Comments (6)
Ah yes that fixed it, thanks!
from gandlf.
Hi @Karol-G , Thanks for using GaNDLF.
It would be great if you comment out the line :
#parallel_compute_command: 'qsub -b y -l gpu -l h_vmem=32G -cwd -o ${outputDir}/\$JOB_ID.stdout -e ${outputDir}/\$JOB_ID.stderr `pwd`/sge_wrapper _correct_location_of_virtual_environment_/venv/bin/python'
so that it doesn't attempt to do a parallel compute, similarly, I would also recommend you to disable folded validation as:
nested_training:
{
testing: 1, # this controls the testing data splits for final model evaluation; use '1' if this is to be disabled
validation: -5 # this controls the validation data splits for model training
}
for running it once.
from gandlf.
I commented out parallel_compute_command
and testing
and validation
instead of nested_training, else I get the error The parameter 'nested_training' needs to be defined
.
But now i get the following error:
Using default folds for testing split: -5
Using default folds for validation split: -5
Using previously saved parameter file ./experiments/2d_classification/output_dir/parameters.pkl
Traceback (most recent call last):
File "gandlf_run", line 75, in <module>
main()
File "gandlf_run", line 70, in main
TrainingManager(dataframe=data_full, headers = headers, outputDir=model_path, parameters=parameters, device=device, reset_prev = reset_prev)
File "/content/GaNDLF-refactor/GANDLF/training_manager.py", line 146, in TrainingManager
device=device, params=parameters, testing_data=testingData)
File "/content/GaNDLF-refactor/GANDLF/training_loop.py", line 322, in training_loop
metrics = params["metrics"]
KeyError: 'metrics'
Seems a metric is missing. How do I define it for a classification task?
from gandlf.
HI @Karol-G,
You can add a key metrics
in the parameter file with the value ['mse']
for classification. I will update this on the testing config for clarity.
Cheers,
Sarthak
from gandlf.
Hmm, I still get the same error with:
metrics:
- mse
or:
metrics: ['mse']
from gandlf.
Ah, I think you need to try with this:
python gandlf_run -config ./experiments/2d_classification/model.yaml -data ./experiments/2d_classification/train.csv -output ./experiments/2d_classification/output_dir/ -train 1 -device cuda \
-reset_prev True # this will remove all writes to disk (such as training/validation data and parameters) from previous run
from gandlf.
Related Issues (20)
- Add CITATION.cff file to reference journal publication
- code for curation of tiles after patching
- remove
- Using patch miner without `Label` header results in failure
- Issue with cuda lib? HOT 11
- Add a few segmentation-specific metrics
- Allow metrics to be generated by passing predictions and ground truths
- RuntimeError: cuDNN error: CUDNN_STATUS_MAPPING_ERROR HOT 1
- Add metrics to compare 2 images
- Ensure the training parameters are captured in the model dictionary itself
- Updated post-training model optimization to ensure the built-in parameters can be used
- Perform penalty calculation after all sanity checks are completed
- Allow penalty and class weights to be taken from the config
- torch.Size error
- `gandlf_collectStats` is not detecting `problem_type` correctly HOT 1
- Add documentation describing how one could federate their models using OpenFL
- Problem with cuda following installing virtual environment HOT 4
- Per-label and per-class metrics should be formatted properly for CSV output
- No need for the history file
- Add focal loss
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from gandlf.