Compare the baseline and the winning setup of one campaign on cost, speed, and answer quality
GET
/
api
/
v1
/
optimization
/
campaigns
/
{id}
/
report
Compare the baseline and the winning setup of one campaign on cost, speed, and answer quality
curl --request GET \
--url http://127.0.0.1:18000/api/v1/optimization/campaigns/{id}/report \
--header 'Authorization: Bearer <token>'import requests
url = "http://127.0.0.1:18000/api/v1/optimization/campaigns/{id}/report"
headers = {"Authorization": "Bearer <token>"}
response = requests.get(url, headers=headers)
print(response.text)const options = {method: 'GET', headers: {Authorization: 'Bearer <token>'}};
fetch('http://127.0.0.1:18000/api/v1/optimization/campaigns/{id}/report', options)
.then(res => res.json())
.then(res => console.log(res))
.catch(err => console.error(err));package main
import (
"fmt"
"net/http"
"io"
)
func main() {
url := "http://127.0.0.1:18000/api/v1/optimization/campaigns/{id}/report"
req, _ := http.NewRequest("GET", url, nil)
req.Header.Add("Authorization", "Bearer <token>")
res, _ := http.DefaultClient.Do(req)
defer res.Body.Close()
body, _ := io.ReadAll(res.Body)
fmt.Println(string(body))
}{
"report": {
"algorithm_version": "<string>",
"appendix": {
"baseline": {
"accelerator": "<string>",
"accelerator_class": "gpu",
"accelerator_count": 123,
"args": [
"<string>"
],
"candidate_id": "<string>",
"engine": "<string>",
"engine_version": "<string>",
"hourly_usd_per_replica": 123,
"label": "<string>",
"max_replicas": 123,
"min_replicas": 123,
"model": "<string>",
"provider": "<string>",
"compute_mode": "<string>",
"deployment_name": "<string>",
"env": {},
"image": "<string>",
"price_source": "<string>",
"region": "<string>",
"revision_id": "<string>"
},
"evidence": [
{
"digest": "<string>",
"kind": "<string>",
"id": "<string>",
"subject": "<string>"
}
],
"winner": {
"accelerator": "<string>",
"accelerator_class": "gpu",
"accelerator_count": 123,
"args": [
"<string>"
],
"candidate_id": "<string>",
"engine": "<string>",
"engine_version": "<string>",
"hourly_usd_per_replica": 123,
"label": "<string>",
"max_replicas": 123,
"min_replicas": 123,
"model": "<string>",
"provider": "<string>",
"compute_mode": "<string>",
"deployment_name": "<string>",
"env": {},
"image": "<string>",
"price_source": "<string>",
"region": "<string>",
"revision_id": "<string>"
}
},
"bottleneck": {
"gpu_telemetry": "<string>",
"signals": [
{
"baseline": 123,
"name": "<string>",
"stat": "<string>",
"winner": 123,
"unit": "<string>"
}
],
"source": "<string>",
"state": "measured",
"reason": "<string>"
},
"campaign_id": "<string>",
"campaign_state": "<string>",
"candidates": [
{
"cost_per_million_output_tokens": 123,
"error_rate_percent": 123,
"id": "<string>",
"label": "<string>",
"latency_p95_ms": 123,
"output_tokens_per_request": 123,
"output_tokens_per_second": 123,
"requests_per_second": 123,
"role": "baseline",
"state": "<string>",
"tpot_p95_ms": 123,
"ttft_p95_ms": 123,
"versus_baseline": {
"change_percent": 123,
"decision": "improved",
"metric": "<string>",
"noise_floor_percent": 123,
"basis": "interleaved_control"
},
"failure_code": "<string>",
"guards": [
{
"baseline": 123,
"basis": "same_window",
"code": "measured_error_rate_regression",
"metric": "error_rate_percent",
"reason": "<string>",
"value": 123
}
],
"reason": "<string>",
"regressions": [
{
"change_percent": 123,
"decision": "improved",
"metric": "<string>",
"noise_floor_percent": 123,
"basis": "interleaved_control"
}
]
}
],
"finished": true,
"intent": "<string>",
"metrics": [
{
"baseline": 123,
"basis": "interleaved_control",
"change_percent": 123,
"decision": "improved",
"higher_is_better": true,
"key": "cost_per_million_output_tokens",
"label": "<string>",
"noise_floor_percent": 123,
"unit": "<string>",
"winner": 123,
"reason": "<string>"
}
],
"model_identity": "<string>",
