mradermacher/medgemma-4b-it-heretic-GGUF
Same weights, converted (fewer layers) vs parziva1/medgemma-4b-it-heretic. mradermacher/medgemma-4b-it-heretic-GGUF matches part of parziva1/medgemma-4b-it-heretic (444 of its 883 tensors; the rest of parziva1/medgemma-4b-it-heretic has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/205, MLP 0/102, Attention 0/136