Instructions to use LHC88/XPurpose-ClownCar-v0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use LHC88/XPurpose-ClownCar-v0 with Transformers:

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

pipe = pipeline("text-generation", model="LHC88/XPurpose-ClownCar-v0")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("LHC88/XPurpose-ClownCar-v0")
model = AutoModelForCausalLM.from_pretrained("LHC88/XPurpose-ClownCar-v0", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use LHC88/XPurpose-ClownCar-v0 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "LHC88/XPurpose-ClownCar-v0"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "LHC88/XPurpose-ClownCar-v0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/LHC88/XPurpose-ClownCar-v0

SGLang

How to use LHC88/XPurpose-ClownCar-v0 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 "LHC88/XPurpose-ClownCar-v0" \
    --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": "LHC88/XPurpose-ClownCar-v0",
		"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 "LHC88/XPurpose-ClownCar-v0" \
        --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": "LHC88/XPurpose-ClownCar-v0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use LHC88/XPurpose-ClownCar-v0 with Docker Model Runner:
```
docker model run hf.co/LHC88/XPurpose-ClownCar-v0
```

Chat & support: LHC's LinkedIn

Want to contribute? LHC's Github Sponsors

XPurpose-ClownCar-v0 is a multi-purpose MoE-model with the following expert configuration.

base_model: openaccess-ai-collective/DPOpenHermes-7B
dtype: bfloat16
experts:
- positive_prompts:
  - instruction
  - solutions
  - chat
  - questions
  - comprehension
  source_model: teknium/OpenHermes-2.5-Mistral-7B
- negative_prompts:
  - chat
  - questions
  - python
  positive_prompts:
  - coding
  - programming
  - code
  - programming language
  source_model: codellama/CodeLlama-13b-hf
- negative_prompts:
  - chat
  - questions
  positive_prompts:
  - python
  - pip
  - coding
  - programming
  - code
  - programming language
  source_model: codellama/CodeLlama-13b-Python-hf
- negative_prompts:
  - chat
  - questions
  positive_prompts:
  - mathematics
  - optimization
  - step-by-step
  - science
  source_model: cognitivecomputations/dolphin-2.6-mistral-7b-dpo
- negative_prompts:
  - chat
  - questions
  positive_prompts:
  - bedtime story
  - Once upon a time
  - storytelling
  - narrator
  source_model: tom92119/llama-2-7b-bedtime-story
- negative_prompts:
  - chat
  - questions
  positive_prompts:
  - story
  - Once upon a time
  - storytelling
  - narrator
  source_model: Norquinal/Mistral-7B-storywriter
- negative_prompts:
  - chat
  - questions
  - instruction
  - solutions
  - chat
  - comprehension
  - mathematics
  - optimization
  - code
  - step-by-step
  - science
  positive_prompts:
  - function calls
  - functions
  - constrained grammar
  - API calls
  - LLM Tools
  source_model: meetkai/functionary-small-v2.2
- positive_prompts:
  - indonesian
  - indonesia
  source_model: azale-ai/Starstreak-7b-beta
- positive_prompts:
  - arabic
  - arab
  source_model: gagan3012/Mistral_arabic_dpo
- positive_prompts:
  - korean
  - korea
  source_model: davidkim205/komt-mistral-7b-v1
- positive_prompts:
  - chinese
  - china
  source_model: OpenBuddy/openbuddy-zephyr-7b-v14.1
- positive_prompts:
  - hindi
  - india
  source_model: manishiitg/open-aditi-hi-v1
- positive_prompts:
  - german
  - deutsch
  - Germany
  source_model: VAGOsolutions/SauerkrautLM-7b-v1-mistral
- positive_prompts:
  - Norway
  - Norwegian
  - Norsk
  source_model: bineric/NorskGPT-Mistral-7b
- positive_prompts:
  - Russian
  - Russia
  - "\u0420\u0443\u0441\u0441\u043A\u0438\u0439"
  - "\u0420\u043E\u0441\u0441\u0438\u044F"
  source_model: Droidfanat/llama-2-7b-custom-russian
gate_mode: hidden

Prompt template: ChatML

<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

Downloads last month: 7

Safetensors

Model size

86B params

Tensor type

BF16