Upload README.md with huggingface_hub

README.md

@@ -2,45 +2,23 @@
 tags:
 - ocr
 - document-processing
-- glm-ocr
+- lighton-ocr-2
 - markdown
 - uv-script
 - generated
-dataset_info:
-  features:
-  - name: image
-    dtype: image
-  - name: text
-    dtype: string
-  - name: lighton2
-    dtype: string
-  - name: inference_info
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 7940742
-    num_examples: 50
-  download_size: 7922609
-  dataset_size: 7940742
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
 ---
 
-# Document OCR using GLM-OCR
+# Document OCR using LightOnOCR-2-1B
 
-This dataset contains OCR results from images in [technocreep/ussr_typewriter](https://huggingface.co/datasets/technocreep/ussr_typewriter) using GLM-OCR, a compact 0.9B OCR model achieving SOTA performance.
+This dataset contains OCR results from images in [technocreep/ussr_typewriter](https://huggingface.co/datasets/technocreep/ussr_typewriter) using LightOnOCR-2, a fast and compact 1B OCR model trained with RLVR.
 
 ## Processing Details
 
 - **Source Dataset**: [technocreep/ussr_typewriter](https://huggingface.co/datasets/technocreep/ussr_typewriter)
-- **Model**: [zai-org/GLM-OCR](https://huggingface.co/zai-org/GLM-OCR)
-- **Task**: text recognition
+- **Model**: [lightonai/LightOnOCR-2-1B](https://huggingface.co/lightonai/LightOnOCR-2-1B)
 - **Number of Samples**: 50
-- **Processing Time**: 1.5 min
-- **Processing Date**: 2026-02-25 11:52 UTC
+- **Processing Time**: 2.0 min
+- **Processing Date**: 2026-02-25 11:55 UTC
 
 ### Configuration
 
@@ -48,37 +26,76 @@ This dataset contains OCR results from images in [technocreep/ussr_typewriter](h
 - **Output Column**: `markdown`
 - **Dataset Split**: `train`
 - **Batch Size**: 16
+- **Target Image Size**: 1540px (longest dimension)
 - **Max Model Length**: 8,192 tokens
-- **Max Output Tokens**: 8,192
-- **Temperature**: 0.01
-- **Top P**: 1e-05
+- **Max Output Tokens**: 4,096
+- **Temperature**: 0.2
+- **Top P**: 0.9
 - **GPU Memory Utilization**: 80.0%
 
 ## Model Information
 
-GLM-OCR is a compact, high-performance OCR model:
-- 0.9B parameters
-- 94.62% on OmniDocBench V1.5
-- CogViT visual encoder + GLM-0.5B language decoder
-- Multi-Token Prediction (MTP) loss for efficiency
-- Multilingual: zh, en, fr, es, ru, de, ja, ko
-- MIT licensed
+LightOnOCR-2 is a next-generation fast, compact OCR model that excels at:
+- ⚡ **Fastest Speed** - 42.8 pages/second on H100 GPU (7× faster than v1)
+- 🎯 **High Accuracy** - 83.2 ± 0.9% on OlmOCR-Bench (+7.1% vs v1)
+- 🧠 **RLVR Training** - Eliminates repetition loops and formatting errors
+- 📚 **Better Dataset** - 2.5× larger training data with cleaner annotations
+- 📐 **LaTeX formulas** - Mathematical notation in LaTeX format
+- 📊 **Tables** - Extracted and formatted as markdown
+- 📝 **Document structure** - Hierarchy and layout preservation
+- 🌍 **Multilingual** - Optimized for European languages
+- 💪 **Production-ready** - Outperforms models 9× larger
+
+### Key Improvements over v1
+
+- **7.5× faster**: 42.8 vs 5.71 pages/sec on H100
+- **+7.1% accuracy**: 83.2% vs 76.1% on benchmarks
+- **Better quality**: RLVR training eliminates common OCR errors
+- **Cleaner output**: No repetition loops or formatting glitches
+- **Simpler**: Single model (no vocabulary variants)
 
 ## Dataset Structure
 
 The dataset contains all original columns plus:
-- `markdown`: The extracted text in markdown format
+- `markdown`: The extracted text in markdown format with LaTeX formulas
 - `inference_info`: JSON list tracking all OCR models applied to this dataset
 
+## Usage
+
+```python
+from datasets import load_dataset
+import json
+
+# Load the dataset
+dataset = load_dataset("{output_dataset_id}", split="train")
+
+# Access the markdown text
+for example in dataset:
+    print(example["markdown"])
+    break
+
+# View all OCR models applied to this dataset
+inference_info = json.loads(dataset[0]["inference_info"])
+for info in inference_info:
+    print(f"Column: {info['column_name']} - Model: {info['model_id']}")
+```
+
 ## Reproduction
 
+This dataset was generated using the [uv-scripts/ocr](https://huggingface.co/datasets/uv-scripts/ocr) LightOnOCR-2 script:
+
 ```bash
-uv run https://huggingface.co/datasets/uv-scripts/ocr/raw/main/glm-ocr.py \
+uv run https://huggingface.co/datasets/uv-scripts/ocr/raw/main/lighton-ocr2.py \
     technocreep/ussr_typewriter \
     <output-dataset> \
     --image-column image \
-    --batch-size 16 \
-    --task ocr
+    --batch-size 16
 ```
 
-Generated with [UV Scripts](https://huggingface.co/uv-scripts)
+## Performance
+
+- **Processing Speed**: ~0.42 images/second
+- **Benchmark Score**: 83.2 ± 0.9% on OlmOCR-Bench
+- **Training**: RLVR (Reinforcement Learning with Verifiable Rewards)
+
+Generated with 🤖 [UV Scripts](https://huggingface.co/uv-scripts)