End of training

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.25,
-    "eval_accuracy": 0.9032258064516129,
-    "eval_loss": 0.3167099952697754,
-    "eval_runtime": 46.9133,
-    "eval_samples_per_second": 3.304,
-    "eval_steps_per_second": 1.663
+    "eval_accuracy": 0.8903225806451613,
+    "eval_loss": 0.3532361686229706,
+    "eval_runtime": 44.555,
+    "eval_samples_per_second": 3.479,
+    "eval_steps_per_second": 1.751
 }

runs/Nov30_08-17-15_4a534e3efb22/events.out.tfevents.1732955614.4a534e3efb22.2277.1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d281512f3ed3599dfe022520b5dc10c2cb3c6c975f009679933bdbc51588da1d
-size 411
+oid sha256:e9c9f05f3103978db8c6cb195b8a0eb46ff2e83c40800ea521827b3feedb5288
+size 734

test_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.25,
-    "eval_accuracy": 0.9032258064516129,
-    "eval_loss": 0.3167099952697754,
-    "eval_runtime": 46.9133,
-    "eval_samples_per_second": 3.304,
-    "eval_steps_per_second": 1.663
+    "eval_accuracy": 0.8903225806451613,
+    "eval_loss": 0.3532361686229706,
+    "eval_runtime": 44.555,
+    "eval_samples_per_second": 3.479,
+    "eval_steps_per_second": 1.751
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.9142857142857143,
+  "best_metric": 0.8714285714285714,
   "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-600",
   "epoch": 3.25,
   "eval_steps": 500,
@@ -10,485 +10,485 @@
   "log_history": [
     {
       "epoch": 0.016666666666666666,
-      "grad_norm": 13.878811836242676,
+      "grad_norm": 14.917113304138184,
       "learning_rate": 8.333333333333334e-06,
-      "loss": 2.3864,
+      "loss": 2.4251,
       "step": 10
     },
     {
       "epoch": 0.03333333333333333,
-      "grad_norm": 12.958414077758789,
+      "grad_norm": 12.213376998901367,
       "learning_rate": 1.6666666666666667e-05,
-      "loss": 2.2972,
+      "loss": 2.3123,
       "step": 20
     },
     {
       "epoch": 0.05,
-      "grad_norm": 13.340024948120117,
+      "grad_norm": 13.984800338745117,
       "learning_rate": 2.5e-05,
-      "loss": 2.3691,
+      "loss": 2.3116,
       "step": 30
     },
     {
       "epoch": 0.06666666666666667,
-      "grad_norm": 14.514385223388672,
+      "grad_norm": 14.666156768798828,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.2344,
+      "loss": 2.2155,
       "step": 40
     },
     {
       "epoch": 0.08333333333333333,
-      "grad_norm": 14.681234359741211,
+      "grad_norm": 15.621269226074219,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 2.1617,
+      "loss": 2.1373,
       "step": 50
     },
     {
       "epoch": 0.1,
-      "grad_norm": 15.572694778442383,
+      "grad_norm": 14.200703620910645,
       "learning_rate": 5e-05,
-      "loss": 2.2188,
+      "loss": 2.1844,
       "step": 60
     },
     {
       "epoch": 0.11666666666666667,
-      "grad_norm": 16.609718322753906,
+      "grad_norm": 15.562609672546387,
       "learning_rate": 4.9074074074074075e-05,
-      "loss": 2.0842,
+      "loss": 2.1803,
       "step": 70
     },
     {
       "epoch": 0.13333333333333333,
-      "grad_norm": 16.88509750366211,
+      "grad_norm": 16.233795166015625,
       "learning_rate": 4.814814814814815e-05,
-      "loss": 1.8851,
+      "loss": 1.9119,
       "step": 80
     },
     {
       "epoch": 0.15,
-      "grad_norm": 16.43447494506836,
+      "grad_norm": 17.751785278320312,
       "learning_rate": 4.722222222222222e-05,
-      "loss": 2.1666,
+      "loss": 1.9925,
       "step": 90
     },
     {
       "epoch": 0.16666666666666666,
-      "grad_norm": 16.14389419555664,
+      "grad_norm": 16.339580535888672,
       "learning_rate": 4.62962962962963e-05,
-      "loss": 2.0483,
+      "loss": 1.7039,
       "step": 100
     },
     {
       "epoch": 0.18333333333333332,
-      "grad_norm": 12.657491683959961,
+      "grad_norm": 12.999968528747559,
       "learning_rate": 4.5370370370370374e-05,
-      "loss": 1.9418,
+      "loss": 1.7949,
       "step": 110
     },
     {
       "epoch": 0.2,
-      "grad_norm": 17.41259002685547,
+      "grad_norm": 14.487344741821289,
       "learning_rate": 4.4444444444444447e-05,
-      "loss": 1.869,
+      "loss": 1.5218,
       "step": 120
     },
     {
       "epoch": 0.21666666666666667,
-      "grad_norm": 28.190662384033203,
+      "grad_norm": 11.726049423217773,
       "learning_rate": 4.351851851851852e-05,
-      "loss": 1.6983,
+      "loss": 1.6772,
       "step": 130
     },
     {
       "epoch": 0.23333333333333334,
-      "grad_norm": 15.854883193969727,
+      "grad_norm": 13.748116493225098,
       "learning_rate": 4.259259259259259e-05,
-      "loss": 1.9582,
+      "loss": 1.6632,
       "step": 140
     },
     {
       "epoch": 0.25,
-      "grad_norm": 12.764759063720703,
+      "grad_norm": 16.4953556060791,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 1.3516,
+      "loss": 1.2945,
       "step": 150
     },
     {
       "epoch": 0.25,
-      "eval_accuracy": 0.5571428571428572,
-      "eval_loss": 1.3335753679275513,
-      "eval_runtime": 21.7132,
-      "eval_samples_per_second": 3.224,
-      "eval_steps_per_second": 1.612,
+      "eval_accuracy": 0.5142857142857142,
+      "eval_loss": 1.1367460489273071,
+      "eval_runtime": 19.6454,
+      "eval_samples_per_second": 3.563,
+      "eval_steps_per_second": 1.782,
       "step": 150
     },
     {
       "epoch": 1.0166666666666666,
-      "grad_norm": 17.67679786682129,
+      "grad_norm": 17.966445922851562,
       "learning_rate": 4.074074074074074e-05,
-      "loss": 1.083,
+      "loss": 0.8598,
       "step": 160
     },
     {
       "epoch": 1.0333333333333334,
-      "grad_norm": 28.23631477355957,
+      "grad_norm": 13.563071250915527,
       "learning_rate": 3.981481481481482e-05,
-      "loss": 1.5288,
+      "loss": 1.3691,
       "step": 170
     },
     {
       "epoch": 1.05,
-      "grad_norm": 18.444087982177734,
+      "grad_norm": 14.51318073272705,
       "learning_rate": 3.888888888888889e-05,
-      "loss": 1.2424,
+      "loss": 0.8532,
       "step": 180
     },
     {
       "epoch": 1.0666666666666667,
-      "grad_norm": 7.792266368865967,
+      "grad_norm": 6.940643310546875,
       "learning_rate": 3.7962962962962964e-05,
-      "loss": 1.1327,
+      "loss": 0.8452,
       "step": 190
     },
     {
       "epoch": 1.0833333333333333,
-      "grad_norm": 35.91878890991211,
+      "grad_norm": 39.494544982910156,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 1.3238,
+      "loss": 1.2556,
       "step": 200
     },
     {
       "epoch": 1.1,
-      "grad_norm": 14.137838363647461,
+      "grad_norm": 12.343575477600098,
       "learning_rate": 3.611111111111111e-05,
-      "loss": 0.8591,
+      "loss": 0.7716,
       "step": 210
     },
     {
       "epoch": 1.1166666666666667,
-      "grad_norm": 14.379003524780273,
+      "grad_norm": 15.665731430053711,
       "learning_rate": 3.518518518518519e-05,
-      "loss": 0.8904,
+      "loss": 0.9313,
       "step": 220
     },
     {
       "epoch": 1.1333333333333333,
-      "grad_norm": 35.3349723815918,
+      "grad_norm": 13.885952949523926,
       "learning_rate": 3.425925925925926e-05,
-      "loss": 1.0365,
+      "loss": 0.7732,
       "step": 230
     },
     {
       "epoch": 1.15,
-      "grad_norm": 4.670235633850098,
+      "grad_norm": 2.170562982559204,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 1.0926,
+      "loss": 0.7213,
