korkmazumut/orkhon-tonyuk

Orkhon / tonyuk

22M unified assistant tuned from the Tonyukuk story base.

This repository is one member of the Orkhon model family, an auditable from-scratch LLM stack covering tokenizer, pretraining, post-training, evaluation, serving, and Hugging Face export.

• Family: Orkhon
• Member: tonyuk
• Kind: instruct
• Interface: chat
• Parameters: ~21.9M
• Source code: https://github.com/UmutKorkmaz/orkhon
• Local model-zoo folder: models/tonyuk-20260620

Family Context

Member	Role	HF repo
bumin-mini	4M unified assistant smoke model for EN/TR/Kokturk behavior.	korkmazumut/orkhon-bumin-mini
tonyuk current	22M unified assistant tuned from the Tonyukuk story base.	korkmazumut/orkhon-tonyuk
tegin	22M unified assistant tuned from the former Kultegin instruct model.	korkmazumut/orkhon-tegin
istem	51M unified assistant tuned from the FineWeb-Edu base.	korkmazumut/orkhon-istem
kashgar	135M imported assistant slot once local weights are archived.	korkmazumut/orkhon-kashgar
bunghu	57M unified EN/TR/Kokturk assistant tuned from the bilingual branch.	korkmazumut/orkhon-bunghu
tangri	100M unified EN/TR/Kokturk assistant trained from the mixed Tangri base.	korkmazumut/orkhon-tangri
qaghan	Future larger unified assistant slot.	korkmazumut/orkhon-qaghan

Members without archived local weights are skipped by the preparation script.

Intended Use

• Inspecting and reproducing the Orkhon training/export path.
• Running small local demos on CPU/MPS/CUDA.
• Comparing Orkhon family members by training stage and data mix.
• Rune-to-Latin Old Turkic transliteration demos.

Not Intended For

• Reliable factual QA.
• Safety-critical decisions.
• Claims of state-of-the-art Turkish or general LLM performance.
• Treating Old Turkic transliteration as modern Turkish translation.

Evaluation

Metric	Value
checkpoint	runs/tonyuk_unified
tokenizer	models/tonyukuk-20260614/tokenizer
mode	chat
tag	last
device	mps
deterministic_router	True
params	21895680
step	1800
passed	True
elapsed_seconds	2.58

Current benchmark reports in the source repo are smoke baselines unless marked otherwise. Do not treat limit-20 benchmark runs as headline capability claims.

Example Prompt

What can you help me with?

Usage

Install Orkhon from the source repo, then load this exported folder:

pip install git+https://github.com/UmutKorkmaz/orkhon

from huggingface_hub import snapshot_download
from orkhon.export.to_hf import load_exported_model
from orkhon.tokenizer import load_tokenizer

path = snapshot_download("korkmazumut/orkhon-tonyuk")
model, cfg = load_exported_model(path, device="cpu")
tok = load_tokenizer(path)

Use orkhon.serve.chat_cli.reply for the unified assistant members.

Files

• model.safetensors - exported inference weights
• config.json - Orkhon architecture config
• tokenizer.json, tokenizer_config.json, special_tokens_map.json
• manifest.json - model-zoo metadata from the source repo
• samples.txt - saved local samples when available
• orkhon_model_card.md - original local model card

Limitations

This is a small research/education model family. The models can repeat, confabulate, and fail outside their narrow data scale. Public demos should keep the claim narrow: Orkhon is an auditable from-scratch stack with a Turkic branch, not a frontier assistant.