OneScience-Group/PrecipDD

1{2  "model_name": "PrecipDD",3  "model_type": "precipdd",4  "architectures": ["PrecipDD"],5  "framework": "PyTorch",6  "domain": "climate",7  "task": "daily-precipitation-anomaly-to-agmt-regression",8  "implementation": {"entry_point": "model/precipdd.py", "scope": "dimension-faithful reduced-sample engineering reproduction", "train_script": "scripts/train.py", "inference_script": "scripts/inference.py", "evaluation_script": "scripts/result.py", "synthetic_data_script": "scripts/fake_data.py"},9  "architecture": {"input_shape": [1,55,160], "conv_layers": 5, "filters": [8,8,16,16,16], "pooled_feature_shape": [16,14,40], "flattened_features": 8960, "dense": [32,1], "output_activation": "linear"},10  "data": {"datasets": ["CESM2 Large Ensemble"], "format_version": "precipdd_v1", "layout": "NCHW", "paper_members": 80, "paper_period": [1850,2100], "synthetic": true},11  "configuration_sources": ["conf/config.yaml","model/precipdd.py","scripts/fake_data.py","scripts/train.py","scripts/inference.py","scripts/result.py"]12}13