Volavion/eval_temperatures_bench
收藏Hugging Face2026-03-19 更新2026-03-29 收录
下载链接:
https://hf-mirror.com/datasets/Volavion/eval_temperatures_bench
下载链接
链接失效反馈官方服务:
资源简介:
---
dataset_info:
features:
- name: run_key
dtype: large_string
- name: question_id
dtype: int64
- name: model_a
dtype: large_string
- name: model_b
dtype: large_string
- name: human_winner
dtype: large_string
- name: judge_model
dtype: large_string
- name: judge_type
dtype: large_string
- name: prompt_variant
dtype: large_string
- name: temperature
dtype: float64
- name: repeat_id
dtype: int64
- name: seed
dtype: int64
- name: which_response
dtype: large_string
- name: score_a
dtype: float64
- name: score_b
dtype: float64
- name: judge_reason_a
dtype: large_string
- name: judge_reason_b
dtype: large_string
- name: raw_output_a
dtype: large_string
- name: raw_output_b
dtype: large_string
- name: raw_output
dtype: large_string
- name: judge_reason
dtype: large_string
- name: pairwise_winner
dtype: large_string
splits:
- name: train
num_bytes: 286765777
num_examples: 179998
download_size: 118749922
dataset_size: 286765777
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---
提供机构:
Volavion



