mirror of
https://github.com/meta-llama/llama-stack.git
synced 2025-08-07 19:12:09 +00:00
better params fields
This commit is contained in:
Xi Yan
parent
93c131ed5f
commit
a7abe6df74
3 changed files with 763 additions and 571 deletions
452
docs/_static/llama-stack-spec.html
vendored
452
docs/_static/llama-stack-spec.html
vendored
|
|
@ -6386,6 +6386,14 @@
|
|||
"description": "An agent candidate for evaluation."
|
||||
},
|
||||
"AnswerCorrectnessScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "answer_correctness",
|
||||
"default": "answer_correctness"
|
||||
},
|
||||
"answer_correctness": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
|
|
@ -6402,50 +6410,28 @@
|
|||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "answer_correctness",
|
||||
"default": "answer_correctness"
|
||||
"additionalProperties": false,
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
"type",
|
||||
"answer_correctness"
|
||||
],
|
||||
"title": "AnswerCorrectnessScoringFnParams"
|
||||
},
|
||||
"AnswerRelevancyScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string",
|
||||
"enum": [
|
||||
"average",
|
||||
"median",
|
||||
"categorical_count",
|
||||
"accuracy"
|
||||
],
|
||||
"title": "AggregationFunctionType",
|
||||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "answer_relevancy",
|
||||
"default": "answer_relevancy"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
],
|
||||
"title": "AnswerRelevancyScoringFnParams"
|
||||
},
|
||||
"AnswerSimilarityScoringFnParams": {
|
||||
"answer_relevancy": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
|
|
@ -6462,16 +6448,54 @@
|
|||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "answer_similarity",
|
||||
"default": "answer_similarity"
|
||||
"additionalProperties": false,
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
"type",
|
||||
"answer_relevancy"
|
||||
],
|
||||
"title": "AnswerRelevancyScoringFnParams"
|
||||
},
|
||||
"AnswerSimilarityScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "answer_similarity",
|
||||
"default": "answer_similarity"
|
||||
},
|
||||
"answer_similarity": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string",
|
||||
"enum": [
|
||||
"average",
|
||||
"median",
|
||||
"categorical_count",
|
||||
"accuracy"
|
||||
],
|
||||
"title": "AggregationFunctionType",
|
||||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type",
|
||||
"answer_similarity"
|
||||
],
|
||||
"title": "AnswerSimilarityScoringFnParams"
|
||||
},
|
||||
|
|
@ -6503,6 +6527,14 @@
|
|||
"description": "A benchmark configuration for evaluation."
|
||||
},
|
||||
"ContextEntityRecallScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "context_entity_recall",
|
||||
"default": "context_entity_recall"
|
||||
},
|
||||
"context_entity_recall": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
|
|
@ -6519,20 +6551,28 @@
|
|||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "context_entity_recall",
|
||||
"default": "context_entity_recall"
|
||||
"additionalProperties": false,
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
"type",
|
||||
"context_entity_recall"
|
||||
],
|
||||
"title": "ContextEntityRecallScoringFnParams"
|
||||
},
|
||||
"ContextPrecisionScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "context_precision",
|
||||
"default": "context_precision"
|
||||
},
|
||||
"context_precision": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
|
|
@ -6549,20 +6589,28 @@
|
|||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "context_precision",
|
||||
"default": "context_precision"
|
||||
"additionalProperties": false,
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
"type",
|
||||
"context_precision"
|
||||
],
|
||||
"title": "ContextPrecisionScoringFnParams"
|
||||
},
|
||||
"ContextRecallScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "context_recall",
|
||||
"default": "context_recall"
|
||||
},
|
||||
"context_recall": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
|
|
@ -6579,50 +6627,28 @@
|
|||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "context_recall",
|
||||
"default": "context_recall"
