Instructions to use meta-llama/Llama-3.2-3B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use meta-llama/Llama-3.2-3B-Instruct with Transformers:

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

pipe = pipeline("text-generation", model="meta-llama/Llama-3.2-3B-Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-3.2-3B-Instruct")
model = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-3.2-3B-Instruct", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use meta-llama/Llama-3.2-3B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "meta-llama/Llama-3.2-3B-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "meta-llama/Llama-3.2-3B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/meta-llama/Llama-3.2-3B-Instruct

SGLang

How to use meta-llama/Llama-3.2-3B-Instruct 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 "meta-llama/Llama-3.2-3B-Instruct" \
    --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": "meta-llama/Llama-3.2-3B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "meta-llama/Llama-3.2-3B-Instruct" \
        --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": "meta-llama/Llama-3.2-3B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use meta-llama/Llama-3.2-3B-Instruct with Docker Model Runner:
```
docker model run hf.co/meta-llama/Llama-3.2-3B-Instruct
```

Why does this model have no biases?

#48

by Inoob - opened Dec 22, 2024

Discussion

Inoob

Dec 22, 2024

I used this code to extract the weight and biases of the model:

from transformers import AutoModelForCausalLM

model_name = "./Llama-3.2-3b-Instruct"
model = AutoModelForCausalLM.from_pretrained(model_name)
model_weights = model.state_dict()
weights = {}
biases = {}

for key, value in model_weights.items():
    if 'weight' in key:
        weights[key] = value
    else:
        biases[key] = value

print("Weights:", weights)
print("Biases:", biases)

However, the biases dict is empty.

Is this intentional to not have biases?

Jeffsimpsons

Dec 26, 2024

I was working on this today, and noticed the same thing. Apparently there are no biases because the model worked better without them, so it is intentional. I'm trying to duplicate the model in pytorch to finetune and checked the parameters using:
from safetensors.torch import load_file
SAFE_TENSORS_PATH = "model.safetensors"
safetensors = load_file(SAFE_TENSORS_PATH)
for key, tensor in safetensors.items():
print(f"{key}: {tensor.shape}")
The parameter count of these tensors matches the model card "size" exactly with no biases only weights.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment