CultriX/Qwen2.5-14B-HyperSeekv5
Full fine-tune vs CultriX/Qwen2.5-14B-Coder_r128. Fine-tune: 99% of the weights changed, across 48 of 48 layers.
Tensors changed: Norms 59/97, MLP 143/144, Attention 318/336
2 derivatives checked
- mradermacher/Qwen2.5-14B-HyperSeekv5-i1-GGUF: Quantization
- mradermacher/Qwen2.5-14B-HyperSeekv5-GGUF: Quantization