Lambent/Iris-12B-v1.3.2
LoRA-style vs Lambent/Iris-12B-v1.2. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 48 of 48 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/12, Norms 0/289, other 0/48, MLP 144/144, Attention 184/184
3 derivatives checked
- mradermacher/Iris-12B-v1.3.2-i1-GGUF: Quantization
- mradermacher/Iris-12B-v1.3.2-GGUF: Quantization
- Lambent/Iris-12B-v1.3.3: LoRA-style