Add files using upload-large-folder tool

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7245339299030574,0.8523489832878113,0.7760616773552657,0.7859297620265

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7531692766592095,0.8523489832878113,0.7760616773552657,0.8082589513973091

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7530201342281879,0.8523489832878113,0.7760616773552657,0.7910938892378713

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7881431767337808,0.8523489832878113,0.7760616773552657,0.8083466027785488

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0004/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7777315007635708,0.8293384313583374,0.7494576865623174,0.7369909722673248

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0004/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7720394280161045,0.8293384313583374,0.7494576865623174,0.7546567513151913

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0004/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7734971539636263,0.8293384313583374,0.7494576865623174,0.7493635123725401

revision_runs/heldout_v3/trainval/runs/base_2/bce_lr0.0004/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only,100.0,0.0,0.0,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7712064417603777,0.8293384313583374,0.7494576865623174,0.7545463401961421

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep30_wd0.0001/model_roberta/dataset_cola/head_seed_1/LoraConfig_cola_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,cola,train,True,2,0.8293384313583374,0.7663473552686381,0.7461745794635318,0.0003,30,1.0,400,0.0001

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep30_wd0.0001/model_roberta/dataset_cola/head_seed_2/LoraConfig_cola_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,cola,train,True,2,0.8293384313583374,0.7726641677078995,0.7449730466974086,0.0003,30,1.0,400,0.0001

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep30_wd0.0001/model_roberta/dataset_cola/head_seed_3/LoraConfig_cola_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,cola,train,True,2,0.8293384313583374,0.7738442315701791,0.7401928947197506,0.0003,30,1.0,400,0.0001

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep30_wd0.0001/model_roberta/dataset_cola/head_seed_4/LoraConfig_cola_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,cola,train,True,2,0.8293384313583374,0.7858531167569069,0.7502273170098072,0.0003,30,1.0,400,0.0001

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep50_wd0.0001/model_electra/dataset_cola/head_seed_1/LoraConfig_cola_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
+google/electra-base-discriminator,LoraConfig,cola,train,True,2,0.8523489832878113,0.7686800894854586,0.8120936993265452,0.0003,50,1.0,400,0.0001

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep50_wd0.0001/model_electra/dataset_cola/head_seed_2/LoraConfig_cola_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
+google/electra-base-discriminator,LoraConfig,cola,train,True,2,0.8523489832878113,0.7748695003728561,0.8077403473916409,0.0003,50,1.0,400,0.0001

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep50_wd0.0001/model_electra/dataset_cola/head_seed_3/LoraConfig_cola_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
+google/electra-base-discriminator,LoraConfig,cola,train,True,2,0.8523489832878113,0.771290082028337,0.8063014038829561,0.0003,50,1.0,400,0.0001

revision_runs/heldout_v3/trainval/runs/base_2/confidnet_lr0.0003_ep50_wd0.0001/model_electra/dataset_cola/head_seed_4/LoraConfig_cola_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
+google/electra-base-discriminator,LoraConfig,cola,train,True,2,0.8523489832878113,0.7760626398210291,0.803281083371072,0.0003,50,1.0,400,0.0001

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.01/model_electra/dataset_cola/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,cola,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.01,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7299776286353468,0.8523489832878113,0.7760616773552657,0.7950820270842769

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.01/model_electra/dataset_cola/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,cola,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.01,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7464578672632364,0.8523489832878113,0.7760616773552657,0.7962507121674725

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.01/model_electra/dataset_cola/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,cola,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.01,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7051454138702461,0.8523489832878113,0.7760616773552657,0.7846953384073745

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.01/model_electra/dataset_cola/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,cola,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.01,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.722296793437733,0.8523489832878113,0.7760616773552657,0.8000562429696287

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.01/model_roberta/dataset_cola/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,cola,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.01,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.777523254199639,0.8293384313583374,0.7494576865623174,0.7545918035981034

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.01/model_roberta/dataset_cola/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,cola,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.01,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7724559211439678,0.8293384313583374,0.7494576865623174,0.753110995648503

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.01/model_roberta/dataset_cola/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,cola,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.01,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7732194918783839,0.8293384313583374,0.7494576865623174,0.7564622978502307

revision_runs/heldout_v3/trainval/runs/base_2/l2sp_lr0.0004_b0.01/model_roberta/dataset_cola/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,cola,train,2,entropy_response_finetune_l2sp_only,100.0,0.0,0.01,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7631542412883522,0.8293384313583374,0.7494576865623174,0.7492855751120349

