CLI Commands
This document provides a comprehensive guide to the command-line interface (CLI) commands available in the Easy Model Deployer (EMD) tool.
Overview
EMD provides the following main commands:
| Command | Description |
|---|---|
bootstrap |
Initialize AWS resources for model deployment |
deploy |
Deploy models to AWS infrastructure |
status |
Display status of deployed models |
invoke |
Test deployed models with sample requests |
example |
Generate sample code for API integration |
destroy |
Remove deployed models and clean up resources |
list-supported-models |
Display available models |
profile |
Configure AWS profile credentials |
version |
Display tool version information |
Command Details
bootstrap
Initialize AWS resources required for model deployment.
emd bootstrap [OPTIONS]
Options:
| Option | Description |
|---|---|
--skip-confirm |
Skip confirmation prompts |
Example:
emd bootstrap
This command creates the necessary AWS resources, including an S3 bucket and CloudFormation stack, required for model deployment.
deploy
Deploy models to AWS infrastructure.
emd deploy [OPTIONS]
Options:
| Option | Description |
|---|---|
--model-id TEXT |
Model ID to deploy |
-i, --instance-type TEXT |
The instance type to use |
-e, --engine-type TEXT |
The name of the inference engine |
-s, --service-type TEXT |
The name of the service |
--framework-type TEXT |
The name of the framework |
--model-tag TEXT |
Custom tag for the model deployment |
--extra-params TEXT |
Extra parameters in JSON format |
--skip-confirm |
Skip confirmation prompts |
--force-update-env-stack |
Force update environment stack |
--allow-local-deploy |
Allow local instance deployment |
--only-allow-local-deploy |
Only allow local instance deployment |
--dockerfile-local-path TEXT |
Custom Dockerfile path for building the model image |
--local-gpus TEXT |
Local GPU IDs to deploy the model (e.g., 0,1,2) |
Examples:
Deploy a model with interactive prompts:
emd deploy
Deploy a specific model with parameters:
emd deploy --model-id Qwen2.5-7B-Instruct --instance-type g5.2xlarge --engine-type vllm --service-type sagemaker_realtime
Deploy a model locally:
emd deploy --allow-local-deploy
Deploy with custom parameters:
emd deploy --model-id Qwen2.5-7B-Instruct --extra-params '{"engine_params": {"cli_args": "--max_model_len 16000 --max_num_seqs 4"}}'
status
Display the status of deployed models.
emd status [MODEL_ID] [MODEL_TAG]
Arguments:
| Argument | Description |
|---|---|
MODEL_ID |
Optional model ID to check status for |
MODEL_TAG |
Optional model tag (defaults to "dev") |
Examples:
Check status of all deployed models:
emd status
Check status of a specific model:
emd status Qwen2.5-7B-Instruct
Check status of a specific model with a custom tag:
emd status Qwen2.5-7B-Instruct custom-tag
invoke
Test deployed models with sample requests.
emd invoke MODEL_ID [MODEL_TAG]
Arguments:
| Argument | Description |
|---|---|
MODEL_ID |
Model ID to invoke |
MODEL_TAG |
Optional model tag (defaults to "dev") |
Examples:
Invoke a model:
emd invoke DeepSeek-R1-Distill-Qwen-7B
Invoke a model with a custom tag:
emd invoke DeepSeek-R1-Distill-Qwen-7B custom-tag
example
Generate sample code for API integration with a deployed model.
emd example MODEL_ID/MODEL_TAG
Arguments:
| Argument | Description |
|---|---|
MODEL_ID/MODEL_TAG |
Model ID and optional tag (separated by "/") |
Examples:
Generate examples for a model:
emd example Qwen2.5-7B-Instruct
Generate examples for a model with a custom tag:
emd example Qwen2.5-7B-Instruct/custom-tag
destroy
Remove deployed models and clean up resources.
emd destroy MODEL_ID [MODEL_TAG]
Arguments:
| Argument | Description |
|---|---|
MODEL_ID |
Model ID to destroy |
MODEL_TAG |
Optional model tag (defaults to "dev") |
Examples:
Destroy a model:
emd destroy Qwen2.5-7B-Instruct
Destroy a model with a custom tag:
emd destroy Qwen2.5-7B-Instruct custom-tag
list-supported-models
Display available models that can be deployed.
emd list-supported-models [MODEL_ID] [OPTIONS]
Arguments:
| Argument | Description |
|---|---|
MODEL_ID |
Optional model ID to filter results |
Options:
| Option | Description |
|---|---|
-a, --detail |
Output model information in detail |
Examples:
List all supported models:
emd list-supported-models
List detailed information for all models:
emd list-supported-models --detail
List information for a specific model:
emd list-supported-models Qwen2.5-7B-Instruct
profile
Configure AWS profile credentials for deployment.
emd profile COMMAND [ARGS]
Commands:
| Command | Description |
|---|---|
set-default-profile-name [NAME] |
Set the default profile name for deployment |
show-default-profile-name |
Show current default profile |
remove-default-profile-name |
Remove the default profile |
Examples:
Set a default AWS profile:
emd profile set-default-profile-name my-profile
Show the current default profile:
emd profile show-default-profile-name
Remove the default profile:
emd profile remove-default-profile-name
version
Display the current version of the EMD tool.
emd version
Example:
emd version
Environment Variables
LOCAL_DEPLOY_PORT: Local deployment port (default:8080)