NVIDIA TAO 7 is now available

Accelerated Computing Intelligent Video Analytics TAO Toolkit
,
1

NVIDIA TAO 7 is now generally available, adding a suite of agent skills and tools for fine-tuning vision AI models with coding agents and natural language prompts.

What’s new:

  • Agent skills and tools to fine-tune open models, including Cosmos 3, using natural language with any coding agent (Codex, Claude Code, NemoClaw)

  • AutoML optimization with LLM-guided hyperparameter tuning that converges 2x faster toward accuracy targets

  • Fine-tuning for any HuggingFace CV/VLM model (full or LoRA) on local NVIDIA GPUs inside an NGC PyTorch container.

  • Data Enhanced Fine-Tuning (developer preview): autonomous workflow that diagnoses model weaknesses and selects the right data strategy to improve accuracy

Resources: