Finetune

Package ai is Hanzo AI — the model API on /v1 (/v1/chat/completions, /v1/messages, /v1/models and the rest of hanzoai/ai's surface) — mounted into a cloud binary with the money, ingest and telemetry…

All API references → · Six flows, four surfaces →


Base URL	`https://api.hanzo.ai`
Operations	9
Auth	`Authorization: Bearer $HANZO_API_KEY`

`POST /v1/finetune/cancel`

Deletes the TrainJob CR, meters the GPU-hours used so far, and marks the job cancelled. ?id= or ?name=

`POST /v1/finetune/deploy`

Serves a completed job's checkpoints and registers the result as a routable model on api.hanzo.ai. ?id= or ?name=

`GET /v1/finetune/hf/datasets`

Proxies a HuggingFace dataset search (dataset picker).

`GET /v1/finetune/hf/models`

Proxies a HuggingFace model search (base-model picker).

`GET /v1/finetune/hf/repo`

Returns a repo's detail (files, gated/private state). ?id=&kind=model|dataset

`GET /v1/finetune/job`

Returns one job with refreshed live status. ?id=owner/name or ?name=

`GET /v1/finetune/jobs`

Returns the org's jobs, refreshing live status for active ones.

Validates the request, resolves efficient defaults, persists the job, and submits a real TrainJob CR. A submit failure (e.g. no cluster wired) is surfaced honestly: the job is saved with status "failed" + the reason, never faked.

`GET /v1/finetune/presets`

Returns the new-job catalog plus, when a selection is passed (?baseModel&method&task&preset[&datasetExamples]), the recommended config so the console can render "Recommended" as a one-click, ready-to-run default.

All Hanzo APIs · Interactive reference