Deployment and upgradesOfficial documentation5 min readDeployment and upgrades

Deploy ChatGLM2-M3E Custom Model Image for FastGPT

Available Container Images Two pre-built container images are provided for this custom model deployment.

Available Container Images

Two pre-built container images are provided for this custom model deployment. The primary public image is stawky/chatglm2-m3e:latest. For users in mainland China with limited international network access, an official mirrored image is hosted at registry.cn-hangzhou.aliyuncs.com/fastgpt_docker/chatglm2-m3e:latest. Pull the appropriate image for your environment using the standard Docker pull command, substituting the correct image name as needed.

Core Configuration Parameters

All required and configurable parameters for this model container are defined below:

ParameterDefault ValueUsage Notes
Exposed Port6006Fixed port for API communication between the model container and FastGPT platform
Default Security Token[REDACTED_CREDENTIAL]Static token that functions as the channel key when integrating with OneAPI
Custom Token Overridesk-keyEnvironment variable used to set a unique security token, replacing the default value

Full Deployment Command

A complete Docker run command for deploying the model is shown below. Adjust the image name and custom token to match your environment:

docker run -d \
  --name chatglm2-m3e \
  -p 6006:6006 \
  -e sk-key=your-custom-token \
  stawky/chatglm2-m3e:latest

If using the China mainland mirrored image, replace the final image argument with registry.cn-hangzhou.aliyuncs.com/fastgpt_docker/chatglm2-m3e:latest. Omit the -e flag entirely to use the default security token, though using a custom token is recommended for secure deployments. After starting the container, confirm it is running with docker ps and validate connectivity to the API endpoint on port 6006 to ensure successful deployment.

Source: FastGPT official source

Applicability and version scope

Use this page for the documented Deployment and upgrades scenario. Confirm the FastGPT, dependency, API, and deployment versions in the official source before applying a change.

Safety guardrails

Use [REDACTED_CREDENTIAL] for credentials and private data. Confirm the documented environment and version before review.

Rollback guidance

Restore the prior technical-content authority snapshot. Restore saved configuration and data snapshots, then repeat the smallest verification scenario.