Lythri/Lythri-7B-A4B
Full fine-tune (fewer layers) vs google/gemma-4-E4B. Lythri/Lythri-7B-A4B matches part of google/gemma-4-E4B (665 of its 2130 tensors; the rest of google/gemma-4-E4B has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 42 of 42 layers.
Tensors changed: Embeddings 2/2, Norms 130/278, other 85/127, MLP 126/126, Attention 132/132
4 derivatives checked
- mradermacher/Lythri-7B-A4B-i1-GGUF: Quantization
- mradermacher/Lythri-7B-A4B-GGUF: Same weights, converted
- Lythri/Lythri-7B-A4B-GGUF: Quantization
- Lythri/Lythri-7B-A4B-Q4_0_8-GGUF: Quantization