End of training

README.md

@@ -33,12 +33,12 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swin-base-patch4-window7-224-in22k](https://huggingface.co/microsoft/swin-base-patch4-window7-224-in22k) on the imagefolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3619
+- Loss: 0.3584
 - Accuracy: 0.8600
-- Auc Overall: 0.9320
-- Auc Class 0: 0.9628
-- Auc Class 1: 0.9083
-- Auc Class 2: 0.9249
+- Auc Overall: 0.9327
+- Auc Class 0: 0.9624
+- Auc Class 1: 0.9106
+- Auc Class 2: 0.9253
 
 ## Model description
 

all_results.json

@@ -1,17 +1,17 @@
 {
-    "epoch": 6.885245901639344,
-    "eval_accuracy": 0.8677685950413223,
-    "eval_auc_class_0": 0.9631310096153847,
-    "eval_auc_class_1": 0.9189268008165645,
-    "eval_auc_class_2": 0.9322997435239486,
-    "eval_auc_overall": 0.938119184651966,
-    "eval_loss": 0.34411147236824036,
-    "eval_runtime": 207.0344,
-    "eval_samples_per_second": 9.351,
-    "eval_steps_per_second": 0.077,
-    "total_flos": 4.180411553219629e+18,
-    "train_loss": 0.4214494603020804,
-    "train_runtime": 12181.9025,
-    "train_samples_per_second": 4.448,
+    "epoch": 7.0,
+    "eval_accuracy": 0.8600206611570248,
+    "eval_auc_class_0": 0.9623798076923077,
+    "eval_auc_class_1": 0.9106044580248364,
+    "eval_auc_class_2": 0.9252550195053197,
+    "eval_auc_overall": 0.932746428407488,
+    "eval_loss": 0.358356237411499,
+    "eval_runtime": 205.5157,
+    "eval_samples_per_second": 9.42,
+    "eval_steps_per_second": 0.078,
+    "total_flos": 4.143902092898771e+18,
+    "train_loss": 0.37870263712746755,
+    "train_runtime": 11542.069,
+    "train_samples_per_second": 4.583,
     "train_steps_per_second": 0.009
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 6.885245901639344,
-    "eval_accuracy": 0.8677685950413223,
-    "eval_auc_class_0": 0.9631310096153847,
-    "eval_auc_class_1": 0.9189268008165645,
-    "eval_auc_class_2": 0.9322997435239486,
-    "eval_auc_overall": 0.938119184651966,
-    "eval_loss": 0.34411147236824036,
-    "eval_runtime": 207.0344,
-    "eval_samples_per_second": 9.351,
-    "eval_steps_per_second": 0.077
+    "epoch": 7.0,
+    "eval_accuracy": 0.8600206611570248,
+    "eval_auc_class_0": 0.9623798076923077,
+    "eval_auc_class_1": 0.9106044580248364,
+    "eval_auc_class_2": 0.9252550195053197,
+    "eval_auc_overall": 0.932746428407488,
+    "eval_loss": 0.358356237411499,
+    "eval_runtime": 205.5157,
+    "eval_samples_per_second": 9.42,
+    "eval_steps_per_second": 0.078
 }

runs/Sep23_07-38-42_data-science-wbi/events.out.tfevents.1727088886.data-science-wbi

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9bd2868be35a73e31d4a9ce520d716280e3dbcb46be6d1ee9419479fc2e6b864
+size 300

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 6.885245901639344,
-    "total_flos": 4.180411553219629e+18,
-    "train_loss": 0.4214494603020804,
-    "train_runtime": 12181.9025,
-    "train_samples_per_second": 4.448,
+    "epoch": 7.0,
+    "total_flos": 4.143902092898771e+18,
+    "train_loss": 0.37870263712746755,
+    "train_runtime": 11542.069,
+    "train_samples_per_second": 4.583,
     "train_steps_per_second": 0.009
 }

trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.8677685950413223,
-  "best_model_checkpoint": "swin-base-patch4-window7-224-in22k-MM/NMM_Classification_base_V10/checkpoint-76",
-  "epoch": 6.885245901639344,
+  "best_metric": 0.8600206611570248,
+  "best_model_checkpoint": "swin-base-patch4-window7-224-in22k-MM/NMM_Classification_base_V10/checkpoint-90",
+  "epoch": 7.0,
   "eval_steps": 500,
   "global_step": 105,
   "is_hyper_param_search": false,
@@ -9,264 +9,264 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.6557377049180327,
-      "grad_norm": 4.473873615264893,
+      "epoch": 0.6666666666666666,
+      "grad_norm": 4.08638334274292,
       "learning_rate": 4.545454545454546e-05,
-      "loss": 0.8448,
+      "loss": 0.8429,
       "step": 10
     },
     {
-      "epoch": 0.9836065573770492,
+      "epoch": 1.0,
       "step": 15,
-      "train_accuracy": 0.8035137579124144,
-      "train_auc_class_0": 0.9088436086296741,
-      "train_auc_class_1": 0.8379446511466175,
-      "train_auc_class_2": 0.8758602364377229,
-      "train_auc_overall": 0.8742161654046715,
-      "train_loss": 0.49683675169944763,
-      "train_runtime": 777.5048,
-      "train_samples_per_second": 9.956,
-      "train_steps_per_second": 0.078
+      "train_accuracy": 0.828215987294865,
+      "train_auc_class_0": 0.9164869255677786,
+      "train_auc_class_1": 0.770497463263411,
+      "train_auc_class_2": 0.8879816559016815,
+      "train_auc_overall": 0.858322014910957,
+      "train_loss": 0.4563606083393097,
+      "train_runtime": 687.587,
+      "train_samples_per_second": 10.989,
+      "train_steps_per_second": 0.087
     },
     {
-      "epoch": 0.9836065573770492,
-      "eval_accuracy": 0.8362603305785123,
-      "eval_auc_class_0": 0.9403517055860806,
-      "eval_auc_class_1": 0.8605572313411073,
-      "eval_auc_class_2": 0.8997462510986076,
-      "eval_auc_overall": 0.9002183960085985,
-      "eval_loss": 0.44267916679382324,
-      "eval_runtime": 267.5178,
-      "eval_samples_per_second": 7.237,
-      "eval_steps_per_second": 0.06,
+      "epoch": 1.0,
+      "eval_accuracy": 0.8274793388429752,
+      "eval_auc_class_0": 0.9340616414835164,
+      "eval_auc_class_1": 0.7507519271533845,
+      "eval_auc_class_2": 0.8937217290626166,
+      "eval_auc_overall": 0.8595117658998391,
+      "eval_loss": 0.49912166595458984,
+      "eval_runtime": 208.8157,
+      "eval_samples_per_second": 9.271,
+      "eval_steps_per_second": 0.077,
       "step": 15
     },
     {
-      "epoch": 1.3114754098360657,
-      "grad_norm": 2.1386613845825195,
+      "epoch": 1.3333333333333333,
+      "grad_norm": 1.5998913049697876,
       "learning_rate": 4.5212765957446815e-05,
-      "loss": 0.512,
+      "loss": 0.4716,
       "step": 20
     },
     {
-      "epoch": 1.9672131147540983,
-      "grad_norm": 1.435038685798645,
+      "epoch": 2.0,
+      "grad_norm": 2.4967057704925537,
       "learning_rate": 3.9893617021276594e-05,
-      "loss": 0.4434,
+      "loss": 0.3887,
       "step": 30
     },
     {
-      "epoch": 1.9672131147540983,
+      "epoch": 2.0,
       "step": 30,
-      "train_accuracy": 0.840201524350859,
-      "train_auc_class_0": 0.9414490637347536,
-      "train_auc_class_1": 0.8997057391727663,
-      "train_auc_class_2": 0.9139346212557102,
-      "train_auc_overall": 0.9183631413877434,
-      "train_loss": 0.4013007879257202,
-      "train_runtime": 714.2886,
-      "train_samples_per_second": 10.837,
-      "train_steps_per_second": 0.085
+      "train_accuracy": 0.8553467443091582,
+      "train_auc_class_0": 0.9472708924769339,
+      "train_auc_class_1": 0.9283657851250956,
+      "train_auc_class_2": 0.9247901532600991,
+      "train_auc_overall": 0.9334756102873761,
+      "train_loss": 0.35952258110046387,
+      "train_runtime": 685.7612,
+      "train_samples_per_second": 11.018,
+      "train_steps_per_second": 0.087
     },
     {
-      "epoch": 1.9672131147540983,
-      "eval_accuracy": 0.8445247933884298,
-      "eval_auc_class_0": 0.9560139079670331,
-      "eval_auc_class_1": 0.9042625890669138,
-      "eval_auc_class_2": 0.9195875061825728,
-      "eval_auc_overall": 0.9266213344055066,
-      "eval_loss": 0.3872886896133423,
-      "eval_runtime": 214.551,
-      "eval_samples_per_second": 9.023,
-      "eval_steps_per_second": 0.075,
+      "epoch": 2.0,
+      "eval_accuracy": 0.8502066115702479,
+      "eval_auc_class_0": 0.9477621336996336,
+      "eval_auc_class_1": 0.9035008683616039,
+      "eval_auc_class_2": 0.9096346486873113,
+      "eval_auc_overall": 0.9202992169161829,
+      "eval_loss": 0.3957720398902893,
+      "eval_runtime": 208.9311,
+      "eval_samples_per_second": 9.266,
+      "eval_steps_per_second": 0.077,
       "step": 30
     },
     {
-      "epoch": 2.6229508196721314,
-      "grad_norm": 2.210986852645874,
+      "epoch": 2.6666666666666665,
+      "grad_norm": 1.716306209564209,
       "learning_rate": 3.4574468085106386e-05,
-      "loss": 0.3979,
+      "loss": 0.3533,
       "step": 40
     },
     {
-      "epoch": 2.9508196721311473,
+      "epoch": 3.0,
       "step": 45,
-      "train_accuracy": 0.8537656633509882,
-      "train_auc_class_0": 0.9565036613124893,
-      "train_auc_class_1": 0.9187875975798461,
-      "train_auc_class_2": 0.9285740636485835,
-      "train_auc_overall": 0.9346217741803063,
-      "train_loss": 0.3611737787723541,
-      "train_runtime": 704.8242,
-      "train_samples_per_second": 10.983,
-      "train_steps_per_second": 0.087
+      "train_accuracy": 0.8750661725780836,
+      "train_auc_class_0": 0.9582223466702152,
+      "train_auc_class_1": 0.9392220276138971,
+      "train_auc_class_2": 0.9390688589921182,
+      "train_auc_overall": 0.9455044110920768,
+      "train_loss": 0.3131944239139557,
+      "train_runtime": 681.8829,
+      "train_samples_per_second": 11.081,
+      "train_steps_per_second": 0.088
     },
     {
-      "epoch": 2.9508196721311473,
-      "eval_accuracy": 0.8584710743801653,
-      "eval_auc_class_0": 0.9593907394688643,
-      "eval_auc_class_1": 0.9080973087319308,
-      "eval_auc_class_2": 0.9256975615561118,
-      "eval_auc_overall": 0.931061869918969,
-      "eval_loss": 0.36146315932273865,
-      "eval_runtime": 207.9281,
-      "eval_samples_per_second": 9.311,
+      "epoch": 3.0,
+      "eval_accuracy": 0.8589876033057852,
+      "eval_auc_class_0": 0.9578196543040294,
+      "eval_auc_class_1": 0.9092855930322143,
+      "eval_auc_class_2": 0.920674647298944,
+      "eval_auc_overall": 0.9292599648783959,
+      "eval_loss": 0.37213757634162903,
+      "eval_runtime": 207.6482,
+      "eval_samples_per_second": 9.323,
       "eval_steps_per_second": 0.077,
       "step": 45
     },
     {
-      "epoch": 3.278688524590164,
-      "grad_norm": 2.5246520042419434,
+      "epoch": 3.3333333333333335,
+      "grad_norm": 1.727489948272705,
       "learning_rate": 2.925531914893617e-05,
-      "loss": 0.3733,
+      "loss": 0.3208,
       "step": 50
     },
     {
-      "epoch": 3.9344262295081966,
-      "grad_norm": 2.381304979324341,
+      "epoch": 4.0,
+      "grad_norm": 10.33216381072998,
       "learning_rate": 2.393617021276596e-05,
-      "loss": 0.3593,
+      "loss": 0.3092,
       "step": 60
     },
     {
       "epoch": 4.0,
-      "step": 61,
-      "train_accuracy": 0.865779615036817,
-      "train_auc_class_0": 0.9630790814865717,
-      "train_auc_class_1": 0.9318475950588521,
-      "train_auc_class_2": 0.9373109719652593,
-      "train_auc_overall": 0.9440792161702277,
-      "train_loss": 0.33730149269104004,
-      "train_runtime": 698.4574,
-      "train_samples_per_second": 11.083,
-      "train_steps_per_second": 0.087
+      "step": 60,
+      "train_accuracy": 0.8848597141344626,
+      "train_auc_class_0": 0.9659353652856636,
+      "train_auc_class_1": 0.9407955383480827,
+      "train_auc_class_2": 0.946874731546507,
+      "train_auc_overall": 0.9512018783934177,
+      "train_loss": 0.29102450609207153,
+      "train_runtime": 680.3084,
+      "train_samples_per_second": 11.107,
+      "train_steps_per_second": 0.088
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.862603305785124,
-      "eval_auc_class_0": 0.9629407051282051,
-      "eval_auc_class_1": 0.9201237904963373,
-      "eval_auc_class_2": 0.9316191956643276,
-      "eval_auc_overall": 0.9382278970962901,
-      "eval_loss": 0.34607622027397156,
-      "eval_runtime": 206.0574,
-      "eval_samples_per_second": 9.395,
-      "eval_steps_per_second": 0.078,
-      "step": 61
+      "eval_accuracy": 0.8579545454545454,
+      "eval_auc_class_0": 0.9606026785714286,
+      "eval_auc_class_1": 0.9083715281858423,
+      "eval_auc_class_2": 0.9231885836547031,
+      "eval_auc_overall": 0.9307209301373248,
+      "eval_loss": 0.37202903628349304,
+      "eval_runtime": 208.7004,
+      "eval_samples_per_second": 9.276,
+      "eval_steps_per_second": 0.077,
+      "step": 60
     },
     {
-      "epoch": 4.590163934426229,
-      "grad_norm": 2.1240530014038086,
+      "epoch": 4.666666666666667,
+      "grad_norm": 2.016211748123169,
       "learning_rate": 1.8617021276595745e-05,
-      "loss": 0.3473,
+      "loss": 0.2891,
       "step": 70
     },
     {
-      "epoch": 4.983606557377049,
-      "step": 76,
-      "train_accuracy": 0.8688799896654179,
-      "train_auc_class_0": 0.9654619767508447,
-      "train_auc_class_1": 0.9391473390475861,
-      "train_auc_class_2": 0.9405490167752112,
-      "train_auc_overall": 0.9483861108578807,
-      "train_loss": 0.32362109422683716,
-      "train_runtime": 706.7637,
-      "train_samples_per_second": 10.953,
-      "train_steps_per_second": 0.086
+      "epoch": 5.0,
+      "step": 75,
+      "train_accuracy": 0.8913446267866596,
+      "train_auc_class_0": 0.9681433160338301,
+      "train_auc_class_1": 0.9542337553261225,
+      "train_auc_class_2": 0.9506045992123575,
+      "train_auc_overall": 0.9576605568574367,
+      "train_loss": 0.2791108191013336,
+      "train_runtime": 690.7325,
+      "train_samples_per_second": 10.939,
+      "train_steps_per_second": 0.087
     },
     {
-      "epoch": 4.983606557377049,
-      "eval_accuracy": 0.8677685950413223,
-      "eval_auc_class_0": 0.9631310096153847,
-      "eval_auc_class_1": 0.9189268008165645,
-      "eval_auc_class_2": 0.9322997435239486,
-      "eval_auc_overall": 0.938119184651966,
-      "eval_loss": 0.34411147236824036,
-      "eval_runtime": 207.2323,
-      "eval_samples_per_second": 9.342,
-      "eval_steps_per_second": 0.077,
-      "step": 76
+      "epoch": 5.0,
+      "eval_accuracy": 0.8584710743801653,
+      "eval_auc_class_0": 0.9613524496336998,
+      "eval_auc_class_1": 0.9090984273731954,
+      "eval_auc_class_2": 0.9242633286351978,
+      "eval_auc_overall": 0.9315714018806975,
+      "eval_loss": 0.36684560775756836,
+      "eval_runtime": 211.7912,
+      "eval_samples_per_second": 9.141,
+      "eval_steps_per_second": 0.076,
+      "step": 75
     },
     {
-      "epoch": 5.245901639344262,
-      "grad_norm": 2.395565986633301,
+      "epoch": 5.333333333333333,
+      "grad_norm": 2.7578961849212646,
       "learning_rate": 1.3297872340425532e-05,
-      "loss": 0.3381,
+      "loss": 0.301,
       "step": 80
     },
     {
-      "epoch": 5.901639344262295,
-      "grad_norm": 1.6295080184936523,
+      "epoch": 6.0,
+      "grad_norm": 2.526148557662964,
       "learning_rate": 7.97872340425532e-06,
-      "loss": 0.3318,
+      "loss": 0.2852,
       "step": 90
     },
     {
-      "epoch": 5.967213114754099,
-      "step": 91,
-      "train_accuracy": 0.8766309262369203,
-      "train_auc_class_0": 0.9675600734409895,
-      "train_auc_class_1": 0.9360009866648122,
-      "train_auc_class_2": 0.9450547274524197,
-      "train_auc_overall": 0.9495385958527405,
-      "train_loss": 0.31677091121673584,
-      "train_runtime": 712.6133,
-      "train_samples_per_second": 10.863,
