End of training

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.94,
     "total_flos": 1.1265748270391624e+18,
-    "train_loss": 0.6647836283633584,
-    "train_runtime": 6473.0815,
-    "train_samples_per_second": 2.286,
-    "train_steps_per_second": 0.018
+    "train_loss": 0.2613685036960401,
+    "train_runtime": 863.1868,
+    "train_samples_per_second": 17.141,
+    "train_steps_per_second": 0.132
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.94,
     "total_flos": 1.1265748270391624e+18,
-    "train_loss": 0.6647836283633584,
-    "train_runtime": 6473.0815,
-    "train_samples_per_second": 2.286,
-    "train_steps_per_second": 0.018
+    "train_loss": 0.2613685036960401,
+    "train_runtime": 863.1868,
+    "train_samples_per_second": 17.141,
+    "train_steps_per_second": 0.132
 }

trainer_state.json

@@ -11,125 +11,125 @@
     {
       "epoch": 0.03,
       "learning_rate": 4.166666666666667e-06,
-      "loss": 1.3431,
+      "loss": 0.3991,
       "step": 1
     },
     {
       "epoch": 0.31,
       "learning_rate": 5e-05,
-      "loss": 1.226,
+      "loss": 0.3808,
       "step": 12
     },
     {
       "epoch": 0.62,
       "learning_rate": 4.411764705882353e-05,
-      "loss": 0.9888,
+      "loss": 0.3474,
       "step": 24
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.8235294117647055e-05,
-      "loss": 0.8637,
+      "loss": 0.3093,
       "step": 36
     },
     {
       "epoch": 0.98,
-      "eval_Macro F1": 0.5661973571975384,
-      "eval_Macro Precision": 0.7510967291822963,
-      "eval_Macro Recall": 0.6473652579844594,
-      "eval_Micro F1": 0.6778127458693942,
-      "eval_Micro Precision": 0.6778127458693942,
-      "eval_Micro Recall": 0.6778127458693942,
-      "eval_Weighted F1": 0.5949172262865288,
-      "eval_Weighted Precision": 0.7505686242573377,
-      "eval_Weighted Recall": 0.6778127458693942,
-      "eval_accuracy": 0.6778127458693942,
-      "eval_loss": 0.598620593547821,
-      "eval_runtime": 1926.4978,
-      "eval_samples_per_second": 1.319,
-      "eval_steps_per_second": 0.042,
+      "eval_Macro F1": 0.8820015735129666,
+      "eval_Macro Precision": 0.899385909210207,
+      "eval_Macro Recall": 0.8738252833106378,
+      "eval_Micro F1": 0.8890637293469709,
+      "eval_Micro Precision": 0.8890637293469709,
+      "eval_Micro Recall": 0.8890637293469709,
+      "eval_Weighted F1": 0.8879137250311478,
+      "eval_Weighted Precision": 0.8952120340834557,
+      "eval_Weighted Recall": 0.8890637293469709,
+      "eval_accuracy": 0.8890637293469709,
+      "eval_loss": 0.2252058982849121,
+      "eval_runtime": 64.7982,
+      "eval_samples_per_second": 39.229,
+      "eval_steps_per_second": 1.235,
       "step": 38
     },
     {
       "epoch": 1.24,
       "learning_rate": 3.235294117647059e-05,
-      "loss": 0.6928,
+      "loss": 0.2358,
       "step": 48
     },
     {
       "epoch": 1.55,
       "learning_rate": 2.647058823529412e-05,
-      "loss": 0.6054,
+      "loss": 0.2391,
       "step": 60
     },
     {
       "epoch": 1.86,
       "learning_rate": 2.058823529411765e-05,
-      "loss": 0.5018,
+      "loss": 0.2278,
       "step": 72
     },
     {
       "epoch": 1.99,
-      "eval_Macro F1": 0.8376972138667833,
-      "eval_Macro Precision": 0.8622787078319065,
-      "eval_Macro Recall": 0.8344563153022619,
-      "eval_Micro F1": 0.8516915814319432,
-      "eval_Micro Precision": 0.8516915814319433,
-      "eval_Micro Recall": 0.8516915814319433,
-      "eval_Weighted F1": 0.8452584084591334,
-      "eval_Weighted Precision": 0.8588424441061323,
-      "eval_Weighted Recall": 0.8516915814319433,
-      "eval_accuracy": 0.8516915814319433,
-      "eval_loss": 0.270693838596344,
-      "eval_runtime": 63.8305,
-      "eval_samples_per_second": 39.824,
-      "eval_steps_per_second": 1.253,
+      "eval_Macro F1": 0.9220395057673447,
+      "eval_Macro Precision": 0.9240847889405631,
+      "eval_Macro Recall": 0.9220981981645411,
+      "eval_Micro F1": 0.9291896144767899,
+      "eval_Micro Precision": 0.9291896144767899,
+      "eval_Micro Recall": 0.9291896144767899,
+      "eval_Weighted F1": 0.929177891947533,
+      "eval_Weighted Precision": 0.9309640796953544,
+      "eval_Weighted Recall": 0.9291896144767899,
+      "eval_accuracy": 0.9291896144767899,
+      "eval_loss": 0.16484124958515167,
+      "eval_runtime": 64.4008,
+      "eval_samples_per_second": 39.472,
+      "eval_steps_per_second": 1.242,
       "step": 77
     },
     {
       "epoch": 2.17,
       "learning_rate": 1.4705882352941177e-05,
-      "loss": 0.4449,
+      "loss": 0.1989,
       "step": 84
     },
     {
       "epoch": 2.48,
       "learning_rate": 8.823529411764707e-06,
-      "loss": 0.4168,
+      "loss": 0.2251,
       "step": 96
     },
     {
       "epoch": 2.79,
       "learning_rate": 2.9411764705882355e-06,
-      "loss": 0.3761,
+      "loss": 0.2066,
       "step": 108
     },
     {
       "epoch": 2.94,
-      "eval_Macro F1": 0.8640166757623045,
-      "eval_Macro Precision": 0.8744992381555277,
-      "eval_Macro Recall": 0.8626385930557097,
-      "eval_Micro F1": 0.8741148701809599,
-      "eval_Micro Precision": 0.8741148701809599,
-      "eval_Micro Recall": 0.8741148701809599,
-      "eval_Weighted F1": 0.8719933502245091,
-      "eval_Weighted Precision": 0.8780283114822369,
-      "eval_Weighted Recall": 0.8741148701809599,
-      "eval_accuracy": 0.8741148701809599,
-      "eval_loss": 0.22845804691314697,
-      "eval_runtime": 63.2765,
-      "eval_samples_per_second": 40.173,
-      "eval_steps_per_second": 1.264,
+      "eval_Macro F1": 0.9386216440513059,
+      "eval_Macro Precision": 0.9382250515595165,
+      "eval_Macro Recall": 0.9403801182311695,
+      "eval_Micro F1": 0.9425649095200629,
+      "eval_Micro Precision": 0.9425649095200629,
+      "eval_Micro Recall": 0.9425649095200629,
+      "eval_Weighted F1": 0.9426485199157074,
+      "eval_Weighted Precision": 0.9439679024820588,
+      "eval_Weighted Recall": 0.9425649095200629,
+      "eval_accuracy": 0.9425649095200629,
+      "eval_loss": 0.13585534691810608,
+      "eval_runtime": 63.8941,
+      "eval_samples_per_second": 39.785,
+      "eval_steps_per_second": 1.252,
       "step": 114
     },
     {
       "epoch": 2.94,
       "step": 114,
       "total_flos": 1.1265748270391624e+18,
-      "train_loss": 0.6647836283633584,
-      "train_runtime": 6473.0815,
-      "train_samples_per_second": 2.286,
-      "train_steps_per_second": 0.018
+      "train_loss": 0.2613685036960401,
+      "train_runtime": 863.1868,
+      "train_samples_per_second": 17.141,
+      "train_steps_per_second": 0.132
     }
   ],
   "logging_steps": 12,