ribesstefano's picture
Training in progress, step 10000
9df4b86 verified
raw
history blame
9.87 kB
{
"best_metric": 0.0,
"best_model_checkpoint": "/mimer/NOBACKUP/groups/naiss2023-6-290/stefano/models//PROTAC-Splitter-EncoderDecoder-lr_cosine_restarts-opt25/trial-number=2-learning_rate=4.1e-04-warmup_ratio=0.050-num_cycles=5/checkpoint-5000",
"epoch": 0.9863878477017163,
"eval_steps": 2500,
"global_step": 10000,
"is_hyper_param_search": true,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.049319392385085814,
"grad_norm": 0.35638532042503357,
"learning_rate": 0.0003674335849278369,
"loss": 0.8941,
"step": 500
},
{
"epoch": 0.09863878477017163,
"grad_norm": 0.30781424045562744,
"learning_rate": 0.00023269820894080274,
"loss": 0.1879,
"step": 1000
},
{
"epoch": 0.14795817715525744,
"grad_norm": 0.33434826135635376,
"learning_rate": 8.048701487059137e-05,
"loss": 0.1669,
"step": 1500
},
{
"epoch": 0.19727756954034326,
"grad_norm": 0.43194010853767395,
"learning_rate": 1.6777303215323511e-06,
"loss": 0.1584,
"step": 2000
},
{
"epoch": 0.24659696192542907,
"grad_norm": 0.2756250500679016,
"learning_rate": 0.00036353243224784815,
"loss": 0.1812,
"step": 2500
},
{
"epoch": 0.2959163543105149,
"grad_norm": 0.29849982261657715,
"learning_rate": 0.000226296285743809,
"loss": 0.164,
"step": 3000
},
{
"epoch": 0.3452357466956007,
"grad_norm": 0.359710693359375,
"learning_rate": 7.540659083914886e-05,
"loss": 0.1552,
"step": 3500
},
{
"epoch": 0.3945551390806865,
"grad_norm": 0.41441383957862854,
"learning_rate": 9.520737646127763e-07,
"loss": 0.1501,
"step": 4000
},
{
"epoch": 0.4438745314657723,
"grad_norm": 0.7078486680984497,
"learning_rate": 0.00036029524368022244,
"loss": 1.6264,
"step": 4500
},
{
"epoch": 0.49319392385085814,
"grad_norm": 1.0192649364471436,
"learning_rate": 0.0002211578578707995,
"loss": 3.5719,
"step": 5000
},
{
"epoch": 0.49319392385085814,
"eval_all_ligands_equal": 0.0,
"eval_e3_equal": 0.0,
"eval_e3_graph_edit_distance": 9.999999999999999e+63,
"eval_e3_graph_edit_distance_norm": 1.0,
"eval_e3_has_attachment_point(s)": 0.0,
"eval_e3_heavy_atoms_difference": 28.086433937251435,
"eval_e3_heavy_atoms_difference_norm": 1.0,
"eval_e3_tanimoto_similarity": 0.0,
"eval_e3_valid": 0.0,
"eval_has_all_attachment_points": 0.0,
"eval_has_three_substructures": 0.0,
"eval_heavy_atoms_difference": 75.67370746796288,
"eval_heavy_atoms_difference_norm": 1.0,
"eval_linker_equal": 0.0,
"eval_linker_graph_edit_distance": 9.999999999999999e+63,
"eval_linker_graph_edit_distance_norm": 1.0,
"eval_linker_has_attachment_point(s)": 0.0,
"eval_linker_heavy_atoms_difference": 16.430402121078213,
"eval_linker_heavy_atoms_difference_norm": 1.0,
"eval_linker_tanimoto_similarity": 0.0,
"eval_linker_valid": 0.0,
"eval_loss": 3.6144495010375977,
"eval_num_fragments": 1.0093680954485196,
"eval_poi_equal": 0.0,
"eval_poi_graph_edit_distance": 9.999999999999999e+63,
"eval_poi_graph_edit_distance_norm": 1.0,
"eval_poi_has_attachment_point(s)": 0.0,
"eval_poi_heavy_atoms_difference": 31.15687140963323,
"eval_poi_heavy_atoms_difference_norm": 1.0,
"eval_poi_tanimoto_similarity": 0.0,
"eval_poi_valid": 0.0,
"eval_reassembly": 0.0,
"eval_reassembly_nostereo": 0.0,
"eval_runtime": 5677.767,
"eval_samples_per_second": 1.993,
"eval_steps_per_second": 0.031,
"eval_tanimoto_similarity": 0.0,
"eval_valid": 0.0,
"step": 5000
},
{
"epoch": 0.542513316235944,
"grad_norm": 1.571471929550171,
"learning_rate": 7.14348231763617e-05,
"loss": 3.5644,
"step": 5500
},
{
"epoch": 0.5918327086210298,
"grad_norm": 2.053893804550171,
"learning_rate": 5.183110871239787e-07,
"loss": 3.5542,
"step": 6000
},
{
"epoch": 0.6411521010061156,
"grad_norm": 1.0490036010742188,
"learning_rate": 0.00035610684750814095,
"loss": 3.5695,
"step": 6500
},
{
"epoch": 0.6904714933912014,
"grad_norm": 1.165826678276062,
"learning_rate": 0.00021471914406652085,
"loss": 3.5649,
"step": 7000
},
{
"epoch": 0.7397908857762873,
"grad_norm": 1.7424224615097046,
"learning_rate": 6.659002726278426e-05,
"loss": 3.5572,
"step": 7500
},
{
"epoch": 0.7397908857762873,
