---
title: nvidia/Flora-Rewarder
url: https://recentmodel.com/models/nvidia/Flora-Rewarder/
status: main board
published: 2026-10-02T18:01:10Z
updated: 2026-10-20T23:23:28Z
source: https://huggingface.co/nvidia/Flora-Rewarder
---

# nvidia/Flora-Rewarder

Training labels are a combination of human-written annotations, automatic success signals recorded with the data, and VLM-generated annotations. Dense, 28B params. Context: 256K tokens (config). License: nvidia-license.


- **Org:** nvidia
- **Released:** 2026-10-02
- **Status:** main board
- **License:** nvidia-license (restricted)
- **Parameters:** 27.8B
- **Architecture:** dense · Qwen3_5ForConditionalGeneration
- **Context:** 262,144 tokens
- **Task:** video-text-to-text · vision
- **Signal:** 17 HF likes · 1024 downloads

## Runs on

estimate from params; sizing calculation, not tested.

- Q4 ~19.8 GB (estimated): 24GB GPU (3090 / 4090), 32GB GPU (5090), 24GB GPU (7900 XTX), Mac 36GB, CPU only, 32GB RAM (slow)
- Q8 ~32.6 GB (estimated): 48GB GPU (RTX 6000 Ada / A6000), Mac 48GB, Mac 64GB (M1 Max), CPU only, 64GB RAM (slow)
- FP16 ~58.7 GB (estimated): 96GB GPU (RTX Pro 6000 Blackwell), 1× DGX Spark (128GB), 2× DGX Spark (256GB), Strix Halo 128GB unified (~96GB usable), Mac 96GB, Mac 128GB, Mac 192GB, Mac 256GB, Mac 512GB, CPU only, 128GB RAM (slow), CPU only, 256GB RAM (slow)

## Links

- [Hugging Face](https://huggingface.co/nvidia/Flora-Rewarder)