Instructions to use FallenMerick/MN-Violet-Lotus-12B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use FallenMerick/MN-Violet-Lotus-12B with Transformers:

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

pipe = pipeline("text-generation", model="FallenMerick/MN-Violet-Lotus-12B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("FallenMerick/MN-Violet-Lotus-12B")
model = AutoModelForCausalLM.from_pretrained("FallenMerick/MN-Violet-Lotus-12B")
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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use FallenMerick/MN-Violet-Lotus-12B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "FallenMerick/MN-Violet-Lotus-12B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "FallenMerick/MN-Violet-Lotus-12B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/FallenMerick/MN-Violet-Lotus-12B

SGLang

How to use FallenMerick/MN-Violet-Lotus-12B 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 "FallenMerick/MN-Violet-Lotus-12B" \
    --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": "FallenMerick/MN-Violet-Lotus-12B",
		"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 "FallenMerick/MN-Violet-Lotus-12B" \
        --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": "FallenMerick/MN-Violet-Lotus-12B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use FallenMerick/MN-Violet-Lotus-12B with Docker Model Runner:
```
docker model run hf.co/FallenMerick/MN-Violet-Lotus-12B
```

Feedback

by EloyOn - opened Nov 23, 2024

Discussion

EloyOn

Nov 23, 2024

This is an interesting merge, I'm liking it more, the more I use it. The models used are all good in their own way.

It dulled the charming personality of Violet Twilight, but in exchange the messages are two to three paragraphs, which is perfect.

I'm using ChatML, so far so good. How is the world knowledge of this model, did you keep a good percentage? It's great to be able to get mostly accurate info when asking he AI about various topics.

FallenMerick

Owner Nov 23, 2024

Glad to hear you're liking the model! I think dulling the personality from Violet Twilight is an unfortunate downside of SLERPing it into the recipe, though I also find the tradeoff of even higher EQ and output quality to be worth it.

I couldn't comment at all on the world knowledge, as that isn't something I test for in any of the original models or in my resulting merges. My testing focus is on roleplaying capability and writing quality.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment