Update a Customization Target#

Prerequisites#

Before you can update a customization target’s details, make sure that you have:

  • The namespace the target belongs to.

  • The name of the target.

  • Set the CUSTOMIZER_BASE_URL environment variable to your NeMo Customizer service endpoint

export CUSTOMIZER_BASE_URL="https://your-customizer-service-url"

You can get this information for all targets by listing the available targets.

To Update a Customization Target#

Choose one of the following options to update a customization target’s metadata or enable/disable it. The following examples show you how to disable a customization target.

import os
from nemo_microservices import NeMoMicroservices

# Initialize the client
client = NeMoMicroservices(
    base_url=os.environ['CUSTOMIZER_BASE_URL']
)

# Update customization target (disable it)
updated_target = client.customization.targets.update(
    target_name="llama-3.1-8b-instruct@2.0",
    namespace="default",
    enabled=False
)

print(f"Updated target: {updated_target.name}")
print(f"Enabled: {updated_target.enabled}")
print(f"Status: {updated_target.status}")

curl -X PATCH "${CUSTOMIZER_BASE_URL}/customization/targets/default/llama-3.1-8b-instruct@2.0" \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
   "enabled": false
  }' | jq
Example Response
{
    "id": "cust-target-abc123",
    "name": "llama-3.1-8b-instruct@2.0",
    "namespace": "meta",
    "description": "Customization target for Meta Llama 3.1 8B",
    "enabled": false,
    "base_model": "meta/llama-3.1-8b",
    "model_path": "llama-3_1-8b-instruct_2_0",
    "model_uri": "ngc://nvidia/nemo/llama-3_1-8b:2.0",
    "hf_endpoint": "<your HF endpoint>",
    "tokenizer": {},
    "num_parameters": 123456789,
    "precision": "bf16-mixed",
    "status": "created",
    "ownership": {},
    "custom_fields": {},
    "created_at": "2024-05-08T12:00:00Z",
    "updated_at": "2024-05-08T12:00:00Z"
}

Note

If you are updating the model_path, model_uri or re-enabling a target, please contact your admin to verify that the model download job doesn’t exist for this target. If it does, admin needs to update ttlSecondsAfterFinished for modelDownloader to 30 seconds or less (default TTL is set to 7200 seconds) and delete the job before proceeding.