JaniShubh/gemma2b_FT
Full fine-tune vs google/gemma-2b. JaniShubh/gemma2b_FT is a GGUF (Q4_K_M) of different weights. Fine-tune: 100% of the weights changed, across 18 of 18 layers.
Tensors changed: Embeddings 1/1, Norms 17/37, MLP 54/54, Attention 72/72