Add files using upload-large-folder tool

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0001/model_electra/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0001,entropy,0,0.8467966318130493,0.7159688995215311,0.7473086124401914,0.8180851340293884,0.7450779093377136,0.7184237142487775

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0001/model_electra/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0001,entropy,0,0.8467966318130493,0.7159688995215311,0.7552631578947369,0.8180851340293884,0.7450779093377136,0.7298534589616652

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0001/model_electra/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0001,entropy,0,0.8467966318130493,0.7159688995215311,0.7332535885167464,0.8180851340293884,0.7450779093377136,0.7461045331143202

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0001/model_electra/dataset_toxigen/head_seed_4/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0001,entropy,0,0.8467966318130493,0.7159688995215311,0.7406100478468899,0.8180851340293884,0.7450779093377136,0.7398611396284381

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0001/model_roberta/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7913719943422913,0.7797872424125671,0.7452893607766375,0.7446929104797305

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0001/model_roberta/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7898986327204149,0.7797872424125671,0.7452893607766375,0.7361844316586591

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0001/model_roberta/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7913719943422913,0.7797872424125671,0.7452893607766375,0.7280845707205514

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0001/model_roberta/dataset_toxigen/head_seed_4/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7896628948609146,0.7797872424125671,0.7452893607766375,0.7442711113747356

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep50_wd0.0/model_roberta/dataset_toxigen/head_seed_1/LoraConfig_toxigen_seed_2_normalization_True_confidnet.csv

@@ -0,0 +1,2 @@
+model,adapter,dataset,train_on_dataset,normalization,seed,test-acc,valid-confidnet-roc-auc,confidnet_roc_auc,best_lr,best_epoch,best_weighting,best_hidden,best_weight_decay
+roberta-base,LoraConfig,toxigen,train,True,2,0.7797872424125671,0.8081683168316831,0.7546776861682846,0.0003,50,1.0,400,0.0

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep50_wd0.0/model_roberta/dataset_toxigen/head_seed_2/LoraConfig_toxigen_seed_2_normalization_True_confidnet.csv

@@ -0,0 +1,2 @@
+model,adapter,dataset,train_on_dataset,normalization,seed,test-acc,valid-confidnet-roc-auc,confidnet_roc_auc,best_lr,best_epoch,best_weighting,best_hidden,best_weight_decay
+roberta-base,LoraConfig,toxigen,train,True,2,0.7797872424125671,0.7865983026874117,0.7535243292405639,0.0003,50,1.0,400,0.0

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep50_wd0.0/model_roberta/dataset_toxigen/head_seed_3/LoraConfig_toxigen_seed_2_normalization_True_confidnet.csv

@@ -0,0 +1,2 @@
+model,adapter,dataset,train_on_dataset,normalization,seed,test-acc,valid-confidnet-roc-auc,confidnet_roc_auc,best_lr,best_epoch,best_weighting,best_hidden,best_weight_decay
+roberta-base,LoraConfig,toxigen,train,True,2,0.7797872424125671,0.8135313531353136,0.7452333405830054,0.0003,50,1.0,400,0.0

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep50_wd0.0/model_roberta/dataset_toxigen/head_seed_4/LoraConfig_toxigen_seed_2_normalization_True_confidnet.csv

@@ -0,0 +1,2 @@
+model,adapter,dataset,train_on_dataset,normalization,seed,test-acc,valid-confidnet-roc-auc,confidnet_roc_auc,best_lr,best_epoch,best_weighting,best_hidden,best_weight_decay
+roberta-base,LoraConfig,toxigen,train,True,2,0.7797872424125671,0.8061056105610561,0.7480673033196907,0.0003,50,1.0,400,0.0

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0001_b1.0/model_roberta/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,1.0,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7953206034889203,0.7797872424125671,0.7452893607766375,0.7443633799289533

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0001_b1.0/model_roberta/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,1.0,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7934936350777936,0.7797872424125671,0.7452893607766375,0.7494743987715101

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0001_b1.0/model_roberta/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,1.0,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7887199434229137,0.7797872424125671,0.7452893607766375,0.7362767002128767

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0001_b1.0/model_roberta/dataset_toxigen/head_seed_4/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,1.0,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7946133899104195,0.7797872424125671,0.7452893607766375,0.7447654072009017

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.1/model_electra/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.1,cls_head,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.749043062200957,0.8180851340293884,0.7450779093377136,0.7381120768979232

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.1/model_electra/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.1,cls_head,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7510167464114832,0.8180851340293884,0.7450779093377136,0.7285682780857649

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.1/model_electra/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.1,cls_head,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7476076555023924,0.8180851340293884,0.7450779093377136,0.6937619297485151

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.1/model_electra/dataset_toxigen/head_seed_4/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.1,cls_head,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7295454545454545,0.8180851340293884,0.7450779093377136,0.7450398862348763

