Delta-Vector/Austral-GLM4-SFT
LoRA-style vs Delta-Vector/GLM-4-32B-Tulu-Instruct. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, k_proj, o_proj, q_proj, v_proj changed, in 61 of 61 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/245, MLP 61/122, Attention 244/244, lm_head 0/1
3 derivatives checked
- mradermacher/Austral-GLM4-SFT-i1-GGUF: Quantization
- mradermacher/Austral-GLM4-SFT-GGUF: Quantization
- Delta-Vector/Austral-GLM4-KTO: LoRA-style