       "step": 240
     },
     {
       "epoch": 1.1666666666666667,
-      "grad_norm": 24.737224578857422,
+      "grad_norm": 31.940528869628906,
       "learning_rate": 3.240740740740741e-05,
-      "loss": 0.6525,
+      "loss": 0.6287,
       "step": 250
     },
     {
       "epoch": 1.1833333333333333,
-      "grad_norm": 7.743138313293457,
+      "grad_norm": 5.167452335357666,
       "learning_rate": 3.148148148148148e-05,
-      "loss": 0.6751,
+      "loss": 0.575,
       "step": 260
     },
     {
       "epoch": 1.2,
-      "grad_norm": 19.620027542114258,
+      "grad_norm": 22.906959533691406,
       "learning_rate": 3.055555555555556e-05,
-      "loss": 0.9504,
+      "loss": 0.4796,
       "step": 270
     },
     {
       "epoch": 1.2166666666666668,
-      "grad_norm": 11.127130508422852,
+      "grad_norm": 11.629034042358398,
       "learning_rate": 2.962962962962963e-05,
-      "loss": 0.3769,
+      "loss": 0.3411,
       "step": 280
     },
     {
       "epoch": 1.2333333333333334,
-      "grad_norm": 15.655902862548828,
+      "grad_norm": 8.874262809753418,
       "learning_rate": 2.8703703703703706e-05,
-      "loss": 1.0,
+      "loss": 0.6513,
       "step": 290
     },
     {
       "epoch": 1.25,
-      "grad_norm": 3.454685926437378,
+      "grad_norm": 12.997471809387207,
       "learning_rate": 2.777777777777778e-05,
-      "loss": 1.0445,
+      "loss": 0.8621,
       "step": 300
     },
     {
       "epoch": 1.25,
-      "eval_accuracy": 0.7714285714285715,
-      "eval_loss": 0.804114580154419,
-      "eval_runtime": 21.0006,
-      "eval_samples_per_second": 3.333,
-      "eval_steps_per_second": 1.667,
+      "eval_accuracy": 0.7857142857142857,
+      "eval_loss": 0.7921156287193298,
+      "eval_runtime": 19.5302,
+      "eval_samples_per_second": 3.584,
+      "eval_steps_per_second": 1.792,
       "step": 300
     },
     {
       "epoch": 2.0166666666666666,
-      "grad_norm": 10.995410919189453,
+      "grad_norm": 17.919185638427734,
       "learning_rate": 2.6851851851851855e-05,
-      "loss": 0.4714,
+      "loss": 0.4647,
       "step": 310
     },
     {
       "epoch": 2.033333333333333,
-      "grad_norm": 1.7594047784805298,
+      "grad_norm": 0.9270963072776794,
       "learning_rate": 2.5925925925925925e-05,
-      "loss": 0.1752,
+      "loss": 0.1672,
       "step": 320
     },
     {
       "epoch": 2.05,
-      "grad_norm": 13.059496879577637,
+      "grad_norm": 13.211724281311035,
       "learning_rate": 2.5e-05,
-      "loss": 0.4913,
+      "loss": 0.5385,
       "step": 330
     },
     {
       "epoch": 2.066666666666667,
-      "grad_norm": 12.100168228149414,
+      "grad_norm": 61.24877166748047,
       "learning_rate": 2.4074074074074074e-05,
-      "loss": 0.2387,
+      "loss": 0.445,
       "step": 340
     },
     {
       "epoch": 2.0833333333333335,
-      "grad_norm": 14.437403678894043,
+      "grad_norm": 8.559224128723145,
       "learning_rate": 2.314814814814815e-05,
-      "loss": 0.6051,
+      "loss": 0.5108,
       "step": 350
     },
     {
       "epoch": 2.1,
-      "grad_norm": 18.535968780517578,
+      "grad_norm": 5.133317470550537,
       "learning_rate": 2.2222222222222223e-05,
-      "loss": 0.5252,
+      "loss": 0.4387,
       "step": 360
     },
     {
       "epoch": 2.1166666666666667,
-      "grad_norm": 27.02277183532715,
+      "grad_norm": 0.2432795614004135,
       "learning_rate": 2.1296296296296296e-05,
-      "loss": 0.5674,
+      "loss": 0.5434,
       "step": 370
     },
     {
       "epoch": 2.1333333333333333,
-      "grad_norm": 2.0069992542266846,
+      "grad_norm": 1.517449140548706,
       "learning_rate": 2.037037037037037e-05,
