Kiln Documentation

HTTP API contract

Look up every production HTTP operation, transport, request and response schema, handler owner, error, and trust boundary.

Canonical inventory of every production HTTP route, method, transport, handler owner, request type, response type, status, media type, path/header/query input, and structured error envelope. Every named payload component delegates its complete field contract to a canonical generated schema. Kiln does not authenticate HTTP requests. The server binds to loopback by default; put any remotely accessible deployment behind an authenticated reverse proxy on a trusted network.

Contract status

This contract contains 111 paths and 125 operations. The aggregate field status is complete; every operation and transport remains canonical while the remaining payload components are migrated. Of 144 top-level payload components, 144 are field-complete and 0 remain migration pending.

OpenAPI identity

Property Value
Title Kiln HTTP API
Version 1.0.0
OpenAPI 3.1.1
JSON Schema dialect https://json-schema.org/draft/2020-12/schema
Servers http://127.0.0.1:8420
Authentication none declared; protect remote deployments with an authenticated reverse proxy
Method counts DELETE 13, GET 59, POST 52, PUT 1

ui

Embedded dashboard, redirects, and static assets.

Method Path Summary Request inputs Successful response Owner
GET / Redirect the server root to the embedded dashboard none 303: Redirect (text/html) ui::redirect_root_to_ui
GET /ui Redirect /ui to /ui/ none 307: Redirect (text/html) ui::redirect_ui_to_slash
GET /ui/ Serve the embedded dashboard none 200: HtmlDocument (text/html) ui::serve_ui
GET /ui/app.js Serve the dashboard JavaScript none 200: JavaScriptSource (application/javascript) ui::serve_app_js
GET /ui/demo.js Serve the dashboard demo-fixture JavaScript none 200: JavaScriptSource (application/javascript) ui::serve_demo_js
GET /ui/fonts/InterVariable.woff2 Serve the Inter Variable UI typeface (weights 100–900) none 200: BinaryArchive (font/woff2) ui::serve_font_inter
GET /ui/fonts/JetBrainsMonoVariable.ttf Serve the JetBrains Mono Variable code typeface (weights 100–800) none 200: BinaryArchive (font/ttf) ui::serve_font_jbmono
GET /ui/styles.css Serve the dashboard stylesheet none 200: CssStylesheet (text/css) ui::serve_styles_css
GET /ui/vendor/xterm-addon-fit.js Serve the bundled xterm fit add-on none 200: JavaScriptSource (application/javascript) ui::serve_xterm_fit_js
GET /ui/vendor/xterm.css Serve the bundled xterm stylesheet none 200: CssStylesheet (text/css) ui::serve_xterm_css
GET /ui/vendor/xterm.js Serve the bundled xterm JavaScript none 200: JavaScriptSource (application/javascript) ui::serve_xterm_js

observability

Health, metrics, configuration, model identity, and runtime statistics.

Method Path Summary Request inputs Successful response Owner
GET /health Get server readiness and runtime diagnostics none 200: HealthResponse (application/json) health::health
GET /metrics Export Prometheus metrics none 200: PrometheusText (text/plain) metrics::metrics_handler
GET /v1/config Get the resolved runtime configuration none 200: ConfigResponse (application/json) config::get_config
GET /v1/debug/model-state Get trusted model-state diagnostics none 200: ModelStateResponse (application/json) debug_model_state::model_state
GET /v1/health Get server readiness and runtime diagnostics (v1 alias) none 200: HealthResponse (application/json) health::health
GET /v1/models List models available through the inference API none 200: ModelsResponse (application/json) models::list_models
GET /v1/stats/decode Get aggregate decode statistics none 200: DecodeStatsSnapshot (application/json) stats::get_decode_stats
GET /v1/stats/recent-requests List recent inference request records none 200: Vec (application/json) stats::get_recent_requests

inference

OpenAI-compatible chat, prompt scoring, and batch text generation.

Method Path Summary Request inputs Successful response Owner
POST /v1/chat/completions Create an OpenAI-compatible chat completion body: ChatCompletionRequest (application/json) 200: ChatCompletionResponse (application/json, text/event-stream) completions::chat_completions
POST /v1/completions Score prompt log probabilities body: TextCompletionRequest (application/json) 200: TextCompletionResponse (application/json) completions::completions
POST /v1/completions/batch Generate text for a batch of prompts body: BatchCompletionRequest (application/json) 200: BatchCompletionResponse (application/json) completions::batch_completions

adapters

Adapter discovery, loading, transfer, merge, deletion, and receipts.

Method Path Summary Request inputs Successful response Owner
GET /v1/adapters List installed adapters none 200: AdaptersResponse (application/json) adapters::list_adapters
POST /v1/adapters/load Load an adapter for inference body: LoadAdapterRequest (application/json) 200: LoadAdapterResponse (application/json) adapters::load_adapter
POST /v1/adapters/merge Merge adapters into a composed adapter body: MergeAdapterRequest (application/json) 200: MergeAdapterResponse (application/json) adapters::merge_adapters
POST /v1/adapters/unload Unload the active adapter none 200: UnloadAdapterResponse (application/json) adapters::unload_adapter
POST /v1/adapters/upload Upload and validate an adapter archive body: AdapterUploadMultipart (multipart/form-data) 200: UploadAdapterResponse (application/json) adapters::upload_adapter
DELETE /v1/adapters/{name} Delete a stored adapter path: name 200: DeleteAdapterResponse (application/json) adapters::delete_adapter
GET /v1/adapters/{name}/detail Get adapter metadata and verification detail path: name 200: AdapterDetail (application/json) adapters::adapter_detail
GET /v1/adapters/{name}/download Download an adapter archive path: name 200: BinaryArchive (application/gzip) adapters::download_adapter
GET /v1/adapters/{name}/receipt Get an adapter training receipt path: name 200: kiln_train::AdapterReceipt (application/json) adapters::adapter_receipt

corrections

Stored correction rows and their training status.

Method Path Summary Request inputs Successful response Owner
GET /v1/corrections List stored corrections query: ListQuery 200: ListResponse (application/json) corrections::list_corrections
POST /v1/corrections Create or replace a correction body: CorrectionRow (application/json) 200: CorrectionRow (application/json) corrections::upsert_correction
DELETE /v1/corrections Delete every stored correction none 200: ClearCorrectionsResponse (application/json) corrections::clear_corrections
POST /v1/corrections/mark_trained Mark corrections as used for training body: MarkTrainedRequest (application/json) 200: MarkTrainedResponse (application/json) corrections::mark_trained
DELETE /v1/corrections/{request_id} Delete one correction by request ID path: request_id 200: DeleteCorrectionResponse (application/json) corrections::delete_correction

teachers

Identity-bound remote teachers and teacher-logit cache artifacts.

Method Path Summary Request inputs Successful response Owner
GET /v1/cache/export Download the teacher-logit cache archive none 200: BinaryArchive (application/gzip) cache::cache_export
GET /v1/cache/stats Get teacher-logit cache statistics none 200: CacheStatsResponse (application/json) cache::cache_stats
GET /v1/teachers List registered remote teachers none 200: TeachersListResponse (application/json) teachers::list_teachers
POST /v1/teachers Register an identity-bound remote teacher body: RegisterTeacherRequest (application/json) 200: TeacherEntry (application/json) teachers::register_teacher
DELETE /v1/teachers/{alias} Delete a registered remote teacher path: alias 200: DeleteTeacherResponse (application/json) teachers::delete_teacher

agents

Embedded agent runs, imported traces, terminal access, and self-improvement workflows.

