diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,38606 @@
+{
+  "best_global_step": 50000,
+  "best_metric": 0.4009660835826073,
+  "best_model_checkpoint": "./whisper-translate-bn-or\\checkpoint-50000",
+  "epoch": 5.508839585315771,
+  "eval_steps": 10000,
+  "global_step": 55000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00010016527269995493,
+      "grad_norm": 1.0699678659439087,
+      "learning_rate": 0.0,
+      "loss": 3.4658,
+      "step": 1
+    },
+    {
+      "epoch": 0.0010016527269995493,
+      "grad_norm": 0.946382999420166,
+      "learning_rate": 4.5e-06,
+      "loss": 3.1873,
+      "step": 10
+    },
+    {
+      "epoch": 0.0020033054539990987,
+      "grad_norm": 1.0159204006195068,
+      "learning_rate": 9.5e-06,
+      "loss": 3.3418,
+      "step": 20
+    },
+    {
+      "epoch": 0.0030049581809986478,
+      "grad_norm": 0.963729202747345,
+      "learning_rate": 1.45e-05,
+      "loss": 3.2141,
+      "step": 30
+    },
+    {
+      "epoch": 0.004006610907998197,
+      "grad_norm": 1.0494544506072998,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 3.2094,
+      "step": 40
+    },
+    {
+      "epoch": 0.005008263634997746,
+      "grad_norm": 1.1693147420883179,
+      "learning_rate": 2.45e-05,
+      "loss": 3.1064,
+      "step": 50
+    },
+    {
+      "epoch": 0.0060099163619972956,
+      "grad_norm": 1.0876061916351318,
+      "learning_rate": 2.95e-05,
+      "loss": 2.8635,
+      "step": 60
+    },
+    {
+      "epoch": 0.007011569088996845,
+      "grad_norm": 1.2751100063323975,
+      "learning_rate": 3.45e-05,
+      "loss": 2.7471,
+      "step": 70
+    },
+    {
+      "epoch": 0.008013221815996395,
+      "grad_norm": 1.3546741008758545,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 2.6155,
+      "step": 80
+    },
+    {
+      "epoch": 0.009014874542995942,
+      "grad_norm": 1.6993346214294434,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 2.5273,
+      "step": 90
+    },
+    {
+      "epoch": 0.010016527269995492,
+      "grad_norm": 1.595080852508545,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 2.413,
+      "step": 100
+    },
+    {
+      "epoch": 0.011018179996995042,
+      "grad_norm": 1.627835750579834,
+      "learning_rate": 4.9999995973874405e-05,
+      "loss": 2.1941,
+      "step": 110
+    },
+    {
+      "epoch": 0.012019832723994591,
+      "grad_norm": 1.5041255950927734,
+      "learning_rate": 4.999998205640485e-05,
+      "loss": 2.1507,
+      "step": 120
+    },
+    {
+      "epoch": 0.01302148545099414,
+      "grad_norm": 1.970567226409912,
+      "learning_rate": 4.999995819789164e-05,
+      "loss": 2.0218,
+      "step": 130
+    },
+    {
+      "epoch": 0.01402313817799369,
+      "grad_norm": 1.9568697214126587,
+      "learning_rate": 4.999992439834424e-05,
+      "loss": 1.6866,
+      "step": 140
+    },
+    {
+      "epoch": 0.015024790904993238,
+      "grad_norm": 2.1707658767700195,
+      "learning_rate": 4.99998806577761e-05,
+      "loss": 1.6941,
+      "step": 150
+    },
+    {
+      "epoch": 0.01602644363199279,
+      "grad_norm": 2.4342153072357178,
+      "learning_rate": 4.999982697620461e-05,
+      "loss": 1.7162,
+      "step": 160
+    },
+    {
+      "epoch": 0.01702809635899234,
+      "grad_norm": 2.1355204582214355,
+      "learning_rate": 4.9999763353651117e-05,
+      "loss": 1.6787,
+      "step": 170
+    },
+    {
+      "epoch": 0.018029749085991885,
+      "grad_norm": 2.1012508869171143,
+      "learning_rate": 4.999968979014093e-05,
+      "loss": 1.5662,
+      "step": 180
+    },
+    {
+      "epoch": 0.019031401812991434,
+      "grad_norm": 1.8109233379364014,
+      "learning_rate": 4.999960628570328e-05,
+      "loss": 1.5796,
+      "step": 190
+    },
+    {
+      "epoch": 0.020033054539990984,
+      "grad_norm": 2.47389554977417,
+      "learning_rate": 4.9999512840371386e-05,
+      "loss": 1.6256,
+      "step": 200
+    },
+    {
+      "epoch": 0.021034707266990534,
+      "grad_norm": 1.9195002317428589,
+      "learning_rate": 4.999940945418241e-05,
+      "loss": 1.5094,
+      "step": 210
+    },
+    {
+      "epoch": 0.022036359993990083,
+      "grad_norm": 2.2418041229248047,
+      "learning_rate": 4.9999296127177453e-05,
+      "loss": 1.5252,
+      "step": 220
+    },
+    {
+      "epoch": 0.023038012720989633,
+      "grad_norm": 2.5955166816711426,
+      "learning_rate": 4.999917285940158e-05,
+      "loss": 1.5686,
+      "step": 230
+    },
+    {
+      "epoch": 0.024039665447989182,
+      "grad_norm": 1.936550498008728,
+      "learning_rate": 4.999903965090381e-05,
+      "loss": 1.5285,
+      "step": 240
+    },
+    {
+      "epoch": 0.025041318174988732,
+      "grad_norm": 1.894722819328308,
+      "learning_rate": 4.9998896501737116e-05,
+      "loss": 1.5413,
+      "step": 250
+    },
+    {
+      "epoch": 0.02604297090198828,
+      "grad_norm": 2.192305326461792,
+      "learning_rate": 4.99987434119584e-05,
+      "loss": 1.5218,
+      "step": 260
+    },
+    {
+      "epoch": 0.02704462362898783,
+      "grad_norm": 2.490818500518799,
+      "learning_rate": 4.999858038162857e-05,
+      "loss": 1.4895,
+      "step": 270
+    },
+    {
+      "epoch": 0.02804627635598738,
+      "grad_norm": 2.2005741596221924,
+      "learning_rate": 4.999840741081243e-05,
+      "loss": 1.481,
+      "step": 280
+    },
+    {
+      "epoch": 0.02904792908298693,
+      "grad_norm": 2.2879180908203125,
+      "learning_rate": 4.999822449957876e-05,
+      "loss": 1.3541,
+      "step": 290
+    },
+    {
+      "epoch": 0.030049581809986476,
+      "grad_norm": 1.9570573568344116,
+      "learning_rate": 4.999803164800031e-05,
+      "loss": 1.4484,
+      "step": 300
+    },
+    {
+      "epoch": 0.031051234536986026,
+      "grad_norm": 2.203392744064331,
+      "learning_rate": 4.999782885615375e-05,
+      "loss": 1.4385,
+      "step": 310
+    },
+    {
+      "epoch": 0.03205288726398558,
+      "grad_norm": 2.16286563873291,
+      "learning_rate": 4.999761612411972e-05,
+      "loss": 1.4522,
+      "step": 320
+    },
+    {
+      "epoch": 0.03305453999098513,
+      "grad_norm": 2.594715118408203,
+      "learning_rate": 4.999739345198282e-05,
+      "loss": 1.4169,
+      "step": 330
+    },
+    {
+      "epoch": 0.03405619271798468,
+      "grad_norm": 2.793673515319824,
+      "learning_rate": 4.999716083983159e-05,
+      "loss": 1.525,
+      "step": 340
+    },
+    {
+      "epoch": 0.03505784544498423,
+      "grad_norm": 2.3903636932373047,
+      "learning_rate": 4.999691828775853e-05,
+      "loss": 1.4475,
+      "step": 350
+    },
+    {
+      "epoch": 0.03605949817198377,
+      "grad_norm": 2.7492942810058594,
+      "learning_rate": 4.999666579586008e-05,
+      "loss": 1.397,
+      "step": 360
+    },
+    {
+      "epoch": 0.03706115089898332,
+      "grad_norm": 2.458517551422119,
+      "learning_rate": 4.9996403364236645e-05,
+      "loss": 1.3987,
+      "step": 370
+    },
+    {
+      "epoch": 0.03806280362598287,
+      "grad_norm": 2.7841598987579346,
+      "learning_rate": 4.999613099299258e-05,
+      "loss": 1.3781,
+      "step": 380
+    },
+    {
+      "epoch": 0.03906445635298242,
+      "grad_norm": 2.5349690914154053,
+      "learning_rate": 4.9995848682236194e-05,
+      "loss": 1.3281,
+      "step": 390
+    },
+    {
+      "epoch": 0.04006610907998197,
+      "grad_norm": 2.858383893966675,
+      "learning_rate": 4.999555643207974e-05,
+      "loss": 1.3173,
+      "step": 400
+    },
+    {
+      "epoch": 0.04106776180698152,
+      "grad_norm": 2.612161874771118,
+      "learning_rate": 4.999525424263943e-05,
+      "loss": 1.4123,
+      "step": 410
+    },
+    {
+      "epoch": 0.04206941453398107,
+      "grad_norm": 2.9477760791778564,
+      "learning_rate": 4.9994942114035435e-05,
+      "loss": 1.3716,
+      "step": 420
+    },
+    {
+      "epoch": 0.04307106726098062,
+      "grad_norm": 2.3722777366638184,
+      "learning_rate": 4.999462004639186e-05,
+      "loss": 1.3067,
+      "step": 430
+    },
+    {
+      "epoch": 0.044072719987980166,
+      "grad_norm": 2.594200611114502,
+      "learning_rate": 4.999428803983678e-05,
+      "loss": 1.3702,
+      "step": 440
+    },
+    {
+      "epoch": 0.045074372714979716,
+      "grad_norm": 2.670445203781128,
+      "learning_rate": 4.99939460945022e-05,
+      "loss": 1.3728,
+      "step": 450
+    },
+    {
+      "epoch": 0.046076025441979265,
+      "grad_norm": 2.574885606765747,
+      "learning_rate": 4.9993594210524115e-05,
+      "loss": 1.351,
+      "step": 460
+    },
+    {
+      "epoch": 0.047077678168978815,
+      "grad_norm": 2.1980414390563965,
+      "learning_rate": 4.999323238804243e-05,
+      "loss": 1.425,
+      "step": 470
+    },
+    {
+      "epoch": 0.048079330895978364,
+      "grad_norm": 2.689714193344116,
+      "learning_rate": 4.9992860627201034e-05,
+      "loss": 1.3985,
+      "step": 480
+    },
+    {
+      "epoch": 0.049080983622977914,
+      "grad_norm": 2.989649772644043,
+      "learning_rate": 4.999247892814775e-05,
+      "loss": 1.3723,
+      "step": 490
+    },
+    {
+      "epoch": 0.050082636349977463,
+      "grad_norm": 2.6088790893554688,
+      "learning_rate": 4.999208729103435e-05,
+      "loss": 1.2535,
+      "step": 500
+    },
+    {
+      "epoch": 0.05108428907697701,
+      "grad_norm": 2.829699754714966,
+      "learning_rate": 4.999168571601658e-05,
+      "loss": 1.2934,
+      "step": 510
+    },
+    {
+      "epoch": 0.05208594180397656,
+      "grad_norm": 1.9769593477249146,
+      "learning_rate": 4.999127420325411e-05,
+      "loss": 1.1993,
+      "step": 520
+    },
+    {
+      "epoch": 0.05308759453097611,
+      "grad_norm": 2.4946117401123047,
+      "learning_rate": 4.9990852752910576e-05,
+      "loss": 1.2418,
+      "step": 530
+    },
+    {
+      "epoch": 0.05408924725797566,
+      "grad_norm": 2.7048685550689697,
+      "learning_rate": 4.999042136515358e-05,
+      "loss": 1.2653,
+      "step": 540
+    },
+    {
+      "epoch": 0.05509089998497521,
+      "grad_norm": 2.877206802368164,
+      "learning_rate": 4.998998004015464e-05,
+      "loss": 1.2766,
+      "step": 550
+    },
+    {
+      "epoch": 0.05609255271197476,
+      "grad_norm": 2.246899366378784,
+      "learning_rate": 4.998952877808925e-05,
+      "loss": 1.3626,
+      "step": 560
+    },
+    {
+      "epoch": 0.05709420543897431,
+      "grad_norm": 2.2098586559295654,
+      "learning_rate": 4.998906757913686e-05,
+      "loss": 1.2818,
+      "step": 570
+    },
+    {
+      "epoch": 0.05809585816597386,
+      "grad_norm": 3.2610485553741455,
+      "learning_rate": 4.998859644348085e-05,
+      "loss": 1.2254,
+      "step": 580
+    },
+    {
+      "epoch": 0.05909751089297341,
+      "grad_norm": 3.1614911556243896,
+      "learning_rate": 4.998811537130857e-05,
+      "loss": 1.2474,
+      "step": 590
+    },
+    {
+      "epoch": 0.06009916361997295,
+      "grad_norm": 2.5031402111053467,
+      "learning_rate": 4.9987624362811324e-05,
+      "loss": 1.1746,
+      "step": 600
+    },
+    {
+      "epoch": 0.0611008163469725,
+      "grad_norm": 2.7153284549713135,
+      "learning_rate": 4.9987123418184344e-05,
+      "loss": 1.285,
+      "step": 610
+    },
+    {
+      "epoch": 0.06210246907397205,
+      "grad_norm": 2.6688461303710938,
+      "learning_rate": 4.998661253762683e-05,
+      "loss": 1.2516,
+      "step": 620
+    },
+    {
+      "epoch": 0.06310412180097161,
+      "grad_norm": 2.5687406063079834,
+      "learning_rate": 4.998609172134192e-05,
+      "loss": 1.1905,
+      "step": 630
+    },
+    {
+      "epoch": 0.06410577452797116,
+      "grad_norm": 2.876481533050537,
+      "learning_rate": 4.9985560969536736e-05,
+      "loss": 1.192,
+      "step": 640
+    },
+    {
+      "epoch": 0.0651074272549707,
+      "grad_norm": 3.5918831825256348,
+      "learning_rate": 4.9985020282422304e-05,
+      "loss": 1.3036,
+      "step": 650
+    },
+    {
+      "epoch": 0.06610907998197026,
+      "grad_norm": 2.464890718460083,
+      "learning_rate": 4.998446966021364e-05,
+      "loss": 1.1751,
+      "step": 660
+    },
+    {
+      "epoch": 0.0671107327089698,
+      "grad_norm": 2.5518927574157715,
+      "learning_rate": 4.998390910312969e-05,
+      "loss": 1.2133,
+      "step": 670
+    },
+    {
+      "epoch": 0.06811238543596936,
+      "grad_norm": 2.709120512008667,
+      "learning_rate": 4.998333861139335e-05,
+      "loss": 1.2233,
+      "step": 680
+    },
+    {
+      "epoch": 0.0691140381629689,
+      "grad_norm": 2.6371023654937744,
+      "learning_rate": 4.998275818523147e-05,
+      "loss": 1.1879,
+      "step": 690
+    },
+    {
+      "epoch": 0.07011569088996845,
+      "grad_norm": 2.7895636558532715,
+      "learning_rate": 4.998216782487486e-05,
+      "loss": 1.1956,
+      "step": 700
+    },
+    {
+      "epoch": 0.071117343616968,
+      "grad_norm": 2.9015183448791504,
+      "learning_rate": 4.998156753055826e-05,
+      "loss": 1.1716,
+      "step": 710
+    },
+    {
+      "epoch": 0.07211899634396754,
+      "grad_norm": 3.0008327960968018,
+      "learning_rate": 4.9980957302520396e-05,
+      "loss": 1.2089,
+      "step": 720
+    },
+    {
+      "epoch": 0.07312064907096709,
+      "grad_norm": 2.6571829319000244,
+      "learning_rate": 4.9980337141003895e-05,
+      "loss": 1.1451,
+      "step": 730
+    },
+    {
+      "epoch": 0.07412230179796664,
+      "grad_norm": 2.932692766189575,
+      "learning_rate": 4.9979707046255376e-05,
+      "loss": 1.1582,
+      "step": 740
+    },
+    {
+      "epoch": 0.07512395452496619,
+      "grad_norm": 2.8995957374572754,
+      "learning_rate": 4.9979067018525383e-05,
+      "loss": 1.1619,
+      "step": 750
+    },
+    {
+      "epoch": 0.07612560725196574,
+      "grad_norm": 2.709153890609741,
+      "learning_rate": 4.9978417058068414e-05,
+      "loss": 1.202,
+      "step": 760
+    },
+    {
+      "epoch": 0.07712725997896529,
+      "grad_norm": 2.4261250495910645,
+      "learning_rate": 4.997775716514293e-05,
+      "loss": 1.2184,
+      "step": 770
+    },
+    {
+      "epoch": 0.07812891270596484,
+      "grad_norm": 2.8053817749023438,
+      "learning_rate": 4.997708734001133e-05,
+      "loss": 1.2656,
+      "step": 780
+    },
+    {
+      "epoch": 0.07913056543296439,
+      "grad_norm": 2.360339879989624,
+      "learning_rate": 4.997640758293996e-05,
+      "loss": 1.1717,
+      "step": 790
+    },
+    {
+      "epoch": 0.08013221815996394,
+      "grad_norm": 2.579951524734497,
+      "learning_rate": 4.9975717894199126e-05,
+      "loss": 1.2521,
+      "step": 800
+    },
+    {
+      "epoch": 0.08113387088696349,
+      "grad_norm": 2.9729485511779785,
+      "learning_rate": 4.997501827406307e-05,
+      "loss": 1.1597,
+      "step": 810
+    },
+    {
+      "epoch": 0.08213552361396304,
+      "grad_norm": 3.1533076763153076,
+      "learning_rate": 4.997430872280999e-05,
+      "loss": 1.1519,
+      "step": 820
+    },
+    {
+      "epoch": 0.08313717634096258,
+      "grad_norm": 4.105657577514648,
+      "learning_rate": 4.997358924072205e-05,
+      "loss": 1.1856,
+      "step": 830
+    },
+    {
+      "epoch": 0.08413882906796213,
+      "grad_norm": 2.565356731414795,
+      "learning_rate": 4.997285982808533e-05,
+      "loss": 1.232,
+      "step": 840
+    },
+    {
+      "epoch": 0.08514048179496168,
+      "grad_norm": 3.074532985687256,
+      "learning_rate": 4.9972120485189874e-05,
+      "loss": 1.1723,
+      "step": 850
+    },
+    {
+      "epoch": 0.08614213452196123,
+      "grad_norm": 2.80393123626709,
+      "learning_rate": 4.997137121232969e-05,
+      "loss": 1.2643,
+      "step": 860
+    },
+    {
+      "epoch": 0.08714378724896078,
+      "grad_norm": 2.783311128616333,
+      "learning_rate": 4.99706120098027e-05,
+      "loss": 1.1941,
+      "step": 870
+    },
+    {
+      "epoch": 0.08814543997596033,
+      "grad_norm": 2.6996560096740723,
+      "learning_rate": 4.9969842877910814e-05,
+      "loss": 1.18,
+      "step": 880
+    },
+    {
+      "epoch": 0.08914709270295988,
+      "grad_norm": 3.0572733879089355,
+      "learning_rate": 4.996906381695986e-05,
+      "loss": 1.1936,
+      "step": 890
+    },
+    {
+      "epoch": 0.09014874542995943,
+      "grad_norm": 2.5547292232513428,
+      "learning_rate": 4.996827482725963e-05,
+      "loss": 1.2188,
+      "step": 900
+    },
+    {
+      "epoch": 0.09115039815695898,
+      "grad_norm": 2.564838171005249,
+      "learning_rate": 4.996747590912386e-05,
+      "loss": 1.1149,
+      "step": 910
+    },
+    {
+      "epoch": 0.09215205088395853,
+      "grad_norm": 2.6100144386291504,
+      "learning_rate": 4.9966667062870235e-05,
+      "loss": 1.1286,
+      "step": 920
+    },
+    {
+      "epoch": 0.09315370361095808,
+      "grad_norm": 2.742168664932251,
+      "learning_rate": 4.996584828882038e-05,
+      "loss": 1.1574,
+      "step": 930
+    },
+    {
+      "epoch": 0.09415535633795763,
+      "grad_norm": 2.3155078887939453,
+      "learning_rate": 4.9965019587299874e-05,
+      "loss": 1.0794,
+      "step": 940
+    },
+    {
+      "epoch": 0.09515700906495718,
+      "grad_norm": 3.218953847885132,
+      "learning_rate": 4.996418095863824e-05,
+      "loss": 1.1358,
+      "step": 950
+    },
+    {
+      "epoch": 0.09615866179195673,
+      "grad_norm": 2.9602880477905273,
+      "learning_rate": 4.996333240316897e-05,
+      "loss": 1.1506,
+      "step": 960
+    },
+    {
+      "epoch": 0.09716031451895628,
+      "grad_norm": 2.4367001056671143,
+      "learning_rate": 4.996247392122948e-05,
+      "loss": 1.1119,
+      "step": 970
+    },
+    {
+      "epoch": 0.09816196724595583,
+      "grad_norm": 2.4974889755249023,
+      "learning_rate": 4.996160551316112e-05,
+      "loss": 1.1244,
+      "step": 980
+    },
+    {
+      "epoch": 0.09916361997295538,
+      "grad_norm": 3.0747549533843994,
+      "learning_rate": 4.9960727179309216e-05,
+      "loss": 1.1237,
+      "step": 990
+    },
+    {
+      "epoch": 0.10016527269995493,
+      "grad_norm": 3.293057441711426,
+      "learning_rate": 4.9959838920023036e-05,
+      "loss": 1.1125,
+      "step": 1000
+    },
+    {
+      "epoch": 0.10116692542695448,
+      "grad_norm": 3.2793030738830566,
+      "learning_rate": 4.995894073565579e-05,
+      "loss": 1.1349,
+      "step": 1010
+    },
+    {
+      "epoch": 0.10216857815395403,
+      "grad_norm": 2.5710489749908447,
+      "learning_rate": 4.995803262656463e-05,
+      "loss": 1.1671,
+      "step": 1020
+    },
+    {
+      "epoch": 0.10317023088095358,
+      "grad_norm": 2.81728458404541,
+      "learning_rate": 4.995711459311065e-05,
+      "loss": 1.098,
+      "step": 1030
+    },
+    {
+      "epoch": 0.10417188360795313,
+      "grad_norm": 2.4683609008789062,
+      "learning_rate": 4.995618663565892e-05,
+      "loss": 1.0697,
+      "step": 1040
+    },
+    {
+      "epoch": 0.10517353633495267,
+      "grad_norm": 2.7064619064331055,
+      "learning_rate": 4.995524875457841e-05,
+      "loss": 1.1379,
+      "step": 1050
+    },
+    {
+      "epoch": 0.10617518906195222,
+      "grad_norm": 2.810743570327759,
+      "learning_rate": 4.995430095024207e-05,
+      "loss": 1.1032,
+      "step": 1060
+    },
+    {
+      "epoch": 0.10717684178895177,
+      "grad_norm": 2.8417282104492188,
+      "learning_rate": 4.9953343223026796e-05,
+      "loss": 1.1375,
+      "step": 1070
+    },
+    {
+      "epoch": 0.10817849451595132,
+      "grad_norm": 2.6498005390167236,
+      "learning_rate": 4.9952375573313415e-05,
+      "loss": 1.0805,
+      "step": 1080
+    },
+    {
+      "epoch": 0.10918014724295087,
+      "grad_norm": 2.90468692779541,
+      "learning_rate": 4.9951398001486704e-05,
+      "loss": 1.1367,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11018179996995042,
+      "grad_norm": 2.9180171489715576,
+      "learning_rate": 4.995041050793538e-05,
+      "loss": 1.0368,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11118345269694997,
+      "grad_norm": 2.695122480392456,
+      "learning_rate": 4.994941309305212e-05,
+      "loss": 1.1418,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11218510542394952,
+      "grad_norm": 2.9390275478363037,
+      "learning_rate": 4.994840575723354e-05,
+      "loss": 1.1555,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11318675815094907,
+      "grad_norm": 3.146312713623047,
+      "learning_rate": 4.99473885008802e-05,
+      "loss": 1.105,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11418841087794862,
+      "grad_norm": 2.8160715103149414,
+      "learning_rate": 4.99463613243966e-05,
+      "loss": 1.0991,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11519006360494817,
+      "grad_norm": 2.578988790512085,
+      "learning_rate": 4.994532422819117e-05,
+      "loss": 1.0957,
+      "step": 1150
+    },
+    {
+      "epoch": 0.11619171633194772,
+      "grad_norm": 2.783162832260132,
+      "learning_rate": 4.994427721267634e-05,
+      "loss": 1.0774,
+      "step": 1160
+    },
+    {
+      "epoch": 0.11719336905894727,
+      "grad_norm": 2.6223232746124268,
+      "learning_rate": 4.9943220278268424e-05,
+      "loss": 1.1457,
+      "step": 1170
+    },
+    {
+      "epoch": 0.11819502178594682,
+      "grad_norm": 3.61311411857605,
+      "learning_rate": 4.994215342538771e-05,
+      "loss": 1.1417,
+      "step": 1180
+    },
+    {
+      "epoch": 0.11919667451294637,
+      "grad_norm": 2.6800315380096436,
+      "learning_rate": 4.994107665445841e-05,
+      "loss": 1.1383,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1201983272399459,
+      "grad_norm": 2.0521180629730225,
+      "learning_rate": 4.993998996590872e-05,
+      "loss": 0.992,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12119997996694545,
+      "grad_norm": 3.1667087078094482,
+      "learning_rate": 4.9938893360170726e-05,
+      "loss": 1.0216,
+      "step": 1210
+    },
+    {
+      "epoch": 0.122201632693945,
+      "grad_norm": 3.2581300735473633,
+      "learning_rate": 4.99377868376805e-05,
+      "loss": 1.0763,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12320328542094455,
+      "grad_norm": 2.8601009845733643,
+      "learning_rate": 4.993667039887804e-05,
+      "loss": 1.1149,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1242049381479441,
+      "grad_norm": 2.900925636291504,
+      "learning_rate": 4.99355440442073e-05,
+      "loss": 1.0522,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12520659087494365,
+      "grad_norm": 2.9990346431732178,
+      "learning_rate": 4.993440777411613e-05,
+      "loss": 1.0632,
+      "step": 1250
+    },
+    {
+      "epoch": 0.12620824360194322,
+      "grad_norm": 2.717073917388916,
+      "learning_rate": 4.99332615890564e-05,
+      "loss": 1.0612,
+      "step": 1260
+    },
+    {
+      "epoch": 0.12720989632894275,
+      "grad_norm": 3.6020913124084473,
+      "learning_rate": 4.993210548948385e-05,
+      "loss": 0.9953,
+      "step": 1270
+    },
+    {
+      "epoch": 0.12821154905594231,
+      "grad_norm": 2.9735605716705322,
+      "learning_rate": 4.993093947585822e-05,
+      "loss": 1.1802,
+      "step": 1280
+    },
+    {
+      "epoch": 0.12921320178294185,
+      "grad_norm": 2.907518148422241,
+      "learning_rate": 4.9929763548643145e-05,
+      "loss": 1.1365,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1302148545099414,
+      "grad_norm": 2.708824396133423,
+      "learning_rate": 4.9928577708306235e-05,
+      "loss": 1.0376,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13121650723694095,
+      "grad_norm": 3.0437746047973633,
+      "learning_rate": 4.992738195531902e-05,
+      "loss": 1.1522,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1322181599639405,
+      "grad_norm": 3.0373198986053467,
+      "learning_rate": 4.9926176290157e-05,
+      "loss": 1.0515,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13321981269094005,
+      "grad_norm": 2.457960367202759,
+      "learning_rate": 4.9924960713299575e-05,
+      "loss": 1.0844,
+      "step": 1330
+    },
+    {
+      "epoch": 0.1342214654179396,
+      "grad_norm": 3.0508358478546143,
+      "learning_rate": 4.9923735225230127e-05,
+      "loss": 1.0746,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13522311814493915,
+      "grad_norm": 2.9356350898742676,
+      "learning_rate": 4.992249982643595e-05,
+      "loss": 1.1505,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1362247708719387,
+      "grad_norm": 3.3241400718688965,
+      "learning_rate": 4.99212545174083e-05,
+      "loss": 1.1369,
+      "step": 1360
+    },
+    {
+      "epoch": 0.13722642359893825,
+      "grad_norm": 3.232041120529175,
+      "learning_rate": 4.9919999298642364e-05,
+      "loss": 0.9972,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1382280763259378,
+      "grad_norm": 2.7437665462493896,
+      "learning_rate": 4.991873417063726e-05,
+      "loss": 1.0838,
+      "step": 1380
+    },
+    {
+      "epoch": 0.13922972905293735,
+      "grad_norm": 2.6859841346740723,
+      "learning_rate": 4.991745913389606e-05,
+      "loss": 1.0474,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1402313817799369,
+      "grad_norm": 2.495213508605957,
+      "learning_rate": 4.9916174188925776e-05,
+      "loss": 1.0574,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14123303450693644,
+      "grad_norm": 2.9001944065093994,
+      "learning_rate": 4.991487933623736e-05,
+      "loss": 0.9758,
+      "step": 1410
+    },
+    {
+      "epoch": 0.142234687233936,
+      "grad_norm": 2.9604761600494385,
+      "learning_rate": 4.991357457634569e-05,
+      "loss": 1.0665,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14323633996093554,
+      "grad_norm": 2.8431453704833984,
+      "learning_rate": 4.9912259909769595e-05,
+      "loss": 1.1466,
+      "step": 1430
+    },
+    {
+      "epoch": 0.14423799268793508,
+      "grad_norm": 3.013521909713745,
+      "learning_rate": 4.991093533703184e-05,
+      "loss": 1.0455,
+      "step": 1440
+    },
+    {
+      "epoch": 0.14523964541493464,
+      "grad_norm": 3.180718421936035,
+      "learning_rate": 4.9909600858659146e-05,
+      "loss": 1.1061,
+      "step": 1450
+    },
+    {
+      "epoch": 0.14624129814193418,
+      "grad_norm": 2.9873087406158447,
+      "learning_rate": 4.9908256475182133e-05,
+      "loss": 0.9961,
+      "step": 1460
+    },
+    {
+      "epoch": 0.14724295086893374,
+      "grad_norm": 2.9176833629608154,
+      "learning_rate": 4.990690218713541e-05,
+      "loss": 1.0316,
+      "step": 1470
+    },
+    {
+      "epoch": 0.14824460359593328,
+      "grad_norm": 2.420295476913452,
+      "learning_rate": 4.990553799505748e-05,
+      "loss": 1.03,
+      "step": 1480
+    },
+    {
+      "epoch": 0.14924625632293284,
+      "grad_norm": 2.4549472332000732,
+      "learning_rate": 4.990416389949081e-05,
+      "loss": 1.0582,
+      "step": 1490
+    },
+    {
+      "epoch": 0.15024790904993238,
+      "grad_norm": 3.2612345218658447,
+      "learning_rate": 4.99027799009818e-05,
+      "loss": 1.0485,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15124956177693194,
+      "grad_norm": 2.7688887119293213,
+      "learning_rate": 4.9901386000080785e-05,
+      "loss": 1.0013,
+      "step": 1510
+    },
+    {
+      "epoch": 0.15225121450393148,
+      "grad_norm": 2.7964026927948,
+      "learning_rate": 4.989998219734203e-05,
+      "loss": 0.9652,
+      "step": 1520
+    },
+    {
+      "epoch": 0.15325286723093104,
+      "grad_norm": 2.564159393310547,
+      "learning_rate": 4.989856849332376e-05,
+      "loss": 0.9539,
+      "step": 1530
+    },
+    {
+      "epoch": 0.15425451995793057,
+      "grad_norm": 2.9798643589019775,
+      "learning_rate": 4.989714488858812e-05,
+      "loss": 1.0204,
+      "step": 1540
+    },
+    {
+      "epoch": 0.15525617268493014,
+      "grad_norm": 3.4313089847564697,
+      "learning_rate": 4.989571138370118e-05,
+      "loss": 0.9848,
+      "step": 1550
+    },
+    {
+      "epoch": 0.15625782541192967,
+      "grad_norm": 3.4639475345611572,
+      "learning_rate": 4.989426797923299e-05,
+      "loss": 1.0276,
+      "step": 1560
+    },
+    {
+      "epoch": 0.15725947813892924,
+      "grad_norm": 2.630809783935547,
+      "learning_rate": 4.989281467575748e-05,
+      "loss": 1.0028,
+      "step": 1570
+    },
+    {
+      "epoch": 0.15826113086592877,
+      "grad_norm": 3.15830397605896,
+      "learning_rate": 4.989135147385255e-05,
+      "loss": 0.9953,
+      "step": 1580
+    },
+    {
+      "epoch": 0.15926278359292834,
+      "grad_norm": 2.3013908863067627,
+      "learning_rate": 4.9889878374100054e-05,
+      "loss": 0.9445,
+      "step": 1590
+    },
+    {
+      "epoch": 0.16026443631992787,
+      "grad_norm": 2.9990906715393066,
+      "learning_rate": 4.9888395377085734e-05,
+      "loss": 1.0629,
+      "step": 1600
+    },
+    {
+      "epoch": 0.16126608904692744,
+      "grad_norm": 2.9761180877685547,
+      "learning_rate": 4.98869024833993e-05,
+      "loss": 1.0075,
+      "step": 1610
+    },
+    {
+      "epoch": 0.16226774177392697,
+      "grad_norm": 2.7043395042419434,
+      "learning_rate": 4.9885399693634385e-05,
+      "loss": 1.0862,
+      "step": 1620
+    },
+    {
+      "epoch": 0.16326939450092653,
+      "grad_norm": 2.7392871379852295,
+      "learning_rate": 4.988388700838856e-05,
+      "loss": 1.0467,
+      "step": 1630
+    },
+    {
+      "epoch": 0.16427104722792607,
+      "grad_norm": 2.6225688457489014,
+      "learning_rate": 4.988236442826335e-05,
+      "loss": 1.0083,
+      "step": 1640
+    },
+    {
+      "epoch": 0.16527269995492563,
+      "grad_norm": 2.732118606567383,
+      "learning_rate": 4.988083195386418e-05,
+      "loss": 1.0408,
+      "step": 1650
+    },
+    {
+      "epoch": 0.16627435268192517,
+      "grad_norm": 3.136059284210205,
+      "learning_rate": 4.987928958580043e-05,
+      "loss": 1.0722,
+      "step": 1660
+    },
+    {
+      "epoch": 0.16727600540892473,
+      "grad_norm": 3.0781443119049072,
+      "learning_rate": 4.987773732468541e-05,
+      "loss": 1.0305,
+      "step": 1670
+    },
+    {
+      "epoch": 0.16827765813592427,
+      "grad_norm": 2.469252824783325,
+      "learning_rate": 4.9876175171136366e-05,
+      "loss": 1.0357,
+      "step": 1680
+    },
+    {
+      "epoch": 0.16927931086292383,
+      "grad_norm": 2.5065886974334717,
+      "learning_rate": 4.9874603125774465e-05,
+      "loss": 0.9225,
+      "step": 1690
+    },
+    {
+      "epoch": 0.17028096358992337,
+      "grad_norm": 2.7271981239318848,
+      "learning_rate": 4.987302118922484e-05,
+      "loss": 0.9579,
+      "step": 1700
+    },
+    {
+      "epoch": 0.17128261631692293,
+      "grad_norm": 3.1236214637756348,
+      "learning_rate": 4.9871429362116517e-05,
+      "loss": 0.9902,
+      "step": 1710
+    },
+    {
+      "epoch": 0.17228426904392247,
+      "grad_norm": 2.5821967124938965,
+      "learning_rate": 4.986982764508248e-05,
+      "loss": 0.9573,
+      "step": 1720
+    },
+    {
+      "epoch": 0.17328592177092203,
+      "grad_norm": 2.606112480163574,
+      "learning_rate": 4.986821603875964e-05,
+      "loss": 1.0657,
+      "step": 1730
+    },
+    {
+      "epoch": 0.17428757449792157,
+      "grad_norm": 2.6190571784973145,
+      "learning_rate": 4.986659454378885e-05,
+      "loss": 1.0047,
+      "step": 1740
+    },
+    {
+      "epoch": 0.17528922722492113,
+      "grad_norm": 2.8123345375061035,
+      "learning_rate": 4.986496316081486e-05,
+      "loss": 0.994,
+      "step": 1750
+    },
+    {
+      "epoch": 0.17629087995192066,
+      "grad_norm": 2.665658712387085,
+      "learning_rate": 4.9863321890486386e-05,
+      "loss": 0.956,
+      "step": 1760
+    },
+    {
+      "epoch": 0.17729253267892023,
+      "grad_norm": 2.5470478534698486,
+      "learning_rate": 4.986167073345608e-05,
+      "loss": 1.0184,
+      "step": 1770
+    },
+    {
+      "epoch": 0.17829418540591976,
+      "grad_norm": 3.3314783573150635,
+      "learning_rate": 4.986000969038049e-05,
+      "loss": 0.9555,
+      "step": 1780
+    },
+    {
+      "epoch": 0.17929583813291933,
+      "grad_norm": 3.4173240661621094,
+      "learning_rate": 4.985833876192013e-05,
+      "loss": 0.9164,
+      "step": 1790
+    },
+    {
+      "epoch": 0.18029749085991886,
+      "grad_norm": 3.0542538166046143,
+      "learning_rate": 4.985665794873944e-05,
+      "loss": 0.992,
+      "step": 1800
+    },
+    {
+      "epoch": 0.18129914358691843,
+      "grad_norm": 3.4689574241638184,
+      "learning_rate": 4.9854967251506755e-05,
+      "loss": 1.075,
+      "step": 1810
+    },
+    {
+      "epoch": 0.18230079631391796,
+      "grad_norm": 2.872692346572876,
+      "learning_rate": 4.98532666708944e-05,
+      "loss": 1.0528,
+      "step": 1820
+    },
+    {
+      "epoch": 0.18330244904091753,
+      "grad_norm": 2.7932639122009277,
+      "learning_rate": 4.9851556207578565e-05,
+      "loss": 0.9843,
+      "step": 1830
+    },
+    {
+      "epoch": 0.18430410176791706,
+      "grad_norm": 2.615530014038086,
+      "learning_rate": 4.9849835862239426e-05,
+      "loss": 0.8914,
+      "step": 1840
+    },
+    {
+      "epoch": 0.18530575449491662,
+      "grad_norm": 2.508965253829956,
+      "learning_rate": 4.984810563556106e-05,
+      "loss": 1.0039,
+      "step": 1850
+    },
+    {
+      "epoch": 0.18630740722191616,
+      "grad_norm": 2.9795517921447754,
+      "learning_rate": 4.9846365528231466e-05,
+      "loss": 0.9747,
+      "step": 1860
+    },
+    {
+      "epoch": 0.18730905994891572,
+      "grad_norm": 3.470461845397949,
+      "learning_rate": 4.9844615540942584e-05,
+      "loss": 0.9412,
+      "step": 1870
+    },
+    {
+      "epoch": 0.18831071267591526,
+      "grad_norm": 3.2644646167755127,
+      "learning_rate": 4.9842855674390306e-05,
+      "loss": 0.9621,
+      "step": 1880
+    },
+    {
+      "epoch": 0.18931236540291482,
+      "grad_norm": 2.799272298812866,
+      "learning_rate": 4.9841085929274404e-05,
+      "loss": 1.0094,
+      "step": 1890
+    },
+    {
+      "epoch": 0.19031401812991436,
+      "grad_norm": 2.402174949645996,
+      "learning_rate": 4.983930630629862e-05,
+      "loss": 0.9504,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1913156708569139,
+      "grad_norm": 3.4706664085388184,
+      "learning_rate": 4.983751680617059e-05,
+      "loss": 0.9323,
+      "step": 1910
+    },
+    {
+      "epoch": 0.19231732358391346,
+      "grad_norm": 3.1237800121307373,
+      "learning_rate": 4.9835717429601905e-05,
+      "loss": 0.9726,
+      "step": 1920
+    },
+    {
+      "epoch": 0.193318976310913,
+      "grad_norm": 2.986846446990967,
+      "learning_rate": 4.983390817730808e-05,
+      "loss": 0.9088,
+      "step": 1930
+    },
+    {
+      "epoch": 0.19432062903791256,
+      "grad_norm": 2.4418344497680664,
+      "learning_rate": 4.983208905000854e-05,
+      "loss": 1.0038,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1953222817649121,
+      "grad_norm": 3.0399439334869385,
+      "learning_rate": 4.9830260048426636e-05,
+      "loss": 1.0384,
+      "step": 1950
+    },
+    {
+      "epoch": 0.19632393449191166,
+      "grad_norm": 2.5777769088745117,
+      "learning_rate": 4.9828421173289676e-05,
+      "loss": 0.9796,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1973255872189112,
+      "grad_norm": 3.4346256256103516,
+      "learning_rate": 4.982657242532886e-05,
+      "loss": 0.99,
+      "step": 1970
+    },
+    {
+      "epoch": 0.19832723994591075,
+      "grad_norm": 2.746309518814087,
+      "learning_rate": 4.982471380527934e-05,
+      "loss": 0.9916,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1993288926729103,
+      "grad_norm": 2.7143187522888184,
+      "learning_rate": 4.982284531388017e-05,
+      "loss": 0.9572,
+      "step": 1990
+    },
+    {
+      "epoch": 0.20033054539990985,
+      "grad_norm": 2.336798667907715,
+      "learning_rate": 4.9820966951874345e-05,
+      "loss": 0.9508,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2013321981269094,
+      "grad_norm": 2.9657061100006104,
+      "learning_rate": 4.981907872000878e-05,
+      "loss": 1.0059,
+      "step": 2010
+    },
+    {
+      "epoch": 0.20233385085390895,
+      "grad_norm": 2.5726466178894043,
+      "learning_rate": 4.9817180619034324e-05,
+      "loss": 0.9964,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2033355035809085,
+      "grad_norm": 2.542393207550049,
+      "learning_rate": 4.9815272649705733e-05,
+      "loss": 0.9314,
+      "step": 2030
+    },
+    {
+      "epoch": 0.20433715630790805,
+      "grad_norm": 2.635509729385376,
+      "learning_rate": 4.98133548127817e-05,
+      "loss": 0.9909,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2053388090349076,
+      "grad_norm": 2.7643823623657227,
+      "learning_rate": 4.981142710902482e-05,
+      "loss": 0.9538,
+      "step": 2050
+    },
+    {
+      "epoch": 0.20634046176190715,
+      "grad_norm": 2.9941859245300293,
+      "learning_rate": 4.980948953920166e-05,
+      "loss": 1.0004,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2073421144889067,
+      "grad_norm": 2.7872562408447266,
+      "learning_rate": 4.980754210408266e-05,
+      "loss": 0.9699,
+      "step": 2070
+    },
+    {
+      "epoch": 0.20834376721590625,
+      "grad_norm": 4.475282192230225,
+      "learning_rate": 4.98055848044422e-05,
+      "loss": 1.0183,
+      "step": 2080
+    },
+    {
+      "epoch": 0.20934541994290579,
+      "grad_norm": 2.9876139163970947,
+      "learning_rate": 4.980361764105859e-05,
+      "loss": 0.9116,
+      "step": 2090
+    },
+    {
+      "epoch": 0.21034707266990535,
+      "grad_norm": 3.074251890182495,
+      "learning_rate": 4.980164061471405e-05,
+      "loss": 0.9749,
+      "step": 2100
+    },
+    {
+      "epoch": 0.21134872539690489,
+      "grad_norm": 3.2771337032318115,
+      "learning_rate": 4.979965372619475e-05,
+      "loss": 0.9717,
+      "step": 2110
+    },
+    {
+      "epoch": 0.21235037812390445,
+      "grad_norm": 3.6136999130249023,
+      "learning_rate": 4.9797656976290736e-05,
+      "loss": 0.95,
+      "step": 2120
+    },
+    {
+      "epoch": 0.21335203085090398,
+      "grad_norm": 3.1407272815704346,
+      "learning_rate": 4.979565036579601e-05,
+      "loss": 0.9831,
+      "step": 2130
+    },
+    {
+      "epoch": 0.21435368357790355,
+      "grad_norm": 2.501585006713867,
+      "learning_rate": 4.979363389550849e-05,
+      "loss": 0.9725,
+      "step": 2140
+    },
+    {
+      "epoch": 0.21535533630490308,
+      "grad_norm": 2.292156219482422,
+      "learning_rate": 4.979160756623e-05,
+      "loss": 0.8937,
+      "step": 2150
+    },
+    {
+      "epoch": 0.21635698903190265,
+      "grad_norm": 2.627570629119873,
+      "learning_rate": 4.978957137876629e-05,
+      "loss": 0.9211,
+      "step": 2160
+    },
+    {
+      "epoch": 0.21735864175890218,
+      "grad_norm": 3.1033174991607666,
+      "learning_rate": 4.978752533392705e-05,
+      "loss": 0.9802,
+      "step": 2170
+    },
+    {
+      "epoch": 0.21836029448590175,
+      "grad_norm": 2.952439069747925,
+      "learning_rate": 4.978546943252586e-05,
+      "loss": 0.8931,
+      "step": 2180
+    },
+    {
+      "epoch": 0.21936194721290128,
+      "grad_norm": 2.534576416015625,
+      "learning_rate": 4.978340367538023e-05,
+      "loss": 1.0088,
+      "step": 2190
+    },
+    {
+      "epoch": 0.22036359993990084,
+      "grad_norm": 2.862668752670288,
+      "learning_rate": 4.9781328063311614e-05,
+      "loss": 0.9563,
+      "step": 2200
+    },
+    {
+      "epoch": 0.22136525266690038,
+      "grad_norm": 2.849451780319214,
+      "learning_rate": 4.977924259714534e-05,
+      "loss": 0.986,
+      "step": 2210
+    },
+    {
+      "epoch": 0.22236690539389994,
+      "grad_norm": 2.9716901779174805,
+      "learning_rate": 4.977714727771069e-05,
+      "loss": 0.8527,
+      "step": 2220
+    },
+    {
+      "epoch": 0.22336855812089948,
+      "grad_norm": 2.634556770324707,
+      "learning_rate": 4.977504210584084e-05,
+      "loss": 0.9426,
+      "step": 2230
+    },
+    {
+      "epoch": 0.22437021084789904,
+      "grad_norm": 2.3726024627685547,
+      "learning_rate": 4.97729270823729e-05,
+      "loss": 0.9265,
+      "step": 2240
+    },
+    {
+      "epoch": 0.22537186357489858,
+      "grad_norm": 3.005441665649414,
+      "learning_rate": 4.977080220814789e-05,
+      "loss": 0.919,
+      "step": 2250
+    },
+    {
+      "epoch": 0.22637351630189814,
+      "grad_norm": 3.0273005962371826,
+      "learning_rate": 4.976866748401075e-05,
+      "loss": 0.9681,
+      "step": 2260
+    },
+    {
+      "epoch": 0.22737516902889768,
+      "grad_norm": 2.832097291946411,
+      "learning_rate": 4.976652291081035e-05,
+      "loss": 0.9547,
+      "step": 2270
+    },
+    {
+      "epoch": 0.22837682175589724,
+      "grad_norm": 2.40688157081604,
+      "learning_rate": 4.976436848939944e-05,
+      "loss": 0.949,
+      "step": 2280
+    },
+    {
+      "epoch": 0.22937847448289678,
+      "grad_norm": 2.177398681640625,
+      "learning_rate": 4.976220422063473e-05,
+      "loss": 0.9046,
+      "step": 2290
+    },
+    {
+      "epoch": 0.23038012720989634,
+      "grad_norm": 2.506649971008301,
+      "learning_rate": 4.97600301053768e-05,
+      "loss": 0.9084,
+      "step": 2300
+    },
+    {
+      "epoch": 0.23138177993689588,
+      "grad_norm": 2.7216312885284424,
+      "learning_rate": 4.975784614449019e-05,
+      "loss": 0.8898,
+      "step": 2310
+    },
+    {
+      "epoch": 0.23238343266389544,
+      "grad_norm": 2.622244358062744,
+      "learning_rate": 4.975565233884333e-05,
+      "loss": 0.9973,
+      "step": 2320
+    },
+    {
+      "epoch": 0.23338508539089498,
+      "grad_norm": 2.5111606121063232,
+      "learning_rate": 4.9753448689308557e-05,
+      "loss": 0.9258,
+      "step": 2330
+    },
+    {
+      "epoch": 0.23438673811789454,
+      "grad_norm": 3.298383951187134,
+      "learning_rate": 4.975123519676215e-05,
+      "loss": 0.9868,
+      "step": 2340
+    },
+    {
+      "epoch": 0.23538839084489407,
+      "grad_norm": 3.6510188579559326,
+      "learning_rate": 4.974901186208428e-05,
+      "loss": 0.9603,
+      "step": 2350
+    },
+    {
+      "epoch": 0.23639004357189364,
+      "grad_norm": 2.3483967781066895,
+      "learning_rate": 4.9746778686159045e-05,
+      "loss": 0.8984,
+      "step": 2360
+    },
+    {
+      "epoch": 0.23739169629889317,
+      "grad_norm": 2.734997510910034,
+      "learning_rate": 4.9744535669874435e-05,
+      "loss": 0.9412,
+      "step": 2370
+    },
+    {
+      "epoch": 0.23839334902589274,
+      "grad_norm": 2.5782012939453125,
+      "learning_rate": 4.974228281412239e-05,
+      "loss": 1.002,
+      "step": 2380
+    },
+    {
+      "epoch": 0.23939500175289227,
+      "grad_norm": 2.8419687747955322,
+      "learning_rate": 4.974002011979871e-05,
+      "loss": 0.9378,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2403966544798918,
+      "grad_norm": 2.9137983322143555,
+      "learning_rate": 4.973774758780316e-05,
+      "loss": 0.9227,
+      "step": 2400
+    },
+    {
+      "epoch": 0.24139830720689137,
+      "grad_norm": 2.580446481704712,
+      "learning_rate": 4.9735465219039396e-05,
+      "loss": 0.9542,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2423999599338909,
+      "grad_norm": 3.0914855003356934,
+      "learning_rate": 4.973317301441497e-05,
+      "loss": 0.9186,
+      "step": 2420
+    },
+    {
+      "epoch": 0.24340161266089047,
+      "grad_norm": 2.856234073638916,
+      "learning_rate": 4.973087097484136e-05,
+      "loss": 0.9466,
+      "step": 2430
+    },
+    {
+      "epoch": 0.24440326538789,
+      "grad_norm": 3.1362180709838867,
+      "learning_rate": 4.972855910123397e-05,
+      "loss": 0.8701,
+      "step": 2440
+    },
+    {
+      "epoch": 0.24540491811488957,
+      "grad_norm": 2.7053310871124268,
+      "learning_rate": 4.972623739451208e-05,
+      "loss": 0.9137,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2464065708418891,
+      "grad_norm": 2.5193698406219482,
+      "learning_rate": 4.9723905855598904e-05,
+      "loss": 0.8837,
+      "step": 2460
+    },
+    {
+      "epoch": 0.24740822356888867,
+      "grad_norm": 2.841562509536743,
+      "learning_rate": 4.972156448542156e-05,
+      "loss": 0.9857,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2484098762958882,
+      "grad_norm": 2.9049248695373535,
+      "learning_rate": 4.9719213284911084e-05,
+      "loss": 0.947,
+      "step": 2480
+    },
+    {
+      "epoch": 0.24941152902288777,
+      "grad_norm": 2.958031415939331,
+      "learning_rate": 4.971685225500239e-05,
+      "loss": 0.9208,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2504131817498873,
+      "grad_norm": 2.562629222869873,
+      "learning_rate": 4.9714481396634346e-05,
+      "loss": 0.8968,
+      "step": 2500
+    },
+    {
+      "epoch": 0.25141483447688684,
+      "grad_norm": 2.949819803237915,
+      "learning_rate": 4.97121007107497e-05,
+      "loss": 0.856,
+      "step": 2510
+    },
+    {
+      "epoch": 0.25241648720388643,
+      "grad_norm": 3.606531858444214,
+      "learning_rate": 4.97097101982951e-05,
+      "loss": 0.9165,
+      "step": 2520
+    },
+    {
+      "epoch": 0.25341813993088597,
+      "grad_norm": 3.1925840377807617,
+      "learning_rate": 4.970730986022113e-05,
+      "loss": 0.9481,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2544197926578855,
+      "grad_norm": 3.001221179962158,
+      "learning_rate": 4.9704899697482256e-05,
+      "loss": 0.9386,
+      "step": 2540
+    },
+    {
+      "epoch": 0.25542144538488504,
+      "grad_norm": 3.6826331615448,
+      "learning_rate": 4.9702479711036864e-05,
+      "loss": 0.8899,
+      "step": 2550
+    },
+    {
+      "epoch": 0.25642309811188463,
+      "grad_norm": 3.200007200241089,
+      "learning_rate": 4.970004990184724e-05,
+      "loss": 0.8934,
+      "step": 2560
+    },
+    {
+      "epoch": 0.25742475083888416,
+      "grad_norm": 2.6780619621276855,
+      "learning_rate": 4.969761027087957e-05,
+      "loss": 0.8705,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2584264035658837,
+      "grad_norm": 2.7149860858917236,
+      "learning_rate": 4.969516081910397e-05,
+      "loss": 0.9574,
+      "step": 2580
+    },
+    {
+      "epoch": 0.25942805629288324,
+      "grad_norm": 2.5190060138702393,
+      "learning_rate": 4.969270154749444e-05,
+      "loss": 0.9411,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2604297090198828,
+      "grad_norm": 2.75197696685791,
+      "learning_rate": 4.9690232457028887e-05,
+      "loss": 0.9248,
+      "step": 2600
+    },
+    {
+      "epoch": 0.26143136174688236,
+      "grad_norm": 2.462170362472534,
+      "learning_rate": 4.968775354868912e-05,
+      "loss": 0.9209,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2624330144738819,
+      "grad_norm": 2.7664783000946045,
+      "learning_rate": 4.9685264823460866e-05,
+      "loss": 0.8915,
+      "step": 2620
+    },
+    {
+      "epoch": 0.26343466720088143,
+      "grad_norm": 2.3150320053100586,
+      "learning_rate": 4.968276628233374e-05,
+      "loss": 0.9826,
+      "step": 2630
+    },
+    {
+      "epoch": 0.264436319927881,
+      "grad_norm": 2.6080338954925537,
+      "learning_rate": 4.9680257926301274e-05,
+      "loss": 0.8408,
+      "step": 2640
+    },
+    {
+      "epoch": 0.26543797265488056,
+      "grad_norm": 3.5915043354034424,
+      "learning_rate": 4.967773975636088e-05,
+      "loss": 0.844,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2664396253818801,
+      "grad_norm": 2.8301284313201904,
+      "learning_rate": 4.967521177351392e-05,
+      "loss": 0.9263,
+      "step": 2660
+    },
+    {
+      "epoch": 0.26744127810887963,
+      "grad_norm": 2.7187178134918213,
+      "learning_rate": 4.9672673978765594e-05,
+      "loss": 0.8942,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2684429308358792,
+      "grad_norm": 2.7909884452819824,
+      "learning_rate": 4.9670126373125056e-05,
+      "loss": 0.9328,
+      "step": 2680
+    },
+    {
+      "epoch": 0.26944458356287876,
+      "grad_norm": 2.38917875289917,
+      "learning_rate": 4.9667568957605324e-05,
+      "loss": 0.9005,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2704462362898783,
+      "grad_norm": 2.494300127029419,
+      "learning_rate": 4.966500173322335e-05,
+      "loss": 0.8757,
+      "step": 2700
+    },
+    {
+      "epoch": 0.27144788901687783,
+      "grad_norm": 2.545513868331909,
+      "learning_rate": 4.966242470099997e-05,
+      "loss": 0.9322,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2724495417438774,
+      "grad_norm": 2.4247043132781982,
+      "learning_rate": 4.96598378619599e-05,
+      "loss": 0.9301,
+      "step": 2720
+    },
+    {
+      "epoch": 0.27345119447087696,
+      "grad_norm": 2.355764865875244,
+      "learning_rate": 4.96572412171318e-05,
+      "loss": 0.8569,
+      "step": 2730
+    },
+    {
+      "epoch": 0.2744528471978765,
+      "grad_norm": 3.5294244289398193,
+      "learning_rate": 4.965463476754819e-05,
+      "loss": 0.9178,
+      "step": 2740
+    },
+    {
+      "epoch": 0.27545449992487603,
+      "grad_norm": 3.509269952774048,
+      "learning_rate": 4.9652018514245514e-05,
+      "loss": 0.861,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2764561526518756,
+      "grad_norm": 2.9568450450897217,
+      "learning_rate": 4.96493924582641e-05,
+      "loss": 0.8833,
+      "step": 2760
+    },
+    {
+      "epoch": 0.27745780537887516,
+      "grad_norm": 2.9352688789367676,
+      "learning_rate": 4.964675660064817e-05,
+      "loss": 0.8888,
+      "step": 2770
+    },
+    {
+      "epoch": 0.2784594581058747,
+      "grad_norm": 2.9790830612182617,
+      "learning_rate": 4.9644110942445864e-05,
+      "loss": 0.9142,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2794611108328742,
+      "grad_norm": 2.514970064163208,
+      "learning_rate": 4.9641455484709196e-05,
+      "loss": 0.9056,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2804627635598738,
+      "grad_norm": 3.5085692405700684,
+      "learning_rate": 4.9638790228494106e-05,
+      "loss": 0.857,
+      "step": 2800
+    },
+    {
+      "epoch": 0.28146441628687335,
+      "grad_norm": 4.000670433044434,
+      "learning_rate": 4.9636115174860396e-05,
+      "loss": 0.8086,
+      "step": 2810
+    },
+    {
+      "epoch": 0.2824660690138729,
+      "grad_norm": 2.9119763374328613,
+      "learning_rate": 4.9633430324871776e-05,
+      "loss": 0.8537,
+      "step": 2820
+    },
+    {
+      "epoch": 0.2834677217408724,
+      "grad_norm": 1.9512786865234375,
+      "learning_rate": 4.963073567959587e-05,
+      "loss": 0.7311,
+      "step": 2830
+    },
+    {
+      "epoch": 0.284469374467872,
+      "grad_norm": 2.5956673622131348,
+      "learning_rate": 4.962803124010417e-05,
+      "loss": 0.8748,
+      "step": 2840
+    },
+    {
+      "epoch": 0.28547102719487155,
+      "grad_norm": 2.648887872695923,
+      "learning_rate": 4.9625317007472085e-05,
+      "loss": 0.911,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2864726799218711,
+      "grad_norm": 2.7080471515655518,
+      "learning_rate": 4.96225929827789e-05,
+      "loss": 0.8843,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2874743326488706,
+      "grad_norm": 2.9496068954467773,
+      "learning_rate": 4.961985916710781e-05,
+      "loss": 0.8772,
+      "step": 2870
+    },
+    {
+      "epoch": 0.28847598537587016,
+      "grad_norm": 3.3540945053100586,
+      "learning_rate": 4.961711556154588e-05,
+      "loss": 0.8951,
+      "step": 2880
+    },
+    {
+      "epoch": 0.28947763810286975,
+      "grad_norm": 2.877737045288086,
+      "learning_rate": 4.961436216718409e-05,
+      "loss": 0.9708,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2904792908298693,
+      "grad_norm": 2.5713319778442383,
+      "learning_rate": 4.961159898511732e-05,
+      "loss": 0.8817,
+      "step": 2900
+    },
+    {
+      "epoch": 0.2914809435568688,
+      "grad_norm": 3.155395746231079,
+      "learning_rate": 4.960882601644431e-05,
+      "loss": 0.9464,
+      "step": 2910
+    },
+    {
+      "epoch": 0.29248259628386836,
+      "grad_norm": 2.9532198905944824,
+      "learning_rate": 4.960604326226771e-05,
+      "loss": 0.8817,
+      "step": 2920
+    },
+    {
+      "epoch": 0.29348424901086795,
+      "grad_norm": 3.1550869941711426,
+      "learning_rate": 4.960325072369407e-05,
+      "loss": 0.8659,
+      "step": 2930
+    },
+    {
+      "epoch": 0.2944859017378675,
+      "grad_norm": 2.6398537158966064,
+      "learning_rate": 4.960044840183381e-05,
+      "loss": 0.9133,
+      "step": 2940
+    },
+    {
+      "epoch": 0.295487554464867,
+      "grad_norm": 2.6897711753845215,
+      "learning_rate": 4.959763629780126e-05,
+      "loss": 0.8332,
+      "step": 2950
+    },
+    {
+      "epoch": 0.29648920719186656,
+      "grad_norm": 3.5697755813598633,
+      "learning_rate": 4.959481441271462e-05,
+      "loss": 0.9082,
+      "step": 2960
+    },
+    {
+      "epoch": 0.29749085991886615,
+      "grad_norm": 3.7671542167663574,
+      "learning_rate": 4.9591982747696006e-05,
+      "loss": 0.9394,
+      "step": 2970
+    },
+    {
+      "epoch": 0.2984925126458657,
+      "grad_norm": 2.823279857635498,
+      "learning_rate": 4.958914130387139e-05,
+      "loss": 0.8998,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2994941653728652,
+      "grad_norm": 2.3585197925567627,
+      "learning_rate": 4.958629008237066e-05,
+      "loss": 0.8573,
+      "step": 2990
+    },
+    {
+      "epoch": 0.30049581809986475,
+      "grad_norm": 2.3386313915252686,
+      "learning_rate": 4.958342908432757e-05,
+      "loss": 0.7891,
+      "step": 3000
+    },
+    {
+      "epoch": 0.30149747082686434,
+      "grad_norm": 2.2583234310150146,
+      "learning_rate": 4.958055831087979e-05,
+      "loss": 0.912,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3024991235538639,
+      "grad_norm": 3.0650525093078613,
+      "learning_rate": 4.957767776316885e-05,
+      "loss": 0.9714,
+      "step": 3020
+    },
+    {
+      "epoch": 0.3035007762808634,
+      "grad_norm": 2.90987491607666,
+      "learning_rate": 4.9574787442340185e-05,
+      "loss": 0.7965,
+      "step": 3030
+    },
+    {
+      "epoch": 0.30450242900786295,
+      "grad_norm": 2.936350107192993,
+      "learning_rate": 4.95718873495431e-05,
+      "loss": 0.8288,
+      "step": 3040
+    },
+    {
+      "epoch": 0.30550408173486254,
+      "grad_norm": 3.2421929836273193,
+      "learning_rate": 4.956897748593079e-05,
+      "loss": 0.8186,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3065057344618621,
+      "grad_norm": 2.5979058742523193,
+      "learning_rate": 4.956605785266034e-05,
+      "loss": 0.875,
+      "step": 3060
+    },
+    {
+      "epoch": 0.3075073871888616,
+      "grad_norm": 2.861781597137451,
+      "learning_rate": 4.956312845089274e-05,
+      "loss": 0.8187,
+      "step": 3070
+    },
+    {
+      "epoch": 0.30850903991586115,
+      "grad_norm": 2.5205078125,
+      "learning_rate": 4.9560189281792815e-05,
+      "loss": 0.8718,
+      "step": 3080
+    },
+    {
+      "epoch": 0.30951069264286074,
+      "grad_norm": 2.3308043479919434,
+      "learning_rate": 4.955724034652931e-05,
+      "loss": 0.9551,
+      "step": 3090
+    },
+    {
+      "epoch": 0.3105123453698603,
+      "grad_norm": 2.916992425918579,
+      "learning_rate": 4.955428164627486e-05,
+      "loss": 0.828,
+      "step": 3100
+    },
+    {
+      "epoch": 0.3115139980968598,
+      "grad_norm": 2.490170478820801,
+      "learning_rate": 4.955131318220595e-05,
+      "loss": 0.8135,
+      "step": 3110
+    },
+    {
+      "epoch": 0.31251565082385935,
+      "grad_norm": 3.505068302154541,
+      "learning_rate": 4.954833495550297e-05,
+      "loss": 0.8689,
+      "step": 3120
+    },
+    {
+      "epoch": 0.31351730355085894,
+      "grad_norm": 2.6903154850006104,
+      "learning_rate": 4.9545346967350204e-05,
+      "loss": 0.9086,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3145189562778585,
+      "grad_norm": 3.43782377243042,
+      "learning_rate": 4.954234921893579e-05,
+      "loss": 0.8903,
+      "step": 3140
+    },
+    {
+      "epoch": 0.315520609004858,
+      "grad_norm": 2.6618101596832275,
+      "learning_rate": 4.953934171145175e-05,
+      "loss": 0.9068,
+      "step": 3150
+    },
+    {
+      "epoch": 0.31652226173185755,
+      "grad_norm": 3.6636927127838135,
+      "learning_rate": 4.953632444609401e-05,
+      "loss": 0.8745,
+      "step": 3160
+    },
+    {
+      "epoch": 0.31752391445885714,
+      "grad_norm": 2.4644505977630615,
+      "learning_rate": 4.953329742406235e-05,
+      "loss": 0.9156,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3185255671858567,
+      "grad_norm": 2.7352774143218994,
+      "learning_rate": 4.9530260646560455e-05,
+      "loss": 0.8799,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3195272199128562,
+      "grad_norm": 2.6652235984802246,
+      "learning_rate": 4.952721411479587e-05,
+      "loss": 0.8353,
+      "step": 3190
+    },
+    {
+      "epoch": 0.32052887263985574,
+      "grad_norm": 2.7542600631713867,
+      "learning_rate": 4.952415782998001e-05,
+      "loss": 0.7625,
+      "step": 3200
+    },
+    {
+      "epoch": 0.32153052536685534,
+      "grad_norm": 2.849294662475586,
+      "learning_rate": 4.9521091793328204e-05,
+      "loss": 0.8022,
+      "step": 3210
+    },
+    {
+      "epoch": 0.32253217809385487,
+      "grad_norm": 2.48455548286438,
+      "learning_rate": 4.951801600605962e-05,
+      "loss": 0.8215,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3235338308208544,
+      "grad_norm": 3.339585304260254,
+      "learning_rate": 4.9514930469397326e-05,
+      "loss": 0.8675,
+      "step": 3230
+    },
+    {
+      "epoch": 0.32453548354785394,
+      "grad_norm": 2.703068494796753,
+      "learning_rate": 4.951183518456827e-05,
+      "loss": 0.7836,
+      "step": 3240
+    },
+    {
+      "epoch": 0.32553713627485353,
+      "grad_norm": 2.545600414276123,
+      "learning_rate": 4.950873015280325e-05,
+      "loss": 0.8031,
+      "step": 3250
+    },
+    {
+      "epoch": 0.32653878900185307,
+      "grad_norm": 3.298454523086548,
+      "learning_rate": 4.9505615375336965e-05,
+      "loss": 0.8976,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3275404417288526,
+      "grad_norm": 2.814692735671997,
+      "learning_rate": 4.950249085340799e-05,
+      "loss": 0.8475,
+      "step": 3270
+    },
+    {
+      "epoch": 0.32854209445585214,
+      "grad_norm": 2.775646448135376,
+      "learning_rate": 4.949935658825876e-05,
+      "loss": 0.8412,
+      "step": 3280
+    },
+    {
+      "epoch": 0.32954374718285173,
+      "grad_norm": 2.9351887702941895,
+      "learning_rate": 4.9496212581135585e-05,
+      "loss": 0.8447,
+      "step": 3290
+    },
+    {
+      "epoch": 0.33054539990985127,
+      "grad_norm": 2.701782464981079,
+      "learning_rate": 4.9493058833288666e-05,
+      "loss": 0.9201,
+      "step": 3300
+    },
+    {
+      "epoch": 0.3315470526368508,
+      "grad_norm": 3.1329259872436523,
+      "learning_rate": 4.948989534597205e-05,
+      "loss": 0.7931,
+      "step": 3310
+    },
+    {
+      "epoch": 0.33254870536385034,
+      "grad_norm": 3.1399502754211426,
+      "learning_rate": 4.948672212044369e-05,
+      "loss": 0.8213,
+      "step": 3320
+    },
+    {
+      "epoch": 0.33355035809084993,
+      "grad_norm": 2.487262725830078,
+      "learning_rate": 4.948353915796538e-05,
+      "loss": 0.82,
+      "step": 3330
+    },
+    {
+      "epoch": 0.33455201081784947,
+      "grad_norm": 2.3617560863494873,
+      "learning_rate": 4.948034645980281e-05,
+      "loss": 0.8816,
+      "step": 3340
+    },
+    {
+      "epoch": 0.335553663544849,
+      "grad_norm": 2.443876266479492,
+      "learning_rate": 4.947714402722552e-05,
+      "loss": 0.8592,
+      "step": 3350
+    },
+    {
+      "epoch": 0.33655531627184854,
+      "grad_norm": 3.451190948486328,
+      "learning_rate": 4.947393186150694e-05,
+      "loss": 0.8691,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3375569689988481,
+      "grad_norm": 2.79589581489563,
+      "learning_rate": 4.947070996392435e-05,
+      "loss": 0.8485,
+      "step": 3370
+    },
+    {
+      "epoch": 0.33855862172584766,
+      "grad_norm": 3.133854389190674,
+      "learning_rate": 4.946747833575894e-05,
+      "loss": 0.8401,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3395602744528472,
+      "grad_norm": 2.8713510036468506,
+      "learning_rate": 4.946423697829572e-05,
+      "loss": 0.7918,
+      "step": 3390
+    },
+    {
+      "epoch": 0.34056192717984674,
+      "grad_norm": 3.4451394081115723,
+      "learning_rate": 4.946098589282359e-05,
+      "loss": 0.8507,
+      "step": 3400
+    },
+    {
+      "epoch": 0.34156357990684627,
+      "grad_norm": 2.4211623668670654,
+      "learning_rate": 4.945772508063531e-05,
+      "loss": 0.9028,
+      "step": 3410
+    },
+    {
+      "epoch": 0.34256523263384586,
+      "grad_norm": 3.622356414794922,
+      "learning_rate": 4.945445454302754e-05,
+      "loss": 0.7338,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3435668853608454,
+      "grad_norm": 2.8370401859283447,
+      "learning_rate": 4.945117428130076e-05,
+      "loss": 0.8819,
+      "step": 3430
+    },
+    {
+      "epoch": 0.34456853808784493,
+      "grad_norm": 2.372068405151367,
+      "learning_rate": 4.944788429675936e-05,
+      "loss": 0.817,
+      "step": 3440
+    },
+    {
+      "epoch": 0.34557019081484447,
+      "grad_norm": 2.5194711685180664,
+      "learning_rate": 4.944458459071156e-05,
+      "loss": 0.8744,
+      "step": 3450
+    },
+    {
+      "epoch": 0.34657184354184406,
+      "grad_norm": 2.253473997116089,
+      "learning_rate": 4.944127516446947e-05,
+      "loss": 0.8461,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3475734962688436,
+      "grad_norm": 2.651766300201416,
+      "learning_rate": 4.9437956019349054e-05,
+      "loss": 0.8544,
+      "step": 3470
+    },
+    {
+      "epoch": 0.34857514899584313,
+      "grad_norm": 2.0998165607452393,
+      "learning_rate": 4.943462715667015e-05,
+      "loss": 0.8272,
+      "step": 3480
+    },
+    {
+      "epoch": 0.34957680172284267,
+      "grad_norm": 3.30053973197937,
+      "learning_rate": 4.9431288577756446e-05,
+      "loss": 0.8739,
+      "step": 3490
+    },
+    {
+      "epoch": 0.35057845444984226,
+      "grad_norm": 2.7177000045776367,
+      "learning_rate": 4.9427940283935504e-05,
+      "loss": 0.8858,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3515801071768418,
+      "grad_norm": 2.158338785171509,
+      "learning_rate": 4.9424582276538746e-05,
+      "loss": 0.7827,
+      "step": 3510
+    },
+    {
+      "epoch": 0.35258175990384133,
+      "grad_norm": 3.1199424266815186,
+      "learning_rate": 4.9421214556901454e-05,
+      "loss": 0.8491,
+      "step": 3520
+    },
+    {
+      "epoch": 0.35358341263084087,
+      "grad_norm": 2.8127048015594482,
+      "learning_rate": 4.941783712636278e-05,
+      "loss": 0.8002,
+      "step": 3530
+    },
+    {
+      "epoch": 0.35458506535784046,
+      "grad_norm": 2.9365146160125732,
+      "learning_rate": 4.941444998626573e-05,
+      "loss": 0.9608,
+      "step": 3540
+    },
+    {
+      "epoch": 0.35558671808484,
+      "grad_norm": 3.4413721561431885,
+      "learning_rate": 4.9411053137957174e-05,
+      "loss": 0.8685,
+      "step": 3550
+    },
+    {
+      "epoch": 0.35658837081183953,
+      "grad_norm": 2.9965898990631104,
+      "learning_rate": 4.940764658278785e-05,
+      "loss": 0.8144,
+      "step": 3560
+    },
+    {
+      "epoch": 0.35759002353883906,
+      "grad_norm": 2.883406162261963,
+      "learning_rate": 4.9404230322112324e-05,
+      "loss": 0.8437,
+      "step": 3570
+    },
+    {
+      "epoch": 0.35859167626583865,
+      "grad_norm": 3.0368754863739014,
+      "learning_rate": 4.940080435728907e-05,
+      "loss": 0.8707,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3595933289928382,
+      "grad_norm": 2.8644607067108154,
+      "learning_rate": 4.939736868968038e-05,
+      "loss": 0.8184,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3605949817198377,
+      "grad_norm": 2.6353414058685303,
+      "learning_rate": 4.9393923320652424e-05,
+      "loss": 0.8343,
+      "step": 3600
+    },
+    {
+      "epoch": 0.36159663444683726,
+      "grad_norm": 3.469896078109741,
+      "learning_rate": 4.939046825157523e-05,
+      "loss": 0.845,
+      "step": 3610
+    },
+    {
+      "epoch": 0.36259828717383685,
+      "grad_norm": 2.8434600830078125,
+      "learning_rate": 4.9387003483822666e-05,
+      "loss": 0.8833,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3635999399008364,
+      "grad_norm": 3.1709251403808594,
+      "learning_rate": 4.938352901877249e-05,
+      "loss": 0.826,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3646015926278359,
+      "grad_norm": 2.7404329776763916,
+      "learning_rate": 4.938004485780628e-05,
+      "loss": 0.7455,
+      "step": 3640
+    },
+    {
+      "epoch": 0.36560324535483546,
+      "grad_norm": 4.057974338531494,
+      "learning_rate": 4.937655100230949e-05,
+      "loss": 0.7923,
+      "step": 3650
+    },
+    {
+      "epoch": 0.36660489808183505,
+      "grad_norm": 2.943912982940674,
+      "learning_rate": 4.937304745367143e-05,
+      "loss": 0.8386,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3676065508088346,
+      "grad_norm": 3.354377269744873,
+      "learning_rate": 4.936953421328524e-05,
+      "loss": 0.8257,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3686082035358341,
+      "grad_norm": 2.800428628921509,
+      "learning_rate": 4.936601128254794e-05,
+      "loss": 0.7587,
+      "step": 3680
+    },
+    {
+      "epoch": 0.36960985626283366,
+      "grad_norm": 3.1578376293182373,
+      "learning_rate": 4.936247866286041e-05,
+      "loss": 0.7953,
+      "step": 3690
+    },
+    {
+      "epoch": 0.37061150898983325,
+      "grad_norm": 2.962111711502075,
+      "learning_rate": 4.935893635562735e-05,
+      "loss": 0.826,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3716131617168328,
+      "grad_norm": 2.9009153842926025,
+      "learning_rate": 4.935538436225733e-05,
+      "loss": 0.7948,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3726148144438323,
+      "grad_norm": 2.9375569820404053,
+      "learning_rate": 4.9351822684162787e-05,
+      "loss": 0.8348,
+      "step": 3720
+    },
+    {
+      "epoch": 0.37361646717083186,
+      "grad_norm": 3.2307522296905518,
+      "learning_rate": 4.934825132275999e-05,
+      "loss": 0.8755,
+      "step": 3730
+    },
+    {
+      "epoch": 0.37461811989783145,
+      "grad_norm": 3.187584161758423,
+      "learning_rate": 4.934467027946905e-05,
+      "loss": 0.8315,
+      "step": 3740
+    },
+    {
+      "epoch": 0.375619772624831,
+      "grad_norm": 2.4784653186798096,
+      "learning_rate": 4.9341079555713946e-05,
+      "loss": 0.8161,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3766214253518305,
+      "grad_norm": 3.0355725288391113,
+      "learning_rate": 4.9337479152922504e-05,
+      "loss": 0.7234,
+      "step": 3760
+    },
+    {
+      "epoch": 0.37762307807883005,
+      "grad_norm": 5.005615711212158,
+      "learning_rate": 4.93338690725264e-05,
+      "loss": 0.872,
+      "step": 3770
+    },
+    {
+      "epoch": 0.37862473080582965,
+      "grad_norm": 2.9800662994384766,
+      "learning_rate": 4.933024931596114e-05,
+      "loss": 0.8399,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3796263835328292,
+      "grad_norm": 3.00419282913208,
+      "learning_rate": 4.93266198846661e-05,
+      "loss": 0.7835,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3806280362598287,
+      "grad_norm": 2.776589870452881,
+      "learning_rate": 4.93229807800845e-05,
+      "loss": 0.8586,
+      "step": 3800
+    },
+    {
+      "epoch": 0.38162968898682825,
+      "grad_norm": 2.228154420852661,
+      "learning_rate": 4.9319332003663385e-05,
+      "loss": 0.7737,
+      "step": 3810
+    },
+    {
+      "epoch": 0.3826313417138278,
+      "grad_norm": 2.2705202102661133,
+      "learning_rate": 4.931567355685368e-05,
+      "loss": 0.8327,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3836329944408274,
+      "grad_norm": 3.1992833614349365,
+      "learning_rate": 4.9312005441110126e-05,
+      "loss": 0.7732,
+      "step": 3830
+    },
+    {
+      "epoch": 0.3846346471678269,
+      "grad_norm": 2.792924642562866,
+      "learning_rate": 4.930832765789132e-05,
+      "loss": 0.7673,
+      "step": 3840
+    },
+    {
+      "epoch": 0.38563629989482645,
+      "grad_norm": 2.705148935317993,
+      "learning_rate": 4.9304640208659714e-05,
+      "loss": 0.7368,
+      "step": 3850
+    },
+    {
+      "epoch": 0.386637952621826,
+      "grad_norm": 2.6847712993621826,
+      "learning_rate": 4.930094309488158e-05,
+      "loss": 0.8845,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3876396053488256,
+      "grad_norm": 2.3294968605041504,
+      "learning_rate": 4.929723631802705e-05,
+      "loss": 0.8204,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3886412580758251,
+      "grad_norm": 3.4525415897369385,
+      "learning_rate": 4.9293519879570095e-05,
+      "loss": 0.7846,
+      "step": 3880
+    },
+    {
+      "epoch": 0.38964291080282465,
+      "grad_norm": 3.896613597869873,
+      "learning_rate": 4.9289793780988526e-05,
+      "loss": 0.7868,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3906445635298242,
+      "grad_norm": 2.876314878463745,
+      "learning_rate": 4.928605802376399e-05,
+      "loss": 0.7856,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3916462162568238,
+      "grad_norm": 3.1712610721588135,
+      "learning_rate": 4.9282312609382004e-05,
+      "loss": 0.748,
+      "step": 3910
+    },
+    {
+      "epoch": 0.3926478689838233,
+      "grad_norm": 3.0286600589752197,
+      "learning_rate": 4.927855753933188e-05,
+      "loss": 0.8701,
+      "step": 3920
+    },
+    {
+      "epoch": 0.39364952171082285,
+      "grad_norm": 2.6831576824188232,
+      "learning_rate": 4.9274792815106794e-05,
+      "loss": 0.8515,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3946511744378224,
+      "grad_norm": 2.632242202758789,
+      "learning_rate": 4.9271018438203766e-05,
+      "loss": 0.8772,
+      "step": 3940
+    },
+    {
+      "epoch": 0.395652827164822,
+      "grad_norm": 3.1001410484313965,
+      "learning_rate": 4.9267234410123644e-05,
+      "loss": 0.7569,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3966544798918215,
+      "grad_norm": 3.0256540775299072,
+      "learning_rate": 4.9263440732371116e-05,
+      "loss": 0.8183,
+      "step": 3960
+    },
+    {
+      "epoch": 0.39765613261882105,
+      "grad_norm": 3.7380168437957764,
+      "learning_rate": 4.925963740645471e-05,
+      "loss": 0.8241,
+      "step": 3970
+    },
+    {
+      "epoch": 0.3986577853458206,
+      "grad_norm": 3.397099256515503,
+      "learning_rate": 4.925582443388679e-05,
+      "loss": 0.8335,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3996594380728202,
+      "grad_norm": 2.583353281021118,
+      "learning_rate": 4.925200181618354e-05,
+      "loss": 0.7761,
+      "step": 3990
+    },
+    {
+      "epoch": 0.4006610907998197,
+      "grad_norm": 2.2222886085510254,
+      "learning_rate": 4.924816955486501e-05,
+      "loss": 0.7627,
+      "step": 4000
+    },
+    {
+      "epoch": 0.40166274352681924,
+      "grad_norm": 2.9799551963806152,
+      "learning_rate": 4.9244327651455065e-05,
+      "loss": 0.8719,
+      "step": 4010
+    },
+    {
+      "epoch": 0.4026643962538188,
+      "grad_norm": 2.8081564903259277,
+      "learning_rate": 4.9240476107481405e-05,
+      "loss": 0.858,
+      "step": 4020
+    },
+    {
+      "epoch": 0.40366604898081837,
+      "grad_norm": 3.4415876865386963,
+      "learning_rate": 4.923661492447556e-05,
+      "loss": 0.828,
+      "step": 4030
+    },
+    {
+      "epoch": 0.4046677017078179,
+      "grad_norm": 3.00347900390625,
+      "learning_rate": 4.92327441039729e-05,
+      "loss": 0.7376,
+      "step": 4040
+    },
+    {
+      "epoch": 0.40566935443481744,
+      "grad_norm": 2.860119581222534,
+      "learning_rate": 4.922886364751263e-05,
+      "loss": 0.8286,
+      "step": 4050
+    },
+    {
+      "epoch": 0.406671007161817,
+      "grad_norm": 2.8311233520507812,
+      "learning_rate": 4.92249735566378e-05,
+      "loss": 0.8347,
+      "step": 4060
+    },
+    {
+      "epoch": 0.40767265988881657,
+      "grad_norm": 2.2550930976867676,
+      "learning_rate": 4.922107383289524e-05,
+      "loss": 0.8662,
+      "step": 4070
+    },
+    {
+      "epoch": 0.4086743126158161,
+      "grad_norm": 2.128415584564209,
+      "learning_rate": 4.921716447783566e-05,
+      "loss": 0.7438,
+      "step": 4080
+    },
+    {
+      "epoch": 0.40967596534281564,
+      "grad_norm": 2.728203773498535,
+      "learning_rate": 4.921324549301359e-05,
+      "loss": 0.8708,
+      "step": 4090
+    },
+    {
+      "epoch": 0.4106776180698152,
+      "grad_norm": 2.9716641902923584,
+      "learning_rate": 4.9209316879987374e-05,
+      "loss": 0.8483,
+      "step": 4100
+    },
+    {
+      "epoch": 0.41167927079681477,
+      "grad_norm": 2.1702470779418945,
+      "learning_rate": 4.92053786403192e-05,
+      "loss": 0.7749,
+      "step": 4110
+    },
+    {
+      "epoch": 0.4126809235238143,
+      "grad_norm": 2.6090810298919678,
+      "learning_rate": 4.9201430775575074e-05,
+      "loss": 0.7221,
+      "step": 4120
+    },
+    {
+      "epoch": 0.41368257625081384,
+      "grad_norm": 3.643763780593872,
+      "learning_rate": 4.9197473287324835e-05,
+      "loss": 0.8035,
+      "step": 4130
+    },
+    {
+      "epoch": 0.4146842289778134,
+      "grad_norm": 2.5547125339508057,
+      "learning_rate": 4.919350617714215e-05,
+      "loss": 0.779,
+      "step": 4140
+    },
+    {
+      "epoch": 0.41568588170481297,
+      "grad_norm": 2.8796565532684326,
+      "learning_rate": 4.9189529446604484e-05,
+      "loss": 0.8606,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4166875344318125,
+      "grad_norm": 2.778782606124878,
+      "learning_rate": 4.918554309729318e-05,
+      "loss": 0.8208,
+      "step": 4160
+    },
+    {
+      "epoch": 0.41768918715881204,
+      "grad_norm": 3.649463176727295,
+      "learning_rate": 4.9181547130793385e-05,
+      "loss": 0.8072,
+      "step": 4170
+    },
+    {
+      "epoch": 0.41869083988581157,
+      "grad_norm": 2.776176929473877,
+      "learning_rate": 4.917754154869403e-05,
+      "loss": 0.7387,
+      "step": 4180
+    },
+    {
+      "epoch": 0.41969249261281116,
+      "grad_norm": 3.0687742233276367,
+      "learning_rate": 4.917352635258794e-05,
+      "loss": 0.8521,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4206941453398107,
+      "grad_norm": 2.9265615940093994,
+      "learning_rate": 4.916950154407169e-05,
+      "loss": 0.8117,
+      "step": 4200
+    },
+    {
+      "epoch": 0.42169579806681023,
+      "grad_norm": 3.030052900314331,
+      "learning_rate": 4.916546712474573e-05,
+      "loss": 0.8334,
+      "step": 4210
+    },
+    {
+      "epoch": 0.42269745079380977,
+      "grad_norm": 3.2427847385406494,
+      "learning_rate": 4.916142309621432e-05,
+      "loss": 0.8247,
+      "step": 4220
+    },
+    {
+      "epoch": 0.42369910352080936,
+      "grad_norm": 2.7868199348449707,
+      "learning_rate": 4.9157369460085535e-05,
+      "loss": 0.8202,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4247007562478089,
+      "grad_norm": 2.5167086124420166,
+      "learning_rate": 4.915330621797126e-05,
+      "loss": 0.8389,
+      "step": 4240
+    },
+    {
+      "epoch": 0.42570240897480843,
+      "grad_norm": 2.5550220012664795,
+      "learning_rate": 4.914923337148722e-05,
+      "loss": 0.8128,
+      "step": 4250
+    },
+    {
+      "epoch": 0.42670406170180797,
+      "grad_norm": 2.4748544692993164,
+      "learning_rate": 4.9145150922252944e-05,
+      "loss": 0.826,
+      "step": 4260
+    },
+    {
+      "epoch": 0.42770571442880756,
+      "grad_norm": 3.1754608154296875,
+      "learning_rate": 4.9141058871891793e-05,
+      "loss": 0.7927,
+      "step": 4270
+    },
+    {
+      "epoch": 0.4287073671558071,
+      "grad_norm": 3.4101622104644775,
+      "learning_rate": 4.9136957222030934e-05,
+      "loss": 0.8662,
+      "step": 4280
+    },
+    {
+      "epoch": 0.42970901988280663,
+      "grad_norm": 2.1540229320526123,
+      "learning_rate": 4.9132845974301357e-05,
+      "loss": 0.7403,
+      "step": 4290
+    },
+    {
+      "epoch": 0.43071067260980617,
+      "grad_norm": 3.1899423599243164,
+      "learning_rate": 4.912872513033786e-05,
+      "loss": 0.7887,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4317123253368057,
+      "grad_norm": 2.9181227684020996,
+      "learning_rate": 4.912459469177907e-05,
+      "loss": 0.7683,
+      "step": 4310
+    },
+    {
+      "epoch": 0.4327139780638053,
+      "grad_norm": 2.837829351425171,
+      "learning_rate": 4.9120454660267426e-05,
+      "loss": 0.7659,
+      "step": 4320
+    },
+    {
+      "epoch": 0.43371563079080483,
+      "grad_norm": 2.8570311069488525,
+      "learning_rate": 4.911630503744916e-05,
+      "loss": 0.8271,
+      "step": 4330
+    },
+    {
+      "epoch": 0.43471728351780436,
+      "grad_norm": 2.584602117538452,
+      "learning_rate": 4.911214582497436e-05,
+      "loss": 0.8455,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4357189362448039,
+      "grad_norm": 3.5033724308013916,
+      "learning_rate": 4.91079770244969e-05,
+      "loss": 0.8245,
+      "step": 4350
+    },
+    {
+      "epoch": 0.4367205889718035,
+      "grad_norm": 2.954213857650757,
+      "learning_rate": 4.910379863767446e-05,
+      "loss": 0.832,
+      "step": 4360
+    },
+    {
+      "epoch": 0.437722241698803,
+      "grad_norm": 3.0140154361724854,
+      "learning_rate": 4.909961066616855e-05,
+      "loss": 0.7357,
+      "step": 4370
+    },
+    {
+      "epoch": 0.43872389442580256,
+      "grad_norm": 2.633049726486206,
+      "learning_rate": 4.909541311164448e-05,
+      "loss": 0.7821,
+      "step": 4380
+    },
+    {
+      "epoch": 0.4397255471528021,
+      "grad_norm": 3.462796688079834,
+      "learning_rate": 4.909120597577137e-05,
+      "loss": 0.759,
+      "step": 4390
+    },
+    {
+      "epoch": 0.4407271998798017,
+      "grad_norm": 2.973219633102417,
+      "learning_rate": 4.9086989260222165e-05,
+      "loss": 0.7813,
+      "step": 4400
+    },
+    {
+      "epoch": 0.4417288526068012,
+      "grad_norm": 2.684271812438965,
+      "learning_rate": 4.90827629666736e-05,
+      "loss": 0.7966,
+      "step": 4410
+    },
+    {
+      "epoch": 0.44273050533380076,
+      "grad_norm": 2.947772264480591,
+      "learning_rate": 4.9078527096806225e-05,
+      "loss": 0.7481,
+      "step": 4420
+    },
+    {
+      "epoch": 0.4437321580608003,
+      "grad_norm": 3.4349887371063232,
+      "learning_rate": 4.907428165230441e-05,
+      "loss": 0.7553,
+      "step": 4430
+    },
+    {
+      "epoch": 0.4447338107877999,
+      "grad_norm": 2.8952155113220215,
+      "learning_rate": 4.907002663485632e-05,
+      "loss": 0.7569,
+      "step": 4440
+    },
+    {
+      "epoch": 0.4457354635147994,
+      "grad_norm": 2.7803821563720703,
+      "learning_rate": 4.9065762046153914e-05,
+      "loss": 0.8073,
+      "step": 4450
+    },
+    {
+      "epoch": 0.44673711624179896,
+      "grad_norm": 3.2252511978149414,
+      "learning_rate": 4.9061487887892985e-05,
+      "loss": 0.7818,
+      "step": 4460
+    },
+    {
+      "epoch": 0.4477387689687985,
+      "grad_norm": 3.153841018676758,
+      "learning_rate": 4.905720416177312e-05,
+      "loss": 0.839,
+      "step": 4470
+    },
+    {
+      "epoch": 0.4487404216957981,
+      "grad_norm": 2.7690632343292236,
+      "learning_rate": 4.9052910869497704e-05,
+      "loss": 0.7685,
+      "step": 4480
+    },
+    {
+      "epoch": 0.4497420744227976,
+      "grad_norm": 2.3396732807159424,
+      "learning_rate": 4.904860801277392e-05,
+      "loss": 0.6467,
+      "step": 4490
+    },
+    {
+      "epoch": 0.45074372714979716,
+      "grad_norm": 2.70893931388855,
+      "learning_rate": 4.904429559331279e-05,
+      "loss": 0.7481,
+      "step": 4500
+    },
+    {
+      "epoch": 0.4517453798767967,
+      "grad_norm": 3.1398801803588867,
+      "learning_rate": 4.9039973612829094e-05,
+      "loss": 0.7929,
+      "step": 4510
+    },
+    {
+      "epoch": 0.4527470326037963,
+      "grad_norm": 2.2686514854431152,
+      "learning_rate": 4.903564207304143e-05,
+      "loss": 0.8276,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4537486853307958,
+      "grad_norm": 2.6630866527557373,
+      "learning_rate": 4.903130097567222e-05,
+      "loss": 0.7651,
+      "step": 4530
+    },
+    {
+      "epoch": 0.45475033805779536,
+      "grad_norm": 2.9019227027893066,
+      "learning_rate": 4.902695032244765e-05,
+      "loss": 0.8739,
+      "step": 4540
+    },
+    {
+      "epoch": 0.4557519907847949,
+      "grad_norm": 2.878720760345459,
+      "learning_rate": 4.9022590115097723e-05,
+      "loss": 0.8215,
+      "step": 4550
+    },
+    {
+      "epoch": 0.4567536435117945,
+      "grad_norm": 2.4830129146575928,
+      "learning_rate": 4.9018220355356246e-05,
+      "loss": 0.7917,
+      "step": 4560
+    },
+    {
+      "epoch": 0.457755296238794,
+      "grad_norm": 2.447023630142212,
+      "learning_rate": 4.901384104496083e-05,
+      "loss": 0.7182,
+      "step": 4570
+    },
+    {
+      "epoch": 0.45875694896579355,
+      "grad_norm": 2.4233357906341553,
+      "learning_rate": 4.900945218565285e-05,
+      "loss": 0.8353,
+      "step": 4580
+    },
+    {
+      "epoch": 0.4597586016927931,
+      "grad_norm": 2.6926865577697754,
+      "learning_rate": 4.900505377917751e-05,
+      "loss": 0.7566,
+      "step": 4590
+    },
+    {
+      "epoch": 0.4607602544197927,
+      "grad_norm": 3.171412229537964,
+      "learning_rate": 4.900064582728381e-05,
+      "loss": 0.8012,
+      "step": 4600
+    },
+    {
+      "epoch": 0.4617619071467922,
+      "grad_norm": 2.605421781539917,
+      "learning_rate": 4.899622833172452e-05,
+      "loss": 0.7585,
+      "step": 4610
+    },
+    {
+      "epoch": 0.46276355987379175,
+      "grad_norm": 2.7635443210601807,
+      "learning_rate": 4.899180129425625e-05,
+      "loss": 0.7919,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4637652126007913,
+      "grad_norm": 3.7377936840057373,
+      "learning_rate": 4.8987364716639346e-05,
+      "loss": 0.7324,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4647668653277909,
+      "grad_norm": 2.2805068492889404,
+      "learning_rate": 4.898291860063799e-05,
+      "loss": 0.7262,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4657685180547904,
+      "grad_norm": 2.1092188358306885,
+      "learning_rate": 4.897846294802014e-05,
+      "loss": 0.7749,
+      "step": 4650
+    },
+    {
+      "epoch": 0.46677017078178995,
+      "grad_norm": 2.417418956756592,
+      "learning_rate": 4.8973997760557566e-05,
+      "loss": 0.8886,
+      "step": 4660
+    },
+    {
+      "epoch": 0.4677718235087895,
+      "grad_norm": 2.4715187549591064,
+      "learning_rate": 4.89695230400258e-05,
+      "loss": 0.7872,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4687734762357891,
+      "grad_norm": 3.4084720611572266,
+      "learning_rate": 4.8965038788204185e-05,
+      "loss": 0.7903,
+      "step": 4680
+    },
+    {
+      "epoch": 0.4697751289627886,
+      "grad_norm": 2.9078190326690674,
+      "learning_rate": 4.8960545006875844e-05,
+      "loss": 0.7957,
+      "step": 4690
+    },
+    {
+      "epoch": 0.47077678168978815,
+      "grad_norm": 3.02535080909729,
+      "learning_rate": 4.895604169782769e-05,
+      "loss": 0.7782,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4717784344167877,
+      "grad_norm": 2.9437060356140137,
+      "learning_rate": 4.8951528862850444e-05,
+      "loss": 0.8061,
+      "step": 4710
+    },
+    {
+      "epoch": 0.4727800871437873,
+      "grad_norm": 2.7083654403686523,
+      "learning_rate": 4.894700650373858e-05,
+      "loss": 0.6672,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4737817398707868,
+      "grad_norm": 2.350159168243408,
+      "learning_rate": 4.894247462229038e-05,
+      "loss": 0.7149,
+      "step": 4730
+    },
+    {
+      "epoch": 0.47478339259778635,
+      "grad_norm": 2.328615665435791,
+      "learning_rate": 4.893793322030793e-05,
+      "loss": 0.734,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4757850453247859,
+      "grad_norm": 2.403383255004883,
+      "learning_rate": 4.8933382299597063e-05,
+      "loss": 0.7973,
+      "step": 4750
+    },
+    {
+      "epoch": 0.4767866980517855,
+      "grad_norm": 3.2092747688293457,
+      "learning_rate": 4.892882186196742e-05,
+      "loss": 0.7937,
+      "step": 4760
+    },
+    {
+      "epoch": 0.477788350778785,
+      "grad_norm": 2.68188214302063,
+      "learning_rate": 4.892425190923242e-05,
+      "loss": 0.7354,
+      "step": 4770
+    },
+    {
+      "epoch": 0.47879000350578454,
+      "grad_norm": 2.7987594604492188,
+      "learning_rate": 4.891967244320929e-05,
+      "loss": 0.8261,
+      "step": 4780
+    },
+    {
+      "epoch": 0.4797916562327841,
+      "grad_norm": 2.9749321937561035,
+      "learning_rate": 4.8915083465718984e-05,
+      "loss": 0.7946,
+      "step": 4790
+    },
+    {
+      "epoch": 0.4807933089597836,
+      "grad_norm": 2.480320692062378,
+      "learning_rate": 4.891048497858629e-05,
+      "loss": 0.7868,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4817949616867832,
+      "grad_norm": 3.2900009155273438,
+      "learning_rate": 4.8905876983639775e-05,
+      "loss": 0.7455,
+      "step": 4810
+    },
+    {
+      "epoch": 0.48279661441378274,
+      "grad_norm": 2.6279869079589844,
+      "learning_rate": 4.8901259482711744e-05,
+      "loss": 0.7786,
+      "step": 4820
+    },
+    {
+      "epoch": 0.4837982671407823,
+      "grad_norm": 2.4610986709594727,
+      "learning_rate": 4.8896632477638324e-05,
+      "loss": 0.7257,
+      "step": 4830
+    },
+    {
+      "epoch": 0.4847999198677818,
+      "grad_norm": 2.5319132804870605,
+      "learning_rate": 4.88919959702594e-05,
+      "loss": 0.8303,
+      "step": 4840
+    },
+    {
+      "epoch": 0.4858015725947814,
+      "grad_norm": 2.8233323097229004,
+      "learning_rate": 4.888734996241865e-05,
+      "loss": 0.7286,
+      "step": 4850
+    },
+    {
+      "epoch": 0.48680322532178094,
+      "grad_norm": 2.4413650035858154,
+      "learning_rate": 4.8882694455963516e-05,
+      "loss": 0.7694,
+      "step": 4860
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 3.778494358062744,
+      "learning_rate": 4.887802945274523e-05,
+      "loss": 0.77,
+      "step": 4870
+    },
+    {
+      "epoch": 0.48880653077578,
+      "grad_norm": 3.3177435398101807,
+      "learning_rate": 4.887335495461879e-05,
+      "loss": 0.8672,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4898081835027796,
+      "grad_norm": 2.7683284282684326,
+      "learning_rate": 4.886867096344296e-05,
+      "loss": 0.7497,
+      "step": 4890
+    },
+    {
+      "epoch": 0.49080983622977914,
+      "grad_norm": 3.1127943992614746,
+      "learning_rate": 4.886397748108031e-05,
+      "loss": 0.7698,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4918114889567787,
+      "grad_norm": 2.4643614292144775,
+      "learning_rate": 4.885927450939716e-05,
+      "loss": 0.8021,
+      "step": 4910
+    },
+    {
+      "epoch": 0.4928131416837782,
+      "grad_norm": 3.0037858486175537,
+      "learning_rate": 4.8854562050263614e-05,
+      "loss": 0.8111,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4938147944107778,
+      "grad_norm": 2.6661083698272705,
+      "learning_rate": 4.8849840105553536e-05,
+      "loss": 0.7426,
+      "step": 4930
+    },
+    {
+      "epoch": 0.49481644713777734,
+      "grad_norm": 2.461570978164673,
+      "learning_rate": 4.8845108677144565e-05,
+      "loss": 0.8039,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4958180998647769,
+      "grad_norm": 2.258551836013794,
+      "learning_rate": 4.884036776691813e-05,
+      "loss": 0.6808,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4968197525917764,
+      "grad_norm": 2.733013153076172,
+      "learning_rate": 4.8835617376759405e-05,
+      "loss": 0.8166,
+      "step": 4960
+    },
+    {
+      "epoch": 0.497821405318776,
+      "grad_norm": 2.9279303550720215,
+      "learning_rate": 4.8830857508557344e-05,
+      "loss": 0.7596,
+      "step": 4970
+    },
+    {
+      "epoch": 0.49882305804577554,
+      "grad_norm": 2.5595335960388184,
+      "learning_rate": 4.882608816420467e-05,
+      "loss": 0.7144,
+      "step": 4980
+    },
+    {
+      "epoch": 0.49982471077277507,
+      "grad_norm": 1.9373695850372314,
+      "learning_rate": 4.88213093455979e-05,
+      "loss": 0.7014,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5008263634997746,
+      "grad_norm": 3.3078603744506836,
+      "learning_rate": 4.8816521054637264e-05,
+      "loss": 0.7673,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5018280162267742,
+      "grad_norm": 3.2751052379608154,
+      "learning_rate": 4.8811723293226796e-05,
+      "loss": 0.7277,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5028296689537737,
+      "grad_norm": 2.650498628616333,
+      "learning_rate": 4.880691606327429e-05,
+      "loss": 0.8175,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5038313216807733,
+      "grad_norm": 3.0442254543304443,
+      "learning_rate": 4.88020993666913e-05,
+      "loss": 0.7676,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5048329744077729,
+      "grad_norm": 2.6046295166015625,
+      "learning_rate": 4.8797273205393144e-05,
+      "loss": 0.6961,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5058346271347723,
+      "grad_norm": 2.5837275981903076,
+      "learning_rate": 4.8792437581298923e-05,
+      "loss": 0.7531,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5068362798617719,
+      "grad_norm": 2.275136947631836,
+      "learning_rate": 4.8787592496331456e-05,
+      "loss": 0.7709,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5078379325887715,
+      "grad_norm": 3.156888961791992,
+      "learning_rate": 4.8782737952417376e-05,
+      "loss": 0.818,
+      "step": 5070
+    },
+    {
+      "epoch": 0.508839585315771,
+      "grad_norm": 3.180203914642334,
+      "learning_rate": 4.877787395148705e-05,
+      "loss": 0.8009,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5098412380427706,
+      "grad_norm": 2.5815541744232178,
+      "learning_rate": 4.877300049547461e-05,
+      "loss": 0.766,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5108428907697701,
+      "grad_norm": 3.7498693466186523,
+      "learning_rate": 4.876811758631793e-05,
+      "loss": 0.6966,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5118445434967697,
+      "grad_norm": 2.6118204593658447,
+      "learning_rate": 4.8763225225958686e-05,
+      "loss": 0.7633,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5128461962237693,
+      "grad_norm": 2.1438217163085938,
+      "learning_rate": 4.875832341634227e-05,
+      "loss": 0.7183,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5138478489507687,
+      "grad_norm": 3.222095012664795,
+      "learning_rate": 4.875341215941784e-05,
+      "loss": 0.7274,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5148495016777683,
+      "grad_norm": 2.6931872367858887,
+      "learning_rate": 4.874849145713833e-05,
+      "loss": 0.7493,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5158511544047679,
+      "grad_norm": 2.759004831314087,
+      "learning_rate": 4.8743561311460424e-05,
+      "loss": 0.7586,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5168528071317674,
+      "grad_norm": 2.3311071395874023,
+      "learning_rate": 4.873862172434455e-05,
+      "loss": 0.7245,
+      "step": 5160
+    },
+    {
+      "epoch": 0.517854459858767,
+      "grad_norm": 2.3808846473693848,
+      "learning_rate": 4.8733672697754884e-05,
+      "loss": 0.7359,
+      "step": 5170
+    },
+    {
+      "epoch": 0.5188561125857665,
+      "grad_norm": 2.6579818725585938,
+      "learning_rate": 4.8728714233659376e-05,
+      "loss": 0.7077,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5198577653127661,
+      "grad_norm": 2.8846435546875,
+      "learning_rate": 4.872374633402972e-05,
+      "loss": 0.7361,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5208594180397657,
+      "grad_norm": 3.370232105255127,
+      "learning_rate": 4.871876900084137e-05,
+      "loss": 0.7259,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5218610707667651,
+      "grad_norm": 2.5193967819213867,
+      "learning_rate": 4.871378223607351e-05,
+      "loss": 0.7675,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5228627234937647,
+      "grad_norm": 2.9328339099884033,
+      "learning_rate": 4.870878604170909e-05,
+      "loss": 0.8851,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5238643762207643,
+      "grad_norm": 3.101991891860962,
+      "learning_rate": 4.870378041973481e-05,
+      "loss": 0.776,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5248660289477638,
+      "grad_norm": 3.0238535404205322,
+      "learning_rate": 4.8698765372141106e-05,
+      "loss": 0.7957,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5258676816747634,
+      "grad_norm": 2.859778642654419,
+      "learning_rate": 4.8693740900922193e-05,
+      "loss": 0.7584,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5268693344017629,
+      "grad_norm": 2.8081979751586914,
+      "learning_rate": 4.868870700807599e-05,
+      "loss": 0.8051,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5278709871287625,
+      "grad_norm": 3.409069061279297,
+      "learning_rate": 4.86836636956042e-05,
+      "loss": 0.7246,
+      "step": 5270
+    },
+    {
+      "epoch": 0.528872639855762,
+      "grad_norm": 2.7888481616973877,
+      "learning_rate": 4.867861096551224e-05,
+      "loss": 0.7656,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5298742925827615,
+      "grad_norm": 2.872657299041748,
+      "learning_rate": 4.867354881980931e-05,
+      "loss": 0.8549,
+      "step": 5290
+    },
+    {
+      "epoch": 0.5308759453097611,
+      "grad_norm": 3.0871007442474365,
+      "learning_rate": 4.8668477260508304e-05,
+      "loss": 0.757,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5318775980367606,
+      "grad_norm": 2.824735641479492,
+      "learning_rate": 4.866339628962591e-05,
+      "loss": 0.8681,
+      "step": 5310
+    },
+    {
+      "epoch": 0.5328792507637602,
+      "grad_norm": 3.060173273086548,
+      "learning_rate": 4.8658305909182535e-05,
+      "loss": 0.7545,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5338809034907598,
+      "grad_norm": 2.845705509185791,
+      "learning_rate": 4.865320612120231e-05,
+      "loss": 0.766,
+      "step": 5330
+    },
+    {
+      "epoch": 0.5348825562177593,
+      "grad_norm": 2.0810883045196533,
+      "learning_rate": 4.8648096927713135e-05,
+      "loss": 0.7974,
+      "step": 5340
+    },
+    {
+      "epoch": 0.5358842089447589,
+      "grad_norm": 2.593435764312744,
+      "learning_rate": 4.864297833074665e-05,
+      "loss": 0.758,
+      "step": 5350
+    },
+    {
+      "epoch": 0.5368858616717584,
+      "grad_norm": 2.5074310302734375,
+      "learning_rate": 4.863785033233821e-05,
+      "loss": 0.7201,
+      "step": 5360
+    },
+    {
+      "epoch": 0.5378875143987579,
+      "grad_norm": 2.603663444519043,
+      "learning_rate": 4.863271293452693e-05,
+      "loss": 0.6588,
+      "step": 5370
+    },
+    {
+      "epoch": 0.5388891671257575,
+      "grad_norm": 2.7700934410095215,
+      "learning_rate": 4.862756613935565e-05,
+      "loss": 0.6865,
+      "step": 5380
+    },
+    {
+      "epoch": 0.539890819852757,
+      "grad_norm": 2.748194932937622,
+      "learning_rate": 4.862240994887096e-05,
+      "loss": 0.7789,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5408924725797566,
+      "grad_norm": 2.4352834224700928,
+      "learning_rate": 4.8617244365123174e-05,
+      "loss": 0.7766,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5418941253067562,
+      "grad_norm": 2.7042815685272217,
+      "learning_rate": 4.8612069390166344e-05,
+      "loss": 0.7305,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5428957780337557,
+      "grad_norm": 2.7817211151123047,
+      "learning_rate": 4.860688502605826e-05,
+      "loss": 0.8304,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5438974307607553,
+      "grad_norm": 2.6968953609466553,
+      "learning_rate": 4.860169127486043e-05,
+      "loss": 0.6718,
+      "step": 5430
+    },
+    {
+      "epoch": 0.5448990834877548,
+      "grad_norm": 2.5667612552642822,
+      "learning_rate": 4.859648813863813e-05,
+      "loss": 0.7263,
+      "step": 5440
+    },
+    {
+      "epoch": 0.5459007362147543,
+      "grad_norm": 2.4719300270080566,
+      "learning_rate": 4.859127561946033e-05,
+      "loss": 0.7459,
+      "step": 5450
+    },
+    {
+      "epoch": 0.5469023889417539,
+      "grad_norm": 2.9818973541259766,
+      "learning_rate": 4.858605371939976e-05,
+      "loss": 0.7407,
+      "step": 5460
+    },
+    {
+      "epoch": 0.5479040416687534,
+      "grad_norm": 2.792579412460327,
+      "learning_rate": 4.8580822440532845e-05,
+      "loss": 0.6898,
+      "step": 5470
+    },
+    {
+      "epoch": 0.548905694395753,
+      "grad_norm": 2.4535489082336426,
+      "learning_rate": 4.857558178493979e-05,
+      "loss": 0.7494,
+      "step": 5480
+    },
+    {
+      "epoch": 0.5499073471227526,
+      "grad_norm": 2.920201539993286,
+      "learning_rate": 4.857033175470448e-05,
+      "loss": 0.7872,
+      "step": 5490
+    },
+    {
+      "epoch": 0.5509089998497521,
+      "grad_norm": 2.750412702560425,
+      "learning_rate": 4.856507235191454e-05,
+      "loss": 0.784,
+      "step": 5500
+    },
+    {
+      "epoch": 0.5519106525767516,
+      "grad_norm": 3.8967442512512207,
+      "learning_rate": 4.8559803578661356e-05,
+      "loss": 0.7216,
+      "step": 5510
+    },
+    {
+      "epoch": 0.5529123053037512,
+      "grad_norm": 2.851120948791504,
+      "learning_rate": 4.855452543704e-05,
+      "loss": 0.7928,
+      "step": 5520
+    },
+    {
+      "epoch": 0.5539139580307507,
+      "grad_norm": 2.98593807220459,
+      "learning_rate": 4.8549237929149275e-05,
+      "loss": 0.8081,
+      "step": 5530
+    },
+    {
+      "epoch": 0.5549156107577503,
+      "grad_norm": 3.0606110095977783,
+      "learning_rate": 4.854394105709173e-05,
+      "loss": 0.7919,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5559172634847498,
+      "grad_norm": 2.4815075397491455,
+      "learning_rate": 4.8538634822973616e-05,
+      "loss": 0.8171,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5569189162117494,
+      "grad_norm": 3.970489740371704,
+      "learning_rate": 4.853331922890492e-05,
+      "loss": 0.7901,
+      "step": 5560
+    },
+    {
+      "epoch": 0.557920568938749,
+      "grad_norm": 3.3668770790100098,
+      "learning_rate": 4.852799427699934e-05,
+      "loss": 0.7935,
+      "step": 5570
+    },
+    {
+      "epoch": 0.5589222216657485,
+      "grad_norm": 2.571666955947876,
+      "learning_rate": 4.8522659969374303e-05,
+      "loss": 0.7554,
+      "step": 5580
+    },
+    {
+      "epoch": 0.559923874392748,
+      "grad_norm": 2.3285953998565674,
+      "learning_rate": 4.851731630815095e-05,
+      "loss": 0.7063,
+      "step": 5590
+    },
+    {
+      "epoch": 0.5609255271197476,
+      "grad_norm": 2.6997599601745605,
+      "learning_rate": 4.8511963295454156e-05,
+      "loss": 0.6977,
+      "step": 5600
+    },
+    {
+      "epoch": 0.5619271798467471,
+      "grad_norm": 3.021113395690918,
+      "learning_rate": 4.8506600933412494e-05,
+      "loss": 0.8443,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5629288325737467,
+      "grad_norm": 2.5922207832336426,
+      "learning_rate": 4.850122922415827e-05,
+      "loss": 0.7096,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5639304853007462,
+      "grad_norm": 2.5463130474090576,
+      "learning_rate": 4.84958481698275e-05,
+      "loss": 0.7234,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5649321380277458,
+      "grad_norm": 3.00300931930542,
+      "learning_rate": 4.8490457772559915e-05,
+      "loss": 0.7499,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5659337907547454,
+      "grad_norm": 1.9227794408798218,
+      "learning_rate": 4.848505803449897e-05,
+      "loss": 0.7959,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5669354434817448,
+      "grad_norm": 2.412663698196411,
+      "learning_rate": 4.847964895779181e-05,
+      "loss": 0.6528,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5679370962087444,
+      "grad_norm": 2.4437978267669678,
+      "learning_rate": 4.847423054458933e-05,
+      "loss": 0.7367,
+      "step": 5670
+    },
+    {
+      "epoch": 0.568938748935744,
+      "grad_norm": 2.260493040084839,
+      "learning_rate": 4.846880279704612e-05,
+      "loss": 0.7545,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5699404016627435,
+      "grad_norm": 2.8233468532562256,
+      "learning_rate": 4.846336571732046e-05,
+      "loss": 0.7347,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5709420543897431,
+      "grad_norm": 2.773327350616455,
+      "learning_rate": 4.845791930757438e-05,
+      "loss": 0.7908,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5719437071167426,
+      "grad_norm": 3.0723698139190674,
+      "learning_rate": 4.84524635699736e-05,
+      "loss": 0.7355,
+      "step": 5710
+    },
+    {
+      "epoch": 0.5729453598437422,
+      "grad_norm": 3.2027089595794678,
+      "learning_rate": 4.844699850668756e-05,
+      "loss": 0.792,
+      "step": 5720
+    },
+    {
+      "epoch": 0.5739470125707418,
+      "grad_norm": 2.687241554260254,
+      "learning_rate": 4.844152411988937e-05,
+      "loss": 0.6883,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5749486652977412,
+      "grad_norm": 3.0531158447265625,
+      "learning_rate": 4.8436040411755887e-05,
+      "loss": 0.7003,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5759503180247408,
+      "grad_norm": 3.0482845306396484,
+      "learning_rate": 4.8430547384467684e-05,
+      "loss": 0.701,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5769519707517403,
+      "grad_norm": 2.8414599895477295,
+      "learning_rate": 4.8425045040209e-05,
+      "loss": 0.7466,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5779536234787399,
+      "grad_norm": 2.379384994506836,
+      "learning_rate": 4.84195333811678e-05,
+      "loss": 0.7467,
+      "step": 5770
+    },
+    {
+      "epoch": 0.5789552762057395,
+      "grad_norm": 2.5822062492370605,
+      "learning_rate": 4.8414012409535755e-05,
+      "loss": 0.7882,
+      "step": 5780
+    },
+    {
+      "epoch": 0.579956928932739,
+      "grad_norm": 2.840921640396118,
+      "learning_rate": 4.840848212750824e-05,
+      "loss": 0.7207,
+      "step": 5790
+    },
+    {
+      "epoch": 0.5809585816597386,
+      "grad_norm": 2.9414560794830322,
+      "learning_rate": 4.840294253728431e-05,
+      "loss": 0.754,
+      "step": 5800
+    },
+    {
+      "epoch": 0.5819602343867382,
+      "grad_norm": 2.658170223236084,
+      "learning_rate": 4.8397393641066755e-05,
+      "loss": 0.7069,
+      "step": 5810
+    },
+    {
+      "epoch": 0.5829618871137376,
+      "grad_norm": 2.7854273319244385,
+      "learning_rate": 4.839183544106205e-05,
+      "loss": 0.7554,
+      "step": 5820
+    },
+    {
+      "epoch": 0.5839635398407372,
+      "grad_norm": 2.4555301666259766,
+      "learning_rate": 4.838626793948037e-05,
+      "loss": 0.7387,
+      "step": 5830
+    },
+    {
+      "epoch": 0.5849651925677367,
+      "grad_norm": 2.213578939437866,
+      "learning_rate": 4.838069113853557e-05,
+      "loss": 0.6798,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5859668452947363,
+      "grad_norm": 3.014767646789551,
+      "learning_rate": 4.837510504044523e-05,
+      "loss": 0.6855,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5869684980217359,
+      "grad_norm": 2.637495994567871,
+      "learning_rate": 4.836950964743063e-05,
+      "loss": 0.7603,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5879701507487354,
+      "grad_norm": 2.4447574615478516,
+      "learning_rate": 4.8363904961716726e-05,
+      "loss": 0.7416,
+      "step": 5870
+    },
+    {
+      "epoch": 0.588971803475735,
+      "grad_norm": 2.8144404888153076,
+      "learning_rate": 4.835829098553217e-05,
+      "loss": 0.7945,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5899734562027346,
+      "grad_norm": 2.2470645904541016,
+      "learning_rate": 4.8352667721109314e-05,
+      "loss": 0.6977,
+      "step": 5890
+    },
+    {
+      "epoch": 0.590975108929734,
+      "grad_norm": 4.678529739379883,
+      "learning_rate": 4.834703517068422e-05,
+      "loss": 0.695,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5919767616567336,
+      "grad_norm": 3.0645904541015625,
+      "learning_rate": 4.83413933364966e-05,
+      "loss": 0.7326,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5929784143837331,
+      "grad_norm": 3.1756017208099365,
+      "learning_rate": 4.833574222078991e-05,
+      "loss": 0.7561,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5939800671107327,
+      "grad_norm": 2.3154969215393066,
+      "learning_rate": 4.833008182581127e-05,
+      "loss": 0.7144,
+      "step": 5930
+    },
+    {
+      "epoch": 0.5949817198377323,
+      "grad_norm": 3.0473766326904297,
+      "learning_rate": 4.832441215381147e-05,
+      "loss": 0.6792,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5959833725647318,
+      "grad_norm": 2.6138970851898193,
+      "learning_rate": 4.8318733207045026e-05,
+      "loss": 0.7414,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5969850252917314,
+      "grad_norm": 2.8817873001098633,
+      "learning_rate": 4.831304498777012e-05,
+      "loss": 0.6992,
+      "step": 5960
+    },
+    {
+      "epoch": 0.597986678018731,
+      "grad_norm": 2.776374340057373,
+      "learning_rate": 4.830734749824863e-05,
+      "loss": 0.7556,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5989883307457304,
+      "grad_norm": 3.001796245574951,
+      "learning_rate": 4.830164074074612e-05,
+      "loss": 0.7315,
+      "step": 5980
+    },
+    {
+      "epoch": 0.59998998347273,
+      "grad_norm": 2.222621202468872,
+      "learning_rate": 4.8295924717531833e-05,
+      "loss": 0.6979,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6009916361997295,
+      "grad_norm": 2.8218321800231934,
+      "learning_rate": 4.82901994308787e-05,
+      "loss": 0.6922,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6019932889267291,
+      "grad_norm": 2.5931153297424316,
+      "learning_rate": 4.828446488306333e-05,
+      "loss": 0.6913,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6029949416537287,
+      "grad_norm": 2.5786752700805664,
+      "learning_rate": 4.827872107636604e-05,
+      "loss": 0.7638,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6039965943807282,
+      "grad_norm": 3.696964740753174,
+      "learning_rate": 4.8272968013070785e-05,
+      "loss": 0.726,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6049982471077278,
+      "grad_norm": 3.4040944576263428,
+      "learning_rate": 4.8267205695465236e-05,
+      "loss": 0.712,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6059998998347274,
+      "grad_norm": 2.7188923358917236,
+      "learning_rate": 4.8261434125840735e-05,
+      "loss": 0.6717,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6070015525617268,
+      "grad_norm": 2.0458145141601562,
+      "learning_rate": 4.825565330649229e-05,
+      "loss": 0.6694,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6080032052887264,
+      "grad_norm": 2.1476712226867676,
+      "learning_rate": 4.8249863239718604e-05,
+      "loss": 0.7041,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6090048580157259,
+      "grad_norm": 2.831986904144287,
+      "learning_rate": 4.824406392782206e-05,
+      "loss": 0.6909,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6100065107427255,
+      "grad_norm": 2.525928497314453,
+      "learning_rate": 4.82382553731087e-05,
+      "loss": 0.6722,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6110081634697251,
+      "grad_norm": 2.2550671100616455,
+      "learning_rate": 4.823243757788825e-05,
+      "loss": 0.7381,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6120098161967246,
+      "grad_norm": 2.4147746562957764,
+      "learning_rate": 4.822661054447411e-05,
+      "loss": 0.8251,
+      "step": 6110
+    },
+    {
+      "epoch": 0.6130114689237242,
+      "grad_norm": 2.8840737342834473,
+      "learning_rate": 4.822077427518335e-05,
+      "loss": 0.7597,
+      "step": 6120
+    },
+    {
+      "epoch": 0.6140131216507237,
+      "grad_norm": 2.4112513065338135,
+      "learning_rate": 4.821492877233672e-05,
+      "loss": 0.7331,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6150147743777232,
+      "grad_norm": 2.4079248905181885,
+      "learning_rate": 4.8209074038258636e-05,
+      "loss": 0.8029,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6160164271047228,
+      "grad_norm": 2.258690595626831,
+      "learning_rate": 4.8203210075277194e-05,
+      "loss": 0.6666,
+      "step": 6150
+    },
+    {
+      "epoch": 0.6170180798317223,
+      "grad_norm": 2.800661325454712,
+      "learning_rate": 4.819733688572414e-05,
+      "loss": 0.702,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6180197325587219,
+      "grad_norm": 2.231372833251953,
+      "learning_rate": 4.81914544719349e-05,
+      "loss": 0.6428,
+      "step": 6170
+    },
+    {
+      "epoch": 0.6190213852857215,
+      "grad_norm": 2.4702677726745605,
+      "learning_rate": 4.818556283624858e-05,
+      "loss": 0.6676,
+      "step": 6180
+    },
+    {
+      "epoch": 0.620023038012721,
+      "grad_norm": 2.390310287475586,
+      "learning_rate": 4.817966198100794e-05,
+      "loss": 0.715,
+      "step": 6190
+    },
+    {
+      "epoch": 0.6210246907397206,
+      "grad_norm": 2.378634214401245,
+      "learning_rate": 4.81737519085594e-05,
+      "loss": 0.7538,
+      "step": 6200
+    },
+    {
+      "epoch": 0.62202634346672,
+      "grad_norm": 2.6578643321990967,
+      "learning_rate": 4.816783262125306e-05,
+      "loss": 0.7171,
+      "step": 6210
+    },
+    {
+      "epoch": 0.6230279961937196,
+      "grad_norm": 2.7009196281433105,
+      "learning_rate": 4.816190412144268e-05,
+      "loss": 0.7951,
+      "step": 6220
+    },
+    {
+      "epoch": 0.6240296489207192,
+      "grad_norm": 3.3209683895111084,
+      "learning_rate": 4.8155966411485676e-05,
+      "loss": 0.7156,
+      "step": 6230
+    },
+    {
+      "epoch": 0.6250313016477187,
+      "grad_norm": 2.5062389373779297,
+      "learning_rate": 4.8150019493743125e-05,
+      "loss": 0.7081,
+      "step": 6240
+    },
+    {
+      "epoch": 0.6260329543747183,
+      "grad_norm": 2.785179376602173,
+      "learning_rate": 4.8144063370579785e-05,
+      "loss": 0.6611,
+      "step": 6250
+    },
+    {
+      "epoch": 0.6270346071017179,
+      "grad_norm": 3.0619423389434814,
+      "learning_rate": 4.8138098044364056e-05,
+      "loss": 0.7037,
+      "step": 6260
+    },
+    {
+      "epoch": 0.6280362598287174,
+      "grad_norm": 3.4450790882110596,
+      "learning_rate": 4.8132123517467995e-05,
+      "loss": 0.7167,
+      "step": 6270
+    },
+    {
+      "epoch": 0.629037912555717,
+      "grad_norm": 3.5104589462280273,
+      "learning_rate": 4.8126139792267334e-05,
+      "loss": 0.6591,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6300395652827164,
+      "grad_norm": 2.983975410461426,
+      "learning_rate": 4.812014687114145e-05,
+      "loss": 0.6971,
+      "step": 6290
+    },
+    {
+      "epoch": 0.631041218009716,
+      "grad_norm": 3.098531484603882,
+      "learning_rate": 4.811414475647337e-05,
+      "loss": 0.6375,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6320428707367156,
+      "grad_norm": 2.6577069759368896,
+      "learning_rate": 4.81081334506498e-05,
+      "loss": 0.7211,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6330445234637151,
+      "grad_norm": 2.649841547012329,
+      "learning_rate": 4.8102112956061105e-05,
+      "loss": 0.6979,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6340461761907147,
+      "grad_norm": 3.851295232772827,
+      "learning_rate": 4.809608327510125e-05,
+      "loss": 0.7051,
+      "step": 6330
+    },
+    {
+      "epoch": 0.6350478289177143,
+      "grad_norm": 2.3493716716766357,
+      "learning_rate": 4.8090044410167914e-05,
+      "loss": 0.7381,
+      "step": 6340
+    },
+    {
+      "epoch": 0.6360494816447138,
+      "grad_norm": 2.7770872116088867,
+      "learning_rate": 4.8083996363662386e-05,
+      "loss": 0.7313,
+      "step": 6350
+    },
+    {
+      "epoch": 0.6370511343717133,
+      "grad_norm": 2.8262245655059814,
+      "learning_rate": 4.8077939137989645e-05,
+      "loss": 0.736,
+      "step": 6360
+    },
+    {
+      "epoch": 0.6380527870987128,
+      "grad_norm": 2.4160664081573486,
+      "learning_rate": 4.807187273555828e-05,
+      "loss": 0.6895,
+      "step": 6370
+    },
+    {
+      "epoch": 0.6390544398257124,
+      "grad_norm": 2.7470526695251465,
+      "learning_rate": 4.8065797158780556e-05,
+      "loss": 0.7013,
+      "step": 6380
+    },
+    {
+      "epoch": 0.640056092552712,
+      "grad_norm": 2.770486354827881,
+      "learning_rate": 4.8059712410072376e-05,
+      "loss": 0.8161,
+      "step": 6390
+    },
+    {
+      "epoch": 0.6410577452797115,
+      "grad_norm": 2.5913543701171875,
+      "learning_rate": 4.80536184918533e-05,
+      "loss": 0.6284,
+      "step": 6400
+    },
+    {
+      "epoch": 0.6420593980067111,
+      "grad_norm": 2.855409860610962,
+      "learning_rate": 4.804751540654651e-05,
+      "loss": 0.695,
+      "step": 6410
+    },
+    {
+      "epoch": 0.6430610507337107,
+      "grad_norm": 3.0769102573394775,
+      "learning_rate": 4.8041403156578864e-05,
+      "loss": 0.7523,
+      "step": 6420
+    },
+    {
+      "epoch": 0.6440627034607102,
+      "grad_norm": 3.437361240386963,
+      "learning_rate": 4.803528174438084e-05,
+      "loss": 0.7152,
+      "step": 6430
+    },
+    {
+      "epoch": 0.6450643561877097,
+      "grad_norm": 3.7562596797943115,
+      "learning_rate": 4.802915117238657e-05,
+      "loss": 0.7299,
+      "step": 6440
+    },
+    {
+      "epoch": 0.6460660089147092,
+      "grad_norm": 2.92394757270813,
+      "learning_rate": 4.8023011443033835e-05,
+      "loss": 0.7394,
+      "step": 6450
+    },
+    {
+      "epoch": 0.6470676616417088,
+      "grad_norm": 2.9074056148529053,
+      "learning_rate": 4.8016862558764034e-05,
+      "loss": 0.7502,
+      "step": 6460
+    },
+    {
+      "epoch": 0.6480693143687084,
+      "grad_norm": 2.7694075107574463,
+      "learning_rate": 4.801070452202224e-05,
+      "loss": 0.7541,
+      "step": 6470
+    },
+    {
+      "epoch": 0.6490709670957079,
+      "grad_norm": 3.2710673809051514,
+      "learning_rate": 4.800453733525714e-05,
+      "loss": 0.6723,
+      "step": 6480
+    },
+    {
+      "epoch": 0.6500726198227075,
+      "grad_norm": 2.673231601715088,
+      "learning_rate": 4.7998361000921055e-05,
+      "loss": 0.7593,
+      "step": 6490
+    },
+    {
+      "epoch": 0.6510742725497071,
+      "grad_norm": 2.528330087661743,
+      "learning_rate": 4.7992175521469975e-05,
+      "loss": 0.7758,
+      "step": 6500
+    },
+    {
+      "epoch": 0.6520759252767065,
+      "grad_norm": 2.416775703430176,
+      "learning_rate": 4.798598089936349e-05,
+      "loss": 0.7003,
+      "step": 6510
+    },
+    {
+      "epoch": 0.6530775780037061,
+      "grad_norm": 3.5243756771087646,
+      "learning_rate": 4.7979777137064854e-05,
+      "loss": 0.6824,
+      "step": 6520
+    },
+    {
+      "epoch": 0.6540792307307056,
+      "grad_norm": 2.3242297172546387,
+      "learning_rate": 4.7973564237040936e-05,
+      "loss": 0.6347,
+      "step": 6530
+    },
+    {
+      "epoch": 0.6550808834577052,
+      "grad_norm": 2.6728193759918213,
+      "learning_rate": 4.7967342201762244e-05,
+      "loss": 0.6672,
+      "step": 6540
+    },
+    {
+      "epoch": 0.6560825361847048,
+      "grad_norm": 2.3104145526885986,
+      "learning_rate": 4.7961111033702933e-05,
+      "loss": 0.6489,
+      "step": 6550
+    },
+    {
+      "epoch": 0.6570841889117043,
+      "grad_norm": 2.7398927211761475,
+      "learning_rate": 4.795487073534077e-05,
+      "loss": 0.6653,
+      "step": 6560
+    },
+    {
+      "epoch": 0.6580858416387039,
+      "grad_norm": 3.224385976791382,
+      "learning_rate": 4.794862130915716e-05,
+      "loss": 0.6914,
+      "step": 6570
+    },
+    {
+      "epoch": 0.6590874943657035,
+      "grad_norm": 2.406184196472168,
+      "learning_rate": 4.794236275763714e-05,
+      "loss": 0.6799,
+      "step": 6580
+    },
+    {
+      "epoch": 0.6600891470927029,
+      "grad_norm": 3.6518681049346924,
+      "learning_rate": 4.793609508326936e-05,
+      "loss": 0.7738,
+      "step": 6590
+    },
+    {
+      "epoch": 0.6610907998197025,
+      "grad_norm": 3.039585828781128,
+      "learning_rate": 4.7929818288546136e-05,
+      "loss": 0.641,
+      "step": 6600
+    },
+    {
+      "epoch": 0.662092452546702,
+      "grad_norm": 3.637732744216919,
+      "learning_rate": 4.792353237596336e-05,
+      "loss": 0.7518,
+      "step": 6610
+    },
+    {
+      "epoch": 0.6630941052737016,
+      "grad_norm": 2.7439217567443848,
+      "learning_rate": 4.7917237348020594e-05,
+      "loss": 0.6375,
+      "step": 6620
+    },
+    {
+      "epoch": 0.6640957580007012,
+      "grad_norm": 2.8755624294281006,
+      "learning_rate": 4.7910933207220985e-05,
+      "loss": 0.6667,
+      "step": 6630
+    },
+    {
+      "epoch": 0.6650974107277007,
+      "grad_norm": 2.3648972511291504,
+      "learning_rate": 4.790461995607135e-05,
+      "loss": 0.7319,
+      "step": 6640
+    },
+    {
+      "epoch": 0.6660990634547003,
+      "grad_norm": 2.573380470275879,
+      "learning_rate": 4.789829759708209e-05,
+      "loss": 0.7848,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6671007161816999,
+      "grad_norm": 2.4021103382110596,
+      "learning_rate": 4.789196613276723e-05,
+      "loss": 0.6436,
+      "step": 6660
+    },
+    {
+      "epoch": 0.6681023689086993,
+      "grad_norm": 2.582141399383545,
+      "learning_rate": 4.7885625565644444e-05,
+      "loss": 0.7283,
+      "step": 6670
+    },
+    {
+      "epoch": 0.6691040216356989,
+      "grad_norm": 2.6833086013793945,
+      "learning_rate": 4.7879275898235e-05,
+      "loss": 0.6683,
+      "step": 6680
+    },
+    {
+      "epoch": 0.6701056743626984,
+      "grad_norm": 2.5321731567382812,
+      "learning_rate": 4.787291713306379e-05,
+      "loss": 0.7005,
+      "step": 6690
+    },
+    {
+      "epoch": 0.671107327089698,
+      "grad_norm": 2.673696279525757,
+      "learning_rate": 4.786654927265933e-05,
+      "loss": 0.7181,
+      "step": 6700
+    },
+    {
+      "epoch": 0.6721089798166976,
+      "grad_norm": 2.5562915802001953,
+      "learning_rate": 4.7860172319553753e-05,
+      "loss": 0.6797,
+      "step": 6710
+    },
+    {
+      "epoch": 0.6731106325436971,
+      "grad_norm": 2.444516897201538,
+      "learning_rate": 4.78537862762828e-05,
+      "loss": 0.6879,
+      "step": 6720
+    },
+    {
+      "epoch": 0.6741122852706967,
+      "grad_norm": 2.7865781784057617,
+      "learning_rate": 4.7847391145385834e-05,
+      "loss": 0.6981,
+      "step": 6730
+    },
+    {
+      "epoch": 0.6751139379976961,
+      "grad_norm": 2.3912670612335205,
+      "learning_rate": 4.784098692940582e-05,
+      "loss": 0.6845,
+      "step": 6740
+    },
+    {
+      "epoch": 0.6761155907246957,
+      "grad_norm": 3.1096134185791016,
+      "learning_rate": 4.7834573630889333e-05,
+      "loss": 0.7959,
+      "step": 6750
+    },
+    {
+      "epoch": 0.6771172434516953,
+      "grad_norm": 2.7713048458099365,
+      "learning_rate": 4.78281512523866e-05,
+      "loss": 0.7643,
+      "step": 6760
+    },
+    {
+      "epoch": 0.6781188961786948,
+      "grad_norm": 2.5605435371398926,
+      "learning_rate": 4.782171979645141e-05,
+      "loss": 0.6974,
+      "step": 6770
+    },
+    {
+      "epoch": 0.6791205489056944,
+      "grad_norm": 2.641727924346924,
+      "learning_rate": 4.7815279265641186e-05,
+      "loss": 0.7517,
+      "step": 6780
+    },
+    {
+      "epoch": 0.680122201632694,
+      "grad_norm": 2.469717502593994,
+      "learning_rate": 4.780882966251694e-05,
+      "loss": 0.7026,
+      "step": 6790
+    },
+    {
+      "epoch": 0.6811238543596935,
+      "grad_norm": 2.858966827392578,
+      "learning_rate": 4.7802370989643324e-05,
+      "loss": 0.7318,
+      "step": 6800
+    },
+    {
+      "epoch": 0.6821255070866931,
+      "grad_norm": 2.377610206604004,
+      "learning_rate": 4.779590324958857e-05,
+      "loss": 0.7191,
+      "step": 6810
+    },
+    {
+      "epoch": 0.6831271598136925,
+      "grad_norm": 2.691115379333496,
+      "learning_rate": 4.7789426444924525e-05,
+      "loss": 0.7618,
+      "step": 6820
+    },
+    {
+      "epoch": 0.6841288125406921,
+      "grad_norm": 2.288198947906494,
+      "learning_rate": 4.778294057822663e-05,
+      "loss": 0.7858,
+      "step": 6830
+    },
+    {
+      "epoch": 0.6851304652676917,
+      "grad_norm": 2.3845489025115967,
+      "learning_rate": 4.7776445652073944e-05,
+      "loss": 0.7049,
+      "step": 6840
+    },
+    {
+      "epoch": 0.6861321179946912,
+      "grad_norm": 2.66279673576355,
+      "learning_rate": 4.776994166904913e-05,
+      "loss": 0.7543,
+      "step": 6850
+    },
+    {
+      "epoch": 0.6871337707216908,
+      "grad_norm": 2.5185487270355225,
+      "learning_rate": 4.776342863173844e-05,
+      "loss": 0.7063,
+      "step": 6860
+    },
+    {
+      "epoch": 0.6881354234486904,
+      "grad_norm": 2.798473834991455,
+      "learning_rate": 4.775690654273172e-05,
+      "loss": 0.7993,
+      "step": 6870
+    },
+    {
+      "epoch": 0.6891370761756899,
+      "grad_norm": 4.163158893585205,
+      "learning_rate": 4.775037540462245e-05,
+      "loss": 0.6505,
+      "step": 6880
+    },
+    {
+      "epoch": 0.6901387289026895,
+      "grad_norm": 2.7145819664001465,
+      "learning_rate": 4.774383522000766e-05,
+      "loss": 0.7308,
+      "step": 6890
+    },
+    {
+      "epoch": 0.6911403816296889,
+      "grad_norm": 2.461956024169922,
+      "learning_rate": 4.7737285991488027e-05,
+      "loss": 0.6925,
+      "step": 6900
+    },
+    {
+      "epoch": 0.6921420343566885,
+      "grad_norm": 2.0583384037017822,
+      "learning_rate": 4.7730727721667776e-05,
+      "loss": 0.6763,
+      "step": 6910
+    },
+    {
+      "epoch": 0.6931436870836881,
+      "grad_norm": 3.1688992977142334,
+      "learning_rate": 4.7724160413154764e-05,
+      "loss": 0.6261,
+      "step": 6920
+    },
+    {
+      "epoch": 0.6941453398106876,
+      "grad_norm": 3.163015604019165,
+      "learning_rate": 4.771758406856043e-05,
+      "loss": 0.6836,
+      "step": 6930
+    },
+    {
+      "epoch": 0.6951469925376872,
+      "grad_norm": 2.387488842010498,
+      "learning_rate": 4.7710998690499794e-05,
+      "loss": 0.7193,
+      "step": 6940
+    },
+    {
+      "epoch": 0.6961486452646868,
+      "grad_norm": 3.6488964557647705,
+      "learning_rate": 4.770440428159149e-05,
+      "loss": 0.6738,
+      "step": 6950
+    },
+    {
+      "epoch": 0.6971502979916863,
+      "grad_norm": 2.8245675563812256,
+      "learning_rate": 4.769780084445773e-05,
+      "loss": 0.7247,
+      "step": 6960
+    },
+    {
+      "epoch": 0.6981519507186859,
+      "grad_norm": 2.353736162185669,
+      "learning_rate": 4.769118838172432e-05,
+      "loss": 0.6718,
+      "step": 6970
+    },
+    {
+      "epoch": 0.6991536034456853,
+      "grad_norm": 2.815807819366455,
+      "learning_rate": 4.7684566896020645e-05,
+      "loss": 0.7584,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7001552561726849,
+      "grad_norm": 3.2099483013153076,
+      "learning_rate": 4.767793638997969e-05,
+      "loss": 0.6991,
+      "step": 6990
+    },
+    {
+      "epoch": 0.7011569088996845,
+      "grad_norm": 2.4797520637512207,
+      "learning_rate": 4.7671296866238025e-05,
+      "loss": 0.7686,
+      "step": 7000
+    },
+    {
+      "epoch": 0.702158561626684,
+      "grad_norm": 2.62703275680542,
+      "learning_rate": 4.76646483274358e-05,
+      "loss": 0.7068,
+      "step": 7010
+    },
+    {
+      "epoch": 0.7031602143536836,
+      "grad_norm": 2.9379310607910156,
+      "learning_rate": 4.765799077621677e-05,
+      "loss": 0.696,
+      "step": 7020
+    },
+    {
+      "epoch": 0.7041618670806832,
+      "grad_norm": 2.635833740234375,
+      "learning_rate": 4.765132421522823e-05,
+      "loss": 0.6948,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7051635198076827,
+      "grad_norm": 2.5286707878112793,
+      "learning_rate": 4.7644648647121096e-05,
+      "loss": 0.6741,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7061651725346823,
+      "grad_norm": 2.699763059616089,
+      "learning_rate": 4.7637964074549865e-05,
+      "loss": 0.721,
+      "step": 7050
+    },
+    {
+      "epoch": 0.7071668252616817,
+      "grad_norm": 2.9536404609680176,
+      "learning_rate": 4.763127050017259e-05,
+      "loss": 0.7539,
+      "step": 7060
+    },
+    {
+      "epoch": 0.7081684779886813,
+      "grad_norm": 1.9956881999969482,
+      "learning_rate": 4.762456792665093e-05,
+      "loss": 0.6176,
+      "step": 7070
+    },
+    {
+      "epoch": 0.7091701307156809,
+      "grad_norm": 2.333440065383911,
+      "learning_rate": 4.761785635665009e-05,
+      "loss": 0.7214,
+      "step": 7080
+    },
+    {
+      "epoch": 0.7101717834426804,
+      "grad_norm": 2.3856465816497803,
+      "learning_rate": 4.76111357928389e-05,
+      "loss": 0.681,
+      "step": 7090
+    },
+    {
+      "epoch": 0.71117343616968,
+      "grad_norm": 2.928725242614746,
+      "learning_rate": 4.760440623788972e-05,
+      "loss": 0.738,
+      "step": 7100
+    },
+    {
+      "epoch": 0.7121750888966796,
+      "grad_norm": 2.650521993637085,
+      "learning_rate": 4.75976676944785e-05,
+      "loss": 0.6664,
+      "step": 7110
+    },
+    {
+      "epoch": 0.7131767416236791,
+      "grad_norm": 2.572051525115967,
+      "learning_rate": 4.7590920165284785e-05,
+      "loss": 0.7096,
+      "step": 7120
+    },
+    {
+      "epoch": 0.7141783943506786,
+      "grad_norm": 2.118530750274658,
+      "learning_rate": 4.758416365299166e-05,
+      "loss": 0.7019,
+      "step": 7130
+    },
+    {
+      "epoch": 0.7151800470776781,
+      "grad_norm": 2.7526917457580566,
+      "learning_rate": 4.757739816028581e-05,
+      "loss": 0.6642,
+      "step": 7140
+    },
+    {
+      "epoch": 0.7161816998046777,
+      "grad_norm": 2.8893725872039795,
+      "learning_rate": 4.757062368985748e-05,
+      "loss": 0.685,
+      "step": 7150
+    },
+    {
+      "epoch": 0.7171833525316773,
+      "grad_norm": 2.5351388454437256,
+      "learning_rate": 4.756384024440047e-05,
+      "loss": 0.6915,
+      "step": 7160
+    },
+    {
+      "epoch": 0.7181850052586768,
+      "grad_norm": 2.42851185798645,
+      "learning_rate": 4.7557047826612176e-05,
+      "loss": 0.7623,
+      "step": 7170
+    },
+    {
+      "epoch": 0.7191866579856764,
+      "grad_norm": 3.844846725463867,
+      "learning_rate": 4.7550246439193546e-05,
+      "loss": 0.7131,
+      "step": 7180
+    },
+    {
+      "epoch": 0.7201883107126759,
+      "grad_norm": 2.4313747882843018,
+      "learning_rate": 4.7543436084849094e-05,
+      "loss": 0.6911,
+      "step": 7190
+    },
+    {
+      "epoch": 0.7211899634396755,
+      "grad_norm": 2.404658317565918,
+      "learning_rate": 4.7536616766286915e-05,
+      "loss": 0.6961,
+      "step": 7200
+    },
+    {
+      "epoch": 0.722191616166675,
+      "grad_norm": 2.1430392265319824,
+      "learning_rate": 4.752978848621863e-05,
+      "loss": 0.7149,
+      "step": 7210
+    },
+    {
+      "epoch": 0.7231932688936745,
+      "grad_norm": 2.6081392765045166,
+      "learning_rate": 4.7522951247359484e-05,
+      "loss": 0.7602,
+      "step": 7220
+    },
+    {
+      "epoch": 0.7241949216206741,
+      "grad_norm": 2.632668972015381,
+      "learning_rate": 4.751610505242822e-05,
+      "loss": 0.6545,
+      "step": 7230
+    },
+    {
+      "epoch": 0.7251965743476737,
+      "grad_norm": 2.9545376300811768,
+      "learning_rate": 4.750924990414719e-05,
+      "loss": 0.7311,
+      "step": 7240
+    },
+    {
+      "epoch": 0.7261982270746732,
+      "grad_norm": 2.353555679321289,
+      "learning_rate": 4.7502385805242286e-05,
+      "loss": 0.6705,
+      "step": 7250
+    },
+    {
+      "epoch": 0.7271998798016728,
+      "grad_norm": 2.0468590259552,
+      "learning_rate": 4.749551275844297e-05,
+      "loss": 0.6565,
+      "step": 7260
+    },
+    {
+      "epoch": 0.7282015325286723,
+      "grad_norm": 2.693467855453491,
+      "learning_rate": 4.748863076648224e-05,
+      "loss": 0.7045,
+      "step": 7270
+    },
+    {
+      "epoch": 0.7292031852556718,
+      "grad_norm": 2.921161651611328,
+      "learning_rate": 4.748173983209667e-05,
+      "loss": 0.6481,
+      "step": 7280
+    },
+    {
+      "epoch": 0.7302048379826714,
+      "grad_norm": 2.5260236263275146,
+      "learning_rate": 4.74748399580264e-05,
+      "loss": 0.6944,
+      "step": 7290
+    },
+    {
+      "epoch": 0.7312064907096709,
+      "grad_norm": 2.695580244064331,
+      "learning_rate": 4.746793114701508e-05,
+      "loss": 0.7933,
+      "step": 7300
+    },
+    {
+      "epoch": 0.7322081434366705,
+      "grad_norm": 1.944222092628479,
+      "learning_rate": 4.7461013401809974e-05,
+      "loss": 0.683,
+      "step": 7310
+    },
+    {
+      "epoch": 0.7332097961636701,
+      "grad_norm": 2.3878674507141113,
+      "learning_rate": 4.745408672516184e-05,
+      "loss": 0.6057,
+      "step": 7320
+    },
+    {
+      "epoch": 0.7342114488906696,
+      "grad_norm": 2.2860217094421387,
+      "learning_rate": 4.744715111982504e-05,
+      "loss": 0.6008,
+      "step": 7330
+    },
+    {
+      "epoch": 0.7352131016176692,
+      "grad_norm": 2.850374937057495,
+      "learning_rate": 4.744020658855745e-05,
+      "loss": 0.6024,
+      "step": 7340
+    },
+    {
+      "epoch": 0.7362147543446687,
+      "grad_norm": 2.4920895099639893,
+      "learning_rate": 4.7433253134120515e-05,
+      "loss": 0.7168,
+      "step": 7350
+    },
+    {
+      "epoch": 0.7372164070716682,
+      "grad_norm": 2.360792875289917,
+      "learning_rate": 4.742629075927921e-05,
+      "loss": 0.6677,
+      "step": 7360
+    },
+    {
+      "epoch": 0.7382180597986678,
+      "grad_norm": 2.3476812839508057,
+      "learning_rate": 4.7419319466802074e-05,
+      "loss": 0.7269,
+      "step": 7370
+    },
+    {
+      "epoch": 0.7392197125256673,
+      "grad_norm": 2.4540038108825684,
+      "learning_rate": 4.7412339259461194e-05,
+      "loss": 0.7336,
+      "step": 7380
+    },
+    {
+      "epoch": 0.7402213652526669,
+      "grad_norm": 2.768291473388672,
+      "learning_rate": 4.740535014003218e-05,
+      "loss": 0.6893,
+      "step": 7390
+    },
+    {
+      "epoch": 0.7412230179796665,
+      "grad_norm": 2.2983901500701904,
+      "learning_rate": 4.73983521112942e-05,
+      "loss": 0.6626,
+      "step": 7400
+    },
+    {
+      "epoch": 0.742224670706666,
+      "grad_norm": 3.18436336517334,
+      "learning_rate": 4.739134517602998e-05,
+      "loss": 0.7355,
+      "step": 7410
+    },
+    {
+      "epoch": 0.7432263234336656,
+      "grad_norm": 2.746232748031616,
+      "learning_rate": 4.738432933702575e-05,
+      "loss": 0.6544,
+      "step": 7420
+    },
+    {
+      "epoch": 0.744227976160665,
+      "grad_norm": 2.6315596103668213,
+      "learning_rate": 4.737730459707132e-05,
+      "loss": 0.6887,
+      "step": 7430
+    },
+    {
+      "epoch": 0.7452296288876646,
+      "grad_norm": 2.708827495574951,
+      "learning_rate": 4.737027095896002e-05,
+      "loss": 0.6689,
+      "step": 7440
+    },
+    {
+      "epoch": 0.7462312816146642,
+      "grad_norm": 2.332124948501587,
+      "learning_rate": 4.73632284254887e-05,
+      "loss": 0.6914,
+      "step": 7450
+    },
+    {
+      "epoch": 0.7472329343416637,
+      "grad_norm": 2.743593692779541,
+      "learning_rate": 4.73561769994578e-05,
+      "loss": 0.7856,
+      "step": 7460
+    },
+    {
+      "epoch": 0.7482345870686633,
+      "grad_norm": 2.3313615322113037,
+      "learning_rate": 4.7349116683671247e-05,
+      "loss": 0.6917,
+      "step": 7470
+    },
+    {
+      "epoch": 0.7492362397956629,
+      "grad_norm": 2.5874178409576416,
+      "learning_rate": 4.7342047480936516e-05,
+      "loss": 0.6347,
+      "step": 7480
+    },
+    {
+      "epoch": 0.7502378925226624,
+      "grad_norm": 2.8353426456451416,
+      "learning_rate": 4.733496939406462e-05,
+      "loss": 0.7379,
+      "step": 7490
+    },
+    {
+      "epoch": 0.751239545249662,
+      "grad_norm": 3.2078945636749268,
+      "learning_rate": 4.73278824258701e-05,
+      "loss": 0.7179,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7522411979766614,
+      "grad_norm": 2.3015921115875244,
+      "learning_rate": 4.732078657917105e-05,
+      "loss": 0.745,
+      "step": 7510
+    },
+    {
+      "epoch": 0.753242850703661,
+      "grad_norm": 2.150043487548828,
+      "learning_rate": 4.7313681856789054e-05,
+      "loss": 0.5916,
+      "step": 7520
+    },
+    {
+      "epoch": 0.7542445034306606,
+      "grad_norm": 2.9892406463623047,
+      "learning_rate": 4.7306568261549264e-05,
+      "loss": 0.7063,
+      "step": 7530
+    },
+    {
+      "epoch": 0.7552461561576601,
+      "grad_norm": 2.488659143447876,
+      "learning_rate": 4.7299445796280345e-05,
+      "loss": 0.7205,
+      "step": 7540
+    },
+    {
+      "epoch": 0.7562478088846597,
+      "grad_norm": 2.3893909454345703,
+      "learning_rate": 4.729231446381448e-05,
+      "loss": 0.7307,
+      "step": 7550
+    },
+    {
+      "epoch": 0.7572494616116593,
+      "grad_norm": 2.462712049484253,
+      "learning_rate": 4.7285174266987395e-05,
+      "loss": 0.6747,
+      "step": 7560
+    },
+    {
+      "epoch": 0.7582511143386588,
+      "grad_norm": 3.0485661029815674,
+      "learning_rate": 4.727802520863832e-05,
+      "loss": 0.6703,
+      "step": 7570
+    },
+    {
+      "epoch": 0.7592527670656584,
+      "grad_norm": 3.7530717849731445,
+      "learning_rate": 4.727086729161003e-05,
+      "loss": 0.689,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7602544197926578,
+      "grad_norm": 2.519653081893921,
+      "learning_rate": 4.726370051874882e-05,
+      "loss": 0.7283,
+      "step": 7590
+    },
+    {
+      "epoch": 0.7612560725196574,
+      "grad_norm": 2.17928409576416,
+      "learning_rate": 4.725652489290449e-05,
+      "loss": 0.6723,
+      "step": 7600
+    },
+    {
+      "epoch": 0.762257725246657,
+      "grad_norm": 2.4222123622894287,
+      "learning_rate": 4.724934041693036e-05,
+      "loss": 0.6834,
+      "step": 7610
+    },
+    {
+      "epoch": 0.7632593779736565,
+      "grad_norm": 2.6158435344696045,
+      "learning_rate": 4.724214709368331e-05,
+      "loss": 0.5397,
+      "step": 7620
+    },
+    {
+      "epoch": 0.7642610307006561,
+      "grad_norm": 2.603689193725586,
+      "learning_rate": 4.723494492602368e-05,
+      "loss": 0.7108,
+      "step": 7630
+    },
+    {
+      "epoch": 0.7652626834276556,
+      "grad_norm": 2.116384744644165,
+      "learning_rate": 4.722773391681536e-05,
+      "loss": 0.6177,
+      "step": 7640
+    },
+    {
+      "epoch": 0.7662643361546552,
+      "grad_norm": 2.3073859214782715,
+      "learning_rate": 4.722051406892577e-05,
+      "loss": 0.713,
+      "step": 7650
+    },
+    {
+      "epoch": 0.7672659888816548,
+      "grad_norm": 2.8427605628967285,
+      "learning_rate": 4.7213285385225803e-05,
+      "loss": 0.7059,
+      "step": 7660
+    },
+    {
+      "epoch": 0.7682676416086542,
+      "grad_norm": 3.5749852657318115,
+      "learning_rate": 4.72060478685899e-05,
+      "loss": 0.7576,
+      "step": 7670
+    },
+    {
+      "epoch": 0.7692692943356538,
+      "grad_norm": 2.713200569152832,
+      "learning_rate": 4.7198801521895985e-05,
+      "loss": 0.7588,
+      "step": 7680
+    },
+    {
+      "epoch": 0.7702709470626534,
+      "grad_norm": 3.1419548988342285,
+      "learning_rate": 4.7191546348025526e-05,
+      "loss": 0.7121,
+      "step": 7690
+    },
+    {
+      "epoch": 0.7712725997896529,
+      "grad_norm": 2.1218645572662354,
+      "learning_rate": 4.718428234986348e-05,
+      "loss": 0.6079,
+      "step": 7700
+    },
+    {
+      "epoch": 0.7722742525166525,
+      "grad_norm": 2.9028546810150146,
+      "learning_rate": 4.717700953029833e-05,
+      "loss": 0.7087,
+      "step": 7710
+    },
+    {
+      "epoch": 0.773275905243652,
+      "grad_norm": 2.609346389770508,
+      "learning_rate": 4.7169727892222023e-05,
+      "loss": 0.6647,
+      "step": 7720
+    },
+    {
+      "epoch": 0.7742775579706516,
+      "grad_norm": 2.2379143238067627,
+      "learning_rate": 4.716243743853008e-05,
+      "loss": 0.6019,
+      "step": 7730
+    },
+    {
+      "epoch": 0.7752792106976512,
+      "grad_norm": 2.9337775707244873,
+      "learning_rate": 4.7155138172121475e-05,
+      "loss": 0.6523,
+      "step": 7740
+    },
+    {
+      "epoch": 0.7762808634246506,
+      "grad_norm": 2.9232068061828613,
+      "learning_rate": 4.7147830095898704e-05,
+      "loss": 0.7737,
+      "step": 7750
+    },
+    {
+      "epoch": 0.7772825161516502,
+      "grad_norm": 2.630415916442871,
+      "learning_rate": 4.714051321276776e-05,
+      "loss": 0.6862,
+      "step": 7760
+    },
+    {
+      "epoch": 0.7782841688786498,
+      "grad_norm": 2.902622938156128,
+      "learning_rate": 4.7133187525638156e-05,
+      "loss": 0.7157,
+      "step": 7770
+    },
+    {
+      "epoch": 0.7792858216056493,
+      "grad_norm": 2.3229422569274902,
+      "learning_rate": 4.7125853037422885e-05,
+      "loss": 0.6656,
+      "step": 7780
+    },
+    {
+      "epoch": 0.7802874743326489,
+      "grad_norm": 3.199389934539795,
+      "learning_rate": 4.711850975103844e-05,
+      "loss": 0.6862,
+      "step": 7790
+    },
+    {
+      "epoch": 0.7812891270596484,
+      "grad_norm": 2.6107218265533447,
+      "learning_rate": 4.711115766940484e-05,
+      "loss": 0.6955,
+      "step": 7800
+    },
+    {
+      "epoch": 0.782290779786648,
+      "grad_norm": 2.342616558074951,
+      "learning_rate": 4.710379679544557e-05,
+      "loss": 0.7143,
+      "step": 7810
+    },
+    {
+      "epoch": 0.7832924325136476,
+      "grad_norm": 2.832531213760376,
+      "learning_rate": 4.709642713208762e-05,
+      "loss": 0.6895,
+      "step": 7820
+    },
+    {
+      "epoch": 0.784294085240647,
+      "grad_norm": 3.0061943531036377,
+      "learning_rate": 4.7089048682261485e-05,
+      "loss": 0.6991,
+      "step": 7830
+    },
+    {
+      "epoch": 0.7852957379676466,
+      "grad_norm": 2.8207926750183105,
+      "learning_rate": 4.7081661448901136e-05,
+      "loss": 0.7095,
+      "step": 7840
+    },
+    {
+      "epoch": 0.7862973906946462,
+      "grad_norm": 2.1883671283721924,
+      "learning_rate": 4.707426543494407e-05,
+      "loss": 0.7691,
+      "step": 7850
+    },
+    {
+      "epoch": 0.7872990434216457,
+      "grad_norm": 2.096417188644409,
+      "learning_rate": 4.7066860643331226e-05,
+      "loss": 0.6957,
+      "step": 7860
+    },
+    {
+      "epoch": 0.7883006961486453,
+      "grad_norm": 2.4886560440063477,
+      "learning_rate": 4.705944707700708e-05,
+      "loss": 0.7012,
+      "step": 7870
+    },
+    {
+      "epoch": 0.7893023488756448,
+      "grad_norm": 2.392404079437256,
+      "learning_rate": 4.705202473891957e-05,
+      "loss": 0.6981,
+      "step": 7880
+    },
+    {
+      "epoch": 0.7903040016026444,
+      "grad_norm": 1.8706591129302979,
+      "learning_rate": 4.704459363202012e-05,
+      "loss": 0.7211,
+      "step": 7890
+    },
+    {
+      "epoch": 0.791305654329644,
+      "grad_norm": 2.4096221923828125,
+      "learning_rate": 4.703715375926367e-05,
+      "loss": 0.6804,
+      "step": 7900
+    },
+    {
+      "epoch": 0.7923073070566434,
+      "grad_norm": 2.4984166622161865,
+      "learning_rate": 4.7029705123608604e-05,
+      "loss": 0.6623,
+      "step": 7910
+    },
+    {
+      "epoch": 0.793308959783643,
+      "grad_norm": 2.8774752616882324,
+      "learning_rate": 4.7022247728016836e-05,
+      "loss": 0.6281,
+      "step": 7920
+    },
+    {
+      "epoch": 0.7943106125106426,
+      "grad_norm": 2.1365795135498047,
+      "learning_rate": 4.7014781575453724e-05,
+      "loss": 0.6289,
+      "step": 7930
+    },
+    {
+      "epoch": 0.7953122652376421,
+      "grad_norm": 2.5661661624908447,
+      "learning_rate": 4.700730666888813e-05,
+      "loss": 0.6251,
+      "step": 7940
+    },
+    {
+      "epoch": 0.7963139179646417,
+      "grad_norm": 3.090698480606079,
+      "learning_rate": 4.69998230112924e-05,
+      "loss": 0.7738,
+      "step": 7950
+    },
+    {
+      "epoch": 0.7973155706916412,
+      "grad_norm": 2.6969759464263916,
+      "learning_rate": 4.699233060564233e-05,
+      "loss": 0.685,
+      "step": 7960
+    },
+    {
+      "epoch": 0.7983172234186408,
+      "grad_norm": 2.0235366821289062,
+      "learning_rate": 4.698482945491722e-05,
+      "loss": 0.6514,
+      "step": 7970
+    },
+    {
+      "epoch": 0.7993188761456403,
+      "grad_norm": 3.036466360092163,
+      "learning_rate": 4.6977319562099866e-05,
+      "loss": 0.662,
+      "step": 7980
+    },
+    {
+      "epoch": 0.8003205288726398,
+      "grad_norm": 2.331421375274658,
+      "learning_rate": 4.696980093017649e-05,
+      "loss": 0.6494,
+      "step": 7990
+    },
+    {
+      "epoch": 0.8013221815996394,
+      "grad_norm": 2.20649790763855,
+      "learning_rate": 4.6962273562136826e-05,
+      "loss": 0.6467,
+      "step": 8000
+    },
+    {
+      "epoch": 0.802323834326639,
+      "grad_norm": 2.333871603012085,
+      "learning_rate": 4.6954737460974074e-05,
+      "loss": 0.6229,
+      "step": 8010
+    },
+    {
+      "epoch": 0.8033254870536385,
+      "grad_norm": 2.6986920833587646,
+      "learning_rate": 4.69471926296849e-05,
+      "loss": 0.6443,
+      "step": 8020
+    },
+    {
+      "epoch": 0.8043271397806381,
+      "grad_norm": 2.474196434020996,
+      "learning_rate": 4.6939639071269454e-05,
+      "loss": 0.6341,
+      "step": 8030
+    },
+    {
+      "epoch": 0.8053287925076376,
+      "grad_norm": 2.205517053604126,
+      "learning_rate": 4.6932076788731336e-05,
+      "loss": 0.5968,
+      "step": 8040
+    },
+    {
+      "epoch": 0.8063304452346371,
+      "grad_norm": 2.337341785430908,
+      "learning_rate": 4.6924505785077645e-05,
+      "loss": 0.6785,
+      "step": 8050
+    },
+    {
+      "epoch": 0.8073320979616367,
+      "grad_norm": 3.4576077461242676,
+      "learning_rate": 4.6916926063318914e-05,
+      "loss": 0.7103,
+      "step": 8060
+    },
+    {
+      "epoch": 0.8083337506886362,
+      "grad_norm": 2.9819705486297607,
+      "learning_rate": 4.6909337626469155e-05,
+      "loss": 0.6929,
+      "step": 8070
+    },
+    {
+      "epoch": 0.8093354034156358,
+      "grad_norm": 2.672879457473755,
+      "learning_rate": 4.6901740477545874e-05,
+      "loss": 0.6572,
+      "step": 8080
+    },
+    {
+      "epoch": 0.8103370561426354,
+      "grad_norm": 3.234699010848999,
+      "learning_rate": 4.6894134619569996e-05,
+      "loss": 0.6925,
+      "step": 8090
+    },
+    {
+      "epoch": 0.8113387088696349,
+      "grad_norm": 2.5342490673065186,
+      "learning_rate": 4.688652005556593e-05,
+      "loss": 0.6874,
+      "step": 8100
+    },
+    {
+      "epoch": 0.8123403615966345,
+      "grad_norm": 4.199151515960693,
+      "learning_rate": 4.687889678856156e-05,
+      "loss": 0.7003,
+      "step": 8110
+    },
+    {
+      "epoch": 0.813342014323634,
+      "grad_norm": 2.5348429679870605,
+      "learning_rate": 4.6871264821588214e-05,
+      "loss": 0.6502,
+      "step": 8120
+    },
+    {
+      "epoch": 0.8143436670506335,
+      "grad_norm": 2.564166307449341,
+      "learning_rate": 4.686362415768066e-05,
+      "loss": 0.7122,
+      "step": 8130
+    },
+    {
+      "epoch": 0.8153453197776331,
+      "grad_norm": 3.15704607963562,
+      "learning_rate": 4.685597479987718e-05,
+      "loss": 0.6718,
+      "step": 8140
+    },
+    {
+      "epoch": 0.8163469725046326,
+      "grad_norm": 3.303502082824707,
+      "learning_rate": 4.684831675121947e-05,
+      "loss": 0.6489,
+      "step": 8150
+    },
+    {
+      "epoch": 0.8173486252316322,
+      "grad_norm": 2.5538973808288574,
+      "learning_rate": 4.6840650014752675e-05,
+      "loss": 0.6453,
+      "step": 8160
+    },
+    {
+      "epoch": 0.8183502779586317,
+      "grad_norm": 2.57222580909729,
+      "learning_rate": 4.683297459352544e-05,
+      "loss": 0.6897,
+      "step": 8170
+    },
+    {
+      "epoch": 0.8193519306856313,
+      "grad_norm": 2.3636786937713623,
+      "learning_rate": 4.68252904905898e-05,
+      "loss": 0.6485,
+      "step": 8180
+    },
+    {
+      "epoch": 0.8203535834126309,
+      "grad_norm": 2.2532460689544678,
+      "learning_rate": 4.6817597709001314e-05,
+      "loss": 0.7237,
+      "step": 8190
+    },
+    {
+      "epoch": 0.8213552361396304,
+      "grad_norm": 2.7108314037323,
+      "learning_rate": 4.680989625181893e-05,
+      "loss": 0.6412,
+      "step": 8200
+    },
+    {
+      "epoch": 0.8223568888666299,
+      "grad_norm": 2.512598991394043,
+      "learning_rate": 4.6802186122105084e-05,
+      "loss": 0.6971,
+      "step": 8210
+    },
+    {
+      "epoch": 0.8233585415936295,
+      "grad_norm": 2.5848026275634766,
+      "learning_rate": 4.6794467322925636e-05,
+      "loss": 0.6953,
+      "step": 8220
+    },
+    {
+      "epoch": 0.824360194320629,
+      "grad_norm": 2.2770142555236816,
+      "learning_rate": 4.678673985734992e-05,
+      "loss": 0.7248,
+      "step": 8230
+    },
+    {
+      "epoch": 0.8253618470476286,
+      "grad_norm": 1.9408689737319946,
+      "learning_rate": 4.677900372845069e-05,
+      "loss": 0.6522,
+      "step": 8240
+    },
+    {
+      "epoch": 0.8263634997746281,
+      "grad_norm": 1.805517315864563,
+      "learning_rate": 4.677125893930416e-05,
+      "loss": 0.7298,
+      "step": 8250
+    },
+    {
+      "epoch": 0.8273651525016277,
+      "grad_norm": 2.5754191875457764,
+      "learning_rate": 4.676350549298998e-05,
+      "loss": 0.6288,
+      "step": 8260
+    },
+    {
+      "epoch": 0.8283668052286273,
+      "grad_norm": 2.7753493785858154,
+      "learning_rate": 4.675574339259125e-05,
+      "loss": 0.6635,
+      "step": 8270
+    },
+    {
+      "epoch": 0.8293684579556267,
+      "grad_norm": 2.473208427429199,
+      "learning_rate": 4.674797264119451e-05,
+      "loss": 0.6893,
+      "step": 8280
+    },
+    {
+      "epoch": 0.8303701106826263,
+      "grad_norm": 5.45206356048584,
+      "learning_rate": 4.674019324188973e-05,
+      "loss": 0.7061,
+      "step": 8290
+    },
+    {
+      "epoch": 0.8313717634096259,
+      "grad_norm": 2.2079737186431885,
+      "learning_rate": 4.673240519777033e-05,
+      "loss": 0.7112,
+      "step": 8300
+    },
+    {
+      "epoch": 0.8323734161366254,
+      "grad_norm": 2.702819347381592,
+      "learning_rate": 4.6724608511933166e-05,
+      "loss": 0.6834,
+      "step": 8310
+    },
+    {
+      "epoch": 0.833375068863625,
+      "grad_norm": 2.0012543201446533,
+      "learning_rate": 4.671680318747852e-05,
+      "loss": 0.6652,
+      "step": 8320
+    },
+    {
+      "epoch": 0.8343767215906245,
+      "grad_norm": 2.19185209274292,
+      "learning_rate": 4.670898922751012e-05,
+      "loss": 0.6717,
+      "step": 8330
+    },
+    {
+      "epoch": 0.8353783743176241,
+      "grad_norm": 2.6093597412109375,
+      "learning_rate": 4.670116663513514e-05,
+      "loss": 0.7139,
+      "step": 8340
+    },
+    {
+      "epoch": 0.8363800270446237,
+      "grad_norm": 2.486466884613037,
+      "learning_rate": 4.6693335413464156e-05,
+      "loss": 0.6394,
+      "step": 8350
+    },
+    {
+      "epoch": 0.8373816797716231,
+      "grad_norm": 2.484245777130127,
+      "learning_rate": 4.6685495565611196e-05,
+      "loss": 0.7008,
+      "step": 8360
+    },
+    {
+      "epoch": 0.8383833324986227,
+      "grad_norm": 2.261580228805542,
+      "learning_rate": 4.66776470946937e-05,
+      "loss": 0.7157,
+      "step": 8370
+    },
+    {
+      "epoch": 0.8393849852256223,
+      "grad_norm": 2.329270362854004,
+      "learning_rate": 4.666979000383257e-05,
+      "loss": 0.6162,
+      "step": 8380
+    },
+    {
+      "epoch": 0.8403866379526218,
+      "grad_norm": 2.7080113887786865,
+      "learning_rate": 4.666192429615211e-05,
+      "loss": 0.6352,
+      "step": 8390
+    },
+    {
+      "epoch": 0.8413882906796214,
+      "grad_norm": 3.4409689903259277,
+      "learning_rate": 4.665404997478004e-05,
+      "loss": 0.6912,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8423899434066209,
+      "grad_norm": 2.870689868927002,
+      "learning_rate": 4.6646167042847545e-05,
+      "loss": 0.6864,
+      "step": 8410
+    },
+    {
+      "epoch": 0.8433915961336205,
+      "grad_norm": 3.0154449939727783,
+      "learning_rate": 4.663827550348919e-05,
+      "loss": 0.649,
+      "step": 8420
+    },
+    {
+      "epoch": 0.8443932488606201,
+      "grad_norm": 2.5777080059051514,
+      "learning_rate": 4.663037535984299e-05,
+      "loss": 0.6965,
+      "step": 8430
+    },
+    {
+      "epoch": 0.8453949015876195,
+      "grad_norm": 2.8088622093200684,
+      "learning_rate": 4.6622466615050386e-05,
+      "loss": 0.5746,
+      "step": 8440
+    },
+    {
+      "epoch": 0.8463965543146191,
+      "grad_norm": 2.371058464050293,
+      "learning_rate": 4.6614549272256216e-05,
+      "loss": 0.6209,
+      "step": 8450
+    },
+    {
+      "epoch": 0.8473982070416187,
+      "grad_norm": 3.050855875015259,
+      "learning_rate": 4.660662333460874e-05,
+      "loss": 0.7257,
+      "step": 8460
+    },
+    {
+      "epoch": 0.8483998597686182,
+      "grad_norm": 2.714768409729004,
+      "learning_rate": 4.659868880525966e-05,
+      "loss": 0.6487,
+      "step": 8470
+    },
+    {
+      "epoch": 0.8494015124956178,
+      "grad_norm": 2.225267171859741,
+      "learning_rate": 4.659074568736407e-05,
+      "loss": 0.615,
+      "step": 8480
+    },
+    {
+      "epoch": 0.8504031652226173,
+      "grad_norm": 2.1432557106018066,
+      "learning_rate": 4.658279398408049e-05,
+      "loss": 0.6767,
+      "step": 8490
+    },
+    {
+      "epoch": 0.8514048179496169,
+      "grad_norm": 2.810725688934326,
+      "learning_rate": 4.657483369857085e-05,
+      "loss": 0.6305,
+      "step": 8500
+    },
+    {
+      "epoch": 0.8524064706766165,
+      "grad_norm": 2.351372003555298,
+      "learning_rate": 4.65668648340005e-05,
+      "loss": 0.6995,
+      "step": 8510
+    },
+    {
+      "epoch": 0.8534081234036159,
+      "grad_norm": 2.379453420639038,
+      "learning_rate": 4.6558887393538185e-05,
+      "loss": 0.6905,
+      "step": 8520
+    },
+    {
+      "epoch": 0.8544097761306155,
+      "grad_norm": 2.038651466369629,
+      "learning_rate": 4.655090138035607e-05,
+      "loss": 0.6281,
+      "step": 8530
+    },
+    {
+      "epoch": 0.8554114288576151,
+      "grad_norm": 2.4542579650878906,
+      "learning_rate": 4.654290679762975e-05,
+      "loss": 0.6665,
+      "step": 8540
+    },
+    {
+      "epoch": 0.8564130815846146,
+      "grad_norm": 2.596134662628174,
+      "learning_rate": 4.653490364853818e-05,
+      "loss": 0.6725,
+      "step": 8550
+    },
+    {
+      "epoch": 0.8574147343116142,
+      "grad_norm": 2.8084211349487305,
+      "learning_rate": 4.652689193626377e-05,
+      "loss": 0.7161,
+      "step": 8560
+    },
+    {
+      "epoch": 0.8584163870386137,
+      "grad_norm": 2.6709377765655518,
+      "learning_rate": 4.651887166399229e-05,
+      "loss": 0.7052,
+      "step": 8570
+    },
+    {
+      "epoch": 0.8594180397656133,
+      "grad_norm": 2.4577574729919434,
+      "learning_rate": 4.6510842834912966e-05,
+      "loss": 0.6728,
+      "step": 8580
+    },
+    {
+      "epoch": 0.8604196924926129,
+      "grad_norm": 3.887608766555786,
+      "learning_rate": 4.650280545221838e-05,
+      "loss": 0.577,
+      "step": 8590
+    },
+    {
+      "epoch": 0.8614213452196123,
+      "grad_norm": 2.4187510013580322,
+      "learning_rate": 4.649475951910454e-05,
+      "loss": 0.7278,
+      "step": 8600
+    },
+    {
+      "epoch": 0.8624229979466119,
+      "grad_norm": 2.964956760406494,
+      "learning_rate": 4.6486705038770836e-05,
+      "loss": 0.6633,
+      "step": 8610
+    },
+    {
+      "epoch": 0.8634246506736114,
+      "grad_norm": 3.0552287101745605,
+      "learning_rate": 4.647864201442008e-05,
+      "loss": 0.6312,
+      "step": 8620
+    },
+    {
+      "epoch": 0.864426303400611,
+      "grad_norm": 2.2803544998168945,
+      "learning_rate": 4.647057044925847e-05,
+      "loss": 0.6599,
+      "step": 8630
+    },
+    {
+      "epoch": 0.8654279561276106,
+      "grad_norm": 2.118612766265869,
+      "learning_rate": 4.646249034649559e-05,
+      "loss": 0.6478,
+      "step": 8640
+    },
+    {
+      "epoch": 0.8664296088546101,
+      "grad_norm": 2.494878053665161,
+      "learning_rate": 4.645440170934443e-05,
+      "loss": 0.7454,
+      "step": 8650
+    },
+    {
+      "epoch": 0.8674312615816097,
+      "grad_norm": 3.2278289794921875,
+      "learning_rate": 4.6446304541021384e-05,
+      "loss": 0.6687,
+      "step": 8660
+    },
+    {
+      "epoch": 0.8684329143086092,
+      "grad_norm": 3.0382726192474365,
+      "learning_rate": 4.6438198844746216e-05,
+      "loss": 0.7565,
+      "step": 8670
+    },
+    {
+      "epoch": 0.8694345670356087,
+      "grad_norm": 3.102384328842163,
+      "learning_rate": 4.643008462374209e-05,
+      "loss": 0.6778,
+      "step": 8680
+    },
+    {
+      "epoch": 0.8704362197626083,
+      "grad_norm": 3.0376174449920654,
+      "learning_rate": 4.6421961881235565e-05,
+      "loss": 0.6931,
+      "step": 8690
+    },
+    {
+      "epoch": 0.8714378724896078,
+      "grad_norm": 2.6684823036193848,
+      "learning_rate": 4.641383062045659e-05,
+      "loss": 0.602,
+      "step": 8700
+    },
+    {
+      "epoch": 0.8724395252166074,
+      "grad_norm": 2.0593066215515137,
+      "learning_rate": 4.640569084463849e-05,
+      "loss": 0.6166,
+      "step": 8710
+    },
+    {
+      "epoch": 0.873441177943607,
+      "grad_norm": 2.141124725341797,
+      "learning_rate": 4.639754255701798e-05,
+      "loss": 0.5958,
+      "step": 8720
+    },
+    {
+      "epoch": 0.8744428306706065,
+      "grad_norm": 2.7273616790771484,
+      "learning_rate": 4.638938576083517e-05,
+      "loss": 0.5938,
+      "step": 8730
+    },
+    {
+      "epoch": 0.875444483397606,
+      "grad_norm": 2.745358467102051,
+      "learning_rate": 4.638122045933353e-05,
+      "loss": 0.6457,
+      "step": 8740
+    },
+    {
+      "epoch": 0.8764461361246056,
+      "grad_norm": 2.2381041049957275,
+      "learning_rate": 4.637304665575994e-05,
+      "loss": 0.6258,
+      "step": 8750
+    },
+    {
+      "epoch": 0.8774477888516051,
+      "grad_norm": 2.4562036991119385,
+      "learning_rate": 4.6364864353364645e-05,
+      "loss": 0.6121,
+      "step": 8760
+    },
+    {
+      "epoch": 0.8784494415786047,
+      "grad_norm": 1.8846955299377441,
+      "learning_rate": 4.6356673555401274e-05,
+      "loss": 0.6452,
+      "step": 8770
+    },
+    {
+      "epoch": 0.8794510943056042,
+      "grad_norm": 1.8423177003860474,
+      "learning_rate": 4.6348474265126836e-05,
+      "loss": 0.6773,
+      "step": 8780
+    },
+    {
+      "epoch": 0.8804527470326038,
+      "grad_norm": 2.811655044555664,
+      "learning_rate": 4.63402664858017e-05,
+      "loss": 0.6714,
+      "step": 8790
+    },
+    {
+      "epoch": 0.8814543997596034,
+      "grad_norm": 2.5115389823913574,
+      "learning_rate": 4.633205022068963e-05,
+      "loss": 0.6618,
+      "step": 8800
+    },
+    {
+      "epoch": 0.8824560524866029,
+      "grad_norm": 2.644765615463257,
+      "learning_rate": 4.632382547305777e-05,
+      "loss": 0.6575,
+      "step": 8810
+    },
+    {
+      "epoch": 0.8834577052136025,
+      "grad_norm": 2.793731212615967,
+      "learning_rate": 4.6315592246176606e-05,
+      "loss": 0.6772,
+      "step": 8820
+    },
+    {
+      "epoch": 0.884459357940602,
+      "grad_norm": 1.6667630672454834,
+      "learning_rate": 4.630735054332003e-05,
+      "loss": 0.6543,
+      "step": 8830
+    },
+    {
+      "epoch": 0.8854610106676015,
+      "grad_norm": 3.318857192993164,
+      "learning_rate": 4.629910036776528e-05,
+      "loss": 0.668,
+      "step": 8840
+    },
+    {
+      "epoch": 0.8864626633946011,
+      "grad_norm": 2.305460214614868,
+      "learning_rate": 4.629084172279298e-05,
+      "loss": 0.6429,
+      "step": 8850
+    },
+    {
+      "epoch": 0.8874643161216006,
+      "grad_norm": 2.154136896133423,
+      "learning_rate": 4.628257461168711e-05,
+      "loss": 0.6032,
+      "step": 8860
+    },
+    {
+      "epoch": 0.8884659688486002,
+      "grad_norm": 2.2056803703308105,
+      "learning_rate": 4.627429903773502e-05,
+      "loss": 0.6624,
+      "step": 8870
+    },
+    {
+      "epoch": 0.8894676215755998,
+      "grad_norm": 2.4072306156158447,
+      "learning_rate": 4.626601500422743e-05,
+      "loss": 0.689,
+      "step": 8880
+    },
+    {
+      "epoch": 0.8904692743025993,
+      "grad_norm": 2.359260320663452,
+      "learning_rate": 4.6257722514458426e-05,
+      "loss": 0.6717,
+      "step": 8890
+    },
+    {
+      "epoch": 0.8914709270295988,
+      "grad_norm": 2.8342933654785156,
+      "learning_rate": 4.624942157172543e-05,
+      "loss": 0.6374,
+      "step": 8900
+    },
+    {
+      "epoch": 0.8924725797565984,
+      "grad_norm": 2.6110384464263916,
+      "learning_rate": 4.624111217932926e-05,
+      "loss": 0.6617,
+      "step": 8910
+    },
+    {
+      "epoch": 0.8934742324835979,
+      "grad_norm": 2.4755423069000244,
+      "learning_rate": 4.623279434057408e-05,
+      "loss": 0.6272,
+      "step": 8920
+    },
+    {
+      "epoch": 0.8944758852105975,
+      "grad_norm": 1.939832091331482,
+      "learning_rate": 4.6224468058767414e-05,
+      "loss": 0.6798,
+      "step": 8930
+    },
+    {
+      "epoch": 0.895477537937597,
+      "grad_norm": 2.495572566986084,
+      "learning_rate": 4.621613333722013e-05,
+      "loss": 0.6241,
+      "step": 8940
+    },
+    {
+      "epoch": 0.8964791906645966,
+      "grad_norm": 1.9067161083221436,
+      "learning_rate": 4.620779017924648e-05,
+      "loss": 0.5824,
+      "step": 8950
+    },
+    {
+      "epoch": 0.8974808433915962,
+      "grad_norm": 2.223557472229004,
+      "learning_rate": 4.619943858816403e-05,
+      "loss": 0.5958,
+      "step": 8960
+    },
+    {
+      "epoch": 0.8984824961185957,
+      "grad_norm": 2.710803747177124,
+      "learning_rate": 4.619107856729376e-05,
+      "loss": 0.6501,
+      "step": 8970
+    },
+    {
+      "epoch": 0.8994841488455952,
+      "grad_norm": 2.848140001296997,
+      "learning_rate": 4.618271011995994e-05,
+      "loss": 0.6274,
+      "step": 8980
+    },
+    {
+      "epoch": 0.9004858015725948,
+      "grad_norm": 2.5017592906951904,
+      "learning_rate": 4.617433324949021e-05,
+      "loss": 0.6688,
+      "step": 8990
+    },
+    {
+      "epoch": 0.9014874542995943,
+      "grad_norm": 2.8840415477752686,
+      "learning_rate": 4.616594795921558e-05,
+      "loss": 0.6324,
+      "step": 9000
+    },
+    {
+      "epoch": 0.9024891070265939,
+      "grad_norm": 2.0160205364227295,
+      "learning_rate": 4.61575542524704e-05,
+      "loss": 0.6435,
+      "step": 9010
+    },
+    {
+      "epoch": 0.9034907597535934,
+      "grad_norm": 2.5726449489593506,
+      "learning_rate": 4.6149152132592346e-05,
+      "loss": 0.6423,
+      "step": 9020
+    },
+    {
+      "epoch": 0.904492412480593,
+      "grad_norm": 2.614830493927002,
+      "learning_rate": 4.6140741602922466e-05,
+      "loss": 0.5871,
+      "step": 9030
+    },
+    {
+      "epoch": 0.9054940652075926,
+      "grad_norm": 2.377392530441284,
+      "learning_rate": 4.6132322666805125e-05,
+      "loss": 0.6431,
+      "step": 9040
+    },
+    {
+      "epoch": 0.906495717934592,
+      "grad_norm": 3.235520839691162,
+      "learning_rate": 4.612389532758806e-05,
+      "loss": 0.7138,
+      "step": 9050
+    },
+    {
+      "epoch": 0.9074973706615916,
+      "grad_norm": 2.785496473312378,
+      "learning_rate": 4.611545958862233e-05,
+      "loss": 0.6979,
+      "step": 9060
+    },
+    {
+      "epoch": 0.9084990233885911,
+      "grad_norm": 2.0526037216186523,
+      "learning_rate": 4.610701545326234e-05,
+      "loss": 0.6125,
+      "step": 9070
+    },
+    {
+      "epoch": 0.9095006761155907,
+      "grad_norm": 2.4618897438049316,
+      "learning_rate": 4.609856292486583e-05,
+      "loss": 0.6538,
+      "step": 9080
+    },
+    {
+      "epoch": 0.9105023288425903,
+      "grad_norm": 2.852219343185425,
+      "learning_rate": 4.6090102006793875e-05,
+      "loss": 0.5938,
+      "step": 9090
+    },
+    {
+      "epoch": 0.9115039815695898,
+      "grad_norm": 2.0149478912353516,
+      "learning_rate": 4.608163270241092e-05,
+      "loss": 0.6686,
+      "step": 9100
+    },
+    {
+      "epoch": 0.9125056342965894,
+      "grad_norm": 3.2436070442199707,
+      "learning_rate": 4.6073155015084676e-05,
+      "loss": 0.7302,
+      "step": 9110
+    },
+    {
+      "epoch": 0.913507287023589,
+      "grad_norm": 3.005521059036255,
+      "learning_rate": 4.606466894818625e-05,
+      "loss": 0.6218,
+      "step": 9120
+    },
+    {
+      "epoch": 0.9145089397505884,
+      "grad_norm": 2.209559679031372,
+      "learning_rate": 4.6056174505090066e-05,
+      "loss": 0.7016,
+      "step": 9130
+    },
+    {
+      "epoch": 0.915510592477588,
+      "grad_norm": 1.9661232233047485,
+      "learning_rate": 4.604767168917386e-05,
+      "loss": 0.6632,
+      "step": 9140
+    },
+    {
+      "epoch": 0.9165122452045875,
+      "grad_norm": 2.880694627761841,
+      "learning_rate": 4.603916050381871e-05,
+      "loss": 0.6389,
+      "step": 9150
+    },
+    {
+      "epoch": 0.9175138979315871,
+      "grad_norm": 2.2712185382843018,
+      "learning_rate": 4.603064095240902e-05,
+      "loss": 0.64,
+      "step": 9160
+    },
+    {
+      "epoch": 0.9185155506585867,
+      "grad_norm": 2.218029022216797,
+      "learning_rate": 4.6022113038332534e-05,
+      "loss": 0.5949,
+      "step": 9170
+    },
+    {
+      "epoch": 0.9195172033855862,
+      "grad_norm": 2.7205657958984375,
+      "learning_rate": 4.6013576764980293e-05,
+      "loss": 0.6492,
+      "step": 9180
+    },
+    {
+      "epoch": 0.9205188561125858,
+      "grad_norm": 2.42616605758667,
+      "learning_rate": 4.6005032135746684e-05,
+      "loss": 0.5645,
+      "step": 9190
+    },
+    {
+      "epoch": 0.9215205088395854,
+      "grad_norm": 2.109893560409546,
+      "learning_rate": 4.5996479154029406e-05,
+      "loss": 0.7036,
+      "step": 9200
+    },
+    {
+      "epoch": 0.9225221615665848,
+      "grad_norm": 2.179466485977173,
+      "learning_rate": 4.59879178232295e-05,
+      "loss": 0.6462,
+      "step": 9210
+    },
+    {
+      "epoch": 0.9235238142935844,
+      "grad_norm": 2.5310440063476562,
+      "learning_rate": 4.597934814675129e-05,
+      "loss": 0.6169,
+      "step": 9220
+    },
+    {
+      "epoch": 0.9245254670205839,
+      "grad_norm": 3.850490093231201,
+      "learning_rate": 4.597077012800245e-05,
+      "loss": 0.7126,
+      "step": 9230
+    },
+    {
+      "epoch": 0.9255271197475835,
+      "grad_norm": 1.9445955753326416,
+      "learning_rate": 4.596218377039397e-05,
+      "loss": 0.6866,
+      "step": 9240
+    },
+    {
+      "epoch": 0.9265287724745831,
+      "grad_norm": 2.2818901538848877,
+      "learning_rate": 4.595358907734013e-05,
+      "loss": 0.6533,
+      "step": 9250
+    },
+    {
+      "epoch": 0.9275304252015826,
+      "grad_norm": 3.0567245483398438,
+      "learning_rate": 4.5944986052258555e-05,
+      "loss": 0.6954,
+      "step": 9260
+    },
+    {
+      "epoch": 0.9285320779285822,
+      "grad_norm": 2.6508970260620117,
+      "learning_rate": 4.5936374698570154e-05,
+      "loss": 0.6175,
+      "step": 9270
+    },
+    {
+      "epoch": 0.9295337306555818,
+      "grad_norm": 2.2970590591430664,
+      "learning_rate": 4.5927755019699175e-05,
+      "loss": 0.5485,
+      "step": 9280
+    },
+    {
+      "epoch": 0.9305353833825812,
+      "grad_norm": 2.39508056640625,
+      "learning_rate": 4.591912701907316e-05,
+      "loss": 0.6005,
+      "step": 9290
+    },
+    {
+      "epoch": 0.9315370361095808,
+      "grad_norm": 2.541376829147339,
+      "learning_rate": 4.591049070012297e-05,
+      "loss": 0.6507,
+      "step": 9300
+    },
+    {
+      "epoch": 0.9325386888365803,
+      "grad_norm": 2.181727647781372,
+      "learning_rate": 4.590184606628276e-05,
+      "loss": 0.6837,
+      "step": 9310
+    },
+    {
+      "epoch": 0.9335403415635799,
+      "grad_norm": 2.3271026611328125,
+      "learning_rate": 4.589319312099001e-05,
+      "loss": 0.6608,
+      "step": 9320
+    },
+    {
+      "epoch": 0.9345419942905795,
+      "grad_norm": 2.447262763977051,
+      "learning_rate": 4.588453186768549e-05,
+      "loss": 0.662,
+      "step": 9330
+    },
+    {
+      "epoch": 0.935543647017579,
+      "grad_norm": 2.4956793785095215,
+      "learning_rate": 4.587586230981327e-05,
+      "loss": 0.6311,
+      "step": 9340
+    },
+    {
+      "epoch": 0.9365452997445786,
+      "grad_norm": 2.4352951049804688,
+      "learning_rate": 4.5867184450820746e-05,
+      "loss": 0.6221,
+      "step": 9350
+    },
+    {
+      "epoch": 0.9375469524715782,
+      "grad_norm": 2.322591543197632,
+      "learning_rate": 4.5858498294158594e-05,
+      "loss": 0.6063,
+      "step": 9360
+    },
+    {
+      "epoch": 0.9385486051985776,
+      "grad_norm": 3.020512819290161,
+      "learning_rate": 4.584980384328078e-05,
+      "loss": 0.6755,
+      "step": 9370
+    },
+    {
+      "epoch": 0.9395502579255772,
+      "grad_norm": 2.604862928390503,
+      "learning_rate": 4.5841101101644604e-05,
+      "loss": 0.6647,
+      "step": 9380
+    },
+    {
+      "epoch": 0.9405519106525767,
+      "grad_norm": 3.1676621437072754,
+      "learning_rate": 4.5832390072710634e-05,
+      "loss": 0.6585,
+      "step": 9390
+    },
+    {
+      "epoch": 0.9415535633795763,
+      "grad_norm": 2.4377002716064453,
+      "learning_rate": 4.582367075994274e-05,
+      "loss": 0.648,
+      "step": 9400
+    },
+    {
+      "epoch": 0.9425552161065759,
+      "grad_norm": 2.3394460678100586,
+      "learning_rate": 4.581494316680809e-05,
+      "loss": 0.5972,
+      "step": 9410
+    },
+    {
+      "epoch": 0.9435568688335754,
+      "grad_norm": 2.6361212730407715,
+      "learning_rate": 4.580620729677714e-05,
+      "loss": 0.6573,
+      "step": 9420
+    },
+    {
+      "epoch": 0.944558521560575,
+      "grad_norm": 3.2919678688049316,
+      "learning_rate": 4.5797463153323625e-05,
+      "loss": 0.6253,
+      "step": 9430
+    },
+    {
+      "epoch": 0.9455601742875746,
+      "grad_norm": 2.6798956394195557,
+      "learning_rate": 4.578871073992461e-05,
+      "loss": 0.6699,
+      "step": 9440
+    },
+    {
+      "epoch": 0.946561827014574,
+      "grad_norm": 3.6646926403045654,
+      "learning_rate": 4.577995006006042e-05,
+      "loss": 0.6509,
+      "step": 9450
+    },
+    {
+      "epoch": 0.9475634797415736,
+      "grad_norm": 2.593172788619995,
+      "learning_rate": 4.577118111721464e-05,
+      "loss": 0.6777,
+      "step": 9460
+    },
+    {
+      "epoch": 0.9485651324685731,
+      "grad_norm": 2.2215142250061035,
+      "learning_rate": 4.576240391487421e-05,
+      "loss": 0.6076,
+      "step": 9470
+    },
+    {
+      "epoch": 0.9495667851955727,
+      "grad_norm": 2.540949583053589,
+      "learning_rate": 4.575361845652928e-05,
+      "loss": 0.6286,
+      "step": 9480
+    },
+    {
+      "epoch": 0.9505684379225723,
+      "grad_norm": 2.5423927307128906,
+      "learning_rate": 4.574482474567334e-05,
+      "loss": 0.607,
+      "step": 9490
+    },
+    {
+      "epoch": 0.9515700906495718,
+      "grad_norm": 2.4569149017333984,
+      "learning_rate": 4.573602278580313e-05,
+      "loss": 0.662,
+      "step": 9500
+    },
+    {
+      "epoch": 0.9525717433765714,
+      "grad_norm": 2.495696783065796,
+      "learning_rate": 4.572721258041868e-05,
+      "loss": 0.5999,
+      "step": 9510
+    },
+    {
+      "epoch": 0.953573396103571,
+      "grad_norm": 2.62209153175354,
+      "learning_rate": 4.57183941330233e-05,
+      "loss": 0.6722,
+      "step": 9520
+    },
+    {
+      "epoch": 0.9545750488305704,
+      "grad_norm": 2.378762722015381,
+      "learning_rate": 4.5709567447123577e-05,
+      "loss": 0.6478,
+      "step": 9530
+    },
+    {
+      "epoch": 0.95557670155757,
+      "grad_norm": 4.010514259338379,
+      "learning_rate": 4.5700732526229364e-05,
+      "loss": 0.6729,
+      "step": 9540
+    },
+    {
+      "epoch": 0.9565783542845695,
+      "grad_norm": 2.489445447921753,
+      "learning_rate": 4.5691889373853806e-05,
+      "loss": 0.6039,
+      "step": 9550
+    },
+    {
+      "epoch": 0.9575800070115691,
+      "grad_norm": 2.4253251552581787,
+      "learning_rate": 4.56830379935133e-05,
+      "loss": 0.5937,
+      "step": 9560
+    },
+    {
+      "epoch": 0.9585816597385687,
+      "grad_norm": 2.035104513168335,
+      "learning_rate": 4.567417838872754e-05,
+      "loss": 0.6291,
+      "step": 9570
+    },
+    {
+      "epoch": 0.9595833124655682,
+      "grad_norm": 2.7530062198638916,
+      "learning_rate": 4.566531056301948e-05,
+      "loss": 0.6301,
+      "step": 9580
+    },
+    {
+      "epoch": 0.9605849651925678,
+      "grad_norm": 2.7969632148742676,
+      "learning_rate": 4.565643451991533e-05,
+      "loss": 0.6635,
+      "step": 9590
+    },
+    {
+      "epoch": 0.9615866179195672,
+      "grad_norm": 3.570622444152832,
+      "learning_rate": 4.564755026294457e-05,
+      "loss": 0.6416,
+      "step": 9600
+    },
+    {
+      "epoch": 0.9625882706465668,
+      "grad_norm": 2.4569382667541504,
+      "learning_rate": 4.563865779563997e-05,
+      "loss": 0.6606,
+      "step": 9610
+    },
+    {
+      "epoch": 0.9635899233735664,
+      "grad_norm": 2.543860912322998,
+      "learning_rate": 4.562975712153754e-05,
+      "loss": 0.6704,
+      "step": 9620
+    },
+    {
+      "epoch": 0.9645915761005659,
+      "grad_norm": 1.9001827239990234,
+      "learning_rate": 4.562084824417657e-05,
+      "loss": 0.6436,
+      "step": 9630
+    },
+    {
+      "epoch": 0.9655932288275655,
+      "grad_norm": 3.1932127475738525,
+      "learning_rate": 4.56119311670996e-05,
+      "loss": 0.5471,
+      "step": 9640
+    },
+    {
+      "epoch": 0.9665948815545651,
+      "grad_norm": 2.821469306945801,
+      "learning_rate": 4.560300589385243e-05,
+      "loss": 0.6969,
+      "step": 9650
+    },
+    {
+      "epoch": 0.9675965342815646,
+      "grad_norm": 2.833065986633301,
+      "learning_rate": 4.559407242798413e-05,
+      "loss": 0.6397,
+      "step": 9660
+    },
+    {
+      "epoch": 0.9685981870085641,
+      "grad_norm": 2.193622350692749,
+      "learning_rate": 4.558513077304703e-05,
+      "loss": 0.5948,
+      "step": 9670
+    },
+    {
+      "epoch": 0.9695998397355636,
+      "grad_norm": 2.1319234371185303,
+      "learning_rate": 4.557618093259668e-05,
+      "loss": 0.6201,
+      "step": 9680
+    },
+    {
+      "epoch": 0.9706014924625632,
+      "grad_norm": 2.4496283531188965,
+      "learning_rate": 4.5567222910191945e-05,
+      "loss": 0.6134,
+      "step": 9690
+    },
+    {
+      "epoch": 0.9716031451895628,
+      "grad_norm": 2.4668614864349365,
+      "learning_rate": 4.5558256709394884e-05,
+      "loss": 0.665,
+      "step": 9700
+    },
+    {
+      "epoch": 0.9726047979165623,
+      "grad_norm": 2.3747925758361816,
+      "learning_rate": 4.554928233377086e-05,
+      "loss": 0.6614,
+      "step": 9710
+    },
+    {
+      "epoch": 0.9736064506435619,
+      "grad_norm": 2.282470941543579,
+      "learning_rate": 4.5540299786888443e-05,
+      "loss": 0.6946,
+      "step": 9720
+    },
+    {
+      "epoch": 0.9746081033705615,
+      "grad_norm": 2.3198509216308594,
+      "learning_rate": 4.553130907231947e-05,
+      "loss": 0.7002,
+      "step": 9730
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 2.980468511581421,
+      "learning_rate": 4.552231019363904e-05,
+      "loss": 0.6124,
+      "step": 9740
+    },
+    {
+      "epoch": 0.9766114088245605,
+      "grad_norm": 2.8307383060455322,
+      "learning_rate": 4.551330315442549e-05,
+      "loss": 0.6939,
+      "step": 9750
+    },
+    {
+      "epoch": 0.97761306155156,
+      "grad_norm": 2.3871123790740967,
+      "learning_rate": 4.5504287958260376e-05,
+      "loss": 0.6207,
+      "step": 9760
+    },
+    {
+      "epoch": 0.9786147142785596,
+      "grad_norm": 2.3071370124816895,
+      "learning_rate": 4.5495264608728535e-05,
+      "loss": 0.6086,
+      "step": 9770
+    },
+    {
+      "epoch": 0.9796163670055592,
+      "grad_norm": 2.3521180152893066,
+      "learning_rate": 4.548623310941802e-05,
+      "loss": 0.7043,
+      "step": 9780
+    },
+    {
+      "epoch": 0.9806180197325587,
+      "grad_norm": 2.123138904571533,
+      "learning_rate": 4.547719346392015e-05,
+      "loss": 0.6656,
+      "step": 9790
+    },
+    {
+      "epoch": 0.9816196724595583,
+      "grad_norm": 2.3112244606018066,
+      "learning_rate": 4.546814567582946e-05,
+      "loss": 0.6143,
+      "step": 9800
+    },
+    {
+      "epoch": 0.9826213251865579,
+      "grad_norm": 4.398324966430664,
+      "learning_rate": 4.545908974874371e-05,
+      "loss": 0.6633,
+      "step": 9810
+    },
+    {
+      "epoch": 0.9836229779135574,
+      "grad_norm": 3.7052175998687744,
+      "learning_rate": 4.5450025686263955e-05,
+      "loss": 0.6285,
+      "step": 9820
+    },
+    {
+      "epoch": 0.9846246306405569,
+      "grad_norm": 2.8410427570343018,
+      "learning_rate": 4.544095349199442e-05,
+      "loss": 0.7186,
+      "step": 9830
+    },
+    {
+      "epoch": 0.9856262833675564,
+      "grad_norm": 2.3758535385131836,
+      "learning_rate": 4.5431873169542596e-05,
+      "loss": 0.6215,
+      "step": 9840
+    },
+    {
+      "epoch": 0.986627936094556,
+      "grad_norm": 2.0713741779327393,
+      "learning_rate": 4.542278472251921e-05,
+      "loss": 0.6362,
+      "step": 9850
+    },
+    {
+      "epoch": 0.9876295888215556,
+      "grad_norm": 2.075610637664795,
+      "learning_rate": 4.54136881545382e-05,
+      "loss": 0.5963,
+      "step": 9860
+    },
+    {
+      "epoch": 0.9886312415485551,
+      "grad_norm": 2.8744397163391113,
+      "learning_rate": 4.5404583469216756e-05,
+      "loss": 0.6866,
+      "step": 9870
+    },
+    {
+      "epoch": 0.9896328942755547,
+      "grad_norm": 2.514857769012451,
+      "learning_rate": 4.539547067017528e-05,
+      "loss": 0.6553,
+      "step": 9880
+    },
+    {
+      "epoch": 0.9906345470025543,
+      "grad_norm": 2.687924385070801,
+      "learning_rate": 4.538634976103738e-05,
+      "loss": 0.6553,
+      "step": 9890
+    },
+    {
+      "epoch": 0.9916361997295537,
+      "grad_norm": 3.2299506664276123,
+      "learning_rate": 4.5377220745429954e-05,
+      "loss": 0.6194,
+      "step": 9900
+    },
+    {
+      "epoch": 0.9926378524565533,
+      "grad_norm": 2.1460533142089844,
+      "learning_rate": 4.536808362698305e-05,
+      "loss": 0.6697,
+      "step": 9910
+    },
+    {
+      "epoch": 0.9936395051835528,
+      "grad_norm": 2.8532016277313232,
+      "learning_rate": 4.535893840932999e-05,
+      "loss": 0.6531,
+      "step": 9920
+    },
+    {
+      "epoch": 0.9946411579105524,
+      "grad_norm": 2.4520788192749023,
+      "learning_rate": 4.5349785096107275e-05,
+      "loss": 0.6417,
+      "step": 9930
+    },
+    {
+      "epoch": 0.995642810637552,
+      "grad_norm": 2.3643345832824707,
+      "learning_rate": 4.534062369095467e-05,
+      "loss": 0.6163,
+      "step": 9940
+    },
+    {
+      "epoch": 0.9966444633645515,
+      "grad_norm": 2.5428478717803955,
+      "learning_rate": 4.5331454197515126e-05,
+      "loss": 0.6432,
+      "step": 9950
+    },
+    {
+      "epoch": 0.9976461160915511,
+      "grad_norm": 2.155560255050659,
+      "learning_rate": 4.5322276619434814e-05,
+      "loss": 0.5903,
+      "step": 9960
+    },
+    {
+      "epoch": 0.9986477688185507,
+      "grad_norm": 2.393480062484741,
+      "learning_rate": 4.531309096036313e-05,
+      "loss": 0.6368,
+      "step": 9970
+    },
+    {
+      "epoch": 0.9996494215455501,
+      "grad_norm": 2.6624627113342285,
+      "learning_rate": 4.530389722395268e-05,
+      "loss": 0.5995,
+      "step": 9980
+    },
+    {
+      "epoch": 1.0006009916361998,
+      "grad_norm": 2.460616111755371,
+      "learning_rate": 4.529469541385928e-05,
+      "loss": 0.6235,
+      "step": 9990
+    },
+    {
+      "epoch": 1.0016026443631993,
+      "grad_norm": 2.41973614692688,
+      "learning_rate": 4.5285485533741946e-05,
+      "loss": 0.6751,
+      "step": 10000
+    },
+    {
+      "epoch": 1.0016026443631993,
+      "eval_bleu": 0.08263337231259318,
+      "eval_loss": 0.6337834000587463,
+      "eval_rouge1": 0.4374259000948164,
+      "eval_rouge2": 0.32409193917532053,
+      "eval_rougeL": 0.41067563698366927,
+      "eval_runtime": 106969.7121,
+      "eval_samples_per_second": 0.166,
+      "eval_steps_per_second": 0.021,
+      "eval_wer": 4.417791876035146,
+      "step": 10000
+    },
+    {
+      "epoch": 1.002604297090199,
+      "grad_norm": 2.241828680038452,
+      "learning_rate": 4.527626758726292e-05,
+      "loss": 0.5698,
+      "step": 10010
+    },
+    {
+      "epoch": 1.0036059498171983,
+      "grad_norm": 2.0964417457580566,
+      "learning_rate": 4.526704157808765e-05,
+      "loss": 0.6447,
+      "step": 10020
+    },
+    {
+      "epoch": 1.004607602544198,
+      "grad_norm": 2.1923632621765137,
+      "learning_rate": 4.525780750988479e-05,
+      "loss": 0.5937,
+      "step": 10030
+    },
+    {
+      "epoch": 1.0056092552711975,
+      "grad_norm": 2.193221092224121,
+      "learning_rate": 4.524856538632617e-05,
+      "loss": 0.5717,
+      "step": 10040
+    },
+    {
+      "epoch": 1.006610907998197,
+      "grad_norm": 1.8794662952423096,
+      "learning_rate": 4.5239315211086865e-05,
+      "loss": 0.574,
+      "step": 10050
+    },
+    {
+      "epoch": 1.0076125607251967,
+      "grad_norm": 2.7534232139587402,
+      "learning_rate": 4.523005698784514e-05,
+      "loss": 0.607,
+      "step": 10060
+    },
+    {
+      "epoch": 1.008614213452196,
+      "grad_norm": 1.9754960536956787,
+      "learning_rate": 4.522079072028243e-05,
+      "loss": 0.5808,
+      "step": 10070
+    },
+    {
+      "epoch": 1.0096158661791956,
+      "grad_norm": 2.343451499938965,
+      "learning_rate": 4.5211516412083416e-05,
+      "loss": 0.6064,
+      "step": 10080
+    },
+    {
+      "epoch": 1.0106175189061952,
+      "grad_norm": 2.7020604610443115,
+      "learning_rate": 4.520223406693594e-05,
+      "loss": 0.6037,
+      "step": 10090
+    },
+    {
+      "epoch": 1.0116191716331948,
+      "grad_norm": 2.809910774230957,
+      "learning_rate": 4.5192943688531056e-05,
+      "loss": 0.5117,
+      "step": 10100
+    },
+    {
+      "epoch": 1.0126208243601944,
+      "grad_norm": 2.667957305908203,
+      "learning_rate": 4.5183645280563e-05,
+      "loss": 0.583,
+      "step": 10110
+    },
+    {
+      "epoch": 1.0136224770871938,
+      "grad_norm": 2.3873534202575684,
+      "learning_rate": 4.517433884672924e-05,
+      "loss": 0.6762,
+      "step": 10120
+    },
+    {
+      "epoch": 1.0146241298141934,
+      "grad_norm": 2.1108696460723877,
+      "learning_rate": 4.516502439073037e-05,
+      "loss": 0.556,
+      "step": 10130
+    },
+    {
+      "epoch": 1.015625782541193,
+      "grad_norm": 2.5046160221099854,
+      "learning_rate": 4.5155701916270215e-05,
+      "loss": 0.5785,
+      "step": 10140
+    },
+    {
+      "epoch": 1.0166274352681925,
+      "grad_norm": 2.8896470069885254,
+      "learning_rate": 4.5146371427055803e-05,
+      "loss": 0.5929,
+      "step": 10150
+    },
+    {
+      "epoch": 1.0176290879951921,
+      "grad_norm": 1.979096531867981,
+      "learning_rate": 4.513703292679731e-05,
+      "loss": 0.6493,
+      "step": 10160
+    },
+    {
+      "epoch": 1.0186307407221915,
+      "grad_norm": 2.9013965129852295,
+      "learning_rate": 4.5127686419208125e-05,
+      "loss": 0.5868,
+      "step": 10170
+    },
+    {
+      "epoch": 1.019632393449191,
+      "grad_norm": 1.9182281494140625,
+      "learning_rate": 4.511833190800481e-05,
+      "loss": 0.5256,
+      "step": 10180
+    },
+    {
+      "epoch": 1.0206340461761907,
+      "grad_norm": 3.2245006561279297,
+      "learning_rate": 4.510896939690711e-05,
+      "loss": 0.6408,
+      "step": 10190
+    },
+    {
+      "epoch": 1.0216356989031903,
+      "grad_norm": 2.499907970428467,
+      "learning_rate": 4.509959888963795e-05,
+      "loss": 0.5626,
+      "step": 10200
+    },
+    {
+      "epoch": 1.0226373516301899,
+      "grad_norm": 1.939091682434082,
+      "learning_rate": 4.509022038992345e-05,
+      "loss": 0.5804,
+      "step": 10210
+    },
+    {
+      "epoch": 1.0236390043571895,
+      "grad_norm": 2.0463554859161377,
+      "learning_rate": 4.5080833901492884e-05,
+      "loss": 0.549,
+      "step": 10220
+    },
+    {
+      "epoch": 1.0246406570841888,
+      "grad_norm": 2.2921488285064697,
+      "learning_rate": 4.507143942807872e-05,
+      "loss": 0.6272,
+      "step": 10230
+    },
+    {
+      "epoch": 1.0256423098111884,
+      "grad_norm": 2.6805219650268555,
+      "learning_rate": 4.506203697341659e-05,
+      "loss": 0.6537,
+      "step": 10240
+    },
+    {
+      "epoch": 1.026643962538188,
+      "grad_norm": 2.7690038681030273,
+      "learning_rate": 4.505262654124531e-05,
+      "loss": 0.6039,
+      "step": 10250
+    },
+    {
+      "epoch": 1.0276456152651876,
+      "grad_norm": 2.6545157432556152,
+      "learning_rate": 4.504320813530687e-05,
+      "loss": 0.554,
+      "step": 10260
+    },
+    {
+      "epoch": 1.0286472679921872,
+      "grad_norm": 2.342416763305664,
+      "learning_rate": 4.5033781759346406e-05,
+      "loss": 0.6261,
+      "step": 10270
+    },
+    {
+      "epoch": 1.0296489207191866,
+      "grad_norm": 2.439129590988159,
+      "learning_rate": 4.502434741711226e-05,
+      "loss": 0.6274,
+      "step": 10280
+    },
+    {
+      "epoch": 1.0306505734461862,
+      "grad_norm": 3.0521368980407715,
+      "learning_rate": 4.501490511235591e-05,
+      "loss": 0.6079,
+      "step": 10290
+    },
+    {
+      "epoch": 1.0316522261731857,
+      "grad_norm": 2.864180326461792,
+      "learning_rate": 4.5005454848832014e-05,
+      "loss": 0.6524,
+      "step": 10300
+    },
+    {
+      "epoch": 1.0326538789001853,
+      "grad_norm": 2.0836775302886963,
+      "learning_rate": 4.4995996630298395e-05,
+      "loss": 0.5631,
+      "step": 10310
+    },
+    {
+      "epoch": 1.033655531627185,
+      "grad_norm": 1.885198712348938,
+      "learning_rate": 4.4986530460516054e-05,
+      "loss": 0.5759,
+      "step": 10320
+    },
+    {
+      "epoch": 1.0346571843541843,
+      "grad_norm": 1.9586544036865234,
+      "learning_rate": 4.497705634324912e-05,
+      "loss": 0.5971,
+      "step": 10330
+    },
+    {
+      "epoch": 1.035658837081184,
+      "grad_norm": 2.376250982284546,
+      "learning_rate": 4.49675742822649e-05,
+      "loss": 0.6532,
+      "step": 10340
+    },
+    {
+      "epoch": 1.0366604898081835,
+      "grad_norm": 2.651041269302368,
+      "learning_rate": 4.495808428133387e-05,
+      "loss": 0.658,
+      "step": 10350
+    },
+    {
+      "epoch": 1.037662142535183,
+      "grad_norm": 2.460192918777466,
+      "learning_rate": 4.494858634422965e-05,
+      "loss": 0.5724,
+      "step": 10360
+    },
+    {
+      "epoch": 1.0386637952621827,
+      "grad_norm": 2.251406669616699,
+      "learning_rate": 4.493908047472901e-05,
+      "loss": 0.5986,
+      "step": 10370
+    },
+    {
+      "epoch": 1.0396654479891823,
+      "grad_norm": 2.525148868560791,
+      "learning_rate": 4.49295666766119e-05,
+      "loss": 0.5119,
+      "step": 10380
+    },
+    {
+      "epoch": 1.0406671007161816,
+      "grad_norm": 2.9440059661865234,
+      "learning_rate": 4.492004495366139e-05,
+      "loss": 0.6081,
+      "step": 10390
+    },
+    {
+      "epoch": 1.0416687534431812,
+      "grad_norm": 2.5194528102874756,
+      "learning_rate": 4.491051530966372e-05,
+      "loss": 0.5246,
+      "step": 10400
+    },
+    {
+      "epoch": 1.0426704061701808,
+      "grad_norm": 2.512968063354492,
+      "learning_rate": 4.490097774840829e-05,
+      "loss": 0.6149,
+      "step": 10410
+    },
+    {
+      "epoch": 1.0436720588971804,
+      "grad_norm": 3.91227650642395,
+      "learning_rate": 4.489143227368763e-05,
+      "loss": 0.6073,
+      "step": 10420
+    },
+    {
+      "epoch": 1.04467371162418,
+      "grad_norm": 2.4288504123687744,
+      "learning_rate": 4.488187888929741e-05,
+      "loss": 0.6019,
+      "step": 10430
+    },
+    {
+      "epoch": 1.0456753643511794,
+      "grad_norm": 3.138383150100708,
+      "learning_rate": 4.487231759903647e-05,
+      "loss": 0.6624,
+      "step": 10440
+    },
+    {
+      "epoch": 1.046677017078179,
+      "grad_norm": 2.4413421154022217,
+      "learning_rate": 4.486274840670677e-05,
+      "loss": 0.5642,
+      "step": 10450
+    },
+    {
+      "epoch": 1.0476786698051785,
+      "grad_norm": 2.7690060138702393,
+      "learning_rate": 4.485317131611344e-05,
+      "loss": 0.6147,
+      "step": 10460
+    },
+    {
+      "epoch": 1.0486803225321781,
+      "grad_norm": 2.3034732341766357,
+      "learning_rate": 4.484358633106471e-05,
+      "loss": 0.5251,
+      "step": 10470
+    },
+    {
+      "epoch": 1.0496819752591777,
+      "grad_norm": 2.441673994064331,
+      "learning_rate": 4.483399345537199e-05,
+      "loss": 0.551,
+      "step": 10480
+    },
+    {
+      "epoch": 1.050683627986177,
+      "grad_norm": 2.078007936477661,
+      "learning_rate": 4.482439269284981e-05,
+      "loss": 0.6121,
+      "step": 10490
+    },
+    {
+      "epoch": 1.0516852807131767,
+      "grad_norm": 2.7018327713012695,
+      "learning_rate": 4.4814784047315836e-05,
+      "loss": 0.7006,
+      "step": 10500
+    },
+    {
+      "epoch": 1.0526869334401763,
+      "grad_norm": 2.6600396633148193,
+      "learning_rate": 4.480516752259086e-05,
+      "loss": 0.5935,
+      "step": 10510
+    },
+    {
+      "epoch": 1.0536885861671759,
+      "grad_norm": 2.5032167434692383,
+      "learning_rate": 4.4795543122498826e-05,
+      "loss": 0.6069,
+      "step": 10520
+    },
+    {
+      "epoch": 1.0546902388941755,
+      "grad_norm": 2.078220844268799,
+      "learning_rate": 4.478591085086681e-05,
+      "loss": 0.5766,
+      "step": 10530
+    },
+    {
+      "epoch": 1.055691891621175,
+      "grad_norm": 2.401156187057495,
+      "learning_rate": 4.477627071152498e-05,
+      "loss": 0.6457,
+      "step": 10540
+    },
+    {
+      "epoch": 1.0566935443481744,
+      "grad_norm": 2.45200514793396,
+      "learning_rate": 4.476662270830668e-05,
+      "loss": 0.6415,
+      "step": 10550
+    },
+    {
+      "epoch": 1.057695197075174,
+      "grad_norm": 2.3690593242645264,
+      "learning_rate": 4.4756966845048355e-05,
+      "loss": 0.5754,
+      "step": 10560
+    },
+    {
+      "epoch": 1.0586968498021736,
+      "grad_norm": 1.8617242574691772,
+      "learning_rate": 4.474730312558959e-05,
+      "loss": 0.5331,
+      "step": 10570
+    },
+    {
+      "epoch": 1.0596985025291732,
+      "grad_norm": 2.688246250152588,
+      "learning_rate": 4.473763155377307e-05,
+      "loss": 0.5913,
+      "step": 10580
+    },
+    {
+      "epoch": 1.0607001552561728,
+      "grad_norm": 2.441681146621704,
+      "learning_rate": 4.472795213344464e-05,
+      "loss": 0.6329,
+      "step": 10590
+    },
+    {
+      "epoch": 1.0617018079831722,
+      "grad_norm": 1.5429996252059937,
+      "learning_rate": 4.471826486845323e-05,
+      "loss": 0.5835,
+      "step": 10600
+    },
+    {
+      "epoch": 1.0627034607101717,
+      "grad_norm": 2.474956512451172,
+      "learning_rate": 4.4708569762650904e-05,
+      "loss": 0.6397,
+      "step": 10610
+    },
+    {
+      "epoch": 1.0637051134371713,
+      "grad_norm": 2.1213226318359375,
+      "learning_rate": 4.4698866819892846e-05,
+      "loss": 0.6741,
+      "step": 10620
+    },
+    {
+      "epoch": 1.064706766164171,
+      "grad_norm": 2.361389636993408,
+      "learning_rate": 4.4689156044037363e-05,
+      "loss": 0.6327,
+      "step": 10630
+    },
+    {
+      "epoch": 1.0657084188911705,
+      "grad_norm": 2.966655731201172,
+      "learning_rate": 4.4679437438945855e-05,
+      "loss": 0.6742,
+      "step": 10640
+    },
+    {
+      "epoch": 1.0667100716181699,
+      "grad_norm": 2.1844279766082764,
+      "learning_rate": 4.466971100848285e-05,
+      "loss": 0.6256,
+      "step": 10650
+    },
+    {
+      "epoch": 1.0677117243451695,
+      "grad_norm": 2.297964334487915,
+      "learning_rate": 4.465997675651599e-05,
+      "loss": 0.6274,
+      "step": 10660
+    },
+    {
+      "epoch": 1.068713377072169,
+      "grad_norm": 2.126434087753296,
+      "learning_rate": 4.465023468691601e-05,
+      "loss": 0.6694,
+      "step": 10670
+    },
+    {
+      "epoch": 1.0697150297991687,
+      "grad_norm": 2.612492799758911,
+      "learning_rate": 4.464048480355679e-05,
+      "loss": 0.5821,
+      "step": 10680
+    },
+    {
+      "epoch": 1.0707166825261683,
+      "grad_norm": 2.2019944190979004,
+      "learning_rate": 4.463072711031529e-05,
+      "loss": 0.6195,
+      "step": 10690
+    },
+    {
+      "epoch": 1.0717183352531676,
+      "grad_norm": 2.5340359210968018,
+      "learning_rate": 4.462096161107154e-05,
+      "loss": 0.67,
+      "step": 10700
+    },
+    {
+      "epoch": 1.0727199879801672,
+      "grad_norm": 2.8773465156555176,
+      "learning_rate": 4.4611188309708766e-05,
+      "loss": 0.5325,
+      "step": 10710
+    },
+    {
+      "epoch": 1.0737216407071668,
+      "grad_norm": 2.5338494777679443,
+      "learning_rate": 4.460140721011321e-05,
+      "loss": 0.674,
+      "step": 10720
+    },
+    {
+      "epoch": 1.0747232934341664,
+      "grad_norm": 2.4730725288391113,
+      "learning_rate": 4.459161831617426e-05,
+      "loss": 0.5787,
+      "step": 10730
+    },
+    {
+      "epoch": 1.075724946161166,
+      "grad_norm": 1.9959347248077393,
+      "learning_rate": 4.458182163178439e-05,
+      "loss": 0.5417,
+      "step": 10740
+    },
+    {
+      "epoch": 1.0767265988881656,
+      "grad_norm": 2.4486429691314697,
+      "learning_rate": 4.4572017160839176e-05,
+      "loss": 0.6154,
+      "step": 10750
+    },
+    {
+      "epoch": 1.077728251615165,
+      "grad_norm": 2.2543139457702637,
+      "learning_rate": 4.4562204907237274e-05,
+      "loss": 0.5951,
+      "step": 10760
+    },
+    {
+      "epoch": 1.0787299043421645,
+      "grad_norm": 4.040517807006836,
+      "learning_rate": 4.455238487488047e-05,
+      "loss": 0.5494,
+      "step": 10770
+    },
+    {
+      "epoch": 1.0797315570691641,
+      "grad_norm": 2.089376926422119,
+      "learning_rate": 4.454255706767361e-05,
+      "loss": 0.5377,
+      "step": 10780
+    },
+    {
+      "epoch": 1.0807332097961637,
+      "grad_norm": 2.3366119861602783,
+      "learning_rate": 4.453272148952464e-05,
+      "loss": 0.6139,
+      "step": 10790
+    },
+    {
+      "epoch": 1.0817348625231633,
+      "grad_norm": 2.543785333633423,
+      "learning_rate": 4.4522878144344606e-05,
+      "loss": 0.5641,
+      "step": 10800
+    },
+    {
+      "epoch": 1.0827365152501627,
+      "grad_norm": 2.731062173843384,
+      "learning_rate": 4.451302703604763e-05,
+      "loss": 0.6759,
+      "step": 10810
+    },
+    {
+      "epoch": 1.0837381679771623,
+      "grad_norm": 2.7335755825042725,
+      "learning_rate": 4.4503168168550934e-05,
+      "loss": 0.6713,
+      "step": 10820
+    },
+    {
+      "epoch": 1.0847398207041619,
+      "grad_norm": 2.4380624294281006,
+      "learning_rate": 4.449330154577481e-05,
+      "loss": 0.6745,
+      "step": 10830
+    },
+    {
+      "epoch": 1.0857414734311615,
+      "grad_norm": 2.1496427059173584,
+      "learning_rate": 4.4483427171642643e-05,
+      "loss": 0.6305,
+      "step": 10840
+    },
+    {
+      "epoch": 1.086743126158161,
+      "grad_norm": 2.2780048847198486,
+      "learning_rate": 4.4473545050080915e-05,
+      "loss": 0.5717,
+      "step": 10850
+    },
+    {
+      "epoch": 1.0877447788851604,
+      "grad_norm": 2.7438337802886963,
+      "learning_rate": 4.446365518501915e-05,
+      "loss": 0.5682,
+      "step": 10860
+    },
+    {
+      "epoch": 1.08874643161216,
+      "grad_norm": 2.0671546459198,
+      "learning_rate": 4.4453757580389984e-05,
+      "loss": 0.6222,
+      "step": 10870
+    },
+    {
+      "epoch": 1.0897480843391596,
+      "grad_norm": 2.4361698627471924,
+      "learning_rate": 4.444385224012912e-05,
+      "loss": 0.5047,
+      "step": 10880
+    },
+    {
+      "epoch": 1.0907497370661592,
+      "grad_norm": 2.4023778438568115,
+      "learning_rate": 4.443393916817535e-05,
+      "loss": 0.5796,
+      "step": 10890
+    },
+    {
+      "epoch": 1.0917513897931588,
+      "grad_norm": 2.4326682090759277,
+      "learning_rate": 4.442401836847051e-05,
+      "loss": 0.5218,
+      "step": 10900
+    },
+    {
+      "epoch": 1.0927530425201581,
+      "grad_norm": 1.928910493850708,
+      "learning_rate": 4.441408984495953e-05,
+      "loss": 0.5586,
+      "step": 10910
+    },
+    {
+      "epoch": 1.0937546952471577,
+      "grad_norm": 2.215165138244629,
+      "learning_rate": 4.4404153601590415e-05,
+      "loss": 0.5602,
+      "step": 10920
+    },
+    {
+      "epoch": 1.0947563479741573,
+      "grad_norm": 2.9248368740081787,
+      "learning_rate": 4.439420964231422e-05,
+      "loss": 0.6545,
+      "step": 10930
+    },
+    {
+      "epoch": 1.095758000701157,
+      "grad_norm": 2.902667999267578,
+      "learning_rate": 4.43842579710851e-05,
+      "loss": 0.5476,
+      "step": 10940
+    },
+    {
+      "epoch": 1.0967596534281565,
+      "grad_norm": 2.431466579437256,
+      "learning_rate": 4.437429859186025e-05,
+      "loss": 0.5683,
+      "step": 10950
+    },
+    {
+      "epoch": 1.097761306155156,
+      "grad_norm": 1.8508896827697754,
+      "learning_rate": 4.436433150859993e-05,
+      "loss": 0.5298,
+      "step": 10960
+    },
+    {
+      "epoch": 1.0987629588821555,
+      "grad_norm": 2.972052812576294,
+      "learning_rate": 4.4354356725267485e-05,
+      "loss": 0.6163,
+      "step": 10970
+    },
+    {
+      "epoch": 1.099764611609155,
+      "grad_norm": 3.3024284839630127,
+      "learning_rate": 4.4344374245829294e-05,
+      "loss": 0.5949,
+      "step": 10980
+    },
+    {
+      "epoch": 1.1007662643361547,
+      "grad_norm": 2.4315567016601562,
+      "learning_rate": 4.433438407425482e-05,
+      "loss": 0.5789,
+      "step": 10990
+    },
+    {
+      "epoch": 1.1017679170631542,
+      "grad_norm": 2.4305036067962646,
+      "learning_rate": 4.4324386214516576e-05,
+      "loss": 0.5941,
+      "step": 11000
+    },
+    {
+      "epoch": 1.1027695697901538,
+      "grad_norm": 2.330115795135498,
+      "learning_rate": 4.4314380670590125e-05,
+      "loss": 0.6021,
+      "step": 11010
+    },
+    {
+      "epoch": 1.1037712225171532,
+      "grad_norm": 2.429572582244873,
+      "learning_rate": 4.4304367446454084e-05,
+      "loss": 0.5992,
+      "step": 11020
+    },
+    {
+      "epoch": 1.1047728752441528,
+      "grad_norm": 2.6118149757385254,
+      "learning_rate": 4.429434654609016e-05,
+      "loss": 0.5482,
+      "step": 11030
+    },
+    {
+      "epoch": 1.1057745279711524,
+      "grad_norm": 2.079375982284546,
+      "learning_rate": 4.428431797348306e-05,
+      "loss": 0.6004,
+      "step": 11040
+    },
+    {
+      "epoch": 1.106776180698152,
+      "grad_norm": 2.0178675651550293,
+      "learning_rate": 4.4274281732620574e-05,
+      "loss": 0.5709,
+      "step": 11050
+    },
+    {
+      "epoch": 1.1077778334251516,
+      "grad_norm": 2.3021092414855957,
+      "learning_rate": 4.426423782749352e-05,
+      "loss": 0.5602,
+      "step": 11060
+    },
+    {
+      "epoch": 1.1087794861521512,
+      "grad_norm": 2.4679481983184814,
+      "learning_rate": 4.42541862620958e-05,
+      "loss": 0.6025,
+      "step": 11070
+    },
+    {
+      "epoch": 1.1097811388791505,
+      "grad_norm": 2.0906708240509033,
+      "learning_rate": 4.424412704042432e-05,
+      "loss": 0.5415,
+      "step": 11080
+    },
+    {
+      "epoch": 1.1107827916061501,
+      "grad_norm": 2.103607416152954,
+      "learning_rate": 4.423406016647906e-05,
+      "loss": 0.5777,
+      "step": 11090
+    },
+    {
+      "epoch": 1.1117844443331497,
+      "grad_norm": 2.2250356674194336,
+      "learning_rate": 4.422398564426303e-05,
+      "loss": 0.5818,
+      "step": 11100
+    },
+    {
+      "epoch": 1.1127860970601493,
+      "grad_norm": 2.02903151512146,
+      "learning_rate": 4.421390347778228e-05,
+      "loss": 0.6417,
+      "step": 11110
+    },
+    {
+      "epoch": 1.113787749787149,
+      "grad_norm": 2.283987283706665,
+      "learning_rate": 4.420381367104591e-05,
+      "loss": 0.603,
+      "step": 11120
+    },
+    {
+      "epoch": 1.1147894025141483,
+      "grad_norm": 2.4817943572998047,
+      "learning_rate": 4.419371622806604e-05,
+      "loss": 0.5143,
+      "step": 11130
+    },
+    {
+      "epoch": 1.1157910552411479,
+      "grad_norm": 1.9237391948699951,
+      "learning_rate": 4.4183611152857854e-05,
+      "loss": 0.5663,
+      "step": 11140
+    },
+    {
+      "epoch": 1.1167927079681474,
+      "grad_norm": 2.5406248569488525,
+      "learning_rate": 4.417349844943953e-05,
+      "loss": 0.564,
+      "step": 11150
+    },
+    {
+      "epoch": 1.117794360695147,
+      "grad_norm": 2.506094217300415,
+      "learning_rate": 4.416337812183233e-05,
+      "loss": 0.5951,
+      "step": 11160
+    },
+    {
+      "epoch": 1.1187960134221466,
+      "grad_norm": 2.317397117614746,
+      "learning_rate": 4.415325017406051e-05,
+      "loss": 0.5939,
+      "step": 11170
+    },
+    {
+      "epoch": 1.119797666149146,
+      "grad_norm": 2.1029767990112305,
+      "learning_rate": 4.4143114610151374e-05,
+      "loss": 0.5199,
+      "step": 11180
+    },
+    {
+      "epoch": 1.1207993188761456,
+      "grad_norm": 2.195216417312622,
+      "learning_rate": 4.413297143413523e-05,
+      "loss": 0.6576,
+      "step": 11190
+    },
+    {
+      "epoch": 1.1218009716031452,
+      "grad_norm": 2.5397441387176514,
+      "learning_rate": 4.412282065004546e-05,
+      "loss": 0.5166,
+      "step": 11200
+    },
+    {
+      "epoch": 1.1228026243301448,
+      "grad_norm": 2.8908276557922363,
+      "learning_rate": 4.4112662261918415e-05,
+      "loss": 0.6163,
+      "step": 11210
+    },
+    {
+      "epoch": 1.1238042770571444,
+      "grad_norm": 2.3265693187713623,
+      "learning_rate": 4.41024962737935e-05,
+      "loss": 0.5597,
+      "step": 11220
+    },
+    {
+      "epoch": 1.1248059297841437,
+      "grad_norm": 1.9278279542922974,
+      "learning_rate": 4.4092322689713164e-05,
+      "loss": 0.6404,
+      "step": 11230
+    },
+    {
+      "epoch": 1.1258075825111433,
+      "grad_norm": 2.3198344707489014,
+      "learning_rate": 4.408214151372283e-05,
+      "loss": 0.6297,
+      "step": 11240
+    },
+    {
+      "epoch": 1.126809235238143,
+      "grad_norm": 2.27579402923584,
+      "learning_rate": 4.407195274987096e-05,
+      "loss": 0.5715,
+      "step": 11250
+    },
+    {
+      "epoch": 1.1278108879651425,
+      "grad_norm": 2.439666748046875,
+      "learning_rate": 4.4061756402209047e-05,
+      "loss": 0.6156,
+      "step": 11260
+    },
+    {
+      "epoch": 1.128812540692142,
+      "grad_norm": 2.574446201324463,
+      "learning_rate": 4.4051552474791585e-05,
+      "loss": 0.6065,
+      "step": 11270
+    },
+    {
+      "epoch": 1.1298141934191417,
+      "grad_norm": 2.493682384490967,
+      "learning_rate": 4.404134097167608e-05,
+      "loss": 0.517,
+      "step": 11280
+    },
+    {
+      "epoch": 1.130815846146141,
+      "grad_norm": 2.36503529548645,
+      "learning_rate": 4.403112189692305e-05,
+      "loss": 0.5445,
+      "step": 11290
+    },
+    {
+      "epoch": 1.1318174988731406,
+      "grad_norm": 2.6429598331451416,
+      "learning_rate": 4.402089525459604e-05,
+      "loss": 0.5817,
+      "step": 11300
+    },
+    {
+      "epoch": 1.1328191516001402,
+      "grad_norm": 2.0238144397735596,
+      "learning_rate": 4.401066104876158e-05,
+      "loss": 0.6503,
+      "step": 11310
+    },
+    {
+      "epoch": 1.1338208043271398,
+      "grad_norm": 1.9457943439483643,
+      "learning_rate": 4.4000419283489234e-05,
+      "loss": 0.5834,
+      "step": 11320
+    },
+    {
+      "epoch": 1.1348224570541394,
+      "grad_norm": 3.049051523208618,
+      "learning_rate": 4.3990169962851556e-05,
+      "loss": 0.5707,
+      "step": 11330
+    },
+    {
+      "epoch": 1.1358241097811388,
+      "grad_norm": 2.3645284175872803,
+      "learning_rate": 4.39799130909241e-05,
+      "loss": 0.6106,
+      "step": 11340
+    },
+    {
+      "epoch": 1.1368257625081384,
+      "grad_norm": 2.4572527408599854,
+      "learning_rate": 4.3969648671785427e-05,
+      "loss": 0.576,
+      "step": 11350
+    },
+    {
+      "epoch": 1.137827415235138,
+      "grad_norm": 2.7629027366638184,
+      "learning_rate": 4.395937670951712e-05,
+      "loss": 0.5461,
+      "step": 11360
+    },
+    {
+      "epoch": 1.1388290679621376,
+      "grad_norm": 1.8488332033157349,
+      "learning_rate": 4.3949097208203715e-05,
+      "loss": 0.5847,
+      "step": 11370
+    },
+    {
+      "epoch": 1.1398307206891372,
+      "grad_norm": 2.3692805767059326,
+      "learning_rate": 4.3938810171932795e-05,
+      "loss": 0.58,
+      "step": 11380
+    },
+    {
+      "epoch": 1.1408323734161367,
+      "grad_norm": 2.6989333629608154,
+      "learning_rate": 4.392851560479492e-05,
+      "loss": 0.5461,
+      "step": 11390
+    },
+    {
+      "epoch": 1.1418340261431361,
+      "grad_norm": 2.201425552368164,
+      "learning_rate": 4.3918213510883624e-05,
+      "loss": 0.5628,
+      "step": 11400
+    },
+    {
+      "epoch": 1.1428356788701357,
+      "grad_norm": 2.2745614051818848,
+      "learning_rate": 4.390790389429546e-05,
+      "loss": 0.5372,
+      "step": 11410
+    },
+    {
+      "epoch": 1.1438373315971353,
+      "grad_norm": 1.8337105512619019,
+      "learning_rate": 4.3897586759129975e-05,
+      "loss": 0.5933,
+      "step": 11420
+    },
+    {
+      "epoch": 1.144838984324135,
+      "grad_norm": 2.181349992752075,
+      "learning_rate": 4.388726210948969e-05,
+      "loss": 0.6146,
+      "step": 11430
+    },
+    {
+      "epoch": 1.1458406370511343,
+      "grad_norm": 2.9680917263031006,
+      "learning_rate": 4.387692994948012e-05,
+      "loss": 0.6383,
+      "step": 11440
+    },
+    {
+      "epoch": 1.1468422897781338,
+      "grad_norm": 2.7716708183288574,
+      "learning_rate": 4.386659028320975e-05,
+      "loss": 0.6112,
+      "step": 11450
+    },
+    {
+      "epoch": 1.1478439425051334,
+      "grad_norm": 2.357621908187866,
+      "learning_rate": 4.385624311479009e-05,
+      "loss": 0.5798,
+      "step": 11460
+    },
+    {
+      "epoch": 1.148845595232133,
+      "grad_norm": 2.055521249771118,
+      "learning_rate": 4.3845888448335596e-05,
+      "loss": 0.5666,
+      "step": 11470
+    },
+    {
+      "epoch": 1.1498472479591326,
+      "grad_norm": 2.8949215412139893,
+      "learning_rate": 4.3835526287963726e-05,
+      "loss": 0.5668,
+      "step": 11480
+    },
+    {
+      "epoch": 1.1508489006861322,
+      "grad_norm": 2.6228830814361572,
+      "learning_rate": 4.38251566377949e-05,
+      "loss": 0.6357,
+      "step": 11490
+    },
+    {
+      "epoch": 1.1518505534131316,
+      "grad_norm": 2.118854284286499,
+      "learning_rate": 4.381477950195254e-05,
+      "loss": 0.5745,
+      "step": 11500
+    },
+    {
+      "epoch": 1.1528522061401312,
+      "grad_norm": 2.6163575649261475,
+      "learning_rate": 4.380439488456301e-05,
+      "loss": 0.5824,
+      "step": 11510
+    },
+    {
+      "epoch": 1.1538538588671308,
+      "grad_norm": 2.5866751670837402,
+      "learning_rate": 4.3794002789755705e-05,
+      "loss": 0.6282,
+      "step": 11520
+    },
+    {
+      "epoch": 1.1548555115941304,
+      "grad_norm": 1.9551018476486206,
+      "learning_rate": 4.3783603221662925e-05,
+      "loss": 0.5497,
+      "step": 11530
+    },
+    {
+      "epoch": 1.15585716432113,
+      "grad_norm": 2.2837166786193848,
+      "learning_rate": 4.3773196184419996e-05,
+      "loss": 0.5256,
+      "step": 11540
+    },
+    {
+      "epoch": 1.1568588170481293,
+      "grad_norm": 2.673213005065918,
+      "learning_rate": 4.376278168216518e-05,
+      "loss": 0.638,
+      "step": 11550
+    },
+    {
+      "epoch": 1.157860469775129,
+      "grad_norm": 2.9592199325561523,
+      "learning_rate": 4.375235971903973e-05,
+      "loss": 0.5934,
+      "step": 11560
+    },
+    {
+      "epoch": 1.1588621225021285,
+      "grad_norm": 2.324190855026245,
+      "learning_rate": 4.374193029918786e-05,
+      "loss": 0.5472,
+      "step": 11570
+    },
+    {
+      "epoch": 1.159863775229128,
+      "grad_norm": 3.0013883113861084,
+      "learning_rate": 4.3731493426756734e-05,
+      "loss": 0.5461,
+      "step": 11580
+    },
+    {
+      "epoch": 1.1608654279561277,
+      "grad_norm": 2.3880512714385986,
+      "learning_rate": 4.372104910589649e-05,
+      "loss": 0.6383,
+      "step": 11590
+    },
+    {
+      "epoch": 1.1618670806831273,
+      "grad_norm": 2.222216844558716,
+      "learning_rate": 4.371059734076024e-05,
+      "loss": 0.5615,
+      "step": 11600
+    },
+    {
+      "epoch": 1.1628687334101266,
+      "grad_norm": 2.4552252292633057,
+      "learning_rate": 4.3700138135504044e-05,
+      "loss": 0.5334,
+      "step": 11610
+    },
+    {
+      "epoch": 1.1638703861371262,
+      "grad_norm": 2.393084764480591,
+      "learning_rate": 4.3689671494286913e-05,
+      "loss": 0.6087,
+      "step": 11620
+    },
+    {
+      "epoch": 1.1648720388641258,
+      "grad_norm": 2.7061352729797363,
+      "learning_rate": 4.367919742127083e-05,
+      "loss": 0.5929,
+      "step": 11630
+    },
+    {
+      "epoch": 1.1658736915911254,
+      "grad_norm": 2.2179784774780273,
+      "learning_rate": 4.366871592062073e-05,
+      "loss": 0.5751,
+      "step": 11640
+    },
+    {
+      "epoch": 1.1668753443181248,
+      "grad_norm": 2.616370677947998,
+      "learning_rate": 4.3658226996504494e-05,
+      "loss": 0.6019,
+      "step": 11650
+    },
+    {
+      "epoch": 1.1678769970451244,
+      "grad_norm": 2.4288666248321533,
+      "learning_rate": 4.364773065309296e-05,
+      "loss": 0.5673,
+      "step": 11660
+    },
+    {
+      "epoch": 1.168878649772124,
+      "grad_norm": 2.0479538440704346,
+      "learning_rate": 4.36372268945599e-05,
+      "loss": 0.6108,
+      "step": 11670
+    },
+    {
+      "epoch": 1.1698803024991236,
+      "grad_norm": 2.7070276737213135,
+      "learning_rate": 4.362671572508207e-05,
+      "loss": 0.5864,
+      "step": 11680
+    },
+    {
+      "epoch": 1.1708819552261232,
+      "grad_norm": 3.1016364097595215,
+      "learning_rate": 4.3616197148839155e-05,
+      "loss": 0.5442,
+      "step": 11690
+    },
+    {
+      "epoch": 1.1718836079531227,
+      "grad_norm": 2.680513620376587,
+      "learning_rate": 4.360567117001377e-05,
+      "loss": 0.5797,
+      "step": 11700
+    },
+    {
+      "epoch": 1.172885260680122,
+      "grad_norm": 1.764623999595642,
+      "learning_rate": 4.359513779279149e-05,
+      "loss": 0.5252,
+      "step": 11710
+    },
+    {
+      "epoch": 1.1738869134071217,
+      "grad_norm": 2.167689323425293,
+      "learning_rate": 4.358459702136083e-05,
+      "loss": 0.5704,
+      "step": 11720
+    },
+    {
+      "epoch": 1.1748885661341213,
+      "grad_norm": 3.4920172691345215,
+      "learning_rate": 4.3574048859913247e-05,
+      "loss": 0.6426,
+      "step": 11730
+    },
+    {
+      "epoch": 1.1758902188611209,
+      "grad_norm": 2.359330415725708,
+      "learning_rate": 4.3563493312643125e-05,
+      "loss": 0.6073,
+      "step": 11740
+    },
+    {
+      "epoch": 1.1768918715881205,
+      "grad_norm": 2.428713798522949,
+      "learning_rate": 4.3552930383747806e-05,
+      "loss": 0.5856,
+      "step": 11750
+    },
+    {
+      "epoch": 1.1778935243151198,
+      "grad_norm": 2.476402521133423,
+      "learning_rate": 4.354236007742754e-05,
+      "loss": 0.5963,
+      "step": 11760
+    },
+    {
+      "epoch": 1.1788951770421194,
+      "grad_norm": 2.6675846576690674,
+      "learning_rate": 4.3531782397885534e-05,
+      "loss": 0.5157,
+      "step": 11770
+    },
+    {
+      "epoch": 1.179896829769119,
+      "grad_norm": 1.9533215761184692,
+      "learning_rate": 4.352119734932791e-05,
+      "loss": 0.6073,
+      "step": 11780
+    },
+    {
+      "epoch": 1.1808984824961186,
+      "grad_norm": 2.8943727016448975,
+      "learning_rate": 4.351060493596375e-05,
+      "loss": 0.6105,
+      "step": 11790
+    },
+    {
+      "epoch": 1.1819001352231182,
+      "grad_norm": 3.005053758621216,
+      "learning_rate": 4.350000516200501e-05,
+      "loss": 0.557,
+      "step": 11800
+    },
+    {
+      "epoch": 1.1829017879501178,
+      "grad_norm": 2.9034831523895264,
+      "learning_rate": 4.348939803166664e-05,
+      "loss": 0.5801,
+      "step": 11810
+    },
+    {
+      "epoch": 1.1839034406771172,
+      "grad_norm": 2.1721954345703125,
+      "learning_rate": 4.347878354916645e-05,
+      "loss": 0.6232,
+      "step": 11820
+    },
+    {
+      "epoch": 1.1849050934041168,
+      "grad_norm": 2.2449421882629395,
+      "learning_rate": 4.346816171872522e-05,
+      "loss": 0.5813,
+      "step": 11830
+    },
+    {
+      "epoch": 1.1859067461311164,
+      "grad_norm": 2.2640445232391357,
+      "learning_rate": 4.345753254456663e-05,
+      "loss": 0.6179,
+      "step": 11840
+    },
+    {
+      "epoch": 1.186908398858116,
+      "grad_norm": 1.8900947570800781,
+      "learning_rate": 4.34468960309173e-05,
+      "loss": 0.5509,
+      "step": 11850
+    },
+    {
+      "epoch": 1.1879100515851155,
+      "grad_norm": 2.5752103328704834,
+      "learning_rate": 4.343625218200674e-05,
+      "loss": 0.6068,
+      "step": 11860
+    },
+    {
+      "epoch": 1.188911704312115,
+      "grad_norm": 2.4647738933563232,
+      "learning_rate": 4.342560100206739e-05,
+      "loss": 0.5853,
+      "step": 11870
+    },
+    {
+      "epoch": 1.1899133570391145,
+      "grad_norm": 2.0032401084899902,
+      "learning_rate": 4.3414942495334634e-05,
+      "loss": 0.5757,
+      "step": 11880
+    },
+    {
+      "epoch": 1.190915009766114,
+      "grad_norm": 2.42807674407959,
+      "learning_rate": 4.340427666604671e-05,
+      "loss": 0.6069,
+      "step": 11890
+    },
+    {
+      "epoch": 1.1919166624931137,
+      "grad_norm": 2.145582914352417,
+      "learning_rate": 4.3393603518444803e-05,
+      "loss": 0.5653,
+      "step": 11900
+    },
+    {
+      "epoch": 1.1929183152201133,
+      "grad_norm": 2.337334632873535,
+      "learning_rate": 4.338292305677303e-05,
+      "loss": 0.5728,
+      "step": 11910
+    },
+    {
+      "epoch": 1.1939199679471129,
+      "grad_norm": 2.395955801010132,
+      "learning_rate": 4.337223528527836e-05,
+      "loss": 0.5418,
+      "step": 11920
+    },
+    {
+      "epoch": 1.1949216206741122,
+      "grad_norm": 2.5528903007507324,
+      "learning_rate": 4.3361540208210725e-05,
+      "loss": 0.5833,
+      "step": 11930
+    },
+    {
+      "epoch": 1.1959232734011118,
+      "grad_norm": 1.9089552164077759,
+      "learning_rate": 4.335083782982293e-05,
+      "loss": 0.556,
+      "step": 11940
+    },
+    {
+      "epoch": 1.1969249261281114,
+      "grad_norm": 2.391150951385498,
+      "learning_rate": 4.334012815437069e-05,
+      "loss": 0.6625,
+      "step": 11950
+    },
+    {
+      "epoch": 1.197926578855111,
+      "grad_norm": 2.3361966609954834,
+      "learning_rate": 4.3329411186112616e-05,
+      "loss": 0.5594,
+      "step": 11960
+    },
+    {
+      "epoch": 1.1989282315821104,
+      "grad_norm": 1.8724137544631958,
+      "learning_rate": 4.3318686929310235e-05,
+      "loss": 0.5774,
+      "step": 11970
+    },
+    {
+      "epoch": 1.19992988430911,
+      "grad_norm": 2.7264559268951416,
+      "learning_rate": 4.330795538822795e-05,
+      "loss": 0.5614,
+      "step": 11980
+    },
+    {
+      "epoch": 1.2009315370361096,
+      "grad_norm": 2.320357322692871,
+      "learning_rate": 4.3297216567133085e-05,
+      "loss": 0.5901,
+      "step": 11990
+    },
+    {
+      "epoch": 1.2019331897631091,
+      "grad_norm": 2.6622066497802734,
+      "learning_rate": 4.328647047029584e-05,
+      "loss": 0.6711,
+      "step": 12000
+    },
+    {
+      "epoch": 1.2029348424901087,
+      "grad_norm": 2.5785880088806152,
+      "learning_rate": 4.3275717101989316e-05,
+      "loss": 0.6132,
+      "step": 12010
+    },
+    {
+      "epoch": 1.2039364952171083,
+      "grad_norm": 2.7682085037231445,
+      "learning_rate": 4.3264956466489504e-05,
+      "loss": 0.5557,
+      "step": 12020
+    },
+    {
+      "epoch": 1.2049381479441077,
+      "grad_norm": 2.4161806106567383,
+      "learning_rate": 4.325418856807529e-05,
+      "loss": 0.6732,
+      "step": 12030
+    },
+    {
+      "epoch": 1.2059398006711073,
+      "grad_norm": 2.803616523742676,
+      "learning_rate": 4.324341341102843e-05,
+      "loss": 0.5855,
+      "step": 12040
+    },
+    {
+      "epoch": 1.2069414533981069,
+      "grad_norm": 2.3754525184631348,
+      "learning_rate": 4.3232630999633595e-05,
+      "loss": 0.6923,
+      "step": 12050
+    },
+    {
+      "epoch": 1.2079431061251065,
+      "grad_norm": 2.4744246006011963,
+      "learning_rate": 4.3221841338178316e-05,
+      "loss": 0.646,
+      "step": 12060
+    },
+    {
+      "epoch": 1.208944758852106,
+      "grad_norm": 2.2677700519561768,
+      "learning_rate": 4.321104443095302e-05,
+      "loss": 0.5519,
+      "step": 12070
+    },
+    {
+      "epoch": 1.2099464115791054,
+      "grad_norm": 2.31657075881958,
+      "learning_rate": 4.3200240282251005e-05,
+      "loss": 0.5602,
+      "step": 12080
+    },
+    {
+      "epoch": 1.210948064306105,
+      "grad_norm": 2.340644598007202,
+      "learning_rate": 4.3189428896368456e-05,
+      "loss": 0.6619,
+      "step": 12090
+    },
+    {
+      "epoch": 1.2119497170331046,
+      "grad_norm": 2.118570566177368,
+      "learning_rate": 4.317861027760444e-05,
+      "loss": 0.5768,
+      "step": 12100
+    },
+    {
+      "epoch": 1.2129513697601042,
+      "grad_norm": 2.9038562774658203,
+      "learning_rate": 4.3167784430260895e-05,
+      "loss": 0.5889,
+      "step": 12110
+    },
+    {
+      "epoch": 1.2139530224871038,
+      "grad_norm": 2.5005056858062744,
+      "learning_rate": 4.3156951358642626e-05,
+      "loss": 0.5664,
+      "step": 12120
+    },
+    {
+      "epoch": 1.2149546752141034,
+      "grad_norm": 2.4353575706481934,
+      "learning_rate": 4.314611106705732e-05,
+      "loss": 0.6251,
+      "step": 12130
+    },
+    {
+      "epoch": 1.2159563279411028,
+      "grad_norm": 2.5719215869903564,
+      "learning_rate": 4.313526355981554e-05,
+      "loss": 0.6044,
+      "step": 12140
+    },
+    {
+      "epoch": 1.2169579806681023,
+      "grad_norm": 2.214345932006836,
+      "learning_rate": 4.3124408841230696e-05,
+      "loss": 0.5722,
+      "step": 12150
+    },
+    {
+      "epoch": 1.217959633395102,
+      "grad_norm": 2.1811490058898926,
+      "learning_rate": 4.3113546915619095e-05,
+      "loss": 0.5788,
+      "step": 12160
+    },
+    {
+      "epoch": 1.2189612861221015,
+      "grad_norm": 2.520916700363159,
+      "learning_rate": 4.3102677787299886e-05,
+      "loss": 0.6075,
+      "step": 12170
+    },
+    {
+      "epoch": 1.219962938849101,
+      "grad_norm": 2.66408634185791,
+      "learning_rate": 4.309180146059509e-05,
+      "loss": 0.5751,
+      "step": 12180
+    },
+    {
+      "epoch": 1.2209645915761005,
+      "grad_norm": 2.4396579265594482,
+      "learning_rate": 4.3080917939829604e-05,
+      "loss": 0.5761,
+      "step": 12190
+    },
+    {
+      "epoch": 1.2219662443031,
+      "grad_norm": 1.9858520030975342,
+      "learning_rate": 4.3070027229331155e-05,
+      "loss": 0.617,
+      "step": 12200
+    },
+    {
+      "epoch": 1.2229678970300997,
+      "grad_norm": 2.1379098892211914,
+      "learning_rate": 4.305912933343037e-05,
+      "loss": 0.5834,
+      "step": 12210
+    },
+    {
+      "epoch": 1.2239695497570993,
+      "grad_norm": 2.563939094543457,
+      "learning_rate": 4.304822425646069e-05,
+      "loss": 0.569,
+      "step": 12220
+    },
+    {
+      "epoch": 1.2249712024840989,
+      "grad_norm": 2.471806287765503,
+      "learning_rate": 4.303731200275844e-05,
+      "loss": 0.5577,
+      "step": 12230
+    },
+    {
+      "epoch": 1.2259728552110982,
+      "grad_norm": 2.36328387260437,
+      "learning_rate": 4.302639257666279e-05,
+      "loss": 0.6297,
+      "step": 12240
+    },
+    {
+      "epoch": 1.2269745079380978,
+      "grad_norm": 2.4426753520965576,
+      "learning_rate": 4.3015465982515765e-05,
+      "loss": 0.5597,
+      "step": 12250
+    },
+    {
+      "epoch": 1.2279761606650974,
+      "grad_norm": 2.5594425201416016,
+      "learning_rate": 4.300453222466224e-05,
+      "loss": 0.5633,
+      "step": 12260
+    },
+    {
+      "epoch": 1.228977813392097,
+      "grad_norm": 2.1706836223602295,
+      "learning_rate": 4.299359130744993e-05,
+      "loss": 0.5809,
+      "step": 12270
+    },
+    {
+      "epoch": 1.2299794661190966,
+      "grad_norm": 2.1926965713500977,
+      "learning_rate": 4.298264323522941e-05,
+      "loss": 0.5436,
+      "step": 12280
+    },
+    {
+      "epoch": 1.230981118846096,
+      "grad_norm": 3.1730353832244873,
+      "learning_rate": 4.297168801235409e-05,
+      "loss": 0.6721,
+      "step": 12290
+    },
+    {
+      "epoch": 1.2319827715730955,
+      "grad_norm": 2.2085063457489014,
+      "learning_rate": 4.296072564318023e-05,
+      "loss": 0.5455,
+      "step": 12300
+    },
+    {
+      "epoch": 1.2329844243000951,
+      "grad_norm": 2.361359119415283,
+      "learning_rate": 4.2949756132066924e-05,
+      "loss": 0.551,
+      "step": 12310
+    },
+    {
+      "epoch": 1.2339860770270947,
+      "grad_norm": 2.3133087158203125,
+      "learning_rate": 4.29387794833761e-05,
+      "loss": 0.6182,
+      "step": 12320
+    },
+    {
+      "epoch": 1.2349877297540943,
+      "grad_norm": 2.7090024948120117,
+      "learning_rate": 4.2927795701472564e-05,
+      "loss": 0.6011,
+      "step": 12330
+    },
+    {
+      "epoch": 1.235989382481094,
+      "grad_norm": 2.7313599586486816,
+      "learning_rate": 4.291680479072391e-05,
+      "loss": 0.6484,
+      "step": 12340
+    },
+    {
+      "epoch": 1.2369910352080933,
+      "grad_norm": 2.4200663566589355,
+      "learning_rate": 4.290580675550059e-05,
+      "loss": 0.6089,
+      "step": 12350
+    },
+    {
+      "epoch": 1.2379926879350929,
+      "grad_norm": 2.2249794006347656,
+      "learning_rate": 4.2894801600175885e-05,
+      "loss": 0.609,
+      "step": 12360
+    },
+    {
+      "epoch": 1.2389943406620925,
+      "grad_norm": 2.8315131664276123,
+      "learning_rate": 4.2883789329125894e-05,
+      "loss": 0.551,
+      "step": 12370
+    },
+    {
+      "epoch": 1.239995993389092,
+      "grad_norm": 2.9285271167755127,
+      "learning_rate": 4.287276994672959e-05,
+      "loss": 0.5966,
+      "step": 12380
+    },
+    {
+      "epoch": 1.2409976461160916,
+      "grad_norm": 2.0900344848632812,
+      "learning_rate": 4.286174345736871e-05,
+      "loss": 0.488,
+      "step": 12390
+    },
+    {
+      "epoch": 1.241999298843091,
+      "grad_norm": 2.371091604232788,
+      "learning_rate": 4.285070986542787e-05,
+      "loss": 0.5586,
+      "step": 12400
+    },
+    {
+      "epoch": 1.2430009515700906,
+      "grad_norm": 2.1431174278259277,
+      "learning_rate": 4.283966917529448e-05,
+      "loss": 0.5699,
+      "step": 12410
+    },
+    {
+      "epoch": 1.2440026042970902,
+      "grad_norm": 2.3268375396728516,
+      "learning_rate": 4.282862139135879e-05,
+      "loss": 0.5179,
+      "step": 12420
+    },
+    {
+      "epoch": 1.2450042570240898,
+      "grad_norm": 2.512864112854004,
+      "learning_rate": 4.281756651801386e-05,
+      "loss": 0.6027,
+      "step": 12430
+    },
+    {
+      "epoch": 1.2460059097510894,
+      "grad_norm": 3.174145221710205,
+      "learning_rate": 4.280650455965557e-05,
+      "loss": 0.5509,
+      "step": 12440
+    },
+    {
+      "epoch": 1.247007562478089,
+      "grad_norm": 2.1517832279205322,
+      "learning_rate": 4.279543552068263e-05,
+      "loss": 0.618,
+      "step": 12450
+    },
+    {
+      "epoch": 1.2480092152050883,
+      "grad_norm": 2.6362404823303223,
+      "learning_rate": 4.278435940549653e-05,
+      "loss": 0.5599,
+      "step": 12460
+    },
+    {
+      "epoch": 1.249010867932088,
+      "grad_norm": 2.2011988162994385,
+      "learning_rate": 4.277327621850162e-05,
+      "loss": 0.5423,
+      "step": 12470
+    },
+    {
+      "epoch": 1.2500125206590875,
+      "grad_norm": 2.244523286819458,
+      "learning_rate": 4.2762185964105045e-05,
+      "loss": 0.5017,
+      "step": 12480
+    },
+    {
+      "epoch": 1.2510141733860871,
+      "grad_norm": 2.230452060699463,
+      "learning_rate": 4.275108864671674e-05,
+      "loss": 0.5543,
+      "step": 12490
+    },
+    {
+      "epoch": 1.2520158261130865,
+      "grad_norm": 2.8010337352752686,
+      "learning_rate": 4.273998427074948e-05,
+      "loss": 0.6166,
+      "step": 12500
+    },
+    {
+      "epoch": 1.253017478840086,
+      "grad_norm": 2.4748897552490234,
+      "learning_rate": 4.2728872840618814e-05,
+      "loss": 0.61,
+      "step": 12510
+    },
+    {
+      "epoch": 1.2540191315670857,
+      "grad_norm": 2.357698917388916,
+      "learning_rate": 4.271775436074313e-05,
+      "loss": 0.5702,
+      "step": 12520
+    },
+    {
+      "epoch": 1.2550207842940853,
+      "grad_norm": 2.1585423946380615,
+      "learning_rate": 4.270662883554361e-05,
+      "loss": 0.5865,
+      "step": 12530
+    },
+    {
+      "epoch": 1.2560224370210848,
+      "grad_norm": 2.516031503677368,
+      "learning_rate": 4.2695496269444196e-05,
+      "loss": 0.6007,
+      "step": 12540
+    },
+    {
+      "epoch": 1.2570240897480844,
+      "grad_norm": 2.1451737880706787,
+      "learning_rate": 4.2684356666871696e-05,
+      "loss": 0.6683,
+      "step": 12550
+    },
+    {
+      "epoch": 1.2580257424750838,
+      "grad_norm": 2.237550735473633,
+      "learning_rate": 4.267321003225567e-05,
+      "loss": 0.5501,
+      "step": 12560
+    },
+    {
+      "epoch": 1.2590273952020834,
+      "grad_norm": 2.1487345695495605,
+      "learning_rate": 4.266205637002849e-05,
+      "loss": 0.6097,
+      "step": 12570
+    },
+    {
+      "epoch": 1.260029047929083,
+      "grad_norm": 1.9155800342559814,
+      "learning_rate": 4.2650895684625325e-05,
+      "loss": 0.5589,
+      "step": 12580
+    },
+    {
+      "epoch": 1.2610307006560826,
+      "grad_norm": 2.6508138179779053,
+      "learning_rate": 4.263972798048413e-05,
+      "loss": 0.624,
+      "step": 12590
+    },
+    {
+      "epoch": 1.262032353383082,
+      "grad_norm": 3.7485527992248535,
+      "learning_rate": 4.262855326204565e-05,
+      "loss": 0.5505,
+      "step": 12600
+    },
+    {
+      "epoch": 1.2630340061100815,
+      "grad_norm": 2.4273035526275635,
+      "learning_rate": 4.2617371533753445e-05,
+      "loss": 0.5796,
+      "step": 12610
+    },
+    {
+      "epoch": 1.2640356588370811,
+      "grad_norm": 2.114297866821289,
+      "learning_rate": 4.2606182800053806e-05,
+      "loss": 0.614,
+      "step": 12620
+    },
+    {
+      "epoch": 1.2650373115640807,
+      "grad_norm": 2.203754425048828,
+      "learning_rate": 4.259498706539586e-05,
+      "loss": 0.5348,
+      "step": 12630
+    },
+    {
+      "epoch": 1.2660389642910803,
+      "grad_norm": 2.6585121154785156,
+      "learning_rate": 4.258378433423152e-05,
+      "loss": 0.5813,
+      "step": 12640
+    },
+    {
+      "epoch": 1.26704061701808,
+      "grad_norm": 3.2380688190460205,
+      "learning_rate": 4.257257461101542e-05,
+      "loss": 0.6336,
+      "step": 12650
+    },
+    {
+      "epoch": 1.2680422697450795,
+      "grad_norm": 2.2499895095825195,
+      "learning_rate": 4.256135790020506e-05,
+      "loss": 0.6303,
+      "step": 12660
+    },
+    {
+      "epoch": 1.2690439224720789,
+      "grad_norm": 1.870112419128418,
+      "learning_rate": 4.255013420626064e-05,
+      "loss": 0.622,
+      "step": 12670
+    },
+    {
+      "epoch": 1.2700455751990785,
+      "grad_norm": 2.667515277862549,
+      "learning_rate": 4.2538903533645206e-05,
+      "loss": 0.5994,
+      "step": 12680
+    },
+    {
+      "epoch": 1.271047227926078,
+      "grad_norm": 2.1761505603790283,
+      "learning_rate": 4.252766588682452e-05,
+      "loss": 0.5713,
+      "step": 12690
+    },
+    {
+      "epoch": 1.2720488806530776,
+      "grad_norm": 2.370919704437256,
+      "learning_rate": 4.251642127026715e-05,
+      "loss": 0.5416,
+      "step": 12700
+    },
+    {
+      "epoch": 1.273050533380077,
+      "grad_norm": 1.990342617034912,
+      "learning_rate": 4.2505169688444435e-05,
+      "loss": 0.5408,
+      "step": 12710
+    },
+    {
+      "epoch": 1.2740521861070766,
+      "grad_norm": 2.1336588859558105,
+      "learning_rate": 4.2493911145830464e-05,
+      "loss": 0.6172,
+      "step": 12720
+    },
+    {
+      "epoch": 1.2750538388340762,
+      "grad_norm": 2.30049467086792,
+      "learning_rate": 4.248264564690212e-05,
+      "loss": 0.5551,
+      "step": 12730
+    },
+    {
+      "epoch": 1.2760554915610758,
+      "grad_norm": 2.0460050106048584,
+      "learning_rate": 4.247137319613904e-05,
+      "loss": 0.5742,
+      "step": 12740
+    },
+    {
+      "epoch": 1.2770571442880754,
+      "grad_norm": 2.2391974925994873,
+      "learning_rate": 4.246009379802361e-05,
+      "loss": 0.5554,
+      "step": 12750
+    },
+    {
+      "epoch": 1.278058797015075,
+      "grad_norm": 2.327256917953491,
+      "learning_rate": 4.2448807457041006e-05,
+      "loss": 0.5254,
+      "step": 12760
+    },
+    {
+      "epoch": 1.2790604497420746,
+      "grad_norm": 3.1748154163360596,
+      "learning_rate": 4.243751417767915e-05,
+      "loss": 0.5259,
+      "step": 12770
+    },
+    {
+      "epoch": 1.280062102469074,
+      "grad_norm": 2.3406453132629395,
+      "learning_rate": 4.2426213964428704e-05,
+      "loss": 0.5984,
+      "step": 12780
+    },
+    {
+      "epoch": 1.2810637551960735,
+      "grad_norm": 2.08475399017334,
+      "learning_rate": 4.241490682178314e-05,
+      "loss": 0.5541,
+      "step": 12790
+    },
+    {
+      "epoch": 1.282065407923073,
+      "grad_norm": 2.3705079555511475,
+      "learning_rate": 4.240359275423863e-05,
+      "loss": 0.6299,
+      "step": 12800
+    },
+    {
+      "epoch": 1.2830670606500727,
+      "grad_norm": 2.232518196105957,
+      "learning_rate": 4.239227176629413e-05,
+      "loss": 0.5389,
+      "step": 12810
+    },
+    {
+      "epoch": 1.284068713377072,
+      "grad_norm": 2.1422078609466553,
+      "learning_rate": 4.238094386245134e-05,
+      "loss": 0.5609,
+      "step": 12820
+    },
+    {
+      "epoch": 1.2850703661040717,
+      "grad_norm": 2.5519092082977295,
+      "learning_rate": 4.236960904721472e-05,
+      "loss": 0.5631,
+      "step": 12830
+    },
+    {
+      "epoch": 1.2860720188310713,
+      "grad_norm": 3.0536038875579834,
+      "learning_rate": 4.2358267325091456e-05,
+      "loss": 0.5331,
+      "step": 12840
+    },
+    {
+      "epoch": 1.2870736715580708,
+      "grad_norm": 2.136219024658203,
+      "learning_rate": 4.2346918700591497e-05,
+      "loss": 0.6137,
+      "step": 12850
+    },
+    {
+      "epoch": 1.2880753242850704,
+      "grad_norm": 2.4962804317474365,
+      "learning_rate": 4.2335563178227544e-05,
+      "loss": 0.5247,
+      "step": 12860
+    },
+    {
+      "epoch": 1.28907697701207,
+      "grad_norm": 3.05644154548645,
+      "learning_rate": 4.232420076251501e-05,
+      "loss": 0.5721,
+      "step": 12870
+    },
+    {
+      "epoch": 1.2900786297390694,
+      "grad_norm": 2.8292741775512695,
+      "learning_rate": 4.231283145797208e-05,
+      "loss": 0.5837,
+      "step": 12880
+    },
+    {
+      "epoch": 1.291080282466069,
+      "grad_norm": 3.002732753753662,
+      "learning_rate": 4.2301455269119665e-05,
+      "loss": 0.5213,
+      "step": 12890
+    },
+    {
+      "epoch": 1.2920819351930686,
+      "grad_norm": 2.6589910984039307,
+      "learning_rate": 4.229007220048142e-05,
+      "loss": 0.5508,
+      "step": 12900
+    },
+    {
+      "epoch": 1.2930835879200682,
+      "grad_norm": 2.5175070762634277,
+      "learning_rate": 4.227868225658373e-05,
+      "loss": 0.5646,
+      "step": 12910
+    },
+    {
+      "epoch": 1.2940852406470675,
+      "grad_norm": 2.4398906230926514,
+      "learning_rate": 4.226728544195572e-05,
+      "loss": 0.5605,
+      "step": 12920
+    },
+    {
+      "epoch": 1.2950868933740671,
+      "grad_norm": 1.8536179065704346,
+      "learning_rate": 4.225588176112922e-05,
+      "loss": 0.5898,
+      "step": 12930
+    },
+    {
+      "epoch": 1.2960885461010667,
+      "grad_norm": 2.7143056392669678,
+      "learning_rate": 4.224447121863885e-05,
+      "loss": 0.6005,
+      "step": 12940
+    },
+    {
+      "epoch": 1.2970901988280663,
+      "grad_norm": 2.5145957469940186,
+      "learning_rate": 4.223305381902189e-05,
+      "loss": 0.5204,
+      "step": 12950
+    },
+    {
+      "epoch": 1.298091851555066,
+      "grad_norm": 1.898524522781372,
+      "learning_rate": 4.222162956681839e-05,
+      "loss": 0.5403,
+      "step": 12960
+    },
+    {
+      "epoch": 1.2990935042820655,
+      "grad_norm": 2.350985527038574,
+      "learning_rate": 4.221019846657112e-05,
+      "loss": 0.5891,
+      "step": 12970
+    },
+    {
+      "epoch": 1.300095157009065,
+      "grad_norm": 2.1936440467834473,
+      "learning_rate": 4.219876052282555e-05,
+      "loss": 0.5486,
+      "step": 12980
+    },
+    {
+      "epoch": 1.3010968097360645,
+      "grad_norm": 2.114114999771118,
+      "learning_rate": 4.21873157401299e-05,
+      "loss": 0.6083,
+      "step": 12990
+    },
+    {
+      "epoch": 1.302098462463064,
+      "grad_norm": 2.300907850265503,
+      "learning_rate": 4.2175864123035085e-05,
+      "loss": 0.6078,
+      "step": 13000
+    },
+    {
+      "epoch": 1.3031001151900636,
+      "grad_norm": 2.4883675575256348,
+      "learning_rate": 4.2164405676094766e-05,
+      "loss": 0.5808,
+      "step": 13010
+    },
+    {
+      "epoch": 1.3041017679170632,
+      "grad_norm": 2.3520798683166504,
+      "learning_rate": 4.215294040386528e-05,
+      "loss": 0.6279,
+      "step": 13020
+    },
+    {
+      "epoch": 1.3051034206440626,
+      "grad_norm": 2.238900661468506,
+      "learning_rate": 4.214146831090572e-05,
+      "loss": 0.6265,
+      "step": 13030
+    },
+    {
+      "epoch": 1.3061050733710622,
+      "grad_norm": 2.766570806503296,
+      "learning_rate": 4.2129989401777876e-05,
+      "loss": 0.6319,
+      "step": 13040
+    },
+    {
+      "epoch": 1.3071067260980618,
+      "grad_norm": 2.1312856674194336,
+      "learning_rate": 4.211850368104623e-05,
+      "loss": 0.5227,
+      "step": 13050
+    },
+    {
+      "epoch": 1.3081083788250614,
+      "grad_norm": 2.324946165084839,
+      "learning_rate": 4.210701115327799e-05,
+      "loss": 0.5999,
+      "step": 13060
+    },
+    {
+      "epoch": 1.309110031552061,
+      "grad_norm": 2.1552188396453857,
+      "learning_rate": 4.2095511823043064e-05,
+      "loss": 0.5514,
+      "step": 13070
+    },
+    {
+      "epoch": 1.3101116842790606,
+      "grad_norm": 2.5594635009765625,
+      "learning_rate": 4.208400569491408e-05,
+      "loss": 0.5999,
+      "step": 13080
+    },
+    {
+      "epoch": 1.31111333700606,
+      "grad_norm": 1.8165738582611084,
+      "learning_rate": 4.2072492773466366e-05,
+      "loss": 0.5742,
+      "step": 13090
+    },
+    {
+      "epoch": 1.3121149897330595,
+      "grad_norm": 2.341984987258911,
+      "learning_rate": 4.2060973063277924e-05,
+      "loss": 0.5683,
+      "step": 13100
+    },
+    {
+      "epoch": 1.313116642460059,
+      "grad_norm": 2.9836859703063965,
+      "learning_rate": 4.204944656892948e-05,
+      "loss": 0.6223,
+      "step": 13110
+    },
+    {
+      "epoch": 1.3141182951870587,
+      "grad_norm": 2.9429287910461426,
+      "learning_rate": 4.203791329500446e-05,
+      "loss": 0.5961,
+      "step": 13120
+    },
+    {
+      "epoch": 1.315119947914058,
+      "grad_norm": 2.433025598526001,
+      "learning_rate": 4.202637324608897e-05,
+      "loss": 0.5167,
+      "step": 13130
+    },
+    {
+      "epoch": 1.3161216006410577,
+      "grad_norm": 2.171271800994873,
+      "learning_rate": 4.2014826426771825e-05,
+      "loss": 0.5361,
+      "step": 13140
+    },
+    {
+      "epoch": 1.3171232533680572,
+      "grad_norm": 2.2172558307647705,
+      "learning_rate": 4.2003272841644525e-05,
+      "loss": 0.5201,
+      "step": 13150
+    },
+    {
+      "epoch": 1.3181249060950568,
+      "grad_norm": 2.2057085037231445,
+      "learning_rate": 4.199171249530125e-05,
+      "loss": 0.5465,
+      "step": 13160
+    },
+    {
+      "epoch": 1.3191265588220564,
+      "grad_norm": 2.7693426609039307,
+      "learning_rate": 4.1980145392338896e-05,
+      "loss": 0.586,
+      "step": 13170
+    },
+    {
+      "epoch": 1.320128211549056,
+      "grad_norm": 2.1529502868652344,
+      "learning_rate": 4.196857153735702e-05,
+      "loss": 0.5835,
+      "step": 13180
+    },
+    {
+      "epoch": 1.3211298642760556,
+      "grad_norm": 2.1191604137420654,
+      "learning_rate": 4.195699093495788e-05,
+      "loss": 0.6044,
+      "step": 13190
+    },
+    {
+      "epoch": 1.322131517003055,
+      "grad_norm": 1.6969443559646606,
+      "learning_rate": 4.194540358974639e-05,
+      "loss": 0.5782,
+      "step": 13200
+    },
+    {
+      "epoch": 1.3231331697300546,
+      "grad_norm": 2.8751461505889893,
+      "learning_rate": 4.19338095063302e-05,
+      "loss": 0.5474,
+      "step": 13210
+    },
+    {
+      "epoch": 1.3241348224570542,
+      "grad_norm": 1.8890577554702759,
+      "learning_rate": 4.192220868931958e-05,
+      "loss": 0.5827,
+      "step": 13220
+    },
+    {
+      "epoch": 1.3251364751840538,
+      "grad_norm": 2.4971513748168945,
+      "learning_rate": 4.1910601143327496e-05,
+      "loss": 0.566,
+      "step": 13230
+    },
+    {
+      "epoch": 1.3261381279110531,
+      "grad_norm": 2.1435883045196533,
+      "learning_rate": 4.1898986872969626e-05,
+      "loss": 0.5873,
+      "step": 13240
+    },
+    {
+      "epoch": 1.3271397806380527,
+      "grad_norm": 2.1177799701690674,
+      "learning_rate": 4.188736588286426e-05,
+      "loss": 0.5642,
+      "step": 13250
+    },
+    {
+      "epoch": 1.3281414333650523,
+      "grad_norm": 2.0428977012634277,
+      "learning_rate": 4.187573817763242e-05,
+      "loss": 0.5201,
+      "step": 13260
+    },
+    {
+      "epoch": 1.329143086092052,
+      "grad_norm": 2.122833490371704,
+      "learning_rate": 4.1864103761897746e-05,
+      "loss": 0.5834,
+      "step": 13270
+    },
+    {
+      "epoch": 1.3301447388190515,
+      "grad_norm": 2.3523943424224854,
+      "learning_rate": 4.185246264028659e-05,
+      "loss": 0.5832,
+      "step": 13280
+    },
+    {
+      "epoch": 1.331146391546051,
+      "grad_norm": 1.596888780593872,
+      "learning_rate": 4.184081481742794e-05,
+      "loss": 0.6354,
+      "step": 13290
+    },
+    {
+      "epoch": 1.3321480442730507,
+      "grad_norm": 2.279982328414917,
+      "learning_rate": 4.182916029795346e-05,
+      "loss": 0.5331,
+      "step": 13300
+    },
+    {
+      "epoch": 1.33314969700005,
+      "grad_norm": 2.457702875137329,
+      "learning_rate": 4.181749908649748e-05,
+      "loss": 0.6325,
+      "step": 13310
+    },
+    {
+      "epoch": 1.3341513497270496,
+      "grad_norm": 1.8950936794281006,
+      "learning_rate": 4.180583118769699e-05,
+      "loss": 0.5559,
+      "step": 13320
+    },
+    {
+      "epoch": 1.3351530024540492,
+      "grad_norm": 2.870814561843872,
+      "learning_rate": 4.179415660619164e-05,
+      "loss": 0.5552,
+      "step": 13330
+    },
+    {
+      "epoch": 1.3361546551810488,
+      "grad_norm": 1.7458791732788086,
+      "learning_rate": 4.178247534662372e-05,
+      "loss": 0.5206,
+      "step": 13340
+    },
+    {
+      "epoch": 1.3371563079080482,
+      "grad_norm": 2.9492814540863037,
+      "learning_rate": 4.17707874136382e-05,
+      "loss": 0.499,
+      "step": 13350
+    },
+    {
+      "epoch": 1.3381579606350478,
+      "grad_norm": 2.308802604675293,
+      "learning_rate": 4.1759092811882696e-05,
+      "loss": 0.6022,
+      "step": 13360
+    },
+    {
+      "epoch": 1.3391596133620474,
+      "grad_norm": 1.9808357954025269,
+      "learning_rate": 4.174739154600746e-05,
+      "loss": 0.5856,
+      "step": 13370
+    },
+    {
+      "epoch": 1.340161266089047,
+      "grad_norm": 2.5292766094207764,
+      "learning_rate": 4.173568362066542e-05,
+      "loss": 0.6017,
+      "step": 13380
+    },
+    {
+      "epoch": 1.3411629188160465,
+      "grad_norm": 1.9373970031738281,
+      "learning_rate": 4.172396904051215e-05,
+      "loss": 0.5282,
+      "step": 13390
+    },
+    {
+      "epoch": 1.3421645715430461,
+      "grad_norm": 3.495178699493408,
+      "learning_rate": 4.1712247810205824e-05,
+      "loss": 0.5122,
+      "step": 13400
+    },
+    {
+      "epoch": 1.3431662242700455,
+      "grad_norm": 2.2471678256988525,
+      "learning_rate": 4.170051993440733e-05,
+      "loss": 0.5307,
+      "step": 13410
+    },
+    {
+      "epoch": 1.344167876997045,
+      "grad_norm": 2.055947780609131,
+      "learning_rate": 4.1688785417780155e-05,
+      "loss": 0.5769,
+      "step": 13420
+    },
+    {
+      "epoch": 1.3451695297240447,
+      "grad_norm": 2.5820884704589844,
+      "learning_rate": 4.167704426499042e-05,
+      "loss": 0.5912,
+      "step": 13430
+    },
+    {
+      "epoch": 1.3461711824510443,
+      "grad_norm": 2.5380706787109375,
+      "learning_rate": 4.1665296480706917e-05,
+      "loss": 0.6127,
+      "step": 13440
+    },
+    {
+      "epoch": 1.3471728351780436,
+      "grad_norm": 2.453075885772705,
+      "learning_rate": 4.1653542069601055e-05,
+      "loss": 0.5635,
+      "step": 13450
+    },
+    {
+      "epoch": 1.3481744879050432,
+      "grad_norm": 2.359921932220459,
+      "learning_rate": 4.164178103634688e-05,
+      "loss": 0.5789,
+      "step": 13460
+    },
+    {
+      "epoch": 1.3491761406320428,
+      "grad_norm": 2.4070520401000977,
+      "learning_rate": 4.163001338562108e-05,
+      "loss": 0.6129,
+      "step": 13470
+    },
+    {
+      "epoch": 1.3501777933590424,
+      "grad_norm": 2.7085323333740234,
+      "learning_rate": 4.1618239122102965e-05,
+      "loss": 0.6108,
+      "step": 13480
+    },
+    {
+      "epoch": 1.351179446086042,
+      "grad_norm": 2.0704357624053955,
+      "learning_rate": 4.160645825047447e-05,
+      "loss": 0.6244,
+      "step": 13490
+    },
+    {
+      "epoch": 1.3521810988130416,
+      "grad_norm": 2.2796645164489746,
+      "learning_rate": 4.159467077542016e-05,
+      "loss": 0.5491,
+      "step": 13500
+    },
+    {
+      "epoch": 1.3531827515400412,
+      "grad_norm": 2.2350540161132812,
+      "learning_rate": 4.158287670162725e-05,
+      "loss": 0.5411,
+      "step": 13510
+    },
+    {
+      "epoch": 1.3541844042670406,
+      "grad_norm": 2.277599334716797,
+      "learning_rate": 4.1571076033785556e-05,
+      "loss": 0.5953,
+      "step": 13520
+    },
+    {
+      "epoch": 1.3551860569940402,
+      "grad_norm": 2.4963531494140625,
+      "learning_rate": 4.155926877658751e-05,
+      "loss": 0.5321,
+      "step": 13530
+    },
+    {
+      "epoch": 1.3561877097210397,
+      "grad_norm": 1.7574398517608643,
+      "learning_rate": 4.154745493472817e-05,
+      "loss": 0.5299,
+      "step": 13540
+    },
+    {
+      "epoch": 1.3571893624480393,
+      "grad_norm": 2.508462905883789,
+      "learning_rate": 4.1535634512905225e-05,
+      "loss": 0.5877,
+      "step": 13550
+    },
+    {
+      "epoch": 1.3581910151750387,
+      "grad_norm": 3.9622373580932617,
+      "learning_rate": 4.152380751581897e-05,
+      "loss": 0.5962,
+      "step": 13560
+    },
+    {
+      "epoch": 1.3591926679020383,
+      "grad_norm": 2.56437087059021,
+      "learning_rate": 4.151197394817231e-05,
+      "loss": 0.5113,
+      "step": 13570
+    },
+    {
+      "epoch": 1.360194320629038,
+      "grad_norm": 3.0979878902435303,
+      "learning_rate": 4.150013381467078e-05,
+      "loss": 0.5881,
+      "step": 13580
+    },
+    {
+      "epoch": 1.3611959733560375,
+      "grad_norm": 2.280622959136963,
+      "learning_rate": 4.148828712002252e-05,
+      "loss": 0.6048,
+      "step": 13590
+    },
+    {
+      "epoch": 1.362197626083037,
+      "grad_norm": 2.422917366027832,
+      "learning_rate": 4.147643386893825e-05,
+      "loss": 0.6643,
+      "step": 13600
+    },
+    {
+      "epoch": 1.3631992788100367,
+      "grad_norm": 1.9158912897109985,
+      "learning_rate": 4.146457406613134e-05,
+      "loss": 0.5696,
+      "step": 13610
+    },
+    {
+      "epoch": 1.364200931537036,
+      "grad_norm": 2.052349328994751,
+      "learning_rate": 4.145270771631773e-05,
+      "loss": 0.5746,
+      "step": 13620
+    },
+    {
+      "epoch": 1.3652025842640356,
+      "grad_norm": 3.113779306411743,
+      "learning_rate": 4.144083482421599e-05,
+      "loss": 0.5859,
+      "step": 13630
+    },
+    {
+      "epoch": 1.3662042369910352,
+      "grad_norm": 2.7686078548431396,
+      "learning_rate": 4.1428955394547286e-05,
+      "loss": 0.5829,
+      "step": 13640
+    },
+    {
+      "epoch": 1.3672058897180348,
+      "grad_norm": 2.467316150665283,
+      "learning_rate": 4.141706943203537e-05,
+      "loss": 0.51,
+      "step": 13650
+    },
+    {
+      "epoch": 1.3682075424450342,
+      "grad_norm": 2.354332447052002,
+      "learning_rate": 4.140517694140661e-05,
+      "loss": 0.5914,
+      "step": 13660
+    },
+    {
+      "epoch": 1.3692091951720338,
+      "grad_norm": 2.3805387020111084,
+      "learning_rate": 4.1393277927389946e-05,
+      "loss": 0.6216,
+      "step": 13670
+    },
+    {
+      "epoch": 1.3702108478990334,
+      "grad_norm": 1.6980246305465698,
+      "learning_rate": 4.138137239471693e-05,
+      "loss": 0.65,
+      "step": 13680
+    },
+    {
+      "epoch": 1.371212500626033,
+      "grad_norm": 2.8044466972351074,
+      "learning_rate": 4.136946034812171e-05,
+      "loss": 0.5768,
+      "step": 13690
+    },
+    {
+      "epoch": 1.3722141533530325,
+      "grad_norm": 2.284212112426758,
+      "learning_rate": 4.135754179234102e-05,
+      "loss": 0.548,
+      "step": 13700
+    },
+    {
+      "epoch": 1.3732158060800321,
+      "grad_norm": 2.20108699798584,
+      "learning_rate": 4.134561673211417e-05,
+      "loss": 0.5497,
+      "step": 13710
+    },
+    {
+      "epoch": 1.3742174588070317,
+      "grad_norm": 2.3220138549804688,
+      "learning_rate": 4.133368517218305e-05,
+      "loss": 0.6195,
+      "step": 13720
+    },
+    {
+      "epoch": 1.375219111534031,
+      "grad_norm": 2.1386282444000244,
+      "learning_rate": 4.132174711729217e-05,
+      "loss": 0.563,
+      "step": 13730
+    },
+    {
+      "epoch": 1.3762207642610307,
+      "grad_norm": 3.0128397941589355,
+      "learning_rate": 4.130980257218861e-05,
+      "loss": 0.5758,
+      "step": 13740
+    },
+    {
+      "epoch": 1.3772224169880303,
+      "grad_norm": 2.1034698486328125,
+      "learning_rate": 4.129785154162201e-05,
+      "loss": 0.536,
+      "step": 13750
+    },
+    {
+      "epoch": 1.3782240697150299,
+      "grad_norm": 2.1103596687316895,
+      "learning_rate": 4.12858940303446e-05,
+      "loss": 0.5476,
+      "step": 13760
+    },
+    {
+      "epoch": 1.3792257224420292,
+      "grad_norm": 2.607357978820801,
+      "learning_rate": 4.1273930043111185e-05,
+      "loss": 0.5823,
+      "step": 13770
+    },
+    {
+      "epoch": 1.3802273751690288,
+      "grad_norm": 2.7858710289001465,
+      "learning_rate": 4.1261959584679156e-05,
+      "loss": 0.5756,
+      "step": 13780
+    },
+    {
+      "epoch": 1.3812290278960284,
+      "grad_norm": 2.6473164558410645,
+      "learning_rate": 4.124998265980848e-05,
+      "loss": 0.6113,
+      "step": 13790
+    },
+    {
+      "epoch": 1.382230680623028,
+      "grad_norm": 2.2737371921539307,
+      "learning_rate": 4.1237999273261676e-05,
+      "loss": 0.6199,
+      "step": 13800
+    },
+    {
+      "epoch": 1.3832323333500276,
+      "grad_norm": 2.420541524887085,
+      "learning_rate": 4.1226009429803836e-05,
+      "loss": 0.5491,
+      "step": 13810
+    },
+    {
+      "epoch": 1.3842339860770272,
+      "grad_norm": 1.7622172832489014,
+      "learning_rate": 4.121401313420264e-05,
+      "loss": 0.4744,
+      "step": 13820
+    },
+    {
+      "epoch": 1.3852356388040268,
+      "grad_norm": 2.3782007694244385,
+      "learning_rate": 4.1202010391228306e-05,
+      "loss": 0.6206,
+      "step": 13830
+    },
+    {
+      "epoch": 1.3862372915310262,
+      "grad_norm": 2.8256185054779053,
+      "learning_rate": 4.1190001205653636e-05,
+      "loss": 0.5888,
+      "step": 13840
+    },
+    {
+      "epoch": 1.3872389442580257,
+      "grad_norm": 2.546830177307129,
+      "learning_rate": 4.117798558225399e-05,
+      "loss": 0.5943,
+      "step": 13850
+    },
+    {
+      "epoch": 1.3882405969850253,
+      "grad_norm": 3.37953782081604,
+      "learning_rate": 4.116596352580728e-05,
+      "loss": 0.5787,
+      "step": 13860
+    },
+    {
+      "epoch": 1.389242249712025,
+      "grad_norm": 2.2097134590148926,
+      "learning_rate": 4.1153935041093974e-05,
+      "loss": 0.5856,
+      "step": 13870
+    },
+    {
+      "epoch": 1.3902439024390243,
+      "grad_norm": 2.6662843227386475,
+      "learning_rate": 4.114190013289712e-05,
+      "loss": 0.5243,
+      "step": 13880
+    },
+    {
+      "epoch": 1.3912455551660239,
+      "grad_norm": 1.8162366151809692,
+      "learning_rate": 4.112985880600229e-05,
+      "loss": 0.5903,
+      "step": 13890
+    },
+    {
+      "epoch": 1.3922472078930235,
+      "grad_norm": 2.9583468437194824,
+      "learning_rate": 4.111781106519763e-05,
+      "loss": 0.5232,
+      "step": 13900
+    },
+    {
+      "epoch": 1.393248860620023,
+      "grad_norm": 2.8035693168640137,
+      "learning_rate": 4.1105756915273826e-05,
+      "loss": 0.5892,
+      "step": 13910
+    },
+    {
+      "epoch": 1.3942505133470227,
+      "grad_norm": 2.036820650100708,
+      "learning_rate": 4.10936963610241e-05,
+      "loss": 0.5729,
+      "step": 13920
+    },
+    {
+      "epoch": 1.3952521660740222,
+      "grad_norm": 2.5496792793273926,
+      "learning_rate": 4.108162940724427e-05,
+      "loss": 0.5305,
+      "step": 13930
+    },
+    {
+      "epoch": 1.3962538188010216,
+      "grad_norm": 2.4634103775024414,
+      "learning_rate": 4.1069556058732624e-05,
+      "loss": 0.6228,
+      "step": 13940
+    },
+    {
+      "epoch": 1.3972554715280212,
+      "grad_norm": 1.6737140417099,
+      "learning_rate": 4.105747632029006e-05,
+      "loss": 0.5499,
+      "step": 13950
+    },
+    {
+      "epoch": 1.3982571242550208,
+      "grad_norm": 2.705306053161621,
+      "learning_rate": 4.104539019671997e-05,
+      "loss": 0.5735,
+      "step": 13960
+    },
+    {
+      "epoch": 1.3992587769820204,
+      "grad_norm": 2.2632908821105957,
+      "learning_rate": 4.103329769282832e-05,
+      "loss": 0.5479,
+      "step": 13970
+    },
+    {
+      "epoch": 1.4002604297090198,
+      "grad_norm": 1.8262296915054321,
+      "learning_rate": 4.10211988134236e-05,
+      "loss": 0.5874,
+      "step": 13980
+    },
+    {
+      "epoch": 1.4012620824360194,
+      "grad_norm": 2.7786033153533936,
+      "learning_rate": 4.100909356331682e-05,
+      "loss": 0.5826,
+      "step": 13990
+    },
+    {
+      "epoch": 1.402263735163019,
+      "grad_norm": 2.1953024864196777,
+      "learning_rate": 4.099698194732154e-05,
+      "loss": 0.5816,
+      "step": 14000
+    },
+    {
+      "epoch": 1.4032653878900185,
+      "grad_norm": 2.1544125080108643,
+      "learning_rate": 4.098486397025386e-05,
+      "loss": 0.5336,
+      "step": 14010
+    },
+    {
+      "epoch": 1.4042670406170181,
+      "grad_norm": 2.1569459438323975,
+      "learning_rate": 4.097273963693239e-05,
+      "loss": 0.5122,
+      "step": 14020
+    },
+    {
+      "epoch": 1.4052686933440177,
+      "grad_norm": 2.009190559387207,
+      "learning_rate": 4.096060895217826e-05,
+      "loss": 0.5175,
+      "step": 14030
+    },
+    {
+      "epoch": 1.4062703460710173,
+      "grad_norm": 2.4000799655914307,
+      "learning_rate": 4.094847192081516e-05,
+      "loss": 0.5778,
+      "step": 14040
+    },
+    {
+      "epoch": 1.4072719987980167,
+      "grad_norm": 1.9345402717590332,
+      "learning_rate": 4.0936328547669264e-05,
+      "loss": 0.5855,
+      "step": 14050
+    },
+    {
+      "epoch": 1.4082736515250163,
+      "grad_norm": 2.416372060775757,
+      "learning_rate": 4.09241788375693e-05,
+      "loss": 0.627,
+      "step": 14060
+    },
+    {
+      "epoch": 1.4092753042520159,
+      "grad_norm": 2.3259639739990234,
+      "learning_rate": 4.091202279534651e-05,
+      "loss": 0.5788,
+      "step": 14070
+    },
+    {
+      "epoch": 1.4102769569790155,
+      "grad_norm": 2.430366277694702,
+      "learning_rate": 4.089986042583465e-05,
+      "loss": 0.5928,
+      "step": 14080
+    },
+    {
+      "epoch": 1.4112786097060148,
+      "grad_norm": 2.045180320739746,
+      "learning_rate": 4.088769173386996e-05,
+      "loss": 0.6082,
+      "step": 14090
+    },
+    {
+      "epoch": 1.4122802624330144,
+      "grad_norm": 2.674597978591919,
+      "learning_rate": 4.0875516724291255e-05,
+      "loss": 0.5751,
+      "step": 14100
+    },
+    {
+      "epoch": 1.413281915160014,
+      "grad_norm": 2.7040843963623047,
+      "learning_rate": 4.0863335401939815e-05,
+      "loss": 0.5585,
+      "step": 14110
+    },
+    {
+      "epoch": 1.4142835678870136,
+      "grad_norm": 2.3586766719818115,
+      "learning_rate": 4.085114777165945e-05,
+      "loss": 0.519,
+      "step": 14120
+    },
+    {
+      "epoch": 1.4152852206140132,
+      "grad_norm": 1.9052765369415283,
+      "learning_rate": 4.0838953838296464e-05,
+      "loss": 0.5796,
+      "step": 14130
+    },
+    {
+      "epoch": 1.4162868733410128,
+      "grad_norm": 1.9968295097351074,
+      "learning_rate": 4.08267536066997e-05,
+      "loss": 0.5678,
+      "step": 14140
+    },
+    {
+      "epoch": 1.4172885260680121,
+      "grad_norm": 2.5665669441223145,
+      "learning_rate": 4.081454708172047e-05,
+      "loss": 0.5907,
+      "step": 14150
+    },
+    {
+      "epoch": 1.4182901787950117,
+      "grad_norm": 2.1774866580963135,
+      "learning_rate": 4.080233426821259e-05,
+      "loss": 0.5392,
+      "step": 14160
+    },
+    {
+      "epoch": 1.4192918315220113,
+      "grad_norm": 2.24665904045105,
+      "learning_rate": 4.079011517103241e-05,
+      "loss": 0.5701,
+      "step": 14170
+    },
+    {
+      "epoch": 1.420293484249011,
+      "grad_norm": 2.3016774654388428,
+      "learning_rate": 4.0777889795038736e-05,
+      "loss": 0.5526,
+      "step": 14180
+    },
+    {
+      "epoch": 1.4212951369760103,
+      "grad_norm": 3.0300395488739014,
+      "learning_rate": 4.07656581450929e-05,
+      "loss": 0.5284,
+      "step": 14190
+    },
+    {
+      "epoch": 1.4222967897030099,
+      "grad_norm": 2.790093421936035,
+      "learning_rate": 4.0753420226058724e-05,
+      "loss": 0.5558,
+      "step": 14200
+    },
+    {
+      "epoch": 1.4232984424300095,
+      "grad_norm": 2.152724027633667,
+      "learning_rate": 4.074117604280252e-05,
+      "loss": 0.6086,
+      "step": 14210
+    },
+    {
+      "epoch": 1.424300095157009,
+      "grad_norm": 2.25683856010437,
+      "learning_rate": 4.0728925600193076e-05,
+      "loss": 0.5255,
+      "step": 14220
+    },
+    {
+      "epoch": 1.4253017478840087,
+      "grad_norm": 3.285686492919922,
+      "learning_rate": 4.07166689031017e-05,
+      "loss": 0.5968,
+      "step": 14230
+    },
+    {
+      "epoch": 1.4263034006110082,
+      "grad_norm": 2.408635377883911,
+      "learning_rate": 4.070440595640217e-05,
+      "loss": 0.5599,
+      "step": 14240
+    },
+    {
+      "epoch": 1.4273050533380078,
+      "grad_norm": 2.2016947269439697,
+      "learning_rate": 4.069213676497073e-05,
+      "loss": 0.5839,
+      "step": 14250
+    },
+    {
+      "epoch": 1.4283067060650072,
+      "grad_norm": 2.815704107284546,
+      "learning_rate": 4.067986133368614e-05,
+      "loss": 0.6123,
+      "step": 14260
+    },
+    {
+      "epoch": 1.4293083587920068,
+      "grad_norm": 2.9127774238586426,
+      "learning_rate": 4.0667579667429625e-05,
+      "loss": 0.5208,
+      "step": 14270
+    },
+    {
+      "epoch": 1.4303100115190064,
+      "grad_norm": 2.2251622676849365,
+      "learning_rate": 4.0655291771084896e-05,
+      "loss": 0.5773,
+      "step": 14280
+    },
+    {
+      "epoch": 1.431311664246006,
+      "grad_norm": 2.434687614440918,
+      "learning_rate": 4.064299764953813e-05,
+      "loss": 0.5782,
+      "step": 14290
+    },
+    {
+      "epoch": 1.4323133169730053,
+      "grad_norm": 2.1257598400115967,
+      "learning_rate": 4.0630697307678e-05,
+      "loss": 0.5693,
+      "step": 14300
+    },
+    {
+      "epoch": 1.433314969700005,
+      "grad_norm": 2.1355576515197754,
+      "learning_rate": 4.061839075039562e-05,
+      "loss": 0.5607,
+      "step": 14310
+    },
+    {
+      "epoch": 1.4343166224270045,
+      "grad_norm": 2.9941141605377197,
+      "learning_rate": 4.060607798258459e-05,
+      "loss": 0.6015,
+      "step": 14320
+    },
+    {
+      "epoch": 1.4353182751540041,
+      "grad_norm": 1.9984222650527954,
+      "learning_rate": 4.059375900914102e-05,
+      "loss": 0.5565,
+      "step": 14330
+    },
+    {
+      "epoch": 1.4363199278810037,
+      "grad_norm": 2.6559393405914307,
+      "learning_rate": 4.058143383496341e-05,
+      "loss": 0.5428,
+      "step": 14340
+    },
+    {
+      "epoch": 1.4373215806080033,
+      "grad_norm": 2.177220106124878,
+      "learning_rate": 4.05691024649528e-05,
+      "loss": 0.607,
+      "step": 14350
+    },
+    {
+      "epoch": 1.438323233335003,
+      "grad_norm": 2.5836830139160156,
+      "learning_rate": 4.055676490401264e-05,
+      "loss": 0.5665,
+      "step": 14360
+    },
+    {
+      "epoch": 1.4393248860620023,
+      "grad_norm": 3.489238977432251,
+      "learning_rate": 4.0544421157048875e-05,
+      "loss": 0.5957,
+      "step": 14370
+    },
+    {
+      "epoch": 1.4403265387890019,
+      "grad_norm": 2.5777053833007812,
+      "learning_rate": 4.053207122896989e-05,
+      "loss": 0.5574,
+      "step": 14380
+    },
+    {
+      "epoch": 1.4413281915160014,
+      "grad_norm": 2.3161325454711914,
+      "learning_rate": 4.0519715124686535e-05,
+      "loss": 0.5757,
+      "step": 14390
+    },
+    {
+      "epoch": 1.4423298442430008,
+      "grad_norm": 2.5164365768432617,
+      "learning_rate": 4.050735284911212e-05,
+      "loss": 0.5554,
+      "step": 14400
+    },
+    {
+      "epoch": 1.4433314969700004,
+      "grad_norm": 2.887172222137451,
+      "learning_rate": 4.049498440716241e-05,
+      "loss": 0.6013,
+      "step": 14410
+    },
+    {
+      "epoch": 1.444333149697,
+      "grad_norm": 1.7889225482940674,
+      "learning_rate": 4.0482609803755604e-05,
+      "loss": 0.5402,
+      "step": 14420
+    },
+    {
+      "epoch": 1.4453348024239996,
+      "grad_norm": 2.003990650177002,
+      "learning_rate": 4.047022904381238e-05,
+      "loss": 0.5742,
+      "step": 14430
+    },
+    {
+      "epoch": 1.4463364551509992,
+      "grad_norm": 2.028975248336792,
+      "learning_rate": 4.045784213225584e-05,
+      "loss": 0.5762,
+      "step": 14440
+    },
+    {
+      "epoch": 1.4473381078779988,
+      "grad_norm": 3.170283317565918,
+      "learning_rate": 4.0445449074011535e-05,
+      "loss": 0.5331,
+      "step": 14450
+    },
+    {
+      "epoch": 1.4483397606049984,
+      "grad_norm": 2.5726683139801025,
+      "learning_rate": 4.0433049874007475e-05,
+      "loss": 0.5711,
+      "step": 14460
+    },
+    {
+      "epoch": 1.4493414133319977,
+      "grad_norm": 2.1386616230010986,
+      "learning_rate": 4.042064453717411e-05,
+      "loss": 0.5832,
+      "step": 14470
+    },
+    {
+      "epoch": 1.4503430660589973,
+      "grad_norm": 3.0277740955352783,
+      "learning_rate": 4.040823306844431e-05,
+      "loss": 0.5874,
+      "step": 14480
+    },
+    {
+      "epoch": 1.451344718785997,
+      "grad_norm": 2.120304822921753,
+      "learning_rate": 4.039581547275339e-05,
+      "loss": 0.5645,
+      "step": 14490
+    },
+    {
+      "epoch": 1.4523463715129965,
+      "grad_norm": 1.4980655908584595,
+      "learning_rate": 4.038339175503914e-05,
+      "loss": 0.5484,
+      "step": 14500
+    },
+    {
+      "epoch": 1.4533480242399959,
+      "grad_norm": 2.6482741832733154,
+      "learning_rate": 4.037096192024171e-05,
+      "loss": 0.573,
+      "step": 14510
+    },
+    {
+      "epoch": 1.4543496769669955,
+      "grad_norm": 2.6279139518737793,
+      "learning_rate": 4.035852597330375e-05,
+      "loss": 0.5177,
+      "step": 14520
+    },
+    {
+      "epoch": 1.455351329693995,
+      "grad_norm": 2.7452750205993652,
+      "learning_rate": 4.034608391917032e-05,
+      "loss": 0.489,
+      "step": 14530
+    },
+    {
+      "epoch": 1.4563529824209946,
+      "grad_norm": 1.965122103691101,
+      "learning_rate": 4.033363576278889e-05,
+      "loss": 0.575,
+      "step": 14540
+    },
+    {
+      "epoch": 1.4573546351479942,
+      "grad_norm": 2.4850306510925293,
+      "learning_rate": 4.0321181509109374e-05,
+      "loss": 0.5551,
+      "step": 14550
+    },
+    {
+      "epoch": 1.4583562878749938,
+      "grad_norm": 2.552098274230957,
+      "learning_rate": 4.0308721163084105e-05,
+      "loss": 0.5781,
+      "step": 14560
+    },
+    {
+      "epoch": 1.4593579406019934,
+      "grad_norm": 2.414721727371216,
+      "learning_rate": 4.029625472966785e-05,
+      "loss": 0.5155,
+      "step": 14570
+    },
+    {
+      "epoch": 1.4603595933289928,
+      "grad_norm": 2.572413444519043,
+      "learning_rate": 4.028378221381778e-05,
+      "loss": 0.5722,
+      "step": 14580
+    },
+    {
+      "epoch": 1.4613612460559924,
+      "grad_norm": 2.4709455966949463,
+      "learning_rate": 4.027130362049348e-05,
+      "loss": 0.6083,
+      "step": 14590
+    },
+    {
+      "epoch": 1.462362898782992,
+      "grad_norm": 2.268739700317383,
+      "learning_rate": 4.025881895465699e-05,
+      "loss": 0.541,
+      "step": 14600
+    },
+    {
+      "epoch": 1.4633645515099916,
+      "grad_norm": 2.681468963623047,
+      "learning_rate": 4.024632822127271e-05,
+      "loss": 0.5514,
+      "step": 14610
+    },
+    {
+      "epoch": 1.464366204236991,
+      "grad_norm": 2.1005759239196777,
+      "learning_rate": 4.023383142530751e-05,
+      "loss": 0.5443,
+      "step": 14620
+    },
+    {
+      "epoch": 1.4653678569639905,
+      "grad_norm": 2.0630645751953125,
+      "learning_rate": 4.022132857173061e-05,
+      "loss": 0.5303,
+      "step": 14630
+    },
+    {
+      "epoch": 1.4663695096909901,
+      "grad_norm": 2.8516407012939453,
+      "learning_rate": 4.0208819665513684e-05,
+      "loss": 0.5339,
+      "step": 14640
+    },
+    {
+      "epoch": 1.4673711624179897,
+      "grad_norm": 2.4817099571228027,
+      "learning_rate": 4.0196304711630805e-05,
+      "loss": 0.5363,
+      "step": 14650
+    },
+    {
+      "epoch": 1.4683728151449893,
+      "grad_norm": 3.3822922706604004,
+      "learning_rate": 4.0183783715058444e-05,
+      "loss": 0.528,
+      "step": 14660
+    },
+    {
+      "epoch": 1.4693744678719889,
+      "grad_norm": 2.3686294555664062,
+      "learning_rate": 4.017125668077546e-05,
+      "loss": 0.5148,
+      "step": 14670
+    },
+    {
+      "epoch": 1.4703761205989883,
+      "grad_norm": 2.2232017517089844,
+      "learning_rate": 4.0158723613763145e-05,
+      "loss": 0.5897,
+      "step": 14680
+    },
+    {
+      "epoch": 1.4713777733259878,
+      "grad_norm": 2.1292519569396973,
+      "learning_rate": 4.014618451900517e-05,
+      "loss": 0.59,
+      "step": 14690
+    },
+    {
+      "epoch": 1.4723794260529874,
+      "grad_norm": 2.4203102588653564,
+      "learning_rate": 4.013363940148759e-05,
+      "loss": 0.5647,
+      "step": 14700
+    },
+    {
+      "epoch": 1.473381078779987,
+      "grad_norm": 1.9357800483703613,
+      "learning_rate": 4.0121088266198906e-05,
+      "loss": 0.5342,
+      "step": 14710
+    },
+    {
+      "epoch": 1.4743827315069864,
+      "grad_norm": 2.3694651126861572,
+      "learning_rate": 4.0108531118129934e-05,
+      "loss": 0.5526,
+      "step": 14720
+    },
+    {
+      "epoch": 1.475384384233986,
+      "grad_norm": 2.351116180419922,
+      "learning_rate": 4.009596796227396e-05,
+      "loss": 0.489,
+      "step": 14730
+    },
+    {
+      "epoch": 1.4763860369609856,
+      "grad_norm": 2.0364739894866943,
+      "learning_rate": 4.0083398803626606e-05,
+      "loss": 0.6545,
+      "step": 14740
+    },
+    {
+      "epoch": 1.4773876896879852,
+      "grad_norm": 3.0939629077911377,
+      "learning_rate": 4.00708236471859e-05,
+      "loss": 0.5798,
+      "step": 14750
+    },
+    {
+      "epoch": 1.4783893424149848,
+      "grad_norm": 1.8895127773284912,
+      "learning_rate": 4.005824249795225e-05,
+      "loss": 0.5247,
+      "step": 14760
+    },
+    {
+      "epoch": 1.4793909951419844,
+      "grad_norm": 2.274951934814453,
+      "learning_rate": 4.004565536092845e-05,
+      "loss": 0.5154,
+      "step": 14770
+    },
+    {
+      "epoch": 1.480392647868984,
+      "grad_norm": 2.3646650314331055,
+      "learning_rate": 4.0033062241119676e-05,
+      "loss": 0.5363,
+      "step": 14780
+    },
+    {
+      "epoch": 1.4813943005959833,
+      "grad_norm": 2.129894256591797,
+      "learning_rate": 4.002046314353348e-05,
+      "loss": 0.537,
+      "step": 14790
+    },
+    {
+      "epoch": 1.482395953322983,
+      "grad_norm": 2.5647120475769043,
+      "learning_rate": 4.000785807317981e-05,
+      "loss": 0.5885,
+      "step": 14800
+    },
+    {
+      "epoch": 1.4833976060499825,
+      "grad_norm": 1.9728856086730957,
+      "learning_rate": 3.999524703507095e-05,
+      "loss": 0.5233,
+      "step": 14810
+    },
+    {
+      "epoch": 1.484399258776982,
+      "grad_norm": 2.2294158935546875,
+      "learning_rate": 3.998263003422159e-05,
+      "loss": 0.5812,
+      "step": 14820
+    },
+    {
+      "epoch": 1.4854009115039815,
+      "grad_norm": 2.4784843921661377,
+      "learning_rate": 3.997000707564877e-05,
+      "loss": 0.5809,
+      "step": 14830
+    },
+    {
+      "epoch": 1.486402564230981,
+      "grad_norm": 1.835519552230835,
+      "learning_rate": 3.995737816437192e-05,
+      "loss": 0.5403,
+      "step": 14840
+    },
+    {
+      "epoch": 1.4874042169579806,
+      "grad_norm": 2.5397074222564697,
+      "learning_rate": 3.994474330541282e-05,
+      "loss": 0.6364,
+      "step": 14850
+    },
+    {
+      "epoch": 1.4884058696849802,
+      "grad_norm": 2.101902961730957,
+      "learning_rate": 3.9932102503795616e-05,
+      "loss": 0.5697,
+      "step": 14860
+    },
+    {
+      "epoch": 1.4894075224119798,
+      "grad_norm": 2.234866142272949,
+      "learning_rate": 3.991945576454683e-05,
+      "loss": 0.4622,
+      "step": 14870
+    },
+    {
+      "epoch": 1.4904091751389794,
+      "grad_norm": 2.1409013271331787,
+      "learning_rate": 3.990680309269534e-05,
+      "loss": 0.5951,
+      "step": 14880
+    },
+    {
+      "epoch": 1.4914108278659788,
+      "grad_norm": 1.8268654346466064,
+      "learning_rate": 3.9894144493272376e-05,
+      "loss": 0.6112,
+      "step": 14890
+    },
+    {
+      "epoch": 1.4924124805929784,
+      "grad_norm": 2.455508232116699,
+      "learning_rate": 3.988147997131152e-05,
+      "loss": 0.6239,
+      "step": 14900
+    },
+    {
+      "epoch": 1.493414133319978,
+      "grad_norm": 2.3702023029327393,
+      "learning_rate": 3.986880953184874e-05,
+      "loss": 0.5383,
+      "step": 14910
+    },
+    {
+      "epoch": 1.4944157860469776,
+      "grad_norm": 2.136279821395874,
+      "learning_rate": 3.985613317992231e-05,
+      "loss": 0.6343,
+      "step": 14920
+    },
+    {
+      "epoch": 1.495417438773977,
+      "grad_norm": 2.7083253860473633,
+      "learning_rate": 3.98434509205729e-05,
+      "loss": 0.5874,
+      "step": 14930
+    },
+    {
+      "epoch": 1.4964190915009765,
+      "grad_norm": 2.658341407775879,
+      "learning_rate": 3.9830762758843496e-05,
+      "loss": 0.5574,
+      "step": 14940
+    },
+    {
+      "epoch": 1.497420744227976,
+      "grad_norm": 1.8282076120376587,
+      "learning_rate": 3.981806869977945e-05,
+      "loss": 0.5252,
+      "step": 14950
+    },
+    {
+      "epoch": 1.4984223969549757,
+      "grad_norm": 2.072915554046631,
+      "learning_rate": 3.980536874842846e-05,
+      "loss": 0.5425,
+      "step": 14960
+    },
+    {
+      "epoch": 1.4994240496819753,
+      "grad_norm": 2.2089407444000244,
+      "learning_rate": 3.979266290984055e-05,
+      "loss": 0.5542,
+      "step": 14970
+    },
+    {
+      "epoch": 1.5004257024089749,
+      "grad_norm": 1.9677778482437134,
+      "learning_rate": 3.97799511890681e-05,
+      "loss": 0.59,
+      "step": 14980
+    },
+    {
+      "epoch": 1.5014273551359745,
+      "grad_norm": 1.9966034889221191,
+      "learning_rate": 3.976723359116583e-05,
+      "loss": 0.616,
+      "step": 14990
+    },
+    {
+      "epoch": 1.502429007862974,
+      "grad_norm": 2.4626691341400146,
+      "learning_rate": 3.975451012119078e-05,
+      "loss": 0.5608,
+      "step": 15000
+    },
+    {
+      "epoch": 1.5034306605899734,
+      "grad_norm": 1.9694550037384033,
+      "learning_rate": 3.974178078420234e-05,
+      "loss": 0.5136,
+      "step": 15010
+    },
+    {
+      "epoch": 1.504432313316973,
+      "grad_norm": 2.206510066986084,
+      "learning_rate": 3.9729045585262235e-05,
+      "loss": 0.5868,
+      "step": 15020
+    },
+    {
+      "epoch": 1.5054339660439724,
+      "grad_norm": 1.8147963285446167,
+      "learning_rate": 3.9716304529434504e-05,
+      "loss": 0.5395,
+      "step": 15030
+    },
+    {
+      "epoch": 1.506435618770972,
+      "grad_norm": 2.563148021697998,
+      "learning_rate": 3.970355762178555e-05,
+      "loss": 0.5693,
+      "step": 15040
+    },
+    {
+      "epoch": 1.5074372714979716,
+      "grad_norm": 3.0714714527130127,
+      "learning_rate": 3.9690804867384046e-05,
+      "loss": 0.6563,
+      "step": 15050
+    },
+    {
+      "epoch": 1.5084389242249712,
+      "grad_norm": 2.1805124282836914,
+      "learning_rate": 3.967804627130105e-05,
+      "loss": 0.5471,
+      "step": 15060
+    },
+    {
+      "epoch": 1.5094405769519708,
+      "grad_norm": 2.3275880813598633,
+      "learning_rate": 3.9665281838609905e-05,
+      "loss": 0.4971,
+      "step": 15070
+    },
+    {
+      "epoch": 1.5104422296789703,
+      "grad_norm": 2.1147117614746094,
+      "learning_rate": 3.9652511574386286e-05,
+      "loss": 0.5394,
+      "step": 15080
+    },
+    {
+      "epoch": 1.51144388240597,
+      "grad_norm": 2.3367505073547363,
+      "learning_rate": 3.9639735483708195e-05,
+      "loss": 0.5847,
+      "step": 15090
+    },
+    {
+      "epoch": 1.5124455351329695,
+      "grad_norm": 2.4513463973999023,
+      "learning_rate": 3.9626953571655926e-05,
+      "loss": 0.5493,
+      "step": 15100
+    },
+    {
+      "epoch": 1.513447187859969,
+      "grad_norm": 2.5679237842559814,
+      "learning_rate": 3.961416584331212e-05,
+      "loss": 0.6568,
+      "step": 15110
+    },
+    {
+      "epoch": 1.5144488405869685,
+      "grad_norm": 2.4463155269622803,
+      "learning_rate": 3.960137230376171e-05,
+      "loss": 0.5491,
+      "step": 15120
+    },
+    {
+      "epoch": 1.515450493313968,
+      "grad_norm": 3.207108497619629,
+      "learning_rate": 3.958857295809195e-05,
+      "loss": 0.6069,
+      "step": 15130
+    },
+    {
+      "epoch": 1.5164521460409675,
+      "grad_norm": 2.2444918155670166,
+      "learning_rate": 3.957576781139238e-05,
+      "loss": 0.5662,
+      "step": 15140
+    },
+    {
+      "epoch": 1.517453798767967,
+      "grad_norm": 3.2347822189331055,
+      "learning_rate": 3.9562956868754884e-05,
+      "loss": 0.6197,
+      "step": 15150
+    },
+    {
+      "epoch": 1.5184554514949666,
+      "grad_norm": 2.22012996673584,
+      "learning_rate": 3.955014013527363e-05,
+      "loss": 0.6081,
+      "step": 15160
+    },
+    {
+      "epoch": 1.5194571042219662,
+      "grad_norm": 2.368472099304199,
+      "learning_rate": 3.9537317616045075e-05,
+      "loss": 0.4841,
+      "step": 15170
+    },
+    {
+      "epoch": 1.5204587569489658,
+      "grad_norm": 1.771808385848999,
+      "learning_rate": 3.952448931616801e-05,
+      "loss": 0.5171,
+      "step": 15180
+    },
+    {
+      "epoch": 1.5214604096759654,
+      "grad_norm": 2.661496877670288,
+      "learning_rate": 3.9511655240743494e-05,
+      "loss": 0.5657,
+      "step": 15190
+    },
+    {
+      "epoch": 1.522462062402965,
+      "grad_norm": 2.4911489486694336,
+      "learning_rate": 3.949881539487489e-05,
+      "loss": 0.5156,
+      "step": 15200
+    },
+    {
+      "epoch": 1.5234637151299646,
+      "grad_norm": 2.6235358715057373,
+      "learning_rate": 3.948596978366787e-05,
+      "loss": 0.5028,
+      "step": 15210
+    },
+    {
+      "epoch": 1.524465367856964,
+      "grad_norm": 2.052980661392212,
+      "learning_rate": 3.9473118412230406e-05,
+      "loss": 0.5424,
+      "step": 15220
+    },
+    {
+      "epoch": 1.5254670205839636,
+      "grad_norm": 2.1379051208496094,
+      "learning_rate": 3.9460261285672716e-05,
+      "loss": 0.6054,
+      "step": 15230
+    },
+    {
+      "epoch": 1.5264686733109631,
+      "grad_norm": 2.0419297218322754,
+      "learning_rate": 3.944739840910733e-05,
+      "loss": 0.5623,
+      "step": 15240
+    },
