| import os |
| import sys |
| import argparse |
| import time |
| import torch |
|
|
| project_root = os.path.abspath(os.path.join(os.path.dirname(__file__), "..")) |
| sys.path.insert(0, project_root) |
|
|
| from arthaml.inference import Translator |
|
|
| def main(): |
| parser = argparse.ArgumentParser() |
| parser.add_argument('text', nargs='?', type=str, help='English text to translate') |
| parser.add_argument('--interactive', action='store_true', help='Run in interactive mode') |
| args = parser.parse_args() |
| |
| device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') |
| |
| model_path = os.path.join(project_root, 'models/best_model.pth') |
| vocab_en_path = os.path.join(project_root, 'data/processed/vocab_en.json') |
| vocab_kn_path = os.path.join(project_root, 'data/processed/vocab_kn.json') |
| |
| if not os.path.exists(model_path): |
| print(f"Error: Could not find model at {model_path}") |
| return |
| |
| print("Loading Translator...") |
| start_load = time.time() |
| translator = Translator(model_path, vocab_en_path, vocab_kn_path, device) |
| print(f"Loaded in {(time.time() - start_load)*1000:.0f}ms") |
| |
| if args.interactive: |
| print("Interactive mode started. Enter 'q' to quit.") |
| while True: |
| try: |
| text = input("Enter English text (q to quit): ") |
| if text.lower().strip() == 'q': |
| break |
| |
| start = time.time() |
| translation = translator.translate(text) |
| elapsed = (time.time() - start) * 1000 |
| print(f"Kannada translation: {translation}") |
| print(f"Time taken: {elapsed:.2f} ms") |
| except (KeyboardInterrupt, EOFError): |
| break |
| elif args.text: |
| start = time.time() |
| translation = translator.translate(args.text) |
| elapsed = (time.time() - start) * 1000 |
| print(f"Kannada translation: {translation}") |
| print(f"Time taken: {elapsed:.2f} ms") |
| else: |
| parser.print_help() |
|
|
| if __name__ == "__main__": |
| main() |
|
|
