datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
chat_formatted_exampleslmsys-chat-1m-formattedHealthCareMagic-100k-Chat-Format-enlmsys-chat-1m-chat-formattedchat-formattedqasper-chat-format
QASPER (Chat-Format Preparation)
This dataset is a chat-format preparation of QASPER for supervised fine-tuning (SFT).
Format
This format is commonly referred to as:
chat-format SFT data
instruction-tuning conversations
OpenAI-style messages format
Included files
train.jsonl
validation.jsonl
stats.json
prepare_qasper_unsloth.py
Source
Base dataset: allenai/qasper
Original Dataset Highlights
Original dataset: allenai/qasper
Focus:… See the full description on the dataset page: https://huggingface.co/datasets/ali5341/qasper-chat-format.t5gemma2-indonesia-chat-formatted
T5Gemma-2 Indonesian Chat & QA Dataset
A high-quality Indonesian language multi-turn conversation and reading comprehension dataset, specifically formatted for instruction tuning of sequence-to-sequence (Seq2Seq) models like T5-Gemma / T5-Gemma-2.
Dataset Description
This dataset contains over 7,400 multi-turn conversations and document-based Q&A in Bahasa Indonesia. It covers diverse topics including everyday life, technology, general knowledge, and structured… See the full description on the dataset page: https://huggingface.co/datasets/daruokta/t5gemma2-indonesia-chat-formatted.pile-chat-format-with-assistant-qwen1.5-0.5b-chatShareGPT-Unfiltered-RedPajama-Chat-format
ShareGPT unfiltered dataset in RedPajama-Chat format
This dataset was created by converting The alpaca-lora formatted ShareGPT dataset to the format required by RedPajama-Chat.
This script was used for the conversion: https://github.com/fredi-python/Alpaca2INCITE-Dataset-Converter/blob/main/convert.py
WARNING: Only the first human and gpt text of each conversation from the original dataset is included in the dataset.
The format
{"text": "<human>: hello\n<bot>: Hello!… See the full description on the dataset page: https://huggingface.co/datasets/Fredithefish/ShareGPT-Unfiltered-RedPajama-Chat-format.chat-format-index
Chat Format Index
A catalog of verified 2026 instruct chat templates (plus a few still-current late-2025 flagships), not training data. Each row is one public transformers checkpoint: the jinja the tokenizer actually used, a SHA-256 of that string, thinking/tools flags, and dummy apply_chat_template renders. v1 is 2026-08-31: 80 rows, 57 distinct template hashes, 40 inferred families.
Use it to compare templates before you load a tokenizer. It is not a replacement for the… See the full description on the dataset page: https://huggingface.co/datasets/asaverren/chat-format-index.hh-rlhf_chat_formatllama31_no_additional_chat_formatSynthetic-Japanese-Roleplay-NSFW-gpt-5-chat-5k-formatted
Synthetic-Japanese-Roleplay-NSFW-gpt-5-chat-5k-formatted
概要
OpenRouterのgpt-5-chatを用いて作成した日本語ロールプレイデータセットであるAratako/Synthetic-Japanese-Roleplay-NSFW-gpt-5-chat-5kをOpenAI messages形式に整形したデータセットです。
データの詳細については元データセットのREADMEを参照してください。
ライセンス
CC-BY-NC-SA 4.0の元配布します。
また、OpenAIの利用規約に記載のある通り、このデータを使ってOpenAIのサービスやモデルと競合するようなモデルを開発することは禁止されています。
llama3_additional_rr40k_non_delete_sft_chat_formatscitldr-chat-format
SciTLDR (Chat-Format Preparation)
This dataset is a chat-format preparation of SciTLDR for summarization SFT.
Format
This format is commonly referred to as:
chat-format SFT data
instruction-tuning conversations
OpenAI-style messages format
Included files
train.jsonl
validation.jsonl
stats.json
prepare_scitldr_unsloth.py
Source
Base dataset: allenai/scitldr
Variants used:
A
AIC
FullText
Original Dataset Highlights
Original dataset:… See the full description on the dataset page: https://huggingface.co/datasets/ali5341/scitldr-chat-format.tuluv2-expanded-150k-part_v0-chat-format-syn-knowledgevi-self-chat-sharegpt-format
🇻🇳 Vietnamese Self-Chat Dataset
This dataset is designed to enhance the model's ability to engage in multi-turn conversations with humans.
