ollama

README

Ollama LLM Inference Provider

Build up your YoMo AI architecture with fully open-source dependencies.

1. Install Ollama

Follow the Ollama doc:

https://github.com/ollama/ollama?tab=readme-ov-file#ollama

2. Run the Mistral model

Notice that only the Mistral v0.3+ models are supported currently.

ollama run mistral:7b

3. Start YoMo Zipper

By default, ollama server will be listening at the port 11434.

Then the config file could be:

name: Service
host: 0.0.0.0
port: 9000

bridge:
  ai:
    server:
      addr: "localhost:8000"
      provider: ollama
    providers:
      ollama:
        api_endpoint: "http://localhost:11434/"

yomo serve -c config.yml

4. Start YoMo serverless function

example

Documentation

Index

• type Provider
  • func NewProvider(endpoint string) *Provider
  • func (p *Provider) GetChatCompletions(ctx context.Context, req openai.ChatCompletionRequest, _ metadata.M) (openai.ChatCompletionResponse, error)
  • func (p *Provider) GetChatCompletionsStream(ctx context.Context, req openai.ChatCompletionRequest, _ metadata.M) (ai.ResponseRecver, error)
  • func (p *Provider) Name() string

Types

type Provider

type Provider struct {
	Endpoint string
}

Provider is the provider for Ollama

func NewProvider

func NewProvider(endpoint string) *Provider

NewProvider creates a new OllamaProvider

func (*Provider) GetChatCompletions

func (p *Provider) GetChatCompletions(ctx context.Context, req openai.ChatCompletionRequest, _ metadata.M) (openai.ChatCompletionResponse, error)

GetChatCompletions implements ai.LLMProvider.

func (*Provider) GetChatCompletionsStream

func (p *Provider) GetChatCompletionsStream(ctx context.Context, req openai.ChatCompletionRequest, _ metadata.M) (ai.ResponseRecver, error)

GetChatCompletionsStream implements ai.LLMProvider.

func (*Provider) Name

func (p *Provider) Name() string

Name implements ai.LLMProvider.