File size: 5,883 Bytes
eba8a37 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 |
import gradio as gr
import pandas as pd
from dataset import get_dataframe
from markdown import GUIDELINES, PANEL_MARKDOWN
df = get_dataframe()
def filter_dataframe(dataframe, eval_dataset, cont_source, checkboxes):
"""
Filter the dataframe based on the provided evaluation dataset, contaminated source, and checkboxes.
Args:
dataframe (pandas.DataFrame): The input dataframe to filter.
eval_dataset (str): The evaluation dataset to filter by.
cont_source (str): The contaminated source to filter by.
checkboxes (list): The checkboxes to filter by.
Returns:
pandas.DataFrame: The filtered dataframe.
"""
if isinstance(eval_dataset, str):
dataframe = dataframe[
dataframe["Evaluation Dataset"].str.contains(eval_dataset)
]
if isinstance(cont_source, str):
dataframe = dataframe[
dataframe["Contaminated Source"].str.contains(cont_source)
]
if isinstance(checkboxes, list) and "Exclude model-based evidences" in checkboxes:
dataframe = dataframe[dataframe["Approach"] != "model-based"]
if isinstance(checkboxes, list) and "Show only contaminated" in checkboxes:
dataframe = dataframe[
(dataframe["Train Split"] > 0.0)
| (dataframe["Development Split"] > 0.0)
| (dataframe["Test Split"] > 0.0)
]
return dataframe
def filter_dataframe_corpus(*args, **kwargs) -> pd.DataFrame:
"""
Filter the dataframe for corpus contamination.
Returns:
pandas.DataFrame: The filtered dataframe for corpus contamination.
"""
# Get rows in which the column Model or corpus is equal to dataset
filtered_df = df[df["Model or corpus"] == "corpus"]
filtered_df = filtered_df.drop(columns=["Model or corpus"])
return filter_dataframe(filtered_df, *args, **kwargs)
def filter_dataframe_model(*args, **kwargs) -> pd.DataFrame:
"""
Filter the dataframe for model contamination.
Returns:
pandas.DataFrame: The filtered dataframe for model contamination.
"""
# Get rows in which the column Model or corpus is equal to dataset
filtered_df = df[df["Model or corpus"] == "model"]
filtered_df = filtered_df.drop(columns=["Model or corpus"])
return filter_dataframe(filtered_df, *args, **kwargs)
theme = gr.themes.Soft(
primary_hue="emerald",
secondary_hue="red",
text_size="sm",
spacing_size="sm",
font=[
gr.themes.GoogleFont("Poppins"),
gr.themes.GoogleFont("Poppins"),
gr.themes.GoogleFont("Poppins"),
gr.themes.GoogleFont("Poppins"),
],
).set(block_background_fill="*neutral_50", block_background_fill_dark="*neutral_950")
with gr.Blocks(
theme=theme,
title="💨 Data Contamination Report",
analytics_enabled=False,
) as demo:
gr.Markdown(PANEL_MARKDOWN)
with gr.Tab("Corpus contamination") as tab_corpus:
with gr.Row(variant="compact"):
with gr.Column():
eval_dataset_corpus = gr.Textbox(
placeholder="Evaluation dataset",
label="Evaluation dataset",
value="",
)
cont_corpora = gr.Textbox(
placeholder="Pre-training corpora",
label="Pre-training corpora",
value="",
)
with gr.Column():
checkboxes_corpus = gr.CheckboxGroup(
["Exclude model-based evidences", "Show only contaminated"],
label="Search options",
value=[],
)
filter_corpus_btn = gr.Button("Filter")
corpus_dataframe = gr.DataFrame(
value=filter_dataframe_corpus(
eval_dataset_corpus, cont_corpora, checkboxes_corpus
).style.format(precision=2),
headers=df.columns.to_list(),
datatype=[
"markdown",
"markdown",
"number",
"number",
"number",
"str",
"markdown",
"markdown",
],
)
with gr.Tab("Model contamination") as tab_model:
with gr.Row(variant="compact"):
with gr.Column():
eval_dataset_model = gr.Textbox(
placeholder="Evaluation dataset",
label="Evaluation dataset",
value="",
)
cont_model = gr.Textbox(
placeholder="Model", label="Pre-training corpora", value=""
)
with gr.Column():
checkboxes_model = gr.CheckboxGroup(
["Exclude model-based evidences", "Show only contaminated"],
label="Search options",
value=[],
)
filter_model_btn = gr.Button("Filter")
model_dataframe = gr.DataFrame(
value=filter_dataframe_model(
eval_dataset_model, cont_model, checkboxes_model
),
headers=df.columns.to_list(),
datatype=[
"markdown",
"markdown",
"number",
"number",
"number",
"str",
"markdown",
"markdown",
],
)
filter_corpus_btn.click(
filter_dataframe_corpus,
inputs=[eval_dataset_corpus, cont_corpora, checkboxes_corpus],
outputs=corpus_dataframe,
)
filter_model_btn.click(
filter_dataframe_model,
inputs=[eval_dataset_model, cont_model, checkboxes_model],
outputs=model_dataframe,
)
with gr.Tab("Guidelines") as tab_guidelines:
gr.Markdown(GUIDELINES)
demo.launch()
|