CausalLmChatClientConfig

Configuration for adapting a Toro causal language model to IChatClient.

Record fields

NameDescription
CreateDecoderOptional factory for request-local incremental decoding.
DecodeDecode generated token IDs into response text.
DefaultMaxOutputTokensMaximum number of generated tokens when a request does not specify a limit.
EncodeEncode a formatted prompt into token IDs.
FormatPromptModel-specific formatter for validated text-only chat messages.
ModelBound causal language model used for generation.
ModelIdModel identifier reported through Microsoft.Extensions.AI responses and metadata.

CausalLmChatClient

Factory for Microsoft.Extensions.AI chat clients backed by Toro causal language models.

FunctionDescription
create configCreate a stateless IChatClient. The returned client does not own or dispose the model.

create

create config

Create a stateless IChatClient. The returned client does not own or dispose the model.

Parameters

  • config : CausalLmChatClientConfig<'a>

Returns IChatClient