sh2orc/Llama-3.1-Korean-8B-Instruct
Full fine-tune vs meta-llama/Llama-3.1-8B-Instruct. Fine-tune: 93% of the weights changed, across 32 of 32 layers.
Tensors changed: Embeddings 0/1, Norms 12/65, MLP 96/96, Attention 128/128, lm_head 1/1
6 derivatives checked
- featherless-ai-quants/sh2orc-Llama-3.1-Korean-8B-Instruct-GGUF: Quantization
- mradermacher/Llama-3.1-Korean-8B-Instruct-GGUF: Full fine-tune
- tensorblock/Llama-3.1-Korean-8B-Instruct-GGUF: Quantization
- BCCard/Llama-3.1-Kor-BCCard-Finance-8B: Full fine-tune
- lemon-mint/LLaMa-3.1-Korean-Reasoning-8B-Instruct: Full fine-tune
- fanchul/aris_llm2: LoRA-style
2 other repos hold byte-for-byte identical weight files.