HsuanLLM/BreezeDolphin-SLERP-0.1

Full fine-tune vs MediaTek-Research/Breeze-7B-Instruct-v1_0. Fine-tune: 97% of the weights changed, across 32 of 32 layers.

Tensors changed: Norms 20/65, MLP 93/96, Attention 124/128

1 derivatives checked

• mradermacher/BreezeDolphin-SLERP-0.1-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse