Skip to content

api-and-cli.md

Latest commit

 

History

History
109 lines (88 loc) · 3.64 KB

api-and-cli.md

File metadata and controls

109 lines (88 loc) · 3.64 KB

CLI And API Guide

CLI Overview

casecrawler --help

Important commands:

  • generate-dataset: generate synthetic healthcare records.
  • generate-release-package: generate, benchmark, export, and verify a multimodal package.
  • verify-split-package: verify split package integrity and release readiness.
  • datasets: list, inspect, quality-check, benchmark-plan, and manage datasets.
  • reviews: queue and mark human review decisions.
  • export-dataset: export a stored dataset to one file.
  • export-dataset-splits: export train/validation/test split packages.
  • benchmark-dataset: compare a generated dataset against a reference dataset.
  • reference-datasets: list reference datasets.
  • import-reference-dataset: import Hugging Face or local references.
  • import-synthea and run-synthea: import or run Synthea output.
  • imaging-models, clinical-text-models, timeseries-models: inspect adapter contracts.

Common CLI Workflows

Generate and export:

casecrawler generate-dataset "heart failure exacerbation" --count 100 --complexity complex
casecrawler datasets quality <dataset_id>
casecrawler export-dataset --dataset-id <dataset_id> --format sft_jsonl --output train.jsonl

Generate a multimodal cohort:

casecrawler generate-dataset "pulmonary embolism" \
  --count 50 \
  --modalities structured_ehr,clinical_text,labs,vitals,time_series,imaging \
  --age-min 45 \
  --age-max 85 \
  --sexes female,male

Create a release package:

casecrawler generate-release-package "mixed acute care cohort" \
  --count 25 \
  --output-dir release-package \
  --format multimodal_jsonl \
  --max-validation-retries 2

casecrawler verify-split-package --require-multimodal-release release-package

Import references and benchmark:

casecrawler import-reference-dataset synthchex_75k --dataset-id ds-synthchex-ref --limit 100
casecrawler benchmark-dataset \
  --dataset-id <dataset_id> \
  --reference-dataset-id ds-synthchex-ref \
  --min-overall-score 0.8 \
  --min-metric-score 0.5

API Overview

Start the server:

casecrawler serve

Core endpoints:

Endpoint Method Purpose
/api/datasets/capabilities GET Modalities, export formats, release requirements, model profiles, validators
/api/datasets/generate POST Generate synthetic records
/api/datasets/release-package POST Generate and download a release package zip
/api/datasets/reference-catalog GET List reference datasets
/api/datasets/reference-import POST Import Hugging Face or local references
/api/datasets/synthea-import POST Import Synthea output
/api/datasets/{dataset_id}/quality GET Dataset quality and export readiness
/api/datasets/{dataset_id}/benchmark GET Benchmark against a reference dataset
/api/datasets/{dataset_id}/benchmark-plan GET Recommended reference readiness
/api/datasets/{dataset_id}/reference-fixtures POST Seed bundled reference fixtures
/api/datasets/{dataset_id}/export GET Stream export records
/api/datasets/{dataset_id}/export-splits GET Download split package zip
/api/datasets/{dataset_id}/reviews GET Review queue
/api/records/{record_id}/review POST Save review decision

Example release-package API call:

curl -X POST http://localhost:8000/api/datasets/release-package \
  -H 'Content-Type: application/json' \
  -o release-package.zip \
  -d '{
    "topic": "mixed acute care cohort",
    "count": 25,
    "recipe": "full_multimodal_acute_care",
    "export_format": "multimodal_jsonl",
    "max_validation_retries": 2,
    "seed": "casecrawler"
  }'