modeldapi

type ActiveModel struct {
	ModelName  string        `json:"modelName,omitempty" example:"qwen3-8b"`
	Type       string        `json:"type,omitempty" example:"llama"`
	Digest     string        `json:"digest,omitempty" example:"sha256:abcdef"`
	Adapters   []AdapterInfo `json:"adapters,omitempty" openapi_include_type:"modeldapi.AdapterInfo"`
	Config     RuntimeConfig `json:"config"`
	Generation uint64        `json:"generation" example:"3"`
}

type AdapterInfo struct {
	Name   string  `json:"name,omitempty" example:"style"`
	Digest string  `json:"digest,omitempty" example:"sha256:abcdef"`
	Scale  float32 `json:"scale,omitempty" example:"1"`
}

type CapacityDevice struct {
	Index            int    `json:"index"`
	Name             string `json:"name,omitempty"`
	Description      string `json:"description,omitempty"`
	Type             string `json:"type,omitempty"`
	MemoryFree       int64  `json:"memoryFree,omitempty"`
	MemoryTotal      int64  `json:"memoryTotal,omitempty"`
	MemoryFreeKnown  bool   `json:"memoryFreeKnown,omitempty"`
	MemoryTotalKnown bool   `json:"memoryTotalKnown,omitempty"`
}

type CapacityInfo struct {
	ModelMaxContext                     int              `json:"modelMaxContext"`
	EffectiveContext                    int              `json:"effectiveContext"`
	MemoryContextTokens                 int              `json:"memoryContextTokens,omitempty"`
	HotContextTokens                    int              `json:"hotContextTokens,omitempty"`
	PlannerEffectiveContext             int              `json:"plannerEffectiveContext,omitempty"`
	KVBytesPerToken                     int64            `json:"kvBytesPerToken,omitempty"`
	FreeBytes                           int64            `json:"freeBytes,omitempty"`
	WeightsBytes                        int64            `json:"weightsBytes,omitempty"`
	OverheadBytes                       int64            `json:"overheadBytes,omitempty"`
	ReservedBytes                       int64            `json:"reservedBytes,omitempty"`
	UserLimitBytes                      int64            `json:"userLimitBytes,omitempty"`
	MinFreeBytes                        int64            `json:"minFreeBytes,omitempty"`
	HostColdBudgetBytes                 int64            `json:"hostColdBudgetBytes,omitempty"`
	UsableBytes                         int64            `json:"usableBytes,omitempty"`
	RequiredBytes                       int64            `json:"requiredBytes,omitempty"`
	Clamped                             bool             `json:"clamped,omitempty"`
	Reason                              string           `json:"reason,omitempty"`
	DeviceKind                          string           `json:"deviceKind,omitempty"`
	DeviceID                            string           `json:"deviceId,omitempty"`
	DeviceTotalBytes                    int64            `json:"deviceTotalBytes,omitempty"`
	SharedWithDisplay                   bool             `json:"sharedWithDisplay,omitempty"`
	RequestedGpuLayers                  int              `json:"requestedGpuLayers,omitempty"`
	ResolvedGpuLayers                   int              `json:"resolvedGpuLayers,omitempty"`
	SparseAttention                     bool             `json:"sparseAttention,omitempty"`
	SlidingWindowAttentionTokens        int              `json:"slidingWindowAttentionTokens,omitempty"`
	ChatTemplateFormat                  string           `json:"chatTemplateFormat,omitempty"`
	ChatTemplateThinkingStartTag        string           `json:"chatTemplateThinkingStartTag,omitempty"`
	ChatTemplateReasoningFormat         string           `json:"chatTemplateReasoningFormat,omitempty"`
	ChatTemplateSupportsToolCalls       bool             `json:"chatTemplateSupportsToolCalls,omitempty"`
	ChatTemplateSupportsThinking        bool             `json:"chatTemplateSupportsThinking,omitempty"`
	ChatTemplateSupportsReasoningEffort bool             `json:"chatTemplateSupportsReasoningEffort,omitempty"`
	RuntimeName                         string           `json:"runtimeName,omitempty"`
	RuntimeDigest                       string           `json:"runtimeDigest,omitempty"`
	RuntimeSystemInfo                   string           `json:"runtimeSystemInfo,omitempty"`
	SupportsGPUOffload                  bool             `json:"supportsGpuOffload,omitempty"`
	Devices                             []CapacityDevice `json:"devices,omitempty" openapi_include_type:"modeldapi.CapacityDevice"`
}

