평가 결과 쿼리

평가 결과 쿼리

curl --request POST \
  --url https://api.example.com/v2/{entity}/{project}/eval_results/query \
  --header 'Content-Type: application/json' \
  --data '
{
  "evaluation_call_ids": [
    "<string>"
  ],
  "evaluation_run_ids": [
    "<string>"
  ],
  "filter_logic_operator": "or",
  "filters": [
    {
      "query": {
        "$expr": {
          "$and": [
            {
              "$literal": "<string>"
            }
          ]
        }
      },
      "evaluation_call_id": "<string>"
    }
  ],
  "include_costs": false,
  "include_predict_and_score_children": true,
  "include_raw_data_rows": false,
  "include_rows": true,
  "include_summary": false,
  "limit": 123,
  "offset": 0,
  "require_intersection": false,
  "resolve_row_refs": false,
  "sort_by": [
    {
      "field": "<string>",
      "evaluation_call_id": "<string>",
      "mode": "value"
    }
  ],
  "summary_require_intersection": true
}
'

import requests

url = "https://api.example.com/v2/{entity}/{project}/eval_results/query"

payload = {
    "evaluation_call_ids": ["<string>"],
    "evaluation_run_ids": ["<string>"],
    "filter_logic_operator": "or",
    "filters": [
        {
            "query": { "$expr": { "$and": [{ "$literal": "<string>" }] } },
            "evaluation_call_id": "<string>"
        }
    ],
    "include_costs": False,
    "include_predict_and_score_children": True,
    "include_raw_data_rows": False,
    "include_rows": True,
    "include_summary": False,
    "limit": 123,
    "offset": 0,
    "require_intersection": False,
    "resolve_row_refs": False,
    "sort_by": [
        {
            "field": "<string>",
            "evaluation_call_id": "<string>",
            "mode": "value"
        }
    ],
    "summary_require_intersection": True
}
headers = {"Content-Type": "application/json"}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({
    evaluation_call_ids: ['<string>'],
    evaluation_run_ids: ['<string>'],
    filter_logic_operator: 'or',
    filters: [
      {
        query: {$expr: {$and: [{$literal: '<string>'}]}},
        evaluation_call_id: '<string>'
      }
    ],
    include_costs: false,
    include_predict_and_score_children: true,
    include_raw_data_rows: false,
    include_rows: true,
    include_summary: false,
    limit: 123,
    offset: 0,
    require_intersection: false,
    resolve_row_refs: false,
    sort_by: [{field: '<string>', evaluation_call_id: '<string>', mode: 'value'}],
    summary_require_intersection: true
  })
};

fetch('https://api.example.com/v2/{entity}/{project}/eval_results/query', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.example.com/v2/{entity}/{project}/eval_results/query",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'evaluation_call_ids' => [
        '<string>'
    ],
    'evaluation_run_ids' => [
        '<string>'
    ],
    'filter_logic_operator' => 'or',
    'filters' => [
        [
                'query' => [
                                '$expr' => [
                                                                '$and' => [
                                                                                                                                [
                                                                                                                                                                                                                                                                '$literal' => '<string>'
                                                                                                                                ]
                                                                ]
                                ]
                ],
                'evaluation_call_id' => '<string>'
        ]
    ],
    'include_costs' => false,
    'include_predict_and_score_children' => true,
    'include_raw_data_rows' => false,
    'include_rows' => true,
    'include_summary' => false,
    'limit' => 123,
    'offset' => 0,
    'require_intersection' => false,
    'resolve_row_refs' => false,
    'sort_by' => [
        [
                'field' => '<string>',
                'evaluation_call_id' => '<string>',
                'mode' => 'value'
        ]
    ],
    'summary_require_intersection' => true
  ]),
  CURLOPT_HTTPHEADER => [
    "Content-Type: application/json"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.example.com/v2/{entity}/{project}/eval_results/query"

	payload := strings.NewReader("{\n  \"evaluation_call_ids\": [\n    \"<string>\"\n  ],\n  \"evaluation_run_ids\": [\n    \"<string>\"\n  ],\n  \"filter_logic_operator\": \"or\",\n  \"filters\": [\n    {\n      \"query\": {\n        \"$expr\": {\n          \"$and\": [\n            {\n              \"$literal\": \"<string>\"\n            }\n          ]\n        }\n      },\n      \"evaluation_call_id\": \"<string>\"\n    }\n  ],\n  \"include_costs\": false,\n  \"include_predict_and_score_children\": true,\n  \"include_raw_data_rows\": false,\n  \"include_rows\": true,\n  \"include_summary\": false,\n  \"limit\": 123,\n  \"offset\": 0,\n  \"require_intersection\": false,\n  \"resolve_row_refs\": false,\n  \"sort_by\": [\n    {\n      \"field\": \"<string>\",\n      \"evaluation_call_id\": \"<string>\",\n      \"mode\": \"value\"\n    }\n  ],\n  \"summary_require_intersection\": true\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://api.example.com/v2/{entity}/{project}/eval_results/query")
  .header("Content-Type", "application/json")
  .body("{\n  \"evaluation_call_ids\": [\n    \"<string>\"\n  ],\n  \"evaluation_run_ids\": [\n    \"<string>\"\n  ],\n  \"filter_logic_operator\": \"or\",\n  \"filters\": [\n    {\n      \"query\": {\n        \"$expr\": {\n          \"$and\": [\n            {\n              \"$literal\": \"<string>\"\n            }\n          ]\n        }\n      },\n      \"evaluation_call_id\": \"<string>\"\n    }\n  ],\n  \"include_costs\": false,\n  \"include_predict_and_score_children\": true,\n  \"include_raw_data_rows\": false,\n  \"include_rows\": true,\n  \"include_summary\": false,\n  \"limit\": 123,\n  \"offset\": 0,\n  \"require_intersection\": false,\n  \"resolve_row_refs\": false,\n  \"sort_by\": [\n    {\n      \"field\": \"<string>\",\n      \"evaluation_call_id\": \"<string>\",\n      \"mode\": \"value\"\n    }\n  ],\n  \"summary_require_intersection\": true\n}")
  .asString();

require 'uri'
require 'net/http'

url = URI("https://api.example.com/v2/{entity}/{project}/eval_results/query")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Content-Type"] = 'application/json'
request.body = "{\n  \"evaluation_call_ids\": [\n    \"<string>\"\n  ],\n  \"evaluation_run_ids\": [\n    \"<string>\"\n  ],\n  \"filter_logic_operator\": \"or\",\n  \"filters\": [\n    {\n      \"query\": {\n        \"$expr\": {\n          \"$and\": [\n            {\n              \"$literal\": \"<string>\"\n            }\n          ]\n        }\n      },\n      \"evaluation_call_id\": \"<string>\"\n    }\n  ],\n  \"include_costs\": false,\n  \"include_predict_and_score_children\": true,\n  \"include_raw_data_rows\": false,\n  \"include_rows\": true,\n  \"include_summary\": false,\n  \"limit\": 123,\n  \"offset\": 0,\n  \"require_intersection\": false,\n  \"resolve_row_refs\": false,\n  \"sort_by\": [\n    {\n      \"field\": \"<string>\",\n      \"evaluation_call_id\": \"<string>\",\n      \"mode\": \"value\"\n    }\n  ],\n  \"summary_require_intersection\": true\n}"

response = http.request(request)
puts response.read_body

{
  "rows": [
    {
      "row_digest": "<string>",
      "evaluations": [
        {
          "evaluation_call_id": "<string>",
          "trials": [
            {
              "predict_and_score_call_id": "<string>",
              "genai_span_ref": [
                {
                  "span_id": "<string>",
                  "trace_id": "<string>"
                }
              ],
              "model_latency_seconds": 123,
              "model_output": null,
              "predict_call_id": "<string>",
              "scorer_call_ids": {},
              "scores": {},
              "total_cost": 123,
              "total_tokens": 123
            }
          ]
        }
      ],
      "raw_data_row": null
    }
  ],
  "total_rows": 123,
  "summary": {
    "evaluations": [
      {
        "evaluation_call_id": "<string>",
        "display_name": "<string>",
        "evaluation_ref": "<string>",
        "model_ref": "<string>",
        "predict_total_cost": 123,
        "predict_total_tokens": 123,
        "scorer_stats": [
          {
            "scorer_key": "<string>",
            "numeric_count": 0,
            "numeric_mean": 123,
            "pass_known_count": 0,
            "pass_rate": 123,
            "pass_signal_coverage": 123,
            "pass_true_count": 0,
            "path": "<string>",
            "trial_count": 0
          }
        ],
        "started_at": "<string>",
        "trace_id": "<string>",
        "trial_count": 0
      }
    ],
    "row_count": 0
  },
  "warnings": [
    "<string>"
  ]
}

