End of training

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.25,
-    "eval_accuracy": 0.8387096774193549,
-    "eval_loss": 0.6151520609855652,
-    "eval_runtime": 18.1208,
-    "eval_samples_per_second": 8.554,
-    "eval_steps_per_second": 8.554
+    "eval_accuracy": 0.8709677419354839,
+    "eval_loss": 0.41407766938209534,
+    "eval_runtime": 44.3003,
+    "eval_samples_per_second": 3.499,
+    "eval_steps_per_second": 3.499
 }

runs/Feb16_05-22-13_de4bb4676102/events.out.tfevents.1708062338.de4bb4676102.1456.2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a7b4a65f5dd858eb91f7803678b3b4274f86bfbc4b49ee343f2bdcca74bfac5
-size 411
+oid sha256:bd1bd5be46b9dac4107ffc16862dc51585a976f51b7049e441720e82a08c0e61
+size 734

test_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.25,
-    "eval_accuracy": 0.8387096774193549,
-    "eval_loss": 0.6151520609855652,
-    "eval_runtime": 18.1208,
-    "eval_samples_per_second": 8.554,
-    "eval_steps_per_second": 8.554
+    "eval_accuracy": 0.8709677419354839,
+    "eval_loss": 0.41407766938209534,
+    "eval_runtime": 44.3003,
+    "eval_samples_per_second": 3.499,
+    "eval_steps_per_second": 3.499
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.8571428571428571,
-  "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-900",
+  "best_metric": 0.8857142857142857,
+  "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-1200",
   "epoch": 3.25,
   "eval_steps": 500,
   "global_step": 1200,
@@ -11,784 +11,784 @@
     {
       "epoch": 0.01,
       "learning_rate": 4.166666666666667e-06,
-      "loss": 2.5348,
+      "loss": 2.329,
       "step": 10
     },
     {
       "epoch": 0.02,
       "learning_rate": 8.333333333333334e-06,
-      "loss": 2.2499,
+      "loss": 2.2466,
       "step": 20
     },
     {
       "epoch": 0.03,
       "learning_rate": 1.25e-05,
-      "loss": 2.5052,
+      "loss": 2.2363,
       "step": 30
     },
     {
       "epoch": 0.03,
       "learning_rate": 1.6666666666666667e-05,
-      "loss": 2.6626,
+      "loss": 2.328,
       "step": 40
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.0833333333333336e-05,
-      "loss": 2.2731,
+      "loss": 2.1135,
       "step": 50
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.5e-05,
-      "loss": 2.3693,
+      "loss": 2.2845,
       "step": 60
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.916666666666667e-05,
-      "loss": 2.2157,
+      "loss": 2.2225,
       "step": 70
     },
     {
       "epoch": 0.07,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.1835,
+      "loss": 2.163,
       "step": 80
     },
     {
       "epoch": 0.07,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 2.023,
+      "loss": 2.1187,
       "step": 90
     },
     {
       "epoch": 0.08,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 2.1581,
+      "loss": 2.2381,
       "step": 100
     },
     {
       "epoch": 0.09,
       "learning_rate": 4.5833333333333334e-05,
-      "loss": 2.1551,
+      "loss": 2.1272,
       "step": 110
     },
     {
       "epoch": 0.1,
       "learning_rate": 5e-05,
-      "loss": 1.8589,
+      "loss": 1.8531,
       "step": 120
     },
     {
       "epoch": 0.11,
       "learning_rate": 4.9537037037037035e-05,
-      "loss": 2.6667,
+      "loss": 2.5397,
       "step": 130
     },
     {
       "epoch": 0.12,
       "learning_rate": 4.9074074074074075e-05,
-      "loss": 2.0886,
+      "loss": 2.1318,
       "step": 140
     },
     {
       "epoch": 0.12,
       "learning_rate": 4.8611111111111115e-05,
-      "loss": 2.2096,
+      "loss": 2.3615,
       "step": 150
     },
     {
       "epoch": 0.13,
       "learning_rate": 4.814814814814815e-05,
-      "loss": 2.0123,
+      "loss": 2.0472,
       "step": 160
     },
     {
       "epoch": 0.14,
       "learning_rate": 4.768518518518519e-05,
-      "loss": 1.8907,
+      "loss": 2.0705,
       "step": 170
     },
     {
       "epoch": 0.15,
       "learning_rate": 4.722222222222222e-05,
-      "loss": 1.7564,
+      "loss": 1.8354,
       "step": 180
     },
     {
       "epoch": 0.16,
       "learning_rate": 4.675925925925926e-05,
-      "loss": 1.7499,
+      "loss": 2.1276,
       "step": 190
     },
     {
       "epoch": 0.17,
       "learning_rate": 4.62962962962963e-05,
-      "loss": 1.9893,
+      "loss": 2.1269,
       "step": 200
     },
     {
       "epoch": 0.17,
       "learning_rate": 4.5833333333333334e-05,
-      "loss": 1.9123,
+      "loss": 1.9699,
       "step": 210
     },
     {
       "epoch": 0.18,
       "learning_rate": 4.5370370370370374e-05,
-      "loss": 2.0462,
+      "loss": 2.2254,
       "step": 220
     },
     {
       "epoch": 0.19,
       "learning_rate": 4.490740740740741e-05,
-      "loss": 1.8366,
+      "loss": 2.1619,
       "step": 230
     },
     {
       "epoch": 0.2,
       "learning_rate": 4.4444444444444447e-05,
-      "loss": 2.006,
+      "loss": 2.0637,
       "step": 240
     },
     {
       "epoch": 0.21,
       "learning_rate": 4.3981481481481486e-05,
-      "loss": 1.5225,
+      "loss": 1.5704,
       "step": 250
     },
     {
       "epoch": 0.22,
       "learning_rate": 4.351851851851852e-05,
-      "loss": 1.4465,
+      "loss": 1.6747,
       "step": 260
     },
     {
       "epoch": 0.23,
       "learning_rate": 4.305555555555556e-05,
-      "loss": 1.9501,
+      "loss": 2.4152,
       "step": 270
     },
     {
       "epoch": 0.23,
       "learning_rate": 4.259259259259259e-05,
-      "loss": 1.5535,
+      "loss": 1.3747,
       "step": 280
     },
     {
       "epoch": 0.24,
       "learning_rate": 4.212962962962963e-05,
-      "loss": 1.5687,
+      "loss": 1.5412,
       "step": 290
     },
     {
       "epoch": 0.25,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 1.4673,
+      "loss": 1.7031,
       "step": 300
     },
     {
       "epoch": 0.25,
-      "eval_accuracy": 0.35714285714285715,
-      "eval_loss": 1.834161639213562,
-      "eval_runtime": 8.1144,
-      "eval_samples_per_second": 8.627,
-      "eval_steps_per_second": 8.627,
+      "eval_accuracy": 0.15714285714285714,
+      "eval_loss": 2.6308274269104004,
+      "eval_runtime": 19.445,
+      "eval_samples_per_second": 3.6,
+      "eval_steps_per_second": 3.6,
       "step": 300
     },
     {
       "epoch": 1.01,
       "learning_rate": 4.1203703703703705e-05,
-      "loss": 1.9569,
+      "loss": 2.6236,
       "step": 310
     },
     {
       "epoch": 1.02,
       "learning_rate": 4.074074074074074e-05,
-      "loss": 1.6084,
+      "loss": 1.4325,
       "step": 320
     },
     {
       "epoch": 1.02,
       "learning_rate": 4.027777777777778e-05,
-      "loss": 0.9802,
+      "loss": 1.1836,
       "step": 330
     },
     {
       "epoch": 1.03,
       "learning_rate": 3.981481481481482e-05,
-      "loss": 1.5647,
+      "loss": 1.4357,
       "step": 340
     },
     {
       "epoch": 1.04,
       "learning_rate": 3.935185185185186e-05,
-      "loss": 1.1866,
+      "loss": 1.4766,
       "step": 350
     },
     {
       "epoch": 1.05,
       "learning_rate": 3.888888888888889e-05,
-      "loss": 0.6345,
+      "loss": 0.9234,
       "step": 360
     },
     {
       "epoch": 1.06,
       "learning_rate": 3.8425925925925924e-05,
-      "loss": 0.9199,
+      "loss": 0.9579,
       "step": 370
     },
     {
       "epoch": 1.07,
       "learning_rate": 3.7962962962962964e-05,
-      "loss": 1.0264,
+      "loss": 1.9765,
       "step": 380
     },
     {
       "epoch": 1.07,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 1.2867,
+      "loss": 1.3833,
       "step": 390
     },
     {
       "epoch": 1.08,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.8381,
+      "loss": 1.1666,
       "step": 400
     },
     {
       "epoch": 1.09,
       "learning_rate": 3.6574074074074076e-05,
-      "loss": 1.6984,
+      "loss": 1.5377,
       "step": 410
     },
     {
       "epoch": 1.1,
       "learning_rate": 3.611111111111111e-05,
-      "loss": 0.994,
+      "loss": 0.9717,
       "step": 420
     },
     {
       "epoch": 1.11,
       "learning_rate": 3.564814814814815e-05,
-      "loss": 1.5705,
+      "loss": 1.9949,
       "step": 430
     },
     {
       "epoch": 1.12,
       "learning_rate": 3.518518518518519e-05,
-      "loss": 0.1452,
+      "loss": 0.8711,
       "step": 440
     },
     {
       "epoch": 1.12,
       "learning_rate": 3.472222222222222e-05,
-      "loss": 0.6541,
+      "loss": 1.5926,
       "step": 450
     },
     {
       "epoch": 1.13,
       "learning_rate": 3.425925925925926e-05,
-      "loss": 1.0954,
+      "loss": 0.774,
       "step": 460
     },
     {
       "epoch": 1.14,
       "learning_rate": 3.3796296296296295e-05,
-      "loss": 1.6305,
+      "loss": 0.9463,
       "step": 470
     },
     {
       "epoch": 1.15,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.5766,
+      "loss": 1.7684,
       "step": 480
     },
     {
       "epoch": 1.16,
       "learning_rate": 3.2870370370370375e-05,
-      "loss": 1.4136,
+      "loss": 1.159,
       "step": 490
     },
     {
       "epoch": 1.17,
       "learning_rate": 3.240740740740741e-05,
-      "loss": 0.3009,
+      "loss": 0.3306,
       "step": 500
     },
     {
       "epoch": 1.18,
       "learning_rate": 3.194444444444444e-05,
-      "loss": 1.3079,
+      "loss": 1.6143,
       "step": 510
     },
     {
       "epoch": 1.18,
       "learning_rate": 3.148148148148148e-05,
-      "loss": 1.6593,
+      "loss": 1.3537,
       "step": 520
     },
     {
       "epoch": 1.19,
       "learning_rate": 3.101851851851852e-05,
-      "loss": 0.8459,
+      "loss": 1.1058,
       "step": 530
     },
     {
       "epoch": 1.2,
       "learning_rate": 3.055555555555556e-05,
-      "loss": 0.609,
+      "loss": 1.2422,
       "step": 540
     },
     {
       "epoch": 1.21,
       "learning_rate": 3.0092592592592593e-05,
-      "loss": 0.8575,
+      "loss": 1.121,
       "step": 550
     },
     {
       "epoch": 1.22,
       "learning_rate": 2.962962962962963e-05,
-      "loss": 0.425,
+      "loss": 0.3986,
       "step": 560
     },
     {
       "epoch": 1.23,
       "learning_rate": 2.916666666666667e-05,
-      "loss": 1.4892,
+      "loss": 0.9468,
       "step": 570
     },
     {
       "epoch": 1.23,
       "learning_rate": 2.8703703703703706e-05,
-      "loss": 0.9041,
+      "loss": 0.665,
       "step": 580
     },
     {
       "epoch": 1.24,
       "learning_rate": 2.824074074074074e-05,
-      "loss": 0.5981,
+      "loss": 1.8967,
       "step": 590
     },
     {
       "epoch": 1.25,
       "learning_rate": 2.777777777777778e-05,
-      "loss": 1.3231,
+      "loss": 1.465,
       "step": 600
     },
     {
       "epoch": 1.25,
-      "eval_accuracy": 0.5571428571428572,
-      "eval_loss": 1.3841789960861206,
-      "eval_runtime": 8.1929,
-      "eval_samples_per_second": 8.544,
-      "eval_steps_per_second": 8.544,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.529031753540039,
+      "eval_runtime": 20.4162,
+      "eval_samples_per_second": 3.429,
+      "eval_steps_per_second": 3.429,
       "step": 600
     },
     {
       "epoch": 2.01,
       "learning_rate": 2.7314814814814816e-05,
-      "loss": 0.6209,
+      "loss": 0.6396,
       "step": 610
     },
     {
       "epoch": 2.02,
       "learning_rate": 2.6851851851851855e-05,
-      "loss": 2.3085,
+      "loss": 2.567,
       "step": 620
     },
     {
       "epoch": 2.02,
       "learning_rate": 2.6388888888888892e-05,
-      "loss": 1.2361,
+      "loss": 1.4652,
       "step": 630
     },
     {
       "epoch": 2.03,
       "learning_rate": 2.5925925925925925e-05,
-      "loss": 0.4469,
+      "loss": 1.3357,
       "step": 640
     },
     {
       "epoch": 2.04,
       "learning_rate": 2.5462962962962965e-05,
-      "loss": 0.7314,
+      "loss": 0.6774,
       "step": 650
     },
     {
       "epoch": 2.05,
       "learning_rate": 2.5e-05,
-      "loss": 0.8052,
+      "loss": 0.8569,
       "step": 660
     },
     {
       "epoch": 2.06,
       "learning_rate": 2.4537037037037038e-05,
-      "loss": 0.8913,
+      "loss": 1.5074,
       "step": 670
     },
     {
       "epoch": 2.07,
       "learning_rate": 2.4074074074074074e-05,
-      "loss": 0.501,
+      "loss": 0.1759,
       "step": 680
     },
     {
       "epoch": 2.08,
       "learning_rate": 2.361111111111111e-05,
-      "loss": 0.2084,
+      "loss": 0.1035,
       "step": 690
     },
     {
       "epoch": 2.08,
       "learning_rate": 2.314814814814815e-05,
-      "loss": 1.2433,
+      "loss": 1.5577,
       "step": 700
     },
     {
       "epoch": 2.09,
       "learning_rate": 2.2685185185185187e-05,
-      "loss": 0.7629,
+      "loss": 1.3958,
       "step": 710
     },
     {
       "epoch": 2.1,
       "learning_rate": 2.2222222222222223e-05,
-      "loss": 0.6061,
+      "loss": 0.6997,
       "step": 720
     },
     {
       "epoch": 2.11,
       "learning_rate": 2.175925925925926e-05,
-      "loss": 0.087,
+      "loss": 0.1684,
       "step": 730
     },
     {
       "epoch": 2.12,
       "learning_rate": 2.1296296296296296e-05,
-      "loss": 1.1573,
+      "loss": 1.0568,
       "step": 740
     },
     {
       "epoch": 2.12,
       "learning_rate": 2.0833333333333336e-05,
-      "loss": 0.3543,
+      "loss": 0.2304,
       "step": 750
     },
     {
       "epoch": 2.13,
       "learning_rate": 2.037037037037037e-05,
-      "loss": 0.1171,
+      "loss": 0.3393,
       "step": 760
     },
     {
       "epoch": 2.14,
       "learning_rate": 1.990740740740741e-05,
-      "loss": 0.043,
+      "loss": 0.1672,
       "step": 770
     },
     {
       "epoch": 2.15,
       "learning_rate": 1.9444444444444445e-05,
-      "loss": 0.3266,
+      "loss": 0.2065,
       "step": 780
     },
     {
       "epoch": 2.16,
       "learning_rate": 1.8981481481481482e-05,
-      "loss": 1.0509,
+      "loss": 0.8764,
       "step": 790
     },
     {
       "epoch": 2.17,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.0125,
+      "loss": 0.6443,
       "step": 800
     },
     {
       "epoch": 2.17,
       "learning_rate": 1.8055555555555555e-05,
-      "loss": 0.534,
+      "loss": 0.9863,
       "step": 810
     },
     {
       "epoch": 2.18,
       "learning_rate": 1.7592592592592595e-05,
-      "loss": 1.2978,
+      "loss": 1.2478,
       "step": 820
     },
     {
       "epoch": 2.19,
       "learning_rate": 1.712962962962963e-05,
-      "loss": 0.205,
+      "loss": 0.5116,
       "step": 830
     },
     {
       "epoch": 2.2,
       "learning_rate": 1.6666666666666667e-05,
-      "loss": 0.2479,
+      "loss": 0.4251,
       "step": 840
     },
     {
       "epoch": 2.21,
       "learning_rate": 1.6203703703703704e-05,
-      "loss": 0.2374,
+      "loss": 0.3013,
       "step": 850
     },
     {
       "epoch": 2.22,
       "learning_rate": 1.574074074074074e-05,
-      "loss": 0.4977,
+      "loss": 0.5455,
       "step": 860
     },
     {
       "epoch": 2.23,
       "learning_rate": 1.527777777777778e-05,
-      "loss": 0.164,
+      "loss": 0.236,
       "step": 870
     },
     {
       "epoch": 2.23,
       "learning_rate": 1.4814814814814815e-05,
-      "loss": 0.5244,
+      "loss": 0.2929,
       "step": 880
     },
     {
       "epoch": 2.24,
       "learning_rate": 1.4351851851851853e-05,
-      "loss": 0.7385,
+      "loss": 0.6095,
       "step": 890
     },
     {
       "epoch": 2.25,
       "learning_rate": 1.388888888888889e-05,
-      "loss": 0.1856,
+      "loss": 0.1843,
       "step": 900
     },
     {
       "epoch": 2.25,
-      "eval_accuracy": 0.8571428571428571,
-      "eval_loss": 0.4796161651611328,
-      "eval_runtime": 8.0524,
-      "eval_samples_per_second": 8.693,
-      "eval_steps_per_second": 8.693,
+      "eval_accuracy": 0.8714285714285714,
+      "eval_loss": 0.41582703590393066,
+      "eval_runtime": 19.8934,
+      "eval_samples_per_second": 3.519,
+      "eval_steps_per_second": 3.519,
       "step": 900
     },
     {
       "epoch": 3.01,
       "learning_rate": 1.3425925925925928e-05,
-      "loss": 0.0568,
+      "loss": 0.2904,
       "step": 910
     },
     {
       "epoch": 3.02,
       "learning_rate": 1.2962962962962962e-05,
-      "loss": 0.264,
+      "loss": 0.6225,
       "step": 920
     },
     {
       "epoch": 3.02,
       "learning_rate": 1.25e-05,
-      "loss": 0.2365,
+      "loss": 0.1852,
       "step": 930
     },
     {
       "epoch": 3.03,
       "learning_rate": 1.2037037037037037e-05,
-      "loss": 0.0268,
+      "loss": 0.0468,
       "step": 940
     },
     {
       "epoch": 3.04,
       "learning_rate": 1.1574074074074075e-05,
-      "loss": 0.2831,
+      "loss": 0.3321,
       "step": 950
     },
     {
       "epoch": 3.05,
       "learning_rate": 1.1111111111111112e-05,
-      "loss": 0.3282,
+      "loss": 0.3365,
       "step": 960
     },
     {
       "epoch": 3.06,
       "learning_rate": 1.0648148148148148e-05,
-      "loss": 0.547,
+      "loss": 0.5674,
       "step": 970
     },
     {
       "epoch": 3.07,
       "learning_rate": 1.0185185185185185e-05,
-      "loss": 0.3342,
+      "loss": 0.6881,
       "step": 980
     },
     {
       "epoch": 3.08,
       "learning_rate": 9.722222222222223e-06,
-      "loss": 0.3915,
+      "loss": 0.317,
       "step": 990
     },
     {
       "epoch": 3.08,
       "learning_rate": 9.259259259259259e-06,
-      "loss": 0.8406,
+      "loss": 0.6281,
       "step": 1000
     },
     {
       "epoch": 3.09,
       "learning_rate": 8.796296296296297e-06,
-      "loss": 0.1584,
+      "loss": 0.4502,
       "step": 1010
     },
     {
       "epoch": 3.1,
       "learning_rate": 8.333333333333334e-06,
-      "loss": 0.0057,
+      "loss": 0.0073,
       "step": 1020
     },
     {
       "epoch": 3.11,
       "learning_rate": 7.87037037037037e-06,
-      "loss": 0.0056,
+      "loss": 0.0168,
       "step": 1030
     },
     {
       "epoch": 3.12,
       "learning_rate": 7.4074074074074075e-06,
-      "loss": 0.0188,
+      "loss": 0.0054,
       "step": 1040
     },
     {
       "epoch": 3.12,
       "learning_rate": 6.944444444444445e-06,
-      "loss": 0.3416,
+      "loss": 0.3653,
       "step": 1050
     },
     {
       "epoch": 3.13,
       "learning_rate": 6.481481481481481e-06,
-      "loss": 0.0079,
+      "loss": 0.0088,
       "step": 1060
     },
     {
       "epoch": 3.14,
       "learning_rate": 6.0185185185185185e-06,
-      "loss": 0.0132,
+      "loss": 0.014,
       "step": 1070