-      "train_steps_per_second": 0.086
+      "epoch": 6.0,
+      "step": 90,
+      "train_accuracy": 0.8955796717840127,
+      "train_auc_class_0": 0.9719376552519517,
+      "train_auc_class_1": 0.9603400183000109,
+      "train_auc_class_2": 0.9560110428344394,
+      "train_auc_overall": 0.9627629054621339,
+      "train_loss": 0.2651444673538208,
+      "train_runtime": 689.8251,
+      "train_samples_per_second": 10.954,
+      "train_steps_per_second": 0.087
     },
     {
-      "epoch": 5.967213114754099,
-      "eval_accuracy": 0.8667355371900827,
-      "eval_auc_class_0": 0.9631009615384616,
-      "eval_auc_class_1": 0.916550232215998,
-      "eval_auc_class_2": 0.9317034893882878,
-      "eval_auc_overall": 0.9371182277142491,
-      "eval_loss": 0.34614038467407227,
-      "eval_runtime": 209.8829,
-      "eval_samples_per_second": 9.224,
+      "epoch": 6.0,
+      "eval_accuracy": 0.8600206611570248,
+      "eval_auc_class_0": 0.9623798076923077,
+      "eval_auc_class_1": 0.9106044580248364,
+      "eval_auc_class_2": 0.9252550195053197,
+      "eval_auc_overall": 0.932746428407488,
+      "eval_loss": 0.358356237411499,
+      "eval_runtime": 209.7236,
+      "eval_samples_per_second": 9.231,
       "eval_steps_per_second": 0.076,
-      "step": 91
+      "step": 90
     },
     {
-      "epoch": 6.557377049180328,
-      "grad_norm": 3.0411829948425293,
+      "epoch": 6.666666666666667,
+      "grad_norm": 1.6473604440689087,
       "learning_rate": 2.6595744680851065e-06,
-      "loss": 0.3147,
+      "loss": 0.2804,
       "step": 100
     },
     {
-      "epoch": 6.885245901639344,
+      "epoch": 7.0,
       "step": 105,
-      "train_accuracy": 0.8770184730654954,
-      "train_auc_class_0": 0.9688894018782569,
-      "train_auc_class_1": 0.9353519937583671,
-      "train_auc_class_2": 0.9460088191382243,
-      "train_auc_overall": 0.9500834049249495,
-      "train_loss": 0.3142475485801697,
-      "train_runtime": 690.6226,
-      "train_samples_per_second": 11.209,
-      "train_steps_per_second": 0.088
+      "train_accuracy": 0.8926680783483325,
+      "train_auc_class_0": 0.9692773945765316,
+      "train_auc_class_1": 0.9595504889107397,
+      "train_auc_class_2": 0.9538150093690271,
+      "train_auc_overall": 0.9608809642854328,
+      "train_loss": 0.2725497782230377,
+      "train_runtime": 674.8314,
+      "train_samples_per_second": 11.197,
+      "train_steps_per_second": 0.089
     },
     {
-      "epoch": 6.885245901639344,
-      "eval_accuracy": 0.862603305785124,
-      "eval_auc_class_0": 0.9628863324175823,
-      "eval_auc_class_1": 0.9189616223345215,
-      "eval_auc_class_2": 0.93194521403788,
-      "eval_auc_overall": 0.937931056263328,
-      "eval_loss": 0.3451487123966217,
-      "eval_runtime": 207.6543,
-      "eval_samples_per_second": 9.323,
-      "eval_steps_per_second": 0.077,
+      "epoch": 7.0,
+      "eval_accuracy": 0.8600206611570248,
+      "eval_auc_class_0": 0.9628448374542126,
+      "eval_auc_class_1": 0.90828447439095,
+      "eval_auc_class_2": 0.924919084223066,
+      "eval_auc_overall": 0.9320161320227428,
+      "eval_loss": 0.3619421422481537,
+      "eval_runtime": 204.9165,
+      "eval_samples_per_second": 9.448,
+      "eval_steps_per_second": 0.078,
       "step": 105
     },
     {
-      "epoch": 6.885245901639344,
+      "epoch": 7.0,
       "step": 105,
-      "total_flos": 4.180411553219629e+18,
-      "train_loss": 0.4214494603020804,
-      "train_runtime": 12181.9025,
-      "train_samples_per_second": 4.448,
+      "total_flos": 4.143902092898771e+18,
+      "train_loss": 0.37870263712746755,
+      "train_runtime": 11542.069,
+      "train_samples_per_second": 4.583,
       "train_steps_per_second": 0.009
     }
   ],
@@ -287,7 +287,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.180411553219629e+18,
+  "total_flos": 4.143902092898771e+18,
   "train_batch_size": 128,
   "trial_name": null,
   "trial_params": null