|
Download README.md from dwidlee/systemone-lite-phase2: direct link, hf CLI and curl.
- Browser
- Download file 2.21 kB
-
https://huggingface.co/datasets/dwidlee/systemone-lite-phase2/resolve/main/README.md
- Command line
-
hf download hf://datasets/dwidlee/systemone-lite-phase2/README.md
-
curl -L -o README.md https://huggingface.co/datasets/dwidlee/systemone-lite-phase2/resolve/main/README.md
2.21 kB
| license: apache-2.0 | |
| pretty_name: systemone-lite Phase 2 distill | |
| task_categories: | |
| - text-classification | |
| - multiple-choice | |
| tags: | |
| - system-one | |
| - alias-ce | |
| - spatial | |
| - cloze | |
| - zero-leakage | |
| - action-v2 | |
| - staged-chess | |
| size_categories: | |
| - 100K<n<1M | |
| # systemone-lite-phase2 | |
| Typed System One distill rows (`task` / `state` / `instructions` / criteria / | |
| `label_alias`) for [`systemone-lite`](https://github.com/fritzprix/systemone-lite). | |
| ## Revision (2026-09-26) | |
| Paired with Hub model revision **action-v2-qwen** | |
| (`dwidlee/systemone-lite-0.5b`). | |
| | Change | Detail | | |
| |---|---| | |
| | Chess | `staged_v1` — piece + destination, option caps ≤8 | | |
| | Spatial | `action_v2` — legal-only actions; Connect4 drop≤3 + win_now | | |
| | Sokoban `test` | Deadlock alerts balanced 75/75 yes/no; remap_alert_prob≈0.35 | | |
| Postmortem: repo `docs/NOTE_ACTION_V2_QWEN_POSTMORTEM.md`. | |
| ## Critical: train / test hygiene | |
| **0.00%** train∩test overlap on `state_task` fingerprints | |
| (`scripts/audit_train_eval_overlap.py`). Earlier mixes had severe leakage; | |
| do not regress. | |
| | Mechanism | Detail | | |
| |---|---| | |
| | `word_games` | Disjoint TRAIN / EVAL vocabularies | | |
| | `debate_judge` | Held-out topics (eval never sees train motions) | | |
| | Spatial / chess / CA | Eval states rejected against train state hashes | | |
| | `nlp_cloze` | WikiText-2 **document-disjoint** (train split → train, test split → eval) | | |
| **Do not mix `train` rows into `test` (or vice versa) when building local evals.** | |
| ## Splits | |
| | Split | Rows | Role | | |
| |---|---:|---| | |
| | `train` | 240800 | SFT | | |
| | `test` | 4700 | Held-out eval only | | |
| Gym counts (train): `{"connect4": 35000, "game2048": 35000, "gridworld": 35000, "sokoban": 35000, "chess": 32400, "cellular_automata": 12000, "nlp_cloze": 12000, "word_games": 12000, "debate_judge": 10800, "resource_allocator": 10800, "ticket_dungeon": 10800}` | |
| ## Schema | |
| Flat columns for Hub friendliness: `state` and `meta` are JSON strings; | |
| `criteria_keys` / `criteria_values` parallel lists. | |
| ## Cite / notes | |
| - Leakage + cloze: `docs/NOTE_S1_CLOZE_AND_LEAKAGE_2026-09-22.md` | |
| - This revision: `docs/NOTE_ACTION_V2_QWEN_POSTMORTEM.md` | |
| - Not affiliated with TypeSafe AI / JevBench official leaderboards. | |