Skip to content

Latest esrun configs #218

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

Open
wants to merge 3 commits into
base: master
Choose a base branch
from
Open

Latest esrun configs #218

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
8f0db13
Latest esrun configs
joshhvulcan Oct 3, 2025
8743ce5
Update to 0.5 grid
joshhvulcan Oct 3, 2025
e739a43
Add missing env vars
joshhvulcan Oct 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 4 additions & 11 deletions esrun_data/satlas/solar_farm_oe/esrun.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,15 +1,11 @@
inference_results_config:
data_type: raster

partition_strategies:
partition_request_geometry:
class_path: esrun.runner.tools.partitioners.grid_partitioner.GridPartitioner
init_args:
grid_size: 0.15
output_projection:
class_path: rslearn.utils.geometry.Projection
init_args:
crs: EPSG:3857
x_resolution: 10
y_resolution: -10
use_utm: true
grid_size: 0.5

prepare_window_geometries:
class_path: esrun.runner.tools.partitioners.grid_partitioner.GridPartitioner
Expand All @@ -32,6 +28,3 @@ postprocessing_strategies:

process_window:
class_path: esrun.runner.tools.postprocessors.noop_raster.NoopRaster

inference_results_config:
data_type: RASTER
67 changes: 32 additions & 35 deletions esrun_data/satlas/solar_farm_oe/model.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,35 +1,7 @@
trainer:
callbacks:
- class_path: lightning.pytorch.callbacks.LearningRateMonitor
init_args:
logging_interval: epoch
- class_path: lightning.pytorch.callbacks.ModelCheckpoint
init_args:
monitor: val_class/accuracy
save_last: true
save_top_k: 1
mode: max
- class_path: rslearn.train.callbacks.freeze_unfreeze.FreezeUnfreeze
init_args:
module_selector:
- model
- encoder
- 0
unfreeze_at_epoch: 20
unfreeze_lr_factor: 10.0
- class_path: rslearn.train.prediction_writer.RslearnWriter
init_args:
path: placeholder
output_layer: ${PREDICTION_OUTPUT_LAYER}
selector: ["class"]
max_epochs: 500
default_root_dir: ${TRAINER_DATA_PATH}
logger:
class_path: lightning.pytorch.loggers.WandbLogger
init_args:
project: ${WANDB_PROJECT}
name: ${WANDB_NAME}
entity: ${WANDB_ENTITY}
# FIXME ${TRAINER_DATA_PATH}
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What is the reason to move it from configuring things properly to having it in a comment?

# FIXME ${WANDB_PROJECT}: wandb project for the trainer to log metrics to
# FIXME ${WANDB_NAME}: wandb name for the trainer to log metrics to
# FIXME ${WANDB_ENTITY}: wandb entity for the trainer to log metrics to
model:
class_path: rslearn.train.lightning_module.RslearnLightningModule
init_args:
Expand All @@ -39,7 +11,9 @@ model:
encoder:
- class_path: rslp.helios.model.Helios
init_args:
checkpoint_path: ${EXTRA_FILES_PATH}/yawenzzzz/latent_mim_cross_random_per_modality_patchdisc_add_contrastive_0.1_1/step400000
checkpoint_path: gs://earth-system-run-dev/models/02ce432e-98e7-46d7-8a67-a1d9b346e746/stage_0/helios/step300000
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't think we should workaround things just by putting variables like EXTRA_FILES_PATH in a comment. If there is support to pass EXTRA_FILES_PATH to the local runner then we should do that, the path is gs://ai2-rslearn-projects-data/helios/checkpoints/.

# checkpoint_path: ${EXTRA_FILES_PATH}/helios/step300000
random_initialization: true
selector:
- encoder
forward_kwargs:
Expand Down Expand Up @@ -343,5 +317,28 @@ data:
- B12
- B01
- B09
# rslp_project: 2025_06_06_helios_finetuning
# rslp_experiment: v2_satlas_solar_farm_128_ts_helios_per_mod_patchdisc_contrastive_fix
trainer:
callbacks:
- class_path: lightning.pytorch.callbacks.LearningRateMonitor
init_args:
logging_interval: epoch
- class_path: lightning.pytorch.callbacks.ModelCheckpoint
init_args:
monitor: val_class/accuracy
save_last: true
save_top_k: 1
mode: max
- class_path: rslearn.train.callbacks.freeze_unfreeze.FreezeUnfreeze
init_args:
module_selector:
- model
- encoder
- 0
unfreeze_at_epoch: 20
unfreeze_lr_factor: 10.0
- class_path: rslearn.train.prediction_writer.RslearnWriter
init_args:
path: ${DATASET_PATH}
output_layer: ${PREDICTION_OUTPUT_LAYER}
selector: ["class"]
max_epochs: 500
Loading