Metadata-Version: 2.1
Name: glmtuner
Version: 0.1.1
Summary: Fine-tuning ChatGLM-6B with PEFT
Home-page: https://github.com/hiyouga/ChatGLM-Efficient-Tuning
Author: hiyouga
Author-email: hiyouga@buaa.edu.cn
License: Apache 2.0 License
Keywords: ChatGLM,LLM,ChatGPT,transformer,pytorch,deep learning
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Education
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.8.0
Description-Content-Type: text/markdown
License-File: LICENSE

# ChatGLM Efficient Tuning

[![GitHub Repo stars](https://img.shields.io/github/stars/hiyouga/ChatGLM-Efficient-Tuning?style=social)](https://github.com/hiyouga/ChatGLM-Efficient-Tuning/stargazers)
[![GitHub Code License](https://img.shields.io/github/license/hiyouga/ChatGLM-Efficient-Tuning)](LICENSE)
[![GitHub last commit](https://img.shields.io/github/last-commit/hiyouga/ChatGLM-Efficient-Tuning)](https://github.com/hiyouga/ChatGLM-Efficient-Tuning/commits/main)
[![PyPI](https://img.shields.io/pypi/v/glmtuner)](https://pypi.org/project/glmtuner/)
[![GitHub pull request](https://img.shields.io/badge/PRs-welcome-blue)](https://github.com/hiyouga/ChatGLM-Efficient-Tuning/pulls)

Fine-tuning 🤖[ChatGLM-6B](https://github.com/THUDM/ChatGLM-6B) model with 🤗[PEFT](https://github.com/huggingface/peft).

👋 Join our [WeChat](assets/wechat.jpg).

\[ English | [中文](README_zh.md) \]

## Changelog

[23/07/15] Now we develop an all-in-one Web UI for training, evaluation and inference. Try `train_web.py` to fine-tune ChatGLM-6B model in your Web browser. Thank [@KanadeSiina](https://github.com/KanadeSiina) for his efforts in the development.

[23/07/09] Now we release [FastEdit](https://github.com/hiyouga/FastEdit)⚡🩹, an easy-to-use package for editing the factual knowledge of large language models efficiently. Please follow [FastEdit](https://github.com/hiyouga/FastEdit) if you are interested.

[23/06/25] Now we align the [demo API](src/api_demo.py) with the [OpenAI's](https://platform.openai.com/docs/api-reference/chat) format where you can insert the fine-tuned model in arbitrary ChatGPT-based applications.

[23/06/25] Now we support fine-tuning the [ChatGLM2-6B](https://github.com/THUDM/ChatGLM2-6B) model with our framework!

[23/06/05] Now we support 4-bit LoRA training (aka [QLoRA](https://github.com/artidoro/qlora)). Try `--quantization_bit 4` argument to work with 4-bit quantized model. (experimental feature)

[23/06/01] We implemented a framework supporting the efficient tuning of LLaMA and BLOOM models. Please follow [LLaMA-Efficient-Tuning](https://github.com/hiyouga/LLaMA-Efficient-Tuning) if you are interested.

[23/05/19] Now we support using the development set to evaluate the model while training. Try `--dev_ratio` argument to specify the size of development set.

[23/04/29] Now we support training ChatGLM with **Reinforcement Learning with Human Feedback (RLHF)** ! We provide several examples to run RLHF training, please refer to the `examples` folder for details.

[23/04/20] Our repo achieved 100 stars within 12 days! Congratulations!

[23/04/19] Now we support **merging the weights** of fine-tuned models trained by LoRA! Try `--checkpoint_dir checkpoint1,checkpoint2` argument for continually fine-tuning the models.

[23/04/18] Now we support training the **quantized models** using three fine-tuning methods! Try `quantization_bit` argument for training the model in 4/8 bits.

[23/04/12] Now we support **training from checkpoints**! Use `--checkpoint_dir` argument to specify the checkpoint model to fine-tune from.

[23/04/11] Now we support training with **combined datasets**! Try `--dataset dataset1,dataset2` argument for training with multiple datasets.

