mradermacher/ThinkMix-Gemma3-4B-GGUF
Same weights, converted (fewer layers) vs theprint/ThinkMix-Gemma3-4B. mradermacher/ThinkMix-Gemma3-4B-GGUF matches part of theprint/ThinkMix-Gemma3-4B (444 of its 883 tensors; the rest of theprint/ThinkMix-Gemma3-4B 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