gloamxun/modeling_zhaocui
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Train ATI-HO and finalize project outputs

Browse Source
This commit is contained in:
gloamxun
2026-09-26 16:05:44 +08:00
parent a86560da64
commit 9cdd604117
358 changed files with 10540 additions and 173 deletions
Download Patch File Download Diff File Expand all files Collapse all files
final/experiments/q3/ati_ho/models/A0/seed_2026/model_best.pt
BIN
View File
Binary file not shown.
final/experiments/q3/ati_ho/models/A0/seed_2026/training_history.csv
+8
View File
@@ -0,0 +1,8 @@
method,seed,epoch,train_loss,valid_selection_loss,valid_clean_loss,lambda_interaction,lambda_mask
A0,2026,1,1.0065954625606537,0.9363003446833118,0.931940136375008,0.001,0.0
A0,2026,2,0.8319815562831031,0.8692282438278198,0.8585401066056975,0.001,0.0
A0,2026,3,0.7417113295307866,0.8601733071135951,0.8481150000959963,0.001,0.0
A0,2026,4,0.6949168112542894,0.8574190991265433,0.8474381146850166,0.001,0.0
A0,2026,5,0.6569236450725131,0.8673936297277827,0.8559725572774698,0.001,0.0
A0,2026,6,0.6017359914603057,0.9111922525770062,0.9027906433566586,0.001,0.0
A0,2026,7,0.5584396239784029,0.9282659668843826,0.9144766841615949,0.001,0.0
1 method seed epoch train_loss valid_selection_loss valid_clean_loss lambda_interaction lambda_mask
2 A0 2026 1 1.0065954625606537 0.9363003446833118 0.931940136375008 0.001 0.0
3 A0 2026 2 0.8319815562831031 0.8692282438278198 0.8585401066056975 0.001 0.0
4 A0 2026 3 0.7417113295307866 0.8601733071135951 0.8481150000959963 0.001 0.0
5 A0 2026 4 0.6949168112542894 0.8574190991265433 0.8474381146850166 0.001 0.0
6 A0 2026 5 0.6569236450725131 0.8673936297277827 0.8559725572774698 0.001 0.0
7 A0 2026 6 0.6017359914603057 0.9111922525770062 0.9027906433566586 0.001 0.0
8 A0 2026 7 0.5584396239784029 0.9282659668843826 0.9144766841615949 0.001 0.0
final/experiments/q3/ati_ho/models/A0/seed_2026/training_manifest.json
+137
View File
@@ -0,0 +1,137 @@
{
"method": "A0",
"seed": 2026,
"best_epoch": 4,
"best_selection_loss": 0.8574190991265433,
"elapsed_seconds": 6.77071833203081,
"batch_size": 64,
"epoch_limit": 12,
"patience": 3,
"optimizer": "AdamW",
"learning_rate": 0.0003,
"weight_decay": 0.001,
"gradient_clip_norm": 1.0,
"training_mask_rates": [
0.0,
0.1,
0.3,
0.5,
0.7
],
"training_mask_patterns": [
"single",
"sync",
"partial",
"async"
],
"training_mask_seed_base": 20261227,
"same_orders_and_masks_across_methods_for_same_seed": true,
"config": {
"name": "A0_main_effects",
"low_rank": false,
"cross_attention": false,
"anchored": true,
"lambda_interaction": 0.001,
"lambda_mask": 0.0,
"rank": 4,
"hidden": 64,
"gru_hidden_per_direction": 32,
"attention_heads": 4,
"attention_ffn": 128,
"eta_init": 0.1
},
"history": [
{
"method": "A0",
"seed": 2026,
"epoch": 1,
"train_loss": 1.0065954625606537,
"valid_selection_loss": 0.9363003446833118,
"valid_clean_loss": 0.931940136375008,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 2026,
"epoch": 2,
"train_loss": 0.8319815562831031,
"valid_selection_loss": 0.8692282438278198,
