File size: 3,215 Bytes
2f2195a
e1a6930
 
cdeeed6
2f2195a
53d5dd8
d8f2ec7
fbb8c61
 
d8f2ec7
e1a6930
 
d8f2ec7
afa5fdc
fbb8c61
90b2246
 
 
 
afa5fdc
fbb8c61
 
 
 
 
 
 
 
 
 
 
 
e1a6930
53d5dd8
8ad245f
d35fe98
afa5fdc
 
fbb8c61
e1a6930
b05c242
fbb8c61
 
 
73afbb1
fbb8c61
53d5dd8
 
 
 
 
 
 
e1a6930
 
 
 
cca1790
 
 
 
 
 
 
7fa11aa
cca1790
 
 
 
 
 
 
 
 
 
fbb8c61
cca1790
fbb8c61
cca1790
 
 
 
 
 
 
fbb8c61
cca1790
 
 
 
 
b18c9e2
723cce8
cca1790
 
 
afa5fdc
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
import gradio as gr
import plotly.graph_objects as go
import numpy as np
from src.dataloading import get_leaderboard_models_cached, get_leaderboard_datasets

def create_heatmap(selected_models, selected_dataset):
    if not selected_models or not selected_dataset:
        return None  # Return None to hide the plot if inputs are missing

    # Generate random similarity matrix
    size = len(selected_models)
    similarities = np.random.rand(size, size)
    similarities = (similarities + similarities.T) / 2  # Make symmetric
    similarities = np.round(similarities, 2)  # Round for clarity

    print(f"Generated heatmap with {len(selected_models)} models")
    print("Sample coordinates:", selected_models[:2])
    print("Sample similarity value:", similarities[0][0])

    # Create the heatmap figure
    fig = go.Figure(data=go.Heatmap(
        z=similarities,
        x=selected_models,
        y=selected_models,
        colorscale='Viridis',
        zmin=0,
        zmax=1,
        text=similarities,
        hoverinfo="text"
    ))

    # Update layout with axis titles and margins
    fig.update_layout(
        title=f"Similarity Matrix for {selected_dataset}",
        xaxis_title="Models",
        yaxis_title="Models",
        width=800 + 20 * len(selected_models),
        height=800 + 20 * len(selected_models),
        margin=dict(b=100, l=100)
    )

    # Force the axes to use category types so that model names appear
    fig.update_xaxes(type="category", tickangle=45, automargin=True)
    fig.update_yaxes(type="category", automargin=True)

    return fig  # Return only the figure

def validate_inputs(selected_models, selected_dataset):
    if not selected_models:
        raise gr.Error("Please select at least one model!")
    if not selected_dataset:
        raise gr.Error("Please select a dataset!")

with gr.Blocks(title="LLM Similarity Analyzer") as demo:
    gr.Markdown("## Model Similarity Comparison Tool")
    
    with gr.Row():
        dataset_dropdown = gr.Dropdown(
            choices=get_leaderboard_datasets(),
            label="Select Dataset",
            filterable=True,
            interactive=True,
            info="Leaderboard benchmark datasets"
        )

        model_dropdown = gr.Dropdown(
            choices=get_leaderboard_models_cached(),
            label="Select Models",
            multiselect=True,
            filterable=True,
            allow_custom_value=False,
            info="Search and select multiple models"
        )
    
    generate_btn = gr.Button("Generate Heatmap", variant="primary")
    heatmap = gr.Plot(label="Similarity Heatmap", visible=True)
    
    # Event handling: Validate first then create the heatmap
    generate_btn.click(
        fn=validate_inputs,
        inputs=[model_dropdown, dataset_dropdown],
        queue=False
    ).then(
        fn=create_heatmap,
        inputs=[model_dropdown, dataset_dropdown],
        outputs=heatmap  # Only one output is needed
    )
    
    # Clear button
    clear_btn = gr.Button("Clear Selection")
    clear_btn.click(
        lambda: [None, None, None],
        outputs=[model_dropdown, dataset_dropdown, heatmap]
    )

if __name__ == "__main__":
    demo.launch()