cerebras/GLM-4.5-Air-REAP-82B-A12B
LoRA-style vs zai-org/GLM-4.5-Air. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, up_proj changed, in 45 of 47 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/2, Norms 0/98, MLP 1687/1887, Attention 0/329, lm_head 0/1, other 0/2
4 derivatives checked
- bartowski/cerebras_GLM-4.5-Air-REAP-82B-A12B-GGUF: Quantization
- mradermacher/GLM-4.5-Air-REAP-82B-A12B-i1-GGUF: Quantization
- mradermacher/GLM-4.5-Air-REAP-82B-A12B-GGUF: Quantization
- DevQuasar/cerebras.GLM-4.5-Air-REAP-82B-A12B-GGUF: Same weights, converted