feat: Add configurable timeout for OpenRouter embedding model

Author: rostilos

deployment/config/rag-pipeline/.env.sample

@@ -50,6 +50,7 @@ REDIS_URL=redis://redis:6379/1
 #OLLAMA_BATCH_SIZE=100
 
 ## Request timeout in seconds (increase for slow CPU)
+#OPENROUTER_TIMEOUT=300
 #OLLAMA_TIMEOUT=120
 
 ## CPU Threading Optimization (set based on your CPU cores)

python-ecosystem/rag-pipeline/src/rag_pipeline/core/embedding_factory.py

@@ -4,6 +4,7 @@
 """
 
 import logging
+import os
 from typing import Union
 
 from llama_index.core.base.embeddings.base import BaseEmbedding
@@ -37,12 +38,13 @@ def create_embedding_model(config: RAGConfig) -> BaseEmbedding:
         )
 
     elif provider == "openrouter":
-        logger.info(f"Creating OpenRouter embedding model: {config.openrouter_model}")
+        timeout = float(os.getenv("OPENROUTER_TIMEOUT", "300"))
+        logger.info(f"Creating OpenRouter embedding model: {config.openrouter_model} (timeout={timeout}s)")
         return OpenRouterEmbedding(
             api_key=config.openrouter_api_key,
             model=config.openrouter_model,
             api_base=config.openrouter_base_url,
-            timeout=60.0,
+            timeout=timeout,
             max_retries=3,
             expected_dim=config.embedding_dim
         )