manishiitg/human_eval

---
dataset_info:
features:
- name: type
dtype: string
- name: messages
dtype: string
- name: answer
dtype: string
- name: answer_english
dtype: string
- name: lang
dtype: string
- name: hi_answer
dtype: string
- name: mt_question
sequence: string
splits:
- name: train
num_bytes: 756440
num_examples: 734
download_size: 219666
dataset_size: 756440
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---

Evaluation on hindi and english prompts borrowed from teknimum, airoboros, https://huggingface.co/datasets/HuggingFaceH4/mt_bench_prompts, https://huggingface.co/datasets/ai4bharat/human-eval and other sources

Mainly used to evalaution on written tasks through LLM JUDGE

https://github.com/lm-sys/FastChat/blob/main/fastchat/llm_judge/README.md

原始信息汇总

数据集信息

特征

• type: 字符串类型
• messages: 字符串类型
• answer: 字符串类型
• answer_english: 字符串类型
• lang: 字符串类型
• hi_answer: 字符串类型
• mt_question: 序列字符串类型

数据分割

• train:
  • 字节数: 756440
  • 样本数: 734

数据大小

• 下载大小: 219666
• 数据集大小: 756440

配置

• default:
  • 数据文件:
    • 分割: train
    • 路径: data/train-*