Llama-3-Typhoon-1.5X-70B-instruct is a 70 billion parameter instruct model designed for Thai 🇹🇭 language. It demonstrates competitive performance with GPT-4-0612, and is optimized for application use cases, Retrieval-Augmented Generation (RAG), constrained generation, and reasoning tasks.

Built on Typhoon 1.5 70B (not yet released) and Llama 3 70B Instruct. this model is a result of our experiment on cross-lingual transfer. It utilizes the task-arithmetic model editing technique, combining the Thai understanding capability of Typhoon with the human alignment performance of Llama 3 Instruct.

Remark: To acknowledge Meta's efforts in creating the foundation model and comply with the license, we explicitly include "llama-3" in the model name.

Model Description

Model type: A 70B instruct decoder-only model based on the Llama architecture
Requirement: vllm (https://pypi.org/project/vllm/) 0.3.2 or newer.
Primary Language(s): Thai 🇹🇭 and English 🇬🇧
License: Llama 3 Community License

Performance

We evaluated the model's performance in Language & Knowledge Capabilities and Instruction Following Capabilities.

Language & Knowledge Capabilities:
- Assessed using multiple-choice question-answering datasets such as ThaiExam and MMLU.
Instruction Following Capabilities:
- Evaluated based on beta users' feedback, focusing on two factors:
  - Human Alignment & Reasoning: Ability to generate responses that are clear and logically structured across multiple steps.
    - Evaluated using MT-Bench — How LLMs can align with human needs.
  - Instruction-following: Ability to adhere to specified constraints in the instructions.
    - Evaluated using IFEval — How LLMs can follow specified constraints, such as formatting and brevity.
Agentic Capabilities:
- Evaluated in agent use-cases using Hugging Face's Transformer Agents and the associated benchmark.

Remark: We developed the Thai (TH) pairs by translating the original datasets into Thai through machine and human methods.

ThaiExam

Model	ONET	IC	TGAT	TPAT-1	A-Level	Average (ThaiExam)	MMLU
Typhoon-1.5X 70B	0.565	0.68	0.778	0.517	0.56	0.620	0.7945
gpt-4-0612	0.493	0.69	0.744	0.509	0.616	0.610	0.864**
---	---	---	---	---	---	---	---
gpt-4o	0.62	0.63	0.789	0.56	0.623	0.644	0.887**

** We report the MMLU score that is reported in GPT-4o Tech Report.

MT-Bench

Model	MT-Bench Thai	MT-Bench English
Typhoon-1.5X 70B	8.029	8.797
gpt-4-0612	7.801	8.671
---	---	---
gpt-4o	8.514	9.184

IFEval

Model	IFEval Thai	IFEval English
Typhoon-1.5X 70B	0.645	0.810
gpt-4-0612	0.612	0.793*
---	---	---
gpt-4o	0.737	0.871

We report the number from IFEval paper.

Agent

Model	GAIA - Thai/English	GSM8K - Thai/English	HotpotQA - Thai/English
gpt-3.5-turbo-0125	18.42/37.5	70/80	39.56/59
Typhoon-1.5X 70B	17.10/36.25	80/95	52.7/65.83
gpt-4-0612	17.10/38.75	90/100	56.41/76.25
---	---	---	---
gpt-4o	44.73/57.5	100/100	71.64/76.58

Insight

We utilized model editing techniques and found that the most critical feature for generating accurate Thai answers is located in the backend (the upper layers of the transformer block). Accordingly, we incorporated a high ratio of Typhoon components in these backend layers to enhance our model’s performance.

Usage Example

python

from transformers import AutoTokenizer
from vllm import LLM, SamplingParams
quant_path = "scb10x/llama-3-typhoon-v1.5x-70b-instruct-awq"
llm = LLM(model=quant_path, quantization='awq', max_model_len=8192)
tokenizer = AutoTokenizer.from_pretrained(quant_path)

messages = [
    // messages here
]
prompts = tokenizer.apply_chat_template(
    messages,
    add_generation_prompt=True,
    tokenize=False
)
sampling_params = SamplingParams(repetition_penalty=1.05, top_p=0.6, temperature=0.9, max_tokens=1024, stop=['<|eot_id|>', '<|start_header_id|>', '<|end_header_id|>'])
outputs = llm.generate(prompts, sampling_params=sampling_params)
print(outputs[0].outputs)

Chat Template

We use the Llama 3 chat template.

{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}

Intended Uses & Limitations

This model is experimental and might not be fully evaluated for all use cases. Developers should assess risks in the context of their specific applications.

https://twitter.com/opentyphoon

Support

https://discord.gg/us5gAYmrxw

SCB 10X Typhoon Team

Kunat Pipatanakul, Potsawee Manakul, Sittipong Sripaisarnmongkol, Natapong Nitarach, Pathomporn Chokchainant, Kasima Tharnpipitchai
If you find Typhoon-1.5X useful for your work, please cite it using:

@article{pipatanakul2023typhoon,
    title={Typhoon: Thai Large Language Models}, 
    author={Kunat Pipatanakul and Phatrasek Jirabovonvisut and Potsawee Manakul and Sittipong Sripaisarnmongkol and Ruangsak Patomwong and Pathomporn Chokchainant and Kasima Tharnpipitchai},
    year={2023},
    journal={arXiv preprint arXiv:2312.13951},
    url={https://arxiv.org/abs/2312.13951}
}

Contact Us

General & Collaboration: kasima@scb10x.com, pathomporn@scb10x.com
Technical: kunat@scb10x.com

typhoon-ai/llama-3-typhoon-v1.5x-70b-instruct-awq