diff --git "a/checkpoint-17500/trainer_state.json" "b/checkpoint-17500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-17500/trainer_state.json"
@@ -0,0 +1,5033 @@
+{
+  "best_metric": 1.9262617826461792,
+  "best_model_checkpoint": "Paraphrase-v3/checkpoint-17500",
+  "epoch": 7.0,
+  "eval_steps": 500,
+  "global_step": 17500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.671796798706055,
+      "learning_rate": 3.2e-07,
+      "loss": 3.3655,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.20463752746582,
+      "learning_rate": 6.533333333333334e-07,
+      "loss": 3.2816,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.489435195922852,
+      "learning_rate": 9.733333333333335e-07,
+      "loss": 3.2049,
+      "step": 75
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.335665702819824,
+      "learning_rate": 1.2933333333333334e-06,
+      "loss": 3.2663,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.564672470092773,
+      "learning_rate": 1.6133333333333333e-06,
+      "loss": 3.0361,
+      "step": 125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.445804595947266,
+      "learning_rate": 1.946666666666667e-06,
+      "loss": 3.0074,
+      "step": 150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.889507293701172,
+      "learning_rate": 2.28e-06,
+      "loss": 2.9401,
+      "step": 175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.881171226501465,
+      "learning_rate": 2.6133333333333334e-06,
+      "loss": 2.7827,
+      "step": 200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8032336235046387,
+      "learning_rate": 2.9466666666666667e-06,
+      "loss": 2.7671,
+      "step": 225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.653895616531372,
+      "learning_rate": 3.2800000000000004e-06,
+      "loss": 2.6844,
+      "step": 250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.168716907501221,
+      "learning_rate": 3.613333333333334e-06,
+      "loss": 2.5707,
+      "step": 275
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.516740322113037,
+      "learning_rate": 3.9466666666666664e-06,
+      "loss": 2.5428,
+      "step": 300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4983417987823486,
+      "learning_rate": 4.28e-06,
+      "loss": 2.57,
+      "step": 325
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.452354669570923,
+      "learning_rate": 4.613333333333334e-06,
+      "loss": 2.5167,
+      "step": 350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4193527698516846,
+      "learning_rate": 4.946666666666667e-06,
+      "loss": 2.4788,
+      "step": 375
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7596585750579834,
+      "learning_rate": 5.28e-06,
+      "loss": 2.4994,
+      "step": 400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1304233074188232,
+      "learning_rate": 5.6133333333333335e-06,
+      "loss": 2.4324,
+      "step": 425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9708311557769775,
+      "learning_rate": 5.946666666666667e-06,
+      "loss": 2.4117,
+      "step": 450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2388558387756348,
+      "learning_rate": 6.28e-06,
+      "loss": 2.3727,
+      "step": 475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3298180103302,
+      "learning_rate": 6.613333333333334e-06,
+      "loss": 2.4732,
+      "step": 500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0654354095458984,
+      "learning_rate": 6.9466666666666665e-06,
+      "loss": 2.4922,
+      "step": 525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9771661758422852,
+      "learning_rate": 7.280000000000001e-06,
+      "loss": 2.3858,
+      "step": 550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3070411682128906,
+      "learning_rate": 7.613333333333334e-06,
+      "loss": 2.3785,
+      "step": 575
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0033750534057617,
+      "learning_rate": 7.946666666666668e-06,
+      "loss": 2.369,
+      "step": 600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.960777997970581,
+      "learning_rate": 8.28e-06,
+      "loss": 2.4246,
+      "step": 625
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8081718683242798,
+      "learning_rate": 8.613333333333334e-06,
+      "loss": 2.3665,
+      "step": 650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8548564910888672,
+      "learning_rate": 8.946666666666667e-06,
+      "loss": 2.3306,
+      "step": 675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.222216844558716,
+      "learning_rate": 9.28e-06,
+      "loss": 2.3112,
+      "step": 700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7288415431976318,
+      "learning_rate": 9.613333333333333e-06,
+      "loss": 2.3098,
+      "step": 725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0263378620147705,
+      "learning_rate": 9.946666666666667e-06,
+      "loss": 2.3304,
+      "step": 750
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5405957698822021,
+      "learning_rate": 1.0280000000000002e-05,
+      "loss": 2.3302,
+      "step": 775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8870110511779785,
+      "learning_rate": 1.0613333333333334e-05,
+      "loss": 2.3241,
+      "step": 800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8393750190734863,
+      "learning_rate": 1.0946666666666668e-05,
+      "loss": 2.3361,
+      "step": 825
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.870672583580017,
+      "learning_rate": 1.128e-05,
+      "loss": 2.3178,
+      "step": 850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7924401760101318,
+      "learning_rate": 1.1613333333333335e-05,
+      "loss": 2.2558,
+      "step": 875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6780457496643066,
+      "learning_rate": 1.1946666666666667e-05,
+      "loss": 2.2928,
+      "step": 900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8860801458358765,
+      "learning_rate": 1.2280000000000001e-05,
+      "loss": 2.2523,
+      "step": 925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7925626039505005,
+      "learning_rate": 1.2613333333333332e-05,
+      "loss": 2.2885,
+      "step": 950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.916331171989441,
+      "learning_rate": 1.2946666666666668e-05,
+      "loss": 2.3369,
+      "step": 975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6720538139343262,
+      "learning_rate": 1.3280000000000002e-05,
+      "loss": 2.2789,
+      "step": 1000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8012834787368774,
+      "learning_rate": 1.3613333333333334e-05,
+      "loss": 2.2818,
+      "step": 1025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.82932710647583,
+      "learning_rate": 1.3946666666666666e-05,
+      "loss": 2.3173,
+      "step": 1050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8059712648391724,
+      "learning_rate": 1.4280000000000002e-05,
+      "loss": 2.266,
+      "step": 1075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4668495655059814,
+      "learning_rate": 1.4613333333333335e-05,
+      "loss": 2.2639,
+      "step": 1100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7411282062530518,
+      "learning_rate": 1.4946666666666667e-05,
+      "loss": 2.2816,
+      "step": 1125
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5545501708984375,
+      "learning_rate": 1.528e-05,
+      "loss": 2.2578,
+      "step": 1150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6876237392425537,
+      "learning_rate": 1.5613333333333335e-05,
+      "loss": 2.3282,
+      "step": 1175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6718631982803345,
+      "learning_rate": 1.5946666666666668e-05,
+      "loss": 2.2713,
+      "step": 1200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.628440499305725,
+      "learning_rate": 1.628e-05,
+      "loss": 2.2581,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8468382358551025,
+      "learning_rate": 1.6613333333333332e-05,
+      "loss": 2.3533,
+      "step": 1250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8317992687225342,
+      "learning_rate": 1.6946666666666665e-05,
+      "loss": 2.2398,
+      "step": 1275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.673600673675537,
+      "learning_rate": 1.728e-05,
+      "loss": 2.2879,
+      "step": 1300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5316749811172485,
+      "learning_rate": 1.7613333333333333e-05,
+      "loss": 2.2502,
+      "step": 1325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6259863376617432,
+      "learning_rate": 1.794666666666667e-05,
+      "loss": 2.2399,
+      "step": 1350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4696708917617798,
+      "learning_rate": 1.828e-05,
+      "loss": 2.233,
+      "step": 1375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9751217365264893,
+      "learning_rate": 1.8613333333333337e-05,
+      "loss": 2.2855,
+      "step": 1400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.541063666343689,
+      "learning_rate": 1.894666666666667e-05,
+      "loss": 2.2782,
+      "step": 1425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.525913953781128,
+      "learning_rate": 1.9280000000000002e-05,
+      "loss": 2.238,
+      "step": 1450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6467074155807495,
+      "learning_rate": 1.9613333333333334e-05,
+      "loss": 2.2921,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6107714176177979,
+      "learning_rate": 1.9946666666666667e-05,
+      "loss": 2.2359,
+      "step": 1500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6940182447433472,
+      "learning_rate": 2.0280000000000002e-05,
+      "loss": 2.2137,
+      "step": 1525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7330864667892456,
+      "learning_rate": 2.0613333333333335e-05,
+      "loss": 2.211,
+      "step": 1550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.847303032875061,
+      "learning_rate": 2.0946666666666667e-05,
+      "loss": 2.2547,
+      "step": 1575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3495903015136719,
+      "learning_rate": 2.128e-05,
+      "loss": 2.217,
+      "step": 1600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4377658367156982,
+      "learning_rate": 2.1613333333333335e-05,
+      "loss": 2.2179,
+      "step": 1625
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9955512285232544,
+      "learning_rate": 2.1946666666666668e-05,
+      "loss": 2.2818,
+      "step": 1650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3620418310165405,
+      "learning_rate": 2.228e-05,
+      "loss": 2.2304,
+      "step": 1675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4833818674087524,
+      "learning_rate": 2.2613333333333333e-05,
+      "loss": 2.2016,
+      "step": 1700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4675543308258057,
+      "learning_rate": 2.294666666666667e-05,
+      "loss": 2.1752,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5722428560256958,
+      "learning_rate": 2.328e-05,
+      "loss": 2.2749,
+      "step": 1750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.711739182472229,
+      "learning_rate": 2.3613333333333333e-05,
+      "loss": 2.1855,
+      "step": 1775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5259678363800049,
+      "learning_rate": 2.394666666666667e-05,
+      "loss": 2.1757,
+      "step": 1800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4229743480682373,
+      "learning_rate": 2.428e-05,
+      "loss": 2.2109,
+      "step": 1825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5702102184295654,
+      "learning_rate": 2.4613333333333337e-05,
+      "loss": 2.1032,
+      "step": 1850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7006195783615112,
+      "learning_rate": 2.494666666666667e-05,
+      "loss": 2.1867,
+      "step": 1875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5293747186660767,
+      "learning_rate": 2.5280000000000005e-05,
+      "loss": 2.2308,
+      "step": 1900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1540281772613525,
+      "learning_rate": 2.5613333333333334e-05,
+      "loss": 2.184,
+      "step": 1925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6256271600723267,
+      "learning_rate": 2.594666666666667e-05,
+      "loss": 2.2499,
+      "step": 1950
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.394745111465454,
+      "learning_rate": 2.628e-05,
+      "loss": 2.1345,
+      "step": 1975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5799375772476196,
+      "learning_rate": 2.6613333333333335e-05,
+      "loss": 2.1641,
+      "step": 2000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4012045860290527,
+      "learning_rate": 2.694666666666667e-05,
+      "loss": 2.169,
+      "step": 2025
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5689955949783325,
+      "learning_rate": 2.728e-05,
+      "loss": 2.1793,
+      "step": 2050
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5047041177749634,
+      "learning_rate": 2.7613333333333335e-05,
+      "loss": 2.1984,
+      "step": 2075
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.399196743965149,
+      "learning_rate": 2.7946666666666664e-05,
+      "loss": 2.1996,
+      "step": 2100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7386200428009033,
+      "learning_rate": 2.828e-05,
+      "loss": 2.1799,
+      "step": 2125
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6265965700149536,
+      "learning_rate": 2.8613333333333336e-05,
+      "loss": 2.1946,
+      "step": 2150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3947643041610718,
+      "learning_rate": 2.8946666666666665e-05,
+      "loss": 2.206,
+      "step": 2175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4571387767791748,
+      "learning_rate": 2.928e-05,
+      "loss": 2.1895,
+      "step": 2200
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3659110069274902,
+      "learning_rate": 2.9613333333333337e-05,
+      "loss": 2.1662,
+      "step": 2225
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6103181838989258,
+      "learning_rate": 2.9946666666666666e-05,
+      "loss": 2.2149,
+      "step": 2250
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4823683500289917,
+      "learning_rate": 3.028e-05,
+      "loss": 2.2413,
+      "step": 2275
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4936267137527466,
+      "learning_rate": 3.0613333333333334e-05,
+      "loss": 2.1555,
+      "step": 2300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.487349271774292,
+      "learning_rate": 3.0946666666666666e-05,
+      "loss": 2.194,
+      "step": 2325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.626238226890564,
+      "learning_rate": 3.1280000000000005e-05,
+      "loss": 2.1924,
+      "step": 2350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4786629676818848,
+      "learning_rate": 3.161333333333333e-05,
+      "loss": 2.1831,
+      "step": 2375
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.430049419403076,
+      "learning_rate": 3.194666666666667e-05,
+      "loss": 2.1984,
+      "step": 2400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4396685361862183,
+      "learning_rate": 3.2279999999999996e-05,
+      "loss": 2.1214,
+      "step": 2425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5446324348449707,
+      "learning_rate": 3.2613333333333335e-05,
+      "loss": 2.1869,
+      "step": 2450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.591899037361145,
+      "learning_rate": 3.294666666666667e-05,
+      "loss": 2.166,
+      "step": 2475
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7222836017608643,
+      "learning_rate": 3.328e-05,
+      "loss": 2.1121,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0,
+      "eval_gen_len": 13.0372,
+      "eval_loss": 2.038074016571045,
+      "eval_rouge1": 50.5825,
+      "eval_rouge2": 24.8269,
+      "eval_rougeL": 46.4789,
+      "eval_rougeLsum": 46.4745,
+      "eval_runtime": 194.3582,
+      "eval_samples_per_second": 102.903,
+      "eval_steps_per_second": 1.61,
+      "step": 2500
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6080988645553589,
+      "learning_rate": 3.361333333333333e-05,
+      "loss": 2.1864,
+      "step": 2525
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.4214496612548828,
+      "learning_rate": 3.394666666666667e-05,
+      "loss": 2.1115,
+      "step": 2550
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.4816479682922363,
+      "learning_rate": 3.4280000000000004e-05,
+      "loss": 2.1319,
+      "step": 2575
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5511168241500854,
+      "learning_rate": 3.4613333333333336e-05,
+      "loss": 2.1639,
+      "step": 2600
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.5296896696090698,
+      "learning_rate": 3.494666666666667e-05,
+      "loss": 2.1616,
+      "step": 2625
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.540246605873108,
+      "learning_rate": 3.528e-05,
+      "loss": 2.1357,
+      "step": 2650
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5819247961044312,
+      "learning_rate": 3.561333333333334e-05,
+      "loss": 2.1471,
+      "step": 2675
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.4258027076721191,
+      "learning_rate": 3.5946666666666666e-05,
+      "loss": 2.113,
+      "step": 2700
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.356101393699646,
+      "learning_rate": 3.6280000000000005e-05,
+      "loss": 2.1027,
+      "step": 2725
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.754785418510437,
+      "learning_rate": 3.661333333333333e-05,
+      "loss": 2.1101,
+      "step": 2750
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.111976146697998,
+      "learning_rate": 3.6933333333333334e-05,
+      "loss": 2.1733,
+      "step": 2775
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.40675687789917,
+      "learning_rate": 3.726666666666667e-05,
+      "loss": 2.1599,
+      "step": 2800
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.312162160873413,
+      "learning_rate": 3.76e-05,
+      "loss": 2.092,
+      "step": 2825
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 1.3412392139434814,
+      "learning_rate": 3.793333333333334e-05,
+      "loss": 2.1503,
+      "step": 2850
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.3180032968521118,
+      "learning_rate": 3.8266666666666664e-05,
+      "loss": 2.1323,
+      "step": 2875
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.4078686237335205,
+      "learning_rate": 3.86e-05,
+      "loss": 2.1663,
+      "step": 2900
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.4193419218063354,
+      "learning_rate": 3.8933333333333336e-05,
+      "loss": 2.1407,
+      "step": 2925
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.5237839221954346,
+      "learning_rate": 3.926666666666667e-05,
+      "loss": 2.1884,
+      "step": 2950
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.3546311855316162,
+      "learning_rate": 3.960000000000001e-05,
+      "loss": 2.0769,
+      "step": 2975
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.1709140539169312,
+      "learning_rate": 3.993333333333333e-05,
+      "loss": 1.9888,
+      "step": 3000
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.3492522239685059,
+      "learning_rate": 4.026666666666667e-05,
+      "loss": 2.0922,
+      "step": 3025
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.5517680644989014,
+      "learning_rate": 4.0600000000000004e-05,
+      "loss": 2.1161,
+      "step": 3050
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.3444249629974365,
+      "learning_rate": 4.093333333333334e-05,
+      "loss": 2.1351,
+      "step": 3075
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.4553472995758057,
+      "learning_rate": 4.126666666666667e-05,
+      "loss": 2.1889,
+      "step": 3100
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4636095762252808,
+      "learning_rate": 4.16e-05,
+      "loss": 2.1161,
+      "step": 3125
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.3863240480422974,
+      "learning_rate": 4.1933333333333334e-05,
+      "loss": 2.0402,
+      "step": 3150
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.2597577571868896,
+      "learning_rate": 4.226666666666667e-05,
+      "loss": 2.0852,
+      "step": 3175
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.3540936708450317,
+      "learning_rate": 4.26e-05,
+      "loss": 2.1439,
+      "step": 3200
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.4448434114456177,
+      "learning_rate": 4.293333333333334e-05,
+      "loss": 2.1271,
+      "step": 3225
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.352243185043335,
+      "learning_rate": 4.3266666666666664e-05,
+      "loss": 2.1263,
+      "step": 3250
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.5224635601043701,
+      "learning_rate": 4.36e-05,
+      "loss": 2.1373,
+      "step": 3275
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.51206636428833,
+      "learning_rate": 4.3933333333333335e-05,
+      "loss": 2.1521,
+      "step": 3300
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.4777268171310425,
+      "learning_rate": 4.426666666666667e-05,
+      "loss": 2.0391,
+      "step": 3325
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.278085708618164,
+      "learning_rate": 4.46e-05,
+      "loss": 2.1313,
+      "step": 3350
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.4700874090194702,
+      "learning_rate": 4.493333333333333e-05,
+      "loss": 2.0469,
+      "step": 3375
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 1.6795507669448853,
+      "learning_rate": 4.526666666666667e-05,
+      "loss": 2.1422,
+      "step": 3400
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.496156096458435,
+      "learning_rate": 4.5600000000000004e-05,
+      "loss": 2.0771,
+      "step": 3425
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.4133617877960205,
+      "learning_rate": 4.5933333333333336e-05,
+      "loss": 2.0397,
+      "step": 3450
+    },
+    {
+      "epoch": 1.3900000000000001,
+      "grad_norm": 1.3693889379501343,
+      "learning_rate": 4.626666666666667e-05,
+      "loss": 2.1713,
+      "step": 3475
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.5390167236328125,
+      "learning_rate": 4.660000000000001e-05,
+      "loss": 2.1348,
+      "step": 3500
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7128788232803345,
+      "learning_rate": 4.6933333333333333e-05,
+      "loss": 2.0817,
+      "step": 3525
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.4634878635406494,
+      "learning_rate": 4.726666666666667e-05,
+      "loss": 2.1121,
+      "step": 3550
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.623915672302246,
+      "learning_rate": 4.76e-05,
+      "loss": 2.1968,
+      "step": 3575
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.4591940641403198,
+      "learning_rate": 4.793333333333334e-05,
+      "loss": 2.0789,
+      "step": 3600
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.2940136194229126,
+      "learning_rate": 4.826666666666667e-05,
+      "loss": 2.147,
+      "step": 3625
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.283172845840454,
+      "learning_rate": 4.86e-05,
+      "loss": 2.0765,
+      "step": 3650
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.3233940601348877,
+      "learning_rate": 4.8933333333333335e-05,
+      "loss": 2.1018,
+      "step": 3675
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.6195733547210693,
+      "learning_rate": 4.926666666666667e-05,
+      "loss": 2.1608,
+      "step": 3700
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.5443871021270752,
+      "learning_rate": 4.96e-05,
+      "loss": 2.0899,
+      "step": 3725
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.3754618167877197,
+      "learning_rate": 4.993333333333334e-05,
+      "loss": 2.1288,
+      "step": 3750
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.2181512117385864,
+      "learning_rate": 4.997037037037037e-05,
+      "loss": 2.1582,
+      "step": 3775
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.4862068891525269,
+      "learning_rate": 4.993333333333334e-05,
+      "loss": 2.1475,
+      "step": 3800
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.3613662719726562,
+      "learning_rate": 4.9896296296296293e-05,
+      "loss": 2.1625,
+      "step": 3825
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.3887194395065308,
+      "learning_rate": 4.985925925925926e-05,
+      "loss": 2.1284,
+      "step": 3850
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.468736171722412,
+      "learning_rate": 4.982222222222222e-05,
+      "loss": 2.0792,
+      "step": 3875
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.4086352586746216,
+      "learning_rate": 4.978518518518519e-05,
+      "loss": 2.0997,
+      "step": 3900
+    },
+    {
+      "epoch": 1.5699999999999998,
+      "grad_norm": 1.3831346035003662,
+      "learning_rate": 4.9748148148148146e-05,
+      "loss": 2.0768,
+      "step": 3925
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.2189925909042358,
+      "learning_rate": 4.9711111111111115e-05,
+      "loss": 2.0809,
+      "step": 3950
+    },
+    {
+      "epoch": 1.5899999999999999,
+      "grad_norm": 1.2166587114334106,
+      "learning_rate": 4.9674074074074076e-05,
+      "loss": 2.128,
+      "step": 3975
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.5292308330535889,
+      "learning_rate": 4.963703703703704e-05,
+      "loss": 2.1093,
+      "step": 4000
+    },
+    {
+      "epoch": 1.6099999999999999,
+      "grad_norm": 1.492801308631897,
+      "learning_rate": 4.96e-05,
+      "loss": 2.116,
+      "step": 4025
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.4455024003982544,
+      "learning_rate": 4.956296296296297e-05,
+      "loss": 2.1292,
+      "step": 4050
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.3066167831420898,
+      "learning_rate": 4.952592592592592e-05,
+      "loss": 2.1011,
+      "step": 4075
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 1.496256947517395,
+      "learning_rate": 4.949037037037037e-05,
+      "loss": 2.1056,
+      "step": 4100
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.2948390245437622,
+      "learning_rate": 4.9453333333333336e-05,
+      "loss": 2.0569,
+      "step": 4125
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 1.4984357357025146,
+      "learning_rate": 4.94162962962963e-05,
+      "loss": 2.1433,
+      "step": 4150
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.7721185684204102,
+      "learning_rate": 4.9379259259259266e-05,
+      "loss": 2.1098,
+      "step": 4175
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 1.3611931800842285,
+      "learning_rate": 4.934222222222222e-05,
+      "loss": 2.0652,
+      "step": 4200
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.3092522621154785,
+      "learning_rate": 4.930518518518519e-05,
+      "loss": 2.0888,
+      "step": 4225
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.3880172967910767,
+      "learning_rate": 4.926814814814815e-05,
+      "loss": 2.0751,
+      "step": 4250
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.361522912979126,
+      "learning_rate": 4.923111111111111e-05,
+      "loss": 2.1263,
+      "step": 4275
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.357974886894226,
+      "learning_rate": 4.9194074074074074e-05,
+      "loss": 2.0967,
+      "step": 4300
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.3499886989593506,
+      "learning_rate": 4.915703703703704e-05,
+      "loss": 2.0426,
+      "step": 4325
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.4301838874816895,
+      "learning_rate": 4.9120000000000004e-05,
+      "loss": 2.0596,
+      "step": 4350
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.4266048669815063,
+      "learning_rate": 4.9082962962962966e-05,
+      "loss": 2.0574,
+      "step": 4375
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.4211664199829102,
+      "learning_rate": 4.904592592592593e-05,
+      "loss": 2.0839,
+      "step": 4400
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.2608351707458496,
+      "learning_rate": 4.9008888888888896e-05,
+      "loss": 2.037,
+      "step": 4425
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.35879385471344,
+      "learning_rate": 4.897185185185185e-05,
+      "loss": 2.0807,
+      "step": 4450
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4060136079788208,
+      "learning_rate": 4.893481481481482e-05,
+      "loss": 2.1474,
+      "step": 4475
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.3391823768615723,
+      "learning_rate": 4.889777777777778e-05,
+      "loss": 2.115,
+      "step": 4500
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6455671787261963,
+      "learning_rate": 4.886074074074075e-05,
+      "loss": 2.0943,
+      "step": 4525
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 1.3512789011001587,
+      "learning_rate": 4.88237037037037e-05,
+      "loss": 2.0675,
+      "step": 4550
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.3271268606185913,
+      "learning_rate": 4.878666666666667e-05,
+      "loss": 2.139,
+      "step": 4575
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 1.3180307149887085,
+      "learning_rate": 4.874962962962963e-05,
+      "loss": 2.0442,
+      "step": 4600
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.322908878326416,
+      "learning_rate": 4.8712592592592595e-05,
+      "loss": 2.1164,
+      "step": 4625
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 1.2810012102127075,
+      "learning_rate": 4.8675555555555556e-05,
+      "loss": 2.0138,
+      "step": 4650
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.3733242750167847,
+      "learning_rate": 4.8638518518518525e-05,
+      "loss": 2.0639,
+      "step": 4675
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.3254153728485107,
+      "learning_rate": 4.860148148148148e-05,
+      "loss": 2.0502,
+      "step": 4700
+    },
+    {
+      "epoch": 1.8900000000000001,
+      "grad_norm": 1.3326213359832764,
+      "learning_rate": 4.856444444444445e-05,
+      "loss": 2.0394,
+      "step": 4725
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.3088740110397339,
+      "learning_rate": 4.852740740740741e-05,
+      "loss": 2.1614,
+      "step": 4750
+    },
+    {
+      "epoch": 1.9100000000000001,
+      "grad_norm": 1.4830114841461182,
+      "learning_rate": 4.849037037037038e-05,
+      "loss": 2.0836,
+      "step": 4775
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.5083235502243042,
+      "learning_rate": 4.845333333333333e-05,
+      "loss": 2.084,
+      "step": 4800
+    },
+    {
+      "epoch": 1.9300000000000002,
+      "grad_norm": 1.4989155530929565,
+      "learning_rate": 4.84162962962963e-05,
+      "loss": 2.1183,
+      "step": 4825
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.2256449460983276,
+      "learning_rate": 4.837925925925926e-05,
+      "loss": 2.1002,
+      "step": 4850
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2824608087539673,
+      "learning_rate": 4.8342222222222224e-05,
+      "loss": 2.0369,
+      "step": 4875
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.2830954790115356,
+      "learning_rate": 4.8305185185185185e-05,
+      "loss": 2.0361,
+      "step": 4900
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.4407249689102173,
+      "learning_rate": 4.8268148148148154e-05,
+      "loss": 2.0535,
+      "step": 4925
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.3975064754486084,
+      "learning_rate": 4.8231111111111115e-05,
+      "loss": 2.0805,
+      "step": 4950
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.2193372249603271,
+      "learning_rate": 4.819407407407408e-05,
+      "loss": 2.0748,
+      "step": 4975
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4137203693389893,
+      "learning_rate": 4.815703703703704e-05,
+      "loss": 2.0812,
+      "step": 5000
+    },
+    {
+      "epoch": 2.0,
+      "eval_gen_len": 13.1104,
+      "eval_loss": 1.972951054573059,
+      "eval_rouge1": 51.236,
+      "eval_rouge2": 25.4764,
+      "eval_rougeL": 47.1309,
+      "eval_rougeLsum": 47.1388,
+      "eval_runtime": 195.1159,
+      "eval_samples_per_second": 102.503,
+      "eval_steps_per_second": 1.604,
+      "step": 5000
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.292515754699707,
+      "learning_rate": 4.812000000000001e-05,
+      "loss": 1.9586,
+      "step": 5025
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.329329490661621,
+      "learning_rate": 4.808296296296296e-05,
+      "loss": 2.0068,
+      "step": 5050
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.5336360931396484,
+      "learning_rate": 4.804592592592593e-05,
+      "loss": 1.986,
+      "step": 5075
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.313008427619934,
+      "learning_rate": 4.800888888888889e-05,
+      "loss": 2.0093,
+      "step": 5100
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.390008807182312,
+      "learning_rate": 4.797185185185186e-05,
+      "loss": 2.0478,
+      "step": 5125
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.4343863725662231,
+      "learning_rate": 4.7934814814814815e-05,
+      "loss": 2.0325,
+      "step": 5150
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.3213175535202026,
+      "learning_rate": 4.789777777777778e-05,
+      "loss": 2.0362,
+      "step": 5175
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.2570987939834595,
+      "learning_rate": 4.7860740740740745e-05,
+      "loss": 1.9476,
+      "step": 5200
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.377868413925171,
+      "learning_rate": 4.7823703703703706e-05,
+      "loss": 1.9655,
+      "step": 5225
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.1855210065841675,
+      "learning_rate": 4.778666666666667e-05,
+      "loss": 2.0382,
+      "step": 5250
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.3284786939620972,
+      "learning_rate": 4.7749629629629636e-05,
+      "loss": 2.0808,
+      "step": 5275
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.4279937744140625,
+      "learning_rate": 4.77125925925926e-05,
+      "loss": 2.0179,
+      "step": 5300
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5963534116744995,
+      "learning_rate": 4.767555555555556e-05,
+      "loss": 1.9902,
+      "step": 5325
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.3769519329071045,
+      "learning_rate": 4.763851851851852e-05,
+      "loss": 1.9842,
+      "step": 5350
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.3761203289031982,
+      "learning_rate": 4.760148148148149e-05,
+      "loss": 2.0256,
+      "step": 5375
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4661396741867065,
+      "learning_rate": 4.7564444444444444e-05,
+      "loss": 2.0704,
+      "step": 5400
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.301453709602356,
+      "learning_rate": 4.752740740740741e-05,
+      "loss": 1.9957,
+      "step": 5425
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.3096243143081665,
+      "learning_rate": 4.7490370370370374e-05,
+      "loss": 2.0023,
+      "step": 5450
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.3345898389816284,
+      "learning_rate": 4.7453333333333335e-05,
+      "loss": 2.0472,
+      "step": 5475
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.1880935430526733,
+      "learning_rate": 4.74162962962963e-05,
+      "loss": 1.9565,
+      "step": 5500
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.2483596801757812,
+      "learning_rate": 4.7379259259259265e-05,
+      "loss": 2.0419,
+      "step": 5525
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.3003907203674316,
+      "learning_rate": 4.734222222222223e-05,
+      "loss": 2.0411,
+      "step": 5550
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.3677852153778076,
+      "learning_rate": 4.730518518518519e-05,
+      "loss": 2.0454,
+      "step": 5575
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.3494675159454346,
+      "learning_rate": 4.726814814814815e-05,
+      "loss": 1.9652,
+      "step": 5600
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.2623995542526245,
+      "learning_rate": 4.723111111111112e-05,
+      "loss": 2.0516,
+      "step": 5625
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.4136385917663574,
+      "learning_rate": 4.719407407407407e-05,
+      "loss": 1.9735,
+      "step": 5650
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.2677298784255981,
+      "learning_rate": 4.715703703703704e-05,
+      "loss": 1.9916,
+      "step": 5675
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 1.422560453414917,
+      "learning_rate": 4.712e-05,
+      "loss": 2.0216,
+      "step": 5700
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4796884059906006,
+      "learning_rate": 4.7082962962962964e-05,
+      "loss": 2.0134,
+      "step": 5725
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.347684383392334,
+      "learning_rate": 4.7045925925925926e-05,
+      "loss": 2.0073,
+      "step": 5750
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.238133192062378,
+      "learning_rate": 4.7008888888888894e-05,
+      "loss": 2.0002,
+      "step": 5775
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.271411657333374,
+      "learning_rate": 4.6971851851851856e-05,
+      "loss": 1.9866,
+      "step": 5800
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.4188841581344604,
+      "learning_rate": 4.693481481481482e-05,
+      "loss": 1.96,
+      "step": 5825
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.3498203754425049,
+      "learning_rate": 4.689777777777778e-05,
+      "loss": 2.0338,
+      "step": 5850
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.224448800086975,
+      "learning_rate": 4.686074074074074e-05,
+      "loss": 2.0572,
+      "step": 5875
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.4224382638931274,
+      "learning_rate": 4.682370370370371e-05,
+      "loss": 2.0333,
+      "step": 5900
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.553787350654602,
+      "learning_rate": 4.678666666666667e-05,
+      "loss": 1.9975,
+      "step": 5925
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.377954125404358,
+      "learning_rate": 4.674962962962963e-05,
+      "loss": 2.0317,
+      "step": 5950
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4469361305236816,
+      "learning_rate": 4.6712592592592594e-05,
+      "loss": 2.0382,
+      "step": 5975
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.3143268823623657,
+      "learning_rate": 4.6675555555555555e-05,
+      "loss": 2.0107,
+      "step": 6000
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.322283148765564,
+      "learning_rate": 4.6638518518518523e-05,
+      "loss": 1.9968,
+      "step": 6025
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.5277537107467651,
+      "learning_rate": 4.6601481481481485e-05,
+      "loss": 2.0148,
+      "step": 6050
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.3555701971054077,
+      "learning_rate": 4.6564444444444447e-05,
+      "loss": 1.9547,
+      "step": 6075
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.3030173778533936,
+      "learning_rate": 4.652740740740741e-05,
+      "loss": 2.0186,
+      "step": 6100
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.1996219158172607,
+      "learning_rate": 4.6491851851851854e-05,
+      "loss": 1.9938,
+      "step": 6125
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.2713743448257446,
+      "learning_rate": 4.6454814814814815e-05,
+      "loss": 2.0135,
+      "step": 6150
+    },
+    {
+      "epoch": 2.4699999999999998,
+      "grad_norm": 1.4466146230697632,
+      "learning_rate": 4.6417777777777784e-05,
+      "loss": 2.0027,
+      "step": 6175
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.3012115955352783,
+      "learning_rate": 4.638074074074074e-05,
+      "loss": 2.0462,
+      "step": 6200
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.4161940813064575,
+      "learning_rate": 4.634370370370371e-05,
+      "loss": 2.0416,
+      "step": 6225
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.382034420967102,
+      "learning_rate": 4.630666666666667e-05,
+      "loss": 1.9795,
+      "step": 6250
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.4618980884552002,
+      "learning_rate": 4.626962962962963e-05,
+      "loss": 2.0163,
+      "step": 6275
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.304030179977417,
+      "learning_rate": 4.623259259259259e-05,
+      "loss": 1.9821,
+      "step": 6300
+    },
+    {
+      "epoch": 2.5300000000000002,
+      "grad_norm": 1.5101556777954102,
+      "learning_rate": 4.619555555555556e-05,
+      "loss": 1.9863,
+      "step": 6325
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.3217978477478027,
+      "learning_rate": 4.615851851851852e-05,
+      "loss": 2.0009,
+      "step": 6350
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.367722749710083,
+      "learning_rate": 4.612148148148148e-05,
+      "loss": 2.0459,
+      "step": 6375
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5014063119888306,
+      "learning_rate": 4.6084444444444444e-05,
+      "loss": 2.0129,
+      "step": 6400
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4557509422302246,
+      "learning_rate": 4.604740740740741e-05,
+      "loss": 1.9869,
+      "step": 6425
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.3464524745941162,
+      "learning_rate": 4.601037037037037e-05,
+      "loss": 2.0133,
+      "step": 6450
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.41274893283844,
+      "learning_rate": 4.5973333333333336e-05,
+      "loss": 1.996,
+      "step": 6475
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.3834503889083862,
+      "learning_rate": 4.59362962962963e-05,
+      "loss": 2.0057,
+      "step": 6500
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.4154237508773804,
+      "learning_rate": 4.5899259259259266e-05,
+      "loss": 2.0168,
+      "step": 6525
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.2246060371398926,
+      "learning_rate": 4.586222222222222e-05,
+      "loss": 1.9689,
+      "step": 6550
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.5142685174942017,
+      "learning_rate": 4.582518518518519e-05,
+      "loss": 2.0025,
+      "step": 6575
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.6933802366256714,
+      "learning_rate": 4.578814814814815e-05,
+      "loss": 2.0064,
+      "step": 6600
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.3439992666244507,
+      "learning_rate": 4.575111111111111e-05,
+      "loss": 1.9638,
+      "step": 6625
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.4206079244613647,
+      "learning_rate": 4.5714074074074074e-05,
+      "loss": 1.9888,
+      "step": 6650
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.3695896863937378,
+      "learning_rate": 4.567703703703704e-05,
+      "loss": 1.9735,
+      "step": 6675
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.2460039854049683,
+      "learning_rate": 4.564e-05,
+      "loss": 2.0707,
+      "step": 6700
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.2507898807525635,
+      "learning_rate": 4.5602962962962965e-05,
+      "loss": 2.0131,
+      "step": 6725
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.2939226627349854,
+      "learning_rate": 4.5565925925925927e-05,
+      "loss": 1.9902,
+      "step": 6750
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.4121019840240479,
+      "learning_rate": 4.5528888888888895e-05,
+      "loss": 1.9906,
+      "step": 6775
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 1.1183571815490723,
+      "learning_rate": 4.549185185185185e-05,
+      "loss": 2.0259,
+      "step": 6800
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4369187355041504,
+      "learning_rate": 4.545481481481482e-05,
+      "loss": 1.963,
+      "step": 6825
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.1758921146392822,
+      "learning_rate": 4.541777777777778e-05,
+      "loss": 1.9743,
+      "step": 6850
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.3526884317398071,
+      "learning_rate": 4.538074074074074e-05,
+      "loss": 2.0403,
+      "step": 6875
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.2723559141159058,
+      "learning_rate": 4.53437037037037e-05,
+      "loss": 1.958,
+      "step": 6900
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4061169624328613,
+      "learning_rate": 4.530666666666667e-05,
+      "loss": 2.0213,
+      "step": 6925
+    },
+    {
+      "epoch": 2.7800000000000002,
+      "grad_norm": 1.292668104171753,
+      "learning_rate": 4.526962962962963e-05,
+      "loss": 2.0301,
+      "step": 6950
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.2787526845932007,
+      "learning_rate": 4.5232592592592594e-05,
+      "loss": 1.9822,
+      "step": 6975
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.2812767028808594,
+      "learning_rate": 4.5195555555555556e-05,
+      "loss": 1.9919,
+      "step": 7000
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.2720341682434082,
+      "learning_rate": 4.5158518518518524e-05,
+      "loss": 2.0142,
+      "step": 7025
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.3361945152282715,
+      "learning_rate": 4.512148148148148e-05,
+      "loss": 2.0006,
+      "step": 7050
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.1397879123687744,
+      "learning_rate": 4.508444444444445e-05,
+      "loss": 1.9674,
+      "step": 7075
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.3024892807006836,
+      "learning_rate": 4.504740740740741e-05,
+      "loss": 1.969,
+      "step": 7100
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.2149879932403564,
+      "learning_rate": 4.501037037037038e-05,
+      "loss": 1.9873,
+      "step": 7125
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.2626320123672485,
+      "learning_rate": 4.497333333333333e-05,
+      "loss": 1.9552,
+      "step": 7150
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.3262602090835571,
+      "learning_rate": 4.49362962962963e-05,
+      "loss": 1.9485,
+      "step": 7175
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4364970922470093,
+      "learning_rate": 4.489925925925926e-05,
+      "loss": 2.0553,
+      "step": 7200
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.2992265224456787,
+      "learning_rate": 4.486222222222222e-05,
+      "loss": 2.0324,
+      "step": 7225
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.246061086654663,
+      "learning_rate": 4.4825185185185185e-05,
+      "loss": 1.9706,
+      "step": 7250
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.625594139099121,
+      "learning_rate": 4.478814814814815e-05,
+      "loss": 1.9419,
+      "step": 7275
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.2219802141189575,
+      "learning_rate": 4.4751111111111115e-05,
+      "loss": 2.0337,
+      "step": 7300
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.436552882194519,
+      "learning_rate": 4.4714074074074076e-05,
+      "loss": 1.9975,
+      "step": 7325
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.3120185136795044,
+      "learning_rate": 4.467703703703704e-05,
+      "loss": 2.0306,
+      "step": 7350
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.2584844827651978,
+      "learning_rate": 4.4640000000000006e-05,
+      "loss": 2.0335,
+      "step": 7375
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4038913249969482,
+      "learning_rate": 4.460296296296296e-05,
+      "loss": 1.9072,
+      "step": 7400
+    },
+    {
+      "epoch": 2.9699999999999998,
+      "grad_norm": 1.2124029397964478,
+      "learning_rate": 4.456592592592593e-05,
+      "loss": 2.0269,
+      "step": 7425
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.45619797706604,
+      "learning_rate": 4.452888888888889e-05,
+      "loss": 1.9908,
+      "step": 7450
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.2446014881134033,
+      "learning_rate": 4.449185185185185e-05,
+      "loss": 2.0035,
+      "step": 7475
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.4367069005966187,
+      "learning_rate": 4.4454814814814814e-05,
+      "loss": 2.0286,
+      "step": 7500
+    },
+    {
+      "epoch": 3.0,
+      "eval_gen_len": 13.1299,
+      "eval_loss": 1.9462145566940308,
+      "eval_rouge1": 51.2672,
+      "eval_rouge2": 25.5438,
+      "eval_rougeL": 47.1691,
+      "eval_rougeLsum": 47.1653,
+      "eval_runtime": 194.6381,
+      "eval_samples_per_second": 102.755,
+      "eval_steps_per_second": 1.608,
+      "step": 7500
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 1.4803961515426636,
+      "learning_rate": 4.441777777777778e-05,
+      "loss": 1.8803,
+      "step": 7525
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.3261098861694336,
+      "learning_rate": 4.4380740740740744e-05,
+      "loss": 1.9219,
+      "step": 7550
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.2701349258422852,
+      "learning_rate": 4.4343703703703706e-05,
+      "loss": 1.9066,
+      "step": 7575
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.4366850852966309,
+      "learning_rate": 4.430666666666667e-05,
+      "loss": 1.9186,
+      "step": 7600
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.304621934890747,
+      "learning_rate": 4.4269629629629635e-05,
+      "loss": 1.9499,
+      "step": 7625
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.4453222751617432,
+      "learning_rate": 4.423259259259259e-05,
+      "loss": 1.8863,
+      "step": 7650
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 1.1800676584243774,
+      "learning_rate": 4.419555555555556e-05,
+      "loss": 1.8879,
+      "step": 7675
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 1.5625019073486328,
+      "learning_rate": 4.415851851851852e-05,
+      "loss": 1.8945,
+      "step": 7700
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 1.702916145324707,
+      "learning_rate": 4.412148148148149e-05,
+      "loss": 1.9361,
+      "step": 7725
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.641830325126648,
+      "learning_rate": 4.408444444444444e-05,
+      "loss": 1.9031,
+      "step": 7750
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 1.6276681423187256,
+      "learning_rate": 4.404740740740741e-05,
+      "loss": 1.9207,
+      "step": 7775
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 1.4015640020370483,
+      "learning_rate": 4.401037037037037e-05,
+      "loss": 1.91,
+      "step": 7800
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 1.1643470525741577,
+      "learning_rate": 4.3973333333333335e-05,
+      "loss": 1.94,
+      "step": 7825
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 1.436637282371521,
+      "learning_rate": 4.393777777777778e-05,
+      "loss": 1.9152,
+      "step": 7850
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.2453664541244507,
+      "learning_rate": 4.390074074074074e-05,
+      "loss": 1.8912,
+      "step": 7875
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.507677674293518,
+      "learning_rate": 4.386370370370371e-05,
+      "loss": 1.9148,
+      "step": 7900
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 1.3336275815963745,
+      "learning_rate": 4.382666666666667e-05,
+      "loss": 1.9583,
+      "step": 7925
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.5578241348266602,
+      "learning_rate": 4.378962962962963e-05,
+      "loss": 1.9273,
+      "step": 7950
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 1.299078345298767,
+      "learning_rate": 4.3752592592592595e-05,
+      "loss": 1.9137,
+      "step": 7975
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.4711804389953613,
+      "learning_rate": 4.3715555555555556e-05,
+      "loss": 1.8916,
+      "step": 8000
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.258434534072876,
+      "learning_rate": 4.367851851851852e-05,
+      "loss": 1.8674,
+      "step": 8025
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.4030392169952393,
+      "learning_rate": 4.3641481481481486e-05,
+      "loss": 1.9207,
+      "step": 8050
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.3217052221298218,
+      "learning_rate": 4.360444444444445e-05,
+      "loss": 1.9356,
+      "step": 8075
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.4329230785369873,
+      "learning_rate": 4.356740740740741e-05,
+      "loss": 1.9849,
+      "step": 8100
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 1.4400358200073242,
+      "learning_rate": 4.353037037037037e-05,
+      "loss": 1.966,
+      "step": 8125
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.449614405632019,
+      "learning_rate": 4.349333333333334e-05,
+      "loss": 1.9351,
+      "step": 8150
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 1.3323086500167847,
+      "learning_rate": 4.34562962962963e-05,
+      "loss": 1.9247,
+      "step": 8175
+    },
+    {
+      "epoch": 3.2800000000000002,
+      "grad_norm": 1.267240047454834,
+      "learning_rate": 4.341925925925926e-05,
+      "loss": 1.902,
+      "step": 8200
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.276930332183838,
+      "learning_rate": 4.3382222222222224e-05,
+      "loss": 1.938,
+      "step": 8225
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.3871114253997803,
+      "learning_rate": 4.3345185185185186e-05,
+      "loss": 1.8964,
+      "step": 8250
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 1.4275935888290405,
+      "learning_rate": 4.330814814814815e-05,
+      "loss": 1.8912,
+      "step": 8275
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.4508986473083496,
+      "learning_rate": 4.3271111111111115e-05,
+      "loss": 1.959,
+      "step": 8300
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 1.4686108827590942,
+      "learning_rate": 4.323407407407408e-05,
+      "loss": 1.9674,
+      "step": 8325
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 1.2445565462112427,
+      "learning_rate": 4.319703703703704e-05,
+      "loss": 1.9429,
+      "step": 8350
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.3624173402786255,
+      "learning_rate": 4.316e-05,
+      "loss": 1.943,
+      "step": 8375
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.9012771844863892,
+      "learning_rate": 4.312296296296296e-05,
+      "loss": 1.9657,
+      "step": 8400
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.4263502359390259,
+      "learning_rate": 4.308592592592593e-05,
+      "loss": 1.9606,
+      "step": 8425
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.3634259700775146,
+      "learning_rate": 4.304888888888889e-05,
+      "loss": 1.8931,
+      "step": 8450
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 1.498611569404602,
+      "learning_rate": 4.301185185185185e-05,
+      "loss": 1.9311,
+      "step": 8475
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.5958226919174194,
+      "learning_rate": 4.2974814814814815e-05,
+      "loss": 1.9499,
+      "step": 8500
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 1.4740561246871948,
+      "learning_rate": 4.293777777777778e-05,
+      "loss": 1.9545,
+      "step": 8525
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.3004882335662842,
+      "learning_rate": 4.2900740740740745e-05,
+      "loss": 1.9883,
+      "step": 8550
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.7882790565490723,
+      "learning_rate": 4.2863703703703706e-05,
+      "loss": 1.9759,
+      "step": 8575
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.3567942380905151,
+      "learning_rate": 4.282666666666667e-05,
+      "loss": 1.9219,
+      "step": 8600
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.2835100889205933,
+      "learning_rate": 4.278962962962963e-05,
+      "loss": 1.9568,
+      "step": 8625
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 1.3128582239151,
+      "learning_rate": 4.275259259259259e-05,
+      "loss": 1.9076,
+      "step": 8650
+    },
+    {
+      "epoch": 3.4699999999999998,
+      "grad_norm": 1.4423998594284058,
+      "learning_rate": 4.271555555555556e-05,
+      "loss": 1.953,
+      "step": 8675
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.6110066175460815,
+      "learning_rate": 4.267851851851852e-05,
+      "loss": 1.9314,
+      "step": 8700
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 1.235214114189148,
+      "learning_rate": 4.264148148148148e-05,
+      "loss": 1.8981,
+      "step": 8725
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.223358392715454,
+      "learning_rate": 4.2604444444444444e-05,
+      "loss": 1.9607,
+      "step": 8750
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.7454544305801392,
+      "learning_rate": 4.256740740740741e-05,
+      "loss": 1.9374,
+      "step": 8775
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.6236810684204102,
+      "learning_rate": 4.253185185185186e-05,
+      "loss": 1.9503,
+      "step": 8800
+    },
+    {
+      "epoch": 3.5300000000000002,
+      "grad_norm": 1.454771637916565,
+      "learning_rate": 4.249481481481481e-05,
+      "loss": 1.9947,
+      "step": 8825
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.4015847444534302,
+      "learning_rate": 4.245777777777778e-05,
+      "loss": 1.9912,
+      "step": 8850
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.4323807954788208,
+      "learning_rate": 4.242074074074074e-05,
+      "loss": 1.9142,
+      "step": 8875
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.426339864730835,