Method Path Summary Request inputs Successful response Owner
POST /v1/agent/judge_distill Distill a judge adapter from captured judgments body: JudgeDistillRequest (application/json) 200: JudgeDistillResponse (application/json) self_improve::judge_distill
POST /v1/agent/judge_drift_check Validate a judge drift-check request (scoring is unavailable) body: JudgeDriftCheckRequest (application/json) unspecified self_improve::judge_drift_check
GET /v1/agent/runs List embedded agent runs query: ListRunsQuery 200: AgentRunListResponse (application/json) agent_runs::list_runs
POST /v1/agent/runs Create an embedded agent run body: CreateRunRequest (application/json) 200: AgentRunRecord (application/json) agent_runs::create_run
GET /v1/agent/runs/status Get aggregate agent-run status none 200: AgentRunsStatusResponse (application/json) agent_runs::runs_status
GET /v1/agent/runs/{id} Get one agent run path: id 200: AgentRunRecord (application/json) agent_runs::get_run
POST /v1/agent/runs/{id}/abort Abort an agent run path: id 200: AgentRunAbortResponse (application/json) agent_runs::abort_run
GET /v1/agent/runs/{id}/events List events for an agent run path: id; query: EventsQuery 200: AgentRunEventsResponse (application/json) agent_runs::run_events
POST /v1/agent/runs/{id}/follow_up Queue a follow-up message for an agent run path: id; body: MessageRequest (application/json) 200: AgentRunQueuedResponse (application/json) agent_runs::follow_up_run
POST /v1/agent/runs/{id}/steer Steer an active agent run path: id; body: MessageRequest (application/json) 200: AgentRunQueuedResponse (application/json) agent_runs::steer_run
POST /v1/agent/self_improve Run the self-improvement workflow body: SelfImproveRequest (application/json) 200: SelfImproveResponse (application/json) self_improve::self_improve
GET /v1/agent/traces List discovered agent traces none 200: AgentTracesListResponse (application/json) agent_traces::list_traces
POST /v1/agent/traces/discover Discover agent traces from the configured session root body: DiscoverRequest (application/json) 200: DiscoverResponse (application/json) agent_traces::discover_traces
GET /v1/agent/traces/{id} Get one agent trace path: id 200: AgentTrace (application/json) agent_traces::get_trace
GET /v1/terminal/status Get embedded-terminal availability and access status none 200: TerminalStatusResponse (application/json) terminal::terminal_status
GET /v1/terminal/ws Open the embedded-terminal WebSocket when access policy permits (WebSocket) headers: Origin 101: WebSocketUpgrade (application/octet-stream) terminal::terminal_ws

training

Native SFT, GRPO, OPD, distillation, queue, and job lifecycle.

Method Path Summary Request inputs Successful response Owner
POST /v1/adapters/distill_merge Submit an adapter-distillation merge job body: DistillMergeRequest (application/json) 200: TrainingResponse (application/json) training::submit_distill_merge
POST /v1/distill/pump Submit a frontier-teacher distillation job body: DistillPumpRequest (application/json) 200: TrainingResponse (application/json) training::submit_distill_pump
POST /v1/distill/refresh Submit a dataset-refresh distillation job body: DistillRefreshRequest (application/json) 200: TrainingResponse (application/json) training::submit_distill_refresh
POST /v1/distill/self Submit a self-distillation job body: DistillSelfRequest (application/json) 200: TrainingResponse (application/json) training::submit_distill_self
POST /v1/train/agentic Submit an agentic GRPO training job body: GrpoRequest (application/json) 200: TrainingResponse (application/json) training::submit_grpo
POST /v1/train/grpo Submit a GRPO training job body: GrpoRequest (application/json) 200: TrainingResponse (application/json) training::submit_grpo
GET /v1/train/jobs/{job_id} Get training-job configuration, status, and artifacts path: job_id 200: TrainingJobDetail (application/json) training::job_detail
DELETE /v1/train/jobs/{job_id} Delete an archived training job path: job_id 200: DeleteTrainingJobResponse (application/json) training::delete_archived_job
POST /v1/train/opd Submit an online policy distillation job body: OpdRequest (application/json) 200: TrainingResponse (application/json) training::submit_opd
GET /v1/train/queue List queued and running training jobs none 200: QueueResponse (application/json) training::list_queue
DELETE /v1/train/queue/{job_id} Cancel a queued training job path: job_id 200: CancelTrainingJobResponse (application/json) training::cancel_queued_job
POST /v1/train/sft Submit a supervised fine-tuning job body: SftRequest (application/json) 200: TrainingResponse (application/json) training::submit_sft
GET /v1/train/status List tracked training-job statuses none 200: Vec (application/json) training::training_status
GET /v1/train/status/{job_id} Get one training-job status path: job_id 200: TrainingStatus (application/json) training::job_status
POST /v1/training/grpo Submit a GRPO training job (compatibility alias) body: GrpoRequest (application/json) 200: TrainingResponse (application/json) training::submit_grpo

openenv

Native OpenEnv discovery, stateful rollout collection, replay artifacts, and GRPO handoff.

