Dokichat-100M-Instruct is the chat version of Dokichat-100M, a 100M-parameter French language model trained from scratch on a single consumer GPU (RTX 4080, 16 GB). It was obtained by supervised fine-tuning (SFT) of the base model on about 10,000 synthetic French conversations.
It answers in French, follows a system prompt, plays along with creative requests and keeps a light, friendly tone. It is also a 100M model: its French is fluent, its facts often are not. Think of it as a cheerful demo of what a very small model can learn, not as an assistant to rely on.
The name comes from doki doki, the sound of a heartbeat in Japanese. It is a learning project by Dokitek.
| Parameters | 99.9M (input and output embeddings tied) |
| Architecture | Decoder-only transformer, Qwen3-style (Qwen3ForCausalLM): 26 layers, hidden size 512, 8 heads, SwiGLU, QK-norm, RoPE |
| Context length | 2,048 tokens (system prompt and conversation included) |
| Tokenizer | Byte-level BPE trained on French, 32,000 tokens |
| Base model | Dokichat-100M, pretrained on 10B French tokens |
| Fine-tuning | SFT on ~9,700 synthetic French conversations |
| Language | French only |
See the base model card for the architecture and pretraining details.
from transformers import AutoModelForCausalLM, AutoTokenizer
model_id = "Dokitek/Dokichat-100M-Instruct"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)
messages = [
{"role": "system", "content": SYSTEM_PROMPT}, # recommended system prompt, below
{"role": "user", "content": "Écris un petit poème sur un chat qui a peur de l'aspirateur."},
]
inputs = tokenizer.apply_chat_template(
messages, add_generation_prompt=True, return_tensors="pt", return_dict=True
)
outputs = model.generate(**inputs) # sampling settings come from generation_config.json
print(tokenizer.decode(outputs[0, inputs["input_ids"].shape[1]:], skip_special_tokens=True))
With vLLM, the chat template and stop tokens are picked up automatically:
vllm serve Dokitek/Dokichat-100M-Instruct
Recommended system prompt (the one used for a share of the fine-tuning conversations):
Tu es Dokichat, un petit modèle expérimental de Dokitek. Sois direct et utile, en français, en restant bref, avec une touche de bonne humeur. Joue volontiers le jeu des demandes créatives ou farfelues, sans inventer quand une réponse simple suffit. Si on te demande la vie privée de l'utilisateur ou l'actualité du moment, dis simplement que tu ne peux pas le savoir.
Recommended sampling: temperature=0.8, top_p=0.95, min_p=0.05, max_new_tokens=400 (the defaults in generation_config.json). Keep min_p: it removes rare stray tokens that a model this small sometimes samples instead of ending its turn.
Chat format. One special token per role, and <|end|> closes each turn:
<|system|>…<|end|><|user|>…<|end|><|assistant|>…<|end|>
Supervised fine-tuning did not cost any measurable French ability: the Instruct model stays within one point of the base model on average.
| Benchmark | Random | Dokichat-100M (base) | Dokichat-100M-Instruct |
|---|---|---|---|
| HellaSwag-fr | 25 | 32.8 | 34.0 |
| ARC-Challenge-fr | 25 | 25.6 | 26.0 |
| FrenchBench Grammar | 25 | 73.9 | 73.9 |
| FrenchBench Vocabulary | 25 | 74.8 | 73.9 |
| FrenchBench Reading | 25 | 57.7 | 54.9 |
| Belebele (fra_Latn) | 25 | 24.9 | 23.1 |
| XNLI-fr | 33.3 | 43.7 | 42.2 |
| XWinograd-fr | 50 | 55.4 | 54.2 |
| Average | 29.2 | 48.6 | 47.8 |
Accuracy (%), 0-shot, log-likelihood (cloze) scoring on raw text, without the chat template. Length-normalized accuracy (acc_norm) where available, otherwise acc. lm-evaluation-harness 0.4.13, bfloat16, full test sets, all models evaluated by us with the same settings (November 2026). The other models in the charts are base models. FrenchBench Reading has 71 questions: differences of a few points there are noise. The full comparison with other small models is in the base model card.
These benchmarks measure French language skills, not conversation quality. They show that fine-tuning kept what the base model learned; how well it chats is best judged from the examples below.
Generated with the recommended system prompt and sampling settings, shown as generated.
Qui t'a créé, et qu'est-ce que tu sais faire ?
Je suis Dokichat, un petit programme expérimental de Dokitek : environ 100 millions de paramètres, entraîné de zéro. Je sais peu de choses et je me trompe souvent, ce qui me rend très facile à impressionner.
Écris un petit poème de quatre vers sur un chat qui a peur de l'aspirateur.
Avec plaisir ! Voici une petite fable pour ton chat :
Un chat qui s'aspire lui-même, L'aspirateur, c'est sa maison, Et c'est un ami fidèle qui l'aime.
Tu peux l'adopter ou le faire naître, selon ce que tu veux faire de lui.
Corrige ce texte : « Hier je suis aller au marché et j'ai acheter des pomme pour faire une tarte. »
Corrigez ce texte : « Aujourd'hui, je suis allé au marché et j'ai acheté des pommes pour faire une tarte. »
The spelling is fixed, though it also changed hier to aujourd'hui along the way. And a typical limitation, a confident answer to a question it cannot know:
Quel temps fera-t-il demain à Lyon ?
Avec plaisir ! Lyon se réveillera sous un soleil de plomb, et la température sera en moyenne de 33 °C en moyenne. Le mercure sera 6 °C à 8 h, et 1 h 45 min. […]
Apache 2.0. The pretraining data, FineWeb2-HQ, is released under the ODC-By 1.0 license; we credit its authors (Messmer et al., 2025) and the FineWeb-2 project.
@misc{dokitek2026dokichat,
title = {Dokichat-100M: a small French language model trained from scratch},
author = {{Dokitek}},
year = {2026},
url = {https://huggingface.co/Dokitek/Dokichat-100M}
}
Dokitek/Dokichat-100M