mradermacher/Gemma-3-4B-it-RP-Test-GGUF
Same weights, converted (fewer layers) vs jeiku/Gemma-3-4B-it-RP-Test. mradermacher/Gemma-3-4B-it-RP-Test-GGUF matches part of jeiku/Gemma-3-4B-it-RP-Test (444 of its 883 tensors; the rest of jeiku/Gemma-3-4B-it-RP-Test 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