Training in progress, step 100

all_results.json

@@ -5,8 +5,8 @@
     "eval_samples_per_second": 9.337,
     "eval_steps_per_second": 0.294,
     "total_flos": 7.020774259082035e+17,
-    "train_loss": 0.1566001436166596,
-    "train_runtime": 2850.1536,
-    "train_samples_per_second": 3.2,
-    "train_steps_per_second": 0.1
+    "train_loss": 0.1542712652892397,
+    "train_runtime": 2868.9315,
+    "train_samples_per_second": 3.179,
+    "train_steps_per_second": 0.099
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42cf82b62b6d823e14720710a7cb252c7dc6fa8a4e53005a3e2b0a32eb494d92
+oid sha256:9e06478adc942e5c98492dfcf5e1ccecd7a5d4aabc4b6628ba6e29d768ea6d33
 size 343223968

runs/Jul23_12-57-29_REDLGLAUCORE01/events.out.tfevents.1721767562.REDLGLAUCORE01.547362.21

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5ec0c80ec2f2bfcf3f2424c0204299377e8a20a44bbf1d60db29f1f0cd5183f
+size 1057

runs/Jul23_14-08-36_REDLGLAUCORE01/events.out.tfevents.1721768917.REDLGLAUCORE01.547362.23

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9b0af0c18bd11971e76662931066af4f2eaaba517de1f4b9be7ba03c3de5983
+size 7437

