End of training

all_results.json

@@ -1,18 +1,18 @@
 {
     "epoch": 3.0,
-    "eval_gen_len": 47.36,
-    "eval_loss": 2.074782371520996,
-    "eval_rouge1": 27.0446,
-    "eval_rouge2": 7.9355,
-    "eval_rougeL": 18.5537,
-    "eval_rougeLsum": 23.2386,
-    "eval_runtime": 31.8414,
-    "eval_samples": 50,
-    "eval_samples_per_second": 1.57,
-    "eval_steps_per_second": 0.408,
-    "train_loss": 2.2608707682291667,
-    "train_runtime": 447.225,
-    "train_samples": 500,
-    "train_samples_per_second": 3.354,
-    "train_steps_per_second": 0.839
+    "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,
+    "eval_samples": 1500,
+    "eval_samples_per_second": 2.678,
+    "eval_steps_per_second": 0.669,
+    "train_loss": 2.4165630696614584,
+    "train_runtime": 3505.1584,
+    "train_samples": 7000,
+    "train_samples_per_second": 5.991,
+    "train_steps_per_second": 0.749
 }

eval_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 3.0,
-    "eval_gen_len": 47.36,
-    "eval_loss": 2.074782371520996,
-    "eval_rouge1": 27.0446,
-    "eval_rouge2": 7.9355,
-    "eval_rougeL": 18.5537,
-    "eval_rougeLsum": 23.2386,
-    "eval_runtime": 31.8414,
-    "eval_samples": 50,
-    "eval_samples_per_second": 1.57,
-    "eval_steps_per_second": 0.408
+    "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,
+    "eval_samples": 1500,
+    "eval_samples_per_second": 2.678,
+    "eval_steps_per_second": 0.669
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
-    "train_loss": 2.2608707682291667,
-    "train_runtime": 447.225,
-    "train_samples": 500,
-    "train_samples_per_second": 3.354,
-    "train_steps_per_second": 0.839
+    "train_loss": 2.4165630696614584,
+    "train_runtime": 3505.1584,
+    "train_samples": 7000,
+    "train_samples_per_second": 5.991,
+    "train_steps_per_second": 0.749
 }

trainer_state.json

@@ -2,24 +2,54 @@
   "best_metric": null,
   "best_model_checkpoint": null,
   "epoch": 3.0,
-  "global_step": 375,
+  "global_step": 2625,
   "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,
+      "step": 500
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 2.489,
+      "step": 1000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 2.3848,
+      "step": 1500
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 2.3436,
+      "step": 2000
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 2.3011,
+      "step": 2500
+    },
     {
       "epoch": 3.0,
-      "step": 375,
-      "total_flos": 909288310210560.0,
-      "train_loss": 2.2608707682291667,
-      "train_runtime": 447.225,
-      "train_samples_per_second": 3.354,
-      "train_steps_per_second": 0.839
+      "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
     }
   ],
-  "max_steps": 375,
+  "max_steps": 2625,
   "num_train_epochs": 3,
-  "total_flos": 909288310210560.0,
+  "total_flos": 1.278595580424192e+16,
   "trial_name": null,
   "trial_params": null
 }