Model Intelligence Sheet
1bit-MONSTER/gpt-neox-20b-GGUF overview
gpt neox 20b — GGUF Q4 K M EleutherAI's own architecture, community Q4 K M GGUF, re hosted with measured performance for the 1bit engine https://github.com/1bi…
Runs locally from ~12.23 GB disk (16 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
1 GGUF files detected
Direct downloads for local inference
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| gpt-neox-20b.Q4_K_M.gguf | GGUF | GGUF | 12.23 GB | Download |
Model Details
| Model ID | 1bit-MONSTER/gpt-neox-20b-GGUF |
|---|---|
| Author | 1bit-MONSTER |
| Pipeline | — |
| License | apache-2.0 |
| Base model | EleutherAI/gpt-neox-20b |
| Last modified | 2026-09-26T16:19:55.000Z |
Model README
---
license: apache-2.0
base_model:
- EleutherAI/gpt-neox-20b
tags:
- gguf
---
gpt-neox-20b — GGUF (Q4_K_M)
EleutherAI's own architecture, community Q4_K_M GGUF, re-hosted with measured
performance for the 1bit engine on Strix Halo.
A legacy but architecturally distinct large open pretrained model (not instruction-tuned).
Contents
gpt-neox-20b.Q4_K_M.gguf
Measured performance (Strix Halo, Vulkan)
pp512: 292 tok/s · tg128: 14.6 tok/s
Running it
1bit serve -m gpt-neox-20b.Q4_K_M.gguf --device vulkanAttribution
- Model: EleutherAI/gpt-neox-20b, Apache 2.0.
- Quantization: mradermacher/gpt-neox-20b-GGUF.
- License: Apache 2.0, inherited from the base model.
Run 1bit-MONSTER/gpt-neox-20b-GGUF with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models