datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
daytrader-benchmarksbenchmarks
OpenChainBench Crypto Infrastructure Benchmarks
Daily snapshots of every public benchmark on
openchainbench.com, released as
Hive-partitioned Parquet under CC-BY-4.0.
OCB measures latency, cost, coverage and accuracy of crypto
infrastructure (RPCs, oracles, bridges, data APIs, Polymarket adapters,
Hyperliquid builders). Every snapshot here mirrors the
/api/citable,
/api/stat/<slug>,
and /api/series/<slug>
JSON feeds at the time of capture.
Latest snapshot: 2026-09-26 (captured… See the full description on the dataset page: https://huggingface.co/datasets/OpenChainBench/benchmarks.benchmarks
BitRouter Benchmarks
This dataset release contains the private-router Terminal-Bench 2.1 routing study and its audit evidence. July 2026 experiments are preserved under v0/; the current study, raw Harbor traces, run-scoped BitRouter SQLite files, sanitized production-usage backfill, reproducible joins, and Parquet tables are under terminal-bench-2.1/router-random-study/.
Main result
All valid-case metrics below use the same strict intersection of 80 tasks. “Harbor… See the full description on the dataset page: https://huggingface.co/datasets/BitRouterAI/benchmarks.mlx-benchmarks
MLX Benchmarks
Structured benchmark results for MLX-quantized and other locally-hosted
LLMs on Apple Silicon. Covers throughput, time-to-first-token, tool-calling,
code generation, reasoning, knowledge, and math suites.
Results are produced by a sweep harness that wires upstream evaluation tools
against a local vllm-mlx inference server:
EleutherAI/lm-evaluation-harness — coding, reasoning, knowledge, math
linusvwe/MLXBench — throughput and time-to-first-token
vllm… See the full description on the dataset page: https://huggingface.co/datasets/JacobPEvans/mlx-benchmarks.LEM-benchmarks
LEM-benchmarks
Canonical 8-PAC benchmark results for the Lemma model family.
This dataset is an aggregated store of per-round evaluation data produced by
lthn/LEM-Eval. Every row
represents one model's answer to one question in one round of a paired A/B
run against its unmodified base, and the dataset grows monotonically as more
workers contribute — different machines, different sampling states, different
hardware paths — which is the whole point of 8-PAC: multiple independent… See the full description on the dataset page: https://huggingface.co/datasets/lthn/LEM-benchmarks.Benchmarks_CyberSec_RedSageMCQ
Dataset Card for RedSage-MCQ
Dataset Summary
RedSage-MCQ is a large-scale, high-quality multiple-choice question (MCQ) benchmark designed to evaluate the cybersecurity knowledge, skills, and tool proficiency of Large Language Models (LLMs). It is a component of the RedSage-Bench suite introduced in the paper "RedSage: A Cybersecurity Generalist LLM".
The dataset comprises 30,000 questions derived from RedSage-Seed, a curated collection of authoritative… See the full description on the dataset page: https://huggingface.co/datasets/RISys-Lab/Benchmarks_CyberSec_RedSageMCQ.post-training-benchmarks-viewerdemo-tabular-benchmarks
📊 Carla HQ Tabular Foundation Model Benchmarks
Centralized benchmark repository of canonical tabular datasets curated for Carla HQ and TabICL (In-Context Learning foundation models for tabular data).
Each dataset is hosted as an independent subset/config with native Parquet storage, schema qualities, OpenML source links, and synchronized Google Sheets for live spreadsheet experimentation.
🚀 Quickstart & Download Options
Option 1: Using… See the full description on the dataset page: https://huggingface.co/datasets/carlahq/demo-tabular-benchmarks.m5-retail-demand-forecasting-benchmarks
M5 Retail Demand Forecasting & Inventory Risk Benchmarks
This dataset contains the heavily processed artifacts, extracted time-series features, baseline benchmarks, and model artifacts for the M5 Retail Demand Forecasting dataset.
It includes:
Over 1GB of highly engineered temporal, pricing, and calendar features.
Volatility and shortfall risk metrics for 42,840 time series.
XGBoost, Prophet, and SARIMA predictions (point + 95% intervals).
Isolation Forest anomaly detection… See the full description on the dataset page: https://huggingface.co/datasets/snchakri/m5-retail-demand-forecasting-benchmarks.llmfit-benchmarks
llmfit Real-World LLM Inference Benchmarks
An evolving dataset of real-world LLM inference measurements across consumer, workstation, datacenter, and unified-memory hardware. It combines benchmarks from an external community source with community benchmarks contributed directly to llmfit.
The initial release contains 1,501 normalized observations:
1,010 unique external-community observations from the repository's 2026-08-10 snapshot.
491 llmfit-community observations from 61… See the full description on the dataset page: https://huggingface.co/datasets/axjns/llmfit-benchmarks.multimodal-benchmarks-miniapp-review-complaint-benchmarks
App Review Complaint Benchmarks — Category Baselines + App Complaint Profiles
Complaint-topic benchmarks for 5,592 apps across 47 app-store categories, computed from real Apple App Store and Google Play review data: complaint rates by topic (crashes, ads, billing, UX…) per app vs. category baseline, plus per-topic worst-offender leaderboards.
