forBalanced method
Configure for balanced performance and memory usage
Provides a good balance between performance and resource usage. Suitable for most general-purpose applications.
Implementation
OllamaBuilder forBalanced() {
return numGpu(20) // Partial GPU usage
.numCtx(2048) // Moderate context window
.numBatch(256) // Moderate batch size
.keepAlive("30m"); // Moderate keep-alive
}