File size: 9,278 Bytes
72f90b1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 |
"""
Language Detection Backend
This module provides the main LanguageDetector class and ModelRegistry
for managing multiple language detection models organized by architecture and dataset.
Model Architecture:
- Model A: XLM-RoBERTa based architectures
- Model B: BERT based architectures
Training Datasets:
- Dataset A: Standard multilingual language detection dataset
- Dataset B: Enhanced/specialized language detection dataset
"""
import logging
from typing import Dict, List, Any
from .models import (
BaseLanguageModel,
ModelADatasetA,
ModelBDatasetA,
ModelADatasetB,
ModelBDatasetB,
get_all_model_configs,
get_language_name,
LANGUAGE_MAPPINGS
)
class ModelRegistry:
"""
Registry for managing available language detection models.
This class handles the registration and creation of language detection models
organized by model architecture (A: XLM-RoBERTa, B: BERT) and training
dataset (A: standard, B: enhanced).
"""
def __init__(self):
"""Initialize the model registry with available models."""
# Get model configurations from centralized config
self.model_configs = get_all_model_configs()
# Map model keys to their implementation classes
self.model_classes = {
"model-a-dataset-a": ModelADatasetA, # XLM-RoBERTa + Dataset A
"model-b-dataset-a": ModelBDatasetA, # BERT + Dataset A
"model-a-dataset-b": ModelADatasetB, # XLM-RoBERTa + Dataset B
"model-b-dataset-b": ModelBDatasetB, # BERT + Dataset B
}
# Build models registry by combining configs with classes
self.models = {}
# Add the new organized models
for model_key, config in self.model_configs.items():
if model_key in self.model_classes:
self.models[model_key] = {
"class": self.model_classes[model_key],
"display_name": config["display_name"],
"description": config["description"],
"status": config["status"]
}
def get_available_models(self) -> Dict[str, Dict[str, Any]]:
"""
Get all registered models.
Returns:
Dict containing all model information
"""
return self.models.copy()
def create_model(self, model_key: str) -> BaseLanguageModel:
"""
Create an instance of the specified model.
Args:
model_key (str): Key of the model to create
Returns:
BaseLanguageModel: Instance of the requested model
Raises:
ValueError: If the model key is not found
"""
if model_key not in self.models:
available_keys = list(self.models.keys())
raise ValueError(f"Unknown model: {model_key}. Available models: {available_keys}")
model_class = self.models[model_key]["class"]
return model_class()
class LanguageDetector:
"""
Main language detection class that orchestrates model predictions.
This class provides a unified interface for language detection using
different model architectures and training datasets. It handles model
switching and provides consistent output formatting.
"""
def __init__(self, model_key: str = "model-a-dataset-a"):
"""
Initialize the language detector.
Args:
model_key (str): Key of the model to use from the registry
- "model-a-dataset-a": XLM-RoBERTa + standard dataset
- "model-b-dataset-a": BERT + standard dataset
- "model-a-dataset-b": XLM-RoBERTa + enhanced dataset
- "model-b-dataset-b": BERT + enhanced dataset
"""
self.registry = ModelRegistry()
self.current_model_key = model_key
self.model = self.registry.create_model(model_key)
# Use centralized language mappings
self.language_names = LANGUAGE_MAPPINGS
def switch_model(self, model_key: str):
"""
Switch to a different model.
Args:
model_key (str): Key of the new model to use
Raises:
Exception: If model switching fails
"""
try:
self.model = self.registry.create_model(model_key)
self.current_model_key = model_key
logging.info(f"Successfully switched to model: {model_key}")
except Exception as e:
logging.error(f"Failed to switch to model {model_key}: {e}")
raise
def get_current_model_info(self) -> Dict[str, Any]:
"""
Get information about the currently selected model.
Returns:
Dict containing current model information
"""
return self.model.get_model_info()
def get_available_models(self) -> Dict[str, Dict[str, Any]]:
"""
Get all available models.
Returns:
Dict containing all available models
"""
return self.registry.get_available_models()
def detect_language(self, text: str) -> Dict[str, Any]:
"""
Detect the language of the input text.
Args:
text (str): Input text to analyze
Returns:
Dict containing:
- language: Main predicted language name
- language_code: Main predicted language code
- confidence: Confidence score for main prediction
- top_predictions: List of top 5 predictions with details
- metadata: Additional information about the prediction
Raises:
ValueError: If input text is empty
RuntimeError: If model prediction fails
"""
if not text or not text.strip():
raise ValueError("Input text cannot be empty")
# Get predictions from the current model
model_result = self.model.predict(text.strip())
predictions = model_result['predictions']
if not predictions:
raise RuntimeError("Model returned no predictions")
# Extract main prediction
top_prediction = predictions[0]
main_language_code = top_prediction['language_code']
main_confidence = top_prediction['confidence']
# Get human-readable language name using centralized function
main_language_name = get_language_name(main_language_code)
# Format top predictions (limit to 5)
top_predictions = []
for pred in predictions[:5]:
lang_code = pred['language_code']
lang_name = get_language_name(lang_code)
top_predictions.append({
'language': lang_name,
'language_code': lang_code,
'confidence': pred['confidence']
})
# Prepare metadata
metadata = {
'text_length': model_result.get('text_length', len(text)),
'model_name': model_result.get('model_version', 'unknown'),
'model_type': model_result.get('model_type', 'unknown'),
'current_model_key': self.current_model_key,
'model_info': self.get_current_model_info()
}
return {
'language': main_language_name,
'language_code': main_language_code,
'confidence': main_confidence,
'top_predictions': top_predictions,
'metadata': metadata
}
def get_supported_languages(self) -> Dict[str, str]:
"""
Get dictionary of supported language codes and names.
Returns:
Dict mapping language codes to language names
"""
supported_codes = self.model.get_supported_languages()
return {
code: get_language_name(code)
for code in supported_codes
}
# Example usage and testing
if __name__ == "__main__":
# Initialize detector with default model (Model A Dataset A)
detector = LanguageDetector()
# Test with sample texts
test_texts = [
"Hello, how are you today?",
"Bonjour, comment allez-vous?",
"Hola, ¿cómo estás?",
"Guten Tag, wie geht es Ihnen?"
]
print("Language Detection Test - Model A Dataset A")
print("=" * 60)
for text in test_texts:
try:
result = detector.detect_language(text)
print(f"Text: {text}")
print(f"Detected: {result['language']} ({result['language_code']}) - {result['confidence']:.3f}")
print("---")
except Exception as e:
print(f"Error detecting language for '{text}': {e}")
print("---")
# Show available models
print("\nAvailable Models:")
models = detector.get_available_models()
for key, info in models.items():
status = "✅" if info["status"] == "available" else "🚧"
print(f"{status} {info['display_name']} ({key}): {info['description']}") |