rawsh/mirrorqwen2.5-0.5b-ORPO-1
Partial fine-tune vs rawsh/mirrorqwen2.5-0.5b-SFT. Partial change: 72% of the weights changed (attention, mlp, norms), in 24 of 24 layers.
Tensors changed: Embeddings 0/1, Norms 12/49, MLP 72/72, Attention 163/168
2 derivatives checked
- mradermacher/mirrorqwen2.5-0.5b-ORPO-1-GGUF: Same weights, converted
- rawsh/mirrorqwen2.5-0.5b-ORPO-2: Partial fine-tune