|
||||
"additionalProperties": false,
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
"type",
|
||||
"context_recall"
|
||||
],
|
||||
"title": "ContextRecallScoringFnParams"
|
||||
},
|
||||
"ContextRelevancyScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string",
|
||||
"enum": [
|
||||
"average",
|
||||
"median",
|
||||
"categorical_count",
|
||||
"accuracy"
|
||||
],
|
||||
"title": "AggregationFunctionType",
|
||||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "context_relevancy",
|
||||
"default": "context_relevancy"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
],
|
||||
"title": "ContextRelevancyScoringFnParams"
|
||||
},
|
||||
"EqualityScoringFnParams": {
|
||||
"context_relevancy": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
|
|
@ -6639,16 +6665,98 @@
|
|||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "equality",
|
||||
"default": "equality"
|
||||
"additionalProperties": false,
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
"type",
|
||||
"context_relevancy"
|
||||
],
|
||||
"title": "ContextRelevancyScoringFnParams"
|
||||
},
|
||||
"CustomLLMAsJudgeScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "custom_llm_as_judge",
|
||||
"default": "custom_llm_as_judge"
|
||||
},
|
||||
"custom_llm_as_judge": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "custom_llm_as_judge",
|
||||
"default": "custom_llm_as_judge"
|
||||
},
|
||||
"judge_model": {
|
||||
"type": "string"
|
||||
},
|
||||
"prompt_template": {
|
||||
"type": "string"
|
||||
},
|
||||
"judge_score_regexes": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string"
|
||||
}
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type",
|
||||
"judge_model"
|
||||
],
|
||||
"title": "CustomLLMAsJudgeScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type",
|
||||
"custom_llm_as_judge"
|
||||
],
|
||||
"title": "CustomLLMAsJudgeScoringFnParams"
|
||||
},
|
||||
"EqualityScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "equality",
|
||||
"default": "equality"
|
||||
},
|
||||
"equality": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string",
|
||||
"enum": [
|
||||
"average",
|
||||
"median",
|
||||
"categorical_count",
|
||||
"accuracy"
|
||||
],
|
||||
"title": "AggregationFunctionType",
|
||||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type",
|
||||
"equality"
|
||||
],
|
||||
"title": "EqualityScoringFnParams"
|
||||
},
|
||||
|
|
@ -6670,6 +6778,14 @@
|
|||
}
|
||||
},
|
||||
"FactualityScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "factuality",
|
||||
"default": "factuality"
|
||||
},
|
||||
"factuality": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
|
|
@ -6686,20 +6802,28 @@
|
|||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "factuality",
|
||||
"default": "factuality"
|
||||
"additionalProperties": false,
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
"type",
|
||||
"factuality"
|
||||
],
|
||||
"title": "FactualityScoringFnParams"
|
||||
},
|
||||
"FaithfulnessScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "faithfulness",
|
||||
"default": "faithfulness"
|
||||
},
|
||||
"faithfulness": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
|
|
@ -6716,60 +6840,18 @@
|
|||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "faithfulness",
|
||||
"default": "faithfulness"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
],
|
||||
"title": "FaithfulnessScoringFnParams"
|
||||
},
|
||||
"LLMAsJudgeScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "custom_llm_as_judge",
|
||||
"default": "custom_llm_as_judge"
|
||||
},
|
||||
"judge_model": {
|
||||
"type": "string"
|
||||
},
|
||||
"prompt_template": {
|
||||
"type": "string"
|
||||
},
|
||||
"judge_score_regexes": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string"
|
||||
}
|
||||
},
|
||||
"aggregation_functions": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string",
|
||||
"enum": [
|
||||
"average",
|
||||
"median",
|
||||
"categorical_count",
|
||||
"accuracy"
|
||||
],
|
||||
"title": "AggregationFunctionType",
|
||||
"description": "A type of aggregation function."
|
||||
}
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type",
|
||||
"judge_model"
|
||||
"faithfulness"
|
||||
],
|
||||
"title": "LLMAsJudgeScoringFnParams"
|
||||
"title": "FaithfulnessScoringFnParams"
|
||||
},
|
||||
"ModelCandidate": {
|
||||
"type": "object",
|
||||
|
|
@ -6804,42 +6886,12 @@
|
|||
"RegexParserMathScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"parsing_regexes": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string"
|
||||
},
|
||||
"description": "(Optional) Regexes to extract the answer from generated response."
|
||||
},
|
||||
"aggregation_functions": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string",
|
||||
"enum": [
|
||||
"average",
|
||||
"median",
|
||||
"categorical_count",
|
||||
"accuracy"
|
||||
],
|
||||
"title": "AggregationFunctionType",
|
||||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "regex_parser_math_response",
|
||||
"default": "regex_parser_math_response"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"parsing_regexes",
|
||||
"type"
|
||||
],
|
||||
"title": "RegexParserMathScoringFnParams"
|
||||
},
|
||||
"RegexParserScoringFnParams": {
|
||||
"regex_parser_math_response": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"parsing_regexes": {
|
||||
|
|
@ -6863,24 +6915,74 @@
|
|||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "regex_parser",
|
||||
"default": "regex_parser"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"parsing_regexes",
|
||||
"type"
|
||||
"parsing_regexes"
|
||||
],
|
||||
"title": "RegexParserScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type",
|
||||
"regex_parser_math_response"
|
||||
],
|
||||
"title": "RegexParserMathScoringFnParams"
|
||||
},
|
||||
"RegexParserScoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "regex_parser",
|
||||
"default": "regex_parser"
|
||||
},
|
||||
"regex_parser": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"parsing_regexes": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string"
|
||||
},
|
||||
"description": "(Optional) Regexes to extract the answer from generated response."
|
||||
},
|
||||
"aggregation_functions": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"type": "string",
|
||||
"enum": [
|
||||
"average",
|
||||
"median",
|
||||
"categorical_count",
|
||||
"accuracy"
|
||||
],
|
||||
"title": "AggregationFunctionType",
|
||||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"parsing_regexes"
|
||||
],
|
||||
"title": "RegexParserScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type",
|
||||
"regex_parser"
|
||||
],
|
||||
"title": "RegexParserScoringFnParams"
|
||||
},
|
||||
"ScoringFnParams": {
|
||||
"oneOf": [
|
||||
{
|
||||
"$ref": "#/components/schemas/LLMAsJudgeScoringFnParams"
|
||||
"$ref": "#/components/schemas/CustomLLMAsJudgeScoringFnParams"
|
||||
},
|
||||
{
|
||||
"$ref": "#/components/schemas/RegexParserScoringFnParams"
|
||||
|
|
@ -6925,7 +7027,7 @@
|
|||
"discriminator": {
|
||||
"propertyName": "type",
|
||||
"mapping": {
|
||||
"custom_llm_as_judge": "#/components/schemas/LLMAsJudgeScoringFnParams",
|
||||
"custom_llm_as_judge": "#/components/schemas/CustomLLMAsJudgeScoringFnParams",
|
||||
"regex_parser": "#/components/schemas/RegexParserScoringFnParams",
|
||||
"regex_parser_math_response": "#/components/schemas/RegexParserMathScoringFnParams",
|
||||
"equality": "#/components/schemas/EqualityScoringFnParams",
|
||||
|
|
@ -6943,6 +7045,14 @@
|
|||
}
|
||||
},
|
||||
"SubsetOfcoringFnParams": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "subset_of",
|
||||
"default": "subset_of"
|
||||
},
|
||||
"subset_of": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"aggregation_functions": {
|
||||
|
|
@ -6959,16 +7069,16 @@
|
|||
"description": "A type of aggregation function."
|
||||
},
|
||||
"description": "(Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed."
