retrival mit Jina hinzugefügt

math-tutor/backend/app/api/retrieval.py

@@ -65,7 +65,7 @@ def ingest(request: IngestRequest) -> dict:
 
     embedder = _get_embedder()
     docs = vector_store.load_docs(base_dir)
-    embeddings = vector_store.embed_passages(embedder, [doc.markdown for doc in docs])
+    embeddings = vector_store.embed_documents(embedder, [doc.markdown for doc in docs])
     upserted = vector_store.upsert_docs(pg_url, docs, embeddings)
     return {"status": "ok", "upserted": upserted}
 

math-tutor/backend/app/config.py

@@ -19,7 +19,7 @@ def get_embedding_settings() -> EmbeddingSettings:
     if embedding_type == "sentence-transformer":
         return EmbeddingSettings(
             embedding_type=embedding_type,
-            model=os.getenv("SENTENCE_TRANSFORMER_MODEL", "all-MiniLM-L6-v2"),
+            model=os.getenv("SENTENCE_TRANSFORMER_MODEL", "tencent/KaLM-Embedding-Gemma3-12B-2511"),
             target_dim=int(os.getenv("SENTENCE_TRANSFORMER_TARGET_DIM", "1024")),
         )
     if embedding_type == "openai-like":

math-tutor/backend/app/deterministic_services/embeddings.py

@@ -7,6 +7,7 @@ from enum import Enum
 import httpx
 from pydantic import BaseModel, Field
 from sentence_transformers import SentenceTransformer
+import torch
 
 
 # -----------------------------
@@ -29,7 +30,7 @@ class OpenAILikeConfig(BaseModel):
 class SentenceTransformerConfig(BaseModel):
     """Konfiguration für lokale SentenceTransformer-Modelle."""
     model: str = Field(..., description="Name des SentenceTransformer-Modells (z. B. 'all-MiniLM-L6-v2')")
-    target_dim: int = Field(1024, description="Ziel-Dimension der Embeddings")
+    target_dim: int = Field(384, description="Ziel-Dimension der Embeddings")
 
 
 class EmbeddingConfig(BaseModel):
@@ -46,7 +47,7 @@ class BaseEmbeddings:
     Basisklasse für Embedding-Generierung mit gemeinsamen Methoden.
     """
 
-    def __init__(self, target_dim: int = 1024) -> None:
+    def __init__(self, target_dim: int = 384) -> None:
         self.target_dim = target_dim
 
     def _normalize(self, vec: List[float]) -> List[float]:
@@ -148,18 +149,28 @@ class SentenceTransformerEmbeddings(BaseEmbeddings):
     def model(self) -> SentenceTransformer:
         """Liefert das SentenceTransformer-Modell (lazy load)."""
         if self._model is None:
-            self._model = SentenceTransformer(self.model_name)
+            self._model = SentenceTransformer(self.model_name, trust_remote_code=True)
+            self._model.max_seq_length = 512
         return self._model
 
-    def _embed(self, inputs: List[str] | str) -> List[List[float]]:
-        """Generiert Embeddings mit dem lokalen SentenceTransformer-Modell."""
-        embeddings = self.model.encode(inputs)
-        # Konvertiere in Liste von Listen (falls nötig)
-        if isinstance(embeddings, list) and all(isinstance(x, (int, float)) for x in embeddings[0]):
-            # Einzelner Vektor
-            return [self._truncate(embeddings)]
-        # Mehrere Vektoren
-        return [self._truncate(vec) for vec in embeddings]
+    def embed_documents(self, texts: List[str]) -> List[List[float]]:
+        """Generiert Embeddings für eine Liste von Texten."""
+        passage_embeddings = self.model.encode(
+            sentences=texts,
+            task="retrieval",
+            prompt_name="passage",
+        )
+        return [self._truncate([float(x) for x in emb]) for emb in passage_embeddings]
+
+    def embed_query(self, text: str) -> List[float]:
+        """Generiert ein Embedding für einen einzelnen Text."""
+        query_embeddings = self.model.encode(
+            sentences=[text],
+            task="retrieval",
+            prompt_name="query",
+        )
+        return self._truncate([float(x) for x in query_embeddings[0]])
+
 
 
 # -----------------------------
@@ -190,4 +201,4 @@ class EmbeddingFactory:
         elif config.embedding_type == EmbeddingType.SENTENCE_TRANSFORMER:
             return SentenceTransformerEmbeddings(config=config)
         else:
-            raise ValueError(f"Unsupported embedding type: {config.embedding_type}")
\ No newline at end of file
+            raise ValueError(f"Unsupported embedding type: {config.embedding_type}")

math-tutor/backend/app/deterministic_services/vector_store.py

@@ -108,13 +108,12 @@ def load_docs(base_dir: Path) -> List[DocRecord]:
     return docs
 
 
-def embed_passages(embedder: EmbeddingLike, texts: List[str]) -> List[List[float]]:
-    prefixed = [f"passage: {t}" for t in texts]
-    return embedder.embed_documents(prefixed)
+def embed_documents(embedder: EmbeddingLike, texts: List[str]) -> List[List[float]]:
+    return embedder.embed_documents(texts)
 
 
 def embed_query(embedder: EmbeddingLike, text: str) -> List[float]:
-    return embedder.embed_query(f"query: {text}")
+    return embedder.embed_query(text)
 
 
 DDL = """
@@ -135,7 +134,7 @@ CREATE TABLE IF NOT EXISTS docs (
   path TEXT NOT NULL,
   markdown TEXT NOT NULL,
 
-  embedding VECTOR(1024) NOT NULL
+  embedding VECTOR(512) NOT NULL
 );
 
 CREATE INDEX IF NOT EXISTS docs_embedding_cos_idx

math-tutor/backend/requirements.txt

@@ -10,3 +10,9 @@ sympy
 psycopg[binary]
 pgvector
 pyyaml
+
+sentence-transformers
+transformers
+torch
+peft
+torchvision

math-tutor/backend/scripts/retrieval_cli.py

@@ -36,7 +36,7 @@ def cli_ingest(args: argparse.Namespace) -> None:
     embedder = build_embedder()
     base_dir = Path(args.base)
     docs = vector_store.load_docs(base_dir)
-    embeddings = vector_store.embed_passages(embedder, [doc.markdown for doc in docs])
+    embeddings = vector_store.embed_documents(embedder, [doc.markdown for doc in docs])
     pg_url = args.pg or os.getenv("POSTGRES_URL")
     if not pg_url:
         raise ValueError("Missing POSTGRES_URL")