phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-10-04 04:04:14 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

reward scoring

Browse source
This commit is contained in:
Raghotham Murthy 2024-07-10 21:56:16 -07:00
parent 69ecf55de2
commit ebb59aa35f
3 changed files with 117 additions and 79 deletions
Download patch file Download diff file Expand all files Collapse all files

100
source/openapi.html
View file

@ -1960,15 +1960,18 @@
"$ref": "#/components/schemas/Message"
}
},
"generation": {
"$ref": "#/components/schemas/Message"
"k_generations": {
"type": "array",
"items": {
"$ref": "#/components/schemas/Message"
}
}
},
"additionalProperties": false,
"required": [
"prompt",
"message_history",
"generation"
"k_generations"
]
}
},
@ -1983,46 +1986,49 @@
],
"title": "Request to score a reward function. A list of prompts and a list of responses per prompt."
},
"KScoredPromptGenerations": {
"type": "object",
"properties": {
"prompt": {
"$ref": "#/components/schemas/Message"
},
"k_scored_generations": {
"type": "array",
"items": {
"$ref": "#/components/schemas/MessageScore"
}
}
},
"additionalProperties": false,
"required": [
"prompt",
"k_scored_generations"
]
},
"MessageScore": {
"type": "object",
"properties": {
"message": {
"$ref": "#/components/schemas/Message"
},
"score": {
"type": "number"
}
},
"additionalProperties": false,
"required": [
"message",
"score"
],
"title": "A single message and its score."
},
"RewardScoringResponse": {
"type": "object",
"properties": {
"scored_generations": {
"type": "array",
"items": {
"type": "object",
"properties": {
"prompt_generation": {
"type": "object",
"properties": {
"prompt": {
"$ref": "#/components/schemas/Message"
},
"message_history": {
"type": "array",
"items": {
"$ref": "#/components/schemas/Message"
}
},
"generation": {
"$ref": "#/components/schemas/Message"
}
},
"additionalProperties": false,
"required": [
"prompt",
"message_history",
"generation"
]
},
"score": {
"type": "number"
}
},
"additionalProperties": false,
"required": [
"prompt_generation",
"score"
]
"$ref": "#/components/schemas/KScoredPromptGenerations"
}
}
},
@ -2306,11 +2312,14 @@
}
],
"tags": [
{
"name": "Inference"
},
{
"name": "SyntheticDataGeneration"
},
{
"name": "RewardScoring"
"name": "Datasets"
},
{
"name": "AgenticSystem"
@ -2319,10 +2328,7 @@
"name": "Finetuning"
},
{
"name": "Inference"
},
{
"name": "Datasets"
"name": "RewardScoring"
},
{
"name": "ShieldConfig",
@ -2416,6 +2422,14 @@
"name": "RewardScoringRequest",
"description": "Request to score a reward function. A list of prompts and a list of responses per prompt.\n\n<SchemaDefinition schemaRef=\"#/components/schemas/RewardScoringRequest\" />"
},
{
"name": "KScoredPromptGenerations",
"description": "<SchemaDefinition schemaRef=\"#/components/schemas/KScoredPromptGenerations\" />"
},
{
"name": "MessageScore",
"description": "A single message and its score.\n\n<SchemaDefinition schemaRef=\"#/components/schemas/MessageScore\" />"
},
{
"name": "RewardScoringResponse",
"description": "Response from the reward scoring. Batch of (prompt, response, score) tuples that pass the threshold.\n\n<SchemaDefinition schemaRef=\"#/components/schemas/RewardScoringResponse\" />"
@ -2470,8 +2484,10 @@
"FinetuningJobLogStream",
"FinetuningJobStatusResponse",
"FinetuningTrainRequest",
"KScoredPromptGenerations",
"LoraFinetuningConfig",
"Message",
"MessageScore",
"OptimizerConfig",
"RewardScoringRequest",
"RewardScoringResponse",