Datasets:

IO-AIR
/

filterme-sft

Tasks:

Question Answering

Text Generation

Modalities:

Formats:

Languages:

Size:

Tags:

instruction-tuning

reverse-engineering

Libraries:

License:

Dataset card Data Studio Files Files and versions

5.65 GB

Ctrl+K

Ctrl+K

1 contributor

History: 75 commits

IO-AIR's picture

bigbuild v1.8: shard svg_code (8000 rows)

db3e9d3 verified about 1 hour ago

build
bigbuild v1.8: shard svg_code (8000 rows) about 1 hour ago
shards
bigbuild v1.7: shard longctx_forge (451 rows) about 12 hours ago
state
bigbuild v1.7 finals: state/qa_report about 11 hours ago
.gitattributes

2.5 kB
initial commit about 14 hours ago
DATASET_CARD.md

5.07 kB
v1.7 card set rebuilt from build state: polished README (version history + QA + license ledger), DATASET_CARD appendix, ENSEMBLE_CARD; restore root qa_report + state/; drop junk train.jsonl about 11 hours ago
ENSEMBLE_CARD.json

7.91 kB
v1.7 card set rebuilt from build state: polished README (version history + QA + license ledger), DATASET_CARD appendix, ENSEMBLE_CARD; restore root qa_report + state/; drop junk train.jsonl about 11 hours ago
README.md

11.5 kB
v1.7 card: legacy trainable-flag note + schema key variants documented about 11 hours ago
images.tar.gz

72.4 MB
xet

bigbuild v1.7 finals: images about 11 hours ago
manifest.jsonl.gz

4.05 MB
xet

bigbuild v1.7 finals: manifest about 11 hours ago
qa_report.json

5.65 kB
v1.7 card set rebuilt from build state: polished README (version history + QA + license ledger), DATASET_CARD appendix, ENSEMBLE_CARD; restore root qa_report + state/; drop junk train.jsonl about 11 hours ago
train.jsonl.gz

1.42 GB
xet

bigbuild v1.7 finals: train.jsonl.gz (165,545 rows / 2,004,534,369 gpt2) about 11 hours ago