mradermacher/Gemma4-E4B-8B-Heretic-Ultra-GGUF
Same weights, converted (fewer layers) vs dopaemon/Gemma4-E4B-8B-Heretic-Ultra. mradermacher/Gemma4-E4B-8B-Heretic-Ultra-GGUF matches part of dopaemon/Gemma4-E4B-8B-Heretic-Ultra (677 of its 2130 tensors; the rest of dopaemon/Gemma4-E4B-8B-Heretic-Ultra has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/BF16), plus 43 new tensors.
Tensors changed: Embeddings 0/2, Norms 0/296, MLP 0/126, other 0/85, Attention 0/168