elinas/Llama-3-13B-Instruct
Partial fine-tune (fewer layers) vs meta-llama/Meta-Llama-3-8B-Instruct. meta-llama/Meta-Llama-3-8B-Instruct matches part of elinas/Llama-3-13B-Instruct (291 of its 498 tensors; the rest of elinas/Llama-3-13B-Instruct has no counterpart). On the shared part: Partial change: 60% of the weights changed (attention, mlp, norms), in 22 of 32 layers.
Tensors changed: Embeddings 0/1, Norms 44/65, MLP 66/96, Attention 88/128, lm_head 0/1
3 derivatives checked
- mradermacher/Llama-3-13B-Instruct-i1-GGUF: Quantization
- mradermacher/Llama-3-13B-Instruct-GGUF: Quantization
- elinas/Llama-3-13B-Instruct-ft: Full fine-tune