GR00T-N1.5-3B fully fine-tuned on the concatenated so101_grasp_1 (210 pick and place episodes across 15+ objects) and so101_grasp_2 (780 pick and place episodes across 15+ objects) datasets.

Fine-tuned for ~30k steps, which took around 24 hrs on an RTX Ada 6000 on RunPod.

Downloads last month: -

Safetensors

Model size

3B params

Tensor type

F32

BF16

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for 5hadytru/so101_GR00T-N1.5-3B_v3

Base model

nvidia/GR00T-N1.5-3B

Finetuned

(33)

this model

5hadytru
/

so101_GR00T-N1.5-3B_v3

Model tree for 5hadytru/so101_GR00T-N1.5-3B_v3

Datasets used to train 5hadytru/so101_GR00T-N1.5-3B_v3