Improve models configuration

src/aichat/config.yaml

@@ -1,9 +1,9 @@
 # see https://github.com/sigoden/aichat/blob/main/config.example.yaml
 
-rag_embedding_model: ollama:nomic-embed-text
+rag_embedding_model: __EMBEDDING_MODEL__
 rag_chunk_size: 8192
 rag_chunk_overlap: 409
-model: ollama:gemma3:12b-it-qat
+model: __DEFAULT_MODEL__
 temperature: 0
 clients:
 - type: openai-compatible
@@ -11,21 +11,3 @@ clients:
   api_base: http://localhost:11434/v1
   api_key: __LLM_API_KEY__
   models:
-    - name: gemma3:4b
-      supports_function_calling: true
-      max_input_tokens: 128000
-    - name: gemma3:12b-it-qat
-      supports_function_calling: true
-      max_input_tokens: 128000
-    - name: gemma3:27b-it-qat
-      supports_function_calling: true
-      max_input_tokens: 128000
-    - name: gpt-oss:20b
-      supports_function_calling: true
-      max_input_tokens: 128000
-    - name: llama3.1:8b
-      supports_function_calling: true
-      max_input_tokens: 128000
-    - name: qwen2.5:7b
-      supports_function_calling: true
-      max_input_tokens: 128000