kueizen/Marco-Mini-Instruct-REAP60
LoRA-style (fewer layers) vs ATH-MaaS/Marco-Mini-Instruct. kueizen/Marco-Mini-Instruct-REAP60 matches part of ATH-MaaS/Marco-Mini-Instruct (8878 of its 21758 tensors; the rest of ATH-MaaS/Marco-Mini-Instruct has no counterpart). On the shared part: 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 28 of 28 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, MLP 1106/1112, Norms 0/113, Attention 0/112