update training results, hyper, issues

README.md

@@ -295,7 +295,7 @@ You can find a [table](training/results/README.md) of common benchmark datasets
 
 Some known issues to be aware of, if using and making new models or layers with `kgcnn`.
 * Jagged or nested Tensors loading into models for PyTorch backend is not working.
-* ForceModel does not support all backends and does not have Scaler layer yet.
+* ForceModel does not support all backends.
 * BatchNormalization layer dos not support padding yet.
 * Keras AUC metric does not seem to work for torch cuda.
 

training/hyper/hyper_qm9_energies.py

@@ -316,7 +316,7 @@
             "cross_validation": {"class_name": "KFold",
                                  "config": {"n_splits": 5, "random_state": 42, "shuffle": True}},
             "fit": {
-                "batch_size": 10, "epochs": 600, "validation_freq": 10, "verbose": 2, "callbacks": []
+                "batch_size": 32, "epochs": 600, "validation_freq": 10, "verbose": 2, "callbacks": []
             },
             "compile": {
                 "optimizer": {
@@ -442,7 +442,7 @@
                                   "angle_attributes": "angle_attributes_2"}}
                 ]
             },
-            "data_unit": "eV"
+            # "data_unit": "eV"
         },
         "info": {
             "postfix": "",
@@ -522,7 +522,7 @@
                     {"map_list": {"method": "set_range", "max_distance": 10, "max_neighbours": 10000}}
                 ]
             },
-            "data_unit": "eV"
+            # "data_unit": "eV"
         },
         "info": {
             "postfix": "",

training/hyper/hyper_qm9_orbitals.py

@@ -310,7 +310,7 @@
             "cross_validation": {"class_name": "KFold",
                                  "config": {"n_splits": 5, "random_state": 42, "shuffle": True}},
             "fit": {
-                "batch_size": 10, "epochs": 600, "validation_freq": 10, "verbose": 2, "callbacks": []
+                "batch_size": 32, "epochs": 600, "validation_freq": 10, "verbose": 2, "callbacks": []
             },
             "compile": {
                 "optimizer": {
@@ -434,7 +434,7 @@
                                   "angle_attributes": "angle_attributes_2"}}
                 ]
             },
-            "data_unit": "eV"
+            # "data_unit": "eV"
         },
         "info": {
             "postfix": "",
@@ -515,7 +515,7 @@
                     {"map_list": {"method": "set_range", "max_distance": 10, "max_neighbours": 10000}}
                 ]
             },
-            "data_unit": "eV"
+            # "data_unit": "eV"
         },
         "info": {
             "postfix": "",

training/results/ESOLDataset/HamNet/HamNet_ESOLDataset_score.yaml

@@ -0,0 +1,137 @@
+OS: posix_linux
+backend: tensorflow
+cuda_available: 'True'
+data_unit: mol/L
+date_time: '2024-01-16 21:12:16'
+device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'',
+  device_type=''GPU'')]'
+device_memory: '[]'
+device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V100-SXM2-32GB''}]'
+epochs:
+- 400
+- 400
+- 400
+- 400
+- 400
+execute_folds: null
+kgcnn_version: 4.0.0
+loss:
+- 0.0194633137434721
+- 0.027471264824271202
+- 0.024131199344992638
+- 0.023608962073922157
+- 0.024778731167316437
+max_loss:
+- 0.9549863934516907
+- 0.9460381865501404
+- 0.9562951922416687
+- 0.9522734880447388
+- 0.93553626537323
+max_scaled_mean_absolute_error:
+- 1.6129345893859863
+- 1.6329313516616821
+- 1.6552671194076538
+- 1.6240527629852295
+- 1.6248692274093628
+max_scaled_root_mean_squared_error:
+- 2.0324018001556396
+- 2.0522613525390625
+- 2.0783069133758545
+- 2.0291409492492676
+- 2.0363149642944336
+max_val_loss:
+- 0.9976205229759216
+- 0.736245334148407
+- 0.7551329135894775
+- 1.004599928855896
+- 0.8924134373664856
+max_val_scaled_mean_absolute_error:
+- 1.6911805868148804
+- 1.443051815032959
+- 1.4413104057312012
+- 1.5957467555999756
+- 1.5963668823242188
+max_val_scaled_root_mean_squared_error:
+- 2.069248914718628
+- 1.8094627857208252
+- 1.879884958267212
+- 2.082547664642334
+- 1.98152756690979
+min_loss:
+- 0.01858583278954029
+- 0.02215714007616043
+- 0.0208753552287817
+- 0.01921209879219532
+- 0.019109230488538742
+min_scaled_mean_absolute_error:
+- 0.19970214366912842
+- 0.22037352621555328
+- 0.21708956360816956
+- 0.20531746745109558
+- 0.20399823784828186
+min_scaled_root_mean_squared_error:
+- 0.28398433327674866
+- 0.31362035870552063
+- 0.3070819079875946
+- 0.28848695755004883
+- 0.29025784134864807
+min_val_loss:
+- 0.11382653564214706
+- 0.1074131429195404
+- 0.12269075959920883
+- 0.11106088012456894
+- 0.10852447152137756
+min_val_scaled_mean_absolute_error:
+- 0.5177220106124878
+- 0.4899892807006836
+- 0.5187506079673767
+- 0.5088114142417908
+- 0.5194295644760132
+min_val_scaled_root_mean_squared_error:
+- 0.6913321614265442
+- 0.6793058514595032
+- 0.7476562261581421
+- 0.6959748864173889
+- 0.6894879937171936
+model_class: make_model
+model_name: HamNet
+model_version: '2023-09-07'
+multi_target_indices: null
+number_histories: 5
+scaled_mean_absolute_error:
+- 0.2109655737876892
+- 0.24282591044902802
+- 0.22718660533428192
+- 0.2207445502281189
+- 0.22914427518844604
+scaled_root_mean_squared_error:
+- 0.2910397946834564
+- 0.3478170335292816
+- 0.3304538130760193
+- 0.3202421963214874
+- 0.33203455805778503
+seed: 42
+time_list:
+- '0:03:35.549083'
+- '0:03:45.836377'
+- '0:03:44.205630'
+- '0:03:43.842392'
+- '0:03:40.104215'
+val_loss:
+- 0.13075850903987885
+- 0.11916299909353256
+- 0.13162805140018463
+- 0.13982142508029938
+- 0.11291807889938354
+val_scaled_mean_absolute_error:
+- 0.5587661862373352
+- 0.5337684750556946
+- 0.5608881115913391
+- 0.5584719777107239
+- 0.5274047255516052
+val_scaled_root_mean_squared_error:
+- 0.746735155582428
+- 0.7206915616989136
+- 0.7735453248023987
+- 0.7712847590446472
+- 0.696090579032898

training/results/ESOLDataset/HamNet/HamNet_hyper.json

@@ -0,0 +1 @@
+{"model": {"class_name": "make_model", "module_name": "kgcnn.literature.HamNet", "config": {"name": "HamNet", "inputs": [{"shape": [null, 41], "name": "node_attributes", "dtype": "float32"}, {"shape": [null, 3], "name": "node_coordinates", "dtype": "float32"}, {"shape": [null, 11], "name": "edge_attributes", "dtype": "float32"}, {"shape": [null, 2], "name": "edge_indices", "dtype": "int64"}, {"shape": [], "name": "total_nodes", "dtype": "int64"}, {"shape": [], "name": "total_edges", "dtype": "int64"}], "input_embedding": null, "input_node_embedding": {"input_dim": 95, "output_dim": 64}, "input_edge_embedding": {"input_dim": 5, "output_dim": 64}, "message_kwargs": {"units": 200, "units_edge": 200, "rate": 0.5, "use_dropout": true}, "fingerprint_kwargs": {"units": 200, "units_attend": 200, "rate": 0.5, "use_dropout": true, "depth": 3}, "gru_kwargs": {"units": 200}, "verbose": 10, "depth": 3, "union_type_node": "gru", "union_type_edge": "None", "given_coordinates": true, "output_embedding": "graph", "output_mlp": {"use_bias": [true, false], "units": [200, 1], "activation": ["relu", "linear"], "use_dropout": [true, false], "rate": [0.5, 0.0]}}}, "training": {"fit": {"batch_size": 40, "epochs": 400, "validation_freq": 1, "verbose": 2, "callbacks": []}, "compile": {"optimizer": {"class_name": "AdamW", "config": {"learning_rate": 0.001, "weight_decay": 1e-05}}, "loss": "mean_squared_error"}, "cross_validation": {"class_name": "KFold", "config": {"n_splits": 5, "random_state": 42, "shuffle": true}}, "scaler": {"class_name": "StandardLabelScaler", "config": {"with_std": true, "with_mean": true, "copy": true}}}, "data": {"data_unit": "mol/L"}, "info": {"postfix": "", "postfix_file": "", "kgcnn_version": "4.0.0"}, "dataset": {"class_name": "ESOLDataset", "module_name": "kgcnn.data.datasets.ESOLDataset", "config": {}, "methods": [{"set_attributes": {}}, {"map_list": {"method": "count_nodes_and_edges"}}]}}

