A comprehensive dataset of 101,926 training examples from 5 AgentGym environments, formatted for supervised fine-tuning (SFT) of language models on interactive tasks.

Dataset Details

Total Samples: 101,926 training examples Environments: 5 AgentGym environments with balanced representation Format: ChatML-style messages with environment labels File Size: 459MB

Environments Included

AlfWorld (household tasks) BabyAI (grid navigation)… See the full description on the dataset page: https://huggingface.co/datasets/bunnybhaiya/agentgym-sft-trajectories.