Instructions to use xiaodongguaAIGC/llama-3-debug with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use xiaodongguaAIGC/llama-3-debug with Transformers:

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

pipe = pipeline("text-generation", model="xiaodongguaAIGC/llama-3-debug")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("xiaodongguaAIGC/llama-3-debug")
model = AutoModelForCausalLM.from_pretrained("xiaodongguaAIGC/llama-3-debug")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use xiaodongguaAIGC/llama-3-debug with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "xiaodongguaAIGC/llama-3-debug"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "xiaodongguaAIGC/llama-3-debug",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/xiaodongguaAIGC/llama-3-debug

SGLang

How to use xiaodongguaAIGC/llama-3-debug 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 "xiaodongguaAIGC/llama-3-debug" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "xiaodongguaAIGC/llama-3-debug",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "xiaodongguaAIGC/llama-3-debug" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "xiaodongguaAIGC/llama-3-debug",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use xiaodongguaAIGC/llama-3-debug with Docker Model Runner:
```
docker model run hf.co/xiaodongguaAIGC/llama-3-debug
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

llama-3-debug

This model use for debug, the parameter is random.

It's small only '~32MB' memory size, that is efficent for you to download and debug.

llama-3-debug model config modified as follow

config.intermediate_size = 128
config.hidden_size = 64
config.num_attention_heads = 2
config.num_key_value_heads = 2
config.num_hidden_layers = 1

If you want to load it by this code

from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = 'xiaodongguaAIGC/llama-3-debug'
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16)
tokenizer = AutoTokenizer.from_pretrained(model_name)
print(model)
print(tokenizer)

Downloads last month: 37

Safetensors

Model size

16.5M params

Tensor type

BF16

Model tree for xiaodongguaAIGC/llama-3-debug

Quantizations

1 model