allenai/Olmo-3-7B-Think
Full fine-tune vs allenai/Olmo-3-7B-Think-DPO. Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Embeddings 1/1, MLP 96/96, Norms 128/129, Attention 128/128, lm_head 1/1
17 derivatives checked
- unsloth/Olmo-3-7B-Think-GGUF: Same weights, converted
- mradermacher/Olmo-3-7B-Think-i1-GGUF: Quantization
- bartowski/allenai_Olmo-3-7B-Think-GGUF: Same weights, converted
- Tuwhy/Olmo-3-7B-Think-OPSA: Full fine-tune
- eaddario/Olmo-3-7B-Think-GGUF: Same weights, converted
- richardyoung/Olmo-3-7B-Think-heretic: Abliterated
- lmstudio-community/Olmo-3-7B-Think-GGUF: Quantization
- unsloth/Olmo-3-7B-Think: Identical copy
- jkim96/Olmo-3-7B-Think-DASHQ-Q2-GGUF: Quantization
- DevQuasar/allenai.Olmo-3-7B-Think-GGUF: Same weights, converted
- mradermacher/Olmo-3-7B-Think-GGUF: Same weights, converted
- MaziyarPanahi/Olmo-3-7B-Think-GGUF: Same weights, converted
- unsloth/Olmo-3-7B-Think-unsloth-bnb-4bit: Quantization
- Samll/olmo3-7b-think-thinkprune-b4000: LoRA-style
- kaitchup/Olmo-3-7B-Think-awq-w4a16-asym: Quantization
- lmstudio-community/Olmo-3-7B-Think-MLX-8bit: Quantization
- AIML-TUDA/Olmo-3.1-7B-Think: Full fine-tune
5 other repos hold byte-for-byte identical weight files.