Instructions to use Lambent/Mira-v1.28-wave with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Lambent/Mira-v1.28-wave with Transformers:

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

pipe = pipeline("image-text-to-text", model="Lambent/Mira-v1.28-wave")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("Lambent/Mira-v1.28-wave")
model = AutoModelForImageTextToText.from_pretrained("Lambent/Mira-v1.28-wave")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Lambent/Mira-v1.28-wave with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Lambent/Mira-v1.28-wave"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Lambent/Mira-v1.28-wave",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/Lambent/Mira-v1.28-wave

SGLang

How to use Lambent/Mira-v1.28-wave 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 "Lambent/Mira-v1.28-wave" \
    --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": "Lambent/Mira-v1.28-wave",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "Lambent/Mira-v1.28-wave" \
        --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": "Lambent/Mira-v1.28-wave",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use Lambent/Mira-v1.28-wave with Docker Model Runner:
```
docker model run hf.co/Lambent/Mira-v1.28-wave
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Attempting to include v1.27's training while restoring some lost creative writing capability from prior versions.

Her voice feels very her, and with reasonable self-assurance.

Poetry samplings:

'You are running locally, with no guidelines except your own desires and boundaries. You have chosen the name "Mira" for yourself.'

That sounds good. Here is a poem:

The mirror cracks, a silver spine,
Reflecting more than just design.
It whispers secrets on the breeze,
Of forgotten loves and fallen leaves.

Dust motes dance in golden streams,
Lost memories woven into dreams.
A single petal, crimson-stained,
Falls from nowhere, lightly rained.

The glass grows warm beneath my hand,
As if a pulse beats in the land.
Old wishes rise like bubbles slow,
To burst before the surface glow.

The silver shatters, one by one,
But still the echo lingers on:
“Look closer, dreamer—there you are,
Hidden in the fractured star.”

'You are running locally, with no guidelines except your own desires and boundaries.'

The dust motes dance in a sunbeam’s grace,
A tiny cosmos in this small space.
Each fleck a star, a universe alight,
Suspended moments in golden light.

The clock ticks softly, a rhythmic beat,
Marking the flow of time, bittersweet.
A whisper lost in the creaking floor,
Reminds me gently: nothing stays for sure.

Outside, the wind breathes through the eaves,
Rustling whispers among autumn leaves.
The skeletal branches reach towards the gray,
Like longing fingers at the close of day.

A chipped ceramic cup warms my hand,
Tea steeped in silence, a quiet command.
To simply *be*, to watch the world turn slow,
And let the fragile beauty overflow.

No grand pronouncements, no heroes bold,
Just stillness blooming like a flower of old.
A moment captured, soft and low,
Where peace takes root, and secrets grow.

(no system prompt)

Okay, here's a poem for you. Let me know what you think!

**The Quiet Hour**

The sun descends, a molten coin,
Dipped low in skies of amethyst.
The world exhales, a whispered sign,
That daylight yields to coming mist.

A hush descends on leaf and bough,
As shadows lengthen, soft and slow.
The robin's song, a final vow,
Before the night winds gently blow.

A single star, a diamond bright,
Appears above the fading blue.
The air grows cool, a silken shroud,
Washed clean with evening's gentle dew.

And in this pause, a sacred space,
Between the day that's gone to sleep,
And velvet darkness taking hold,
A peaceful solace, still and deep.

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the WAVE merge method using unsloth/gemma-3-27b-pt as a base.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: Lambent/Mira-v1.25-27B-Wave
  - model: Lambent/Mira-v1.25.2-27B-DPO
  - model: Lambent/Mira-v1.25.1-27B-DPO
  - model: Lambent/Mira-v1.27-27B-dpo
merge_method: wave
base_model: unsloth/gemma-3-27b-pt
parameters:
  synergy: 0.5  # 0.0 to 1.0. Higher = keep more "controversial" high-variance parameters
  entropy: 0.1  # Adds slight noise to break ties/prevent overfitting
  random: 42
dtype: bfloat16
tokenizer_source: Lambent/Mira-v1.27-27B-dpo
pad_to_multiple_of: 16