获取训练运行 - Fastino API Docs

GET

https://api.fastino.ai

training

jobs

{job_id}

curl https://api.fastino.ai/v1/training/jobs/YOUR_JOB_ID \
  -H "X-API-Key: $FASTINO_API_KEY"

{
  "type": "interactive",
  "training_run_id": "YOUR_SESSION_ID:train:0",
  "base_model": "Qwen/Qwen3.6-35B-A3B",
  "model_owner": "YOUR_TEAM_ID",
  "is_lora": true,
  "corrupted": false,
  "lora_rank": 32,
  "last_request_time": "2026-10-09T18:42:10Z",
  "last_checkpoint": {
    "checkpoint_id": "weights/step-100",
    "checkpoint_type": "training",
    "time": "2026-10-09T18:40:00Z",
    "tinker_path": "tinker://YOUR_SESSION_ID:train:0/weights/step-100",
    "size_bytes": 503316480,
    "public": false,
    "user_metadata": null,
    "expires_at": null
  },
  "last_sampler_checkpoint": {
    "checkpoint_id": "sampler_weights/step-100",
    "checkpoint_type": "sampler",
    "time": "2026-10-09T18:40:02Z",
    "tinker_path": "tinker://YOUR_SESSION_ID:train:0/sampler_weights/step-100",
    "size_bytes": 167772160,
    "public": false,
    "user_metadata": null,
    "expires_at": null
  },
  "user_metadata": {"experiment": "baseline"}
}

GET /v1/training/jobs/{job_id} 返回调用方可见的单个训练运行，以及其最新的训练状态检查点和采样检查点。

请求

string

必填

您的 Fastino API 密钥。

string

必填

训练运行 ID。即兼容 Tinker 的 SDK 在创建训练客户端时返回的模型 ID，格式为 SESSION_ID:train:N。

curl https://api.fastino.ai/v1/training/jobs/YOUR_JOB_ID \
  -H "X-API-Key: $FASTINO_API_KEY"

响应

string

始终为 interactive：您通过兼容 Tinker 的 SDK 驱动的训练运行。

string

训练运行 ID。

string

该运行训练的基础模型。

string

拥有该运行的团队 ID。

boolean

该运行是否训练 LoRA 适配器。

boolean

该运行的训练状态是否不可用，且无法再执行训练步骤。

integer | null

LoRA 秩（当该运行训练 LoRA 适配器时）。

string

该运行最近一次 SDK 请求的 ISO 8601 时间戳。

object | null

最新的训练状态检查点，或 null。检查点字段见列出训练运行检查点。

object | null

最新的采样权重检查点，或 null。

object | null

创建运行时附加的字符串键值元数据。

{
  "type": "interactive",
  "training_run_id": "YOUR_SESSION_ID:train:0",
  "base_model": "Qwen/Qwen3.6-35B-A3B",
  "model_owner": "YOUR_TEAM_ID",
  "is_lora": true,
  "corrupted": false,
  "lora_rank": 32,
  "last_request_time": "2026-10-09T18:42:10Z",
  "last_checkpoint": {
    "checkpoint_id": "weights/step-100",
    "checkpoint_type": "training",
    "time": "2026-10-09T18:40:00Z",
    "tinker_path": "tinker://YOUR_SESSION_ID:train:0/weights/step-100",
    "size_bytes": 503316480,
    "public": false,
    "user_metadata": null,
    "expires_at": null
  },
  "last_sampler_checkpoint": {
    "checkpoint_id": "sampler_weights/step-100",
    "checkpoint_type": "sampler",
    "time": "2026-10-09T18:40:02Z",
    "tinker_path": "tinker://YOUR_SESSION_ID:train:0/sampler_weights/step-100",
    "size_bytes": 167772160,
    "public": false,
    "user_metadata": null,
    "expires_at": null
  },
  "user_metadata": {"experiment": "baseline"}
}

未知的运行或您无权查看的运行会返回 404。

⌘I

curl https://api.fastino.ai/v1/training/jobs/YOUR_JOB_ID \
  -H "X-API-Key: $FASTINO_API_KEY"

{
  "type": "interactive",
  "training_run_id": "YOUR_SESSION_ID:train:0",
  "base_model": "Qwen/Qwen3.6-35B-A3B",
  "model_owner": "YOUR_TEAM_ID",
  "is_lora": true,
  "corrupted": false,
  "lora_rank": 32,
  "last_request_time": "2026-10-09T18:42:10Z",
  "last_checkpoint": {
    "checkpoint_id": "weights/step-100",
    "checkpoint_type": "training",
    "time": "2026-10-09T18:40:00Z",
    "tinker_path": "tinker://YOUR_SESSION_ID:train:0/weights/step-100",
    "size_bytes": 503316480,
    "public": false,
    "user_metadata": null,
    "expires_at": null
  },
  "last_sampler_checkpoint": {
    "checkpoint_id": "sampler_weights/step-100",
    "checkpoint_type": "sampler",
    "time": "2026-10-09T18:40:02Z",
    "tinker_path": "tinker://YOUR_SESSION_ID:train:0/sampler_weights/step-100",
    "size_bytes": 167772160,
    "public": false,
    "user_metadata": null,
    "expires_at": null
  },
  "user_metadata": {"experiment": "baseline"}
}

​请求

​响应

请求

响应