Bedhoy/Tinman-gemma4-companion-gguf
Same weights, converted (fewer layers) vs Tinman-Lab/Tinman-gemma4-companion-merged. Bedhoy/Tinman-gemma4-companion-gguf matches part of Tinman-Lab/Tinman-gemma4-companion-merged (623 of its 2076 tensors; the rest of Tinman-Lab/Tinman-gemma4-companion-merged has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → BF16/F32), plus 43 new tensors.
Tensors changed: Embeddings 0/2, Norms 0/278, MLP 0/126, other 0/85, Attention 0/132