End of training

all_results.json

@@ -1,18 +1,18 @@
 {
-    "epoch": 3.0,
-    "eval_gen_len": 55.018,
-    "eval_loss": 2.2143824100494385,
-    "eval_rouge1": 26.012,
-    "eval_rouge2": 7.0961,
-    "eval_rougeL": 18.1094,
-    "eval_rougeLsum": 22.507,
-    "eval_runtime": 560.2144,
+    "epoch": 10.0,
+    "eval_gen_len": 52.717333333333336,
+    "eval_loss": 2.173374652862549,
+    "eval_rouge1": 26.8011,
+    "eval_rouge2": 8.2533,
+    "eval_rougeL": 18.7776,
+    "eval_rougeLsum": 23.2859,
+    "eval_runtime": 369.7996,
     "eval_samples": 1500,
-    "eval_samples_per_second": 2.678,
-    "eval_steps_per_second": 0.669,
-    "train_loss": 2.4165630696614584,
-    "train_runtime": 3505.1584,
+    "eval_samples_per_second": 4.056,
+    "eval_steps_per_second": 0.508,
+    "train_loss": 2.152468052455357,
+    "train_runtime": 10217.243,
     "train_samples": 7000,
-    "train_samples_per_second": 5.991,
-    "train_steps_per_second": 0.749
+    "train_samples_per_second": 6.851,
+    "train_steps_per_second": 0.856
 }

eval_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 3.0,
-    "eval_gen_len": 55.018,
-    "eval_loss": 2.2143824100494385,
-    "eval_rouge1": 26.012,
-    "eval_rouge2": 7.0961,
-    "eval_rougeL": 18.1094,
-    "eval_rougeLsum": 22.507,
-    "eval_runtime": 560.2144,
+    "epoch": 10.0,
+    "eval_gen_len": 52.717333333333336,
+    "eval_loss": 2.173374652862549,
+    "eval_rouge1": 26.8011,
+    "eval_rouge2": 8.2533,
+    "eval_rougeL": 18.7776,
+    "eval_rougeLsum": 23.2859,
+    "eval_runtime": 369.7996,
     "eval_samples": 1500,
-    "eval_samples_per_second": 2.678,
-    "eval_steps_per_second": 0.669
+    "eval_samples_per_second": 4.056,
+    "eval_steps_per_second": 0.508
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "train_loss": 2.4165630696614584,
-    "train_runtime": 3505.1584,
+    "epoch": 10.0,
+    "train_loss": 2.152468052455357,
+    "train_runtime": 10217.243,
     "train_samples": 7000,
-    "train_samples_per_second": 5.991,
-    "train_steps_per_second": 0.749
+    "train_samples_per_second": 6.851,
+    "train_steps_per_second": 0.856
 }

trainer_state.json

@@ -1,55 +1,127 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.0,
-  "global_step": 2625,
+  "epoch": 10.0,
+  "global_step": 8750,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.57,
-      "learning_rate": 4.047619047619048e-05,
-      "loss": 2.5903,
+      "learning_rate": 4.714285714285714e-05,
+      "loss": 2.5639,
       "step": 500
     },
     {
       "epoch": 1.14,
-      "learning_rate": 3.095238095238095e-05,
-      "loss": 2.489,
+      "learning_rate": 4.428571428571428e-05,
+      "loss": 2.4439,
       "step": 1000
     },
     {
       "epoch": 1.71,
-      "learning_rate": 2.1428571428571428e-05,
-      "loss": 2.3848,
+      "learning_rate": 4.1428571428571437e-05,
+      "loss": 2.3715,
       "step": 1500
     },
     {
       "epoch": 2.29,
-      "learning_rate": 1.1904761904761905e-05,
-      "loss": 2.3436,
+      "learning_rate": 3.857142857142858e-05,
+      "loss": 2.2965,
       "step": 2000
     },
     {
       "epoch": 2.86,
-      "learning_rate": 2.3809523809523808e-06,
-      "loss": 2.3011,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 2.2701,
       "step": 2500
     },
     {
-      "epoch": 3.0,
-      "step": 2625,
-      "total_flos": 1.278595580424192e+16,
-      "train_loss": 2.4165630696614584,
-      "train_runtime": 3505.1584,
-      "train_samples_per_second": 5.991,
-      "train_steps_per_second": 0.749
+      "epoch": 3.43,
+      "learning_rate": 3.285714285714286e-05,
+      "loss": 2.2025,
+      "step": 3000
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3e-05,
+      "loss": 2.1752,
+      "step": 3500
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.714285714285714e-05,
+      "loss": 2.1169,
+      "step": 4000
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 2.4285714285714288e-05,
+      "loss": 2.1151,
+      "step": 4500
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 2.0713,
+      "step": 5000
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 1.8571428571428572e-05,
+      "loss": 2.0616,
+      "step": 5500
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 2.0385,
+      "step": 6000
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 1.2857142857142857e-05,
+      "loss": 2.0265,
+      "step": 6500
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 1e-05,
+      "loss": 1.9998,
+      "step": 7000
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 1.9831,
+      "step": 7500
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 4.285714285714286e-06,
+      "loss": 1.9831,
+      "step": 8000
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 1.9695,
+      "step": 8500
+    },
+    {
+      "epoch": 10.0,
+      "step": 8750,
+      "total_flos": 4.262140044877824e+16,
+      "train_loss": 2.152468052455357,
+      "train_runtime": 10217.243,
+      "train_samples_per_second": 6.851,
+      "train_steps_per_second": 0.856
     }
   ],
-  "max_steps": 2625,
-  "num_train_epochs": 3,
-  "total_flos": 1.278595580424192e+16,
+  "max_steps": 8750,
+  "num_train_epochs": 10,
+  "total_flos": 4.262140044877824e+16,
   "trial_name": null,
   "trial_params": null
 }