import gradio as gr
from datasets import load_dataset
import pandas as pd
db = load_dataset("nicholasKluge/model-library", split='main')
db = db.to_pandas()
def display_model_information(value):
"""
This function will display the model information for the selected model
"""
# If the value is empty, return None
if value == '':
return None, None
# Get the model information
info = db.iloc[int(db[db.model_name_string == value].index.values)]
# Create the model details and model info
model_details = f"""## Model Details
- Name: {info.model_name_url}
- Model Size: {info.model_size_string}
- Dataset: {info.dataset}
- Input/Output Format: {info.data_type}
- Research Field: {info.research_field}
- Contains an Impact Assessment: {info.risks_and_limitations}
- Associated Risks: ☣️ {info.risk_types} ☣️
- Date of Publication: {info.publication_date}
- Organization: {info.organization_and_url} ({info.institution_type})
- Country/Origin: {info.country}
- License: {info.license}
- Publication: {info.paper_name_url}
"""
model_info = f"""## Description
{info.model_description}
## Organization
{info.organization_info}
"""
return model_details, model_info
with open('risks_list.md', 'rb') as f:
risk_text = f.read().decode('utf-8')[44:]
with gr.Blocks(theme='HaleyCH/HaleyCH_Theme') as demo:
gr.Markdown("""
Model Library
""")
gr.HTML("""
""")
gr.HTML(f"
The Model Library is a project that maps the risks associated with modern machine \
learning systems. Here, we assess some of the most recent and capable AI systems ever created. \
We have already mapped {len(db)} models from the AI community!
")
dropdown = gr.Dropdown(
choices=db.model_name_string.tolist(),
label="Choose a model",
info="These are the models we have already produced reports."
)
display = gr.Button(value="Display")
with gr.Row():
with gr.Column(scale=1):
model_details = gr.Markdown()
with gr.Column(scale=4):
model_info = gr.Markdown()
with gr.Accordion(label="Mapped Risks", open=False):
gr.Markdown(risk_text)
gr.HTML(f"
If you would like to add a model, read our\
documentation and submit a PR on GitHub!
