hotfix 2.2

app.py

@@ -1,37 +1,28 @@
 import spaces
 import torch
-
 import gradio as gr
 import yt_dlp as youtube_dl
-from transformers import pipeline
+from faster_whisper import WhisperModel
 from transformers.pipelines.audio_utils import ffmpeg_read
-
 import tempfile
 import os
 
-MODEL_NAME = "openai/whisper-large-v3-turbo"
+MODEL_NAME = "large-v3"
 BATCH_SIZE = 8
 FILE_LIMIT_MB = 1000
 YT_LENGTH_LIMIT_S = 3600  # limit to 1 hour YouTube files
 
-device = 0 if torch.cuda.is_available() else "cpu"
-
-pipe = pipeline(
-    task="automatic-speech-recognition",
-    model=MODEL_NAME,
-    chunk_length_s=30,
-    device=device,
-)
-
+device = "cuda" if torch.cuda.is_available() else "cpu"
+model = WhisperModel(MODEL_NAME, device=device, compute_type="float16" if torch.cuda.is_available() else "int8")
 
 @spaces.GPU
 def transcribe(inputs, task):
     if inputs is None:
         raise gr.Error("No audio file submitted! Please upload or record an audio file before submitting your request.")
-
-    text = pipe(inputs, batch_size=BATCH_SIZE, generate_kwargs={"task": task}, return_timestamps=True)["text"]
-    return  text
-
+    
+    segments, _ = model.transcribe(inputs, task=task)
+    text = " ".join([segment.text for segment in segments])
+    return text
 
 def _return_yt_html_embed(yt_url):
     video_id = yt_url.split("?v=")[-1]
@@ -49,23 +40,11 @@ def download_yt_audio(yt_url, filename):
     except youtube_dl.utils.DownloadError as err:
         raise gr.Error(str(err))
     
-    file_length = info["duration_string"]
-    file_h_m_s = file_length.split(":")
-    file_h_m_s = [int(sub_length) for sub_length in file_h_m_s]
-    
-    if len(file_h_m_s) == 1:
-        file_h_m_s.insert(0, 0)
-    if len(file_h_m_s) == 2:
-        file_h_m_s.insert(0, 0)
-    file_length_s = file_h_m_s[0] * 3600 + file_h_m_s[1] * 60 + file_h_m_s[2]
-    
+    file_length_s = info["duration"]
     if file_length_s > YT_LENGTH_LIMIT_S:
-        yt_length_limit_hms = time.strftime("%HH:%MM:%SS", time.gmtime(YT_LENGTH_LIMIT_S))
-        file_length_hms = time.strftime("%HH:%MM:%SS", time.gmtime(file_length_s))
-        raise gr.Error(f"Maximum YouTube length is {yt_length_limit_hms}, got {file_length_hms} YouTube video.")
-    
-    ydl_opts = {"outtmpl": filename, "format": "worstvideo[ext=mp4]+bestaudio[ext=m4a]/best[ext=mp4]/best"}
+        raise gr.Error(f"Maximum YouTube length is {YT_LENGTH_LIMIT_S} seconds, got {file_length_s} seconds.")
     
+    ydl_opts = {"outtmpl": filename, "format": "bestaudio/best"}
     with youtube_dl.YoutubeDL(ydl_opts) as ydl:
         try:
             ydl.download([yt_url])
@@ -73,7 +52,7 @@ def download_yt_audio(yt_url, filename):
             raise gr.Error(str(err))
 
 @spaces.GPU
-def yt_transcribe(yt_url, task, max_filesize=75.0):
+def yt_transcribe(yt_url, task):
     html_embed_str = _return_yt_html_embed(yt_url)
 
     with tempfile.TemporaryDirectory() as tmpdirname:
@@ -81,15 +60,13 @@ def yt_transcribe(yt_url, task, max_filesize=75.0):
         download_yt_audio(yt_url, filepath)
         with open(filepath, "rb") as f:
             inputs = f.read()
-
-    inputs = ffmpeg_read(inputs, pipe.feature_extractor.sampling_rate)
-    inputs = {"array": inputs, "sampling_rate": pipe.feature_extractor.sampling_rate}
-
-    text = pipe(inputs, batch_size=BATCH_SIZE, generate_kwargs={"task": task}, return_timestamps=True)["text"]
-
+    
+    inputs = ffmpeg_read(inputs, 16000)  # Convertir en 16kHz
+    segments, _ = model.transcribe(inputs, task=task)
+    text = " ".join([segment.text for segment in segments])
+    
     return html_embed_str, text
 
-
 demo = gr.Blocks(theme=gr.themes.Ocean())
 
 mf_transcribe = gr.Interface(
@@ -101,9 +78,7 @@ mf_transcribe = gr.Interface(
     outputs="text",
     title="VerbaLend Demo 1 : Prototype",
     description=(
-        "Transcribe long-form microphone or audio inputs with the click of a button! Demo uses the"
-        f" checkpoint [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe audio files"
-        " of arbitrary length."
+        "Transcribe long-form microphone or audio inputs with the click of a button! Demo uses Faster Whisper"
     ),
     allow_flagging="never",
 )
@@ -116,11 +91,7 @@ file_transcribe = gr.Interface(
     ],
     outputs="text",
     title="VerbaLend Demo 1 : Prototype",
-    description=(
-        "Transcribe long-form microphone or audio inputs with the click of a button! Demo uses the"
-        f" checkpoint [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe audio files"
-        " of arbitrary length."
-    ),
+    description="Transcribe uploaded audio files with Faster Whisper.",
     allow_flagging="never",
 )
 
@@ -132,11 +103,7 @@ yt_transcribe = gr.Interface(
     ],
     outputs=["html", "text"],
     title="VerbaLend Demo 1 : Prototype",
-    description=(
-        "Transcribe long-form YouTube videos with the click of a button! Demo uses the checkpoint"
-        f" [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe video files of"
-        " arbitrary length."
-    ),
+    description="Transcribe YouTube videos using Faster Whisper.",
     allow_flagging="never",
 )
 
@@ -144,3 +111,4 @@ with demo:
     gr.TabbedInterface([mf_transcribe, file_transcribe, yt_transcribe], ["Microphone", "Audio file", "YouTube"])
 
 demo.queue().launch(ssr_mode=False)
+