mradermacher/gemma-4-e4b-cymraeg-v4-GGUF

Same weights, converted (fewer layers) vs EryriLabs/gemma-4-e4b-cymraeg-v4. mradermacher/gemma-4-e4b-cymraeg-v4-GGUF matches part of EryriLabs/gemma-4-e4b-cymraeg-v4 (624 of its 2077 tensors; the rest of EryriLabs/gemma-4-e4b-cymraeg-v4 has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/BF16), plus 43 new tensors.

Tensors changed: Embeddings 0/2, lm_head 0/1, Norms 0/278, MLP 0/126, other 0/85, Attention 0/132

On the Hub · Downloads on Model Pulse