ngxson HF Staff commited on
Commit
c065523
·
verified ·
1 Parent(s): 48666fa

tiny test slice

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
LICENSE ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ Apache License
3
+ Version 2.0, January 2004
4
+ http://www.apache.org/licenses/
5
+
6
+ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
7
+
8
+ 1. Definitions.
9
+
10
+ "License" shall mean the terms and conditions for use, reproduction,
11
+ and distribution as defined by Sections 1 through 9 of this document.
12
+
13
+ "Licensor" shall mean the copyright owner or entity authorized by
14
+ the copyright owner that is granting the License.
15
+
16
+ "Legal Entity" shall mean the union of the acting entity and all
17
+ other entities that control, are controlled by, or are under common
18
+ control with that entity. For the purposes of this definition,
19
+ "control" means (i) the power, direct or indirect, to cause the
20
+ direction or management of such entity, whether by contract or
21
+ otherwise, or (ii) ownership of fifty percent (50%) or more of the
22
+ outstanding shares, or (iii) beneficial ownership of such entity.
23
+
24
+ "You" (or "Your") shall mean an individual or Legal Entity
25
+ exercising permissions granted by this License.
26
+
27
+ "Source" form shall mean the preferred form for making modifications,
28
+ including but not limited to software source code, documentation
29
+ source, and configuration files.
30
+
31
+ "Object" form shall mean any form resulting from mechanical
32
+ transformation or translation of a Source form, including but
33
+ not limited to compiled object code, generated documentation,
34
+ and conversions to other media types.
35
+
36
+ "Work" shall mean the work of authorship, whether in Source or
37
+ Object form, made available under the License, as indicated by a
38
+ copyright notice that is included in or attached to the work
39
+ (an example is provided in the Appendix below).
40
+
41
+ "Derivative Works" shall mean any work, whether in Source or Object
42
+ form, that is based on (or derived from) the Work and for which the
43
+ editorial revisions, annotations, elaborations, or other modifications
44
+ represent, as a whole, an original work of authorship. For the purposes
45
+ of this License, Derivative Works shall not include works that remain
46
+ separable from, or merely link (or bind by name) to the interfaces of,
47
+ the Work and Derivative Works thereof.
48
+
49
+ "Contribution" shall mean any work of authorship, including
50
+ the original version of the Work and any modifications or additions
51
+ to that Work or Derivative Works thereof, that is intentionally
52
+ submitted to Licensor for inclusion in the Work by the copyright owner
53
+ or by an individual or Legal Entity authorized to submit on behalf of
54
+ the copyright owner. For the purposes of this definition, "submitted"
55
+ means any form of electronic, verbal, or written communication sent
56
+ to the Licensor or its representatives, including but not limited to
57
+ communication on electronic mailing lists, source code control systems,
58
+ and issue tracking systems that are managed by, or on behalf of, the
59
+ Licensor for the purpose of discussing and improving the Work, but
60
+ excluding communication that is conspicuously marked or otherwise
61
+ designated in writing by the copyright owner as "Not a Contribution."
62
+
63
+ "Contributor" shall mean Licensor and any individual or Legal Entity
64
+ on behalf of whom a Contribution has been received by Licensor and
65
+ subsequently incorporated within the Work.
66
+
67
+ 2. Grant of Copyright License. Subject to the terms and conditions of
68
+ this License, each Contributor hereby grants to You a perpetual,
69
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
70
+ copyright license to reproduce, prepare Derivative Works of,
71
+ publicly display, publicly perform, sublicense, and distribute the
72
+ Work and such Derivative Works in Source or Object form.
73
+
74
+ 3. Grant of Patent License. Subject to the terms and conditions of
75
+ this License, each Contributor hereby grants to You a perpetual,
76
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
77
+ (except as stated in this section) patent license to make, have made,
78
+ use, offer to sell, sell, import, and otherwise transfer the Work,
79
+ where such license applies only to those patent claims licensable
80
+ by such Contributor that are necessarily infringed by their
81
+ Contribution(s) alone or by combination of their Contribution(s)
82
+ with the Work to which such Contribution(s) was submitted. If You
83
+ institute patent litigation against any entity (including a
84
+ cross-claim or counterclaim in a lawsuit) alleging that the Work
85
+ or a Contribution incorporated within the Work constitutes direct
86
+ or contributory patent infringement, then any patent licenses
87
+ granted to You under this License for that Work shall terminate
88
+ as of the date such litigation is filed.
