intervitens-archive/internlm2-limarp-chat-20b-GGUF
Full fine-tune vs intervitens/internlm2-base-20b-llama. intervitens-archive/internlm2-limarp-chat-20b-GGUF is a GGUF (Q8_0) of different weights. Fine-tune: 100% of the weights changed, across 48 of 48 layers.
Tensors changed: Norms 97/97, MLP 144/144, Attention 192/192, lm_head 1/1, Embeddings 1/1