mlx-community/translategemma-27b-it-bf16
Same weights, repacked (fewer layers) vs google/translategemma-27b-it. mlx-community/translategemma-27b-it-bf16 matches part of google/translategemma-27b-it (808 of its 1247 tensors; the rest of google/translategemma-27b-it has no counterpart). On the shared part: Same weights in different files (re-saved, re-sharded or new metadata).
Tensors changed: Embeddings 0/1, Norms 0/373, MLP 0/186, Attention 0/248