| """ | |
| SALMONN Example Usage | |
| This script demonstrates how to use the SALMONN inference API. | |
| """ | |
| import sys | |
| sys.path.insert(0, '..') | |
| from inference import SALMONNInference | |
| def main(): | |
| # Initialize model | |
| print("Loading SALMONN model...") | |
| model = SALMONNInference(config_path="../config.yaml") | |
| model.load() | |
| # Example audio file | |
| audio_file = "sample_audio.wav" | |
| print("\n" + "="*50) | |
| print("SALMONN Example Usage") | |
| print("="*50) | |
| # 1. Transcribe audio | |
| print("\n1. Transcription:") | |
| text = model.transcribe(audio_file) | |
| print(f" {text}") | |
| # 2. Ask questions | |
| print("\n2. Question Answering:") | |
| questions = [ | |
| "What language is being spoken?", | |
| "What is the tone of the speaker?", | |
| "Is this audio about weather?", | |
| ] | |
| for q in questions: | |
| answer = model.chat(audio_file, q) | |
| print(f" Q: {q}") | |
| print(f" A: {answer}") | |
| print() | |
| # 3. Describe audio | |
| print("3. Audio Description:") | |
| description = model.describe(audio_file) | |
| print(f" {description}") | |
| print("\n" + "="*50) | |
| print("Done!") | |
| if __name__ == "__main__": | |
| main() | |