Update app.py

app.py

@@ -1,12 +1,266 @@
-from git import Repo
-import os
+import gradio as gr
+import whisper
+from transformers import pipeline
+import gradio as gr
+import pandas as pd
+from io import StringIO
+import os,re
+from langchain.llms import OpenAI
+import pandas as pd
 
-GITHUB_REPO_URL = os.environ.get("GITHUB_REPO_URL")
+from langchain.document_loaders import UnstructuredPDFLoader
 
-if not os.path.exists('repo_directory'):  
-    Repo.clone_from(GITHUB_REPO_URL, 'repo_directory')
+from langchain.prompts import PromptTemplate
+from langchain.chains import LLMChain
 
+from langchain.embeddings.openai import OpenAIEmbeddings
+from langchain.vectorstores import Chroma
+from langchain.text_splitter import CharacterTextSplitter
+from langchain.llms import OpenAI
+from langchain.chains import RetrievalQA
+from langchain.document_loaders import TextLoader
+from langchain.prompts import PromptTemplate
+from langchain.callbacks.stdout import StdOutCallbackHandler
+from langchain.chat_models.openai import ChatOpenAI
+from langchain.prompts.prompt import PromptTemplate
 
-from repo_directory.AudioSentimentAnalysis.app import *
+from langchain.llms import OpenAI
+from langchain.prompts import PromptTemplate
+from langchain.chains import LLMChain
+model = whisper.load_model("base")
+sentiment_analysis = pipeline("sentiment-analysis", framework="pt", model="SamLowe/roberta-base-go_emotions")
 
-main()
+def predict(text):
+    # loader = UnstructuredPDFLoader(file_obj.orig_name)
+    # data = loader.load()
+    # text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
+    # texts = text_splitter.split_documents(data)    
+  
+    # embeddings = OpenAIEmbeddings()
+    # docsearch = Chroma.from_documents(texts, embeddings)
+    # qa = RetrievalQA.from_chain_type(llm=OpenAI(), chain_type="map_reduce", retriever=docsearch.as_retriever())
+    
+    prompt_template = """Ignore all previous instructions. You are the world's hearing aid company markerting agent.
+    I am going to give you a text of a customer. Analyze it and you have 4 products in list which you have to suggest to the customer: 
+    ampli-mini it is mainly works for Maximum comfort and discretion, ampli-connect it is mainly works for Connected to the things you love, 
+    ampli-energy it is mainly works for Full of energy, like you, ampli-easy it is mainly works for Allow yourself to hear well.
+    You can also be creative, funny, or show emotions at time.
+    also share the book a appointment link of your company https://www.amplifon.com/uk/book-an-appointment
+Question: {question}
+Product details:"""
+
+    prompt_template_lang = """
+    You are the world's best languages translator. Will give you some text or paragraph which you have to convert into Tamil, Hindi, Kannada
+    and French.
+    Input Text: {text}
+    Tamil: 
+    Hindi:
+    Kannada:
+    French:
+    """
+    PROMPT = PromptTemplate(
+        template=prompt_template, input_variables=["question"]
+    )
+    PROMPT_lang = PromptTemplate(
+        template=prompt_template_lang, input_variables=["text"]
+    )
+    # chain_type_kwargs = {"prompt": PROMPT}
+    # qa = RetrievalQA.from_chain_type(llm=OpenAI(), chain_type="stuff", retriever=docsearch.as_retriever(), chain_type_kwargs=chain_type_kwargs)
+
+    #Actually, Hi, how are you doing? Actually, I am looking for the hearing aid for my grandfather. He has like age around 62, 65 year old and one of the like major thing that I am looking for the hearing aid product which is like maximum comfort. So if you have anything in that category, so can you please tell me? Thank you.
+
+    llm = OpenAI()
+    # prompt = PromptTemplate(
+    #     input_variables=["product"],
+    #     template="What is a good name for a company that makes {product}?",
+    # )
+    chain = LLMChain(llm=llm, prompt=PROMPT)
+    chain_lang = LLMChain(llm=llm, prompt=PROMPT_lang)
+    
+    resp = chain.run(question=text)
+    resp_lang = chain_lang.run(text=resp)
+    
+    # print(resp)
+    
+    # response = []
+    # category = ["ampli-mini", "ampli-connect", "ampli-energy", "ampli-easy"]
+    # for value in category:
+
+        # response.append({value:ai(qa, value)})
+
+    # html_output = ""
+    # for obj in response:
+    # # Loop through the key-value pairs in the object
+    #     for key, value in obj.items():
+    #         value = re.sub(r'[\d\.]+', '', value)
+    #         value_list = value.strip().split('\n')
+    #         value_html = "<ol>"
+    #         for item in value_list:
+    #             value_html += "<li>{}</li>".format(item.strip())
+    #         value_html += "</ol>"
+    #         html_output += "<h2>{}</h2>".format(key)
+    #         html_output += value_html
+
+    
+    return [resp, resp_lang]
+    
+# def ai(qa,category):
+#     query = "please suggest "+ category +" interview questions"
+#     data = list(filter(None, qa.run(query).split('\n')))
+#     results = list(filter(lambda x: x != ' ', data))
+#     results = "\n".join(results)
+    
+#     return results
+
+
+def analyze_sentiment(text):
+    results = sentiment_analysis(text)
+    sentiment_results = {result['label']: result['score'] for result in results}
+    return sentiment_results
+
+def get_sentiment_emoji(sentiment):
+    # Define the emojis corresponding to each sentiment
+    emoji_mapping = {
+        "disappointment": "😞",
+        "sadness": "😢",
+        "annoyance": "😠",
+        "neutral": "😐",
+        "disapproval": "👎",
+        "realization": "😮",
+        "nervousness": "😬",
+        "approval": "👍",
+        "joy": "😄",
+        "anger": "😡",
+        "embarrassment": "😳",
+        "caring": "🤗",
+        "remorse": "😔",
+        "disgust": "🤢",
+        "grief": "😥",
+        "confusion": "😕",
+        "relief": "😌",
+        "desire": "😍",
+        "admiration": "😌",
+        "optimism": "😊",
+        "fear": "😨",
+        "love": "❤️",
+        "excitement": "🎉",
+        "curiosity": "🤔",
+        "amusement": "😄",
+        "surprise": "😲",
+        "gratitude": "🙏",
+        "pride": "🦁"
+    }
+    return emoji_mapping.get(sentiment, "")
+
+def display_sentiment_results(sentiment_results, option):
+    sentiment_text = ""
+    for sentiment, score in sentiment_results.items():
+        emoji = get_sentiment_emoji(sentiment)
+        if option == "Sentiment Only":
+            sentiment_text += f"{sentiment} {emoji}\n"
+        elif option == "Sentiment + Score":
+            sentiment_text += f"{sentiment} {emoji}: {score}\n"
+    return sentiment_text
+
+def inference(audio, sentiment_option):
+    audio = whisper.load_audio(audio)
+    audio = whisper.pad_or_trim(audio)
+
+    mel = whisper.log_mel_spectrogram(audio).to(model.device)
+
+    _, probs = model.detect_language(mel)
+    lang = max(probs, key=probs.get)
+
+    options = whisper.DecodingOptions(fp16=False)
+    result = whisper.decode(model, mel, options)
+
+    sentiment_results = analyze_sentiment(result.text)
+    print(result.text)
+    prediction  = predict(result.text)
+    sentiment_output = display_sentiment_results(sentiment_results, sentiment_option)
+
+    return lang.upper(), result.text, sentiment_output, prediction[0], prediction[1]
+
+
+def main():
+    title = """<h1 align="center">🎤 Multilingual ASR 💬</h1>"""
+    description = """
+    💻 This demo showcases a general-purpose speech recognition model called Whisper. It is trained on a large dataset of diverse audio and supports multilingual speech recognition, speech translation, and language identification tasks.<br><br>
+    <br>
+    ⚙️ Components of the tool:<br>
+    <br>
+    &nbsp;&nbsp;&nbsp;&nbsp; - Real-time multilingual speech recognition<br>
+    &nbsp;&nbsp;&nbsp;&nbsp; - Language identification<br>
+    &nbsp;&nbsp;&nbsp;&nbsp; - Sentiment analysis of the transcriptions<br>
+    <br>
+    🎯 The sentiment analysis results are provided as a dictionary with different emotions and their corresponding scores.<br>
+    <br>
+    😃 The sentiment analysis results are displayed with emojis representing the corresponding sentiment.<br>
+    <br>
+    ✅ The higher the score for a specific emotion, the stronger the presence of that emotion in the transcribed text.<br>
+    <br>
+    ❓ Use the microphone for real-time speech recognition.<br>
+    <br>
+    ⚡️ The model will transcribe the audio and perform sentiment analysis on the transcribed text.<br>
+    """
+
+    custom_css = """
+    #banner-image {
+        display: block;
+        margin-left: auto;
+        margin-right: auto;
+    }
+    #chat-message {
+        font-size: 14px;
+        min-height: 300px;
+    }
+    """
+
+    block = gr.Blocks(css=custom_css)
+
+    with block:
+        gr.HTML(title)
+
+        with gr.Row():
+            with gr.Column():
+                gr.HTML(description)
+
+        with gr.Group():
+            with gr.Box():
+
+                audio = gr.Audio(
+                    label="Input Audio",
+                    show_label=False,
+                    source="microphone",
+                    type="filepath"
+                )
+    
+                sentiment_option = gr.Radio(
+                    choices=["Sentiment Only", "Sentiment + Score"],
+                    label="Select an option",
+                    default="Sentiment Only"
+                )
+
+            btn = gr.Button("Transcribe")
+
+            lang_str = gr.Textbox(label="Language")
+
+            text = gr.Textbox(label="Transcription")
+
+            sentiment_output = gr.Textbox(label="Sentiment Analysis Results", output=True)
+
+            prediction = gr.Textbox(label="Prediction")
+            
+            language_translation = gr.Textbox(label="Language Translation")
+            
+
+            btn.click(inference, inputs=[audio, sentiment_option], outputs=[lang_str, text, sentiment_output, prediction,language_translation])
+
+            # gr.HTML('''
+            # <div class="footer">
+            #     <p>Model by <a href="https://github.com/openai/whisper" style="text-decoration: underline;" target="_blank">OpenAI</a>
+            #     </p>
+            # </div>
+            # ''') 
+
+    block.launch()