89
+
90
+ 4. Redistribution. You may reproduce and distribute copies of the
91
+ Work or Derivative Works thereof in any medium, with or without
92
+ modifications, and in Source or Object form, provided that You
93
+ meet the following conditions:
94
+
95
+ (a) You must give any other recipients of the Work or
96
+ Derivative Works a copy of this License; and
97
+
98
+ (b) You must cause any modified files to carry prominent notices
99
+ stating that You changed the files; and
100
+
101
+ (c) You must retain, in the Source form of any Derivative Works
102
+ that You distribute, all copyright, patent, trademark, and
103
+ attribution notices from the Source form of the Work,
104
+ excluding those notices that do not pertain to any part of
105
+ the Derivative Works; and
106
+
107
+ (d) If the Work includes a "NOTICE" text file as part of its
108
+ distribution, then any Derivative Works that You distribute must
109
+ include a readable copy of the attribution notices contained
110
+ within such NOTICE file, excluding those notices that do not
111
+ pertain to any part of the Derivative Works, in at least one
112
+ of the following places: within a NOTICE text file distributed
113
+ as part of the Derivative Works; within the Source form or
114
+ documentation, if provided along with the Derivative Works; or,
115
+ within a display generated by the Derivative Works, if and
116
+ wherever such third-party notices normally appear. The contents
117
+ of the NOTICE file are for informational purposes only and
118
+ do not modify the License. You may add Your own attribution
119
+ notices within Derivative Works that You distribute, alongside
120
+ or as an addendum to the NOTICE text from the Work, provided
121
+ that such additional attribution notices cannot be construed
122
+ as modifying the License.
123
+
124
+ You may add Your own copyright statement to Your modifications and
125
+ may provide additional or different license terms and conditions
126
+ for use, reproduction, or distribution of Your modifications, or
127
+ for any such Derivative Works as a whole, provided Your use,
128
+ reproduction, and distribution of the Work otherwise complies with
129
+ the conditions stated in this License.
130
+
131
+ 5. Submission of Contributions. Unless You explicitly state otherwise,
132
+ any Contribution intentionally submitted for inclusion in the Work
133
+ by You to the Licensor shall be under the terms and conditions of
134
+ this License, without any additional terms or conditions.
135
+ Notwithstanding the above, nothing herein shall supersede or modify
136
+ the terms of any separate license agreement you may have executed
137
+ with Licensor regarding such Contributions.
138
+
139
+ 6. Trademarks. This License does not grant permission to use the trade
140
+ names, trademarks, service marks, or product names of the Licensor,
141
+ except as required for reasonable and customary use in describing the
142
+ origin of the Work and reproducing the content of the NOTICE file.
143
+
144
+ 7. Disclaimer of Warranty. Unless required by applicable law or
145
+ agreed to in writing, Licensor provides the Work (and each
146
+ Contributor provides its Contributions) on an "AS IS" BASIS,
147
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
148
+ implied, including, without limitation, any warranties or conditions
149
+ of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
150
+ PARTICULAR PURPOSE. You are solely responsible for determining the
151
+ appropriateness of using or redistributing the Work and assume any
152
+ risks associated with Your exercise of permissions under this License.
153
+
154
+ 8. Limitation of Liability. In no event and under no legal theory,
155
+ whether in tort (including negligence), contract, or otherwise,
156
+ unless required by applicable law (such as deliberate and grossly
157
+ negligent acts) or agreed to in writing, shall any Contributor be
158
+ liable to You for damages, including any direct, indirect, special,
159
+ incidental, or consequential damages of any character arising as a
160
+ result of this License or out of the use or inability to use the
161
+ Work (including but not limited to damages for loss of goodwill,
162
+ work stoppage, computer failure or malfunction, or any and all
163
+ other commercial damages or losses), even if such Contributor
164
+ has been advised of the possibility of such damages.
165
+
166
+ 9. Accepting Warranty or Additional Liability. While redistributing
167
+ the Work or Derivative Works thereof, You may choose to offer,
168
+ and charge a fee for, acceptance of support, warranty, indemnity,
169
+ or other liability obligations and/or rights consistent with this
170
+ License. However, in accepting such obligations, You may act only
171
+ on Your own behalf and on Your sole responsibility, not on behalf
172
+ of any other Contributor, and only if You agree to indemnify,
173
+ defend, and hold each Contributor harmless for any liability
174
+ incurred by, or claims asserted against, such Contributor by reason
175
+ of your accepting any such warranty or additional liability.
176
+
177
+ END OF TERMS AND CONDITIONS
178
+
179
+ APPENDIX: How to apply the Apache License to your work.
180
+
181
+ To apply the Apache License to your work, attach the following
182
+ boilerplate notice, with the fields enclosed by brackets "[]"
183
+ replaced with your own identifying information. (Don't include
184
+ the brackets!) The text should be enclosed in the appropriate
185
+ comment syntax for the file format. We also recommend that a
186
+ file or class name and description of purpose be included on the
187
+ same "printed page" as the copyright notice for easier
188
+ identification within third-party archives.
