# Hermes-3-Llama-3.1-8B GPU requirements

Hermes-3-Llama-3.1-8B has 8.0B parameters. To serve it, plan for about 19 GB of GPU memory in 16-bit, 9.6 GB in 8-bit or 4.8 GB in 4-bit, including about 20% for the KV cache and activations.

[Deploy it on Arrakis](/console/#/models?model=NousResearch/Hermes-3-Llama-3.1-8B).

## What it takes to serve it

| Precision | Memory and the smallest setup |
|---|---|
| 16-bit | 19 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |
| 8-bit | 9.6 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |
| 4-bit | 4.8 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |

## What it takes to fine-tune it

| Method | Memory and the smallest setup |
|---|---|
| QLoRA | 5.6 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |
| LoRA | 19 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |
| Full fine-tuning | 128 GB: one [H200](/gpus/h200.html.md), from $4.03 per GPU-hour |

## About the model

| | |
|---|---|
| Parameters | `8.0B` |
| Context window (tokens) | `131,072` |
| Published in | `BF16` |
| License | `llama3` |
| Released | `2024-07-28` |
| Weights | [NousResearch/Hermes-3-Llama-3.1-8B](https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B) |

## Serve it yourself

```
vllm serve NousResearch/Hermes-3-Llama-3.1-8B --tensor-parallel-size 1
```

## Questions

**How much GPU memory does Hermes-3-Llama-3.1-8B need?** To serve it, about 19 GB in 16-bit, 10 GB in 8-bit, 5 GB in 4-bit, including about 20% for the KV cache and activations. To fine-tune it, about 6 GB with QLoRA, 19 GB with LoRA and 128 GB for full fine-tuning with Adam, before activations.

**Can I run Hermes-3-Llama-3.1-8B on one GPU?** Yes. In 16-bit it needs about 19 GB, which fits on one RTX 4090 or RTX 3090.

**What does it cost to run Hermes-3-Llama-3.1-8B?** In 16-bit it needs one RTX 4090 or RTX 3090; in 4-bit, one RTX 4090 or RTX 3090. Each setup on this page shows the lowest price per GPU-hour on Arrakis right now, and you pay by the hour from a prepaid balance.

**How do I deploy Hermes-3-Llama-3.1-8B on Arrakis?** Open Models in the Arrakis console, paste NousResearch/Hermes-3-Llama-3.1-8B, choose vLLM, SGLang or llama.cpp, and rent a GPU it fits on. You get an OpenAI-compatible endpoint with its own API key, and root SSH to the machine.


More models: [/models](/models.html.md).