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 5.0,
     "total_flos": 7.020774259082035e+17,
-    "train_loss": 0.1566001436166596,
-    "train_runtime": 2850.1536,
-    "train_samples_per_second": 3.2,
-    "train_steps_per_second": 0.1
+    "train_loss": 0.1542712652892397,
+    "train_runtime": 2868.9315,
+    "train_samples_per_second": 3.179,
+    "train_steps_per_second": 0.099
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.10913512110710144,
+  "best_metric": 0.10512479394674301,
   "best_model_checkpoint": "/home/projects/Project - Fundus Photo Glaucoma Detection (AI)//model_results/vit-base-gpu/checkpoint-200",
   "epoch": 5.0,
   "eval_steps": 100,
@@ -10,72 +10,72 @@
   "log_history": [
     {
       "epoch": 0.17543859649122806,
-      "grad_norm": 0.9124495983123779,
+      "grad_norm": 1.762933373451233,
       "learning_rate": 9.649122807017545e-06,
-      "loss": 0.6422,
+      "loss": 0.5817,
       "step": 10
     },
     {
       "epoch": 0.3508771929824561,
-      "grad_norm": 1.400277853012085,
+      "grad_norm": 1.567679762840271,
       "learning_rate": 9.298245614035088e-06,
-      "loss": 0.5107,
+      "loss": 0.4467,
       "step": 20
     },
     {
       "epoch": 0.5263157894736842,
-      "grad_norm": 1.2670001983642578,
+      "grad_norm": 1.3810224533081055,
       "learning_rate": 8.947368421052632e-06,
-      "loss": 0.4099,
+      "loss": 0.3667,
       "step": 30
     },
     {
       "epoch": 0.7017543859649122,
-      "grad_norm": 1.2379003763198853,
+      "grad_norm": 1.4549124240875244,
       "learning_rate": 8.596491228070176e-06,
-      "loss": 0.3261,
+      "loss": 0.2947,
       "step": 40
     },
     {
       "epoch": 0.8771929824561403,
-      "grad_norm": 1.3035060167312622,
+      "grad_norm": 0.9504640102386475,
       "learning_rate": 8.24561403508772e-06,
-      "loss": 0.278,
+      "loss": 0.2705,
       "step": 50
     },
     {
       "epoch": 1.0526315789473684,
-      "grad_norm": 0.7487425804138184,
+      "grad_norm": 0.7635705471038818,
       "learning_rate": 7.894736842105265e-06,
-      "loss": 0.1874,
+      "loss": 0.1962,
       "step": 60
     },
     {
       "epoch": 1.2280701754385965,
-      "grad_norm": 1.2215185165405273,
+      "grad_norm": 0.7457385063171387,
       "learning_rate": 7.5438596491228074e-06,
-      "loss": 0.1672,
+      "loss": 0.1838,
       "step": 70
     },
     {
       "epoch": 1.4035087719298245,
-      "grad_norm": 0.8201051950454712,
+      "grad_norm": 0.7596029043197632,
       "learning_rate": 7.192982456140352e-06,
-      "loss": 0.1454,
+      "loss": 0.1583,
       "step": 80
     },
     {
       "epoch": 1.5789473684210527,
-      "grad_norm": 1.6430377960205078,
+      "grad_norm": 1.0756629705429077,
       "learning_rate": 6.842105263157896e-06,
-      "loss": 0.1211,
+      "loss": 0.1294,
       "step": 90
     },
     {
       "epoch": 1.7543859649122808,
-      "grad_norm": 3.366523504257202,
+      "grad_norm": 0.8500545024871826,
       "learning_rate": 6.491228070175439e-06,
-      "loss": 0.124,
+      "loss": 0.1311,
       "step": 100
     },
     {
@@ -91,165 +91,165 @@
           161
         ]
       ],
-      "eval_loss": 0.16817055642604828,
-      "eval_runtime": 25.0148,
-      "eval_samples_per_second": 9.075,
-      "eval_steps_per_second": 0.32,
+      "eval_loss": 0.15690478682518005,
+      "eval_runtime": 24.1707,
+      "eval_samples_per_second": 9.392,
+      "eval_steps_per_second": 0.331,
       "step": 100
     },
     {
       "epoch": 1.9298245614035088,
-      "grad_norm": 5.223181247711182,
+      "grad_norm": 2.896561861038208,
       "learning_rate": 6.140350877192983e-06,
-      "loss": 0.1256,
+      "loss": 0.1334,
       "step": 110
     },
     {
       "epoch": 2.1052631578947367,
-      "grad_norm": 0.456160306930542,
+      "grad_norm": 0.3675668239593506,
       "learning_rate": 5.789473684210527e-06,
-      "loss": 0.1073,
+      "loss": 0.1151,
       "step": 120
     },
     {
       "epoch": 2.280701754385965,
-      "grad_norm": 0.40346741676330566,
+      "grad_norm": 0.46107763051986694,
       "learning_rate": 5.438596491228071e-06,
-      "loss": 0.1071,
+      "loss": 0.1217,
       "step": 130
     },
     {
       "epoch": 2.456140350877193,
-      "grad_norm": 0.3802075684070587,
+      "grad_norm": 0.46047243475914,
       "learning_rate": 5.087719298245615e-06,
-      "loss": 0.0769,
+      "loss": 0.0809,
       "step": 140
     },
     {
       "epoch": 2.6315789473684212,
-      "grad_norm": 0.29936113953590393,
+      "grad_norm": 0.33362969756126404,
       "learning_rate": 4.736842105263158e-06,
-      "loss": 0.0904,
+      "loss": 0.0935,
       "step": 150
     },
     {
       "epoch": 2.807017543859649,
-      "grad_norm": 0.5365840792655945,
+      "grad_norm": 0.4287916421890259,
       "learning_rate": 4.385964912280702e-06,
-      "loss": 0.0815,
+      "loss": 0.0855,
       "step": 160
     },
     {
       "epoch": 2.982456140350877,
-      "grad_norm": 3.9614758491516113,
+      "grad_norm": 0.4844302237033844,
       "learning_rate": 4.035087719298246e-06,
-      "loss": 0.0836,
+      "loss": 0.0845,
       "step": 170
     },
     {
       "epoch": 3.1578947368421053,
-      "grad_norm": 0.2877553701400757,
+      "grad_norm": 0.3423462212085724,
       "learning_rate": 3.6842105263157896e-06,
-      "loss": 0.086,
+      "loss": 0.0911,
       "step": 180
     },
     {
       "epoch": 3.3333333333333335,
-      "grad_norm": 0.27285733819007874,
+      "grad_norm": 0.288198322057724,
       "learning_rate": 3.3333333333333333e-06,
-      "loss": 0.0808,
+      "loss": 0.0855,
       "step": 190
     },
     {
       "epoch": 3.5087719298245617,
-      "grad_norm": 1.5225030183792114,
+      "grad_norm": 0.5208536982536316,
       "learning_rate": 2.9824561403508774e-06,
-      "loss": 0.0881,
+      "loss": 0.0929,
       "step": 200
     },
     {
       "epoch": 3.5087719298245617,
-      "eval_accuracy": 0.9779735682819384,
+      "eval_accuracy": 0.9823788546255506,
       "eval_confusion_matrix": [
         [
-          61,
-          5
+          62,
+          4
         ],
         [
           0,
           161
         ]
       ],
-      "eval_loss": 0.10913512110710144,
-      "eval_runtime": 24.033,
-      "eval_samples_per_second": 9.445,
+      "eval_loss": 0.10512479394674301,
+      "eval_runtime": 24.0408,
+      "eval_samples_per_second": 9.442,
       "eval_steps_per_second": 0.333,
       "step": 200
     },
     {
       "epoch": 3.6842105263157894,
-      "grad_norm": 0.2580512762069702,
+      "grad_norm": 0.3295786380767822,
       "learning_rate": 2.631578947368421e-06,
-      "loss": 0.1032,
+      "loss": 0.1019,
       "step": 210
     },
     {
       "epoch": 3.8596491228070176,
-      "grad_norm": 0.5844484567642212,
+      "grad_norm": 0.4086381196975708,
       "learning_rate": 2.280701754385965e-06,
-      "loss": 0.0662,
+      "loss": 0.069,
       "step": 220
     },
     {
       "epoch": 4.035087719298246,
-      "grad_norm": 0.8993027210235596,
+      "grad_norm": 0.7247630953788757,
       "learning_rate": 1.929824561403509e-06,
-      "loss": 0.0954,
+      "loss": 0.1078,
       "step": 230
     },
     {
       "epoch": 4.2105263157894735,
-      "grad_norm": 0.35451096296310425,
+      "grad_norm": 0.37916073203086853,
       "learning_rate": 1.5789473684210526e-06,
-      "loss": 0.0752,
+      "loss": 0.0757,
       "step": 240
     },
     {
       "epoch": 4.385964912280702,
-      "grad_norm": 0.27109700441360474,
+      "grad_norm": 0.26061660051345825,
       "learning_rate": 1.2280701754385965e-06,
-      "loss": 0.0563,
+      "loss": 0.0572,
       "step": 250
     },
     {
       "epoch": 4.56140350877193,
-      "grad_norm": 0.24192890524864197,
+      "grad_norm": 0.259589284658432,
       "learning_rate": 8.771929824561404e-07,
-      "loss": 0.0535,
+      "loss": 0.0545,
       "step": 260
     },
     {
       "epoch": 4.7368421052631575,
-      "grad_norm": 0.2452133148908615,
+      "grad_norm": 0.25950363278388977,
       "learning_rate": 5.263157894736843e-07,
-      "loss": 0.0405,
+      "loss": 0.0447,
       "step": 270
     },
     {
       "epoch": 4.912280701754386,
-      "grad_norm": 0.4464854598045349,
+      "grad_norm": 0.4733501970767975,
       "learning_rate": 1.7543859649122808e-07,
-      "loss": 0.0869,
+      "loss": 0.0958,
       "step": 280
     },
     {
       "epoch": 5.0,
       "step": 285,
       "total_flos": 7.020774259082035e+17,
-      "train_loss": 0.1566001436166596,
-      "train_runtime": 2850.1536,
-      "train_samples_per_second": 3.2,
-      "train_steps_per_second": 0.1
+      "train_loss": 0.1542712652892397,
+      "train_runtime": 2868.9315,
+      "train_samples_per_second": 3.179,
+      "train_steps_per_second": 0.099
     }
   ],
   "logging_steps": 10,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a918e9d2444d2d031c19a1c82f4000ad8b839b20b58e9435864a7279430830b
+oid sha256:7c1c361d3cbac2a92ed11a391a07e944283c5f829d07e39a2674baa3e68a4b34
 size 5240