"valid_clean_loss": 0.8585401066056975,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 2026,
"epoch": 3,
"train_loss": 0.7417113295307866,
"valid_selection_loss": 0.8601733071135951,
"valid_clean_loss": 0.8481150000959963,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 2026,
"epoch": 4,
"train_loss": 0.6949168112542894,
"valid_selection_loss": 0.8574190991265433,
"valid_clean_loss": 0.8474381146850166,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 2026,
"epoch": 5,
"train_loss": 0.6569236450725131,
"valid_selection_loss": 0.8673936297277827,
"valid_clean_loss": 0.8559725572774698,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 2026,
"epoch": 6,
"train_loss": 0.6017359914603057,
"valid_selection_loss": 0.9111922525770062,
"valid_clean_loss": 0.9027906433566586,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 2026,
"epoch": 7,
"train_loss": 0.5584396239784029,
"valid_selection_loss": 0.9282659668843826,
"valid_clean_loss": 0.9144766841615949,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
}
],
"mask_counts": {
"0.1/async": 1149,
"0.0/async": 1181,
"0.5/sync": 1205,
"0.0/sync": 1177,
"0.5/single": 1154,
"0.0/single": 1223,
"0.7/async": 1223,
"0.0/partial": 1254,
"0.7/sync": 1182,
"0.5/partial": 1182,
"0.3/partial": 1153,
"0.1/partial": 1216,
"0.3/single": 1178,
"0.3/async": 1180,
"0.5/async": 1182,
"0.1/single": 1208,
"0.7/single": 1129,
"0.3/sync": 1220,
"0.1/sync": 1187,
"0.7/partial": 1182
}
}
final/experiments/q3/ati_ho/models/A0/seed_3407/model_best.pt
BIN
View File
Binary file not shown.
final/experiments/q3/ati_ho/models/A0/seed_3407/training_history.csv
+7
View File
@@ -0,0 +1,7 @@
method,seed,epoch,train_loss,valid_selection_loss,valid_clean_loss,lambda_interaction,lambda_mask
A0,3407,1,1.0028652648131053,0.9436100423336029,0.9391928321712619,0.001,0.0
A0,3407,2,0.8238278020311285,0.8820641582811272,0.8730134918139532,0.001,0.0
A0,3407,3,0.74309202587163,0.8756466648735842,0.8646609815922413,0.001,0.0
A0,3407,4,0.6922215135009201,0.8815795684253777,0.8710692741058685,0.001,0.0
A0,3407,5,0.6518548451088093,0.9090294297579881,0.8968785254509895,0.001,0.0
A0,3407,6,0.6016097575150154,0.9234253629878326,0.9113659098908141,0.001,0.0
1 method seed epoch train_loss valid_selection_loss valid_clean_loss lambda_interaction lambda_mask
2 A0 3407 1 1.0028652648131053 0.9436100423336029 0.9391928321712619 0.001 0.0
3 A0 3407 2 0.8238278020311285 0.8820641582811272 0.8730134918139532 0.001 0.0
4 A0 3407 3 0.74309202587163 0.8756466648735842 0.8646609815922413 0.001 0.0
5 A0 3407 4 0.6922215135009201 0.8815795684253777 0.8710692741058685 0.001 0.0
6 A0 3407 5 0.6518548451088093 0.9090294297579881 0.8968785254509895 0.001 0.0
7 A0 3407 6 0.6016097575150154 0.9234253629878326 0.9113659098908141 0.001 0.0
final/experiments/q3/ati_ho/models/A0/seed_3407/training_manifest.json
+127
View File
@@ -0,0 +1,127 @@
{
"method": "A0",
"seed": 3407,
"best_epoch": 3,
"best_selection_loss": 0.8756466648735842,
"elapsed_seconds": 5.952770586998668,
"batch_size": 64,
"epoch_limit": 12,
"patience": 3,
"optimizer": "AdamW",
"learning_rate": 0.0003,
"weight_decay": 0.001,
