FEAT: [model] MinerU2.5-2509-1.2B support

doc/source/models/builtin/llm/index.rst

@@ -346,6 +346,11 @@ The following is a list of built-in LLM in Xinference:
      - 32768
      - Marco-o1: Towards Open Reasoning Models for Open-Ended Solutions
 
+   * - :ref:`mineru2.5-2509-1.2b <models_llm_mineru2.5-2509-1.2b>`
+     - chat, vision
+     - 32768
+     - MinerU2.5-2509-1.2B is a vision language model for document understanding.
+
    * - :ref:`minicpm-2b-dpo-bf16 <models_llm_minicpm-2b-dpo-bf16>`
      - chat
      - 4096
@@ -868,6 +873,8 @@ The following is a list of built-in LLM in Xinference:
 
    marco-o1
 
+   mineru2.5-2509-1.2b
+
    minicpm-2b-dpo-bf16
 
    minicpm-2b-dpo-fp16

doc/source/models/builtin/llm/mineru2.5-2509-1.2b.rst

@@ -0,0 +1,31 @@
+.. _models_llm_mineru2.5-2509-1.2b:
+
+========================================
+MinerU2.5-2509-1.2B
+========================================
+
+- **Context Length:** 32768
+- **Model Name:** MinerU2.5-2509-1.2B
+- **Languages:** en, zh
+- **Abilities:** chat, vision
+- **Description:** MinerU2.5-2509-1.2B is a vision language model for document understanding.
+
+Specifications
+^^^^^^^^^^^^^^
+
+
+Model Spec 1 (pytorch, 1_2 Billion)
+++++++++++++++++++++++++++++++++++++++++
+
+- **Model Format:** pytorch
+- **Model Size (in billions):** 1_2
+- **Quantizations:** none
+- **Engines**: Transformers
+- **Model ID:** opendatalab/MinerU2.5-2509-1.2B
+- **Model Hubs**:  `Hugging Face <https://huggingface.co/opendatalab/MinerU2.5-2509-1.2B>`__, `ModelScope <https://modelscope.cn/models/opendatalab/MinerU2.5-2509-1.2B>`__
+
+Execute the following command to launch the model, remember to replace ``${quantization}`` with your
+chosen quantization method from the options listed above::
+
+   xinference launch --model-engine ${engine} --model-name MinerU2.5-2509-1.2B --size-in-billions 1_2 --model-format pytorch --quantization ${quantization}
+

xinference/model/image/ocr/__init__.py

@@ -15,23 +15,20 @@
 from .deepseek_ocr import DeepSeekOCRModel
 from .got_ocr2 import GotOCR2Model
 from .hunyuan_ocr import HunyuanOCRModel
-from .mineru import MinerUModel
 from .mlx import MLXDeepSeekOCRModel
 from .ocr_family import SUPPORTED_ENGINES
 from .paddleocr_vl import PaddleOCRVLModel
 from .vllm import (
     VLLMDeepSeekOCRModel,
     VLLMGotOCR2Model,
     VLLMHunyuanOCRModel,
-    VLLMMinerUModel,
     VLLMPaddleOCRVLModel,
 )
 
 __all__ = [
     "DeepSeekOCRModel",
     "GotOCR2Model",
     "HunyuanOCRModel",
-    "MinerUModel",
     "PaddleOCRVLModel",
 ]
 
@@ -41,12 +38,10 @@ def register_builtin_ocr_engines() -> None:
         DeepSeekOCRModel,
         GotOCR2Model,
         HunyuanOCRModel,
-        MinerUModel,
         PaddleOCRVLModel,
     ]
     SUPPORTED_ENGINES["vllm"] = [
         VLLMDeepSeekOCRModel,
         VLLMHunyuanOCRModel,
-        VLLMMinerUModel,
     ]
     SUPPORTED_ENGINES["mlx"] = [MLXDeepSeekOCRModel]