EvalSet
Held-out Q&A pairs frozen on a pipeline's first run; every version is scored on them.
Held-out Q&A pairs frozen on a pipeline's first run; every version is scored on them.
| Property | Type | Required | Description |
|---|---|---|---|
file_id | string | Yes | Project file with the eval pairs (prompt + reference answer). |
chunk_hashes | string[] | Yes | Chunks the eval pairs came from, excluded from training. |
size | integer | Yes | Number of eval pairs. |
version | integer | No | Incremented whenever the eval set is rebuilt. Default: 1. |
built_at | integer | Yes |
We use cookies on our website. We use them to ensure proper functioning of the site and, if you agree, for purposes such as analytics, marketing, and targeting ads.