fix: untangle pkg/grpc and core/schema for Transcription (#3419)

untangle pkg/grpc and core/schema in Transcribe

Signed-off-by: Dave Lee <dave@gray101.com>

backend/go/transcribe/transcript.go

@@ -1,104 +0,0 @@
-package main
-
-import (
-	"fmt"
-	"os"
-	"os/exec"
-	"path/filepath"
-
-	"github.com/ggerganov/whisper.cpp/bindings/go/pkg/whisper"
-	"github.com/go-audio/wav"
-	"github.com/mudler/LocalAI/core/schema"
-)
-
-func ffmpegCommand(args []string) (string, error) {
-	cmd := exec.Command("ffmpeg", args...) // Constrain this to ffmpeg to permit security scanner to see that the command is safe.
-	cmd.Env = os.Environ()
-	out, err := cmd.CombinedOutput()
-	return string(out), err
-}
-
-// AudioToWav converts audio to wav for transcribe.
-// TODO: use https://github.com/mccoyst/ogg?
-func audioToWav(src, dst string) error {
-	commandArgs := []string{"-i", src, "-format", "s16le", "-ar", "16000", "-ac", "1", "-acodec", "pcm_s16le", dst}
-	out, err := ffmpegCommand(commandArgs)
-	if err != nil {
-		return fmt.Errorf("error: %w out: %s", err, out)
-	}
-	return nil
-}
-
-func Transcript(model whisper.Model, audiopath, language string, translate bool, threads uint) (schema.TranscriptionResult, error) {
-	res := schema.TranscriptionResult{}
-
-	dir, err := os.MkdirTemp("", "whisper")
-	if err != nil {
-		return res, err
-	}
-	defer os.RemoveAll(dir)
-
-	convertedPath := filepath.Join(dir, "converted.wav")
-
-	if err := audioToWav(audiopath, convertedPath); err != nil {
-		return res, err
-	}
-
-	// Open samples
-	fh, err := os.Open(convertedPath)
-	if err != nil {
-		return res, err
-	}
-	defer fh.Close()
-
-	// Read samples
-	d := wav.NewDecoder(fh)
-	buf, err := d.FullPCMBuffer()
-	if err != nil {
-		return res, err
-	}
-
-	data := buf.AsFloat32Buffer().Data
-
-	// Process samples
-	context, err := model.NewContext()
-	if err != nil {
-		return res, err
-
-	}
-
-	context.SetThreads(threads)
-
-	if language != "" {
-		context.SetLanguage(language)
-	} else {
-		context.SetLanguage("auto")
-	}
-
-	if translate {
-		context.SetTranslate(true)
-	}
-
-	if err := context.Process(data, nil, nil); err != nil {
-		return res, err
-	}
-
-	for {
-		s, err := context.NextSegment()
-		if err != nil {
-			break
-		}
-
-		var tokens []int
-		for _, t := range s.Tokens {
-			tokens = append(tokens, t.Id)
-		}
-
-		segment := schema.Segment{Id: s.Num, Text: s.Text, Start: s.Start, End: s.End, Tokens: tokens}
-		res.Segments = append(res.Segments, segment)
-
-		res.Text += s.Text
-	}
-
-	return res, nil
-}

backend/go/transcribe/whisper.go

@@ -1,26 +0,0 @@
-package main
-
-// This is a wrapper to statisfy the GRPC service interface
-// It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
-import (
-	"github.com/ggerganov/whisper.cpp/bindings/go/pkg/whisper"
-	"github.com/mudler/LocalAI/core/schema"
-	"github.com/mudler/LocalAI/pkg/grpc/base"
-	pb "github.com/mudler/LocalAI/pkg/grpc/proto"
-)
-
-type Whisper struct {
-	base.SingleThread
-	whisper whisper.Model
-}
-
-func (sd *Whisper) Load(opts *pb.ModelOptions) error {
-	// Note: the Model here is a path to a directory containing the model files
-	w, err := whisper.New(opts.ModelFile)
-	sd.whisper = w
-	return err
-}
-
-func (sd *Whisper) AudioTranscription(opts *pb.TranscriptRequest) (schema.TranscriptionResult, error) {
-	return Transcript(sd.whisper, opts.Dst, opts.Language, opts.Translate, uint(opts.Threads))
-}

backend/go/transcribe/whisper/whisper.go

@@ -0,0 +1,105 @@
+package main
+
+// This is a wrapper to statisfy the GRPC service interface
+// It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
+import (
+	"os"
+	"path/filepath"
+
+	"github.com/ggerganov/whisper.cpp/bindings/go/pkg/whisper"
+	"github.com/go-audio/wav"
+	"github.com/mudler/LocalAI/pkg/grpc/base"
+	pb "github.com/mudler/LocalAI/pkg/grpc/proto"
+	"github.com/mudler/LocalAI/pkg/utils"
+)
+
+type Whisper struct {
+	base.SingleThread
+	whisper whisper.Model
+}
+
+func (sd *Whisper) Load(opts *pb.ModelOptions) error {
+	// Note: the Model here is a path to a directory containing the model files
+	w, err := whisper.New(opts.ModelFile)
+	sd.whisper = w
+	return err
+}
+
+func (sd *Whisper) AudioTranscription(opts *pb.TranscriptRequest) (pb.TranscriptResult, error) {
+
+	dir, err := os.MkdirTemp("", "whisper")
+	if err != nil {
+		return pb.TranscriptResult{}, err
+	}
+	defer os.RemoveAll(dir)
+
+	convertedPath := filepath.Join(dir, "converted.wav")
+
+	if err := utils.AudioToWav(opts.Dst, convertedPath); err != nil {
+		return pb.TranscriptResult{}, err
+	}
+
+	// Open samples
+	fh, err := os.Open(convertedPath)
+	if err != nil {
+		return pb.TranscriptResult{}, err
+	}
+	defer fh.Close()
+
+	// Read samples
+	d := wav.NewDecoder(fh)
+	buf, err := d.FullPCMBuffer()
+	if err != nil {
+		return pb.TranscriptResult{}, err
+	}
+
+	data := buf.AsFloat32Buffer().Data
+
+	// Process samples
+	context, err := sd.whisper.NewContext()
+	if err != nil {
+		return pb.TranscriptResult{}, err
+
+	}
+
+	context.SetThreads(uint(opts.Threads))
+
+	if opts.Language != "" {
+		context.SetLanguage(opts.Language)
+	} else {
+		context.SetLanguage("auto")
+	}
+
+	if opts.Translate {
+		context.SetTranslate(true)
+	}
+
+	if err := context.Process(data, nil, nil); err != nil {
+		return pb.TranscriptResult{}, err
+	}
+
+	segments := []*pb.TranscriptSegment{}
+	text := ""
+	for {
+		s, err := context.NextSegment()
+		if err != nil {
+			break
+		}
+
+		var tokens []int32
+		for _, t := range s.Tokens {
+			tokens = append(tokens, int32(t.Id))
+		}
+
+		segment := &pb.TranscriptSegment{Id: int32(s.Num), Text: s.Text, Start: int64(s.Start), End: int64(s.End), Tokens: tokens}
+		segments = append(segments, segment)
+
+		text += s.Text
+	}
+
+	return pb.TranscriptResult{
+		Segments: segments,
+		Text:     text,
+	}, nil
+
+}