README.md

---
license: apache-2.0
language:
  - en
  - ja
programming_language:
  - C
  - C++
  - C#
  - Go
  - Java
  - JavaScript
  - Lua
  - PHP
  - Python
  - Ruby
  - Rust
  - Scala
  - TypeScript
library_name: peft
pipeline_tag: text-generation
inference: false
---
# llm-jp-13b-instruct-lora-dolly-oasst-v1.0

This repository provides large language models developed by [LLM-jp](https://llm-jp.nii.ac.jp/), a collaborative project launched in Japan.

| Model Variant | 
| :--- |
|**Instruction models**|
| [llm-jp-13b-instruct-full-jaster-v1.0](https://huggingface.co./llm-jp/llm-jp-13b-instruct-full-jaster-v1.0) |
| [llm-jp-13b-instruct-full-jaster-dolly-oasst-v1.0](https://huggingface.co./llm-jp/llm-jp-13b-instruct-full-jaster-dolly-oasst-v1.0) |
| [llm-jp-13b-instruct-full-dolly-oasst-v1.0](https://huggingface.co./llm-jp/llm-jp-13b-instruct-full-dolly-oasst-v1.0) |
| [llm-jp-13b-instruct-lora-jaster-v1.0](https://huggingface.co./llm-jp/llm-jp-13b-instruct-lora-jaster-v1.0) | 
| [llm-jp-13b-instruct-lora-jaster-dolly-oasst-v1.0](https://huggingface.co./llm-jp/llm-jp-13b-instruct-lora-jaster-dolly-oasst-v1.0) | 
| [llm-jp-13b-instruct-lora-dolly-oasst-v1.0](https://huggingface.co./llm-jp/llm-jp-13b-instruct-lora-dolly-oasst-v1.0) | 


|  | 
| :--- |
|**Pre-trained models**|
| [llm-jp-13b-v1.0](https://huggingface.co./llm-jp/llm-jp-13b-v1.0) | 
| [llm-jp-1.3b-v1.0](https://huggingface.co./llm-jp/llm-jp-1.3b-v1.0) | 
Checkpoints format: `transformers` (Megatron-DeepSpeed format available [here](https://huggingface.co./llm-jp/llm-jp-13b-v1.0-mdsfmt))


## Required Libraries and Their Versions

- torch>=2.0.0
- transformers>=4.34.0
- tokenizers>=0.14.0
- peft==0.5.0

## Usage

```python
import torch
from peft import PeftModel, PeftConfig
from transformers import AutoTokenizer, AutoModelForCausalLM
peft_model_name = "llm-jp/llm-jp-13b-instruct-lora-dolly-oasst-v1.0"
tokenizer = AutoTokenizer.from_pretrained(peft_model_name)
config = PeftConfig.from_pretrained(peft_model_name)
model = AutoModelForCausalLM.from_pretrained(config.base_model_name_or_path, torch_dtype=torch.float16)
model = PeftModel.from_pretrained(model, peft_model_name)
text = "自然言語処理とは何か"
text = text + "### 回答："
tokenized_input = tokenizer(text, add_special_tokens=False, return_tensors="pt").to(model.device)
with torch.no_grad():
    output = model.generate(
        **tokenized_input,
        max_new_tokens=100,
        do_sample=True,
        top_p=0.95,
        temperature=0.7,
    )[0]
print(tokenizer.decode(output))
```


## Model Details

- **Model type:** Transformer-based Language Model
- **Total seen tokens:** 270B+

|Model|Params|Layers|Hidden size|Heads|Context length|
|:---:|:---:|:---:|:---:|:---:|:---:|
|13b model|13b|40|5120|40|2048|
|1.3b model|1.3b|24|2048|16|2048|


## Training

- **Pre-training:**
  - **Hardware:** 96 A100 40GB GPUs ([mdx cluster](https://mdx.jp/en/))
  - **Software:** Megatron-DeepSpeed

- **Instruction tuning:**
  - **Hardware:** 8 A100 40GB GPUs ([mdx cluster](https://mdx.jp/en/))
  - **Software:** [TRL](https://github.com/huggingface/trl), [PEFT](https://github.com/huggingface/peft), and [DeepSpeed](https://github.com/microsoft/DeepSpeed)


## Tokenizer
The tokenizer of this model is based on [huggingface/tokenizers](https://github.com/huggingface/tokenizers) Unigram byte-fallback model.
The vocab entries were converted from [`llm-jp-tokenizer v2.1 (50k)`](https://github.com/llm-jp/llm-jp-tokenizer/releases/tag/v2.1).
Please refer to [README.md](https://github.com/llm-jp/llm-jp-tokenizer) of `llm-ja-tokenizer` for the details of vocab constuction steps.
- **Model:** Hugging Face Fast Tokenizer using Unigram byte-fallback model which requires `tokenizers>=0.14.0`
- **Training algorithm:** SentencePiece Unigram byte-fallback
- **Training data:** A subset of the datasets for model pre-training
- **Vocabulary size:** 50,570 (mixed vocabulary of Japanese, English, and source code)


## Datasets

### Pre-training

The models have been pre-trained on approximately 287.5B tokens, sourced from a blend of the following datasets.

| Language | Dataset | Tokens|
|:---:|:---:|:---:|
|Japanese|[Wikipedia](https://huggingface.co./datasets/wikipedia)|1.5B
||[mC4](https://huggingface.co./datasets/mc4)|136B
|English|[Wikipedia](https://huggingface.co./datasets/wikipedia)|5B
||[The Pile](https://huggingface.co./datasets/EleutherAI/pile)|135B
|Codes|[The Stack](https://huggingface.co./datasets/bigcode/the-stack)|10B

Pretraining was done by 10-hold shards that consists approx. 27-28B tokens. We further finalized the pretraining with additional cleaned 27B tokens data.

### Instruction tuning

The models have been fine-tuned on the following datasets.
 
| Language | Dataset | description |
|:---|:---:|:---:|
|Japanese|[jaster](https://github.com/llm-jp/llm-jp-eval)| An automatically transformed data from the existing Japanese NLP datasets |
||[databricks-dolly-15k](https://huggingface.co./datasets/databricks/databricks-dolly-15k)| A translated one by DeepL in LLM-jp |
||[OpenAssistant Conversations Dataset](https://huggingface.co./datasets/OpenAssistant/oasst1)| A translated one by DeepL in LLM-jp |


## Evaluation
You can view the evaluation results of several LLMs on this [leaderboard](http://wandb.me/llm-jp-leaderboard). We used [llm-jp-eval](https://github.com/llm-jp/llm-jp-eval) for the evaluation.

## Risks and Limitations

The models released here are still in the early stages of our research and development and have not been tuned to ensure outputs align with human intent and safety considerations.


## Send Questions to

llm-jp(at)nii.ac.jp


## License

[Apache License, Version 2.0](https://www.apache.org/licenses/LICENSE-2.0)


## Model Card Authors
*The names are listed in alphabetical order.*

Namgi Han, Hirokazu Kiyomaru, Hiroshi Matsuda, Shota Sasaki, Shuhei Kurita,　Taishi Nakamura, Takumi Okamoto.