diff --git "a/checkpoint-7000/trainer_state.json" "b/checkpoint-7000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-7000/trainer_state.json"
@@ -0,0 +1,5014 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 8.0,
+  "eval_steps": 500,
+  "global_step": 7000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.011428571428571429,
+      "grad_norm": 2.3897457122802734,
+      "learning_rate": 4.9935714285714285e-05,
+      "loss": 5.5799,
+      "step": 10
+    },
+    {
+      "epoch": 0.022857142857142857,
+      "grad_norm": 1.2946193218231201,
+      "learning_rate": 4.986428571428572e-05,
+      "loss": 2.9562,
+      "step": 20
+    },
+    {
+      "epoch": 0.03428571428571429,
+      "grad_norm": 1.235308051109314,
+      "learning_rate": 4.9792857142857146e-05,
+      "loss": 2.2701,
+      "step": 30
+    },
+    {
+      "epoch": 0.045714285714285714,
+      "grad_norm": 1.25638747215271,
+      "learning_rate": 4.972142857142858e-05,
+      "loss": 1.6642,
+      "step": 40
+    },
+    {
+      "epoch": 0.05714285714285714,
+      "grad_norm": 1.275929570198059,
+      "learning_rate": 4.965e-05,
+      "loss": 1.2349,
+      "step": 50
+    },
+    {
+      "epoch": 0.06857142857142857,
+      "grad_norm": 1.2436413764953613,
+      "learning_rate": 4.957857142857143e-05,
+      "loss": 0.9634,
+      "step": 60
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1808298826217651,
+      "learning_rate": 4.9507142857142855e-05,
+      "loss": 0.7025,
+      "step": 70
+    },
+    {
+      "epoch": 0.09142857142857143,
+      "grad_norm": 1.0625460147857666,
+      "learning_rate": 4.943571428571429e-05,
+      "loss": 0.5385,
+      "step": 80
+    },
+    {
+      "epoch": 0.10285714285714286,
+      "grad_norm": 0.8773331046104431,
+      "learning_rate": 4.936428571428572e-05,
+      "loss": 0.418,
+      "step": 90
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 0.8841960430145264,
+      "learning_rate": 4.929285714285715e-05,
+      "loss": 0.3878,
+      "step": 100
+    },
+    {
+      "epoch": 0.12571428571428572,
+      "grad_norm": 0.7253878116607666,
+      "learning_rate": 4.922142857142857e-05,
+      "loss": 0.3189,
+      "step": 110
+    },
+    {
+      "epoch": 0.13714285714285715,
+      "grad_norm": 0.6557711958885193,
+      "learning_rate": 4.915e-05,
+      "loss": 0.2695,
+      "step": 120
+    },
+    {
+      "epoch": 0.14857142857142858,
+      "grad_norm": 0.6099198460578918,
+      "learning_rate": 4.9078571428571426e-05,
+      "loss": 0.2424,
+      "step": 130
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6418436169624329,
+      "learning_rate": 4.900714285714286e-05,
+      "loss": 0.2242,
+      "step": 140
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 0.6211830973625183,
+      "learning_rate": 4.893571428571429e-05,
+      "loss": 0.1925,
+      "step": 150
+    },
+    {
+      "epoch": 0.18285714285714286,
+      "grad_norm": 0.5805016756057739,
+      "learning_rate": 4.886428571428572e-05,
+      "loss": 0.1856,
+      "step": 160
+    },
+    {
+      "epoch": 0.19428571428571428,
+      "grad_norm": 0.5227402448654175,
+      "learning_rate": 4.879285714285715e-05,
+      "loss": 0.1676,
+      "step": 170
+    },
+    {
+      "epoch": 0.2057142857142857,
+      "grad_norm": 0.5591551661491394,
+      "learning_rate": 4.872142857142857e-05,
+      "loss": 0.1706,
+      "step": 180
+    },
+    {
+      "epoch": 0.21714285714285714,
+      "grad_norm": 0.47619232535362244,
+      "learning_rate": 4.8650000000000003e-05,
+      "loss": 0.1589,
+      "step": 190
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 0.48772159218788147,
+      "learning_rate": 4.857857142857143e-05,
+      "loss": 0.1654,
+      "step": 200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48056119680404663,
+      "learning_rate": 4.8507142857142865e-05,
+      "loss": 0.1501,
+      "step": 210
+    },
+    {
+      "epoch": 0.25142857142857145,
+      "grad_norm": 0.4571131467819214,
+      "learning_rate": 4.843571428571429e-05,
+      "loss": 0.1413,
+      "step": 220
+    },
+    {
+      "epoch": 0.26285714285714284,
+      "grad_norm": 0.5747288465499878,
+      "learning_rate": 4.836428571428572e-05,
+      "loss": 0.1258,
+      "step": 230
+    },
+    {
+      "epoch": 0.2742857142857143,
+      "grad_norm": 0.4191080629825592,
+      "learning_rate": 4.8292857142857143e-05,
+      "loss": 0.1182,
+      "step": 240
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.4493291974067688,
+      "learning_rate": 4.8221428571428574e-05,
+      "loss": 0.1258,
+      "step": 250
+    },
+    {
+      "epoch": 0.29714285714285715,
+      "grad_norm": 0.3831436038017273,
+      "learning_rate": 4.815e-05,
+      "loss": 0.1273,
+      "step": 260
+    },
+    {
+      "epoch": 0.30857142857142855,
+      "grad_norm": 0.3865627944469452,
+      "learning_rate": 4.807857142857143e-05,
+      "loss": 0.1119,
+      "step": 270
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42649543285369873,
+      "learning_rate": 4.800714285714286e-05,
+      "loss": 0.1166,
+      "step": 280
+    },
+    {
+      "epoch": 0.3314285714285714,
+      "grad_norm": 0.4094502925872803,
+      "learning_rate": 4.793571428571429e-05,
+      "loss": 0.1142,
+      "step": 290
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 0.3270905315876007,
+      "learning_rate": 4.7864285714285714e-05,
+      "loss": 0.1123,
+      "step": 300
+    },
+    {
+      "epoch": 0.35428571428571426,
+      "grad_norm": 0.35536834597587585,
+      "learning_rate": 4.7792857142857145e-05,
+      "loss": 0.1003,
+      "step": 310
+    },
+    {
+      "epoch": 0.3657142857142857,
+      "grad_norm": 0.41495510935783386,
+      "learning_rate": 4.7721428571428576e-05,
+      "loss": 0.099,
+      "step": 320
+    },
+    {
+      "epoch": 0.37714285714285717,
+      "grad_norm": 0.470359206199646,
+      "learning_rate": 4.765e-05,
+      "loss": 0.0969,
+      "step": 330
+    },
+    {
+      "epoch": 0.38857142857142857,
+      "grad_norm": 0.4018060266971588,
+      "learning_rate": 4.757857142857143e-05,
+      "loss": 0.0955,
+      "step": 340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34157299995422363,
+      "learning_rate": 4.750714285714286e-05,
+      "loss": 0.0979,
+      "step": 350
+    },
+    {
+      "epoch": 0.4114285714285714,
+      "grad_norm": 0.31011414527893066,
+      "learning_rate": 4.743571428571429e-05,
+      "loss": 0.0869,
+      "step": 360
+    },
+    {
+      "epoch": 0.4228571428571429,
+      "grad_norm": 0.26191529631614685,
+      "learning_rate": 4.7364285714285716e-05,
+      "loss": 0.0846,
+      "step": 370
+    },
+    {
+      "epoch": 0.4342857142857143,
+      "grad_norm": 0.30978405475616455,
+      "learning_rate": 4.7292857142857146e-05,
+      "loss": 0.087,
+      "step": 380
+    },
+    {
+      "epoch": 0.44571428571428573,
+      "grad_norm": 0.38050106167793274,
+      "learning_rate": 4.722142857142857e-05,
+      "loss": 0.0786,
+      "step": 390
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.24444682896137238,
+      "learning_rate": 4.715e-05,
+      "loss": 0.0811,
+      "step": 400
+    },
+    {
+      "epoch": 0.4685714285714286,
+      "grad_norm": 0.31481674313545227,
+      "learning_rate": 4.707857142857143e-05,
+      "loss": 0.0818,
+      "step": 410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3219773471355438,
+      "learning_rate": 4.700714285714286e-05,
+      "loss": 0.0775,
+      "step": 420
+    },
+    {
+      "epoch": 0.49142857142857144,
+      "grad_norm": 0.2883884310722351,
+      "learning_rate": 4.6935714285714286e-05,
+      "loss": 0.0697,
+      "step": 430
+    },
+    {
+      "epoch": 0.5028571428571429,
+      "grad_norm": 0.33833104372024536,
+      "learning_rate": 4.686428571428572e-05,
+      "loss": 0.0797,
+      "step": 440
+    },
+    {
+      "epoch": 0.5142857142857142,
+      "grad_norm": 0.23762065172195435,
+      "learning_rate": 4.679285714285714e-05,
+      "loss": 0.0755,
+      "step": 450
+    },
+    {
+      "epoch": 0.5257142857142857,
+      "grad_norm": 0.32968372106552124,
+      "learning_rate": 4.672142857142857e-05,
+      "loss": 0.0795,
+      "step": 460
+    },
+    {
+      "epoch": 0.5371428571428571,
+      "grad_norm": 0.29430899024009705,
+      "learning_rate": 4.665e-05,
+      "loss": 0.0705,
+      "step": 470
+    },
+    {
+      "epoch": 0.5485714285714286,
+      "grad_norm": 0.3451765179634094,
+      "learning_rate": 4.657857142857143e-05,
+      "loss": 0.067,
+      "step": 480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30210795998573303,
+      "learning_rate": 4.650714285714286e-05,
+      "loss": 0.0701,
+      "step": 490
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.2613590359687805,
+      "learning_rate": 4.643571428571429e-05,
+      "loss": 0.0649,
+      "step": 500
+    },
+    {
+      "epoch": 0.5828571428571429,
+      "grad_norm": 0.43155479431152344,
+      "learning_rate": 4.636428571428572e-05,
+      "loss": 0.0648,
+      "step": 510
+    },
+    {
+      "epoch": 0.5942857142857143,
+      "grad_norm": 0.38308286666870117,
+      "learning_rate": 4.629285714285714e-05,
+      "loss": 0.0584,
+      "step": 520
+    },
+    {
+      "epoch": 0.6057142857142858,
+      "grad_norm": 0.2277926206588745,
+      "learning_rate": 4.622142857142857e-05,
+      "loss": 0.0708,
+      "step": 530
+    },
+    {
+      "epoch": 0.6171428571428571,
+      "grad_norm": 0.21378019452095032,
+      "learning_rate": 4.6150000000000004e-05,
+      "loss": 0.0624,
+      "step": 540
+    },
+    {
+      "epoch": 0.6285714285714286,
+      "grad_norm": 0.23331303894519806,
+      "learning_rate": 4.6078571428571434e-05,
+      "loss": 0.062,
+      "step": 550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2659645080566406,
+      "learning_rate": 4.600714285714286e-05,
+      "loss": 0.0658,
+      "step": 560
+    },
+    {
+      "epoch": 0.6514285714285715,
+      "grad_norm": 0.30504903197288513,
+      "learning_rate": 4.593571428571429e-05,
+      "loss": 0.0605,
+      "step": 570
+    },
+    {
+      "epoch": 0.6628571428571428,
+      "grad_norm": 0.3272831439971924,
+      "learning_rate": 4.586428571428571e-05,
+      "loss": 0.0667,
+      "step": 580
+    },
+    {
+      "epoch": 0.6742857142857143,
+      "grad_norm": 0.22067314386367798,
+      "learning_rate": 4.5792857142857144e-05,
+      "loss": 0.0649,
+      "step": 590
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 0.3015865087509155,
+      "learning_rate": 4.5721428571428574e-05,
+      "loss": 0.0639,
+      "step": 600
+    },
+    {
+      "epoch": 0.6971428571428572,
+      "grad_norm": 0.2650837302207947,
+      "learning_rate": 4.5650000000000005e-05,
+      "loss": 0.0614,
+      "step": 610
+    },
+    {
+      "epoch": 0.7085714285714285,
+      "grad_norm": 0.2098853588104248,
+      "learning_rate": 4.557857142857143e-05,
+      "loss": 0.0622,
+      "step": 620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.21571612358093262,
+      "learning_rate": 4.550714285714286e-05,
+      "loss": 0.0557,
+      "step": 630
+    },
+    {
+      "epoch": 0.7314285714285714,
+      "grad_norm": 0.23635227978229523,
+      "learning_rate": 4.5435714285714284e-05,
+      "loss": 0.0566,
+      "step": 640
+    },
+    {
+      "epoch": 0.7428571428571429,
+      "grad_norm": 0.20776158571243286,
+      "learning_rate": 4.5364285714285714e-05,
+      "loss": 0.06,
+      "step": 650
+    },
+    {
+      "epoch": 0.7542857142857143,
+      "grad_norm": 0.2526012063026428,
+      "learning_rate": 4.5292857142857145e-05,
+      "loss": 0.0628,
+      "step": 660
+    },
+    {
+      "epoch": 0.7657142857142857,
+      "grad_norm": 0.18527157604694366,
+      "learning_rate": 4.5221428571428576e-05,
+      "loss": 0.0582,
+      "step": 670
+    },
+    {
+      "epoch": 0.7771428571428571,
+      "grad_norm": 0.2718740701675415,
+      "learning_rate": 4.5150000000000006e-05,
+      "loss": 0.0582,
+      "step": 680
+    },
+    {
+      "epoch": 0.7885714285714286,
+      "grad_norm": 0.19608113169670105,
+      "learning_rate": 4.507857142857143e-05,
+      "loss": 0.0589,
+      "step": 690
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2381214052438736,
+      "learning_rate": 4.500714285714286e-05,
+      "loss": 0.0528,
+      "step": 700
+    },
+    {
+      "epoch": 0.8114285714285714,
+      "grad_norm": 0.22083961963653564,
+      "learning_rate": 4.4935714285714285e-05,
+      "loss": 0.0619,
+      "step": 710
+    },
+    {
+      "epoch": 0.8228571428571428,
+      "grad_norm": 0.253137469291687,
+      "learning_rate": 4.4864285714285716e-05,
+      "loss": 0.0589,
+      "step": 720
+    },
+    {
+      "epoch": 0.8342857142857143,
+      "grad_norm": 0.24962878227233887,
+      "learning_rate": 4.4792857142857146e-05,
+      "loss": 0.0599,
+      "step": 730
+    },
+    {
+      "epoch": 0.8457142857142858,
+      "grad_norm": 0.2471620887517929,
+      "learning_rate": 4.472142857142858e-05,
+      "loss": 0.0607,
+      "step": 740
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.22191166877746582,
+      "learning_rate": 4.465e-05,
+      "loss": 0.0563,
+      "step": 750
+    },
+    {
+      "epoch": 0.8685714285714285,
+      "grad_norm": 0.24587008357048035,
+      "learning_rate": 4.457857142857143e-05,
+      "loss": 0.0552,
+      "step": 760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2442178726196289,
+      "learning_rate": 4.4507142857142856e-05,
+      "loss": 0.0573,
+      "step": 770
+    },
+    {
+      "epoch": 0.8914285714285715,
+      "grad_norm": 0.17938025295734406,
+      "learning_rate": 4.4435714285714286e-05,
+      "loss": 0.0581,
+      "step": 780
+    },
+    {
+      "epoch": 0.9028571428571428,
+      "grad_norm": 0.20533467829227448,
+      "learning_rate": 4.436428571428572e-05,
+      "loss": 0.0551,
+      "step": 790
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 0.18826663494110107,
+      "learning_rate": 4.429285714285715e-05,
+      "loss": 0.0585,
+      "step": 800
+    },
+    {
+      "epoch": 0.9257142857142857,
+      "grad_norm": 0.2727915346622467,
+      "learning_rate": 4.422142857142857e-05,
+      "loss": 0.0569,
+      "step": 810
+    },
+    {
+      "epoch": 0.9371428571428572,
+      "grad_norm": 0.300370454788208,
+      "learning_rate": 4.415e-05,
+      "loss": 0.0549,
+      "step": 820
+    },
+    {
+      "epoch": 0.9485714285714286,
+      "grad_norm": 0.3136334717273712,
+      "learning_rate": 4.407857142857143e-05,
+      "loss": 0.0551,
+      "step": 830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25469839572906494,
+      "learning_rate": 4.400714285714286e-05,
+      "loss": 0.0499,
+      "step": 840
+    },
+    {
+      "epoch": 0.9714285714285714,
+      "grad_norm": 0.17637017369270325,
+      "learning_rate": 4.393571428571429e-05,
+      "loss": 0.0544,
+      "step": 850
+    },
+    {
+      "epoch": 0.9828571428571429,
+      "grad_norm": 0.2047278881072998,
+      "learning_rate": 4.386428571428572e-05,
+      "loss": 0.0534,
+      "step": 860
+    },
+    {
+      "epoch": 0.9942857142857143,
+      "grad_norm": 0.19194772839546204,
+      "learning_rate": 4.379285714285715e-05,
+      "loss": 0.0499,
+      "step": 870
+    },
+    {
+      "epoch": 1.0,
+      "eval_bleu": 0.07,
+      "eval_exact_match": 0.0,
+      "eval_loss": 0.039771415293216705,
+      "eval_runtime": 343.2775,
+      "eval_samples_per_second": 17.479,
+      "eval_steps_per_second": 1.092,
+      "step": 875
+    },
+    {
+      "epoch": 1.0057142857142858,
+      "grad_norm": 0.25923219323158264,
+      "learning_rate": 4.372142857142857e-05,
+      "loss": 0.0535,
+      "step": 880
+    },
+    {
+      "epoch": 1.0171428571428571,
+      "grad_norm": 0.19547821581363678,
+      "learning_rate": 4.3650000000000004e-05,
+      "loss": 0.0516,
+      "step": 890
+    },
+    {
+      "epoch": 1.0285714285714285,
+      "grad_norm": 0.16015152633190155,
+      "learning_rate": 4.357857142857143e-05,
+      "loss": 0.0502,
+      "step": 900
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19731080532073975,
+      "learning_rate": 4.350714285714286e-05,
+      "loss": 0.0482,
+      "step": 910
+    },
+    {
+      "epoch": 1.0514285714285714,
+      "grad_norm": 0.17309032380580902,
+      "learning_rate": 4.343571428571428e-05,
+      "loss": 0.0461,
+      "step": 920
+    },
+    {
+      "epoch": 1.062857142857143,
+      "grad_norm": 0.17293015122413635,
+      "learning_rate": 4.336428571428572e-05,
+      "loss": 0.0521,
+      "step": 930
+    },
+    {
+      "epoch": 1.0742857142857143,
+      "grad_norm": 0.1969359666109085,
+      "learning_rate": 4.3292857142857144e-05,
+      "loss": 0.0499,
+      "step": 940
+    },
+    {
+      "epoch": 1.0857142857142856,
+      "grad_norm": 0.14096097648143768,
+      "learning_rate": 4.3221428571428575e-05,
+      "loss": 0.0559,
+      "step": 950
+    },
+    {
+      "epoch": 1.0971428571428572,
+      "grad_norm": 0.24003499746322632,
+      "learning_rate": 4.315e-05,
+      "loss": 0.0541,
+      "step": 960
+    },
+    {
+      "epoch": 1.1085714285714285,
+      "grad_norm": 0.1646154224872589,
+      "learning_rate": 4.307857142857143e-05,
+      "loss": 0.0554,
+      "step": 970
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.15991230309009552,
+      "learning_rate": 4.300714285714286e-05,
+      "loss": 0.052,
+      "step": 980
+    },
+    {
+      "epoch": 1.1314285714285715,
+      "grad_norm": 0.1888715773820877,
+      "learning_rate": 4.293571428571429e-05,
+      "loss": 0.0478,
+      "step": 990
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.15449759364128113,
+      "learning_rate": 4.2864285714285715e-05,
+      "loss": 0.0496,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1542857142857144,
+      "grad_norm": 0.12944145500659943,
+      "learning_rate": 4.2792857142857145e-05,
+      "loss": 0.0482,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1657142857142857,
+      "grad_norm": 0.14871393144130707,
+      "learning_rate": 4.2721428571428576e-05,
+      "loss": 0.0512,
+      "step": 1020
+    },
+    {
+      "epoch": 1.177142857142857,
+      "grad_norm": 0.17820103466510773,
+      "learning_rate": 4.265e-05,
+      "loss": 0.0475,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1885714285714286,
+      "grad_norm": 0.1732417643070221,
+      "learning_rate": 4.257857142857143e-05,
+      "loss": 0.0503,
+      "step": 1040
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19619877636432648,
+      "learning_rate": 4.2507142857142855e-05,
+      "loss": 0.0481,
+      "step": 1050
+    },
+    {
+      "epoch": 1.2114285714285715,
+      "grad_norm": 0.35855549573898315,
+      "learning_rate": 4.243571428571429e-05,
+      "loss": 0.0525,
+      "step": 1060
+    },
+    {
+      "epoch": 1.2228571428571429,
+      "grad_norm": 0.22570298612117767,
+      "learning_rate": 4.2364285714285716e-05,
+      "loss": 0.0507,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2342857142857142,
+      "grad_norm": 0.26217854022979736,
+      "learning_rate": 4.229285714285715e-05,
+      "loss": 0.0506,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2457142857142858,
+      "grad_norm": 0.19021452963352203,
+      "learning_rate": 4.222142857142857e-05,
+      "loss": 0.0487,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2571428571428571,
+      "grad_norm": 0.2223762720823288,
+      "learning_rate": 4.215e-05,
+      "loss": 0.0491,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2685714285714287,
+      "grad_norm": 0.13785573840141296,
+      "learning_rate": 4.2078571428571425e-05,
+      "loss": 0.0462,
+      "step": 1110
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.15734604001045227,
+      "learning_rate": 4.200714285714286e-05,
+      "loss": 0.0448,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2914285714285714,
+      "grad_norm": 0.5288769006729126,
+      "learning_rate": 4.193571428571429e-05,
+      "loss": 0.0501,
+      "step": 1130
+    },
+    {
+      "epoch": 1.302857142857143,
+      "grad_norm": 0.17690952122211456,
+      "learning_rate": 4.186428571428572e-05,
+      "loss": 0.0453,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3142857142857143,
+      "grad_norm": 0.17811287939548492,
+      "learning_rate": 4.179285714285715e-05,
+      "loss": 0.0478,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3257142857142856,
+      "grad_norm": 0.19255459308624268,
+      "learning_rate": 4.172142857142857e-05,
+      "loss": 0.0477,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3371428571428572,
+      "grad_norm": 0.16269567608833313,
+      "learning_rate": 4.165e-05,
+      "loss": 0.0438,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3485714285714285,
+      "grad_norm": 0.16998130083084106,
+      "learning_rate": 4.157857142857143e-05,
+      "loss": 0.046,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.21180035173892975,
+      "learning_rate": 4.1507142857142864e-05,
+      "loss": 0.0443,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3714285714285714,
+      "grad_norm": 0.24145734310150146,
+      "learning_rate": 4.143571428571429e-05,
+      "loss": 0.0438,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3828571428571428,
+      "grad_norm": 0.13968069851398468,
+      "learning_rate": 4.136428571428572e-05,
+      "loss": 0.0471,
+      "step": 1210
+    },
+    {
+      "epoch": 1.3942857142857144,
+      "grad_norm": 0.16085991263389587,
+      "learning_rate": 4.129285714285714e-05,
+      "loss": 0.0472,
+      "step": 1220
+    },
+    {
+      "epoch": 1.4057142857142857,
+      "grad_norm": 0.3225364685058594,
+      "learning_rate": 4.1221428571428573e-05,
+      "loss": 0.044,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4171428571428573,
+      "grad_norm": 0.11656186729669571,
+      "learning_rate": 4.115e-05,
+      "loss": 0.0447,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.16806724667549133,
+      "learning_rate": 4.1078571428571435e-05,
+      "loss": 0.0441,
+      "step": 1250
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.22123603522777557,
+      "learning_rate": 4.100714285714286e-05,
+      "loss": 0.043,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4514285714285715,
+      "grad_norm": 0.14784376323223114,
+      "learning_rate": 4.093571428571429e-05,
+      "loss": 0.0439,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4628571428571429,
+      "grad_norm": 0.1381123811006546,
+      "learning_rate": 4.0864285714285713e-05,
+      "loss": 0.0409,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4742857142857142,
+      "grad_norm": 0.14749744534492493,
+      "learning_rate": 4.0792857142857144e-05,
+      "loss": 0.0412,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4857142857142858,
+      "grad_norm": 0.16648408770561218,
+      "learning_rate": 4.0721428571428575e-05,
+      "loss": 0.0429,
+      "step": 1300
+    },
+    {
+      "epoch": 1.497142857142857,
+      "grad_norm": 0.11807863414287567,
+      "learning_rate": 4.065e-05,
+      "loss": 0.0415,
+      "step": 1310
+    },
+    {
+      "epoch": 1.5085714285714285,
+      "grad_norm": 0.1766592413187027,
+      "learning_rate": 4.057857142857143e-05,
+      "loss": 0.0424,
+      "step": 1320
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.17224718630313873,
+      "learning_rate": 4.050714285714286e-05,
+      "loss": 0.0443,
+      "step": 1330
+    },
+    {
+      "epoch": 1.5314285714285716,
+      "grad_norm": 0.25334063172340393,
+      "learning_rate": 4.043571428571429e-05,
+      "loss": 0.0444,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5428571428571427,
+      "grad_norm": 0.1791314333677292,
+      "learning_rate": 4.0364285714285715e-05,
+      "loss": 0.0429,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5542857142857143,
+      "grad_norm": 0.1268022358417511,
+      "learning_rate": 4.0292857142857146e-05,
+      "loss": 0.0429,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5657142857142858,
+      "grad_norm": 0.1954212486743927,
+      "learning_rate": 4.022142857142857e-05,
+      "loss": 0.0455,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5771428571428572,
+      "grad_norm": 0.17540448904037476,
+      "learning_rate": 4.015000000000001e-05,
+      "loss": 0.0459,
+      "step": 1380
+    },
+    {
+      "epoch": 1.5885714285714285,
+      "grad_norm": 0.12704719603061676,
+      "learning_rate": 4.007857142857143e-05,
+      "loss": 0.0429,
+      "step": 1390
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2572875916957855,
+      "learning_rate": 4.000714285714286e-05,
+      "loss": 0.0412,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6114285714285714,
+      "grad_norm": 0.2871057689189911,
+      "learning_rate": 3.9935714285714285e-05,
+      "loss": 0.0424,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6228571428571428,
+      "grad_norm": 0.14216271042823792,
+      "learning_rate": 3.9864285714285716e-05,
+      "loss": 0.041,
+      "step": 1420
+    },
+    {
+      "epoch": 1.6342857142857143,
+      "grad_norm": 0.2252836674451828,
+      "learning_rate": 3.979285714285714e-05,
+      "loss": 0.0421,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6457142857142857,
+      "grad_norm": 0.24771438539028168,
+      "learning_rate": 3.972142857142858e-05,
+      "loss": 0.0434,
+      "step": 1440
+    },
+    {
+      "epoch": 1.657142857142857,
+      "grad_norm": 0.15339438617229462,
+      "learning_rate": 3.965e-05,
+      "loss": 0.0452,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6685714285714286,
+      "grad_norm": 0.13443227112293243,
+      "learning_rate": 3.957857142857143e-05,
+      "loss": 0.0401,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.15537743270397186,
+      "learning_rate": 3.9507142857142856e-05,
+      "loss": 0.0414,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6914285714285713,
+      "grad_norm": 0.14956708252429962,
+      "learning_rate": 3.943571428571429e-05,
+      "loss": 0.0423,
+      "step": 1480
+    },
+    {
+      "epoch": 1.7028571428571428,
+      "grad_norm": 0.13108207285404205,
+      "learning_rate": 3.936428571428572e-05,
+      "loss": 0.0385,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.2588536739349365,
+      "learning_rate": 3.929285714285714e-05,
+      "loss": 0.0449,
+      "step": 1500
+    },
+    {
+      "epoch": 1.7257142857142858,
+      "grad_norm": 0.13190333545207977,
+      "learning_rate": 3.922142857142858e-05,
+      "loss": 0.0415,
+      "step": 1510
+    },
+    {
+      "epoch": 1.737142857142857,
+      "grad_norm": 0.15415117144584656,
+      "learning_rate": 3.915e-05,
+      "loss": 0.0404,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7485714285714287,
+      "grad_norm": 0.22507691383361816,
+      "learning_rate": 3.9078571428571434e-05,
+      "loss": 0.0412,
+      "step": 1530
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.14862731099128723,
+      "learning_rate": 3.900714285714286e-05,
+      "loss": 0.0412,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7714285714285714,
+      "grad_norm": 0.19868846237659454,
+      "learning_rate": 3.893571428571429e-05,
+      "loss": 0.0413,
+      "step": 1550
+    },
+    {
+      "epoch": 1.782857142857143,
+      "grad_norm": 0.11937547475099564,
+      "learning_rate": 3.886428571428571e-05,
+      "loss": 0.0416,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7942857142857143,
+      "grad_norm": 0.17745395004749298,
+      "learning_rate": 3.879285714285715e-05,
+      "loss": 0.0412,
+      "step": 1570
+    },
+    {
+      "epoch": 1.8057142857142856,
+      "grad_norm": 0.11944162100553513,
+      "learning_rate": 3.8721428571428574e-05,
+      "loss": 0.0382,
+      "step": 1580
+    },
+    {
+      "epoch": 1.8171428571428572,
+      "grad_norm": 0.3294443190097809,
+      "learning_rate": 3.8650000000000004e-05,
+      "loss": 0.0409,
+      "step": 1590
+    },
+    {
+      "epoch": 1.8285714285714287,
+      "grad_norm": 0.19752420485019684,
+      "learning_rate": 3.857857142857143e-05,
+      "loss": 0.0386,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.18311238288879395,
+      "learning_rate": 3.850714285714286e-05,
+      "loss": 0.04,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8514285714285714,
+      "grad_norm": 0.12520070374011993,
+      "learning_rate": 3.843571428571428e-05,
+      "loss": 0.0393,
+      "step": 1620
+    },
+    {
+      "epoch": 1.862857142857143,
+      "grad_norm": 0.15205688774585724,
+      "learning_rate": 3.8364285714285714e-05,
+      "loss": 0.0385,
+      "step": 1630
+    },
+    {
+      "epoch": 1.8742857142857143,
+      "grad_norm": 0.22413566708564758,
+      "learning_rate": 3.8292857142857144e-05,
+      "loss": 0.0394,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8857142857142857,
+      "grad_norm": 0.17256265878677368,
+      "learning_rate": 3.8221428571428575e-05,
+      "loss": 0.0364,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8971428571428572,
+      "grad_norm": 0.20736004412174225,
+      "learning_rate": 3.8150000000000006e-05,
+      "loss": 0.039,
+      "step": 1660
+    },
+    {
+      "epoch": 1.9085714285714286,
+      "grad_norm": 0.13663062453269958,
+      "learning_rate": 3.807857142857143e-05,
+      "loss": 0.0387,
+      "step": 1670
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.14069513976573944,
+      "learning_rate": 3.800714285714286e-05,
+      "loss": 0.0381,
+      "step": 1680
+    },
+    {
+      "epoch": 1.9314285714285715,
+      "grad_norm": 0.11960676312446594,
+      "learning_rate": 3.7935714285714284e-05,
+      "loss": 0.0383,
+      "step": 1690
+    },
+    {
+      "epoch": 1.9428571428571428,
+      "grad_norm": 0.14744821190834045,
+      "learning_rate": 3.786428571428572e-05,
+      "loss": 0.0375,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9542857142857142,
+      "grad_norm": 0.10939270257949829,
+      "learning_rate": 3.7792857142857146e-05,
+      "loss": 0.0393,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9657142857142857,
+      "grad_norm": 0.1914082169532776,
+      "learning_rate": 3.7721428571428576e-05,
+      "loss": 0.0378,
+      "step": 1720
+    },
+    {
+      "epoch": 1.977142857142857,
+      "grad_norm": 0.12824489176273346,
+      "learning_rate": 3.765e-05,
+      "loss": 0.0375,
+      "step": 1730
+    },
+    {
+      "epoch": 1.9885714285714284,
+      "grad_norm": 0.1403360813856125,
+      "learning_rate": 3.757857142857143e-05,
+      "loss": 0.0375,
+      "step": 1740
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.12333709001541138,
+      "learning_rate": 3.7507142857142855e-05,
+      "loss": 0.0349,
+      "step": 1750
+    },
+    {
+      "epoch": 2.0,
+      "eval_bleu": 0.07,
+      "eval_exact_match": 0.0,
+      "eval_loss": 0.030519695952534676,
+      "eval_runtime": 343.7355,
+      "eval_samples_per_second": 17.455,
+      "eval_steps_per_second": 1.091,
+      "step": 1750
+    },
+    {
+      "epoch": 2.0114285714285716,
+      "grad_norm": 0.20832888782024384,
+      "learning_rate": 3.7435714285714286e-05,
+      "loss": 0.0387,
+      "step": 1760
+    },
+    {
+      "epoch": 2.0228571428571427,
+      "grad_norm": 0.2904718518257141,
+      "learning_rate": 3.7364285714285716e-05,
+      "loss": 0.0406,
+      "step": 1770
+    },
+    {
+      "epoch": 2.0342857142857143,
+      "grad_norm": 0.18017873167991638,
+      "learning_rate": 3.729285714285715e-05,
+      "loss": 0.0382,
+      "step": 1780
+    },
+    {
+      "epoch": 2.045714285714286,
+      "grad_norm": 0.17356900870800018,
+      "learning_rate": 3.722142857142857e-05,
+      "loss": 0.0372,
+      "step": 1790
+    },
+    {
+      "epoch": 2.057142857142857,
+      "grad_norm": 0.11434105038642883,
+      "learning_rate": 3.715e-05,
+      "loss": 0.039,
+      "step": 1800
+    },
+    {
+      "epoch": 2.0685714285714285,
+      "grad_norm": 0.14525334537029266,
+      "learning_rate": 3.707857142857143e-05,
+      "loss": 0.0389,
+      "step": 1810
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.12170791625976562,
+      "learning_rate": 3.7007142857142856e-05,
+      "loss": 0.0362,
+      "step": 1820
+    },
+    {
+      "epoch": 2.0914285714285716,
+      "grad_norm": 0.12874464690685272,
+      "learning_rate": 3.693571428571429e-05,
+      "loss": 0.0377,
+      "step": 1830
+    },
+    {
+      "epoch": 2.1028571428571428,
+      "grad_norm": 0.13844266533851624,
+      "learning_rate": 3.686428571428572e-05,
+      "loss": 0.0387,
+      "step": 1840
+    },
+    {
+      "epoch": 2.1142857142857143,
+      "grad_norm": 0.12960191071033478,
+      "learning_rate": 3.679285714285715e-05,
+      "loss": 0.0384,
+      "step": 1850
+    },
+    {
+      "epoch": 2.125714285714286,
+      "grad_norm": 0.15162068605422974,
+      "learning_rate": 3.672142857142857e-05,
+      "loss": 0.0375,
+      "step": 1860
+    },
+    {
+      "epoch": 2.137142857142857,
+      "grad_norm": 0.2595318555831909,
+      "learning_rate": 3.665e-05,
+      "loss": 0.0389,
+      "step": 1870
+    },
+    {
+      "epoch": 2.1485714285714286,
+      "grad_norm": 0.19488683342933655,
+      "learning_rate": 3.657857142857143e-05,
+      "loss": 0.0405,
+      "step": 1880
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.17449724674224854,
+      "learning_rate": 3.650714285714286e-05,
+      "loss": 0.0376,
+      "step": 1890
+    },
+    {
+      "epoch": 2.1714285714285713,
+      "grad_norm": 0.17534580826759338,
+      "learning_rate": 3.643571428571429e-05,
+      "loss": 0.0356,
+      "step": 1900
+    },
+    {
+      "epoch": 2.182857142857143,
+      "grad_norm": 0.11023643612861633,
+      "learning_rate": 3.636428571428572e-05,
+      "loss": 0.0363,
+      "step": 1910
+    },
+    {
+      "epoch": 2.1942857142857144,
+      "grad_norm": 0.1318758726119995,
+      "learning_rate": 3.629285714285714e-05,
+      "loss": 0.0363,
+      "step": 1920
+    },
+    {
+      "epoch": 2.2057142857142855,
+      "grad_norm": 0.13930299878120422,
+      "learning_rate": 3.6221428571428574e-05,
+      "loss": 0.0355,
+      "step": 1930
+    },
+    {
+      "epoch": 2.217142857142857,
+      "grad_norm": 0.13110798597335815,
+      "learning_rate": 3.615e-05,
+      "loss": 0.0366,
+      "step": 1940
+    },
+    {
+      "epoch": 2.2285714285714286,
+      "grad_norm": 0.1405479907989502,
+      "learning_rate": 3.607857142857143e-05,
+      "loss": 0.0369,
+      "step": 1950
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.15569846332073212,
+      "learning_rate": 3.600714285714286e-05,
+      "loss": 0.0377,
+      "step": 1960
+    },
+    {
+      "epoch": 2.2514285714285713,
+      "grad_norm": 0.11618765443563461,
+      "learning_rate": 3.593571428571429e-05,
+      "loss": 0.0343,
+      "step": 1970
+    },
+    {
+      "epoch": 2.262857142857143,
+      "grad_norm": 0.13145925104618073,
+      "learning_rate": 3.5864285714285714e-05,
+      "loss": 0.0344,
+      "step": 1980
+    },
+    {
+      "epoch": 2.2742857142857145,
+      "grad_norm": 0.10836513340473175,
+      "learning_rate": 3.5792857142857145e-05,
+      "loss": 0.0356,
+      "step": 1990
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.12755240499973297,
+      "learning_rate": 3.5721428571428575e-05,
+      "loss": 0.0369,
+      "step": 2000
+    },
+    {
+      "epoch": 2.297142857142857,
+      "grad_norm": 0.11500225961208344,
+      "learning_rate": 3.565e-05,
+      "loss": 0.0364,
+      "step": 2010
+    },
+    {
+      "epoch": 2.3085714285714287,
+      "grad_norm": 0.1484687179327011,
+      "learning_rate": 3.557857142857143e-05,
+      "loss": 0.037,
+      "step": 2020
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.13004055619239807,
+      "learning_rate": 3.550714285714286e-05,
+      "loss": 0.036,
+      "step": 2030
+    },
+    {
+      "epoch": 2.3314285714285714,
+      "grad_norm": 0.1361243575811386,
+      "learning_rate": 3.543571428571429e-05,
+      "loss": 0.0374,
+      "step": 2040
+    },
+    {
+      "epoch": 2.342857142857143,
+      "grad_norm": 0.10585059970617294,
+      "learning_rate": 3.5364285714285715e-05,
+      "loss": 0.0368,
+      "step": 2050
+    },
+    {
+      "epoch": 2.354285714285714,
+      "grad_norm": 0.0975913405418396,
+      "learning_rate": 3.5292857142857146e-05,
+      "loss": 0.035,
+      "step": 2060
+    },
+    {
+      "epoch": 2.3657142857142857,
+      "grad_norm": 0.13559779524803162,
+      "learning_rate": 3.522142857142857e-05,
+      "loss": 0.034,
+      "step": 2070
+    },
+    {
+      "epoch": 2.3771428571428572,
+      "grad_norm": 0.11691150814294815,
+      "learning_rate": 3.515e-05,
+      "loss": 0.0361,
+      "step": 2080
+    },
+    {
+      "epoch": 2.388571428571429,
+      "grad_norm": 0.09625136852264404,
+      "learning_rate": 3.507857142857143e-05,
+      "loss": 0.0356,
+      "step": 2090
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.15719136595726013,
+      "learning_rate": 3.500714285714286e-05,
+      "loss": 0.036,
+      "step": 2100
+    },
+    {
+      "epoch": 2.4114285714285715,
+      "grad_norm": 0.2051057517528534,
+      "learning_rate": 3.4935714285714286e-05,
+      "loss": 0.0356,
+      "step": 2110
+    },
+    {
+      "epoch": 2.422857142857143,
+      "grad_norm": 0.13482578098773956,
+      "learning_rate": 3.486428571428572e-05,
+      "loss": 0.0349,
+      "step": 2120
+    },
+    {
+      "epoch": 2.434285714285714,
+      "grad_norm": 0.13397566974163055,
+      "learning_rate": 3.479285714285714e-05,
+      "loss": 0.0364,
+      "step": 2130
+    },
+    {
+      "epoch": 2.4457142857142857,
+      "grad_norm": 0.1820402443408966,
+      "learning_rate": 3.472142857142857e-05,
+      "loss": 0.0347,
+      "step": 2140
+    },
+    {
+      "epoch": 2.4571428571428573,
+      "grad_norm": 0.13529136776924133,
+      "learning_rate": 3.465e-05,
+      "loss": 0.0364,
+      "step": 2150
+    },
+    {
+      "epoch": 2.4685714285714284,
+      "grad_norm": 0.08887185156345367,
+      "learning_rate": 3.457857142857143e-05,
+      "loss": 0.0361,
+      "step": 2160
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.12572938203811646,
+      "learning_rate": 3.4507142857142863e-05,
+      "loss": 0.0369,
+      "step": 2170
+    },
+    {
+      "epoch": 2.4914285714285715,
+      "grad_norm": 0.10877826064825058,
+      "learning_rate": 3.443571428571429e-05,
+      "loss": 0.0361,
+      "step": 2180
+    },
+    {
+      "epoch": 2.5028571428571427,
+      "grad_norm": 0.11405471712350845,
+      "learning_rate": 3.436428571428572e-05,
+      "loss": 0.0339,
+      "step": 2190
+    },
+    {
+      "epoch": 2.5142857142857142,
+      "grad_norm": 0.12478823214769363,
+      "learning_rate": 3.429285714285714e-05,
+      "loss": 0.0386,
+      "step": 2200
+    },
+    {
+      "epoch": 2.525714285714286,
+      "grad_norm": 0.1045498475432396,
+      "learning_rate": 3.422142857142857e-05,
+      "loss": 0.036,
+      "step": 2210
+    },
+    {
+      "epoch": 2.5371428571428574,
+      "grad_norm": 0.11122453212738037,
+      "learning_rate": 3.415e-05,
+      "loss": 0.0336,
+      "step": 2220
+    },
+    {
+      "epoch": 2.5485714285714285,
+      "grad_norm": 0.1618603616952896,
+      "learning_rate": 3.4078571428571434e-05,
+      "loss": 0.0352,
+      "step": 2230
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.10025645792484283,
+      "learning_rate": 3.400714285714286e-05,
+      "loss": 0.0339,
+      "step": 2240
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.12444077432155609,
+      "learning_rate": 3.393571428571429e-05,
+      "loss": 0.033,
+      "step": 2250
+    },
+    {
+      "epoch": 2.5828571428571427,
+      "grad_norm": 0.1822001338005066,
+      "learning_rate": 3.386428571428571e-05,
+      "loss": 0.035,
+      "step": 2260
+    },
+    {
+      "epoch": 2.5942857142857143,
+      "grad_norm": 0.09717897325754166,
+      "learning_rate": 3.379285714285714e-05,
+      "loss": 0.0342,
+      "step": 2270
+    },
+    {
+      "epoch": 2.605714285714286,
+      "grad_norm": 0.11690513044595718,
+      "learning_rate": 3.372142857142857e-05,
+      "loss": 0.0358,
+      "step": 2280
+    },
+    {
+      "epoch": 2.617142857142857,
+      "grad_norm": 0.09562231600284576,
+      "learning_rate": 3.3650000000000005e-05,
+      "loss": 0.0334,
+      "step": 2290
+    },
+    {
+      "epoch": 2.6285714285714286,
+      "grad_norm": 0.1744600236415863,
+      "learning_rate": 3.357857142857143e-05,
+      "loss": 0.0349,
+      "step": 2300
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2031196504831314,
+      "learning_rate": 3.350714285714286e-05,
+      "loss": 0.0354,
+      "step": 2310
+    },
+    {
+      "epoch": 2.6514285714285712,
+      "grad_norm": 0.16061373054981232,
+      "learning_rate": 3.343571428571429e-05,
+      "loss": 0.0337,
+      "step": 2320
+    },
+    {
+      "epoch": 2.662857142857143,
+      "grad_norm": 0.11644838005304337,
+      "learning_rate": 3.3364285714285714e-05,
+      "loss": 0.0345,
+      "step": 2330
+    },
+    {
+      "epoch": 2.6742857142857144,
+      "grad_norm": 0.14671610295772552,
+      "learning_rate": 3.3292857142857145e-05,
+      "loss": 0.0335,
+      "step": 2340
+    },
+    {
+      "epoch": 2.685714285714286,
+      "grad_norm": 0.09919189661741257,
+      "learning_rate": 3.3221428571428575e-05,
+      "loss": 0.035,
+      "step": 2350
+    },
+    {
+      "epoch": 2.697142857142857,
+      "grad_norm": 0.10584178566932678,
+      "learning_rate": 3.3150000000000006e-05,
+      "loss": 0.0337,
+      "step": 2360
+    },
+    {
+      "epoch": 2.7085714285714286,
+      "grad_norm": 0.09700863063335419,
+      "learning_rate": 3.307857142857143e-05,
+      "loss": 0.0362,
+      "step": 2370
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.14599616825580597,
+      "learning_rate": 3.300714285714286e-05,
+      "loss": 0.0355,
+      "step": 2380
+    },
+    {
+      "epoch": 2.7314285714285713,
+      "grad_norm": 0.11282289028167725,
+      "learning_rate": 3.2935714285714285e-05,
+      "loss": 0.0324,
+      "step": 2390
+    },
+    {
+      "epoch": 2.742857142857143,
+      "grad_norm": 0.11761584132909775,
+      "learning_rate": 3.2864285714285715e-05,
+      "loss": 0.0348,
+      "step": 2400
+    },
+    {
+      "epoch": 2.7542857142857144,
+      "grad_norm": 0.14849688112735748,
+      "learning_rate": 3.279285714285714e-05,
+      "loss": 0.0338,
+      "step": 2410
+    },
+    {
+      "epoch": 2.7657142857142856,
+      "grad_norm": 0.10622404515743256,
+      "learning_rate": 3.272142857142858e-05,
+      "loss": 0.0327,
+      "step": 2420
+    },
+    {
+      "epoch": 2.777142857142857,
+      "grad_norm": 0.14599359035491943,
+      "learning_rate": 3.265e-05,
+      "loss": 0.0335,
+      "step": 2430
+    },
+    {
+      "epoch": 2.7885714285714287,
+      "grad_norm": 0.11975230276584625,
+      "learning_rate": 3.257857142857143e-05,
+      "loss": 0.033,
+      "step": 2440
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.15992556512355804,
+      "learning_rate": 3.2507142857142855e-05,
+      "loss": 0.0356,
+      "step": 2450
+    },
+    {
+      "epoch": 2.8114285714285714,
+      "grad_norm": 0.08963726460933685,
+      "learning_rate": 3.2435714285714286e-05,
+      "loss": 0.0341,
+      "step": 2460
+    },
+    {
+      "epoch": 2.822857142857143,
+      "grad_norm": 0.10056816041469574,
+      "learning_rate": 3.236428571428572e-05,
+      "loss": 0.0349,
+      "step": 2470
+    },
+    {
+      "epoch": 2.8342857142857145,
+      "grad_norm": 0.12423935532569885,
+      "learning_rate": 3.229285714285715e-05,
+      "loss": 0.0341,
+      "step": 2480
+    },
+    {
+      "epoch": 2.8457142857142856,
+      "grad_norm": 0.12661360204219818,
+      "learning_rate": 3.222142857142857e-05,
+      "loss": 0.0339,
+      "step": 2490
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.08652866631746292,
+      "learning_rate": 3.215e-05,
+      "loss": 0.0345,
+      "step": 2500
+    },
+    {
+      "epoch": 2.8685714285714283,
+      "grad_norm": 0.10129117220640182,
+      "learning_rate": 3.207857142857143e-05,
+      "loss": 0.0364,
+      "step": 2510
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.09667441248893738,
+      "learning_rate": 3.200714285714286e-05,
+      "loss": 0.0335,
+      "step": 2520
+    },
+    {
+      "epoch": 2.8914285714285715,
+      "grad_norm": 0.10019873082637787,
+      "learning_rate": 3.193571428571429e-05,
+      "loss": 0.034,
+      "step": 2530
+    },
+    {
+      "epoch": 2.902857142857143,
+      "grad_norm": 0.109789177775383,
+      "learning_rate": 3.186428571428571e-05,
+      "loss": 0.0325,
+      "step": 2540
+    },
+    {
+      "epoch": 2.914285714285714,
+      "grad_norm": 0.13727444410324097,
+      "learning_rate": 3.179285714285715e-05,
+      "loss": 0.0322,
+      "step": 2550
+    },
+    {
+      "epoch": 2.9257142857142857,
+      "grad_norm": 0.09896359592676163,
+      "learning_rate": 3.172142857142857e-05,
+      "loss": 0.0346,
+      "step": 2560
+    },
+    {
+      "epoch": 2.9371428571428573,
+      "grad_norm": 0.09976238012313843,
+      "learning_rate": 3.1650000000000004e-05,
+      "loss": 0.0323,
+      "step": 2570
+    },
+    {
+      "epoch": 2.9485714285714284,
+      "grad_norm": 0.10362710803747177,
+      "learning_rate": 3.157857142857143e-05,
+      "loss": 0.0347,
+      "step": 2580
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.11956743150949478,
+      "learning_rate": 3.150714285714286e-05,
+      "loss": 0.035,
+      "step": 2590
+    },
+    {
+      "epoch": 2.9714285714285715,
+      "grad_norm": 0.1660848706960678,
+      "learning_rate": 3.143571428571428e-05,
+      "loss": 0.0341,
+      "step": 2600
+    },
+    {
+      "epoch": 2.982857142857143,
+      "grad_norm": 0.0876573994755745,
+      "learning_rate": 3.136428571428572e-05,
+      "loss": 0.0349,
+      "step": 2610
+    },
+    {
+      "epoch": 2.994285714285714,
+      "grad_norm": 0.10691452026367188,
+      "learning_rate": 3.1292857142857144e-05,
+      "loss": 0.0349,
+      "step": 2620
+    },
+    {
+      "epoch": 3.0,
+      "eval_bleu": 0.07,
+      "eval_exact_match": 0.0,
+      "eval_loss": 0.029123220592737198,
+      "eval_runtime": 344.2459,
+      "eval_samples_per_second": 17.429,
+      "eval_steps_per_second": 1.089,
+      "step": 2625
+    },
+    {
+      "epoch": 3.005714285714286,
+      "grad_norm": 0.14447887241840363,
+      "learning_rate": 3.1221428571428574e-05,
+      "loss": 0.0358,
+      "step": 2630
+    },
+    {
+      "epoch": 3.0171428571428573,
+      "grad_norm": 0.0858982652425766,
