feat(llm): Add retry mechanism for LLM API calls (#400)

* feat(llm): Add retry mechanism for LLM API calls

* feat: configure max_retries for LLM calls via conf.yaml

---------

Co-authored-by: Willem Jiang <willem.jiang@gmail.com>

conf.yaml.example

@@ -10,6 +10,7 @@ BASIC_MODEL:
   base_url: https://ark.cn-beijing.volces.com/api/v3
   model: "doubao-1-5-pro-32k-250115"
   api_key: xxxx
+  # max_retries: 3 # Maximum number of retries for LLM calls
   # verify_ssl: false  # Uncomment this line to disable SSL certificate verification for self-signed certificates
  
 # Reasoning model is optional.
@@ -20,6 +21,8 @@ BASIC_MODEL:
 #   base_url: https://ark-cn-beijing.bytedance.net/api/v3
 #   model: "doubao-1-5-thinking-pro-m-250428"
 #   api_key: xxxx
+#   max_retries: 3 # Maximum number of retries for LLM calls
+
 
 # OTHER SETTINGS:
 # Search engine configuration (Only supports Tavily currently)

src/llms/llm.py

@@ -70,6 +70,10 @@ def _create_llm_use_conf(
     if not merged_conf:
         raise ValueError(f"No configuration found for LLM type: {llm_type}")
 
+    # Add max_retries to handle rate limit errors
+    if "max_retries" not in merged_conf:
+        merged_conf["max_retries"] = 3
+
     if llm_type == "reasoning":
         merged_conf["api_base"] = merged_conf.pop("base_url", None)