prithivMLmods/Taurus-Opus-7B
Different architecture vs Qwen/Qwen2.5-7B-Instruct. Different architecture: only 0 of 339/255 tensors match by name and shape.
3 derivatives checked
- mradermacher/Taurus-Opus-7B-i1-GGUF: Quantization
- mradermacher/Taurus-Opus-7B-GGUF: Same weights, converted
- Sakalti/Centauri-7B: Partial fine-tune (fewer layers)