jablonkagroup/corral_score
收藏Hugging Face2026-01-12 更新2026-02-07 收录
下载链接：
https://hf-mirror.com/datasets/jablonkagroup/corral_score
下载链接
链接失效反馈官方服务：
资源简介：
---
dataset_info:
config_name: 1.0.0
features:
- name: model
dtype: string
- name: environment
dtype: string
- name: agent
dtype: string
- name: subtype
dtype: string
- name: level
dtype: string
- name: verbosity
dtype: string
- name: task_type
dtype: string
- name: prompt_tokens
dtype: int64
- name: completion_tokens
dtype: int64
- name: total_tokens
dtype: int64
- name: average_score
dtype: float64
- name: overall_success_rate
dtype: float64
- name: pass@1
dtype: float64
- name: pass@2
dtype: float64
- name: pass@3
dtype: float64
- name: pass@4
dtype: float64
- name: pass@5
dtype: float64
- name: pass^1
dtype: float64
- name: pass^2
dtype: float64
- name: pass^3
dtype: float64
- name: pass^4
dtype: float64
- name: pass^5
dtype: float64
- name: total_tasks
dtype: int64
- name: tool_verbosity
dtype: string
- name: total_tool_calls
dtype: int64
- name: successful_tool_calls
dtype: int64
- name: failed_tool_calls
dtype: int64
- name: surrendered_trials
dtype: float64
- name: total_tool_execution_duration
dtype: float64
- name: total_benchmark_duration
dtype: float64
- name: qa_score
dtype: float64
splits:
- name: train
num_bytes: 134186
num_examples: 492
download_size: 56808
dataset_size: 134186
configs:
- config_name: 1.0.0
data_files:
- split: train
path: 1.0.0/train-*
---
提供机构：
jablonkagroup
