End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 40.0,
-    "eval_accuracy": 0.9767441860465116,
-    "eval_loss": 0.14032752811908722,
-    "eval_runtime": 2.9988,
-    "eval_samples_per_second": 14.339,
-    "eval_steps_per_second": 0.667,
-    "total_flos": 3.7685890970025984e+17,
-    "train_loss": 0.15792271693547566,
-    "train_runtime": 1762.7321,
-    "train_samples_per_second": 8.6,
-    "train_steps_per_second": 0.068
+    "epoch": 39.92,
+    "eval_accuracy": 0.9555555555555556,
+    "eval_loss": 0.20596002042293549,
+    "eval_runtime": 3.0656,
+    "eval_samples_per_second": 14.679,
+    "eval_steps_per_second": 0.652,
+    "total_flos": 3.992815176520827e+17,
+    "train_loss": 0.1958512246608734,
+    "train_runtime": 1830.7134,
+    "train_samples_per_second": 8.783,
+    "train_steps_per_second": 0.066
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 40.0,
-    "eval_accuracy": 0.9767441860465116,
-    "eval_loss": 0.14032752811908722,
-    "eval_runtime": 2.9988,
-    "eval_samples_per_second": 14.339,
-    "eval_steps_per_second": 0.667
+    "epoch": 39.92,
+    "eval_accuracy": 0.9555555555555556,
+    "eval_loss": 0.20596002042293549,
+    "eval_runtime": 3.0656,
+    "eval_samples_per_second": 14.679,
+    "eval_steps_per_second": 0.652
 }

runs/Oct06_12-26-35_ab34123e7ac3/events.out.tfevents.1665061361.ab34123e7ac3.68.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eeeff0eb56b194444022352812e4d3963610013dee4aca0fde8064365b66121e
+size 357

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 40.0,
-    "total_flos": 3.7685890970025984e+17,
-    "train_loss": 0.15792271693547566,
-    "train_runtime": 1762.7321,
-    "train_samples_per_second": 8.6,
-    "train_steps_per_second": 0.068
+    "epoch": 39.92,
+    "total_flos": 3.992815176520827e+17,
+    "train_loss": 0.1958512246608734,
+    "train_runtime": 1830.7134,
+    "train_samples_per_second": 8.783,
+    "train_steps_per_second": 0.066
 }

trainer_state.json

@@ -1,421 +1,421 @@
 {
-  "best_metric": 0.9767441860465116,
-  "best_model_checkpoint": "delivery_truck_classification/checkpoint-9",
-  "epoch": 40.0,
+  "best_metric": 0.9555555555555556,
+  "best_model_checkpoint": "delivery_truck_classification/checkpoint-27",
+  "epoch": 39.92307692307692,
   "global_step": 120,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 1.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.1490669548511505,
-      "eval_runtime": 2.9917,
-      "eval_samples_per_second": 14.373,
-      "eval_steps_per_second": 0.669,
+      "epoch": 0.92,
+      "eval_accuracy": 0.9111111111111111,
+      "eval_loss": 0.3269350230693817,
+      "eval_runtime": 3.0493,
+      "eval_samples_per_second": 14.757,
+      "eval_steps_per_second": 0.656,
       "step": 3
     },
     {
-      "epoch": 2.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.14621073007583618,
-      "eval_runtime": 3.3898,
-      "eval_samples_per_second": 12.685,
-      "eval_steps_per_second": 0.59,
+      "epoch": 1.92,
+      "eval_accuracy": 0.9333333333333333,
+      "eval_loss": 0.28137677907943726,
+      "eval_runtime": 3.5544,
+      "eval_samples_per_second": 12.66,
+      "eval_steps_per_second": 0.563,
       "step": 6
     },
     {
-      "epoch": 3.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.14032752811908722,
-      "eval_runtime": 3.2019,
-      "eval_samples_per_second": 13.429,
-      "eval_steps_per_second": 0.625,
+      "epoch": 2.92,
+      "eval_accuracy": 0.9333333333333333,
+      "eval_loss": 0.26246681809425354,
+      "eval_runtime": 3.3113,
+      "eval_samples_per_second": 13.59,
+      "eval_steps_per_second": 0.604,
       "step": 9
     },
     {
-      "epoch": 4.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.14311185479164124,
-      "eval_runtime": 2.9749,
-      "eval_samples_per_second": 14.455,
-      "eval_steps_per_second": 0.672,
+      "epoch": 3.92,
+      "eval_accuracy": 0.9333333333333333,
+      "eval_loss": 0.2771083414554596,
+      "eval_runtime": 2.9987,
+      "eval_samples_per_second": 15.006,
+      "eval_steps_per_second": 0.667,
       "step": 12
     },
     {
-      "epoch": 5.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.17613790929317474,
-      "eval_runtime": 3.2124,
-      "eval_samples_per_second": 13.386,
-      "eval_steps_per_second": 0.623,
+      "epoch": 4.92,
+      "eval_accuracy": 0.9333333333333333,
+      "eval_loss": 0.24194401502609253,
+      "eval_runtime": 3.2862,
+      "eval_samples_per_second": 13.694,
+      "eval_steps_per_second": 0.609,
       "step": 15
     },
     {
-      "epoch": 6.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.19303567707538605,
-      "eval_runtime": 3.2899,
-      "eval_samples_per_second": 13.07,
-      "eval_steps_per_second": 0.608,
+      "epoch": 5.92,
+      "eval_accuracy": 0.9111111111111111,
+      "eval_loss": 0.2263958752155304,
+      "eval_runtime": 3.8859,
+      "eval_samples_per_second": 11.58,
+      "eval_steps_per_second": 0.515,
       "step": 18
     },
     {
-      "epoch": 6.67,
+      "epoch": 6.62,
       "learning_rate": 4.62962962962963e-05,
-      "loss": 0.2637,
+      "loss": 0.3207,
       "step": 20
     },
     {
-      "epoch": 7.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.16774171590805054,
-      "eval_runtime": 2.9265,
-      "eval_samples_per_second": 14.693,
-      "eval_steps_per_second": 0.683,
+      "epoch": 6.92,
+      "eval_accuracy": 0.9333333333333333,
+      "eval_loss": 0.25299549102783203,
+      "eval_runtime": 3.0094,
+      "eval_samples_per_second": 14.953,
+      "eval_steps_per_second": 0.665,
       "step": 21
     },
     {
-      "epoch": 8.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.18350879848003387,
-      "eval_runtime": 3.3322,
-      "eval_samples_per_second": 12.904,
-      "eval_steps_per_second": 0.6,
+      "epoch": 7.92,
+      "eval_accuracy": 0.9333333333333333,
+      "eval_loss": 0.22416652739048004,
+      "eval_runtime": 3.3422,
+      "eval_samples_per_second": 13.464,
+      "eval_steps_per_second": 0.598,
       "step": 24
     },
     {
-      "epoch": 9.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.18035028874874115,
-      "eval_runtime": 3.1848,
-      "eval_samples_per_second": 13.501,
-      "eval_steps_per_second": 0.628,
+      "epoch": 8.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.20596002042293549,
+      "eval_runtime": 3.2888,
+      "eval_samples_per_second": 13.683,
+      "eval_steps_per_second": 0.608,
       "step": 27
     },
     {
-      "epoch": 10.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.18557493388652802,
-      "eval_runtime": 3.0604,
-      "eval_samples_per_second": 14.051,
-      "eval_steps_per_second": 0.654,
+      "epoch": 9.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.18088287115097046,
+      "eval_runtime": 3.0438,
+      "eval_samples_per_second": 14.784,
+      "eval_steps_per_second": 0.657,
       "step": 30
     },
     {
-      "epoch": 11.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.17186354100704193,
-      "eval_runtime": 3.2354,
-      "eval_samples_per_second": 13.291,
-      "eval_steps_per_second": 0.618,
+      "epoch": 10.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.20699164271354675,
+      "eval_runtime": 3.347,
+      "eval_samples_per_second": 13.445,
+      "eval_steps_per_second": 0.598,
       "step": 33
     },
     {
-      "epoch": 12.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.16796864569187164,
-      "eval_runtime": 3.1706,
-      "eval_samples_per_second": 13.562,
-      "eval_steps_per_second": 0.631,
+      "epoch": 11.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.1999157965183258,
+      "eval_runtime": 3.2411,
+      "eval_samples_per_second": 13.884,
+      "eval_steps_per_second": 0.617,
       "step": 36
     },
     {
-      "epoch": 13.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.15706905722618103,
-      "eval_runtime": 2.9572,
-      "eval_samples_per_second": 14.541,
-      "eval_steps_per_second": 0.676,
+      "epoch": 12.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.20134444534778595,
+      "eval_runtime": 3.001,
+      "eval_samples_per_second": 14.995,
+      "eval_steps_per_second": 0.666,
       "step": 39
     },
     {
-      "epoch": 13.33,
+      "epoch": 13.31,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.1687,
+      "loss": 0.2066,
       "step": 40
     },
     {
-      "epoch": 14.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.13334713876247406,
-      "eval_runtime": 3.2093,
-      "eval_samples_per_second": 13.398,
-      "eval_steps_per_second": 0.623,
+      "epoch": 13.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.2026776820421219,
+      "eval_runtime": 3.2971,
+      "eval_samples_per_second": 13.648,
+      "eval_steps_per_second": 0.607,
       "step": 42
     },
     {
-      "epoch": 15.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.12847872078418732,
-      "eval_runtime": 3.2531,
-      "eval_samples_per_second": 13.218,
-      "eval_steps_per_second": 0.615,
+      "epoch": 14.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.1808823049068451,
+      "eval_runtime": 3.3018,
+      "eval_samples_per_second": 13.629,
+      "eval_steps_per_second": 0.606,
       "step": 45
     },
     {
-      "epoch": 16.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.12928344309329987,
-      "eval_runtime": 3.0101,
-      "eval_samples_per_second": 14.285,
-      "eval_steps_per_second": 0.664,
+      "epoch": 15.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.16574886441230774,
+      "eval_runtime": 3.1007,
+      "eval_samples_per_second": 14.513,
+      "eval_steps_per_second": 0.645,
       "step": 48
     },
     {
-      "epoch": 17.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.12077976763248444,
-      "eval_runtime": 3.2048,
-      "eval_samples_per_second": 13.417,
-      "eval_steps_per_second": 0.624,
+      "epoch": 16.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.17283517122268677,
+      "eval_runtime": 3.305,
+      "eval_samples_per_second": 13.616,
+      "eval_steps_per_second": 0.605,
       "step": 51
     },
     {
-      "epoch": 18.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.1060628667473793,
-      "eval_runtime": 3.2421,
-      "eval_samples_per_second": 13.263,
-      "eval_steps_per_second": 0.617,
+      "epoch": 17.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.20132465660572052,
+      "eval_runtime": 3.258,
+      "eval_samples_per_second": 13.812,
+      "eval_steps_per_second": 0.614,
       "step": 54
     },
     {
-      "epoch": 19.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.09775999933481216,
-      "eval_runtime": 2.9563,
-      "eval_samples_per_second": 14.545,
-      "eval_steps_per_second": 0.677,
+      "epoch": 18.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.22262603044509888,
+      "eval_runtime": 3.0138,
+      "eval_samples_per_second": 14.931,
+      "eval_steps_per_second": 0.664,
       "step": 57
     },
     {
-      "epoch": 20.0,
+      "epoch": 19.92,
       "learning_rate": 2.777777777777778e-05,
-      "loss": 0.1435,
+      "loss": 0.1894,
       "step": 60
     },
     {
-      "epoch": 20.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.11002050340175629,
-      "eval_runtime": 3.164,
-      "eval_samples_per_second": 13.59,
-      "eval_steps_per_second": 0.632,
+      "epoch": 19.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.20911100506782532,
+      "eval_runtime": 3.2366,
+      "eval_samples_per_second": 13.903,
+      "eval_steps_per_second": 0.618,
       "step": 60
     },
     {
-      "epoch": 21.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.1205144003033638,
-      "eval_runtime": 3.223,
-      "eval_samples_per_second": 13.342,
-      "eval_steps_per_second": 0.621,
+      "epoch": 20.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.19400011003017426,
+      "eval_runtime": 3.2491,
+      "eval_samples_per_second": 13.85,
+      "eval_steps_per_second": 0.616,
       "step": 63
     },
     {
-      "epoch": 22.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.10268723964691162,
-      "eval_runtime": 2.9505,
-      "eval_samples_per_second": 14.574,
-      "eval_steps_per_second": 0.678,
+      "epoch": 21.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.19758270680904388,
+      "eval_runtime": 3.0406,
+      "eval_samples_per_second": 14.8,
+      "eval_steps_per_second": 0.658,
       "step": 66
     },
     {
-      "epoch": 23.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.10407420247793198,
-      "eval_runtime": 3.2576,
-      "eval_samples_per_second": 13.2,
-      "eval_steps_per_second": 0.614,
+      "epoch": 22.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.22320419549942017,
+      "eval_runtime": 3.2519,
+      "eval_samples_per_second": 13.838,
