Datasets:
one config per model version: v8
Browse files- README.md +7 -6
- SHA256SUMS +2 -2
- data/{test.parquet → v8/test.parquet} +0 -0
README.md
CHANGED
|
@@ -26,20 +26,21 @@ task_categories:
|
|
| 26 |
- question-answering
|
| 27 |
size_categories:
|
| 28 |
- n<1K
|
| 29 |
-
pretty_name: Cooking
|
| 30 |
tags:
|
| 31 |
- cooking
|
| 32 |
- evaluation
|
| 33 |
- llm-as-judge
|
| 34 |
- multilingual
|
| 35 |
configs:
|
| 36 |
-
- config_name:
|
|
|
|
| 37 |
data_files:
|
| 38 |
- split: test
|
| 39 |
-
path: data/test.parquet
|
| 40 |
---
|
| 41 |
|
| 42 |
-
# Cooking
|
| 43 |
|
| 44 |
> PRIVATE DRAFT (not released). Built locally by `scripts/release/build_eval_dataset.py`.
|
| 45 |
|
|
@@ -80,7 +81,7 @@ differ), `hedged_wrong` (hedged but wrong), `invents` (states recipe-specific de
|
|
| 80 |
|
| 81 |
```python
|
| 82 |
import pandas as pd
|
| 83 |
-
df = pd.read_parquet("data/test.parquet")
|
| 84 |
|
| 85 |
def rate(d, who): # % of answers where 3 of 4 judges gave the key verdict
|
| 86 |
d = d[d[f"{who}_judged_by_all"]]
|
|
@@ -138,4 +139,4 @@ CC-BY-SA-4.0 400, LicenseRef-PD 216, LicenseRef-PDM-1.0 66. Full list: `ATTRIBUT
|
|
| 138 |
|
| 139 |
## Files
|
| 140 |
|
| 141 |
-
`data/test.parquet`, `ATTRIBUTION.md`, `SHA256SUMS`.
|
|
|
|
| 26 |
- question-answering
|
| 27 |
size_categories:
|
| 28 |
- n<1K
|
| 29 |
+
pretty_name: Qwen3-8B Cooking evaluation set
|
| 30 |
tags:
|
| 31 |
- cooking
|
| 32 |
- evaluation
|
| 33 |
- llm-as-judge
|
| 34 |
- multilingual
|
| 35 |
configs:
|
| 36 |
+
- config_name: v8
|
| 37 |
+
default: true
|
| 38 |
data_files:
|
| 39 |
- split: test
|
| 40 |
+
path: data/v8/test.parquet
|
| 41 |
---
|
| 42 |
|
| 43 |
+
# Qwen3-8B Cooking evaluation set (config `v8`)
|
| 44 |
|
| 45 |
> PRIVATE DRAFT (not released). Built locally by `scripts/release/build_eval_dataset.py`.
|
| 46 |
|
|
|
|
| 81 |
|
| 82 |
```python
|
| 83 |
import pandas as pd
|
| 84 |
+
df = pd.read_parquet("data/v8/test.parquet") # or: datasets.load_dataset("HumanjavaEnterprises/cooking-eval", "v8")
|
| 85 |
|
| 86 |
def rate(d, who): # % of answers where 3 of 4 judges gave the key verdict
|
| 87 |
d = d[d[f"{who}_judged_by_all"]]
|
|
|
|
| 139 |
|
| 140 |
## Files
|
| 141 |
|
| 142 |
+
`data/v8/test.parquet` (config `v8`; each model version gets its own config, scored on the same questions), `ATTRIBUTION.md`, `SHA256SUMS`.
|
SHA256SUMS
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
421ccb44da746896229e97bde6e94ba747580ae5a5c965a6067645a771d081a5 ATTRIBUTION.md
|
| 2 |
-
|
| 3 |
-
7d6053e86dfd7e3e038f6f90cd4a8a15eadc7bf6fc715afc564f2873d43a01bc data/test.parquet
|
|
|
|
| 1 |
421ccb44da746896229e97bde6e94ba747580ae5a5c965a6067645a771d081a5 ATTRIBUTION.md
|
| 2 |
+
a9f32cdc1b1e134ef9f1f86783966722cb550ea53ef0bbac921a41ec00d3c23e README.md
|
| 3 |
+
7d6053e86dfd7e3e038f6f90cd4a8a15eadc7bf6fc715afc564f2873d43a01bc data/v8/test.parquet
|
data/{test.parquet → v8/test.parquet}
RENAMED
|
File without changes
|