mazz68/VisionPsy-Nano-460M

Full fine-tune vs lusxvr/nanoVLM-460M-8k. Fine-tune: 100% of the weights changed, across 32 of 32 layers.

Tensors changed: other 3/3, Attention 176/176, MLP 80/80, Norms 115/115, Embeddings 3/3

3 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse