Instructions to use c-bone/CrystaLLM-pi_alex_mp_20_base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use c-bone/CrystaLLM-pi_alex_mp_20_base with Transformers:

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

pipe = pipeline("text-generation", model="c-bone/CrystaLLM-pi_alex_mp_20_base")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("c-bone/CrystaLLM-pi_alex_mp_20_base")
model = AutoModelForCausalLM.from_pretrained("c-bone/CrystaLLM-pi_alex_mp_20_base", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use c-bone/CrystaLLM-pi_alex_mp_20_base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "c-bone/CrystaLLM-pi_alex_mp_20_base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "c-bone/CrystaLLM-pi_alex_mp_20_base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/c-bone/CrystaLLM-pi_alex_mp_20_base

SGLang

How to use c-bone/CrystaLLM-pi_alex_mp_20_base 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 "c-bone/CrystaLLM-pi_alex_mp_20_base" \
    --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": "c-bone/CrystaLLM-pi_alex_mp_20_base",
		"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 "c-bone/CrystaLLM-pi_alex_mp_20_base" \
        --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": "c-bone/CrystaLLM-pi_alex_mp_20_base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use c-bone/CrystaLLM-pi_alex_mp_20_base with Docker Model Runner:
```
docker model run hf.co/c-bone/CrystaLLM-pi_alex_mp_20_base
```

Model Card for CrystaLLM-pi_alex_mp_20_base

Model Details

Model Description

CrystaLLM-pi_alex_mp_20_base is an unconditional generative model for crystal structures. It is a GPT-2 decoder-only model that generates Crystallographic Information Files (CIFs) from the patterns learned during training, with no property or diffraction conditioning attached.

The model was trained from scratch on Alex-MP-20 CIF text as an unconditional baseline for LeMat-Bench. For fine-tuning, use c-bone/CrystaLLM-pi_ft_alex_mp_20-text, which was initialised from the LeMat-Bulk pretrained model.

Developed by: Bone et al. (University College London)
Model type: Autoregressive Transformer, unconditional (~25.9M parameters)
Language(s): CIF (Crystallographic Information File) syntax
License: MIT
Finetuned from model: None, trained from scratch

Model Sources

Repository: GitHub: CrystaLLM-pi (generation)
Dataset: HuggingFace: c-bone/alex_mp_20

Uses

Direct Use

Unconditional or prompt-steered generation of inorganic crystal structures, and evaluation as the Alex-MP-20 baseline in LeMat-Bench.

Out-of-Scope Use

Property targeting: the model has no conditioning channel. Use a conditional model to request a property value.
Disordered Systems: No native handling of partial occupancies or significant disorder.
Organic/MOFs: Training data is inorganic, so organic frameworks are out of distribution.
Extremely Large Unit Cells: the context window is 1024 tokens, so large cells will not fit.

Bias, Risks, and Limitations

Distribution matching: output closely reflects the Alex-MP-20 distribution, so novelty is limited by design. Raise the generation temperature to move away from it.
Validity is not stability: generated CIFs passing structural checks are not necessarily thermodynamically stable. Screen with an energy model before drawing conclusions.
Computed data: training structures are relaxed DFT results rather than experimentally determined ones.

Getting started

For generation, use T2_load_and_generate.ipynb in CrystaLLM-pi. The training config is alex-mp-20-text.jsonc.

Citation

@misc{bone2025discoveryrecoverycrystallinematerials,
      title={Discovery and recovery of crystalline materials with property-conditioned transformers},
      author={Cyprien Bone and Matthew Walker and Bradley A. A. Martin and Kuangdai Leng and Luis M. Antunes and Ricardo Grau-Crespo and Amil Aligayev and Javier Dominguez and Keith T. Butler},
      year={2025},
      eprint={2511.21299},
      archivePrefix={arXiv},
      primaryClass={cond-mat.mtrl-sci},
      url={https://arxiv.org/abs/2511.21299},
}

Downloads last month: 185

Safetensors

Model size

25.9M params

Tensor type

F32

Dataset used to train c-bone/CrystaLLM-pi_alex_mp_20_base

Collection including c-bone/CrystaLLM-pi_alex_mp_20_base

CrystaLLM-pi

Collection

Models and data from our 'Discovery and recovery of crystalline materials with property-conditioned transformers' Paper and Package • 25 items • Updated 4 days ago

Paper for c-bone/CrystaLLM-pi_alex_mp_20_base

Discovery and recovery of crystalline materials with property-conditioned transformers

Paper • 2511.21299 • Published Nov 26, 2025