leonMW/DeepSeek-R1-Distill-Qwen-14B-GSPO-Basic

Partial fine-tune vs deepseek-ai/DeepSeek-R1-Distill-Qwen-14B. Partial change: 89% of the weights changed (attention, mlp, norms), in 48 of 48 layers.

Tensors changed: Embeddings 0/1, Attention 336/336, MLP 143/144, Norms 92/97, lm_head 0/1

2 derivatives checked

• mradermacher/DeepSeek-R1-Distill-Qwen-14B-GSPO-Basic-i1-GGUF: Full fine-tune
• mradermacher/DeepSeek-R1-Distill-Qwen-14B-GSPO-Basic-GGUF: Full fine-tune

On the Hub · Downloads on Model Pulse