mirror of
https://github.com/lordmathis/llamactl.git
synced 2025-11-05 16:44:22 +00:00
322 lines
8.4 KiB
Go
322 lines
8.4 KiB
Go
package backends
|
|
|
|
import (
|
|
"encoding/json"
|
|
"fmt"
|
|
"llamactl/pkg/config"
|
|
"llamactl/pkg/validation"
|
|
"maps"
|
|
)
|
|
|
|
type BackendType string
|
|
|
|
const (
|
|
BackendTypeLlamaCpp BackendType = "llama_cpp"
|
|
BackendTypeMlxLm BackendType = "mlx_lm"
|
|
BackendTypeVllm BackendType = "vllm"
|
|
// BackendTypeMlxVlm BackendType = "mlx_vlm" // Future expansion
|
|
)
|
|
|
|
type Options struct {
|
|
BackendType BackendType `json:"backend_type"`
|
|
BackendOptions map[string]any `json:"backend_options,omitempty"`
|
|
|
|
// Backend-specific options
|
|
LlamaServerOptions *LlamaServerOptions `json:"-"`
|
|
MlxServerOptions *MlxServerOptions `json:"-"`
|
|
VllmServerOptions *VllmServerOptions `json:"-"`
|
|
}
|
|
|
|
func (o *Options) UnmarshalJSON(data []byte) error {
|
|
// Use anonymous struct to avoid recursion
|
|
type Alias Options
|
|
aux := &struct {
|
|
*Alias
|
|
}{
|
|
Alias: (*Alias)(o),
|
|
}
|
|
|
|
if err := json.Unmarshal(data, aux); err != nil {
|
|
return err
|
|
}
|
|
|
|
// Parse backend-specific options
|
|
switch o.BackendType {
|
|
case BackendTypeLlamaCpp:
|
|
if o.BackendOptions != nil {
|
|
// Convert map to JSON and then unmarshal to LlamaServerOptions
|
|
optionsData, err := json.Marshal(o.BackendOptions)
|
|
if err != nil {
|
|
return fmt.Errorf("failed to marshal backend options: %w", err)
|
|
}
|
|
|
|
o.LlamaServerOptions = &LlamaServerOptions{}
|
|
if err := json.Unmarshal(optionsData, o.LlamaServerOptions); err != nil {
|
|
return fmt.Errorf("failed to unmarshal llama.cpp options: %w", err)
|
|
}
|
|
}
|
|
case BackendTypeMlxLm:
|
|
if o.BackendOptions != nil {
|
|
optionsData, err := json.Marshal(o.BackendOptions)
|
|
if err != nil {
|
|
return fmt.Errorf("failed to marshal backend options: %w", err)
|
|
}
|
|
|
|
o.MlxServerOptions = &MlxServerOptions{}
|
|
if err := json.Unmarshal(optionsData, o.MlxServerOptions); err != nil {
|
|
return fmt.Errorf("failed to unmarshal MLX options: %w", err)
|
|
}
|
|
}
|
|
case BackendTypeVllm:
|
|
if o.BackendOptions != nil {
|
|
optionsData, err := json.Marshal(o.BackendOptions)
|
|
if err != nil {
|
|
return fmt.Errorf("failed to marshal backend options: %w", err)
|
|
}
|
|
|
|
o.VllmServerOptions = &VllmServerOptions{}
|
|
if err := json.Unmarshal(optionsData, o.VllmServerOptions); err != nil {
|
|
return fmt.Errorf("failed to unmarshal vLLM options: %w", err)
|
|
}
|
|
}
|
|
}
|
|
|
|
return nil
|
|
}
|
|
|
|
func (o *Options) MarshalJSON() ([]byte, error) {
|
|
// Use anonymous struct to avoid recursion
|
|
type Alias Options
|
|
aux := &struct {
|
|
*Alias
|
|
}{
|
|
Alias: (*Alias)(o),
|
|
}
|
|
|
|
// Prepare BackendOptions map
|
|
if o.BackendOptions == nil {
|
|
o.BackendOptions = make(map[string]any)
|
|
}
|
|
|
|
// Populate BackendOptions based on backend-specific options
|
|
switch o.BackendType {
|
|
case BackendTypeLlamaCpp:
|
|
if o.LlamaServerOptions != nil {
|
|
optionsData, err := json.Marshal(o.LlamaServerOptions)
|
|
if err != nil {
|
|
return nil, fmt.Errorf("failed to marshal llama.cpp options: %w", err)
|
|
}
|
|
if err := json.Unmarshal(optionsData, &o.BackendOptions); err != nil {
|
|
return nil, fmt.Errorf("failed to unmarshal llama.cpp options to map: %w", err)
|
|
}
|
|
}
|
|
case BackendTypeMlxLm:
|
|
if o.MlxServerOptions != nil {
|
|
optionsData, err := json.Marshal(o.MlxServerOptions)
|
|
if err != nil {
|
|
return nil, fmt.Errorf("failed to marshal MLX options: %w", err)
|
|
}
|
|
if err := json.Unmarshal(optionsData, &o.BackendOptions); err != nil {
|
|
return nil, fmt.Errorf("failed to unmarshal MLX options to map: %w", err)
|
|
}
|
|
}
|
|
case BackendTypeVllm:
|
|
if o.VllmServerOptions != nil {
|
|
optionsData, err := json.Marshal(o.VllmServerOptions)
|
|
if err != nil {
|
|
return nil, fmt.Errorf("failed to marshal vLLM options: %w", err)
|
|
}
|
|
if err := json.Unmarshal(optionsData, &o.BackendOptions); err != nil {
|
|
return nil, fmt.Errorf("failed to unmarshal vLLM options to map: %w", err)
|
|
}
|
|
}
|
|
}
|
|
|
|
return json.Marshal(aux)
|
|
}
|
|
|
|
func getBackendSettings(o *Options, backendConfig *config.BackendConfig) *config.BackendSettings {
|
|
switch o.BackendType {
|
|
case BackendTypeLlamaCpp:
|
|
return &backendConfig.LlamaCpp
|
|
case BackendTypeMlxLm:
|
|
return &backendConfig.MLX
|
|
case BackendTypeVllm:
|
|
return &backendConfig.VLLM
|
|
default:
|
|
return nil
|
|
}
|
|
}
|
|
|
|
func (o *Options) isDockerEnabled(backend *config.BackendSettings) bool {
|
|
if backend.Docker != nil && backend.Docker.Enabled && o.BackendType != BackendTypeMlxLm {
|
|
return true
|
|
}
|
|
return false
|
|
}
|
|
|
|
func (o *Options) IsDockerEnabled(backendConfig *config.BackendConfig) bool {
|
|
backendSettings := getBackendSettings(o, backendConfig)
|
|
return o.isDockerEnabled(backendSettings)
|
|
}
|
|
|
|
// GetCommand builds the command to run the backend
|
|
func (o *Options) GetCommand(backendConfig *config.BackendConfig) string {
|
|
|
|
backendSettings := getBackendSettings(o, backendConfig)
|
|
|
|
if o.isDockerEnabled(backendSettings) {
|
|
return "docker"
|
|
}
|
|
|
|
return backendSettings.Command
|
|
}
|
|
|
|
// buildCommandArgs builds command line arguments for the backend
|
|
func (o *Options) BuildCommandArgs(backendConfig *config.BackendConfig) []string {
|
|
|
|
var args []string
|
|
|
|
backendSettings := getBackendSettings(o, backendConfig)
|
|
|
|
if o.isDockerEnabled(backendSettings) {
|
|
// For Docker, start with Docker args
|
|
args = append(args, backendSettings.Docker.Args...)
|
|
args = append(args, backendSettings.Docker.Image)
|
|
|
|
switch o.BackendType {
|
|
case BackendTypeLlamaCpp:
|
|
if o.LlamaServerOptions != nil {
|
|
args = append(args, o.LlamaServerOptions.BuildDockerArgs()...)
|
|
}
|
|
case BackendTypeVllm:
|
|
if o.VllmServerOptions != nil {
|
|
args = append(args, o.VllmServerOptions.BuildDockerArgs()...)
|
|
}
|
|
}
|
|
|
|
} else {
|
|
// For native execution, start with backend args
|
|
args = append(args, backendSettings.Args...)
|
|
|
|
switch o.BackendType {
|
|
case BackendTypeLlamaCpp:
|
|
if o.LlamaServerOptions != nil {
|
|
args = append(args, o.LlamaServerOptions.BuildCommandArgs()...)
|
|
}
|
|
case BackendTypeMlxLm:
|
|
if o.MlxServerOptions != nil {
|
|
args = append(args, o.MlxServerOptions.BuildCommandArgs()...)
|
|
}
|
|
case BackendTypeVllm:
|
|
if o.VllmServerOptions != nil {
|
|
args = append(args, o.VllmServerOptions.BuildCommandArgs()...)
|
|
}
|
|
}
|
|
}
|
|
|
|
return args
|
|
}
|
|
|
|
// BuildEnvironment builds the environment variables for the backend process
|
|
func (o *Options) BuildEnvironment(backendConfig *config.BackendConfig, environment map[string]string) map[string]string {
|
|
|
|
backendSettings := getBackendSettings(o, backendConfig)
|
|
env := map[string]string{}
|
|
|
|
if backendSettings.Environment != nil {
|
|
maps.Copy(env, backendSettings.Environment)
|
|
}
|
|
|
|
if o.isDockerEnabled(backendSettings) {
|
|
if backendSettings.Docker.Environment != nil {
|
|
maps.Copy(env, backendSettings.Docker.Environment)
|
|
}
|
|
}
|
|
|
|
if environment != nil {
|
|
maps.Copy(env, environment)
|
|
}
|
|
|
|
return env
|
|
}
|
|
|
|
func (o *Options) GetPort() int {
|
|
if o != nil {
|
|
switch o.BackendType {
|
|
case BackendTypeLlamaCpp:
|
|
if o.LlamaServerOptions != nil {
|
|
return o.LlamaServerOptions.Port
|
|
}
|
|
case BackendTypeMlxLm:
|
|
if o.MlxServerOptions != nil {
|
|
return o.MlxServerOptions.Port
|
|
}
|
|
case BackendTypeVllm:
|
|
if o.VllmServerOptions != nil {
|
|
return o.VllmServerOptions.Port
|
|
}
|
|
}
|
|
}
|
|
return 0
|
|
}
|
|
|
|
func (o *Options) SetPort(port int) {
|
|
if o != nil {
|
|
switch o.BackendType {
|
|
case BackendTypeLlamaCpp:
|
|
if o.LlamaServerOptions != nil {
|
|
o.LlamaServerOptions.Port = port
|
|
}
|
|
case BackendTypeMlxLm:
|
|
if o.MlxServerOptions != nil {
|
|
o.MlxServerOptions.Port = port
|
|
}
|
|
case BackendTypeVllm:
|
|
if o.VllmServerOptions != nil {
|
|
o.VllmServerOptions.Port = port
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
func (o *Options) GetHost() string {
|
|
if o != nil {
|
|
switch o.BackendType {
|
|
case BackendTypeLlamaCpp:
|
|
if o.LlamaServerOptions != nil {
|
|
return o.LlamaServerOptions.Host
|
|
}
|
|
case BackendTypeMlxLm:
|
|
if o.MlxServerOptions != nil {
|
|
return o.MlxServerOptions.Host
|
|
}
|
|
case BackendTypeVllm:
|
|
if o.VllmServerOptions != nil {
|
|
return o.VllmServerOptions.Host
|
|
}
|
|
}
|
|
}
|
|
return "localhost"
|
|
}
|
|
|
|
func (o *Options) GetResponseHeaders(backendConfig *config.BackendConfig) map[string]string {
|
|
backendSettings := getBackendSettings(o, backendConfig)
|
|
return backendSettings.ResponseHeaders
|
|
}
|
|
|
|
// ValidateInstanceOptions performs validation based on backend type
|
|
func (o *Options) ValidateInstanceOptions() error {
|
|
// Validate based on backend type
|
|
switch o.BackendType {
|
|
case BackendTypeLlamaCpp:
|
|
return validateLlamaCppOptions(o.LlamaServerOptions)
|
|
case BackendTypeMlxLm:
|
|
return validateMlxOptions(o.MlxServerOptions)
|
|
case BackendTypeVllm:
|
|
return validateVllmOptions(o.VllmServerOptions)
|
|
default:
|
|
return validation.ValidationError(fmt.Errorf("unsupported backend type: %s", o.BackendType))
|
|
}
|
|
}
|