pdbbindfull.out

Some weights of the model checkpoint at data/saved_models/our_90epochs_saved_model were not used when initializing RobertaForSelfiesClassification: ['lm_head.dense.bias', 'lm_head.bias', 'lm_head.layer_norm.bias', 'lm_head.layer_norm.weight', 'lm_head.dense.weight']
- This IS expected if you are initializing RobertaForSelfiesClassification from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
- This IS NOT expected if you are initializing RobertaForSelfiesClassification from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
Some weights of RobertaForSelfiesClassification were not initialized from the model checkpoint at data/saved_models/our_90epochs_saved_model and are newly initialized: ['classifier.dense.bias', 'classifier.dense.weight', 'classifier.out_proj.weight', 'classifier.out_proj.bias', 'roberta.pooler.dense.bias', 'roberta.pooler.dense.weight']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
Not using a scaler.
***** Running training *****
  Num examples = 7898
  Num Epochs = 50
  Instantaneous batch size per device = 8
  Total train batch size (w. parallel, distributed & accumulation) = 8
  Gradient Accumulation steps = 1
  Total optimization steps = 49400
{'loss': 4.6519, 'learning_rate': 9.898785425101214e-06, 'epoch': 0.51}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.3328564167022705, 'eval_mse': 2.3328561782836914, 'eval_rmse': 1.5273690223693848, 'eval_mae': 1.2115871906280518, 'eval_runtime': 2.139, 'eval_samples_per_second': 460.504, 'eval_steps_per_second': 57.972, 'epoch': 1.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-988
Configuration saved in our_pdbbindfull_model/checkpoint-988/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-988/pytorch_model.bin
{'loss': 1.9813, 'learning_rate': 9.79757085020243e-06, 'epoch': 1.01}
{'loss': 1.7364, 'learning_rate': 9.696356275303645e-06, 'epoch': 1.52}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.25815749168396, 'eval_mse': 2.25815749168396, 'eval_rmse': 1.5027166604995728, 'eval_mae': 1.1845210790634155, 'eval_runtime': 2.2991, 'eval_samples_per_second': 428.436, 'eval_steps_per_second': 53.935, 'epoch': 2.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-1976
Configuration saved in our_pdbbindfull_model/checkpoint-1976/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-1976/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-988] due to args.save_total_limit
{'loss': 1.6471, 'learning_rate': 9.595141700404859e-06, 'epoch': 2.02}
{'loss': 1.4318, 'learning_rate': 9.493927125506074e-06, 'epoch': 2.53}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2711498737335205, 'eval_mse': 2.2711498737335205, 'eval_rmse': 1.5070334672927856, 'eval_mae': 1.195154070854187, 'eval_runtime': 1.9177, 'eval_samples_per_second': 513.633, 'eval_steps_per_second': 64.66, 'epoch': 3.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-2964
Configuration saved in our_pdbbindfull_model/checkpoint-2964/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-2964/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-1976] due to args.save_total_limit
{'loss': 1.3929, 'learning_rate': 9.392712550607288e-06, 'epoch': 3.04}
{'loss': 1.2099, 'learning_rate': 9.291497975708503e-06, 'epoch': 3.54}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.5615780353546143, 'eval_mse': 2.5615780353546143, 'eval_rmse': 1.60049307346344, 'eval_mae': 1.2618587017059326, 'eval_runtime': 2.4694, 'eval_samples_per_second': 398.879, 'eval_steps_per_second': 50.214, 'epoch': 4.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-3952
Configuration saved in our_pdbbindfull_model/checkpoint-3952/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-3952/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-2964] due to args.save_total_limit
{'loss': 1.1585, 'learning_rate': 9.190283400809717e-06, 'epoch': 4.05}
{'loss': 0.9984, 'learning_rate': 9.089068825910932e-06, 'epoch': 4.55}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.386498212814331, 'eval_mse': 2.386498212814331, 'eval_rmse': 1.5448294878005981, 'eval_mae': 1.2156014442443848, 'eval_runtime': 2.1201, 'eval_samples_per_second': 464.611, 'eval_steps_per_second': 58.489, 'epoch': 5.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-4940
Configuration saved in our_pdbbindfull_model/checkpoint-4940/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-4940/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-3952] due to args.save_total_limit
{'loss': 1.0114, 'learning_rate': 8.987854251012147e-06, 'epoch': 5.06}
