Minimal card: sync with relational-transformer repo

Card: drop the column-statistics note

Model card: neutral column-statistics note

Model card: point at the verified rt-j quickstart

Update the model card

LICENSE: CC BY 4.0, replacing CC BY-NC-SA 4.0 (issue #9)

Card: CC BY 4.0 licence (issues #9/#10)

the live step-8k checkpoint, not the swa one

single checkpoint: best swa at step 8k of the 26-09-09 plurel pretraining (mask 0, filtered task list)

Merge stanford-star/rt-p into this repo; move original PluRel-paper checkpoints under paper/

Trim leaderboard section intro line

Download section: new repo id + config.json-first flow

Update README.md

Remove no-op countDownloads metadata

Add config.json with model architecture (also enables download tracking)

Count .pt fetches as downloads

RelBench leaderboard checkpoints + README (flat layout)

Add pipeline tag and link to original Relational Transformer paper (#1)

Update README.md

Update README.md

Update README.md

Rename syn-pretrain_rdb_1024_size_4b.pt to synthetic-pretrain_rdb_1024_size_4b.pt

Rename syn-pretrain_rdb_512_size_16b.pt to synthetic-pretrain_rdb_512_size_16b.pt

Rename syn-pretrain_d_seeds_14000_2000_512_100_max_steps_128001.pt to syn-pretrain_rdb_512_size_16b.pt

Rename syn-pretrain_d_seeds_14000_2000_1024_100_max_steps_32001.pt to syn-pretrain_rdb_1024_size_4b.pt

Upload folder using huggingface_hub

initial commit