PleasedPenguin/A2R-30B-A3B
LoRA-style vs Qwen/Qwen3-Omni-30B-A3B-Instruct. 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: other 0/148, MLP 774/1557, Norms 0/562, Attention 192/696, Embeddings 0/21, lm_head 0/16
2 derivatives checked
- mradermacher/A2R-30B-A3B-i1-GGUF: Quantization
- mradermacher/A2R-30B-A3B-GGUF: Quantization