Reza2kn/Bina-0.1-Koochik-FP16

Partial fine-tune vs datalab-to/surya-ocr-2. Partial change: 90% of the weights changed (attention, embeddings, mlp), in 24 of 24 layers.

Tensors changed: Embeddings 1/4, Norms 0/129, Attention 138/216, MLP 98/124

1 derivatives checked

• Reza2kn/Bina-0.1-Koochik-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse