Training in progress, epoch 1

all_results.json

@@ -1,26 +1,26 @@
 {
-    "epoch": 21.0,
-    "eval_gen_len": 17.2444,
-    "eval_loss": 2.5511105060577393,
-    "eval_rouge-1": 65.3788,
-    "eval_rouge-2": 55.3744,
-    "eval_rouge-l": 65.0206,
-    "eval_runtime": 65.3718,
+    "epoch": 13.0,
+    "eval_gen_len": 15.7444,
+    "eval_loss": 2.75215744972229,
+    "eval_rouge-1": 42.9172,
+    "eval_rouge-2": 25.6169,
+    "eval_rouge-l": 42.4864,
+    "eval_runtime": 62.7032,
     "eval_samples": 450,
-    "eval_samples_per_second": 6.884,
-    "eval_steps_per_second": 0.872,
-    "predict_gen_len": 16.777,
-    "predict_loss": 2.622779369354248,
-    "predict_rouge-1": 65.0566,
-    "predict_rouge-2": 53.9337,
-    "predict_rouge-l": 64.5625,
-    "predict_runtime": 58.2188,
+    "eval_samples_per_second": 7.177,
+    "eval_steps_per_second": 0.909,
+    "predict_gen_len": 15.4868,
+    "predict_loss": 2.8386902809143066,
+    "predict_rouge-1": 42.8629,
+    "predict_rouge-2": 25.4871,
+    "predict_rouge-l": 42.0716,
+    "predict_runtime": 55.2936,
     "predict_samples": 417,
-    "predict_samples_per_second": 7.163,
-    "predict_steps_per_second": 0.91,
-    "train_loss": 2.2238484863498553,
-    "train_runtime": 6515.1509,
+    "predict_samples_per_second": 7.542,
+    "predict_steps_per_second": 0.959,
+    "train_loss": 3.136590390419063,
+    "train_runtime": 4227.0028,
     "train_samples": 5133,
-    "train_samples_per_second": 39.393,
-    "train_steps_per_second": 4.927
+    "train_samples_per_second": 60.717,
+    "train_steps_per_second": 7.594
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 21.0,
-    "eval_gen_len": 17.2444,
-    "eval_loss": 2.5511105060577393,
-    "eval_rouge-1": 65.3788,
-    "eval_rouge-2": 55.3744,
-    "eval_rouge-l": 65.0206,
-    "eval_runtime": 65.3718,
+    "epoch": 13.0,
+    "eval_gen_len": 15.7444,
+    "eval_loss": 2.75215744972229,
+    "eval_rouge-1": 42.9172,
+    "eval_rouge-2": 25.6169,
+    "eval_rouge-l": 42.4864,
+    "eval_runtime": 62.7032,
     "eval_samples": 450,
-    "eval_samples_per_second": 6.884,
-    "eval_steps_per_second": 0.872
+    "eval_samples_per_second": 7.177,
+    "eval_steps_per_second": 0.909
 }

predict_results.json

@@ -1,11 +1,11 @@
 {
-    "predict_gen_len": 16.777,
-    "predict_loss": 2.622779369354248,
-    "predict_rouge-1": 65.0566,
-    "predict_rouge-2": 53.9337,
-    "predict_rouge-l": 64.5625,
-    "predict_runtime": 58.2188,
+    "predict_gen_len": 15.4868,
+    "predict_loss": 2.8386902809143066,
+    "predict_rouge-1": 42.8629,
+    "predict_rouge-2": 25.4871,
+    "predict_rouge-l": 42.0716,
+    "predict_runtime": 55.2936,
     "predict_samples": 417,
-    "predict_samples_per_second": 7.163,
-    "predict_steps_per_second": 0.91
+    "predict_samples_per_second": 7.542,
+    "predict_steps_per_second": 0.959
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 21.0,
-    "train_loss": 2.2238484863498553,
-    "train_runtime": 6515.1509,
+    "epoch": 13.0,
+    "train_loss": 3.136590390419063,
+    "train_runtime": 4227.0028,
     "train_samples": 5133,
-    "train_samples_per_second": 39.393,
-    "train_steps_per_second": 4.927
+    "train_samples_per_second": 60.717,
+    "train_steps_per_second": 7.594
 }

trainer_state.json

@@ -1,292 +1,190 @@
 {
-  "best_metric": 2.5511105060577393,
-  "best_model_checkpoint": "PEFT/adapters-lib/output/parallel/dataset-5400/checkpoint-10272",
-  "epoch": 21.0,
+  "best_metric": 2.75215744972229,
+  "best_model_checkpoint": "PEFT/adapters-lib/output/parallel/dataset-5400/checkpoint-5136",
+  "epoch": 13.0,
   "eval_steps": 500,
-  "global_step": 13482,
+  "global_step": 8346,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_gen_len": 19.7933,
-      "eval_loss": 3.2734224796295166,
-      "eval_rouge-1": 30.2297,
-      "eval_rouge-2": 14.9139,
-      "eval_rouge-l": 29.2882,
-      "eval_runtime": 72.7236,
-      "eval_samples_per_second": 6.188,
-      "eval_steps_per_second": 0.784,
+      "eval_gen_len": 16.2889,
+      "eval_loss": 3.750108242034912,
+      "eval_rouge-1": 18.3573,
+      "eval_rouge-2": 7.2689,
+      "eval_rouge-l": 18.0787,
+      "eval_runtime": 67.5801,
+      "eval_samples_per_second": 6.659,
+      "eval_steps_per_second": 0.843,
       "step": 642
     },
     {
       "epoch": 2.0,
-      "eval_gen_len": 17.6289,
-      "eval_loss": 2.915264844894409,
-      "eval_rouge-1": 37.0442,
-      "eval_rouge-2": 19.9571,
-      "eval_rouge-l": 36.4521,
-      "eval_runtime": 67.6182,
-      "eval_samples_per_second": 6.655,
-      "eval_steps_per_second": 0.843,
+      "eval_gen_len": 18.9889,
+      "eval_loss": 3.052328109741211,
+      "eval_rouge-1": 35.6298,
+      "eval_rouge-2": 18.8236,
+      "eval_rouge-l": 35.061,
+      "eval_runtime": 75.9845,
+      "eval_samples_per_second": 5.922,
+      "eval_steps_per_second": 0.75,
       "step": 1284
     },
     {
       "epoch": 3.0,
-      "eval_gen_len": 17.6533,
-      "eval_loss": 2.798626184463501,
-      "eval_rouge-1": 40.7073,
-      "eval_rouge-2": 23.2477,
-      "eval_rouge-l": 40.1741,
-      "eval_runtime": 66.5464,
-      "eval_samples_per_second": 6.762,
-      "eval_steps_per_second": 0.857,
+      "eval_gen_len": 17.7733,
+      "eval_loss": 2.960456371307373,
+      "eval_rouge-1": 38.5671,
+      "eval_rouge-2": 21.0729,
+      "eval_rouge-l": 37.9242,
+      "eval_runtime": 70.7318,
+      "eval_samples_per_second": 6.362,
+      "eval_steps_per_second": 0.806,
       "step": 1926
     },
     {
       "epoch": 4.0,
-      "eval_gen_len": 17.1178,
-      "eval_loss": 2.71970534324646,
-      "eval_rouge-1": 43.788,
-      "eval_rouge-2": 26.5027,
-      "eval_rouge-l": 43.155,
-      "eval_runtime": 64.8425,
-      "eval_samples_per_second": 6.94,
-      "eval_steps_per_second": 0.879,
+      "eval_gen_len": 17.3978,
+      "eval_loss": 2.899169683456421,
+      "eval_rouge-1": 40.4934,
+      "eval_rouge-2": 22.7632,
+      "eval_rouge-l": 39.8547,
+      "eval_runtime": 69.4305,
+      "eval_samples_per_second": 6.481,
+      "eval_steps_per_second": 0.821,
       "step": 2568
     },
     {
       "epoch": 5.0,
-      "eval_gen_len": 16.6556,
-      "eval_loss": 2.6767334938049316,
-      "eval_rouge-1": 45.1804,
-      "eval_rouge-2": 28.1457,
-      "eval_rouge-l": 44.708,
-      "eval_runtime": 62.17,
-      "eval_samples_per_second": 7.238,
-      "eval_steps_per_second": 0.917,
+      "eval_gen_len": 17.2022,
+      "eval_loss": 2.8467295169830322,
+      "eval_rouge-1": 41.7604,
+      "eval_rouge-2": 24.0399,
+      "eval_rouge-l": 41.092,
+      "eval_runtime": 70.3771,
+      "eval_samples_per_second": 6.394,
+      "eval_steps_per_second": 0.81,
       "step": 3210
     },
     {
       "epoch": 6.0,
-      "eval_gen_len": 16.8711,
-      "eval_loss": 2.6528820991516113,
-      "eval_rouge-1": 46.9075,
-      "eval_rouge-2": 31.1211,
-      "eval_rouge-l": 46.3224,
-      "eval_runtime": 64.1515,
-      "eval_samples_per_second": 7.015,
-      "eval_steps_per_second": 0.889,
+      "eval_gen_len": 16.8956,
+      "eval_loss": 2.812547206878662,
+      "eval_rouge-1": 41.5256,
+      "eval_rouge-2": 23.8131,
+      "eval_rouge-l": 40.8416,
+      "eval_runtime": 68.6799,
+      "eval_samples_per_second": 6.552,
+      "eval_steps_per_second": 0.83,
       "step": 3852
     },
     {
       "epoch": 7.0,
-      "eval_gen_len": 16.86,
-      "eval_loss": 2.6175150871276855,
-      "eval_rouge-1": 50.0575,
-      "eval_rouge-2": 34.9346,
-      "eval_rouge-l": 49.5262,
-      "eval_runtime": 64.4679,
-      "eval_samples_per_second": 6.98,
-      "eval_steps_per_second": 0.884,
+      "eval_gen_len": 16.1022,
+      "eval_loss": 2.783888816833496,
+      "eval_rouge-1": 42.7127,
+      "eval_rouge-2": 24.6471,
+      "eval_rouge-l": 42.1972,
+      "eval_runtime": 63.7095,
+      "eval_samples_per_second": 7.063,
+      "eval_steps_per_second": 0.895,
       "step": 4494
     },
     {
       "epoch": 7.79,
       "learning_rate": 8.715112540192926e-05,
-      "loss": 2.7978,
+      "loss": 3.4332,
       "step": 5000
     },
     {
       "epoch": 8.0,
-      "eval_gen_len": 16.7378,
-      "eval_loss": 2.5928914546966553,
-      "eval_rouge-1": 53.5853,
-      "eval_rouge-2": 38.6997,
-      "eval_rouge-l": 52.982,
-      "eval_runtime": 65.4577,
-      "eval_samples_per_second": 6.875,
-      "eval_steps_per_second": 0.871,
+      "eval_gen_len": 15.7444,
+      "eval_loss": 2.75215744972229,
+      "eval_rouge-1": 42.9172,
+      "eval_rouge-2": 25.6169,
+      "eval_rouge-l": 42.4864,
+      "eval_runtime": 62.162,
+      "eval_samples_per_second": 7.239,
+      "eval_steps_per_second": 0.917,
       "step": 5136
     },
     {
       "epoch": 9.0,
-      "eval_gen_len": 16.5556,
-      "eval_loss": 2.575561046600342,
-      "eval_rouge-1": 54.7716,
-      "eval_rouge-2": 40.1626,
-      "eval_rouge-l": 54.1312,
-      "eval_runtime": 63.6914,
-      "eval_samples_per_second": 7.065,
-      "eval_steps_per_second": 0.895,
+      "eval_gen_len": 19.3733,
+      "eval_loss": 3.0449774265289307,
+      "eval_rouge-1": 33.8509,
+      "eval_rouge-2": 18.7234,
+      "eval_rouge-l": 33.1807,
+      "eval_runtime": 80.7729,
+      "eval_samples_per_second": 5.571,
+      "eval_steps_per_second": 0.706,
       "step": 5778
     },
     {
       "epoch": 10.0,
-      "eval_gen_len": 16.6044,
-      "eval_loss": 2.5771408081054688,
-      "eval_rouge-1": 57.2781,
-      "eval_rouge-2": 44.4179,
-      "eval_rouge-l": 56.9181,
-      "eval_runtime": 63.0905,
-      "eval_samples_per_second": 7.133,
-      "eval_steps_per_second": 0.903,
+      "eval_gen_len": 18.5022,
+      "eval_loss": 2.987593650817871,
+      "eval_rouge-1": 34.5358,
+      "eval_rouge-2": 18.8882,
+      "eval_rouge-l": 33.5787,
+      "eval_runtime": 70.1569,
+      "eval_samples_per_second": 6.414,
+      "eval_steps_per_second": 0.812,
       "step": 6420
     },
     {
       "epoch": 11.0,
-      "eval_gen_len": 16.6333,
-      "eval_loss": 2.570962905883789,
-      "eval_rouge-1": 59.0229,
-      "eval_rouge-2": 46.6584,
-      "eval_rouge-l": 58.5764,
-      "eval_runtime": 63.325,
-      "eval_samples_per_second": 7.106,
-      "eval_steps_per_second": 0.9,
+      "eval_gen_len": 18.0133,
+      "eval_loss": 2.951963424682617,
+      "eval_rouge-1": 36.0373,
+      "eval_rouge-2": 20.1364,
+      "eval_rouge-l": 35.0084,
+      "eval_runtime": 67.9106,
+      "eval_samples_per_second": 6.626,
+      "eval_steps_per_second": 0.839,
       "step": 7062
     },
     {
       "epoch": 12.0,
-      "eval_gen_len": 16.6,
-      "eval_loss": 2.552370309829712,
-      "eval_rouge-1": 59.7409,
-      "eval_rouge-2": 48.0446,
-      "eval_rouge-l": 59.4073,
-      "eval_runtime": 63.9445,
-      "eval_samples_per_second": 7.037,
-      "eval_steps_per_second": 0.891,
+      "eval_gen_len": 17.6444,
+      "eval_loss": 2.919402599334717,
+      "eval_rouge-1": 36.9999,
+      "eval_rouge-2": 22.0519,
+      "eval_rouge-l": 36.0277,
+      "eval_runtime": 67.2477,
+      "eval_samples_per_second": 6.692,
+      "eval_steps_per_second": 0.848,
       "step": 7704
     },
     {
       "epoch": 13.0,
-      "eval_gen_len": 17.0356,
-      "eval_loss": 2.5687365531921387,
-      "eval_rouge-1": 61.6703,
-      "eval_rouge-2": 49.8271,
-      "eval_rouge-l": 61.2833,
-      "eval_runtime": 65.4144,
-      "eval_samples_per_second": 6.879,
-      "eval_steps_per_second": 0.871,
+      "eval_gen_len": 18.0467,
+      "eval_loss": 2.8770110607147217,
+      "eval_rouge-1": 39.3542,
+      "eval_rouge-2": 23.7167,
+      "eval_rouge-l": 38.2318,
+      "eval_runtime": 68.3111,
+      "eval_samples_per_second": 6.588,
+      "eval_steps_per_second": 0.834,
       "step": 8346
     },
     {
-      "epoch": 14.0,
-      "eval_gen_len": 16.5644,
-      "eval_loss": 2.5709228515625,
-      "eval_rouge-1": 63.2241,
-      "eval_rouge-2": 52.278,
-      "eval_rouge-l": 62.7525,
-      "eval_runtime": 62.9575,
-      "eval_samples_per_second": 7.148,
-      "eval_steps_per_second": 0.905,
-      "step": 8988
-    },
-    {
-      "epoch": 15.0,
-      "eval_gen_len": 17.0111,
-      "eval_loss": 2.5851821899414062,
-      "eval_rouge-1": 63.9802,
-      "eval_rouge-2": 53.6175,
-      "eval_rouge-l": 63.571,
-      "eval_runtime": 65.0883,
-      "eval_samples_per_second": 6.914,
-      "eval_steps_per_second": 0.876,
-      "step": 9630
-    },
-    {
-      "epoch": 15.58,
-      "learning_rate": 7.109003215434083e-05,
-      "loss": 1.9492,
-      "step": 10000
-    },
-    {
-      "epoch": 16.0,
-      "eval_gen_len": 17.2444,
-      "eval_loss": 2.5511105060577393,
-      "eval_rouge-1": 65.3788,
-      "eval_rouge-2": 55.3744,
-      "eval_rouge-l": 65.0206,
-      "eval_runtime": 66.1393,
-      "eval_samples_per_second": 6.804,
-      "eval_steps_per_second": 0.862,
-      "step": 10272
-    },
-    {
-      "epoch": 17.0,
-      "eval_gen_len": 16.4956,
-      "eval_loss": 2.56447434425354,
-      "eval_rouge-1": 65.8379,
-      "eval_rouge-2": 56.0679,
-      "eval_rouge-l": 65.3959,
-      "eval_runtime": 62.5339,
-      "eval_samples_per_second": 7.196,
-      "eval_steps_per_second": 0.912,
-      "step": 10914
-    },
-    {
-      "epoch": 18.0,
-      "eval_gen_len": 16.5267,
-      "eval_loss": 2.5657262802124023,
-      "eval_rouge-1": 65.5539,
-      "eval_rouge-2": 55.8889,
-      "eval_rouge-l": 65.1224,
-      "eval_runtime": 62.5783,
-      "eval_samples_per_second": 7.191,
-      "eval_steps_per_second": 0.911,
-      "step": 11556
-    },
-    {
-      "epoch": 19.0,
-      "eval_gen_len": 17.1111,
-      "eval_loss": 2.5901031494140625,
-      "eval_rouge-1": 66.8166,
-      "eval_rouge-2": 57.2767,
-      "eval_rouge-l": 66.3378,
-      "eval_runtime": 64.711,
-      "eval_samples_per_second": 6.954,
-      "eval_steps_per_second": 0.881,
-      "step": 12198
-    },
-    {
-      "epoch": 20.0,
-      "eval_gen_len": 16.6711,
-      "eval_loss": 2.5617997646331787,
-      "eval_rouge-1": 66.652,
-      "eval_rouge-2": 57.0521,
-      "eval_rouge-l": 66.27,
-      "eval_runtime": 63.1464,
-      "eval_samples_per_second": 7.126,
-      "eval_steps_per_second": 0.903,
-      "step": 12840
-    },
-    {
-      "epoch": 21.0,
-      "eval_gen_len": 17.1711,
-      "eval_loss": 2.5742971897125244,
-      "eval_rouge-1": 67.1727,
-      "eval_rouge-2": 57.2985,
-      "eval_rouge-l": 66.5791,
-      "eval_runtime": 65.0484,
-      "eval_samples_per_second": 6.918,
-      "eval_steps_per_second": 0.876,
-      "step": 13482
-    },
-    {
-      "epoch": 21.0,
-      "step": 13482,
-      "total_flos": 2.9462758234914816e+16,
-      "train_loss": 2.2238484863498553,
-      "train_runtime": 6515.1509,
-      "train_samples_per_second": 39.393,
-      "train_steps_per_second": 4.927
+      "epoch": 13.0,
+      "step": 8346,
+      "total_flos": 1.840140038622413e+16,
+      "train_loss": 3.136590390419063,
+      "train_runtime": 4227.0028,
+      "train_samples_per_second": 60.717,
+      "train_steps_per_second": 7.594
     }
   ],
   "logging_steps": 5000,
   "max_steps": 32100,
   "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 2.9462758234914816e+16,
+  "total_flos": 1.840140038622413e+16,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2b7dc2513d2cd9ce4df66296643c3d78258835df0a1ca563307f80e32185948a
+oid sha256:6279d4630ce072544b087cf592fd0ef6232685fbe7ffdef5416d022bb17bc941
 size 4792