-      "loss": 0.1657,
+      "loss": 0.1532,
       "step": 380
     },
     {
       "epoch": 2.15,
-      "grad_norm": 0.19569337368011475,
+      "grad_norm": 0.15583747625350952,
       "learning_rate": 1.9444444444444445e-05,
-      "loss": 0.461,
+      "loss": 0.4374,
       "step": 390
     },
     {
       "epoch": 2.1666666666666665,
-      "grad_norm": 0.5586942434310913,
+      "grad_norm": 1.5011425018310547,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.4876,
+      "loss": 0.5595,
       "step": 400
     },
     {
       "epoch": 2.183333333333333,
-      "grad_norm": 1.0278728008270264,
+      "grad_norm": 2.154327154159546,
       "learning_rate": 1.7592592592592595e-05,
-      "loss": 0.5508,
+      "loss": 0.4602,
       "step": 410
     },
     {
       "epoch": 2.2,
-      "grad_norm": 1.1786044836044312,
+      "grad_norm": 0.3127913177013397,
       "learning_rate": 1.6666666666666667e-05,
-      "loss": 0.1829,
+      "loss": 0.1026,
       "step": 420
     },
     {
       "epoch": 2.216666666666667,
-      "grad_norm": 5.859825611114502,
+      "grad_norm": 10.559721946716309,
       "learning_rate": 1.574074074074074e-05,
-      "loss": 0.2612,
+      "loss": 0.243,
       "step": 430
     },
     {
       "epoch": 2.2333333333333334,
-      "grad_norm": 0.49297359585762024,
+      "grad_norm": 0.7269019484519958,
       "learning_rate": 1.4814814814814815e-05,
-      "loss": 0.1112,
+      "loss": 0.0711,
       "step": 440
     },
     {
       "epoch": 2.25,
-      "grad_norm": 29.458885192871094,
+      "grad_norm": 48.36089324951172,
       "learning_rate": 1.388888888888889e-05,
-      "loss": 0.1752,
+      "loss": 0.2108,
       "step": 450
     },
     {
       "epoch": 2.25,
-      "eval_accuracy": 0.7285714285714285,
-      "eval_loss": 0.7214392423629761,
-      "eval_runtime": 21.636,
-      "eval_samples_per_second": 3.235,
-      "eval_steps_per_second": 1.618,
+      "eval_accuracy": 0.7857142857142857,
+      "eval_loss": 0.5835635662078857,
+      "eval_runtime": 19.6437,
+      "eval_samples_per_second": 3.563,
+      "eval_steps_per_second": 1.782,
       "step": 450
     },
     {
       "epoch": 3.0166666666666666,
-      "grad_norm": 0.2373313158750534,
+      "grad_norm": 0.2267579734325409,
       "learning_rate": 1.2962962962962962e-05,
-      "loss": 0.0994,
+      "loss": 0.0858,
       "step": 460
     },
     {
       "epoch": 3.033333333333333,
-      "grad_norm": 3.266122341156006,
+      "grad_norm": 1.7032139301300049,
       "learning_rate": 1.2037037037037037e-05,
-      "loss": 0.1565,
+      "loss": 0.0633,
       "step": 470
     },
     {
       "epoch": 3.05,
-      "grad_norm": 0.10638546943664551,
+      "grad_norm": 0.07878056168556213,
       "learning_rate": 1.1111111111111112e-05,
-      "loss": 0.0355,
+      "loss": 0.02,
       "step": 480
     },
     {
       "epoch": 3.066666666666667,
-      "grad_norm": 0.10515712946653366,
+      "grad_norm": 0.10859273374080658,
       "learning_rate": 1.0185185185185185e-05,
-      "loss": 0.066,
+      "loss": 0.046,
       "step": 490
     },
     {
       "epoch": 3.0833333333333335,
-      "grad_norm": 0.4172205328941345,
+      "grad_norm": 0.13690859079360962,
       "learning_rate": 9.259259259259259e-06,
-      "loss": 0.0395,
+      "loss": 0.0805,
       "step": 500
     },
     {
       "epoch": 3.1,
-      "grad_norm": 38.50326156616211,
+      "grad_norm": 11.356720924377441,
       "learning_rate": 8.333333333333334e-06,
-      "loss": 0.3863,
+      "loss": 0.0611,
       "step": 510
     },
     {
       "epoch": 3.1166666666666667,
-      "grad_norm": 0.11968211829662323,
+      "grad_norm": 0.12047996371984482,
       "learning_rate": 7.4074074074074075e-06,
