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/A1/seed_2026/model_best.pt
BIN
View File
Binary file not shown.
final/experiments/q3/ati_ho/models/A1/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
A1,2026,1,1.0153538673012346,0.9397767204177248,0.9357580549114353,0.001,0.0
A1,2026,2,0.8344417214393616,0.8647092963968004,0.8555307774753361,0.001,0.0
A1,2026,3,0.7413759132226309,0.8561831976358707,0.8470297724336058,0.001,0.0
A1,2026,4,0.6923596881054066,0.8557776766163963,0.8481834858328432,0.001,0.0
A1,2026,5,0.6530446432254933,0.8681209758742825,0.8606502813297313,0.001,0.0
A1,2026,6,0.5979297994463532,0.9125843056283154,0.9086607209928743,0.001,0.0
A1,2026,7,0.5576094913261908,0.9276157835355172,0.9214578185762677,0.001,0.0
1 method seed epoch train_loss valid_selection_loss valid_clean_loss lambda_interaction lambda_mask
2 A1 2026 1 1.0153538673012346 0.9397767204177248 0.9357580549114353 0.001 0.0
3 A1 2026 2 0.8344417214393616 0.8647092963968004 0.8555307774753361 0.001 0.0
4 A1 2026 3 0.7413759132226309 0.8561831976358707 0.8470297724336058 0.001 0.0
5 A1 2026 4 0.6923596881054066 0.8557776766163963 0.8481834858328432 0.001 0.0
6 A1 2026 5 0.6530446432254933 0.8681209758742825 0.8606502813297313 0.001 0.0
7 A1 2026 6 0.5979297994463532 0.9125843056283154 0.9086607209928743 0.001 0.0
8 A1 2026 7 0.5576094913261908 0.9276157835355172 0.9214578185762677 0.001 0.0
final/experiments/q3/ati_ho/models/A1/seed_2026/training_manifest.json
+137
View File
@@ -0,0 +1,137 @@
{
"method": "A1",
"seed": 2026,
"best_epoch": 4,
"best_selection_loss": 0.8557776766163963,
"elapsed_seconds": 7.712336958968081,
"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": "A1_low_rank_pairs",
"low_rank": true,
"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": "A1",
"seed": 2026,
"epoch": 1,
"train_loss": 1.0153538673012346,
"valid_selection_loss": 0.9397767204177248,
"valid_clean_loss": 0.9357580549114353,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 2026,
"epoch": 2,
"train_loss": 0.8344417214393616,
"valid_selection_loss": 0.8647092963968004,
"valid_clean_loss": 0.8555307774753361,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 2026,
"epoch": 3,
"train_loss": 0.7413759132226309,
"valid_selection_loss": 0.8561831976358707,
"valid_clean_loss": 0.8470297724336058,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 2026,
"epoch": 4,
"train_loss": 0.6923596881054066,
"valid_selection_loss": 0.8557776766163963,
"valid_clean_loss": 0.8481834858328432,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 2026,
"epoch": 5,
"train_loss": 0.6530446432254933,
"valid_selection_loss": 0.8681209758742825,
"valid_clean_loss": 0.8606502813297313,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 2026,
"epoch": 6,
"train_loss": 0.5979297994463532,
"valid_selection_loss": 0.9125843056283154,
"valid_clean_loss": 0.9086607209928743,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 2026,
"epoch": 7,
"train_loss": 0.5576094913261908,
"valid_selection_loss": 0.9276157835355172,
"valid_clean_loss": 0.9214578185762677,
"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/A1/seed_3407/model_best.pt
BIN
View File
Binary file not shown.
