Note RL data used in Nemo-3-nano (30B-A3B). With difficulty score. Paper found curr helps!

Note aggregates high-quality agent trajectories from various environments including web browsing, code generation, household tasks, knowledge base querying, and software engineering. The dataset is collected through methods described in Agent Data Protocol.

Note SFT bootstrapping: covers different capabilities, such as text editing, creative writing, coding, reading comprehension, etc