Duplicate from AI-DHD/chatGPT_voice

Co-authored-by: Alexander McLennan <AI-DHD@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: ChatGPT Voice
+emoji: 📊
+colorFrom: blue
+colorTo: blue
+sdk: gradio
+sdk_version: 3.12.0
+app_file: app.py
+pinned: false
+duplicated_from: AI-DHD/chatGPT_voice
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,196 @@
+from pyChatGPT import ChatGPT
+import gradio as gr
+import os, json
+from loguru import logger
+import random
+from transformers import pipeline
+import torch
+
+session_token = os.environ.get('SessionToken')      
+
+device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
+
+whisper_model = pipeline(
+    task="automatic-speech-recognition",
+    model="openai/whisper-large-v2",
+    chunk_length_s=30,
+    device=device,
+)
+
+all_special_ids = whisper_model.tokenizer.all_special_ids
+transcribe_token_id = all_special_ids[-5]
+translate_token_id = all_special_ids[-6]
+
+def get_api():
+    api = ChatGPT(session_token)
+    return api
+    
+def translate_or_transcribe(audio, task):
+    whisper_model.model.config.forced_decoder_ids = [[2, transcribe_token_id if task=="Transcribe in Spoken Language" else translate_token_id]]
+    text = whisper_model(audio)["text"]
+    return text
+    
+def get_response_from_chatbot(api,text):
+    if api is None:
+        return "Sorry, the chatGPT API has some issues. Please try again later"
+    try:
+      resp = api.send_message(text)    
+      api.refresh_auth()
+      # api.reset_conversation() 
+      response = resp['message']
+    except:
+      response = "Sorry, the chatGPT queue is full. Please try again later"
+    return response
+
+def chat(api,message, chat_history):      
+    out_chat = []
+    if chat_history != '':
+        out_chat = json.loads(chat_history)
+    response = get_response_from_chatbot(api,message)
+    out_chat.append((message, response))
+    chat_history = json.dumps(out_chat)
+    logger.info(f"out_chat_: {len(out_chat)}")
+    return api,out_chat, chat_history
+
+start_work = """async() => {
+    function isMobile() {
+        try {
+            document.createEvent("TouchEvent"); return true;
+        } catch(e) {
+            return false; 
+        }
+    }
+	function getClientHeight()
+	{
+	  var clientHeight=0;
+	  if(document.body.clientHeight&&document.documentElement.clientHeight) {
+		var clientHeight = (document.body.clientHeight<document.documentElement.clientHeight)?document.body.clientHeight:document.documentElement.clientHeight;
+	  } else {
+		var clientHeight = (document.body.clientHeight>document.documentElement.clientHeight)?document.body.clientHeight:document.documentElement.clientHeight;
+	  }
+	  return clientHeight;
+	}
+ 
+    function setNativeValue(element, value) {
+      const valueSetter = Object.getOwnPropertyDescriptor(element.__proto__, 'value').set;
+      const prototype = Object.getPrototypeOf(element);
+      const prototypeValueSetter = Object.getOwnPropertyDescriptor(prototype, 'value').set;
+      
+      if (valueSetter && valueSetter !== prototypeValueSetter) {
+            prototypeValueSetter.call(element, value);
+      } else {
+            valueSetter.call(element, value);
+      }
+    }
+    var gradioEl = document.querySelector('body > gradio-app').shadowRoot;
+    if (!gradioEl) {
+        gradioEl = document.querySelector('body > gradio-app');
+    }
+    
+    if (typeof window['gradioEl'] === 'undefined') {
+        window['gradioEl'] = gradioEl;
+       
+        const page1 = window['gradioEl'].querySelectorAll('#page_1')[0];
+        const page2 = window['gradioEl'].querySelectorAll('#page_2')[0]; 
+    
+        page1.style.display = "none";
+        page2.style.display = "block"; 
+
+        window['div_count'] = 0;
+        window['chat_bot'] = window['gradioEl'].querySelectorAll('#chat_bot')[0];
+        window['chat_bot1'] = window['gradioEl'].querySelectorAll('#chat_bot1')[0];   
+        chat_row = window['gradioEl'].querySelectorAll('#chat_row')[0]; 
+        prompt_row = window['gradioEl'].querySelectorAll('#prompt_row')[0]; 
+        window['chat_bot1'].children[1].textContent = '';
+        
+        clientHeight = getClientHeight();
+        new_height = (clientHeight-300) + 'px';
+        chat_row.style.height = new_height;
+        window['chat_bot'].style.height = new_height;
+        window['chat_bot'].children[2].style.height = new_height;
+        window['chat_bot1'].style.height = new_height;
+        window['chat_bot1'].children[2].style.height = new_height;
+        prompt_row.children[0].style.flex = 'auto';
+        prompt_row.children[0].style.width = '100%';
+        
+        window['checkChange'] = function checkChange() {
+            try {
+                if (window['chat_bot'].children[2].children[0].children.length > window['div_count']) {
+                    new_len = window['chat_bot'].children[2].children[0].children.length - window['div_count'];
+                    for (var i = 0; i < new_len; i++) { 