|
||||
}
|
||||
},
|
||||
"type": {
|
||||
"type": "string",
|
||||
"const": "subset_of",
|
||||
"default": "subset_of"
|
||||
"additionalProperties": false,
|
||||
"title": "BasicScoringFnParamsFields"
|
||||
}
|
||||
},
|
||||
"additionalProperties": false,
|
||||
"required": [
|
||||
"type"
|
||||
"type",
|
||||
"subset_of"
|
||||
],
|
||||
"title": "SubsetOfcoringFnParams"
|
||||
},
|
||||
|
|
|
|||
501
docs/_static/llama-stack-spec.yaml
vendored
501
docs/_static/llama-stack-spec.yaml
vendored
|
|
@ -4451,54 +4451,42 @@ components:
|
|||
AnswerCorrectnessScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: answer_correctness
|
||||
default: answer_correctness
|
||||
answer_correctness:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- answer_correctness
|
||||
title: AnswerCorrectnessScoringFnParams
|
||||
AnswerRelevancyScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: answer_relevancy
|
||||
default: answer_relevancy
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
title: AnswerRelevancyScoringFnParams
|
||||
AnswerSimilarityScoringFnParams:
|
||||
answer_relevancy:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
|
|
@ -4513,15 +4501,45 @@ components:
|
|||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- answer_relevancy
|
||||
title: AnswerRelevancyScoringFnParams
|
||||
AnswerSimilarityScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
type:
|
||||
type: string
|
||||
const: answer_similarity
|
||||
default: answer_similarity
|
||||
answer_similarity:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- answer_similarity
|
||||
title: AnswerSimilarityScoringFnParams
|
||||
BenchmarkConfig:
|
||||
type: object
|
||||
|
|
@ -4551,188 +4569,135 @@ components:
|
|||
ContextEntityRecallScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: context_entity_recall
|
||||
default: context_entity_recall
|
||||
context_entity_recall:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- context_entity_recall
|
||||
title: ContextEntityRecallScoringFnParams
|
||||
ContextPrecisionScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: context_precision
|
||||
default: context_precision
|
||||
context_precision:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- context_precision
|
||||
title: ContextPrecisionScoringFnParams
|
||||
ContextRecallScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: context_recall
|
||||
default: context_recall
|
||||
context_recall:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- context_recall
|
||||
title: ContextRecallScoringFnParams
|
||||
ContextRelevancyScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: context_relevancy
|
||||
default: context_relevancy
|
||||
context_relevancy:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- context_relevancy
|
||||
title: ContextRelevancyScoringFnParams
|
||||
EqualityScoringFnParams:
|
||||
CustomLLMAsJudgeScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: equality
|
||||
default: equality
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
title: EqualityScoringFnParams
|
||||
EvalCandidate:
|
||||
oneOf:
|
||||
- $ref: '#/components/schemas/ModelCandidate'
|
||||
- $ref: '#/components/schemas/AgentCandidate'
|
||||
discriminator:
|
||||
propertyName: type
|
||||
mapping:
|
||||
model: '#/components/schemas/ModelCandidate'
|
||||
agent: '#/components/schemas/AgentCandidate'
|
||||
FactualityScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: factuality
|
||||
default: factuality
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
title: FactualityScoringFnParams
|
||||
FaithfulnessScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: faithfulness
|
||||
default: faithfulness
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
title: FaithfulnessScoringFnParams
|
||||
LLMAsJudgeScoringFnParams:
|
||||
const: custom_llm_as_judge
|
||||
default: custom_llm_as_judge
|
||||
custom_llm_as_judge:
|
||||
type: object
|
||||
properties:
|
||||
type:
|
||||
|
|
@ -4747,6 +4712,26 @@ components:
|
|||
type: array
|
||||
items:
|
||||
type: string
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- judge_model
|
||||
title: CustomLLMAsJudgeScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- custom_llm_as_judge
|
||||
title: CustomLLMAsJudgeScoringFnParams
|
||||
EqualityScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
type:
|
||||
type: string
|
||||
const: equality
|
||||
