Instructions to use xxond/disco-limbic-dialogue with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use xxond/disco-limbic-dialogue with Transformers:

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

pipe = pipeline("text-generation", model="xxond/disco-limbic-dialogue", trust_remote_code=True)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("xxond/disco-limbic-dialogue", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use xxond/disco-limbic-dialogue with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/xxond/disco-limbic-dialogue

SGLang

How to use xxond/disco-limbic-dialogue 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 "xxond/disco-limbic-dialogue" \
    --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": "xxond/disco-limbic-dialogue",
		"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 "xxond/disco-limbic-dialogue" \
        --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": "xxond/disco-limbic-dialogue",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use xxond/disco-limbic-dialogue with Docker Model Runner:
```
docker model run hf.co/xxond/disco-limbic-dialogue
```

Project Overview

This project features an AI model, specifically trained on a subset of dialogues from the game "Disco Elysium". The model, developed by xxond, mimics the game's distinct style, often characterized by its impolite and candid dialogue.

Technical Specifications

Base Model: microsoft/phi-2
License: Adheres to the licensing terms of the base model.

Development and Purpose

Developed as a personal project, the primary objective is to explore and experiment with the capabilities of language models in replicating specific narrative styles. This project serves as an innovative playground for AI enthusiasts and is not intended for commercial use. Additional information and project updates can be found on the accompanying GitHub repository.

Training Data

The AI was trained on dialogue data extracted from "Disco Elysium", which underwent extensive preprocessing to align with the model's requirements.

Contact and Legal Disclaimer

For further details on the project's methodology, or to address any legal queries, please feel free to contact me. While the legal aspects of utilizing game dialogue for AI training are still being explored, your input and inquiries are welcomed to ensure compliance and ethical development.

Downloads last month: 5

Safetensors

Model size

3B params

Tensor type

BF16