Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- .bandit +3 -0
- .editorconfig +14 -0
- .flake8 +5 -0
- .gitattributes +44 -35
- .github/CODE_OF_CONDUCT.md +129 -0
- .github/CONTRIBUTING.md +76 -0
- .github/FUNDING.yml +13 -0
- .github/ISSUE_TEMPLATE/bug-report.yaml +113 -0
- .github/ISSUE_TEMPLATE/config.yml +7 -0
- .github/ISSUE_TEMPLATE/docs.yml +46 -0
- .github/ISSUE_TEMPLATE/feature-request.yaml +63 -0
- .github/PULL_REQUEST_TEMPLATE.md +27 -0
- .github/SECURITY.md +9 -0
- .github/SUPPORT.md +10 -0
- .github/release-drafter.yml +31 -0
- .github/workflows/base.yml +58 -0
- .github/workflows/docs.yml +31 -0
- .github/workflows/lint.yml +22 -0
- .github/workflows/main.yml +119 -0
- .github/workflows/pypi.yml +45 -0
- .github/workflows/tests.yml +100 -0
- .gitignore +177 -0
- .isort.cfg +3 -0
- .mypy.ini +51 -0
- .pre-commit-config.yaml +43 -0
- .pylintrc +14 -0
- FAQS.md +7 -0
- LICENSE +202 -0
- README.md +1356 -0
- TODO.md +10 -0
- _quarto.yml +51 -0
- axol/.gitignore +2 -0
- axol/bin/activate +87 -0
- axol/bin/activate.csh +55 -0
- axol/bin/activate.fish +103 -0
- axol/bin/activate.nu +96 -0
- axol/bin/activate.ps1 +61 -0
- axol/bin/activate_this.py +37 -0
- axol/bin/pip +8 -0
- axol/bin/pip-3.10 +8 -0
- axol/bin/pip3 +8 -0
- axol/bin/pip3.10 +8 -0
- axol/bin/python +3 -0
- axol/bin/python3 +3 -0
- axol/bin/python3.10 +3 -0
- axol/bin/wheel +8 -0
- axol/bin/wheel-3.10 +8 -0
- axol/bin/wheel3 +8 -0
- axol/bin/wheel3.10 +8 -0
- axol/pyvenv.cfg +8 -0
.bandit
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
[bandit]
|
2 |
+
exclude = tests
|
3 |
+
skips = B101
|
.editorconfig
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
root = true
|
2 |
+
|
3 |
+
[*]
|
4 |
+
end_of_line = lf
|
5 |
+
insert_final_newline = true
|
6 |
+
trim_trailing_whitespace = true
|
7 |
+
|
8 |
+
[*.py]
|
9 |
+
indent_style = space
|
10 |
+
indent_size = 4
|
11 |
+
|
12 |
+
[**.yml]
|
13 |
+
indent_style = space
|
14 |
+
indent_size = 2
|
.flake8
ADDED
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[flake8]
|
2 |
+
max-line-length = 88
|
3 |
+
|
4 |
+
select = C,E,F,W,B,B950
|
5 |
+
extend-ignore = E203, E501, W503
|
.gitattributes
CHANGED
@@ -1,35 +1,44 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
data/*.jsonl filter=lfs diff=lfs merge=lfs -text
|
2 |
+
axol/bin/python filter=lfs diff=lfs merge=lfs -text
|
3 |
+
axol/bin/python3 filter=lfs diff=lfs merge=lfs -text
|
4 |
+
axol/bin/python3.10 filter=lfs diff=lfs merge=lfs -text
|
5 |
+
git-lfs-3.2.0/git-lfs filter=lfs diff=lfs merge=lfs -text
|
6 |
+
git-lfs-linux-amd64-v3.2.0.tar.gz filter=lfs diff=lfs merge=lfs -text
|
7 |
+
out/adapter_model.bin filter=lfs diff=lfs merge=lfs -text
|
8 |
+
out/adapter_model.safetensors filter=lfs diff=lfs merge=lfs -text
|
9 |
+
out/checkpoint-126/adapter_model.safetensors filter=lfs diff=lfs merge=lfs -text
|
10 |
+
out/checkpoint-126/optimizer.pt filter=lfs diff=lfs merge=lfs -text
|
11 |
+
out/checkpoint-126/rng_state_0.pth filter=lfs diff=lfs merge=lfs -text
|
12 |
+
out/checkpoint-126/rng_state_1.pth filter=lfs diff=lfs merge=lfs -text
|
13 |
+
out/checkpoint-126/scheduler.pt filter=lfs diff=lfs merge=lfs -text
|
14 |
+
out/checkpoint-126/tokenizer.model filter=lfs diff=lfs merge=lfs -text
|
15 |
+
out/checkpoint-126/training_args.bin filter=lfs diff=lfs merge=lfs -text
|
16 |
+
out/checkpoint-189/adapter_model.safetensors filter=lfs diff=lfs merge=lfs -text
|
17 |
+
out/checkpoint-189/optimizer.pt filter=lfs diff=lfs merge=lfs -text
|
18 |
+
out/checkpoint-189/rng_state_0.pth filter=lfs diff=lfs merge=lfs -text
|
19 |
+
out/checkpoint-189/rng_state_1.pth filter=lfs diff=lfs merge=lfs -text
|
20 |
+
out/checkpoint-189/scheduler.pt filter=lfs diff=lfs merge=lfs -text
|
21 |
+
out/checkpoint-189/tokenizer.model filter=lfs diff=lfs merge=lfs -text
|
22 |
+
out/checkpoint-189/training_args.bin filter=lfs diff=lfs merge=lfs -text
|
23 |
+
out/checkpoint-230/adapter_model.safetensors filter=lfs diff=lfs merge=lfs -text
|
24 |
+
out/checkpoint-230/optimizer.pt filter=lfs diff=lfs merge=lfs -text
|
25 |
+
out/checkpoint-230/rng_state_0.pth filter=lfs diff=lfs merge=lfs -text
|
26 |
+
out/checkpoint-230/rng_state_1.pth filter=lfs diff=lfs merge=lfs -text
|
27 |
+
out/checkpoint-230/scheduler.pt filter=lfs diff=lfs merge=lfs -text
|
28 |
+
out/checkpoint-230/tokenizer.model filter=lfs diff=lfs merge=lfs -text
|
29 |
+
out/checkpoint-230/training_args.bin filter=lfs diff=lfs merge=lfs -text
|
30 |
+
out/checkpoint-345/adapter_model.safetensors filter=lfs diff=lfs merge=lfs -text
|
31 |
+
out/checkpoint-345/optimizer.pt filter=lfs diff=lfs merge=lfs -text
|
32 |
+
out/checkpoint-345/rng_state_0.pth filter=lfs diff=lfs merge=lfs -text
|
33 |
+
out/checkpoint-345/rng_state_1.pth filter=lfs diff=lfs merge=lfs -text
|
34 |
+
out/checkpoint-345/scheduler.pt filter=lfs diff=lfs merge=lfs -text
|
35 |
+
out/checkpoint-345/tokenizer.model filter=lfs diff=lfs merge=lfs -text
|
36 |
+
out/checkpoint-345/training_args.bin filter=lfs diff=lfs merge=lfs -text
|
37 |
+
out/merged/git-lfs-3.2.0/git-lfs filter=lfs diff=lfs merge=lfs -text
|
38 |
+
out/merged/git-lfs-linux-amd64-v3.2.0.tar.gz.1 filter=lfs diff=lfs merge=lfs -text
|
39 |
+
out/merged/pytorch_model-00001-of-00003.bin filter=lfs diff=lfs merge=lfs -text
|
40 |
+
out/merged/pytorch_model-00002-of-00003.bin filter=lfs diff=lfs merge=lfs -text
|
41 |
+
out/merged/pytorch_model-00003-of-00003.bin filter=lfs diff=lfs merge=lfs -text
|
42 |
+
out/merged/tokenizer.model filter=lfs diff=lfs merge=lfs -text
|
43 |
+
out/tokenizer.model filter=lfs diff=lfs merge=lfs -text
|
44 |
+
out/training_args.bin filter=lfs diff=lfs merge=lfs -text
|
.github/CODE_OF_CONDUCT.md
ADDED
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Contributor Covenant Code of Conduct
|
2 |
+
|
3 |
+
## Our Pledge
|
4 |
+
|
5 |
+
We as members, contributors, and leaders pledge to make participation in our
|
6 |
+
community a harassment-free experience for everyone, regardless of age, body
|
7 |
+
size, visible or invisible disability, ethnicity, sex characteristics, gender
|
8 |
+
identity and expression, level of experience, education, socio-economic status,
|
9 |
+
nationality, personal appearance, race, religion, or sexual identity
|
10 |
+
and orientation.
|
11 |
+
|
12 |
+
We pledge to act and interact in ways that contribute to an open, welcoming,
|
13 |
+
diverse, inclusive, and healthy community.
|
14 |
+
|
15 |
+
## Our Standards
|
16 |
+
|
17 |
+
Examples of behavior that contributes to a positive environment for our
|
18 |
+
community include:
|
19 |
+
|
20 |
+
* Demonstrating empathy and kindness toward other people
|
21 |
+
* Being respectful of differing opinions, viewpoints, and experiences
|
22 |
+
* Giving and gracefully accepting constructive feedback
|
23 |
+
* Accepting responsibility and apologizing to those affected by our mistakes,
|
24 |
+
and learning from the experience
|
25 |
+
* Focusing on what is best not just for us as individuals, but for the
|
26 |
+
overall community
|
27 |
+
|
28 |
+
Examples of unacceptable behavior include:
|
29 |
+
|
30 |
+
* The use of sexualized language or imagery, and sexual attention or
|
31 |
+
advances of any kind
|
32 |
+
* Trolling, insulting or derogatory comments, and personal or political attacks
|
33 |
+
* Public or private harassment
|
34 |
+
* Publishing others' private information, such as a physical or email
|
35 |
+
address, without their explicit permission
|
36 |
+
* Other conduct which could reasonably be considered inappropriate in a
|
37 |
+
professional setting
|
38 |
+
|
39 |
+
## Enforcement Responsibilities
|
40 |
+
|
41 |
+
Community leaders are responsible for clarifying and enforcing our standards of
|
42 |
+
acceptable behavior and will take appropriate and fair corrective action in
|
43 |
+
response to any behavior that they deem inappropriate, threatening, offensive,
|
44 |
+
or harmful.
|
45 |
+
|
46 |
+
Community leaders have the right and responsibility to remove, edit, or reject
|
47 |
+
comments, commits, code, wiki edits, issues, and other contributions that are
|
48 |
+
not aligned to this Code of Conduct, and will communicate reasons for moderation
|
49 |
+
decisions when appropriate.
|
50 |
+
|
51 |
+
## Scope
|
52 |
+
|
53 |
+
This Code of Conduct applies within all community spaces, and also applies when
|
54 |
+
an individual is officially representing the community in public spaces.
|
55 |
+
Examples of representing our community include using an official e-mail address,
|
56 |
+
posting via an official social media account, or acting as an appointed
|
57 |
+
representative at an online or offline event.
|
58 |
+
|
59 |
+
## Enforcement
|
60 |
+
|
61 |
+
Instances of abusive, harassing, or otherwise unacceptable behavior may be
|
62 |
+
reported to the community leaders responsible for enforcement on Discord
|
63 |
+
at https://discord.gg/QYF8QrtEUm
|
64 |
+
|
65 |
+
All complaints will be reviewed and investigated promptly and fairly.
|
66 |
+
|
67 |
+
All community leaders are obligated to respect the privacy and security of the
|
68 |
+
reporter of any incident.
|
69 |
+
|
70 |
+
## Enforcement Guidelines
|
71 |
+
|
72 |
+
Community leaders will follow these Community Impact Guidelines in determining
|
73 |
+
the consequences for any action they deem in violation of this Code of Conduct:
|
74 |
+
|
75 |
+
### 1. Correction
|
76 |
+
|
77 |
+
**Community Impact**: Use of inappropriate language or other behavior deemed
|
78 |
+
unprofessional or unwelcome in the community.
|
79 |
+
|
80 |
+
**Consequence**: A private, written warning from community leaders, providing
|
81 |
+
clarity around the nature of the violation and an explanation of why the
|
82 |
+
behavior was inappropriate. A public apology may be requested.
|
83 |
+
|
84 |
+
### 2. Warning
|
85 |
+
|
86 |
+
**Community Impact**: A violation through a single incident or series
|
87 |
+
of actions.
|
88 |
+
|
89 |
+
**Consequence**: A warning with consequences for continued behavior. No
|
90 |
+
interaction with the people involved, including unsolicited interaction with
|
91 |
+
those enforcing the Code of Conduct, for a specified period of time. This
|
92 |
+
includes avoiding interactions in community spaces as well as external channels
|
93 |
+
like social media. Violating these terms may lead to a temporary or
|
94 |
+
permanent ban.
|
95 |
+
|
96 |
+
### 3. Temporary Ban
|
97 |
+
|
98 |
+
**Community Impact**: A serious violation of community standards, including
|
99 |
+
sustained inappropriate behavior.
|
100 |
+
|
101 |
+
**Consequence**: A temporary ban from any sort of interaction or public
|
102 |
+
communication with the community for a specified period of time. No public or
|
103 |
+
private interaction with the people involved, including unsolicited interaction
|
104 |
+
with those enforcing the Code of Conduct, is allowed during this period.
|
105 |
+
Violating these terms may lead to a permanent ban.
|
106 |
+
|
107 |
+
### 4. Permanent Ban
|
108 |
+
|
109 |
+
**Community Impact**: Demonstrating a pattern of violation of community
|
110 |
+
standards, including sustained inappropriate behavior, harassment of an
|
111 |
+
individual, or aggression toward or disparagement of classes of individuals.
|
112 |
+
|
113 |
+
**Consequence**: A permanent ban from any sort of public interaction within
|
114 |
+
the community.
|
115 |
+
|
116 |
+
## Attribution
|
117 |
+
|
118 |
+
This Code of Conduct is adapted from the [Contributor Covenant][homepage],
|
119 |
+
version 2.0, available at
|
120 |
+
https://www.contributor-covenant.org/version/2/0/code_of_conduct.html.
|
121 |
+
|
122 |
+
Community Impact Guidelines were inspired by [Mozilla's code of conduct
|
123 |
+
enforcement ladder](https://github.com/mozilla/diversity).
|
124 |
+
|
125 |
+
[homepage]: https://www.contributor-covenant.org
|
126 |
+
|
127 |
+
For answers to common questions about this code of conduct, see the FAQ at
|
128 |
+
https://www.contributor-covenant.org/faq. Translations are available at
|
129 |
+
https://www.contributor-covenant.org/translations.
|
.github/CONTRIBUTING.md
ADDED
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Contributing to axolotl
|
2 |
+
|
3 |
+
First of all, thank you for your interest in contributing to axolotl! We appreciate the time and effort you're willing to invest in making our project better. This document provides guidelines and information to make the contribution process as smooth as possible.
|
4 |
+
|
5 |
+
## Table of Contents
|
6 |
+
|
7 |
+
- [Code of Conduct](#code-of-conduct)
|
8 |
+
- [Getting Started](#getting-started)
|
9 |
+
- [How to Contribute](#how-to-contribute)
|
10 |
+
- [Reporting Bugs](#reporting-bugs)
|
11 |
+
- [Suggesting Enhancements](#suggesting-enhancements)
|
12 |
+
- [Submitting Pull Requests](#submitting-pull-requests)
|
13 |
+
- [Style Guidelines](#style-guidelines)
|
14 |
+
- [Code Style](#code-style)
|
15 |
+
- [Commit Messages](#commit-messages)
|
16 |
+
- [Additional Resources](#additional-resources)
|
17 |
+
|
18 |
+
## Code of Conductcode
|
19 |
+
|
20 |
+
All contributors are expected to adhere to our [Code of Conduct](CODE_OF_CONDUCT.md). Please read it before participating in the axolotl community.
|
21 |
+
|
22 |
+
## Getting Started
|
23 |
+
|
24 |
+
Bugs? Please check for open issue else create a new [Issue](https://github.com/OpenAccess-AI-Collective/axolotl/issues/new).
|
25 |
+
|
26 |
+
PRs are **greatly welcome**!
|
27 |
+
|
28 |
+
1. Fork the repository and clone it to your local machine.
|
29 |
+
2. Set up the development environment by following the instructions in the [README.md](https://github.com/OpenAccess-AI-Collective/axolotl/tree/main/README.md) file.
|
30 |
+
3. Explore the codebase, run tests, and verify that everything works as expected.
|
31 |
+
|
32 |
+
Please run below to setup env
|
33 |
+
```bash
|
34 |
+
pip3 install -r requirements-dev.txt -r requirements-tests.txt
|
35 |
+
pre-commit install
|
36 |
+
|
37 |
+
# test
|
38 |
+
pytest tests/
|
39 |
+
```
|
40 |
+
|
41 |
+
## How to Contribute
|
42 |
+
|
43 |
+
### Reporting Bugs
|
44 |
+
|
45 |
+
If you encounter a bug or issue while using axolotl, please open a new issue on the [GitHub Issues](https://github.com/OpenAccess-AI-Collective/axolotl/issues) page. Provide a clear and concise description of the problem, steps to reproduce it, and any relevant error messages or logs.
|
46 |
+
|
47 |
+
### Suggesting Enhancements
|
48 |
+
|
49 |
+
We welcome ideas for improvements and new features. To suggest an enhancement, open a new issue on the [GitHub Issues](https://github.com/OpenAccess-AI-Collective/axolotl/issues) page. Describe the enhancement in detail, explain the use case, and outline the benefits it would bring to the project.
|
50 |
+
|
51 |
+
### Submitting Pull Requests
|
52 |
+
|
53 |
+
1. Create a new branch for your feature or bugfix. Use a descriptive name like `feature/your-feature-name` or `fix/your-bugfix-name`.
|
54 |
+
2. Make your changes, following the [Style Guidelines](#style-guidelines) below.
|
55 |
+
3. Test your changes and ensure that they don't introduce new issues or break existing functionality.
|
56 |
+
4. Commit your changes, following the [commit message guidelines](#commit-messages).
|
57 |
+
5. Push your branch to your fork on GitHub.
|
58 |
+
6. Open a new pull request against the `main` branch of the axolotl repository. Include a clear and concise description of your changes, referencing any related issues.
|
59 |
+
|
60 |
+
## Style Guidelines
|
61 |
+
|
62 |
+
### Code Style
|
63 |
+
|
64 |
+
axolotl uses [{codestyle}]({URLofCodestyle}) as its code style guide. Please ensure that your code follows these guidelines.
|
65 |
+
|
66 |
+
### Commit Messages
|
67 |
+
|
68 |
+
Write clear and concise commit messages that briefly describe the changes made in each commit. Use the imperative mood and start with a capitalized verb, e.g., "Add new feature" or "Fix bug in function".
|
69 |
+
|
70 |
+
## Additional Resources
|
71 |
+
|
72 |
+
- [GitHub Help](https://help.github.com/)
|
73 |
+
- [GitHub Pull Request Documentation](https://docs.github.com/en/github/collaborating-with-issues-and-pull-requests)
|
74 |
+
- [{codestyle}]({URLofCodestyle})
|
75 |
+
|
76 |
+
Thank you once again for your interest in contributing to axolotl. We look forward to collaborating with you and creating an even better project together!
|
.github/FUNDING.yml
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# These are supported funding model platforms
|
2 |
+
|
3 |
+
github: [winglian, OpenAccess-AI-Collective] # Replace with up to 4 GitHub Sponsors-enabled usernames e.g., [user1, user2]
|
4 |
+
patreon: # Replace with a single Patreon username
|
5 |
+
open_collective: # Replace with a single Open Collective username
|
6 |
+
ko_fi: axolotl_ai # Replace with a single Ko-fi username
|
7 |
+
tidelift: # Replace with a single Tidelift platform-name/package-name e.g., npm/babel
|
8 |
+
community_bridge: # Replace with a single Community Bridge project-name e.g., cloud-foundry
|
9 |
+
liberapay: # Replace with a single Liberapay username
|
10 |
+
issuehunt: # Replace with a single IssueHunt username
|
11 |
+
otechie: # Replace with a single Otechie username
|
12 |
+
lfx_crowdfunding: # Replace with a single LFX Crowdfunding project-name e.g., cloud-foundry
|
13 |
+
custom: ['https://quickchart.io/qr?text=bitcoin%3Abc1qxlgwlqwfea5s2cxm42xqsfmwjct0rj8w8ea5np&size=480¢erImageUrl=https%3A%2F%2Fupload.wikimedia.org%2Fwikipedia%2Fcommons%2Fthumb%2F4%2F46%2FBitcoin.svg%2F64px-Bitcoin.svg.png'] # Replace with up to 4 custom sponsorship URLs e.g., ['link1', 'link2']
|
.github/ISSUE_TEMPLATE/bug-report.yaml
ADDED
@@ -0,0 +1,113 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name: Bug Report
|
2 |
+
description: File a bug report
|
3 |
+
labels: ["bug", "needs triage"]
|
4 |
+
body:
|
5 |
+
- type: markdown
|
6 |
+
attributes:
|
7 |
+
value: |
|
8 |
+
## Before you start
|
9 |
+
Please **make sure you are on the latest version.**
|
10 |
+
If you encountered the issue after you installed, updated, or reloaded, **please try restarting before reporting the bug**.
|
11 |
+
|
12 |
+
- type: checkboxes
|
13 |
+
id: no-duplicate-issues
|
14 |
+
attributes:
|
15 |
+
label: "Please check that this issue hasn't been reported before."
|
16 |
+
description: "The **Label filters** may help make your search more focussed."
|
17 |
+
options:
|
18 |
+
- label: "I searched previous [Bug Reports](https://github.com/OpenAccess-AI-Collective/axolotl/labels/bug) didn't find any similar reports."
|
19 |
+
required: true
|
20 |
+
|
21 |
+
- type: textarea
|
22 |
+
id: expected
|
23 |
+
attributes:
|
24 |
+
label: Expected Behavior
|
25 |
+
description: Tell us what **should** happen.
|
26 |
+
validations:
|
27 |
+
required: true
|
28 |
+
|
29 |
+
- type: textarea
|
30 |
+
id: what-happened
|
31 |
+
attributes:
|
32 |
+
label: Current behaviour
|
33 |
+
description: |
|
34 |
+
Tell us what happens instead of the expected behavior.
|
35 |
+
Provide stacktrace and/or screenshots.
|
36 |
+
validations:
|
37 |
+
required: true
|
38 |
+
|
39 |
+
- type: textarea
|
40 |
+
id: reproduce
|
41 |
+
attributes:
|
42 |
+
label: Steps to reproduce
|
43 |
+
description: |
|
44 |
+
Which exact steps can a developer take to reproduce the issue?
|
45 |
+
The more detail you provide, the easier it will be to narrow down and fix the bug.
|
46 |
+
Please paste in tasks and/or queries **as text, not screenshots**.
