• SLIM models: Structured Language Instruction Models from LLMWare
• Designed for AI agents and function calling tasks
• Generate structured outputs for complex automation workflows
• Multi-step, multi-process, and multi-model LLM-based
• Consists of 10 models: addition, arithmetic, comparison, concatenation, division, function-calling, multiplication, subtraction, sentiment-tool, variable assignment
• Slim-sentiment-tool: A notable example, a 4_K_M quantized GGUF version of slim-sentiment
• Specialized decoder-based LLMs, fine-tuned for function calling
• Fast and efficient inference
• Parallelizable for multi-model concurrent deployment on CPUs

demo on how it works:

all the analysis tools to use: