End of training

all_results.json

@@ -1,14 +1,14 @@
 {
-    "epoch": 8.0,
-    "eval_accuracy": 0.7581227436823105,
-    "eval_loss": 0.6365415453910828,
-    "eval_runtime": 0.5623,
+    "epoch": 10.0,
+    "eval_accuracy": 0.7978339350180506,
+    "eval_loss": 0.5446364283561707,
+    "eval_runtime": 0.6251,
     "eval_samples": 277,
-    "eval_samples_per_second": 492.657,
-    "eval_steps_per_second": 62.249,
-    "train_loss": 0.42912168074876833,
-    "train_runtime": 173.8183,
+    "eval_samples_per_second": 443.162,
+    "eval_steps_per_second": 55.995,
+    "train_loss": 0.31504388589125415,
+    "train_runtime": 295.0697,
     "train_samples": 2490,
-    "train_samples_per_second": 143.253,
-    "train_steps_per_second": 8.975
+    "train_samples_per_second": 84.387,
+    "train_steps_per_second": 5.287
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 8.0,
-    "eval_accuracy": 0.7581227436823105,
-    "eval_loss": 0.6365415453910828,
-    "eval_runtime": 0.5623,
+    "epoch": 10.0,
+    "eval_accuracy": 0.7978339350180506,
+    "eval_loss": 0.5446364283561707,
+    "eval_runtime": 0.6251,
     "eval_samples": 277,
-    "eval_samples_per_second": 492.657,
-    "eval_steps_per_second": 62.249
+    "eval_samples_per_second": 443.162,
+    "eval_steps_per_second": 55.995
 }

runs/Jun20_21-46-24_pikachu/events.out.tfevents.1655733724.pikachu.2660395.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dce7652753a7417a338897d10bc5af62e211396cdbc60c3fabdcfc6d9e616208
+size 363

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 8.0,
-    "train_loss": 0.42912168074876833,
-    "train_runtime": 173.8183,
+    "epoch": 10.0,
+    "train_loss": 0.31504388589125415,
+    "train_runtime": 295.0697,
     "train_samples": 2490,
-    "train_samples_per_second": 143.253,
-    "train_steps_per_second": 8.975
+    "train_samples_per_second": 84.387,
+    "train_steps_per_second": 5.287
 }

