部署训练检查点 - Fastino API Docs

POST

https://api.fastino.ai

training

jobs

{job_id}

checkpoints

{checkpoint_id}

deploy

curl -X POST \
  https://api.fastino.ai/v1/training/jobs/YOUR_TRAINING_JOB_ID/checkpoints/CHECKPOINT_ID/deploy \
  -H "X-API-Key: $FASTINO_API_KEY"

{
  "success": true,
  "message": "Checkpoint deployed successfully",
  "job_id": "YOUR_TRAINING_JOB_ID",
  "checkpoint_id": "CHECKPOINT_ID",
  "deployment_path": "PROVIDER_MODEL_PATH",
  "mme_path": "PROVIDER_MODEL_PATH"
}

POST /v1/training/jobs/{job_id}/checkpoints/{checkpoint_id}/deploy 将一个可部署的检查点部署到托管的多模型推理端点。

请求

string

必填

您的 Fastino API 密钥。

string

必填

训练任务 UUID。

string

必填

由检查点列表端点返回的检查点 ID。

curl -X POST \
  https://api.fastino.ai/v1/training/jobs/YOUR_TRAINING_JOB_ID/checkpoints/CHECKPOINT_ID/deploy \
  -H "X-API-Key: $FASTINO_API_KEY"

响应

{
  "success": true,
  "message": "Checkpoint deployed successfully",
  "job_id": "YOUR_TRAINING_JOB_ID",
  "checkpoint_id": "CHECKPOINT_ID",
  "deployment_path": "PROVIDER_MODEL_PATH",
  "mme_path": "PROVIDER_MODEL_PATH"
}

deployment_path 是规范的已部署对象路径。mme_path 是已弃用的兼容性别名。未知的任务或检查点返回 404。不可部署的检查点返回请求相关的 4xx 错误。

授权

由于此操作会更改实时部署，因此需要团队的 Owner 或 Admin 角色。其他角色会收到 403 Forbidden。

curl -X POST \
  https://api.fastino.ai/v1/training/jobs/YOUR_TRAINING_JOB_ID/checkpoints/CHECKPOINT_ID/deploy \
  -H "X-API-Key: $FASTINO_API_KEY"

{
  "success": true,
  "message": "Checkpoint deployed successfully",
  "job_id": "YOUR_TRAINING_JOB_ID",
  "checkpoint_id": "CHECKPOINT_ID",
  "deployment_path": "PROVIDER_MODEL_PATH",
  "mme_path": "PROVIDER_MODEL_PATH"
}

​请求

​响应

​授权

请求

响应

授权