Upload README.md with huggingface_hub

README.md

@@ -22,60 +22,25 @@ configs:
     path: data/test-*
   - split: test_hard
     path: data/test_hard-*
-  - split: all
-    path: data/all-*
   - split: full
     path: data/full-*
-dataset_info:
-  features:
-  - name: task
-    dtype: string
-  - name: state
-    dtype: string
-  - name: instructions
-    dtype: string
-  - name: criteria_keys
-    list: string
-  - name: criteria_values
-    list: string
-  - name: label_alias
-    dtype: string
-  - name: label_key
-    dtype: string
-  - name: meta
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 25088586
-    num_examples: 32400
-  - name: test
-    num_bytes: 2793606
-    num_examples: 3600
-  - name: test_hard
-    num_bytes: 4279035
-    num_examples: 5400
-  - name: full
-    num_bytes: 27882192
-    num_examples: 36000
-  download_size: 16241205
-  dataset_size: 60043419
 ---
 
 # systemone-lite-general
 
 Synthetic **System One–style** typed decision examples for
-[`systemone-lite`](https://github.com/fritzprix/systemone-lite)
-(local Jev-shaped `state + questions → probabilities` training).
+[`systemone-lite`](https://github.com/fritzprix/systemone-lite).
 
 **Not affiliated with TypeSafe AI / Jev.** Rule-based labels only.
 
 ## Splits
 
-| Split | Rows |
-|------|------|
-| `train` | 16,200 |
-| `test` | 1,800 |
-| `all` | 18,000 |
+| Split | Rows | Notes |
+|------|------|------|
+| `train` | 32,400 | Stratified mix of 3 gyms |
+| `test` | 3,600 | iid held-out by task |
+| `test_hard` | 5,400 | layout / paraphrase / option-subset shift |
+| `full` | 36,000 | train + iid test pool |
 
 ## Gyms / tasks
 
@@ -83,37 +48,12 @@ Synthetic **System One–style** typed decision examples for
 - **ResourceAllocator**: `alloc.fund_next`, `alloc.can_fund_all`, `alloc.pressure`
 - **DebateJudge**: `debate.winner`, `debate.enough_evidence`, `debate.confidence`
 
-Noul/score are encoded as single-letter **choice** options (`criteria_keys` / `label_alias`)
-for causal LM next-token SFT.
-
-## Fields
-
-| Field | Description |
-|------|-------------|
-| `task` | Task id |
-| `state` | JSON string of application state |
-| `instructions` | Question text |
-| `criteria_keys` / `criteria_values` | Option aliases and descriptions |
-| `label_alias` | Gold option letter (e.g. `A`) |
-| `label_key` | Gold semantic key (e.g. `billing`) |
-| `meta` | JSON string (gym, schema hints, …) |
+Noul/score are encoded as single-letter **choice** options for causal LM next-token SFT.
 
 ## Load
 
 ```python
 from datasets import load_dataset
-import json
-
 ds = load_dataset("dwidlee/systemone-lite-general")
-row = ds["train"][0]
-state = json.loads(row["state"])
-criteria = dict(zip(row["criteria_keys"], row["criteria_values"]))
-```
-
-## Reproduce locally
-
-```bash
-git clone https://github.com/fritzprix/systemone-lite
-cd systemone-lite
-python scripts/build_general_distill.py --episodes 2000 --seed 0
+print(ds)
 ```