WorldCereal presto-worldcereal issues

WorldCereal / presto-worldcereal

10 stars 0 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Inconsistent use of `device` when running inference on system with GPU availability

#118 kvantricht opened 6 days ago
0
Valid month and mask debugging

#115 cbutsko closed 1 month ago
0
Timestep position debugging

#114 cbutsko closed 1 month ago
0
Check handling of `Dataloader` in eval task

#113 kvantricht opened 1 month ago
1
Merge `dekadal-for-merge` branch into `croptype`

#112 kvantricht opened 1 month ago
1
`available_timesteps` might be wrongly computed (and needs some documentation)

#111 kvantricht opened 1 month ago
1
Augmentation sometimes fails with `ValueError: low >= high` in `np.random.randint`

#110 kvantricht opened 1 month ago
4
Augmentation not performed for monthly dataset

#109 kvantricht opened 1 month ago
1
Avoid global `matplotlib` and `CLASS_MAPPINGS` loading in `utils.py`

#108 kvantricht closed 1 month ago
3
Temporal jittering on the `valid_date` token

#107 kvantricht opened 1 month ago
0
105-unify-inference

#106 kvantricht closed 2 months ago
0
Align `WorldCerealInferenceDataset` and `inference.py`

#105 kvantricht closed 2 months ago
1
Using new parquet in train

#104 cbutsko opened 2 months ago
3
Updated WorldCerealInferenceDataset

#103 kvantricht closed 2 months ago
6
Masking during finetuning

#102 gabrieltseng closed 2 months ago
6
#98 Remove duplicate band remapping

#99 kvantricht closed 2 months ago
0
Remove duplicate band name remapping in inference

#98 kvantricht closed 2 months ago
0
Wc-dataset-fix

#97 kvantricht closed 2 months ago
0
Wc-dataset-fix

#96 kvantricht closed 2 months ago
1
Accept optional date in `process_parquet` for more precise subsetting

#95 kvantricht opened 2 months ago
1
Inference-bugfixes

#94 kvantricht closed 2 months ago
0
Fix DEM/slope handling during inference

#93 kvantricht closed 2 months ago
2
Finetune Presto with masking of specific input channels

#92 kvantricht closed 2 months ago
1
Fixed band mapping :facepalm:

#91 kvantricht closed 2 months ago
0
Fixed band mapping :facepalm:

#90 kvantricht closed 2 months ago
1
Add warnings when expected input datastreams are missing in inference

#89 kvantricht closed 2 months ago
1
Test compilation for inference

#88 kvantricht closed 2 months ago
1
Small improvement computing encodings

#87 kvantricht closed 3 months ago
0
Isolated `prep_dataframe` method

#86 kvantricht closed 4 months ago
0
Use proper nodata value for B8A in parquet processing

#85 kvantricht closed 4 months ago
0
Layout of worldcereal-presto paper outline

#84 kvantricht opened 4 months ago
0
Implement augmentation during Presto finetuning

#83 kvantricht closed 3 weeks ago
1
Implement batch balancing for presto croptype finetuning

#82 kvantricht closed 3 weeks ago
0
Balancing class representation in the dataset for crop type classification

#81 gabrieltseng opened 4 months ago
0
Use torch==2.3.1

#80 kvantricht closed 4 months ago
0
Embeddings for new parquet/dataframe input

#79 cbutsko closed 4 months ago
0
Make `batch_size` configurable for inference

#78 kvantricht closed 5 months ago
0
Blackified

#77 GriffinBabe closed 5 months ago
0
72 lightweight dependencies

#76 kvantricht closed 5 months ago
0
Decadal ss training

#75 giollimirgia opened 5 months ago
2
Simplify requirements

#74 gabrieltseng closed 5 months ago
0
72-lightweight-dependencies

#73 kvantricht closed 5 months ago
0
Lightweight dependencies for presto-worldcereal in inference mode

#72 kvantricht closed 5 months ago
0
Croptype

#71 gabrieltseng opened 5 months ago
3
Added inference module for openeo

#70 kvantricht closed 5 months ago
0
Create experiment log for Presto runs

#69 cbutsko closed 5 months ago
0
Collect all the relevant Presto and CatBoost models for the future use

#68 cbutsko closed 4 months ago
2
Implement Presto finetuning for croptype task

#67 cbutsko opened 5 months ago
1
update Presto for Self-Supervised training on 10D WorldCereal data

#66 giollimirgia closed 4 months ago
1
fix: latlon handling in inference dataset

#65 kvantricht closed 2 months ago
8