training/results/MD17Dataset/PAiNN_EnergyForceModel/PAiNN_MD17Dataset_score_aspirin_ccsd.yaml

@@ -0,0 +1,58 @@
+OS: posix_linux
+backend: tensorflow
+cuda_available: 'True'
+data_unit: ''
+date_time: '2024-01-19 19:17:46'
+device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'',
+  device_type=''GPU'')]'
+device_memory: '[]'
+device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V100-SXM2-32GB''}]'
+energy_scaled_mean_absolute_error:
+- 0.20170779526233673
+epochs:
+- 1000
+execute_folds: null
+force_scaled_mean_absolute_error:
+- 0.27192622423171997
+kgcnn_version: 4.0.0
+loss:
+- 0.040751248598098755
+max_energy_scaled_mean_absolute_error:
+- 12.22606086730957
+max_force_scaled_mean_absolute_error:
+- 19.689008712768555
+max_loss:
+- 2.975248336791992
+max_val_energy_scaled_mean_absolute_error:
+- 9.95431900024414
+max_val_force_scaled_mean_absolute_error:
+- 14.166427612304688
+max_val_loss:
+- 2.147376298904419
+min_energy_scaled_mean_absolute_error:
+- 0.20170779526233673
+min_force_scaled_mean_absolute_error:
+- 0.27192622423171997
+min_loss:
+- 0.040751248598098755
+min_val_energy_scaled_mean_absolute_error:
+- 0.2632357180118561
+min_val_force_scaled_mean_absolute_error:
+- 0.8550477623939514
+min_val_loss:
+- 0.12949016690254211
+model_class: EnergyForceModel
+model_name: PAiNN
+model_version: ''
+multi_target_indices: null
+number_histories: 1
+seed: 42
+time_list:
+- '0:20:41.188479'
+trajectory_name: aspirin_ccsd
+val_energy_scaled_mean_absolute_error:
+- 0.2693638503551483
+val_force_scaled_mean_absolute_error:
+- 0.8550565242767334
+val_loss:
+- 0.12949059903621674

training/results/MD17Dataset/PAiNN_EnergyForceModel/PAiNN_MD17Dataset_score_malonaldehyde_ccsd_t.yaml

@@ -0,0 +1,58 @@
+OS: posix_linux
+backend: tensorflow
+cuda_available: 'True'
+data_unit: ''
+date_time: '2024-01-25 10:55:36'
+device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'',
+  device_type=''GPU'')]'
+device_memory: '[]'
+device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V100-SXM2-32GB''}]'
+energy_scaled_mean_absolute_error:
+- 0.108326256275177
+epochs:
+- 1000
+execute_folds: null
+force_scaled_mean_absolute_error:
+- 0.1324000060558319
+kgcnn_version: 4.0.0
+loss:
+- 0.02997162938117981
+max_energy_scaled_mean_absolute_error:
+- 9.238640785217285
+max_force_scaled_mean_absolute_error:
+- 19.9771671295166
+max_loss:
+- 4.552161693572998
+max_val_energy_scaled_mean_absolute_error:
+- 18.43704605102539
+max_val_force_scaled_mean_absolute_error:
+- 14.447487831115723
+max_val_loss:
+- 3.3687236309051514
+min_energy_scaled_mean_absolute_error:
+- 0.10826544463634491
+min_force_scaled_mean_absolute_error:
+- 0.13237859308719635
+min_loss:
+- 0.02996666543185711
+min_val_energy_scaled_mean_absolute_error:
+- 0.14528214931488037
+min_val_force_scaled_mean_absolute_error:
+- 0.7742980718612671
+min_val_loss:
+- 0.1764606386423111
+model_class: EnergyForceModel
+model_name: PAiNN
+model_version: ''
+multi_target_indices: null
+number_histories: 1
+seed: 42
+time_list:
+- '0:17:02.875278'
+trajectory_name: malonaldehyde_ccsd_t
+val_energy_scaled_mean_absolute_error:
+- 0.14697757363319397
+val_force_scaled_mean_absolute_error:
+- 0.7749027013778687
+val_loss:
+- 0.1765810251235962