trainer_state.json

@@ -1,8 +1,8 @@
 {
-  "best_metric": 0.7581227436823105,
+  "best_metric": 0.7978339350180506,
   "best_model_checkpoint": "./fine-tune/roberta-base/rte/checkpoint-780",
-  "epoch": 8.0,
-  "global_step": 1248,
+  "epoch": 10.0,
+  "global_step": 1560,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -10,100 +10,124 @@
     {
       "epoch": 1.0,
       "eval_accuracy": 0.4729241877256318,
-      "eval_loss": 0.7072076797485352,
-      "eval_runtime": 0.4902,
-      "eval_samples_per_second": 565.102,
-      "eval_steps_per_second": 71.403,
+      "eval_loss": 0.7023295164108276,
+      "eval_runtime": 0.5782,
+      "eval_samples_per_second": 479.05,
+      "eval_steps_per_second": 60.53,
       "step": 156
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.5270758122743683,
-      "eval_loss": 0.6958328485488892,
-      "eval_runtime": 0.689,
-      "eval_samples_per_second": 402.039,
-      "eval_steps_per_second": 50.799,
+      "eval_accuracy": 0.6895306859205776,
+      "eval_loss": 0.6356249451637268,
+      "eval_runtime": 0.6365,
+      "eval_samples_per_second": 435.208,
+      "eval_steps_per_second": 54.99,
       "step": 312
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.6462093862815884,
-      "eval_loss": 0.6193148493766785,
-      "eval_runtime": 0.5121,
-      "eval_samples_per_second": 540.927,
-      "eval_steps_per_second": 68.348,
+      "eval_accuracy": 0.7617328519855595,
+      "eval_loss": 0.5177018046379089,
+      "eval_runtime": 0.6111,
+      "eval_samples_per_second": 453.318,
+      "eval_steps_per_second": 57.278,
       "step": 468
     },
     {
       "epoch": 3.21,
-      "learning_rate": 1.446111869031378e-05,
-      "loss": 0.6759,
+      "learning_rate": 7.23055934515689e-06,
+      "loss": 0.6131,
       "step": 500
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.7075812274368231,
-      "eval_loss": 0.6046352386474609,
-      "eval_runtime": 0.5274,
-      "eval_samples_per_second": 525.205,
-      "eval_steps_per_second": 66.362,
+      "eval_accuracy": 0.7472924187725631,
+      "eval_loss": 0.6237531304359436,
+      "eval_runtime": 0.6069,
+      "eval_samples_per_second": 456.426,
+      "eval_steps_per_second": 57.671,
       "step": 624
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.7581227436823105,
-      "eval_loss": 0.6365415453910828,
-      "eval_runtime": 0.681,
-      "eval_samples_per_second": 406.732,
-      "eval_steps_per_second": 51.392,
+      "eval_accuracy": 0.7978339350180506,
+      "eval_loss": 0.5446364283561707,
+      "eval_runtime": 0.6019,
+      "eval_samples_per_second": 460.195,
+      "eval_steps_per_second": 58.147,
       "step": 780
     },
     {
       "epoch": 6.0,
       "eval_accuracy": 0.7545126353790613,
-      "eval_loss": 0.897487223148346,
-      "eval_runtime": 0.6584,
-      "eval_samples_per_second": 420.704,
-      "eval_steps_per_second": 53.157,
+      "eval_loss": 0.9697452783584595,
+      "eval_runtime": 0.6225,
+      "eval_samples_per_second": 444.97,
+      "eval_steps_per_second": 56.224,
       "step": 936
     },
     {
       "epoch": 6.41,
-      "learning_rate": 7.639836289222374e-06,
-      "loss": 0.3194,
+      "learning_rate": 3.819918144611187e-06,
+      "loss": 0.2528,
       "step": 1000
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.7581227436823105,
-      "eval_loss": 1.2031357288360596,
-      "eval_runtime": 0.5817,
-      "eval_samples_per_second": 476.173,
-      "eval_steps_per_second": 60.166,
+      "eval_accuracy": 0.7689530685920578,
+      "eval_loss": 1.100365161895752,
+      "eval_runtime": 0.6174,
+      "eval_samples_per_second": 448.691,
+      "eval_steps_per_second": 56.694,
       "step": 1092
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.7581227436823105,
-      "eval_loss": 1.2942094802856445,
-      "eval_runtime": 0.6818,
-      "eval_samples_per_second": 406.254,
-      "eval_steps_per_second": 51.332,
+      "eval_accuracy": 0.7725631768953068,
+      "eval_loss": 1.193679928779602,
+      "eval_runtime": 0.6461,
+      "eval_samples_per_second": 428.72,
+      "eval_steps_per_second": 54.17,
       "step": 1248
     },
     {
-      "epoch": 8.0,
-      "step": 1248,
-      "total_flos": 1310293055692800.0,
-      "train_loss": 0.42912168074876833,
-      "train_runtime": 173.8183,
-      "train_samples_per_second": 143.253,
-      "train_steps_per_second": 8.975
+      "epoch": 9.0,
+      "eval_accuracy": 0.7725631768953068,
+      "eval_loss": 1.331286072731018,
+      "eval_runtime": 0.5642,
+      "eval_samples_per_second": 490.964,
+      "eval_steps_per_second": 62.035,
+      "step": 1404
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 4.092769440654844e-07,
+      "loss": 0.1073,
+      "step": 1500
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.7725631768953068,
+      "eval_loss": 1.353358268737793,
+      "eval_runtime": 0.6152,
+      "eval_samples_per_second": 450.281,
+      "eval_steps_per_second": 56.895,
+      "step": 1560
+    },
+    {
+      "epoch": 10.0,
+      "step": 1560,
+      "total_flos": 1637866319616000.0,
+      "train_loss": 0.31504388589125415,
+      "train_runtime": 295.0697,
+      "train_samples_per_second": 84.387,
+      "train_steps_per_second": 5.287
     }
   ],
   "max_steps": 1560,
   "num_train_epochs": 10,
-  "total_flos": 1310293055692800.0,
+  "total_flos": 1637866319616000.0,
   "trial_name": null,
   "trial_params": null
 }