final/experiments/q3/ati_ho/models/A1/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
A1,3407,1,0.993991059285623,0.9350224992076119,0.9297198895569686,0.001,0.0
A1,3407,2,0.8231704179887418,0.8853522272227885,0.8782615517521952,0.001,0.0
A1,3407,3,0.7412442289016865,0.8810990981675766,0.872995504966149,0.001,0.0
A1,3407,4,0.6908998235508248,0.8885695379186462,0.8816041920211289,0.001,0.0
A1,3407,5,0.6549228540173283,0.9160289600655273,0.9081327443594461,0.001,0.0
A1,3407,6,0.6015714597370889,0.9299172356233493,0.9222097593349415,0.001,0.0
1 method seed epoch train_loss valid_selection_loss valid_clean_loss lambda_interaction lambda_mask
2 A1 3407 1 0.993991059285623 0.9350224992076119 0.9297198895569686 0.001 0.0
3 A1 3407 2 0.8231704179887418 0.8853522272227885 0.8782615517521952 0.001 0.0
4 A1 3407 3 0.7412442289016865 0.8810990981675766 0.872995504966149 0.001 0.0
5 A1 3407 4 0.6908998235508248 0.8885695379186462 0.8816041920211289 0.001 0.0
6 A1 3407 5 0.6549228540173283 0.9160289600655273 0.9081327443594461 0.001 0.0
7 A1 3407 6 0.6015714597370889 0.9299172356233493 0.9222097593349415 0.001 0.0
final/experiments/q3/ati_ho/models/A1/seed_3407/training_manifest.json
+127
View File
@@ -0,0 +1,127 @@
{
"method": "A1",
"seed": 3407,
"best_epoch": 3,
"best_selection_loss": 0.8810990981675766,
"elapsed_seconds": 6.539086283999495,
"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": "A1_low_rank_pairs",
"low_rank": true,
"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": "A1",
"seed": 3407,
"epoch": 1,
"train_loss": 0.993991059285623,
"valid_selection_loss": 0.9350224992076119,
"valid_clean_loss": 0.9297198895569686,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 3407,
"epoch": 2,
"train_loss": 0.8231704179887418,
"valid_selection_loss": 0.8853522272227885,
"valid_clean_loss": 0.8782615517521952,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 3407,
"epoch": 3,
"train_loss": 0.7412442289016865,
"valid_selection_loss": 0.8810990981675766,
"valid_clean_loss": 0.872995504966149,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 3407,
"epoch": 4,
"train_loss": 0.6908998235508248,
"valid_selection_loss": 0.8885695379186462,
"valid_clean_loss": 0.8816041920211289,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 3407,
"epoch": 5,
"train_loss": 0.6549228540173283,
"valid_selection_loss": 0.9160289600655273,
"valid_clean_loss": 0.9081327443594461,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 3407,
"epoch": 6,
"train_loss": 0.6015714597370889,
"valid_selection_loss": 0.9299172356233493,
"valid_clean_loss": 0.9222097593349415,
"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/A1/seed_42/model_best.pt
BIN
View File
Binary file not shown.
final/experiments/q3/ati_ho/models/A1/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
A1,42,1,0.9995549983448453,0.9426855438358182,0.9384826362788022,0.001,0.0
A1,42,2,0.8219450292763887,0.8703031135456902,0.8586705464583176,0.001,0.0
A1,42,3,0.7375812182823817,0.8694013199963413,0.8590540899025215,0.001,0.0
A1,42,4,0.6825026941520197,0.8649765662439577,0.8543586613057734,0.001,0.0
A1,42,5,0.648727031217681,0.877989024087623,0.8682481325589694,0.001,0.0
A1,42,6,0.6008665031856961,0.9173483715935067,0.9082090225848523,0.001,0.0
A1,42,7,0.5538152815015228,0.9287053037148255,0.9185833616571112,0.001,0.0
1 method seed epoch train_loss valid_selection_loss valid_clean_loss lambda_interaction lambda_mask
2 A1 42 1 0.9995549983448453 0.9426855438358182 0.9384826362788022 0.001 0.0
3 A1 42 2 0.8219450292763887 0.8703031135456902 0.8586705464583176 0.001 0.0
4 A1 42 3 0.7375812182823817 0.8694013199963413 0.8590540899025215 0.001 0.0
5 A1 42 4 0.6825026941520197 0.8649765662439577 0.8543586613057734 0.001 0.0
6 A1 42 5 0.648727031217681 0.877989024087623 0.8682481325589694 0.001 0.0
7 A1 42 6 0.6008665031856961 0.9173483715935067 0.9082090225848523 0.001 0.0
8 A1 42 7 0.5538152815015228 0.9287053037148255 0.9185833616571112 0.001 0.0
final/experiments/q3/ati_ho/models/A1/seed_42/training_manifest.json
+137
View File
@@ -0,0 +1,137 @@
{
"method": "A1",
"seed": 42,
"best_epoch": 4,
"best_selection_loss": 0.8649765662439577,
"elapsed_seconds": 7.483738042006735,
"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": "A1_low_rank_pairs",
"low_rank": true,
"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": "A1",
"seed": 42,
"epoch": 1,
"train_loss": 0.9995549983448453,
"valid_selection_loss": 0.9426855438358182,
"valid_clean_loss": 0.9384826362788022,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 42,
"epoch": 2,
"train_loss": 0.8219450292763887,
"valid_selection_loss": 0.8703031135456902,
"valid_clean_loss": 0.8586705464583176,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 42,
"epoch": 3,
"train_loss": 0.7375812182823817,
"valid_selection_loss": 0.8694013199963413,
"valid_clean_loss": 0.8590540899025215,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 42,
"epoch": 4,
"train_loss": 0.6825026941520197,
"valid_selection_loss": 0.8649765662439577,
"valid_clean_loss": 0.8543586613057734,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 42,
"epoch": 5,
"train_loss": 0.648727031217681,
"valid_selection_loss": 0.877989024087623,
"valid_clean_loss": 0.8682481325589694,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 42,
"epoch": 6,
"train_loss": 0.6008665031856961,
"valid_selection_loss": 0.9173483715935067,
"valid_clean_loss": 0.9082090225848523,
"lambda_interaction": 0.001,
"lambda_mask": 0.0
},
{
"method": "A1",
"seed": 42,
"epoch": 7,
"train_loss": 0.5538152815015228,
"valid_selection_loss": 0.9287053037148255,
"valid_clean_loss": 0.9185833616571112,
"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
}
}