omarabb315/OCR__merged_nanonets_3b

Full fine-tune vs nanonets/Nanonets-OCR-s. Fine-tune: 100% of the weights changed, across 36 of 36 layers.

Tensors changed: Embeddings 2/2, Norms 74/138, MLP 273/304, Attention 369/380

1 derivatives checked

• mradermacher/OCR__merged_nanonets_3b-GGUF: Same weights, converted (fewer layers)

On the Hub · Downloads on Model Pulse