default: equality
|
||||
equality:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
|
|
@ -4758,11 +4743,87 @@ components:
|
|||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- judge_model
|
||||
title: LLMAsJudgeScoringFnParams
|
||||
- equality
|
||||
title: EqualityScoringFnParams
|
||||
EvalCandidate:
|
||||
oneOf:
|
||||
- $ref: '#/components/schemas/ModelCandidate'
|
||||
- $ref: '#/components/schemas/AgentCandidate'
|
||||
discriminator:
|
||||
propertyName: type
|
||||
mapping:
|
||||
model: '#/components/schemas/ModelCandidate'
|
||||
agent: '#/components/schemas/AgentCandidate'
|
||||
FactualityScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
type:
|
||||
type: string
|
||||
const: factuality
|
||||
default: factuality
|
||||
factuality:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- factuality
|
||||
title: FactualityScoringFnParams
|
||||
FaithfulnessScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
type:
|
||||
type: string
|
||||
const: faithfulness
|
||||
default: faithfulness
|
||||
faithfulness:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- faithfulness
|
||||
title: FaithfulnessScoringFnParams
|
||||
ModelCandidate:
|
||||
type: object
|
||||
properties:
|
||||
|
|
@ -4791,36 +4852,11 @@ components:
|
|||
RegexParserMathScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
parsing_regexes:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
description: >-
|
||||
(Optional) Regexes to extract the answer from generated response.
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: regex_parser_math_response
|
||||
default: regex_parser_math_response
|
||||
additionalProperties: false
|
||||
required:
|
||||
- parsing_regexes
|
||||
- type
|
||||
title: RegexParserMathScoringFnParams
|
||||
RegexParserScoringFnParams:
|
||||
regex_parser_math_response:
|
||||
type: object
|
||||
properties:
|
||||
parsing_regexes:
|
||||
|
|
@ -4841,20 +4877,59 @@ components:
|
|||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
required:
|
||||
- parsing_regexes
|
||||
title: RegexParserScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- regex_parser_math_response
|
||||
title: RegexParserMathScoringFnParams
|
||||
RegexParserScoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
type:
|
||||
type: string
|
||||
const: regex_parser
|
||||
default: regex_parser
|
||||
regex_parser:
|
||||
type: object
|
||||
properties:
|
||||
parsing_regexes:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
description: >-
|
||||
(Optional) Regexes to extract the answer from generated response.
|
||||
aggregation_functions:
|
||||
type: array
|
||||
items:
|
||||
type: string
|
||||
enum:
|
||||
- average
|
||||
- median
|
||||
- categorical_count
|
||||
- accuracy
|
||||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
required:
|
||||
- parsing_regexes
|
||||
title: RegexParserScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- regex_parser
|
||||
title: RegexParserScoringFnParams
|
||||
ScoringFnParams:
|
||||
oneOf:
|
||||
- $ref: '#/components/schemas/LLMAsJudgeScoringFnParams'
|
||||
- $ref: '#/components/schemas/CustomLLMAsJudgeScoringFnParams'
|
||||
- $ref: '#/components/schemas/RegexParserScoringFnParams'
|
||||
- $ref: '#/components/schemas/RegexParserMathScoringFnParams'
|
||||
- $ref: '#/components/schemas/EqualityScoringFnParams'
|
||||
|
|
@ -4871,7 +4946,7 @@ components:
|
|||
discriminator:
|
||||
propertyName: type
|
||||
mapping:
|
||||
custom_llm_as_judge: '#/components/schemas/LLMAsJudgeScoringFnParams'
|
||||
custom_llm_as_judge: '#/components/schemas/CustomLLMAsJudgeScoringFnParams'
|
||||
regex_parser: '#/components/schemas/RegexParserScoringFnParams'
|
||||
regex_parser_math_response: '#/components/schemas/RegexParserMathScoringFnParams'
|
||||
equality: '#/components/schemas/EqualityScoringFnParams'
|
||||
|
|
@ -4886,6 +4961,13 @@ components:
|
|||
context_recall: '#/components/schemas/ContextRecallScoringFnParams'
|
||||
context_relevancy: '#/components/schemas/ContextRelevancyScoringFnParams'
|
||||
SubsetOfcoringFnParams:
|
||||
type: object
|
||||
properties:
|
||||
type:
|
||||
type: string
|
||||
const: subset_of
|
||||
default: subset_of
|
||||
subset_of:
|
||||
type: object
|
||||
properties:
|
||||
aggregation_functions:
|
||||
|
|
@ -4900,15 +4982,14 @@ components:
|
|||
title: AggregationFunctionType
|
||||
description: A type of aggregation function.