+                        new_div = window['chat_bot'].children[2].children[0].children[window['div_count'] + i].cloneNode(true);
+                        window['chat_bot1'].children[2].children[0].appendChild(new_div);
+                    }
+                    window['div_count'] = chat_bot.children[2].children[0].children.length;
+                }
+                if (window['chat_bot'].children[0].children.length > 1) {
+                     window['chat_bot1'].children[1].textContent = window['chat_bot'].children[0].children[1].textContent;
+                } else {
+                    window['chat_bot1'].children[1].textContent = '';
+                }
+              
+            } catch(e) {
+            }        
+        }
+        window['checkChange_interval'] = window.setInterval("window.checkChange()", 500);         
+    }
+   
+    return false;
+}"""
+
+
+with gr.Blocks(title='Talk to chatGPT') as demo:
+    gr.Markdown("## Talk to chatGPT with your voice in your native language ! ##")
+    gr.HTML("<p>You can duplicate this space and use your own session token: <a style='display:inline-block' href='https://huggingface.co/spaces/yizhangliu/chatGPT?duplicate=true'><img src='https://img.shields.io/badge/-Duplicate%20Space-blue?labelColor=white&style=flat&logo=data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAAAXNSR0IArs4c6QAAAP5JREFUOE+lk7FqAkEURY+ltunEgFXS2sZGIbXfEPdLlnxJyDdYB62sbbUKpLbVNhyYFzbrrA74YJlh9r079973psed0cvUD4A+4HoCjsA85X0Dfn/RBLBgBDxnQPfAEJgBY+A9gALA4tcbamSzS4xq4FOQAJgCDwV2CPKV8tZAJcAjMMkUe1vX+U+SMhfAJEHasQIWmXNN3abzDwHUrgcRGmYcgKe0bxrblHEB4E/pndMazNpSZGcsZdBlYJcEL9Afo75molJyM2FxmPgmgPqlWNLGfwZGG6UiyEvLzHYDmoPkDDiNm9JR9uboiONcBXrpY1qmgs21x1QwyZcpvxt9NS09PlsPAAAAAElFTkSuQmCC&logoWidth=14' alt='Duplicate Space'></a></p>")
+    gr.HTML("<p> Instruction on how to get session token can be seen in video <a style='display:inline-block' href='https://www.youtube.com/watch?v=TdNSj_qgdFk'><font style='color:blue;weight:bold;'>here</font></a>. Add your session token by going to settings and add under secrets. </p>")
+    with gr.Group(elem_id="page_1", visible=True) as page_1:
+        with gr.Box():            
+            with gr.Row():
+                start_button = gr.Button("Let's talk to chatGPT!", elem_id="start-btn", visible=True) 
+                start_button.click(fn=None, inputs=[], outputs=[], _js=start_work)
+                
+    with gr.Group(elem_id="page_2", visible=False) as page_2:        
+        with gr.Row(elem_id="chat_row"):
+            chatbot = gr.Chatbot(elem_id="chat_bot", visible=False).style(color_map=("green", "blue"))
+            chatbot1 = gr.Chatbot(elem_id="chat_bot1").style(color_map=("green", "blue"))    
+        with gr.Row():
+            prompt_input_audio = gr.Audio(
+                                        source="microphone",
+                                        type="filepath",
+                                        label="Record Audio Input",
+                                        
+                                    )
+            translate_btn = gr.Button("Check Whisper first ? 👍")
+            
+        whisper_task = gr.Radio(["Translate to English", "Transcribe in Spoken Language"], value="Translate to English", show_label=False)
+        with gr.Row(elem_id="prompt_row"):
+            prompt_input = gr.Textbox(lines=2, label="Input text",show_label=True)
+            chat_history = gr.Textbox(lines=4, label="prompt", visible=False)
+            submit_btn = gr.Button(value = "Send to chatGPT",elem_id="submit-btn").style(
+                    margin=True,
+                    rounded=(True, True, True, True),
+                    width=100
+                )
+            
+             
+
+        translate_btn.click(fn=translate_or_transcribe, 
+                         inputs=[prompt_input_audio,whisper_task], 
+                         outputs=prompt_input
+                        )
+        
+        api = gr.State(value=get_api())    
+        submit_btn.click(fn=chat, 
+                             inputs=[api,prompt_input, chat_history], 
+                             outputs=[api,chatbot, chat_history],
+                            )
+    gr.HTML('''
+         <p>Note: Please be aware that audio records from iOS devices will not be decoded as expected by Gradio. For the best experience, record your voice from a computer instead of your smartphone ;)</p>
+        <div class="footer">
+                    <p>Whisper Model by <a href="https://github.com/openai/whisper" style="text-decoration: underline;" target="_blank">OpenAI</a> -   
+                    <a href="https://chat.openai.com/chat" target="_blank">chatGPT</a> by <a href="https://openai.com/" style="text-decoration: underline;" target="_blank">OpenAI</a> 
+                    </p>
+        </div>
+        ''')
+
+demo.launch(debug = True)

requirements.txt

@@ -0,0 +1,6 @@
+openai
+pyChatGPT
+loguru
+--extra-index-url https://download.pytorch.org/whl/cu113
+torch
+transformers