ggml-org/tinyopenjev-for-testing-gguf overview
tinyopenjev for testing gguf Q8 0 GGUF text model + mmproj of a very small slice of openjev/openjev https://huggingface.co/openjev/openjev , for testing purpos…
Runs locally from ~3.1 MB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
Model Details
Model README
---
license: cc-by-nc-4.0
base_model: openjev/openjev
tags: [testing, gguf]
---
tinyopenjev-for-testing-gguf
Q8_0 GGUF (text model + mmproj) of a very small slice of openjev/openjev, for testing purposes only.
The slice keeps the first 4 of 64 text layers and the first 2 of 27 vision blocks, and also cuts the width (hidden size 5120 -> 64, fewer attention heads, smaller FFN). About 34M parameters, most of them in the token embedding and output matrices.
The output is not meaningful. Do not use this model for anything other than testing a loader or a runtime.
OpenJev weights are released under CC BY-NC 4.0 and are derived from Qwen/Qwen3.8-27B (Apache 2.0); see NOTICE, LICENSE and LICENSE-APACHE-2.0.
Run ggml-org/tinyopenjev-for-testing-gguf with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models