training/results/MD17Dataset/PAiNN_EnergyForceModel/PAiNN_hyper_aspirin_ccsd.json

@@ -0,0 +1 @@
+{"model": {"class_name": "EnergyForceModel", "module_name": "kgcnn.models.force", "config": {"name": "PAiNN", "nested_model_config": true, "output_to_tensor": false, "output_squeeze_states": true, "coordinate_input": 1, "inputs": [{"shape": [null], "name": "atomic_number", "dtype": "int32"}, {"shape": [null, 3], "name": "node_coordinates", "dtype": "float32"}, {"shape": [null, 2], "name": "range_indices", "dtype": "int64"}, {"shape": [], "name": "total_nodes", "dtype": "int64"}, {"shape": [], "name": "total_ranges", "dtype": "int64"}], "model_energy": {"class_name": "make_model", "module_name": "kgcnn.literature.PAiNN", "config": {"name": "PAiNNEnergy", "inputs": [{"shape": [null], "name": "atomic_number", "dtype": "int32"}, {"shape": [null, 3], "name": "node_coordinates", "dtype": "float32"}, {"shape": [null, 2], "name": "range_indices", "dtype": "int64"}, {"shape": [], "name": "total_nodes", "dtype": "int64"}, {"shape": [], "name": "total_ranges", "dtype": "int64"}], "input_embedding": null, "input_node_embedding": {"input_dim": 95, "output_dim": 128}, "equiv_initialize_kwargs": {"dim": 3, "method": "eps"}, "bessel_basis": {"num_radial": 20, "cutoff": 5.0, "envelope_exponent": 5}, "pooling_args": {"pooling_method": "scatter_sum"}, "conv_args": {"units": 128, "cutoff": null}, "update_args": {"units": 128}, "depth": 3, "verbose": 10, "output_embedding": "graph", "output_mlp": {"use_bias": [true, true], "units": [128, 1], "activation": ["swish", "linear"]}}}, "outputs": {"energy": {"name": "energy", "shape": [1]}, "force": {"name": "force", "shape": [null, 3]}}}}, "training": {"fit": {"batch_size": 32, "epochs": 1000, "validation_freq": 1, "verbose": 2, "callbacks": []}, "compile": {"optimizer": {"class_name": "Adam", "config": {"learning_rate": {"class_name": "kgcnn>LinearWarmupExponentialDecay", "config": {"learning_rate": 0.001, "warmup_steps": 150.0, "decay_steps": 20000.0, "decay_rate": 0.01}}, "amsgrad": true, "use_ema": true}}, "loss_weights": {"energy": 0.02, "force": 0.98}}, "scaler": {"class_name": "EnergyForceExtensiveLabelScaler", "config": {"standardize_scale": true}}}, "data": {}, "dataset": {"class_name": "MD17Dataset", "module_name": "kgcnn.data.datasets.MD17Dataset", "config": {"trajectory_name": "aspirin_ccsd"}, "methods": [{"rename_property_on_graphs": {"old_property_name": "E", "new_property_name": "energy"}}, {"rename_property_on_graphs": {"old_property_name": "F", "new_property_name": "force"}}, {"rename_property_on_graphs": {"old_property_name": "z", "new_property_name": "atomic_number"}}, {"rename_property_on_graphs": {"old_property_name": "R", "new_property_name": "node_coordinates"}}, {"map_list": {"method": "set_range", "max_distance": 5, "max_neighbours": 10000, "node_coordinates": "node_coordinates"}}, {"map_list": {"method": "count_nodes_and_edges", "total_edges": "total_ranges", "count_edges": "range_indices", "count_nodes": "atomic_number", "total_nodes": "total_nodes"}}]}, "info": {"postfix": "", "postfix_file": "_aspirin_ccsd", "kgcnn_version": "4.0.0"}}

training/results/MD17Dataset/PAiNN_EnergyForceModel/PAiNN_hyper_malonaldehyde_ccsd_t.json

