Instructions to use liemo/open-lilm-v2-q4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use liemo/open-lilm-v2-q4 with Transformers:

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

pipe = pipeline("text-generation", model="liemo/open-lilm-v2-q4")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("liemo/open-lilm-v2-q4")
model = AutoModelForCausalLM.from_pretrained("liemo/open-lilm-v2-q4")
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

vLLM

How to use liemo/open-lilm-v2-q4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "liemo/open-lilm-v2-q4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "liemo/open-lilm-v2-q4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/liemo/open-lilm-v2-q4

SGLang

How to use liemo/open-lilm-v2-q4 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 "liemo/open-lilm-v2-q4" \
    --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": "liemo/open-lilm-v2-q4",
		"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 "liemo/open-lilm-v2-q4" \
        --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": "liemo/open-lilm-v2-q4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use liemo/open-lilm-v2-q4 with Docker Model Runner:
```
docker model run hf.co/liemo/open-lilm-v2-q4
```

open-lilm-v2-q4

This is simply a quantized version of open-lilm-v2 without other modification. This model is only intended for research or entertainment purposes as the original model.

Warning: Due to the nature of the training data, this model is highly likely to return violent, racist and discriminative content. DO NOT USE IN PRODUCTION ENVIRONMENT.

Model Details

Name: open-lilm-v2-q4
Quantization: 4-bit quantization
Base Model: 0xtaipoian/open-lilm-v2

Usage

This model can be used with Hugging Face's Transformers library:

from transformers import AutoTokenizer, AutoModelForCausalLM

model_name = "liemo/open-lilm-v2-q4"

model = AutoModelForCausalLM.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)

def chat(messages, temperature=0.9, max_new_tokens=200):
    input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt').to('cuda:0')
    output_ids = quantized_model.generate(input_ids, max_new_tokens=max_new_tokens, temperature=temperature, do_sample=True)

    chatml = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
    print(chatml)

    response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=False)

    return response

messages = [
    {"role": "user",
     "content": """
    INPUT_CONTENT_HERE
     """}
]

result = chat(messages, max_new_tokens=200, temperature=1)
print(result)

Downloads last month: 3

Safetensors

Model size

6B params

Tensor type

F32

F16

Model tree for liemo/open-lilm-v2-q4

Base model

hon9kon9ize/CantoneseLLMChat-v0.5

Finetuned

0xtaipoian/open-lilm-v2

Quantized

(1)

this model