your own custom SLMs hosted locally (QA, classification, function calling, ...)

Using the distil labs platform you can create your own small language model overnight. All you need is a prompt and a few dozen examples, the platform handles the rest (synthetic data generation, model fine-tuning, model evaluation). Models provided by distil labs are task-specific and despite their small size, then can achieve comparative performance to LLMs. At the end of the training process you can download the model and deploy it locally, leaving you with the ultimate control and agency over your data.