Instructions to use KamilaMila/FastVLM-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use KamilaMila/FastVLM-7B with Transformers:

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

pipe = pipeline("image-text-to-text", model="KamilaMila/FastVLM-7B")
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 AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("KamilaMila/FastVLM-7B")
model = AutoModelForMultimodalLM.from_pretrained("KamilaMila/FastVLM-7B", 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 = 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 KamilaMila/FastVLM-7B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KamilaMila/FastVLM-7B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KamilaMila/FastVLM-7B",
		"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/KamilaMila/FastVLM-7B

SGLang

How to use KamilaMila/FastVLM-7B 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 "KamilaMila/FastVLM-7B" \
    --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": "KamilaMila/FastVLM-7B",
		"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 "KamilaMila/FastVLM-7B" \
        --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": "KamilaMila/FastVLM-7B",
		"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 KamilaMila/FastVLM-7B with Docker Model Runner:
```
docker model run hf.co/KamilaMila/FastVLM-7B
```

FastVLM-7B

File size: 2,829 Bytes

{
  "architectures": [
    "FastVlmForConditionalGeneration"
  ],
  "dtype": "bfloat16",
  "image_seq_length": 256,
  "image_token_index": 151646,
  "model_type": "fast_vlm",
  "multimodal_projector_bias": true,
  "projector_hidden_act": "gelu",
  "text_config": {
    "_name_or_path": "Qwen/Qwen2-7B",
    "architectures": [
      "Qwen2ForCausalLM"
    ],
    "attention_dropout": 0.0,
    "bos_token_id": 151643,
    "dtype": "bfloat16",
    "eos_token_id": 151643,
    "hidden_act": "silu",
    "hidden_size": 3584,
    "initializer_range": 0.02,
    "intermediate_size": 18944,
    "layer_types": [
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention"
    ],
    "max_position_embeddings": 131072,
    "max_window_layers": 28,
    "model_type": "qwen2",
    "num_attention_heads": 28,
    "num_hidden_layers": 28,
    "num_key_value_heads": 4,
    "rms_norm_eps": 1e-06,
    "rope_scaling": null,
    "rope_theta": 1000000.0,
    "sliding_window": null,
    "use_cache": true,
    "use_sliding_window": false,
    "vocab_size": 152128
  },
  "transformers_version": "4.56.0.dev0",
  "vision_config": {
    "_name_or_path": "timm/fastvit_mci3.apple_mclip2_dfndr2b",
    "architecture": "fastvit_mci3",
    "do_pooling": true,
    "global_pool": "avg",
    "hidden_size": 3072,
    "initializer_range": 0.02,
    "label_names": [],
    "model_args": {
      "inference_mode": true
    },
    "model_type": "timm_wrapper",
    "num_classes": 0,
    "num_features": 3072,
    "pretrained_cfg": {
      "classifier": "head.fc",
      "crop_mode": "center",
      "crop_pct": 0.95,
      "custom_load": false,
      "first_conv": [
        "stem.0.conv_kxk.0.conv",
        "stem.0.conv_scale.conv"
      ],
      "fixed_input_size": false,
      "input_size": [
        3,
        256,
        256
      ],
      "interpolation": "bicubic",
      "license": "apple-amlr",
      "mean": [
        0.48145466,
        0.4578275,
        0.40821073
      ],
      "pool_size": [
        8,
        8
      ],
      "std": [
        0.26862954,
        0.26130258,
        0.27577711
      ],
      "tag": "apple_mclip2_dfndr2b"
    }
  },
  "vision_feature_layer": -1,
  "vision_feature_select_strategy": "full"
}