GMLinux 9B
GMLinux 9B is a 9-billion-parameter merged vision-language model sponsored by Galactic Mandate Linux.
Model files
| File | Format | Description |
|---|---|---|
| Gmlinux9b.safetensors | BF16 Safetensors | Transformers model weights |
| GMlinuxQ8.gguf | Q8_0 GGUF | Quantized model for compatible runtimes |
The repository also includes the configuration, tokenizer, processor, and chat-template files used by the Transformers release.
Context and vision
The configuration advertises a context length of up to 262,144 tokens. Practical context depends on the inference runtime and available memory. Local use has been tested at 68,096 tokens; longer contexts have not been validated here.
For image input, use the Transformers model and processor files. The Q8 GGUF may require a separate multimodal projector, depending on the runtime. Check the runtime's vision setup before expecting image support from that file.
Evaluation
On Galactic Mandate Linux's internal knowledge and reasoning benchmarks, GMLinux 9B scores 30% higher than Qwen3.5-9B and Ornith-1.5-9B.
General coding performance has not been established as superior to the base models. Evaluate it on your intended tasks before relying on it.
Transformers
import torch
from transformers import AutoModelForImageTextToText, AutoProcessor
model_id = "emperorofrome/GMLINUX-9B"
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForImageTextToText.from_pretrained(
model_id,
dtype=torch.bfloat16,
device_map="auto",
)
For LM Studio or llama.cpp, load GMlinuxQ8.gguf and set the context length according to available memory and runtime support.
Upstream models and licensing
The listed upstream repositories are Qwen3.5-9B (Apache-2.0) and Ornith-1.5-9B (MIT). This repository retains its existing Apache-2.0 metadata. Review the upstream license notices before using or redistributing the model.