+      "learning_rate": 3.115e-05,
+      "loss": 0.0323,
+      "step": 2640
+    },
+    {
+      "epoch": 3.0285714285714285,
+      "grad_norm": 0.08900221437215805,
+      "learning_rate": 3.107857142857143e-05,
+      "loss": 0.0334,
+      "step": 2650
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.11133663356304169,
+      "learning_rate": 3.100714285714286e-05,
+      "loss": 0.0338,
+      "step": 2660
+    },
+    {
+      "epoch": 3.0514285714285716,
+      "grad_norm": 0.08960743993520737,
+      "learning_rate": 3.0935714285714284e-05,
+      "loss": 0.0329,
+      "step": 2670
+    },
+    {
+      "epoch": 3.0628571428571427,
+      "grad_norm": 0.1239769235253334,
+      "learning_rate": 3.086428571428572e-05,
+      "loss": 0.0326,
+      "step": 2680
+    },
+    {
+      "epoch": 3.0742857142857143,
+      "grad_norm": 0.1012355387210846,
+      "learning_rate": 3.0792857142857145e-05,
+      "loss": 0.0346,
+      "step": 2690
+    },
+    {
+      "epoch": 3.085714285714286,
+      "grad_norm": 0.08918215334415436,
+      "learning_rate": 3.0721428571428576e-05,
+      "loss": 0.0332,
+      "step": 2700
+    },
+    {
+      "epoch": 3.097142857142857,
+      "grad_norm": 0.11924552172422409,
+      "learning_rate": 3.065e-05,
+      "loss": 0.0325,
+      "step": 2710
+    },
+    {
+      "epoch": 3.1085714285714285,
+      "grad_norm": 0.08870343118906021,
+      "learning_rate": 3.057857142857143e-05,
+      "loss": 0.0348,
+      "step": 2720
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.10869447886943817,
+      "learning_rate": 3.0507142857142858e-05,
+      "loss": 0.0316,
+      "step": 2730
+    },
+    {
+      "epoch": 3.1314285714285712,
+      "grad_norm": 0.15603742003440857,
+      "learning_rate": 3.043571428571429e-05,
+      "loss": 0.0319,
+      "step": 2740
+    },
+    {
+      "epoch": 3.142857142857143,
+      "grad_norm": 0.0926317349076271,
+      "learning_rate": 3.0364285714285716e-05,
+      "loss": 0.0337,
+      "step": 2750
+    },
+    {
+      "epoch": 3.1542857142857144,
+      "grad_norm": 0.1345146745443344,
+      "learning_rate": 3.0292857142857146e-05,
+      "loss": 0.0355,
+      "step": 2760
+    },
+    {
+      "epoch": 3.1657142857142855,
+      "grad_norm": 0.1272844821214676,
+      "learning_rate": 3.0221428571428574e-05,
+      "loss": 0.0357,
+      "step": 2770
+    },
+    {
+      "epoch": 3.177142857142857,
+      "grad_norm": 0.1036539152264595,
+      "learning_rate": 3.015e-05,
+      "loss": 0.0341,
+      "step": 2780
+    },
+    {
+      "epoch": 3.1885714285714286,
+      "grad_norm": 0.1159033477306366,
+      "learning_rate": 3.007857142857143e-05,
+      "loss": 0.0332,
+      "step": 2790
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.15586401522159576,
+      "learning_rate": 3.0007142857142856e-05,
+      "loss": 0.0326,
+      "step": 2800
+    },
+    {
+      "epoch": 3.2114285714285713,
+      "grad_norm": 0.11249198019504547,
+      "learning_rate": 2.993571428571429e-05,
+      "loss": 0.0345,
+      "step": 2810
+    },
+    {
+      "epoch": 3.222857142857143,
+      "grad_norm": 0.08775516599416733,
+      "learning_rate": 2.9864285714285717e-05,
+      "loss": 0.0328,
+      "step": 2820
+    },
+    {
+      "epoch": 3.2342857142857144,
+      "grad_norm": 0.1125701293349266,
+      "learning_rate": 2.9792857142857144e-05,
+      "loss": 0.0335,
+      "step": 2830
+    },
+    {
+      "epoch": 3.2457142857142856,
+      "grad_norm": 0.09027904272079468,
+      "learning_rate": 2.9721428571428572e-05,
+      "loss": 0.0322,
+      "step": 2840
+    },
+    {
+      "epoch": 3.257142857142857,
+      "grad_norm": 0.09299852699041367,
+      "learning_rate": 2.965e-05,
+      "loss": 0.0346,
+      "step": 2850
+    },
+    {
+      "epoch": 3.2685714285714287,
+      "grad_norm": 0.08893876522779465,
+      "learning_rate": 2.9578571428571426e-05,
+      "loss": 0.0326,
+      "step": 2860
+    },
+    {
+      "epoch": 3.2800000000000002,
+      "grad_norm": 0.11996601521968842,
+      "learning_rate": 2.950714285714286e-05,
+      "loss": 0.0339,
+      "step": 2870
+    },
+    {
+      "epoch": 3.2914285714285714,
+      "grad_norm": 0.10866489261388779,
+      "learning_rate": 2.9435714285714288e-05,
+      "loss": 0.0351,
+      "step": 2880
+    },
+    {
+      "epoch": 3.302857142857143,
+      "grad_norm": 0.09160767495632172,
+      "learning_rate": 2.9364285714285715e-05,
+      "loss": 0.0327,
+      "step": 2890
+    },
+    {
+      "epoch": 3.314285714285714,
+      "grad_norm": 0.12444985657930374,
+      "learning_rate": 2.9292857142857146e-05,
+      "loss": 0.0355,
+      "step": 2900
+    },
+    {
+      "epoch": 3.3257142857142856,
+      "grad_norm": 0.07536906003952026,
+      "learning_rate": 2.9221428571428573e-05,
+      "loss": 0.0336,
+      "step": 2910
+    },
+    {
+      "epoch": 3.337142857142857,
+      "grad_norm": 0.08127142488956451,
+      "learning_rate": 2.915e-05,
+      "loss": 0.033,
+      "step": 2920
+    },
+    {
+      "epoch": 3.3485714285714288,
+      "grad_norm": 0.0742766261100769,
+      "learning_rate": 2.9078571428571428e-05,
+      "loss": 0.0328,
+      "step": 2930
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.08692529797554016,
+      "learning_rate": 2.9007142857142862e-05,
+      "loss": 0.0329,
+      "step": 2940
+    },
+    {
+      "epoch": 3.3714285714285714,
+      "grad_norm": 0.08498033881187439,
+      "learning_rate": 2.893571428571429e-05,
+      "loss": 0.0329,
+      "step": 2950
+    },
+    {
+      "epoch": 3.382857142857143,
+      "grad_norm": 0.07162804156541824,
+      "learning_rate": 2.8864285714285716e-05,
+      "loss": 0.0356,
+      "step": 2960
+    },
+    {
+      "epoch": 3.394285714285714,
+      "grad_norm": 0.12961073219776154,
+      "learning_rate": 2.8792857142857144e-05,
+      "loss": 0.0335,
+      "step": 2970
+    },
+    {
+      "epoch": 3.4057142857142857,
+      "grad_norm": 0.09309065341949463,
+      "learning_rate": 2.872142857142857e-05,
+      "loss": 0.0312,
+      "step": 2980
+    },
+    {
+      "epoch": 3.4171428571428573,
+      "grad_norm": 0.09326250106096268,
+      "learning_rate": 2.865e-05,
+      "loss": 0.0334,
+      "step": 2990
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.23755627870559692,
+      "learning_rate": 2.8578571428571433e-05,
+      "loss": 0.0339,
+      "step": 3000
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.32597196102142334,
+      "learning_rate": 2.850714285714286e-05,
+      "loss": 0.0324,
+      "step": 3010
+    },
+    {
+      "epoch": 3.4514285714285715,
+      "grad_norm": 0.1102445125579834,
+      "learning_rate": 2.8435714285714287e-05,
+      "loss": 0.0317,
+      "step": 3020
+    },
+    {
+      "epoch": 3.4628571428571426,
+      "grad_norm": 0.09068254381418228,
+      "learning_rate": 2.8364285714285714e-05,
+      "loss": 0.0336,
+      "step": 3030
+    },
+    {
+      "epoch": 3.474285714285714,
+      "grad_norm": 0.10059994459152222,
+      "learning_rate": 2.8292857142857142e-05,
+      "loss": 0.0337,
+      "step": 3040
+    },
+    {
+      "epoch": 3.4857142857142858,
+      "grad_norm": 0.08563591539859772,
+      "learning_rate": 2.8221428571428573e-05,
+      "loss": 0.0339,
+      "step": 3050
+    },
+    {
+      "epoch": 3.4971428571428573,
+      "grad_norm": 0.23142589628696442,
+      "learning_rate": 2.815e-05,
+      "loss": 0.0331,
+      "step": 3060
+    },
+    {
+      "epoch": 3.5085714285714285,
+      "grad_norm": 0.10148648172616959,
+      "learning_rate": 2.807857142857143e-05,
+      "loss": 0.0324,
+      "step": 3070
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.10303816944360733,
+      "learning_rate": 2.800714285714286e-05,
+      "loss": 0.0342,
+      "step": 3080
+    },
+    {
+      "epoch": 3.5314285714285716,
+      "grad_norm": 0.1418079286813736,
+      "learning_rate": 2.793571428571429e-05,
+      "loss": 0.0335,
+      "step": 3090
+    },
+    {
+      "epoch": 3.5428571428571427,
+      "grad_norm": 0.10841654986143112,
+      "learning_rate": 2.7864285714285716e-05,
+      "loss": 0.0327,
+      "step": 3100
+    },
+    {
+      "epoch": 3.5542857142857143,
+      "grad_norm": 0.07732987403869629,
+      "learning_rate": 2.7792857142857143e-05,
+      "loss": 0.0312,
+      "step": 3110
+    },
+    {
+      "epoch": 3.565714285714286,
+      "grad_norm": 0.08883870393037796,
+      "learning_rate": 2.772142857142857e-05,
+      "loss": 0.0318,
+      "step": 3120
+    },
+    {
+      "epoch": 3.5771428571428574,
+      "grad_norm": 0.08147554099559784,
+      "learning_rate": 2.7650000000000005e-05,
+      "loss": 0.033,
+      "step": 3130
+    },
+    {
+      "epoch": 3.5885714285714285,
+      "grad_norm": 0.14047172665596008,
+      "learning_rate": 2.7578571428571432e-05,
+      "loss": 0.033,
+      "step": 3140
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.0935438871383667,
+      "learning_rate": 2.750714285714286e-05,
+      "loss": 0.0314,
+      "step": 3150
+    },
+    {
+      "epoch": 3.611428571428571,
+      "grad_norm": 0.08656948059797287,
+      "learning_rate": 2.7435714285714287e-05,
+      "loss": 0.0317,
+      "step": 3160
+    },
+    {
+      "epoch": 3.6228571428571428,
+      "grad_norm": 0.07801087200641632,
+      "learning_rate": 2.7364285714285714e-05,
+      "loss": 0.031,
+      "step": 3170
+    },
+    {
+      "epoch": 3.6342857142857143,
+      "grad_norm": 0.10883987694978714,
+      "learning_rate": 2.729285714285714e-05,
+      "loss": 0.031,
+      "step": 3180
+    },
+    {
+      "epoch": 3.645714285714286,
+      "grad_norm": 0.1345929652452469,
+      "learning_rate": 2.7221428571428575e-05,
+      "loss": 0.033,
+      "step": 3190
+    },
+    {
+      "epoch": 3.657142857142857,
+      "grad_norm": 0.07864482700824738,
+      "learning_rate": 2.7150000000000003e-05,
+      "loss": 0.0332,
+      "step": 3200
+    },
+    {
+      "epoch": 3.6685714285714286,
+      "grad_norm": 0.07242916524410248,
+      "learning_rate": 2.707857142857143e-05,
+      "loss": 0.0322,
+      "step": 3210
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.30902132391929626,
+      "learning_rate": 2.7007142857142857e-05,
+      "loss": 0.0327,
+      "step": 3220
+    },
+    {
+      "epoch": 3.6914285714285713,
+      "grad_norm": 0.13460323214530945,
+      "learning_rate": 2.6935714285714288e-05,
+      "loss": 0.0335,
+      "step": 3230
+    },
+    {
+      "epoch": 3.702857142857143,
+      "grad_norm": 0.10607478767633438,
+      "learning_rate": 2.6864285714285715e-05,
+      "loss": 0.0316,
+      "step": 3240
+    },
+    {
+      "epoch": 3.7142857142857144,
+      "grad_norm": 0.12986110150814056,
+      "learning_rate": 2.6792857142857143e-05,
+      "loss": 0.0326,
+      "step": 3250
+    },
+    {
+      "epoch": 3.725714285714286,
+      "grad_norm": 0.06864488869905472,
+      "learning_rate": 2.6721428571428577e-05,
+      "loss": 0.0316,
+      "step": 3260
+    },
+    {
+      "epoch": 3.737142857142857,
+      "grad_norm": 0.12121660262346268,
+      "learning_rate": 2.6650000000000004e-05,
+      "loss": 0.0319,
+      "step": 3270
+    },
+    {
+      "epoch": 3.7485714285714287,
+      "grad_norm": 0.07145349681377411,
+      "learning_rate": 2.657857142857143e-05,
+      "loss": 0.0332,
+      "step": 3280
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.12492495775222778,
+      "learning_rate": 2.650714285714286e-05,
+      "loss": 0.0331,
+      "step": 3290
+    },
+    {
+      "epoch": 3.7714285714285714,
+      "grad_norm": 0.06998569518327713,
+      "learning_rate": 2.6435714285714286e-05,
+      "loss": 0.0314,
+      "step": 3300
+    },
+    {
+      "epoch": 3.782857142857143,
+      "grad_norm": 0.09135474264621735,
+      "learning_rate": 2.6364285714285713e-05,
+      "loss": 0.0325,
+      "step": 3310
+    },
+    {
+      "epoch": 3.7942857142857145,
+      "grad_norm": 0.16541588306427002,
+      "learning_rate": 2.6292857142857147e-05,
+      "loss": 0.0342,
+      "step": 3320
+    },
+    {
+      "epoch": 3.8057142857142856,
+      "grad_norm": 0.08637196570634842,
+      "learning_rate": 2.6221428571428575e-05,
+      "loss": 0.0311,
+      "step": 3330
+    },
+    {
+      "epoch": 3.817142857142857,
+      "grad_norm": 0.09347192198038101,
+      "learning_rate": 2.6150000000000002e-05,
+      "loss": 0.0332,
+      "step": 3340
+    },
+    {
+      "epoch": 3.8285714285714287,
+      "grad_norm": 0.10765931010246277,
+      "learning_rate": 2.607857142857143e-05,
+      "loss": 0.032,
+      "step": 3350
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.07229693233966827,
+      "learning_rate": 2.6007142857142857e-05,
+      "loss": 0.0322,
+      "step": 3360
+    },
+    {
+      "epoch": 3.8514285714285714,
+      "grad_norm": 0.08713468909263611,
+      "learning_rate": 2.5935714285714284e-05,
+      "loss": 0.0345,
+      "step": 3370
+    },
+    {
+      "epoch": 3.862857142857143,
+      "grad_norm": 0.13004758954048157,
+      "learning_rate": 2.5864285714285715e-05,
+      "loss": 0.0328,
+      "step": 3380
+    },
+    {
+      "epoch": 3.8742857142857146,
+      "grad_norm": 0.0932050496339798,
+      "learning_rate": 2.5792857142857145e-05,
+      "loss": 0.0337,
+      "step": 3390
+    },
+    {
+      "epoch": 3.8857142857142857,
+      "grad_norm": 0.11469905078411102,
+      "learning_rate": 2.5721428571428573e-05,
+      "loss": 0.0334,
+      "step": 3400
+    },
+    {
+      "epoch": 3.8971428571428572,
+      "grad_norm": 0.06456197798252106,
+      "learning_rate": 2.5650000000000003e-05,
+      "loss": 0.0308,
+      "step": 3410
+    },
+    {
+      "epoch": 3.9085714285714284,
+      "grad_norm": 0.06196196377277374,
+      "learning_rate": 2.557857142857143e-05,
+      "loss": 0.0311,
+      "step": 3420
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.07760277390480042,
+      "learning_rate": 2.5507142857142858e-05,
+      "loss": 0.0335,
+      "step": 3430
+    },
+    {
+      "epoch": 3.9314285714285715,
+      "grad_norm": 0.13347361981868744,
+      "learning_rate": 2.5435714285714285e-05,
+      "loss": 0.0322,
+      "step": 3440
+    },
+    {
+      "epoch": 3.942857142857143,
+      "grad_norm": 0.07051094621419907,
+      "learning_rate": 2.536428571428572e-05,
+      "loss": 0.032,
+      "step": 3450
+    },
+    {
+      "epoch": 3.954285714285714,
+      "grad_norm": 0.07086417078971863,
+      "learning_rate": 2.5292857142857147e-05,
+      "loss": 0.0339,
+      "step": 3460
+    },
+    {
+      "epoch": 3.9657142857142857,
+      "grad_norm": 0.07837912440299988,
+      "learning_rate": 2.5221428571428574e-05,
+      "loss": 0.0316,
+      "step": 3470
+    },
+    {
+      "epoch": 3.977142857142857,
+      "grad_norm": 0.07793241739273071,
+      "learning_rate": 2.515e-05,
+      "loss": 0.0335,
+      "step": 3480
+    },
+    {
+      "epoch": 3.9885714285714284,
+      "grad_norm": 0.10782703012228012,
+      "learning_rate": 2.507857142857143e-05,
+      "loss": 0.0327,
+      "step": 3490
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.1338101178407669,
+      "learning_rate": 2.5007142857142856e-05,
+      "loss": 0.0329,
+      "step": 3500
+    },
+    {
+      "epoch": 4.0,
+      "eval_bleu": 0.07,
+      "eval_exact_match": 0.0,
+      "eval_loss": 0.02901034988462925,
+      "eval_runtime": 343.4078,
+      "eval_samples_per_second": 17.472,
+      "eval_steps_per_second": 1.092,
+      "step": 3500
+    },
+    {
+      "epoch": 4.011428571428572,
+      "grad_norm": 0.10343596339225769,
+      "learning_rate": 2.4935714285714287e-05,
+      "loss": 0.0323,
+      "step": 3510
+    },
+    {
+      "epoch": 4.022857142857143,
+      "grad_norm": 0.089602530002594,
+      "learning_rate": 2.4864285714285714e-05,
+      "loss": 0.0336,
+      "step": 3520
+    },
+    {
+      "epoch": 4.034285714285715,
+      "grad_norm": 0.06534019112586975,
+      "learning_rate": 2.479285714285714e-05,
+      "loss": 0.031,
+      "step": 3530
+    },
+    {
+      "epoch": 4.045714285714285,
+      "grad_norm": 0.07555046677589417,
+      "learning_rate": 2.4721428571428572e-05,
+      "loss": 0.0317,
+      "step": 3540
+    },
+    {
+      "epoch": 4.057142857142857,
+      "grad_norm": 0.07456432282924652,
+      "learning_rate": 2.465e-05,
+      "loss": 0.0332,
+      "step": 3550
+    },
+    {
+      "epoch": 4.0685714285714285,
+      "grad_norm": 0.07345906645059586,
+      "learning_rate": 2.457857142857143e-05,
+      "loss": 0.0311,
+      "step": 3560
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.07652904838323593,
+      "learning_rate": 2.4507142857142857e-05,
+      "loss": 0.0317,
+      "step": 3570
+    },
+    {
+      "epoch": 4.091428571428572,
+      "grad_norm": 0.08077405393123627,
+      "learning_rate": 2.4435714285714288e-05,
+      "loss": 0.0351,
+      "step": 3580
+    },
+    {
+      "epoch": 4.102857142857143,
+      "grad_norm": 0.13678324222564697,
+      "learning_rate": 2.4364285714285716e-05,
+      "loss": 0.032,
+      "step": 3590
+    },
+    {
+      "epoch": 4.114285714285714,
+      "grad_norm": 0.08344317227602005,
+      "learning_rate": 2.4292857142857146e-05,
+      "loss": 0.0339,
+      "step": 3600
+    },
+    {
+      "epoch": 4.1257142857142854,
+      "grad_norm": 0.07824575155973434,
+      "learning_rate": 2.4221428571428574e-05,
+      "loss": 0.0318,
+      "step": 3610
+    },
+    {
+      "epoch": 4.137142857142857,
+      "grad_norm": 0.1025475487112999,
+      "learning_rate": 2.415e-05,
+      "loss": 0.0323,
+      "step": 3620
+    },
+    {
+      "epoch": 4.148571428571429,
+      "grad_norm": 0.17045310139656067,
+      "learning_rate": 2.407857142857143e-05,
+      "loss": 0.0329,
+      "step": 3630
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.07618770748376846,
+      "learning_rate": 2.400714285714286e-05,
+      "loss": 0.0307,
+      "step": 3640
+    },
+    {
+      "epoch": 4.171428571428572,
+      "grad_norm": 0.09664011001586914,
+      "learning_rate": 2.3935714285714286e-05,
+      "loss": 0.0326,
+      "step": 3650
+    },
+    {
+      "epoch": 4.182857142857143,
+      "grad_norm": 0.0686192512512207,
+      "learning_rate": 2.3864285714285717e-05,
+      "loss": 0.0317,
+      "step": 3660
+    },
+    {
+      "epoch": 4.194285714285714,
+      "grad_norm": 0.06639493256807327,
+      "learning_rate": 2.3792857142857144e-05,
+      "loss": 0.0335,
+      "step": 3670
+    },
+    {
+      "epoch": 4.2057142857142855,
+      "grad_norm": 0.12928254902362823,
+      "learning_rate": 2.372142857142857e-05,
+      "loss": 0.0314,
+      "step": 3680
+    },
+    {
+      "epoch": 4.217142857142857,
+      "grad_norm": 0.17191903293132782,
+      "learning_rate": 2.365e-05,
+      "loss": 0.0312,
+      "step": 3690
+    },
+    {
+      "epoch": 4.228571428571429,
+      "grad_norm": 0.0797748863697052,
+      "learning_rate": 2.357857142857143e-05,
+      "loss": 0.0328,
+      "step": 3700
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.08241511881351471,
+      "learning_rate": 2.3507142857142857e-05,
+      "loss": 0.0321,
+      "step": 3710
+    },
+    {
+      "epoch": 4.251428571428572,
+      "grad_norm": 0.12593980133533478,
+      "learning_rate": 2.3435714285714284e-05,
+      "loss": 0.0308,
+      "step": 3720
+    },
+    {
+      "epoch": 4.2628571428571425,
+      "grad_norm": 0.09482847154140472,
+      "learning_rate": 2.3364285714285715e-05,
+      "loss": 0.0323,
+      "step": 3730
+    },
+    {
+      "epoch": 4.274285714285714,
+      "grad_norm": 0.08406007289886475,
+      "learning_rate": 2.3292857142857146e-05,
+      "loss": 0.0325,
+      "step": 3740
+    },
+    {
+      "epoch": 4.285714285714286,
+      "grad_norm": 0.10658417642116547,
+      "learning_rate": 2.3221428571428573e-05,
+      "loss": 0.0313,
+      "step": 3750
+    },
+    {
+      "epoch": 4.297142857142857,
+      "grad_norm": 0.09814857691526413,
+      "learning_rate": 2.3150000000000004e-05,
+      "loss": 0.0327,
+      "step": 3760
+    },
+    {
+      "epoch": 4.308571428571429,
+      "grad_norm": 0.14591078460216522,
+      "learning_rate": 2.307857142857143e-05,
+      "loss": 0.0336,
+      "step": 3770
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.11772993952035904,
+      "learning_rate": 2.3007142857142858e-05,
+      "loss": 0.0318,
+      "step": 3780
+    },
+    {
+      "epoch": 4.331428571428571,
+      "grad_norm": 0.08087225258350372,
+      "learning_rate": 2.293571428571429e-05,
+      "loss": 0.0313,
+      "step": 3790
+    },
+    {
+      "epoch": 4.3428571428571425,
+      "grad_norm": 0.11614172905683517,
+      "learning_rate": 2.2864285714285716e-05,
+      "loss": 0.0311,
+      "step": 3800
+    },
+    {
+      "epoch": 4.354285714285714,
+      "grad_norm": 0.10768585652112961,
+      "learning_rate": 2.2792857142857144e-05,
+      "loss": 0.0328,
+      "step": 3810
+    },
+    {
+      "epoch": 4.365714285714286,
+      "grad_norm": 0.07184877246618271,
+      "learning_rate": 2.272142857142857e-05,
+      "loss": 0.0307,
+      "step": 3820
+    },
+    {
+      "epoch": 4.377142857142857,
+      "grad_norm": 0.07980121672153473,
+      "learning_rate": 2.265e-05,
+      "loss": 0.0323,
+      "step": 3830
+    },
+    {
+      "epoch": 4.388571428571429,
+      "grad_norm": 0.1039133369922638,
+      "learning_rate": 2.257857142857143e-05,
+      "loss": 0.0333,
+      "step": 3840
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.12529844045639038,
+      "learning_rate": 2.2507142857142856e-05,
+      "loss": 0.0325,
+      "step": 3850
+    },
+    {
+      "epoch": 4.411428571428571,
+      "grad_norm": 0.09552208334207535,
+      "learning_rate": 2.2435714285714287e-05,
+      "loss": 0.0321,
+      "step": 3860
+    },
+    {
+      "epoch": 4.422857142857143,
+      "grad_norm": 0.086667500436306,
+      "learning_rate": 2.2364285714285714e-05,
+      "loss": 0.032,
+      "step": 3870
+    },
+    {
+      "epoch": 4.434285714285714,
+      "grad_norm": 0.09163305163383484,
+      "learning_rate": 2.229285714285714e-05,
+      "loss": 0.0333,
+      "step": 3880
+    },
+    {
+      "epoch": 4.445714285714286,
+      "grad_norm": 0.06276048719882965,
+      "learning_rate": 2.2221428571428572e-05,
+      "loss": 0.031,
+      "step": 3890
+    },
+    {
+      "epoch": 4.457142857142857,
+      "grad_norm": 0.07514110952615738,
+      "learning_rate": 2.215e-05,
+      "loss": 0.0315,
+      "step": 3900
+    },
+    {
+      "epoch": 4.468571428571429,
+      "grad_norm": 0.0655631348490715,
+      "learning_rate": 2.207857142857143e-05,
+      "loss": 0.0316,
+      "step": 3910
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.09730257093906403,
+      "learning_rate": 2.200714285714286e-05,
+      "loss": 0.0312,
+      "step": 3920
+    },
+    {
+      "epoch": 4.491428571428571,
+      "grad_norm": 0.07015182077884674,
+      "learning_rate": 2.193571428571429e-05,
+      "loss": 0.0329,
+      "step": 3930
+    },
+    {
+      "epoch": 4.502857142857143,
+      "grad_norm": 0.09095067530870438,
+      "learning_rate": 2.1864285714285716e-05,
+      "loss": 0.0333,
+      "step": 3940
+    },
+    {
+      "epoch": 4.514285714285714,
+      "grad_norm": 0.08112490177154541,
+      "learning_rate": 2.1792857142857143e-05,
+      "loss": 0.0326,
+      "step": 3950
+    },
+    {
+      "epoch": 4.525714285714286,
+      "grad_norm": 0.11766567826271057,
+      "learning_rate": 2.1721428571428574e-05,
+      "loss": 0.0314,
+      "step": 3960
+    },
+    {
+      "epoch": 4.537142857142857,
+      "grad_norm": 0.0992494523525238,
+      "learning_rate": 2.165e-05,
+      "loss": 0.031,
+      "step": 3970
+    },
+    {
+      "epoch": 4.548571428571429,
+      "grad_norm": 0.2512616813182831,
+      "learning_rate": 2.157857142857143e-05,
+      "loss": 0.0321,
+      "step": 3980
+    },
+    {
+      "epoch": 4.5600000000000005,
+      "grad_norm": 0.13177117705345154,
+      "learning_rate": 2.150714285714286e-05,
+      "loss": 0.0312,
+      "step": 3990
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 0.14732351899147034,
+      "learning_rate": 2.1435714285714286e-05,
+      "loss": 0.0332,
+      "step": 4000
+    },
+    {
+      "epoch": 4.582857142857143,
+      "grad_norm": 0.08171796053647995,
+      "learning_rate": 2.1364285714285714e-05,
+      "loss": 0.0325,
+      "step": 4010
+    },
+    {
+      "epoch": 4.594285714285714,
+      "grad_norm": 0.08084321767091751,
+      "learning_rate": 2.1292857142857144e-05,
+      "loss": 0.0316,
+      "step": 4020
+    },
+    {
+      "epoch": 4.605714285714286,
+      "grad_norm": 0.11409573256969452,
+      "learning_rate": 2.1221428571428572e-05,
+      "loss": 0.0329,
+      "step": 4030
+    },
+    {
+      "epoch": 4.617142857142857,
+      "grad_norm": 0.07737822085618973,
+      "learning_rate": 2.115e-05,
+      "loss": 0.0308,
+      "step": 4040
+    },
+    {
+      "epoch": 4.628571428571428,
+      "grad_norm": 0.08457338809967041,
+      "learning_rate": 2.107857142857143e-05,
+      "loss": 0.0319,
+      "step": 4050
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.11584480851888657,
+      "learning_rate": 2.1007142857142857e-05,
+      "loss": 0.0321,
+      "step": 4060
+    },
+    {
+      "epoch": 4.651428571428571,
+      "grad_norm": 0.0774964839220047,
+      "learning_rate": 2.0935714285714288e-05,
+      "loss": 0.0309,
+      "step": 4070
+    },
+    {
+      "epoch": 4.662857142857143,
+      "grad_norm": 0.1350235491991043,
+      "learning_rate": 2.0864285714285715e-05,
+      "loss": 0.031,
+      "step": 4080
+    },
+    {
+      "epoch": 4.674285714285714,
+      "grad_norm": 0.07618366181850433,
+      "learning_rate": 2.0792857142857146e-05,
+      "loss": 0.031,
+      "step": 4090
+    },
+    {
+      "epoch": 4.685714285714286,
+      "grad_norm": 0.06407870352268219,
+      "learning_rate": 2.0721428571428573e-05,
+      "loss": 0.034,
+      "step": 4100
+    },
+    {
+      "epoch": 4.6971428571428575,
+      "grad_norm": 0.09820050001144409,
+      "learning_rate": 2.065e-05,
+      "loss": 0.0325,
+      "step": 4110
+    },
+    {
+      "epoch": 4.708571428571428,
+      "grad_norm": 0.07984957098960876,
+      "learning_rate": 2.057857142857143e-05,
+      "loss": 0.0312,
+      "step": 4120
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.06861667335033417,
+      "learning_rate": 2.050714285714286e-05,
+      "loss": 0.0313,
+      "step": 4130
+    },
+    {
+      "epoch": 4.731428571428571,
+      "grad_norm": 0.07516956329345703,
+      "learning_rate": 2.0435714285714286e-05,
+      "loss": 0.0336,
+      "step": 4140
+    },
+    {
+      "epoch": 4.742857142857143,
+      "grad_norm": 0.06301568448543549,
+      "learning_rate": 2.0364285714285717e-05,
+      "loss": 0.0317,
+      "step": 4150
+    },
+    {
+      "epoch": 4.7542857142857144,
+      "grad_norm": 0.2740742266178131,
+      "learning_rate": 2.0292857142857144e-05,
+      "loss": 0.0323,
+      "step": 4160
+    },
+    {
+      "epoch": 4.765714285714286,
+      "grad_norm": 0.15055391192436218,
+      "learning_rate": 2.022142857142857e-05,
+      "loss": 0.0316,
+      "step": 4170
+    },
+    {
+      "epoch": 4.777142857142858,
+      "grad_norm": 0.06952224671840668,
+      "learning_rate": 2.0150000000000002e-05,
+      "loss": 0.0311,
+      "step": 4180
+    },
+    {
+      "epoch": 4.788571428571428,
+      "grad_norm": 0.06311614066362381,
+      "learning_rate": 2.007857142857143e-05,
+      "loss": 0.0306,
+      "step": 4190
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.05969397723674774,
+      "learning_rate": 2.0007142857142857e-05,
+      "loss": 0.0303,
+      "step": 4200
+    },
+    {
+      "epoch": 4.811428571428571,
+      "grad_norm": 0.1038677766919136,
+      "learning_rate": 1.9935714285714284e-05,
+      "loss": 0.032,
+      "step": 4210
+    },
+    {
+      "epoch": 4.822857142857143,
+      "grad_norm": 0.06804952770471573,
+      "learning_rate": 1.9864285714285715e-05,
+      "loss": 0.0304,
+      "step": 4220
+    },
+    {
+      "epoch": 4.8342857142857145,
+      "grad_norm": 0.0827961415052414,
+      "learning_rate": 1.9792857142857142e-05,
+      "loss": 0.031,
+      "step": 4230
+    },
+    {
+      "epoch": 4.845714285714286,
+      "grad_norm": 0.07299306988716125,
+      "learning_rate": 1.9721428571428573e-05,
+      "loss": 0.0308,
+      "step": 4240
+    },
+    {
+      "epoch": 4.857142857142857,
+      "grad_norm": 0.16776058077812195,
+      "learning_rate": 1.9650000000000003e-05,
+      "loss": 0.0313,
+      "step": 4250
+    },
+    {
+      "epoch": 4.868571428571428,
+      "grad_norm": 0.13184590637683868,
+      "learning_rate": 1.957857142857143e-05,
+      "loss": 0.0329,
+      "step": 4260
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.08228486776351929,
+      "learning_rate": 1.9507142857142858e-05,
+      "loss": 0.0304,
+      "step": 4270
+    },
+    {
+      "epoch": 4.8914285714285715,
+      "grad_norm": 0.12768115103244781,
+      "learning_rate": 1.943571428571429e-05,
+      "loss": 0.0308,
+      "step": 4280
+    },
+    {
+      "epoch": 4.902857142857143,
+      "grad_norm": 0.09318844974040985,
+      "learning_rate": 1.9364285714285716e-05,
+      "loss": 0.031,
+      "step": 4290
+    },
+    {
+      "epoch": 4.914285714285715,
+      "grad_norm": 0.07075022161006927,
+      "learning_rate": 1.9292857142857143e-05,
+      "loss": 0.0307,
+      "step": 4300
+    },
+    {
+      "epoch": 4.925714285714285,
+      "grad_norm": 0.0722036212682724,
+      "learning_rate": 1.9221428571428574e-05,
+      "loss": 0.0316,
+      "step": 4310
+    },
+    {
+      "epoch": 4.937142857142857,
+      "grad_norm": 0.0727299302816391,
+      "learning_rate": 1.915e-05,
+      "loss": 0.0309,
+      "step": 4320
+    },
+    {
+      "epoch": 4.948571428571428,
+      "grad_norm": 0.061714913696050644,
+      "learning_rate": 1.907857142857143e-05,
+      "loss": 0.0311,
+      "step": 4330
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.0752352699637413,
+      "learning_rate": 1.9007142857142856e-05,
+      "loss": 0.0303,
+      "step": 4340
+    },
+    {
+      "epoch": 4.9714285714285715,
+      "grad_norm": 0.07696497440338135,
+      "learning_rate": 1.8935714285714287e-05,
+      "loss": 0.0314,
+      "step": 4350
+    },
+    {
+      "epoch": 4.982857142857143,
+      "grad_norm": 0.07310967147350311,
+      "learning_rate": 1.8864285714285714e-05,
+      "loss": 0.0318,
+      "step": 4360
+    },
+    {
+      "epoch": 4.994285714285715,
+      "grad_norm": 0.0728888288140297,
+      "learning_rate": 1.879285714285714e-05,
+      "loss": 0.0305,
+      "step": 4370
+    },
+    {
+      "epoch": 5.0,
+      "eval_bleu": 0.07,
+      "eval_exact_match": 0.0,
+      "eval_loss": 0.028437696397304535,
+      "eval_runtime": 342.7815,
+      "eval_samples_per_second": 17.504,
+      "eval_steps_per_second": 1.094,
+      "step": 4375
+    },
+    {
+      "epoch": 5.005714285714285,
+      "grad_norm": 0.09679137915372849,
+      "learning_rate": 1.8721428571428572e-05,
+      "loss": 0.0325,
+      "step": 4380
+    },
+    {
+      "epoch": 5.017142857142857,
