TABULAR & TIME SERIES BENCHMARK SUITE

41 Real-World Datasets & In-Browser Benchmarks

Comprehensive performance evaluation of TabICLv2 and tabular/temporal foundation models executing 100% client-side inside Google Sheets via WebGPU tensor kernels. Zero cloud roundtrips, zero preprocessing, and native integration with LLMs via Model Context Protocol (MCP).

41Evaluation Datasets

14 / 2 / 5 / 20Bin / Multi / Reg / Forecast

100%In-Browser WebGPU

$0.00Cloud Compute Incurred

0 bytesData Exfiltrated

START WITH YOUR BUSINESS PROBLEM

Machine learning use cases by domain

Start with a business question. Each guide connects it to useful data, real benchmark results, and a Google Sheets workflow you can try.

Finance & BankingCredit risk · customer acquisition · valuation · forecasting Healthcare & Life SciencesRisk prediction · diagnosis support · patient outcomes · resource planning Operations & Workforce AnalyticsDemand prediction · employee analytics · access decisions · quality control Commerce & IndustryCustomer behaviour · retention · filtering · inventory planning Automotive & MobilityVehicle classification · valuation · maintenance · fleet decisions Public Services & RiskPolicy analysis · risk triage · fairness · resource allocation

BENCHMARK MATRIX

Benchmark Results & Dataset Directory

Precomputed, deterministic IID, non-IID, and rolling window splits evaluate in-browser foundation model latency, throughput, and out-of-sample accuracy against reference baselines.

Tabular Foundation Models (TabICLv2 vs TabPFN vs EXAONE) →Time Series Forecasting Models (Timestral vs TimesFM vs Chronos) →