|
||||
description: >-
|
||||
(Optional) Aggregation functions to apply to the scores of each row. If
|
||||
not provided, no aggregation will be performed.
|
||||
type:
|
||||
type: string
|
||||
const: subset_of
|
||||
default: subset_of
|
||||
(Optional) Aggregation functions to apply to the scores of each row.
|
||||
If not provided, no aggregation will be performed.
|
||||
additionalProperties: false
|
||||
title: BasicScoringFnParamsFields
|
||||
additionalProperties: false
|
||||
required:
|
||||
- type
|
||||
- subset_of
|
||||
title: SubsetOfcoringFnParams
|
||||
EvaluateRowsRequest:
|
||||
type: object
|
||||
|
|
|
|||
|
|
@ -67,7 +67,7 @@ class AggregationFunctionType(Enum):
|
|||
accuracy = "accuracy"
|
||||
|
||||
|
||||
class BasicScoringFnParamsCommon(BaseModel):
|
||||
class BasicScoringFnParamsFields(BaseModel):
|
||||
"""
|
||||
:param aggregation_functions: (Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed.
|
||||
"""
|
||||
|
|
@ -78,7 +78,7 @@ class BasicScoringFnParamsCommon(BaseModel):
|
|||
)
|
||||
|
||||
|
||||
class RegexParserScoringFnParamsCommon(BaseModel):
|
||||
class RegexParserScoringFnParamsFields(BaseModel):
|
||||
"""
|
||||
:param parsing_regexes: (Optional) Regexes to extract the answer from generated response.
|
||||
:param aggregation_functions: (Optional) Aggregation functions to apply to the scores of each row. If not provided, no aggregation will be performed.
|
||||
|
|
@ -93,74 +93,7 @@ class RegexParserScoringFnParamsCommon(BaseModel):
|
|||
default_factory=list,
|
||||
)
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class RegexParserScoringFnParams(RegexParserScoringFnParamsCommon):
|
||||
type: Literal["regex_parser"] = "regex_parser"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class RegexParserMathScoringFnParams(RegexParserScoringFnParamsCommon):
|
||||
type: Literal["regex_parser_math_response"] = "regex_parser_math_response"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class EqualityScoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["equality"] = "equality"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class SubsetOfcoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["subset_of"] = "subset_of"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class FactualityScoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["factuality"] = "factuality"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class FaithfulnessScoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["faithfulness"] = "faithfulness"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class AnswerCorrectnessScoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["answer_correctness"] = "answer_correctness"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class AnswerRelevancyScoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["answer_relevancy"] = "answer_relevancy"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class AnswerSimilarityScoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["answer_similarity"] = "answer_similarity"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class ContextEntityRecallScoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["context_entity_recall"] = "context_entity_recall"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class ContextPrecisionScoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["context_precision"] = "context_precision"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class ContextRecallScoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["context_recall"] = "context_recall"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class ContextRelevancyScoringFnParams(BasicScoringFnParamsCommon):
|
||||
type: Literal["context_relevancy"] = "context_relevancy"
