1bit-MONSTER/Llama-3.3-70B-Instruct-GGUF overview
Llama 3.3 70B Instruct — GGUF Q4 K M bartowski's Q4 K M GGUF, re hosted with measured performance for the 1bit engine https://github.com/1bit MONSTER/engine on…
Runs locally from ~39.60 GB disk (32 GB+ VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| Llama-3.3-70B-Instruct-Q4_K_M.gguf | GGUF | Q4_K_M | 39.60 GB | Download |
Model Details
| Model ID | 1bit-MONSTER/Llama-3.3-70B-Instruct-GGUF |
|---|---|
| Author | 1bit-MONSTER |
| Pipeline | — |
| License | llama3.3 |
| Base model | meta-llama/Llama-3.3-70B-Instruct |
| Last modified | 2026-09-26T14:50:02.000Z |
Model README
---
license: llama3.3
base_model:
- meta-llama/Llama-3.3-70B-Instruct
tags:
- gguf
---
Llama-3.3-70B-Instruct — GGUF (Q4_K_M)
bartowski's Q4_K_M GGUF, re-hosted with measured performance for the
1bit engine on Strix Halo.
Licensed under the Llama 3.3 Community License —
"Built with Llama" attribution required; a separate license is needed if your
product/service exceeds 700M monthly active users.
Contents
Llama-3.3-70B-Instruct-Q4_K_M.gguf(39.6 GiB)
Measured performance (Strix Halo, Vulkan)
pp512: 88 tok/s · tg128: 4.6 tok/s
The largest dense model in this collection — usable, but the slowest. Needs ~40 GiB
of resident memory.
Running it
1bit serve -m Llama-3.3-70B-Instruct-Q4_K_M.gguf --device vulkanAttribution
Built with Llama. Base model meta-llama/Llama-3.3-70B-Instruct;
quantization by bartowski.
Run 1bit-MONSTER/Llama-3.3-70B-Instruct-GGUF with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models