Hugging Face's logo

BlueSR
/
HerbaScan_Weights

Joblib
ONNX
Safetensors
GGUF
conversational
Model card Files
xet
 Community
12
MicWJJ commited on
Commit
be9c826
·
verified ·
1 Parent(s): cf6f482

Model upload for OCR

Browse files
Files changed (22) hide show
  1. .gitattributes +2 -0
  2. models/glm-ocr/chat_template.jinja +140 -0
  3. models/glm-ocr/config.json +70 -0
  4. models/glm-ocr/generation_config.json +10 -0
  5. models/glm-ocr/model.safetensors +3 -0
  6. models/glm-ocr/processor_config.json +63 -0
  7. models/glm-ocr/tokenizer.json +0 -0
  8. models/glm-ocr/tokenizer_config.json +50 -0
  9. models/qwen-vl-2b/chat_template.jinja +7 -0
  10. models/qwen-vl-2b/config.json +93 -0
  11. models/qwen-vl-2b/generation_config.json +14 -0
  12. models/qwen-vl-2b/model.safetensors +3 -0
  13. models/qwen-vl-2b/processor_config.json +59 -0
  14. models/qwen-vl-2b/tokenizer.json +3 -0
  15. models/qwen-vl-2b/tokenizer_config.json +31 -0
  16. models/qwen-vl/chat_template.jinja +7 -0
  17. models/qwen-vl/config.json +109 -0
  18. models/qwen-vl/generation_config.json +12 -0
  19. models/qwen-vl/model.safetensors +3 -0
  20. models/qwen-vl/processor_config.json +63 -0
  21. models/qwen-vl/tokenizer.json +3 -0
  22. models/qwen-vl/tokenizer_config.json +30 -0
.gitattributes CHANGED
@@ -655,3 +655,5 @@ measurement_pipeline/herbarium_rulers/ruler_crops_normalized/L.1934962_Dillen_ru
655
  measurement_pipeline/herbarium_rulers/ruler_crops_normalized/L__0007231_Dillen_ruler_block-10-Naturalis_0.jpg filter=lfs diff=lfs merge=lfs -text
656
  measurement_pipeline/herbarium_rulers/ruler_crops_normalized/L__0007231_Dillen_ruler_block-10-Naturalis_1.jpg filter=lfs diff=lfs merge=lfs -text
657
  measurement_pipeline/plantsam/BBS2_1024_2_epoch5.torch filter=lfs diff=lfs merge=lfs -text
 
 
 
655
  measurement_pipeline/herbarium_rulers/ruler_crops_normalized/L__0007231_Dillen_ruler_block-10-Naturalis_0.jpg filter=lfs diff=lfs merge=lfs -text
656
  measurement_pipeline/herbarium_rulers/ruler_crops_normalized/L__0007231_Dillen_ruler_block-10-Naturalis_1.jpg filter=lfs diff=lfs merge=lfs -text
657
  measurement_pipeline/plantsam/BBS2_1024_2_epoch5.torch filter=lfs diff=lfs merge=lfs -text
658
+ models/qwen-vl-2b/tokenizer.json filter=lfs diff=lfs merge=lfs -text
659
+ models/qwen-vl/tokenizer.json filter=lfs diff=lfs merge=lfs -text
models/glm-ocr/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 -%}
models/glm-ocr/config.json ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "GlmOcrForConditionalGeneration"
4
+ ],
5
+ "dtype": "bfloat16",
6
+ "image_end_token_id": 59257,
7
+ "image_start_token_id": 59256,
8
+ "image_token_id": 59280,
9
+ "model_type": "glm_ocr",
10
+ "text_config": {
11
+ "attention_bias": false,
12
+ "attention_dropout": 0.0,
13
+ "dtype": "bfloat16",
14
+ "eos_token_id": [
15
+ 59246,
16
+ 59253
17
+ ],
18
+ "head_dim": 128,
19
+ "hidden_act": "silu",
20
+ "hidden_size": 1536,
21
+ "initializer_range": 0.02,
22
+ "intermediate_size": 4608,
23
+ "max_position_embeddings": 131072,
24
+ "model_type": "glm_ocr_text",
25
+ "num_attention_heads": 16,
26
+ "num_hidden_layers": 16,
27
+ "num_key_value_heads": 8,
28
+ "num_nextn_predict_layers": 1,
29
+ "pad_token_id": 59246,
30
+ "rms_norm_eps": 1e-05,
31
+ "rope_parameters": {
32
+ "mrope_section": [
33
+ 16,
34
+ 24,
35
+ 24
36
+ ],
37
+ "partial_rotary_factor": 1.0,
38
+ "rope_theta": 10000,
39
+ "rope_type": "default"
40
+ },
41
+ "tie_word_embeddings": false,
42
+ "use_cache": true,
43
+ "vocab_size": 59392
44
+ },
45
+ "tie_word_embeddings": false,
46
+ "transformers_version": "5.8.0.dev0",
47
+ "video_end_token_id": 59259,
48
+ "video_start_token_id": 59258,
49
+ "video_token_id": 59281,
50
+ "vision_config": {
51
+ "attention_bias": true,
52
+ "attention_dropout": 0.0,
53
+ "depth": 24,
54
+ "dtype": "bfloat16",
55
+ "hidden_act": "silu",
56
+ "hidden_dropout_prob": 0.0,
57
+ "hidden_size": 1024,
58
+ "image_size": 336,
59
+ "in_channels": 3,
60
+ "initializer_range": 0.02,
61
+ "intermediate_size": 4096,
62
+ "model_type": "glm_ocr_vision",
63
+ "num_heads": 16,
64
+ "out_hidden_size": 1536,
65
+ "patch_size": 14,
66
+ "rms_norm_eps": 1e-05,
67
+ "spatial_merge_size": 2,
68
+ "temporal_patch_size": 2
69
+ }
70
+ }
models/glm-ocr/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": "5.8.0.dev0"
10
+ }
models/glm-ocr/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7deb53e3f51baa7f61b73536cefe70f3075acdf4f58506e18e59490943d0942b
3
+ size 2214872552
models/glm-ocr/processor_config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_processor": {
3
+ "do_convert_rgb": true,
4
+ "do_normalize": true,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "image_mean": [
8
+ 0.48145466,
9
+ 0.4578275,
10
+ 0.40821073
11
+ ],
12
+ "image_processor_type": "Glm46VImageProcessor",
13
+ "image_std": [
14
+ 0.26862954,
15
+ 0.26130258,
16
+ 0.27577711
17
+ ],
18
+ "merge_size": 2,
19
+ "patch_size": 14,
20
+ "resample": 3,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "longest_edge": 9633792,
24
+ "shortest_edge": 12544
25
+ },
26
+ "temporal_patch_size": 2
27
+ },
28
+ "processor_class": "Glm46VProcessor",
29
+ "video_processor": {
30
+ "do_convert_rgb": true,
31
+ "do_normalize": true,
32
+ "do_rescale": true,
33
+ "do_resize": true,
34
+ "do_sample_frames": true,
35
+ "fps": 2,
36
+ "image_mean": [
37
+ 0.48145466,
38
+ 0.4578275,
39
+ 0.40821073
40
+ ],
41
+ "image_std": [
42
+ 0.26862954,
43
+ 0.26130258,
44
+ 0.27577711
45
+ ],
46
+ "max_duration": 300,
47
+ "max_image_size": {
48
+ "longest_edge": 47040000
49
+ },
50
+ "merge_size": 2,
51
+ "num_frames": 16,
52
+ "patch_size": 14,
53
+ "resample": 3,
54
+ "rescale_factor": 0.00392156862745098,
55
+ "return_metadata": false,
56
+ "size": {
57
+ "longest_edge": 9633792,
58
+ "shortest_edge": 12544
59
+ },
60
+ "temporal_patch_size": 2,
61
+ "video_processor_type": "Glm46VVideoProcessor"
62
+ }
63
+ }
models/glm-ocr/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
models/glm-ocr/tokenizer_config.json ADDED
@@ -0,0 +1,50 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": false,
44
+ "local_files_only": false,
45
+ "model_max_length": 655380,
46
+ "pad_token": "<|endoftext|>",
47
+ "padding_side": "left",
48
+ "processor_class": "Glm46VProcessor",
49
+ "tokenizer_class": "TokenizersBackend"
50
+ }
models/qwen-vl-2b/chat_template.jinja ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system
2
+ You are a helpful assistant.<|im_end|>
3
+ {% endif %}<|im_start|>{{ message['role'] }}
4
+ {% if message['content'] is string %}{{ message['content'] }}<|im_end|>
5
+ {% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>
6
+ {% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant
7
+ {% endif %}
models/qwen-vl-2b/config.json ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2VLForConditionalGeneration"
4
+ ],
5
+ "dtype": "bfloat16",
6
+ "image_token_id": 151655,
7
+ "model_type": "qwen2_vl",
8
+ "text_config": {
9
+ "attention_dropout": 0.0,
10
+ "bos_token_id": 151643,
11
+ "dtype": "bfloat16",
12
+ "eos_token_id": 151645,
13
+ "hidden_act": "silu",
14
+ "hidden_size": 1536,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 8960,
17
+ "layer_types": [
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention"
46
+ ],
47
+ "max_position_embeddings": 32768,
48
+ "max_window_layers": 28,
49
+ "model_type": "qwen2_vl_text",
50
+ "num_attention_heads": 12,
51
+ "num_hidden_layers": 28,
52
+ "num_key_value_heads": 2,
53
+ "pad_token_id": null,
54
+ "rms_norm_eps": 1e-06,
55
+ "rope_parameters": {
56
+ "mrope_section": [
57
+ 16,
58
+ 24,
59
+ 24
60
+ ],
61
+ "rope_theta": 1000000.0,
62
+ "rope_type": "default",
63
+ "type": "mrope"
64
+ },
65
+ "sliding_window": null,
66
+ "use_cache": true,
67
+ "use_sliding_window": false,
68
+ "vocab_size": 151936
69
+ },
70
+ "tie_word_embeddings": true,
71
+ "transformers_version": "5.5.0",
72
+ "video_token_id": 151656,
73
+ "vision_config": {
74
+ "depth": 32,
75
+ "dtype": "bfloat16",
76
+ "embed_dim": 1280,
77
+ "hidden_act": "quick_gelu",
78
+ "hidden_size": 1536,
79
+ "in_channels": 3,
80
+ "in_chans": 3,
81
+ "initializer_range": 0.02,
82
+ "mlp_ratio": 4,
83
+ "model_type": "qwen2_vl",
84
+ "num_heads": 16,
85
+ "patch_size": 14,
86
+ "spatial_merge_size": 2,
87
+ "spatial_patch_size": 14,
88
+ "temporal_patch_size": 2
89
+ },
90
+ "vision_end_token_id": 151653,
91
+ "vision_start_token_id": 151652,
92
+ "vision_token_id": 151654
93
+ }
models/qwen-vl-2b/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 151643,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "repetition_penalty": 1.0,
10
+ "temperature": 0.01,
11
+ "top_k": 1,
12
+ "top_p": 0.001,
13
+ "transformers_version": "5.5.0"
14
+ }
models/qwen-vl-2b/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5484e3bf8c529e19f7d948297e39888d3aad66583e61b3747fc7934bf1664b7
3
+ size 4418053192
models/qwen-vl-2b/processor_config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_processor": {
3
+ "do_convert_rgb": true,
4
+ "do_normalize": true,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "image_mean": [
8
+ 0.48145466,
9
+ 0.4578275,
10
+ 0.40821073
11
+ ],
12
+ "image_processor_type": "Qwen2VLImageProcessor",
13
+ "image_std": [
14
+ 0.26862954,
15
+ 0.26130258,
16
+ 0.27577711
17
+ ],
18
+ "merge_size": 2,
19
+ "patch_size": 14,
20
+ "resample": 3,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "longest_edge": 12845056,
24
+ "shortest_edge": 3136
25
+ },
26
+ "temporal_patch_size": 2
27
+ },
28
+ "processor_class": "Qwen2VLProcessor",
29
+ "video_processor": {
30
+ "do_convert_rgb": true,
31
+ "do_normalize": true,
32
+ "do_rescale": true,
33
+ "do_resize": true,
34
+ "do_sample_frames": false,
35
+ "image_mean": [
36
+ 0.48145466,
37
+ 0.4578275,
38
+ 0.40821073
39
+ ],
40
+ "image_std": [
41
+ 0.26862954,
42
+ 0.26130258,
43
+ 0.27577711
44
+ ],
45
+ "max_frames": 768,
46
+ "merge_size": 2,
47
+ "min_frames": 4,
48
+ "patch_size": 14,
49
+ "resample": 3,
50
+ "rescale_factor": 0.00392156862745098,
51
+ "return_metadata": false,
52
+ "size": {
53
+ "longest_edge": 12845056,
54
+ "shortest_edge": 3136
55
+ },
56
+ "temporal_patch_size": 2,
57
+ "video_processor_type": "Qwen2VLVideoProcessor"
58
+ }
59
+ }
models/qwen-vl-2b/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff8cce547abc110590d19c6b5b6e0c6a7b4c8d1012d78b9c42131bae7f494a02
3
+ size 11420367
models/qwen-vl-2b/tokenizer_config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|im_start|>",
10
+ "<|im_end|>",
11
+ "<|object_ref_start|>",
12
+ "<|object_ref_end|>",
13
+ "<|box_start|>",
14
+ "<|box_end|>",
15
+ "<|quad_start|>",
16
+ "<|quad_end|>",
17
+ "<|vision_start|>",
18
+ "<|vision_end|>",
19
+ "<|vision_pad|>",
20
+ "<|image_pad|>",
21
+ "<|video_pad|>"
22
+ ],
23
+ "is_local": false,
24
+ "model_max_length": 32768,
25
+ "pad_token": "<|endoftext|>",
26
+ "padding_side": "left",
27
+ "processor_class": "Qwen2VLProcessor",
28
+ "split_special_tokens": false,
29
+ "tokenizer_class": "Qwen2Tokenizer",
30
+ "unk_token": null
31
+ }
models/qwen-vl/chat_template.jinja ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system
2
+ You are a helpful assistant.<|im_end|>
3
+ {% endif %}<|im_start|>{{ message['role'] }}
4
+ {% if message['content'] is string %}{{ message['content'] }}<|im_end|>
5
+ {% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>
6
+ {% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant
7
+ {% endif %}
models/qwen-vl/config.json ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2_5_VLForConditionalGeneration"
4
+ ],
5
+ "dtype": "bfloat16",
6
+ "image_token_id": 151655,
7
+ "model_type": "qwen2_5_vl",
8
+ "text_config": {
9
+ "attention_dropout": 0.0,
10
+ "bos_token_id": 151643,
11
+ "dtype": "bfloat16",
12
+ "eos_token_id": 151645,
13
+ "hidden_act": "silu",
14
+ "hidden_size": 2048,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 11008,
17
+ "layer_types": [
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention",
52
+ "full_attention",
53
+ "full_attention"
54
+ ],
55
+ "max_position_embeddings": 128000,
56
+ "max_window_layers": 70,
57
+ "model_type": "qwen2_5_vl_text",
58
+ "num_attention_heads": 16,
59
+ "num_hidden_layers": 36,
60
+ "num_key_value_heads": 2,
61
+ "pad_token_id": null,
62
+ "rms_norm_eps": 1e-06,
63
+ "rope_parameters": {
64
+ "mrope_section": [
65
+ 16,
66
+ 24,
67
+ 24
68
+ ],
69
+ "rope_theta": 1000000.0,
70
+ "rope_type": "default",
71
+ "type": "mrope"
72
