testing

.gitignore

@@ -126,5 +126,5 @@ dmypy.json
 
 
 data/
-secrets.yaml
+*secrets.yaml
 *.csv

README.md

@@ -1 +1 @@
-Simple Transcription tool using Huggigface models.
+Simple Transcription tool using Huggingface models.

amondin/__init__.py

@@ -4,4 +4,4 @@
 from .diarize_speakers import diarize_speakers
 from .speech2text import speech2text
 from .tools import get_secret, convert_audio_to_wav
-from .transcription_tool import transcribe
+from .main import transcribe

amondin/transcription_tool.py → amondin/main.py

@@ -9,12 +9,13 @@
 
 
 def transcribe(
-        file_path: str, hf_token: str, language: str = "german", num_speakers: int = None,
+        input_file_path: str, output_file_path: str, hf_token: str, language: str = "german", num_speakers: int = None,
         s2t_model: str = "openai/whisper-tiny"
 ):
     """
     Transcribe a give audio.wav file.
-    :param file_path:
+    :param output_file_path:
+    :param input_file_path:
     :param hf_token:
     :param language: Set the language for improved performance. None results in language detection.
     :param num_speakers: Set the number of speakers for improved performance. None results in
@@ -24,7 +25,7 @@ def transcribe(
     """
     print("Diarizing speakers...")
     diarized_speakers = diarize_speakers(
-        file_path,
+        input_file_path,
         hf_token=hf_token,
         num_speakers=num_speakers,
     )

setup.py

@@ -13,10 +13,10 @@
         "pyannote.core==5.0.0",
         "pyyaml==6.0.1",
         "ffmpeg-python==0.2.0",
-        "pandas==2.2.2",
+        "pandas==2.2.1",
         "librosa==0.10.1",
         "soundfile==0.12.1",
         "numpy==1.26.4",
-        "torch==2.2.2",
+        "torch==2.2.1",
     ]
 )