daimon-provider-local 0.23.0

Locally-hosted model providers (Ollama, llama.cpp, llama-rs, OpenAI-compatible) for the Daimon AI agent framework
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

//! Locally-hosted model providers for the Daimon agent framework.
//!
//! Covers every runtime you can point at a machine under your own control:
//!
//! - [`ollama`] — [Ollama](https://ollama.com), via its native `/api/chat` API
//! - [`llamacpp`] — [llama.cpp](https://github.com/ggml-org/llama.cpp)'s
//!   `llama-server`, over its OpenAI-compatible API plus llama.cpp-native
//!   sampling extras
//! - [`llamars`] — [llama-rs](https://github.com/Lexmata/llama-rs)'s server,
//!   over its OpenAI-compatible API
//! - [`generic`] — any other OpenAI-compatible server (vLLM, LM Studio,
//!   llamafile, LocalAI, …)
//!
//! All four are HTTP clients; none embed inference in-process.

mod openai_compat;

pub mod generic;
pub mod llamacpp;
pub mod llamacpp_embed;
pub mod llamars;
pub mod ollama;
pub mod ollama_embed;

pub use generic::{OpenAiCompatible, OpenAiCompatibleEmbedding};
pub use llamacpp::LlamaCpp;
pub use llamacpp_embed::LlamaCppEmbedding;
pub use llamars::{LlamaRs, LlamaRsEmbedding};
pub use ollama::Ollama;
pub use ollama_embed::OllamaEmbedding;