Update app.py

app.py

@@ -23,6 +23,45 @@ from google.oauth2 import service_account
 from googleapiclient.discovery import build
 import wget
 import urllib.request
+import sqlite3
+import pandas as pd
+import pandasql as ps
+
+def clean(value):
+    val = value.replace("'",'').replace("[",'').replace("]",'')
+    return val
+
+def save_uploadedfile(uploadedfile):
+    with open(uploadedfile.name,"wb") as f:
+        f.write(uploadedfile.getbuffer())
+
+def gpt3(texts):
+    # openai.api_key = os.environ["Secret"]
+    openai.api_key = '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'#'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' #'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'
+    response = openai.Completion.create(
+    engine="text-davinci-002",
+    prompt= texts,
+        temperature=temp,
+        max_tokens=750,
+        top_p=1,
+        frequency_penalty=0.0,          
+        presence_penalty=0.0,
+        stop = (";", "/*", "</code>"))
+    x = response.choices[0].text 
+    return x
+
+def warning(sqlOutput):
+    dl = []
+    lst = ['DELETE','DROP','TRUNCATE','MERGE','ALTER','UPDATE','INSERT']
+    op2 = " ".join(sqlOutput.split())
+    op3 = op2.split(' ')
+    op4 = list(map(lambda x: x.upper(), op3))
+    for i in op4:
+        if i in lst: 
+            dl.append(i)
+    for i in dl:
+        st.warning("This query will " + i + " the data ",icon="⚠️")
+
 
 stability_api = client.StabilityInference(
     key=st.secrets["STABILITY_KEY"], #os.environ("STABILITY_KEY"), # key=os.environ['STABILITY_KEY'], # API Key reference.
@@ -138,209 +177,201 @@ def openai_response(PROMPT):
 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
+Usage = st.radio(
+    "I want to ask:",
+    ('Random Questions', 'Questions based on custom CSV data')
+)
 
-        if ("gen_draw" in string_temp):
+if Usage == 'Questions based on custom CSV data':
+    option = ['Reset','Upload_csv']
+    res = st.selectbox('Select the Upload_csv option:',option)
+    if res == 'Upload_csv':
+        uploaded_file = st.file_uploader("Add dataset (csv) ",type=['csv'])
+        if uploaded_file is not None:
+            st.write("File Uploaded")
+            file_name=uploaded_file.name     
+            ext=file_name.split(".")[0]
+            st.write(ext)
+            df=pd.read_csv(uploaded_file)
+            save_uploadedfile(uploaded_file)
+            col= df.columns
             try:
-                try:
-                    wget.download(openai_response(prompt))
-                    img2 = Image.open(wget.download(openai_response(prompt)))
-                    img2.show()
-                    rx = 'Image returned'
-                    g_sheet_log(mytext, rx)
-                except:
-                    urllib.request.urlretrieve(openai_response(prompt),"img_ret.png")
-                    img = Image.open("img_ret.png")
-                    img.show()
-                    rx = 'Image returned'
-                    g_sheet_log(mytext, rx)
+                columns = str((df.columns).tolist())
+                column = clean(columns)
+                st.write('Columns:' )
+                st.text(col)
             except:
-                # 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.
-                            rx = 'Image returned'
-                            g_sheet_log(mytext, rx)
-                            
-            # 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)
-            ry = 'Youtube link and video returned'
-            g_sheet_log(mytext, ry)
-
-        elif ("don't" in string_temp or "internet" in string_temp):
-            st.write('searching internet ')
-            search_internet(question)
-            rz = 'Internet result returned'
-            g_sheet_log(mytext, rz)
-
-        else:
-            st.write(string_temp)
-            g_sheet_log(mytext, string_temp)
+                pass
     
-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();
-        """))
+        temp = st.slider('Temperature: ', 0.0, 1.0, 0.0)
+        
+        if st.checkbox('Use Prompt'):
+            with st.form(key='columns_in_form2'):
+                col3, col4 = st.columns(2)
+            with col3:
+                userPrompt = st.text_area("Input Prompt",'Enter Natural Language Query')
+                submitButton = st.form_submit_button(label = 'Submit')
+                if submitButton:
+                    try:
+                        col_p ="Create SQL statement from instruction.  "+ext+" " " (" + column +")." +" Request:" + userPrompt  + "SQL statement:" 
+                        result = gpt3(col_p)
+                    except:    
+                        results = gpt3(userPrompt)
+                    st.success('loaded')
+            with col4:
+                try:
+                    sqlOutput = st.text_area('SQL Query', value=gpt3(col_p))
+                    warning(sqlOutput) 
+                    cars=pd.read_csv('cars.csv')
+                    result_tab2=ps.sqldf(sqlOutput)
+                    st.write(result_tab2)
+                    with open("fewshot_matplot.txt", "r") as file:
+                                text_plot = file.read()
+    
+                    result_tab = result_tab2.reset_index(drop=True)
+                    result_tab_string = result_tab.to_string()
+                    gr_prompt = text_plot + userPrompt + result_tab_string + "Plot graph for: " 
+                    
+                    if len(gr_prompt) > 4097:
+                        st.write('OVERWHELMING DATA!!! You have given me more than 4097 tokens! ^_^')
+                        st.write('As of today, the NLP model text-davinci-003 that I run on takes in inputs that have less than 4097 tokens. Kindly retry ^_^')
+                    
+                    elif len(result_tab2.columns) < 2:
+                        st.write("I need more data to conduct analysis and provide visualizations for you... ^_^") 
+    
+                    else:
+                        st.success("Plotting...")
+                        response_graph = openai.Completion.create(
+                                engine="text-davinci-003",
+                                prompt = gr_prompt,
+                                max_tokens=1024,
+                                n=1,
+                                stop=None,
+                                temperature=0.5,
+                            )
+                        
+                        if response_graph['choices'][0]['text'] != "":    
+                            print(response_graph['choices'][0]['text'])
+                            exec(response_graph['choices'][0]['text'])
+                        
+                        else:
+                            print('Retry! Graph could not be plotted *_*')
+    
+                except:
+                    pass
     
-    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")
+elif Usage == 'Random Questions':
+    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 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
+              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):
-                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 
-            
+                try:
+                    try:
+                        wget.download(openai_response(prompt))
+                        img2 = Image.open(wget.download(openai_response(prompt)))
+                        img2.show()
+                        rx = 'Image returned'
+                        g_sheet_log(mytext, rx)
+                    except:
+                        urllib.request.urlretrieve(openai_response(prompt),"img_ret.png")
+                        img = Image.open("img_ret.png")
+                        img.show()
+                        rx = 'Image returned'
+                        g_sheet_log(mytext, rx)
+                except:
+                    # 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.
+                                rx = 'Image returned'
+                                g_sheet_log(mytext, rx)
+                                
+                # 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(question)
+                s = Search(mytext)
                 search_res = s.results
                 first_vid = search_res[0]
                 print(first_vid)
@@ -351,9 +382,102 @@ elif Input_type == 'SPEECH':
                 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*')
+                ry = 'Youtube link and video returned'
+                g_sheet_log(mytext, ry)
+    
+            elif ("don't" in string_temp or "internet" in string_temp):
+                st.write('searching internet ')
                 search_internet(question)
+                rz = 'Internet result returned'
+                g_sheet_log(mytext, rz)
+    
             else:
                 st.write(string_temp)
+                g_sheet_log(mytext, 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)