## Datasets

Our script now supports the following datasets:

- [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca)
- [Stanford Alpaca (Chinese)](https://github.com/ymcui/Chinese-LLaMA-Alpaca)
- [GPT-4 Generated Data](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
- [BELLE 2M](https://huggingface.co/datasets/BelleGroup/train_2M_CN)
- [BELLE 1M](https://huggingface.co/datasets/BelleGroup/train_1M_CN)
- [BELLE 0.5M](https://huggingface.co/datasets/BelleGroup/train_0.5M_CN)
- [BELLE Dialogue 0.4M](https://huggingface.co/datasets/BelleGroup/generated_chat_0.4M)
- [BELLE School Math 0.25M](https://huggingface.co/datasets/BelleGroup/school_math_0.25M)
- [BELLE Multiturn Chat 0.8M](https://huggingface.co/datasets/BelleGroup/multiturn_chat_0.8M)
- [Guanaco Dataset](https://huggingface.co/datasets/JosephusCheung/GuanacoDataset)
- [Firefly 1.1M](https://huggingface.co/datasets/YeungNLP/firefly-train-1.1M)
- [CodeAlpaca 20k](https://huggingface.co/datasets/sahil2801/CodeAlpaca-20k)
- [Alpaca CoT](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT)
- [Web QA (Chinese)](https://huggingface.co/datasets/suolyer/webqa)
- [UltraChat](https://github.com/thunlp/UltraChat)
- [WebNovel (Chinese)](https://huggingface.co/datasets/zxbsmk/webnovel_cn)

Please refer to [data/README.md](data/README.md) for details.

Some datasets require confirmation before using them, so we recommend logging in with your HuggingFace account using these commands.

```bash
pip install --upgrade huggingface_hub
huggingface-cli login
```

## Fine-Tuning Methods

Our script now supports the following fine-tuning methods:

- [LoRA](https://arxiv.org/abs/2106.09685)
  - Fine-tuning the low-rank adapters of the model.
- [P-Tuning V2](https://github.com/THUDM/P-tuning-v2)
  - Fine-tuning the prefix encoder of the model.
- [Freeze](https://arxiv.org/abs/2012.14913)
  - Fine-tuning the MLPs in the last n blocks of the model.
- Full Tuning
  - Fine-tuning all the parameters of the model.

## Requirement

- Python 3.8+ and PyTorch 1.13.1+
- 🤗Transformers, Datasets, Accelerate, PEFT and TRL
- fire, protobuf, cpm-kernels and sentencepiece
- jieba, rouge-chinese and nltk (used at evaluation)
- gradio and matplotlib (used in train_web.py)
- uvicorn, fastapi and sse-starlette (used in api_demo.py)

And **powerful GPUs**!

## Getting Started

### Data Preparation (optional)

Please refer to `data/example_dataset` for checking the details about the format of dataset files. You can either use a single `.json` file or a [dataset loading script](https://huggingface.co/docs/datasets/dataset_script) with multiple files to create a custom dataset.

Note: please update `data/dataset_info.json` to use your custom dataset. About the format of this file, please refer to `data/README.md`.

### Dependence Installation (optional)

```bash
git clone https://github.com/hiyouga/ChatGLM-Efficient-Tuning.git
conda create -n chatglm_etuning python=3.10
conda activate chatglm_etuning
cd ChatGLM-Efficient-Tuning
pip install -r requirements.txt
```

If you want to enable LoRA(QLoRA) or Freeze quantization on Windows, you will be required to install a pre-built version of `bitsandbytes` library, which supports CUDA 11.1 to 12.1.

```bash
pip install https://github.com/jllllll/bitsandbytes-windows-webui/releases/download/wheels/bitsandbytes-0.39.1-py3-none-win_amd64.whl
```

### Fine-tuning with a Single GPU

```bash
CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
    --stage sft \
    --model_name_or_path path_to_your_chatglm_model \
    --do_train \
    --dataset alpaca_gpt4_en \
    --finetuning_type lora \
    --output_dir path_to_sft_checkpoint \
    --per_device_train_batch_size 4 \
    --gradient_accumulation_steps 4 \
    --lr_scheduler_type cosine \
    --logging_steps 10 \
    --save_steps 1000 \
    --learning_rate 5e-5 \
    --num_train_epochs 3.0 \
    --fp16
```

Please refer to our [Wiki](https://github.com/hiyouga/ChatGLM-Efficient-Tuning/wiki) about the details of the arguments.

