Bizuayeu/GLM-5.3-Flash-NVFP4-attn-lmhead-W4A16
Same weights, repacked vs nvidia/GLM-5.3-Flash-NVFP4. Same weights in different files (re-saved, re-sharded or new metadata), plus 746 new tensors.
Tensors changed: MLP 0/1283, Norms 0/277, Attention 0/455, other 0/4, Embeddings 0/3