qhz991029/spaceom-3b
Full fine-tune vs UCSC-VLAA/VLAA-Thinker-Qwen2.5VL-3B. Fine-tune: 100% of the weights changed, across 36 of 36 layers.
Tensors changed: Embeddings 2/2, Norms 74/138, MLP 274/304, Attention 379/380
1 derivatives checked
- mradermacher/spaceom-3b-GGUF: Same weights, converted (fewer layers)