{
  "detail": [
    {
      "loc": [
        "<string>"
      ],
      "msg": "<string>",
      "type": "<string>"
    }
  ]
}

POST

{entity}

{project}

eval_results

query

평가 결과 쿼리

curl --request POST \
  --url https://api.example.com/v2/{entity}/{project}/eval_results/query \
  --header 'Content-Type: application/json' \
  --data '
{
  "evaluation_call_ids": [
    "<string>"
  ],
  "evaluation_run_ids": [
    "<string>"
  ],
  "filter_logic_operator": "or",
  "filters": [
    {
      "query": {
        "$expr": {
          "$and": [
            {
              "$literal": "<string>"
            }
          ]
        }
      },
      "evaluation_call_id": "<string>"
    }
  ],
  "include_costs": false,
  "include_predict_and_score_children": true,
  "include_raw_data_rows": false,
  "include_rows": true,
  "include_summary": false,
  "limit": 123,
  "offset": 0,
  "require_intersection": false,
  "resolve_row_refs": false,
  "sort_by": [
    {
      "field": "<string>",
      "evaluation_call_id": "<string>",
      "mode": "value"
    }
  ],
  "summary_require_intersection": true
}
'

import requests

url = "https://api.example.com/v2/{entity}/{project}/eval_results/query"

payload = {
    "evaluation_call_ids": ["<string>"],
    "evaluation_run_ids": ["<string>"],
    "filter_logic_operator": "or",
    "filters": [
        {
            "query": { "$expr": { "$and": [{ "$literal": "<string>" }] } },
            "evaluation_call_id": "<string>"
        }
    ],
    "include_costs": False,
    "include_predict_and_score_children": True,
    "include_raw_data_rows": False,
    "include_rows": True,
    "include_summary": False,
    "limit": 123,
    "offset": 0,
    "require_intersection": False,
    "resolve_row_refs": False,
    "sort_by": [
        {
            "field": "<string>",
            "evaluation_call_id": "<string>",
            "mode": "value"
        }
    ],
    "summary_require_intersection": True
}
headers = {"Content-Type": "application/json"}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({
    evaluation_call_ids: ['<string>'],
    evaluation_run_ids: ['<string>'],
    filter_logic_operator: 'or',
    filters: [
      {
        query: {$expr: {$and: [{$literal: '<string>'}]}},
        evaluation_call_id: '<string>'
      }
    ],
    include_costs: false,
    include_predict_and_score_children: true,
    include_raw_data_rows: false,
    include_rows: true,
    include_summary: false,
    limit: 123,
    offset: 0,
    require_intersection: false,
    resolve_row_refs: false,
    sort_by: [{field: '<string>', evaluation_call_id: '<string>', mode: 'value'}],
    summary_require_intersection: true
  })
};

fetch('https://api.example.com/v2/{entity}/{project}/eval_results/query', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.example.com/v2/{entity}/{project}/eval_results/query",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'evaluation_call_ids' => [
        '<string>'
    ],
    'evaluation_run_ids' => [
        '<string>'
    ],
    'filter_logic_operator' => 'or',
    'filters' => [
        [
                'query' => [
                                '$expr' => [
                                                                '$and' => [
                                                                                                                                [
                                                                                                                                                                                                                                                                '$literal' => '<string>'
                                                                                                                                ]
                                                                ]
                                ]
                ],
                'evaluation_call_id' => '<string>'
        ]
    ],
    'include_costs' => false,
    'include_predict_and_score_children' => true,
    'include_raw_data_rows' => false,
    'include_rows' => true,
    'include_summary' => false,
    'limit' => 123,
    'offset' => 0,
    'require_intersection' => false,
    'resolve_row_refs' => false,
    'sort_by' => [
        [
                'field' => '<string>',
                'evaluation_call_id' => '<string>',
                'mode' => 'value'
        ]
    ],
    'summary_require_intersection' => true
  ]),
  CURLOPT_HTTPHEADER => [
    "Content-Type: application/json"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.example.com/v2/{entity}/{project}/eval_results/query"

