Jun 15 21:24 Re-extract: full statement+proof, normalized schema, provenance 21:08 Re-extract: full statement+proof, normalized schema, provenance Jun 10 05:18 Re-extract: full statement+proof, normalized schema, provenance 01:52 Re-extract: full statement+proof, normalized schema, provenance 00:04 Re-extract: full statement+proof, normalized schema, provenance Jun 9 20:41 Re-extract: full statement+proof, normalized schema, provenance 20:25 Re-extract: full statement+proof, normalized schema, provenance 19:52 Re-extract: full statement+proof, normalized schema, provenance 19:38 Re-extract: full statement+proof, normalized schema, provenance 05:41 Re-extract: full statement+proof, normalized schema, provenance 05:28 Re-extract: full statement+proof, normalized schema, provenance 03:34 Re-extract: full statement+proof, normalized schema, provenance Jun 8 22:08 Re-extract: full statement+proof, normalized schema, provenance 21:05 Re-extract: full statement+proof, normalized schema, provenance 20:54 Re-extract: full statement+proof, normalized schema, provenance 18:33 Re-extract: full statement+proof, normalized schema, provenance 16:08 Re-extract: full statement+proof, normalized schema, provenance Jan 13 23:31 Add placeholder docstring column for schema consistency - Added empty string docstring column as placeholder
- Ensures compatibility when training across all prover datasets
- Real docstrings to be extracted and added in future update
- Prevents training errors from schema mismatch Jan 10 14:07 Initial upload: 9,572 declarations from Dafny