|
47 |
+
placeholder: |
|
48 |
+
Example of the level of detail needed to reproduce any bugs efficiently and reliably.
|
49 |
+
1. Go to the '...' page.
|
50 |
+
2. Click on the '...' button.
|
51 |
+
3. Scroll down to '...'.
|
52 |
+
4. Observe the error.
|
53 |
+
validations:
|
54 |
+
required: true
|
55 |
+
|
56 |
+
- type: textarea
|
57 |
+
id: config
|
58 |
+
attributes:
|
59 |
+
label: Config yaml
|
60 |
+
description: |
|
61 |
+
Please attach the config yaml!
|
62 |
+
render: yaml
|
63 |
+
|
64 |
+
- type: textarea
|
65 |
+
id: possible-solution
|
66 |
+
attributes:
|
67 |
+
label: Possible solution
|
68 |
+
description: |
|
69 |
+
Not obligatory, but please suggest a fix or reason for the bug, if you have an idea.
|
70 |
+
|
71 |
+
|
72 |
+
- type: checkboxes
|
73 |
+
id: operating-systems
|
74 |
+
attributes:
|
75 |
+
label: Which Operating Systems are you using?
|
76 |
+
description: You may select more than one.
|
77 |
+
options:
|
78 |
+
- label: Linux
|
79 |
+
- label: macOS
|
80 |
+
- label: Windows
|
81 |
+
|
82 |
+
- type: input
|
83 |
+
id: Python-version
|
84 |
+
attributes:
|
85 |
+
label: Python Version
|
86 |
+
description: Which {Programming} version are you using?
|
87 |
+
placeholder: 3.10 / please change accordingly
|
88 |
+
validations:
|
89 |
+
required: true
|
90 |
+
|
91 |
+
- type: input
|
92 |
+
id: axolotl-branch-commit
|
93 |
+
attributes:
|
94 |
+
label: axolotl branch-commit
|
95 |
+
description: On which branch/commit are you?
|
96 |
+
placeholder: main/4d6490b
|
97 |
+
validations:
|
98 |
+
required: true
|
99 |
+
|
100 |
+
- type: checkboxes
|
101 |
+
id: acknowledgements
|
102 |
+
attributes:
|
103 |
+
label: 'Acknowledgements'
|
104 |
+
description: 'Please confirm the following:'
|
105 |
+
options:
|
106 |
+
- label: 'My issue title is concise, descriptive, and in title casing.'
|
107 |
+
required: true
|
108 |
+
- label: 'I have searched the existing issues to make sure this bug has not been reported yet.'
|
109 |
+
required: true
|
110 |
+
- label: 'I am using the latest version of axolotl.'
|
111 |
+
required: true
|
112 |
+
- label: 'I have provided enough information for the maintainers to reproduce and diagnose the issue.'
|
113 |
+
required: true
|
.github/ISSUE_TEMPLATE/config.yml
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
blank_issues_enabled: false
|
2 |
+
contact_links:
|
3 |
+
- name: Ask a question
|
4 |
+
url: https://github.com/OpenAccess-AI-Collective/axolotl/discussions/categories/q-a
|
5 |
+
about: Ask questions and discuss with other community members
|
6 |
+
- name: Discuss the Project in Discord
|
7 |
+
url: https://discord.gg/HhrNrHJPRb
|
.github/ISSUE_TEMPLATE/docs.yml
ADDED
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name: Documentation Improvement / Clarity
|
2 |
+
description: Make a suggestion to improve the project documentation.
|
3 |
+
labels: ['needs triage', 'docs']
|
4 |
+
body:
|
5 |
+
- type: markdown
|
6 |
+
attributes:
|
7 |
+
value: '## :book: Documentation :book:'
|
8 |
+
- type: markdown
|
9 |
+
attributes:
|
10 |
+
value: |
|
11 |
+
* Ask questions in [Discord](https://discord.gg/HhrNrHJPRb).
|
12 |
+
* Before you file an issue read the [Contributing guide](./CONTRIBUTING.md).
|
13 |
+
* Check to make sure someone hasn't already opened a [similar issue](https://github.com/OpenAccess-AI-Collective/axolotl/issues).
|
14 |
+
- type: textarea
|
15 |
+
attributes:
|
16 |
+
label: What piece of documentation is affected?
|
17 |
+
description: Please link to the article you'd like to see updated.
|
18 |
+
validations:
|
19 |
+
required: true
|
20 |
+
- type: textarea
|
21 |
+
attributes:
|
22 |
+
label: What part(s) of the article would you like to see updated?
|
23 |
+
description: |
|
24 |
+
- Give as much detail as you can to help us understand the change you want to see.
|
25 |
+
- Why should the docs be changed? What use cases does it support?
|
26 |
+
- What is the expected outcome?
|
27 |
+
validations:
|
28 |
+
required: true
|
29 |
+
- type: textarea
|
30 |
+
attributes:
|
31 |
+
label: Additional Information
|
32 |
+
description: Add any other context or screenshots about the feature request here.
|
33 |
+
validations:
|
34 |
+
required: false
|
35 |
+
- type: checkboxes
|
36 |
+
id: acknowledgements
|
37 |
+
attributes:
|
38 |
+
label: 'Acknowledgements'
|
39 |
+
description: 'Please confirm the following:'
|
40 |
+
options:
|
41 |
+
- label: 'My issue title is concise, descriptive, and in title casing.'
|
42 |
+
required: true
|
43 |
+
- label: 'I have searched the existing issues to make sure this feature has not been requested yet.'
|
44 |
+
required: true
|
45 |
+
- label: 'I have provided enough information for the maintainers to understand and evaluate this request.'
|
46 |
+
required: true
|
.github/ISSUE_TEMPLATE/feature-request.yaml
ADDED
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name: Feature Request / Enhancement
|
2 |
+
description: Suggest a new feature or feature enhancement for the project
|
3 |
+
labels: ["enhancement", "needs triage"]
|
4 |
+
body:
|
5 |
+
- type: checkboxes
|
6 |
+
id: no-duplicate-issues
|
7 |
+
attributes:
|
8 |
+
label: "⚠️ Please check that this feature request hasn't been suggested before."
|
9 |
+
description: "There are two locations for previous feature requests. Please search in both. Thank you. The **Label filters** may help make your search more focussed."
|
10 |
+
options:
|
11 |
+
- label: "I searched previous [Ideas in Discussions](https://github.com/OpenAccess-AI-Collective/axolotl/discussions/categories/ideas) didn't find any similar feature requests."
|
12 |
+
required: true
|
13 |
+
- label: "I searched previous [Issues](https://github.com/OpenAccess-AI-Collective/axolotl/labels/enhancement) didn't find any similar feature requests."
|
14 |
+
required: true
|
15 |
+
|
16 |
+
- type: textarea
|
17 |
+
id: feature-description
|
18 |
+
validations:
|
19 |
+
required: true
|
20 |
+
attributes:
|
21 |
+
label: "🔖 Feature description"
|
22 |
+
description: "A clear and concise description of what the feature request is."
|
23 |
+
placeholder: "You should add ..."
|
24 |
+
|
25 |
+
- type: textarea
|
26 |
+
id: solution
|
27 |
+
validations:
|
28 |
+
required: true
|
29 |
+
attributes:
|
30 |
+
label: "✔️ Solution"
|
31 |
+
description: "A clear and concise description of what you want to happen, and why."
|
32 |
+
placeholder: "In my use-case, ..."
|
33 |
+
|
34 |
+
- type: textarea
|
35 |
+
id: alternatives
|
36 |
+
validations:
|
37 |
+
required: false
|
38 |
+
attributes:
|
39 |
+
label: "❓ Alternatives"
|
40 |
+
description: "A clear and concise description of any alternative solutions or features you've considered."
|
41 |
+
placeholder: "I have considered ..."
|
42 |
+
|
43 |
+
- type: textarea
|
44 |
+
id: additional-context
|
45 |
+
validations:
|
46 |
+
required: false
|
47 |
+
attributes:
|
48 |
+
label: "📝 Additional Context"
|
49 |
+
description: "Add any other context or screenshots about the feature request here."
|
50 |
+
placeholder: "..."
|
51 |
+
|
52 |
+
- type: checkboxes
|
53 |
+
id: acknowledgements
|
54 |
+
attributes:
|
55 |
+
label: 'Acknowledgements'
|
56 |
+
description: 'Please confirm the following:'
|
57 |
+
options:
|
58 |
+
- label: 'My issue title is concise, descriptive, and in title casing.'
|
59 |
+
required: true
|
60 |
+
- label: 'I have searched the existing issues to make sure this feature has not been requested yet.'
|
61 |
+
required: true
|
62 |
+
- label: 'I have provided enough information for the maintainers to understand and evaluate this request.'
|
63 |
+
required: true
|
.github/PULL_REQUEST_TEMPLATE.md
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
<!--- Provide a general summary of your changes in the Title above -->
|
2 |
+
|
3 |
+
# Description
|
4 |
+
|
5 |
+
<!--- Describe your changes in detail -->
|
6 |
+
|
7 |
+
## Motivation and Context
|
8 |
+
|
9 |
+
<!--- Why is this change required? What problem does it solve? -->
|
10 |
+
<!--- If it fixes an open issue, please link to the issue here. -->
|
11 |
+
|
12 |
+
## How has this been tested?
|
13 |
+
|
14 |
+
<!--- Please describe in detail how you tested your changes. -->
|
15 |
+
<!--- Include details of your testing environment, tests ran to see how -->
|
16 |
+
<!--- your change affects other areas of the code, etc. -->
|
17 |
+
|
18 |
+
## Screenshots (if appropriate)
|
19 |
+
|
20 |
+
## Types of changes
|
21 |
+
|
22 |
+
<!--- What types of changes does your code introduce? Put an `x` in all the boxes that apply: -->
|
23 |
+
|
24 |
+
## Social Handles (Optional)
|
25 |
+
|
26 |
+
<!-- Thanks for submitting a bugfix or enhancement. -->
|
27 |
+
<!-- We'd love to show our thanks to you on Twitter & Discord if you provide your handle -->
|
.github/SECURITY.md
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Security Policy
|
2 |
+
|
3 |
+
## Supported Versions
|
4 |
+
|
5 |
+
Due to the nature of the fast development that is happening in this project, only the latest released version can be supported.
|
6 |
+
|
7 |
+
## Reporting a Vulnerability
|
8 |
+
|
9 |
+
If you find a vulnerability, please contact us on [Discord](https://discord.gg/xcu3ECkH9a) rather than creating a GitHub issue to allow us some time to fix it before it is a known vulnerability to others.
|
.github/SUPPORT.md
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Support
|
2 |
+
|
3 |
+
If you need help with this project or have questions, please:
|
4 |
+
|
5 |
+
1. Check the documentation.
|
6 |
+
2. Search the existing issues and pull requests.
|
7 |
+
3. Create a new issue if your question is not answered or your problem is not solved.
|
8 |
+
4. Have a look in the [Discord server](https://discord.gg/HhrNrHJPRb)
|
9 |
+
|
10 |
+
Please note that this project is maintained by volunteers who have limited availability. We'll do our best to address your questions and concerns in a timely manner.
|
.github/release-drafter.yml
ADDED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name-template: 'v$RESOLVED_VERSION'
|
2 |
+
tag-template: 'v$RESOLVED_VERSION'
|
3 |
+
categories:
|
4 |
+
- title: '🚀 Features'
|
5 |
+
labels:
|
6 |
+
- 'feature'
|
7 |
+
- 'enhancement'
|
8 |
+
- title: '🐛 Bug Fixes'
|
9 |
+
labels:
|
10 |
+
- 'fix'
|
11 |
+
- 'bugfix'
|
12 |
+
- 'bug'
|
13 |
+
- title: '🧰 Maintenance'
|
14 |
+
label: 'chore'
|
15 |
+
change-template: '- $TITLE @$AUTHOR (#$NUMBER)'
|
16 |
+
change-title-escapes: '\<*_&' # You can add # and @ to disable mentions, and add ` to disable code blocks.
|
17 |
+
version-resolver:
|
18 |
+
major:
|
19 |
+
labels:
|
20 |
+
- 'major'
|
21 |
+
minor:
|
22 |
+
labels:
|
23 |
+
- 'minor'
|
24 |
+
patch:
|
25 |
+
labels:
|
26 |
+
- 'patch'
|
27 |
+
default: patch
|
28 |
+
template: |
|
29 |
+
## What’s Changed
|
30 |
+
|
31 |
+
$CHANGES
|
.github/workflows/base.yml
ADDED
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name: ci-cd-base
|
2 |
+
|
3 |
+
on:
|
4 |
+
workflow_dispatch:
|
5 |
+
|
6 |
+
jobs:
|
7 |
+
build-base:
|
8 |
+
if: github.repository_owner == 'OpenAccess-AI-Collective'
|
9 |
+
# this job needs to be run on self-hosted GPU runners...
|
10 |
+
runs-on: axolotl-gpu-runner
|
11 |
+
strategy:
|
12 |
+
fail-fast: false
|
13 |
+
matrix:
|
14 |
+
include:
|
15 |
+
- cuda: "118"
|
16 |
+
cuda_version: 11.8.0
|
17 |
+
python_version: "3.10"
|
18 |
+
pytorch: 2.1.2
|
19 |
+
torch_cuda_arch_list: "7.0 7.5 8.0 8.6 9.0+PTX"
|
20 |
+
- cuda: "121"
|
21 |
+
cuda_version: 12.1.0
|
22 |
+
python_version: "3.10"
|
23 |
+
pytorch: 2.1.2
|
24 |
+
torch_cuda_arch_list: "7.0 7.5 8.0 8.6 9.0+PTX"
|
25 |
+
- cuda: "121"
|
26 |
+
cuda_version: 12.1.0
|
27 |
+
python_version: "3.11"
|
28 |
+
pytorch: 2.1.2
|
29 |
+
torch_cuda_arch_list: "7.0 7.5 8.0 8.6 9.0+PTX"
|
30 |
+
steps:
|
31 |
+
- name: Checkout
|
32 |
+
uses: actions/checkout@v3
|
33 |
+
- name: Docker metadata
|
34 |
+
id: metadata
|
35 |
+
uses: docker/metadata-action@v3
|
36 |
+
with:
|
37 |
+
images: winglian/axolotl-base
|
38 |
+
- name: Login to Docker Hub
|
39 |
+
uses: docker/login-action@v2
|
40 |
+
with:
|
41 |
+
username: ${{ secrets.DOCKERHUB_USERNAME }}
|
42 |
+
password: ${{ secrets.DOCKERHUB_TOKEN }}
|
43 |
+
- name: Set up Docker Buildx
|
44 |
+
uses: docker/setup-buildx-action@v2
|
45 |
+
- name: Build
|
46 |
+
uses: docker/build-push-action@v4
|
47 |
+
with:
|
48 |
+
context: .
|
49 |
+
file: ./docker/Dockerfile-base
|
50 |
+
push: ${{ github.event_name != 'pull_request' }}
|
51 |
+
tags: ${{ steps.metadata.outputs.tags }}-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
|
52 |
+
labels: ${{ steps.metadata.outputs.labels }}
|
53 |
+
build-args: |
|
54 |
+
CUDA_VERSION=${{ matrix.cuda_version }}
|
55 |
+
CUDA=${{ matrix.cuda }}
|
56 |
+
PYTHON_VERSION=${{ matrix.python_version }}
|
57 |
+
PYTORCH_VERSION=${{ matrix.pytorch }}
|
58 |
+
TORCH_CUDA_ARCH_LIST=${{ matrix.torch_cuda_arch_list }}
|
.github/workflows/docs.yml
ADDED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name: Publish Docs
|
2 |
+
on:
|
3 |
+
push:
|
4 |
+
branches:
|
5 |
+
- main
|
6 |
+
|
7 |
+
permissions:
|
8 |
+
contents: write
|
9 |
+
pages: write
|
10 |
+
|
11 |
+
jobs:
|
12 |
+
build-deploy:
|
13 |
+
runs-on: ubuntu-latest
|
14 |
+
steps:
|
15 |
+
- name: Check out repository
|
16 |
+
uses: actions/checkout@v4
|
17 |
+
- name: Set up Quarto
|
18 |
+
uses: quarto-dev/quarto-actions/setup@v2
|
19 |
+
- name: Setup Python
|
20 |
+
uses: actions/setup-python@v3
|
21 |
+
with:
|
22 |
+
python-version: '3.10'
|
23 |
+
- name: install dependencies
|
24 |
+
run: |
|
25 |
+
python3 -m pip install jupyter
|
26 |
+
- name: Publish to GitHub Pages (and render)
|
27 |
+
uses: quarto-dev/quarto-actions/publish@v2
|
28 |
+
with:
|
29 |
+
target: gh-pages
|
30 |
+
env:
|
31 |
+
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
|
.github/workflows/lint.yml
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name: lint
|
2 |
+
on:
|
3 |
+
# check on PRs, and manual triggers
|
4 |
+
pull_request:
|
5 |
+
paths:
|
6 |
+
- '**.py'
|
7 |
+
- 'requirements.txt'
|
8 |
+
- '.github/workflows/*.yml'
|
9 |
+
- "*.md"
|
10 |
+
workflow_dispatch:
|
11 |
+
|
12 |
+
jobs:
|
13 |
+
pre-commit:
|
14 |
+
name: pre-commit
|
15 |
+
runs-on: ubuntu-latest
|
16 |
+
steps:
|
17 |
+
- uses: actions/checkout@v3
|
18 |
+
- uses: actions/setup-python@v4
|
19 |
+
with:
|
20 |
+
python-version: "3.10"
|
21 |
+
cache: 'pip' # caching pip dependencies
|
22 |
+
- uses: pre-commit/action@v3.0.0
|
.github/workflows/main.yml
ADDED
@@ -0,0 +1,119 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name: ci-cd
|
2 |
+
|
3 |
+
on:
|
4 |
+
push:
|
5 |
+
branches:
|
6 |
+
- "main"
|
7 |
+
workflow_dispatch:
|
8 |
+
|
9 |
+
jobs:
|
10 |
+
build-axolotl:
|
11 |
+
if: ${{ ! contains(github.event.commits[0].message, '[skip docker]]') && github.repository_owner == 'OpenAccess-AI-Collective' }}
|
12 |
+
strategy:
|
13 |
+
fail-fast: false
|
14 |
+
matrix:
|
15 |
+
include:
|
16 |
+
- cuda: 118
|
17 |
+
cuda_version: 11.8.0
|
18 |
+
python_version: "3.10"
|
19 |
+
pytorch: 2.1.2
|
20 |
+
axolotl_extras:
|
21 |
+
axolotl_args: "--extra-index-url https://download.pytorch.org/whl/cu118"
|
22 |
+
is_latest: true
|
23 |
+
- cuda: 121
|
24 |
+
cuda_version: 12.1.0
|
25 |
+
python_version: "3.10"
|
26 |
+
pytorch: 2.1.2
|
27 |
+
axolotl_extras:
|
28 |
+
- cuda: 121
|
29 |
+
cuda_version: 12.1.0
|
30 |
+
python_version: "3.11"
|
31 |
+
pytorch: 2.1.2
|
32 |
+
axolotl_extras:
|
33 |
+
runs-on: axolotl-gpu-runner
|
34 |
+
steps:
|
35 |
+
- name: Checkout
|
36 |
+
uses: actions/checkout@v4
|
37 |
+
- name: Docker metadata
|
38 |
+
id: metadata
|
39 |
+
uses: docker/metadata-action@v5
|
40 |
+
with:
|
41 |
+
images: winglian/axolotl
|
42 |
+
- name: Set up Docker Buildx
|
43 |
+
uses: docker/setup-buildx-action@v3
|
44 |
+
- name: Login to Docker Hub
|
45 |
+
uses: docker/login-action@v3
|
46 |
+
with:
|
47 |
+
username: ${{ secrets.DOCKERHUB_USERNAME }}
|
48 |
+
password: ${{ secrets.DOCKERHUB_TOKEN }}
|
49 |
+
# guidance for testing before pushing: https://docs.docker.com/build/ci/github-actions/test-before-push/
|
50 |
+
- name: Build and export to Docker
|
51 |
+
uses: docker/build-push-action@v5
|
52 |
+
with:
|
53 |
+
context: .
|
54 |
+
build-args: |
|
55 |
+
BASE_TAG=${{ github.ref_name }}-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}
|
56 |
+
CUDA=${{ matrix.cuda }}
|
57 |
+
PYTORCH_VERSION=${{ matrix.pytorch }}
|
58 |
+
AXOLOTL_ARGS=${{ matrix.axolotl_args }}
|
59 |
+
file: ./docker/Dockerfile
|
60 |
+
push: ${{ github.event_name != 'pull_request' }}
|
61 |
+
tags: |
|
62 |
+
${{ steps.metadata.outputs.tags }}-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
|
63 |
+
${{ (matrix.is_latest) && format('{0}-latest', steps.metadata.outputs.tags) || '' }}
|
64 |
+
labels: ${{ steps.metadata.outputs.labels }}
|
65 |
+
|
66 |
+
build-axolotl-runpod:
|
67 |
+
needs: build-axolotl
|
68 |
+
if: ${{ ! contains(github.event.commits[0].message, '[skip docker]]') && github.repository_owner == 'OpenAccess-AI-Collective' }}
|
69 |
+
# this job needs to be run on self-hosted GPU runners...
|
70 |
+
strategy:
|
71 |
+
matrix:
|
72 |
+
include:
|
73 |
+
- cuda: 118
|
74 |
+
cuda_version: 11.8.0
|
75 |
+
python_version: "3.10"
|
76 |
+
pytorch: 2.1.2
|
77 |
+
axolotl_extras:
|
78 |
+
is_latest: true
|
79 |
+
- cuda: 121
|
80 |
+
cuda_version: 12.1.0
|
81 |
+
python_version: "3.10"
|
82 |
+
pytorch: 2.1.2
|
83 |
+
axolotl_extras:
|
84 |
+
- cuda: 121
|
85 |
+
cuda_version: 12.1.0
|
86 |
+
python_version: "3.11"
|
87 |
+
pytorch: 2.1.2
|
88 |
+
axolotl_extras:
|
89 |
+
runs-on: axolotl-gpu-runner
|
90 |
+
steps:
|
91 |
+
- name: Checkout
|
92 |
+
uses: actions/checkout@v4
|
93 |
+
- name: Docker metadata
|
94 |
+
id: metadata
|
95 |
+
uses: docker/metadata-action@v5
|
96 |
+
with:
|
97 |
+
images: winglian/axolotl-cloud
|
98 |
+
- name: Login to Docker Hub
|
99 |
+
uses: docker/login-action@v3
|
100 |
+
with:
|
101 |
+
username: ${{ secrets.DOCKERHUB_USERNAME }}
|
102 |
+
password: ${{ secrets.DOCKERHUB_TOKEN }}
|
103 |
+
- name: Set up Docker Buildx
|
104 |
+
uses: docker/setup-buildx-action@v2
|
105 |
+
- name: Build
|
106 |
+
uses: docker/build-push-action@v5
|
107 |
+
with:
|
108 |
+
context: .
