Merge branch 'main' into vector-store-chunks

llama_stack/providers/inline/vector_io/faiss/faiss.py

@@ -160,8 +160,11 @@ class FaissIndex(EmbeddingIndex):
         for d, i in zip(distances[0], indices[0], strict=False):
             if i < 0:
                 continue
+            score = 1.0 / float(d) if d != 0 else float("inf")
+            if score < score_threshold:
+                continue
             chunks.append(self.chunk_by_index[int(i)])
-            scores.append(1.0 / float(d) if d != 0 else float("inf"))
+            scores.append(score)
 
         return QueryChunksResponse(chunks=chunks, scores=scores)
 

llama_stack/providers/remote/datasetio/nvidia/README.md

@@ -20,7 +20,7 @@ This provider enables dataset management using NVIDIA's NeMo Customizer service.
 Build the NVIDIA environment:
 
 ```bash
-llama stack build --template nvidia --image-type conda
+llama stack build --template nvidia --image-type venv
 ```
 
 ### Basic Usage using the LlamaStack Python Client

llama_stack/providers/remote/inference/nvidia/NVIDIA.md

@@ -18,7 +18,7 @@ This provider enables running inference using NVIDIA NIM.
 Build the NVIDIA environment:
 
 ```bash
-llama stack build --template nvidia --image-type conda
+llama stack build --template nvidia --image-type venv
 ```
 
 ### Basic Usage using the LlamaStack Python Client

llama_stack/providers/remote/post_training/nvidia/README.md

@@ -22,7 +22,7 @@ This provider enables fine-tuning of LLMs using NVIDIA's NeMo Customizer service
 Build the NVIDIA environment:
 
 ```bash
-llama stack build --template nvidia --image-type conda
+llama stack build --template nvidia --image-type venv
 ```
 
 ### Basic Usage using the LlamaStack Python Client

llama_stack/providers/remote/safety/nvidia/README.md

@@ -19,7 +19,7 @@ This provider enables safety checks and guardrails for LLM interactions using NV
 Build the NVIDIA environment:
 
 ```bash
-llama stack build --template nvidia --image-type conda
+llama stack build --template nvidia --image-type venv
 ```
 
 ### Basic Usage using the LlamaStack Python Client

llama_stack/providers/remote/vector_io/pgvector/pgvector.py

@@ -132,8 +132,11 @@ class PGVectorIndex(EmbeddingIndex):
             chunks = []
             scores = []
             for doc, dist in results:
+                score = 1.0 / float(dist) if dist != 0 else float("inf")
+                if score < score_threshold:
+                    continue
                 chunks.append(Chunk(**doc))
-                scores.append(1.0 / float(dist) if dist != 0 else float("inf"))
+                scores.append(score)
 
             return QueryChunksResponse(chunks=chunks, scores=scores)
 

llama_stack/providers/remote/vector_io/weaviate/weaviate.py

@@ -105,8 +105,12 @@ class WeaviateIndex(EmbeddingIndex):
                 log.exception(f"Failed to parse document: {chunk_json}")
                 continue
 
+            score = 1.0 / doc.metadata.distance if doc.metadata.distance != 0 else float("inf")
+            if score < score_threshold:
+                continue
+
             chunks.append(chunk)
-            scores.append(1.0 / doc.metadata.distance if doc.metadata.distance != 0 else float("inf"))
+            scores.append(score)
 
         return QueryChunksResponse(chunks=chunks, scores=scores)
 

llama_stack/providers/utils/memory/openai_vector_store_mixin.py

@@ -444,10 +444,6 @@ class OpenAIVectorStoreMixin(ABC):
             # Convert response to OpenAI format
             data = []
             for chunk, score in zip(response.chunks, response.scores, strict=False):
-                # Apply score based filtering
-                if score < score_threshold:
-                    continue
-
                 # Apply filters if provided
                 if filters:
                     # Simple metadata filtering