Update app.py

app.py

@@ -1,19 +1,13 @@
-import spaces
 import torch
-
 import gradio as gr
-import yt_dlp as youtube_dl
 from transformers import pipeline
-from transformers.pipelines.audio_utils import ffmpeg_read
-
+import openai  # Import OpenAI for GPT-4 API integration
+import os  # 確保導入 os
 import tempfile
-import os
 
+# 使用 Whisper Large 模型進行語音轉錄
 MODEL_NAME = "openai/whisper-large-v3-turbo"
 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(
@@ -23,83 +17,42 @@ pipe = pipeline(
     device=device,
 )
 
+openai_api_key = os.getenv('OPENAI_API_KEY')  # Load OpenAI API key
 
-@spaces.GPU
+# 語音轉文字的功能
 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
-
+    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()
+# 增加翻譯功能，呼叫 GPT-4 API
+def translate_text(text, target_language):
+    prompt = f"Translate the following text to {target_language}:\n\n{text}"
     
     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))
-
-@spaces.GPU
-def yt_transcribe(yt_url, task, 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}
-
-    text = pipe(inputs, batch_size=BATCH_SIZE, generate_kwargs={"task": task}, return_timestamps=True)["text"]
-
-    return html_embed_str, text
-
-
-demo = gr.Blocks(theme=gr.themes.Ocean())
+        response = openai.ChatCompletion.create(
+            model="gpt-4o",  # 使用 GPT-4o 模型
+            messages=[{"role": "user", "content": prompt}],
+            max_tokens=500
+        )
+        translation = response.choices[0].message["content"]
+        return translation
+    except Exception as e:
+        return f"翻譯時出錯: {str(e)}"
+
+# Gradio 介面
+demo = gr.Blocks()
 
 mf_transcribe = gr.Interface(
     fn=transcribe,
     inputs=[
-        gr.Audio(sources="microphone", type="filepath"),
+        gr.Audio(source="microphone", type="filepath"),
         gr.Radio(["transcribe", "translate"], label="Task", value="transcribe"),
     ],
     outputs="text",
-    title="清華大學多模態課程＆廖老師嫡傳弟子-第二組 「語音轉文字」model",
+    title="Whisper Large V3 Turbo: Transcribe Audio",
     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"
@@ -111,7 +64,7 @@ mf_transcribe = gr.Interface(
 file_transcribe = gr.Interface(
     fn=transcribe,
     inputs=[
-        gr.Audio(sources="upload", type="filepath", label="Audio file"),
+        gr.Audio(source="upload", type="filepath", label="Audio file"),
         gr.Radio(["transcribe", "translate"], label="Task", value="transcribe"),
     ],
     outputs="text",
@@ -124,23 +77,47 @@ file_transcribe = gr.Interface(
     allow_flagging="never",
 )
 
-yt_transcribe = gr.Interface(
-    fn=yt_transcribe,
+# 增加翻譯���能，讓使用者選擇要翻譯的語言
+def transcribe_and_translate(inputs, task, target_language):
+    text = transcribe(inputs, task)
+    if target_language != "None":
+        translated_text = translate_text(text, target_language)
+        return text, translated_text
+    return text, None
+
+# 介面結合了轉錄和翻譯功能
+mf_transcribe_and_translate = gr.Interface(
+    fn=transcribe_and_translate,
+    inputs=[
+        gr.Audio(source="microphone", type="filepath"),
+        gr.Radio(["transcribe", "translate"], label="Task", value="transcribe"),
+        gr.Dropdown(choices=["French", "German", "Spanish", "Chinese", "None"], label="Translate to Language", value="None")
+    ],
+    outputs=["text", "text"],  # 兩個輸出，一個是原文，一個是翻譯後的文字
+    title="Whisper Large V3 Turbo: Transcribe and Translate",
+    description=(
+        "Transcribe audio from microphone inputs and optionally translate it to a selected language using OpenAI GPT-4."
+    ),
+    allow_flagging="never",
+)
+
+file_transcribe_and_translate = gr.Interface(
+    fn=transcribe_and_translate,
     inputs=[
-        gr.Textbox(lines=1, placeholder="Paste the URL to a YouTube video here", label="YouTube URL"),
-        gr.Radio(["transcribe", "translate"], label="Task", value="transcribe")
+        gr.Audio(source="upload", type="filepath", label="Audio file"),
+        gr.Radio(["transcribe", "translate"], label="Task", value="transcribe"),
+        gr.Dropdown(choices=["French", "German", "Spanish", "Chinese", "None"], label="Translate to Language", value="None")
     ],
-    outputs=["html", "text"],
-    title="Whisper Large V3: Transcribe YouTube",
+    outputs=["text", "text"],  # 兩個輸出，一個是原文，一個是翻譯後的文字
+    title="Whisper Large V3: Transcribe and Translate",
     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."
+        "Transcribe audio from uploaded files and optionally translate it to a selected language using OpenAI GPT-4."
     ),
     allow_flagging="never",
 )
 
+# 結合 Gradio 介面
 with demo:
-    gr.TabbedInterface([mf_transcribe, file_transcribe, yt_transcribe], ["Microphone", "Audio file", "YouTube"])
+    gr.TabbedInterface([mf_transcribe_and_translate, file_transcribe_and_translate], ["Microphone Transcription", "File Transcription"])
 
-demo.queue().launch(ssr_mode=False)
+demo.queue().launch()