shenzhi-wang/Gemma-2-27B-Chinese-Chat

Full fine-tune vs google/gemma-2-27b-it. Fine-tune: 100% of the weights changed, across 46 of 46 layers.

Tensors changed: Embeddings 1/1, Norms 112/185, MLP 138/138, Attention 184/184

2 derivatives checked

• mradermacher/Gemma-2-27B-Chinese-Chat-i1-GGUF: Quantization
• mradermacher/Gemma-2-27B-Chinese-Chat-GGUF: Quantization

On the Hub · Downloads on Model Pulse