "model.safetensors",
958
+ "vision_tower.patch_embed.proj.weight": "model.safetensors",
959
+ "vision_tower.post_layernorm.weight": "model.safetensors"
960
+ }
961
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "size": {
3
+ "shortest_edge": 12544,
4
+ "longest_edge": 9633792
5
+ },
6
+ "do_rescale": true,
7
+ "patch_size": 14,
8
+ "temporal_patch_size": 2,
9
+ "merge_size": 2,
10
+ "image_mean": [
11
+ 0.48145466,
12
+ 0.4578275,
13
+ 0.40821073
14
+ ],
15
+ "image_std": [
16
+ 0.26862954,
17
+ 0.26130258,
18
+ 0.27577711
19
+ ],
20
+ "image_processor_type": "GlmOcrImageProcessor",
21
+ "processor_class": "GlmOcrProcessor"
22
+ }
processor_config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_processor": {
3
+ "data_format": "channels_first",
4
+ "do_convert_rgb": true,
5
+ "do_normalize": true,
6
+ "do_rescale": true,
7
+ "do_resize": true,
8
+ "image_mean": [
9
+ 0.48145466,
10
+ 0.4578275,
11
+ 0.40821073
12
+ ],
13
+ "image_processor_type": "GlmOcrImageProcessorFast",
14
+ "image_std": [
15
+ 0.26862954,
16
+ 0.26130258,
17
+ 0.27577711
18
+ ],
19
+ "merge_size": 2,
20
+ "patch_size": 14,
21
+ "resample": 3,
22
+ "rescale_factor": 0.00392156862745098,
23
+ "size": {
24
+ "longest_edge": 9633792,
25
+ "shortest_edge": 12544
26
+ },
27
+ "temporal_patch_size": 2
28
+ },
29
+ "processor_class": "GlmOcrProcessor"
30
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "clean_up_tokenization_spaces": false,
4
+ "eos_token": "<|endoftext|>",
5
+ "extra_special_tokens": [
6
+ "<|endoftext|>",
7
+ "[MASK]",
8
+ "[gMASK]",
9
+ "[sMASK]",
10
+ "<sop>",
11
+ "<eop>",
12
+ "<|system|>",
13
+ "<|user|>",
14
+ "<|assistant|>",
15
+ "<|observation|>",
16
+ "<|begin_of_image|>",
17
+ "<|end_of_image|>",
18
+ "<|begin_of_video|>",
19
+ "<|end_of_video|>",
20
+ "<|begin_of_audio|>",
21
+ "<|end_of_audio|>",
22
+ "<|begin_of_transcription|>",
23
+ "<|end_of_transcription|>",
24
+ "<|code_prefix|>",
25
+ "<|code_middle|>",
26
+ "<|code_suffix|>",
27
+ "<think>",
28
+ "</think>",
29
+ "<tool_call>",
30
+ "</tool_call>",
31
+ "<tool_response>",
32
+ "</tool_response>",
33
+ "<arg_key>",
34
+ "</arg_key>",
35
+ "<arg_value>",
36
+ "</arg_value>",
37
+ "/nothink",
38
+ "<|begin_of_box|>",
39
+ "<|end_of_box|>",
40
+ "<|image|>",
41
+ "<|video|>"
42
+ ],
43
+ "is_local": true,
44
+ "model_max_length": 655380,
45
+ "pad_token": "<|endoftext|>",
46
+ "padding_side": "left",
47
+ "processor_class": "GlmOcrProcessor",
48
+ "tokenizer_class": "TokenizersBackend"
49
+ }