|
109 |
+
build-args: |
|
110 |
+
BASE_TAG=${{ github.ref_name }}-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
|
111 |
+
CUDA=${{ matrix.cuda }}
|
112 |
+
file: ./docker/Dockerfile-cloud
|
113 |
+
push: ${{ github.event_name != 'pull_request' }}
|
114 |
+
tags: |
|
115 |
+
${{ steps.metadata.outputs.tags }}-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
|
116 |
+
winglian/axolotl-runpod:main-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
|
117 |
+
${{ (matrix.is_latest) && format('{0}-latest', steps.metadata.outputs.tags) || '' }}
|
118 |
+
${{ (matrix.is_latest) && format('{0}-latest', 'winglian/axolotl-runpod:main') || '' }}
|
119 |
+
labels: ${{ steps.metadata.outputs.labels }}
|
.github/workflows/pypi.yml
ADDED
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name: publish pypi
|
2 |
+
|
3 |
+
on:
|
4 |
+
push:
|
5 |
+
tags:
|
6 |
+
- '*'
|
7 |
+
|
8 |
+
jobs:
|
9 |
+
pypi-publish:
|
10 |
+
name: Upload release to PyPI
|
11 |
+
runs-on: ubuntu-latest
|
12 |
+
environment:
|
13 |
+
name: pypi
|
14 |
+
url: https://pypi.org/p/axolotl
|
15 |
+
permissions:
|
16 |
+
id-token: write # IMPORTANT: this permission is mandatory for trusted publishing
|
17 |
+
steps:
|
18 |
+
- name: Check out repository code
|
19 |
+
uses: actions/checkout@v3
|
20 |
+
|
21 |
+
- name: Setup Python
|
22 |
+
uses: actions/setup-python@v4
|
23 |
+
with:
|
24 |
+
python-version: "3.10"
|
25 |
+
|
26 |
+
- name: Install dependencies
|
27 |
+
run: |
|
28 |
+
pip3 install wheel
|
29 |
+
pip3 install -e .
|
30 |
+
pip3 install -r requirements-tests.txt
|
31 |
+
|
32 |
+
- name: Extract tag name
|
33 |
+
id: tag
|
34 |
+
run: echo ::set-output name=TAG_NAME::$(echo $GITHUB_REF | cut -d / -f 3)
|
35 |
+
|
36 |
+
- name: Update version in setup.py
|
37 |
+
run: |
|
38 |
+
sed -i -E 's/version="([0-9.]+)",/version="${{ steps.tag.outputs.TAG_NAME }}",/g' setup.py
|
39 |
+
|
40 |
+
- name: Build a binary wheel
|
41 |
+
run: |
|
42 |
+
python setup.py sdist bdist_wheel
|
43 |
+
|
44 |
+
- name: Publish package distributions to PyPI
|
45 |
+
uses: pypa/gh-action-pypi-publish@release/v1
|
.github/workflows/tests.yml
ADDED
@@ -0,0 +1,100 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name: Tests
|
2 |
+
on:
|
3 |
+
# check on push/merge to main, PRs, and manual triggers
|
4 |
+
push:
|
5 |
+
branches:
|
6 |
+
- "main"
|
7 |
+
paths:
|
8 |
+
- '**.py'
|
9 |
+
- 'requirements.txt'
|
10 |
+
- '.github/workflows/*.yml'
|
11 |
+
pull_request:
|
12 |
+
paths:
|
13 |
+
- '**.py'
|
14 |
+
- 'requirements.txt'
|
15 |
+
- '.github/workflows/*.yml'
|
16 |
+
workflow_dispatch:
|
17 |
+
|
18 |
+
jobs:
|
19 |
+
pre-commit:
|
20 |
+
name: pre-commit
|
21 |
+
runs-on: ubuntu-latest
|
22 |
+
steps:
|
23 |
+
- uses: actions/checkout@v3
|
24 |
+
- uses: actions/setup-python@v4
|
25 |
+
with:
|
26 |
+
python-version: "3.10"
|
27 |
+
cache: 'pip' # caching pip dependencies
|
28 |
+
- uses: pre-commit/action@v3.0.0
|
29 |
+
|
30 |
+
pytest:
|
31 |
+
name: PyTest
|
32 |
+
runs-on: ubuntu-latest
|
33 |
+
strategy:
|
34 |
+
fail-fast: false
|
35 |
+
matrix:
|
36 |
+
python_version: ["3.10", "3.11"]
|
37 |
+
timeout-minutes: 20
|
38 |
+
|
39 |
+
steps:
|
40 |
+
- name: Check out repository code
|
41 |
+
uses: actions/checkout@v3
|
42 |
+
|
43 |
+
- name: Setup Python
|
44 |
+
uses: actions/setup-python@v4
|
45 |
+
with:
|
46 |
+
python-version: ${{ matrix.python_version }}
|
47 |
+
cache: 'pip' # caching pip dependencies
|
48 |
+
|
49 |
+
- name: Install dependencies
|
50 |
+
run: |
|
51 |
+
pip3 install -U -e .
|
52 |
+
pip3 install -r requirements-tests.txt
|
53 |
+
|
54 |
+
- name: Run tests
|
55 |
+
run: |
|
56 |
+
pytest --ignore=tests/e2e/ tests/
|
57 |
+
|
58 |
+
docker-e2e-tests:
|
59 |
+
if: github.repository_owner == 'OpenAccess-AI-Collective'
|
60 |
+
# this job needs to be run on self-hosted GPU runners...
|
61 |
+
runs-on: [self-hosted, modal]
|
62 |
+
timeout-minutes: 60
|
63 |
+
needs: [pre-commit, pytest]
|
64 |
+
|
65 |
+
strategy:
|
66 |
+
fail-fast: false
|
67 |
+
matrix:
|
68 |
+
include:
|
69 |
+
- cuda: 118
|
70 |
+
cuda_version: 11.8.0
|
71 |
+
python_version: "3.10"
|
72 |
+
pytorch: 2.1.2
|
73 |
+
axolotl_args: "--extra-index-url https://download.pytorch.org/whl/cu118"
|
74 |
+
num_gpus: 1
|
75 |
+
- cuda: 121
|
76 |
+
cuda_version: 12.1.0
|
77 |
+
python_version: "3.10"
|
78 |
+
pytorch: 2.1.2
|
79 |
+
num_gpus: 1
|
80 |
+
steps:
|
81 |
+
- name: Checkout
|
82 |
+
uses: actions/checkout@v4
|
83 |
+
- name: Install Python
|
84 |
+
uses: actions/setup-python@v5
|
85 |
+
with:
|
86 |
+
python-version: "3.10"
|
87 |
+
- name: Install Modal
|
88 |
+
run: |
|
89 |
+
python -m pip install --upgrade pip
|
90 |
+
pip install modal jinja2
|
91 |
+
- name: Update env vars
|
92 |
+
run: |
|
93 |
+
echo "BASE_TAG=main-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
|
94 |
+
echo "PYTORCH_VERSION=${{ matrix.pytorch}}" >> $GITHUB_ENV
|
95 |
+
echo "AXOLOTL_ARGS=${{ matrix.axolotl_args}}" >> $GITHUB_ENV
|
96 |
+
echo "CUDA=${{ matrix.cuda }}" >> $GITHUB_ENV
|
97 |
+
echo "N_GPUS=${{ matrix.num_gpus }}" >> $GITHUB_ENV
|
98 |
+
- name: Run tests job on Modal
|
99 |
+
run: |
|
100 |
+
modal run cicd.tests
|
.gitignore
ADDED
@@ -0,0 +1,177 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
**/axolotl.egg-info
|
2 |
+
configs
|
3 |
+
last_run_prepared/
|
4 |
+
.vscode
|
5 |
+
_site/
|
6 |
+
|
7 |
+
# Byte-compiled / optimized / DLL files
|
8 |
+
__pycache__/
|
9 |
+
*.py[cod]
|
10 |
+
*$py.class
|
11 |
+
|
12 |
+
# C extensions
|
13 |
+
*.so
|
14 |
+
|
15 |
+
# Distribution / packaging
|
16 |
+
.Python
|
17 |
+
build/
|
18 |
+
develop-eggs/
|
19 |
+
dist/
|
20 |
+
downloads/
|
21 |
+
eggs/
|
22 |
+
.eggs/
|
23 |
+
lib/
|
24 |
+
lib64/
|
25 |
+
parts/
|
26 |
+
sdist/
|
27 |
+
var/
|
28 |
+
wheels/
|
29 |
+
share/python-wheels/
|
30 |
+
*.egg-info/
|
31 |
+
.installed.cfg
|
32 |
+
*.egg
|
33 |
+
MANIFEST
|
34 |
+
|
35 |
+
# PyInstaller
|
36 |
+
# Usually these files are written by a python script from a template
|
37 |
+
# before PyInstaller builds the exe, so as to inject date/other infos into it.
|
38 |
+
*.manifest
|
39 |
+
*.spec
|
40 |
+
|
41 |
+
# Installer logs
|
42 |
+
pip-log.txt
|
43 |
+
pip-delete-this-directory.txt
|
44 |
+
|
45 |
+
# Unit test / coverage reports
|
46 |
+
htmlcov/
|
47 |
+
.tox/
|
48 |
+
.nox/
|
49 |
+
.coverage
|
50 |
+
.coverage.*
|
51 |
+
.cache
|
52 |
+
nosetests.xml
|
53 |
+
coverage.xml
|
54 |
+
*.cover
|
55 |
+
*.py,cover
|
56 |
+
.hypothesis/
|
57 |
+
.pytest_cache/
|
58 |
+
cover/
|
59 |
+
|
60 |
+
# Translations
|
61 |
+
*.mo
|
62 |
+
*.pot
|
63 |
+
|
64 |
+
# Django stuff:
|
65 |
+
*.log
|
66 |
+
local_settings.py
|
67 |
+
db.sqlite3
|
68 |
+
db.sqlite3-journal
|
69 |
+
|
70 |
+
# Flask stuff:
|
71 |
+
instance/
|
72 |
+
.webassets-cache
|
73 |
+
|
74 |
+
# Scrapy stuff:
|
75 |
+
.scrapy
|
76 |
+
|
77 |
+
# Sphinx documentation
|
78 |
+
docs/_build/
|
79 |
+
|
80 |
+
# PyBuilder
|
81 |
+
.pybuilder/
|
82 |
+
target/
|
83 |
+
|
84 |
+
# Jupyter Notebook
|
85 |
+
.ipynb_checkpoints
|
86 |
+
|
87 |
+
# IPython
|
88 |
+
profile_default/
|
89 |
+
ipython_config.py
|
90 |
+
|
91 |
+
# pyenv
|
92 |
+
# For a library or package, you might want to ignore these files since the code is
|
93 |
+
# intended to run in multiple environments; otherwise, check them in:
|
94 |
+
# .python-version
|
95 |
+
|
96 |
+
# pipenv
|
97 |
+
# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
|
98 |
+
# However, in case of collaboration, if having platform-specific dependencies or dependencies
|
99 |
+
# having no cross-platform support, pipenv may install dependencies that don't work, or not
|
100 |
+
# install all needed dependencies.
|
101 |
+
#Pipfile.lock
|
102 |
+
|
103 |
+
# poetry
|
104 |
+
# Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
|
105 |
+
# This is especially recommended for binary packages to ensure reproducibility, and is more
|
106 |
+
# commonly ignored for libraries.
|
107 |
+
# https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
|
108 |
+
#poetry.lock
|
109 |
+
|
110 |
+
# pdm
|
111 |
+
# Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
|
112 |
+
#pdm.lock
|
113 |
+
# pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
|
114 |
+
# in version control.
|
115 |
+
# https://pdm.fming.dev/#use-with-ide
|
116 |
+
.pdm.toml
|
117 |
+
|
118 |
+
# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
|
119 |
+
__pypackages__/
|
120 |
+
|
121 |
+
# Celery stuff
|
122 |
+
celerybeat-schedule
|
123 |
+
celerybeat.pid
|
124 |
+
|
125 |
+
# SageMath parsed files
|
126 |
+
*.sage.py
|
127 |
+
|
128 |
+
# Environments
|
129 |
+
.env
|
130 |
+
.venv
|
131 |
+
env/
|
132 |
+
venv/
|
133 |
+
ENV/
|
134 |
+
env.bak/
|
135 |
+
venv.bak/
|
136 |
+
|
137 |
+
# Spyder project settings
|
138 |
+
.spyderproject
|
139 |
+
.spyproject
|
140 |
+
|
141 |
+
# Rope project settings
|
142 |
+
.ropeproject
|
143 |
+
|
144 |
+
# mkdocs documentation
|
145 |
+
/site
|
146 |
+
|
147 |
+
# mypy
|
148 |
+
.mypy_cache/
|
149 |
+
.dmypy.json
|
150 |
+
dmypy.json
|
151 |
+
|
152 |
+
# Pyre type checker
|
153 |
+
.pyre/
|
154 |
+
|
155 |
+
# pytype static type analyzer
|
156 |
+
.pytype/
|
157 |
+
|
158 |
+
# Cython debug symbols
|
159 |
+
cython_debug/
|
160 |
+
|
161 |
+
# PyCharm
|
162 |
+
# JetBrains specific template is maintained in a separate JetBrains.gitignore that can
|
163 |
+
# be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
|
164 |
+
# and can be added to the global gitignore or merged into this file. For a more nuclear
|
165 |
+
# option (not recommended) you can uncomment the following to ignore the entire idea folder.
|
166 |
+
.idea/
|
167 |
+
|
168 |
+
# WandB
|
169 |
+
# wandb creates a folder to store logs for training runs
|
170 |
+
wandb
|
171 |
+
|
172 |
+
# Runs
|
173 |
+
lora-out/*
|
174 |
+
qlora-out/*
|
175 |
+
mlruns/*
|
176 |
+
|
177 |
+
/.quarto/
|
.isort.cfg
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
[settings]
|
2 |
+
profile=black
|
3 |
+
known_third_party=wandb
|
.mypy.ini
ADDED
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[mypy]
|
2 |
+
plugins = pydantic.mypy
|
3 |
+
exclude = venv
|
4 |
+
|
5 |
+
[mypy-alpaca_lora_4bit.*]
|
6 |
+
ignore_missing_imports = True
|
7 |
+
|
8 |
+
[mypy-axolotl.monkeypatch.*]
|
9 |
+
ignore_errors = True
|
10 |
+
|
11 |
+
[mypy-axolotl.models.mixtral.*]
|
12 |
+
ignore_errors = True
|
13 |
+
|
14 |
+
[mypy-axolotl.models.phi.*]
|
15 |
+
ignore_errors = True
|
16 |
+
|
17 |
+
[mypy-flash_attn.*]
|
18 |
+
ignore_missing_imports = True
|
19 |
+
|
20 |
+
[mypy-huggingface_hub]
|
21 |
+
ignore_missing_imports = True
|
22 |
+
|
23 |
+
[mypy-transformers.*]
|
24 |
+
ignore_missing_imports = True
|
25 |
+
|
26 |
+
[mypy-peft]
|
27 |
+
ignore_missing_imports = True
|
28 |
+
|
29 |
+
[mypy-wandb]
|
30 |
+
ignore_missing_imports = True
|
31 |
+
|
32 |
+
[mypy-bitsandbytes]
|
33 |
+
ignore_missing_imports = True
|
34 |
+
|
35 |
+
[mypy-requests]
|
36 |
+
ignore_missing_imports = True
|
37 |
+
|
38 |
+
[mypy-datasets]
|
39 |
+
ignore_missing_imports = True
|
40 |
+
|
41 |
+
[mypy-fire]
|
42 |
+
ignore_missing_imports = True
|
43 |
+
|
44 |
+
[mypy-setuptools]
|
45 |
+
ignore_missing_imports = True
|
46 |
+
|
47 |
+
[mypy-addict]
|
48 |
+
ignore_missing_imports = True
|
49 |
+
|
50 |
+
[mypy-xformers.*]
|
51 |
+
ignore_missing_imports = True
|
.pre-commit-config.yaml
ADDED
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
default_language_version:
|
2 |
+
python: python3
|
3 |
+
|
4 |
+
repos:
|
5 |
+
- repo: https://github.com/pre-commit/pre-commit-hooks
|
6 |
+
rev: v4.4.0
|
7 |
+
hooks:
|
8 |
+
- id: check-yaml
|
9 |
+
- id: end-of-file-fixer
|
10 |
+
- id: trailing-whitespace
|
11 |
+
- repo: https://github.com/psf/black
|
12 |
+
rev: 23.3.0
|
13 |
+
hooks:
|
14 |
+
- id: black
|
15 |
+
- repo: https://github.com/pycqa/isort
|
16 |
+
rev: 5.12.0
|
17 |
+
hooks:
|
18 |
+
- id: isort
|
19 |
+
- repo: https://github.com/PyCQA/flake8
|
20 |
+
rev: 6.0.0
|
21 |
+
hooks:
|
22 |
+
- id: flake8
|
23 |
+
- repo: https://github.com/PyCQA/pylint
|
24 |
+
rev: v2.17.4
|
25 |
+
hooks:
|
26 |
+
- id: pylint
|
27 |
+
- repo: https://github.com/pre-commit/mirrors-mypy
|
28 |
+
rev: v1.3.0
|
29 |
+
hooks:
|
30 |
+
- id: mypy
|
31 |
+
additional_dependencies:
|
32 |
+
[
|
33 |
+
'types-PyYAML',
|
34 |
+
'pydantic>=2.5.3',
|
35 |
+
]
|
36 |
+
- repo: https://github.com/PyCQA/bandit
|
37 |
+
rev: 1.7.5
|
38 |
+
hooks:
|
39 |
+
- id: bandit
|
40 |
+
args: [
|
41 |
+
'--ini',
|
42 |
+
'.bandit',
|
43 |
+
]
|
.pylintrc
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[MASTER]
|
2 |
+
init-hook="from pylint.config import find_pylintrc; import os, sys; sys.path.append(os.path.dirname(find_pylintrc()))"
|
3 |
+
|
4 |
+
[TYPECHECK]
|
5 |
+
|
6 |
+
# List of members which are set dynamically and missed by Pylint inference
|
7 |
+
# system, and so shouldn't trigger E1101 when accessed.
|
8 |
+
generated-members=numpy.*, torch.*
|
9 |
+
|
10 |
+
|
11 |
+
[pylint.messages_control]
|
12 |
+
disable=missing-function-docstring, line-too-long, import-error,
|
13 |
+
too-many-arguments, too-many-locals, too-many-statements, too-many-branches, too-few-public-methods,
|
14 |
+
too-many-instance-attributes, fixme, import-outside-toplevel, logging-fstring-interpolation,
|
FAQS.md
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# FAQs
|
2 |
+
|
3 |
+
- Can you train StableLM with this? Yes, but only with a single GPU atm. Multi GPU support is coming soon! Just waiting on this [PR](https://github.com/huggingface/transformers/pull/22874)
|
4 |
+
- Will this work with Deepspeed? That's still a WIP, but setting `export ACCELERATE_USE_DEEPSPEED=true` should work in some cases
|
5 |
+
- `Error invalid argument at line 359 in file /workspace/bitsandbytes/csrc/pythonInterface.c`
|
6 |
+
`/arrow/cpp/src/arrow/filesystem/s3fs.cc:2598: arrow::fs::FinalizeS3 was not called even though S3 was initialized.`
|
7 |
+
This could lead to a segmentation fault at exit. Try reinstalling bitsandbytes and transformers from source.
|
LICENSE
ADDED
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
Apache License
|
3 |
+
Version 2.0, January 2004
|
4 |
+
http://www.apache.org/licenses/
|
5 |
+
|
6 |
+
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
|
7 |
+
|
8 |
+
1. Definitions.
|
9 |
+
|
10 |
+
"License" shall mean the terms and conditions for use, reproduction,
|
11 |
+
and distribution as defined by Sections 1 through 9 of this document.
|
12 |
+
|
13 |
+
"Licensor" shall mean the copyright owner or entity authorized by
|
14 |
+
the copyright owner that is granting the License.
|
15 |
+
|
16 |
+
"Legal Entity" shall mean the union of the acting entity and all
|
17 |
+
other entities that control, are controlled by, or are under common
|
18 |
+
control with that entity. For the purposes of this definition,
|
19 |
+
"control" means (i) the power, direct or indirect, to cause the
|
20 |
+
direction or management of such entity, whether by contract or
|
21 |
+
otherwise, or (ii) ownership of fifty percent (50%) or more of the
|
22 |
+
outstanding shares, or (iii) beneficial ownership of such entity.
|
23 |
+
|
24 |
+
"You" (or "Your") shall mean an individual or Legal Entity
|
25 |
+
exercising permissions granted by this License.
|
26 |
+
|
27 |
+
"Source" form shall mean the preferred form for making modifications,
|
28 |
+
including but not limited to software source code, documentation
|
29 |
+
source, and configuration files.
|
30 |
+
|
31 |
+
"Object" form shall mean any form resulting from mechanical
|
32 |
+
transformation or translation of a Source form, including but
|
33 |
+
not limited to compiled object code, generated documentation,
|
34 |
+
and conversions to other media types.
|
35 |
+
|
36 |
+
"Work" shall mean the work of authorship, whether in Source or
|
37 |
+
Object form, made available under the License, as indicated by a
|
38 |
+
copyright notice that is included in or attached to the work
|
39 |
+
(an example is provided in the Appendix below).
|
40 |
+
|
41 |
+
"Derivative Works" shall mean any work, whether in Source or Object
|
42 |
+
form, that is based on (or derived from) the Work and for which the
|
43 |
+
editorial revisions, annotations, elaborations, or other modifications
|
44 |
+
represent, as a whole, an original work of authorship. For the purposes
|
45 |
+
of this License, Derivative Works shall not include works that remain
|
46 |
+
separable from, or merely link (or bind by name) to the interfaces of,
|
47 |
+
the Work and Derivative Works thereof.
|
48 |
+
|
49 |
+
"Contribution" shall mean any work of authorship, including
|
50 |
+
the original version of the Work and any modifications or additions
|
51 |
+
to that Work or Derivative Works thereof, that is intentionally
|
52 |
+
submitted to Licensor for inclusion in the Work by the copyright owner
|
53 |
+
or by an individual or Legal Entity authorized to submit on behalf of
|
54 |
+
the copyright owner. For the purposes of this definition, "submitted"
|
55 |
+
means any form of electronic, verbal, or written communication sent
|
56 |
+
to the Licensor or its representatives, including but not limited to
|
57 |
+
communication on electronic mailing lists, source code control systems,
|
58 |
+
and issue tracking systems that are managed by, or on behalf of, the
|
59 |
+
Licensor for the purpose of discussing and improving the Work, but
|
60 |
+
excluding communication that is conspicuously marked or otherwise
|
61 |
+
designated in writing by the copyright owner as "Not a Contribution."
|
62 |
+
|
63 |
+
"Contributor" shall mean Licensor and any individual or Legal Entity
|
64 |
+
on behalf of whom a Contribution has been received by Licensor and
|
65 |
+
subsequently incorporated within the Work.
|
66 |
+
|
67 |
+
2. Grant of Copyright License. Subject to the terms and conditions of
|
68 |
+
this License, each Contributor hereby grants to You a perpetual,
|
69 |
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
70 |
+
copyright license to reproduce, prepare Derivative Works of,
|
71 |
+
publicly display, publicly perform, sublicense, and distribute the
|
72 |
+
Work and such Derivative Works in Source or Object form.
|
73 |
+
|
74 |
+
3. Grant of Patent License. Subject to the terms and conditions of
|
75 |
+
this License, each Contributor hereby grants to You a perpetual,
|
76 |
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
77 |
+
(except as stated in this section) patent license to make, have made,
|
78 |
+
use, offer to sell, sell, import, and otherwise transfer the Work,
|
79 |
+
where such license applies only to those patent claims licensable
|
80 |
+
by such Contributor that are necessarily infringed by their
|
81 |
+
Contribution(s) alone or by combination of their Contribution(s)
|
82 |
+
with the Work to which such Contribution(s) was submitted. If You
|
83 |
+
institute patent litigation against any entity (including a
|
84 |
+
cross-claim or counterclaim in a lawsuit) alleging that the Work
|
85 |
+
or a Contribution incorporated within the Work constitutes direct
|
86 |
+
or contributory patent infringement, then any patent licenses
|
87 |
+
granted to You under this License for that Work shall terminate
|
88 |
+
as of the date such litigation is filed.
|
89 |
+
|
90 |
+
4. Redistribution. You may reproduce and distribute copies of the
|
91 |
+
Work or Derivative Works thereof in any medium, with or without
|
92 |
+
modifications, and in Source or Object form, provided that You
|
93 |
+
meet the following conditions:
|
94 |
+
|
95 |
+
(a) You must give any other recipients of the Work or
|
96 |
+
Derivative Works a copy of this License; and
|
97 |
+
|
98 |
+
(b) You must cause any modified files to carry prominent notices
|
99 |
+
stating that You changed the files; and
|
100 |
+
|
101 |
+
(c) You must retain, in the Source form of any Derivative Works
|
102 |
+
that You distribute, all copyright, patent, trademark, and
|
103 |
+
attribution notices from the Source form of the Work,
|
104 |
+
excluding those notices that do not pertain to any part of
|
105 |
+
the Derivative Works; and
|
106 |
+
|
107 |
+
(d) If the Work includes a "NOTICE" text file as part of its
|
108 |
+
distribution, then any Derivative Works that You distribute must
|
109 |
+
include a readable copy of the attribution notices contained
|
110 |
+
within such NOTICE file, excluding those notices that do not
|
111 |
+
pertain to any part of the Derivative Works, in at least one
|
112 |
+
of the following places: within a NOTICE text file distributed
|
113 |
+
as part of the Derivative Works; within the Source form or
|
114 |
+
documentation, if provided along with the Derivative Works; or,
|
115 |
+
within a display generated by the Derivative Works, if and
|
116 |
+
wherever such third-party notices normally appear. The contents
|
117 |
+
of the NOTICE file are for informational purposes only and
|
118 |
+
do not modify the License. You may add Your own attribution
|
119 |
+
notices within Derivative Works that You distribute, alongside
|
120 |
+
or as an addendum to the NOTICE text from the Work, provided
|
121 |
+
that such additional attribution notices cannot be construed
|
122 |
+
as modifying the License.
|
123 |
+
|
124 |
+
You may add Your own copyright statement to Your modifications and
|
125 |
+
may provide additional or different license terms and conditions
|
126 |
+
for use, reproduction, or distribution of Your modifications, or
|
127 |
+
for any such Derivative Works as a whole, provided Your use,
|
128 |
+
reproduction, and distribution of the Work otherwise complies with
|
129 |
+
the conditions stated in this License.
|
130 |
+
|
131 |
+
5. Submission of Contributions. Unless You explicitly state otherwise,
|
132 |
+
any Contribution intentionally submitted for inclusion in the Work
|
133 |
+
by You to the Licensor shall be under the terms and conditions of
|
134 |
+
this License, without any additional terms or conditions.
|
135 |
+
Notwithstanding the above, nothing herein shall supersede or modify
|
136 |
+
the terms of any separate license agreement you may have executed
|
137 |
+
with Licensor regarding such Contributions.
|
138 |
+
|
139 |
+
6. Trademarks. This License does not grant permission to use the trade
|
140 |
+
names, trademarks, service marks, or product names of the Licensor,
|
141 |
+
except as required for reasonable and customary use in describing the
|
142 |
+
origin of the Work and reproducing the content of the NOTICE file.
|
143 |
+
|
144 |
+
7. Disclaimer of Warranty. Unless required by applicable law or
|
145 |
+
agreed to in writing, Licensor provides the Work (and each
|
146 |
+
Contributor provides its Contributions) on an "AS IS" BASIS,
|
147 |
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
|
148 |
+
implied, including, without limitation, any warranties or conditions
|
149 |
+
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
|
150 |
+
PARTICULAR PURPOSE. You are solely responsible for determining the
|
151 |
+
appropriateness of using or redistributing the Work and assume any
|
152 |
+
risks associated with Your exercise of permissions under this License.
|
153 |
+
|
154 |
+
8. Limitation of Liability. In no event and under no legal theory,
|
155 |
+
whether in tort (including negligence), contract, or otherwise,
|
156 |
+
unless required by applicable law (such as deliberate and grossly
|
157 |
+
negligent acts) or agreed to in writing, shall any Contributor be
|
158 |
+
liable to You for damages, including any direct, indirect, special,
|
159 |
+
incidental, or consequential damages of any character arising as a
|
160 |
+
result of this License or out of the use or inability to use the
|
161 |
+
Work (including but not limited to damages for loss of goodwill,
|
162 |
+
work stoppage, computer failure or malfunction, or any and all
|
163 |
+
other commercial damages or losses), even if such Contributor
|
164 |
+
has been advised of the possibility of such damages.
|
165 |
+
|
166 |
+
9. Accepting Warranty or Additional Liability. While redistributing
|
167 |
+
the Work or Derivative Works thereof, You may choose to offer,
|
168 |
+
and charge a fee for, acceptance of support, warranty, indemnity,
|
169 |
+
or other liability obligations and/or rights consistent with this
|
170 |
+
License. However, in accepting such obligations, You may act only
|
171 |
+
on Your own behalf and on Your sole responsibility, not on behalf
|
172 |
+
of any other Contributor, and only if You agree to indemnify,
|
173 |
+
defend, and hold each Contributor harmless for any liability
|
174 |
+
incurred by, or claims asserted against, such Contributor by reason
|
175 |
+
of your accepting any such warranty or additional liability.
|
176 |
+
|
177 |
+
END OF TERMS AND CONDITIONS
|
178 |
+
|
179 |
+
APPENDIX: How to apply the Apache License to your work.
|
180 |
+
|
181 |
+
To apply the Apache License to your work, attach the following
|
182 |
+
boilerplate notice, with the fields enclosed by brackets "[]"
|
183 |
+
replaced with your own identifying information. (Don't include
|
184 |
+
the brackets!) The text should be enclosed in the appropriate
|
185 |
+
comment syntax for the file format. We also recommend that a
|
186 |
+
file or class name and description of purpose be included on the
|
187 |
+
same "printed page" as the copyright notice for easier
|
188 |
+
identification within third-party archives.
|
189 |
+
|
190 |
+
Copyright [yyyy] [name of copyright owner]
|
191 |
+
|
192 |
+
Licensed under the Apache License, Version 2.0 (the "License");
|
193 |
+
you may not use this file except in compliance with the License.
|
194 |
+
You may obtain a copy of the License at
|
195 |
+
|
196 |
+
http://www.apache.org/licenses/LICENSE-2.0
|
197 |
+
|
198 |
+
Unless required by applicable law or agreed to in writing, software
|
199 |
+
distributed under the License is distributed on an "AS IS" BASIS,
|
200 |
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
201 |
+
See the License for the specific language governing permissions and
|
202 |
+
limitations under the License.
|
README.md
ADDED
@@ -0,0 +1,1356 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Axolotl
|
2 |
+
|
3 |
+
Axolotl is a tool designed to streamline the fine-tuning of various AI models, offering support for multiple configurations and architectures.
|
4 |
+
|
5 |
+
Features:
|
6 |
+
- Train various Huggingface models such as llama, pythia, falcon, mpt
|
7 |
+
- Supports fullfinetune, lora, qlora, relora, and gptq
|
8 |
+
- Customize configurations using a simple yaml file or CLI overwrite
|
9 |
+
- Load different dataset formats, use custom formats, or bring your own tokenized datasets
|
10 |
+
- Integrated with xformer, flash attention, rope scaling, and multipacking
|
11 |
+
- Works with single GPU or multiple GPUs via FSDP or Deepspeed
|
12 |
+
- Easily run with Docker locally or on the cloud
|
13 |
+
- Log results and optionally checkpoints to wandb or mlflow
|
14 |
+
- And more!
|
15 |
+
|
16 |
+
<a href="https://www.phorm.ai/query?projectId=e315ba4a-4e14-421f-ab05-38a1f9076f25">
|
17 |
+
<img alt="phorm.ai" src="https://img.shields.io/badge/Phorm-Ask_AI-%23F2777A.svg?&logo=data:image/svg+xml;base64,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">
|
18 |
+
</a>
|
19 |
+
|
20 |
+
<table>
|
21 |
+
<tr>
|
22 |
+
<td>
|
23 |
+
|
24 |
+
## Table of Contents
|
25 |
+
- [Introduction](#axolotl)
|
26 |
+
- [Supported Features](#axolotl-supports)
|
27 |
+
- [Quickstart](#quickstart-)
|
28 |
+
- [Environment](#environment)
|
29 |
+
- [Docker](#docker)
|
30 |
+
- [Conda/Pip venv](#condapip-venv)
|
31 |
+
- [Cloud GPU](#cloud-gpu) - Latitude.sh, JarvisLabs, RunPod
|
32 |
+
- [Bare Metal Cloud GPU](#bare-metal-cloud-gpu)
|
33 |
+
- [Windows](#windows)
|
34 |
+
- [Mac](#mac)
|
35 |
+
- [Google Colab](#google-colab)
|
36 |
+
- [Launching on public clouds via SkyPilot](#launching-on-public-clouds-via-skypilot)
|
37 |
+
- [Dataset](#dataset)
|
38 |
+
- [How to Add Custom Prompts](#how-to-add-custom-prompts)
|
39 |
+
- [How to Use Custom Pretokenized Dataset](#how-to-use-your-custom-pretokenized-dataset)
|
40 |
+
- [Config](#config)
|
41 |
+
- [Train](#train)
|
42 |
+
- [Inference](#inference-playground)
|
43 |
+
- [Merge LORA to Base](#merge-lora-to-base)
|
44 |
+
- [Special Tokens](#special-tokens)
|
45 |
+
- Advanced Topics
|
46 |
+
- [Multipack](./docs/multipack.qmd)<svg width="24" height="24" viewBox="0 0 24 24" xmlns="http://www.w3.org/2000/svg"><path d="M17 13.5v6H5v-12h6m3-3h6v6m0-6-9 9" class="icon_svg-stroke" stroke="#666" stroke-width="1.5" fill="none" fill-rule="evenodd" stroke-linecap="round" stroke-linejoin="round"></path></svg>
|
47 |
+
- [RLHF & DPO](./docs/rlhf.qmd)<svg width="24" height="24" viewBox="0 0 24 24" xmlns="http://www.w3.org/2000/svg"><path d="M17 13.5v6H5v-12h6m3-3h6v6m0-6-9 9" class="icon_svg-stroke" stroke="#666" stroke-width="1.5" fill="none" fill-rule="evenodd" stroke-linecap="round" stroke-linejoin="round"></path></svg>
|
48 |
+
- [Common Errors](#common-errors-)
|
49 |
+
- [Tokenization Mismatch b/w Training & Inference](#tokenization-mismatch-bw-inference--training)
|
50 |
+
- [Debugging Axolotl](#debugging-axolotl)
|
51 |
+
- [Need Help?](#need-help-)
|
52 |
+
- [Badge](#badge-)
|
53 |
+
- [Community Showcase](#community-showcase)
|
54 |
+
- [Contributing](#contributing-)
|
55 |
+
- [Sponsors](#sponsors-)
|
56 |
+
|
57 |
+
</td>
|
58 |
+
<td>
|
59 |
+
|
60 |
+
<div align="center">
|
61 |
+
<img src="image/axolotl.png" alt="axolotl" width="160">
|
62 |
+
<div>
|
63 |
+
<p>
|
64 |
+
<b>Axolotl provides a unified repository for fine-tuning <br />a variety of AI models with ease</b>
|
65 |
+
</p>
|
66 |
+
<p>
|
67 |
+
Go ahead and Axolotl questions!!
|
68 |
+
</p>
|
69 |
+
<img src="https://github.com/OpenAccess-AI-Collective/axolotl/actions/workflows/pre-commit.yml/badge.svg?branch=main" alt="pre-commit">
|
70 |
+
<img alt="PyTest Status" src="https://github.com/OpenAccess-AI-Collective/axolotl/actions/workflows/tests.yml/badge.svg?branch=main">
|
71 |
+
</div>
|
72 |
+
</div>
|
73 |
+
|
74 |
+
</td>
|
75 |
+
</tr>
|
76 |
+
</table>
|
77 |
+
|
78 |
+
## Axolotl supports
|
79 |
+
|
80 |
+
| | fp16/fp32 | lora | qlora | gptq | gptq w/flash attn | flash attn | xformers attn |
|
81 |
+
|-------------|:----------|:-----|-------|------|-------------------|------------|--------------|
|
82 |
+
| llama | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
|
83 |
+
| Mistral | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
|
84 |
+
| Mixtral-MoE | ✅ | ✅ | ✅ | ❓ | ❓ | ❓ | ❓ |
|
85 |
+
| Pythia | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ | ❓ |
|
86 |
+
| cerebras | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ | ❓ |
|
87 |
+
| btlm | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ | ❓ |
|
88 |
+
| mpt | ✅ | ❌ | ❓ | ❌ | ❌ | ❌ | ❓ |
|
89 |
+
| falcon | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ | ❓ |
|
90 |
+
| gpt-j | ✅ | ✅ | ✅ | ❌ | ❌ | ❓ | ❓ |
|
91 |
+
| XGen | ✅ | ❓ | ✅ | ❓ | ❓ | ❓ | ✅ |
|
92 |
+
| phi | ✅ | ✅ | ✅ | ❓ | ❓ | ❓ | ❓ |
|
93 |
+
| RWKV | ✅ | ❓ | ❓ | ❓ | ❓ | ❓ | ❓ |
|
94 |
+
| Qwen | ✅ | ✅ | ✅ | ❓ | ❓ | ❓ | ❓ |
|
95 |
+
| Gemma | ✅ | ✅ | ✅ | ❓ | ❓ | ✅ | ❓ |
|
96 |
+
|
97 |
+
✅: supported
|
98 |
+
❌: not supported
|
99 |
+
❓: untested
|
100 |
+
|
101 |
+
## Quickstart ⚡
|
102 |
+
|
103 |
+
Get started with Axolotl in just a few steps! This quickstart guide will walk you through setting up and running a basic fine-tuning task.
|
104 |
+
|
105 |
+
**Requirements**: Python >=3.10 and Pytorch >=2.1.1.
|
106 |
+
|
107 |
+
```bash
|
108 |
+
git clone https://github.com/OpenAccess-AI-Collective/axolotl
|
109 |
+
cd axolotl
|
110 |
+
|
111 |
+
pip3 install packaging
|
112 |
+
pip3 install -e '.[flash-attn,deepspeed]'
|
113 |
+
```
|
114 |
+
|
115 |
+
### Usage
|
116 |
+
```bash
|
117 |
+
# preprocess datasets - optional but recommended
|
118 |
+
CUDA_VISIBLE_DEVICES="" python -m axolotl.cli.preprocess examples/openllama-3b/lora.yml
|
119 |
+
|
120 |
+
# finetune lora
|
121 |
+
accelerate launch -m axolotl.cli.train examples/openllama-3b/lora.yml
|
122 |
+
|
123 |
+
# inference
|
124 |
+
accelerate launch -m axolotl.cli.inference examples/openllama-3b/lora.yml \
|
125 |
+
--lora_model_dir="./lora-out"
|
126 |
+
|
127 |
+
# gradio
|
128 |
+
accelerate launch -m axolotl.cli.inference examples/openllama-3b/lora.yml \
|
129 |
+
--lora_model_dir="./lora-out" --gradio
|
130 |
+
|
131 |
+
# remote yaml files - the yaml config can be hosted on a public URL
|
132 |
+
# Note: the yaml config must directly link to the **raw** yaml
|
133 |
+
accelerate launch -m axolotl.cli.train https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/examples/openllama-3b/lora.yml
|
134 |
+
```
|
135 |
+
|
136 |
+
## Advanced Setup
|
137 |
+
|
138 |
+
### Environment
|
139 |
+
|
140 |
+
#### Docker
|
141 |
+
|
142 |
+
```bash
|
143 |
+
docker run --gpus '"all"' --rm -it winglian/axolotl:main-latest
|
144 |
+
```
|
145 |
+
|
146 |
+
Or run on the current files for development:
|
147 |
+
|
148 |
+
```sh
|
149 |
+
docker compose up -d
|
150 |
+
```
|
151 |
+
|
152 |
+
>[!Tip]
|
153 |
+
> If you want to debug axolotl or prefer to use Docker as your development environment, see the [debugging guide's section on Docker](docs/debugging.qmd#debugging-with-docker).
|
154 |
+
|
155 |
+
<details>
|
156 |
+
|
157 |
+
<summary>Docker advanced</summary>
|
158 |
+
|
159 |
+
A more powerful Docker command to run would be this:
|
160 |
+
|
161 |
+
```bash
|
162 |
+
docker run --privileged --gpus '"all"' --shm-size 10g --rm -it --name axolotl --ipc=host --ulimit memlock=-1 --ulimit stack=67108864 --mount type=bind,src="${PWD}",target=/workspace/axolotl -v ${HOME}/.cache/huggingface:/root/.cache/huggingface winglian/axolotl:main-latest
|
163 |
+
```
|
164 |
+
|
165 |
+
It additionally:
|
166 |
+
* Prevents memory issues when running e.g. deepspeed (e.g. you could hit SIGBUS/signal 7 error) through `--ipc` and `--ulimit` args.
|
167 |
+
* Persists the downloaded HF data (models etc.) and your modifications to axolotl code through `--mount`/`-v` args.
|
168 |
+
* The `--name` argument simply makes it easier to refer to the container in vscode (`Dev Containers: Attach to Running Container...`) or in your terminal.
|
169 |
+
* The `--privileged` flag gives all capabilities to the container.
|
170 |
+
* The `--shm-size 10g` argument increases the shared memory size. Use this if you see `exitcode: -7` errors using deepspeed.
|
171 |
+
|
172 |
+
[More information on nvidia website](https://docs.nvidia.com/deeplearning/frameworks/user-guide/index.html#setincshmem)
|
173 |
+
|
174 |
+
</details>
|
175 |
+
|
176 |
+
#### Conda/Pip venv
|
177 |
+
1. Install python >=**3.10**
|
178 |
+
|
179 |
+
2. Install pytorch stable https://pytorch.org/get-started/locally/
|
180 |
+
|
181 |
+
3. Install Axolotl along with python dependencies
|
182 |
+
```bash
|
183 |
+
pip3 install packaging
|
184 |
+
pip3 install -e '.[flash-attn,deepspeed]'
|
185 |
+
```
|
186 |
+
4. (Optional) Login to Huggingface to use gated models/datasets.
|
187 |
+
```bash
|
188 |
+
huggingface-cli login
|
189 |
+
```
|
190 |
+
Get the token at huggingface.co/settings/tokens
|
191 |
+
|
192 |
+
#### Cloud GPU
|
193 |
+
|
194 |
+
For cloud GPU providers that support docker images, use [`winglian/axolotl-cloud:main-latest`](https://hub.docker.com/r/winglian/axolotl-cloud/tags)
|
195 |
+
|
196 |
+
- on Latitude.sh use this [direct link](https://latitude.sh/blueprint/989e0e79-3bf6-41ea-a46b-1f246e309d5c)
|
197 |
+
- on JarvisLabs.ai use this [direct link](https://jarvislabs.ai/templates/axolotl)
|
198 |
+
- on RunPod use this [direct link](https://runpod.io/gsc?template=v2ickqhz9s&ref=6i7fkpdz)
|
199 |
+
|
200 |
+
#### Bare Metal Cloud GPU
|
201 |
+
|
202 |
+
##### LambdaLabs
|
203 |
+
|
204 |
+
<details>
|
205 |
+
|
206 |
+
<summary>Click to Expand</summary>
|
207 |
+
|
208 |
+
1. Install python
|
209 |
+
```bash
|
210 |
+
sudo apt update
|
211 |
+
sudo apt install -y python3.10
|
212 |
+
|
213 |
+
sudo update-alternatives --install /usr/bin/python python /usr/bin/python3.10 1
|
214 |
+
sudo update-alternatives --config python # pick 3.10 if given option
|
215 |
+
python -V # should be 3.10
|
216 |
+
|
217 |
+
```
|
218 |
+
|
219 |
+
2. Install pip
|
220 |
+
```bash
|
221 |
+
wget https://bootstrap.pypa.io/get-pip.py
|
222 |
+
python get-pip.py
|
223 |
+
```
|
224 |
+
|
225 |
+
3. Install torch
|
226 |
+
```bash
|
227 |
+
pip3 install -U torch --index-url https://download.pytorch.org/whl/cu118
|
228 |
+
```
|
229 |
+
|
230 |
+
4. Axolotl
|
231 |
+
```bash
|
232 |
+
git clone https://github.com/OpenAccess-AI-Collective/axolotl
|
233 |
+
cd axolotl
|
234 |
+
|
235 |
+
pip3 install packaging
|
236 |
+
pip3 install -e '.[flash-attn,deepspeed]'
|
237 |
+
pip3 install protobuf==3.20.3
|
238 |
+
pip3 install -U --ignore-installed requests Pillow psutil scipy
|
239 |
+
```
|
240 |
+
|
241 |
+
5. Set path
|
242 |
+
```bash
|
243 |
+
export LD_LIBRARY_PATH=/usr/lib/x86_64-linux-gnu:$LD_LIBRARY_PATH
|
244 |
+
```
|
245 |
+
</details>
|
246 |
+
|
247 |
+
##### GCP
|
248 |
+
|
249 |
+
<details>
|
250 |
+
|
251 |
+
<summary>Click to Expand</summary>
|
252 |
+
|
253 |
+
Use a Deeplearning linux OS with cuda and pytorch installed. Then follow instructions on quickstart.
|
254 |
+
|
255 |
+
Make sure to run the below to uninstall xla.
|
256 |
+
```bash
|
257 |
+
pip uninstall -y torch_xla[tpu]
|
258 |
+
```
|
259 |
+
|
260 |
+
</details>
|
261 |
+
|
262 |
+
#### Windows
|
263 |
+
Please use WSL or Docker!
|
264 |
+
|
265 |
+
#### Mac
|
266 |
+
|
267 |
+
Use the below instead of the install method in QuickStart.
|
268 |
+
```
|
269 |
+
pip3 install -e '.'
|
270 |
+
```
|
271 |
+
More info: [mac.md](/docs/mac.qmd)
|
272 |
+
|
273 |
+
#### Google Colab
|
274 |
+
|
275 |
+
Please use this example [notebook](examples/colab-notebooks/colab-axolotl-example.ipynb).
|
276 |
+
|
277 |
+
#### Launching on public clouds via SkyPilot
|
278 |
+
To launch on GPU instances (both on-demand and spot instances) on 7+ clouds (GCP, AWS, Azure, OCI, and more), you can use [SkyPilot](https://skypilot.readthedocs.io/en/latest/index.html):
|
279 |
+
|
280 |
+
```bash
|
281 |
+
pip install "skypilot-nightly[gcp,aws,azure,oci,lambda,kubernetes,ibm,scp]" # choose your clouds
|
282 |
+
sky check
|
283 |
+
```
|
284 |
+
|
285 |
+
Get the [example YAMLs](https://github.com/skypilot-org/skypilot/tree/master/llm/axolotl) of using Axolotl to finetune `mistralai/Mistral-7B-v0.1`:
|
286 |
+
```
|
287 |
+
git clone https://github.com/skypilot-org/skypilot.git
|
288 |
+
cd skypilot/llm/axolotl
|
289 |
+
```
|
290 |
+
|
291 |
+
Use one command to launch:
|
292 |
+
```bash
|
293 |
+
# On-demand
|
294 |
+
HF_TOKEN=xx sky launch axolotl.yaml --env HF_TOKEN
|
295 |
+
|
296 |
+
# Managed spot (auto-recovery on preemption)
|
297 |
+
HF_TOKEN=xx BUCKET=<unique-name> sky spot launch axolotl-spot.yaml --env HF_TOKEN --env BUCKET
|
298 |
+
```
|
299 |
+
|
300 |
+
### Dataset
|
301 |
+
|
302 |
+
Axolotl supports a variety of dataset formats. Below are some of the formats you can use.
|
303 |
+
Have dataset(s) in one of the following format (JSONL recommended):
|
304 |
+
|
305 |
+
#### Pretraining
|
306 |
+
|
307 |
+
- `completion`: raw corpus
|
308 |
+
```json
|
309 |
+
{"text": "..."}
|
310 |
+
```
|
311 |
+
|
312 |
+
Note: Axolotl usually loads the entire dataset into memory. This will be challenging for large datasets. Use the following config to enable streaming:
|
313 |
+
|
314 |
+
```yaml
|
315 |
+
pretraining_dataset: # hf path only
|
316 |
+
```
|
317 |
+
|
318 |
+
#### Supervised finetuning
|
319 |
+
|
320 |
+
##### Instruction
|
321 |
+
|
322 |
+
- `alpaca`: instruction; input(optional)
|
323 |
+
```json
|
324 |
+
{"instruction": "...", "input": "...", "output": "..."}
|
325 |
+
```
|
326 |
+
|
327 |
+
<details>
|
328 |
+
|
329 |
+
<summary>See other formats</summary>
|
330 |
+
|
331 |
+
- `jeopardy`: question and answer
|
332 |
+
```json
|
333 |
+
{"question": "...", "category": "...", "answer": "..."}
|
334 |
+
```
|
335 |
+
- `oasst`: instruction
|
336 |
+
```json
|
337 |
+
{"INSTRUCTION": "...", "RESPONSE": "..."}
|
338 |
+
```
|
339 |
+
- `gpteacher`: instruction; input(optional)
|
340 |
+
```json
|
341 |
+
{"instruction": "...", "input": "...", "response": "..."}
|
342 |
+
```
|
343 |
+
- `reflection`: instruction with reflect; input(optional)
|
344 |
+
```json
|
345 |
+
{"instruction": "...", "input": "...", "output": "...", "reflection": "...", "corrected": "..."}
|
346 |
+
```
|
347 |
+
- `explainchoice`: question, choices, (solution OR explanation)
|
348 |
+
```json
|
349 |
+
{"question": "...", "choices": ["..."], "solution": "...", "explanation": "..."}
|
350 |
+
```
|
351 |
+
- `concisechoice`: question, choices, (solution OR explanation)
|
352 |
+
```json
|
353 |
+
{"question": "...", "choices": ["..."], "solution": "...", "explanation": "..."}
|
354 |
+
```
|
355 |
+
- `summarizetldr`: article and summary
|
356 |
+
```json
|
357 |
+
{"article": "...", "summary": "..."}
|
358 |
+
```
|
359 |
+
- `alpaca_chat`: basic instruct for alpaca chat
|
360 |
+
```json
|
361 |
+
{"instruction": "...", "input": "...", "response": "..."}
|
362 |
+
```
|
363 |
+
- `alpaca_chat.load_qa`: question and answer for alpaca chat
|
364 |
+
```json
|
365 |
+
{"question": "...", "answer": "..."}
|
366 |
+
```
|
367 |
+
- `alpaca_chat.load_concise`: question and answer for alpaca chat, for concise answers
|
368 |
+
```json
|
369 |
+
{"instruction": "...", "input": "...", "response": "..."}
|
370 |
+
```
|
371 |
+
- `alpaca_chat.load_camel_ai`: question and answer for alpaca chat, for load_camel_ai
|
372 |
+
```json
|
373 |
+
{"message_1": "...", "message_2": "..."}
|
374 |
+
```
|
375 |
+
- `alpaca_w_system.load_open_orca`: support for open orca datasets with included system prompts, instruct
|
376 |
+
```json
|
377 |
+
{"system_prompt": "...", "question": "...", "response": "..."}
|
378 |
+
```
|
379 |
+
- `context_qa`: in context question answering from an article
|
380 |
+
```json
|
381 |
+
{"article": "...", "question": "...", "answer": "..."}
|
382 |
+
```
|
383 |
+
- `context_qa.load_v2`: in context question answering (alternate)
|
384 |
+
```json
|
385 |
+
{"context": "...", "question": "...", "answer": "..."}
|
386 |
+
```
|
387 |
+
- `context_qa.load_404`: in context question answering from an article, with default response for no answer from context
|
388 |
+
```json
|
389 |
+
{"article": "...", "unanswerable_question": "..."}
|
390 |
+
```
|
391 |
+
- `creative_acr.load_answer`: instruction and revision
|
392 |
+
```json
|
393 |
+
{"instruction": "...", "revision": "..."}
|
394 |
+
```
|
395 |
+
- `creative_acr.load_critique`: critique
|
396 |
+
```json
|
397 |
+
{"scores": "...", "critiques": "...", "instruction": "...", "answer": "..."}
|
398 |
+
```
|
399 |
+
- `creative_acr.load_revise`: critique and revise
|
400 |
+
```json
|
401 |
+
{"scores": "...", "critiques": "...", "instruction": "...", "answer": "...", "revision": "..."}
|
402 |
+
```
|
403 |
+
- `metharme`: instruction, adds additional eos tokens
|
404 |
+
```json
|
405 |
+
{"prompt": "...", "generation": "..."}
|
406 |
+
```
|
407 |
+
|
408 |
+
</details>
|
409 |
+
|
410 |
+
##### Template-Free
|
411 |
+
|
412 |
+
- `input_output`: template-free prompt construction
|
413 |
+
```json
|
414 |
+
{"segments": [{"label": true|false, "text": "..."}]}
|
415 |
+
```
|
416 |
+
|
417 |
+
This is a special format that allows you to construct prompts without using templates. This is for advanced users who want more freedom with prompt construction. See [these docs](docs/input_output.qmd) for more details.
|
418 |
+
|
419 |
+
##### Conversation
|
420 |
+
|
421 |
+
- `sharegpt`: conversations where `from` is `human`/`gpt`. (optional: first row with role `system` to override default system prompt)
|
422 |
+
```json
|
423 |
+
{"conversations": [{"from": "...", "value": "..."}]}
|
424 |
+
```
|
425 |
+
|
426 |
+
<details>
|
427 |
+
|
428 |
+
<summary>See other formats</summary>
|
429 |
+
|
430 |
+
- `pygmalion`: pygmalion
|
431 |
+
```json
|
432 |
+
{"conversations": [{"role": "...", "value": "..."}]}
|
433 |
+
```
|
434 |
+
- `sharegpt.load_role`: conversations where `role` is used instead of `from`
|
435 |
+
```json
|
436 |
+
{"conversations": [{"role": "...", "value": "..."}]}
|
437 |
+
```
|
438 |
+
- `sharegpt.load_guanaco`: conversations where `from` is `prompter`/`assistant` instead of default sharegpt
|
439 |
+
```json
|
440 |
+
{"conversations": [{"from": "...", "value": "..."}]}
|
441 |
+
```
|
442 |
+
- `sharegpt_jokes`: creates a chat where bot is asked to tell a joke, then explain why the joke is funny
|
443 |
+
```json
|
444 |
+
{"conversations": [{"title": "...", "text": "...", "explanation": "..."}]}
|
445 |
+
```
|
446 |
+
|
447 |
+
</details>
|
448 |
+
|
449 |
+
Note: `type: sharegpt` opens a special config `conversation:` that enables conversions to many Conversation types. See dataset section under [all yaml options](#all-yaml-options).
|
450 |
+
|
451 |
+
#### How to add custom prompts
|
452 |
+
|
453 |
+
For a dataset that is preprocessed for instruction purposes:
|
454 |
+
|
455 |
+
```json
|
456 |
+
{"input": "...", "output": "..."}
|
457 |
+
```
|
458 |
+
|
459 |
+
You can use this example in your YAML config:
|
460 |
+
|
461 |
+
```yaml
|
462 |
+
datasets:
|
463 |
+
- path: repo
|
464 |
+
type:
|
465 |
+
system_prompt: ""
|
466 |
+
field_system: system
|
467 |
+
field_instruction: input
|
468 |
+
field_output: output
|
469 |
+
format: "[INST] {instruction} [/INST]"
|
470 |
+
no_input_format: "[INST] {instruction} [/INST]"
|
471 |
+
```
|
472 |
+
See full config options under [all yaml options](#all-yaml-options).
|
473 |
+
|
474 |
+
#### How to use your custom pretokenized dataset
|
475 |
+
|
476 |
+
- Do not pass a `type:`
|
477 |
+
- Columns in Dataset must be exactly `input_ids`, `attention_mask`, `labels`
|
478 |
+
|
479 |
+
```yaml
|
480 |
+
- path: ...
|
481 |
+
```
|
482 |
+
|
483 |
+
### Config
|
484 |
+
|
485 |
+
See [examples](examples) for quick start. It is recommended to duplicate and modify to your needs. The most important options are:
|
486 |
+
|
487 |
+
- model
|
488 |
+
```yaml
|
489 |
+
base_model: ./llama-7b-hf # local or huggingface repo
|
490 |
+
```
|
491 |
+
Note: The code will load the right architecture.
|
492 |
+
|
493 |
+
- dataset
|
494 |
+
```yaml
|
495 |
+
datasets:
|
496 |
+
# huggingface repo
|
497 |
+
- path: vicgalle/alpaca-gpt4
|
498 |
+
type: alpaca
|
499 |
+
|
500 |
+
# huggingface repo with specific configuration/subset
|
501 |
+
- path: EleutherAI/pile
|
502 |
+
name: enron_emails
|
503 |
+
type: completion # format from earlier
|
504 |
+
field: text # Optional[str] default: text, field to use for completion data
|
505 |
+
|
506 |
+
# huggingface repo with multiple named configurations/subsets
|
507 |
+
- path: bigcode/commitpackft
|
508 |
+
name:
|
509 |
+
- ruby
|
510 |
+
- python
|
511 |
+
- typescript
|
512 |
+
type: ... # unimplemented custom format
|
513 |
+
|
514 |
+
# fastchat conversation
|
515 |
+
# See 'conversation' options: https://github.com/lm-sys/FastChat/blob/main/fastchat/conversation.py
|
516 |
+
- path: ...
|
517 |
+
type: sharegpt
|
518 |
+
conversation: chatml # default: vicuna_v1.1
|
519 |
+
|
520 |
+
# local
|
521 |
+
- path: data.jsonl # or json
|
522 |
+
ds_type: json # see other options below
|
523 |
+
type: alpaca
|
524 |
+
|
525 |
+
# dataset with splits, but no train split
|
526 |
+
- path: knowrohit07/know_sql
|
527 |
+
type: context_qa.load_v2
|
528 |
+
train_on_split: validation
|
529 |
+
|
530 |
+
# loading from s3 or gcs
|
531 |
+
# s3 creds will be loaded from the system default and gcs only supports public access
|
532 |
+
- path: s3://path_to_ds # Accepts folder with arrow/parquet or file path like above. Supports s3, gcs.
|
533 |
+
...
|
534 |
+
|
535 |
+
# Loading Data From a Public URL
|
536 |
+
# - The file format is `json` (which includes `jsonl`) by default. For different formats, adjust the `ds_type` option accordingly.
|
537 |
+
- path: https://some.url.com/yourdata.jsonl # The URL should be a direct link to the file you wish to load. URLs must use HTTPS protocol, not HTTP.
|
538 |
+
ds_type: json # this is the default, see other options below.
|
539 |
+
```
|
540 |
+
|
541 |
+
- loading
|
542 |
+
```yaml
|
543 |
+
load_in_4bit: true
|
544 |
+
load_in_8bit: true
|
545 |
+
|
546 |
+
bf16: auto # require >=ampere, auto will detect if your GPU supports this and choose automatically.
|
547 |
+
fp16: # leave empty to use fp16 when bf16 is 'auto'. set to false if you want to fallback to fp32
|
548 |
+
tf32: true # require >=ampere
|
549 |
+
|
550 |
+
bfloat16: true # require >=ampere, use instead of bf16 when you don't want AMP (automatic mixed precision)
|
551 |
+
float16: true # use instead of fp16 when you don't want AMP
|
552 |
+
```
|
553 |
+
Note: Repo does not do 4-bit quantization.
|
554 |
+
|
555 |
+
- lora
|
556 |
+
```yaml
|
557 |
+
adapter: lora # 'qlora' or leave blank for full finetune
|
558 |
+
lora_r: 8
|
559 |
+
lora_alpha: 16
|
560 |
+
lora_dropout: 0.05
|
561 |
+
lora_target_modules:
|
562 |
+
- q_proj
|
563 |
+
- v_proj
|
564 |
+
```
|
565 |
+
|
566 |
+
<details id="all-yaml-options">
|
567 |
+
|
568 |
+
<summary>All yaml options (click to expand)</summary>
|
569 |
+
|
570 |
+
```yaml
|
571 |
+
# This is the huggingface model that contains *.pt, *.safetensors, or *.bin files
|
572 |
+
# This can also be a relative path to a model on disk
|
573 |
+
base_model: ./llama-7b-hf
|
574 |
+
# You can specify an ignore pattern if the model repo contains more than 1 model type (*.pt, etc)
|
575 |
+
base_model_ignore_patterns:
|
576 |
+
# If the base_model repo on hf hub doesn't include configuration .json files,
|
577 |
+
# You can set that here, or leave this empty to default to base_model
|
578 |
+
base_model_config: ./llama-7b-hf
|
579 |
+
# You can specify to choose a specific model revision from huggingface hub
|
580 |
+
revision_of_model:
|
581 |
+
# Optional tokenizer configuration path in case you want to use a different tokenizer
|
582 |
+
# than the one defined in the base model
|
583 |
+
tokenizer_config:
|
584 |
+
# If you want to specify the type of model to load, AutoModelForCausalLM is a good choice too
|
585 |
+
model_type: AutoModelForCausalLM
|
586 |
+
# Corresponding tokenizer for the model AutoTokenizer is a good choice
|
587 |
+
tokenizer_type: AutoTokenizer
|
588 |
+
# Trust remote code for untrusted source
|
589 |
+
trust_remote_code:
|
590 |
+
# use_fast option for tokenizer loading from_pretrained, default to True
|
591 |
+
tokenizer_use_fast:
|
592 |
+
# Whether to use the legacy tokenizer setting, defaults to True
|
593 |
+
tokenizer_legacy:
|
594 |
+
# Resize the model embeddings when new tokens are added to multiples of 32
|
595 |
+
# This is reported to improve training speed on some models
|
596 |
+
resize_token_embeddings_to_32x:
|
597 |
+
|
598 |
+
# (Internal use only)
|
599 |
+
# Used to identify which the model is based on
|
600 |
+
is_falcon_derived_model:
|
601 |
+
is_llama_derived_model:
|
602 |
+
is_qwen_derived_model:
|
603 |
+
# Please note that if you set this to true, `padding_side` will be set to "left" by default
|
604 |
+
is_mistral_derived_model:
|
605 |
+
|
606 |
+
# optional overrides to the base model configuration
|
607 |
+
overrides_of_model_config:
|
608 |
+
# RoPE Scaling https://github.com/huggingface/transformers/pull/24653
|
609 |
+
rope_scaling:
|
610 |
+
type: # linear | dynamic
|
611 |
+
factor: # float
|
612 |
+
|
613 |
+
# optional overrides to the bnb 4bit quantization configuration
|
614 |
+
# https://huggingface.co/docs/transformers/main/main_classes/quantization#transformers.BitsAndBytesConfig
|
615 |
+
bnb_config_kwargs:
|
616 |
+
# These are default values
|
617 |
+
llm_int8_has_fp16_weight: false
|
618 |
+
bnb_4bit_quant_type: nf4
|
619 |
+
bnb_4bit_use_double_quant: true
|
620 |
+
|
621 |
+
|
622 |
+
# Whether you are training a 4-bit GPTQ quantized model
|
623 |
+
gptq: true
|
624 |
+
|
625 |
+
# This will attempt to quantize the model down to 8 bits and use adam 8 bit optimizer
|
626 |
+
load_in_8bit: true
|
627 |
+
# Use bitsandbytes 4 bit
|
628 |
+
load_in_4bit:
|
629 |
+
|
630 |
+
# Use CUDA bf16
|
631 |
+
bf16: true # bool or 'full' for `bf16_full_eval`. require >=ampere
|
632 |
+
# Use CUDA fp16
|
633 |
+
fp16: true
|
634 |
+
# Use CUDA tf32
|
635 |
+
tf32: true # require >=ampere
|
636 |
+
|
637 |
+
# No AMP (automatic mixed precision)
|
638 |
+
bfloat16: true # require >=ampere
|
639 |
+
float16: true
|
640 |
+
|
641 |
+
# Limit the memory for all available GPUs to this amount (if an integer, expressed in gigabytes); default: unset
|
642 |
+
gpu_memory_limit: 20GiB
|
643 |
+
# Do the LoRA/PEFT loading on CPU -- this is required if the base model is so large it takes up most or all of the available GPU VRAM, e.g. during a model and LoRA merge
|
644 |
+
lora_on_cpu: true
|
645 |
+
|
646 |
+
# A list of one or more datasets to finetune the model with
|
647 |
+
datasets:
|
648 |
+
# HuggingFace dataset repo | s3://,gs:// path | "json" for local dataset, make sure to fill data_files
|
649 |
+
- path: vicgalle/alpaca-gpt4
|
650 |
+
# The type of prompt to use for training. [alpaca, sharegpt, gpteacher, oasst, reflection]
|
651 |
+
type: alpaca # format | format:<prompt_style> (chat/instruct) | <prompt_strategies>.load_<load_fn>
|
652 |
+
ds_type: # Optional[str] (json|arrow|parquet|text|csv) defines the datatype when path is a file
|
653 |
+
data_files: # Optional[str] path to source data files
|
654 |
+
shards: # Optional[int] number of shards to split data into
|
655 |
+
name: # Optional[str] name of dataset configuration to load
|
656 |
+
train_on_split: train # Optional[str] name of dataset split to load from
|
657 |
+
|
658 |
+
# Optional[str] fastchat conversation type, only used with type: sharegpt
|
659 |
+
conversation: # Options (see Conversation 'name'): https://github.com/lm-sys/FastChat/blob/main/fastchat/conversation.py
|
660 |
+
field_human: # Optional[str]. Human key to use for conversation.
|
661 |
+
field_model: # Optional[str]. Assistant key to use for conversation.
|
662 |
+
# Add additional keys from your dataset as input or output roles
|
663 |
+
roles:
|
664 |
+
input: # Optional[List[str]]. These will be masked based on train_on_input
|
665 |
+
output: # Optional[List[str]].
|
666 |
+
|
667 |
+
# Custom user instruction prompt
|
668 |
+
- path: repo
|
669 |
+
type:
|
670 |
+
# The below are defaults. only set what's needed if you use a different column name.
|
671 |
+
system_prompt: ""
|
672 |
+
system_format: "{system}"
|
673 |
+
field_system: system
|
674 |
+
field_instruction: instruction
|
675 |
+
field_input: input
|
676 |
+
field_output: output
|
677 |
+
|
678 |
+
# Customizable to be single line or multi-line
|
679 |
+
# Use {instruction}/{input} as key to be replaced
|
680 |
+
# 'format' can include {input}
|
681 |
+
format: |-
|
682 |
+
User: {instruction} {input}
|
683 |
+
Assistant:
|
684 |
+
# 'no_input_format' cannot include {input}
|
685 |
+
no_input_format: "{instruction} "
|
686 |
+
|
687 |
+
# For `completion` datsets only, uses the provided field instead of `text` column
|
688 |
+
field:
|
689 |
+
|
690 |
+
# If false, the datasets will not be shuffled and will keep their original order in `datasets`.
|
691 |
+
# The same applies to the `test_datasets` option and the `pretraining_dataset` option. Default is true.
|
692 |
+
shuffle_merged_datasets: true
|
693 |
+
|
694 |
+
# A list of one or more datasets to eval the model with.
|
695 |
+
# You can use either test_datasets, or val_set_size, but not both.
|
696 |
+
test_datasets:
|
697 |
+
- path: /workspace/data/eval.jsonl
|
698 |
+
ds_type: json
|
699 |
+
# You need to specify a split. For "json" datasets the default split is called "train".
|
700 |
+
split: train
|
701 |
+
type: completion
|
702 |
+
data_files:
|
703 |
+
- /workspace/data/eval.jsonl
|
704 |
+
|
705 |
+
# use RL training: 'dpo', 'ipo', 'kto_pair'
|
706 |
+
rl:
|
707 |
+
|
708 |
+
# Saves the desired chat template to the tokenizer_config.json for easier inferencing
|
709 |
+
# Currently supports chatml and inst (mistral/mixtral)
|
710 |
+
chat_template: chatml
|
711 |
+
# Changes the default system message
|
712 |
+
default_system_message: You are a helpful assistant. Please give a long and detailed answer. # Currently only supports chatml.