189
+
190
+ Copyright [yyyy] [name of copyright owner]
191
+
192
+ Licensed under the Apache License, Version 2.0 (the "License");
193
+ you may not use this file except in compliance with the License.
194
+ You may obtain a copy of the License at
195
+
196
+ http://www.apache.org/licenses/LICENSE-2.0
197
+
198
+ Unless required by applicable law or agreed to in writing, software
199
+ distributed under the License is distributed on an "AS IS" BASIS,
200
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
201
+ See the License for the specific language governing permissions and
202
+ limitations under the License.
README.md ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: openbmb/MiniCPM-OCR-Jev
4
+ tags:
5
+ - testing
6
+ ---
7
+
8
+ # tiny-MiniCPM-OCR-Jev-for-testing
9
+
10
+ > [!WARNING]
11
+ > This is a **modified, damaged copy** of [openbmb/MiniCPM-OCR-Jev](https://huggingface.co/openbmb/MiniCPM-OCR-Jev), made only to test software. Its output is gibberish. Do not use it for anything else.
12
+
13
+ It keeps the architecture, tokenizer, chat template, image processor config and 2-class score head of MiniCPM-OCR-Jev, so it runs the same code paths as the real model in llama.cpp (`openbmb-minicpm-ocr-jev` decision type: Qwen3.5 trunk with linear and full attention, class scores read at the last token, 2 images per request), but it is small enough for CI.
14
+
15
+ GGUF version: [ggml-org-dev/tiny-MiniCPM-OCR-Jev-for-testing-GGUF](https://huggingface.co/ggml-org-dev/tiny-MiniCPM-OCR-Jev-for-testing-GGUF)
16
+
17
+ ## Changes from the original model
18
+
19
+ Sliced from the safetensors with [slice_minicpm_ocr_jev.py](slice_minicpm_ocr_jev.py), which follows the `slice-model` skill of llama.cpp:
20
+
21
+ - Text model: 4 of 24 layers (0-3: 3 linear attention + 1 full attention), hidden size 2048 -> 32, FFN 6144 -> 128, heads 8 -> 2, KV heads 2 -> 1, linear attention key and value heads 16 -> 2 (head dims kept)
22
+ - Score head (`score.weight`, [inconsistent, consistent]): hidden size 2048 -> 32
23
+ - Vision encoder: 2 of 24 blocks (0, 1), hidden size 1024 -> 128, heads 16 -> 2, FFN 4096 -> 256, output size 2048 -> 32
24
+ - `config.json` declares `MiniCPMOCRForSequenceClassification` instead of `Qwen3_5ForConditionalGeneration`, so that `convert_hf_to_gguf.py` converts it as the sequence classifier
25
+ - The original repo has no LICENSE file, the Apache License 2.0 text is added
26
+
27
+ ## License
28
+
29
+ Derived from MiniCPM-OCR-Jev by OpenBMB, under the [Apache License 2.0](LICENSE).
chat_template.jinja ADDED
@@ -0,0 +1,154 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- set image_count = namespace(value=0) %}
2
+ {%- set video_count = namespace(value=0) %}
3
+ {%- macro render_content(content, do_vision_count, is_system_content=false) %}
4
+ {%- if content is string %}
5
+ {{- content }}
6
+ {%- elif content is iterable and content is not mapping %}
7
+ {%- for item in content %}
8
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
9
+ {%- if is_system_content %}
10
+ {{- raise_exception('System message cannot contain images.') }}
11
+ {%- endif %}
12
+ {%- if do_vision_count %}
13
+ {%- set image_count.value = image_count.value + 1 %}
14
+ {%- endif %}
15
+ {%- if add_vision_id %}
16
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
17
+ {%- endif %}
18
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
19
+ {%- elif 'video' in item or item.type == 'video' %}
20
+ {%- if is_system_content %}
21
+ {{- raise_exception('System message cannot contain videos.') }}
22
+ {%- endif %}
23
+ {%- if do_vision_count %}
24
+ {%- set video_count.value = video_count.value + 1 %}
25
+ {%- endif %}
26
+ {%- if add_vision_id %}
27
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
28
+ {%- endif %}
29
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
30
+ {%- elif 'text' in item %}
31
+ {{- item.text }}
32
+ {%- else %}
33
+ {{- raise_exception('Unexpected item type in content.') }}
34
+ {%- endif %}
35
+ {%- endfor %}
36
+ {%- elif content is none or content is undefined %}
37
+ {{- '' }}
38
+ {%- else %}
39
+ {{- raise_exception('Unexpected content type.') }}
40
+ {%- endif %}
41
+ {%- endmacro %}
42
+ {%- if not messages %}
43
+ {{- raise_exception('No messages provided.') }}
44
+ {%- endif %}
45