Method Path Summary Request inputs Successful response Owner
POST /v1/openenv/inspect Discover and content-address OpenEnv servers body: OpenEnvInspectRequest (application/json) 200: OpenEnvInspectResponse (application/json) openenv::inspect
GET /v1/openenv/runs List retained OpenEnv rollout and training runs none 200: OpenEnvRunList (application/json) openenv::list_runs
POST /v1/openenv/runs Create or idempotently recover a persisted OpenEnv rollout or rollout-and-train run body: OpenEnvRunRequest (application/json) 200: OpenEnvRunStatus (application/json); 202: OpenEnvRunStatus (application/json) openenv::create_run
GET /v1/openenv/runs/{run_id} Get one OpenEnv run path: run_id 200: OpenEnvRunStatus (application/json) openenv::get_run
DELETE /v1/openenv/runs/{run_id} Cooperatively cancel an active OpenEnv run path: run_id 200: OpenEnvRunStatus (application/json) openenv::cancel_run
GET /v1/openenv/runs/{run_id}/artifacts/{kind} Verify and stream a manifest-published OpenEnv training or held-out-evaluation artifact path: run_id, kind 200: Body (application/json, application/x-ndjson) openenv::download_artifact
POST /v1/openenv/tasks Discover and page optional OpenEnv Task API catalogs body: OpenEnvTaskCatalogRequest (application/json) 200: OpenEnvTaskCatalogResponse (application/json) openenv::inspect_tasks
POST /v1/openenv/training/preflight Preflight direct OpenEnv training before environment collection body: OpenEnvTrainingPreflightRequest (application/json) 200: OpenEnvTrainingPreflightReceipt (application/json) openenv::preflight_training

hf-trl

Immutable Hugging Face/TRL exports and verified PEFT imports.

Method Path Summary Request inputs Successful response Owner
GET /v1/train/hf/exports List immutable Hugging Face training exports none 200: ExportList (application/json) hf_trl::list_exports
GET /v1/train/hf/exports/{name} Get a Hugging Face training export manifest path: name 200: ExportDetail (application/json) hf_trl::get_export
DELETE /v1/train/hf/exports/{name} Delete a Hugging Face training export path: name; headers: If-Match 200: DeleteExportResponse (application/json) hf_trl::delete_export
GET /v1/train/hf/exports/{name}/download Download a Hugging Face training export path: name 200: BinaryArchive (application/gzip) hf_trl::download_export
POST /v1/train/hf/grpo/exports Create an immutable GRPO export for Hugging Face body: GrpoExportRequest (application/json) 201: ExportSummary (application/json) hf_trl::create_grpo_export
POST /v1/train/hf/peft/imports/{name} Import a verified HF/TRL PEFT result path: name; headers: Content-Type; body: BinaryArchive (application/gzip) 201: ImportPeftResponse (application/json) hf_trl_import::import_peft
POST /v1/train/hf/sft/exports Create an immutable SFT export for Hugging Face body: SftExportRequest (application/json) 201: ExportSummary (application/json) hf_trl::create_sft_export

evals

Eval suites, datasets, jobs, synthesis, and judgment datasets.

