GGUFGuy/floppyx3-MEGAmodel-merged-GGUF overview
floppyx3 MEGAmodel merged GGUF GGUF conversions of NILKNARFGonzo/floppyx3 MEGAmodel merged https://huggingface.co/NILKNARFGonzo/floppyx3 MEGAmodel merged . Ava…
Runs locally from ~2.5 MB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
Model Details
| Model ID | GGUFGuy/floppyx3-MEGAmodel-merged-GGUF |
|---|---|
| Author | GGUFGuy |
| Pipeline | text-generation |
| License | cc-by-sa-4.0 |
| Base model | NILKNARFGonzo/floppyx3-MEGAmodel-merged |
| Last modified | 2026-09-05T23:39:31.000Z |
Model README
---
base_model:
- NILKNARFGonzo/floppyx3-MEGAmodel-merged
datasets:
- ssmi153/Capybara-ShareGPT
language:
- en
pipeline_tag: text-generation
tags:
- base
- gguf
- ggufguy
- megamodel
- floppy
license: cc-by-sa-4.0
---
floppyx3-MEGAmodel-merged GGUF
GGUF conversions of NILKNARFGonzo/floppyx3-MEGAmodel-merged.
Available formats:
- F32
- F16
- BF16
Q8_0 is not included because the model's 106-dimensional embedding tensors are not compatible with Q8_0's 32-value quantization blocks, causing llama.cpp's quantizer to fail.
The original model is a very small ~1.1M-parameter GPT-2-style base model.
Run GGUFGuy/floppyx3-MEGAmodel-merged-GGUF with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models