From e90f97f111b517056426c52590135cdbc02cdc3d Mon Sep 17 00:00:00 2001 From: PatReis Date: Fri, 3 May 2024 09:09:35 +0200 Subject: [PATCH] Updated training results --- .../DimeNetPP_QM9Dataset_score_H.yaml | 66 ++++- .../DimeNetPP_QM9Dataset_score_HOMO.yaml | 66 ++++- .../DimeNetPP_QM9Dataset_score_U0.yaml | 139 +++++++++++ .../DimeNetPP/DimeNetPP_hyper_U0.json | 1 + .../EGNN/EGNN_QM9Dataset_score_H.yaml | 75 +++++- .../EGNN/EGNN_QM9Dataset_score_HOMO.yaml | 75 +++++- .../EGNN/EGNN_QM9Dataset_score_U0.yaml | 157 ++++++++++++ .../QM9Dataset/EGNN/EGNN_hyper_U0.json | 1 + .../MXMNet/MXMNet_QM9Dataset_score_H.yaml | 111 ++++++++- .../MXMNet/MXMNet_QM9Dataset_score_HOMO.yaml | 111 ++++++++- .../MXMNet/MXMNet_QM9Dataset_score_U0.yaml | 229 ++++++++++++++++++ .../QM9Dataset/MXMNet/MXMNet_hyper_U0.json | 1 + .../Megnet/Megnet_QM9Dataset_score_H.yaml | 75 +++++- .../Megnet/Megnet_QM9Dataset_score_HOMO.yaml | 75 +++++- .../Megnet/Megnet_QM9Dataset_score_U0.yaml | 157 ++++++++++++ .../QM9Dataset/Megnet/Megnet_hyper_U0.json | 1 + .../NMPN/NMPN_QM9Dataset_score_H.yaml | 75 +++++- .../NMPN/NMPN_QM9Dataset_score_HOMO.yaml | 75 +++++- .../NMPN/NMPN_QM9Dataset_score_U0.yaml | 157 ++++++++++++ .../QM9Dataset/NMPN/NMPN_hyper_U0.json | 1 + training/results/README.md | 24 +- 21 files changed, 1635 insertions(+), 37 deletions(-) create mode 100644 training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_U0.yaml create mode 100644 training/results/QM9Dataset/DimeNetPP/DimeNetPP_hyper_U0.json create mode 100644 training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_U0.yaml create mode 100644 training/results/QM9Dataset/EGNN/EGNN_hyper_U0.json create mode 100644 training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_U0.yaml create mode 100644 training/results/QM9Dataset/MXMNet/MXMNet_hyper_U0.json create mode 100644 training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_U0.yaml create mode 100644 training/results/QM9Dataset/Megnet/Megnet_hyper_U0.json create mode 100644 training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_U0.yaml create mode 100644 training/results/QM9Dataset/NMPN/NMPN_hyper_U0.json diff --git a/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_H.yaml b/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_H.yaml index a1488c9e..b0cabb9f 100644 --- a/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_H.yaml +++ b/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_H.yaml @@ -2,7 +2,7 @@ OS: posix_linux backend: tensorflow cuda_available: 'True' data_unit: '[''eV'']' -date_time: '2024-03-25 04:26:18' +date_time: '2024-04-06 03:35:47' device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'', device_type=''GPU'')]' device_memory: '[]' @@ -10,46 +10,88 @@ device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V10 epochs: - 600 - 600 +- 600 +- 600 +- 600 execute_folds: -- 3 +- 0 kgcnn_version: 4.0.1 loss: +- 0.0028725878801196814 +- 0.0026110385078936815 +- 0.0027430588379502296 - 0.0030547985807061195 - 0.0031594946049153805 max_loss: +- 0.183021679520607 +- 0.18046464025974274 +- 0.18171830475330353 - 0.18200941383838654 - 0.18039806187152863 max_scaled_mean_absolute_error: +- 0.19949166476726532 +- 0.19701921939849854 +- 0.19832226634025574 - 0.19871577620506287 - 0.1971033662557602 max_scaled_root_mean_squared_error: +- 0.3373342752456665 +- 0.3346664607524872 +- 0.3374268114566803 - 0.3368868827819824 - 0.3338407874107361 max_val_loss: +- 0.028285875916481018 +- 0.026908351108431816 +- 0.03353552892804146 - 0.026954108849167824 - 0.031196754425764084 max_val_scaled_mean_absolute_error: +- 0.030688278377056122 +- 0.029262633994221687 +- 0.03644924983382225 - 0.029313750565052032 - 0.03397288918495178 max_val_scaled_root_mean_squared_error: +- 0.05506062135100365 +- 0.05016918107867241 +- 0.062191907316446304 - 0.05636412277817726 - 0.059521134942770004 min_loss: +- 0.002750977175310254 +- 0.0026110385078936815 +- 0.0027430588379502296 - 0.0026739316526800394 - 0.0028502270579338074 min_scaled_mean_absolute_error: +- 0.0029984619468450546 +- 0.0028504282236099243 +- 0.002993666799739003 - 0.0029193637892603874 - 0.0031141082290560007 min_scaled_root_mean_squared_error: +- 0.004070690367370844 +- 0.0038469485007226467 +- 0.004056457430124283 - 0.004010347183793783 - 0.004236821085214615 min_val_loss: +- 0.008693299256265163 +- 0.008540540002286434 +- 0.00872416514903307 - 0.008577571250498295 - 0.008637722581624985 min_val_scaled_mean_absolute_error: +- 0.009394599124789238 +- 0.009247254580259323 +- 0.00943470187485218 - 0.009276311844587326 - 0.009365281090140343 min_val_scaled_root_mean_squared_error: +- 0.03038923628628254 +- 0.02790170907974243 +- 0.030592383816838264 - 0.03182884305715561 - 0.032217614352703094 model_class: make_model @@ -57,23 +99,41 @@ model_name: DimeNetPP model_version: '2023-12-04' multi_target_indices: - 12 -number_histories: 2 +number_histories: 5 scaled_mean_absolute_error: +- 0.003131032455712557 +- 0.0028504282236099243 +- 0.002993666799739003 - 0.003335044253617525 - 0.0034518814645707607 scaled_root_mean_squared_error: +- 0.0042451610788702965 +- 0.0038469485007226467 +- 0.004056457430124283 - 0.00451664999127388 - 0.004660776816308498 seed: 42 time_list: +- 1 day, 8:10:40.988993 +- 1 day, 10:07:29.205856 +- 1 day, 9:27:36.518466 - 1 day, 10:50:05.903201 - 1 day, 12:02:47.644100 val_loss: +- 0.009649361483752728 +- 0.008774645626544952 +- 0.009712413884699345 - 0.009079432114958763 - 0.009007935412228107 val_scaled_mean_absolute_error: +- 0.01044066809117794 +- 0.009502683766186237 +- 0.010507510975003242 - 0.009823390282690525 - 0.009767130017280579 val_scaled_root_mean_squared_error: +- 0.03147687390446663 +- 0.028618615120649338 +- 0.031682539731264114 - 0.0325394906103611 - 0.032723262906074524 diff --git a/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_HOMO.yaml b/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_HOMO.yaml index 56ea1212..44748b78 100644 --- a/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_HOMO.yaml +++ b/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_HOMO.yaml @@ -2,7 +2,7 @@ OS: posix_linux backend: tensorflow cuda_available: 'True' data_unit: '[''eV'']' -date_time: '2024-03-24 18:18:12' +date_time: '2024-04-05 17:42:42' device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'', device_type=''GPU'')]' device_memory: '[]' @@ -10,46 +10,88 @@ device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V10 epochs: - 600 - 600 +- 600 +- 600 +- 600 execute_folds: -- 3 +- 0 kgcnn_version: 4.0.1 loss: +- 0.005713232327252626 +- 0.005560360848903656 +- 0.005917480681091547 - 0.005621917080134153 - 0.00561577919870615 max_loss: +- 0.32745715975761414 +- 0.33054694533348083 +- 0.33204618096351624 - 0.3294568657875061 - 0.33053842186927795 max_scaled_mean_absolute_error: +- 0.19764377176761627 +- 0.19908972084522247 +- 0.19949591159820557 - 0.19851791858673096 - 0.19896641373634338 max_scaled_root_mean_squared_error: +- 0.2918851971626282 +- 0.29421722888946533 +- 0.29252153635025024 - 0.2926676273345947 - 0.29251179099082947 max_val_loss: +- 0.11165573447942734 +- 0.09737082570791245 +- 0.11071251332759857 - 0.10414715856313705 - 0.1003655269742012 max_val_scaled_mean_absolute_error: +- 0.06727948784828186 +- 0.05857941880822182 +- 0.0663936585187912 - 0.06269031018018723 - 0.06037885695695877 max_val_scaled_root_mean_squared_error: +- 0.09674684703350067 +- 0.08620665967464447 +- 0.10019882768392563 - 0.09375123679637909 - 0.08823802322149277 min_loss: +- 0.005554972682148218 +- 0.005560360848903656 +- 0.005845641251653433 - 0.005608322098851204 - 0.005492119584232569 min_scaled_mean_absolute_error: +- 0.003352767089381814 +- 0.003348978003486991 +- 0.0035118910018354654 - 0.0033793302718549967 - 0.0033058912958949804 min_scaled_root_mean_squared_error: +- 0.004644325468689203 +- 0.00463532842695713 +- 0.004878734238445759 - 0.0046747042797505856 - 0.0046160463243722916 min_val_loss: +- 0.04813332483172417 +- 0.047232165932655334 +- 0.04910600185394287 - 0.04799223318696022 - 0.04828876256942749 min_val_scaled_mean_absolute_error: +- 0.028955692425370216 +- 0.028396088629961014 +- 0.029412847012281418 - 0.02886221557855606 - 0.029051648452878 min_val_scaled_root_mean_squared_error: +- 0.0533822663128376 +- 0.04810069501399994 +- 0.052283599972724915 - 0.04968913644552231 - 0.05029897391796112 model_class: make_model @@ -57,23 +99,41 @@ model_name: DimeNetPP model_version: '2023-12-04' multi_target_indices: - 5 -number_histories: 2 +number_histories: 5 scaled_mean_absolute_error: +- 0.0034483682829886675 +- 0.003348978003486991 +- 0.003555207047611475 - 0.0033875287044793367 - 0.0033803347032517195 scaled_root_mean_squared_error: +- 0.00481433467939496 +- 0.00463532842695713 +- 0.004918843973428011 - 0.004697032272815704 - 0.004739508498460054 seed: 42 time_list: +- 1 day, 8:39:58.267289 +- 1 day, 8:06:06.323344 +- 1 day, 7:41:10.019637 - 1 day, 9:46:40.387562 - 1 day, 9:59:13.148100 val_loss: +- 0.048546645790338516 +- 0.04728297144174576 +- 0.049677859991788864 - 0.04799223318696022 - 0.04854681342840195 val_scaled_mean_absolute_error: +- 0.029205242171883583 +- 0.028426313772797585 +- 0.02975715510547161 - 0.02886221557855606 - 0.029207797721028328 val_scaled_root_mean_squared_error: +- 0.05359412357211113 +- 0.04821906238794327 +- 0.0529012605547905 - 0.0497230626642704 - 0.050450123846530914 diff --git a/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_U0.yaml b/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_U0.yaml new file mode 100644 index 00000000..dd7175f2 --- /dev/null +++ b/training/results/QM9Dataset/DimeNetPP/DimeNetPP_QM9Dataset_score_U0.yaml @@ -0,0 +1,139 @@ +OS: posix_linux +backend: tensorflow +cuda_available: 'True' +data_unit: '[''eV'']' +date_time: '2024-04-28 21:02:41' +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: +- 600 +- 600 +- 600 +- 600 +- 600 +execute_folds: +- 4 +kgcnn_version: 4.0.1 +loss: +- 0.0026392473373562098 +- 0.002731921384111047 +- 0.0028873435221612453 +- 0.002945298096165061 +- 0.003149635624140501 +max_loss: +- 0.18285708129405975 +- 0.18223541975021362 +- 0.18204642832279205 +- 0.18123295903205872 +- 0.1777096539735794 +max_scaled_mean_absolute_error: +- 