datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
snips_built_in_intents
Dataset Card for Snips Built In Intents
Dataset Summary
Snips' built in intents dataset was initially used to compare different voice assistants and released as a public dataset hosted at
https://github.com/sonos/nlu-benchmark in folder 2016-12-built-in-intents. The dataset contains 328 utterances over 10 intent classes.
A related Medium post is https://medium.com/snips-ai/benchmarking-natural-language-understanding-systems-d35be6ce568d.
Supported Tasks and… See the full description on the dataset page: https://huggingface.co/datasets/sonos-nlu-benchmark/snips_built_in_intents.IntentGrasp
IntentGrasp: A Comprehensive Benchmark for Intent Understanding
Paper: https://arxiv.org/abs/2605.06832
Authors: Yuwei Yin, Chuyuan Li, Giuseppe Carenini
Institute: UBC NLP Group, Department of Computer Science, University of British Columbia
Keywords: Intent Understanding, Dataset, Benchmark, LLM, Evaluation, Intentional Fine-Tuning
Abstract:
Accurately understanding the intent behind speech, conversation, and writing is crucial to the development of helpful Large Language… See the full description on the dataset page: https://huggingface.co/datasets/yuweiyin/IntentGrasp.Movens-Intent Movens-Intent
An omni-modal benchmark for evaluating intent-to-humanoid motion generation
Evaluation-only split. Movens-Intent contains fixed benchmark samples selected from the Movens training-data pool. It is intended to evaluate models that were not trained on these exact samples. Any overlap with a model's training data must be disclosed.
Overview
Movens-Intent evaluates whether a humanoid motion model can follow intent expressed through four input… See the full description on the dataset page: https://huggingface.co/datasets/wendell0218/Movens-Intent.peak-intent-50slurp_slu_intentmassive_intentfunction-calling-intent-eval-v1This dataset contains the intent evaluation of fw function calling mode vs GPT-4. The dataset contains both
fw model responses under completion
GPT-4 model responses under previous_completion
GPT-4 acts as a teach and is given the following instructions.
GPT-4 teacher respones are stored under
validation_result
completion_reason/completion_score - GPT-4's reason for giving completion_score to the fw function calling model.
previous_completion_reason/previous_completion_score - GPT-4's… See the full description on the dataset page: https://huggingface.co/datasets/fireworks-ai/function-calling-intent-eval-v1.llm-eval-massive_intentegoobject-intention
EgoObject Intention Dataset
Egocentric images with human intention annotations for object interaction.
Fields
Field
Description
id
Sample ID
image
Egocentric view image (1920×1080)
target_category
Target object category (e.g., "sink", "charger")
bbox
Bounding box [x, y, w, h] in COCO format
scene_reasoning
Scene context description
intention_1/2/3
Three plausible interaction intentions
BBox Format
COCO format: [x, y, width, height]… See the full description on the dataset page: https://huggingface.co/datasets/Nanase1234/egoobject-intention.call-transcript-intent-data-v2
Call Transcript Intent Dataset
Multimodal Hindi/Hinglish customer utterance dataset for loan/EMI/payment call intent classification.
Dataset Summary
Metric
Value
Total examples
139,348
Total audio duration
51.04 h
Number of intents
17
Split Statistics
Split
Examples
Duration
Hours
train
126,848
2755.14 min
45.92 h
validation
10,000
219.03 min
3.65 h
eval
2,500
88.37 min
1.47 h
Class Distribution… See the full description on the dataset page: https://huggingface.co/datasets/kapturecx/call-transcript-intent-data-v2.massive-intent-vie-classification
MassiveIntent_vie_Classification
Deduplicated copy of kornwtp/massive-intent-vie-classification.
Splits
split
rows
test
2,935
train
11,126
validation
2,020
shopping_intent
Dataset Card for "shopping_intent"
More Information needed
medical-intent-audio-datasetcoco-outdoor-intention
COCO Outdoor Intention Dataset
Outdoor/street scene images with human intention annotations for object interaction.
Dataset Description
This dataset is derived from COCO 2017, focusing on 27 outdoor object categories (vehicles, street furniture, sports equipment, etc.).
Each sample contains an outdoor scene image, a target object, and 3 GPT-generated plausible human intentions for interacting with that object.
Splits
Split
Images/Queries
Source… See the full description on the dataset page: https://huggingface.co/datasets/Nanase1234/coco-outdoor-intention.massive-intent-ind-classification
MassiveIntent_ind_Classification
Deduplicated copy of kornwtp/massive-intent-ind-classification.
Splits
split
rows
test
2,913
train
10,998
validation
2,008
slurp_slu_intent_with_transcriptionmassive-intent-tha-classification
MassiveIntent_tha_Classification
Deduplicated copy of kornwtp/massive-intent-tha-classification.
Splits
split
rows
test
2,906
train
10,912
validation
2,004
task607_sbic_intentional_offense_binary_classification
Dataset Card for Natural Instructions (https://github.com/allenai/natural-instructions) Task: task607_sbic_intentional_offense_binary_classification
Additional Information
Citation Information
The following paper introduces the corpus in detail. If you use the corpus in published work, please cite it:
@misc{wang2022supernaturalinstructionsgeneralizationdeclarativeinstructions,
title={Super-NaturalInstructions: Generalization via Declarative Instructions on… See the full description on the dataset page: https://huggingface.co/datasets/Lots-of-LoRAs/task607_sbic_intentional_offense_binary_classification.telecom-intent-config-sft-10k
Telecom Intent→Config SFT Dataset (10K)
The first open SFT dataset for training LLMs to translate natural language network intents into structured 5G/6G configurations.
This dataset addresses the #1 gap identified in the telecom LLM research landscape: there is no public training dataset for intent-to-policy translation. All existing telecom datasets (TeleQnA, ORANBench-13K, 6G-Bench) are MCQ evaluation benchmarks — not instruction-following format. This dataset fills that gap.… See the full description on the dataset page: https://huggingface.co/datasets/nraptisss/telecom-intent-config-sft-10k.massive-intent-fil-classification
MassiveIntent_fil_Classification
Deduplicated copy of kornwtp/massive-intent-fil-classification.
Splits
split
rows
test
2,943
train
11,173
validation
2,014
Research-Intent-Curated
ethicalabs/Research-Intent-Curated
▶️ Watch the Video
A curated, multi-source dataset for research paper intent classification — 5-class taxonomy trained into
Echo-DSRN-v0.1.3-Research-Intent-CLF.
Each record pairs a paper's title and abstract with one of five research-intent labels, sourced from a tiered pipeline of gold, near-gold, and silver sources spanning PubMed, Semantic Scholar, Papers With Code, arXiv, and the OpenAIRE Graph.
OpenAIRE AI Hackathon 2026… See the full description on the dataset page: https://huggingface.co/datasets/ethicalabs/Research-Intent-Curated.massive-intent-khm-classification
MassiveIntent_khm_Classification
Deduplicated copy of kornwtp/massive-intent-khm-classification.
Splits
split
rows
test
2,786
train
10,320
validation
1,932
massive-intent-zsm-classification
MassiveIntent_zsm_Classification
Deduplicated copy of kornwtp/massive-intent-zsm-classification.
Splits
split
rows
test
2,930
train
11,151
validation
2,014
TMF921-intent-to-config-25k
TMF921-Grounded Intent-to-Network-Configuration Dataset (25K)
The most comprehensive open dataset for training LLMs to translate natural language network intents into spec-compliant 5G/6G configurations.
25,000 samples (22,500 train / 2,500 test) of natural language intents paired with structured network configurations across 6 target specification layers and 8 lifecycle operations, all grounded in real telecom standards.
What Makes This Dataset Unique… See the full description on the dataset page: https://huggingface.co/datasets/nraptisss/TMF921-intent-to-config-25k.banking_intenttask456_matres_intention_classification
Dataset Card for Natural Instructions (https://github.com/allenai/natural-instructions) Task: task456_matres_intention_classification
Additional Information
Citation Information
The following paper introduces the corpus in detail. If you use the corpus in published work, please cite it:
@misc{wang2022supernaturalinstructionsgeneralizationdeclarativeinstructions,
title={Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP… See the full description on the dataset page: https://huggingface.co/datasets/Lots-of-LoRAs/task456_matres_intention_classification.massive-intent-tam-classification
MassiveIntent_tam_Classification
Deduplicated copy of kornwtp/massive-intent-tam-classification.
Splits
split
rows
test
2,946
train
11,196
validation
2,013
vira-intents-live
Dataset Card for "vira-intents-live"
More Information needed
aims-safety-intents
AIMS: Annotated Intents for Model Safety
AIMS is a human-annotated dataset of user intents for LLM safety classification. Each example pairs a difficult safety prompt with a concise, human-written description of the user's underlying intent and a human-assigned harm label. The dataset is built to study a single question: can safety classifiers be improved by modeling why a user is asking something, rather than relying on surface-level text cues?
It contains 1,724 annotated… See the full description on the dataset page: https://huggingface.co/datasets/Jazhyc/aims-safety-intents.truevoice-intent-tha-multilabelclassification
TrueVoiceIntent_tha_MultiLabelClassification
Deduplicated copy of kornwtp/truevoice-intent-tha-multilabelclassification.
Splits
split
rows
train
13,355