+ {%- if tools and tools is iterable and tools is not mapping %}
46
+ {{- '<|im_start|>system\n' }}
47
+ {{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
48
+ {%- for tool in tools %}
49
+ {{- "\n" }}
50
+ {{- tool | tojson }}
51
+ {%- endfor %}
52
+ {{- "\n</tools>" }}
53
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
54
+ {%- if messages[0].role == 'system' %}
55
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
56
+ {%- if content %}
57
+ {{- '\n\n' + content }}
58
+ {%- endif %}
59
+ {%- endif %}
60
+ {{- '<|im_end|>\n' }}
61
+ {%- else %}
62
+ {%- if messages[0].role == 'system' %}
63
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
64
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
65
+ {%- endif %}
66
+ {%- endif %}
67
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
68
+ {%- for message in messages[::-1] %}
69
+ {%- set index = (messages|length - 1) - loop.index0 %}
70
+ {%- if ns.multi_step_tool and message.role == "user" %}
71
+ {%- set content = render_content(message.content, false)|trim %}
72
+ {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
73
+ {%- set ns.multi_step_tool = false %}
74
+ {%- set ns.last_query_index = index %}
75
+ {%- endif %}
76
+ {%- endif %}
77
+ {%- endfor %}
78
+ {%- if ns.multi_step_tool %}
79
+ {{- raise_exception('No user query found in messages.') }}
80
+ {%- endif %}
81
+ {%- for message in messages %}
82
+ {%- set content = render_content(message.content, true)|trim %}
83
+ {%- if message.role == "system" %}
84
+ {%- if not loop.first %}
85
+ {{- raise_exception('System message must be at the beginning.') }}
86
+ {%- endif %}
87
+ {%- elif message.role == "user" %}
88
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
89
+ {%- elif message.role == "assistant" %}
90
+ {%- set reasoning_content = '' %}
91
+ {%- if message.reasoning_content is string %}
92
+ {%- set reasoning_content = message.reasoning_content %}
93
+ {%- else %}
94
+ {%- if '</think>' in content %}
95
+ {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
96
+ {%- set content = content.split('</think>')[-1].lstrip('\n') %}
97
+ {%- endif %}
98
+ {%- endif %}
99
+ {%- set reasoning_content = reasoning_content|trim %}
100
+ {%- if loop.index0 > ns.last_query_index %}
101
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
102
+ {%- else %}
103
+ {{- '<|im_start|>' + message.role + '\n' + content }}
104
+ {%- endif %}
105
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
106
+ {%- for tool_call in message.tool_calls %}
107
+ {%- if tool_call.function is defined %}
108
+ {%- set tool_call = tool_call.function %}
109
+ {%- endif %}
110
+ {%- if loop.first %}
111
+ {%- if content|trim %}
112
+ {{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
113
+ {%- else %}
114
+ {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
115
+ {%- endif %}
116
+ {%- else %}
117
+ {{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
118
+ {%- endif %}
119
+ {%- if tool_call.arguments is defined %}
120
+ {%- for args_name, args_value in tool_call.arguments|items %}
121
+ {{- '<parameter=' + args_name + '>\n' }}
122
+ {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
123
+ {{- args_value }}
124
+ {{- '\n</parameter>\n' }}
125
+ {%- endfor %}
126
+ {%- endif %}
127
+ {{- '</function>\n</tool_call>' }}
128
+ {%- endfor %}
129
+ {%- endif %}
130
+ {{- '<|im_end|>\n' }}
131
+ {%- elif message.role == "tool" %}
132
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
133
+ {{- '<|im_start|>user' }}
134
+ {%- endif %}
135
+ {{- '\n<tool_response>\n' }}
136
+ {{- content }}
137
+ {{- '\n</tool_response>' }}
138
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
139
+ {{- '<|im_end|>\n' }}
140
+ {%- elif loop.last %}
141
+ {{- '<|im_end|>\n' }}
142
+ {%- endif %}
143
+ {%- else %}
144
+ {{- raise_exception('Unexpected message role.') }}
145
+ {%- endif %}
146
+ {%- endfor %}
147
+ {%- if add_generation_prompt %}
148
+ {{- '<|im_start|>assistant\n' }}
149
+ {%- if enable_thinking is defined and enable_thinking is true %}
150
+ {{- '<think>\n' }}
151
+ {%- else %}
152
+ {{- '<think>\n\n</think>\n\n' }}
153
+ {%- endif %}
154
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,86 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "MiniCPMOCRForSequenceClassification"
4
+ ],
5
+ "dtype": "bfloat16",
