Files
LocalAI/pkg/grpc/transcription_live_test.go
Richard Palethorpe 5d0c43ec6e feat(realtime): Semantic VAD EOU token (#10444)
* feat(realtime): EOU-driven semantic_vad turn detection

Add a `semantic_vad` turn-detection mode to the realtime API that feeds
the transcription model live and decides "the user finished speaking"
from the `<EOU>` end-of-utterance token rather than from silence alone.
When EOU fires the turn commits immediately (~0.3s); otherwise it falls
back to an eagerness-scaled silence threshold (low/med/high = 8/4/2s).

Plumbing, bottom to top:

- proto: `AudioTranscriptionLive` bidirectional RPC (config-first oneof,
  mono float PCM @16k, ready-ack / Unimplemented degrade signal) plus
  `TranscriptResult.eou` for the unary retranscribe gate.
- pkg/grpc: client/server/base/embed scaffolding for the bidi stream,
  modeled on AudioTransformStream; release stream conns on terminal Recv.
- parakeet-cpp: live transcription RPC with per-C-call engine locking
  (one live stream per turn, finalize+free at commit); bump parakeet.cpp
  to ABI v5 — incremental StreamingMel (no more quadratic per-feed mel
  recompute that delayed EOU on long turns) and the <EOU>/<EOB> split;
  strip the literal <EOU>/<EOB> from offline text and set Eou.
- core/backend: LiveTranscriptionSession wrapper + pipeline
  `turn_detection:` config block (type/eagerness/retranscribe).
- realtime: semantic_vad integration — live input captions streamed as
  transcription deltas while the user speaks, EOU-immediate commit with
  eagerness fallback, optional retranscribe gate (batch re-decode must
  also end in <EOU> to confirm), clause synthesis off the LLM token
  callback, and per-turn live-transcription / model_load telemetry.
- UI: show the realtime pipeline components as a vertical list.

Docs and tests included; opt-in via the pipeline YAML or per-session
`session.update`. Non-streaming STT backends degrade to silence-only.

Assisted-by: Claude Code:claude-opus-4-8 [Read] [Edit] [Write] [Bash]
Assisted-by: Claude Code:claude-fable-5 [Read] [Edit] [Bash]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* feat(realtime): explicit formally-verified state machines + parakeet streaming driver

The realtime API had several implicit state machines whose state was inferred
from scattered booleans, channels, and five separate mutexes, leaving
illegal/inconsistent states reachable. Make them explicit and keep the
implementation in step with a formal design; rework the parakeet streaming
backend along the same lines.

Realtime state machines (M1-M5). Each is a sealed sum-type State/Event/Effect
with a total, pure Next(state,event)->(state,[]effect) behind a single-writer
Coordinator:

  M1 conncoord    connection lifecycle: VAD toggle + once-only teardown
                  (replaces vadServerStarted + a `done` channel closed from
                  two sites).
  M2 turncoord    turn detection: collapses speechStarted and the live-stream
                  "turn open" flag into one state, so discardTurn can no longer
                  desync them and suppress the next onset.
  M3 respcoord    response coordination: serializes the dual-writer
                  start/cancel so at most one response is live; one
                  response.done per response.create.
  M4 compactcoord conversation compaction: single-flight (replaces the
                  `compacting atomic.Bool` CAS).
  M5 ttscoord     TTS pipeline: open->closing->closed, idempotent wait(),
                  rejects enqueue-after-close (was a silent drop).

The Coordinator/Sink/Next plumbing — only the sealed types and Next differed
per machine — is extracted once into core/http/endpoints/openai/coordinator as
a generic Coordinator[S,E,F]; each machine keeps its public API via type
aliases, so no sink, call-site, or test moved.

Hierarchy. session_lifecycle.fizz models M1 as the parent region with its
children (M2/M3/M4) as one statechart and asserts ChildrenDieWithParent (conn
torn => all children terminal, none start after teardown). respcoord and
compactcoord gain an absorbing Terminated state + Shutdown event; conncoord's
teardown drives the children terminal. This closes a compaction teardown gap: a
fire-and-forget compaction could outlive a torn session — compactionSink now
takes a session-scoped cancellable context + WaitGroup and joins the in-flight
summarize+evict on shutdown.

Formal verification. formal-verification/ holds one authoritative FizzBee spec
per machine plus the composition spec, each with an always-assertion and a
documented one-line edit that makes the checker fail (verified non-vacuous).
scripts/realtime-conformance.sh is fail-closed: all Go conformance suites under
-race AND a model-check of every .fizz spec; a missing FizzBee is a hard error
(only the loud REALTIME_CONFORMANCE_SKIP_FIZZBEE=1 bypasses it, never in CI).
FizzBee is pinned by sha256 and installed via scripts/install-fizzbee.sh into
.tools/ (gitignored). Wired as make test-realtime-conformance, a CI workflow,
and a pre-commit path filter. Go conformance tests are Ginkgo/Gomega (per the
repo's forbidigo lint): transition tables + fixed-seed property walks +
concurrent/-race specs, no rapid dependency. Design map:
docs/design/realtime-state-machines.md.

Parakeet streaming backend. The same treatment applied to the parakeet-cpp
streaming paths:
- AudioTranscriptionStream returns codes.Unimplemented for non-streaming models
  instead of decoding offline and emitting it as one delta + final. A client
  that asked for streaming learns the model cannot stream rather than receiving
  a batch result shaped like a stream. New grpcerrors.StreamTranscriptionUnsupported
  carries that signal; the HTTP /v1/audio/transcriptions stream path surfaces it
  as an SSE error event. Mirrors AudioTranscriptionLive, which already did this.
- utteranceBoundary (boundary.go): a single definition of the end-of-utterance
  latch, replacing three open-coded finalEou toggles. Modelled as a two-valued
  type so illegal states are unrepresentable.
- Shared decode driver (driver.go): streamFeedResult (one per-feed event) +
  feedChunk (hides the ABI v4 JSON vs text-only split) + feedSlices + flushTail.
  The feed loop is written once.
- AudioTranscriptionLive becomes a bidi adapter: it streams the per-feed
  {delta,eou,eob,words} the realtime turn detector consumes and a terminal
  FinalResult carrying only Text. Segments/duration/eou are offline-only and no
  longer produced (nor read) on the live path; liveTraceState drops the terminal
  eou and keeps the per-feed eou_events count.
- AudioTranscriptionStream + streamJSON merge into one driver-based function;
  streamSegmenter is generalized to the unified event with a text-only fallback
  that preserves the legacy (no-words) library's per-utterance segmentation.

Verified: build/vet/gofumpt clean, golangci-lint 0 issues, all coordinator and
parakeet packages under -race, the fail-closed conformance gate green, and
make test-realtime (12 e2e WS+WebRTC).

Assisted-by: Claude:claude-opus-4-8 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

