Add OLLAMA_BASE_URL support for chat and embeddings

src/chat_service.py

@@ -13,6 +13,7 @@ from src.ingest import get_or_build_vectorstore
 
 MAX_MEMORY_TURNS = int(os.getenv("RAG_MEMORY_TURNS", "6"))
 LLM_MODEL = os.getenv("LLM_MODEL", "hf.co/LiquidAI/LFM2-1.2B-RAG-GGUF:Q5_K_M")
+OLLAMA_BASE_URL = os.getenv("OLLAMA_BASE_URL", "http://localhost:11434")
 
 
 @dataclass
@@ -59,7 +60,11 @@ class RagChatService:
 
     def _get_llm(self) -> ChatOllama:
         if self._llm is None:
-            self._llm = ChatOllama(model=LLM_MODEL, temperature=0.2)
+            self._llm = ChatOllama(
+                model=LLM_MODEL,
+                temperature=0.2,
+                base_url=OLLAMA_BASE_URL,
+            )
         return self._llm
 
     def _format_context(self, question: str) -> str:

src/ingest.py

@@ -23,6 +23,7 @@ FAQ_PATH = _RAG_DIR / "faq.json"
 CHROMA_DIR = _RAG_DIR / "chroma_db"
 
 EMBED_MODEL = os.getenv("EMBED_MODEL", "embeddinggemma:latest")
+OLLAMA_BASE_URL = os.getenv("OLLAMA_BASE_URL", "http://localhost:11434")
 COLLECTION_NAME = "naijalingo_faq"
 
 
@@ -50,7 +51,7 @@ def build_vectorstore(
     embed_model: str = EMBED_MODEL,
 ) -> Chroma:
     docs = load_faq_documents(faq_path)
-    embeddings = OllamaEmbeddings(model=embed_model)
+    embeddings = OllamaEmbeddings(model=embed_model, base_url=OLLAMA_BASE_URL)
 
     chroma_dir.mkdir(parents=True, exist_ok=True)
     vectorstore = Chroma.from_documents(
@@ -67,7 +68,7 @@ def load_vectorstore(
     chroma_dir: Path = CHROMA_DIR,
     embed_model: str = EMBED_MODEL,
 ) -> Chroma:
-    embeddings = OllamaEmbeddings(model=embed_model)
+    embeddings = OllamaEmbeddings(model=embed_model, base_url=OLLAMA_BASE_URL)
     return Chroma(
         collection_name=COLLECTION_NAME,
         embedding_function=embeddings,