--- license: apache-2.0 language: - th - en library_name: transformers pipeline_tag: text-generation tags: - openthaigpt - llama --- # ðđð OpenThaiGPT 70b 1.0.0  [More Info](https://openthaigpt.aieat.or.th/) ðđð **OpenThaiGPT 70b Version 1.0.0** is an advanced 70-billion-parameter Thai language chat model based on LLaMA v2. It has been specifically fine-tuned for Thai instructions and enhanced by incorporating over 10,000 of the most commonly used Thai words into the large language model's (LLM) dictionary, significantly boosting its response speed. ## Highlights - **Leading-edge Thai language LLM**, setting new benchmarks by achieving the highest average scores across 9 Thai language evaluations when compared to all other open-source Thai LLMs. - **Support for extended conversations** across multiple turns. - Integration of **Retrieval Augmented Generation (RAG)** for enriched response generation. - **Generation speeds increased by tenfold**, thanks to the addition of 10,000 frequently used Thai words to the model's dictionary. - Built upon a foundation of **more than 65 billion Thai language words** and meticulously fine-tuned with over 1 million Thai instruction examples. - Capable of understanding and processing **input contexts of up to 4096 Thai words**, allowing for detailed and complex instructions. ## Benchmark by Multiple Choices Thai Exams **Please take a look at ``OTG70b (March2024)`` column for this model's evaluation result** | **Exams** | **OTG 7b (Aug 2023)** | **OTG 13b (Dec 2023)** | **OTG 7b (March 2024)** | **OTG 13b (March 2024)** | **OTG 70b (March 2024)** | **SeaLLM 7b v1** | **SeaLLM 7b v2** | **TyphoonGPT 7b** | **SeaLion 7b** | **WanchanGLM 7b** | **Sailor-7B-Chat** | **GPT3.5** | **GPT4** | **Gemini Pro** | **Gemini 1.5** | **Claude 3 Haiku** | **Claude 3 Sonnet** | **Claude 3 Opus** | |----------------------------------|-----------------------|------------------------|-------------------------|--------------------------|--------------------------|------------------|------------------|--------------------|----------------|-------------------|--------------------|------------|----------|----------------|----------------|--------------------|---------------------|-------------------| | **A-Level** | 17.50% | 34.17% | 25.00% | 30.83% | **45.83%** | 18.33% | 34.17% | N/A* | 21.67% | 17.50% | 40.00% | 38.33% | 65.83% | 56.67% | 55.83% | 58.33% | 59.17% | 77.50% | | **TGAT** | 24.00% | 22.00% | 22.00% | 36.00% | **36.00%** | 14.00% | 28.00% | N/A* | 24.00% | 16.00% | 34.00% | 28.00% | 44.00% | 22.00% | 28.00% | 36.00% | 34.00% | 46.00% | | **TPAT1** | 22.50% | 47.50% | 42.50% | 27.50% | **62.50%** | 22.50% | 27.50% | N/A* | 22.50% | 17.50% | 40.00% | 45.00% | 52.50% | 52.50% | 50.00% | 52.50% | 50.00% | 62.50% | | **ic_all_test** | 8.00% | 28.00% | 76.00% | 84.00% | **68.00%** | 16.00% | 28.00% | N/A* | 24.00% | 16.00% | 24.00% | 40.00% | 64.00% | 52.00% | 32.00% | 44.00% | 64.00% | 72.00% | | **facebook_beleble_tha** | 25.00% | 45.00% | 34.50% | 39.50% | **70.00%** | 13.50% | 51.00% | N/A* | 27.00% | 24.50% | 63.00% | 50.00% | 72.50% | 65.00% | 74.00% | 63.50% | 77.00% | 90.00% | | **xcopa_th_200** | 45.00% | 56.50% | 49.50% | 51.50% | **74.50%** | 26.50% | 47.00% | N/A* | 51.50% | 48.50% | 68.50% | 64.00% | 82.00% | 68.00% | 74.00% | 64.00% | 80.00% | 86.00% | | **xnli2.0_tha** | 33.50% | 34.50% | 39.50% | 31.00% | **47.00%** | 21.00% | 43.00% | N/A* | 37.50% | 33.50% | 16.00% | 50.00% | 69.00% | 53.00% | 54.50% | 50.00% | 68.00% | 68.50% | | **ONET M3** | 17.85% | 38.86% | 34.11% | 39.36% | **56.15%** | 15.58% | 23.92% | N/A* | 21.79% | 19.56% | 21.37% | 37.91% | 49.97% | 55.99% | 57.41% | 52.73% | 40.60% | 63.87% | | **ONET M6** | 21.14% | 28.87% | 22.53% | 23.32% | **42.85%** | 15.09% | 19.48% | N/A* | 16.96% | 20.67% | 28.64% | 34.44% | 46.29% | 45.53% | 50.23% | 34.79% | 38.49% | 48.56% | | **Average Score** | 23.83% | 37.27% | 38.40% | 40.33% | **55.87%** | 18.06% | 33.56% | N/A* | 27.44% | 23.75% | 37.28% | 43.07% | 60.68% | 52.30% | 52.89% | 50.65% | 56.81% | 68.32% | \*Waiting for accessing TyphoonGPT Instruct Model. (More benchmark is coming soon) ### Benchmark Configuration - Clearly instruct model to answer by select from possible choices and followed by an explanation. - Zero shot only - Tested on Unseen test set only - Detect a multi-choice answer on (A),(B),(C),(D),(E) at the beginning of the answer (First priority) and at the end of the answer (Second priority) ## Licenses **Source Code**: License Apache Software License 2.0.<br> **Weight**: Research and **Commercial uses**.<br> ## Sponsors <img src="https://cdn-uploads.huggingface.co/production/uploads/5fcd9c426d942eaf4d1ebd30/42d-GioSs4evIdNuMAaPB.png" width="600px"> ## Supports - Official website: https://openthaigpt.aieat.or.th - Facebook page: https://web.facebook.com/groups/openthaigpt - A Discord server for discussion and support [here](https://discord.gg/rUTp6dfVUF) - E-mail: kobkrit@aieat.or.th ## Prompt Format Prompt format is based on Llama2 with a small modification (Adding "###" to specify the context part) ``` <s>[INST] <<SYS> {system_prompt} <</SYS>> {human_turn1}###{context_turn1} [/INST]{assistant_turn1}</s><s>{human_turn2}###{context_turn2} [/INST] ... ``` Practically, when usually used "\n" for a new line so, ``` <s>[INST] <<SYS>\n{system_prompt}\n<</SYS>>\n\n{human_turn1}###{context_turn1} [/INST]{assistant_turn1}</s><s>{human_turn2}###{context_turn2} [/INST] ... ``` ### System prompt: ``` You are a question answering assistant. Answer the question as truthful and helpful as possible āļāļļāļāļāļ·āļāļāļđāđāļāđāļ§āļĒāļāļāļāļāļģāļāļēāļĄ āļāļāļāļāļāļāļģāļāļēāļĄāļāļĒāđāļēāļāļāļđāļāļāđāļāļāđāļĨāļ°āļĄāļĩāļāļĢāļ°āđāļĒāļāļāđāļāļĩāđāļŠāļļāļ ``` ### Examples #### Single Turn Conversation Example ``` <s>[INST] <<SYS> You are a question answering assistant. Answer the question as truthful and helpful as possible āļāļļāļāļāļ·āļāļāļđāđāļāđāļ§āļĒāļāļāļāļāļģāļāļēāļĄ āļāļāļāļāļāļāļģāļāļēāļĄāļāļĒāđāļēāļāļāļđāļāļāđāļāļāđāļĨāļ°āļĄāļĩāļāļĢāļ°āđāļĒāļāļāđāļāļĩāđāļŠāļļāļ <</SYS>> āļŠāļ§āļąāļŠāļāļĩāļāļĢāļąāļ [/INST] ``` #### Single Turn Conversation with Context (RAG) Example ``` <s>[INST] <<SYS> You are a question answering assistant. Answer the question as truthful and helpful as possible āļāļļāļāļāļ·āļāļāļđāđāļāđāļ§āļĒāļāļāļāļāļģāļāļēāļĄ āļāļāļāļāļāļāļģāļāļēāļĄāļāļĒāđāļēāļāļāļđāļāļāđāļāļāđāļĨāļ°āļĄāļĩāļāļĢāļ°āđāļĒāļāļāđāļāļĩāđāļŠāļļāļ <</SYS>> āļāļĢāļļāļāđāļāļāļĄāļĩāļāļ·āđāļāļāļĩāđāđāļāđāļēāđāļĢāđ###āļāļĢāļļāļāđāļāļāļĄāļŦāļēāļāļāļĢ āđāļāđāļāđāļĄāļ·āļāļāļŦāļĨāļ§āļ āļāļāļĢāđāļĨāļ°āļĄāļŦāļēāļāļāļĢāļāļĩāđāļĄāļĩāļāļĢāļ°āļāļēāļāļĢāļĄāļēāļāļāļĩāđāļŠāļļāļāļāļāļāļāļĢāļ°āđāļāļĻāđāļāļĒ āļāļĢāļļāļāđāļāļāļĄāļŦāļēāļāļāļĢāļĄāļĩāļāļ·āđāļāļāļĩāđāļāļąāđāļāļŦāļĄāļ 1,568.737 āļāļĢ.