File size: 6,356 Bytes
c8c8070
558e9ac
 
c8c8070
 
 
 
3a30fdd
c8c8070
 
558e9ac
 
 
 
 
 
f0321ce
c8c8070
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bca8716
c8c8070
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bca8716
c8c8070
 
bca8716
 
f8fa014
 
c8c8070
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bca8716
c8c8070
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3b68ae6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
import gradio as gr
import os
import sys
import json 
import requests

#Streaming endpoint 
API_URL = os.getenv("API_URL")

#Testing with my Open AI Key 
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

#Supress errors
def exception_handler(exception_type, exception, traceback):
    print("%s: %s" % (exception_type.__name__, exception))
sys.excepthook = exception_handler
sys.tracebacklimit = 0

def predict(inputs, top_p, temperature, chat_counter, chatbot=[], history=[]):  

    payload = {
    "model": "gpt-4",
    "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": "gpt-4",
        "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_code = f"{response}"
    if response_code.strip() != "<Response [200]>":
        #print(f"response code - {response}")
        raise Exception(f"Sorry, hitting rate limit. Please try again later. {response}")
    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, response  # resembles {chatbot: chat, state: history}  
    print(json.dumps({"chat_counter": chat_counter, "payload": payload, "partial_words": partial_words, "token_counter": token_counter, "counter": counter}))
                   

def reset_textbox():
    return gr.update(value='')

title = """<h1 align="center">🔥GPT4 with ChatCompletions API +🚀Gradio-Streaming</h1>"""
description = """Language models can be conditioned to act like dialogue agents through a conversational prompt that typically takes the form:
```
User: <utterance>
Assistant: <utterance>
User: <utterance>
Assistant: <utterance>
...
```
In this app, you can explore the outputs of a gpt-4 LLM.
"""

theme = gr.themes.Default(primary_hue="green")                

with gr.Blocks(css = """#col_container { margin-left: auto; margin-right: auto;}
                #chatbot {height: 520px; overflow: auto;}""",
              theme=theme) as demo:
    gr.HTML(title)
    gr.HTML("""<h3 align="center">🔥This Huggingface Gradio Demo provides you full access to GPT4 API (4096 token limit). 🎉🥳🎉You don't need any OPENAI API key🙌</h1>""")
    gr.HTML('''<center><a href="https://huggingface.co/spaces/ysharma/ChatGPT4?duplicate=true"><img src="https://bit.ly/3gLdBN6" alt="Duplicate Space"></a>Duplicate the Space and run securely with your OpenAI API Key</center>''')
    with gr.Column(elem_id = "col_container"):
        #GPT4 API Key is provided by Huggingface 
        #openai_api_key = gr.Textbox(type='password', label="Enter only your GPT4 OpenAI API key here")
        chatbot = gr.Chatbot(elem_id='chatbot') #c
        inputs = gr.Textbox(placeholder= "Hi there!", label= "Type an input and press Enter") #t
        state = gr.State([]) #s
        with gr.Row():
            with gr.Column(scale=7):
                b1 = gr.Button().style(full_width=True)
            with gr.Column(scale=3):
                server_status_code = gr.Textbox(label="Status code from OpenAI server", )
    
        #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, [inputs, top_p, temperature, chat_counter, chatbot, state], [chatbot, state, chat_counter, server_status_code],)  #openai_api_key
    b1.click( predict, [inputs, top_p, temperature, chat_counter, chatbot, state], [chatbot, state, chat_counter, server_status_code],)  #openai_api_key
    b1.click(reset_textbox, [], [inputs])
    inputs.submit(reset_textbox, [], [inputs])
                    
    #gr.Markdown(description)
    demo.queue(max_size=20, concurrency_count=10).launch()