last-checkpoint/trainer_state.json

{
  "best_metric": 0.0002651347768448962,
  "best_model_checkpoint": "/mimer/NOBACKUP/groups/naiss2023-6-290/stefano/models//PROTAC-Splitter-EncoderDecoder-lr_cosine-opt25/trial-number=8-learning_rate=6.7e-06-warmup_ratio=0.040-num_cycles=6.500/checkpoint-5000",
  "epoch": 0.49319392385085814,
  "eval_steps": 2500,
  "global_step": 5000,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.049319392385085814,
      "grad_norm": 1.609957218170166,
      "learning_rate": 4.764287234716251e-06,
      "loss": 4.427,
      "step": 500
    },
    {
      "epoch": 0.09863878477017163,
      "grad_norm": 1.3062211275100708,
      "learning_rate": 4.151261952929334e-06,
      "loss": 1.1114,
      "step": 1000
    },
    {
      "epoch": 0.14795817715525744,
      "grad_norm": 1.2533338069915771,
      "learning_rate": 2.321578953844436e-07,
      "loss": 0.7793,
      "step": 1500
    },
    {
      "epoch": 0.19727756954034326,
      "grad_norm": 1.1933494806289673,
      "learning_rate": 6.133081099363204e-06,
      "loss": 0.7138,
      "step": 2000
    },
    {
      "epoch": 0.24659696192542907,
      "grad_norm": 1.1922088861465454,
      "learning_rate": 3.1497864683804228e-06,
      "loss": 0.5344,
      "step": 2500
    },
    {
      "epoch": 0.2959163543105149,
      "grad_norm": 1.0372626781463623,
      "learning_rate": 7.384678968603421e-07,
      "loss": 0.451,
      "step": 3000
    },
    {
      "epoch": 0.3452357466956007,
      "grad_norm": 0.9294341206550598,
      "learning_rate": 6.542318154539015e-06,
      "loss": 0.4202,
      "step": 3500
    },
    {
      "epoch": 0.3945551390806865,
      "grad_norm": 0.7656099796295166,
      "learning_rate": 2.164785723897086e-06,
      "loss": 0.3395,
      "step": 4000
    },
    {
      "epoch": 0.4438745314657723,
      "grad_norm": 0.7425525188446045,
      "learning_rate": 1.481401038093183e-06,
      "loss": 0.3034,
      "step": 4500
    },
    {
      "epoch": 0.49319392385085814,
      "grad_norm": 0.9094532132148743,
      "learning_rate": 6.658298821578928e-06,
      "loss": 0.2848,
      "step": 5000
    },
    {
      "epoch": 0.49319392385085814,
      "eval_all_ligands_equal": 0.0002651347768448962,
      "eval_e3_equal": 0.17516570923552807,
      "eval_e3_graph_edit_distance": 9.999999999999999e+63,
      "eval_e3_graph_edit_distance_norm": 1.0,
      "eval_e3_has_attachment_point(s)": 0.9882456915598763,
      "eval_e3_heavy_atoms_difference": -0.993813521873619,
      "eval_e3_heavy_atoms_difference_norm": -0.06273710988010914,
      "eval_e3_tanimoto_similarity": 0.0,
      "eval_e3_valid": 0.9882456915598763,
      "eval_has_all_attachment_points": 0.8872293415819709,
      "eval_has_three_substructures": 0.9905435262925321,
      "eval_heavy_atoms_difference": 36.26442775077331,
      "eval_heavy_atoms_difference_norm": 0.46364946102044235,
      "eval_linker_equal": 0.0036235086168802475,
      "eval_linker_graph_edit_distance": 9.999999999999999e+63,
      "eval_linker_graph_edit_distance_norm": 1.0,
      "eval_linker_has_attachment_point(s)": 0.979584622182943,
      "eval_linker_heavy_atoms_difference": 6.370835174547062,
      "eval_linker_heavy_atoms_difference_norm": 0.2822560976043188,
      "eval_linker_tanimoto_similarity": 0.0,
      "eval_linker_valid": 0.979584622182943,
      "eval_loss": 0.42363283038139343,
      "eval_num_fragments": 2.9927529827662394,
      "eval_poi_equal": 0.10720282810428634,
      "eval_poi_graph_edit_distance": 9.999999999999999e+63,
      "eval_poi_graph_edit_distance_norm": 1.0,
      "eval_poi_has_attachment_point(s)": 0.6131683605832965,
      "eval_poi_heavy_atoms_difference": 14.577286787450287,
      "eval_poi_heavy_atoms_difference_norm": 0.42516346148688866,
      "eval_poi_tanimoto_similarity": 0.0,
      "eval_poi_valid": 0.6131683605832965,
      "eval_reassembly": 0.00035351303579319487,
      "eval_reassembly_nostereo": 0.00035351303579319487,
      "eval_runtime": 1482.7962,
      "eval_samples_per_second": 7.631,
      "eval_steps_per_second": 0.119,
      "eval_tanimoto_similarity": 0.0,
      "eval_valid": 0.6063632346442775,
      "step": 5000
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.33139477262336e+16,
  "train_batch_size": 128,
  "trial_name": "trial-number=8-learning_rate=6.7e-06-warmup_ratio=0.040-num_cycles=6.500",
  "trial_params": {
    "learning_rate": 6.660473554133319e-06,
    "num_cycles": 6.5,
    "warmup_ratio": 0.04
  }
}