diff --git "a/checkpoint-6000/trainer_state.json" "b/checkpoint-6000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-6000/trainer_state.json"
@@ -0,0 +1,4261 @@
+{
+  "best_global_step": 6000,
+  "best_metric": 0.4967807301369415,
+  "best_model_checkpoint": "./SALAMA_NEW8/checkpoint-6000",
+  "epoch": 2.396245256640703,
+  "eval_steps": 2000,
+  "global_step": 6000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003994407829039345,
+      "grad_norm": 2.2177462577819824,
+      "learning_rate": 1.8e-07,
+      "loss": 0.022,
+      "step": 10
+    },
+    {
+      "epoch": 0.00798881565807869,
+      "grad_norm": 2.724672794342041,
+      "learning_rate": 3.8e-07,
+      "loss": 0.0249,
+      "step": 20
+    },
+    {
+      "epoch": 0.011983223487118035,
+      "grad_norm": 2.5243964195251465,
+      "learning_rate": 5.800000000000001e-07,
+      "loss": 0.0215,
+      "step": 30
+    },
+    {
+      "epoch": 0.01597763131615738,
+      "grad_norm": 2.4517173767089844,
+      "learning_rate": 7.8e-07,
+      "loss": 0.0229,
+      "step": 40
+    },
+    {
+      "epoch": 0.019972039145196723,
+      "grad_norm": 1.1104662418365479,
+      "learning_rate": 9.800000000000001e-07,
+      "loss": 0.0212,
+      "step": 50
+    },
+    {
+      "epoch": 0.02396644697423607,
+      "grad_norm": 3.2835190296173096,
+      "learning_rate": 1.1800000000000001e-06,
+      "loss": 0.0177,
+      "step": 60
+    },
+    {
+      "epoch": 0.027960854803275415,
+      "grad_norm": 2.933971405029297,
+      "learning_rate": 1.3800000000000001e-06,
+      "loss": 0.0181,
+      "step": 70
+    },
+    {
+      "epoch": 0.03195526263231476,
+      "grad_norm": 2.667775869369507,
+      "learning_rate": 1.5800000000000001e-06,
+      "loss": 0.0243,
+      "step": 80
+    },
+    {
+      "epoch": 0.035949670461354104,
+      "grad_norm": 2.8503150939941406,
+      "learning_rate": 1.7800000000000001e-06,
+      "loss": 0.0158,
+      "step": 90
+    },
+    {
+      "epoch": 0.039944078290393446,
+      "grad_norm": 2.6173582077026367,
+      "learning_rate": 1.98e-06,
+      "loss": 0.0217,
+      "step": 100
+    },
+    {
+      "epoch": 0.043938486119432796,
+      "grad_norm": 2.148167371749878,
+      "learning_rate": 2.1800000000000003e-06,
+      "loss": 0.0216,
+      "step": 110
+    },
+    {
+      "epoch": 0.04793289394847214,
+      "grad_norm": 1.6248950958251953,
+      "learning_rate": 2.38e-06,
+      "loss": 0.0211,
+      "step": 120
+    },
+    {
+      "epoch": 0.05192730177751148,
+      "grad_norm": 2.5831820964813232,
+      "learning_rate": 2.5800000000000003e-06,
+      "loss": 0.0282,
+      "step": 130
+    },
+    {
+      "epoch": 0.05592170960655083,
+      "grad_norm": 2.1337149143218994,
+      "learning_rate": 2.7800000000000005e-06,
+      "loss": 0.0215,
+      "step": 140
+    },
+    {
+      "epoch": 0.05991611743559017,
+      "grad_norm": 2.95695161819458,
+      "learning_rate": 2.9800000000000003e-06,
+      "loss": 0.0177,
+      "step": 150
+    },
+    {
+      "epoch": 0.06391052526462952,
+      "grad_norm": 2.277665853500366,
+      "learning_rate": 3.1800000000000005e-06,
+      "loss": 0.0217,
+      "step": 160
+    },
+    {
+      "epoch": 0.06790493309366886,
+      "grad_norm": 2.531909704208374,
+      "learning_rate": 3.3800000000000007e-06,
+      "loss": 0.0375,
+      "step": 170
+    },
+    {
+      "epoch": 0.07189934092270821,
+      "grad_norm": 2.5161631107330322,
+      "learning_rate": 3.58e-06,
+      "loss": 0.0262,
+      "step": 180
+    },
+    {
+      "epoch": 0.07589374875174755,
+      "grad_norm": 2.9737093448638916,
+      "learning_rate": 3.7800000000000002e-06,
+      "loss": 0.0237,
+      "step": 190
+    },
+    {
+      "epoch": 0.07988815658078689,
+      "grad_norm": 2.2957046031951904,
+      "learning_rate": 3.980000000000001e-06,
+      "loss": 0.0211,
+      "step": 200
+    },
+    {
+      "epoch": 0.08388256440982625,
+      "grad_norm": 1.9400784969329834,
+      "learning_rate": 4.18e-06,
+      "loss": 0.0241,
+      "step": 210
+    },
+    {
+      "epoch": 0.08787697223886559,
+      "grad_norm": 3.7293314933776855,
+      "learning_rate": 4.38e-06,
+      "loss": 0.0263,
+      "step": 220
+    },
+    {
+      "epoch": 0.09187138006790493,
+      "grad_norm": 4.266306400299072,
+      "learning_rate": 4.58e-06,
+      "loss": 0.0387,
+      "step": 230
+    },
+    {
+      "epoch": 0.09586578789694428,
+      "grad_norm": 2.510282278060913,
+      "learning_rate": 4.78e-06,
+      "loss": 0.029,
+      "step": 240
+    },
+    {
+      "epoch": 0.09986019572598362,
+      "grad_norm": 1.4938842058181763,
+      "learning_rate": 4.980000000000001e-06,
+      "loss": 0.0154,
+      "step": 250
+    },
+    {
+      "epoch": 0.10385460355502296,
+      "grad_norm": 2.0076608657836914,
+      "learning_rate": 5.18e-06,
+      "loss": 0.0263,
+      "step": 260
+    },
+    {
+      "epoch": 0.10784901138406232,
+      "grad_norm": 2.252916097640991,
+      "learning_rate": 5.380000000000001e-06,
+      "loss": 0.0273,
+      "step": 270
+    },
+    {
+      "epoch": 0.11184341921310166,
+      "grad_norm": 0.9760869145393372,
+      "learning_rate": 5.580000000000001e-06,
+      "loss": 0.0227,
+      "step": 280
+    },
+    {
+      "epoch": 0.115837827042141,
+      "grad_norm": 1.7485207319259644,
+      "learning_rate": 5.78e-06,
+      "loss": 0.0234,
+      "step": 290
+    },
+    {
+      "epoch": 0.11983223487118035,
+      "grad_norm": 2.540903329849243,
+      "learning_rate": 5.98e-06,
+      "loss": 0.0216,
+      "step": 300
+    },
+    {
+      "epoch": 0.12382664270021969,
+      "grad_norm": 2.970639228820801,
+      "learning_rate": 6.18e-06,
+      "loss": 0.0244,
+      "step": 310
+    },
+    {
+      "epoch": 0.12782105052925904,
+      "grad_norm": 4.563722610473633,
+      "learning_rate": 6.380000000000001e-06,
+      "loss": 0.0332,
+      "step": 320
+    },
+    {
+      "epoch": 0.1318154583582984,
+      "grad_norm": 3.670563220977783,
+      "learning_rate": 6.5800000000000005e-06,
+      "loss": 0.0259,
+      "step": 330
+    },
+    {
+      "epoch": 0.13580986618733773,
+      "grad_norm": 3.6714515686035156,
+      "learning_rate": 6.780000000000001e-06,
+      "loss": 0.032,
+      "step": 340
+    },
+    {
+      "epoch": 0.13980427401637707,
+      "grad_norm": 2.029430389404297,
+      "learning_rate": 6.98e-06,
+      "loss": 0.0264,
+      "step": 350
+    },
+    {
+      "epoch": 0.14379868184541642,
+      "grad_norm": 2.609748363494873,
+      "learning_rate": 7.180000000000001e-06,
+      "loss": 0.0369,
+      "step": 360
+    },
+    {
+      "epoch": 0.14779308967445576,
+      "grad_norm": 1.438359260559082,
+      "learning_rate": 7.3800000000000005e-06,
+      "loss": 0.0242,
+      "step": 370
+    },
+    {
+      "epoch": 0.1517874975034951,
+      "grad_norm": 2.1897497177124023,
+      "learning_rate": 7.58e-06,
+      "loss": 0.02,
+      "step": 380
+    },
+    {
+      "epoch": 0.15578190533253444,
+      "grad_norm": 1.9179186820983887,
+      "learning_rate": 7.78e-06,
+      "loss": 0.0366,
+      "step": 390
+    },
+    {
+      "epoch": 0.15977631316157379,
+      "grad_norm": 2.6540606021881104,
+      "learning_rate": 7.980000000000002e-06,
+      "loss": 0.0253,
+      "step": 400
+    },
+    {
+      "epoch": 0.16377072099061313,
+      "grad_norm": 2.5292742252349854,
+      "learning_rate": 8.18e-06,
+      "loss": 0.0356,
+      "step": 410
+    },
+    {
+      "epoch": 0.1677651288196525,
+      "grad_norm": 2.4729959964752197,
+      "learning_rate": 8.380000000000001e-06,
+      "loss": 0.0327,
+      "step": 420
+    },
+    {
+      "epoch": 0.17175953664869184,
+      "grad_norm": 3.639272689819336,
+      "learning_rate": 8.580000000000001e-06,
+      "loss": 0.032,
+      "step": 430
+    },
+    {
+      "epoch": 0.17575394447773118,
+      "grad_norm": 3.2385575771331787,
+      "learning_rate": 8.78e-06,
+      "loss": 0.0359,
+      "step": 440
+    },
+    {
+      "epoch": 0.17974835230677053,
+      "grad_norm": 3.476755142211914,
+      "learning_rate": 8.98e-06,
+      "loss": 0.0286,
+      "step": 450
+    },
+    {
+      "epoch": 0.18374276013580987,
+      "grad_norm": 2.612086772918701,
+      "learning_rate": 9.180000000000002e-06,
+      "loss": 0.0228,
+      "step": 460
+    },
+    {
+      "epoch": 0.1877371679648492,
+      "grad_norm": 3.5403027534484863,
+      "learning_rate": 9.38e-06,
+      "loss": 0.0243,
+      "step": 470
+    },
+    {
+      "epoch": 0.19173157579388855,
+      "grad_norm": 2.7765207290649414,
+      "learning_rate": 9.58e-06,
+      "loss": 0.0348,
+      "step": 480
+    },
+    {
+      "epoch": 0.1957259836229279,
+      "grad_norm": 3.589587450027466,
+      "learning_rate": 9.780000000000001e-06,
+      "loss": 0.0323,
+      "step": 490
+    },
+    {
+      "epoch": 0.19972039145196724,
+      "grad_norm": 1.8719650506973267,
+      "learning_rate": 9.980000000000001e-06,
+      "loss": 0.0275,
+      "step": 500
+    },
+    {
+      "epoch": 0.20371479928100658,
+      "grad_norm": 2.750931739807129,
+      "learning_rate": 9.983636363636364e-06,
+      "loss": 0.0348,
+      "step": 510
+    },
+    {
+      "epoch": 0.20770920711004592,
+      "grad_norm": 1.9383965730667114,
+      "learning_rate": 9.965454545454546e-06,
+      "loss": 0.0282,
+      "step": 520
+    },
+    {
+      "epoch": 0.2117036149390853,
+      "grad_norm": 3.6504414081573486,
+      "learning_rate": 9.947272727272728e-06,
+      "loss": 0.0316,
+      "step": 530
+    },
+    {
+      "epoch": 0.21569802276812464,
+      "grad_norm": 3.26383376121521,
+      "learning_rate": 9.92909090909091e-06,
+      "loss": 0.0273,
+      "step": 540
+    },
+    {
+      "epoch": 0.21969243059716398,
+      "grad_norm": 4.769904613494873,
+      "learning_rate": 9.910909090909092e-06,
+      "loss": 0.0295,
+      "step": 550
+    },
+    {
+      "epoch": 0.22368683842620332,
+      "grad_norm": 1.143069863319397,
+      "learning_rate": 9.892727272727273e-06,
+      "loss": 0.027,
+      "step": 560
+    },
+    {
+      "epoch": 0.22768124625524266,
+      "grad_norm": 2.988460063934326,
+      "learning_rate": 9.874545454545455e-06,
+      "loss": 0.0349,
+      "step": 570
+    },
+    {
+      "epoch": 0.231675654084282,
+      "grad_norm": 2.6448323726654053,
+      "learning_rate": 9.856363636363637e-06,
+      "loss": 0.036,
+      "step": 580
+    },
+    {
+      "epoch": 0.23567006191332135,
+      "grad_norm": 3.077092409133911,
+      "learning_rate": 9.838181818181819e-06,
+      "loss": 0.0491,
+      "step": 590
+    },
+    {
+      "epoch": 0.2396644697423607,
+      "grad_norm": 6.214341163635254,
+      "learning_rate": 9.820000000000001e-06,
+      "loss": 0.0313,
+      "step": 600
+    },
+    {
+      "epoch": 0.24365887757140003,
+      "grad_norm": 3.1780426502227783,
+      "learning_rate": 9.801818181818183e-06,
+      "loss": 0.0251,
+      "step": 610
+    },
+    {
+      "epoch": 0.24765328540043938,
+      "grad_norm": 1.7672063112258911,
+      "learning_rate": 9.783636363636365e-06,
+      "loss": 0.0259,
+      "step": 620
+    },
+    {
+      "epoch": 0.2516476932294787,
+      "grad_norm": 3.9317915439605713,
+      "learning_rate": 9.765454545454546e-06,
+      "loss": 0.0263,
+      "step": 630
+    },
+    {
+      "epoch": 0.2556421010585181,
+      "grad_norm": 2.23710298538208,
+      "learning_rate": 9.747272727272728e-06,
+      "loss": 0.0335,
+      "step": 640
+    },
+    {
+      "epoch": 0.2596365088875574,
+      "grad_norm": 3.24867844581604,
+      "learning_rate": 9.72909090909091e-06,
+      "loss": 0.0383,
+      "step": 650
+    },
+    {
+      "epoch": 0.2636309167165968,
+      "grad_norm": 3.0995874404907227,
+      "learning_rate": 9.710909090909092e-06,
+      "loss": 0.0296,
+      "step": 660
+    },
+    {
+      "epoch": 0.2676253245456361,
+      "grad_norm": 1.4767210483551025,
+      "learning_rate": 9.692727272727274e-06,
+      "loss": 0.04,
+      "step": 670
+    },
+    {
+      "epoch": 0.27161973237467546,
+      "grad_norm": 1.7182132005691528,
+      "learning_rate": 9.674545454545456e-06,
+      "loss": 0.0294,
+      "step": 680
+    },
+    {
+      "epoch": 0.2756141402037148,
+      "grad_norm": 3.0770034790039062,
+      "learning_rate": 9.656363636363637e-06,
+      "loss": 0.0393,
+      "step": 690
+    },
+    {
+      "epoch": 0.27960854803275415,
+      "grad_norm": 2.452723503112793,
+      "learning_rate": 9.63818181818182e-06,
+      "loss": 0.0321,
+      "step": 700
+    },
+    {
+      "epoch": 0.2836029558617935,
+      "grad_norm": 2.9560225009918213,
+      "learning_rate": 9.620000000000001e-06,
+      "loss": 0.0292,
+      "step": 710
+    },
+    {
+      "epoch": 0.28759736369083283,
+      "grad_norm": 2.8095853328704834,
+      "learning_rate": 9.601818181818183e-06,
+      "loss": 0.028,
+      "step": 720
+    },
+    {
+      "epoch": 0.2915917715198722,
+      "grad_norm": 3.4851746559143066,
+      "learning_rate": 9.583636363636365e-06,
+      "loss": 0.0296,
+      "step": 730
+    },
+    {
+      "epoch": 0.2955861793489115,
+      "grad_norm": 4.2011027336120605,
+      "learning_rate": 9.565454545454547e-06,
+      "loss": 0.0353,
+      "step": 740
+    },
+    {
+      "epoch": 0.2995805871779509,
+      "grad_norm": 5.23146390914917,
+      "learning_rate": 9.547272727272728e-06,
+      "loss": 0.0393,
+      "step": 750
+    },
+    {
+      "epoch": 0.3035749950069902,
+      "grad_norm": 2.8136441707611084,
+      "learning_rate": 9.52909090909091e-06,
+      "loss": 0.0339,
+      "step": 760
+    },
+    {
+      "epoch": 0.30756940283602957,
+      "grad_norm": 3.0712337493896484,
+      "learning_rate": 9.510909090909092e-06,
+      "loss": 0.0414,
+      "step": 770
+    },
+    {
+      "epoch": 0.3115638106650689,
+      "grad_norm": 2.1798360347747803,
+      "learning_rate": 9.492727272727274e-06,
+      "loss": 0.03,
+      "step": 780
+    },
+    {
+      "epoch": 0.31555821849410826,
+      "grad_norm": 3.7893567085266113,
+      "learning_rate": 9.474545454545456e-06,
+      "loss": 0.0325,
+      "step": 790
+    },
+    {
+      "epoch": 0.31955262632314757,
+      "grad_norm": 2.0599608421325684,
+      "learning_rate": 9.456363636363638e-06,
+      "loss": 0.0389,
+      "step": 800
+    },
+    {
+      "epoch": 0.32354703415218694,
+      "grad_norm": 2.7613537311553955,
+      "learning_rate": 9.438181818181818e-06,
+      "loss": 0.0284,
+      "step": 810
+    },
+    {
+      "epoch": 0.32754144198122626,
+      "grad_norm": 2.630587577819824,
+      "learning_rate": 9.42e-06,
+      "loss": 0.0308,
+      "step": 820
+    },
+    {
+      "epoch": 0.3315358498102656,
+      "grad_norm": 2.8372304439544678,
+      "learning_rate": 9.401818181818183e-06,
+      "loss": 0.029,
+      "step": 830
+    },
+    {
+      "epoch": 0.335530257639305,
+      "grad_norm": 2.21663236618042,
+      "learning_rate": 9.383636363636365e-06,
+      "loss": 0.0367,
+      "step": 840
+    },
+    {
+      "epoch": 0.3395246654683443,
+      "grad_norm": 3.325446367263794,
+      "learning_rate": 9.365454545454547e-06,
+      "loss": 0.0256,
+      "step": 850
+    },
+    {
+      "epoch": 0.3435190732973837,
+      "grad_norm": 2.7013704776763916,
+      "learning_rate": 9.347272727272729e-06,
+      "loss": 0.0279,
+      "step": 860
+    },
+    {
+      "epoch": 0.347513481126423,
+      "grad_norm": 1.6712983846664429,
+      "learning_rate": 9.32909090909091e-06,
+      "loss": 0.0256,
+      "step": 870
+    },
+    {
+      "epoch": 0.35150788895546237,
+      "grad_norm": 3.0498671531677246,
+      "learning_rate": 9.310909090909092e-06,
+      "loss": 0.0351,
+      "step": 880
+    },
+    {
+      "epoch": 0.3555022967845017,
+      "grad_norm": 2.060898542404175,
+      "learning_rate": 9.292727272727272e-06,
+      "loss": 0.0349,
+      "step": 890
+    },
+    {
+      "epoch": 0.35949670461354105,
+      "grad_norm": 2.191037178039551,
+      "learning_rate": 9.274545454545454e-06,
+      "loss": 0.0236,
+      "step": 900
+    },
+    {
+      "epoch": 0.36349111244258037,
+      "grad_norm": 2.7442283630371094,
+      "learning_rate": 9.256363636363636e-06,
+      "loss": 0.0339,
+      "step": 910
+    },
+    {
+      "epoch": 0.36748552027161974,
+      "grad_norm": 2.151813268661499,
+      "learning_rate": 9.23818181818182e-06,
+      "loss": 0.0272,
+      "step": 920
+    },
+    {
+      "epoch": 0.37147992810065905,
+      "grad_norm": 3.0903069972991943,
+      "learning_rate": 9.220000000000002e-06,
+      "loss": 0.0343,
+      "step": 930
+    },
+    {
+      "epoch": 0.3754743359296984,
+      "grad_norm": 3.2857322692871094,
+      "learning_rate": 9.201818181818183e-06,
+      "loss": 0.028,
+      "step": 940
+    },
+    {
+      "epoch": 0.3794687437587378,
+      "grad_norm": 2.7144224643707275,
+      "learning_rate": 9.183636363636365e-06,
+      "loss": 0.0224,
+      "step": 950
+    },
+    {
+      "epoch": 0.3834631515877771,
+      "grad_norm": 2.985253095626831,
+      "learning_rate": 9.165454545454547e-06,
+      "loss": 0.0297,
+      "step": 960
+    },
+    {
+      "epoch": 0.3874575594168165,
+      "grad_norm": 2.6801540851593018,
+      "learning_rate": 9.147272727272727e-06,
+      "loss": 0.0306,
+      "step": 970
+    },
+    {
+      "epoch": 0.3914519672458558,
+      "grad_norm": 2.944761276245117,
+      "learning_rate": 9.129090909090909e-06,
+      "loss": 0.0324,
+      "step": 980
+    },
+    {
+      "epoch": 0.39544637507489516,
+      "grad_norm": 2.6776340007781982,
+      "learning_rate": 9.11090909090909e-06,
+      "loss": 0.04,
+      "step": 990
+    },
+    {
+      "epoch": 0.3994407829039345,
+      "grad_norm": 4.007734775543213,
+      "learning_rate": 9.092727272727273e-06,
+      "loss": 0.0286,
+      "step": 1000
+    },
+    {
+      "epoch": 0.40343519073297385,
+      "grad_norm": 3.652127265930176,
+      "learning_rate": 9.074545454545455e-06,
+      "loss": 0.0373,
+      "step": 1010
+    },
+    {
+      "epoch": 0.40742959856201316,
+      "grad_norm": 1.8009936809539795,
+      "learning_rate": 9.056363636363638e-06,
+      "loss": 0.024,
+      "step": 1020
+    },
+    {
+      "epoch": 0.41142400639105253,
+      "grad_norm": 2.0933914184570312,
+      "learning_rate": 9.03818181818182e-06,
+      "loss": 0.0341,
+      "step": 1030
+    },
+    {
+      "epoch": 0.41541841422009185,
+      "grad_norm": 2.683140277862549,
+      "learning_rate": 9.020000000000002e-06,
+      "loss": 0.0473,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4194128220491312,
+      "grad_norm": 2.7582523822784424,
+      "learning_rate": 9.001818181818182e-06,
+      "loss": 0.0322,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4234072298781706,
+      "grad_norm": 3.6831445693969727,
+      "learning_rate": 8.983636363636364e-06,
+      "loss": 0.0382,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4274016377072099,
+      "grad_norm": 2.053619623184204,
+      "learning_rate": 8.965454545454546e-06,
+      "loss": 0.0304,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4313960455362493,
+      "grad_norm": 2.5573840141296387,
+      "learning_rate": 8.947272727272727e-06,
+      "loss": 0.0567,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4353904533652886,
+      "grad_norm": 3.5455336570739746,
+      "learning_rate": 8.92909090909091e-06,
+      "loss": 0.0338,
+      "step": 1090
+    },
+    {
+      "epoch": 0.43938486119432796,
+      "grad_norm": 2.8842337131500244,
+      "learning_rate": 8.910909090909091e-06,
+      "loss": 0.0314,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4433792690233673,
+      "grad_norm": 2.779266834259033,
+      "learning_rate": 8.892727272727275e-06,
+      "loss": 0.0486,
+      "step": 1110
+    },
+    {
+      "epoch": 0.44737367685240664,
+      "grad_norm": 3.959299325942993,
+      "learning_rate": 8.874545454545456e-06,
+      "loss": 0.0396,
+      "step": 1120
+    },
+    {
+      "epoch": 0.45136808468144596,
+      "grad_norm": 2.053091526031494,
+      "learning_rate": 8.856363636363637e-06,
+      "loss": 0.0329,
+      "step": 1130
+    },
+    {
+      "epoch": 0.45536249251048533,
+      "grad_norm": 1.1432011127471924,
+      "learning_rate": 8.838181818181818e-06,
+      "loss": 0.0268,
+      "step": 1140
+    },
+    {
+      "epoch": 0.45935690033952464,
+      "grad_norm": 2.738510847091675,
+      "learning_rate": 8.82e-06,
+      "loss": 0.0388,
+      "step": 1150
+    },
+    {
+      "epoch": 0.463351308168564,
+      "grad_norm": 1.9435960054397583,
+      "learning_rate": 8.801818181818182e-06,
+      "loss": 0.0321,
+      "step": 1160
+    },
+    {
+      "epoch": 0.46734571599760333,
+      "grad_norm": 2.462301015853882,
+      "learning_rate": 8.783636363636364e-06,
+      "loss": 0.0344,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4713401238266427,
+      "grad_norm": 2.8463451862335205,
+      "learning_rate": 8.765454545454546e-06,
+      "loss": 0.0372,
+      "step": 1180
+    },
+    {
+      "epoch": 0.47533453165568207,
+      "grad_norm": 4.407367706298828,
+      "learning_rate": 8.747272727272728e-06,
+      "loss": 0.0399,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4793289394847214,
+      "grad_norm": 2.5179121494293213,
+      "learning_rate": 8.72909090909091e-06,
+      "loss": 0.0331,
+      "step": 1200
+    },
+    {
+      "epoch": 0.48332334731376075,
+      "grad_norm": 3.2438509464263916,
+      "learning_rate": 8.710909090909091e-06,
+      "loss": 0.0348,
+      "step": 1210
+    },
+    {
+      "epoch": 0.48731775514280007,
+      "grad_norm": 2.54004168510437,
+      "learning_rate": 8.692727272727273e-06,
+      "loss": 0.0317,
+      "step": 1220
+    },
+    {
+      "epoch": 0.49131216297183944,
+      "grad_norm": 2.8079185485839844,
+      "learning_rate": 8.674545454545455e-06,
+      "loss": 0.0386,
+      "step": 1230
+    },
+    {
+      "epoch": 0.49530657080087875,
+      "grad_norm": 4.827033519744873,
+      "learning_rate": 8.656363636363637e-06,
+      "loss": 0.0338,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4993009786299181,
+      "grad_norm": 2.919968843460083,
+      "learning_rate": 8.638181818181819e-06,
+      "loss": 0.0265,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5032953864589574,
+      "grad_norm": 3.6935982704162598,
+      "learning_rate": 8.62e-06,
+      "loss": 0.0327,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5072897942879968,
+      "grad_norm": 4.243749141693115,
+      "learning_rate": 8.601818181818182e-06,
+      "loss": 0.0282,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5112842021170362,
+      "grad_norm": 3.979485273361206,
+      "learning_rate": 8.583636363636364e-06,
+      "loss": 0.0242,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5152786099460755,
+      "grad_norm": 3.32183575630188,
+      "learning_rate": 8.565454545454546e-06,
+      "loss": 0.0356,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5192730177751148,
+      "grad_norm": 2.474187135696411,
+      "learning_rate": 8.547272727272728e-06,
+      "loss": 0.0201,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5232674256041542,
+      "grad_norm": 3.150151252746582,
+      "learning_rate": 8.52909090909091e-06,
+      "loss": 0.0304,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5272618334331935,
+      "grad_norm": 2.039504289627075,
+      "learning_rate": 8.510909090909092e-06,
+      "loss": 0.045,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5312562412622329,
+      "grad_norm": 2.7388851642608643,
+      "learning_rate": 8.492727272727273e-06,
+      "loss": 0.0268,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5352506490912722,
+      "grad_norm": 2.9932122230529785,
+      "learning_rate": 8.474545454545455e-06,
+      "loss": 0.0331,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5392450569203115,
+      "grad_norm": 3.3818037509918213,
+      "learning_rate": 8.456363636363637e-06,
+      "loss": 0.0365,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5432394647493509,
+      "grad_norm": 3.251274585723877,
+      "learning_rate": 8.438181818181819e-06,
+      "loss": 0.0395,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5472338725783903,
+      "grad_norm": 3.46567964553833,
+      "learning_rate": 8.42e-06,
+      "loss": 0.0287,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5512282804074295,
+      "grad_norm": 2.459820032119751,
+      "learning_rate": 8.401818181818183e-06,
+      "loss": 0.0329,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5552226882364689,
+      "grad_norm": 1.735729694366455,
+      "learning_rate": 8.383636363636364e-06,
+      "loss": 0.0313,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5592170960655083,
+      "grad_norm": 2.2887370586395264,
+      "learning_rate": 8.365454545454546e-06,
+      "loss": 0.0215,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5632115038945477,
+      "grad_norm": 3.110576868057251,
+      "learning_rate": 8.347272727272728e-06,
+      "loss": 0.0213,
+      "step": 1410
+    },
+    {
+      "epoch": 0.567205911723587,
+      "grad_norm": 3.1144895553588867,
+      "learning_rate": 8.32909090909091e-06,
+      "loss": 0.0312,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5712003195526263,
+      "grad_norm": 2.9777989387512207,
+      "learning_rate": 8.310909090909092e-06,
+      "loss": 0.0322,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5751947273816657,
+      "grad_norm": 2.2770936489105225,
+      "learning_rate": 8.292727272727274e-06,
+      "loss": 0.0304,
+      "step": 1440
+    },
+    {
+      "epoch": 0.579189135210705,
+      "grad_norm": 3.5356180667877197,
+      "learning_rate": 8.274545454545455e-06,
+      "loss": 0.0347,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5831835430397444,
+      "grad_norm": 3.6384565830230713,
+      "learning_rate": 8.256363636363637e-06,
+      "loss": 0.0289,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5871779508687837,
+      "grad_norm": 3.053424119949341,
+      "learning_rate": 8.238181818181819e-06,
+      "loss": 0.0356,
+      "step": 1470
+    },
+    {
+      "epoch": 0.591172358697823,
+      "grad_norm": 4.9105963706970215,
+      "learning_rate": 8.220000000000001e-06,
+      "loss": 0.0315,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5951667665268624,
+      "grad_norm": 3.0485212802886963,
+      "learning_rate": 8.201818181818183e-06,
+      "loss": 0.0298,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5991611743559018,
+      "grad_norm": 3.3632636070251465,
+      "learning_rate": 8.183636363636365e-06,
+      "loss": 0.0268,
+      "step": 1500
+    },
+    {
+      "epoch": 0.603155582184941,
+      "grad_norm": 2.593235969543457,
+      "learning_rate": 8.165454545454546e-06,
+      "loss": 0.0286,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6071499900139804,
+      "grad_norm": 2.542865753173828,
+      "learning_rate": 8.147272727272728e-06,
+      "loss": 0.0317,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6111443978430198,
+      "grad_norm": 3.246321201324463,
+      "learning_rate": 8.12909090909091e-06,
+      "loss": 0.0358,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6151388056720591,
+      "grad_norm": 4.592155456542969,
+      "learning_rate": 8.110909090909092e-06,
+      "loss": 0.0295,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6191332135010985,
+      "grad_norm": 2.1040351390838623,
+      "learning_rate": 8.092727272727274e-06,
+      "loss": 0.044,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6231276213301378,
+      "grad_norm": 2.832470417022705,
+      "learning_rate": 8.074545454545456e-06,
+      "loss": 0.0335,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6271220291591771,
+      "grad_norm": 3.7737035751342773,
+      "learning_rate": 8.056363636363636e-06,
+      "loss": 0.0289,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6311164369882165,
+      "grad_norm": 2.9322657585144043,
+      "learning_rate": 8.038181818181818e-06,
+      "loss": 0.0209,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6351108448172559,
+      "grad_norm": 3.0506591796875,
+      "learning_rate": 8.020000000000001e-06,
+      "loss": 0.0277,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6391052526462951,
+      "grad_norm": 1.679126501083374,
+      "learning_rate": 8.001818181818183e-06,
+      "loss": 0.0313,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6430996604753345,
+      "grad_norm": 1.4025191068649292,
+      "learning_rate": 7.983636363636365e-06,
+      "loss": 0.0342,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6470940683043739,
+      "grad_norm": 3.2698376178741455,
+      "learning_rate": 7.965454545454547e-06,
+      "loss": 0.0334,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6510884761334133,
+      "grad_norm": 2.007560968399048,
+      "learning_rate": 7.947272727272728e-06,
+      "loss": 0.0361,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6550828839624525,
+      "grad_norm": 3.021299123764038,
+      "learning_rate": 7.92909090909091e-06,
+      "loss": 0.0316,
+      "step": 1640
+    },
+    {
+      "epoch": 0.6590772917914919,
+      "grad_norm": 3.4559459686279297,
+      "learning_rate": 7.91090909090909e-06,
+      "loss": 0.0336,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6630716996205313,
+      "grad_norm": 3.786959648132324,
+      "learning_rate": 7.892727272727272e-06,
+      "loss": 0.0337,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6670661074495706,
+      "grad_norm": 2.5222368240356445,
+      "learning_rate": 7.874545454545454e-06,
+      "loss": 0.026,
+      "step": 1670
+    },
+    {
+      "epoch": 0.67106051527861,
+      "grad_norm": 2.880535364151001,
+      "learning_rate": 7.856363636363638e-06,
+      "loss": 0.023,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6750549231076493,
+      "grad_norm": 3.375427007675171,
+      "learning_rate": 7.83818181818182e-06,
+      "loss": 0.025,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6790493309366886,
+      "grad_norm": 1.8976999521255493,
+      "learning_rate": 7.820000000000001e-06,
+      "loss": 0.0267,
+      "step": 1700
+    },
+    {
+      "epoch": 0.683043738765728,
+      "grad_norm": 3.0969624519348145,
+      "learning_rate": 7.801818181818183e-06,
+      "loss": 0.0334,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6870381465947674,
+      "grad_norm": 2.949564218521118,
+      "learning_rate": 7.783636363636365e-06,
+      "loss": 0.0317,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6910325544238066,
+      "grad_norm": 3.1288681030273438,
+      "learning_rate": 7.765454545454545e-06,
+      "loss": 0.036,
+      "step": 1730
+    },
+    {
+      "epoch": 0.695026962252846,
+      "grad_norm": 2.337353229522705,
+      "learning_rate": 7.747272727272727e-06,
+      "loss": 0.0233,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6990213700818854,
+      "grad_norm": 2.593578577041626,
+      "learning_rate": 7.729090909090909e-06,
+      "loss": 0.0338,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7030157779109247,
+      "grad_norm": 2.1465232372283936,
+      "learning_rate": 7.71090909090909e-06,
+      "loss": 0.0289,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7070101857399641,
+      "grad_norm": 2.932419538497925,
+      "learning_rate": 7.692727272727273e-06,
+      "loss": 0.0263,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7110045935690034,
+      "grad_norm": 4.221918106079102,
+      "learning_rate": 7.674545454545456e-06,
+      "loss": 0.0262,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7149990013980427,
+      "grad_norm": 3.3574588298797607,
+      "learning_rate": 7.656363636363638e-06,
+      "loss": 0.0298,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7189934092270821,
+      "grad_norm": 3.61956787109375,
+      "learning_rate": 7.63818181818182e-06,
+      "loss": 0.0361,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7229878170561215,
+      "grad_norm": 3.8640527725219727,
+      "learning_rate": 7.620000000000001e-06,
+      "loss": 0.0379,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7269822248851607,
+      "grad_norm": 3.021359920501709,
+      "learning_rate": 7.6018181818181826e-06,
+      "loss": 0.0256,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7309766327142001,
+      "grad_norm": 3.4771692752838135,
+      "learning_rate": 7.583636363636364e-06,
+      "loss": 0.0317,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7349710405432395,
+      "grad_norm": 1.799800992012024,
+      "learning_rate": 7.565454545454546e-06,
+      "loss": 0.0318,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7389654483722788,
+      "grad_norm": 2.8619134426116943,
+      "learning_rate": 7.547272727272727e-06,
+      "loss": 0.0351,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7429598562013181,
+      "grad_norm": 1.9078686237335205,
+      "learning_rate": 7.529090909090909e-06,
+      "loss": 0.0223,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7469542640303575,
+      "grad_norm": 3.3508458137512207,
+      "learning_rate": 7.510909090909092e-06,
+      "loss": 0.0219,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7509486718593968,
+      "grad_norm": 2.5900681018829346,
+      "learning_rate": 7.492727272727274e-06,
+      "loss": 0.0231,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7549430796884362,
+      "grad_norm": 2.878042697906494,
+      "learning_rate": 7.4745454545454554e-06,
+      "loss": 0.0269,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7589374875174756,
+      "grad_norm": 2.814326524734497,
+      "learning_rate": 7.456363636363637e-06,
+      "loss": 0.027,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7629318953465148,
+      "grad_norm": 3.0135231018066406,
+      "learning_rate": 7.438181818181819e-06,
+      "loss": 0.0363,
+      "step": 1910
+    },
+    {
+      "epoch": 0.7669263031755542,
+      "grad_norm": 4.630500793457031,
+      "learning_rate": 7.420000000000001e-06,
+      "loss": 0.0318,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7709207110045936,
+      "grad_norm": 2.1792924404144287,
+      "learning_rate": 7.401818181818182e-06,
+      "loss": 0.0299,
+      "step": 1930
+    },
+    {
+      "epoch": 0.774915118833633,
+      "grad_norm": 2.716294527053833,
+      "learning_rate": 7.383636363636364e-06,
+      "loss": 0.028,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7789095266626722,
+      "grad_norm": 2.591440439224243,
+      "learning_rate": 7.365454545454546e-06,
+      "loss": 0.0311,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7829039344917116,
+      "grad_norm": 1.3216569423675537,
+      "learning_rate": 7.3472727272727275e-06,
+      "loss": 0.0213,
+      "step": 1960
+    },
+    {
+      "epoch": 0.786898342320751,
+      "grad_norm": 1.4774867296218872,
+      "learning_rate": 7.32909090909091e-06,
+      "loss": 0.0218,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7908927501497903,
+      "grad_norm": 2.8911545276641846,
+      "learning_rate": 7.310909090909092e-06,
+      "loss": 0.0251,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7948871579788296,
+      "grad_norm": 4.558359146118164,
+      "learning_rate": 7.292727272727274e-06,
+      "loss": 0.023,
+      "step": 1990
+    },
+    {
+      "epoch": 0.798881565807869,
+      "grad_norm": 2.78124737739563,
+      "learning_rate": 7.274545454545456e-06,
+      "loss": 0.0361,
+      "step": 2000
+    },
+    {
+      "epoch": 0.798881565807869,
+      "eval_loss": 0.017742320895195007,
+      "eval_runtime": 7487.3562,
+      "eval_samples_per_second": 2.675,
+      "eval_steps_per_second": 0.334,
+      "eval_wer": 1.5431796145591616,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8028759736369083,
+      "grad_norm": 3.3901844024658203,
+      "learning_rate": 7.256363636363637e-06,
+      "loss": 0.0319,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8068703814659477,
+      "grad_norm": 3.4345223903656006,
+      "learning_rate": 7.2381818181818185e-06,
+      "loss": 0.0313,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8108647892949871,
+      "grad_norm": 2.4429991245269775,
+      "learning_rate": 7.22e-06,
+      "loss": 0.0255,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8148591971240263,
+      "grad_norm": 1.69983971118927,
+      "learning_rate": 7.201818181818182e-06,
+      "loss": 0.0326,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8188536049530657,
+      "grad_norm": 1.1749235391616821,
+      "learning_rate": 7.183636363636364e-06,
+      "loss": 0.0265,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8228480127821051,
+      "grad_norm": 3.216322422027588,
+      "learning_rate": 7.165454545454547e-06,
+      "loss": 0.0297,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8268424206111444,
+      "grad_norm": 2.813265562057495,
+      "learning_rate": 7.1472727272727285e-06,
+      "loss": 0.0284,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8308368284401837,
+      "grad_norm": 2.927987575531006,
+      "learning_rate": 7.12909090909091e-06,
+      "loss": 0.0397,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8348312362692231,
+      "grad_norm": 3.2884976863861084,
+      "learning_rate": 7.110909090909091e-06,
+      "loss": 0.0319,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8388256440982624,
+      "grad_norm": 2.057528495788574,
+      "learning_rate": 7.092727272727273e-06,
+      "loss": 0.0305,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8428200519273018,
+      "grad_norm": 2.372114658355713,
+      "learning_rate": 7.074545454545455e-06,
+      "loss": 0.03,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8468144597563412,
+      "grad_norm": 2.805551052093506,
+      "learning_rate": 7.056363636363637e-06,
+      "loss": 0.0305,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8508088675853804,
+      "grad_norm": 2.1154088973999023,
+      "learning_rate": 7.038181818181819e-06,
+      "loss": 0.026,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8548032754144198,
+      "grad_norm": 3.1023805141448975,
+      "learning_rate": 7.0200000000000006e-06,
+      "loss": 0.033,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8587976832434592,
+      "grad_norm": 2.8123090267181396,
+      "learning_rate": 7.0018181818181815e-06,
+      "loss": 0.0273,
+      "step": 2150
+    },
+    {
+      "epoch": 0.8627920910724985,
+      "grad_norm": 2.0984227657318115,
+      "learning_rate": 6.983636363636365e-06,
+      "loss": 0.0279,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8667864989015378,
+      "grad_norm": 4.393746376037598,
+      "learning_rate": 6.965454545454546e-06,
+      "loss": 0.0318,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8707809067305772,
+      "grad_norm": 3.9992425441741943,
+      "learning_rate": 6.947272727272728e-06,
+      "loss": 0.0353,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8747753145596165,
+      "grad_norm": 1.945391297340393,
+      "learning_rate": 6.92909090909091e-06,
+      "loss": 0.0228,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8787697223886559,
+      "grad_norm": 3.7365753650665283,
+      "learning_rate": 6.910909090909092e-06,
+      "loss": 0.0299,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8827641302176952,
+      "grad_norm": 2.2218515872955322,
+      "learning_rate": 6.892727272727273e-06,
+      "loss": 0.025,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8867585380467345,
+      "grad_norm": 1.6459863185882568,
+      "learning_rate": 6.874545454545455e-06,
+      "loss": 0.0272,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8907529458757739,
+      "grad_norm": 1.9987223148345947,
+      "learning_rate": 6.856363636363636e-06,
+      "loss": 0.0316,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8947473537048133,
+      "grad_norm": 1.8636083602905273,
+      "learning_rate": 6.838181818181818e-06,
+      "loss": 0.0296,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8987417615338527,
+      "grad_norm": 4.068820476531982,
+      "learning_rate": 6.820000000000001e-06,
+      "loss": 0.0377,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9027361693628919,
+      "grad_norm": 2.758310079574585,
+      "learning_rate": 6.801818181818183e-06,
+      "loss": 0.0255,
+      "step": 2260
+    },
+    {
+      "epoch": 0.9067305771919313,
+      "grad_norm": 2.86716890335083,
+      "learning_rate": 6.7836363636363644e-06,
+      "loss": 0.0302,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9107249850209707,
+      "grad_norm": 2.4357964992523193,
+      "learning_rate": 6.765454545454546e-06,
+      "loss": 0.0302,
+      "step": 2280
+    },
+    {
+      "epoch": 0.91471939285001,
+      "grad_norm": 2.8732433319091797,
+      "learning_rate": 6.747272727272728e-06,
+      "loss": 0.0289,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9187138006790493,
+      "grad_norm": 4.2874836921691895,
+      "learning_rate": 6.72909090909091e-06,
+      "loss": 0.0303,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9227082085080887,
+      "grad_norm": 2.6381747722625732,
+      "learning_rate": 6.710909090909091e-06,
+      "loss": 0.029,
+      "step": 2310
+    },
+    {
+      "epoch": 0.926702616337128,
+      "grad_norm": 2.819223403930664,
+      "learning_rate": 6.692727272727273e-06,
+      "loss": 0.0261,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9306970241661674,
+      "grad_norm": 2.7608511447906494,
+      "learning_rate": 6.674545454545455e-06,
+      "loss": 0.025,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9346914319952067,
+      "grad_norm": 2.7692387104034424,
+      "learning_rate": 6.6563636363636365e-06,
+      "loss": 0.0247,
+      "step": 2340
+    },
+    {
+      "epoch": 0.938685839824246,
+      "grad_norm": 3.5204415321350098,
+      "learning_rate": 6.638181818181819e-06,
+      "loss": 0.0377,
+      "step": 2350
+    },
+    {
+      "epoch": 0.9426802476532854,
+      "grad_norm": 3.3943400382995605,
+      "learning_rate": 6.620000000000001e-06,
+      "loss": 0.0315,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9466746554823248,
+      "grad_norm": 3.179027795791626,
+      "learning_rate": 6.601818181818183e-06,
+      "loss": 0.0269,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9506690633113641,
+      "grad_norm": 1.89017653465271,
+      "learning_rate": 6.583636363636365e-06,
+      "loss": 0.0294,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9546634711404034,
+      "grad_norm": 3.284336566925049,
+      "learning_rate": 6.565454545454546e-06,
+      "loss": 0.0395,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9586578789694428,
+      "grad_norm": 2.3796751499176025,
+      "learning_rate": 6.5472727272727275e-06,
+      "loss": 0.0337,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9626522867984821,
+      "grad_norm": 2.710698366165161,
+      "learning_rate": 6.529090909090909e-06,
+      "loss": 0.0276,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9666466946275215,
+      "grad_norm": Infinity,
+      "learning_rate": 6.510909090909091e-06,
+      "loss": 0.0459,
+      "step": 2420
+    },
+    {
+      "epoch": 0.9706411024565608,
+      "grad_norm": 3.24204421043396,
+      "learning_rate": 6.492727272727273e-06,
+      "loss": 0.0345,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9746355102856001,
+      "grad_norm": 4.349393844604492,
+      "learning_rate": 6.474545454545456e-06,
+      "loss": 0.0355,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9786299181146395,
+      "grad_norm": 2.544041395187378,
+      "learning_rate": 6.4563636363636375e-06,
+      "loss": 0.0326,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9826243259436789,
+      "grad_norm": 1.9386078119277954,
+      "learning_rate": 6.438181818181819e-06,
+      "loss": 0.028,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9866187337727182,
+      "grad_norm": 2.767906904220581,
+      "learning_rate": 6.42e-06,
+      "loss": 0.0244,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9906131416017575,
+      "grad_norm": 3.3528952598571777,
+      "learning_rate": 6.401818181818182e-06,
+      "loss": 0.03,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9946075494307969,
+      "grad_norm": 2.875234365463257,
+      "learning_rate": 6.383636363636364e-06,
+      "loss": 0.0323,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9986019572598362,
+      "grad_norm": 3.0124990940093994,
+      "learning_rate": 6.365454545454546e-06,
+      "loss": 0.0373,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0023966446974235,
+      "grad_norm": 1.3243356943130493,
+      "learning_rate": 6.347272727272728e-06,
+      "loss": 0.0143,
+      "step": 2510
+    },
+    {
+      "epoch": 1.006391052526463,
+      "grad_norm": 1.5087229013442993,
+      "learning_rate": 6.3290909090909096e-06,
+      "loss": 0.0078,
+      "step": 2520
+    },
+    {
+      "epoch": 1.0103854603555023,
+      "grad_norm": 2.5617003440856934,
+      "learning_rate": 6.3109090909090905e-06,
+      "loss": 0.0155,
+      "step": 2530
+    },
+    {
+      "epoch": 1.0143798681845417,
+      "grad_norm": 0.9475475549697876,
+      "learning_rate": 6.292727272727274e-06,
+      "loss": 0.0091,
+      "step": 2540
+    },
+    {
+      "epoch": 1.018374276013581,
+      "grad_norm": 1.712641716003418,
+      "learning_rate": 6.274545454545455e-06,
+      "loss": 0.013,
+      "step": 2550
+    },
+    {
+      "epoch": 1.0223686838426203,
+      "grad_norm": 0.9213377237319946,
+      "learning_rate": 6.256363636363637e-06,
+      "loss": 0.009,
+      "step": 2560
+    },
+    {
+      "epoch": 1.0263630916716597,
+      "grad_norm": 1.4909203052520752,
+      "learning_rate": 6.238181818181819e-06,
+      "loss": 0.0093,
+      "step": 2570
+    },
+    {
+      "epoch": 1.030357499500699,
+      "grad_norm": 0.989854633808136,
+      "learning_rate": 6.220000000000001e-06,
+      "loss": 0.0068,
+      "step": 2580
+    },
+    {
+      "epoch": 1.0343519073297385,
+      "grad_norm": 1.8446928262710571,
+      "learning_rate": 6.2018181818181824e-06,
+      "loss": 0.0125,
+      "step": 2590
+    },
+    {
+      "epoch": 1.0383463151587777,
+      "grad_norm": 0.5228179693222046,
+      "learning_rate": 6.183636363636364e-06,
+      "loss": 0.0074,
+      "step": 2600
+    },
+    {
+      "epoch": 1.042340722987817,
+      "grad_norm": 1.9997162818908691,
+      "learning_rate": 6.165454545454545e-06,
+      "loss": 0.0078,
+      "step": 2610
+    },
+    {
+      "epoch": 1.0463351308168565,
+      "grad_norm": 0.7548797130584717,
+      "learning_rate": 6.147272727272727e-06,
+      "loss": 0.0132,
+      "step": 2620
+    },
+    {
+      "epoch": 1.0503295386458957,
+      "grad_norm": 1.5827348232269287,
+      "learning_rate": 6.12909090909091e-06,
+      "loss": 0.0067,
+      "step": 2630
+    },
+    {
+      "epoch": 1.054323946474935,
+      "grad_norm": 1.7578275203704834,
+      "learning_rate": 6.110909090909092e-06,
+      "loss": 0.012,
+      "step": 2640
+    },
+    {
+      "epoch": 1.0583183543039745,
+      "grad_norm": 1.5093387365341187,
+      "learning_rate": 6.0927272727272735e-06,
+      "loss": 0.008,
+      "step": 2650
+    },
+    {
+      "epoch": 1.0623127621330137,
+      "grad_norm": 2.2290687561035156,
+      "learning_rate": 6.074545454545455e-06,
+      "loss": 0.0055,
+      "step": 2660
+    },
+    {
+      "epoch": 1.0663071699620532,
+      "grad_norm": 2.303945779800415,
+      "learning_rate": 6.056363636363637e-06,
+      "loss": 0.0099,
+      "step": 2670
+    },
+    {
+      "epoch": 1.0703015777910925,
+      "grad_norm": 0.6369355320930481,
+      "learning_rate": 6.038181818181819e-06,
+      "loss": 0.0063,
+      "step": 2680
+    },
+    {
+      "epoch": 1.0742959856201317,
+      "grad_norm": 0.5402636528015137,
+      "learning_rate": 6.02e-06,
+      "loss": 0.0113,
+      "step": 2690
+    },
+    {
+      "epoch": 1.0782903934491712,
+      "grad_norm": 1.1112077236175537,
+      "learning_rate": 6.001818181818182e-06,
+      "loss": 0.0121,
+      "step": 2700
+    },
+    {
+      "epoch": 1.0822848012782105,
+      "grad_norm": 1.1533619165420532,
+      "learning_rate": 5.983636363636364e-06,
+      "loss": 0.0093,
+      "step": 2710
+    },
+    {
+      "epoch": 1.08627920910725,
+      "grad_norm": 0.7646775245666504,
+      "learning_rate": 5.965454545454546e-06,
+      "loss": 0.0115,
+      "step": 2720
+    },
+    {
+      "epoch": 1.0902736169362892,
+      "grad_norm": 0.8578062057495117,
+      "learning_rate": 5.947272727272728e-06,
+      "loss": 0.0086,
+      "step": 2730
+    },
+    {
+      "epoch": 1.0942680247653285,
+      "grad_norm": 3.2737674713134766,
+      "learning_rate": 5.92909090909091e-06,
+      "loss": 0.0073,
+      "step": 2740
+    },
+    {
+      "epoch": 1.098262432594368,
+      "grad_norm": 2.4335756301879883,
+      "learning_rate": 5.910909090909092e-06,
+      "loss": 0.0101,
+      "step": 2750
+    },
+    {
+      "epoch": 1.1022568404234072,
+      "grad_norm": 1.416908860206604,
+      "learning_rate": 5.892727272727274e-06,
+      "loss": 0.0072,
+      "step": 2760
+    },
+    {
+      "epoch": 1.1062512482524465,
+      "grad_norm": 0.8457365036010742,
+      "learning_rate": 5.874545454545455e-06,
+      "loss": 0.0088,
+      "step": 2770
+    },
+    {
+      "epoch": 1.110245656081486,
+      "grad_norm": 1.1250872611999512,
+      "learning_rate": 5.8563636363636365e-06,
+      "loss": 0.0064,
+      "step": 2780
+    },
+    {
+      "epoch": 1.1142400639105252,
+      "grad_norm": 1.291493535041809,
+      "learning_rate": 5.838181818181818e-06,
+      "loss": 0.0099,
+      "step": 2790
+    },
+    {
+      "epoch": 1.1182344717395647,
+      "grad_norm": 0.5688849091529846,
+      "learning_rate": 5.82e-06,
+      "loss": 0.0085,
+      "step": 2800
+    },
+    {
+      "epoch": 1.122228879568604,
+      "grad_norm": 1.1171317100524902,
+      "learning_rate": 5.801818181818182e-06,
+      "loss": 0.0067,
+      "step": 2810
+    },
+    {
+      "epoch": 1.1262232873976432,
+      "grad_norm": 2.204411506652832,
+      "learning_rate": 5.783636363636365e-06,
+      "loss": 0.0097,
+      "step": 2820
+    },
+    {
+      "epoch": 1.1302176952266827,
+      "grad_norm": 0.7467852830886841,
+      "learning_rate": 5.7654545454545465e-06,
+      "loss": 0.0069,
+      "step": 2830
+    },
+    {
+      "epoch": 1.134212103055722,
+      "grad_norm": 1.518353819847107,
+      "learning_rate": 5.747272727272728e-06,
+      "loss": 0.0103,
+      "step": 2840
+    },
+    {
+      "epoch": 1.1382065108847614,
+      "grad_norm": 3.4078903198242188,
+      "learning_rate": 5.729090909090909e-06,
+      "loss": 0.0136,
+      "step": 2850
+    },
+    {
+      "epoch": 1.1422009187138007,
+      "grad_norm": 0.7704038619995117,
+      "learning_rate": 5.710909090909091e-06,
+      "loss": 0.006,
+      "step": 2860
+    },
+    {
+      "epoch": 1.14619532654284,
+      "grad_norm": 0.5476235151290894,
+      "learning_rate": 5.692727272727273e-06,
+      "loss": 0.0069,
+      "step": 2870
+    },
+    {
+      "epoch": 1.1501897343718794,
+      "grad_norm": 0.3564358353614807,
+      "learning_rate": 5.674545454545455e-06,
+      "loss": 0.0122,
+      "step": 2880
+    },
+    {
+      "epoch": 1.1541841422009187,
+      "grad_norm": 1.082381010055542,
+      "learning_rate": 5.656363636363637e-06,
+      "loss": 0.0078,
+      "step": 2890
+    },
+    {
+      "epoch": 1.158178550029958,
+      "grad_norm": 1.2779443264007568,
+      "learning_rate": 5.6381818181818186e-06,
+      "loss": 0.0101,
+      "step": 2900
+    },
+    {
+      "epoch": 1.1621729578589974,
+      "grad_norm": 1.3106609582901,
+      "learning_rate": 5.620000000000001e-06,
+      "loss": 0.0099,
+      "step": 2910
+    },
+    {
+      "epoch": 1.1661673656880367,
+      "grad_norm": 1.0210970640182495,
+      "learning_rate": 5.601818181818183e-06,
+      "loss": 0.0081,
+      "step": 2920
+    },
+    {
+      "epoch": 1.1701617735170762,
+      "grad_norm": 2.3123538494110107,
+      "learning_rate": 5.583636363636364e-06,
+      "loss": 0.0072,
+      "step": 2930
+    },
+    {
+      "epoch": 1.1741561813461154,
+      "grad_norm": 1.2326884269714355,
+      "learning_rate": 5.565454545454546e-06,
+      "loss": 0.0072,
+      "step": 2940
+    },
+    {
+      "epoch": 1.1781505891751547,
+      "grad_norm": 1.6470218896865845,
+      "learning_rate": 5.547272727272728e-06,
+      "loss": 0.0117,
+      "step": 2950
+    },
+    {
+      "epoch": 1.1821449970041942,
+      "grad_norm": 1.5907762050628662,
+      "learning_rate": 5.52909090909091e-06,
+      "loss": 0.0096,
+      "step": 2960
+    },
+    {
+      "epoch": 1.1861394048332334,
+      "grad_norm": 1.7873153686523438,
+      "learning_rate": 5.5109090909090914e-06,
+      "loss": 0.0076,
+      "step": 2970
+    },
+    {
+      "epoch": 1.190133812662273,
+      "grad_norm": 0.8777304291725159,
+      "learning_rate": 5.492727272727273e-06,
+      "loss": 0.0094,
+      "step": 2980
+    },
+    {
+      "epoch": 1.1941282204913122,
+      "grad_norm": 1.556298851966858,
+      "learning_rate": 5.474545454545454e-06,
+      "loss": 0.0098,
+      "step": 2990
+    },
+    {
+      "epoch": 1.1981226283203514,
+      "grad_norm": 2.023460865020752,
+      "learning_rate": 5.456363636363636e-06,
+      "loss": 0.0103,
+      "step": 3000
+    },
+    {
+      "epoch": 1.202117036149391,
+      "grad_norm": 0.8088117837905884,
+      "learning_rate": 5.438181818181819e-06,
+      "loss": 0.0091,
+      "step": 3010
+    },
+    {
+      "epoch": 1.2061114439784302,
+      "grad_norm": 1.834544062614441,
+      "learning_rate": 5.420000000000001e-06,
+      "loss": 0.0071,
+      "step": 3020
+    },
+    {
+      "epoch": 1.2101058518074694,
+      "grad_norm": 1.0826184749603271,
+      "learning_rate": 5.4018181818181825e-06,
+      "loss": 0.007,
+      "step": 3030
+    },
+    {
+      "epoch": 1.214100259636509,
+      "grad_norm": 1.198883056640625,
+      "learning_rate": 5.383636363636364e-06,
+      "loss": 0.0074,
+      "step": 3040
+    },
+    {
+      "epoch": 1.2180946674655482,
+      "grad_norm": 1.476663589477539,
+      "learning_rate": 5.365454545454546e-06,
+      "loss": 0.0084,
+      "step": 3050
+    },
+    {
+      "epoch": 1.2220890752945877,
+      "grad_norm": 2.082937717437744,
+      "learning_rate": 5.347272727272728e-06,
+      "loss": 0.0081,
+      "step": 3060
+    },
+    {
+      "epoch": 1.226083483123627,
+      "grad_norm": 1.7027249336242676,
+      "learning_rate": 5.329090909090909e-06,
+      "loss": 0.0055,
+      "step": 3070
+    },
+    {
+      "epoch": 1.2300778909526662,
+      "grad_norm": 1.8248317241668701,
+      "learning_rate": 5.310909090909091e-06,
+      "loss": 0.0106,
+      "step": 3080
+    },
+    {
+      "epoch": 1.2340722987817057,
+      "grad_norm": 0.6236146092414856,
+      "learning_rate": 5.292727272727273e-06,
+      "loss": 0.0064,
+      "step": 3090
+    },
+    {
+      "epoch": 1.238066706610745,
+      "grad_norm": 1.895790934562683,
+      "learning_rate": 5.274545454545455e-06,
+      "loss": 0.0062,
+      "step": 3100
+    },
+    {
+      "epoch": 1.2420611144397844,
+      "grad_norm": 2.1608922481536865,
+      "learning_rate": 5.256363636363637e-06,
+      "loss": 0.0077,
+      "step": 3110
+    },
+    {
+      "epoch": 1.2460555222688237,
+      "grad_norm": 0.884495735168457,
+      "learning_rate": 5.238181818181819e-06,
+      "loss": 0.005,
+      "step": 3120
+    },
+    {
+      "epoch": 1.250049930097863,
+      "grad_norm": 1.3650517463684082,
+      "learning_rate": 5.220000000000001e-06,
+      "loss": 0.0071,
+      "step": 3130
+    },
+    {
+      "epoch": 1.2540443379269024,
+      "grad_norm": 1.8378918170928955,
+      "learning_rate": 5.201818181818183e-06,
+      "loss": 0.0078,
+      "step": 3140
+    },
+    {
+      "epoch": 1.2580387457559417,
+      "grad_norm": 0.3236289322376251,
+      "learning_rate": 5.183636363636364e-06,
+      "loss": 0.0106,
+      "step": 3150
+    },
+    {
+      "epoch": 1.262033153584981,
+      "grad_norm": 1.896931529045105,
+      "learning_rate": 5.1654545454545455e-06,
+      "loss": 0.0103,
+      "step": 3160
+    },
+    {
+      "epoch": 1.2660275614140204,
+      "grad_norm": 0.38708317279815674,
+      "learning_rate": 5.147272727272727e-06,
+      "loss": 0.0101,
+      "step": 3170
+    },
+    {
+      "epoch": 1.2700219692430597,
+      "grad_norm": 0.8924418687820435,
+      "learning_rate": 5.129090909090909e-06,
+      "loss": 0.0084,
+      "step": 3180
+    },
+    {
+      "epoch": 1.2740163770720991,
+      "grad_norm": 1.9992389678955078,
+      "learning_rate": 5.110909090909091e-06,
+      "loss": 0.008,
+      "step": 3190
+    },
+    {
+      "epoch": 1.2780107849011384,
+      "grad_norm": 6.68002462387085,
+      "learning_rate": 5.092727272727274e-06,
+      "loss": 0.0079,
+      "step": 3200
+    },
+    {
+      "epoch": 1.2820051927301779,
+      "grad_norm": 0.4097543954849243,
+      "learning_rate": 5.0745454545454555e-06,
+      "loss": 0.007,
+      "step": 3210
+    },
+    {
+      "epoch": 1.2859996005592171,
+      "grad_norm": 2.126607656478882,
+      "learning_rate": 5.056363636363637e-06,
+      "loss": 0.0085,
+      "step": 3220
+    },
+    {
+      "epoch": 1.2899940083882564,
+      "grad_norm": 0.7770309448242188,
+      "learning_rate": 5.038181818181818e-06,
+      "loss": 0.0092,
+      "step": 3230
+    },
+    {
+      "epoch": 1.2939884162172959,
+      "grad_norm": 1.09266996383667,
+      "learning_rate": 5.02e-06,
+      "loss": 0.0154,
+      "step": 3240
+    },
+    {
+      "epoch": 1.2979828240463351,
+      "grad_norm": 0.644081175327301,
+      "learning_rate": 5.001818181818182e-06,
+      "loss": 0.0109,
+      "step": 3250
+    },
+    {
+      "epoch": 1.3019772318753744,
+      "grad_norm": 0.40339845418930054,
+      "learning_rate": 4.983636363636364e-06,
+      "loss": 0.0136,
+      "step": 3260
+    },
+    {
+      "epoch": 1.3059716397044139,
+      "grad_norm": 1.0117156505584717,
+      "learning_rate": 4.965454545454546e-06,
+      "loss": 0.0079,
+      "step": 3270
+    },
+    {
+      "epoch": 1.3099660475334531,
+      "grad_norm": 0.6380696296691895,
+      "learning_rate": 4.9472727272727276e-06,
+      "loss": 0.0055,
+      "step": 3280
+    },
+    {
+      "epoch": 1.3139604553624924,
+      "grad_norm": 2.3137929439544678,
+      "learning_rate": 4.929090909090909e-06,
+      "loss": 0.0081,
+      "step": 3290
+    },
+    {
+      "epoch": 1.3179548631915319,
+      "grad_norm": 1.307806372642517,
+      "learning_rate": 4.910909090909091e-06,
+      "loss": 0.0107,
+      "step": 3300
+    },
+    {
+      "epoch": 1.3219492710205711,
+      "grad_norm": 1.1640592813491821,
+      "learning_rate": 4.892727272727273e-06,
+      "loss": 0.0071,
+      "step": 3310
+    },
+    {
+      "epoch": 1.3259436788496106,
+      "grad_norm": 0.8527657389640808,
+      "learning_rate": 4.874545454545455e-06,
+      "loss": 0.0075,
+      "step": 3320
+    },
+    {
+      "epoch": 1.3299380866786499,
+      "grad_norm": 1.5638917684555054,
+      "learning_rate": 4.856363636363637e-06,
+      "loss": 0.0077,
+      "step": 3330
+    },
+    {
+      "epoch": 1.3339324945076894,
+      "grad_norm": 3.4176924228668213,
+      "learning_rate": 4.838181818181819e-06,
+      "loss": 0.0134,
+      "step": 3340
+    },
+    {
+      "epoch": 1.3379269023367286,
+      "grad_norm": 1.353973388671875,
+      "learning_rate": 4.8200000000000004e-06,
+      "loss": 0.0063,
+      "step": 3350
+    },
+    {
+      "epoch": 1.3419213101657679,
+      "grad_norm": 1.280344843864441,
+      "learning_rate": 4.801818181818182e-06,
+      "loss": 0.0131,
+      "step": 3360
+    },
+    {
+      "epoch": 1.3459157179948074,
+      "grad_norm": 1.0167405605316162,
+      "learning_rate": 4.783636363636364e-06,
+      "loss": 0.0091,
+      "step": 3370
+    },
+    {
+      "epoch": 1.3499101258238466,
+      "grad_norm": 0.347282350063324,
+      "learning_rate": 4.765454545454546e-06,
+      "loss": 0.0066,
+      "step": 3380
+    },
+    {
+      "epoch": 1.3539045336528859,
+      "grad_norm": 0.9135898351669312,
+      "learning_rate": 4.747272727272728e-06,
+      "loss": 0.0093,
+      "step": 3390
+    },
+    {
+      "epoch": 1.3578989414819254,
+      "grad_norm": 0.4328613877296448,
+      "learning_rate": 4.72909090909091e-06,
+      "loss": 0.0089,
+      "step": 3400
+    },
+    {
+      "epoch": 1.3618933493109646,
+      "grad_norm": 0.7179979681968689,
+      "learning_rate": 4.7109090909090915e-06,
+      "loss": 0.0078,
+      "step": 3410
+    },
+    {
+      "epoch": 1.3658877571400039,
+      "grad_norm": 1.748172402381897,
+      "learning_rate": 4.692727272727273e-06,
+      "loss": 0.0062,
+      "step": 3420
+    },
+    {
+      "epoch": 1.3698821649690434,
+      "grad_norm": 0.30844706296920776,
+      "learning_rate": 4.674545454545455e-06,
+      "loss": 0.0041,
+      "step": 3430
+    },
+    {
+      "epoch": 1.3738765727980826,
+      "grad_norm": 1.167373538017273,
+      "learning_rate": 4.656363636363637e-06,
+      "loss": 0.008,
+      "step": 3440
+    },
+    {
+      "epoch": 1.377870980627122,
+      "grad_norm": 1.5947142839431763,
+      "learning_rate": 4.638181818181818e-06,
+      "loss": 0.0067,
+      "step": 3450
+    },
+    {
+      "epoch": 1.3818653884561614,
+      "grad_norm": 1.5891377925872803,
+      "learning_rate": 4.620000000000001e-06,
+      "loss": 0.007,
+      "step": 3460
+    },
+    {
+      "epoch": 1.3858597962852008,
+      "grad_norm": 2.0473077297210693,
+      "learning_rate": 4.6018181818181825e-06,
+      "loss": 0.0096,
+      "step": 3470
+    },
+    {
+      "epoch": 1.38985420411424,
+      "grad_norm": 2.16921067237854,
+      "learning_rate": 4.583636363636364e-06,
+      "loss": 0.0084,
+      "step": 3480
+    },
+    {
+      "epoch": 1.3938486119432794,
+      "grad_norm": 2.0252652168273926,
+      "learning_rate": 4.565454545454545e-06,
+      "loss": 0.0081,
+      "step": 3490
+    },
+    {
+      "epoch": 1.3978430197723188,
+      "grad_norm": 1.2650831937789917,
+      "learning_rate": 4.547272727272727e-06,
+      "loss": 0.0065,
+      "step": 3500
+    },
+    {
+      "epoch": 1.401837427601358,
+      "grad_norm": 1.4343613386154175,
+      "learning_rate": 4.52909090909091e-06,
+      "loss": 0.0067,
+      "step": 3510
+    },
+    {
+      "epoch": 1.4058318354303974,
+      "grad_norm": 0.9264397025108337,
+      "learning_rate": 4.510909090909092e-06,
+      "loss": 0.0084,
+      "step": 3520
+    },
+    {
+      "epoch": 1.4098262432594368,
+      "grad_norm": 0.47705429792404175,
+      "learning_rate": 4.492727272727273e-06,
+      "loss": 0.0147,
+      "step": 3530
+    },
+    {
+      "epoch": 1.413820651088476,
+      "grad_norm": 0.33390942215919495,
+      "learning_rate": 4.4745454545454545e-06,
+      "loss": 0.0067,
+      "step": 3540
+    },
+    {
+      "epoch": 1.4178150589175154,
+      "grad_norm": 0.7790824174880981,
+      "learning_rate": 4.456363636363637e-06,
+      "loss": 0.0073,
+      "step": 3550
+    },
+    {
+      "epoch": 1.4218094667465548,
+      "grad_norm": 1.8825984001159668,
+      "learning_rate": 4.438181818181819e-06,
+      "loss": 0.0119,
+      "step": 3560
+    },
+    {
+      "epoch": 1.425803874575594,
+      "grad_norm": 2.6809818744659424,
+      "learning_rate": 4.42e-06,
+      "loss": 0.0076,
+      "step": 3570
+    },
+    {
+      "epoch": 1.4297982824046336,
+      "grad_norm": 1.4570934772491455,
+      "learning_rate": 4.401818181818182e-06,
+      "loss": 0.0073,
+      "step": 3580
+    },
+    {
+      "epoch": 1.4337926902336728,
+      "grad_norm": 0.7681152820587158,
+      "learning_rate": 4.383636363636364e-06,
+      "loss": 0.0044,
+      "step": 3590
+    },
+    {
+      "epoch": 1.4377870980627123,
+      "grad_norm": 0.2887667715549469,
+      "learning_rate": 4.365454545454546e-06,
+      "loss": 0.0058,
+      "step": 3600
+    },
+    {
+      "epoch": 1.4417815058917516,
+      "grad_norm": 1.9216909408569336,
+      "learning_rate": 4.347272727272727e-06,
+      "loss": 0.0079,
+      "step": 3610
+    },
+    {
+      "epoch": 1.4457759137207908,
+      "grad_norm": 1.959386944770813,
+      "learning_rate": 4.329090909090909e-06,
+      "loss": 0.007,
+      "step": 3620
+    },
+    {
+      "epoch": 1.4497703215498303,
+      "grad_norm": 1.421069860458374,
+      "learning_rate": 4.310909090909091e-06,
+      "loss": 0.0064,
+      "step": 3630
+    },
+    {
+      "epoch": 1.4537647293788696,
+      "grad_norm": 1.9688565731048584,
+      "learning_rate": 4.292727272727273e-06,
+      "loss": 0.0065,
+      "step": 3640
+    },
+    {
+      "epoch": 1.4577591372079088,
+      "grad_norm": 0.46297717094421387,
+      "learning_rate": 4.274545454545455e-06,
+      "loss": 0.0045,
+      "step": 3650
+    },
+    {
+      "epoch": 1.4617535450369483,
+      "grad_norm": 0.6084499359130859,
+      "learning_rate": 4.256363636363637e-06,
+      "loss": 0.0063,
+      "step": 3660
+    },
+    {
+      "epoch": 1.4657479528659876,
+      "grad_norm": 2.4733595848083496,
+      "learning_rate": 4.238181818181818e-06,
+      "loss": 0.0082,
+      "step": 3670
+    },
+    {
+      "epoch": 1.4697423606950268,
+      "grad_norm": 0.8938449025154114,
+      "learning_rate": 4.22e-06,
+      "loss": 0.0056,
+      "step": 3680
+    },
+    {
+      "epoch": 1.4737367685240663,
+      "grad_norm": 1.3058514595031738,
+      "learning_rate": 4.201818181818182e-06,
+      "loss": 0.0079,
+      "step": 3690
+    },
+    {
+      "epoch": 1.4777311763531056,
+      "grad_norm": 0.42573487758636475,
+      "learning_rate": 4.183636363636364e-06,
+      "loss": 0.0083,
+      "step": 3700
+    },
+    {
+      "epoch": 1.481725584182145,
+      "grad_norm": 2.336344003677368,
+      "learning_rate": 4.165454545454546e-06,
+      "loss": 0.0105,
+      "step": 3710
+    },
+    {
+      "epoch": 1.4857199920111843,
+      "grad_norm": 0.4930185079574585,
+      "learning_rate": 4.147272727272728e-06,
+      "loss": 0.0083,
+      "step": 3720
+    },
+    {
+      "epoch": 1.4897143998402238,
+      "grad_norm": 1.1534149646759033,
+      "learning_rate": 4.1290909090909094e-06,
+      "loss": 0.0057,
+      "step": 3730
+    },
+    {
+      "epoch": 1.493708807669263,
+      "grad_norm": 1.0539729595184326,
+      "learning_rate": 4.110909090909091e-06,
+      "loss": 0.0075,
+      "step": 3740
+    },
+    {
+      "epoch": 1.4977032154983023,
+      "grad_norm": 1.9535585641860962,
+      "learning_rate": 4.092727272727273e-06,
+      "loss": 0.0081,
+      "step": 3750
+    },
+    {
+      "epoch": 1.5016976233273418,
+      "grad_norm": 1.5772063732147217,
+      "learning_rate": 4.074545454545455e-06,
+      "loss": 0.0074,
+      "step": 3760
+    },
+    {
+      "epoch": 1.505692031156381,
+      "grad_norm": 1.9321742057800293,
+      "learning_rate": 4.056363636363637e-06,
+      "loss": 0.0076,
+      "step": 3770
+    },
+    {
+      "epoch": 1.5096864389854203,
+      "grad_norm": 1.391487717628479,
+      "learning_rate": 4.038181818181819e-06,
+      "loss": 0.0054,
+      "step": 3780
+    },
+    {
+      "epoch": 1.5136808468144598,
+      "grad_norm": 0.29940593242645264,
+      "learning_rate": 4.0200000000000005e-06,
+      "loss": 0.0075,
+      "step": 3790
+    },
+    {
+      "epoch": 1.517675254643499,
+      "grad_norm": 0.7316534519195557,
+      "learning_rate": 4.001818181818182e-06,
+      "loss": 0.0059,
+      "step": 3800
+    },
+    {
+      "epoch": 1.5216696624725383,
+      "grad_norm": 0.8725746273994446,
+      "learning_rate": 3.983636363636364e-06,
+      "loss": 0.0082,
+      "step": 3810
+    },
+    {
+      "epoch": 1.5256640703015778,
+      "grad_norm": 1.1659518480300903,
+      "learning_rate": 3.965454545454546e-06,
+      "loss": 0.0056,
+      "step": 3820
+    },
+    {
+      "epoch": 1.5296584781306173,
+      "grad_norm": 0.9450637698173523,
+      "learning_rate": 3.947272727272727e-06,
+      "loss": 0.0078,
+      "step": 3830
+    },
+    {
+      "epoch": 1.5336528859596563,
+      "grad_norm": 0.7942542433738708,
+      "learning_rate": 3.92909090909091e-06,
+      "loss": 0.0056,
+      "step": 3840
+    },
+    {
+      "epoch": 1.5376472937886958,
+      "grad_norm": 2.9248387813568115,
+      "learning_rate": 3.9109090909090915e-06,
+      "loss": 0.0065,
+      "step": 3850
+    },
+    {
+      "epoch": 1.5416417016177353,
+      "grad_norm": 0.9460102319717407,
+      "learning_rate": 3.892727272727273e-06,
+      "loss": 0.0063,
+      "step": 3860
+    },
+    {
+      "epoch": 1.5456361094467745,
+      "grad_norm": 1.5365135669708252,
+      "learning_rate": 3.874545454545454e-06,
+      "loss": 0.0091,
+      "step": 3870
+    },
+    {
+      "epoch": 1.5496305172758138,
+      "grad_norm": 1.6475716829299927,
+      "learning_rate": 3.856363636363636e-06,
+      "loss": 0.005,
+      "step": 3880
+    },
+    {
+      "epoch": 1.5536249251048533,
+      "grad_norm": 0.8353217244148254,
+      "learning_rate": 3.838181818181819e-06,
+      "loss": 0.0043,
+      "step": 3890
+    },
+    {
+      "epoch": 1.5576193329338925,
+      "grad_norm": 0.36758750677108765,
+      "learning_rate": 3.820000000000001e-06,
+      "loss": 0.004,
+      "step": 3900
+    },
+    {
+      "epoch": 1.5616137407629318,
+      "grad_norm": 0.2298198938369751,
+      "learning_rate": 3.801818181818182e-06,
+      "loss": 0.0084,
+      "step": 3910
+    },
+    {
+      "epoch": 1.5656081485919713,
+      "grad_norm": 1.170965313911438,
+      "learning_rate": 3.783636363636364e-06,
+      "loss": 0.0067,
+      "step": 3920
+    },
+    {
+      "epoch": 1.5696025564210105,
+      "grad_norm": 0.3110823631286621,
+      "learning_rate": 3.765454545454546e-06,
+      "loss": 0.0072,
+      "step": 3930
+    },
+    {
+      "epoch": 1.5735969642500498,
+      "grad_norm": 1.0501469373703003,
+      "learning_rate": 3.7472727272727276e-06,
+      "loss": 0.0101,
+      "step": 3940
+    },
+    {
+      "epoch": 1.5775913720790893,
+      "grad_norm": 1.3563507795333862,
+      "learning_rate": 3.7290909090909095e-06,
+      "loss": 0.0063,
+      "step": 3950
+    },
+    {
+      "epoch": 1.5815857799081288,
+      "grad_norm": 1.415356159210205,
+      "learning_rate": 3.7109090909090913e-06,
+      "loss": 0.0048,
+      "step": 3960
+    },
+    {
+      "epoch": 1.5855801877371678,
+      "grad_norm": 1.6954106092453003,
+      "learning_rate": 3.6927272727272727e-06,
+      "loss": 0.0083,
+      "step": 3970
+    },
+    {
+      "epoch": 1.5895745955662073,
+      "grad_norm": 0.9005348086357117,
+      "learning_rate": 3.674545454545455e-06,
+      "loss": 0.0093,
+      "step": 3980
+    },
+    {
+      "epoch": 1.5935690033952468,
+      "grad_norm": 0.7900477647781372,
+      "learning_rate": 3.656363636363637e-06,
+      "loss": 0.0087,
+      "step": 3990
+    },
+    {
+      "epoch": 1.597563411224286,
+      "grad_norm": 1.1566689014434814,
+      "learning_rate": 3.6381818181818187e-06,
+      "loss": 0.0072,
+      "step": 4000
+    },
+    {
+      "epoch": 1.597563411224286,
+      "eval_loss": 0.009391536004841328,
+      "eval_runtime": 7517.9662,
+      "eval_samples_per_second": 2.664,
+      "eval_steps_per_second": 0.333,
+      "eval_wer": 0.7816670479676657,
+      "step": 4000
+    },
+    {
+      "epoch": 1.6015578190533253,
+      "grad_norm": 1.0169645547866821,
+      "learning_rate": 3.62e-06,
+      "loss": 0.0093,
+      "step": 4010
+    },
+    {
+      "epoch": 1.6055522268823648,
+      "grad_norm": 1.4514238834381104,
+      "learning_rate": 3.601818181818182e-06,
+      "loss": 0.0079,
+      "step": 4020
+    },
+    {
+      "epoch": 1.609546634711404,
+      "grad_norm": 0.7925217747688293,
+      "learning_rate": 3.583636363636364e-06,
+      "loss": 0.0063,
+      "step": 4030
+    },
+    {
+      "epoch": 1.6135410425404433,
+      "grad_norm": 1.5382804870605469,
+      "learning_rate": 3.565454545454546e-06,
+      "loss": 0.008,
+      "step": 4040
+    },
+    {
+      "epoch": 1.6175354503694828,
+      "grad_norm": 1.5981707572937012,
+      "learning_rate": 3.5472727272727274e-06,
+      "loss": 0.0052,
+      "step": 4050
+    },
+    {
+      "epoch": 1.621529858198522,
+      "grad_norm": 0.7755517363548279,
+      "learning_rate": 3.5290909090909093e-06,
+      "loss": 0.0058,
+      "step": 4060
+    },
+    {
+      "epoch": 1.6255242660275613,
+      "grad_norm": 0.9957447648048401,
+      "learning_rate": 3.510909090909091e-06,
+      "loss": 0.0084,
+      "step": 4070
+    },
+    {
+      "epoch": 1.6295186738566008,
+      "grad_norm": 0.39248886704444885,
+      "learning_rate": 3.4927272727272734e-06,
+      "loss": 0.0091,
+      "step": 4080
+    },
+    {
+      "epoch": 1.6335130816856402,
+      "grad_norm": 0.7498955130577087,
+      "learning_rate": 3.4745454545454548e-06,
+      "loss": 0.0045,
+      "step": 4090
+    },
+    {
+      "epoch": 1.6375074895146795,
+      "grad_norm": 1.9002658128738403,
+      "learning_rate": 3.4563636363636366e-06,
+      "loss": 0.0053,
+      "step": 4100
+    },
+    {
+      "epoch": 1.6415018973437188,
+      "grad_norm": 1.1334868669509888,
+      "learning_rate": 3.4381818181818185e-06,
+      "loss": 0.0082,
+      "step": 4110
+    },
+    {
+      "epoch": 1.6454963051727582,
+      "grad_norm": 0.7285448908805847,
+      "learning_rate": 3.4200000000000007e-06,
+      "loss": 0.005,
+      "step": 4120
+    },
+    {
+      "epoch": 1.6494907130017975,
+      "grad_norm": 0.21141596138477325,
+      "learning_rate": 3.401818181818182e-06,
+      "loss": 0.0066,
+      "step": 4130
+    },
+    {
+      "epoch": 1.6534851208308368,
+      "grad_norm": 1.2618212699890137,
+      "learning_rate": 3.383636363636364e-06,
+      "loss": 0.0062,
+      "step": 4140
+    },
+    {
+      "epoch": 1.6574795286598762,
+      "grad_norm": 0.4843064844608307,
+      "learning_rate": 3.365454545454546e-06,
+      "loss": 0.0113,
+      "step": 4150
+    },
+    {
+      "epoch": 1.6614739364889155,
+      "grad_norm": 1.6873738765716553,
+      "learning_rate": 3.3472727272727272e-06,
+      "loss": 0.0074,
+      "step": 4160
+    },
+    {
+      "epoch": 1.6654683443179548,
+      "grad_norm": 1.6466562747955322,
+      "learning_rate": 3.3290909090909095e-06,
+      "loss": 0.0069,
+      "step": 4170
+    },
+    {
+      "epoch": 1.6694627521469942,
+      "grad_norm": 1.5858378410339355,
+      "learning_rate": 3.3109090909090913e-06,
+      "loss": 0.0052,
+      "step": 4180
+    },
+    {
+      "epoch": 1.6734571599760335,
+      "grad_norm": 0.9765921235084534,
+      "learning_rate": 3.292727272727273e-06,
+      "loss": 0.0077,
+      "step": 4190
+    },
+    {
+      "epoch": 1.6774515678050728,
+      "grad_norm": 0.858818769454956,
+      "learning_rate": 3.2745454545454546e-06,
+      "loss": 0.0064,
+      "step": 4200
+    },
+    {
+      "epoch": 1.6814459756341122,
+      "grad_norm": 2.0127484798431396,
+      "learning_rate": 3.2563636363636364e-06,
+      "loss": 0.0051,
+      "step": 4210
+    },
+    {
+      "epoch": 1.6854403834631517,
+      "grad_norm": 0.588580846786499,
+      "learning_rate": 3.2381818181818187e-06,
+      "loss": 0.0067,
+      "step": 4220
+    },
+    {
+      "epoch": 1.689434791292191,
+      "grad_norm": 0.42130523920059204,
+      "learning_rate": 3.2200000000000005e-06,
+      "loss": 0.0083,
+      "step": 4230
+    },
+    {
+      "epoch": 1.6934291991212302,
+      "grad_norm": 0.5209999680519104,
+      "learning_rate": 3.201818181818182e-06,
+      "loss": 0.0068,
+      "step": 4240
+    },
+    {
+      "epoch": 1.6974236069502697,
+      "grad_norm": 0.638425886631012,
+      "learning_rate": 3.1836363636363638e-06,
+      "loss": 0.0078,
+      "step": 4250
+    },
+    {
+      "epoch": 1.701418014779309,
+      "grad_norm": 2.0498859882354736,
+      "learning_rate": 3.1654545454545456e-06,
+      "loss": 0.0064,
+      "step": 4260
+    },
+    {
+      "epoch": 1.7054124226083482,
+      "grad_norm": 0.8182183504104614,
+      "learning_rate": 3.147272727272728e-06,
+      "loss": 0.0059,
+      "step": 4270
+    },
+    {
+      "epoch": 1.7094068304373877,
+      "grad_norm": 2.9371860027313232,
+      "learning_rate": 3.1290909090909093e-06,
+      "loss": 0.0072,
+      "step": 4280
+    },
+    {
+      "epoch": 1.713401238266427,
+      "grad_norm": 1.737328052520752,
+      "learning_rate": 3.110909090909091e-06,
+      "loss": 0.0069,
+      "step": 4290
+    },
+    {
+      "epoch": 1.7173956460954662,
+      "grad_norm": 0.9384961724281311,
+      "learning_rate": 3.092727272727273e-06,
+      "loss": 0.0064,
+      "step": 4300
+    },
+    {
+      "epoch": 1.7213900539245057,
+      "grad_norm": 2.61478328704834,
+      "learning_rate": 3.0745454545454552e-06,
+      "loss": 0.0066,
+      "step": 4310
+    },
+    {
+      "epoch": 1.725384461753545,
+      "grad_norm": 0.6870206594467163,
+      "learning_rate": 3.0563636363636366e-06,
+      "loss": 0.0073,
+      "step": 4320
+    },
+    {
+      "epoch": 1.7293788695825842,
+      "grad_norm": 1.5444024801254272,
+      "learning_rate": 3.0381818181818185e-06,
+      "loss": 0.0071,
+      "step": 4330
+    },
+    {
+      "epoch": 1.7333732774116237,
+      "grad_norm": 1.0944797992706299,
+      "learning_rate": 3.0200000000000003e-06,
+      "loss": 0.0042,
+      "step": 4340
+    },
+    {
+      "epoch": 1.7373676852406632,
+      "grad_norm": 1.1846562623977661,
+      "learning_rate": 3.0018181818181817e-06,
+      "loss": 0.008,
+      "step": 4350
+    },
+    {
+      "epoch": 1.7413620930697025,
+      "grad_norm": 0.3231483995914459,
+      "learning_rate": 2.983636363636364e-06,
+      "loss": 0.0056,
+      "step": 4360
+    },
+    {
+      "epoch": 1.7453565008987417,
+      "grad_norm": 1.571178913116455,
+      "learning_rate": 2.965454545454546e-06,
+      "loss": 0.0059,
+      "step": 4370
+    },
+    {
+      "epoch": 1.7493509087277812,
+      "grad_norm": 1.2889162302017212,
+      "learning_rate": 2.9472727272727277e-06,
+      "loss": 0.0076,
+      "step": 4380
+    },
+    {
+      "epoch": 1.7533453165568205,
+      "grad_norm": 0.7713314890861511,
+      "learning_rate": 2.929090909090909e-06,
+      "loss": 0.0066,
+      "step": 4390
+    },
+    {
+      "epoch": 1.7573397243858597,
+      "grad_norm": 2.216398000717163,
+      "learning_rate": 2.910909090909091e-06,
+      "loss": 0.0057,
+      "step": 4400
+    },
+    {
+      "epoch": 1.7613341322148992,
+      "grad_norm": 2.5534586906433105,
+      "learning_rate": 2.892727272727273e-06,
+      "loss": 0.0098,
+      "step": 4410
+    },
+    {
+      "epoch": 1.7653285400439385,
+      "grad_norm": 1.9392699003219604,
+      "learning_rate": 2.874545454545455e-06,
+      "loss": 0.0064,
+      "step": 4420
+    },
+    {
+      "epoch": 1.7693229478729777,
+      "grad_norm": 0.7885441780090332,
+      "learning_rate": 2.8563636363636364e-06,
+      "loss": 0.0072,
+      "step": 4430
+    },
+    {
+      "epoch": 1.7733173557020172,
+      "grad_norm": 0.9948071241378784,
+      "learning_rate": 2.8381818181818183e-06,
+      "loss": 0.0071,
+      "step": 4440
+    },
+    {
+      "epoch": 1.7773117635310565,
+      "grad_norm": 0.9726841449737549,
+      "learning_rate": 2.82e-06,
+      "loss": 0.0096,
+      "step": 4450
+    },
+    {
+      "epoch": 1.7813061713600957,
+      "grad_norm": 1.1720293760299683,
+      "learning_rate": 2.8018181818181824e-06,
+      "loss": 0.006,
+      "step": 4460
+    },
+    {
+      "epoch": 1.7853005791891352,
+      "grad_norm": 0.894111692905426,
+      "learning_rate": 2.7836363636363638e-06,
+      "loss": 0.0079,
+      "step": 4470
+    },
+    {
+      "epoch": 1.7892949870181747,
+      "grad_norm": 1.6590758562088013,
+      "learning_rate": 2.7654545454545456e-06,
+      "loss": 0.009,
+      "step": 4480
+    },
+    {
+      "epoch": 1.793289394847214,
+      "grad_norm": 2.0424230098724365,
+      "learning_rate": 2.7472727272727275e-06,
+      "loss": 0.0064,
+      "step": 4490
+    },
+    {
+      "epoch": 1.7972838026762532,
+      "grad_norm": 2.1726737022399902,
+      "learning_rate": 2.7290909090909097e-06,
+      "loss": 0.0078,
+      "step": 4500
+    },
+    {
+      "epoch": 1.8012782105052927,
+      "grad_norm": 0.4029232859611511,
+      "learning_rate": 2.710909090909091e-06,
+      "loss": 0.0105,
+      "step": 4510
+    },
+    {
+      "epoch": 1.805272618334332,
+      "grad_norm": 1.8009533882141113,
+      "learning_rate": 2.692727272727273e-06,
+      "loss": 0.0085,
+      "step": 4520
+    },
+    {
+      "epoch": 1.8092670261633712,
+      "grad_norm": 1.1566264629364014,
+      "learning_rate": 2.674545454545455e-06,
+      "loss": 0.0039,
+      "step": 4530
+    },
+    {
+      "epoch": 1.8132614339924107,
+      "grad_norm": 1.4425307512283325,
+      "learning_rate": 2.6563636363636362e-06,
+      "loss": 0.0074,
+      "step": 4540
+    },
+    {
+      "epoch": 1.81725584182145,
+      "grad_norm": 0.8721437454223633,
+      "learning_rate": 2.6381818181818185e-06,
+      "loss": 0.0064,
+      "step": 4550
+    },
+    {
+      "epoch": 1.8212502496504892,
+      "grad_norm": 0.976674497127533,
+      "learning_rate": 2.6200000000000003e-06,
+      "loss": 0.0033,
+      "step": 4560
+    },
+    {
+      "epoch": 1.8252446574795287,
+      "grad_norm": 0.8307082056999207,
+      "learning_rate": 2.601818181818182e-06,
+      "loss": 0.0053,
+      "step": 4570
+    },
+    {
+      "epoch": 1.8292390653085682,
+      "grad_norm": 0.25643059611320496,
+      "learning_rate": 2.5836363636363636e-06,
+      "loss": 0.0079,
+      "step": 4580
+    },
+    {
+      "epoch": 1.8332334731376072,
+      "grad_norm": 2.7886087894439697,
+      "learning_rate": 2.5654545454545454e-06,
+      "loss": 0.0088,
+      "step": 4590
+    },
+    {
+      "epoch": 1.8372278809666467,
+      "grad_norm": 0.3735600411891937,
+      "learning_rate": 2.5472727272727277e-06,
+      "loss": 0.0071,
+      "step": 4600
+    },
+    {
+      "epoch": 1.8412222887956862,
+      "grad_norm": 0.8093706369400024,
+      "learning_rate": 2.5290909090909095e-06,
+      "loss": 0.0086,
+      "step": 4610
+    },
+    {
+      "epoch": 1.8452166966247254,
+      "grad_norm": 3.601065158843994,
+      "learning_rate": 2.510909090909091e-06,
+      "loss": 0.0061,
+      "step": 4620
+    },
+    {
+      "epoch": 1.8492111044537647,
+      "grad_norm": 0.9476358890533447,
+      "learning_rate": 2.492727272727273e-06,
+      "loss": 0.0083,
+      "step": 4630
+    },
+    {
+      "epoch": 1.8532055122828042,
+      "grad_norm": 1.1815894842147827,
+      "learning_rate": 2.4745454545454546e-06,
+      "loss": 0.0108,
+      "step": 4640
+    },
+    {
+      "epoch": 1.8571999201118434,
+      "grad_norm": 0.382880300283432,
+      "learning_rate": 2.4563636363636364e-06,
+      "loss": 0.0042,
+      "step": 4650
+    },
+    {
+      "epoch": 1.8611943279408827,
+      "grad_norm": 0.46276092529296875,
+      "learning_rate": 2.4381818181818183e-06,
+      "loss": 0.0071,
+      "step": 4660
+    },
+    {
+      "epoch": 1.8651887357699222,
+      "grad_norm": 0.8990813493728638,
+      "learning_rate": 2.42e-06,
+      "loss": 0.0061,
+      "step": 4670
+    },
+    {
+      "epoch": 1.8691831435989614,
+      "grad_norm": 0.8797413110733032,
+      "learning_rate": 2.401818181818182e-06,
+      "loss": 0.005,
+      "step": 4680
+    },
+    {
+      "epoch": 1.8731775514280007,
+      "grad_norm": 2.5706422328948975,
+      "learning_rate": 2.383636363636364e-06,
+      "loss": 0.0064,
+      "step": 4690
+    },
+    {
+      "epoch": 1.8771719592570402,
+      "grad_norm": 1.353135108947754,
+      "learning_rate": 2.3654545454545456e-06,
+      "loss": 0.0042,
+      "step": 4700
+    },
+    {
+      "epoch": 1.8811663670860796,
+      "grad_norm": 0.8188295960426331,
+      "learning_rate": 2.3472727272727275e-06,
+      "loss": 0.01,
+      "step": 4710
+    },
+    {
+      "epoch": 1.8851607749151187,
+      "grad_norm": 1.070620059967041,
+      "learning_rate": 2.3290909090909093e-06,
+      "loss": 0.0089,
+      "step": 4720
+    },
+    {
+      "epoch": 1.8891551827441582,
+      "grad_norm": 0.45555347204208374,
+      "learning_rate": 2.310909090909091e-06,
+      "loss": 0.006,
+      "step": 4730
+    },
+    {
+      "epoch": 1.8931495905731976,
+      "grad_norm": 0.24032829701900482,
+      "learning_rate": 2.292727272727273e-06,
+      "loss": 0.0048,
+      "step": 4740
+    },
+    {
+      "epoch": 1.897143998402237,
+      "grad_norm": 1.4816186428070068,
+      "learning_rate": 2.274545454545455e-06,
+      "loss": 0.006,
+      "step": 4750
+    },
+    {
+      "epoch": 1.9011384062312762,
+      "grad_norm": 0.911095380783081,
+      "learning_rate": 2.2563636363636367e-06,
+      "loss": 0.0055,
+      "step": 4760
+    },
+    {
+      "epoch": 1.9051328140603156,
+      "grad_norm": 1.3121198415756226,
+      "learning_rate": 2.238181818181818e-06,
+      "loss": 0.0076,
+      "step": 4770
+    },
+    {
+      "epoch": 1.909127221889355,
+      "grad_norm": 0.4576851725578308,
+      "learning_rate": 2.2200000000000003e-06,
+      "loss": 0.0076,
+      "step": 4780
+    },
+    {
+      "epoch": 1.9131216297183942,
+      "grad_norm": 0.4526568055152893,
+      "learning_rate": 2.2018181818181818e-06,
+      "loss": 0.0049,
+      "step": 4790
+    },
+    {
+      "epoch": 1.9171160375474336,
+      "grad_norm": 2.19838285446167,
+      "learning_rate": 2.183636363636364e-06,
+      "loss": 0.007,
+      "step": 4800
+    },
+    {
+      "epoch": 1.921110445376473,
+      "grad_norm": 1.7481430768966675,
+      "learning_rate": 2.1654545454545454e-06,
+      "loss": 0.0075,
+      "step": 4810
+    },
+    {
+      "epoch": 1.9251048532055122,
+      "grad_norm": 1.511225938796997,
+      "learning_rate": 2.1472727272727277e-06,
+      "loss": 0.0053,
+      "step": 4820
+    },
+    {
+      "epoch": 1.9290992610345516,
+      "grad_norm": 0.44884830713272095,
+      "learning_rate": 2.129090909090909e-06,
+      "loss": 0.0054,
+      "step": 4830
+    },
+    {
+      "epoch": 1.9330936688635911,
+      "grad_norm": 1.4792394638061523,
+      "learning_rate": 2.110909090909091e-06,
+      "loss": 0.0063,
+      "step": 4840
+    },
+    {
+      "epoch": 1.9370880766926302,
+      "grad_norm": 1.228148102760315,
+      "learning_rate": 2.092727272727273e-06,
+      "loss": 0.0049,
+      "step": 4850
+    },
+    {
+      "epoch": 1.9410824845216696,
+      "grad_norm": 1.785239338874817,
+      "learning_rate": 2.0745454545454546e-06,
+      "loss": 0.0082,
+      "step": 4860
+    },
+    {
+      "epoch": 1.9450768923507091,
+      "grad_norm": 1.5273386240005493,
+      "learning_rate": 2.0563636363636365e-06,
+      "loss": 0.0049,
+      "step": 4870
+    },
+    {
+      "epoch": 1.9490713001797484,
+      "grad_norm": 1.5013206005096436,
+      "learning_rate": 2.0381818181818183e-06,
+      "loss": 0.0064,
+      "step": 4880
+    },
+    {
+      "epoch": 1.9530657080087876,
+      "grad_norm": 2.1962521076202393,
+      "learning_rate": 2.02e-06,
+      "loss": 0.0046,
+      "step": 4890
+    },
+    {
+      "epoch": 1.9570601158378271,
+      "grad_norm": 1.8651431798934937,
+      "learning_rate": 2.001818181818182e-06,
+      "loss": 0.0067,
+      "step": 4900
+    },
+    {
+      "epoch": 1.9610545236668664,
+      "grad_norm": 0.9547003507614136,
+      "learning_rate": 1.983636363636364e-06,
+      "loss": 0.0062,
+      "step": 4910
+    },
+    {
+      "epoch": 1.9650489314959056,
+      "grad_norm": 1.2471400499343872,
+      "learning_rate": 1.9654545454545457e-06,
+      "loss": 0.005,
+      "step": 4920
+    },
+    {
+      "epoch": 1.9690433393249451,
+      "grad_norm": 0.28791195154190063,
+      "learning_rate": 1.9472727272727275e-06,
+      "loss": 0.007,
+      "step": 4930
+    },
+    {
+      "epoch": 1.9730377471539844,
+      "grad_norm": 1.5757160186767578,
+      "learning_rate": 1.9290909090909093e-06,
+      "loss": 0.0069,
+      "step": 4940
+    },
+    {
+      "epoch": 1.9770321549830236,
+      "grad_norm": 0.5068025588989258,
+      "learning_rate": 1.910909090909091e-06,
+      "loss": 0.0077,
+      "step": 4950
+    },
+    {
+      "epoch": 1.9810265628120631,
+      "grad_norm": 0.2998751699924469,
+      "learning_rate": 1.8927272727272728e-06,
+      "loss": 0.0045,
+      "step": 4960
+    },
+    {
+      "epoch": 1.9850209706411026,
+      "grad_norm": 0.4755394756793976,
+      "learning_rate": 1.8745454545454546e-06,
+      "loss": 0.0044,
+      "step": 4970
+    },
+    {
+      "epoch": 1.9890153784701416,
+      "grad_norm": 1.0987839698791504,
+      "learning_rate": 1.8563636363636365e-06,
+      "loss": 0.0075,
+      "step": 4980
+    },
+    {
+      "epoch": 1.9930097862991811,
+      "grad_norm": 1.1204299926757812,
+      "learning_rate": 1.8381818181818183e-06,
+      "loss": 0.0066,
+      "step": 4990
+    },
+    {
+      "epoch": 1.9970041941282206,
+      "grad_norm": 0.4197748005390167,
+      "learning_rate": 1.8200000000000002e-06,
+      "loss": 0.0055,
+      "step": 5000
+    },
+    {
+      "epoch": 2.000798881565808,
+      "grad_norm": 0.19804427027702332,
+      "learning_rate": 1.801818181818182e-06,
+      "loss": 0.0044,
+      "step": 5010
+    },
+    {
+      "epoch": 2.004793289394847,
+      "grad_norm": 0.36684417724609375,
+      "learning_rate": 1.7836363636363638e-06,
+      "loss": 0.0027,
+      "step": 5020
+    },
+    {
+      "epoch": 2.0087876972238865,
+      "grad_norm": 0.6657086610794067,
+      "learning_rate": 1.7654545454545455e-06,
+      "loss": 0.0017,
+      "step": 5030
+    },
+    {
+      "epoch": 2.012782105052926,
+      "grad_norm": 0.05520480126142502,
+      "learning_rate": 1.7472727272727275e-06,
+      "loss": 0.0016,
+      "step": 5040
+    },
+    {
+      "epoch": 2.0167765128819655,
+      "grad_norm": 0.09664568305015564,
+      "learning_rate": 1.7290909090909091e-06,
+      "loss": 0.0019,
+      "step": 5050
+    },
+    {
+      "epoch": 2.0207709207110045,
+      "grad_norm": 0.37269479036331177,
+      "learning_rate": 1.7109090909090912e-06,
+      "loss": 0.0018,
+      "step": 5060
+    },
+    {
+      "epoch": 2.024765328540044,
+      "grad_norm": 0.11714975535869598,
+      "learning_rate": 1.6927272727272728e-06,
+      "loss": 0.0013,
+      "step": 5070
+    },
+    {
+      "epoch": 2.0287597363690835,
+      "grad_norm": 0.24795569479465485,
+      "learning_rate": 1.6745454545454549e-06,
+      "loss": 0.0014,
+      "step": 5080
+    },
+    {
+      "epoch": 2.0327541441981225,
+      "grad_norm": 0.12845009565353394,
+      "learning_rate": 1.6563636363636365e-06,
+      "loss": 0.0026,
+      "step": 5090
+    },
+    {
+      "epoch": 2.036748552027162,
+      "grad_norm": 0.2782312035560608,
+      "learning_rate": 1.6381818181818181e-06,
+      "loss": 0.0017,
+      "step": 5100
+    },
+    {
+      "epoch": 2.0407429598562015,
+      "grad_norm": 0.15887074172496796,
+      "learning_rate": 1.6200000000000002e-06,
+      "loss": 0.0014,
+      "step": 5110
+    },
+    {
+      "epoch": 2.0447373676852405,
+      "grad_norm": 0.13874128460884094,
+      "learning_rate": 1.6018181818181818e-06,
+      "loss": 0.0023,
+      "step": 5120
+    },
+    {
+      "epoch": 2.04873177551428,
+      "grad_norm": 0.34651249647140503,
+      "learning_rate": 1.5836363636363638e-06,
+      "loss": 0.0036,
+      "step": 5130
+    },
+    {
+      "epoch": 2.0527261833433195,
+      "grad_norm": 0.1842481791973114,
+      "learning_rate": 1.5654545454545455e-06,
+      "loss": 0.0016,
+      "step": 5140
+    },
+    {
+      "epoch": 2.0567205911723585,
+      "grad_norm": 0.16887687146663666,
+      "learning_rate": 1.5472727272727275e-06,
+      "loss": 0.0031,
+      "step": 5150
+    },
+    {
+      "epoch": 2.060714999001398,
+      "grad_norm": 0.24830514192581177,
+      "learning_rate": 1.5290909090909091e-06,
+      "loss": 0.0014,
+      "step": 5160
+    },
+    {
+      "epoch": 2.0647094068304375,
+      "grad_norm": 0.08864196389913559,
+      "learning_rate": 1.510909090909091e-06,
+      "loss": 0.0014,
+      "step": 5170
+    },
+    {
+      "epoch": 2.068703814659477,
+      "grad_norm": 0.2347860485315323,
+      "learning_rate": 1.4927272727272728e-06,
+      "loss": 0.0015,
+      "step": 5180
+    },
+    {
+      "epoch": 2.072698222488516,
+      "grad_norm": 0.17117437720298767,
+      "learning_rate": 1.4745454545454547e-06,
+      "loss": 0.0025,
+      "step": 5190
+    },
+    {
+      "epoch": 2.0766926303175555,
+      "grad_norm": 1.6268408298492432,
+      "learning_rate": 1.4563636363636365e-06,
+      "loss": 0.004,
+      "step": 5200
+    },
+    {
+      "epoch": 2.080687038146595,
+      "grad_norm": 0.3034919202327728,
+      "learning_rate": 1.4381818181818183e-06,
+      "loss": 0.0015,
+      "step": 5210
+    },
+    {
+      "epoch": 2.084681445975634,
+      "grad_norm": 0.15098127722740173,
+      "learning_rate": 1.42e-06,
+      "loss": 0.0041,
+      "step": 5220
+    },
+    {
+      "epoch": 2.0886758538046735,
+      "grad_norm": 0.5514845252037048,
+      "learning_rate": 1.401818181818182e-06,
+      "loss": 0.001,
+      "step": 5230
+    },
+    {
+      "epoch": 2.092670261633713,
+      "grad_norm": 0.9603298306465149,
+      "learning_rate": 1.3836363636363636e-06,
+      "loss": 0.0017,
+      "step": 5240
+    },
+    {
+      "epoch": 2.096664669462752,
+      "grad_norm": 0.23426704108715057,
+      "learning_rate": 1.3654545454545457e-06,
+      "loss": 0.0016,
+      "step": 5250
+    },
+    {
+      "epoch": 2.1006590772917915,
+      "grad_norm": 0.1383952647447586,
+      "learning_rate": 1.3472727272727273e-06,
+      "loss": 0.0013,
+      "step": 5260
+    },
+    {
+      "epoch": 2.104653485120831,
+      "grad_norm": 0.22570370137691498,
+      "learning_rate": 1.3290909090909094e-06,
+      "loss": 0.002,
+      "step": 5270
+    },
+    {
+      "epoch": 2.10864789294987,
+      "grad_norm": 0.2061236947774887,
+      "learning_rate": 1.310909090909091e-06,
+      "loss": 0.001,
+      "step": 5280
+    },
+    {
+      "epoch": 2.1126423007789095,
+      "grad_norm": 0.1093958392739296,
+      "learning_rate": 1.2927272727272728e-06,
+      "loss": 0.0026,
+      "step": 5290
+    },
+    {
+      "epoch": 2.116636708607949,
+      "grad_norm": 0.11976289749145508,
+      "learning_rate": 1.2745454545454547e-06,
+      "loss": 0.0033,
+      "step": 5300
+    },
+    {
+      "epoch": 2.1206311164369884,
+      "grad_norm": 0.18352824449539185,
+      "learning_rate": 1.2563636363636365e-06,
+      "loss": 0.0011,
+      "step": 5310
+    },
+    {
+      "epoch": 2.1246255242660275,
+      "grad_norm": 0.11029355227947235,
+      "learning_rate": 1.2381818181818183e-06,
+      "loss": 0.0013,
+      "step": 5320
+    },
+    {
+      "epoch": 2.128619932095067,
+      "grad_norm": 0.8353503942489624,
+      "learning_rate": 1.2200000000000002e-06,
+      "loss": 0.0023,
+      "step": 5330
+    },
+    {
+      "epoch": 2.1326143399241064,
+      "grad_norm": 0.30112984776496887,
+      "learning_rate": 1.201818181818182e-06,
+      "loss": 0.0012,
+      "step": 5340
+    },
+    {
+      "epoch": 2.1366087477531455,
+      "grad_norm": 0.4989093542098999,
+      "learning_rate": 1.1836363636363639e-06,
+      "loss": 0.0019,
+      "step": 5350
+    },
+    {
+      "epoch": 2.140603155582185,
+      "grad_norm": 0.09659893065690994,
+      "learning_rate": 1.1654545454545457e-06,
+      "loss": 0.002,
+      "step": 5360
+    },
+    {
+      "epoch": 2.1445975634112244,
+      "grad_norm": 0.10450714826583862,
+      "learning_rate": 1.1472727272727275e-06,
+      "loss": 0.0033,
+      "step": 5370
+    },
+    {
+      "epoch": 2.1485919712402635,
+      "grad_norm": 0.1421971321105957,
+      "learning_rate": 1.1290909090909092e-06,
+      "loss": 0.0011,
+      "step": 5380
+    },
+    {
+      "epoch": 2.152586379069303,
+      "grad_norm": 0.13540399074554443,
+      "learning_rate": 1.110909090909091e-06,
+      "loss": 0.0015,
+      "step": 5390
+    },
+    {
+      "epoch": 2.1565807868983424,
+      "grad_norm": 0.13781139254570007,
+      "learning_rate": 1.0927272727272728e-06,
+      "loss": 0.0015,
+      "step": 5400
+    },
+    {
+      "epoch": 2.1605751947273815,
+      "grad_norm": 0.12392697483301163,
+      "learning_rate": 1.0745454545454547e-06,
+      "loss": 0.0024,
+      "step": 5410
+    },
+    {
+      "epoch": 2.164569602556421,
+      "grad_norm": 0.326977014541626,
+      "learning_rate": 1.0563636363636365e-06,
+      "loss": 0.0023,
+      "step": 5420
+    },
+    {
+      "epoch": 2.1685640103854604,
+      "grad_norm": 0.1042676791548729,
+      "learning_rate": 1.0381818181818183e-06,
+      "loss": 0.0018,
+      "step": 5430
+    },
+    {
+      "epoch": 2.1725584182145,
+      "grad_norm": 0.16718903183937073,
+      "learning_rate": 1.02e-06,
+      "loss": 0.0013,
+      "step": 5440
+    },
+    {
+      "epoch": 2.176552826043539,
+      "grad_norm": 1.135487675666809,
+      "learning_rate": 1.0018181818181818e-06,
+      "loss": 0.0022,
+      "step": 5450
+    },
+    {
+      "epoch": 2.1805472338725784,
+      "grad_norm": 0.1465596705675125,
+      "learning_rate": 9.836363636363637e-07,
+      "loss": 0.002,
+      "step": 5460
+    },
+    {
+      "epoch": 2.184541641701618,
+      "grad_norm": 0.18476231396198273,
+      "learning_rate": 9.654545454545455e-07,
+      "loss": 0.0016,
+      "step": 5470
+    },
+    {
+      "epoch": 2.188536049530657,
+      "grad_norm": 0.1280352771282196,
+      "learning_rate": 9.472727272727273e-07,
+      "loss": 0.0018,
+      "step": 5480
+    },
+    {
+      "epoch": 2.1925304573596964,
+      "grad_norm": 0.3329474627971649,
+      "learning_rate": 9.290909090909092e-07,
+      "loss": 0.0018,
+      "step": 5490
+    },
+    {
+      "epoch": 2.196524865188736,
+      "grad_norm": 0.4752912223339081,
+      "learning_rate": 9.10909090909091e-07,
+      "loss": 0.0015,
+      "step": 5500
+    },
+    {
+      "epoch": 2.200519273017775,
+      "grad_norm": 0.10243528336286545,
+      "learning_rate": 8.927272727272727e-07,
+      "loss": 0.002,
+      "step": 5510
+    },
+    {
+      "epoch": 2.2045136808468144,
+      "grad_norm": 0.09885286539793015,
+      "learning_rate": 8.745454545454546e-07,
+      "loss": 0.0021,
+      "step": 5520
+    },
+    {
+      "epoch": 2.208508088675854,
+      "grad_norm": 0.08831353485584259,
+      "learning_rate": 8.563636363636364e-07,
+      "loss": 0.0013,
+      "step": 5530
+    },
+    {
+      "epoch": 2.212502496504893,
+      "grad_norm": 0.32018667459487915,
+      "learning_rate": 8.381818181818183e-07,
+      "loss": 0.002,
+      "step": 5540
+    },
+    {
+      "epoch": 2.2164969043339324,
+      "grad_norm": 0.17237108945846558,
+      "learning_rate": 8.200000000000001e-07,
+      "loss": 0.0018,
+      "step": 5550
+    },
+    {
+      "epoch": 2.220491312162972,
+      "grad_norm": 0.2744125723838806,
+      "learning_rate": 8.018181818181819e-07,
+      "loss": 0.001,
+      "step": 5560
+    },
+    {
+      "epoch": 2.2244857199920114,
+      "grad_norm": 0.134715735912323,
+      "learning_rate": 7.836363636363637e-07,
+      "loss": 0.001,
+      "step": 5570
+    },
+    {
+      "epoch": 2.2284801278210504,
+      "grad_norm": 0.40728604793548584,
+      "learning_rate": 7.654545454545455e-07,
+      "loss": 0.0011,
+      "step": 5580
+    },
+    {
+      "epoch": 2.23247453565009,
+      "grad_norm": 0.28443050384521484,
+      "learning_rate": 7.472727272727273e-07,
+      "loss": 0.0018,
+      "step": 5590
+    },
+    {
+      "epoch": 2.2364689434791294,
+      "grad_norm": 0.27793049812316895,
+      "learning_rate": 7.290909090909092e-07,
+      "loss": 0.0017,
+      "step": 5600
+    },
+    {
+      "epoch": 2.2404633513081684,
+      "grad_norm": 0.10264703631401062,
+      "learning_rate": 7.10909090909091e-07,
+      "loss": 0.0044,
+      "step": 5610
+    },
+    {
+      "epoch": 2.244457759137208,
+      "grad_norm": 0.12262337654829025,
+      "learning_rate": 6.927272727272729e-07,
+      "loss": 0.0011,
+      "step": 5620
+    },
+    {
+      "epoch": 2.2484521669662474,
+      "grad_norm": 0.17834536731243134,
+      "learning_rate": 6.745454545454547e-07,
+      "loss": 0.0012,
+      "step": 5630
+    },
+    {
+      "epoch": 2.2524465747952864,
+      "grad_norm": 0.25186392664909363,
+      "learning_rate": 6.563636363636363e-07,
+      "loss": 0.0032,
+      "step": 5640
+    },
+    {
+      "epoch": 2.256440982624326,
+      "grad_norm": 0.23524995148181915,
+      "learning_rate": 6.381818181818183e-07,
+      "loss": 0.0036,
+      "step": 5650
+    },
+    {
+      "epoch": 2.2604353904533654,
+      "grad_norm": 0.08354029059410095,
+      "learning_rate": 6.200000000000001e-07,
+      "loss": 0.0009,
+      "step": 5660
+    },
+    {
+      "epoch": 2.2644297982824044,
+      "grad_norm": 0.5130342841148376,
+      "learning_rate": 6.018181818181819e-07,
+      "loss": 0.0012,
+      "step": 5670
+    },
+    {
+      "epoch": 2.268424206111444,
+      "grad_norm": 0.17740464210510254,
+      "learning_rate": 5.836363636363637e-07,
+      "loss": 0.0045,
+      "step": 5680
+    },
+    {
+      "epoch": 2.2724186139404834,
+      "grad_norm": 0.07126139849424362,
+      "learning_rate": 5.654545454545455e-07,
+      "loss": 0.0017,
+      "step": 5690
+    },
+    {
+      "epoch": 2.276413021769523,
+      "grad_norm": 0.07142548263072968,
+      "learning_rate": 5.472727272727273e-07,
+      "loss": 0.0032,
+      "step": 5700
+    },
+    {
+      "epoch": 2.280407429598562,
+      "grad_norm": 0.1244952380657196,
+      "learning_rate": 5.290909090909092e-07,
+      "loss": 0.0011,
+      "step": 5710
+    },
+    {
+      "epoch": 2.2844018374276014,
+      "grad_norm": 0.6580970883369446,
+      "learning_rate": 5.109090909090909e-07,
+      "loss": 0.0015,
+      "step": 5720
+    },
+    {
+      "epoch": 2.288396245256641,
+      "grad_norm": 0.06932102888822556,
+      "learning_rate": 4.927272727272728e-07,
+      "loss": 0.0025,
+      "step": 5730
+    },
+    {
+      "epoch": 2.29239065308568,
+      "grad_norm": 1.0498089790344238,
+      "learning_rate": 4.745454545454546e-07,
+      "loss": 0.0037,
+      "step": 5740
+    },
+    {
+      "epoch": 2.2963850609147194,
+      "grad_norm": 0.13281786441802979,
+      "learning_rate": 4.563636363636364e-07,
+      "loss": 0.0035,
+      "step": 5750
+    },
+    {
+      "epoch": 2.300379468743759,
+      "grad_norm": 0.37985411286354065,
+      "learning_rate": 4.381818181818182e-07,
+      "loss": 0.0012,
+      "step": 5760
+    },
+    {
+      "epoch": 2.304373876572798,
+      "grad_norm": 0.20376817882061005,
+      "learning_rate": 4.2000000000000006e-07,
+      "loss": 0.0021,
+      "step": 5770
+    },
+    {
+      "epoch": 2.3083682844018374,
+      "grad_norm": 0.04566574841737747,
+      "learning_rate": 4.0181818181818184e-07,
+      "loss": 0.0018,
+      "step": 5780
+    },
+    {
+      "epoch": 2.312362692230877,
+      "grad_norm": 0.12633280456066132,
+      "learning_rate": 3.836363636363637e-07,
+      "loss": 0.002,
+      "step": 5790
+    },
+    {
+      "epoch": 2.316357100059916,
+      "grad_norm": 0.08685902506113052,
+      "learning_rate": 3.654545454545455e-07,
+      "loss": 0.0012,
+      "step": 5800
+    },
+    {
+      "epoch": 2.3203515078889554,
+      "grad_norm": 0.18134891986846924,
+      "learning_rate": 3.4727272727272725e-07,
+      "loss": 0.0022,
+      "step": 5810
+    },
+    {
+      "epoch": 2.324345915717995,
+      "grad_norm": 0.06675967574119568,
+      "learning_rate": 3.290909090909091e-07,
+      "loss": 0.0012,
+      "step": 5820
+    },
+    {
+      "epoch": 2.3283403235470344,
+      "grad_norm": 0.14449624717235565,
+      "learning_rate": 3.109090909090909e-07,
+      "loss": 0.0022,
+      "step": 5830
+    },
+    {
+      "epoch": 2.3323347313760734,
+      "grad_norm": 0.24351002275943756,
+      "learning_rate": 2.9272727272727276e-07,
+      "loss": 0.0021,
+      "step": 5840
+    },
+    {
+      "epoch": 2.336329139205113,
+      "grad_norm": 0.314222514629364,
+      "learning_rate": 2.7454545454545455e-07,
+      "loss": 0.0027,
+      "step": 5850
+    },
+    {
+      "epoch": 2.3403235470341524,
+      "grad_norm": 0.5329357981681824,
+      "learning_rate": 2.563636363636364e-07,
+      "loss": 0.0016,
+      "step": 5860
+    },
+    {
+      "epoch": 2.3443179548631914,
+      "grad_norm": 0.6777558922767639,
+      "learning_rate": 2.381818181818182e-07,
+      "loss": 0.0018,
+      "step": 5870
+    },
+    {
+      "epoch": 2.348312362692231,
+      "grad_norm": 0.5858981609344482,
+      "learning_rate": 2.2e-07,
+      "loss": 0.0015,
+      "step": 5880
+    },
+    {
+      "epoch": 2.3523067705212704,
+      "grad_norm": 0.29430916905403137,
+      "learning_rate": 2.0181818181818185e-07,
+      "loss": 0.0023,
+      "step": 5890
+    },
+    {
+      "epoch": 2.3563011783503094,
+      "grad_norm": 0.07274141162633896,
+      "learning_rate": 1.8363636363636363e-07,
+      "loss": 0.0015,
+      "step": 5900
+    },
+    {
+      "epoch": 2.360295586179349,
+      "grad_norm": 0.14015686511993408,
+      "learning_rate": 1.6545454545454545e-07,
+      "loss": 0.0016,
+      "step": 5910
+    },
+    {
+      "epoch": 2.3642899940083884,
+      "grad_norm": 0.13865520060062408,
+      "learning_rate": 1.4727272727272728e-07,
+      "loss": 0.0024,
+      "step": 5920
+    },
+    {
+      "epoch": 2.3682844018374274,
+      "grad_norm": 0.3300738036632538,
+      "learning_rate": 1.290909090909091e-07,
+      "loss": 0.004,
+      "step": 5930
+    },
+    {
+      "epoch": 2.372278809666467,
+      "grad_norm": 0.10867658257484436,
+      "learning_rate": 1.1090909090909091e-07,
+      "loss": 0.0016,
+      "step": 5940
+    },
+    {
+      "epoch": 2.3762732174955064,
+      "grad_norm": 0.06421338766813278,
+      "learning_rate": 9.272727272727273e-08,
+      "loss": 0.0019,
+      "step": 5950
+    },
+    {
+      "epoch": 2.380267625324546,
+      "grad_norm": 0.09091343730688095,
+      "learning_rate": 7.454545454545455e-08,
+      "loss": 0.0009,
+      "step": 5960
+    },
+    {
+      "epoch": 2.384262033153585,
+      "grad_norm": 0.06275226920843124,
+      "learning_rate": 5.636363636363637e-08,
+      "loss": 0.001,
+      "step": 5970
+    },
+    {
+      "epoch": 2.3882564409826244,
+      "grad_norm": 0.13375839591026306,
+      "learning_rate": 3.818181818181819e-08,
+      "loss": 0.0011,
+      "step": 5980
+    },
+    {
+      "epoch": 2.392250848811664,
+      "grad_norm": 0.1499566286802292,
+      "learning_rate": 2e-08,
+      "loss": 0.0011,
+      "step": 5990
+    },
+    {
+      "epoch": 2.396245256640703,
+      "grad_norm": 0.47475066781044006,
+      "learning_rate": 1.8181818181818182e-09,
+      "loss": 0.0017,
+      "step": 6000
+    },
+    {
+      "epoch": 2.396245256640703,
+      "eval_loss": 0.006691531278192997,
+      "eval_runtime": 7534.0645,
+      "eval_samples_per_second": 2.658,
+      "eval_steps_per_second": 0.332,
+      "eval_wer": 0.4967807301369415,
+      "step": 6000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 6000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.539223607902208e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}