Cactus-Compute/gemma-4-e2b-it-hybrid
Same weights, repacked (fewer layers) vs google/gemma-4-E2B-it. Cactus-Compute/gemma-4-e2b-it-hybrid matches part of google/gemma-4-E2B-it (600 of its 2011 tensors; the rest of google/gemma-4-E2B-it has no counterpart). On the shared part: Same weights in different files (re-saved, re-sharded or new metadata), plus 10 new tensors.
Tensors changed: Embeddings 0/2, Norms 0/247, other 0/106, MLP 0/105, Attention 0/140
2 derivatives checked
- Cactus-Compute/gemma-4-e2b-it-hybrid-GGUF: Same weights, converted
- Cactus-Compute/gemma-4-e2b-it-hybrid-mlx: Quantization