diff --git a/lyrics_transcriber/transcriber.py b/lyrics_transcriber/transcriber.py index 56af92d..81546d4 100644 --- a/lyrics_transcriber/transcriber.py +++ b/lyrics_transcriber/transcriber.py @@ -29,7 +29,7 @@ def __init__( log_level=logging.DEBUG, log_formatter=None, transcription_model="medium", - llm_model="gpt-3.5-turbo-1106", + llm_model="gpt-4-1106-preview", render_video=False, video_resolution="360p", video_background_image=None, @@ -743,7 +743,7 @@ def transcribe(self): self.outputs["transcription_data_dict"] = json.load(cache_file) return - self.logger.debug(f"no cached transcription file found, running whisper transcribe") + self.logger.debug(f"no cached transcription file found, running whisper transcribe with model: {self.transcription_model}") audio = whisper.load_audio(self.audio_filepath) model = whisper.load_model(self.transcription_model, device="cpu") result = whisper.transcribe(model, audio, language="en") diff --git a/lyrics_transcriber/utils/cli.py b/lyrics_transcriber/utils/cli.py index 3b670ce..ac36e5c 100755 --- a/lyrics_transcriber/utils/cli.py +++ b/lyrics_transcriber/utils/cli.py @@ -66,7 +66,7 @@ def main(): parser.add_argument( "--llm_model", - default="gpt-3.5-turbo-1106", + default="gpt-4-1106-preview", help="Optional: LLM model to use (currently only supports OpenAI chat completion models, e.g. gpt-4-1106-preview). Default: gpt-3.5-turbo-1106", ) diff --git a/pyproject.toml b/pyproject.toml index 73b4c88..75298c6 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -1,6 +1,6 @@ [tool.poetry] name = "lyrics-transcriber" -version = "0.12.3" +version = "0.12.4" description = "Automatically create synchronised lyrics files in ASS and MidiCo LRC formats with word-level timestamps, using Whisper and lyrics from Genius and Spotify" authors = ["Andrew Beveridge "] license = "MIT"