6
+ "eos_token_id": 248046,
7
+ "hidden_size": 32,
8
+ "image_token_id": 248056,
9
+ "model_type": "qwen3_5",
10
+ "pad_token_id": 248044,
11
+ "problem_type": "single_label_classification",
12
+ "text_config": {
13
+ "attention_bias": false,
14
+ "attention_dropout": 0.0,
15
+ "attn_output_gate": true,
16
+ "bos_token_id": null,
17
+ "dtype": "bfloat16",
18
+ "eos_token_id": 248044,
19
+ "full_attention_interval": 4,
20
+ "head_dim": 256,
21
+ "hidden_act": "silu",
22
+ "hidden_size": 32,
23
+ "initializer_range": 0.02,
24
+ "intermediate_size": 128,
25
+ "layer_types": [
26
+ "linear_attention",
27
+ "linear_attention",
28
+ "linear_attention",
29
+ "full_attention"
30
+ ],
31
+ "linear_conv_kernel_dim": 4,
32
+ "linear_key_head_dim": 128,
33
+ "linear_num_key_heads": 2,
34
+ "linear_num_value_heads": 2,
35
+ "linear_value_head_dim": 128,
36
+ "mamba_ssm_dtype": "float32",
37
+ "max_position_embeddings": 262144,
38
+ "mlp_only_layers": [],
39
+ "model_type": "qwen3_5_text",
40
+ "mtp_num_hidden_layers": 1,
41
+ "mtp_use_dedicated_embeddings": false,
42
+ "num_attention_heads": 2,
43
+ "num_hidden_layers": 4,
44
+ "num_key_value_heads": 1,
45
+ "pad_token_id": 248044,
46
+ "partial_rotary_factor": 0.25,
47
+ "rms_norm_eps": 1e-06,
48
+ "rope_parameters": {
49
+ "mrope_interleaved": true,
50
+ "mrope_section": [
51
+ 11,
52
+ 11,
53
+ 10
54
+ ],
55
+ "partial_rotary_factor": 0.25,
56
+ "rope_theta": 10000000,
57
+ "rope_type": "default"
58
+ },
59
+ "tie_word_embeddings": false,
60
+ "use_cache": false,
61
+ "vocab_size": 248320
62
+ },
63
+ "tie_word_embeddings": false,
64
+ "transformers_version": "5.10.4",
65
+ "use_cache": false,
66
+ "video_token_id": 248057,
67
+ "vision_config": {
68
+ "deepstack_visual_indexes": [],
69
+ "depth": 2,
70
+ "dtype": "bfloat16",
71
+ "hidden_act": "gelu_pytorch_tanh",
72
+ "hidden_size": 128,
73
+ "in_channels": 3,
74
+ "initializer_range": 0.02,
75
+ "intermediate_size": 256,
76
+ "model_type": "qwen3_5_vision",
77
+ "num_heads": 2,
78
+ "num_position_embeddings": 2304,
79
+ "out_hidden_size": 32,
80
+ "patch_size": 16,
81
+ "spatial_merge_size": 2,
82
+ "temporal_patch_size": 2
83
+ },
84
+ "vision_end_token_id": 248054,
85
+ "vision_start_token_id": 248053
86
+ }
generation_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "eos_token_id": [
4
+ 248046,
5
+ 248044,
6
+ 248046
7
+ ],
8
+ "pad_token_id": 248044,
9
+ "transformers_version": "5.10.4",
10
+ "use_cache": true
11
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:988f5365d7e17e8bfb0e6750eca4c2301d282fcac3d7b123641759bcac261438
3
+ size 18471648
preprocessor_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "size": {
3
+ "longest_edge": 16777216,
4
+ "shortest_edge": 65536
5
+ },
6
+ "patch_size": 16,
7
+ "temporal_patch_size": 2,
8
+ "merge_size": 2,
9
+ "image_mean": [
10
+ 0.5,
11
+ 0.5,
12
+ 0.5
13
+ ],
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "processor_class": "Qwen3VLProcessor",
20
+ "image_processor_type": "Qwen2VLImageProcessorFast"
21
+ }
processor_config.json ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ 0.5,
9
+ 0.5,
10
+ 0.5
11
+ ],
12
+ "image_processor_type": "Qwen2VLImageProcessor",
13
+ "image_std": [
14
+ 0.5,
15
+ 0.5,
16
+ 0.5
17
+ ],
18
+ "merge_size": 2,
19
+ "patch_size": 16,
20
+ "resample": 3,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "longest_edge": 16777216,
24
+ "shortest_edge": 65536
25
+ },
26
+ "temporal_patch_size": 2
27
+ },
28
+ "processor_class": "Qwen3VLProcessor",
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
+ 0.5,
38
+ 0.5,
39
+ 0.5
40
+ ],
41
+ "image_std": [
42
+ 0.5,
43
+ 0.5,
44
+ 0.5
45
+ ],
46
+ "max_frames": 768,
47
+ "merge_size": 2,
48
+ "min_frames": 4,
49
+ "patch_size": 16,
50
+ "resample": 3,
51
+ "rescale_factor": 0.00392156862745098,
52
+ "return_metadata": false,
53
+ "size": {
54
+ "longest_edge": 25165824,
55
+ "shortest_edge": 4096
56
+ },
57
+ "temporal_patch_size": 2,
58
+ "video_processor_type": "Qwen3VLVideoProcessor"
59
+ }
60
+ }
slice_minicpm_ocr_jev.py ADDED
@@ -0,0 +1,213 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ #!/usr/bin/env python3
2
+ """
3
+ Build a tiny slice of openbmb/MiniCPM-OCR-Jev (openbmb-minicpm-ocr-jev decision model) for testing. Output is gibberish, only the shapes and code paths are real.
