End of training

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.23,
-    "eval_accuracy": 0.7870967741935484,
-    "eval_loss": 0.4868558943271637,
-    "eval_runtime": 149.4696,
-    "eval_samples_per_second": 1.037,
-    "eval_steps_per_second": 0.134
+    "eval_accuracy": 0.8516129032258064,
+    "eval_loss": 0.4354906678199768,
+    "eval_runtime": 104.5631,
+    "eval_samples_per_second": 1.482,
+    "eval_steps_per_second": 0.191
 }

runs/Jul03_15-16-36_e0f8396b976e/events.out.tfevents.1688400695.e0f8396b976e.796.1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eb7e9382ba8ddf43421e952986d1a49a76ab7015bd1e2b2b3ece8f20c94a2029
-size 411
+oid sha256:47b1f245da788e2cdbbbdcb211b26336f171c77bbd4c4571879c30a2175d83e2
+size 734

test_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.23,
-    "eval_accuracy": 0.7870967741935484,
-    "eval_loss": 0.4868558943271637,
-    "eval_runtime": 149.4696,
-    "eval_samples_per_second": 1.037,
-    "eval_steps_per_second": 0.134
+    "eval_accuracy": 0.8516129032258064,
+    "eval_loss": 0.4354906678199768,
+    "eval_runtime": 104.5631,
+    "eval_samples_per_second": 1.482,
+    "eval_steps_per_second": 0.191
 }

trainer_state.json

@@ -10,148 +10,148 @@
     {
       "epoch": 0.07,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.2791,
+      "loss": 2.2591,
       "step": 10
     },
     {
       "epoch": 0.14,
       "learning_rate": 4.81203007518797e-05,
-      "loss": 2.1704,
+      "loss": 2.2132,
       "step": 20
     },
     {
       "epoch": 0.2,
       "learning_rate": 4.43609022556391e-05,
-      "loss": 2.1335,
+      "loss": 2.1351,
       "step": 30
     },
     {
       "epoch": 0.26,
-      "eval_accuracy": 0.5142857142857142,
-      "eval_loss": 1.8132479190826416,
-      "eval_runtime": 45.5282,
-      "eval_samples_per_second": 1.538,
-      "eval_steps_per_second": 0.198,
+      "eval_accuracy": 0.6285714285714286,
+      "eval_loss": 1.6582270860671997,
+      "eval_runtime": 48.1772,
+      "eval_samples_per_second": 1.453,
+      "eval_steps_per_second": 0.187,
       "step": 38
     },
     {
       "epoch": 1.01,
       "learning_rate": 4.0601503759398494e-05,
-      "loss": 2.0233,
+      "loss": 1.9093,
       "step": 40
     },
     {
       "epoch": 1.08,
       "learning_rate": 3.6842105263157895e-05,
-      "loss": 1.5915,
+      "loss": 1.5111,
       "step": 50
     },
     {
       "epoch": 1.15,
       "learning_rate": 3.3082706766917295e-05,
-      "loss": 1.2445,
+      "loss": 1.1256,
       "step": 60
     },
     {
       "epoch": 1.22,
       "learning_rate": 2.9323308270676693e-05,
-      "loss": 0.9301,
+      "loss": 0.7409,
       "step": 70
     },
     {
       "epoch": 1.26,
-      "eval_accuracy": 0.7714285714285715,
-      "eval_loss": 0.8595510125160217,
-      "eval_runtime": 66.0517,
-      "eval_samples_per_second": 1.06,
-      "eval_steps_per_second": 0.136,
+      "eval_accuracy": 0.7142857142857143,
+      "eval_loss": 0.8406534194946289,
+      "eval_runtime": 39.133,
+      "eval_samples_per_second": 1.789,
+      "eval_steps_per_second": 0.23,
       "step": 76
     },
     {
       "epoch": 2.03,
       "learning_rate": 2.556390977443609e-05,
-      "loss": 0.8006,
+      "loss": 0.8077,
       "step": 80
     },
     {
       "epoch": 2.09,
       "learning_rate": 2.1804511278195487e-05,
-      "loss": 0.664,
+      "loss": 0.5975,
       "step": 90
     },
     {
       "epoch": 2.16,
       "learning_rate": 1.8045112781954888e-05,
-      "loss": 0.5401,
+      "loss": 0.6191,
       "step": 100
     },
     {
       "epoch": 2.23,
       "learning_rate": 1.4285714285714285e-05,
-      "loss": 0.4422,
+      "loss": 0.4333,
       "step": 110
     },
     {
       "epoch": 2.26,
-      "eval_accuracy": 0.8714285714285714,
-      "eval_loss": 0.500382125377655,
-      "eval_runtime": 69.7054,
-      "eval_samples_per_second": 1.004,
-      "eval_steps_per_second": 0.129,
+      "eval_accuracy": 0.8142857142857143,
+      "eval_loss": 0.5106571316719055,
+      "eval_runtime": 41.2785,
+      "eval_samples_per_second": 1.696,
+      "eval_steps_per_second": 0.218,
       "step": 114
     },
     {
       "epoch": 3.04,
       "learning_rate": 1.0526315789473684e-05,
-      "loss": 0.3282,
+      "loss": 0.3766,
       "step": 120
     },
     {
       "epoch": 3.11,
       "learning_rate": 6.766917293233083e-06,
-      "loss": 0.3011,
+      "loss": 0.265,
       "step": 130
     },
     {
       "epoch": 3.18,
       "learning_rate": 3.007518796992481e-06,
-      "loss": 0.2828,
+      "loss": 0.2766,
       "step": 140
     },
     {
       "epoch": 3.23,
       "eval_accuracy": 0.9142857142857143,
-      "eval_loss": 0.37309029698371887,
-      "eval_runtime": 70.3365,
-      "eval_samples_per_second": 0.995,
-      "eval_steps_per_second": 0.128,
+      "eval_loss": 0.3579496741294861,
+      "eval_runtime": 49.2971,
+      "eval_samples_per_second": 1.42,
+      "eval_steps_per_second": 0.183,
       "step": 148
     },
     {
       "epoch": 3.23,
       "step": 148,
       "total_flos": 1.460491890402263e+18,
-      "train_loss": 1.0764089355597626,
-      "train_runtime": 3191.3076,
-      "train_samples_per_second": 0.371,
-      "train_steps_per_second": 0.046
+      "train_loss": 1.0465768640105788,
+      "train_runtime": 3053.3069,
+      "train_samples_per_second": 0.388,
+      "train_steps_per_second": 0.048
     },
     {
       "epoch": 3.23,
-      "eval_accuracy": 0.7870967741935484,
-      "eval_loss": 0.4868559241294861,
-      "eval_runtime": 150.5785,
-      "eval_samples_per_second": 1.029,
-      "eval_steps_per_second": 0.133,
+      "eval_accuracy": 0.8516129032258064,
+      "eval_loss": 0.43549060821533203,
+      "eval_runtime": 105.4615,
+      "eval_samples_per_second": 1.47,
+      "eval_steps_per_second": 0.19,
       "step": 148
     },
     {
       "epoch": 3.23,
-      "eval_accuracy": 0.7870967741935484,
-      "eval_loss": 0.4868558943271637,
-      "eval_runtime": 149.4696,
-      "eval_samples_per_second": 1.037,
-      "eval_steps_per_second": 0.134,
+      "eval_accuracy": 0.8516129032258064,
+      "eval_loss": 0.4354906678199768,
+      "eval_runtime": 104.5631,
+      "eval_samples_per_second": 1.482,
+      "eval_steps_per_second": 0.191,
       "step": 148
     }
   ],