-      "loss": 0.1334,
+      "loss": 0.1968,
       "step": 520
     },
     {
       "epoch": 3.1333333333333333,
-      "grad_norm": 7.747541427612305,
+      "grad_norm": 0.7145315408706665,
       "learning_rate": 6.481481481481481e-06,
-      "loss": 0.0893,
+      "loss": 0.2474,
       "step": 530
     },
     {
       "epoch": 3.15,
-      "grad_norm": 0.860350489616394,
+      "grad_norm": 0.5600405931472778,
       "learning_rate": 5.555555555555556e-06,
-      "loss": 0.3064,
+      "loss": 0.2428,
       "step": 540
     },
     {
       "epoch": 3.1666666666666665,
-      "grad_norm": 1.3779197931289673,
+      "grad_norm": 0.8212262988090515,
       "learning_rate": 4.6296296296296296e-06,
-      "loss": 0.2478,
+      "loss": 0.2353,
       "step": 550
     },
     {
       "epoch": 3.183333333333333,
-      "grad_norm": 0.1085130125284195,
+      "grad_norm": 0.07675948739051819,
       "learning_rate": 3.7037037037037037e-06,
-      "loss": 0.302,
+      "loss": 0.3411,
       "step": 560
     },
     {
       "epoch": 3.2,
-      "grad_norm": 0.2623406648635864,
+      "grad_norm": 0.13475623726844788,
       "learning_rate": 2.777777777777778e-06,
-      "loss": 0.0376,
+      "loss": 0.2851,
       "step": 570
     },
     {
       "epoch": 3.216666666666667,
-      "grad_norm": 0.7496708631515503,
+      "grad_norm": 0.40810149908065796,
       "learning_rate": 1.8518518518518519e-06,
-      "loss": 0.0716,
+      "loss": 0.1894,
       "step": 580
     },
     {
       "epoch": 3.2333333333333334,
-      "grad_norm": 0.5140882134437561,
+      "grad_norm": 0.12464094907045364,
       "learning_rate": 9.259259259259259e-07,
-      "loss": 0.4109,
+      "loss": 0.4717,
       "step": 590
     },
     {
       "epoch": 3.25,
-      "grad_norm": 0.12395908683538437,
+      "grad_norm": 0.10543236881494522,
       "learning_rate": 0.0,
-      "loss": 0.1347,
+      "loss": 0.0447,
       "step": 600
     },
     {
       "epoch": 3.25,
-      "eval_accuracy": 0.9142857142857143,
-      "eval_loss": 0.28625765442848206,
-      "eval_runtime": 22.6931,
-      "eval_samples_per_second": 3.085,
-      "eval_steps_per_second": 1.542,
+      "eval_accuracy": 0.8714285714285714,
+      "eval_loss": 0.3807193338871002,
+      "eval_runtime": 20.385,
+      "eval_samples_per_second": 3.434,
+      "eval_steps_per_second": 1.717,
       "step": 600
     },
     {
       "epoch": 3.25,
       "step": 600,
       "total_flos": 1.495384188125184e+18,
-      "train_loss": 0.8924319915970167,
-      "train_runtime": 896.196,
-      "train_samples_per_second": 1.339,
-      "train_steps_per_second": 0.669
+      "train_loss": 0.8200334679583708,
+      "train_runtime": 930.7335,
+      "train_samples_per_second": 1.289,
+      "train_steps_per_second": 0.645
     },
     {
       "epoch": 3.25,
-      "eval_accuracy": 0.9032258064516129,
-      "eval_loss": 0.3167099952697754,
-      "eval_runtime": 48.193,
-      "eval_samples_per_second": 3.216,
-      "eval_steps_per_second": 1.618,
+      "eval_accuracy": 0.8903225806451613,
+      "eval_loss": 0.35323619842529297,
+      "eval_runtime": 44.4479,
+      "eval_samples_per_second": 3.487,
+      "eval_steps_per_second": 1.755,
       "step": 600
     },
     {
       "epoch": 3.25,
-      "eval_accuracy": 0.9032258064516129,
-      "eval_loss": 0.3167099952697754,
-      "eval_runtime": 46.9133,
-      "eval_samples_per_second": 3.304,
-      "eval_steps_per_second": 1.663,
+      "eval_accuracy": 0.8903225806451613,
+      "eval_loss": 0.3532361686229706,
+      "eval_runtime": 44.555,
+      "eval_samples_per_second": 3.479,
+      "eval_steps_per_second": 1.751,
       "step": 600
     }
   ],