mradermacher/TW3-JRGL-v2-GGUF
Full fine-tune vs paloalma/TW3-JRGL-v2. mradermacher/TW3-JRGL-v2-GGUF is a GGUF (Q5_K_S) of different weights. Fine-tune: 100% of the weights changed, across 80 of 80 layers.
Tensors changed: MLP 240/240, Norms 0/161, Attention 480/640, Embeddings 1/1, lm_head 1/1