thox-micro-125m-GGUF

<div align="center"> <img src="hero-family.png" alt="The THOX four-device family" width="100%"> <h1>thox-micro-125m-GGUF</h1>

GGUF Builds of the From-Scratch THOX 125M — GGUF builds (Q4_0, Q4_K_M, Q8_0, F16) of Thox-ai/thox-micro-125m, THOX's from-scratch 123M base model, for llama.cpp and on-device use

GGUF: thox-micro-125m.f16.gguf, thox-micro-125m.q4_0.gguf, thox-micro-125m.q4_k_m.gguf, thox-micro-125m.q8_0.gguf

</div>

Your AI. Your Data. Your Rules.™

Status: PUBLISHED — GGUF BUILDS. GGUF builds (Q4_0, Q4_K_M, Q8_0, F16) of Thox-ai/thox-micro-125m, THOX's from-scratch 123M base model, for llama.cpp and on-device use. The model card of record is the main repository's. See Files for exactly what this repo contains.

Summary

GGUF builds (Q4_0, Q4_K_M, Q8_0, F16) of Thox-ai/thox-micro-125m, THOX's from-scratch 123M base model, for llama.cpp and on-device use. Q4_0 is available because the FFN dimension (2048) is divisible by 32. A SHA256SUMS.txt is included in the repo for verification.

Highlights

• GGUF builds (Q4_0, Q4_K_M, Q8_0, F16) of Thox-ai/thox-micro-125m, THOX's from-scratch 123M base model, for llama.cpp and on-device use.
• bash hf download Thox-ai/thox-micro-125m-GGUF --include '*.gguf' --local-dir ./thox-micro-125m-GGUF llama-cli -m ./thox-micro-125m-GGUF/thox-micro-125m.q4_0.gguf -p "Hello"

Specifications

Files

Run it

hf download Thox-ai/thox-micro-125m-GGUF --include '*.gguf' --local-dir ./thox-micro-125m-GGUF
llama-cli -m ./thox-micro-125m-GGUF/thox-micro-125m.q4_0.gguf -p "Hello"

ollama run hf.co/Thox-ai/thox-micro-125m-GGUF:Q4_K_M

This is a base (non-instruct) model: use plain completion prompts, not chat.

Platforms

Intended use

• On-device and edge text completion within the THOX stack.
• Not a safety-aligned public assistant unless deployed behind THOX guardrails.
• Out of scope: medical, legal, financial, safety-critical, or other high-stakes decisions without human review; unattended control of physical devices without a confirmation step.

Training data and procedure

Trained by THOX from scratch as part of the THOX small-model research line. The corpus and procedure are as described in this card at release; no external base model was used.

Evaluation

No THOX GATE49 result is published for this model.

Limitations

• A small base model: weak factual reliability; verify important answers.
• Behaviour and bias reflect the training data, which is mostly English; quality in other languages is lower.
• Quantization trades accuracy for size; lower-bit files degrade more.

License

Apache-2.0. Quantized from a THOX model trained from scratch by THOX.ai LLC; covered under Apache-2.0.

THOX.ai — Your AI. Your Data. Your Rules.™

Provenance

Published by THOX.ai as Thox-ai/thox-micro-125m-GGUF. THOX from-scratch research line; no upstream base model. This card was conformed to THOX Model Card Standard v1.0 on 2026-10-31 (frontmatter, tags, header, files, and section structure); the technical content is as published at release.

Citation

@misc{thoxai2026thoxmicro125mgguf,
      title={thox-micro-125m-GGUF},
      author={THOX.ai},
      year={2026},
      url={https://huggingface.co/Thox-ai/thox-micro-125m-GGUF}
}