End of training

README.md

@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 2.0503
-- Accuracy: 0.6439
+- Loss: 2.0762
+- Accuracy: 0.6393
 
 ## Model description
 

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.2442307692307693,
-    "eval_accuracy": 0.6320582877959927,
-    "eval_loss": 1.9572811126708984,
-    "eval_runtime": 146.1218,
-    "eval_samples_per_second": 3.757,
-    "eval_steps_per_second": 0.472
+    "epoch": 2.3282051282051284,
+    "eval_accuracy": 0.639344262295082,
+    "eval_loss": 2.076159954071045,
+    "eval_runtime": 145.7908,
+    "eval_samples_per_second": 3.766,
+    "eval_steps_per_second": 0.473
 }

runs/Jan08_20-28-15_5a3016f64bbf/events.out.tfevents.1736370286.5a3016f64bbf.332.4

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b8da9ed9c73210f936b2678c82a6f6f55bea6a60ff4c717f06a4dc1419425a65
-size 411
+oid sha256:47b19653c1c6cdcd71b52310712cadc42fd870acbae5547ad645d07c2dd317a4
+size 734

test_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.2442307692307693,
-    "eval_accuracy": 0.6320582877959927,
-    "eval_loss": 1.9572811126708984,
-    "eval_runtime": 146.1218,
-    "eval_samples_per_second": 3.757,
-    "eval_steps_per_second": 0.472
+    "epoch": 2.3282051282051284,
+    "eval_accuracy": 0.639344262295082,
+    "eval_loss": 2.076159954071045,
+    "eval_runtime": 145.7908,
+    "eval_samples_per_second": 3.766,
+    "eval_steps_per_second": 0.473
 }

trainer_state.json

@@ -1,443 +1,343 @@
 {
-  "best_metric": 0.820627802690583,
-  "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-520",
-  "epoch": 3.2442307692307693,
+  "best_metric": 0.8475336322869955,
+  "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-390",
+  "epoch": 2.3282051282051284,
   "eval_steps": 500,
-  "global_step": 520,
+  "global_step": 390,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.019230769230769232,
-      "grad_norm": 8.085018157958984,
-      "learning_rate": 9.615384615384616e-06,
-      "loss": 3.4301,
+      "epoch": 0.02564102564102564,
+      "grad_norm": 9.303430557250977,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 0.191,
       "step": 10
     },
     {
-      "epoch": 0.038461538461538464,
-      "grad_norm": 8.687586784362793,
-      "learning_rate": 1.923076923076923e-05,
-      "loss": 3.3945,
+      "epoch": 0.05128205128205128,
+      "grad_norm": 17.580381393432617,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 0.1535,
       "step": 20
     },
     {
-      "epoch": 0.057692307692307696,
-      "grad_norm": 9.751335144042969,
-      "learning_rate": 2.8846153846153845e-05,
-      "loss": 3.426,
+      "epoch": 0.07692307692307693,
+      "grad_norm": 3.0345261096954346,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.1446,
       "step": 30
     },
     {
-      "epoch": 0.07692307692307693,
-      "grad_norm": 8.339034080505371,
-      "learning_rate": 3.846153846153846e-05,
-      "loss": 3.4135,
+      "epoch": 0.10256410256410256,
+      "grad_norm": 5.7094550132751465,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 0.1564,
       "step": 40
     },
     {
-      "epoch": 0.09615384615384616,
-      "grad_norm": 7.60875129699707,
-      "learning_rate": 4.8076923076923084e-05,
-      "loss": 3.295,
+      "epoch": 0.1282051282051282,
+      "grad_norm": 1.9542657136917114,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 0.1486,
       "step": 50
     },
     {
-      "epoch": 0.11538461538461539,
-      "grad_norm": 8.210917472839355,
-      "learning_rate": 4.9145299145299147e-05,
-      "loss": 3.2379,
+      "epoch": 0.15384615384615385,
+      "grad_norm": 8.122542381286621,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.1577,
       "step": 60
     },
     {
-      "epoch": 0.1346153846153846,
-      "grad_norm": 8.948013305664062,
-      "learning_rate": 4.8076923076923084e-05,
-      "loss": 3.181,
+      "epoch": 0.1794871794871795,
+      "grad_norm": 9.375791549682617,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 0.1942,
       "step": 70
     },
     {
-      "epoch": 0.15384615384615385,
-      "grad_norm": 9.540766716003418,
-      "learning_rate": 4.700854700854701e-05,
-      "loss": 3.0669,
+      "epoch": 0.20512820512820512,
+      "grad_norm": 11.179394721984863,
+      "learning_rate": 4.967948717948718e-06,
+      "loss": 0.2429,
       "step": 80
     },
     {
-      "epoch": 0.17307692307692307,
-      "grad_norm": 10.327016830444336,
-      "learning_rate": 4.594017094017094e-05,
-      "loss": 2.9399,
+      "epoch": 0.23076923076923078,
+      "grad_norm": 11.36959457397461,
+      "learning_rate": 4.807692307692308e-06,
+      "loss": 0.1481,
       "step": 90
     },
     {
-      "epoch": 0.19230769230769232,
-      "grad_norm": 10.140628814697266,
-      "learning_rate": 4.4871794871794874e-05,
-      "loss": 2.7217,
+      "epoch": 0.2564102564102564,
+      "grad_norm": 1.9242831468582153,
+      "learning_rate": 4.647435897435898e-06,
+      "loss": 0.1894,
       "step": 100
     },
     {
-      "epoch": 0.21153846153846154,
-      "grad_norm": 10.16087818145752,
-      "learning_rate": 4.3803418803418805e-05,
-      "loss": 2.5126,
+      "epoch": 0.28205128205128205,
+      "grad_norm": 3.764237642288208,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 0.1557,
       "step": 110
     },
     {
-      "epoch": 0.23076923076923078,
-      "grad_norm": 9.189348220825195,
-      "learning_rate": 4.2735042735042735e-05,
-      "loss": 2.4065,
+      "epoch": 0.3076923076923077,
+      "grad_norm": 2.1338346004486084,
+      "learning_rate": 4.326923076923077e-06,
+      "loss": 0.1373,
       "step": 120
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 11.86257553100586,
-      "learning_rate": 4.166666666666667e-05,
-      "loss": 2.1315,
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.327778935432434,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.1658,
       "step": 130
     },
     {
-      "epoch": 0.2519230769230769,
-      "eval_accuracy": 0.38565022421524664,
-      "eval_loss": 2.2071354389190674,
-      "eval_runtime": 59.4454,
-      "eval_samples_per_second": 3.751,
-      "eval_steps_per_second": 0.471,
+      "epoch": 0.33589743589743587,
+      "eval_accuracy": 0.8251121076233184,
+      "eval_loss": 0.5804738998413086,
+      "eval_runtime": 56.2864,
+      "eval_samples_per_second": 3.962,
+      "eval_steps_per_second": 0.497,
       "step": 131
     },
     {
-      "epoch": 1.0173076923076922,
-      "grad_norm": 10.583504676818848,
-      "learning_rate": 4.05982905982906e-05,
-      "loss": 1.9809,
+      "epoch": 1.023076923076923,
+      "grad_norm": 0.7002243399620056,
+      "learning_rate": 4.006410256410257e-06,