     },
     {
       "epoch": 3.15,
       "learning_rate": 5.555555555555556e-06,
-      "loss": 0.3391,
+      "loss": 0.1096,
       "step": 1080
     },
     {
       "epoch": 3.16,
       "learning_rate": 5.092592592592592e-06,
-      "loss": 0.702,
+      "loss": 0.3679,
       "step": 1090
     },
     {
       "epoch": 3.17,
       "learning_rate": 4.6296296296296296e-06,
-      "loss": 0.2409,
+      "loss": 0.0451,
       "step": 1100
     },
     {
       "epoch": 3.17,
       "learning_rate": 4.166666666666667e-06,
-      "loss": 0.0807,
+      "loss": 0.5222,
       "step": 1110
     },
     {
       "epoch": 3.18,
       "learning_rate": 3.7037037037037037e-06,
-      "loss": 0.0162,
+      "loss": 0.0134,
       "step": 1120
     },
     {
       "epoch": 3.19,
       "learning_rate": 3.2407407407407406e-06,
-      "loss": 0.4397,
+      "loss": 0.4009,
       "step": 1130
     },
     {
       "epoch": 3.2,
       "learning_rate": 2.777777777777778e-06,
-      "loss": 0.0262,
+      "loss": 0.1051,
       "step": 1140
     },
     {
       "epoch": 3.21,
       "learning_rate": 2.3148148148148148e-06,
-      "loss": 0.0113,
+      "loss": 0.1059,
       "step": 1150
     },
     {
       "epoch": 3.22,
       "learning_rate": 1.8518518518518519e-06,
-      "loss": 0.0421,
+      "loss": 0.5751,
       "step": 1160
     },
     {
       "epoch": 3.23,
       "learning_rate": 1.388888888888889e-06,
-      "loss": 0.0142,
+      "loss": 0.0371,
       "step": 1170
     },
     {
       "epoch": 3.23,
       "learning_rate": 9.259259259259259e-07,
-      "loss": 0.0231,
+      "loss": 0.039,
       "step": 1180
     },
     {
       "epoch": 3.24,
       "learning_rate": 4.6296296296296297e-07,
-      "loss": 0.0173,
+      "loss": 0.1572,
       "step": 1190
     },
     {
       "epoch": 3.25,
       "learning_rate": 0.0,
-      "loss": 0.3141,
+      "loss": 0.1232,
       "step": 1200
     },
     {
       "epoch": 3.25,
-      "eval_accuracy": 0.8142857142857143,
-      "eval_loss": 0.49507665634155273,
-      "eval_runtime": 8.1862,
-      "eval_samples_per_second": 8.551,
-      "eval_steps_per_second": 8.551,
+      "eval_accuracy": 0.8857142857142857,
+      "eval_loss": 0.37796080112457275,
+      "eval_runtime": 19.8608,
+      "eval_samples_per_second": 3.525,
+      "eval_steps_per_second": 3.525,
       "step": 1200
     },
     {
       "epoch": 3.25,
       "step": 1200,
       "total_flos": 1.495384188125184e+18,
-      "train_loss": 0.9936102718735734,
-      "train_runtime": 297.6864,
-      "train_samples_per_second": 4.031,
-      "train_steps_per_second": 4.031
+      "train_loss": 1.0834253594931216,
+      "train_runtime": 831.8803,
+      "train_samples_per_second": 1.443,
+      "train_steps_per_second": 1.443
     },
     {
       "epoch": 3.25,
-      "eval_accuracy": 0.8387096774193549,
-      "eval_loss": 0.6151520609855652,
-      "eval_runtime": 18.5662,
-      "eval_samples_per_second": 8.348,
-      "eval_steps_per_second": 8.348,
+      "eval_accuracy": 0.8709677419354839,
+      "eval_loss": 0.41407766938209534,
+      "eval_runtime": 52.821,
+      "eval_samples_per_second": 2.934,
+      "eval_steps_per_second": 2.934,
       "step": 1200
     },
     {
       "epoch": 3.25,
-      "eval_accuracy": 0.8387096774193549,
-      "eval_loss": 0.6151520609855652,
-      "eval_runtime": 18.1208,
-      "eval_samples_per_second": 8.554,
-      "eval_steps_per_second": 8.554,
+      "eval_accuracy": 0.8709677419354839,
+      "eval_loss": 0.41407766938209534,
+      "eval_runtime": 44.3003,
+      "eval_samples_per_second": 3.499,
+      "eval_steps_per_second": 3.499,
       "step": 1200
     }
   ],