Documentation
¶
Overview ¶
Package langchaingo provides an adapter for using LangChainGo (github.com/tmc/langchaingo) models within AgentMesh workflows.
This adapter enables integration with LangChainGo's 50+ model providers (OpenAI, Anthropic, Google AI, Cohere, local models, etc.) while using the AgentMesh model interface for agents and graph execution.
Example usage:
import (
"github.com/hupe1980/agentmesh/pkg/model/langchaingo"
"github.com/tmc/langchaingo/llms/openai"
)
// Create a LangChainGo model
llm, err := openai.New(openai.WithModel("gpt-4"))
if err != nil {
log.Fatal(err)
}
// Wrap it as an AgentMesh model
model, _ := langchaingo.NewModel(llm)
// Use with AgentMesh agents
agent, err := agent.NewReAct(model, tools)
Package langchaingo provides sentinel errors for the LangChain Go model package.
Index ¶
Constants ¶
This section is empty.
Variables ¶
var ( // ErrNoMessages is returned when generate is called without messages. ErrNoMessages = errors.New("model/langchaingo: generate requires at least one message") )
Functions ¶
This section is empty.
Types ¶
type Model ¶
type Model struct {
// contains filtered or unexported fields
}
Model wraps a LangChainGo llms.Model to implement the AgentMesh model.Model interface.
func MustNewModel ¶
MustNewModel creates a new model adapter, panicking on error. Use this only when you can guarantee the llm is non-nil.
func NewModel ¶
NewModel creates a new AgentMesh model adapter from a LangChainGo model. Returns an error if the llm parameter is nil.
func (*Model) BindTools ¶
BindTools returns a new model with the specified tools bound for function calling.
func (*Model) Capabilities ¶
func (m *Model) Capabilities() model.Capabilities
Capabilities returns the features supported by this model adapter. Note: LangChainGo doesn't expose capability introspection, so we return conservative defaults. Streaming support depends on the underlying model.
type Option ¶
type Option func(*Options)
Option is a function that configures Options.
func WithMaxTokens ¶
WithMaxTokens sets the maximum tokens to generate.
func WithStopWords ¶
WithStopWords sets the stop sequences for generation.
func WithStreaming ¶
WithStreaming enables or disables streaming mode.
func WithTemperature ¶
WithTemperature sets the temperature for generation.
type Options ¶
type Options struct {
// Temperature controls randomness in output (0.0 to 1.0).
// Higher values produce more random output.
Temperature float64
// MaxTokens is the maximum number of tokens to generate.
MaxTokens int
// StopWords are sequences that will stop generation when encountered.
StopWords []string
// Streaming enables streaming mode when true.
Streaming bool
}
Options configures the LangChainGo model adapter behavior.
Source Files