allenai/Olmo-3-32B-Think-DPO

Full fine-tune vs allenai/Olmo-3-32B-Think-SFT. Fine-tune: 100% of the weights changed, across 64 of 64 layers.

Tensors changed: lm_head 1/1, Embeddings 1/1, MLP 192/192, Norms 257/257, Attention 256/256

4 derivatives checked

• allenai/Olmo-3.1-32B-Think: Full fine-tune
• allenai/Olmo-3-32B-Think: Full fine-tune
• mradermacher/Olmo-3-32B-Think-DPO-i1-GGUF: Quantization
• mradermacher/Olmo-3-32B-Think-DPO-GGUF: Quantization

On the Hub · Downloads on Model Pulse