mradermacher/EGM-8B-GGUF
Same weights, converted (fewer layers) vs Efficient-Large-Model/EGM-8B. mradermacher/EGM-8B-GGUF matches part of Efficient-Large-Model/EGM-8B (399 of its 750 tensors; the rest of Efficient-Large-Model/EGM-8B has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: MLP 0/108, Norms 0/145, Attention 0/144, lm_head 0/1, Embeddings 0/1