Method Path Summary Request inputs Successful response Owner
POST /v1/eval/compare Submit a paired eval comparison body: EvalCompareSpec (application/json) 200: EvalRunResponse (application/json) eval::submit_compare
GET /v1/eval/datasets List eval datasets none 200: DatasetListResponse (application/json) eval::list_datasets
POST /v1/eval/datasets/upload Upload and validate an eval dataset body: DatasetUploadMultipart (multipart/form-data) 200: DatasetManifest (application/json) eval::upload_dataset
GET /v1/eval/datasets/{name} Get an eval dataset manifest path: name 200: DatasetManifest (application/json) eval::get_dataset
DELETE /v1/eval/datasets/{name} Delete an eval dataset path: name 200: DeleteDatasetResponse (application/json) eval::delete_dataset
POST /v1/eval/datasets/{name}/preview Preview synthetic rows without saving them path: name; body: SynthesisPreviewBody (application/json) 200: SynthesisPreview (application/json) eval::synthesize_dataset_preview
GET /v1/eval/datasets/{name}/rows List rows from an eval dataset path: name; query: DatasetRowsQuery 200: Vec<serde_json::Value> (application/json) eval::dataset_rows
GET /v1/eval/datasets/{name}/split Get an eval dataset's split manifest path: name 200: DatasetSplitManifest (application/json) eval::get_dataset_split
PUT /v1/eval/datasets/{name}/split Replace an eval dataset's split policy path: name; body: DatasetSplitConfig (application/json) 200: DatasetSplitManifest (application/json) eval::update_dataset_split
POST /v1/eval/datasets/{name}/synthesize Generate and save synthetic dataset rows path: name; body: SynthesizeBody (application/json) 200: SynthesizeDatasetResponse (application/json) eval::synthesize_dataset
GET /v1/eval/jobs List eval jobs none 200: EvalJobListResponse (application/json) eval::list_jobs
GET /v1/eval/jobs/{job_id} Get one eval job and its results path: job_id 200: EvalResult (application/json) eval::get_job
DELETE /v1/eval/jobs/{job_id} Cancel a queued or running eval job path: job_id 200: CancelEvalJobResponse (application/json) eval::cancel_job
POST /v1/eval/jobs/{job_id}/replay Strictly replay one retained eval run path: job_id; body: ReplayBody (application/json) 200: EvalRunResponse (application/json) eval::replay_job
POST /v1/eval/jobs/{job_id}/rerun Rerun the failed examples from an eval job path: job_id; body: RerunBody (application/json) 200: EvalRunResponse (application/json) eval::rerun_job
POST /v1/eval/run Submit an eval job body: EvalRunRequest (application/json) 200: EvalRunResponse (application/json) eval::submit_eval
GET /v1/eval/suites List eval suites none 200: SuiteListResponse (application/json) eval::list_suites
POST /v1/eval/suites Create or replace an eval suite query: SuiteUpsertQuery; body: EvalSuite (application/json) 200: SuiteSaveResponse (application/json) eval::save_suite
GET /v1/eval/suites/{name} Get an eval suite path: name 200: EvalSuite (application/json) eval::get_suite
DELETE /v1/eval/suites/{name} Delete an eval suite path: name 200: DeleteSuiteResponse (application/json) eval::delete_suite
GET /v1/judgments List judgment datasets none 200: JudgmentListResponse (application/json) eval::list_judgments
POST /v1/judgments Create a judgment dataset body: CreateJudgmentBody (application/json) 200: JudgmentManifest (application/json) eval::create_judgment_dataset
POST /v1/judgments/render_prompt Render the prompt for a proposed judgment row body: AppendJudgmentBody (application/json) 200: RenderJudgmentPromptResponse (application/json) eval::render_judgment_prompt
DELETE /v1/judgments/{name} Delete a judgment dataset path: name 200: DeleteJudgmentResponse (application/json) eval::delete_judgment_dataset
POST /v1/judgments/{name}/compile Compile a judgment dataset for training path: name; body: CompileJudgmentBody (application/json) 200: CompileJudgmentResponse (application/json) eval::compile_judgment
POST /v1/judgments/{name}/rows Append a row to a judgment dataset path: name; body: AppendJudgmentBody (application/json) 200: AppendJudgmentResponse (application/json) eval::append_judgment
DELETE /v1/judgments/{name}/rows/{judgment_id} Delete one row from a judgment dataset path: name, judgment_id 200: JudgmentManifest (application/json) eval::remove_judgment_row
POST /v1/judgments/{name}/validate Validate an adapter against a judgment dataset path: name; body: PromoteJudgmentBody (application/json) 200: ValidateJudgmentResponse (application/json) eval::validate_judgment_adapter

preflight

Training compatibility, capacity estimates, tier defaults, and intent routing.

