datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
slue-sqa-code-l22-c500
SLUE-SQA-5 HuBERT Layer-22 K=500 Discrete Units
Packed discrete-unit files for SpeechGR experiments on SLUE-SQA-5.
The units were produced with HuBERT layer 22 and a K=500 k-means model, then deduplicated with consecutive counts retained. The packed format avoids one .code and .cnt file per utterance.
Files
documents.npz: packed document/passage units
train.npz: packed train question units
validation.npz: packed validation question units
test.npz: packed test question… See the full description on the dataset page: https://huggingface.co/datasets/dodofk/slue-sqa-code-l22-c500.qrecc
QReCC Topics
This repository hosts the QReCC topics with passage relevance.
This dataset complements the QReCC retrieval setup outlined in the Apple ML-QReCC GitHub repository.
Train split has 63501 examples, and test split 16451 examples. Relevant passages are in the field "Truth_passages".
from datasets import load_dataset
def main():
# 1. Load the dataset
dataset = load_dataset("slupart/qrecc")
# 2. Show the available splits
print("Available splits:"… See the full description on the dataset page: https://huggingface.co/datasets/slupart/qrecc.autonomous-cloud-gpu-slurm-serving-suite
⚡ Autonomous Cloud GPU Infrastructure, Slurm Orchestration & Distributed Serving Suite (2026)
A Production-Grade, Verifiable Synthetic Corpus for Training Autonomous AI Supercomputing & LLM Serving Agents
⚡ Overview & Industry Problem
Operating massive AI supercomputers (thousands of NVIDIA H100/H200 and Blackwell GPUs) requires coordinating Slurm cluster schedules, topology-aware NVLink cliques, NCCL AllReduce rings, RoCE v2 lossless fabrics… See the full description on the dataset page: https://huggingface.co/datasets/beatsprom/autonomous-cloud-gpu-slurm-serving-suite.M3-SLU-sample
M3-SLU-sample
This repository provides 100-sample subsets of the M3-SLU benchmark for two tasks:
Task 1: Speaker-Attributed Question Answering
Task 2: Speaker Attribution Matching (True/False)
Each example includes multi-speaker conversational audio, instruction/question/answer pairs, transcripts, and metadata (n_speakers, data_source).The full dataset (10K+ samples) will be released later in 2025.