Instructions to use BakeLab/Kallisti-35B-A3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use BakeLab/Kallisti-35B-A3B with Transformers:

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

pipe = pipeline("image-text-to-text", model="BakeLab/Kallisti-35B-A3B")
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("BakeLab/Kallisti-35B-A3B")
model = AutoModelForImageTextToText.from_pretrained("BakeLab/Kallisti-35B-A3B")
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 BakeLab/Kallisti-35B-A3B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "BakeLab/Kallisti-35B-A3B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "BakeLab/Kallisti-35B-A3B",
		"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/BakeLab/Kallisti-35B-A3B

SGLang

How to use BakeLab/Kallisti-35B-A3B 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 "BakeLab/Kallisti-35B-A3B" \
    --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": "BakeLab/Kallisti-35B-A3B",
		"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 "BakeLab/Kallisti-35B-A3B" \
        --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": "BakeLab/Kallisti-35B-A3B",
		"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 BakeLab/Kallisti-35B-A3B with Docker Model Runner:
```
docker model run hf.co/BakeLab/Kallisti-35B-A3B
```

Kallisti-35B-A3B

File size: 1,778 Bytes

aed26a9
 
 
 
 
 
1f306bf
aed26a9
 
cc12a45
aed26a9
 
 
 
 
 
 
cc12a45
aed26a9
cc12a45
aed26a9
cc12a45
10c9ae1
cc12a45
d1f12b7
 
aed26a9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cc12a45

---
library_name: transformers
license: apache-2.0
base_model: Qwen/Qwen3.5-35B-A3B
tags:
- llama-factory
- full
- generated_from_trainer
model-index:
- name: Kallisti-35B-A3B
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->


# 🍎 Kallisti-35B-A3B

This model is a fine-tuned version of [Qwen/Qwen3.5-35B-A3B](https://huggingface.co/Qwen/Qwen3.5-35B-A3B).

- 🌐 [Project Website](https://vab.bakelab.ai/) - Learn more about Visual Aesthetic Benchmark
- 📖 [Technical Report](https://arxiv.org/abs/2605.12684) - Discover technical details behind VAB
- 🔧 [GitHub Repo](https://github.com/BakeLab/Visual-Aesthetic-Benchmark) - Evaluation scripts and benchmark tooling
- 🤗 [Visual Aesthetic Benchmark](https://huggingface.co/datasets/BakeLab/Visual-Aesthetic-Benchmark) - HF Datasets
- 🤗 [Kallisti-35B-A3B](https://huggingface.co/BakeLab/Kallisti-35B-A3B) - Finetune model [📍| You are here!]

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 1e-06
- train_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- total_train_batch_size: 32
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 10.0
- num_epochs: 3.0

### Framework versions

- Transformers 5.2.0
- Pytorch 2.10.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2

## License

[Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0)

## Contact

Please contact [Yichen](mailto:yfeng42@uw.edu) by email.