End of training

all_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 8.0,
-    "train_loss": 1.7074453571236545,
-    "train_runtime": 28108.9431,
-    "train_samples_per_second": 28.461,
-    "train_steps_per_second": 0.222
+    "epoch": 16.0,
+    "train_loss": 0.8017544373965294,
+    "train_runtime": 29679.9237,
+    "train_samples_per_second": 53.908,
+    "train_steps_per_second": 0.421
 }

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 8.0,
-    "train_loss": 1.7074453571236545,
-    "train_runtime": 28108.9431,
-    "train_samples_per_second": 28.461,
-    "train_steps_per_second": 0.222
+    "epoch": 16.0,
+    "train_loss": 0.8017544373965294,
+    "train_runtime": 29679.9237,
+    "train_samples_per_second": 53.908,
+    "train_steps_per_second": 0.421
 }

trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 7.99744,
+  "epoch": 15.99712,
   "eval_steps": 500,
-  "global_step": 6248,
+  "global_step": 12496,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -195,35 +195,235 @@
     {
       "epoch": 8.0,
       "eval_f1": 0.9123,
-      "eval_gen_len": 25.38509090909091,
-      "eval_loss": 1.540870189666748,
+      "eval_gen_len": 25.383636363636363,
+      "eval_loss": 1.5380274057388306,
       "eval_precision": 0.9131,
-      "eval_recall": 0.9117,
-      "eval_rouge1": 0.4869,
-      "eval_rouge2": 0.2373,
-      "eval_rougeL": 0.406,
-      "eval_rougeLsum": 0.4058,
-      "eval_runtime": 863.7281,
-      "eval_samples_per_second": 6.368,
-      "eval_steps_per_second": 0.398,
+      "eval_recall": 0.9118,
+      "eval_rouge1": 0.4872,
+      "eval_rouge2": 0.2387,
+      "eval_rougeL": 0.407,
+      "eval_rougeLsum": 0.4071,
+      "eval_runtime": 863.1004,
+      "eval_samples_per_second": 6.372,
+      "eval_steps_per_second": 0.399,
       "step": 6248
     },
     {
-      "epoch": 8.0,
-      "step": 6248,
-      "total_flos": 1.1554159142381814e+18,
-      "train_loss": 1.7074453571236545,
-      "train_runtime": 28108.9431,
-      "train_samples_per_second": 28.461,
-      "train_steps_per_second": 0.222
+      "epoch": 8.32,
+      "learning_rate": 9.596670934699104e-06,
+      "loss": 1.5886,
+      "step": 6500
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 8.796414852752882e-06,
+      "loss": 1.5836,
+      "step": 7000
+    },
+    {
+      "epoch": 9.0,
+      "eval_f1": 0.9126,
+      "eval_gen_len": 25.499454545454544,
+      "eval_loss": 1.527321696281433,
+      "eval_precision": 0.9133,
+      "eval_recall": 0.9122,
+      "eval_rouge1": 0.4891,
+      "eval_rouge2": 0.2399,
+      "eval_rougeL": 0.4088,
+      "eval_rougeLsum": 0.4089,
+      "eval_runtime": 859.8962,
+      "eval_samples_per_second": 6.396,
+      "eval_steps_per_second": 0.4,
+      "step": 7029
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 7.996158770806658e-06,
+      "loss": 1.5667,
+      "step": 7500
+    },
+    {
+      "epoch": 10.0,
+      "eval_f1": 0.9127,
+      "eval_gen_len": 25.386727272727274,
+      "eval_loss": 1.5195879936218262,
+      "eval_precision": 0.9135,
+      "eval_recall": 0.9123,
+      "eval_rouge1": 0.4906,
+      "eval_rouge2": 0.2416,
+      "eval_rougeL": 0.411,
+      "eval_rougeLsum": 0.4112,
+      "eval_runtime": 855.4017,
+      "eval_samples_per_second": 6.43,
+      "eval_steps_per_second": 0.402,
+      "step": 7810
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 7.195902688860435e-06,
+      "loss": 1.5577,
+      "step": 8000
+    },
+    {
+      "epoch": 10.88,
+      "learning_rate": 6.395646606914213e-06,
+      "loss": 1.5521,
+      "step": 8500
+    },
+    {
+      "epoch": 11.0,
+      "eval_f1": 0.9127,
+      "eval_gen_len": 25.21909090909091,
+      "eval_loss": 1.5124093294143677,
+      "eval_precision": 0.9137,
+      "eval_recall": 0.912,
+      "eval_rouge1": 0.4899,
+      "eval_rouge2": 0.2406,
+      "eval_rougeL": 0.4102,
+      "eval_rougeLsum": 0.4103,
+      "eval_runtime": 851.6688,
+      "eval_samples_per_second": 6.458,
+      "eval_steps_per_second": 0.404,
+      "step": 8592
+    },
+    {
+      "epoch": 11.52,
+      "learning_rate": 5.595390524967991e-06,
+      "loss": 1.5413,
+      "step": 9000
+    },
+    {
+      "epoch": 12.0,
+      "eval_f1": 0.9128,
+      "eval_gen_len": 25.349090909090908,
+      "eval_loss": 1.5083255767822266,
+      "eval_precision": 0.9137,
+      "eval_recall": 0.9123,
+      "eval_rouge1": 0.4914,
+      "eval_rouge2": 0.2416,
+      "eval_rougeL": 0.4118,
+      "eval_rougeLsum": 0.412,
+      "eval_runtime": 856.2804,
+      "eval_samples_per_second": 6.423,
+      "eval_steps_per_second": 0.402,
+      "step": 9373
+    },
+    {
+      "epoch": 12.16,
+      "learning_rate": 4.795134443021768e-06,
+      "loss": 1.5354,
+      "step": 9500
+    },
+    {
+      "epoch": 12.8,
+      "learning_rate": 3.9948783610755446e-06,
+      "loss": 1.5291,
+      "step": 10000
+    },
+    {
+      "epoch": 13.0,
+      "eval_f1": 0.913,
+      "eval_gen_len": 25.208181818181817,
+      "eval_loss": 1.5044068098068237,
+      "eval_precision": 0.914,
+      "eval_recall": 0.9123,
+      "eval_rouge1": 0.4913,
+      "eval_rouge2": 0.2419,
+      "eval_rougeL": 0.4118,
+      "eval_rougeLsum": 0.4119,
+      "eval_runtime": 858.747,
+      "eval_samples_per_second": 6.405,
+      "eval_steps_per_second": 0.401,
+      "step": 10154
+    },
+    {
+      "epoch": 13.44,
+      "learning_rate": 3.194622279129322e-06,
+      "loss": 1.527,
+      "step": 10500
+    },
+    {
+      "epoch": 14.0,
+      "eval_f1": 0.913,
+      "eval_gen_len": 25.10690909090909,
+      "eval_loss": 1.5025616884231567,
+      "eval_precision": 0.9141,
+      "eval_recall": 0.9123,
+      "eval_rouge1": 0.4917,
+      "eval_rouge2": 0.2426,
+      "eval_rougeL": 0.4126,
+      "eval_rougeLsum": 0.4128,
+      "eval_runtime": 849.8377,
+      "eval_samples_per_second": 6.472,
+      "eval_steps_per_second": 0.405,
+      "step": 10935
+    },
+    {
+      "epoch": 14.08,
+      "learning_rate": 2.3943661971830984e-06,
+      "loss": 1.5206,
+      "step": 11000
+    },
+    {
+      "epoch": 14.72,
+      "learning_rate": 1.594110115236876e-06,
+      "loss": 1.5203,
+      "step": 11500
+    },
+    {
+      "epoch": 15.0,
+      "eval_f1": 0.9131,
+      "eval_gen_len": 25.106181818181817,
+      "eval_loss": 1.5006238222122192,
+      "eval_precision": 0.9143,
+      "eval_recall": 0.9123,
+      "eval_rouge1": 0.4921,
+      "eval_rouge2": 0.243,
+      "eval_rougeL": 0.4135,
+      "eval_rougeLsum": 0.4136,
+      "eval_runtime": 854.2433,
+      "eval_samples_per_second": 6.438,
+      "eval_steps_per_second": 0.403,
+      "step": 11717
+    },
+    {
+      "epoch": 15.36,
+      "learning_rate": 7.93854033290653e-07,
+      "loss": 1.5126,
+      "step": 12000
+    },
+    {
+      "epoch": 16.0,
+      "eval_f1": 0.9132,
+      "eval_gen_len": 25.133454545454544,
+      "eval_loss": 1.5003753900527954,
+      "eval_precision": 0.9143,
+      "eval_recall": 0.9124,
+      "eval_rouge1": 0.4923,
+      "eval_rouge2": 0.2429,
+      "eval_rougeL": 0.4134,
+      "eval_rougeLsum": 0.4134,
+      "eval_runtime": 851.4198,
+      "eval_samples_per_second": 6.46,
+      "eval_steps_per_second": 0.404,
+      "step": 12496
+    },
+    {
+      "epoch": 16.0,
+      "step": 12496,
+      "total_flos": 2.310831828476363e+18,
+      "train_loss": 0.8017544373965294,
+      "train_runtime": 29679.9237,
+      "train_samples_per_second": 53.908,
+      "train_steps_per_second": 0.421
     }
   ],
   "logging_steps": 500,
-  "max_steps": 6248,
+  "max_steps": 12496,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 8,
+  "num_train_epochs": 16,
   "save_steps": 500,
-  "total_flos": 1.1554159142381814e+18,
+  "total_flos": 2.310831828476363e+18,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null