+    {
+      "epoch": 1.5274703260379625,
+      "grad_norm": 1.9566359519958496,
+      "learning_rate": 3.9434529787649096e-05,
+      "loss": 0.579,
+      "step": 15250
+    },
+    {
+      "epoch": 1.528471978764962,
+      "grad_norm": 2.9252963066101074,
+      "learning_rate": 3.9421655426415094e-05,
+      "loss": 0.5774,
+      "step": 15260
+    },
+    {
+      "epoch": 1.5294736314919617,
+      "grad_norm": 1.8522799015045166,
+      "learning_rate": 3.940877533052473e-05,
+      "loss": 0.546,
+      "step": 15270
+    },
+    {
+      "epoch": 1.5304752842189613,
+      "grad_norm": 1.8291130065917969,
+      "learning_rate": 3.939588950509966e-05,
+      "loss": 0.4684,
+      "step": 15280
+    },
+    {
+      "epoch": 1.5314769369459609,
+      "grad_norm": 2.20719313621521,
+      "learning_rate": 3.9382997955263826e-05,
+      "loss": 0.6081,
+      "step": 15290
+    },
+    {
+      "epoch": 1.5324785896729605,
+      "grad_norm": 2.3221805095672607,
+      "learning_rate": 3.937010068614346e-05,
+      "loss": 0.5657,
+      "step": 15300
+    },
+    {
+      "epoch": 1.53348024239996,
+      "grad_norm": 2.302234172821045,
+      "learning_rate": 3.935719770286706e-05,
+      "loss": 0.6058,
+      "step": 15310
+    },
+    {
+      "epoch": 1.5344818951269596,
+      "grad_norm": 1.9883896112442017,
+      "learning_rate": 3.934428901056538e-05,
+      "loss": 0.5818,
+      "step": 15320
+    },
+    {
+      "epoch": 1.535483547853959,
+      "grad_norm": 2.019928216934204,
+      "learning_rate": 3.9331374614371485e-05,
+      "loss": 0.5142,
+      "step": 15330
+    },
+    {
+      "epoch": 1.5364852005809586,
+      "grad_norm": 2.7909915447235107,
+      "learning_rate": 3.931845451942065e-05,
+      "loss": 0.5924,
+      "step": 15340
+    },
+    {
+      "epoch": 1.537486853307958,
+      "grad_norm": 2.2816619873046875,
+      "learning_rate": 3.930552873085047e-05,
+      "loss": 0.4988,
+      "step": 15350
+    },
+    {
+      "epoch": 1.5384885060349576,
+      "grad_norm": 2.1884818077087402,
+      "learning_rate": 3.929259725380077e-05,
+      "loss": 0.4728,
+      "step": 15360
+    },
+    {
+      "epoch": 1.5394901587619572,
+      "grad_norm": 1.680711269378662,
+      "learning_rate": 3.927966009341365e-05,
+      "loss": 0.5373,
+      "step": 15370
+    },
+    {
+      "epoch": 1.5404918114889568,
+      "grad_norm": 1.9891623258590698,
+      "learning_rate": 3.9266717254833475e-05,
+      "loss": 0.4977,
+      "step": 15380
+    },
+    {
+      "epoch": 1.5414934642159563,
+      "grad_norm": 2.236722946166992,
+      "learning_rate": 3.9253768743206867e-05,
+      "loss": 0.5669,
+      "step": 15390
+    },
+    {
+      "epoch": 1.542495116942956,
+      "grad_norm": 2.8090734481811523,
+      "learning_rate": 3.924081456368268e-05,
+      "loss": 0.6334,
+      "step": 15400
+    },
+    {
+      "epoch": 1.5434967696699555,
+      "grad_norm": 2.2476439476013184,
+      "learning_rate": 3.922785472141205e-05,
+      "loss": 0.5412,
+      "step": 15410
+    },
+    {
+      "epoch": 1.5444984223969551,
+      "grad_norm": 2.49711012840271,
+      "learning_rate": 3.9214889221548365e-05,
+      "loss": 0.5622,
+      "step": 15420
+    },
+    {
+      "epoch": 1.5455000751239545,
+      "grad_norm": 2.0667357444763184,
+      "learning_rate": 3.920191806924723e-05,
+      "loss": 0.5434,
+      "step": 15430
+    },
+    {
+      "epoch": 1.546501727850954,
+      "grad_norm": 2.0976450443267822,
+      "learning_rate": 3.9188941269666544e-05,
+      "loss": 0.5828,
+      "step": 15440
+    },
+    {
+      "epoch": 1.5475033805779537,
+      "grad_norm": 2.21097469329834,
+      "learning_rate": 3.9175958827966416e-05,
+      "loss": 0.5162,
+      "step": 15450
+    },
+    {
+      "epoch": 1.548505033304953,
+      "grad_norm": 2.5535941123962402,
+      "learning_rate": 3.9162970749309207e-05,
+      "loss": 0.6123,
+      "step": 15460
+    },
+    {
+      "epoch": 1.5495066860319526,
+      "grad_norm": 2.2607061862945557,
+      "learning_rate": 3.9149977038859534e-05,
+      "loss": 0.6232,
+      "step": 15470
+    },
+    {
+      "epoch": 1.5505083387589522,
+      "grad_norm": 1.9447360038757324,
+      "learning_rate": 3.913697770178423e-05,
+      "loss": 0.5005,
+      "step": 15480
+    },
+    {
+      "epoch": 1.5515099914859518,
+      "grad_norm": 1.894853115081787,
+      "learning_rate": 3.9123972743252394e-05,
+      "loss": 0.5458,
+      "step": 15490
+    },
+    {
+      "epoch": 1.5525116442129514,
+      "grad_norm": 1.628859281539917,
+      "learning_rate": 3.9110962168435315e-05,
+      "loss": 0.5476,
+      "step": 15500
+    },
+    {
+      "epoch": 1.553513296939951,
+      "grad_norm": 2.6020190715789795,
+      "learning_rate": 3.9097945982506584e-05,
+      "loss": 0.5331,
+      "step": 15510
+    },
+    {
+      "epoch": 1.5545149496669506,
+      "grad_norm": 2.6441216468811035,
+      "learning_rate": 3.908492419064196e-05,
+      "loss": 0.578,
+      "step": 15520
+    },
+    {
+      "epoch": 1.5555166023939502,
+      "grad_norm": 1.7780492305755615,
+      "learning_rate": 3.907189679801945e-05,
+      "loss": 0.501,
+      "step": 15530
+    },
+    {
+      "epoch": 1.5565182551209495,
+      "grad_norm": 2.1867527961730957,
+      "learning_rate": 3.90588638098193e-05,
+      "loss": 0.5687,
+      "step": 15540
+    },
+    {
+      "epoch": 1.5575199078479491,
+      "grad_norm": 2.3934149742126465,
+      "learning_rate": 3.904582523122398e-05,
+      "loss": 0.595,
+      "step": 15550
+    },
+    {
+      "epoch": 1.5585215605749485,
+      "grad_norm": 2.72482967376709,
+      "learning_rate": 3.9032781067418176e-05,
+      "loss": 0.5956,
+      "step": 15560
+    },
+    {
+      "epoch": 1.559523213301948,
+      "grad_norm": 2.589207410812378,
+      "learning_rate": 3.9019731323588785e-05,
+      "loss": 0.5599,
+      "step": 15570
+    },
+    {
+      "epoch": 1.5605248660289477,
+      "grad_norm": 2.120452404022217,
+      "learning_rate": 3.900667600492494e-05,
+      "loss": 0.5554,
+      "step": 15580
+    },
+    {
+      "epoch": 1.5615265187559473,
+      "grad_norm": 2.300694704055786,
+      "learning_rate": 3.8993615116617985e-05,
+      "loss": 0.5926,
+      "step": 15590
+    },
+    {
+      "epoch": 1.5625281714829469,
+      "grad_norm": 2.3215324878692627,
+      "learning_rate": 3.8980548663861485e-05,
+      "loss": 0.5243,
+      "step": 15600
+    },
+    {
+      "epoch": 1.5635298242099465,
+      "grad_norm": 2.2662432193756104,
+      "learning_rate": 3.8967476651851196e-05,
+      "loss": 0.5805,
+      "step": 15610
+    },
+    {
+      "epoch": 1.564531476936946,
+      "grad_norm": 2.3407816886901855,
+      "learning_rate": 3.895439908578511e-05,
+      "loss": 0.5136,
+      "step": 15620
+    },
+    {
+      "epoch": 1.5655331296639456,
+      "grad_norm": 2.640747547149658,
+      "learning_rate": 3.894131597086341e-05,
+      "loss": 0.5318,
+      "step": 15630
+    },
+    {
+      "epoch": 1.566534782390945,
+      "grad_norm": 2.426098585128784,
+      "learning_rate": 3.89282273122885e-05,
+      "loss": 0.5826,
+      "step": 15640
+    },
+    {
+      "epoch": 1.5675364351179446,
+      "grad_norm": 2.3364737033843994,
+      "learning_rate": 3.891513311526498e-05,
+      "loss": 0.4968,
+      "step": 15650
+    },
+    {
+      "epoch": 1.5685380878449442,
+      "grad_norm": 2.151834011077881,
+      "learning_rate": 3.890203338499965e-05,
+      "loss": 0.5438,
+      "step": 15660
+    },
+    {
+      "epoch": 1.5695397405719436,
+      "grad_norm": 2.491666555404663,
+      "learning_rate": 3.8888928126701515e-05,
+      "loss": 0.5787,
+      "step": 15670
+    },
+    {
+      "epoch": 1.5705413932989432,
+      "grad_norm": 2.0937247276306152,
+      "learning_rate": 3.887581734558177e-05,
+      "loss": 0.5191,
+      "step": 15680
+    },
+    {
+      "epoch": 1.5715430460259427,
+      "grad_norm": 2.5927858352661133,
+      "learning_rate": 3.886270104685382e-05,
+      "loss": 0.5813,
+      "step": 15690
+    },
+    {
+      "epoch": 1.5725446987529423,
+      "grad_norm": 1.6065778732299805,
+      "learning_rate": 3.884957923573325e-05,
+      "loss": 0.5401,
+      "step": 15700
+    },
+    {
+      "epoch": 1.573546351479942,
+      "grad_norm": 3.2194879055023193,
+      "learning_rate": 3.883645191743786e-05,
+      "loss": 0.5814,
+      "step": 15710
+    },
+    {
+      "epoch": 1.5745480042069415,
+      "grad_norm": 2.245382308959961,
+      "learning_rate": 3.88233190971876e-05,
+      "loss": 0.586,
+      "step": 15720
+    },
+    {
+      "epoch": 1.5755496569339411,
+      "grad_norm": 2.3317959308624268,
+      "learning_rate": 3.8810180780204645e-05,
+      "loss": 0.5584,
+      "step": 15730
+    },
+    {
+      "epoch": 1.5765513096609407,
+      "grad_norm": 2.8339760303497314,
+      "learning_rate": 3.8797036971713344e-05,
+      "loss": 0.5879,
+      "step": 15740
+    },
+    {
+      "epoch": 1.57755296238794,
+      "grad_norm": 2.1983373165130615,
+      "learning_rate": 3.8783887676940225e-05,
+      "loss": 0.6056,
+      "step": 15750
+    },
+    {
+      "epoch": 1.5785546151149397,
+      "grad_norm": 2.6038060188293457,
+      "learning_rate": 3.8770732901113994e-05,
+      "loss": 0.5227,
+      "step": 15760
+    },
+    {
+      "epoch": 1.5795562678419393,
+      "grad_norm": 3.548153877258301,
+      "learning_rate": 3.875757264946555e-05,
+      "loss": 0.5142,
+      "step": 15770
+    },
+    {
+      "epoch": 1.5805579205689386,
+      "grad_norm": 2.2880027294158936,
+      "learning_rate": 3.874440692722796e-05,
+      "loss": 0.5454,
+      "step": 15780
+    },
+    {
+      "epoch": 1.5815595732959382,
+      "grad_norm": 2.094909191131592,
+      "learning_rate": 3.8731235739636476e-05,
+      "loss": 0.5399,
+      "step": 15790
+    },
+    {
+      "epoch": 1.5825612260229378,
+      "grad_norm": 2.4892055988311768,
+      "learning_rate": 3.87180590919285e-05,
+      "loss": 0.5292,
+      "step": 15800
+    },
+    {
+      "epoch": 1.5835628787499374,
+      "grad_norm": 2.3183560371398926,
+      "learning_rate": 3.870487698934363e-05,
+      "loss": 0.5417,
+      "step": 15810
+    },
+    {
+      "epoch": 1.584564531476937,
+      "grad_norm": 2.3660850524902344,
+      "learning_rate": 3.869168943712362e-05,
+      "loss": 0.5847,
+      "step": 15820
+    },
+    {
+      "epoch": 1.5855661842039366,
+      "grad_norm": 2.3487000465393066,
+      "learning_rate": 3.8678496440512415e-05,
+      "loss": 0.5391,
+      "step": 15830
+    },
+    {
+      "epoch": 1.5865678369309362,
+      "grad_norm": 1.8460687398910522,
+      "learning_rate": 3.8665298004756075e-05,
+      "loss": 0.5236,
+      "step": 15840
+    },
+    {
+      "epoch": 1.5875694896579355,
+      "grad_norm": 2.310218334197998,
+      "learning_rate": 3.8652094135102865e-05,
+      "loss": 0.5462,
+      "step": 15850
+    },
+    {
+      "epoch": 1.5885711423849351,
+      "grad_norm": 2.01802659034729,
+      "learning_rate": 3.8638884836803205e-05,
+      "loss": 0.5358,
+      "step": 15860
+    },
+    {
+      "epoch": 1.5895727951119347,
+      "grad_norm": 1.9128429889678955,
+      "learning_rate": 3.8625670115109667e-05,
+      "loss": 0.6124,
+      "step": 15870
+    },
+    {
+      "epoch": 1.590574447838934,
+      "grad_norm": 2.289938449859619,
+      "learning_rate": 3.8612449975276965e-05,
+      "loss": 0.5916,
+      "step": 15880
+    },
+    {
+      "epoch": 1.5915761005659337,
+      "grad_norm": 3.483588457107544,
+      "learning_rate": 3.8599224422561997e-05,
+      "loss": 0.5341,
+      "step": 15890
+    },
+    {
+      "epoch": 1.5925777532929333,
+      "grad_norm": 2.2982561588287354,
+      "learning_rate": 3.858599346222379e-05,
+      "loss": 0.5238,
+      "step": 15900
+    },
+    {
+      "epoch": 1.5935794060199329,
+      "grad_norm": 3.378920555114746,
+      "learning_rate": 3.857275709952354e-05,
+      "loss": 0.5332,
+      "step": 15910
+    },
+    {
+      "epoch": 1.5945810587469325,
+      "grad_norm": 1.8684322834014893,
+      "learning_rate": 3.855951533972457e-05,
+      "loss": 0.5602,
+      "step": 15920
+    },
+    {
+      "epoch": 1.595582711473932,
+      "grad_norm": 2.5684077739715576,
+      "learning_rate": 3.854626818809237e-05,
+      "loss": 0.6066,
+      "step": 15930
+    },
+    {
+      "epoch": 1.5965843642009316,
+      "grad_norm": 2.02518630027771,
+      "learning_rate": 3.853301564989455e-05,
+      "loss": 0.5548,
+      "step": 15940
+    },
+    {
+      "epoch": 1.5975860169279312,
+      "grad_norm": 2.17681884765625,
+      "learning_rate": 3.8519757730400894e-05,
+      "loss": 0.5176,
+      "step": 15950
+    },
+    {
+      "epoch": 1.5985876696549306,
+      "grad_norm": 2.2103607654571533,
+      "learning_rate": 3.85064944348833e-05,
+      "loss": 0.5658,
+      "step": 15960
+    },
+    {
+      "epoch": 1.5995893223819302,
+      "grad_norm": 1.7236950397491455,
+      "learning_rate": 3.849322576861582e-05,
+      "loss": 0.5217,
+      "step": 15970
+    },
+    {
+      "epoch": 1.6005909751089298,
+      "grad_norm": 1.8692692518234253,
+      "learning_rate": 3.847995173687461e-05,
+      "loss": 0.6032,
+      "step": 15980
+    },
+    {
+      "epoch": 1.6015926278359292,
+      "grad_norm": 2.6671016216278076,
+      "learning_rate": 3.8466672344938005e-05,
+      "loss": 0.6187,
+      "step": 15990
+    },
+    {
+      "epoch": 1.6025942805629287,
+      "grad_norm": 2.8001585006713867,
+      "learning_rate": 3.845338759808644e-05,
+      "loss": 0.5646,
+      "step": 16000
+    },
+    {
+      "epoch": 1.6035959332899283,
+      "grad_norm": 2.768728733062744,
+      "learning_rate": 3.844009750160249e-05,
+      "loss": 0.6015,
+      "step": 16010
+    },
+    {
+      "epoch": 1.604597586016928,
+      "grad_norm": 3.4323201179504395,
+      "learning_rate": 3.842680206077086e-05,
+      "loss": 0.5414,
+      "step": 16020
+    },
+    {
+      "epoch": 1.6055992387439275,
+      "grad_norm": 2.223552942276001,
+      "learning_rate": 3.841350128087837e-05,
+      "loss": 0.5509,
+      "step": 16030
+    },
+    {
+      "epoch": 1.606600891470927,
+      "grad_norm": 2.1531848907470703,
+      "learning_rate": 3.840019516721398e-05,
+      "loss": 0.6104,
+      "step": 16040
+    },
+    {
+      "epoch": 1.6076025441979267,
+      "grad_norm": 2.558938503265381,
+      "learning_rate": 3.8386883725068745e-05,
+      "loss": 0.5257,
+      "step": 16050
+    },
+    {
+      "epoch": 1.6086041969249263,
+      "grad_norm": 2.4752094745635986,
+      "learning_rate": 3.837356695973586e-05,
+      "loss": 0.5968,
+      "step": 16060
+    },
+    {
+      "epoch": 1.6096058496519257,
+      "grad_norm": 2.3830080032348633,
+      "learning_rate": 3.836024487651064e-05,
+      "loss": 0.5585,
+      "step": 16070
+    },
+    {
+      "epoch": 1.6106075023789252,
+      "grad_norm": 2.419825315475464,
+      "learning_rate": 3.834691748069049e-05,
+      "loss": 0.5552,
+      "step": 16080
+    },
+    {
+      "epoch": 1.6116091551059246,
+      "grad_norm": 2.2571637630462646,
+      "learning_rate": 3.833358477757496e-05,
+      "loss": 0.5709,
+      "step": 16090
+    },
+    {
+      "epoch": 1.6126108078329242,
+      "grad_norm": 2.4724700450897217,
+      "learning_rate": 3.8320246772465674e-05,
+      "loss": 0.553,
+      "step": 16100
+    },
+    {
+      "epoch": 1.6136124605599238,
+      "grad_norm": 2.4371488094329834,
+      "learning_rate": 3.8306903470666385e-05,
+      "loss": 0.5119,
+      "step": 16110
+    },
+    {
+      "epoch": 1.6146141132869234,
+      "grad_norm": 2.0440852642059326,
+      "learning_rate": 3.829355487748297e-05,
+      "loss": 0.6072,
+      "step": 16120
+    },
+    {
+      "epoch": 1.615615766013923,
+      "grad_norm": 1.7289458513259888,
+      "learning_rate": 3.828020099822338e-05,
+      "loss": 0.5656,
+      "step": 16130
+    },
+    {
+      "epoch": 1.6166174187409226,
+      "grad_norm": 2.366450071334839,
+      "learning_rate": 3.826684183819768e-05,
+      "loss": 0.5754,
+      "step": 16140
+    },
+    {
+      "epoch": 1.6176190714679222,
+      "grad_norm": 1.9177955389022827,
+      "learning_rate": 3.825347740271802e-05,
+      "loss": 0.5357,
+      "step": 16150
+    },
+    {
+      "epoch": 1.6186207241949218,
+      "grad_norm": 2.2635581493377686,
+      "learning_rate": 3.824010769709868e-05,
+      "loss": 0.5734,
+      "step": 16160
+    },
+    {
+      "epoch": 1.6196223769219211,
+      "grad_norm": 2.226444959640503,
+      "learning_rate": 3.8226732726656005e-05,
+      "loss": 0.5329,
+      "step": 16170
+    },
+    {
+      "epoch": 1.6206240296489207,
+      "grad_norm": 1.7379987239837646,
+      "learning_rate": 3.821335249670845e-05,
+      "loss": 0.5442,
+      "step": 16180
+    },
+    {
+      "epoch": 1.6216256823759203,
+      "grad_norm": 2.6636769771575928,
+      "learning_rate": 3.8199967012576566e-05,
+      "loss": 0.5654,
+      "step": 16190
+    },
+    {
+      "epoch": 1.6226273351029197,
+      "grad_norm": 2.491895914077759,
+      "learning_rate": 3.818657627958296e-05,
+      "loss": 0.5794,
+      "step": 16200
+    },
+    {
+      "epoch": 1.6236289878299193,
+      "grad_norm": 2.4376542568206787,
+      "learning_rate": 3.817318030305238e-05,
+      "loss": 0.5393,
+      "step": 16210
+    },
+    {
+      "epoch": 1.6246306405569189,
+      "grad_norm": 2.03060245513916,
+      "learning_rate": 3.815977908831161e-05,
+      "loss": 0.5238,
+      "step": 16220
+    },
+    {
+      "epoch": 1.6256322932839185,
+      "grad_norm": 1.902113914489746,
+      "learning_rate": 3.8146372640689536e-05,
+      "loss": 0.5333,
+      "step": 16230
+    },
+    {
+      "epoch": 1.626633946010918,
+      "grad_norm": 2.258549928665161,
+      "learning_rate": 3.8132960965517135e-05,
+      "loss": 0.5551,
+      "step": 16240
+    },
+    {
+      "epoch": 1.6276355987379176,
+      "grad_norm": 1.7955327033996582,
+      "learning_rate": 3.811954406812744e-05,
+      "loss": 0.5322,
+      "step": 16250
+    },
+    {
+      "epoch": 1.6286372514649172,
+      "grad_norm": 1.9605530500411987,
+      "learning_rate": 3.810612195385558e-05,
+      "loss": 0.5111,
+      "step": 16260
+    },
+    {
+      "epoch": 1.6296389041919168,
+      "grad_norm": 2.7651565074920654,
+      "learning_rate": 3.8092694628038764e-05,
+      "loss": 0.472,
+      "step": 16270
+    },
+    {
+      "epoch": 1.6306405569189162,
+      "grad_norm": 2.146113395690918,
+      "learning_rate": 3.807926209601624e-05,
+      "loss": 0.6197,
+      "step": 16280
+    },
+    {
+      "epoch": 1.6316422096459158,
+      "grad_norm": 2.629794120788574,
+      "learning_rate": 3.806582436312936e-05,
+      "loss": 0.5674,
+      "step": 16290
+    },
+    {
+      "epoch": 1.6326438623729151,
+      "grad_norm": 1.815111756324768,
+      "learning_rate": 3.805238143472154e-05,
+      "loss": 0.5137,
+      "step": 16300
+    },
+    {
+      "epoch": 1.6336455150999147,
+      "grad_norm": 2.728011131286621,
+      "learning_rate": 3.8038933316138225e-05,
+      "loss": 0.5547,
+      "step": 16310
+    },
+    {
+      "epoch": 1.6346471678269143,
+      "grad_norm": 1.9159232378005981,
+      "learning_rate": 3.802548001272698e-05,
+      "loss": 0.5681,
+      "step": 16320
+    },
+    {
+      "epoch": 1.635648820553914,
+      "grad_norm": 2.153987407684326,
+      "learning_rate": 3.801202152983738e-05,
+      "loss": 0.5581,
+      "step": 16330
+    },
+    {
+      "epoch": 1.6366504732809135,
+      "grad_norm": 2.767010450363159,
+      "learning_rate": 3.7998557872821104e-05,
+      "loss": 0.5055,
+      "step": 16340
+    },
+    {
+      "epoch": 1.637652126007913,
+      "grad_norm": 2.1452322006225586,
+      "learning_rate": 3.798508904703186e-05,
+      "loss": 0.5336,
+      "step": 16350
+    },
+    {
+      "epoch": 1.6386537787349127,
+      "grad_norm": 1.8051600456237793,
+      "learning_rate": 3.797161505782543e-05,
+      "loss": 0.566,
+      "step": 16360
+    },
+    {
+      "epoch": 1.6396554314619123,
+      "grad_norm": 2.4228758811950684,
+      "learning_rate": 3.795813591055961e-05,
+      "loss": 0.5281,
+      "step": 16370
+    },
+    {
+      "epoch": 1.6406570841889117,
+      "grad_norm": 2.133131742477417,
+      "learning_rate": 3.794465161059431e-05,
+      "loss": 0.5489,
+      "step": 16380
+    },
+    {
+      "epoch": 1.6416587369159112,
+      "grad_norm": 2.345156192779541,
+      "learning_rate": 3.793116216329143e-05,
+      "loss": 0.5207,
+      "step": 16390
+    },
+    {
+      "epoch": 1.6426603896429108,
+      "grad_norm": 2.2094056606292725,
+      "learning_rate": 3.791766757401495e-05,
+      "loss": 0.5303,
+      "step": 16400
+    },
+    {
+      "epoch": 1.6436620423699102,
+      "grad_norm": 2.3264756202697754,
+      "learning_rate": 3.790416784813088e-05,
+      "loss": 0.5847,
+      "step": 16410
+    },
+    {
+      "epoch": 1.6446636950969098,
+      "grad_norm": 2.0210537910461426,
+      "learning_rate": 3.7890662991007294e-05,
+      "loss": 0.6641,
+      "step": 16420
+    },
+    {
+      "epoch": 1.6456653478239094,
+      "grad_norm": 2.2783915996551514,
+      "learning_rate": 3.7877153008014275e-05,
+      "loss": 0.6113,
+      "step": 16430
+    },
+    {
+      "epoch": 1.646667000550909,
+      "grad_norm": 2.1935906410217285,
+      "learning_rate": 3.7863637904523956e-05,
+      "loss": 0.5113,
+      "step": 16440
+    },
+    {
+      "epoch": 1.6476686532779086,
+      "grad_norm": 2.478006601333618,
+      "learning_rate": 3.7850117685910535e-05,
+      "loss": 0.5166,
+      "step": 16450
+    },
+    {
+      "epoch": 1.6486703060049082,
+      "grad_norm": 2.1115570068359375,
+      "learning_rate": 3.783659235755019e-05,
+      "loss": 0.5136,
+      "step": 16460
+    },
+    {
+      "epoch": 1.6496719587319078,
+      "grad_norm": 2.320053815841675,
+      "learning_rate": 3.782306192482119e-05,
+      "loss": 0.5142,
+      "step": 16470
+    },
+    {
+      "epoch": 1.6506736114589073,
+      "grad_norm": 3.074125051498413,
+      "learning_rate": 3.7809526393103785e-05,
+      "loss": 0.5026,
+      "step": 16480
+    },
+    {
+      "epoch": 1.6516752641859067,
+      "grad_norm": 2.644097328186035,
+      "learning_rate": 3.779598576778026e-05,
+      "loss": 0.5428,
+      "step": 16490
+    },
+    {
+      "epoch": 1.6526769169129063,
+      "grad_norm": 2.510568857192993,
+      "learning_rate": 3.7782440054234966e-05,
+      "loss": 0.5376,
+      "step": 16500
+    },
+    {
+      "epoch": 1.653678569639906,
+      "grad_norm": 1.993364691734314,
+      "learning_rate": 3.7768889257854224e-05,
+      "loss": 0.5449,
+      "step": 16510
+    },
+    {
+      "epoch": 1.6546802223669053,
+      "grad_norm": 2.8113162517547607,
+      "learning_rate": 3.775533338402641e-05,
+      "loss": 0.5261,
+      "step": 16520
+    },
+    {
+      "epoch": 1.6556818750939049,
+      "grad_norm": 2.3818821907043457,
+      "learning_rate": 3.7741772438141916e-05,
+      "loss": 0.6036,
+      "step": 16530
+    },
+    {
+      "epoch": 1.6566835278209044,
+      "grad_norm": 2.2668466567993164,
+      "learning_rate": 3.7728206425593126e-05,
+      "loss": 0.5026,
+      "step": 16540
+    },
+    {
+      "epoch": 1.657685180547904,
+      "grad_norm": 2.132610321044922,
+      "learning_rate": 3.771463535177447e-05,
+      "loss": 0.5779,
+      "step": 16550
+    },
+    {
+      "epoch": 1.6586868332749036,
+      "grad_norm": 2.991370439529419,
+      "learning_rate": 3.770105922208239e-05,
+      "loss": 0.6101,
+      "step": 16560
+    },
+    {
+      "epoch": 1.6596884860019032,
+      "grad_norm": 2.392225742340088,
+      "learning_rate": 3.768747804191529e-05,
+      "loss": 0.5587,
+      "step": 16570
+    },
+    {
+      "epoch": 1.6606901387289028,
+      "grad_norm": 2.6654226779937744,
+      "learning_rate": 3.767389181667365e-05,
+      "loss": 0.4945,
+      "step": 16580
+    },
+    {
+      "epoch": 1.6616917914559024,
+      "grad_norm": 2.315359115600586,
+      "learning_rate": 3.766030055175991e-05,
+      "loss": 0.6415,
+      "step": 16590
+    },
+    {
+      "epoch": 1.6626934441829018,
+      "grad_norm": 3.1376960277557373,
+      "learning_rate": 3.764670425257853e-05,
+      "loss": 0.5494,
+      "step": 16600
+    },
+    {
+      "epoch": 1.6636950969099014,
+      "grad_norm": 2.255483865737915,
+      "learning_rate": 3.763310292453597e-05,
+      "loss": 0.5318,
+      "step": 16610
+    },
+    {
+      "epoch": 1.6646967496369007,
+      "grad_norm": 1.7115904092788696,
+      "learning_rate": 3.761949657304068e-05,
+      "loss": 0.5176,
+      "step": 16620
+    },
+    {
+      "epoch": 1.6656984023639003,
+      "grad_norm": 1.925130009651184,
+      "learning_rate": 3.760588520350315e-05,
+      "loss": 0.5681,
+      "step": 16630
+    },
+    {
+      "epoch": 1.6667000550909,
+      "grad_norm": 3.0387139320373535,
+      "learning_rate": 3.75922688213358e-05,
+      "loss": 0.4607,
+      "step": 16640
+    },
+    {
+      "epoch": 1.6677017078178995,
+      "grad_norm": 1.6796802282333374,
+      "learning_rate": 3.7578647431953086e-05,
+      "loss": 0.6343,
+      "step": 16650
+    },
+    {
+      "epoch": 1.668703360544899,
+      "grad_norm": 2.335080146789551,
+      "learning_rate": 3.756502104077145e-05,
+      "loss": 0.5031,
+      "step": 16660
+    },
+    {
+      "epoch": 1.6697050132718987,
+      "grad_norm": 1.8291679620742798,
+      "learning_rate": 3.75513896532093e-05,
+      "loss": 0.5351,
+      "step": 16670
+    },
+    {
+      "epoch": 1.6707066659988983,
+      "grad_norm": 2.4488725662231445,
+      "learning_rate": 3.753775327468708e-05,
+      "loss": 0.5344,
+      "step": 16680
+    },
+    {
+      "epoch": 1.6717083187258979,
+      "grad_norm": 2.563654661178589,
+      "learning_rate": 3.7524111910627157e-05,
+      "loss": 0.6013,
+      "step": 16690
+    },
+    {
+      "epoch": 1.6727099714528972,
+      "grad_norm": 2.365469455718994,
+      "learning_rate": 3.7510465566453924e-05,
+      "loss": 0.4762,
+      "step": 16700
+    },
+    {
+      "epoch": 1.6737116241798968,
+      "grad_norm": 2.369333028793335,
+      "learning_rate": 3.749681424759374e-05,
+      "loss": 0.6089,
+      "step": 16710
+    },
+    {
+      "epoch": 1.6747132769068964,
+      "grad_norm": 3.0419299602508545,
+      "learning_rate": 3.748315795947495e-05,
+      "loss": 0.5973,
+      "step": 16720
+    },
+    {
+      "epoch": 1.6757149296338958,
+      "grad_norm": 2.051492929458618,
+      "learning_rate": 3.7469496707527854e-05,
+      "loss": 0.5512,
+      "step": 16730
+    },
+    {
+      "epoch": 1.6767165823608954,
+      "grad_norm": 2.6813406944274902,
+      "learning_rate": 3.745583049718475e-05,
+      "loss": 0.5789,
+      "step": 16740
+    },
+    {
+      "epoch": 1.677718235087895,
+      "grad_norm": 1.9882440567016602,
+      "learning_rate": 3.74421593338799e-05,
+      "loss": 0.5064,
+      "step": 16750
+    },
+    {
+      "epoch": 1.6787198878148946,
+      "grad_norm": 2.8968048095703125,
+      "learning_rate": 3.742848322304952e-05,
+      "loss": 0.5373,
+      "step": 16760
+    },
+    {
+      "epoch": 1.6797215405418942,
+      "grad_norm": 2.5871849060058594,
+      "learning_rate": 3.741480217013182e-05,
+      "loss": 0.5037,
+      "step": 16770
+    },
+    {
+      "epoch": 1.6807231932688937,
+      "grad_norm": 2.5383193492889404,
+      "learning_rate": 3.7401116180566954e-05,
+      "loss": 0.4799,
+      "step": 16780
+    },
+    {
+      "epoch": 1.6817248459958933,
+      "grad_norm": 2.1125566959381104,
+      "learning_rate": 3.738742525979705e-05,
+      "loss": 0.469,
+      "step": 16790
+    },
+    {
+      "epoch": 1.682726498722893,
+      "grad_norm": 2.907052755355835,
+      "learning_rate": 3.737372941326619e-05,
+      "loss": 0.6347,
+      "step": 16800
+    },
+    {
+      "epoch": 1.6837281514498923,
+      "grad_norm": 2.6465771198272705,
+      "learning_rate": 3.736002864642042e-05,
+      "loss": 0.5437,
+      "step": 16810
+    },
+    {
+      "epoch": 1.6847298041768919,
+      "grad_norm": 2.14072322845459,
+      "learning_rate": 3.7346322964707744e-05,
+      "loss": 0.4944,
+      "step": 16820
+    },
+    {
+      "epoch": 1.6857314569038913,
+      "grad_norm": 2.4284651279449463,
+      "learning_rate": 3.733261237357812e-05,
+      "loss": 0.5874,
+      "step": 16830
+    },
+    {
+      "epoch": 1.6867331096308908,
+      "grad_norm": 2.144977569580078,
+      "learning_rate": 3.731889687848344e-05,
+      "loss": 0.5592,
+      "step": 16840
+    },
+    {
+      "epoch": 1.6877347623578904,
+      "grad_norm": 1.864293098449707,
+      "learning_rate": 3.730517648487758e-05,
+      "loss": 0.5409,
+      "step": 16850
+    },
+    {
+      "epoch": 1.68873641508489,
+      "grad_norm": 2.6154680252075195,
+      "learning_rate": 3.7291451198216334e-05,
+      "loss": 0.5337,
+      "step": 16860
+    },
+    {
+      "epoch": 1.6897380678118896,
+      "grad_norm": 3.94486403465271,
+      "learning_rate": 3.727772102395745e-05,
+      "loss": 0.4953,
+      "step": 16870
+    },
+    {
+      "epoch": 1.6907397205388892,
+      "grad_norm": 2.5961368083953857,
+      "learning_rate": 3.726398596756063e-05,
+      "loss": 0.5295,
+      "step": 16880
+    },
+    {
+      "epoch": 1.6917413732658888,
+      "grad_norm": 2.7672832012176514,
+      "learning_rate": 3.725024603448751e-05,
+      "loss": 0.5942,
+      "step": 16890
+    },
+    {
+      "epoch": 1.6927430259928884,
+      "grad_norm": 3.2715678215026855,
+      "learning_rate": 3.723650123020166e-05,
+      "loss": 0.6036,
+      "step": 16900
+    },
+    {
+      "epoch": 1.6937446787198878,
+      "grad_norm": 2.5985188484191895,
+      "learning_rate": 3.72227515601686e-05,
+      "loss": 0.5415,
+      "step": 16910
+    },
+    {
+      "epoch": 1.6947463314468874,
+      "grad_norm": 2.263279676437378,
+      "learning_rate": 3.7208997029855764e-05,
+      "loss": 0.5924,
+      "step": 16920
+    },
+    {
+      "epoch": 1.695747984173887,
+      "grad_norm": 2.0555026531219482,
+      "learning_rate": 3.7195237644732545e-05,
+      "loss": 0.5864,
+      "step": 16930
+    },
+    {
+      "epoch": 1.6967496369008863,
+      "grad_norm": 2.1603214740753174,
+      "learning_rate": 3.718147341027024e-05,
+      "loss": 0.5368,
+      "step": 16940
+    },
+    {
+      "epoch": 1.697751289627886,
+      "grad_norm": 3.1869144439697266,
+      "learning_rate": 3.71677043319421e-05,
+      "loss": 0.6791,
+      "step": 16950
+    },
+    {
+      "epoch": 1.6987529423548855,
+      "grad_norm": 2.106552839279175,
+      "learning_rate": 3.715393041522328e-05,
+      "loss": 0.5239,
+      "step": 16960
+    },
+    {
+      "epoch": 1.699754595081885,
+      "grad_norm": 2.6369104385375977,
+      "learning_rate": 3.714015166559087e-05,
+      "loss": 0.5722,
+      "step": 16970
+    },
+    {
+      "epoch": 1.7007562478088847,
+      "grad_norm": 2.095924139022827,
+      "learning_rate": 3.7126368088523884e-05,
+      "loss": 0.597,
+      "step": 16980
+    },
+    {
+      "epoch": 1.7017579005358843,
+      "grad_norm": 2.8147194385528564,
+      "learning_rate": 3.711257968950325e-05,
+      "loss": 0.5293,
+      "step": 16990
+    },
+    {
+      "epoch": 1.7027595532628839,
+      "grad_norm": 2.2241532802581787,
+      "learning_rate": 3.709878647401181e-05,
+      "loss": 0.5732,
+      "step": 17000
+    },
+    {
+      "epoch": 1.7037612059898835,
+      "grad_norm": 2.0934462547302246,
+      "learning_rate": 3.708498844753433e-05,
+      "loss": 0.5241,
+      "step": 17010
+    },
+    {
+      "epoch": 1.7047628587168828,
+      "grad_norm": 1.8266611099243164,
+      "learning_rate": 3.707118561555748e-05,
+      "loss": 0.4946,
+      "step": 17020
+    },
+    {
+      "epoch": 1.7057645114438824,
+      "grad_norm": 1.8321797847747803,
+      "learning_rate": 3.705737798356985e-05,
+      "loss": 0.6098,
+      "step": 17030
+    },
+    {
+      "epoch": 1.706766164170882,
+      "grad_norm": 2.4191246032714844,
+      "learning_rate": 3.704356555706195e-05,
+      "loss": 0.5116,
+      "step": 17040
+    },
+    {
+      "epoch": 1.7077678168978814,
+      "grad_norm": 2.142313003540039,
+      "learning_rate": 3.702974834152616e-05,
+      "loss": 0.5581,
+      "step": 17050
+    },
+    {
+      "epoch": 1.708769469624881,
+      "grad_norm": 2.1364362239837646,
+      "learning_rate": 3.70159263424568e-05,
+      "loss": 0.4971,
+      "step": 17060
+    },
+    {
+      "epoch": 1.7097711223518806,
+      "grad_norm": 2.3862712383270264,
+      "learning_rate": 3.7002099565350053e-05,
+      "loss": 0.5418,
+      "step": 17070
+    },
+    {
+      "epoch": 1.7107727750788801,
+      "grad_norm": 2.108258008956909,
+      "learning_rate": 3.698826801570406e-05,
+      "loss": 0.5097,
+      "step": 17080
+    },
+    {
+      "epoch": 1.7117744278058797,
+      "grad_norm": 2.0370752811431885,
+      "learning_rate": 3.6974431699018806e-05,
+      "loss": 0.5491,
+      "step": 17090
+    },
+    {
+      "epoch": 1.7127760805328793,
+      "grad_norm": 1.8179024457931519,
+      "learning_rate": 3.69605906207962e-05,
+      "loss": 0.5095,
+      "step": 17100
+    },
+    {
+      "epoch": 1.713777733259879,
+      "grad_norm": 2.070310592651367,
+      "learning_rate": 3.694674478654003e-05,
+      "loss": 0.579,
+      "step": 17110
+    },
+    {
+      "epoch": 1.7147793859868785,
+      "grad_norm": 2.2030398845672607,
+      "learning_rate": 3.693289420175599e-05,
+      "loss": 0.5857,
+      "step": 17120
+    },
+    {
+      "epoch": 1.7157810387138779,
+      "grad_norm": 2.334590196609497,
+      "learning_rate": 3.691903887195165e-05,
+      "loss": 0.5118,
+      "step": 17130
+    },
+    {
+      "epoch": 1.7167826914408775,
+      "grad_norm": 2.140634298324585,
+      "learning_rate": 3.690517880263647e-05,
+      "loss": 0.5775,
+      "step": 17140
+    },
+    {
+      "epoch": 1.7177843441678768,
+      "grad_norm": 1.9049276113510132,
+      "learning_rate": 3.68913139993218e-05,
+      "loss": 0.5149,
+      "step": 17150
+    },
+    {
+      "epoch": 1.7187859968948764,
+      "grad_norm": 2.7127888202667236,
+      "learning_rate": 3.687744446752086e-05,
+      "loss": 0.5779,
+      "step": 17160
+    },
+    {
+      "epoch": 1.719787649621876,
+      "grad_norm": 2.346618413925171,
+      "learning_rate": 3.686357021274877e-05,
+      "loss": 0.5229,
+      "step": 17170
+    },
+    {
+      "epoch": 1.7207893023488756,
+      "grad_norm": 2.411379337310791,
+      "learning_rate": 3.684969124052251e-05,
+      "loss": 0.5378,
+      "step": 17180
+    },
+    {
+      "epoch": 1.7217909550758752,
+      "grad_norm": 2.139772891998291,
+      "learning_rate": 3.683580755636094e-05,
+      "loss": 0.5433,
+      "step": 17190
+    },
+    {
+      "epoch": 1.7227926078028748,
+      "grad_norm": 2.1813013553619385,
+      "learning_rate": 3.682191916578481e-05,
+      "loss": 0.5352,
+      "step": 17200
+    },
+    {
+      "epoch": 1.7237942605298744,
+      "grad_norm": 2.3435051441192627,
+      "learning_rate": 3.680802607431673e-05,
+      "loss": 0.5517,
+      "step": 17210
+    },
+    {
+      "epoch": 1.724795913256874,
+      "grad_norm": 2.307100534439087,
+      "learning_rate": 3.6794128287481136e-05,
+      "loss": 0.5181,
+      "step": 17220
+    },
+    {
+      "epoch": 1.7257975659838733,
+      "grad_norm": 1.8775631189346313,
+      "learning_rate": 3.6780225810804426e-05,
+      "loss": 0.5047,
+      "step": 17230
+    },
+    {
+      "epoch": 1.726799218710873,
+      "grad_norm": 2.518465042114258,
+      "learning_rate": 3.676631864981478e-05,
+      "loss": 0.4953,
+      "step": 17240
+    },
+    {
+      "epoch": 1.7278008714378725,
+      "grad_norm": 2.4192512035369873,
+      "learning_rate": 3.675240681004227e-05,
+      "loss": 0.5995,
+      "step": 17250
+    },
+    {
+      "epoch": 1.728802524164872,
+      "grad_norm": 2.5736687183380127,
+      "learning_rate": 3.673849029701883e-05,
+      "loss": 0.6249,
+      "step": 17260
+    },
+    {
+      "epoch": 1.7298041768918715,
+      "grad_norm": 1.8964473009109497,
+      "learning_rate": 3.672456911627826e-05,
+      "loss": 0.6039,
+      "step": 17270
+    },
+    {
+      "epoch": 1.730805829618871,
+      "grad_norm": 1.9748671054840088,
+      "learning_rate": 3.6710643273356206e-05,
+      "loss": 0.5387,
+      "step": 17280
+    },
+    {
+      "epoch": 1.7318074823458707,
+      "grad_norm": 2.3679747581481934,
+      "learning_rate": 3.669671277379016e-05,
+      "loss": 0.5977,
+      "step": 17290
+    },
+    {
+      "epoch": 1.7328091350728703,
+      "grad_norm": 2.5027241706848145,
+      "learning_rate": 3.6682777623119474e-05,
+      "loss": 0.5304,
+      "step": 17300
+    },
+    {
+      "epoch": 1.7338107877998699,
+      "grad_norm": 1.9013116359710693,
+      "learning_rate": 3.666883782688535e-05,
+      "loss": 0.5139,
+      "step": 17310
+    },
+    {
+      "epoch": 1.7348124405268694,
+      "grad_norm": 2.220552682876587,
+      "learning_rate": 3.665489339063085e-05,
+      "loss": 0.5624,
+      "step": 17320
+    },
+    {
+      "epoch": 1.735814093253869,
+      "grad_norm": 1.7051708698272705,
+      "learning_rate": 3.664094431990085e-05,
+      "loss": 0.4726,
+      "step": 17330
+    },
+    {
+      "epoch": 1.7368157459808684,
+      "grad_norm": 1.9945650100708008,
+      "learning_rate": 3.662699062024209e-05,
+      "loss": 0.5907,
+      "step": 17340
+    },
+    {
+      "epoch": 1.737817398707868,
+      "grad_norm": 2.306936502456665,
+      "learning_rate": 3.661303229720316e-05,
+      "loss": 0.5114,
+      "step": 17350
+    },
+    {
+      "epoch": 1.7388190514348674,
+      "grad_norm": 2.131807565689087,
+      "learning_rate": 3.659906935633446e-05,
+      "loss": 0.5478,
+      "step": 17360
+    },
+    {
+      "epoch": 1.739820704161867,
+      "grad_norm": 2.530008554458618,
+      "learning_rate": 3.658510180318826e-05,
+      "loss": 0.558,
+      "step": 17370
+    },
+    {
+      "epoch": 1.7408223568888666,
+      "grad_norm": 1.6221373081207275,
+      "learning_rate": 3.657112964331862e-05,
+      "loss": 0.5436,
+      "step": 17380
+    },
+    {
+      "epoch": 1.7418240096158661,
+      "grad_norm": 2.6965410709381104,
+      "learning_rate": 3.6557152882281497e-05,
+      "loss": 0.5593,
+      "step": 17390
+    },
+    {
+      "epoch": 1.7428256623428657,
+      "grad_norm": 2.156054735183716,
+      "learning_rate": 3.65431715256346e-05,
+      "loss": 0.522,
+      "step": 17400
+    },
+    {
+      "epoch": 1.7438273150698653,
+      "grad_norm": 2.559577465057373,
+      "learning_rate": 3.652918557893753e-05,
+      "loss": 0.5416,
+      "step": 17410
+    },
+    {
+      "epoch": 1.744828967796865,
+      "grad_norm": 2.3171913623809814,
+      "learning_rate": 3.651519504775167e-05,
+      "loss": 0.5473,
+      "step": 17420
+    },
+    {
+      "epoch": 1.7458306205238645,
+      "grad_norm": 2.0408565998077393,
+      "learning_rate": 3.650119993764025e-05,
+      "loss": 0.4801,
+      "step": 17430
+    },
+    {
+      "epoch": 1.7468322732508639,
+      "grad_norm": 2.4444737434387207,
+      "learning_rate": 3.648720025416832e-05,
+      "loss": 0.5292,
+      "step": 17440
+    },
+    {
+      "epoch": 1.7478339259778635,
+      "grad_norm": 3.833693027496338,
+      "learning_rate": 3.647319600290273e-05,
+      "loss": 0.5295,
+      "step": 17450
+    },
+    {
+      "epoch": 1.748835578704863,
+      "grad_norm": 2.1103272438049316,
+      "learning_rate": 3.6459187189412175e-05,
+      "loss": 0.5395,
+      "step": 17460
+    },
+    {
+      "epoch": 1.7498372314318624,
+      "grad_norm": 2.4942052364349365,
+      "learning_rate": 3.6445173819267133e-05,
+      "loss": 0.5076,
+      "step": 17470
+    },
+    {
+      "epoch": 1.750838884158862,
+      "grad_norm": 2.432206869125366,
+      "learning_rate": 3.643115589803992e-05,
+      "loss": 0.5246,
+      "step": 17480
+    },
+    {
+      "epoch": 1.7518405368858616,
+      "grad_norm": 2.308864116668701,
+      "learning_rate": 3.641713343130465e-05,
+      "loss": 0.6066,
+      "step": 17490
+    },
+    {
+      "epoch": 1.7528421896128612,
+      "grad_norm": 2.465198516845703,
+      "learning_rate": 3.640310642463723e-05,
+      "loss": 0.5812,
+      "step": 17500
+    },
+    {
+      "epoch": 1.7538438423398608,
+      "grad_norm": 2.269287347793579,
+      "learning_rate": 3.6389074883615395e-05,
+      "loss": 0.6508,
+      "step": 17510
+    },
+    {
+      "epoch": 1.7548454950668604,
+      "grad_norm": 2.149580717086792,
+      "learning_rate": 3.637503881381869e-05,
+      "loss": 0.5685,
+      "step": 17520
+    },
+    {
+      "epoch": 1.75584714779386,
+      "grad_norm": 2.5178351402282715,
+      "learning_rate": 3.6360998220828436e-05,
+      "loss": 0.5466,
+      "step": 17530
+    },
+    {
+      "epoch": 1.7568488005208596,
+      "grad_norm": 2.1612465381622314,
+      "learning_rate": 3.634695311022775e-05,
+      "loss": 0.4816,
+      "step": 17540
+    },
+    {
+      "epoch": 1.757850453247859,
+      "grad_norm": 2.286977529525757,
+      "learning_rate": 3.6332903487601584e-05,
+      "loss": 0.5219,
+      "step": 17550
+    },
+    {
+      "epoch": 1.7588521059748585,
+      "grad_norm": 1.9696747064590454,
+      "learning_rate": 3.6318849358536635e-05,
+      "loss": 0.4647,
+      "step": 17560
+    },
+    {
+      "epoch": 1.7598537587018581,
+      "grad_norm": 2.1928446292877197,
+      "learning_rate": 3.630479072862143e-05,
+      "loss": 0.5435,
+      "step": 17570
+    },
+    {
+      "epoch": 1.7608554114288575,
+      "grad_norm": 2.580597162246704,
+      "learning_rate": 3.629072760344627e-05,
+      "loss": 0.5458,
+      "step": 17580
+    },
+    {
+      "epoch": 1.761857064155857,
+      "grad_norm": 1.8262009620666504,
+      "learning_rate": 3.6276659988603234e-05,
+      "loss": 0.5767,
+      "step": 17590
+    },
+    {
+      "epoch": 1.7628587168828567,
+      "grad_norm": 2.46189284324646,
+      "learning_rate": 3.6262587889686205e-05,
+      "loss": 0.5401,
+      "step": 17600
+    },
+    {
+      "epoch": 1.7638603696098563,
+      "grad_norm": 3.054136037826538,
+      "learning_rate": 3.624851131229084e-05,
+      "loss": 0.4778,
+      "step": 17610
+    },
+    {
+      "epoch": 1.7648620223368559,
+      "grad_norm": 2.439973831176758,
+      "learning_rate": 3.6234430262014594e-05,
+      "loss": 0.5884,
+      "step": 17620
+    },
+    {
+      "epoch": 1.7658636750638554,
+      "grad_norm": 1.986525058746338,
+      "learning_rate": 3.622034474445665e-05,
+      "loss": 0.5668,
+      "step": 17630
+    },
+    {
+      "epoch": 1.766865327790855,
+      "grad_norm": 1.8597817420959473,
+      "learning_rate": 3.620625476521803e-05,
+      "loss": 0.5563,
+      "step": 17640
+    },
+    {
+      "epoch": 1.7678669805178546,
+      "grad_norm": 1.821752905845642,
+      "learning_rate": 3.6192160329901484e-05,
+      "loss": 0.5131,
+      "step": 17650
+    },
+    {
+      "epoch": 1.768868633244854,
+      "grad_norm": 2.221463918685913,
+      "learning_rate": 3.617806144411156e-05,
+      "loss": 0.545,
+      "step": 17660
+    },
+    {
+      "epoch": 1.7698702859718536,
+      "grad_norm": 2.033578395843506,
+      "learning_rate": 3.6163958113454574e-05,
+      "loss": 0.5321,
+      "step": 17670
+    },
+    {
+      "epoch": 1.770871938698853,
+      "grad_norm": 1.9562311172485352,
+      "learning_rate": 3.61498503435386e-05,
+      "loss": 0.5289,
+      "step": 17680
+    },
+    {
+      "epoch": 1.7718735914258525,
+      "grad_norm": 2.811387062072754,
+      "learning_rate": 3.6135738139973466e-05,
+      "loss": 0.5456,
+      "step": 17690
+    },
+    {
+      "epoch": 1.7728752441528521,
+      "grad_norm": 1.598136067390442,
+      "learning_rate": 3.6121621508370805e-05,
+      "loss": 0.5147,
+      "step": 17700
+    },
+    {
+      "epoch": 1.7738768968798517,
+      "grad_norm": 1.8884917497634888,
+      "learning_rate": 3.610750045434396e-05,
+      "loss": 0.5152,
+      "step": 17710
+    },
+    {
+      "epoch": 1.7748785496068513,
+      "grad_norm": 1.971439003944397,
+      "learning_rate": 3.609337498350805e-05,
+      "loss": 0.5127,
+      "step": 17720
+    },
+    {
+      "epoch": 1.775880202333851,
+      "grad_norm": 2.07023286819458,
+      "learning_rate": 3.607924510147998e-05,
+      "loss": 0.5321,
+      "step": 17730
+    },
+    {
+      "epoch": 1.7768818550608505,
+      "grad_norm": 2.4199936389923096,
+      "learning_rate": 3.6065110813878365e-05,
+      "loss": 0.5243,
+      "step": 17740
+    },
+    {
+      "epoch": 1.77788350778785,
+      "grad_norm": 2.1403119564056396,
+      "learning_rate": 3.6050972126323615e-05,
+      "loss": 0.553,
+      "step": 17750
+    },
+    {
+      "epoch": 1.7788851605148495,
+      "grad_norm": 3.023754596710205,
+      "learning_rate": 3.6036829044437835e-05,
+      "loss": 0.5912,
+      "step": 17760
+    },
+    {
+      "epoch": 1.779886813241849,
+      "grad_norm": 1.9545767307281494,
+      "learning_rate": 3.602268157384493e-05,
+      "loss": 0.6121,
+      "step": 17770
+    },
+    {
+      "epoch": 1.7808884659688486,
+      "grad_norm": 1.6822097301483154,
+      "learning_rate": 3.6008529720170524e-05,
+      "loss": 0.5864,
+      "step": 17780
+    },
+    {
+      "epoch": 1.781890118695848,
+      "grad_norm": 2.0851354598999023,
+      "learning_rate": 3.5994373489041995e-05,
+      "loss": 0.5074,
+      "step": 17790
+    },
+    {
+      "epoch": 1.7828917714228476,
+      "grad_norm": 1.7990679740905762,
+      "learning_rate": 3.598021288608845e-05,
+      "loss": 0.4877,
+      "step": 17800
+    },
+    {
+      "epoch": 1.7838934241498472,
+      "grad_norm": 2.6405606269836426,
+      "learning_rate": 3.5966047916940734e-05,
+      "loss": 0.5953,
+      "step": 17810
+    },
+    {
+      "epoch": 1.7848950768768468,
+      "grad_norm": 2.476152181625366,
+      "learning_rate": 3.595187858723144e-05,
+      "loss": 0.532,
+      "step": 17820
+    },
+    {
+      "epoch": 1.7858967296038464,
+      "grad_norm": 2.792933940887451,
+      "learning_rate": 3.593770490259489e-05,
+      "loss": 0.5781,
+      "step": 17830
+    },
+    {
+      "epoch": 1.786898382330846,
+      "grad_norm": 2.29030179977417,
+      "learning_rate": 3.592352686866713e-05,
+      "loss": 0.5617,
+      "step": 17840
+    },
+    {
+      "epoch": 1.7879000350578456,
+      "grad_norm": 2.300529956817627,
+      "learning_rate": 3.590934449108596e-05,
+      "loss": 0.5806,
+      "step": 17850
+    },
+    {
+      "epoch": 1.7889016877848452,
+      "grad_norm": 2.4189414978027344,
+      "learning_rate": 3.589515777549087e-05,
+      "loss": 0.5626,
+      "step": 17860
+    },
+    {
+      "epoch": 1.7899033405118445,
+      "grad_norm": 1.9713430404663086,
+      "learning_rate": 3.588096672752309e-05,
+      "loss": 0.5583,
+      "step": 17870
+    },
+    {
+      "epoch": 1.7909049932388441,
+      "grad_norm": 2.270739793777466,
+      "learning_rate": 3.5866771352825605e-05,
+      "loss": 0.5607,
+      "step": 17880
+    },
+    {
+      "epoch": 1.7919066459658435,
+      "grad_norm": 2.855412006378174,
+      "learning_rate": 3.5852571657043075e-05,
+      "loss": 0.5981,
+      "step": 17890
+    },
+    {
+      "epoch": 1.792908298692843,
+      "grad_norm": 2.324159622192383,
+      "learning_rate": 3.583836764582189e-05,
+      "loss": 0.4861,
+      "step": 17900
+    },
+    {
+      "epoch": 1.7939099514198427,
+      "grad_norm": 2.3632209300994873,
+      "learning_rate": 3.5824159324810175e-05,
+      "loss": 0.5487,
+      "step": 17910
+    },
+    {
+      "epoch": 1.7949116041468423,
+      "grad_norm": 2.1716160774230957,
+      "learning_rate": 3.580994669965774e-05,
+      "loss": 0.4606,
+      "step": 17920
+    },
+    {
+      "epoch": 1.7959132568738418,
+      "grad_norm": 2.514984607696533,
+      "learning_rate": 3.579572977601615e-05,
+      "loss": 0.4827,
+      "step": 17930
+    },
+    {
+      "epoch": 1.7969149096008414,
+      "grad_norm": 2.0306577682495117,
+      "learning_rate": 3.578150855953861e-05,
+      "loss": 0.6076,
+      "step": 17940
+    },
+    {
+      "epoch": 1.797916562327841,
+      "grad_norm": 2.062478542327881,
+      "learning_rate": 3.576728305588012e-05,
+      "loss": 0.558,
+      "step": 17950
+    },
+    {
+      "epoch": 1.7989182150548406,
+      "grad_norm": 2.077491283416748,
+      "learning_rate": 3.57530532706973e-05,
+      "loss": 0.5423,
+      "step": 17960
+    },
+    {
+      "epoch": 1.79991986778184,
+      "grad_norm": 1.7659790515899658,
+      "learning_rate": 3.573881920964853e-05,
+      "loss": 0.5647,
+      "step": 17970
+    },
+    {
+      "epoch": 1.8009215205088396,
+      "grad_norm": 2.1615078449249268,
+      "learning_rate": 3.5724580878393867e-05,
+      "loss": 0.5863,
+      "step": 17980
+    },
+    {
+      "epoch": 1.8019231732358392,
+      "grad_norm": 1.5051380395889282,
+      "learning_rate": 3.571033828259507e-05,
+      "loss": 0.5236,
+      "step": 17990
+    },
+    {
+      "epoch": 1.8029248259628385,
+      "grad_norm": 2.394617795944214,
+      "learning_rate": 3.569609142791559e-05,
+      "loss": 0.5616,
+      "step": 18000
+    },
+    {
+      "epoch": 1.8039264786898381,
+      "grad_norm": 2.699385643005371,
+      "learning_rate": 3.5681840320020585e-05,
+      "loss": 0.6394,
+      "step": 18010
+    },
+    {
+      "epoch": 1.8049281314168377,
+      "grad_norm": 2.4347870349884033,
+      "learning_rate": 3.566758496457688e-05,
+      "loss": 0.5286,
+      "step": 18020
+    },
+    {
+      "epoch": 1.8059297841438373,
+      "grad_norm": 1.4760379791259766,
+      "learning_rate": 3.5653325367253016e-05,
+      "loss": 0.5071,
+      "step": 18030
+    },
+    {
+      "epoch": 1.806931436870837,
+      "grad_norm": 2.0530753135681152,
+      "learning_rate": 3.56390615337192e-05,
+      "loss": 0.5589,
+      "step": 18040
+    },
+    {
+      "epoch": 1.8079330895978365,
+      "grad_norm": 2.031733751296997,
+      "learning_rate": 3.5624793469647344e-05,
+      "loss": 0.4645,
+      "step": 18050
+    },
+    {
+      "epoch": 1.808934742324836,
+      "grad_norm": 2.617835283279419,
+      "learning_rate": 3.5610521180711015e-05,
+      "loss": 0.6479,
+      "step": 18060
+    },
+    {
+      "epoch": 1.8099363950518357,
+      "grad_norm": 2.419201374053955,
+      "learning_rate": 3.559624467258548e-05,
+      "loss": 0.4964,
+      "step": 18070
+    },
+    {
+      "epoch": 1.810938047778835,
+      "grad_norm": 1.7701033353805542,
+      "learning_rate": 3.5581963950947686e-05,
+      "loss": 0.525,
+      "step": 18080
+    },
+    {
+      "epoch": 1.8119397005058346,
+      "grad_norm": 2.738837480545044,
+      "learning_rate": 3.556767902147623e-05,
+      "loss": 0.5093,
+      "step": 18090
+    },
+    {
+      "epoch": 1.8129413532328342,
+      "grad_norm": 1.944037914276123,
+      "learning_rate": 3.5553389889851426e-05,
+      "loss": 0.5076,
+      "step": 18100
+    },
+    {
+      "epoch": 1.8139430059598336,
+      "grad_norm": 2.223052978515625,
+      "learning_rate": 3.553909656175522e-05,
+      "loss": 0.5436,
+      "step": 18110
+    },
+    {
+      "epoch": 1.8149446586868332,
+      "grad_norm": 2.251912832260132,
+      "learning_rate": 3.552479904287123e-05,
+      "loss": 0.5097,
+      "step": 18120
+    },
+    {
+      "epoch": 1.8159463114138328,
+      "grad_norm": 2.422960042953491,
+      "learning_rate": 3.5510497338884774e-05,
+      "loss": 0.5101,
+      "step": 18130
+    },
+    {
+      "epoch": 1.8169479641408324,
+      "grad_norm": 2.5609405040740967,
+      "learning_rate": 3.549619145548279e-05,
+      "loss": 0.542,
+      "step": 18140
+    },
+    {
+      "epoch": 1.817949616867832,
+      "grad_norm": 2.37107515335083,
+      "learning_rate": 3.54818813983539e-05,
+      "loss": 0.5574,
+      "step": 18150
+    },
+    {
+      "epoch": 1.8189512695948316,
+      "grad_norm": 2.355593681335449,
+      "learning_rate": 3.54675671731884e-05,
+      "loss": 0.5213,
+      "step": 18160
+    },
+    {
+      "epoch": 1.8199529223218311,
+      "grad_norm": 2.0474841594696045,
+      "learning_rate": 3.545324878567821e-05,
+      "loss": 0.5093,
+      "step": 18170
+    },
+    {
+      "epoch": 1.8209545750488307,
+      "grad_norm": 2.3665642738342285,
+      "learning_rate": 3.543892624151693e-05,
+      "loss": 0.5607,
+      "step": 18180
+    },
+    {
+      "epoch": 1.82195622777583,
+      "grad_norm": 2.511904239654541,
+      "learning_rate": 3.542459954639981e-05,
+      "loss": 0.5602,
+      "step": 18190
+    },
+    {
+      "epoch": 1.8229578805028297,
+      "grad_norm": 2.2814455032348633,
+      "learning_rate": 3.541026870602375e-05,
+      "loss": 0.5188,
+      "step": 18200
+    },
+    {
+      "epoch": 1.823959533229829,
+      "grad_norm": 2.252317428588867,
+      "learning_rate": 3.539593372608727e-05,
+      "loss": 0.4779,
+      "step": 18210
+    },
+    {
+      "epoch": 1.8249611859568287,
+      "grad_norm": 1.9406018257141113,
+      "learning_rate": 3.538159461229059e-05,
+      "loss": 0.6143,
+      "step": 18220
+    },
+    {
+      "epoch": 1.8259628386838282,
+      "grad_norm": 2.396653175354004,
+      "learning_rate": 3.5367251370335526e-05,
+      "loss": 0.5132,
+      "step": 18230
+    },
+    {
+      "epoch": 1.8269644914108278,
+      "grad_norm": 2.2335658073425293,
+      "learning_rate": 3.535290400592556e-05,
+      "loss": 0.5352,
+      "step": 18240
+    },
+    {
+      "epoch": 1.8279661441378274,
+      "grad_norm": 2.568760395050049,
+      "learning_rate": 3.53385525247658e-05,
+      "loss": 0.5214,
+      "step": 18250
+    },
+    {
+      "epoch": 1.828967796864827,
+      "grad_norm": 1.8453973531723022,
+      "learning_rate": 3.532419693256301e-05,
+      "loss": 0.5491,
+      "step": 18260
+    },
+    {
+      "epoch": 1.8299694495918266,
+      "grad_norm": 2.2260847091674805,
+      "learning_rate": 3.5309837235025574e-05,
+      "loss": 0.5602,
+      "step": 18270
+    },
+    {
+      "epoch": 1.8309711023188262,
+      "grad_norm": 1.965449333190918,
+      "learning_rate": 3.5295473437863505e-05,
+      "loss": 0.554,
+      "step": 18280
+    },
+    {
+      "epoch": 1.8319727550458256,
+      "grad_norm": 2.2575039863586426,
+      "learning_rate": 3.528110554678846e-05,
+      "loss": 0.597,
+      "step": 18290
+    },
+    {
+      "epoch": 1.8329744077728252,
+      "grad_norm": 1.884130835533142,
+      "learning_rate": 3.526673356751371e-05,
+      "loss": 0.5049,
+      "step": 18300
+    },
+    {
+      "epoch": 1.8339760604998248,
+      "grad_norm": 2.2805416584014893,
+      "learning_rate": 3.525235750575416e-05,
+      "loss": 0.5269,
+      "step": 18310
+    },
+    {
+      "epoch": 1.8349777132268241,
+      "grad_norm": 2.192514419555664,
+      "learning_rate": 3.523797736722634e-05,
+      "loss": 0.5541,
+      "step": 18320
+    },
+    {
+      "epoch": 1.8359793659538237,
+      "grad_norm": 2.4717044830322266,
+      "learning_rate": 3.52235931576484e-05,
+      "loss": 0.5149,
+      "step": 18330
+    },
+    {
+      "epoch": 1.8369810186808233,
+      "grad_norm": 2.0693359375,
+      "learning_rate": 3.520920488274009e-05,
+      "loss": 0.5642,
+      "step": 18340
+    },
+    {
+      "epoch": 1.837982671407823,
+      "grad_norm": 2.423137903213501,
+      "learning_rate": 3.51948125482228e-05,
+      "loss": 0.5392,
+      "step": 18350
+    },
+    {
+      "epoch": 1.8389843241348225,
+      "grad_norm": 2.4728684425354004,
+      "learning_rate": 3.518041615981954e-05,
+      "loss": 0.5382,
+      "step": 18360
+    },
+    {
+      "epoch": 1.839985976861822,
+      "grad_norm": 2.3842051029205322,
+      "learning_rate": 3.516601572325491e-05,
+      "loss": 0.5356,
+      "step": 18370
+    },
+    {
+      "epoch": 1.8409876295888217,
+      "grad_norm": 1.8995606899261475,
+      "learning_rate": 3.515161124425513e-05,
+      "loss": 0.505,
+      "step": 18380
+    },
+    {
+      "epoch": 1.8419892823158213,
+      "grad_norm": 2.8482911586761475,
+      "learning_rate": 3.513720272854802e-05,
+      "loss": 0.5478,
+      "step": 18390
+    },
+    {
+      "epoch": 1.8429909350428206,
+      "grad_norm": 1.9546759128570557,
+      "learning_rate": 3.5122790181863017e-05,
+      "loss": 0.4782,
+      "step": 18400
+    },
+    {
+      "epoch": 1.8439925877698202,
+      "grad_norm": 2.6709511280059814,
+      "learning_rate": 3.510837360993116e-05,
+      "loss": 0.5357,
+      "step": 18410
+    },
+    {
+      "epoch": 1.8449942404968196,
+      "grad_norm": 2.3356592655181885,
+      "learning_rate": 3.5093953018485076e-05,
+      "loss": 0.5344,
+      "step": 18420
+    },
+    {
+      "epoch": 1.8459958932238192,
+      "grad_norm": 2.1896042823791504,
+      "learning_rate": 3.507952841325899e-05,
+      "loss": 0.6113,
+      "step": 18430
+    },
+    {
+      "epoch": 1.8469975459508188,
+      "grad_norm": 1.9698134660720825,
+      "learning_rate": 3.5065099799988766e-05,
+      "loss": 0.4655,
+      "step": 18440
+    },
+    {
+      "epoch": 1.8479991986778184,
+      "grad_norm": 2.549419403076172,
+      "learning_rate": 3.505066718441179e-05,
+      "loss": 0.4673,
+      "step": 18450
+    },
+    {
+      "epoch": 1.849000851404818,
+      "grad_norm": 2.0355772972106934,
+      "learning_rate": 3.503623057226709e-05,
+      "loss": 0.5131,
+      "step": 18460
+    },
+    {
+      "epoch": 1.8500025041318175,
+      "grad_norm": 2.1255834102630615,
+      "learning_rate": 3.502178996929527e-05,
+      "loss": 0.5656,
+      "step": 18470
+    },
+    {
+      "epoch": 1.8510041568588171,
+      "grad_norm": 2.1015443801879883,
+      "learning_rate": 3.500734538123852e-05,
+      "loss": 0.5569,
+      "step": 18480
+    },
+    {
+      "epoch": 1.8520058095858167,
+      "grad_norm": 2.9126319885253906,
+      "learning_rate": 3.4992896813840624e-05,
+      "loss": 0.5279,
+      "step": 18490
+    },
+    {
+      "epoch": 1.853007462312816,
+      "grad_norm": 2.6113879680633545,
+      "learning_rate": 3.497844427284693e-05,
+      "loss": 0.6152,
+      "step": 18500
+    },
+    {
+      "epoch": 1.8540091150398157,
+      "grad_norm": 2.1508989334106445,
+      "learning_rate": 3.496398776400437e-05,
+      "loss": 0.5262,
+      "step": 18510
+    },
+    {
+      "epoch": 1.8550107677668153,
+      "grad_norm": 2.867258071899414,
+      "learning_rate": 3.4949527293061475e-05,
+      "loss": 0.6089,
+      "step": 18520
+    },
+    {
+      "epoch": 1.8560124204938147,
+      "grad_norm": 2.2705745697021484,
+      "learning_rate": 3.493506286576832e-05,
+      "loss": 0.5411,
+      "step": 18530
+    },
+    {
+      "epoch": 1.8570140732208142,
+      "grad_norm": 2.0853490829467773,
+      "learning_rate": 3.492059448787659e-05,
+      "loss": 0.5184,
+      "step": 18540
+    },
+    {
+      "epoch": 1.8580157259478138,
+      "grad_norm": 2.6328513622283936,
+      "learning_rate": 3.4906122165139496e-05,
+      "loss": 0.5298,
+      "step": 18550
+    },
+    {
+      "epoch": 1.8590173786748134,
+      "grad_norm": 2.7920444011688232,
+      "learning_rate": 3.489164590331186e-05,
+      "loss": 0.5228,
+      "step": 18560
+    },
+    {
+      "epoch": 1.860019031401813,
+      "grad_norm": 1.7941895723342896,
+      "learning_rate": 3.487716570815004e-05,
+      "loss": 0.5057,
+      "step": 18570
+    },
+    {
+      "epoch": 1.8610206841288126,
+      "grad_norm": 1.9120043516159058,
+      "learning_rate": 3.4862681585411984e-05,
+      "loss": 0.5185,
+      "step": 18580
+    },
+    {
+      "epoch": 1.8620223368558122,
+      "grad_norm": 1.4897717237472534,
+      "learning_rate": 3.484819354085717e-05,
+      "loss": 0.4982,
+      "step": 18590
+    },
+    {
+      "epoch": 1.8630239895828118,
+      "grad_norm": 1.8527815341949463,
+      "learning_rate": 3.483370158024667e-05,
+      "loss": 0.5612,
+      "step": 18600
+    },
+    {
+      "epoch": 1.8640256423098112,
+      "grad_norm": 2.3453290462493896,
+      "learning_rate": 3.481920570934308e-05,
+      "loss": 0.4998,
+      "step": 18610
+    },
+    {
+      "epoch": 1.8650272950368108,
+      "grad_norm": 2.2166287899017334,
+      "learning_rate": 3.48047059339106e-05,
+      "loss": 0.5478,
+      "step": 18620
+    },
+    {
+      "epoch": 1.8660289477638103,
+      "grad_norm": 1.6502530574798584,
+      "learning_rate": 3.479020225971491e-05,
+      "loss": 0.535,
+      "step": 18630
+    },
+    {
+      "epoch": 1.8670306004908097,
+      "grad_norm": 1.860781192779541,
+      "learning_rate": 3.4775694692523306e-05,
+      "loss": 0.4861,
+      "step": 18640
+    },
+    {
+      "epoch": 1.8680322532178093,
+      "grad_norm": 2.397423505783081,
+      "learning_rate": 3.476118323810459e-05,
+      "loss": 0.5373,
+      "step": 18650
+    },
+    {
+      "epoch": 1.869033905944809,
+      "grad_norm": 2.4832894802093506,
+      "learning_rate": 3.474666790222914e-05,
+      "loss": 0.5306,
+      "step": 18660
+    },
+    {
+      "epoch": 1.8700355586718085,
+      "grad_norm": 2.5092248916625977,
+      "learning_rate": 3.4732148690668866e-05,
+      "loss": 0.5647,
+      "step": 18670
+    },
+    {
+      "epoch": 1.871037211398808,
+      "grad_norm": 2.138489007949829,
+      "learning_rate": 3.471762560919719e-05,
+      "loss": 0.5775,
+      "step": 18680
+    },
+    {
+      "epoch": 1.8720388641258077,
+      "grad_norm": 2.200510263442993,
+      "learning_rate": 3.470309866358914e-05,
+      "loss": 0.5667,
+      "step": 18690
+    },
+    {
+      "epoch": 1.8730405168528073,
+      "grad_norm": 2.08661150932312,
+      "learning_rate": 3.46885678596212e-05,
+      "loss": 0.5655,
+      "step": 18700
+    },
+    {
+      "epoch": 1.8740421695798066,
+      "grad_norm": 2.4580740928649902,
+      "learning_rate": 3.4674033203071464e-05,
+      "loss": 0.522,
+      "step": 18710
+    },
+    {
+      "epoch": 1.8750438223068062,
+      "grad_norm": 2.8262064456939697,
+      "learning_rate": 3.46594946997195e-05,
+      "loss": 0.5338,
+      "step": 18720
+    },
+    {
+      "epoch": 1.8760454750338058,
+      "grad_norm": 2.0510544776916504,
+      "learning_rate": 3.4644952355346435e-05,
+      "loss": 0.588,
+      "step": 18730
+    },
+    {
+      "epoch": 1.8770471277608052,
+      "grad_norm": 2.456479549407959,
+      "learning_rate": 3.463040617573491e-05,
+      "loss": 0.6353,
+      "step": 18740
+    },
+    {
+      "epoch": 1.8780487804878048,
+      "grad_norm": 2.268458127975464,
+      "learning_rate": 3.461585616666911e-05,
+      "loss": 0.5018,
+      "step": 18750
+    },
+    {
+      "epoch": 1.8790504332148044,
+      "grad_norm": 2.170693874359131,
+      "learning_rate": 3.460130233393472e-05,
+      "loss": 0.5805,
+      "step": 18760
+    },
+    {
+      "epoch": 1.880052085941804,
+      "grad_norm": 2.764228582382202,
+      "learning_rate": 3.458674468331896e-05,
+      "loss": 0.5206,
+      "step": 18770
+    },
+    {
+      "epoch": 1.8810537386688035,
+      "grad_norm": 1.9646960496902466,
+      "learning_rate": 3.457218322061056e-05,
+      "loss": 0.4927,
+      "step": 18780
+    },
+    {
+      "epoch": 1.8820553913958031,
+      "grad_norm": 2.303696870803833,
+      "learning_rate": 3.455761795159978e-05,
+      "loss": 0.5437,
+      "step": 18790
+    },
+    {
+      "epoch": 1.8830570441228027,
+      "grad_norm": 2.2272274494171143,
+      "learning_rate": 3.454304888207837e-05,
+      "loss": 0.6098,
+      "step": 18800
+    },
+    {
+      "epoch": 1.8840586968498023,
+      "grad_norm": 2.036489248275757,
+      "learning_rate": 3.452847601783959e-05,
+      "loss": 0.5595,
+      "step": 18810
+    },
+    {
+      "epoch": 1.8850603495768017,
+      "grad_norm": 2.238173007965088,
+      "learning_rate": 3.451389936467827e-05,
+      "loss": 0.5055,
+      "step": 18820
+    },
+    {
+      "epoch": 1.8860620023038013,
+      "grad_norm": 2.1000025272369385,
+      "learning_rate": 3.4499318928390665e-05,
+      "loss": 0.5292,
+      "step": 18830
+    },
+    {
+      "epoch": 1.8870636550308009,
+      "grad_norm": 2.115056037902832,
+      "learning_rate": 3.448473471477457e-05,
+      "loss": 0.5193,
+      "step": 18840
+    },
+    {
+      "epoch": 1.8880653077578002,
+      "grad_norm": 2.3937647342681885,
+      "learning_rate": 3.44701467296293e-05,
+      "loss": 0.5607,
+      "step": 18850
+    },
+    {
+      "epoch": 1.8890669604847998,
+      "grad_norm": 2.403609037399292,
+      "learning_rate": 3.4455554978755634e-05,
+      "loss": 0.4612,
+      "step": 18860
+    },
+    {
+      "epoch": 1.8900686132117994,
+      "grad_norm": 2.042008399963379,
+      "learning_rate": 3.444095946795587e-05,
+      "loss": 0.5366,
+      "step": 18870
+    },
+    {
+      "epoch": 1.891070265938799,
+      "grad_norm": 2.9627082347869873,
+      "learning_rate": 3.44263602030338e-05,
+      "loss": 0.5807,
+      "step": 18880
+    },
+    {
+      "epoch": 1.8920719186657986,
+      "grad_norm": 1.7297526597976685,
+      "learning_rate": 3.4411757189794703e-05,
+      "loss": 0.4538,
+      "step": 18890
+    },
+    {
+      "epoch": 1.8930735713927982,
+      "grad_norm": 2.134171724319458,
+      "learning_rate": 3.439715043404535e-05,
+      "loss": 0.5695,
+      "step": 18900
+    },
+    {
+      "epoch": 1.8940752241197978,
+      "grad_norm": 3.0094010829925537,
+      "learning_rate": 3.4382539941594e-05,
+      "loss": 0.4792,
+      "step": 18910
+    },
+    {
+      "epoch": 1.8950768768467974,
+      "grad_norm": 1.9440160989761353,
+      "learning_rate": 3.4367925718250405e-05,
+      "loss": 0.5362,
+      "step": 18920
+    },
+    {
+      "epoch": 1.8960785295737967,
+      "grad_norm": 2.718689203262329,
+      "learning_rate": 3.4353307769825794e-05,
+      "loss": 0.4992,
+      "step": 18930
+    },
+    {
+      "epoch": 1.8970801823007963,
+      "grad_norm": 2.141523599624634,
+      "learning_rate": 3.433868610213286e-05,
+      "loss": 0.5421,
+      "step": 18940
+    },
+    {
+      "epoch": 1.8980818350277957,
+      "grad_norm": 2.2745018005371094,
+      "learning_rate": 3.4324060720985815e-05,
+      "loss": 0.5296,
+      "step": 18950
+    },
+    {
+      "epoch": 1.8990834877547953,
+      "grad_norm": 2.313880443572998,
+      "learning_rate": 3.4309431632200325e-05,
+      "loss": 0.5812,
+      "step": 18960
+    },
+    {
+      "epoch": 1.9000851404817949,
+      "grad_norm": 2.571535110473633,
+      "learning_rate": 3.429479884159351e-05,
+      "loss": 0.5244,
+      "step": 18970
+    },
+    {
+      "epoch": 1.9010867932087945,
+      "grad_norm": 2.148982286453247,
+      "learning_rate": 3.4280162354984e-05,
+      "loss": 0.4673,
+      "step": 18980
+    },
+    {
+      "epoch": 1.902088445935794,
+      "grad_norm": 3.1783385276794434,
+      "learning_rate": 3.426552217819187e-05,
+      "loss": 0.4956,
+      "step": 18990
+    },
+    {
+      "epoch": 1.9030900986627937,
+      "grad_norm": 2.790367841720581,
+      "learning_rate": 3.425087831703868e-05,
+      "loss": 0.5036,
+      "step": 19000
+    },
+    {
+      "epoch": 1.9040917513897933,
+      "grad_norm": 2.6144115924835205,
+      "learning_rate": 3.423623077734743e-05,
+      "loss": 0.5183,
+      "step": 19010
+    },
+    {
+      "epoch": 1.9050934041167928,
+      "grad_norm": 1.9311896562576294,
+      "learning_rate": 3.4221579564942604e-05,
+      "loss": 0.5641,
+      "step": 19020
+    },
+    {
+      "epoch": 1.9060950568437922,
+      "grad_norm": 2.1071672439575195,
+      "learning_rate": 3.4206924685650143e-05,
+      "loss": 0.5145,
+      "step": 19030
+    },
+    {
+      "epoch": 1.9070967095707918,
+      "grad_norm": 2.4376401901245117,
+      "learning_rate": 3.419226614529744e-05,
+      "loss": 0.4721,
+      "step": 19040
+    },
+    {
+      "epoch": 1.9080983622977914,
+      "grad_norm": 2.126660108566284,
+      "learning_rate": 3.417760394971335e-05,
+      "loss": 0.4731,
+      "step": 19050
+    },
+    {
+      "epoch": 1.9091000150247908,
+      "grad_norm": 2.3037705421447754,
+      "learning_rate": 3.4162938104728165e-05,
+      "loss": 0.5206,
+      "step": 19060
+    },
+    {
+      "epoch": 1.9101016677517904,
+      "grad_norm": 2.035919427871704,
+      "learning_rate": 3.4148268616173655e-05,
+      "loss": 0.564,
+      "step": 19070
+    },
+    {
+      "epoch": 1.91110332047879,
+      "grad_norm": 2.29209566116333,
+      "learning_rate": 3.413359548988303e-05,
+      "loss": 0.4901,
+      "step": 19080
+    },
+    {
+      "epoch": 1.9121049732057895,
+      "grad_norm": 3.5236032009124756,
+      "learning_rate": 3.4118918731690925e-05,
+      "loss": 0.564,
+      "step": 19090
+    },
+    {
+      "epoch": 1.9131066259327891,
+      "grad_norm": 3.52905011177063,
+      "learning_rate": 3.410423834743345e-05,
+      "loss": 0.6002,
+      "step": 19100
+    },
+    {
+      "epoch": 1.9141082786597887,
+      "grad_norm": 2.6410086154937744,
+      "learning_rate": 3.408955434294813e-05,
+      "loss": 0.5538,
+      "step": 19110
+    },
+    {
+      "epoch": 1.9151099313867883,
+      "grad_norm": 2.0407655239105225,
+      "learning_rate": 3.407486672407395e-05,
+      "loss": 0.5388,
+      "step": 19120
+    },
+    {
+      "epoch": 1.916111584113788,
+      "grad_norm": 2.2420051097869873,
+      "learning_rate": 3.406017549665134e-05,
+      "loss": 0.567,
+      "step": 19130
+    },
+    {
+      "epoch": 1.9171132368407873,
+      "grad_norm": 2.550873041152954,
+      "learning_rate": 3.404548066652211e-05,
+      "loss": 0.5538,
+      "step": 19140
+    },
+    {
+      "epoch": 1.9181148895677869,
+      "grad_norm": 2.334690570831299,
+      "learning_rate": 3.403078223952959e-05,
+      "loss": 0.5259,
+      "step": 19150
+    },
+    {
+      "epoch": 1.9191165422947862,
+      "grad_norm": 1.7644654512405396,
+      "learning_rate": 3.4016080221518455e-05,
+      "loss": 0.5255,
+      "step": 19160
+    },
+    {
+      "epoch": 1.9201181950217858,
+      "grad_norm": 2.645530939102173,
+      "learning_rate": 3.4001374618334856e-05,
+      "loss": 0.5159,
+      "step": 19170
+    },
+    {
+      "epoch": 1.9211198477487854,
+      "grad_norm": 2.168233871459961,
+      "learning_rate": 3.398666543582637e-05,
+      "loss": 0.5738,
+      "step": 19180
+    },
+    {
+      "epoch": 1.922121500475785,
+      "grad_norm": 2.2151002883911133,
+      "learning_rate": 3.397195267984197e-05,
+      "loss": 0.558,
+      "step": 19190
+    },
+    {
+      "epoch": 1.9231231532027846,
+      "grad_norm": 1.8769047260284424,
+      "learning_rate": 3.395723635623208e-05,
+      "loss": 0.5495,
+      "step": 19200
+    },
+    {
+      "epoch": 1.9241248059297842,
+      "grad_norm": 2.8317298889160156,
+      "learning_rate": 3.394251647084852e-05,
+      "loss": 0.5709,
+      "step": 19210
+    },
+    {
+      "epoch": 1.9251264586567838,
+      "grad_norm": 2.330803871154785,
+      "learning_rate": 3.392779302954454e-05,
+      "loss": 0.5585,
+      "step": 19220
+    },
+    {
+      "epoch": 1.9261281113837834,
+      "grad_norm": 2.302245616912842,
+      "learning_rate": 3.39130660381748e-05,
+      "loss": 0.5349,
+      "step": 19230
+    },
+    {
+      "epoch": 1.9271297641107827,
+      "grad_norm": 2.2176613807678223,
+      "learning_rate": 3.389833550259536e-05,
+      "loss": 0.47,
+      "step": 19240
+    },
+    {
+      "epoch": 1.9281314168377823,
+      "grad_norm": 1.9405122995376587,
+      "learning_rate": 3.388360142866371e-05,
+      "loss": 0.5174,
+      "step": 19250
+    },
+    {
+      "epoch": 1.929133069564782,
+      "grad_norm": 2.862607479095459,
+      "learning_rate": 3.386886382223874e-05,
+      "loss": 0.5595,
+      "step": 19260
+    },
+    {
+      "epoch": 1.9301347222917813,
+      "grad_norm": 1.8453691005706787,
+      "learning_rate": 3.385412268918073e-05,
+      "loss": 0.5051,
+      "step": 19270
+    },
+    {
+      "epoch": 1.9311363750187809,
+      "grad_norm": 2.0955848693847656,
+      "learning_rate": 3.383937803535139e-05,
+      "loss": 0.5681,
+      "step": 19280
+    },
+    {
+      "epoch": 1.9321380277457805,
+      "grad_norm": 2.8637776374816895,
+      "learning_rate": 3.3824629866613795e-05,
+      "loss": 0.5354,
+      "step": 19290
+    },
+    {
+      "epoch": 1.93313968047278,
+      "grad_norm": 1.8948688507080078,
+      "learning_rate": 3.380987818883245e-05,
+      "loss": 0.5965,
+      "step": 19300
+    },
+    {
+      "epoch": 1.9341413331997797,
+      "grad_norm": 2.108222246170044,
+      "learning_rate": 3.379512300787324e-05,
+      "loss": 0.5391,
+      "step": 19310
+    },
+    {
+      "epoch": 1.9351429859267792,
+      "grad_norm": 2.2994465827941895,
+      "learning_rate": 3.3780364329603445e-05,
+      "loss": 0.532,
+      "step": 19320
+    },
+    {
+      "epoch": 1.9361446386537788,
+      "grad_norm": 2.433121681213379,
+      "learning_rate": 3.376560215989174e-05,
+      "loss": 0.5057,
+      "step": 19330
+    },
+    {
+      "epoch": 1.9371462913807784,
+      "grad_norm": 2.2182536125183105,
+      "learning_rate": 3.3750836504608176e-05,
+      "loss": 0.5601,
+      "step": 19340
+    },
+    {
+      "epoch": 1.9381479441077778,
+      "grad_norm": 3.7288570404052734,
+      "learning_rate": 3.373606736962419e-05,
+      "loss": 0.587,
+      "step": 19350
+    },
+    {
+      "epoch": 1.9391495968347774,
+      "grad_norm": 2.0662009716033936,
+      "learning_rate": 3.372129476081264e-05,
+      "loss": 0.5102,
+      "step": 19360
+    },
+    {
+      "epoch": 1.940151249561777,
+      "grad_norm": 2.209134817123413,
+      "learning_rate": 3.37065186840477e-05,
+      "loss": 0.5374,
+      "step": 19370
+    },
+    {
+      "epoch": 1.9411529022887763,
+      "grad_norm": 1.951745867729187,
+      "learning_rate": 3.369173914520499e-05,
+      "loss": 0.5269,
+      "step": 19380
+    },
+    {
+      "epoch": 1.942154555015776,
+      "grad_norm": 2.849954605102539,
+      "learning_rate": 3.367695615016146e-05,
+      "loss": 0.5075,
+      "step": 19390
+    },
+    {
+      "epoch": 1.9431562077427755,
+      "grad_norm": 2.189919948577881,
+      "learning_rate": 3.3662169704795454e-05,
+      "loss": 0.5444,
+      "step": 19400
+    },
+    {
+      "epoch": 1.9441578604697751,
+      "grad_norm": 2.3285562992095947,
+      "learning_rate": 3.364737981498668e-05,
+      "loss": 0.527,
+      "step": 19410
+    },
+    {
+      "epoch": 1.9451595131967747,
+      "grad_norm": 2.5581185817718506,
+      "learning_rate": 3.363258648661623e-05,
+      "loss": 0.5201,
+      "step": 19420
+    },
+    {
+      "epoch": 1.9461611659237743,
+      "grad_norm": 2.3942644596099854,
+      "learning_rate": 3.361778972556655e-05,
+      "loss": 0.5137,
+      "step": 19430
+    },
+    {
+      "epoch": 1.947162818650774,
+      "grad_norm": 2.7398712635040283,
+      "learning_rate": 3.360298953772144e-05,
+      "loss": 0.597,
+      "step": 19440
+    },
+    {
+      "epoch": 1.9481644713777735,
+      "grad_norm": 2.240510940551758,
+      "learning_rate": 3.358818592896609e-05,
+      "loss": 0.5167,
+      "step": 19450
+    },
+    {
+      "epoch": 1.9491661241047729,
+      "grad_norm": 2.2413768768310547,
+      "learning_rate": 3.357337890518704e-05,
+      "loss": 0.4973,
+      "step": 19460
+    },
+    {
+      "epoch": 1.9501677768317724,
+      "grad_norm": 2.2250759601593018,
+      "learning_rate": 3.355856847227217e-05,
+      "loss": 0.5633,
+      "step": 19470
+    },
+    {
+      "epoch": 1.9511694295587718,
+      "grad_norm": 2.3619704246520996,
+      "learning_rate": 3.3543754636110755e-05,
+      "loss": 0.5079,
+      "step": 19480
+    },
+    {
+      "epoch": 1.9521710822857714,
+      "grad_norm": 1.8453508615493774,
+      "learning_rate": 3.3528937402593375e-05,
+      "loss": 0.535,
+      "step": 19490
+    },
+    {
+      "epoch": 1.953172735012771,
+      "grad_norm": 1.8270654678344727,
+      "learning_rate": 3.351411677761199e-05,
+      "loss": 0.4869,
+      "step": 19500
+    },
+    {
+      "epoch": 1.9541743877397706,
+      "grad_norm": 1.9440797567367554,
+      "learning_rate": 3.349929276705992e-05,
+      "loss": 0.5365,
+      "step": 19510
+    },
+    {
+      "epoch": 1.9551760404667702,
+      "grad_norm": 2.2624425888061523,
+      "learning_rate": 3.3484465376831784e-05,
+      "loss": 0.5449,
+      "step": 19520
+    },
+    {
+      "epoch": 1.9561776931937698,
+      "grad_norm": 2.4637839794158936,
+      "learning_rate": 3.3469634612823616e-05,
+      "loss": 0.5565,
+      "step": 19530
+    },
+    {
+      "epoch": 1.9571793459207694,
+      "grad_norm": 2.3287465572357178,
+      "learning_rate": 3.345480048093272e-05,
+      "loss": 0.5509,
+      "step": 19540
+    },
+    {
+      "epoch": 1.958180998647769,
+      "grad_norm": 1.6043740510940552,
+      "learning_rate": 3.3439962987057774e-05,
+      "loss": 0.5079,
+      "step": 19550
+    },
+    {
+      "epoch": 1.9591826513747683,
+      "grad_norm": 2.930209159851074,
+      "learning_rate": 3.3425122137098794e-05,
+      "loss": 0.5571,
+      "step": 19560
+    },
+    {
+      "epoch": 1.960184304101768,
+      "grad_norm": 2.760525941848755,
+      "learning_rate": 3.341027793695713e-05,
+      "loss": 0.5042,
+      "step": 19570
+    },
+    {
+      "epoch": 1.9611859568287675,
+      "grad_norm": 2.3335533142089844,
+      "learning_rate": 3.3395430392535455e-05,
+      "loss": 0.549,
+      "step": 19580
+    },
+    {
+      "epoch": 1.9621876095557669,
+      "grad_norm": 2.1368236541748047,
+      "learning_rate": 3.338057950973778e-05,
+      "loss": 0.4718,
+      "step": 19590
+    },
+    {
+      "epoch": 1.9631892622827665,
+      "grad_norm": 1.9596712589263916,
+      "learning_rate": 3.336572529446944e-05,
+      "loss": 0.5775,
+      "step": 19600
+    },
+    {
+      "epoch": 1.964190915009766,
+      "grad_norm": 2.1493420600891113,
+      "learning_rate": 3.335086775263709e-05,
+      "loss": 0.5564,
+      "step": 19610
+    },
+    {
+      "epoch": 1.9651925677367656,
+      "grad_norm": 2.2132809162139893,
+      "learning_rate": 3.333600689014872e-05,
+      "loss": 0.5046,
+      "step": 19620
+    },
+    {
+      "epoch": 1.9661942204637652,
+      "grad_norm": 2.2998745441436768,
+      "learning_rate": 3.3321142712913625e-05,
+      "loss": 0.4809,
+      "step": 19630
+    },
+    {
+      "epoch": 1.9671958731907648,
+      "grad_norm": 1.8581064939498901,
+      "learning_rate": 3.330627522684244e-05,
+      "loss": 0.5627,
+      "step": 19640
+    },
+    {
+      "epoch": 1.9681975259177644,
+      "grad_norm": 2.277087926864624,
+      "learning_rate": 3.329140443784709e-05,
+      "loss": 0.5622,
+      "step": 19650
+    },
+    {
+      "epoch": 1.969199178644764,
+      "grad_norm": 2.815183162689209,
+      "learning_rate": 3.3276530351840825e-05,
+      "loss": 0.4914,
+      "step": 19660
+    },
+    {
+      "epoch": 1.9702008313717634,
+      "grad_norm": 3.2514102458953857,
+      "learning_rate": 3.326165297473821e-05,
+      "loss": 0.5576,
+      "step": 19670
+    },
+    {
+      "epoch": 1.971202484098763,
+      "grad_norm": 1.6029939651489258,
+      "learning_rate": 3.324677231245512e-05,
+      "loss": 0.5235,
+      "step": 19680
+    },
+    {
+      "epoch": 1.9722041368257623,
+      "grad_norm": 1.8514877557754517,
+      "learning_rate": 3.323188837090874e-05,
+      "loss": 0.5727,
+      "step": 19690
+    },
+    {
+      "epoch": 1.973205789552762,
+      "grad_norm": 2.3635377883911133,
+      "learning_rate": 3.3217001156017526e-05,
+      "loss": 0.4909,
+      "step": 19700
+    },
+    {
+      "epoch": 1.9742074422797615,
+      "grad_norm": 2.3811750411987305,
+      "learning_rate": 3.320211067370128e-05,
+      "loss": 0.5762,
+      "step": 19710
+    },
+    {
+      "epoch": 1.9752090950067611,
+      "grad_norm": 2.026413679122925,
+      "learning_rate": 3.318721692988108e-05,
+      "loss": 0.4796,
+      "step": 19720
+    },
+    {
+      "epoch": 1.9762107477337607,
+      "grad_norm": 2.4324848651885986,
+      "learning_rate": 3.317231993047929e-05,
+      "loss": 0.4812,
+      "step": 19730
+    },
+    {
+      "epoch": 1.9772124004607603,
+      "grad_norm": 3.1195809841156006,
+      "learning_rate": 3.3157419681419616e-05,
+      "loss": 0.5272,
+      "step": 19740
+    },
+    {
+      "epoch": 1.97821405318776,
+      "grad_norm": 1.83283269405365,
+      "learning_rate": 3.314251618862699e-05,
+      "loss": 0.5282,
+      "step": 19750
+    },
+    {
+      "epoch": 1.9792157059147595,
+      "grad_norm": 1.9941926002502441,
+      "learning_rate": 3.3127609458027675e-05,
+      "loss": 0.4591,
+      "step": 19760
+    },
+    {
+      "epoch": 1.9802173586417589,
+      "grad_norm": 2.339540958404541,
+      "learning_rate": 3.311269949554923e-05,
+      "loss": 0.4986,
+      "step": 19770
+    },
+    {
+      "epoch": 1.9812190113687584,
+      "grad_norm": 1.8773863315582275,
+      "learning_rate": 3.309778630712047e-05,
+      "loss": 0.6094,
+      "step": 19780
+    },
+    {
+      "epoch": 1.982220664095758,
+      "grad_norm": 2.326773166656494,
+      "learning_rate": 3.30828698986715e-05,
+      "loss": 0.4437,
+      "step": 19790
+    },
+    {
+      "epoch": 1.9832223168227574,
+      "grad_norm": 2.306333541870117,
+      "learning_rate": 3.3067950276133716e-05,
+      "loss": 0.4891,
+      "step": 19800
+    },
+    {
+      "epoch": 1.984223969549757,
+      "grad_norm": 2.0173776149749756,
+      "learning_rate": 3.30530274454398e-05,
+      "loss": 0.5067,
+      "step": 19810
+    },
+    {
+      "epoch": 1.9852256222767566,
+      "grad_norm": 2.5315053462982178,
+      "learning_rate": 3.303810141252368e-05,
+      "loss": 0.4984,
+      "step": 19820
+    },
+    {
+      "epoch": 1.9862272750037562,
+      "grad_norm": 2.567328453063965,
+      "learning_rate": 3.302317218332058e-05,
+      "loss": 0.6134,
+      "step": 19830
+    },
+    {
+      "epoch": 1.9872289277307558,
+      "grad_norm": 2.731470823287964,
+      "learning_rate": 3.300823976376699e-05,
+      "loss": 0.4655,
+      "step": 19840
+    },
+    {
+      "epoch": 1.9882305804577554,
+      "grad_norm": 1.868723750114441,
+      "learning_rate": 3.2993304159800666e-05,
+      "loss": 0.5736,
+      "step": 19850
+    },
+    {
+      "epoch": 1.989232233184755,
+      "grad_norm": 2.778196334838867,
+      "learning_rate": 3.2978365377360625e-05,
+      "loss": 0.5701,
+      "step": 19860
+    },
+    {
+      "epoch": 1.9902338859117545,
+      "grad_norm": 2.3369123935699463,
+      "learning_rate": 3.2963423422387175e-05,
+      "loss": 0.5677,
+      "step": 19870
+    },
+    {
+      "epoch": 1.991235538638754,
+      "grad_norm": 2.2308642864227295,
+      "learning_rate": 3.294847830082184e-05,
+      "loss": 0.5402,
+      "step": 19880
+    },
+    {
+      "epoch": 1.9922371913657535,
+      "grad_norm": 3.2330994606018066,
+      "learning_rate": 3.293353001860745e-05,
+      "loss": 0.5364,
+      "step": 19890
+    },
+    {
+      "epoch": 1.993238844092753,
+      "grad_norm": 2.73323392868042,
+      "learning_rate": 3.291857858168805e-05,
+      "loss": 0.5763,
+      "step": 19900
+    },
+    {
+      "epoch": 1.9942404968197525,
+      "grad_norm": 1.7553918361663818,
+      "learning_rate": 3.2903623996008984e-05,
+      "loss": 0.5238,
+      "step": 19910
+    },
+    {
+      "epoch": 1.995242149546752,
+      "grad_norm": 1.860754370689392,
+      "learning_rate": 3.2888666267516806e-05,
+      "loss": 0.5092,
+      "step": 19920
+    },
+    {
+      "epoch": 1.9962438022737516,
+      "grad_norm": 2.701542377471924,
+      "learning_rate": 3.287370540215934e-05,
+      "loss": 0.4812,
+      "step": 19930
+    },
+    {
+      "epoch": 1.9972454550007512,
+      "grad_norm": 2.274592638015747,
+      "learning_rate": 3.285874140588566e-05,
+      "loss": 0.5499,
+      "step": 19940
+    },
+    {
+      "epoch": 1.9982471077277508,
+      "grad_norm": 2.010769844055176,
+      "learning_rate": 3.2843774284646074e-05,
+      "loss": 0.4987,
+      "step": 19950
+    },
+    {
+      "epoch": 1.9992487604547504,
+      "grad_norm": 1.8435958623886108,
+      "learning_rate": 3.282880404439214e-05,
+      "loss": 0.5644,
+      "step": 19960
+    },
+    {
+      "epoch": 2.0002003305454,
+      "grad_norm": 2.158094644546509,
+      "learning_rate": 3.281383069107666e-05,
+      "loss": 0.5283,
+      "step": 19970
+    },
+    {
+      "epoch": 2.0012019832723995,
+      "grad_norm": 1.6558252573013306,
+      "learning_rate": 3.279885423065367e-05,
+      "loss": 0.4891,
+      "step": 19980
+    },
+    {
+      "epoch": 2.002203635999399,
+      "grad_norm": 2.214022159576416,
+      "learning_rate": 3.278387466907843e-05,
+      "loss": 0.5128,
+      "step": 19990
+    },
+    {
+      "epoch": 2.0032052887263987,
+      "grad_norm": 2.0052435398101807,
+      "learning_rate": 3.276889201230745e-05,
+      "loss": 0.4326,
+      "step": 20000
+    },
+    {
+      "epoch": 2.0032052887263987,
+      "eval_bleu": 0.2002503205345812,
+      "eval_loss": 0.5508657097816467,
+      "eval_rouge1": 0.6375570480064021,
+      "eval_rouge2": 0.4898024330131999,
+      "eval_rougeL": 0.6005766639745327,
+      "eval_runtime": 89842.9594,
+      "eval_samples_per_second": 0.198,
+      "eval_steps_per_second": 0.025,
+      "eval_wer": 1.7028008589955927,
+      "step": 20000
+    },
+    {
+      "epoch": 2.0042069414533983,
+      "grad_norm": 2.0601089000701904,
+      "learning_rate": 3.275390626629846e-05,
+      "loss": 0.453,
+      "step": 20010
+    },
+    {
+      "epoch": 2.005208594180398,
+      "grad_norm": 2.3491313457489014,
+      "learning_rate": 3.273891743701043e-05,
+      "loss": 0.4925,
+      "step": 20020
+    },
+    {
+      "epoch": 2.006210246907397,
+      "grad_norm": 2.14555025100708,
+      "learning_rate": 3.272392553040354e-05,
+      "loss": 0.5125,
+      "step": 20030
+    },
+    {
+      "epoch": 2.0072118996343966,
+      "grad_norm": 2.575279712677002,
+      "learning_rate": 3.270893055243921e-05,
+      "loss": 0.5094,
+      "step": 20040
+    },
+    {
+      "epoch": 2.008213552361396,
+      "grad_norm": 2.2042462825775146,
+      "learning_rate": 3.2693932509080066e-05,
+      "loss": 0.4826,
+      "step": 20050
+    },
+    {
+      "epoch": 2.009215205088396,
+      "grad_norm": 1.9596309661865234,
+      "learning_rate": 3.267893140628996e-05,
+      "loss": 0.5047,
+      "step": 20060
+    },
+    {
+      "epoch": 2.0102168578153954,
+      "grad_norm": 1.733544111251831,
+      "learning_rate": 3.266392725003398e-05,
+      "loss": 0.539,
+      "step": 20070
+    },
+    {
+      "epoch": 2.011218510542395,
+      "grad_norm": 2.1472814083099365,
+      "learning_rate": 3.264892004627838e-05,
+      "loss": 0.5984,
+      "step": 20080
+    },
+    {
+      "epoch": 2.0122201632693946,
+      "grad_norm": 2.2403430938720703,
+      "learning_rate": 3.2633909800990666e-05,
+      "loss": 0.5175,
+      "step": 20090
+    },
+    {
+      "epoch": 2.013221815996394,
+      "grad_norm": 1.9223722219467163,
+      "learning_rate": 3.261889652013955e-05,
+      "loss": 0.4314,
+      "step": 20100
+    },
+    {
+      "epoch": 2.0142234687233938,
+      "grad_norm": 2.1457340717315674,
+      "learning_rate": 3.2603880209694946e-05,
+      "loss": 0.4796,
+      "step": 20110
+    },
+    {
+      "epoch": 2.0152251214503933,
+      "grad_norm": 2.3787331581115723,
+      "learning_rate": 3.258886087562796e-05,
+      "loss": 0.4678,
+      "step": 20120
+    },
+    {
+      "epoch": 2.0162267741773925,
+      "grad_norm": 2.1419434547424316,
+      "learning_rate": 3.2573838523910905e-05,
+      "loss": 0.5041,
+      "step": 20130
+    },
+    {
+      "epoch": 2.017228426904392,
+      "grad_norm": 2.3345863819122314,
+      "learning_rate": 3.255881316051732e-05,
+      "loss": 0.5159,
+      "step": 20140
+    },
+    {
+      "epoch": 2.0182300796313917,
+      "grad_norm": 2.441622734069824,
+      "learning_rate": 3.25437847914219e-05,
+      "loss": 0.4615,
+      "step": 20150
+    },
+    {
+      "epoch": 2.0192317323583913,
+      "grad_norm": 1.5690834522247314,
+      "learning_rate": 3.2528753422600564e-05,
+      "loss": 0.4809,
+      "step": 20160
+    },
+    {
+      "epoch": 2.020233385085391,
+      "grad_norm": 1.9578380584716797,
+      "learning_rate": 3.251371906003042e-05,
+      "loss": 0.5748,
+      "step": 20170
+    },
+    {
+      "epoch": 2.0212350378123904,
+      "grad_norm": 1.7934218645095825,
+      "learning_rate": 3.249868170968977e-05,
+      "loss": 0.4488,
+      "step": 20180
+    },
+    {
+      "epoch": 2.02223669053939,
+      "grad_norm": 2.1038122177124023,
+      "learning_rate": 3.248364137755808e-05,
+      "loss": 0.5006,
+      "step": 20190
+    },
+    {
+      "epoch": 2.0232383432663896,
+      "grad_norm": 1.5591011047363281,
+      "learning_rate": 3.2468598069616016e-05,
+      "loss": 0.4804,
+      "step": 20200
+    },
+    {
+      "epoch": 2.024239995993389,
+      "grad_norm": 2.1266472339630127,
+      "learning_rate": 3.245355179184545e-05,
+      "loss": 0.4946,
+      "step": 20210
+    },
+    {
+      "epoch": 2.025241648720389,
+      "grad_norm": 2.56298565864563,
+      "learning_rate": 3.24385025502294e-05,
+      "loss": 0.4698,
+      "step": 20220
+    },
+    {
+      "epoch": 2.0262433014473884,
+      "grad_norm": 1.669932246208191,
+      "learning_rate": 3.242345035075208e-05,
+      "loss": 0.4983,
+      "step": 20230
+    },
+    {
+      "epoch": 2.0272449541743875,
+      "grad_norm": 2.1388626098632812,
+      "learning_rate": 3.240839519939887e-05,
+      "loss": 0.491,
+      "step": 20240
+    },
+    {
+      "epoch": 2.028246606901387,
+      "grad_norm": 2.082549810409546,
+      "learning_rate": 3.239333710215635e-05,
+      "loss": 0.4602,
+      "step": 20250
+    },
+    {
+      "epoch": 2.0292482596283867,
+      "grad_norm": 3.176657199859619,
+      "learning_rate": 3.237827606501223e-05,
+      "loss": 0.5682,
+      "step": 20260
+    },
+    {
+      "epoch": 2.0302499123553863,
+      "grad_norm": 2.1459405422210693,
+      "learning_rate": 3.236321209395543e-05,
+      "loss": 0.4811,
+      "step": 20270
+    },
+    {
+      "epoch": 2.031251565082386,
+      "grad_norm": 2.0369656085968018,
+      "learning_rate": 3.2348145194976e-05,
+      "loss": 0.507,
+      "step": 20280
+    },
+    {
+      "epoch": 2.0322532178093855,
+      "grad_norm": 2.6629416942596436,
+      "learning_rate": 3.233307537406519e-05,
+      "loss": 0.4818,
+      "step": 20290
+    },
+    {
+      "epoch": 2.033254870536385,
+      "grad_norm": 2.0255343914031982,
+      "learning_rate": 3.231800263721538e-05,
