Add application file

packages.txt

@@ -0,0 +1 @@
+ffmpeg

.gitattributes

@@ -32,4 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.xz filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1 @@
+.idea

app.py

@@ -0,0 +1,184 @@
+import torch
+
+import time
+
+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 = "openai/whisper-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,
+)
+
+
+def chunks_to_srt(chunks):
+    srt_format = ""
+    for i, chunk in enumerate(chunks, 1):
+        start_time, end_time = chunk['timestamp']
+        start_time_hms = "{:02}:{:02}:{:02},{:03}".format(int(start_time // 3600), int((start_time % 3600) // 60),
+                                                          int(start_time % 60), int((start_time % 1) * 1000))
+        end_time_hms = "{:02}:{:02}:{:02},{:03}".format(int(end_time // 3600), int((end_time % 3600) // 60),
+                                                        int(end_time % 60), int((end_time % 1) * 1000))
+        srt_format += f"{i}\n{start_time_hms} --> {end_time_hms}\n{chunk['text']}\n\n"
+    return srt_format
+
+
+def transcribe(inputs, task, return_timestamps, language):
+    if inputs is None:
+        raise gr.Error("No audio file submitted! Please upload or record an audio file before submitting your request.")
+
+    # Map the language names to their corresponding codes
+    language_codes = {"English": "en", "Uzbek": "uz"}
+    language_code = language_codes.get(language, "en")  # Default to "en" if the language is not found
+    result = pipe(inputs, batch_size=BATCH_SIZE, generate_kwargs={"task": task, "language": f"<|{language_code}|>"},
+                  return_timestamps=return_timestamps)
+
+    if return_timestamps:
+        return chunks_to_srt(result['chunks'])
+    else:
+        return result['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))
+
+
+def yt_transcribe(yt_url, task, return_timestamps, language, max_filesize=75.0):
+    html_embed_str = _return_yt_html_embed(yt_url)
+
+    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()
+
+    inputs = ffmpeg_read(inputs, pipe.feature_extractor.sampling_rate)
+    inputs = {"array": inputs, "sampling_rate": pipe.feature_extractor.sampling_rate}
+
+    # Map the language names to their corresponding codes
+    language_codes = {"English": "en", "Uzbek": "uz"}
+    language_code = language_codes.get(language, "en")  # Default to "en" if the language is not found
+
+    result = pipe(inputs, batch_size=BATCH_SIZE, generate_kwargs={"task": task, "language": f"<|{language_code}|>"},
+                  return_timestamps=return_timestamps)
+
+    if return_timestamps:
+        return html_embed_str, chunks_to_srt(result['chunks'])
+    else:
+        return html_embed_str, result['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"),
+        gr.inputs.Checkbox(label="Return timestamps"),
+        gr.inputs.Dropdown(choices=["English", "Uzbek"], label="Language"),
+    ],
+    outputs="text",
+    layout="horizontal",
+    theme="huggingface",
+    title="Whisper Large V3: Transcribe Audio",
+    description=(
+        "\n\n"
+        "<center>⭐️Brought to you by <a href='https://note.com/sangmin/n/n9813f2064a6a'>Chiomirai School</a>⭐️</center>"
+    ),
+    allow_flagging="never",
+)
+
+file_transcribe = gr.Interface(
+    fn=transcribe,
+    inputs=[
+        gr.inputs.Audio(source="upload", type="filepath", optional=True, label="Audio file"),
+        gr.inputs.Radio(["transcribe", "translate"], label="Task", default="transcribe"),
+        gr.inputs.Checkbox(label="Return timestamps"),
+        gr.inputs.Dropdown(choices=["English", "Uzbek"], label="Language"),
+    ],
+    outputs="text",
+    layout="horizontal",
+    theme="huggingface",
+    title="Whisper Large V3: Transcribe Audio File",
+    description=(
+        "\n\n"
+        "<center>⭐️Brought to you by <a href='https://note.com/sangmin/n/n9813f2064a6a'>Chiomirai School</a>⭐️</center>"
+    ),
+    allow_flagging="never",
+)
+
+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"),
+        gr.inputs.Checkbox(label="Return timestamps"),
+        gr.inputs.Dropdown(choices=["English", "Uzbek"], label="Language"),
+    ],
+    outputs=["html", "text"],
+    layout="horizontal",
+    theme="huggingface",
+    title="Whisper Large V3: Transcribe YouTube",
+    description=(
+        "\n\n"
+        "<center>⭐️Brought to you by <a href='https://note.com/sangmin/n/n9813f2064a6a'>Chiomirai School</a>⭐️</center>"
+    ),
+    allow_flagging="never",
+)
+
+with demo:
+    gr.TabbedInterface([mf_transcribe, file_transcribe, yt_transcribe], ["Microphone", "Audio file", "YouTube"])
+
+demo.launch(enable_queue=True)

requirements.txt

@@ -0,0 +1,3 @@
+git+https://github.com/huggingface/transformers
+torch
+yt-dlp