Duplicate from remzicam/voicebot_german

Co-authored-by: rc <remzicam@users.noreply.huggingface.co>

.gitattributes

@@ -0,0 +1,34 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm 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

README.md

@@ -0,0 +1,13 @@
+---
+title: Voicebot German
+emoji: 🚀
+colorFrom: indigo
+colorTo: blue
+sdk: gradio
+sdk_version: 3.12.0
+app_file: app.py
+pinned: false
+duplicated_from: remzicam/voicebot_german
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,57 @@
+"""Deploying AI Voice Chatbot Gradio App."""
+from gradio import Audio, Interface, Textbox
+from typing import Tuple
+
+from utils import (TextGenerationPipeline, from_en_translation,
+                   html_audio_autoplay, stt, to_en_translation, tts,
+                   tts_to_bytesio)
+
+max_answer_length = 100
+desired_language = "de"
+response_generator_pipe = TextGenerationPipeline(max_length=max_answer_length)
+
+
+def main(audio: object) -> Tuple[str, str, str, object]:
+    """Calls functions for deploying gradio app.
+
+    It responds both verbally and in text
+    by taking voice input from user.
+
+    Args:
+        audio (object): recorded speech of user
+
+    Returns:
+        tuple containing
+
+        - user_speech_text (str) : recognized speech
+        - bot_response_de (str) : translated answer of bot
+        - bot_response_en (str) : bot's original answer
+        - html (object) : autoplayer for bot's speech
+    """
+    user_speech_text = stt(audio, desired_language)
+    tranlated_text = to_en_translation(user_speech_text, desired_language)
+    bot_response_en = response_generator_pipe(tranlated_text)
+    bot_response_de = from_en_translation(bot_response_en, desired_language)
+    bot_voice = tts(bot_response_de, desired_language)
+    bot_voice_bytes = tts_to_bytesio(bot_voice)
+    html = html_audio_autoplay(bot_voice_bytes)
+    return user_speech_text, bot_response_de, bot_response_en, html
+
+
+Interface(
+    fn=main,
+    inputs=[
+        Audio(
+            source="microphone",
+            type="filepath",
+        ),
+    ],
+    outputs=[
+        Textbox(label="You said: "),
+        Textbox(label="AI said: "),
+        Textbox(label="AI said (English): "),
+        "html",
+    ],
+    live=True,
+    allow_flagging="never",
+).launch(debug=True)

requirements.txt

@@ -0,0 +1,7 @@
+--find-links https://download.pytorch.org/whl/torch_stable.html
+torch==1.13.1+cpu
+gradio==3.14.0
+SpeechRecognition==3.9.0
+mtranslate==1.8
+gTTS==2.3.0
+transformers==4.25.1

utils.py

@@ -0,0 +1,157 @@
+"""Some utility functions for the app."""
+from base64 import b64encode
+from io import BytesIO
+
+from gtts import gTTS
+from mtranslate import translate
+from speech_recognition import AudioFile, Recognizer
+from transformers import (BlenderbotSmallForConditionalGeneration,
+                          BlenderbotSmallTokenizer)
+
+
+def stt(audio: object, language: str) -> str:
+    """Converts speech to text.
+
+    Args:
+        audio: record of user speech
+
+    Returns:
+        text (str): recognized speech of user
+    """
+    r = Recognizer()
+    # open the audio file
+    with AudioFile(audio) as source:
+        # listen for the data (load audio to memory)
+        audio_data = r.record(source)
+        # recognize (convert from speech to text)
+        text = r.recognize_google(audio_data, language=language)
+    return text
+
+
+def to_en_translation(text: str, language: str) -> str:
+    """Translates text from specified language to English.
+
+    Args:
+        text (str): input text
+        language (str): desired language
+
+    Returns:
+        str: translated text
+    """
+    return translate(text, "en", language)
+
+
+def from_en_translation(text: str, language: str) -> str:
+    """Translates text from english to specified language.
+
+    Args:
+        text (str): input text
+        language (str): desired language
+
+    Returns:
+        str: translated text
+    """
+    return translate(text, language, "en")
+
+
+class TextGenerationPipeline:
+    """Pipeline for text generation of blenderbot model.
+
+    Returns:
+        str: generated text
+    """
+
+    # load tokenizer and the model
+    model_name = "facebook/blenderbot_small-90M"
+    tokenizer = BlenderbotSmallTokenizer.from_pretrained(model_name)
+    model = BlenderbotSmallForConditionalGeneration.from_pretrained(model_name)
+
+    def __init__(self, **kwargs):
+        """Specififying text generation parameters.
+
+        For example: max_length=100 which generates text shorter than
+        100 tokens. Visit:
+        https://huggingface.co/docs/transformers/main_classes/text_generation
+        for more parameters
+        """
+        self.__dict__.update(kwargs)
+
+    def preprocess(self, text) -> str:
+        """Tokenizes input text.
+
+        Args:
+            text (str): user specified text
+
+        Returns:
+            torch.Tensor (obj): text representation as tensors
+        """
+        return self.tokenizer(text, return_tensors="pt")
+
+    def postprocess(self, outputs) -> str:
+        """Converts tensors into text.
+
+        Args:
+            outputs (torch.Tensor obj): model text generation output
+
+        Returns:
+            str: generated text
+        """
+        return self.tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+    def __call__(self, text: str) -> str:
+        """Generates text from input text.
+
+        Args:
+            text (str): user specified text
+
+        Returns:
+            str: generated text
+        """
+        tokenized_text = self.preprocess(text)
+        output = self.model.generate(**tokenized_text, **self.__dict__)
+        return self.postprocess(output)
+
+
+def tts(text: str, language: str) -> object:
+    """Converts text into audio object.
+
+    Args:
+        text (str): generated answer of bot
+
+    Returns:
+        object: text to speech object
+    """
+    return gTTS(text=text, lang=language, slow=False)
+
+
+def tts_to_bytesio(tts_object: object) -> bytes:
+    """Converts tts object to bytes.
+
+    Args:
+        tts_object (object): audio object obtained from gtts
+
+    Returns:
+        bytes: audio bytes
+    """
+    bytes_object = BytesIO()
+    tts_object.write_to_fp(bytes_object)
+    bytes_object.seek(0)
+    return bytes_object.getvalue()
+
+
+def html_audio_autoplay(bytes: bytes) -> object:
+    """Creates html object for autoplaying audio at gradio app.
+
+    Args:
+        bytes (bytes): audio bytes
+
+    Returns:
+        object: html object that provides audio autoplaying
+    """
+    b64 = b64encode(bytes).decode()
+    html = f"""
+    <audio controls autoplay>
+    <source src="data:audio/wav;base64,{b64}" type="audio/wav">
+    </audio>
+    """
+    return html