"monthly": {
"baseline_usd": 123,
"basis": "replayed_traffic",
"capture_cut_off": true,
"capture_window_seconds": 123,
"change_percent": 123,
"change_usd": 123,
"decision": "improved",
"hours_per_month": 123,
"monthly_output_tokens": 123,
"output_tokens_in_window": 123,
"requests_in_window": 123,
"window_seconds": 123,
"winner_usd": 123,
"assumption": "<string>",
"reason": "<string>",
"trace_digest": "<string>"
},
"noise": {
"basis": "interleaved_control",
"candidate_runs": 123,
"control_drift_percent": 123,
"control_runs": 123,
"drift_limit_percent": 123,
"objective_change_percent": 123,
"objective_decision": "improved",
"objective_noise_floor_percent": 123,
"candidate_id": "<string>",
"objective_metric": "<string>",
"reason": "<string>"
},
"objective": "<string>",
"outcome": {
"activated": true,
"baseline_candidate_id": "<string>",
"baseline_label": "<string>",
"headline": "<string>",
"kind": "winner_selected",
"winner_candidate_id": "<string>",
"winner_label": "<string>"
},
"quality": {
"baseline_score": 123,
"basis": "release_guard",
"block_threshold_percent": 123,
"delta": 123,
"evals": [
{
"baseline_score": 123,
"delta": 123,
"evaluator": "<string>",
"message": "<string>",
"regression_percent": 123,
"suite": "<string>",
"verdict": "PASS",
"winner_score": 123,
"follow_up": "<string>"
}
],
"regression_percent": 123,
"verdict": "PASS",
"watch_threshold_percent": 123,
"winner_score": 123,
"eval": "<string>",
"reason": "<string>",
"release_guard_evaluation_id": "<string>",
"source": "infercrane"
},
"report_digest": "<string>",
"schema_version": "infercrane.optimization-report/v1",
"simulated_accelerator": true,
"unknown": [
"<string>"
],
"warnings": [
"<string>"
],
"workload": {
"concurrency": 123,
"requests_per_run": 123,
"source": "<string>",
"method": "<string>",
"profile": "<string>"
},
"banner": "<string>"
}
}{
"error": {
"code": "<string>",
"message": "<string>",
"category": "<string>",
"remediation": "<string>",
"request_id": "<string>",
"retryable": true
}
}⌘I
Compare the baseline and the winning setup of one campaign on cost, speed, and answer quality
curl --request GET \
--url http://127.0.0.1:18000/api/v1/optimization/campaigns/{id}/report \
--header 'Authorization: Bearer <token>'import requests
url = "http://127.0.0.1:18000/api/v1/optimization/campaigns/{id}/report"
headers = {"Authorization": "Bearer <token>"}
response = requests.get(url, headers=headers)
print(response.text)const options = {method: 'GET', headers: {Authorization: 'Bearer <token>'}};
fetch('http://127.0.0.1:18000/api/v1/optimization/campaigns/{id}/report', options)
.then(res => res.json())
.then(res => console.log(res))
.catch(err => console.error(err));package main
import (
"fmt"
"net/http"
"io"
)
func main() {
url := "http://127.0.0.1:18000/api/v1/optimization/campaigns/{id}/report"
req, _ := http.NewRequest("GET", url, nil)
req.Header.Add("Authorization", "Bearer <token>")
res, _ := http.DefaultClient.Do(req)
defer res.Body.Close()
body, _ := io.ReadAll(res.Body)
fmt.Println(string(body))
}{
"report": {
"algorithm_version": "<string>",
"appendix": {
"baseline": {
"accelerator": "<string>",
"accelerator_class": "gpu",
"accelerator_count": 123,
"args": [
"<string>"
],
"candidate_id": "<string>",
"engine": "<string>",
"engine_version": "<string>",
"hourly_usd_per_replica": 123,
"label": "<string>",
"max_replicas": 123,
"min_replicas": 123,
"model": "<string>",
"provider": "<string>",
"compute_mode": "<string>",
"deployment_name": "<string>",
"env": {},
"image": "<string>",
"price_source": "<string>",
"region": "<string>",
"revision_id": "<string>"
},
"evidence": [
{
"digest": "<string>",
"kind": "<string>",
"id": "<string>",
"subject": "<string>"
}
],
"winner": {
"accelerator": "<string>",
"accelerator_class": "gpu",
"accelerator_count": 123,
"args": [
"<string>"
],
"candidate_id": "<string>",
