language:
- th
- en
license: llama2
datasets:
- HuggingFaceH4/ultrachat_200k
- HuggingFaceH4/ultrafeedback_binarized
- HuggingFaceH4/cai-conversation-harmless
model-index:
- name: SambaLingo-Thai-Chat
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 52.73
name: normalized accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=sambanovasystems/SambaLingo-Thai-Chat
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 78.42
name: normalized accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=sambanovasystems/SambaLingo-Thai-Chat
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 43.95
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=sambanovasystems/SambaLingo-Thai-Chat
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 40.84
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=sambanovasystems/SambaLingo-Thai-Chat
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 72.22
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=sambanovasystems/SambaLingo-Thai-Chat
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 8.57
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=sambanovasystems/SambaLingo-Thai-Chat
name: Open LLM Leaderboard
SambaLingo-Thai-Chat
SambaLingo-Thai-Chat is a human aligned chat model trained in Thai and English. It is trained using direct preference optimization on top the base model SambaLingo-Thai-Base. The base model adapts Llama-2-7b to Thai by training on 38 billion tokens from the Thai split of the Cultura-X dataset. Try This Model at SambaLingo-chat-space.
Model Description
- Developed by: SambaNova Systems
- Model type: Language Model
- Language(s): Thai, English
- Finetuned from model: Llama-2-7b
- Try This Model: SambaLingo-chat-space
- Paper: SambaLingo: Teaching Large Language Models New Languages
- Blog Post: sambalingo-open-source-language-experts
Getting Started
Loading Model With Hugging Face
Please make sure to set use_fast=False when loading the tokenizer.
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("sambanovasystems/SambaLingo-Thai-Chat", use_fast=False)
model = AutoModelForCausalLM.from_pretrained("sambanovasystems/SambaLingo-Thai-Chat", device_map="auto", torch_dtype="auto")
Interacting With Model Pipeline
Please make sure to set use_fast=False when loading the tokenizer.
from transformers import pipeline
pipe = pipeline("text-generation", model="sambanovasystems/SambaLingo-Thai-Chat", device_map="auto", use_fast=False)
messages = [
{"role": "user", "content": {YOUR_QUESTION}},
]
prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipe(prompt)[0]
outputs = outputs["generated_text"]
Suggested Inference Parameters
- Temperature: 0.8
- Repetition penalty: 1.0
- Top-p: 0.9
Prompting Guidelines
To prompt this model, please use the following chat template:
<|user|>\n{question}</s>\n<|assistant|>\n
Example Prompts and Generations
<|user|>
ปรัชญาทางเศรษฐกิจที่พระบาทสมเด็จพระมหาภูมิพลอดุลยเดชมหาราช บรมนาถบพิตรมีพระราชดำรัสแก่ชาวไทย มีชื่อว่าอะไร</s>
<|assistant|>
ปรัชญาทางเศรษฐกิจที่พระบาทสมเด็จพระปรมินทรมหาภูมิพลอดุลยเดชทรงมีพระราชดำรัสแก่ชาวไทยเรียกว่า "ปรัชญาเศรษฐกิจพอเพียง" ปรัชญานี้เน้นความสำคัญของการพัฒนาแบบองค์รวมแ ละความสมดุล โดยคำนึงถึงสภาพแวดล้อมทางสังคม เศรษฐกิจ และสิ่งแวดล้อม ปรัชญาเศรษฐกิจพอเพียงเน้นความสำคัญของการพึ่งพาตนเอง ความมีเหตุผล และการประหยัด ตลอดจนความสำคัญของก ารพัฒนาจิตใจและจิตวิญญาณ ปรัชญานี้เน้นย้ำถึงความสำคัญของการดำเนินชีวิตอย่างพอประมาณ มีเหตุผล และมีคุณธรรม ตลอดจนความสำคัญของการพัฒนาความรู้และทักษะอย่างต่อเนื่อง ปรัชญาเศรษฐกิ จพอเพียงได้รับการยกย่องอย่างกว้างขวางว่าเป็นแนวทางแก้ไขปัญหาการพัฒนาที่ยั่งยืนและเป็นแนวทางในการดำเนินชีวิตที่นำไปสู่ความสุขและความเป็นอยู่ที่ดี
Training Details
The alignment phase follows the recipe for Zephyr-7B, and comprises two stages: supervised fine-tuning (SFT) and Direct Performance Optimization (DPO).
The SFT phase was done on the ultrachat_200k dataset mixed with the Google translated version of the ultrachat_200k dataset. It was trained for one epoch with global batch size 512 and max sequence length 2048 tokens. We used a linear decay learning rate of 2e-5 and 10% warmup.
The DPO phase was done on the ultrafeedback dataset and cai-conversation-harmless dataset, mixed with 10% of the data Google translated. It was trained with global batch size 32 and for three epochs. We used a linear decay learning rate of 5e-7, 10% warmup and β=0.1 as the regularization factor for DPO.
Tokenizer Details
We extended the vocabulary of the base llama model from 32,000 tokens to 57,000 tokens by adding up to 25,000 non-overlapping tokens from the new language.
Evaluation
For evaluation results see our paper: SambaLingo: Teaching Large Language Models New Languages
Uses
Direct Use
Use of this model is governed by the Meta’s Llama 2 Community License Agreement. Please review and accept the license before downloading the model weights.
Out-of-Scope Use
SambaLingo should NOT be used for:
- Mission-critical applications
- Applications that involve the safety of others
- Making highly important decisions
Bias, Risks, and Limitations
Like all LLMs, SambaLingo has certain limitations:
- Hallucination: Model may sometimes generate responses that contain plausible-sounding but factually incorrect or irrelevant information.
- Code Switching: The model might unintentionally switch between languages or dialects within a single response, affecting the coherence and understandability of the output.
- Repetition: The Model may produce repetitive phrases or sentences, leading to less engaging and informative responses.
- Coding and Math: The model's performance in generating accurate code or solving complex mathematical problems may be limited.
- Toxicity: The model could inadvertently generate responses containing inappropriate or harmful content.
Acknowledgments
We extend our heartfelt gratitude to the open-source AI community; this endeavor would not have been possible without open source. SambaNova embraces the open-source community and aspires to actively contribute to this initiative.
We would like to give a special thanks to the following groups:
- Meta for open sourcing LLama 2 and open sourcing FLORES-200 dataset
- Nguyen et al for open sourcing CulturaX dataset
- CohereAI for releasing AYA-101 and open sourcing a multilingual instruction tuning dataset
- EleutherAI for their open source evaluation framework
- Hugging Face-H4 team for open source the zephyr training recipe and alignment handbook repo
Cite SambaLingo
@misc{csaki2024sambalingo,
title={SambaLingo: Teaching Large Language Models New Languages},
author={Zoltan Csaki and Bo Li and Jonathan Li and Qiantong Xu and Pian Pawakapan and Leon Zhang and Yun Du and Hengyu Zhao and Changran Hu and Urmish Thakker},
year={2024},
eprint={2404.05829},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 49.45 |
| AI2 Reasoning Challenge (25-Shot) | 52.73 |
| HellaSwag (10-Shot) | 78.42 |
| MMLU (5-Shot) | 43.95 |
| TruthfulQA (0-shot) | 40.84 |
| Winogrande (5-shot) | 72.22 |
| GSM8k (5-shot) | 8.57 |