Create litellm_config.yaml

litellm_config.yaml

@@ -0,0 +1,20 @@
+model_list:
+  - model_name: hf-mistral-7b-instruct-v03
+    litellm_params:
+      model: huggingface/mistralai/Mistral-7B-Instruct-v0.3
+      api_key: os.environ/HUGGINGFACE_API_KEY
+      api_base: "https://api-inference.huggingface.co/"
+      #api_base: "https://my-endpoint.endpoints.huggingface.cloud"
+  - model_name: awan-llama3-8b
+    litellm_params:
+      model: openai/Meta-Llama-3-8B-Instruct
+      api_key: os.environ/OTHER_AWANLLM_API_KEY
+      api_base: https://api.awanllm.com/v1
+      rpm: 10
+
+litellm_settings:
+  success_callback: ["langfuse"]
+  failure_callback: ["langfuse"]
+  num_retries: 3
+  request_timeout: 20
+  allowed_fails: 3