talon-community/PanDomain-V1.3
收藏Hugging Face2025-04-19 更新2025-11-29 收录
下载链接：
https://hf-mirror.com/datasets/talon-community/PanDomain-V1.3
下载链接
链接失效反馈官方服务：
资源简介：
---
dataset_info:
features:
- name: id
dtype: string
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 166994655
num_examples: 19940
download_size: 72972871
dataset_size: 166994655
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
license: cc-by-4.0
task_categories:
- question-answering
language:
- en
tags:
- chemistry
- biology
- code
size_categories:
- 10K<n<100K
---
PanDomain-V1 is a high-quality, fully English dataset designed for training generalist language models across all major domains. It serves as the foundational training corpus for the Talon model family, built to support broad capabilities in both reasoning and generation.
Every model sees everything.
提供机构：
talon-community
