hunarbatra/SpatialThinker-30B

Partial fine-tune vs Qwen/Qwen3-VL-30B-A3B-Instruct. Partial change: 78% of the weights changed (attention, lm_head, mlp), in 48 of 48 layers.

Tensors changed: Embeddings 0/4, Norms 0/309, MLP 70/268, Attention 185/300, lm_head 1/1

2 derivatives checked