"eval_all_ligands_equal": 0.0,
"eval_e3_equal": 0.0,
"eval_e3_graph_edit_distance": 9.999999999999999e+63,
"eval_e3_graph_edit_distance_norm": 1.0,
"eval_e3_has_attachment_point(s)": 0.0,
"eval_e3_heavy_atoms_difference": 28.086433937251435,
"eval_e3_heavy_atoms_difference_norm": 1.0,
"eval_e3_tanimoto_similarity": 0.0,
"eval_e3_valid": 0.0,
"eval_has_all_attachment_points": 0.0,
"eval_has_three_substructures": 0.0,
"eval_heavy_atoms_difference": 75.67370746796288,
"eval_heavy_atoms_difference_norm": 1.0,
"eval_linker_equal": 0.0,
"eval_linker_graph_edit_distance": 9.999999999999999e+63,
"eval_linker_graph_edit_distance_norm": 1.0,
"eval_linker_has_attachment_point(s)": 0.0,
"eval_linker_heavy_atoms_difference": 16.430402121078213,
"eval_linker_heavy_atoms_difference_norm": 1.0,
"eval_linker_tanimoto_similarity": 0.0,
"eval_linker_valid": 0.0,
"eval_loss": 3.5743491649627686,
"eval_num_fragments": 1.002032699955811,
"eval_poi_equal": 0.0,
"eval_poi_graph_edit_distance": 9.999999999999999e+63,
"eval_poi_graph_edit_distance_norm": 1.0,
"eval_poi_has_attachment_point(s)": 0.0,
"eval_poi_heavy_atoms_difference": 31.15687140963323,
"eval_poi_heavy_atoms_difference_norm": 1.0,
"eval_poi_tanimoto_similarity": 0.0,
"eval_poi_valid": 0.0,
"eval_reassembly": 0.0,
"eval_reassembly_nostereo": 0.0,
"eval_runtime": 5648.6414,
"eval_samples_per_second": 2.003,
"eval_steps_per_second": 0.031,
"eval_tanimoto_similarity": 0.0,
"eval_valid": 0.0,
"step": 7500
},
{
"epoch": 0.789110278161373,
"grad_norm": 1.879319190979004,
"learning_rate": 1.6001955252906495e-07,
"loss": 3.5524,
"step": 8000
},
{
"epoch": 0.8384296705464589,
"grad_norm": 1.0135574340820312,
"learning_rate": 0.0003517647166225557,
"loss": 3.5653,
"step": 8500
},
{
"epoch": 0.8877490629315447,
"grad_norm": 1.2660466432571411,
"learning_rate": 0.00020826906098118177,
"loss": 3.5623,
"step": 9000
},
{
"epoch": 0.9370684553166305,
"grad_norm": 1.7765446901321411,
"learning_rate": 6.18830155324002e-05,
"loss": 3.5552,
"step": 9500
},
{
"epoch": 0.9863878477017163,
"grad_norm": 1.804037094116211,
"learning_rate": 6.401587861099511e-09,
"loss": 3.5521,
"step": 10000
},
{
"epoch": 0.9863878477017163,
"eval_all_ligands_equal": 0.0,
"eval_e3_equal": 0.0,
"eval_e3_graph_edit_distance": 9.999999999999999e+63,
"eval_e3_graph_edit_distance_norm": 1.0,
"eval_e3_has_attachment_point(s)": 0.0,
"eval_e3_heavy_atoms_difference": 28.086433937251435,
"eval_e3_heavy_atoms_difference_norm": 1.0,
"eval_e3_tanimoto_similarity": 0.0,
"eval_e3_valid": 0.0,
"eval_has_all_attachment_points": 0.0,
"eval_has_three_substructures": 8.837825894829872e-05,
"eval_heavy_atoms_difference": 75.67370746796288,
"eval_heavy_atoms_difference_norm": 1.0,
"eval_linker_equal": 0.0,
"eval_linker_graph_edit_distance": 9.999999999999999e+63,
"eval_linker_graph_edit_distance_norm": 1.0,
"eval_linker_has_attachment_point(s)": 0.0,
"eval_linker_heavy_atoms_difference": 16.430402121078213,
"eval_linker_heavy_atoms_difference_norm": 1.0,
"eval_linker_tanimoto_similarity": 0.0,
"eval_linker_valid": 0.0,
"eval_loss": 3.57191801071167,
"eval_num_fragments": 1.002032699955811,
"eval_poi_equal": 0.0,
"eval_poi_graph_edit_distance": 9.999999999999999e+63,
"eval_poi_graph_edit_distance_norm": 1.0,
"eval_poi_has_attachment_point(s)": 0.0,
"eval_poi_heavy_atoms_difference": 31.15687140963323,
"eval_poi_heavy_atoms_difference_norm": 1.0,
"eval_poi_tanimoto_similarity": 0.0,
"eval_poi_valid": 0.0,
"eval_reassembly": 0.0,
"eval_reassembly_nostereo": 0.0,
"eval_runtime": 5651.5021,
"eval_samples_per_second": 2.002,
"eval_steps_per_second": 0.031,
"eval_tanimoto_similarity": 0.0,
"eval_valid": 0.0,
"step": 10000
}
],
"logging_steps": 500,
"max_steps": 10000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 5000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 8.663036348367667e+16,
"train_batch_size": 128,
"trial_name": "trial-number=2-learning_rate=4.1e-04-warmup_ratio=0.050-num_cycles=5",
"trial_params": {
"learning_rate": 0.00040685588858105847,
"num_cycles": 5,
"warmup_ratio": 0.05
}
}