End of training

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 59.84,
-    "eval_accuracy": 0.9846153846153847,
-    "eval_loss": 0.1295621246099472,
-    "eval_runtime": 4.3947,
-    "eval_samples_per_second": 14.791,
-    "eval_steps_per_second": 0.683,
+    "eval_accuracy": 0.9692307692307692,
+    "eval_loss": 0.22925466299057007,
+    "eval_runtime": 4.2919,
+    "eval_samples_per_second": 15.145,
+    "eval_steps_per_second": 0.699,
     "total_flos": 8.648855308501955e+17,
-    "train_loss": 0.577343902985255,
-    "train_runtime": 4213.9216,
-    "train_samples_per_second": 8.273,
-    "train_steps_per_second": 0.057
+    "train_loss": 0.5746380070845286,
+    "train_runtime": 4313.6028,
+    "train_samples_per_second": 8.081,
+    "train_steps_per_second": 0.056
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 59.84,
-    "eval_accuracy": 0.9846153846153847,
-    "eval_loss": 0.1295621246099472,
-    "eval_runtime": 4.3947,
-    "eval_samples_per_second": 14.791,
-    "eval_steps_per_second": 0.683
+    "eval_accuracy": 0.9692307692307692,
+    "eval_loss": 0.22925466299057007,
+    "eval_runtime": 4.2919,
+    "eval_samples_per_second": 15.145,
+    "eval_steps_per_second": 0.699
 }

runs/Jan05_16-34-55_23b52696f101/events.out.tfevents.1672941018.23b52696f101.141.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb4a80be8131bd20dc4f6cf0c578659ccb798beab4a070879c84ccaca3ed2797
+size 363

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 59.84,
     "total_flos": 8.648855308501955e+17,
-    "train_loss": 0.577343902985255,
-    "train_runtime": 4213.9216,
-    "train_samples_per_second": 8.273,
-    "train_steps_per_second": 0.057
+    "train_loss": 0.5746380070845286,
+    "train_runtime": 4313.6028,
+    "train_samples_per_second": 8.081,
+    "train_steps_per_second": 0.056
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.9846153846153847,
-  "best_model_checkpoint": "delivery_truck_classification/checkpoint-104",
+  "best_metric": 0.9692307692307692,
+  "best_model_checkpoint": "delivery_truck_classification/checkpoint-56",
   "epoch": 59.8421052631579,
   "global_step": 240,
   "is_hyper_param_search": false,
@@ -9,624 +9,624 @@
   "log_history": [
     {
       "epoch": 0.84,
-      "eval_accuracy": 0.3230769230769231,
-      "eval_loss": 1.819914698600769,
-      "eval_runtime": 4.4893,
-      "eval_samples_per_second": 14.479,
-      "eval_steps_per_second": 0.668,
+      "eval_accuracy": 0.18461538461538463,
+      "eval_loss": 1.9335094690322876,
+      "eval_runtime": 4.2505,
+      "eval_samples_per_second": 15.292,
+      "eval_steps_per_second": 0.706,
       "step": 4
     },
     {
       "epoch": 1.84,
-      "eval_accuracy": 0.4153846153846154,
-      "eval_loss": 1.7274595499038696,
-      "eval_runtime": 5.1038,
-      "eval_samples_per_second": 12.736,
-      "eval_steps_per_second": 0.588,
+      "eval_accuracy": 0.26153846153846155,
+      "eval_loss": 1.83644700050354,
+      "eval_runtime": 4.7909,
+      "eval_samples_per_second": 13.567,
+      "eval_steps_per_second": 0.626,
       "step": 8
     },
     {
       "epoch": 2.84,
-      "eval_accuracy": 0.46153846153846156,
-      "eval_loss": 1.628129005432129,
-      "eval_runtime": 4.4017,
-      "eval_samples_per_second": 14.767,
-      "eval_steps_per_second": 0.682,
+      "eval_accuracy": 0.38461538461538464,
+      "eval_loss": 1.7054301500320435,
+      "eval_runtime": 4.2339,
+      "eval_samples_per_second": 15.352,
+      "eval_steps_per_second": 0.709,
       "step": 12
     },
     {
       "epoch": 3.84,
-      "eval_accuracy": 0.46153846153846156,
-      "eval_loss": 1.5271515846252441,
-      "eval_runtime": 4.7528,
-      "eval_samples_per_second": 13.676,
-      "eval_steps_per_second": 0.631,
+      "eval_accuracy": 0.4153846153846154,
+      "eval_loss": 1.5629212856292725,
+      "eval_runtime": 4.6,
+      "eval_samples_per_second": 14.13,
+      "eval_steps_per_second": 0.652,
       "step": 16
     },
     {
       "epoch": 4.84,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 1.9537,
+      "loss": 2.0106,
       "step": 20
     },
     {
       "epoch": 4.84,
-      "eval_accuracy": 0.5076923076923077,
-      "eval_loss": 1.3667770624160767,
-      "eval_runtime": 4.3866,
-      "eval_samples_per_second": 14.818,
-      "eval_steps_per_second": 0.684,
+      "eval_accuracy": 0.47692307692307695,
+      "eval_loss": 1.3906540870666504,
+      "eval_runtime": 4.2108,
+      "eval_samples_per_second": 15.437,
+      "eval_steps_per_second": 0.712,
       "step": 20
     },
     {
       "epoch": 5.84,
-      "eval_accuracy": 0.6,
-      "eval_loss": 1.0963521003723145,
-      "eval_runtime": 4.7064,
-      "eval_samples_per_second": 13.811,
-      "eval_steps_per_second": 0.637,
+      "eval_accuracy": 0.5692307692307692,
+      "eval_loss": 1.1983743906021118,
+      "eval_runtime": 4.4922,
+      "eval_samples_per_second": 14.47,
+      "eval_steps_per_second": 0.668,
       "step": 24
     },
     {
       "epoch": 6.84,
-      "eval_accuracy": 0.7846153846153846,
-      "eval_loss": 0.7691364884376526,
-      "eval_runtime": 4.2153,
-      "eval_samples_per_second": 15.42,
-      "eval_steps_per_second": 0.712,
+      "eval_accuracy": 0.6615384615384615,
+      "eval_loss": 0.951930582523346,
+      "eval_runtime": 4.3052,
+      "eval_samples_per_second": 15.098,
+      "eval_steps_per_second": 0.697,
       "step": 28
     },
     {
       "epoch": 7.84,
-      "eval_accuracy": 0.8307692307692308,
-      "eval_loss": 0.6369762420654297,
-      "eval_runtime": 4.7258,
-      "eval_samples_per_second": 13.754,
-      "eval_steps_per_second": 0.635,
+      "eval_accuracy": 0.7846153846153846,
+      "eval_loss": 0.7509785294532776,
+      "eval_runtime": 4.6224,
+      "eval_samples_per_second": 14.062,
+      "eval_steps_per_second": 0.649,
       "step": 32
     },
     {
       "epoch": 8.84,
-      "eval_accuracy": 0.9076923076923077,
-      "eval_loss": 0.4328678548336029,
-      "eval_runtime": 4.7157,
-      "eval_samples_per_second": 13.784,
-      "eval_steps_per_second": 0.636,
+      "eval_accuracy": 0.8615384615384616,
+      "eval_loss": 0.5749186277389526,
+      "eval_runtime": 5.8947,
+      "eval_samples_per_second": 11.027,
+      "eval_steps_per_second": 0.509,
       "step": 36
     },
     {
       "epoch": 9.84,
       "learning_rate": 4.62962962962963e-05,
-      "loss": 1.0682,
+      "loss": 1.1009,
       "step": 40
     },
     {
       "epoch": 9.84,
-      "eval_accuracy": 0.9076923076923077,
-      "eval_loss": 0.3517535626888275,
-      "eval_runtime": 4.7127,
-      "eval_samples_per_second": 13.792,
-      "eval_steps_per_second": 0.637,
+      "eval_accuracy": 0.9384615384615385,
+      "eval_loss": 0.42441025376319885,
+      "eval_runtime": 4.917,
+      "eval_samples_per_second": 13.219,
+      "eval_steps_per_second": 0.61,
       "step": 40
     },
     {
       "epoch": 10.84,
       "eval_accuracy": 0.8923076923076924,
-      "eval_loss": 0.3229323625564575,
-      "eval_runtime": 4.2434,
-      "eval_samples_per_second": 15.318,
-      "eval_steps_per_second": 0.707,
+      "eval_loss": 0.3652417063713074,
+      "eval_runtime": 4.3232,
+      "eval_samples_per_second": 15.035,
+      "eval_steps_per_second": 0.694,
       "step": 44
     },
     {
       "epoch": 11.84,
-      "eval_accuracy": 0.9384615384615385,
-      "eval_loss": 0.2324085831642151,
-      "eval_runtime": 4.734,
-      "eval_samples_per_second": 13.731,
-      "eval_steps_per_second": 0.634,
+      "eval_accuracy": 0.9538461538461539,
+      "eval_loss": 0.2734673023223877,
+      "eval_runtime": 4.5868,
+      "eval_samples_per_second": 14.171,
+      "eval_steps_per_second": 0.654,
       "step": 48
     },
     {
       "epoch": 12.84,
-      "eval_accuracy": 0.9384615384615385,
-      "eval_loss": 0.23686641454696655,
-      "eval_runtime": 4.3354,
-      "eval_samples_per_second": 14.993,
-      "eval_steps_per_second": 0.692,
+      "eval_accuracy": 0.8923076923076924,
+      "eval_loss": 0.29086846113204956,
+      "eval_runtime": 4.5389,
+      "eval_samples_per_second": 14.321,
+      "eval_steps_per_second": 0.661,
       "step": 52
     },
     {
       "epoch": 13.84,
-      "eval_accuracy": 0.9384615384615385,
-      "eval_loss": 0.21190524101257324,
-      "eval_runtime": 4.793,
-      "eval_samples_per_second": 13.561,
-      "eval_steps_per_second": 0.626,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.22925466299057007,
+      "eval_runtime": 4.2293,
+      "eval_samples_per_second": 15.369,
+      "eval_steps_per_second": 0.709,
       "step": 56
     },
     {
       "epoch": 14.84,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 0.6335,
+      "loss": 0.6329,
       "step": 60
     },
     {
       "epoch": 14.84,
-      "eval_accuracy": 0.9384615384615385,
-      "eval_loss": 0.18046662211418152,
-      "eval_runtime": 4.2602,
-      "eval_samples_per_second": 15.257,
-      "eval_steps_per_second": 0.704,
+      "eval_accuracy": 0.9076923076923077,
+      "eval_loss": 0.2562795877456665,
+      "eval_runtime": 4.6895,
+      "eval_samples_per_second": 13.861,
+      "eval_steps_per_second": 0.64,
       "step": 60
     },
     {
       "epoch": 15.84,
-      "eval_accuracy": 0.9076923076923077,
-      "eval_loss": 0.21350829303264618,
-      "eval_runtime": 4.7427,
-      "eval_samples_per_second": 13.705,
-      "eval_steps_per_second": 0.633,
+      "eval_accuracy": 0.9230769230769231,
+      "eval_loss": 0.22179557383060455,
+      "eval_runtime": 4.4085,
+      "eval_samples_per_second": 14.744,
+      "eval_steps_per_second": 0.681,
       "step": 64
     },
     {
       "epoch": 16.84,
-      "eval_accuracy": 0.9230769230769231,
-      "eval_loss": 0.1889328807592392,
-      "eval_runtime": 4.2616,
-      "eval_samples_per_second": 15.253,
-      "eval_steps_per_second": 0.704,
+      "eval_accuracy": 0.9538461538461539,
+      "eval_loss": 0.21023423969745636,
+      "eval_runtime": 4.6409,
+      "eval_samples_per_second": 14.006,
+      "eval_steps_per_second": 0.646,
       "step": 68
     },
     {
       "epoch": 17.84,
-      "eval_accuracy": 0.9538461538461539,
-      "eval_loss": 0.16009008884429932,
-      "eval_runtime": 4.7629,
-      "eval_samples_per_second": 13.647,
-      "eval_steps_per_second": 0.63,
+      "eval_accuracy": 0.9230769230769231,
+      "eval_loss": 0.18287315964698792,
+      "eval_runtime": 4.2374,
+      "eval_samples_per_second": 15.339,
+      "eval_steps_per_second": 0.708,
       "step": 72
     },
     {
       "epoch": 18.84,
-      "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.14124387502670288,
-      "eval_runtime": 4.3251,
-      "eval_samples_per_second": 15.029,
-      "eval_steps_per_second": 0.694,
+      "eval_accuracy": 0.9230769230769231,
+      "eval_loss": 0.1991574615240097,
+      "eval_runtime": 4.6374,
+      "eval_samples_per_second": 14.016,
+      "eval_steps_per_second": 0.647,
       "step": 76
     },
     {
       "epoch": 19.84,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.5133,
+      "loss": 0.497,
       "step": 80
     },
     {
       "epoch": 19.84,
-      "eval_accuracy": 0.9538461538461539,
-      "eval_loss": 0.14968368411064148,
-      "eval_runtime": 4.7085,
