Create config.yaml

config.yaml

@@ -0,0 +1,52 @@
+# LocalAI Configuration for fast lightweight models
+name: tinyllama
+backend: llama
+parameters:
+  # Model file - will be downloaded automatically
+  model: tinyllama-1.1b-chat-v1.0.Q4_K_M.gguf
+  
+  # Performance settings
+  threads: 4
+  context_size: 2048
+  batch: 512
+  
+  # Speed optimizations
+  f16: true
+  low_vram: true
+  
+  # Generation settings for faster responses
+  temperature: 0.1
+  top_p: 0.9
+  top_k: 10
+  repeat_penalty: 1.1
+  
+  # Template for chat format
+  template:
+    chat: |
+      <|system|>
+      You are a helpful assistant.</s>
+      <|user|>
+      {{.Input}}</s>
+      <|assistant|>
+    completion: |
+      {{.Input}}
+
+---
+name: qwen2-0.5b
+backend: llama
+parameters:
+  model: qwen2-0_5b-instruct-q4_k_m.gguf
+  threads: 4
+  context_size: 1024
+  batch: 256
+  f16: true
+  low_vram: true
+  temperature: 0.1
+  top_p: 0.9
+  template:
+    chat: |
+      <|im_start|>system
+      You are a helpful assistant.<|im_end|>
+      <|im_start|>user
+      {{.Input}}<|im_end|>
+      <|im_start|>assistant