`ecg-rvq-32x`¶

Overview¶

Modality: ECG
Structure: codec
Compression ratio: 32×
Sample rate: 256 Hz
Recipe: train-ecg-rvq
Config: configs/ecg_rvq_256hz_32x_golden.yaml
Run name: ecg_rvq_256hz_32x_golden
HuggingFace: Ambiq/compressionkit-ecg-32x-v1.0

Dataset & License¶

Dataset: PTB-XL (dataset_id: ptb-xl)
License: CC BY 4.0 (open)
Notes: Auto-downloaded on first use.

The lifecycle runner pre-flights dataset availability before training (see #26 and the dataset contract).

Reproduction¶

# Single command, end-to-end.
uv run compressionkit golden run ecg-rvq-32x

# Publish the deploy package to HuggingFace (requires HF_TOKEN).
uv run compressionkit golden run ecg-rvq-32x --publish

Results land under results/ecg_rvq_256hz_32x_golden/; deploy artifacts under results/ecg_rvq_256hz_32x_golden/deploy/.

Evaluation Metrics¶

See the modality model zoo for the full metrics table:

ECG models

Each run writes quality_scorecard.json and summary.json under its results/<run>/.

Deploy Artifacts¶

Every successful run produces the canonical edge deploy package:

encoder.tflite / encoder.h — INT8 encoder.
encoder.keras — float32 Python reference encoder.
decoder.tflite / decoder.h — decoder (float32 + optional INT8).
decoder.keras — float32 Python reference decoder.
codebook.npz / codebook.h — RVQ codebook tables.
sample_data.npz — license-safe input/target/reconstruction reference frames (published to HuggingFace as sample_stimulus.npz).
model_card.json, deploy_manifest.json — metadata.

Customization Notes¶

Tweak the YAML to explore neighbouring operating points; copy the file before editing.
For new recipes, prefer the compressionkit/recipes/ package recipes as a starting point.
To resume publishing without retraining, pass --skip-train to compressionkit golden run ecg-rvq-32x --publish.

ecg-rvq-32x¶