The thuml/bytesized32-world-model-sft galaxy

2 models are built on thuml/bytesized32-world-model-sft: 0 quantized, 2 fine-tuned, 0 adapters and 0 merges. Together they were downloaded 108 times in the last 30 days.

Most downloaded direct derivatives

• thuml/bytesized32-world-model-rlvr-binary-reward (finetune), 40 downloads in 30 days
• thuml/bytesized32-world-model-rlvr-task-specific-reward (finetune), 34 downloads in 30 days