Implementation:Cohere ai Cohere python Finetuning BaseModel Type

Knowledge Sources	Cohere Python SDK
Domains	SDK, Fine_Tuning
Last Updated	2026-02-15 14:00 GMT

Overview

The BaseModel class defines the data model for representing base models used in Cohere fine-tuning operations.

Description

BaseModel is a Pydantic-based data class that encapsulates the properties of a base model upon which fine-tuning is performed. It includes the model's name, version (read-only), base type (such as generative, classification, rerank, or chat), and an optional fine-tuning strategy. The class extends UncheckedBaseModel and supports both Pydantic v1 and v2 through a compatibility layer, allowing extra fields to be passed without raising validation errors.

Usage

Use this class when you need to specify or inspect the base model configuration for a fine-tuning job. It is typically used as a nested field within a broader fine-tuned model definition, representing which foundation model the fine-tuning is built upon.

Code Reference

Source Location

Repository: Cohere Python SDK
File: src/cohere/finetuning/finetuning/types/base_model.py

Signature

class BaseModel(UncheckedBaseModel):
    """
    The base model used for fine-tuning.
    """

    name: typing.Optional[str] = pydantic.Field(default=None)
    version: typing.Optional[str] = pydantic.Field(default=None)
    base_type: BaseType = pydantic.Field()
    strategy: typing.Optional[Strategy] = pydantic.Field(default=None)

Import

from cohere.finetuning.finetuning.types import BaseModel

I/O Contract

Fields

Field	Type	Required	Default	Description
`name`	`typing.Optional[str]`	No	`None`	The name of the base model.
`version`	`typing.Optional[str]`	No	`None`	Read-only. The version of the base model.
`base_type`	`BaseType`	Yes	(none)	The type of the base model. One of: `"BASE_TYPE_UNSPECIFIED"`, `"BASE_TYPE_GENERATIVE"`, `"BASE_TYPE_CLASSIFICATION"`, `"BASE_TYPE_RERANK"`, `"BASE_TYPE_CHAT"`.
`strategy`	`typing.Optional[Strategy]`	No	`None`	Deprecated. The fine-tuning strategy. One of: `"STRATEGY_UNSPECIFIED"`, `"STRATEGY_VANILLA"`, `"STRATEGY_TFEW"`.

Usage Examples

Creating a BaseModel instance

from cohere.finetuning.finetuning.types import BaseModel

# Minimal usage with only the required field
base_model = BaseModel(base_type="BASE_TYPE_CHAT")

print(base_model.base_type)  # "BASE_TYPE_CHAT"
print(base_model.name)       # None
print(base_model.version)    # None

Creating a fully specified BaseModel

from cohere.finetuning.finetuning.types import BaseModel

base_model = BaseModel(
    name="command",
    version="2024-01",
    base_type="BASE_TYPE_GENERATIVE",
    strategy="STRATEGY_VANILLA",
)

print(base_model.name)       # "command"
print(base_model.version)    # "2024-01"
print(base_model.base_type)  # "BASE_TYPE_GENERATIVE"
print(base_model.strategy)   # "STRATEGY_VANILLA"

Related Pages

Environment:Cohere_ai_Cohere_python_Python_SDK_Runtime
Cohere_ai_Cohere_python_Finetuning_Event - Lifecycle events during fine-tuned model training
Cohere_ai_Cohere_python_TrainingStepMetrics - Evaluation metrics at a training step
Cohere_ai_Cohere_python_WandbConfig - Weights & Biases integration configuration

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment