Instructions to use Tsunami-th/Tsunami-0.5x-7B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Tsunami-th/Tsunami-0.5x-7B-Instruct with Transformers:

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

pipe = pipeline("text-generation", model="Tsunami-th/Tsunami-0.5x-7B-Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Tsunami-th/Tsunami-0.5x-7B-Instruct")
model = AutoModelForCausalLM.from_pretrained("Tsunami-th/Tsunami-0.5x-7B-Instruct")
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

vLLM

How to use Tsunami-th/Tsunami-0.5x-7B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Tsunami-th/Tsunami-0.5x-7B-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": "Tsunami-th/Tsunami-0.5x-7B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Tsunami-th/Tsunami-0.5x-7B-Instruct

SGLang

How to use Tsunami-th/Tsunami-0.5x-7B-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 "Tsunami-th/Tsunami-0.5x-7B-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": "Tsunami-th/Tsunami-0.5x-7B-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 "Tsunami-th/Tsunami-0.5x-7B-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": "Tsunami-th/Tsunami-0.5x-7B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Tsunami-th/Tsunami-0.5x-7B-Instruct with Docker Model Runner:
```
docker model run hf.co/Tsunami-th/Tsunami-0.5x-7B-Instruct
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Tsunami-0.5x-7B-Instruct

TSUNAMI: Transformative Semantic Understanding and Natural Augmentation Model for Intelligence.

TSUNAMI full name was created by ChatGPT.

infomation

Tsunami-0.5x-7B-Instruct is Thai Large Language Model that fine-tuned from Qwen2.5-7B around 100,000 rows in Thai dataset.

Prompt Template

This model uses ChatML prompt template:

<|im_start|>system
{System}<|im_end|>
<|im_start|>user
{User}<|im_end|>
<|im_start|>assistant
{Assistant}

How to use


from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_name = "Tsunami-th/Tsunami-0.5x-7B-Instruct"

model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)

messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "สวัสดีครับ"}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)

inputs = tokenizer(text, return_tensors="pt")
inputs = inputs.to(model.device)
with torch.no_grad():
   output = model.generate(**inputs, max_new_tokens=512)

response = tokenizer.decode(output[0, len(inputs['input_ids'][0]):], skip_special_tokens=True)

Author

Pollakrit Lorprasertkul | game.pollakrit@gmail.com

Tsunami-0.5x-7B-Instruct is the version 0.5x that did not train on the whole dataset.
Tsunami-1.0-7B-Instruct is coming soon.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	29.80
IFEval (0-Shot)	70.99
BBH (3-Shot)	37.36
MATH Lvl 5 (4-Shot)	4.83
GPQA (0-shot)	8.61
MuSR (0-shot)	18.57
MMLU-PRO (5-shot)	38.42

Downloads last month: 16

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for Tsunami-th/Tsunami-0.5x-7B-Instruct

Base model

Qwen/Qwen2.5-7B

Finetuned

(957)

this model

Merges

18 models

Quantizations

3 models

Evaluation results

strict accuracy on IFEval (0-Shot)
Open LLM Leaderboard

70.990
normalized accuracy on BBH (3-Shot)
Open LLM Leaderboard

37.360
exact match on MATH Lvl 5 (4-Shot)
Open LLM Leaderboard

4.830
acc_norm on GPQA (0-shot)
Open LLM Leaderboard

8.610
acc_norm on MuSR (0-shot)
Open LLM Leaderboard

18.570
accuracy on MMLU-PRO (5-shot)
test set Open LLM Leaderboard

38.420