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_electra/dataset_toxigen/head_seed_0/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune,100.0,0.0,1.0,cls_head,0.0001,entropy,0,0.8467966318130493,0.7159688995215311,0.7528708133971291,0.8180851340293884,0.7450779093377136,0.7366824082312412

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_electra/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune,100.0,0.0,1.0,cls_head,0.0001,entropy,0,0.8467966318130493,0.7159688995215311,0.747787081339713,0.8180851340293884,0.7450779093377136,0.7398003026638985

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_electra/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune,100.0,0.0,1.0,cls_head,0.0001,entropy,0,0.8467966318130493,0.7159688995215311,0.7447966507177033,0.8180851340293884,0.7450779093377136,0.7434657297774128

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_electra/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune,100.0,0.0,1.0,cls_head,0.0001,entropy,0,0.8467966318130493,0.7159688995215311,0.7359449760765551,0.8180851340293884,0.7450779093377136,0.7453440710575746

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_roberta/dataset_toxigen/head_seed_0/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune,100.0,0.0,0.1,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7940829797265441,0.7797872424125671,0.7452893607766375,0.7469139463919701

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_roberta/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune,100.0,0.0,0.1,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7907237152286658,0.7797872424125671,0.7452893607766375,0.743144116891077

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_roberta/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune,100.0,0.0,0.1,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7944955209806694,0.7797872424125671,0.7452893607766375,0.7473818797740739

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_roberta/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune,100.0,0.0,0.1,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7901343705799151,0.7797872424125671,0.7452893607766375,0.7462351134573685

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_roberta/dataset_toxigen/head_seed_4/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune,100.0,0.0,0.1,cls_head,0.0001,entropy,0,0.844011127948761,0.7931105610561056,0.7862446958981614,0.7797872424125671,0.7452893607766375,0.7281109331646137

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_electra/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7264952153110047,0.8180851340293884,0.7450779093377136,0.6469706993969535

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_electra/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7272727272727273,0.8180851340293884,0.7450779093377136,0.6309629731024571

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_electra/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7327751196172249,0.8180851340293884,0.7450779093377136,0.628864097825839

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_electra/dataset_toxigen/head_seed_4/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7229665071770335,0.8180851340293884,0.7450779093377136,0.6303089757336557

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_roberta/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.7587223008015087,0.7797872424125671,0.7452893607766375,0.715048342131799

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_roberta/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.7309052333804809,0.7797872424125671,0.7452893607766375,0.6436852060554534

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_roberta/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.6844648750589345,0.7797872424125671,0.7452893607766375,0.5929968167348795

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_roberta/dataset_toxigen/head_seed_4/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.7597241867043847,0.7797872424125671,0.7452893607766375,0.7080952475103965

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr0.0004/model_electra/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7069976076555025,0.8180851340293884,0.7450779093377136,0.645495403006867

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr0.0004/model_roberta/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.7382720414898633,0.7797872424125671,0.7452893607766375,0.6535974850228364

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr0.0004/model_roberta/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.735914662894861,0.7797872424125671,0.7452893607766375,0.6570311933619366

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr0.0004/model_roberta/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.7317892503536068,0.7797872424125671,0.7452893607766375,0.655759205435936

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr0.0004/model_roberta/dataset_toxigen/head_seed_4/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.7357967939651108,0.7797872424125671,0.7452893607766375,0.6506448912878713

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_electra/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7454545454545455,0.8180851340293884,0.7450779093377136,0.7416710393234929

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_electra/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7321770334928229,0.8180851340293884,0.7450779093377136,0.6623472421843513

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_electra/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7501196172248803,0.8180851340293884,0.7450779093377136,0.7292070662134312

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_electra/dataset_toxigen/head_seed_4/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,google/electra-base-discriminator,LoraConfig,toxigen,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.8467966318130493,0.7159688995215311,0.7468301435406699,0.8180851340293884,0.7450779093377136,0.7269180754226268

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_roberta/dataset_toxigen/head_seed_1/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.7659712399811409,0.7797872424125671,0.7452893607766375,0.7287634036551528

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_roberta/dataset_toxigen/head_seed_2/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.8003889674681753,0.7797872424125671,0.7452893607766375,0.6910387462021604

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_roberta/dataset_toxigen/head_seed_3/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.79508486562942,0.7797872424125671,0.7452893607766375,0.7431638887241238

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_roberta/dataset_toxigen/head_seed_4/best_run_metric_df.csv

@@ -0,0 +1,2 @@
+,model,adapter,dataset,train_on_dataset,seed,grid,lam,reg_alpha,l2sp_alpha,load_weights,lr,head_type,trial_number,valid-acc,valid-base-max-prob-roc-auc,valid-fine-tune-max-prob-roc-auc,test-acc,test-base-max-prob-roc-auc,test-fine-tune-max-prob-roc-auc
+0,roberta-base,LoraConfig,toxigen,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.844011127948761,0.7931105610561056,0.7909005186232909,0.7797872424125671,0.7452893607766375,0.7400597109358009