up

run_demo_multi_models.py

@@ -14,9 +14,8 @@ disable_progress_bar()
 DEFAULT_MODEL_NAME = "bofenghuang/whisper-large-v2-cv11-french"
 # make sure no OOM
 MODEL_NAMES = [
-    # "bofenghuang/whisper-medium-cv11-french",
+    "bofenghuang/whisper-medium-cv11-french",
     "bofenghuang/whisper-large-v2-cv11-french",
-    "bofenghuang/whisper-large-v2-french",
 ]
 CHUNK_LENGTH_S = 30
 # STRIDE_LENGTH_S = 0

run_demo_openai_merged.py

@@ -123,7 +123,7 @@ def transcribe(microphone, file_upload, yt_url, with_timestamps, model_name=DEFA
             "The uploaded audio will be used and the YouTube URL will be discarded.\n"
         )
 
-    elif (microphone is None) and (file_upload is None) or (yt_url is None):
+    elif (microphone is None) and (file_upload is None) and (yt_url is None):
         return "ERROR: You have to either use the microphone, upload an audio file or paste a YouTube URL"
 
     if microphone is not None:
@@ -162,7 +162,7 @@ demo = gr.Interface(
     theme="huggingface",
     title="Whisper French Demo 🇫🇷 : Transcribe Audio",
     description=(
-        "Transcribe long-form microphone or audio inputs with the click of a button! Demo uses the the fine-tuned"
+        "**Transcribe long-form microphone, audio inputs or YouTube videos with the click of a button!** \n\nDemo uses the the fine-tuned"
         f" checkpoint [{DEFAULT_MODEL_NAME}](https://huggingface.co/{DEFAULT_MODEL_NAME}) and 🤗 Transformers to transcribe audio files"
         " of arbitrary length."
     ),