Writing a simple vAccel Go application🔗

For a simple example of using the vaccel Go package, you can replicate in Go an image classification application similar to the image classification example from Running the examples.

Initialize a new go-vaccel-classify project and module:

mkdir go-vaccel-classify
cd go-vaccel-classify
go mod init go-vaccel-classify

Create a new Go file main.go with the following content:

main.go

package main

import (
        "fmt"
        "os"

        "github.com/nubificus/go-vaccel/vaccel"
)

func main() {

        if len(os.Args) < 2 {
                fmt.Println("Usage: ./main <filename>")
                return
        }

        /* Get the filename from command line argument */
        filePath := os.Args[1]

        /* Session */
        var session vaccel.Session

        err := vaccel.SessionInit(&session, 0)

        if err != 0 {
                fmt.Println("error initializing session")
                os.Exit(err)
        }

        var outText string

        /* Read the image-bytes */
        imageBytes, e := os.ReadFile(filePath)
        if e != nil {
                fmt.Printf("Error reading file: %s\n", e)
                os.Exit(1)
        }

        /* Perform Image Classification */
        outText, err = vaccel.ImageClassification(&session, imageBytes)

        if err != 0 {
                fmt.Println("Image Classification failed")
                os.Exit(err)
        }

        fmt.Println("Output: ", outText)

        /* Free Session */
        if vaccel.SessionRelease(&session) != 0 {
                fmt.Println("An error occurred while releasing the session")
        }
}

Add all Go dependencies for your package:

go get .

And you are ready to run your application.

Same as with the upstream Go classify, configure vAccel:

export VACCEL_PLUGINS=libvaccel-noop.so

and run with:

$ go run main.go /usr/local/share/vaccel/images/example.jpg
Output:  This is a dummy classification tag!