mradermacher/gemma4-tamil-e4b-it-GGUF
Same weights, converted (fewer layers) vs prasadvittaldev/gemma4-tamil-e4b-it. mradermacher/gemma4-tamil-e4b-it-GGUF matches part of prasadvittaldev/gemma4-tamil-e4b-it (677 of its 2130 tensors; the rest of prasadvittaldev/gemma4-tamil-e4b-it 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, Norms 0/296, MLP 0/126, other 0/85, Attention 0/168