Instructions to use monsoon-nlp/mGPT-13B-quantized with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use monsoon-nlp/mGPT-13B-quantized with Transformers:

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

pipe = pipeline("text-generation", model="monsoon-nlp/mGPT-13B-quantized")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("monsoon-nlp/mGPT-13B-quantized")
model = AutoModelForCausalLM.from_pretrained("monsoon-nlp/mGPT-13B-quantized")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use monsoon-nlp/mGPT-13B-quantized with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "monsoon-nlp/mGPT-13B-quantized"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "monsoon-nlp/mGPT-13B-quantized",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/monsoon-nlp/mGPT-13B-quantized

SGLang

How to use monsoon-nlp/mGPT-13B-quantized 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 "monsoon-nlp/mGPT-13B-quantized" \
    --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": "monsoon-nlp/mGPT-13B-quantized",
		"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 "monsoon-nlp/mGPT-13B-quantized" \
        --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": "monsoon-nlp/mGPT-13B-quantized",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use monsoon-nlp/mGPT-13B-quantized with Docker Model Runner:
```
docker model run hf.co/monsoon-nlp/mGPT-13B-quantized
```

mGPT-quantized

The concept: 8-bit quantized version of mGPT-13B, an LLM released by AI-Forever / Sberbank AI in 2022-2023.

On the GPT scale, it is between the # of parameters for GPT-2 and GPT-3, but comparison is tricky after training on 60+ languages.

My goal is to evaluate this on Hindi and Indonesian tasks, where there are fewer autoregressive language models in this size range.

For English: use a GPT model or LLaMa2-7B

For Arabic: in August 2023 I would recommend the bilingual JAIS model, which is also 13B parameters can be quantized.

In August 2023 AI-Forever added 1.3B-param models for 20+ languages. If your language is Mongolian, for example, it might be better to use mGPT-1.3B-mongol and not this one.

They also have a 1.3B param model for all languages, which I further quantized here: https://huggingface.co/monsoon-nlp/mGPT-quantized

How was the model created?

Quantization of mGPT-13B was done using bitsandbytes library, CoLab Pro with an A100 GPU, and a lot of space on Google Drive.

from transformers import BitsAndBytesConfig, GPT2LMHeadModel

quantization_config = BitsAndBytesConfig(
    load_in_8bit=True,
    bnb_8bit_compute_dtype=torch.bfloat16,
    bnb_8bit_use_double_quant=True,
    bnb_8bit_quant_type="nf4",
)

qmodel = GPT2LMHeadModel.from_pretrained(
    "ai-forever/mGPT-13B",
    load_in_8bit=True,
    torch_dtype=torch.bfloat16,
    quantization_config=quantization_config,
    device_map="auto"
)

qmodel.save_pretrained("model_name")

Future steps

mGPT could be further quantized (4-bit), but model.save_pretrained() currently throws a NotImplementedError error.

Downloads last month: 5