Method Path Summary Request inputs Successful response Owner
POST /v1/preflight/capacity Estimate whether a training request fits body: CapacityRequest (application/json) 200: CapacityResponse (application/json) pit_of_success::capacity
GET /v1/preflight/compatibility Check model, backend, and training compatibility query: CompatibilityQuery 200: CompatibilityResponse (application/json) pit_of_success::compatibility
GET /v1/preflight/tier_defaults Get defaults for one training tier query: TierQuery 200: TierDefaultsResponse (application/json) pit_of_success::tier_defaults_endpoint
GET /v1/preflight/tiers List built-in training tiers and their defaults none 200: TierDefaultsListResponse (application/json) pit_of_success::tier_defaults_list
POST /v1/train Validate and route a training request body: FrontDoorRequest (application/json) 200: FrontDoorResponse (application/json) pit_of_success::submit_front_door

recipes

Built-in multi-step training recipes.

Method Path Summary Request inputs Successful response Owner
GET /v1/recipes List built-in training recipes none 200: RecipesListResponse (application/json) recipes::list_recipes
POST /v1/recipes/run Start a built-in training recipe body: RecipeRunRequest (application/json) 200: RecipeRunResponse (application/json) recipes::run_recipe

library

Adapter-library contract surface. Remote discovery, installation, and publication are not operational.

Method Path Summary Request inputs Successful response Owner
GET /v1/library Report adapter-library status (remote listing is unavailable) none 200: LibraryListResponse (application/json) library::list_library
POST /v1/library/install/{id} Report that remote adapter installation is unavailable path: id unspecified library::install_from_library
POST /v1/library/publish/{name} Validate an adapter for future library publication path: name; body: PublishPayload (application/json) 200: PublishToLibraryResponse (application/json) library::publish_to_library

Payload components

Open components are explicit migration markers, not an assertion that arbitrary fields are accepted by the runtime.