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class LLMAsJudgeScoringFnParams(BaseModel):
|
||||
class CustomLLMAsJudgeScoringFnParamsFields(BaseModel):
|
||||
type: Literal["custom_llm_as_judge"] = "custom_llm_as_judge"
|
||||
judge_model: str
|
||||
prompt_template: Optional[str] = None
|
||||
|
|
@ -168,16 +101,84 @@ class LLMAsJudgeScoringFnParams(BaseModel):
|
|||
description="Regexes to extract the answer from generated response",
|
||||
default_factory=list,
|
||||
)
|
||||
aggregation_functions: Optional[List[AggregationFunctionType]] = Field(
|
||||
description="Aggregation functions to apply to the scores of each row",
|
||||
default_factory=list,
|
||||
)
|
||||
|
||||
@json_schema_type
|
||||
class RegexParserScoringFnParams(BaseModel):
|
||||
type: Literal["regex_parser"] = "regex_parser"
|
||||
regex_parser: RegexParserScoringFnParamsFields
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class RegexParserMathScoringFnParams(BaseModel):
|
||||
type: Literal["regex_parser_math_response"] = "regex_parser_math_response"
|
||||
regex_parser_math_response: RegexParserScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class EqualityScoringFnParams(BaseModel):
|
||||
type: Literal["equality"] = "equality"
|
||||
equality: BasicScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class SubsetOfcoringFnParams(BaseModel):
|
||||
type: Literal["subset_of"] = "subset_of"
|
||||
subset_of: BasicScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class FactualityScoringFnParams(BaseModel):
|
||||
type: Literal["factuality"] = "factuality"
|
||||
factuality: BasicScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class FaithfulnessScoringFnParams(BaseModel):
|
||||
type: Literal["faithfulness"] = "faithfulness"
|
||||
faithfulness: BasicScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class AnswerCorrectnessScoringFnParams(BaseModel):
|
||||
type: Literal["answer_correctness"] = "answer_correctness"
|
||||
answer_correctness: BasicScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class AnswerRelevancyScoringFnParams(BaseModel):
|
||||
type: Literal["answer_relevancy"] = "answer_relevancy"
|
||||
answer_relevancy: BasicScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class AnswerSimilarityScoringFnParams(BaseModel):
|
||||
type: Literal["answer_similarity"] = "answer_similarity"
|
||||
answer_similarity: BasicScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class ContextEntityRecallScoringFnParams(BaseModel):
|
||||
type: Literal["context_entity_recall"] = "context_entity_recall"
|
||||
context_entity_recall: BasicScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class ContextPrecisionScoringFnParams(BaseModel):
|
||||
type: Literal["context_precision"] = "context_precision"
|
||||
context_precision: BasicScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class ContextRecallScoringFnParams(BaseModel):
|
||||
type: Literal["context_recall"] = "context_recall"
|
||||
context_recall: BasicScoringFnParamsFields
|
||||
|
||||
@json_schema_type
|
||||
class ContextRelevancyScoringFnParams(BaseModel):
|
||||
type: Literal["context_relevancy"] = "context_relevancy"
|
||||
context_relevancy: BasicScoringFnParamsFields
|
||||
|
||||
|
||||
@json_schema_type
|
||||
class CustomLLMAsJudgeScoringFnParams(BaseModel):
|
||||
type: Literal["custom_llm_as_judge"] = "custom_llm_as_judge"
|
||||
custom_llm_as_judge: CustomLLMAsJudgeScoringFnParamsFields
|
||||
|
||||
|
||||
ScoringFnParams = register_schema(
|
||||
Annotated[
|
||||
Union[
|
||||
LLMAsJudgeScoringFnParams,
|
||||
CustomLLMAsJudgeScoringFnParams,
|
||||
RegexParserScoringFnParams,
|
||||
RegexParserMathScoringFnParams,
|
||||
EqualityScoringFnParams,
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue