datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
difficulty-aime_2025-generations
Generations Dataset: aime_2025
Paper: LLMs Encode Their Failures: Predicting Success from Pre-Generation ActivationsCode: GitHub
LLM-generated solutions across train/validation/test splits for multiple models.
Columns
Column
Type
Description
problem
str
Problem statement
generated_solutions
list
Generated solutions with scores
success_rate
float
Fraction of correct generations
majority_vote_is_correct
int (0/1)
Whether majority vote is correct
k… See the full description on the dataset page: https://huggingface.co/datasets/CoffeeGitta/difficulty-aime_2025-generations.pika-math-generations
PIKA MATH Generations Dataset
A comprehensive dataset of MATH problem solutions generated by different language models with various sampling parameters.
Paper (arXiv) | GitHub Repository
Dataset Description
This dataset contains code generation results from the MATH Dataset evaluated across multiple models. It was created to support the PIKA (Probe-Informed K-Aware Routing) project, which investigates how LLMs encode their own likelihood of success in their internal… See the full description on the dataset page: https://huggingface.co/datasets/CoffeeGitta/pika-math-generations.coffeescript-code-suite
CoffeeScript Code Suite
CoffeeScript Code Suite is a public code dataset built from permissively licensed open-source CoffeeScript repositories. It is designed for three practical uses:
CoffeeScript domain adaptation and continued pretraining through raw_corpus examples.
CoffeeScript completion training through completion examples.
CoffeeScript and JavaScript translation training through coffee_to_js and js_to_coffee examples.
The dataset was assembled automatically from public… See the full description on the dataset page: https://huggingface.co/datasets/NickIBrody/coffeescript-code-suite.coffee-order-zhtw
本資料集由OllaForge生成
Coffee Order Dataset (繁體中文/Traditional Chinese)
專為咖啡點餐場景設計的繁體中文多輪對話資料集，適用於訓練任務導向對話系統。
資料集描述
此資料集包含模擬咖啡店點餐場景的多輪對話，涵蓋各種點餐情境，包括：
基本點餐流程
訂單修改與取消
處理菜單外品項
處理超出限制的請求（如加兩份濃縮）
口語化表達理解
語言
繁體中文（台灣）
包含台灣口語表達（如「ㄋㄟㄋㄟ」）
資料規模
項目
數量
總對話數
2,939
平均輪數
2-6 輪
語言
繁體中文
資料格式
每筆資料為 JSONL 格式，包含 conversations 欄位：
{
"conversations": [
{
"role": "system",
"content":… See the full description on the dataset page: https://huggingface.co/datasets/renhehuang/coffee-order-zhtw.coffee-barista-dataset
☕ Coffee Barista Training Dataset
Qwen2.5-3B-Instruct 金牌咖啡师 SFT 数据集。
快速加载
from datasets import load_dataset
ds = load_dataset("ynanxiu/coffee-barista-dataset", split="train")
# 默认加载 coffee_latest.jsonl
仓库结构
coffee-barista-dataset/
├── coffee_latest.jsonl ← 最新版 (当前 v21, 1392条)
├── archive/
│ └── coffee_v21.jsonl ← v21 历史快照
│ └── ... ← 未来迭代
└── README.md
版本说明
版本
条数
说明
v21
1392
去… See the full description on the dataset page: https://huggingface.co/datasets/ynanxiu/coffee-barista-dataset.Turka-fo-making-coffee-C3G-225ml-reviewsCoffee-Making-Test-Dataset
Dataset Card for Coffee-Making-Test-Dataset
This dataset has been created with distilabel.
Dataset Summary
This dataset contains a pipeline.yaml which can be used to reproduce the pipeline that generated it in distilabel using the distilabel CLI:
distilabel pipeline run --config "https://huggingface.co/datasets/Egrigor/Coffee-Making-Test-Dataset/raw/main/pipeline.yaml"
or explore the configuration:
distilabel pipeline info --config… See the full description on the dataset page: https://huggingface.co/datasets/Egrigor/Coffee-Making-Test-Dataset.