### Distributed Fine-tuning with Multiple GPUs

```bash
accelerate config # configure the environment
accelerate launch src/train_bash.py # arguments (same as above)
```

### Training Reward Model

```bash
CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
    --stage rm \
    --model_name_or_path path_to_your_chatglm_model \
    --do_train \
    --dataset comparison_gpt4_en \
    --finetuning_type lora \
    --output_dir path_to_rm_checkpoint \
    --per_device_train_batch_size 4 \
    --gradient_accumulation_steps 4 \
    --lr_scheduler_type cosine \
    --logging_steps 10 \
    --save_steps 1000 \
    --learning_rate 1e-5 \
    --num_train_epochs 1.0 \
    --fp16
```

### Training with RLHF

```bash
CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
    --stage ppo \
    --model_name_or_path path_to_your_chatglm_model \
    --do_train \
    --dataset alpaca_gpt4_en \
    --finetuning_type lora \
    --resume_lora_training False \
    --checkpoint_dir path_to_sft_checkpoint \
    --reward_model path_to_rm_checkpoint \
    --output_dir path_to_ppo_checkpoint \
    --per_device_train_batch_size 2 \
    --gradient_accumulation_steps 4 \
    --lr_scheduler_type cosine \
    --logging_steps 10 \
    --save_steps 1000 \
    --learning_rate 1e-5 \
    --num_train_epochs 1.0 \
    --fp16
```

### Evaluation (BLEU and ROUGE_CHINESE)

```bash
CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
    --stage sft \
    --model_name_or_path path_to_your_chatglm_model \
    --do_eval \
    --dataset alpaca_gpt4_en \
    --checkpoint_dir path_to_checkpoint \
    --output_dir path_to_eval_result \
    --per_device_eval_batch_size 8 \
    --max_samples 50 \
    --predict_with_generate
```

### Predict
```bash
CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
    --stage sft \
    --model_name_or_path path_to_your_chatglm_model \
    --do_predict \
    --dataset alpaca_gpt4_en \
    --checkpoint_dir path_to_checkpoint \
    --output_dir path_to_predict_result \
    --per_device_eval_batch_size 8 \
    --max_samples 50 \
    --predict_with_generate
```

### API / CLI Demo

```bash
python src/xxx_demo.py \
    --model_name_or_path path_to_your_chatglm_model \
    --checkpoint_dir path_to_checkpoint
```

### All-in-one Web UI

```bash
python src/train_web.py
```

### Export model

```bash
python src/export_model.py \
    --model_name_or_path path_to_your_chatglm_model \
    --checkpoint_dir path_to_checkpoint \
    --output_dir path_to_export
```

### Hardware Requirements

| Fine-tune method | Batch size | Mode |  GRAM  | Speed |
| ---------------- | ---------- | ---- | ------ | ----- |
| LoRA (r=8)       |     16     | FP16 |  28GB  | 8ex/s |
| LoRA (r=8)       |     8      | FP16 |  24GB  | 8ex/s |
| LoRA (r=8)       |     4      | FP16 |  20GB  | 8ex/s |
| LoRA (r=8)       |     4      | INT8 |  10GB  | 8ex/s |
| LoRA (r=8)       |     4      | INT4 |   8GB  | 8ex/s |
| P-Tuning (p=16)  |     4      | FP16 |  20GB  | 8ex/s |
| P-Tuning (p=16)  |     4      | INT8 |  16GB  | 8ex/s |
| P-Tuning (p=16)  |     4      | INT4 |  12GB  | 8ex/s |
| Freeze (l=3)     |     4      | FP16 |  24GB  | 8ex/s |
| Freeze (l=3)     |     4      | INT8 |  12GB  | 8ex/s |

