GleghornLab/optimal_ph_rigor_DPLM2-3B_2026-04-27-19-40_RTHS

About Protify

Why Protify?

• Benchmark multiple models efficiently.
• Flexible for all skill levels.
• Accessible computing with support for precomputed embeddings.
• Cost-effective workflows for training and evaluation.

Training Run

• dataset: optimal_ph_rigor
• model: DPLM2-3B
• run_id: 2026-04-27-19-40_RTHS
• task_type: regression
• num_runs: 1

Dataset Statistics

• train_size: 8015
• valid_size: 430
• test_size: 439

Validation Metrics

• balanced_bin_borders: [5.0, 9.0]
• balanced_binned_f1_mean: 0.312960
• balanced_binned_f1_per_bin: [0.43742, 0.50147, 0.0]
• balanced_binned_mcc: 0.247530
• balanced_binned_roc_auc_mean: 0.566050
• balanced_binned_roc_auc_per_bin: [0.63725, 0.5609, 0.5]
• balanced_n_bins: 3
• balanced_n_resamples: 100
• balanced_weighted_pearson_rho: 0.653130
• balanced_weighted_pearson_rho_std: 0.025310
• balanced_weighted_r_squared: 0.409610
• balanced_weighted_rmse: 1.715870
• balanced_weighted_spearman_rho: 0.639160
• balanced_weighted_spearman_rho_std: 0.031450
• epoch: 16.000000
• eval_balanced_bin_borders: [5.0, 9.0]
• eval_balanced_binned_f1_mean: 0.312960
• eval_balanced_binned_f1_per_bin: [0.43742, 0.50147, 0.0]
• eval_balanced_binned_mcc: 0.247530
• eval_balanced_binned_roc_auc_mean: 0.566050
• eval_balanced_binned_roc_auc_per_bin: [0.63725, 0.5609, 0.5]
• eval_balanced_n_bins: 3
• eval_balanced_n_resamples: 100
• eval_balanced_weighted_pearson_rho: 0.653130
• eval_balanced_weighted_pearson_rho_std: 0.025310
• eval_balanced_weighted_r_squared: 0.409610
• eval_balanced_weighted_rmse: 1.715870
• eval_balanced_weighted_spearman_rho: 0.639160
• eval_balanced_weighted_spearman_rho_std: 0.031450
• eval_loss: 1.121172
• eval_mae: 0.805240
• eval_model_preparation_time: 0.002100
• eval_mse: 1.121170
• eval_pear_pval: 0.000000
• eval_pearson_rho: 0.557960
• eval_r_squared: 0.290330
• eval_rmse: 1.058850
• eval_runtime: 0.793400
• eval_samples_per_second: 541.939000
• eval_spear_pval: 0.000000
• eval_spearman_rho: 0.504080
• eval_steps_per_second: 6.302000

Test Metrics

• balanced_bin_borders: [5.0, 9.0]
• balanced_binned_f1_mean: 0.260300
• balanced_binned_f1_per_bin: [0.32703, 0.45387, 0.0]
• balanced_binned_mcc: 0.183320
• balanced_binned_roc_auc_mean: 0.543310
• balanced_binned_roc_auc_per_bin: [0.59296, 0.53697, 0.5]
• balanced_n_bins: 3
• balanced_n_resamples: 100
• balanced_weighted_pearson_rho: 0.554600
• balanced_weighted_pearson_rho_std: 0.025790
• balanced_weighted_r_squared: 0.287740
• balanced_weighted_rmse: 1.932500
• balanced_weighted_spearman_rho: 0.593750
• balanced_weighted_spearman_rho_std: 0.026270
• test_loss: 1.297019
• test_mae: 0.850050
• test_model_preparation_time: 0.002100
• test_mse: 1.297020
• test_pear_pval: 0.000000
• test_pearson_rho: 0.525320
• test_r_squared: 0.248250
• test_rmse: 1.138870
• test_runtime: 0.903100
• test_samples_per_second: 486.111000
• test_spear_pval: 0.000000
• test_spearman_rho: 0.495680
• test_steps_per_second: 5.537000
• training_time_seconds: 228.701500