nvidia/NVIDIA-Nemotron-Nano-12B-v2
Full fine-tune vs nvidia/NVIDIA-Nemotron-Nano-12B-v2-Base. Fine-tune: 100% of the weights changed, across 62 of 62 layers.
Tensors changed: Embeddings 1/1, other 196/196, Norms 91/91, MLP 56/56, Attention 24/24, lm_head 1/1
9 derivatives checked
- nvidia/NVIDIA-Nemotron-Nano-9B-v2: Different architecture
- MaziyarPanahi/NVIDIA-Nemotron-Nano-12B-v2-GGUF: Same weights, converted
- bartowski/nvidia_NVIDIA-Nemotron-Nano-12B-v2-GGUF: Same weights, converted
- gabriellarson/NVIDIA-Nemotron-Nano-12B-v2-GGUF: Same weights, converted
- DevQuasar/nvidia.NVIDIA-Nemotron-Nano-12B-v2-GGUF: Same weights, converted
- cyankiwi/NVIDIA-Nemotron-Nano-12B-v2-AWQ-4bit: Quantization
- cyankiwi/NVIDIA-Nemotron-Nano-12B-v2-AWQ-8bit: Quantization
- unsloth/NVIDIA-Nemotron-Nano-9B-v2: Different architecture
- cpagac/Nemotron-Nano-9B-v2-heretic: Different architecture