Instructions to use MaziyarPanahi/Smaug-72B-v0.1-AWQ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MaziyarPanahi/Smaug-72B-v0.1-AWQ with Transformers:

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

pipe = pipeline("text-generation", model="MaziyarPanahi/Smaug-72B-v0.1-AWQ")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MaziyarPanahi/Smaug-72B-v0.1-AWQ")
model = AutoModelForCausalLM.from_pretrained("MaziyarPanahi/Smaug-72B-v0.1-AWQ", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use MaziyarPanahi/Smaug-72B-v0.1-AWQ with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MaziyarPanahi/Smaug-72B-v0.1-AWQ"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MaziyarPanahi/Smaug-72B-v0.1-AWQ",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/MaziyarPanahi/Smaug-72B-v0.1-AWQ

SGLang

How to use MaziyarPanahi/Smaug-72B-v0.1-AWQ 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 "MaziyarPanahi/Smaug-72B-v0.1-AWQ" \
    --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": "MaziyarPanahi/Smaug-72B-v0.1-AWQ",
		"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 "MaziyarPanahi/Smaug-72B-v0.1-AWQ" \
        --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": "MaziyarPanahi/Smaug-72B-v0.1-AWQ",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use MaziyarPanahi/Smaug-72B-v0.1-AWQ with Docker Model Runner:
```
docker model run hf.co/MaziyarPanahi/Smaug-72B-v0.1-AWQ
```

Smaug-72B-v0.1-AWQ

File size: 1,019 Bytes

690bb4d

{
  "_name_or_path": "/home/maziyar/.cache/huggingface/hub/models--abacusai--Smaug-72B-v0.1/snapshots/39d5b3d505599ca7adcb171a79b48b0f499fe1ca",
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": true,
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "eos_token_id": 151643,
  "hidden_act": "silu",
  "hidden_size": 8192,
  "initializer_range": 0.02,
  "intermediate_size": 24576,
  "max_position_embeddings": 32768,
  "model_type": "llama",
  "num_attention_heads": 64,
  "num_hidden_layers": 80,
  "num_key_value_heads": 64,
  "pad_token_id": 151643,
  "pretraining_tp": 1,
  "quantization_config": {
    "bits": 4,
    "group_size": 128,
    "modules_to_not_convert": null,
    "quant_method": "awq",
    "version": "gemm",
    "zero_point": true
  },
  "rms_norm_eps": 1e-06,
  "rope_scaling": null,
  "rope_theta": 1000000,
  "seq_length": 32768,
  "tie_word_embeddings": false,
  "torch_dtype": "float16",
  "transformers_version": "4.38.0.dev0",
  "use_cache": true,
  "vocab_size": 152064
}