nbeerbower/Dumpling-Qwen2.5-32B-v2
LoRA-style vs nbeerbower/Rombos-EVAGutenberg-TIES-Qwen2.5-32B. 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 64 of 64 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/129, MLP 191/192, Attention 256/448
3 derivatives checked
- bartowski/nbeerbower_Dumpling-Qwen2.5-32B-v2-GGUF: Quantization
- mradermacher/Dumpling-Qwen2.5-32B-v2-i1-GGUF: Quantization
- mradermacher/Dumpling-Qwen2.5-32B-v2-GGUF: Quantization