0.19973328709602356 +- 0.1993834376335144 +- 0.19911372661590576 +- 0.19830676913261414 +- 0.19458016753196716 +max_scaled_root_mean_squared_error: +- 0.33743229508399963 +- 0.33675724267959595 +- 0.33958911895751953 +- 0.33760958909988403 +- 0.3328561782836914 +max_val_loss: +- 0.028299810364842415 +- 0.03416652977466583 +- 0.02581869624555111 +- 0.029098335653543472 +- 0.032180193811655045 +max_val_scaled_mean_absolute_error: +- 0.03079795278608799 +- 0.037262458354234695 +- 0.028094956651329994 +- 0.031712841242551804 +- 0.03513992950320244 +max_val_scaled_root_mean_squared_error: +- 0.053669072687625885 +- 0.05935506895184517 +- 0.053301841020584106 +- 0.0567391999065876 +- 0.057167280465364456 +min_loss: +- 0.0024848871398717165 +- 0.002731921384111047 +- 0.0027678008191287518 +- 0.0027585651259869337 +- 0.002822358161211014 +min_scaled_mean_absolute_error: +- 0.002714161528274417 +- 0.0029889612924307585 +- 0.0030272812582552433 +- 0.003018327057361603 +- 0.0030901264399290085 +min_scaled_root_mean_squared_error: +- 0.0036881815176457167 +- 0.004039817489683628 +- 0.004107831511646509 +- 0.004117818083614111 +- 0.004227353259921074 +min_val_loss: +- 0.008630308322608471 +- 0.008305891416966915 +- 0.008546438068151474 +- 0.008595584891736507 +- 0.008666712790727615 +min_val_scaled_mean_absolute_error: +- 0.009331987239420414 +- 0.009004950523376465 +- 0.009255052544176579 +- 0.009304068982601166 +- 0.00941450335085392 +min_val_scaled_root_mean_squared_error: +- 0.033357586711645126 +- 0.028126170858740807 +- 0.031101016327738762 +- 0.03380666300654411 +- 0.030644871294498444 +model_class: make_model +model_name: DimeNetPP +model_version: '2023-12-04' +multi_target_indices: +- 10 +number_histories: 5 +scaled_mean_absolute_error: +- 0.0028827479109168053 +- 0.0029889612924307585 +- 0.0031579574570059776 +- 0.003222643630579114 +- 0.0034485801588743925 +scaled_root_mean_squared_error: +- 0.003917039837688208 +- 0.004039817489683628 +- 0.0042478726245462894 +- 0.004393288865685463 +- 0.00470581091940403 +seed: 42 +time_list: +- 1 day, 9:08:30.203211 +- 1 day, 8:47:51.767858 +- 1 day, 7:02:47.637455 +- 1 day, 9:40:50.219178 +- 1 day, 10:28:46.536116 +val_loss: +- 0.008863898925483227 +- 0.009176176972687244 +- 0.008725205436348915 +- 0.009186572395265102 +- 0.008666712790727615 +val_scaled_mean_absolute_error: +- 0.009587355889379978 +- 0.009963677264750004 +- 0.009450665675103664 +- 0.009953288361430168 +- 0.00941450335085392 +val_scaled_root_mean_squared_error: +- 0.03364449739456177 +- 0.028510110452771187 +- 0.031466495245695114 +- 0.034984342753887177 +- 0.030661096796393394 diff --git a/training/results/QM9Dataset/DimeNetPP/DimeNetPP_hyper_U0.json b/training/results/QM9Dataset/DimeNetPP/DimeNetPP_hyper_U0.json new file mode 100644 index 00000000..1db3c6db --- /dev/null +++ b/training/results/QM9Dataset/DimeNetPP/DimeNetPP_hyper_U0.json @@ -0,0 +1 @@ +{"model": {"class_name": "make_model", "module_name": "kgcnn.literature.DimeNetPP", "config": {"name": "DimeNetPP", "inputs": [{"shape": [null], "name": "node_number", "dtype": "int64", "ragged": true}, {"shape": [null, 3], "name": "node_coordinates", "dtype": "float32", "ragged": true}, {"shape": [null, 2], "name": "range_indices", "dtype": "int64", "ragged": true}, {"shape": [null, 2], "name": "angle_indices", "dtype": "int64", "ragged": true}], "input_tensor_type": "ragged", "input_embedding": null, "input_node_embedding": {"input_dim": 95, "output_dim": 128, "embeddings_initializer": {"class_name": "RandomUniform", "config": {"minval": -1.7320508075688772, "maxval": 1.7320508075688772}}}, "emb_size": 128, "out_emb_size": 256, "int_emb_size": 64, "basis_emb_size": 8, "num_blocks": 4, "num_spherical": 7, "num_radial": 6, "cutoff": 5.0, "envelope_exponent": 5, "num_before_skip": 1, "num_after_skip": 2, "num_dense_output": 3, "num_targets": 1, "extensive": true, "output_init": "zeros", "activation": "swish", "verbose": 10, "output_embedding": "graph", "use_output_mlp": false, "output_mlp": {}}}, "training": {"cross_validation": {"class_name": "KFold", "config": {"n_splits": 5, "random_state": 42, "shuffle": true}}, "fit": {"batch_size": 32, "epochs": 600, "validation_freq": 10, "verbose": 2, "callbacks": []}, "compile": {"optimizer": {"class_name": "Adam", "config": {"learning_rate": {"class_name": "kgcnn>LinearWarmupExponentialDecay", "config": {"learning_rate": 0.001, "warmup_steps": 3000.0, "decay_steps": 4000000.0, "decay_rate": 0.01}}, "use_ema": true, "amsgrad": true}}, "loss": "mean_absolute_error"}, "scaler": {"class_name": "QMGraphLabelScaler", "config": {"atomic_number": "node_number", "scaler": [{"class_name": "ExtensiveMolecularLabelScaler", "config": {}}]}}, "multi_target_indices": [10]}, "data": {}, "info": {"postfix": "", "postfix_file": "_U0", "kgcnn_version": "4.0.0"}, "dataset": {"class_name": "QM9Dataset", "module_name": "kgcnn.data.datasets.QM9Dataset", "config": {}, "methods": [{"map_list": {"method": "set_range", "max_distance": 5, "max_neighbours": 1000}}, {"map_list": {"method": "set_angle"}}]}} \ No newline at end of file diff --git a/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_H.yaml b/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_H.yaml index 7c738e1e..5f50ad12 100644 --- a/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_H.yaml +++ b/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_H.yaml @@ -2,7 +2,7 @@ OS: posix_linux backend: tensorflow cuda_available: 'True' data_unit: '[''eV'']' -date_time: '2024-03-23 17:02:52' +date_time: '2024-04-04 07:43:46' device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'', device_type=''GPU'')]' device_memory: '[]' @@ -10,55 +10,106 @@ device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V10 epochs: - 800 - 800 +- 800 +- 800 +- 800 execute_folds: -- 3 +- 0 kgcnn_version: 4.0.1 learning_rate: - 1.927654702527093e-09 - 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 loss: +- 0.0010894415900111198 +- 0.0010499722557142377 +- 0.0011891224421560764 - 0.0010756702395156026 - 0.0011008529691025615 max_learning_rate: - 0.0005000000237487257 - 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 max_loss: +- 6.503025531768799 +- 4.95654821395874 +- 3.308363437652588 - 4.407474517822266 - 3.408555746078491 max_scaled_mean_absolute_error: +- 7.090329170227051 +- 5.412769794464111 +- 3.611560821533203 - 4.813329219818115 - 3.725175380706787 max_scaled_root_mean_squared_error: +- 161.2344207763672 +- 99.43590545654297 +- 82.41368865966797 - 95.80032348632812 - 75.86785125732422 max_val_loss: +- 0.6937893629074097 +- 0.6822652220726013 +- 0.6792652606964111 - 0.6639952659606934 - 0.7039309740066528 max_val_scaled_mean_absolute_error: +- 0.7561386227607727 +- 0.7447496056556702 +- 0.7412288188934326 - 0.7249158024787903 - 0.7690656185150146 max_val_scaled_root_mean_squared_error: +- 0.9559394717216492 +- 0.9621168971061707 +- 0.9378513693809509 - 0.9143601059913635 - 0.9589051008224487 min_learning_rate: - 1.927654702527093e-09 - 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 min_loss: +- 0.0010894415900111198 +- 0.0010499722557142377 +- 0.0011891224421560764 - 0.0010756702395156026 - 0.0011008529691025615 min_scaled_mean_absolute_error: +- 0.0011876070639118552 +- 0.0011463997652754188 +- 0.0012979528401046991 - 0.0011744957882910967 - 0.0012029350036755204 min_scaled_root_mean_squared_error: +- 0.00166812923271209 +- 0.0015968704828992486 +- 0.0018322744872421026 - 0.0016492840368300676 - 0.0016721710562705994 min_val_loss: +- 0.00916870404034853 +- 0.00850798562169075 +- 0.009117784909904003 - 0.008818797767162323 - 0.009190385229885578 min_val_scaled_mean_absolute_error: +- 0.00997067615389824 +- 0.009267368353903294 +- 0.009916488081216812 - 0.009606427513062954 - 0.0100205447524786 min_val_scaled_root_mean_squared_error: +- 0.03467392548918724 +- 0.027963019907474518 +- 0.038177039474248886 - 0.03540008142590523 - 0.032759107649326324 model_class: make_model @@ -66,23 +117,41 @@ model_name: EGNN model_version: '2023-12-04' multi_target_indices: - 12 -number_histories: 2 +number_histories: 5 scaled_mean_absolute_error: +- 0.0011876070639118552 +- 0.0011463997652754188 +- 0.0012979528401046991 - 0.0011744957882910967 - 0.0012029350036755204 scaled_root_mean_squared_error: +- 0.00166812923271209 +- 0.0015968704828992486 +- 0.0018322744872421026 - 0.0016492840368300676 - 0.0016721710562705994 seed: 42 time_list: +- '8:57:24.824495' +- '9:07:35.840109' +- '9:06:36.365104' - '8:53:14.618473' - '9:22:14.404742' val_loss: +- 0.009168821386992931 +- 0.00850798562169075 +- 0.009117973037064075 - 0.008818797767162323 - 0.009190516546368599 val_scaled_mean_absolute_error: +- 0.009970810264348984 +- 0.009267368353903294 +- 0.009916694834828377 - 0.009606427513062954 - 0.010020695626735687 val_scaled_root_mean_squared_error: +- 0.035058338195085526 +- 0.02809295244514942 +- 0.03893856331706047 - 0.03576206788420677 - 0.03538968786597252 diff --git a/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_HOMO.yaml b/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_HOMO.yaml index 3844c37e..2dd18c72 100644 --- a/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_HOMO.yaml +++ b/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_HOMO.yaml @@ -2,7 +2,7 @@ OS: posix_linux backend: tensorflow cuda_available: 'True' data_unit: '[''eV'']' -date_time: '2024-03-23 17:32:45' +date_time: '2024-04-04 07:51:33' device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'', device_type=''GPU'')]' device_memory: '[]' @@ -10,55 +10,106 @@ device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V10 epochs: - 800 - 800 +- 800 +- 800 +- 800 execute_folds: -- 3 +- 0 kgcnn_version: 4.0.1 learning_rate: - 1.927654702527093e-09 - 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 loss: +- 0.002737970557063818 +- 0.002600967651233077 +- 0.002760219154879451 - 0.0028957126196473837 - 0.002597656799480319 max_learning_rate: - 0.0005000000237487257 - 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 max_loss: +- 6.36514139175415 +- 6.3354082107543945 +- 3.2120323181152344 - 3.657465934753418 - 3.212603807449341 max_scaled_mean_absolute_error: +- 3.8430910110473633 +- 3.817063093185425 +- 1.9303600788116455 - 2.2045230865478516 - 1.9343459606170654 max_scaled_root_mean_squared_error: +- 91.58289337158203 +- 73.41930389404297 +- 45.41462707519531 - 50.40476989746094 - 41.87397003173828 max_val_loss: +- 0.6599817872047424 +- 0.5521999597549438 +- 0.6278929114341736 - 0.5486903190612793 - 0.5359625816345215 max_val_scaled_mean_absolute_error: +- 0.39830490946769714 +- 0.33254238963127136 +- 0.3772137463092804 - 0.33057093620300293 - 0.3225601315498352 max_val_scaled_root_mean_squared_error: +- 0.4824778139591217 +- 0.45746099948883057 +- 0.486127644777298 - 0.452626496553421 - 0.44354674220085144 min_learning_rate: - 1.927654702527093e-09 - 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 min_loss: +- 0.0025574713945388794 +- 0.0023802428040653467 +- 0.0025175956543534994 - 0.0025854564737528563 - 0.0024895048700273037 min_scaled_mean_absolute_error: +- 0.001543749705888331 +- 0.0014338576002046466 +- 0.0015127232763916254 - 0.0015581249026581645 - 0.0014986685710027814 min_scaled_root_mean_squared_error: +- 0.0023054105695337057 +- 0.0021260655485093594 +- 0.002259395783767104 - 0.00232700165361166 - 0.0021728877909481525 min_val_loss: +- 0.05351784825325012 +- 0.05056117847561836 +- 0.05054227635264397 - 0.050690535455942154 - 0.04970072954893112 min_val_scaled_mean_absolute_error: +- 0.03227536380290985 +- 0.030437694862484932 +- 0.03034353442490101 - 0.030524469912052155 - 0.029908040538430214 min_val_scaled_root_mean_squared_error: +- 0.06082373484969139 +- 0.05166103318333626 +- 0.05559256672859192 - 0.05220561474561691 - 0.05357697978615761 model_class: make_model @@ -66,23 +117,41 @@ model_name: EGNN model_version: '2023-12-04' multi_target_indices: - 5 -number_histories: 2 +number_histories: 5 scaled_mean_absolute_error: +- 0.0016526500694453716 +- 0.0015668243868276477 +- 0.0016584444092586637 - 0.0017449988517910242 - 0.0015636744210496545 scaled_root_mean_squared_error: +- 0.002306584268808365 +- 0.0021525220945477486 +- 0.0023082816042006016 - 0.002412783447653055 - 0.0021728877909481525 seed: 42 time_list: +- '9:05:28.705985' +- '9:03:43.029214' +- '9:01:12.607599' - '9:22:51.255236' - '9:03:34.039838' val_loss: +- 0.05351785570383072 +- 0.05056117847561836 +- 0.05054479464888573 - 0.05069223418831825 - 0.04970075935125351 val_scaled_mean_absolute_error: +- 0.032275374978780746 +- 0.030437694862484932 +- 0.030345039442181587 - 0.030525507405400276 - 0.029908040538430214 val_scaled_root_mean_squared_error: +- 0.06082373857498169 +- 0.051684316247701645 +- 0.05568939074873924 - 0.05222340673208237 - 0.05357697978615761 diff --git a/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_U0.yaml b/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_U0.yaml new file mode 100644 index 00000000..8620a3db --- /dev/null +++ b/training/results/QM9Dataset/EGNN/EGNN_QM9Dataset_score_U0.yaml @@ -0,0 +1,157 @@ +OS: posix_linux +backend: tensorflow +cuda_available: 'True' +data_unit: '[''eV'']' +date_time: '2024-04-26 11:49:59' +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: +- 800 +- 800 +- 800 +- 800 +- 800 +execute_folds: +- 4 +kgcnn_version: 4.0.1 +learning_rate: +- 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 +loss: +- 0.0012311397586017847 +- 0.0010912681464105844 +- 0.001204933738335967 +- 0.001095940126106143 +- 0.0011091696796938777 +max_learning_rate: +- 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 +max_loss: +- 6.500751495361328 +- 4.947936058044434 +- 3.295323371887207 +- 4.41763162612915 +- 3.3878307342529297 +max_scaled_mean_absolute_error: +- 7.102834701538086 +- 5.415079593658447 +- 3.6051416397094727 +- 4.835123538970947 +- 3.7104332447052 +max_scaled_root_mean_squared_error: +- 161.5733184814453 +- 99.65192413330078 +- 82.59259033203125 +- 96.01168823242188 +- 76.02869415283203 +max_val_loss: +- 0.7016094326972961 +- 0.6603164076805115 +- 0.6916307806968689 +- 0.7157029509544373 +- 0.7800574898719788 +max_val_scaled_mean_absolute_error: +- 0.7662613391876221 +- 0.7223361730575562 +- 0.7563048601150513 +- 0.7830995321273804 +- 0.8539851307868958 +max_val_scaled_root_mean_squared_error: +- 0.9794498085975647 +- 0.9319915771484375 +- 0.9697304368019104 +- 0.9709248542785645 +- 1.0801883935928345 +min_learning_rate: +- 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 +- 1.927654702527093e-09 +min_loss: +- 0.0012311397586017847 +- 0.0010912681464105844 +- 0.001204933738335967 +- 0.001095940126106143 +- 0.0011091696796938777 +min_scaled_mean_absolute_error: +- 0.001344874850474298 +- 0.0011941180564463139 +- 0.001318076392635703 +- 0.0011992636136710644 +- 0.0012146378867328167 +min_scaled_root_mean_squared_error: +- 0.0018997651059180498 +- 0.0029807600658386946 +- 0.0018503061728551984 +- 0.0016830191016197205 +- 0.001691434532403946 +min_val_loss: +- 0.009048271924257278 +- 0.008530893363058567 +- 0.008978086523711681 +- 0.00858803279697895 +- 0.008887539617717266 +min_val_scaled_mean_absolute_error: +- 0.009855221025645733 +- 0.009315433911979198 +- 0.009793723002076149 +- 0.009375786408782005 +- 0.009721021167933941 +min_val_scaled_root_mean_squared_error: +- 0.03671259060502052 +- 0.026049809530377388 +- 0.035086072981357574 +- 0.033445317298173904 +- 0.03285631537437439 +model_class: make_model +model_name: EGNN +model_version: '2023-12-04' +multi_target_indices: +- 10 +number_histories: 5 +scaled_mean_absolute_error: +- 0.001344874850474298 +- 0.0011941180564463139 +- 0.001318076392635703 +- 0.0011992636136710644 +- 0.0012146378867328167 +scaled_root_mean_squared_error: +- 0.0018997651059180498 +- 0.0029807600658386946 +- 0.0018503061728551984 +- 0.0016830191016197205 +- 0.001691434532403946 +seed: 42 +time_list: +- '9:13:04.203417' +- '8:55:52.139003' +- '9:13:45.449592' +- '8:52:20.830477' +- '8:57:48.913931' +val_loss: +- 0.009048271924257278 +- 0.008530893363058567 +- 0.008978086523711681 +- 0.00858811754733324 +- 0.00888855755329132 +val_scaled_mean_absolute_error: +- 0.009855221025645733 +- 0.009315433911979198 +- 0.009793723002076149 +- 0.009375888854265213 +- 0.009722139686346054 +val_scaled_root_mean_squared_error: +- 0.036932267248630524 +- 0.026062658056616783 +- 0.03688986599445343 +- 0.03466796875 +- 0.032981641590595245 diff --git a/training/results/QM9Dataset/EGNN/EGNN_hyper_U0.json b/training/results/QM9Dataset/EGNN/EGNN_hyper_U0.json new file mode 100644 index 00000000..682a3803 --- /dev/null +++ b/training/results/QM9Dataset/EGNN/EGNN_hyper_U0.json @@ -0,0 +1 @@ +{"model": {"class_name": "make_model", "module_name": "kgcnn.literature.EGNN", "config": {"name": "EGNN", "inputs": [{"shape": [null, 15], "name": "node_attributes", "dtype": "float32", "ragged": true}, {"shape": [null, 3], "name": "node_coordinates", "dtype": "float32", "ragged": true}, {"shape": [null, 1], "name": "range_attributes", "dtype": "float32", "ragged": true}, {"shape": [null, 2], "name": "range_indices", "dtype": "int64", "ragged": true}], "input_tensor_type": "ragged", "input_embedding": null, "input_node_embedding": {"input_dim": 95, "output_dim": 128}, "input_edge_embedding": {"input_dim": 95, "output_dim": 128}, "depth": 7, "node_mlp_initialize": {"units": 128, "activation": "linear"}, "euclidean_norm_kwargs": {"keepdims": true, "axis": 1, "square_norm": true}, "use_edge_attributes": false, "edge_mlp_kwargs": {"units": [128, 128], "activation": ["swish", "swish"]}, "edge_attention_kwargs": {"units": 1, "activation": "sigmoid"}, "use_normalized_difference": false, "expand_distance_kwargs": null, "coord_mlp_kwargs": null, "pooling_coord_kwargs": null, "pooling_edge_kwargs": {"pooling_method": "sum"}, "node_normalize_kwargs": null, "use_node_attributes": false, "node_mlp_kwargs": {"units": [128, 128], "activation": ["swish", "linear"]}, "use_skip": true, "verbose": 10, "node_decoder_kwargs": {"units": [128, 128], "activation": ["swish", "linear"]}, "node_pooling_kwargs": {"pooling_method": "sum"}, "output_embedding": "graph", "output_to_tensor": true, "output_mlp": {"use_bias": [true, true], "units": [128, 1], "activation": ["swish", "linear"]}}}, "training": {"cross_validation": {"class_name": "KFold", "config": {"n_splits": 5, "random_state": 42, "shuffle": true}}, "fit": {"batch_size": 96, "epochs": 800, "validation_freq": 1, "verbose": 2, "callbacks": [{"class_name": "kgcnn>CosineAnnealingLRScheduler", "config": {"lr_start": 0.0005, "lr_min": 0.0, "epoch_max": 800, "verbose": 1}}]}, "compile": {"optimizer": {"class_name": "Adam", "config": {"learning_rate": 0.0005}}, "loss": "mean_absolute_error"}, "scaler": {"class_name": "QMGraphLabelScaler", "config": {"atomic_number": "node_number", "scaler": [{"class_name": "ExtensiveMolecularLabelScaler", "config": {}}]}}, "multi_target_indices": [10]}, "data": {}, "info": {"postfix": "", "postfix_file": "_U0", "kgcnn_version": "4.0.0"}, "dataset": {"class_name": "QM9Dataset", "module_name": "kgcnn.data.datasets.QM9Dataset", "config": {}, "methods": [{"map_list": {"method": "atomic_charge_representation"}}, {"map_list": {"method": "set_range", "max_distance": 10, "max_neighbours": 10000}}]}} \ No newline at end of file diff --git a/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_H.yaml b/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_H.yaml index f620f5a2..1bc34422 100644 --- a/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_H.yaml +++ b/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_H.yaml @@ -2,7 +2,7 @@ OS: posix_linux backend: tensorflow cuda_available: 'True' data_unit: '[''eV'']' -date_time: '2024-03-24 14:07:25' +date_time: '2024-04-05 15:52:45' device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'', device_type=''GPU'')]' device_memory: '[]' @@ -10,85 +10,166 @@ device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V10 epochs: - 900 - 900 +- 900 +- 900 +- 900 execute_folds: -- 3 +- 0 kgcnn_version: 4.0.1 learning_rate: - 3.1866063636698527e-06 - 3.1866063636698527e-06 +- 3.1866063636698527e-06 +- 3.1866063636698527e-06 +- 3.1866063636698527e-06 loss: +- 0.002681844402104616 +- 0.0032064171973615885 +- 0.0025959820486605167 - 0.002213386818766594 - 0.0026501708198338747 max_learning_rate: - 9.999999747378752e-05 - 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 max_loss: +- 1554695.0 +- 1553981.375 +- 1556007.25 - 