+      "eval_steps_per_second": 0.615,
       "step": 69
     },
     {
-      "epoch": 24.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.10208708792924881,
-      "eval_runtime": 3.1867,
-      "eval_samples_per_second": 13.494,
-      "eval_steps_per_second": 0.628,
+      "epoch": 23.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.2381269484758377,
+      "eval_runtime": 3.3206,
+      "eval_samples_per_second": 13.552,
+      "eval_steps_per_second": 0.602,
       "step": 72
     },
     {
-      "epoch": 25.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.09740080684423447,
-      "eval_runtime": 2.9153,
-      "eval_samples_per_second": 14.75,
-      "eval_steps_per_second": 0.686,
+      "epoch": 24.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.24054782092571259,
+      "eval_runtime": 2.9599,
+      "eval_samples_per_second": 15.203,
+      "eval_steps_per_second": 0.676,
       "step": 75
     },
     {
-      "epoch": 26.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.10056695342063904,
-      "eval_runtime": 3.2051,
-      "eval_samples_per_second": 13.416,
-      "eval_steps_per_second": 0.624,
+      "epoch": 25.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.224700927734375,
+      "eval_runtime": 3.266,
+      "eval_samples_per_second": 13.778,
+      "eval_steps_per_second": 0.612,
       "step": 78
     },
     {
-      "epoch": 26.67,
+      "epoch": 26.62,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.1361,
+      "loss": 0.1713,
       "step": 80
     },
     {
-      "epoch": 27.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.10110510140657425,
-      "eval_runtime": 3.2233,
-      "eval_samples_per_second": 13.34,
-      "eval_steps_per_second": 0.62,
+      "epoch": 26.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.18945971131324768,
+      "eval_runtime": 3.2466,
+      "eval_samples_per_second": 13.861,
+      "eval_steps_per_second": 0.616,
       "step": 81
     },
     {
-      "epoch": 28.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.09926853328943253,
-      "eval_runtime": 2.9231,
-      "eval_samples_per_second": 14.711,
-      "eval_steps_per_second": 0.684,
+      "epoch": 27.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.18361297249794006,
+      "eval_runtime": 2.9994,
+      "eval_samples_per_second": 15.003,
+      "eval_steps_per_second": 0.667,
       "step": 84
     },
     {
-      "epoch": 29.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.09514892101287842,
-      "eval_runtime": 3.1715,
-      "eval_samples_per_second": 13.558,
-      "eval_steps_per_second": 0.631,
+      "epoch": 28.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.19853755831718445,
+      "eval_runtime": 3.2778,
+      "eval_samples_per_second": 13.729,
+      "eval_steps_per_second": 0.61,
       "step": 87
     },
     {
-      "epoch": 30.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.09706470370292664,
-      "eval_runtime": 3.2972,
-      "eval_samples_per_second": 13.041,
-      "eval_steps_per_second": 0.607,
+      "epoch": 29.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.2126913219690323,
+      "eval_runtime": 3.2397,
+      "eval_samples_per_second": 13.89,
+      "eval_steps_per_second": 0.617,
       "step": 90
     },
     {
-      "epoch": 31.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.10362450033426285,
-      "eval_runtime": 2.9574,
-      "eval_samples_per_second": 14.54,
-      "eval_steps_per_second": 0.676,
+      "epoch": 30.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.20982736349105835,
+      "eval_runtime": 2.9929,
+      "eval_samples_per_second": 15.035,
+      "eval_steps_per_second": 0.668,
       "step": 93
     },
     {
-      "epoch": 32.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.10851310193538666,
-      "eval_runtime": 3.2605,
-      "eval_samples_per_second": 13.188,
+      "epoch": 31.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.20032857358455658,
+      "eval_runtime": 3.2633,
+      "eval_samples_per_second": 13.79,
       "eval_steps_per_second": 0.613,
       "step": 96
     },
     {
-      "epoch": 33.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.10988349467515945,
-      "eval_runtime": 3.2152,
-      "eval_samples_per_second": 13.374,
-      "eval_steps_per_second": 0.622,
+      "epoch": 32.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.18492181599140167,
+      "eval_runtime": 3.3119,
+      "eval_samples_per_second": 13.588,
+      "eval_steps_per_second": 0.604,
       "step": 99
     },
     {
-      "epoch": 33.33,
+      "epoch": 33.31,
       "learning_rate": 9.259259259259259e-06,
-      "loss": 0.1221,
+      "loss": 0.1428,
       "step": 100
     },
     {
-      "epoch": 34.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.11153779178857803,
-      "eval_runtime": 2.9113,
-      "eval_samples_per_second": 14.77,
-      "eval_steps_per_second": 0.687,
+      "epoch": 33.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.1842741221189499,
+      "eval_runtime": 2.9406,
+      "eval_samples_per_second": 15.303,
+      "eval_steps_per_second": 0.68,
       "step": 102
     },
     {
-      "epoch": 35.0,
-      "eval_accuracy": 0.9767441860465116,
-      "eval_loss": 0.11329730600118637,
-      "eval_runtime": 3.208,
-      "eval_samples_per_second": 13.404,
-      "eval_steps_per_second": 0.623,
+      "epoch": 34.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.1899583339691162,
+      "eval_runtime": 3.2029,
+      "eval_samples_per_second": 14.05,
+      "eval_steps_per_second": 0.624,
       "step": 105
     },
     {
-      "epoch": 36.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.11842045933008194,
-      "eval_runtime": 3.2236,
-      "eval_samples_per_second": 13.339,
-      "eval_steps_per_second": 0.62,
+      "epoch": 35.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.19720764458179474,
+      "eval_runtime": 3.2124,
+      "eval_samples_per_second": 14.008,
+      "eval_steps_per_second": 0.623,
       "step": 108
     },
     {
-      "epoch": 37.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.12152226269245148,
-      "eval_runtime": 2.9306,
-      "eval_samples_per_second": 14.673,
-      "eval_steps_per_second": 0.682,
+      "epoch": 36.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.20228949189186096,
+      "eval_runtime": 2.9912,
+      "eval_samples_per_second": 15.044,
+      "eval_steps_per_second": 0.669,
       "step": 111
     },
     {
-      "epoch": 38.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.12237626314163208,
-      "eval_runtime": 3.2687,
-      "eval_samples_per_second": 13.155,
+      "epoch": 37.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.20600810647010803,
+      "eval_runtime": 3.269,
+      "eval_samples_per_second": 13.766,
       "eval_steps_per_second": 0.612,
       "step": 114
     },
     {
-      "epoch": 39.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.12215745449066162,
-      "eval_runtime": 3.2407,
-      "eval_samples_per_second": 13.269,
-      "eval_steps_per_second": 0.617,
+      "epoch": 38.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.20927244424819946,
+      "eval_runtime": 3.214,
+      "eval_samples_per_second": 14.001,
+      "eval_steps_per_second": 0.622,
       "step": 117
     },
     {
-      "epoch": 40.0,
+      "epoch": 39.92,
       "learning_rate": 0.0,
-      "loss": 0.1135,
+      "loss": 0.1443,
       "step": 120
     },
     {
-      "epoch": 40.0,
-      "eval_accuracy": 0.9534883720930233,
-      "eval_loss": 0.1217384859919548,
-      "eval_runtime": 2.9059,
-      "eval_samples_per_second": 14.797,
-      "eval_steps_per_second": 0.688,
+      "epoch": 39.92,
+      "eval_accuracy": 0.9555555555555556,
+      "eval_loss": 0.21063438057899475,
+      "eval_runtime": 3.0071,
+      "eval_samples_per_second": 14.965,
+      "eval_steps_per_second": 0.665,
       "step": 120
     },
     {
-      "epoch": 40.0,
+      "epoch": 39.92,
       "step": 120,
-      "total_flos": 3.7685890970025984e+17,
-      "train_loss": 0.15792271693547566,
-      "train_runtime": 1762.7321,
-      "train_samples_per_second": 8.6,
-      "train_steps_per_second": 0.068
+      "total_flos": 3.992815176520827e+17,
+      "train_loss": 0.1958512246608734,
+      "train_runtime": 1830.7134,
+      "train_samples_per_second": 8.783,
+      "train_steps_per_second": 0.066
     }
   ],
   "max_steps": 120,
   "num_train_epochs": 40,
-  "total_flos": 3.7685890970025984e+17,
+  "total_flos": 3.992815176520827e+17,
   "trial_name": null,
   "trial_params": null
 }