"engine": "<string>",
"engine_version": "<string>",
"hourly_usd_per_replica": 123,
"label": "<string>",
"max_replicas": 123,
"min_replicas": 123,
"model": "<string>",
"provider": "<string>",
"compute_mode": "<string>",
"deployment_name": "<string>",
"env": {},
"image": "<string>",
"price_source": "<string>",
"region": "<string>",
"revision_id": "<string>"
}
},
"bottleneck": {
"gpu_telemetry": "<string>",
"signals": [
{
"baseline": 123,
"name": "<string>",
"stat": "<string>",
"winner": 123,
"unit": "<string>"
}
],
"source": "<string>",
"state": "measured",
"reason": "<string>"
},
"campaign_id": "<string>",
"campaign_state": "<string>",
"candidates": [
{
"cost_per_million_output_tokens": 123,
"error_rate_percent": 123,
"id": "<string>",
"label": "<string>",
"latency_p95_ms": 123,
"output_tokens_per_request": 123,
"output_tokens_per_second": 123,
"requests_per_second": 123,
"role": "baseline",
"state": "<string>",
"tpot_p95_ms": 123,
"ttft_p95_ms": 123,
"versus_baseline": {
"change_percent": 123,
"decision": "improved",
"metric": "<string>",
"noise_floor_percent": 123,
"basis": "interleaved_control"
},
"failure_code": "<string>",
"guards": [
{
"baseline": 123,
"basis": "same_window",
"code": "measured_error_rate_regression",
"metric": "error_rate_percent",
"reason": "<string>",
"value": 123
}
],
"reason": "<string>",
"regressions": [
{
"change_percent": 123,
"decision": "improved",
"metric": "<string>",
"noise_floor_percent": 123,
"basis": "interleaved_control"
}
]
}
],
"finished": true,
"intent": "<string>",
"metrics": [
{
"baseline": 123,
"basis": "interleaved_control",
"change_percent": 123,
"decision": "improved",
"higher_is_better": true,
"key": "cost_per_million_output_tokens",
"label": "<string>",
"noise_floor_percent": 123,
"unit": "<string>",
"winner": 123,
"reason": "<string>"
}
],
"model_identity": "<string>",
"monthly": {
"baseline_usd": 123,
"basis": "replayed_traffic",
"capture_cut_off": true,
"capture_window_seconds": 123,
"change_percent": 123,
"change_usd": 123,
"decision": "improved",
"hours_per_month": 123,
"monthly_output_tokens": 123,
"output_tokens_in_window": 123,
"requests_in_window": 123,
"window_seconds": 123,
"winner_usd": 123,
"assumption": "<string>",
"reason": "<string>",
"trace_digest": "<string>"
},
"noise": {
"basis": "interleaved_control",
"candidate_runs": 123,
"control_drift_percent": 123,
"control_runs": 123,
"drift_limit_percent": 123,
"objective_change_percent": 123,
"objective_decision": "improved",
"objective_noise_floor_percent": 123,
"candidate_id": "<string>",
"objective_metric": "<string>",
"reason": "<string>"
},
"objective": "<string>",
"outcome": {
"activated": true,
"baseline_candidate_id": "<string>",
"baseline_label": "<string>",
"headline": "<string>",
"kind": "winner_selected",
"winner_candidate_id": "<string>",
"winner_label": "<string>"
},
"quality": {
"baseline_score": 123,
"basis": "release_guard",
"block_threshold_percent": 123,
"delta": 123,
"evals": [
{
"baseline_score": 123,
"delta": 123,
"evaluator": "<string>",
"message": "<string>",
"regression_percent": 123,
"suite": "<string>",
"verdict": "PASS",
"winner_score": 123,
"follow_up": "<string>"
}
],
"regression_percent": 123,
"verdict": "PASS",
"watch_threshold_percent": 123,
"winner_score": 123,
"eval": "<string>",
"reason": "<string>",
"release_guard_evaluation_id": "<string>",
"source": "infercrane"
},
"report_digest": "<string>",
"schema_version": "infercrane.optimization-report/v1",
"simulated_accelerator": true,
"unknown": [
"<string>"
],
"warnings": [
"<string>"
],
"workload": {
"concurrency": 123,
"requests_per_run": 123,
"source": "<string>",
"method": "<string>",
"profile": "<string>"
},
"banner": "<string>"
}
}{
"error": {
"code": "<string>",
"message": "<string>",
"category": "<string>",
"remediation": "<string>",
"request_id": "<string>",
"retryable": true
}
}