Instructions to use nothingiisreal/MN-12B-Celeste-V1.9 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nothingiisreal/MN-12B-Celeste-V1.9 with Transformers:

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

pipe = pipeline("text-generation", model="nothingiisreal/MN-12B-Celeste-V1.9")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("nothingiisreal/MN-12B-Celeste-V1.9")
model = AutoModelForCausalLM.from_pretrained("nothingiisreal/MN-12B-Celeste-V1.9")
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 Settings

vLLM

How to use nothingiisreal/MN-12B-Celeste-V1.9 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "nothingiisreal/MN-12B-Celeste-V1.9"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nothingiisreal/MN-12B-Celeste-V1.9",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/nothingiisreal/MN-12B-Celeste-V1.9

SGLang

How to use nothingiisreal/MN-12B-Celeste-V1.9 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 "nothingiisreal/MN-12B-Celeste-V1.9" \
    --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": "nothingiisreal/MN-12B-Celeste-V1.9",
		"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 "nothingiisreal/MN-12B-Celeste-V1.9" \
        --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": "nothingiisreal/MN-12B-Celeste-V1.9",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use nothingiisreal/MN-12B-Celeste-V1.9 with Docker Model Runner:
```
docker model run hf.co/nothingiisreal/MN-12B-Celeste-V1.9
```

<|im_end|> token at the end of every message

by ProjectXMP - opened Aug 1, 2024

Discussion

ProjectXMP

Aug 1, 2024

The token is at the end of every single assistant response:

mrjackspade

Aug 1, 2024

Thats one of the ChatML tokens referenced in the text:

It's also trained with ChatML tokens so there should be no EOS bleeding whatsoever.

You can set it as a stop token, but I've found the model performs better for me just using the ChatML template now

ProjectXMP

Aug 1, 2024

I have the chatML in place in VLLM, and the token still bleeds through.

AuriAetherwiing

Aug 1, 2024

Hm, that's quite weird, model was tested extensively with vLLM, and special token bleeding never happened. It was tested with text completion, not chat completion though, but jinja2 template in tokenizer_config.json does account for <|im_end|> as it should.

sheliak

Aug 3, 2024

•

edited Aug 3, 2024

I can confirm that I am experiencing this as well when using KoboldCPP and the built in ChatML template.

djuna

Aug 3, 2024

I think we should include 15 in eos token in the config file?

Memphisto

Aug 14, 2024

i had the same problem.
the fix for me was to activate "skip special tokens" in SillyTavern.
I have no idea what i am doing or what other effects this has.
But after i activated that option the Token is gone.

aaronday3

Nothing is Real org Aug 20, 2024

Ah that might just do it @Memphisto I had it on this entire time

Luni

Aug 27, 2024

This is due to a fault in the training. They did not add the token for the chatml to the special tokens, for whatever reason. So during training it kept spitting out <|im_end|> and other as strings rather as the EOS etc. token
Hopefully they will learn from their mistake because this model as a partial merge is a fine contribution

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