chore: optimize CPU thread allocation and context limits for free instances

Dockerfile

@@ -11,5 +11,7 @@ RUN huggingface-cli download Qwen/Qwen2.5-1.5B-Instruct-GGUF qwen2.5-1.5b-instru
 
 EXPOSE 7860
 
-# Start the OpenAI-compatible API server
-CMD ["python3", "-m", "llama_cpp.server", "--model", "/model/qwen2.5-1.5b-instruct-q4_k_m.gguf", "--host", "0.0.0.0", "--port", "7860", "--n_ctx", "2048"]
+# Start the OpenAI-compatible API server with explicit limits for free CPU instances
+# - n_threads: 2 (matches 2 vCPU limit)
+# - n_ctx: 2048 (limits context memory usage)
+CMD ["python3", "-m", "llama_cpp.server", "--model", "/model/qwen2.5-1.5b-instruct-q4_k_m.gguf", "--host", "0.0.0.0", "--port", "7860", "--n_ctx", "2048", "--n_threads", "2", "--n_threads_batch", "2"]