<span id="模型微调"></span>

# Model fine-tuning

Choose an API for the resource you need to operate on, then prepare the identity, IDs, and parameters listed on its action page.

<span id="选择任务"></span>

## Choose a task

| Task | Page | Description |
| --- | --- | --- |
| Cancel a nonterminal fine-tuning job | [Cancel a nonterminal fine-tuning job](cancel.md) | If a real workflow binding exists, request workflow cancellation before persisting cancelled. |
| Create a fine-tuning job | [Create a fine-tuning job](create.md) | Set `BASE_URL` to the product API prefix `https://api.moi.matrixorigin.cn/v5`; use the corresponding product API URL for a dedicated deployment. |
| Get fine-tuning job status and artifacts | [Get fine-tuning job status and artifacts](get.md) | Read current job state, training inputs, workflow associations, errors, and adapter URI. |
| Start a created fine-tuning job | [Start a created fine-tuning job](launch.md) | Prepare training inputs and schedule a workflow for a nonterminal job. |
| Creation, scheduling, and result recovery | [Creation, scheduling, and result recovery](lifecycle.md) | Save a queryable job first, then launch and read its terminal state. |
| Fine-tuning job objects | [Fine-tuning job objects](objects.md) | The four control APIs return the same job DTO. Success uses string code=OK, msg=OK, and data. |
| Training inputs and execution environment | [Training inputs and execution environment](training-inputs.md) | These APIs form the fine-tuning training-node control plane. |

```{toctree}
:hidden:
:maxdepth: 1

Cancel a nonterminal fine-tuning job <cancel>
Create a fine-tuning job <create>
Get fine-tuning job status and artifacts <get>
Start a created fine-tuning job <launch>
Creation, scheduling, and result recovery <lifecycle>
Fine-tuning job objects <objects>
Training inputs and execution environment <training-inputs>
```