	payload := strings.NewReader("{\n  \"evaluation_call_ids\": [\n    \"<string>\"\n  ],\n  \"evaluation_run_ids\": [\n    \"<string>\"\n  ],\n  \"filter_logic_operator\": \"or\",\n  \"filters\": [\n    {\n      \"query\": {\n        \"$expr\": {\n          \"$and\": [\n            {\n              \"$literal\": \"<string>\"\n            }\n          ]\n        }\n      },\n      \"evaluation_call_id\": \"<string>\"\n    }\n  ],\n  \"include_costs\": false,\n  \"include_predict_and_score_children\": true,\n  \"include_raw_data_rows\": false,\n  \"include_rows\": true,\n  \"include_summary\": false,\n  \"limit\": 123,\n  \"offset\": 0,\n  \"require_intersection\": false,\n  \"resolve_row_refs\": false,\n  \"sort_by\": [\n    {\n      \"field\": \"<string>\",\n      \"evaluation_call_id\": \"<string>\",\n      \"mode\": \"value\"\n    }\n  ],\n  \"summary_require_intersection\": true\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://api.example.com/v2/{entity}/{project}/eval_results/query")
  .header("Content-Type", "application/json")
  .body("{\n  \"evaluation_call_ids\": [\n    \"<string>\"\n  ],\n  \"evaluation_run_ids\": [\n    \"<string>\"\n  ],\n  \"filter_logic_operator\": \"or\",\n  \"filters\": [\n    {\n      \"query\": {\n        \"$expr\": {\n          \"$and\": [\n            {\n              \"$literal\": \"<string>\"\n            }\n          ]\n        }\n      },\n      \"evaluation_call_id\": \"<string>\"\n    }\n  ],\n  \"include_costs\": false,\n  \"include_predict_and_score_children\": true,\n  \"include_raw_data_rows\": false,\n  \"include_rows\": true,\n  \"include_summary\": false,\n  \"limit\": 123,\n  \"offset\": 0,\n  \"require_intersection\": false,\n  \"resolve_row_refs\": false,\n  \"sort_by\": [\n    {\n      \"field\": \"<string>\",\n      \"evaluation_call_id\": \"<string>\",\n      \"mode\": \"value\"\n    }\n  ],\n  \"summary_require_intersection\": true\n}")
  .asString();

require 'uri'
require 'net/http'

url = URI("https://api.example.com/v2/{entity}/{project}/eval_results/query")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Content-Type"] = 'application/json'
request.body = "{\n  \"evaluation_call_ids\": [\n    \"<string>\"\n  ],\n  \"evaluation_run_ids\": [\n    \"<string>\"\n  ],\n  \"filter_logic_operator\": \"or\",\n  \"filters\": [\n    {\n      \"query\": {\n        \"$expr\": {\n          \"$and\": [\n            {\n              \"$literal\": \"<string>\"\n            }\n          ]\n        }\n      },\n      \"evaluation_call_id\": \"<string>\"\n    }\n  ],\n  \"include_costs\": false,\n  \"include_predict_and_score_children\": true,\n  \"include_raw_data_rows\": false,\n  \"include_rows\": true,\n  \"include_summary\": false,\n  \"limit\": 123,\n  \"offset\": 0,\n  \"require_intersection\": false,\n  \"resolve_row_refs\": false,\n  \"sort_by\": [\n    {\n      \"field\": \"<string>\",\n      \"evaluation_call_id\": \"<string>\",\n      \"mode\": \"value\"\n    }\n  ],\n  \"summary_require_intersection\": true\n}"

response = http.request(request)
puts response.read_body

{
  "rows": [
    {
      "row_digest": "<string>",
      "evaluations": [
        {
          "evaluation_call_id": "<string>",
          "trials": [
            {
              "predict_and_score_call_id": "<string>",
              "genai_span_ref": [
                {
                  "span_id": "<string>",
                  "trace_id": "<string>"
                }
              ],
              "model_latency_seconds": 123,
              "model_output": null,
              "predict_call_id": "<string>",
              "scorer_call_ids": {},
              "scores": {},
              "total_cost": 123,
              "total_tokens": 123
            }
          ]
        }
      ],
      "raw_data_row": null
    }
  ],
  "total_rows": 123,
  "summary": {
    "evaluations": [
      {
        "evaluation_call_id": "<string>",
        "display_name": "<string>",
        "evaluation_ref": "<string>",
        "model_ref": "<string>",
        "predict_total_cost": 123,
        "predict_total_tokens": 123,
        "scorer_stats": [
          {
            "scorer_key": "<string>",
            "numeric_count": 0,
            "numeric_mean": 123,
            "pass_known_count": 0,
            "pass_rate": 123,
            "pass_signal_coverage": 123,
            "pass_true_count": 0,
            "path": "<string>",
            "trial_count": 0
          }
        ],
        "started_at": "<string>",
        "trace_id": "<string>",
        "trial_count": 0
      }
    ],
    "row_count": 0
  },
  "warnings": [
    "<string>"
  ]
}