+      "loss": 0.5273,
+      "step": 20300
+    },
+    {
+      "epoch": 2.0342565232633847,
+      "grad_norm": 1.8463060855865479,
+      "learning_rate": 3.230292699042014e-05,
+      "loss": 0.4931,
+      "step": 20310
+    },
+    {
+      "epoch": 2.0352581759903843,
+      "grad_norm": 2.6563429832458496,
+      "learning_rate": 3.228784843967416e-05,
+      "loss": 0.5148,
+      "step": 20320
+    },
+    {
+      "epoch": 2.036259828717384,
+      "grad_norm": 2.2401204109191895,
+      "learning_rate": 3.2272766990973316e-05,
+      "loss": 0.4384,
+      "step": 20330
+    },
+    {
+      "epoch": 2.037261481444383,
+      "grad_norm": 2.7477822303771973,
+      "learning_rate": 3.225768265031463e-05,
+      "loss": 0.5228,
+      "step": 20340
+    },
+    {
+      "epoch": 2.0382631341713826,
+      "grad_norm": 2.590658187866211,
+      "learning_rate": 3.2242595423696277e-05,
+      "loss": 0.4807,
+      "step": 20350
+    },
+    {
+      "epoch": 2.039264786898382,
+      "grad_norm": 2.0068235397338867,
+      "learning_rate": 3.222750531711755e-05,
+      "loss": 0.5156,
+      "step": 20360
+    },
+    {
+      "epoch": 2.040266439625382,
+      "grad_norm": 2.5355372428894043,
+      "learning_rate": 3.2212412336578926e-05,
+      "loss": 0.6042,
+      "step": 20370
+    },
+    {
+      "epoch": 2.0412680923523814,
+      "grad_norm": 2.3344709873199463,
+      "learning_rate": 3.2197316488081995e-05,
+      "loss": 0.5159,
+      "step": 20380
+    },
+    {
+      "epoch": 2.042269745079381,
+      "grad_norm": 2.0762157440185547,
+      "learning_rate": 3.218221777762952e-05,
+      "loss": 0.4592,
+      "step": 20390
+    },
+    {
+      "epoch": 2.0432713978063806,
+      "grad_norm": 2.093581199645996,
+      "learning_rate": 3.216711621122537e-05,
+      "loss": 0.4531,
+      "step": 20400
+    },
+    {
+      "epoch": 2.04427305053338,
+      "grad_norm": 2.9248385429382324,
+      "learning_rate": 3.215201179487456e-05,
+      "loss": 0.469,
+      "step": 20410
+    },
+    {
+      "epoch": 2.0452747032603797,
+      "grad_norm": 1.7484266757965088,
+      "learning_rate": 3.213690453458325e-05,
+      "loss": 0.4237,
+      "step": 20420
+    },
+    {
+      "epoch": 2.0462763559873793,
+      "grad_norm": 1.9926973581314087,
+      "learning_rate": 3.212179443635872e-05,
+      "loss": 0.471,
+      "step": 20430
+    },
+    {
+      "epoch": 2.047278008714379,
+      "grad_norm": 1.6716692447662354,
+      "learning_rate": 3.210668150620937e-05,
+      "loss": 0.4656,
+      "step": 20440
+    },
+    {
+      "epoch": 2.048279661441378,
+      "grad_norm": 1.3834606409072876,
+      "learning_rate": 3.209156575014475e-05,
+      "loss": 0.4969,
+      "step": 20450
+    },
+    {
+      "epoch": 2.0492813141683777,
+      "grad_norm": 1.8815782070159912,
+      "learning_rate": 3.207644717417551e-05,
+      "loss": 0.4847,
+      "step": 20460
+    },
+    {
+      "epoch": 2.0502829668953773,
+      "grad_norm": 1.73691987991333,
+      "learning_rate": 3.2061325784313436e-05,
+      "loss": 0.4458,
+      "step": 20470
+    },
+    {
+      "epoch": 2.051284619622377,
+      "grad_norm": 2.1354963779449463,
+      "learning_rate": 3.204620158657144e-05,
+      "loss": 0.4935,
+      "step": 20480
+    },
+    {
+      "epoch": 2.0522862723493764,
+      "grad_norm": 1.9221725463867188,
+      "learning_rate": 3.20310745869635e-05,
+      "loss": 0.4741,
+      "step": 20490
+    },
+    {
+      "epoch": 2.053287925076376,
+      "grad_norm": 2.065915822982788,
+      "learning_rate": 3.20159447915048e-05,
+      "loss": 0.5215,
+      "step": 20500
+    },
+    {
+      "epoch": 2.0542895778033756,
+      "grad_norm": 2.0788304805755615,
+      "learning_rate": 3.200081220621154e-05,
+      "loss": 0.5446,
+      "step": 20510
+    },
+    {
+      "epoch": 2.055291230530375,
+      "grad_norm": 2.2969303131103516,
+      "learning_rate": 3.1985676837101095e-05,
+      "loss": 0.5335,
+      "step": 20520
+    },
+    {
+      "epoch": 2.056292883257375,
+      "grad_norm": 1.6240125894546509,
+      "learning_rate": 3.1970538690191914e-05,
+      "loss": 0.5019,
+      "step": 20530
+    },
+    {
+      "epoch": 2.0572945359843744,
+      "grad_norm": 2.1084020137786865,
+      "learning_rate": 3.195539777150356e-05,
+      "loss": 0.4362,
+      "step": 20540
+    },
+    {
+      "epoch": 2.058296188711374,
+      "grad_norm": 1.850347876548767,
+      "learning_rate": 3.194025408705671e-05,
+      "loss": 0.4695,
+      "step": 20550
+    },
+    {
+      "epoch": 2.059297841438373,
+      "grad_norm": 2.2284724712371826,
+      "learning_rate": 3.192510764287311e-05,
+      "loss": 0.4978,
+      "step": 20560
+    },
+    {
+      "epoch": 2.0602994941653727,
+      "grad_norm": 2.4977900981903076,
+      "learning_rate": 3.1909958444975636e-05,
+      "loss": 0.5228,
+      "step": 20570
+    },
+    {
+      "epoch": 2.0613011468923723,
+      "grad_norm": 3.1662728786468506,
+      "learning_rate": 3.189480649938825e-05,
+      "loss": 0.4964,
+      "step": 20580
+    },
+    {
+      "epoch": 2.062302799619372,
+      "grad_norm": 1.9692362546920776,
+      "learning_rate": 3.1879651812135995e-05,
+      "loss": 0.4977,
+      "step": 20590
+    },
+    {
+      "epoch": 2.0633044523463715,
+      "grad_norm": 1.8094409704208374,
+      "learning_rate": 3.1864494389245016e-05,
+      "loss": 0.4916,
+      "step": 20600
+    },
+    {
+      "epoch": 2.064306105073371,
+      "grad_norm": 2.69331431388855,
+      "learning_rate": 3.184933423674254e-05,
+      "loss": 0.4704,
+      "step": 20610
+    },
+    {
+      "epoch": 2.0653077578003707,
+      "grad_norm": 1.9121180772781372,
+      "learning_rate": 3.183417136065686e-05,
+      "loss": 0.5054,
+      "step": 20620
+    },
+    {
+      "epoch": 2.0663094105273703,
+      "grad_norm": 1.9425513744354248,
+      "learning_rate": 3.1819005767017406e-05,
+      "loss": 0.437,
+      "step": 20630
+    },
+    {
+      "epoch": 2.06731106325437,
+      "grad_norm": 2.331561326980591,
+      "learning_rate": 3.180383746185464e-05,
+      "loss": 0.4489,
+      "step": 20640
+    },
+    {
+      "epoch": 2.0683127159813695,
+      "grad_norm": 2.1345226764678955,
+      "learning_rate": 3.178866645120012e-05,
+      "loss": 0.488,
+      "step": 20650
+    },
+    {
+      "epoch": 2.0693143687083686,
+      "grad_norm": 2.7039108276367188,
+      "learning_rate": 3.1773492741086474e-05,
+      "loss": 0.4713,
+      "step": 20660
+    },
+    {
+      "epoch": 2.070316021435368,
+      "grad_norm": 2.335880756378174,
+      "learning_rate": 3.1758316337547414e-05,
+      "loss": 0.5457,
+      "step": 20670
+    },
+    {
+      "epoch": 2.071317674162368,
+      "grad_norm": 1.7798826694488525,
+      "learning_rate": 3.1743137246617714e-05,
+      "loss": 0.5122,
+      "step": 20680
+    },
+    {
+      "epoch": 2.0723193268893674,
+      "grad_norm": 1.9185101985931396,
+      "learning_rate": 3.172795547433321e-05,
+      "loss": 0.5018,
+      "step": 20690
+    },
+    {
+      "epoch": 2.073320979616367,
+      "grad_norm": 2.214864730834961,
+      "learning_rate": 3.171277102673082e-05,
+      "loss": 0.4785,
+      "step": 20700
+    },
+    {
+      "epoch": 2.0743226323433666,
+      "grad_norm": 2.6873626708984375,
+      "learning_rate": 3.169758390984852e-05,
+      "loss": 0.4523,
+      "step": 20710
+    },
+    {
+      "epoch": 2.075324285070366,
+      "grad_norm": 1.6275378465652466,
+      "learning_rate": 3.168239412972534e-05,
+      "loss": 0.5009,
+      "step": 20720
+    },
+    {
+      "epoch": 2.0763259377973657,
+      "grad_norm": 1.7685691118240356,
+      "learning_rate": 3.166720169240138e-05,
+      "loss": 0.4612,
+      "step": 20730
+    },
+    {
+      "epoch": 2.0773275905243653,
+      "grad_norm": 2.1459972858428955,
+      "learning_rate": 3.165200660391779e-05,
+      "loss": 0.441,
+      "step": 20740
+    },
+    {
+      "epoch": 2.078329243251365,
+      "grad_norm": 2.075395107269287,
+      "learning_rate": 3.1636808870316775e-05,
+      "loss": 0.4499,
+      "step": 20750
+    },
+    {
+      "epoch": 2.0793308959783645,
+      "grad_norm": 2.5253570079803467,
+      "learning_rate": 3.1621608497641605e-05,
+      "loss": 0.5045,
+      "step": 20760
+    },
+    {
+      "epoch": 2.0803325487053637,
+      "grad_norm": 2.685187816619873,
+      "learning_rate": 3.1606405491936566e-05,
+      "loss": 0.5111,
+      "step": 20770
+    },
+    {
+      "epoch": 2.0813342014323633,
+      "grad_norm": 2.0310208797454834,
+      "learning_rate": 3.159119985924702e-05,
+      "loss": 0.4889,
+      "step": 20780
+    },
+    {
+      "epoch": 2.082335854159363,
+      "grad_norm": 2.0942888259887695,
+      "learning_rate": 3.157599160561937e-05,
+      "loss": 0.4823,
+      "step": 20790
+    },
+    {
+      "epoch": 2.0833375068863624,
+      "grad_norm": 2.0800859928131104,
+      "learning_rate": 3.156078073710105e-05,
+      "loss": 0.4514,
+      "step": 20800
+    },
+    {
+      "epoch": 2.084339159613362,
+      "grad_norm": 2.227480411529541,
+      "learning_rate": 3.1545567259740554e-05,
+      "loss": 0.4826,
+      "step": 20810
+    },
+    {
+      "epoch": 2.0853408123403616,
+      "grad_norm": 2.061187744140625,
+      "learning_rate": 3.153035117958739e-05,
+      "loss": 0.4673,
+      "step": 20820
+    },
+    {
+      "epoch": 2.086342465067361,
+      "grad_norm": 1.741713047027588,
+      "learning_rate": 3.15151325026921e-05,
+      "loss": 0.4825,
+      "step": 20830
+    },
+    {
+      "epoch": 2.087344117794361,
+      "grad_norm": 4.484399795532227,
+      "learning_rate": 3.14999112351063e-05,
+      "loss": 0.5185,
+      "step": 20840
+    },
+    {
+      "epoch": 2.0883457705213604,
+      "grad_norm": 1.8805603981018066,
+      "learning_rate": 3.148468738288258e-05,
+      "loss": 0.4914,
+      "step": 20850
+    },
+    {
+      "epoch": 2.08934742324836,
+      "grad_norm": 2.6285297870635986,
+      "learning_rate": 3.146946095207459e-05,
+      "loss": 0.4523,
+      "step": 20860
+    },
+    {
+      "epoch": 2.090349075975359,
+      "grad_norm": 2.502845525741577,
+      "learning_rate": 3.1454231948737e-05,
+      "loss": 0.4848,
+      "step": 20870
+    },
+    {
+      "epoch": 2.0913507287023587,
+      "grad_norm": 2.0391061305999756,
+      "learning_rate": 3.1439000378925496e-05,
+      "loss": 0.5003,
+      "step": 20880
+    },
+    {
+      "epoch": 2.0923523814293583,
+      "grad_norm": 2.1008460521698,
+      "learning_rate": 3.1423766248696804e-05,
+      "loss": 0.4715,
+      "step": 20890
+    },
+    {
+      "epoch": 2.093354034156358,
+      "grad_norm": 2.340632200241089,
+      "learning_rate": 3.1408529564108644e-05,
+      "loss": 0.5281,
+      "step": 20900
+    },
+    {
+      "epoch": 2.0943556868833575,
+      "grad_norm": 2.45475172996521,
+      "learning_rate": 3.139329033121977e-05,
+      "loss": 0.4497,
+      "step": 20910
+    },
+    {
+      "epoch": 2.095357339610357,
+      "grad_norm": 1.7948698997497559,
+      "learning_rate": 3.137804855608993e-05,
+      "loss": 0.4581,
+      "step": 20920
+    },
+    {
+      "epoch": 2.0963589923373567,
+      "grad_norm": 2.1844940185546875,
+      "learning_rate": 3.1362804244779906e-05,
+      "loss": 0.447,
+      "step": 20930
+    },
+    {
+      "epoch": 2.0973606450643563,
+      "grad_norm": 2.17244553565979,
+      "learning_rate": 3.134755740335147e-05,
+      "loss": 0.5166,
+      "step": 20940
+    },
+    {
+      "epoch": 2.098362297791356,
+      "grad_norm": 2.8408572673797607,
+      "learning_rate": 3.133230803786741e-05,
+      "loss": 0.4978,
+      "step": 20950
+    },
+    {
+      "epoch": 2.0993639505183554,
+      "grad_norm": 2.3953614234924316,
+      "learning_rate": 3.131705615439152e-05,
+      "loss": 0.4383,
+      "step": 20960
+    },
+    {
+      "epoch": 2.100365603245355,
+      "grad_norm": 1.9969464540481567,
+      "learning_rate": 3.130180175898857e-05,
+      "loss": 0.4647,
+      "step": 20970
+    },
+    {
+      "epoch": 2.101367255972354,
+      "grad_norm": 1.826015591621399,
+      "learning_rate": 3.1286544857724366e-05,
+      "loss": 0.4327,
+      "step": 20980
+    },
+    {
+      "epoch": 2.1023689086993538,
+      "grad_norm": 2.493840217590332,
+      "learning_rate": 3.1271285456665696e-05,
+      "loss": 0.4594,
+      "step": 20990
+    },
+    {
+      "epoch": 2.1033705614263534,
+      "grad_norm": 2.08193039894104,
+      "learning_rate": 3.125602356188032e-05,
+      "loss": 0.4489,
+      "step": 21000
+    },
+    {
+      "epoch": 2.104372214153353,
+      "grad_norm": 2.3533523082733154,
+      "learning_rate": 3.124075917943704e-05,
+      "loss": 0.4333,
+      "step": 21010
+    },
+    {
+      "epoch": 2.1053738668803526,
+      "grad_norm": 2.025791645050049,
+      "learning_rate": 3.122549231540558e-05,
+      "loss": 0.4217,
+      "step": 21020
+    },
+    {
+      "epoch": 2.106375519607352,
+      "grad_norm": 2.909153699874878,
+      "learning_rate": 3.121022297585672e-05,
+      "loss": 0.5087,
+      "step": 21030
+    },
+    {
+      "epoch": 2.1073771723343517,
+      "grad_norm": 3.3912322521209717,
+      "learning_rate": 3.1194951166862165e-05,
+      "loss": 0.4623,
+      "step": 21040
+    },
+    {
+      "epoch": 2.1083788250613513,
+      "grad_norm": 2.3475584983825684,
+      "learning_rate": 3.117967689449464e-05,
+      "loss": 0.5099,
+      "step": 21050
+    },
+    {
+      "epoch": 2.109380477788351,
+      "grad_norm": 2.320850133895874,
+      "learning_rate": 3.116440016482784e-05,
+      "loss": 0.4373,
+      "step": 21060
+    },
+    {
+      "epoch": 2.1103821305153505,
+      "grad_norm": 1.9819415807724,
+      "learning_rate": 3.1149120983936434e-05,
+      "loss": 0.469,
+      "step": 21070
+    },
+    {
+      "epoch": 2.11138378324235,
+      "grad_norm": 1.4946792125701904,
+      "learning_rate": 3.1133839357896055e-05,
+      "loss": 0.4708,
+      "step": 21080
+    },
+    {
+      "epoch": 2.1123854359693492,
+      "grad_norm": 2.409613847732544,
+      "learning_rate": 3.1118555292783336e-05,
+      "loss": 0.513,
+      "step": 21090
+    },
+    {
+      "epoch": 2.113387088696349,
+      "grad_norm": 2.309664011001587,
+      "learning_rate": 3.110326879467585e-05,
+      "loss": 0.521,
+      "step": 21100
+    },
+    {
+      "epoch": 2.1143887414233484,
+      "grad_norm": 1.7967244386672974,
+      "learning_rate": 3.108797986965217e-05,
+      "loss": 0.4797,
+      "step": 21110
+    },
+    {
+      "epoch": 2.115390394150348,
+      "grad_norm": 2.344731330871582,
+      "learning_rate": 3.107268852379179e-05,
+      "loss": 0.5341,
+      "step": 21120
+    },
+    {
+      "epoch": 2.1163920468773476,
+      "grad_norm": 1.8803640604019165,
+      "learning_rate": 3.105739476317521e-05,
+      "loss": 0.4447,
+      "step": 21130
+    },
+    {
+      "epoch": 2.117393699604347,
+      "grad_norm": 2.1133954524993896,
+      "learning_rate": 3.104209859388387e-05,
+      "loss": 0.5363,
+      "step": 21140
+    },
+    {
+      "epoch": 2.118395352331347,
+      "grad_norm": 2.5146517753601074,
+      "learning_rate": 3.102680002200017e-05,
+      "loss": 0.5577,
+      "step": 21150
+    },
+    {
+      "epoch": 2.1193970050583464,
+      "grad_norm": 1.94539213180542,
+      "learning_rate": 3.1011499053607445e-05,
+      "loss": 0.4741,
+      "step": 21160
+    },
+    {
+      "epoch": 2.120398657785346,
+      "grad_norm": 2.0953662395477295,
+      "learning_rate": 3.0996195694790037e-05,
+      "loss": 0.4667,
+      "step": 21170
+    },
+    {
+      "epoch": 2.1214003105123456,
+      "grad_norm": 2.344572067260742,
+      "learning_rate": 3.098088995163318e-05,
+      "loss": 0.5503,
+      "step": 21180
+    },
+    {
+      "epoch": 2.1224019632393447,
+      "grad_norm": 2.016023874282837,
+      "learning_rate": 3.096558183022309e-05,
+      "loss": 0.5102,
+      "step": 21190
+    },
+    {
+      "epoch": 2.1234036159663443,
+      "grad_norm": 1.793877363204956,
+      "learning_rate": 3.095027133664692e-05,
+      "loss": 0.4622,
+      "step": 21200
+    },
+    {
+      "epoch": 2.124405268693344,
+      "grad_norm": 1.8648747205734253,
+      "learning_rate": 3.093495847699276e-05,
+      "loss": 0.5166,
+      "step": 21210
+    },
+    {
+      "epoch": 2.1254069214203435,
+      "grad_norm": 2.274259328842163,
+      "learning_rate": 3.091964325734965e-05,
+      "loss": 0.503,
+      "step": 21220
+    },
+    {
+      "epoch": 2.126408574147343,
+      "grad_norm": 2.119114398956299,
+      "learning_rate": 3.090432568380757e-05,
+      "loss": 0.4296,
+      "step": 21230
+    },
+    {
+      "epoch": 2.1274102268743427,
+      "grad_norm": 2.1176578998565674,
+      "learning_rate": 3.088900576245742e-05,
+      "loss": 0.4331,
+      "step": 21240
+    },
+    {
+      "epoch": 2.1284118796013423,
+      "grad_norm": 2.4417192935943604,
+      "learning_rate": 3.087368349939106e-05,
+      "loss": 0.5897,
+      "step": 21250
+    },
+    {
+      "epoch": 2.129413532328342,
+      "grad_norm": 2.3847012519836426,
+      "learning_rate": 3.0858358900701254e-05,
+      "loss": 0.4446,
+      "step": 21260
+    },
+    {
+      "epoch": 2.1304151850553414,
+      "grad_norm": 2.534501791000366,
+      "learning_rate": 3.0843031972481706e-05,
+      "loss": 0.5521,
+      "step": 21270
+    },
+    {
+      "epoch": 2.131416837782341,
+      "grad_norm": 2.3288180828094482,
+      "learning_rate": 3.082770272082706e-05,
+      "loss": 0.4715,
+      "step": 21280
+    },
+    {
+      "epoch": 2.13241849050934,
+      "grad_norm": 2.3620693683624268,
+      "learning_rate": 3.081237115183285e-05,
+      "loss": 0.4759,
+      "step": 21290
+    },
+    {
+      "epoch": 2.1334201432363398,
+      "grad_norm": 1.644264578819275,
+      "learning_rate": 3.079703727159556e-05,
+      "loss": 0.4878,
+      "step": 21300
+    },
+    {
+      "epoch": 2.1344217959633394,
+      "grad_norm": 1.9679802656173706,
+      "learning_rate": 3.07817010862126e-05,
+      "loss": 0.4896,
+      "step": 21310
+    },
+    {
+      "epoch": 2.135423448690339,
+      "grad_norm": 1.9164067506790161,
+      "learning_rate": 3.0766362601782276e-05,
+      "loss": 0.4766,
+      "step": 21320
+    },
+    {
+      "epoch": 2.1364251014173385,
+      "grad_norm": 3.025406837463379,
+      "learning_rate": 3.075102182440379e-05,
+      "loss": 0.5287,
+      "step": 21330
+    },
+    {
+      "epoch": 2.137426754144338,
+      "grad_norm": 2.551084280014038,
+      "learning_rate": 3.073567876017732e-05,
+      "loss": 0.5036,
+      "step": 21340
+    },
+    {
+      "epoch": 2.1384284068713377,
+      "grad_norm": 1.9237477779388428,
+      "learning_rate": 3.072033341520388e-05,
+      "loss": 0.5097,
+      "step": 21350
+    },
+    {
+      "epoch": 2.1394300595983373,
+      "grad_norm": 1.697118878364563,
+      "learning_rate": 3.070498579558543e-05,
+      "loss": 0.4359,
+      "step": 21360
+    },
+    {
+      "epoch": 2.140431712325337,
+      "grad_norm": 1.9887810945510864,
+      "learning_rate": 3.0689635907424845e-05,
+      "loss": 0.4565,
+      "step": 21370
+    },
+    {
+      "epoch": 2.1414333650523365,
+      "grad_norm": 2.0479846000671387,
+      "learning_rate": 3.067428375682587e-05,
+      "loss": 0.4848,
+      "step": 21380
+    },
+    {
+      "epoch": 2.142435017779336,
+      "grad_norm": 1.8660268783569336,
+      "learning_rate": 3.0658929349893174e-05,
+      "loss": 0.4491,
+      "step": 21390
+    },
+    {
+      "epoch": 2.1434366705063352,
+      "grad_norm": 2.399012327194214,
+      "learning_rate": 3.0643572692732305e-05,
+      "loss": 0.5697,
+      "step": 21400
+    },
+    {
+      "epoch": 2.144438323233335,
+      "grad_norm": 2.653744697570801,
+      "learning_rate": 3.062821379144973e-05,
+      "loss": 0.5021,
+      "step": 21410
+    },
+    {
+      "epoch": 2.1454399759603344,
+      "grad_norm": 2.3516993522644043,
+      "learning_rate": 3.061285265215279e-05,
+      "loss": 0.4619,
+      "step": 21420
+    },
+    {
+      "epoch": 2.146441628687334,
+      "grad_norm": 2.4360647201538086,
+      "learning_rate": 3.059748928094971e-05,
+      "loss": 0.4099,
+      "step": 21430
+    },
+    {
+      "epoch": 2.1474432814143336,
+      "grad_norm": 1.9208985567092896,
+      "learning_rate": 3.058212368394962e-05,
+      "loss": 0.4671,
+      "step": 21440
+    },
+    {
+      "epoch": 2.148444934141333,
+      "grad_norm": 1.6549501419067383,
+      "learning_rate": 3.056675586726252e-05,
+      "loss": 0.5079,
+      "step": 21450
+    },
+    {
+      "epoch": 2.149446586868333,
+      "grad_norm": 2.041652202606201,
+      "learning_rate": 3.05513858369993e-05,
+      "loss": 0.4768,
+      "step": 21460
+    },
+    {
+      "epoch": 2.1504482395953324,
+      "grad_norm": 2.64847731590271,
+      "learning_rate": 3.053601359927174e-05,
+      "loss": 0.499,
+      "step": 21470
+    },
+    {
+      "epoch": 2.151449892322332,
+      "grad_norm": 2.6299538612365723,
+      "learning_rate": 3.0520639160192484e-05,
+      "loss": 0.4828,
+      "step": 21480
+    },
+    {
+      "epoch": 2.1524515450493316,
+      "grad_norm": 2.515505790710449,
+      "learning_rate": 3.0505262525875045e-05,
+      "loss": 0.4939,
+      "step": 21490
+    },
+    {
+      "epoch": 2.153453197776331,
+      "grad_norm": 2.789552927017212,
+      "learning_rate": 3.048988370243384e-05,
+      "loss": 0.4844,
+      "step": 21500
+    },
+    {
+      "epoch": 2.1544548505033303,
+      "grad_norm": 2.123347282409668,
+      "learning_rate": 3.0474502695984113e-05,
+      "loss": 0.5176,
+      "step": 21510
+    },
+    {
+      "epoch": 2.15545650323033,
+      "grad_norm": 2.637828826904297,
+      "learning_rate": 3.045911951264202e-05,
+      "loss": 0.4876,
+      "step": 21520
+    },
+    {
+      "epoch": 2.1564581559573295,
+      "grad_norm": 1.7636555433273315,
+      "learning_rate": 3.0443734158524538e-05,
+      "loss": 0.4821,
+      "step": 21530
+    },
+    {
+      "epoch": 2.157459808684329,
+      "grad_norm": 2.0353643894195557,
+      "learning_rate": 3.042834663974955e-05,
+      "loss": 0.4496,
+      "step": 21540
+    },
+    {
+      "epoch": 2.1584614614113287,
+      "grad_norm": 1.786335825920105,
+      "learning_rate": 3.0412956962435773e-05,
+      "loss": 0.4764,
+      "step": 21550
+    },
+    {
+      "epoch": 2.1594631141383283,
+      "grad_norm": 1.59462571144104,
+      "learning_rate": 3.039756513270278e-05,
+      "loss": 0.4608,
+      "step": 21560
+    },
+    {
+      "epoch": 2.160464766865328,
+      "grad_norm": 2.1465108394622803,
+      "learning_rate": 3.0382171156671028e-05,
+      "loss": 0.5007,
+      "step": 21570
+    },
+    {
+      "epoch": 2.1614664195923274,
+      "grad_norm": 1.9091776609420776,
+      "learning_rate": 3.0366775040461802e-05,
+      "loss": 0.5191,
+      "step": 21580
+    },
+    {
+      "epoch": 2.162468072319327,
+      "grad_norm": 2.044081449508667,
+      "learning_rate": 3.035137679019724e-05,
+      "loss": 0.4763,
+      "step": 21590
+    },
+    {
+      "epoch": 2.1634697250463266,
+      "grad_norm": 1.9604840278625488,
+      "learning_rate": 3.033597641200034e-05,
+      "loss": 0.4886,
+      "step": 21600
+    },
+    {
+      "epoch": 2.164471377773326,
+      "grad_norm": 2.2669661045074463,
+      "learning_rate": 3.032057391199493e-05,
+      "loss": 0.4859,
+      "step": 21610
+    },
+    {
+      "epoch": 2.1654730305003254,
+      "grad_norm": 2.0096001625061035,
+      "learning_rate": 3.0305169296305695e-05,
+      "loss": 0.5297,
+      "step": 21620
+    },
+    {
+      "epoch": 2.166474683227325,
+      "grad_norm": 2.0376088619232178,
+      "learning_rate": 3.0289762571058167e-05,
+      "loss": 0.47,
+      "step": 21630
+    },
+    {
+      "epoch": 2.1674763359543245,
+      "grad_norm": 2.465101718902588,
+      "learning_rate": 3.0274353742378697e-05,
+      "loss": 0.5087,
+      "step": 21640
+    },
+    {
+      "epoch": 2.168477988681324,
+      "grad_norm": 1.4533342123031616,
+      "learning_rate": 3.0258942816394492e-05,
+      "loss": 0.52,
+      "step": 21650
+    },
+    {
+      "epoch": 2.1694796414083237,
+      "grad_norm": 1.972916841506958,
+      "learning_rate": 3.0243529799233572e-05,
+      "loss": 0.4282,
+      "step": 21660
+    },
+    {
+      "epoch": 2.1704812941353233,
+      "grad_norm": 1.6645925045013428,
+      "learning_rate": 3.0228114697024813e-05,
+      "loss": 0.4742,
+      "step": 21670
+    },
+    {
+      "epoch": 2.171482946862323,
+      "grad_norm": 2.133258581161499,
+      "learning_rate": 3.0212697515897893e-05,
+      "loss": 0.5186,
+      "step": 21680
+    },
+    {
+      "epoch": 2.1724845995893225,
+      "grad_norm": 1.7237110137939453,
+      "learning_rate": 3.0197278261983343e-05,
+      "loss": 0.4855,
+      "step": 21690
+    },
+    {
+      "epoch": 2.173486252316322,
+      "grad_norm": 1.968004822731018,
+      "learning_rate": 3.0181856941412512e-05,
+      "loss": 0.4972,
+      "step": 21700
+    },
+    {
+      "epoch": 2.1744879050433217,
+      "grad_norm": 1.9006356000900269,
+      "learning_rate": 3.0166433560317543e-05,
+      "loss": 0.4153,
+      "step": 21710
+    },
+    {
+      "epoch": 2.175489557770321,
+      "grad_norm": 2.670501708984375,
+      "learning_rate": 3.0151008124831438e-05,
+      "loss": 0.5024,
+      "step": 21720
+    },
+    {
+      "epoch": 2.1764912104973204,
+      "grad_norm": 1.9438066482543945,
+      "learning_rate": 3.0135580641088e-05,
+      "loss": 0.4693,
+      "step": 21730
+    },
+    {
+      "epoch": 2.17749286322432,
+      "grad_norm": 2.635740280151367,
+      "learning_rate": 3.0120151115221834e-05,
+      "loss": 0.5429,
+      "step": 21740
+    },
+    {
+      "epoch": 2.1784945159513196,
+      "grad_norm": 1.8196395635604858,
+      "learning_rate": 3.010471955336838e-05,
+      "loss": 0.4438,
+      "step": 21750
+    },
+    {
+      "epoch": 2.179496168678319,
+      "grad_norm": 2.196422576904297,
+      "learning_rate": 3.008928596166386e-05,
+      "loss": 0.5454,
+      "step": 21760
+    },
+    {
+      "epoch": 2.180497821405319,
+      "grad_norm": 2.2442145347595215,
+      "learning_rate": 3.0073850346245337e-05,
+      "loss": 0.4403,
+      "step": 21770
+    },
+    {
+      "epoch": 2.1814994741323184,
+      "grad_norm": 1.8484165668487549,
+      "learning_rate": 3.005841271325065e-05,
+      "loss": 0.4638,
+      "step": 21780
+    },
+    {
+      "epoch": 2.182501126859318,
+      "grad_norm": 1.8003365993499756,
+      "learning_rate": 3.004297306881845e-05,
+      "loss": 0.4501,
+      "step": 21790
+    },
+    {
+      "epoch": 2.1835027795863176,
+      "grad_norm": 2.756234884262085,
+      "learning_rate": 3.002753141908819e-05,
+      "loss": 0.4455,
+      "step": 21800
+    },
+    {
+      "epoch": 2.184504432313317,
+      "grad_norm": 1.779693841934204,
+      "learning_rate": 3.001208777020012e-05,
+      "loss": 0.4788,
+      "step": 21810
+    },
+    {
+      "epoch": 2.1855060850403163,
+      "grad_norm": 1.9188098907470703,
+      "learning_rate": 2.999664212829528e-05,
+      "loss": 0.4916,
+      "step": 21820
+    },
+    {
+      "epoch": 2.186507737767316,
+      "grad_norm": 2.3162620067596436,
+      "learning_rate": 2.998119449951552e-05,
+      "loss": 0.4194,
+      "step": 21830
+    },
+    {
+      "epoch": 2.1875093904943155,
+      "grad_norm": 2.515458822250366,
+      "learning_rate": 2.996574489000345e-05,
+      "loss": 0.4606,
+      "step": 21840
+    },
+    {
+      "epoch": 2.188511043221315,
+      "grad_norm": 1.8557404279708862,
+      "learning_rate": 2.9950293305902494e-05,
+      "loss": 0.5494,
+      "step": 21850
+    },
+    {
+      "epoch": 2.1895126959483147,
+      "grad_norm": 1.8735612630844116,
+      "learning_rate": 2.9934839753356846e-05,
+      "loss": 0.4404,
+      "step": 21860
+    },
+    {
+      "epoch": 2.1905143486753142,
+      "grad_norm": 1.6917238235473633,
+      "learning_rate": 2.9919384238511484e-05,
+      "loss": 0.5173,
+      "step": 21870
+    },
+    {
+      "epoch": 2.191516001402314,
+      "grad_norm": 1.8293862342834473,
+      "learning_rate": 2.990392676751219e-05,
+      "loss": 0.4723,
+      "step": 21880
+    },
+    {
+      "epoch": 2.1925176541293134,
+      "grad_norm": 1.5702310800552368,
+      "learning_rate": 2.9888467346505478e-05,
+      "loss": 0.4483,
+      "step": 21890
+    },
+    {
+      "epoch": 2.193519306856313,
+      "grad_norm": 1.7046412229537964,
+      "learning_rate": 2.987300598163868e-05,
+      "loss": 0.4452,
+      "step": 21900
+    },
+    {
+      "epoch": 2.1945209595833126,
+      "grad_norm": 2.7863757610321045,
+      "learning_rate": 2.985754267905989e-05,
+      "loss": 0.574,
+      "step": 21910
+    },
+    {
+      "epoch": 2.195522612310312,
+      "grad_norm": 2.4093217849731445,
+      "learning_rate": 2.9842077444917944e-05,
+      "loss": 0.4968,
+      "step": 21920
+    },
+    {
+      "epoch": 2.1965242650373114,
+      "grad_norm": 2.1676666736602783,
+      "learning_rate": 2.9826610285362494e-05,
+      "loss": 0.4581,
+      "step": 21930
+    },
+    {
+      "epoch": 2.197525917764311,
+      "grad_norm": 2.1943066120147705,
+      "learning_rate": 2.9811141206543914e-05,
+      "loss": 0.4902,
+      "step": 21940
+    },
+    {
+      "epoch": 2.1985275704913105,
+      "grad_norm": 2.6076924800872803,
+      "learning_rate": 2.979567021461337e-05,
+      "loss": 0.5178,
+      "step": 21950
+    },
+    {
+      "epoch": 2.19952922321831,
+      "grad_norm": 1.8468619585037231,
+      "learning_rate": 2.978019731572278e-05,
+      "loss": 0.4675,
+      "step": 21960
+    },
+    {
+      "epoch": 2.2005308759453097,
+      "grad_norm": 2.3149871826171875,
+      "learning_rate": 2.976472251602481e-05,
+      "loss": 0.4665,
+      "step": 21970
+    },
+    {
+      "epoch": 2.2015325286723093,
+      "grad_norm": 2.723015308380127,
+      "learning_rate": 2.97492458216729e-05,
+      "loss": 0.4283,
+      "step": 21980
+    },
+    {
+      "epoch": 2.202534181399309,
+      "grad_norm": 2.4012863636016846,
+      "learning_rate": 2.9733767238821224e-05,
+      "loss": 0.5706,
+      "step": 21990
+    },
+    {
+      "epoch": 2.2035358341263085,
+      "grad_norm": 1.7116522789001465,
+      "learning_rate": 2.9718286773624733e-05,
+      "loss": 0.5375,
+      "step": 22000
+    },
+    {
+      "epoch": 2.204537486853308,
+      "grad_norm": 2.074054002761841,
+      "learning_rate": 2.9702804432239094e-05,
+      "loss": 0.4885,
+      "step": 22010
+    },
+    {
+      "epoch": 2.2055391395803077,
+      "grad_norm": 2.2811458110809326,
+      "learning_rate": 2.9687320220820748e-05,
+      "loss": 0.5012,
+      "step": 22020
+    },
+    {
+      "epoch": 2.2065407923073073,
+      "grad_norm": 1.5797665119171143,
+      "learning_rate": 2.967183414552686e-05,
+      "loss": 0.4347,
+      "step": 22030
+    },
+    {
+      "epoch": 2.2075424450343064,
+      "grad_norm": 1.8701937198638916,
+      "learning_rate": 2.965634621251535e-05,
+      "loss": 0.5473,
+      "step": 22040
+    },
+    {
+      "epoch": 2.208544097761306,
+      "grad_norm": 2.81683349609375,
+      "learning_rate": 2.9640856427944863e-05,
+      "loss": 0.511,
+      "step": 22050
+    },
+    {
+      "epoch": 2.2095457504883056,
+      "grad_norm": 4.241177082061768,
+      "learning_rate": 2.962536479797481e-05,
+      "loss": 0.4851,
+      "step": 22060
+    },
+    {
+      "epoch": 2.210547403215305,
+      "grad_norm": 2.56923246383667,
+      "learning_rate": 2.9609871328765288e-05,
+      "loss": 0.5112,
+      "step": 22070
+    },
+    {
+      "epoch": 2.2115490559423048,
+      "grad_norm": 1.703288197517395,
+      "learning_rate": 2.9594376026477173e-05,
+      "loss": 0.4989,
+      "step": 22080
+    },
+    {
+      "epoch": 2.2125507086693044,
+      "grad_norm": 2.1868696212768555,
+      "learning_rate": 2.9578878897272022e-05,
+      "loss": 0.4711,
+      "step": 22090
+    },
+    {
+      "epoch": 2.213552361396304,
+      "grad_norm": 2.30442214012146,
+      "learning_rate": 2.9563379947312176e-05,
+      "loss": 0.5014,
+      "step": 22100
+    },
+    {
+      "epoch": 2.2145540141233035,
+      "grad_norm": 1.7584575414657593,
+      "learning_rate": 2.954787918276065e-05,
+      "loss": 0.4397,
+      "step": 22110
+    },
+    {
+      "epoch": 2.215555666850303,
+      "grad_norm": 2.35834002494812,
+      "learning_rate": 2.9532376609781197e-05,
+      "loss": 0.4921,
+      "step": 22120
+    },
+    {
+      "epoch": 2.2165573195773027,
+      "grad_norm": 2.399610757827759,
+      "learning_rate": 2.9516872234538306e-05,
+      "loss": 0.4985,
+      "step": 22130
+    },
+    {
+      "epoch": 2.2175589723043023,
+      "grad_norm": 2.3085238933563232,
+      "learning_rate": 2.950136606319716e-05,
+      "loss": 0.5122,
+      "step": 22140
+    },
+    {
+      "epoch": 2.2185606250313015,
+      "grad_norm": 1.8348264694213867,
+      "learning_rate": 2.9485858101923664e-05,
+      "loss": 0.4883,
+      "step": 22150
+    },
+    {
+      "epoch": 2.219562277758301,
+      "grad_norm": 1.838392734527588,
+      "learning_rate": 2.9470348356884436e-05,
+      "loss": 0.4828,
+      "step": 22160
+    },
+    {
+      "epoch": 2.2205639304853007,
+      "grad_norm": 2.5261988639831543,
+      "learning_rate": 2.945483683424681e-05,
+      "loss": 0.4862,
+      "step": 22170
+    },
+    {
+      "epoch": 2.2215655832123002,
+      "grad_norm": 2.360555410385132,
+      "learning_rate": 2.9439323540178803e-05,
+      "loss": 0.4755,
+      "step": 22180
+    },
+    {
+      "epoch": 2.2225672359393,
+      "grad_norm": 1.8836420774459839,
+      "learning_rate": 2.9423808480849162e-05,
+      "loss": 0.5563,
+      "step": 22190
+    },
+    {
+      "epoch": 2.2235688886662994,
+      "grad_norm": 2.629767417907715,
+      "learning_rate": 2.9408291662427334e-05,
+      "loss": 0.4914,
+      "step": 22200
+    },
+    {
+      "epoch": 2.224570541393299,
+      "grad_norm": 1.7553353309631348,
+      "learning_rate": 2.939277309108345e-05,
+      "loss": 0.4442,
+      "step": 22210
+    },
+    {
+      "epoch": 2.2255721941202986,
+      "grad_norm": 1.9337244033813477,
+      "learning_rate": 2.937725277298834e-05,
+      "loss": 0.4718,
+      "step": 22220
+    },
+    {
+      "epoch": 2.226573846847298,
+      "grad_norm": 2.2966341972351074,
+      "learning_rate": 2.936173071431354e-05,
+      "loss": 0.4718,
+      "step": 22230
+    },
+    {
+      "epoch": 2.227575499574298,
+      "grad_norm": 1.795206069946289,
+      "learning_rate": 2.934620692123129e-05,
+      "loss": 0.4941,
+      "step": 22240
+    },
+    {
+      "epoch": 2.228577152301297,
+      "grad_norm": 2.291245698928833,
+      "learning_rate": 2.933068139991447e-05,
+      "loss": 0.47,
+      "step": 22250
+    },
+    {
+      "epoch": 2.2295788050282965,
+      "grad_norm": 2.183180570602417,
+      "learning_rate": 2.931515415653671e-05,
+      "loss": 0.4611,
+      "step": 22260
+    },
+    {
+      "epoch": 2.230580457755296,
+      "grad_norm": 2.6957666873931885,
+      "learning_rate": 2.9299625197272274e-05,
+      "loss": 0.5005,
+      "step": 22270
+    },
+    {
+      "epoch": 2.2315821104822957,
+      "grad_norm": 2.4472970962524414,
+      "learning_rate": 2.9284094528296147e-05,
+      "loss": 0.5073,
+      "step": 22280
+    },
+    {
+      "epoch": 2.2325837632092953,
+      "grad_norm": 2.379964590072632,
+      "learning_rate": 2.9268562155783957e-05,
+      "loss": 0.4925,
+      "step": 22290
+    },
+    {
+      "epoch": 2.233585415936295,
+      "grad_norm": 2.3397207260131836,
+      "learning_rate": 2.9253028085912044e-05,
+      "loss": 0.4374,
+      "step": 22300
+    },
+    {
+      "epoch": 2.2345870686632945,
+      "grad_norm": 1.8619310855865479,
+      "learning_rate": 2.9237492324857397e-05,
+      "loss": 0.4889,
+      "step": 22310
+    },
+    {
+      "epoch": 2.235588721390294,
+      "grad_norm": 1.8362969160079956,
+      "learning_rate": 2.9221954878797686e-05,
+      "loss": 0.43,
+      "step": 22320
+    },
+    {
+      "epoch": 2.2365903741172937,
+      "grad_norm": 3.204434633255005,
+      "learning_rate": 2.9206415753911266e-05,
+      "loss": 0.5093,
+      "step": 22330
+    },
+    {
+      "epoch": 2.2375920268442933,
+      "grad_norm": 2.4659273624420166,
+      "learning_rate": 2.9190874956377136e-05,
+      "loss": 0.4225,
+      "step": 22340
+    },
+    {
+      "epoch": 2.2385936795712924,
+      "grad_norm": 1.955161213874817,
+      "learning_rate": 2.9175332492374963e-05,
+      "loss": 0.5238,
+      "step": 22350
+    },
+    {
+      "epoch": 2.239595332298292,
+      "grad_norm": 1.948196291923523,
+      "learning_rate": 2.91597883680851e-05,
+      "loss": 0.4562,
+      "step": 22360
+    },
+    {
+      "epoch": 2.2405969850252916,
+      "grad_norm": 2.1601054668426514,
+      "learning_rate": 2.9144242589688535e-05,
+      "loss": 0.4433,
+      "step": 22370
+    },
+    {
+      "epoch": 2.241598637752291,
+      "grad_norm": 2.4899299144744873,
+      "learning_rate": 2.9128695163366927e-05,
+      "loss": 0.4281,
+      "step": 22380
+    },
+    {
+      "epoch": 2.2426002904792908,
+      "grad_norm": 2.4548087120056152,
+      "learning_rate": 2.9113146095302584e-05,
+      "loss": 0.5069,
+      "step": 22390
+    },
+    {
+      "epoch": 2.2436019432062904,
+      "grad_norm": 2.362183094024658,
+      "learning_rate": 2.9097595391678467e-05,
+      "loss": 0.4856,
+      "step": 22400
+    },
+    {
+      "epoch": 2.24460359593329,
+      "grad_norm": 2.1862354278564453,
+      "learning_rate": 2.9082043058678192e-05,
+      "loss": 0.4262,
+      "step": 22410
+    },
+    {
+      "epoch": 2.2456052486602895,
+      "grad_norm": 2.056276559829712,
+      "learning_rate": 2.9066489102486023e-05,
+      "loss": 0.4686,
+      "step": 22420
+    },
+    {
+      "epoch": 2.246606901387289,
+      "grad_norm": 2.3078675270080566,
+      "learning_rate": 2.9050933529286857e-05,
+      "loss": 0.5046,
+      "step": 22430
+    },
+    {
+      "epoch": 2.2476085541142887,
+      "grad_norm": 2.5429131984710693,
+      "learning_rate": 2.9035376345266256e-05,
+      "loss": 0.5805,
+      "step": 22440
+    },
+    {
+      "epoch": 2.2486102068412883,
+      "grad_norm": 2.4503445625305176,
+      "learning_rate": 2.90198175566104e-05,
+      "loss": 0.4483,
+      "step": 22450
+    },
+    {
+      "epoch": 2.2496118595682875,
+      "grad_norm": 2.3516530990600586,
+      "learning_rate": 2.9004257169506128e-05,
+      "loss": 0.4998,
+      "step": 22460
+    },
+    {
+      "epoch": 2.250613512295287,
+      "grad_norm": 2.0752251148223877,
+      "learning_rate": 2.898869519014089e-05,
+      "loss": 0.4373,
+      "step": 22470
+    },
+    {
+      "epoch": 2.2516151650222866,
+      "grad_norm": 1.8537724018096924,
+      "learning_rate": 2.8973131624702798e-05,
+      "loss": 0.4856,
+      "step": 22480
+    },
+    {
+      "epoch": 2.2526168177492862,
+      "grad_norm": 1.905326247215271,
+      "learning_rate": 2.8957566479380576e-05,
+      "loss": 0.49,
+      "step": 22490
+    },
+    {
+      "epoch": 2.253618470476286,
+      "grad_norm": 1.725106954574585,
+      "learning_rate": 2.8941999760363563e-05,
+      "loss": 0.4265,
+      "step": 22500
+    },
+    {
+      "epoch": 2.2546201232032854,
+      "grad_norm": 2.606820821762085,
+      "learning_rate": 2.8926431473841763e-05,
+      "loss": 0.4936,
+      "step": 22510
+    },
+    {
+      "epoch": 2.255621775930285,
+      "grad_norm": 2.2708587646484375,
+      "learning_rate": 2.8910861626005776e-05,
+      "loss": 0.4706,
+      "step": 22520
+    },
+    {
+      "epoch": 2.2566234286572846,
+      "grad_norm": 1.8929194211959839,
+      "learning_rate": 2.889529022304682e-05,
+      "loss": 0.4891,
+      "step": 22530
+    },
+    {
+      "epoch": 2.257625081384284,
+      "grad_norm": 2.43183970451355,
+      "learning_rate": 2.887971727115674e-05,
+      "loss": 0.5013,
+      "step": 22540
+    },
+    {
+      "epoch": 2.258626734111284,
+      "grad_norm": 1.686383605003357,
+      "learning_rate": 2.8864142776528007e-05,
+      "loss": 0.4561,
+      "step": 22550
+    },
+    {
+      "epoch": 2.2596283868382834,
+      "grad_norm": 1.827078104019165,
+      "learning_rate": 2.8848566745353683e-05,
+      "loss": 0.423,
+      "step": 22560
+    },
+    {
+      "epoch": 2.2606300395652825,
+      "grad_norm": 2.1075212955474854,
+      "learning_rate": 2.8832989183827464e-05,
+      "loss": 0.4196,
+      "step": 22570
+    },
+    {
+      "epoch": 2.261631692292282,
+      "grad_norm": 2.4093353748321533,
+      "learning_rate": 2.881741009814363e-05,
+      "loss": 0.5551,
+      "step": 22580
+    },
+    {
+      "epoch": 2.2626333450192817,
+      "grad_norm": 2.049708843231201,
+      "learning_rate": 2.8801829494497095e-05,
+      "loss": 0.4111,
+      "step": 22590
+    },
+    {
+      "epoch": 2.2636349977462813,
+      "grad_norm": 2.1682515144348145,
+      "learning_rate": 2.878624737908335e-05,
+      "loss": 0.516,
+      "step": 22600
+    },
+    {
+      "epoch": 2.264636650473281,
+      "grad_norm": 2.5212790966033936,
+      "learning_rate": 2.8770663758098494e-05,
+      "loss": 0.4576,
+      "step": 22610
+    },
+    {
+      "epoch": 2.2656383032002805,
+      "grad_norm": 2.4226863384246826,
+      "learning_rate": 2.8755078637739253e-05,
+      "loss": 0.5454,
+      "step": 22620
+    },
+    {
+      "epoch": 2.26663995592728,
+      "grad_norm": 1.4106111526489258,
+      "learning_rate": 2.8739492024202895e-05,
+      "loss": 0.3773,
+      "step": 22630
+    },
+    {
+      "epoch": 2.2676416086542797,
+      "grad_norm": 2.756021022796631,
+      "learning_rate": 2.8723903923687345e-05,
+      "loss": 0.4958,
+      "step": 22640
+    },
+    {
+      "epoch": 2.2686432613812793,
+      "grad_norm": 2.1808178424835205,
+      "learning_rate": 2.8708314342391062e-05,
+      "loss": 0.4745,
+      "step": 22650
+    },
+    {
+      "epoch": 2.269644914108279,
+      "grad_norm": 2.1560256481170654,
+      "learning_rate": 2.8692723286513133e-05,
+      "loss": 0.465,
+      "step": 22660
+    },
+    {
+      "epoch": 2.2706465668352784,
+      "grad_norm": 2.243764638900757,
+      "learning_rate": 2.8677130762253212e-05,
+      "loss": 0.4736,
+      "step": 22670
+    },
+    {
+      "epoch": 2.2716482195622776,
+      "grad_norm": 2.5886261463165283,
+      "learning_rate": 2.8661536775811542e-05,
+      "loss": 0.4174,
+      "step": 22680
+    },
+    {
+      "epoch": 2.272649872289277,
+      "grad_norm": 2.1652519702911377,
+      "learning_rate": 2.8645941333388948e-05,
+      "loss": 0.5013,
+      "step": 22690
+    },
+    {
+      "epoch": 2.2736515250162768,
+      "grad_norm": 2.6116833686828613,
+      "learning_rate": 2.863034444118683e-05,
+      "loss": 0.4973,
+      "step": 22700
+    },
+    {
+      "epoch": 2.2746531777432764,
+      "grad_norm": 2.0481340885162354,
+      "learning_rate": 2.8614746105407177e-05,
+      "loss": 0.4399,
+      "step": 22710
+    },
+    {
+      "epoch": 2.275654830470276,
+      "grad_norm": 2.0342276096343994,
+      "learning_rate": 2.859914633225253e-05,
+      "loss": 0.5249,
+      "step": 22720
+    },
+    {
+      "epoch": 2.2766564831972755,
+      "grad_norm": 2.232968330383301,
+      "learning_rate": 2.8583545127926025e-05,
+      "loss": 0.4847,
+      "step": 22730
+    },
+    {
+      "epoch": 2.277658135924275,
+      "grad_norm": 3.0498735904693604,
+      "learning_rate": 2.856794249863135e-05,
+      "loss": 0.5188,
+      "step": 22740
+    },
+    {
+      "epoch": 2.2786597886512747,
+      "grad_norm": 1.7310694456100464,
+      "learning_rate": 2.8552338450572768e-05,
+      "loss": 0.4896,
+      "step": 22750
+    },
+    {
+      "epoch": 2.2796614413782743,
+      "grad_norm": 2.2194247245788574,
+      "learning_rate": 2.8536732989955105e-05,
+      "loss": 0.4247,
+      "step": 22760
+    },
+    {
+      "epoch": 2.2806630941052735,
+      "grad_norm": 1.4236329793930054,
+      "learning_rate": 2.852112612298376e-05,
+      "loss": 0.4623,
+      "step": 22770
+    },
+    {
+      "epoch": 2.2816647468322735,
+      "grad_norm": 1.7755547761917114,
+      "learning_rate": 2.850551785586466e-05,
+      "loss": 0.4402,
+      "step": 22780
+    },
+    {
+      "epoch": 2.2826663995592726,
+      "grad_norm": 2.213015079498291,
+      "learning_rate": 2.848990819480431e-05,
+      "loss": 0.4715,
+      "step": 22790
+    },
+    {
+      "epoch": 2.2836680522862722,
+      "grad_norm": 1.86104154586792,
+      "learning_rate": 2.847429714600979e-05,
+      "loss": 0.4849,
+      "step": 22800
+    },
+    {
+      "epoch": 2.284669705013272,
+      "grad_norm": 2.5028235912323,
+      "learning_rate": 2.8458684715688676e-05,
+      "loss": 0.5567,
+      "step": 22810
+    },
+    {
+      "epoch": 2.2856713577402714,
+      "grad_norm": 2.2891948223114014,
+      "learning_rate": 2.8443070910049153e-05,
+      "loss": 0.5308,
+      "step": 22820
+    },
+    {
+      "epoch": 2.286673010467271,
+      "grad_norm": 2.1193156242370605,
+      "learning_rate": 2.8427455735299908e-05,
+      "loss": 0.5333,
+      "step": 22830
+    },
+    {
+      "epoch": 2.2876746631942706,
+      "grad_norm": 1.979323148727417,
+      "learning_rate": 2.841183919765021e-05,
+      "loss": 0.4608,
+      "step": 22840
+    },
+    {
+      "epoch": 2.28867631592127,
+      "grad_norm": 2.817578077316284,
+      "learning_rate": 2.8396221303309832e-05,
+      "loss": 0.3857,
+      "step": 22850
+    },
+    {
+      "epoch": 2.28967796864827,
+      "grad_norm": 2.252032518386841,
+      "learning_rate": 2.8380602058489115e-05,
+      "loss": 0.5232,
+      "step": 22860
+    },
+    {
+      "epoch": 2.2906796213752694,
+      "grad_norm": 4.845706939697266,
+      "learning_rate": 2.8364981469398925e-05,
+      "loss": 0.4782,
+      "step": 22870
+    },
+    {
+      "epoch": 2.2916812741022685,
+      "grad_norm": 2.3078505992889404,
+      "learning_rate": 2.834935954225067e-05,
+      "loss": 0.4634,
+      "step": 22880
+    },
+    {
+      "epoch": 2.292682926829268,
+      "grad_norm": 2.475879192352295,
+      "learning_rate": 2.8333736283256272e-05,
+      "loss": 0.4872,
+      "step": 22890
+    },
+    {
+      "epoch": 2.2936845795562677,
+      "grad_norm": 2.6143455505371094,
+      "learning_rate": 2.8318111698628214e-05,
+      "loss": 0.4772,
+      "step": 22900
+    },
+    {
+      "epoch": 2.2946862322832673,
+      "grad_norm": 2.3077473640441895,
+      "learning_rate": 2.830248579457947e-05,
+      "loss": 0.4862,
+      "step": 22910
+    },
+    {
+      "epoch": 2.295687885010267,
+      "grad_norm": 1.7800365686416626,
+      "learning_rate": 2.8286858577323566e-05,
+      "loss": 0.4625,
+      "step": 22920
+    },
+    {
+      "epoch": 2.2966895377372665,
+      "grad_norm": 2.5691211223602295,
+      "learning_rate": 2.827123005307454e-05,
+      "loss": 0.4215,
+      "step": 22930
+    },
+    {
+      "epoch": 2.297691190464266,
+      "grad_norm": 2.73599910736084,
+      "learning_rate": 2.8255600228046935e-05,
+      "loss": 0.4423,
+      "step": 22940
+    },
+    {
+      "epoch": 2.2986928431912657,
+      "grad_norm": 1.8676958084106445,
+      "learning_rate": 2.8239969108455856e-05,
+      "loss": 0.4174,
+      "step": 22950
+    },
+    {
+      "epoch": 2.2996944959182652,
+      "grad_norm": 1.835666537284851,
+      "learning_rate": 2.8224336700516873e-05,
+      "loss": 0.4988,
+      "step": 22960
+    },
+    {
+      "epoch": 2.300696148645265,
+      "grad_norm": 2.497913122177124,
+      "learning_rate": 2.8208703010446086e-05,
+      "loss": 0.5001,
+      "step": 22970
+    },
+    {
+      "epoch": 2.3016978013722644,
+      "grad_norm": 2.3494608402252197,
+      "learning_rate": 2.8193068044460118e-05,
+      "loss": 0.4715,
+      "step": 22980
+    },
+    {
+      "epoch": 2.3026994540992636,
+      "grad_norm": 1.8175679445266724,
+      "learning_rate": 2.8177431808776078e-05,
+      "loss": 0.4408,
+      "step": 22990
+    },
+    {
+      "epoch": 2.303701106826263,
+      "grad_norm": 2.166266918182373,
+      "learning_rate": 2.8161794309611612e-05,
+      "loss": 0.3983,
+      "step": 23000
+    },
+    {
+      "epoch": 2.3047027595532628,
+      "grad_norm": 1.8629862070083618,
+      "learning_rate": 2.8146155553184812e-05,
+      "loss": 0.4278,
+      "step": 23010
+    },
+    {
+      "epoch": 2.3057044122802623,
+      "grad_norm": 2.0270631313323975,
+      "learning_rate": 2.813051554571433e-05,
+      "loss": 0.5,
+      "step": 23020
+    },
+    {
+      "epoch": 2.306706065007262,
+      "grad_norm": 2.0581462383270264,
+      "learning_rate": 2.811487429341929e-05,
+      "loss": 0.4549,
+      "step": 23030
+    },
+    {
+      "epoch": 2.3077077177342615,
+      "grad_norm": 2.967930793762207,
+      "learning_rate": 2.8099231802519304e-05,
+      "loss": 0.5308,
+      "step": 23040
+    },
+    {
+      "epoch": 2.308709370461261,
+      "grad_norm": 1.926661491394043,
+      "learning_rate": 2.8083588079234485e-05,
+      "loss": 0.4796,
+      "step": 23050
+    },
+    {
+      "epoch": 2.3097110231882607,
+      "grad_norm": 2.6114249229431152,
+      "learning_rate": 2.806794312978544e-05,
+      "loss": 0.5267,
+      "step": 23060
+    },
+    {
+      "epoch": 2.3107126759152603,
+      "grad_norm": 1.7948778867721558,
+      "learning_rate": 2.805229696039325e-05,
+      "loss": 0.461,
+      "step": 23070
+    },
+    {
+      "epoch": 2.31171432864226,
+      "grad_norm": 1.9823604822158813,
+      "learning_rate": 2.8036649577279495e-05,
+      "loss": 0.4477,
+      "step": 23080
+    },
+    {
+      "epoch": 2.3127159813692595,
+      "grad_norm": 2.148059844970703,
+      "learning_rate": 2.8021000986666235e-05,
+      "loss": 0.4411,
+      "step": 23090
+    },
+    {
+      "epoch": 2.3137176340962586,
+      "grad_norm": 2.309586763381958,
+      "learning_rate": 2.8005351194775996e-05,
+      "loss": 0.5383,
+      "step": 23100
+    },
+    {
+      "epoch": 2.3147192868232582,
+      "grad_norm": 2.083111047744751,
+      "learning_rate": 2.7989700207831803e-05,
+      "loss": 0.4957,
+      "step": 23110
+    },
+    {
+      "epoch": 2.315720939550258,
+      "grad_norm": 2.7374930381774902,
+      "learning_rate": 2.7974048032057142e-05,
+      "loss": 0.466,
+      "step": 23120
+    },
+    {
+      "epoch": 2.3167225922772574,
+      "grad_norm": 2.315837860107422,
+      "learning_rate": 2.7958394673675986e-05,
+      "loss": 0.5647,
+      "step": 23130
+    },
+    {
+      "epoch": 2.317724245004257,
+      "grad_norm": 2.1665873527526855,
+      "learning_rate": 2.7942740138912748e-05,
+      "loss": 0.47,
+      "step": 23140
+    },
+    {
+      "epoch": 2.3187258977312566,
+      "grad_norm": 2.3246545791625977,
+      "learning_rate": 2.7927084433992355e-05,
+      "loss": 0.4727,
+      "step": 23150
+    },
+    {
+      "epoch": 2.319727550458256,
+      "grad_norm": 2.2014048099517822,
+      "learning_rate": 2.791142756514015e-05,
+      "loss": 0.4729,
+      "step": 23160
+    },
+    {
+      "epoch": 2.3207292031852558,
+      "grad_norm": 2.3125619888305664,
+      "learning_rate": 2.7895769538581973e-05,
+      "loss": 0.4782,
+      "step": 23170
+    },
+    {
+      "epoch": 2.3217308559122554,
+      "grad_norm": 1.8085143566131592,
+      "learning_rate": 2.788011036054412e-05,
+      "loss": 0.4332,
+      "step": 23180
+    },
+    {
+      "epoch": 2.322732508639255,
+      "grad_norm": 2.0354082584381104,
+      "learning_rate": 2.7864450037253316e-05,
+      "loss": 0.483,
+      "step": 23190
+    },
+    {
+      "epoch": 2.3237341613662545,
+      "grad_norm": 1.992011308670044,
+      "learning_rate": 2.7848788574936787e-05,
+      "loss": 0.4718,
+      "step": 23200
+    },
+    {
+      "epoch": 2.3247358140932537,
+      "grad_norm": 2.2824182510375977,
+      "learning_rate": 2.783312597982218e-05,
+      "loss": 0.5439,
+      "step": 23210
+    },
+    {
+      "epoch": 2.3257374668202533,
+      "grad_norm": 2.152175188064575,
+      "learning_rate": 2.78174622581376e-05,
+      "loss": 0.4938,
+      "step": 23220
+    },
+    {
+      "epoch": 2.326739119547253,
+      "grad_norm": 1.9049335718154907,
+      "learning_rate": 2.7801797416111597e-05,
+      "loss": 0.5244,
+      "step": 23230
+    },
+    {
+      "epoch": 2.3277407722742525,
+      "grad_norm": 1.704810619354248,
+      "learning_rate": 2.7786131459973185e-05,
+      "loss": 0.4052,
+      "step": 23240
+    },
+    {
+      "epoch": 2.328742425001252,
+      "grad_norm": 1.6601033210754395,
+      "learning_rate": 2.7770464395951788e-05,
+      "loss": 0.4639,
+      "step": 23250
+    },
+    {
+      "epoch": 2.3297440777282516,
+      "grad_norm": 2.091416597366333,
+      "learning_rate": 2.7754796230277307e-05,
+      "loss": 0.4594,
+      "step": 23260
+    },
+    {
+      "epoch": 2.3307457304552512,
+      "grad_norm": 2.0445423126220703,
+      "learning_rate": 2.7739126969180053e-05,
+      "loss": 0.4626,
+      "step": 23270
+    },
+    {
+      "epoch": 2.331747383182251,
+      "grad_norm": 2.797914981842041,
+      "learning_rate": 2.7723456618890782e-05,
+      "loss": 0.5565,
+      "step": 23280
+    },
+    {
+      "epoch": 2.3327490359092504,
+      "grad_norm": 2.423527956008911,
+      "learning_rate": 2.770778518564069e-05,
+      "loss": 0.5019,
+      "step": 23290
+    },
+    {
+      "epoch": 2.3337506886362496,
+      "grad_norm": 2.7250008583068848,
+      "learning_rate": 2.7692112675661385e-05,
+      "loss": 0.4707,
+      "step": 23300
+    },
+    {
+      "epoch": 2.3347523413632496,
+      "grad_norm": 2.50681734085083,
+      "learning_rate": 2.7676439095184936e-05,
+      "loss": 0.4785,
+      "step": 23310
+    },
+    {
+      "epoch": 2.3357539940902488,
+      "grad_norm": 2.4695465564727783,
+      "learning_rate": 2.76607644504438e-05,
+      "loss": 0.4693,
+      "step": 23320
+    },
+    {
+      "epoch": 2.3367556468172483,
+      "grad_norm": 2.140570640563965,
+      "learning_rate": 2.7645088747670885e-05,
+      "loss": 0.4847,
+      "step": 23330
+    },
+    {
+      "epoch": 2.337757299544248,
+      "grad_norm": 1.7873948812484741,
+      "learning_rate": 2.7629411993099497e-05,
+      "loss": 0.4984,
+      "step": 23340
+    },
+    {
+      "epoch": 2.3387589522712475,
+      "grad_norm": 2.5162179470062256,
+      "learning_rate": 2.7613734192963385e-05,
+      "loss": 0.5601,
+      "step": 23350
+    },
+    {
+      "epoch": 2.339760604998247,
+      "grad_norm": 1.9141145944595337,
+      "learning_rate": 2.7598055353496705e-05,
+      "loss": 0.4532,
+      "step": 23360
+    },
+    {
+      "epoch": 2.3407622577252467,
+      "grad_norm": 2.3677682876586914,
+      "learning_rate": 2.7582375480934e-05,
+      "loss": 0.4797,
+      "step": 23370
+    },
+    {
+      "epoch": 2.3417639104522463,
+      "grad_norm": 2.3838820457458496,
+      "learning_rate": 2.7566694581510272e-05,
+      "loss": 0.4768,
+      "step": 23380
+    },
+    {
+      "epoch": 2.342765563179246,
+      "grad_norm": 1.4900152683258057,
+      "learning_rate": 2.7551012661460897e-05,
+      "loss": 0.4558,
+      "step": 23390
+    },
+    {
+      "epoch": 2.3437672159062455,
+      "grad_norm": 2.4450631141662598,
+      "learning_rate": 2.7535329727021653e-05,
+      "loss": 0.4752,
+      "step": 23400
+    },
+    {
+      "epoch": 2.3447688686332446,
+      "grad_norm": 2.6570770740509033,
+      "learning_rate": 2.751964578442875e-05,
+      "loss": 0.4175,
+      "step": 23410
+    },
+    {
+      "epoch": 2.345770521360244,
+      "grad_norm": 2.1034457683563232,
+      "learning_rate": 2.7503960839918775e-05,
+      "loss": 0.5427,
+      "step": 23420
+    },
+    {
+      "epoch": 2.346772174087244,
+      "grad_norm": 1.7023869752883911,
+      "learning_rate": 2.7488274899728728e-05,
+      "loss": 0.4666,
+      "step": 23430
+    },
+    {
+      "epoch": 2.3477738268142434,
+      "grad_norm": 1.739460825920105,
+      "learning_rate": 2.7472587970095985e-05,
+      "loss": 0.479,
+      "step": 23440
+    },
+    {
+      "epoch": 2.348775479541243,
+      "grad_norm": 1.9922562837600708,
+      "learning_rate": 2.7456900057258344e-05,
+      "loss": 0.4813,
+      "step": 23450
+    },
+    {
+      "epoch": 2.3497771322682426,
+      "grad_norm": 2.244210720062256,
+      "learning_rate": 2.7441211167453973e-05,
+      "loss": 0.4797,
+      "step": 23460
+    },
+    {
+      "epoch": 2.350778784995242,
+      "grad_norm": 2.837597131729126,
+      "learning_rate": 2.7425521306921427e-05,
+      "loss": 0.4473,
+      "step": 23470
+    },
+    {
+      "epoch": 2.3517804377222418,
+      "grad_norm": 2.2177488803863525,
+      "learning_rate": 2.740983048189966e-05,
+      "loss": 0.5151,
+      "step": 23480
+    },
+    {
+      "epoch": 2.3527820904492414,
+      "grad_norm": 1.891855239868164,
+      "learning_rate": 2.7394138698628e-05,
+      "loss": 0.4739,
+      "step": 23490
+    },
+    {
+      "epoch": 2.353783743176241,
+      "grad_norm": 2.4679808616638184,
+      "learning_rate": 2.7378445963346165e-05,
+      "loss": 0.4497,
+      "step": 23500
+    },
+    {
+      "epoch": 2.3547853959032405,
+      "grad_norm": 1.8632603883743286,
+      "learning_rate": 2.7362752282294252e-05,
+      "loss": 0.4815,
+      "step": 23510
+    },
+    {
+      "epoch": 2.3557870486302397,
+      "grad_norm": 1.9275087118148804,
+      "learning_rate": 2.7347057661712706e-05,
+      "loss": 0.4674,
+      "step": 23520
+    },
+    {
+      "epoch": 2.3567887013572393,
+      "grad_norm": 2.283627986907959,
+      "learning_rate": 2.7331362107842388e-05,
+      "loss": 0.4662,
+      "step": 23530
+    },
+    {
+      "epoch": 2.357790354084239,
+      "grad_norm": 2.1892151832580566,
+      "learning_rate": 2.7315665626924515e-05,
+      "loss": 0.3666,
+      "step": 23540
+    },
+    {
+      "epoch": 2.3587920068112385,
+      "grad_norm": 1.8962342739105225,
+      "learning_rate": 2.7299968225200638e-05,
+      "loss": 0.4528,
+      "step": 23550
+    },
+    {
+      "epoch": 2.359793659538238,
+      "grad_norm": 2.1917147636413574,
+      "learning_rate": 2.7284269908912737e-05,
+      "loss": 0.5034,
+      "step": 23560
+    },
+    {
+      "epoch": 2.3607953122652376,
+      "grad_norm": 1.915441632270813,
+      "learning_rate": 2.7268570684303095e-05,
+      "loss": 0.5584,
+      "step": 23570
+    },
+    {
+      "epoch": 2.3617969649922372,
+      "grad_norm": 2.1486260890960693,
+      "learning_rate": 2.7252870557614402e-05,
+      "loss": 0.4875,
+      "step": 23580
+    },
+    {
+      "epoch": 2.362798617719237,
+      "grad_norm": 2.0003814697265625,
+      "learning_rate": 2.7237169535089686e-05,
+      "loss": 0.4692,
+      "step": 23590
+    },
+    {
+      "epoch": 2.3638002704462364,
+      "grad_norm": 2.1308910846710205,
+      "learning_rate": 2.7221467622972326e-05,
+      "loss": 0.4914,
+      "step": 23600
+    },
+    {
+      "epoch": 2.364801923173236,
+      "grad_norm": 2.5318915843963623,
+      "learning_rate": 2.7205764827506064e-05,
+      "loss": 0.5162,
+      "step": 23610
+    },
+    {
+      "epoch": 2.3658035759002356,
+      "grad_norm": 1.9081600904464722,
+      "learning_rate": 2.7190061154934993e-05,
+      "loss": 0.4971,
+      "step": 23620
+    },
+    {
+      "epoch": 2.3668052286272347,
+      "grad_norm": 3.0676238536834717,
+      "learning_rate": 2.717435661150356e-05,
+      "loss": 0.474,
+      "step": 23630
+    },
+    {
+      "epoch": 2.3678068813542343,
+      "grad_norm": 2.3040099143981934,
+      "learning_rate": 2.7158651203456542e-05,
+      "loss": 0.4463,
+      "step": 23640
+    },
+    {
+      "epoch": 2.368808534081234,
+      "grad_norm": 2.0134572982788086,
+      "learning_rate": 2.7142944937039072e-05,
+      "loss": 0.4961,
+      "step": 23650
+    },
+    {
+      "epoch": 2.3698101868082335,
+      "grad_norm": 2.1416115760803223,
+      "learning_rate": 2.712723781849662e-05,
+      "loss": 0.4916,
+      "step": 23660
+    },
+    {
+      "epoch": 2.370811839535233,
+      "grad_norm": 2.0588133335113525,
+      "learning_rate": 2.7111529854075002e-05,
+      "loss": 0.4562,
+      "step": 23670
+    },
+    {
+      "epoch": 2.3718134922622327,
+      "grad_norm": 1.623829960823059,
+      "learning_rate": 2.7095821050020358e-05,
+      "loss": 0.4011,
+      "step": 23680
+    },
+    {
+      "epoch": 2.3728151449892323,
+      "grad_norm": 2.447624921798706,
+      "learning_rate": 2.708011141257918e-05,
+      "loss": 0.4624,
+      "step": 23690
+    },
+    {
+      "epoch": 2.373816797716232,
+      "grad_norm": 3.0367507934570312,
+      "learning_rate": 2.7064400947998263e-05,
+      "loss": 0.4924,
+      "step": 23700
+    },
+    {
+      "epoch": 2.3748184504432315,
+      "grad_norm": 3.0571627616882324,
+      "learning_rate": 2.7048689662524767e-05,
+      "loss": 0.4647,
+      "step": 23710
+    },
+    {
+      "epoch": 2.375820103170231,
+      "grad_norm": 2.4752557277679443,
+      "learning_rate": 2.7032977562406147e-05,
+      "loss": 0.5044,
+      "step": 23720
+    },
+    {
+      "epoch": 2.3768217558972307,
+      "grad_norm": 2.2401139736175537,
+      "learning_rate": 2.7017264653890202e-05,
+      "loss": 0.5055,
+      "step": 23730
+    },
+    {
+      "epoch": 2.37782340862423,
+      "grad_norm": 2.167793035507202,
+      "learning_rate": 2.700155094322504e-05,
+      "loss": 0.4909,
+      "step": 23740
+    },
+    {
+      "epoch": 2.3788250613512294,
+      "grad_norm": 2.418596029281616,
+      "learning_rate": 2.6985836436659084e-05,
+      "loss": 0.5,
+      "step": 23750
+    },
+    {
+      "epoch": 2.379826714078229,
+      "grad_norm": 2.262390613555908,
+      "learning_rate": 2.6970121140441097e-05,
+      "loss": 0.5482,
+      "step": 23760
+    },
+    {
+      "epoch": 2.3808283668052286,
+      "grad_norm": 2.523616075515747,
+      "learning_rate": 2.6954405060820138e-05,
+      "loss": 0.4713,
+      "step": 23770
+    },
+    {
+      "epoch": 2.381830019532228,
+      "grad_norm": 1.510984182357788,
+      "learning_rate": 2.6938688204045582e-05,
+      "loss": 0.42,
+      "step": 23780
+    },
+    {
+      "epoch": 2.3828316722592278,
+      "grad_norm": 1.5943301916122437,
+      "learning_rate": 2.69229705763671e-05,
+      "loss": 0.4441,
+      "step": 23790
+    },
+    {
+      "epoch": 2.3838333249862274,
+      "grad_norm": 2.2615034580230713,
+      "learning_rate": 2.6907252184034697e-05,
+      "loss": 0.4397,
+      "step": 23800
+    },
+    {
+      "epoch": 2.384834977713227,
+      "grad_norm": 2.052757740020752,
+      "learning_rate": 2.6891533033298656e-05,
+      "loss": 0.4832,
+      "step": 23810
+    },
+    {
+      "epoch": 2.3858366304402265,
+      "grad_norm": 1.8588956594467163,
+      "learning_rate": 2.687581313040958e-05,
+      "loss": 0.4632,
+      "step": 23820
+    },
+    {
+      "epoch": 2.3868382831672257,
+      "grad_norm": 2.4952921867370605,
+      "learning_rate": 2.6860092481618355e-05,
+      "loss": 0.4788,
+      "step": 23830
+    },
+    {
+      "epoch": 2.3878399358942257,
+      "grad_norm": 2.3003437519073486,
+      "learning_rate": 2.684437109317618e-05,
+      "loss": 0.4824,
+      "step": 23840
+    },
+    {
+      "epoch": 2.388841588621225,
+      "grad_norm": 2.297075033187866,
+      "learning_rate": 2.682864897133453e-05,
+      "loss": 0.517,
+      "step": 23850
+    },
+    {
+      "epoch": 2.3898432413482245,
+      "grad_norm": 2.4082040786743164,
+      "learning_rate": 2.6812926122345185e-05,
+      "loss": 0.5695,
+      "step": 23860
+    },
+    {
+      "epoch": 2.390844894075224,
+      "grad_norm": 2.1202104091644287,
+      "learning_rate": 2.679720255246022e-05,
+      "loss": 0.5065,
+      "step": 23870
+    },
+    {
+      "epoch": 2.3918465468022236,
+      "grad_norm": 2.0936365127563477,
+      "learning_rate": 2.6781478267931975e-05,
+      "loss": 0.4594,
+      "step": 23880
+    },
+    {
+      "epoch": 2.3928481995292232,
+      "grad_norm": 2.3779218196868896,
+      "learning_rate": 2.67657532750131e-05,
+      "loss": 0.4314,
+      "step": 23890
+    },
+    {
+      "epoch": 2.393849852256223,
+      "grad_norm": 1.846227765083313,
+      "learning_rate": 2.6750027579956493e-05,
+      "loss": 0.4417,
+      "step": 23900
+    },
+    {
+      "epoch": 2.3948515049832224,
+      "grad_norm": 2.4832935333251953,
+      "learning_rate": 2.6734301189015363e-05,
+      "loss": 0.483,
+      "step": 23910
+    },
+    {
+      "epoch": 2.395853157710222,
+      "grad_norm": 2.117427110671997,
+      "learning_rate": 2.6718574108443196e-05,
+      "loss": 0.4926,
+      "step": 23920
+    },
+    {
+      "epoch": 2.3968548104372216,
+      "grad_norm": 1.825005292892456,
+      "learning_rate": 2.6702846344493714e-05,
+      "loss": 0.4941,
+      "step": 23930
+    },
+    {
+      "epoch": 2.3978564631642207,
+      "grad_norm": 2.119313955307007,
+      "learning_rate": 2.668711790342096e-05,
+      "loss": 0.4495,
+      "step": 23940
+    },
+    {
+      "epoch": 2.3988581158912203,
+      "grad_norm": 2.2307143211364746,
+      "learning_rate": 2.667138879147921e-05,
+      "loss": 0.5252,
+      "step": 23950
+    },
+    {
+      "epoch": 2.39985976861822,
+      "grad_norm": 2.309399127960205,
+      "learning_rate": 2.665565901492303e-05,
+      "loss": 0.5204,
+      "step": 23960
+    },
+    {
+      "epoch": 2.4008614213452195,
+      "grad_norm": 2.689640522003174,
+      "learning_rate": 2.663992858000723e-05,
+      "loss": 0.4835,
+      "step": 23970
+    },
+    {
+      "epoch": 2.401863074072219,
+      "grad_norm": 2.8619892597198486,
+      "learning_rate": 2.6624197492986897e-05,
+      "loss": 0.4268,
+      "step": 23980
+    },
+    {
+      "epoch": 2.4028647267992187,
+      "grad_norm": 1.9714316129684448,
+      "learning_rate": 2.6608465760117374e-05,
+      "loss": 0.4982,
+      "step": 23990
+    },
+    {
+      "epoch": 2.4038663795262183,
+      "grad_norm": 1.962856650352478,
+      "learning_rate": 2.6592733387654255e-05,
+      "loss": 0.4649,
+      "step": 24000
+    },
+    {
+      "epoch": 2.404868032253218,
+      "grad_norm": 2.0655150413513184,
+      "learning_rate": 2.65770003818534e-05,
+      "loss": 0.4452,
+      "step": 24010
+    },
+    {
+      "epoch": 2.4058696849802175,
+      "grad_norm": 2.481144905090332,
+      "learning_rate": 2.6561266748970902e-05,
+      "loss": 0.4823,
+      "step": 24020
+    },
+    {
+      "epoch": 2.406871337707217,
+      "grad_norm": 2.099876642227173,
+      "learning_rate": 2.6545532495263125e-05,
+      "loss": 0.4664,
+      "step": 24030
+    },
+    {
+      "epoch": 2.4078729904342167,
+      "grad_norm": 2.909708261489868,
+      "learning_rate": 2.6529797626986662e-05,
+      "loss": 0.5461,
+      "step": 24040
+    },
+    {
+      "epoch": 2.408874643161216,
+      "grad_norm": 2.3196239471435547,
+      "learning_rate": 2.651406215039837e-05,
+      "loss": 0.4675,
+      "step": 24050
+    },
+    {
+      "epoch": 2.4098762958882154,
+      "grad_norm": 2.3955135345458984,
+      "learning_rate": 2.649832607175532e-05,
+      "loss": 0.493,
+      "step": 24060
+    },
+    {
+      "epoch": 2.410877948615215,
+      "grad_norm": 1.9178582429885864,
+      "learning_rate": 2.6482589397314854e-05,
+      "loss": 0.4758,
+      "step": 24070
+    },
+    {
+      "epoch": 2.4118796013422146,
+      "grad_norm": 2.7419323921203613,
+      "learning_rate": 2.646685213333452e-05,
+      "loss": 0.4915,
+      "step": 24080
+    },
+    {
+      "epoch": 2.412881254069214,
+      "grad_norm": 2.2376959323883057,
+      "learning_rate": 2.645111428607212e-05,
+      "loss": 0.4918,
+      "step": 24090
+    },
+    {
+      "epoch": 2.4138829067962138,
+      "grad_norm": 1.6830248832702637,
+      "learning_rate": 2.6435375861785692e-05,
+      "loss": 0.4887,
+      "step": 24100
+    },
+    {
+      "epoch": 2.4148845595232133,
+      "grad_norm": 1.9309697151184082,
+      "learning_rate": 2.641963686673349e-05,
+      "loss": 0.498,
+      "step": 24110
+    },
+    {
+      "epoch": 2.415886212250213,
+      "grad_norm": 1.619993805885315,
+      "learning_rate": 2.6403897307173996e-05,
+      "loss": 0.3869,
+      "step": 24120
+    },
+    {
+      "epoch": 2.4168878649772125,
+      "grad_norm": 2.886514663696289,
+      "learning_rate": 2.638815718936593e-05,
+      "loss": 0.4794,
+      "step": 24130
+    },
+    {
+      "epoch": 2.417889517704212,
+      "grad_norm": 2.1846656799316406,
+      "learning_rate": 2.6372416519568216e-05,
+      "loss": 0.4546,
+      "step": 24140
+    },
+    {
+      "epoch": 2.4188911704312117,
+      "grad_norm": 2.40488338470459,
+      "learning_rate": 2.6356675304040003e-05,
+      "loss": 0.4387,
+      "step": 24150
+    },
+    {
+      "epoch": 2.419892823158211,
+      "grad_norm": 2.2383806705474854,
+      "learning_rate": 2.634093354904067e-05,
+      "loss": 0.49,
+      "step": 24160
+    },
+    {
+      "epoch": 2.4208944758852105,
+      "grad_norm": 2.0622711181640625,
+      "learning_rate": 2.6325191260829795e-05,
+      "loss": 0.5343,
+      "step": 24170
+    },
+    {
+      "epoch": 2.42189612861221,
+      "grad_norm": 1.8681621551513672,
+      "learning_rate": 2.6309448445667172e-05,
+      "loss": 0.4797,
+      "step": 24180
+    },
+    {
+      "epoch": 2.4228977813392096,
+      "grad_norm": 2.4106242656707764,
+      "learning_rate": 2.6293705109812806e-05,
+      "loss": 0.4516,
+      "step": 24190
+    },
+    {
+      "epoch": 2.4238994340662092,
+      "grad_norm": 2.238723039627075,
+      "learning_rate": 2.627796125952692e-05,
+      "loss": 0.4675,
+      "step": 24200
+    },
+    {
+      "epoch": 2.424901086793209,
+      "grad_norm": 2.0002214908599854,
+      "learning_rate": 2.6262216901069913e-05,
+      "loss": 0.4723,
+      "step": 24210
+    },
+    {
+      "epoch": 2.4259027395202084,
+      "grad_norm": 1.454474687576294,
+      "learning_rate": 2.6246472040702413e-05,
+      "loss": 0.4336,
+      "step": 24220
+    },
+    {
+      "epoch": 2.426904392247208,
+      "grad_norm": 2.259845495223999,
+      "learning_rate": 2.6230726684685246e-05,
+      "loss": 0.4863,
+      "step": 24230
+    },
+    {
+      "epoch": 2.4279060449742076,
+      "grad_norm": 2.448485851287842,
+      "learning_rate": 2.621498083927941e-05,
+      "loss": 0.4837,
+      "step": 24240
+    },
+    {
+      "epoch": 2.428907697701207,
+      "grad_norm": 1.6946465969085693,
+      "learning_rate": 2.619923451074613e-05,
+      "loss": 0.4887,
+      "step": 24250
+    },
+    {
+      "epoch": 2.4299093504282068,
+      "grad_norm": 2.0357892513275146,
+      "learning_rate": 2.6183487705346794e-05,
+      "loss": 0.5002,
+      "step": 24260
+    },
+    {
+      "epoch": 2.430911003155206,
+      "grad_norm": 2.7066149711608887,
+      "learning_rate": 2.6167740429343006e-05,
+      "loss": 0.4957,
+      "step": 24270
+    },
+    {
+      "epoch": 2.4319126558822055,
+      "grad_norm": 2.0033223628997803,
+      "learning_rate": 2.6151992688996546e-05,
+      "loss": 0.462,
+      "step": 24280
+    },
+    {
+      "epoch": 2.432914308609205,
+      "grad_norm": 2.564603328704834,
+      "learning_rate": 2.6136244490569366e-05,
+      "loss": 0.4292,
+      "step": 24290
+    },
+    {
+      "epoch": 2.4339159613362047,
+      "grad_norm": 3.0108890533447266,
+      "learning_rate": 2.6120495840323628e-05,
+      "loss": 0.537,
+      "step": 24300
+    },
+    {
+      "epoch": 2.4349176140632043,
+      "grad_norm": 2.422274351119995,
+      "learning_rate": 2.610474674452163e-05,
+      "loss": 0.4835,
+      "step": 24310
+    },
+    {
+      "epoch": 2.435919266790204,
+      "grad_norm": 2.4366953372955322,
+      "learning_rate": 2.6088997209425903e-05,
+      "loss": 0.4794,
+      "step": 24320
+    },
+    {
+      "epoch": 2.4369209195172035,
+      "grad_norm": 2.602489709854126,
+      "learning_rate": 2.607324724129911e-05,
+      "loss": 0.5207,
+      "step": 24330
+    },
+    {
+      "epoch": 2.437922572244203,
+      "grad_norm": 2.0761661529541016,
+      "learning_rate": 2.6057496846404105e-05,
+      "loss": 0.4881,
+      "step": 24340
+    },
+    {
+      "epoch": 2.4389242249712026,
+      "grad_norm": 2.906827211380005,
+      "learning_rate": 2.60417460310039e-05,
+      "loss": 0.5105,
+      "step": 24350
+    },
+    {
+      "epoch": 2.439925877698202,
+      "grad_norm": 1.975074052810669,
+      "learning_rate": 2.602599480136169e-05,
+      "loss": 0.4768,
+      "step": 24360
+    },
+    {
+      "epoch": 2.440927530425202,
+      "grad_norm": 2.0563461780548096,
+      "learning_rate": 2.6010243163740815e-05,
+      "loss": 0.5195,
+      "step": 24370
+    },
+    {
+      "epoch": 2.441929183152201,
+      "grad_norm": 1.6481271982192993,
+      "learning_rate": 2.599449112440481e-05,
+      "loss": 0.4575,
+      "step": 24380
+    },
+    {
+      "epoch": 2.4429308358792006,
+      "grad_norm": 1.454416275024414,
+      "learning_rate": 2.5978738689617322e-05,
+      "loss": 0.5094,
+      "step": 24390
+    },
+    {
+      "epoch": 2.4439324886062,
+      "grad_norm": 2.3923304080963135,
+      "learning_rate": 2.5962985865642198e-05,
+      "loss": 0.5516,
+      "step": 24400
+    },
+    {
+      "epoch": 2.4449341413331998,
+      "grad_norm": 2.3013486862182617,
+      "learning_rate": 2.5947232658743414e-05,
+      "loss": 0.4856,
+      "step": 24410
+    },
+    {
+      "epoch": 2.4459357940601993,
+      "grad_norm": 2.057035207748413,
+      "learning_rate": 2.593147907518511e-05,
+      "loss": 0.4566,
+      "step": 24420
+    },
+    {
+      "epoch": 2.446937446787199,
+      "grad_norm": 1.8558269739151,
+      "learning_rate": 2.5915725121231582e-05,
+      "loss": 0.4738,
+      "step": 24430
+    },
+    {
+      "epoch": 2.4479390995141985,
+      "grad_norm": 1.7031394243240356,
+      "learning_rate": 2.5899970803147246e-05,
+      "loss": 0.4891,
+      "step": 24440
+    },
+    {
+      "epoch": 2.448940752241198,
+      "grad_norm": 1.8119549751281738,
+      "learning_rate": 2.5884216127196694e-05,
+      "loss": 0.4586,
+      "step": 24450
+    },
+    {
+      "epoch": 2.4499424049681977,
+      "grad_norm": 2.4061672687530518,
+      "learning_rate": 2.5868461099644642e-05,
+      "loss": 0.5631,
+      "step": 24460
+    },
+    {
+      "epoch": 2.450944057695197,
+      "grad_norm": 1.9223402738571167,
+      "learning_rate": 2.5852705726755954e-05,
+      "loss": 0.5479,
+      "step": 24470
+    },
+    {
+      "epoch": 2.4519457104221964,
+      "grad_norm": 2.5272395610809326,
+      "learning_rate": 2.583695001479563e-05,
+      "loss": 0.5184,
+      "step": 24480
+    },
+    {
+      "epoch": 2.452947363149196,
+      "grad_norm": 2.009592056274414,
+      "learning_rate": 2.5821193970028806e-05,
+      "loss": 0.4291,
+      "step": 24490
+    },
+    {
+      "epoch": 2.4539490158761956,
+      "grad_norm": 2.079152822494507,
+      "learning_rate": 2.580543759872074e-05,
+      "loss": 0.4355,
+      "step": 24500
+    },
+    {
+      "epoch": 2.454950668603195,
+      "grad_norm": 1.9347796440124512,
+      "learning_rate": 2.5789680907136826e-05,
+      "loss": 0.5151,
+      "step": 24510
+    },
+    {
+      "epoch": 2.455952321330195,
+      "grad_norm": 3.365255117416382,
+      "learning_rate": 2.5773923901542597e-05,
+      "loss": 0.4749,
+      "step": 24520
+    },
+    {
+      "epoch": 2.4569539740571944,
+      "grad_norm": 2.0016565322875977,
+      "learning_rate": 2.575816658820369e-05,
+      "loss": 0.4134,
+      "step": 24530
+    },
+    {
+      "epoch": 2.457955626784194,
+      "grad_norm": 2.206012725830078,
+      "learning_rate": 2.5742408973385883e-05,
+      "loss": 0.4936,
+      "step": 24540
+    },
+    {
+      "epoch": 2.4589572795111936,
+      "grad_norm": 1.8435074090957642,
+      "learning_rate": 2.5726651063355057e-05,
+      "loss": 0.5471,
+      "step": 24550
+    },
+    {
+      "epoch": 2.459958932238193,
+      "grad_norm": 3.3343753814697266,
+      "learning_rate": 2.5710892864377234e-05,
+      "loss": 0.5009,
+      "step": 24560
+    },
+    {
+      "epoch": 2.4609605849651928,
+      "grad_norm": 2.1347665786743164,
+      "learning_rate": 2.5695134382718522e-05,
+      "loss": 0.4714,
+      "step": 24570
+    },
+    {
+      "epoch": 2.461962237692192,
+      "grad_norm": 2.361144542694092,
+      "learning_rate": 2.567937562464517e-05,
+      "loss": 0.531,
+      "step": 24580
+    },
+    {
+      "epoch": 2.4629638904191915,
+      "grad_norm": 1.9751992225646973,
+      "learning_rate": 2.5663616596423508e-05,
+      "loss": 0.4586,
+      "step": 24590
+    },
+    {
+      "epoch": 2.463965543146191,
+      "grad_norm": 2.1158523559570312,
+      "learning_rate": 2.564785730431999e-05,
+      "loss": 0.598,
+      "step": 24600
+    },
+    {
+      "epoch": 2.4649671958731907,
+      "grad_norm": 2.8670387268066406,
+      "learning_rate": 2.5632097754601185e-05,
+      "loss": 0.5161,
+      "step": 24610
+    },
+    {
+      "epoch": 2.4659688486001903,
+      "grad_norm": 2.2353668212890625,
+      "learning_rate": 2.5616337953533736e-05,
+      "loss": 0.5001,
+      "step": 24620
+    },
+    {
+      "epoch": 2.46697050132719,
+      "grad_norm": 2.321228504180908,
+      "learning_rate": 2.560057790738442e-05,
+      "loss": 0.5423,
+      "step": 24630
+    },
+    {
+      "epoch": 2.4679721540541895,
+      "grad_norm": 2.136357307434082,
+      "learning_rate": 2.5584817622420077e-05,
+      "loss": 0.4286,
+      "step": 24640
+    },
+    {
+      "epoch": 2.468973806781189,
+      "grad_norm": 1.9053417444229126,
+      "learning_rate": 2.5569057104907668e-05,
+      "loss": 0.4948,
+      "step": 24650
+    },
+    {
+      "epoch": 2.4699754595081886,
+      "grad_norm": 1.6119698286056519,
+      "learning_rate": 2.555329636111424e-05,
+      "loss": 0.5038,
+      "step": 24660
+    },
+    {
+      "epoch": 2.4709771122351882,
+      "grad_norm": 2.262643814086914,
+      "learning_rate": 2.5537535397306916e-05,
+      "loss": 0.5284,
+      "step": 24670
+    },
+    {
+      "epoch": 2.471978764962188,
+      "grad_norm": 2.5736396312713623,
+      "learning_rate": 2.5521774219752926e-05,
+      "loss": 0.4789,
+      "step": 24680
+    },
+    {
+      "epoch": 2.472980417689187,
+      "grad_norm": 1.8912774324417114,
+      "learning_rate": 2.5506012834719577e-05,
+      "loss": 0.5158,
+      "step": 24690
+    },
+    {
+      "epoch": 2.4739820704161866,
+      "grad_norm": 2.2454535961151123,
+      "learning_rate": 2.5490251248474257e-05,
+      "loss": 0.4481,
+      "step": 24700
+    },
+    {
+      "epoch": 2.474983723143186,
+      "grad_norm": 1.9818633794784546,
+      "learning_rate": 2.547448946728443e-05,
+      "loss": 0.4488,
+      "step": 24710
+    },
+    {
+      "epoch": 2.4759853758701857,
+      "grad_norm": 1.9652526378631592,
+      "learning_rate": 2.545872749741765e-05,
+      "loss": 0.472,
+      "step": 24720
+    },
+    {
+      "epoch": 2.4769870285971853,
+      "grad_norm": 2.1900923252105713,
+      "learning_rate": 2.5442965345141533e-05,
+      "loss": 0.426,
+      "step": 24730
+    },
+    {
+      "epoch": 2.477988681324185,
+      "grad_norm": 1.747668743133545,
+      "learning_rate": 2.5427203016723782e-05,
+      "loss": 0.4571,
+      "step": 24740
+    },
+    {
+      "epoch": 2.4789903340511845,
+      "grad_norm": 2.722621202468872,
+      "learning_rate": 2.5411440518432145e-05,
+      "loss": 0.533,
+      "step": 24750
+    },
+    {
+      "epoch": 2.479991986778184,
+      "grad_norm": 1.841064453125,
+      "learning_rate": 2.5395677856534477e-05,
+      "loss": 0.3963,
+      "step": 24760
+    },
+    {
+      "epoch": 2.4809936395051837,
+      "grad_norm": 1.9657102823257446,
+      "learning_rate": 2.537991503729865e-05,
+      "loss": 0.4979,
+      "step": 24770
+    },
+    {
+      "epoch": 2.4819952922321833,
+      "grad_norm": 2.6763949394226074,
+      "learning_rate": 2.536415206699264e-05,
+      "loss": 0.5105,
+      "step": 24780
+    },
+    {
+      "epoch": 2.482996944959183,
+      "grad_norm": 2.6003458499908447,
+      "learning_rate": 2.5348388951884472e-05,
+      "loss": 0.4752,
+      "step": 24790
+    },
+    {
+      "epoch": 2.483998597686182,
+      "grad_norm": 3.0447092056274414,
+      "learning_rate": 2.5332625698242195e-05,
+      "loss": 0.5319,
+      "step": 24800
+    },
+    {
+      "epoch": 2.4850002504131816,
+      "grad_norm": 1.5755362510681152,
+      "learning_rate": 2.5316862312333978e-05,
+      "loss": 0.4119,
+      "step": 24810
+    },
+    {
+      "epoch": 2.486001903140181,
+      "grad_norm": 1.8942782878875732,
+      "learning_rate": 2.530109880042797e-05,
+      "loss": 0.5605,
+      "step": 24820
+    },
+    {
+      "epoch": 2.487003555867181,
+      "grad_norm": 2.6547932624816895,
+      "learning_rate": 2.5285335168792435e-05,
+      "loss": 0.4788,
+      "step": 24830
+    },
+    {
+      "epoch": 2.4880052085941804,
+      "grad_norm": 2.3786137104034424,
+      "learning_rate": 2.5269571423695638e-05,
+      "loss": 0.4729,
+      "step": 24840
+    },
+    {
+      "epoch": 2.48900686132118,
+      "grad_norm": 2.6195545196533203,
+      "learning_rate": 2.5253807571405913e-05,
+      "loss": 0.5661,
+      "step": 24850
+    },
+    {
+      "epoch": 2.4900085140481796,
+      "grad_norm": 2.390591621398926,
+      "learning_rate": 2.523804361819163e-05,
+      "loss": 0.4755,
+      "step": 24860
+    },
+    {
+      "epoch": 2.491010166775179,
+      "grad_norm": 2.008957624435425,
+      "learning_rate": 2.52222795703212e-05,
+      "loss": 0.5102,
+      "step": 24870
+    },
+    {
+      "epoch": 2.4920118195021788,
+      "grad_norm": 1.8178975582122803,
+      "learning_rate": 2.520651543406307e-05,
+      "loss": 0.5363,
+      "step": 24880
+    },
+    {
+      "epoch": 2.493013472229178,
+      "grad_norm": 1.8573118448257446,
+      "learning_rate": 2.5190751215685727e-05,
+      "loss": 0.4536,
+      "step": 24890
+    },
+    {
+      "epoch": 2.494015124956178,
+      "grad_norm": 1.995599627494812,
+      "learning_rate": 2.5174986921457677e-05,
+      "loss": 0.5196,
+      "step": 24900
+    },
+    {
+      "epoch": 2.495016777683177,
+      "grad_norm": 2.5662119388580322,
+      "learning_rate": 2.5159222557647473e-05,
+      "loss": 0.4686,
+      "step": 24910
+    },
+    {
+      "epoch": 2.4960184304101767,
+      "grad_norm": 2.2614381313323975,
+      "learning_rate": 2.514345813052369e-05,
+      "loss": 0.4819,
+      "step": 24920
+    },
+    {
+      "epoch": 2.4970200831371763,
+      "grad_norm": 2.1325461864471436,
+      "learning_rate": 2.5127693646354915e-05,
+      "loss": 0.414,
+      "step": 24930
+    },
+    {
+      "epoch": 2.498021735864176,
+      "grad_norm": 2.0967817306518555,
+      "learning_rate": 2.5111929111409795e-05,
+      "loss": 0.4791,
+      "step": 24940
+    },
+    {
+      "epoch": 2.4990233885911755,
+      "grad_norm": 2.1351726055145264,
+      "learning_rate": 2.509616453195694e-05,
+      "loss": 0.4869,
+      "step": 24950
+    },
+    {
+      "epoch": 2.500025041318175,
+      "grad_norm": 2.286760091781616,
+      "learning_rate": 2.5080399914265034e-05,
+      "loss": 0.5563,
+      "step": 24960
+    },
+    {
+      "epoch": 2.5010266940451746,
+      "grad_norm": 2.3411989212036133,
+      "learning_rate": 2.506463526460274e-05,
+      "loss": 0.5398,
+      "step": 24970
+    },
+    {
+      "epoch": 2.5020283467721742,
+      "grad_norm": 2.220675230026245,
+      "learning_rate": 2.504887058923875e-05,
+      "loss": 0.5249,
+      "step": 24980
+    },
+    {
+      "epoch": 2.503029999499174,
+      "grad_norm": 2.0709569454193115,
+      "learning_rate": 2.503310589444176e-05,
+      "loss": 0.5202,
+      "step": 24990
+    },
+    {
+      "epoch": 2.504031652226173,
+      "grad_norm": 2.3817501068115234,
+      "learning_rate": 2.501734118648047e-05,
+      "loss": 0.4466,
+      "step": 25000
+    },
+    {
+      "epoch": 2.505033304953173,
+      "grad_norm": 2.626307487487793,
+      "learning_rate": 2.50015764716236e-05,
+      "loss": 0.4598,
+      "step": 25010
+    },
+    {
+      "epoch": 2.506034957680172,
+      "grad_norm": 2.288289785385132,
+      "learning_rate": 2.498581175613986e-05,
+      "loss": 0.4345,
+      "step": 25020
+    },
+    {
+      "epoch": 2.5070366104071717,
+      "grad_norm": 2.8735311031341553,
+      "learning_rate": 2.497004704629796e-05,
+      "loss": 0.4907,
+      "step": 25030
+    },
+    {
+      "epoch": 2.5080382631341713,
+      "grad_norm": 2.262436866760254,
+      "learning_rate": 2.4954282348366613e-05,
+      "loss": 0.5129,
+      "step": 25040
+    },
+    {
+      "epoch": 2.509039915861171,
+      "grad_norm": 2.4624216556549072,
+      "learning_rate": 2.493851766861452e-05,
+      "loss": 0.43,
+      "step": 25050
+    },
+    {
+      "epoch": 2.5100415685881705,
+      "grad_norm": 2.532715082168579,
+      "learning_rate": 2.4922753013310397e-05,
+      "loss": 0.4568,
+      "step": 25060
+    },
+    {
+      "epoch": 2.51104322131517,
+      "grad_norm": 2.2034213542938232,
+      "learning_rate": 2.490698838872292e-05,
+      "loss": 0.4944,
+      "step": 25070
+    },
+    {
+      "epoch": 2.5120448740421697,
+      "grad_norm": 1.553830623626709,
+      "learning_rate": 2.489122380112076e-05,
+      "loss": 0.4666,
+      "step": 25080
+    },
+    {
+      "epoch": 2.5130465267691693,
+      "grad_norm": 2.5888707637786865,
+      "learning_rate": 2.48754592567726e-05,
+      "loss": 0.4994,
+      "step": 25090
+    },
+    {
+      "epoch": 2.514048179496169,
+      "grad_norm": 2.0383970737457275,
+      "learning_rate": 2.4859694761947074e-05,
+      "loss": 0.4831,
+      "step": 25100
+    },
+    {
+      "epoch": 2.515049832223168,
+      "grad_norm": 2.4339680671691895,
+      "learning_rate": 2.4843930322912795e-05,
+      "loss": 0.4834,
+      "step": 25110
+    },
+    {
+      "epoch": 2.5160514849501676,
+      "grad_norm": 2.0037882328033447,
+      "learning_rate": 2.482816594593839e-05,
+      "loss": 0.5053,
+      "step": 25120
+    },
+    {
+      "epoch": 2.517053137677167,
+      "grad_norm": 2.148824691772461,
+      "learning_rate": 2.481240163729243e-05,
+      "loss": 0.4823,
+      "step": 25130
+    },
+    {
+      "epoch": 2.518054790404167,
+      "grad_norm": 2.539546251296997,
+      "learning_rate": 2.4796637403243462e-05,
+      "loss": 0.5396,
+      "step": 25140
+    },
+    {
+      "epoch": 2.5190564431311664,
+      "grad_norm": 2.2066774368286133,
+      "learning_rate": 2.478087325006e-05,
+      "loss": 0.4696,
+      "step": 25150
+    },
+    {
+      "epoch": 2.520058095858166,
+      "grad_norm": 2.336604595184326,
+      "learning_rate": 2.476510918401056e-05,
+      "loss": 0.4451,
+      "step": 25160
+    },
+    {
+      "epoch": 2.5210597485851656,
+      "grad_norm": 2.245626211166382,
+      "learning_rate": 2.4749345211363574e-05,
+      "loss": 0.4704,
+      "step": 25170
+    },
+    {
+      "epoch": 2.522061401312165,
+      "grad_norm": 2.162360429763794,
+      "learning_rate": 2.473358133838746e-05,
+      "loss": 0.4521,
+      "step": 25180
+    },
+    {
+      "epoch": 2.5230630540391648,
+      "grad_norm": 1.6677041053771973,
+      "learning_rate": 2.4717817571350617e-05,
+      "loss": 0.4513,
+      "step": 25190
+    },
+    {
+      "epoch": 2.524064706766164,
+      "grad_norm": 1.8102422952651978,
+      "learning_rate": 2.470205391652137e-05,
+      "loss": 0.4525,
+      "step": 25200
+    },
+    {
+      "epoch": 2.525066359493164,
+      "grad_norm": 3.0435376167297363,
+      "learning_rate": 2.468629038016799e-05,
+      "loss": 0.5204,
+      "step": 25210
+    },
+    {
+      "epoch": 2.526068012220163,
+      "grad_norm": 2.122325897216797,
+      "learning_rate": 2.4670526968558755e-05,
+      "loss": 0.4684,
+      "step": 25220
+    },
+    {
+      "epoch": 2.5270696649471627,
+      "grad_norm": 2.4607677459716797,
+      "learning_rate": 2.465476368796184e-05,
+      "loss": 0.554,
+      "step": 25230
+    },
+    {
+      "epoch": 2.5280713176741623,
+      "grad_norm": 1.8819490671157837,
+      "learning_rate": 2.4639000544645403e-05,
+      "loss": 0.4483,
+      "step": 25240
+    },
+    {
+      "epoch": 2.529072970401162,
+      "grad_norm": 2.0729873180389404,
+      "learning_rate": 2.4623237544877514e-05,
+      "loss": 0.4738,
+      "step": 25250
+    },
+    {
+      "epoch": 2.5300746231281614,
+      "grad_norm": 1.9538005590438843,
+      "learning_rate": 2.4607474694926213e-05,
+      "loss": 0.4464,
+      "step": 25260
+    },
+    {
+      "epoch": 2.531076275855161,
+      "grad_norm": 2.4094536304473877,
+      "learning_rate": 2.459171200105948e-05,
+      "loss": 0.4938,
+      "step": 25270
+    },
+    {
+      "epoch": 2.5320779285821606,
+      "grad_norm": 2.4971518516540527,
+      "learning_rate": 2.4575949469545213e-05,
+      "loss": 0.5492,
+      "step": 25280
+    },
+    {
+      "epoch": 2.53307958130916,
+      "grad_norm": 2.3595423698425293,
+      "learning_rate": 2.4560187106651257e-05,
+      "loss": 0.4369,
+      "step": 25290
+    },
+    {
+      "epoch": 2.53408123403616,
+      "grad_norm": 1.8320460319519043,
+      "learning_rate": 2.4544424918645396e-05,
+      "loss": 0.4955,
+      "step": 25300
+    },
+    {
+      "epoch": 2.535082886763159,
+      "grad_norm": 2.384916305541992,
+      "learning_rate": 2.4528662911795342e-05,
+      "loss": 0.4669,
+      "step": 25310
+    },
+    {
+      "epoch": 2.536084539490159,
+      "grad_norm": 2.4493815898895264,
+      "learning_rate": 2.451290109236872e-05,
+      "loss": 0.4963,
+      "step": 25320
+    },
+    {
+      "epoch": 2.537086192217158,
+      "grad_norm": 1.902748942375183,
+      "learning_rate": 2.449713946663309e-05,
+      "loss": 0.4792,
+      "step": 25330
+    },
+    {
+      "epoch": 2.5380878449441577,
+      "grad_norm": 2.6030337810516357,
+      "learning_rate": 2.4481378040855955e-05,
+      "loss": 0.4986,
+      "step": 25340
+    },
+    {
+      "epoch": 2.5390894976711573,
+      "grad_norm": 3.346127510070801,
+      "learning_rate": 2.446561682130471e-05,
+      "loss": 0.4734,
+      "step": 25350
+    },
+    {
+      "epoch": 2.540091150398157,
+      "grad_norm": 2.2294840812683105,
+      "learning_rate": 2.4449855814246668e-05,
+      "loss": 0.443,
+      "step": 25360
+    },
+    {
+      "epoch": 2.5410928031251565,
+      "grad_norm": 1.9707425832748413,
+      "learning_rate": 2.4434095025949085e-05,
+      "loss": 0.5229,
+      "step": 25370
+    },
+    {
+      "epoch": 2.542094455852156,
+      "grad_norm": 2.2785727977752686,
+      "learning_rate": 2.4418334462679112e-05,
+      "loss": 0.4674,
+      "step": 25380
+    },
+    {
+      "epoch": 2.5430961085791557,
+      "grad_norm": 2.2725393772125244,
+      "learning_rate": 2.4402574130703795e-05,
+      "loss": 0.4049,
+      "step": 25390
+    },
+    {
+      "epoch": 2.5440977613061553,
+      "grad_norm": 1.964576244354248,
+      "learning_rate": 2.438681403629012e-05,
+      "loss": 0.5268,
+      "step": 25400
+    },
+    {
+      "epoch": 2.545099414033155,
+      "grad_norm": 2.273085594177246,
+      "learning_rate": 2.437105418570495e-05,
+      "loss": 0.4817,
+      "step": 25410
+    },
+    {
+      "epoch": 2.546101066760154,
+      "grad_norm": 2.092212438583374,
+      "learning_rate": 2.4355294585215084e-05,
+      "loss": 0.4439,
+      "step": 25420
+    },
+    {
+      "epoch": 2.547102719487154,
+      "grad_norm": 2.514829635620117,
+      "learning_rate": 2.433953524108718e-05,
+      "loss": 0.492,
+      "step": 25430
+    },
+    {
+      "epoch": 2.548104372214153,
+      "grad_norm": 2.0971784591674805,
+      "learning_rate": 2.4323776159587828e-05,
+      "loss": 0.4162,
+      "step": 25440
+    },
+    {
+      "epoch": 2.549106024941153,
+      "grad_norm": 2.2613589763641357,
+      "learning_rate": 2.4308017346983508e-05,
+      "loss": 0.5042,
+      "step": 25450
+    },
+    {
+      "epoch": 2.5501076776681524,
+      "grad_norm": 2.3998570442199707,
+      "learning_rate": 2.429225880954056e-05,
+      "loss": 0.4472,
+      "step": 25460
+    },
+    {
+      "epoch": 2.551109330395152,
+      "grad_norm": 2.4489238262176514,
+      "learning_rate": 2.4276500553525267e-05,
+      "loss": 0.4959,
+      "step": 25470
+    },
+    {
+      "epoch": 2.5521109831221516,
+      "grad_norm": 2.075310468673706,
+      "learning_rate": 2.4260742585203755e-05,
+      "loss": 0.5041,
+      "step": 25480
+    },
+    {
+      "epoch": 2.553112635849151,
+      "grad_norm": 2.417213201522827,
+      "learning_rate": 2.4244984910842076e-05,
+      "loss": 0.4575,
+      "step": 25490
+    },
+    {
+      "epoch": 2.5541142885761507,
+      "grad_norm": 2.008910655975342,
+      "learning_rate": 2.4229227536706126e-05,
+      "loss": 0.4455,
+      "step": 25500
+    },
+    {
+      "epoch": 2.5551159413031503,
+      "grad_norm": 2.7786033153533936,
+      "learning_rate": 2.42134704690617e-05,
+      "loss": 0.4653,
+      "step": 25510
+    },
+    {
+      "epoch": 2.55611759403015,
+      "grad_norm": 1.7164369821548462,
+      "learning_rate": 2.4197713714174477e-05,
+      "loss": 0.4826,
+      "step": 25520
+    },
+    {
+      "epoch": 2.557119246757149,
+      "grad_norm": 2.7447457313537598,
+      "learning_rate": 2.418195727831001e-05,
+      "loss": 0.4789,
+      "step": 25530
+    },
+    {
+      "epoch": 2.558120899484149,
+      "grad_norm": 1.7751262187957764,
+      "learning_rate": 2.4166201167733705e-05,
+      "loss": 0.5007,
+      "step": 25540
+    },
+    {
+      "epoch": 2.5591225522111483,
+      "grad_norm": 2.5949177742004395,
+      "learning_rate": 2.415044538871086e-05,
+      "loss": 0.475,
+      "step": 25550
+    },
+    {
+      "epoch": 2.560124204938148,
+      "grad_norm": 2.2382097244262695,
+      "learning_rate": 2.413468994750665e-05,
+      "loss": 0.4627,
+      "step": 25560
+    },
+    {
+      "epoch": 2.5611258576651474,
+      "grad_norm": 3.2741479873657227,
+      "learning_rate": 2.4118934850386076e-05,
+      "loss": 0.4902,
+      "step": 25570
+    },
+    {
+      "epoch": 2.562127510392147,
+      "grad_norm": 1.8402247428894043,
+      "learning_rate": 2.4103180103614042e-05,
+      "loss": 0.5062,
+      "step": 25580
+    },
+    {
+      "epoch": 2.5631291631191466,
+      "grad_norm": 3.088942766189575,
+      "learning_rate": 2.408742571345529e-05,
+      "loss": 0.452,
+      "step": 25590
+    },
+    {
+      "epoch": 2.564130815846146,
+      "grad_norm": 2.6764485836029053,
+      "learning_rate": 2.4071671686174444e-05,
+      "loss": 0.4863,
+      "step": 25600
+    },
+    {
+      "epoch": 2.565132468573146,
+      "grad_norm": 1.7545393705368042,
+      "learning_rate": 2.4055918028035944e-05,
+      "loss": 0.5092,
+      "step": 25610
+    },
+    {
+      "epoch": 2.5661341213001454,
+      "grad_norm": 2.4561171531677246,
+      "learning_rate": 2.404016474530412e-05,
+      "loss": 0.4584,
+      "step": 25620
+    },
+    {
+      "epoch": 2.567135774027145,
+      "grad_norm": 2.4661591053009033,
+      "learning_rate": 2.4024411844243136e-05,
+      "loss": 0.47,
+      "step": 25630
+    },
+    {
+      "epoch": 2.568137426754144,
+      "grad_norm": 2.1460041999816895,
+      "learning_rate": 2.4008659331117018e-05,
+      "loss": 0.5056,
+      "step": 25640
+    },
+    {
+      "epoch": 2.5691390794811437,
+      "grad_norm": 2.1262660026550293,
+      "learning_rate": 2.3992907212189615e-05,
+      "loss": 0.511,
+      "step": 25650
+    },
+    {
+      "epoch": 2.5701407322081433,
+      "grad_norm": 2.1861934661865234,
+      "learning_rate": 2.3977155493724624e-05,
+      "loss": 0.4645,
+      "step": 25660
+    },
+    {
+      "epoch": 2.571142384935143,
+      "grad_norm": 4.291914939880371,
+      "learning_rate": 2.3961404181985613e-05,
+      "loss": 0.5166,
+      "step": 25670
+    },
+    {
+      "epoch": 2.5721440376621425,
+      "grad_norm": 2.78710675239563,
+      "learning_rate": 2.394565328323595e-05,
+      "loss": 0.5688,
+      "step": 25680
+    },
+    {
+      "epoch": 2.573145690389142,
+      "grad_norm": 2.88307523727417,
+      "learning_rate": 2.3929902803738852e-05,
+      "loss": 0.4476,
+      "step": 25690
+    },
+    {
+      "epoch": 2.5741473431161417,
+      "grad_norm": 2.102853298187256,
+      "learning_rate": 2.391415274975738e-05,
+      "loss": 0.522,
+      "step": 25700
+    },
+    {
+      "epoch": 2.5751489958431413,
+      "grad_norm": 2.1062817573547363,
+      "learning_rate": 2.3898403127554423e-05,
+      "loss": 0.5218,
+      "step": 25710
+    },
+    {
+      "epoch": 2.576150648570141,
+      "grad_norm": 2.3642542362213135,
+      "learning_rate": 2.3882653943392678e-05,
+      "loss": 0.4886,
+      "step": 25720
+    },
+    {
+      "epoch": 2.57715230129714,
+      "grad_norm": 1.5191824436187744,
+      "learning_rate": 2.3866905203534693e-05,
+      "loss": 0.4415,
+      "step": 25730
+    },
+    {
+      "epoch": 2.57815395402414,
+      "grad_norm": 2.072801113128662,
+      "learning_rate": 2.385115691424283e-05,
+      "loss": 0.4764,
+      "step": 25740
+    },
+    {
+      "epoch": 2.579155606751139,
+      "grad_norm": 2.228182554244995,
+      "learning_rate": 2.3835409081779252e-05,
+      "loss": 0.5007,
+      "step": 25750
+    },
+    {
+      "epoch": 2.580157259478139,
+      "grad_norm": 1.7731989622116089,
+      "learning_rate": 2.3819661712405983e-05,
+      "loss": 0.4685,
+      "step": 25760
+    },
+    {
+      "epoch": 2.5811589122051384,
+      "grad_norm": 1.9319185018539429,
+      "learning_rate": 2.3803914812384823e-05,
+      "loss": 0.4991,
+      "step": 25770
+    },
+    {
+      "epoch": 2.582160564932138,
+      "grad_norm": 2.0118401050567627,
+      "learning_rate": 2.3788168387977424e-05,
+      "loss": 0.5158,
+      "step": 25780
+    },
+    {
+      "epoch": 2.5831622176591376,
+      "grad_norm": 1.7415335178375244,
+      "learning_rate": 2.377242244544519e-05,
+      "loss": 0.4717,
+      "step": 25790
+    },
+    {
+      "epoch": 2.584163870386137,
+      "grad_norm": 2.1140804290771484,
+      "learning_rate": 2.3756676991049402e-05,
+      "loss": 0.4628,
+      "step": 25800
+    },
+    {
+      "epoch": 2.5851655231131367,
+      "grad_norm": 1.9852862358093262,
+      "learning_rate": 2.3740932031051092e-05,
+      "loss": 0.5444,
+      "step": 25810
+    },
+    {
+      "epoch": 2.5861671758401363,
+      "grad_norm": 1.569001317024231,
+      "learning_rate": 2.3725187571711138e-05,
+      "loss": 0.478,
+      "step": 25820
+    },
+    {
+      "epoch": 2.587168828567136,
+      "grad_norm": 1.7986372709274292,
+      "learning_rate": 2.3709443619290184e-05,
+      "loss": 0.4702,
+      "step": 25830
+    },
+    {
+      "epoch": 2.588170481294135,
+      "grad_norm": 2.3105223178863525,
+      "learning_rate": 2.3693700180048688e-05,
+      "loss": 0.4755,
+      "step": 25840
+    },
+    {
+      "epoch": 2.589172134021135,
+      "grad_norm": 1.8925827741622925,
+      "learning_rate": 2.3677957260246917e-05,
+      "loss": 0.4492,
+      "step": 25850
+    },
+    {
+      "epoch": 2.5901737867481343,
+      "grad_norm": 2.353822946548462,
+      "learning_rate": 2.36622148661449e-05,
+      "loss": 0.5658,
+      "step": 25860
+    },
+    {
+      "epoch": 2.591175439475134,
+      "grad_norm": 2.317485809326172,
+      "learning_rate": 2.364647300400248e-05,
+      "loss": 0.4531,
+      "step": 25870
+    },
+    {
+      "epoch": 2.5921770922021334,
+      "grad_norm": 1.9144796133041382,
+      "learning_rate": 2.363073168007929e-05,
+      "loss": 0.4664,
+      "step": 25880
+    },
+    {
+      "epoch": 2.593178744929133,
+      "grad_norm": 2.278697967529297,
+      "learning_rate": 2.361499090063474e-05,
+      "loss": 0.4907,
+      "step": 25890
+    },
+    {
+      "epoch": 2.5941803976561326,
+      "grad_norm": 2.4261505603790283,
+      "learning_rate": 2.3599250671928014e-05,
+      "loss": 0.4243,
+      "step": 25900
+    },
+    {
+      "epoch": 2.595182050383132,
+      "grad_norm": 2.083456039428711,
+      "learning_rate": 2.3583511000218103e-05,
+      "loss": 0.4674,
+      "step": 25910
+    },
+    {
+      "epoch": 2.596183703110132,
+      "grad_norm": 2.447664737701416,
+      "learning_rate": 2.3567771891763763e-05,
+      "loss": 0.493,
+      "step": 25920
+    },
+    {
+      "epoch": 2.5971853558371314,
+      "grad_norm": 2.034817934036255,
+      "learning_rate": 2.3552033352823505e-05,
+      "loss": 0.4827,
+      "step": 25930
+    },
+    {
+      "epoch": 2.598187008564131,
+      "grad_norm": 2.444237470626831,
+      "learning_rate": 2.3536295389655654e-05,
+      "loss": 0.4507,
+      "step": 25940
+    },
+    {
+      "epoch": 2.59918866129113,
+      "grad_norm": 2.6944849491119385,
+      "learning_rate": 2.3520558008518273e-05,
+      "loss": 0.5128,
+      "step": 25950
+    },
+    {
+      "epoch": 2.60019031401813,
+      "grad_norm": 1.7680457830429077,
+      "learning_rate": 2.3504821215669228e-05,
+      "loss": 0.4819,
+      "step": 25960
+    },
+    {
+      "epoch": 2.6011919667451293,
+      "grad_norm": 2.6250157356262207,
+      "learning_rate": 2.3489085017366097e-05,
+      "loss": 0.5471,
+      "step": 25970
+    },
+    {
+      "epoch": 2.602193619472129,
+      "grad_norm": 1.8260471820831299,
+      "learning_rate": 2.3473349419866275e-05,
+      "loss": 0.4382,
+      "step": 25980
+    },
+    {
+      "epoch": 2.6031952721991285,
+      "grad_norm": 2.2221004962921143,
+      "learning_rate": 2.345761442942689e-05,
+      "loss": 0.4822,
+      "step": 25990
+    },
+    {
+      "epoch": 2.604196924926128,
+      "grad_norm": 1.932762622833252,
+      "learning_rate": 2.3441880052304842e-05,
+      "loss": 0.5148,
+      "step": 26000
+    },
+    {
+      "epoch": 2.6051985776531277,
+      "grad_norm": 2.2098388671875,
+      "learning_rate": 2.3426146294756774e-05,
+      "loss": 0.5548,
+      "step": 26010
+    },
+    {
+      "epoch": 2.6062002303801273,
+      "grad_norm": 2.6943869590759277,
+      "learning_rate": 2.3410413163039088e-05,
+      "loss": 0.4284,
+      "step": 26020
+    },
+    {
+      "epoch": 2.607201883107127,
+      "grad_norm": 1.8194222450256348,
+      "learning_rate": 2.3394680663407954e-05,
+      "loss": 0.4428,
+      "step": 26030
+    },
+    {
+      "epoch": 2.6082035358341265,
+      "grad_norm": 2.8261239528656006,
+      "learning_rate": 2.3378948802119254e-05,
+      "loss": 0.5257,
+      "step": 26040
+    },
+    {
+      "epoch": 2.609205188561126,
+      "grad_norm": 2.3305418491363525,
+      "learning_rate": 2.3363217585428644e-05,
+      "loss": 0.3921,
+      "step": 26050
+    },
+    {
+      "epoch": 2.610206841288125,
+      "grad_norm": 1.934168815612793,
+      "learning_rate": 2.3347487019591524e-05,
+      "loss": 0.4497,
+      "step": 26060
+    },
+    {
+      "epoch": 2.6112084940151252,
+      "grad_norm": 2.6037657260894775,
+      "learning_rate": 2.3331757110863028e-05,
+      "loss": 0.5347,
+      "step": 26070
+    },
+    {
+      "epoch": 2.6122101467421244,
+      "grad_norm": 2.024768829345703,
+      "learning_rate": 2.3316027865498017e-05,
+      "loss": 0.4866,
+      "step": 26080
+    },
+    {
+      "epoch": 2.613211799469124,
+      "grad_norm": 2.0269052982330322,
+      "learning_rate": 2.330029928975111e-05,
+      "loss": 0.4654,
+      "step": 26090
+    },
+    {
+      "epoch": 2.6142134521961236,
+      "grad_norm": 2.378782033920288,
+      "learning_rate": 2.3284571389876643e-05,
+      "loss": 0.5363,
+      "step": 26100
+    },
+    {
+      "epoch": 2.615215104923123,
+      "grad_norm": 2.4418351650238037,
+      "learning_rate": 2.3268844172128703e-05,
+      "loss": 0.4381,
+      "step": 26110
+    },
+    {
+      "epoch": 2.6162167576501227,
+      "grad_norm": 2.0797080993652344,
+      "learning_rate": 2.325311764276108e-05,
+      "loss": 0.4586,
+      "step": 26120
+    },
+    {
+      "epoch": 2.6172184103771223,
+      "grad_norm": 2.9233202934265137,
+      "learning_rate": 2.323739180802731e-05,
+      "loss": 0.5051,
+      "step": 26130
+    },
+    {
+      "epoch": 2.618220063104122,
+      "grad_norm": 1.9914222955703735,
+      "learning_rate": 2.3221666674180647e-05,
+      "loss": 0.4812,
+      "step": 26140
+    },
+    {
+      "epoch": 2.6192217158311215,
+      "grad_norm": 1.946094036102295,
+      "learning_rate": 2.320594224747405e-05,
+      "loss": 0.5021,
+      "step": 26150
+    },
+    {
+      "epoch": 2.620223368558121,
+      "grad_norm": 2.0537545680999756,
+      "learning_rate": 2.3190218534160233e-05,
+      "loss": 0.4171,
+      "step": 26160
+    },
+    {
+      "epoch": 2.6212250212851202,
+      "grad_norm": 2.107255697250366,
+      "learning_rate": 2.3174495540491588e-05,
+      "loss": 0.4683,
+      "step": 26170
+    },
+    {
+      "epoch": 2.62222667401212,
+      "grad_norm": 1.7646708488464355,
+      "learning_rate": 2.3158773272720254e-05,
+      "loss": 0.4256,
+      "step": 26180
+    },
+    {
+      "epoch": 2.6232283267391194,
+      "grad_norm": 1.94224214553833,
+      "learning_rate": 2.3143051737098054e-05,
+      "loss": 0.4832,
+      "step": 26190
+    },
+    {
+      "epoch": 2.624229979466119,
+      "grad_norm": 2.5952892303466797,
+      "learning_rate": 2.312733093987653e-05,
+      "loss": 0.4565,
+      "step": 26200
+    },
+    {
+      "epoch": 2.6252316321931186,
+      "grad_norm": 1.9447942972183228,
+      "learning_rate": 2.3111610887306946e-05,
+      "loss": 0.4124,
+      "step": 26210
+    },
+    {
+      "epoch": 2.626233284920118,
+      "grad_norm": 2.0823614597320557,
+      "learning_rate": 2.3095891585640246e-05,
+      "loss": 0.5009,
+      "step": 26220
+    },
+    {
+      "epoch": 2.627234937647118,
+      "grad_norm": 2.4931070804595947,
+      "learning_rate": 2.3080173041127074e-05,
+      "loss": 0.4633,
+      "step": 26230
+    },
+    {
+      "epoch": 2.6282365903741174,
+      "grad_norm": 2.088278293609619,
+      "learning_rate": 2.3064455260017803e-05,
+      "loss": 0.4155,
+      "step": 26240
+    },
+    {
+      "epoch": 2.629238243101117,
+      "grad_norm": 2.1908822059631348,
+      "learning_rate": 2.3048738248562478e-05,
+      "loss": 0.3935,
+      "step": 26250
+    },
+    {
+      "epoch": 2.630239895828116,
+      "grad_norm": 2.7435402870178223,
+      "learning_rate": 2.3033022013010834e-05,
+      "loss": 0.4813,
+      "step": 26260
+    },
+    {
+      "epoch": 2.631241548555116,
+      "grad_norm": 2.0322299003601074,
+      "learning_rate": 2.301730655961232e-05,
+      "loss": 0.4314,
+      "step": 26270
+    },
+    {
+      "epoch": 2.6322432012821153,
+      "grad_norm": 2.4347589015960693,
+      "learning_rate": 2.300159189461605e-05,
+      "loss": 0.4644,
+      "step": 26280
+    },
+    {
+      "epoch": 2.633244854009115,
+      "grad_norm": 2.0065510272979736,
+      "learning_rate": 2.298587802427085e-05,
+      "loss": 0.4897,
+      "step": 26290
+    },
+    {
+      "epoch": 2.6342465067361145,
+      "grad_norm": 2.5837438106536865,
+      "learning_rate": 2.2970164954825192e-05,
+      "loss": 0.469,
+      "step": 26300
+    },
+    {
+      "epoch": 2.635248159463114,
+      "grad_norm": 2.4908740520477295,
+      "learning_rate": 2.2954452692527277e-05,
+      "loss": 0.4813,
+      "step": 26310
+    },
+    {
+      "epoch": 2.6362498121901137,
+      "grad_norm": 2.1361074447631836,
+      "learning_rate": 2.293874124362495e-05,
+      "loss": 0.4876,
+      "step": 26320
+    },
+    {
+      "epoch": 2.6372514649171133,
+      "grad_norm": 2.353395462036133,
+      "learning_rate": 2.2923030614365735e-05,
+      "loss": 0.4718,
+      "step": 26330
+    },
+    {
+      "epoch": 2.638253117644113,
+      "grad_norm": 1.8883154392242432,
+      "learning_rate": 2.290732081099685e-05,
+      "loss": 0.5121,
+      "step": 26340
+    },
+    {
+      "epoch": 2.6392547703711124,
+      "grad_norm": 2.065039873123169,
+      "learning_rate": 2.289161183976517e-05,
+      "loss": 0.488,
+      "step": 26350
+    },
+    {
+      "epoch": 2.640256423098112,
+      "grad_norm": 2.287727117538452,
+      "learning_rate": 2.287590370691725e-05,
+      "loss": 0.4939,
+      "step": 26360
+    },
+    {
+      "epoch": 2.641258075825111,
+      "grad_norm": 2.5591814517974854,
+      "learning_rate": 2.2860196418699293e-05,
+      "loss": 0.4275,
+      "step": 26370
+    },
+    {
+      "epoch": 2.642259728552111,
+      "grad_norm": 2.074220657348633,
+      "learning_rate": 2.284448998135717e-05,
+      "loss": 0.4809,
+      "step": 26380
+    },
+    {
+      "epoch": 2.6432613812791104,
+      "grad_norm": 1.7956701517105103,
+      "learning_rate": 2.282878440113644e-05,
+      "loss": 0.4729,
+      "step": 26390
+    },
+    {
+      "epoch": 2.64426303400611,
+      "grad_norm": 2.273275375366211,
+      "learning_rate": 2.28130796842823e-05,
+      "loss": 0.5459,
+      "step": 26400
+    },
+    {
+      "epoch": 2.6452646867331095,
+      "grad_norm": 2.206364393234253,
+      "learning_rate": 2.2797375837039586e-05,
+      "loss": 0.5378,
+      "step": 26410
+    },
+    {
+      "epoch": 2.646266339460109,
+      "grad_norm": 2.165149688720703,
+      "learning_rate": 2.2781672865652824e-05,
+      "loss": 0.4299,
+      "step": 26420
+    },
+    {
+      "epoch": 2.6472679921871087,
+      "grad_norm": 2.0546000003814697,
+      "learning_rate": 2.276597077636618e-05,
+      "loss": 0.4466,
+      "step": 26430
+    },
+    {
+      "epoch": 2.6482696449141083,
+      "grad_norm": 1.8035660982131958,
+      "learning_rate": 2.2750269575423447e-05,
+      "loss": 0.4712,
+      "step": 26440
+    },
+    {
+      "epoch": 2.649271297641108,
+      "grad_norm": 2.0470290184020996,
+      "learning_rate": 2.2734569269068092e-05,
+      "loss": 0.4758,
+      "step": 26450
+    },
+    {
+      "epoch": 2.6502729503681075,
+      "grad_norm": 2.314500570297241,
+      "learning_rate": 2.271886986354322e-05,
+      "loss": 0.5292,
+      "step": 26460
+    },
+    {
+      "epoch": 2.651274603095107,
+      "grad_norm": 1.8078278303146362,
+      "learning_rate": 2.2703171365091577e-05,
+      "loss": 0.4468,
+      "step": 26470
+    },
+    {
+      "epoch": 2.6522762558221062,
+      "grad_norm": 2.535794734954834,
+      "learning_rate": 2.2687473779955526e-05,
+      "loss": 0.4847,
+      "step": 26480
+    },
+    {
+      "epoch": 2.6532779085491063,
+      "grad_norm": 2.081583023071289,
+      "learning_rate": 2.267177711437711e-05,
+      "loss": 0.4684,
+      "step": 26490
+    },
+    {
+      "epoch": 2.6542795612761054,
+      "grad_norm": 2.319704294204712,
+      "learning_rate": 2.2656081374597976e-05,
+      "loss": 0.4822,
+      "step": 26500
+    },
+    {
+      "epoch": 2.655281214003105,
+      "grad_norm": 2.5577144622802734,
+      "learning_rate": 2.2640386566859398e-05,
+      "loss": 0.4674,
+      "step": 26510
+    },
+    {
+      "epoch": 2.6562828667301046,
+      "grad_norm": 1.8538458347320557,
+      "learning_rate": 2.2624692697402306e-05,
+      "loss": 0.4069,
+      "step": 26520
+    },
+    {
+      "epoch": 2.657284519457104,
+      "grad_norm": 2.406271457672119,
+      "learning_rate": 2.2608999772467228e-05,
+      "loss": 0.5484,
+      "step": 26530
+    },
+    {
+      "epoch": 2.658286172184104,
+      "grad_norm": 2.2202024459838867,
+      "learning_rate": 2.2593307798294355e-05,
+      "loss": 0.3974,
+      "step": 26540
+    },
+    {
+      "epoch": 2.6592878249111034,
+      "grad_norm": 1.886047601699829,
+      "learning_rate": 2.257761678112345e-05,
+      "loss": 0.4842,
+      "step": 26550
+    },
+    {
+      "epoch": 2.660289477638103,
+      "grad_norm": 2.4025936126708984,
+      "learning_rate": 2.256192672719393e-05,
+      "loss": 0.466,
+      "step": 26560
+    },
+    {
+      "epoch": 2.6612911303651026,
+      "grad_norm": 2.0745341777801514,
+      "learning_rate": 2.254623764274482e-05,
+      "loss": 0.4589,
+      "step": 26570
+    },
+    {
+      "epoch": 2.662292783092102,
+      "grad_norm": 1.9398627281188965,
+      "learning_rate": 2.2530549534014772e-05,
+      "loss": 0.507,
+      "step": 26580
+    },
+    {
+      "epoch": 2.6632944358191013,
+      "grad_norm": 2.9563961029052734,
+      "learning_rate": 2.2514862407242012e-05,
+      "loss": 0.4202,
+      "step": 26590
+    },
+    {
+      "epoch": 2.6642960885461013,
+      "grad_norm": 2.6384031772613525,
+      "learning_rate": 2.249917626866442e-05,
+      "loss": 0.4569,
+      "step": 26600
+    },
+    {
+      "epoch": 2.6652977412731005,
+      "grad_norm": 2.049692153930664,
+      "learning_rate": 2.2483491124519465e-05,
+      "loss": 0.4576,
+      "step": 26610
+    },
+    {
+      "epoch": 2.6662993940001,
+      "grad_norm": 2.3128302097320557,
+      "learning_rate": 2.2467806981044207e-05,
+      "loss": 0.4881,
+      "step": 26620
+    },
+    {
+      "epoch": 2.6673010467270997,
+      "grad_norm": 1.87588369846344,
+      "learning_rate": 2.2452123844475322e-05,
+      "loss": 0.4082,
+      "step": 26630
+    },
+    {
+      "epoch": 2.6683026994540993,
+      "grad_norm": 2.448084592819214,
+      "learning_rate": 2.2436441721049095e-05,
+      "loss": 0.4678,
+      "step": 26640
+    },
+    {
+      "epoch": 2.669304352181099,
+      "grad_norm": 2.088381290435791,
+      "learning_rate": 2.2420760617001395e-05,
+      "loss": 0.4423,
+      "step": 26650
+    },
+    {
+      "epoch": 2.6703060049080984,
+      "grad_norm": 2.1151883602142334,
+      "learning_rate": 2.2405080538567673e-05,
+      "loss": 0.4899,
+      "step": 26660
+    },
+    {
+      "epoch": 2.671307657635098,
+      "grad_norm": 2.232419729232788,
+      "learning_rate": 2.238940149198301e-05,
+      "loss": 0.5182,
+      "step": 26670
+    },
+    {
+      "epoch": 2.6723093103620976,
+      "grad_norm": 1.9108939170837402,
+      "learning_rate": 2.237372348348203e-05,
+      "loss": 0.4394,
+      "step": 26680
+    },
+    {
+      "epoch": 2.673310963089097,
+      "grad_norm": 2.0978970527648926,
+      "learning_rate": 2.2358046519298997e-05,
+      "loss": 0.5118,
+      "step": 26690
+    },
+    {
+      "epoch": 2.6743126158160964,
+      "grad_norm": 2.155327558517456,
+      "learning_rate": 2.234237060566771e-05,
+      "loss": 0.5335,
+      "step": 26700
+    },
+    {
+      "epoch": 2.675314268543096,
+      "grad_norm": 2.3690075874328613,
+      "learning_rate": 2.2326695748821565e-05,
+      "loss": 0.4261,
+      "step": 26710
+    },
+    {
+      "epoch": 2.6763159212700955,
+      "grad_norm": 1.9474866390228271,
+      "learning_rate": 2.2311021954993572e-05,
+      "loss": 0.5525,
+      "step": 26720
+    },
+    {
+      "epoch": 2.677317573997095,
+      "grad_norm": 2.135596990585327,
+      "learning_rate": 2.2295349230416266e-05,
+      "loss": 0.5035,
+      "step": 26730
+    },
+    {
+      "epoch": 2.6783192267240947,
+      "grad_norm": 2.132472276687622,
+      "learning_rate": 2.227967758132178e-05,
+      "loss": 0.5243,
+      "step": 26740
+    },
+    {
+      "epoch": 2.6793208794510943,
+      "grad_norm": 1.8780027627944946,
+      "learning_rate": 2.226400701394184e-05,
+      "loss": 0.4771,
+      "step": 26750
+    },
+    {
+      "epoch": 2.680322532178094,
+      "grad_norm": 1.9652684926986694,
+      "learning_rate": 2.224833753450771e-05,
+      "loss": 0.4811,
+      "step": 26760
+    },
+    {
+      "epoch": 2.6813241849050935,
+      "grad_norm": 2.099169969558716,
+      "learning_rate": 2.2232669149250227e-05,
+      "loss": 0.4777,
+      "step": 26770
+    },
+    {
+      "epoch": 2.682325837632093,
+      "grad_norm": 2.4007065296173096,
+      "learning_rate": 2.221700186439981e-05,
+      "loss": 0.4176,
+      "step": 26780
+    },
+    {
+      "epoch": 2.6833274903590922,
+      "grad_norm": 2.2021305561065674,
+      "learning_rate": 2.2201335686186435e-05,
+      "loss": 0.5119,
+      "step": 26790
+    },
+    {
+      "epoch": 2.6843291430860923,
+      "grad_norm": 2.159282922744751,
+      "learning_rate": 2.218567062083962e-05,
+      "loss": 0.4317,
+      "step": 26800
+    },
+    {
+      "epoch": 2.6853307958130914,
+      "grad_norm": 2.309873342514038,
+      "learning_rate": 2.217000667458845e-05,
+      "loss": 0.5505,
+      "step": 26810
+    },
+    {
+      "epoch": 2.686332448540091,
+      "grad_norm": 2.6905899047851562,
+      "learning_rate": 2.2154343853661586e-05,
+      "loss": 0.4556,
+      "step": 26820
+    },
+    {
+      "epoch": 2.6873341012670906,
+      "grad_norm": 1.9428925514221191,
+      "learning_rate": 2.2138682164287217e-05,
+      "loss": 0.4622,
+      "step": 26830
+    },
+    {
+      "epoch": 2.68833575399409,
+      "grad_norm": 2.6776959896087646,
+      "learning_rate": 2.212302161269308e-05,
+      "loss": 0.5363,
+      "step": 26840
+    },
+    {
+      "epoch": 2.68933740672109,
+      "grad_norm": 2.054520606994629,
+      "learning_rate": 2.2107362205106477e-05,
+      "loss": 0.4442,
+      "step": 26850
+    },
+    {
+      "epoch": 2.6903390594480894,
+      "grad_norm": 1.8479695320129395,
+      "learning_rate": 2.209170394775424e-05,
+      "loss": 0.4855,
+      "step": 26860
+    },
+    {
+      "epoch": 2.691340712175089,
+      "grad_norm": 2.017530679702759,
+      "learning_rate": 2.2076046846862767e-05,
+      "loss": 0.4774,
+      "step": 26870
+    },
+    {
+      "epoch": 2.6923423649020886,
+      "grad_norm": 2.2654614448547363,
+      "learning_rate": 2.2060390908657964e-05,
+      "loss": 0.5063,
+      "step": 26880
+    },
+    {
+      "epoch": 2.693344017629088,
+      "grad_norm": 2.8104443550109863,
+      "learning_rate": 2.2044736139365286e-05,
+      "loss": 0.4963,
+      "step": 26890
+    },
+    {
+      "epoch": 2.6943456703560873,
+      "grad_norm": 2.534173011779785,
+      "learning_rate": 2.2029082545209748e-05,
+      "loss": 0.5132,
+      "step": 26900
+    },
+    {
+      "epoch": 2.6953473230830873,
+      "grad_norm": 2.722679376602173,
+      "learning_rate": 2.2013430132415858e-05,
+      "loss": 0.4832,
+      "step": 26910
+    },
+    {
+      "epoch": 2.6963489758100865,
+      "grad_norm": 1.6218446493148804,
+      "learning_rate": 2.1997778907207673e-05,
+      "loss": 0.4437,
+      "step": 26920
+    },
+    {
+      "epoch": 2.697350628537086,
+      "grad_norm": 2.0946335792541504,
+      "learning_rate": 2.198212887580879e-05,
+      "loss": 0.4733,
+      "step": 26930
+    },
+    {
+      "epoch": 2.6983522812640857,
+      "grad_norm": 2.202590227127075,
+      "learning_rate": 2.196648004444232e-05,
+      "loss": 0.4577,
+      "step": 26940
+    },
+    {
+      "epoch": 2.6993539339910853,
+      "grad_norm": 2.196267604827881,
+      "learning_rate": 2.195083241933088e-05,
+      "loss": 0.5106,
+      "step": 26950
+    },
+    {
+      "epoch": 2.700355586718085,
+      "grad_norm": 2.427388906478882,
+      "learning_rate": 2.1935186006696634e-05,
+      "loss": 0.4277,
+      "step": 26960
+    },
+    {
+      "epoch": 2.7013572394450844,
+      "grad_norm": 2.214708089828491,
+      "learning_rate": 2.191954081276126e-05,
+      "loss": 0.4305,
+      "step": 26970
+    },
+    {
+      "epoch": 2.702358892172084,
+      "grad_norm": 2.6666109561920166,
+      "learning_rate": 2.1903896843745932e-05,
+      "loss": 0.4622,
+      "step": 26980
+    },
+    {
+      "epoch": 2.7033605448990836,
+      "grad_norm": 2.6057868003845215,
+      "learning_rate": 2.188825410587135e-05,
+      "loss": 0.4764,
+      "step": 26990
+    },
+    {
+      "epoch": 2.704362197626083,
+      "grad_norm": 2.430042266845703,
+      "learning_rate": 2.1872612605357735e-05,
+      "loss": 0.4984,
+      "step": 27000
+    },
+    {
+      "epoch": 2.7053638503530824,
+      "grad_norm": 2.1184325218200684,
+      "learning_rate": 2.1856972348424805e-05,
+      "loss": 0.4949,
+      "step": 27010
+    },
+    {
+      "epoch": 2.7063655030800824,
+      "grad_norm": 1.98492431640625,
+      "learning_rate": 2.184133334129176e-05,
+      "loss": 0.5301,
+      "step": 27020
+    },
+    {
+      "epoch": 2.7073671558070815,
+      "grad_norm": 1.8999614715576172,
+      "learning_rate": 2.1825695590177355e-05,
+      "loss": 0.4537,
+      "step": 27030
+    },
+    {
+      "epoch": 2.708368808534081,
+      "grad_norm": 3.519404649734497,
+      "learning_rate": 2.1810059101299802e-05,
+      "loss": 0.4338,
+      "step": 27040
+    },
+    {
+      "epoch": 2.7093704612610807,
+      "grad_norm": 2.756481409072876,
+      "learning_rate": 2.1794423880876842e-05,
+      "loss": 0.472,
+      "step": 27050
+    },
+    {
+      "epoch": 2.7103721139880803,
+      "grad_norm": 2.2912352085113525,
+      "learning_rate": 2.1778789935125674e-05,
+      "loss": 0.4935,
+      "step": 27060
+    },
+    {
+      "epoch": 2.71137376671508,
+      "grad_norm": 1.8153929710388184,
+      "learning_rate": 2.1763157270263017e-05,
+      "loss": 0.4422,
+      "step": 27070
+    },
+    {
+      "epoch": 2.7123754194420795,
+      "grad_norm": 2.11661434173584,
+      "learning_rate": 2.1747525892505094e-05,
+      "loss": 0.484,
+      "step": 27080
+    },
+    {
+      "epoch": 2.713377072169079,
+      "grad_norm": 2.561905860900879,
+      "learning_rate": 2.1731895808067583e-05,
+      "loss": 0.4715,
+      "step": 27090
+    },
+    {
+      "epoch": 2.7143787248960787,
+      "grad_norm": 2.2325327396392822,
+      "learning_rate": 2.171626702316565e-05,
+      "loss": 0.4948,
+      "step": 27100
+    },
+    {
+      "epoch": 2.7153803776230783,
+      "grad_norm": 2.3941023349761963,
+      "learning_rate": 2.170063954401398e-05,
+      "loss": 0.5549,
+      "step": 27110
+    },
+    {
+      "epoch": 2.7163820303500774,
+      "grad_norm": 1.9785363674163818,
+      "learning_rate": 2.1685013376826715e-05,
+      "loss": 0.4756,
+      "step": 27120
+    },
+    {
+      "epoch": 2.7173836830770774,
+      "grad_norm": 2.551440715789795,
+      "learning_rate": 2.166938852781746e-05,
+      "loss": 0.5079,
+      "step": 27130
+    },
+    {
+      "epoch": 2.7183853358040766,
+      "grad_norm": 2.1827147006988525,
+      "learning_rate": 2.1653765003199315e-05,
+      "loss": 0.4197,
+      "step": 27140
+    },
+    {
+      "epoch": 2.719386988531076,
+      "grad_norm": 1.9968968629837036,
+      "learning_rate": 2.163814280918486e-05,
+      "loss": 0.4311,
+      "step": 27150
+    },
+    {
+      "epoch": 2.720388641258076,
+      "grad_norm": 1.8684368133544922,
+      "learning_rate": 2.162252195198613e-05,
+      "loss": 0.416,
+      "step": 27160
+    },
+    {
+      "epoch": 2.7213902939850754,
+      "grad_norm": 2.0815813541412354,
+      "learning_rate": 2.160690243781463e-05,
+      "loss": 0.4611,
+      "step": 27170
+    },
+    {
+      "epoch": 2.722391946712075,
+      "grad_norm": 2.146352767944336,
+      "learning_rate": 2.159128427288134e-05,
+      "loss": 0.4603,
+      "step": 27180
+    },
+    {
+      "epoch": 2.7233935994390746,
+      "grad_norm": 2.8626322746276855,
+      "learning_rate": 2.157566746339671e-05,
+      "loss": 0.484,
+      "step": 27190
+    },
+    {
+      "epoch": 2.724395252166074,
+      "grad_norm": 1.959716796875,
+      "learning_rate": 2.1560052015570613e-05,
+      "loss": 0.4973,
+      "step": 27200
+    },
+    {
+      "epoch": 2.7253969048930737,
+      "grad_norm": 2.7703990936279297,
+      "learning_rate": 2.154443793561243e-05,
+      "loss": 0.4741,
+      "step": 27210
+    },
+    {
+      "epoch": 2.7263985576200733,
+      "grad_norm": 2.6336097717285156,
+      "learning_rate": 2.1528825229730966e-05,
+      "loss": 0.4971,
+      "step": 27220
+    },
+    {
+      "epoch": 2.7274002103470725,
+      "grad_norm": 2.6301822662353516,
+      "learning_rate": 2.1513213904134496e-05,
+      "loss": 0.4954,
+      "step": 27230
+    },
+    {
+      "epoch": 2.728401863074072,
+      "grad_norm": 2.201345920562744,
+      "learning_rate": 2.1497603965030736e-05,
+      "loss": 0.4635,
+      "step": 27240
+    },
+    {
+      "epoch": 2.7294035158010717,
+      "grad_norm": 2.9680652618408203,
+      "learning_rate": 2.1481995418626842e-05,
+      "loss": 0.4896,
+      "step": 27250
+    },
+    {
+      "epoch": 2.7304051685280712,
+      "grad_norm": 1.5185983180999756,
+      "learning_rate": 2.1466388271129458e-05,
+      "loss": 0.4887,
+      "step": 27260
+    },
+    {
+      "epoch": 2.731406821255071,
+      "grad_norm": 2.366058111190796,
+      "learning_rate": 2.1450782528744605e-05,
+      "loss": 0.4411,
+      "step": 27270
+    },
+    {
+      "epoch": 2.7324084739820704,
+      "grad_norm": 2.3551275730133057,
+      "learning_rate": 2.143517819767781e-05,
+      "loss": 0.3853,
+      "step": 27280
+    },
+    {
+      "epoch": 2.73341012670907,
+      "grad_norm": 1.7083995342254639,
+      "learning_rate": 2.1419575284133993e-05,
+      "loss": 0.4732,
+      "step": 27290
+    },
+    {
+      "epoch": 2.7344117794360696,
+      "grad_norm": 2.4988021850585938,
+      "learning_rate": 2.1403973794317543e-05,
+      "loss": 0.546,
+      "step": 27300
+    },
+    {
+      "epoch": 2.735413432163069,
+      "grad_norm": 2.1923186779022217,
+      "learning_rate": 2.1388373734432256e-05,
+      "loss": 0.4854,
+      "step": 27310
+    },
+    {
+      "epoch": 2.7364150848900683,
+      "grad_norm": 2.622400999069214,
+      "learning_rate": 2.1372775110681376e-05,
+      "loss": 0.4919,
+      "step": 27320
+    },
+    {
+      "epoch": 2.7374167376170684,
+      "grad_norm": 2.5335519313812256,
+      "learning_rate": 2.1357177929267574e-05,
+      "loss": 0.5083,
+      "step": 27330
+    },
+    {
+      "epoch": 2.7384183903440675,
+      "grad_norm": 2.4166882038116455,
+      "learning_rate": 2.134158219639295e-05,
+      "loss": 0.4475,
+      "step": 27340
+    },
+    {
+      "epoch": 2.739420043071067,
+      "grad_norm": 2.1072611808776855,
+      "learning_rate": 2.1325987918259006e-05,
+      "loss": 0.498,
+      "step": 27350
+    },
+    {
+      "epoch": 2.7404216957980667,
+      "grad_norm": 2.3170204162597656,
+      "learning_rate": 2.1310395101066692e-05,
+      "loss": 0.4514,
+      "step": 27360
+    },
+    {
+      "epoch": 2.7414233485250663,
+      "grad_norm": 2.4249660968780518,
+      "learning_rate": 2.129480375101638e-05,
+      "loss": 0.4603,
+      "step": 27370
+    },
+    {
+      "epoch": 2.742425001252066,
+      "grad_norm": 1.9688012599945068,
+      "learning_rate": 2.1279213874307818e-05,
+      "loss": 0.4447,
+      "step": 27380
+    },
+    {
+      "epoch": 2.7434266539790655,
+      "grad_norm": 2.5289130210876465,
+      "learning_rate": 2.126362547714022e-05,
+      "loss": 0.4885,
+      "step": 27390
+    },
+    {
+      "epoch": 2.744428306706065,
+      "grad_norm": 1.826878309249878,
+      "learning_rate": 2.1248038565712175e-05,
+      "loss": 0.4917,
+      "step": 27400
+    },
+    {
+      "epoch": 2.7454299594330647,
+      "grad_norm": 2.649292230606079,
+      "learning_rate": 2.123245314622171e-05,
+      "loss": 0.4552,
+      "step": 27410
+    },
+    {
+      "epoch": 2.7464316121600643,
+      "grad_norm": 3.107866048812866,
+      "learning_rate": 2.1216869224866226e-05,
+      "loss": 0.4791,
+      "step": 27420
+    },
+    {
+      "epoch": 2.7474332648870634,
+      "grad_norm": 2.3746352195739746,
+      "learning_rate": 2.120128680784255e-05,
+      "loss": 0.4838,
+      "step": 27430
+    },
+    {
+      "epoch": 2.7484349176140634,
+      "grad_norm": 1.829545497894287,
+      "learning_rate": 2.118570590134691e-05,
+      "loss": 0.4865,
+      "step": 27440
+    },
+    {
+      "epoch": 2.7494365703410626,
+      "grad_norm": 1.9630305767059326,
+      "learning_rate": 2.1170126511574927e-05,
+      "loss": 0.4979,
+      "step": 27450
+    },
+    {
+      "epoch": 2.750438223068062,
+      "grad_norm": 2.2091236114501953,
+      "learning_rate": 2.115454864472162e-05,
+      "loss": 0.5135,
+      "step": 27460
+    },
+    {
+      "epoch": 2.7514398757950618,
+      "grad_norm": 2.330111265182495,
+      "learning_rate": 2.1138972306981398e-05,
+      "loss": 0.4824,
+      "step": 27470
+    },
+    {
+      "epoch": 2.7524415285220614,
+      "grad_norm": 2.7530877590179443,
+      "learning_rate": 2.1123397504548087e-05,
+      "loss": 0.4567,
+      "step": 27480
+    },
+    {
+      "epoch": 2.753443181249061,
+      "grad_norm": 2.259148597717285,
+      "learning_rate": 2.1107824243614865e-05,
+      "loss": 0.4989,
+      "step": 27490
+    },
+    {
+      "epoch": 2.7544448339760605,
+      "grad_norm": 1.6459332704544067,
+      "learning_rate": 2.109225253037431e-05,
+      "loss": 0.4808,
+      "step": 27500
+    },
+    {
+      "epoch": 2.75544648670306,
+      "grad_norm": 2.2297933101654053,
+      "learning_rate": 2.107668237101841e-05,
+      "loss": 0.4954,
+      "step": 27510
+    },
+    {
+      "epoch": 2.7564481394300597,
+      "grad_norm": 2.1366751194000244,
+      "learning_rate": 2.106111377173851e-05,
+      "loss": 0.4906,
+      "step": 27520
+    },
+    {
+      "epoch": 2.7574497921570593,
+      "grad_norm": 2.337979555130005,
+      "learning_rate": 2.104554673872532e-05,
+      "loss": 0.5064,
+      "step": 27530
+    },
+    {
+      "epoch": 2.7584514448840585,
+      "grad_norm": 2.3288912773132324,
+      "learning_rate": 2.102998127816897e-05,
+      "loss": 0.4822,
+      "step": 27540
+    },
+    {
+      "epoch": 2.7594530976110585,
+      "grad_norm": 2.4183363914489746,
+      "learning_rate": 2.1014417396258936e-05,
+      "loss": 0.5117,
+      "step": 27550
+    },
+    {
+      "epoch": 2.7604547503380577,
+      "grad_norm": 2.034332752227783,
+      "learning_rate": 2.0998855099184058e-05,
+      "loss": 0.4364,
+      "step": 27560
+    },
+    {
+      "epoch": 2.7614564030650572,
+      "grad_norm": 2.2769012451171875,
+      "learning_rate": 2.0983294393132575e-05,
+      "loss": 0.4577,
+      "step": 27570
+    },
+    {
+      "epoch": 2.762458055792057,
+      "grad_norm": 2.3711085319519043,
+      "learning_rate": 2.0967735284292065e-05,
+      "loss": 0.5053,
+      "step": 27580
+    },
+    {
+      "epoch": 2.7634597085190564,
+      "grad_norm": 1.7482993602752686,
+      "learning_rate": 2.09521777788495e-05,
+      "loss": 0.4199,
+      "step": 27590
+    },
+    {
+      "epoch": 2.764461361246056,
+      "grad_norm": 1.9056187868118286,
+      "learning_rate": 2.0936621882991186e-05,
+      "loss": 0.4214,
+      "step": 27600
+    },
+    {
+      "epoch": 2.7654630139730556,
+      "grad_norm": 1.977642297744751,
+      "learning_rate": 2.0921067602902804e-05,
+      "loss": 0.4373,
+      "step": 27610
+    },
+    {
+      "epoch": 2.766464666700055,
+      "grad_norm": 2.014859199523926,
+      "learning_rate": 2.0905514944769382e-05,
+      "loss": 0.4806,
+      "step": 27620
+    },
+    {
+      "epoch": 2.767466319427055,
+      "grad_norm": 3.8795015811920166,
+      "learning_rate": 2.0889963914775333e-05,
+      "loss": 0.4404,
+      "step": 27630
+    },
+    {
+      "epoch": 2.7684679721540544,
+      "grad_norm": 2.2615413665771484,
+      "learning_rate": 2.0874414519104375e-05,
+      "loss": 0.5025,
+      "step": 27640
+    },
+    {
+      "epoch": 2.7694696248810535,
+      "grad_norm": 3.039480209350586,
+      "learning_rate": 2.085886676393961e-05,
+      "loss": 0.4841,
+      "step": 27650
+    },
+    {
+      "epoch": 2.7704712776080536,
+      "grad_norm": 2.475024700164795,
+      "learning_rate": 2.084332065546349e-05,
+      "loss": 0.5154,
+      "step": 27660
+    },
+    {
+      "epoch": 2.7714729303350527,
+      "grad_norm": 2.402113437652588,
+      "learning_rate": 2.082777619985778e-05,
+      "loss": 0.4713,
+      "step": 27670
+    },
+    {
+      "epoch": 2.7724745830620523,
+      "grad_norm": 1.7097605466842651,
+      "learning_rate": 2.0812233403303625e-05,
+      "loss": 0.4631,
+      "step": 27680
+    },
+    {
+      "epoch": 2.773476235789052,
+      "grad_norm": 2.578178644180298,
+      "learning_rate": 2.079669227198149e-05,
+      "loss": 0.4864,
+      "step": 27690
+    },
+    {
+      "epoch": 2.7744778885160515,
+      "grad_norm": 2.374117374420166,
+      "learning_rate": 2.0781152812071187e-05,
+      "loss": 0.4303,
+      "step": 27700
+    },
+    {
+      "epoch": 2.775479541243051,
+      "grad_norm": 2.412297010421753,
+      "learning_rate": 2.0765615029751843e-05,
+      "loss": 0.5018,
+      "step": 27710
+    },
+    {
+      "epoch": 2.7764811939700507,
+      "grad_norm": 2.166706085205078,
+      "learning_rate": 2.075007893120195e-05,
+      "loss": 0.4195,
+      "step": 27720
+    },
+    {
+      "epoch": 2.7774828466970503,
+      "grad_norm": 2.3247227668762207,
+      "learning_rate": 2.0734544522599298e-05,
+      "loss": 0.4862,
+      "step": 27730
+    },
+    {
+      "epoch": 2.77848449942405,
+      "grad_norm": 2.5825328826904297,
+      "learning_rate": 2.0719011810121043e-05,
+      "loss": 0.5212,
+      "step": 27740
+    },
+    {
+      "epoch": 2.7794861521510494,
+      "grad_norm": 3.512326955795288,
+      "learning_rate": 2.070348079994363e-05,
+      "loss": 0.5204,
+      "step": 27750
+    },
+    {
+      "epoch": 2.7804878048780486,
+      "grad_norm": 2.1720283031463623,
+      "learning_rate": 2.0687951498242833e-05,
+      "loss": 0.4784,
+      "step": 27760
+    },
+    {
+      "epoch": 2.781489457605048,
+      "grad_norm": 2.6881589889526367,
+      "learning_rate": 2.067242391119378e-05,
+      "loss": 0.4718,
+      "step": 27770
+    },
+    {
+      "epoch": 2.7824911103320478,
+      "grad_norm": 1.799422025680542,
+      "learning_rate": 2.0656898044970866e-05,
+      "loss": 0.4398,
+      "step": 27780
+    },
+    {
+      "epoch": 2.7834927630590474,
+      "grad_norm": 2.7559547424316406,
+      "learning_rate": 2.064137390574784e-05,
+      "loss": 0.5155,
+      "step": 27790
+    },
+    {
+      "epoch": 2.784494415786047,
+      "grad_norm": 2.2045767307281494,
+      "learning_rate": 2.0625851499697747e-05,
+      "loss": 0.5234,
+      "step": 27800
+    },
+    {
+      "epoch": 2.7854960685130465,
+      "grad_norm": 3.2407376766204834,
+      "learning_rate": 2.0610330832992966e-05,
+      "loss": 0.514,
+      "step": 27810
+    },
+    {
+      "epoch": 2.786497721240046,
+      "grad_norm": 1.689829707145691,
+      "learning_rate": 2.0594811911805145e-05,
+      "loss": 0.3743,
+      "step": 27820
+    },
+    {
+      "epoch": 2.7874993739670457,
+      "grad_norm": 2.45869517326355,
+      "learning_rate": 2.0579294742305263e-05,
+      "loss": 0.5228,
+      "step": 27830
+    },
+    {
+      "epoch": 2.7885010266940453,
+      "grad_norm": 2.170257806777954,
+      "learning_rate": 2.0563779330663614e-05,
+      "loss": 0.4958,
+      "step": 27840
+    },
+    {
+      "epoch": 2.7895026794210445,
+      "grad_norm": 2.5842323303222656,
+      "learning_rate": 2.0548265683049763e-05,
+      "loss": 0.5162,
+      "step": 27850
+    },
+    {
+      "epoch": 2.7905043321480445,
+      "grad_norm": 2.090120315551758,
+      "learning_rate": 2.0532753805632588e-05,
+      "loss": 0.4321,
+      "step": 27860
+    },
+    {
+      "epoch": 2.7915059848750436,
+      "grad_norm": 1.8928121328353882,
+      "learning_rate": 2.051724370458027e-05,
+      "loss": 0.4296,
+      "step": 27870
+    },
+    {
+      "epoch": 2.7925076376020432,
+      "grad_norm": 2.0915029048919678,
+      "learning_rate": 2.050173538606028e-05,
+      "loss": 0.4516,
+      "step": 27880
+    },
+    {
+      "epoch": 2.793509290329043,
+      "grad_norm": 2.43007230758667,
+      "learning_rate": 2.0486228856239366e-05,
+      "loss": 0.4801,
+      "step": 27890
+    },
+    {
+      "epoch": 2.7945109430560424,
+      "grad_norm": 1.7025607824325562,
+      "learning_rate": 2.0470724121283584e-05,
+      "loss": 0.5191,
+      "step": 27900
+    },
+    {
+      "epoch": 2.795512595783042,
+      "grad_norm": 2.6428115367889404,
+      "learning_rate": 2.0455221187358258e-05,
+      "loss": 0.4427,
+      "step": 27910
+    },
+    {
+      "epoch": 2.7965142485100416,
+      "grad_norm": 2.2131521701812744,
+      "learning_rate": 2.043972006062803e-05,
+      "loss": 0.5153,
+      "step": 27920
+    },
+    {
+      "epoch": 2.797515901237041,
+      "grad_norm": 2.5547220706939697,
+      "learning_rate": 2.0424220747256777e-05,
+      "loss": 0.4719,
+      "step": 27930
+    },
+    {
+      "epoch": 2.798517553964041,
+      "grad_norm": 2.3055248260498047,
+      "learning_rate": 2.040872325340769e-05,
+      "loss": 0.4596,
+      "step": 27940
+    },
+    {
+      "epoch": 2.7995192066910404,
+      "grad_norm": 2.9861836433410645,
+      "learning_rate": 2.039322758524322e-05,
+      "loss": 0.4723,
+      "step": 27950
+    },
+    {
+      "epoch": 2.8005208594180395,
+      "grad_norm": 1.8417468070983887,
+      "learning_rate": 2.0377733748925082e-05,
+      "loss": 0.5091,
+      "step": 27960
+    },
+    {
+      "epoch": 2.8015225121450396,
+      "grad_norm": 2.0634090900421143,
+      "learning_rate": 2.0362241750614298e-05,
+      "loss": 0.4419,
+      "step": 27970
+    },
+    {
+      "epoch": 2.8025241648720387,
+      "grad_norm": 1.7758311033248901,
+      "learning_rate": 2.034675159647112e-05,
+      "loss": 0.5149,
+      "step": 27980
+    },
+    {
+      "epoch": 2.8035258175990383,
+      "grad_norm": 2.2538533210754395,
+      "learning_rate": 2.0331263292655106e-05,
+      "loss": 0.5028,
+      "step": 27990
+    },
+    {
+      "epoch": 2.804527470326038,
+      "grad_norm": 2.3173768520355225,
+      "learning_rate": 2.0315776845325038e-05,
+      "loss": 0.4044,
+      "step": 28000
+    },
+    {
+      "epoch": 2.8055291230530375,
+      "grad_norm": 1.9004391431808472,
+      "learning_rate": 2.030029226063898e-05,
+      "loss": 0.4568,
+      "step": 28010
+    },
+    {
+      "epoch": 2.806530775780037,
+      "grad_norm": 2.0229578018188477,
+      "learning_rate": 2.028480954475427e-05,
+      "loss": 0.4042,
+      "step": 28020
+    },
+    {
+      "epoch": 2.8075324285070367,
+      "grad_norm": 2.288827657699585,
+      "learning_rate": 2.0269328703827468e-05,
+      "loss": 0.4943,
+      "step": 28030
+    },
+    {
+      "epoch": 2.8085340812340363,
+      "grad_norm": 2.060798168182373,
+      "learning_rate": 2.0253849744014405e-05,
+      "loss": 0.4906,
+      "step": 28040
+    },
+    {
+      "epoch": 2.809535733961036,
+      "grad_norm": 1.997622013092041,
+      "learning_rate": 2.0238372671470186e-05,
+      "loss": 0.4681,
+      "step": 28050
+    },
+    {
+      "epoch": 2.8105373866880354,
+      "grad_norm": 2.37577748298645,
+      "learning_rate": 2.0222897492349137e-05,
+      "loss": 0.4847,
+      "step": 28060
+    },
+    {
+      "epoch": 2.8115390394150346,
+      "grad_norm": 2.299891471862793,
+      "learning_rate": 2.0207424212804822e-05,
+      "loss": 0.4673,
+      "step": 28070
+    },
+    {
+      "epoch": 2.8125406921420346,
+      "grad_norm": 1.9214786291122437,
+      "learning_rate": 2.0191952838990093e-05,
+      "loss": 0.4824,
+      "step": 28080
+    },
+    {
+      "epoch": 2.8135423448690338,
+      "grad_norm": 2.1179656982421875,
+      "learning_rate": 2.0176483377056997e-05,
+      "loss": 0.5231,
+      "step": 28090
+    },
+    {
+      "epoch": 2.8145439975960334,
+      "grad_norm": 1.6549978256225586,
+      "learning_rate": 2.0161015833156877e-05,
+      "loss": 0.5406,
+      "step": 28100
+    },
+    {
+      "epoch": 2.815545650323033,
+      "grad_norm": 1.342764973640442,
+      "learning_rate": 2.0145550213440233e-05,
+      "loss": 0.3921,
+      "step": 28110
+    },
+    {
+      "epoch": 2.8165473030500325,
+      "grad_norm": 2.5125224590301514,
+      "learning_rate": 2.0130086524056873e-05,
+      "loss": 0.476,
+      "step": 28120
+    },
+    {
+      "epoch": 2.817548955777032,
+      "grad_norm": 2.1776862144470215,
+      "learning_rate": 2.011462477115581e-05,
+      "loss": 0.5087,
+      "step": 28130
+    },
+    {
+      "epoch": 2.8185506085040317,
+      "grad_norm": 2.026686668395996,
+      "learning_rate": 2.009916496088527e-05,
+      "loss": 0.461,
+      "step": 28140
+    },
+    {
+      "epoch": 2.8195522612310313,
+      "grad_norm": 2.4303576946258545,
+      "learning_rate": 2.008370709939274e-05,
+      "loss": 0.4916,
+      "step": 28150
+    },
+    {
+      "epoch": 2.820553913958031,
+      "grad_norm": 2.582199811935425,
+      "learning_rate": 2.0068251192824904e-05,
+      "loss": 0.4763,
+      "step": 28160
+    },
+    {
+      "epoch": 2.8215555666850305,
+      "grad_norm": 2.5607872009277344,
+      "learning_rate": 2.00527972473277e-05,
+      "loss": 0.5032,
+      "step": 28170
+    },
+    {
+      "epoch": 2.8225572194120296,
+      "grad_norm": 2.378352165222168,
+      "learning_rate": 2.003734526904624e-05,
+      "loss": 0.4109,
+      "step": 28180
+    },
+    {
+      "epoch": 2.8235588721390297,
+      "grad_norm": 1.7307512760162354,
+      "learning_rate": 2.0021895264124892e-05,
+      "loss": 0.4661,
+      "step": 28190
+    },
+    {
+      "epoch": 2.824560524866029,
+      "grad_norm": 2.8087072372436523,
+      "learning_rate": 2.000644723870723e-05,
+      "loss": 0.4496,
+      "step": 28200
+    },
+    {
+      "epoch": 2.8255621775930284,
+      "grad_norm": 1.670059323310852,
+      "learning_rate": 1.9991001198936043e-05,
+      "loss": 0.4691,
+      "step": 28210
+    },
+    {
+      "epoch": 2.826563830320028,
+      "grad_norm": 2.1640727519989014,
+      "learning_rate": 1.9975557150953307e-05,
+      "loss": 0.4691,
+      "step": 28220
+    },
+    {
+      "epoch": 2.8275654830470276,
+      "grad_norm": 2.104672431945801,
+      "learning_rate": 1.996011510090024e-05,
+      "loss": 0.4978,
+      "step": 28230
+    },
+    {
+      "epoch": 2.828567135774027,
+      "grad_norm": 2.842073917388916,
+      "learning_rate": 1.994467505491725e-05,
+      "loss": 0.4459,
+      "step": 28240
+    },
+    {
+      "epoch": 2.8295687885010268,
+      "grad_norm": 2.6502578258514404,
+      "learning_rate": 1.9929237019143937e-05,
+      "loss": 0.4711,
+      "step": 28250
+    },
+    {
+      "epoch": 2.8305704412280264,
+      "grad_norm": 1.7367525100708008,
+      "learning_rate": 1.9913800999719122e-05,
+      "loss": 0.4424,
+      "step": 28260
+    },
+    {
+      "epoch": 2.8315720939550255,
+      "grad_norm": 1.9925302267074585,
+      "learning_rate": 1.989836700278081e-05,
+      "loss": 0.4808,
+      "step": 28270
+    },
+    {
+      "epoch": 2.8325737466820256,
+      "grad_norm": 2.3260669708251953,
+      "learning_rate": 1.988293503446623e-05,
+      "loss": 0.401,
+      "step": 28280
+    },
+    {
+      "epoch": 2.8335753994090247,
+      "grad_norm": 2.675114154815674,
+      "learning_rate": 1.9867505100911744e-05,
+      "loss": 0.4942,
+      "step": 28290
+    },
+    {
+      "epoch": 2.8345770521360243,
+      "grad_norm": 2.485140085220337,
+      "learning_rate": 1.985207720825296e-05,
+      "loss": 0.5136,
+      "step": 28300
+    },
+    {
+      "epoch": 2.835578704863024,
+      "grad_norm": 2.7628536224365234,
+      "learning_rate": 1.9836651362624666e-05,
+      "loss": 0.4945,
+      "step": 28310
+    },
+    {
+      "epoch": 2.8365803575900235,
+      "grad_norm": 2.1196231842041016,
+      "learning_rate": 1.9821227570160804e-05,
+      "loss": 0.5113,
+      "step": 28320
+    },
+    {
+      "epoch": 2.837582010317023,
+      "grad_norm": 1.7910830974578857,
+      "learning_rate": 1.9805805836994546e-05,
+      "loss": 0.4488,
+      "step": 28330
+    },
+    {
+      "epoch": 2.8385836630440227,
+      "grad_norm": 2.0061752796173096,
+      "learning_rate": 1.9790386169258197e-05,
+      "loss": 0.4598,
+      "step": 28340
+    },
+    {
+      "epoch": 2.8395853157710222,
+      "grad_norm": 2.4097883701324463,
+      "learning_rate": 1.9774968573083294e-05,
+      "loss": 0.4771,
+      "step": 28350
+    },
+    {
+      "epoch": 2.840586968498022,
+      "grad_norm": 1.9780919551849365,
+      "learning_rate": 1.97595530546005e-05,
+      "loss": 0.5228,
+      "step": 28360
+    },
+    {
+      "epoch": 2.8415886212250214,
+      "grad_norm": 2.0265541076660156,
+      "learning_rate": 1.9744139619939673e-05,
+      "loss": 0.4618,
+      "step": 28370
+    },
+    {
+      "epoch": 2.8425902739520206,
+      "grad_norm": 2.245145082473755,
+      "learning_rate": 1.972872827522985e-05,
+      "loss": 0.5279,
+      "step": 28380
+    },
+    {
+      "epoch": 2.8435919266790206,
+      "grad_norm": 2.6057050228118896,
+      "learning_rate": 1.9713319026599235e-05,
+      "loss": 0.4666,
+      "step": 28390
+    },
+    {
+      "epoch": 2.8445935794060198,
+      "grad_norm": 1.8463797569274902,
+      "learning_rate": 1.9697911880175178e-05,
+      "loss": 0.4472,
+      "step": 28400
+    },
+    {
+      "epoch": 2.8455952321330193,
+      "grad_norm": 2.043485403060913,
+      "learning_rate": 1.9682506842084218e-05,
+      "loss": 0.4854,
+      "step": 28410
+    },
+    {
+      "epoch": 2.846596884860019,
+      "grad_norm": 2.281378746032715,
+      "learning_rate": 1.966710391845205e-05,
+      "loss": 0.499,
+      "step": 28420
+    },
+    {
+      "epoch": 2.8475985375870185,
+      "grad_norm": 1.9262113571166992,
+      "learning_rate": 1.9651703115403514e-05,
+      "loss": 0.5275,
+      "step": 28430
+    },
+    {
+      "epoch": 2.848600190314018,
+      "grad_norm": 1.9217075109481812,
+      "learning_rate": 1.9636304439062613e-05,
+      "loss": 0.4581,
+      "step": 28440
+    },
+    {
+      "epoch": 2.8496018430410177,
+      "grad_norm": 2.031644105911255,
+      "learning_rate": 1.9620907895552515e-05,
+      "loss": 0.4833,
+      "step": 28450
+    },
+    {
+      "epoch": 2.8506034957680173,
+      "grad_norm": 1.8989890813827515,
+      "learning_rate": 1.9605513490995543e-05,
+      "loss": 0.4247,
+      "step": 28460
+    },
+    {
+      "epoch": 2.851605148495017,
+      "grad_norm": 2.3920555114746094,
+      "learning_rate": 1.9590121231513137e-05,
+      "loss": 0.5115,
+      "step": 28470
+    },
+    {
+      "epoch": 2.8526068012220165,
+      "grad_norm": 2.1428468227386475,
+      "learning_rate": 1.9574731123225922e-05,
+      "loss": 0.5196,
+      "step": 28480
+    },
+    {
+      "epoch": 2.8536084539490156,
+      "grad_norm": 1.9174599647521973,
+      "learning_rate": 1.9559343172253645e-05,
+      "loss": 0.5146,
+      "step": 28490
+    },
+    {
+      "epoch": 2.8546101066760157,
+      "grad_norm": 1.9081904888153076,
+      "learning_rate": 1.9543957384715213e-05,
+      "loss": 0.4258,
+      "step": 28500
+    },
+    {
+      "epoch": 2.855611759403015,
+      "grad_norm": 2.5541326999664307,
+      "learning_rate": 1.952857376672865e-05,
+      "loss": 0.5465,
+      "step": 28510
+    },
+    {
+      "epoch": 2.8566134121300144,
+      "grad_norm": 2.1741418838500977,
+      "learning_rate": 1.951319232441113e-05,
+      "loss": 0.5138,
+      "step": 28520
+    },
+    {
+      "epoch": 2.857615064857014,
+      "grad_norm": 2.374300479888916,
+      "learning_rate": 1.9497813063878977e-05,
+      "loss": 0.5121,
+      "step": 28530
+    },
+    {
+      "epoch": 2.8586167175840136,
+      "grad_norm": 2.0950210094451904,
+      "learning_rate": 1.948243599124761e-05,
+      "loss": 0.4595,
+      "step": 28540
+    },
+    {
+      "epoch": 2.859618370311013,
+      "grad_norm": 1.7163071632385254,
+      "learning_rate": 1.94670611126316e-05,
+      "loss": 0.5161,
+      "step": 28550
+    },
+    {
+      "epoch": 2.8606200230380128,
+      "grad_norm": 2.6274149417877197,
+      "learning_rate": 1.945168843414466e-05,
+      "loss": 0.4492,
+      "step": 28560
+    },
+    {
+      "epoch": 2.8616216757650124,
+      "grad_norm": 2.2445178031921387,
+      "learning_rate": 1.9436317961899612e-05,
+      "loss": 0.4481,
+      "step": 28570
+    },
+    {
+      "epoch": 2.862623328492012,
+      "grad_norm": 2.0394294261932373,
+      "learning_rate": 1.9420949702008384e-05,
+      "loss": 0.4574,
+      "step": 28580
+    },
+    {
+      "epoch": 2.8636249812190115,
+      "grad_norm": 2.9116296768188477,
+      "learning_rate": 1.9405583660582062e-05,
+      "loss": 0.4636,
+      "step": 28590
+    },
+    {
+      "epoch": 2.8646266339460107,
+      "grad_norm": 2.4173495769500732,
+      "learning_rate": 1.9390219843730825e-05,
+      "loss": 0.4435,
+      "step": 28600
+    },
+    {
+      "epoch": 2.8656282866730107,
+      "grad_norm": 2.1900784969329834,
+      "learning_rate": 1.9374858257563968e-05,
+      "loss": 0.4313,
+      "step": 28610
+    },
+    {
+      "epoch": 2.86662993940001,
+      "grad_norm": 2.1137681007385254,
+      "learning_rate": 1.9359498908189893e-05,
+      "loss": 0.4938,
+      "step": 28620
+    },
+    {
+      "epoch": 2.8676315921270095,
+      "grad_norm": 2.0387117862701416,
+      "learning_rate": 1.934414180171614e-05,
+      "loss": 0.4846,
+      "step": 28630
+    },
+    {
+      "epoch": 2.868633244854009,
+      "grad_norm": 2.3030261993408203,
+      "learning_rate": 1.9328786944249347e-05,
+      "loss": 0.444,
+      "step": 28640
+    },
+    {
+      "epoch": 2.8696348975810086,
+      "grad_norm": 2.0840537548065186,
+      "learning_rate": 1.9313434341895224e-05,
+      "loss": 0.4349,
+      "step": 28650
+    },
+    {
+      "epoch": 2.8706365503080082,
+      "grad_norm": 2.0024921894073486,
+      "learning_rate": 1.929808400075863e-05,
+      "loss": 0.4305,
+      "step": 28660
+    },
+    {
+      "epoch": 2.871638203035008,
+      "grad_norm": 2.2334063053131104,
+      "learning_rate": 1.9282735926943492e-05,
+      "loss": 0.4807,
+      "step": 28670
+    },
+    {
+      "epoch": 2.8726398557620074,
+      "grad_norm": 2.153905153274536,
+      "learning_rate": 1.9267390126552876e-05,
+      "loss": 0.4767,
+      "step": 28680
+    },
+    {
+      "epoch": 2.873641508489007,
+      "grad_norm": 1.7628467082977295,
+      "learning_rate": 1.925204660568889e-05,
+      "loss": 0.4862,
+      "step": 28690
+    },
+    {
+      "epoch": 2.8746431612160066,
+      "grad_norm": 2.5055811405181885,
+      "learning_rate": 1.923670537045277e-05,
+      "loss": 0.5051,
+      "step": 28700
+    },
+    {
+      "epoch": 2.8756448139430058,
+      "grad_norm": 2.421543836593628,
+      "learning_rate": 1.9221366426944845e-05,
+      "loss": 0.4746,
+      "step": 28710
+    },
+    {
+      "epoch": 2.876646466670006,
+      "grad_norm": 2.5517852306365967,
+      "learning_rate": 1.9206029781264513e-05,
+      "loss": 0.4647,
+      "step": 28720
+    },
+    {
+      "epoch": 2.877648119397005,
+      "grad_norm": 1.869573712348938,
+      "learning_rate": 1.919069543951027e-05,
+      "loss": 0.4866,
+      "step": 28730
+    },
+    {
+      "epoch": 2.8786497721240045,
+      "grad_norm": 2.2313554286956787,
+      "learning_rate": 1.91753634077797e-05,
+      "loss": 0.4412,
+      "step": 28740
+    },
+    {
+      "epoch": 2.879651424851004,
+      "grad_norm": 2.8452532291412354,
+      "learning_rate": 1.916003369216947e-05,
+      "loss": 0.5655,
+      "step": 28750
+    },
+    {
+      "epoch": 2.8806530775780037,
+      "grad_norm": 2.1433467864990234,
+      "learning_rate": 1.91447062987753e-05,
+      "loss": 0.4176,
+      "step": 28760
+    },
+    {
+      "epoch": 2.8816547303050033,
+      "grad_norm": 1.7829129695892334,
+      "learning_rate": 1.9129381233692014e-05,
+      "loss": 0.4787,
+      "step": 28770
+    },
+    {
+      "epoch": 2.882656383032003,
+      "grad_norm": 2.395498275756836,
+      "learning_rate": 1.911405850301351e-05,
+      "loss": 0.4389,
+      "step": 28780
+    },
+    {
+      "epoch": 2.8836580357590025,
+      "grad_norm": 1.7302987575531006,
+      "learning_rate": 1.909873811283275e-05,
+      "loss": 0.459,
+      "step": 28790
+    },
+    {
+      "epoch": 2.8846596884860016,
+      "grad_norm": 2.646073341369629,
+      "learning_rate": 1.9083420069241747e-05,
+      "loss": 0.4666,
+      "step": 28800
+    },
+    {
+      "epoch": 2.8856613412130017,
+      "grad_norm": 1.9419676065444946,
+      "learning_rate": 1.9068104378331615e-05,
+      "loss": 0.4968,
+      "step": 28810
+    },
+    {
+      "epoch": 2.886662993940001,
+      "grad_norm": 1.7777568101882935,
+      "learning_rate": 1.905279104619252e-05,
+      "loss": 0.4824,
+      "step": 28820
+    },
+    {
+      "epoch": 2.8876646466670004,
+      "grad_norm": 1.9487704038619995,
+      "learning_rate": 1.903748007891367e-05,
+      "loss": 0.5306,
+      "step": 28830
+    },
+    {
+      "epoch": 2.888666299394,
+      "grad_norm": 2.5456743240356445,
+      "learning_rate": 1.902217148258336e-05,
+      "loss": 0.4403,
+      "step": 28840
+    },
+    {
+      "epoch": 2.8896679521209996,
+      "grad_norm": 3.1180431842803955,
+      "learning_rate": 1.9006865263288926e-05,
+      "loss": 0.531,
+      "step": 28850
+    },
+    {
+      "epoch": 2.890669604847999,
+      "grad_norm": 3.0304436683654785,
+      "learning_rate": 1.8991561427116777e-05,
+      "loss": 0.5187,
+      "step": 28860
+    },
+    {
+      "epoch": 2.8916712575749988,
+      "grad_norm": 2.2932333946228027,
+      "learning_rate": 1.8976259980152348e-05,
+      "loss": 0.4541,
+      "step": 28870
+    },
+    {
+      "epoch": 2.8926729103019984,
+      "grad_norm": 2.4942545890808105,
+      "learning_rate": 1.8960960928480136e-05,
+      "loss": 0.4583,
+      "step": 28880
+    },
+    {
+      "epoch": 2.893674563028998,
+      "grad_norm": 1.900305986404419,
+      "learning_rate": 1.8945664278183704e-05,
+      "loss": 0.413,
+      "step": 28890
+    },
+    {
+      "epoch": 2.8946762157559975,
+      "grad_norm": 2.3043212890625,
+      "learning_rate": 1.8930370035345623e-05,
+      "loss": 0.5049,
+      "step": 28900
+    },
+    {
+      "epoch": 2.8956778684829967,
+      "grad_norm": 3.232800006866455,
+      "learning_rate": 1.8915078206047535e-05,
+      "loss": 0.4768,
+      "step": 28910
+    },
+    {
+      "epoch": 2.8966795212099967,
+      "grad_norm": 1.8843035697937012,
+      "learning_rate": 1.889978879637012e-05,
+      "loss": 0.4959,
+      "step": 28920
+    },
+    {
+      "epoch": 2.897681173936996,
+      "grad_norm": 2.1841742992401123,
+      "learning_rate": 1.8884501812393086e-05,
+      "loss": 0.4738,
+      "step": 28930
+    },
+    {
+      "epoch": 2.8986828266639955,
+      "grad_norm": 2.26435923576355,
+      "learning_rate": 1.8869217260195178e-05,
+      "loss": 0.4499,
+      "step": 28940
+    },
+    {
+      "epoch": 2.899684479390995,
+      "grad_norm": 2.054779529571533,
+      "learning_rate": 1.885393514585417e-05,
+      "loss": 0.4213,
+      "step": 28950
+    },
+    {
+      "epoch": 2.9006861321179946,
+      "grad_norm": 2.1258368492126465,
+      "learning_rate": 1.883865547544688e-05,
+      "loss": 0.5051,
+      "step": 28960
+    },
+    {
+      "epoch": 2.9016877848449942,
+      "grad_norm": 1.691569447517395,
+      "learning_rate": 1.882337825504916e-05,
+      "loss": 0.4256,
+      "step": 28970
+    },
+    {
+      "epoch": 2.902689437571994,
+      "grad_norm": 2.8432114124298096,
+      "learning_rate": 1.880810349073585e-05,
+      "loss": 0.512,
+      "step": 28980
+    },
+    {
+      "epoch": 2.9036910902989934,
+      "grad_norm": 2.043947696685791,
+      "learning_rate": 1.8792831188580856e-05,
+      "loss": 0.5151,
+      "step": 28990
+    },
+    {
+      "epoch": 2.904692743025993,
+      "grad_norm": 1.4768218994140625,
+      "learning_rate": 1.8777561354657087e-05,
+      "loss": 0.4742,
+      "step": 29000
+    },
+    {
+      "epoch": 2.9056943957529926,
+      "grad_norm": 2.128981590270996,
+      "learning_rate": 1.8762293995036457e-05,
+      "loss": 0.5125,
+      "step": 29010
+    },
+    {
+      "epoch": 2.9066960484799917,
+      "grad_norm": 2.0752673149108887,
+      "learning_rate": 1.874702911578992e-05,
+      "loss": 0.5301,
+      "step": 29020
+    },
+    {
+      "epoch": 2.907697701206992,
+      "grad_norm": 2.5276637077331543,
+      "learning_rate": 1.873176672298743e-05,
+      "loss": 0.4494,
+      "step": 29030
+    },
+    {
+      "epoch": 2.908699353933991,
+      "grad_norm": 2.664680004119873,
+      "learning_rate": 1.871650682269797e-05,
+      "loss": 0.4918,
+      "step": 29040
+    },
+    {
+      "epoch": 2.9097010066609905,
+      "grad_norm": 2.063596725463867,
+      "learning_rate": 1.87012494209895e-05,
+      "loss": 0.4775,
+      "step": 29050
+    },
+    {
+      "epoch": 2.91070265938799,
+      "grad_norm": 1.8587756156921387,
+      "learning_rate": 1.8685994523929005e-05,
+      "loss": 0.4373,
+      "step": 29060
+    },
+    {
+      "epoch": 2.9117043121149897,
+      "grad_norm": 2.288957118988037,
+      "learning_rate": 1.8670742137582486e-05,
+      "loss": 0.5238,
+      "step": 29070
+    },
+    {
+      "epoch": 2.9127059648419893,
+      "grad_norm": 1.6254380941390991,
+      "learning_rate": 1.8655492268014922e-05,
+      "loss": 0.4024,
+      "step": 29080
+    },
+    {
+      "epoch": 2.913707617568989,
+      "grad_norm": 1.8339122533798218,
+      "learning_rate": 1.864024492129031e-05,
+      "loss": 0.429,
+      "step": 29090
+    },
+    {
+      "epoch": 2.9147092702959885,
+      "grad_norm": 2.6196937561035156,
+      "learning_rate": 1.8625000103471622e-05,
+      "loss": 0.5356,
+      "step": 29100
+    },
+    {
+      "epoch": 2.915710923022988,
+      "grad_norm": 2.130065679550171,
+      "learning_rate": 1.8609757820620856e-05,
+      "loss": 0.4308,
+      "step": 29110
+    },
+    {
+      "epoch": 2.9167125757499877,
+      "grad_norm": 2.1206037998199463,
+      "learning_rate": 1.8594518078798976e-05,
+      "loss": 0.4566,
+      "step": 29120
+    },
+    {
+      "epoch": 2.917714228476987,
+      "grad_norm": 2.1794984340667725,
+      "learning_rate": 1.8579280884065935e-05,
+      "loss": 0.4251,
+      "step": 29130
+    },
+    {
+      "epoch": 2.918715881203987,
+      "grad_norm": 1.4547064304351807,
+      "learning_rate": 1.8564046242480697e-05,
+      "loss": 0.4447,
+      "step": 29140
+    },
+    {
+      "epoch": 2.919717533930986,
+      "grad_norm": 2.497053623199463,
+      "learning_rate": 1.8548814160101197e-05,
+      "loss": 0.4754,
+      "step": 29150
+    },
+    {
+      "epoch": 2.9207191866579856,
+      "grad_norm": 2.3862662315368652,
+      "learning_rate": 1.8533584642984334e-05,
+      "loss": 0.5176,
+      "step": 29160
+    },
+    {
+      "epoch": 2.921720839384985,
+      "grad_norm": 2.6100847721099854,
+      "learning_rate": 1.8518357697186017e-05,
+      "loss": 0.4517,
+      "step": 29170
+    },
+    {
+      "epoch": 2.9227224921119848,
+      "grad_norm": 1.9027708768844604,
+      "learning_rate": 1.8503133328761123e-05,
+      "loss": 0.4236,
+      "step": 29180
+    },
+    {
+      "epoch": 2.9237241448389844,
+      "grad_norm": 2.5158753395080566,
+      "learning_rate": 1.8487911543763486e-05,
+      "loss": 0.4196,
+      "step": 29190
+    },
+    {
+      "epoch": 2.924725797565984,
+      "grad_norm": 2.278364896774292,
+      "learning_rate": 1.847269234824594e-05,
+      "loss": 0.5142,
+      "step": 29200
+    },
+    {
+      "epoch": 2.9257274502929835,
+      "grad_norm": 2.325521469116211,
+      "learning_rate": 1.8457475748260268e-05,
+      "loss": 0.524,
+      "step": 29210
+    },
+    {
+      "epoch": 2.926729103019983,
+      "grad_norm": 2.00109601020813,
+      "learning_rate": 1.844226174985725e-05,
+      "loss": 0.4382,
+      "step": 29220
+    },
+    {
+      "epoch": 2.9277307557469827,
+      "grad_norm": 2.5737013816833496,
+      "learning_rate": 1.8427050359086586e-05,
+      "loss": 0.4711,
+      "step": 29230
+    },
+    {
+      "epoch": 2.928732408473982,
+      "grad_norm": 2.1871094703674316,
+      "learning_rate": 1.8411841581996966e-05,
+      "loss": 0.4493,
+      "step": 29240
+    },
+    {
+      "epoch": 2.9297340612009815,
+      "grad_norm": 2.6685447692871094,
+      "learning_rate": 1.8396635424636056e-05,
+      "loss": 0.4532,
+      "step": 29250
+    },
+    {
+      "epoch": 2.930735713927981,
+      "grad_norm": 2.19049334526062,
+      "learning_rate": 1.838143189305046e-05,
+      "loss": 0.5213,
+      "step": 29260
+    },
+    {
+      "epoch": 2.9317373666549806,
+      "grad_norm": 2.1203505992889404,
+      "learning_rate": 1.836623099328573e-05,
+      "loss": 0.5022,
+      "step": 29270
+    },
+    {
+      "epoch": 2.9327390193819802,
+      "grad_norm": 2.115341901779175,
+      "learning_rate": 1.8351032731386385e-05,
+      "loss": 0.4569,
+      "step": 29280
+    },
+    {
+      "epoch": 2.93374067210898,
+      "grad_norm": 2.712913990020752,
+      "learning_rate": 1.833583711339591e-05,
+      "loss": 0.4849,
+      "step": 29290
+    },
+    {
+      "epoch": 2.9347423248359794,
+      "grad_norm": 2.8189609050750732,
+      "learning_rate": 1.8320644145356697e-05,
+      "loss": 0.5019,
+      "step": 29300
+    },
+    {
+      "epoch": 2.935743977562979,
+      "grad_norm": 1.6259757280349731,
+      "learning_rate": 1.830545383331012e-05,
+      "loss": 0.4915,
+      "step": 29310
+    },
+    {
+      "epoch": 2.9367456302899786,
+      "grad_norm": 2.498462677001953,
+      "learning_rate": 1.8290266183296483e-05,
+      "loss": 0.4389,
+      "step": 29320
+    },
+    {
+      "epoch": 2.9377472830169777,
+      "grad_norm": 2.082543134689331,
+      "learning_rate": 1.8275081201355044e-05,
+      "loss": 0.4655,
+      "step": 29330
+    },
+    {
+      "epoch": 2.9387489357439778,
+      "grad_norm": 2.135136842727661,
+      "learning_rate": 1.8259898893523976e-05,
+      "loss": 0.4485,
+      "step": 29340
+    },
+    {
+      "epoch": 2.939750588470977,
+      "grad_norm": 2.071990489959717,
+      "learning_rate": 1.824471926584041e-05,
+      "loss": 0.4787,
+      "step": 29350
+    },
+    {
+      "epoch": 2.9407522411979765,
+      "grad_norm": 1.6802020072937012,
+      "learning_rate": 1.822954232434041e-05,
+      "loss": 0.4796,
+      "step": 29360
+    },
+    {
+      "epoch": 2.941753893924976,
+      "grad_norm": 1.772538661956787,
+      "learning_rate": 1.8214368075058946e-05,
+      "loss": 0.4682,
+      "step": 29370
+    },
+    {
+      "epoch": 2.9427555466519757,
+      "grad_norm": 2.284114122390747,
+      "learning_rate": 1.8199196524029956e-05,
+      "loss": 0.4882,
+      "step": 29380
+    },
+    {
+      "epoch": 2.9437571993789753,
+      "grad_norm": 2.2266476154327393,
+      "learning_rate": 1.8184027677286275e-05,
+      "loss": 0.494,
+      "step": 29390
+    },
+    {
+      "epoch": 2.944758852105975,
+      "grad_norm": 3.315500497817993,
+      "learning_rate": 1.8168861540859693e-05,
+      "loss": 0.5365,
+      "step": 29400
+    },
+    {
+      "epoch": 2.9457605048329745,
+      "grad_norm": 1.9478274583816528,
+      "learning_rate": 1.815369812078088e-05,
+      "loss": 0.4552,
+      "step": 29410
+    },
+    {
+      "epoch": 2.946762157559974,
+      "grad_norm": 2.8757264614105225,
+      "learning_rate": 1.8138537423079463e-05,
+      "loss": 0.5344,
+      "step": 29420
+    },
+    {
+      "epoch": 2.9477638102869737,
+      "grad_norm": 2.388357400894165,
+      "learning_rate": 1.8123379453783962e-05,
+      "loss": 0.5292,
+      "step": 29430
+    },
+    {
+      "epoch": 2.948765463013973,
+      "grad_norm": 2.107295036315918,
+      "learning_rate": 1.8108224218921838e-05,
+      "loss": 0.4461,
+      "step": 29440
+    },
+    {
+      "epoch": 2.949767115740973,
+      "grad_norm": 2.5148425102233887,
+      "learning_rate": 1.809307172451944e-05,
+      "loss": 0.4258,
+      "step": 29450
+    },
+    {
+      "epoch": 2.950768768467972,
+      "grad_norm": 2.1600284576416016,
+      "learning_rate": 1.8077921976602025e-05,
+      "loss": 0.5051,
+      "step": 29460
+    },
+    {
+      "epoch": 2.9517704211949716,
+      "grad_norm": 1.8847415447235107,
+      "learning_rate": 1.8062774981193796e-05,
+      "loss": 0.4117,
+      "step": 29470
+    },
+    {
+      "epoch": 2.952772073921971,
+      "grad_norm": 1.3785723447799683,
+      "learning_rate": 1.8047630744317817e-05,
+      "loss": 0.4374,
+      "step": 29480
+    },
+    {
+      "epoch": 2.9537737266489708,
+      "grad_norm": 3.6868884563446045,
+      "learning_rate": 1.8032489271996073e-05,
+      "loss": 0.5149,
+      "step": 29490
+    },
+    {
+      "epoch": 2.9547753793759703,
+      "grad_norm": 2.474139451980591,
+      "learning_rate": 1.801735057024946e-05,
+      "loss": 0.4437,
+      "step": 29500
+    },
+    {
+      "epoch": 2.95577703210297,
+      "grad_norm": 2.2893331050872803,
+      "learning_rate": 1.8002214645097755e-05,
+      "loss": 0.4841,
+      "step": 29510
+    },
+    {
+      "epoch": 2.9567786848299695,
+      "grad_norm": 1.97756826877594,
+      "learning_rate": 1.7987081502559634e-05,
+      "loss": 0.4465,
+      "step": 29520
+    },
+    {
+      "epoch": 2.957780337556969,
+      "grad_norm": 2.1581039428710938,
+      "learning_rate": 1.7971951148652676e-05,
+      "loss": 0.4425,
+      "step": 29530
+    },
+    {
+      "epoch": 2.9587819902839687,
+      "grad_norm": 1.6784934997558594,
+      "learning_rate": 1.7956823589393345e-05,
+      "loss": 0.4253,
+      "step": 29540
+    },
+    {
+      "epoch": 2.959783643010968,
+      "grad_norm": 1.9420301914215088,
+      "learning_rate": 1.7941698830797005e-05,
+      "loss": 0.4639,
+      "step": 29550
+    },
+    {
+      "epoch": 2.960785295737968,
+      "grad_norm": 2.0235326290130615,
+      "learning_rate": 1.7926576878877876e-05,
+      "loss": 0.5391,
+      "step": 29560
+    },
+    {
+      "epoch": 2.961786948464967,
+      "grad_norm": 1.8747906684875488,
+      "learning_rate": 1.791145773964909e-05,
+      "loss": 0.4593,
+      "step": 29570
+    },
+    {
+      "epoch": 2.9627886011919666,
+      "grad_norm": 2.1775217056274414,
+      "learning_rate": 1.7896341419122665e-05,
+      "loss": 0.4645,
+      "step": 29580
+    },
+    {
+      "epoch": 2.9637902539189662,
+      "grad_norm": 2.069146156311035,
+      "learning_rate": 1.788122792330946e-05,
+      "loss": 0.4522,
+      "step": 29590
+    },
+    {
+      "epoch": 2.964791906645966,
+      "grad_norm": 2.7413077354431152,
+      "learning_rate": 1.786611725821926e-05,
+      "loss": 0.4479,
+      "step": 29600
+    },
+    {
+      "epoch": 2.9657935593729654,
+      "grad_norm": 2.3349380493164062,
+      "learning_rate": 1.785100942986068e-05,
+      "loss": 0.4827,
+      "step": 29610
+    },
+    {
+      "epoch": 2.966795212099965,
+      "grad_norm": 1.798128604888916,
+      "learning_rate": 1.783590444424125e-05,
+      "loss": 0.4739,
+      "step": 29620
+    },
+    {
+      "epoch": 2.9677968648269646,
+      "grad_norm": 3.001051425933838,
+      "learning_rate": 1.782080230736733e-05,
+      "loss": 0.448,
+      "step": 29630
+    },
+    {
+      "epoch": 2.968798517553964,
+      "grad_norm": 2.069577693939209,
+      "learning_rate": 1.7805703025244163e-05,
+      "loss": 0.378,
+      "step": 29640
+    },
+    {
+      "epoch": 2.9698001702809638,
+      "grad_norm": 1.8276804685592651,
+      "learning_rate": 1.779060660387588e-05,
+      "loss": 0.4602,
+      "step": 29650
+    },
+    {
+      "epoch": 2.970801823007963,
+      "grad_norm": 2.161472797393799,
+      "learning_rate": 1.7775513049265427e-05,
+      "loss": 0.454,
+      "step": 29660
+    },
+    {
+      "epoch": 2.971803475734963,
+      "grad_norm": 2.3388967514038086,
+      "learning_rate": 1.7760422367414643e-05,
+      "loss": 0.4632,
+      "step": 29670
+    },
+    {
+      "epoch": 2.972805128461962,
+      "grad_norm": 2.109787702560425,
+      "learning_rate": 1.774533456432423e-05,
+      "loss": 0.4374,
+      "step": 29680
+    },
+    {
+      "epoch": 2.9738067811889617,
+      "grad_norm": 2.1075479984283447,
+      "learning_rate": 1.7730249645993725e-05,
+      "loss": 0.4088,
+      "step": 29690
+    },
+    {
+      "epoch": 2.9748084339159613,
+      "grad_norm": 2.297384262084961,
+      "learning_rate": 1.7715167618421512e-05,
+      "loss": 0.5332,
+      "step": 29700
+    },
+    {
+      "epoch": 2.975810086642961,
+      "grad_norm": 2.0056192874908447,
+      "learning_rate": 1.7700088487604857e-05,
+      "loss": 0.4708,
+      "step": 29710
+    },
+    {
+      "epoch": 2.9768117393699605,
+      "grad_norm": 2.1155338287353516,
+      "learning_rate": 1.7685012259539847e-05,
+      "loss": 0.4678,
+      "step": 29720
+    },
+    {
+      "epoch": 2.97781339209696,
+      "grad_norm": 2.3071138858795166,
+      "learning_rate": 1.7669938940221438e-05,
+      "loss": 0.4392,
+      "step": 29730
+    },
+    {
+      "epoch": 2.9788150448239596,
+      "grad_norm": 2.3638877868652344,
+      "learning_rate": 1.7654868535643397e-05,
+      "loss": 0.4035,
+      "step": 29740
+    },
+    {
+      "epoch": 2.9798166975509592,
+      "grad_norm": 2.3334412574768066,
+      "learning_rate": 1.763980105179835e-05,
+      "loss": 0.494,
+      "step": 29750
+    },
+    {
+      "epoch": 2.980818350277959,
+      "grad_norm": 2.2591774463653564,
+      "learning_rate": 1.762473649467779e-05,
+      "loss": 0.3932,
+      "step": 29760
+    },
+    {
+      "epoch": 2.981820003004958,
+      "grad_norm": 2.066215991973877,
+      "learning_rate": 1.7609674870271985e-05,
+      "loss": 0.4833,
+      "step": 29770
+    },
+    {
+      "epoch": 2.9828216557319576,
+      "grad_norm": 1.8637421131134033,
+      "learning_rate": 1.7594616184570085e-05,
+      "loss": 0.4515,
+      "step": 29780
+    },
+    {
+      "epoch": 2.983823308458957,
+      "grad_norm": 2.29803729057312,
+      "learning_rate": 1.7579560443560045e-05,
+      "loss": 0.5153,
+      "step": 29790
+    },
+    {
+      "epoch": 2.9848249611859567,
+      "grad_norm": 2.6893725395202637,
+      "learning_rate": 1.7564507653228685e-05,
+      "loss": 0.4788,
+      "step": 29800
+    },
+    {
+      "epoch": 2.9858266139129563,
+      "grad_norm": 2.405869483947754,
+      "learning_rate": 1.7549457819561603e-05,
+      "loss": 0.4965,
+      "step": 29810
+    },
+    {
+      "epoch": 2.986828266639956,
+      "grad_norm": 2.2893760204315186,
+      "learning_rate": 1.7534410948543253e-05,
+      "loss": 0.4852,
+      "step": 29820
+    },
+    {
+      "epoch": 2.9878299193669555,
+      "grad_norm": 2.5346555709838867,
+      "learning_rate": 1.751936704615691e-05,
+      "loss": 0.52,
+      "step": 29830
+    },
+    {
+      "epoch": 2.988831572093955,
+      "grad_norm": 2.542680025100708,
+      "learning_rate": 1.7504326118384658e-05,
+      "loss": 0.433,
+      "step": 29840
+    },
+    {
+      "epoch": 2.9898332248209547,
+      "grad_norm": 3.150050163269043,
+      "learning_rate": 1.7489288171207394e-05,
+      "loss": 0.5064,
+      "step": 29850
+    },
+    {
+      "epoch": 2.990834877547954,
+      "grad_norm": 1.9536211490631104,
+      "learning_rate": 1.7474253210604847e-05,
+      "loss": 0.4579,
+      "step": 29860
+    },
+    {
+      "epoch": 2.991836530274954,
+      "grad_norm": 2.1609601974487305,
+      "learning_rate": 1.7459221242555553e-05,
+      "loss": 0.4633,
+      "step": 29870
+    },
+    {
+      "epoch": 2.992838183001953,
+      "grad_norm": 1.8260406255722046,
+      "learning_rate": 1.744419227303684e-05,
+      "loss": 0.4563,
+      "step": 29880
+    },
+    {
+      "epoch": 2.9938398357289526,
+      "grad_norm": 2.1514317989349365,
+      "learning_rate": 1.742916630802487e-05,
+      "loss": 0.4483,
+      "step": 29890
+    },
+    {
+      "epoch": 2.994841488455952,
+      "grad_norm": 2.2385141849517822,
+      "learning_rate": 1.7414143353494585e-05,
+      "loss": 0.4668,
+      "step": 29900
+    },
+    {
+      "epoch": 2.995843141182952,
+      "grad_norm": 2.0614023208618164,
+      "learning_rate": 1.739912341541977e-05,
+      "loss": 0.4083,
+      "step": 29910
+    },
+    {
+      "epoch": 2.9968447939099514,
+      "grad_norm": 2.2477142810821533,
+      "learning_rate": 1.7384106499772954e-05,
+      "loss": 0.5019,
+      "step": 29920
+    },
+    {
+      "epoch": 2.997846446636951,
+      "grad_norm": 1.8166592121124268,
+      "learning_rate": 1.736909261252551e-05,
+      "loss": 0.4157,
+      "step": 29930
+    },
+    {
+      "epoch": 2.9988480993639506,
+      "grad_norm": 1.802172303199768,
+      "learning_rate": 1.735408175964759e-05,
+      "loss": 0.4273,
+      "step": 29940
+    },
+    {
+      "epoch": 2.99984975209095,
+      "grad_norm": 2.223151445388794,
+      "learning_rate": 1.733907394710813e-05,
+      "loss": 0.3838,
+      "step": 29950
+    },
+    {
+      "epoch": 3.0008013221815997,
+      "grad_norm": 1.856826901435852,
+      "learning_rate": 1.7324069180874874e-05,
+      "loss": 0.3971,
+      "step": 29960
+    },
+    {
+      "epoch": 3.0018029749085993,
+      "grad_norm": 2.1584417819976807,
+      "learning_rate": 1.730906746691435e-05,
+      "loss": 0.4797,
+      "step": 29970
+    },
+    {
+      "epoch": 3.002804627635599,
+      "grad_norm": 2.594388484954834,
+      "learning_rate": 1.729406881119187e-05,
+      "loss": 0.4686,
+      "step": 29980
+    },
+    {
+      "epoch": 3.0038062803625984,
+      "grad_norm": 2.0008304119110107,
+      "learning_rate": 1.7279073219671525e-05,
+      "loss": 0.4361,
+      "step": 29990
+    },
+    {
+      "epoch": 3.004807933089598,
+      "grad_norm": 2.1979317665100098,
+      "learning_rate": 1.7264080698316186e-05,
+      "loss": 0.4022,
+      "step": 30000
+    },
+    {
+      "epoch": 3.004807933089598,
+      "eval_bleu": 0.3025730108532601,
+      "eval_loss": 0.5179290175437927,
+      "eval_rouge1": 0.6867038326183534,
+      "eval_rouge2": 0.5271660350662968,
+      "eval_rougeL": 0.6468106582725407,
+      "eval_runtime": 82898.3122,
+      "eval_samples_per_second": 0.214,
+      "eval_steps_per_second": 0.027,
+      "eval_wer": 0.9895713443561743,
+      "step": 30000
+    },
+    {
+      "epoch": 3.005809585816597,
+      "grad_norm": 1.4881905317306519,
+      "learning_rate": 1.7249091253087528e-05,
+      "loss": 0.4422,
+      "step": 30010
+    },
+    {
+      "epoch": 3.0068112385435968,
+      "grad_norm": 2.210646390914917,
+      "learning_rate": 1.7234104889945983e-05,
+      "loss": 0.435,
+      "step": 30020
+    },
+    {
+      "epoch": 3.0078128912705964,
+      "grad_norm": 2.2528607845306396,
+      "learning_rate": 1.7219121614850737e-05,
+      "loss": 0.4465,
+      "step": 30030
+    },
+    {
+      "epoch": 3.008814543997596,
+      "grad_norm": 1.7737597227096558,
+      "learning_rate": 1.720414143375979e-05,
+      "loss": 0.4228,
+      "step": 30040
+    },
+    {
+      "epoch": 3.0098161967245955,
+      "grad_norm": 2.0847411155700684,
+      "learning_rate": 1.7189164352629895e-05,
+      "loss": 0.3812,
+      "step": 30050
+    },
+    {
+      "epoch": 3.010817849451595,
+      "grad_norm": 2.101858615875244,
+      "learning_rate": 1.7174190377416545e-05,
+      "loss": 0.4121,
+      "step": 30060
+    },
+    {
+      "epoch": 3.0118195021785947,
+      "grad_norm": 2.346134662628174,
+      "learning_rate": 1.7159219514074042e-05,
+      "loss": 0.4031,
+      "step": 30070
+    },
+    {
+      "epoch": 3.0128211549055943,
+      "grad_norm": 2.2245075702667236,
+      "learning_rate": 1.714425176855542e-05,
+      "loss": 0.3885,
+      "step": 30080
+    },
+    {
+      "epoch": 3.013822807632594,
+      "grad_norm": 2.2482717037200928,
+      "learning_rate": 1.71292871468125e-05,
+      "loss": 0.4431,
+      "step": 30090
+    },
+    {
+      "epoch": 3.0148244603595935,
+      "grad_norm": 2.3442859649658203,
+      "learning_rate": 1.7114325654795823e-05,
+      "loss": 0.4431,
+      "step": 30100
+    },
+    {
+      "epoch": 3.015826113086593,
+      "grad_norm": 2.165501356124878,
+      "learning_rate": 1.7099367298454723e-05,
+      "loss": 0.4719,
+      "step": 30110
+    },
+    {
+      "epoch": 3.0168277658135922,
+      "grad_norm": 2.0025136470794678,
+      "learning_rate": 1.708441208373727e-05,
+      "loss": 0.4177,
+      "step": 30120
+    },
+    {
+      "epoch": 3.017829418540592,
+      "grad_norm": 1.7397187948226929,
+      "learning_rate": 1.7069460016590272e-05,
+      "loss": 0.4252,
+      "step": 30130
+    },
+    {
+      "epoch": 3.0188310712675914,
+      "grad_norm": 1.9813027381896973,
+      "learning_rate": 1.7054511102959315e-05,
+      "loss": 0.4546,
+      "step": 30140
+    },
+    {
+      "epoch": 3.019832723994591,
+      "grad_norm": 1.9185618162155151,
+      "learning_rate": 1.70395653487887e-05,
+      "loss": 0.4019,
+      "step": 30150
+    },
+    {
+      "epoch": 3.0208343767215906,
+      "grad_norm": 1.8481942415237427,
+      "learning_rate": 1.7024622760021508e-05,
+      "loss": 0.3715,
+      "step": 30160
+    },
+    {
+      "epoch": 3.02183602944859,
+      "grad_norm": 2.2238407135009766,
+      "learning_rate": 1.7009683342599526e-05,
+      "loss": 0.5105,
+      "step": 30170
+    },
+    {
+      "epoch": 3.02283768217559,
+      "grad_norm": 2.2952382564544678,
+      "learning_rate": 1.6994747102463292e-05,
+      "loss": 0.4362,
+      "step": 30180
+    },
+    {
+      "epoch": 3.0238393349025894,
+      "grad_norm": 2.0249791145324707,
+      "learning_rate": 1.6979814045552095e-05,
+      "loss": 0.4306,
+      "step": 30190
+    },
+    {
+      "epoch": 3.024840987629589,
+      "grad_norm": 2.3426079750061035,
+      "learning_rate": 1.6964884177803942e-05,
+      "loss": 0.4174,
+      "step": 30200
+    },
+    {
+      "epoch": 3.0258426403565886,
+      "grad_norm": 2.315671920776367,
+      "learning_rate": 1.694995750515556e-05,
+      "loss": 0.4406,
+      "step": 30210
+    },
+    {
+      "epoch": 3.0268442930835877,
+      "grad_norm": 2.683622121810913,
+      "learning_rate": 1.6935034033542443e-05,
+      "loss": 0.4369,
+      "step": 30220
+    },
+    {
+      "epoch": 3.0278459458105873,
+      "grad_norm": 2.1116085052490234,
+      "learning_rate": 1.692011376889879e-05,
+      "loss": 0.4556,
+      "step": 30230
+    },
+    {
+      "epoch": 3.028847598537587,
+      "grad_norm": 2.1275312900543213,
+      "learning_rate": 1.6905196717157505e-05,
+      "loss": 0.4808,
+      "step": 30240
+    },
+    {
+      "epoch": 3.0298492512645865,
+      "grad_norm": 1.7960904836654663,
+      "learning_rate": 1.6890282884250247e-05,
+      "loss": 0.4133,
+      "step": 30250
+    },
+    {
+      "epoch": 3.030850903991586,
+      "grad_norm": 1.9978631734848022,
+      "learning_rate": 1.687537227610739e-05,
+      "loss": 0.4625,
+      "step": 30260
+    },
+    {
+      "epoch": 3.0318525567185857,
+      "grad_norm": 2.269608497619629,
+      "learning_rate": 1.6860464898658013e-05,
+      "loss": 0.4151,
+      "step": 30270
+    },
+    {
+      "epoch": 3.0328542094455853,
+      "grad_norm": 2.2099547386169434,
+      "learning_rate": 1.6845560757829903e-05,
+      "loss": 0.4048,
+      "step": 30280
+    },
+    {
+      "epoch": 3.033855862172585,
+      "grad_norm": 2.1496224403381348,
+      "learning_rate": 1.6830659859549593e-05,
+      "loss": 0.3884,
+      "step": 30290
+    },
+    {
+      "epoch": 3.0348575148995844,
+      "grad_norm": 2.702627420425415,
+      "learning_rate": 1.681576220974229e-05,
+      "loss": 0.4569,
+      "step": 30300
+    },
+    {
+      "epoch": 3.035859167626584,
+      "grad_norm": 2.0695064067840576,
+      "learning_rate": 1.6800867814331943e-05,
+      "loss": 0.4137,
+      "step": 30310
+    },
+    {
+      "epoch": 3.0368608203535836,
+      "grad_norm": 2.1617956161499023,
+      "learning_rate": 1.678597667924118e-05,
+      "loss": 0.4098,
+      "step": 30320
+    },
+    {
+      "epoch": 3.0378624730805828,
+      "grad_norm": 2.5029983520507812,
+      "learning_rate": 1.6771088810391332e-05,
+      "loss": 0.4801,
+      "step": 30330
+    },
+    {
+      "epoch": 3.0388641258075824,
+      "grad_norm": 1.9497227668762207,
+      "learning_rate": 1.6756204213702467e-05,
+      "loss": 0.4275,
+      "step": 30340
+    },
+    {
+      "epoch": 3.039865778534582,
+      "grad_norm": 1.6746639013290405,
+      "learning_rate": 1.6741322895093304e-05,
+      "loss": 0.4753,
+      "step": 30350
+    },
+    {
+      "epoch": 3.0408674312615815,
+      "grad_norm": 1.430365800857544,
+      "learning_rate": 1.6726444860481283e-05,
+      "loss": 0.3356,
+      "step": 30360
+    },
+    {
+      "epoch": 3.041869083988581,
+      "grad_norm": 1.6194040775299072,
+      "learning_rate": 1.6711570115782542e-05,
+      "loss": 0.3804,
+      "step": 30370
+    },
+    {
+      "epoch": 3.0428707367155807,
+      "grad_norm": 2.530052423477173,
+      "learning_rate": 1.6696698666911913e-05,
+      "loss": 0.4464,
+      "step": 30380
+    },
+    {
+      "epoch": 3.0438723894425803,
+      "grad_norm": 1.8114992380142212,
+      "learning_rate": 1.668183051978289e-05,
+      "loss": 0.4295,
+      "step": 30390
+    },
+    {
+      "epoch": 3.04487404216958,
+      "grad_norm": 2.182450532913208,
+      "learning_rate": 1.666696568030769e-05,
+      "loss": 0.3924,
+      "step": 30400
+    },
+    {
+      "epoch": 3.0458756948965795,
+      "grad_norm": 1.8115897178649902,
+      "learning_rate": 1.665210415439719e-05,
+      "loss": 0.4182,
+      "step": 30410
+    },
+    {
+      "epoch": 3.046877347623579,
+      "grad_norm": 1.8325234651565552,
+      "learning_rate": 1.6637245947960955e-05,
+      "loss": 0.4205,
+      "step": 30420
+    },
+    {
+      "epoch": 3.0478790003505782,
+      "grad_norm": 2.4362452030181885,
+      "learning_rate": 1.662239106690723e-05,