1554053.125 - 1555721.75 max_mean_absolute_error: +- 1554695.75 +- 1553994.875 +- 1555946.5 - 1554094.625 - 1555668.125 max_mean_squared_error: +- 3148369887232.0 +- 3145550266368.0 +- 3146034970624.0 - 3137068072960.0 - 3148559941632.0 max_scaled_mean_absolute_error: +- 1554695.75 +- 1553994.875 +- 1555946.5 - 1554094.625 - 1555668.125 max_scaled_root_mean_squared_error: +- 1774364.625 +- 1773569.875 +- 1773706.5 - 1771177.0 - 1774418.25 max_val_loss: +- 0.1499040722846985 +- 0.1765926629304886 +- 0.43149346113204956 - 0.2166866958141327 - 0.09389010071754456 max_val_mean_absolute_error: +- 0.14993922412395477 +- 0.1766568273305893 +- 0.4307798147201538 - 0.21607866883277893 - 0.09375178813934326 max_val_mean_squared_error: +- 0.03740379214286804 +- 0.038885850459337234 +- 0.2188498079776764 - 0.056355491280555725 - 1.902894377708435 max_val_scaled_mean_absolute_error: +- 0.14993922412395477 +- 0.1766568273305893 +- 0.4307798147201538 - 0.21607866883277893 - 0.09375178813934326 max_val_scaled_root_mean_squared_error: +- 0.19340060651302338 +- 0.19719496369361877 +- 0.4678138494491577 - 0.2373931109905243 - 1.3794543743133545 mean_absolute_error: +- 0.002682383405044675 +- 0.0032065522391349077 +- 0.0025931824930012226 - 0.0022136522457003593 - 0.0026506620924919844 mean_squared_error: +- 1.280460855923593e-05 +- 1.831956979003735e-05 +- 1.2003599294985179e-05 - 9.020595825859345e-06 - 1.2643344234675169e-05 min_learning_rate: - 0.0 - 0.0 +- 0.0 +- 0.0 +- 0.0 min_loss: +- 0.002328281057998538 +- 0.0024495236575603485 +- 0.0023000878281891346 - 0.002213386818766594 - 0.0022141484078019857 min_mean_absolute_error: +- 0.002328545320779085 +- 0.002449710387736559 +- 0.0023000547662377357 - 0.0022136522457003593 - 0.002214109757915139 min_mean_squared_error: +- 9.832401701714844e-06 +- 1.0698954611143563e-05 +- 9.578599929227494e-06 - 9.020595825859345e-06 - 8.868021723174024e-06 min_scaled_mean_absolute_error: +- 0.002328545320779085 +- 0.002449710387736559 +- 0.0023000547662377357 - 0.0022136522457003593 - 0.002214109757915139 min_scaled_root_mean_squared_error: +- 0.0031356660183519125 +- 0.003270925721153617 +- 0.0030949313659220934 - 0.003003430785611272 - 0.002977922325953841 min_val_loss: +- 0.007864161394536495 +- 0.007485698908567429 +- 0.007987338118255138 - 0.00743128964677453 - 0.016484372317790985 min_val_mean_absolute_error: +- 0.007765881717205048 +- 0.00744463037699461 +- 0.0077869100496172905 - 0.0074112010188400745 - 0.01621674932539463 min_val_mean_squared_error: +- 0.0005204707267694175 +- 0.0003245097759645432 +- 0.0008214145782403648 - 0.0003903789329342544 - 0.011060231365263462 min_val_scaled_mean_absolute_error: +- 0.007765881717205048 +- 0.00744463037699461 +- 0.0077869100496172905 - 0.0074112010188400745 - 0.01621674932539463 min_val_scaled_root_mean_squared_error: +- 0.0228138267993927 +- 0.018014153465628624 +- 0.028660330921411514 - 0.019758008420467377 - 0.10516763478517532 model_class: make_model @@ -96,29 +177,53 @@ model_name: MXMNet model_version: '2023-12-09' multi_target_indices: - 17 -number_histories: 2 +number_histories: 5 scaled_mean_absolute_error: +- 0.002682383405044675 +- 0.0032065522391349077 +- 0.0025931824930012226 - 0.0022136522457003593 - 0.0026506620924919844 scaled_root_mean_squared_error: +- 0.003578352741897106 +- 0.0042801364324986935 +- 0.003464621026068926 - 0.003003430785611272 - 0.003555747913196683 seed: 42 time_list: +- 1 day, 4:09:35.261125 +- 1 day, 4:29:49.657195 +- 1 day, 4:43:42.024086 - 1 day, 4:46:25.801978 - 1 day, 4:29:22.767528 val_loss: +- 0.007864161394536495 +- 0.007485698908567429 +- 0.008842719718813896 - 0.00743128964677453 - 0.016571108251810074 val_mean_absolute_error: +- 0.007765881717205048 +- 0.00744463037699461 +- 0.008645348250865936 - 0.0074112010188400745 - 0.016321230679750443 val_mean_squared_error: +- 0.0005204707267694175 +- 0.0003245097759645432 +- 0.0014210563385859132 - 0.0003907398786395788 - 1.902894377708435 val_scaled_mean_absolute_error: +- 0.007765881717205048 +- 0.00744463037699461 +- 0.008645348250865936 - 0.0074112010188400745 - 0.016321230679750443 val_scaled_root_mean_squared_error: +- 0.0228138267993927 +- 0.018014153465628624 +- 0.037696901708841324 - 0.019767140969634056 - 1.3794543743133545 diff --git a/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_HOMO.yaml b/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_HOMO.yaml index 5eab1a04..7a194fdb 100644 --- a/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_HOMO.yaml +++ b/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_HOMO.yaml @@ -2,7 +2,7 @@ OS: posix_linux backend: tensorflow cuda_available: 'True' data_unit: '[''eV'']' -date_time: '2024-03-24 13:11:12' +date_time: '2024-04-05 12:53:24' device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'', device_type=''GPU'')]' device_memory: '[]' @@ -10,85 +10,166 @@ device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V10 epochs: - 900 - 900 +- 900 +- 900 +- 900 execute_folds: -- 3 +- 0 kgcnn_version: 4.0.1 learning_rate: - 3.186606409144588e-05 - 3.186606409144588e-05 +- 3.186606409144588e-05 +- 3.186606409144588e-05 +- 3.186606409144588e-05 loss: +- 0.012575064785778522 +- 0.007334146182984114 +- 0.0039936513639986515 - 0.01078067161142826 - 0.00478322384878993 max_learning_rate: - 0.0010000000474974513 - 0.0010000000474974513 +- 0.0010000000474974513 +- 0.0010000000474974513 +- 0.0010000000474974513 max_loss: +- 3605502.5 +- 3605368.5 +- 3608746.0 - 3608025.25 - 3608167.0 max_mean_absolute_error: +- 3605553.0 +- 3605363.0 +- 3608634.75 - 3608055.0 - 3608097.75 max_mean_squared_error: +- 15796611842048.0 +- 15971303555072.0 +- 15891316080640.0 - 16038961872896.0 - 16010225647616.0 max_scaled_mean_absolute_error: +- 3605553.0 +- 3605363.0 +- 3608634.75 - 3608055.0 - 3608097.75 max_scaled_root_mean_squared_error: +- 3974495.25 +- 3996411.25 +- 3986391.25 - 4004867.25 - 4001278.0 max_val_loss: +- 0.13663585484027863 +- 0.11844862997531891 +- 0.2528819441795349 - 0.11132214218378067 - 0.24343882501125336 max_val_mean_absolute_error: +- 0.13590598106384277 +- 0.11841301620006561 +- 0.25335535407066345 - 0.11099009960889816 - 0.24326354265213013 max_val_mean_squared_error: +- 3.3152921199798584 +- 0.021277446299791336 +- 676.0514526367188 - 0.020909026265144348 - 1.187813639640808 max_val_scaled_mean_absolute_error: +- 0.13590598106384277 +- 0.11841301620006561 +- 0.25335535407066345 - 0.11099009960889816 - 0.24326354265213013 max_val_scaled_root_mean_squared_error: +- 1.8207943439483643 +- 0.14586789906024933 +- 26.00098991394043 - 0.1445995420217514 - 1.0898686647415161 mean_absolute_error: +- 0.012575104832649231 +- 0.007334934081882238 +- 0.0039945305325090885 - 0.010778913274407387 - 0.004784419666975737 mean_squared_error: +- 0.000245674978941679 +- 8.993221854325384e-05 +- 3.07419613818638e-05 - 0.0001861106138676405 - 4.498466296354309e-05 min_learning_rate: - 0.0 - 0.0 +- 0.0 +- 0.0 +- 0.0 min_loss: +- 0.00655166432261467 +- 0.004847946111112833 +- 0.0037461391184479 - 0.008195515722036362 - 0.003713286481797695 min_mean_absolute_error: +- 0.00655211228877306 +- 0.004848274402320385 +- 0.003746886271983385 - 0.008196063339710236 - 0.00371303828433156 min_mean_squared_error: +- 7.887562969699502e-05 +- 4.32189044659026e-05 +- 2.6835808967007324e-05 - 0.0001453463191865012 - 2.6165618692175485e-05 min_scaled_mean_absolute_error: +- 0.00655211228877306 +- 0.004848274402320385 +- 0.003746886271983385 - 0.008196063339710236 - 0.00371303828433156 min_scaled_root_mean_squared_error: +- 0.008881195448338985 +- 0.0065741087310016155 +- 0.005180329084396362 - 0.012055966071784496 - 0.005115233827382326 min_val_loss: +- 0.031404219567775726 +- 0.029028331860899925 +- 0.0336117148399353 - 0.031849466264247894 - 0.031604230403900146 min_val_mean_absolute_error: +- 0.03133499249815941 +- 0.02896134927868843 +- 0.03353891521692276 - 0.03171669691801071 - 0.031424228101968765 min_val_mean_squared_error: +- 0.002995939925312996 +- 0.002373416442424059 +- 0.0026221447624266148 - 0.0022724990267306566 - 0.00259683420881629 min_val_scaled_mean_absolute_error: +- 0.03133499249815941 +- 0.02896134927868843 +- 0.03353891521692276 - 0.03171669691801071 - 0.031424228101968765 min_val_scaled_root_mean_squared_error: +- 0.054735179990530014 +- 0.04871772229671478 +- 0.05120687931776047 - 0.04767073318362236 - 0.05095914006233215 model_class: make_model @@ -96,29 +177,53 @@ model_name: MXMNet model_version: '2023-12-09' multi_target_indices: - 5 -number_histories: 2 +number_histories: 5 scaled_mean_absolute_error: +- 0.012575104832649231 +- 0.007334934081882238 +- 0.0039945305325090885 - 0.010778913274407387 - 0.004784419666975737 scaled_root_mean_squared_error: +- 0.01567402109503746 +- 0.009483260102570057 +- 0.005544543266296387 - 0.013642235659062862 - 0.006707060616463423 seed: 42 time_list: +- 1 day, 4:53:34.726124 +- 1 day, 5:05:14.178755 +- 1 day, 5:05:16.826940 - 1 day, 4:56:14.212757 - 1 day, 4:47:31.396423 val_loss: +- 0.036044564098119736 +- 0.032148946076631546 +- 0.0336117148399353 - 0.036266177892684937 - 0.03359147906303406 val_mean_absolute_error: +- 0.03597966954112053 +- 0.032088425010442734 +- 0.03353891521692276 - 0.036107782274484634 - 0.03340762108564377 val_mean_squared_error: +- 0.0035048355348408222 +- 0.002828829223290086 +- 0.002737524686381221 - 0.0028413906693458557 - 0.002803423907607794 val_scaled_mean_absolute_error: +- 0.03597966954112053 +- 0.032088425010442734 +- 0.03353891521692276 - 0.036107782274484634 - 0.03340762108564377 val_scaled_root_mean_squared_error: +- 0.05920165032148361 +- 0.05318674072623253 +- 0.052321359515190125 - 0.05330469459295273 - 0.05294736847281456 diff --git a/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_U0.yaml b/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_U0.yaml new file mode 100644 index 00000000..9822d67f --- /dev/null +++ b/training/results/QM9Dataset/MXMNet/MXMNet_QM9Dataset_score_U0.yaml @@ -0,0 +1,229 @@ +OS: posix_linux +backend: tensorflow +cuda_available: 'True' +data_unit: '[''eV'']' +date_time: '2024-04-27 06:53:22' +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: +- 900 +- 900 +- 900 +- 900 +- 900 +execute_folds: +- 4 +kgcnn_version: 4.0.1 +learning_rate: +- 3.1866063636698527e-06 +- 3.1866063636698527e-06 +- 3.1866063636698527e-06 +- 3.1866063636698527e-06 +- 3.1866063636698527e-06 +loss: +- 0.0036107641644775867 +- 0.00348224607296288 +- 0.0023768290411680937 +- 0.003605922218412161 +- 0.0032897398341447115 +max_learning_rate: +- 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 +max_loss: +- 1554696.0 +- 1553981.75 +- 1556008.375 +- 1554054.125 +- 1555722.75 +max_mean_absolute_error: +- 1554696.625 +- 1553995.25 +- 1555947.5 +- 1554095.5 +- 1555669.25 +max_mean_squared_error: +- 3148371460096.0 +- 3145553936384.0 +- 3146037854208.0 +- 3137070956544.0 +- 3148562563072.0 +max_scaled_mean_absolute_error: +- 1554696.625 +- 1553995.25 +- 1555947.5 +- 1554095.5 +- 1555669.25 +max_scaled_root_mean_squared_error: +- 1774365.125 +- 1773571.0 +- 1773707.375 +- 1771177.875 +- 1774418.875 +max_val_loss: +- 0.2533656358718872 +- 0.27102017402648926 +- 0.2515379786491394 +- 0.22129854559898376 +- 0.08242256194353104 +max_val_mean_absolute_error: +- 0.25325536727905273 +- 0.2710208296775818 +- 0.251488596200943 +- 0.22158773243427277 +- 0.08239153772592545 +max_val_mean_squared_error: +- 0.08039408177137375 +- 0.08280055224895477 +- 0.07148066908121109 +- 0.05774109065532684 +- 4.138404369354248 +max_val_scaled_mean_absolute_error: +- 0.25325536727905273 +- 0.2710208296775818 +- 0.251488596200943 +- 0.22158773243427277 +- 0.08239153772592545 +max_val_scaled_root_mean_squared_error: +- 0.28353849053382874 +- 0.28775084018707275 +- 0.2673586905002594 +- 0.24029375612735748 +- 2.034306764602661 +mean_absolute_error: +- 0.0036107918713241816 +- 0.0034829918295145035 +- 0.0023774299770593643 +- 0.003605607198551297 +- 0.0032900075893849134 +mean_squared_error: +- 2.1307201677700505e-05 +- 2.0705885617644526e-05 +- 1.0227241546090227e-05 +- 2.22941416723188e-05 +- 1.802921724447515e-05 +min_learning_rate: +- 0.0 +- 0.0 +- 0.0 +- 0.0 +- 0.0 +min_loss: +- 0.0025323976296931505 +- 0.00238243886269629 +- 0.0023768290411680937 +- 0.0024545544292777777 +- 0.00229792227037251 +min_mean_absolute_error: +- 0.0025328765623271465 +- 0.00238246563822031 +- 0.0023774299770593643 +- 0.0024545362684875727 +- 0.0022984619718044996 +min_mean_squared_error: +- 1.1545389497769065e-05 +- 1.0232446584268473e-05 +- 1.0227241546090227e-05 +- 1.0932703844446223e-05 +- 9.613416295906063e-06 +min_scaled_mean_absolute_error: +- 0.0025328765623271465 +- 0.00238246563822031 +- 0.0023774299770593643 +- 0.0024545362684875727 +- 0.0022984619718044996 +min_scaled_root_mean_squared_error: +- 0.0033978507854044437 +- 0.0031988194677978754 +- 0.003198005724698305 +- 0.003306464059278369 +- 0.003100550966337323 +min_val_loss: +- 0.00812331773340702 +- 0.007428425829857588 +- 0.007924382574856281 +- 0.007561167702078819 +- 0.01865706406533718 +min_val_mean_absolute_error: +- 0.008039351552724838 +- 0.007383247837424278 +- 0.007725783158093691 +- 0.007546260952949524 +- 0.018481571227312088 +min_val_mean_squared_error: +- 0.00047742994502186775 +- 0.0003663378010969609 +- 0.000760754628572613 +- 0.0003624066594056785 +- 0.008477309718728065 +min_val_scaled_mean_absolute_error: +- 0.008039351552724838 +- 0.007383247837424278 +- 0.007725783158093691 +- 0.007546260952949524 +- 0.018481571227312088 +min_val_scaled_root_mean_squared_error: +- 0.02185017056763172 +- 0.019139952957630157 +- 0.027581781148910522 +- 0.019036980345845222 +- 0.09207230806350708 +model_class: make_model +model_name: MXMNet +model_version: '2023-12-09' +multi_target_indices: +- 15 +number_histories: 5 +scaled_mean_absolute_error: +- 0.0036107918713241816 +- 0.0034829918295145035 +- 0.0023774299770593643 +- 0.003605607198551297 +- 0.0032900075893849134 +scaled_root_mean_squared_error: +- 0.00461597228422761 +- 0.004550372250378132 +- 0.003198005724698305 +- 0.004721667151898146 +- 0.0042460826225578785 +seed: 42 +time_list: +- 1 day, 4:07:30.718964 +- 1 day, 2:17:06.929488 +- 1 day, 3:59:19.817712 +- 1 day, 5:03:11.501750 +- 1 day, 3:55:13.703010 +val_loss: +- 0.00812331773340702 +- 0.008302227593958378 +- 0.008040172979235649 +- 0.007561167702078819 +- 0.02075917460024357 +val_mean_absolute_error: +- 0.008039351552724838 +- 0.008259082213044167 +- 0.007835005410015583 +- 0.007546260952949524 +- 0.020614054054021835 +val_mean_squared_error: +- 0.00047742994502186775 +- 0.0003844579041469842 +- 0.0008069894975051284 +- 0.0003624066594056785 +- 4.138404369354248 +val_scaled_mean_absolute_error: +- 0.008039351552724838 +- 0.008259082213044167 +- 0.007835005410015583 +- 0.007546260952949524 +- 0.020614054054021835 +val_scaled_root_mean_squared_error: +- 0.02185017056763172 +- 0.019607597962021828 +- 0.028407560661435127 +- 0.019036980345845222 +- 2.034306764602661 diff --git a/training/results/QM9Dataset/MXMNet/MXMNet_hyper_U0.json b/training/results/QM9Dataset/MXMNet/MXMNet_hyper_U0.json new file mode 100644 index 00000000..228ba5d2 --- /dev/null +++ b/training/results/QM9Dataset/MXMNet/MXMNet_hyper_U0.json @@ -0,0 +1 @@ +{"model": {"class_name": "make_model", "module_name": "kgcnn.literature.MXMNet", "config": {"name": "MXMNet", "inputs": [{"shape": [null], "name": "node_number", "dtype": "int64", "ragged": true}, {"shape": [null, 3], "name": "node_coordinates", "dtype": "float32", "ragged": true}, {"shape": [null, 1], "name": "edge_weights", "dtype": "float32", "ragged": true}, {"shape": [null, 2], "name": "edge_indices", "dtype": "int64", "ragged": true}, {"shape": [null, 2], "name": "range_indices", "dtype": "int64", "ragged": true}, {"shape": [null, 2], "name": "angle_indices_1", "dtype": "int64", "ragged": true}, {"shape": [null, 2], "name": "angle_indices_2", "dtype": "int64", "ragged": true}], "input_tensor_type": "ragged", "input_embedding": null, "input_node_embedding": {"input_dim": 95, "output_dim": 128, "embeddings_initializer": {"class_name": "RandomUniform", "config": {"minval": -1.7320508075688772, "maxval": 1.7320508075688772}}}, "input_edge_embedding": {"input_dim": 32, "output_dim": 128}, "bessel_basis_local": {"num_radial": 16, "cutoff": 5.0, "envelope_exponent": 5}, "bessel_basis_global": {"num_radial": 16, "cutoff": 5.0, "envelope_exponent": 5}, "spherical_basis_local": {"num_spherical": 7, "num_radial": 6, "cutoff": 5.0, "envelope_exponent": 5}, "mlp_rbf_kwargs": {"units": 128, "activation": "swish"}, "mlp_sbf_kwargs": {"units": 128, "activation": "swish"}, "global_mp_kwargs": {"units": 128}, "local_mp_kwargs": {"units": 128, "output_units": 1, "output_kernel_initializer": "glorot_uniform"}, "use_edge_attributes": false, "depth": 6, "verbose": 10, "node_pooling_args": {"pooling_method": "sum"}, "output_embedding": "graph", "output_to_tensor": true, "use_output_mlp": false, "output_mlp": {"use_bias": [true], "units": [1], "activation": ["linear"]}}}, "training": {"cross_validation": {"class_name": "KFold", "config": {"n_splits": 5, "random_state": 42, "shuffle": true}}, "fit": {"batch_size": 128, "epochs": 900, "validation_freq": 10, "verbose": 2, "callbacks": [{"class_name": "kgcnn>LinearWarmupExponentialLRScheduler", "config": {"lr_start": 0.0001, "gamma": 0.9961697, "epo_warmup": 1, "verbose": 1}}]}, "compile": {"optimizer": {"class_name": "Adam", "config": {"learning_rate": 0.0001, "global_clipnorm": 1000}}, "loss": "mean_absolute_error", "metrics": ["mean_absolute_error", "mean_squared_error", {"class_name": "RootMeanSquaredError", "config": {"name": "scaled_root_mean_squared_error"}}, {"class_name": "MeanAbsoluteError", "config": {"name": "scaled_mean_absolute_error"}}]}, "multi_target_indices": [15]}, "data": {}, "info": {"postfix": "", "postfix_file": "_U0", "kgcnn_version": "4.0.0"}, "dataset": {"class_name": "QM9Dataset", "module_name": "kgcnn.data.datasets.QM9Dataset", "config": {}, "methods": [{"remove_uncharacterized": {}}, {"map_list": {"method": "set_edge_weights_uniform"}}, {"map_list": {"method": "set_range", "max_distance": 5, "max_neighbours": 1000}}, {"map_list": {"method": "set_angle", "range_indices": "edge_indices", "edge_pairing": "jk", "angle_indices": "angle_indices_1", "angle_indices_nodes": "angle_indices_nodes_1", "angle_attributes": "angle_attributes_1"}}, {"map_list": {"method": "set_angle", "range_indices": "edge_indices", "edge_pairing": "ik", "allow_self_edges": true, "angle_indices": "angle_indices_2", "angle_indices_nodes": "angle_indices_nodes_2", "angle_attributes": "angle_attributes_2"}}]}} \ No newline at end of file diff --git a/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_H.yaml b/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_H.yaml index f7a32500..b7fa81c8 100644 --- a/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_H.yaml +++ b/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_H.yaml @@ -2,7 +2,7 @@ OS: posix_linux backend: tensorflow cuda_available: 'True' data_unit: '[''eV'']' -date_time: '2024-03-24 21:09:56' +date_time: '2024-04-08 09:24:29' device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'', device_type=''GPU'')]' device_memory: '[]' @@ -10,55 +10,106 @@ device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V10 epochs: - 800 - 800 +- 800 +- 800 +- 800 execute_folds: -- 3 +- 0 kgcnn_version: 4.0.1 learning_rate: - 1.06999996205559e-05 - 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 loss: +- 0.00449918769299984 +- 0.00480709969997406 +- 0.004801088944077492 - 0.004764853976666927 - 0.004947832319885492 max_learning_rate: - 0.0005000000237487257 - 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 max_loss: +- 0.7754255533218384 +- 0.7777360677719116 +- 0.7746722102165222 - 0.7773099541664124 - 0.7771309018135071 max_scaled_mean_absolute_error: +- 0.8451758027076721 +- 0.8490518927574158 +- 0.8454299569129944 - 0.8486399054527283 - 0.8490580916404724 max_scaled_root_mean_squared_error: +- 1.0735920667648315 +- 1.0769978761672974 +- 1.0732311010360718 - 1.0785257816314697 - 1.077731966972351 max_val_loss: +- 0.113815538585186 +- 0.0949530228972435 +- 0.10097459703683853 - 0.10990142077207565 - 0.11022701114416122 max_val_scaled_mean_absolute_error: +- 0.12372852116823196 +- 0.10345017164945602 +- 0.10990025103092194 - 0.11967755109071732 - 0.12017706781625748 max_val_scaled_root_mean_squared_error: +- 0.19564726948738098 +- 0.153788760304451 +- 0.17892277240753174 - 0.19413302838802338 - 0.2035401165485382 min_learning_rate: - 1.06999996205559e-05 - 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 min_loss: +- 0.00449918769299984 +- 0.00480709969997406 +- 0.004791752900928259 - 0.004764853976666927 - 0.004947832319885492 min_scaled_mean_absolute_error: +- 0.0049039628356695175 +- 0.005248055327683687 +- 0.005229531787335873 - 0.005201948340982199 - 0.005405828822404146 min_scaled_root_mean_squared_error: +- 0.0068879942409694195 +- 0.009614086709916592 +- 0.007560875732451677 - 0.007367674261331558 - 0.007595034781843424 min_val_loss: +- 0.01575741171836853 +- 0.014592286199331284 +- 0.015833746641874313 - 0.014941912144422531 - 0.015913689509034157 min_val_scaled_mean_absolute_error: +- 0.01704293303191662 +- 0.015818098559975624 +- 0.01713581755757332 - 0.016232075169682503 - 0.017248570919036865 min_val_scaled_root_mean_squared_error: +- 0.04706299304962158 +- 0.0392644889652729 +- 0.047808073461055756 - 0.037356771528720856 - 0.04759074002504349 model_class: make_model @@ -66,23 +117,41 @@ model_name: Megnet model_version: '2023-12-05' multi_target_indices: - 12 -number_histories: 2 +number_histories: 5 scaled_mean_absolute_error: +- 0.0049039628356695175 +- 0.005248055327683687 +- 0.005239762365818024 - 0.005201948340982199 - 0.005405828822404146 scaled_root_mean_squared_error: +- 0.0068879942409694195 +- 0.009614086709916592 +- 0.007560875732451677 - 0.007367674261331558 - 0.007595034781843424 seed: 42 time_list: +- 1 day, 1:10:40.905836 +- 1 day, 1:12:21.414727 +- 1 day, 2:00:33.539044 - 1 day, 3:27:14.314629 - 1 day, 3:02:53.012669 val_loss: +- 0.01575741171836853 +- 0.014592286199331284 +- 0.015861116349697113 - 0.015188874676823616 - 0.015913689509034157 val_scaled_mean_absolute_error: +- 0.01704293303191662 +- 0.015818098559975624 +- 0.01716645061969757 - 0.016505751758813858 - 0.017248570919036865 val_scaled_root_mean_squared_error: +- 0.04706299304962158 +- 0.03932986035943031 +- 0.04781997948884964 - 0.03748267516493797 - 0.051103245466947556 diff --git a/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_HOMO.yaml b/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_HOMO.yaml index 6c427f41..b957540a 100644 --- a/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_HOMO.yaml +++ b/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_HOMO.yaml @@ -2,7 +2,7 @@ OS: posix_linux backend: tensorflow cuda_available: 'True' data_unit: '[''eV'']' -date_time: '2024-03-24 08:48:20' +date_time: '2024-04-05 12:14:38' device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'', device_type=''GPU'')]' device_memory: '[]' @@ -10,55 +10,106 @@ device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V10 epochs: - 800 - 800 +- 800 +- 800 +- 800 execute_folds: -- 3 +- 0 kgcnn_version: 4.0.1 learning_rate: - 1.06999996205559e-05 - 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 loss: +- 0.016325801610946655 +- 0.015983140096068382 +- 0.0162623580545187 - 0.015953239053487778 - 0.016383972018957138 max_learning_rate: - 0.0005000000237487257 - 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 max_loss: +- 0.5663787722587585 +- 0.571689784526825 +- 0.5686174631118774 - 0.571775496006012 - 0.5746840238571167 max_scaled_mean_absolute_error: +- 0.3418545126914978 +- 0.344331294298172 +- 0.3416273295879364 - 0.34452834725379944 - 0.34592726826667786 max_scaled_root_mean_squared_error: +- 0.47598567605018616 +- 0.47961050271987915 +- 0.47551852464675903 - 0.48018452525138855 - 0.48139113187789917 max_val_loss: +- 0.2139911651611328 +- 0.19688135385513306 +- 0.1938907355070114 - 0.1906891018152237 - 0.20106293261051178 max_val_scaled_mean_absolute_error: +- 0.1289941966533661 +- 0.11848942935466766 +- 0.11635471135377884 - 0.11481630057096481 - 0.12096023559570312 max_val_scaled_root_mean_squared_error: +- 0.18448910117149353 +- 0.16506318747997284 +- 0.1690875142812729 - 0.16864857077598572 - 0.17896375060081482 min_learning_rate: - 1.06999996205559e-05 - 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 min_loss: +- 0.016224723309278488 +- 0.015833277255296707 +- 0.016133960336446762 - 0.015802957117557526 - 0.016289122402668 min_scaled_mean_absolute_error: +- 0.009792533703148365 +- 0.009536445140838623 +- 0.009693225845694542 - 0.009522140957415104 - 0.009805066511034966 min_scaled_root_mean_squared_error: +- 0.015724705532193184 +- 0.01566649042069912 +- 0.015286288224160671 - 0.014821133576333523 - 0.017847513779997826 min_val_loss: +- 0.08089471608400345 +- 0.07886076718568802 +- 0.08195757120847702 - 0.0758657306432724 - 0.07961893826723099 min_val_scaled_mean_absolute_error: +- 0.04872087389230728 +- 0.04744846746325493 +- 0.04914051666855812 - 0.0456601046025753 - 0.047898560762405396 min_val_scaled_root_mean_squared_error: +- 0.0768187940120697 +- 0.0734473317861557 +- 0.0770028606057167 - 0.06968275457620621 - 0.07626539468765259 model_class: make_model @@ -66,23 +117,41 @@ model_name: Megnet model_version: '2023-12-05' multi_target_indices: - 5 -number_histories: 2 +number_histories: 5 scaled_mean_absolute_error: +- 0.009853425435721874 +- 0.009626748040318489 +- 0.009770313277840614 - 0.009612714871764183 - 0.009862092323601246 scaled_root_mean_squared_error: +- 0.015761153772473335 +- 0.01571161113679409 +- 0.015349210239946842 - 0.014880934730172157 - 0.017847513779997826 seed: 42 time_list: +- 1 day, 1:26:57.340834 +- 1 day, 0:43:42.751471 +- 1 day, 0:28:35.469120 - 1 day, 0:39:02.240716 - 1 day, 3:19:02.522863 val_loss: +- 0.08089471608400345 +- 0.07918484508991241 +- 0.08195757120847702 - 0.0758657306432724 - 0.07961893826723099 val_scaled_mean_absolute_error: +- 0.04872087389230728 +- 0.04764195904135704 +- 0.04914051666855812 - 0.0456601046025753 - 0.047898560762405396 val_scaled_root_mean_squared_error: +- 0.07903549075126648 +- 0.07467254996299744 +- 0.0779530256986618 - 0.07025036215782166 - 0.07659842818975449 diff --git a/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_U0.yaml b/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_U0.yaml new file mode 100644 index 00000000..3104f284 --- /dev/null +++ b/training/results/QM9Dataset/Megnet/Megnet_QM9Dataset_score_U0.yaml @@ -0,0 +1,157 @@ +OS: posix_linux +backend: tensorflow +cuda_available: 'True' +data_unit: '[''eV'']' +date_time: '2024-04-28 12:42:09' +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: +- 800 +- 800 +- 800 +- 800 +- 800 +execute_folds: +- 4 +kgcnn_version: 4.0.1 +learning_rate: +- 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 +loss: +- 0.004634249955415726 +- 0.004807903431355953 +- 0.004945005290210247 +- 0.004850754048675299 +- 0.00491912430152297 +max_learning_rate: +- 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 +- 0.0005000000237487257 +max_loss: +- 0.7744609713554382 +- 0.7779005765914917 +- 0.7765160799026489 +- 0.7639061212539673 +- 0.7753151059150696 +max_scaled_mean_absolute_error: +- 0.8459039330482483 +- 0.851077675819397 +- 0.8492878079414368 +- 0.8358521461486816 +- 0.848880410194397 +max_scaled_root_mean_squared_error: +- 1.0753504037857056 +- 1.0803871154785156 +- 1.0795201063156128 +- 1.0663139820098877 +- 1.0791212320327759 +max_val_loss: +- 0.08474229276180267 +- 0.10189788043498993 +- 0.10797122865915298 +- 0.09738004207611084 +- 0.14591144025325775 +max_val_scaled_mean_absolute_error: +- 0.09228112548589706 +- 0.11122322827577591 +- 0.11768129467964172 +- 0.10624174028635025 +- 0.15947864949703217 +max_val_scaled_root_mean_squared_error: +- 0.14865222573280334 +- 0.1764373481273651 +- 0.19590848684310913 +- 0.18263152241706848 +- 0.2433769255876541 +min_learning_rate: +- 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 +- 1.06999996205559e-05 +min_loss: +- 0.004634249955415726 +- 0.004807903431355953 +- 0.004945005290210247 +- 0.004850754048675299 +- 0.00491912430152297 +min_scaled_mean_absolute_error: +- 0.00506167346611619 +- 0.005260199774056673 +- 0.005408469121903181 +- 0.005307466723024845 +- 0.005385955795645714 +min_scaled_root_mean_squared_error: +- 0.0071287816390395164 +- 0.0076753743924200535 +- 0.008121642284095287 +- 0.007745463866740465 +- 0.007510432042181492 +min_val_loss: +- 0.01565733551979065 +- 0.014823234640061855 +- 0.01604733243584633 +- 0.015279361046850681 +- 0.015920881181955338 +min_val_scaled_mean_absolute_error: +- 0.01696264185011387 +- 0.01611301302909851 +- 0.017375729978084564 +- 0.016619566828012466 +- 0.017303109169006348 +min_val_scaled_root_mean_squared_error: +- 0.04555272310972214 +- 0.04017811268568039 +- 0.04914117604494095 +- 0.0424349345266819 +- 0.05245567858219147 +model_class: make_model +model_name: Megnet +model_version: '2023-12-05' +multi_target_indices: +- 10 +number_histories: 5 +scaled_mean_absolute_error: +- 0.00506167346611619 +- 0.005260199774056673 +- 0.005408469121903181 +- 0.005307466723024845 +- 0.005385955795645714 +scaled_root_mean_squared_error: +- 0.0071287816390395164 +- 0.0076753743924200535 +- 0.008121642284095287 +- 0.007745463866740465 +- 0.007510432042181492 +seed: 42 +time_list: +- 1 day, 0:41:09.347128 +- 1 day, 1:38:26.593325 +- 1 day, 1:12:28.815031 +- 1 day, 1:57:32.792615 +- 1 day, 2:31:27.109517 +val_loss: +- 0.015873653814196587 +- 0.014855638146400452 +- 0.01604733243584633 +- 0.015279361046850681 +- 0.015959978103637695 +val_scaled_mean_absolute_error: +- 0.017198842018842697 +- 0.01614818349480629 +- 0.017375729978084564 +- 0.016619566828012466 +- 0.017347270622849464 +val_scaled_root_mean_squared_error: +- 0.04585716500878334 +- 0.04040386155247688 +- 0.04916248843073845 +- 0.042563583701848984 +- 0.05361202359199524 diff --git a/training/results/QM9Dataset/Megnet/Megnet_hyper_U0.json b/training/results/QM9Dataset/Megnet/Megnet_hyper_U0.json new file mode 100644 index 00000000..582b91f1 --- /dev/null +++ b/training/results/QM9Dataset/Megnet/Megnet_hyper_U0.json @@ -0,0 +1 @@ +{"model": {"class_name": "make_model", "module_name": "kgcnn.literature.Megnet", "config": {"name": "Megnet", "inputs": [{"shape": [null], "name": "node_number", "dtype": "int64", "ragged": true}, {"shape": [null, 3], "name": "node_coordinates", "dtype": "float32", "ragged": true}, {"shape": [null, 