{"id":51018271,"url":"https://github.com/raunak-agarwal/instruction-datasets","last_synced_at":"2026-07-09T14:00:39.329Z","repository":{"id":149496502,"uuid":"621900912","full_name":"raunak-agarwal/instruction-datasets","owner":"raunak-agarwal","description":"All available datasets for Instruction Tuning of Large Language Models","archived":false,"fork":false,"pushed_at":"2023-11-30T18:48:50.000Z","size":52,"stargazers_count":185,"open_issues_count":0,"forks_count":10,"subscribers_count":7,"default_branch":"main","last_synced_at":"2023-11-30T19:40:42.838Z","etag":null,"topics":["alpaca","chain-of-thought","chatgpt","dataset","gpt-3","gpt-4","gpt4all","instruction-tuning","large-language-models","multi-task-learning","open-assistant","sharegpt"],"latest_commit_sha":null,"homepage":"","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/raunak-agarwal.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2023-03-31T16:23:53.000Z","updated_at":"2023-11-29T14:07:11.000Z","dependencies_parsed_at":null,"dependency_job_id":"4cdb51cb-52df-46e7-9316-beee8b39cbe2","html_url":"https://github.com/raunak-agarwal/instruction-datasets","commit_stats":null,"previous_names":[],"tags_count":0,"template":null,"template_full_name":null,"purl":"pkg:github/raunak-agarwal/instruction-datasets","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/raunak-agarwal%2Finstruction-datasets","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/raunak-agarwal%2Finstruction-datasets/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/raunak-agarwal%2Finstruction-datasets/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/raunak-agarwal%2Finstruction-datasets/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/raunak-agarwal","download_url":"https://codeload.github.com/raunak-agarwal/instruction-datasets/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/raunak-agarwal%2Finstruction-datasets/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":35301501,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-07-09T02:00:07.329Z","response_time":57,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["alpaca","chain-of-thought","chatgpt","dataset","gpt-3","gpt-4","gpt4all","instruction-tuning","large-language-models","multi-task-learning","open-assistant","sharegpt"],"created_at":"2026-06-21T14:00:21.545Z","updated_at":"2026-07-09T14:00:39.309Z","avatar_url":"https://github.com/raunak-agarwal.png","language":null,"funding_links":[],"categories":["SFT Statistics"],"sub_categories":["General SFT"],"readme":"# Instruction Tuning Datasets\nAll available datasets for Instruction Tuning of Large Language Models\n\n### Gold standard datasets \n- P3: https://github.com/bigscience-workshop/promptsource, https://huggingface.co/datasets/bigscience/P3\n  - Collection of prompted English datasets covering a diverse set of NLP tasks\n  - 2000 prompt types over 270 datasets\n- xP3: https://huggingface.co/datasets/bigscience/xP3mt\n  - Mixture of 13 training tasks in 46 languages with prompts in 20 languages (machine-translated from English)\n- Natural Instructions v2: https://github.com/allenai/natural-instructions\n  - A benchmark of 1,616 diverse NLP tasks and their expert-written instructions, covering 76 distinct task types and 55 different languages.\n- The Flan Collection: https://github.com/google-research/FLAN/tree/main/flan/v2 \n  - superset of some of the datasets here\n  -  1836 Tasks, 15m examples \n- Open Assistant: https://huggingface.co/datasets/OpenAssistant/oasst1\n  - Human-annotated assistant-style conversation corpus consisting of 161,443 messages distributed across 66,497 conversation trees, in 35 different languages, annotated with 461,292 quality ratings\n- LIMA: 1K high-quality instructions\n  - https://huggingface.co/datasets/GAIR/lima\n- databricks-dolly-15k: https://github.com/databrickslabs/dolly/tree/master/data\n- PRESTO: https://github.com/google-research-datasets/presto\n  - 550K contextual multilingual conversations between humans and virtual assistants\n- BB3x: https://parl.ai/projects/bb3x/\n- InstructCTG: https://github.com/MichaelZhouwang/InstructCTG\n  - Framework for controlled generation https://arxiv.org/abs/2304.14293\n- CrossFit: https://github.com/INK-USC/CrossFit\n- tasksource: https://arxiv.org/abs/2301.05948\n- ExMix: https://arxiv.org/abs/2111.10952\n- InstructEval: https://github.com/declare-lab/instruct-eval\n- M3IT: https://huggingface.co/datasets/MMInstruction/M3IT\n  - https://arxiv.org/abs/2306.04387\n  - 2.4M multi-modal instances and 400 instructions across 40 tasks and 80 languages\n- MIMIC-IT: Multi-Modal In-Context Instruction Tuning : https://arxiv.org/abs/2306.05425\n- MultiInstruct: https://github.com/VT-NLP/MultiInstruct\n- COLLIE: https://github.com/princeton-nlp/Collie\n- Mind2Web: Towards a Generalist Agent for the Web https://osu-nlp-group.github.io/Mind2Web/    \n- Android in the Wild: A Large-Scale Dataset for Android Device Control: https://github.com/google-research/google-research/tree/master/android_in_the_wild\n- FLASK: Fine-grained Language Model Evaluation Based on Alignment Skill Sets https://github.com/kaistAI/FLASK\n- Safe-RLHF: https://arxiv.org/abs/2310.12773\n  - https://arxiv.org/pdf/2310.12773.pdf#https%3A//github.com/PKU-Alignment/safe-rlhf\n- HelpSteer: https://huggingface.co/datasets/nvidia/HelpSteer \n\n\n### Silver standard/Generated using LM\n\n- Self-Instruct: https://github.com/yizhongw/self-instruct\n- Unnatural Instructions: https://github.com/orhonovich/unnatural-instructions\n- Alpaca: https://huggingface.co/datasets/tatsu-lab/alpaca\n  - Alpaca-Clean: https://github.com/gururise/AlpacaDataCleaned\n- Code Alpaca: https://github.com/sahil280114/codealpaca\n- AlpacaGPT3.5Customized: https://huggingface.co/datasets/whitefox44/AlpacaGPT3.5Customized\n- GPT4All: https://github.com/nomic-ai/gpt4all\n  - GPT4All-pruned: https://huggingface.co/datasets/Nebulous/gpt4all_pruned\n- ShareGPT: https://huggingface.co/datasets/RyokoAI/ShareGPT52K\n- GPTeacher: https://github.com/teknium1/GPTeacher\n- CAMEL🐪: https://www.camel-ai.org/\n- Human ChatGPT Comparison Corpus: https://github.com/Hello-SimpleAI/chatgpt-comparison-detection\n- InstructionWild: https://github.com/XueFuzhao/InstructionWild\n- Instruction Tuning with GPT-4: https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM\n- Guanaco: https://huggingface.co/datasets/JosephusCheung/GuanacoDataset\n- The LongForm Dataset: https://github.com/akoksal/LongForm/tree/main/dataset\n  - LLM instruction generation for a diverse set of corpus samples (27,739 instructions and long text pairs)\n- UltraChat: https://huggingface.co/datasets/stingning/ultrachat\n- LLaVA Visual Instruct 150K: https://huggingface.co/datasets/liuhaotian/LLaVA-Instruct-150K\n  - GPT-generated multimodal instruction-following data\n- GPT4Tools: https://github.com/StevenGrove/GPT4Tools\n  - Instruction data to make API calls to several multi-modal models\n- LaMini-Instruction: https://huggingface.co/datasets/MBZUAI/LaMini-instruction\n  - 2.58M pairs of instructions and responses\n- Evol-Instruct 70k: https://github.com/nlpxucan/WizardLM\n- Dynosaur: https://dynosaur-it.github.io/\n- Alpaca-Farm: https://github.com/tatsu-lab/alpaca_farm\n  - https://huggingface.co/datasets/tatsu-lab/alpaca_farm\n- ign_clean_instruct_dataset_500k: https://huggingface.co/datasets/ignmilton/ign_clean_instruct_dataset_500k\n- airoboros: https://github.com/jondurbin/airoboros\n- UltraFeedback: https://huggingface.co/datasets/openbmb/UltraFeedback\n- WildChat: Corpus of 570K real-world user-ChatGPT interactions https://wildchat.allen.ai/\n- Feedback Collection: https://arxiv.org/abs/2310.08491\n  - https://huggingface.co/datasets/kaist-ai/Feedback-Collection\n\n### Preference Datasets (can be used to train the reward model)\n- HH-RLHF: https://huggingface.co/datasets/Anthropic/hh-rlhf\n  - Contains human ratings of harmfulness and helpfulness of model outputs. The dataset contains ~160K human-rated examples, where each example in this dataset consists of a pair of responses from a chatbot, one of which is preferred by humans.\n- OpenAI WebGPT: https://huggingface.co/datasets/openai/webgpt_comparisons\n  - Includes a total of around 20K comparisons where each example comprises a question, a pair of model answers, and metadata. The answers are rated by humans with a preference score.\n- OpenAI Summarization: https://huggingface.co/datasets/openai/summarize_from_feedback\n  - Contains ~93K examples, each example consists of feedback from humans regarding the summarizations generated by a model. Human evaluators chose the superior summary from two options.\n- Stanford Human Preferences Dataset (SHP): https://huggingface.co/datasets/stanfordnlp/SHP\n  - 385K collective human preferences over responses to questions/instructions in 18 different subject areas\n- Stack Exchange Preferences: https://huggingface.co/datasets/HuggingFaceH4/stack-exchange-preferences\n- SLF5K: https://huggingface.co/datasets/JeremyAlain/SLF5K\n- qa-from-hf: https://github.com/lil-lab/qa-from-hf\n- Nectar: https://huggingface.co/datasets/berkeley-nest/Nectar\n- JudgeLM-100K: https://huggingface.co/datasets/BAAI/JudgeLM-100K\n- UltraFeedback: https://huggingface.co/datasets/openbmb/UltraFeedback\n- \n\n### Misc\n- OIG: https://huggingface.co/datasets/laion/OIG\n  - Superset of some of the datasets here\n- oa_leet10k: https://huggingface.co/datasets/ehartford/oa_leet10k\n  - LeetCode problems solved in multiple programming languages\n- ProSocial Dialog: https://huggingface.co/datasets/allenai/prosocial-dialog\n- ConvoKit: https://convokit.cornell.edu/documentation/datasets.html\n- CoT-Collection: https://github.com/kaist-lklab/CoT-Collection\n- DialogStudio: https://github.com/salesforce/DialogStudio\n- Chatbot Arena Conversations https://huggingface.co/datasets/lmsys/chatbot_arena_conversations\n- lmsys 1M: https://huggingface.co/datasets/lmsys/lmsys-chat-1m\n- Conversation Chronicles: https://conversation-chronicles.github.io/\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fraunak-agarwal%2Finstruction-datasets","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fraunak-agarwal%2Finstruction-datasets","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fraunak-agarwal%2Finstruction-datasets/lists"}