{
  "detail": [
    {
      "loc": [
        "<string>"
      ],
      "msg": "<string>",
      "type": "<string>"
    }
  ]
}

경로 매개변수

entity

string

필수

project

string

필수

본문

application/json

evaluation_call_ids

string[] | null

포함할 Evaluation 루트 호출 ID입니다.

evaluation_run_ids

string[] | null

Evaluation Runs API의 evaluation 호출 ID에 대한 별칭입니다.

filter_logic_operator

enum<string>

기본값:or

evaluations 전반에서 필터를 결합하는 방법입니다. 'and'(모두 일치 - 행이 모든 eval에서 일치해야 함) 또는 'or'(하나라도 일치 - 행이 하나 이상의 eval에서 일치해야 함)를 사용합니다. 기본값은 'or'(하나라도 일치)입니다.

사용 가능한 옵션:

and,

or

filters

EvalResultsFilter · object[] | null

그룹화된 행에 적용되는 필터입니다. 여러 필터는 AND로 결합됩니다.

Show child attributes

include_costs

boolean

기본값:false

true이면 요약에서 predict 전용 predict_total_cost를 보고할 수 있도록 predict-and-score 하위 Call에 비용 정보를 추가합니다. 옵트인 방식이므로 다른 호출자에서는 비용 계산을 건너뜁니다.

include_predict_and_score_children

boolean

기본값:true

true(기본값)로 설정하면 각 predict_and_score Call의 하위 Call(predict/score)을 가져와 predict_call_id, scorer_call_ids 및 더 정확한 지연 시간/토큰 데이터를 채웁니다. false이면 이러한 필드는 predict_and_score Call 자체에서 파생됩니다(predict_call_id 및 scorer_call_ids는 null/empty가 됨).

include_raw_data_rows

boolean

기본값:false

true이면 각 결과 행의 raw_data_row를 채웁니다. 인라인 행은 해당 dict 값으로 반환되며, dataset를 참조하는 행은 resolve_row_refs도 true가 아닌 한 ref 문자열로 반환됩니다.

include_rows

boolean

기본값:true

true이면 그룹화된 행/트라이얼 데이터를 rows에 포함하고, 요청된 행 수준 뷰에 대해 total_rows를 계산합니다.

include_summary

boolean

기본값:false

true이면 집계된 scorer/evaluation summary 데이터를 summary에 포함합니다.

limit

integer | null

그룹화와 교집합 적용 후 적용되는 선택적 행 수준 페이지 크기입니다.

offset

integer

기본값:0

그룹화와 교집합 적용 후 적용되는 선택적 행 수준 페이지 오프셋입니다.

require_intersection

boolean

기본값:false

true이면 요청된 모든 evaluation에 있는 행만 포함합니다.

resolve_row_refs

boolean

기본값:false

true이면(include_raw_data_rows=True 필요) 테이블 조회를 통해 dataset 행 참조 문자열을 실제 행 데이터로 해석합니다. false이면 dataset 행 ref를 있는 그대로 반환합니다.

sort_by

EvalResultsSortBy · object[] | null

결과 행의 정렬 지정입니다. 지원되는 필드 접두사: scores., inputs., outputs.. null이면 행이 row_digest ASC 기준으로 정렬됩니다.

Show child attributes

summary_require_intersection

boolean | null

summary 섹션에 대한 선택적 교집합 동작입니다. null이면 require_intersection 값을 사용합니다.

응답

성공 응답

rows

EvalResultsRow · object[]

필수

Show child attributes

total_rows

integer

필수

summary

EvalResultsSummaryRes · object | null

Show child attributes

warnings

string[]

치명적이지 않은 경고입니다(예: dataset 행 ref를 해석하지 못한 경우).

데이터셋 조회

평가 run 목록

W&B Models

W&B Weave

Serverless Inference

Serverless Training

경로 매개변수

본문

응답