Llama-3_3-Nemotron-Super-49B-v1
Llama-3_3-Nemotron-Super-49B-v1
Llama-3.3-Nemotron-Super-49B-v1 is a NVIDIA model derived from Llama-3.1-70B through Neural Architecture Search (NAS)-based pruning and knowledge distillation, resulting in a 49B model with strong reasoning capabilities. It uses the DeciLMForCausalLM architecture.
Use this page as a checkpoint and architecture reference. Set up NeMo AutoModel with the latest container or follow the installation instructions.