| RM  method       | Batch size | Mode |  GRAM  | Speed |
| ---------------- | ---------- | ---- | ------ | ----- |
| LoRA (r=8) + rm  |     4      | FP16 |  22GB  | -     |
| LoRA (r=8) + rm  |     1      | INT8 |  11GB  | -     |

| RLHF method      | Batch size | Mode |  GRAM  | Speed |
| ---------------- | ---------- | ---- | ------ | ----- |
| LoRA (r=8) + ppo |     4      | FP16 |  23GB  | -     |
| LoRA (r=8) + ppo |     1      | INT8 |  12GB  | -     |

> Note: `r` is the lora rank, `p` is the number of prefix tokens, `l` is the number of trainable layers, `ex/s` is the examples per second at training. The `gradient_accumulation_steps` is set to `1`. All are evaluated on a single Tesla V100 (32G) GPU, they are approximated values and may vary in different GPUs.

## Fine-tuning ChatGLM: A Case

### Training Results

We use the whole `alpaca_gpt4_zh` dataset to fine-tune the ChatGLM model with LoRA (r=8) for one epoch, using the default hyper-parameters. The loss curve during training is presented below.

![training loss](assets/trainer_state.jpg)

### Evaluation Results

We select 100 instances in the `alpaca_gpt4_zh` dataset to evaluate the fine-tuned ChatGLM model and compute the BLEU and ROUGE scores. The results are presented below.

|   Score   | Original | FZ (l=2) | PT (p=16) | LoRA (r=8) |
| --------- | -------- | ----- | ----- | ----------------- |
| BLEU-4    |  15.75   | 16.85 | 16.06 | 17.01 (**+1.26**) |
| Rouge-1   |  34.51   | 36.62 | 34.80 | 36.77 (**+2.26**) |
| Rouge-2   |  15.11   | 17.04 | 15.32 | 16.83 (**+1.72**) |
| Rouge-l   |  26.18   | 28.17 | 26.35 | 28.86 (**+2.68**) |
| Params (%)|  /       | 4.35% | 0.06% | 0.06%             |

> FZ: freeze tuning, PT: P-Tuning V2 (we use `pre_seq_len=16` for fair comparison with LoRA), Params: the percentange of trainable parameters.

## Compared with Existing Implementations

- [THUDM/ChatGLM-6B](https://github.com/THUDM/ChatGLM-6B/tree/main/ptuning)
  - Official implementation of fine-tuning ChatGLM with [P-Tuning v2](https://github.com/THUDM/P-tuning-v2) on the [ADGEN](https://aclanthology.org/D19-1321.pdf) dataset.
  - Our fine-tuning script is largely depend on it. We further implement the [LoRA](https://arxiv.org/abs/2106.09685) tuning method. Additionally, we **dynamically** pad the inputs to the longest sequence in the batch instead of the maximum length, to accelerate the fine-tuning.
- [mymusise/ChatGLM-Tuning](https://github.com/mymusise/ChatGLM-Tuning)
  - An unoffical implementation of fine-tuning ChatGLM with [LoRA](https://arxiv.org/abs/2106.09685) on the [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca) dataset.
  - We borrowed some ideas from it. Our fine-tuning script **integrates** the data pre-processing part into the training procedure, so we need not generate a pre-processed dataset before training.
- [ssbuild/chatglm_finetuning](https://github.com/ssbuild/chatglm_finetuning)
  - An unofficial implementation of fine-tuning ChatGLM with several PEFT methods on the [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca) dataset.
  - Our fine-tuning script is implemented **purely** with [Huggingface transformers](https://github.com/huggingface/transformers) and is independent of the [deep_training](https://github.com/ssbuild/deep_training) framework.
- [lich99/ChatGLM-finetune-LoRA](https://github.com/lich99/ChatGLM-finetune-LoRA)
  - An unofficial implementation of fine-tuning ChatGLM with [LoRA](https://arxiv.org/abs/2106.09685) on the [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca) dataset.
  - We use the [Huggingface PEFT](https://github.com/huggingface/peft) to provide the state-of-the-art PEFT methods.
- [liucongg/ChatGLM-Finetuning](https://github.com/liucongg/ChatGLM-Finetuning)
  - An unofficial implementation of fine-tuning ChatGLM with several methods including Freeze, LoRA and P-Tuning on the industrial dataset.
  - We are aim to incorporate more instruction-following datasets for fine-tuning the ChatGLM model.
- [yanqiangmiffy/InstructGLM](https://github.com/yanqiangmiffy/InstructGLM)
  - An unofficial implementation of fine-tuning ChatGLM that explores the ChatGLM's ability on the instruction-following datasets.
  - Our fine-tuning script integrates the data pre-processing part in to the training procedure.

