247 lines
5.7 KiB
JSON
247 lines
5.7 KiB
JSON
{
|
|
"$schema": "https://json-schema.org/draft/2020-12/schema",
|
|
"title": "CorpusItem",
|
|
"description": "Key: `item:<source-slug>:<n>`.",
|
|
"type": "object",
|
|
"properties": {
|
|
"admitted_on": {
|
|
"type": "string"
|
|
},
|
|
"group": {
|
|
"description": "The unit the split hashes: a scene, a fixture, a work.",
|
|
"type": "string"
|
|
},
|
|
"id": {
|
|
"type": "string"
|
|
},
|
|
"language": {
|
|
"type": "string"
|
|
},
|
|
"names": {
|
|
"description": "Which entities the text names, and where. Fixed with the text at intake.",
|
|
"type": "array",
|
|
"items": {
|
|
"$ref": "#/$defs/NameSpan"
|
|
}
|
|
},
|
|
"prior_turns": {
|
|
"type": "array",
|
|
"items": {
|
|
"$ref": "#/$defs/Turn"
|
|
}
|
|
},
|
|
"source": {
|
|
"$ref": "#/$defs/Source"
|
|
},
|
|
"speaker": {
|
|
"$ref": "#/$defs/EntityId"
|
|
},
|
|
"supersedes": {
|
|
"description": "Key of the item this one corrects.",
|
|
"type": [
|
|
"string",
|
|
"null"
|
|
]
|
|
},
|
|
"text": {
|
|
"type": "string"
|
|
},
|
|
"withdrawn": {
|
|
"type": "boolean"
|
|
}
|
|
},
|
|
"required": [
|
|
"id",
|
|
"text",
|
|
"language",
|
|
"speaker",
|
|
"group",
|
|
"prior_turns",
|
|
"source",
|
|
"admitted_on",
|
|
"withdrawn",
|
|
"names"
|
|
],
|
|
"$defs": {
|
|
"EntityId": {
|
|
"type": "string",
|
|
"pattern": "^entity:[a-z0-9]([a-z0-9-]{0,46}[a-z0-9])?$"
|
|
},
|
|
"ModelFamily": {
|
|
"description": "The family of the model that wrote a line.",
|
|
"oneOf": [
|
|
{
|
|
"type": "string",
|
|
"enum": [
|
|
"Anthropic",
|
|
"OpenAi",
|
|
"Google",
|
|
"Alibaba",
|
|
"Meta",
|
|
"Mistral"
|
|
]
|
|
},
|
|
{
|
|
"description": "Nothing records which family wrote the line.",
|
|
"type": "string",
|
|
"const": "Unrecorded"
|
|
}
|
|
]
|
|
},
|
|
"NameSpan": {
|
|
"description": "A byte range `[start, end)` of an item's text that names an entity. The\nstore admits it only when both ends sit on extended grapheme cluster\nboundaries (UAX #29) of the NFC text and the range's bytes equal one of the\nentity's forms in the item's language.",
|
|
"type": "object",
|
|
"properties": {
|
|
"end": {
|
|
"type": "integer",
|
|
"format": "uint32",
|
|
"minimum": 0
|
|
},
|
|
"entity": {
|
|
"$ref": "#/$defs/EntityId"
|
|
},
|
|
"start": {
|
|
"type": "integer",
|
|
"format": "uint32",
|
|
"minimum": 0
|
|
}
|
|
},
|
|
"required": [
|
|
"entity",
|
|
"start",
|
|
"end"
|
|
]
|
|
},
|
|
"Source": {
|
|
"oneOf": [
|
|
{
|
|
"type": "string",
|
|
"enum": [
|
|
"OperatorWritten"
|
|
]
|
|
},
|
|
{
|
|
"description": "A line from a cast member's hand-authored fixture.",
|
|
"type": "object",
|
|
"properties": {
|
|
"CastAuthored": {
|
|
"type": "object",
|
|
"properties": {
|
|
"fixture": {
|
|
"type": "string"
|
|
},
|
|
"persona": {
|
|
"type": "string"
|
|
}
|
|
},
|
|
"required": [
|
|
"persona",
|
|
"fixture"
|
|
]
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"required": [
|
|
"CastAuthored"
|
|
]
|
|
},
|
|
{
|
|
"type": "object",
|
|
"properties": {
|
|
"OperatorApproved": {
|
|
"type": "object",
|
|
"properties": {
|
|
"note": {
|
|
"type": "string"
|
|
}
|
|
},
|
|
"required": [
|
|
"note"
|
|
]
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"required": [
|
|
"OperatorApproved"
|
|
]
|
|
},
|
|
{
|
|
"type": "object",
|
|
"properties": {
|
|
"PublicDomain": {
|
|
"type": "object",
|
|
"properties": {
|
|
"author": {
|
|
"type": "string"
|
|
},
|
|
"edition_url": {
|
|
"type": "string"
|
|
},
|
|
"work": {
|
|
"type": "string"
|
|
}
|
|
},
|
|
"required": [
|
|
"work",
|
|
"author",
|
|
"edition_url"
|
|
]
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"required": [
|
|
"PublicDomain"
|
|
]
|
|
},
|
|
{
|
|
"description": "Labelled model output; never in the test split, never judged by its own family.\n`family` is `Unrecorded` when nothing records one. `generator` is the\nstring the fixture or run wrote, kept verbatim.",
|
|
"type": "object",
|
|
"properties": {
|
|
"ModelWritten": {
|
|
"type": "object",
|
|
"properties": {
|
|
"family": {
|
|
"$ref": "#/$defs/ModelFamily"
|
|
},
|
|
"generator": {
|
|
"type": [
|
|
"string",
|
|
"null"
|
|
]
|
|
},
|
|
"prompt_digest": {
|
|
"type": [
|
|
"string",
|
|
"null"
|
|
]
|
|
}
|
|
},
|
|
"required": [
|
|
"family"
|
|
]
|
|
}
|
|
},
|
|
"additionalProperties": false,
|
|
"required": [
|
|
"ModelWritten"
|
|
]
|
|
}
|
|
]
|
|
},
|
|
"Turn": {
|
|
"type": "object",
|
|
"properties": {
|
|
"speaker": {
|
|
"$ref": "#/$defs/EntityId"
|
|
},
|
|
"text": {
|
|
"type": "string"
|
|
}
|
|
},
|
|
"required": [
|
|
"speaker",
|
|
"text"
|
|
]
|
|
}
|
|
}
|
|
}
|