End of training

all_results.json

@@ -1,8 +1,8 @@
 {
  "epoch": 3.25,
- "eval_accuracy": 0.9096774193548387,
- "eval_loss": 0.27532899379730225,
- "eval_runtime": 44.6134,
- "eval_samples_per_second": 3.474,
- "eval_steps_per_second": 1.748
+ "eval_accuracy": 0.8284023668639053,
+ "eval_loss": 0.45989519357681274,
+ "eval_runtime": 94.1469,
+ "eval_samples_per_second": 1.795,
+ "eval_steps_per_second": 0.903
 }

runs/May11_18-07-17_7ffe87b6d6e8/events.out.tfevents.1683829879.7ffe87b6d6e8.6406.2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a89529d8cbce5e82369fc4f7c1a80fb8967a944339aeb5e1f9317c696f140fcd
-size 411
+oid sha256:432986d9b7a1a03f4af6f7b7c987efb96e6b53b13898824fdd0674920286b1ce
+size 734

test_results.json

@@ -1,8 +1,8 @@
 {
  "epoch": 3.25,
- "eval_accuracy": 0.9096774193548387,
- "eval_loss": 0.27532899379730225,
- "eval_runtime": 44.6134,
- "eval_samples_per_second": 3.474,
- "eval_steps_per_second": 1.748
+ "eval_accuracy": 0.8284023668639053,
+ "eval_loss": 0.45989519357681274,
+ "eval_runtime": 94.1469,
+ "eval_samples_per_second": 1.795,
+ "eval_steps_per_second": 0.903
 }

trainer_state.json

@@ -1,439 +1,439 @@
 {
- "best_metric": 0.9,
- "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-600",
+ "best_metric": 0.7738095238095238,
+ "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-604",
  "epoch": 3.25,
- "global_step": 600,
+ "global_step": 604,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
  {
  "epoch": 0.02,
- "learning_rate": 8.333333333333334e-06,
- "loss": 2.3492,
+ "learning_rate": 8.196721311475409e-06,
+ "loss": 2.4329,
  "step": 10
  },
  {
  "epoch": 0.03,
- "learning_rate": 1.6666666666666667e-05,
- "loss": 2.3466,
+ "learning_rate": 1.6393442622950818e-05,
+ "loss": 2.4383,
  "step": 20
  },
  {
  "epoch": 0.05,
- "learning_rate": 2.5e-05,
- "loss": 2.3064,
+ "learning_rate": 2.459016393442623e-05,
+ "loss": 2.428,
  "step": 30
  },
  {
  "epoch": 0.07,
- "learning_rate": 3.3333333333333335e-05,
- "loss": 2.1908,
+ "learning_rate": 3.2786885245901635e-05,
+ "loss": 2.3514,
  "step": 40
  },
  {
  "epoch": 0.08,
- "learning_rate": 4.166666666666667e-05,
- "loss": 2.1392,
+ "learning_rate": 4.098360655737705e-05,
+ "loss": 2.4083,
  "step": 50
  },
  {
  "epoch": 0.1,
- "learning_rate": 5e-05,
- "loss": 2.207,
+ "learning_rate": 4.918032786885246e-05,
+ "loss": 2.2609,
  "step": 60
  },
  {
  "epoch": 0.12,
- "learning_rate": 4.9074074074074075e-05,
- "loss": 2.1197,
+ "learning_rate": 4.9171270718232046e-05,
+ "loss": 2.1355,
  "step": 70
  },
  {
  "epoch": 0.13,
- "learning_rate": 4.814814814814815e-05,
- "loss": 1.7111,
+ "learning_rate": 4.825046040515654e-05,
+ "loss": 2.0781,
  "step": 80
  },
  {
  "epoch": 0.15,
- "learning_rate": 4.722222222222222e-05,
- "loss": 1.7897,
+ "learning_rate": 4.732965009208104e-05,
+ "loss": 1.686,
  "step": 90
  },
  {
  "epoch": 0.17,
- "learning_rate": 4.62962962962963e-05,
- "loss": 1.7259,
+ "learning_rate": 4.6408839779005524e-05,
+ "loss": 1.5192,
  "step": 100
  },
  {
  "epoch": 0.18,
- "learning_rate": 4.5370370370370374e-05,
- "loss": 1.491,
+ "learning_rate": 4.5488029465930024e-05,
+ "loss": 1.7072,
  "step": 110
  },
  {
  "epoch": 0.2,
- "learning_rate": 4.4444444444444447e-05,
- "loss": 1.6251,
+ "learning_rate": 4.4567219152854516e-05,
+ "loss": 1.93,
  "step": 120
  },
  {
  "epoch": 0.22,
- "learning_rate": 4.351851851851852e-05,
- "loss": 1.5884,
+ "learning_rate": 4.364640883977901e-05,
+ "loss": 1.4418,
  "step": 130
  },
  {
  "epoch": 0.23,
- "learning_rate": 4.259259259259259e-05,
- "loss": 1.3184,
+ "learning_rate": 4.27255985267035e-05,
+ "loss": 1.6683,
  "step": 140
  },
  {
  "epoch": 0.25,
- "learning_rate": 4.166666666666667e-05,
- "loss": 1.5351,
+ "learning_rate": 4.1804788213627995e-05,
+ "loss": 1.3861,
  "step": 150
  },
  {
  "epoch": 0.25,
- "eval_accuracy": 0.5285714285714286,
- "eval_loss": 1.7103750705718994,
- "eval_runtime": 20.2519,
- "eval_samples_per_second": 3.456,
- "eval_steps_per_second": 1.728,
- "step": 150
+ "eval_accuracy": 0.40476190476190477,
+ "eval_loss": 1.518173336982727,
+ "eval_runtime": 74.2638,
+ "eval_samples_per_second": 1.131,
+ "eval_steps_per_second": 0.566,
+ "step": 151
  },
  {
- "epoch": 1.02,
- "learning_rate": 4.074074074074074e-05,
- "loss": 1.3053,
+ "epoch": 1.01,
+ "learning_rate": 4.088397790055249e-05,
+ "loss": 1.1002,
  "step": 160
  },
  {
  "epoch": 1.03,
- "learning_rate": 3.981481481481482e-05,
- "loss": 1.1629,
+ "learning_rate": 3.996316758747698e-05,
+ "loss": 1.0042,
  "step": 170
  },
  {
  "epoch": 1.05,
- "learning_rate": 3.888888888888889e-05,
- "loss": 0.9645,
+ "learning_rate": 3.904235727440147e-05,
+ "loss": 0.6699,
  "step": 180
  },
  {
- "epoch": 1.07,
- "learning_rate": 3.7962962962962964e-05,
- "loss": 0.7846,
+ "epoch": 1.06,
+ "learning_rate": 3.812154696132597e-05,
+ "loss": 0.7446,
  "step": 190
  },
  {
  "epoch": 1.08,
- "learning_rate": 3.7037037037037037e-05,
- "loss": 0.9741,
+ "learning_rate": 3.720073664825046e-05,
+ "loss": 0.575,
  "step": 200
  },
  {
  "epoch": 1.1,
- "learning_rate": 3.611111111111111e-05,
- "loss": 0.6873,
+ "learning_rate": 3.627992633517496e-05,
+ "loss": 1.0777,
  "step": 210
  },
  {
- "epoch": 1.12,
- "learning_rate": 3.518518518518519e-05,
- "loss": 0.9336,
+ "epoch": 1.11,
+ "learning_rate": 3.535911602209945e-05,
+ "loss": 0.8046,
  "step": 220
  },
  {
  "epoch": 1.13,
- "learning_rate": 3.425925925925926e-05,
- "loss": 0.8429,
+ "learning_rate": 3.4438305709023944e-05,
+ "loss": 1.3932,
  "step": 230
  },
  {
  "epoch": 1.15,
- "learning_rate": 3.3333333333333335e-05,
- "loss": 0.7939,
+ "learning_rate": 3.3517495395948436e-05,
+ "loss": 1.0903,
  "step": 240
  },
  {
- "epoch": 1.17,
- "learning_rate": 3.240740740740741e-05,
- "loss": 0.6433,
+ "epoch": 1.16,
+ "learning_rate": 3.259668508287293e-05,
+ "loss": 0.9612,
  "step": 250
  },
  {
  "epoch": 1.18,
- "learning_rate": 3.148148148148148e-05,
- "loss": 0.6427,
+ "learning_rate": 3.167587476979743e-05,
+ "loss": 1.0058,
  "step": 260
  },
  {
  "epoch": 1.2,
- "learning_rate": 3.055555555555556e-05,
- "loss": 0.8235,
+ "learning_rate": 3.0755064456721915e-05,
+ "loss": 1.0088,
  "step": 270
  },
  {
- "epoch": 1.22,
- "learning_rate": 2.962962962962963e-05,
- "loss": 0.3465,
+ "epoch": 1.21,
+ "learning_rate": 2.983425414364641e-05,
+ "loss": 0.5327,
  "step": 280
  },
  {
  "epoch": 1.23,
- "learning_rate": 2.8703703703703706e-05,
- "loss": 0.77,
+ "learning_rate": 2.8913443830570904e-05,
+ "loss": 0.5463,
  "step": 290
  },
  {
  "epoch": 1.25,
- "learning_rate": 2.777777777777778e-05,
- "loss": 0.8621,
+ "learning_rate": 2.7992633517495396e-05,
+ "loss": 0.6672,
  "step": 300
  },
  {
  "epoch": 1.25,
- "eval_accuracy": 0.7428571428571429,
- "eval_loss": 0.7072036266326904,
- "eval_runtime": 19.9162,
- "eval_samples_per_second": 3.515,
- "eval_steps_per_second": 1.757,
- "step": 300
+ "eval_accuracy": 0.7023809523809523,
+ "eval_loss": 0.9395186901092529,
+ "eval_runtime": 73.8401,
+ "eval_samples_per_second": 1.138,
+ "eval_steps_per_second": 0.569,
+ "step": 302
  },
  {
- "epoch": 2.02,
- "learning_rate": 2.6851851851851855e-05,
- "loss": 0.309,
+ "epoch": 2.01,
+ "learning_rate": 2.707182320441989e-05,
+ "loss": 0.1746,
  "step": 310
  },
  {
  "epoch": 2.03,
- "learning_rate": 2.5925925925925925e-05,
- "loss": 0.1178,
+ "learning_rate": 2.6151012891344385e-05,
+ "loss": 0.2385,
  "step": 320
  },
  {
  "epoch": 2.05,
- "learning_rate": 2.5e-05,
- "loss": 0.3755,
+ "learning_rate": 2.523020257826888e-05,
+ "loss": 0.7082,
  "step": 330
  },
  {
- "epoch": 2.07,
- "learning_rate": 2.4074074074074074e-05,
- "loss": 0.2543,
+ "epoch": 2.06,
+ "learning_rate": 2.430939226519337e-05,
+ "loss": 0.4966,
  "step": 340
  },
  {
  "epoch": 2.08,
- "learning_rate": 2.314814814814815e-05,
- "loss": 0.2794,
+ "learning_rate": 2.3388581952117867e-05,
+ "loss": 0.74,
  "step": 350
  },
  {
  "epoch": 2.1,
- "learning_rate": 2.2222222222222223e-05,
- "loss": 0.3993,
+ "learning_rate": 2.246777163904236e-05,
+ "loss": 0.5615,
  "step": 360
  },
  {
- "epoch": 2.12,
- "learning_rate": 2.1296296296296296e-05,
- "loss": 0.1797,
+ "epoch": 2.11,
+ "learning_rate": 2.1546961325966852e-05,
+ "loss": 0.406,
  "step": 370
  },
  {
  "epoch": 2.13,
- "learning_rate": 2.037037037037037e-05,
- "loss": 0.2349,
+ "learning_rate": 2.0626151012891345e-05,
+ "loss": 0.2123,
  "step": 380
  },
  {
  "epoch": 2.15,
- "learning_rate": 1.9444444444444445e-05,
- "loss": 0.2238,
+ "learning_rate": 1.9705340699815838e-05,
+ "loss": 0.0971,
  "step": 390
  },
  {
- "epoch": 2.17,
- "learning_rate": 1.8518518518518518e-05,
- "loss": 0.3756,
+ "epoch": 2.16,
+ "learning_rate": 1.878453038674033e-05,
+ "loss": 0.2095,
  "step": 400
  },
  {
  "epoch": 2.18,
- "learning_rate": 1.7592592592592595e-05,
- "loss": 0.1291,
+ "learning_rate": 1.7863720073664823e-05,
+ "loss": 0.0643,
  "step": 410
  },
  {
  "epoch": 2.2,
- "learning_rate": 1.6666666666666667e-05,
- "loss": 0.2602,
+ "learning_rate": 1.694290976058932e-05,
+ "loss": 0.1121,
  "step": 420
  },
  {
- "epoch": 2.22,
- "learning_rate": 1.574074074074074e-05,
- "loss": 0.3749,
+ "epoch": 2.21,
+ "learning_rate": 1.6022099447513812e-05,
+ "loss": 0.4411,
  "step": 430
  },
  {
  "epoch": 2.23,
- "learning_rate": 1.4814814814814815e-05,
- "loss": 0.2041,
+ "learning_rate": 1.5101289134438307e-05,
+ "loss": 0.2766,
  "step": 440
  },
  {
  "epoch": 2.25,
- "learning_rate": 1.388888888888889e-05,
- "loss": 0.2002,
+ "learning_rate": 1.41804788213628e-05,
+ "loss": 0.1478,
  "step": 450
  },
  {
  "epoch": 2.25,
- "eval_accuracy": 0.8285714285714286,
- "eval_loss": 0.5905740261077881,
- "eval_runtime": 19.9694,
- "eval_samples_per_second": 3.505,
- "eval_steps_per_second": 1.753,
- "step": 450
+ "eval_accuracy": 0.7380952380952381,
+ "eval_loss": 0.7313230633735657,
+ "eval_runtime": 75.0778,
+ "eval_samples_per_second": 1.119,
+ "eval_steps_per_second": 0.559,
+ "step": 453
  },
  {
- "epoch": 3.02,
- "learning_rate": 1.2962962962962962e-05,
- "loss": 0.4304,
+ "epoch": 3.01,
+ "learning_rate": 1.3259668508287292e-05,
+ "loss": 0.2004,
  "step": 460
  },
  {
  "epoch": 3.03,
- "learning_rate": 1.2037037037037037e-05,
- "loss": 0.1548,
+ "learning_rate": 1.2338858195211787e-05,
+ "loss": 0.0466,
  "step": 470
  },
  {
- "epoch": 3.05,
- "learning_rate": 1.1111111111111112e-05,
- "loss": 0.0379,
+ "epoch": 3.04,
+ "learning_rate": 1.1418047882136281e-05,
+ "loss": 0.2506,
  "step": 480
  },
  {
- "epoch": 3.07,
- "learning_rate": 1.0185185185185185e-05,
- "loss": 0.0551,
+ "epoch": 3.06,
+ "learning_rate": 1.0497237569060774e-05,
+ "loss": 0.2664,
  "step": 490
  },
  {
  "epoch": 3.08,
- "learning_rate": 9.259259259259259e-06,
- "loss": 0.0582,
+ "learning_rate": 9.576427255985269e-06,
+ "loss": 0.3079,
  "step": 500
  },
  {
- "epoch": 3.1,
- "learning_rate": 8.333333333333334e-06,
- "loss": 0.1102,
+ "epoch": 3.09,
+ "learning_rate": 8.655616942909761e-06,
+ "loss": 0.3008,
  "step": 510
  },
  {
- "epoch": 3.12,
- "learning_rate": 7.4074074074074075e-06,
- "loss": 0.1212,
+ "epoch": 3.11,
+ "learning_rate": 7.734806629834254e-06,
+ "loss": 0.1139,
  "step": 520
  },
  {
  "epoch": 3.13,
- "learning_rate": 6.481481481481481e-06,
- "loss": 0.136,
+ "learning_rate": 6.8139963167587485e-06,
+ "loss": 0.271,
  "step": 530
  },
  {
- "epoch": 3.15,
- "learning_rate": 5.555555555555556e-06,
- "loss": 0.133,
+ "epoch": 3.14,
+ "learning_rate": 5.893186003683242e-06,
+ "loss": 0.104,
  "step": 540
  },
  {
- "epoch": 3.17,
- "learning_rate": 4.6296296296296296e-06,
- "loss": 0.2359,
+ "epoch": 3.16,
+ "learning_rate": 4.972375690607735e-06,
+ "loss": 0.191,
  "step": 550
  },
  {
  "epoch": 3.18,
- "learning_rate": 3.7037037037037037e-06,
- "loss": 0.2602,
+ "learning_rate": 4.0515653775322285e-06,
+ "loss": 0.1103,
  "step": 560
  },
  {
- "epoch": 3.2,
- "learning_rate": 2.777777777777778e-06,
- "loss": 0.1343,
+ "epoch": 3.19,
+ "learning_rate": 3.1307550644567225e-06,
+ "loss": 0.1846,
  "step": 570
  },
  {
- "epoch": 3.22,
- "learning_rate": 1.8518518518518519e-06,
- "loss": 0.137,
+ "epoch": 3.21,
+ "learning_rate": 2.2099447513812157e-06,
+ "loss": 0.1183,
  "step": 580
  },
  {
  "epoch": 3.23,
- "learning_rate": 9.259259259259259e-07,
- "loss": 0.2331,
+ "learning_rate": 1.289134438305709e-06,
+ "loss": 0.0612,
  "step": 590
  },
  {
- "epoch": 3.25,
- "learning_rate": 0.0,
- "loss": 0.4786,
+ "epoch": 3.24,
+ "learning_rate": 3.6832412523020263e-07,
+ "loss": 0.2596,
  "step": 600
  },
  {
  "epoch": 3.25,
- "eval_accuracy": 0.9,
- "eval_loss": 0.285513311624527,
- "eval_runtime": 19.8769,
- "eval_samples_per_second": 3.522,
- "eval_steps_per_second": 1.761,
- "step": 600
+ "eval_accuracy": 0.7738095238095238,
+ "eval_loss": 0.543230414390564,
+ "eval_runtime": 74.0438,
+ "eval_samples_per_second": 1.134,
+ "eval_steps_per_second": 0.567,
+ "step": 604
  },
  {
  "epoch": 3.25,
- "step": 600,
- "total_flos": 1.495384188125184e+18,
- "train_loss": 0.7935705941418807,
- "train_runtime": 854.0599,
- "train_samples_per_second": 1.405,
- "train_steps_per_second": 0.703
+ "step": 604,
+ "total_flos": 1.5053668400415375e+18,
+ "train_loss": 0.8410116481958636,
+ "train_runtime": 1073.1012,
+ "train_samples_per_second": 1.126,
+ "train_steps_per_second": 0.563
  },
  {
  "epoch": 3.25,
- "eval_accuracy": 0.9096774193548387,
- "eval_loss": 0.27532899379730225,
- "eval_runtime": 46.5951,
- "eval_samples_per_second": 3.327,
- "eval_steps_per_second": 1.674,
- "step": 600
+ "eval_accuracy": 0.8284023668639053,
+ "eval_loss": 0.4598952829837799,
+ "eval_runtime": 100.3611,
+ "eval_samples_per_second": 1.684,
+ "eval_steps_per_second": 0.847,
+ "step": 604
  },
  {
  "epoch": 3.25,
- "eval_accuracy": 0.9096774193548387,
- "eval_loss": 0.27532899379730225,
- "eval_runtime": 44.6134,
- "eval_samples_per_second": 3.474,
- "eval_steps_per_second": 1.748,
- "step": 600
+ "eval_accuracy": 0.8284023668639053,
+ "eval_loss": 0.45989519357681274,
+ "eval_runtime": 94.1469,
+ "eval_samples_per_second": 1.795,
+ "eval_steps_per_second": 0.903,
+ "step": 604
  }
  ],
- "max_steps": 600,
+ "max_steps": 604,
  "num_train_epochs": 9223372036854775807,
- "total_flos": 1.495384188125184e+18,
+ "total_flos": 1.5053668400415375e+18,
  "trial_name": null,
  "trial_params": null
 }