### Quick Start: Entity Extraction

Source: https://github.com/macpaw/gliner2swift/blob/main/README.md

Load a pre-trained model and extract entities from text. The model is automatically downloaded from HuggingFace.

```swift
import GLiNER2Swift

// Load model (downloads automatically from HuggingFace)
let model = try await GLiNER2.fromPretrained("fastino/gliner2-base-v1")

// Extract entities
let text = "Tim Cook is CEO of Apple in Cupertino."
let entities = try model.extractEntities(
    from: text,
    labels: ["person", "company", "location"]
)

for entity in entities {
    print("\(entity.label): \(entity.text) [\(entity.start)-\(entity.end)]")
}
// Output:
// person: Tim Cook [0-8]
// company: Apple [23-28]
// location: Cupertino [32-41]
```

--------------------------------

### Load LoRA Adapter with GLiNER2Swift

Source: https://github.com/macpaw/gliner2swift/blob/main/README.md

Demonstrates two methods for loading a GLiNER2Swift model with a LoRA adapter: a one-step process combining base model and adapter loading, and a two-step process loading the base model first, then attaching the adapter. Ensure the adapter path points to a valid LoRA adapter directory.

```swift
let model = try await GLiNER2.fromPretrained(
    "fastino/gliner2-base-v1",
    adapterPath: "/path/to/adapter"
)
```

```swift
let model = try await GLiNER2.fromPretrained("fastino/gliner2-base-v1")
try model.loadAdapter(from: "/path/to/adapter")
```

--------------------------------

### API: Structured Extraction

Source: https://github.com/macpaw/gliner2swift/blob/main/README.md

Create a schema for structured extraction, defining entities and classification tasks.

```swift
let schema = model.createSchema()
    .entities(["person", "company"])
    .classification(task: "sentiment", labels: ["positive", "negative"])

let result = try model.extract(from: text, schema: schema)
```

--------------------------------

### Add Gliner2Swift to Package.swift

Source: https://github.com/macpaw/gliner2swift/blob/main/README.md

Add the Gliner2Swift package to your project's Package.swift file to manage dependencies.

```swift
dependencies: [
    .package(url: "https://github.com/MacPaw/Gliner2Swift", branch: "main"),
]
```

--------------------------------

### API: Entity Extraction

Source: https://github.com/macpaw/gliner2swift/blob/main/README.md

Extract entities from a given text using specified labels.

```swift
let entities = try model.extractEntities(
    from: "Your text here",
    labels: ["person", "organization", "location"]
)
```

--------------------------------

### API: Text Classification

Source: https://github.com/macpaw/gliner2swift/blob/main/README.md

Classify the sentiment of a given text into predefined categories.

```swift
let classification = try model.classifyText(
    "Great product, highly recommend!",
    labels: ["positive", "negative", "neutral"]
)
```

--------------------------------

### Structured Extraction

Source: https://github.com/macpaw/gliner2swift/blob/main/README.md

Performs structured extraction from text based on a defined schema, which can include entities and text classification tasks.

```APIDOC
## Structured Extraction

### Description
Performs structured extraction from text based on a defined schema, which can include entities and text classification tasks.

### Method Signature
```swift
func extract(from text: String, schema: Schema) async throws -> ExtractionResult
```

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
None

### Request Example
```swift
let schema = model.createSchema()
    .entities(["person", "company"])
    .classification(task: "sentiment", labels: ["positive", "negative"])

let result = try model.extract(from: text, schema: schema)
```

### Response
#### Success Response
A `ExtractionResult` object containing the structured extraction results based on the provided schema.

#### Response Example
```json
{
  "entities": [
    {
      "label": "person",
      "text": "Tim Cook",
      "start": 0,
      "end": 8
    }
  ],
  "classification": {
    "sentiment": {
      "label": "positive",
      "score": 0.95
    }
  }
}
```
```

--------------------------------

### Entity Extraction

Source: https://github.com/macpaw/gliner2swift/blob/main/README.md

Extracts entities from a given text based on a list of labels. This method is useful for Named Entity Recognition (NER) tasks.

```APIDOC
## Entity Extraction

### Description
Extracts entities from a given text based on a list of labels. This method is useful for Named Entity Recognition (NER) tasks.

### Method Signature
```swift
func extractEntities(from text: String, labels: [String]) async throws -> [Entity]
```

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
None

### Request Example
```swift
let entities = try model.extractEntities(
    from: "Your text here",
    labels: ["person", "organization", "location"]
)
```

### Response
#### Success Response
An array of `Entity` objects, where each entity has a `label`, `text`, `start`, and `end` property.

#### Response Example
```json
[
  {
    "label": "person",
    "text": "Tim Cook",
    "start": 0,
    "end": 8
  }
]
```
```

--------------------------------

### Text Classification

Source: https://github.com/macpaw/gliner2swift/blob/main/README.md

Classifies a given text into one of the provided labels. This is useful for tasks like sentiment analysis or topic categorization.

```APIDOC
## Text Classification

### Description
Classifies a given text into one of the provided labels. This is useful for tasks like sentiment analysis or topic categorization.

### Method Signature
```swift
func classifyText(_ text: String, labels: [String]) async throws -> Classification
```

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
None

### Request Example
```swift
let classification = try model.classifyText(
    "Great product, highly recommend!",
    labels: ["positive", "negative", "neutral"]
)
```

### Response
#### Success Response
A `Classification` object containing the predicted label and its confidence score.

#### Response Example
```json
{
  "label": "positive",
  "score": 0.95
}
```
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.