{'loss': 0.897, 'learning_rate': 8.886639676113361e-06, 'epoch': 5.57}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.3048605918884277, 'eval_mse': 2.3048605918884277, 'eval_rmse': 1.5181766748428345, 'eval_mae': 1.186144471168518, 'eval_runtime': 2.4807, 'eval_samples_per_second': 397.063, 'eval_steps_per_second': 49.986, 'epoch': 6.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-5928
Configuration saved in our_pdbbindfull_model/checkpoint-5928/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-5928/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-4940] due to args.save_total_limit
{'loss': 0.912, 'learning_rate': 8.785425101214575e-06, 'epoch': 6.07}
{'loss': 0.7973, 'learning_rate': 8.68421052631579e-06, 'epoch': 6.58}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.423027992248535, 'eval_mse': 2.4230284690856934, 'eval_rmse': 1.556607961654663, 'eval_mae': 1.223141074180603, 'eval_runtime': 2.3801, 'eval_samples_per_second': 413.851, 'eval_steps_per_second': 52.099, 'epoch': 7.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-6916
Configuration saved in our_pdbbindfull_model/checkpoint-6916/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-6916/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-5928] due to args.save_total_limit
{'loss': 0.8141, 'learning_rate': 8.582995951417005e-06, 'epoch': 7.09}
{'loss': 0.7115, 'learning_rate': 8.481781376518219e-06, 'epoch': 7.59}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.51936936378479, 'eval_mse': 2.51936936378479, 'eval_rmse': 1.587252140045166, 'eval_mae': 1.2446390390396118, 'eval_runtime': 2.2497, 'eval_samples_per_second': 437.844, 'eval_steps_per_second': 55.119, 'epoch': 8.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-7904
Configuration saved in our_pdbbindfull_model/checkpoint-7904/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-7904/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-6916] due to args.save_total_limit
{'loss': 0.7455, 'learning_rate': 8.380566801619434e-06, 'epoch': 8.1}
{'loss': 0.6548, 'learning_rate': 8.279352226720648e-06, 'epoch': 8.6}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.3894152641296387, 'eval_mse': 2.3894152641296387, 'eval_rmse': 1.5457733869552612, 'eval_mae': 1.20915687084198, 'eval_runtime': 2.012, 'eval_samples_per_second': 489.572, 'eval_steps_per_second': 61.631, 'epoch': 9.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-8892
Configuration saved in our_pdbbindfull_model/checkpoint-8892/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-8892/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-7904] due to args.save_total_limit
{'loss': 0.6625, 'learning_rate': 8.178137651821862e-06, 'epoch': 9.11}
{'loss': 0.6466, 'learning_rate': 8.076923076923077e-06, 'epoch': 9.62}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.3797826766967773, 'eval_mse': 2.3797826766967773, 'eval_rmse': 1.5426543951034546, 'eval_mae': 1.209175944328308, 'eval_runtime': 1.9282, 'eval_samples_per_second': 510.852, 'eval_steps_per_second': 64.31, 'epoch': 10.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-9880
Configuration saved in our_pdbbindfull_model/checkpoint-9880/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-9880/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-8892] due to args.save_total_limit
{'loss': 0.6429, 'learning_rate': 7.975708502024292e-06, 'epoch': 10.12}
{'loss': 0.6218, 'learning_rate': 7.874493927125508e-06, 'epoch': 10.63}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.325528383255005, 'eval_mse': 2.325528383255005, 'eval_rmse': 1.5249682664871216, 'eval_mae': 1.197866678237915, 'eval_runtime': 2.2819, 'eval_samples_per_second': 431.658, 'eval_steps_per_second': 54.341, 'epoch': 11.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-10868
Configuration saved in our_pdbbindfull_model/checkpoint-10868/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-10868/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-9880] due to args.save_total_limit
{'loss': 0.596, 'learning_rate': 7.773279352226721e-06, 'epoch': 11.13}
{'loss': 0.5498, 'learning_rate': 7.672064777327935e-06, 'epoch': 11.64}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.391416072845459, 'eval_mse': 2.391416072845459, 'eval_rmse': 1.5464204549789429, 'eval_mae': 1.2055319547653198, 'eval_runtime': 2.3484, 'eval_samples_per_second': 419.443, 'eval_steps_per_second': 52.803, 'epoch': 12.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-11856
