End of training

all_results.json

@@ -1,13 +1,14 @@
 {
     "epoch": 6.885245901639344,
-    "eval_accuracy": 0.8729338842975206,
-    "eval_loss": 0.34267309308052063,
-    "eval_runtime": 203.1004,
-    "eval_samples_per_second": 9.532,
+    "eval_accuracy": 0.8824443293630243,
+    "eval_auc": 0.957557852599524,
+    "eval_loss": 0.30561181902885437,
+    "eval_runtime": 203.3286,
+    "eval_samples_per_second": 9.497,
     "eval_steps_per_second": 0.079,
-    "total_flos": 4.180411553219629e+18,
-    "train_loss": 0.4180879854020618,
-    "train_runtime": 6613.2188,
-    "train_samples_per_second": 8.194,
+    "total_flos": 4.1785312376666235e+18,
+    "train_loss": 0.38780884515671504,
+    "train_runtime": 6680.4521,
+    "train_samples_per_second": 8.107,
     "train_steps_per_second": 0.016
 }

eval_results.json

@@ -1,8 +1,9 @@
 {
     "epoch": 6.885245901639344,
-    "eval_accuracy": 0.8729338842975206,
-    "eval_loss": 0.34267309308052063,
-    "eval_runtime": 203.1004,
-    "eval_samples_per_second": 9.532,
+    "eval_accuracy": 0.8824443293630243,
+    "eval_auc": 0.957557852599524,
+    "eval_loss": 0.30561181902885437,
+    "eval_runtime": 203.3286,
+    "eval_samples_per_second": 9.497,
     "eval_steps_per_second": 0.079
 }

runs/Aug28_16-52-19_data-science-wbi/events.out.tfevents.1724870839.data-science-wbi

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b08144f96c5b4de882fba4b0c1b99cff56e739f68a99dc5afe31dc0e5892055
+size 87

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 6.885245901639344,
-    "total_flos": 4.180411553219629e+18,
-    "train_loss": 0.4180879854020618,
-    "train_runtime": 6613.2188,
-    "train_samples_per_second": 8.194,
+    "total_flos": 4.1785312376666235e+18,
+    "train_loss": 0.38780884515671504,
+    "train_runtime": 6680.4521,
+    "train_samples_per_second": 8.107,
     "train_steps_per_second": 0.016
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.8729338842975206,
-  "best_model_checkpoint": "swin-base-patch4-window7-224-in22k-MM_Classification_base_V10/checkpoint-91",
+  "best_metric": 0.8824443293630243,
+  "best_model_checkpoint": "swin-base-patch4-window7-224-in22k-MM_Classification_base_V10/checkpoint-105",
   "epoch": 6.885245901639344,
   "eval_steps": 500,
   "global_step": 105,
@@ -10,144 +10,151 @@
   "log_history": [
     {
       "epoch": 0.6557377049180327,
-      "grad_norm": 1.847607135772705,
+      "grad_norm": 5.335843086242676,
       "learning_rate": 4.545454545454546e-05,
-      "loss": 0.7948,
+      "loss": 0.8134,
       "step": 10
     },
     {
       "epoch": 0.9836065573770492,
-      "eval_accuracy": 0.8352272727272727,
-      "eval_loss": 0.4498441815376282,
-      "eval_runtime": 204.2611,
-      "eval_samples_per_second": 9.478,
+      "eval_accuracy": 0.8430864836872087,
+      "eval_auc": 0.9250639596370771,
+      "eval_loss": 0.419461190700531,
+      "eval_runtime": 205.2459,
+      "eval_samples_per_second": 9.408,
       "eval_steps_per_second": 0.078,
       "step": 15
     },
     {
       "epoch": 1.3114754098360657,
-      "grad_norm": 2.3756892681121826,
+      "grad_norm": 2.8889973163604736,
       "learning_rate": 4.5212765957446815e-05,
-      "loss": 0.5062,
+      "loss": 0.5101,
       "step": 20
     },
     {
       "epoch": 1.9672131147540983,
-      "grad_norm": 1.4792882204055786,
+      "grad_norm": 1.6868778467178345,
       "learning_rate": 3.9893617021276594e-05,
-      "loss": 0.4439,
+      "loss": 0.3996,
       "step": 30
     },
     {
       "epoch": 1.9672131147540983,
-      "eval_accuracy": 0.8589876033057852,
-      "eval_loss": 0.38362234830856323,
-      "eval_runtime": 210.2217,
-      "eval_samples_per_second": 9.209,
-      "eval_steps_per_second": 0.076,
+      "eval_accuracy": 0.8570688762299327,
+      "eval_auc": 0.9437553079992403,
+      "eval_loss": 0.3682312071323395,
+      "eval_runtime": 204.3267,
+      "eval_samples_per_second": 9.451,
+      "eval_steps_per_second": 0.078,
       "step": 30
     },
     {
       "epoch": 2.6229508196721314,
-      "grad_norm": 2.130309581756592,
+      "grad_norm": 1.9163786172866821,
       "learning_rate": 3.4574468085106386e-05,
-      "loss": 0.4024,
+      "loss": 0.3693,
       "step": 40
     },
     {
       "epoch": 2.9508196721311473,
-      "eval_accuracy": 0.8600206611570248,
-      "eval_loss": 0.36524850130081177,
-      "eval_runtime": 204.4105,
-      "eval_samples_per_second": 9.471,
+      "eval_accuracy": 0.877783531848783,
+      "eval_auc": 0.9529726135197429,
+      "eval_loss": 0.324381947517395,
+      "eval_runtime": 203.8679,
+      "eval_samples_per_second": 9.472,
       "eval_steps_per_second": 0.078,
       "step": 45
     },
     {
       "epoch": 3.278688524590164,
-      "grad_norm": 2.3738443851470947,
+      "grad_norm": 2.268683433532715,
       "learning_rate": 2.925531914893617e-05,
-      "loss": 0.3801,
+      "loss": 0.3326,
       "step": 50
     },
     {
       "epoch": 3.9344262295081966,
-      "grad_norm": 3.785062074661255,
+      "grad_norm": 2.837725877761841,
       "learning_rate": 2.393617021276596e-05,
-      "loss": 0.3562,
+      "loss": 0.3147,
       "step": 60
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8641528925619835,
-      "eval_loss": 0.34744057059288025,
-      "eval_runtime": 203.8324,
-      "eval_samples_per_second": 9.498,
+      "eval_accuracy": 0.8793371310201968,
+      "eval_auc": 0.9557452970649672,
+      "eval_loss": 0.31093424558639526,
+      "eval_runtime": 204.9066,
+      "eval_samples_per_second": 9.424,
       "eval_steps_per_second": 0.078,
       "step": 61
     },
     {
       "epoch": 4.590163934426229,
-      "grad_norm": 3.449556827545166,
+      "grad_norm": 1.7413060665130615,
       "learning_rate": 1.8617021276595745e-05,
-      "loss": 0.345,
+      "loss": 0.3166,
       "step": 70
     },
     {
       "epoch": 4.983606557377049,
-      "eval_accuracy": 0.868801652892562,
-      "eval_loss": 0.34286215901374817,
-      "eval_runtime": 203.6664,
-      "eval_samples_per_second": 9.506,
-      "eval_steps_per_second": 0.079,
+      "eval_accuracy": 0.8803728638011393,
+      "eval_auc": 0.9561359266346766,
+      "eval_loss": 0.3134761154651642,
+      "eval_runtime": 204.4948,
+      "eval_samples_per_second": 9.443,
+      "eval_steps_per_second": 0.078,
       "step": 76
     },
     {
       "epoch": 5.245901639344262,
-      "grad_norm": 2.114426851272583,
+      "grad_norm": 1.7945168018341064,
       "learning_rate": 1.3297872340425532e-05,
-      "loss": 0.3374,
+      "loss": 0.2921,
       "step": 80
     },
     {
       "epoch": 5.901639344262295,
-      "grad_norm": 4.8909759521484375,
+      "grad_norm": 1.865674376487732,
       "learning_rate": 7.97872340425532e-06,
-      "loss": 0.3379,
+      "loss": 0.2918,
       "step": 90
     },
     {
       "epoch": 5.967213114754099,
-      "eval_accuracy": 0.8729338842975206,
-      "eval_loss": 0.34267309308052063,
-      "eval_runtime": 204.1334,
-      "eval_samples_per_second": 9.484,
+      "eval_accuracy": 0.8746763335059554,
+      "eval_auc": 0.9568584001596542,
+      "eval_loss": 0.3116800785064697,
+      "eval_runtime": 205.4646,
+      "eval_samples_per_second": 9.398,
       "eval_steps_per_second": 0.078,
       "step": 91
     },
     {
       "epoch": 6.557377049180328,
-      "grad_norm": 3.155640125274658,
+      "grad_norm": 4.410208225250244,
       "learning_rate": 2.6595744680851065e-06,
-      "loss": 0.3213,
+      "loss": 0.2849,
       "step": 100
     },
     {
       "epoch": 6.885245901639344,
-      "eval_accuracy": 0.8708677685950413,
-      "eval_loss": 0.3443421423435211,
-      "eval_runtime": 203.5472,
-      "eval_samples_per_second": 9.511,
+      "eval_accuracy": 0.8824443293630243,
+      "eval_auc": 0.957557852599524,
+      "eval_loss": 0.30561181902885437,
+      "eval_runtime": 203.0816,
+      "eval_samples_per_second": 9.508,
       "eval_steps_per_second": 0.079,
       "step": 105
     },
     {
       "epoch": 6.885245901639344,
       "step": 105,
-      "total_flos": 4.180411553219629e+18,
-      "train_loss": 0.4180879854020618,
-      "train_runtime": 6613.2188,
-      "train_samples_per_second": 8.194,
+      "total_flos": 4.1785312376666235e+18,
+      "train_loss": 0.38780884515671504,
+      "train_runtime": 6680.4521,
+      "train_samples_per_second": 8.107,
       "train_steps_per_second": 0.016
     }
   ],
@@ -168,7 +175,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.180411553219629e+18,
+  "total_flos": 4.1785312376666235e+18,
   "train_batch_size": 128,
   "trial_name": null,
   "trial_params": null