trinity-nano-stheno — GGUF

realoperator42/trinity-nano-stheno-lora (attention-only LoRA, r=16, alpha=32) merged into

Architecture is afmoe, so you need a llama.cpp build that includes AFMOE support.

The embedded chat template is the adapter's corrected ChatML, with the training-only

{% generation %} markers removed so --jinja parses cleanly.

llama-cli -m trinity-nano-stheno-Q4_K_M.gguf --jinja -cnv

Files

| File | Size |

| --- | --- |

| trinity-nano-stheno-Q4_K_M.gguf | 3.8 GB |

| trinity-nano-stheno-Q8_0.gguf | 6.5 GB |

Question 2

What license applies to realoperator42/trinity-nano-stheno-GGUF?

Accepted Answer

License: apache-2.0. Verify terms on Hugging Face before commercial use.

Question 3

How do I run realoperator42/trinity-nano-stheno-GGUF locally?

Accepted Answer

Download a GGUF file from this page and load it in guIDE or llama.cpp. Pipeline task: text-generation.

Question 4

How much VRAM or disk space does realoperator42/trinity-nano-stheno-GGUF need?

Accepted Answer

Runs locally from ~3.50 GB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).

File	Type	Quantization	Size	Link
trinity-nano-stheno-Q4_K_M.gguf	GGUF	Q4_K_M	3.50 GB	Download
trinity-nano-stheno-Q8_0.gguf	GGUF	Q8_0	6.08 GB	Download

Model ID	realoperator42/trinity-nano-stheno-GGUF
Author	realoperator42
Pipeline	—
License	apache-2.0
Base model	arcee-ai/Trinity-Nano-Preview
Last modified	2026-08-08T22:10:02.000Z