"gradient_clip_norm": 1.0,
"training_mask_rates": [
0.0,
0.1,
0.3,
0.5,
0.7
],
"training_mask_patterns": [
"single",
"sync",
"partial",
"async"
],
"training_mask_seed_base": 20261227,
"same_orders_and_masks_across_methods_for_same_seed": true,
"config": {
"name": "A0_main_effects",
"low_rank": false,
"cross_attention": false,
"anchored": true,
"lambda_interaction": 0.001,
"lambda_mask": 0.0,
"rank": 4,
"hidden": 64,
"gru_hidden_per_direction": 32,
"attention_heads": 4,
"attention_ffn": 128,
"eta_init": 0.1
},
"history": [
{
"method": "A0",
"seed": 3407,
"epoch": 1,
"train_loss": 1.0028652648131053,
"valid_selection_loss": 0.9436100423336029,
"valid_clean_loss": 0.9391928321712619,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 3407,
"epoch": 2,
"train_loss": 0.8238278020311285,
"valid_selection_loss": 0.8820641582811272,
"valid_clean_loss": 0.8730134918139532,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 3407,
"epoch": 3,
"train_loss": 0.74309202587163,
"valid_selection_loss": 0.8756466648735842,
"valid_clean_loss": 0.8646609815922413,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 3407,
"epoch": 4,
"train_loss": 0.6922215135009201,
"valid_selection_loss": 0.8815795684253777,
"valid_clean_loss": 0.8710692741058685,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 3407,
"epoch": 5,
"train_loss": 0.6518548451088093,
"valid_selection_loss": 0.9090294297579881,
"valid_clean_loss": 0.8968785254509895,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 3407,
"epoch": 6,
"train_loss": 0.6016097575150154,
"valid_selection_loss": 0.9234253629878326,
"valid_clean_loss": 0.9113659098908141,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
}
],
"mask_counts": {
"0.7/async": 1003,
"0.0/async": 1014,
"0.5/sync": 1033,
"0.1/sync": 1024,
"0.5/partial": 996,
"0.3/partial": 1019,
"0.5/async": 1053,
"0.3/single": 1003,
"0.0/sync": 1042,
"0.7/partial": 1079,
"0.7/sync": 986,
"0.0/single": 1029,
"0.1/async": 959,
"0.3/sync": 1055,
"0.0/partial": 1016,
"0.1/single": 1011,
"0.1/partial": 988,
"0.5/single": 1076,
"0.7/single": 997,
"0.3/async": 987
}
}
final/experiments/q3/ati_ho/models/A0/seed_42/model_best.pt
BIN
View File
Binary file not shown.
final/experiments/q3/ati_ho/models/A0/seed_42/training_history.csv
+8
View File
@@ -0,0 +1,8 @@
method,seed,epoch,train_loss,valid_selection_loss,valid_clean_loss,lambda_interaction,lambda_mask
A0,42,1,0.9974622616061458,0.9360404843157465,0.9310548764008743,0.001,0.0
A0,42,2,0.8197665159349088,0.8663458909307207,0.8535200449136587,0.001,0.0
A0,42,3,0.7383816402267527,0.8669271827726575,0.8542619085573888,0.001,0.0
A0,42,4,0.6841275570569215,0.8643234267339601,0.850496660222064,0.001,0.0
A0,42,5,0.6515898009141287,0.8761576001460736,0.8627192581092918,0.001,0.0
A0,42,6,0.6044590015102316,0.9111036181777388,0.8968614366028335,0.001,0.0
A0,42,7,0.5584419562860772,0.9260857973124955,0.9113180768358838,0.001,0.0
1 method seed epoch train_loss valid_selection_loss valid_clean_loss lambda_interaction lambda_mask
2 A0 42 1 0.9974622616061458 0.9360404843157465 0.9310548764008743 0.001 0.0
