clean up README and add CONTRIBUTING guide

CONTRIBUTING.md

@@ -0,0 +1,99 @@
+# Contributing
+
+Thanks for taking the time to contribute to this project. 🙏🏼
+
+This document covers how to get started working on this project, and proposing changes to make it better.
+
+## Start with an issue
+
+Rather than jump straight to an implementation in a pull request, please start with an issue.
+
+This gives everyone a chance to weigh in on the idea, and for you to get some early feedback. It also helps avoid wasting time on something that the maintainers don't think is a good fit for this project.
+
+If you've found a bug, please include steps to reproduce it.
+
+## Development
+
+Once you've found an issue to work on, it's time to jump into making changes.
+
+Before submitting a PR, format the code and run the linter locally.
+
+```shell
+pip install -r requirements-test.txt  # only need to do this once
+ruff format
+ruff check
+```
+
+## Open a pull request
+
+We use the [GitHub Flow](https://docs.github.com/en/get-started/quickstart/github-flow) model for managing changes.
+
+1. Fork the repository
+2. Create a new branch
+3. Make your changes
+4. Add or update tests, if applicable
+5. Add or update documentation, if applicable
+6. Open a pull request, and include a reference to the original issue
+
+---
+
+## Deploy to production
+
+> [!NOTE]
+> 👇 The rest of this document is for internal Replicate use. It contains instructions for deploying the model to production and updating hotswap bases. If you're an external contributor, you can safely ignore it.
+
+### Push model
+
+Manually trigger the [Github Push action](https://github.com/replicate/flux-fine-tuner/actions/workflows/push.yaml). Uncheck the "Test only, without pushing to prod" button. You might also have to uncheck the "Compare outputs..." button.
+
+The push action takes half an hour since it tests both training and inference. But it's worth doing to be safe.
+
+Once you've deployed to production, make a test fine-tune and run predictions on the trained model, just to be super sure it works.
+
+### Update hotswap bases
+
+When you've tested the pushed model, you need to update all existing fine-tuned versions to use the predictor you just pushed.
+
+First, only update a single model or a subset of models to test that this step works.
+
+In your local checkout of [web](https://github.com/replicate/web), run
+
+```shell
+script/manage-prod update_hotswap_base_version \
+    --from-model ostris/flux-dev-lora-trainer \
+    --to-latest \
+    --trained-version-filter="<your-username>/<your-model-name>"
+```
+
+This is a dry run to list the versions that will be updated. The `--trained-version-filter` can be just "<your-username>" if you want to test all your models.
+
+When you're happy to with the list of models that will be updated, run
+
+```shell
+script/manage-prod update_hotswap_base_version \
+    --from-model ostris/flux-dev-lora-trainer \
+    --to-latest \
+    --trained-version-filter="<your-username>/<your-model-name>" \
+    --force
+```
+
+Now you can test that the predictor works for the updated model(s).
+
+When you're happy with that, do the same thing but for all models. First, a dry run:
+
+```shell
+script/manage-prod update_hotswap_base_version \
+    --from-model ostris/flux-dev-lora-trainer \
+    --to-latest \
+```
+
+And then actually update all the trained Flux versions:
+
+```shell
+script/manage-prod update_hotswap_base_version \
+    --from-model ostris/flux-dev-lora-trainer \
+    --to-latest \
+    --force
+```
+
+This process is being improved -- soon it will be possible to configure base versions in Django admin.

README.md

@@ -1,69 +1,29 @@
 # flux-fine-tuner
 
-## Development
+This is a [Cog](https://cog.run) training model that creates LoRA-based fine-tunes for the [FLUX.1](https://replicate.com/blog/flux-state-of-the-art-image-generation) family of image generation models.
 
-Before submitting a PR, format the code and run the linter locally.
+It's live at [replicate.com/ostris/flux-dev-lora-trainer/train](https://replicate.com/ostris/flux-dev-lora-trainer).
 
-```shell
-pip install -r requirements-test.txt  # only need to do this once
-ruff format
-ruff check
-```
+## Fine-tuning Flux
 
-## Deploy to production
+If you're just looking to create a fine-tuned Flux model, you don't need to do anything with this codebase.
 
-### Push model
+To get started fine-tuning Flux on Replicate, check out these guides:
 
-Manually trigger the [Github Push action](https://github.com/replicate/flux-fine-tuner/actions/workflows/push.yaml). Uncheck the "Test only, without pushing to prod" button. You might also have to uncheck the "Compare outputs..." button.
+- [Fine-tune FlUX.1](https://replicate.com/blog/fine-tune-flux)
+- [Fine-tune FLUX.1 to create images of yourself](https://replicate.com/blog/fine-tune-flux-with-faces)
+- [Fine-tune FLUX.1 with an API](https://replicate.com/blog/fine-tune-flux-with-an-api)
 
-The push action takes half an hour since it tests both training and inference. But it's worth doing to be safe.
+## Contributing
 
-Once you've deployed to production, make a test fine-tune and run predictions on the trained model, just to be super sure it works.
+Want to help make this project better? Check out the [contributing guide](CONTRIBUTING.md) to get started.
 
-### Update hotswap bases
+## Credits
 
-When you've tested the pushed model, you need to update all existing fine-tuned versions to use the predictor you just pushed.
+This project is based on the [ai-toolkit](https://github.com/ostris/ai-toolkit) project, which was created by [@ostris](https://github.com/ostris). ❤️
 
-First, only update a single model or a subset of models to test that this step works.
+## License
 
-In your local checkout of [web](https://github.com/replicate/web), run
+Flux Dev falls under the [`FLUX.1 [dev]` Non-Commercial License](https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/LICENSE.md).
 
-```shell
-script/manage-prod update_hotswap_base_version \
-    --from-model ostris/flux-dev-lora-trainer \
-    --to-latest \
-    --trained-version-filter="<your-username>/<your-model-name>"
-```
-
-This is a dry run to list the versions that will be updated. The `--trained-version-filter` can be just "<your-username>" if you want to test all your models.
-
-When you're happy to with the list of models that will be updated, run
-
-```shell
-script/manage-prod update_hotswap_base_version \
-    --from-model ostris/flux-dev-lora-trainer \
-    --to-latest \
-    --trained-version-filter="<your-username>/<your-model-name>" \
-    --force
-```
-
-Now you can test that the predictor works for the updated model(s).
-
-When you're happy with that, do the same thing but for all models. First, a dry run:
-
-```shell
-script/manage-prod update_hotswap_base_version \
-    --from-model ostris/flux-dev-lora-trainer \
-    --to-latest \
-```
-
-And then actually update all the trained Flux versions:
-
-```shell
-script/manage-prod update_hotswap_base_version \
-    --from-model ostris/flux-dev-lora-trainer \
-    --to-latest \
-    --force
-```
-
-This process is being improved -- soon it will be possible to configure base versions in Django admin.
+Flux Schnell falls under the [Apache-2.0 License](https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/apache-2.0.md).