+ },
73
+ "sliding_window": null,
74
+ "use_cache": true,
75
+ "use_sliding_window": false,
76
+ "vocab_size": 151936
77
+ },
78
+ "tie_word_embeddings": true,
79
+ "transformers_version": "5.3.0",
80
+ "video_token_id": 151656,
81
+ "vision_config": {
82
+ "depth": 32,
83
+ "dtype": "bfloat16",
84
+ "fullatt_block_indexes": [
85
+ 7,
86
+ 15,
87
+ 23,
88
+ 31
89
+ ],
90
+ "hidden_act": "silu",
91
+ "hidden_size": 1280,
92
+ "in_channels": 3,
93
+ "in_chans": 3,
94
+ "initializer_range": 0.02,
95
+ "intermediate_size": 3420,
96
+ "model_type": "qwen2_5_vl",
97
+ "num_heads": 16,
98
+ "out_hidden_size": 2048,
99
+ "patch_size": 14,
100
+ "spatial_merge_size": 2,
101
+ "spatial_patch_size": 14,
102
+ "temporal_patch_size": 2,
103
+ "tokens_per_second": 2,
104
+ "window_size": 112
105
+ },
106
+ "vision_end_token_id": 151653,
107
+ "vision_start_token_id": 151652,
108
+ "vision_token_id": 151654
109
+ }
models/qwen-vl/generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 151643,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "repetition_penalty": 1.05,
10
+ "temperature": 1e-06,
11
+ "transformers_version": "5.3.0"
12
+ }
models/qwen-vl/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b5b5c5088f4e72eb82dc11890c38596bda270628ed17bc42dcd699928add753
3
+ size 7509338080
models/qwen-vl/processor_config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "Qwen2VLImageProcessorFast",
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": 12845056,
25
+ "shortest_edge": 3136
26
+ },
27
+ "temporal_patch_size": 2
28
+ },
29
+ "processor_class": "Qwen2_5_VLProcessor",
30
+ "video_processor": {
31
+ "data_format": "channels_first",
32
+ "default_to_square": true,
33
+ "do_convert_rgb": true,
34
+ "do_normalize": true,
35
+ "do_rescale": true,
36
+ "do_resize": true,
37
+ "do_sample_frames": false,
38
+ "image_mean": [
39
+ 0.48145466,
40
+ 0.4578275,
41
+ 0.40821073
42
+ ],
43
+ "image_processor_type": "Qwen2VLImageProcessor",
44
+ "image_std": [
45
+ 0.26862954,
46
+ 0.26130258,
47
+ 0.27577711
48
+ ],
49
+ "max_frames": 768,
50
+ "merge_size": 2,
51
+ "min_frames": 4,
52
+ "patch_size": 14,
53
+ "resample": 3,
54
+ "rescale_factor": 0.00392156862745098,
55
+ "return_metadata": false,
56
+ "size": {
57
+ "longest_edge": 12845056,
58
+ "shortest_edge": 3136
59
+ },
60
+ "temporal_patch_size": 2,
61
+ "video_processor_type": "Qwen2VLVideoProcessor"
62
+ }
63
+ }
models/qwen-vl/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
3
+ size 11421892
models/qwen-vl/tokenizer_config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|im_start|>",
10
+ "<|im_end|>",
11
+ "<|object_ref_start|>",
12
+ "<|object_ref_end|>",
13
+ "<|box_start|>",
14
+ "<|box_end|>",
15
+ "<|quad_start|>",
16
+ "<|quad_end|>",
17
+ "<|vision_start|>",
18
+ "<|vision_end|>",
19
+ "<|vision_pad|>",
20
+ "<|image_pad|>",
21
+ "<|video_pad|>"
22
+ ],
23
+ "is_local": false,
24
+ "model_max_length": 131072,
25
+ "pad_token": "<|endoftext|>",
26
+ "processor_class": "Qwen2_5_VLProcessor",
27
+ "split_special_tokens": false,
28
+ "tokenizer_class": "Qwen2Tokenizer",
29
+ "unk_token": null
30
+ }