+      "loss": 0.1178,
       "step": 140
     },
     {
-      "epoch": 1.0365384615384616,
-      "grad_norm": 12.537158966064453,
-      "learning_rate": 3.952991452991453e-05,
-      "loss": 1.8472,
+      "epoch": 1.0487179487179488,
+      "grad_norm": 1.249237298965454,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.1735,
       "step": 150
     },
     {
-      "epoch": 1.0557692307692308,
-      "grad_norm": 13.388022422790527,
-      "learning_rate": 3.846153846153846e-05,
-      "loss": 1.6871,
+      "epoch": 1.0743589743589743,
+      "grad_norm": 8.169507026672363,
+      "learning_rate": 3.6858974358974363e-06,
+      "loss": 0.1527,
       "step": 160
     },
     {
-      "epoch": 1.075,
-      "grad_norm": 20.201147079467773,
-      "learning_rate": 3.739316239316239e-05,
-      "loss": 1.7488,
+      "epoch": 1.1,
+      "grad_norm": 4.84148645401001,
+      "learning_rate": 3.5256410256410263e-06,
+      "loss": 0.1313,
       "step": 170
     },
     {
-      "epoch": 1.0942307692307693,
-      "grad_norm": 14.368993759155273,
-      "learning_rate": 3.6324786324786323e-05,
-      "loss": 1.6513,
+      "epoch": 1.1256410256410256,
+      "grad_norm": 2.1306402683258057,
+      "learning_rate": 3.365384615384616e-06,
+      "loss": 0.0957,
       "step": 180
     },
     {
-      "epoch": 1.1134615384615385,
-      "grad_norm": 11.838800430297852,
-      "learning_rate": 3.525641025641026e-05,
-      "loss": 1.5155,
+      "epoch": 1.1512820512820512,
+      "grad_norm": 1.54770827293396,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 0.1137,
       "step": 190
     },
     {
-      "epoch": 1.1326923076923077,
-      "grad_norm": 12.411248207092285,
-      "learning_rate": 3.418803418803419e-05,
-      "loss": 1.6197,
+      "epoch": 1.176923076923077,
+      "grad_norm": 13.978232383728027,
+      "learning_rate": 3.044871794871795e-06,
+      "loss": 0.1326,
       "step": 200
     },
     {
-      "epoch": 1.1519230769230768,
-      "grad_norm": 9.437255859375,
-      "learning_rate": 3.311965811965812e-05,
-      "loss": 1.356,
+      "epoch": 1.2025641025641025,
+      "grad_norm": 4.250465393066406,
+      "learning_rate": 2.8846153846153845e-06,
+      "loss": 0.0616,
       "step": 210
     },
     {
-      "epoch": 1.1711538461538462,
-      "grad_norm": 9.415695190429688,
-      "learning_rate": 3.205128205128206e-05,
-      "loss": 1.2385,
+      "epoch": 1.2282051282051283,
+      "grad_norm": 1.337228775024414,
+      "learning_rate": 2.7243589743589744e-06,
+      "loss": 0.0948,
       "step": 220
     },
     {
-      "epoch": 1.1903846153846154,
-      "grad_norm": 14.387253761291504,
-      "learning_rate": 3.098290598290599e-05,
-      "loss": 1.1676,
+      "epoch": 1.2538461538461538,
+      "grad_norm": 11.519142150878906,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 0.0906,
       "step": 230
     },
     {
-      "epoch": 1.2096153846153845,
-      "grad_norm": 18.609485626220703,
-      "learning_rate": 2.9914529914529915e-05,
-      "loss": 1.0883,
+      "epoch": 1.2794871794871794,
+      "grad_norm": 9.989028930664062,
+      "learning_rate": 2.403846153846154e-06,
+      "loss": 0.1384,
       "step": 240
     },
     {
-      "epoch": 1.228846153846154,
-      "grad_norm": 8.951836585998535,
-      "learning_rate": 2.8846153846153845e-05,
