app.py

import os
import openai
import wget
import streamlit as st
from PIL import Image
from serpapi import GoogleSearch
import torch
from diffusers import StableDiffusionPipeline
from bokeh.models.widgets import Button
from bokeh.models import CustomJS
from streamlit_bokeh_events import streamlit_bokeh_events
import base64
from streamlit_player import st_player
from pytube import YouTube
from pytube import Search
import io
import warnings
from PIL import Image
from stability_sdk import client
import stability_sdk.interfaces.gooseai.generation.generation_pb2 as generation

os.environ['STABILITY_HOST'] = 'grpc.stability.ai:443'
os.environ['STABILITY_KEY'] = 'sk-Ndzkpi6OYwM5fQgEJAVwRPbFZSMNyFk0GoZw1EvNtqVExGdi'
os.environ['GOOGLE_API'] = 'e77d5416608a110ea2babd7b2e33ede48b0c4159ade5cfd5cebbc7483c513ff3'
os.environ['OPENAI_KEY'] = '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'

stability_api = client.StabilityInference(
    key=os.environ['STABILITY_KEY'], # API Key reference.
    verbose=True, # Print debug messages.
    engine="stable-diffusion-v1-5", # Set the engine to use for generation.
    # Available engines: stable-diffusion-v1 stable-diffusion-v1-5 stable-diffusion-512-v2-0 stable-diffusion-768-v2-0
    # stable-diffusion-512-v2-1 stable-diffusion-768-v2-1 stable-inpainting-v1-0 stable-inpainting-512-v2-0
)

def search_internet(question):
    params = {
        "q": question,
        "location": "Bengaluru, Karnataka, India",
        "hl": "hi",
        "gl": "in",
        "google_domain": "google.co.in",
        # "api_key": ""
        "api_key": os.environ['GOOGLE_API']
    }

    params = {
        "q": question,
        "location": "Bengaluru, Karnataka, India",
        "hl": "hi",
        "gl": "in",
        "google_domain": "google.co.in",
        # "api_key": ""
        "api_key": os.environ['GOOGLE_API']
    }

    search = GoogleSearch(params)
    results = search.get_dict()
    organic_results = results["organic_results"]
  

    snippets = ""
    counter = 1
    for item in organic_results:
        snippets += str(counter) + ". " + item.get("snippet", "") + '\n' + item['about_this_result']['source']['source_info_link'] + '\n'
        counter += 1

    # snippets

    response = openai.Completion.create(
        model="text-davinci-003",
        prompt=f'''following are snippets from google search with these as knowledge base only answer questions and print  reference link as well followed by answer. \n\n {snippets}\n\n question-{question}\n\nAnswer-''',
        temperature=0.49,
        max_tokens=256,
        top_p=1,
        frequency_penalty=0,
        presence_penalty=0)

    string_temp = response.choices[0].text

    st.write(string_temp)
    st.write(snippets)

# openai.api_key = ""
openai.api_key = os.environ['OPENAI_KEY']

def openai_response(PROMPT):
    response = openai.Image.create(
    prompt=PROMPT,
    n=1,
    size="256x256",
)
    return response["data"][0]["url"]

#page_bg_img = """
#<style>
#[data-testid="stAppViewContainer"] {
#background-color: #ffffff;
#opacity: 0.8;
#background-image:  repeating-radial-gradient( circle at 0 0, transparent 0, #ffffff 40px ), repeating-linear-gradient( #55a6f655, #55a6f6 );
#}
#</style>
#"""

#st.markdown(page_bg_img, unsafe_allow_html=True)
st.title("Welcome to :red[_HyperChat_]!!🤖")
st.title("How can I help?")

Input_type = st.radio(
    "**Input type:**",
    ('TEXT', 'SPEECH')
    )

if Input_type == 'TEXT':
    #page_bg_img2 = """
    #<style>
    #[data-testid="stAppViewContainer"] {
    #background-color: #e5e5f7;
    #opacity: 0.8;
    #background-size: 20px 20px;
    #background-image:  repeating-linear-gradient(0deg, #32d947, #32d947 1px, #e5e5f7 1px, #e5e5f7);
    #}
    #</style>
    #"""
    #st.markdown(page_bg_img, unsafe_allow_html=True)
    st.write('**You are now in Text input mode**')
    mytext = st.text_input('**Go on! Ask me anything:**')
    if st.button("SUBMIT"):
        question=mytext
        response = openai.Completion.create(
          model="text-davinci-003",
          prompt=f'''Your name is alexa and  knowledge cutoff date is 2021-09, and it is not aware of any events after that time. if the  
                    Answer to following questions is not  from your knowledge base or in case of queries like weather 
                      updates / stock updates / current news Etc which requires you to have internet connection  then print i don't have access to internet to answer your question, 
                      if  question is related to  image or  painting or drawing generation then print ipython type output   function gen_draw("detailed prompt of image to be generated")
                      if the question is related to playing a song or video or music of a singer then print ipython type output  function vid_tube("relevent search query")
                      if the question is related to operating home appliances then print ipython type output function home_app(" action(ON/Off),appliance(TV,Geaser,Fridge,Lights,fans,AC)") . 
                      if question is realted to sending mail or sms then print ipython type output function messenger_app(" message of us ,messenger(email,sms)")
                      \nQuestion-{question}
                      \nAnswer -''',
          temperature=0.49,
          max_tokens=256,
          top_p=1,
          frequency_penalty=0,
          presence_penalty=0
        )
        string_temp=response.choices[0].text

        if ("gen_draw" in string_temp):
            try:
                # Set up our initial generation parameters.
                answers = stability_api.generate(
                prompt = mytext,
                seed=992446758, # If a seed is provided, the resulting generated image will be deterministic.
                        # What this means is that as long as all generation parameters remain the same, you can always recall the same image simply by generating it again.
                        # Note: This isn't quite the case for Clip Guided generations, which we'll tackle in a future example notebook.
                steps=30, # Amount of inference steps performed on image generation. Defaults to 30.
                cfg_scale=8.0, # Influences how strongly your generation is guided to match your prompt.
                    # Setting this value higher increases the strength in which it tries to match your prompt.
                    # Defaults to 7.0 if not specified.
                width=512, # Generation width, defaults to 512 if not included.
                height=512, # Generation height, defaults to 512 if not included.
                samples=1, # Number of images to generate, defaults to 1 if not included.
                sampler=generation.SAMPLER_K_DPMPP_2M # Choose which sampler we want to denoise our generation with.
                                                    # Defaults to k_dpmpp_2m if not specified. Clip Guidance only supports ancestral samplers.
                                                    # (Available Samplers: ddim, plms, k_euler, k_euler_ancestral, k_heun, k_dpm_2, k_dpm_2_ancestral, k_dpmpp_2s_ancestral, k_lms, k_dpmpp_2m)
                )

                # Set up our warning to print to the console if the adult content classifier is tripped.
                # If adult content classifier is not tripped, save generated images.
                for resp in answers:
                    for artifact in resp.artifacts:
                        if artifact.finish_reason == generation.FILTER:
                            warnings.warn(
                                "Your request activated the API's safety filters and could not be processed."
                                "Please modify the prompt and try again.")
                        if artifact.type == generation.ARTIFACT_IMAGE:
                            img = Image.open(io.BytesIO(artifact.binary))
                            st.image(img)
                            img.save(str(artifact.seed)+ ".png") # Save our generated images with their seed number as the filename.
            except:        
                st.write('image is being generated please wait...')
                def extract_image_description(input_string):
                    return input_string.split('gen_draw("')[1].split('")')[0]
                prompt=extract_image_description(string_temp)
                # model_id = "CompVis/stable-diffusion-v1-4"
                model_id='runwayml/stable-diffusion-v1-5'
                device = "cuda"


                pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
                pipe = pipe.to(device)

