haive.core.models.llm.providers.cohere

Cohere Provider Module.

This module implements the Cohere language model provider for the Haive framework, supporting Cohere’s Command models with advanced generation and reasoning capabilities.

The provider handles API key management, model configuration, and safe imports of the langchain-cohere package dependencies.

Examples

Basic usage:

from haive.core.models.llm.providers.cohere import CohereProvider

provider = CohereProvider(
    model="command-r-plus",
    temperature=0.7,
    max_tokens=1000
)
llm = provider.instantiate()

With custom parameters:

provider = CohereProvider(
    model="command-r",
    temperature=0.1,
    k=40,
    p=0.9,
    frequency_penalty=0.1
)

CohereProvider(*[, requests_per_second, ...])

Cohere language model provider configuration.

Classes

CohereProvider

Cohere language model provider configuration.

Module Contents

class haive.core.models.llm.providers.cohere.CohereProvider(/, **data)

Bases: haive.core.models.llm.providers.base.BaseLLMProvider

Cohere language model provider configuration.

This provider supports Cohere’s Command models including Command R+, Command R, and other models optimized for reasoning, generation, and multilingual tasks.

Parameters:

• data (Any)

• requests_per_second (float | None)

• tokens_per_second (int | None)

• tokens_per_minute (int | None)

• max_retries (int)

• retry_delay (float)

• check_every_n_seconds (float | None)

• burst_size (int | None)

• provider (LLMProvider)

• model (str | None)

• name (str | None)

• api_key (SecretStr)

• cache_enabled (bool)

• cache_ttl (int | None)

• extra_params (dict[str, Any] | None)

• debug (bool)

• temperature (float | None)

• max_tokens (int | None)

• k (int | None)

• p (float | None)

• frequency_penalty (float | None)

• presence_penalty (float | None)

• stop_sequences (list[str] | None)

provider

Always LLMProvider.COHERE

Type:

LLMProvider

model

The Cohere model to use

Type:

str

temperature

Sampling temperature (0.0-5.0)

Type:

float

max_tokens

Maximum tokens in response

Type:

int

k

Top-k sampling parameter

Type:

int

p

Top-p nucleus sampling parameter

Type:

float

frequency_penalty

Frequency penalty parameter

Type:

float

presence_penalty

Presence penalty parameter

Type:

float

stop_sequences

Stop sequences for generation

Type:

list

Examples

High-performance reasoning:

provider = CohereProvider(
    model="command-r-plus",
    temperature=0.3,
    max_tokens=2000,
    k=40
)

Creative writing:

provider = CohereProvider(
    model="command-r",
    temperature=0.9,
    p=0.95,
    frequency_penalty=0.2
)

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

classmethod get_models()

Get available Cohere models.

Return type:

list[str]

max_tokens: int | None = None

Get maximum total tokens for this model.