Instructions to use beyoru/KAT-Coder-V2.5-Dev-VL with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use beyoru/KAT-Coder-V2.5-Dev-VL with Transformers:

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

pipe = pipeline("image-text-to-text", model="beyoru/KAT-Coder-V2.5-Dev-VL")
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, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("beyoru/KAT-Coder-V2.5-Dev-VL")
model = AutoModelForMultimodalLM.from_pretrained("beyoru/KAT-Coder-V2.5-Dev-VL", device_map="auto")
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 Settings

vLLM

How to use beyoru/KAT-Coder-V2.5-Dev-VL with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "beyoru/KAT-Coder-V2.5-Dev-VL"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "beyoru/KAT-Coder-V2.5-Dev-VL",
		"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/beyoru/KAT-Coder-V2.5-Dev-VL

SGLang

How to use beyoru/KAT-Coder-V2.5-Dev-VL 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 "beyoru/KAT-Coder-V2.5-Dev-VL" \
    --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": "beyoru/KAT-Coder-V2.5-Dev-VL",
		"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 "beyoru/KAT-Coder-V2.5-Dev-VL" \
        --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": "beyoru/KAT-Coder-V2.5-Dev-VL",
		"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 beyoru/KAT-Coder-V2.5-Dev-VL with Docker Model Runner:
```
docker model run hf.co/beyoru/KAT-Coder-V2.5-Dev-VL
```

KAT-Coder-V2.5-Dev-VL

File size: 2,254 Bytes

# vLLM — cau hinh DA KIEM CHUNG cho beyoru/KAT-Coder-V2.5-Dev-VL (bf16).
#
# Da chay that: vllm/vllm-openai:v0.26.0, 1x H200 143GB, TP1, ~2 phut nap nguoi (14 shard,
# 65.4 GiB), FlashAttention 3 cho LM + FLASH_ATTN cho ViT, MoE backend TRITON tu chon.
#
# ⚠️ KHONG truyen --language-model-only. Co do danh cho ban text-only cua Kwaipilot;
#    truyen vao day se tat vision tower va ban quay ve dung model goc.
services:
  kat-vl:
    image: vllm/vllm-openai:v0.26.0
    container_name: kat-vl
    ipc: host
    ports:
      - "8000:8000"
    volumes:
      # Doi sang thu muc chua checkpoint cua ban, hoac bo volume va dung thang
      # --model beyoru/KAT-Coder-V2.5-Dev-VL de vLLM tu tai ve HF cache.
      - /path/to/models:/models
      - ${HOME}/.cache/huggingface:/root/.cache/huggingface
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              device_ids: ["0"]
              capabilities: [gpu]
    command: >
      --model /models/KAT-Coder-V2.5-Dev-VL
      --served-model-name kat-vl
      --max-model-len 32768
      --max-num-batched-tokens 8192
      --gpu-memory-utilization 0.90
      --trust-remote-code

# Bo nho: weight bf16 la 70.2 GB. Mot card 80 GB KHONG du cho ca weight lan KV cache
# dung duoc. Dung mot card 141/143 GB, hoac them --tensor-parallel-size 2 tren 2x80 GB.
#
# Context: native 262144. Chi nang --max-model-len sau khi da kiem bo nho con lai;
# KV cache o context day rat lon. Muon vuot native thi ap YaRN y het huong dan cua
# Qwen/Qwen3.6-35B-A3B (tham so RoPE o day khong doi so voi upstream) — luu y moi
# framework deu cai YaRN TINH nen bat len se lam giam chat luong prompt ngan.

# ⚠️ `--max-num-batched-tokens 8192` la BAT BUOC, dung bo di.
#
# Kien truc nay lai attention + Gated DeltaNet. Khi prefix caching bat (mac dinh o
# vLLM moi), vLLM ep `mamba_cache_mode='align'`, roi nang attention block size len
# 2096 de "attention page size >= mamba page size". Mac dinh max_num_batched_tokens
# chi 2048 => assert vo ngay luc khoi tao KV cache:
#
#   AssertionError: In Mamba cache align mode,
#                   block_size (2096) must be <= max_num_batched_tokens (2048)
#
# Bat ky gia tri nao >= 2096 deu qua; 8192 con loi cho prefill.