Libraries & frameworks
labs-molt
nvidia-nemo/labs-molt
An agentic-first PyTorch-native RL training framework for multimodal, multi-turn research at frontier MoE scale.
Topics
Related projects
hands-on-modern-rl4.9kAn open hands-on curriculum covering classical reinforcement learning through LLM alignment, RLVR, and agentic systems with runnable experiments.
heretic40.4kA fully automatic command-line tool that removes safety alignment from transformers via optimized directional ablation with minimal quality loss.
toon25.6kA compact, token-efficient serialization of JSON for LLM prompts with TypeScript SDK, CLI, and lossless round-trips.
pandas-ai23.9kA Python library for conversational data analysis, letting users chat with SQL, CSV, and dataframes to query and chart.
peft21.9kA Python library for parameter-efficient fine-tuning like LoRA, adapting large pretrained models by training only small adapters.
ms-swift16.1kA full-pipeline framework for fine-tuning, aligning, evaluating, and deploying hundreds of LLMs and multimodal models.