End of training

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.94,
-    "total_flos": 1.1265748270391624e+18,
-    "train_loss": 0.16734615334293299,
-    "train_runtime": 848.5959,
-    "train_samples_per_second": 17.436,
-    "train_steps_per_second": 0.134
+    "epoch": 3.0,
+    "total_flos": 2.320189226545914e+18,
+    "train_loss": 0.18204877430047745,
+    "train_runtime": 4369.7043,
+    "train_samples_per_second": 6.851,
+    "train_steps_per_second": 0.054
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.94,
-    "total_flos": 1.1265748270391624e+18,
-    "train_loss": 0.16734615334293299,
-    "train_runtime": 848.5959,
-    "train_samples_per_second": 17.436,
-    "train_steps_per_second": 0.134
+    "epoch": 3.0,
+    "total_flos": 2.320189226545914e+18,
+    "train_loss": 0.18204877430047745,
+    "train_runtime": 4369.7043,
+    "train_samples_per_second": 6.851,
+    "train_steps_per_second": 0.054
 }

trainer_state.json

@@ -1,142 +1,202 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.9419354838709677,
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 114,
+  "global_step": 234,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.03,
-      "learning_rate": 4.166666666666667e-06,
-      "loss": 0.1438,
+      "epoch": 0.01,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.1764,
       "step": 1
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 5e-05,
-      "loss": 0.146,
+      "epoch": 0.15,
+      "learning_rate": 2.5e-05,
+      "loss": 0.2288,
       "step": 12
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.411764705882353e-05,
-      "loss": 0.1822,
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2324,
       "step": 24
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 3.8235294117647055e-05,
-      "loss": 0.1916,
+      "epoch": 0.46,
+      "learning_rate": 4.714285714285714e-05,
+      "loss": 0.2094,
       "step": 36
     },
     {
-      "epoch": 0.98,
-      "eval_Macro F1": 0.9408451121862865,
-      "eval_Macro Precision": 0.9411664777049882,
-      "eval_Macro Recall": 0.9427118741264285,
-      "eval_Micro F1": 0.9461054287962234,
-      "eval_Micro Precision": 0.9461054287962234,
-      "eval_Micro Recall": 0.9461054287962234,
-      "eval_Weighted F1": 0.9464960091472664,
-      "eval_Weighted Precision": 0.9487006349374144,
-      "eval_Weighted Recall": 0.9461054287962234,
-      "eval_accuracy": 0.9461054287962234,
-      "eval_loss": 0.13961628079414368,
-      "eval_runtime": 63.2832,
-      "eval_samples_per_second": 40.169,
-      "eval_steps_per_second": 1.264,
-      "step": 38
-    },
-    {
-      "epoch": 1.24,
-      "learning_rate": 3.235294117647059e-05,
-      "loss": 0.1565,
+      "epoch": 0.62,
+      "learning_rate": 4.428571428571428e-05,
+      "loss": 0.2219,
       "step": 48
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 2.647058823529412e-05,
-      "loss": 0.1692,
+      "epoch": 0.77,
+      "learning_rate": 4.1428571428571437e-05,
+      "loss": 0.2048,
       "step": 60
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 2.058823529411765e-05,
-      "loss": 0.1597,
+      "epoch": 0.92,
+      "learning_rate": 3.857142857142858e-05,
+      "loss": 0.2337,
       "step": 72
     },
     {
-      "epoch": 1.99,
-      "eval_Macro F1": 0.9484522652918805,
-      "eval_Macro Precision": 0.9471910825399321,
-      "eval_Macro Recall": 0.95148334257645,
-      "eval_Micro F1": 0.9520062942564909,
-      "eval_Micro Precision": 0.952006294256491,
-      "eval_Micro Recall": 0.952006294256491,
-      "eval_Weighted F1": 0.9522977862410118,
-      "eval_Weighted Precision": 0.9540997275393984,
-      "eval_Weighted Recall": 0.952006294256491,
-      "eval_accuracy": 0.952006294256491,
-      "eval_loss": 0.12265925854444504,
-      "eval_runtime": 64.9371,
-      "eval_samples_per_second": 39.146,
-      "eval_steps_per_second": 1.232,
-      "step": 77
-    },
-    {
-      "epoch": 2.17,
-      "learning_rate": 1.4705882352941177e-05,
-      "loss": 0.1619,
+      "epoch": 1.0,
+      "eval_Macro F1": 0.9058241520350262,
+      "eval_Macro Precision": 0.9220297823287216,
+      "eval_Macro Recall": 0.9040021601099654,
+      "eval_Micro F1": 0.9087221095334685,
+      "eval_Micro Precision": 0.9087221095334685,
+      "eval_Micro Recall": 0.9087221095334685,
+      "eval_Weighted F1": 0.9097572645217704,
+      "eval_Weighted Precision": 0.9228816407309685,
+      "eval_Weighted Recall": 0.9087221095334685,
+      "eval_accuracy": 0.9087221095334685,
+      "eval_loss": 0.26681017875671387,
+      "eval_runtime": 509.7519,
+      "eval_samples_per_second": 3.869,
+      "eval_steps_per_second": 0.122,
+      "step": 78
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.1943,
       "step": 84
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 8.823529411764707e-06,
-      "loss": 0.1611,
+      "epoch": 1.23,
+      "learning_rate": 3.285714285714286e-05,
+      "loss": 0.2022,
       "step": 96
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.9411764705882355e-06,
-      "loss": 0.1722,
+      "epoch": 1.38,
+      "learning_rate": 3e-05,
+      "loss": 0.1662,
       "step": 108
     },
     {
-      "epoch": 2.94,
-      "eval_Macro F1": 0.947301058449996,
-      "eval_Macro Precision": 0.9458287584858323,
-      "eval_Macro Recall": 0.9498167521420487,
-      "eval_Micro F1": 0.9512195121951219,
-      "eval_Micro Precision": 0.9512195121951219,
-      "eval_Micro Recall": 0.9512195121951219,
-      "eval_Weighted F1": 0.9515018097434871,
-      "eval_Weighted Precision": 0.9527033488356239,
-      "eval_Weighted Recall": 0.9512195121951219,
-      "eval_accuracy": 0.9512195121951219,
-      "eval_loss": 0.11822786182165146,
-      "eval_runtime": 63.9617,
-      "eval_samples_per_second": 39.743,
-      "eval_steps_per_second": 1.251,
-      "step": 114
-    },
-    {
-      "epoch": 2.94,
-      "step": 114,
-      "total_flos": 1.1265748270391624e+18,
-      "train_loss": 0.16734615334293299,
-      "train_runtime": 848.5959,
-      "train_samples_per_second": 17.436,
-      "train_steps_per_second": 0.134
+      "epoch": 1.54,
+      "learning_rate": 2.714285714285714e-05,
+      "loss": 0.1906,
+      "step": 120
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.4285714285714288e-05,
+      "loss": 0.1614,
+      "step": 132
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.16,
+      "step": 144
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8571428571428572e-05,
+      "loss": 0.1711,
+      "step": 156
+    },
+    {
+      "epoch": 2.0,
+      "eval_Macro F1": 0.9330778801188252,
+      "eval_Macro Precision": 0.9291584729568021,
+      "eval_Macro Recall": 0.9402683529435533,
+      "eval_Micro F1": 0.9376267748478702,
+      "eval_Micro Precision": 0.9376267748478702,
+      "eval_Micro Recall": 0.9376267748478702,
+      "eval_Weighted F1": 0.9380245851334839,
+      "eval_Weighted Precision": 0.9415569462618286,
+      "eval_Weighted Recall": 0.9376267748478702,
+      "eval_accuracy": 0.9376267748478702,
+      "eval_loss": 0.1819760948419571,
+      "eval_runtime": 52.791,
+      "eval_samples_per_second": 37.355,
+      "eval_steps_per_second": 1.174,
+      "step": 156
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.1743,
+      "step": 168
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.2857142857142857e-05,
+      "loss": 0.1607,
+      "step": 180
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1e-05,
+      "loss": 0.1675,
+      "step": 192
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.1393,
+      "step": 204
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.285714285714286e-06,
+      "loss": 0.1308,
+      "step": 216
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.1297,
+      "step": 228
+    },
+    {
+      "epoch": 3.0,
+      "eval_Macro F1": 0.9504520666030198,
+      "eval_Macro Precision": 0.9506141996378501,
+      "eval_Macro Recall": 0.9523141162334021,
+      "eval_Micro F1": 0.9523326572008114,
+      "eval_Micro Precision": 0.9523326572008114,
+      "eval_Micro Recall": 0.9523326572008114,
+      "eval_Weighted F1": 0.9523782212526937,
+      "eval_Weighted Precision": 0.9543848061355097,
+      "eval_Weighted Recall": 0.9523326572008114,
+      "eval_accuracy": 0.9523326572008114,
+      "eval_loss": 0.1481797993183136,
+      "eval_runtime": 51.55,
+      "eval_samples_per_second": 38.254,
+      "eval_steps_per_second": 1.203,
+      "step": 234
+    },
+    {
+      "epoch": 3.0,
+      "step": 234,
+      "total_flos": 2.320189226545914e+18,
+      "train_loss": 0.18204877430047745,
+      "train_runtime": 4369.7043,
+      "train_samples_per_second": 6.851,
+      "train_steps_per_second": 0.054
     }
   ],
   "logging_steps": 12,
-  "max_steps": 114,
+  "max_steps": 234,
   "num_train_epochs": 3,
   "save_steps": 500,
-  "total_flos": 1.1265748270391624e+18,
+  "total_flos": 2.320189226545914e+18,
   "trial_name": null,
   "trial_params": null
 }