---------

Signed-off-by: Richard Palethorpe <io@richiejp.com>
2026-06-30 09:01:22 +02:00

121 lines
3.9 KiB
Go

package grpc
import (
"context"
"errors"
"fmt"
"io"
"strings"
"github.com/mudler/LocalAI/pkg/grpc/base"
"github.com/mudler/LocalAI/pkg/grpc/grpcerrors"
pb "github.com/mudler/LocalAI/pkg/grpc/proto"
. "github.com/onsi/ginkgo/v2"
. "github.com/onsi/gomega"
)
// echoLiveASRModel is a minimal AIModel exercising the AudioTranscriptionLive
// contract: ready ack after the config, one delta per audio frame (eou set on
// frames whose first sample is negative), final_result on close.
type echoLiveASRModel struct {
base.SingleThread
}
func (m *echoLiveASRModel) AudioTranscriptionLive(in <-chan *pb.TranscriptLiveRequest, out chan<- *pb.TranscriptLiveResponse) error {
defer close(out)
first, ok := <-in
if !ok || first.GetConfig() == nil {
return errors.New("first message must carry a config")
}
out <- &pb.TranscriptLiveResponse{Ready: true}
var full strings.Builder
for req := range in {
audio := req.GetAudio()
if audio == nil {
continue
}
delta := fmt.Sprintf("[%d]", len(audio.Pcm))
full.WriteString(delta)
out <- &pb.TranscriptLiveResponse{
Delta: delta,
Eou: len(audio.Pcm) > 0 && audio.Pcm[0] < 0,
}
}
out <- &pb.TranscriptLiveResponse{FinalResult: &pb.TranscriptResult{Text: full.String(), Eou: true}}
return nil
}
var _ = Describe("AudioTranscriptionLive RPC (in-process)", func() {
It("acks the config, streams deltas with eou flags, and finalizes on CloseSend", func() {
addr := "test://transcription-live-echo"
Provide(addr, &echoLiveASRModel{})
c := NewClient(addr, true, nil, false)
stream, err := c.AudioTranscriptionLive(context.Background())
Expect(err).NotTo(HaveOccurred())
Expect(stream.Send(&pb.TranscriptLiveRequest{
Payload: &pb.TranscriptLiveRequest_Config{Config: &pb.TranscriptLiveConfig{Language: "en"}},
})).To(Succeed())
// The ready ack must arrive before any audio is sent — this is what
// lets callers degrade synchronously when live ASR is unsupported.
ack, err := stream.Recv()
Expect(err).NotTo(HaveOccurred())
Expect(ack.Ready).To(BeTrue())
Expect(stream.Send(&pb.TranscriptLiveRequest{
Payload: &pb.TranscriptLiveRequest_Audio{Audio: &pb.TranscriptLiveAudio{Pcm: []float32{0.1, 0.2}}},
})).To(Succeed())
r1, err := stream.Recv()
Expect(err).NotTo(HaveOccurred())
Expect(r1.Delta).To(Equal("[2]"))
Expect(r1.Eou).To(BeFalse())
Expect(stream.Send(&pb.TranscriptLiveRequest{
Payload: &pb.TranscriptLiveRequest_Audio{Audio: &pb.TranscriptLiveAudio{Pcm: []float32{-0.5, 0.0, 0.5}}},
})).To(Succeed())
r2, err := stream.Recv()
Expect(err).NotTo(HaveOccurred())
Expect(r2.Delta).To(Equal("[3]"))
Expect(r2.Eou).To(BeTrue())
Expect(stream.CloseSend()).To(Succeed())
final, err := stream.Recv()
Expect(err).NotTo(HaveOccurred())
Expect(final.FinalResult).NotTo(BeNil())
Expect(final.FinalResult.Text).To(Equal("[2][3]"))
Expect(final.FinalResult.Eou).To(BeTrue())
_, err = stream.Recv()
Expect(errors.Is(err, io.EOF)).To(BeTrue())
})
It("surfaces Unimplemented from base.Base on the first Recv without CloseSend", func() {
// The ready-ack contract: a caller that sent its config and blocks on
// the first Recv must get the unsupported signal immediately — NOT
// after it gives up and closes the send side.
addr := "test://transcription-live-base"
Provide(addr, &base.SingleThread{})
c := NewClient(addr, true, nil, false)
stream, err := c.AudioTranscriptionLive(context.Background())
Expect(err).NotTo(HaveOccurred())
Expect(stream.Send(&pb.TranscriptLiveRequest{
Payload: &pb.TranscriptLiveRequest_Config{Config: &pb.TranscriptLiveConfig{}},
})).To(Succeed())
_, err = stream.Recv()
Expect(err).To(HaveOccurred())
Expect(grpcerrors.IsLiveTranscriptionUnsupported(err)).To(BeTrue())
// Degrading callers close the send side; in embed mode this is also
// what unwinds the server-side recv pump.
Expect(stream.CloseSend()).To(Succeed())
})
})