Component Rust type Shape Field status
AdapterDetail AdapterDetail AdapterDetail complete
AdapterUploadMultipart AdapterUploadMultipart AdapterUploadMultipart complete
AdaptersResponse AdaptersResponse AdaptersResponse complete
AgentRunAbortResponse AgentRunAbortResponse AgentRunAbortResponse complete
AgentRunEventsResponse AgentRunEventsResponse AgentRunEventsResponse complete
AgentRunListResponse AgentRunListResponse AgentRunListResponse complete
AgentRunQueuedResponse AgentRunQueuedResponse AgentRunQueuedResponse complete
AgentRunRecord AgentRunRecord AgentRunRecord complete
AgentRunsStatusResponse AgentRunsStatusResponse AgentRunsStatusResponse complete
AgentTrace AgentTrace AgentTrace complete
AgentTracesListResponse AgentTracesListResponse AgentTracesListResponse complete
ApiError ApiError object; closed complete
AppendJudgmentBody AppendJudgmentBody AppendJudgmentBody complete
AppendJudgmentResponse AppendJudgmentResponse AppendJudgmentResponse complete
BatchCompletionRequest BatchCompletionRequest BatchCompletionRequest complete
BatchCompletionResponse BatchCompletionResponse BatchCompletionResponse complete
BinaryArchive BinaryArchive string complete
CacheStatsResponse CacheStatsResponse CacheStatsResponse complete
CancelEvalJobResponse CancelEvalJobResponse CancelEvalJobResponse complete
CancelTrainingJobResponse CancelTrainingJobResponse CancelTrainingJobResponse complete
CapacityRequest CapacityRequest CapacityRequest complete
CapacityResponse CapacityResponse CapacityResponse complete
ChatCompletionChunkStream ChatCompletionChunkStream ChatCompletionChunkStream complete
ChatCompletionRequest ChatCompletionRequest ChatCompletionRequest complete
ChatCompletionResponse ChatCompletionResponse ChatCompletionResponse complete
ClearCorrectionsResponse ClearCorrectionsResponse ClearCorrectionsResponse complete
CompatibilityResponse CompatibilityResponse CompatibilityResponse complete
CompileJudgmentBody CompileJudgmentBody CompileJudgmentBody complete
CompileJudgmentResponse CompileJudgmentResponse CompileJudgmentResponse complete
ConfigResponse ConfigResponse ConfigResponse complete
CorrectionRow CorrectionRow CorrectionRow complete
CorrectionRowInput CorrectionRow CorrectionRowInput complete
CreateJudgmentBody CreateJudgmentBody CreateJudgmentBody complete
CreateRunRequest CreateRunRequest CreateRunRequest complete
CssStylesheet CssStylesheet string complete
DatasetListResponse DatasetListResponse DatasetListResponse complete
DatasetManifest DatasetManifest DatasetManifest complete
DatasetSplitConfig DatasetSplitConfig DatasetSplitConfig complete
DatasetSplitManifest DatasetSplitManifest DatasetSplitManifest complete
DatasetUploadMultipart DatasetUploadMultipart DatasetUploadMultipart complete
DebugDisabledResponse DebugDisabledResponse DebugDisabledResponse complete
DebugProvenanceErrorResponse serde_json::Value DebugProvenanceErrorResponse complete
DecodeStatsSnapshot DecodeStatsSnapshot DecodeStatsSnapshot complete
DeleteAdapterResponse DeleteAdapterResponse DeleteAdapterResponse complete
DeleteCorrectionResponse DeleteCorrectionResponse DeleteCorrectionResponse complete
DeleteDatasetResponse DeleteDatasetResponse DeleteDatasetResponse complete
DeleteExportResponse DeleteExportResponse DeleteExportResponse complete
DeleteJudgmentResponse DeleteJudgmentResponse DeleteJudgmentResponse complete
DeleteSuiteResponse DeleteSuiteResponse DeleteSuiteResponse complete
DeleteTeacherResponse DeleteTeacherResponse DeleteTeacherResponse complete
DeleteTrainingJobResponse DeleteTrainingJobResponse DeleteTrainingJobResponse complete
DiscoverRequest DiscoverRequest DiscoverRequest complete
DiscoverResponse DiscoverResponse DiscoverResponse complete
DistillMergeRequest DistillMergeRequest DistillMergeRequest complete
DistillPumpRequest DistillPumpRequest DistillPumpRequest complete
DistillRefreshRequest DistillRefreshRequest DistillRefreshRequest complete
DistillSelfRequest DistillSelfRequest DistillSelfRequest complete
EvalCompareSpec EvalCompareSpec EvalCompareSpec complete
EvalJobListResponse EvalJobListResponse EvalJobListResponse complete
EvalResult EvalResult EvalResult complete
EvalRunRequest EvalRunRequest EvalRunRequest complete
EvalRunResponse EvalRunResponse EvalRunResponse complete
EvalSuite EvalSuite EvalSuite complete
ExportDetail ExportDetail ExportDetail complete
ExportList ExportList ExportList complete
ExportSummary ExportSummary ExportSummary complete
FrontDoorRequest FrontDoorRequest FrontDoorRequest complete
FrontDoorResponse FrontDoorResponse FrontDoorResponse complete
GrpoExportRequest GrpoExportRequest GrpoExportRequest complete
GrpoRequest GrpoRequest GrpoRequest complete
HealthResponse HealthResponse HealthResponse complete
HtmlDocument HtmlDocument string complete
