Instructions to use GautierP/MiniLLM-40M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use GautierP/MiniLLM-40M with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="GautierP/MiniLLM-40M", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("GautierP/MiniLLM-40M", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use GautierP/MiniLLM-40M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "GautierP/MiniLLM-40M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "GautierP/MiniLLM-40M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/GautierP/MiniLLM-40M

SGLang

How to use GautierP/MiniLLM-40M with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "GautierP/MiniLLM-40M" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "GautierP/MiniLLM-40M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "GautierP/MiniLLM-40M" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "GautierP/MiniLLM-40M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use GautierP/MiniLLM-40M with Docker Model Runner:
```
docker model run hf.co/GautierP/MiniLLM-40M
```

Mini-LLM Assistant 40M

A ~40M parameter conversational assistant with a full training pipeline implemented end-to-end in PyTorch: BPE tokenizer, pretraining and supervised fine-tuning, no pretrained backbone.

This repository hosts the SFT assistant. The base model it was fine-tuned from is published separately: GautierP/MiniLLM-40M-base.

Model details


Parameters	42,476,032 (input embedding shared with the output head)
Architecture	Transformer decoder: RMSNorm, RoPE, causal attention, SwiGLU
Context length	1,024 tokens, twice the 512 of the earlier 1.0-generation runs
Vocabulary	32,768 (BPE)
Pretraining	~~4.60B tokens (FineWeb-Edu + Wikipedia), 35,096 steps (~~115 tokens/param), base val loss 3.202
Fine-tuning	2 epochs of SFT on Dolly-15k, No Robots, OpenAssistant oasst2, alpaca-cleaned and identity/small-talk examples, val loss 3.058
Efficiency	~115 tokens/param, zero-shot scores at or above the family 70M (1.44B tokens) with 40% fewer parameters

Base model

The base repository contains the pretrained checkpoint (mini-llm-base-40M-35096steps.pt) before SFT: 42,476,032 parameters, 32,768-token BPE vocabulary, ~4.60B tokens (FineWeb-Edu + Wikipedia), 35,096 steps, base val loss 3.202. It is a raw completion model, not an assistant: the base ships without a chat template and without the assistant's sampling defaults.

from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "GautierP/MiniLLM-40M-base"
tok = AutoTokenizer.from_pretrained(repo, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(repo, trust_remote_code=True)

inputs = tok("The capital of France is", return_tensors="pt")
out = model.generate(**inputs, max_new_tokens=64, do_sample=True, temperature=0.8, top_k=50)
print(tok.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "GautierP/MiniLLM-40M"
tok = AutoTokenizer.from_pretrained(repo, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(repo, trust_remote_code=True)

messages = [{"role": "user", "content": "What is the capital of France?"}]
inputs = tok.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt")
out = model.generate(inputs, max_new_tokens=128)
print(tok.decode(out[0][inputs.shape[1]:], skip_special_tokens=True))

trust_remote_code=True is required, the architecture ships with the repository (modeling_mini_llm.py, configuration_mini_llm.py). Sampling defaults (temperature 0.5, top_k 40, repetition penalty 1.15) live in generation_config.json.

The training format is single-turn and fixed; chat_template.jinja implements it exactly:

User: <question>
Assistant: <answer>

Each question is processed independently, no conversation history is attached.

Limitations

Factual accuracy is limited by model scale: fluent, well-formatted answers do not imply verified facts, and any factual statement should be treated as unconfirmed. Generation quality is tracked by the Mini-Eval suite. English only.

Collection including GautierP/MiniLLM-40M

Mini-LLM

Collection

3 items • Updated 4 days ago

GautierP
/

MiniLLM-40M