README.md

metadata

pretty_name: CaseForge generated dataset
tags:
  - synthetic
  - templating
  - reproducible
license: cc-by-4.0
dataset_info:
  features:
    - name: id
      dtype: string
    - name: run_id
      dtype: string
    - name: case_type
      dtype: string
    - name: seed
      dtype: int64
    - name: question
      dtype: string
    - name: gold
      dtype: string
  splits:
    - name: train
      num_bytes: 5910639
      num_examples: 8000
    - name: validation
      num_bytes: 739560
      num_examples: 1000
    - name: test
      num_bytes: 740632
      num_examples: 1000
  download_size: 918995
  dataset_size: 7390831
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*

caseforge dataset

Run metadata

• run_id: c9655dcd543f
• code_version: 0.1.dev27+ga3f30cc39.d20260218
• git_sha: 94af0c6
• python_version: 3.13.12
• generation_timestamp: 2026-03-03T12:47:11.357275+00:00
• templates_source: https://github.com/victorlavrenko/caseforge-templates

Build options

option	value
debug	False
n_total	10000
run_id	(auto)
seed	1337
splits	0.8,0.1,0.1
templates_git	https://github.com/victorlavrenko/caseforge-templates
templates_local	(none)
templates_subdir	.

Case catalog

case_type	template_path
orl-conductive-acute	orl/conductive/acute.md
orl-ssnhl-acute	orl/ssnhl/acute.md