win10/Llama-3-Taiwan-13.3B-Instruct
Partial fine-tune (fewer layers) vs yentinglin/Llama-3-Taiwan-8B-Instruct. yentinglin/Llama-3-Taiwan-8B-Instruct matches part of win10/Llama-3-Taiwan-13.3B-Instruct (291 of its 507 tensors; the rest of win10/Llama-3-Taiwan-13.3B-Instruct has no counterpart). On the shared part: Partial change: 65% of the weights changed (attention, mlp, norms), in 24 of 32 layers.
Tensors changed: Embeddings 0/1, Norms 48/65, MLP 72/96, Attention 96/128, lm_head 0/1
2 derivatives checked
- mradermacher/Llama-3-Taiwan-13.3B-Instruct-i1-GGUF: Quantization
- mradermacher/Llama-3-Taiwan-13.3B-Instruct-GGUF: Quantization