Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Model choice: maximize Hydra's features #246

Closed
remtav opened this issue Jan 25, 2022 · 0 comments · Fixed by #295
Closed

Model choice: maximize Hydra's features #246

remtav opened this issue Jan 25, 2022 · 0 comments · Fixed by #295
Assignees
@remtav
Labels
enhancement New feature or request

Comments

@remtav
Copy link
Collaborator

remtav commented Jan 25, 2022

Currently, choosing a model in GDL with model_choice.py makes use of a series of disorganized if-else statements.

As mentionned in this Towards Data Science article, choosing models based on a series of if-else conditions is to be avoided. Now that hydra is implemented with PR #208, choosing a model can be much more straight forward by using config files for each relevant architecture and pointing to the model class directly in this config file.

This was already done by @CharlesAuthier with fastrcnn.
This was referenced Feb 1, 2022
Closed
Merged
Merged
remtav added a commit to remtav/geo-deep-learning that referenced this issue Feb 11, 2022
@remtav
config/hydra/default.yaml: implement optuna hyperparameter optimizati…
1bbd7f1 
…on (add returns to train_segmentation.py and run_gdl())

metrics.py: add try/except if missing key
model_choice.py: temporarily raise gpu threshold to 100% until issue NRCan#246 is addressed
train_segmentation.py:
- fix bug for dontcare value in metrics
- always calculate metrics at test time
inference_segmentation.py: warn if inference has only background values
test hyperparameter optimization in CI with github actions
@remtav remtav mentioned this issue Feb 11, 2022
Closed
@remtav remtav added the enhancement New feature or request label Mar 1, 2022
@remtav remtav mentioned this issue Mar 1, 2022
Closed
This was referenced Mar 14, 2022
Closed
Closed
@remtav remtav self-assigned this Mar 14, 2022
This was referenced Mar 21, 2022
Closed
Closed
remtav added a commit to remtav/geo-deep-learning that referenced this issue Mar 22, 2022
@remtav
fixes NRCan#293 NRCan#246
b1f1f4c 
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models
@remtav remtav mentioned this issue Mar 22, 2022
Merged
remtav added a commit that referenced this issue Mar 24, 2022
@remtav
Use hydra for instantiating models and optimizers (#295)
a1e322d 
* fixes #293 #246
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models

* minor typo fixes

* name model yamls as close as possible to upcoming naming convention

* fix model name
remtav added a commit to remtav/geo-deep-learning that referenced this issue Mar 24, 2022
@remtav
fixes NRCan#293 NRCan#246
9000883 
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models
remtav added a commit to remtav/geo-deep-learning that referenced this issue Mar 24, 2022
@remtav
fixes NRCan#293 NRCan#246
6275fb0 
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models
remtav added a commit that referenced this issue Mar 29, 2022
@remtav
Override inference parameters (model architecture, num bands, num cla…
aafaa17 
…sses) from checkpoint (#298)

* fixes #293 #246
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models

* minor typo fixes

* implement overriding model params from checkpoint with minimal error handling for checkpoints from different gdl versions
fixes #183

* name model yamls as close as possible to upcoming naming convention

* fix model name

* implement overriding model params from checkpoint with minimal error handling for checkpoints from different gdl versions
fixes #183

* small bugfix for pointing to parameters inside checkpoint

* model_choice.py: add update checkpoint utility

* fixes #293 #246
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models

* minor typo fixes

* name model yamls as close as possible to upcoming naming convention

* small bugfix for pointing to parameters inside checkpoint

* remove deeplabv3 dualhead warning and add link for deeplabv3_dualhead.py

* fixes #293 #246
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models

* name model yamls as close as possible to upcoming naming convention

* minor typo fixes

* update to PR 295

* GDL.py: restore to previous commit based on cauthier's comment
remtav added a commit to remtav/geo-deep-learning that referenced this issue Jul 5, 2022
@remtav
Use hydra for instantiating models and optimizers (NRCan#295)
096575e 
* fixes NRCan#293 NRCan#246
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models

* minor typo fixes

* name model yamls as close as possible to upcoming naming convention

* fix model name
remtav added a commit to remtav/geo-deep-learning that referenced this issue Jul 5, 2022
@remtav
Override inference parameters (model architecture, num bands, num cla…
5b6cc3c 
…sses) from checkpoint (NRCan#298)

* fixes NRCan#293 NRCan#246
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models

* minor typo fixes

* implement overriding model params from checkpoint with minimal error handling for checkpoints from different gdl versions
fixes NRCan#183

* name model yamls as close as possible to upcoming naming convention

* fix model name

* implement overriding model params from checkpoint with minimal error handling for checkpoints from different gdl versions
fixes NRCan#183

* small bugfix for pointing to parameters inside checkpoint

* model_choice.py: add update checkpoint utility

* fixes NRCan#293 NRCan#246
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models

* minor typo fixes

* name model yamls as close as possible to upcoming naming convention

* small bugfix for pointing to parameters inside checkpoint

* remove deeplabv3 dualhead warning and add link for deeplabv3_dualhead.py

* fixes NRCan#293 NRCan#246
add tests for optimizer instantiation in test_optimizers.py
adapt our unet models (models/unet.py) to expect same parameter names as smp models

* name model yamls as close as possible to upcoming naming convention

* minor typo fixes

* update to PR 295

* GDL.py: restore to previous commit based on cauthier's comment
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@remtav remtav

Labels
enhancement New feature or request
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Use hydra for instantiating models and optimizers remtav/geo-deep-learning
1 participant
@remtav