|
--- |
|
thumbnail: https://github.com/rinnakk/japanese-pretrained-models/blob/master/rinna.png |
|
license: llama3 |
|
datasets: |
|
- CohereForAI/aya_dataset |
|
- kunishou/databricks-dolly-15k-ja |
|
- kunishou/HelpSteer-35k-ja |
|
- kunishou/HelpSteer2-20k-ja |
|
- kunishou/hh-rlhf-49k-ja |
|
- kunishou/oasst1-chat-44k-ja |
|
- kunishou/oasst2-chat-68k-ja |
|
- meta-math/MetaMathQA |
|
- OpenAssistant/oasst1 |
|
- OpenAssistant/oasst2 |
|
- sahil2801/CodeAlpaca-20k |
|
language: |
|
- ja |
|
- en |
|
tags: |
|
- llama |
|
- llama-3 |
|
inference: false |
|
base_model: |
|
- rinna/llama-3-youko-8b |
|
- meta-llama/Meta-Llama-3-8B |
|
- meta-llama/Meta-Llama-3-8B-Instruct |
|
base_model_relation: merge |
|
--- |
|
|
|
# `Llama 3 Youko 8B Instruct (rinna/llama-3-youko-8b-instruct)` |
|
|
|
 |
|
|
|
# Overview |
|
|
|
The model is the instruction-tuned version of [rinna/llama-3-youko-8b](https://huggingface.co/rinna/llama-3-youko-8b), using supervised fine-tuning (SFT), Chat Vector, and direct preference optimization (DPO). It adpots the Llama-3 chat format. |
|
|
|
| Size | Continual Pre-Training | Instruction-Tuning | |
|
| :- | :- | :- | |
|
| 8B | Llama 3 Youko 8B [[HF]](https://huggingface.co/rinna/llama-3-youko-8b) [[GPTQ]](https://huggingface.co/rinna/llama-3-youko-8b-gptq) | Llama 3 Youko 8B Instruct [[HF]](https://huggingface.co/rinna/llama-3-youko-8b-instruct) [[GPTQ]](https://huggingface.co/rinna/llama-3-youko-8b-instruct-gptq) | |
|
| 70B | Llama 3 Youko 70B [[HF]](https://huggingface.co/rinna/llama-3-youko-70b) [[GPTQ]](https://huggingface.co/rinna/llama-3-youko-70b-gptq) | Llama 3 Youko 70B Instruct [[HF]](https://huggingface.co/rinna/llama-3-youko-70b-instruct) [[GPTQ]](https://huggingface.co/rinna/llama-3-youko-70b-instruct-gptq) | |
|
|
|
* **Model architecture** |
|
|
|
A 32-layer, 4096-hidden-size transformer-based language model. Refer to the [Llama 3 Model Card](https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md) for architecture details. |
|
|
|
* **Training: Built with Meta Llama 3** |
|
|
|
**Supervised fine-tuning.** The supervised fine-tuning data is a subset of the following datasets. |
|
|
|
- [CohereForAI/aya_dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset) |
|
- The JPN subset was used. |
|
- [FLAN](https://github.com/google-research/FLAN/tree/main/flan/v2) |
|
- [kunishou/databricks-dolly-15k-ja](https://huggingface.co/datasets/kunishou/databricks-dolly-15k-ja) |
|
- [kunishou/hh-rlhf-49k-ja](https://huggingface.co/datasets/kunishou/hh-rlhf-49k-ja) |
|
- [kunishou/oasst1-chat-44k-ja](https://huggingface.co/datasets/kunishou/oasst1-chat-44k-ja) |
|
- [kunishou/oasst2-chat-68k-ja](https://huggingface.co/datasets/kunishou/oasst2-chat-68k-ja) |
|
- [meta-math/MetaMathQA](https://huggingface.co/datasets/meta-math/MetaMathQA) |
|
- The following sections were used: MATH_AnsAug, MATH_Rephrased, MATH_SV, and MATH_FOBAR. |
|
- The remaining sections, containing augmented data from commonly used evaluation corpora, were skipped for preventing any possibility of data leak. |
|
- [OpenAssistant/oasst1](https://huggingface.co/datasets/OpenAssistant/oasst1) |
|
- The EN and JA subsets were used. |
|
- [OpenAssistant/oasst2](https://huggingface.co/datasets/OpenAssistant/oasst2) |
|
- The EN and JA subsets were used. |
|
- [sahil2801/CodeAlpaca-20k](https://huggingface.co/datasets/sahil2801/CodeAlpaca-20k) |
|
- rinna Dataset |
|
|
|
**Model merging.** The fine-tuned model (llama-3-youko-8b-sft) has been enhanced through the following chat vector addition. The chat vector was obtained by subtracting the parameter vectors of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) from those of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct). |
|
|
|
~~~~text |
|
llama-3-youko-8b-sft + 0.5 * (meta-llama/Meta-Llama-3-8B-Instruct - meta-llama/Meta-Llama-3-8B) |
|
~~~~ |
|
|
|
Here, the embedding layer was skipped while subtracting and adding the parameter vectors. |
|
|
|
**Direct preference optimization** was then applied with a subset of the following datasets to build this instruct model. |
|
|
|
- [kunishou/HelpSteer-35k-ja](https://huggingface.co/datasets/kunishou/HelpSteer-35k-ja) |
|
- [kunishou/HelpSteer2-20k-ja](https://huggingface.co/datasets/kunishou/HelpSteer2-20k-ja) |
|
- rinna Dataset |
|
|
|
* **Contributors** |
|
|
|
- [Xinqi Chen](https://huggingface.co/Keely0419) |
|
- [Koh Mitsuda](https://huggingface.co/mitsu-koh) |
|
- [Toshiaki Wakatsuki](https://huggingface.co/t-w) |
|
- [Kei Sawada](https://huggingface.co/keisawada) |
|
|
|
--- |
|
|
|
# Benchmarking |
|
|
|
Please refer to [rinna's LM benchmark page](https://rinnakk.github.io/research/benchmarks/lm/index.html). |
|
|
|
--- |
|
|
|
# How to use the model |
|
|
|
We found this instruction-tuned model tends to generate repeated text more often than its base counterpart, and thus we set repetition_penalty=1.1 for better generation performance. The same repetition penalty was applied to the instruction-tuned model in the aforementioned evaluation experiments. |
|
|
|
~~~~python |
|
import torch |
|
from transformers import AutoTokenizer, AutoModelForCausalLM |
|
|
|
model_id = "rinna/llama-3-youko-8b-instruct" |
|
|
|
tokenizer = AutoTokenizer.from_pretrained(model_id) |
|
model = AutoModelForCausalLM.from_pretrained( |
|
model_id, |
|
torch_dtype=torch.bfloat16, |
|
device_map="auto", |
|
) |
|
|
|
messages = [ |
|
{"role": "system", "content": "あなたは誠実で優秀なアシスタントです。どうか、簡潔かつ正直に答えてください。"}, |
|
{"role": "user", "content": "西田幾多郎とはどんな人物ですか？"}, |
|
] |
|
|
|
input_ids = tokenizer.apply_chat_template( |
|
messages, |
|
add_generation_prompt=True, |
|
return_tensors="pt" |
|
).to(model.device) |
|
|
|
terminators = [ |
|
tokenizer.convert_tokens_to_ids("<|end_of_text|>"), |
|
tokenizer.convert_tokens_to_ids("<|eot_id|>") |
|
] |
|
|
|
outputs = model.generate( |
|
input_ids, |
|
max_new_tokens=512, |
|
eos_token_id=terminators, |
|
do_sample=True, |
|
temperature=0.6, |
|
top_p=0.9, |
|
repetition_penalty=1.1, |
|
) |
|
|
|
response = outputs[0][input_ids.shape[-1]:] |
|
response = tokenizer.decode(response, skip_special_tokens=True) |
|
print(response) |
|
~~~~ |
|
|
|
--- |
|
|
|
# Tokenization |
|
The model uses the original [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) tokenizer. |
|
|
|
--- |
|
|
|
# How to cite |
|
```bibtex |
|
@misc{rinna-llama-3-youko-8b-instruct, |
|
title = {rinna/llama-3-youko-8b-instruct}, |
|
author = {Chen, Xinqi and Mitsuda, Koh and Wakatsuki, Toshiaki and Sawada, Kei}, |
|
url = {https://huggingface.co/rinna/llama-3-youko-8b-instruct} |
|
} |
|
|
|
@inproceedings{sawada2024release, |
|
title = {Release of Pre-Trained Models for the {J}apanese Language}, |
|
author = {Sawada, Kei and Zhao, Tianyu and Shing, Makoto and Mitsui, Kentaro and Kaga, Akio and Hono, Yukiya and Wakatsuki, Toshiaki and Mitsuda, Koh}, |
|
booktitle = {Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)}, |
|
month = {5}, |
|
year = {2024}, |
|
pages = {13898--13905}, |
|
url = {https://aclanthology.org/2024.lrec-main.1213}, |
|
note = {\url{https://arxiv.org/abs/2404.01657}} |
|
} |
|
``` |
|
--- |
|
|
|
# References |
|
```bibtex |
|
@article{llama3modelcard, |
|
title = {Llama 3 Model Card}, |
|
author = {AI@Meta}, |
|
year = {2024}, |
|
url = {https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md} |
|
} |
|
|
|
@article{huang2023chat, |
|
title = {Chat Vector: A Simple Approach to Equip LLMs with Instruction Following and Model Alignment in New Languages}, |
|
author = {Huang, Shih-Cheng and Li, Pin-Zu and Hsu, Yu-Chi and Chen, Kuang-Ming and Lin, Yu Tung and Hsiao, Shih-Kai and Tzong-Han Tsai, Richard and Lee, Hung-yi}, |
|
year = {2023}, |
|
url = {https://arxiv.org/abs/2310.04799} |
|
} |
|
``` |
|
--- |
|
|
|
# License |
|
[Meta Llama 3 Community License](https://llama.meta.com/llama3/license/) |
