Instructions to use Mewband12/seallm-7b-v2-chat-th with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Mewband12/seallm-7b-v2-chat-th with Transformers:

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

pipe = pipeline("text-generation", model="Mewband12/seallm-7b-v2-chat-th")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Mewband12/seallm-7b-v2-chat-th")
model = AutoModelForCausalLM.from_pretrained("Mewband12/seallm-7b-v2-chat-th", device_map="auto")
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]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Mewband12/seallm-7b-v2-chat-th with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Mewband12/seallm-7b-v2-chat-th"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Mewband12/seallm-7b-v2-chat-th",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Mewband12/seallm-7b-v2-chat-th

SGLang

How to use Mewband12/seallm-7b-v2-chat-th 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 "Mewband12/seallm-7b-v2-chat-th" \
    --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": "Mewband12/seallm-7b-v2-chat-th",
		"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 "Mewband12/seallm-7b-v2-chat-th" \
        --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": "Mewband12/seallm-7b-v2-chat-th",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Mewband12/seallm-7b-v2-chat-th with Docker Model Runner:
```
docker model run hf.co/Mewband12/seallm-7b-v2-chat-th
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Template

<|im_start|>system
You are a helpful assistant. You will only answer based on provided document if relevant to users question.
<|im_start|>user
คุณจะได้รับรายละเอียดเอกสาร และคำถามจากผู้ใช้ใน ''' ด้านล่าง
'''
DOCUMENT: {reference}
QUESTION: {question}
'''
ตอบคำถาม ใช้รายละเอียดเกี่ยวกับปัญหาทั่วไปจาก DOCUMENT หากคำถามไม่เกี่ยวข้องกับ DOCUMENT ที่ให้ไว้ โปรดตอบเท่านั้น
“ฉันไม่ทราบคำตอบ ไม่ได้เป็นส่วนหนึ่งของเอกสาร DOCUMENT ที่ให้มา”
ห้ามใช้ข้อมูลใดๆ นอกเหนือ DOCUMENT ที่ให้ไว้</s>
<|im_start|>user
QUESTION: {question} </s><|im_start|>assistant
ANSWER: {answer}</s>

Step	Training Loss	Validation Loss
1000	0.978900	0.980906
2000	0.699000	0.612407
3000	0.478000	0.490171
4000	0.362000	0.437152
5000	0.333700	0.413056
6000	0.333200	0.393786
7000	0.324200	0.379746
8000	0.297500	0.372649

QLORA FT using thai dataset ~ 4K on RAG instruction. Fine tuned on https://huggingface.co/SeaLLMs/SeaLLM-7B-v2

Purpose: To help LLM avoid to answer what's not in the document better (reduce hallucination for RAG)

Downloads last month: 12

Safetensors

Model size

7B params

Tensor type

F16

Model tree for Mewband12/seallm-7b-v2-chat-th

Quantizations

1 model