mradermacher/olmOCR-7B-grpo-v3-GGUF

Partial fine-tune (fewer layers) vs suv11235/olmOCR-7B-grpo-v3. mradermacher/olmOCR-7B-grpo-v3-GGUF matches part of suv11235/olmOCR-7B-grpo-v3 (339 of its 729 tensors; the rest of suv11235/olmOCR-7B-grpo-v3 has no counterpart). On the shared part: mradermacher/olmOCR-7B-grpo-v3-GGUF is a GGUF (F16) of different weights. Partial change: 86% of the weights changed (attention, mlp, norms), in 28 of 28 layers.

Tensors changed: Embeddings 0/1, Norms 37/57, MLP 84/84, Attention 166/196, lm_head 0/1

On the Hub · Downloads on Model Pulse