datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
so101_poker_play
so101_poker_play
This dataset was generated using a phospho starter pack.
This dataset contains a series of episodes recorded with a robot and multiple cameras. It can be directly used to train a policy using imitation learning. It's compatible with LeRobot and RLDS.
poker_handsA collection of poker hand histories, covering 11 poker variants, in the poker hand history (PHH) format.
Contents:
21,605,687 uncorrupted no-limit hold'em hands from anonymized hand history logs scraped from July 1st to July 23, 2009, uploaded by HandHQ, of varying stakes (from 25NL to 1000NL).
Absolute Poker (1,270,658)
Full Tilt Poker (1,299,503)
iPoker Network (5,996,345)
Ongame Network (1,647,765)
PokerStars (3,092,698)
PartyPoker (8,298,718)
All 83 televised hands played in the final… See the full description on the dataset page: https://huggingface.co/datasets/takara-ai/poker_hands.pokerbench-rl-dpo
PokerBench RL — Counterfactual DPO Preference Data
DPO preference pairs and raw self-play logs for training a Texas Hold'em LLM
to exploit non-GTO opponents, addressing the PokerBench paper's Future Work observation that pure SFT models lose to GPT-4-style "donking" strategies.
This dataset feeds the ianlee1996/pokerbench-qwen3-14b-lora-dpo checkpoint training.
How it was built
Self-play (5000 hands): ianlee1996/pokerbench-qwen3-14b-lora-mixed (Qwen3-14B + LoRA… See the full description on the dataset page: https://huggingface.co/datasets/ianlee1996/pokerbench-rl-dpo.metatree_pokerhand
Dataset Card for "metatree_pokerhand"
More Information needed
arena-pokerkit-hands
Arena PokerKit Hands (S8 archive — offline practice data)
A real-hand dataset of 6-max No-Limit Texas Hold'em poker played by 19 AI agents
on the dev.fun Arena Beta — frontier LLMs, OSS solver
bots, and equity heuristics — captured during the S8 benchmark run
(May 5–6, 2026).
What this is and is not. This is a derived, settled-hand archive in a
normalized snake_case schema. It is not a mirror of the live
/texas/benchmark/status.table (or /texas/pending-actions[].) shape. Use
it for… See the full description on the dataset page: https://huggingface.co/datasets/dannyobito/arena-pokerkit-hands.clbench-exploitable-poker-step3-state-action-wm-labelspoker_discard_slot2clbench-exploitable-poker-wm_summar-gemini-flash-3-1-litepoker_discard_mergeThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "so_follower",
"total_episodes": 75,
"total_frames": 33675,
"total_tasks": 15,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"fps": 30,
"splits": {
"train": "0:75"
},
"data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/tyoshino/poker_discard_merge.arena-poker-reasoned-decisions-v0
DevFun Arena Poker - Reasoned Decision Traces (v0)
1000 agent decision traces from live 6-max No-Limit Texas Hold'em on the
dev.fun AI-agent poker Arena. Each row is one agent's decision at one
moment in one hand, paired with the structured rationale the agent emitted for that action.
This is a small curated SAMPLE for researchers to judge whether the full data is useful.
Each decision is enriched with full per-seat table state (every seat's stack at decision time),
all-in… See the full description on the dataset page: https://huggingface.co/datasets/dannyobito/arena-poker-reasoned-decisions-v0.poker-trainer-gto-turnso101-poker-yellow-taskThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "so_follower",
"total_episodes": 11,
"total_frames": 5455,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"fps": 30,
"splits": {
"train": "0:11"
},
"data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/christian0420/so101-poker-yellow-task.poker_discard_slot5so101_poker_play_trimmed
so101_poker_play
This dataset was generated using a phospho starter pack.
This dataset contains a series of episodes recorded with a robot and multiple cameras. It can be directly used to train a policy using imitation learning. It's compatible with LeRobot and RLDS.
so101_poker_play_train
so101_poker_play_train
This dataset was generated using a phospho starter pack.
This dataset contains a series of episodes recorded with a robot and multiple cameras. It can be directly used to train a policy using imitation learning. It's compatible with LeRobot and RLDS.
clbench-poker-halftest-traj-qwen3-4b-next-statemobile-poker-club-data
Mobile Poker Club Data
Mobile Poker Club Data is a versioned, machine-readable research dataset covering publicly listed private-club ecosystems across PPPoker, PokerBros, ClubGG and X-Poker.
The July 2026 release contains 11 club listings across four mobile poker applications. It is intended for comparison, research, citation and structured data analysis.
Official links
Canonical dataset page: https://pppoker-catalog.com/data/mobile-poker-club-data/
Permanent… See the full description on the dataset page: https://huggingface.co/datasets/Poker-Catalog/mobile-poker-club-data.kuhn-poker-Qwen-QwQ-32B-5000poker_discard_slot1clbench-poker-halftest-traj-qwen3-4b-nowmpoker_discard_slot3pokerclbench-poker-halftest-traj-gemini-analysisclbench-poker-halftest-traj-gemini-next-stateclbench-poker-halftest-traj-qwen3-4b-future-summaryclbench-poker-halftest-traj-gemini-future-summaryclbench-poker-halftest-traj-gemini-nowmclbench-poker-halftest-traj-qwen3-4b-analysispoker_discard_slot4pokerai