Configuration saved in our_pdbbindfull_model/checkpoint-11856/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-11856/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-10868] due to args.save_total_limit
{'loss': 0.5617, 'learning_rate': 7.570850202429151e-06, 'epoch': 12.15}
{'loss': 0.5403, 'learning_rate': 7.469635627530365e-06, 'epoch': 12.65}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.4148483276367188, 'eval_mse': 2.4148483276367188, 'eval_rmse': 1.5539782047271729, 'eval_mae': 1.2168691158294678, 'eval_runtime': 2.3371, 'eval_samples_per_second': 421.463, 'eval_steps_per_second': 53.057, 'epoch': 13.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-12844
Configuration saved in our_pdbbindfull_model/checkpoint-12844/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-12844/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-11856] due to args.save_total_limit
{'loss': 0.5424, 'learning_rate': 7.368421052631579e-06, 'epoch': 13.16}
{'loss': 0.5392, 'learning_rate': 7.267206477732795e-06, 'epoch': 13.66}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.429582357406616, 'eval_mse': 2.429582357406616, 'eval_rmse': 1.5587117671966553, 'eval_mae': 1.2232253551483154, 'eval_runtime': 1.971, 'eval_samples_per_second': 499.752, 'eval_steps_per_second': 62.913, 'epoch': 14.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-13832
Configuration saved in our_pdbbindfull_model/checkpoint-13832/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-13832/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-12844] due to args.save_total_limit
{'loss': 0.5105, 'learning_rate': 7.165991902834008e-06, 'epoch': 14.17}
{'loss': 0.5224, 'learning_rate': 7.064777327935223e-06, 'epoch': 14.68}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.279872417449951, 'eval_mse': 2.279872417449951, 'eval_rmse': 1.5099246501922607, 'eval_mae': 1.1799190044403076, 'eval_runtime': 2.2712, 'eval_samples_per_second': 433.694, 'eval_steps_per_second': 54.597, 'epoch': 15.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-14820
Configuration saved in our_pdbbindfull_model/checkpoint-14820/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-14820/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-13832] due to args.save_total_limit
{'loss': 0.5082, 'learning_rate': 6.963562753036438e-06, 'epoch': 15.18}
{'loss': 0.4868, 'learning_rate': 6.862348178137653e-06, 'epoch': 15.69}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.309877634048462, 'eval_mse': 2.309877634048462, 'eval_rmse': 1.519828200340271, 'eval_mae': 1.1822220087051392, 'eval_runtime': 1.9368, 'eval_samples_per_second': 508.564, 'eval_steps_per_second': 64.022, 'epoch': 16.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-15808
Configuration saved in our_pdbbindfull_model/checkpoint-15808/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-15808/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-14820] due to args.save_total_limit
{'loss': 0.5021, 'learning_rate': 6.761133603238867e-06, 'epoch': 16.19}
{'loss': 0.4702, 'learning_rate': 6.6599190283400816e-06, 'epoch': 16.7}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.205157518386841, 'eval_mse': 2.205157518386841, 'eval_rmse': 1.4849772453308105, 'eval_mae': 1.1669542789459229, 'eval_runtime': 2.2927, 'eval_samples_per_second': 429.627, 'eval_steps_per_second': 54.085, 'epoch': 17.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-16796
Configuration saved in our_pdbbindfull_model/checkpoint-16796/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-16796/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-15808] due to args.save_total_limit
{'loss': 0.4905, 'learning_rate': 6.558704453441296e-06, 'epoch': 17.21}
{'loss': 0.4665, 'learning_rate': 6.457489878542511e-06, 'epoch': 17.71}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.323201894760132, 'eval_mse': 2.323201894760132, 'eval_rmse': 1.5242053270339966, 'eval_mae': 1.1936134099960327, 'eval_runtime': 2.1851, 'eval_samples_per_second': 450.772, 'eval_steps_per_second': 56.747, 'epoch': 18.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-17784
Configuration saved in our_pdbbindfull_model/checkpoint-17784/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-17784/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-16796] due to args.save_total_limit
{'loss': 0.4431, 'learning_rate': 6.356275303643725e-06, 'epoch': 18.22}
{'loss': 0.483, 'learning_rate': 6.2550607287449395e-06, 'epoch': 18.72}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.3569979667663574, 'eval_mse': 2.3569979667663574, 'eval_rmse': 1.5352517366409302, 'eval_mae': 1.2059428691864014, 'eval_runtime': 6.1451, 'eval_samples_per_second': 160.29, 'eval_steps_per_second': 20.179, 'epoch': 19.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-18772
Configuration saved in our_pdbbindfull_model/checkpoint-18772/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-18772/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-17784] due to args.save_total_limit
{'loss': 0.4433, 'learning_rate': 6.153846153846155e-06, 'epoch': 19.23}
{'loss': 0.4561, 'learning_rate': 6.0526315789473685e-06, 'epoch': 19.74}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.313786268234253, 'eval_mse': 2.313786268234253, 'eval_rmse': 1.5211135149002075, 'eval_mae': 1.190987229347229, 'eval_runtime': 6.1814, 'eval_samples_per_second': 159.348, 'eval_steps_per_second': 20.06, 'epoch': 20.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-19760
Configuration saved in our_pdbbindfull_model/checkpoint-19760/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-19760/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-18772] due to args.save_total_limit
{'loss': 0.4688, 'learning_rate': 5.951417004048583e-06, 'epoch': 20.24}
{'loss': 0.4577, 'learning_rate': 5.850202429149798e-06, 'epoch': 20.75}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2783560752868652, 'eval_mse': 2.2783560752868652, 'eval_rmse': 1.5094224214553833, 'eval_mae': 1.1627191305160522, 'eval_runtime': 6.5272, 'eval_samples_per_second': 150.907, 'eval_steps_per_second': 18.997, 'epoch': 21.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-20748
Configuration saved in our_pdbbindfull_model/checkpoint-20748/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-20748/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-19760] due to args.save_total_limit
{'loss': 0.4261, 'learning_rate': 5.748987854251013e-06, 'epoch': 21.26}
{'loss': 0.4251, 'learning_rate': 5.6477732793522265e-06, 'epoch': 21.76}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.3060412406921387, 'eval_mse': 2.3060412406921387, 'eval_rmse': 1.5185655355453491, 'eval_mae': 1.176581621170044, 'eval_runtime': 6.411, 'eval_samples_per_second': 153.641, 'eval_steps_per_second': 19.342, 'epoch': 22.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-21736
Configuration saved in our_pdbbindfull_model/checkpoint-21736/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-21736/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-20748] due to args.save_total_limit
{'loss': 0.4235, 'learning_rate': 5.546558704453442e-06, 'epoch': 22.27}
{'loss': 0.4477, 'learning_rate': 5.445344129554656e-06, 'epoch': 22.77}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.3399007320404053, 'eval_mse': 2.339900493621826, 'eval_rmse': 1.5296733379364014, 'eval_mae': 1.1934117078781128, 'eval_runtime': 6.5358, 'eval_samples_per_second': 150.708, 'eval_steps_per_second': 18.972, 'epoch': 23.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-22724
Configuration saved in our_pdbbindfull_model/checkpoint-22724/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-22724/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-21736] due to args.save_total_limit
{'loss': 0.3849, 'learning_rate': 5.344129554655872e-06, 'epoch': 23.28}
{'loss': 0.4274, 'learning_rate': 5.242914979757085e-06, 'epoch': 23.79}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2604634761810303, 'eval_mse': 2.2604634761810303, 'eval_rmse': 1.503483772277832, 'eval_mae': 1.1710097789764404, 'eval_runtime': 6.4581, 'eval_samples_per_second': 152.522, 'eval_steps_per_second': 19.201, 'epoch': 24.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-23712
Configuration saved in our_pdbbindfull_model/checkpoint-23712/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-23712/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-22724] due to args.save_total_limit
{'loss': 0.402, 'learning_rate': 5.1417004048583e-06, 'epoch': 24.29}
{'loss': 0.4176, 'learning_rate': 5.040485829959515e-06, 'epoch': 24.8}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.277299165725708, 'eval_mse': 2.277299404144287, 'eval_rmse': 1.5090724229812622, 'eval_mae': 1.1737443208694458, 'eval_runtime': 6.5489, 'eval_samples_per_second': 150.407, 'eval_steps_per_second': 18.935, 'epoch': 25.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-24700
Configuration saved in our_pdbbindfull_model/checkpoint-24700/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-24700/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-23712] due to args.save_total_limit
{'loss': 0.4278, 'learning_rate': 4.939271255060729e-06, 'epoch': 25.3}
{'loss': 0.4105, 'learning_rate': 4.838056680161944e-06, 'epoch': 25.81}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.3140110969543457, 'eval_mse': 2.3140110969543457, 'eval_rmse': 1.521187424659729, 'eval_mae': 1.1842072010040283, 'eval_runtime': 6.5703, 'eval_samples_per_second': 149.918, 'eval_steps_per_second': 18.873, 'epoch': 26.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-25688
Configuration saved in our_pdbbindfull_model/checkpoint-25688/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-25688/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-24700] due to args.save_total_limit
{'loss': 0.404, 'learning_rate': 4.736842105263158e-06, 'epoch': 26.32}
{'loss': 0.3843, 'learning_rate': 4.635627530364373e-06, 'epoch': 26.82}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2196788787841797, 'eval_mse': 2.2196786403656006, 'eval_rmse': 1.489858627319336, 'eval_mae': 1.1630769968032837, 'eval_runtime': 4.2474, 'eval_samples_per_second': 231.905, 'eval_steps_per_second': 29.194, 'epoch': 27.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-26676
Configuration saved in our_pdbbindfull_model/checkpoint-26676/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-26676/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-25688] due to args.save_total_limit
{'loss': 0.4395, 'learning_rate': 4.534412955465588e-06, 'epoch': 27.33}
{'loss': 0.3985, 'learning_rate': 4.433198380566802e-06, 'epoch': 27.83}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2451844215393066, 'eval_mse': 2.2451844215393066, 'eval_rmse': 1.4983938932418823, 'eval_mae': 1.1653474569320679, 'eval_runtime': 4.5968, 'eval_samples_per_second': 214.279, 'eval_steps_per_second': 26.975, 'epoch': 28.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-27664
Configuration saved in our_pdbbindfull_model/checkpoint-27664/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-27664/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-26676] due to args.save_total_limit
{'loss': 0.3633, 'learning_rate': 4.3319838056680166e-06, 'epoch': 28.34}
{'loss': 0.4055, 'learning_rate': 4.230769230769231e-06, 'epoch': 28.85}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.1678428649902344, 'eval_mse': 2.1678428649902344, 'eval_rmse': 1.4723596572875977, 'eval_mae': 1.1424471139907837, 'eval_runtime': 3.4579, 'eval_samples_per_second': 284.854, 'eval_steps_per_second': 35.86, 'epoch': 29.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-28652
Configuration saved in our_pdbbindfull_model/checkpoint-28652/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-28652/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-27664] due to args.save_total_limit
{'loss': 0.4101, 'learning_rate': 4.1295546558704455e-06, 'epoch': 29.35}
{'loss': 0.3916, 'learning_rate': 4.028340080971661e-06, 'epoch': 29.86}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2453699111938477, 'eval_mse': 2.2453699111938477, 'eval_rmse': 1.4984558820724487, 'eval_mae': 1.166300654411316, 'eval_runtime': 2.9933, 'eval_samples_per_second': 329.066, 'eval_steps_per_second': 41.426, 'epoch': 30.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-29640
Configuration saved in our_pdbbindfull_model/checkpoint-29640/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-29640/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-28652] due to args.save_total_limit
{'loss': 0.3769, 'learning_rate': 3.9271255060728745e-06, 'epoch': 30.36}
{'loss': 0.3905, 'learning_rate': 3.825910931174089e-06, 'epoch': 30.87}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.278787136077881, 'eval_mse': 2.278787136077881, 'eval_rmse': 1.5095652341842651, 'eval_mae': 1.170925259590149, 'eval_runtime': 2.1743, 'eval_samples_per_second': 453.018, 'eval_steps_per_second': 57.03, 'epoch': 31.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-30628
Configuration saved in our_pdbbindfull_model/checkpoint-30628/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-30628/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-29640] due to args.save_total_limit
{'loss': 0.3638, 'learning_rate': 3.724696356275304e-06, 'epoch': 31.38}
{'loss': 0.4033, 'learning_rate': 3.6234817813765184e-06, 'epoch': 31.88}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.218869686126709, 'eval_mse': 2.218869686126709, 'eval_rmse': 1.4895870685577393, 'eval_mae': 1.1555440425872803, 'eval_runtime': 2.8668, 'eval_samples_per_second': 343.59, 'eval_steps_per_second': 43.254, 'epoch': 32.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-31616
Configuration saved in our_pdbbindfull_model/checkpoint-31616/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-31616/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-30628] due to args.save_total_limit
{'loss': 0.3432, 'learning_rate': 3.522267206477733e-06, 'epoch': 32.39}
{'loss': 0.3603, 'learning_rate': 3.421052631578948e-06, 'epoch': 32.89}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2483489513397217, 'eval_mse': 2.2483489513397217, 'eval_rmse': 1.4994494915008545, 'eval_mae': 1.1578136682510376, 'eval_runtime': 2.7175, 'eval_samples_per_second': 362.47, 'eval_steps_per_second': 45.631, 'epoch': 33.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-32604
Configuration saved in our_pdbbindfull_model/checkpoint-32604/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-32604/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-31616] due to args.save_total_limit
{'loss': 0.3832, 'learning_rate': 3.3198380566801623e-06, 'epoch': 33.4}
{'loss': 0.3955, 'learning_rate': 3.218623481781377e-06, 'epoch': 33.91}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2034738063812256, 'eval_mse': 2.2034738063812256, 'eval_rmse': 1.484410285949707, 'eval_mae': 1.148417353630066, 'eval_runtime': 2.6975, 'eval_samples_per_second': 365.15, 'eval_steps_per_second': 45.968, 'epoch': 34.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-33592
Configuration saved in our_pdbbindfull_model/checkpoint-33592/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-33592/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-32604] due to args.save_total_limit
{'loss': 0.3517, 'learning_rate': 3.1174089068825913e-06, 'epoch': 34.41}
{'loss': 0.3802, 'learning_rate': 3.0161943319838062e-06, 'epoch': 34.92}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2388076782226562, 'eval_mse': 2.238807201385498, 'eval_rmse': 1.4962644577026367, 'eval_mae': 1.1580203771591187, 'eval_runtime': 2.781, 'eval_samples_per_second': 354.184, 'eval_steps_per_second': 44.588, 'epoch': 35.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-34580
Configuration saved in our_pdbbindfull_model/checkpoint-34580/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-34580/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-33592] due to args.save_total_limit
{'loss': 0.3543, 'learning_rate': 2.9149797570850203e-06, 'epoch': 35.43}
{'loss': 0.3818, 'learning_rate': 2.8137651821862348e-06, 'epoch': 35.93}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.3153350353240967, 'eval_mse': 2.3153350353240967, 'eval_rmse': 1.5216225385665894, 'eval_mae': 1.178168773651123, 'eval_runtime': 2.7673, 'eval_samples_per_second': 355.946, 'eval_steps_per_second': 44.81, 'epoch': 36.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-35568
Configuration saved in our_pdbbindfull_model/checkpoint-35568/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-35568/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-34580] due to args.save_total_limit
{'loss': 0.3451, 'learning_rate': 2.7125506072874497e-06, 'epoch': 36.44}
{'loss': 0.3837, 'learning_rate': 2.6113360323886646e-06, 'epoch': 36.94}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2114505767822266, 'eval_mse': 2.2114510536193848, 'eval_rmse': 1.4870948791503906, 'eval_mae': 1.1500502824783325, 'eval_runtime': 4.6473, 'eval_samples_per_second': 211.95, 'eval_steps_per_second': 26.682, 'epoch': 37.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-36556
Configuration saved in our_pdbbindfull_model/checkpoint-36556/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-36556/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-35568] due to args.save_total_limit
{'loss': 0.3335, 'learning_rate': 2.5101214574898787e-06, 'epoch': 37.45}
{'loss': 0.3724, 'learning_rate': 2.408906882591093e-06, 'epoch': 37.96}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.190290927886963, 'eval_mse': 2.190290689468384, 'eval_rmse': 1.4799630641937256, 'eval_mae': 1.1481947898864746, 'eval_runtime': 4.4593, 'eval_samples_per_second': 220.888, 'eval_steps_per_second': 27.807, 'epoch': 38.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-37544
Configuration saved in our_pdbbindfull_model/checkpoint-37544/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-37544/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-36556] due to args.save_total_limit
{'loss': 0.3631, 'learning_rate': 2.307692307692308e-06, 'epoch': 38.46}
