README.md

metadata

configs:
  - config_name: default
    data_files:
      - split: hfopenllm_v2
        path: data/hfopenllm_v2.parquet
      - split: livecodebenchpro
        path: data/livecodebenchpro.parquet

Every Eval Ever Dataset

Evaluation results from various AI model leaderboards.

Usage

from datasets import load_dataset

# Load specific leaderboard
dataset = load_dataset("deepmage121/eee_test", split="hfopenllm_v2")

# Load all
dataset = load_dataset("deepmage121/eee_test")

Available Leaderboards (Splits)

• hfopenllm_v2
• livecodebenchpro

Schema

• model_name, model_id, model_developer: Model information
• evaluation_source_name: Leaderboard name
• evaluation_results: JSON string with all metrics
• Additional metadata for reproducibility

Auto-updated via GitHub Actions.