+      "learning_rate": 4.2383703703703704e-05,
+      "loss": 1.9208,
+      "step": 8900
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.2626959085464478,
+      "learning_rate": 4.2346666666666666e-05,
+      "loss": 1.9237,
+      "step": 8925
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 1.3355377912521362,
+      "learning_rate": 4.2309629629629634e-05,
+      "loss": 1.9616,
+      "step": 8950
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.3246771097183228,
+      "learning_rate": 4.2272592592592595e-05,
+      "loss": 1.9595,
+      "step": 8975
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.3291839361190796,
+      "learning_rate": 4.223555555555556e-05,
+      "loss": 1.9718,
+      "step": 9000
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.4041240215301514,
+      "learning_rate": 4.219851851851852e-05,
+      "loss": 1.9147,
+      "step": 9025
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.5576542615890503,
+      "learning_rate": 4.216148148148149e-05,
+      "loss": 1.9467,
+      "step": 9050
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.355878472328186,
+      "learning_rate": 4.212444444444444e-05,
+      "loss": 1.932,
+      "step": 9075
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 1.3441921472549438,
+      "learning_rate": 4.208740740740741e-05,
+      "loss": 1.9606,
+      "step": 9100
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.5347744226455688,
+      "learning_rate": 4.205037037037037e-05,
+      "loss": 1.9081,
+      "step": 9125
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.4844331741333008,
+      "learning_rate": 4.201333333333334e-05,
+      "loss": 1.847,
+      "step": 9150
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.4049714803695679,
+      "learning_rate": 4.1976296296296295e-05,
+      "loss": 1.9042,
+      "step": 9175
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 1.4935232400894165,
+      "learning_rate": 4.193925925925926e-05,
+      "loss": 1.9575,
+      "step": 9200
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.6459118127822876,
+      "learning_rate": 4.1902222222222225e-05,
+      "loss": 1.9779,
+      "step": 9225
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.6347599029541016,
+      "learning_rate": 4.1865185185185186e-05,
+      "loss": 1.9265,
+      "step": 9250
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.5071406364440918,
+      "learning_rate": 4.182814814814815e-05,
+      "loss": 1.9869,
+      "step": 9275
+    },
+    {
+      "epoch": 3.7199999999999998,
+      "grad_norm": 1.521414041519165,
+      "learning_rate": 4.1791111111111116e-05,
+      "loss": 1.9607,
+      "step": 9300
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.6524264812469482,
+      "learning_rate": 4.175407407407408e-05,
+      "loss": 1.9068,
+      "step": 9325
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 1.5275657176971436,
+      "learning_rate": 4.171703703703704e-05,
+      "loss": 1.9225,
+      "step": 9350
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.2827988862991333,
+      "learning_rate": 4.168e-05,
+      "loss": 1.9199,
+      "step": 9375
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.5955214500427246,
+      "learning_rate": 4.164296296296297e-05,
+      "loss": 1.8854,
+      "step": 9400
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.4321446418762207,
+      "learning_rate": 4.1605925925925924e-05,
+      "loss": 1.9488,
+      "step": 9425
+    },
+    {
+      "epoch": 3.7800000000000002,
+      "grad_norm": 1.3754669427871704,
+      "learning_rate": 4.156888888888889e-05,
+      "loss": 1.9209,
+      "step": 9450
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.1604045629501343,
+      "learning_rate": 4.1531851851851854e-05,
+      "loss": 1.9515,
+      "step": 9475
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.3413041830062866,
+      "learning_rate": 4.1494814814814815e-05,
+      "loss": 1.9457,
+      "step": 9500
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.349363088607788,
+      "learning_rate": 4.145777777777778e-05,
+      "loss": 1.9789,
+      "step": 9525
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.3260554075241089,
+      "learning_rate": 4.1420740740740745e-05,
+      "loss": 1.8999,
+      "step": 9550
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.289332628250122,
+      "learning_rate": 4.138370370370371e-05,
+      "loss": 1.9856,
+      "step": 9575
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.4211158752441406,
+      "learning_rate": 4.134666666666667e-05,
+      "loss": 1.919,
+      "step": 9600
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 1.4039026498794556,
+      "learning_rate": 4.130962962962963e-05,
+      "loss": 1.9073,
+      "step": 9625
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.2093830108642578,
+      "learning_rate": 4.12725925925926e-05,
+      "loss": 1.9501,
+      "step": 9650
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.323862075805664,
+      "learning_rate": 4.123555555555555e-05,
+      "loss": 1.8516,
+      "step": 9675
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.4265086650848389,
+      "learning_rate": 4.119851851851852e-05,
+      "loss": 1.9498,
+      "step": 9700
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 1.4013090133666992,
+      "learning_rate": 4.116148148148148e-05,
+      "loss": 1.9302,
+      "step": 9725
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 1.44480562210083,
+      "learning_rate": 4.112444444444445e-05,
+      "loss": 1.9059,
+      "step": 9750
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.6620755195617676,
+      "learning_rate": 4.1087407407407406e-05,
+      "loss": 1.9037,
+      "step": 9775
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.4004356861114502,
+      "learning_rate": 4.1050370370370374e-05,
+      "loss": 1.8987,
+      "step": 9800
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.536922812461853,
+      "learning_rate": 4.1013333333333336e-05,
+      "loss": 2.0043,
+      "step": 9825
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.4472166299819946,
+      "learning_rate": 4.09762962962963e-05,
+      "loss": 1.9224,
+      "step": 9850
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.3354836702346802,
+      "learning_rate": 4.093925925925926e-05,
+      "loss": 1.934,
+      "step": 9875
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.3124749660491943,
+      "learning_rate": 4.090222222222223e-05,
+      "loss": 1.9879,
+      "step": 9900
+    },
+    {
+      "epoch": 3.9699999999999998,
+      "grad_norm": 1.283579707145691,
+      "learning_rate": 4.086518518518519e-05,
+      "loss": 1.8905,
+      "step": 9925
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.672229290008545,
+      "learning_rate": 4.082814814814815e-05,
+      "loss": 1.8554,
+      "step": 9950
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 1.5786038637161255,
+      "learning_rate": 4.079111111111111e-05,
+      "loss": 1.8784,
+      "step": 9975
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.361163854598999,
+      "learning_rate": 4.075407407407408e-05,
+      "loss": 1.9669,
+      "step": 10000
+    },
+    {
+      "epoch": 4.0,
+      "eval_gen_len": 13.1328,
+      "eval_loss": 1.9336023330688477,
+      "eval_rouge1": 51.5525,
+      "eval_rouge2": 25.7439,
+      "eval_rougeL": 47.3861,
+      "eval_rougeLsum": 47.3849,
+      "eval_runtime": 193.2674,
+      "eval_samples_per_second": 103.484,
+      "eval_steps_per_second": 1.62,
+      "step": 10000
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 1.6996824741363525,
+      "learning_rate": 4.0717037037037035e-05,
+      "loss": 1.8428,
+      "step": 10025
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 1.6396452188491821,
+      "learning_rate": 4.0680000000000004e-05,
+      "loss": 1.9144,
+      "step": 10050
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.508521556854248,
+      "learning_rate": 4.0642962962962965e-05,
+      "loss": 1.8983,
+      "step": 10075
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.3133175373077393,
+      "learning_rate": 4.0605925925925933e-05,
+      "loss": 1.8159,
+      "step": 10100
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.4404352903366089,
+      "learning_rate": 4.056888888888889e-05,
+      "loss": 1.8468,
+      "step": 10125
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 1.410372257232666,
+      "learning_rate": 4.0531851851851857e-05,
+      "loss": 1.9628,
+      "step": 10150
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 1.509523630142212,
+      "learning_rate": 4.049481481481482e-05,
+      "loss": 1.8723,
+      "step": 10175
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 1.488282561302185,
+      "learning_rate": 4.045777777777778e-05,
+      "loss": 1.8414,
+      "step": 10200
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.3858217000961304,
+      "learning_rate": 4.042074074074074e-05,
+      "loss": 1.8241,
+      "step": 10225
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.5078978538513184,
+      "learning_rate": 4.038370370370371e-05,
+      "loss": 1.9011,
+      "step": 10250
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 1.340116262435913,
+      "learning_rate": 4.0346666666666664e-05,
+      "loss": 1.8832,
+      "step": 10275
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 1.4269579648971558,
+      "learning_rate": 4.030962962962963e-05,
+      "loss": 1.8314,
+      "step": 10300
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 1.3984671831130981,
+      "learning_rate": 4.0272592592592594e-05,
+      "loss": 1.8422,
+      "step": 10325
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 1.5651813745498657,
+      "learning_rate": 4.023555555555556e-05,
+      "loss": 1.8475,
+      "step": 10350
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.5041767358779907,
+      "learning_rate": 4.019851851851852e-05,
+      "loss": 1.8392,
+      "step": 10375
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.6004717350006104,
+      "learning_rate": 4.0161481481481486e-05,
+      "loss": 1.8364,
+      "step": 10400
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.3751362562179565,
+      "learning_rate": 4.012444444444445e-05,
+      "loss": 1.8457,
+      "step": 10425
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 1.4664337635040283,
+      "learning_rate": 4.008740740740741e-05,
+      "loss": 1.9372,
+      "step": 10450
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 1.3487550020217896,
+      "learning_rate": 4.005037037037037e-05,
+      "loss": 1.9049,
+      "step": 10475
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 1.2509174346923828,
+      "learning_rate": 4.001333333333334e-05,
+      "loss": 1.8987,
+      "step": 10500
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.5854827165603638,
+      "learning_rate": 3.99762962962963e-05,
+      "loss": 1.8423,
+      "step": 10525
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 1.4388275146484375,
+      "learning_rate": 3.993925925925926e-05,
+      "loss": 1.8349,
+      "step": 10550
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.4520167112350464,
+      "learning_rate": 3.9902222222222223e-05,
+      "loss": 1.8283,
+      "step": 10575
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.44745671749115,
+      "learning_rate": 3.986518518518519e-05,
+      "loss": 1.9163,
+      "step": 10600
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.4306520223617554,
+      "learning_rate": 3.9828148148148147e-05,
+      "loss": 1.8803,
+      "step": 10625
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 1.3885631561279297,
+      "learning_rate": 3.9791111111111115e-05,
+      "loss": 1.8763,
+      "step": 10650
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 1.3905619382858276,
+      "learning_rate": 3.9754074074074076e-05,
+      "loss": 1.8828,
+      "step": 10675
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.3806442022323608,
+      "learning_rate": 3.9717037037037045e-05,
+      "loss": 1.908,
+      "step": 10700
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.4566986560821533,
+      "learning_rate": 3.968e-05,
+      "loss": 1.8479,
+      "step": 10725
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.4372694492340088,
+      "learning_rate": 3.964296296296297e-05,
+      "loss": 1.9098,
+      "step": 10750
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.5970664024353027,
+      "learning_rate": 3.960592592592593e-05,
+      "loss": 1.9132,
+      "step": 10775
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.3332023620605469,
+      "learning_rate": 3.956888888888889e-05,
+      "loss": 1.8479,
+      "step": 10800
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.4356377124786377,
+      "learning_rate": 3.953185185185185e-05,
+      "loss": 1.8216,
+      "step": 10825
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 1.3846675157546997,
+      "learning_rate": 3.949481481481482e-05,
+      "loss": 1.9176,
+      "step": 10850
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.4137978553771973,
+      "learning_rate": 3.945777777777778e-05,
+      "loss": 1.8499,
+      "step": 10875
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 1.456305980682373,
+      "learning_rate": 3.9420740740740744e-05,
+      "loss": 1.8801,
+      "step": 10900
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 1.4478751420974731,
+      "learning_rate": 3.9383703703703706e-05,
+      "loss": 1.8548,
+      "step": 10925
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 1.2786507606506348,
+      "learning_rate": 3.9346666666666674e-05,
+      "loss": 1.8149,
+      "step": 10950
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 1.5575779676437378,
+      "learning_rate": 3.930962962962963e-05,
+      "loss": 1.8687,
+      "step": 10975
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.6050180196762085,
+      "learning_rate": 3.92725925925926e-05,
+      "loss": 1.8918,
+      "step": 11000
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 1.4834775924682617,
+      "learning_rate": 3.923555555555556e-05,
+      "loss": 1.9042,
+      "step": 11025
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 1.3932939767837524,
+      "learning_rate": 3.919851851851852e-05,
+      "loss": 1.8931,
+      "step": 11050
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 1.4138563871383667,