{'loss': 0.3525, 'learning_rate': 2.2064777327935226e-06, 'epoch': 38.97}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.208022117614746, 'eval_mse': 2.208022117614746, 'eval_rmse': 1.4859415292739868, 'eval_mae': 1.1508959531784058, 'eval_runtime': 4.5419, 'eval_samples_per_second': 216.868, 'eval_steps_per_second': 27.301, 'epoch': 39.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-38532
Configuration saved in our_pdbbindfull_model/checkpoint-38532/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-38532/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-37544] due to args.save_total_limit
{'loss': 0.3589, 'learning_rate': 2.105263157894737e-06, 'epoch': 39.47}
{'loss': 0.3477, 'learning_rate': 2.0040485829959516e-06, 'epoch': 39.98}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2591028213500977, 'eval_mse': 2.2591030597686768, 'eval_rmse': 1.5030312538146973, 'eval_mae': 1.1669268608093262, 'eval_runtime': 4.628, 'eval_samples_per_second': 212.837, 'eval_steps_per_second': 26.794, 'epoch': 40.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-39520
Configuration saved in our_pdbbindfull_model/checkpoint-39520/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-39520/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-38532] due to args.save_total_limit
{'loss': 0.3239, 'learning_rate': 1.902834008097166e-06, 'epoch': 40.49}
{'loss': 0.3891, 'learning_rate': 1.8016194331983807e-06, 'epoch': 40.99}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2003180980682373, 'eval_mse': 2.2003180980682373, 'eval_rmse': 1.483346939086914, 'eval_mae': 1.1513614654541016, 'eval_runtime': 4.722, 'eval_samples_per_second': 208.597, 'eval_steps_per_second': 26.26, 'epoch': 41.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-40508
Configuration saved in our_pdbbindfull_model/checkpoint-40508/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-40508/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-39520] due to args.save_total_limit
{'loss': 0.3474, 'learning_rate': 1.7004048582995952e-06, 'epoch': 41.5}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2396390438079834, 'eval_mse': 2.2396390438079834, 'eval_rmse': 1.4965423345565796, 'eval_mae': 1.1607612371444702, 'eval_runtime': 4.5927, 'eval_samples_per_second': 214.471, 'eval_steps_per_second': 26.999, 'epoch': 42.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-41496
Configuration saved in our_pdbbindfull_model/checkpoint-41496/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-41496/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-40508] due to args.save_total_limit
{'loss': 0.347, 'learning_rate': 1.59919028340081e-06, 'epoch': 42.0}
{'loss': 0.3173, 'learning_rate': 1.4979757085020244e-06, 'epoch': 42.51}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2095425128936768, 'eval_mse': 2.2095425128936768, 'eval_rmse': 1.4864530563354492, 'eval_mae': 1.151437520980835, 'eval_runtime': 4.6099, 'eval_samples_per_second': 213.67, 'eval_steps_per_second': 26.899, 'epoch': 43.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-42484
Configuration saved in our_pdbbindfull_model/checkpoint-42484/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-42484/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-41496] due to args.save_total_limit
{'loss': 0.3691, 'learning_rate': 1.3967611336032391e-06, 'epoch': 43.02}
{'loss': 0.335, 'learning_rate': 1.2955465587044536e-06, 'epoch': 43.52}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.1894824504852295, 'eval_mse': 2.1894824504852295, 'eval_rmse': 1.4796899557113647, 'eval_mae': 1.146257758140564, 'eval_runtime': 4.6282, 'eval_samples_per_second': 212.828, 'eval_steps_per_second': 26.793, 'epoch': 44.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-43472
Configuration saved in our_pdbbindfull_model/checkpoint-43472/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-43472/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-42484] due to args.save_total_limit
{'loss': 0.3441, 'learning_rate': 1.1943319838056681e-06, 'epoch': 44.03}
{'loss': 0.3443, 'learning_rate': 1.0931174089068828e-06, 'epoch': 44.53}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2246272563934326, 'eval_mse': 2.2246272563934326, 'eval_rmse': 1.491518497467041, 'eval_mae': 1.1584477424621582, 'eval_runtime': 4.7245, 'eval_samples_per_second': 208.487, 'eval_steps_per_second': 26.246, 'epoch': 45.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-44460
