Instructions to use Leon-LLM/Leon-Chess-71k-20E with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Leon-LLM/Leon-Chess-71k-20E with Transformers:

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

pipe = pipeline("text-generation", model="Leon-LLM/Leon-Chess-71k-20E")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Leon-LLM/Leon-Chess-71k-20E")
model = AutoModelForCausalLM.from_pretrained("Leon-LLM/Leon-Chess-71k-20E")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Leon-LLM/Leon-Chess-71k-20E with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Leon-LLM/Leon-Chess-71k-20E"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Leon-LLM/Leon-Chess-71k-20E",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Leon-LLM/Leon-Chess-71k-20E

SGLang

How to use Leon-LLM/Leon-Chess-71k-20E 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 "Leon-LLM/Leon-Chess-71k-20E" \
    --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": "Leon-LLM/Leon-Chess-71k-20E",
		"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 "Leon-LLM/Leon-Chess-71k-20E" \
        --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": "Leon-LLM/Leon-Chess-71k-20E",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Leon-LLM/Leon-Chess-71k-20E with Docker Model Runner:
```
docker model run hf.co/Leon-LLM/Leon-Chess-71k-20E
```

Chess Language Model

This model is a GPT-2-based language model trained on chess game sequences using our unique xLAN notation, focusing on predicting legal chess moves and understanding game dynamics. For more information see GitHub

Model Details

Model Description

Developed by: Schmid Lars, Maag Jerome
Model type: GPT-2 adaptation for chess language understanding
Language(s) (NLP): xLAN (Chess Notation)

Model Sources

Repository: Leon LLM Chess Research

Uses

Direct Use

The model is intended for predicting chess moves, analyzing game positions, and studying chess strategies.

Out-of-Scope Use

This model is not designed for general language understanding or tasks unrelated to chess.

Bias, Risks, and Limitations

The model reflects the strategies and styles present in the training dataset and may not encompass all possible chess scenarios.

How to Get Started with the Model

To use the model use the Notebooks on GitHub

Training Details

Training Data

The model was trained on the "Leon-LLM/Leon-Chess-Dataset-71k" dataset, sourced from Lichess database (September 2023).

Training Procedure

Preprocessing

Chess games from Lichess were converted from PGN into xLAN as part of the preprocessing.

Training Hyperparameters

Batch Size: 23
Epochs: 20
Learning Rate: 0.0001

Evaluation

Metrics

The model was tested on 3 metrics:

Average Number of Correct Plies: Measures the model's ability to simulate chess games, evaluating the average number of correctly generated plies in 100 games.
Hard Position Accuracy: Assesses the model's handling of 67 challenging chess positions, including unusual castling, pawn promotions, and checkmate scenarios. Success is measured by the model's ability to generate legal moves in these complex positions.
Legal Piece Moves Accuracy: Examines the model's proficiency in keeping the board state various situations, including checks, pinned pieces, and pawn promotions. The metric focuses on the model's understanding of the board state.

Results

Model Architecture

GPT2 config with the following changes:

VOCAB_SIZE = 75
N_POSITION = 512
PAD_TOKEN_ID = 0
EOS_TOKEN_ID = 74

Downloads last month: 5

Collection including Leon-LLM/Leon-Chess-71k-20E

Leon LLM Chess

Collection

Collection of Leon Chess models and datasets - https://github.zhaw.ch/schmila7/leon-llm • 14 items • Updated Mar 2 • 1