+      "learning_rate": 3.916148148148148e-05,
+      "loss": 1.8471,
+      "step": 11075
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.5365270376205444,
+      "learning_rate": 3.912444444444445e-05,
+      "loss": 1.8746,
+      "step": 11100
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.2787660360336304,
+      "learning_rate": 3.908740740740741e-05,
+      "loss": 1.857,
+      "step": 11125
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 1.3813387155532837,
+      "learning_rate": 3.905037037037037e-05,
+      "loss": 1.8436,
+      "step": 11150
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 1.3845306634902954,
+      "learning_rate": 3.9013333333333335e-05,
+      "loss": 1.8494,
+      "step": 11175
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.4810049533843994,
+      "learning_rate": 3.8976296296296296e-05,
+      "loss": 1.9129,
+      "step": 11200
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 1.5942555665969849,
+      "learning_rate": 3.893925925925926e-05,
+      "loss": 1.8596,
+      "step": 11225
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 1.3388746976852417,
+      "learning_rate": 3.8902222222222226e-05,
+      "loss": 1.8152,
+      "step": 11250
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.2811378240585327,
+      "learning_rate": 3.886518518518519e-05,
+      "loss": 1.8035,
+      "step": 11275
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 1.4130557775497437,
+      "learning_rate": 3.882814814814815e-05,
+      "loss": 1.8349,
+      "step": 11300
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 1.3981435298919678,
+      "learning_rate": 3.879111111111111e-05,
+      "loss": 1.9117,
+      "step": 11325
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.4052165746688843,
+      "learning_rate": 3.875407407407408e-05,
+      "loss": 1.9413,
+      "step": 11350
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.462985634803772,
+      "learning_rate": 3.871703703703704e-05,
+      "loss": 1.8346,
+      "step": 11375
+    },
+    {
+      "epoch": 4.5600000000000005,
+      "grad_norm": 1.417001485824585,
+      "learning_rate": 3.868e-05,
+      "loss": 1.8613,
+      "step": 11400
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 1.3185068368911743,
+      "learning_rate": 3.8642962962962964e-05,
+      "loss": 1.8832,
+      "step": 11425
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 1.4845627546310425,
+      "learning_rate": 3.8605925925925925e-05,
+      "loss": 1.8526,
+      "step": 11450
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.5781909227371216,
+      "learning_rate": 3.8568888888888894e-05,
+      "loss": 1.9009,
+      "step": 11475
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.4118759632110596,
+      "learning_rate": 3.8531851851851855e-05,
+      "loss": 1.8705,
+      "step": 11500
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 1.604973316192627,
+      "learning_rate": 3.849481481481482e-05,
+      "loss": 1.8933,
+      "step": 11525
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.2688302993774414,
+      "learning_rate": 3.845777777777778e-05,
+      "loss": 1.8558,
+      "step": 11550
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 1.3213862180709839,
+      "learning_rate": 3.842074074074074e-05,
+      "loss": 1.8655,
+      "step": 11575
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 1.5168383121490479,
+      "learning_rate": 3.838370370370371e-05,
+      "loss": 1.8862,
+      "step": 11600
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.30950927734375,
+      "learning_rate": 3.834666666666667e-05,
+      "loss": 1.8862,
+      "step": 11625
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.4738622903823853,
+      "learning_rate": 3.830962962962963e-05,
+      "loss": 1.8185,
+      "step": 11650
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 1.4453907012939453,
+      "learning_rate": 3.827259259259259e-05,
+      "loss": 1.8367,
+      "step": 11675
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 1.5001963376998901,
+      "learning_rate": 3.8235555555555555e-05,
+      "loss": 1.9168,
+      "step": 11700
+    },
+    {
+      "epoch": 4.6899999999999995,
+      "grad_norm": 1.4097387790679932,
+      "learning_rate": 3.819851851851852e-05,
+      "loss": 1.8582,
+      "step": 11725
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.5029624700546265,
+      "learning_rate": 3.8161481481481485e-05,
+      "loss": 1.87,
+      "step": 11750
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.4353328943252563,
+      "learning_rate": 3.8124444444444446e-05,
+      "loss": 1.8595,
+      "step": 11775
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.6048336029052734,
+      "learning_rate": 3.808740740740741e-05,
+      "loss": 1.9087,
+      "step": 11800
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 1.3120425939559937,
+      "learning_rate": 3.805037037037037e-05,
+      "loss": 1.8873,
+      "step": 11825
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 1.6301329135894775,
+      "learning_rate": 3.801333333333333e-05,
+      "loss": 1.8895,
+      "step": 11850
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 1.480413556098938,
+      "learning_rate": 3.79762962962963e-05,
+      "loss": 1.8713,
+      "step": 11875
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 1.2750059366226196,
+      "learning_rate": 3.793925925925926e-05,
+      "loss": 1.8706,
+      "step": 11900
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.5305309295654297,
+      "learning_rate": 3.790222222222222e-05,
+      "loss": 1.9087,
+      "step": 11925
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 1.3535616397857666,
+      "learning_rate": 3.7865185185185184e-05,
+      "loss": 1.8658,
+      "step": 11950
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 1.5504143238067627,
+      "learning_rate": 3.782814814814815e-05,
+      "loss": 1.8742,
+      "step": 11975
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.7946292161941528,
+      "learning_rate": 3.7791111111111114e-05,
+      "loss": 1.8456,
+      "step": 12000
+    },
+    {
+      "epoch": 4.8100000000000005,
+      "grad_norm": 1.395272135734558,
+      "learning_rate": 3.7754074074074075e-05,
+      "loss": 1.931,
+      "step": 12025
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 1.4061551094055176,
+      "learning_rate": 3.771703703703704e-05,
+      "loss": 1.8836,
+      "step": 12050
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 1.3979151248931885,
+      "learning_rate": 3.7680000000000005e-05,
+      "loss": 1.8869,
+      "step": 12075
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.4107104539871216,
+      "learning_rate": 3.764296296296296e-05,
+      "loss": 1.8736,
+      "step": 12100
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 1.321219801902771,
+      "learning_rate": 3.760592592592593e-05,
+      "loss": 1.8761,
+      "step": 12125
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.5014673471450806,
+      "learning_rate": 3.756888888888889e-05,
+      "loss": 1.8445,
+      "step": 12150
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.4317951202392578,
+      "learning_rate": 3.753185185185185e-05,
+      "loss": 1.848,
+      "step": 12175
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 1.4096531867980957,
+      "learning_rate": 3.749481481481481e-05,
+      "loss": 1.8933,
+      "step": 12200
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.2559773921966553,
+      "learning_rate": 3.745777777777778e-05,
+      "loss": 1.8033,
+      "step": 12225
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 1.3311185836791992,
+      "learning_rate": 3.742074074074074e-05,
+      "loss": 1.8796,
+      "step": 12250
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 1.4260166883468628,
+      "learning_rate": 3.7383703703703704e-05,
+      "loss": 1.8633,
+      "step": 12275
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.3146923780441284,
+      "learning_rate": 3.7346666666666666e-05,
+      "loss": 1.888,
+      "step": 12300
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 1.3045905828475952,
+      "learning_rate": 3.7309629629629634e-05,
+      "loss": 1.8274,
+      "step": 12325
+    },
+    {
+      "epoch": 4.9399999999999995,
+      "grad_norm": 1.3906166553497314,
+      "learning_rate": 3.727259259259259e-05,
+      "loss": 1.9098,
+      "step": 12350
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 1.518526554107666,
+      "learning_rate": 3.723555555555556e-05,
+      "loss": 1.9287,
+      "step": 12375
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 1.3113497495651245,
+      "learning_rate": 3.719851851851852e-05,
+      "loss": 1.8677,
+      "step": 12400
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 1.6457515954971313,
+      "learning_rate": 3.716148148148149e-05,
+      "loss": 1.9183,
+      "step": 12425
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.5027391910552979,
+      "learning_rate": 3.712444444444444e-05,
+      "loss": 1.8613,
+      "step": 12450
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 1.561333179473877,
+      "learning_rate": 3.708740740740741e-05,
+      "loss": 1.9073,
+      "step": 12475
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.6902166604995728,
+      "learning_rate": 3.705037037037037e-05,
+      "loss": 1.8617,
+      "step": 12500
+    },
+    {
+      "epoch": 5.0,
+      "eval_gen_len": 13.1348,
+      "eval_loss": 1.9302258491516113,
+      "eval_rouge1": 51.4633,
+      "eval_rouge2": 25.7335,
+      "eval_rougeL": 47.3103,
+      "eval_rougeLsum": 47.305,
+      "eval_runtime": 194.3038,
+      "eval_samples_per_second": 102.932,
+      "eval_steps_per_second": 1.611,
+      "step": 12500
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 1.3787955045700073,
+      "learning_rate": 3.7013333333333334e-05,
+      "loss": 1.763,
+      "step": 12525
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 1.3770140409469604,
+      "learning_rate": 3.6976296296296295e-05,
+      "loss": 1.8189,
+      "step": 12550
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 1.486107587814331,
+      "learning_rate": 3.6939259259259263e-05,
+      "loss": 1.8043,
+      "step": 12575
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 1.3387417793273926,
+      "learning_rate": 3.690222222222222e-05,
+      "loss": 1.8575,
+      "step": 12600
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 1.340170979499817,
+      "learning_rate": 3.6865185185185187e-05,
+      "loss": 1.8634,
+      "step": 12625
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 1.2771445512771606,
+      "learning_rate": 3.682814814814815e-05,
+      "loss": 1.8065,
+      "step": 12650
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 1.3724429607391357,
+      "learning_rate": 3.6791111111111116e-05,
+      "loss": 1.8152,
+      "step": 12675
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 1.2865967750549316,
+      "learning_rate": 3.675407407407407e-05,
+      "loss": 1.8625,
+      "step": 12700
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 1.4663174152374268,
+      "learning_rate": 3.671703703703704e-05,
+      "loss": 1.7772,
+      "step": 12725
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 1.6799187660217285,
+      "learning_rate": 3.668e-05,
+      "loss": 1.8082,
+      "step": 12750
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 1.3404431343078613,
+      "learning_rate": 3.664296296296296e-05,
+      "loss": 1.8315,
+      "step": 12775
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 1.4227215051651,
+      "learning_rate": 3.6605925925925924e-05,
+      "loss": 1.7863,
+      "step": 12800
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 1.4327188730239868,
+      "learning_rate": 3.656888888888889e-05,
+      "loss": 1.8599,
+      "step": 12825
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 1.4590907096862793,
+      "learning_rate": 3.6531851851851854e-05,
+      "loss": 1.8372,
+      "step": 12850
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 1.4968684911727905,
+      "learning_rate": 3.6494814814814816e-05,
+      "loss": 1.8266,
+      "step": 12875
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 1.3817492723464966,
+      "learning_rate": 3.645925925925926e-05,
+      "loss": 1.8532,
+      "step": 12900
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 1.5503273010253906,
+      "learning_rate": 3.642222222222222e-05,
+      "loss": 1.8062,
+      "step": 12925
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 1.5043991804122925,
+      "learning_rate": 3.638518518518519e-05,
+      "loss": 1.8428,
+      "step": 12950
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 1.3269332647323608,
+      "learning_rate": 3.6348148148148146e-05,
+      "loss": 1.7959,
+      "step": 12975
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 1.3784312009811401,
+      "learning_rate": 3.6311111111111114e-05,
+      "loss": 1.8489,
+      "step": 13000
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 1.557144284248352,
+      "learning_rate": 3.6274074074074076e-05,
+      "loss": 1.8175,
+      "step": 13025
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 1.7027719020843506,
+      "learning_rate": 3.623703703703704e-05,
+      "loss": 1.8363,
+      "step": 13050
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 1.415287733078003,
+      "learning_rate": 3.62e-05,
+      "loss": 1.8242,
+      "step": 13075
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 1.4795790910720825,
+      "learning_rate": 3.616296296296297e-05,
+      "loss": 1.7947,
+      "step": 13100
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 1.703626036643982,
+      "learning_rate": 3.612592592592593e-05,
+      "loss": 1.8156,
+      "step": 13125
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 1.5055326223373413,
+      "learning_rate": 3.608888888888889e-05,
+      "loss": 1.8225,
+      "step": 13150
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 1.5992326736450195,
+      "learning_rate": 3.605185185185185e-05,
+      "loss": 1.8556,
+      "step": 13175
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 1.630657434463501,
+      "learning_rate": 3.601481481481482e-05,
+      "loss": 1.8139,
+      "step": 13200
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 1.4624261856079102,
+      "learning_rate": 3.5977777777777775e-05,
+      "loss": 1.8808,
+      "step": 13225
