adrida/tracer

Replaces 90% of LLM classification calls with traditional ML

View on index · View in 3D Map

Related repos

• run large language models on any device
• evaluate and track LLM experiments and AI agents
• Benchmarks LLMs for end-to-end machine learning workflows
• OpenTelemetry observability for AI agents and LLMs
• Multimodal LLM inference engine for mobile devices
• Optimizes LLM reasoning without model training
• Validates financial transactions with sub-100ms decisions
• Benchmark for LLMs on software engineering tasks
• Optimizes AI agents through execution traces
• Fine-tunes LLMs on Apple silicon