ai-forever/FRIDA-Decisions
LoRA-style vs ai-forever/FRIDA. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of k, o, q, v changed, in 24 of 24 layers; norms, biases and embeddings are untouched.
Tensors changed: Attention 96/97, Norms 0/49, other 0/73