Instructions to use unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit with Transformers:

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

pipe = pipeline("image-text-to-text", model="unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit")
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("unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit")
model = AutoModelForMultimodalLM.from_pretrained("unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit", 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 unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit",
		"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/unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit

SGLang

How to use unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit 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 "unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit" \
    --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": "unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit",
		"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 "unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit" \
        --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": "unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit",
		"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 unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit with Docker Model Runner:
```
docker model run hf.co/unsloth/Llama-4-Scout-17B-16E-Instruct-unsloth-bnb-4bit
```

danielhanchen commited on Apr 6, 2025

Commit

df0ad74

1 Parent(s): dbaaceb

Add files using upload-large-folder tool

Browse files

Files changed (4) hide show

config.json +2 -10
generation_config.json +1 -1
model.safetensors.index.json +0 -0
tokenizer_config.json +1 -4

config.json CHANGED Viewed

@@ -3,12 +3,9 @@
     "Llama4ForConditionalGeneration"
   ],
   "boi_token_index": 200080,
-  "bos_token_id": 200000,
   "eoi_token_index": 200081,
-  "eos_token_id": 200008,
   "image_token_index": 200092,
   "model_type": "llama4",
-  "pad_token_id": 201134,
   "quantization_config": {
     "_load_in_4bit": true,
     "_load_in_8bit": false,
@@ -19,12 +16,8 @@
     "llm_int8_enable_fp32_cpu_offload": false,
     "llm_int8_has_fp16_weight": false,
     "llm_int8_skip_modules": [
-      "lm_head",
-      "multi_modal_projector",
-      "merger",
-      "modality_projection",
-      "vision_model.model.layers.33.mlp",
-      "language_model.model.layers.43.self_attn"
     ],
     "llm_int8_threshold": 6.0,
     "load_in_4bit": true,
@@ -181,7 +174,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.51.0",
-  "unsloth_fixed": true,
   "vision_config": {
     "_attn_implementation_autoset": true,
     "attention_dropout": 0.0,

     "Llama4ForConditionalGeneration"
   ],
   "boi_token_index": 200080,
   "eoi_token_index": 200081,
   "image_token_index": 200092,
   "model_type": "llama4",
   "quantization_config": {
     "_load_in_4bit": true,
     "_load_in_8bit": false,
     "llm_int8_enable_fp32_cpu_offload": false,
     "llm_int8_has_fp16_weight": false,
     "llm_int8_skip_modules": [
+      "vision_model",
+      "multi_modal_projector"
     ],
     "llm_int8_threshold": 6.0,
     "load_in_4bit": true,
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.51.0",
   "vision_config": {
     "_attn_implementation_autoset": true,
     "attention_dropout": 0.0,

generation_config.json CHANGED Viewed

@@ -6,7 +6,7 @@
     200007,
     200008
   ],
-  "pad_token_id": 201134,
   "temperature": 0.6,
   "top_p": 0.9,
   "transformers_version": "4.51.0"

     200007,
     200008
   ],
+  "pad_token_id": 200018,
   "temperature": 0.6,
   "top_p": 0.9,
   "transformers_version": "4.51.0"

model.safetensors.index.json CHANGED Viewed

The diff for this file is too large to render. See raw diff

tokenizer_config.json CHANGED Viewed

@@ -1,5 +1,4 @@
 {
-  "add_bos_token": true,
   "added_tokens_decoder": {
     "200000": {
       "content": "<|begin_of_text|>",
@@ -9093,8 +9092,6 @@
   ],
   "model_max_length": 10485760,
   "pad_token": "<|finetune_right_pad_id|>",
-  "padding_side": "left",
   "processor_class": "Llama4Processor",
-  "tokenizer_class": "PreTrainedTokenizer",
-  "unk_token": null
 }

 {
   "added_tokens_decoder": {
     "200000": {
       "content": "<|begin_of_text|>",
   ],
   "model_max_length": 10485760,
   "pad_token": "<|finetune_right_pad_id|>",
   "processor_class": "Llama4Processor",
+  "tokenizer_class": "PreTrainedTokenizer"
 }