Closed
Conversation
* rm model_forward assignment in val * rm clutter from diffusion branch * reverse if order
* Fix bug with diagnostic streams * Avoid that empty decoders are allocated
* Doing something wrong * Make fine-tuning work * Rename sensibly
* Enable multiple student views for one target * Improved readability
* add pin mem to IOReaderData * add pin mem to sample & modelbatch class * add pin mem to stream data * add pin mem to training loop * run /scripts/actions.sh lint * run ./scripts/actions.sh unit-test * ignore check torch import in package * move pinning to MultiStreamDataSampler * add _pin_tensor & _pin_tensor_list helper func * ruff the code * move back pin mem. to train loop * Remove the ignore-import-error rule and revert to the state before the change * create protocol for pinnable obj * remove pin_mem from IOReaderData class * add pin_memory to Trainer.validate * remove pin_memory from loader_params * Rever export/export_inference.py to state before c3fc9a7 * change name * revise Pinnable class description * add memory_pinning in config, train & va loop * use getattr to avoid CICD warning * use setattr to avoid CICD warning * disable pylint for self.source_tokens_lens * Fixed issues with memory pinning due to rebasing and also adjusted config position of flag * Reverting unadvert changes --------- Co-authored-by: Javad Kasravi <j.kasravi@fz-juelich.de> Co-authored-by: Javad Kasravi <jkasravi@santis-ln002.cscs.ch> Co-authored-by: Javad kasravi <kasravi66@gmail.com>
* split WeatherGenReader functionality to allow reading only JSON adding weathergen JSON reader to develop * informative error when metrics are not there * restore JSONreader after rebase * JSONreader mostly restored * MLFlow logging independent of JSON/zarr * linting, properly cheking fsteps, ens, samples in JSONreader * tiny change to restore the MergeReader * lint * enabling JSONreader to skip plots and missing scores gracefully * required reformatting * move skipping of metrics to the reader class * slighly more explicit formulations --------- Co-authored-by: Sebastian Buschow <sbuschow@santis-ln001.cscs.ch> Co-authored-by: Sebastian Buschow <sbuschow@santis-ln002.cscs.ch> Co-authored-by: iluise <72020169+iluise@users.noreply.github.com> Co-authored-by: Ilaria Luise <luise.ilaria@gmail.com>
* Add target type value error * Remove type * Remove unused code * Commit what shall have been committed * Remove target readout type from config * Add computing stream names to embedding engine --------- Co-authored-by: Christian Lessig <christian.lessig@ecmwf.int>
* add default streams + fix lead time error * update config * Correct a bug creating aggr issues on scores (#1685) --------- Co-authored-by: Savvas Melidonis <79579567+SavvasMel@users.noreply.github.com>
* add default streams + fix lead time error * update config * update ratio plots and bar plots for single run * fix title * Update config Added support information for forecast_step configuration. --------- Co-authored-by: Savvas Melidonis <79579567+SavvasMel@users.noreply.github.com>
* add argument * check stage argument * removed unnecessary code * arbitrary position arguments * Fix error text * get stage info from environment variable. * Update run_train.py --------- Co-authored-by: Simon Grasse <s.grasse@fz-juelich.de>
* caching get_shared_wg_path()
* renaming get_path_output to get_path_results
* model and results paths from get_shared_wg_path() and removed _get_config_attribute()
* marking get_shared_wg_path() as private
* removing set_path()
* fixed call to _get_shared_wg_path
* fixed import, code clean-up, change caching decorator
* changed way of caching _get_shared_wg_base_path
* fixed typing error
* changes in Refactor shared WG path handling and model config I/O
- Simplify get_path_model/get_path_run to always resolve via _get_shared_wg_path()
- Change _get_shared_wg_path() to cached, argument-free helper returning the shared working dir from private config
- Adjust model config save/load to build filenames relative to the run’s model directory instead of passing parent paths around
- Update load_run_config and load_merge_configs to use new path helpers and improve assertion/log messages
- Replace internal _get_shared_wg_path("results") usages with get_path_run() in wegen_reader and train_logger
* fixed base_path in metrics_path
* fixed forgotten config.general
* fixed lint raised issues
* Improve path handling and add missing docstrings
- Add docstrings to 10+ utility functions for better documentation
- Refactor load_run_config to improve path construction logic
- Move mini_epoch string formatting from _get_model_config_file_read_name
to caller for better separation of concerns
- Add validation for mini_epoch_str format with descriptive error messages
- Fix multi-line docstring format in _load_private_conf
* fixed line too long
* reverting to previous _get_model_config_file_read_name()
* pretty fix for _get_model_config_file_read_name
* pretty fix for _get_model_config_file_read_name
* removed unused/undefined path
* replace '_' with '-' * cli options underscore to dash * change underscores to hyphens * rename options in cli unit test
Co-authored-by: Savvas Melidonis <79579567+SavvasMel@users.noreply.github.com>
* rename write_num_samples to num_samples * Fixing linting --------- Co-authored-by: Christian Lessig <christian.lessig@ecmwf.int>
* remove misleading logging of mini_epoch * add forecast_steps logging
* Fix duplicate run_id in results and runplots paths. Linting. * remove duplicate run_id also from metrics directory * Linting
iluise
changed the title
clessig
reviewed
Feb 3, 2026
|
|
||
| def get_evaluation_config(run_id, verbose=False): | ||
| """Create evaluation configuration for multiple streams.""" | ||
| cfg = omegaconf.OmegaConf.create( |
Collaborator
There was a problem hiding this comment.
We should use yaml config files here and not hard-code configs in the code.
Contributor
Author
There was a problem hiding this comment.
do you want it as part of the main config then? Can we do it in a separate PR?
Contributor
|
Hey @iluise, thanks so much for working on this! Is there a summary or high-level overview of the changes or is the best thing I can do, going through the code line-by-line? |
Contributor
Author
|
I think you should mainly test if it does what we want for the ssl analysis (aka are the plots and the numbers what we want to test the pre-training?) and feedbacks are more than welcome. most of the changes are to remove verbosity, so not very interesting overall. the main novelty is the ssl_analysis.py code. Feedbacks on that are also welcome as people from pre-training might touch it in the future to change their analysis I guess (nb. the config is now hardcoded but we will change that in the future). |
6 tasks
tjhunter
reviewed
Feb 25, 2026
| ] | ||
|
|
||
| [project.scripts] | ||
| ssl_analysis= "weathergen.evaluate.ssl.ssl_eval:ssl_analysis" |
Collaborator
There was a problem hiding this comment.
can we fold that into an evaluation sub-task? I am happy to discuss reasons for that.
Contributor
Author
|
closing as implemented elsewhere by Simone. |
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
11 participants
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.
Description
add code for standard inference + evaluation for jepa/dinov3 etc..
usage:
Issue Number
Closes #1746
Is this PR a draft? Mark it as draft.
Checklist before asking for review
./scripts/actions.sh lint./scripts/actions.sh unit-test./scripts/actions.sh integration-testlaunch-slurm.py --time 60