type CapacityResponse struct {
	Model LocalModel   `json:"model" openapi_include_type:"modeldapi.LocalModel"`
	Info  CapacityInfo `json:"info" openapi_include_type:"modeldapi.CapacityInfo"`
}

type LoadRequest struct {
	Model              string  `json:"model" example:"llama:qwen3-8b"`
	ExpectedGeneration *uint64 `json:"expectedGeneration,omitempty" example:"3"`
}

type LoadResponse struct {
	Loaded             bool        `json:"loaded" example:"true"`
	ExpectedGeneration uint64      `json:"expectedGeneration,omitempty" example:"3"`
	Active             ActiveModel `json:"active" openapi_include_type:"modeldapi.ActiveModel"`
}

type LocalModel struct {
	ID              string `json:"id" example:"llama:qwen3-8b"`
	Model           string `json:"model" example:"qwen3-8b"`
	Name            string `json:"name,omitempty" example:"Qwen3 8B"`
	BackendID       string `json:"backendId,omitempty" example:"b7d9e1a3-8f0c-4a7d-9b1e-2f3a4b5c6d7e"`
	BackendName     string `json:"backendName,omitempty" example:"llama"`
	BackendType     string `json:"backendType" example:"llama"`
	Digest          string `json:"digest,omitempty" example:"sha256:abcdef"`
	ContextLength   int    `json:"contextLength,omitempty" example:"8192"`
	MaxOutputTokens int    `json:"maxOutputTokens,omitempty" example:"4096"`
	CanChat         bool   `json:"canChat"`
	CanEmbed        bool   `json:"canEmbed"`
	CanPrompt       bool   `json:"canPrompt"`
	CanStream       bool   `json:"canStream"`
	CanThink        bool   `json:"canThink,omitempty"`
}

type Option func(*handler)

type RuntimeConfig struct {
	NumCtx                  int       `json:"numCtx,omitempty"`
	HotContextTokens        int       `json:"hotContextTokens,omitempty"`
	PlannerEffectiveContext int       `json:"plannerEffectiveContext,omitempty"`
	NumBatch                int       `json:"numBatch,omitempty"`
	NumThreads              int       `json:"numThreads,omitempty"`
	NumGpuLayers            int       `json:"numGpuLayers,omitempty"`
	TensorSplit             []float32 `json:"tensorSplit,omitempty"`
	FlashAttn               bool      `json:"flashAttn,omitempty"`
	KVCacheType             string    `json:"kvCacheType,omitempty"`
	PromptFormat            string    `json:"promptFormat,omitempty"`
	PromptTemplateDigest    string    `json:"promptTemplateDigest,omitempty"`
	DisableBOS              bool      `json:"disableBOS,omitempty"`
	ReasoningFormat         string    `json:"reasoningFormat,omitempty"`
}

type SlotStatus struct {
	OwnerInstanceID string       `json:"ownerInstanceId,omitempty" example:"5f2a23ad-3d9f-46dd-bc21-4c6c2f901e22"`
	Backend         string       `json:"backend,omitempty" example:"llama"`
	State           string       `json:"state,omitempty" example:"Ready"`
	Active          *ActiveModel `json:"active,omitempty" openapi_include_type:"modeldapi.ActiveModel"`
	BusyOperation   string       `json:"busyOperation,omitempty" example:"load"`
	LastError       string       `json:"lastError,omitempty" example:"model does not fit"`
}

type StatusResponse struct {
	State              string      `json:"state" example:"running"`
	Available          bool        `json:"available" example:"true"`
	Binary             string      `json:"binary,omitempty" example:"/home/user/.contenox/modeld/v0.1.0/linux-amd64/modeld"`
	Endpoint           string      `json:"endpoint,omitempty" example:"127.0.0.1:42001"`
	Instance           string      `json:"instance,omitempty" example:"5f2a23ad-3d9f-46dd-bc21-4c6c2f901e22"`
	Backend            string      `json:"backend,omitempty" example:"llama"`
	Error              string      `json:"error,omitempty" example:"modeld is not running"`
	RuntimeProtocol    int         `json:"runtimeProtocol" example:"1"`
	MinRuntimeProtocol int         `json:"minRuntimeProtocol" example:"1"`
	Slot               *SlotStatus `json:"slot,omitempty" openapi_include_type:"modeldapi.SlotStatus"`
}

type UnloadRequest struct {
	ExpectedGeneration *uint64 `json:"expectedGeneration" example:"3"`
}

type UnloadResponse struct {
	Unloaded           bool   `json:"unloaded" example:"true"`
	ExpectedGeneration uint64 `json:"expectedGeneration" example:"3"`
}