+      "grad_norm": 0.13325752317905426,
+      "learning_rate": 1.865e-05,
+      "loss": 0.0311,
+      "step": 4390
+    },
+    {
+      "epoch": 5.0285714285714285,
+      "grad_norm": 0.0718117207288742,
+      "learning_rate": 1.857857142857143e-05,
+      "loss": 0.0314,
+      "step": 4400
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.07288797199726105,
+      "learning_rate": 1.8507142857142857e-05,
+      "loss": 0.0319,
+      "step": 4410
+    },
+    {
+      "epoch": 5.051428571428572,
+      "grad_norm": 0.0838438868522644,
+      "learning_rate": 1.8435714285714288e-05,
+      "loss": 0.0317,
+      "step": 4420
+    },
+    {
+      "epoch": 5.062857142857143,
+      "grad_norm": 0.11004387587308884,
+      "learning_rate": 1.8364285714285715e-05,
+      "loss": 0.0318,
+      "step": 4430
+    },
+    {
+      "epoch": 5.074285714285715,
+      "grad_norm": 0.06953708082437515,
+      "learning_rate": 1.8292857142857146e-05,
+      "loss": 0.0315,
+      "step": 4440
+    },
+    {
+      "epoch": 5.085714285714285,
+      "grad_norm": 0.08132421225309372,
+      "learning_rate": 1.8221428571428573e-05,
+      "loss": 0.033,
+      "step": 4450
+    },
+    {
+      "epoch": 5.097142857142857,
+      "grad_norm": 0.09449005872011185,
+      "learning_rate": 1.815e-05,
+      "loss": 0.0309,
+      "step": 4460
+    },
+    {
+      "epoch": 5.1085714285714285,
+      "grad_norm": 0.1384226381778717,
+      "learning_rate": 1.807857142857143e-05,
+      "loss": 0.0313,
+      "step": 4470
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.11961528658866882,
+      "learning_rate": 1.800714285714286e-05,
+      "loss": 0.0323,
+      "step": 4480
+    },
+    {
+      "epoch": 5.131428571428572,
+      "grad_norm": 0.07924534380435944,
+      "learning_rate": 1.7935714285714286e-05,
+      "loss": 0.0309,
+      "step": 4490
+    },
+    {
+      "epoch": 5.142857142857143,
+      "grad_norm": 0.09376318007707596,
+      "learning_rate": 1.7864285714285713e-05,
+      "loss": 0.0305,
+      "step": 4500
+    },
+    {
+      "epoch": 5.154285714285714,
+      "grad_norm": 0.13669812679290771,
+      "learning_rate": 1.7792857142857144e-05,
+      "loss": 0.0301,
+      "step": 4510
+    },
+    {
+      "epoch": 5.1657142857142855,
+      "grad_norm": 0.11170660704374313,
+      "learning_rate": 1.772142857142857e-05,
+      "loss": 0.031,
+      "step": 4520
+    },
+    {
+      "epoch": 5.177142857142857,
+      "grad_norm": 0.08319266140460968,
+      "learning_rate": 1.765e-05,
+      "loss": 0.0304,
+      "step": 4530
+    },
+    {
+      "epoch": 5.188571428571429,
+      "grad_norm": 0.08556800335645676,
+      "learning_rate": 1.757857142857143e-05,
+      "loss": 0.0308,
+      "step": 4540
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.10389435291290283,
+      "learning_rate": 1.7507142857142857e-05,
+      "loss": 0.0306,
+      "step": 4550
+    },
+    {
+      "epoch": 5.211428571428572,
+      "grad_norm": 0.07358275353908539,
+      "learning_rate": 1.7435714285714287e-05,
+      "loss": 0.0319,
+      "step": 4560
+    },
+    {
+      "epoch": 5.222857142857142,
+      "grad_norm": 0.07893013209104538,
+      "learning_rate": 1.7364285714285715e-05,
+      "loss": 0.0313,
+      "step": 4570
+    },
+    {
+      "epoch": 5.234285714285714,
+      "grad_norm": 0.0908748209476471,
+      "learning_rate": 1.7292857142857145e-05,
+      "loss": 0.0284,
+      "step": 4580
+    },
+    {
+      "epoch": 5.2457142857142856,
+      "grad_norm": 0.07306382060050964,
+      "learning_rate": 1.7221428571428573e-05,
+      "loss": 0.0309,
+      "step": 4590
+    },
+    {
+      "epoch": 5.257142857142857,
+      "grad_norm": 0.06851589679718018,
+      "learning_rate": 1.7150000000000004e-05,
+      "loss": 0.0301,
+      "step": 4600
+    },
+    {
+      "epoch": 5.268571428571429,
+      "grad_norm": 0.10181226581335068,
+      "learning_rate": 1.707857142857143e-05,
+      "loss": 0.032,
+      "step": 4610
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.08728600293397903,
+      "learning_rate": 1.7007142857142858e-05,
+      "loss": 0.0324,
+      "step": 4620
+    },
+    {
+      "epoch": 5.291428571428572,
+      "grad_norm": 0.09001002460718155,
+      "learning_rate": 1.6935714285714285e-05,
+      "loss": 0.0304,
+      "step": 4630
+    },
+    {
+      "epoch": 5.3028571428571425,
+      "grad_norm": 0.09468323737382889,
+      "learning_rate": 1.6864285714285716e-05,
+      "loss": 0.0303,
+      "step": 4640
+    },
+    {
+      "epoch": 5.314285714285714,
+      "grad_norm": 0.07029464840888977,
+      "learning_rate": 1.6792857142857143e-05,
+      "loss": 0.0313,
+      "step": 4650
+    },
+    {
+      "epoch": 5.325714285714286,
+      "grad_norm": 0.0702148973941803,
+      "learning_rate": 1.672142857142857e-05,
+      "loss": 0.0312,
+      "step": 4660
+    },
+    {
+      "epoch": 5.337142857142857,
+      "grad_norm": 0.06833337247371674,
+      "learning_rate": 1.665e-05,
+      "loss": 0.0319,
+      "step": 4670
+    },
+    {
+      "epoch": 5.348571428571429,
+      "grad_norm": 0.10571616888046265,
+      "learning_rate": 1.657857142857143e-05,
+      "loss": 0.0306,
+      "step": 4680
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.07842124998569489,
+      "learning_rate": 1.6507142857142856e-05,
+      "loss": 0.0306,
+      "step": 4690
+    },
+    {
+      "epoch": 5.371428571428572,
+      "grad_norm": 0.07739285379648209,
+      "learning_rate": 1.6435714285714287e-05,
+      "loss": 0.0304,
+      "step": 4700
+    },
+    {
+      "epoch": 5.382857142857143,
+      "grad_norm": 0.07978440076112747,
+      "learning_rate": 1.6364285714285714e-05,
+      "loss": 0.0307,
+      "step": 4710
+    },
+    {
+      "epoch": 5.394285714285714,
+      "grad_norm": 0.07032512873411179,
+      "learning_rate": 1.629285714285714e-05,
+      "loss": 0.0313,
+      "step": 4720
+    },
+    {
+      "epoch": 5.405714285714286,
+      "grad_norm": 0.06472887843847275,
+      "learning_rate": 1.6221428571428572e-05,
+      "loss": 0.0295,
+      "step": 4730
+    },
+    {
+      "epoch": 5.417142857142857,
+      "grad_norm": 0.08794354647397995,
+      "learning_rate": 1.6150000000000003e-05,
+      "loss": 0.0305,
+      "step": 4740
+    },
+    {
+      "epoch": 5.428571428571429,
+      "grad_norm": 0.08242463320493698,
+      "learning_rate": 1.607857142857143e-05,
+      "loss": 0.0307,
+      "step": 4750
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.0753536969423294,
+      "learning_rate": 1.6007142857142858e-05,
+      "loss": 0.0301,
+      "step": 4760
+    },
+    {
+      "epoch": 5.451428571428571,
+      "grad_norm": 0.06398053467273712,
+      "learning_rate": 1.5935714285714288e-05,
+      "loss": 0.0299,
+      "step": 4770
+    },
+    {
+      "epoch": 5.462857142857143,
+      "grad_norm": 0.07519324123859406,
+      "learning_rate": 1.5864285714285716e-05,
+      "loss": 0.0307,
+      "step": 4780
+    },
+    {
+      "epoch": 5.474285714285714,
+      "grad_norm": 0.0712146982550621,
+      "learning_rate": 1.5792857142857143e-05,
+      "loss": 0.0316,
+      "step": 4790
+    },
+    {
+      "epoch": 5.485714285714286,
+      "grad_norm": 0.06212683022022247,
+      "learning_rate": 1.5721428571428574e-05,
+      "loss": 0.0298,
+      "step": 4800
+    },
+    {
+      "epoch": 5.497142857142857,
+      "grad_norm": 0.060816798359155655,
+      "learning_rate": 1.565e-05,
+      "loss": 0.0316,
+      "step": 4810
+    },
+    {
+      "epoch": 5.508571428571429,
+      "grad_norm": 0.06750506162643433,
+      "learning_rate": 1.5578571428571428e-05,
+      "loss": 0.0307,
+      "step": 4820
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.09742510318756104,
+      "learning_rate": 1.550714285714286e-05,
+      "loss": 0.0326,
+      "step": 4830
+    },
+    {
+      "epoch": 5.531428571428571,
+      "grad_norm": 0.061222221702337265,
+      "learning_rate": 1.5435714285714286e-05,
+      "loss": 0.0289,
+      "step": 4840
+    },
+    {
+      "epoch": 5.542857142857143,
+      "grad_norm": 0.08668530732393265,
+      "learning_rate": 1.5364285714285714e-05,
+      "loss": 0.0315,
+      "step": 4850
+    },
+    {
+      "epoch": 5.554285714285714,
+      "grad_norm": 0.0962333232164383,
+      "learning_rate": 1.5292857142857144e-05,
+      "loss": 0.0316,
+      "step": 4860
+    },
+    {
+      "epoch": 5.565714285714286,
+      "grad_norm": 0.06368666142225266,
+      "learning_rate": 1.5221428571428573e-05,
+      "loss": 0.031,
+      "step": 4870
+    },
+    {
+      "epoch": 5.577142857142857,
+      "grad_norm": 0.0649942010641098,
+      "learning_rate": 1.515e-05,
+      "loss": 0.03,
+      "step": 4880
+    },
+    {
+      "epoch": 5.588571428571429,
+      "grad_norm": 0.08651057630777359,
+      "learning_rate": 1.5078571428571428e-05,
+      "loss": 0.0301,
+      "step": 4890
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.06237189471721649,
+      "learning_rate": 1.5007142857142859e-05,
+      "loss": 0.0305,
+      "step": 4900
+    },
+    {
+      "epoch": 5.611428571428571,
+      "grad_norm": 0.06680475175380707,
+      "learning_rate": 1.4935714285714286e-05,
+      "loss": 0.0311,
+      "step": 4910
+    },
+    {
+      "epoch": 5.622857142857143,
+      "grad_norm": 0.07316766679286957,
+      "learning_rate": 1.4864285714285713e-05,
+      "loss": 0.0308,
+      "step": 4920
+    },
+    {
+      "epoch": 5.634285714285714,
+      "grad_norm": 0.09247075021266937,
+      "learning_rate": 1.4792857142857144e-05,
+      "loss": 0.0304,
+      "step": 4930
+    },
+    {
+      "epoch": 5.645714285714286,
+      "grad_norm": 0.2784198820590973,
+      "learning_rate": 1.4721428571428571e-05,
+      "loss": 0.0294,
+      "step": 4940
+    },
+    {
+      "epoch": 5.6571428571428575,
+      "grad_norm": 0.09237693250179291,
+      "learning_rate": 1.465e-05,
+      "loss": 0.0313,
+      "step": 4950
+    },
+    {
+      "epoch": 5.668571428571429,
+      "grad_norm": 0.07822994887828827,
+      "learning_rate": 1.4578571428571431e-05,
+      "loss": 0.0301,
+      "step": 4960
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 0.10892365127801895,
+      "learning_rate": 1.4507142857142858e-05,
+      "loss": 0.0314,
+      "step": 4970
+    },
+    {
+      "epoch": 5.691428571428571,
+      "grad_norm": 0.07287736982107162,
+      "learning_rate": 1.4435714285714286e-05,
+      "loss": 0.0309,
+      "step": 4980
+    },
+    {
+      "epoch": 5.702857142857143,
+      "grad_norm": 0.0734676644206047,
+      "learning_rate": 1.4364285714285716e-05,
+      "loss": 0.0319,
+      "step": 4990
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 0.06947411596775055,
+      "learning_rate": 1.4292857142857144e-05,
+      "loss": 0.0326,
+      "step": 5000
+    },
+    {
+      "epoch": 5.725714285714286,
+      "grad_norm": 0.08650074899196625,
+      "learning_rate": 1.4221428571428571e-05,
+      "loss": 0.0318,
+      "step": 5010
+    },
+    {
+      "epoch": 5.737142857142857,
+      "grad_norm": 0.059798020869493484,
+      "learning_rate": 1.415e-05,
+      "loss": 0.0315,
+      "step": 5020
+    },
+    {
+      "epoch": 5.748571428571428,
+      "grad_norm": 0.1155044287443161,
+      "learning_rate": 1.4078571428571429e-05,
+      "loss": 0.0305,
+      "step": 5030
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.06892979145050049,
+      "learning_rate": 1.4007142857142858e-05,
+      "loss": 0.0314,
+      "step": 5040
+    },
+    {
+      "epoch": 5.771428571428571,
+      "grad_norm": 0.08073502033948898,
+      "learning_rate": 1.3935714285714285e-05,
+      "loss": 0.0304,
+      "step": 5050
+    },
+    {
+      "epoch": 5.782857142857143,
+      "grad_norm": 0.07314231246709824,
+      "learning_rate": 1.3864285714285716e-05,
+      "loss": 0.03,
+      "step": 5060
+    },
+    {
+      "epoch": 5.7942857142857145,
+      "grad_norm": 0.08082398027181625,
+      "learning_rate": 1.3792857142857143e-05,
+      "loss": 0.0307,
+      "step": 5070
+    },
+    {
+      "epoch": 5.805714285714286,
+      "grad_norm": 0.08921355754137039,
+      "learning_rate": 1.372142857142857e-05,
+      "loss": 0.0315,
+      "step": 5080
+    },
+    {
+      "epoch": 5.817142857142857,
+      "grad_norm": 0.05853136256337166,
+      "learning_rate": 1.3650000000000001e-05,
+      "loss": 0.0297,
+      "step": 5090
+    },
+    {
+      "epoch": 5.828571428571428,
+      "grad_norm": 0.09137912839651108,
+      "learning_rate": 1.3578571428571429e-05,
+      "loss": 0.0304,
+      "step": 5100
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.0859987661242485,
+      "learning_rate": 1.3507142857142858e-05,
+      "loss": 0.0313,
+      "step": 5110
+    },
+    {
+      "epoch": 5.851428571428571,
+      "grad_norm": 0.06775084137916565,
+      "learning_rate": 1.3435714285714287e-05,
+      "loss": 0.029,
+      "step": 5120
+    },
+    {
+      "epoch": 5.862857142857143,
+      "grad_norm": 0.1020933985710144,
+      "learning_rate": 1.3364285714285716e-05,
+      "loss": 0.0318,
+      "step": 5130
+    },
+    {
+      "epoch": 5.8742857142857146,
+      "grad_norm": 0.07362332940101624,
+      "learning_rate": 1.3292857142857143e-05,
+      "loss": 0.0298,
+      "step": 5140
+    },
+    {
+      "epoch": 5.885714285714286,
+      "grad_norm": 0.07384783774614334,
+      "learning_rate": 1.322142857142857e-05,
+      "loss": 0.0325,
+      "step": 5150
+    },
+    {
+      "epoch": 5.897142857142857,
+      "grad_norm": 0.07200124859809875,
+      "learning_rate": 1.3150000000000001e-05,
+      "loss": 0.0296,
+      "step": 5160
+    },
+    {
+      "epoch": 5.908571428571428,
+      "grad_norm": 0.09490924328565598,
+      "learning_rate": 1.3078571428571428e-05,
+      "loss": 0.0321,
+      "step": 5170
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.07857917249202728,
+      "learning_rate": 1.3007142857142856e-05,
+      "loss": 0.031,
+      "step": 5180
+    },
+    {
+      "epoch": 5.9314285714285715,
+      "grad_norm": 0.08538127690553665,
+      "learning_rate": 1.2935714285714286e-05,
+      "loss": 0.0302,
+      "step": 5190
+    },
+    {
+      "epoch": 5.942857142857143,
+      "grad_norm": 0.13723719120025635,
+      "learning_rate": 1.2864285714285716e-05,
+      "loss": 0.0301,
+      "step": 5200
+    },
+    {
+      "epoch": 5.954285714285715,
+      "grad_norm": 0.06784668564796448,
+      "learning_rate": 1.2792857142857143e-05,
+      "loss": 0.0326,
+      "step": 5210
+    },
+    {
+      "epoch": 5.965714285714286,
+      "grad_norm": 0.07009400427341461,
+      "learning_rate": 1.2721428571428574e-05,
+      "loss": 0.0307,
+      "step": 5220
+    },
+    {
+      "epoch": 5.977142857142857,
+      "grad_norm": 0.06899885833263397,
+      "learning_rate": 1.2650000000000001e-05,
+      "loss": 0.0321,
+      "step": 5230
+    },
+    {
+      "epoch": 5.988571428571428,
+      "grad_norm": 0.06912403553724289,
+      "learning_rate": 1.2578571428571428e-05,
+      "loss": 0.031,
+      "step": 5240
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.0813736692070961,
+      "learning_rate": 1.2507142857142859e-05,
+      "loss": 0.0317,
+      "step": 5250
+    },
+    {
+      "epoch": 6.0,
+      "eval_bleu": 0.07,
+      "eval_exact_match": 0.0,
+      "eval_loss": 0.028217371553182602,
+      "eval_runtime": 343.1777,
+      "eval_samples_per_second": 17.484,
+      "eval_steps_per_second": 1.093,
+      "step": 5250
+    },
+    {
+      "epoch": 6.011428571428572,
+      "grad_norm": 0.06998402625322342,
+      "learning_rate": 1.2435714285714286e-05,