ImportPeftResponse ImportPeftResponse ImportPeftResponse complete
JavaScriptSource JavaScriptSource string complete
JsonValue serde_json::Value any complete
JsonValueArray Vec<serde_json::Value> array complete
JudgeDistillRequest JudgeDistillRequest JudgeDistillRequest complete
JudgeDistillResponse JudgeDistillResponse JudgeDistillResponse complete
JudgeDriftCheckRequest JudgeDriftCheckRequest JudgeDriftCheckRequest complete
JudgmentListResponse JudgmentListResponse JudgmentListResponse complete
JudgmentManifest JudgmentManifest JudgmentManifest complete
LibraryListResponse LibraryListResponse LibraryListResponse complete
ListResponse ListResponse ListResponse complete
LoadAdapterRequest LoadAdapterRequest LoadAdapterRequest complete
LoadAdapterResponse LoadAdapterResponse LoadAdapterResponse complete
MarkTrainedRequest MarkTrainedRequest MarkTrainedRequest complete
MarkTrainedResponse MarkTrainedResponse MarkTrainedResponse complete
MergeAdapterRequest MergeAdapterRequest MergeAdapterRequest complete
MergeAdapterResponse MergeAdapterResponse MergeAdapterResponse complete
MessageRequest MessageRequest MessageRequest complete
ModelStateResponse ModelStateResponse ModelStateResponse complete
ModelsResponse ModelsResponse ModelsResponse complete
OpdRequest OpdRequest OpdRequest complete
OpenEnvInspectRequest OpenEnvInspectRequest OpenEnvInspectRequest complete
OpenEnvInspectResponse OpenEnvInspectResponse OpenEnvInspectResponse complete
OpenEnvRunList OpenEnvRunList OpenEnvRunList complete
OpenEnvRunRequest OpenEnvRunRequest OpenEnvRunRequest complete
OpenEnvRunStatus OpenEnvRunStatus OpenEnvRunStatus complete
OpenEnvTaskCatalogRequest OpenEnvTaskCatalogRequest OpenEnvTaskCatalogRequest complete
OpenEnvTaskCatalogResponse OpenEnvTaskCatalogResponse OpenEnvTaskCatalogResponse complete
OpenEnvTrainingCapacitySnapshot OpenEnvTrainingCapacitySnapshot OpenEnvTrainingCapacitySnapshot complete
OpenEnvTrainingPreflightReceipt OpenEnvTrainingPreflightReceipt OpenEnvTrainingPreflightReceipt complete
OpenEnvTrainingPreflightRequest OpenEnvTrainingPreflightRequest OpenEnvTrainingPreflightRequest complete
PrometheusText PrometheusText string complete
PromoteJudgmentBody PromoteJudgmentBody PromoteJudgmentBody complete
PublishPayload PublishPayload PublishPayload complete
PublishToLibraryResponse PublishToLibraryResponse PublishToLibraryResponse complete
QueueResponse QueueResponse QueueResponse complete
RecipeRunRequest RecipeRunRequest RecipeRunRequest complete
RecipeRunResponse RecipeRunResponse RecipeRunResponse complete
RecipesListResponse RecipesListResponse RecipesListResponse complete
Redirect Redirect string complete
RegisterTeacherRequest RegisterTeacherRequest RegisterTeacherRequest complete
RenderJudgmentPromptResponse RenderJudgmentPromptResponse RenderJudgmentPromptResponse complete
ReplayBody ReplayBody ReplayBody complete
RequestRecord RequestRecord RequestRecord complete
RerunBody RerunBody RerunBody complete
SelfImproveRequest SelfImproveRequest SelfImproveRequest complete
SelfImproveResponse SelfImproveResponse SelfImproveResponse complete
SftExportRequest SftExportRequest SftExportRequest complete
SftRequest SftRequest SftRequest complete
SuiteListResponse SuiteListResponse SuiteListResponse complete
SuiteSaveResponse SuiteSaveResponse SuiteSaveResponse complete
SynthesisPreview SynthesisPreview SynthesisPreview complete
SynthesisPreviewBody SynthesisPreviewBody SynthesisPreviewBody complete
SynthesizeBody SynthesizeBody SynthesizeBody complete
SynthesizeDatasetResponse SynthesizeDatasetResponse SynthesizeDatasetResponse complete
TeacherEntry TeacherEntry TeacherEntry complete
TeachersListResponse TeachersListResponse TeachersListResponse complete
TerminalStatusResponse TerminalStatusResponse TerminalStatusResponse complete
TextCompletionRequest TextCompletionRequest TextCompletionRequest complete
TextCompletionResponse TextCompletionResponse TextCompletionResponse complete
TierDefaultsListResponse TierDefaultsListResponse TierDefaultsListResponse complete
TierDefaultsResponse TierDefaultsResponse TierDefaultsResponse complete
TrainingJobDetail TrainingJobDetail TrainingJobDetail complete
TrainingResponse TrainingResponse TrainingResponse complete
TrainingStatus TrainingStatus TrainingStatus complete
UnloadAdapterResponse UnloadAdapterResponse UnloadAdapterResponse complete
UploadAdapterResponse UploadAdapterResponse UploadAdapterResponse complete
ValidateJudgmentResponse ValidateJudgmentResponse ValidateJudgmentResponse complete
Vec_RequestRecord Vec<RequestRecord> Vec_RequestRecord complete
Vec_TrainingStatus Vec<TrainingStatus> Vec_TrainingStatus complete
WebSocketUpgrade WebSocketUpgrade string complete
kiln_train_AdapterReceipt kiln_train::AdapterReceipt kiln_train_AdapterReceipt complete