Improve model card: add transformers library, pipeline tag, and paper link

README.md

@@ -1,12 +1,14 @@
 ---
+base_model:
+- tencent/Youtu-LLM-2B
 license: other
 license_name: youtu-parsing
 license_link: https://huggingface.co/tencent/Youtu-Parsing/blob/main/LICENSE.txt
-pipeline_tag: image-text-to-text
-base_model:
-  - tencent/Youtu-LLM-2B
+pipeline_tag: image-segmentation
+library_name: transformers
 base_model_relation: finetune
 ---
+
 <div align="center">
 
 # <img src="assets/youtu-parsing-logo.png" alt="Youtu-Parsing Logo" height="100px">
@@ -22,7 +24,9 @@ base_model_relation: finetune
 
 ## 🎯 Introduction
 
-**Youtu-Parsing** is a specialized document parsing model built upon the open-source Youtu-LLM 2B foundation. By extending the capabilities of the base model with a prompt-guided framework and NaViT-style dynamic visual encoder, Youtu-Parsing offers enhanced parsing capabilities for diverse document elements including text, tables, formulas, and charts. The model incorporates an efficient parallel decoding mechanism that significantly accelerates inference, making it practical for real-world document analysis applications. We share Youtu-Parsing with the community to facilitate research and development in document understanding.
+**Youtu-Parsing** is a specialized document parsing model built upon the open-source Youtu-LLM 2B foundation, as presented in the paper [Youtu-VL: Unleashing Visual Potential via Unified Vision-Language Supervision](https://huggingface.co/papers/2601.19798). 
+
+By extending the capabilities of the base model with a prompt-guided framework and NaViT-style dynamic visual encoder, Youtu-Parsing offers enhanced parsing capabilities for diverse document elements including text, tables, formulas, and charts. The model incorporates an efficient parallel decoding mechanism that significantly accelerates inference, making it practical for real-world document analysis applications. We share Youtu-Parsing with the community to facilitate research and development in document understanding.
 
 
 ## ✨ Key Features
@@ -62,33 +66,70 @@ base_model_relation: finetune
 <a id="quickstart"></a>
 
 ## 🚀 Quick Start
-### Install packages
-```bash
-conda create -n youtu_parsing python=3.10
-conda activate youtu_parsing
-pip install git+https://github.com/TencentCloudADP/youtu-parsing.git#subdirectory=youtu_hf_parser
-
-# install the flash-attn2
-# For CUDA 12.x + PyTorch 2.6 + Python 3.10 + Linux x86_64:
-pip install https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiFALSE-cp310-cp310-linux_x86_64.whl
 
-# Alternative: Install from PyPI
-pip install flash-attn==2.7.0
+### Installation
+Ensure your Python environment has the `transformers` library installed:
+```bash
+pip install "transformers>=4.56.0,<=4.57.1" torch accelerate pillow torchvision opencv-python-headless
 ```
 
-### Usage with transformers
+### Usage with Transformers
+You can interact with the model using the `transformers` library:
+
 ```python
-from youtu_hf_parser import YoutuOCRParserHF
+from transformers import AutoProcessor, AutoModelForCausalLM
+import torch
+
+model = AutoModelForCausalLM.from_pretrained(
+    "tencent/Youtu-VL-4B-Instruct", 
+    attn_implementation="flash_attention_2", 
+    torch_dtype="auto", 
+    device_map="cuda", 
+    trust_remote_code=True
+).eval()
+
+processor = AutoProcessor.from_pretrained(
+    "tencent/Youtu-VL-4B-Instruct", 
+    use_fast=True, 
+    trust_remote_code=True
+)
 
-# Initialize the parser
-parser = YoutuOCRParserHF(
-    model_path=model_path,
-    enable_angle_correct=True,  # Set to False to disable angle correction
-    angle_correct_model_path=angle_correct_model_path
+img_path = "path/to/your/image.png"
+messages = [
+    {
+        "role": "user",
+        "content": [
+            {"type": "image", "image": img_path},
+            {"type": "text",  "text": "Describe the image"},
+        ],
+    }
+]
+
+inputs = processor.apply_chat_template(
+    messages,
+    tokenize=True,
+    add_generation_prompt=True,
+    return_dict=True,
+    return_tensors="pt"
+).to(model.device)
+
+generated_ids = model.generate(
+    **inputs,
+    temperature=0.1,
+    top_p=0.001,
+    repetition_penalty=1.05,
+    do_sample=True,
+    max_new_tokens=32768,
 )
 
-# Parse an image
-parser.parse_file(input_path=image_path, output_dir=output_dir)
+generated_ids_trimmed = [
+    out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
+]
+outputs = processor.batch_decode(
+    generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
+)
+print(f"Youtu-VL output:
+{outputs[0]}")
 ```
 
 ## 🎨 Visualization
@@ -138,16 +179,6 @@ We would like to thank [Youtu-LLM](https://github.com/TencentCloudADP/youtu-tip/
 
 If you find our work useful in your research, please consider citing the following paper:
 ```
-@article{youtu-parsing,
-  title={Youtu-Parsing: Perception, Structuring and Recognition via High-Parallelism Decoding},
-  author={Tencent Youtu Lab},
-  year={2026},
-  eprint={},
-  archivePrefix={},
-  primaryClass={},
-  url={}, 
-}
-
 @article{youtu-vl,
   title={Youtu-VL: Unleashing Visual Potential via Unified Vision-Language Supervision},
   author={Tencent Youtu Lab},
@@ -158,6 +189,16 @@ If you find our work useful in your research, please consider citing the followi
   url={https://arxiv.org/abs/2601.19798}, 
 }
 
+@article{youtu-parsing,
+  title={Youtu-Parsing: Perception, Structuring and Recognition via High-Parallelism Decoding},
+  author={Tencent Youtu Lab},
+  year={2026},
+  eprint={},
+  archivePrefix={},
+  primaryClass={},
+  url={}, 
+}
+
 @article{youtu-llm,
   title={Youtu-LLM: Unlocking the Native Agentic Potential for Lightweight Large Language Models},
   author={Tencent Youtu Lab},
@@ -167,4 +208,4 @@ If you find our work useful in your research, please consider citing the followi
   primaryClass={cs.CL},
   url={https://arxiv.org/abs/2512.24618}, 
 }
-```
+```