hackhackhack66666 commited on
Commit
23af585
·
verified ·
1 Parent(s): 925e4f9

Add counterfactual overfit watcher reports

Browse files
overfit_watcher/early_stop_report.jsonl ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"status": "waiting_min_epoch", "reason": "epoch=78 < min_epoch=150", "last_epoch": 78, "last_global_step": 37842, "last_train_loss": 3.3077609539031982, "check_time_utc": "2026-05-30T21:23:02+00:00", "counterfactual_only": true, "training_not_stopped": true}
2
+ {"status": "active", "reason": "counterfactual stop: overfit pattern on recent val points", "verdict": "would_stop_overfit", "last_epoch": 81, "last_global_step": 39402, "last_train_loss": 3.3337106704711914, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [60, 70, 80], "recent_train_loss": [3.491044759750366, 3.394482374191284, 3.309861421585083], "recent_val_loss": [4.061497211456299, 4.0887556076049805, 4.122718334197998], "recent_reconst_mse": [0.07266432046890259, 0.07204309105873108, 0.07246608287096024], "delta_train_loss": 0.1811833381652832, "delta_val_loss": 0.06122112274169922, "delta_reconst_mse": -0.0001982375979423523, "signals": {"overfit_val": true, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-30T21:28:39+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "low"}
3
+ {"status": "active", "reason": "counterfactual stop: overfit pattern on recent val points", "verdict": "would_stop_overfit", "last_epoch": 81, "last_global_step": 39402, "last_train_loss": 3.3337106704711914, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [60, 70, 80], "recent_train_loss": [3.491044759750366, 3.394482374191284, 3.309861421585083], "recent_val_loss": [4.061497211456299, 4.0887556076049805, 4.122718334197998], "recent_reconst_mse": [0.07266432046890259, 0.07204309105873108, 0.07246608287096024], "delta_train_loss": 0.1811833381652832, "delta_val_loss": 0.06122112274169922, "delta_reconst_mse": -0.0001982375979423523, "signals": {"overfit_val": true, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-30T21:28:39+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "low"}
4
+ {"status": "active", "reason": "counterfactual stop: overfit pattern on recent val points", "verdict": "would_stop_overfit", "last_epoch": 116, "last_global_step": 56498, "last_train_loss": 3.054319381713867, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [90, 100, 110], "recent_train_loss": [3.2307076454162598, 3.163587808609009, 3.0954320430755615], "recent_val_loss": [4.165990352630615, 4.210514545440674, 4.258655071258545], "recent_reconst_mse": [0.07257667183876038, 0.07221361249685287, 0.0730852484703064], "delta_train_loss": 0.13527560234069824, "delta_val_loss": 0.09266471862792969, "delta_reconst_mse": 0.0005085766315460205, "signals": {"overfit_val": true, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-30T22:28:39+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "low"}
5
+ {"status": "active", "reason": "counterfactual stop: overfit pattern on recent val points", "verdict": "would_stop_overfit", "last_epoch": 151, "last_global_step": 73551, "last_train_loss": 2.810685157775879, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [130, 140, 150], "recent_train_loss": [2.980478286743164, 2.930964708328247, 2.886359930038452], "recent_val_loss": [4.360464572906494, 4.4101457595825195, 4.458889961242676], "recent_reconst_mse": [0.0725029855966568, 0.07244156301021576, 0.07230886816978455], "delta_train_loss": 0.09411835670471191, "delta_val_loss": 0.09842538833618164, "delta_reconst_mse": -0.00019411742687225342, "signals": {"overfit_val": true, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-30T23:28:40+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
6
+ {"status": "active", "reason": "counterfactual stop: overfit pattern on recent val points", "verdict": "would_stop_overfit", "last_epoch": 187, "last_global_step": 90645, "last_train_loss": 2.6836342811584473, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [160, 170, 180], "recent_train_loss": [2.8381547927856445, 2.796478033065796, 2.75876522064209], "recent_val_loss": [4.506368160247803, 4.555228233337402, 4.6020002365112305], "recent_reconst_mse": [0.0723370835185051, 0.07282605767250061, 0.07308126986026764], "delta_train_loss": 0.07938957214355469, "delta_val_loss": 0.09563207626342773, "delta_reconst_mse": 0.0007441863417625427, "signals": {"overfit_val": true, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T00:28:40+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
7
+ {"status": "active", "reason": "counterfactual stop: overfit pattern on recent val points", "verdict": "would_stop_overfit", "last_epoch": 222, "last_global_step": 107708, "last_train_loss": 2.542912244796753, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [200, 210, 220], "recent_train_loss": [2.689443588256836, 2.6581196784973145, 2.625403881072998], "recent_val_loss": [4.691418170928955, 4.734548091888428, 4.77445650100708], "recent_reconst_mse": [0.07321396470069885, 0.07309829443693161, 0.07285675406455994], "delta_train_loss": 0.06403970718383789, "delta_val_loss": 0.083038330078125, "delta_reconst_mse": -0.000357210636138916, "signals": {"overfit_val": true, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T01:28:41+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
8
+ {"status": "active", "reason": "counterfactual stop: overfit pattern on recent val points", "verdict": "would_stop_overfit", "last_epoch": 257, "last_global_step": 124757, "last_train_loss": 2.515023708343506, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [230, 240, 250], "recent_train_loss": [2.5961344242095947, 2.5717735290527344, 2.550201892852783], "recent_val_loss": [4.818082809448242, 4.856466293334961, 4.892910957336426], "recent_reconst_mse": [0.07246249914169312, 0.07315405458211899, 0.07362166792154312], "delta_train_loss": 0.04593253135681152, "delta_val_loss": 0.0748281478881836, "delta_reconst_mse": 0.001159168779850006, "signals": {"overfit_val": true, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T02:28:41+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
9
+ {"status": "active", "reason": "counterfactual stop: overfit pattern on recent val points", "verdict": "would_stop_overfit", "last_epoch": 293, "last_global_step": 141825, "last_train_loss": 2.4282307624816895, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [270, 280, 290], "recent_train_loss": [2.49985933303833, 2.481078863143921, 2.4613029956817627], "recent_val_loss": [4.961987018585205, 4.994965076446533, 5.026979446411133], "recent_reconst_mse": [0.07352404296398163, 0.07357458025217056, 0.07387738674879074], "delta_train_loss": 0.03855633735656738, "delta_val_loss": 0.06499242782592773, "delta_reconst_mse": 0.00035334378480911255, "signals": {"overfit_val": true, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T03:28:42+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
10
+ {"status": "active", "reason": "counterfactual stop: overfit pattern on recent val points", "verdict": "would_stop_overfit", "last_epoch": 328, "last_global_step": 158924, "last_train_loss": 2.3348381519317627, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [300, 310, 320], "recent_train_loss": [2.4420902729034424, 2.422183036804199, 2.4052648544311523], "recent_val_loss": [5.05718469619751, 5.085014820098877, 5.114203929901123], "recent_reconst_mse": [0.07346421480178833, 0.07343598455190659, 0.0729653462767601], "delta_train_loss": 0.03682541847229004, "delta_val_loss": 0.05701923370361328, "delta_reconst_mse": -0.0004988685250282288, "signals": {"overfit_val": true, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T04:28:43+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
11
+ {"status": "active", "reason": "no counterfactual stop signal", "verdict": "continue", "last_epoch": 363, "last_global_step": 176014, "last_train_loss": 2.3974239826202393, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [340, 350, 360], "recent_train_loss": [2.374774217605591, 2.3610122203826904, 2.344843626022339], "recent_val_loss": [5.167078495025635, 5.1935296058654785, 5.215641021728516], "recent_reconst_mse": [0.0737115740776062, 0.0739370658993721, 0.07308915257453918], "delta_train_loss": 0.029930591583251953, "delta_val_loss": 0.04856252670288086, "delta_reconst_mse": -0.0006224215030670166, "signals": {"overfit_val": false, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T05:28:43+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
12
+ {"status": "active", "reason": "no counterfactual stop signal", "verdict": "continue", "last_epoch": 399, "last_global_step": 193153, "last_train_loss": 2.301030397415161, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [370, 380, 390], "recent_train_loss": [2.329850435256958, 2.3168551921844482, 2.3045976161956787], "recent_val_loss": [5.239645481109619, 5.2570414543151855, 5.283329486846924], "recent_reconst_mse": [0.07331275194883347, 0.07300858944654465, 0.07387711107730865], "delta_train_loss": 0.025252819061279297, "delta_val_loss": 0.04368400573730469, "delta_reconst_mse": 0.0005643591284751892, "signals": {"overfit_val": false, "overfit_reconst": false, "offline_plateau": false, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T06:28:44+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
13
+ {"status": "active", "reason": "counterfactual review: offline metrics plateau", "verdict": "would_review_plateau", "last_epoch": 434, "last_global_step": 210234, "last_train_loss": 2.2520642280578613, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [410, 420, 430], "recent_train_loss": [2.279146194458008, 2.2710607051849365, 2.260478973388672], "recent_val_loss": [5.323919773101807, 5.346829414367676, 5.362071514129639], "recent_reconst_mse": [0.07390313595533371, 0.07285141199827194, 0.07314528524875641], "delta_train_loss": 0.018667221069335938, "delta_val_loss": 0.03815174102783203, "delta_reconst_mse": -0.000757850706577301, "signals": {"overfit_val": false, "overfit_reconst": false, "offline_plateau": true, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T07:28:45+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
14
+ {"status": "active", "reason": "counterfactual review: offline metrics plateau", "verdict": "would_review_plateau", "last_epoch": 469, "last_global_step": 227362, "last_train_loss": 2.182297468185425, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [440, 450, 460], "recent_train_loss": [2.2482223510742188, 2.2399418354034424, 2.229154586791992], "recent_val_loss": [5.3797993659973145, 5.395876884460449, 5.414102554321289], "recent_reconst_mse": [0.07340233027935028, 0.07343350350856781, 0.07325354963541031], "delta_train_loss": 0.019067764282226562, "delta_val_loss": 0.03430318832397461, "delta_reconst_mse": -0.00014878064393997192, "signals": {"overfit_val": false, "overfit_reconst": false, "offline_plateau": true, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T08:28:46+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
15
+ {"status": "active", "reason": "counterfactual review: offline metrics plateau", "verdict": "would_review_plateau", "last_epoch": 505, "last_global_step": 244457, "last_train_loss": 2.1715638637542725, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [480, 490, 500], "recent_train_loss": [2.2127134799957275, 2.204120397567749, 2.196258068084717], "recent_val_loss": [5.452072620391846, 5.46730375289917, 5.482973098754883], "recent_reconst_mse": [0.07327774167060852, 0.07316216826438904, 0.07313615828752518], "delta_train_loss": 0.016455411911010742, "delta_val_loss": 0.03090047836303711, "delta_reconst_mse": -0.0001415833830833435, "signals": {"overfit_val": false, "overfit_reconst": false, "offline_plateau": true, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T09:28:47+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}
16
+ {"status": "active", "reason": "counterfactual review: offline metrics plateau", "verdict": "would_review_plateau", "last_epoch": 540, "last_global_step": 261628, "last_train_loss": 2.2408196926116943, "last_val_loss": null, "last_reconst_mse": null, "recent_eval_epochs": [510, 520, 530], "recent_train_loss": [2.1868083477020264, 2.1790850162506104, 2.1684956550598145], "recent_val_loss": [5.498061180114746, 5.516268730163574, 5.5319976806640625], "recent_reconst_mse": [0.07329904288053513, 0.073332779109478, 0.07386821508407593], "delta_train_loss": 0.018312692642211914, "delta_val_loss": 0.033936500549316406, "delta_reconst_mse": 0.000569172203540802, "signals": {"overfit_val": false, "overfit_reconst": false, "offline_plateau": true, "counterfactual_only": true, "training_not_stopped": true}, "check_time_utc": "2026-05-31T10:28:48+00:00", "counterfactual_only": true, "training_not_stopped": true, "confidence": "high"}