hfl/llama-3-chinese-8b-instruct

Full fine-tune vs hfl/llama-3-chinese-8b. Fine-tune: 93% of the weights changed, across 32 of 32 layers.

Tensors changed: Norms 0/65, MLP 96/96, Attention 128/128, Embeddings 0/1, lm_head 1/1

2 derivatives checked

• featherless-ai-quants/hfl-llama-3-chinese-8b-instruct-GGUF: Quantization
• tensorblock/llama-3-chinese-8b-instruct-GGUF: Quantization

On the Hub · Downloads on Model Pulse