nbeerbower/Mahou-1.5-Qwen2.5-1.5B-E4
LoRA-style vs nbeerbower/Dumpling-Qwen2.5-1.5B-v2. 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 28 of 28 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/57, MLP 84/84, Attention 112/196
2 derivatives checked
- mradermacher/Mahou-1.5-Qwen2.5-1.5B-E4-i1-GGUF: Quantization
- mradermacher/Mahou-1.5-Qwen2.5-1.5B-E4-GGUF: Same weights, converted