openmhc-toto-fc

OpenMHC Forecasting — Toto

Track 3 (forecasting) reference checkpoint for the MyHeartCounts / OpenMHC wearable-health benchmark.

This checkpoint is a Toto model. Toto (Time-series Optimized Transformer for Observability) is Datadog's decoder-only time-series foundation model with a probabilistic output head.

Fine-tuned (full) from the pretrained Datadog/Toto-Open-Base-1.0 base on the MHC training split.

• Checkpoint format: PyTorch Lightning checkpoint (model.ckpt)
• Forecasting task: 24-hour-ahead, 19 sensor channels, hourly resolution.

Intended use

Intended use: research, methodological comparison, and reproduction of the OpenMHC benchmark.

Out of scope: diagnosis, screening, prognosis, prospective risk prediction, clinical decision support, or any other medical decision-making. OpenMHC's medical condition labels are self-reported and have no diagnosis dates, and benchmark performance does not establish clinical utility. Any diagnostic, prognostic, or screening claim requires independent validation using temporally aligned, clinically adjudicated outcomes in the target population.

Model & implementation

• Toto (official implementation)
• Base model Datadog/Toto-Open-Base-1.0
• Paper: Toto: Time Series Optimized Transformer for Observability (Cohen et al., 2024)

Usage

import openmhc
from openmhc.forecasters import TotoForecaster

# pip install "openmhc[toto]"
fc = TotoForecaster.from_release("hf://MyHeartCounts/[email protected]")
results = openmhc.evaluate_forecasting(fc, version="full")

The same bundle also loads in the evaluation harness via model.release_dir=hf://MyHeartCounts/[email protected]. See openmhc_manifest.json for provenance (training run, base model, fine-tuning details) and architecture metadata.

License

Released under the BigScience OpenRAIL-M license. The use restrictions in its Attachment A apply to this checkpoint and any derivatives; among other things, they prohibit providing medical advice and medical results interpretation.

This checkpoint is fine-tuned from Datadog/Toto-Open-Base-1.0, which is licensed under Apache-2.0; its license and notices continue to apply to the base model.

Citation

If you use this checkpoint, please cite the OpenMHC benchmark and the original Toto work (linked above).