Finetuning API

List all fine_tuning checkpoints.

get

Header parameters

x-userstringRequired

username

Responses

200

Checkpoints list.

application/json

string[]Optional

Checkpoint name.

400

Unknown APIs.

application/json

get/v1/inference/checkpoints

GET /v1/inference/checkpoints HTTP/1.1
x-user: text
Accept: */*

[
  "text"
]

Delete the checkpoint

get

Path parameters

idstringRequired

Header parameters

x-userstringRequired

username

Responses

200

Checkpoint is deleted.

No content

400

Unknown APIs.

application/json

get/v1/inference/checkpoints/{id}/del

GET /v1/inference/checkpoints/{id}/del HTTP/1.1
x-user: text
Accept: */*

No content

List inference tasks

get

Header parameters

x-userstringRequired

username

Responses

200

Inference tasks list.

application/json

Task information.

namestringOptional

Task name.

idstringOptional

Task ID.

400

Unknown APIs.

application/json

get/v1/inference/tasks

GET /v1/inference/tasks HTTP/1.1
x-user: text
Accept: */*

[
  {
    "name": "text",
    "id": "text"
  }
]

Create inference task

post

Header parameters

x-userstringRequired

username

Body

task_namestringOptional

Task name.

namespacestringOptional

Task Name.

Default: gpu-scheduler

priorityintegerOptional

Task priority.

modelstringOptional

Model name.

checkpointstringOptional

Checkpoint name.

ngpuintegerOptional

Number of GPU to be used by the inference task.

Responses

200

Inference task is created.

application/json

namestringOptional

Task name.

idstringOptional

Task ID.

400

Unknown APIs.

application/json

post/v1/inference/tasks

POST /v1/inference/tasks HTTP/1.1
x-user: text
Content-Type: application/json
Accept: */*
Content-Length: 105

{
  "task_name": "text",
  "namespace": "gpu-scheduler",
  "priority": 1,
  "model": "text",
  "checkpoint": "text",
  "ngpu": 1
}

{
  "name": "text",
  "id": "text"
}

Get inference task information.

get

Path parameters

idstringRequired

Header parameters

x-userstringRequired

username

Responses

200

Inference task information.

application/json

namestringOptional

Task name.

namespacestringOptional

Task namespace.

idstringOptional

Task ID.

prioritystringOptional

Task priority.

statusstringOptional

Task status.

inference_svc_namestringOptional

Inference service name.

inference_svc_urlstringOptional

Inference service url.

400

Unknown APIs.

application/json

get/v1/inference/tasks/{id}

GET /v1/inference/tasks/{id} HTTP/1.1
x-user: text
Accept: */*

{
  "name": "text",
  "namespace": "text",
  "id": "text",
  "priority": "text",
  "status": "text",
  "inference_svc_name": "text",
  "inference_svc_url": "text"
}

Cancel inference task.

get

Path parameters

idstringRequired

Header parameters

x-userstringRequired

username

Responses

200

Inference task is canceled

No content

400

Unknown APIs.

application/json

get/v1/inference/tasks/{id}/cancel

GET /v1/inference/tasks/{id}/cancel HTTP/1.1
x-user: text
Accept: */*

No content