Create .eval_results/typed-decisions.yaml
Browse files
.eval_results/typed-decisions.yaml
ADDED
|
@@ -0,0 +1,52 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
- dataset:
|
| 2 |
+
id: LocalLLaMA/typed-decisions
|
| 3 |
+
task_id: accuracy
|
| 4 |
+
revision: 1c9479de422cc0117abcddfadc3afe2cd8383e40
|
| 5 |
+
value: 0.7735
|
| 6 |
+
date: '2026-10-31'
|
| 7 |
+
notes: 'Full test split (400 cases); zero-shot chat-JSON adapter, self-reported
|
| 8 |
+
probabilities (not native System One classifier/logits); low reasoning, max_completion_tokens=1024;
|
| 9 |
+
one generated response per case; official scoring functions; 0 invalid and 0 truncated
|
| 10 |
+
outputs retained. Public model-generated labels; self-report, not independently
|
| 11 |
+
verified. No training on this benchmark was performed by this runner; upstream
|
| 12 |
+
training exposure is unknown. Organization: 江苏银行 (Bank of Jiangsu); public model
|
| 13 |
+
display name: intellisu-coder-pro.'
|
| 14 |
+
- dataset:
|
| 15 |
+
id: LocalLLaMA/typed-decisions
|
| 16 |
+
task_id: kl_from_gold
|
| 17 |
+
revision: 1c9479de422cc0117abcddfadc3afe2cd8383e40
|
| 18 |
+
value: 0.30256444354916573
|
| 19 |
+
date: '2026-10-31'
|
| 20 |
+
notes: 'Full test split (400 cases); zero-shot chat-JSON adapter, self-reported
|
| 21 |
+
probabilities (not native System One classifier/logits); low reasoning, max_completion_tokens=1024;
|
| 22 |
+
one generated response per case; official scoring functions; 0 invalid and 0 truncated
|
| 23 |
+
outputs retained. Public model-generated labels; self-report, not independently
|
| 24 |
+
verified. No training on this benchmark was performed by this runner; upstream
|
| 25 |
+
training exposure is unknown. Organization: 江苏银行 (Bank of Jiangsu); public model
|
| 26 |
+
display name: intellisu-coder-pro.'
|
| 27 |
+
- dataset:
|
| 28 |
+
id: LocalLLaMA/typed-decisions
|
| 29 |
+
task_id: brier
|
| 30 |
+
revision: 1c9479de422cc0117abcddfadc3afe2cd8383e40
|
| 31 |
+
value: 0.1164229197998335
|
| 32 |
+
date: '2026-10-31'
|
| 33 |
+
notes: 'Full test split (400 cases); zero-shot chat-JSON adapter, self-reported
|
| 34 |
+
probabilities (not native System One classifier/logits); low reasoning, max_completion_tokens=1024;
|
| 35 |
+
one generated response per case; official scoring functions; 0 invalid and 0 truncated
|
| 36 |
+
outputs retained. Public model-generated labels; self-report, not independently
|
| 37 |
+
verified. No training on this benchmark was performed by this runner; upstream
|
| 38 |
+
training exposure is unknown. Organization: 江苏银行 (Bank of Jiangsu); public model
|
| 39 |
+
display name: intellisu-coder-pro.'
|
| 40 |
+
- dataset:
|
| 41 |
+
id: LocalLLaMA/typed-decisions
|
| 42 |
+
task_id: ece
|
| 43 |
+
revision: 1c9479de422cc0117abcddfadc3afe2cd8383e40
|
| 44 |
+
value: 0.011837000000000002
|
| 45 |
+
date: '2026-10-31'
|
| 46 |
+
notes: 'Full test split (400 cases); zero-shot chat-JSON adapter, self-reported
|
| 47 |
+
probabilities (not native System One classifier/logits); low reasoning, max_completion_tokens=1024;
|
| 48 |
+
one generated response per case; official scoring functions; 0 invalid and 0 truncated
|
| 49 |
+
outputs retained. Public model-generated labels; self-report, not independently
|
| 50 |
+
verified. No training on this benchmark was performed by this runner; upstream
|
| 51 |
+
training exposure is unknown. Organization: 江苏银行 (Bank of Jiangsu); public model
|
| 52 |
+
display name: intellisu-coder-pro.'
|