Update app.py

app.py

@@ -1,12 +1,20 @@
-import gradio as gr
 import torch
-import torchaudio
-from transformers import AutoProcessor, AutoModelForSpeechSeq2Seq, pipeline
-import numpy as np
 
-# Load the Whisper model
-MODEL_NAME = "chiyo123/whisper-small-bemba"
-device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+import gradio as gr
+import yt_dlp as youtube_dl
+from transformers import pipeline
+from transformers.pipelines.audio_utils import ffmpeg_read
+
+import tempfile
+import os
+
+MODEL_NAME = "chiyo123/whisper-small-tonga"
+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,
@@ -14,50 +22,129 @@ pipe = pipeline(
     device=device,
 )
 
-def transcribe(audio):
-    if audio is None:
-        return "No audio recorded. Please record your audio."
 
-    # Extract the sampling rate and audio data
-    sampling_rate, audio_data = audio
+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
+
+
+def _return_yt_html_embed(yt_url):
+    video_id = yt_url.split("?v=")[-1]
+    HTML_str = (
+        f'<center> <iframe width="500" height="320" src="https://www.youtube.com/embed/{video_id}"> </iframe>'
+        " </center>"
+    )
+    return HTML_str
+
+def download_yt_audio(yt_url, filename):
+    info_loader = youtube_dl.YoutubeDL()
+    
+    try:
+        info = info_loader.extract_info(yt_url, download=False)
+    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]
+    
+    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"}
+    
+    with youtube_dl.YoutubeDL(ydl_opts) as ydl:
+        try:
+            ydl.download([yt_url])
+        except youtube_dl.utils.ExtractorError as err:
+            raise gr.Error(str(err))
 
-    # Convert the numpy array to a PyTorch tensor
-    audio_tensor = torch.tensor(audio_data).unsqueeze(0)
 
-    # Resample the audio to 16 kHz
-    resampler = torchaudio.transforms.Resample(orig_freq=sampling_rate, new_freq=16000)
-    audio_tensor = resampler(audio_tensor)
+def yt_transcribe(yt_url, task, max_filesize=75.0):
+    html_embed_str = _return_yt_html_embed(yt_url)
 
-    # Remove the channel dimension if it exists
-    if audio_tensor.shape[0] > 1:
-        audio_tensor = torch.mean(audio_tensor, dim=0, keepdim=True)
+    with tempfile.TemporaryDirectory() as tmpdirname:
+        filepath = os.path.join(tmpdirname, "video.mp4")
+        download_yt_audio(yt_url, filepath)
+        with open(filepath, "rb") as f:
+            inputs = f.read()
 
-    # Prepare the input for the model
-    inputs = {"array": audio_tensor.squeeze().numpy(), "sampling_rate": 16000}
+    inputs = ffmpeg_read(inputs, pipe.feature_extractor.sampling_rate)
+    inputs = {"array": inputs, "sampling_rate": pipe.feature_extractor.sampling_rate}
 
-    # Perform transcription
-    result = pipe(inputs)
-    return result["text"]
+    text = pipe(inputs, batch_size=BATCH_SIZE, generate_kwargs={"task": task}, return_timestamps=True)["text"]
 
-# Create Gradio interface with waveform visualization
-input_audio = gr.Audio(
-    sources=["microphone"],
-    type="filepath",
-    waveform_options=gr.WaveformOptions(
-        waveform_color="#01C6FF",
-        waveform_progress_color="#0066B4",
-        skip_length=2,
-        show_controls=True,
+    return html_embed_str, text
+
+
+demo = gr.Blocks()
+
+mf_transcribe = gr.Interface(
+    fn=transcribe,
+    inputs=[
+        gr.inputs.Audio(source="microphone", type="filepath", optional=True),
+        gr.inputs.Radio(["transcribe", "translate"], label="Task", default="transcribe"),
+    ],
+    outputs="text",
+    layout="horizontal",
+    theme="huggingface",
+    title="chiyo123/whisper-small-tonga",
+    description=(
+        "Transcribe long-form microphone or audio inputs with the click of a button! Demo uses the OpenAI Whisper"
+        f" checkpoint [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe audio files"
+        " of arbitrary length."
     ),
+    allow_flagging="never",
 )
 
-interface = gr.Interface(
+file_transcribe = gr.Interface(
     fn=transcribe,
-    inputs=input_audio,
-    outputs=gr.Textbox(label="Transcription"),
-    title="Tonga Speech-to-Text Transcriber",
-    description="Speak into the microphone and get a transcription using the Whisper Small Tonga model."
+    inputs=[
+        gr.inputs.Audio(source="upload", type="filepath", optional=True, label="Audio file"),
+        gr.inputs.Radio(["transcribe", "translate"], label="Task", default="transcribe"),
+    ],
+    outputs="text",
+    layout="horizontal",
+    theme="huggingface",
+    title="Whisper Large V3: Transcribe Audio",
+    description=(
+        "Transcribe long-form microphone or audio inputs with the click of a button! Demo uses the OpenAI Whisper"
+        f" checkpoint [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe audio files"
+        " of arbitrary length."
+    ),
+    allow_flagging="never",
 )
 
-if __name__ == "__main__":
-    interface.launch()
+yt_transcribe = gr.Interface(
+    fn=yt_transcribe,
+    inputs=[
+        gr.inputs.Textbox(lines=1, placeholder="Paste the URL to a YouTube video here", label="YouTube URL"),
+        gr.inputs.Radio(["transcribe", "translate"], label="Task", default="transcribe")
+    ],
+    outputs=["html", "text"],
+    layout="horizontal",
+    theme="huggingface",
+    title="Whisper Large V3: Transcribe YouTube",
+    description=(
+        "Transcribe long-form YouTube videos with the click of a button! Demo uses the OpenAI Whisper checkpoint"
+        f" [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe video files of"
+        " arbitrary length."
+    ),
+    allow_flagging="never",
+)
+
+with demo:
+    gr.TabbedInterface([mf_transcribe, file_transcribe, yt_transcribe], ["Microphone", "Audio file", "YouTube"])
+
+demo.launch(enable_queue=True)