Documentation
¶
Overview ¶
Package ollama wires local Ollama chat models into loom-mcp planners.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type Client ¶
type Client struct {
// contains filtered or unexported fields
}
Client implements model.Client using Ollama's /api/chat endpoint.
type Options ¶
type Options struct {
// HTTPClient is used for requests. When nil, a client is created from Timeout.
HTTPClient *http.Client
// ServerURL is the local Ollama base URL, for example http://localhost:11434.
ServerURL string
// DefaultModel is the model identifier used when Request.Model is empty.
DefaultModel string
// HighModel is used when Request.ModelClass is high-reasoning and
// Request.Model is empty.
HighModel string
// SmallModel is used when Request.ModelClass is small and Request.Model is
// empty.
SmallModel string
// MaxTokens is mapped to Ollama's num_predict option when Request.MaxTokens
// is unset.
MaxTokens int
// Temperature is the default sampling temperature when Request.Temperature
// is unset.
Temperature float32
// Timeout configures the response header timeout when HTTPClient is nil.
// Request lifetime deadlines should be owned by the request context.
Timeout time.Duration
}
Options configures the Ollama adapter.
Click to show internal directories.
Click to hide internal directories.