yevvonlim/Llada-8B-Instruct-Kor

Full fine-tune vs GSAI-ML/LLaDA-8B-Instruct. Fine-tune: 100% of the weights changed, across 32 of 32 layers.

Tensors changed: Attention 128/128, Norms 0/65, Embeddings 1/1, other 65/65, MLP 32/32

2 derivatives checked

• mradermacher/Llada-8B-Instruct-Kor-i1-GGUF: Quantization
• mradermacher/Llada-8B-Instruct-Kor-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse