Utility Modules

Documentation for utility modules in wimarka/utils/ that provide supporting functionality.

Overview

Utility modules handle cross-cutting concerns:

  • helper.py - File I/O and language tag management

  • logger.py - Logging configuration

  • model.py - Model loading and caching

  • cache.py - Response caching

  • torch.py - PyTorch device management

helper Module

File: wimarka/utils/helper.py

Functions

check_tag

def check_tag(src_lang: str, tgt_lang: str) -> None

Validates language codes.

Raises: ValueError if language codes are invalid

add_tag

def add_tag(sentence: str, lang: str) -> str

Adds language tag to sentence.

Example:

add_tag("Good morning!", "EN")
# Returns: "[EN] Good morning!"

printEvaluationResults

def printEvaluationResults(results: dict) -> None

Prints formatted evaluation results to console.

logger Module

File: wimarka/utils/logger.py

setup_logger

def setup_logger() -> logging.Logger

Configures and returns logger instance.

Configuration:

  • Format: [LEVEL] - message

  • Handler: Console (stdout)

  • Level: INFO

Usage:

from wimarka.utils.logger import setup_logger

logger = setup_logger()
logger.info("Processing...")

model Module

File: wimarka/utils/model.py

Model loading and management utilities.

Key Functions:

  • Load models from HuggingFace Hub

  • Manage model cache

  • Handle model initialization

cache Module

File: wimarka/utils/cache.py

Caching utilities for LLM responses.

torch Module

File: wimarka/utils/torch.py

PyTorch device management.

Content:

import torch

device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

See Also