Mirror of mlboydaisuke/GLM-OCR-CoreAI

.gitattributes

@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+decoder/glm_ocr_decode_int8hu_s1.aimodel/main.mlirb filter=lfs diff=lfs merge=lfs -text
+vision/glm_ocr_vision.aimodel/main.mlirb filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,74 @@
+---
+license: mit
+library_name: coreai
+pipeline_tag: image-to-text
+tags:
+  - core-ai
+  - on-device
+  - apple
+  - ocr
+  - document-understanding
+  - glm
+base_model: zai-org/GLM-OCR
+---
+
+> **Mirror** of [`mlboydaisuke/GLM-OCR-CoreAI`](https://huggingface.co/mlboydaisuke/GLM-OCR-CoreAI) — the canonical repo ([CoreAI Model Zoo](https://github.com/john-rocky/coreai-model-zoo)). Updates land there first.
+
+
+# GLM-OCR → Core AI
+
+On-device document OCR, running entirely on Apple's **Core AI** (Neural Engine / GPU).
+A port of [`zai-org/GLM-OCR`](https://huggingface.co/zai-org/GLM-OCR) (0.9B, **MIT**) — a small,
+SOTA-quality document recognizer (OmniDocBench v1.5 **94.62**, #1 with its layout pipeline).
+Prompt it with `Text Recognition:` / `Table Recognition:` / `Formula Recognition:` and get back
+plain text (reading order), HTML tables (`<table>…`), or LaTeX. zh / en / fr / es / ru / de / **ja** / ko.
+
+GLM-OCR is a small OCR variant of **GLM-4.V** (`Glm4v`): a CogViT vision tower + a 16-layer GLM text
+decoder with sectioned 3D M-RoPE. This port reuses the shipped Qwen3-VL vision idiom and GLM text
+decode — no R-SWA, no MoE, no MLA.
+
+## Bundles
+
+| dir | what | precision | size |
+|---|---|---|---|
+| `vision/` | CogViT encoder → `image_embeds [N, 1536]` | fp16 | 829 MB |
+| `decoder/` | GLM text decoder, S=1 pipelined, M-RoPE + image injection | int8hu (body int8 per-block-32 + untied head absmax) | 764 MB |
+| `tokenizer/` | `tokenizer.json` etc. | — | — |
+
+The decoder rides three static graph inputs — `image_embeds [682,1536]` f16, `rope_shift_start [1]`,
+`rope_shift_amount [1]` — so the vision tower runs once, its output is injected at the image
+placeholder positions (`V + slot`, row-major over the merged grid), and the text decodes on top.
+`N` (visual-token count) is fixed at export by the chosen input resolution (here 682 = a 22×31 merged
+grid); resize the page to that grid host-side.
+
+## Verified (M4 Max, GPU, Core AI pipelined engine)
+
+- **End-to-end real generation on the engine: 40/40 tokens identical to the fp32 HF reference** — a
+  synthetic document read verbatim (*"Quarterly Report / On-device inference shipped across all
+  product lines this quarter…"*), **~375 tok/s** decode.
+- Torch ladder vs HF: decoder logits cos **1.000020**, vision `image_embeds` cos **1.000061**,
+  full-VLM argmax **694/694**.
+- Engine gate: vision `image_embeds` cos **0.9998**; decoder argmax exact over the sampled positions.
+- int8hu vs fp16: **7 / 694** argmax flips, all at visual-token positions (0 in the text region), the
+  generation-driving position exact — i.e. the OCR text is preserved.
+
+## Run it
+
+The decoder is a standard Core AI pipelined LLM bundle with three multimodal static inputs. Drive it
+with the pipelined engine (S=1, `COREAI_CHUNK_THRESHOLD=1`; feed the prompt with the image
+placeholders rewritten to `V+slot`, bind `image_embeds` from the vision tower, set
+`rope_shift_start = img_start + N`, `rope_shift_amount = N − max(gh, gw)`). The full conversion recipe
+and the host contract (with the exact static-input values) are in the
+[Core AI model zoo](https://github.com/john-rocky/coreai-model-zoo) —
+`conversion/export_glm_ocr_pipelined.py`, `zoo/glm-ocr.md`, `knowledge/glm-ocr-port.md`.
+
+## Scope / honesty
+
+- This is the **recognition** model: per-prompt text / table / formula. Whole-page auto-structuring
+  (the 94.62 full-pipeline number) additionally needs a layout detector (PP-DocLayoutV3) that is not
+  part of this port.
+- int4 is not shipped (weight-only int4 without QAT risks a quality cliff on a 0.9B model).
+
+## License
+
+**MIT** (inherited from `zai-org/GLM-OCR`). *Community port — not affiliated with Apple or Z.ai.*

