Models: Choosing from different LLMs and embedding models

Currently, many different LLMs are emerging. Edgechain offers integrations to a wide range of models and a streamlined interface to all of them. Edgechain differentiates between three types of models that differ in their inputs and outputs:

-LLMs take a string as an input (prompt) and output a string (completion).

@PostMapping("/chat")
  public ArkResponse chat(ArkRequest arkRequest) {

    OpenAiEndpoint gpt4Endpoint =
            new OpenAiEndpoint(
                    OPENAI_CHAT_COMPLETION_API,
                    OPENAI_AUTH_KEY,
                    "gpt-3.5-turbo",
                    "user",
                    0.7,
                    new ExponentialDelay(3, 5, 2, TimeUnit.SECONDS));

    String prompt = "Arakoo has a cat. What animal is Arakoo;s pet";

    EdgeChain<ChatCompletionResponse> chatChain = 
            new EdgeChain<>(gpt4Endpoint.chatCompletion(prompt, "ChatChain", arkRequest));

    return chatChain.getArkResponse();
  }

-Chat models are similar to LLMs. They take a list of chat messages as input and return a chat message.

-Text embedding models take text input and return a list of floats (embeddings), which are the numerical representation of the input text. Embeddings help extract information from a text. This information can then be later used, e.g., for calculating similarities between texts (e.g., movie summaries).

@PostMapping("/embedding")
  public ArkResponse embedding(ArkRequest arkRequest) {

    OpenAiEndpoint ada002Embeddings =
            new OpenAiEndpoint(
                    OPENAI_EMBEDDINGS_API,
                    OPENAI_AUTH_KEY,
                    "",//orgId
                    "text-embedding-ada-002",
                    "user",
                    0.7,
                    false,
                    new ExponentialDelay(3, 3, 2, TimeUnit.SECONDS));;

    String input = "Alice has a parrot. What animal is Alice;s pet";

    EdgeChain<ChatCompletionResponse> embeddingChain =
            new EdgeChain<>(gpt4Endpoint.embeddings(input, "EmbeddingChianChain", arkRequest));

    return embeddingChain.getArkResponse();
  }