4
+
5
+ Adapted from the slice-model skill of llama.cpp (skills/slice-model): tensors are fetched with HTTP range requests and config.json is
6
+ patched to match, then convert_hf_to_gguf.py does the conversion. The stock script only drops layers, which cannot go below the
7
+ 248320 x 2048 token embedding, so this one also narrows the hidden sizes of the trunk, of the vision encoder and of the score head.
8
+
9
+ The config.json of the upstream repo declares Qwen3_5ForConditionalGeneration, the slice declares MiniCPMOCRForSequenceClassification
10
+ so that convert_hf_to_gguf.py picks the sequence classifier (the upstream repo has no LICENSE file, it is Apache-2.0 per its model card).
11
+
12
+ usage: slice_minicpm_ocr_jev.py <output dir>
13
+ """
14
+ import json, re, struct, sys
15
+ from collections import defaultdict
16
+ from concurrent.futures import ThreadPoolExecutor
17
+ from pathlib import Path
18
+
19
+ import requests
20
+ import torch
21
+ from huggingface_hub import hf_hub_url
22
+ from safetensors.torch import save_file
23
+
24
+ REPO = 'openbmb/MiniCPM-OCR-Jev'
25
+ AUX = ['tokenizer.json', 'tokenizer_config.json', 'chat_template.jinja', 'processor_config.json', 'preprocessor_config.json', 'generation_config.json']
26
+
27
+ # text: hidden, ffn, attention heads, kv heads (head_dim stays 256), linear attention key / value heads (head dims stay 128)
28
+ E, FF, H, KV, LK, LV = 32, 128, 2, 1, 2, 2 # upstream has as many key heads as value heads
29
+ HD, LKD, LVD = 256, 128, 128
30
+ TEXT_LAYERS = [0, 1, 2, 3] # 3 linear attention + 1 full attention
31
+ # vision: hidden, ffn, heads (head_dim stays 64)
32
+ V, VFF, VH = 128, 256, 2
33
+ VISION_LAYERS = [0, 1]
34
+ MERGE = 4 # 2 x 2 patches are merged
35
+
36
+ DTYPES = {'BF16': torch.bfloat16, 'F32': torch.float32, 'F16': torch.float16}
37
+
38
+
39
+ def read_header(url):
40
+ def rng(a, b):
41
+ r = requests.get(url, headers={'Range': f'bytes={a}-{b}'}, timeout=120)
42
+ r.raise_for_status()
43
+ return r.content
44
+ n = struct.unpack('<Q', rng(0, 7))[0]
45
+ h = json.loads(rng(8, 8 + n - 1))
46
+ h.pop('__metadata__', None)
47
+ return h, 8 + n
48
+
49
+
50
+ def fetch(url, meta, base):
51
+ a, b = meta['data_offsets']
52
+ if a == b:
53
+ return torch.zeros(meta['shape'], dtype=DTYPES[meta['dtype']])
54
+ r = requests.get(url, headers={'Range': f'bytes={base + a}-{base + b - 1}'}, timeout=600)
55
+ r.raise_for_status()
56
+ assert len(r.content) == b - a
57
+ return torch.frombuffer(bytearray(r.content), dtype=DTYPES[meta['dtype']]).reshape(meta['shape'])
58
+
59
+
60
+ def renumber(name, prefix, keep):
61
+ m = re.match(re.escape(prefix) + r'\.(\d+)\.(.*)', name)
62
+ if not m:
63
+ return name
64
+ i = int(m[1])
65
+ return f'{prefix}.{keep.index(i)}.{m[2]}' if i in keep else None
66
+
67
+
68
+ def rename_main(n):
69
+ for prefix, keep in (('model.language_model.layers', TEXT_LAYERS), ('model.visual.blocks', VISION_LAYERS)):
70
+ if n.startswith(prefix + '.'):
71
+ return renumber(n, prefix, keep)
72
+ if n.startswith('mtp.'):
73
+ return None
74
+ return n
75
+
76
+
77
+ def linear_qkv_rows(t, nk, nv):
78
+ """in_proj_qkv / conv1d: q (nk * dk), k (nk * dk), v (nv * dv) stacked on dim 0, v heads grouped by key head."""
