geti-using-the-pipeline

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Using the Geti pipeline (application)

使用Geti流程（应用层）

The Geti application is a FastAPI server (

application/backend/

, the

geti

package) that exposes a REST API for the full computer-vision workflow: create a project, upload and annotate media, train a model as an async job, then configure and enable a live inference pipeline (source → model → sink). This skill is about using that API; to change backend code use the

geti-backend-dev

skill instead.

Start the server from

application/backend/

with

just run-server

(default

https://localhost:7860

). The full endpoint reference is

application/docs/api.md

; interactive docs are the generated OpenAPI spec.

Geti应用是一个FastAPI服务器（位于

application/backend/

目录，对应

geti

包），它暴露REST API以支持完整的计算机视觉工作流：创建项目，上传并标注媒体资源，以异步任务形式训练模型，然后配置并启用实时推理流程（数据源→模型→输出端）。本技能聚焦于该API的使用；若需修改后端代码，请使用

geti-backend-dev

技能。

从

application/backend/

目录执行

just run-server

启动服务器（默认地址为

https://localhost:7860

）。完整的接口参考文档位于

application/docs/api.md

；交互式文档为生成的OpenAPI规范。

End-to-end pipeline

端到端流程

mermaid

flowchart LR
    A[Create project] --> B[Upload media]
    B --> C[Annotate media]
    C --> D[Train job]
    D --> E[Configure pipeline: source, model, sink]
    E --> F[Enable pipeline / live inference]

Create a project with a task type and labels.
- ```
POST /api/projects
```
  (name, task, labels) → project info.
- Done when:
```
GET /api/projects/<id>
```
  returns the project with its labels.
Upload media (images/videos) to the project dataset.
- ```
POST /api/projects/<id>/dataset/media
```
  (binary) → media info.
- Done when:
```
GET /api/projects/<id>/dataset/media
```
  lists the uploaded item.
Annotate media so the dataset is trainable.
- ```
POST /api/projects/<id>/dataset/media/<media_id>/annotations
```
  (annotation info).
- Done when:
```
GET .../annotations
```
  returns the saved annotation.
- (Optional) import an existing dataset instead via the dataset jobs below.
Train a model as an async job.
- ```
POST /api/jobs
```
  with job type
```
train
```
  → job id.
- Track it:
```
GET /api/jobs/<id>
```
  , stream
```
GET /api/jobs/<id>/status
```
  and
```
GET /api/jobs/<id>/logs
```
  ; cancel with
```
POST /api/jobs/<id>:cancel
```
  .
- Done when: the job reaches a finished state and
```
GET /api/projects/<id>/models
```
  lists the new model.
(Optional) Quantize the trained model for faster inference.
- ```
POST /api/jobs
```
  with job type
```
quantize
```
  .
- Done when: the quantized model variant appears under the project's models.
Configure the inference pipeline — bind a source, the model, and a sink.
- Sources:
```
POST /api/sources
```
  ; sinks:
```
POST /api/sinks
```
  .
- ```
PATCH /api/projects/<id>/pipeline
```
  with the ids of source, sink, and model.
- Done when:
```
GET /api/projects/<id>/pipeline
```
  shows the wired components.
Enable live inference and monitor it.
- ```
POST /api/projects/<id>/pipeline:enable
```
  (disable with
```
:disable
```
  ).
- Metrics:
```
GET /api/projects/<id>/pipeline/metrics
```
  (latency, throughput).
- ```
POST /api/projects/<id>/pipeline:capture
```
  collects the next frame into the dataset for continued annotation/retraining.
- Done when: the pipeline reports active and metrics update.

mermaid

flowchart LR
    A[Create project] --> B[Upload media]
    B --> C[Annotate media]
    C --> D[Train job]
    D --> E[Configure pipeline: source, model, sink]
    E --> F[Enable pipeline / live inference]

创建项目，指定任务类型和标签。
- 调用
```
POST /api/projects
```
  （传入名称、任务类型、标签）→ 返回项目信息。
- 完成标志：
```
GET /api/projects/<id>
```
  返回包含标签的项目详情。
上传媒体资源（图片/视频）到项目数据集。
- 调用
```
POST /api/projects/<id>/dataset/media
```
  （传入二进制文件）→ 返回媒体资源信息。
- 完成标志：
```
GET /api/projects/<id>/dataset/media
```
  列出已上传的资源。
标注媒体资源，使数据集可用于训练。
- 调用
```
POST /api/projects/<id>/dataset/media/<media_id>/annotations
```
  （传入标注信息）。
- 完成标志：
```
GET .../annotations
```
  返回已保存的标注内容。
- （可选）通过下方的数据集任务导入现有数据集，替代手动标注。
以异步任务形式训练模型。
- 调用
```
POST /api/jobs
```
  并指定任务类型为
```
train
```
  → 返回任务ID。
- 跟踪任务：调用
```
GET /api/jobs/<id>
```
  ，或流式获取
```
GET /api/jobs/<id>/status
```
  和
```
GET /api/jobs/<id>/logs
```
  ；调用
```
POST /api/jobs/<id>:cancel
```
  可取消任务。
- 完成标志：任务进入完成状态，且
```
GET /api/projects/<id>/models
```
  列出新模型。
（可选）量化训练后的模型以提升推理速度。
- 调用
```
POST /api/jobs
```
  并指定任务类型为
```
quantize
```
  。
- 完成标志：量化后的模型变体出现在项目模型列表中。
配置推理流程——绑定数据源、模型和输出端。
- 创建数据源：
```
POST /api/sources
```
  ；创建输出端：
```
POST /api/sinks
```
  。
- 调用
```
PATCH /api/projects/<id>/pipeline
```
  并传入数据源、输出端和模型的ID。
- 完成标志：
```
GET /api/projects/<id>/pipeline
```
  显示已绑定的组件。
启用实时推理并监控。
- 调用
```
POST /api/projects/<id>/pipeline:enable
```
  （调用
```
:disable
```
  可关闭）。
- 查看指标：
```
GET /api/projects/<id>/pipeline/metrics
```
  （延迟、吞吐量）。
- 调用
```
POST /api/projects/<id>/pipeline:capture
```
  可将下一帧采集到数据集，用于后续标注/再训练。
- 完成标志：流程报告处于活跃状态，且指标持续更新。

The async job model

异步任务模型

Long-running work runs as jobs (

POST /api/jobs

), keeping the API responsive. Job types:

train

quantize

prepare_dataset_for_import

import_dataset_to_existing_project

import_dataset_as_new_project

export_dataset

stage_dataset

. Poll

GET /api/jobs/<id>

or stream

/status

and

/logs

; jobs are cancelable.

耗时操作以任务形式运行（调用

POST /api/jobs

），确保API响应性。任务类型包括：

train

、

quantize

、

prepare_dataset_for_import

、

import_dataset_to_existing_project

、

import_dataset_as_new_project

、

export_dataset

、

stage_dataset

。可轮询

GET /api/jobs/<id>

或流式获取

/status

和

/logs

；任务支持取消操作。

Datasets: import instead of manual annotation

数据集：导入替代手动标注

To bring in an existing dataset rather than annotating from scratch:

Upload an archive to staging:
```
POST /api/staged_datasets
```
.

Then submit an import job (

import_dataset_as_new_project

import_dataset_to_existing_project

) via

POST /api/jobs

Export a project's dataset with the
```
export_dataset
```
job.

若无需从头标注，可导入现有数据集：

将归档文件上传到暂存区：
```
POST /api/staged_datasets
```
。

然后通过

POST /api/jobs

提交导入任务（

import_dataset_as_new_project

或

import_dataset_to_existing_project

）。

调用
```
export_dataset
```
任务可导出项目数据集。

Notes

注意事项

Training and quantization jobs run out-of-process and call into the
```
getitune
```
library; the underlying capabilities map to the
```
getitune-training-a-model
```
and
```
getitune-optimizing-a-model
```
skills.
This skill covers API usage; endpoint paths and payloads are the contract in
```
application/docs/api.md
```
. To add or change endpoints, use
```
geti-backend-dev
```
and
```
geti-openapi-sync
```
.

训练和量化任务在进程外运行，并调用
```
getitune
```
库；其底层能力对应
```
getitune-training-a-model
```
和
```
getitune-optimizing-a-model
```
技能。
本技能涵盖API使用；接口路径和请求体是
```
application/docs/api.md
```
中定义的契约。若需添加或修改接口，请使用
```
geti-backend-dev
```
和
```
geti-openapi-sync
```
技能。