āļāļĄ. āļĄāļĩāļāļĢāļ°āļāļēāļāļĢāļāļēāļĄāļāļ°āđāļāļĩāļĒāļāļĢāļēāļĐāļāļĢāļāļ§āđāļē 8 āļĨāđāļēāļāļāļ [/INST] ``` #### Multi Turn Conversation Example ##### First turn ``` <s>[INST] <<SYS> You are a question answering assistant. Answer the question as truthful and helpful as possible āļāļļāļāļāļ·āļāļāļđāđāļāđāļ§āļĒāļāļāļāļāļģāļāļēāļĄ āļāļāļāļāļāļāļģāļāļēāļĄāļāļĒāđāļēāļāļāļđāļāļāđāļāļāđāļĨāļ°āļĄāļĩāļāļĢāļ°āđāļĒāļāļāđāļāļĩāđāļŠāļļāļ <</SYS>> āļŠāļ§āļąāļŠāļāļĩāļāļĢāļąāļ [/INST] ``` ##### Second turn ``` <s>[INST] <<SYS> You are a question answering assistant. Answer the question as truthful and helpful as possible āļāļļāļāļāļ·āļāļāļđāđāļāđāļ§āļĒāļāļāļāļāļģāļāļēāļĄ āļāļāļāļāļāļāļģāļāļēāļĄāļāļĒāđāļēāļāļāļđāļāļāđāļāļāđāļĨāļ°āļĄāļĩāļāļĢāļ°āđāļĒāļāļāđāļāļĩāđāļŠāļļāļ <</SYS>> āļŠāļ§āļąāļŠāļāļĩāļāļĢāļąāļ [/INST]āļŠāļ§āļąāļŠāļāļĩāļāđāļ° āļĄāļĩāļāļģāļāļēāļĄāļāļ°āđāļĢ āļāļēāļĄāđāļāđāđāļĨāļĒ</s><s>āļāļāļŠāļđāļāļĢāļāļģāļŠāđāļĄāļāļģāļŦāļāđāļāļĒ [/INST] ``` ##### Third turn ``` <s>[INST] <<SYS> You are a question answering assistant. Answer the question as truthful and helpful as possible āļāļļāļāļāļ·āļāļāļđāđāļāđāļ§āļĒāļāļāļāļāļģāļāļēāļĄ āļāļāļāļāļāļāļģāļāļēāļĄāļāļĒāđāļēāļāļāļđāļāļāđāļāļāđāļĨāļ°āļĄāļĩāļāļĢāļ°āđāļĒāļāļāđāļāļĩāđāļŠāļļāļ <</SYS>> āļŠāļ§āļąāļŠāļāļĩāļāļĢāļąāļ [/INST]āļŠāļ§āļąāļŠāļāļĩāļāđāļ° āļĄāļĩāļāļģāļāļēāļĄāļāļ°āđāļĢ āļāļēāļĄāđāļāđāđāļĨāļĒ</s><s>āļāļāļŠāļđāļāļĢāļāļģāļŠāđāļĄāļāļģāļŦāļāđāļāļĒ [/INST]āđāļāđāđāļĨāļĒāļāđāļ° āļŠāđāļĄāļāļģāđāļāđāļāđāļĄāļāļđāļāļĩāđāļāļģāļāđāļēāļĒāđāļĨāļ°āļāļĢāđāļāļĒ āļĄāļēāđāļĢāļīāđāļĄāļāļąāļāđāļĨāļĒāļāļ°āļāļ°</s><s>āđāļĢāļīāđāļĄāđāļāđāđāļĨāļĒ [/INST] ``` ##### Fourth turn ``` <s>[INST] <<SYS> You are a question answering assistant. Answer the question as truthful and helpful as possible āļāļļāļāļāļ·āļāļāļđāđāļāđāļ§āļĒāļāļāļāļāļģāļāļēāļĄ āļāļāļāļāļāļāļģāļāļēāļĄāļāļĒāđāļēāļāļāļđāļāļāđāļāļāđāļĨāļ°āļĄāļĩāļāļĢāļ°āđāļĒāļāļāđāļāļĩāđāļŠāļļāļ <</SYS>> āļŠāļ§āļąāļŠāļāļĩāļāļĢāļąāļ [/INST]āļŠāļ§āļąāļŠāļāļĩāļāđāļ° āļĄāļĩāļāļģāļāļēāļĄāļāļ°āđāļĢ āļāļēāļĄāđāļāđāđāļĨāļĒ</s><s>āļāļāļŠāļđāļāļĢāļāļģāļŠāđāļĄāļāļģāļŦāļāđāļāļĒ [/INST]āđāļāđāđāļĨāļĒāļāđāļ° āļŠāđāļĄāļāļģāđāļāđāļāđāļĄāļāļđāļāļĩāđāļāļģāļāđāļēāļĒāđāļĨāļ°āļāļĢāđāļāļĒ āļĄāļēāđāļĢāļīāđāļĄāļāļąāļāđāļĨāļĒāļāļ°āļāļ°</s><s>āđāļĢāļīāđāļĄāđāļāđāđāļĨāļĒ [/INST] āļāļģāđāļāļĒāđāļāđāļāđāļĄāļāļđāļāļĩāđāđāļāļĢāđ āļāđāļāļļāđāļāđāļāļĒ āļāđāļ§āļĒāļĢāļŠāļāļēāļāļīāļāļĩāđāļāļĨāļĄāļāļĨāđāļāļĄ āļŦāļ§āļēāļ āđāļāļĢāļĩāđāļĒāļ§ āđāļāđāļĄ āđāļāđāļ āļāļĢāļāļĢāļŠ āļ§āļąāļāļāļĩāđāđāļĢāļēāļāļ°āļĄāļēāļāļģāļŠāđāļĄāļāļģāļāđāļ§āļĒāļāļąāļāļāđāļ° āļŠāļīāđāļāđāļĢāļāļāļĩāđāļāđāļāļāļĄāļĩāļāļ·āļāđāļāļĢāļ·āđāļāļāļāļĢāļļāļāļāļąāļāļāļĩāđ - āļĄāļ°āļĨāļ°āļāļāļāļīāļ 1 āļĨāļđāļ - āļāļļāđāļāđāļŦāđāļ 1/2 āļāđāļ§āļĒ - āļāļĢāļ°āđāļāļĩāļĒāļĄ 3 āļāļĨāļĩāļ - āļāļĢāļīāļāļāļĩāđāļŦāļāļđ 3 āđāļĄāđāļ - āļāđāļģāļāļēāļĨāļāļĩāđāļ 1 āļāđāļāļāđāļāđāļ° - āļāđāļģāļāļĨāļē 2 āļāđāļāļāđāļāđāļ° - āļĄāļ°āļāļēāļ§ 1 āļĨāļđāļ - āļāļēāļāļēāļ§āļāļąāđāļ§ 1/4 āļāđāļ§āļĒ āļ§āļīāļāļĩāļāļģāļĄāļĩāļāļąāļāļāļĩāđāļāđāļ° 1. āđāļĢāļīāđāļĄāļāļēāļāļĨāđāļēāļāļĄāļ°āļĨāļ°āļāļāđāļŦāđāļŠāļ°āļāļēāļ āđāļĨāđāļ§āđāļāđāļĄāļĩāļāļāļāļāđāļāļĨāļ·āļāļ āđāļāļēāđāļŠāđāļāļāļ āļŦāļąāđāļāđāļāđāļāđāļŠāđāļāļāļēāļāđ āđāļāļĢāļĩāļĒāļĄāđāļ§āđ 2. āļāļģāļāļļāđāļāđāļŦāđāļāđāļŦāđāļĨāļ°āđāļāļĩāļĒāļ āđāļĨāđāļ§āļāļąāļāļāļķāđāļāļāļąāļāđāļ§āđ 3. āđāļāđāļāļĢāļāļŦāļīāļāļŦāļĢāļ·āļāđāļāļĢāļ·āđāļāļāļāļąāđāļ āļāļāļāļĢāļīāļāļāļĩāđāļŦāļāļđāļāļąāļāļāļĢāļ°āđāļāļĩāļĒāļĄāđāļŦāđāļĨāļ°āđāļāļĩāļĒāļ 4. āđāļŠāđāļāļļāđāļāđāļŦāđāļāļāļĩāđāļāļģāđāļĨāđāļ§āļĨāļāđāļāļāļŠāļĄ āļāļēāļĄāļāđāļ§āļĒāļāđāļģāļāļēāļĨāļāļĩāđāļ āļāđāļģāļāļĨāļē āļĄāļ°āļāļēāļ§ āđāļĨāļ°āđāļŠāđāļāļĄāļ°āļĨāļ°āļāļ āļāļĨāļļāļāđāļāļĨāđāļēāđāļŦāđāđāļāđāļēāļāļąāļ 5. āļāļīāļĄāļĢāļŠāđāļŦāđāđāļāđāļĢāļŠāļŦāļ§āļēāļ āđāļāļĢāļĩāđāļĒāļ§ āđāļāđāļĄ āđāļāđāļ āļāļēāļāļāļąāđāļāļāļąāļāļāļķāđāļāđāļŠāļīāļĢāđāļāļāļĢāđāļāļĄāļāļąāļāļŠāļ āļāļēāļāļī āļāļ°āļŦāļĨāđāļģāļāļĨāļĩ āļāļąāđāļ§āļāļāļ āđāļāļĢāļāļ āļāļąāļāļāļļāđāļ</s><s>āļāļāļāļāļļāļāļāļĢāļąāļ [/INST] ``` #### Multi Turn Conversation with Context (RAG) Example ``` <s>[INST] <<SYS> You are a question answering assistant. Answer the question as truthful and helpful as possible āļāļļāļāļāļ·āļāļāļđāđāļāđāļ§āļĒāļāļāļāļāļģāļāļēāļĄ āļāļāļāļāļāļāļģāļāļēāļĄāļāļĒāđāļēāļāļāļđāļāļāđāļāļāđāļĨāļ°āļĄāļĩāļāļĢāļ°āđāļĒāļāļāđāļāļĩāđāļŠāļļāļ <</SYS>> āļāļĢāļļāļāđāļāļāļĄāļĩāļāļ·āđāļāļāļĩāđāđāļāđāļēāđāļĢāđ###āļāļĢāļļāļāđāļāļāļĄāļŦāļēāļāļāļĢ āđāļāđāļāđāļĄāļ·āļāļāļŦāļĨāļ§āļ āļāļāļĢāđāļĨāļ°āļĄāļŦāļēāļāļāļĢāļāļĩāđāļĄāļĩāļāļĢāļ°āļāļēāļāļĢāļĄāļēāļāļāļĩāđāļŠāļļāļāļāļāļāļāļĢāļ°āđāļāļĻāđāļāļĒ āļāļĢāļļāļāđāļāļāļĄāļŦāļēāļāļāļĢāļĄāļĩāļāļ·āđāļāļāļĩāđāļāļąāđāļāļŦāļĄāļ 1,568.737 āļāļĢ.