79
+ q, k, v = torch.split(t, [nk * LKD, nk * LKD, nv * LVD], dim=0)
80
+ return torch.cat([q[:LK * LKD], k[:LK * LKD], v[:LV * LVD]], dim=0)
81
+
82
+
83
+ def narrow_text_layer(key, t, nk, nv):
84
+ if key.endswith(('input_layernorm.weight', 'post_attention_layernorm.weight')):
85
+ return t[:E]
86
+ if key.endswith(('mlp.gate_proj.weight', 'mlp.up_proj.weight')):
87
+ return t[:FF, :E]
88
+ if key.endswith('mlp.down_proj.weight'):
89
+ return t[:E, :FF]
90
+ # full attention, q_proj has the output gate: [n_head, 2 * head_dim] per head
91
+ if key.endswith('self_attn.q_proj.weight'):
92
+ return t[:H * 2 * HD, :E]
93
+ if key.endswith(('self_attn.k_proj.weight', 'self_attn.v_proj.weight')):
94
+ return t[:KV * HD, :E]
95
+ if key.endswith('self_attn.o_proj.weight'):
96
+ return t[:E, :H * HD]
97
+ if key.endswith(('self_attn.q_norm.weight', 'self_attn.k_norm.weight', 'linear_attn.norm.weight')):
98
+ return t
99
+ # linear attention (gated delta net)
100
+ if key.endswith('linear_attn.in_proj_qkv.weight'):
101
+ return linear_qkv_rows(t, nk, nv)[:, :E]
102
+ if key.endswith('linear_attn.conv1d.weight'):
103
+ return linear_qkv_rows(t, nk, nv)
104
+ if key.endswith('linear_attn.in_proj_z.weight'):
105
+ return t[:LV * LVD, :E]
106
+ if key.endswith(('linear_attn.in_proj_a.weight', 'linear_attn.in_proj_b.weight')):
107
+ return t[:LV, :E]
108
+ if key.endswith(('linear_attn.A_log', 'linear_attn.dt_bias')):
109
+ return t[:LV]
110
+ if key.endswith('linear_attn.out_proj.weight'):
111
+ return t[:E, :LV * LVD]
112
+ raise SystemExit(f'unhandled tensor {key}')
113
+
114
+
115
+ def narrow_vision_block(key, t, vo):
116
+ if key.endswith('attn.qkv.weight'):
117
+ return t.reshape(3, vo, vo)[:, :V, :V].reshape(3 * V, V)
118
+ if key.endswith('attn.qkv.bias'):
119
+ return t.reshape(3, vo)[:, :V].reshape(-1)
120
+ if key.endswith('attn.proj.weight'):
121
+ return t[:V, :V]
122
+ if key.endswith('mlp.linear_fc1.weight'):
123
+ return t[:VFF, :V]
124
+ if key.endswith('mlp.linear_fc1.bias'):
125
+ return t[:VFF]
126
+ if key.endswith('mlp.linear_fc2.weight'):
127
+ return t[:V, :VFF]
128
+ if key.endswith(('attn.proj.bias', 'mlp.linear_fc2.bias', 'norm1.weight', 'norm1.bias', 'norm2.weight', 'norm2.bias')):
129
+ return t[:V]
130
+ raise SystemExit(f'unhandled tensor {key}')
131
+
132
+
133
+ def make_narrow(text_cfg, vision_cfg):
134
+ nk, nv = text_cfg['linear_num_key_heads'], text_cfg['linear_num_value_heads']
135
+ vo = vision_cfg['hidden_size']
136
+
137
+ def narrow(name, t):
138
+ if name.startswith('model.language_model.layers.'):
139
+ return narrow_text_layer(name, t, nk, nv)
140
+ if name.startswith('model.visual.blocks.'):
141
+ return narrow_vision_block(name, t, vo)
142
+ table = {
143
+ 'model.language_model.embed_tokens.weight': lambda: t[:, :E],
144
+ 'score.weight': lambda: t[:, :E], # [inconsistent, consistent] x hidden
145
+ 'model.language_model.norm.weight': lambda: t[:E],
146
+ 'model.visual.patch_embed.proj.weight': lambda: t[:V],
147
+ 'model.visual.patch_embed.proj.bias': lambda: t[:V],
148
+ 'model.visual.pos_embed.weight': lambda: t[:, :V],
149
+ 'model.visual.merger.norm.weight': lambda: t[:V],
150
+ 'model.visual.merger.norm.bias': lambda: t[:V],
151
+ # the input of linear_fc1 is MERGE patches of the vision hidden size, its output has the same size
152
+ 'model.visual.merger.linear_fc1.weight': lambda: t[:MERGE * V].reshape(MERGE * V, MERGE, vo)[:, :, :V].reshape(MERGE * V, MERGE * V),
153
+ 'model.visual.merger.linear_fc1.bias': lambda: t[:MERGE * V],
154