Configuration saved in our_pdbbindfull_model/checkpoint-44460/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-44460/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-43472] due to args.save_total_limit
{'loss': 0.3242, 'learning_rate': 9.919028340080973e-07, 'epoch': 45.04}
{'loss': 0.3604, 'learning_rate': 8.906882591093118e-07, 'epoch': 45.55}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2161009311676025, 'eval_mse': 2.2161009311676025, 'eval_rmse': 1.4886574745178223, 'eval_mae': 1.1535395383834839, 'eval_runtime': 4.6251, 'eval_samples_per_second': 212.97, 'eval_steps_per_second': 26.81, 'epoch': 46.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-45448
Configuration saved in our_pdbbindfull_model/checkpoint-45448/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-45448/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-44460] due to args.save_total_limit
{'loss': 0.3321, 'learning_rate': 7.894736842105263e-07, 'epoch': 46.05}
{'loss': 0.3422, 'learning_rate': 6.882591093117409e-07, 'epoch': 46.56}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2106308937072754, 'eval_mse': 2.2106311321258545, 'eval_rmse': 1.4868191480636597, 'eval_mae': 1.1536109447479248, 'eval_runtime': 4.7086, 'eval_samples_per_second': 209.193, 'eval_steps_per_second': 26.335, 'epoch': 47.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-46436
Configuration saved in our_pdbbindfull_model/checkpoint-46436/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-46436/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-45448] due to args.save_total_limit
{'loss': 0.3223, 'learning_rate': 5.870445344129555e-07, 'epoch': 47.06}
{'loss': 0.3253, 'learning_rate': 4.858299595141701e-07, 'epoch': 47.57}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2084391117095947, 'eval_mse': 2.2084388732910156, 'eval_rmse': 1.4860817193984985, 'eval_mae': 1.1512514352798462, 'eval_runtime': 4.6784, 'eval_samples_per_second': 210.541, 'eval_steps_per_second': 26.505, 'epoch': 48.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-47424
Configuration saved in our_pdbbindfull_model/checkpoint-47424/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-47424/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-46436] due to args.save_total_limit
{'loss': 0.3418, 'learning_rate': 3.846153846153847e-07, 'epoch': 48.08}
{'loss': 0.3309, 'learning_rate': 2.834008097165992e-07, 'epoch': 48.58}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2103824615478516, 'eval_mse': 2.2103824615478516, 'eval_rmse': 1.486735463142395, 'eval_mae': 1.1521849632263184, 'eval_runtime': 3.2286, 'eval_samples_per_second': 305.083, 'eval_steps_per_second': 38.406, 'epoch': 49.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-48412
Configuration saved in our_pdbbindfull_model/checkpoint-48412/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-48412/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-47424] due to args.save_total_limit
{'loss': 0.3491, 'learning_rate': 1.8218623481781377e-07, 'epoch': 49.09}
{'loss': 0.3148, 'learning_rate': 8.097165991902835e-08, 'epoch': 49.6}
***** Running Evaluation *****
  Num examples = 985
  Batch size = 8
{'eval_loss': 2.2166781425476074, 'eval_mse': 2.2166783809661865, 'eval_rmse': 1.4888513088226318, 'eval_mae': 1.1540616750717163, 'eval_runtime': 4.5752, 'eval_samples_per_second': 215.291, 'eval_steps_per_second': 27.103, 'epoch': 50.0}
Saving model checkpoint to our_pdbbindfull_model/checkpoint-49400
Configuration saved in our_pdbbindfull_model/checkpoint-49400/config.json
Model weights saved in our_pdbbindfull_model/checkpoint-49400/pytorch_model.bin
Deleting older checkpoint [our_pdbbindfull_model/checkpoint-48412] due to args.save_total_limit


Training completed. Do not forget to share your model on huggingface.co/models =)


{'train_runtime': 6118.0305, 'train_samples_per_second': 64.547, 'train_steps_per_second': 8.074, 'train_loss': 0.5750620862829541, 'epoch': 50.0}
Metrics
TrainOutput(global_step=49400, training_loss=0.5750620862829541, metrics={'train_runtime': 6118.0305, 'train_samples_per_second': 64.547, 'train_steps_per_second': 8.074, 'train_loss': 0.5750620862829541, 'epoch': 50.0})
Saving model checkpoint to our_pdbbindfull_model
Configuration saved in our_pdbbindfull_model/config.json
Model weights saved in our_pdbbindfull_model/pytorch_model.bin
***** Running Prediction *****
  Num examples = 988
  Batch size = 8

Mean Squared Error (MSE): 2.2154239492054857
Root Mean Square Error (RMSE): 1.488430028320272
Mean Absolute Error (MAE): 1.1369174286276704