forked from mudler/LocalAI
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathmain.go
More file actions
149 lines (140 loc) · 4.21 KB
/
main.go
File metadata and controls
149 lines (140 loc) · 4.21 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
package main
import (
"fmt"
"os"
"path/filepath"
api "github.com/go-skynet/LocalAI/api"
model "github.com/go-skynet/LocalAI/pkg/model"
"github.com/rs/zerolog"
"github.com/rs/zerolog/log"
"github.com/urfave/cli/v2"
)
func main() {
log.Logger = log.Output(zerolog.ConsoleWriter{Out: os.Stderr})
path, err := os.Getwd()
if err != nil {
log.Error().Msgf("error: %s", err.Error())
os.Exit(1)
}
app := &cli.App{
Name: "LocalAI",
Usage: "OpenAI compatible API for running LLaMA/GPT models locally on CPU with consumer grade hardware.",
Flags: []cli.Flag{
&cli.BoolFlag{
Name: "f16",
EnvVars: []string{"F16"},
},
&cli.BoolFlag{
Name: "debug",
EnvVars: []string{"DEBUG"},
},
&cli.BoolFlag{
Name: "cors",
EnvVars: []string{"CORS"},
},
&cli.StringFlag{
Name: "cors-allow-origins",
EnvVars: []string{"CORS_ALLOW_ORIGINS"},
},
&cli.IntFlag{
Name: "threads",
DefaultText: "Number of threads used for parallel computation. Usage of the number of physical cores in the system is suggested.",
EnvVars: []string{"THREADS"},
Value: 4,
},
&cli.StringFlag{
Name: "models-path",
DefaultText: "Path containing models used for inferencing",
EnvVars: []string{"MODELS_PATH"},
Value: filepath.Join(path, "models"),
},
&cli.StringFlag{
Name: "preload-models",
DefaultText: "A List of models to apply in JSON at start",
EnvVars: []string{"PRELOAD_MODELS"},
},
&cli.StringFlag{
Name: "preload-models-config",
DefaultText: "A List of models to apply at startup. Path to a YAML config file",
EnvVars: []string{"PRELOAD_MODELS_CONFIG"},
},
&cli.StringFlag{
Name: "config-file",
DefaultText: "Config file",
EnvVars: []string{"CONFIG_FILE"},
},
&cli.StringFlag{
Name: "address",
DefaultText: "Bind address for the API server.",
EnvVars: []string{"ADDRESS"},
Value: ":8080",
},
&cli.StringFlag{
Name: "image-path",
DefaultText: "Image directory",
EnvVars: []string{"IMAGE_PATH"},
Value: "",
},
&cli.StringFlag{
Name: "backend-assets-path",
DefaultText: "Path used to extract libraries that are required by some of the backends in runtime.",
EnvVars: []string{"BACKEND_ASSETS_PATH"},
Value: "/tmp/localai/backend_data",
},
&cli.IntFlag{
Name: "context-size",
DefaultText: "Default context size of the model",
EnvVars: []string{"CONTEXT_SIZE"},
Value: 512,
},
&cli.IntFlag{
Name: "upload-limit",
DefaultText: "Default upload-limit. MB",
EnvVars: []string{"UPLOAD_LIMIT"},
Value: 15,
},
},
Description: `
LocalAI is a drop-in replacement OpenAI API which runs inference locally.
Some of the models compatible are:
- Vicuna
- Koala
- GPT4ALL
- GPT4ALL-J
- Cerebras
- Alpaca
- StableLM (ggml quantized)
It uses llama.cpp, ggml and gpt4all as backend with golang c bindings.
`,
UsageText: `local-ai [options]`,
Copyright: "go-skynet authors",
Action: func(ctx *cli.Context) error {
fmt.Printf("Starting LocalAI using %d threads, with models path: %s\n", ctx.Int("threads"), ctx.String("models-path"))
app, err := api.App(
api.WithConfigFile(ctx.String("config-file")),
api.WithJSONStringPreload(ctx.String("preload-models")),
api.WithYAMLConfigPreload(ctx.String("preload-models-config")),
api.WithModelLoader(model.NewModelLoader(ctx.String("models-path"))),
api.WithContextSize(ctx.Int("context-size")),
api.WithDebug(ctx.Bool("debug")),
api.WithImageDir(ctx.String("image-path")),
api.WithF16(ctx.Bool("f16")),
api.WithDisableMessage(false),
api.WithCors(ctx.Bool("cors")),
api.WithCorsAllowOrigins(ctx.String("cors-allow-origins")),
api.WithThreads(ctx.Int("threads")),
api.WithBackendAssets(backendAssets),
api.WithBackendAssetsOutput(ctx.String("backend-assets-path")),
api.WithUploadLimitMB(ctx.Int("upload-limit")))
if err != nil {
return err
}
return app.Listen(ctx.String("address"))
},
}
err = app.Run(os.Args)
if err != nil {
log.Error().Msgf("error: %s", err.Error())
os.Exit(1)
}
}