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_roberta/dataset_cola/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,cola,train,2,entropy_response_finetune,100.0,0.0,0.01,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.777523254199639,0.8293384313583374,0.7494576865623174,0.7545918035981034

revision_runs/heldout_v3/trainval/runs/base_2/loo_noreg/model_roberta/dataset_cola/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,cola,train,2,entropy_response_finetune,100.0,0.0,0.01,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7724559211439678,0.8293384313583374,0.7494576865623174,0.753110995648503

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.5516778523489932,0.8523489832878113,0.7760616773552657,0.6338911369845003

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7484712900820284,0.8523489832878113,0.7760616773552657,0.825168363694798

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7539895600298284,0.8523489832878113,0.7760616773552657,0.8233276846887646

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7542878448918717,0.8523489832878113,0.7760616773552657,0.8211802258483922

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.6161321671525753,0.8293384313583374,0.7494576865623174,0.6397122816133014

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.6056504234346799,0.8293384313583374,0.7494576865623174,0.625956355134117

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7623906705539358,0.8293384313583374,0.7494576865623174,0.7395758914074169

revision_runs/heldout_v3/trainval/runs/base_2/randcls_lr0.0004/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_cls_random,100.0,0.0,0.0,cls_random,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.6802721088435374,0.8293384313583374,0.7494576865623174,0.6624634669091382

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7424310216256526,0.8523489832878113,0.7760616773552657,0.7725702306692184

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7255033557046979,0.8523489832878113,0.7760616773552657,0.7714453712766424

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7352721849366145,0.8523489832878113,0.7760616773552657,0.7770550596759821

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr0.0004/model_electra/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7413124533929903,0.8523489832878113,0.7760616773552657,0.7724606664426688

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr1e-05/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,1e-05,entropy,0,0.8571428656578064,0.762529501596557,0.6906150215188116,0.8293384313583374,0.7494576865623174,0.6418945249074495

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr1e-05/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,1e-05,entropy,0,0.8571428656578064,0.762529501596557,0.6077328890739969,0.8293384313583374,0.7494576865623174,0.6225401052153017

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr1e-05/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,1e-05,entropy,0,0.8571428656578064,0.762529501596557,0.6379980563654033,0.8293384313583374,0.7494576865623174,0.6538838734818471

revision_runs/heldout_v3/trainval/runs/base_2/randlin_lr1e-05/model_roberta/dataset_cola/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,cola,train,2,ablation_entropy_finetune_bce_only_linear_random,100.0,0.0,0.0,linear_random,1e-05,entropy,0,0.8571428656578064,0.762529501596557,0.6424406497292795,0.8293384313583374,0.7494576865623174,0.6462979801259986

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a0.1/model_electra/dataset_cola/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,cola,train,2,entropy_response_finetune_reg_only,100.0,0.1,0.0,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7313199105145414,0.8523489832878113,0.7760616773552657,0.8177545176982747

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a0.1/model_electra/dataset_cola/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,cola,train,2,entropy_response_finetune_reg_only,100.0,0.1,0.0,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7875466070096943,0.8523489832878113,0.7760616773552657,0.777080624662177

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a0.1/model_electra/dataset_cola/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,cola,train,2,entropy_response_finetune_reg_only,100.0,0.1,0.0,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7651006711409395,0.8523489832878113,0.7760616773552657,0.8032555183848773

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a0.1/model_electra/dataset_cola/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,cola,train,2,entropy_response_finetune_reg_only,100.0,0.1,0.0,cls_head,0.0004,entropy,0,0.8688046932220459,0.7010067114093959,0.7510812826249068,0.8523489832878113,0.7760616773552657,0.7702401647845967

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_roberta/dataset_cola/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,cola,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7797445508815771,0.8293384313583374,0.7494576865623174,0.7558517893096056

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_roberta/dataset_cola/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,cola,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7775926697209496,0.8293384313583374,0.7494576865623174,0.7535071767227381

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_roberta/dataset_cola/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,cola,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7725947521865889,0.8293384313583374,0.7494576865623174,0.7542605702409559

revision_runs/heldout_v3/trainval/runs/base_2/reg_lr0.0004_a1.0/model_roberta/dataset_cola/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,cola,train,2,entropy_response_finetune_reg_only,100.0,1.0,0.0,cls_head,0.0004,entropy,0,0.8571428656578064,0.762529501596557,0.7733583229210051,0.8293384313583374,0.7494576865623174,0.7542865493277912