anchor-chain 0.4.2

A statically typed async framework for building LLM-based applications.
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
31
32
33
34
35

use std::collections::HashMap;

use anchor_chain::{
    to_boxed_future, ChainBuilder, Claude3Bedrock, OpenAIModel, ParallelNode, Prompt,
};

#[tokio::main]
async fn main() {
    let gpt3 =
        Box::new(OpenAIModel::new_gpt3_5_turbo("You are a helpful assistant".to_string()).await);
    let claude3 = Box::new(Claude3Bedrock::new("You are a helpful assistant".to_string()).await);

    let concat_fn = to_boxed_future(|outputs: Vec<String>| {
        Ok(outputs
            .iter()
            .enumerate()
            .map(|(i, output)| format!("Output {}:\n```\n{}\n```\n", i + 1, output))
            .collect::<Vec<String>>()
            .concat())
    });

    let chain = ChainBuilder::new()
        .link(Prompt::new("{{ input }}"))
        .link(ParallelNode::new(vec![gpt3, claude3], concat_fn))
        .build();

    let output = chain
        .process(HashMap::from([(
            "input".to_string(),
            "Write a hello world program in Rust".to_string(),
        )]))
        .await
        .expect("Error processing chain");
    println!("{}", output);
}