-      "eval_samples_per_second": 13.805,
-      "eval_steps_per_second": 0.637,
+      "eval_accuracy": 0.9230769230769231,
+      "eval_loss": 0.18140976130962372,
+      "eval_runtime": 4.3083,
+      "eval_samples_per_second": 15.087,
+      "eval_steps_per_second": 0.696,
       "step": 80
     },
     {
       "epoch": 20.84,
-      "eval_accuracy": 0.9538461538461539,
-      "eval_loss": 0.15449711680412292,
-      "eval_runtime": 4.2803,
-      "eval_samples_per_second": 15.186,
-      "eval_steps_per_second": 0.701,
+      "eval_accuracy": 0.9384615384615385,
+      "eval_loss": 0.18073710799217224,
+      "eval_runtime": 4.692,
+      "eval_samples_per_second": 13.853,
+      "eval_steps_per_second": 0.639,
       "step": 84
     },
     {
       "epoch": 21.84,
       "eval_accuracy": 0.9538461538461539,
-      "eval_loss": 0.12975920736789703,
-      "eval_runtime": 6.2107,
-      "eval_samples_per_second": 10.466,
-      "eval_steps_per_second": 0.483,
+      "eval_loss": 0.17651371657848358,
+      "eval_runtime": 4.3195,
+      "eval_samples_per_second": 15.048,
+      "eval_steps_per_second": 0.695,
       "step": 88
     },
     {
       "epoch": 22.84,
-      "eval_accuracy": 0.9538461538461539,
-      "eval_loss": 0.14153267443180084,
-      "eval_runtime": 4.2538,
-      "eval_samples_per_second": 15.28,
-      "eval_steps_per_second": 0.705,
+      "eval_accuracy": 0.9230769230769231,
+      "eval_loss": 0.18682582676410675,
+      "eval_runtime": 4.7596,
+      "eval_samples_per_second": 13.657,
+      "eval_steps_per_second": 0.63,
       "step": 92
     },
     {
       "epoch": 23.84,
-      "eval_accuracy": 0.9230769230769231,
-      "eval_loss": 0.16845214366912842,
-      "eval_runtime": 4.6805,
-      "eval_samples_per_second": 13.888,
-      "eval_steps_per_second": 0.641,
+      "eval_accuracy": 0.9384615384615385,
+      "eval_loss": 0.20892775058746338,
+      "eval_runtime": 4.337,
+      "eval_samples_per_second": 14.987,
+      "eval_steps_per_second": 0.692,
       "step": 96
     },
     {
       "epoch": 24.84,
       "learning_rate": 3.240740740740741e-05,
-      "loss": 0.4383,
+      "loss": 0.4198,
       "step": 100
     },
     {
       "epoch": 24.84,
       "eval_accuracy": 0.9384615384615385,
-      "eval_loss": 0.13808518648147583,
-      "eval_runtime": 4.2829,
-      "eval_samples_per_second": 15.177,
-      "eval_steps_per_second": 0.7,
+      "eval_loss": 0.18977122008800507,
+      "eval_runtime": 4.6813,
+      "eval_samples_per_second": 13.885,
+      "eval_steps_per_second": 0.641,
       "step": 100
     },
     {
       "epoch": 25.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.1295621246099472,
-      "eval_runtime": 4.717,
-      "eval_samples_per_second": 13.78,
-      "eval_steps_per_second": 0.636,
+      "eval_accuracy": 0.9230769230769231,
+      "eval_loss": 0.20645342767238617,
+      "eval_runtime": 4.2709,
+      "eval_samples_per_second": 15.219,
+      "eval_steps_per_second": 0.702,
       "step": 104
     },
     {
       "epoch": 26.84,
-      "eval_accuracy": 0.9538461538461539,
-      "eval_loss": 0.1106828972697258,
-      "eval_runtime": 4.4246,
-      "eval_samples_per_second": 14.691,
-      "eval_steps_per_second": 0.678,
+      "eval_accuracy": 0.9230769230769231,
+      "eval_loss": 0.18446099758148193,
+      "eval_runtime": 4.7129,
+      "eval_samples_per_second": 13.792,
+      "eval_steps_per_second": 0.637,
       "step": 108
     },
     {
       "epoch": 27.84,
-      "eval_accuracy": 0.9384615384615385,
-      "eval_loss": 0.12369797378778458,
-      "eval_runtime": 4.7653,
-      "eval_samples_per_second": 13.64,
-      "eval_steps_per_second": 0.63,
+      "eval_accuracy": 0.9230769230769231,
