ulises-c/SocratTeachLLM
LoRA-style vs zai-org/glm-4-9b-chat. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of dense, dense_4h_to_h, dense_h_to_4h, query_key_value changed, in 40 of 40 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/81, MLP 80/80, Attention 80/120, other 0/2
1 other repos hold byte-for-byte identical weight files.