AI-MO/Kimina-Prover-Distill-1.7B
Full fine-tune vs Qwen/Qwen3-1.7B. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 109/113, MLP 84/84, Attention 112/112, lm_head 1/1
7 derivatives checked
- mradermacher/Kimina-Prover-Distill-1.7B-i1-GGUF: Quantization
- prithivMLmods/Kimina-Prover-Distill-1.7B-F32-GGUF: Same weights, converted
- gabriellarson/Kimina-Prover-Distill-1.7B-GGUF: Quantization
- DevQuasar/AI-MO.Kimina-Prover-Distill-1.7B-GGUF: Same weights, converted
- anonymous-submission-ICLR2027/Gobble-Prover-1.7B: LoRA-style
- mradermacher/Kimina-Prover-Distill-1.7B-GGUF: Same weights, converted
- AI-MO/Kimina-Prover-RL-1.7B: Partial fine-tune