Lambent/Zora-9B-v1
Partial fine-tune vs Lambent/Zora-9B-v0. Partial change: 57% of the weights changed (attention, embeddings, mlp, norms), in 32 of 32 layers.
Tensors changed: lm_head 0/1, Embeddings 1/4, Norms 2/215, Attention 32/332, MLP 100/208
3 derivatives checked
- mradermacher/Zora-9B-v1-GGUF: Same weights, converted
- mradermacher/Zora-9B-v1-i1-GGUF: Quantization
- Lambent/Zora-9B-v2: Full fine-tune