datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
long-context-retrieval-training-pool
Long context retrieval training pool
Long prompts with short, checkable answers. Each row is one complete message: a task instruction,
a long body of text that hides what the question is about, and the question itself, together with
every string an answer has to contain for it to be right. The bodies run from four thousand to
thirty-two thousand tokens. Three sources, laid out twice. Train on either layer or on
both.
pool.jsonl
Every source rewritten into one… See the full description on the dataset page: https://huggingface.co/datasets/Emulated-Inc/long-context-retrieval-training-pool.tool-reasoning-sft-RESEARCH-rlvr-env-retrieval-source
Tool-Reasoning SFT — RLVR Retrieval Source Trajectories
156,381 multi-turn agentic retrieval trajectories across three document corpora, in a strict reasoning + tool-call format with validated FSM transitions. Each trajectory records a model searching a corpus, opening documents, and citing relevant passages to answer a question.
Author: Aman Priyanshu
Source Environments
Trajectories were collected against three RLVR retrieval environments from the FORMAT: Search -… See the full description on the dataset page: https://huggingface.co/datasets/AmanPriyanshu/tool-reasoning-sft-RESEARCH-rlvr-env-retrieval-source.retrieval_grounding
Dino Data Retrieval Grounding Preview
What This Dataset Is
This dataset is a focused retrieval-grounding preview built from four Dino Data capability slices:
search trigger detection
grounded search integration
history search trigger
history search integration
The goal is to train or inspect assistant behavior around two connected problems:
deciding when retrieval or history lookup is needed
generating answers that stay grounded to supplied evidence or prior thread… See the full description on the dataset page: https://huggingface.co/datasets/DinoDS/retrieval_grounding.passkey-retrieval
ReactiveAI / passkey-retrieval (Interactions Format)
Conversational (in RxLM Interactions Format) retrieval (Passkey / Needle In a Haystack type) dataset, filtered and transformed from grimulkan/passkey-retrieval
Subsets
to-4k - 3-step instruct examples with first (context) message with up to 4k tokens
to-4k-reasoning - 3-step reasoning examples with first (context) query with up to 4k tokens and all the interaction (with reasoning) up to 8k tokens
to-8k - 3-step… See the full description on the dataset page: https://huggingface.co/datasets/ReactiveAI/passkey-retrieval.stage3-synthetic-structured-retrieval
Stage 3 Synthetic Structured-Retrieval Agents
Native search-tool trajectories generated by
Qwen/Qwen3-235B-A22B-Instruct-2507 for LCLM Stage-3 agent post-training.
The default config contains only traces that passed programmatic evidence and
answer verification.
Harvest
Accepted traces: 82
Native search calls: 179
Compressed tool-observation traces: 41
Uncompressed traces: 41
Task-ID overlap between pilot and collection batch: 0
Family counts are 20 latest-state… See the full description on the dataset page: https://huggingface.co/datasets/leonli66/stage3-synthetic-structured-retrieval.Retrieval-SFT-Chat
Retrieval-Based Multi-Turn Chat SFT Synthetic Data
A year ago, we released CausalLM/Refined-Anime-Text, a thematic subset of a dataset generated using the then state-of-the-art LLMs. This dataset comprises 1 million entries synthesized through long-context models that rewrote multi-document web text inputs, intended for continued pre-training. We are pleased to note that this data has been employed in various training scenarios and in studies concerning data and internet culture.
In… See the full description on the dataset page: https://huggingface.co/datasets/CausalLM/Retrieval-SFT-Chat.MoA_Long_RetrievalRetrieval-Augmented-Question-Answering
🇰🇿 Retrieval-Augmented Question Answering in Kazakh Context
Dataset Summary
Retrieval-Augmented Question Answering (RAG), Kazakh Context is a specialized dataset designed to train Large Language Models (LLMs) to accurately answer complex questions by drawing strictly from provided external knowledge sources in the Kazakh language.
This dataset teaches models to synthesize information from multiple retrieved documents, compare concepts, and ground their answers… See the full description on the dataset page: https://huggingface.co/datasets/farabi-lab/Retrieval-Augmented-Question-Answering.API_Discovery_Retrieval_Augmented_Calling
🇰🇿 Kazakh API Discovery and Tool Retrieval Dataset
Dataset Summary
Kazakh API Discovery and Tool Retrieval Dataset is a Kazakh-language dataset designed for training and evaluating Large Language Models (LLMs) in agentic AI workflows that require API discovery, tool documentation retrieval, function calling, and multi-step tool execution.
The dataset focuses on scenarios where the assistant must first inspect or retrieve API documentation before calling the… See the full description on the dataset page: https://huggingface.co/datasets/farabi-lab/API_Discovery_Retrieval_Augmented_Calling.