xx18/Composition-RL-14B
Different architecture vs Qwen/Qwen3-8B. Different architecture: only 72 of 399/443 tensors match by name and shape.
2 derivatives checked
- mradermacher/Composition-RL-14B-i1-GGUF: Quantization
- mradermacher/Composition-RL-14B-GGUF: Quantization