-      "loss": 1.0067,
+      "epoch": 1.3051282051282052,
+      "grad_norm": 0.850464940071106,
+      "learning_rate": 2.243589743589744e-06,
+      "loss": 0.1329,
       "step": 250
     },
     {
-      "epoch": 1.248076923076923,
-      "grad_norm": 9.234614372253418,
-      "learning_rate": 2.777777777777778e-05,
-      "loss": 1.1581,
+      "epoch": 1.3307692307692307,
+      "grad_norm": 2.935410976409912,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.1039,
       "step": 260
     },
     {
-      "epoch": 1.251923076923077,
-      "eval_accuracy": 0.6547085201793722,
-      "eval_loss": 1.158636212348938,
-      "eval_runtime": 59.7594,
-      "eval_samples_per_second": 3.732,
-      "eval_steps_per_second": 0.469,
+      "epoch": 1.3358974358974358,
+      "eval_accuracy": 0.8251121076233184,
+      "eval_loss": 0.5684685707092285,
+      "eval_runtime": 60.4317,
+      "eval_samples_per_second": 3.69,
+      "eval_steps_per_second": 0.463,
       "step": 262
     },
     {
-      "epoch": 2.0153846153846153,
-      "grad_norm": 9.139333724975586,
-      "learning_rate": 2.670940170940171e-05,
-      "loss": 0.8048,
+      "epoch": 2.0205128205128204,
+      "grad_norm": 2.507692575454712,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.1267,
       "step": 270
     },
     {
-      "epoch": 2.0346153846153845,
-      "grad_norm": 19.066368103027344,
-      "learning_rate": 2.564102564102564e-05,
-      "loss": 1.0916,
+      "epoch": 2.046153846153846,
+      "grad_norm": 9.046551704406738,
+      "learning_rate": 1.7628205128205131e-06,
+      "loss": 0.0909,
       "step": 280
     },
     {
-      "epoch": 2.0538461538461537,
-      "grad_norm": 5.820519924163818,
-      "learning_rate": 2.4572649572649573e-05,
-      "loss": 0.8838,
+      "epoch": 2.071794871794872,
+      "grad_norm": 1.059535264968872,
+      "learning_rate": 1.602564102564103e-06,
+      "loss": 0.1108,
       "step": 290
     },
     {
-      "epoch": 2.0730769230769233,
-      "grad_norm": 9.551618576049805,
-      "learning_rate": 2.3504273504273504e-05,
-      "loss": 0.7372,
+      "epoch": 2.0974358974358975,
+      "grad_norm": 7.851165294647217,
+      "learning_rate": 1.4423076923076922e-06,
+      "loss": 0.0695,
       "step": 300
     },
     {
-      "epoch": 2.0923076923076924,
-      "grad_norm": 5.678286075592041,
-      "learning_rate": 2.2435897435897437e-05,
-      "loss": 0.7344,
+      "epoch": 2.123076923076923,
+      "grad_norm": 1.073171615600586,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 0.1031,
       "step": 310
     },
     {
-      "epoch": 2.1115384615384616,
-      "grad_norm": 7.4141411781311035,
-      "learning_rate": 2.1367521367521368e-05,
-      "loss": 0.5431,
+      "epoch": 2.1487179487179486,
+      "grad_norm": 2.39035964012146,
+      "learning_rate": 1.121794871794872e-06,
+      "loss": 0.115,
       "step": 320
     },
     {
-      "epoch": 2.1307692307692307,
-      "grad_norm": 12.246278762817383,
-      "learning_rate": 2.02991452991453e-05,
-      "loss": 0.7999,
+      "epoch": 2.174358974358974,
+      "grad_norm": 1.7656793594360352,
+      "learning_rate": 9.615384615384617e-07,
+      "loss": 0.0919,
       "step": 330
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 9.39754581451416,
-      "learning_rate": 1.923076923076923e-05,
-      "loss": 0.7329,
+      "epoch": 2.2,
+      "grad_norm": 19.47994613647461,
+      "learning_rate": 8.012820512820515e-07,
+      "loss": 0.1017,
       "step": 340
     },
     {
-      "epoch": 2.169230769230769,
-      "grad_norm": 11.727778434753418,
-      "learning_rate": 1.8162393162393162e-05,
-      "loss": 0.5766,
+      "epoch": 2.2256410256410257,
+      "grad_norm": 14.330218315124512,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 0.0826,
       "step": 350
     },
     {
-      "epoch": 2.1884615384615387,
-      "grad_norm": 5.652421474456787,
-      "learning_rate": 1.7094017094017095e-05,
-      "loss": 0.6079,
+      "epoch": 2.2512820512820513,
+      "grad_norm": 3.4437031745910645,
+      "learning_rate": 4.807692307692308e-07,
+      "loss": 0.1271,
       "step": 360
     },
     {
-      "epoch": 2.207692307692308,
-      "grad_norm": 18.59898567199707,
-      "learning_rate": 1.602564102564103e-05,
-      "loss": 0.5848,
+      "epoch": 2.276923076923077,
+      "grad_norm": 3.954479932785034,
+      "learning_rate": 3.205128205128205e-07,
+      "loss": 0.1132,
       "step": 370
     },
     {
-      "epoch": 2.226923076923077,
-      "grad_norm": 9.42827033996582,
-      "learning_rate": 1.4957264957264958e-05,
-      "loss": 0.4852,
+      "epoch": 2.3025641025641024,
+      "grad_norm": 0.6898751854896545,
+      "learning_rate": 1.6025641025641025e-07,
+      "loss": 0.0906,
       "step": 380
     },
     {
-      "epoch": 2.246153846153846,
-      "grad_norm": 10.546754837036133,
-      "learning_rate": 1.388888888888889e-05,
-      "loss": 0.6854,
+      "epoch": 2.3282051282051284,
+      "grad_norm": 2.044902801513672,
+      "learning_rate": 0.0,
+      "loss": 0.0832,
       "step": 390
     },
     {
-      "epoch": 2.251923076923077,
-      "eval_accuracy": 0.7623318385650224,
-      "eval_loss": 0.816370964050293,
-      "eval_runtime": 60.0828,
-      "eval_samples_per_second": 3.712,
-      "eval_steps_per_second": 0.466,
-      "step": 393
-    },
-    {
-      "epoch": 3.0134615384615384,
-      "grad_norm": 7.366909980773926,
-      "learning_rate": 1.282051282051282e-05,
-      "loss": 0.4249,
-      "step": 400
-    },
-    {
-      "epoch": 3.0326923076923076,
-      "grad_norm": 19.081708908081055,
-      "learning_rate": 1.1752136752136752e-05,
-      "loss": 0.4456,
-      "step": 410
-    },
-    {
-      "epoch": 3.0519230769230767,
-      "grad_norm": 6.918426990509033,
-      "learning_rate": 1.0683760683760684e-05,
-      "loss": 0.4338,
-      "step": 420
-    },
-    {
-      "epoch": 3.0711538461538463,
-      "grad_norm": 20.953861236572266,
-      "learning_rate": 9.615384615384616e-06,
-      "loss": 0.4404,
-      "step": 430
-    },
-    {
-      "epoch": 3.0903846153846155,
-      "grad_norm": 7.814371585845947,
-      "learning_rate": 8.547008547008548e-06,
-      "loss": 0.415,
-      "step": 440
-    },
-    {
-      "epoch": 3.1096153846153847,
-      "grad_norm": 5.119196891784668,
-      "learning_rate": 7.478632478632479e-06,
-      "loss": 0.387,
-      "step": 450
-    },
-    {
-      "epoch": 3.128846153846154,
-      "grad_norm": 12.346413612365723,
-      "learning_rate": 6.41025641025641e-06,
-      "loss": 0.3826,
-      "step": 460
-    },
-    {
-      "epoch": 3.148076923076923,
-      "grad_norm": 13.700543403625488,
-      "learning_rate": 5.341880341880342e-06,
-      "loss": 0.2908,
-      "step": 470
-    },
-    {
-      "epoch": 3.167307692307692,
-      "grad_norm": 11.321188926696777,
-      "learning_rate": 4.273504273504274e-06,
-      "loss": 0.4446,
-      "step": 480
-    },
-    {
-      "epoch": 3.1865384615384613,
-      "grad_norm": 11.503890037536621,
-      "learning_rate": 3.205128205128205e-06,
-      "loss": 0.3246,
-      "step": 490
+      "epoch": 2.3282051282051284,
+      "eval_accuracy": 0.8475336322869955,
+      "eval_loss": 0.5565493702888489,
+      "eval_runtime": 62.3378,
+      "eval_samples_per_second": 3.577,
+      "eval_steps_per_second": 0.449,
+      "step": 390
     },
     {
-      "epoch": 3.205769230769231,
-      "grad_norm": 9.901238441467285,
-      "learning_rate": 2.136752136752137e-06,
-      "loss": 0.3634,
-      "step": 500
+      "epoch": 2.3282051282051284,
+      "step": 390,
+      "total_flos": 3.8861995625761997e+18,
+      "train_loss": 0.12900116260235125,
+      "train_runtime": 2004.3611,
+      "train_samples_per_second": 1.557,
+      "train_steps_per_second": 0.195
     },
     {
-      "epoch": 3.225,
-      "grad_norm": 4.689971446990967,
-      "learning_rate": 1.0683760683760685e-06,
-      "loss": 0.3183,
-      "step": 510
+      "epoch": 2.3282051282051284,
+      "eval_accuracy": 0.6438848920863309,
+      "eval_loss": 2.050309658050537,
+      "eval_runtime": 149.2566,
+      "eval_samples_per_second": 3.725,
+      "eval_steps_per_second": 0.469,
+      "step": 390
     },
     {
-      "epoch": 3.2442307692307693,
-      "grad_norm": 17.20209503173828,
-      "learning_rate": 0.0,
-      "loss": 0.3349,
-      "step": 520
-    },
-    {
-      "epoch": 3.2442307692307693,
-      "eval_accuracy": 0.820627802690583,
-      "eval_loss": 0.6745598912239075,
-      "eval_runtime": 60.6402,
-      "eval_samples_per_second": 3.677,
-      "eval_steps_per_second": 0.462,
-      "step": 520
-    },
-    {
-      "epoch": 3.2442307692307693,
-      "step": 520,
-      "total_flos": 5.181183958187704e+18,
-      "train_loss": 1.3941553766910846,
-      "train_runtime": 2772.4544,
-      "train_samples_per_second": 1.5,
-      "train_steps_per_second": 0.188
-    },
-    {
-      "epoch": 3.2442307692307693,
-      "eval_accuracy": 0.6320582877959927,
-      "eval_loss": 1.9572809934616089,
-      "eval_runtime": 147.8534,
-      "eval_samples_per_second": 3.713,
-      "eval_steps_per_second": 0.467,
-      "step": 520
-    },
-    {
-      "epoch": 3.2442307692307693,
-      "eval_accuracy": 0.6320582877959927,
-      "eval_loss": 1.9572811126708984,
-      "eval_runtime": 146.1218,
-      "eval_samples_per_second": 3.757,
-      "eval_steps_per_second": 0.472,
-      "step": 520
+      "epoch": 2.3282051282051284,
+      "eval_accuracy": 0.639344262295082,
+      "eval_loss": 2.076159954071045,
+      "eval_runtime": 145.7908,
+      "eval_samples_per_second": 3.766,
+      "eval_steps_per_second": 0.473,
+      "step": 390
     }
   ],
   "logging_steps": 10,
-  "max_steps": 520,
+  "max_steps": 390,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 9223372036854775807,
   "save_steps": 500,
@@ -453,7 +353,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 5.181183958187704e+18,
+  "total_flos": 3.8861995625761997e+18,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null