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 1.6484681367874146,
+      "learning_rate": 3.5940740740740743e-05,
+      "loss": 1.7941,
+      "step": 13250
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 1.5216662883758545,
+      "learning_rate": 3.5903703703703705e-05,
+      "loss": 1.7918,
+      "step": 13275
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 1.578412413597107,
+      "learning_rate": 3.586666666666667e-05,
+      "loss": 1.8681,
+      "step": 13300
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 1.547946572303772,
+      "learning_rate": 3.582962962962963e-05,
+      "loss": 1.8343,
+      "step": 13325
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 1.491708517074585,
+      "learning_rate": 3.5792592592592596e-05,
+      "loss": 1.7636,
+      "step": 13350
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 1.4895257949829102,
+      "learning_rate": 3.575555555555556e-05,
+      "loss": 1.8088,
+      "step": 13375
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 1.5237321853637695,
+      "learning_rate": 3.571851851851852e-05,
+      "loss": 1.8399,
+      "step": 13400
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 1.5580129623413086,
+      "learning_rate": 3.568148148148148e-05,
+      "loss": 1.8515,
+      "step": 13425
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 1.3289860486984253,
+      "learning_rate": 3.564444444444445e-05,
+      "loss": 1.7867,
+      "step": 13450
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 1.3806110620498657,
+      "learning_rate": 3.560740740740741e-05,
+      "loss": 1.8141,
+      "step": 13475
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 1.393282175064087,
+      "learning_rate": 3.557037037037037e-05,
+      "loss": 1.7835,
+      "step": 13500
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 1.3642027378082275,
+      "learning_rate": 3.5533333333333334e-05,
+      "loss": 1.797,
+      "step": 13525
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 1.378496766090393,
+      "learning_rate": 3.54962962962963e-05,
+      "loss": 1.8129,
+      "step": 13550
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 1.7297312021255493,
+      "learning_rate": 3.545925925925926e-05,
+      "loss": 1.8175,
+      "step": 13575
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 1.3795442581176758,
+      "learning_rate": 3.5422222222222226e-05,
+      "loss": 1.8089,
+      "step": 13600
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 1.577501654624939,
+      "learning_rate": 3.538518518518519e-05,
+      "loss": 1.8813,
+      "step": 13625
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 1.478531002998352,
+      "learning_rate": 3.5348148148148156e-05,
+      "loss": 1.7873,
+      "step": 13650
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 1.4849592447280884,
+      "learning_rate": 3.531111111111111e-05,
+      "loss": 1.8269,
+      "step": 13675
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 1.413120985031128,
+      "learning_rate": 3.527407407407408e-05,
+      "loss": 1.8449,
+      "step": 13700
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 1.6044095754623413,
+      "learning_rate": 3.523703703703704e-05,
+      "loss": 1.7789,
+      "step": 13725
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 1.4970237016677856,
+      "learning_rate": 3.52e-05,
+      "loss": 1.8088,
+      "step": 13750
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 1.735145926475525,
+      "learning_rate": 3.516296296296296e-05,
+      "loss": 1.8242,
+      "step": 13775
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 1.592650055885315,
+      "learning_rate": 3.512592592592593e-05,
+      "loss": 1.8115,
+      "step": 13800
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 1.3803372383117676,
+      "learning_rate": 3.5088888888888886e-05,
+      "loss": 1.7575,
+      "step": 13825
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 1.5937862396240234,
+      "learning_rate": 3.5051851851851855e-05,
+      "loss": 1.7927,
+      "step": 13850
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 1.6011449098587036,
+      "learning_rate": 3.5014814814814816e-05,
+      "loss": 1.8037,
+      "step": 13875
+    },
+    {
+      "epoch": 5.5600000000000005,
+      "grad_norm": 1.4640209674835205,
+      "learning_rate": 3.4977777777777785e-05,
+      "loss": 1.7909,
+      "step": 13900
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 1.577725887298584,
+      "learning_rate": 3.494074074074074e-05,
+      "loss": 1.8111,
+      "step": 13925
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 1.4513524770736694,
+      "learning_rate": 3.490370370370371e-05,
+      "loss": 1.7605,
+      "step": 13950
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 1.6748623847961426,
+      "learning_rate": 3.486666666666667e-05,
+      "loss": 1.8199,
+      "step": 13975
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 1.5236202478408813,
+      "learning_rate": 3.482962962962963e-05,
+      "loss": 1.8162,
+      "step": 14000
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 1.5020064115524292,
+      "learning_rate": 3.479259259259259e-05,
+      "loss": 1.8282,
+      "step": 14025
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 1.5203158855438232,
+      "learning_rate": 3.475555555555556e-05,
+      "loss": 1.8871,
+      "step": 14050
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 1.4594610929489136,
+      "learning_rate": 3.471851851851852e-05,
+      "loss": 1.818,
+      "step": 14075
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 1.5922948122024536,
+      "learning_rate": 3.4681481481481484e-05,
+      "loss": 1.8806,
+      "step": 14100
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 1.5797369480133057,
+      "learning_rate": 3.4644444444444446e-05,
+      "loss": 1.8307,
+      "step": 14125
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 1.5268616676330566,
+      "learning_rate": 3.4607407407407414e-05,
+      "loss": 1.8021,
+      "step": 14150
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 1.6109580993652344,
+      "learning_rate": 3.457037037037037e-05,
+      "loss": 1.8541,
+      "step": 14175
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 1.4869205951690674,
+      "learning_rate": 3.453333333333334e-05,
+      "loss": 1.8282,
+      "step": 14200
+    },
+    {
+      "epoch": 5.6899999999999995,
+      "grad_norm": 1.5621397495269775,
+      "learning_rate": 3.44962962962963e-05,
+      "loss": 1.8665,
+      "step": 14225
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 1.6197333335876465,
+      "learning_rate": 3.445925925925926e-05,
+      "loss": 1.9229,
+      "step": 14250
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 1.4683600664138794,
+      "learning_rate": 3.442222222222222e-05,
+      "loss": 1.8309,
+      "step": 14275
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 1.4849534034729004,
+      "learning_rate": 3.438518518518519e-05,
+      "loss": 1.8315,
+      "step": 14300
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 1.2944457530975342,
+      "learning_rate": 3.434814814814815e-05,
+      "loss": 1.7922,
+      "step": 14325
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 1.4732681512832642,
+      "learning_rate": 3.431111111111111e-05,
+      "loss": 1.8181,
+      "step": 14350
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 1.362735629081726,
+      "learning_rate": 3.4274074074074075e-05,
+      "loss": 1.8194,
+      "step": 14375
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 1.5724767446517944,
+      "learning_rate": 3.423703703703704e-05,
+      "loss": 1.9212,
+      "step": 14400
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 1.430247187614441,
+      "learning_rate": 3.4200000000000005e-05,
+      "loss": 1.8141,
+      "step": 14425
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 1.4439537525177002,
+      "learning_rate": 3.4162962962962966e-05,
+      "loss": 1.8393,
+      "step": 14450
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 1.6095616817474365,
+      "learning_rate": 3.412592592592593e-05,
+      "loss": 1.8514,
+      "step": 14475
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 1.3736265897750854,
+      "learning_rate": 3.408888888888889e-05,
+      "loss": 1.8419,
+      "step": 14500
+    },
+    {
+      "epoch": 5.8100000000000005,
+      "grad_norm": 1.5470609664916992,
+      "learning_rate": 3.405185185185185e-05,
+      "loss": 1.8313,
+      "step": 14525
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 1.4805546998977661,
+      "learning_rate": 3.401481481481482e-05,
+      "loss": 1.8075,
+      "step": 14550
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 1.3933120965957642,
+      "learning_rate": 3.397777777777778e-05,
+      "loss": 1.7556,
+      "step": 14575
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 1.3401881456375122,
+      "learning_rate": 3.394074074074074e-05,
+      "loss": 1.7676,
+      "step": 14600
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 1.713220238685608,
+      "learning_rate": 3.3903703703703704e-05,
+      "loss": 1.8106,
+      "step": 14625
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 1.4203695058822632,
+      "learning_rate": 3.3866666666666665e-05,
+      "loss": 1.8415,
+      "step": 14650
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 1.406247854232788,
+      "learning_rate": 3.3829629629629634e-05,
+      "loss": 1.7898,
+      "step": 14675
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 1.422446846961975,
+      "learning_rate": 3.3792592592592595e-05,
+      "loss": 1.7958,
+      "step": 14700
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 1.5542668104171753,
+      "learning_rate": 3.375555555555556e-05,
+      "loss": 1.8299,
+      "step": 14725
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 1.426229476928711,
+      "learning_rate": 3.371851851851852e-05,
+      "loss": 1.8498,
+      "step": 14750
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 1.4984149932861328,
+      "learning_rate": 3.368148148148148e-05,
+      "loss": 1.8642,
+      "step": 14775
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 1.347084879875183,
+      "learning_rate": 3.364444444444445e-05,
+      "loss": 1.8321,
+      "step": 14800
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 1.4737646579742432,
+      "learning_rate": 3.360740740740741e-05,
+      "loss": 1.9252,
+      "step": 14825
+    },
+    {
+      "epoch": 5.9399999999999995,
+      "grad_norm": 1.3631387948989868,
+      "learning_rate": 3.357037037037037e-05,
+      "loss": 1.8539,
+      "step": 14850
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 1.3909265995025635,
+      "learning_rate": 3.353333333333333e-05,
+      "loss": 1.8415,
+      "step": 14875
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 1.398703932762146,
+      "learning_rate": 3.3496296296296295e-05,
+      "loss": 1.823,
+      "step": 14900
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 1.5495917797088623,
+      "learning_rate": 3.345925925925926e-05,
+      "loss": 1.7733,
+      "step": 14925
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 1.7243595123291016,
+      "learning_rate": 3.3422222222222224e-05,
+      "loss": 1.8468,
+      "step": 14950
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 1.5244386196136475,
+      "learning_rate": 3.3385185185185186e-05,
+      "loss": 1.8229,
+      "step": 14975
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.4378917217254639,
+      "learning_rate": 3.334814814814815e-05,
+      "loss": 1.7693,
+      "step": 15000
+    },
+    {
+      "epoch": 6.0,
+      "eval_gen_len": 13.2205,
+      "eval_loss": 1.9290603399276733,
+      "eval_rouge1": 51.6382,
+      "eval_rouge2": 25.9065,
+      "eval_rougeL": 47.4714,
+      "eval_rougeLsum": 47.4661,
+      "eval_runtime": 195.9112,
+      "eval_samples_per_second": 102.087,
+      "eval_steps_per_second": 1.598,
+      "step": 15000
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 1.421264886856079,
+      "learning_rate": 3.3311111111111116e-05,
+      "loss": 1.7225,
+      "step": 15025
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 1.5342766046524048,
+      "learning_rate": 3.327407407407408e-05,
+      "loss": 1.8321,
+      "step": 15050
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 1.4197121858596802,
+      "learning_rate": 3.323703703703704e-05,
+      "loss": 1.7555,
+      "step": 15075
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 1.6157759428024292,
+      "learning_rate": 3.32e-05,
+      "loss": 1.7768,
+      "step": 15100
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 1.5571355819702148,
+      "learning_rate": 3.316296296296296e-05,
+      "loss": 1.7692,
+      "step": 15125
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 1.5974923372268677,
+      "learning_rate": 3.3125925925925924e-05,
+      "loss": 1.7249,
+      "step": 15150
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 1.3685996532440186,
+      "learning_rate": 3.308888888888889e-05,
+      "loss": 1.7672,
+      "step": 15175
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 1.4658602476119995,
+      "learning_rate": 3.3051851851851854e-05,
+      "loss": 1.7717,
+      "step": 15200
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 1.399776577949524,
+      "learning_rate": 3.3014814814814815e-05,
+      "loss": 1.7609,
+      "step": 15225
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 1.6905499696731567,
+      "learning_rate": 3.297777777777778e-05,
+      "loss": 1.7796,
+      "step": 15250
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 1.7223976850509644,
+      "learning_rate": 3.2940740740740745e-05,
+      "loss": 1.7806,
+      "step": 15275
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 1.4597574472427368,
+      "learning_rate": 3.29037037037037e-05,
+      "loss": 1.7689,
+      "step": 15300
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 1.7513933181762695,
+      "learning_rate": 3.286666666666667e-05,
+      "loss": 1.8005,
+      "step": 15325
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 1.7440515756607056,
+      "learning_rate": 3.282962962962963e-05,
+      "loss": 1.7965,
+      "step": 15350
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 1.6327168941497803,
+      "learning_rate": 3.279259259259259e-05,
+      "loss": 1.7875,
+      "step": 15375
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 1.5745080709457397,