|
713 |
+
# Axolotl attempts to save the dataset as an arrow after packing the data together so
|
714 |
+
# subsequent training attempts load faster, relative path
|
715 |
+
dataset_prepared_path: data/last_run_prepared
|
716 |
+
# Push prepared dataset to hub
|
717 |
+
push_dataset_to_hub: # repo path
|
718 |
+
# The maximum number of processes to use while preprocessing your input dataset. This defaults to `os.cpu_count()`
|
719 |
+
# if not set.
|
720 |
+
dataset_processes: # defaults to os.cpu_count() if not set
|
721 |
+
# Keep dataset in memory while preprocessing
|
722 |
+
# Only needed if cached dataset is taking too much storage
|
723 |
+
dataset_keep_in_memory:
|
724 |
+
# push checkpoints to hub
|
725 |
+
hub_model_id: # private repo path to push finetuned model
|
726 |
+
# how to push checkpoints to hub
|
727 |
+
# https://huggingface.co/docs/transformers/v4.31.0/en/main_classes/trainer#transformers.TrainingArguments.hub_strategy
|
728 |
+
hub_strategy:
|
729 |
+
# Whether to use hf `use_auth_token` for loading datasets. Useful for fetching private datasets
|
730 |
+
# Required to be true when used in combination with `push_dataset_to_hub`
|
731 |
+
hf_use_auth_token: # boolean
|
732 |
+
# How much of the dataset to set aside as evaluation. 1 = 100%, 0.50 = 50%, etc. 0 for no eval.
|
733 |
+
val_set_size: 0.04
|
734 |
+
# Num shards for whole dataset
|
735 |
+
dataset_shard_num:
|
736 |
+
# Index of shard to use for whole dataset
|
737 |
+
dataset_shard_idx:
|
738 |
+
|
739 |
+
# The maximum length of an input to train with, this should typically be less than 2048
|
740 |
+
# as most models have a token/context limit of 2048
|
741 |
+
sequence_len: 2048
|
742 |
+
# Pad inputs so each step uses constant sized buffers
|
743 |
+
# This will reduce memory fragmentation and may prevent OOMs, by re-using memory more efficiently
|
744 |
+
pad_to_sequence_len:
|
745 |
+
# Use efficient multi-packing with block diagonal attention and per sequence position_ids. Recommend set to 'true'
|
746 |
+
sample_packing:
|
747 |
+
# Set to 'false' if getting errors during eval with sample_packing on.
|
748 |
+
eval_sample_packing:
|
749 |
+
# You can set these packing optimizations AFTER starting a training at least once.
|
750 |
+
# The trainer will provide recommended values for these values.
|
751 |
+
sample_packing_eff_est:
|
752 |
+
total_num_tokens:
|
753 |
+
|
754 |
+
# Passed through to transformers when loading the model when launched without accelerate
|
755 |
+
# Use `sequential` when training w/ model parallelism to limit memory
|
756 |
+
device_map:
|
757 |
+
# Defines the max memory usage per gpu on the system. Passed through to transformers when loading the model.
|
758 |
+
max_memory:
|
759 |
+
|
760 |
+
# If you want to use 'lora' or 'qlora' or leave blank to train all parameters in original model
|
761 |
+
adapter: lora
|
762 |
+
# If you already have a lora model trained that you want to load, put that here.
|
763 |
+
# This means after training, if you want to test the model, you should set this to the value of `output_dir`.
|
764 |
+
# Note that if you merge an adapter to the base model, a new subdirectory `merged` will be created under the `output_dir`.
|
765 |
+
lora_model_dir:
|
766 |
+
|
767 |
+
# LoRA hyperparameters
|
768 |
+
# For more details about the following options, see:
|
769 |
+
# https://www.anyscale.com/blog/fine-tuning-llms-lora-or-full-parameter-an-in-depth-analysis-with-llama-2
|
770 |
+
lora_r: 8
|
771 |
+
lora_alpha: 16
|
772 |
+
lora_dropout: 0.05
|
773 |
+
lora_target_modules:
|
774 |
+
- q_proj
|
775 |
+
- v_proj
|
776 |
+
# - k_proj
|
777 |
+
# - o_proj
|
778 |
+
# - gate_proj
|
779 |
+
# - down_proj
|
780 |
+
# - up_proj
|
781 |
+
lora_target_linear: # If true, will target all linear modules
|
782 |
+
peft_layers_to_transform: # The layer indices to transform, otherwise, apply to all layers
|
783 |
+
|
784 |
+
# If you added new tokens to the tokenizer, you may need to save some LoRA modules because they need to know the new tokens.
|
785 |
+
# For LLaMA and Mistral, you need to save `embed_tokens` and `lm_head`. It may vary for other models.
|
786 |
+
# `embed_tokens` converts tokens to embeddings, and `lm_head` converts embeddings to token probabilities.
|
787 |
+
# https://github.com/huggingface/peft/issues/334#issuecomment-1561727994
|
788 |
+
lora_modules_to_save:
|
789 |
+
# - embed_tokens
|
790 |
+
# - lm_head
|
791 |
+
|
792 |
+
lora_fan_in_fan_out: false
|
793 |
+
|
794 |
+
peft:
|
795 |
+
# Configuration options for loftq initialization for LoRA
|
796 |
+
# https://huggingface.co/docs/peft/developer_guides/quantization#loftq-initialization
|
797 |
+
loftq_config:
|
798 |
+
loftq_bits: # typically 4 bits
|
799 |
+
|
800 |
+
# ReLoRA configuration
|
801 |
+
# Must use either 'lora' or 'qlora' adapter, and does not support fsdp or deepspeed
|
802 |
+
relora_steps: # Number of steps per ReLoRA restart
|
803 |
+
relora_warmup_steps: # Number of per-restart warmup steps
|
804 |
+
relora_anneal_steps: # Number of anneal steps for each relora cycle
|
805 |
+
relora_prune_ratio: # threshold for optimizer magnitude when pruning
|
806 |
+
relora_cpu_offload: # True to perform lora weight merges on cpu during restarts, for modest gpu memory savings
|
807 |
+
|
808 |
+
# wandb configuration if you're using it
|
809 |
+
# Make sure your `WANDB_API_KEY` environment variable is set (recommended) or you login to wandb with `wandb login`.
|
810 |
+
wandb_mode: # "offline" to save run metadata locally and not sync to the server, "disabled" to turn off wandb
|
811 |
+
wandb_project: # Your wandb project name
|
812 |
+
wandb_entity: # A wandb Team name if using a Team
|
813 |
+
wandb_watch:
|
814 |
+
wandb_name: # Set the name of your wandb run
|
815 |
+
wandb_run_id: # Set the ID of your wandb run
|
816 |
+
wandb_log_model: # "checkpoint" to log model to wandb Artifacts every `save_steps` or "end" to log only at the end of training
|
817 |
+
|
818 |
+
# mlflow configuration if you're using it
|
819 |
+
mlflow_tracking_uri: # URI to mlflow
|
820 |
+
mlflow_experiment_name: # Your experiment name
|
821 |
+
hf_mlflow_log_artifacts: # set to true to copy each saved checkpoint on each save to mlflow artifact registry
|
822 |
+
|
823 |
+
# Where to save the full-finetuned model to
|
824 |
+
output_dir: ./completed-model
|
825 |
+
|
826 |
+
# Whether to use torch.compile and which backend to use
|
827 |
+
torch_compile: # bool
|
828 |
+
torch_compile_backend: # Optional[str]
|
829 |
+
|
830 |
+
# Training hyperparameters
|
831 |
+
|
832 |
+
# If greater than 1, backpropagation will be skipped and the gradients will be accumulated for the given number of steps.
|
833 |
+
gradient_accumulation_steps: 1
|
834 |
+
# The number of samples to include in each batch. This is the number of samples sent to each GPU.
|
835 |
+
micro_batch_size: 2
|
836 |
+
eval_batch_size:
|
837 |
+
num_epochs: 4
|
838 |
+
warmup_steps: 100 # cannot use with warmup_ratio
|
839 |
+
warmup_ratio: 0.05 # cannot use with warmup_steps
|
840 |
+
learning_rate: 0.00003
|
841 |
+
lr_quadratic_warmup:
|
842 |
+
logging_steps:
|
843 |
+
eval_steps: # Leave empty to eval at each epoch, integers for every N steps. decimal for fraction of total steps
|
844 |
+
evals_per_epoch: # number of times per epoch to run evals, mutually exclusive with eval_steps
|
845 |
+
save_strategy: # Set to `no` to skip checkpoint saves
|
846 |
+
save_steps: # Leave empty to save at each epoch
|
847 |
+
saves_per_epoch: # number of times per epoch to save a checkpoint, mutually exclusive with save_steps
|
848 |
+
save_total_limit: # Checkpoints saved at a time
|
849 |
+
# Maximum number of iterations to train for. It precedes num_epochs which means that
|
850 |
+
# if both are set, num_epochs will not be guaranteed.
|
851 |
+
# e.g., when 1 epoch is 1000 steps => `num_epochs: 2` and `max_steps: 100` will train for 100 steps
|
852 |
+
max_steps:
|
853 |
+
|
854 |
+
eval_table_size: # Approximate number of predictions sent to wandb depending on batch size. Enabled above 0. Default is 0
|
855 |
+
eval_max_new_tokens: # Total number of tokens generated for predictions sent to wandb. Default is 128
|
856 |
+
eval_causal_lm_metrics: # HF evaluate metrics used during evaluation. Default is ["sacrebleu", "comet", "ter", chrf]
|
857 |
+
|
858 |
+
loss_watchdog_threshold: # High loss value, indicating the learning has broken down (a good estimate is ~2 times the loss at the start of training)
|
859 |
+
loss_watchdog_patience: # Number of high-loss steps in a row before the trainer aborts (default: 3)
|
860 |
+
|
861 |
+
# Save model as safetensors (require safetensors package)
|
862 |
+
save_safetensors:
|
863 |
+
|
864 |
+
# Whether to mask out or include the human's prompt from the training labels
|
865 |
+
train_on_inputs: false
|
866 |
+
# Group similarly sized data to minimize padding.
|
867 |
+
# May be slower to start, as it must download and sort the entire dataset.
|
868 |
+
# Note that training loss may have an oscillating pattern with this enabled.
|
869 |
+
group_by_length: false
|
870 |
+
|
871 |
+
# Whether to use gradient checkpointing https://huggingface.co/docs/transformers/v4.18.0/en/performance#gradient-checkpointing
|
872 |
+
gradient_checkpointing: false
|
873 |
+
# additional kwargs to pass to the trainer for gradient checkpointing
|
874 |
+
# gradient_checkpointing_kwargs:
|
875 |
+
# use_reentrant: true
|
876 |
+
|
877 |
+
# Stop training after this many evaluation losses have increased in a row
|
878 |
+
# https://huggingface.co/transformers/v4.2.2/_modules/transformers/trainer_callback.html#EarlyStoppingCallback
|
879 |
+
early_stopping_patience: 3
|
880 |
+
|
881 |
+
# Specify a scheduler and kwargs to use with the optimizer
|
882 |
+
lr_scheduler: # 'one_cycle' | 'log_sweep' | empty for cosine
|
883 |
+
lr_scheduler_kwargs:
|
884 |
+
cosine_min_lr_ratio: # decay lr to some percentage of the peak lr, e.g. cosine_min_lr_ratio=0.1 for 10% of peak lr
|
885 |
+
cosine_constant_lr_ratio: # freeze lr at some percentage of the step, e.g. cosine_constant_lr_ratio=0.8 means start cosine_min_lr at 80% of training step (https://arxiv.org/pdf/2308.04014.pdf)
|
886 |
+
|
887 |
+
# For one_cycle optim
|
888 |
+
lr_div_factor: # Learning rate div factor
|
889 |
+
|
890 |
+
# Specify optimizer
|
891 |
+
# Valid values are driven by the Transformers OptimizerNames class, see:
|
892 |
+
# https://github.com/huggingface/transformers/blob/95b374952dc27d8511541d6f5a4e22c9ec11fb24/src/transformers/training_args.py#L134
|
893 |
+
#
|
894 |
+
# Note that not all optimizers may be available in your environment, ex: 'adamw_anyprecision' is part of
|
895 |
+
# torchdistx, 'adamw_bnb_8bit' is part of bnb.optim.Adam8bit, etc. When in doubt, it is recommended to start with the optimizer used
|
896 |
+
# in the examples/ for your model and fine-tuning use case.
|
897 |
+
#
|
898 |
+
# Valid values for 'optimizer' include:
|
899 |
+
# - adamw_hf
|
900 |
+
# - adamw_torch
|
901 |
+
# - adamw_torch_fused
|
902 |
+
# - adamw_torch_xla
|
903 |
+
# - adamw_apex_fused
|
904 |
+
# - adafactor
|
905 |
+
# - adamw_anyprecision
|
906 |
+
# - sgd
|
907 |
+
# - adagrad
|
908 |
+
# - adamw_bnb_8bit
|
909 |
+
# - lion_8bit
|
910 |
+
# - lion_32bit
|
911 |
+
# - paged_adamw_32bit
|
912 |
+
# - paged_adamw_8bit
|
913 |
+
# - paged_lion_32bit
|
914 |
+
# - paged_lion_8bit
|
915 |
+
# - galore_adamw
|
916 |
+
# - galore_adamw_8bit
|
917 |
+
# - galore_adafactor
|
918 |
+
# - galore_adamw_layerwise
|
919 |
+
# - galore_adamw_8bit_layerwise
|
920 |
+
# - galore_adafactor_layerwise
|
921 |
+
optimizer:
|
922 |
+
# Dictionary of arguments to pass to the optimizer
|
923 |
+
optim_args:
|
924 |
+
# For Galore Optimizers the following optim_args are available
|
925 |
+
# rank: # type: int
|
926 |
+
# update_proj_gap # type: int
|
927 |
+
# scale # type: float
|
928 |
+
# proj_type: # type: str, default = std
|
929 |
+
|
930 |
+
# The target modules to optimize, i.e. the module names that you would like to train, right now this is used only for GaLore algorithm
|
931 |
+
optim_target_modules:
|
932 |
+
# - self_attn # for llama
|
933 |
+
# - mlp
|
934 |
+
|
935 |
+
# Specify weight decay
|
936 |
+
weight_decay:
|
937 |
+
# adamw hyperparams
|
938 |
+
adam_beta1:
|
939 |
+
adam_beta2:
|
940 |
+
adam_epsilon:
|
941 |
+
# Gradient clipping max norm
|
942 |
+
max_grad_norm:
|
943 |
+
|
944 |
+
# Augmentation techniques
|
945 |
+
# NEFT https://arxiv.org/abs/2310.05914, set this to a number (paper default is 5) to add noise to embeddings
|
946 |
+
# currently only supported on Llama and Mistral
|
947 |
+
neftune_noise_alpha:
|
948 |
+
|
949 |
+
# Whether to bettertransformers
|
950 |
+
flash_optimum:
|
951 |
+
# Whether to use xformers attention patch https://github.com/facebookresearch/xformers:
|
952 |
+
xformers_attention:
|
953 |
+
# Whether to use flash attention patch https://github.com/Dao-AILab/flash-attention:
|
954 |
+
flash_attention:
|
955 |
+
flash_attn_cross_entropy: # Whether to use flash-attention cross entropy implementation - advanced use only
|
956 |
+
flash_attn_rms_norm: # Whether to use flash-attention rms norm implementation - advanced use only
|
957 |
+
flash_attn_fuse_qkv: # Whether to fuse QKV into a single operation
|
958 |
+
flash_attn_fuse_mlp: # Whether to fuse part of the MLP into a single operation
|
959 |
+
# Whether to use scaled-dot-product attention
|
960 |
+
# https://pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html
|
961 |
+
sdp_attention:
|
962 |
+
# Shifted-sparse attention (only llama) - https://arxiv.org/pdf/2309.12307.pdf
|
963 |
+
s2_attention:
|
964 |
+
# Resume from a specific checkpoint dir
|
965 |
+
resume_from_checkpoint:
|
966 |
+
# If resume_from_checkpoint isn't set and you simply want it to start where it left off.
|
967 |
+
# Be careful with this being turned on between different models.
|
968 |
+
auto_resume_from_checkpoints: false
|
969 |
+
|
970 |
+
# Don't mess with this, it's here for accelerate and torchrun
|
971 |
+
local_rank:
|
972 |
+
|
973 |
+
# Add or change special tokens.
|
974 |
+
# If you add tokens here, you don't need to add them to the `tokens` list.
|
975 |
+
special_tokens:
|
976 |
+
# bos_token: "<s>"
|
977 |
+
# eos_token: "</s>"
|
978 |
+
# unk_token: "<unk>"
|
979 |
+
|
980 |
+
# Add extra tokens.
|
981 |
+
tokens:
|
982 |
+
|
983 |
+
# FSDP
|
984 |
+
fsdp:
|
985 |
+
fsdp_config:
|
986 |
+
|
987 |
+
# Deepspeed config path. e.g., deepspeed_configs/zero3.json
|
988 |
+
deepspeed:
|
989 |
+
|
990 |
+
# Advanced DDP Arguments
|
991 |
+
ddp_timeout:
|
992 |
+
ddp_bucket_cap_mb:
|
993 |
+
ddp_broadcast_buffers:
|
994 |
+
|
995 |
+
# Path to torch distx for optim 'adamw_anyprecision'
|
996 |
+
torchdistx_path:
|
997 |
+
|
998 |
+
# Set to HF dataset for type: 'completion' for streaming instead of pre-tokenize
|
999 |
+
pretraining_dataset:
|
1000 |
+
|
1001 |
+
# Debug mode
|
1002 |
+
debug:
|
1003 |
+
|
1004 |
+
# Seed
|
1005 |
+
seed:
|
1006 |
+
|
1007 |
+
# Allow overwrite yml config using from cli
|
1008 |
+
strict:
|
1009 |
+
```
|
1010 |
+
|
1011 |
+
</details>
|
1012 |
+
|
1013 |
+
<details>
|
1014 |
+
<summary> Understanding of batch size and gradient accumulation steps </summary>
|
1015 |
+
<br/>
|
1016 |
+
Gradient accumulation means accumulating gradients over several mini-batches and updating the model weights afterward. When the samples in each batch are diverse, this technique doesn't significantly impact learning.
|
1017 |
+
|
1018 |
+
This method allows for effective training with larger effective batch sizes without needing proportionally larger memory. Here's why:
|
1019 |
+
|
1020 |
+
1. **Memory Consumption with Batch Size**: The primary reason increasing the batch size impacts memory is due to the storage requirements for intermediate activations. When you forward propagate a batch through a network, you have to store the activations at each layer for each sample in the batch, because these activations are used during backpropagation to compute gradients. Therefore, larger batches mean more activations, leading to greater GPU memory consumption.
|
1021 |
+
|
1022 |
+
2. **Gradient Accumulation**: With gradient accumulation, you're effectively simulating a larger batch size by accumulating gradients over several smaller batches (or micro-batches). However, at any given time, you're only forward and backward propagating a micro-batch. This means you only store activations for the micro-batch, not the full accumulated batch. As a result, you can simulate the effect of a larger batch size without the memory cost of storing activations for a large batch.
|
1023 |
+
|
1024 |
+
**Example 1:**
|
1025 |
+
Micro batch size: 3
|
1026 |
+
Gradient accumulation steps: 2
|
1027 |
+
Number of GPUs: 3
|
1028 |
+
Total batch size = 3 * 2 * 3 = 18
|
1029 |
+
|
1030 |
+
```
|
1031 |
+
| GPU 1 | GPU 2 | GPU 3 |
|
1032 |
+
|----------------|----------------|----------------|
|
1033 |
+
| S1, S2, S3 | S4, S5, S6 | S7, S8, S9 |
|
1034 |
+
| e1, e2, e3 | e4, e5, e6 | e7, e8, e9 |
|
1035 |
+
|----------------|----------------|----------------|
|
1036 |
+
| → (accumulate) | → (accumulate) | → (accumulate) |
|
1037 |
+
|----------------|----------------|----------------|
|
1038 |
+
| S10, S11, S12 | S13, S14, S15 | S16, S17, S18 |
|
1039 |
+
| e10, e11, e12 | e13, e14, e15 | e16, e17, e18 |
|
1040 |
+
|----------------|----------------|----------------|
|
1041 |
+
| → (apply) | → (apply) | → (apply) |
|
1042 |
+
|
1043 |
+
Accumulated gradient for the weight w1 after the second iteration (considering all GPUs):
|
1044 |
+
Total gradient for w1 = e1 + e2 + e3 + e4 + e5 + e6 + e7 + e8 + e9 + e10 + e11 + e12 + e13 + e14 + e15 + e16 + e17 + e18
|
1045 |
+
|
1046 |
+
Weight update for w1:
|
1047 |
+
w1_new = w1_old - learning rate x (Total gradient for w1 / 18)
|
1048 |
+
```
|
1049 |
+
|
1050 |
+
**Example 2:**
|
1051 |
+
Micro batch size: 2
|
1052 |
+
Gradient accumulation steps: 1
|
1053 |
+
Number of GPUs: 3
|
1054 |
+
Total batch size = 2 * 1 * 3 = 6
|
1055 |
+
|
1056 |
+
```
|
1057 |
+
| GPU 1 | GPU 2 | GPU 3 |
|
1058 |
+
|-----------|-----------|-----------|
|
1059 |
+
| S1, S2 | S3, S4 | S5, S6 |
|
1060 |
+
| e1, e2 | e3, e4 | e5, e6 |
|
1061 |
+
|-----------|-----------|-----------|
|
1062 |
+
| → (apply) | → (apply) | → (apply) |
|
1063 |
+
|
1064 |
+
Accumulated gradient for the weight w1 (considering all GPUs):
|
1065 |
+
Total gradient for w1 = e1 + e2 + e3 + e4 + e5 + e6
|
1066 |
+
|
1067 |
+
Weight update for w1:
|
1068 |
+
w1_new = w1_old - learning rate × (Total gradient for w1 / 6)
|
1069 |
+
```
|
1070 |
+
|
1071 |
+
</details>
|
1072 |
+
|
1073 |
+
### Train
|
1074 |
+
|
1075 |
+
Run
|
1076 |
+
```bash
|
1077 |
+
accelerate launch -m axolotl.cli.train your_config.yml
|
1078 |
+
```
|
1079 |
+
|
1080 |
+
> [!TIP]
|
1081 |
+
> You can also reference a config file that is hosted on a public URL, for example `accelerate launch -m axolotl.cli.train https://yourdomain.com/your_config.yml`
|
1082 |
+
|
1083 |
+
#### Preprocess dataset
|
1084 |
+
|
1085 |
+
You can optionally pre-tokenize dataset with the following before finetuning.
|
1086 |
+
This is recommended for large datasets.
|
1087 |
+
|
1088 |
+
- Set `dataset_prepared_path:` to a local folder for saving and loading pre-tokenized dataset.
|
1089 |
+
- (Optional): Set `push_dataset_to_hub: hf_user/repo` to push it to Huggingface.
|
1090 |
+
- (Optional): Use `--debug` to see preprocessed examples.
|
1091 |
+
|
1092 |
+
```bash
|
1093 |
+
python -m axolotl.cli.preprocess your_config.yml
|
1094 |
+
```
|
1095 |
+
|
1096 |
+
#### Multi-GPU
|
1097 |
+
|
1098 |
+
Below are the options available in axolotl for training with multiple GPUs. Note that DeepSpeed
|
1099 |
+
is the recommended multi-GPU option currently because FSDP may experience
|
1100 |
+
[loss instability](https://github.com/huggingface/transformers/issues/26498).
|
1101 |
+
|
1102 |
+
##### DeepSpeed
|
1103 |
+
|
1104 |
+
Deepspeed is an optimization suite for multi-gpu systems allowing you to train much larger models than you
|
1105 |
+
might typically be able to fit into your GPU's VRAM. More information about the various optimization types
|
1106 |
+
for deepspeed is available at https://huggingface.co/docs/accelerate/main/en/usage_guides/deepspeed#what-is-integrated
|
1107 |
+
|
1108 |
+
We provide several default deepspeed JSON configurations for ZeRO stage 1, 2, and 3.
|
1109 |
+
|
1110 |
+
```yaml
|
1111 |
+
deepspeed: deepspeed_configs/zero1.json
|
1112 |
+
```
|
1113 |
+
|
1114 |
+
```shell
|
1115 |
+
accelerate launch -m axolotl.cli.train examples/llama-2/config.py --deepspeed deepspeed_configs/zero1.json
|
1116 |
+
```
|
1117 |
+
|
1118 |
+
##### FSDP
|
1119 |
+
|
1120 |
+
- llama FSDP
|
1121 |
+
```yaml
|
1122 |
+
fsdp:
|
1123 |
+
- full_shard
|
1124 |
+
- auto_wrap
|
1125 |
+
fsdp_config:
|
1126 |
+
fsdp_offload_params: true
|
1127 |
+
fsdp_state_dict_type: FULL_STATE_DICT
|
1128 |
+
fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
|
1129 |
+
```
|
1130 |
+
|
1131 |
+
##### FSDP + QLoRA
|
1132 |
+
|
1133 |
+
Axolotl supports training with FSDP and QLoRA, see [these docs](docs/fsdp_qlora.qmd) for more information.
|
1134 |
+
|
1135 |
+
##### Weights & Biases Logging
|
1136 |
+
|
1137 |
+
Make sure your `WANDB_API_KEY` environment variable is set (recommended) or you login to wandb with `wandb login`.
|
1138 |
+
|
1139 |
+
- wandb options
|
1140 |
+
```yaml
|
1141 |
+
wandb_mode:
|
1142 |
+
wandb_project:
|
1143 |
+
wandb_entity:
|
1144 |
+
wandb_watch:
|
1145 |
+
wandb_name:
|
1146 |
+
wandb_log_model:
|
1147 |
+
```
|
1148 |
+
|
1149 |
+
##### Special Tokens
|
1150 |
+
|
1151 |
+
It is important to have special tokens like delimiters, end-of-sequence, beginning-of-sequence in your tokenizer's vocabulary. This will help you avoid tokenization issues and help your model train better. You can do this in axolotl like this:
|
1152 |
+
|
1153 |
+
```yml
|
1154 |
+
special_tokens:
|
1155 |
+
bos_token: "<s>"
|
1156 |
+
eos_token: "</s>"
|
1157 |
+
unk_token: "<unk>"
|
1158 |
+
tokens: # these are delimiters
|
1159 |
+
- "<|im_start|>"
|
1160 |
+
- "<|im_end|>"
|
1161 |
+
```
|
1162 |
+
|
1163 |
+
When you include these tokens in your axolotl config, axolotl adds these tokens to the tokenizer's vocabulary.
|
1164 |
+
|
1165 |
+
### Inference Playground
|
1166 |
+
|
1167 |
+
Axolotl allows you to load your model in an interactive terminal playground for quick experimentation.
|
1168 |
+
The config file is the same config file used for training.
|
1169 |
+
|
1170 |
+
Pass the appropriate flag to the inference command, depending upon what kind of model was trained:
|
1171 |
+
|
1172 |
+
- Pretrained LORA:
|
1173 |
+
```bash
|
1174 |
+
python -m axolotl.cli.inference examples/your_config.yml --lora_model_dir="./lora-output-dir"
|
1175 |
+
```
|
1176 |
+
- Full weights finetune:
|
1177 |
+
```bash
|
1178 |
+
python -m axolotl.cli.inference examples/your_config.yml --base_model="./completed-model"
|
1179 |
+
```
|
1180 |
+
- Full weights finetune w/ a prompt from a text file:
|
1181 |
+
```bash
|
1182 |
+
cat /tmp/prompt.txt | python -m axolotl.cli.inference examples/your_config.yml \
|
1183 |
+
--base_model="./completed-model" --prompter=None --load_in_8bit=True
|
1184 |
+
```
|
1185 |
+
-- With gradio hosting
|
1186 |
+
```bash
|
1187 |
+
python -m axolotl.cli.inference examples/your_config.yml --gradio
|
1188 |
+
```
|
1189 |
+
|
1190 |
+
Please use `--sample_packing False` if you have it on and receive the error similar to below:
|
1191 |
+
|
1192 |
+
> RuntimeError: stack expects each tensor to be equal size, but got [1, 32, 1, 128] at entry 0 and [1, 32, 8, 128] at entry 1
|
1193 |
+
|
1194 |
+
### Merge LORA to base
|
1195 |
+
|
1196 |
+
The following command will merge your LORA adapater with your base model. You can optionally pass the argument `--lora_model_dir` to specify the directory where your LORA adapter was saved, otherwhise, this will be inferred from `output_dir` in your axolotl config file. The merged model is saved in the sub-directory `{lora_model_dir}/merged`.
|
1197 |
+
|
1198 |
+
```bash
|
1199 |
+
python3 -m axolotl.cli.merge_lora your_config.yml --lora_model_dir="./completed-model"
|
1200 |
+
```
|
1201 |
+
|
1202 |
+
You may need to use the `gpu_memory_limit` and/or `lora_on_cpu` config options to avoid running out of memory. If you still run out of CUDA memory, you can try to merge in system RAM with
|
1203 |
+
|
1204 |
+
```bash
|
1205 |
+
CUDA_VISIBLE_DEVICES="" python3 -m axolotl.cli.merge_lora ...
|
1206 |
+
```
|
1207 |
+
|
1208 |
+
although this will be very slow, and using the config options above are recommended instead.
|
1209 |
+
|
1210 |
+
## Common Errors 🧰
|
1211 |
+
|
1212 |
+
See also the [FAQ's](./docs/faq.qmd) and [debugging guide](docs/debugging.qmd).
|
1213 |
+
|
1214 |
+
> If you encounter a 'Cuda out of memory' error, it means your GPU ran out of memory during the training process. Here's how to resolve it:
|
1215 |
+
|
1216 |
+
Please reduce any below
|
1217 |
+
- `micro_batch_size`
|
1218 |
+
- `eval_batch_size`
|
1219 |
+
- `gradient_accumulation_steps`
|
1220 |
+
- `sequence_len`
|
1221 |
+
|
1222 |
+
If it does not help, try running without deepspeed and without accelerate (replace "accelerate launch" with "python") in the command.
|
1223 |
+
|
1224 |
+
Using adamw_bnb_8bit might also save you some memory.
|
1225 |
+
|
1226 |
+
> `failed (exitcode: -9)`
|
1227 |
+
|
1228 |
+
Usually means your system has run out of system memory.
|
1229 |
+
Similarly, you should consider reducing the same settings as when you run out of VRAM.
|
1230 |
+
Additionally, look into upgrading your system RAM which should be simpler than GPU upgrades.
|
1231 |
+
|
1232 |
+
> RuntimeError: expected scalar type Float but found Half
|
1233 |
+
|
1234 |
+
Try set `fp16: true`
|
1235 |
+
|
1236 |
+
> NotImplementedError: No operator found for `memory_efficient_attention_forward` ...
|
1237 |
+
|
1238 |
+
Try to turn off xformers.
|
1239 |
+
|
1240 |
+
> accelerate config missing
|
1241 |
+
|
1242 |
+
It's safe to ignore it.
|
1243 |
+
|
1244 |
+
> NCCL Timeouts during training
|
1245 |
+
|
1246 |
+
See the [NCCL](docs/nccl.qmd) guide.
|
1247 |
+
|
1248 |
+
|
1249 |
+
### Tokenization Mismatch b/w Inference & Training
|
1250 |
+
|
1251 |
+
For many formats, Axolotl constructs prompts by concatenating token ids _after_ tokenizing strings. The reason for concatenating token ids rather than operating on strings is to maintain precise accounting for attention masks.
|
1252 |
+
|
1253 |
+
If you decode a prompt constructed by axolotl, you might see spaces between tokens (or lack thereof) that you do not expect, especially around delimiters and special tokens. When you are starting out with a new format, you should always do the following:
|
1254 |
+
|
1255 |
+
1. Materialize some data using `python -m axolotl.cli.preprocess your_config.yml --debug`, and then decode the first few rows with your model's tokenizer.
|
1256 |
+
2. During inference, right before you pass a tensor of token ids to your model, decode these tokens back into a string.
|
1257 |
+
3. Make sure the inference string from #2 looks **exactly** like the data you fine tuned on from #1, including spaces and new lines. If they aren't the same, adjust your inference server accordingly.
|
1258 |
+
4. As an additional troubleshooting step, you can look at the token ids between 1 and 2 to make sure they are identical.
|
1259 |
+
|
1260 |
+
Having misalignment between your prompts during training and inference can cause models to perform very poorly, so it is worth checking this. See [this blog post](https://hamel.dev/notes/llm/05_tokenizer_gotchas.html) for a concrete example.
|
1261 |
+
|
1262 |
+
## Debugging Axolotl
|
1263 |
+
|
1264 |
+
See [this debugging guide](docs/debugging.qmd) for tips on debugging Axolotl, along with an example configuration for debugging with VSCode.
|
1265 |
+
|
1266 |
+
## Need help? 🙋
|
1267 |
+
|
1268 |
+
Join our [Discord server](https://discord.gg/HhrNrHJPRb) where we our community members can help you.
|
1269 |
+
|
1270 |
+
Need dedicated support? Please contact us at [✉️wing@openaccessaicollective.org](mailto:wing@openaccessaicollective.org) for dedicated support options.
|
1271 |
+
|
1272 |
+
## Badge ❤🏷️
|
1273 |
+
|
1274 |
+
Building something cool with Axolotl? Consider adding a badge to your model card.
|
1275 |
+
|
1276 |
+
```markdown
|
1277 |
+
[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
|
1278 |
+
```
|
1279 |
+
|
1280 |
+
[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
|
1281 |
+
|
1282 |
+
## Community Showcase
|
1283 |
+
|
1284 |
+
Check out some of the projects and models that have been built using Axolotl! Have a model you'd like to add to our Community Showcase? Open a PR with your model.
|
1285 |
+
|
1286 |
+
Open Access AI Collective
|
1287 |
+
- [Minotaur 13b](https://huggingface.co/openaccess-ai-collective/minotaur-13b-fixed)
|
1288 |
+
- [Manticore 13b](https://huggingface.co/openaccess-ai-collective/manticore-13b)
|
1289 |
+
- [Hippogriff 30b](https://huggingface.co/openaccess-ai-collective/hippogriff-30b-chat)
|
1290 |
+
|
1291 |
+
PocketDoc Labs
|
1292 |
+
- [Dan's PersonalityEngine 13b LoRA](https://huggingface.co/PocketDoc/Dans-PersonalityEngine-13b-LoRA)
|
1293 |
+
|
1294 |
+
## Contributing 🤝
|
1295 |
+
|
1296 |
+
Please read the [contributing guide](./.github/CONTRIBUTING.md)
|
1297 |
+
|
1298 |
+
Bugs? Please check the [open issues](https://github.com/OpenAccess-AI-Collective/axolotl/issues/bug) else create a new Issue.
|
1299 |
+
|
1300 |
+
PRs are **greatly welcome**!
|
1301 |
+
|
1302 |
+
Please run below to setup env
|
1303 |
+
```bash
|
1304 |
+
git clone https://github.com/OpenAccess-AI-Collective/axolotl
|
1305 |
+
cd axolotl
|
1306 |
+
|
1307 |
+
pip3 install packaging
|
1308 |
+
pip3 install -e '.[flash-attn,deepspeed]'
|
1309 |
+
|
1310 |
+
pip3 install -r requirements-dev.txt -r requirements-tests.txt
|
1311 |
+
pre-commit install
|
1312 |
+
|
1313 |
+
# test
|
1314 |
+
pytest tests/
|
1315 |
+
|
1316 |
+
# optional: run against all files
|
1317 |
+
pre-commit run --all-files
|
1318 |
+
```
|
1319 |
+
|
1320 |
+
Thanks to all of our contributors to date. Help drive open source AI progress forward by contributing to Axolotl.
|
1321 |
+
|
1322 |
+
<a href="https://github.com/openaccess-ai-collective/axolotl/graphs/contributors">
|
1323 |
+
<img src="https://contrib.rocks/image?repo=openaccess-ai-collective/axolotl" alt="contributor chart by https://contrib.rocks"/>
|
1324 |
+
</a>
|
1325 |
+
|
1326 |
+
## Sponsors 🤝❤
|
1327 |
+
|
1328 |
+
OpenAccess AI Collective is run by volunteer contributors such as [winglian](https://github.com/winglian),
|
1329 |
+
[NanoCode012](https://github.com/NanoCode012), [tmm1](https://github.com/tmm1),
|
1330 |
+
[mhenrichsen](https://github.com/mhenrichsen), [casper-hansen](https://github.com/casper-hansen),
|
1331 |
+
[hamelsmu](https://github.com/hamelsmu) and many more who help us accelerate forward by fixing bugs, answering
|
1332 |
+
community questions and implementing new features. Axolotl needs donations from sponsors for the compute needed to
|
1333 |
+
run our unit & integration tests, troubleshooting community issues, and providing bounties. If you love axolotl,
|
1334 |
+
consider sponsoring the project via [GitHub Sponsors](https://github.com/sponsors/OpenAccess-AI-Collective),
|
1335 |
+
[Ko-fi](https://ko-fi.com/axolotl_ai) or reach out directly to
|
1336 |
+
[wing@openaccessaicollective.org](mailto:wing@openaccessaicollective.org).
|
1337 |
+
|
1338 |
+
---
|
1339 |
+
|
1340 |
+
#### 💎 Diamond Sponsors - [Contact directly](mailto:wing@openaccessaicollective.org)
|
1341 |
+
|
1342 |
+
---
|
1343 |
+
|
1344 |
+
#### 🥇 Gold Sponsors - $5000/mo
|
1345 |
+
|
1346 |
+
---
|
1347 |
+
|
1348 |
+
#### 🥈 Silver Sponsors - $1000/mo
|
1349 |
+
|
1350 |
+
---
|
1351 |
+
|
1352 |
+
#### 🥉 Bronze Sponsors - $500/mo
|
1353 |
+
|
1354 |
+
- [JarvisLabs.ai](https://jarvislabs.ai)
|
1355 |
+
|
1356 |
+
---
|
TODO.md
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# todo list
|
2 |
+
|
3 |
+
- [] Validation of parameters for combinations that won't work
|
4 |
+
|
5 |
+
|
6 |
+
|
7 |
+
## things that are known not to work
|
8 |
+
|
9 |
+
- FSDP offload and gradient_checkpointing - https://github.com/pytorch/pytorch/issues/82203
|
10 |
+
- adamw_bnb_8bit doesn't play well with FSDP offload
|
_quarto.yml
ADDED
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
project:
|
2 |
+
type: website
|
3 |
+
|
4 |
+
website:
|
5 |
+
title: "Axolotl"
|
6 |
+
description: "Fine-tuning"
|
7 |
+
favicon: favicon.jpg
|
8 |
+
navbar:
|
9 |
+
title: Axolotl
|
10 |
+
background: dark
|
11 |
+
pinned: false
|
12 |
+
collapse: false
|
13 |
+
tools:
|
14 |
+
- icon: twitter
|
15 |
+
href: https://twitter.com/axolotl_ai
|
16 |
+
- icon: github
|
17 |
+
href: https://github.com/OpenAccess-AI-Collective/axolotl/
|
18 |
+
- icon: discord
|
19 |
+
href: https://discord.gg/7m9sfhzaf3
|
20 |
+
|
21 |
+
sidebar:
|
22 |
+
pinned: true
|
23 |
+
collapse-level: 2
|
24 |
+
style: docked
|
25 |
+
contents:
|
26 |
+
- text: Home
|
27 |
+
href: index.qmd
|
28 |
+
- section: "How-To Guides"
|
29 |
+
contents:
|
30 |
+
# TODO Edit folder structure after we have more docs.
|
31 |
+
- docs/debugging.qmd
|
32 |
+
- docs/multipack.qmd
|
33 |
+
- docs/fdsp_qlora.qmd
|
34 |
+
- docs/input_output.qmd
|
35 |
+
- docs/rlhf.qmd
|
36 |
+
- docs/nccl.qmd
|
37 |
+
- docs/mac.qmd
|
38 |
+
- docs/multi-node.qmd
|
39 |
+
- section: "Reference"
|
40 |
+
contents:
|
41 |
+
- docs/config.qmd
|
42 |
+
- docs/faq.qmd
|
43 |
+
|
44 |
+
|
45 |
+
|
46 |
+
|
47 |
+
format:
|
48 |
+
html:
|
49 |
+
theme: materia
|
50 |
+
css: styles.css
|
51 |
+
toc: true
|
axol/.gitignore
ADDED
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
1 |
+
# created by virtualenv automatically
|
2 |
+
*
|
axol/bin/activate
ADDED
@@ -0,0 +1,87 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# This file must be used with "source bin/activate" *from bash*
|
2 |
+
# you cannot run it directly
|
3 |
+
|
4 |
+
|
5 |
+
if [ "${BASH_SOURCE-}" = "$0" ]; then
|
6 |
+
echo "You must source this script: \$ source $0" >&2
|
7 |
+
exit 33
|
8 |
+
fi
|
9 |
+
|
10 |
+
deactivate () {
|
11 |
+
unset -f pydoc >/dev/null 2>&1 || true
|
12 |
+
|
13 |
+
# reset old environment variables
|
14 |
+
# ! [ -z ${VAR+_} ] returns true if VAR is declared at all
|
15 |
+
if ! [ -z "${_OLD_VIRTUAL_PATH:+_}" ] ; then
|
16 |
+
PATH="$_OLD_VIRTUAL_PATH"
|
17 |
+
export PATH
|
18 |
+
unset _OLD_VIRTUAL_PATH
|
19 |
+
fi
|
20 |
+
if ! [ -z "${_OLD_VIRTUAL_PYTHONHOME+_}" ] ; then
|
21 |
+
PYTHONHOME="$_OLD_VIRTUAL_PYTHONHOME"
|
22 |
+
export PYTHONHOME
|
23 |
+
unset _OLD_VIRTUAL_PYTHONHOME
|
24 |
+
fi
|
25 |
+
|
26 |
+
# The hash command must be called to get it to forget past
|
27 |
+
# commands. Without forgetting past commands the $PATH changes
|
28 |
+
# we made may not be respected
|
29 |
+
hash -r 2>/dev/null
|
30 |
+
|
31 |
+
if ! [ -z "${_OLD_VIRTUAL_PS1+_}" ] ; then
|
32 |
+
PS1="$_OLD_VIRTUAL_PS1"
|
33 |
+
export PS1
|
34 |
+
unset _OLD_VIRTUAL_PS1
|
35 |
+
fi
|
36 |
+
|
37 |
+
unset VIRTUAL_ENV
|
38 |
+
unset VIRTUAL_ENV_PROMPT
|
39 |
+
if [ ! "${1-}" = "nondestructive" ] ; then
|
40 |
+
# Self destruct!
|
41 |
+
unset -f deactivate
|
42 |
+
fi
|
43 |
+
}
|
44 |
+
|
45 |
+
# unset irrelevant variables
|
46 |
+
deactivate nondestructive
|
47 |
+
|
48 |
+
VIRTUAL_ENV='/workspace/axolotl/axol'
|
49 |
+
if ([ "$OSTYPE" = "cygwin" ] || [ "$OSTYPE" = "msys" ]) && $(command -v cygpath &> /dev/null) ; then
|
50 |
+
VIRTUAL_ENV=$(cygpath -u "$VIRTUAL_ENV")
|
51 |
+
fi
|
52 |
+
export VIRTUAL_ENV
|
53 |
+
|
54 |
+
_OLD_VIRTUAL_PATH="$PATH"
|
55 |
+
PATH="$VIRTUAL_ENV/bin:$PATH"
|
56 |
+
export PATH
|
57 |
+
|
58 |
+
if [ "x" != x ] ; then
|
59 |
+
VIRTUAL_ENV_PROMPT=""
|
60 |
+
else
|
61 |
+
VIRTUAL_ENV_PROMPT=$(basename "$VIRTUAL_ENV")
|
62 |
+
fi
|
63 |
+
export VIRTUAL_ENV_PROMPT
|
64 |
+
|
65 |
+
# unset PYTHONHOME if set
|
66 |
+
if ! [ -z "${PYTHONHOME+_}" ] ; then
|
67 |
+
_OLD_VIRTUAL_PYTHONHOME="$PYTHONHOME"
|
68 |
+
unset PYTHONHOME
|
69 |
+
fi
|
70 |
+
|
71 |
+
if [ -z "${VIRTUAL_ENV_DISABLE_PROMPT-}" ] ; then
|
72 |
+
_OLD_VIRTUAL_PS1="${PS1-}"
|
73 |
+
PS1="(${VIRTUAL_ENV_PROMPT}) ${PS1-}"
|
74 |
+
export PS1
|
75 |
+
fi
|
76 |
+
|
77 |
+
# Make sure to unalias pydoc if it's already there
|
78 |
+
alias pydoc 2>/dev/null >/dev/null && unalias pydoc || true
|
79 |
+
|
80 |
+
pydoc () {
|
81 |
+
python -m pydoc "$@"
|
82 |
+
}
|
83 |
+
|
84 |
+
# The hash command must be called to get it to forget past
|
85 |
+
# commands. Without forgetting past commands the $PATH changes
|
86 |
+
# we made may not be respected
|
87 |
+
hash -r 2>/dev/null
|
axol/bin/activate.csh
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# This file must be used with "source bin/activate.csh" *from csh*.
|
2 |
+
# You cannot run it directly.
|
3 |
+
# Created by Davide Di Blasi <davidedb@gmail.com>.
|
4 |
+
|
5 |
+
set newline='\
|
6 |
+
'
|
7 |
+
|
8 |
+
alias deactivate 'test $?_OLD_VIRTUAL_PATH != 0 && setenv PATH "$_OLD_VIRTUAL_PATH:q" && unset _OLD_VIRTUAL_PATH; rehash; test $?_OLD_VIRTUAL_PROMPT != 0 && set prompt="$_OLD_VIRTUAL_PROMPT:q" && unset _OLD_VIRTUAL_PROMPT; unsetenv VIRTUAL_ENV; unsetenv VIRTUAL_ENV_PROMPT; test "\!:*" != "nondestructive" && unalias deactivate && unalias pydoc'
|
9 |
+
|
10 |
+
# Unset irrelevant variables.
|
11 |
+
deactivate nondestructive
|
12 |
+
|
13 |
+
setenv VIRTUAL_ENV '/workspace/axolotl/axol'
|
14 |
+
|
15 |
+
set _OLD_VIRTUAL_PATH="$PATH:q"
|
16 |
+
setenv PATH "$VIRTUAL_ENV:q/bin:$PATH:q"
|
17 |
+
|
18 |
+
|
19 |
+
|
20 |
+
if ('' != "") then
|
21 |
+
setenv VIRTUAL_ENV_PROMPT ''
|
22 |
+
else
|
23 |
+
setenv VIRTUAL_ENV_PROMPT "$VIRTUAL_ENV:t:q"
|
24 |
+
endif
|
25 |
+
|
26 |
+
if ( $?VIRTUAL_ENV_DISABLE_PROMPT ) then
|
27 |
+
if ( $VIRTUAL_ENV_DISABLE_PROMPT == "" ) then
|
28 |
+
set do_prompt = "1"
|
29 |
+
else
|
30 |
+
set do_prompt = "0"
|
31 |
+
endif
|
32 |
+
else
|
33 |
+
set do_prompt = "1"
|
34 |
+
endif
|
35 |
+
|
36 |
+
if ( $do_prompt == "1" ) then
|
37 |
+
# Could be in a non-interactive environment,
|
38 |
+
# in which case, $prompt is undefined and we wouldn't
|
39 |
+
# care about the prompt anyway.
|
40 |
+
if ( $?prompt ) then
|
41 |
+
set _OLD_VIRTUAL_PROMPT="$prompt:q"
|
42 |
+
if ( "$prompt:q" =~ *"$newline:q"* ) then
|
43 |
+
:
|
44 |
+
else
|
45 |
+
set prompt = '('"$VIRTUAL_ENV_PROMPT:q"') '"$prompt:q"
|
46 |
+
endif
|
47 |
+
endif
|
48 |
+
endif
|
49 |
+
|
50 |
+
unset env_name
|
51 |
+
unset do_prompt
|
52 |
+
|
53 |
+
alias pydoc python -m pydoc
|
54 |
+
|
55 |
+
rehash
|
axol/bin/activate.fish
ADDED
@@ -0,0 +1,103 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# This file must be used using `source bin/activate.fish` *within a running fish ( http://fishshell.com ) session*.
|
2 |
+
# Do not run it directly.
|
3 |
+
|
4 |
+
function _bashify_path -d "Converts a fish path to something bash can recognize"
|
5 |
+
set fishy_path $argv
|
6 |
+
set bashy_path $fishy_path[1]
|
7 |
+
for path_part in $fishy_path[2..-1]
|
8 |
+
set bashy_path "$bashy_path:$path_part"
|
9 |
+
end
|
10 |
+
echo $bashy_path
|
11 |
+
end
|
12 |
+
|
13 |
+
function _fishify_path -d "Converts a bash path to something fish can recognize"
|
14 |
+
echo $argv | tr ':' '\n'
|
15 |
+
end
|
16 |
+
|
17 |
+
function deactivate -d 'Exit virtualenv mode and return to the normal environment.'
|
18 |
+
# reset old environment variables
|
19 |
+
if test -n "$_OLD_VIRTUAL_PATH"
|
20 |
+
# https://github.com/fish-shell/fish-shell/issues/436 altered PATH handling
|
21 |
+
if test (echo $FISH_VERSION | head -c 1) -lt 3
|
22 |
+
set -gx PATH (_fishify_path "$_OLD_VIRTUAL_PATH")
|
23 |
+
else
|
24 |
+
set -gx PATH $_OLD_VIRTUAL_PATH
|
25 |
+
end
|
26 |
+
set -e _OLD_VIRTUAL_PATH
|
27 |
+
end
|
28 |
+
|
29 |
+
if test -n "$_OLD_VIRTUAL_PYTHONHOME"
|
30 |
+
set -gx PYTHONHOME "$_OLD_VIRTUAL_PYTHONHOME"
|
31 |
+
set -e _OLD_VIRTUAL_PYTHONHOME
|
32 |
+
end
|
33 |
+
|
34 |
+
if test -n "$_OLD_FISH_PROMPT_OVERRIDE"
|
35 |
+
and functions -q _old_fish_prompt
|
36 |
+
# Set an empty local `$fish_function_path` to allow the removal of `fish_prompt` using `functions -e`.
|
37 |
+
set -l fish_function_path
|
38 |
+
|
39 |
+
# Erase virtualenv's `fish_prompt` and restore the original.
|
40 |
+
functions -e fish_prompt
|
41 |
+
functions -c _old_fish_prompt fish_prompt
|
42 |
+
functions -e _old_fish_prompt
|
43 |
+
set -e _OLD_FISH_PROMPT_OVERRIDE
|
44 |
+
end
|
45 |
+
|
46 |
+
set -e VIRTUAL_ENV
|
47 |
+
set -e VIRTUAL_ENV_PROMPT
|
48 |
+
|
49 |
+
if test "$argv[1]" != 'nondestructive'
|
50 |
+
# Self-destruct!
|
51 |
+
functions -e pydoc
|
52 |
+
functions -e deactivate
|
53 |
+
functions -e _bashify_path
|
54 |
+
functions -e _fishify_path
|
55 |
+
end
|
56 |
+
end
|
57 |
+
|
58 |
+
# Unset irrelevant variables.
|
59 |
+
deactivate nondestructive
|
60 |
+
|
61 |
+
set -gx VIRTUAL_ENV '/workspace/axolotl/axol'
|
62 |
+
|
63 |
+
# https://github.com/fish-shell/fish-shell/issues/436 altered PATH handling
|
64 |
+
if test (echo $FISH_VERSION | head -c 1) -lt 3
|
65 |
+
set -gx _OLD_VIRTUAL_PATH (_bashify_path $PATH)
|
66 |
+
else
|
67 |
+
set -gx _OLD_VIRTUAL_PATH $PATH
|
68 |
+
end
|
69 |
+
set -gx PATH "$VIRTUAL_ENV"'/bin' $PATH
|
70 |
+
|
71 |
+
# Prompt override provided?
|
72 |
+
# If not, just use the environment name.
|
73 |
+
if test -n ''
|
74 |
+
set -gx VIRTUAL_ENV_PROMPT ''
|
75 |
+
else
|
76 |
+
set -gx VIRTUAL_ENV_PROMPT (basename "$VIRTUAL_ENV")
|
77 |
+
end
|
78 |
+
|
79 |
+
# Unset `$PYTHONHOME` if set.
|
80 |
+
if set -q PYTHONHOME
|
81 |
+
set -gx _OLD_VIRTUAL_PYTHONHOME $PYTHONHOME
|
82 |
+
set -e PYTHONHOME
|
83 |
+
end
|
84 |
+
|
85 |
+
function pydoc
|
86 |
+
python -m pydoc $argv
|
87 |
+
end
|
88 |
+
|
89 |
+
if test -z "$VIRTUAL_ENV_DISABLE_PROMPT"
|
90 |
+
# Copy the current `fish_prompt` function as `_old_fish_prompt`.
|
91 |
+
functions -c fish_prompt _old_fish_prompt
|
92 |
+
|
93 |
+
function fish_prompt
|
94 |
+
# Run the user's prompt first; it might depend on (pipe)status.
|
95 |
+
set -l prompt (_old_fish_prompt)
|
96 |
+
|
97 |
+
printf '(%s) ' $VIRTUAL_ENV_PROMPT
|
98 |
+
|
99 |
+
string join -- \n $prompt # handle multi-line prompts
|
100 |
+
end
|
101 |
+
|
102 |
+
set -gx _OLD_FISH_PROMPT_OVERRIDE "$VIRTUAL_ENV"
|
103 |
+
end
|
axol/bin/activate.nu
ADDED
@@ -0,0 +1,96 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# virtualenv activation module
|
2 |
+
# Activate with `overlay use activate.nu`
|
3 |
+
# Deactivate with `deactivate`, as usual
|
4 |
+
#
|
5 |
+
# To customize the overlay name, you can call `overlay use activate.nu as foo`,
|
6 |
+
# but then simply `deactivate` won't work because it is just an alias to hide
|
7 |
+
# the "activate" overlay. You'd need to call `overlay hide foo` manually.
|
8 |
+
|
9 |
+
export-env {
|
10 |
+
def is-string [x] {
|
11 |
+
($x | describe) == 'string'
|
12 |
+
}
|
13 |
+
|
14 |
+
def has-env [...names] {
|
15 |
+
$names | each {|n|
|
16 |
+
$n in $env
|
17 |
+
} | all {|i| $i == true}
|
18 |
+
}
|
19 |
+
|
20 |
+
# Emulates a `test -z`, but btter as it handles e.g 'false'
|
21 |
+
def is-env-true [name: string] {
|
22 |
+
if (has-env $name) {
|
23 |
+
# Try to parse 'true', '0', '1', and fail if not convertible
|
24 |
+
let parsed = (do -i { $env | get $name | into bool })
|
25 |
+
if ($parsed | describe) == 'bool' {
|
26 |
+
$parsed
|
27 |
+
} else {
|
28 |
+
not ($env | get -i $name | is-empty)
|
29 |
+
}
|
30 |
+
} else {
|
31 |
+
false
|
32 |
+
}
|
33 |
+
}
|
34 |
+
|
35 |
+
let virtual_env = '/workspace/axolotl/axol'
|
36 |
+
let bin = 'bin'
|
37 |
+
|
38 |
+
let is_windows = ($nu.os-info.family) == 'windows'
|
39 |
+
let path_name = (if (has-env 'Path') {
|
40 |
+
'Path'
|
41 |
+
} else {
|
42 |
+
'PATH'
|
43 |
+
}
|
44 |
+
)
|
45 |
+
|
46 |
+
let venv_path = ([$virtual_env $bin] | path join)
|
47 |
+
let new_path = ($env | get $path_name | prepend $venv_path)
|
48 |
+
|
49 |
+
# If there is no default prompt, then use the env name instead
|
50 |
+
let virtual_env_prompt = (if ('' | is-empty) {
|
51 |
+
($virtual_env | path basename)
|
52 |
+
} else {
|
53 |
+
''
|
54 |
+
})
|
55 |
+
|
56 |
+
let new_env = {
|
57 |
+
$path_name : $new_path
|
58 |
+
VIRTUAL_ENV : $virtual_env
|
59 |
+
VIRTUAL_ENV_PROMPT : $virtual_env_prompt
|
60 |
+
}
|
61 |
+
|
62 |
+
let new_env = (if (is-env-true 'VIRTUAL_ENV_DISABLE_PROMPT') {
|
63 |
+
$new_env
|
64 |
+
} else {
|
65 |
+
# Creating the new prompt for the session
|
66 |
+
let virtual_prefix = $'(char lparen)($virtual_env_prompt)(char rparen) '
|
67 |
+
|
68 |
+
# Back up the old prompt builder
|
69 |
+
let old_prompt_command = (if (has-env 'PROMPT_COMMAND') {
|
70 |
+
$env.PROMPT_COMMAND
|
71 |
+
} else {
|
72 |
+
''
|
73 |
+
})
|
74 |
+
|
75 |
+
let new_prompt = (if (has-env 'PROMPT_COMMAND') {
|
76 |
+
if 'closure' in ($old_prompt_command | describe) {
|
77 |
+
{|| $'($virtual_prefix)(do $old_prompt_command)' }
|
78 |
+
} else {
|
79 |
+
{|| $'($virtual_prefix)($old_prompt_command)' }
|
80 |
+
}
|
81 |
+
} else {
|
82 |
+
{|| $'($virtual_prefix)' }
|
83 |
+
})
|
84 |
+
|
85 |
+
$new_env | merge {
|
86 |
+
PROMPT_COMMAND : $new_prompt
|
87 |
+
VIRTUAL_PREFIX : $virtual_prefix
|
88 |
+
}
|
89 |
+
})
|
90 |
+
|
91 |
+
# Environment variables that will be loaded as the virtual env
|
92 |
+
load-env $new_env
|
93 |
+
}
|
94 |
+
|
95 |
+
export alias pydoc = python -m pydoc
|
96 |
+
export alias deactivate = overlay hide activate
|
axol/bin/activate.ps1
ADDED
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
$script:THIS_PATH = $myinvocation.mycommand.path
|
2 |
+
$script:BASE_DIR = Split-Path (Resolve-Path "$THIS_PATH/..") -Parent
|
3 |
+
|
4 |
+
function global:deactivate([switch] $NonDestructive) {
|
5 |
+
if (Test-Path variable:_OLD_VIRTUAL_PATH) {
|
6 |
+
$env:PATH = $variable:_OLD_VIRTUAL_PATH
|
7 |
+
Remove-Variable "_OLD_VIRTUAL_PATH" -Scope global
|
8 |
+
}
|
9 |
+
|
10 |
+
if (Test-Path function:_old_virtual_prompt) {
|
11 |
+
$function:prompt = $function:_old_virtual_prompt
|
12 |
+
Remove-Item function:\_old_virtual_prompt
|
13 |
+
}
|
14 |
+
|
15 |
+
if ($env:VIRTUAL_ENV) {
|
16 |
+
Remove-Item env:VIRTUAL_ENV -ErrorAction SilentlyContinue
|
17 |
+
}
|
18 |
+
|
19 |
+
if ($env:VIRTUAL_ENV_PROMPT) {
|
20 |
+
Remove-Item env:VIRTUAL_ENV_PROMPT -ErrorAction SilentlyContinue
|
21 |
+
}
|
22 |
+
|
23 |
+
if (!$NonDestructive) {
|
24 |
+
# Self destruct!
|
25 |
+
Remove-Item function:deactivate
|
26 |
+
Remove-Item function:pydoc
|
27 |
+
}
|
28 |
+
}
|
29 |
+
|
30 |
+
function global:pydoc {
|
31 |
+
python -m pydoc $args
|
32 |
+
}
|
33 |
+
|
34 |
+
# unset irrelevant variables
|
35 |
+
deactivate -nondestructive
|
36 |
+
|
37 |
+
$VIRTUAL_ENV = $BASE_DIR
|
38 |
+
$env:VIRTUAL_ENV = $VIRTUAL_ENV
|
39 |
+
|
40 |
+
if ("" -ne "") {
|
41 |
+
$env:VIRTUAL_ENV_PROMPT = ""
|
42 |
+
}
|
43 |
+
else {
|
44 |
+
$env:VIRTUAL_ENV_PROMPT = $( Split-Path $env:VIRTUAL_ENV -Leaf )
|
45 |
+
}
|
46 |
+
|
47 |
+
New-Variable -Scope global -Name _OLD_VIRTUAL_PATH -Value $env:PATH
|
48 |
+
|
49 |
+
$env:PATH = "$env:VIRTUAL_ENV/bin:" + $env:PATH
|
50 |
+
if (!$env:VIRTUAL_ENV_DISABLE_PROMPT) {
|
51 |
+
function global:_old_virtual_prompt {
|
52 |
+
""
|
53 |
+
}
|
54 |
+
$function:_old_virtual_prompt = $function:prompt
|
55 |
+
|
56 |
+
function global:prompt {
|
57 |
+
# Add the custom prefix to the existing prompt
|
58 |
+
$previous_prompt_value = & $function:_old_virtual_prompt
|
59 |
+
("(" + $env:VIRTUAL_ENV_PROMPT + ") " + $previous_prompt_value)
|
60 |
+
}
|
61 |
+
}
|
axol/bin/activate_this.py
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
"""
|
2 |
+
Activate virtualenv for current interpreter:
|
3 |
+
|
4 |
+
Use exec(open(this_file).read(), {'__file__': this_file}).
|
5 |
+
|
6 |
+
This can be used when you must use an existing Python interpreter, not the virtualenv bin/python.
|
7 |
+
""" # noqa: D415
|
8 |
+
|
9 |
+
from __future__ import annotations
|
10 |
+
|
11 |
+
import os
|
12 |
+
import site
|
13 |
+
import sys
|
14 |
+
|
15 |
+
try:
|
16 |
+
abs_file = os.path.abspath(__file__)
|
17 |
+
except NameError as exc:
|
18 |
+
msg = "You must use exec(open(this_file).read(), {'__file__': this_file})"
|
19 |
+
raise AssertionError(msg) from exc
|
20 |
+
|
21 |
+
bin_dir = os.path.dirname(abs_file)
|
22 |
+
base = bin_dir[: -len("bin") - 1] # strip away the bin part from the __file__, plus the path separator
|
23 |
+
|
24 |
+
# prepend bin to PATH (this file is inside the bin directory)
|
25 |
+
os.environ["PATH"] = os.pathsep.join([bin_dir, *os.environ.get("PATH", "").split(os.pathsep)])
|
26 |
+
os.environ["VIRTUAL_ENV"] = base # virtual env is right above bin directory
|
27 |
+
os.environ["VIRTUAL_ENV_PROMPT"] = "" or os.path.basename(base) # noqa: SIM222
|
28 |
+
|
29 |
+
# add the virtual environments libraries to the host python import mechanism
|
30 |
+
prev_length = len(sys.path)
|
31 |
+
for lib in "../lib/python3.10/site-packages".split(os.pathsep):
|
32 |
+
path = os.path.realpath(os.path.join(bin_dir, lib))
|
33 |
+
site.addsitedir(path.decode("utf-8") if "" else path)
|
34 |
+
sys.path[:] = sys.path[prev_length:] + sys.path[0:prev_length]
|
35 |
+
|
36 |
+
sys.real_prefix = sys.prefix
|
37 |
+
sys.prefix = base
|
axol/bin/pip
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/workspace/axolotl/axol/bin/python
|
2 |
+
# -*- coding: utf-8 -*-
|
3 |
+
import re
|
4 |
+
import sys
|
5 |
+
from pip._internal.cli.main import main
|
6 |
+
if __name__ == '__main__':
|
7 |
+
sys.argv[0] = re.sub(r'(-script\.pyw|\.exe)?$', '', sys.argv[0])
|
8 |
+
sys.exit(main())
|
axol/bin/pip-3.10
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/workspace/axolotl/axol/bin/python
|
2 |
+
# -*- coding: utf-8 -*-
|
3 |
+
import re
|
4 |
+
import sys
|
5 |
+
from pip._internal.cli.main import main
|
6 |
+
if __name__ == '__main__':
|
7 |
+
sys.argv[0] = re.sub(r'(-script\.pyw|\.exe)?$', '', sys.argv[0])
|
8 |
+
sys.exit(main())
|
axol/bin/pip3
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/workspace/axolotl/axol/bin/python
|
2 |
+
# -*- coding: utf-8 -*-
|
3 |
+
import re
|
4 |
+
import sys
|
5 |
+
from pip._internal.cli.main import main
|
6 |
+
if __name__ == '__main__':
|
7 |
+
sys.argv[0] = re.sub(r'(-script\.pyw|\.exe)?$', '', sys.argv[0])
|
8 |
+
sys.exit(main())
|
axol/bin/pip3.10
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/workspace/axolotl/axol/bin/python
|
2 |
+
# -*- coding: utf-8 -*-
|
3 |
+
import re
|
4 |
+
import sys
|
5 |
+
from pip._internal.cli.main import main
|
6 |
+
if __name__ == '__main__':
|
7 |
+
sys.argv[0] = re.sub(r'(-script\.pyw|\.exe)?$', '', sys.argv[0])
|
8 |
+
sys.exit(main())
|
axol/bin/python
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c37ac6a3327a3ef707c6e7df9888ed0b6713ccf4c1af3a7088aedbf87132d08a
|
3 |
+
size 17215080
|
axol/bin/python3
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c37ac6a3327a3ef707c6e7df9888ed0b6713ccf4c1af3a7088aedbf87132d08a
|
3 |
+
size 17215080
|
axol/bin/python3.10
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c37ac6a3327a3ef707c6e7df9888ed0b6713ccf4c1af3a7088aedbf87132d08a
|
3 |
+
size 17215080
|
axol/bin/wheel
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/workspace/axolotl/axol/bin/python
|
2 |
+
# -*- coding: utf-8 -*-
|
3 |
+
import re
|
4 |
+
import sys
|
5 |
+
from wheel.cli import main
|
6 |
+
if __name__ == '__main__':
|
7 |
+
sys.argv[0] = re.sub(r'(-script\.pyw|\.exe)?$', '', sys.argv[0])
|
8 |
+
sys.exit(main())
|
axol/bin/wheel-3.10
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/workspace/axolotl/axol/bin/python
|
2 |
+
# -*- coding: utf-8 -*-
|
3 |
+
import re
|
4 |
+
import sys
|
5 |
+
from wheel.cli import main
|
6 |
+
if __name__ == '__main__':
|
7 |
+
sys.argv[0] = re.sub(r'(-script\.pyw|\.exe)?$', '', sys.argv[0])
|
8 |
+
sys.exit(main())
|
axol/bin/wheel3
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/workspace/axolotl/axol/bin/python
|
2 |
+
# -*- coding: utf-8 -*-
|
3 |
+
import re
|
4 |
+
import sys
|
5 |
+
from wheel.cli import main
|
6 |
+
if __name__ == '__main__':
|
7 |
+
sys.argv[0] = re.sub(r'(-script\.pyw|\.exe)?$', '', sys.argv[0])
|
8 |
+
sys.exit(main())
|
axol/bin/wheel3.10
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/workspace/axolotl/axol/bin/python
|
2 |
+
# -*- coding: utf-8 -*-
|
3 |
+
import re
|
4 |
+
import sys
|
5 |
+
from wheel.cli import main
|
6 |
+
if __name__ == '__main__':
|
7 |
+
sys.argv[0] = re.sub(r'(-script\.pyw|\.exe)?$', '', sys.argv[0])
|
8 |
+
sys.exit(main())
|
axol/pyvenv.cfg
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
home = /opt/conda/envs/axol/bin
|
2 |
+
implementation = CPython
|
3 |
+
version_info = 3.10.14.final.0
|
4 |
+
virtualenv = 20.25.1
|
5 |
+
include-system-site-packages = false
|
6 |
+
base-prefix = /opt/conda/envs/axol
|
7 |
+
base-exec-prefix = /opt/conda/envs/axol
|
8 |
+
base-executable = /opt/conda/envs/axol/bin/python3.10
|