feat: add debug route for embedding experiments

Author: winstxnhdw

server/api/debug/embedder.py

@@ -0,0 +1,39 @@
+from typing import Annotated
+
+from litestar import Controller, post
+from litestar.di import Provide
+from litestar.params import Dependency
+from litestar.status_codes import HTTP_200_OK
+
+from server.dependencies import embedder_model
+from server.features.embeddings import Embedder
+from server.schemas.v1 import Embedding
+
+
+class EmbeddingController(Controller):
+    """
+    Summary
+    -------
+    Litestar controller for embedding-related debug endpoints
+    """
+
+    path = '/embedding'
+    dependencies = {'embedder': Provide(embedder_model)}
+
+    @post(status_code=HTTP_200_OK, sync_to_thread=True)
+    def generate(
+        self,
+        embedder: Annotated[Embedder, Dependency()],
+        data: Embedding,
+        encode_query: bool = False,
+    ) -> float:
+        """
+        Summary
+        -------
+        an endpoint for generating text directly from the LLM model
+        """
+        instruction_embedding = (
+            embedder.encode_query(data.instruction) if encode_query else embedder.encode_normalise(data.instruction)
+        )
+
+        return float(instruction_embedding @ embedder.encode_normalise(data.text).T)

server/schemas/v1/__init__.py

@@ -1,4 +1,5 @@
 from server.schemas.v1.benchmark import Benchmark as Benchmark
 from server.schemas.v1.chat import Chat as Chat
+from server.schemas.v1.embedding import Embedding as Embedding
 from server.schemas.v1.files import Files as Files
 from server.schemas.v1.query import Query as Query

server/schemas/v1/embedding.py

@@ -0,0 +1,17 @@
+from msgspec import Struct
+
+
+class Embedding(Struct):
+    """
+    Summary
+    -------
+    the response schema on embedding
+
+    Attributes
+    ----------
+    instruction (str) : the instruction
+    text (str) : the text to compare similarity
+    """
+
+    instruction: str
+    text: str