Enha: embedgemma model

This commit is contained in:
Grail Finder
2025-11-22 14:56:24 +03:00
parent 5fe03fa66c
commit 50d7bfced3
4 changed files with 123 additions and 36 deletions

View File

@@ -148,10 +148,12 @@ func (r *RAG) LoadRAG(fpath string) error {
for w := 0; w < int(r.cfg.RAGWorkers); w++ {
go r.batchToVectorAsync(lock, w, batchCh, vectorCh, errCh, doneCh, path.Base(fpath))
}
// Wait for embedding to be done
<-doneCh
err = <-errCh
if err != nil {
return err
}
// Write vectors to storage
return r.writeVectors(vectorCh)
}
@@ -178,9 +180,11 @@ func (r *RAG) writeVectors(vectorCh chan []models.VectorRow) error {
func (r *RAG) batchToVectorAsync(lock *sync.Mutex, id int, inputCh <-chan map[int][]string,
vectorCh chan<- []models.VectorRow, errCh chan error, doneCh chan bool, filename string) {
var err error
defer func() {
if len(doneCh) == 0 {
doneCh <- true
errCh <- err
}
}()
@@ -201,7 +205,7 @@ func (r *RAG) batchToVectorAsync(lock *sync.Mutex, id int, inputCh <-chan map[in
}
}
lock.Unlock()
case err := <-errCh:
case err = <-errCh:
r.logger.Error("got an error from error channel", "error", err)
lock.Unlock()
return
@@ -215,7 +219,23 @@ func (r *RAG) batchToVectorAsync(lock *sync.Mutex, id int, inputCh <-chan map[in
}
func (r *RAG) fetchEmb(lines []string, errCh chan error, vectorCh chan<- []models.VectorRow, slug, filename string) error {
embeddings, err := r.embedder.Embed(lines)
// Filter out empty lines before sending to embedder
nonEmptyLines := make([]string, 0, len(lines))
for _, line := range lines {
trimmed := strings.TrimSpace(line)
if trimmed != "" {
nonEmptyLines = append(nonEmptyLines, trimmed)
}
}
// Skip if no non-empty lines
if len(nonEmptyLines) == 0 {
// Send empty result but don't error
vectorCh <- []models.VectorRow{}
return nil
}
embeddings, err := r.embedder.EmbedSlice(nonEmptyLines)
if err != nil {
r.logger.Error("failed to embed lines", "err", err.Error())
errCh <- err
@@ -229,15 +249,22 @@ func (r *RAG) fetchEmb(lines []string, errCh chan error, vectorCh chan<- []model
return err
}
vectors := make([]models.VectorRow, len(embeddings))
for i, emb := range embeddings {
vector := models.VectorRow{
Embeddings: emb,
RawText: lines[i],
if len(embeddings) != len(nonEmptyLines) {
err := errors.New("mismatch between number of lines and embeddings returned")
r.logger.Error("embedding mismatch", "err", err.Error())
errCh <- err
return err
}
// Create a VectorRow for each line in the batch
vectors := make([]models.VectorRow, len(nonEmptyLines))
for i, line := range nonEmptyLines {
vectors[i] = models.VectorRow{
Embeddings: embeddings[i],
RawText: line,
Slug: fmt.Sprintf("%s_%d", slug, i),
FileName: filename,
}
vectors[i] = vector
}
vectorCh <- vectors
@@ -245,7 +272,7 @@ func (r *RAG) fetchEmb(lines []string, errCh chan error, vectorCh chan<- []model
}
func (r *RAG) LineToVector(line string) ([]float32, error) {
return r.embedder.EmbedSingle(line)
return r.embedder.Embed(line)
}
func (r *RAG) SearchEmb(emb *models.EmbeddingResp) ([]models.VectorRow, error) {
@@ -259,4 +286,3 @@ func (r *RAG) ListLoaded() ([]string, error) {
func (r *RAG) RemoveFile(filename string) error {
return r.storage.RemoveEmbByFileName(filename)
}