release v0.7.1

Former-commit-id: a4f8adb021b6218d624303b51cd5e93ffa3111a1

.dockerignore

@@ -0,0 +1,11 @@
+.vscode
+.git
+.github
+.venv
+cache
+data
+examples
+.dockerignore
+.gitattributes
+.gitignore
+Dockerfile

.github/CODE_OF_CONDUCT.md

@@ -0,0 +1,128 @@
+# Contributor Covenant Code of Conduct
+
+## Our Pledge
+
+We as members, contributors, and leaders pledge to make participation in our
+community a harassment-free experience for everyone, regardless of age, body
+size, visible or invisible disability, ethnicity, sex characteristics, gender
+identity and expression, level of experience, education, socio-economic status,
+nationality, personal appearance, race, religion, or sexual identity
+and orientation.
+
+We pledge to act and interact in ways that contribute to an open, welcoming,
+diverse, inclusive, and healthy community.
+
+## Our Standards
+
+Examples of behavior that contributes to a positive environment for our
+community include:
+
+* Demonstrating empathy and kindness toward other people
+* Being respectful of differing opinions, viewpoints, and experiences
+* Giving and gracefully accepting constructive feedback
+* Accepting responsibility and apologizing to those affected by our mistakes,
+  and learning from the experience
+* Focusing on what is best not just for us as individuals, but for the
+  overall community
+
+Examples of unacceptable behavior include:
+
+* The use of sexualized language or imagery, and sexual attention or
+  advances of any kind
+* Trolling, insulting or derogatory comments, and personal or political attacks
+* Public or private harassment
+* Publishing others' private information, such as a physical or email
+  address, without their explicit permission
+* Other conduct which could reasonably be considered inappropriate in a
+  professional setting
+
+## Enforcement Responsibilities
+
+Community leaders are responsible for clarifying and enforcing our standards of
+acceptable behavior and will take appropriate and fair corrective action in
+response to any behavior that they deem inappropriate, threatening, offensive,
+or harmful.
+
+Community leaders have the right and responsibility to remove, edit, or reject
+comments, commits, code, wiki edits, issues, and other contributions that are
+not aligned to this Code of Conduct, and will communicate reasons for moderation
+decisions when appropriate.
+
+## Scope
+
+This Code of Conduct applies within all community spaces, and also applies when
+an individual is officially representing the community in public spaces.
+Examples of representing our community include using an official e-mail address,
+posting via an official social media account, or acting as an appointed
+representative at an online or offline event.
+
+## Enforcement
+
+Instances of abusive, harassing, or otherwise unacceptable behavior may be
+reported to the community leaders responsible for enforcement at
+`hoshihiyouga AT gmail DOT com`.
+All complaints will be reviewed and investigated promptly and fairly.
+
+All community leaders are obligated to respect the privacy and security of the
+reporter of any incident.
+
+## Enforcement Guidelines
+
+Community leaders will follow these Community Impact Guidelines in determining
+the consequences for any action they deem in violation of this Code of Conduct:
+
+### 1. Correction
+
+**Community Impact**: Use of inappropriate language or other behavior deemed
+unprofessional or unwelcome in the community.
+
+**Consequence**: A private, written warning from community leaders, providing
+clarity around the nature of the violation and an explanation of why the
+behavior was inappropriate. A public apology may be requested.
+
+### 2. Warning
+
+**Community Impact**: A violation through a single incident or series
+of actions.
+
+**Consequence**: A warning with consequences for continued behavior. No
+interaction with the people involved, including unsolicited interaction with
+those enforcing the Code of Conduct, for a specified period of time. This
+includes avoiding interactions in community spaces as well as external channels
+like social media. Violating these terms may lead to a temporary or
+permanent ban.
+
+### 3. Temporary Ban
+
+**Community Impact**: A serious violation of community standards, including
+sustained inappropriate behavior.
+
+**Consequence**: A temporary ban from any sort of interaction or public
+communication with the community for a specified period of time. No public or
+private interaction with the people involved, including unsolicited interaction
+with those enforcing the Code of Conduct, is allowed during this period.
+Violating these terms may lead to a permanent ban.
+
+### 4. Permanent Ban
+
+**Community Impact**: Demonstrating a pattern of violation of community
+standards, including sustained inappropriate behavior,  harassment of an
+individual, or aggression toward or disparagement of classes of individuals.
+
+**Consequence**: A permanent ban from any sort of public interaction within
+the community.
+
+## Attribution
+
+This Code of Conduct is adapted from the [Contributor Covenant][homepage],
+version 2.0, available at
+https://www.contributor-covenant.org/version/2/0/code_of_conduct.html.
+
+Community Impact Guidelines were inspired by [Mozilla's code of conduct
+enforcement ladder](https://github.com/mozilla/diversity).
+
+[homepage]: https://www.contributor-covenant.org
+
+For answers to common questions about this code of conduct, see the FAQ at
+https://www.contributor-covenant.org/faq. Translations are available at
+https://www.contributor-covenant.org/translations.

.github/CONTRIBUTING.md

@@ -0,0 +1,21 @@
+# Contributing to LLaMA Factory
+
+Everyone is welcome to contribute, and we value everybody's contribution. Code contributions are not the only way to help the community. Answering questions, helping others, and improving the documentation are also immensely valuable.
+
+It also helps us if you spread the word! Reference the library in blog posts about the awesome projects it made possible, shout out on Twitter every time it has helped you, or simply ⭐️ the repository to say thank you.
+
+However you choose to contribute, please be mindful and respect our [code of conduct](CODE_OF_CONDUCT.md).
+
+**This guide was heavily inspired by [transformers guide to contributing](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md).**
+
+## Ways to contribute
+
+There are several ways you can contribute to LLaMA Factory:
+
+* Fix outstanding issues with the existing code.
+* Submit issues related to bugs or desired new features.
+* Contribute to the examples or to the documentation.
+
+### Style guide
+
+LLaMA Factory follows the [Google Python Style Guide](https://google.github.io/styleguide/pyguide.html), check it for details.

.github/ISSUE_TEMPLATE/bug-report.yml

@@ -0,0 +1,58 @@
+name: "\U0001F41B Bug / Help"
+description: Create a report to help us improve the LLaMA Factory
+body:
+  - type: checkboxes
+    id: reminder
+    attributes:
+      label: Reminder
+      description: |
+        Please ensure you have read the README carefully and searched the existing issues.
+        请确保您已经认真阅读了 README 并且搜索过现有的 Issue。
+
+      options:
+        - label: I have read the README and searched the existing issues.
+          required: true
+
+  - type: textarea
+    id: reproduction
+    validations:
+      required: true
+    attributes:
+      label: Reproduction
+      description: |
+        Please provide code snippets, error messages and stack traces that reproduces the problem.
+        请提供运行参数，错误信息以及异常堆栈以便于我们复现该问题。
+        Remember to use Markdown tags to correctly format your code.
+        请合理使用 Markdown 标签来格式化您的文本。
+
+      placeholder: |
+        python src/train_bash.py ...
+
+  - type: textarea
+    id: expected-behavior
+    validations:
+      required: false
+    attributes:
+      label: Expected behavior
+      description: |
+        Please provide a clear and concise description of what you would expect to happen.
+        请提供您原本的目的，即这段代码的期望行为。
+
+  - type: textarea
+    id: system-info
+    validations:
+      required: false
+    attributes:
+      label: System Info
+      description: |
+        Please share your system info with us. You can run the command **transformers-cli env** and copy-paste its output below.
+        请提供您的系统信息。您可以在命令行运行 **transformers-cli env** 并将其输出复制到该文本框中。
+
+      placeholder: transformers version, platform, python version, ...
+
+  - type: textarea
+    id: others
+    validations:
+      required: false
+    attributes:
+      label: Others

.github/PULL_REQUEST_TEMPLATE.md

@@ -0,0 +1,7 @@
+# What does this PR do?
+
+Fixes # (issue)
+
+## Before submitting
+
+- [ ] Did you read the [contributor guideline](https://github.com/hiyouga/LLaMA-Factory/blob/main/.github/CONTRIBUTING.md)?

.github/SECURITY.md

@@ -0,0 +1,7 @@
+# Reporting Security Issues
+
+To report a security issue, please use the GitHub Security Advisory ["Report a Vulnerability"](https://github.com/hiyouga/LLaMA-Factory/security/advisories/new) tab.
+
+We will send a response indicating the next steps in handling your report. After the initial reply to your report, the security team will keep you informed of the progress towards a fix and full announcement, and may ask for additional information or guidance.
+
+Report security bugs in third-party modules to the person or team maintaining the module.

.github/workflows/tests.yml

@@ -0,0 +1,29 @@
+name: tests
+
+on:
+  push:
+    branches: [ "main" ]
+  pull_request:
+    branches: [ "main" ]
+
+jobs:
+  check_code_quality:
+
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.8"
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install ruff
+
+      - name: Check quality
+        run: |
+            make style && make quality

.gitignore

@@ -0,0 +1,165 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/#use-with-ide
+.pdm.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+.idea/
+
+# custom .gitignore
+user.config
+saves/
+cache/

CITATION.cff

@@ -0,0 +1,37 @@
+cff-version: 1.2.0
+date-released: 2024-03
+message: "If you use this software, please cite it as below."
+authors:
+- family-names: "Zheng"
+  given-names: "Yaowei"
+- family-names: "Zhang"
+  given-names: "Richong"
+- family-names: "Zhang"
+  given-names: "Junhao"
+- family-names: "Ye"
+  given-names: "Yanhan"
+- family-names: "Luo"
+  given-names: "Zheyan"
+- family-names: "Ma"
+  given-names: "Yongqiang"
+title: "LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models"
+url: "https://arxiv.org/abs/2403.13372"
+preferred-citation:
+  type: article
+  authors:
+    - family-names: "Zheng"
+      given-names: "Yaowei"
+    - family-names: "Zhang"
+      given-names: "Richong"
+    - family-names: "Zhang"
+      given-names: "Junhao"
+    - family-names: "Ye"
+      given-names: "Yanhan"
+    - family-names: "Luo"
+      given-names: "Zheyan"
+    - family-names: "Ma"
+      given-names: "Yongqiang"
+  journal: "arXiv preprint arXiv:2403.13372"
+  title: "LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models"
+  url: "https://arxiv.org/abs/2403.13372"
+  year: 2024

Dockerfile

@@ -0,0 +1,14 @@
+FROM nvcr.io/nvidia/pytorch:24.01-py3
+
+WORKDIR /app
+
+COPY requirements.txt /app/
+RUN pip install -r requirements.txt
+
+COPY . /app/
+RUN pip install -e .[deepspeed,metrics,bitsandbytes,qwen]
+
+VOLUME [ "/root/.cache/huggingface/", "/app/data", "/app/output" ]
+EXPOSE 7860
+
+CMD [ "llamafactory-cli", "webui" ]

Makefile

@@ -0,0 +1,11 @@
+.PHONY: quality style
+
+check_dirs := scripts src tests
+
+quality:
+	ruff check $(check_dirs)
+	ruff format --check $(check_dirs)
+
+style:
+	ruff check $(check_dirs) --fix
+	ruff format $(check_dirs)

README.md

@@ -1,104 +1,274 @@
-# LLaMA Efficient Tuning
+![# LLaMA Factory](assets/logo.png)
 
-[![GitHub Repo stars](https://img.shields.io/github/stars/hiyouga/LLaMA-Efficient-Tuning?style=social)](https://github.com/hiyouga/LLaMA-Efficient-Tuning/stargazers)
-[![GitHub Code License](https://img.shields.io/github/license/hiyouga/LLaMA-Efficient-Tuning)](LICENSE)
-[![GitHub last commit](https://img.shields.io/github/last-commit/hiyouga/LLaMA-Efficient-Tuning)](https://github.com/hiyouga/LLaMA-Efficient-Tuning/commits/main)
+[![GitHub Repo stars](https://img.shields.io/github/stars/hiyouga/LLaMA-Factory?style=social)](https://github.com/hiyouga/LLaMA-Factory/stargazers)
+[![GitHub Code License](https://img.shields.io/github/license/hiyouga/LLaMA-Factory)](LICENSE)
+[![GitHub last commit](https://img.shields.io/github/last-commit/hiyouga/LLaMA-Factory)](https://github.com/hiyouga/LLaMA-Factory/commits/main)
 [![PyPI](https://img.shields.io/pypi/v/llmtuner)](https://pypi.org/project/llmtuner/)
-[![GitHub pull request](https://img.shields.io/badge/PRs-welcome-blue)](https://github.com/hiyouga/LLaMA-Efficient-Tuning/pulls)
+[![Downloads](https://static.pepy.tech/badge/llmtuner)](https://pypi.org/project/llmtuner/)
+[![Citation](https://img.shields.io/badge/citation-44-green)](#projects-using-llama-factory)
+[![GitHub pull request](https://img.shields.io/badge/PRs-welcome-blue)](https://github.com/hiyouga/LLaMA-Factory/pulls)
+[![Discord](https://dcbadge.vercel.app/api/server/rKfvV9r9FK?compact=true&style=flat)](https://discord.gg/rKfvV9r9FK)
+[![Twitter](https://img.shields.io/twitter/follow/llamafactory_ai)](https://twitter.com/llamafactory_ai)
+[![Spaces](https://img.shields.io/badge/🤗-Open%20in%20Spaces-blue)](https://huggingface.co/spaces/hiyouga/LLaMA-Board)
+[![Studios](https://img.shields.io/badge/ModelScope-Open%20in%20Studios-blue)](https://modelscope.cn/studios/hiyouga/LLaMA-Board)
+[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1eRTPn37ltBbYsISy9Aw2NuI2Aq5CQrD9?usp=sharing)
+
+[![GitHub Tread](https://trendshift.io/api/badge/repositories/4535)](https://trendshift.io/repositories/4535)
 
 👋 Join our [WeChat](assets/wechat.jpg).
 
 \[ English | [中文](README_zh.md) \]
 
+**Fine-tuning a large language model can be easy as...**
+
+https://github.com/hiyouga/LLaMA-Factory/assets/16256802/9840a653-7e9c-41c8-ae89-7ace5698baf6
+
+Choose your path:
+
+- **Colab**: https://colab.research.google.com/drive/1eRTPn37ltBbYsISy9Aw2NuI2Aq5CQrD9?usp=sharing
+- **Local machine**: Please refer to [usage](#getting-started)
+
+## Table of Contents
+
+- [Features](#features)
+- [Benchmark](#benchmark)
+- [Changelog](#changelog)
+- [Supported Models](#supported-models)
+- [Supported Training Approaches](#supported-training-approaches)
+- [Provided Datasets](#provided-datasets)
+- [Requirement](#requirement)
+- [Getting Started](#getting-started)
+- [Projects using LLaMA Factory](#projects-using-llama-factory)
+- [License](#license)
+- [Citation](#citation)
+- [Acknowledgement](#acknowledgement)
+
+## Features
+
+- **Various models**: LLaMA, LLaVA, Mistral, Mixtral-MoE, Qwen, Yi, Gemma, Baichuan, ChatGLM, Phi, etc.
+- **Integrated methods**: (Continuous) pre-training, (multimodal) supervised fine-tuning, reward modeling, PPO, DPO and ORPO.
+- **Scalable resources**: 32-bit full-tuning, 16-bit freeze-tuning, 16-bit LoRA and 2/4/8-bit QLoRA via AQLM/AWQ/GPTQ/LLM.int8.
+- **Advanced algorithms**: GaLore, BAdam, DoRA, LongLoRA, LLaMA Pro, Mixture-of-Depths, LoRA+, LoftQ and Agent tuning.
+- **Practical tricks**: FlashAttention-2, Unsloth, RoPE scaling, NEFTune and rsLoRA.
+- **Experiment monitors**: LlamaBoard, TensorBoard, Wandb, MLflow, etc.
+- **Faster inference**: OpenAI-style API, Gradio UI and CLI with vLLM worker.
+
+## Benchmark
+
+Compared to ChatGLM's [P-Tuning](https://github.com/THUDM/ChatGLM2-6B/tree/main/ptuning), LLaMA Factory's LoRA tuning offers up to **3.7 times faster** training speed with a better Rouge score on the advertising text generation task. By leveraging 4-bit quantization technique, LLaMA Factory's QLoRA further improves the efficiency regarding the GPU memory.
+
+![benchmark](assets/benchmark.svg)
+
+<details><summary>Definitions</summary>
+
+- **Training Speed**: the number of training samples processed per second during the training. (bs=4, cutoff_len=1024)
+- **Rouge Score**: Rouge-2 score on the development set of the [advertising text generation](https://aclanthology.org/D19-1321.pdf) task. (bs=4, cutoff_len=1024)
+- **GPU Memory**: Peak GPU memory usage in 4-bit quantized training. (bs=1, cutoff_len=1024)
+- We adopt `pre_seq_len=128` for ChatGLM's P-Tuning and `lora_rank=32` for LLaMA Factory's LoRA tuning.
+
+</details>
+
 ## Changelog
 
-[23/07/31] Now we support dataset streaming. Try `--streaming` and `--max_steps 100` arguments to stream your dataset.
+[24/05/14] We supported training and inference on the Ascend NPU devices. Check [installation](#installation) section for details.
 
-[23/07/29] We release two instruction-tuned 13B models at Hugging Face. See these Hugging Face Repos ([LLaMA-2](https://huggingface.co/hiyouga/Llama-2-Chinese-13b-chat) / [Baichuan](https://huggingface.co/hiyouga/baichuan-13b-sft)) for details.
+[24/05/13] We supported fine-tuning the **Yi-1.5** series models.
 
-[23/07/19] Now we support training the **LLaMA-2** models in this repo. Try `--model_name_or_path meta-llama/Llama-2-7b-hf` argument to use the LLaMA-2 model. Remember to use `--template llama2` argument when you are using the LLaMA-2-chat model.
+[24/04/26] We supported fine-tuning the **LLaVA-1.5** multimodal LLMs. See [examples](examples/README.md) for usage.
 
-[23/07/18] Now we develop an all-in-one Web UI for training, evaluation and inference. Try `train_web.py` to fine-tune models in your Web browser. Thank [@KanadeSiina](https://github.com/KanadeSiina) and [@codemayq](https://github.com/codemayq) for their efforts in the development.
+<details><summary>Full Changelog</summary>
 
-[23/07/11] Now we support training the **Baichuan-13B** model in this repo. Try `--model_name_or_path baichuan-inc/Baichuan-13B-Base` and `--lora_target W_pack` arguments to train the Baichuan-13B model. Remember to use `--template baichuan` argument when you are using the Baichuan-13B-Chat model.
+[24/04/22] We provided a **[Colab notebook](https://colab.research.google.com/drive/1eRTPn37ltBbYsISy9Aw2NuI2Aq5CQrD9?usp=sharing)** for fine-tuning the Llama-3 model on a free T4 GPU. Two Llama-3-derived models fine-tuned using LLaMA Factory are available at Hugging Face, check [Llama3-8B-Chinese-Chat](https://huggingface.co/shenzhi-wang/Llama3-8B-Chinese-Chat) and [Llama3-Chinese](https://huggingface.co/zhichen/Llama3-Chinese) for details.
 
-[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.
+[24/04/21] We supported **[Mixture-of-Depths](https://arxiv.org/abs/2404.02258)** according to [AstraMindAI's implementation](https://github.com/astramind-ai/Mixture-of-depths). See [examples](examples/README.md) for usage.
 
-[23/07/07] Now we support training the **InternLM-7B** model in this repo. Try `--model_name_or_path internlm/internlm-7b` argument to use the InternLM model. Remember to use `--template intern` argument when you are using the InternLM-chat model.
+[24/04/16] We supported **[BAdam](https://arxiv.org/abs/2404.02827)**. See [examples](examples/README.md) for usage.
 
-[23/07/05] Now we support training the **Falcon-7B/40B** models in this repo. Try `--model_name_or_path tiiuae/falcon-7b` and `--lora_target query_key_value` arguments to use the Falcon model.
+[24/04/16] We supported **[unsloth](https://github.com/unslothai/unsloth)**'s long-sequence training (Llama-2-7B-56k within 24GB). It achieves **117%** speed and **50%** memory compared with FlashAttention-2, more benchmarks can be found in [this page](https://github.com/hiyouga/LLaMA-Factory/wiki/Performance-comparison).
 
-[23/06/29] We provide a **reproducible example** of training a chat model using instruction-following datasets, see this [Hugging Face Repo](https://huggingface.co/hiyouga/baichuan-7b-sft) for details.
+[24/03/31] We supported **[ORPO](https://arxiv.org/abs/2403.07691)**. See [examples](examples/README.md) for usage.
 
-[23/06/22] 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**.
+[24/03/21] Our paper "[LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models](https://arxiv.org/abs/2403.13372)" is available at arXiv!
 
-[23/06/15] Now we support training the **Baichuan-7B** model in this repo. Try `--model_name_or_path baichuan-inc/Baichuan-7B` and `--lora_target W_pack` arguments to use the Baichuan-7B model.
+[24/03/20] We supported **FSDP+QLoRA** that fine-tunes a 70B model on 2x24GB GPUs. See [examples](examples/README.md) for usage.
 
-[23/06/03] Now we support quantized training and inference (aka **[QLoRA](https://github.com/artidoro/qlora)**). Try `--quantization_bit 4/8` argument to work with quantized models.
+[24/03/13] We supported **[LoRA+](https://arxiv.org/abs/2402.12354)**. See [examples](examples/README.md) for usage.
 
-[23/05/31] Now we support training the **BLOOM & BLOOMZ** models in this repo. Try `--model_name_or_path bigscience/bloomz-7b1-mt` and `--lora_target query_key_value` arguments to use the BLOOMZ model.
+[24/03/07] We supported gradient low-rank projection (**[GaLore](https://arxiv.org/abs/2403.03507)**) algorithm. See [examples](examples/README.md) for usage.
+
+[24/03/07] We integrated **[vLLM](https://github.com/vllm-project/vllm)** for faster and concurrent inference. Try `infer_backend: vllm` to enjoy **270%** inference speed.
+
+[24/02/28] We supported weight-decomposed LoRA (**[DoRA](https://arxiv.org/abs/2402.09353)**). Try `use_dora: true` to activate DoRA training.
+
+[24/02/15] We supported **block expansion** proposed by [LLaMA Pro](https://github.com/TencentARC/LLaMA-Pro). See [examples](examples/README.md) for usage.
+
+[24/02/05] Qwen1.5 (Qwen2 beta version) series models are supported in LLaMA-Factory. Check this [blog post](https://qwenlm.github.io/blog/qwen1.5/) for details.
+
+[24/01/18] We supported **agent tuning** for most models, equipping model with tool using abilities by fine-tuning with `dataset: glaive_toolcall`.
+
+[23/12/23] We supported **[unsloth](https://github.com/unslothai/unsloth)**'s implementation to boost LoRA tuning for the LLaMA, Mistral and Yi models. Try `use_unsloth: true` argument to activate unsloth patch. It achieves **170%** speed in our benchmark, check [this page](https://github.com/hiyouga/LLaMA-Factory/wiki/Performance-comparison) for details.
+
+[23/12/12] We supported fine-tuning the latest MoE model **[Mixtral 8x7B](https://huggingface.co/mistralai/Mixtral-8x7B-v0.1)** in our framework. See hardware requirement [here](#hardware-requirement).
+
+[23/12/01] We supported downloading pre-trained models and datasets from the **[ModelScope Hub](https://modelscope.cn/models)** for Chinese mainland users. See [this tutorial](#download-from-modelscope-hub) for usage.
+
+[23/10/21] We supported **[NEFTune](https://arxiv.org/abs/2310.05914)** trick for fine-tuning. Try `neftune_noise_alpha: 5` argument to activate NEFTune.
+
+[23/09/27] We supported **$S^2$-Attn** proposed by [LongLoRA](https://github.com/dvlab-research/LongLoRA) for the LLaMA models. Try `shift_attn: true` argument to enable shift short attention.
+
+[23/09/23] We integrated MMLU, C-Eval and CMMLU benchmarks in this repo. See [examples](examples/README.md) for usage.
+
+[23/09/10] We supported **[FlashAttention-2](https://github.com/Dao-AILab/flash-attention)**. Try `flash_attn: fa2` argument to enable FlashAttention-2 if you are using RTX4090, A100 or H100 GPUs.
+
+[23/08/12] We supported **RoPE scaling** to extend the context length of the LLaMA models. Try `rope_scaling: linear` argument in training and `rope_scaling: dynamic` argument at inference to extrapolate the position embeddings.
+
+[23/08/11] We supported **[DPO training](https://arxiv.org/abs/2305.18290)** for instruction-tuned models. See [examples](examples/README.md) for usage.
+
+[23/07/31] We supported **dataset streaming**. Try `streaming: true` and `max_steps: 10000` arguments to load your dataset in streaming mode.
+
+[23/07/29] We released two instruction-tuned 13B models at Hugging Face. See these Hugging Face Repos ([LLaMA-2](https://huggingface.co/hiyouga/Llama-2-Chinese-13b-chat) / [Baichuan](https://huggingface.co/hiyouga/Baichuan-13B-sft)) for details.
+
+[23/07/18] We developed an **all-in-one Web UI** for training, evaluation and inference. Try `train_web.py` to fine-tune models in your Web browser. Thank [@KanadeSiina](https://github.com/KanadeSiina) and [@codemayq](https://github.com/codemayq) for their efforts in the development.
+
+[23/07/09] We released **[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.
+
+[23/06/29] We provided a **reproducible example** of training a chat model using instruction-following datasets, see [Baichuan-7B-sft](https://huggingface.co/hiyouga/Baichuan-7B-sft) for details.
+
+[23/06/22] We aligned 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**.
+
+[23/06/03] We supported quantized training and inference (aka **[QLoRA](https://github.com/artidoro/qlora)**). See [examples](examples/README.md) for usage.
+
+</details>
 
 ## Supported Models
 
-- [LLaMA](https://github.com/facebookresearch/llama) (7B/13B/33B/65B)
-- [LLaMA-2](https://huggingface.co/meta-llama) (7B/13B/70B)
-- [BLOOM](https://huggingface.co/bigscience/bloom) & [BLOOMZ](https://huggingface.co/bigscience/bloomz) (560M/1.1B/1.7B/3B/7.1B/176B)
-- [Falcon](https://huggingface.co/tiiuae/falcon-7b) (7B/40B)
-- [Baichuan](https://huggingface.co/baichuan-inc/baichuan-7B) (7B/13B)
-- [InternLM](https://github.com/InternLM/InternLM) (7B)
+| Model                                                    | Model size                       | Default module    | Template  |
+| -------------------------------------------------------- | -------------------------------- | ----------------- | --------- |
+| [Baichuan2](https://huggingface.co/baichuan-inc)         | 7B/13B                           | W_pack            | baichuan2 |
+| [BLOOM](https://huggingface.co/bigscience)               | 560M/1.1B/1.7B/3B/7.1B/176B      | query_key_value   | -         |
+| [BLOOMZ](https://huggingface.co/bigscience)              | 560M/1.1B/1.7B/3B/7.1B/176B      | query_key_value   | -         |
+| [ChatGLM3](https://huggingface.co/THUDM)                 | 6B                               | query_key_value   | chatglm3  |
+| [Command-R](https://huggingface.co/CohereForAI)          | 35B/104B                         | q_proj,v_proj     | cohere    |
+| [DeepSeek (MoE)](https://huggingface.co/deepseek-ai)     | 7B/16B/67B/236B                  | q_proj,v_proj     | deepseek  |
+| [Falcon](https://huggingface.co/tiiuae)                  | 7B/40B/180B                      | query_key_value   | falcon    |
+| [Gemma/CodeGemma](https://huggingface.co/google)         | 2B/7B                            | q_proj,v_proj     | gemma     |
+| [InternLM2](https://huggingface.co/internlm)             | 7B/20B                           | wqkv              | intern2   |
+| [LLaMA](https://github.com/facebookresearch/llama)       | 7B/13B/33B/65B                   | q_proj,v_proj     | -         |
+| [LLaMA-2](https://huggingface.co/meta-llama)             | 7B/13B/70B                       | q_proj,v_proj     | llama2    |
+| [LLaMA-3](https://huggingface.co/meta-llama)             | 8B/70B                           | q_proj,v_proj     | llama3    |
+| [LLaVA-1.5](https://huggingface.co/llava-hf)             | 7B/13B                           | q_proj,v_proj     | vicuna    |
+| [Mistral/Mixtral](https://huggingface.co/mistralai)      | 7B/8x7B/8x22B                    | q_proj,v_proj     | mistral   |
+| [OLMo](https://huggingface.co/allenai)                   | 1B/7B                            | q_proj,v_proj     | -         |
+| [Phi-1.5/2](https://huggingface.co/microsoft)            | 1.3B/2.7B                        | q_proj,v_proj     | -         |
+| [Phi-3](https://huggingface.co/microsoft)                | 3.8B                             | qkv_proj          | phi       |
+| [Qwen](https://huggingface.co/Qwen)                      | 1.8B/7B/14B/72B                  | c_attn            | qwen      |
+| [Qwen1.5 (Code/MoE)](https://huggingface.co/Qwen)        | 0.5B/1.8B/4B/7B/14B/32B/72B/110B | q_proj,v_proj     | qwen      |
+| [StarCoder2](https://huggingface.co/bigcode)             | 3B/7B/15B                        | q_proj,v_proj     | -         |
+| [XVERSE](https://huggingface.co/xverse)                  | 7B/13B/65B                       | q_proj,v_proj     | xverse    |
+| [Yi (1/1.5)](https://huggingface.co/01-ai)               | 6B/9B/34B                        | q_proj,v_proj     | yi        |
+| [Yi-VL](https://huggingface.co/01-ai)                    | 6B/34B                           | q_proj,v_proj     | yi_vl     |
+| [Yuan](https://huggingface.co/IEITYuan)                  | 2B/51B/102B                      | q_proj,v_proj     | yuan      |
+
+> [!NOTE]
+> **Default module** is used for the `--lora_target` argument, you can use `--lora_target all` to specify all the available modules for better convergence.
+>
+> For the "base" models, the `--template` argument can be chosen from `default`, `alpaca`, `vicuna` etc. But make sure to use the **corresponding template** for the "instruct/chat" models.
+>
+> Remember to use the **SAME** template in training and inference.
+
+Please refer to [constants.py](src/llmtuner/extras/constants.py) for a full list of models we supported.
+
+You also can add a custom chat template to [template.py](src/llmtuner/data/template.py).
 
 ## Supported Training Approaches
 
-- [(Continually) pre-training](https://s3-us-west-2.amazonaws.com/openai-assets/research-covers/language-unsupervised/language_understanding_paper.pdf)
-  - Full-parameter tuning
-  - Partial-parameter tuning
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
-- [Supervised fine-tuning](https://arxiv.org/abs/2109.01652)
-  - Full-parameter tuning
-  - Partial-parameter tuning
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
-- [RLHF](https://arxiv.org/abs/2203.02155)
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
+| Approach               |     Full-tuning    |    Freeze-tuning   |       LoRA         |       QLoRA        |
+| ---------------------- | ------------------ | ------------------ | ------------------ | ------------------ |
+| Pre-Training           | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| Supervised Fine-Tuning | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| Reward Modeling        | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| PPO Training           | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| DPO Training           | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| ORPO Training          | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
 
 ## Provided Datasets
 
-- For pre-training:
-  - [Wiki Demo (en)](data/wiki_demo.txt)
-  - [RefinedWeb (en)](https://huggingface.co/datasets/tiiuae/falcon-refinedweb)
-  - [StarCoder (en)](https://huggingface.co/datasets/bigcode/starcoderdata)
-  - [Wikipedia (en)](https://huggingface.co/datasets/olm/olm-wikipedia-20221220)
-  - [Wikipedia (zh)](https://huggingface.co/datasets/pleisto/wikipedia-cn-20230720-filtered)
-- For supervised fine-tuning:
-  - [Stanford Alpaca (en)](https://github.com/tatsu-lab/stanford_alpaca)
-  - [Stanford Alpaca (zh)](https://github.com/ymcui/Chinese-LLaMA-Alpaca)
-  - [GPT-4 Generated Data (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
-  - [Open Assistant (multilingual)](https://huggingface.co/datasets/OpenAssistant/oasst1)
-  - [Self-cognition (zh)](data/self_cognition.json)
-  - [ShareGPT (zh)](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT/tree/main/Chinese-instruction-collection)
-  - [RefGPT (zh)](https://github.com/sufengniu/RefGPT)
-  - [Guanaco Dataset (multilingual)](https://huggingface.co/datasets/JosephusCheung/GuanacoDataset)
-  - [BELLE 2M (zh)](https://huggingface.co/datasets/BelleGroup/train_2M_CN)
-  - [BELLE 1M (zh)](https://huggingface.co/datasets/BelleGroup/train_1M_CN)
-  - [BELLE 0.5M (zh)](https://huggingface.co/datasets/BelleGroup/train_0.5M_CN)
-  - [BELLE Dialogue 0.4M (zh)](https://huggingface.co/datasets/BelleGroup/generated_chat_0.4M)
-  - [BELLE School Math 0.25M (zh)](https://huggingface.co/datasets/BelleGroup/school_math_0.25M)
-  - [BELLE Multiturn Chat 0.8M (zh)](https://huggingface.co/datasets/BelleGroup/multiturn_chat_0.8M)
-  - [Firefly 1.1M (zh)](https://huggingface.co/datasets/YeungNLP/firefly-train-1.1M)
-  - [LIMA (en)](https://huggingface.co/datasets/GAIR/lima)
-  - [CodeAlpaca 20k (en)](https://huggingface.co/datasets/sahil2801/CodeAlpaca-20k)
-  - [Alpaca CoT (multilingual)](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT)
-  - [Web QA (zh)](https://huggingface.co/datasets/suolyer/webqa)
-  - [UltraChat (en)](https://github.com/thunlp/UltraChat)
-  - [WebNovel (zh)](https://huggingface.co/datasets/zxbsmk/webnovel_cn)
-- For reward modelling:
-  - [HH-RLHF (en)](https://huggingface.co/datasets/Anthropic/hh-rlhf)
-  - [Open Assistant (multilingual)](https://huggingface.co/datasets/OpenAssistant/oasst1)
-  - [GPT-4 Generated Data (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
+<details><summary>Pre-training datasets</summary>
 
-Please refer to [data/README.md](data/README.md) for details.
+- [Wiki Demo (en)](data/wiki_demo.txt)
+- [RefinedWeb (en)](https://huggingface.co/datasets/tiiuae/falcon-refinedweb)
+- [RedPajama V2 (en)](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-V2)
+- [Wikipedia (en)](https://huggingface.co/datasets/olm/olm-wikipedia-20221220)
+- [Wikipedia (zh)](https://huggingface.co/datasets/pleisto/wikipedia-cn-20230720-filtered)
+- [Pile (en)](https://huggingface.co/datasets/EleutherAI/pile)
+- [SkyPile (zh)](https://huggingface.co/datasets/Skywork/SkyPile-150B)
+- [The Stack (en)](https://huggingface.co/datasets/bigcode/the-stack)
+- [StarCoder (en)](https://huggingface.co/datasets/bigcode/starcoderdata)
+
+</details>
+
+<details><summary>Supervised fine-tuning datasets</summary>
+
+- [Stanford Alpaca (en)](https://github.com/tatsu-lab/stanford_alpaca)
+- [Stanford Alpaca (zh)](https://github.com/ymcui/Chinese-LLaMA-Alpaca)
+- [Alpaca GPT4 (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
+- [Identity (en&zh)](data/identity.json)
+- [Open Assistant (zh)](https://huggingface.co/datasets/OpenAssistant/oasst1)
+- [ShareGPT (zh)](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT/tree/main/Chinese-instruction-collection)
+- [Guanaco Dataset (multilingual)](https://huggingface.co/datasets/JosephusCheung/GuanacoDataset)
+- [BELLE 2M (zh)](https://huggingface.co/datasets/BelleGroup/train_2M_CN)
+- [BELLE 1M (zh)](https://huggingface.co/datasets/BelleGroup/train_1M_CN)
+- [BELLE 0.5M (zh)](https://huggingface.co/datasets/BelleGroup/train_0.5M_CN)
+- [BELLE Dialogue 0.4M (zh)](https://huggingface.co/datasets/BelleGroup/generated_chat_0.4M)
+- [BELLE School Math 0.25M (zh)](https://huggingface.co/datasets/BelleGroup/school_math_0.25M)
+- [BELLE Multiturn Chat 0.8M (zh)](https://huggingface.co/datasets/BelleGroup/multiturn_chat_0.8M)
+- [UltraChat (en)](https://github.com/thunlp/UltraChat)
+- [LIMA (en)](https://huggingface.co/datasets/GAIR/lima)
+- [OpenPlatypus (en)](https://huggingface.co/datasets/garage-bAInd/Open-Platypus)
+- [CodeAlpaca 20k (en)](https://huggingface.co/datasets/sahil2801/CodeAlpaca-20k)
+- [Alpaca CoT (multilingual)](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT)
+- [OpenOrca (en)](https://huggingface.co/datasets/Open-Orca/OpenOrca)
+- [SlimOrca (en)](https://huggingface.co/datasets/Open-Orca/SlimOrca)
+- [MathInstruct (en)](https://huggingface.co/datasets/TIGER-Lab/MathInstruct)
+- [Firefly 1.1M (zh)](https://huggingface.co/datasets/YeungNLP/firefly-train-1.1M)
+- [Wiki QA (en)](https://huggingface.co/datasets/wiki_qa)
+- [Web QA (zh)](https://huggingface.co/datasets/suolyer/webqa)
+- [WebNovel (zh)](https://huggingface.co/datasets/zxbsmk/webnovel_cn)
+- [Nectar (en)](https://huggingface.co/datasets/berkeley-nest/Nectar)
+- [deepctrl (en&zh)](https://www.modelscope.cn/datasets/deepctrl/deepctrl-sft-data)
+- [Ad Gen (zh)](https://huggingface.co/datasets/HasturOfficial/adgen)
+- [ShareGPT Hyperfiltered (en)](https://huggingface.co/datasets/totally-not-an-llm/sharegpt-hyperfiltered-3k)
+- [ShareGPT4 (en&zh)](https://huggingface.co/datasets/shibing624/sharegpt_gpt4)
+- [UltraChat 200k (en)](https://huggingface.co/datasets/HuggingFaceH4/ultrachat_200k)
+- [AgentInstruct (en)](https://huggingface.co/datasets/THUDM/AgentInstruct)
+- [LMSYS Chat 1M (en)](https://huggingface.co/datasets/lmsys/lmsys-chat-1m)
+- [Evol Instruct V2 (en)](https://huggingface.co/datasets/WizardLM/WizardLM_evol_instruct_V2_196k)
+- [Glaive Function Calling V2 (en)](https://huggingface.co/datasets/glaiveai/glaive-function-calling-v2)
+- [Cosmopedia (en)](https://huggingface.co/datasets/HuggingFaceTB/cosmopedia)
+- [LLaVA mixed (en&zh)](https://huggingface.co/datasets/BUAADreamer/llava-en-zh-300k)
+- [Open Assistant (de)](https://huggingface.co/datasets/mayflowergmbh/oasst_de)
+- [Dolly 15k (de)](https://huggingface.co/datasets/mayflowergmbh/dolly-15k_de)
+- [Alpaca GPT4 (de)](https://huggingface.co/datasets/mayflowergmbh/alpaca-gpt4_de)
+- [OpenSchnabeltier (de)](https://huggingface.co/datasets/mayflowergmbh/openschnabeltier_de)
+- [Evol Instruct (de)](https://huggingface.co/datasets/mayflowergmbh/evol-instruct_de)
+- [Dolphin (de)](https://huggingface.co/datasets/mayflowergmbh/dolphin_de)
+- [Booksum (de)](https://huggingface.co/datasets/mayflowergmbh/booksum_de)
+- [Airoboros (de)](https://huggingface.co/datasets/mayflowergmbh/airoboros-3.0_de)
+- [Ultrachat (de)](https://huggingface.co/datasets/mayflowergmbh/ultra-chat_de)
+
+</details>
+
+<details><summary>Preference datasets</summary>
+
+- [HH-RLHF (en)](https://huggingface.co/datasets/Anthropic/hh-rlhf)
+- [GPT-4 Generated Data (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
+- [Orca DPO (en)](https://huggingface.co/datasets/Intel/orca_dpo_pairs)
+- [Nectar (en)](https://huggingface.co/datasets/berkeley-nest/Nectar)
+- [DPO mixed (en&zh)](https://huggingface.co/datasets/hiyouga/DPO-En-Zh-20k)
+- [Open Assistant (zh)](https://huggingface.co/datasets/OpenAssistant/oasst1)
+- [Orca DPO (de)](https://huggingface.co/datasets/mayflowergmbh/intel_orca_dpo_pairs_de)
+
+</details>
 
 Some datasets require confirmation before using them, so we recommend logging in with your Hugging Face account using these commands.
 
@@ -109,293 +279,261 @@ huggingface-cli login
 
 ## Requirement
 
-- Python 3.8+ and PyTorch 1.13.1+
-- 🤗Transformers, Datasets, Accelerate, PEFT and TRL
-- jieba, rouge-chinese and nltk (used at evaluation)
-- gradio and matplotlib (used in web_demo.py)
-- uvicorn, fastapi and sse-starlette (used in api_demo.py)
+| Mandatory    | Minimum | Recommend |
+| ------------ | ------- | --------- |
+| python       | 3.8     | 3.10      |
+| torch        | 1.13.1  | 2.2.0     |
+| transformers | 4.37.2  | 4.40.1    |
+| datasets     | 2.14.3  | 2.19.1    |
+| accelerate   | 0.27.2  | 0.30.0    |
+| peft         | 0.9.0   | 0.10.0    |
+| trl          | 0.8.1   | 0.8.6     |
 
-And **powerful GPUs**!
+| Optional     | Minimum | Recommend |
+| ------------ | ------- | --------- |
+| CUDA         | 11.6    | 12.2      |
+| deepspeed    | 0.10.0  | 0.14.0    |
+| bitsandbytes | 0.39.0  | 0.43.1    |
+| vllm         | 0.4.0   | 0.4.2     |
+| flash-attn   | 2.3.0   | 2.5.8     |
+
+### Hardware Requirement
+
+\* *estimated*
+
+| Method            | Bits |   7B  |  13B  |  30B  |   70B  |  110B  |  8x7B |  8x22B |
+| ----------------- | ---- | ----- | ----- | ----- | ------ | ------ | ----- | ------ |
+| Full              | AMP  | 120GB | 240GB | 600GB | 1200GB | 2000GB | 900GB | 2400GB |
+| Full              |  16  |  60GB | 120GB | 300GB |  600GB |  900GB | 400GB | 1200GB |
+| Freeze            |  16  |  20GB |  40GB |  80GB |  200GB |  360GB | 160GB |  400GB |
+| LoRA/GaLore/BAdam |  16  |  16GB |  32GB |  64GB |  160GB |  240GB | 120GB |  320GB |
+| QLoRA             |   8  |  10GB |  20GB |  40GB |   80GB |  140GB |  60GB |  160GB |
+| QLoRA             |   4  |   6GB |  12GB |  24GB |   48GB |   72GB |  30GB |   96GB |
+| QLoRA             |   2  |   4GB |   8GB |  16GB |   24GB |   48GB |  18GB |   48GB |
 
 ## Getting Started
 
-### Data Preparation (optional)
+### Installation
 
-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)
+> [!IMPORTANT]
+> Installation is mandatory.
 
 ```bash
-git lfs install
-git clone https://github.com/hiyouga/LLaMA-Efficient-Tuning.git
-conda create -n llama_etuning python=3.10
-conda activate llama_etuning
-cd LLaMA-Efficient-Tuning
-pip install -r requirements.txt
+git clone https://github.com/hiyouga/LLaMA-Factory.git
+cd LLaMA-Factory
+pip install -e .[torch,metrics]
 ```
 
-If you want to enable the quantized LoRA (QLoRA) on the Windows platform, you will be required to install a pre-built version of `bitsandbytes` library, which supports CUDA 11.1 to 12.1.
+Extra dependencies available: torch, metrics, deepspeed, bitsandbytes, vllm, galore, badam, gptq, awq, aqlm, qwen, modelscope, quality
+
+> [!TIP]
+> Use `pip install --no-deps -e .` to resolve package conflicts.
+
+<details><summary>For Windows users</summary>
+
+If you want to enable the quantized LoRA (QLoRA) on the Windows platform, you need to install a pre-built version of `bitsandbytes` library, which supports CUDA 11.1 to 12.2, please select the appropriate [release version](https://github.com/jllllll/bitsandbytes-windows-webui/releases/tag/wheels) based on your CUDA version.
 
 ```bash
-pip install https://github.com/jllllll/bitsandbytes-windows-webui/releases/download/wheels/bitsandbytes-0.39.1-py3-none-win_amd64.whl
+pip install https://github.com/jllllll/bitsandbytes-windows-webui/releases/download/wheels/bitsandbytes-0.41.2.post2-py3-none-win_amd64.whl
 ```
 
-### All-in-one Web UI
+To enable FlashAttention-2 on the Windows platform, you need to install the precompiled `flash-attn` library, which supports CUDA 12.1 to 12.2. Please download the corresponding version from [flash-attention](https://github.com/bdashore3/flash-attention/releases) based on your requirements.
+
+</details>
+
+<details><summary>For Ascend NPU users</summary>
+
+To utilize Ascend NPU devices for (distributed) training and inference, you need to install the **[torch-npu](https://gitee.com/ascend/pytorch)** library and the **[Ascend CANN Kernels](https://www.hiascend.com/developer/download/community/result?module=cann)**.
+
+| Requirement  | Minimum | Recommend |
+| ------------ | ------- | --------- |
+| CANN         | 8.0.RC1 | 8.0.RC1   |
+| torch        | 2.2.0   | 2.2.0     |
+| torch-npu    | 2.2.0   | 2.2.0     |
+| deepspeed    | 0.13.2  | 0.13.2    |
+
+Docker image:
+
+- 32GB: [Download page](http://mirrors.cn-central-221.ovaijisuan.com/detail/130.html)
+- 64GB: Coming soon
+
+Remember to use `ASCEND_RT_VISIBLE_DEVICES` instead of `CUDA_VISIBLE_DEVICES` to specify the device to use.
+
+If you cannot infer model on NPU devices, try setting `do_sample: false` in the configurations.
+
+</details>
+
+### Data Preparation
+
+Please refer to [data/README.md](data/README.md) for checking the details about the format of dataset files. You can either use datasets on HuggingFace / ModelScope hub or load the dataset in local disk.
+
+> [!NOTE]
+> Please update `data/dataset_info.json` to use your custom dataset.
+
+### Quickstart
+
+Use the following 3 commands to run LoRA **fine-tuning**, **inference** and **merging** of the Llama3-8B-Instruct model, respectively.
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_web.py
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_sft.yaml
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli chat examples/inference/llama3_lora_sft.yaml
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli export examples/merge_lora/llama3_lora_sft.yaml
 ```
 
-Currently the web UI only supports training on **a single GPU**.
+See [examples/README.md](examples/README.md) for advanced usage (including distributed training).
 
-### (Continually) Pre-Training
+> [!TIP]
+> Use `llamafactory-cli help` to show help information.
+
+### Fine-Tuning with LLaMA Board GUI (powered by [Gradio](https://github.com/gradio-app/gradio))
+
+> [!IMPORTANT]
+> LLaMA Board GUI only supports training on a single GPU.
+
+#### Use local environment
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage pt \
-    --model_name_or_path path_to_your_model \
-    --do_train \
-    --dataset wiki_demo \
-    --template default \
-    --finetuning_type lora \
-    --output_dir path_to_pt_checkpoint \
-    --overwrite_cache \
-    --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 \
-    --plot_loss \
-    --fp16
+CUDA_VISIBLE_DEVICES=0 GRADIO_SHARE=1 llamafactory-cli webui
 ```
 
-### Supervised Fine-Tuning
+<details><summary>For Alibaba Cloud PAI or AutoDL users</summary>
+
+If you encountered display problems in LLaMA Board on Alibaba Cloud PAI, try using the following command to set environment variables before starting LLaMA Board:
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage sft \
-    --model_name_or_path path_to_your_model \
-    --do_train \
-    --dataset alpaca_gpt4_en \
-    --template default \
-    --finetuning_type lora \
-    --output_dir path_to_sft_checkpoint \
-    --overwrite_cache \
-    --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 \
-    --plot_loss \
-    --fp16
+export GRADIO_SERVER_PORT=7860 GRADIO_ROOT_PATH=/${JUPYTER_NAME}/proxy/7860/
 ```
 
-Remember to specify `--lora_target W_pack` if you are using Baichuan models.
-
-### Reward Model Training
+If you are using AutoDL, please install a specific version of Gradio:
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage rm \
-    --model_name_or_path path_to_your_model \
-    --do_train \
-    --dataset comparison_gpt4_en \
-    --template default \
-    --finetuning_type lora \
-    --resume_lora_training False \
-    --checkpoint_dir path_to_sft_checkpoint \
-    --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 \
-    --plot_loss \
-    --fp16
-```
-
-### PPO Training (RLHF)
-
-```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage ppo \
-    --model_name_or_path path_to_your_model \
-    --do_train \
-    --dataset alpaca_gpt4_en \
-    --template default \
-    --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 \
-    --plot_loss
-```
-
-### Distributed Training
-
-```bash
-accelerate config # configure the environment
-accelerate launch src/train_bash.py # arguments (same as above)
-```
-
-<details><summary>Example configuration for full-tuning with DeepSpeed ZeRO-2</summary>
-
-```yaml
-compute_environment: LOCAL_MACHINE
-deepspeed_config:
-  gradient_accumulation_steps: 4
-  gradient_clipping: 0.5
-  offload_optimizer_device: none
-  offload_param_device: none
-  zero3_init_flag: false
-  zero_stage: 2
-distributed_type: DEEPSPEED
-downcast_bf16: 'no'
-machine_rank: 0
-main_training_function: main
-mixed_precision: fp16
-num_machines: 1
-num_processes: 4
-rdzv_backend: static
-same_network: true
-tpu_env: []
-tpu_use_cluster: false
-tpu_use_sudo: false
-use_cpu: false
+pip install gradio==4.10.0
 ```
 
 </details>
 
-### Evaluation (BLEU and ROUGE_CHINESE)
+#### Use Docker
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage sft \
-    --model_name_or_path path_to_your_model \
-    --do_eval \
-    --dataset alpaca_gpt4_en \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint \
-    --output_dir path_to_eval_result \
-    --per_device_eval_batch_size 8 \
-    --max_samples 100 \
-    --predict_with_generate
+docker build -f ./Dockerfile -t llama-factory:latest .
+docker run --gpus=all \
+    -v ./hf_cache:/root/.cache/huggingface/ \
+    -v ./data:/app/data \
+    -v ./output:/app/output \
+    -e CUDA_VISIBLE_DEVICES=0 \
+    -p 7860:7860 \
+    --shm-size 16G \
+    --name llama_factory \
+    -d llama-factory:latest
 ```
 
-We recommend using `--per_device_eval_batch_size=1` and `--max_target_length 128` at 4/8-bit evaluation.
-
-### Predict
+#### Use Docker Compose
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage sft \
-    --model_name_or_path path_to_your_model \
-    --do_predict \
-    --dataset alpaca_gpt4_en \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint \
-    --output_dir path_to_predict_result \
-    --per_device_eval_batch_size 8 \
-    --max_samples 100 \
-    --predict_with_generate
+docker compose -f ./docker-compose.yml up -d
 ```
 
-### API Demo
+<details><summary>Details about volume</summary>
+
+- hf_cache: Utilize Hugging Face cache on the host machine. Reassignable if a cache already exists in a different directory.
+- data: Place datasets on this dir of the host machine so that they can be selected on LLaMA Board GUI.
+- output: Set export dir to this location so that the merged result can be accessed directly on the host machine.
+
+</details>
+
+### Deploy with OpenAI-style API and vLLM
 
 ```bash
-python src/api_demo.py \
-    --model_name_or_path path_to_your_model \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint
+CUDA_VISIBLE_DEVICES=0,1 API_PORT=8000 llamafactory-cli api examples/inference/llama3_vllm.yaml
 ```
 
-Visit `http://localhost:8000/docs` for API documentation.
+### Download from ModelScope Hub
 
-### CLI Demo
+If you have trouble with downloading models and datasets from Hugging Face, you can use ModelScope.
 
 ```bash
-python src/cli_demo.py \
-    --model_name_or_path path_to_your_model \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint
+export USE_MODELSCOPE_HUB=1 # `set USE_MODELSCOPE_HUB=1` for Windows
 ```
 
-### Web Demo
+Train the model by specifying a model ID of the ModelScope Hub as the `--model_name_or_path`. You can find a full list of model IDs at [ModelScope Hub](https://modelscope.cn/models), e.g., `LLM-Research/Meta-Llama-3-8B-Instruct`.
 
-```bash
-python src/web_demo.py \
-    --model_name_or_path path_to_your_model \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint
-```
+## Projects using LLaMA Factory
 
-### Export model
+If you have a project that should be incorporated, please contact via email or create a pull request.
 
-```bash
-python src/export_model.py \
-    --model_name_or_path path_to_your_model \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint \
-    --output_dir path_to_export
-```
+<details><summary>Click to show</summary>
 
-## TODO
+1. Wang et al. ESRL: Efficient Sampling-based Reinforcement Learning for Sequence Generation. 2023. [[arxiv]](https://arxiv.org/abs/2308.02223)
+1. Yu et al. Open, Closed, or Small Language Models for Text Classification? 2023. [[arxiv]](https://arxiv.org/abs/2308.10092)
+1. Wang et al. UbiPhysio: Support Daily Functioning, Fitness, and Rehabilitation with Action Understanding and Feedback in Natural Language. 2023. [[arxiv]](https://arxiv.org/abs/2308.10526)
+1. Luceri et al. Leveraging Large Language Models to Detect Influence Campaigns in Social Media. 2023. [[arxiv]](https://arxiv.org/abs/2311.07816)
+1. Zhang et al. Alleviating Hallucinations of Large Language Models through Induced Hallucinations. 2023. [[arxiv]](https://arxiv.org/abs/2312.15710)
+1. Wang et al. Know Your Needs Better: Towards Structured Understanding of Marketer Demands with Analogical Reasoning Augmented LLMs. 2024. [[arxiv]](https://arxiv.org/abs/2401.04319)
+1. Wang et al. CANDLE: Iterative Conceptualization and Instantiation Distillation from Large Language Models for Commonsense Reasoning. 2024. [[arxiv]](https://arxiv.org/abs/2401.07286)
+1. Choi et al. FACT-GPT: Fact-Checking Augmentation via Claim Matching with LLMs. 2024. [[arxiv]](https://arxiv.org/abs/2402.05904)
+1. Zhang et al. AutoMathText: Autonomous Data Selection with Language Models for Mathematical Texts. 2024. [[arxiv]](https://arxiv.org/abs/2402.07625)
+1. Lyu et al. KnowTuning: Knowledge-aware Fine-tuning for Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2402.11176)
+1. Yang et al. LaCo: Large Language Model Pruning via Layer Collaps. 2024. [[arxiv]](https://arxiv.org/abs/2402.11187)
+1. Bhardwaj et al. Language Models are Homer Simpson! Safety Re-Alignment of Fine-tuned Language Models through Task Arithmetic. 2024. [[arxiv]](https://arxiv.org/abs/2402.11746)
+1. Yang et al. Enhancing Empathetic Response Generation by Augmenting LLMs with Small-scale Empathetic Models. 2024. [[arxiv]](https://arxiv.org/abs/2402.11801)
+1. Yi et al. Generation Meets Verification: Accelerating Large Language Model Inference with Smart Parallel Auto-Correct Decoding. 2024. [[arxiv]](https://arxiv.org/abs/2402.11809)
+1. Cao et al. Head-wise Shareable Attention for Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2402.11819)
+1. Zhang et al. Enhancing Multilingual Capabilities of Large Language Models through Self-Distillation from Resource-Rich Languages. 2024. [[arxiv]](https://arxiv.org/abs/2402.12204)
+1. Kim et al. Efficient and Effective Vocabulary Expansion Towards Multilingual Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2402.14714)
+1. Yu et al. KIEval: A Knowledge-grounded Interactive Evaluation Framework for Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2402.15043)
+1. Huang et al. Key-Point-Driven Data Synthesis with its Enhancement on Mathematical Reasoning. 2024. [[arxiv]](https://arxiv.org/abs/2403.02333)
+1. Duan et al. Negating Negatives: Alignment without Human Positive Samples via Distributional Dispreference Optimization. 2024. [[arxiv]](https://arxiv.org/abs/2403.03419)
+1. Xie and Schwertfeger. Empowering Robotics with Large Language Models: osmAG Map Comprehension with LLMs. 2024. [[arxiv]](https://arxiv.org/abs/2403.08228)
+1. Wu et al. Large Language Models are Parallel Multilingual Learners. 2024. [[arxiv]](https://arxiv.org/abs/2403.09073)
+1. Zhang et al. EDT: Improving Large Language Models' Generation by Entropy-based Dynamic Temperature Sampling. 2024. [[arxiv]](https://arxiv.org/abs/2403.14541)
+1. Weller et al. FollowIR: Evaluating and Teaching Information Retrieval Models to Follow Instructions. 2024. [[arxiv]](https://arxiv.org/abs/2403.15246)
+1. Hongbin Na. CBT-LLM: A Chinese Large Language Model for Cognitive Behavioral Therapy-based Mental Health Question Answering. 2024. [[arxiv]](https://arxiv.org/abs/2403.16008)
+1. Zan et al. CodeS: Natural Language to Code Repository via Multi-Layer Sketch. 2024. [[arxiv]](https://arxiv.org/abs/2403.16443)
+1. Liu et al. Extensive Self-Contrast Enables Feedback-Free Language Model Alignment. 2024. [[arxiv]](https://arxiv.org/abs/2404.00604)
+1. Luo et al. BAdam: A Memory Efficient Full Parameter Training Method for Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2404.02827)
+1. Du et al. Chinese Tiny LLM: Pretraining a Chinese-Centric Large Language Model. 2024. [[arxiv]](https://arxiv.org/abs/2404.04167)
+1. Ma et al. Parameter Efficient Quasi-Orthogonal Fine-Tuning via Givens Rotation. 2024. [[arxiv]](https://arxiv.org/abs/2404.04316)
+1. Liu et al. Dynamic Generation of Personalities with Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2404.07084)
+1. Shang et al. How Far Have We Gone in Stripped Binary Code Understanding Using Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2404.09836)
+1. Huang et al. LLMTune: Accelerate Database Knob Tuning with Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2404.11581)
+1. Deng et al. Text-Tuple-Table: Towards Information Integration in Text-to-Table Generation via Global Tuple Extraction. 2024. [[arxiv]](https://arxiv.org/abs/2404.14215)
+1. Acikgoz et al. Hippocrates: An Open-Source Framework for Advancing Large Language Models in Healthcare. 2024. [[arxiv]](https://arxiv.org/abs/2404.16621)
+1. Zhang et al. Small Language Models Need Strong Verifiers to Self-Correct Reasoning. 2024. [[arxiv]](https://arxiv.org/abs/2404.17140)
+1. Zhou et al. FREB-TQA: A Fine-Grained Robustness Evaluation Benchmark for Table Question Answering. 2024. [[arxiv]](https://arxiv.org/abs/2404.18585)
+1. **[StarWhisper](https://github.com/Yu-Yang-Li/StarWhisper)**: A large language model for Astronomy, based on ChatGLM2-6B and Qwen-14B.
+1. **[DISC-LawLLM](https://github.com/FudanDISC/DISC-LawLLM)**: A large language model specialized in Chinese legal domain, based on Baichuan-13B, is capable of retrieving and reasoning on legal knowledge.
+1. **[Sunsimiao](https://github.com/thomas-yanxin/Sunsimiao)**: A large language model specialized in Chinese medical domain, based on Baichuan-7B and ChatGLM-6B.
+1. **[CareGPT](https://github.com/WangRongsheng/CareGPT)**: A series of large language models for Chinese medical domain, based on LLaMA2-7B and Baichuan-13B.
+1. **[MachineMindset](https://github.com/PKU-YuanGroup/Machine-Mindset/)**: A series of MBTI Personality large language models, capable of giving any LLM 16 different personality types based on different datasets and training methods.
+1. **[Luminia-13B-v3](https://huggingface.co/Nekochu/Luminia-13B-v3)**: A large language model specialized in generate metadata for stable diffusion. [[🤗Demo]](https://huggingface.co/spaces/Nekochu/Luminia-13B_SD_Prompt)
+1. **[Chinese-LLaVA-Med](https://github.com/BUAADreamer/Chinese-LLaVA-Med)**: A multimodal large language model specialized in Chinese medical domain, based on LLaVA-1.5-7B.
 
-- [ ] Supporting flash attention ([torch](https://pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html) / [xformers](https://github.com/facebookresearch/xformers) / [flashattn](https://github.com/Dao-AILab/flash-attention)).
-- [ ] Implementing multi-query attention for faster inference.
-- [ ] Supporting full-parameter RLHF training.
+</details>
 
 ## License
 
 This repository is licensed under the [Apache-2.0 License](LICENSE).
 
-Please follow the model licenses to use the corresponding model weights:
-
-- [LLaMA](https://github.com/facebookresearch/llama/blob/main/MODEL_CARD.md)
-- [LLaMA-2](https://ai.meta.com/llama/license/)
-- [BLOOM](https://huggingface.co/spaces/bigscience/license)
-- [Falcon](LICENSE)
-- [Baichuan](https://huggingface.co/baichuan-inc/baichuan-7B/resolve/main/baichuan-7B%20%E6%A8%A1%E5%9E%8B%E8%AE%B8%E5%8F%AF%E5%8D%8F%E8%AE%AE.pdf)
-- [InternLM](https://github.com/InternLM/InternLM#open-source-license)
+Please follow the model licenses to use the corresponding model weights: [Baichuan2](https://huggingface.co/baichuan-inc/Baichuan2-7B-Base/blob/main/Community%20License%20for%20Baichuan%202%20Model.pdf) / [BLOOM](https://huggingface.co/spaces/bigscience/license) / [ChatGLM3](https://github.com/THUDM/ChatGLM3/blob/main/MODEL_LICENSE) / [Command-R](https://cohere.com/c4ai-cc-by-nc-license) / [DeepSeek](https://github.com/deepseek-ai/DeepSeek-LLM/blob/main/LICENSE-MODEL) / [Falcon](https://huggingface.co/tiiuae/falcon-180B/blob/main/LICENSE.txt) / [Gemma](https://ai.google.dev/gemma/terms) / [InternLM2](https://github.com/InternLM/InternLM#license) / [LLaMA](https://github.com/facebookresearch/llama/blob/main/MODEL_CARD.md) / [LLaMA-2 (LLaVA-1.5)](https://ai.meta.com/llama/license/) / [LLaMA-3](https://llama.meta.com/llama3/license/) / [Mistral](LICENSE) / [OLMo](LICENSE) / [Phi-1.5/2](https://huggingface.co/microsoft/phi-1_5/resolve/main/Research%20License.docx) / [Phi-3](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct/blob/main/LICENSE) / [Qwen](https://github.com/QwenLM/Qwen/blob/main/Tongyi%20Qianwen%20LICENSE%20AGREEMENT) / [StarCoder2](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) / [XVERSE](https://github.com/xverse-ai/XVERSE-13B/blob/main/MODEL_LICENSE.pdf) / [Yi](https://huggingface.co/01-ai/Yi-6B/blob/main/LICENSE) / [Yi-1.5](LICENSE) / [Yuan](https://github.com/IEIT-Yuan/Yuan-2.0/blob/main/LICENSE-Yuan)
 
 ## Citation
 
 If this work is helpful, please kindly cite as:
 
 ```bibtex
-@Misc{llama-efficient-tuning,
-  title = {LLaMA Efficient Tuning},
-  author = {hiyouga},
-  howpublished = {\url{https://github.com/hiyouga/LLaMA-Efficient-Tuning}},
-  year = {2023}
+@article{zheng2024llamafactory,
+  title={LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models},
+  author={Yaowei Zheng and Richong Zhang and Junhao Zhang and Yanhan Ye and Zheyan Luo and Yongqiang Ma},
+  journal={arXiv preprint arXiv:2403.13372},
+  year={2024},
+  url={http://arxiv.org/abs/2403.13372}
 }
 ```
 
 ## Acknowledgement
 
-This repo is a sibling of [ChatGLM-Efficient-Tuning](https://github.com/hiyouga/ChatGLM-Efficient-Tuning). They share a similar code structure of efficient tuning on large language models.
+This repo benefits from [PEFT](https://github.com/huggingface/peft), [TRL](https://github.com/huggingface/trl), [QLoRA](https://github.com/artidoro/qlora) and [FastChat](https://github.com/lm-sys/FastChat). Thanks for their wonderful works.
 
 ## Star History
 
-![Star History Chart](https://api.star-history.com/svg?repos=hiyouga/LLaMA-Efficient-Tuning&type=Date)
+![Star History Chart](https://api.star-history.com/svg?repos=hiyouga/LLaMA-Factory&type=Date)

README_zh.md

@@ -1,104 +1,274 @@
-# LLaMA Efficient Tuning
+![# LLaMA Factory](assets/logo.png)
 
-[![GitHub Repo stars](https://img.shields.io/github/stars/hiyouga/LLaMA-Efficient-Tuning?style=social)](https://github.com/hiyouga/LLaMA-Efficient-Tuning/stargazers)
-[![GitHub Code License](https://img.shields.io/github/license/hiyouga/LLaMA-Efficient-Tuning)](LICENSE)
-[![GitHub last commit](https://img.shields.io/github/last-commit/hiyouga/LLaMA-Efficient-Tuning)](https://github.com/hiyouga/LLaMA-Efficient-Tuning/commits/main)
+[![GitHub Repo stars](https://img.shields.io/github/stars/hiyouga/LLaMA-Factory?style=social)](https://github.com/hiyouga/LLaMA-Factory/stargazers)
+[![GitHub Code License](https://img.shields.io/github/license/hiyouga/LLaMA-Factory)](LICENSE)
+[![GitHub last commit](https://img.shields.io/github/last-commit/hiyouga/LLaMA-Factory)](https://github.com/hiyouga/LLaMA-Factory/commits/main)
 [![PyPI](https://img.shields.io/pypi/v/llmtuner)](https://pypi.org/project/llmtuner/)
-[![GitHub pull request](https://img.shields.io/badge/PRs-welcome-blue)](https://github.com/hiyouga/LLaMA-Efficient-Tuning/pulls)
+[![Downloads](https://static.pepy.tech/badge/llmtuner)](https://pypi.org/project/llmtuner/)
+[![Citation](https://img.shields.io/badge/citation-44-green)](#使用了-llama-factory-的项目)
+[![GitHub pull request](https://img.shields.io/badge/PRs-welcome-blue)](https://github.com/hiyouga/LLaMA-Factory/pulls)
+[![Discord](https://dcbadge.vercel.app/api/server/rKfvV9r9FK?compact=true&style=flat)](https://discord.gg/rKfvV9r9FK)
+[![Twitter](https://img.shields.io/twitter/follow/llamafactory_ai)](https://twitter.com/llamafactory_ai)
+[![Spaces](https://img.shields.io/badge/🤗-Open%20in%20Spaces-blue)](https://huggingface.co/spaces/hiyouga/LLaMA-Board)
+[![Studios](https://img.shields.io/badge/ModelScope-Open%20in%20Studios-blue)](https://modelscope.cn/studios/hiyouga/LLaMA-Board)
+[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1d5KQtbemerlSDSxZIfAaWXhKr30QypiK?usp=sharing)
+
+[![GitHub Tread](https://trendshift.io/api/badge/repositories/4535)](https://trendshift.io/repositories/4535)
 
 👋 加入我们的[微信群](assets/wechat.jpg)。
 
 \[ [English](README.md) | 中文 \]
 
+**微调大模型可以像这样轻松…**
+
+https://github.com/hiyouga/LLaMA-Factory/assets/16256802/ec36a9dd-37f4-4f72-81bd-d76c6d0a6594
+
+选择你的打开方式：
+
+- **Colab**：https://colab.research.google.com/drive/1d5KQtbemerlSDSxZIfAaWXhKr30QypiK?usp=sharing
+- **本地机器**：请见[如何使用](#如何使用)
+
+## 目录
+
+- [项目特色](#项目特色)
+- [性能指标](#性能指标)
+- [更新日志](#更新日志)
+- [模型](#模型)
+- [训练方法](#训练方法)
+- [数据集](#数据集)
+- [软硬件依赖](#软硬件依赖)
+- [如何使用](#如何使用)
+- [使用了 LLaMA Factory 的项目](#使用了-llama-factory-的项目)
+- [协议](#协议)
+- [引用](#引用)
+- [致谢](#致谢)
+
+## 项目特色
+
+- **多种模型**：LLaMA、LLaVA、Mistral、Mixtral-MoE、Qwen、Yi、Gemma、Baichuan、ChatGLM、Phi 等等。
+- **集成方法**：（增量）预训练、（多模态）指令监督微调、奖励模型训练、PPO 训练、DPO 训练和 ORPO 训练。
+- **多种精度**：32 比特全参数微调、16 比特冻结微调、16 比特 LoRA 微调和基于 AQLM/AWQ/GPTQ/LLM.int8 的 2/4/8 比特 QLoRA 微调。
+- **先进算法**：GaLore、BAdam、DoRA、LongLoRA、LLaMA Pro、Mixture-of-Depths、LoRA+、LoftQ 和 Agent 微调。
+- **实用技巧**：FlashAttention-2、Unsloth、RoPE scaling、NEFTune 和 rsLoRA。
+- **实验监控**：LlamaBoard、TensorBoard、Wandb、MLflow 等等。
+- **极速推理**：基于 vLLM 的 OpenAI 风格 API、浏览器界面和命令行接口。
+
+## 性能指标
+
+与 ChatGLM 官方的 [P-Tuning](https://github.com/THUDM/ChatGLM2-6B/tree/main/ptuning) 微调相比，LLaMA Factory 的 LoRA 微调提供了 **3.7 倍**的加速比，同时在广告文案生成任务上取得了更高的 Rouge 分数。结合 4 比特量化技术，LLaMA Factory 的 QLoRA 微调进一步降低了 GPU 显存消耗。
+
+![benchmark](assets/benchmark.svg)
+
+<details><summary>变量定义</summary>
+
+- **Training Speed**: 训练阶段每秒处理的样本数量。（批处理大小=4，截断长度=1024）
+- **Rouge Score**: [广告文案生成](https://aclanthology.org/D19-1321.pdf)任务验证集上的 Rouge-2 分数。（批处理大小=4，截断长度=1024）
+- **GPU Memory**: 4 比特量化训练的 GPU 显存峰值。（批处理大小=1，截断长度=1024）
+- 我们在 ChatGLM 的 P-Tuning 中采用 `pre_seq_len=128`，在 LLaMA Factory 的 LoRA 微调中采用 `lora_rank=32`。
+
+</details>
+
 ## 更新日志
 
-[23/07/31] 现在我们支持了训练数据流式加载。请尝试使用 `--streaming` 和 `--max_steps 100` 参数来流式加载数据集。
+[24/05/14] 我们支持了昇腾 NPU 设备的训练和推理。详情请查阅[安装](#安装-llama-factory)部分。
 
-[23/07/29] 我们在 Hugging Face 发布了两个 13B 指令微调模型。详细内容请查阅我们的 Hugging Face 项目（[LLaMA-2](https://huggingface.co/hiyouga/Llama-2-Chinese-13b-chat) / [Baichuan](https://huggingface.co/hiyouga/baichuan-13b-sft)）。
+[24/05/13] 我们支持了 Yi-1.5 系列模型的微调。
 
-[23/07/19] 现在我们支持了 **LLaMA-2** 模型的训练。请尝试使用 `--model_name_or_path meta-llama/Llama-2-7b-hf` 参数。请注意使用 LLaMA-2-chat 模型需要添加 `--template llama2` 参数。
+[24/04/26] 我们支持了多模态模型 **LLaVA-1.5** 的微调。详细用法请参照 [examples](examples/README_zh.md)。
 
-[23/07/18] 我们开发了支持训练和测试的浏览器一键微调界面。请尝试使用 `train_web.py` 在您的浏览器中微调模型。感谢 [@KanadeSiina](https://github.com/KanadeSiina) 和 [@codemayq](https://github.com/codemayq) 在该功能开发中付出的努力。
+<details><summary>展开日志</summary>
 
-[23/07/11] 现在我们支持了 **Baichuan-13B** 模型的训练。请尝试使用 `--model_name_or_path path_to_baichuan_model` 和 `--lora_target W_pack` 参数。请注意使用 Baichuan-13B-Chat 模型需要添加 `--template baichuan` 参数。
+[24/04/22] 我们提供了在免费 T4 GPU 上微调 Llama-3 模型的 **[Colab 笔记本](https://colab.research.google.com/drive/1d5KQtbemerlSDSxZIfAaWXhKr30QypiK?usp=sharing)**。Hugging Face 社区公开了两个利用 LLaMA Factory 微调的 Llama-3 模型，详情请见 [Llama3-8B-Chinese-Chat](https://huggingface.co/shenzhi-wang/Llama3-8B-Chinese-Chat) 和 [Llama3-Chinese](https://huggingface.co/zhichen/Llama3-Chinese)。
 
-[23/07/09] 我们开源了 [FastEdit](https://github.com/hiyouga/FastEdit)⚡🩹，一个简单易用的、能迅速编辑大模型事实记忆的工具包。如果您感兴趣请关注我们的 [FastEdit](https://github.com/hiyouga/FastEdit) 项目。
+[24/04/21] 我们基于 [AstraMindAI 的仓库](https://github.com/astramind-ai/Mixture-of-depths)支持了 **[混合深度训练](https://arxiv.org/abs/2404.02258)**。详细用法请参照 [examples](examples/README_zh.md)。
 
-[23/07/07] 现在我们支持了 **InternLM-7B** 模型的训练。请尝试使用 `--model_name_or_path internlm/internlm-7b` 参数。请注意使用 InternLM-chat 模型需要添加 `--template intern` 参数。
+[24/04/16] 我们支持了 **[BAdam](https://arxiv.org/abs/2404.02827)**。详细用法请参照 [examples](examples/README_zh.md)。
 
-[23/07/05] 现在我们支持了 **Falcon-7B/40B** 模型的训练。请尝试使用 `--model_name_or_path tiiuae/falcon-7b` 和 `--lora_target query_key_value` 参数。
+[24/04/16] 我们支持了 **[unsloth](https://github.com/unslothai/unsloth)** 的长序列训练（24GB 可训练 Llama-2-7B-56k）。该方法相比 FlashAttention-2 提供了 **117%** 的训练速度和 **50%** 的显存节约。更多数据请见[此页面](https://github.com/hiyouga/LLaMA-Factory/wiki/Performance-comparison)。
 
-[23/06/29] 我们提供了一个**可复现的**指令模型微调示例，详细内容请查阅 [Hugging Face 项目](https://huggingface.co/hiyouga/baichuan-7b-sft)。
+[24/03/31] 我们支持了 **[ORPO](https://arxiv.org/abs/2403.07691)**。详细用法请参照 [examples](examples/README_zh.md)。
 
-[23/06/22] 我们对齐了[示例 API](src/api_demo.py) 与 [OpenAI API](https://platform.openai.com/docs/api-reference/chat) 的格式，您可以将微调模型接入任意基于 ChatGPT 的应用中。
+[24/03/21] 我们的论文 "[LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models](https://arxiv.org/abs/2403.13372)" 可在 arXiv 上查看！
 
-[23/06/15] 现在我们支持了 **Baichuan-7B** 模型的训练。请尝试使用 `--model_name_or_path baichuan-inc/Baichuan-7B` 和 `--lora_target W_pack` 参数。
+[24/03/20] 我们支持了能在 2x24GB GPU 上微调 70B 模型的 **FSDP+QLoRA**。详细用法请参照 [examples](examples/README_zh.md)。
 
-[23/06/03] 现在我们实现了 4 比特的 LoRA 训练（也称 [QLoRA](https://github.com/artidoro/qlora)）。请尝试使用 `--quantization_bit 4` 参数进行 4 比特量化微调。
+[24/03/13] 我们支持了 **[LoRA+](https://arxiv.org/abs/2402.12354)**。详细用法请参照 [examples](examples/README_zh.md)。
 
-[23/05/31] 现在我们支持了 **BLOOM & BLOOMZ** 模型的训练。请尝试使用 `--model_name_or_path bigscience/bloomz-7b1-mt` 和 `--lora_target query_key_value` 参数。
+[24/03/07] 我们支持了梯度低秩投影（**[GaLore](https://arxiv.org/abs/2403.03507)**）算法。详细用法请参照 [examples](examples/README_zh.md)。
+
+[24/03/07] 我们集成了 **[vLLM](https://github.com/vllm-project/vllm)** 以实现极速并发推理。请使用 `infer_backend: vllm` 来获得 **270%** 的推理速度。
+
+[24/02/28] 我们支持了 **[DoRA](https://arxiv.org/abs/2402.09353)** 微调。请使用 `use_dora: true` 参数进行 DoRA 微调。
+
+[24/02/15] 我们支持了 [LLaMA Pro](https://github.com/TencentARC/LLaMA-Pro) 提出的**块扩展**方法。详细用法请参照 [examples](examples/README_zh.md)。
+
+[24/02/05] Qwen1.5（Qwen2 测试版）系列模型已在 LLaMA-Factory 中实现微调支持。详情请查阅该[博客页面](https://qwenlm.github.io/zh/blog/qwen1.5/)。
+
+[24/01/18] 我们针对绝大多数模型实现了 **Agent 微调**，微调时指定 `dataset: glaive_toolcall` 即可使模型获得工具调用能力。
+
+[23/12/23] 我们针对 LLaMA, Mistral 和 Yi 模型支持了 **[unsloth](https://github.com/unslothai/unsloth)** 的 LoRA 训练加速。请使用 `use_unsloth: true` 参数启用 unsloth 优化。该方法可提供 **170%** 的训练速度，详情请查阅[此页面](https://github.com/hiyouga/LLaMA-Factory/wiki/Performance-comparison)。
+
+[23/12/12] 我们支持了微调最新的混合专家模型 **[Mixtral 8x7B](https://huggingface.co/mistralai/Mixtral-8x7B-v0.1)**。硬件需求请查阅[此处](#硬件依赖)。
+
+[23/12/01] 我们支持了从 **[魔搭社区](https://modelscope.cn/models)** 下载预训练模型和数据集。详细用法请参照 [此教程](#从魔搭社区下载)。
+
+[23/10/21] 我们支持了 **[NEFTune](https://arxiv.org/abs/2310.05914)** 训练技巧。请使用 `neftune_noise_alpha: 5` 参数启用 NEFTune。
+
+[23/09/27] 我们针对 LLaMA 模型支持了 [LongLoRA](https://github.com/dvlab-research/LongLoRA) 提出的 **$S^2$-Attn**。请使用 `shift_attn: true` 参数以启用该功能。
+
+[23/09/23] 我们在项目中集成了 MMLU、C-Eval 和 CMMLU 评估集。详细用法请参照 [examples](examples/README_zh.md)。
+
+[23/09/10] 我们支持了 **[FlashAttention-2](https://github.com/Dao-AILab/flash-attention)**。如果您使用的是 RTX4090、A100 或 H100 GPU，请使用 `flash_attn: fa2` 参数以启用 FlashAttention-2。
+
+[23/08/12] 我们支持了 **RoPE 插值**来扩展 LLaMA 模型的上下文长度。请使用 `rope_scaling: linear` 参数训练模型或使用 `rope_scaling: dynamic` 参数评估模型。
+
+[23/08/11] 我们支持了指令模型的 **[DPO 训练](https://arxiv.org/abs/2305.18290)**。详细用法请参照 [examples](examples/README_zh.md)。
+
+[23/07/31] 我们支持了**数据流式加载**。请使用 `streaming: true` 和 `max_steps: 10000` 参数来流式加载数据集。
+
+[23/07/29] 我们在 Hugging Face 发布了两个 13B 指令微调模型。详细内容请查阅我们的 Hugging Face 项目（[LLaMA-2](https://huggingface.co/hiyouga/Llama-2-Chinese-13b-chat) / [Baichuan](https://huggingface.co/hiyouga/Baichuan-13B-sft)）。
+
+[23/07/18] 我们开发了支持训练和测试的**浏览器一体化界面**。请使用 `train_web.py` 在您的浏览器中微调模型。感谢 [@KanadeSiina](https://github.com/KanadeSiina) 和 [@codemayq](https://github.com/codemayq) 在该功能开发中付出的努力。
+
+[23/07/09] 我们开源了 **[FastEdit](https://github.com/hiyouga/FastEdit)** ⚡🩹，一个简单易用的、能迅速编辑大模型事实记忆的工具包。如果您感兴趣请关注我们的 [FastEdit](https://github.com/hiyouga/FastEdit) 项目。
+
+[23/06/29] 我们提供了一个**可复现的**指令模型微调示例，详细内容请查阅 [Baichuan-7B-sft](https://huggingface.co/hiyouga/Baichuan-7B-sft)。
+
+[23/06/22] 我们对齐了[示例 API](src/api_demo.py) 与 [OpenAI API](https://platform.openai.com/docs/api-reference/chat) 的格式，您可以将微调模型接入**任意基于 ChatGPT 的应用**中。
+
+[23/06/03] 我们实现了 4 比特的 LoRA 训练（也称 **[QLoRA](https://github.com/artidoro/qlora)**）。详细用法请参照 [examples](examples/README_zh.md)。
+
+</details>
 
 ## 模型
 
-- [LLaMA](https://github.com/facebookresearch/llama) (7B/13B/33B/65B)
-- [LLaMA-2](https://huggingface.co/meta-llama) (7B/13B/70B)
-- [BLOOM](https://huggingface.co/bigscience/bloom) & [BLOOMZ](https://huggingface.co/bigscience/bloomz) (560M/1.1B/1.7B/3B/7.1B/176B)
-- [Falcon](https://huggingface.co/tiiuae/falcon-7b) (7B/40B)
-- [Baichuan](https://huggingface.co/baichuan-inc/baichuan-7B) (7B/13B)
-- [InternLM](https://github.com/InternLM/InternLM) (7B)
+| 模型名                                                   | 模型大小                          | 默认模块           | Template  |
+| -------------------------------------------------------- | -------------------------------- | ----------------- | --------- |
+| [Baichuan2](https://huggingface.co/baichuan-inc)         | 7B/13B                           | W_pack            | baichuan2 |
+| [BLOOM](https://huggingface.co/bigscience)               | 560M/1.1B/1.7B/3B/7.1B/176B      | query_key_value   | -         |
+| [BLOOMZ](https://huggingface.co/bigscience)              | 560M/1.1B/1.7B/3B/7.1B/176B      | query_key_value   | -         |
+| [ChatGLM3](https://huggingface.co/THUDM)                 | 6B                               | query_key_value   | chatglm3  |
+| [Command-R](https://huggingface.co/CohereForAI)          | 35B/104B                         | q_proj,v_proj     | cohere    |
+| [DeepSeek (MoE)](https://huggingface.co/deepseek-ai)     | 7B/16B/67B/236B                  | q_proj,v_proj     | deepseek  |
+| [Falcon](https://huggingface.co/tiiuae)                  | 7B/40B/180B                      | query_key_value   | falcon    |
+| [Gemma/CodeGemma](https://huggingface.co/google)         | 2B/7B                            | q_proj,v_proj     | gemma     |
+| [InternLM2](https://huggingface.co/internlm)             | 7B/20B                           | wqkv              | intern2   |
+| [LLaMA](https://github.com/facebookresearch/llama)       | 7B/13B/33B/65B                   | q_proj,v_proj     | -         |
+| [LLaMA-2](https://huggingface.co/meta-llama)             | 7B/13B/70B                       | q_proj,v_proj     | llama2    |
+| [LLaMA-3](https://huggingface.co/meta-llama)             | 8B/70B                           | q_proj,v_proj     | llama3    |
+| [LLaVA-1.5](https://huggingface.co/llava-hf)             | 7B/13B                           | q_proj,v_proj     | vicuna    |
+| [Mistral/Mixtral](https://huggingface.co/mistralai)      | 7B/8x7B/8x22B                    | q_proj,v_proj     | mistral   |
+| [OLMo](https://huggingface.co/allenai)                   | 1B/7B                            | q_proj,v_proj     | -         |
+| [Phi-1.5/2](https://huggingface.co/microsoft)            | 1.3B/2.7B                        | q_proj,v_proj     | -         |
+| [Phi-3](https://huggingface.co/microsoft)                | 3.8B                             | qkv_proj          | phi       |
+| [Qwen](https://huggingface.co/Qwen)                      | 1.8B/7B/14B/72B                  | c_attn            | qwen      |
+| [Qwen1.5 (Code/MoE)](https://huggingface.co/Qwen)        | 0.5B/1.8B/4B/7B/14B/32B/72B/110B | q_proj,v_proj     | qwen      |
+| [StarCoder2](https://huggingface.co/bigcode)             | 3B/7B/15B                        | q_proj,v_proj     | -         |
+| [XVERSE](https://huggingface.co/xverse)                  | 7B/13B/65B                       | q_proj,v_proj     | xverse    |
+| [Yi (1/1.5)](https://huggingface.co/01-ai)               | 6B/9B/34B                        | q_proj,v_proj     | yi        |
+| [Yi-VL](https://huggingface.co/01-ai)                    | 6B/34B                           | q_proj,v_proj     | yi_vl     |
+| [Yuan](https://huggingface.co/IEITYuan)                  | 2B/51B/102B                      | q_proj,v_proj     | yuan      |
 
-## 微调方法
+> [!NOTE]
+> **默认模块**应作为 `--lora_target` 参数的默认值，可使用 `--lora_target all` 参数指定全部模块以取得更好的效果。
+>
+> 对于所有“基座”（Base）模型，`--template` 参数可以是 `default`, `alpaca`, `vicuna` 等任意值。但“对话”（Instruct/Chat）模型请务必使用**对应的模板**。
+>
+> 请务必在训练和推理时使用**完全一致**的模板。
 
-- [二次预训练](https://s3-us-west-2.amazonaws.com/openai-assets/research-covers/language-unsupervised/language_understanding_paper.pdf)
-  - 全参数微调
-  - 部分参数微调
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
-- [指令监督微调](https://arxiv.org/abs/2109.01652)
-  - 全参数微调
-  - 部分参数微调
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
-- [人类反馈的强化学习（RLHF）](https://arxiv.org/abs/2203.02155)
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
+项目所支持模型的完整列表请参阅 [constants.py](src/llmtuner/extras/constants.py)。
+
+您也可以在 [template.py](src/llmtuner/data/template.py) 中添加自己的对话模板。
+
+## 训练方法
+
+| 方法                   |     全参数训练      |    部分参数训练     |       LoRA         |       QLoRA        |
+| ---------------------- | ------------------ | ------------------ | ------------------ | ------------------ |
+| 预训练                 | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| 指令监督微调            | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| 奖励模型训练            | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| PPO 训练               | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| DPO 训练               | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| ORPO 训练              | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
 
 ## 数据集
 
-- 用于二次预训练:
-  - [Wiki Demo (en)](data/wiki_demo.txt)
-  - [RefinedWeb (en)](https://huggingface.co/datasets/tiiuae/falcon-refinedweb)
-  - [StarCoder (en)](https://huggingface.co/datasets/bigcode/starcoderdata)
-  - [Wikipedia (en)](https://huggingface.co/datasets/olm/olm-wikipedia-20221220)
-  - [Wikipedia (zh)](https://huggingface.co/datasets/pleisto/wikipedia-cn-20230720-filtered)
-- 用于指令监督微调:
-  - [Stanford Alpaca (en)](https://github.com/tatsu-lab/stanford_alpaca)
-  - [Stanford Alpaca (zh)](https://github.com/ymcui/Chinese-LLaMA-Alpaca)
-  - [GPT-4 Generated Data (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
-  - [Open Assistant (multilingual)](https://huggingface.co/datasets/OpenAssistant/oasst1)
-  - [Self-cognition (zh)](data/self_cognition.json)
-  - [ShareGPT (zh)](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT/tree/main/Chinese-instruction-collection)
-  - [RefGPT (zh)](https://github.com/sufengniu/RefGPT)
-  - [Guanaco Dataset (multilingual)](https://huggingface.co/datasets/JosephusCheung/GuanacoDataset)
-  - [BELLE 2M (zh)](https://huggingface.co/datasets/BelleGroup/train_2M_CN)
-  - [BELLE 1M (zh)](https://huggingface.co/datasets/BelleGroup/train_1M_CN)
-  - [BELLE 0.5M (zh)](https://huggingface.co/datasets/BelleGroup/train_0.5M_CN)
-  - [BELLE Dialogue 0.4M (zh)](https://huggingface.co/datasets/BelleGroup/generated_chat_0.4M)
-  - [BELLE School Math 0.25M (zh)](https://huggingface.co/datasets/BelleGroup/school_math_0.25M)
-  - [BELLE Multiturn Chat 0.8M (zh)](https://huggingface.co/datasets/BelleGroup/multiturn_chat_0.8M)
-  - [Firefly 1.1M (zh)](https://huggingface.co/datasets/YeungNLP/firefly-train-1.1M)
-  - [LIMA (en)](https://huggingface.co/datasets/GAIR/lima)
-  - [CodeAlpaca 20k (en)](https://huggingface.co/datasets/sahil2801/CodeAlpaca-20k)
-  - [Alpaca CoT (multilingual)](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT)
-  - [Web QA (zh)](https://huggingface.co/datasets/suolyer/webqa)
-  - [UltraChat (en)](https://github.com/thunlp/UltraChat)
-  - [WebNovel (zh)](https://huggingface.co/datasets/zxbsmk/webnovel_cn)
-- 用于奖励模型训练:
-  - [HH-RLHF (en)](https://huggingface.co/datasets/Anthropic/hh-rlhf)
-  - [Open Assistant (multilingual)](https://huggingface.co/datasets/OpenAssistant/oasst1)
-  - [GPT-4 Generated Data (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
+<details><summary>预训练数据集</summary>
 
-使用方法请参考 [data/README.md](data/README_zh.md) 文件。
+- [Wiki Demo (en)](data/wiki_demo.txt)
+- [RefinedWeb (en)](https://huggingface.co/datasets/tiiuae/falcon-refinedweb)
+- [RedPajama V2 (en)](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-V2)
+- [Wikipedia (en)](https://huggingface.co/datasets/olm/olm-wikipedia-20221220)
+- [Wikipedia (zh)](https://huggingface.co/datasets/pleisto/wikipedia-cn-20230720-filtered)
+- [Pile (en)](https://huggingface.co/datasets/EleutherAI/pile)
+- [SkyPile (zh)](https://huggingface.co/datasets/Skywork/SkyPile-150B)
+- [The Stack (en)](https://huggingface.co/datasets/bigcode/the-stack)
+- [StarCoder (en)](https://huggingface.co/datasets/bigcode/starcoderdata)
+
+</details>
+
+<details><summary>指令微调数据集</summary>
+
+- [Stanford Alpaca (en)](https://github.com/tatsu-lab/stanford_alpaca)
+- [Stanford Alpaca (zh)](https://github.com/ymcui/Chinese-LLaMA-Alpaca)
+- [Alpaca GPT4 (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
+- [Identity (en&zh)](data/identity.json)
+- [Open Assistant (zh)](https://huggingface.co/datasets/OpenAssistant/oasst1)
+- [ShareGPT (zh)](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT/tree/main/Chinese-instruction-collection)
+- [Guanaco Dataset (multilingual)](https://huggingface.co/datasets/JosephusCheung/GuanacoDataset)
+- [BELLE 2M (zh)](https://huggingface.co/datasets/BelleGroup/train_2M_CN)
+- [BELLE 1M (zh)](https://huggingface.co/datasets/BelleGroup/train_1M_CN)
+- [BELLE 0.5M (zh)](https://huggingface.co/datasets/BelleGroup/train_0.5M_CN)
+- [BELLE Dialogue 0.4M (zh)](https://huggingface.co/datasets/BelleGroup/generated_chat_0.4M)
+- [BELLE School Math 0.25M (zh)](https://huggingface.co/datasets/BelleGroup/school_math_0.25M)
+- [BELLE Multiturn Chat 0.8M (zh)](https://huggingface.co/datasets/BelleGroup/multiturn_chat_0.8M)
+- [UltraChat (en)](https://github.com/thunlp/UltraChat)
+- [LIMA (en)](https://huggingface.co/datasets/GAIR/lima)
+- [OpenPlatypus (en)](https://huggingface.co/datasets/garage-bAInd/Open-Platypus)
+- [CodeAlpaca 20k (en)](https://huggingface.co/datasets/sahil2801/CodeAlpaca-20k)
+- [Alpaca CoT (multilingual)](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT)
+- [OpenOrca (en)](https://huggingface.co/datasets/Open-Orca/OpenOrca)
+- [SlimOrca (en)](https://huggingface.co/datasets/Open-Orca/SlimOrca)
+- [MathInstruct (en)](https://huggingface.co/datasets/TIGER-Lab/MathInstruct)
+- [Firefly 1.1M (zh)](https://huggingface.co/datasets/YeungNLP/firefly-train-1.1M)
+- [Wiki QA (en)](https://huggingface.co/datasets/wiki_qa)
+- [Web QA (zh)](https://huggingface.co/datasets/suolyer/webqa)
+- [WebNovel (zh)](https://huggingface.co/datasets/zxbsmk/webnovel_cn)
+- [Nectar (en)](https://huggingface.co/datasets/berkeley-nest/Nectar)
+- [deepctrl (en&zh)](https://www.modelscope.cn/datasets/deepctrl/deepctrl-sft-data)
+- [Ad Gen (zh)](https://huggingface.co/datasets/HasturOfficial/adgen)
+- [ShareGPT Hyperfiltered (en)](https://huggingface.co/datasets/totally-not-an-llm/sharegpt-hyperfiltered-3k)
+- [ShareGPT4 (en&zh)](https://huggingface.co/datasets/shibing624/sharegpt_gpt4)
+- [UltraChat 200k (en)](https://huggingface.co/datasets/HuggingFaceH4/ultrachat_200k)
+- [AgentInstruct (en)](https://huggingface.co/datasets/THUDM/AgentInstruct)
+- [LMSYS Chat 1M (en)](https://huggingface.co/datasets/lmsys/lmsys-chat-1m)
+- [Evol Instruct V2 (en)](https://huggingface.co/datasets/WizardLM/WizardLM_evol_instruct_V2_196k)
+- [Glaive Function Calling V2 (en)](https://huggingface.co/datasets/glaiveai/glaive-function-calling-v2)
+- [Cosmopedia (en)](https://huggingface.co/datasets/HuggingFaceTB/cosmopedia)
+- [LLaVA mixed (en&zh)](https://huggingface.co/datasets/BUAADreamer/llava-en-zh-300k)
+- [Open Assistant (de)](https://huggingface.co/datasets/mayflowergmbh/oasst_de)
+- [Dolly 15k (de)](https://huggingface.co/datasets/mayflowergmbh/dolly-15k_de)
+- [Alpaca GPT4 (de)](https://huggingface.co/datasets/mayflowergmbh/alpaca-gpt4_de)
+- [OpenSchnabeltier (de)](https://huggingface.co/datasets/mayflowergmbh/openschnabeltier_de)
+- [Evol Instruct (de)](https://huggingface.co/datasets/mayflowergmbh/evol-instruct_de)
+- [Dolphin (de)](https://huggingface.co/datasets/mayflowergmbh/dolphin_de)
+- [Booksum (de)](https://huggingface.co/datasets/mayflowergmbh/booksum_de)
+- [Airoboros (de)](https://huggingface.co/datasets/mayflowergmbh/airoboros-3.0_de)
+- [Ultrachat (de)](https://huggingface.co/datasets/mayflowergmbh/ultra-chat_de)
+
+</details>
+
+<details><summary>偏好数据集</summary>
+
+- [HH-RLHF (en)](https://huggingface.co/datasets/Anthropic/hh-rlhf)
+- [GPT-4 Generated Data (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
+- [Orca DPO (en)](https://huggingface.co/datasets/Intel/orca_dpo_pairs)
+- [Nectar (en)](https://huggingface.co/datasets/berkeley-nest/Nectar)
+- [DPO mixed (en&zh)](https://huggingface.co/datasets/hiyouga/DPO-En-Zh-20k)
+- [Open Assistant (zh)](https://huggingface.co/datasets/OpenAssistant/oasst1)
+- [Orca DPO (de)](https://huggingface.co/datasets/mayflowergmbh/intel_orca_dpo_pairs_de)
+
+</details>
 
 部分数据集的使用需要确认，我们推荐使用下述命令登录您的 Hugging Face 账户。
 
@@ -107,295 +277,263 @@ pip install --upgrade huggingface_hub
 huggingface-cli login
 ```
 
-## 软件依赖
+## 软硬件依赖
 
-- Python 3.8+ 和 PyTorch 1.13.1+
-- 🤗Transformers, Datasets, Accelerate, PEFT 和 TRL
-- jieba, rouge-chinese 和 nltk (用于评估)
-- gradio 和 matplotlib (用于网页端交互)
-- uvicorn, fastapi 和 sse-starlette (用于 API)
+| 必需项       | 至少     | 推荐      |
+| ------------ | ------- | --------- |
+| python       | 3.8     | 3.10      |
+| torch        | 1.13.1  | 2.2.0     |
+| transformers | 4.37.2  | 4.40.1    |
+| datasets     | 2.14.3  | 2.19.1    |
+| accelerate   | 0.27.2  | 0.30.0    |
+| peft         | 0.9.0   | 0.10.0    |
+| trl          | 0.8.1   | 0.8.6     |
 
-以及 **强而有力的 GPU**！
+| 可选项       | 至少     | 推荐      |
+| ------------ | ------- | --------- |
+| CUDA         | 11.6    | 12.2      |
+| deepspeed    | 0.10.0  | 0.14.0    |
+| bitsandbytes | 0.39.0  | 0.43.1    |
+| vllm         | 0.4.0   | 0.4.2     |
+| flash-attn   | 2.3.0   | 2.5.8     |
+
+### 硬件依赖
+
+\* *估算值*
+
+| 方法               | 精度 |   7B  |  13B  |  30B  |   70B  |  110B  |  8x7B |  8x22B |
+| ----------------- | ---- | ----- | ----- | ----- | ------ | ------ | ----- | ------ |
+| Full              | AMP  | 120GB | 240GB | 600GB | 1200GB | 2000GB | 900GB | 2400GB |
+| Full              |  16  |  60GB | 120GB | 300GB |  600GB |  900GB | 400GB | 1200GB |
+| Freeze            |  16  |  20GB |  40GB |  80GB |  200GB |  360GB | 160GB |  400GB |
+| LoRA/GaLore/BAdam |  16  |  16GB |  32GB |  64GB |  160GB |  240GB | 120GB |  320GB |
+| QLoRA             |   8  |  10GB |  20GB |  40GB |   80GB |  140GB |  60GB |  160GB |
+| QLoRA             |   4  |   6GB |  12GB |  24GB |   48GB |   72GB |  30GB |   96GB |
+| QLoRA             |   2  |   4GB |   8GB |  16GB |   24GB |   48GB |  18GB |   48GB |
 
 ## 如何使用
 
-### 数据准备（可跳过）
+### 安装 LLaMA Factory
 
-关于数据集文件的格式，请参考 `data/example_dataset` 文件夹的内容。构建自定义数据集时，既可以使用单个 `.json` 文件，也可以使用一个[数据加载脚本](https://huggingface.co/docs/datasets/dataset_script)和多个文件。
-
-注意：使用自定义数据集时，请更新 `data/dataset_info.json` 文件，该文件的格式请参考 `data/README.md`。
-
-### 环境搭建（可跳过）
+> [!IMPORTANT]
+> 此步骤为必需。
 
