Lambent/qwen2.5-reinstruct-alternate-lumen-14B
Full fine-tune vs Qwen/Qwen2.5-14B-Instruct. Fine-tune: 100% of the weights changed, across 48 of 48 layers.
Tensors changed: Norms 17/97, MLP 144/144, Attention 310/336
3 derivatives checked
- mradermacher/qwen2.5-reinstruct-alternate-lumen-14B-i1-GGUF: Quantization
- mradermacher/qwen2.5-reinstruct-alternate-lumen-14B-GGUF: Quantization
- Lambent/arsenic-v1-qwen2.5-14B: LoRA-style