+      "loss": 0.432,
+      "step": 30430
+    },
+    {
+      "epoch": 3.048880653077578,
+      "grad_norm": 2.317858934402466,
+      "learning_rate": 1.6607539517142952e-05,
+      "loss": 0.4567,
+      "step": 30440
+    },
+    {
+      "epoch": 3.0498823058045774,
+      "grad_norm": 2.2887580394744873,
+      "learning_rate": 1.6592691304573722e-05,
+      "loss": 0.4084,
+      "step": 30450
+    },
+    {
+      "epoch": 3.050883958531577,
+      "grad_norm": 2.171954393386841,
+      "learning_rate": 1.6577846435103794e-05,
+      "loss": 0.4411,
+      "step": 30460
+    },
+    {
+      "epoch": 3.0518856112585766,
+      "grad_norm": 2.362718105316162,
+      "learning_rate": 1.6563004914636132e-05,
+      "loss": 0.4254,
+      "step": 30470
+    },
+    {
+      "epoch": 3.052887263985576,
+      "grad_norm": 2.31545090675354,
+      "learning_rate": 1.6548166749072332e-05,
+      "loss": 0.4505,
+      "step": 30480
+    },
+    {
+      "epoch": 3.053888916712576,
+      "grad_norm": 1.8873096704483032,
+      "learning_rate": 1.6533331944312694e-05,
+      "loss": 0.3795,
+      "step": 30490
+    },
+    {
+      "epoch": 3.0548905694395754,
+      "grad_norm": 2.9692976474761963,
+      "learning_rate": 1.651850050625614e-05,
+      "loss": 0.418,
+      "step": 30500
+    },
+    {
+      "epoch": 3.055892222166575,
+      "grad_norm": 2.100770950317383,
+      "learning_rate": 1.650367244080028e-05,
+      "loss": 0.4898,
+      "step": 30510
+    },
+    {
+      "epoch": 3.0568938748935746,
+      "grad_norm": 2.251727819442749,
+      "learning_rate": 1.6488847753841382e-05,
+      "loss": 0.4811,
+      "step": 30520
+    },
+    {
+      "epoch": 3.057895527620574,
+      "grad_norm": 1.89990234375,
+      "learning_rate": 1.6474026451274356e-05,
+      "loss": 0.4516,
+      "step": 30530
+    },
+    {
+      "epoch": 3.0588971803475733,
+      "grad_norm": 2.172435998916626,
+      "learning_rate": 1.6459208538992772e-05,
+      "loss": 0.3815,
+      "step": 30540
+    },
+    {
+      "epoch": 3.059898833074573,
+      "grad_norm": 2.072516441345215,
+      "learning_rate": 1.6444394022888868e-05,
+      "loss": 0.4441,
+      "step": 30550
+    },
+    {
+      "epoch": 3.0609004858015725,
+      "grad_norm": 2.7983038425445557,
+      "learning_rate": 1.6429582908853515e-05,
+      "loss": 0.3968,
+      "step": 30560
+    },
+    {
+      "epoch": 3.061902138528572,
+      "grad_norm": 1.6153335571289062,
+      "learning_rate": 1.6414775202776224e-05,
+      "loss": 0.4167,
+      "step": 30570
+    },
+    {
+      "epoch": 3.0629037912555717,
+      "grad_norm": 1.913484811782837,
+      "learning_rate": 1.6399970910545165e-05,
+      "loss": 0.4488,
+      "step": 30580
+    },
+    {
+      "epoch": 3.0639054439825713,
+      "grad_norm": 1.8409531116485596,
+      "learning_rate": 1.6385170038047153e-05,
+      "loss": 0.4134,
+      "step": 30590
+    },
+    {
+      "epoch": 3.064907096709571,
+      "grad_norm": 2.3591482639312744,
+      "learning_rate": 1.637037259116764e-05,
+      "loss": 0.3989,
+      "step": 30600
+    },
+    {
+      "epoch": 3.0659087494365704,
+      "grad_norm": 2.223154306411743,
+      "learning_rate": 1.6355578575790696e-05,
+      "loss": 0.4391,
+      "step": 30610
+    },
+    {
+      "epoch": 3.06691040216357,
+      "grad_norm": 1.8484059572219849,
+      "learning_rate": 1.6340787997799058e-05,
+      "loss": 0.4701,
+      "step": 30620
+    },
+    {
+      "epoch": 3.0679120548905696,
+      "grad_norm": 2.000924825668335,
+      "learning_rate": 1.6326000863074084e-05,
+      "loss": 0.4881,
+      "step": 30630
+    },
+    {
+      "epoch": 3.068913707617569,
+      "grad_norm": 2.186718702316284,
+      "learning_rate": 1.6311217177495744e-05,
+      "loss": 0.4545,
+      "step": 30640
+    },
+    {
+      "epoch": 3.0699153603445684,
+      "grad_norm": 2.2570459842681885,
+      "learning_rate": 1.6296436946942666e-05,
+      "loss": 0.4912,
+      "step": 30650
+    },
+    {
+      "epoch": 3.070917013071568,
+      "grad_norm": 1.8138389587402344,
+      "learning_rate": 1.628166017729208e-05,
+      "loss": 0.4317,
+      "step": 30660
+    },
+    {
+      "epoch": 3.0719186657985675,
+      "grad_norm": 2.333362102508545,
+      "learning_rate": 1.6266886874419874e-05,
+      "loss": 0.4869,
+      "step": 30670
+    },
+    {
+      "epoch": 3.072920318525567,
+      "grad_norm": 2.178196907043457,
+      "learning_rate": 1.625211704420051e-05,
+      "loss": 0.5721,
+      "step": 30680
+    },
+    {
+      "epoch": 3.0739219712525667,
+      "grad_norm": 2.0177736282348633,
+      "learning_rate": 1.62373506925071e-05,
+      "loss": 0.4564,
+      "step": 30690
+    },
+    {
+      "epoch": 3.0749236239795663,
+      "grad_norm": 2.001838207244873,
+      "learning_rate": 1.622258782521137e-05,
+      "loss": 0.4209,
+      "step": 30700
+    },
+    {
+      "epoch": 3.075925276706566,
+      "grad_norm": 1.930259108543396,
+      "learning_rate": 1.6207828448183653e-05,
+      "loss": 0.4177,
+      "step": 30710
+    },
+    {
+      "epoch": 3.0769269294335655,
+      "grad_norm": 2.031282663345337,
+      "learning_rate": 1.6193072567292896e-05,
+      "loss": 0.4197,
+      "step": 30720
+    },
+    {
+      "epoch": 3.077928582160565,
+      "grad_norm": 1.463243007659912,
+      "learning_rate": 1.6178320188406665e-05,
+      "loss": 0.4557,
+      "step": 30730
+    },
+    {
+      "epoch": 3.0789302348875647,
+      "grad_norm": 1.8603028059005737,
+      "learning_rate": 1.6163571317391126e-05,
+      "loss": 0.4094,
+      "step": 30740
+    },
+    {
+      "epoch": 3.079931887614564,
+      "grad_norm": 2.061042547225952,
+      "learning_rate": 1.6148825960111038e-05,
+      "loss": 0.3696,
+      "step": 30750
+    },
+    {
+      "epoch": 3.0809335403415634,
+      "grad_norm": 2.072138547897339,
+      "learning_rate": 1.6134084122429778e-05,
+      "loss": 0.4335,
+      "step": 30760
+    },
+    {
+      "epoch": 3.081935193068563,
+      "grad_norm": 1.9868147373199463,
+      "learning_rate": 1.611934581020933e-05,
+      "loss": 0.4321,
+      "step": 30770
+    },
+    {
+      "epoch": 3.0829368457955626,
+      "grad_norm": 2.325610399246216,
+      "learning_rate": 1.610461102931027e-05,
+      "loss": 0.4099,
+      "step": 30780
+    },
+    {
+      "epoch": 3.083938498522562,
+      "grad_norm": 2.9362833499908447,
+      "learning_rate": 1.6089879785591738e-05,
+      "loss": 0.4579,
+      "step": 30790
+    },
+    {
+      "epoch": 3.084940151249562,
+      "grad_norm": 1.9224374294281006,
+      "learning_rate": 1.6075152084911526e-05,
+      "loss": 0.4099,
+      "step": 30800
+    },
+    {
+      "epoch": 3.0859418039765614,
+      "grad_norm": 2.3298308849334717,
+      "learning_rate": 1.6060427933125983e-05,
+      "loss": 0.4423,
+      "step": 30810
+    },
+    {
+      "epoch": 3.086943456703561,
+      "grad_norm": 2.235973834991455,
+      "learning_rate": 1.6045707336090026e-05,
+      "loss": 0.4588,
+      "step": 30820
+    },
+    {
+      "epoch": 3.0879451094305606,
+      "grad_norm": 2.449845790863037,
+      "learning_rate": 1.6030990299657213e-05,
+      "loss": 0.4757,
+      "step": 30830
+    },
+    {
+      "epoch": 3.08894676215756,
+      "grad_norm": 1.9638694524765015,
+      "learning_rate": 1.6016276829679633e-05,
+      "loss": 0.3865,
+      "step": 30840
+    },
+    {
+      "epoch": 3.0899484148845597,
+      "grad_norm": 1.989548921585083,
+      "learning_rate": 1.6001566932008e-05,
+      "loss": 0.393,
+      "step": 30850
+    },
+    {
+      "epoch": 3.090950067611559,
+      "grad_norm": 2.4953360557556152,
+      "learning_rate": 1.598686061249158e-05,
+      "loss": 0.4132,
+      "step": 30860
+    },
+    {
+      "epoch": 3.0919517203385585,
+      "grad_norm": 1.8345521688461304,
+      "learning_rate": 1.597215787697821e-05,
+      "loss": 0.4129,
+      "step": 30870
+    },
+    {
+      "epoch": 3.092953373065558,
+      "grad_norm": 2.3891615867614746,
+      "learning_rate": 1.5957458731314334e-05,
+      "loss": 0.4426,
+      "step": 30880
+    },
+    {
+      "epoch": 3.0939550257925577,
+      "grad_norm": 2.574930429458618,
+      "learning_rate": 1.594276318134495e-05,
+      "loss": 0.4353,
+      "step": 30890
+    },
+    {
+      "epoch": 3.0949566785195572,
+      "grad_norm": 2.4506924152374268,
+      "learning_rate": 1.5928071232913615e-05,
+      "loss": 0.4643,
+      "step": 30900
+    },
+    {
+      "epoch": 3.095958331246557,
+      "grad_norm": 3.072618246078491,
+      "learning_rate": 1.591338289186247e-05,
+      "loss": 0.4184,
+      "step": 30910
+    },
+    {
+      "epoch": 3.0969599839735564,
+      "grad_norm": 2.3220643997192383,
+      "learning_rate": 1.589869816403222e-05,
+      "loss": 0.4504,
+      "step": 30920
+    },
+    {
+      "epoch": 3.097961636700556,
+      "grad_norm": 1.746625542640686,
+      "learning_rate": 1.588401705526213e-05,
+      "loss": 0.4448,
+      "step": 30930
+    },
+    {
+      "epoch": 3.0989632894275556,
+      "grad_norm": 2.5139360427856445,
+      "learning_rate": 1.5869339571390012e-05,
+      "loss": 0.4606,
+      "step": 30940
+    },
+    {
+      "epoch": 3.099964942154555,
+      "grad_norm": 2.294275999069214,
+      "learning_rate": 1.585466571825227e-05,
+      "loss": 0.4509,
+      "step": 30950
+    },
+    {
+      "epoch": 3.1009665948815544,
+      "grad_norm": 2.091242551803589,
+      "learning_rate": 1.5839995501683842e-05,
+      "loss": 0.3922,
+      "step": 30960
+    },
+    {
+      "epoch": 3.101968247608554,
+      "grad_norm": 1.6841497421264648,
+      "learning_rate": 1.5825328927518203e-05,
+      "loss": 0.4624,
+      "step": 30970
+    },
+    {
+      "epoch": 3.1029699003355535,
+      "grad_norm": 2.2461984157562256,
+      "learning_rate": 1.5810666001587422e-05,
+      "loss": 0.4055,
+      "step": 30980
+    },
+    {
+      "epoch": 3.103971553062553,
+      "grad_norm": 2.257538318634033,
+      "learning_rate": 1.579600672972209e-05,
+      "loss": 0.4854,
+      "step": 30990
+    },
+    {
+      "epoch": 3.1049732057895527,
+      "grad_norm": 2.6897525787353516,
+      "learning_rate": 1.5781351117751336e-05,
+      "loss": 0.4658,
+      "step": 31000
+    },
+    {
+      "epoch": 3.1059748585165523,
+      "grad_norm": 2.213277816772461,
+      "learning_rate": 1.5766699171502864e-05,
+      "loss": 0.4478,
+      "step": 31010
+    },
+    {
+      "epoch": 3.106976511243552,
+      "grad_norm": 2.556442975997925,
+      "learning_rate": 1.575205089680289e-05,
+      "loss": 0.385,
+      "step": 31020
+    },
+    {
+      "epoch": 3.1079781639705515,
+      "grad_norm": 2.161545515060425,
+      "learning_rate": 1.5737406299476202e-05,
+      "loss": 0.3858,
+      "step": 31030
+    },
+    {
+      "epoch": 3.108979816697551,
+      "grad_norm": 2.166633367538452,
+      "learning_rate": 1.5722765385346088e-05,
+      "loss": 0.4702,
+      "step": 31040
+    },
+    {
+      "epoch": 3.1099814694245507,
+      "grad_norm": 1.7207388877868652,
+      "learning_rate": 1.5708128160234398e-05,
+      "loss": 0.4262,
+      "step": 31050
+    },
+    {
+      "epoch": 3.1109831221515503,
+      "grad_norm": 2.28712797164917,
+      "learning_rate": 1.569349462996151e-05,
+      "loss": 0.4538,
+      "step": 31060
+    },
+    {
+      "epoch": 3.1119847748785494,
+      "grad_norm": 2.292612075805664,
+      "learning_rate": 1.567886480034634e-05,
+      "loss": 0.41,
+      "step": 31070
+    },
+    {
+      "epoch": 3.112986427605549,
+      "grad_norm": 2.4569859504699707,
+      "learning_rate": 1.5664238677206303e-05,
+      "loss": 0.4326,
+      "step": 31080
+    },
+    {
+      "epoch": 3.1139880803325486,
+      "grad_norm": 2.032771110534668,
+      "learning_rate": 1.5649616266357364e-05,
+      "loss": 0.446,
+      "step": 31090
+    },
+    {
+      "epoch": 3.114989733059548,
+      "grad_norm": 2.4700424671173096,
+      "learning_rate": 1.563499757361403e-05,
+      "loss": 0.4004,
+      "step": 31100
+    },
+    {
+      "epoch": 3.1159913857865478,
+      "grad_norm": 2.9324538707733154,
+      "learning_rate": 1.562038260478928e-05,
+      "loss": 0.3842,
+      "step": 31110
+    },
+    {
+      "epoch": 3.1169930385135474,
+      "grad_norm": 2.291475534439087,
+      "learning_rate": 1.5605771365694653e-05,
+      "loss": 0.3809,
+      "step": 31120
+    },
+    {
+      "epoch": 3.117994691240547,
+      "grad_norm": 1.8152060508728027,
+      "learning_rate": 1.5591163862140193e-05,
+      "loss": 0.4574,
+      "step": 31130
+    },
+    {
+      "epoch": 3.1189963439675465,
+      "grad_norm": 2.637399911880493,
+      "learning_rate": 1.557656009993446e-05,
+      "loss": 0.4402,
+      "step": 31140
+    },
+    {
+      "epoch": 3.119997996694546,
+      "grad_norm": 2.726431131362915,
+      "learning_rate": 1.5561960084884507e-05,
+      "loss": 0.4582,
+      "step": 31150
+    },
+    {
+      "epoch": 3.1209996494215457,
+      "grad_norm": 1.8519651889801025,
+      "learning_rate": 1.5547363822795935e-05,
+      "loss": 0.3868,
+      "step": 31160
+    },
+    {
+      "epoch": 3.1220013021485453,
+      "grad_norm": 1.980515480041504,
+      "learning_rate": 1.5532771319472824e-05,
+      "loss": 0.4291,
+      "step": 31170
+    },
+    {
+      "epoch": 3.1230029548755445,
+      "grad_norm": 2.1525003910064697,
+      "learning_rate": 1.5518182580717756e-05,
+      "loss": 0.4091,
+      "step": 31180
+    },
+    {
+      "epoch": 3.124004607602544,
+      "grad_norm": 2.2667062282562256,
+      "learning_rate": 1.5503597612331834e-05,
+      "loss": 0.4454,
+      "step": 31190
+    },
+    {
+      "epoch": 3.1250062603295437,
+      "grad_norm": 1.6090610027313232,
+      "learning_rate": 1.5489016420114658e-05,
+      "loss": 0.4377,
+      "step": 31200
+    },
+    {
+      "epoch": 3.1260079130565432,
+      "grad_norm": 1.957115650177002,
+      "learning_rate": 1.5474439009864328e-05,
+      "loss": 0.4627,
+      "step": 31210
+    },
+    {
+      "epoch": 3.127009565783543,
+      "grad_norm": 2.4358901977539062,
+      "learning_rate": 1.545986538737742e-05,
+      "loss": 0.4202,
+      "step": 31220
+    },
+    {
+      "epoch": 3.1280112185105424,
+      "grad_norm": 2.5619661808013916,
+      "learning_rate": 1.5445295558449024e-05,
+      "loss": 0.4035,
+      "step": 31230
+    },
+    {
+      "epoch": 3.129012871237542,
+      "grad_norm": 3.102004289627075,
+      "learning_rate": 1.5430729528872722e-05,
+      "loss": 0.4026,
+      "step": 31240
+    },
+    {
+      "epoch": 3.1300145239645416,
+      "grad_norm": 1.8699795007705688,
+      "learning_rate": 1.541616730444058e-05,
+      "loss": 0.4918,
+      "step": 31250
+    },
+    {
+      "epoch": 3.131016176691541,
+      "grad_norm": 1.6903408765792847,
+      "learning_rate": 1.540160889094315e-05,
+      "loss": 0.37,
+      "step": 31260
+    },
+    {
+      "epoch": 3.132017829418541,
+      "grad_norm": 2.113809108734131,
+      "learning_rate": 1.5387054294169455e-05,
+      "loss": 0.4645,
+      "step": 31270
+    },
+    {
+      "epoch": 3.13301948214554,
+      "grad_norm": 1.8396596908569336,
+      "learning_rate": 1.5372503519907035e-05,
+      "loss": 0.3981,
+      "step": 31280
+    },
+    {
+      "epoch": 3.1340211348725395,
+      "grad_norm": 1.8806499242782593,
+      "learning_rate": 1.5357956573941872e-05,
+      "loss": 0.4145,
+      "step": 31290
+    },
+    {
+      "epoch": 3.135022787599539,
+      "grad_norm": 1.8449625968933105,
+      "learning_rate": 1.5343413462058453e-05,
+      "loss": 0.4274,
+      "step": 31300
+    },
+    {
+      "epoch": 3.1360244403265387,
+      "grad_norm": 2.39495849609375,
+      "learning_rate": 1.532887419003973e-05,
+      "loss": 0.4665,
+      "step": 31310
+    },
+    {
+      "epoch": 3.1370260930535383,
+      "grad_norm": 2.1482601165771484,
+      "learning_rate": 1.5314338763667132e-05,
+      "loss": 0.4653,
+      "step": 31320
+    },
+    {
+      "epoch": 3.138027745780538,
+      "grad_norm": 1.9213060140609741,
+      "learning_rate": 1.529980718872054e-05,
+      "loss": 0.4786,
+      "step": 31330
+    },
+    {
+      "epoch": 3.1390293985075375,
+      "grad_norm": 1.8243101835250854,
+      "learning_rate": 1.5285279470978335e-05,
+      "loss": 0.3736,
+      "step": 31340
+    },
+    {
+      "epoch": 3.140031051234537,
+      "grad_norm": 2.4560937881469727,
+      "learning_rate": 1.5270755616217343e-05,
+      "loss": 0.4372,
+      "step": 31350
+    },
+    {
+      "epoch": 3.1410327039615367,
+      "grad_norm": 1.697628140449524,
+      "learning_rate": 1.5256235630212864e-05,
+      "loss": 0.3558,
+      "step": 31360
+    },
+    {
+      "epoch": 3.1420343566885363,
+      "grad_norm": 1.9843086004257202,
+      "learning_rate": 1.5241719518738646e-05,
+      "loss": 0.4401,
+      "step": 31370
+    },
+    {
+      "epoch": 3.1430360094155354,
+      "grad_norm": 2.132075786590576,
+      "learning_rate": 1.5227207287566903e-05,
+      "loss": 0.4759,
+      "step": 31380
+    },
+    {
+      "epoch": 3.144037662142535,
+      "grad_norm": 1.9841755628585815,
+      "learning_rate": 1.521269894246833e-05,
+      "loss": 0.4012,
+      "step": 31390
+    },
+    {
+      "epoch": 3.1450393148695346,
+      "grad_norm": 2.36376690864563,
+      "learning_rate": 1.5198194489212025e-05,
+      "loss": 0.4353,
+      "step": 31400
+    },
+    {
+      "epoch": 3.146040967596534,
+      "grad_norm": 2.0222671031951904,
+      "learning_rate": 1.5183693933565585e-05,
+      "loss": 0.4259,
+      "step": 31410
+    },
+    {
+      "epoch": 3.1470426203235338,
+      "grad_norm": 1.9244353771209717,
+      "learning_rate": 1.5169197281295033e-05,
+      "loss": 0.4191,
+      "step": 31420
+    },
+    {
+      "epoch": 3.1480442730505334,
+      "grad_norm": 2.056471586227417,
+      "learning_rate": 1.515470453816486e-05,
+      "loss": 0.44,
+      "step": 31430
+    },
+    {
+      "epoch": 3.149045925777533,
+      "grad_norm": 2.2160956859588623,
+      "learning_rate": 1.514021570993798e-05,
+      "loss": 0.4678,
+      "step": 31440
+    },
+    {
+      "epoch": 3.1500475785045325,
+      "grad_norm": 2.4954745769500732,
+      "learning_rate": 1.5125730802375754e-05,
+      "loss": 0.4518,
+      "step": 31450
+    },
+    {
+      "epoch": 3.151049231231532,
+      "grad_norm": 2.0969278812408447,
+      "learning_rate": 1.511124982123801e-05,
+      "loss": 0.4325,
+      "step": 31460
+    },
+    {
+      "epoch": 3.1520508839585317,
+      "grad_norm": 2.045363664627075,
+      "learning_rate": 1.5096772772282977e-05,
+      "loss": 0.4275,
+      "step": 31470
+    },
+    {
+      "epoch": 3.1530525366855313,
+      "grad_norm": 2.1558549404144287,
+      "learning_rate": 1.5082299661267338e-05,
+      "loss": 0.4799,
+      "step": 31480
+    },
+    {
+      "epoch": 3.1540541894125305,
+      "grad_norm": 1.989071249961853,
+      "learning_rate": 1.5067830493946222e-05,
+      "loss": 0.444,
+      "step": 31490
+    },
+    {
+      "epoch": 3.15505584213953,
+      "grad_norm": 1.6298166513442993,
+      "learning_rate": 1.5053365276073176e-05,
+      "loss": 0.4246,
+      "step": 31500
+    },
+    {
+      "epoch": 3.1560574948665296,
+      "grad_norm": 1.942826747894287,
+      "learning_rate": 1.5038904013400165e-05,
+      "loss": 0.4383,
+      "step": 31510
+    },
+    {
+      "epoch": 3.1570591475935292,
+      "grad_norm": 2.3768038749694824,
+      "learning_rate": 1.502444671167761e-05,
+      "loss": 0.4235,
+      "step": 31520
+    },
+    {
+      "epoch": 3.158060800320529,
+      "grad_norm": 1.900918960571289,
+      "learning_rate": 1.500999337665433e-05,
+      "loss": 0.3984,
+      "step": 31530
+    },
+    {
+      "epoch": 3.1590624530475284,
+      "grad_norm": 2.272731304168701,
+      "learning_rate": 1.4995544014077595e-05,
+      "loss": 0.4221,
+      "step": 31540
+    },
+    {
+      "epoch": 3.160064105774528,
+      "grad_norm": 2.2893619537353516,
+      "learning_rate": 1.4981098629693066e-05,
+      "loss": 0.4546,
+      "step": 31550
+    },
+    {
+      "epoch": 3.1610657585015276,
+      "grad_norm": 2.42533016204834,
+      "learning_rate": 1.496665722924483e-05,
+      "loss": 0.3829,
+      "step": 31560
+    },
+    {
+      "epoch": 3.162067411228527,
+      "grad_norm": 2.7083005905151367,
+      "learning_rate": 1.4952219818475422e-05,
+      "loss": 0.4093,
+      "step": 31570
+    },
+    {
+      "epoch": 3.163069063955527,
+      "grad_norm": 2.4388954639434814,
+      "learning_rate": 1.4937786403125725e-05,
+      "loss": 0.4436,
+      "step": 31580
+    },
+    {
+      "epoch": 3.1640707166825264,
+      "grad_norm": 2.242216110229492,
+      "learning_rate": 1.4923356988935094e-05,
+      "loss": 0.4942,
+      "step": 31590
+    },
+    {
+      "epoch": 3.1650723694095255,
+      "grad_norm": 2.0267419815063477,
+      "learning_rate": 1.4908931581641267e-05,
+      "loss": 0.4141,
+      "step": 31600
+    },
+    {
+      "epoch": 3.166074022136525,
+      "grad_norm": 2.0003437995910645,
+      "learning_rate": 1.48945101869804e-05,
+      "loss": 0.4125,
+      "step": 31610
+    },
+    {
+      "epoch": 3.1670756748635247,
+      "grad_norm": 2.4626035690307617,
+      "learning_rate": 1.4880092810687034e-05,
+      "loss": 0.3956,
+      "step": 31620
+    },
+    {
+      "epoch": 3.1680773275905243,
+      "grad_norm": 2.163177967071533,
+      "learning_rate": 1.4865679458494123e-05,
+      "loss": 0.4299,
+      "step": 31630
+    },
+    {
+      "epoch": 3.169078980317524,
+      "grad_norm": 2.259563684463501,
+      "learning_rate": 1.485127013613303e-05,
+      "loss": 0.4229,
+      "step": 31640
+    },
+    {
+      "epoch": 3.1700806330445235,
+      "grad_norm": 1.6991698741912842,
+      "learning_rate": 1.4836864849333509e-05,
+      "loss": 0.4263,
+      "step": 31650
+    },
+    {
+      "epoch": 3.171082285771523,
+      "grad_norm": 2.012312889099121,
+      "learning_rate": 1.4822463603823694e-05,
+      "loss": 0.4329,
+      "step": 31660
+    },
+    {
+      "epoch": 3.1720839384985227,
+      "grad_norm": 1.6116600036621094,
+      "learning_rate": 1.480806640533014e-05,
+      "loss": 0.45,
+      "step": 31670
+    },
+    {
+      "epoch": 3.1730855912255223,
+      "grad_norm": 1.7727469205856323,
+      "learning_rate": 1.4793673259577773e-05,
+      "loss": 0.4319,
+      "step": 31680
+    },
+    {
+      "epoch": 3.174087243952522,
+      "grad_norm": 1.567363977432251,
+      "learning_rate": 1.4779284172289904e-05,
+      "loss": 0.4308,
+      "step": 31690
+    },
+    {
+      "epoch": 3.1750888966795214,
+      "grad_norm": 2.1986780166625977,
+      "learning_rate": 1.4764899149188256e-05,
+      "loss": 0.4781,
+      "step": 31700
+    },
+    {
+      "epoch": 3.1760905494065206,
+      "grad_norm": 2.068117380142212,
+      "learning_rate": 1.4750518195992897e-05,
+      "loss": 0.4237,
+      "step": 31710
+    },
+    {
+      "epoch": 3.17709220213352,
+      "grad_norm": 2.2476770877838135,
+      "learning_rate": 1.4736141318422331e-05,
+      "loss": 0.478,
+      "step": 31720
+    },
+    {
+      "epoch": 3.1780938548605198,
+      "grad_norm": 2.5276355743408203,
+      "learning_rate": 1.4721768522193372e-05,
+      "loss": 0.4347,
+      "step": 31730
+    },
+    {
+      "epoch": 3.1790955075875194,
+      "grad_norm": 2.2223894596099854,
+      "learning_rate": 1.4707399813021272e-05,
+      "loss": 0.466,
+      "step": 31740
+    },
+    {
+      "epoch": 3.180097160314519,
+      "grad_norm": 1.6248425245285034,
+      "learning_rate": 1.4693035196619629e-05,
+      "loss": 0.4342,
+      "step": 31750
+    },
+    {
+      "epoch": 3.1810988130415185,
+      "grad_norm": 1.8279471397399902,
+      "learning_rate": 1.4678674678700405e-05,
+      "loss": 0.4708,
+      "step": 31760
+    },
+    {
+      "epoch": 3.182100465768518,
+      "grad_norm": 2.691256523132324,
+      "learning_rate": 1.4664318264973965e-05,
+      "loss": 0.4571,
+      "step": 31770
+    },
+    {
+      "epoch": 3.1831021184955177,
+      "grad_norm": 4.807773590087891,
+      "learning_rate": 1.4649965961149003e-05,
+      "loss": 0.4275,
+      "step": 31780
+    },
+    {
+      "epoch": 3.1841037712225173,
+      "grad_norm": 1.9490879774093628,
+      "learning_rate": 1.4635617772932625e-05,
+      "loss": 0.4218,
+      "step": 31790
+    },
+    {
+      "epoch": 3.185105423949517,
+      "grad_norm": 2.2792632579803467,
+      "learning_rate": 1.462127370603025e-05,
+      "loss": 0.4507,
+      "step": 31800
+    },
+    {
+      "epoch": 3.186107076676516,
+      "grad_norm": 2.2887744903564453,
+      "learning_rate": 1.4606933766145686e-05,
+      "loss": 0.458,
+      "step": 31810
+    },
+    {
+      "epoch": 3.1871087294035156,
+      "grad_norm": 2.0166826248168945,
+      "learning_rate": 1.4592597958981102e-05,
+      "loss": 0.4476,
+      "step": 31820
+    },
+    {
+      "epoch": 3.1881103821305152,
+      "grad_norm": 1.9539391994476318,
+      "learning_rate": 1.457826629023702e-05,
+      "loss": 0.4931,
+      "step": 31830
+    },
+    {
+      "epoch": 3.189112034857515,
+      "grad_norm": 2.3129167556762695,
+      "learning_rate": 1.4563938765612304e-05,
+      "loss": 0.4348,
+      "step": 31840
+    },
+    {
+      "epoch": 3.1901136875845144,
+      "grad_norm": 2.1217124462127686,
+      "learning_rate": 1.4549615390804183e-05,
+      "loss": 0.4267,
+      "step": 31850
+    },
+    {
+      "epoch": 3.191115340311514,
+      "grad_norm": 2.5243582725524902,
+      "learning_rate": 1.453529617150824e-05,
+      "loss": 0.4407,
+      "step": 31860
+    },
+    {
+      "epoch": 3.1921169930385136,
+      "grad_norm": 2.3081367015838623,
+      "learning_rate": 1.4520981113418391e-05,
+      "loss": 0.5121,
+      "step": 31870
+    },
+    {
+      "epoch": 3.193118645765513,
+      "grad_norm": 2.267855167388916,
+      "learning_rate": 1.450667022222691e-05,
+      "loss": 0.4379,
+      "step": 31880
+    },
+    {
+      "epoch": 3.1941202984925128,
+      "grad_norm": 1.9463863372802734,
+      "learning_rate": 1.44923635036244e-05,
+      "loss": 0.4241,
+      "step": 31890
+    },
+    {
+      "epoch": 3.1951219512195124,
+      "grad_norm": 1.9318593740463257,
+      "learning_rate": 1.4478060963299825e-05,
+      "loss": 0.4678,
+      "step": 31900
+    },
+    {
+      "epoch": 3.1961236039465115,
+      "grad_norm": 2.170785665512085,
+      "learning_rate": 1.4463762606940467e-05,
+      "loss": 0.471,
+      "step": 31910
+    },
+    {
+      "epoch": 3.197125256673511,
+      "grad_norm": 2.2391130924224854,
+      "learning_rate": 1.4449468440231961e-05,
+      "loss": 0.4823,
+      "step": 31920
+    },
+    {
+      "epoch": 3.1981269094005107,
+      "grad_norm": 2.5374581813812256,
+      "learning_rate": 1.4435178468858263e-05,
+      "loss": 0.4299,
+      "step": 31930
+    },
+    {
+      "epoch": 3.1991285621275103,
+      "grad_norm": 1.7519168853759766,
+      "learning_rate": 1.4420892698501665e-05,
+      "loss": 0.4503,
+      "step": 31940
+    },
+    {
+      "epoch": 3.20013021485451,
+      "grad_norm": 2.36940860748291,
+      "learning_rate": 1.4406611134842795e-05,
+      "loss": 0.4593,
+      "step": 31950
+    },
+    {
+      "epoch": 3.2011318675815095,
+      "grad_norm": 2.056506872177124,
+      "learning_rate": 1.4392333783560602e-05,
+      "loss": 0.4437,
+      "step": 31960
+    },
+    {
+      "epoch": 3.202133520308509,
+      "grad_norm": 2.2406508922576904,
+      "learning_rate": 1.4378060650332359e-05,
+      "loss": 0.3915,
+      "step": 31970
+    },
+    {
+      "epoch": 3.2031351730355087,
+      "grad_norm": 2.973459482192993,
+      "learning_rate": 1.4363791740833666e-05,
+      "loss": 0.4502,
+      "step": 31980
+    },
+    {
+      "epoch": 3.2041368257625082,
+      "grad_norm": 2.4402339458465576,
+      "learning_rate": 1.4349527060738435e-05,
+      "loss": 0.4597,
+      "step": 31990
+    },
+    {
+      "epoch": 3.205138478489508,
+      "grad_norm": 2.2443811893463135,
+      "learning_rate": 1.43352666157189e-05,
+      "loss": 0.4083,
+      "step": 32000
+    },
+    {
+      "epoch": 3.2061401312165074,
+      "grad_norm": 1.9804131984710693,
+      "learning_rate": 1.432101041144564e-05,
+      "loss": 0.4083,
+      "step": 32010
+    },
+    {
+      "epoch": 3.2071417839435066,
+      "grad_norm": 2.6847102642059326,
+      "learning_rate": 1.4306758453587494e-05,
+      "loss": 0.4574,
+      "step": 32020
+    },
+    {
+      "epoch": 3.208143436670506,
+      "grad_norm": 2.2255048751831055,
+      "learning_rate": 1.4292510747811633e-05,
+      "loss": 0.3964,
+      "step": 32030
+    },
+    {
+      "epoch": 3.2091450893975058,
+      "grad_norm": 1.9805270433425903,
+      "learning_rate": 1.4278267299783576e-05,
+      "loss": 0.454,
+      "step": 32040
+    },
+    {
+      "epoch": 3.2101467421245053,
+      "grad_norm": 2.9171619415283203,
+      "learning_rate": 1.4264028115167091e-05,
+      "loss": 0.4639,
+      "step": 32050
+    },
+    {
+      "epoch": 3.211148394851505,
+      "grad_norm": 1.7909626960754395,
+      "learning_rate": 1.4249793199624284e-05,
+      "loss": 0.4419,
+      "step": 32060
+    },
+    {
+      "epoch": 3.2121500475785045,
+      "grad_norm": 1.8210331201553345,
+      "learning_rate": 1.4235562558815544e-05,
+      "loss": 0.3819,
+      "step": 32070
+    },
+    {
+      "epoch": 3.213151700305504,
+      "grad_norm": 2.2666754722595215,
+      "learning_rate": 1.4221336198399605e-05,
+      "loss": 0.4234,
+      "step": 32080
+    },
+    {
+      "epoch": 3.2141533530325037,
+      "grad_norm": 2.126728057861328,
+      "learning_rate": 1.420711412403343e-05,
+      "loss": 0.4262,
+      "step": 32090
+    },
+    {
+      "epoch": 3.2151550057595033,
+      "grad_norm": 2.2342824935913086,
+      "learning_rate": 1.419289634137232e-05,
+      "loss": 0.4111,
+      "step": 32100
+    },
+    {
+      "epoch": 3.216156658486503,
+      "grad_norm": 2.0711076259613037,
+      "learning_rate": 1.417868285606988e-05,
+      "loss": 0.4585,
+      "step": 32110
+    },
+    {
+      "epoch": 3.2171583112135025,
+      "grad_norm": 2.6071414947509766,
+      "learning_rate": 1.416447367377799e-05,
+      "loss": 0.4229,
+      "step": 32120
+    },
+    {
+      "epoch": 3.2181599639405016,
+      "grad_norm": 2.0745346546173096,
+      "learning_rate": 1.4150268800146787e-05,
+      "loss": 0.4607,
+      "step": 32130
+    },
+    {
+      "epoch": 3.2191616166675012,
+      "grad_norm": 2.3919355869293213,
+      "learning_rate": 1.4136068240824758e-05,
+      "loss": 0.4789,
+      "step": 32140
+    },
+    {
+      "epoch": 3.220163269394501,
+      "grad_norm": 1.9690043926239014,
+      "learning_rate": 1.4121872001458641e-05,
+      "loss": 0.4274,
+      "step": 32150
+    },
+    {
+      "epoch": 3.2211649221215004,
+      "grad_norm": 2.3856241703033447,
+      "learning_rate": 1.4107680087693425e-05,
+      "loss": 0.3834,
+      "step": 32160
+    },
+    {
+      "epoch": 3.2221665748485,
+      "grad_norm": 2.1940131187438965,
+      "learning_rate": 1.4093492505172445e-05,
+      "loss": 0.3788,
+      "step": 32170
+    },
+    {
+      "epoch": 3.2231682275754996,
+      "grad_norm": 2.333587169647217,
+      "learning_rate": 1.407930925953727e-05,
+      "loss": 0.4461,
+      "step": 32180
+    },
+    {
+      "epoch": 3.224169880302499,
+      "grad_norm": 2.39921236038208,
+      "learning_rate": 1.4065130356427753e-05,
+      "loss": 0.4476,
+      "step": 32190
+    },
+    {
+      "epoch": 3.2251715330294988,
+      "grad_norm": 2.1098825931549072,
+      "learning_rate": 1.4050955801482024e-05,
+      "loss": 0.4647,
+      "step": 32200
+    },
+    {
+      "epoch": 3.2261731857564984,
+      "grad_norm": 1.7223210334777832,
+      "learning_rate": 1.4036785600336477e-05,
+      "loss": 0.4261,
+      "step": 32210
+    },
+    {
+      "epoch": 3.227174838483498,
+      "grad_norm": 1.8587815761566162,
+      "learning_rate": 1.4022619758625793e-05,
+      "loss": 0.4039,
+      "step": 32220
+    },
+    {
+      "epoch": 3.2281764912104975,
+      "grad_norm": 2.4933252334594727,
+      "learning_rate": 1.4008458281982879e-05,
+      "loss": 0.4714,
+      "step": 32230
+    },
+    {
+      "epoch": 3.2291781439374967,
+      "grad_norm": 1.8548504114151,
+      "learning_rate": 1.399430117603896e-05,
+      "loss": 0.4457,
+      "step": 32240
+    },
+    {
+      "epoch": 3.2301797966644963,
+      "grad_norm": 1.866041660308838,
+      "learning_rate": 1.3980148446423483e-05,
+      "loss": 0.4801,
+      "step": 32250
+    },
+    {
+      "epoch": 3.231181449391496,
+      "grad_norm": 2.160381317138672,
+      "learning_rate": 1.3966000098764175e-05,
+      "loss": 0.4553,
+      "step": 32260
+    },
+    {
+      "epoch": 3.2321831021184955,
+      "grad_norm": 2.0052473545074463,
+      "learning_rate": 1.3951856138687009e-05,
+      "loss": 0.4554,
+      "step": 32270
+    },
+    {
+      "epoch": 3.233184754845495,
+      "grad_norm": 2.251857280731201,
+      "learning_rate": 1.3937716571816218e-05,
+      "loss": 0.3904,
+      "step": 32280
+    },
+    {
+      "epoch": 3.2341864075724946,
+      "grad_norm": 2.450507164001465,
+      "learning_rate": 1.3923581403774294e-05,
+      "loss": 0.4219,
+      "step": 32290
+    },
+    {
+      "epoch": 3.2351880602994942,
+      "grad_norm": 2.243173599243164,
+      "learning_rate": 1.3909450640181965e-05,
+      "loss": 0.4448,
+      "step": 32300
+    },
+    {
+      "epoch": 3.236189713026494,
+      "grad_norm": 2.0648858547210693,
+      "learning_rate": 1.3895324286658224e-05,
+      "loss": 0.471,
+      "step": 32310
+    },
+    {
+      "epoch": 3.2371913657534934,
+      "grad_norm": 2.1174683570861816,
+      "learning_rate": 1.38812023488203e-05,
+      "loss": 0.4278,
+      "step": 32320
+    },
+    {
+      "epoch": 3.238193018480493,
+      "grad_norm": 2.3641650676727295,
+      "learning_rate": 1.3867084832283667e-05,
+      "loss": 0.4653,
+      "step": 32330
+    },
+    {
+      "epoch": 3.239194671207492,
+      "grad_norm": 1.988904356956482,
+      "learning_rate": 1.3852971742662047e-05,
+      "loss": 0.4166,
+      "step": 32340
+    },
+    {
+      "epoch": 3.2401963239344918,
+      "grad_norm": 2.140962839126587,
+      "learning_rate": 1.3838863085567395e-05,
+      "loss": 0.4228,
+      "step": 32350
+    },
+    {
+      "epoch": 3.2411979766614913,
+      "grad_norm": 2.70149564743042,
+      "learning_rate": 1.3824758866609897e-05,
+      "loss": 0.393,
+      "step": 32360
+    },
+    {
+      "epoch": 3.242199629388491,
+      "grad_norm": 1.729416012763977,
+      "learning_rate": 1.3810659091398012e-05,
+      "loss": 0.5004,
+      "step": 32370
+    },
+    {
+      "epoch": 3.2432012821154905,
+      "grad_norm": 1.8394402265548706,
+      "learning_rate": 1.3796563765538376e-05,
+      "loss": 0.4351,
+      "step": 32380
+    },
+    {
+      "epoch": 3.24420293484249,
+      "grad_norm": 2.235522985458374,
+      "learning_rate": 1.378247289463589e-05,
+      "loss": 0.3838,
+      "step": 32390
+    },
+    {
+      "epoch": 3.2452045875694897,
+      "grad_norm": 1.8229221105575562,
+      "learning_rate": 1.376838648429367e-05,
+      "loss": 0.3934,
+      "step": 32400
+    },
+    {
+      "epoch": 3.2462062402964893,
+      "grad_norm": 1.658739686012268,
+      "learning_rate": 1.3754304540113089e-05,
+      "loss": 0.3869,
+      "step": 32410
+    },
+    {
+      "epoch": 3.247207893023489,
+      "grad_norm": 2.5649220943450928,
+      "learning_rate": 1.3740227067693697e-05,
+      "loss": 0.4916,
+      "step": 32420
+    },
+    {
+      "epoch": 3.2482095457504885,
+      "grad_norm": 1.8066800832748413,
+      "learning_rate": 1.3726154072633285e-05,
+      "loss": 0.456,
+      "step": 32430
+    },
+    {
+      "epoch": 3.2492111984774876,
+      "grad_norm": 2.1143670082092285,
+      "learning_rate": 1.37120855605279e-05,
+      "loss": 0.4308,
+      "step": 32440
+    },
+    {
+      "epoch": 3.250212851204487,
+      "grad_norm": 2.2256264686584473,
+      "learning_rate": 1.3698021536971733e-05,
+      "loss": 0.4165,
+      "step": 32450
+    },
+    {
+      "epoch": 3.251214503931487,
+      "grad_norm": 1.9437214136123657,
+      "learning_rate": 1.3683962007557247e-05,
+      "loss": 0.447,
+      "step": 32460
+    },
+    {
+      "epoch": 3.2522161566584864,
+      "grad_norm": 2.0992989540100098,
+      "learning_rate": 1.3669906977875113e-05,
+      "loss": 0.3604,
+      "step": 32470
+    },
+    {
+      "epoch": 3.253217809385486,
+      "grad_norm": 2.455745220184326,
+      "learning_rate": 1.3655856453514202e-05,
+      "loss": 0.3967,
+      "step": 32480
+    },
+    {
+      "epoch": 3.2542194621124856,
+      "grad_norm": 1.914459228515625,
+      "learning_rate": 1.3641810440061564e-05,
+      "loss": 0.4463,
+      "step": 32490
+    },
+    {
+      "epoch": 3.255221114839485,
+      "grad_norm": 2.0907294750213623,
+      "learning_rate": 1.3627768943102515e-05,
+      "loss": 0.448,
+      "step": 32500
+    },
+    {
+      "epoch": 3.2562227675664848,
+      "grad_norm": 2.3785226345062256,
+      "learning_rate": 1.3613731968220539e-05,
+      "loss": 0.4895,
+      "step": 32510
+    },
+    {
+      "epoch": 3.2572244202934844,
+      "grad_norm": 2.042774200439453,
+      "learning_rate": 1.3599699520997306e-05,
+      "loss": 0.4133,
+      "step": 32520
+    },
+    {
+      "epoch": 3.258226073020484,
+      "grad_norm": 2.2214324474334717,
+      "learning_rate": 1.3585671607012735e-05,
+      "loss": 0.3817,
+      "step": 32530
+    },
+    {
+      "epoch": 3.2592277257474835,
+      "grad_norm": 2.654376983642578,
+      "learning_rate": 1.3571648231844897e-05,
+      "loss": 0.4512,
+      "step": 32540
+    },
+    {
+      "epoch": 3.2602293784744827,
+      "grad_norm": 2.0407917499542236,
+      "learning_rate": 1.3557629401070094e-05,
+      "loss": 0.3947,
+      "step": 32550
+    },
+    {
+      "epoch": 3.2612310312014823,
+      "grad_norm": 1.8769006729125977,
+      "learning_rate": 1.3543615120262764e-05,
+      "loss": 0.4184,
+      "step": 32560
+    },
+    {
+      "epoch": 3.262232683928482,
+      "grad_norm": 1.9162296056747437,
+      "learning_rate": 1.3529605394995613e-05,
+      "loss": 0.3897,
+      "step": 32570
+    },
+    {
+      "epoch": 3.2632343366554815,
+      "grad_norm": 2.0198333263397217,
+      "learning_rate": 1.3515600230839478e-05,
+      "loss": 0.4421,
+      "step": 32580
+    },
+    {
+      "epoch": 3.264235989382481,
+      "grad_norm": 2.189434289932251,
+      "learning_rate": 1.350159963336341e-05,
+      "loss": 0.4541,
+      "step": 32590
+    },
+    {
+      "epoch": 3.2652376421094806,
+      "grad_norm": 1.9533107280731201,
+      "learning_rate": 1.3487603608134627e-05,
+      "loss": 0.4527,
+      "step": 32600
+    },
+    {
+      "epoch": 3.2662392948364802,
+      "grad_norm": 2.1534202098846436,
+      "learning_rate": 1.347361216071854e-05,
+      "loss": 0.4493,
+      "step": 32610
+    },
+    {
+      "epoch": 3.26724094756348,
+      "grad_norm": 1.7690802812576294,
+      "learning_rate": 1.345962529667874e-05,
+      "loss": 0.5048,
+      "step": 32620
+    },
+    {
+      "epoch": 3.2682426002904794,
+      "grad_norm": 2.1077520847320557,
+      "learning_rate": 1.3445643021576987e-05,
+      "loss": 0.43,
+      "step": 32630
+    },
+    {
+      "epoch": 3.269244253017479,
+      "grad_norm": 2.1861956119537354,
+      "learning_rate": 1.3431665340973223e-05,
+      "loss": 0.4492,
+      "step": 32640
+    },
+    {
+      "epoch": 3.2702459057444786,
+      "grad_norm": 2.0920753479003906,
+      "learning_rate": 1.3417692260425564e-05,
+      "loss": 0.4196,
+      "step": 32650
+    },
+    {
+      "epoch": 3.2712475584714777,
+      "grad_norm": 2.3200623989105225,
+      "learning_rate": 1.3403723785490286e-05,
+      "loss": 0.4897,
+      "step": 32660
+    },
+    {
+      "epoch": 3.2722492111984773,
+      "grad_norm": 2.933948516845703,
+      "learning_rate": 1.338975992172185e-05,
+      "loss": 0.4299,
+      "step": 32670
+    },
+    {
+      "epoch": 3.273250863925477,
+      "grad_norm": 2.2343332767486572,
+      "learning_rate": 1.3375800674672872e-05,
+      "loss": 0.3738,
+      "step": 32680
+    },
+    {
+      "epoch": 3.2742525166524765,
+      "grad_norm": 2.4421591758728027,
+      "learning_rate": 1.3361846049894127e-05,
+      "loss": 0.4757,
+      "step": 32690
+    },
+    {
+      "epoch": 3.275254169379476,
+      "grad_norm": 1.903268575668335,
+      "learning_rate": 1.3347896052934591e-05,
+      "loss": 0.3908,
+      "step": 32700
+    },
+    {
+      "epoch": 3.2762558221064757,
+      "grad_norm": 1.6958303451538086,
+      "learning_rate": 1.3333950689341335e-05,
+      "loss": 0.4063,
+      "step": 32710
+    },
+    {
+      "epoch": 3.2772574748334753,
+      "grad_norm": 2.532978057861328,
+      "learning_rate": 1.3320009964659624e-05,
+      "loss": 0.4859,
+      "step": 32720
+    },
+    {
+      "epoch": 3.278259127560475,
+      "grad_norm": 2.2470481395721436,
+      "learning_rate": 1.3306073884432907e-05,
+      "loss": 0.4377,
+      "step": 32730
+    },
+    {
+      "epoch": 3.2792607802874745,
+      "grad_norm": 2.7446441650390625,
+      "learning_rate": 1.3292142454202725e-05,
+      "loss": 0.4472,
+      "step": 32740
+    },
+    {
+      "epoch": 3.280262433014474,
+      "grad_norm": 2.225325107574463,
+      "learning_rate": 1.327821567950881e-05,
+      "loss": 0.4215,
+      "step": 32750
+    },
+    {
+      "epoch": 3.2812640857414737,
+      "grad_norm": 2.3718440532684326,
+      "learning_rate": 1.326429356588902e-05,
+      "loss": 0.4337,
+      "step": 32760
+    },
+    {
+      "epoch": 3.282265738468473,
+      "grad_norm": 1.878641963005066,
+      "learning_rate": 1.3250376118879408e-05,
+      "loss": 0.4516,
+      "step": 32770
+    },
+    {
+      "epoch": 3.2832673911954724,
+      "grad_norm": 2.452392578125,
+      "learning_rate": 1.32364633440141e-05,
+      "loss": 0.4819,
+      "step": 32780
+    },
+    {
+      "epoch": 3.284269043922472,
+      "grad_norm": 3.089430332183838,
+      "learning_rate": 1.3222555246825407e-05,
+      "loss": 0.4242,
+      "step": 32790
+    },
+    {
+      "epoch": 3.2852706966494716,
+      "grad_norm": 1.9483718872070312,
+      "learning_rate": 1.3208651832843796e-05,
+      "loss": 0.4246,
+      "step": 32800
+    },
+    {
+      "epoch": 3.286272349376471,
+      "grad_norm": 2.125678539276123,
+      "learning_rate": 1.3194753107597824e-05,
+      "loss": 0.4367,
+      "step": 32810
+    },
+    {
+      "epoch": 3.2872740021034708,
+      "grad_norm": 2.1187121868133545,
+      "learning_rate": 1.3180859076614205e-05,
+      "loss": 0.4101,
+      "step": 32820
+    },
+    {
+      "epoch": 3.2882756548304704,
+      "grad_norm": 1.9084290266036987,
+      "learning_rate": 1.3166969745417813e-05,
+      "loss": 0.3884,
+      "step": 32830
+    },
+    {
+      "epoch": 3.28927730755747,
+      "grad_norm": 2.143211603164673,
+      "learning_rate": 1.3153085119531624e-05,
+      "loss": 0.4027,
+      "step": 32840
+    },
+    {
+      "epoch": 3.2902789602844695,
+      "grad_norm": 1.8853137493133545,
+      "learning_rate": 1.3139205204476726e-05,
+      "loss": 0.4311,
+      "step": 32850
+    },
+    {
+      "epoch": 3.2912806130114687,
+      "grad_norm": 2.1946933269500732,
+      "learning_rate": 1.3125330005772385e-05,
+      "loss": 0.4407,
+      "step": 32860
+    },
+    {
+      "epoch": 3.2922822657384687,
+      "grad_norm": 2.14318585395813,
+      "learning_rate": 1.3111459528935949e-05,
+      "loss": 0.4017,
+      "step": 32870
+    },
+    {
+      "epoch": 3.293283918465468,
+      "grad_norm": 2.319744110107422,
+      "learning_rate": 1.3097593779482914e-05,
+      "loss": 0.4307,
+      "step": 32880
+    },
+    {
+      "epoch": 3.2942855711924675,
+      "grad_norm": 2.9636361598968506,
+      "learning_rate": 1.3083732762926859e-05,
+      "loss": 0.4675,
+      "step": 32890
+    },
+    {
+      "epoch": 3.295287223919467,
+      "grad_norm": 2.6737558841705322,
+      "learning_rate": 1.306987648477953e-05,
+      "loss": 0.5051,
+      "step": 32900
+    },
+    {
+      "epoch": 3.2962888766464666,
+      "grad_norm": 2.0904157161712646,
+      "learning_rate": 1.3056024950550772e-05,
+      "loss": 0.4611,
+      "step": 32910
+    },
+    {
+      "epoch": 3.2972905293734662,
+      "grad_norm": 2.091607093811035,
+      "learning_rate": 1.3042178165748507e-05,
+      "loss": 0.3842,
+      "step": 32920
+    },
+    {
+      "epoch": 3.298292182100466,
+      "grad_norm": 2.5052433013916016,
+      "learning_rate": 1.3028336135878822e-05,
+      "loss": 0.4381,
+      "step": 32930
+    },
+    {
+      "epoch": 3.2992938348274654,
+      "grad_norm": 1.78681480884552,
+      "learning_rate": 1.3014498866445887e-05,
+      "loss": 0.4373,
+      "step": 32940
+    },
+    {
+      "epoch": 3.300295487554465,
+      "grad_norm": 1.7345925569534302,
+      "learning_rate": 1.3000666362951979e-05,
+      "loss": 0.4163,
+      "step": 32950
+    },
+    {
+      "epoch": 3.3012971402814646,
+      "grad_norm": 2.144805431365967,
+      "learning_rate": 1.298683863089748e-05,
+      "loss": 0.4375,
+      "step": 32960
+    },
+    {
+      "epoch": 3.3022987930084637,
+      "grad_norm": 2.5673115253448486,
+      "learning_rate": 1.2973015675780883e-05,
+      "loss": 0.403,
+      "step": 32970
+    },
+    {
+      "epoch": 3.3033004457354633,
+      "grad_norm": 1.9869693517684937,
+      "learning_rate": 1.2959197503098775e-05,
+      "loss": 0.4217,
+      "step": 32980
+    },
+    {
+      "epoch": 3.304302098462463,
+      "grad_norm": 2.347975969314575,
+      "learning_rate": 1.2945384118345838e-05,
+      "loss": 0.4225,
+      "step": 32990
+    },
+    {
+      "epoch": 3.3053037511894625,
+      "grad_norm": 2.203904390335083,
+      "learning_rate": 1.2931575527014855e-05,
+      "loss": 0.457,
+      "step": 33000
+    },
+    {
+      "epoch": 3.306305403916462,
+      "grad_norm": 2.366241216659546,
+      "learning_rate": 1.2917771734596707e-05,
+      "loss": 0.4346,
+      "step": 33010
+    },
+    {
+      "epoch": 3.3073070566434617,
+      "grad_norm": 2.1328248977661133,
+      "learning_rate": 1.2903972746580361e-05,
+      "loss": 0.4801,
+      "step": 33020
+    },
+    {
+      "epoch": 3.3083087093704613,
+      "grad_norm": 1.9029438495635986,
+      "learning_rate": 1.2890178568452871e-05,
+      "loss": 0.4414,
+      "step": 33030
+    },
+    {
+      "epoch": 3.309310362097461,
+      "grad_norm": 1.8686634302139282,
+      "learning_rate": 1.2876389205699386e-05,
+      "loss": 0.4772,
+      "step": 33040
+    },
+    {
+      "epoch": 3.3103120148244605,
+      "grad_norm": 2.073631525039673,
+      "learning_rate": 1.2862604663803118e-05,
+      "loss": 0.544,
+      "step": 33050
+    },
+    {
+      "epoch": 3.31131366755146,
+      "grad_norm": 1.7073830366134644,
+      "learning_rate": 1.2848824948245423e-05,
+      "loss": 0.4515,
+      "step": 33060
+    },
+    {
+      "epoch": 3.3123153202784597,
+      "grad_norm": 2.569355010986328,
+      "learning_rate": 1.2835050064505655e-05,
+      "loss": 0.4384,
+      "step": 33070
+    },
+    {
+      "epoch": 3.313316973005459,
+      "grad_norm": 1.8745815753936768,
+      "learning_rate": 1.2821280018061305e-05,
+      "loss": 0.4025,
+      "step": 33080
+    },
+    {
+      "epoch": 3.3143186257324584,
+      "grad_norm": 2.1308634281158447,
+      "learning_rate": 1.280751481438791e-05,
+      "loss": 0.4674,
+      "step": 33090
+    },
+    {
+      "epoch": 3.315320278459458,
+      "grad_norm": 2.4393298625946045,
+      "learning_rate": 1.2793754458959109e-05,
+      "loss": 0.4237,
+      "step": 33100
+    },
+    {
+      "epoch": 3.3163219311864576,
+      "grad_norm": 1.8815350532531738,
+      "learning_rate": 1.2779998957246586e-05,
+      "loss": 0.502,
+      "step": 33110
+    },
+    {
+      "epoch": 3.317323583913457,
+      "grad_norm": 2.3304734230041504,
+      "learning_rate": 1.2766248314720102e-05,
+      "loss": 0.4362,
+      "step": 33120
+    },
+    {
+      "epoch": 3.3183252366404568,
+      "grad_norm": 2.8277807235717773,
+      "learning_rate": 1.2752502536847516e-05,
+      "loss": 0.4311,
+      "step": 33130
+    },
+    {
+      "epoch": 3.3193268893674563,
+      "grad_norm": 1.8867182731628418,
+      "learning_rate": 1.2738761629094703e-05,
+      "loss": 0.403,
+      "step": 33140
+    },
+    {
+      "epoch": 3.320328542094456,
+      "grad_norm": 2.2600784301757812,
+      "learning_rate": 1.2725025596925622e-05,
+      "loss": 0.4428,
+      "step": 33150
+    },
+    {
+      "epoch": 3.3213301948214555,
+      "grad_norm": 2.057344436645508,
+      "learning_rate": 1.2711294445802319e-05,
+      "loss": 0.3904,
+      "step": 33160
+    },
+    {
+      "epoch": 3.322331847548455,
+      "grad_norm": 2.2911007404327393,
+      "learning_rate": 1.2697568181184872e-05,
+      "loss": 0.4301,
+      "step": 33170
+    },
+    {
+      "epoch": 3.3233335002754547,
+      "grad_norm": 2.3054580688476562,
+      "learning_rate": 1.2683846808531396e-05,
+      "loss": 0.4788,
+      "step": 33180
+    },
+    {
+      "epoch": 3.324335153002454,
+      "grad_norm": 2.0127499103546143,
+      "learning_rate": 1.2670130333298113e-05,
+      "loss": 0.4351,
+      "step": 33190
+    },
+    {
+      "epoch": 3.3253368057294534,
+      "grad_norm": 1.7581689357757568,
+      "learning_rate": 1.2656418760939267e-05,
+      "loss": 0.3516,
+      "step": 33200
+    },
+    {
+      "epoch": 3.326338458456453,
+      "grad_norm": 2.3614418506622314,
+      "learning_rate": 1.2642712096907145e-05,
+      "loss": 0.4669,
+      "step": 33210
+    },
+    {
+      "epoch": 3.3273401111834526,
+      "grad_norm": 2.0278165340423584,
+      "learning_rate": 1.2629010346652082e-05,
+      "loss": 0.3568,
+      "step": 33220
+    },
+    {
+      "epoch": 3.3283417639104522,
+      "grad_norm": 2.4216668605804443,
+      "learning_rate": 1.2615313515622496e-05,
+      "loss": 0.417,
+      "step": 33230
+    },
+    {
+      "epoch": 3.329343416637452,
+      "grad_norm": 2.564865827560425,
+      "learning_rate": 1.260162160926482e-05,
+      "loss": 0.418,
+      "step": 33240
+    },
+    {
+      "epoch": 3.3303450693644514,
+      "grad_norm": 1.9982346296310425,
+      "learning_rate": 1.2587934633023505e-05,
+      "loss": 0.3987,
+      "step": 33250
+    },
+    {
+      "epoch": 3.331346722091451,
+      "grad_norm": 1.943405032157898,
+      "learning_rate": 1.2574252592341095e-05,
+      "loss": 0.4683,
+      "step": 33260
+    },
+    {
+      "epoch": 3.3323483748184506,
+      "grad_norm": 2.346015453338623,
+      "learning_rate": 1.2560575492658145e-05,
+      "loss": 0.4025,
+      "step": 33270
+    },
+    {
+      "epoch": 3.33335002754545,
+      "grad_norm": 1.8255881071090698,
+      "learning_rate": 1.2546903339413214e-05,
+      "loss": 0.4333,
+      "step": 33280
+    },
+    {
+      "epoch": 3.3343516802724498,
+      "grad_norm": 2.3527069091796875,
+      "learning_rate": 1.2533236138042956e-05,
+      "loss": 0.4259,
+      "step": 33290
+    },
+    {
+      "epoch": 3.335353332999449,
+      "grad_norm": 1.7683374881744385,
+      "learning_rate": 1.2519573893982014e-05,
+      "loss": 0.3926,
+      "step": 33300
+    },
+    {
+      "epoch": 3.3363549857264485,
+      "grad_norm": 1.8756691217422485,
+      "learning_rate": 1.2505916612663071e-05,
+      "loss": 0.4162,
+      "step": 33310
+    },
+    {
+      "epoch": 3.337356638453448,
+      "grad_norm": 2.328777313232422,
+      "learning_rate": 1.2492264299516837e-05,
+      "loss": 0.4315,
+      "step": 33320
+    },
+    {
+      "epoch": 3.3383582911804477,
+      "grad_norm": 2.066019058227539,
+      "learning_rate": 1.2478616959972044e-05,
+      "loss": 0.3843,
+      "step": 33330
+    },
+    {
+      "epoch": 3.3393599439074473,
+      "grad_norm": 2.2497713565826416,
+      "learning_rate": 1.2464974599455451e-05,
+      "loss": 0.4682,
+      "step": 33340
+    },
+    {
+      "epoch": 3.340361596634447,
+      "grad_norm": 2.539682149887085,
+      "learning_rate": 1.245133722339183e-05,
+      "loss": 0.4342,
+      "step": 33350
+    },
+    {
+      "epoch": 3.3413632493614465,
+      "grad_norm": 2.3787872791290283,
+      "learning_rate": 1.2437704837203972e-05,
+      "loss": 0.3957,
+      "step": 33360
+    },
+    {
+      "epoch": 3.342364902088446,
+      "grad_norm": 2.2526967525482178,
+      "learning_rate": 1.2424077446312696e-05,
+      "loss": 0.4469,
+      "step": 33370
+    },
+    {
+      "epoch": 3.3433665548154456,
+      "grad_norm": 2.094153642654419,
+      "learning_rate": 1.2410455056136818e-05,
+      "loss": 0.4025,
+      "step": 33380
+    },
+    {
+      "epoch": 3.344368207542445,
+      "grad_norm": 1.881780982017517,
+      "learning_rate": 1.2396837672093175e-05,
+      "loss": 0.4512,
+      "step": 33390
+    },
+    {
+      "epoch": 3.345369860269445,
+      "grad_norm": 2.2013063430786133,
+      "learning_rate": 1.2383225299596616e-05,
+      "loss": 0.4346,
+      "step": 33400
+    },
+    {
+      "epoch": 3.346371512996444,
+      "grad_norm": 1.754610300064087,
+      "learning_rate": 1.2369617944059983e-05,
+      "loss": 0.3997,
+      "step": 33410
+    },
+    {
+      "epoch": 3.3473731657234436,
+      "grad_norm": 2.1179208755493164,
+      "learning_rate": 1.2356015610894139e-05,
+      "loss": 0.4434,
+      "step": 33420
+    },
+    {
+      "epoch": 3.348374818450443,
+      "grad_norm": 1.6774661540985107,
+      "learning_rate": 1.2342418305507943e-05,
+      "loss": 0.4495,
+      "step": 33430
+    },
+    {
+      "epoch": 3.3493764711774427,
+      "grad_norm": 2.1426808834075928,
+      "learning_rate": 1.2328826033308252e-05,
+      "loss": 0.4384,
+      "step": 33440
+    },
+    {
+      "epoch": 3.3503781239044423,
+      "grad_norm": 1.60723876953125,
+      "learning_rate": 1.2315238799699922e-05,
+      "loss": 0.4383,
+      "step": 33450
+    },
+    {
+      "epoch": 3.351379776631442,
+      "grad_norm": 2.91386079788208,
+      "learning_rate": 1.2301656610085827e-05,
+      "loss": 0.4487,
+      "step": 33460
+    },
+    {
+      "epoch": 3.3523814293584415,
+      "grad_norm": 2.210634708404541,
+      "learning_rate": 1.2288079469866795e-05,
+      "loss": 0.4915,
+      "step": 33470
+    },
+    {
+      "epoch": 3.353383082085441,
+      "grad_norm": 2.1685736179351807,
+      "learning_rate": 1.2274507384441666e-05,
+      "loss": 0.4122,
+      "step": 33480
+    },
+    {
+      "epoch": 3.3543847348124407,
+      "grad_norm": 1.9470291137695312,
+      "learning_rate": 1.2260940359207299e-05,
+      "loss": 0.446,
+      "step": 33490
+    },
+    {
+      "epoch": 3.35538638753944,
+      "grad_norm": 1.978200912475586,
+      "learning_rate": 1.2247378399558488e-05,
+      "loss": 0.4294,
+      "step": 33500
+    },
+    {
+      "epoch": 3.3563880402664394,
+      "grad_norm": 1.9848915338516235,
+      "learning_rate": 1.2233821510888033e-05,
+      "loss": 0.5088,
+      "step": 33510
+    },
+    {
+      "epoch": 3.357389692993439,
+      "grad_norm": 1.9251346588134766,
+      "learning_rate": 1.2220269698586744e-05,
+      "loss": 0.5011,
+      "step": 33520
+    },
+    {
+      "epoch": 3.3583913457204386,
+      "grad_norm": 2.3554296493530273,
+      "learning_rate": 1.2206722968043396e-05,
+      "loss": 0.4617,
+      "step": 33530
+    },
+    {
+      "epoch": 3.359392998447438,
+      "grad_norm": 2.23020339012146,
+      "learning_rate": 1.2193181324644712e-05,
+      "loss": 0.4251,
+      "step": 33540
+    },
+    {
+      "epoch": 3.360394651174438,
+      "grad_norm": 1.555371642112732,
+      "learning_rate": 1.2179644773775422e-05,
+      "loss": 0.4204,
+      "step": 33550
+    },
+    {
+      "epoch": 3.3613963039014374,
+      "grad_norm": 3.28185772895813,
+      "learning_rate": 1.216611332081826e-05,
+      "loss": 0.4155,
+      "step": 33560
+    },
+    {
+      "epoch": 3.362397956628437,
+      "grad_norm": 2.456796884536743,
+      "learning_rate": 1.2152586971153865e-05,
+      "loss": 0.4124,
+      "step": 33570
+    },
+    {
+      "epoch": 3.3633996093554366,
+      "grad_norm": 3.379307270050049,
+      "learning_rate": 1.2139065730160884e-05,
+      "loss": 0.4497,
+      "step": 33580
+    },
+    {
+      "epoch": 3.364401262082436,
+      "grad_norm": 2.1304967403411865,
+      "learning_rate": 1.2125549603215952e-05,
+      "loss": 0.4305,
+      "step": 33590
+    },
+    {
+      "epoch": 3.3654029148094358,
+      "grad_norm": 2.1245455741882324,
+      "learning_rate": 1.2112038595693651e-05,
+      "loss": 0.4025,
+      "step": 33600
+    },
+    {
+      "epoch": 3.366404567536435,
+      "grad_norm": 1.967885971069336,
+      "learning_rate": 1.209853271296649e-05,
+      "loss": 0.4145,
+      "step": 33610
+    },
+    {
+      "epoch": 3.3674062202634345,
+      "grad_norm": 1.7838064432144165,
+      "learning_rate": 1.208503196040501e-05,
+      "loss": 0.3829,
+      "step": 33620
+    },
+    {
+      "epoch": 3.368407872990434,
+      "grad_norm": 2.147975444793701,
+      "learning_rate": 1.2071536343377668e-05,
+      "loss": 0.4521,
+      "step": 33630
+    },
+    {
+      "epoch": 3.3694095257174337,
+      "grad_norm": 2.331613063812256,
+      "learning_rate": 1.2058045867250884e-05,
+      "loss": 0.4155,
+      "step": 33640
+    },
+    {
+      "epoch": 3.3704111784444333,
+      "grad_norm": 2.6464388370513916,
+      "learning_rate": 1.2044560537389044e-05,
+      "loss": 0.4248,
+      "step": 33650
+    },
+    {
+      "epoch": 3.371412831171433,
+      "grad_norm": 1.795331597328186,
+      "learning_rate": 1.2031080359154476e-05,
+      "loss": 0.404,
+      "step": 33660
+    },
+    {
+      "epoch": 3.3724144838984325,
+      "grad_norm": 2.6335437297821045,
+      "learning_rate": 1.2017605337907472e-05,
+      "loss": 0.4048,
+      "step": 33670
+    },
+    {
+      "epoch": 3.373416136625432,
+      "grad_norm": 1.8416485786437988,
+      "learning_rate": 1.2004135479006263e-05,
+      "loss": 0.4342,
+      "step": 33680
+    },
+    {
+      "epoch": 3.3744177893524316,
+      "grad_norm": 2.2963712215423584,
+      "learning_rate": 1.1990670787807031e-05,
+      "loss": 0.4062,
+      "step": 33690
+    },
+    {
+      "epoch": 3.3754194420794312,
+      "grad_norm": 1.744742751121521,
+      "learning_rate": 1.1977211269663904e-05,
+      "loss": 0.3897,
+      "step": 33700
+    },
+    {
+      "epoch": 3.376421094806431,
+      "grad_norm": 1.9296424388885498,
+      "learning_rate": 1.1963756929928952e-05,
+      "loss": 0.4246,
+      "step": 33710
+    },
+    {
+      "epoch": 3.37742274753343,
+      "grad_norm": 2.5108537673950195,
+      "learning_rate": 1.1950307773952185e-05,
+      "loss": 0.4156,
+      "step": 33720
+    },
+    {
+      "epoch": 3.3784244002604296,
+      "grad_norm": 2.305176258087158,
+      "learning_rate": 1.1936863807081555e-05,
+      "loss": 0.4031,
+      "step": 33730
+    },
+    {
+      "epoch": 3.379426052987429,
+      "grad_norm": 2.122509002685547,
+      "learning_rate": 1.1923425034662946e-05,
+      "loss": 0.4498,
+      "step": 33740
+    },
+    {
+      "epoch": 3.3804277057144287,
+      "grad_norm": 2.2472517490386963,
+      "learning_rate": 1.1909991462040182e-05,
+      "loss": 0.4545,
+      "step": 33750
+    },
+    {
+      "epoch": 3.3814293584414283,
+      "grad_norm": 2.068202495574951,
+      "learning_rate": 1.1896563094555014e-05,
+      "loss": 0.4709,
+      "step": 33760
+    },
+    {
+      "epoch": 3.382431011168428,
+      "grad_norm": 2.0109431743621826,
+      "learning_rate": 1.1883139937547122e-05,
+      "loss": 0.4377,
+      "step": 33770
+    },
+    {
+      "epoch": 3.3834326638954275,
+      "grad_norm": 2.1727137565612793,
+      "learning_rate": 1.1869721996354124e-05,
+      "loss": 0.4618,
+      "step": 33780
+    },
+    {
+      "epoch": 3.384434316622427,
+      "grad_norm": 3.1651906967163086,
+      "learning_rate": 1.1856309276311553e-05,
+      "loss": 0.3812,
+      "step": 33790
+    },
+    {
+      "epoch": 3.3854359693494267,
+      "grad_norm": 2.4753785133361816,
+      "learning_rate": 1.1842901782752871e-05,
+      "loss": 0.4006,
+      "step": 33800
+    },
+    {
+      "epoch": 3.3864376220764263,
+      "grad_norm": 2.4785983562469482,
+      "learning_rate": 1.1829499521009452e-05,
+      "loss": 0.457,
+      "step": 33810
+    },
+    {
+      "epoch": 3.387439274803426,
+      "grad_norm": 1.973423719406128,
+      "learning_rate": 1.1816102496410627e-05,
+      "loss": 0.4479,
+      "step": 33820
+    },
+    {
+      "epoch": 3.388440927530425,
+      "grad_norm": 2.3075802326202393,
+      "learning_rate": 1.1802710714283588e-05,
+      "loss": 0.4552,
+      "step": 33830
+    },
+    {
+      "epoch": 3.3894425802574246,
+      "grad_norm": 2.3044092655181885,
+      "learning_rate": 1.1789324179953467e-05,
+      "loss": 0.4096,
+      "step": 33840
+    },
+    {
+      "epoch": 3.390444232984424,
+      "grad_norm": 2.105095863342285,
+      "learning_rate": 1.1775942898743347e-05,
+      "loss": 0.4348,
+      "step": 33850
+    },
+    {
+      "epoch": 3.391445885711424,
+      "grad_norm": 2.2408626079559326,
+      "learning_rate": 1.1762566875974157e-05,
+      "loss": 0.4708,
+      "step": 33860
+    },
+    {
+      "epoch": 3.3924475384384234,
+      "grad_norm": 2.347888469696045,
+      "learning_rate": 1.1749196116964764e-05,
+      "loss": 0.4569,
+      "step": 33870
+    },
+    {
+      "epoch": 3.393449191165423,
+      "grad_norm": 2.3559656143188477,
+      "learning_rate": 1.1735830627031966e-05,
+      "loss": 0.4546,
+      "step": 33880
+    },
+    {
+      "epoch": 3.3944508438924226,
+      "grad_norm": 1.7592754364013672,
+      "learning_rate": 1.1722470411490438e-05,
+      "loss": 0.4055,
+      "step": 33890
+    },
+    {
+      "epoch": 3.395452496619422,
+      "grad_norm": 2.887687921524048,
+      "learning_rate": 1.1709115475652751e-05,
+      "loss": 0.4309,
+      "step": 33900
+    },
+    {
+      "epoch": 3.3964541493464218,
+      "grad_norm": 2.2861642837524414,
+      "learning_rate": 1.1695765824829386e-05,
+      "loss": 0.4514,
+      "step": 33910
+    },
+    {
+      "epoch": 3.397455802073421,
+      "grad_norm": 2.557394504547119,
+      "learning_rate": 1.1682421464328745e-05,
+      "loss": 0.4279,
+      "step": 33920
+    },
+    {
+      "epoch": 3.398457454800421,
+      "grad_norm": 2.420703411102295,
+      "learning_rate": 1.1669082399457105e-05,
+      "loss": 0.4231,
+      "step": 33930
+    },
+    {
+      "epoch": 3.39945910752742,
+      "grad_norm": 1.7291672229766846,
+      "learning_rate": 1.1655748635518609e-05,
+      "loss": 0.4588,
+      "step": 33940
+    },
+    {
+      "epoch": 3.4004607602544197,
+      "grad_norm": 2.1979129314422607,
+      "learning_rate": 1.1642420177815352e-05,
+      "loss": 0.4296,
+      "step": 33950
+    },
+    {
+      "epoch": 3.4014624129814193,
+      "grad_norm": 2.324911594390869,
+      "learning_rate": 1.1629097031647293e-05,
+      "loss": 0.488,
+      "step": 33960
+    },
+    {
+      "epoch": 3.402464065708419,
+      "grad_norm": 2.866677761077881,
+      "learning_rate": 1.1615779202312241e-05,
+      "loss": 0.4562,
+      "step": 33970
+    },
+    {
+      "epoch": 3.4034657184354185,
+      "grad_norm": 1.6479003429412842,
+      "learning_rate": 1.1602466695105956e-05,
+      "loss": 0.4266,
+      "step": 33980
+    },
+    {
+      "epoch": 3.404467371162418,
+      "grad_norm": 2.6396446228027344,
+      "learning_rate": 1.158915951532204e-05,
+      "loss": 0.4387,
+      "step": 33990
+    },
+    {
+      "epoch": 3.4054690238894176,
+      "grad_norm": 2.2791261672973633,
+      "learning_rate": 1.1575857668251988e-05,
+      "loss": 0.4231,
+      "step": 34000
+    },
+    {
+      "epoch": 3.4064706766164172,
+      "grad_norm": 1.8278400897979736,
+      "learning_rate": 1.1562561159185174e-05,
+      "loss": 0.4151,
+      "step": 34010
+    },
+    {
+      "epoch": 3.407472329343417,
+      "grad_norm": 2.3506553173065186,
+      "learning_rate": 1.1549269993408846e-05,
+      "loss": 0.4128,
+      "step": 34020
+    },
+    {
+      "epoch": 3.408473982070416,
+      "grad_norm": 2.4189066886901855,
+      "learning_rate": 1.153598417620813e-05,
+      "loss": 0.4881,
+      "step": 34030
+    },
+    {
+      "epoch": 3.4094756347974156,
+      "grad_norm": 2.4470417499542236,
+      "learning_rate": 1.1522703712866029e-05,
+      "loss": 0.4015,
+      "step": 34040
+    },
+    {
+      "epoch": 3.410477287524415,
+      "grad_norm": 1.7285627126693726,
+      "learning_rate": 1.150942860866341e-05,
+      "loss": 0.4424,
+      "step": 34050
+    },
+    {
+      "epoch": 3.4114789402514147,
+      "grad_norm": 2.0729176998138428,
+      "learning_rate": 1.1496158868879015e-05,
+      "loss": 0.4989,
+      "step": 34060
+    },
+    {
+      "epoch": 3.4124805929784143,
+      "grad_norm": 2.230759382247925,
+      "learning_rate": 1.148289449878945e-05,
+      "loss": 0.4757,
+      "step": 34070
+    },
+    {
+      "epoch": 3.413482245705414,
+      "grad_norm": 1.9001071453094482,
+      "learning_rate": 1.1469635503669186e-05,
+      "loss": 0.4198,
+      "step": 34080
+    },
+    {
+      "epoch": 3.4144838984324135,
+      "grad_norm": 2.2974631786346436,
+      "learning_rate": 1.1456381888790554e-05,
+      "loss": 0.4217,
+      "step": 34090
+    },
+    {
+      "epoch": 3.415485551159413,
+      "grad_norm": 1.4676469564437866,
+      "learning_rate": 1.144313365942375e-05,
+      "loss": 0.4494,
+      "step": 34100
+    },
+    {
+      "epoch": 3.4164872038864127,
+      "grad_norm": 2.2653024196624756,
+      "learning_rate": 1.1429890820836829e-05,
+      "loss": 0.4153,
+      "step": 34110
+    },
+    {
+      "epoch": 3.4174888566134123,
+      "grad_norm": 1.8923325538635254,
+      "learning_rate": 1.1416653378295697e-05,
+      "loss": 0.4202,
+      "step": 34120
+    },
+    {
+      "epoch": 3.418490509340412,
+      "grad_norm": 2.0065672397613525,
+      "learning_rate": 1.1403421337064121e-05,
+      "loss": 0.4641,
+      "step": 34130
+    },
+    {
+      "epoch": 3.419492162067411,
+      "grad_norm": 1.9923065900802612,
+      "learning_rate": 1.1390194702403715e-05,
+      "loss": 0.4349,
+      "step": 34140
+    },
+    {
+      "epoch": 3.4204938147944106,
+      "grad_norm": 2.1380667686462402,
+      "learning_rate": 1.1376973479573942e-05,
+      "loss": 0.45,
+      "step": 34150
+    },
+    {
+      "epoch": 3.42149546752141,
+      "grad_norm": 1.5650336742401123,
+      "learning_rate": 1.136375767383212e-05,
+      "loss": 0.4001,
+      "step": 34160
+    },
+    {
+      "epoch": 3.42249712024841,
+      "grad_norm": 1.5955073833465576,
+      "learning_rate": 1.1350547290433399e-05,
+      "loss": 0.4333,
+      "step": 34170
+    },
+    {
+      "epoch": 3.4234987729754094,
+      "grad_norm": 1.952041506767273,
+      "learning_rate": 1.133734233463081e-05,
+      "loss": 0.4634,
+      "step": 34180
+    },
+    {
+      "epoch": 3.424500425702409,
+      "grad_norm": 2.256084680557251,
+      "learning_rate": 1.1324142811675167e-05,
+      "loss": 0.4104,
+      "step": 34190
+    },
+    {
+      "epoch": 3.4255020784294086,
+      "grad_norm": 1.7892006635665894,
+      "learning_rate": 1.1310948726815162e-05,
+      "loss": 0.4574,
+      "step": 34200
+    },
+    {
+      "epoch": 3.426503731156408,
+      "grad_norm": 2.5087509155273438,
+      "learning_rate": 1.1297760085297332e-05,
+      "loss": 0.4494,
+      "step": 34210
+    },
+    {
+      "epoch": 3.4275053838834078,
+      "grad_norm": 1.9253290891647339,
+      "learning_rate": 1.1284576892366037e-05,
+      "loss": 0.4651,
+      "step": 34220
+    },
+    {
+      "epoch": 3.4285070366104073,
+      "grad_norm": 2.0629913806915283,
+      "learning_rate": 1.1271399153263448e-05,
+      "loss": 0.3922,
+      "step": 34230
+    },
+    {
+      "epoch": 3.429508689337407,
+      "grad_norm": 2.1463217735290527,
+      "learning_rate": 1.1258226873229589e-05,
+      "loss": 0.3902,
+      "step": 34240
+    },
+    {
+      "epoch": 3.430510342064406,
+      "grad_norm": 2.2839205265045166,
+      "learning_rate": 1.124506005750234e-05,
+      "loss": 0.4098,
+      "step": 34250
+    },
+    {
+      "epoch": 3.4315119947914057,
+      "grad_norm": 2.252955198287964,
+      "learning_rate": 1.1231898711317352e-05,
+      "loss": 0.441,
+      "step": 34260
+    },
+    {
+      "epoch": 3.4325136475184053,
+      "grad_norm": 2.093442678451538,
+      "learning_rate": 1.1218742839908131e-05,
+      "loss": 0.3993,
+      "step": 34270
+    },
+    {
+      "epoch": 3.433515300245405,
+      "grad_norm": 2.032712459564209,
+      "learning_rate": 1.1205592448506022e-05,
+      "loss": 0.4433,
+      "step": 34280
+    },
+    {
+      "epoch": 3.4345169529724044,
+      "grad_norm": 2.3281428813934326,
+      "learning_rate": 1.1192447542340173e-05,
+      "loss": 0.4208,
+      "step": 34290
+    },
+    {
+      "epoch": 3.435518605699404,
+      "grad_norm": 2.4409408569335938,
+      "learning_rate": 1.1179308126637524e-05,
+      "loss": 0.3541,
+      "step": 34300
+    },
+    {
+      "epoch": 3.4365202584264036,
+      "grad_norm": 2.3380045890808105,
+      "learning_rate": 1.1166174206622887e-05,
+      "loss": 0.4353,
+      "step": 34310
+    },
+    {
+      "epoch": 3.437521911153403,
+      "grad_norm": 1.6690657138824463,
+      "learning_rate": 1.115304578751886e-05,
+      "loss": 0.4292,
+      "step": 34320
+    },
+    {
+      "epoch": 3.438523563880403,
+      "grad_norm": 2.014461040496826,
+      "learning_rate": 1.1139922874545824e-05,
+      "loss": 0.487,
+      "step": 34330
+    },
+    {
+      "epoch": 3.4395252166074024,
+      "grad_norm": 2.823160409927368,
+      "learning_rate": 1.1126805472922034e-05,
+      "loss": 0.4126,
+      "step": 34340
+    },
+    {
+      "epoch": 3.440526869334402,
+      "grad_norm": 2.1313488483428955,
+      "learning_rate": 1.1113693587863507e-05,
+      "loss": 0.4308,
+      "step": 34350
+    },
+    {
+      "epoch": 3.441528522061401,
+      "grad_norm": 1.7940298318862915,
+      "learning_rate": 1.1100587224584092e-05,
+      "loss": 0.4637,
+      "step": 34360
+    },
+    {
+      "epoch": 3.4425301747884007,
+      "grad_norm": 2.9107887744903564,
+      "learning_rate": 1.1087486388295399e-05,
+      "loss": 0.4597,
+      "step": 34370
+    },
+    {
+      "epoch": 3.4435318275154003,
+      "grad_norm": 1.9696946144104004,
+      "learning_rate": 1.10743910842069e-05,
+      "loss": 0.4005,
+      "step": 34380
+    },
+    {
+      "epoch": 3.4445334802424,
+      "grad_norm": 1.943562388420105,
+      "learning_rate": 1.106130131752583e-05,
+      "loss": 0.4398,
+      "step": 34390
+    },
+    {
+      "epoch": 3.4455351329693995,
+      "grad_norm": 2.523973226547241,
+      "learning_rate": 1.1048217093457228e-05,
+      "loss": 0.4519,
+      "step": 34400
+    },
+    {
+      "epoch": 3.446536785696399,
+      "grad_norm": 2.25036883354187,
+      "learning_rate": 1.1035138417203936e-05,
+      "loss": 0.4734,
+      "step": 34410
+    },
+    {
+      "epoch": 3.4475384384233987,
+      "grad_norm": 2.8198471069335938,
+      "learning_rate": 1.1022065293966579e-05,
+      "loss": 0.4328,
+      "step": 34420
+    },
+    {
+      "epoch": 3.4485400911503983,
+      "grad_norm": 1.8240866661071777,
+      "learning_rate": 1.1008997728943584e-05,
+      "loss": 0.4731,
+      "step": 34430
+    },
+    {
+      "epoch": 3.449541743877398,
+      "grad_norm": 2.043656349182129,
+      "learning_rate": 1.0995935727331166e-05,
+      "loss": 0.439,
+      "step": 34440
+    },
+    {
+      "epoch": 3.450543396604397,
+      "grad_norm": 2.4668407440185547,
+      "learning_rate": 1.0982879294323323e-05,
+      "loss": 0.4616,
+      "step": 34450
+    },
+    {
+      "epoch": 3.451545049331397,
+      "grad_norm": 1.7924633026123047,
+      "learning_rate": 1.0969828435111843e-05,
+      "loss": 0.4282,
+      "step": 34460
+    },
+    {
+      "epoch": 3.452546702058396,
+      "grad_norm": 1.8731104135513306,
+      "learning_rate": 1.0956783154886297e-05,
+      "loss": 0.395,
+      "step": 34470
+    },
+    {
+      "epoch": 3.453548354785396,
+      "grad_norm": 2.430614948272705,
+      "learning_rate": 1.0943743458834033e-05,
+      "loss": 0.4489,
+      "step": 34480
+    },
+    {
+      "epoch": 3.4545500075123954,
+      "grad_norm": 2.1472465991973877,
+      "learning_rate": 1.0930709352140183e-05,
+      "loss": 0.4062,
+      "step": 34490
+    },
+    {
+      "epoch": 3.455551660239395,
+      "grad_norm": 2.722684383392334,
+      "learning_rate": 1.0917680839987652e-05,
+      "loss": 0.4683,
+      "step": 34500
+    },
+    {
+      "epoch": 3.4565533129663946,
+      "grad_norm": 2.144862174987793,
+      "learning_rate": 1.0904657927557144e-05,
+      "loss": 0.4497,
+      "step": 34510
+    },
+    {
+      "epoch": 3.457554965693394,
+      "grad_norm": 2.7352027893066406,
+      "learning_rate": 1.0891640620027094e-05,
+      "loss": 0.4247,
+      "step": 34520
+    },
+    {
+      "epoch": 3.4585566184203937,
+      "grad_norm": 2.6462039947509766,
+      "learning_rate": 1.0878628922573727e-05,
+      "loss": 0.4004,
+      "step": 34530
+    },
+    {
+      "epoch": 3.4595582711473933,
+      "grad_norm": 2.273921012878418,
+      "learning_rate": 1.0865622840371074e-05,
+      "loss": 0.4683,
+      "step": 34540
+    },
+    {
+      "epoch": 3.460559923874393,
+      "grad_norm": 1.9296114444732666,
+      "learning_rate": 1.0852622378590866e-05,
+      "loss": 0.4534,
+      "step": 34550
+    },
+    {
+      "epoch": 3.461561576601392,
+      "grad_norm": 2.1889405250549316,
+      "learning_rate": 1.0839627542402644e-05,
+      "loss": 0.3771,
+      "step": 34560
+    },
+    {
+      "epoch": 3.4625632293283917,
+      "grad_norm": 2.109031915664673,
+      "learning_rate": 1.0826638336973696e-05,
+      "loss": 0.4224,
+      "step": 34570
+    },
+    {
+      "epoch": 3.4635648820553913,
+      "grad_norm": 3.0517172813415527,
+      "learning_rate": 1.08136547674691e-05,
+      "loss": 0.4029,
+      "step": 34580
+    },
+    {
+      "epoch": 3.464566534782391,
+      "grad_norm": 2.070493459701538,
+      "learning_rate": 1.0800676839051641e-05,
+      "loss": 0.4262,
+      "step": 34590
+    },
+    {
+      "epoch": 3.4655681875093904,
+      "grad_norm": 2.03619384765625,
+      "learning_rate": 1.0787704556881888e-05,
+      "loss": 0.4141,
+      "step": 34600
+    },
+    {
+      "epoch": 3.46656984023639,
+      "grad_norm": 2.0710926055908203,
+      "learning_rate": 1.0774737926118198e-05,
+      "loss": 0.4238,
+      "step": 34610
+    },
+    {
+      "epoch": 3.4675714929633896,
+      "grad_norm": 2.2858543395996094,
+      "learning_rate": 1.0761776951916616e-05,
+      "loss": 0.3914,
+      "step": 34620
+    },
+    {
+      "epoch": 3.468573145690389,
+      "grad_norm": 2.7241454124450684,
+      "learning_rate": 1.0748821639430973e-05,
+      "loss": 0.4691,
+      "step": 34630
+    },
+    {
+      "epoch": 3.469574798417389,
+      "grad_norm": 2.1155881881713867,
+      "learning_rate": 1.0735871993812863e-05,
+      "loss": 0.4633,
+      "step": 34640
+    },
+    {
+      "epoch": 3.4705764511443884,
+      "grad_norm": 1.7804661989212036,
+      "learning_rate": 1.0722928020211609e-05,
+      "loss": 0.4208,
+      "step": 34650
+    },
+    {
+      "epoch": 3.471578103871388,
+      "grad_norm": 1.557573676109314,
+      "learning_rate": 1.070998972377425e-05,
+      "loss": 0.4351,
+      "step": 34660
+    },
+    {
+      "epoch": 3.472579756598387,
+      "grad_norm": 2.085498094558716,
+      "learning_rate": 1.0697057109645629e-05,
+      "loss": 0.4311,
+      "step": 34670
+    },
+    {
+      "epoch": 3.4735814093253867,
+      "grad_norm": 2.409001111984253,
+      "learning_rate": 1.0684130182968284e-05,
+      "loss": 0.4173,
+      "step": 34680
+    },
+    {
+      "epoch": 3.4745830620523863,
+      "grad_norm": 1.4353440999984741,
+      "learning_rate": 1.0671208948882514e-05,
+      "loss": 0.441,
+      "step": 34690
+    },
+    {
+      "epoch": 3.475584714779386,
+      "grad_norm": 2.4345576763153076,
+      "learning_rate": 1.0658293412526316e-05,
+      "loss": 0.3898,
+      "step": 34700
+    },
+    {
+      "epoch": 3.4765863675063855,
+      "grad_norm": 2.105102777481079,
+      "learning_rate": 1.0645383579035478e-05,
+      "loss": 0.4628,
+      "step": 34710
+    },
+    {
+      "epoch": 3.477588020233385,
+      "grad_norm": 2.4883739948272705,
+      "learning_rate": 1.0632479453543498e-05,
+      "loss": 0.4361,
+      "step": 34720
+    },
+    {
+      "epoch": 3.4785896729603847,
+      "grad_norm": 1.9684547185897827,
+      "learning_rate": 1.0619581041181565e-05,
+      "loss": 0.4194,
+      "step": 34730
+    },
+    {
+      "epoch": 3.4795913256873843,
+      "grad_norm": 1.775962233543396,
+      "learning_rate": 1.0606688347078659e-05,
+      "loss": 0.4438,
+      "step": 34740
+    },
+    {
+      "epoch": 3.480592978414384,
+      "grad_norm": 2.144451856613159,
+      "learning_rate": 1.0593801376361448e-05,
+      "loss": 0.41,
+      "step": 34750
+    },
+    {
+      "epoch": 3.4815946311413835,
+      "grad_norm": 1.9578640460968018,
+      "learning_rate": 1.0580920134154332e-05,
+      "loss": 0.4127,
+      "step": 34760
+    },
+    {
+      "epoch": 3.482596283868383,
+      "grad_norm": 2.3081228733062744,
+      "learning_rate": 1.0568044625579437e-05,
+      "loss": 0.4418,
+      "step": 34770
+    },
+    {
+      "epoch": 3.483597936595382,
+      "grad_norm": 2.8557839393615723,
+      "learning_rate": 1.0555174855756605e-05,
+      "loss": 0.4509,
+      "step": 34780
+    },
+    {
+      "epoch": 3.484599589322382,
+      "grad_norm": 2.0425710678100586,
+      "learning_rate": 1.0542310829803396e-05,
+      "loss": 0.4134,
+      "step": 34790
+    },
+    {
+      "epoch": 3.4856012420493814,
+      "grad_norm": 1.7663625478744507,
+      "learning_rate": 1.052945255283509e-05,
+      "loss": 0.3944,
+      "step": 34800
+    },
+    {
+      "epoch": 3.486602894776381,
+      "grad_norm": 2.232553243637085,
+      "learning_rate": 1.0516600029964677e-05,
+      "loss": 0.4513,
+      "step": 34810
+    },
+    {
+      "epoch": 3.4876045475033806,
+      "grad_norm": 2.3331985473632812,
+      "learning_rate": 1.0503753266302863e-05,
+      "loss": 0.4078,
+      "step": 34820
+    },
+    {
+      "epoch": 3.48860620023038,
+      "grad_norm": 1.9388011693954468,
+      "learning_rate": 1.0490912266958055e-05,
+      "loss": 0.4758,
+      "step": 34830
+    },
+    {
+      "epoch": 3.4896078529573797,
+      "grad_norm": 2.1196694374084473,
+      "learning_rate": 1.047807703703638e-05,
+      "loss": 0.4602,
+      "step": 34840
+    },
+    {
+      "epoch": 3.4906095056843793,
+      "grad_norm": 1.5808969736099243,
+      "learning_rate": 1.0465247581641663e-05,
+      "loss": 0.3994,
+      "step": 34850
+    },
+    {
+      "epoch": 3.491611158411379,
+      "grad_norm": 1.7989665269851685,
+      "learning_rate": 1.0452423905875425e-05,
+      "loss": 0.407,
+      "step": 34860
+    },
+    {
+      "epoch": 3.4926128111383785,
+      "grad_norm": 2.2902822494506836,
+      "learning_rate": 1.0439606014836924e-05,
+      "loss": 0.3768,
+      "step": 34870
+    },
+    {
+      "epoch": 3.493614463865378,
+      "grad_norm": 2.374018430709839,
+      "learning_rate": 1.042679391362307e-05,
+      "loss": 0.4478,
+      "step": 34880
+    },
+    {
+      "epoch": 3.4946161165923773,
+      "grad_norm": 1.874822735786438,
+      "learning_rate": 1.0413987607328497e-05,
+      "loss": 0.451,
+      "step": 34890
+    },
+    {
+      "epoch": 3.495617769319377,
+      "grad_norm": 2.655895233154297,
+      "learning_rate": 1.0401187101045534e-05,
+      "loss": 0.4629,
+      "step": 34900
+    },
+    {
+      "epoch": 3.4966194220463764,
+      "grad_norm": 2.248385429382324,
+      "learning_rate": 1.03883923998642e-05,
+      "loss": 0.4135,
+      "step": 34910
+    },
+    {
+      "epoch": 3.497621074773376,
+      "grad_norm": 2.5209267139434814,
+      "learning_rate": 1.0375603508872203e-05,
+      "loss": 0.4222,
+      "step": 34920
+    },
+    {
+      "epoch": 3.4986227275003756,
+      "grad_norm": 1.8577488660812378,
+      "learning_rate": 1.0362820433154938e-05,
+      "loss": 0.4049,
+      "step": 34930
+    },
+    {
+      "epoch": 3.499624380227375,
+      "grad_norm": 2.4650468826293945,
+      "learning_rate": 1.0350043177795517e-05,
+      "loss": 0.4289,
+      "step": 34940
+    },
+    {
+      "epoch": 3.500626032954375,
+      "grad_norm": 2.639314889907837,
+      "learning_rate": 1.033727174787469e-05,
+      "loss": 0.4502,
+      "step": 34950
+    },
+    {
+      "epoch": 3.5016276856813744,
+      "grad_norm": 2.0178632736206055,
+      "learning_rate": 1.0324506148470917e-05,
+      "loss": 0.4272,
+      "step": 34960
+    },
+    {
+      "epoch": 3.502629338408374,
+      "grad_norm": 1.9113802909851074,
+      "learning_rate": 1.0311746384660346e-05,
+      "loss": 0.4286,
+      "step": 34970
+    },
+    {
+      "epoch": 3.503630991135373,
+      "grad_norm": 2.1833677291870117,
+      "learning_rate": 1.0298992461516802e-05,
+      "loss": 0.501,
+      "step": 34980
+    },
+    {
+      "epoch": 3.504632643862373,
+      "grad_norm": 2.2047367095947266,
+      "learning_rate": 1.028624438411175e-05,
+      "loss": 0.4528,
+      "step": 34990
+    },
+    {
+      "epoch": 3.5056342965893723,
+      "grad_norm": 1.7950270175933838,
+      "learning_rate": 1.0273502157514394e-05,
+      "loss": 0.4207,
+      "step": 35000
+    },
+    {
+      "epoch": 3.506635949316372,
+      "grad_norm": 1.9958217144012451,
+      "learning_rate": 1.0260765786791577e-05,
+      "loss": 0.3915,
+      "step": 35010
+    },
+    {
+      "epoch": 3.5076376020433715,
+      "grad_norm": 2.9710733890533447,
+      "learning_rate": 1.0248035277007783e-05,
+      "loss": 0.4248,
+      "step": 35020
+    },
+    {
+      "epoch": 3.508639254770371,
+      "grad_norm": 1.977878212928772,
+      "learning_rate": 1.0235310633225232e-05,
+      "loss": 0.4082,
+      "step": 35030
+    },
+    {
+      "epoch": 3.5096409074973707,
+      "grad_norm": 2.591295003890991,
+      "learning_rate": 1.0222591860503761e-05,
+      "loss": 0.4369,
+      "step": 35040
+    },
+    {
+      "epoch": 3.5106425602243703,
+      "grad_norm": 1.7341768741607666,
+      "learning_rate": 1.0209878963900904e-05,
+      "loss": 0.3957,
+      "step": 35050
+    },
+    {
+      "epoch": 3.51164421295137,
+      "grad_norm": 2.3429300785064697,
+      "learning_rate": 1.0197171948471812e-05,
+      "loss": 0.4407,
+      "step": 35060
+    },
+    {
+      "epoch": 3.5126458656783695,
+      "grad_norm": 2.3763067722320557,
+      "learning_rate": 1.0184470819269359e-05,
+      "loss": 0.4635,
+      "step": 35070
+    },
+    {
+      "epoch": 3.513647518405369,
+      "grad_norm": 2.520565986633301,
+      "learning_rate": 1.0171775581344044e-05,
+      "loss": 0.438,
+      "step": 35080
+    },
+    {
+      "epoch": 3.514649171132368,
+      "grad_norm": 2.132378578186035,
+      "learning_rate": 1.0159086239744003e-05,
+      "loss": 0.4628,
+      "step": 35090
+    },
+    {
+      "epoch": 3.5156508238593682,
+      "grad_norm": 2.1328794956207275,
+      "learning_rate": 1.0146402799515079e-05,
+      "loss": 0.449,
+      "step": 35100
+    },
+    {
+      "epoch": 3.5166524765863674,
+      "grad_norm": 2.2346129417419434,
+      "learning_rate": 1.0133725265700728e-05,
+      "loss": 0.4675,
+      "step": 35110
+    },
+    {
+      "epoch": 3.517654129313367,
+      "grad_norm": 2.5284063816070557,
+      "learning_rate": 1.0121053643342074e-05,
+      "loss": 0.4691,
+      "step": 35120
+    },
+    {
+      "epoch": 3.5186557820403666,
+      "grad_norm": 2.44215989112854,
+      "learning_rate": 1.0108387937477886e-05,
+      "loss": 0.4336,
+      "step": 35130
+    },
+    {
+      "epoch": 3.519657434767366,
+      "grad_norm": 2.0735301971435547,
+      "learning_rate": 1.0095728153144579e-05,
+      "loss": 0.4179,
+      "step": 35140
+    },
+    {
+      "epoch": 3.5206590874943657,
+      "grad_norm": 1.7370909452438354,
+      "learning_rate": 1.0083074295376219e-05,
+      "loss": 0.4225,
+      "step": 35150
+    },
+    {
+      "epoch": 3.5216607402213653,
+      "grad_norm": 2.455852746963501,
+      "learning_rate": 1.0070426369204511e-05,
+      "loss": 0.4091,
+      "step": 35160
+    },
+    {
+      "epoch": 3.522662392948365,
+      "grad_norm": 2.051345109939575,
+      "learning_rate": 1.0057784379658797e-05,
+      "loss": 0.4919,
+      "step": 35170
+    },
+    {
+      "epoch": 3.5236640456753645,
+      "grad_norm": 2.2854480743408203,
+      "learning_rate": 1.0045148331766069e-05,
+      "loss": 0.4539,
+      "step": 35180
+    },
+    {
+      "epoch": 3.524665698402364,
+      "grad_norm": 2.3247320652008057,
+      "learning_rate": 1.0032518230550947e-05,
+      "loss": 0.4547,
+      "step": 35190
+    },
+    {
+      "epoch": 3.5256673511293632,
+      "grad_norm": 2.1312015056610107,
+      "learning_rate": 1.0019894081035694e-05,
+      "loss": 0.4494,
+      "step": 35200
+    },
+    {
+      "epoch": 3.526669003856363,
+      "grad_norm": 1.9080195426940918,
+      "learning_rate": 1.00072758882402e-05,
+      "loss": 0.4012,
+      "step": 35210
+    },
+    {
+      "epoch": 3.5276706565833624,
+      "grad_norm": 1.9371165037155151,
+      "learning_rate": 9.994663657181991e-06,
+      "loss": 0.4325,
+      "step": 35220
+    },
+    {
+      "epoch": 3.528672309310362,
+      "grad_norm": 2.2770767211914062,
+      "learning_rate": 9.982057392876215e-06,
+      "loss": 0.4173,
+      "step": 35230
+    },
+    {
+      "epoch": 3.5296739620373616,
+      "grad_norm": 2.719331741333008,
+      "learning_rate": 9.969457100335658e-06,
+      "loss": 0.4215,
+      "step": 35240
+    },
+    {
+      "epoch": 3.530675614764361,
+      "grad_norm": 2.2385475635528564,
+      "learning_rate": 9.956862784570726e-06,
+      "loss": 0.4139,
+      "step": 35250
+    },
+    {
+      "epoch": 3.531677267491361,
+      "grad_norm": 2.103825330734253,
+      "learning_rate": 9.944274450589433e-06,
+      "loss": 0.4366,
+      "step": 35260
+    },
+    {
+      "epoch": 3.5326789202183604,
+      "grad_norm": 2.2733116149902344,
+      "learning_rate": 9.931692103397464e-06,
+      "loss": 0.4787,
+      "step": 35270
+    },
+    {
+      "epoch": 3.53368057294536,
+      "grad_norm": 1.8610239028930664,
+      "learning_rate": 9.919115747998059e-06,
+      "loss": 0.4291,
+      "step": 35280
+    },
+    {
+      "epoch": 3.534682225672359,
+      "grad_norm": 1.7383731603622437,
+      "learning_rate": 9.906545389392105e-06,
+      "loss": 0.4369,
+      "step": 35290
+    },
+    {
+      "epoch": 3.535683878399359,
+      "grad_norm": 3.8377766609191895,
+      "learning_rate": 9.89398103257813e-06,
+      "loss": 0.4582,
+      "step": 35300
+    },
+    {
+      "epoch": 3.5366855311263583,
+      "grad_norm": 1.6542972326278687,
+      "learning_rate": 9.88142268255223e-06,
+      "loss": 0.3911,
+      "step": 35310
+    },
+    {
+      "epoch": 3.537687183853358,
+      "grad_norm": 1.7561484575271606,
+      "learning_rate": 9.868870344308127e-06,
+      "loss": 0.3694,
+      "step": 35320
+    },
+    {
+      "epoch": 3.5386888365803575,
+      "grad_norm": 2.326295852661133,
+      "learning_rate": 9.856324022837177e-06,
+      "loss": 0.42,
+      "step": 35330
+    },
+    {
+      "epoch": 3.539690489307357,
+      "grad_norm": 1.6023072004318237,
+      "learning_rate": 9.843783723128328e-06,
+      "loss": 0.4183,
+      "step": 35340
+    },
+    {
+      "epoch": 3.5406921420343567,
+      "grad_norm": 2.0997536182403564,
+      "learning_rate": 9.8312494501681e-06,
+      "loss": 0.4093,
+      "step": 35350
+    },
+    {
+      "epoch": 3.5416937947613563,
+      "grad_norm": 2.6246097087860107,
+      "learning_rate": 9.818721208940674e-06,
+      "loss": 0.4217,
+      "step": 35360
+    },
+    {
+      "epoch": 3.542695447488356,
+      "grad_norm": 2.3198230266571045,
+      "learning_rate": 9.806199004427807e-06,
+      "loss": 0.4572,
+      "step": 35370
+    },
+    {
+      "epoch": 3.5436971002153554,
+      "grad_norm": 2.1230669021606445,
+      "learning_rate": 9.793682841608836e-06,
+      "loss": 0.4245,
+      "step": 35380
+    },
+    {
+      "epoch": 3.544698752942355,
+      "grad_norm": 2.395850896835327,
+      "learning_rate": 9.781172725460711e-06,
+      "loss": 0.4536,
+      "step": 35390
+    },
+    {
+      "epoch": 3.545700405669354,
+      "grad_norm": 2.717364549636841,
+      "learning_rate": 9.768668660957999e-06,
+      "loss": 0.4774,
+      "step": 35400
+    },
+    {
+      "epoch": 3.546702058396354,
+      "grad_norm": 2.4505763053894043,
+      "learning_rate": 9.756170653072844e-06,
+      "loss": 0.4056,
+      "step": 35410
+    },
+    {
+      "epoch": 3.5477037111233534,
+      "grad_norm": 3.048994779586792,
+      "learning_rate": 9.74367870677495e-06,
+      "loss": 0.4456,
+      "step": 35420
+    },
+    {
+      "epoch": 3.548705363850353,
+      "grad_norm": 2.351593017578125,
+      "learning_rate": 9.731192827031666e-06,
+      "loss": 0.4442,
+      "step": 35430
+    },
+    {
+      "epoch": 3.5497070165773525,
+      "grad_norm": 1.9275217056274414,
+      "learning_rate": 9.718713018807895e-06,
+      "loss": 0.4316,
+      "step": 35440
+    },
+    {
+      "epoch": 3.550708669304352,
+      "grad_norm": 2.4288742542266846,
+      "learning_rate": 9.706239287066132e-06,
+      "loss": 0.4615,
+      "step": 35450
+    },
+    {
+      "epoch": 3.5517103220313517,
+      "grad_norm": 2.2295823097229004,
+      "learning_rate": 9.69377163676646e-06,
+      "loss": 0.4165,
+      "step": 35460
+    },
+    {
+      "epoch": 3.5527119747583513,
+      "grad_norm": 2.53344464302063,
+      "learning_rate": 9.68131007286654e-06,
+      "loss": 0.483,
+      "step": 35470
+    },
+    {
+      "epoch": 3.553713627485351,
+      "grad_norm": 2.220214366912842,
+      "learning_rate": 9.668854600321612e-06,
+      "loss": 0.4092,
+      "step": 35480
+    },
+    {
+      "epoch": 3.5547152802123505,
+      "grad_norm": 2.255557060241699,
+      "learning_rate": 9.656405224084494e-06,
+      "loss": 0.4361,
+      "step": 35490
+    },
+    {
+      "epoch": 3.55571693293935,
+      "grad_norm": 2.5812127590179443,
+      "learning_rate": 9.643961949105584e-06,
+      "loss": 0.4721,
+      "step": 35500
+    },
+    {
