CodeGoat24/UnifiedReward-Edit-qwen35-4b
Partial fine-tune vs CodeGoat24/UnifiedReward-2.0-qwen35-4b. Partial change: 88% of the weights changed (attention, embeddings, lm_head, mlp, norms), in 32 of 32 layers.
Tensors changed: lm_head 1/1, Embeddings 3/4, Norms 147/203, Attention 294/320, MLP 196/196
2 derivatives checked
- mradermacher/UnifiedReward-Edit-qwen35-4b-i1-GGUF: Quantization
- mradermacher/UnifiedReward-Edit-qwen35-4b-GGUF: Same weights, converted