Part of the DataForge Open Data program — full production
packages, free for academic and personal use. Canonical dataset page:… See the full description on the dataset page: https://huggingface.co/datasets/zalizedata/app-review-complaint-benchmarks.Bokeh-Benchmarks-Resultsoctoagent-benchmarks-results
E2E v3 results (Hub splits)
Layout baisbench_celltype (path prefix baisbench_celltype/)
Prepared: datasets.load_dataset(repo, 'baisbench_celltype_prepared', split="task_type_N")
Results: datasets.load_dataset(repo, 'baisbench_celltype', split="task_type_N")
Layout baisbench_codex_gpt55 (path prefix baisbench_codex_gpt55/)
Prepared: datasets.load_dataset(repo, 'baisbench_codex_gpt55_prepared', split="task_type_N")
Results: datasets.load_dataset(repo… See the full description on the dataset page: https://huggingface.co/datasets/ixprzemyslawpietrzak/octoagent-benchmarks-results.uafe-archetype-benchmarkswaveguard-benchmarks
WaveGuard Anomaly Detection Benchmarks
Curated benchmark datasets and comparison results for evaluating anomaly detection models.
Includes labeled training (normal) and test (mixed normal + anomalous) splits, plus
head-to-head comparisons between WaveGuard and traditional methods.
Benchmark Comparisons (benchmark_results/)
WaveGuard vs. IsolationForest, LOF, and OneClassSVM across 12 datasets.
Summary: WaveGuard ranked #1 on all 12 datasets by F1 score.
Dataset… See the full description on the dataset page: https://huggingface.co/datasets/emergentphysicslab/waveguard-benchmarks.SLMS-KD-Benchmarks
SLMS-KD-Benchmarks Dataset
This repository contains the SLMS-KD-Benchmarks dataset, a collection of benchmarks for evaluating smaller language models (SLMs), particularly in knowledge distillation tasks.
This dataset is a curated collection of existing datasets from Hugging Face. We have applied custom preprocessing and new train/validation/test splits to suit our benchmarking needs. We extend our sincere gratitude to the original creators for their invaluable work.… See the full description on the dataset page: https://huggingface.co/datasets/MothMalone/SLMS-KD-Benchmarks.rerandomization-benchmarks
rerandomization-benchmarks
Replication dataset for the benchmark and diagnostic analyses inGoldstein, Jerzak, Kamat & Zhu (2026), “Fast Rerandomization Using Accelerated Computing”.
Project & Paper Links
Paper: https://doi.org/10.1016/j.softx.2026.102508
Software repository: https://github.com/cjerzak/fastrerandomize-software
Package name: fastrerandomize (R)
What’s in this dataset?
The dataset contains simulation-based benchmark results… See the full description on the dataset page: https://huggingface.co/datasets/cjerzak/rerandomization-benchmarks.Mlops-Hardware-Carbon-Benchmarks
MLOps Hardware Benchmarks & Carbon Emissions
Dataset Description
This dataset contains 3,000 empirical, synthetic profiling records tracking large language model execution runs across diverse modern datacenter and consumer accelerators (including NVIDIA H100, A100, RTX 4090, and A10G). It captures token volumes, execution speeds, physical power utilization metrics, and overall computed carbon footprint weights.
Purpose and Impact
As deep learning… See the full description on the dataset page: https://huggingface.co/datasets/sohaibdevv/Mlops-Hardware-Carbon-Benchmarks.TMT-Benchmarks
TMT-Benchmarks
Benchmark dataset for the TemporalMesh Transformer (TMT) paper.
Paper: TemporalMesh Transformer: Dynamic Graph Attention with Temporal Decay and Adaptive Depth RoutingDOI: 10.5281/zenodo.20287197Author: Vigneshwar LK
Key Result
Full TMT achieves PPL 29.4 vs Vanilla 42.1 -- a 30.2% perplexity reduction while using only 48% of the compute (2.1x efficiency gain).
Dataset Subsets
This dataset contains 5 subsets, each measuring a… See the full description on the dataset page: https://huggingface.co/datasets/vigneshwar234/TMT-Benchmarks.cwm-benchmarks-dl4c-generations
CWM Benchmarks — DL4C Generations
Per-model raw generations from the v05_clean evaluation run, scored against the CWM-benchmarks DL4C task suite. Each split contains one model's 435 responses over the same set of (instance, test, side) samples.
Split / model mapping
HuggingFace forbids - in split names (^\w+(\.\w+)*$). The split name replaces - with _ in the model id; the original id is preserved in the model column of every row.
Split name
Model id (model… See the full description on the dataset page: https://huggingface.co/datasets/JetBrains-Research/cwm-benchmarks-dl4c-generations.400zr-coherent-benchmarks
400ZR Coherent Digital Twin Benchmarks
This dataset contains comprehensive benchmarks for 400ZR coherent optical transceiver digital twin implementations.
🏆 Leaderboard (Overall Score)
Rank
Method
Template
Overall Score
Physics
Optimization
Reproducibility
Standards
1
Silicon_Photonics_v12
v12
0.897
0.895
0.692
1.000
1.000
2
Enhanced_400ZR_v11
v11
0.897
0.895
0.692
1.000
1.000
3
Revolutionary_Framework_v14
v14
0.897
0.895
0.691
1.000
1.000
4… See the full description on the dataset page: https://huggingface.co/datasets/englund/400zr-coherent-benchmarks.reasoning_benchmarks_repeatedcwm-benchmarks-dl4c-benchmarkbenchmark_single_cube_v2cwm-benchmarks-dl4c-tracesqwen3-tts-0.6b-en-benchmarks-v1_metricsPredictaBoard_Benchmarks
PredictaBoard
PredictaBoard is a novel benchmark for measuring the predictability of Large Language Models.