TIGER-Lab/One-Shot-CFT-Math-Qwen-7B
Full fine-tune vs Qwen/Qwen2.5-Math-7B. Fine-tune: 93% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 0/1, Norms 53/57, MLP 84/84, Attention 196/196, lm_head 1/1
2 derivatives checked
- mradermacher/One-Shot-CFT-Math-Qwen-7B-i1-GGUF: Quantization
- mradermacher/One-Shot-CFT-Math-Qwen-7B-GGUF: Same weights, converted
1 other repos hold byte-for-byte identical weight files.