 ```bash
-git lfs install
-git clone https://github.com/hiyouga/LLaMA-Efficient-Tuning.git
-conda create -n llama_etuning python=3.10
-conda activate llama_etuning
-cd LLaMA-Efficient-Tuning
-pip install -r requirements.txt
+git clone https://github.com/hiyouga/LLaMA-Factory.git
+cd LLaMA-Factory
+pip install -e .[torch,metrics]
 ```
 
-如果要在 Windows 平台上开启量化 LoRA（QLoRA），需要安装预编译的 `bitsandbytes` 库, 支持 CUDA 11.1 到 12.1.
+可选的额外依赖项：torch、metrics、deepspeed、bitsandbytes、vllm、galore、badam、gptq、awq、aqlm、qwen、modelscope、quality
+
+> [!TIP]
+> 遇到包冲突时，可使用 `pip install --no-deps -e .` 解决。
+
+<details><summary>Windows 用户指南</summary>
+
+如果要在 Windows 平台上开启量化 LoRA（QLoRA），需要安装预编译的 `bitsandbytes` 库, 支持 CUDA 11.1 到 12.2, 请根据您的 CUDA 版本情况选择适合的[发布版本](https://github.com/jllllll/bitsandbytes-windows-webui/releases/tag/wheels)。
 
 ```bash
-pip install https://github.com/jllllll/bitsandbytes-windows-webui/releases/download/wheels/bitsandbytes-0.39.1-py3-none-win_amd64.whl
+pip install https://github.com/jllllll/bitsandbytes-windows-webui/releases/download/wheels/bitsandbytes-0.41.2.post2-py3-none-win_amd64.whl
 ```
 
-### 浏览器一键微调/测试
+如果要在 Windows 平台上开启 FlashAttention-2，需要安装预编译的 `flash-attn` 库，支持 CUDA 12.1 到 12.2，请根据需求到 [flash-attention](https://github.com/bdashore3/flash-attention/releases) 下载对应版本安装。
+
+</details>
+
+<details><summary>昇腾 NPU 用户指南</summary>
+
+如果使用昇腾 NPU 设备进行（分布式）训练或推理，需要安装 **[torch-npu](https://gitee.com/ascend/pytorch)** 库和 **[Ascend CANN Kernels](https://www.hiascend.com/developer/download/community/result?module=cann)**。
+
+| 依赖项       | 至少     | 推荐      |
+| ------------ | ------- | --------- |
+| CANN         | 8.0.RC1 | 8.0.RC1   |
+| torch        | 2.2.0   | 2.2.0     |
+| torch-npu    | 2.2.0   | 2.2.0     |
+| deepspeed    | 0.13.2  | 0.13.2    |
+
+Docker 镜像：
+
+- 32GB：[下载地址](http://mirrors.cn-central-221.ovaijisuan.com/detail/130.html)
+- 64GB：敬请期待
+
+请记得使用 `ASCEND_RT_VISIBLE_DEVICES` 而非 `CUDA_VISIBLE_DEVICES` 来指定您使用的设备。
+
+如果遇到无法正常推理的情况，请尝试设置 `do_sample: false`。
+
+</details>
+
+### 数据准备
+
+关于数据集文件的格式，请参考 [data/README_zh.md](data/README_zh.md) 的内容。你可以使用 HuggingFace / ModelScope 上的数据集或加载本地数据集。
+
+> [!NOTE]
+> 使用自定义数据集时，请更新 `data/dataset_info.json` 文件。
+
+### 快速开始
+
+下面三行命令分别对 Llama3-8B-Instruct 模型进行 LoRA **微调**、**推理**和**合并**。
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_web.py
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_sft.yaml
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli chat examples/inference/llama3_lora_sft.yaml
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli export examples/merge_lora/llama3_lora_sft.yaml
 ```
 
-目前网页 UI 仅支持**单卡训练**。
+高级用法请参考 [examples/README_zh.md](examples/README_zh.md)（包括多 GPU 微调）。
 
-### 二次预训练
+> [!TIP]
+> 使用 `llamafactory-cli help` 显示帮助信息。
+
+### LLaMA Board 可视化微调（由 [Gradio](https://github.com/gradio-app/gradio) 驱动）
+
+> [!IMPORTANT]
+> LLaMA Board 可视化界面目前仅支持单 GPU 训练。
+
+#### 使用本地环境
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage pt \
-    --model_name_or_path path_to_your_model \
-    --do_train \
-    --dataset wiki_demo \
-    --template default \
-    --finetuning_type lora \
-    --output_dir path_to_pt_checkpoint \
-    --overwrite_cache \
-    --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 \
-    --plot_loss \
-    --fp16
+CUDA_VISIBLE_DEVICES=0 GRADIO_SHARE=1 llamafactory-cli webui
 ```
 
-### 指令监督微调
+<details><summary>阿里云 PAI 和 AutoDL 用户指南</summary>
+
+如果您在阿里云 PAI 上使用 LLaMA Board 时遇到显示问题，请尝试在启动前使用以下命令设置环境变量：
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage sft \
-    --model_name_or_path path_to_your_model \
-    --do_train \
-    --dataset alpaca_gpt4_zh \
-    --template default \
-    --finetuning_type lora \
-    --output_dir path_to_sft_checkpoint \
-    --overwrite_cache \
-    --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 \
-    --plot_loss \
-    --fp16
+export GRADIO_SERVER_PORT=7860 GRADIO_ROOT_PATH=/${JUPYTER_NAME}/proxy/7860/
 ```
 
-使用 Baichuan 模型时请指定 `--lora_target W_pack` 参数。
-
-### 奖励模型训练
+如果您正在使用 AutoDL，请安装下述 Gradio 版本：
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage rm \
-    --model_name_or_path path_to_your_model \
-    --do_train \
-    --dataset comparison_gpt4_zh \
-    --template default \
-    --finetuning_type lora \
-    --resume_lora_training False \
-    --checkpoint_dir path_to_sft_checkpoint \
-    --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 \
-    --plot_loss \
-    --fp16
-```
-
-### RLHF 训练
-
-```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage ppo \
-    --model_name_or_path path_to_your_model \
-    --do_train \
-    --dataset alpaca_gpt4_zh \
-    --template default \
-    --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 \
-    --plot_loss
-```
-
-### 多 GPU 分布式训练
-
-```bash
-accelerate config # 首先配置分布式环境
-accelerate launch src/train_bash.py # 参数同上
-```
-
-<details><summary>使用 DeepSpeed ZeRO-2 进行全参数微调的 Accelerate 配置示例</summary>
-
-```yaml
-compute_environment: LOCAL_MACHINE
-deepspeed_config:
-  gradient_accumulation_steps: 4
-  gradient_clipping: 0.5
-  offload_optimizer_device: none
-  offload_param_device: none
-  zero3_init_flag: false
-  zero_stage: 2
-distributed_type: DEEPSPEED
-downcast_bf16: 'no'
-machine_rank: 0
-main_training_function: main
-mixed_precision: fp16
-num_machines: 1
-num_processes: 4
-rdzv_backend: static
-same_network: true
-tpu_env: []
-tpu_use_cluster: false
-tpu_use_sudo: false
-use_cpu: false
+pip install gradio==4.10.0
 ```
 
 </details>
 
-### 指标评估（BLEU分数和汉语ROUGE分数）
+#### 使用 Docker
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage sft \
-    --model_name_or_path path_to_your_model \
-    --do_eval \
-    --dataset alpaca_gpt4_zh \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint \
-    --output_dir path_to_eval_result \
-    --per_device_eval_batch_size 8 \
-    --max_samples 100 \
-    --predict_with_generate
+docker build -f ./Dockerfile -t llama-factory:latest .
+docker run --gpus=all \
+    -v ./hf_cache:/root/.cache/huggingface/ \
+    -v ./data:/app/data \
+    -v ./output:/app/output \
+    -e CUDA_VISIBLE_DEVICES=0 \
+    -p 7860:7860 \
+    --shm-size 16G \
+    --name llama_factory \
+    -d llama-factory:latest
 ```
 
-我们建议在量化模型的评估中使用 `--per_device_eval_batch_size=1` 和 `--max_target_length 128` 参数。
-
-### 模型预测
+#### 使用 Docker Compose
 
 ```bash
-CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
-    --stage sft \
-    --model_name_or_path path_to_your_model \
-    --do_predict \
-    --dataset alpaca_gpt4_zh \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint \
-    --output_dir path_to_predict_result \
-    --per_device_eval_batch_size 8 \
-    --max_samples 100 \
-    --predict_with_generate
+docker compose -f ./docker-compose.yml up -d
 ```
 
-### API 服务
+<details><summary>数据卷详情</summary>
+
+- hf_cache：使用宿主机的 Hugging Face 缓存文件夹，允许更改为新的目录。
+- data：宿主机中存放数据集的文件夹路径。
+- output：将导出目录设置为该路径后，即可在宿主机中访问导出后的模型。
+
+</details>
+
+### 利用 vLLM 部署 OpenAI API
 
 ```bash
-python src/api_demo.py \
-    --model_name_or_path path_to_your_model \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint
+CUDA_VISIBLE_DEVICES=0,1 API_PORT=8000 llamafactory-cli api examples/inference/llama3_vllm.yaml
 ```
 
-关于 API 文档请见 `http://localhost:8000/docs`。
+### 从魔搭社区下载
 
-### 命令行测试
+如果您在 Hugging Face 模型和数据集的下载中遇到了问题，可以通过下述方法使用魔搭社区。
 
 ```bash
-python src/cli_demo.py \
-    --model_name_or_path path_to_your_model \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint
+export USE_MODELSCOPE_HUB=1 # Windows 使用 `set USE_MODELSCOPE_HUB=1`
 ```
 
-### 浏览器测试
+将 `--model_name_or_path` 设置为模型 ID 来加载对应的模型。在[魔搭社区](https://modelscope.cn/models)查看所有可用的模型，例如 `LLM-Research/Meta-Llama-3-8B-Instruct`。
 
-```bash
-python src/web_demo.py \
-    --model_name_or_path path_to_your_model \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint
-```
+## 使用了 LLaMA Factory 的项目
 
-### 导出微调模型
+如果您有项目希望添加至下述列表，请通过邮件联系或者创建一个 PR。
 
-```bash
-python src/export_model.py \
-    --model_name_or_path path_to_your_model \
-    --template default \
-    --finetuning_type lora \
-    --checkpoint_dir path_to_checkpoint \
-    --output_dir path_to_export
-```
+<details><summary>点击显示</summary>
 
-## TODO
+1. Wang et al. ESRL: Efficient Sampling-based Reinforcement Learning for Sequence Generation. 2023. [[arxiv]](https://arxiv.org/abs/2308.02223)
+1. Yu et al. Open, Closed, or Small Language Models for Text Classification? 2023. [[arxiv]](https://arxiv.org/abs/2308.10092)
+1. Wang et al. UbiPhysio: Support Daily Functioning, Fitness, and Rehabilitation with Action Understanding and Feedback in Natural Language. 2023. [[arxiv]](https://arxiv.org/abs/2308.10526)
+1. Luceri et al. Leveraging Large Language Models to Detect Influence Campaigns in Social Media. 2023. [[arxiv]](https://arxiv.org/abs/2311.07816)
+1. Zhang et al. Alleviating Hallucinations of Large Language Models through Induced Hallucinations. 2023. [[arxiv]](https://arxiv.org/abs/2312.15710)
+1. Wang et al. Know Your Needs Better: Towards Structured Understanding of Marketer Demands with Analogical Reasoning Augmented LLMs. 2024. [[arxiv]](https://arxiv.org/abs/2401.04319)
+1. Wang et al. CANDLE: Iterative Conceptualization and Instantiation Distillation from Large Language Models for Commonsense Reasoning. 2024. [[arxiv]](https://arxiv.org/abs/2401.07286)
+1. Choi et al. FACT-GPT: Fact-Checking Augmentation via Claim Matching with LLMs. 2024. [[arxiv]](https://arxiv.org/abs/2402.05904)
+1. Zhang et al. AutoMathText: Autonomous Data Selection with Language Models for Mathematical Texts. 2024. [[arxiv]](https://arxiv.org/abs/2402.07625)
+1. Lyu et al. KnowTuning: Knowledge-aware Fine-tuning for Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2402.11176)
+1. Yang et al. LaCo: Large Language Model Pruning via Layer Collaps. 2024. [[arxiv]](https://arxiv.org/abs/2402.11187)
+1. Bhardwaj et al. Language Models are Homer Simpson! Safety Re-Alignment of Fine-tuned Language Models through Task Arithmetic. 2024. [[arxiv]](https://arxiv.org/abs/2402.11746)
+1. Yang et al. Enhancing Empathetic Response Generation by Augmenting LLMs with Small-scale Empathetic Models. 2024. [[arxiv]](https://arxiv.org/abs/2402.11801)
+1. Yi et al. Generation Meets Verification: Accelerating Large Language Model Inference with Smart Parallel Auto-Correct Decoding. 2024. [[arxiv]](https://arxiv.org/abs/2402.11809)
+1. Cao et al. Head-wise Shareable Attention for Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2402.11819)
+1. Zhang et al. Enhancing Multilingual Capabilities of Large Language Models through Self-Distillation from Resource-Rich Languages. 2024. [[arxiv]](https://arxiv.org/abs/2402.12204)
+1. Kim et al. Efficient and Effective Vocabulary Expansion Towards Multilingual Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2402.14714)
+1. Yu et al. KIEval: A Knowledge-grounded Interactive Evaluation Framework for Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2402.15043)
+1. Huang et al. Key-Point-Driven Data Synthesis with its Enhancement on Mathematical Reasoning. 2024. [[arxiv]](https://arxiv.org/abs/2403.02333)
+1. Duan et al. Negating Negatives: Alignment without Human Positive Samples via Distributional Dispreference Optimization. 2024. [[arxiv]](https://arxiv.org/abs/2403.03419)
+1. Xie and Schwertfeger. Empowering Robotics with Large Language Models: osmAG Map Comprehension with LLMs. 2024. [[arxiv]](https://arxiv.org/abs/2403.08228)
+1. Wu et al. Large Language Models are Parallel Multilingual Learners. 2024. [[arxiv]](https://arxiv.org/abs/2403.09073)
+1. Zhang et al. EDT: Improving Large Language Models' Generation by Entropy-based Dynamic Temperature Sampling. 2024. [[arxiv]](https://arxiv.org/abs/2403.14541)
+1. Weller et al. FollowIR: Evaluating and Teaching Information Retrieval Models to Follow Instructions. 2024. [[arxiv]](https://arxiv.org/abs/2403.15246)
+1. Hongbin Na. CBT-LLM: A Chinese Large Language Model for Cognitive Behavioral Therapy-based Mental Health Question Answering. 2024. [[arxiv]](https://arxiv.org/abs/2403.16008)
+1. Zan et al. CodeS: Natural Language to Code Repository via Multi-Layer Sketch. 2024. [[arxiv]](https://arxiv.org/abs/2403.16443)
+1. Liu et al. Extensive Self-Contrast Enables Feedback-Free Language Model Alignment. 2024. [[arxiv]](https://arxiv.org/abs/2404.00604)
+1. Luo et al. BAdam: A Memory Efficient Full Parameter Training Method for Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2404.02827)
+1. Du et al. Chinese Tiny LLM: Pretraining a Chinese-Centric Large Language Model. 2024. [[arxiv]](https://arxiv.org/abs/2404.04167)
+1. Ma et al. Parameter Efficient Quasi-Orthogonal Fine-Tuning via Givens Rotation. 2024. [[arxiv]](https://arxiv.org/abs/2404.04316)
+1. Liu et al. Dynamic Generation of Personalities with Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2404.07084)
+1. Shang et al. How Far Have We Gone in Stripped Binary Code Understanding Using Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2404.09836)
+1. Huang et al. LLMTune: Accelerate Database Knob Tuning with Large Language Models. 2024. [[arxiv]](https://arxiv.org/abs/2404.11581)
+1. Deng et al. Text-Tuple-Table: Towards Information Integration in Text-to-Table Generation via Global Tuple Extraction. 2024. [[arxiv]](https://arxiv.org/abs/2404.14215)
+1. Acikgoz et al. Hippocrates: An Open-Source Framework for Advancing Large Language Models in Healthcare. 2024. [[arxiv]](https://arxiv.org/abs/2404.16621)
+1. Zhang et al. Small Language Models Need Strong Verifiers to Self-Correct Reasoning. 2024. [[arxiv]](https://arxiv.org/abs/2404.17140)
+1. Zhou et al. FREB-TQA: A Fine-Grained Robustness Evaluation Benchmark for Table Question Answering. 2024. [[arxiv]](https://arxiv.org/abs/2404.18585)
+1. **[StarWhisper](https://github.com/Yu-Yang-Li/StarWhisper)**: 天文大模型 StarWhisper，基于 ChatGLM2-6B 和 Qwen-14B 在天文数据上微调而得。
+1. **[DISC-LawLLM](https://github.com/FudanDISC/DISC-LawLLM)**: 中文法律领域大模型 DISC-LawLLM，基于 Baichuan-13B 微调而得，具有法律推理和知识检索能力。
+1. **[Sunsimiao](https://github.com/thomas-yanxin/Sunsimiao)**: 孙思邈中文医疗大模型 Sumsimiao，基于 Baichuan-7B 和 ChatGLM-6B 在中文医疗数据上微调而得。
+1. **[CareGPT](https://github.com/WangRongsheng/CareGPT)**: 医疗大模型项目 CareGPT，基于 LLaMA2-7B 和 Baichuan-13B 在中文医疗数据上微调而得。
+1. **[MachineMindset](https://github.com/PKU-YuanGroup/Machine-Mindset/)**：MBTI性格大模型项目，根据数据集与训练方式让任意 LLM 拥有 16 个不同的性格类型。
+1. **[Luminia-13B-v3](https://huggingface.co/Nekochu/Luminia-13B-v3)**：一个用于生成 Stable Diffusion 提示词的大型语言模型。[[🤗Demo]](https://huggingface.co/spaces/Nekochu/Luminia-13B_SD_Prompt)
+1. **[Chinese-LLaVA-Med](https://github.com/BUAADreamer/Chinese-LLaVA-Med)**：中文多模态医学大模型，基于 LLaVA-1.5-7B 在中文多模态医疗数据上微调而得。
 
-- [ ] 实现 flash attention ([torch](https://pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html) / [xformers](https://github.com/facebookresearch/xformers) / [flashattn](https://github.com/Dao-AILab/flash-attention))。
-- [ ] 在推理阶段使用 Multi-query attention 进行加速。
-- [ ] 支持 RLHF 的全参数微调。
+</details>
 
 ## 协议
 
 本仓库的代码依照 [Apache-2.0](LICENSE) 协议开源。
 
-使用模型权重时，请遵循对应的模型协议：
-
-- [LLaMA](https://github.com/facebookresearch/llama/blob/main/MODEL_CARD.md)
-- [LLaMA-2](https://ai.meta.com/llama/license/)
-- [BLOOM](https://huggingface.co/spaces/bigscience/license)
-- [Falcon](LICENSE)
-- [Baichuan](https://huggingface.co/baichuan-inc/baichuan-7B/resolve/main/baichuan-7B%20%E6%A8%A1%E5%9E%8B%E8%AE%B8%E5%8F%AF%E5%8D%8F%E8%AE%AE.pdf)
-- [InternLM](https://github.com/InternLM/InternLM#open-source-license)
+使用模型权重时，请遵循对应的模型协议：[Baichuan2](https://huggingface.co/baichuan-inc/Baichuan2-7B-Base/blob/main/Community%20License%20for%20Baichuan%202%20Model.pdf) / [BLOOM](https://huggingface.co/spaces/bigscience/license) / [ChatGLM3](https://github.com/THUDM/ChatGLM3/blob/main/MODEL_LICENSE) / [Command-R](https://cohere.com/c4ai-cc-by-nc-license) / [DeepSeek](https://github.com/deepseek-ai/DeepSeek-LLM/blob/main/LICENSE-MODEL) / [Falcon](https://huggingface.co/tiiuae/falcon-180B/blob/main/LICENSE.txt) / [Gemma](https://ai.google.dev/gemma/terms) / [InternLM2](https://github.com/InternLM/InternLM#license) / [LLaMA](https://github.com/facebookresearch/llama/blob/main/MODEL_CARD.md) / [LLaMA-2 (LLaVA-1.5)](https://ai.meta.com/llama/license/) / [LLaMA-3](https://llama.meta.com/llama3/license/) / [Mistral](LICENSE) / [OLMo](LICENSE) / [Phi-1.5/2](https://huggingface.co/microsoft/phi-1_5/resolve/main/Research%20License.docx) / [Phi-3](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct/blob/main/LICENSE) / [Qwen](https://github.com/QwenLM/Qwen/blob/main/Tongyi%20Qianwen%20LICENSE%20AGREEMENT) / [StarCoder2](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) / [XVERSE](https://github.com/xverse-ai/XVERSE-13B/blob/main/MODEL_LICENSE.pdf) / [Yi](https://huggingface.co/01-ai/Yi-6B/blob/main/LICENSE) / [Yi-1.5](LICENSE) / [Yuan](https://github.com/IEIT-Yuan/Yuan-2.0/blob/main/LICENSE-Yuan)
 
 ## 引用
 
 如果您觉得此项目有帮助，请考虑以下列格式引用
 
 ```bibtex
-@Misc{llama-efficient-tuning,
-  title = {LLaMA Efficient Tuning},
-  author = {hiyouga},
-  howpublished = {\url{https://github.com/hiyouga/LLaMA-Efficient-Tuning}},
-  year = {2023}
+@article{zheng2024llamafactory,
+  title={LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models}, 
+  author={Yaowei Zheng and Richong Zhang and Junhao Zhang and Yanhan Ye and Zheyan Luo and Yongqiang Ma},
+  journal={arXiv preprint arXiv:2403.13372},
+  year={2024},
+  url={http://arxiv.org/abs/2403.13372}
 }
 ```
 
 ## 致谢
 
-本项目是 [ChatGLM-Efficient-Tuning](https://github.com/hiyouga/ChatGLM-Efficient-Tuning) 的同类项目。采用了类似的代码结构和训练方法。
+本项目受益于 [PEFT](https://github.com/huggingface/peft)、[TRL](https://github.com/huggingface/trl)、[QLoRA](https://github.com/artidoro/qlora) 和 [FastChat](https://github.com/lm-sys/FastChat)，感谢以上诸位作者的付出。
 
 ## Star History
 
-![Star History Chart](https://api.star-history.com/svg?repos=hiyouga/LLaMA-Efficient-Tuning&type=Date)
+![Star History Chart](https://api.star-history.com/svg?repos=hiyouga/LLaMA-Factory&type=Date)

data/README.md

@@ -1,18 +1,212 @@
-If you are using a custom dataset, please provide your dataset definition in the following format in `dataset_info.json`.
+If you are using a custom dataset, please add your **dataset description** to `dataset_info.json` according to the following format. We also provide several examples in the next section.
 
 ```json
 "dataset_name": {
-  "hf_hub_url": "the name of the dataset repository on the HuggingFace hub. (if specified, ignore below 3 arguments)",
-  "script_url": "the name of the directory containing a dataset loading script. (if specified, ignore below 2 arguments)",
-  "file_name": "the name of the dataset file in the this directory. (required if above are not specified)",
-  "file_sha1": "the SHA-1 hash value of the dataset file. (optional)",
-  "columns": {
-    "prompt": "the name of the column in the datasets containing the prompts. (default: instruction)",
-    "query": "the name of the column in the datasets containing the queries. (default: input)",
-    "response": "the name of the column in the datasets containing the responses. (default: output)",
-    "history": "the name of the column in the datasets containing the history of chat. (default: None)"
+  "hf_hub_url": "the name of the dataset repository on the Hugging Face hub. (if specified, ignore script_url and file_name)",
+  "ms_hub_url": "the name of the dataset repository on the ModelScope hub. (if specified, ignore script_url and file_name)",
+  "script_url": "the name of the directory containing a dataset loading script. (if specified, ignore file_name)",
+  "file_name": "the name of the dataset file in this directory. (required if above are not specified)",
+  "file_sha1": "the SHA-1 hash value of the dataset file. (optional, does not affect training)",
+  "subset": "the name of the subset. (optional, default: None)",
+  "folder": "the name of the folder of the dataset repository on the Hugging Face hub. (optional, default: None)",
+  "ranking": "whether the dataset is a preference dataset or not. (default: false)",
+  "formatting": "the format of the dataset. (optional, default: alpaca, can be chosen from {alpaca, sharegpt})",
+  "columns (optional)": {
+    "prompt": "the column name in the dataset containing the prompts. (default: instruction)",
+    "query": "the column name in the dataset containing the queries. (default: input)",
+    "response": "the column name in the dataset containing the responses. (default: output)",
+    "history": "the column name in the dataset containing the histories. (default: None)",
+    "messages": "the column name in the dataset containing the messages. (default: conversations)",
+    "system": "the column name in the dataset containing the system prompts. (default: None)",
+    "tools": "the column name in the dataset containing the tool description. (default: None)",
+    "images": "the column name in the dataset containing the image inputs. (default: None)"
+  },
+  "tags (optional, used for the sharegpt format)": {
+    "role_tag": "the key in the message represents the identity. (default: from)",
+    "content_tag": "the key in the message represents the content. (default: value)",
+    "user_tag": "the value of the role_tag represents the user. (default: human)",
+    "assistant_tag": "the value of the role_tag represents the assistant. (default: gpt)",
+    "observation_tag": "the value of the role_tag represents the tool results. (default: observation)",
+    "function_tag": "the value of the role_tag represents the function call. (default: function_call)",
+    "system_tag": "the value of the role_tag represents the system prompt. (default: system, can override system column)"
   }
 }
 ```
 
-where the `prompt` and `response` columns should contain non-empty values. The `query` column will be concatenated with the `prompt` column and used as input for the model. The `history` column should contain a list where each element is a string tuple representing a query-response pair.
+After that, you can load the custom dataset by specifying `--dataset dataset_name`.
+
+----
+
+Currently we support dataset in **alpaca** or **sharegpt** format, the dataset in alpaca format should follow the below format:
+
+```json
+[
+  {
+    "instruction": "user instruction (required)",
+    "input": "user input (optional)",
+    "output": "model response (required)",
+    "system": "system prompt (optional)",
+    "history": [
+      ["user instruction in the first round (optional)", "model response in the first round (optional)"],
+      ["user instruction in the second round (optional)", "model response in the second round (optional)"]
+    ]
+  }
+]
+```
+
+Regarding the above dataset, the description in `dataset_info.json` should be:
+
+```json
+"dataset_name": {
+  "file_name": "data.json",
+  "columns": {
+    "prompt": "instruction",
+    "query": "input",
+    "response": "output",
+    "system": "system",
+    "history": "history"
+  }
+}
+```
+
+The `query` column will be concatenated with the `prompt` column and used as the user prompt, then the user prompt would be `prompt\nquery`. The `response` column represents the model response.
+
+The `system` column will be used as the system prompt. The `history` column is a list consisting string tuples representing prompt-response pairs in the history. Note that the responses in the history **will also be used for training** in supervised fine-tuning.
+
+For the **pre-training datasets**, only the `prompt` column will be used for training, for example:
+
+```json
+[
+  {"text": "document"},
+  {"text": "document"}
+]
+```
+
+Regarding the above dataset, the description in `dataset_info.json` should be:
+
+```json
+"dataset_name": {
+  "file_name": "data.json",
+  "columns": {
+    "prompt": "text"
+  }
+}
+```
+
+For the **preference datasets**, the `response` column should be a string list whose length is 2, with the preferred answers appearing first, for example:
+
+```json
+[
+  {
+    "instruction": "user instruction",
+    "input": "user input",
+    "output": [
+      "chosen answer",
+      "rejected answer"
+    ]
+  }
+]
+```
+
+Regarding the above dataset, the description in `dataset_info.json` should be:
+
+```json
+"dataset_name": {
+  "file_name": "data.json",
+  "ranking": true,
+  "columns": {
+    "prompt": "instruction",
+    "query": "input",
+    "response": "output",
+  }
+}
+```
+
+----
+
+The dataset in **sharegpt** format should follow the below format:
+
+```json
+[
+  {
+    "conversations": [
+      {
+        "from": "human",
+        "value": "user instruction"
+      },
+      {
+        "from": "gpt",
+        "value": "model response"
+      }
+    ],
+    "system": "system prompt (optional)",
+    "tools": "tool description (optional)"
+  }
+]
+```
+
+Regarding the above dataset, the description in `dataset_info.json` should be:
+
+```json
+"dataset_name": {
+  "file_name": "data.json",
+  "formatting": "sharegpt",
+  "columns": {
+    "messages": "conversations",
+    "system": "system",
+    "tools": "tools"
+  },
+  "tags": {
+    "role_tag": "from",
+    "content_tag": "value",
+    "user_tag": "human",
+    "assistant_tag": "gpt"
+  }
+}
+```
+
+where the `messages` column should be a list following the `u/a/u/a/u/a` order.
+
+We also supports the dataset in the **openai** format:
+
+```json
+[
+  {
+    "messages": [
+      {
+        "role": "system",
+        "content": "system prompt (optional)"
+      },
+      {
+        "role": "user",
+        "content": "user instruction"
+      },
+      {
+        "role": "assistant",
+        "content": "model response"
+      }
+    ]
+  }
+]
+```
+
+Regarding the above dataset, the description in `dataset_info.json` should be:
+
+```json
+"dataset_name": {
+  "file_name": "data.json",
+  "formatting": "sharegpt",
+  "columns": {
+    "messages": "messages"
+  },
+  "tags": {
+    "role_tag": "role",
+    "content_tag": "content",
+    "user_tag": "user",
+    "assistant_tag": "assistant",
+    "system_tag": "system"
+  }
+}
+```
+
+Pre-training datasets and preference datasets are **incompatible** with the sharegpt format yet.

data/README_zh.md

@@ -1,18 +1,212 @@
-如果您使用自定义数据集，请务必在 `dataset_info.json` 文件中以如下格式提供您的数据集定义。
+如果您使用自定义数据集，请务必按照以下格式在 `dataset_info.json` 文件中添加**数据集描述**。我们在下面也提供了一些例子。
 
 ```json
 "数据集名称": {
-  "hf_hub_url": "HuggingFace上的项目地址（若指定，则忽略下列三个参数）",
-  "script_url": "包含数据加载脚本的本地文件夹名称（若指定，则忽略下列两个参数）",
+  "hf_hub_url": "Hugging Face 的数据集仓库地址（若指定，则忽略 script_url 和 file_name）",
+  "ms_hub_url": "ModelScope 的数据集仓库地址（若指定，则忽略 script_url 和 file_name）",
+  "script_url": "包含数据加载脚本的本地文件夹名称（若指定，则忽略 file_name）",
   "file_name": "该目录下数据集文件的名称（若上述参数未指定，则此项必需）",
-  "file_sha1": "数据集文件的SHA-1哈希值（可选）",
-  "columns": {
+  "file_sha1": "数据集文件的 SHA-1 哈希值（可选，留空不影响训练）",
+  "subset": "数据集子集的名称（可选，默认：None）",
+  "folder": "Hugging Face 仓库的文件夹名称（可选，默认：None）",
+  "ranking": "是否为偏好数据集（可选，默认：False）",
+  "formatting": "数据集格式（可选，默认：alpaca，可以为 alpaca 或 sharegpt）",
+  "columns（可选）": {
     "prompt": "数据集代表提示词的表头名称（默认：instruction）",
     "query": "数据集代表请求的表头名称（默认：input）",
     "response": "数据集代表回答的表头名称（默认：output）",
-    "history": "数据集代表历史对话的表头名称（默认：None）"
+    "history": "数据集代表历史对话的表头名称（默认：None）",
+    "messages": "数据集代表消息列表的表头名称（默认：conversations）",
+    "system": "数据集代表系统提示的表头名称（默认：None）",
+    "tools": "数据集代表工具描述的表头名称（默认：None）",
+    "images": "数据集代表图像输入的表头名称（默认：None）"
+  },
+  "tags（可选，用于 sharegpt 格式）": {
+    "role_tag": "消息中代表发送者身份的键名（默认：from）",
+    "content_tag": "消息中代表文本内容的键名（默认：value）",
+    "user_tag": "消息中代表用户的 role_tag（默认：human）",
+    "assistant_tag": "消息中代表助手的 role_tag（默认：gpt）",
+    "observation_tag": "消息中代表工具返回结果的 role_tag（默认：observation）",
+    "function_tag": "消息中代表工具调用的 role_tag（默认：function_call）",
+    "system_tag": "消息中代表系统提示的 role_tag（默认：system，会覆盖 system 列）"
   }
 }
 ```
 
-其中 `prompt` 和 `response` 列应当是非空的字符串。`query` 列的内容将会和 `prompt` 列拼接作为模型输入。`history` 列应当是一个列表，其中每个元素是一个字符串二元组，分别代表用户请求和模型答复。
+然后，可通过使用 `--dataset 数据集名称` 参数加载自定义数据集。
+
+----
+
+该项目目前支持两种格式的数据集：**alpaca** 和 **sharegpt**，其中 alpaca 格式的数据集按照以下方式组织：
+
+```json
+[
+  {
+    "instruction": "用户指令（必填）",
+    "input": "用户输入（选填）",
+    "output": "模型回答（必填）",
+    "system": "系统提示词（选填）",
+    "history": [
+      ["第一轮指令（选填）", "第一轮回答（选填）"],
+      ["第二轮指令（选填）", "第二轮回答（选填）"]
+    ]
+  }
+]
+```
+
+对于上述格式的数据，`dataset_info.json` 中的描述应为：
+
+```json
+"数据集名称": {
+  "file_name": "data.json",
+  "columns": {
+    "prompt": "instruction",
+    "query": "input",
+    "response": "output",
+    "system": "system",
+    "history": "history"
+  }
+}
+```
+
+其中 `query` 列对应的内容会与 `prompt` 列对应的内容拼接后作为用户指令，即用户指令为 `prompt\nquery`。`response` 列对应的内容为模型回答。
+
+`system` 列对应的内容将被作为系统提示词。`history` 列是由多个字符串二元组构成的列表，分别代表历史消息中每轮的指令和回答。注意在指令监督学习时，历史消息中的回答**也会被用于训练**。
+
+对于**预训练数据集**，仅 `prompt` 列中的内容会用于模型训练，例如：
+
+```json
+[
+  {"text": "document"},
+  {"text": "document"}
+]
+```
+
+对于上述格式的数据，`dataset_info.json` 中的描述应为：
+
+```json
+"数据集名称": {
+  "file_name": "data.json",
+  "columns": {
+    "prompt": "text"
+  }
+}
+```
+
+对于**偏好数据集**，`response` 列应当是一个长度为 2 的字符串列表，排在前面的代表更优的回答，例如：
+
+```json
+[
+  {
+    "instruction": "用户指令",
+    "input": "用户输入",
+    "output": [
+      "优质回答",
+      "劣质回答"
+    ]
+  }
+]
+```
+
+对于上述格式的数据，`dataset_info.json` 中的描述应为：
+
+```json
+"数据集名称": {
+  "file_name": "data.json",
+  "ranking": true,
+  "columns": {
+    "prompt": "instruction",
+    "query": "input",
+    "response": "output",
+  }
+}
+```
+
+----
+
+而 **sharegpt** 格式的数据集按照以下方式组织：
+
+```json
+[
+  {
+    "conversations": [
+      {
+        "from": "human",
+        "value": "用户指令"
+      },
+      {
+        "from": "gpt",
+        "value": "模型回答"
+      }
+    ],
+    "system": "系统提示词（选填）",
+    "tools": "工具描述（选填）"
+  }
+]
+```
+
+对于上述格式的数据，`dataset_info.json` 中的描述应为：
+
+```json
+"数据集名称": {
+  "file_name": "data.json",
+  "formatting": "sharegpt",
+  "columns": {
+    "messages": "conversations",
+    "system": "system",
+    "tools": "tools"
+  },
+  "tags": {
+    "role_tag": "from",
+    "content_tag": "value",
+    "user_tag": "human",
+    "assistant_tag": "gpt"
+  }
+}
+```
+
+其中 `messages` 列应当是一个列表，且符合 `用户/模型/用户/模型/用户/模型` 的顺序。
+
+我们同样支持 **openai** 格式的数据集：
+
+```json
+[
+  {
+    "messages": [
+      {
+        "role": "system",
+        "content": "系统提示词（选填）"
+      },
+      {
+        "role": "user",
+        "content": "用户指令"
+      },
+      {
+        "role": "assistant",
+        "content": "模型回答"
+      }
+    ]
+  }
+]
+```
+
+对于上述格式的数据，`dataset_info.json` 中的描述应为：
+
+```json
+"数据集名称": {
+  "file_name": "data.json",
+  "formatting": "sharegpt",
+  "columns": {
+    "messages": "messages"
+  },
+  "tags": {
+    "role_tag": "role",
+    "content_tag": "content",
+    "user_tag": "user",
+    "assistant_tag": "assistant",
+    "system_tag": "system"
+  }
+}
+```
+
+预训练数据集和偏好数据集**尚不支持** sharegpt 格式。

data/alpaca_data_zh_51k.json.REMOVED.git-id

@@ -1 +1 @@
-fc9a6a3458caca2af8dafc6181773fe10c6d8657
+a97cf9475291591843976554878568e046d8a46d

data/belle_multiturn/belle_multiturn.py

@@ -1,7 +1,10 @@
 import json
-import datasets
-from typing import Any, Dict, List
+import os
 
+import datasets
+
+
+_HF_ENDPOINT = os.getenv("HF_ENDPOINT", "https://huggingface.co")
 
 _DESCRIPTION = "BELLE multiturn chat dataset."
 
