-
Notifications
You must be signed in to change notification settings - Fork 175
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
v1.0 merge #910
Merged
v1.0 merge #910
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…buckets with crops
…ushing it there; allow user to configure multidatabackend using multi-res and crop by default
…f different model-specific names
…ner into refactor/main-loop
The docs show `instance_prompt` when the value should be `instanceprompt`. ``` 2024-08-29 15:28:24,677 [ERROR] (__main__) Unsupported caption strategy: instance_prompt. Supported: 'filename', 'textfile', 'parquet', 'instanceprompt', traceback: Traceback (most recent call last): ```
Dataloader Docs - Correct caption strategy for instance prompt
…onstant_with_warmup
fix constant_with_warmup not being so constant or warming up
…ile preparing and loading state
follow-up fix for setting last_epoch
fix multigpu schedule issue with LR on resume
… overcome accelerate issue
multiply the resume state step by the number of GPUs in an attempt to overcome accelerate v0.33 issue
default to json/toml before the env file in case multigpu is configured
fix json/toml configs str bool values
bypass some "helpful" diffusers logic that makes random decisions to run on CPU
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Here's a summary of the major changes in version 1.0. These are huge changes, so it would be nice if others can retrieve the main branch and run some old configurations and ensure things still look as you'd expect.
Refactoring and Enhancements:
Refactor
train.py
into a Trainer Class:train.py
has been restructured into aTrainer
class, improving modularity and maintainability.Model Family Unification:
--sd3
,--flux
, etc.) have been replaced with a unified--model_family
argument, streamlining model specification and reducing clutter in configurations.Configuration System Overhaul:
.env
configuration files to JSON (config.json
), with multiple backends supporting JSON configuration loading. This allows more flexible and readable configuration management.Enhanced Argument Handling:
helpers/configuration/cmd_args.py
for better organization.--model_card_safe_for_work
,--flux_schedule_shift
, and--disable_bucket_pruning
.Improved Hugging Face Integration:
configure.py
to avoid asking for Hugging Face model name details unless required.Optimizations and Bug Fixes:
--optimizer
usage.Training and Validation Enhancements:
Enhanced Logging and Monitoring:
Documentation Updates:
Miscellaneous Improvements:
train.sh
to minimal requirements, reducing complexity and streamlining the training process.