nemo_microservices.types.customization_target_param#

Module Contents#

Classes#

CustomizationTargetParam

API#

class nemo_microservices.types.customization_target_param.CustomizationTargetParam#

Bases: typing_extensions.TypedDict

base_model: str#

None

Default to being the same as the the configuration entry name, maps to the name in NIM

custom_fields: Dict[str, str]#

None

A set of custom fields that the user can define and use for various purposes.

description: str#

None

The description of the entity.

enabled: bool#

None

Enable the model for training jobs

hf_endpoint: str#

None

Configure the Hub base URL.

Defaults to NeMo Data Store. Set value as “https://huggingface.co” to download model_uri from HuggingFace.

model_path: str#

None

Path to the model checkpoints to use for training.

Absolute path or local path from the models cache

model_uri: str#

None

The URI of the model to download to the model cache at the model_path directory.

To download from NGC, specify ngc://org/optional-team/model-name:version. To download from Nemo Data Store, specify hf://namespace/model-name@checkpoint-name

name: str#

None

The name of the entity.

Must be unique inside the namespace. If not specified, it will be the same as the automatically generated id.

namespace: str#

None

The namespace of the entity.

You can omit this field for namespace entities or in deployments that don’t use namespaces.

num_parameters: typing_extensions.Required[int]#

None

Number of parameters used for training the model

precision: typing_extensions.Required[nemo_microservices.types.shared.model_precision.ModelPrecision]#

None

Type of model precision.

Values

  • "int8" - 8-bit integer precision

  • "bf16" - Brain floating point precision

  • "fp16" - 16-bit floating point precision

  • "fp32" - 32-bit floating point precision

  • "fp8-mixed" - Mixed 8-bit floating point precision available on Hopper and later architectures.

  • "bf16-mixed" - Mixed Brain floating point precision

project: str#

None

The URN of the project associated with this entity.

tokenizer: Dict[str, object]#

None

Overrides for the model tokenizer