                # prompt = "a photo of an astronaut riding a horse on mars"
                image = pipe(prompt).images[0]  
                
                image.save("astronaut_rides_horse.png")
                st.image(image)
                # image

        elif ("vid_tube" in string_temp):
            s = Search(mytext)
            search_res = s.results
            first_vid = search_res[0]
            print(first_vid)
            string = str(first_vid)
            video_id = string[string.index('=') + 1:-1]
            # print(video_id)
            YoutubeURL = "https://www.youtube.com/watch?v="
            OurURL = YoutubeURL + video_id
            st.write(OurURL)
            st_player(OurURL)

        elif ("don't" in string_temp or "internet" in string_temp  ):
            st.write('searching internet ')
            search_internet(question)

        else:
            st.write(string_temp)
    
elif Input_type == 'SPEECH':
    stt_button = Button(label="Speak", width=100)
    stt_button.js_on_event("button_click", CustomJS(code="""
        var recognition = new webkitSpeechRecognition();
        recognition.continuous = true;
        recognition.interimResults = true;
        recognition.onresult = function (e) {
            var value = "";
            for (var i = e.resultIndex; i < e.results.length; ++i) {
                if (e.results[i].isFinal) {
                    value += e.results[i][0].transcript;
                }
            }
            if ( value != "") {
                document.dispatchEvent(new CustomEvent("GET_TEXT", {detail: value}));
            }
        }
        recognition.start();
        """))
    
    result = streamlit_bokeh_events(
    stt_button,
    events="GET_TEXT",
    key="listen",
    refresh_on_update=False,
    override_height=75,
    debounce_time=0)

    if result:
        if "GET_TEXT" in result:
            st.write(result.get("GET_TEXT"))
            question = result.get("GET_TEXT")
            response = openai.Completion.create(
            model="text-davinci-003",
            prompt=f'''Your knowledge cutoff is 2021-09, and it is not aware of any events after that time. if the  
                      Answer to following questions is not  from your knowledge base or in case of queries like weather 
                        updates / stock updates / current news Etc which requires you to have internet connection  then print i don't have access to internet to answer your question, 
                        if  question is related to  image or  painting or drawing generation then print ipython type output   function gen_draw("detailed prompt of image to be generated")
                        if the question is related to playing a song or video or music of a singer then print ipython type output  function vid_tube("relevent search query")
                        \nQuestion-{question}
                        \nAnswer -''',
            temperature=0.49,
            max_tokens=256,
            top_p=1,
            frequency_penalty=0,
            presence_penalty=0
            )
            string_temp=response.choices[0].text

            if ("gen_draw" in string_temp):
                st.write('*image is being generated please wait..* ')
                def extract_image_description(input_string):
                    return input_string.split('gen_draw("')[1].split('")')[0]
                prompt=extract_image_description(string_temp)
                # model_id = "CompVis/stable-diffusion-v1-4"
                model_id='runwayml/stable-diffusion-v1-5'
                device = "cuda"

                pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
                pipe = pipe.to(device)

                # prompt = "a photo of an astronaut riding a horse on mars"
                image = pipe(prompt).images[0]  
                  
                image.save("astronaut_rides_horse.png")
                st.image(image)
                # image 
            
            elif ("vid_tube" in string_temp):
                s = Search(question)
                search_res = s.results
                first_vid = search_res[0]
                print(first_vid)
                string = str(first_vid)
                video_id = string[string.index('=') + 1:-1]
                # print(video_id)
                YoutubeURL = "https://www.youtube.com/watch?v="
                OurURL = YoutubeURL + video_id
                st.write(OurURL)
                st_player(OurURL)

            elif ("don't" in string_temp or "internet" in string_temp  ):
                st.write('*searching internet*')
                search_internet(question)
            else:
                st.write(string_temp)