Llama-3.3-Nemotron-70B-Reward-Principle

Home/Models/nvidia/Llama-3.3-Nemotron-70B-Reward-Principle

nvidia/Llama-3.3-Nemotron-70B-Reward-Principle

kind: original
quant: bf16 (16-bit)
family: llama
modality: text-generation
root model: meta-llama/Llama-3.3-70B-Instruct
provider: nvidia
license: other
language: en
context: —
library: transformers
vRAM (GB): —
size on disk: —
downloads: 278
likes: 7
created: —
last modified: 2025-10-30T00:01:48+00:00 (2025)
huggingface: https://huggingface.co/nvidia/Llama-3.3-Nemotron-70B-Reward-Principle
tags: transformerssafetensorsllamatext-generationnvidiallama3.3conversationalendataset:nvidia/HelpSteer3arxiv:2410.16184arxiv:2509.21319arxiv:2503.04378arxiv:2505.11475arxiv:2410.01257arxiv:2310.05344arxiv:2311.09528arxiv:2406.08673base_model:meta-llama/Llama-3.3-70B-Instructbase_model:finetune:meta-llama/Llama-3.3-70B-Instructlicense:othertext-generation-inferenceregion:us

benchmark	category	score	context	conf	evidence
MATH	math	81.9		low	`\| [Llama-3.3-Nemotron-70B-Reward-Principle](https://huggingface.co/nvidia/Llama-3.3-Nemotron-70B-Reward-Principle/) \| 85.3 \| 81.9 \| 70.4 \| 96.9 \| 85.5 \| 84.9 \| **80`

none scraped.