## TODO

- [ ] Employing [LangChain](https://github.com/hwchase17/langchain) to easily build applications that are capable of leveraging external knowledge upon fine-tuned ChatGLM models.
- [ ] Implementing the alignment algorithms to align human preferrences.
  - [x] [RLHF](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-chat)
  - [ ] [RRHF](https://github.com/GanjinZero/RRHF)
  - [ ] [RAFT](https://github.com/OptimalScale/LMFlow)
- [ ] Incorporating [Chinese datasets](https://github.com/brightmart/nlp_chinese_corpus) into the training sets.
  - [x] [BELLE](https://github.com/LianjiaTech/BELLE)
  - [ ] [pCLUE](https://github.com/CLUEbenchmark/pCLUE)
  - [ ] [CLUECorpus](https://github.com/CLUEbenchmark/CLUECorpus2020)
  - [x] [GuanacoDataset](https://huggingface.co/datasets/JosephusCheung/GuanacoDataset)
  - [x] [FireflyDataset](https://huggingface.co/datasets/YeungNLP/firefly-train-1.1M)
- [ ] Incorporating [ChatGPT](https://openai.com/blog/chatgpt) & [GPT-4](https://openai.com/research/gpt-4) self-chat data into the training sets.
  - [ ] [Baize](https://github.com/project-baize/baize-chatbot)
  - [x] [GPT-4-LLM](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
- [x] Implementing the Freeze-Tuning and P-Tuning method.
- [x] Supporting Multi-GPUs fine-tuning.
- [x] Adding script for evaluation.
- [x] Loading from checkpoint.
- [x] Fine-tuning the quantized model.
- [x] Writing a guidebook about how to fine-tune ChatGLM with this framework.
- [ ] Combining with state-of-the-art model editing algorithms. (*e.g. [MEND](https://arxiv.org/abs/2110.11309)*)
- [x] Incorporating the [OpenAssistant Conversations Dataset](https://huggingface.co/datasets/OpenAssistant/oasst1) for SFT and alignment.
- [ ] Incorporating the high quality Chinese instruction dataset [COIG](https://huggingface.co/datasets/BAAI/COIG).

## License

This repository is licensed under the [Apache-2.0 License](LICENSE). Please follow the [Model License](https://github.com/THUDM/ChatGLM-6B/blob/main/MODEL_LICENSE) to use ChatGLM-6B model.

## Citation

If this work is helpful, please cite as:

```bibtex
@Misc{chatglm-efficient-tuning,
  title = {ChatGLM Efficient Tuning},
  author = {hiyouga},
  howpublished = {\url{https://github.com/hiyouga/ChatGLM-Efficient-Tuning}},
  year = {2023}
}
```

## Acknowledgement

This repo benefits from [ChatGLM-6B](https://github.com/THUDM/ChatGLM-6B), [ChatGLM-Tuning](https://github.com/mymusise/ChatGLM-Tuning) and [yuanzhoulvpi2017/zero_nlp](https://github.com/yuanzhoulvpi2017/zero_nlp). Thanks for their wonderful works.

## Star History

![Star History Chart](https://api.star-history.com/svg?repos=hiyouga/ChatGLM-Efficient-Tuning&type=Date)
