suv11235/olmOCR-7B-grpo-v6
Partial fine-tune vs allenai/olmOCR-2-7B-1025. Partial change: 87% of the weights changed (attention, embeddings, mlp, norms), in 32 of 32 layers.
Tensors changed: Embeddings 1/2, Norms 39/122, MLP 264/280, Attention 299/324, lm_head 0/1
1 derivatives checked
- mradermacher/olmOCR-7B-grpo-v6-GGUF: Same weights, converted (fewer layers)