config.example.json

{
  "description": "Extract2MD Configuration Schema - Complete example with all available options",
  
  "ocr": {
    "language": "eng",
    "oem": 1,
    "psm": 6,
    "workerPath": "./tesseract-worker.min.js",
    "corePath": "./tesseract-core.wasm.js", 
    "langPath": "./lang-data/",
    "options": {
      "logger": null,
      "errorHandler": null
    }
  },
  
  "webllm": {
    "modelId": "Llama-3.2-1B-Instruct-q4f16_1-MLC",
    "temperature": 0.7,
    "maxTokens": 4000,
    "streamingEnabled": false,
    "customModel": {
      "model": "https://huggingface.co/mlc-ai/custom-model/resolve/main/",
      "model_id": "Custom-Model-ID",
      "model_lib": "https://example.com/path/to/custom-model.wasm",
      "required_features": ["shader-f16"],
      "overrides": {
        "conv_template": "llama"
      }
    }
  },
  
  "systemPrompts": {
    "singleExtraction": "Focus on technical accuracy and preserve all code examples exactly as they appear.",
    "combinedExtraction": "Pay special attention to diagrams and tables that might be better captured in the OCR version."
  },
  
  "processing": {
    "splitPascalCase": false,
    "pdfRenderScale": 2.5,
    "postProcessRules": [
      {
        "find": "\\bAPI\\b",
        "replace": "API"
      },
      {
        "find": "\\bJSON\\b", 
        "replace": "JSON"
      }
    ]
  }
}