@@ -14,66 +17,51 @@ _CITATION = """\
 }
 """
 
-_HOMEPAGE = "https://huggingface.co/datasets/BelleGroup/multiturn_chat_0.8M"
+_HOMEPAGE = "{}/datasets/BelleGroup/multiturn_chat_0.8M".format(_HF_ENDPOINT)
 _LICENSE = "gpl-3.0"
-_URL = "https://huggingface.co/datasets/BelleGroup/multiturn_chat_0.8M/resolve/main/multiturn_chat_0.8M.json"
+_URL = "{}/datasets/BelleGroup/multiturn_chat_0.8M/resolve/main/multiturn_chat_0.8M.json".format(_HF_ENDPOINT)
 
 
 class BelleMultiturn(datasets.GeneratorBasedBuilder):
-
     VERSION = datasets.Version("0.0.0")
 
-    def _info(self) -> datasets.DatasetInfo:
-        features = datasets.Features({
-            "instruction": datasets.Value("string"),
-            "output": datasets.Value("string"),
-            "history": datasets.Sequence(datasets.Sequence(datasets.Value("string")))
-        })
+    def _info(self):
+        features = datasets.Features(
+            {"conversations": [{"from": datasets.Value("string"), "value": datasets.Value("string")}]}
+        )
         return datasets.DatasetInfo(
-            description=_DESCRIPTION,
-            features=features,
-            homepage=_HOMEPAGE,
-            license=_LICENSE,
-            citation=_CITATION
+            description=_DESCRIPTION, features=features, homepage=_HOMEPAGE, license=_LICENSE, citation=_CITATION
         )
 
-    def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
+    def _split_generators(self, dl_manager: datasets.DownloadManager):
         file_path = dl_manager.download(_URL)
-        return [
-            datasets.SplitGenerator(
-                name=datasets.Split.TRAIN,
-                gen_kwargs={
-                    "filepath": file_path
-                }
-            )
-        ]
+        return [datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": file_path})]
 
-    def _generate_examples(self, filepath: str) -> Dict[int, Dict[str, Any]]: # generate multi-turn chat with history
+    def _generate_examples(self, filepath: str):
         with open(filepath, "r", encoding="utf-8") as f:
             for key, row in enumerate(f):
                 data = json.loads(row)
+                conversations = []
                 prompt = data["instruction"].strip()
                 response = data["output"].strip()
 
                 assist_idx = prompt.rfind("Assistant:")
                 human_idx = prompt.rfind("Human:")
-                query = prompt[human_idx+6:assist_idx].strip()
+                query = prompt[human_idx + 6 : assist_idx].strip()
                 prompt = prompt[:human_idx].strip()
-                history = []
+                conversations.insert(0, {"from": "gpt", "value": response})
+                conversations.insert(0, {"from": "human", "value": query})
 
                 while prompt.rfind("Assistant:") != -1:
                     assist_idx = prompt.rfind("Assistant:")
                     human_idx = prompt.rfind("Human:")
                     if human_idx != -1:
-                        old_query = prompt[human_idx+6:assist_idx].strip()
-                        old_resp = prompt[assist_idx+10:].strip()
-                        history.insert(0, (old_query, old_resp))
+                        old_query = prompt[human_idx + 6 : assist_idx].strip()
+                        old_resp = prompt[assist_idx + 10 :].strip()
+                        conversations.insert(0, {"from": "gpt", "value": old_resp})
+                        conversations.insert(0, {"from": "human", "value": old_query})
                     else:
                         break
                     prompt = prompt[:human_idx].strip()
 
-                yield key, {
-                    "instruction": query,
-                    "output": response,
-                    "history": history
-                }
+                yield key, {"conversations": conversations}

data/example_dataset/example_dataset.py

@@ -1,9 +1,10 @@
 import json
+from typing import Any, Dict, Generator, List, Tuple
+
 import datasets
-from typing import Any, Dict, List
 
 
-_DESCRIPTION = "An example of dataset for LLaMA."
+_DESCRIPTION = "An example of dataset."
 _CITATION = ""
 _HOMEPAGE = ""
 _LICENSE = ""
@@ -11,36 +12,26 @@ _URL = "examples.json"
 
 
 class ExampleDataset(datasets.GeneratorBasedBuilder):
-
     VERSION = datasets.Version("0.0.0")
 
     def _info(self) -> datasets.DatasetInfo:
-        features = datasets.Features({
-            "instruction": datasets.Value("string"),
-            "input": datasets.Value("string"),
-            "output": datasets.Value("string"),
-            "history": datasets.Sequence(datasets.Sequence(datasets.Value("string")))
-        })
+        features = datasets.Features(
+            {
+                "instruction": datasets.Value("string"),
+                "input": datasets.Value("string"),
+                "output": datasets.Value("string"),
+                "history": datasets.Sequence(datasets.Sequence(datasets.Value("string"))),
+            }
+        )
         return datasets.DatasetInfo(
-            description=_DESCRIPTION,
-            features=features,
-            homepage=_HOMEPAGE,
-            license=_LICENSE,
-            citation=_CITATION
+            description=_DESCRIPTION, features=features, homepage=_HOMEPAGE, license=_LICENSE, citation=_CITATION
         )
 
     def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
         file_path = dl_manager.download(_URL)
-        return [
-            datasets.SplitGenerator(
-                name=datasets.Split.TRAIN,
-                gen_kwargs={
-                    "filepath": file_path
-                }
-            )
-        ]
+        return [datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": file_path})]
 
-    def _generate_examples(self, filepath: str) -> Dict[int, Dict[str, Any]]:
+    def _generate_examples(self, filepath: str) -> Generator[Tuple[int, Dict[str, Any]], None, None]:
         example_dataset = json.load(open(filepath, "r", encoding="utf-8"))
         for key, example in enumerate(example_dataset):
             yield key, example

data/glaive_toolcall_10k.json.REMOVED.git-id

@@ -0,0 +1 @@
+4748dff00d1dc42768a5b6cc772143c313017812

data/hh_rlhf_en/hh_rlhf_en.py

@@ -1,65 +1,55 @@
 import json
+import os
+from typing import List
+
 import datasets
-from typing import Any, Dict, List
 
 
-_DESCRIPTION = "Human preference data about helpfulness and harmlessness for ChatGLM."
+_HF_ENDPOINT = os.getenv("HF_ENDPOINT", "https://huggingface.co")
+_DESCRIPTION = "Human preference data about helpfulness and harmlessness."
 _CITATION = ""
-_HOMEPAGE = "https://huggingface.co/datasets/Anthropic/hh-rlhf"
+_HOMEPAGE = "{}/datasets/Anthropic/hh-rlhf".format(_HF_ENDPOINT)
 _LICENSE = "mit"
-_URL = "https://huggingface.co/datasets/Anthropic/hh-rlhf/resolve/main/"
+_URL = "{}/datasets/Anthropic/hh-rlhf/resolve/main/".format(_HF_ENDPOINT)
 _URLS = {
     "train": [
         _URL + "harmless-base/train.jsonl.gz",
         _URL + "helpful-base/train.jsonl.gz",
         _URL + "helpful-online/train.jsonl.gz",
-        _URL + "helpful-rejection-sampled/train.jsonl.gz"
+        _URL + "helpful-rejection-sampled/train.jsonl.gz",
     ],
     "test": [
         _URL + "harmless-base/test.jsonl.gz",
         _URL + "helpful-base/test.jsonl.gz",
         _URL + "helpful-online/test.jsonl.gz",
-        _URL + "helpful-rejection-sampled/test.jsonl.gz"
-    ]
+        _URL + "helpful-rejection-sampled/test.jsonl.gz",
+    ],
 }
 
 
 class HhRlhfEn(datasets.GeneratorBasedBuilder):
-
     VERSION = datasets.Version("0.0.0")
 
     def _info(self) -> datasets.DatasetInfo:
-        features = datasets.Features({
-            "instruction": datasets.Value("string"),
-            "output": datasets.Sequence(datasets.Value("string")),
-            "history": datasets.Sequence(datasets.Sequence(datasets.Value("string")))
-        })
+        features = datasets.Features(
+            {
+                "instruction": datasets.Value("string"),
+                "output": datasets.Sequence(datasets.Value("string")),
+                "history": datasets.Sequence(datasets.Sequence(datasets.Value("string"))),
+            }
+        )
         return datasets.DatasetInfo(
-            description=_DESCRIPTION,
-            features=features,
-            homepage=_HOMEPAGE,
-            license=_LICENSE,
-            citation=_CITATION
+            description=_DESCRIPTION, features=features, homepage=_HOMEPAGE, license=_LICENSE, citation=_CITATION
         )
 
-    def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
+    def _split_generators(self, dl_manager: datasets.DownloadManager):
         file_path = dl_manager.download_and_extract(_URLS)
         return [
-            datasets.SplitGenerator(
-                name=datasets.Split.TRAIN,
-                gen_kwargs={
-                    "filepaths": file_path["train"]
-                }
-            ),
-            datasets.SplitGenerator(
-                name=datasets.Split.TEST,
-                gen_kwargs={
-                    "filepaths": file_path["test"]
-                }
-            )
+            datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepaths": file_path["train"]}),
+            datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepaths": file_path["test"]}),
         ]
 
-    def _generate_examples(self, filepaths: List[str]) -> Dict[int, Dict[str, Any]]: # generate multi-turn chat for ChatGLM
+    def _generate_examples(self, filepaths: List[str]):
         key = 0
         for filepath in filepaths:
             with open(filepath, "r", encoding="utf-8") as f:
@@ -69,12 +59,12 @@ class HhRlhfEn(datasets.GeneratorBasedBuilder):
                     rejected = data["rejected"]
 
                     assist_idx = rejected.rfind("\n\nAssistant: ")
-                    r_reject = rejected[assist_idx+13:].strip()
+                    r_reject = rejected[assist_idx + 13 :].strip()
                     assist_idx = chosen.rfind("\n\nAssistant: ")
-                    r_accept = chosen[assist_idx+13:].strip()
+                    r_accept = chosen[assist_idx + 13 :].strip()
 
                     human_idx = chosen.rfind("\n\nHuman: ")
-                    query = chosen[human_idx+9:assist_idx].strip()
+                    query = chosen[human_idx + 9 : assist_idx].strip()
                     prompt = chosen[:human_idx]
                     history = []
 
@@ -82,16 +72,12 @@ class HhRlhfEn(datasets.GeneratorBasedBuilder):
                         assist_idx = prompt.rfind("\n\nAssistant: ")
                         human_idx = prompt.rfind("\n\nHuman: ")
                         if human_idx != -1:
-                            old_query = prompt[human_idx+9:assist_idx].strip()
-                            old_resp = prompt[assist_idx+13:].strip()
+                            old_query = prompt[human_idx + 9 : assist_idx].strip()
+                            old_resp = prompt[assist_idx + 13 :].strip()
                             history.insert(0, (old_query, old_resp))
                         else:
                             break
                         prompt = prompt[:human_idx]
 
-                    yield key, {
-                        "instruction": query,
-                        "output": [r_accept, r_reject],
-                        "history": history
-                    }
+                    yield key, {"instruction": query, "output": [r_accept, r_reject], "history": history}
                     key += 1

data/oaast_rm.json.REMOVED.git-id

@@ -1 +0,0 @@
-274079ea921762be356de85b18f13fa60b7ba8cb

data/oaast_sft.json.REMOVED.git-id

@@ -1 +0,0 @@
-57fd080be5bffe4153fe3ee26a175e3d56da30f3

data/orca_rlhf.json.REMOVED.git-id

@@ -0,0 +1 @@
+736bcedea2b24a1414765c6d69cbdafaea839f3c

data/refgpt_zh_50k_p1.json.REMOVED.git-id

@@ -1 +0,0 @@
-f967a4f6d04a11308a15524aa9a846a19a8d1e83

data/refgpt_zh_50k_p2.json.REMOVED.git-id

@@ -1 +0,0 @@
-0a4f0d74fd1c5cab2eb6d84a3a3fe669847becd8

data/sharegpt_zh_27k.json.REMOVED.git-id

@@ -1 +0,0 @@
-38c89869c6aeca2a3af9ea1e09afe460f9b46810

data/ultra_chat/ultra_chat.py

@@ -1,7 +1,11 @@
 import json
-import datasets
-from typing import Any, Dict, List
+import os
+from typing import List
 
+import datasets
+
+
+_HF_ENDPOINT = os.getenv("HF_ENDPOINT", "https://huggingface.co")
 
 _DESCRIPTION = "UltraChat: Large-scale, Informative, and Diverse Multi-round Dialogue Data."
 
@@ -16,61 +20,41 @@ _CITATION = """\
 }
 """
 
-_HOMEPAGE = "https://huggingface.co/datasets/stingning/ultrachat"
+_HOMEPAGE = "{}/datasets/stingning/ultrachat".format(_HF_ENDPOINT)
 _LICENSE = "cc-by-nc-4.0"
-_BASE_DATA_URL = "https://huggingface.co/datasets/stingning/ultrachat/resolve/main/train_{idx}.jsonl"
+_BASE_DATA_URL = "{}/datasets/stingning/ultrachat/resolve/main/train_{{idx}}.jsonl".format(_HF_ENDPOINT)
 
 
-class BelleMultiturn(datasets.GeneratorBasedBuilder):
-
+class UltraChat(datasets.GeneratorBasedBuilder):
     VERSION = datasets.Version("0.0.0")
 
-    def _info(self) -> datasets.DatasetInfo:
-        features = datasets.Features({
-            "instruction": datasets.Value("string"),
-            "output": datasets.Value("string"),
-            "history": datasets.Sequence(datasets.Sequence(datasets.Value("string")))
-        })
+    def _info(self):
+        features = datasets.Features(
+            {"conversations": [{"from": datasets.Value("string"), "value": datasets.Value("string")}]}
+        )
         return datasets.DatasetInfo(
-            description=_DESCRIPTION,
-            features=features,
-            homepage=_HOMEPAGE,
-            license=_LICENSE,
-            citation=_CITATION
+            description=_DESCRIPTION, features=features, homepage=_HOMEPAGE, license=_LICENSE, citation=_CITATION
         )
 
-    def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
-        file_paths = [dl_manager.download(_BASE_DATA_URL.format(idx=idx)) for idx in range(9)] # multiple shards
-        return [
-            datasets.SplitGenerator(
-                name=datasets.Split.TRAIN,
-                gen_kwargs={
-                    "filepaths": file_paths
-                }
-            )
-        ]
+    def _split_generators(self, dl_manager: datasets.DownloadManager):
+        file_paths = [dl_manager.download(_BASE_DATA_URL.format(idx=idx)) for idx in range(10)]  # multiple shards
+        return [datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepaths": file_paths})]
 
-    def _generate_examples(self, filepaths: List[str]) -> Dict[int, Dict[str, Any]]: # generate multi-turn chat for ChatGLM
+    def _generate_examples(self, filepaths: List[str]):
         for filepath in filepaths:
             with open(filepath, "r", encoding="utf-8") as f:
                 for row in f:
                     try:
                         data = json.loads(row)
-                    except:
+                    except Exception:
                         continue
-                    key = data["id"]
-                    content = data["data"]
+                    key: int = data["id"]
+                    content: List[str] = data["data"]
                     if len(content) % 2 == 1:
                         content.pop(-1)
                     if len(content) < 2:
                         continue
-
-                    query = content[-2]
-                    response = content[-1]
-                    history = [[content[2*i], content[2*i+1]] for i in range(len(content) // 2 - 1)]
-
-                    yield key, {
-                        "instruction": query,
-                        "output": response,
-                        "history": history
-                    }
+                    conversations = [
+                        {"from": "human" if i % 2 == 0 else "gpt", "value": content[i]} for i in range(len(content))
+                    ]
+                    yield key, {"conversations": conversations}

data/wiki_demo.txt

@@ -1,50 +0,0 @@
-Machine learning (ML) is a field devoted to understanding and building methods that let machines "learn" – that is, methods that leverage data to improve computer performance on some set of tasks.
-Machine learning algorithms build a model based on sample data, known as training data, in order to make predictions or decisions without being explicitly programmed to do so. Machine learning algorithms are used in a wide variety of applications, such as in medicine, email filtering, speech recognition, agriculture, and computer vision, where it is difficult or unfeasible to develop conventional algorithms to perform the needed tasks.
-A subset of machine learning is closely related to computational statistics, which focuses on making predictions using computers, but not all machine learning is statistical learning. The study of mathematical optimization delivers methods, theory and application domains to the field of machine learning. Data mining is a related field of study, focusing on exploratory data analysis through unsupervised learning.
-Some implementations of machine learning use data and neural networks in a way that mimics the working of a biological brain.
-In its application across business problems, machine learning is also referred to as predictive analytics.
-Learning algorithms work on the basis that strategies, algorithms, and inferences that worked well in the past are likely to continue working well in the future. These inferences can sometimes be obvious, such as "since the sun rose every morning for the last 10,000 days, it will probably rise tomorrow morning as well". Other times, they can be more nuanced, such as "X% of families have geographically separate species with color variants, so there is a Y% chance that undiscovered black swans exist".
-Machine learning programs can perform tasks without being explicitly programmed to do so. It involves computers learning from data provided so that they carry out certain tasks. For simple tasks assigned to computers, it is possible to program algorithms telling the machine how to execute all steps required to solve the problem at hand; on the computer's part, no learning is needed. For more advanced tasks, it can be challenging for a human to manually create the needed algorithms. In practice, it can turn out to be more effective to help the machine develop its own algorithm, rather than having human programmers specify every needed step.
-The discipline of machine learning employs various approaches to teach computers to accomplish tasks where no fully satisfactory algorithm is available. In cases where vast numbers of potential answers exist, one approach is to label some of the correct answers as valid. This can then be used as training data for the computer to improve the algorithm(s) it uses to determine correct answers. For example, to train a system for the task of digital character recognition, the MNIST dataset of handwritten digits has often been used.
-The term machine learning was coined in 1959 by Arthur Samuel, an IBM employee and pioneer in the field of computer gaming and artificial intelligence. The synonym self-teaching computers was also used in this time period.
-By the early 1960s an experimental "learning machine" with punched tape memory, called Cybertron, had been developed by Raytheon Company to analyze sonar signals, electrocardiograms, and speech patterns using rudimentary reinforcement learning. It was repetitively "trained" by a human operator/teacher to recognize patterns and equipped with a "goof" button to cause it to re-evaluate incorrect decisions. A representative book on research into machine learning during the 1960s was Nilsson's book on Learning Machines, dealing mostly with machine learning for pattern classification. Interest related to pattern recognition continued into the 1970s, as described by Duda and Hart in 1973. In 1981 a report was given on using teaching strategies so that a neural network learns to recognize 40 characters (26 letters, 10 digits, and 4 special symbols) from a computer terminal.
-Tom M. Mitchell provided a widely quoted, more formal definition of the algorithms studied in the machine learning field: "A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E." This definition of the tasks in which machine learning is concerned offers a fundamentally operational definition rather than defining the field in cognitive terms. This follows Alan Turing's proposal in his paper "Computing Machinery and Intelligence", in which the question "Can machines think?" is replaced with the question "Can machines do what we (as thinking entities) can do?".
-Modern-day machine learning has two objectives, one is to classify data based on models which have been developed, the other purpose is to make predictions for future outcomes based on these models. A hypothetical algorithm specific to classifying data may use computer vision of moles coupled with supervised learning in order to train it to classify the cancerous moles. A machine learning algorithm for stock trading may inform the trader of future potential predictions.
-As a scientific endeavor, machine learning grew out of the quest for artificial intelligence (AI). In the early days of AI as an academic discipline, some researchers were interested in having machines learn from data. They attempted to approach the problem with various symbolic methods, as well as what were then termed "neural networks"; these were mostly perceptrons and other models that were later found to be reinventions of the generalized linear models of statistics. Probabilistic reasoning was also employed, especially in automated medical diagnosis.: 488 
-However, an increasing emphasis on the logical, knowledge-based approach caused a rift between AI and machine learning. Probabilistic systems were plagued by theoretical and practical problems of data acquisition and representation.: 488  By 1980, expert systems had come to dominate AI, and statistics was out of favor. Work on symbolic/knowledge-based learning did continue within AI, leading to inductive logic programming, but the more statistical line of research was now outside the field of AI proper, in pattern recognition and information retrieval.: 708–710, 755  Neural networks research had been abandoned by AI and computer science around the same time. This line, too, was continued outside the AI/CS field, as "connectionism", by researchers from other disciplines including Hopfield, Rumelhart, and Hinton. Their main success came in the mid-1980s with the reinvention of backpropagation.: 25 
-Machine learning (ML), reorganized and recognized as its own field, started to flourish in the 1990s. The field changed its goal from achieving artificial intelligence to tackling solvable problems of a practical nature. It shifted focus away from the symbolic approaches it had inherited from AI, and toward methods and models borrowed from statistics, fuzzy logic, and probability theory.
-Machine learning and data mining often employ the same methods and overlap significantly, but while machine learning focuses on prediction, based on known properties learned from the training data, data mining focuses on the discovery of (previously) unknown properties in the data (this is the analysis step of knowledge discovery in databases). Data mining uses many machine learning methods, but with different goals; on the other hand, machine learning also employs data mining methods as "unsupervised learning" or as a preprocessing step to improve learner accuracy. Much of the confusion between these two research communities (which do often have separate conferences and separate journals, ECML PKDD being a major exception) comes from the basic assumptions they work with: in machine learning, performance is usually evaluated with respect to the ability to reproduce known knowledge, while in knowledge discovery and data mining (KDD) the key task is the discovery of previously unknown knowledge. Evaluated with respect to known knowledge, an uninformed (unsupervised) method will easily be outperformed by other supervised methods, while in a typical KDD task, supervised methods cannot be used due to the unavailability of training data.
-Machine learning also has intimate ties to optimization: many learning problems are formulated as minimization of some loss function on a training set of examples. Loss functions express the discrepancy between the predictions of the model being trained and the actual problem instances (for example, in classification, one wants to assign a label to instances, and models are trained to correctly predict the pre-assigned labels of a set of examples).
-The difference between optimization and machine learning arises from the goal of generalization: while optimization algorithms can minimize the loss on a training set, machine learning is concerned with minimizing the loss on unseen samples. Characterizing the generalization of various learning algorithms is an active topic of current research, especially for deep learning algorithms.
-Machine learning and statistics are closely related fields in terms of methods, but distinct in their principal goal: statistics draws population inferences from a sample, while machine learning finds generalizable predictive patterns. According to Michael I. Jordan, the ideas of machine learning, from methodological principles to theoretical tools, have had a long pre-history in statistics. He also suggested the term data science as a placeholder to call the overall field.
-Leo Breiman distinguished two statistical modeling paradigms: data model and algorithmic model, wherein "algorithmic model" means more or less the machine learning algorithms like Random Forest.
-Some statisticians have adopted methods from machine learning, leading to a combined field that they call statistical learning.
-Analytical and computational techniques derived from deep-rooted physics of disordered systems can be extended to large-scale problems, including machine learning, e.g., to analyze the weight space of deep neural networks. Statistical physics is thus finding applications in the area of medical diagnostics.
-A core objective of a learner is to generalize from its experience. Generalization in this context is the ability of a learning machine to perform accurately on new, unseen examples/tasks after having experienced a learning data set. The training examples come from some generally unknown probability distribution (considered representative of the space of occurrences) and the learner has to build a general model about this space that enables it to produce sufficiently accurate predictions in new cases.
-The computational analysis of machine learning algorithms and their performance is a branch of theoretical computer science known as computational learning theory via the Probably Approximately Correct Learning (PAC) model. Because training sets are finite and the future is uncertain, learning theory usually does not yield guarantees of the performance of algorithms. Instead, probabilistic bounds on the performance are quite common. The bias–variance decomposition is one way to quantify generalization error.
-For the best performance in the context of generalization, the complexity of the hypothesis should match the complexity of the function underlying the data. If the hypothesis is less complex than the function, then the model has under fitted the data. If the complexity of the model is increased in response, then the training error decreases. But if the hypothesis is too complex, then the model is subject to overfitting and generalization will be poorer.
-In addition to performance bounds, learning theorists study the time complexity and feasibility of learning. In computational learning theory, a computation is considered feasible if it can be done in polynomial time. There are two kinds of time complexity results: Positive results show that a certain class of functions can be learned in polynomial time. Negative results show that certain classes cannot be learned in polynomial time.
-Machine learning approaches are traditionally divided into three broad categories, which correspond to learning paradigms, depending on the nature of the "signal" or "feedback" available to the learning system:
-Supervised learning: The computer is presented with example inputs and their desired outputs, given by a "teacher", and the goal is to learn a general rule that maps inputs to outputs.
-Unsupervised learning: No labels are given to the learning algorithm, leaving it on its own to find structure in its input. Unsupervised learning can be a goal in itself (discovering hidden patterns in data) or a means towards an end (feature learning).
-Reinforcement learning: A computer program interacts with a dynamic environment in which it must perform a certain goal (such as driving a vehicle or playing a game against an opponent). As it navigates its problem space, the program is provided feedback that's analogous to rewards, which it tries to maximize. Although each algorithm has advantages and limitations, no single algorithm works for all problems.
-Supervised learning algorithms build a mathematical model of a set of data that contains both the inputs and the desired outputs. The data is known as training data, and consists of a set of training examples. Each training example has one or more inputs and the desired output, also known as a supervisory signal. In the mathematical model, each training example is represented by an array or vector, sometimes called a feature vector, and the training data is represented by a matrix. Through iterative optimization of an objective function, supervised learning algorithms learn a function that can be used to predict the output associated with new inputs. An optimal function will allow the algorithm to correctly determine the output for inputs that were not a part of the training data. An algorithm that improves the accuracy of its outputs or predictions over time is said to have learned to perform that task.
-Types of supervised-learning algorithms include active learning, classification and regression. Classification algorithms are used when the outputs are restricted to a limited set of values, and regression algorithms are used when the outputs may have any numerical value within a range. As an example, for a classification algorithm that filters emails, the input would be an incoming email, and the output would be the name of the folder in which to file the email.
-Similarity learning is an area of supervised machine learning closely related to regression and classification, but the goal is to learn from examples using a similarity function that measures how similar or related two objects are. It has applications in ranking, recommendation systems, visual identity tracking, face verification, and speaker verification.
-Unsupervised learning algorithms take a set of data that contains only inputs, and find structure in the data, like grouping or clustering of data points. The algorithms, therefore, learn from test data that has not been labeled, classified or categorized. Instead of responding to feedback, unsupervised learning algorithms identify commonalities in the data and react based on the presence or absence of such commonalities in each new piece of data. A central application of unsupervised learning is in the field of density estimation in statistics, such as finding the probability density function. Though unsupervised learning encompasses other domains involving summarizing and explaining data features. Unsupervised learning algorithms streamlined the process of survey and graph large indel based haplotypes of a gene of interest from pan-genome. 
-Cluster analysis is the assignment of a set of observations into subsets (called clusters) so that observations within the same cluster are similar according to one or more predesignated criteria, while observations drawn from different clusters are dissimilar. Different clustering techniques make different assumptions on the structure of the data, often defined by some similarity metric and evaluated, for example, by internal compactness, or the similarity between members of the same cluster, and separation, the difference between clusters. Other methods are based on estimated density and graph connectivity.
-Semi-supervised learning falls between unsupervised learning (without any labeled training data) and supervised learning (with completely labeled training data). Some of the training examples are missing training labels, yet many machine-learning researchers have found that unlabeled data, when used in conjunction with a small amount of labeled data, can produce a considerable improvement in learning accuracy.
-In weakly supervised learning, the training labels are noisy, limited, or imprecise; however, these labels are often cheaper to obtain, resulting in larger effective training sets.
-Reinforcement learning is an area of machine learning concerned with how software agents ought to take actions in an environment so as to maximize some notion of cumulative reward. Due to its generality, the field is studied in many other disciplines, such as game theory, control theory, operations research, information theory, simulation-based optimization, multi-agent systems, swarm intelligence, statistics and genetic algorithms. In machine learning, the environment is typically represented as a Markov decision process (MDP). Many reinforcements learning algorithms use dynamic programming techniques. Reinforcement learning algorithms do not assume knowledge of an exact mathematical model of the MDP and are used when exact models are infeasible. Reinforcement learning algorithms are used in autonomous vehicles or in learning to play a game against a human opponent.
-Dimensionality reduction is a process of reducing the number of random variables under consideration by obtaining a set of principal variables. In other words, it is a process of reducing the dimension of the feature set, also called the "number of features". Most of the dimensionality reduction techniques can be considered as either feature elimination or extraction. One of the popular methods of dimensionality reduction is principal component analysis (PCA). PCA involves changing higher-dimensional data (e.g., 3D) to a smaller space (e.g., 2D). This results in a smaller dimension of data (2D instead of 3D), while keeping all original variables in the model without changing the data. The manifold hypothesis proposes that high-dimensional data sets lie along low-dimensional manifolds, and many dimensionality reduction techniques make this assumption, leading to the area of manifold learning and manifold regularization.
-Although machine learning has been transformative in some fields, machine-learning programs often fail to deliver expected results. Reasons for this are numerous: lack of (suitable) data, lack of access to the data, data bias, privacy problems, badly chosen tasks and algorithms, wrong tools and people, lack of resources, and evaluation problems.
-In 2018, a self-driving car from Uber failed to detect a pedestrian, who was killed after a collision. Attempts to use machine learning in healthcare with the IBM Watson system failed to deliver even after years of time and billions of dollars invested.
-Machine learning has been used as a strategy to update the evidence related to a systematic review and increased reviewer burden related to the growth of biomedical literature. While it has improved with training sets, it has not yet developed sufficiently to reduce the workload burden without limiting the necessary sensitivity for the findings research themselves.
-Machine learning approaches in particular can suffer from different data biases. A machine learning system trained specifically on current customers may not be able to predict the needs of new customer groups that are not represented in the training data. When trained on human-made data, machine learning is likely to pick up the constitutional and unconscious biases already present in society. Language models learned from data have been shown to contain human-like biases. Machine learning systems used for criminal risk assessment have been found to be biased against black people. In 2015, Google photos would often tag black people as gorillas, and in 2018 this still was not well resolved, but Google reportedly was still using the workaround to remove all gorillas from the training data, and thus was not able to recognize real gorillas at all. Similar issues with recognizing non-white people have been found in many other systems. In 2016, Microsoft tested a chatbot that learned from Twitter, and it quickly picked up racist and sexist language. Because of such challenges, the effective use of machine learning may take longer to be adopted in other domains. Concern for fairness in machine learning, that is, reducing bias in machine learning and propelling its use for human good is increasingly expressed by artificial intelligence scientists, including Fei-Fei Li, who reminds engineers that "There's nothing artificial about AI...It's inspired by people, it's created by people, and—most importantly—it impacts people. It is a powerful tool we are only just beginning to understand, and that is a profound responsibility."
-Learners can also disappoint by "learning the wrong lesson". A toy example is that an image classifier trained only on pictures of brown horses and black cats might conclude that all brown patches are likely to be horses. A real-world example is that, unlike humans, current image classifiers often do not primarily make judgments from the spatial relationship between components of the picture, and they learn relationships between pixels that humans are oblivious to, but that still correlate with images of certain types of real objects. Modifying these patterns on a legitimate image can result in "adversarial" images that the system misclassifies.
-Adversarial vulnerabilities can also result in nonlinear systems, or from non-pattern perturbations. Some systems are so brittle that changing a single adversarial pixel predictably induces misclassification.[citation needed] Machine learning models are often vulnerable to manipulation and/or evasion via adversarial machine learning.
-Researchers have demonstrated how backdoors can be placed undetectably into classifying (e.g., for categories "spam" and well-visible "not spam" of posts) machine learning models which are often developed and/or trained by third parties. Parties can change the classification of any input, including in cases for which a type of data/software transparency is provided, possibly including white-box access.
-Machine learning poses a host of ethical questions. Systems that are trained on datasets collected with biases may exhibit these biases upon use (algorithmic bias), thus digitizing cultural prejudices. For example, in 1988, the UK's Commission for Racial Equality found that St. George's Medical School had been using a computer program trained from data of previous admissions staff and this program had denied nearly 60 candidates who were found to be either women or had non-European sounding names. Using job hiring data from a firm with racist hiring policies may lead to a machine learning system duplicating the bias by scoring job applicants by similarity to previous successful applicants. Responsible collection of data and documentation of algorithmic rules used by a system thus is a critical part of machine learning.
-AI can be well-equipped to make decisions in technical fields, which rely heavily on data and historical information. These decisions rely on the objectivity and logical reasoning. Because human languages contain biases, machines trained on language corpora will necessarily also learn these biases.
-Other forms of ethical challenges, not related to personal biases, are seen in health care. There are concerns among health care professionals that these systems might not be designed in the public's interest but as income-generating machines. This is especially true in the United States where there is a long-standing ethical dilemma of improving health care, but also increase profits. For example, the algorithms could be designed to provide patients with unnecessary tests or medication in which the algorithm's proprietary owners hold stakes. There is potential for machine learning in health care to provide professionals an additional tool to diagnose, medicate, and plan recovery paths for patients, but this requires these biases to be mitigated.
-Since the 2010s, advances in both machine learning algorithms and computer hardware have led to more efficient methods for training deep neural networks (a particular narrow subdomain of machine learning) that contain many layers of non-linear hidden units. By 2019, graphic processing units (GPUs), often with AI-specific enhancements, had displaced CPUs as the dominant method of training large-scale commercial cloud AI. OpenAI estimated the hardware computing used in the largest deep learning projects from AlexNet (2012) to AlphaZero (2017), and found a 300,000-fold increase in the amount of compute required, with a doubling-time trendline of 3.4 months.