Dataset	Domain	Task	Rows	Cols	Splits?Adaptive repeated 3-fold IID evaluation. Smaller datasets receive more repeats to reduce variance; every split is precomputed for identical Python and browser evaluation.Explore Split Protocol →	Score?Primary benchmark error loss: 1 - AUC for binary classification, Log-Loss for multiclass, RMSE for regression (lower is better for all tasks). Evaluated on the dataset's persisted split indices.Explore Methodology Guide →	OpenML?Top-ranked baseline score from traditional tree ensembles (XGBoost, Random Forest) on OpenML.Explore OpenML Reference →	Latency?Mean inference time per persisted evaluation split running 100% locally via WebGPU inside Chrome.Learn about Latency →
Abalone	Agriculture, Forestry & Fishing	REG	4,177	8	3×3	2.03	—	11.7s
Adult	Economics & Public Policy	BIN	48,842	14	3×3	0.0793	0.1067	273.2s
Airfoil Self Noise	UCI	REG	1,503	5	10×3	1.12	—	4.1s
Amazon Employee Access	Information Technology & Enterprise Security	BIN	32,769	9	3×3	0.1483	0.3362	129.2s
Bank Marketing	Finance & Banking	BIN	45,211	15	3×3	0.1970	0.0642	235.4s
Blood Transfusion Service Center	Healthcare & Biomedicine	BIN	748	4	20×3	0.2446	0.2435	1.1s
Breast W	Healthcare & Life Sciences	BIN	699	9	20×3	0.0052	0.0282	1.1s
Car	Automotive & Fleet Management	MULTI	1,728	6	10×3	0.0231	0.0000	1.5s
Compas Two Years	Legal & Public Safety	BIN	5,278	13	3×3	0.2701	—	5.8s
Credit G	Finance & Banking	BIN	1,000	20	10×3	0.2016	0.2219	1.4s
Diabetes	Healthcare & Biomedicine	BIN	768	8	10×3	0.1632	0.1810	1.2s
Employee Salaries	Human Resources & Workforce Analytics	REG	9,228	10	3×3	8,126	—	36.3s
Fitness Club	Fitness, Sports & Recreation	BIN	1,500	6	10×3	0.1795	—	1.3s
House Sales	Real Estate & Property Valuation	REG	21,613	20	3×3	111,524	—	62.7s
Houses	Real Estate & Urban Planning	REG	20,640	8	3×3	40,971	—	50.3s
Monks Problems 2	Data Science & Artificial Intelligence	BIN	601	6	20×3	0.0000	0.0000	1.0s
Phoneme	Speech Processing & Acoustic Engineering	BIN	5,404	5	3×3	0.0278	0.0326	5.1s
Spambase	Cybersecurity & IT Infrastructure	BIN	4,601	57	3×3	0.0075	0.0118	10.9s
Telco Customer Churn	Telecommunications & Subscription Services	BIN	7,043	20	3×3	0.1488	—	10.2s
Titanic	Maritime Safety & Actuarial Risk Analysis	BIN	1,309	11	10×3	0.1229	—	1.3s
Vehicle	Automotive Engineering & Computer Vision	MULTI	846	18	10×3	0.2401	0.0397	1.3s
BizITobs Cloud Infrastructure Telemetry	Technology & Cloud Infrastructure	FORECAST	31,968	8	288h / 20w	0.2887	—	242ms
Boomlet Smart Meter Building Energy	Energy & Utilities	FORECAST	16,384	42	60h / 20w	0.3922	—	4.5s
Boomlet Smart Meter Building Energy	Energy & Utilities	FORECAST	10,463	107	96h / 20w	0.5453	—	1.8s
Boomlet Smart Meter Building Energy	Energy & Utilities	FORECAST	16,384	59	60h / 20w	0.2921	—	900ms
Nord Pool Day-Ahead Electricity Prices	Energy & Utilities	FORECAST	52,416	4	24h / 20w	0.5136	—	139ms
Electricity Transformer Temperature	Energy & Utilities	FORECAST	139,360	8	96h / 20w	0.5180	—	447ms
Electricity Transformer Temperature	Energy & Utilities	FORECAST	34,840	8	168h / 20w	0.8199	—	430ms
Hospital Emergency Admissions	Healthcare & Epidemiology	FORECAST	13,846	2	28h / 20w	0.5541	—	188ms
Hospital Emergency Admissions	Healthcare & Life Sciences	FORECAST	1,968	2	13h / 20w	0.5804	—	45ms
Jena Climate & Weather Telemetry	Meteorology & Climate	FORECAST	8,784	22	24h / 20w	0.3316	—	1.0s
Monash High-Density Time Series Benchmark	Retail & Supply Chain	FORECAST	21,900	2	28h / 20w	0.7053	—	247ms
Global Energy Forecasting Competition (GEFCom 2014)	Energy & Utilities	FORECAST	17,520	3	168h / 20w	0.2756	—	62ms
Rohlik E-Grocery Daily Fulfillment Orders	Retail & Supply Chain	FORECAST	7,710	16	61h / 20w	0.9335	—	2.0s
Rossmann Store Daily & Weekly Sales	Retail & Supply Chain	FORECAST	1,050,330	18	48h / 20w	0.2690	—	5.9s
Rossmann Store Daily & Weekly Sales	Retail & Supply Chain	FORECAST	148,295	17	13h / 20w	0.3050	—	765ms
Solar Photovoltaic Power Output with Weather	Energy & Utilities	FORECAST	49,648	11	24h / 20w	0.6585	—	284ms
UCI Air Quality Chemical Gas Sensor Telemetry	Environmental Science & Public Policy	FORECAST	9,357	14	168h / 20w	0.7126	—	278ms
UK COVID-19 National Epidemiological Surveillance	Healthcare & Epidemiology	FORECAST	2,944	15	28h / 20w	6.4620	—	578ms
US Macroeconomic Consumption & Disposable Income	Finance & Macroeconomics	FORECAST	1,984	2	5h / 20w	3.5365	—	47ms
Global Sovereign Carbon Dioxide (CO2) Emissions	Environmental Science & Macroeconomics	FORECAST	11,460	2	5h / 20w	2.6969	—	207ms

TESTING PROTOCOL

Benchmark Methodology & Runtime Specs

How TabICLv2 in-browser foundation models are evaluated under zero-trust, client-side constraints.

Local WebGPU & WASM Runtime

Inference runs directly on device hardware using WebGPU compute shaders and WebAssembly SIMD kernels. Tabular attention passes execute locally without sending row data to an external server.

IID & Non-IID Evaluation

IID datasets use size-aware repeated 3-fold evaluation: 3, 10, or 20 repeats reduce variance where it matters most. Split indices are persisted once and shared by Python and browser runners for exact, leakage-safe comparisons.

Zero Hyperparameter Tuning

Unlike gradient-boosted trees (XGBoost/LightGBM) that require extensive grid searches and feature encodings, TabICLv2 ingests raw tabular rows in-context with zero manual preprocessing.

Deterministic Local Execution

Benchmark runs execute 100% client-side via WebGPU shader pipelines with fixed random seeds, ensuring reproducible performance within standard browser constraints.

OPEN DATASETS & MODELS

Export, Reproduce & Train Locally

All 41 demo datasets are published as versioned CuratedContainers in the Hugging Face Hub repositories carlahq/demo-tabular-benchmark-containers and carlahq/demo-timeseries-benchmark-containers, with Parquet data, task metadata, and persisted evaluation splits.

View on Hugging Face Hub Explore OpenML Benchmarks →