+      "loss": 0.0307,
+      "step": 5260
+    },
+    {
+      "epoch": 6.022857142857143,
+      "grad_norm": 0.05968615785241127,
+      "learning_rate": 1.2364285714285714e-05,
+      "loss": 0.0318,
+      "step": 5270
+    },
+    {
+      "epoch": 6.034285714285715,
+      "grad_norm": 0.09144943207502365,
+      "learning_rate": 1.2292857142857144e-05,
+      "loss": 0.0319,
+      "step": 5280
+    },
+    {
+      "epoch": 6.045714285714285,
+      "grad_norm": 0.07859035581350327,
+      "learning_rate": 1.2221428571428573e-05,
+      "loss": 0.0304,
+      "step": 5290
+    },
+    {
+      "epoch": 6.057142857142857,
+      "grad_norm": 0.06295861303806305,
+      "learning_rate": 1.215e-05,
+      "loss": 0.0312,
+      "step": 5300
+    },
+    {
+      "epoch": 6.0685714285714285,
+      "grad_norm": 0.12032752484083176,
+      "learning_rate": 1.207857142857143e-05,
+      "loss": 0.0308,
+      "step": 5310
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 0.0782964825630188,
+      "learning_rate": 1.2007142857142857e-05,
+      "loss": 0.0317,
+      "step": 5320
+    },
+    {
+      "epoch": 6.091428571428572,
+      "grad_norm": 0.05909128487110138,
+      "learning_rate": 1.1935714285714286e-05,
+      "loss": 0.0289,
+      "step": 5330
+    },
+    {
+      "epoch": 6.102857142857143,
+      "grad_norm": 0.0750165656208992,
+      "learning_rate": 1.1864285714285715e-05,
+      "loss": 0.0307,
+      "step": 5340
+    },
+    {
+      "epoch": 6.114285714285714,
+      "grad_norm": 0.07260541617870331,
+      "learning_rate": 1.1792857142857142e-05,
+      "loss": 0.0307,
+      "step": 5350
+    },
+    {
+      "epoch": 6.1257142857142854,
+      "grad_norm": 0.07592778652906418,
+      "learning_rate": 1.1721428571428571e-05,
+      "loss": 0.0309,
+      "step": 5360
+    },
+    {
+      "epoch": 6.137142857142857,
+      "grad_norm": 0.08206040412187576,
+      "learning_rate": 1.1650000000000002e-05,
+      "loss": 0.0302,
+      "step": 5370
+    },
+    {
+      "epoch": 6.148571428571429,
+      "grad_norm": 0.08673234283924103,
+      "learning_rate": 1.157857142857143e-05,
+      "loss": 0.0297,
+      "step": 5380
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.06839073449373245,
+      "learning_rate": 1.1507142857142858e-05,
+      "loss": 0.0304,
+      "step": 5390
+    },
+    {
+      "epoch": 6.171428571428572,
+      "grad_norm": 0.0769868940114975,
+      "learning_rate": 1.1435714285714286e-05,
+      "loss": 0.0319,
+      "step": 5400
+    },
+    {
+      "epoch": 6.182857142857143,
+      "grad_norm": 0.07529103755950928,
+      "learning_rate": 1.1364285714285715e-05,
+      "loss": 0.0312,
+      "step": 5410
+    },
+    {
+      "epoch": 6.194285714285714,
+      "grad_norm": 0.06735879182815552,
+      "learning_rate": 1.1292857142857144e-05,
+      "loss": 0.031,
+      "step": 5420
+    },
+    {
+      "epoch": 6.2057142857142855,
+      "grad_norm": 0.06926565617322922,
+      "learning_rate": 1.1221428571428571e-05,
+      "loss": 0.031,
+      "step": 5430
+    },
+    {
+      "epoch": 6.217142857142857,
+      "grad_norm": 0.09316926449537277,
+      "learning_rate": 1.115e-05,
+      "loss": 0.0319,
+      "step": 5440
+    },
+    {
+      "epoch": 6.228571428571429,
+      "grad_norm": 0.07555294036865234,
+      "learning_rate": 1.1078571428571429e-05,
+      "loss": 0.0303,
+      "step": 5450
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.08294390887022018,
+      "learning_rate": 1.1007142857142858e-05,
+      "loss": 0.0297,
+      "step": 5460
+    },
+    {
+      "epoch": 6.251428571428572,
+      "grad_norm": 0.06981640309095383,
+      "learning_rate": 1.0935714285714287e-05,
+      "loss": 0.0305,
+      "step": 5470
+    },
+    {
+      "epoch": 6.2628571428571425,
+      "grad_norm": 0.07442454993724823,
+      "learning_rate": 1.0864285714285714e-05,
+      "loss": 0.031,
+      "step": 5480
+    },
+    {
+      "epoch": 6.274285714285714,
+      "grad_norm": 0.12811383605003357,
+      "learning_rate": 1.0792857142857143e-05,
+      "loss": 0.0299,
+      "step": 5490
+    },
+    {
+      "epoch": 6.285714285714286,
+      "grad_norm": 0.0792556032538414,
+      "learning_rate": 1.0721428571428572e-05,
+      "loss": 0.0315,
+      "step": 5500
+    },
+    {
+      "epoch": 6.297142857142857,
+      "grad_norm": 0.08356291055679321,
+      "learning_rate": 1.065e-05,
+      "loss": 0.0319,
+      "step": 5510
+    },
+    {
+      "epoch": 6.308571428571429,
+      "grad_norm": 0.10948007553815842,
+      "learning_rate": 1.0578571428571429e-05,
+      "loss": 0.0293,
+      "step": 5520
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.0714062824845314,
+      "learning_rate": 1.0507142857142858e-05,
+      "loss": 0.0303,
+      "step": 5530
+    },
+    {
+      "epoch": 6.331428571428571,
+      "grad_norm": 0.058776937425136566,
+      "learning_rate": 1.0435714285714287e-05,
+      "loss": 0.0304,
+      "step": 5540
+    },
+    {
+      "epoch": 6.3428571428571425,
+      "grad_norm": 0.07753952592611313,
+      "learning_rate": 1.0364285714285716e-05,
+      "loss": 0.0311,
+      "step": 5550
+    },
+    {
+      "epoch": 6.354285714285714,
+      "grad_norm": 0.08094929158687592,
+      "learning_rate": 1.0292857142857143e-05,
+      "loss": 0.03,
+      "step": 5560
+    },
+    {
+      "epoch": 6.365714285714286,
+      "grad_norm": 0.07799239456653595,
+      "learning_rate": 1.0221428571428572e-05,
+      "loss": 0.0316,
+      "step": 5570
+    },
+    {
+      "epoch": 6.377142857142857,
+      "grad_norm": 0.07615884393453598,
+      "learning_rate": 1.0150000000000001e-05,
+      "loss": 0.0317,
+      "step": 5580
+    },
+    {
+      "epoch": 6.388571428571429,
+      "grad_norm": 0.06691175699234009,
+      "learning_rate": 1.0078571428571428e-05,
+      "loss": 0.0312,
+      "step": 5590
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.07050418853759766,
+      "learning_rate": 1.0007142857142857e-05,
+      "loss": 0.0314,
+      "step": 5600
+    },
+    {
+      "epoch": 6.411428571428571,
+      "grad_norm": 0.07742037624120712,
+      "learning_rate": 9.935714285714286e-06,
+      "loss": 0.0302,
+      "step": 5610
+    },
+    {
+      "epoch": 6.422857142857143,
+      "grad_norm": 0.07048898935317993,
+      "learning_rate": 9.864285714285715e-06,
+      "loss": 0.0305,
+      "step": 5620
+    },
+    {
+      "epoch": 6.434285714285714,
+      "grad_norm": 0.07033011317253113,
+      "learning_rate": 9.792857142857144e-06,
+      "loss": 0.031,
+      "step": 5630
+    },
+    {
+      "epoch": 6.445714285714286,
+      "grad_norm": 0.1080131009221077,
+      "learning_rate": 9.721428571428572e-06,
+      "loss": 0.032,
+      "step": 5640
+    },
+    {
+      "epoch": 6.457142857142857,
+      "grad_norm": 0.08267289400100708,
+      "learning_rate": 9.65e-06,
+      "loss": 0.0302,
+      "step": 5650
+    },
+    {
+      "epoch": 6.468571428571429,
+      "grad_norm": 0.058059051632881165,
+      "learning_rate": 9.578571428571428e-06,
+      "loss": 0.0304,
+      "step": 5660
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.07850830256938934,
+      "learning_rate": 9.507142857142857e-06,
+      "loss": 0.0308,
+      "step": 5670
+    },
+    {
+      "epoch": 6.491428571428571,
+      "grad_norm": 0.08184772729873657,
+      "learning_rate": 9.435714285714286e-06,
+      "loss": 0.0306,
+      "step": 5680
+    },
+    {
+      "epoch": 6.502857142857143,
+      "grad_norm": 0.06574039161205292,
+      "learning_rate": 9.364285714285715e-06,
+      "loss": 0.0299,
+      "step": 5690
+    },
+    {
+      "epoch": 6.514285714285714,
+      "grad_norm": 0.24083204567432404,
+      "learning_rate": 9.292857142857144e-06,
+      "loss": 0.0316,
+      "step": 5700
+    },
+    {
+      "epoch": 6.525714285714286,
+      "grad_norm": 0.09220234304666519,
+      "learning_rate": 9.221428571428573e-06,
+      "loss": 0.0284,
+      "step": 5710
+    },
+    {
+      "epoch": 6.537142857142857,
+      "grad_norm": 0.07189635932445526,
+      "learning_rate": 9.15e-06,
+      "loss": 0.0331,
+      "step": 5720
+    },
+    {
+      "epoch": 6.548571428571429,
+      "grad_norm": 0.10544195771217346,
+      "learning_rate": 9.07857142857143e-06,
+      "loss": 0.0311,
+      "step": 5730
+    },
+    {
+      "epoch": 6.5600000000000005,
+      "grad_norm": 0.07962433993816376,
+      "learning_rate": 9.007142857142857e-06,
+      "loss": 0.0309,
+      "step": 5740
+    },
+    {
+      "epoch": 6.571428571428571,
+      "grad_norm": 0.07586560398340225,
+      "learning_rate": 8.935714285714286e-06,
+      "loss": 0.0282,
+      "step": 5750
+    },
+    {
+      "epoch": 6.582857142857143,
+      "grad_norm": 0.07862204313278198,
+      "learning_rate": 8.864285714285715e-06,
+      "loss": 0.0313,
+      "step": 5760
+    },
+    {
+      "epoch": 6.594285714285714,
+      "grad_norm": 0.05962090566754341,
+      "learning_rate": 8.792857142857142e-06,
+      "loss": 0.0301,
+      "step": 5770
+    },
+    {
+      "epoch": 6.605714285714286,
+      "grad_norm": 0.06269550323486328,
+      "learning_rate": 8.721428571428573e-06,
+      "loss": 0.0297,
+      "step": 5780
+    },
+    {
+      "epoch": 6.617142857142857,
+      "grad_norm": 0.11576901376247406,
+      "learning_rate": 8.65e-06,
+      "loss": 0.0316,
+      "step": 5790
+    },
+    {
+      "epoch": 6.628571428571428,
+      "grad_norm": 0.06566612422466278,
+      "learning_rate": 8.57857142857143e-06,
+      "loss": 0.0304,
+      "step": 5800
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.06394005566835403,
+      "learning_rate": 8.507142857142858e-06,
+      "loss": 0.0309,
+      "step": 5810
+    },
+    {
+      "epoch": 6.651428571428571,
+      "grad_norm": 0.06856530159711838,
+      "learning_rate": 8.435714285714286e-06,
+      "loss": 0.0313,
+      "step": 5820
+    },
+    {
+      "epoch": 6.662857142857143,
+      "grad_norm": 0.07678523659706116,
+      "learning_rate": 8.364285714285715e-06,
+      "loss": 0.0324,
+      "step": 5830
+    },
+    {
+      "epoch": 6.674285714285714,
+      "grad_norm": 0.10768764466047287,
+      "learning_rate": 8.292857142857144e-06,
+      "loss": 0.03,
+      "step": 5840
+    },
+    {
+      "epoch": 6.685714285714286,
+      "grad_norm": 0.13758371770381927,
+      "learning_rate": 8.221428571428571e-06,
+      "loss": 0.032,
+      "step": 5850
+    },
+    {
+      "epoch": 6.6971428571428575,
+      "grad_norm": 0.06116107106208801,
+      "learning_rate": 8.15e-06,
+      "loss": 0.0323,
+      "step": 5860
+    },
+    {
+      "epoch": 6.708571428571428,
+      "grad_norm": 0.058115601539611816,
+      "learning_rate": 8.078571428571429e-06,
+      "loss": 0.0308,
+      "step": 5870
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.061137229204177856,
+      "learning_rate": 8.007142857142858e-06,
+      "loss": 0.029,
+      "step": 5880
+    },
+    {
+      "epoch": 6.731428571428571,
+      "grad_norm": 0.13177888095378876,
+      "learning_rate": 7.935714285714287e-06,
+      "loss": 0.0305,
+      "step": 5890
+    },
+    {
+      "epoch": 6.742857142857143,
+      "grad_norm": 0.06651601195335388,
+      "learning_rate": 7.864285714285714e-06,
+      "loss": 0.0297,
+      "step": 5900
+    },
+    {
+      "epoch": 6.7542857142857144,
+      "grad_norm": 0.06756425648927689,
+      "learning_rate": 7.792857142857143e-06,
+      "loss": 0.0317,
+      "step": 5910
+    },
+    {
+      "epoch": 6.765714285714286,
+      "grad_norm": 0.11434056609869003,
+      "learning_rate": 7.72142857142857e-06,
+      "loss": 0.0294,
+      "step": 5920
+    },
+    {
+      "epoch": 6.777142857142858,
+      "grad_norm": 0.06794974952936172,
+      "learning_rate": 7.65e-06,
+      "loss": 0.031,
+      "step": 5930
+    },
+    {
+      "epoch": 6.788571428571428,
+      "grad_norm": 0.06978090107440948,
+      "learning_rate": 7.5785714285714295e-06,
+      "loss": 0.03,
+      "step": 5940
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.06427010148763657,
+      "learning_rate": 7.507142857142857e-06,
+      "loss": 0.031,
+      "step": 5950
+    },
+    {
+      "epoch": 6.811428571428571,
+      "grad_norm": 0.06627924740314484,
+      "learning_rate": 7.435714285714286e-06,
+      "loss": 0.0295,
+      "step": 5960
+    },
+    {
+      "epoch": 6.822857142857143,
+      "grad_norm": 0.07513666898012161,
+      "learning_rate": 7.364285714285716e-06,
+      "loss": 0.0309,
+      "step": 5970
+    },
+    {
+      "epoch": 6.8342857142857145,
+      "grad_norm": 0.08213704079389572,
+      "learning_rate": 7.292857142857143e-06,
+      "loss": 0.0321,
+      "step": 5980
+    },
+    {
+      "epoch": 6.845714285714286,
+      "grad_norm": 0.059965796768665314,
+      "learning_rate": 7.221428571428572e-06,
+      "loss": 0.0293,
+      "step": 5990
+    },
+    {
+      "epoch": 6.857142857142857,
+      "grad_norm": 0.10298174619674683,
+      "learning_rate": 7.15e-06,
+      "loss": 0.0311,
+      "step": 6000
+    },
+    {
+      "epoch": 6.868571428571428,
+      "grad_norm": 0.08990275859832764,
+      "learning_rate": 7.078571428571429e-06,
+      "loss": 0.0302,
+      "step": 6010
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.06812401860952377,
+      "learning_rate": 7.007142857142858e-06,
+      "loss": 0.03,
+      "step": 6020
+    },
+    {
+      "epoch": 6.8914285714285715,
+      "grad_norm": 0.08880835771560669,
+      "learning_rate": 6.9357142857142855e-06,
+      "loss": 0.0307,
+      "step": 6030
+    },
+    {
+      "epoch": 6.902857142857143,
+      "grad_norm": 0.06701825559139252,
+      "learning_rate": 6.8642857142857145e-06,
+      "loss": 0.0298,
+      "step": 6040
+    },
+    {
+      "epoch": 6.914285714285715,
+      "grad_norm": 0.11159258335828781,
+      "learning_rate": 6.7928571428571435e-06,
+      "loss": 0.0305,
+      "step": 6050
+    },
+    {
+      "epoch": 6.925714285714285,
+      "grad_norm": 0.07996679842472076,
+      "learning_rate": 6.721428571428572e-06,
+      "loss": 0.0309,
+      "step": 6060
+    },
+    {
+      "epoch": 6.937142857142857,
+      "grad_norm": 0.07035525143146515,
+      "learning_rate": 6.650000000000001e-06,
+      "loss": 0.0299,
+      "step": 6070
+    },
+    {
+      "epoch": 6.948571428571428,
+      "grad_norm": 0.07342403382062912,
+      "learning_rate": 6.578571428571428e-06,
+      "loss": 0.0303,
+      "step": 6080
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.05549318343400955,
+      "learning_rate": 6.507142857142858e-06,
+      "loss": 0.0295,
+      "step": 6090
+    },
+    {
+      "epoch": 6.9714285714285715,
+      "grad_norm": 0.054844338446855545,
+      "learning_rate": 6.435714285714287e-06,
+      "loss": 0.0303,
+      "step": 6100
+    },
+    {
+      "epoch": 6.982857142857143,
+      "grad_norm": 0.07505437731742859,
+      "learning_rate": 6.364285714285714e-06,
+      "loss": 0.0321,
+      "step": 6110
+    },
+    {
+      "epoch": 6.994285714285715,
+      "grad_norm": 0.07832371443510056,
+      "learning_rate": 6.292857142857143e-06,
+      "loss": 0.031,
+      "step": 6120
+    },
+    {
+      "epoch": 7.0,
+      "eval_bleu": 0.07,
+      "eval_exact_match": 0.0,
+      "eval_loss": 0.028141871094703674,
+      "eval_runtime": 343.5619,
+      "eval_samples_per_second": 17.464,
+      "eval_steps_per_second": 1.092,
+      "step": 6125
+    },
+    {
+      "epoch": 7.005714285714285,
+      "grad_norm": 0.06968718022108078,
+      "learning_rate": 6.221428571428571e-06,
+      "loss": 0.0298,
+      "step": 6130
+    },
+    {
+      "epoch": 7.017142857142857,
+      "grad_norm": 0.06143512949347496,
+      "learning_rate": 6.15e-06,
+      "loss": 0.0296,
+      "step": 6140
+    },
+    {
+      "epoch": 7.0285714285714285,
+      "grad_norm": 0.06617894023656845,
+      "learning_rate": 6.0785714285714286e-06,
+      "loss": 0.0304,
+      "step": 6150
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.08548376709222794,
+      "learning_rate": 6.007142857142858e-06,
+      "loss": 0.0316,
+      "step": 6160
+    },
+    {
+      "epoch": 7.051428571428572,
+      "grad_norm": 0.10409008711576462,
+      "learning_rate": 5.935714285714286e-06,
+      "loss": 0.0323,
+      "step": 6170
+    },
+    {
+      "epoch": 7.062857142857143,
+      "grad_norm": 0.06821285933256149,
+      "learning_rate": 5.864285714285715e-06,
+      "loss": 0.0308,
+      "step": 6180
+    },
+    {
+      "epoch": 7.074285714285715,
+      "grad_norm": 0.06576462835073471,
+      "learning_rate": 5.792857142857143e-06,
+      "loss": 0.0286,
+      "step": 6190
+    },
+    {
+      "epoch": 7.085714285714285,
+      "grad_norm": 0.08164945989847183,
+      "learning_rate": 5.721428571428572e-06,
+      "loss": 0.0303,
+      "step": 6200
+    },
+    {
+      "epoch": 7.097142857142857,
+      "grad_norm": 0.060051556676626205,
+      "learning_rate": 5.65e-06,
+      "loss": 0.0296,
+      "step": 6210
+    },
+    {
+      "epoch": 7.1085714285714285,
+      "grad_norm": 0.05957673490047455,
+      "learning_rate": 5.578571428571429e-06,
+      "loss": 0.0311,
+      "step": 6220
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.055859100073575974,
+      "learning_rate": 5.507142857142857e-06,
+      "loss": 0.0293,
+      "step": 6230
+    },
+    {
+      "epoch": 7.131428571428572,
+      "grad_norm": 0.07006938755512238,
+      "learning_rate": 5.4357142857142855e-06,
+      "loss": 0.0308,
+      "step": 6240
+    },
+    {
+      "epoch": 7.142857142857143,
+      "grad_norm": 0.055445924401283264,
+      "learning_rate": 5.3642857142857145e-06,
+      "loss": 0.03,
+      "step": 6250
+    },
+    {
+      "epoch": 7.154285714285714,
+      "grad_norm": 0.07507022470235825,
+      "learning_rate": 5.2928571428571435e-06,
+      "loss": 0.0319,
+      "step": 6260
+    },
+    {
+      "epoch": 7.1657142857142855,
+      "grad_norm": 0.05871541053056717,
+      "learning_rate": 5.221428571428572e-06,
+      "loss": 0.0299,
+      "step": 6270
+    },
+    {
+      "epoch": 7.177142857142857,
+      "grad_norm": 0.05965025722980499,
+      "learning_rate": 5.15e-06,
+      "loss": 0.0299,
+      "step": 6280
+    },
+    {
+      "epoch": 7.188571428571429,
+      "grad_norm": 0.07607734203338623,
+      "learning_rate": 5.078571428571429e-06,
+      "loss": 0.0304,
+      "step": 6290
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.0674389973282814,
+      "learning_rate": 5.007142857142858e-06,
+      "loss": 0.0302,
+      "step": 6300
+    },
+    {
+      "epoch": 7.211428571428572,
+      "grad_norm": 0.14311166107654572,
+      "learning_rate": 4.935714285714286e-06,
+      "loss": 0.0317,
+      "step": 6310
+    },
+    {
+      "epoch": 7.222857142857142,
+      "grad_norm": 0.06374186277389526,
+      "learning_rate": 4.864285714285714e-06,
+      "loss": 0.0291,
+      "step": 6320
+    },
+    {
+      "epoch": 7.234285714285714,
+      "grad_norm": 0.08698999881744385,
+      "learning_rate": 4.792857142857143e-06,
+      "loss": 0.0291,
+      "step": 6330
+    },
+    {
+      "epoch": 7.2457142857142856,
+      "grad_norm": 0.06159118562936783,
+      "learning_rate": 4.721428571428572e-06,
+      "loss": 0.0302,
+      "step": 6340
+    },
+    {
+      "epoch": 7.257142857142857,
+      "grad_norm": 0.07281281054019928,
+      "learning_rate": 4.65e-06,
+      "loss": 0.0302,
+      "step": 6350
+    },
+    {
+      "epoch": 7.268571428571429,
+      "grad_norm": 0.09171772003173828,
+      "learning_rate": 4.5785714285714285e-06,
+      "loss": 0.0298,
+      "step": 6360
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.07112763822078705,
+      "learning_rate": 4.5071428571428576e-06,
+      "loss": 0.0289,
+      "step": 6370
+    },
+    {
+      "epoch": 7.291428571428572,
+      "grad_norm": 0.06696333736181259,
+      "learning_rate": 4.435714285714286e-06,
+      "loss": 0.0301,
+      "step": 6380
+    },
+    {
+      "epoch": 7.3028571428571425,
+      "grad_norm": 0.0801989883184433,
+      "learning_rate": 4.364285714285715e-06,
+      "loss": 0.0313,
+      "step": 6390
+    },
+    {
+      "epoch": 7.314285714285714,
+      "grad_norm": 0.06183260679244995,
+      "learning_rate": 4.292857142857143e-06,
+      "loss": 0.0305,
+      "step": 6400
+    },
+    {
+      "epoch": 7.325714285714286,
+      "grad_norm": 0.06928607076406479,
+      "learning_rate": 4.221428571428571e-06,
+      "loss": 0.0307,
+      "step": 6410
+    },
+    {
+      "epoch": 7.337142857142857,
+      "grad_norm": 0.06948118656873703,
+      "learning_rate": 4.15e-06,
+      "loss": 0.0307,
+      "step": 6420
+    },
+    {
+      "epoch": 7.348571428571429,
+      "grad_norm": 0.06645726412534714,
+      "learning_rate": 4.078571428571429e-06,
+      "loss": 0.0303,
+      "step": 6430
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0641820877790451,
+      "learning_rate": 4.007142857142857e-06,
+      "loss": 0.0291,
+      "step": 6440
+    },
+    {
+      "epoch": 7.371428571428572,
+      "grad_norm": 0.07766878604888916,
+      "learning_rate": 3.935714285714285e-06,
+      "loss": 0.0308,
+      "step": 6450
+    },
+    {
+      "epoch": 7.382857142857143,
+      "grad_norm": 0.07369673252105713,
+      "learning_rate": 3.8642857142857144e-06,
+      "loss": 0.0308,
+      "step": 6460
+    },
+    {
+      "epoch": 7.394285714285714,
+      "grad_norm": 0.0729992687702179,
+      "learning_rate": 3.792857142857143e-06,
+      "loss": 0.0292,
+      "step": 6470
+    },
+    {
+      "epoch": 7.405714285714286,
+      "grad_norm": 0.07691550254821777,
+      "learning_rate": 3.7214285714285716e-06,
+      "loss": 0.0294,
+      "step": 6480
+    },
+    {
+      "epoch": 7.417142857142857,
+      "grad_norm": 0.07256300747394562,
+      "learning_rate": 3.6499999999999998e-06,
+      "loss": 0.0298,
+      "step": 6490
+    },
+    {
+      "epoch": 7.428571428571429,
+      "grad_norm": 0.12671434879302979,
+      "learning_rate": 3.5785714285714292e-06,
+      "loss": 0.0299,
+      "step": 6500
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.09428369998931885,
+      "learning_rate": 3.5071428571428574e-06,
+      "loss": 0.028,
+      "step": 6510
+    },
+    {
+      "epoch": 7.451428571428571,
+      "grad_norm": 0.16297385096549988,
+      "learning_rate": 3.435714285714286e-06,
+      "loss": 0.0305,
+      "step": 6520
+    },
+    {
+      "epoch": 7.462857142857143,
+      "grad_norm": 0.06254955381155014,
+      "learning_rate": 3.364285714285714e-06,
+      "loss": 0.0307,
+      "step": 6530
+    },
+    {
+      "epoch": 7.474285714285714,
+      "grad_norm": 0.13530980050563812,
+      "learning_rate": 3.2928571428571427e-06,
+      "loss": 0.0309,
+      "step": 6540
+    },
+    {
+      "epoch": 7.485714285714286,
+      "grad_norm": 0.06885895878076553,
+      "learning_rate": 3.2214285714285717e-06,
+      "loss": 0.0292,
+      "step": 6550
+    },
+    {
+      "epoch": 7.497142857142857,
+      "grad_norm": 0.06638657301664352,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.03,
+      "step": 6560
+    },
+    {
+      "epoch": 7.508571428571429,
+      "grad_norm": 0.13830330967903137,
+      "learning_rate": 3.0785714285714285e-06,
+      "loss": 0.0322,
+      "step": 6570
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.09429346024990082,
+      "learning_rate": 3.0071428571428575e-06,
+      "loss": 0.0304,
+      "step": 6580
+    },
+    {
+      "epoch": 7.531428571428571,
+      "grad_norm": 0.0697278156876564,
+      "learning_rate": 2.9357142857142857e-06,
+      "loss": 0.0308,
+      "step": 6590
+    },
+    {
+      "epoch": 7.542857142857143,
+      "grad_norm": 0.06435400992631912,
+      "learning_rate": 2.8642857142857147e-06,
+      "loss": 0.0301,
+      "step": 6600
+    },
+    {
+      "epoch": 7.554285714285714,
+      "grad_norm": 0.08088187128305435,
+      "learning_rate": 2.792857142857143e-06,
+      "loss": 0.031,
+      "step": 6610
+    },
+    {
+      "epoch": 7.565714285714286,
+      "grad_norm": 0.06981631368398666,
+      "learning_rate": 2.7214285714285714e-06,
+      "loss": 0.0301,
+      "step": 6620
+    },
+    {
+      "epoch": 7.577142857142857,
+      "grad_norm": 0.10085096210241318,
+      "learning_rate": 2.65e-06,
+      "loss": 0.0311,
+      "step": 6630
+    },
+    {
+      "epoch": 7.588571428571429,
+      "grad_norm": 0.15587645769119263,
+      "learning_rate": 2.5785714285714286e-06,
+      "loss": 0.0295,
+      "step": 6640
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.21229258179664612,
+      "learning_rate": 2.5071428571428572e-06,
+      "loss": 0.032,
+      "step": 6650
+    },
+    {
+      "epoch": 7.611428571428571,
+      "grad_norm": 0.0634523332118988,
+      "learning_rate": 2.435714285714286e-06,
+      "loss": 0.0297,
+      "step": 6660
+    },
+    {
+      "epoch": 7.622857142857143,
+      "grad_norm": 0.08393410593271255,
+      "learning_rate": 2.3642857142857144e-06,
+      "loss": 0.0304,
+      "step": 6670
+    },
+    {
+      "epoch": 7.634285714285714,
+      "grad_norm": 0.07760918885469437,
+      "learning_rate": 2.292857142857143e-06,
+      "loss": 0.0298,
+      "step": 6680
+    },
+    {
+      "epoch": 7.645714285714286,
+      "grad_norm": 0.08117687702178955,
+      "learning_rate": 2.2214285714285716e-06,
+      "loss": 0.031,
+      "step": 6690
+    },
+    {
+      "epoch": 7.6571428571428575,
+      "grad_norm": 0.0747959315776825,
+      "learning_rate": 2.1499999999999997e-06,
+      "loss": 0.0288,
+      "step": 6700
+    },
+    {
+      "epoch": 7.668571428571429,
+      "grad_norm": 0.15194258093833923,
+      "learning_rate": 2.0785714285714288e-06,
+      "loss": 0.0326,
+      "step": 6710
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.0686679482460022,
+      "learning_rate": 2.007142857142857e-06,
+      "loss": 0.0301,
+      "step": 6720
+    },
+    {
+      "epoch": 7.691428571428571,
+      "grad_norm": 0.07142027467489243,
+      "learning_rate": 1.935714285714286e-06,
+      "loss": 0.031,
+      "step": 6730
+    },
+    {
+      "epoch": 7.702857142857143,
+      "grad_norm": 0.07775787264108658,
+      "learning_rate": 1.8642857142857143e-06,
+      "loss": 0.0291,
+      "step": 6740
+    },
+    {
+      "epoch": 7.714285714285714,
+      "grad_norm": 0.07410398125648499,
+      "learning_rate": 1.792857142857143e-06,
+      "loss": 0.0308,
+      "step": 6750
+    },
+    {
+      "epoch": 7.725714285714286,
+      "grad_norm": 0.07670729607343674,
+      "learning_rate": 1.7214285714285715e-06,
+      "loss": 0.0304,
+      "step": 6760
+    },
+    {
+      "epoch": 7.737142857142857,
+      "grad_norm": 0.07260814309120178,
+      "learning_rate": 1.65e-06,
+      "loss": 0.0302,
+      "step": 6770
+    },
+    {
+      "epoch": 7.748571428571428,
+      "grad_norm": 0.06961925327777863,
+      "learning_rate": 1.5785714285714285e-06,
+      "loss": 0.03,
+      "step": 6780
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.09042750298976898,
+      "learning_rate": 1.507142857142857e-06,
+      "loss": 0.0304,
+      "step": 6790
+    },
+    {
+      "epoch": 7.771428571428571,
+      "grad_norm": 0.07849057763814926,
+      "learning_rate": 1.4357142857142856e-06,
+      "loss": 0.0313,
+      "step": 6800
+    },
+    {
+      "epoch": 7.782857142857143,
+      "grad_norm": 0.0724598690867424,
+      "learning_rate": 1.3642857142857142e-06,
+      "loss": 0.0296,
+      "step": 6810
+    },
+    {
+      "epoch": 7.7942857142857145,
+      "grad_norm": 0.0791030079126358,
+      "learning_rate": 1.2928571428571428e-06,
+      "loss": 0.0325,
+      "step": 6820
+    },
+    {
+      "epoch": 7.805714285714286,
+      "grad_norm": 0.06429225951433182,
+      "learning_rate": 1.2214285714285714e-06,
+      "loss": 0.0299,
+      "step": 6830
+    },
+    {
+      "epoch": 7.817142857142857,
+      "grad_norm": 0.06754273921251297,
+      "learning_rate": 1.15e-06,
+      "loss": 0.0295,
+      "step": 6840
+    },
+    {
+      "epoch": 7.828571428571428,
+      "grad_norm": 0.15404383838176727,
+      "learning_rate": 1.0785714285714286e-06,
+      "loss": 0.0307,
+      "step": 6850
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.07046996057033539,
+      "learning_rate": 1.0071428571428572e-06,
+      "loss": 0.029,
+      "step": 6860
+    },
+    {
+      "epoch": 7.851428571428571,
+      "grad_norm": 0.09521915763616562,
+      "learning_rate": 9.357142857142858e-07,
+      "loss": 0.0295,
+      "step": 6870
+    },
+    {
+      "epoch": 7.862857142857143,
+      "grad_norm": 0.08006428182125092,
+      "learning_rate": 8.642857142857144e-07,
+      "loss": 0.0282,
+      "step": 6880
+    },
+    {
+      "epoch": 7.8742857142857146,
+      "grad_norm": 0.08232620358467102,
+      "learning_rate": 7.92857142857143e-07,
+      "loss": 0.0304,
+      "step": 6890
+    },
+    {
+      "epoch": 7.885714285714286,
+      "grad_norm": 0.0687454491853714,
+      "learning_rate": 7.214285714285714e-07,
+      "loss": 0.0301,
+      "step": 6900
+    },
+    {
+      "epoch": 7.897142857142857,
+      "grad_norm": 0.11050191521644592,
+      "learning_rate": 6.5e-07,
+      "loss": 0.0304,
+      "step": 6910
+    },
+    {
+      "epoch": 7.908571428571428,
+      "grad_norm": 0.06012315675616264,
+      "learning_rate": 5.785714285714286e-07,
+      "loss": 0.0309,
+      "step": 6920
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.07466232031583786,
+      "learning_rate": 5.071428571428572e-07,
+      "loss": 0.0312,
+      "step": 6930
+    },
+    {
+      "epoch": 7.9314285714285715,
+      "grad_norm": 0.07779434323310852,
+      "learning_rate": 4.3571428571428574e-07,
+      "loss": 0.0315,
+      "step": 6940
+    },
+    {
+      "epoch": 7.942857142857143,
+      "grad_norm": 0.05634528398513794,
+      "learning_rate": 3.6428571428571433e-07,
+      "loss": 0.0296,
+      "step": 6950
+    },
+    {
+      "epoch": 7.954285714285715,
+      "grad_norm": 0.0750165656208992,
+      "learning_rate": 2.9285714285714287e-07,
+      "loss": 0.0315,
+      "step": 6960
+    },
+    {
+      "epoch": 7.965714285714286,
+      "grad_norm": 0.0592956617474556,
+      "learning_rate": 2.2142857142857143e-07,
+      "loss": 0.0306,
+      "step": 6970
+    },
+    {
+      "epoch": 7.977142857142857,
+      "grad_norm": 0.06860347837209702,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 0.0302,
+      "step": 6980
+    },
+    {
+      "epoch": 7.988571428571428,
+      "grad_norm": 0.06017826497554779,
+      "learning_rate": 7.857142857142858e-08,
+      "loss": 0.0304,
+      "step": 6990
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0971122533082962,
+      "learning_rate": 7.142857142857144e-09,
+      "loss": 0.0311,
+      "step": 7000
+    },
+    {
+      "epoch": 8.0,
+      "eval_bleu": 0.07,
+      "eval_exact_match": 0.0,
+      "eval_loss": 0.028088832274079323,
+      "eval_runtime": 343.24,
+      "eval_samples_per_second": 17.48,
+      "eval_steps_per_second": 1.093,
+      "step": 7000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 7000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.929403150516224e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}