+      "learning_rate": 3.275555555555555e-05,
+      "loss": 1.7659,
+      "step": 15400
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 1.5755482912063599,
+      "learning_rate": 3.271851851851852e-05,
+      "loss": 1.7724,
+      "step": 15425
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 1.6102346181869507,
+      "learning_rate": 3.268148148148148e-05,
+      "loss": 1.7862,
+      "step": 15450
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 1.5103446245193481,
+      "learning_rate": 3.2644444444444444e-05,
+      "loss": 1.7739,
+      "step": 15475
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 1.4357705116271973,
+      "learning_rate": 3.2607407407407406e-05,
+      "loss": 1.7596,
+      "step": 15500
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 1.739065170288086,
+      "learning_rate": 3.2570370370370374e-05,
+      "loss": 1.8472,
+      "step": 15525
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 1.59120512008667,
+      "learning_rate": 3.253333333333333e-05,
+      "loss": 1.8353,
+      "step": 15550
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 1.7250970602035522,
+      "learning_rate": 3.24962962962963e-05,
+      "loss": 1.7692,
+      "step": 15575
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 1.6642422676086426,
+      "learning_rate": 3.245925925925926e-05,
+      "loss": 1.7367,
+      "step": 15600
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 1.4755923748016357,
+      "learning_rate": 3.242222222222223e-05,
+      "loss": 1.7742,
+      "step": 15625
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 1.5769152641296387,
+      "learning_rate": 3.238518518518518e-05,
+      "loss": 1.7956,
+      "step": 15650
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 1.437385082244873,
+      "learning_rate": 3.234814814814815e-05,
+      "loss": 1.7039,
+      "step": 15675
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 1.4990898370742798,
+      "learning_rate": 3.231111111111111e-05,
+      "loss": 1.7824,
+      "step": 15700
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 1.3943312168121338,
+      "learning_rate": 3.2274074074074074e-05,
+      "loss": 1.7593,
+      "step": 15725
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 1.6160963773727417,
+      "learning_rate": 3.2237037037037035e-05,
+      "loss": 1.7873,
+      "step": 15750
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 1.569349765777588,
+      "learning_rate": 3.2200000000000003e-05,
+      "loss": 1.8264,
+      "step": 15775
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 1.4803683757781982,
+      "learning_rate": 3.216444444444445e-05,
+      "loss": 1.7451,
+      "step": 15800
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 1.760981798171997,
+      "learning_rate": 3.212740740740741e-05,
+      "loss": 1.791,
+      "step": 15825
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 1.3866006135940552,
+      "learning_rate": 3.209037037037037e-05,
+      "loss": 1.7865,
+      "step": 15850
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 1.6246627569198608,
+      "learning_rate": 3.2053333333333334e-05,
+      "loss": 1.8029,
+      "step": 15875
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 1.4544792175292969,
+      "learning_rate": 3.20162962962963e-05,
+      "loss": 1.7549,
+      "step": 15900
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 1.4172948598861694,
+      "learning_rate": 3.197925925925926e-05,
+      "loss": 1.7956,
+      "step": 15925
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 1.4404555559158325,
+      "learning_rate": 3.1942222222222225e-05,
+      "loss": 1.7294,
+      "step": 15950
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 1.4792706966400146,
+      "learning_rate": 3.190518518518519e-05,
+      "loss": 1.7118,
+      "step": 15975
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 1.5171838998794556,
+      "learning_rate": 3.186814814814815e-05,
+      "loss": 1.7392,
+      "step": 16000
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 1.5134046077728271,
+      "learning_rate": 3.183111111111111e-05,
+      "loss": 1.7854,
+      "step": 16025
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 1.504233479499817,
+      "learning_rate": 3.179407407407408e-05,
+      "loss": 1.7571,
+      "step": 16050
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 1.272483468055725,
+      "learning_rate": 3.175703703703704e-05,
+      "loss": 1.7626,
+      "step": 16075
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 1.471863031387329,
+      "learning_rate": 3.172e-05,
+      "loss": 1.7756,
+      "step": 16100
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 1.3983781337738037,
+      "learning_rate": 3.168296296296296e-05,
+      "loss": 1.8251,
+      "step": 16125
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 1.3490697145462036,
+      "learning_rate": 3.164592592592593e-05,
+      "loss": 1.7751,
+      "step": 16150
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 1.5252445936203003,
+      "learning_rate": 3.1608888888888886e-05,
+      "loss": 1.7565,
+      "step": 16175
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 1.688368558883667,
+      "learning_rate": 3.1571851851851854e-05,
+      "loss": 1.8372,
+      "step": 16200
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 1.3352307081222534,
+      "learning_rate": 3.1534814814814816e-05,
+      "loss": 1.7646,
+      "step": 16225
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 1.5939990282058716,
+      "learning_rate": 3.1497777777777784e-05,
+      "loss": 1.7409,
+      "step": 16250
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 1.3315916061401367,
+      "learning_rate": 3.146074074074074e-05,
+      "loss": 1.767,
+      "step": 16275
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 1.367167592048645,
+      "learning_rate": 3.142370370370371e-05,
+      "loss": 1.7998,
+      "step": 16300
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 1.7223109006881714,
+      "learning_rate": 3.138666666666667e-05,
+      "loss": 1.8021,
+      "step": 16325
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 1.4448398351669312,
+      "learning_rate": 3.134962962962963e-05,
+      "loss": 1.8232,
+      "step": 16350
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 1.257012128829956,
+      "learning_rate": 3.131259259259259e-05,
+      "loss": 1.7962,
+      "step": 16375
+    },
+    {
+      "epoch": 6.5600000000000005,
+      "grad_norm": 1.2631241083145142,
+      "learning_rate": 3.127555555555556e-05,
+      "loss": 1.7465,
+      "step": 16400
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 1.5594916343688965,
+      "learning_rate": 3.123851851851852e-05,
+      "loss": 1.7723,
+      "step": 16425
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 1.6267319917678833,
+      "learning_rate": 3.1201481481481483e-05,
+      "loss": 1.8153,
+      "step": 16450
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 1.3760591745376587,
+      "learning_rate": 3.1164444444444445e-05,
+      "loss": 1.7863,
+      "step": 16475
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 1.565403699874878,
+      "learning_rate": 3.112740740740741e-05,
+      "loss": 1.7395,
+      "step": 16500
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 1.562738060951233,
+      "learning_rate": 3.109037037037037e-05,
+      "loss": 1.7866,
+      "step": 16525
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 1.7263091802597046,
+      "learning_rate": 3.1053333333333336e-05,
+      "loss": 1.7532,
+      "step": 16550
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 1.4510164260864258,
+      "learning_rate": 3.10162962962963e-05,
+      "loss": 1.7455,
+      "step": 16575
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 1.474307894706726,
+      "learning_rate": 3.0979259259259266e-05,
+      "loss": 1.7317,
+      "step": 16600
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 1.6146974563598633,
+      "learning_rate": 3.094222222222222e-05,
+      "loss": 1.8102,
+      "step": 16625
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 1.4847644567489624,
+      "learning_rate": 3.090518518518519e-05,
+      "loss": 1.774,
+      "step": 16650
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 1.7083697319030762,
+      "learning_rate": 3.086814814814815e-05,
+      "loss": 1.8375,
+      "step": 16675
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 1.3661868572235107,
+      "learning_rate": 3.083111111111111e-05,
+      "loss": 1.8694,
+      "step": 16700
+    },
+    {
+      "epoch": 6.6899999999999995,
+      "grad_norm": 1.7693723440170288,
+      "learning_rate": 3.0794074074074074e-05,
+      "loss": 1.8175,
+      "step": 16725
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 1.5230166912078857,
+      "learning_rate": 3.075703703703704e-05,
+      "loss": 1.7785,
+      "step": 16750
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 1.4941860437393188,
+      "learning_rate": 3.072e-05,
+      "loss": 1.837,
+      "step": 16775
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 1.4451031684875488,
+      "learning_rate": 3.0682962962962966e-05,
+      "loss": 1.7919,
+      "step": 16800
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 1.6617428064346313,
+      "learning_rate": 3.064592592592593e-05,
+      "loss": 1.8227,
+      "step": 16825
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 1.5577620267868042,
+      "learning_rate": 3.0608888888888895e-05,
+      "loss": 1.7545,
+      "step": 16850
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 1.6712673902511597,
+      "learning_rate": 3.057185185185185e-05,
+      "loss": 1.799,
+      "step": 16875
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 1.879231572151184,
+      "learning_rate": 3.053481481481482e-05,
+      "loss": 1.7737,
+      "step": 16900
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 1.4019575119018555,
+      "learning_rate": 3.049777777777778e-05,
+      "loss": 1.7626,
+      "step": 16925
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 1.3805930614471436,
+      "learning_rate": 3.046074074074074e-05,
+      "loss": 1.7682,
+      "step": 16950
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 1.5736984014511108,
+      "learning_rate": 3.0423703703703703e-05,
+      "loss": 1.8448,
+      "step": 16975
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 1.5094330310821533,
+      "learning_rate": 3.0386666666666668e-05,
+      "loss": 1.8033,
+      "step": 17000
+    },
+    {
+      "epoch": 6.8100000000000005,
+      "grad_norm": 1.7193862199783325,
+      "learning_rate": 3.0349629629629633e-05,
+      "loss": 1.7892,
+      "step": 17025
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 1.7383610010147095,
+      "learning_rate": 3.031259259259259e-05,
+      "loss": 1.7517,
+      "step": 17050
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 1.5743800401687622,
+      "learning_rate": 3.0275555555555556e-05,
+      "loss": 1.8145,
+      "step": 17075
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 1.6040315628051758,
+      "learning_rate": 3.023851851851852e-05,
+      "loss": 1.7887,
+      "step": 17100
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 1.7475184202194214,
+      "learning_rate": 3.020148148148148e-05,
+      "loss": 1.7959,
+      "step": 17125
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 1.4680596590042114,
+      "learning_rate": 3.0164444444444444e-05,
+      "loss": 1.7865,
+      "step": 17150
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 1.5900845527648926,
+      "learning_rate": 3.012740740740741e-05,
+      "loss": 1.8171,
+      "step": 17175
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 1.5023775100708008,
+      "learning_rate": 3.0090370370370374e-05,
+      "loss": 1.7596,
+      "step": 17200
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 1.5605974197387695,
+      "learning_rate": 3.0053333333333332e-05,
+      "loss": 1.7812,
+      "step": 17225
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 1.612432837486267,
+      "learning_rate": 3.0016296296296297e-05,
+      "loss": 1.8062,
+      "step": 17250
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 1.664870262145996,
+      "learning_rate": 2.9979259259259262e-05,
+      "loss": 1.7445,
+      "step": 17275
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 1.4798489809036255,
+      "learning_rate": 2.9943703703703708e-05,
+      "loss": 1.7788,
+      "step": 17300
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 1.5232810974121094,
+      "learning_rate": 2.9906666666666666e-05,
+      "loss": 1.8383,
+      "step": 17325
+    },
+    {
+      "epoch": 6.9399999999999995,
+      "grad_norm": 1.4982616901397705,
+      "learning_rate": 2.986962962962963e-05,
+      "loss": 1.7302,
+      "step": 17350
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 1.4312171936035156,
+      "learning_rate": 2.9832592592592596e-05,
+      "loss": 1.7573,
+      "step": 17375
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 1.3876653909683228,
+      "learning_rate": 2.9795555555555554e-05,
+      "loss": 1.7684,
+      "step": 17400
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 1.6542433500289917,
+      "learning_rate": 2.975851851851852e-05,
+      "loss": 1.7529,
+      "step": 17425
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 1.4732959270477295,
+      "learning_rate": 2.9721481481481484e-05,
+      "loss": 1.8128,
+      "step": 17450
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 1.4124960899353027,
+      "learning_rate": 2.968444444444445e-05,
+      "loss": 1.8079,
+      "step": 17475
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 1.466747760772705,
+      "learning_rate": 2.9647407407407407e-05,
+      "loss": 1.8214,
+      "step": 17500
+    },
+    {
+      "epoch": 7.0,
+      "eval_gen_len": 13.125,
+      "eval_loss": 1.9262617826461792,
+      "eval_rouge1": 51.5538,
+      "eval_rouge2": 25.7728,
+      "eval_rougeL": 47.3812,
+      "eval_rougeLsum": 47.3727,
+      "eval_runtime": 196.7996,
+      "eval_samples_per_second": 101.626,
+      "eval_steps_per_second": 1.59,
+      "step": 17500
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 37500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.01
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5719030035841024e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}