Awesome SFT datasets - a HuggingFaceH4 Collection

HuggingFaceH4 's Collections

Scaling Test Time Compute with Open Models

Zephyr 7B Gemma

Papers We've Read

Awesome SFT datasets

Awesome feedback datasets

Awesome reward models

Awesome SFT datasets

updated Apr 12

A curated list of interesting datasets to fine-tune language models with.

bjoernp/ultrachat_de

Viewer • Updated Dec 2, 2023 • 959 • 66 • 7
openchat/openchat_sharegpt4_dataset

Updated Jul 1, 2023 • 202 • 167

Note One of the datasets behind OpenChat-3.5
imone/OpenOrca_FLAN

Viewer • Updated Dec 8, 2023 • 1.58M • 62 • 15

Note One of the datasets behind OpenChat-3.5
tiedong/goat

Viewer • Updated May 25, 2023 • 1.75M • 58 • 34

Note One of the datasets behind OpenChat-3.5, focused on arithmetic
LDJnr/LessWrong-Amplify-Instruct

Viewer • Updated Jun 3 • 663 • 64 • 44

Note One of the datasets behind OpenChat-3.5
LDJnr/Pure-Dove

Viewer • Updated Jun 3 • 3.86k • 257 • 74

Note One of the datasets behind OpenChat-3.5. Possible leakage with MT-Bench prompts
glaiveai/glaive-code-assistant

Viewer • Updated Sep 27, 2023 • 136k • 239 • 88
meta-math/MetaMathQA

Viewer • Updated Dec 21, 2023 • 395k • 7.94k • 341
OpenAssistant/oasst_top1_2023-08-25

Viewer • Updated Aug 28, 2023 • 13.6k • 293 • 62

Note One of the datasets behind OpenChat-3.5
TIGER-Lab/MathInstruct

Viewer • Updated May 15 • 262k • 2.81k • 257
ise-uiuc/Magicoder-Evol-Instruct-110K

Viewer • Updated Dec 28, 2023 • 111k • 1.56k • 152
ise-uiuc/Magicoder-OSS-Instruct-75K

Viewer • Updated Dec 4, 2023 • 75.2k • 1.7k • 140
GAIR/lima

Viewer • Updated Jun 8, 2023 • 1.33k • 1.28k • 415
allenai/tulu-v2-sft-mixture

Viewer • Updated May 24 • 326k • 1.5k • 119

Note Some subsets are NC
glaiveai/glaive-function-calling-v2

Viewer • Updated Sep 27, 2023 • 113k • 601 • 401
glaiveai/glaive-code-assistant-v2

Viewer • Updated Apr 4 • 215k • 63 • 44
openai/gsm8k

Viewer • Updated Jan 4 • 17.6k • 168k • 458
jondurbin/airoboros-3.1

Viewer • Updated Dec 26, 2023 • 59.3k • 84 • 55
teknium/OpenHermes-2.5

Viewer • Updated Apr 15 • 1M • 2.44k • 691
Open-Orca/SlimOrca

Viewer • Updated Oct 12, 2023 • 518k • 576 • 270
lmsys/lmsys-chat-1m

Viewer • Updated Jul 27 • 1M • 4.47k • 617
garage-bAInd/Open-Platypus

Viewer • Updated Jan 24 • 24.9k • 3.23k • 374
LDJnr/Puffin

Viewer • Updated Jun 7 • 3k • 643 • 92
nampdn-ai/tiny-codes

Viewer • Updated Sep 30, 2023 • 1.63M • 316 • 234
shahules786/orca-chat

Viewer • Updated Jul 25, 2023 • 198k • 97 • 111
lmsys/chatbot_arena_conversations

Viewer • Updated Sep 30, 2023 • 33k • 1.24k • 355
cognitivecomputations/dolphin

Viewer • Updated Dec 18, 2023 • 3.73M • 440 • 405
pankajmathur/WizardLM_Orca

Viewer • Updated Jun 26, 2023 • 55k • 52 • 70
nomic-ai/gpt4all-j-prompt-generations

Viewer • Updated Apr 24, 2023 • 809k • 162 • 218
codeparrot/apps

Viewer • Updated Oct 20, 2022 • 20k • 4.98k • 132
allenai/WildChat

Viewer • Updated Oct 17 • 529k • 642 • 129
allenai/WildChat-nontoxic

Viewer • Updated May 6 • 530k • 36 • 23
LDJnr/Capybara

Viewer • Updated Jun 7 • 16k • 206 • 230
OpenAssistant/oasst2

Viewer • Updated Jan 11 • 135k • 1.3k • 219

Note The final version of the OpenAssistant dataset, consisting of 130k messages.
SebastianBodza/LeetCode_UserSubmissions

Viewer • Updated Jan 3 • 138k • 11 • 12
nuprl/CanItEdit

Viewer • Updated Mar 19 • 105 • 612 • 12
totally-not-an-llm/EverythingLM-data-V3

Viewer • Updated Sep 11, 2023 • 1.07k • 52 • 31
OpenMathInstruct-1: A 1.8 Million Math Instruction Tuning Dataset

Paper • 2402.10176 • Published Feb 15 • 36
microsoft/orca-math-word-problems-200k

Viewer • Updated Mar 4 • 200k • 1.14k • 423
qintongli/GSM-Plus-v0

Viewer • Updated Mar 4 • 10.6k • 47 • 6
Vezora/Tested-143k-Python-Alpaca

Viewer • Updated Mar 23 • 143k • 432 • 37