3 A0 42 2 0.8197665159349088 0.8663458909307207 0.8535200449136587 0.001 0.0
4 A0 42 3 0.7383816402267527 0.8669271827726575 0.8542619085573888 0.001 0.0
5 A0 42 4 0.6841275570569215 0.8643234267339601 0.850496660222064 0.001 0.0
6 A0 42 5 0.6515898009141287 0.8761576001460736 0.8627192581092918 0.001 0.0
7 A0 42 6 0.6044590015102316 0.9111036181777388 0.8968614366028335 0.001 0.0
8 A0 42 7 0.5584419562860772 0.9260857973124955 0.9113180768358838 0.001 0.0
final/experiments/q3/ati_ho/models/A0/seed_42/training_manifest.json
+137
View File
@@ -0,0 +1,137 @@
{
"method": "A0",
"seed": 42,
"best_epoch": 4,
"best_selection_loss": 0.8643234267339601,
"elapsed_seconds": 12.891749234986492,
"batch_size": 64,
"epoch_limit": 12,
"patience": 3,
"optimizer": "AdamW",
"learning_rate": 0.0003,
"weight_decay": 0.001,
"gradient_clip_norm": 1.0,
"training_mask_rates": [
0.0,
0.1,
0.3,
0.5,
0.7
],
"training_mask_patterns": [
"single",
"sync",
"partial",
"async"
],
"training_mask_seed_base": 20261227,
"same_orders_and_masks_across_methods_for_same_seed": true,
"config": {
"name": "A0_main_effects",
"low_rank": false,
"cross_attention": false,
"anchored": true,
"lambda_interaction": 0.001,
"lambda_mask": 0.0,
"rank": 4,
"hidden": 64,
"gru_hidden_per_direction": 32,
"attention_heads": 4,
"attention_ffn": 128,
"eta_init": 0.1
},
"history": [
{
"method": "A0",
"seed": 42,
"epoch": 1,
"train_loss": 0.9974622616061458,
"valid_selection_loss": 0.9360404843157465,
"valid_clean_loss": 0.9310548764008743,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 42,
"epoch": 2,
"train_loss": 0.8197665159349088,
"valid_selection_loss": 0.8663458909307207,
"valid_clean_loss": 0.8535200449136587,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 42,
"epoch": 3,
"train_loss": 0.7383816402267527,
"valid_selection_loss": 0.8669271827726575,
"valid_clean_loss": 0.8542619085573888,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 42,
"epoch": 4,
"train_loss": 0.6841275570569215,
"valid_selection_loss": 0.8643234267339601,
"valid_clean_loss": 0.850496660222064,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 42,
"epoch": 5,
"train_loss": 0.6515898009141287,
"valid_selection_loss": 0.8761576001460736,
"valid_clean_loss": 0.8627192581092918,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 42,
"epoch": 6,
"train_loss": 0.6044590015102316,
"valid_selection_loss": 0.9111036181777388,
"valid_clean_loss": 0.8968614366028335,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A0",
"seed": 42,
"epoch": 7,
"train_loss": 0.5584419562860772,
"valid_selection_loss": 0.9260857973124955,
"valid_clean_loss": 0.9113180768358838,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
}
],
"mask_counts": {
"0.0/single": 1169,
"0.3/single": 1187,
"0.5/single": 1158,
"0.7/sync": 1272,
"0.7/partial": 1207,
"0.0/async": 1145,
"0.3/async": 1200,
"0.3/partial": 1209,
"0.3/sync": 1170,
"0.1/single": 1205,
"0.5/async": 1185,
"0.7/single": 1232,
"0.5/sync": 1143,
"0.5/partial": 1236,
"0.0/partial": 1183,
"0.0/sync": 1159,
"0.7/async": 1179,
"0.1/sync": 1202,
"0.1/async": 1192,
"0.1/partial": 1132
}
}