nicoboss/DeepSeek-V2-Lite-Chat-Uncensored-Unbiased-Reasoner
LoRA-style vs nicoboss/DeepSeek-V2-Lite-Chat-Uncensored-Unbiased. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, kv_a_proj_with_mqa, kv_b_proj, o_proj, q_proj, up_proj changed, in 27 of 27 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/82, MLP 2782/2808, Attention 108/108, lm_head 0/1