Instructions to use sbintuitions/sarashina2-13b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use sbintuitions/sarashina2-13b with Transformers:

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

pipe = pipeline("text-generation", model="sbintuitions/sarashina2-13b")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("sbintuitions/sarashina2-13b")
model = AutoModelForCausalLM.from_pretrained("sbintuitions/sarashina2-13b")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use sbintuitions/sarashina2-13b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "sbintuitions/sarashina2-13b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sbintuitions/sarashina2-13b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/sbintuitions/sarashina2-13b

SGLang

How to use sbintuitions/sarashina2-13b 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 "sbintuitions/sarashina2-13b" \
    --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": "sbintuitions/sarashina2-13b",
		"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 "sbintuitions/sarashina2-13b" \
        --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": "sbintuitions/sarashina2-13b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use sbintuitions/sarashina2-13b with Docker Model Runner:
```
docker model run hf.co/sbintuitions/sarashina2-13b
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Sarashina2-13B

This repository provides large language models trained by SB Intuitions.

How to use

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline, set_seed
 
model = AutoModelForCausalLM.from_pretrained("sbintuitions/sarashina2-13b", torch_dtype=torch.bfloat16, device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("sbintuitions/sarashina2-13b")
# If you want to use slow tokenizer
# tokenizer = AutoTokenizer.from_pretrained("sbintuitions/sarashina2-13b", use_fast=False）
generator = pipeline("text-generation", model=model, tokenizer=tokenizer)
set_seed(123)
 
text = generator(
    "おはようございます、今日の天気は",
    max_length=30,
    do_sample=True,
    pad_token_id=tokenizer.pad_token_id,
    num_return_sequences=3,
)

for t in text:
  print(t)

Configuration

Parameters	Vocab size	Training tokens	Architecture	Position type	Layers	Hidden dim	Attention heads
7B	102400	2.1T	Llama2	RoPE	32	4096	32
13B	102400	2.1T	Llama2	RoPE	40	5120	40
70B	102400	2.1T	Llama2	RoPE	80	8192	64

Training Corpus

For our Japanese training data, we used a Japanese portion of the Common Crawl corpus, which is the largest Web corpus, as our training dataset. To clean the training corpus, we used CCNet and HojiChar. After cleaning, our Japanese training data contains about 1T tokens.

For our English training data, we extracted English documents from SlimPajama but we removed books3 corpus due to copyright infringement.

Tokenization

We use a sentencepiece tokenizer with a unigram language model and byte-fallback. We do not apply pre-tokenization with Japanese tokenizer. Thus, a user may directly feed raw sentences into the tokenizer.

Ethical Considerations and Limitations

Sarashina2 has not been tuned to follow an instruction yet. Therefore, sarashina2 might generate some meaningless sequences, some inaccurate instances or biased/objectionable outputs. Before using sarashina2, we would like developers to tune models based on human preferences and safety considerations.