prithivMLmods/Logics-Qwen3-Math-4B
Full fine-tune vs Qwen/Qwen3-4B-Thinking-2507. Fine-tune: 100% of the weights changed, across 36 of 36 layers.
Tensors changed: Norms 141/145, MLP 108/108, Attention 144/144
2 derivatives checked
- mradermacher/Logics-Qwen3-Math-4B-GGUF: Same weights, converted
- mradermacher/Logics-Qwen3-Math-4B-i1-GGUF: Quantization