2], "name": "range_indices", "dtype": "int64", "ragged": true}, {"shape": [2], "name": "graph_attributes", "dtype": "float32", "ragged": false}], "input_tensor_type": "ragged", "input_embedding": null, "input_node_embedding": {"input_dim": 10, "output_dim": 16}, "gauss_args": {"bins": 20, "distance": 4, "offset": 0.0, "sigma": 0.4}, "meg_block_args": {"node_embed": [64, 32, 32], "edge_embed": [64, 32, 32], "env_embed": [64, 32, 32], "activation": "kgcnn>softplus2"}, "set2set_args": {"channels": 16, "T": 3, "pooling_method": "sum", "init_qstar": "0"}, "node_ff_args": {"units": [64, 32], "activation": "kgcnn>softplus2"}, "edge_ff_args": {"units": [64, 32], "activation": "kgcnn>softplus2"}, "state_ff_args": {"units": [64, 32], "activation": "kgcnn>softplus2"}, "nblocks": 3, "has_ff": true, "dropout": null, "use_set2set": true, "verbose": 10, "output_embedding": "graph", "output_mlp": {"use_bias": [true, true, true], "units": [32, 16, 1], "activation": ["kgcnn>softplus2", "kgcnn>softplus2", "linear"]}}}, "training": {"cross_validation": {"class_name": "KFold", "config": {"n_splits": 5, "random_state": 42, "shuffle": true}}, "fit": {"batch_size": 32, "epochs": 800, "validation_freq": 10, "verbose": 2, "callbacks": [{"class_name": "kgcnn>LinearLearningRateScheduler", "config": {"learning_rate_start": 0.0005, "learning_rate_stop": 1e-05, "epo_min": 100, "epo": 800, "verbose": 0}}]}, "compile": {"optimizer": {"class_name": "Adam", "config": {"learning_rate": 0.0005}}, "loss": "mean_absolute_error"}, "scaler": {"class_name": "QMGraphLabelScaler", "config": {"atomic_number": "node_number", "scaler": [{"class_name": "ExtensiveMolecularLabelScaler", "config": {}}]}}, "multi_target_indices": [10]}, "data": {}, "info": {"postfix": "", "postfix_file": "_U0", "kgcnn_version": "4.0.0"}, "dataset": {"class_name": "QM9Dataset", "module_name": "kgcnn.data.datasets.QM9Dataset", "config": {}, "methods": [{"map_list": {"method": "set_range", "max_distance": 4, "max_neighbours": 30}}]}} \ No newline at end of file diff --git a/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_H.yaml b/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_H.yaml index 92855598..1e512c91 100644 --- a/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_H.yaml +++ b/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_H.yaml @@ -2,7 +2,7 @@ OS: posix_linux backend: tensorflow cuda_available: 'True' data_unit: '[''eV'']' -date_time: '2024-03-24 08:08:03' +date_time: '2024-04-06 21:18:44' device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'', device_type=''GPU'')]' device_memory: '[]' @@ -10,55 +10,106 @@ device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V10 epochs: - 700 - 700 +- 700 +- 700 +- 700 execute_folds: -- 3 +- 0 kgcnn_version: 4.0.1 learning_rate: - 1.0138461220776662e-05 - 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 loss: +- 0.01084364764392376 +- 0.01062938291579485 +- 0.01128083374351263 - 0.011429122649133205 - 0.010737859643995762 max_learning_rate: - 9.999999747378752e-05 - 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 max_loss: +- 0.49890589714050293 +- 0.5023810267448425 +- 0.5039345622062683 - 0.5030210018157959 - 0.4944944977760315 max_scaled_mean_absolute_error: +- 0.5437986254692078 +- 0.5484613180160522 +- 0.5499804019927979 - 0.5491960048675537 - 0.5402727723121643 max_scaled_root_mean_squared_error: +- 0.711073637008667 +- 0.7123429179191589 +- 0.7153045535087585 - 0.7163705229759216 - 0.7079967856407166 max_val_loss: +- 0.10999971628189087 +- 0.11447286605834961 +- 0.1153479665517807 - 0.11011811345815659 - 0.10715076327323914 max_val_scaled_mean_absolute_error: +- 0.11969310790300369 +- 0.12478464096784592 +- 0.1256149858236313 - 0.12001536786556244 - 0.11688880622386932 max_val_scaled_root_mean_squared_error: +- 0.17768365144729614 +- 0.18279235064983368 +- 0.1879362165927887 - 0.18209540843963623 - 0.17592087388038635 min_learning_rate: - 1.0138461220776662e-05 - 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 min_loss: +- 0.01084364764392376 +- 0.010527804493904114 +- 0.011239427141845226 - 0.011429122649133205 - 0.010676178149878979 min_scaled_mean_absolute_error: +- 0.011819379404187202 +- 0.011493400670588017 +- 0.012266315519809723 - 0.012477696873247623 - 0.011664663441479206 min_scaled_root_mean_squared_error: +- 0.022167585790157318 +- 0.020761527121067047 +- 0.021678896620869637 - 0.024382978677749634 - 0.022128688171505928 min_val_loss: +- 0.049516864120960236 +- 0.04946356639266014 +- 0.05167530104517937 - 0.04856857657432556 - 0.0511004775762558 min_val_scaled_mean_absolute_error: +- 0.05378863587975502 +- 0.05382430553436279 +- 0.0561690516769886 - 0.05282432585954666 - 0.05567791312932968 min_val_scaled_root_mean_squared_error: +- 0.09276390820741653 +- 0.09207896888256073 +- 0.09773270785808563 - 0.09097553789615631 - 0.0949467346072197 model_class: make_model @@ -66,23 +117,41 @@ model_name: NMPN model_version: '2023-11-22' multi_target_indices: - 12 -number_histories: 2 +number_histories: 5 scaled_mean_absolute_error: +- 0.011819379404187202 +- 0.011604423634707928 +- 0.012311479076743126 - 0.012477696873247623 - 0.011732123792171478 scaled_root_mean_squared_error: +- 0.022167585790157318 +- 0.020898202434182167 +- 0.02173299714922905 - 0.0244058296084404 - 0.022206485271453857 seed: 42 time_list: +- '13:54:36.969605' +- '14:32:08.879593' +- '14:29:22.284908' - '14:40:24.697245' - '14:40:53.357599' val_loss: +- 0.049516864120960236 +- 0.04946356639266014 +- 0.051717836409807205 - 0.049180060625076294 - 0.05113501846790314 val_scaled_mean_absolute_error: +- 0.05378863587975502 +- 0.05382430553436279 +- 0.056221384555101395 - 0.05348145216703415 - 0.05570865049958229 val_scaled_root_mean_squared_error: +- 0.09324630349874496 +- 0.09215664863586426 +- 0.09849289059638977 - 0.09295357763767242 - 0.09528426826000214 diff --git a/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_HOMO.yaml b/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_HOMO.yaml index bd78e232..4f25126f 100644 --- a/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_HOMO.yaml +++ b/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_HOMO.yaml @@ -2,7 +2,7 @@ OS: posix_linux backend: tensorflow cuda_available: 'True' data_unit: '[''eV'']' -date_time: '2024-03-23 22:09:54' +date_time: '2024-04-05 00:45:36' device_id: '[LogicalDevice(name=''/device:CPU:0'', device_type=''CPU''), LogicalDevice(name=''/device:GPU:0'', device_type=''GPU'')]' device_memory: '[]' @@ -10,55 +10,106 @@ device_name: '[{}, {''compute_capability'': (7, 0), ''device_name'': ''Tesla V10 epochs: - 700 - 700 +- 700 +- 700 +- 700 execute_folds: -- 3 +- 0 kgcnn_version: 4.0.1 learning_rate: - 1.0138461220776662e-05 - 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 loss: +- 0.0222546998411417 +- 0.023373976349830627 +- 0.021985197439789772 - 0.02223457396030426 - 0.020672369748353958 max_learning_rate: - 9.999999747378752e-05 - 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 max_loss: +- 0.49209704995155334 +- 0.5004668235778809 +- 0.4979463815689087 - 0.4959501326084137 - 0.4994843304157257 max_scaled_mean_absolute_error: +- 0.29701611399650574 +- 0.3014347553253174 +- 0.29915961623191833 - 0.29884040355682373 - 0.30065858364105225 max_scaled_root_mean_squared_error: +- 0.40698111057281494 +- 0.4120783805847168 +- 0.40800318121910095 - 0.4097099304199219 - 0.4100411832332611 max_val_loss: +- 0.1851343810558319 +- 0.18510058522224426 +- 0.1847764402627945 - 0.18062184751033783 - 0.17841483652591705 max_val_scaled_mean_absolute_error: +- 0.11162623018026352 +- 0.11141566932201385 +- 0.11087964475154877 - 0.10875657200813293 - 0.10735707730054855 max_val_scaled_root_mean_squared_error: +- 0.15837165713310242 +- 0.1556098908185959 +- 0.15638171136379242 - 0.15188486874103546 - 0.14977742731571198 min_learning_rate: - 1.0138461220776662e-05 - 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 min_loss: +- 0.02194303646683693 +- 0.023017769679427147 +- 0.021789031103253365 - 0.021554268896579742 - 0.02036287635564804 min_scaled_mean_absolute_error: +- 0.013244464993476868 +- 0.013863865286111832 +- 0.013090946711599827 - 0.012987556867301464 - 0.012256953865289688 min_scaled_root_mean_squared_error: +- 0.022892586886882782 +- 0.023964742198586464 +- 0.022301338613033295 - 0.021794689819216728 - 0.020948626101017 min_val_loss: +- 0.1305871456861496 +- 0.1303233504295349 +- 0.13231632113456726 - 0.1286155879497528 - 0.13095012307167053 min_val_scaled_mean_absolute_error: +- 0.07872825860977173 +- 0.0784410685300827 +- 0.07936210930347443 - 0.07742822915315628 - 0.07879547774791718 min_val_scaled_root_mean_squared_error: +- 0.11398513615131378 +- 0.10897812247276306 +- 0.11384902894496918 - 0.10747498273849487 - 0.11117482930421829 model_class: make_model @@ -66,23 +117,41 @@ model_name: NMPN model_version: '2023-11-22' multi_target_indices: - 5 -number_histories: 2 +number_histories: 5 scaled_mean_absolute_error: +- 0.013432618230581284 +- 0.014078390784561634 +- 0.013208731077611446 - 0.01339768711477518 - 0.012443263083696365 scaled_root_mean_squared_error: +- 0.023356115445494652 +- 0.024483170360326767 +- 0.022667573764920235 - 0.02263406291604042 - 0.02139880508184433 seed: 42 time_list: +- '13:57:54.729964' +- '14:35:25.818913' +- '14:25:36.413179' - '14:00:34.388450' - '15:16:20.702028' val_loss: +- 0.13915285468101501 +- 0.14142826199531555 +- 0.14092807471752167 - 0.1397639513015747 - 0.14023743569850922 val_scaled_mean_absolute_error: +- 0.08388713747262955 +- 0.08512960374355316 +- 0.0845414474606514 - 0.08415506780147552 - 0.08439762145280838 val_scaled_root_mean_squared_error: +- 0.12028976529836655 +- 0.11769512295722961 +- 0.1194210946559906 - 0.11572442203760147 - 0.11733932793140411 diff --git a/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_U0.yaml b/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_U0.yaml new file mode 100644 index 00000000..17b69b1d --- /dev/null +++ b/training/results/QM9Dataset/NMPN/NMPN_QM9Dataset_score_U0.yaml @@ -0,0 +1,157 @@ +OS: posix_linux +backend: tensorflow +cuda_available: 'True' +data_unit: '[''eV'']' +date_time: '2024-04-28 00:41:45' +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: +- 700 +- 700 +- 700 +- 700 +- 700 +execute_folds: +- 4 +kgcnn_version: 4.0.1 +learning_rate: +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +loss: +- 0.010586600750684738 +- 0.010699751786887646 +- 0.010791690088808537 +- 0.011309382505714893 +- 0.010040332563221455 +max_learning_rate: +- 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 +- 9.999999747378752e-05 +max_loss: +- 0.49615278840065 +- 0.5089646577835083 +- 0.4942590594291687 +- 0.49713581800460815 +- 0.48694780468940735 +max_scaled_mean_absolute_error: +- 0.5419419407844543 +- 0.5568569302558899 +- 0.5405932068824768 +- 0.5439640283584595 +- 0.533162534236908 +max_scaled_root_mean_squared_error: +- 0.7073643207550049 +- 0.7195283770561218 +- 0.70473712682724 +- 0.7108490467071533 +- 0.700431764125824 +max_val_loss: +- 0.10884089022874832 +- 0.12098347395658493 +- 0.12063153088092804 +- 0.11585988849401474 +- 0.11167512834072113 +max_val_scaled_mean_absolute_error: +- 0.11870746314525604 +- 0.13217699527740479 +- 0.1317155957221985 +- 0.12656128406524658 +- 0.12209741026163101 +max_val_scaled_root_mean_squared_error: +- 0.17907515168190002 +- 0.19688613712787628 +- 0.1916581094264984 +- 0.18621160089969635 +- 0.18084511160850525 +min_learning_rate: +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +- 1.0138461220776662e-05 +min_loss: +- 0.010586600750684738 +- 0.010699751786887646 +- 0.01076754555106163 +- 0.011309382505714893 +- 0.010028050281107426 +min_scaled_mean_absolute_error: +- 0.011563435196876526 +- 0.011706218123435974 +- 0.011776852421462536 +- 0.012374410405755043 +- 0.010979754850268364 +min_scaled_root_mean_squared_error: +- 0.020695233717560768 +- 0.020750470459461212 +- 0.0200729388743639 +- 0.020993834361433983 +- 0.01953849568963051 +min_val_loss: +- 0.05041308328509331 +- 0.04753696173429489 +- 0.05048975721001625 +- 0.04851487651467323 +- 0.048815250396728516 +min_val_scaled_mean_absolute_error: +- 0.054866064339876175 +- 0.0518488809466362 +- 0.055013738572597504 +- 0.052893973886966705 +- 0.053317513316869736 +min_val_scaled_root_mean_squared_error: +- 0.09401647001504898 +- 0.08864697813987732 +- 0.09448324143886566 +- 0.09150421619415283 +- 0.08964235335588455 +model_class: make_model +model_name: NMPN +model_version: '2023-11-22' +multi_target_indices: +- 10 +number_histories: 5 +scaled_mean_absolute_error: +- 0.011563435196876526 +- 0.011706218123435974 +- 0.011803199537098408 +- 0.012374410405755043 +- 0.010993118397891521 +scaled_root_mean_squared_error: +- 0.020695233717560768 +- 0.020750470459461212 +- 0.020097341388463974 +- 0.020994573831558228 +- 0.019553957507014275 +seed: 42 +time_list: +- '14:39:14.802031' +- '14:41:58.528623' +- '14:32:09.688854' +- '14:21:11.672350' +- '14:31:05.184236' +val_loss: +- 0.051478415727615356 +- 0.04778595268726349 +- 0.05092955008149147 +- 0.04886075109243393 +- 0.04927417263388634 +val_scaled_mean_absolute_error: +- 0.056019995361566544 +- 0.05212864279747009 +- 0.05549805611371994 +- 0.053264591842889786 +- 0.053824834525585175 +val_scaled_root_mean_squared_error: +- 0.09579446166753769 +- 0.08933930099010468 +- 0.0962802991271019 +- 0.09229935705661774 +- 0.09004057943820953 diff --git a/training/results/QM9Dataset/NMPN/NMPN_hyper_U0.json b/training/results/QM9Dataset/NMPN/NMPN_hyper_U0.json new file mode 100644 index 00000000..1b4fb8a7 --- /dev/null +++ b/training/results/QM9Dataset/NMPN/NMPN_hyper_U0.json @@ -0,0 +1 @@ +{"model": {"class_name": "make_model", "module_name": "kgcnn.literature.NMPN", "config": {"name": "NMPN", "inputs": [{"shape": [null], "name": "node_number", "dtype": "int64", "ragged": true}, {"shape": [null, 3], "name": "node_coordinates", "dtype": "float32", "ragged": true}, {"shape": [null, 2], "name": "edge_indices", "dtype": "int64", "ragged": true}], "input_tensor_type": "ragged", "input_embedding": null, "input_node_embedding": {"input_dim": 95, "output_dim": 64}, "input_edge_embedding": {"input_dim": 5, "output_dim": 64}, "set2set_args": {"channels": 32, "T": 3, "pooling_method": "sum", "init_qstar": "0"}, "pooling_args": {"pooling_method": "scatter_sum"}, "use_set2set": true, "depth": 3, "node_dim": 128, "verbose": 10, "geometric_edge": true, "make_distance": true, "expand_distance": true, "output_embedding": "graph", "output_mlp": {"use_bias": [true, true, false], "units": [25, 25, 1], "activation": ["selu", "selu", "linear"]}}}, "training": {"cross_validation": {"class_name": "KFold", "config": {"n_splits": 5, "random_state": 42, "shuffle": true}}, "fit": {"batch_size": 32, "epochs": 700, "validation_freq": 10, "verbose": 2, "callbacks": [{"class_name": "kgcnn>LinearLearningRateScheduler", "config": {"learning_rate_start": 0.0001, "learning_rate_stop": 1e-05, "epo_min": 50, "epo": 700, "verbose": 0}}]}, "compile": {"optimizer": {"class_name": "Adam", "config": {"learning_rate": 0.0001}}, "loss": "mean_absolute_error"}, "scaler": {"class_name": "QMGraphLabelScaler", "config": {"atomic_number": "node_number", "scaler": [{"class_name": "ExtensiveMolecularLabelScaler", "config": {}}]}}, "multi_target_indices": [10]}, "data": {}, "info": {"postfix": "", "postfix_file": "_U0", "kgcnn_version": "4.0.0"}, "dataset": {"class_name": "QM9Dataset", "module_name": "kgcnn.data.datasets.QM9Dataset", "config": {}, "methods": [{"map_list": {"method": "set_range", "max_distance": 4, "max_neighbours": 30}}]}} \ No newline at end of file diff --git a/training/results/README.md b/training/results/README.md index 4950e442..31d4330b 100644 --- a/training/results/README.md +++ b/training/results/README.md @@ -169,6 +169,8 @@ Materials Project dataset from Matbench with 132752 crystal structures and their | model | kgcnn | epochs | MAE [eV/atom] | RMSE [eV/atom] | |:--------------------------|:--------|---------:|:-----------------------|:-----------------------| | CGCNN.make_crystal_model | 4.0.0 | 1000 | 0.0298 ± 0.0002 | 0.0747 ± 0.0029 | +| Megnet.make_crystal_model | 4.0.1 | 1000 | 0.0272 ± nan | 0.0700 ± nan | +| PAiNN.make_crystal_model | 4.0.1 | 800 | 0.0241 ± 0.0003 | 0.0558 ± 0.0018 | | Schnet.make_crystal_model | 4.0.0 | 800 | **0.0211 ± 0.0003** | **0.0510 ± 0.0024** | #### MatProjectGapDataset @@ -177,7 +179,9 @@ Materials Project dataset from Matbench with 106113 crystal structures and their | model | kgcnn | epochs | MAE [eV] | RMSE [eV] | |:--------------------------|:--------|---------:|:-----------------------|:-----------------------| -| CGCNN.make_crystal_model | 4.0.0 | 1000 | **0.2039 ± 0.0050** | **0.4882 ± 0.0213** | +| CGCNN.make_crystal_model | 4.0.0 | 1000 | 0.2039 ± 0.0050 | 0.4882 ± 0.0213 | +| Megnet.make_crystal_model | 4.0.1 | 1000 | **0.2028 ± 0.0041** | **0.4871 ± 0.0109** | +| PAiNN.make_crystal_model | 4.0.1 | 800 | 0.2250 ± 0.0128 | 1.4224 ± 1.7103 | | Schnet.make_crystal_model | 4.0.0 | 800 | 1.2226 ± 1.0573 | 58.3713 ± 114.2957 | #### MatProjectIsMetalDataset @@ -187,7 +191,8 @@ Materials Project dataset from Matbench with 106113 crystal structures and their | model | kgcnn | epochs | Accuracy | AUC | |:--------------------------|:--------|---------:|:-----------------------|:-----------------------| | CGCNN.make_crystal_model | 4.0.0 | 100 | 0.8910 ± 0.0027 | 0.9406 ± 0.0024 | -| Megnet.make_crystal_model | 4.0.0 | 100 | **0.8966 ± 0.0033** | **0.9506 ± 0.0026** | +| Megnet.make_crystal_model | 4.0.0 | 100 | 0.8966 ± 0.0033 | **0.9506 ± 0.0026** | +| PAiNN.make_crystal_model | 4.0.0 | 80 | **0.8989 ± 0.0034** | 0.9338 ± 0.0035 | | Schnet.make_crystal_model | 4.0.0 | 80 | 0.8953 ± 0.0058 | 0.9506 ± 0.0053 | #### MatProjectJdft2dDataset @@ -312,11 +317,15 @@ QM7 dataset is a subset of GDB-13. Molecules of up to 23 atoms (including 7 heav QM9 dataset of 134k stable small organic molecules made up of C,H,O,N,F. Labels include geometric, energetic, electronic, and thermodynamic properties. We use a random 5-fold cross-validation, but not all splits are evaluated for cheaper evaluation. Test errors are MAE and for energies are given in [eV]. -| model | kgcnn | epochs | HOMO [eV] | LUMO [eV] | U0 [eV] | H [eV] | G [eV] | -|:--------|:--------|---------:|:-------------------|:-----------------------|:-------------------|:-------------------|:-----------------------| -| Megnet | 4.0.0 | 800 | **nan ± nan** | 0.0407 ± 0.0009 | **nan ± nan** | **nan ± nan** | 0.0169 ± 0.0006 | -| PAiNN | 4.0.0 | 872 | 0.0483 ± 0.0275 | **0.0268 ± 0.0002** | 0.0099 ± 0.0003 | 0.0101 ± 0.0003 | **0.0110 ± 0.0002** | -| Schnet | 4.0.0 | 800 | 0.0402 ± 0.0007 | 0.0340 ± 0.0001 | 0.0142 ± 0.0002 | 0.0146 ± 0.0002 | 0.0143 ± 0.0002 | +| model | kgcnn | epochs | HOMO [eV] | LUMO [eV] | U0 [eV] | H [eV] | G [eV] | +|:----------|:--------|---------:|:-----------------------|:-----------------------|:-----------------------|:-----------------------|:-----------------------| +| DimeNetPP | 4.0.1 | 600 | **0.0291 ± 0.0004** | 0.0284 ± 0.0005 | 0.0097 ± 0.0002 | 0.0100 ± 0.0004 | 0.0108 ± 0.0005 | +| EGNN | 4.0.1 | 800 | 0.0307 ± 0.0008 | **0.0265 ± 0.0005** | **0.0096 ± 0.0002** | 0.0098 ± 0.0003 | 0.0108 ± 0.0003 | +| Megnet | 4.0.1 | 800 | 0.0478 ± 0.0012 | 0.0407 ± 0.0009 | 0.0169 ± 0.0005 | 0.0168 ± 0.0005 | 0.0169 ± 0.0006 | +| MXMNet | 4.0.1 | 900 | 0.0342 ± 0.0016 | 0.0292 ± 0.0022 | 0.0105 ± 0.0051 | **0.0095 ± 0.0034** | **0.0102 ± 0.0013** | +| NMPN | 4.0.1 | 700 | 0.0844 ± 0.0004 | 0.0885 ± 0.0019 | 0.0541 ± 0.0014 | 0.0546 ± 0.0011 | 0.0528 ± 0.0010 | +| PAiNN | 4.0.0 | 872 | 0.0483 ± 0.0275 | 0.0268 ± 0.0002 | 0.0099 ± 0.0003 | 0.0101 ± 0.0003 | 0.0110 ± 0.0002 | +| Schnet | 4.0.0 | 800 | 0.0402 ± 0.0007 | 0.0340 ± 0.0001 | 0.0142 ± 0.0002 | 0.0146 ± 0.0002 | 0.0143 ± 0.0002 | #### SIDERDataset @@ -338,6 +347,7 @@ Tox21 (MoleculeNet) consists of 7831 compounds as smiles and 12 different target | model | kgcnn | epochs | Accuracy | AUC(ROC) | BACC | |:----------|:--------|---------:|:-----------------------|:-----------------------|:-----------------------| | DMPNN | 4.0.0 | 50 | **0.9272 ± 0.0024** | **0.8321 ± 0.0103** | **0.6995 ± 0.0130** | +| EGNN | 4.0.1 | 50 | 0.9164 ± 0.0030 | 0.7655 ± 0.0080 | 0.6839 ± 0.0053 | | GAT | 4.0.0 | 50 | 0.9243 ± 0.0022 | 0.8279 ± 0.0092 | 0.6504 ± 0.0074 | | GATv2 | 4.0.0 | 50 | 0.9222 ± 0.0019 | 0.8251 ± 0.0069 | 0.6760 ± 0.0140 | | GIN | 4.0.0 | 50 | 0.9220 ± 0.0024 | 0.7986 ± 0.0180 | 0.6741 ± 0.0143 |