decoder/glm_ocr_decode_int8hu_s1.aimodel/main.hash

@@ -0,0 +1 @@
+����˼�������~	�3q��a��d�OXr�

decoder/glm_ocr_decode_int8hu_s1.aimodel/main.mlirb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9615c20bc0e9cbbcfd91f4968ffffb7e09f43371dffc619a1ce764a44f5872b7
+size 801205551

decoder/glm_ocr_decode_int8hu_s1.aimodel/metadata.json

@@ -0,0 +1,3 @@
+{
+  "assetVersion" : "2.0"
+}

decoder/metadata.json

@@ -0,0 +1,28 @@
+{
+  "metadata_version": "0.2",
+  "kind": "llm",
+  "name": "glm_ocr_decode_int8hu_s1",
+  "assets": {
+    "main": "glm_ocr_decode_int8hu_s1.aimodel"
+  },
+  "language": {
+    "tokenizer": "tokenizer",
+    "vocab_size": 59392,
+    "max_context_length": 4096,
+    "embedded_tokenizer": true,
+    "function_map": {
+      "main": [
+        "main"
+      ]
+    }
+  },
+  "source": {
+    "model_definition": "torch",
+    "hf_model_id": "mlboydaisuke/GLM-OCR-CoreAI"
+  },
+  "compression": null,
+  "compilation": {
+    "date": "2026-07-04T04:02:09.674171+00:00",
+    "targets": []
+  }
+}

decoder/tokenizer/chat_template.jinja

@@ -0,0 +1,140 @@
+[gMASK]<sop>
+{%- if tools -%}
+<|system|>
+# Tools
+
+You may call one or more functions to assist with the user query.
+
+You are provided with function signatures within <tools></tools> XML tags:
+<tools>
+{% for tool in tools %}
+{{ tool | tojson(ensure_ascii=False) }}
+{% endfor %}
+</tools>
+
+For each function call, output the function name and arguments within the following XML format:
+<tool_call>{function-name}
+<arg_key>{arg-key-1}</arg_key>
+<arg_value>{arg-value-1}</arg_value>
+<arg_key>{arg-key-2}</arg_key>
+<arg_value>{arg-value-2}</arg_value>
+...
+</tool_call>{%- endif -%}
+{%- macro visible_text(content) -%}
+    {%- if content is string -%}
+        {{- content }}
+    {%- elif content is iterable and content is not mapping -%}
+        {%- for item in content -%}
+            {%- if item is mapping and item.type == 'text' -%}
+                {{- item.text }}
+            {%- elif item is mapping and (item.type == 'image' or 'image' in item) -%}
+                <|begin_of_image|><|image|><|end_of_image|>
+            {%- elif item is mapping and (item.type == 'video' or 'video' in item) -%}
+                <|begin_of_video|><|video|><|end_of_video|>
+            {%- elif item is string -%}
+                {{- item }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{- content }}
+    {%- endif -%}
+{%- endmacro -%}
+{%- set ns = namespace(last_user_index=-1) %}
+{%- for m in messages %}
+    {%- if m.role == 'user' %}
+        {% set ns.last_user_index = loop.index0 -%}
+    {%- endif %}
+{%- endfor %}
+{% for m in messages %}
+{%- if m.role == 'user' -%}<|user|>
+{% if m.content is string %}
+{{ m.content }}
+{%- else %}
+{%- for item in m.content %}
+{% if item.type == 'video' or 'video' in item %}
+<|begin_of_video|><|video|><|end_of_video|>{% elif item.type == 'image' or 'image' in item %}
+<|begin_of_image|><|image|><|end_of_image|>{% elif item.type == 'text' %}
+{{ item.text }}
+{%- endif %}
+{%- endfor %}
+{%- endif %}
+{{- '/nothink' if (enable_thinking is defined and not enable_thinking and not visible_text(m.content).endswith("/nothink")) else '' -}}
+{%- elif m.role == 'assistant' -%}
+<|assistant|>
+{%- set reasoning_content = '' %}
+{%- set content = visible_text(m.content) %}
+{%- if m.reasoning_content is string %}
+    {%- set reasoning_content = m.reasoning_content %}
+{%- else %}
+    {%- if '</think>' in content %}
+        {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+        {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+    {%- endif %}
+{%- endif %}
+{%- if loop.index0 > ns.last_user_index and reasoning_content -%}
+{{ '\n<think>' + reasoning_content.strip() +  '</think>'}}
+{%- else -%}
+{{ '\n<think></think>' }}
+{%- endif -%}
+{%- if content.strip() -%}
+{{ '\n' + content.strip() }}
+{%- endif -%}
+{% if m.tool_calls %}
+{% for tc in m.tool_calls %}
+{%- if tc.function %}
+    {%- set tc = tc.function %}
+{%- endif %}
+{{ '\n<tool_call>' + tc.name }}
+{% set _args = tc.arguments %}
+{% for k, v in _args.items() %}
+<arg_key>{{ k }}</arg_key>
+<arg_value>{{ v | tojson(ensure_ascii=False) if v is not string else v }}</arg_value>
+{% endfor %}
+</tool_call>{% endfor %}
+{% endif %}
+{%- elif m.role == 'tool' -%}
+{%- if m.content is string -%}
+{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+    {{- '<|observation|>' }}
+{%- endif %}
+{{- '\n<tool_response>\n' }}
+{{- m.content }}
+{{- '\n</tool_response>' }}
+{% elif m.content is iterable and m.content is not mapping %}
+{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+{{- '<|observation|>' }}
+{%- endif %}
+{{- '\n<tool_response>\n' }}
+{%- for tr in m.content -%}
+  {%- if tr is mapping and tr.type is defined -%}
+    {%- set t = tr.type | lower -%}
+    {%- if t == 'text' and tr.text is defined -%}
+{{ tr.text }}
+    {%- elif t in ['image', 'image_url'] -%}
+<|begin_of_image|><|image|><|end_of_image|>
+    {%- elif t in ['video', 'video_url'] -%}
+<|begin_of_video|><|video|><|end_of_video|>
+    {%- else -%}
+{{ tr | tojson(ensure_ascii=False) }}
+    {%- endif -%}
+  {%- else -%}
+{{ tr.output if tr.output is defined else tr }}
+  {%- endif -%}
+{%- endfor -%}
+{{- '\n</tool_response>' }}
+{%- else -%}
+<|observation|>{% for tr in m.content %}
+
+<tool_response>
+{{ tr.output if tr.output is defined else tr }}
+</tool_response>{% endfor -%}
+{% endif -%}
+{%- elif m.role == 'system' -%}
+<|system|>
+{{ visible_text(m.content) }}
+{%- endif -%}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+<|assistant|>
+{{'<think></think>\n' if (enable_thinking is defined and not enable_thinking) else ''}}
+{%- endif -%}

decoder/tokenizer/tokenizer_config.json

@@ -0,0 +1,49 @@
+{
+  "backend": "tokenizers",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "extra_special_tokens": [
+    "<|endoftext|>",
+    "[MASK]",
+    "[gMASK]",
+    "[sMASK]",
+    "<sop>",
+    "<eop>",
+    "<|system|>",
+    "<|user|>",
+    "<|assistant|>",
+    "<|observation|>",
+    "<|begin_of_image|>",
+    "<|end_of_image|>",
+    "<|begin_of_video|>",
+    "<|end_of_video|>",
+    "<|begin_of_audio|>",
+    "<|end_of_audio|>",
+    "<|begin_of_transcription|>",
+    "<|end_of_transcription|>",
+    "<|code_prefix|>",
+    "<|code_middle|>",
+    "<|code_suffix|>",
+    "<think>",
+    "</think>",
+    "<tool_call>",
+    "</tool_call>",
+    "<tool_response>",
+    "</tool_response>",
+    "<arg_key>",
+    "</arg_key>",
+    "<arg_value>",
+    "</arg_value>",
+    "/nothink",
+    "<|begin_of_box|>",
+    "<|end_of_box|>",
+    "<|image|>",
+    "<|video|>"
+  ],
+  "is_local": true,
+  "model_max_length": 655380,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "left",
+  "processor_class": "Glm46VProcessor",
+  "tokenizer_class": "TokenizersBackend"
+}

tokenizer/chat_template.jinja

@@ -0,0 +1,140 @@
+[gMASK]<sop>
+{%- if tools -%}
+<|system|>
+# Tools
+
+You may call one or more functions to assist with the user query.
+
+You are provided with function signatures within <tools></tools> XML tags:
+<tools>
+{% for tool in tools %}
+{{ tool | tojson(ensure_ascii=False) }}
+{% endfor %}
+</tools>
+
+For each function call, output the function name and arguments within the following XML format:
+<tool_call>{function-name}
+<arg_key>{arg-key-1}</arg_key>
+<arg_value>{arg-value-1}</arg_value>
+<arg_key>{arg-key-2}</arg_key>
+<arg_value>{arg-value-2}</arg_value>
+...
+</tool_call>{%- endif -%}
+{%- macro visible_text(content) -%}
+    {%- if content is string -%}
+        {{- content }}
+    {%- elif content is iterable and content is not mapping -%}
+        {%- for item in content -%}
+            {%- if item is mapping and item.type == 'text' -%}
+                {{- item.text }}
+            {%- elif item is mapping and (item.type == 'image' or 'image' in item) -%}
+                <|begin_of_image|><|image|><|end_of_image|>
+            {%- elif item is mapping and (item.type == 'video' or 'video' in item) -%}
+                <|begin_of_video|><|video|><|end_of_video|>
+            {%- elif item is string -%}
+                {{- item }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{- content }}
+    {%- endif -%}
+{%- endmacro -%}
+{%- set ns = namespace(last_user_index=-1) %}
+{%- for m in messages %}
+    {%- if m.role == 'user' %}
+        {% set ns.last_user_index = loop.index0 -%}
+    {%- endif %}
+{%- endfor %}
+{% for m in messages %}
+{%- if m.role == 'user' -%}<|user|>
+{% if m.content is string %}
+{{ m.content }}
+{%- else %}
+{%- for item in m.content %}
+{% if item.type == 'video' or 'video' in item %}
+<|begin_of_video|><|video|><|end_of_video|>{% elif item.type == 'image' or 'image' in item %}
+<|begin_of_image|><|image|><|end_of_image|>{% elif item.type == 'text' %}
+{{ item.text }}
+{%- endif %}
+{%- endfor %}
+{%- endif %}
+{{- '/nothink' if (enable_thinking is defined and not enable_thinking and not visible_text(m.content).endswith("/nothink")) else '' -}}
+{%- elif m.role == 'assistant' -%}
+<|assistant|>
+{%- set reasoning_content = '' %}
+{%- set content = visible_text(m.content) %}
+{%- if m.reasoning_content is string %}
+    {%- set reasoning_content = m.reasoning_content %}
+{%- else %}
+    {%- if '</think>' in content %}
+        {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+        {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+    {%- endif %}
+{%- endif %}
+{%- if loop.index0 > ns.last_user_index and reasoning_content -%}
+{{ '\n<think>' + reasoning_content.strip() +  '</think>'}}
+{%- else -%}
+{{ '\n<think></think>' }}
+{%- endif -%}
+{%- if content.strip() -%}
+{{ '\n' + content.strip() }}
+{%- endif -%}
+{% if m.tool_calls %}
+{% for tc in m.tool_calls %}
+{%- if tc.function %}
+    {%- set tc = tc.function %}
+{%- endif %}
+{{ '\n<tool_call>' + tc.name }}
+{% set _args = tc.arguments %}
+{% for k, v in _args.items() %}
+<arg_key>{{ k }}</arg_key>
+<arg_value>{{ v | tojson(ensure_ascii=False) if v is not string else v }}</arg_value>
+{% endfor %}
+</tool_call>{% endfor %}
+{% endif %}
+{%- elif m.role == 'tool' -%}
+{%- if m.content is string -%}
+{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+    {{- '<|observation|>' }}
+{%- endif %}
+{{- '\n<tool_response>\n' }}
+{{- m.content }}
+{{- '\n</tool_response>' }}
+{% elif m.content is iterable and m.content is not mapping %}
+{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+{{- '<|observation|>' }}
+{%- endif %}
+{{- '\n<tool_response>\n' }}
+{%- for tr in m.content -%}
+  {%- if tr is mapping and tr.type is defined -%}
+    {%- set t = tr.type | lower -%}
+    {%- if t == 'text' and tr.text is defined -%}
+{{ tr.text }}
+    {%- elif t in ['image', 'image_url'] -%}
+<|begin_of_image|><|image|><|end_of_image|>
+    {%- elif t in ['video', 'video_url'] -%}
+<|begin_of_video|><|video|><|end_of_video|>
+    {%- else -%}
+{{ tr | tojson(ensure_ascii=False) }}
+    {%- endif -%}
+  {%- else -%}
+{{ tr.output if tr.output is defined else tr }}
+  {%- endif -%}
+{%- endfor -%}
+{{- '\n</tool_response>' }}
+{%- else -%}
+<|observation|>{% for tr in m.content %}
+
+<tool_response>
+{{ tr.output if tr.output is defined else tr }}
+</tool_response>{% endfor -%}
+{% endif -%}
+{%- elif m.role == 'system' -%}
+<|system|>
+{{ visible_text(m.content) }}
+{%- endif -%}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+<|assistant|>
+{{'<think></think>\n' if (enable_thinking is defined and not enable_thinking) else ''}}
+{%- endif -%}

tokenizer/tokenizer_config.json

@@ -0,0 +1,49 @@
+{
+  "backend": "tokenizers",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "extra_special_tokens": [
+    "<|endoftext|>",
+    "[MASK]",
+    "[gMASK]",
+    "[sMASK]",
+    "<sop>",
+    "<eop>",
+    "<|system|>",
+    "<|user|>",
+    "<|assistant|>",
+    "<|observation|>",
+    "<|begin_of_image|>",
+    "<|end_of_image|>",
+    "<|begin_of_video|>",
+    "<|end_of_video|>",
+    "<|begin_of_audio|>",
+    "<|end_of_audio|>",
+    "<|begin_of_transcription|>",
+    "<|end_of_transcription|>",
+    "<|code_prefix|>",
+    "<|code_middle|>",
+    "<|code_suffix|>",
+    "<think>",
+    "</think>",
+    "<tool_call>",
+    "</tool_call>",
+    "<tool_response>",
+    "</tool_response>",
+    "<arg_key>",
+    "</arg_key>",
+    "<arg_value>",
+    "</arg_value>",
+    "/nothink",
+    "<|begin_of_box|>",
+    "<|end_of_box|>",
+    "<|image|>",
+    "<|video|>"
+  ],
+  "is_local": true,
+  "model_max_length": 655380,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "left",
+  "processor_class": "Glm46VProcessor",
+  "tokenizer_class": "TokenizersBackend"
+}

vision/glm_ocr_vision.aimodel/main.hash

@@ -0,0 +1 @@
+�=v4�2K]m�cH�z�A��.����>�

vision/glm_ocr_vision.aimodel/main.mlirb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a63d76349a324b5d6df1a263488b7a96027f411603b4d72e9119cdc1833e05ab
+size 869139083

vision/glm_ocr_vision.aimodel/metadata.json

@@ -0,0 +1,3 @@
+{
+  "assetVersion" : "2.0"
+}