OLAIR/ko-r1-14b-v2.0.3

Full fine-tune vs deepseek-ai/DeepSeek-R1-Distill-Qwen-14B. Fine-tune: 100% of the weights changed, across 48 of 48 layers.

Tensors changed: Attention 336/336, MLP 144/144, Norms 96/97

2 derivatives checked

• mradermacher/ko-r1-14b-v2.0.3-i1-GGUF: Quantization
• mradermacher/ko-r1-14b-v2.0.3-GGUF: Quantization

On the Hub · Downloads on Model Pulse