In Part 2 of our Models series, I’ll show you how to go beyond “download a model” and start testing models with custom test sets—all no-code and fully on-device in Model HQ. We’ll generate a quick test set (JSON), run it against different models, compare speed + quality, and then level up to a real enterprise use case of how to use your own custom test set: root-cause / classification with a 200-row dataset using one-shot prompting—all running offline once your models are downloaded. ✅ In this video: Generate a custom test set in seconds Run the same test set across different models Review first-token time + total processing time Run a 200-row classification dataset for enterprise-style analysis Compare model responses to gold answers (and refine prompts) Learn more about Model HQ: https://llmware.ai #ModelHQ #LLMWare #OnDeviceAI #PrivateAI #LocalAI #NoCodeAI #AIAgents #RAG #SmallLanguageModels #SLM #EnterpriseAI #EdgeAI