Move to Speech to Speech

README.md

@@ -1,15 +1,14 @@
 ---
 title: Indonesian Whisperer
 emoji: 🇮🇩
-colorFrom: indigo
+colorFrom: purple
 colorTo: red
 sdk: gradio
-sdk_version: 3.9.1
+sdk_version: 3.15.0
 app_file: app.py
-pinned: false
+pinned: true
 tags:
 - whisper-event
-duplicated_from: whisper-event/whisper-demo
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -1,12 +1,20 @@
 import torch
-
 import gradio as gr
-import pytube as pt
 from transformers import pipeline
-from huggingface_hub import model_info
+import tempfile
+from neon_tts_plugin_coqui import CoquiTTS
+from datetime import datetime
+import time
+import psutil
+from mtranslate import translate
+
 
 MODEL_NAME = "cahya/whisper-medium-id" #this always needs to stay in line 8 :D sorry for the hackiness
 lang = "id"
+title = "indonesian Whisperer"
+description = "Cross Language Speech to Speech using OpenAI Whisper and Coqui TTS"
+info = "more info at [indonesian Whisperer](https://github.com/cahya-wirawan/indonesian-whisperer)"
+badge = "https://img.shields.io/badge/Powered%20by-Indonesian%20Whisperer-red"
 
 device = 0 if torch.cuda.is_available() else "cpu"
 
@@ -37,62 +45,58 @@ def transcribe(microphone, file_upload):
     return warn_output + 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>"
+LANGUAGES = list(CoquiTTS.langs.keys())
+default_lang = "en"
+
+coquiTTS = CoquiTTS()
+
+
+def tts(language: str, audio_file: str):
+    print(f"### {datetime.now()} TTS", language, audio_file)
+    transcribed = transcribe(None, audio_file)
+    print(f"### {datetime.now()} transcribed:", transcribed)
+    translation = translate(transcribed, language, "id")
+    # return output
+    with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as fp:
+        coquiTTS.get_tts(translation, fp, speaker={"language": language})
+        print(f"### {datetime.now()} fp.name:", fp.name)
+        return fp.name
+
+
+with gr.Blocks() as blocks:
+    gr.Markdown("<h1 style='text-align: center; margin-bottom: 1rem'>"
+                + title
+                + "</h1>")
+    gr.Markdown(description)
+    with gr.Row():# equal_height=False
+        with gr.Column():# variant="panel"
+            upload = gr.Audio(label="Upload", source="upload", type="filepath", optional=True)
+            print("upload:", upload)
+            radio = gr.Radio(
+                label="Language",
+                choices=LANGUAGES,
+                value=default_lang
+            )
+            with gr.Row(): # mobile_collapse=False
+                submit = gr.Button("Submit", variant="primary")
+        audio = gr.Audio(label="Output", interactive=False)
+    memory = psutil.virtual_memory()
+    gr.Markdown(info)
+    system_status = info = f"""
+            *Memory: {memory.total/(1024*1024*1024):.2f}GB, used: {memory.percent}%, available: {memory.available/(1024*1024*1024):.2f}GB*
+            """
+    gr.Markdown(system_status)
+    gr.Markdown("<center>"
+                +f'<img src={badge} alt="visitors badge"/>'
+                +"</center>")
+
+    # actions
+    submit.click(
+        tts,
+        [radio, upload],
+        [audio],
     )
-    return HTML_str
-
-
-def yt_transcribe(yt_url):
-    yt = pt.YouTube(yt_url)
-    html_embed_str = _return_yt_html_embed(yt_url)
-    stream = yt.streams.filter(only_audio=True)[0]
-    stream.download(filename="audio.mp3")
-
-    text = pipe("audio.mp3")["text"]
-
-    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.Audio(source="upload", type="filepath", optional=True),
-    ],
-    outputs="text",
-    layout="horizontal",
-    theme="huggingface",
-    title="Whisper Demo: Transcribe Audio",
-    description=(
-        "Transcribe long-form microphone or audio inputs with the click of a button! Demo uses the the fine-tuned"
-        f" checkpoint [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe audio files"
-        " of arbitrary length."
-    ),
-    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")],
-    outputs=["html", "text"],
-    layout="horizontal",
-    theme="huggingface",
-    title="Whisper Demo: Transcribe YouTube",
-    description=(
-        "Transcribe long-form YouTube videos with the click of a button! Demo uses the the fine-tuned checkpoint:"
-        f" [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe audio files of"
-        " arbitrary length."
-    ),
-    allow_flagging="never",
-)
+    radio.change(lambda lang: CoquiTTS.langs[lang]["sentence"], radio)
 
-with demo:
-    gr.TabbedInterface([mf_transcribe, yt_transcribe], ["Transcribe Audio", "Transcribe YouTube"])
+blocks.launch()
 
-demo.launch(enable_queue=True)

packages.txt

@@ -1 +1,3 @@
 ffmpeg
+libsndfile1
+espeak-ng

requirements.txt

@@ -1,3 +1,5 @@
 git+https://github.com/huggingface/transformers
 torch
-pytube
+neon-tts-plugin-coqui==0.6.0
+psutil
+mtranslate