+ 'model.visual.merger.linear_fc2.weight': lambda: t[:E, :MERGE * V],
155
+ 'model.visual.merger.linear_fc2.bias': lambda: t[:E],
156
+ }
157
+ if name in table:
158
+ return table[name]()
159
+ raise SystemExit(f'unhandled tensor {name}')
160
+ return narrow
161
+
162
+
163
+ def slice_files(filenames, out_path, rename, transform):
164
+ jobs = []
165
+ n_total = 0
166
+ for filename in filenames:
167
+ url = hf_hub_url(REPO, filename)
168
+ header, base = read_header(url)
169
+ n_total += len(header)
170
+ jobs += [(url, base, n, m, rename(n)) for n, m in header.items()]
171
+ jobs = [j for j in jobs if j[4] is not None]
172
+
173
+ def one(job):
174
+ url, base, src, meta, dst = job
175
+ return dst, transform(src, fetch(url, meta, base)).contiguous().clone()
176
+
177
+ with ThreadPoolExecutor(8) as ex:
178
+ tensors = dict(ex.map(one, jobs))
179
+ out_path.parent.mkdir(parents=True, exist_ok=True)
180
+ save_file(tensors, out_path, metadata={'format': 'pt'})
181
+ print(f'{filenames}: {len(tensors)}/{n_total} tensors, {out_path.stat().st_size / 1e6:.1f} MB')
182
+
183
+
184
+ def get_json(f):
185
+ r = requests.get(hf_hub_url(REPO, f), timeout=120)
186
+ r.raise_for_status()
187
+ return r.json()
188
+
189
+
190
+ def main():
191
+ out = Path(sys.argv[1])
192
+ out.mkdir(parents=True, exist_ok=True)
193
+ cfg = get_json('config.json')
194
+ slice_files(['model.safetensors'], out / 'model.safetensors', rename_main, make_narrow(cfg['text_config'], cfg['vision_config']))
195
+
196
+ for f in AUX:
197
+ r = requests.get(hf_hub_url(REPO, f), timeout=120)
198
+ r.raise_for_status()
199
+ (out / f).write_bytes(r.content)
200
+
201
+ cfg['architectures'] = ['MiniCPMOCRForSequenceClassification']
202
+ cfg['hidden_size'] = E
203
+ t = cfg['text_config']
204
+ t.update(hidden_size=E, intermediate_size=FF, num_attention_heads=H, num_key_value_heads=KV,
205
+ linear_num_key_heads=LK, linear_num_value_heads=LV,
206
+ num_hidden_layers=len(TEXT_LAYERS), layer_types=[t['layer_types'][i] for i in TEXT_LAYERS])
207
+ v = cfg['vision_config']
208
+ v.update(depth=len(VISION_LAYERS), hidden_size=V, intermediate_size=VFF, num_heads=VH, out_hidden_size=E,
209
+ deepstack_visual_indexes=[i for i in v['deepstack_visual_indexes'] if i in VISION_LAYERS])
210
+ (out / 'config.json').write_text(json.dumps(cfg, indent=2) + '\n')
211
+
212
+
213
+ main()
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
3
+ size 19989325
tokenizer_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "audio_bos_token": "<|audio_start|>",
4
+ "audio_eos_token": "<|audio_end|>",
5
+ "audio_token": "<|audio_pad|>",
6
+ "backend": "tokenizers",
7
+ "bos_token": null,
8
+ "clean_up_tokenization_spaces": false,
9
+ "eos_token": "<|im_end|>",
10
+ "errors": "replace",
11
+ "image_token": "<|image_pad|>",
12
+ "is_local": true,
13
+ "local_files_only": false,
14
+ "model_max_length": 262144,
15
+ "model_specific_special_tokens": {
16
+ "audio_bos_token": "<|audio_start|>",
17
+ "audio_eos_token": "<|audio_end|>",
18
+ "audio_token": "<|audio_pad|>",
19
+ "image_token": "<|image_pad|>",
20
+ "video_token": "<|video_pad|>",
21
+ "vision_bos_token": "<|vision_start|>",
22
+ "vision_eos_token": "<|vision_end|>"
23
+ },
24
+ "pad_token": "<|endoftext|>",
25
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
26
+ "processor_class": "Qwen3VLProcessor",
27
+ "split_special_tokens": false,
28
+ "tokenizer_class": "Qwen2Tokenizer",
29
+ "unk_token": null,
30
+ "video_token": "<|video_pad|>",
31
+ "vision_bos_token": "<|vision_start|>",
32
+ "vision_eos_token": "<|vision_end|>"
33
+ }