e35d98c2cd
- Track trainset/train/val/test.jsonl (focused excerpts only, no multi-MB text) so the dataset can be used directly without the ~5h rebuild. - DATASET.md: record schema, both serializations, load snippets (plain Python + HF datasets), a text->triples fine-tuning sketch, eval notes, provenance. - .gitignore: keep only the 100s-of-MB intermediates (samples_full) ignored. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
58 lines
1.4 KiB
Plaintext
58 lines
1.4 KiB
Plaintext
# ---- Large / derived data (reproducible via build_rdf_dataset.py) ----
|
|
# Raw prospectus prose fetched from EDGAR (GBs)
|
|
data/rdf_poc/prose/
|
|
# Intermediate samples that embed the raw multi-MB prospectus text (100s of MB).
|
|
data/rdf_poc/samples.jsonl
|
|
data/rdf_poc/samples_full.jsonl
|
|
# NOTE: the training-ready files below (trainset/train/val/test, ~6 MB total)
|
|
# carry only focused excerpts and ARE committed -- see DATASET.md. Do NOT ignore
|
|
# them. (train/val/test were previously ignored; now tracked.)
|
|
# LLM match inputs/outputs (embed raw text / large)
|
|
data/rdf_poc/match_input.jsonl
|
|
data/rdf_poc/match_all.jsonl
|
|
# diagnostic / scratch files
|
|
data/rdf_poc/_*.jsonl
|
|
# prose backups from iterative fetch-strategy changes
|
|
data/rdf_poc/prose_*_bak/
|
|
# Raw SEC bulk downloads (re-downloadable from sec.gov)
|
|
data/ncen/
|
|
data/nport/
|
|
data/xbrl_rr/
|
|
|
|
# SQLite working DB
|
|
fund_data.db
|
|
fund_data.db-shm
|
|
fund_data.db-wal
|
|
|
|
# Archives
|
|
*.zip
|
|
|
|
# Python
|
|
__pycache__/
|
|
*.pyc
|
|
*.pyo
|
|
|
|
# LaTeX build artifacts
|
|
*.aux
|
|
*.log
|
|
*.out
|
|
*.toc
|
|
*.fls
|
|
*.fdb_latexmk
|
|
*.synctex.gz
|
|
|
|
# OS / editor
|
|
.DS_Store
|
|
.claude/
|
|
data/rdf_poc/samples_labelled.jsonl
|
|
data/rdf_poc/samples_full.jsonl
|
|
data/rdf_poc/labelled_partial.jsonl
|
|
data/rdf_poc/labelled_none.jsonl
|
|
data/rdf_poc/labelled_full.jsonl
|
|
data/rdf_poc/match_input_remaining.jsonl
|
|
data/rdf_poc/match_remaining*.jsonl
|
|
data/rdf_poc/match_all_clean*.jsonl
|
|
data/rdf_poc/match_input_sc.jsonl
|
|
data/rdf_poc/match_remaining.bak
|
|
data/rdf_poc/match_all.jsonl
|