Update app.py

app.py

@@ -1,134 +1,55 @@
-import gradio as gr
-import os 
-import json 
-import requests
+import os
 
-#Streaming endpoint 
-API_URL = "https://api.openai.com/v1/chat/completions" #os.getenv("API_URL") + "/generate_stream"
+from openai import OpenAI
+import gradio as gr
 
-#Testing with my Open AI Key 
-OPENAI_API_KEY = os.getenv("OPENAI_API_KEY") 
+api_key = os.environ.get('OPENAI_API_KEY')
+client = OpenAI(api_key=api_key)
 
 MODELS = [
     'gpt-4o',
     'gpt-4o-mini',
-    'gpt-4-turbo',
     'gpt-4',
+    'gpt-4-turbo',
     'gpt-3.5-turbo',
 ]
 
-def predict(model_name, inputs, top_p, temperature, openai_api_key, chat_counter, chatbot=[], history=[]):  #repetition_penalty, top_k
-
-    payload = {
-    "model": "gpt-3.5-turbo",
-    "messages": [{"role": "user", "content": f"{inputs}"}],
-    "temperature" : 1.0,
-    "top_p":1.0,
-    "n" : 1,
-    "stream": True,
-    "presence_penalty":0,
-    "frequency_penalty":0,
-    }
-
-    headers = {
-    "Content-Type": "application/json",
-    "Authorization": f"Bearer {openai_api_key}"
-    }
-
-    print(f"chat_counter - {chat_counter}")
-    if chat_counter != 0 :
-        messages=[]
-        for data in chatbot:
-          temp1 = {}
-          temp1["role"] = "user" 
-          temp1["content"] = data[0] 
-          temp2 = {}
-          temp2["role"] = "assistant" 
-          temp2["content"] = data[1]
-          messages.append(temp1)
-          messages.append(temp2)
-        temp3 = {}
-        temp3["role"] = "user" 
-        temp3["content"] = inputs
-        messages.append(temp3)
-        #messages
-        payload = {
-        "model": model_name,
-        "messages": messages, #[{"role": "user", "content": f"{inputs}"}],
-        "temperature" : temperature, #1.0,
-        "top_p": top_p, #1.0,
-        "n" : 1,
-        "stream": True,
-        "presence_penalty":0,
-        "frequency_penalty":0,
-        }
-
-    chat_counter+=1
-
-    history.append(inputs)
-    print(f"payload is - {payload}")
-    # make a POST request to the API endpoint using the requests.post method, passing in stream=True
-    response = requests.post(API_URL, headers=headers, json=payload, stream=True)
-    #response = requests.post(API_URL, headers=headers, json=payload, stream=True)
-    token_counter = 0 
-    partial_words = "" 
-
-    counter=0
-    for chunk in response.iter_lines():
-        #Skipping first chunk
-        if counter == 0:
-          counter+=1
-          continue
-        #counter+=1
-        # check whether each line is non-empty
-        if chunk.decode() :
-          chunk = chunk.decode()
-          # decode each line as response data is in bytes
-          if len(chunk) > 12 and "content" in json.loads(chunk[6:])['choices'][0]['delta']:
-              #if len(json.loads(chunk.decode()[6:])['choices'][0]["delta"]) == 0:
-              #  break
-              partial_words = partial_words + json.loads(chunk[6:])['choices'][0]["delta"]["content"]
-              if token_counter == 0:
-                history.append(" " + partial_words)
-              else:
-                history[-1] = partial_words
-              chat = [(history[i], history[i + 1]) for i in range(0, len(history) - 1, 2) ]  # convert to tuples of list
-              token_counter+=1
-              yield chat, history, chat_counter  # resembles {chatbot: chat, state: history}  
-                   
-
-def reset_textbox():
-    return gr.update(value='')
-
-title = """<h1 align="center">Private ChatGPT</h1>"""
-description = """Chat with OpenAI models using their official API. OpenAI <a href="https://platform.openai.com/docs/concepts">promises</a> not to train on input or output of API calls.
-"""
-                
-with gr.Blocks(css = """#col_container {width: 1000px; margin-left: auto; margin-right: auto;}
-                #chatbot {height: 520px; overflow: auto;}""") as demo:
-    gr.HTML(title)
-    gr.HTML(description)
-    with gr.Column(elem_id = "col_container"):
-        openai_api_key = gr.Textbox(type='password', label="OpenAI API key (this space does not store it)", value=OPENAI_API_KEY)
-        model_name = gr.Dropdown(label='model', choices=MODELS, value=MODELS[0], allow_custom_value=True)
-        chatbot = gr.Chatbot(elem_id='chatbot') #c
-        inputs = gr.Textbox(placeholder= "Type here!", label= "Type an input and press Enter") #t
-        state = gr.State([]) #s
-        b1 = gr.Button()
-    
-        #inputs, top_p, temperature, top_k, repetition_penalty
-        with gr.Accordion("Parameters", open=False):
-            top_p = gr.Slider( minimum=-0, maximum=1.0, value=1.0, step=0.05, interactive=True, label="Top-p (nucleus sampling)",)
-            temperature = gr.Slider( minimum=-0, maximum=5.0, value=1.0, step=0.1, interactive=True, label="Temperature",)
-            #top_k = gr.Slider( minimum=1, maximum=50, value=4, step=1, interactive=True, label="Top-k",)
-            #repetition_penalty = gr.Slider( minimum=0.1, maximum=3.0, value=1.03, step=0.01, interactive=True, label="Repetition Penalty", )
-            chat_counter = gr.Number(value=0, visible=False, precision=0)
-    
-    inputs.submit( predict, [model_name, inputs, top_p, temperature, openai_api_key, chat_counter, chatbot, state], [chatbot, state, chat_counter],)
-    b1.click( predict, [model_name, inputs, top_p, temperature, openai_api_key, chat_counter, chatbot, state], [chatbot, state, chat_counter],)
-    b1.click(reset_textbox, [], [inputs])
-    inputs.submit(reset_textbox, [], [inputs])
-                    
-    #gr.Markdown(description)
-    demo.queue().launch(debug=True)
 
+def generate(message, history, model, temperature=1.0):
+    history_openai_format = []
+    for human, assistant in history:
+        history_openai_format.append({"role": "user", "content": human})
+        history_openai_format.append({"role": "assistant", "content": assistant})
+    history_openai_format.append({"role": "user", "content": message})
+
+    response = client.chat.completions.create(model=model,
+                                              messages=history_openai_format,
+                                              temperature=temperature,
+                                              stream=True)
+
+    partial_message = ""
+    for chunk in response:
+        if chunk.choices[0].delta.content is not None:
+            partial_message = partial_message + chunk.choices[0].delta.content
+            yield partial_message
+
+
+chat_interface = gr.ChatInterface(
+    title='Private ChatGPT',
+    description='Chat with OpenAI models using their official API. OpenAI <a href="https://platform.openai.com/docs/concepts">promises</a> not to train on input or output of API calls.',
+    fn=generate,
+    additional_inputs=[
+        gr.Dropdown(label='model',
+                    choices=MODELS,
+                    value=MODELS[0],
+                    allow_custom_value=True),
+        gr.Slider(label="Temperature",
+                  minimum=0.,
+                  maximum=1.2,
+                  step=0.05,
+                  value=1.0),
+    ],
+    analytics_enabled=False,
+    show_progress='full',
+)
+chat_interface.launch()