Instructions to use drooryck/multilingual-macaroni-models with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use drooryck/multilingual-macaroni-models with Transformers:

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

pipe = pipeline("text-generation", model="drooryck/multilingual-macaroni-models")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("drooryck/multilingual-macaroni-models", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use drooryck/multilingual-macaroni-models with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "drooryck/multilingual-macaroni-models"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "drooryck/multilingual-macaroni-models",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/drooryck/multilingual-macaroni-models

SGLang

How to use drooryck/multilingual-macaroni-models 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 "drooryck/multilingual-macaroni-models" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "drooryck/multilingual-macaroni-models",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "drooryck/multilingual-macaroni-models" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "drooryck/multilingual-macaroni-models",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use drooryck/multilingual-macaroni-models with Docker Model Runner:
```
docker model run hf.co/drooryck/multilingual-macaroni-models
```

Add pipeline tag, link to paper and code

by nielsr HF Staff - opened Sep 28

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

+15

-4

Files changed (1) hide show

README.md +15 -4

README.md CHANGED Viewed

@@ -1,12 +1,23 @@
 ---
-language: [en, nl, zh]
-license: mit
 library_name: transformers
-tags: [babylm, babylm-2026, multilingual, code-switching, gpt2]
 ---
 # multilingual-macaroni-models
 All 24 models from our BabyLM 2026 Multilingual-track study of **code-switched pretraining
 curricula** (English / Dutch / Chinese): **8 training conditions × 3 seeds (42/43/44)**, each a
 12-layer GPT-2 (~98M params, 16k vocab, 1024 context) trained from scratch on the 100M
@@ -44,4 +55,4 @@ model = AutoModelForCausalLM.from_pretrained(REPO, revision="curriculum-s43")  #
 ## Recipe
 GPT-2 (12L, 768d, 12 heads, 1024 ctx, 16k vocab); LR 5e-5, cosine-with-min-lr, warmup 0.01,
-batch 16, AdamW, 10 epochs over the corpus. See our paper for full details.

 ---
+language:
+- en
+- nl
+- zh
 library_name: transformers
+license: mit
+pipeline_tag: text-generation
+tags:
+- babylm
+- babylm-2026
+- multilingual
+- code-switching
+- gpt2
 ---
 # multilingual-macaroni-models
+This repository contains the models from the paper [Feeding BabyLMs Macaroni: Code-Switching Curricula Cause Cross-Lingual Convergence](https://huggingface.co/papers/2609.30535). Code and training scripts are available at [GitHub](https://github.com/drooryck/multilingual-macaroni).
 All 24 models from our BabyLM 2026 Multilingual-track study of **code-switched pretraining
 curricula** (English / Dutch / Chinese): **8 training conditions × 3 seeds (42/43/44)**, each a
 12-layer GPT-2 (~98M params, 16k vocab, 1024 context) trained from scratch on the 100M
 ## Recipe
 GPT-2 (12L, 768d, 12 heads, 1024 ctx, 16k vocab); LR 5e-5, cosine-with-min-lr, warmup 0.01,
+batch 16, AdamW, 10 epochs over the corpus. See our paper for full details.