CodeGoat24/UnifiedReward-2.0-qwen-32b
Full fine-tune vs Qwen/Qwen2.5-VL-32B-Instruct. Fine-tune: 98% of the weights changed, across 64 of 64 layers.
Tensors changed: Attention 576/576, MLP 388/388, Norms 192/194, Embeddings 1/2, lm_head 1/1
3 derivatives checked
- mradermacher/UnifiedReward-2.0-qwen-32b-i1-GGUF: Full fine-tune (fewer layers)
- mradermacher/UnifiedReward-2.0-qwen-32b-GGUF: Full fine-tune (fewer layers)
- CodeGoat24/UnifiedReward-Edit-qwen-32b: Full fine-tune