File size: 6,621 Bytes
d47b0e7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import token
from gradio import Interface, Textbox, Markdown, Slider, Chatbot, ChatMessage, Blocks, Row, Column, ClearButton, Button, Number,Label
from final_graph import create_final_graph, render_bettafish_report
from forum_graph import create_forum_graph
from graph_states import FinalState, ForumState
from langchain_core.callbacks import UsageMetadataCallbackHandler
import uuid 
import time

graph = None
async def init_graph():
    global graph
    if graph is None:
        graph = await create_final_graph()
    
async def process_inputs(query, refinements, max_rounds):

    state = FinalState(
            query=query,
            vector_store="",
            query_limit=int(refinements),
            max_rounds=int(max_rounds),
            current_round=1,
            step=0,
            round_digests=[],
            running_mermaid_graph="",
            debate_history=[],
            final_report=None,
            messages = [],
            references = []
        )

    await init_graph()

    messages = []
    thread_id = str(uuid.uuid4())
    usage_callback = UsageMetadataCallbackHandler()
    config = {"configurable":{"thread_id":thread_id},"recursion_limit":100,"callbacks":[usage_callback]}
    step = 0
    current_round = 1
    current_round_md = "### Current Round: {current_round}"
    start_time = time.time()
    token_used = 0
    input_tokens = 0
    output_tokens = 0
    role = "user"
    status = "Searching the web for relevant information..."
    try:
        async for parent,child in graph.astream(state,config=config,stream_mode="updates",subgraphs=True):
            print("Parent:",parent)
            print("Child:",child,end="\n\n")
            if not parent or len(parent)==0:
                continue 
            parent_node = parent[0].split(":")[0]
            curr_node = list(child.keys())[0]
            values = child.get(curr_node,{})
            if parent_node == "Research":
                if curr_node == "Query Refiner":
                    status = "Searching with refined queries..."
            elif parent_node == "Forum Graph":
                if curr_node == "Debate Supervisor":
                    step = values.get("step",step)
                    msgs = values.get("messages",[])

                    if step%2 == 0:
                        role = "assistant"
                    else:
                        role = "user"

                    if len(msgs)>0:
                        messages.append(ChatMessage(role=role,content=msgs[-1].content,metadata={"title":"Moderator Instructions"}))
                    status = "Analyzing and creating arguments..."
                elif curr_node == "Persona Agent":
                    msgs = values.get("messages",[])
                    if msgs:
                        tool_call = False
                        curr_message = msgs[-1]
                        if len(curr_message.content.strip()) == 0 and len(curr_message.tool_calls)>0:
                            tool_call = True
                        
                        if not tool_call:
                            messages.append(ChatMessage(role=role,content=curr_message.content))
                        else:
                            messages.append(ChatMessage(role=role,content="Tool Call Made. Awaiting Response...",metadata={"title":"Fetching researched data"}))
                    status = "Engaging in debate..."
                elif curr_node == "Round Digest":
                    current_round += 1
                    status = "Preparing for next round..."

            for model, usage_data in usage_callback.usage_metadata.items():
                input_tokens += usage_data.get("input_tokens",0)
                output_tokens += usage_data.get("output_tokens",0)
                token_used += usage_data.get("total_tokens",0)

            total_time_taken = round((time.time()-start_time),2)
            yield messages, "" , status, current_round_md.format(current_round=current_round),total_time_taken,input_tokens,output_tokens, token_used
        
    except Exception as e:
        print("Error during running the graph:",e)
        yield messages, "" , f" An error occurred during the graph execution: {e} ", current_round_md.format(current_round=current_round),total_time_taken,input_tokens,output_tokens, token_used
        return

    try:

        snapshot = graph.get_state(config={"configurable":{"thread_id":thread_id}})
        if snapshot.values:
            report=render_bettafish_report(snapshot.values)
        else:
            report = "No state found. Did the graph finish?"


        yield messages, report , " Graph Execution Completed! Report Generated Below. ", current_round_md.format(current_round=current_round),total_time_taken,input_tokens,output_tokens, token_used
            
    except Exception as e:
        print("Error during report generation:",e)
        yield messages, "" , f" An error occurred during report generation: {e} ", current_round_md.format(current_round=current_round),total_time_taken,input_tokens,output_tokens, token_used
        return

    

with Blocks() as demo:
    Markdown("<div style='text-align:center;'> <h1> Strategic Debate Assistant </h1>  </div>")
    with Row():
        with Column():
            with Row():
                refinements = Slider(0, 3, label="Max Query Refinements", step=1, value=0)
                max_rounds = Slider(1, 10, label="Max Debate Rounds",step=1, value=1)
            query = Textbox(label="Enter your strategic question here", lines=2)
            with Row():
                ClearButton([query,refinements,max_rounds], value="Reset")
                submit_btn = Button("Submit",variant="primary")
            with Row():
                input_tokens = Number(label=" Input Tokens Used", value=0,interactive=False)
                output_tokens = Number(label="Output Tokens Used", value=0,interactive=False)
                total_tokens = Number(label="Total Tokens Used", value=0,interactive=False)
                total_time = Number(label="Total Time Taken (s)", value=0,interactive=False)
                label = Textbox(value=" Report will be generated below upon submission", label="Status", interactive=False )
        with Column():
            current_round = Markdown(label="Current round")
            chatbot = Chatbot(label="Debate Conversation")
    report_md = Markdown(label="Strategic Report")

    submit_btn.click(process_inputs, inputs=[query, refinements, max_rounds], outputs=[chatbot, report_md,label, current_round, total_time,input_tokens,output_tokens,total_tokens])



demo.launch()