GLiNER.js is a TypeScript-based inference engine for running GLiNER (Generalist and Lightweight Named Entity Recognition) models. GLiNER can identify any entity type using a bidirectional transformer encoder, offering a practical alternative to traditional NER models and large language models.

📄 Paper   •   📢 Discord   •   🤗 Demo   •   🤗 Available models   •   🧬 Official Repo

• Flexible entity recognition without predefined categories
• Lightweight and fast inference
• Easy integration with web applications
• TypeScript support for better developer experience

npm install gliner

const gliner = new Gliner({
  tokenizerPath: "onnx-community/gliner_small-v2",
  onnxSettings: {
    modelPath: "public/model.onnx",
    executionContext: "web",
    executionProvider: "webgpu",
  },
  maxWidth: 12,
});

await gliner.initialize();

const input_text = "Your input text here";
const texts = [input_text];
const entities = ["city", "country", "person"];
const threshold = 0.1;

const decoded = await gliner.inference({ texts, entities, threshold });
console.log(decoded);

• modelPath: can be either a URL to a local model as in the basic example, or it can also be the Model itself as an array of binary data.
• executionProvider: these are the same providers that ONNX web supports, currently we allow webgpu (recommended), cpu, wasm, webgl but more can be added
• wasmPaths: Path to the wasm binaries, this can be either a URL to the binaries like a CDN url, or a local path to a folder with the binaries.
• multiThread: wether to multithread at all, only relevent for wasm and cpu exeuction providers.
• multiThread: When choosing the wasm or cpu provider, multiThread will allow you to specify the number of cores you want to use.
• fetchBinary: will prefetch the binary from the default or provided wasm paths

To use GLiNER models in a web environment, you need an ONNX format model. You can:

1. Search for pre-converted models on HuggingFace
2. Convert a model yourself using the official Python script

Use the convert_to_onnx.py script with the following arguments:

• model_path: Location of the GLiNER model
• save_path: Where to save the ONNX file
• quantize: Set to True for IntU8 quantization (optional)

Example:

python convert_to_onnx.py --model_path /path/to/your/model --save_path /path/to/save/onnx --quantize True

GLiNER.js offers versatile entity recognition capabilities across various domains:

1. Enhanced Search Query Understanding
2. Real-time PII Detection
3. Intelligent Document Parsing
4. Content Summarization and Insight Extraction
5. Automated Content Tagging and Categorization ...

• Further optimize inference speed
• Add support for token-based GLiNER architecture
• Implement bi-encoder GLiNER architecture for better scalability
• Enable model training capabilities
• Provide more usage examples

• for any changes, remember to run pnpm changeset, otherwise there will not be a version bump and the PR Github Action will fail.

• GLiNER original authors
• ONNX Runtime Web
• Transformers.js

For questions and support, please join our Discord community or open an issue on GitHub.