āļāļĄ. āļĄāļĩāļāļĢāļ°āļāļēāļāļĢāļāļēāļĄāļāļ°āđāļāļĩāļĒāļāļĢāļēāļĐāļāļĢāļāļ§āđāļē 8 āļĨāđāļēāļāļāļ [/INST] āļāļĢāļļāļāđāļāļāļĄāļŦāļēāļāļāļĢāļĄāļĩāļāļ·āđāļāļāļĩāđāļāļąāđāļāļŦāļĄāļ 1,568.737 āļāļĢ.āļāļĄ.</s><s>āđāļĨāļ°āļāļĢāļ°āļāļēāļāļĢāļĨāđāļ° [/INST] ``` ## How to use ### vLLM Engine for float16 model 1. install VLLM (https://github.com/vllm-project/vllm) 2. python -m vllm.entrypoints.api_server --model /path/to/model --tensor-parallel-size num_gpus 3. run inference (CURL example) ``` curl --request POST \ --url http://localhost:8000/generate \ --header "Content-Type: application/json" \ --data '{"prompt": "<s>[INST] <<SYS>>\nYou are a question answering assistant. Answer the question as truthful and helpful as possible āļāļļāļāļāļ·āļāļāļđāđāļāđāļ§āļĒāļāļāļāļāļģāļāļēāļĄ āļāļāļāļāļāļāļģāļāļēāļĄāļāļĒāđāļēāļāļāļđāļāļāđāļāļāđāļĨāļ°āļĄāļĩāļāļĢāļ°āđāļĒāļāļāđāļāļĩāđāļŠāļļāļ\n<</SYS>>\n\nāļāļĒāļēāļāļĨāļāļāļ§āļēāļĄāļāđāļ§āļāļāđāļāļāļāļģāļāļĒāđāļēāļāđāļĢ [/INST]","use_beam_search": false, "temperature": 0.1, "max_tokens": 512, "top_p": 0.75, "top_k": 40, "frequency_penalty": 0.3 "stop": "</s>"}' ``` ### LlamaCPP Engine for 4 bit model ### Authors * Kobkrit Viriyayudhakorn (kobkrit@aieat.or.th) * Sumeth Yuenyong (sumeth.yue@mahidol.edu) * Thaweewat Rugsujarit (thaweewr@scg.com) * Jillaphat Jaroenkantasima (autsadang41@gmail.com) * Norapat Buppodom (new@norapat.com) * Koravich Sangkaew (kwankoravich@gmail.com) * Peerawat Rojratchadakorn (peerawat.roj@gmail.com) * Surapon Nonesung (nonesungsurapon@gmail.com) * Chanon Utupon (chanon.utupon@gmail.com) * Sadhis Wongprayoon (sadhis.tae@gmail.com) * Nucharee Thongthungwong (nuchhub@hotmail.com) * Chawakorn Phiantham (mondcha1507@gmail.com) * Patteera Triamamornwooth (patt.patteera@gmail.com) * Nattarika Juntarapaoraya (natt.juntara@gmail.com) * Kriangkrai Saetan (kraitan.ss21@gmail.com) * Pitikorn Khlaisamniang (pitikorn32@gmail.com) <i>Disclaimer: Provided responses are not guaranteed.</i>