To construct this dataset, we follow a two-step process:
Step 1: Instruction Generation
We employ the methodology outlined in the Self-Instruct paper to craft a diverse set of instructions. This paper serves as a guide for aligning pretrained language models with specific instructions, providing a structured foundation for subsequent dialogue… See the full description on the dataset page: https://huggingface.co/datasets/bkai-foundation-models/vi-self-chat-sharegpt-format.oasst2_thai_top1_chat_format
Open Assistant 2 Top-1 Thai
Dataset Details
Dataset Description
A top-1 Thai dataset taken from the top scoring https://huggingface.co/datasets/OpenAssistant/oasst2 conversations. Saved in HF Chat format.
License: Apache 2.0
Script: https://github.com/wannaphong/deep_4_all/tree/main/datasets/oasst
Dataset Structure
We structure the dataset using the format commonly used as input into Hugging Face Chat Templates:
[
{'content':… See the full description on the dataset page: https://huggingface.co/datasets/pythainlp/oasst2_thai_top1_chat_format.MATH_chat_formattokenized-chat-ml-format-vicgalle-alpaca-gpt4-1koasst2_top1_chat_format
OpenAssistant TOP-1 Conversation Threads in huggingface chat format
Export of oasst2 only top 1 threads in huggingface chat format
Script
The convert script can be find here
samvaad-hi-v1-chat-formatorca-chat-chat-format
Dataset Card for "orca-chat-chat-format"
More Information needed
scifact-chat-format
SciFact (Chat-Format Preparation)
This dataset is a chat-format preparation of SciFact for supervised fine-tuning (SFT).
Format
This format is commonly referred to as:
chat-format SFT data
instruction-tuning conversations
OpenAI-style messages format
Included files
train.jsonl
validation.jsonl
stats.json
prepare_scifact_unsloth.py
Source
Base dataset: allenai/scifact
Original Dataset Highlights
Original dataset: allenai/scifact… See the full description on the dataset page: https://huggingface.co/datasets/ali5341/scifact-chat-format.arxivqa-chat-format
ArxivQA (Chat-Format Preparation)
This dataset is a multimodal chat-format preparation of ArxivQA for supervised fine-tuning (SFT).
Format
This format is commonly referred to as:
chat-format SFT data
instruction-tuning conversations
OpenAI-style messages format
multimodal chat format (image + text)
Included files
train.jsonl
validation.jsonl
stats.json
prepare_arxivqa_unsloth.py
Source
Base dataset: MMInstruction/ArxivQA
Original… See the full description on the dataset page: https://huggingface.co/datasets/ali5341/arxivqa-chat-format.longwriter-8b-context-synthesis-chat-formatMathInstruct_RedPajama_Chat_Formatreasoning-and-chat-harmony-format
Open Paws Reasoning And Conversational Finetuning Harmony Format
This dataset is part of the Open Paws initiative to develop AI training data aligned with animal liberation and advocacy principles. Created to train AI systems that understand and promote animal welfare, rights, and liberation.
Dataset Details
Dataset Type: Reasoning Data
Format: JSONL (JSON Lines)
Languages: Multilingual (primarily English)
Focus: Animal advocacy and ethical reasoning
Organization: Open… See the full description on the dataset page: https://huggingface.co/datasets/open-paws/reasoning-and-chat-harmony-format.llama3_regular_balanced_sft_chat_formatspiqa-chat-format
SPIQA (Chat-Format Preparation)
This dataset is a chat-format preparation of SPIQA for multimodal supervised fine-tuning (SFT).
Format
This format is commonly referred to as:
chat-format SFT data
instruction-tuning conversations
OpenAI-style messages format
Included files
train.jsonl
validation.jsonl
stats.json
prepare_spiqa_unsloth.py
Source
Base dataset: google/spiqa
Original Dataset Highlights
Original dataset: google/spiqa… See the full description on the dataset page: https://huggingface.co/datasets/ali5341/spiqa-chat-format.