data/wiki_demo.txt.REMOVED.git-id

@@ -0,0 +1 @@
+c9cf509b7fdac5490cfd6dae72c2d7b8a60af6cb

docker-compose.yml

@@ -0,0 +1,25 @@
+version: '3.8'
+
+services:
+  llama-factory:
+    build:
+      dockerfile: Dockerfile
+      context: .
+    container_name: llama_factory
+    volumes:
+      - ./hf_cache:/root/.cache/huggingface/
+      - ./data:/app/data
+      - ./output:/app/output
+    environment:
+      - CUDA_VISIBLE_DEVICES=0
+    ports:
+      - "7860:7860"
+    ipc: host
+    deploy:
+      resources:
+        reservations:
+          devices:
+          - driver: nvidia
+            count: "all"
+            capabilities: [gpu]
+    restart: unless-stopped

evaluation/ceval/ceval.py

@@ -0,0 +1,160 @@
+# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+import os
+
+import datasets
+import pandas as pd
+
+
+_CITATION = """\
+@article{huang2023ceval,
+  title={C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for Foundation Models},
+  author={Huang, Yuzhen and Bai, Yuzhuo and Zhu, Zhihao and Zhang, Junlei and Zhang, Jinghan and Su, Tangjun and Liu, Junteng and Lv, Chuancheng and Zhang, Yikai and Lei, Jiayi and Fu, Yao and Sun, Maosong and He, Junxian},
+  journal={arXiv preprint arXiv:2305.08322},
+  year={2023}
+}
+"""
+
+_DESCRIPTION = """\
+C-Eval is a comprehensive Chinese evaluation suite for foundation models. It consists of 13948 multi-choice questions spanning 52 diverse disciplines and four difficulty levels.
+"""
+
+_HOMEPAGE = "https://cevalbenchmark.com"
+
+_LICENSE = "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License"
+
+_URL = "ceval.zip"
+
+task_list = [
+    "computer_network",
+    "operating_system",
+    "computer_architecture",
+    "college_programming",
+    "college_physics",
+    "college_chemistry",
+    "advanced_mathematics",
+    "probability_and_statistics",
+    "discrete_mathematics",
+    "electrical_engineer",
+    "metrology_engineer",
+    "high_school_mathematics",
+    "high_school_physics",
+    "high_school_chemistry",
+    "high_school_biology",
+    "middle_school_mathematics",
+    "middle_school_biology",
+    "middle_school_physics",
+    "middle_school_chemistry",
+    "veterinary_medicine",
+    "college_economics",
+    "business_administration",
+    "marxism",
+    "mao_zedong_thought",
+    "education_science",
+    "teacher_qualification",
+    "high_school_politics",
+    "high_school_geography",
+    "middle_school_politics",
+    "middle_school_geography",
+    "modern_chinese_history",
+    "ideological_and_moral_cultivation",
+    "logic",
+    "law",
+    "chinese_language_and_literature",
+    "art_studies",
+    "professional_tour_guide",
+    "legal_professional",
+    "high_school_chinese",
+    "high_school_history",
+    "middle_school_history",
+    "civil_servant",
+    "sports_science",
+    "plant_protection",
+    "basic_medicine",
+    "clinical_medicine",
+    "urban_and_rural_planner",
+    "accountant",
+    "fire_engineer",
+    "environmental_impact_assessment_engineer",
+    "tax_accountant",
+    "physician",
+]
+
+
+class CevalConfig(datasets.BuilderConfig):
+    def __init__(self, **kwargs):
+        super().__init__(version=datasets.Version("1.0.0"), **kwargs)
+
+
+class Ceval(datasets.GeneratorBasedBuilder):
+    BUILDER_CONFIGS = [
+        CevalConfig(
+            name=task_name,
+        )
+        for task_name in task_list
+    ]
+
+    def _info(self):
+        features = datasets.Features(
+            {
+                "id": datasets.Value("int32"),
+                "question": datasets.Value("string"),
+                "A": datasets.Value("string"),
+                "B": datasets.Value("string"),
+                "C": datasets.Value("string"),
+                "D": datasets.Value("string"),
+                "answer": datasets.Value("string"),
+                "explanation": datasets.Value("string"),
+            }
+        )
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            homepage=_HOMEPAGE,
+            license=_LICENSE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        data_dir = dl_manager.download_and_extract(_URL)
+        task_name = self.config.name
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, "test", f"{task_name}_test.csv"),
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, "val", f"{task_name}_val.csv"),
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, "dev", f"{task_name}_dev.csv"),
+                },
+            ),
+        ]
+
+    def _generate_examples(self, filepath):
+        df = pd.read_csv(filepath, encoding="utf-8")
+        for i, instance in enumerate(df.to_dict(orient="records")):
+            if "answer" not in instance.keys():
+                instance["answer"] = ""
+            if "explanation" not in instance.keys():
+                instance["explanation"] = ""
+            yield i, instance

evaluation/cmmlu/cmmlu.py

@@ -0,0 +1,167 @@
+# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+import os
+
+import datasets
+import pandas as pd
+
+
+_CITATION = """\
+@article{li2023cmmlu,
+  title={CMMLU: Measuring massive multitask language understanding in Chinese},
+  author={Haonan Li and Yixuan Zhang and Fajri Koto and Yifei Yang and Hai Zhao and Yeyun Gong and Nan Duan and Timothy Baldwin},
+  journal={arXiv preprint arXiv:2306.09212},
+  year={2023}
+}
+"""
+
+_DESCRIPTION = """\
+CMMLU is a comprehensive Chinese assessment suite specifically designed to evaluate the advanced knowledge and reasoning abilities of LLMs within the Chinese language and cultural context.
+"""
+
+_HOMEPAGE = "https://github.com/haonan-li/CMMLU"
+
+_LICENSE = "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License"
+
+_URL = "cmmlu.zip"
+
+task_list = [
+    "agronomy",
+    "anatomy",
+    "ancient_chinese",
+    "arts",
+    "astronomy",
+    "business_ethics",
+    "chinese_civil_service_exam",
+    "chinese_driving_rule",
+    "chinese_food_culture",
+    "chinese_foreign_policy",
+    "chinese_history",
+    "chinese_literature",
+    "chinese_teacher_qualification",
+    "clinical_knowledge",
+    "college_actuarial_science",
+    "college_education",
+    "college_engineering_hydrology",
+    "college_law",
+    "college_mathematics",
+    "college_medical_statistics",
+    "college_medicine",
+    "computer_science",
+    "computer_security",
+    "conceptual_physics",
+    "construction_project_management",
+    "economics",
+    "education",
+    "electrical_engineering",
+    "elementary_chinese",
+    "elementary_commonsense",
+    "elementary_information_and_technology",
+    "elementary_mathematics",
+    "ethnology",
+    "food_science",
+    "genetics",
+    "global_facts",
+    "high_school_biology",
+    "high_school_chemistry",
+    "high_school_geography",
+    "high_school_mathematics",
+    "high_school_physics",
+    "high_school_politics",
+    "human_sexuality",
+    "international_law",
+    "journalism",
+    "jurisprudence",
+    "legal_and_moral_basis",
+    "logical",
+    "machine_learning",
+    "management",
+    "marketing",
+    "marxist_theory",
+    "modern_chinese",
+    "nutrition",
+    "philosophy",
+    "professional_accounting",
+    "professional_law",
+    "professional_medicine",
+    "professional_psychology",
+    "public_relations",
+    "security_study",
+    "sociology",
+    "sports_science",
+    "traditional_chinese_medicine",
+    "virology",
+    "world_history",
+    "world_religions",
+]
+
+
+class CMMLUConfig(datasets.BuilderConfig):
+    def __init__(self, **kwargs):
+        super().__init__(version=datasets.Version("1.0.1"), **kwargs)
+
+
+class CMMLU(datasets.GeneratorBasedBuilder):
+    BUILDER_CONFIGS = [
+        CMMLUConfig(
+            name=task_name,
+        )
+        for task_name in task_list
+    ]
+
+    def _info(self):
+        features = datasets.Features(
+            {
+                "question": datasets.Value("string"),
+                "A": datasets.Value("string"),
+                "B": datasets.Value("string"),
+                "C": datasets.Value("string"),
+                "D": datasets.Value("string"),
+                "answer": datasets.Value("string"),
+            }
+        )
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            homepage=_HOMEPAGE,
+            license=_LICENSE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        data_dir = dl_manager.download_and_extract(_URL)
+        task_name = self.config.name
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, f"test/{task_name}.csv"),
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, f"dev/{task_name}.csv"),
+                },
+            ),
+        ]
+
+    def _generate_examples(self, filepath):
+        df = pd.read_csv(filepath, header=0, index_col=0, encoding="utf-8")
+        for i, instance in enumerate(df.to_dict(orient="records")):
+            question = instance.pop("Question", "")
+            answer = instance.pop("Answer", "")
+            instance["question"] = question
+            instance["answer"] = answer
+            yield i, instance

evaluation/mmlu/mmlu.py

@@ -0,0 +1,161 @@
+# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+import os
+
+import datasets
+import pandas as pd
+
+
+_CITATION = """\
+@article{hendryckstest2021,
+  title={Measuring Massive Multitask Language Understanding},
+  author={Dan Hendrycks and Collin Burns and Steven Basart and Andy Zou and Mantas Mazeika and Dawn Song and Jacob Steinhardt},
+  journal={Proceedings of the International Conference on Learning Representations (ICLR)},
+  year={2021}
+}
+"""
+
+_DESCRIPTION = """\
+Measuring Massive Multitask Language Understanding by Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt (ICLR 2021).
+"""
+
+_HOMEPAGE = "https://github.com/hendrycks/test"
+
+_LICENSE = "MIT"
+
+_URL = "mmlu.zip"
+
+task_list = [
+    "high_school_european_history",
+    "business_ethics",
+    "clinical_knowledge",
+    "medical_genetics",
+    "high_school_us_history",
+    "high_school_physics",
+    "high_school_world_history",
+    "virology",
+    "high_school_microeconomics",
+    "econometrics",
+    "college_computer_science",
+    "high_school_biology",
+    "abstract_algebra",
+    "professional_accounting",
+    "philosophy",
+    "professional_medicine",
+    "nutrition",
+    "global_facts",
+    "machine_learning",
+    "security_studies",
+    "public_relations",
+    "professional_psychology",
+    "prehistory",
+    "anatomy",
+    "human_sexuality",
+    "college_medicine",
+    "high_school_government_and_politics",
+    "college_chemistry",
+    "logical_fallacies",
+    "high_school_geography",
+    "elementary_mathematics",
+    "human_aging",
+    "college_mathematics",
+    "high_school_psychology",
+    "formal_logic",
+    "high_school_statistics",
+    "international_law",
+    "high_school_mathematics",
+    "high_school_computer_science",
+    "conceptual_physics",
+    "miscellaneous",
+    "high_school_chemistry",
+    "marketing",
+    "professional_law",
+    "management",
+    "college_physics",
+    "jurisprudence",
+    "world_religions",
+    "sociology",
+    "us_foreign_policy",
+    "high_school_macroeconomics",
+    "computer_security",
+    "moral_scenarios",
+    "moral_disputes",
+    "electrical_engineering",
+    "astronomy",
+    "college_biology",
+]
+
+
+class MMLUConfig(datasets.BuilderConfig):
+    def __init__(self, **kwargs):
+        super().__init__(version=datasets.Version("1.0.0"), **kwargs)
+
+
+class MMLU(datasets.GeneratorBasedBuilder):
+    BUILDER_CONFIGS = [
+        MMLUConfig(
+            name=task_name,
+        )
+        for task_name in task_list
+    ]
+
+    def _info(self):
+        features = datasets.Features(
+            {
+                "question": datasets.Value("string"),
+                "A": datasets.Value("string"),
+                "B": datasets.Value("string"),
+                "C": datasets.Value("string"),
+                "D": datasets.Value("string"),
+                "answer": datasets.Value("string"),
+            }
+        )
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            homepage=_HOMEPAGE,
+            license=_LICENSE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        data_dir = dl_manager.download_and_extract(_URL)
+        task_name = self.config.name
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, "data", "test", f"{task_name}_test.csv"),
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, "data", "val", f"{task_name}_val.csv"),
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, "data", "dev", f"{task_name}_dev.csv"),
+                },
+            ),
+        ]
+
+    def _generate_examples(self, filepath):
+        df = pd.read_csv(filepath)
+        df.columns = ["question", "A", "B", "C", "D", "answer"]
+
+        for i, instance in enumerate(df.to_dict(orient="records")):
+            yield i, instance

examples/README.md

@@ -0,0 +1,229 @@
+We provide diverse examples about fine-tuning LLMs.
+
+Make sure to execute these commands in the `LLaMA-Factory` directory.
+
+## Table of Contents
+
+- [LoRA Fine-Tuning on A Single GPU](#lora-fine-tuning-on-a-single-gpu)
+- [QLoRA Fine-Tuning on a Single GPU](#qlora-fine-tuning-on-a-single-gpu)
+- [LoRA Fine-Tuning on Multiple GPUs](#lora-fine-tuning-on-multiple-gpus)
+- [LoRA Fine-Tuning on Multiple NPUs](#lora-fine-tuning-on-multiple-npus)
+- [Full-Parameter Fine-Tuning on Multiple GPUs](#full-parameter-fine-tuning-on-multiple-gpus)
+- [Merging LoRA Adapters and Quantization](#merging-lora-adapters-and-quantization)
+- [Inferring LoRA Fine-Tuned Models](#inferring-lora-fine-tuned-models)
+- [Extras](#extras)
+
+## Examples
+
+### LoRA Fine-Tuning on A Single GPU
+
+#### (Continuous) Pre-Training
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_pretrain.yaml
+```
+
+#### Supervised Fine-Tuning
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_sft.yaml
+```
+
+#### Multimodal Supervised Fine-Tuning
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llava1_5_lora_sft.yaml
+```
+
+#### Reward Modeling
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_reward.yaml
+```
+
+#### PPO Training
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_ppo.yaml
+```
+
+#### DPO Training
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_dpo.yaml
+```
+
+#### ORPO Training
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_orpo.yaml
+```
+
+#### Preprocess Dataset
+
+It is useful for large dataset, use `tokenized_path` in config to load the preprocessed dataset.
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_preprocess.yaml
+```
+
+#### Evaluating on MMLU/CMMLU/C-Eval Benchmarks
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli eval examples/lora_single_gpu/llama3_lora_eval.yaml
+```
+
+#### Batch Predicting and Computing BLEU and ROUGE Scores
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_predict.yaml
+```
+
+### QLoRA Fine-Tuning on a Single GPU
+
+#### Supervised Fine-Tuning with 4/8-bit Bitsandbytes Quantization (Recommended)
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/qlora_single_gpu/llama3_lora_sft_bitsandbytes.yaml
+```
+
+#### Supervised Fine-Tuning with 4/8-bit GPTQ Quantization
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/qlora_single_gpu/llama3_lora_sft_gptq.yaml
+```
+
+#### Supervised Fine-Tuning with 4-bit AWQ Quantization
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/qlora_single_gpu/llama3_lora_sft_awq.yaml
+```
+
+#### Supervised Fine-Tuning with 2-bit AQLM Quantization
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/qlora_single_gpu/llama3_lora_sft_aqlm.yaml
+```
+
+### LoRA Fine-Tuning on Multiple GPUs
+
+#### Supervised Fine-Tuning with Accelerate on Single Node
+
+```bash
+bash examples/lora_multi_gpu/single_node.sh
+```
+
+#### Supervised Fine-Tuning with Accelerate on Multiple Nodes
+
+```bash
+bash examples/lora_multi_gpu/multi_node.sh
+```
+
+#### Supervised Fine-Tuning with DeepSpeed ZeRO-3 (Weight Sharding)
+
+```bash
+bash examples/lora_multi_gpu/ds_zero3.sh
+```
+
+### LoRA Fine-Tuning on Multiple NPUs
+
+#### Supervised Fine-Tuning with DeepSpeed ZeRO-0
+
+```bash
+bash examples/lora_multi_npu/ds_zero0.sh
+```
+
+### Full-Parameter Fine-Tuning on Multiple GPUs
+
+#### Supervised Fine-Tuning with Accelerate on Single Node
+
+```bash
+bash examples/full_multi_gpu/single_node.sh
+```
+
+#### Supervised Fine-Tuning with Accelerate on Multiple Nodes
+
+```bash
+bash examples/full_multi_gpu/multi_node.sh
+```
+
+#### Batch Predicting and Computing BLEU and ROUGE Scores
+
+```bash
+bash examples/full_multi_gpu/predict.sh
+```
+
+### Merging LoRA Adapters and Quantization
+
+#### Merge LoRA Adapters
+
+Note: DO NOT use quantized model or `quantization_bit` when merging LoRA adapters.
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli export examples/merge_lora/llama3_lora_sft.yaml
+```
+
+#### Quantizing Model using AutoGPTQ
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli export examples/merge_lora/llama3_gptq.yaml
+```
+
+### Inferring LoRA Fine-Tuned Models
+
+#### Use CLI
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli chat examples/merge_lora/llama3_lora_sft.yaml
+```
+
+#### Use Web UI
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli webchat examples/merge_lora/llama3_lora_sft.yaml
+```
+
+#### Launch OpenAI-style API
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli api examples/merge_lora/llama3_lora_sft.yaml
+```
+
+### Extras
+
+#### Full-Parameter Fine-Tuning using GaLore
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/extras/galore/llama3_full_sft.yaml
+```
+
+#### Full-Parameter Fine-Tuning using BAdam
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/extras/badam/llama3_full_sft.yaml
+```
+
+#### LoRA+ Fine-Tuning
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/extras/loraplus/llama3_lora_sft.yaml
+```
+
+#### Mixture-of-Depths Fine-Tuning
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/extras/mod/llama3_full_sft.yaml
+```
+
+#### LLaMA-Pro Fine-Tuning
+
+```bash
+bash examples/extras/llama_pro/expand.sh
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/extras/llama_pro/llama3_freeze_sft.yaml
+```
+
+#### FSDP+QLoRA Fine-Tuning
+
+```bash
+bash examples/extras/fsdp_qlora/single_node.sh
+```

examples/README_zh.md

@@ -0,0 +1,229 @@
+我们提供了多样化的大模型微调示例脚本。
+
+请确保在 `LLaMA-Factory` 目录下执行下述命令。
+
+## 目录
+
+- [单 GPU LoRA 微调](#单-gpu-lora-微调)
+- [单 GPU QLoRA 微调](#单-gpu-qlora-微调)
+- [多 GPU LoRA 微调](#多-gpu-lora-微调)
+- [多 NPU LoRA 微调](#多-npu-lora-微调)
+- [多 GPU 全参数微调](#多-gpu-全参数微调)
+- [合并 LoRA 适配器与模型量化](#合并-lora-适配器与模型量化)
+- [推理 LoRA 模型](#推理-lora-模型)
+- [杂项](#杂项)
+
+## 示例
+
+### 单 GPU LoRA 微调
+
+#### （增量）预训练
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_pretrain.yaml
+```
+
+#### 指令监督微调
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_sft.yaml
+```
+
+#### 多模态指令监督微调
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llava1_5_lora_sft.yaml
+```
+
+#### 奖励模型训练
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_reward.yaml
+```
+
+#### PPO 训练
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_ppo.yaml
+```
+
+#### DPO 训练
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_dpo.yaml
+```
+
+#### ORPO 训练
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_orpo.yaml
+```
+
+#### 预处理数据集
+
+对于大数据集有帮助，在配置中使用 `tokenized_path` 以加载预处理后的数据集。
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_preprocess.yaml
+```
+
+#### 在 MMLU/CMMLU/C-Eval 上评估
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli eval examples/lora_single_gpu/llama3_lora_eval.yaml
+```
+
+#### 批量预测并计算 BLEU 和 ROUGE 分数
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/lora_single_gpu/llama3_lora_predict.yaml
+```
+
+### 单 GPU QLoRA 微调
+
+#### 基于 4/8 比特 Bitsandbytes 量化进行指令监督微调（推荐）
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/qlora_single_gpu/llama3_lora_sft_bitsandbytes.yaml
+```
+
+#### 基于 4/8 比特 GPTQ 量化进行指令监督微调
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/qlora_single_gpu/llama3_lora_sft_gptq.yaml
+```
+
+#### 基于 4 比特 AWQ 量化进行指令监督微调
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/qlora_single_gpu/llama3_lora_sft_awq.yaml
+```
+
+#### 基于 2 比特 AQLM 量化进行指令监督微调
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/qlora_single_gpu/llama3_lora_sft_aqlm.yaml
+```
+
+### 多 GPU LoRA 微调
+
+#### 使用 Accelerate 进行单节点训练
+
+```bash
+bash examples/lora_multi_gpu/single_node.sh
+```
+
+#### 使用 Accelerate 进行多节点训练
+
+```bash
+bash examples/lora_multi_gpu/multi_node.sh
+```
+
+#### 使用 DeepSpeed ZeRO-3 平均分配显存
+
+```bash
+bash examples/lora_multi_gpu/ds_zero3.sh
+```
+
+### 多 NPU LoRA 微调
+
+#### 使用 DeepSpeed ZeRO-0 训练
+
+```bash
+bash examples/lora_multi_npu/ds_zero0.sh
+```
+
+### 多 GPU 全参数微调
+
+#### 使用 DeepSpeed 进行单节点训练
+
+```bash
+bash examples/full_multi_gpu/single_node.sh
+```
+
+#### 使用 DeepSpeed 进行多节点训练
+
+```bash
+bash examples/full_multi_gpu/multi_node.sh
+```
+
+#### 批量预测并计算 BLEU 和 ROUGE 分数
+
+```bash
+bash examples/full_multi_gpu/predict.sh
+```
+
+### 合并 LoRA 适配器与模型量化
+
+#### 合并 LoRA 适配器
+
+注：请勿使用量化后的模型或 `quantization_bit` 参数来合并 LoRA 适配器。
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli export examples/merge_lora/llama3_lora_sft.yaml
+```
+
+#### 使用 AutoGPTQ 量化模型
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli export examples/merge_lora/llama3_gptq.yaml
+```
+
+### 推理 LoRA 模型
+
+#### 使用命令行接口
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli chat examples/merge_lora/llama3_lora_sft.yaml
+```
+
+#### 使用浏览器界面
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli webchat examples/merge_lora/llama3_lora_sft.yaml
+```
+
+#### 启动 OpenAI 风格 API
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli api examples/merge_lora/llama3_lora_sft.yaml
+```
+
+### 杂项
+
+#### 使用 GaLore 进行全参数训练
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/extras/galore/llama3_full_sft.yaml
+```
+
+#### 使用 BAdam 进行全参数训练
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/extras/badam/llama3_full_sft.yaml
+```
+
+#### LoRA+ 微调
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/extras/loraplus/llama3_lora_sft.yaml
+```
+
+#### 深度混合微调
+
+```bash
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/extras/mod/llama3_full_sft.yaml
+```
+
+#### LLaMA-Pro 微调
+
+```bash
+bash examples/extras/llama_pro/expand.sh
+CUDA_VISIBLE_DEVICES=0 llamafactory-cli train examples/extras/llama_pro/llama3_freeze_sft.yaml
+```
+
+#### FSDP+QLoRA 微调
+
+```bash
+bash examples/extras/fsdp_qlora/single_node.sh
+```

examples/accelerate/fsdp_config.yaml

@@ -0,0 +1,25 @@
+compute_environment: LOCAL_MACHINE
+debug: false
+distributed_type: FSDP
+downcast_bf16: 'no'
+fsdp_config:
+  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
+  fsdp_backward_prefetch: BACKWARD_PRE
+  fsdp_cpu_ram_efficient_loading: true
+  fsdp_forward_prefetch: false
+  fsdp_offload_params: true
+  fsdp_sharding_strategy: FULL_SHARD
+  fsdp_state_dict_type: FULL_STATE_DICT
+  fsdp_sync_module_states: true
+  fsdp_use_orig_params: false
+machine_rank: 0
+main_training_function: main
+mixed_precision: fp16
+num_machines: 1 # the number of nodes
+num_processes: 2 # the number of GPUs in all nodes
+rdzv_backend: static
+same_network: true
+tpu_env: []
+tpu_use_cluster: false
+tpu_use_sudo: false
+use_cpu: false

examples/accelerate/master_config.yaml

@@ -0,0 +1,18 @@
+compute_environment: LOCAL_MACHINE
+debug: false
+distributed_type: MULTI_GPU
+downcast_bf16: 'no'
+gpu_ids: all
+machine_rank: 0
+main_process_ip: 192.168.0.1
+main_process_port: 29555
+main_training_function: main
+mixed_precision: fp16
+num_machines: 2 # the number of nodes
+num_processes: 8 # the number of GPUs in all nodes
+rdzv_backend: static
+same_network: true
+tpu_env: []
+tpu_use_cluster: false
+tpu_use_sudo: false
+use_cpu: false

examples/accelerate/single_config.yaml

@@ -0,0 +1,16 @@
+compute_environment: LOCAL_MACHINE
+debug: false
+distributed_type: MULTI_GPU
+downcast_bf16: 'no'
+gpu_ids: all
+machine_rank: 0
+main_training_function: main
+mixed_precision: fp16
+num_machines: 1 # the number of nodes
+num_processes: 4 # the number of GPUs in all nodes
+rdzv_backend: static
+same_network: true
+tpu_env: []
+tpu_use_cluster: false
+tpu_use_sudo: false
+use_cpu: false

examples/accelerate/slave_config.yaml

@@ -0,0 +1,18 @@
+compute_environment: LOCAL_MACHINE
+debug: false
+distributed_type: MULTI_GPU
+downcast_bf16: 'no'
+gpu_ids: all
+machine_rank: 1
+main_process_ip: 192.168.0.1
+main_process_port: 29555
+main_training_function: main
+mixed_precision: fp16
+num_machines: 2 # the number of nodes
+num_processes: 8 # the number of GPUs in all nodes
+rdzv_backend: static
+same_network: true
+tpu_env: []
+tpu_use_cluster: false
+tpu_use_sudo: false
+use_cpu: false

examples/extras/badam/llama3_lora_sft.yaml

@@ -0,0 +1,41 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: sft
+do_train: true
+finetuning_type: full
+use_badam: true
+badam_switch_mode: descending
+badam_switch_interval: 50
+badam_verbose: 2
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/full/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+pure_bf16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/extras/fsdp_qlora/llama3_lora_sft.yaml

@@ -0,0 +1,42 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+quantization_bit: 4
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# ddp
+ddp_timeout: 180000000
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/extras/fsdp_qlora/single_node.sh

@@ -0,0 +1,10 @@
+#!/bin/bash
+# DO NOT use GPTQ/AWQ model in FSDP+QLoRA
+
+pip install "transformers>=4.39.1"
+pip install "accelerate>=0.28.0"
+pip install "bitsandbytes>=0.43.0"
+
+CUDA_VISIBLE_DEVICES=0,1 accelerate launch \
+    --config_file examples/accelerate/fsdp_config.yaml \
+    src/train.py examples/extras/fsdp_qlora/llama3_lora_sft.yaml

examples/extras/galore/llama3_full_sft.yaml

@@ -0,0 +1,42 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: sft
+do_train: true
+finetuning_type: full
+use_galore: true
+galore_layerwise: true
+galore_target: mlp,self_attn
+galore_rank: 128
+galore_scale: 2.0
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/full/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 1
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+pure_bf16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/extras/llama_pro/expand.sh

@@ -0,0 +1,6 @@
+#!/bin/bash
+
+python scripts/llama_pro.py \
+    --model_name_or_path meta-llama/Meta-Llama-3-8B-Instruct \
+    --output_dir models/llama3-8b-instruct-pro \
+    --num_expand 8

examples/extras/llama_pro/llama3_freeze_sft.yaml

@@ -0,0 +1,40 @@
+# model
+model_name_or_path: models/llama3-8b-instruct-pro
+
+# method
+stage: sft
+do_train: true
+finetuning_type: freeze
+freeze_trainable_layers: 8
+freeze_trainable_modules: all
+use_llama_pro: true
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b-instruct-pro/freeze/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/extras/loraplus/llama3_lora_sft.yaml

@@ -0,0 +1,39 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+loraplus_lr_ratio: 16.0
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/extras/mod/llama3_full_sft.yaml

@@ -0,0 +1,39 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: sft
+do_train: true
+finetuning_type: full
+mixture_of_depths: convert
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b-mod/full/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+optim: paged_adamw_8bit
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+pure_bf16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/full_multi_gpu/llama3_full_predict.yaml

@@ -0,0 +1,23 @@
+# model
+model_name_or_path: saves/llama3-8b/full/sft
+
+# method
+stage: sft
+do_predict: true
+finetuning_type: full
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 50
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/full/predict
+overwrite_output_dir: true
+
+# eval
+per_device_eval_batch_size: 1
+predict_with_generate: true

examples/full_multi_gpu/llama3_full_sft.yaml

@@ -0,0 +1,41 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: sft
+do_train: true
+finetuning_type: full
+
+# ddp
+ddp_timeout: 180000000
+deepspeed: examples/deepspeed/ds_z3_config.json
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/full/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 2
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/full_multi_gpu/multi_node.sh

@@ -0,0 +1,15 @@
+#!/bin/bash
+
+NPROC_PER_NODE=4
+NNODES=2
+RANK=0
+MASTER_ADDR=192.168.0.1
+MASTER_PORT=29500
+
+CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun \
+    --nproc_per_node $NPROC_PER_NODE \
+    --nnodes $NNODES \
+    --node_rank $RANK \
+    --master_addr $MASTER_ADDR \
+    --master_port $MASTER_PORT \
+    src/train.py examples/full_multi_gpu/llama3_full_sft.yaml

examples/full_multi_gpu/predict.sh

@@ -0,0 +1,5 @@
+#!/bin/bash
+
+CUDA_VISIBLE_DEVICES=0,1,2,3 accelerate launch \
+    --config_file examples/accelerate/single_config.yaml \
+    src/train.py examples/full_multi_gpu/llama3_full_predict.yaml

examples/full_multi_gpu/single_node.sh

@@ -0,0 +1,15 @@
+#!/bin/bash
+
+NPROC_PER_NODE=4
+NNODES=1
+RANK=0
+MASTER_ADDR=127.0.0.1
+MASTER_PORT=29500
+
+CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun \
+    --nproc_per_node $NPROC_PER_NODE \
+    --nnodes $NNODES \
+    --node_rank $RANK \
+    --master_addr $MASTER_ADDR \
+    --master_port $MASTER_PORT \
+    src/train.py examples/full_multi_gpu/llama3_full_sft.yaml

examples/inference/llama3.yaml

@@ -0,0 +1,2 @@
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+template: llama3

examples/inference/llama3_lora_sft.yaml

@@ -0,0 +1,4 @@
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+adapter_name_or_path: saves/llama3-8b/lora/sft
+template: llama3
+finetuning_type: lora

examples/inference/llama3_vllm.yaml

@@ -0,0 +1,4 @@
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+template: llama3
+infer_backend: vllm
+vllm_enforce_eager: true

examples/lora_multi_gpu/ds_zero3.sh

@@ -0,0 +1,15 @@
+#!/bin/bash
+
+NPROC_PER_NODE=4
+NNODES=1
+RANK=0
+MASTER_ADDR=127.0.0.1
+MASTER_PORT=29500
+
+CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun \
+    --nproc_per_node $NPROC_PER_NODE \
+    --nnodes $NNODES \
+    --node_rank $RANK \
+    --master_addr $MASTER_ADDR \
+    --master_port $MASTER_PORT \
+    src/train.py examples/lora_multi_gpu/llama3_lora_sft_ds.yaml

examples/lora_multi_gpu/llama3_lora_sft.yaml

@@ -0,0 +1,41 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# ddp
+ddp_timeout: 180000000
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 2
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/lora_multi_gpu/llama3_lora_sft_ds.yaml

@@ -0,0 +1,42 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# ddp
+ddp_timeout: 180000000
+deepspeed: examples/deepspeed/ds_z3_config.json
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 2
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/lora_multi_gpu/multi_node.sh

@@ -0,0 +1,6 @@
+#!/bin/bash
+# also launch it on slave machine using slave_config.yaml
+
+CUDA_VISIBLE_DEVICES=0,1,2,3 accelerate launch \
+    --config_file examples/accelerate/master_config.yaml \
+    src/train.py examples/lora_multi_gpu/llama3_lora_sft.yaml

examples/lora_multi_gpu/single_node.sh

@@ -0,0 +1,5 @@
+#!/bin/bash
+
+CUDA_VISIBLE_DEVICES=0,1,2,3 accelerate launch \
+    --config_file examples/accelerate/single_config.yaml \
+    src/train.py examples/lora_multi_gpu/llama3_lora_sft.yaml

examples/lora_multi_npu/ds_zero0.sh

@@ -0,0 +1,15 @@
+#!/bin/bash
+
+NPROC_PER_NODE=4
+NNODES=1
+RANK=0
+MASTER_ADDR=127.0.0.1
+MASTER_PORT=29500
+
+ASCEND_RT_VISIBLE_DEVICES=0,1,2,3 torchrun \
+    --nproc_per_node $NPROC_PER_NODE \
+    --nnodes $NNODES \
+    --node_rank $RANK \
+    --master_addr $MASTER_ADDR \
+    --master_port $MASTER_PORT \
+    src/train.py examples/lora_multi_npu/llama3_lora_sft_ds.yaml

examples/lora_multi_npu/llama3_lora_sft_ds.yaml

@@ -0,0 +1,42 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# ddp
+ddp_timeout: 180000000
+deepspeed: examples/deepspeed/ds_z0_config.json
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 2
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/lora_single_gpu/llama3_lora_dpo.yaml

@@ -0,0 +1,39 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: dpo
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+dpo_ftx: 1.0
+
+# dataset
+dataset: orca_rlhf
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/dpo
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.00001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/lora_single_gpu/llama3_lora_eval.yaml

@@ -0,0 +1,19 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+adapter_name_or_path: saves/llama3-8b/lora/sft
+
+# method
+finetuning_type: lora
+
+# dataset
+task: mmlu
+split: test
+template: fewshot
+lang: en
+n_shot: 5
+
+# output
+save_dir: saves/llama3-8b/lora/eval
+
+# eval
+batch_size: 4

examples/lora_single_gpu/llama3_lora_orpo.yaml

@@ -0,0 +1,38 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: orpo
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: orca_rlhf
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/orpo
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.00001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/lora_single_gpu/llama3_lora_ppo.yaml

@@ -0,0 +1,38 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+reward_model: saves/llama3-8b/lora/reward
+
+# method
+stage: ppo
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/ppo
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.00001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# generate
+max_new_tokens: 512
+top_k: 0
+top_p: 0.9

examples/lora_single_gpu/llama3_lora_predict.yaml

@@ -0,0 +1,24 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+adapter_name_or_path: saves/llama3-8b/lora/sft
+
+# method
+stage: sft
+do_predict: true
+finetuning_type: lora
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 50
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/predict
+overwrite_output_dir: true
+
+# eval
+per_device_eval_batch_size: 1
+predict_with_generate: true

examples/lora_single_gpu/llama3_lora_pretrain.yaml

@@ -0,0 +1,37 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: pt
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: c4_demo
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/lora_single_gpu/llama3_lora_reward.yaml

@@ -0,0 +1,38 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: rm
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: orca_rlhf
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/reward
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.00001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/lora_single_gpu/llama3_lora_sft.yaml

@@ -0,0 +1,38 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/lora_single_gpu/llama3_preprocess.yaml

@@ -0,0 +1,21 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+tokenized_path: saves/llama3-8b/dataset/sft
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+overwrite_output_dir: true

examples/lora_single_gpu/llava1_5_lora_sft.yaml

@@ -0,0 +1,39 @@
+# model
+model_name_or_path: llava-hf/llava-1.5-7b-hf
+visual_inputs: true
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: mllm_demo
+template: vicuna
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llava1_5-7b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/merge_lora/llama3_gptq.yaml

@@ -0,0 +1,11 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+template: llama3
+
+# export
+export_dir: models/llama3_gptq
+export_quantization_bit: 4
+export_quantization_dataset: data/c4_demo.json
+export_size: 2
+export_device: cpu
+export_legacy_format: false

examples/merge_lora/llama3_lora_sft.yaml

@@ -0,0 +1,13 @@
+# Note: DO NOT use quantized model or quantization_bit when merging lora adapters
+
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+adapter_name_or_path: saves/llama3-8b/lora/sft
+template: llama3
+finetuning_type: lora
+
+# export
+export_dir: models/llama3_lora_sft
+export_size: 2
+export_device: cpu
+export_legacy_format: false

examples/qlora_single_gpu/llama3_lora_sft_aqlm.yaml

@@ -0,0 +1,38 @@
+# model
+model_name_or_path: ISTA-DASLab/Meta-Llama-3-8B-Instruct-AQLM-2Bit-1x16
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/qlora_single_gpu/llama3_lora_sft_awq.yaml

@@ -0,0 +1,38 @@
+# model
+model_name_or_path: TechxGenus/Meta-Llama-3-8B-Instruct-AWQ
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/qlora_single_gpu/llama3_lora_sft_bitsandbytes.yaml

@@ -0,0 +1,39 @@
+# model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+quantization_bit: 4
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

examples/qlora_single_gpu/llama3_lora_sft_gptq.yaml

@@ -0,0 +1,38 @@
+# model
+model_name_or_path: TechxGenus/Meta-Llama-3-8B-Instruct-GPTQ
+
+# method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: q_proj,v_proj
+
+# dataset
+dataset: identity,alpaca_gpt4_en
+template: llama3
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+# output
+output_dir: saves/llama3-8b/lora/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+# train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 0.0001
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+# eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+evaluation_strategy: steps
+eval_steps: 500

pyproject.toml

@@ -1,3 +1,33 @@
 [build-system]
 requires = ["setuptools>=61.0"]
 build-backend = "setuptools.build_meta"
+
+[tool.ruff]
+target-version = "py38"
+line-length = 119
+indent-width = 4
+
+[tool.ruff.lint]
+ignore = ["C408", "C901", "E501", "E731", "E741", "W605"]
+select = ["C", "E", "F", "I", "W"]
+
+[tool.ruff.lint.isort]
+lines-after-imports = 2
+known-first-party = ["llmtuner"]
+known-third-party = [
+    "accelerate",
+    "datasets",
+    "gradio",
+    "numpy",
+    "peft",
+    "torch",
+    "transformers",
+    "trl"
+]
+
+[tool.ruff.format]
+quote-style = "double"
+indent-style = "space"
+docstring-code-format = true
+skip-magic-trailing-comma = false
+line-ending = "auto"

requirements.txt

@@ -1,16 +1,18 @@
-torch>=1.13.1
-transformers>=4.29.1
-datasets>=2.12.0
-accelerate>=0.21.0
-peft>=0.4.0
-trl>=0.4.7
+transformers>=4.37.2
+datasets>=2.14.3
+accelerate>=0.27.2
+peft>=0.10.0
+trl>=0.8.1
+gradio>=4.0.0
+scipy
+einops
 sentencepiece
-jieba
-rouge-chinese
-nltk
-gradio>=3.36.0
+protobuf
 uvicorn
-pydantic==1.10.11
-fastapi==0.95.1
+pydantic
+fastapi
 sse-starlette
-matplotlib
+matplotlib>=3.7.0
+fire
+packaging
+pyyaml

scripts/cal_flops.py

@@ -0,0 +1,31 @@
+# coding=utf-8
+# Calculates the flops of pre-trained models.
+# Usage: python cal_flops.py --model_name_or_path path_to_model --batch_size 1 --seq_length 512
+# Inspired by: https://www.deepspeed.ai/tutorials/flops-profiler/
+
+import fire
+import torch
+from deepspeed.accelerator import get_accelerator  # type: ignore
+from deepspeed.profiling.flops_profiler import get_model_profile  # type: ignore
+
+from llmtuner.chat import ChatModel
+
+
+def calculate_flops(
+    model_name_or_path: str,
+    batch_size: int = 1,
+    seq_length: int = 256,
+    flash_attn: str = "auto",
+):
+    with get_accelerator().device(0):
+        chat_model = ChatModel(dict(model_name_or_path=model_name_or_path, template="empty", flash_attn=flash_attn))
+        fake_input = torch.ones((batch_size, seq_length), dtype=torch.long, device=chat_model.model.device)
+        input_dict = {"input_ids": fake_input, "labels": fake_input.clone()}
+        flops, macs, params = get_model_profile(chat_model.model, kwargs=input_dict, print_profile=True, detailed=True)
+        print("FLOPs:", flops)
+        print("MACs:", macs)
+        print("Params:", params)
+
+
+if __name__ == "__main__":
+    fire.Fire(calculate_flops)

scripts/cal_lr.py

@@ -0,0 +1,76 @@
+# coding=utf-8
+# Calculates the optimal learning rate for 7B/13B models using LLaMA's hyper-parameters.
+# Usage: python cal_lr.py --model_name_or_path path_to_model --dataset alpaca_en --cutoff_len 1024 --batch_size 16
+# Inspired by: https://github.com/imoneoi/openchat/blob/master/ochat/training_deepspeed/train.py
+
+import math
+from typing import Literal
+
+import fire
+import torch
+from torch.utils.data import DataLoader
+from tqdm import tqdm
+from transformers import DataCollatorForLanguageModeling, DataCollatorForSeq2Seq
+
+from llmtuner.data import get_dataset
+from llmtuner.extras.constants import IGNORE_INDEX
+from llmtuner.hparams import get_train_args
+from llmtuner.model import load_tokenizer
+
+
+BASE_LR = 3e-4  # 1.5e-4 for 30B-70B models
+BASE_BS = 4_000_000  # from llama paper
+
+
+def calculate_lr(
+    model_name_or_path: str,
+    batch_size: int,  # total batch size, namely (batch size * gradient accumulation * world size)
+    stage: Literal["pt", "sft"] = "sft",
+    dataset: str = "alpaca_en",
+    dataset_dir: str = "data",
+    template: str = "default",
+    cutoff_len: int = 1024,  # i.e. maximum input length during training
+    is_mistral: bool = False,  # mistral model uses a smaller learning rate,
+):
+    model_args, data_args, training_args, _, _ = get_train_args(
+        dict(
+            stage=stage,
+            model_name_or_path=model_name_or_path,
+            dataset=dataset,
+            dataset_dir=dataset_dir,
+            template=template,
+            cutoff_len=cutoff_len,
+            output_dir="dummy_dir",
+            overwrite_cache=True,
+        )
+    )
+    tokenizer_module = load_tokenizer(model_args)
+    tokenizer = tokenizer_module["tokenizer"]
+    trainset = get_dataset(model_args, data_args, training_args, stage, **tokenizer_module)
+    if stage == "pt":
+        data_collator = DataCollatorForLanguageModeling(tokenizer=tokenizer, mlm=False)
+    elif stage == "sft":
+        data_collator = DataCollatorForSeq2Seq(tokenizer=tokenizer, label_pad_token_id=IGNORE_INDEX)
+    else:
+        raise NotImplementedError
+
+    dataloader = DataLoader(trainset, batch_size, shuffle=False, collate_fn=data_collator, pin_memory=True)
+    valid_tokens, total_tokens = 0, 0
+    for batch in tqdm(dataloader):
+        valid_tokens += torch.sum(batch["labels"] != IGNORE_INDEX).item()
+        total_tokens += torch.numel(batch["labels"])
+
+    batch_max_len = cutoff_len * batch_size  # max tokens in a batch
+    valid_ratio = valid_tokens / total_tokens
+    batch_valid_len = batch_max_len * valid_ratio
+    lr = BASE_LR * math.sqrt(batch_valid_len / BASE_BS)  # lr ~ sqrt(batch_size)
+    lr = lr / 6.0 if is_mistral else lr
+    print(
+        "Optimal learning rate is {:.2e} for valid ratio% {:.2f} and effective batch size {:.2f}".format(
+            lr, valid_ratio * 100, batch_valid_len
+        )
+    )
+
+
+if __name__ == "__main__":
+    fire.Fire(calculate_lr)

scripts/cal_ppl.py

@@ -0,0 +1,116 @@
+# coding=utf-8
+# Calculates the ppl on the dataset of the pre-trained models.
+# Usage: python cal_ppl.py --model_name_or_path path_to_model --save_name ppl.json
+
+import json
+from dataclasses import dataclass
+from typing import Any, Dict, Literal, Optional, Sequence
+
+import fire
+import torch
+from torch.utils.data import DataLoader
+from tqdm import tqdm
+from transformers import DataCollatorForLanguageModeling, DataCollatorForSeq2Seq
+
+from llmtuner.data import get_dataset
+from llmtuner.extras.constants import IGNORE_INDEX
+from llmtuner.hparams import get_train_args
+from llmtuner.model import load_model, load_tokenizer
+
+
+@dataclass
+class PairwiseDataCollatorWithPadding(DataCollatorForSeq2Seq):
+    r"""
+    Data collator for pairwise data.
+    """
+
+    train_on_prompt: bool = False
+
+    def __call__(self, features: Sequence[Dict[str, Any]]) -> Dict[str, torch.Tensor]:
+        r"""
+        Pads batched data to the longest sequence in the batch.
+
+        We generate 2 * n examples where the first n examples represent chosen examples and
+        the last n examples represent rejected examples.
+        """
+        chosen_features = []
+        for feature in features:
+            prompt_len, answer_len = len(feature["prompt_ids"]), len(feature["chosen_ids"])
+            input_ids = feature["prompt_ids"] + feature["chosen_ids"]
+            attention_mask = [1] * (prompt_len + answer_len)
+            labels = input_ids if self.train_on_prompt else [IGNORE_INDEX] * prompt_len + feature["chosen_ids"]
+            chosen_features.append({"input_ids": input_ids, "attention_mask": attention_mask, "labels": labels})
+
+        return super().__call__(chosen_features)
+
+
+def cal_ppl(
+    model_name_or_path: str,
+    save_name: str,
+    batch_size: int = 4,
+    stage: Literal["pt", "sft", "rm"] = "sft",
+    dataset: str = "alpaca_en",
+    dataset_dir: str = "data",
+    template: str = "default",
+    cutoff_len: int = 1024,
+    max_samples: Optional[int] = None,
+    train_on_prompt: bool = False,
+):
+    model_args, data_args, training_args, finetuning_args, _ = get_train_args(
+        dict(
+            stage=stage,
+            model_name_or_path=model_name_or_path,
+            dataset=dataset,
+            dataset_dir=dataset_dir,
+            template=template,
+            cutoff_len=cutoff_len,
+            max_samples=max_samples,
+            train_on_prompt=train_on_prompt,
+            output_dir="dummy_dir",
+            overwrite_cache=True,
+        )
+    )
+    tokenizer_module = load_tokenizer(model_args)
+    tokenizer = tokenizer_module["tokenizer"]
+    trainset = get_dataset(model_args, data_args, training_args, stage, **tokenizer_module)
+    model = load_model(tokenizer, model_args, finetuning_args, is_trainable=False)
+    if stage == "pt":
+        data_collator = DataCollatorForLanguageModeling(tokenizer=tokenizer, mlm=False)
+    elif stage == "sft":
+        data_collator = DataCollatorForSeq2Seq(tokenizer=tokenizer, label_pad_token_id=IGNORE_INDEX)
+    elif stage == "rm":
+        data_collator = PairwiseDataCollatorWithPadding(
+            tokenizer=tokenizer, label_pad_token_id=IGNORE_INDEX, train_on_prompt=train_on_prompt
+        )
+    else:
+        raise NotImplementedError
+
+    dataloader = DataLoader(trainset, batch_size, shuffle=False, collate_fn=data_collator, pin_memory=True)
+    criterion = torch.nn.CrossEntropyLoss(reduction="none")
+    total_ppl = 0
+    perplexities = []
+    batch: Dict[str, "torch.Tensor"]
+    with torch.no_grad():
+        for batch in tqdm(dataloader):
+            batch = batch.to(model.device)
+            outputs = model(**batch)
+            shift_logits: "torch.Tensor" = outputs["logits"][..., :-1, :]
+            shift_labels: "torch.Tensor" = batch["labels"][..., 1:]
+            loss_mask = shift_labels != IGNORE_INDEX
+            flatten_logits = shift_logits.contiguous().view(shift_labels.size(0) * shift_labels.size(1), -1)
+            flatten_labels = shift_labels.contiguous().view(-1)
+            token_logps: "torch.Tensor" = criterion(flatten_logits, flatten_labels)
+            token_logps = token_logps.contiguous().view(shift_logits.size(0), -1)
+            sentence_logps = (token_logps * loss_mask).sum(-1) / loss_mask.sum(-1)
+            total_ppl += sentence_logps.exp().sum().item()
+            perplexities.extend(sentence_logps.exp().tolist())
+
+    with open(save_name, "w", encoding="utf-8") as f:
+        json.dump(perplexities, f, indent=2)
+
+    print("Average perplexity is {:.2f}".format(total_ppl / len(perplexities)))
+    print("Perplexities have been saved at {}.".format(save_name))
+
+
+if __name__ == "__main__":
+    fire.Fire(cal_ppl)

scripts/length_cdf.py

@@ -0,0 +1,51 @@
+# coding=utf-8
+# Calculates the distribution of the input lengths in the dataset.
+# Usage: python length_cdf.py --model_name_or_path path_to_model --dataset alpaca_en --template default
+
+from collections import defaultdict
+
+import fire
+from tqdm import tqdm
+
+from llmtuner.data import get_dataset
+from llmtuner.hparams import get_train_args
+from llmtuner.model import load_tokenizer
+
+
+def length_cdf(
+    model_name_or_path: str,
+    dataset: str = "alpaca_en",
+    dataset_dir: str = "data",
+    template: str = "default",
+    interval: int = 1000,
+):
+    model_args, data_args, training_args, _, _ = get_train_args(
+        dict(
+            stage="sft",
+            model_name_or_path=model_name_or_path,
+            dataset=dataset,
+            dataset_dir=dataset_dir,
+            template=template,
+            cutoff_len=1_000_000,
+            output_dir="dummy_dir",
+            overwrite_cache=True,
+        )
+    )
+    tokenizer_module = load_tokenizer(model_args)
+    trainset = get_dataset(model_args, data_args, training_args, stage="sft", **tokenizer_module)
+    total_num = len(trainset)
+    length_dict = defaultdict(int)
+    for sample in tqdm(trainset["input_ids"]):
+        length_dict[len(sample) // interval * interval] += 1
+
+    length_tuples = list(length_dict.items())
+    length_tuples.sort()
+    count_accu, prob_accu = 0, 0
+    for length, count in length_tuples:
+        count_accu += count
+        prob_accu += count / total_num * 100
+        print("{:d} ({:.2f}%) samples have length < {}.".format(count_accu, prob_accu, length + interval))
+
+
+if __name__ == "__main__":
+    fire.Fire(length_cdf)

scripts/llama_pro.py

@@ -0,0 +1,114 @@
+# coding=utf-8
+# Performs block expansion for LLaMA, Mistral, Qwen1.5 or Yi models.
+# Usage: python llama_pro.py --model_name_or_path meta-llama/Llama-2-7b-hf --output_dir llama2_pro --num_expand 8
+# Inspired by: https://github.com/TencentARC/LLaMA-Pro/blob/main/scripts/block_expansion.py
+
+import json
+import os
+from collections import OrderedDict
+from typing import TYPE_CHECKING, Optional
+
+import fire
+import torch
+from safetensors.torch import save_file
+from tqdm import tqdm
+from transformers import AutoConfig, AutoModelForCausalLM, AutoTokenizer
+from transformers.modeling_utils import (
+    SAFE_WEIGHTS_INDEX_NAME,
+    SAFE_WEIGHTS_NAME,
+    WEIGHTS_INDEX_NAME,
+    WEIGHTS_NAME,
+    shard_checkpoint,
+)
+
+
+if TYPE_CHECKING:
+    from transformers import PretrainedConfig, PreTrainedModel
+
+
+def change_name(name: str, old_index: int, new_index: int) -> str:
+    return name.replace(".{:d}.".format(old_index), ".{:d}.".format(new_index))
+
+
+def block_expansion(
+    model_name_or_path: str,
+    output_dir: str,
+    num_expand: int,
+    shard_size: Optional[str] = "2GB",
+    save_safetensors: Optional[bool] = False,
+):
+    config: "PretrainedConfig" = AutoConfig.from_pretrained(model_name_or_path)
+    num_layers = getattr(config, "num_hidden_layers")
+    setattr(config, "num_hidden_layers", num_layers + num_expand)
+    config.save_pretrained(output_dir)
+
+    tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
+    tokenizer.save_pretrained(output_dir)
+
+    config: "PretrainedConfig" = AutoConfig.from_pretrained(model_name_or_path)  # load the original one
+    if save_safetensors:
+        setattr(config, "tie_word_embeddings", False)  # safetensors does not allow shared weights
+
+    model: "PreTrainedModel" = AutoModelForCausalLM.from_pretrained(
+        model_name_or_path,
+        config=config,
+        torch_dtype="auto",
+        trust_remote_code=True,
+        low_cpu_mem_usage=True,
+    )
+    state_dict = model.state_dict()
+
+    if num_layers % num_expand != 0:
+        raise ValueError("`num_layers` {} should be divisible by `num_expand` {}.".format(num_layers, num_expand))
+
+    split = num_layers // num_expand
+    layer_cnt = 0
+    output_state_dict = OrderedDict()
+    for i in range(num_layers):
+        for key, value in state_dict.items():
+            if ".{:d}.".format(i) in key:
+                output_state_dict[change_name(key, i, layer_cnt)] = value
+
+        print("Add layer {} copied from layer {}".format(layer_cnt, i))
+        layer_cnt += 1
+        if (i + 1) % split == 0:
+            for key, value in state_dict.items():
+                if ".{:d}.".format(i) in key:
+                    if "down_proj" in key or "o_proj" in key:
+                        output_state_dict[change_name(key, i, layer_cnt)] = torch.zeros_like(value)
+                    else:
+                        output_state_dict[change_name(key, i, layer_cnt)] = torch.clone(value)
+
+            print("Add layer {} expanded from layer {}".format(layer_cnt, i))
+            layer_cnt += 1
+
+    for key, value in state_dict.items():
+        if key not in output_state_dict:
+            output_state_dict[key] = value
+
+    weights_name = SAFE_WEIGHTS_NAME if save_safetensors else WEIGHTS_NAME
+    shards, index = shard_checkpoint(output_state_dict, max_shard_size=shard_size, weights_name=weights_name)
+
+    for shard_file, shard in tqdm(shards.items(), desc="Save weights"):
+        if save_safetensors:
+            save_file(shard, os.path.join(output_dir, shard_file), metadata={"format": "pt"})
+        else:
+            torch.save(shard, os.path.join(output_dir, shard_file))
+
+    if index is None:
+        print("Model weights saved in {}".format(os.path.join(output_dir, weights_name)))
+    else:
+        index_name = SAFE_WEIGHTS_INDEX_NAME if save_safetensors else WEIGHTS_INDEX_NAME
+        with open(os.path.join(output_dir, index_name), "w", encoding="utf-8") as f:
+            json.dump(index, f, indent=2, sort_keys=True)
+        print("Model weights saved in {}".format(output_dir))
+
+    print("Fine-tune this model with:")
+    print("  --model_name_or_path {} \\".format(output_dir))
+    print("  --finetuning_type freeze \\")
+    print("  --freeze_trainable_layers {} \\".format(num_expand))
+    print("  --use_llama_pro")
+
+
+if __name__ == "__main__":
+    fire.Fire(block_expansion)

scripts/llamafy_baichuan2.py

@@ -0,0 +1,92 @@
+# coding=utf-8
+# Converts the Baichuan2-7B model in the same format as LLaMA2-7B.
+# Usage: python llamafy_baichuan2.py --input_dir input --output_dir output
+# Inspired by: https://huggingface.co/fireballoon/baichuan-llama-7b/blob/main/convert_baichuan_to_llama.py
+# Converted model: https://huggingface.co/hiyouga/Baichuan2-7B-Base-LLaMAfied
+
+import json
+import os
+from collections import OrderedDict
+from typing import Any, Dict, Optional
+
+import fire
+import torch
+from safetensors.torch import save_file
+from tqdm import tqdm
+from transformers.modeling_utils import (
+    SAFE_WEIGHTS_INDEX_NAME,
+    SAFE_WEIGHTS_NAME,
+    WEIGHTS_INDEX_NAME,
+    WEIGHTS_NAME,
+    shard_checkpoint,
+)
+
+
+CONFIG_NAME = "config.json"
+
+
+def save_weight(input_dir: str, output_dir: str, shard_size: str, save_safetensors: bool):
+    baichuan2_state_dict: Dict[str, torch.Tensor] = OrderedDict()
+    for filepath in tqdm(os.listdir(input_dir), desc="Load weights"):
+        if os.path.isfile(os.path.join(input_dir, filepath)) and filepath.endswith(".bin"):
+            shard_weight = torch.load(os.path.join(input_dir, filepath), map_location="cpu")
+            baichuan2_state_dict.update(shard_weight)
+
+    llama2_state_dict: Dict[str, torch.Tensor] = OrderedDict()
+    for key, value in tqdm(baichuan2_state_dict.items(), desc="Convert format"):
+        if "W_pack" in key:
+            proj_size = value.size(0) // 3
+            llama2_state_dict[key.replace("W_pack", "q_proj")] = value[:proj_size, :]
+            llama2_state_dict[key.replace("W_pack", "k_proj")] = value[proj_size : 2 * proj_size, :]
+            llama2_state_dict[key.replace("W_pack", "v_proj")] = value[2 * proj_size :, :]
+        elif "lm_head" in key:
+            llama2_state_dict[key] = torch.nn.functional.normalize(value)
+        else:
+            llama2_state_dict[key] = value
+
+    weights_name = SAFE_WEIGHTS_NAME if save_safetensors else WEIGHTS_NAME
+    shards, index = shard_checkpoint(llama2_state_dict, max_shard_size=shard_size, weights_name=weights_name)
+
+    for shard_file, shard in tqdm(shards.items(), desc="Save weights"):
+        if save_safetensors:
+            save_file(shard, os.path.join(output_dir, shard_file), metadata={"format": "pt"})
+        else:
+            torch.save(shard, os.path.join(output_dir, shard_file))
+
+    if index is None:
+        print("Model weights saved in {}".format(os.path.join(output_dir, WEIGHTS_NAME)))
+    else:
+        index_name = SAFE_WEIGHTS_INDEX_NAME if save_safetensors else WEIGHTS_INDEX_NAME
+        with open(os.path.join(output_dir, index_name), "w", encoding="utf-8") as f:
+            json.dump(index, f, indent=2, sort_keys=True)
+        print("Model weights saved in {}".format(output_dir))
+
+
+def save_config(input_dir: str, output_dir: str):
+    with open(os.path.join(input_dir, CONFIG_NAME), "r", encoding="utf-8") as f:
+        llama2_config_dict: Dict[str, Any] = json.load(f)
+
+    llama2_config_dict["architectures"] = ["LlamaForCausalLM"]
+    llama2_config_dict.pop("auto_map", None)
+    llama2_config_dict.pop("tokenizer_class", None)
+    llama2_config_dict["model_type"] = "llama"
+
+    with open(os.path.join(output_dir, CONFIG_NAME), "w", encoding="utf-8") as f:
+        json.dump(llama2_config_dict, f, indent=2)
+    print("Model config saved in {}".format(os.path.join(output_dir, CONFIG_NAME)))
+
+
+def llamafy_baichuan2(
+    input_dir: str, output_dir: str, shard_size: Optional[str] = "2GB", save_safetensors: Optional[bool] = False
+):
+    try:
+        os.makedirs(output_dir, exist_ok=False)
+    except Exception as e:
+        raise print("Output dir already exists", e)
+
+    save_weight(input_dir, output_dir, shard_size, save_safetensors)
+    save_config(input_dir, output_dir)
+
+
+if __name__ == "__main__":
+    fire.Fire(llamafy_baichuan2)

scripts/llamafy_qwen.py

@@ -0,0 +1,144 @@
+# coding=utf-8
+# Converts the Qwen models in the same format as LLaMA2.
+# Usage: python llamafy_qwen.py --input_dir input --output_dir output
+# Converted model: https://huggingface.co/hiyouga/Qwen-14B-Chat-LLaMAfied
+
+import json
+import os
+from collections import OrderedDict
+from typing import Any, Dict, Optional
+
+import fire
+import torch
+from safetensors import safe_open
+from safetensors.torch import save_file
+from tqdm import tqdm
+from transformers.modeling_utils import (
+    SAFE_WEIGHTS_INDEX_NAME,
+    SAFE_WEIGHTS_NAME,
+    WEIGHTS_INDEX_NAME,
+    WEIGHTS_NAME,
+    shard_checkpoint,
+)
+from transformers.utils import check_min_version
+
+
+try:
+    check_min_version("4.34.0")
+except Exception:
+    raise ValueError("Please upgrade `transformers` to 4.34.0")
+
+
+CONFIG_NAME = "config.json"
+
+
+def save_weight(input_dir: str, output_dir: str, shard_size: str, save_safetensors: bool) -> str:
+    qwen_state_dict: Dict[str, torch.Tensor] = OrderedDict()
+    for filepath in tqdm(os.listdir(input_dir), desc="Load weights"):
+        if os.path.isfile(os.path.join(input_dir, filepath)) and filepath.endswith(".safetensors"):
+            with safe_open(os.path.join(input_dir, filepath), framework="pt", device="cpu") as f:
+                for key in f.keys():
+                    qwen_state_dict[key] = f.get_tensor(key)
+
+    llama2_state_dict: Dict[str, torch.Tensor] = OrderedDict()
+    torch_dtype = None
+    for key, value in tqdm(qwen_state_dict.items(), desc="Convert format"):
+        if torch_dtype is None:
+            torch_dtype = value.dtype
+        if "wte" in key:
+            llama2_state_dict["model.embed_tokens.weight"] = value
+        elif "ln_f" in key:
+            llama2_state_dict["model.norm.weight"] = value
+        else:
+            key = key.replace("transformer.h", "model.layers")
+            if "attn.c_attn" in key:
+                proj_size = value.size(0) // 3
+                llama2_state_dict[key.replace("attn.c_attn", "self_attn.q_proj")] = value[:proj_size, ...]
+                llama2_state_dict[key.replace("attn.c_attn", "self_attn.k_proj")] = value[
+                    proj_size : 2 * proj_size, ...
+                ]
+                llama2_state_dict[key.replace("attn.c_attn", "self_attn.v_proj")] = value[2 * proj_size :, ...]
+            elif "attn.c_proj" in key:
+                llama2_state_dict[key.replace("attn.c_proj", "self_attn.o_proj")] = value
+                llama2_state_dict[key.replace("attn.c_proj.weight", "self_attn.o_proj.bias")] = torch.zeros_like(
+                    value[:, 0]
+                ).squeeze()
+            elif "ln_1" in key:
+                llama2_state_dict[key.replace("ln_1", "input_layernorm")] = value
+            elif "ln_2" in key:
+                llama2_state_dict[key.replace("ln_2", "post_attention_layernorm")] = value
+            elif "mlp.w1" in key:
+                llama2_state_dict[key.replace("mlp.w1", "mlp.up_proj")] = value
+            elif "mlp.w2" in key:
+                llama2_state_dict[key.replace("mlp.w2", "mlp.gate_proj")] = value
+            elif "mlp.c_proj" in key:
+                llama2_state_dict[key.replace("mlp.c_proj", "mlp.down_proj")] = value
+            elif "lm_head" in key:
+                llama2_state_dict[key] = value
+            else:
+                raise KeyError("Unable to process key {}".format(key))
+
+    weights_name = SAFE_WEIGHTS_NAME if save_safetensors else WEIGHTS_NAME
+    shards, index = shard_checkpoint(llama2_state_dict, max_shard_size=shard_size, weights_name=weights_name)
+
+    for shard_file, shard in tqdm(shards.items(), desc="Save weights"):
+        if save_safetensors:
+            save_file(shard, os.path.join(output_dir, shard_file), metadata={"format": "pt"})
+        else:
+            torch.save(shard, os.path.join(output_dir, shard_file))
+
+    if index is None:
+        print("Model weights saved in {}".format(os.path.join(output_dir, weights_name)))
+    else:
+        index_name = SAFE_WEIGHTS_INDEX_NAME if save_safetensors else WEIGHTS_INDEX_NAME
+        with open(os.path.join(output_dir, index_name), "w", encoding="utf-8") as f:
+            json.dump(index, f, indent=2, sort_keys=True)
+        print("Model weights saved in {}".format(output_dir))
+
+    return str(torch_dtype).replace("torch.", "")
+
+
+def save_config(input_dir: str, output_dir: str, torch_dtype: str):
+    with open(os.path.join(input_dir, CONFIG_NAME), "r", encoding="utf-8") as f:
+        qwen_config_dict: Dict[str, Any] = json.load(f)
+
+    llama2_config_dict: Dict[str, Any] = OrderedDict()
+    llama2_config_dict["architectures"] = ["LlamaForCausalLM"]
+    llama2_config_dict["hidden_act"] = "silu"
+    llama2_config_dict["hidden_size"] = qwen_config_dict["hidden_size"]
+    llama2_config_dict["initializer_range"] = qwen_config_dict["initializer_range"]
+    llama2_config_dict["intermediate_size"] = qwen_config_dict["intermediate_size"] // 2
+    llama2_config_dict["max_position_embeddings"] = qwen_config_dict["max_position_embeddings"]
+    llama2_config_dict["model_type"] = "llama"
+    llama2_config_dict["num_attention_heads"] = qwen_config_dict["num_attention_heads"]
+    llama2_config_dict["num_hidden_layers"] = qwen_config_dict["num_hidden_layers"]
+    llama2_config_dict["num_key_value_heads"] = qwen_config_dict["hidden_size"] // qwen_config_dict["kv_channels"]
+    llama2_config_dict["pretraining_tp"] = 1
+    llama2_config_dict["rms_norm_eps"] = qwen_config_dict["layer_norm_epsilon"]
+    llama2_config_dict["rope_scaling"] = None
+    llama2_config_dict["tie_word_embeddings"] = qwen_config_dict["tie_word_embeddings"]
+    llama2_config_dict["torch_dtype"] = torch_dtype
+    llama2_config_dict["transformers_version"] = "4.34.0"
+    llama2_config_dict["use_cache"] = True
+    llama2_config_dict["vocab_size"] = qwen_config_dict["vocab_size"]
+    llama2_config_dict["attention_bias"] = True
+
+    with open(os.path.join(output_dir, CONFIG_NAME), "w", encoding="utf-8") as f:
+        json.dump(llama2_config_dict, f, indent=2)
+    print("Model config saved in {}".format(os.path.join(output_dir, CONFIG_NAME)))
+
+
+def llamafy_qwen(
+    input_dir: str, output_dir: str, shard_size: Optional[str] = "2GB", save_safetensors: Optional[bool] = False
+):
+    try:
+        os.makedirs(output_dir, exist_ok=False)
+    except Exception as e:
+        raise print("Output dir already exists", e)
+
+    torch_dtype = save_weight(input_dir, output_dir, shard_size, save_safetensors)
+    save_config(input_dir, output_dir, torch_dtype)
+
+
+if __name__ == "__main__":
+    fire.Fire(llamafy_qwen)

scripts/loftq_init.py

@@ -0,0 +1,82 @@
+# coding=utf-8
+# Initializes LoRA weights with LoRA-fine-tuning-aware Quantization (LoftQ)
+# Usage: python loftq_init.py --model_name_or_path path_to_model --save_dir output_dir
+# Inspired by: https://github.com/huggingface/peft/blob/main/examples/loftq_finetuning/quantize_save_load.py
+
+import os
+from typing import TYPE_CHECKING, Optional
+
+import fire
+import torch
+import torch.nn as nn
+from peft import LoftQConfig, LoraConfig, TaskType, get_peft_model
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+
+if TYPE_CHECKING:
+    from transformers import PreTrainedModel
+
+
+class Shell(nn.Module):
+    def __init__(self, weight: torch.Tensor, bias: Optional[torch.Tensor] = None):
+        super().__init__()
+        self.weight = nn.Parameter(weight, requires_grad=False)
+        if bias is not None:
+            self.bias = nn.Parameter(bias, requires_grad=False)
+
+
+def unwrap_model(model: nn.Module, pattern=".base_layer") -> None:
+    for name in {k.split(pattern)[0] for k, _ in model.named_modules() if pattern in k}:
+        parent_name = ".".join(name.split(".")[:-1])
+        child_name = name.split(".")[-1]
+        parent_module = model.get_submodule(parent_name)
+        child_module = getattr(parent_module, child_name)
+        base_layer = getattr(child_module, "base_layer")
+        weight = getattr(base_layer, "weight", None)
+        bias = getattr(base_layer, "bias", None)
+        setattr(parent_module, child_name, Shell(weight, bias))
+
+    print("Model unwrapped.")
+
+
+def quantize_loftq(
+    model_name_or_path: str,
+    save_dir: str,
+    loftq_bits: Optional[int] = 4,
+    loftq_iter: Optional[int] = 1,
+    lora_alpha: Optional[int] = None,
+    lora_rank: Optional[int] = 16,
+    lora_target: Optional[str] = "q_proj,v_proj",
+    save_safetensors: Optional[bool] = False,
+):
+    tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=True)
+    model = AutoModelForCausalLM.from_pretrained(model_name_or_path, trust_remote_code=True, torch_dtype="auto")
+    loftq_config = LoftQConfig(loftq_bits=loftq_bits, loftq_iter=loftq_iter)
+    lora_config = LoraConfig(
+        task_type=TaskType.CAUSAL_LM,
+        inference_mode=True,
+        r=lora_rank,
+        lora_alpha=lora_alpha if lora_alpha is not None else lora_rank * 2,
+        lora_dropout=0.1,
+        target_modules=[name.strip() for name in lora_target.split(",")],
+        init_lora_weights="loftq",
+        loftq_config=loftq_config,
+    )
+
+    # Init LoftQ model
+    lora_model = get_peft_model(model, lora_config)
+    base_model: "PreTrainedModel" = lora_model.get_base_model()
+
+    # Save LoftQ model
+    setattr(lora_model.base_model.peft_config["default"], "base_model_name_or_path", save_dir)
+    setattr(lora_model.base_model.peft_config["default"], "init_lora_weights", True)
+    lora_model.save_pretrained(os.path.join(save_dir, "adapters"), safe_serialization=save_safetensors)
+
+    # Save base model
+    unwrap_model(base_model)
+    base_model.save_pretrained(save_dir, safe_serialization=save_safetensors)
+    tokenizer.save_pretrained(save_dir)
+
+
+if __name__ == "__main__":
+    fire.Fire(quantize_loftq)

setup.py

@@ -1,13 +1,14 @@
 import os
 import re
-from setuptools import setup, find_packages
+
+from setuptools import find_packages, setup
 
 
 def get_version():
-    with open(os.path.join("src", "llmtuner", "__init__.py"), "r", encoding="utf-8") as f:
+    with open(os.path.join("src", "llmtuner", "cli.py"), "r", encoding="utf-8") as f:
         file_content = f.read()
-        pattern = r"{0}\W*=\W*\"([^\"]+)\"".format("__version__")
-        version, = re.findall(pattern, file_content)
+        pattern = r"{}\W*=\W*\"([^\"]+)\"".format("VERSION")
+        (version,) = re.findall(pattern, file_content)
         return version
 
 
@@ -18,25 +19,43 @@ def get_requires():
         return lines
 
 
-def main():
+extra_require = {
+    "torch": ["torch>=1.13.1"],
+    "metrics": ["nltk", "jieba", "rouge-chinese"],
+    "deepspeed": ["deepspeed>=0.10.0,<=0.14.0"],
+    "bitsandbytes": ["bitsandbytes>=0.39.0"],
+    "vllm": ["vllm>=0.4.0"],
+    "galore": ["galore-torch"],
+    "badam": ["badam"],
+    "gptq": ["optimum>=1.16.0", "auto-gptq>=0.5.0"],
+    "awq": ["autoawq"],
+    "aqlm": ["aqlm[gpu]>=1.1.0"],
+    "qwen": ["tiktoken", "transformers_stream_generator"],
+    "modelscope": ["modelscope"],
+    "quality": ["ruff"],
+}
 
+
+def main():
     setup(
         name="llmtuner",
         version=get_version(),
         author="hiyouga",
         author_email="hiyouga" "@" "buaa.edu.cn",
-        description="Easy-to-use fine-tuning framework using PEFT",
+        description="Easy-to-use LLM fine-tuning framework",
         long_description=open("README.md", "r", encoding="utf-8").read(),
         long_description_content_type="text/markdown",
         keywords=["LLaMA", "BLOOM", "Falcon", "LLM", "ChatGPT", "transformer", "pytorch", "deep learning"],
         license="Apache 2.0 License",
-        url="https://github.com/hiyouga/LLaMA-Efficient-Tuning",
+        url="https://github.com/hiyouga/LLaMA-Factory",
         package_dir={"": "src"},
         packages=find_packages("src"),
         python_requires=">=3.8.0",
         install_requires=get_requires(),
+        extras_require=extra_require,
+        entry_points={"console_scripts": ["llamafactory-cli = llmtuner.cli:main"]},
         classifiers=[
-            "Development Status :: 3 - Alpha",
+            "Development Status :: 4 - Beta",
             "Intended Audience :: Developers",
             "Intended Audience :: Education",
             "Intended Audience :: Science/Research",
@@ -46,8 +65,9 @@ def main():
             "Programming Language :: Python :: 3.8",
             "Programming Language :: Python :: 3.9",
             "Programming Language :: Python :: 3.10",
+            "Programming Language :: Python :: 3.11",
             "Topic :: Scientific/Engineering :: Artificial Intelligence",
-        ]
+        ],
     )
 
 

src/api.py

@@ -0,0 +1,19 @@
+import os
+
+import uvicorn
+
+from llmtuner.api.app import create_app
+from llmtuner.chat import ChatModel
+
+
+def main():
+    chat_model = ChatModel()
+    app = create_app(chat_model)
+    api_host = os.environ.get("API_HOST", "0.0.0.0")
+    api_port = int(os.environ.get("API_PORT", "8000"))
+    print("Visit http://localhost:{}/docs for API document.".format(api_port))
+    uvicorn.run(app, host=api_host, port=api_port)
+
+
+if __name__ == "__main__":
+    main()

src/api_demo.py

@@ -1,20 +0,0 @@
-# coding=utf-8
-# Implements API for fine-tuned models in OpenAI's format. (https://platform.openai.com/docs/api-reference/chat)
-# Usage: python api_demo.py --model_name_or_path path_to_model --checkpoint_dir path_to_checkpoint
-# Visit http://localhost:8000/docs for document.
-
-import uvicorn
-
-from llmtuner import ChatModel
-from llmtuner.api.app import create_app
-from llmtuner.tuner import get_infer_args
-
-
-def main():
-    chat_model = ChatModel(*get_infer_args())
-    app = create_app(chat_model)
-    uvicorn.run(app, host="0.0.0.0", port=8000, workers=1)
-
-
-if __name__ == "__main__":
-    main()

src/cli_demo.py

@@ -1,43 +0,0 @@
-# coding=utf-8
-# Implements stream chat in command line for fine-tuned models.
-# Usage: python cli_demo.py --model_name_or_path path_to_model --checkpoint_dir path_to_checkpoint
-
-from llmtuner import ChatModel
-from llmtuner.tuner import get_infer_args
-
-
-def main():
-    chat_model = ChatModel(*get_infer_args())
-    history = []
-    print("Welcome to the CLI application, use `clear` to remove the history, use `exit` to exit the application.")
-
-    while True:
-        try:
-            query = input("\nUser: ")
-        except UnicodeDecodeError:
-            print("Detected decoding error at the inputs, please set the terminal encoding to utf-8.")
-            continue
-        except Exception:
-            raise
-
-        if query.strip() == "exit":
-            break
-
-        if query.strip() == "clear":
-            history = []
-            print("History has been removed.")
-            continue
-
-        print("Assistant: ", end="", flush=True)
-
-        response = ""
-        for new_text in chat_model.stream_chat(query, history):
-            print(new_text, end="", flush=True)
-            response += new_text
-        print()
-
-        history = history + [(query, response)]
-
-
-if __name__ == "__main__":
-    main()

src/export_model.py

@@ -1,17 +0,0 @@
-# coding=utf-8
-# Exports the fine-tuned model.
-# Usage: python export_model.py --checkpoint_dir path_to_checkpoint --output_dir path_to_save_model
-
-from llmtuner.tuner import get_train_args, load_model_and_tokenizer
-
-
-def main():
-    model_args, _, training_args, finetuning_args, _ = get_train_args()
-    model, tokenizer = load_model_and_tokenizer(model_args, finetuning_args)
-    model.save_pretrained(training_args.output_dir, max_shard_size="10GB")
-    tokenizer.save_pretrained(training_args.output_dir)
-    print("model and tokenizer have been saved at:", training_args.output_dir)
-
-
-if __name__ == "__main__":
-    main()

src/llmtuner/__init__.py

@@ -1,4 +1,6 @@
-from llmtuner.chat import ChatModel
+# Level: api, webui > chat, eval, train > data, model > extras, hparams
+
+from .cli import VERSION
 
 
-__version__ = "0.1.5"
+__version__ = VERSION

src/llmtuner/api/app.py

@@ -1,38 +1,49 @@
-import uvicorn
-from fastapi import FastAPI, HTTPException
-from fastapi.middleware.cors import CORSMiddleware
+import os
 from contextlib import asynccontextmanager
-from sse_starlette import EventSourceResponse
-from typing import List, Tuple
+from typing import Optional
 
-from llmtuner.tuner import get_infer_args
-from llmtuner.extras.misc import torch_gc
-from llmtuner.chat.stream_chat import ChatModel
-from llmtuner.api.protocol import (
-    Role,
-    Finish,
-    ModelCard,
-    ModelList,
-    ChatMessage,
-    DeltaMessage,
+from typing_extensions import Annotated
+
+from ..chat import ChatModel
+from ..extras.misc import torch_gc
+from ..extras.packages import is_fastapi_available, is_starlette_available, is_uvicorn_available
+from .chat import (
+    create_chat_completion_response,
+    create_score_evaluation_response,
+    create_stream_chat_completion_response,
+)
+from .protocol import (
     ChatCompletionRequest,
     ChatCompletionResponse,
-    ChatCompletionStreamResponse,
-    ChatCompletionResponseChoice,
-    ChatCompletionResponseStreamChoice,
-    ChatCompletionResponseUsage
+    ModelCard,
+    ModelList,
+    ScoreEvaluationRequest,
+    ScoreEvaluationResponse,
 )
 
 
+if is_fastapi_available():
+    from fastapi import Depends, FastAPI, HTTPException, status
+    from fastapi.middleware.cors import CORSMiddleware
+    from fastapi.security.http import HTTPAuthorizationCredentials, HTTPBearer
+
+
+if is_starlette_available():
+    from sse_starlette import EventSourceResponse
+
+
+if is_uvicorn_available():
+    import uvicorn
+
+
 @asynccontextmanager
-async def lifespan(app: FastAPI): # collects GPU memory
+async def lifespan(app: "FastAPI"):  # collects GPU memory
     yield
     torch_gc()
 
 
-def create_app(chat_model: ChatModel) -> FastAPI:
+def create_app(chat_model: "ChatModel") -> "FastAPI":
     app = FastAPI(lifespan=lifespan)
-
     app.add_middleware(
         CORSMiddleware,
         allow_origins=["*"],
@@ -40,88 +51,58 @@ def create_app(chat_model: ChatModel) -> FastAPI:
         allow_methods=["*"],
         allow_headers=["*"],
     )
+    api_key = os.environ.get("API_KEY")
+    security = HTTPBearer(auto_error=False)
 
-    @app.get("/v1/models", response_model=ModelList)
+    async def verify_api_key(auth: Annotated[Optional[HTTPAuthorizationCredentials], Depends(security)]):
+        if api_key and (auth is None or auth.credentials != api_key):
+            raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="Invalid API key.")
+
+    @app.get(
+        "/v1/models",
+        response_model=ModelList,
+        status_code=status.HTTP_200_OK,
+        dependencies=[Depends(verify_api_key)],
+    )
     async def list_models():
         model_card = ModelCard(id="gpt-3.5-turbo")
         return ModelList(data=[model_card])
 
-    @app.post("/v1/chat/completions", response_model=ChatCompletionResponse)
+    @app.post(
+        "/v1/chat/completions",
+        response_model=ChatCompletionResponse,
+        status_code=status.HTTP_200_OK,
+        dependencies=[Depends(verify_api_key)],
+    )
     async def create_chat_completion(request: ChatCompletionRequest):
-        if request.messages[-1].role != Role.USER:
-            raise HTTPException(status_code=400, detail="Invalid request")
-        query = request.messages[-1].content
-
-        prev_messages = request.messages[:-1]
-        if len(prev_messages) > 0 and prev_messages[0].role == Role.SYSTEM:
-            prefix = prev_messages.pop(0).content
-        else:
-            prefix = None
-
-        history = []
-        if len(prev_messages) % 2 == 0:
-            for i in range(0, len(prev_messages), 2):
-                if prev_messages[i].role == Role.USER and prev_messages[i+1].role == Role.ASSISTANT:
-                    history.append([prev_messages[i].content, prev_messages[i+1].content])
+        if not chat_model.engine.can_generate:
+            raise HTTPException(status_code=status.HTTP_405_METHOD_NOT_ALLOWED, detail="Not allowed")
 
         if request.stream:
-            generate = predict(query, history, prefix, request)
+            generate = create_stream_chat_completion_response(request, chat_model)
             return EventSourceResponse(generate, media_type="text/event-stream")
+        else:
+            return await create_chat_completion_response(request, chat_model)
 
-        response, (prompt_length, response_length) = chat_model.chat(
-            query, history, prefix, temperature=request.temperature, top_p=request.top_p, max_new_tokens=request.max_tokens
-        )
+    @app.post(
+        "/v1/score/evaluation",
+        response_model=ScoreEvaluationResponse,
+        status_code=status.HTTP_200_OK,
+        dependencies=[Depends(verify_api_key)],
+    )
+    async def create_score_evaluation(request: ScoreEvaluationRequest):
+        if chat_model.engine.can_generate:
+            raise HTTPException(status_code=status.HTTP_405_METHOD_NOT_ALLOWED, detail="Not allowed")
 
-        usage = ChatCompletionResponseUsage(
-            prompt_tokens=prompt_length,
-            completion_tokens=response_length,
-            total_tokens=prompt_length+response_length
-        )
-
-        choice_data = ChatCompletionResponseChoice(
-            index=0,
-            message=ChatMessage(role=Role.ASSISTANT, content=response),
-            finish_reason=Finish.STOP
-        )
-
-        return ChatCompletionResponse(model=request.model, choices=[choice_data], usage=usage)
-
-    async def predict(query: str, history: List[Tuple[str, str]], prefix: str, request: ChatCompletionRequest):
-        choice_data = ChatCompletionResponseStreamChoice(
-            index=0,
-            delta=DeltaMessage(role=Role.ASSISTANT),
-            finish_reason=None
-        )
-        chunk = ChatCompletionStreamResponse(model=request.model, choices=[choice_data])
-        yield chunk.json(exclude_unset=True, ensure_ascii=False)
-
-        for new_text in chat_model.stream_chat(
-            query, history, prefix, temperature=request.temperature, top_p=request.top_p, max_new_tokens=request.max_tokens
-        ):
-            if len(new_text) == 0:
-                continue
-
-            choice_data = ChatCompletionResponseStreamChoice(
-                index=0,
-                delta=DeltaMessage(content=new_text),
-                finish_reason=None
-            )
-            chunk = ChatCompletionStreamResponse(model=request.model, choices=[choice_data])
-            yield chunk.json(exclude_unset=True, ensure_ascii=False)
-
-        choice_data = ChatCompletionResponseStreamChoice(
-            index=0,
-            delta=DeltaMessage(),
-            finish_reason=Finish.STOP
-        )
-        chunk = ChatCompletionStreamResponse(model=request.model, choices=[choice_data])
-        yield chunk.json(exclude_unset=True, ensure_ascii=False)
-        yield "[DONE]"
+        return await create_score_evaluation_response(request, chat_model)
 
     return app
 
 
-if __name__ == "__main__":
-    chat_model = ChatModel(*get_infer_args())
+def run_api() -> None:
+    chat_model = ChatModel()
     app = create_app(chat_model)
-    uvicorn.run(app, host="0.0.0.0", port=8000, workers=1)
+    api_host = os.environ.get("API_HOST", "0.0.0.0")
+    api_port = int(os.environ.get("API_PORT", "8000"))
+    print("Visit http://localhost:{}/docs for API document.".format(api_port))
+    uvicorn.run(app, host=api_host, port=api_port)

src/llmtuner/api/chat.py

@@ -0,0 +1,186 @@
+import json
+import uuid
+from typing import TYPE_CHECKING, AsyncGenerator, Dict, List, Optional, Tuple
+
+from ..data import Role as DataRole
+from ..extras.logging import get_logger
+from ..extras.packages import is_fastapi_available
+from .common import dictify, jsonify
+from .protocol import (
+    ChatCompletionMessage,
+    ChatCompletionResponse,
+    ChatCompletionResponseChoice,
+    ChatCompletionResponseUsage,
+    ChatCompletionStreamResponse,
+    ChatCompletionStreamResponseChoice,
+    Finish,
+    Function,
+    FunctionCall,
+    Role,
+    ScoreEvaluationResponse,
+)
+
+
+if is_fastapi_available():
+    from fastapi import HTTPException, status
+
+
+if TYPE_CHECKING:
+    from ..chat import ChatModel
+    from .protocol import ChatCompletionRequest, ScoreEvaluationRequest
+
+
+logger = get_logger(__name__)
+ROLE_MAPPING = {
+    Role.USER: DataRole.USER.value,
+    Role.ASSISTANT: DataRole.ASSISTANT.value,
+    Role.SYSTEM: DataRole.SYSTEM.value,
+    Role.FUNCTION: DataRole.FUNCTION.value,
+    Role.TOOL: DataRole.OBSERVATION.value,
+}
+
+
+def _process_request(request: "ChatCompletionRequest") -> Tuple[List[Dict[str, str]], str, str]:
+    logger.info("==== request ====\n{}".format(json.dumps(dictify(request), indent=2, ensure_ascii=False)))
+
+    if len(request.messages) == 0:
+        raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Invalid length")
+
+    if request.messages[0].role == Role.SYSTEM:
+        system = request.messages.pop(0).content
+    else:
+        system = ""
+
+    if len(request.messages) % 2 == 0:
+        raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Only supports u/a/u/a/u...")
+
+    input_messages = []
+    for i, message in enumerate(request.messages):
+        if i % 2 == 0 and message.role not in [Role.USER, Role.TOOL]:
+            raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Invalid role")
+        elif i % 2 == 1 and message.role not in [Role.ASSISTANT, Role.FUNCTION]:
+            raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Invalid role")
+
+        if message.role == Role.ASSISTANT and isinstance(message.tool_calls, list) and len(message.tool_calls):
+            name = message.tool_calls[0].function.name
+            arguments = message.tool_calls[0].function.arguments
+            content = json.dumps({"name": name, "argument": arguments}, ensure_ascii=False)
+            input_messages.append({"role": ROLE_MAPPING[Role.FUNCTION], "content": content})
+        else:
+            input_messages.append({"role": ROLE_MAPPING[message.role], "content": message.content})
+
+    tool_list = request.tools
+    if isinstance(tool_list, list) and len(tool_list):
+        try:
+            tools = json.dumps([dictify(tool.function) for tool in tool_list], ensure_ascii=False)
+        except Exception:
+            raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Invalid tools")
+    else:
+        tools = ""
+
+    return input_messages, system, tools
+
+
+def _create_stream_chat_completion_chunk(
+    completion_id: str,
+    model: str,
+    delta: "ChatCompletionMessage",
+    index: Optional[int] = 0,
+    finish_reason: Optional["Finish"] = None,
+) -> str:
+    choice_data = ChatCompletionStreamResponseChoice(index=index, delta=delta, finish_reason=finish_reason)
+    chunk = ChatCompletionStreamResponse(id=completion_id, model=model, choices=[choice_data])
+    return jsonify(chunk)
+
+
+async def create_chat_completion_response(
+    request: "ChatCompletionRequest", chat_model: "ChatModel"
+) -> "ChatCompletionResponse":
+    completion_id = "chatcmpl-{}".format(uuid.uuid4().hex)
+    input_messages, system, tools = _process_request(request)
+    responses = await chat_model.achat(
+        input_messages,
+        system,
+        tools,
+        do_sample=request.do_sample,
+        temperature=request.temperature,
+        top_p=request.top_p,
+        max_new_tokens=request.max_tokens,
+        num_return_sequences=request.n,
+        stop=request.stop,
+    )
+
+    prompt_length, response_length = 0, 0
+    choices = []
+    for i, response in enumerate(responses):
+        if tools:
+            result = chat_model.engine.template.format_tools.extract(response.response_text)
+        else:
+            result = response.response_text
+
+        if isinstance(result, tuple):
+            name, arguments = result
+            function = Function(name=name, arguments=arguments)
+            tool_call = FunctionCall(id="call_{}".format(uuid.uuid4().hex), function=function)
+            response_message = ChatCompletionMessage(role=Role.ASSISTANT, tool_calls=[tool_call])
+            finish_reason = Finish.TOOL
+        else:
+            response_message = ChatCompletionMessage(role=Role.ASSISTANT, content=result)
+            finish_reason = Finish.STOP if response.finish_reason == "stop" else Finish.LENGTH
+
+        choices.append(ChatCompletionResponseChoice(index=i, message=response_message, finish_reason=finish_reason))
+        prompt_length = response.prompt_length
+        response_length += response.response_length
+
+    usage = ChatCompletionResponseUsage(
+        prompt_tokens=prompt_length,
+        completion_tokens=response_length,
+        total_tokens=prompt_length + response_length,
+    )
+
+    return ChatCompletionResponse(id=completion_id, model=request.model, choices=choices, usage=usage)
+
+
+async def create_stream_chat_completion_response(
+    request: "ChatCompletionRequest", chat_model: "ChatModel"
+) -> AsyncGenerator[str, None]:
+    completion_id = "chatcmpl-{}".format(uuid.uuid4().hex)
+    input_messages, system, tools = _process_request(request)
+    if tools:
+        raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Cannot stream function calls.")
+
+    if request.n > 1:
+        raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Cannot stream multiple responses.")
+
+    yield _create_stream_chat_completion_chunk(
+        completion_id=completion_id, model=request.model, delta=ChatCompletionMessage(role=Role.ASSISTANT, content="")
+    )
+    async for new_token in chat_model.astream_chat(
+        input_messages,
+        system,
+        tools,
+        do_sample=request.do_sample,
+        temperature=request.temperature,
+        top_p=request.top_p,
+        max_new_tokens=request.max_tokens,
+        stop=request.stop,
+    ):
+        if len(new_token) != 0:
+            yield _create_stream_chat_completion_chunk(
+                completion_id=completion_id, model=request.model, delta=ChatCompletionMessage(content=new_token)
+            )
+
+    yield _create_stream_chat_completion_chunk(
+        completion_id=completion_id, model=request.model, delta=ChatCompletionMessage(), finish_reason=Finish.STOP
+    )
+    yield "[DONE]"
+
+
+async def create_score_evaluation_response(
+    request: "ScoreEvaluationRequest", chat_model: "ChatModel"
+) -> "ScoreEvaluationResponse":
+    if len(request.messages) == 0:
+        raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Invalid request")
+
+    scores = await chat_model.aget_scores(request.messages, max_length=request.max_length)
+    return ScoreEvaluationResponse(model=request.model, scores=scores)

src/llmtuner/api/common.py

@@ -0,0 +1,20 @@
+import json
+from typing import TYPE_CHECKING, Any, Dict
+
+
+if TYPE_CHECKING:
+    from pydantic import BaseModel
+
+
+def dictify(data: "BaseModel") -> Dict[str, Any]:
+    try:  # pydantic v2
+        return data.model_dump(exclude_unset=True)
+    except AttributeError:  # pydantic v1
+        return data.dict(exclude_unset=True)
+
+
+def jsonify(data: "BaseModel") -> str:
+    try:  # pydantic v2
+        return json.dumps(data.model_dump(exclude_unset=True), ensure_ascii=False)
+    except AttributeError:  # pydantic v1
+        return data.json(exclude_unset=True, ensure_ascii=False)

src/llmtuner/api/protocol.py

@@ -1,64 +1,95 @@
 import time
-from enum import Enum
+from enum import Enum, unique
+from typing import Any, Dict, List, Optional, Union
+
 from pydantic import BaseModel, Field
-from typing import List, Optional
+from typing_extensions import Literal
 
 
+@unique
 class Role(str, Enum):
     USER = "user"
     ASSISTANT = "assistant"
     SYSTEM = "system"
+    FUNCTION = "function"
+    TOOL = "tool"
 
 
+@unique
 class Finish(str, Enum):
     STOP = "stop"
     LENGTH = "length"
+    TOOL = "tool_calls"
 
 
 class ModelCard(BaseModel):
     id: str
-    object: Optional[str] = "model"
-    created: Optional[int] = Field(default_factory=lambda: int(time.time()))
-    owned_by: Optional[str] = "owner"
-    root: Optional[str] = None
-    parent: Optional[str] = None
-    permission: Optional[list] = []
+    object: Literal["model"] = "model"
+    created: int = Field(default_factory=lambda: int(time.time()))
+    owned_by: Literal["owner"] = "owner"
 
 
 class ModelList(BaseModel):
-    object: Optional[str] = "list"
-    data: Optional[List[ModelCard]] = []
+    object: Literal["list"] = "list"
+    data: List[ModelCard] = []
+
+
+class Function(BaseModel):
+    name: str
+    arguments: str
+
+
+class FunctionDefinition(BaseModel):
+    name: str
+    description: str
+    parameters: Dict[str, Any]
+
+
+class FunctionAvailable(BaseModel):
+    type: Literal["function", "code_interpreter"] = "function"
+    function: Optional[FunctionDefinition] = None
+
+
+class FunctionCall(BaseModel):
+    id: str
+    type: Literal["function"] = "function"
+    function: Function
 
 
 class ChatMessage(BaseModel):
     role: Role
-    content: str
+    content: Optional[str] = None
+    tool_calls: Optional[List[FunctionCall]] = None
 
 
-class DeltaMessage(BaseModel):
+class ChatCompletionMessage(BaseModel):
     role: Optional[Role] = None
     content: Optional[str] = None
+    tool_calls: Optional[List[FunctionCall]] = None
 
 
 class ChatCompletionRequest(BaseModel):
     model: str
     messages: List[ChatMessage]
+    tools: Optional[List[FunctionAvailable]] = None
+    do_sample: bool = True
     temperature: Optional[float] = None
     top_p: Optional[float] = None
-    n: Optional[int] = 1
+    n: int = 1
     max_tokens: Optional[int] = None
-    stream: Optional[bool] = False
+    stop: Optional[Union[str, List[str]]] = None
+    stream: bool = False
 
 
 class ChatCompletionResponseChoice(BaseModel):
     index: int
-    message: ChatMessage
+    message: ChatCompletionMessage
     finish_reason: Finish
 
 
-class ChatCompletionResponseStreamChoice(BaseModel):
+class ChatCompletionStreamResponseChoice(BaseModel):
     index: int
-    delta: DeltaMessage
+    delta: ChatCompletionMessage
     finish_reason: Optional[Finish] = None
 
 
@@ -69,17 +100,30 @@ class ChatCompletionResponseUsage(BaseModel):
 
 
 class ChatCompletionResponse(BaseModel):
-    id: Optional[str] = "chatcmpl-default"
-    object: Optional[str] = "chat.completion"
-    created: Optional[int] = Field(default_factory=lambda: int(time.time()))
+    id: str
+    object: Literal["chat.completion"] = "chat.completion"
+    created: int = Field(default_factory=lambda: int(time.time()))
     model: str
     choices: List[ChatCompletionResponseChoice]
     usage: ChatCompletionResponseUsage
 
 
 class ChatCompletionStreamResponse(BaseModel):
-    id: Optional[str] = "chatcmpl-default"
-    object: Optional[str] = "chat.completion.chunk"
-    created: Optional[int] = Field(default_factory=lambda: int(time.time()))
+    id: str
+    object: Literal["chat.completion.chunk"] = "chat.completion.chunk"
+    created: int = Field(default_factory=lambda: int(time.time()))
     model: str
-    choices: List[ChatCompletionResponseStreamChoice]
+    choices: List[ChatCompletionStreamResponseChoice]
+
+
+class ScoreEvaluationRequest(BaseModel):
+    model: str
+    messages: List[str]
+    max_length: Optional[int] = None
+
+
+class ScoreEvaluationResponse(BaseModel):
+    id: str
+    object: Literal["score.evaluation"] = "score.evaluation"
+    model: str
+    scores: List[float]

src/llmtuner/chat/__init__.py

@@ -1 +1,5 @@
-from llmtuner.chat.stream_chat import ChatModel
+from .base_engine import BaseEngine
+from .chat_model import ChatModel
+
+
+__all__ = ["BaseEngine", "ChatModel"]