Instructions to use maldv/Lytta2.5-32B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use maldv/Lytta2.5-32B-Instruct with Transformers:

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

pipe = pipeline("text-generation", model="maldv/Lytta2.5-32B-Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("maldv/Lytta2.5-32B-Instruct")
model = AutoModelForCausalLM.from_pretrained("maldv/Lytta2.5-32B-Instruct")
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 maldv/Lytta2.5-32B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "maldv/Lytta2.5-32B-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "maldv/Lytta2.5-32B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/maldv/Lytta2.5-32B-Instruct

SGLang

How to use maldv/Lytta2.5-32B-Instruct 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 "maldv/Lytta2.5-32B-Instruct" \
    --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": "maldv/Lytta2.5-32B-Instruct",
		"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 "maldv/Lytta2.5-32B-Instruct" \
        --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": "maldv/Lytta2.5-32B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use maldv/Lytta2.5-32B-Instruct with Docker Model Runner:
```
docker model run hf.co/maldv/Lytta2.5-32B-Instruct
```

It's creative writing and prompt following

by SubtleOne - opened Jan 3, 2025

Discussion

SubtleOne

Jan 3, 2025

•

edited Jan 3, 2025

So first of all, I'll say that for creative writing and regular content such as for DnD or other, Qwentile has been my favorite and go-to. It keeps track of characters, plotlines, and generates detailed complex ones with aplomb. It also writes very well. Lytta writes great too, so shares that DNA, and is very creative too. However, at a price. It seriously ignores significant parts of a query and goes out on a limb to produce its own ideas. These ideas are not bad by any means, but it means having many aspects of a longer prompt completely ignored or twisted around. It is a very fun model, so don't get me wrong.

maldv

Owner Jan 3, 2025

Ha, yeah... That was what I noticed as well. It's extremely creative and good at thinking, but it's instruction following ability was completely toasted. I wonder if there is some deeper insight about the relationship between creative thought and obedience here... that may explain why my teachers thought I was disruptive as a student. 😼

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