qihoo360/Light-R1-14B-DS
Full fine-tune vs deepseek-ai/DeepSeek-R1-Distill-Qwen-14B. Fine-tune: 95% of the weights changed, across 48 of 48 layers.
Tensors changed: Embeddings 0/1, Attention 336/336, MLP 144/144, Norms 94/97, lm_head 1/1
4 derivatives checked
- mradermacher/Light-R1-14B-DS-i1-GGUF: Quantization
- mradermacher/Light-R1-14B-DS-GGUF: Quantization
- qihoo360/Light-R1-14B-DS-GGUF: Quantization
- EventHorizon-AI/Light-R1-14B-DS-abliterated: Abliterated
1 other repos hold byte-for-byte identical weight files.