+      "eval_loss": 0.17242665588855743,
+      "eval_runtime": 4.2634,
+      "eval_samples_per_second": 15.246,
+      "eval_steps_per_second": 0.704,
       "step": 112
     },
     {
       "epoch": 28.84,
-      "eval_accuracy": 0.9538461538461539,
-      "eval_loss": 0.1366400122642517,
-      "eval_runtime": 4.3203,
-      "eval_samples_per_second": 15.045,
-      "eval_steps_per_second": 0.694,
+      "eval_accuracy": 0.9384615384615385,
+      "eval_loss": 0.1611722856760025,
+      "eval_runtime": 4.5386,
+      "eval_samples_per_second": 14.322,
+      "eval_steps_per_second": 0.661,
       "step": 116
     },
     {
       "epoch": 29.84,
       "learning_rate": 2.777777777777778e-05,
-      "loss": 0.4149,
+      "loss": 0.368,
       "step": 120
     },
     {
       "epoch": 29.84,
-      "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.13494354486465454,
-      "eval_runtime": 4.6686,
-      "eval_samples_per_second": 13.923,
-      "eval_steps_per_second": 0.643,
+      "eval_accuracy": 0.9538461538461539,
+      "eval_loss": 0.15382908284664154,
+      "eval_runtime": 5.2863,
+      "eval_samples_per_second": 12.296,
+      "eval_steps_per_second": 0.567,
       "step": 120
     },
     {
       "epoch": 30.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.10458508133888245,
-      "eval_runtime": 4.2843,
-      "eval_samples_per_second": 15.172,
-      "eval_steps_per_second": 0.7,
+      "eval_accuracy": 0.9538461538461539,
+      "eval_loss": 0.15675745904445648,
+      "eval_runtime": 4.6707,
+      "eval_samples_per_second": 13.916,
+      "eval_steps_per_second": 0.642,
       "step": 124
     },
     {
       "epoch": 31.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.0881725624203682,
-      "eval_runtime": 4.7053,
-      "eval_samples_per_second": 13.814,
-      "eval_steps_per_second": 0.638,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14750127494335175,
+      "eval_runtime": 4.2141,
+      "eval_samples_per_second": 15.425,
+      "eval_steps_per_second": 0.712,
       "step": 128
     },
     {
       "epoch": 32.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.10219049453735352,
-      "eval_runtime": 4.3211,
-      "eval_samples_per_second": 15.042,
-      "eval_steps_per_second": 0.694,
+      "eval_accuracy": 0.9538461538461539,
+      "eval_loss": 0.14530035853385925,
+      "eval_runtime": 4.5446,
+      "eval_samples_per_second": 14.303,
+      "eval_steps_per_second": 0.66,
       "step": 132
     },
     {
       "epoch": 33.84,
       "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.12066025286912918,
-      "eval_runtime": 4.8023,
-      "eval_samples_per_second": 13.535,
-      "eval_steps_per_second": 0.625,
+      "eval_loss": 0.1576438844203949,
+      "eval_runtime": 4.2804,
+      "eval_samples_per_second": 15.186,
+      "eval_steps_per_second": 0.701,
       "step": 136
     },
     {
       "epoch": 34.84,
       "learning_rate": 2.314814814814815e-05,
-      "loss": 0.3657,
+      "loss": 0.3709,
       "step": 140
     },
     {
       "epoch": 34.84,
-      "eval_accuracy": 0.9538461538461539,
-      "eval_loss": 0.11678420007228851,
-      "eval_runtime": 4.4027,
-      "eval_samples_per_second": 14.764,
-      "eval_steps_per_second": 0.681,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14298613369464874,
+      "eval_runtime": 4.6347,
+      "eval_samples_per_second": 14.025,
+      "eval_steps_per_second": 0.647,
       "step": 140
     },
     {
       "epoch": 35.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.09219826757907867,
-      "eval_runtime": 6.9593,
-      "eval_samples_per_second": 9.34,
-      "eval_steps_per_second": 0.431,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.13841809332370758,
+      "eval_runtime": 4.3512,
+      "eval_samples_per_second": 14.938,
+      "eval_steps_per_second": 0.689,
       "step": 144
     },
     {
       "epoch": 36.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.09314019978046417,
-      "eval_runtime": 4.3,
-      "eval_samples_per_second": 15.116,
-      "eval_steps_per_second": 0.698,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14322978258132935,
+      "eval_runtime": 4.6696,
+      "eval_samples_per_second": 13.92,
+      "eval_steps_per_second": 0.642,
       "step": 148
     },
     {
       "epoch": 37.84,
       "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.10055378824472427,
-      "eval_runtime": 4.7227,
-      "eval_samples_per_second": 13.763,
-      "eval_steps_per_second": 0.635,
+      "eval_loss": 0.13474790751934052,
+      "eval_runtime": 4.2951,
+      "eval_samples_per_second": 15.133,
+      "eval_steps_per_second": 0.698,
       "step": 152
     },
     {
       "epoch": 38.84,
-      "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.09866410493850708,
-      "eval_runtime": 4.3725,
-      "eval_samples_per_second": 14.866,
-      "eval_steps_per_second": 0.686,
+      "eval_accuracy": 0.9538461538461539,
+      "eval_loss": 0.1358633041381836,
+      "eval_runtime": 4.6622,
+      "eval_samples_per_second": 13.942,
+      "eval_steps_per_second": 0.643,
       "step": 156
     },
     {
       "epoch": 39.84,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.3294,
+      "loss": 0.3373,
       "step": 160
     },
     {
       "epoch": 39.84,
-      "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.11275947839021683,
-      "eval_runtime": 4.6903,
-      "eval_samples_per_second": 13.858,
-      "eval_steps_per_second": 0.64,
+      "eval_accuracy": 0.9538461538461539,
+      "eval_loss": 0.15974925458431244,
+      "eval_runtime": 4.2867,
+      "eval_samples_per_second": 15.163,
+      "eval_steps_per_second": 0.7,
       "step": 160
     },
     {
       "epoch": 40.84,
-      "eval_accuracy": 0.9538461538461539,
-      "eval_loss": 0.11520560085773468,
-      "eval_runtime": 4.3303,
-      "eval_samples_per_second": 15.01,
-      "eval_steps_per_second": 0.693,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.15224121510982513,
+      "eval_runtime": 4.6977,
+      "eval_samples_per_second": 13.837,
+      "eval_steps_per_second": 0.639,
       "step": 164
     },
     {
       "epoch": 41.84,
       "eval_accuracy": 0.9538461538461539,
-      "eval_loss": 0.09974207729101181,
-      "eval_runtime": 4.7259,
-      "eval_samples_per_second": 13.754,
-      "eval_steps_per_second": 0.635,
+      "eval_loss": 0.14772085845470428,
+      "eval_runtime": 4.316,
+      "eval_samples_per_second": 15.06,
+      "eval_steps_per_second": 0.695,
       "step": 168
     },
     {
       "epoch": 42.84,
       "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.09684933722019196,
-      "eval_runtime": 4.3525,
-      "eval_samples_per_second": 14.934,
-      "eval_steps_per_second": 0.689,
+      "eval_loss": 0.1480209231376648,
+      "eval_runtime": 4.7164,
+      "eval_samples_per_second": 13.782,
+      "eval_steps_per_second": 0.636,
       "step": 172
     },
     {
       "epoch": 43.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.08187751471996307,
-      "eval_runtime": 4.6538,
-      "eval_samples_per_second": 13.967,
-      "eval_steps_per_second": 0.645,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14724673330783844,
+      "eval_runtime": 4.3123,
+      "eval_samples_per_second": 15.073,
+      "eval_steps_per_second": 0.696,
       "step": 176
     },
     {
       "epoch": 44.84,
       "learning_rate": 1.388888888888889e-05,
-      "loss": 0.3198,
+      "loss": 0.3342,
       "step": 180
     },
     {
       "epoch": 44.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.07292691618204117,
-      "eval_runtime": 4.3483,
-      "eval_samples_per_second": 14.948,
-      "eval_steps_per_second": 0.69,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14734135568141937,
+      "eval_runtime": 4.5929,
+      "eval_samples_per_second": 14.152,
+      "eval_steps_per_second": 0.653,
       "step": 180
     },
     {
       "epoch": 45.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.07438476383686066,
-      "eval_runtime": 4.8323,
-      "eval_samples_per_second": 13.451,
-      "eval_steps_per_second": 0.621,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14583279192447662,
+      "eval_runtime": 4.3738,
+      "eval_samples_per_second": 14.861,
+      "eval_steps_per_second": 0.686,
       "step": 184
     },
     {
       "epoch": 46.84,
       "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.09510225057601929,
-      "eval_runtime": 4.3498,
-      "eval_samples_per_second": 14.943,
-      "eval_steps_per_second": 0.69,
+      "eval_loss": 0.15290141105651855,
+      "eval_runtime": 4.6815,
+      "eval_samples_per_second": 13.884,
+      "eval_steps_per_second": 0.641,
       "step": 188
     },
     {
       "epoch": 47.84,
       "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.09657642990350723,
-      "eval_runtime": 4.7293,
-      "eval_samples_per_second": 13.744,
-      "eval_steps_per_second": 0.634,
+      "eval_loss": 0.15495683252811432,
+      "eval_runtime": 4.3181,
+      "eval_samples_per_second": 15.053,
+      "eval_steps_per_second": 0.695,
       "step": 192
     },
     {
       "epoch": 48.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.08329986035823822,
-      "eval_runtime": 4.2989,
-      "eval_samples_per_second": 15.12,
-      "eval_steps_per_second": 0.698,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14939415454864502,
+      "eval_runtime": 4.7194,
+      "eval_samples_per_second": 13.773,
+      "eval_steps_per_second": 0.636,
       "step": 196
     },
     {
       "epoch": 49.84,
       "learning_rate": 9.259259259259259e-06,
-      "loss": 0.2936,
+      "loss": 0.2914,
       "step": 200
     },
     {
       "epoch": 49.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.06941184401512146,
-      "eval_runtime": 4.7066,
-      "eval_samples_per_second": 13.811,
-      "eval_steps_per_second": 0.637,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.1469651609659195,
+      "eval_runtime": 4.259,
+      "eval_samples_per_second": 15.262,
+      "eval_steps_per_second": 0.704,
       "step": 200
     },
     {
       "epoch": 50.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.06908843666315079,
-      "eval_runtime": 4.3338,
-      "eval_samples_per_second": 14.999,
-      "eval_steps_per_second": 0.692,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.146009162068367,
+      "eval_runtime": 4.6878,
+      "eval_samples_per_second": 13.866,
+      "eval_steps_per_second": 0.64,
       "step": 204
     },
     {
       "epoch": 51.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.0735919326543808,
-      "eval_runtime": 4.7155,
-      "eval_samples_per_second": 13.784,
-      "eval_steps_per_second": 0.636,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14776931703090668,
+      "eval_runtime": 4.265,
+      "eval_samples_per_second": 15.24,
+      "eval_steps_per_second": 0.703,
       "step": 208
     },
     {
       "epoch": 52.84,
       "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.08048100024461746,
-      "eval_runtime": 4.337,
-      "eval_samples_per_second": 14.987,
-      "eval_steps_per_second": 0.692,
+      "eval_loss": 0.1481345295906067,
+      "eval_runtime": 4.5299,
+      "eval_samples_per_second": 14.349,
+      "eval_steps_per_second": 0.662,
       "step": 212
     },
     {
       "epoch": 53.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.08007017523050308,
-      "eval_runtime": 4.765,
-      "eval_samples_per_second": 13.641,
-      "eval_steps_per_second": 0.63,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14605876803398132,
+      "eval_runtime": 4.1658,
+      "eval_samples_per_second": 15.603,
+      "eval_steps_per_second": 0.72,
       "step": 216
     },
     {
       "epoch": 54.84,
       "learning_rate": 4.6296296296296296e-06,
-      "loss": 0.3127,
+      "loss": 0.2736,
       "step": 220
     },
     {
       "epoch": 54.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.0826273187994957,
-      "eval_runtime": 4.2934,
-      "eval_samples_per_second": 15.139,
-      "eval_steps_per_second": 0.699,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14581629633903503,
+      "eval_runtime": 4.4574,
+      "eval_samples_per_second": 14.583,
+      "eval_steps_per_second": 0.673,
       "step": 220
     },
     {
       "epoch": 55.84,
       "eval_accuracy": 0.9692307692307692,
-      "eval_loss": 0.08565446734428406,
-      "eval_runtime": 4.7099,
-      "eval_samples_per_second": 13.801,
-      "eval_steps_per_second": 0.637,
+      "eval_loss": 0.14382527768611908,
+      "eval_runtime": 4.1183,
+      "eval_samples_per_second": 15.783,
+      "eval_steps_per_second": 0.728,
       "step": 224
     },
     {
       "epoch": 56.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.0864274650812149,
-      "eval_runtime": 4.246,
-      "eval_samples_per_second": 15.309,
-      "eval_steps_per_second": 0.707,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14272591471672058,
+      "eval_runtime": 4.5294,
+      "eval_samples_per_second": 14.351,
+      "eval_steps_per_second": 0.662,
       "step": 228
     },
     {
       "epoch": 57.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.08776658028364182,
-      "eval_runtime": 4.758,
-      "eval_samples_per_second": 13.661,
-      "eval_steps_per_second": 0.631,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.14175653457641602,
+      "eval_runtime": 4.1288,
+      "eval_samples_per_second": 15.743,
+      "eval_steps_per_second": 0.727,
       "step": 232
     },
     {
       "epoch": 58.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.08772400766611099,
-      "eval_runtime": 4.2874,
-      "eval_samples_per_second": 15.161,
-      "eval_steps_per_second": 0.7,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.140133798122406,
+      "eval_runtime": 4.5233,
+      "eval_samples_per_second": 14.37,
+      "eval_steps_per_second": 0.663,
       "step": 236
     },
     {
       "epoch": 59.84,
       "learning_rate": 0.0,
-      "loss": 0.285,
+      "loss": 0.2589,
       "step": 240
     },
     {
       "epoch": 59.84,
-      "eval_accuracy": 0.9846153846153847,
-      "eval_loss": 0.08735792338848114,
-      "eval_runtime": 4.7501,
-      "eval_samples_per_second": 13.684,
-      "eval_steps_per_second": 0.632,
+      "eval_accuracy": 0.9692307692307692,
+      "eval_loss": 0.13985498249530792,
+      "eval_runtime": 4.4933,
+      "eval_samples_per_second": 14.466,
+      "eval_steps_per_second": 0.668,
       "step": 240
     },
     {
       "epoch": 59.84,
       "step": 240,
       "total_flos": 8.648855308501955e+17,
-      "train_loss": 0.577343902985255,
-      "train_runtime": 4213.9216,
-      "train_samples_per_second": 8.273,
-      "train_steps_per_second": 0.057
+      "train_loss": 0.5746380070845286,
+      "train_runtime": 4313.6028,
+      "train_samples_per_second": 8.081,
+      "train_steps_per_second": 0.056
     }
   ],
   "max_steps": 240,