Documentation
¶
Index ¶
- type RunPod
- func (r *RunPod) Destroy(ctx context.Context, id string) error
- func (r *RunPod) GPUTypes(ctx context.Context) ([]computedomain.GPUType, error)
- func (r *RunPod) List(ctx context.Context) ([]computedomain.GPUInstance, error)
- func (r *RunPod) Provision(ctx context.Context, req computedomain.ProvisionRequest) (computedomain.GPUInstance, error)
- func (r *RunPod) Status(ctx context.Context, id string) (computedomain.GPUInstance, error)
- func (r *RunPod) URL(id string, port int) string
- func (r *RunPod) WaitReady(ctx context.Context, id string, port int, report func(string)) (computedomain.GPUInstance, error)
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type RunPod ¶
type RunPod struct {
APIKey string
RestURL string // default https://rest.runpod.io/v1
GraphQLURL string // default https://api.runpod.io/graphql
ProxyBase string // overrides URL in WaitReady; tests only
Client *http.Client
// contains filtered or unexported fields
}
RunPod rents GPU instances from RunPod and translates its API into compute's domain. Base URLs are fields so tests can point at httptest.
func (*RunPod) GPUTypes ¶
GPUTypes returns rentable GPU types with live pricing, cheapest first. This is the one non-REST call (RunPod exposes pricing only via GraphQL).
func (*RunPod) List ¶
func (r *RunPod) List(ctx context.Context) ([]computedomain.GPUInstance, error)
List returns the instances infer rented, told by their name prefix. CRUD: read.
func (*RunPod) Provision ¶
func (r *RunPod) Provision(ctx context.Context, req computedomain.ProvisionRequest) (computedomain.GPUInstance, error)
Provision rents a GPU instance. CRUD: create.
func (*RunPod) Status ¶
func (r *RunPod) Status(ctx context.Context, id string) (computedomain.GPUInstance, error)
Status returns a single instance. CRUD: read.
func (*RunPod) WaitReady ¶
func (r *RunPod) WaitReady(ctx context.Context, id string, port int, report func(string)) (computedomain.GPUInstance, error)
WaitReady polls the pod, then the llama.cpp /health endpoint through the proxy until the model answers. /health is unauthenticated in llama.cpp, so this needs no per-session token and works for `gpu status --wait` too.
Source Files
Click to show internal directories.
Click to hide internal directories.