True2456/Qwen3.8-27B-ANE
Partial fine-tune (fewer layers) vs Qwen/Qwen3.8-27B. True2456/Qwen3.8-27B-ANE matches part of Qwen/Qwen3.8-27B (355 of its 1199 tensors; the rest of Qwen/Qwen3.8-27B has no counterpart). On the shared part: Partial change: 0% of the weights changed (norms), in 64 of 64 layers.
Tensors changed: Norms 161/209, Attention 0/144, Embeddings 0/1, lm_head 0/1