CodeGoat24/UnifiedReward-Think-qwen35-9b
Partial fine-tune vs CodeGoat24/UnifiedReward-2.0-qwen35-9b. Partial change: 89% 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 167/215, Attention 302/332, MLP 206/208
3 derivatives checked
- mradermacher/UnifiedReward-Think-qwen35-9b-i1-GGUF: Quantization
- mradermacher/UnifiedReward-Think-qwen35-9b-GGUF: Same weights, converted
- CodeGoat24/UnifiedReward-Flex-qwen35-9b: Partial fine-tune