@@ -0,0 +1 @@
+{"model": {"class_name": "EnergyForceModel", "module_name": "kgcnn.models.force", "config": {"name": "PAiNN", "nested_model_config": true, "output_to_tensor": false, "output_squeeze_states": true, "coordinate_input": 1, "inputs": [{"shape": [null], "name": "atomic_number", "dtype": "int32"}, {"shape": [null, 3], "name": "node_coordinates", "dtype": "float32"}, {"shape": [null, 2], "name": "range_indices", "dtype": "int64"}, {"shape": [], "name": "total_nodes", "dtype": "int64"}, {"shape": [], "name": "total_ranges", "dtype": "int64"}], "model_energy": {"class_name": "make_model", "module_name": "kgcnn.literature.PAiNN", "config": {"name": "PAiNNEnergy", "inputs": [{"shape": [null], "name": "atomic_number", "dtype": "int32"}, {"shape": [null, 3], "name": "node_coordinates", "dtype": "float32"}, {"shape": [null, 2], "name": "range_indices", "dtype": "int64"}, {"shape": [], "name": "total_nodes", "dtype": "int64"}, {"shape": [], "name": "total_ranges", "dtype": "int64"}], "input_embedding": null, "input_node_embedding": {"input_dim": 95, "output_dim": 128}, "equiv_initialize_kwargs": {"dim": 3, "method": "eps"}, "bessel_basis": {"num_radial": 20, "cutoff": 5.0, "envelope_exponent": 5}, "pooling_args": {"pooling_method": "scatter_sum"}, "conv_args": {"units": 128, "cutoff": null}, "update_args": {"units": 128}, "depth": 3, "verbose": 10, "output_embedding": "graph", "output_mlp": {"use_bias": [true, true], "units": [128, 1], "activation": ["swish", "linear"]}}}, "outputs": {"energy": {"name": "energy", "shape": [1]}, "force": {"name": "force", "shape": [null, 3]}}}}, "training": {"fit": {"batch_size": 32, "epochs": 1000, "validation_freq": 1, "verbose": 2, "callbacks": []}, "compile": {"optimizer": {"class_name": "Adam", "config": {"learning_rate": {"class_name": "kgcnn>LinearWarmupExponentialDecay", "config": {"learning_rate": 0.001, "warmup_steps": 150.0, "decay_steps": 20000.0, "decay_rate": 0.01}}, "amsgrad": true, "use_ema": true}}, "loss_weights": {"energy": 0.02, "force": 0.98}}, "scaler": {"class_name": "EnergyForceExtensiveLabelScaler", "config": {"standardize_scale": true}}}, "data": {}, "dataset": {"class_name": "MD17Dataset", "module_name": "kgcnn.data.datasets.MD17Dataset", "config": {"trajectory_name": "malonaldehyde_ccsd_t"}, "methods": [{"rename_property_on_graphs": {"old_property_name": "E", "new_property_name": "energy"}}, {"rename_property_on_graphs": {"old_property_name": "F", "new_property_name": "force"}}, {"rename_property_on_graphs": {"old_property_name": "z", "new_property_name": "atomic_number"}}, {"rename_property_on_graphs": {"old_property_name": "R", "new_property_name": "node_coordinates"}}, {"map_list": {"method": "set_range", "max_distance": 5, "max_neighbours": 10000, "node_coordinates": "node_coordinates"}}, {"map_list": {"method": "count_nodes_and_edges", "total_edges": "total_ranges", "count_edges": "range_indices", "count_nodes": "atomic_number", "total_nodes": "total_nodes"}}]}, "info": {"postfix": "", "postfix_file": "_malonaldehyde_ccsd_t", "kgcnn_version": "4.0.0"}}