+      "epoch": 3.5567185856663492,
+      "grad_norm": 1.8932257890701294,
+      "learning_rate": 9.631524780332851e-06,
+      "loss": 0.4126,
+      "step": 35510
+    },
+    {
+      "epoch": 3.5577202383933493,
+      "grad_norm": 2.0898537635803223,
+      "learning_rate": 9.619093722711833e-06,
+      "loss": 0.4027,
+      "step": 35520
+    },
+    {
+      "epoch": 3.5587218911203484,
+      "grad_norm": 2.0887091159820557,
+      "learning_rate": 9.606668781185646e-06,
+      "loss": 0.4123,
+      "step": 35530
+    },
+    {
+      "epoch": 3.559723543847348,
+      "grad_norm": 2.2083592414855957,
+      "learning_rate": 9.594249960694965e-06,
+      "loss": 0.4965,
+      "step": 35540
+    },
+    {
+      "epoch": 3.5607251965743476,
+      "grad_norm": 2.0207650661468506,
+      "learning_rate": 9.58183726617804e-06,
+      "loss": 0.4269,
+      "step": 35550
+    },
+    {
+      "epoch": 3.561726849301347,
+      "grad_norm": 1.9305944442749023,
+      "learning_rate": 9.569430702570672e-06,
+      "loss": 0.4924,
+      "step": 35560
+    },
+    {
+      "epoch": 3.562728502028347,
+      "grad_norm": 2.5445716381073,
+      "learning_rate": 9.557030274806239e-06,
+      "loss": 0.3865,
+      "step": 35570
+    },
+    {
+      "epoch": 3.5637301547553464,
+      "grad_norm": 2.159574031829834,
+      "learning_rate": 9.54463598781567e-06,
+      "loss": 0.4323,
+      "step": 35580
+    },
+    {
+      "epoch": 3.564731807482346,
+      "grad_norm": 2.1673364639282227,
+      "learning_rate": 9.532247846527453e-06,
+      "loss": 0.4703,
+      "step": 35590
+    },
+    {
+      "epoch": 3.5657334602093456,
+      "grad_norm": 2.177016496658325,
+      "learning_rate": 9.519865855867637e-06,
+      "loss": 0.4123,
+      "step": 35600
+    },
+    {
+      "epoch": 3.566735112936345,
+      "grad_norm": 1.9912105798721313,
+      "learning_rate": 9.507490020759818e-06,
+      "loss": 0.4372,
+      "step": 35610
+    },
+    {
+      "epoch": 3.5677367656633443,
+      "grad_norm": 1.8226884603500366,
+      "learning_rate": 9.495120346125145e-06,
+      "loss": 0.4321,
+      "step": 35620
+    },
+    {
+      "epoch": 3.5687384183903443,
+      "grad_norm": 1.9442039728164673,
+      "learning_rate": 9.48275683688234e-06,
+      "loss": 0.3922,
+      "step": 35630
+    },
+    {
+      "epoch": 3.5697400711173435,
+      "grad_norm": 1.86778724193573,
+      "learning_rate": 9.470399497947633e-06,
+      "loss": 0.4015,
+      "step": 35640
+    },
+    {
+      "epoch": 3.570741723844343,
+      "grad_norm": 2.015185832977295,
+      "learning_rate": 9.458048334234818e-06,
+      "loss": 0.4101,
+      "step": 35650
+    },
+    {
+      "epoch": 3.5717433765713427,
+      "grad_norm": 2.0542356967926025,
+      "learning_rate": 9.445703350655264e-06,
+      "loss": 0.4151,
+      "step": 35660
+    },
+    {
+      "epoch": 3.5727450292983423,
+      "grad_norm": 2.4986319541931152,
+      "learning_rate": 9.433364552117829e-06,
+      "loss": 0.4189,
+      "step": 35670
+    },
+    {
+      "epoch": 3.573746682025342,
+      "grad_norm": 2.3909826278686523,
+      "learning_rate": 9.421031943528941e-06,
+      "loss": 0.4168,
+      "step": 35680
+    },
+    {
+      "epoch": 3.5747483347523414,
+      "grad_norm": 2.0575599670410156,
+      "learning_rate": 9.408705529792577e-06,
+      "loss": 0.4215,
+      "step": 35690
+    },
+    {
+      "epoch": 3.575749987479341,
+      "grad_norm": 2.2897729873657227,
+      "learning_rate": 9.396385315810242e-06,
+      "loss": 0.4642,
+      "step": 35700
+    },
+    {
+      "epoch": 3.5767516402063406,
+      "grad_norm": 2.1848273277282715,
+      "learning_rate": 9.384071306480947e-06,
+      "loss": 0.4217,
+      "step": 35710
+    },
+    {
+      "epoch": 3.57775329293334,
+      "grad_norm": 2.4846177101135254,
+      "learning_rate": 9.371763506701265e-06,
+      "loss": 0.423,
+      "step": 35720
+    },
+    {
+      "epoch": 3.5787549456603394,
+      "grad_norm": 1.7969287633895874,
+      "learning_rate": 9.359461921365307e-06,
+      "loss": 0.4192,
+      "step": 35730
+    },
+    {
+      "epoch": 3.579756598387339,
+      "grad_norm": 2.2867469787597656,
+      "learning_rate": 9.347166555364704e-06,
+      "loss": 0.461,
+      "step": 35740
+    },
+    {
+      "epoch": 3.5807582511143385,
+      "grad_norm": 2.5535647869110107,
+      "learning_rate": 9.334877413588583e-06,
+      "loss": 0.4299,
+      "step": 35750
+    },
+    {
+      "epoch": 3.581759903841338,
+      "grad_norm": 1.96112859249115,
+      "learning_rate": 9.322594500923653e-06,
+      "loss": 0.4267,
+      "step": 35760
+    },
+    {
+      "epoch": 3.5827615565683377,
+      "grad_norm": 3.1883227825164795,
+      "learning_rate": 9.31031782225411e-06,
+      "loss": 0.4415,
+      "step": 35770
+    },
+    {
+      "epoch": 3.5837632092953373,
+      "grad_norm": 2.1661252975463867,
+      "learning_rate": 9.298047382461656e-06,
+      "loss": 0.46,
+      "step": 35780
+    },
+    {
+      "epoch": 3.584764862022337,
+      "grad_norm": 2.134152412414551,
+      "learning_rate": 9.285783186425559e-06,
+      "loss": 0.3904,
+      "step": 35790
+    },
+    {
+      "epoch": 3.5857665147493365,
+      "grad_norm": 2.428616523742676,
+      "learning_rate": 9.273525239022567e-06,
+      "loss": 0.4525,
+      "step": 35800
+    },
+    {
+      "epoch": 3.586768167476336,
+      "grad_norm": 2.4641940593719482,
+      "learning_rate": 9.26127354512696e-06,
+      "loss": 0.4485,
+      "step": 35810
+    },
+    {
+      "epoch": 3.5877698202033352,
+      "grad_norm": 2.321436643600464,
+      "learning_rate": 9.249028109610524e-06,
+      "loss": 0.4743,
+      "step": 35820
+    },
+    {
+      "epoch": 3.5887714729303353,
+      "grad_norm": 2.3309600353240967,
+      "learning_rate": 9.236788937342558e-06,
+      "loss": 0.4444,
+      "step": 35830
+    },
+    {
+      "epoch": 3.5897731256573344,
+      "grad_norm": 1.9578757286071777,
+      "learning_rate": 9.22455603318987e-06,
+      "loss": 0.4182,
+      "step": 35840
+    },
+    {
+      "epoch": 3.590774778384334,
+      "grad_norm": 1.8859965801239014,
+      "learning_rate": 9.212329402016784e-06,
+      "loss": 0.3995,
+      "step": 35850
+    },
+    {
+      "epoch": 3.5917764311113336,
+      "grad_norm": 2.524366855621338,
+      "learning_rate": 9.200109048685112e-06,
+      "loss": 0.4419,
+      "step": 35860
+    },
+    {
+      "epoch": 3.592778083838333,
+      "grad_norm": 2.6382808685302734,
+      "learning_rate": 9.18789497805419e-06,
+      "loss": 0.3921,
+      "step": 35870
+    },
+    {
+      "epoch": 3.593779736565333,
+      "grad_norm": 2.853111743927002,
+      "learning_rate": 9.175687194980839e-06,
+      "loss": 0.459,
+      "step": 35880
+    },
+    {
+      "epoch": 3.5947813892923324,
+      "grad_norm": 2.2526397705078125,
+      "learning_rate": 9.163485704319392e-06,
+      "loss": 0.4678,
+      "step": 35890
+    },
+    {
+      "epoch": 3.595783042019332,
+      "grad_norm": 1.7041195631027222,
+      "learning_rate": 9.15129051092167e-06,
+      "loss": 0.4143,
+      "step": 35900
+    },
+    {
+      "epoch": 3.5967846947463316,
+      "grad_norm": 1.7411357164382935,
+      "learning_rate": 9.139101619636995e-06,
+      "loss": 0.441,
+      "step": 35910
+    },
+    {
+      "epoch": 3.597786347473331,
+      "grad_norm": 1.8171757459640503,
+      "learning_rate": 9.126919035312186e-06,
+      "loss": 0.4283,
+      "step": 35920
+    },
+    {
+      "epoch": 3.5987880002003303,
+      "grad_norm": 2.0983874797821045,
+      "learning_rate": 9.114742762791547e-06,
+      "loss": 0.4392,
+      "step": 35930
+    },
+    {
+      "epoch": 3.5997896529273303,
+      "grad_norm": 2.409411907196045,
+      "learning_rate": 9.102572806916876e-06,
+      "loss": 0.4293,
+      "step": 35940
+    },
+    {
+      "epoch": 3.6007913056543295,
+      "grad_norm": 1.5346509218215942,
+      "learning_rate": 9.090409172527462e-06,
+      "loss": 0.3898,
+      "step": 35950
+    },
+    {
+      "epoch": 3.601792958381329,
+      "grad_norm": 1.5412535667419434,
+      "learning_rate": 9.078251864460074e-06,
+      "loss": 0.3976,
+      "step": 35960
+    },
+    {
+      "epoch": 3.6027946111083287,
+      "grad_norm": 1.62209153175354,
+      "learning_rate": 9.066100887548972e-06,
+      "loss": 0.4361,
+      "step": 35970
+    },
+    {
+      "epoch": 3.6037962638353283,
+      "grad_norm": 2.0511932373046875,
+      "learning_rate": 9.053956246625884e-06,
+      "loss": 0.4037,
+      "step": 35980
+    },
+    {
+      "epoch": 3.604797916562328,
+      "grad_norm": 1.7861130237579346,
+      "learning_rate": 9.041817946520054e-06,
+      "loss": 0.4244,
+      "step": 35990
+    },
+    {
+      "epoch": 3.6057995692893274,
+      "grad_norm": 1.4258286952972412,
+      "learning_rate": 9.029685992058159e-06,
+      "loss": 0.4323,
+      "step": 36000
+    },
+    {
+      "epoch": 3.606801222016327,
+      "grad_norm": 1.9314757585525513,
+      "learning_rate": 9.01756038806437e-06,
+      "loss": 0.4408,
+      "step": 36010
+    },
+    {
+      "epoch": 3.6078028747433266,
+      "grad_norm": 1.7380179166793823,
+      "learning_rate": 9.005441139360358e-06,
+      "loss": 0.3859,
+      "step": 36020
+    },
+    {
+      "epoch": 3.608804527470326,
+      "grad_norm": 2.3640384674072266,
+      "learning_rate": 8.99332825076524e-06,
+      "loss": 0.4027,
+      "step": 36030
+    },
+    {
+      "epoch": 3.6098061801973254,
+      "grad_norm": 3.133868455886841,
+      "learning_rate": 8.981221727095598e-06,
+      "loss": 0.4192,
+      "step": 36040
+    },
+    {
+      "epoch": 3.6108078329243254,
+      "grad_norm": 1.5783082246780396,
+      "learning_rate": 8.969121573165492e-06,
+      "loss": 0.4265,
+      "step": 36050
+    },
+    {
+      "epoch": 3.6118094856513245,
+      "grad_norm": 1.7218824625015259,
+      "learning_rate": 8.957027793786477e-06,
+      "loss": 0.4918,
+      "step": 36060
+    },
+    {
+      "epoch": 3.612811138378324,
+      "grad_norm": 2.190296173095703,
+      "learning_rate": 8.944940393767524e-06,
+      "loss": 0.4599,
+      "step": 36070
+    },
+    {
+      "epoch": 3.6138127911053237,
+      "grad_norm": 1.9028578996658325,
+      "learning_rate": 8.93285937791509e-06,
+      "loss": 0.3851,
+      "step": 36080
+    },
+    {
+      "epoch": 3.6148144438323233,
+      "grad_norm": 2.483093500137329,
+      "learning_rate": 8.920784751033115e-06,
+      "loss": 0.5199,
+      "step": 36090
+    },
+    {
+      "epoch": 3.615816096559323,
+      "grad_norm": 2.389422655105591,
+      "learning_rate": 8.908716517922972e-06,
+      "loss": 0.4206,
+      "step": 36100
+    },
+    {
+      "epoch": 3.6168177492863225,
+      "grad_norm": 2.4605555534362793,
+      "learning_rate": 8.89665468338348e-06,
+      "loss": 0.3872,
+      "step": 36110
+    },
+    {
+      "epoch": 3.617819402013322,
+      "grad_norm": 2.5835015773773193,
+      "learning_rate": 8.884599252210948e-06,
+      "loss": 0.422,
+      "step": 36120
+    },
+    {
+      "epoch": 3.6188210547403217,
+      "grad_norm": 2.2940011024475098,
+      "learning_rate": 8.872550229199128e-06,
+      "loss": 0.4322,
+      "step": 36130
+    },
+    {
+      "epoch": 3.6198227074673213,
+      "grad_norm": 2.135439872741699,
+      "learning_rate": 8.860507619139194e-06,
+      "loss": 0.3873,
+      "step": 36140
+    },
+    {
+      "epoch": 3.6208243601943204,
+      "grad_norm": 2.894531488418579,
+      "learning_rate": 8.848471426819813e-06,
+      "loss": 0.4451,
+      "step": 36150
+    },
+    {
+      "epoch": 3.6218260129213204,
+      "grad_norm": 2.076733350753784,
+      "learning_rate": 8.836441657027078e-06,
+      "loss": 0.4245,
+      "step": 36160
+    },
+    {
+      "epoch": 3.6228276656483196,
+      "grad_norm": 2.015641689300537,
+      "learning_rate": 8.82441831454453e-06,
+      "loss": 0.484,
+      "step": 36170
+    },
+    {
+      "epoch": 3.623829318375319,
+      "grad_norm": 2.4164772033691406,
+      "learning_rate": 8.812401404153153e-06,
+      "loss": 0.4671,
+      "step": 36180
+    },
+    {
+      "epoch": 3.6248309711023188,
+      "grad_norm": 2.1226491928100586,
+      "learning_rate": 8.80039093063138e-06,
+      "loss": 0.3972,
+      "step": 36190
+    },
+    {
+      "epoch": 3.6258326238293184,
+      "grad_norm": 2.2570536136627197,
+      "learning_rate": 8.788386898755075e-06,
+      "loss": 0.4411,
+      "step": 36200
+    },
+    {
+      "epoch": 3.626834276556318,
+      "grad_norm": 2.020395040512085,
+      "learning_rate": 8.776389313297551e-06,
+      "loss": 0.4889,
+      "step": 36210
+    },
+    {
+      "epoch": 3.6278359292833176,
+      "grad_norm": 2.354189157485962,
+      "learning_rate": 8.764398179029546e-06,
+      "loss": 0.4833,
+      "step": 36220
+    },
+    {
+      "epoch": 3.628837582010317,
+      "grad_norm": 1.986047625541687,
+      "learning_rate": 8.752413500719242e-06,
+      "loss": 0.3904,
+      "step": 36230
+    },
+    {
+      "epoch": 3.6298392347373167,
+      "grad_norm": 1.9563875198364258,
+      "learning_rate": 8.740435283132253e-06,
+      "loss": 0.4306,
+      "step": 36240
+    },
+    {
+      "epoch": 3.6308408874643163,
+      "grad_norm": 2.950375556945801,
+      "learning_rate": 8.72846353103162e-06,
+      "loss": 0.4219,
+      "step": 36250
+    },
+    {
+      "epoch": 3.6318425401913155,
+      "grad_norm": 2.0507826805114746,
+      "learning_rate": 8.716498249177814e-06,
+      "loss": 0.4712,
+      "step": 36260
+    },
+    {
+      "epoch": 3.632844192918315,
+      "grad_norm": 2.2555625438690186,
+      "learning_rate": 8.704539442328736e-06,
+      "loss": 0.4081,
+      "step": 36270
+    },
+    {
+      "epoch": 3.6338458456453147,
+      "grad_norm": 1.7647645473480225,
+      "learning_rate": 8.692587115239706e-06,
+      "loss": 0.425,
+      "step": 36280
+    },
+    {
+      "epoch": 3.6348474983723142,
+      "grad_norm": 2.237157106399536,
+      "learning_rate": 8.680641272663479e-06,
+      "loss": 0.3712,
+      "step": 36290
+    },
+    {
+      "epoch": 3.635849151099314,
+      "grad_norm": 1.650541067123413,
+      "learning_rate": 8.668701919350216e-06,
+      "loss": 0.4203,
+      "step": 36300
+    },
+    {
+      "epoch": 3.6368508038263134,
+      "grad_norm": 2.149442672729492,
+      "learning_rate": 8.656769060047504e-06,
+      "loss": 0.3975,
+      "step": 36310
+    },
+    {
+      "epoch": 3.637852456553313,
+      "grad_norm": 1.5350333452224731,
+      "learning_rate": 8.64484269950037e-06,
+      "loss": 0.3873,
+      "step": 36320
+    },
+    {
+      "epoch": 3.6388541092803126,
+      "grad_norm": 2.2094063758850098,
+      "learning_rate": 8.632922842451212e-06,
+      "loss": 0.3724,
+      "step": 36330
+    },
+    {
+      "epoch": 3.639855762007312,
+      "grad_norm": 2.3353826999664307,
+      "learning_rate": 8.621009493639867e-06,
+      "loss": 0.4148,
+      "step": 36340
+    },
+    {
+      "epoch": 3.6408574147343113,
+      "grad_norm": 2.6725997924804688,
+      "learning_rate": 8.60910265780361e-06,
+      "loss": 0.4197,
+      "step": 36350
+    },
+    {
+      "epoch": 3.6418590674613114,
+      "grad_norm": 2.070082664489746,
+      "learning_rate": 8.597202339677071e-06,
+      "loss": 0.3991,
+      "step": 36360
+    },
+    {
+      "epoch": 3.6428607201883105,
+      "grad_norm": 2.181931972503662,
+      "learning_rate": 8.585308543992329e-06,
+      "loss": 0.4527,
+      "step": 36370
+    },
+    {
+      "epoch": 3.64386237291531,
+      "grad_norm": 1.89093017578125,
+      "learning_rate": 8.573421275478844e-06,
+      "loss": 0.4001,
+      "step": 36380
+    },
+    {
+      "epoch": 3.6448640256423097,
+      "grad_norm": 3.096177101135254,
+      "learning_rate": 8.561540538863522e-06,
+      "loss": 0.4488,
+      "step": 36390
+    },
+    {
+      "epoch": 3.6458656783693093,
+      "grad_norm": 2.1697757244110107,
+      "learning_rate": 8.549666338870619e-06,
+      "loss": 0.4258,
+      "step": 36400
+    },
+    {
+      "epoch": 3.646867331096309,
+      "grad_norm": 2.5357141494750977,
+      "learning_rate": 8.537798680221808e-06,
+      "loss": 0.4039,
+      "step": 36410
+    },
+    {
+      "epoch": 3.6478689838233085,
+      "grad_norm": 2.1486408710479736,
+      "learning_rate": 8.525937567636208e-06,
+      "loss": 0.4074,
+      "step": 36420
+    },
+    {
+      "epoch": 3.648870636550308,
+      "grad_norm": 2.1422061920166016,
+      "learning_rate": 8.514083005830259e-06,
+      "loss": 0.449,
+      "step": 36430
+    },
+    {
+      "epoch": 3.6498722892773077,
+      "grad_norm": 2.3769538402557373,
+      "learning_rate": 8.502234999517839e-06,
+      "loss": 0.4385,
+      "step": 36440
+    },
+    {
+      "epoch": 3.6508739420043073,
+      "grad_norm": 2.6000874042510986,
+      "learning_rate": 8.490393553410225e-06,
+      "loss": 0.4101,
+      "step": 36450
+    },
+    {
+      "epoch": 3.6518755947313064,
+      "grad_norm": 2.888254165649414,
+      "learning_rate": 8.478558672216078e-06,
+      "loss": 0.4509,
+      "step": 36460
+    },
+    {
+      "epoch": 3.6528772474583064,
+      "grad_norm": 1.998104214668274,
+      "learning_rate": 8.46673036064142e-06,
+      "loss": 0.3462,
+      "step": 36470
+    },
+    {
+      "epoch": 3.6538789001853056,
+      "grad_norm": 1.9908857345581055,
+      "learning_rate": 8.454908623389706e-06,
+      "loss": 0.4169,
+      "step": 36480
+    },
+    {
+      "epoch": 3.654880552912305,
+      "grad_norm": 2.254089593887329,
+      "learning_rate": 8.44309346516175e-06,
+      "loss": 0.4894,
+      "step": 36490
+    },
+    {
+      "epoch": 3.6558822056393048,
+      "grad_norm": 2.5244266986846924,
+      "learning_rate": 8.431284890655752e-06,
+      "loss": 0.5155,
+      "step": 36500
+    },
+    {
+      "epoch": 3.6568838583663044,
+      "grad_norm": 2.26857328414917,
+      "learning_rate": 8.419482904567305e-06,
+      "loss": 0.4682,
+      "step": 36510
+    },
+    {
+      "epoch": 3.657885511093304,
+      "grad_norm": 3.812756299972534,
+      "learning_rate": 8.407687511589366e-06,
+      "loss": 0.4426,
+      "step": 36520
+    },
+    {
+      "epoch": 3.6588871638203035,
+      "grad_norm": 2.134424924850464,
+      "learning_rate": 8.395898716412293e-06,
+      "loss": 0.4639,
+      "step": 36530
+    },
+    {
+      "epoch": 3.659888816547303,
+      "grad_norm": 1.799182653427124,
+      "learning_rate": 8.384116523723778e-06,
+      "loss": 0.4578,
+      "step": 36540
+    },
+    {
+      "epoch": 3.6608904692743027,
+      "grad_norm": 2.2345874309539795,
+      "learning_rate": 8.372340938208948e-06,
+      "loss": 0.4465,
+      "step": 36550
+    },
+    {
+      "epoch": 3.6618921220013023,
+      "grad_norm": 1.7349672317504883,
+      "learning_rate": 8.36057196455025e-06,
+      "loss": 0.445,
+      "step": 36560
+    },
+    {
+      "epoch": 3.6628937747283015,
+      "grad_norm": 2.1134767532348633,
+      "learning_rate": 8.348809607427535e-06,
+      "loss": 0.4594,
+      "step": 36570
+    },
+    {
+      "epoch": 3.6638954274553015,
+      "grad_norm": 2.5449306964874268,
+      "learning_rate": 8.337053871518005e-06,
+      "loss": 0.4286,
+      "step": 36580
+    },
+    {
+      "epoch": 3.6648970801823006,
+      "grad_norm": 2.225423812866211,
+      "learning_rate": 8.325304761496234e-06,
+      "loss": 0.4111,
+      "step": 36590
+    },
+    {
+      "epoch": 3.6658987329093002,
+      "grad_norm": 1.876946210861206,
+      "learning_rate": 8.313562282034162e-06,
+      "loss": 0.419,
+      "step": 36600
+    },
+    {
+      "epoch": 3.6669003856363,
+      "grad_norm": 2.1918551921844482,
+      "learning_rate": 8.301826437801097e-06,
+      "loss": 0.4192,
+      "step": 36610
+    },
+    {
+      "epoch": 3.6679020383632994,
+      "grad_norm": 2.6870341300964355,
+      "learning_rate": 8.290097233463697e-06,
+      "loss": 0.386,
+      "step": 36620
+    },
+    {
+      "epoch": 3.668903691090299,
+      "grad_norm": 1.8027386665344238,
+      "learning_rate": 8.278374673685993e-06,
+      "loss": 0.4481,
+      "step": 36630
+    },
+    {
+      "epoch": 3.6699053438172986,
+      "grad_norm": 1.7095636129379272,
+      "learning_rate": 8.266658763129362e-06,
+      "loss": 0.4075,
+      "step": 36640
+    },
+    {
+      "epoch": 3.670906996544298,
+      "grad_norm": 1.323539137840271,
+      "learning_rate": 8.254949506452548e-06,
+      "loss": 0.4143,
+      "step": 36650
+    },
+    {
+      "epoch": 3.671908649271298,
+      "grad_norm": 2.6137101650238037,
+      "learning_rate": 8.243246908311639e-06,
+      "loss": 0.4067,
+      "step": 36660
+    },
+    {
+      "epoch": 3.6729103019982974,
+      "grad_norm": 2.2960166931152344,
+      "learning_rate": 8.231550973360072e-06,
+      "loss": 0.4166,
+      "step": 36670
+    },
+    {
+      "epoch": 3.6739119547252965,
+      "grad_norm": 1.6058300733566284,
+      "learning_rate": 8.219861706248672e-06,
+      "loss": 0.4128,
+      "step": 36680
+    },
+    {
+      "epoch": 3.6749136074522966,
+      "grad_norm": 1.9703116416931152,
+      "learning_rate": 8.208179111625552e-06,
+      "loss": 0.4412,
+      "step": 36690
+    },
+    {
+      "epoch": 3.6759152601792957,
+      "grad_norm": 2.711176633834839,
+      "learning_rate": 8.196503194136213e-06,
+      "loss": 0.4466,
+      "step": 36700
+    },
+    {
+      "epoch": 3.6769169129062953,
+      "grad_norm": 2.1494076251983643,
+      "learning_rate": 8.184833958423494e-06,
+      "loss": 0.4272,
+      "step": 36710
+    },
+    {
+      "epoch": 3.677918565633295,
+      "grad_norm": 2.330552339553833,
+      "learning_rate": 8.173171409127566e-06,
+      "loss": 0.4899,
+      "step": 36720
+    },
+    {
+      "epoch": 3.6789202183602945,
+      "grad_norm": 2.2153825759887695,
+      "learning_rate": 8.161515550885959e-06,
+      "loss": 0.3723,
+      "step": 36730
+    },
+    {
+      "epoch": 3.679921871087294,
+      "grad_norm": 1.7207276821136475,
+      "learning_rate": 8.149866388333515e-06,
+      "loss": 0.397,
+      "step": 36740
+    },
+    {
+      "epoch": 3.6809235238142937,
+      "grad_norm": 2.033440351486206,
+      "learning_rate": 8.138223926102462e-06,
+      "loss": 0.4444,
+      "step": 36750
+    },
+    {
+      "epoch": 3.6819251765412933,
+      "grad_norm": 2.009399175643921,
+      "learning_rate": 8.126588168822303e-06,
+      "loss": 0.4054,
+      "step": 36760
+    },
+    {
+      "epoch": 3.682926829268293,
+      "grad_norm": 1.952618956565857,
+      "learning_rate": 8.114959121119903e-06,
+      "loss": 0.3859,
+      "step": 36770
+    },
+    {
+      "epoch": 3.6839284819952924,
+      "grad_norm": 1.7253168821334839,
+      "learning_rate": 8.10333678761948e-06,
+      "loss": 0.3954,
+      "step": 36780
+    },
+    {
+      "epoch": 3.6849301347222916,
+      "grad_norm": 1.8752772808074951,
+      "learning_rate": 8.091721172942566e-06,
+      "loss": 0.4315,
+      "step": 36790
+    },
+    {
+      "epoch": 3.685931787449291,
+      "grad_norm": 1.8806124925613403,
+      "learning_rate": 8.080112281707985e-06,
+      "loss": 0.4244,
+      "step": 36800
+    },
+    {
+      "epoch": 3.6869334401762908,
+      "grad_norm": 2.6830246448516846,
+      "learning_rate": 8.068510118531949e-06,
+      "loss": 0.4661,
+      "step": 36810
+    },
+    {
+      "epoch": 3.6879350929032904,
+      "grad_norm": 2.2413887977600098,
+      "learning_rate": 8.056914688027964e-06,
+      "loss": 0.4661,
+      "step": 36820
+    },
+    {
+      "epoch": 3.68893674563029,
+      "grad_norm": 2.0021140575408936,
+      "learning_rate": 8.045325994806838e-06,
+      "loss": 0.4098,
+      "step": 36830
+    },
+    {
+      "epoch": 3.6899383983572895,
+      "grad_norm": 1.5996559858322144,
+      "learning_rate": 8.033744043476749e-06,
+      "loss": 0.4021,
+      "step": 36840
+    },
+    {
+      "epoch": 3.690940051084289,
+      "grad_norm": 3.1006476879119873,
+      "learning_rate": 8.022168838643152e-06,
+      "loss": 0.4296,
+      "step": 36850
+    },
+    {
+      "epoch": 3.6919417038112887,
+      "grad_norm": 2.16410231590271,
+      "learning_rate": 8.010600384908848e-06,
+      "loss": 0.4847,
+      "step": 36860
+    },
+    {
+      "epoch": 3.6929433565382883,
+      "grad_norm": 2.522000312805176,
+      "learning_rate": 7.99903868687392e-06,
+      "loss": 0.4312,
+      "step": 36870
+    },
+    {
+      "epoch": 3.6939450092652875,
+      "grad_norm": 1.8807637691497803,
+      "learning_rate": 7.9874837491358e-06,
+      "loss": 0.5074,
+      "step": 36880
+    },
+    {
+      "epoch": 3.6949466619922875,
+      "grad_norm": 1.5786900520324707,
+      "learning_rate": 7.975935576289218e-06,
+      "loss": 0.3864,
+      "step": 36890
+    },
+    {
+      "epoch": 3.6959483147192866,
+      "grad_norm": 2.4154186248779297,
+      "learning_rate": 7.964394172926206e-06,
+      "loss": 0.4542,
+      "step": 36900
+    },
+    {
+      "epoch": 3.6969499674462862,
+      "grad_norm": 1.6693038940429688,
+      "learning_rate": 7.952859543636116e-06,
+      "loss": 0.4141,
+      "step": 36910
+    },
+    {
+      "epoch": 3.697951620173286,
+      "grad_norm": 2.0245635509490967,
+      "learning_rate": 7.941331693005599e-06,
+      "loss": 0.4179,
+      "step": 36920
+    },
+    {
+      "epoch": 3.6989532729002854,
+      "grad_norm": 1.8546513319015503,
+      "learning_rate": 7.929810625618612e-06,
+      "loss": 0.4161,
+      "step": 36930
+    },
+    {
+      "epoch": 3.699954925627285,
+      "grad_norm": 1.8944015502929688,
+      "learning_rate": 7.91829634605642e-06,
+      "loss": 0.432,
+      "step": 36940
+    },
+    {
+      "epoch": 3.7009565783542846,
+      "grad_norm": 2.0617423057556152,
+      "learning_rate": 7.906788858897579e-06,
+      "loss": 0.4466,
+      "step": 36950
+    },
+    {
+      "epoch": 3.701958231081284,
+      "grad_norm": 2.174147129058838,
+      "learning_rate": 7.895288168717951e-06,
+      "loss": 0.4087,
+      "step": 36960
+    },
+    {
+      "epoch": 3.702959883808284,
+      "grad_norm": 2.1996078491210938,
+      "learning_rate": 7.883794280090698e-06,
+      "loss": 0.3939,
+      "step": 36970
+    },
+    {
+      "epoch": 3.7039615365352834,
+      "grad_norm": 1.9552979469299316,
+      "learning_rate": 7.872307197586271e-06,
+      "loss": 0.4228,
+      "step": 36980
+    },
+    {
+      "epoch": 3.7049631892622825,
+      "grad_norm": 2.0555949211120605,
+      "learning_rate": 7.860826925772414e-06,
+      "loss": 0.4539,
+      "step": 36990
+    },
+    {
+      "epoch": 3.7059648419892826,
+      "grad_norm": 2.5038888454437256,
+      "learning_rate": 7.849353469214165e-06,
+      "loss": 0.4057,
+      "step": 37000
+    },
+    {
+      "epoch": 3.7069664947162817,
+      "grad_norm": 2.1329166889190674,
+      "learning_rate": 7.837886832473859e-06,
+      "loss": 0.4751,
+      "step": 37010
+    },
+    {
+      "epoch": 3.7079681474432813,
+      "grad_norm": 2.378232479095459,
+      "learning_rate": 7.826427020111107e-06,
+      "loss": 0.4125,
+      "step": 37020
+    },
+    {
+      "epoch": 3.708969800170281,
+      "grad_norm": 2.8863255977630615,
+      "learning_rate": 7.814974036682814e-06,
+      "loss": 0.4706,
+      "step": 37030
+    },
+    {
+      "epoch": 3.7099714528972805,
+      "grad_norm": 1.9137957096099854,
+      "learning_rate": 7.80352788674317e-06,
+      "loss": 0.4149,
+      "step": 37040
+    },
+    {
+      "epoch": 3.71097310562428,
+      "grad_norm": 1.9049749374389648,
+      "learning_rate": 7.792088574843643e-06,
+      "loss": 0.3909,
+      "step": 37050
+    },
+    {
+      "epoch": 3.7119747583512797,
+      "grad_norm": 2.699704647064209,
+      "learning_rate": 7.780656105532983e-06,
+      "loss": 0.4085,
+      "step": 37060
+    },
+    {
+      "epoch": 3.7129764110782792,
+      "grad_norm": 2.358980178833008,
+      "learning_rate": 7.769230483357212e-06,
+      "loss": 0.4699,
+      "step": 37070
+    },
+    {
+      "epoch": 3.713978063805279,
+      "grad_norm": 2.233934164047241,
+      "learning_rate": 7.757811712859665e-06,
+      "loss": 0.4224,
+      "step": 37080
+    },
+    {
+      "epoch": 3.7149797165322784,
+      "grad_norm": 2.1187474727630615,
+      "learning_rate": 7.7463997985809e-06,
+      "loss": 0.4722,
+      "step": 37090
+    },
+    {
+      "epoch": 3.7159813692592776,
+      "grad_norm": 2.188882350921631,
+      "learning_rate": 7.734994745058771e-06,
+      "loss": 0.4373,
+      "step": 37100
+    },
+    {
+      "epoch": 3.7169830219862776,
+      "grad_norm": 2.354186773300171,
+      "learning_rate": 7.723596556828434e-06,
+      "loss": 0.4367,
+      "step": 37110
+    },
+    {
+      "epoch": 3.7179846747132768,
+      "grad_norm": 2.162371873855591,
+      "learning_rate": 7.71220523842226e-06,
+      "loss": 0.3975,
+      "step": 37120
+    },
+    {
+      "epoch": 3.7189863274402764,
+      "grad_norm": 2.0304088592529297,
+      "learning_rate": 7.700820794369923e-06,
+      "loss": 0.4807,
+      "step": 37130
+    },
+    {
+      "epoch": 3.719987980167276,
+      "grad_norm": 1.8520346879959106,
+      "learning_rate": 7.689443229198365e-06,
+      "loss": 0.4204,
+      "step": 37140
+    },
+    {
+      "epoch": 3.7209896328942755,
+      "grad_norm": 1.8495360612869263,
+      "learning_rate": 7.678072547431787e-06,
+      "loss": 0.4181,
+      "step": 37150
+    },
+    {
+      "epoch": 3.721991285621275,
+      "grad_norm": 2.0318777561187744,
+      "learning_rate": 7.666708753591626e-06,
+      "loss": 0.4456,
+      "step": 37160
+    },
+    {
+      "epoch": 3.7229929383482747,
+      "grad_norm": 2.1289327144622803,
+      "learning_rate": 7.655351852196627e-06,
+      "loss": 0.434,
+      "step": 37170
+    },
+    {
+      "epoch": 3.7239945910752743,
+      "grad_norm": 2.1364686489105225,
+      "learning_rate": 7.644001847762774e-06,
+      "loss": 0.4215,
+      "step": 37180
+    },
+    {
+      "epoch": 3.724996243802274,
+      "grad_norm": 2.0084264278411865,
+      "learning_rate": 7.63265874480329e-06,
+      "loss": 0.3936,
+      "step": 37190
+    },
+    {
+      "epoch": 3.7259978965292735,
+      "grad_norm": 1.8063205480575562,
+      "learning_rate": 7.621322547828663e-06,
+      "loss": 0.3598,
+      "step": 37200
+    },
+    {
+      "epoch": 3.7269995492562726,
+      "grad_norm": 2.089787006378174,
+      "learning_rate": 7.609993261346668e-06,
+      "loss": 0.4059,
+      "step": 37210
+    },
+    {
+      "epoch": 3.7280012019832727,
+      "grad_norm": 1.835331678390503,
+      "learning_rate": 7.598670889862297e-06,
+      "loss": 0.4411,
+      "step": 37220
+    },
+    {
+      "epoch": 3.729002854710272,
+      "grad_norm": 2.2048423290252686,
+      "learning_rate": 7.587355437877777e-06,
+      "loss": 0.4427,
+      "step": 37230
+    },
+    {
+      "epoch": 3.7300045074372714,
+      "grad_norm": 2.705153703689575,
+      "learning_rate": 7.576046909892637e-06,
+      "loss": 0.4109,
+      "step": 37240
+    },
+    {
+      "epoch": 3.731006160164271,
+      "grad_norm": 1.9492063522338867,
+      "learning_rate": 7.564745310403612e-06,
+      "loss": 0.3932,
+      "step": 37250
+    },
+    {
+      "epoch": 3.7320078128912706,
+      "grad_norm": 2.270702600479126,
+      "learning_rate": 7.553450643904692e-06,
+      "loss": 0.3898,
+      "step": 37260
+    },
+    {
+      "epoch": 3.73300946561827,
+      "grad_norm": 1.9737622737884521,
+      "learning_rate": 7.542162914887111e-06,
+      "loss": 0.4094,
+      "step": 37270
+    },
+    {
+      "epoch": 3.7340111183452698,
+      "grad_norm": 2.8277175426483154,
+      "learning_rate": 7.530882127839348e-06,
+      "loss": 0.3824,
+      "step": 37280
+    },
+    {
+      "epoch": 3.7350127710722694,
+      "grad_norm": 2.938880205154419,
+      "learning_rate": 7.519608287247113e-06,
+      "loss": 0.4285,
+      "step": 37290
+    },
+    {
+      "epoch": 3.736014423799269,
+      "grad_norm": 2.2028396129608154,
+      "learning_rate": 7.508341397593363e-06,
+      "loss": 0.3687,
+      "step": 37300
+    },
+    {
+      "epoch": 3.7370160765262685,
+      "grad_norm": 2.1324620246887207,
+      "learning_rate": 7.497081463358286e-06,
+      "loss": 0.4436,
+      "step": 37310
+    },
+    {
+      "epoch": 3.7380177292532677,
+      "grad_norm": 2.0066702365875244,
+      "learning_rate": 7.485828489019303e-06,
+      "loss": 0.369,
+      "step": 37320
+    },
+    {
+      "epoch": 3.7390193819802673,
+      "grad_norm": 2.5639231204986572,
+      "learning_rate": 7.47458247905107e-06,
+      "loss": 0.4182,
+      "step": 37330
+    },
+    {
+      "epoch": 3.740021034707267,
+      "grad_norm": 2.4227943420410156,
+      "learning_rate": 7.463343437925477e-06,
+      "loss": 0.4117,
+      "step": 37340
+    },
+    {
+      "epoch": 3.7410226874342665,
+      "grad_norm": 1.719071626663208,
+      "learning_rate": 7.452111370111636e-06,
+      "loss": 0.4173,
+      "step": 37350
+    },
+    {
+      "epoch": 3.742024340161266,
+      "grad_norm": 2.1645774841308594,
+      "learning_rate": 7.440886280075887e-06,
+      "loss": 0.4216,
+      "step": 37360
+    },
+    {
+      "epoch": 3.7430259928882657,
+      "grad_norm": 2.005495309829712,
+      "learning_rate": 7.429668172281803e-06,
+      "loss": 0.4413,
+      "step": 37370
+    },
+    {
+      "epoch": 3.7440276456152652,
+      "grad_norm": 2.228274345397949,
+      "learning_rate": 7.418457051190173e-06,
+      "loss": 0.3805,
+      "step": 37380
+    },
+    {
+      "epoch": 3.745029298342265,
+      "grad_norm": 2.2738568782806396,
+      "learning_rate": 7.407252921259008e-06,
+      "loss": 0.4104,
+      "step": 37390
+    },
+    {
+      "epoch": 3.7460309510692644,
+      "grad_norm": 2.142463445663452,
+      "learning_rate": 7.396055786943543e-06,
+      "loss": 0.4443,
+      "step": 37400
+    },
+    {
+      "epoch": 3.7470326037962636,
+      "grad_norm": 2.024423122406006,
+      "learning_rate": 7.3848656526962295e-06,
+      "loss": 0.3953,
+      "step": 37410
+    },
+    {
+      "epoch": 3.7480342565232636,
+      "grad_norm": 2.5991528034210205,
+      "learning_rate": 7.373682522966735e-06,
+      "loss": 0.4175,
+      "step": 37420
+    },
+    {
+      "epoch": 3.7490359092502628,
+      "grad_norm": 1.819032907485962,
+      "learning_rate": 7.362506402201932e-06,
+      "loss": 0.46,
+      "step": 37430
+    },
+    {
+      "epoch": 3.7500375619772623,
+      "grad_norm": 2.4464993476867676,
+      "learning_rate": 7.351337294845942e-06,
+      "loss": 0.4255,
+      "step": 37440
+    },
+    {
+      "epoch": 3.751039214704262,
+      "grad_norm": 2.157233715057373,
+      "learning_rate": 7.340175205340044e-06,
+      "loss": 0.4405,
+      "step": 37450
+    },
+    {
+      "epoch": 3.7520408674312615,
+      "grad_norm": 2.148372173309326,
+      "learning_rate": 7.329020138122761e-06,
+      "loss": 0.4301,
+      "step": 37460
+    },
+    {
+      "epoch": 3.753042520158261,
+      "grad_norm": 2.149197578430176,
+      "learning_rate": 7.317872097629836e-06,
+      "loss": 0.4719,
+      "step": 37470
+    },
+    {
+      "epoch": 3.7540441728852607,
+      "grad_norm": 2.0958821773529053,
+      "learning_rate": 7.306731088294172e-06,
+      "loss": 0.4272,
+      "step": 37480
+    },
+    {
+      "epoch": 3.7550458256122603,
+      "grad_norm": 2.766040086746216,
+      "learning_rate": 7.295597114545907e-06,
+      "loss": 0.4578,
+      "step": 37490
+    },
+    {
+      "epoch": 3.75604747833926,
+      "grad_norm": 2.5304043292999268,
+      "learning_rate": 7.284470180812392e-06,
+      "loss": 0.4636,
+      "step": 37500
+    },
+    {
+      "epoch": 3.7570491310662595,
+      "grad_norm": 2.753812313079834,
+      "learning_rate": 7.2733502915181604e-06,
+      "loss": 0.4124,
+      "step": 37510
+    },
+    {
+      "epoch": 3.7580507837932586,
+      "grad_norm": 1.706272840499878,
+      "learning_rate": 7.262237451084938e-06,
+      "loss": 0.3704,
+      "step": 37520
+    },
+    {
+      "epoch": 3.7590524365202587,
+      "grad_norm": 2.44985294342041,
+      "learning_rate": 7.251131663931652e-06,
+      "loss": 0.4415,
+      "step": 37530
+    },
+    {
+      "epoch": 3.760054089247258,
+      "grad_norm": 1.9321026802062988,
+      "learning_rate": 7.240032934474447e-06,
+      "loss": 0.4082,
+      "step": 37540
+    },
+    {
+      "epoch": 3.7610557419742574,
+      "grad_norm": 2.1816911697387695,
+      "learning_rate": 7.228941267126646e-06,
+      "loss": 0.4284,
+      "step": 37550
+    },
+    {
+      "epoch": 3.762057394701257,
+      "grad_norm": 2.1587181091308594,
+      "learning_rate": 7.217856666298736e-06,
+      "loss": 0.4536,
+      "step": 37560
+    },
+    {
+      "epoch": 3.7630590474282566,
+      "grad_norm": 2.2277698516845703,
+      "learning_rate": 7.206779136398445e-06,
+      "loss": 0.3842,
+      "step": 37570
+    },
+    {
+      "epoch": 3.764060700155256,
+      "grad_norm": 2.4164507389068604,
+      "learning_rate": 7.195708681830665e-06,
+      "loss": 0.3984,
+      "step": 37580
+    },
+    {
+      "epoch": 3.7650623528822558,
+      "grad_norm": 2.4073264598846436,
+      "learning_rate": 7.18464530699745e-06,
+      "loss": 0.3684,
+      "step": 37590
+    },
+    {
+      "epoch": 3.7660640056092554,
+      "grad_norm": 2.5888216495513916,
+      "learning_rate": 7.1735890162980855e-06,
+      "loss": 0.4355,
+      "step": 37600
+    },
+    {
+      "epoch": 3.767065658336255,
+      "grad_norm": 2.342641592025757,
+      "learning_rate": 7.162539814129013e-06,
+      "loss": 0.4305,
+      "step": 37610
+    },
+    {
+      "epoch": 3.7680673110632545,
+      "grad_norm": 2.2944300174713135,
+      "learning_rate": 7.151497704883855e-06,
+      "loss": 0.4233,
+      "step": 37620
+    },
+    {
+      "epoch": 3.7690689637902537,
+      "grad_norm": 2.190720796585083,
+      "learning_rate": 7.1404626929534206e-06,
+      "loss": 0.4602,
+      "step": 37630
+    },
+    {
+      "epoch": 3.7700706165172537,
+      "grad_norm": 2.474369764328003,
+      "learning_rate": 7.129434782725697e-06,
+      "loss": 0.4556,
+      "step": 37640
+    },
+    {
+      "epoch": 3.771072269244253,
+      "grad_norm": 3.5644686222076416,
+      "learning_rate": 7.118413978585839e-06,
+      "loss": 0.4676,
+      "step": 37650
+    },
+    {
+      "epoch": 3.7720739219712525,
+      "grad_norm": 2.2803492546081543,
+      "learning_rate": 7.107400284916185e-06,
+      "loss": 0.4021,
+      "step": 37660
+    },
+    {
+      "epoch": 3.773075574698252,
+      "grad_norm": 2.6325793266296387,
+      "learning_rate": 7.0963937060962435e-06,
+      "loss": 0.4832,
+      "step": 37670
+    },
+    {
+      "epoch": 3.7740772274252516,
+      "grad_norm": 2.1822357177734375,
+      "learning_rate": 7.085394246502692e-06,
+      "loss": 0.4158,
+      "step": 37680
+    },
+    {
+      "epoch": 3.7750788801522512,
+      "grad_norm": 2.2507176399230957,
+      "learning_rate": 7.074401910509376e-06,
+      "loss": 0.4426,
+      "step": 37690
+    },
+    {
+      "epoch": 3.776080532879251,
+      "grad_norm": 1.774174690246582,
+      "learning_rate": 7.063416702487313e-06,
+      "loss": 0.3955,
+      "step": 37700
+    },
+    {
+      "epoch": 3.7770821856062504,
+      "grad_norm": 2.155637264251709,
+      "learning_rate": 7.05243862680468e-06,
+      "loss": 0.3967,
+      "step": 37710
+    },
+    {
+      "epoch": 3.77808383833325,
+      "grad_norm": 2.492680311203003,
+      "learning_rate": 7.041467687826825e-06,
+      "loss": 0.4521,
+      "step": 37720
+    },
+    {
+      "epoch": 3.7790854910602496,
+      "grad_norm": 1.9159152507781982,
+      "learning_rate": 7.030503889916254e-06,
+      "loss": 0.4482,
+      "step": 37730
+    },
+    {
+      "epoch": 3.7800871437872487,
+      "grad_norm": 2.340388298034668,
+      "learning_rate": 7.0195472374326335e-06,
+      "loss": 0.4927,
+      "step": 37740
+    },
+    {
+      "epoch": 3.781088796514249,
+      "grad_norm": 2.2713913917541504,
+      "learning_rate": 7.008597734732786e-06,
+      "loss": 0.4053,
+      "step": 37750
+    },
+    {
+      "epoch": 3.782090449241248,
+      "grad_norm": 1.636338710784912,
+      "learning_rate": 6.997655386170698e-06,
+      "loss": 0.4071,
+      "step": 37760
+    },
+    {
+      "epoch": 3.7830921019682475,
+      "grad_norm": 1.8299113512039185,
+      "learning_rate": 6.986720196097507e-06,
+      "loss": 0.4461,
+      "step": 37770
+    },
+    {
+      "epoch": 3.784093754695247,
+      "grad_norm": 2.3852038383483887,
+      "learning_rate": 6.9757921688615e-06,
+      "loss": 0.4631,
+      "step": 37780
+    },
+    {
+      "epoch": 3.7850954074222467,
+      "grad_norm": 2.396932363510132,
+      "learning_rate": 6.964871308808118e-06,
+      "loss": 0.4139,
+      "step": 37790
+    },
+    {
+      "epoch": 3.7860970601492463,
+      "grad_norm": 1.9309535026550293,
+      "learning_rate": 6.953957620279971e-06,
+      "loss": 0.4465,
+      "step": 37800
+    },
+    {
+      "epoch": 3.787098712876246,
+      "grad_norm": 2.1654818058013916,
+      "learning_rate": 6.943051107616785e-06,
+      "loss": 0.3709,
+      "step": 37810
+    },
+    {
+      "epoch": 3.7881003656032455,
+      "grad_norm": 3.310075283050537,
+      "learning_rate": 6.93215177515544e-06,
+      "loss": 0.4081,
+      "step": 37820
+    },
+    {
+      "epoch": 3.789102018330245,
+      "grad_norm": 2.255657434463501,
+      "learning_rate": 6.921259627229989e-06,
+      "loss": 0.4686,
+      "step": 37830
+    },
+    {
+      "epoch": 3.7901036710572447,
+      "grad_norm": 1.773685336112976,
+      "learning_rate": 6.910374668171607e-06,
+      "loss": 0.4184,
+      "step": 37840
+    },
+    {
+      "epoch": 3.791105323784244,
+      "grad_norm": 1.9942299127578735,
+      "learning_rate": 6.899496902308592e-06,
+      "loss": 0.4079,
+      "step": 37850
+    },
+    {
+      "epoch": 3.7921069765112434,
+      "grad_norm": 2.451260805130005,
+      "learning_rate": 6.888626333966405e-06,
+      "loss": 0.4615,
+      "step": 37860
+    },
+    {
+      "epoch": 3.793108629238243,
+      "grad_norm": 2.6214842796325684,
+      "learning_rate": 6.877762967467666e-06,
+      "loss": 0.4257,
+      "step": 37870
+    },
+    {
+      "epoch": 3.7941102819652426,
+      "grad_norm": 2.6001267433166504,
+      "learning_rate": 6.866906807132079e-06,
+      "loss": 0.4699,
+      "step": 37880
+    },
+    {
+      "epoch": 3.795111934692242,
+      "grad_norm": 2.7137014865875244,
+      "learning_rate": 6.856057857276507e-06,
+      "loss": 0.4248,
+      "step": 37890
+    },
+    {
+      "epoch": 3.7961135874192418,
+      "grad_norm": 2.153409957885742,
+      "learning_rate": 6.845216122214973e-06,
+      "loss": 0.4262,
+      "step": 37900
+    },
+    {
+      "epoch": 3.7971152401462414,
+      "grad_norm": 2.0230278968811035,
+      "learning_rate": 6.834381606258597e-06,
+      "loss": 0.4167,
+      "step": 37910
+    },
+    {
+      "epoch": 3.798116892873241,
+      "grad_norm": 1.8823751211166382,
+      "learning_rate": 6.823554313715619e-06,
+      "loss": 0.4419,
+      "step": 37920
+    },
+    {
+      "epoch": 3.7991185456002405,
+      "grad_norm": 2.529693126678467,
+      "learning_rate": 6.81273424889145e-06,
+      "loss": 0.4636,
+      "step": 37930
+    },
+    {
+      "epoch": 3.8001201983272397,
+      "grad_norm": 2.7344167232513428,
+      "learning_rate": 6.801921416088597e-06,
+      "loss": 0.417,
+      "step": 37940
+    },
+    {
+      "epoch": 3.8011218510542397,
+      "grad_norm": 2.9459967613220215,
+      "learning_rate": 6.791115819606689e-06,
+      "loss": 0.4455,
+      "step": 37950
+    },
+    {
+      "epoch": 3.802123503781239,
+      "grad_norm": 2.277332305908203,
+      "learning_rate": 6.780317463742492e-06,
+      "loss": 0.4097,
+      "step": 37960
+    },
+    {
+      "epoch": 3.8031251565082385,
+      "grad_norm": 2.0806617736816406,
+      "learning_rate": 6.769526352789882e-06,
+      "loss": 0.4105,
+      "step": 37970
+    },
+    {
+      "epoch": 3.804126809235238,
+      "grad_norm": 2.3580572605133057,
+      "learning_rate": 6.75874249103986e-06,
+      "loss": 0.4319,
+      "step": 37980
+    },
+    {
+      "epoch": 3.8051284619622376,
+      "grad_norm": 2.0618369579315186,
+      "learning_rate": 6.7479658827805435e-06,
+      "loss": 0.4417,
+      "step": 37990
+    },
+    {
+      "epoch": 3.8061301146892372,
+      "grad_norm": 2.297889471054077,
+      "learning_rate": 6.737196532297163e-06,
+      "loss": 0.4363,
+      "step": 38000
+    },
+    {
+      "epoch": 3.807131767416237,
+      "grad_norm": 1.9872561693191528,
+      "learning_rate": 6.726434443872071e-06,
+      "loss": 0.4912,
+      "step": 38010
+    },
+    {
+      "epoch": 3.8081334201432364,
+      "grad_norm": 1.8478449583053589,
+      "learning_rate": 6.7156796217847155e-06,
+      "loss": 0.3763,
+      "step": 38020
+    },
+    {
+      "epoch": 3.809135072870236,
+      "grad_norm": 1.7361677885055542,
+      "learning_rate": 6.7049320703116756e-06,
+      "loss": 0.4135,
+      "step": 38030
+    },
+    {
+      "epoch": 3.8101367255972356,
+      "grad_norm": 2.6249921321868896,
+      "learning_rate": 6.694191793726623e-06,
+      "loss": 0.3936,
+      "step": 38040
+    },
+    {
+      "epoch": 3.8111383783242347,
+      "grad_norm": 1.9055705070495605,
+      "learning_rate": 6.683458796300349e-06,
+      "loss": 0.4196,
+      "step": 38050
+    },
+    {
+      "epoch": 3.812140031051235,
+      "grad_norm": 1.5801509618759155,
+      "learning_rate": 6.672733082300739e-06,
+      "loss": 0.3927,
+      "step": 38060
+    },
+    {
+      "epoch": 3.813141683778234,
+      "grad_norm": 2.310429573059082,
+      "learning_rate": 6.662014655992791e-06,
+      "loss": 0.4781,
+      "step": 38070
+    },
+    {
+      "epoch": 3.8141433365052335,
+      "grad_norm": 2.7247142791748047,
+      "learning_rate": 6.651303521638599e-06,
+      "loss": 0.4488,
+      "step": 38080
+    },
+    {
+      "epoch": 3.815144989232233,
+      "grad_norm": 1.6246259212493896,
+      "learning_rate": 6.640599683497364e-06,
+      "loss": 0.3836,
+      "step": 38090
+    },
+    {
+      "epoch": 3.8161466419592327,
+      "grad_norm": 1.9836045503616333,
+      "learning_rate": 6.629903145825378e-06,
+      "loss": 0.3884,
+      "step": 38100
+    },
+    {
+      "epoch": 3.8171482946862323,
+      "grad_norm": 1.8660048246383667,
+      "learning_rate": 6.619213912876038e-06,
+      "loss": 0.3775,
+      "step": 38110
+    },
+    {
+      "epoch": 3.818149947413232,
+      "grad_norm": 2.0846478939056396,
+      "learning_rate": 6.608531988899822e-06,
+      "loss": 0.4793,
+      "step": 38120
+    },
+    {
+      "epoch": 3.8191516001402315,
+      "grad_norm": 2.101534366607666,
+      "learning_rate": 6.5978573781443346e-06,
+      "loss": 0.4689,
+      "step": 38130
+    },
+    {
+      "epoch": 3.820153252867231,
+      "grad_norm": 2.1078479290008545,
+      "learning_rate": 6.587190084854228e-06,
+      "loss": 0.4624,
+      "step": 38140
+    },
+    {
+      "epoch": 3.8211549055942307,
+      "grad_norm": 1.6181825399398804,
+      "learning_rate": 6.576530113271265e-06,
+      "loss": 0.4424,
+      "step": 38150
+    },
+    {
+      "epoch": 3.82215655832123,
+      "grad_norm": 1.6215962171554565,
+      "learning_rate": 6.565877467634324e-06,
+      "loss": 0.4434,
+      "step": 38160
+    },
+    {
+      "epoch": 3.82315821104823,
+      "grad_norm": 2.2983250617980957,
+      "learning_rate": 6.5552321521793195e-06,
+      "loss": 0.4694,
+      "step": 38170
+    },
+    {
+      "epoch": 3.824159863775229,
+      "grad_norm": 2.1482555866241455,
+      "learning_rate": 6.5445941711392845e-06,
+      "loss": 0.4466,
+      "step": 38180
+    },
+    {
+      "epoch": 3.8251615165022286,
+      "grad_norm": 2.240410804748535,
+      "learning_rate": 6.533963528744319e-06,
+      "loss": 0.4196,
+      "step": 38190
+    },
+    {
+      "epoch": 3.826163169229228,
+      "grad_norm": 2.4457499980926514,
+      "learning_rate": 6.523340229221639e-06,
+      "loss": 0.4329,
+      "step": 38200
+    },
+    {
+      "epoch": 3.8271648219562278,
+      "grad_norm": 2.4264252185821533,
+      "learning_rate": 6.51272427679549e-06,
+      "loss": 0.3891,
+      "step": 38210
+    },
+    {
+      "epoch": 3.8281664746832273,
+      "grad_norm": 1.8337993621826172,
+      "learning_rate": 6.5021156756872175e-06,
+      "loss": 0.3847,
+      "step": 38220
+    },
+    {
+      "epoch": 3.829168127410227,
+      "grad_norm": 3.0118589401245117,
+      "learning_rate": 6.491514430115278e-06,
+      "loss": 0.4514,
+      "step": 38230
+    },
+    {
+      "epoch": 3.8301697801372265,
+      "grad_norm": 2.0712578296661377,
+      "learning_rate": 6.480920544295141e-06,
+      "loss": 0.4276,
+      "step": 38240
+    },
+    {
+      "epoch": 3.831171432864226,
+      "grad_norm": 1.8239521980285645,
+      "learning_rate": 6.4703340224393895e-06,
+      "loss": 0.3844,
+      "step": 38250
+    },
+    {
+      "epoch": 3.8321730855912257,
+      "grad_norm": 2.277188301086426,
+      "learning_rate": 6.459754868757675e-06,
+      "loss": 0.475,
+      "step": 38260
+    },
+    {
+      "epoch": 3.833174738318225,
+      "grad_norm": 2.219277858734131,
+      "learning_rate": 6.449183087456723e-06,
+      "loss": 0.4694,
+      "step": 38270
+    },
+    {
+      "epoch": 3.834176391045225,
+      "grad_norm": 2.4775564670562744,
+      "learning_rate": 6.4386186827402884e-06,
+      "loss": 0.4562,
+      "step": 38280
+    },
+    {
+      "epoch": 3.835178043772224,
+      "grad_norm": 2.250697612762451,
+      "learning_rate": 6.428061658809248e-06,
+      "loss": 0.4658,
+      "step": 38290
+    },
+    {
+      "epoch": 3.8361796964992236,
+      "grad_norm": 2.2010769844055176,
+      "learning_rate": 6.41751201986151e-06,
+      "loss": 0.4638,
+      "step": 38300
+    },
+    {
+      "epoch": 3.8371813492262232,
+      "grad_norm": 2.2128689289093018,
+      "learning_rate": 6.406969770092056e-06,
+      "loss": 0.4676,
+      "step": 38310
+    },
+    {
+      "epoch": 3.838183001953223,
+      "grad_norm": 2.170311450958252,
+      "learning_rate": 6.396434913692928e-06,
+      "loss": 0.4089,
+      "step": 38320
+    },
+    {
+      "epoch": 3.8391846546802224,
+      "grad_norm": 2.4297749996185303,
+      "learning_rate": 6.385907454853224e-06,
+      "loss": 0.4224,
+      "step": 38330
+    },
+    {
+      "epoch": 3.840186307407222,
+      "grad_norm": 1.6787548065185547,
+      "learning_rate": 6.375387397759114e-06,
+      "loss": 0.4002,
+      "step": 38340
+    },
+    {
+      "epoch": 3.8411879601342216,
+      "grad_norm": 2.9497814178466797,
+      "learning_rate": 6.364874746593793e-06,
+      "loss": 0.4372,
+      "step": 38350
+    },
+    {
+      "epoch": 3.8421896128612207,
+      "grad_norm": 2.112044334411621,
+      "learning_rate": 6.354369505537555e-06,
+      "loss": 0.4423,
+      "step": 38360
+    },
+    {
+      "epoch": 3.8431912655882208,
+      "grad_norm": 1.6409683227539062,
+      "learning_rate": 6.343871678767715e-06,
+      "loss": 0.4345,
+      "step": 38370
+    },
+    {
+      "epoch": 3.84419291831522,
+      "grad_norm": 2.136765241622925,
+      "learning_rate": 6.333381270458655e-06,
+      "loss": 0.4651,
+      "step": 38380
+    },
+    {
+      "epoch": 3.8451945710422195,
+      "grad_norm": 2.24963641166687,
+      "learning_rate": 6.322898284781801e-06,
+      "loss": 0.4305,
+      "step": 38390
+    },
+    {
+      "epoch": 3.846196223769219,
+      "grad_norm": 2.395167112350464,
+      "learning_rate": 6.312422725905626e-06,
+      "loss": 0.427,
+      "step": 38400
+    },
+    {
+      "epoch": 3.8471978764962187,
+      "grad_norm": 2.715379238128662,
+      "learning_rate": 6.3019545979956545e-06,
+      "loss": 0.483,
+      "step": 38410
+    },
+    {
+      "epoch": 3.8481995292232183,
+      "grad_norm": 2.4242758750915527,
+      "learning_rate": 6.291493905214454e-06,
+      "loss": 0.3938,
+      "step": 38420
+    },
+    {
+      "epoch": 3.849201181950218,
+      "grad_norm": 2.016488790512085,
+      "learning_rate": 6.281040651721637e-06,
+      "loss": 0.4412,
+      "step": 38430
+    },
+    {
+      "epoch": 3.8502028346772175,
+      "grad_norm": 2.018977403640747,
+      "learning_rate": 6.270594841673852e-06,
+      "loss": 0.4535,
+      "step": 38440
+    },
+    {
+      "epoch": 3.851204487404217,
+      "grad_norm": 2.2235140800476074,
+      "learning_rate": 6.260156479224797e-06,
+      "loss": 0.4212,
+      "step": 38450
+    },
+    {
+      "epoch": 3.8522061401312166,
+      "grad_norm": 2.009124279022217,
+      "learning_rate": 6.2497255685251995e-06,
+      "loss": 0.3964,
+      "step": 38460
+    },
+    {
+      "epoch": 3.853207792858216,
+      "grad_norm": 2.90751576423645,
+      "learning_rate": 6.239302113722833e-06,
+      "loss": 0.4423,
+      "step": 38470
+    },
+    {
+      "epoch": 3.854209445585216,
+      "grad_norm": 2.9282796382904053,
+      "learning_rate": 6.228886118962493e-06,
+      "loss": 0.4727,
+      "step": 38480
+    },
+    {
+      "epoch": 3.855211098312215,
+      "grad_norm": 2.084012985229492,
+      "learning_rate": 6.218477588386035e-06,
+      "loss": 0.456,
+      "step": 38490
+    },
+    {
+      "epoch": 3.8562127510392146,
+      "grad_norm": 2.1381776332855225,
+      "learning_rate": 6.208076526132306e-06,
+      "loss": 0.4246,
+      "step": 38500
+    },
+    {
+      "epoch": 3.857214403766214,
+      "grad_norm": 1.9279459714889526,
+      "learning_rate": 6.197682936337218e-06,
+      "loss": 0.404,
+      "step": 38510
+    },
+    {
+      "epoch": 3.8582160564932138,
+      "grad_norm": 1.9992539882659912,
+      "learning_rate": 6.187296823133698e-06,
+      "loss": 0.4197,
+      "step": 38520
+    },
+    {
+      "epoch": 3.8592177092202133,
+      "grad_norm": 1.7238306999206543,
+      "learning_rate": 6.1769181906516955e-06,
+      "loss": 0.3974,
+      "step": 38530
+    },
+    {
+      "epoch": 3.860219361947213,
+      "grad_norm": 1.9684339761734009,
+      "learning_rate": 6.1665470430181975e-06,
+      "loss": 0.4424,
+      "step": 38540
+    },
+    {
+      "epoch": 3.8612210146742125,
+      "grad_norm": 2.5566136837005615,
+      "learning_rate": 6.1561833843571975e-06,
+      "loss": 0.4095,
+      "step": 38550
+    },
+    {
+      "epoch": 3.862222667401212,
+      "grad_norm": 2.300997495651245,
+      "learning_rate": 6.1458272187897455e-06,
+      "loss": 0.4618,
+      "step": 38560
+    },
+    {
+      "epoch": 3.8632243201282117,
+      "grad_norm": 2.2941536903381348,
+      "learning_rate": 6.135478550433865e-06,
+      "loss": 0.4177,
+      "step": 38570
+    },
+    {
+      "epoch": 3.864225972855211,
+      "grad_norm": 1.9589506387710571,
+      "learning_rate": 6.125137383404622e-06,
+      "loss": 0.4183,
+      "step": 38580
+    },
+    {
+      "epoch": 3.865227625582211,
+      "grad_norm": 1.2186977863311768,
+      "learning_rate": 6.114803721814114e-06,
+      "loss": 0.4134,
+      "step": 38590
+    },
+    {
+      "epoch": 3.86622927830921,
+      "grad_norm": 1.7170480489730835,
+      "learning_rate": 6.104477569771439e-06,
+      "loss": 0.3788,
+      "step": 38600
+    },
+    {
+      "epoch": 3.8672309310362096,
+      "grad_norm": 3.361743211746216,
+      "learning_rate": 6.094158931382685e-06,
+      "loss": 0.4188,
+      "step": 38610
+    },
+    {
+      "epoch": 3.868232583763209,
+      "grad_norm": 2.1883580684661865,
+      "learning_rate": 6.083847810751004e-06,
+      "loss": 0.4335,
+      "step": 38620
+    },
+    {
+      "epoch": 3.869234236490209,
+      "grad_norm": 1.248789668083191,
+      "learning_rate": 6.073544211976523e-06,
+      "loss": 0.4231,
+      "step": 38630
+    },
+    {
+      "epoch": 3.8702358892172084,
+      "grad_norm": 2.8954248428344727,
+      "learning_rate": 6.063248139156372e-06,
+      "loss": 0.4248,
+      "step": 38640
+    },
+    {
+      "epoch": 3.871237541944208,
+      "grad_norm": 1.9824678897857666,
+      "learning_rate": 6.052959596384719e-06,
+      "loss": 0.4393,
+      "step": 38650
+    },
+    {
+      "epoch": 3.8722391946712076,
+      "grad_norm": 2.0736277103424072,
+      "learning_rate": 6.042678587752718e-06,
+      "loss": 0.3752,
+      "step": 38660
+    },
+    {
+      "epoch": 3.873240847398207,
+      "grad_norm": 2.2819983959198,
+      "learning_rate": 6.032405117348533e-06,
+      "loss": 0.4167,
+      "step": 38670
+    },
+    {
+      "epoch": 3.8742425001252068,
+      "grad_norm": 2.657353639602661,
+      "learning_rate": 6.022139189257306e-06,
+      "loss": 0.4253,
+      "step": 38680
+    },
+    {
+      "epoch": 3.875244152852206,
+      "grad_norm": 2.045306921005249,
+      "learning_rate": 6.011880807561227e-06,
+      "loss": 0.398,
+      "step": 38690
+    },
+    {
+      "epoch": 3.876245805579206,
+      "grad_norm": 1.6201101541519165,
+      "learning_rate": 6.001629976339448e-06,
+      "loss": 0.4137,
+      "step": 38700
+    },
+    {
+      "epoch": 3.877247458306205,
+      "grad_norm": 2.2877581119537354,
+      "learning_rate": 5.991386699668136e-06,
+      "loss": 0.3722,
+      "step": 38710
+    },
+    {
+      "epoch": 3.8782491110332047,
+      "grad_norm": 1.7606736421585083,
+      "learning_rate": 5.981150981620443e-06,
+      "loss": 0.3913,
+      "step": 38720
+    },
+    {
+      "epoch": 3.8792507637602043,
+      "grad_norm": 2.2930450439453125,
+      "learning_rate": 5.970922826266523e-06,
+      "loss": 0.4395,
+      "step": 38730
+    },
+    {
+      "epoch": 3.880252416487204,
+      "grad_norm": 2.470388650894165,
+      "learning_rate": 5.960702237673521e-06,
+      "loss": 0.4097,
+      "step": 38740
+    },
+    {
+      "epoch": 3.8812540692142035,
+      "grad_norm": 2.244858503341675,
+      "learning_rate": 5.950489219905572e-06,
+      "loss": 0.4404,
+      "step": 38750
+    },
+    {
+      "epoch": 3.882255721941203,
+      "grad_norm": 2.1141445636749268,
+      "learning_rate": 5.9402837770238e-06,
+      "loss": 0.437,
+      "step": 38760
+    },
+    {
+      "epoch": 3.8832573746682026,
+      "grad_norm": 2.3806586265563965,
+      "learning_rate": 5.930085913086322e-06,
+      "loss": 0.432,
+      "step": 38770
+    },
+    {
+      "epoch": 3.8842590273952022,
+      "grad_norm": 2.2542169094085693,
+      "learning_rate": 5.919895632148231e-06,
+      "loss": 0.3765,
+      "step": 38780
+    },
+    {
+      "epoch": 3.885260680122202,
+      "grad_norm": 2.2765283584594727,
+      "learning_rate": 5.909712938261616e-06,
+      "loss": 0.4707,
+      "step": 38790
+    },
+    {
+      "epoch": 3.886262332849201,
+      "grad_norm": 1.9423617124557495,
+      "learning_rate": 5.899537835475544e-06,
+      "loss": 0.4707,
+      "step": 38800
+    },
+    {
+      "epoch": 3.887263985576201,
+      "grad_norm": 2.1070590019226074,
+      "learning_rate": 5.889370327836061e-06,
+      "loss": 0.4737,
+      "step": 38810
+    },
+    {
+      "epoch": 3.8882656383032,
+      "grad_norm": 2.498217821121216,
+      "learning_rate": 5.879210419386197e-06,
+      "loss": 0.3988,
+      "step": 38820
+    },
+    {
+      "epoch": 3.8892672910301997,
+      "grad_norm": 2.292095184326172,
+      "learning_rate": 5.869058114165956e-06,
+      "loss": 0.444,
+      "step": 38830
+    },
+    {
+      "epoch": 3.8902689437571993,
+      "grad_norm": 2.990748882293701,
+      "learning_rate": 5.858913416212325e-06,
+      "loss": 0.4561,
+      "step": 38840
+    },
+    {
+      "epoch": 3.891270596484199,
+      "grad_norm": 2.656426429748535,
+      "learning_rate": 5.848776329559261e-06,
+      "loss": 0.4223,
+      "step": 38850
+    },
+    {
+      "epoch": 3.8922722492111985,
+      "grad_norm": 2.1715846061706543,
+      "learning_rate": 5.838646858237695e-06,
+      "loss": 0.4408,
+      "step": 38860
+    },
+    {
+      "epoch": 3.893273901938198,
+      "grad_norm": 2.254647970199585,
+      "learning_rate": 5.828525006275532e-06,
+      "loss": 0.4076,
+      "step": 38870
+    },
+    {
+      "epoch": 3.8942755546651977,
+      "grad_norm": 3.026210308074951,
+      "learning_rate": 5.818410777697639e-06,
+      "loss": 0.4982,
+      "step": 38880
+    },
+    {
+      "epoch": 3.895277207392197,
+      "grad_norm": 1.857161521911621,
+      "learning_rate": 5.808304176525875e-06,
+      "loss": 0.458,
+      "step": 38890
+    },
+    {
+      "epoch": 3.896278860119197,
+      "grad_norm": 1.916955828666687,
+      "learning_rate": 5.798205206779033e-06,
+      "loss": 0.3936,
+      "step": 38900
+    },
+    {
+      "epoch": 3.897280512846196,
+      "grad_norm": 2.5912342071533203,
+      "learning_rate": 5.788113872472886e-06,
+      "loss": 0.4142,
+      "step": 38910
+    },
+    {
+      "epoch": 3.8982821655731956,
+      "grad_norm": 2.2807843685150146,
+      "learning_rate": 5.778030177620198e-06,
+      "loss": 0.471,
+      "step": 38920
+    },
+    {
+      "epoch": 3.899283818300195,
+      "grad_norm": 1.8373711109161377,
+      "learning_rate": 5.767954126230641e-06,
+      "loss": 0.4649,
+      "step": 38930
+    },
+    {
+      "epoch": 3.900285471027195,
+      "grad_norm": 2.2042641639709473,
+      "learning_rate": 5.757885722310882e-06,
+      "loss": 0.4439,
+      "step": 38940
+    },
+    {
+      "epoch": 3.9012871237541944,
+      "grad_norm": 1.9236149787902832,
+      "learning_rate": 5.747824969864554e-06,
+      "loss": 0.4517,
+      "step": 38950
+    },
+    {
+      "epoch": 3.902288776481194,
+      "grad_norm": 2.47882080078125,
+      "learning_rate": 5.7377718728922365e-06,
+      "loss": 0.4453,
+      "step": 38960
+    },
+    {
+      "epoch": 3.9032904292081936,
+      "grad_norm": 1.873042106628418,
+      "learning_rate": 5.72772643539144e-06,
+      "loss": 0.4618,
+      "step": 38970
+    },
+    {
+      "epoch": 3.904292081935193,
+      "grad_norm": 2.677980661392212,
+      "learning_rate": 5.7176886613566735e-06,
+      "loss": 0.4366,
+      "step": 38980
+    },
+    {
+      "epoch": 3.9052937346621928,
+      "grad_norm": 2.4638803005218506,
+      "learning_rate": 5.707658554779374e-06,
+      "loss": 0.4097,
+      "step": 38990
+    },
+    {
+      "epoch": 3.906295387389192,
+      "grad_norm": 2.025956869125366,
+      "learning_rate": 5.697636119647939e-06,
+      "loss": 0.4124,
+      "step": 39000
+    },
+    {
+      "epoch": 3.907297040116192,
+      "grad_norm": 1.7988700866699219,
+      "learning_rate": 5.68762135994769e-06,
+      "loss": 0.3835,
+      "step": 39010
+    },
+    {
+      "epoch": 3.908298692843191,
+      "grad_norm": 1.9764509201049805,
+      "learning_rate": 5.677614279660934e-06,
+      "loss": 0.4574,
+      "step": 39020
+    },
+    {
+      "epoch": 3.9093003455701907,
+      "grad_norm": 2.4781711101531982,
+      "learning_rate": 5.667614882766908e-06,
+      "loss": 0.4962,
+      "step": 39030
+    },
+    {
+      "epoch": 3.9103019982971903,
+      "grad_norm": 1.637224555015564,
+      "learning_rate": 5.6576231732417745e-06,
+      "loss": 0.4241,
+      "step": 39040
+    },
+    {
+      "epoch": 3.91130365102419,
+      "grad_norm": 2.451122522354126,
+      "learning_rate": 5.647639155058676e-06,
+      "loss": 0.412,
+      "step": 39050
+    },
+    {
+      "epoch": 3.9123053037511895,
+      "grad_norm": 1.8591203689575195,
+      "learning_rate": 5.63766283218767e-06,
+      "loss": 0.4624,
+      "step": 39060
+    },
+    {
+      "epoch": 3.913306956478189,
+      "grad_norm": 2.224898099899292,
+      "learning_rate": 5.627694208595763e-06,
+      "loss": 0.4359,
+      "step": 39070
+    },
+    {
+      "epoch": 3.9143086092051886,
+      "grad_norm": 2.6811466217041016,
+      "learning_rate": 5.617733288246898e-06,
+      "loss": 0.3711,
+      "step": 39080
+    },
+    {
+      "epoch": 3.9153102619321882,
+      "grad_norm": 2.6933958530426025,
+      "learning_rate": 5.607780075101956e-06,
+      "loss": 0.4493,
+      "step": 39090
+    },
+    {
+      "epoch": 3.916311914659188,
+      "grad_norm": 1.5979939699172974,
+      "learning_rate": 5.597834573118754e-06,
+      "loss": 0.3847,
+      "step": 39100
+    },
+    {
+      "epoch": 3.917313567386187,
+      "grad_norm": 2.687382698059082,
+      "learning_rate": 5.587896786252039e-06,
+      "loss": 0.4065,
+      "step": 39110
+    },
+    {
+      "epoch": 3.918315220113187,
+      "grad_norm": 2.6313157081604004,
+      "learning_rate": 5.577966718453495e-06,
+      "loss": 0.4204,
+      "step": 39120
+    },
+    {
+      "epoch": 3.919316872840186,
+      "grad_norm": 2.1213088035583496,
+      "learning_rate": 5.5680443736717325e-06,
+      "loss": 0.4447,
+      "step": 39130
+    },
+    {
+      "epoch": 3.9203185255671857,
+      "grad_norm": 2.800734281539917,
+      "learning_rate": 5.558129755852295e-06,
+      "loss": 0.42,
+      "step": 39140
+    },
+    {
+      "epoch": 3.9213201782941853,
+      "grad_norm": 2.896596908569336,
+      "learning_rate": 5.548222868937649e-06,
+      "loss": 0.453,
+      "step": 39150
+    },
+    {
+      "epoch": 3.922321831021185,
+      "grad_norm": 2.1709372997283936,
+      "learning_rate": 5.538323716867194e-06,
+      "loss": 0.4739,
+      "step": 39160
+    },
+    {
+      "epoch": 3.9233234837481845,
+      "grad_norm": 2.2188522815704346,
+      "learning_rate": 5.528432303577244e-06,
+      "loss": 0.432,
+      "step": 39170
+    },
+    {
+      "epoch": 3.924325136475184,
+      "grad_norm": 2.2879672050476074,
+      "learning_rate": 5.518548633001039e-06,
+      "loss": 0.5457,
+      "step": 39180
+    },
+    {
+      "epoch": 3.9253267892021837,
+      "grad_norm": 2.0803894996643066,
+      "learning_rate": 5.508672709068746e-06,
+      "loss": 0.3965,
+      "step": 39190
+    },
+    {
+      "epoch": 3.9263284419291833,
+      "grad_norm": 1.8538846969604492,
+      "learning_rate": 5.498804535707447e-06,
+      "loss": 0.4526,
+      "step": 39200
+    },
+    {
+      "epoch": 3.927330094656183,
+      "grad_norm": 2.3934788703918457,
+      "learning_rate": 5.488944116841144e-06,
+      "loss": 0.3509,
+      "step": 39210
+    },
+    {
+      "epoch": 3.928331747383182,
+      "grad_norm": 1.6396085023880005,
+      "learning_rate": 5.4790914563907474e-06,
+      "loss": 0.4732,
+      "step": 39220
+    },
+    {
+      "epoch": 3.929333400110182,
+      "grad_norm": 1.8581525087356567,
+      "learning_rate": 5.469246558274096e-06,
+      "loss": 0.4159,
+      "step": 39230
+    },
+    {
+      "epoch": 3.930335052837181,
+      "grad_norm": 2.8066928386688232,
+      "learning_rate": 5.459409426405926e-06,
+      "loss": 0.493,
+      "step": 39240
+    },
+    {
+      "epoch": 3.931336705564181,
+      "grad_norm": 2.4748148918151855,
+      "learning_rate": 5.449580064697915e-06,
+      "loss": 0.4374,
+      "step": 39250
+    },
+    {
+      "epoch": 3.9323383582911804,
+      "grad_norm": 2.191162586212158,
+      "learning_rate": 5.43975847705861e-06,
+      "loss": 0.4382,
+      "step": 39260
+    },
+    {
+      "epoch": 3.93334001101818,
+      "grad_norm": 2.5699470043182373,
+      "learning_rate": 5.429944667393486e-06,
+      "loss": 0.3827,
+      "step": 39270
+    },
+    {
+      "epoch": 3.9343416637451796,
+      "grad_norm": 2.012481689453125,
+      "learning_rate": 5.420138639604947e-06,
+      "loss": 0.4211,
+      "step": 39280
+    },
+    {
+      "epoch": 3.935343316472179,
+      "grad_norm": 2.005993127822876,
+      "learning_rate": 5.410340397592262e-06,
+      "loss": 0.4031,
+      "step": 39290
+    },
+    {
+      "epoch": 3.9363449691991788,
+      "grad_norm": 2.065284490585327,
+      "learning_rate": 5.4005499452516234e-06,
+      "loss": 0.4853,
+      "step": 39300
+    },
+    {
+      "epoch": 3.9373466219261783,
+      "grad_norm": 3.1377053260803223,
+      "learning_rate": 5.3907672864761395e-06,
+      "loss": 0.4342,
+      "step": 39310
+    },
+    {
+      "epoch": 3.938348274653178,
+      "grad_norm": 1.8640254735946655,
+      "learning_rate": 5.380992425155809e-06,
+      "loss": 0.4401,
+      "step": 39320
+    },
+    {
+      "epoch": 3.939349927380177,
+      "grad_norm": 2.006674289703369,
+      "learning_rate": 5.371225365177513e-06,
+      "loss": 0.3803,
+      "step": 39330
+    },
+    {
+      "epoch": 3.9403515801071767,
+      "grad_norm": 1.887819766998291,
+      "learning_rate": 5.361466110425045e-06,
+      "loss": 0.4175,
+      "step": 39340
+    },
+    {
+      "epoch": 3.9413532328341763,
+      "grad_norm": 1.9000790119171143,
+      "learning_rate": 5.351714664779106e-06,
+      "loss": 0.4634,
+      "step": 39350
+    },
+    {
+      "epoch": 3.942354885561176,
+      "grad_norm": 2.495184898376465,
+      "learning_rate": 5.341971032117285e-06,
+      "loss": 0.4234,
+      "step": 39360
+    },
+    {
+      "epoch": 3.9433565382881755,
+      "grad_norm": 2.152071237564087,
+      "learning_rate": 5.332235216314035e-06,
+      "loss": 0.3666,
+      "step": 39370
+    },
+    {
+      "epoch": 3.944358191015175,
+      "grad_norm": 2.0491833686828613,
+      "learning_rate": 5.32250722124075e-06,
+      "loss": 0.4581,
+      "step": 39380
+    },
+    {
+      "epoch": 3.9453598437421746,
+      "grad_norm": 2.057877540588379,
+      "learning_rate": 5.312787050765689e-06,
+      "loss": 0.4422,
+      "step": 39390
+    },
+    {
+      "epoch": 3.9463614964691742,
+      "grad_norm": 2.5603151321411133,
+      "learning_rate": 5.303074708753977e-06,
+      "loss": 0.4315,
+      "step": 39400
+    },
+    {
+      "epoch": 3.947363149196174,
+      "grad_norm": 2.4202048778533936,
+      "learning_rate": 5.293370199067671e-06,
+      "loss": 0.4272,
+      "step": 39410
+    },
+    {
+      "epoch": 3.948364801923173,
+      "grad_norm": 1.896180510520935,
+      "learning_rate": 5.283673525565688e-06,
+      "loss": 0.3761,
+      "step": 39420
+    },
+    {
+      "epoch": 3.949366454650173,
+      "grad_norm": 2.138786554336548,
+      "learning_rate": 5.273984692103831e-06,
+      "loss": 0.418,
+      "step": 39430
+    },
+    {
+      "epoch": 3.950368107377172,
+      "grad_norm": 2.09519100189209,
+      "learning_rate": 5.264303702534784e-06,
+      "loss": 0.4182,
+      "step": 39440
+    },
+    {
+      "epoch": 3.9513697601041717,
+      "grad_norm": 2.606905460357666,
+      "learning_rate": 5.254630560708123e-06,
+      "loss": 0.4222,
+      "step": 39450
+    },
+    {
+      "epoch": 3.9523714128311713,
+      "grad_norm": 2.679440498352051,
+      "learning_rate": 5.24496527047029e-06,
+      "loss": 0.4478,
+      "step": 39460
+    },
+    {
+      "epoch": 3.953373065558171,
+      "grad_norm": 2.5487256050109863,
+      "learning_rate": 5.23530783566461e-06,
+      "loss": 0.373,
+      "step": 39470
+    },
+    {
+      "epoch": 3.9543747182851705,
+      "grad_norm": 2.0131003856658936,
+      "learning_rate": 5.225658260131289e-06,
+      "loss": 0.3875,
+      "step": 39480
+    },
+    {
+      "epoch": 3.95537637101217,
+      "grad_norm": 1.9472322463989258,
+      "learning_rate": 5.216016547707403e-06,
+      "loss": 0.38,
+      "step": 39490
+    },
+    {
+      "epoch": 3.9563780237391697,
+      "grad_norm": 1.9929423332214355,
+      "learning_rate": 5.2063827022269e-06,
+      "loss": 0.4123,
+      "step": 39500
+    },
+    {
+      "epoch": 3.9573796764661693,
+      "grad_norm": 2.648130416870117,
+      "learning_rate": 5.196756727520602e-06,
+      "loss": 0.4668,
+      "step": 39510
+    },
+    {
+      "epoch": 3.958381329193169,
+      "grad_norm": 2.119204521179199,
+      "learning_rate": 5.187138627416202e-06,
+      "loss": 0.3989,
+      "step": 39520
+    },
+    {
+      "epoch": 3.959382981920168,
+      "grad_norm": 2.0803627967834473,
+      "learning_rate": 5.177528405738261e-06,
+      "loss": 0.3887,
+      "step": 39530
+    },
+    {
+      "epoch": 3.960384634647168,
+      "grad_norm": 2.0360963344573975,
+      "learning_rate": 5.167926066308207e-06,
+      "loss": 0.4147,
+      "step": 39540
+    },
+    {
+      "epoch": 3.961386287374167,
+      "grad_norm": 2.736513376235962,
+      "learning_rate": 5.158331612944337e-06,
+      "loss": 0.4734,
+      "step": 39550
+    },
+    {
+      "epoch": 3.962387940101167,
+      "grad_norm": 2.068869113922119,
+      "learning_rate": 5.1487450494618004e-06,
+      "loss": 0.4155,
+      "step": 39560
+    },
+    {
+      "epoch": 3.9633895928281664,
+      "grad_norm": 2.3239376544952393,
+      "learning_rate": 5.139166379672627e-06,
+      "loss": 0.4444,
+      "step": 39570
+    },
+    {
+      "epoch": 3.964391245555166,
+      "grad_norm": 2.593562602996826,
+      "learning_rate": 5.129595607385693e-06,
+      "loss": 0.4024,
+      "step": 39580
+    },
+    {
+      "epoch": 3.9653928982821656,
+      "grad_norm": 1.6795233488082886,
+      "learning_rate": 5.120032736406744e-06,
+      "loss": 0.4005,
+      "step": 39590
+    },
+    {
+      "epoch": 3.966394551009165,
+      "grad_norm": 1.533920407295227,
+      "learning_rate": 5.110477770538366e-06,
+      "loss": 0.5235,
+      "step": 39600
+    },
+    {
+      "epoch": 3.9673962037361648,
+      "grad_norm": 2.483339309692383,
+      "learning_rate": 5.100930713580044e-06,
+      "loss": 0.4192,
+      "step": 39610
+    },
+    {
+      "epoch": 3.9683978564631643,
+      "grad_norm": 2.7518630027770996,
+      "learning_rate": 5.09139156932806e-06,
+      "loss": 0.4448,
+      "step": 39620
+    },
+    {
+      "epoch": 3.969399509190164,
+      "grad_norm": 2.8066458702087402,
+      "learning_rate": 5.081860341575584e-06,
+      "loss": 0.3969,
+      "step": 39630
+    },
+    {
+      "epoch": 3.970401161917163,
+      "grad_norm": 2.323917865753174,
+      "learning_rate": 5.072337034112645e-06,
+      "loss": 0.4005,
+      "step": 39640
+    },
+    {
+      "epoch": 3.971402814644163,
+      "grad_norm": 1.8697916269302368,
+      "learning_rate": 5.062821650726113e-06,
+      "loss": 0.361,
+      "step": 39650
+    },
+    {
+      "epoch": 3.9724044673711623,
+      "grad_norm": 2.3262734413146973,
+      "learning_rate": 5.053314195199685e-06,
+      "loss": 0.4364,
+      "step": 39660
+    },
+    {
+      "epoch": 3.973406120098162,
+      "grad_norm": 2.201371192932129,
+      "learning_rate": 5.043814671313932e-06,
+      "loss": 0.4244,
+      "step": 39670
+    },
+    {
+      "epoch": 3.9744077728251614,
+      "grad_norm": 2.123957633972168,
+      "learning_rate": 5.0343230828462764e-06,
+      "loss": 0.366,
+      "step": 39680
+    },
+    {
+      "epoch": 3.975409425552161,
+      "grad_norm": 2.200098752975464,
+      "learning_rate": 5.02483943357096e-06,
+      "loss": 0.503,
+      "step": 39690
+    },
+    {
+      "epoch": 3.9764110782791606,
+      "grad_norm": 1.9379470348358154,
+      "learning_rate": 5.015363727259076e-06,
+      "loss": 0.4664,
+      "step": 39700
+    },
+    {
+      "epoch": 3.97741273100616,
+      "grad_norm": 1.8062063455581665,
+      "learning_rate": 5.0058959676785785e-06,
+      "loss": 0.4212,
+      "step": 39710
+    },
+    {
+      "epoch": 3.97841438373316,
+      "grad_norm": 2.2486565113067627,
+      "learning_rate": 4.996436158594245e-06,
+      "loss": 0.4427,
+      "step": 39720
+    },
+    {
+      "epoch": 3.9794160364601594,
+      "grad_norm": 2.03916597366333,
+      "learning_rate": 4.986984303767675e-06,
+      "loss": 0.4555,
+      "step": 39730
+    },
+    {
+      "epoch": 3.980417689187159,
+      "grad_norm": 1.792839527130127,
+      "learning_rate": 4.9775404069573425e-06,
+      "loss": 0.3701,
+      "step": 39740
+    },
+    {
+      "epoch": 3.981419341914158,
+      "grad_norm": 2.415205240249634,
+      "learning_rate": 4.968104471918533e-06,
+      "loss": 0.5279,
+      "step": 39750
+    },
+    {
+      "epoch": 3.982420994641158,
+      "grad_norm": 1.9716744422912598,
+      "learning_rate": 4.958676502403367e-06,
+      "loss": 0.4309,
+      "step": 39760
+    },
+    {
+      "epoch": 3.9834226473681573,
+      "grad_norm": 1.593368649482727,
+      "learning_rate": 4.949256502160804e-06,
+      "loss": 0.4196,
+      "step": 39770
+    },
+    {
+      "epoch": 3.984424300095157,
+      "grad_norm": 2.1170709133148193,
+      "learning_rate": 4.939844474936634e-06,
+      "loss": 0.4474,
+      "step": 39780
+    },
+    {
+      "epoch": 3.9854259528221565,
+      "grad_norm": 2.5782203674316406,
+      "learning_rate": 4.930440424473467e-06,
+      "loss": 0.4556,
+      "step": 39790
+    },
+    {
+      "epoch": 3.986427605549156,
+      "grad_norm": 2.213399887084961,
+      "learning_rate": 4.921044354510759e-06,
+      "loss": 0.4286,
+      "step": 39800
+    },
+    {
+      "epoch": 3.9874292582761557,
+      "grad_norm": 2.359858989715576,
+      "learning_rate": 4.911656268784775e-06,
+      "loss": 0.4304,
+      "step": 39810
+    },
+    {
+      "epoch": 3.9884309110031553,
+      "grad_norm": 2.288480281829834,
+      "learning_rate": 4.902276171028617e-06,
+      "loss": 0.3837,
+      "step": 39820
+    },
+    {
+      "epoch": 3.989432563730155,
+      "grad_norm": 2.2940833568573,
+      "learning_rate": 4.8929040649722e-06,
+      "loss": 0.4507,
+      "step": 39830
+    },
+    {
+      "epoch": 3.9904342164571545,
+      "grad_norm": 2.6430327892303467,
+      "learning_rate": 4.883539954342276e-06,
+      "loss": 0.3704,
+      "step": 39840
+    },
+    {
+      "epoch": 3.991435869184154,
+      "grad_norm": 2.3419792652130127,
+      "learning_rate": 4.874183842862401e-06,
+      "loss": 0.4236,
+      "step": 39850
+    },
+    {
+      "epoch": 3.992437521911153,
+      "grad_norm": 2.221893310546875,
+      "learning_rate": 4.864835734252962e-06,
+      "loss": 0.4163,
+      "step": 39860
+    },
+    {
+      "epoch": 3.993439174638153,
+      "grad_norm": 2.3750827312469482,
+      "learning_rate": 4.855495632231161e-06,
+      "loss": 0.3974,
+      "step": 39870
+    },
+    {
+      "epoch": 3.9944408273651524,
+      "grad_norm": 1.812092900276184,
+      "learning_rate": 4.846163540511011e-06,
+      "loss": 0.43,
+      "step": 39880
+    },
+    {
+      "epoch": 3.995442480092152,
+      "grad_norm": 2.505802631378174,
+      "learning_rate": 4.836839462803347e-06,
+      "loss": 0.3909,
+      "step": 39890
+    },
+    {
+      "epoch": 3.9964441328191516,
+      "grad_norm": 2.1492509841918945,
+      "learning_rate": 4.827523402815815e-06,
+      "loss": 0.4526,
+      "step": 39900
+    },
+    {
+      "epoch": 3.997445785546151,
+      "grad_norm": 2.6524550914764404,
+      "learning_rate": 4.818215364252871e-06,
+      "loss": 0.4235,
+      "step": 39910
+    },
+    {
+      "epoch": 3.9984474382731507,
+      "grad_norm": 2.4931113719940186,
+      "learning_rate": 4.808915350815779e-06,
+      "loss": 0.4216,
+      "step": 39920
+    },
+    {
+      "epoch": 3.9994490910001503,
+      "grad_norm": 2.9122989177703857,
+      "learning_rate": 4.799623366202615e-06,
+      "loss": 0.4228,
+      "step": 39930
+    },
+    {
+      "epoch": 4.0004006610908,
+      "grad_norm": 2.093388319015503,
+      "learning_rate": 4.790339414108278e-06,
+      "loss": 0.3735,
+      "step": 39940
+    },
+    {
+      "epoch": 4.001402313817799,
+      "grad_norm": 2.304086685180664,
+      "learning_rate": 4.781063498224439e-06,
+      "loss": 0.4207,
+      "step": 39950
+    },
+    {
+      "epoch": 4.002403966544799,
+      "grad_norm": 1.982155442237854,
+      "learning_rate": 4.771795622239592e-06,
+      "loss": 0.3568,
+      "step": 39960
+    },
+    {
+      "epoch": 4.003405619271798,
+      "grad_norm": 2.5733442306518555,
+      "learning_rate": 4.762535789839054e-06,
+      "loss": 0.4407,
+      "step": 39970
+    },
+    {
+      "epoch": 4.004407271998798,
+      "grad_norm": 2.412712812423706,
+      "learning_rate": 4.753284004704902e-06,
+      "loss": 0.4066,
+      "step": 39980
+    },
+    {
+      "epoch": 4.005408924725797,
+      "grad_norm": 1.9959261417388916,
+      "learning_rate": 4.7440402705160425e-06,
+      "loss": 0.3478,
+      "step": 39990
+    },
+    {
+      "epoch": 4.006410577452797,
+      "grad_norm": 2.1611344814300537,
+      "learning_rate": 4.734804590948169e-06,
+      "loss": 0.4195,
+      "step": 40000
+    },
+    {
+      "epoch": 4.006410577452797,
+      "eval_bleu": 0.3662565942705902,
+      "eval_loss": 0.5082083940505981,
+      "eval_rouge1": 0.7014509543822172,
+      "eval_rouge2": 0.5362548944715757,
+      "eval_rougeL": 0.6602146413463208,
+      "eval_runtime": 86792.7268,
+      "eval_samples_per_second": 0.204,
+      "eval_steps_per_second": 0.026,
+      "eval_wer": 0.7582390590348932,
+      "step": 40000
+    },
+    {
+      "epoch": 4.0074122301797965,
+      "grad_norm": 2.3167426586151123,
+      "learning_rate": 4.725576969673789e-06,
+      "loss": 0.4305,
+      "step": 40010
+    },
+    {
+      "epoch": 4.008413882906797,
+      "grad_norm": 2.5307557582855225,
+      "learning_rate": 4.7163574103621825e-06,
+      "loss": 0.3887,
+      "step": 40020
+    },
+    {
+      "epoch": 4.009415535633796,
+      "grad_norm": 1.9515817165374756,
+      "learning_rate": 4.707145916679426e-06,
+      "loss": 0.4566,
+      "step": 40030
+    },
+    {
+      "epoch": 4.010417188360796,
+      "grad_norm": 2.6875641345977783,
+      "learning_rate": 4.6979424922884096e-06,
+      "loss": 0.3808,
+      "step": 40040
+    },
+    {
+      "epoch": 4.011418841087795,
+      "grad_norm": 2.634347677230835,
+      "learning_rate": 4.688747140848807e-06,
+      "loss": 0.3669,
+      "step": 40050
+    },
+    {
+      "epoch": 4.012420493814794,
+      "grad_norm": 2.083259344100952,
+      "learning_rate": 4.679559866017052e-06,
+      "loss": 0.4264,
+      "step": 40060
+    },
+    {
+      "epoch": 4.013422146541794,
+      "grad_norm": 2.376185655593872,
+      "learning_rate": 4.670380671446412e-06,
+      "loss": 0.379,
+      "step": 40070
+    },
+    {
+      "epoch": 4.014423799268793,
+      "grad_norm": 2.035813570022583,
+      "learning_rate": 4.661209560786922e-06,
+      "loss": 0.3986,
+      "step": 40080
+    },
+    {
+      "epoch": 4.015425451995793,
+      "grad_norm": 1.9220411777496338,
+      "learning_rate": 4.652046537685381e-06,
+      "loss": 0.3928,
+      "step": 40090
+    },
+    {
+      "epoch": 4.016427104722792,
+      "grad_norm": 2.600371837615967,
+      "learning_rate": 4.642891605785413e-06,
+      "loss": 0.3932,
+      "step": 40100
+    },
+    {
+      "epoch": 4.017428757449792,
+      "grad_norm": 1.7663367986679077,
+      "learning_rate": 4.633744768727394e-06,
+      "loss": 0.3915,
+      "step": 40110
+    },
+    {
+      "epoch": 4.018430410176792,
+      "grad_norm": 1.809329867362976,
+      "learning_rate": 4.624606030148493e-06,
+      "loss": 0.3715,
+      "step": 40120
+    },
+    {
+      "epoch": 4.019432062903792,
+      "grad_norm": 2.3340373039245605,
+      "learning_rate": 4.615475393682655e-06,
+      "loss": 0.3847,
+      "step": 40130
+    },
+    {
+      "epoch": 4.020433715630791,
+      "grad_norm": 2.9561421871185303,
+      "learning_rate": 4.606352862960606e-06,
+      "loss": 0.4141,
+      "step": 40140
+    },
+    {
+      "epoch": 4.02143536835779,
+      "grad_norm": 2.7550601959228516,
+      "learning_rate": 4.597238441609855e-06,
+      "loss": 0.3774,
+      "step": 40150
+    },
+    {
+      "epoch": 4.02243702108479,
+      "grad_norm": 2.112523317337036,
+      "learning_rate": 4.588132133254658e-06,
+      "loss": 0.3566,
+      "step": 40160
+    },
+    {
+      "epoch": 4.023438673811789,
+      "grad_norm": 2.55678653717041,
+      "learning_rate": 4.579033941516087e-06,
+      "loss": 0.4126,
+      "step": 40170
+    },
+    {
+      "epoch": 4.024440326538789,
+      "grad_norm": 2.198493719100952,
+      "learning_rate": 4.569943870011956e-06,
+      "loss": 0.3802,
+      "step": 40180
+    },
+    {
+      "epoch": 4.025441979265788,
+      "grad_norm": 2.324162006378174,
+      "learning_rate": 4.560861922356863e-06,
+      "loss": 0.4629,
+      "step": 40190
+    },
+    {
+      "epoch": 4.026443631992788,
+      "grad_norm": 1.9970417022705078,
+      "learning_rate": 4.551788102162172e-06,
+      "loss": 0.4426,
+      "step": 40200
+    },
+    {
+      "epoch": 4.0274452847197875,
+      "grad_norm": 2.143714427947998,
+      "learning_rate": 4.54272241303601e-06,
+      "loss": 0.4101,
+      "step": 40210
+    },
+    {
+      "epoch": 4.0284469374467875,
+      "grad_norm": 1.9812605381011963,
+      "learning_rate": 4.5336648585832835e-06,
+      "loss": 0.3751,
+      "step": 40220
+    },
+    {
+      "epoch": 4.029448590173787,
+      "grad_norm": 2.0631392002105713,
+      "learning_rate": 4.524615442405652e-06,
+      "loss": 0.3368,
+      "step": 40230
+    },
+    {
+      "epoch": 4.030450242900787,
+      "grad_norm": 1.7811130285263062,
+      "learning_rate": 4.5155741681015465e-06,
+      "loss": 0.3955,
+      "step": 40240
+    },
+    {
+      "epoch": 4.031451895627786,
+      "grad_norm": 2.1916298866271973,
+      "learning_rate": 4.506541039266154e-06,
+      "loss": 0.4335,
+      "step": 40250
+    },
+    {
+      "epoch": 4.032453548354785,
+      "grad_norm": 2.283879518508911,
+      "learning_rate": 4.49751605949143e-06,
+      "loss": 0.4194,
+      "step": 40260
+    },
+    {
+      "epoch": 4.033455201081785,
+      "grad_norm": 2.2963759899139404,
+      "learning_rate": 4.4884992323660835e-06,
+      "loss": 0.4499,
+      "step": 40270
+    },
+    {
+      "epoch": 4.034456853808784,
+      "grad_norm": 1.6694716215133667,
+      "learning_rate": 4.479490561475585e-06,
+      "loss": 0.4056,
+      "step": 40280
+    },
+    {
+      "epoch": 4.035458506535784,
+      "grad_norm": 2.5013225078582764,
+      "learning_rate": 4.470490050402154e-06,
+      "loss": 0.4222,
+      "step": 40290
+    },
+    {
+      "epoch": 4.036460159262783,
+      "grad_norm": 2.4829888343811035,
+      "learning_rate": 4.4614977027247924e-06,
+      "loss": 0.4054,
+      "step": 40300
+    },
+    {
+      "epoch": 4.037461811989783,
+      "grad_norm": 2.05537748336792,
+      "learning_rate": 4.452513522019214e-06,
+      "loss": 0.4458,
+      "step": 40310
+    },
+    {
+      "epoch": 4.0384634647167825,
+      "grad_norm": 2.26316499710083,
+      "learning_rate": 4.443537511857915e-06,
+      "loss": 0.4931,
+      "step": 40320
+    },
+    {
+      "epoch": 4.039465117443783,
+      "grad_norm": 2.514526605606079,
+      "learning_rate": 4.434569675810132e-06,
+      "loss": 0.4126,
+      "step": 40330
+    },
+    {
+      "epoch": 4.040466770170782,
+      "grad_norm": 1.9930896759033203,
+      "learning_rate": 4.425610017441855e-06,
+      "loss": 0.4017,
+      "step": 40340
+    },
+    {
+      "epoch": 4.041468422897782,
+      "grad_norm": 2.2072863578796387,
+      "learning_rate": 4.416658540315824e-06,
+      "loss": 0.4262,
+      "step": 40350
+    },
+    {
+      "epoch": 4.042470075624781,
+      "grad_norm": 2.070967435836792,
+      "learning_rate": 4.4077152479915115e-06,
+      "loss": 0.4335,
+      "step": 40360
+    },
+    {
+      "epoch": 4.04347172835178,
+      "grad_norm": 2.4259519577026367,
+      "learning_rate": 4.398780144025169e-06,
+      "loss": 0.3856,
+      "step": 40370
+    },
+    {
+      "epoch": 4.04447338107878,
+      "grad_norm": 2.0865633487701416,
+      "learning_rate": 4.3898532319697455e-06,
+      "loss": 0.3952,
+      "step": 40380
+    },
+    {
+      "epoch": 4.045475033805779,
+      "grad_norm": 2.032640218734741,
+      "learning_rate": 4.38093451537496e-06,
+      "loss": 0.3963,
+      "step": 40390
+    },
+    {
+      "epoch": 4.046476686532779,
+      "grad_norm": 2.034989595413208,
+      "learning_rate": 4.372023997787284e-06,
+      "loss": 0.3822,
+      "step": 40400
+    },
+    {
+      "epoch": 4.047478339259778,
+      "grad_norm": 2.2075562477111816,
+      "learning_rate": 4.363121682749907e-06,
+      "loss": 0.4047,
+      "step": 40410
+    },
+    {
+      "epoch": 4.048479991986778,
+      "grad_norm": 1.8768295049667358,
+      "learning_rate": 4.354227573802752e-06,
+      "loss": 0.3965,
+      "step": 40420
+    },
+    {
+      "epoch": 4.049481644713778,
+      "grad_norm": 1.8903871774673462,
+      "learning_rate": 4.345341674482503e-06,
+      "loss": 0.4633,
+      "step": 40430
+    },
+    {
+      "epoch": 4.050483297440778,
+      "grad_norm": 1.9325599670410156,
+      "learning_rate": 4.336463988322572e-06,
+      "loss": 0.405,
+      "step": 40440
+    },
+    {
+      "epoch": 4.051484950167777,
+      "grad_norm": 2.1987829208374023,
+      "learning_rate": 4.327594518853081e-06,
+      "loss": 0.4235,
+      "step": 40450
+    },
+    {
+      "epoch": 4.052486602894777,
+      "grad_norm": 2.2805888652801514,
+      "learning_rate": 4.318733269600919e-06,
+      "loss": 0.4807,
+      "step": 40460
+    },
+    {
+      "epoch": 4.053488255621776,
+      "grad_norm": 2.4149937629699707,
+      "learning_rate": 4.30988024408969e-06,
+      "loss": 0.4326,
+      "step": 40470
+    },
+    {
+      "epoch": 4.054489908348775,
+      "grad_norm": 1.7438945770263672,
+      "learning_rate": 4.3010354458397295e-06,
+      "loss": 0.4199,
+      "step": 40480
+    },
+    {
+      "epoch": 4.055491561075775,
+      "grad_norm": 1.9871585369110107,
+      "learning_rate": 4.292198878368086e-06,
+      "loss": 0.3935,
+      "step": 40490
+    },
+    {
+      "epoch": 4.056493213802774,
+      "grad_norm": 2.012375831604004,
+      "learning_rate": 4.2833705451885696e-06,
+      "loss": 0.3794,
+      "step": 40500
+    },
+    {
+      "epoch": 4.057494866529774,
+      "grad_norm": 2.4525644779205322,
+      "learning_rate": 4.274550449811687e-06,
+      "loss": 0.439,
+      "step": 40510
+    },
+    {
+      "epoch": 4.0584965192567735,
+      "grad_norm": 1.6219289302825928,
+      "learning_rate": 4.265738595744681e-06,
+      "loss": 0.4129,
+      "step": 40520
+    },
+    {
+      "epoch": 4.0594981719837735,
+      "grad_norm": 1.9297707080841064,
+      "learning_rate": 4.2569349864915175e-06,
+      "loss": 0.4051,
+      "step": 40530
+    },
+    {
+      "epoch": 4.060499824710773,
+      "grad_norm": 2.0720765590667725,
+      "learning_rate": 4.248139625552877e-06,
+      "loss": 0.3447,
+      "step": 40540
+    },
+    {
+      "epoch": 4.061501477437773,
+      "grad_norm": 1.5587975978851318,
+      "learning_rate": 4.239352516426167e-06,
+      "loss": 0.3451,
+      "step": 40550
+    },
+    {
+      "epoch": 4.062503130164772,
+      "grad_norm": 2.4509339332580566,
+      "learning_rate": 4.230573662605511e-06,
+      "loss": 0.4148,
+      "step": 40560
+    },
+    {
+      "epoch": 4.063504782891771,
+      "grad_norm": 2.0733141899108887,
+      "learning_rate": 4.221803067581751e-06,
+      "loss": 0.4208,
+      "step": 40570
+    },
+    {
+      "epoch": 4.064506435618771,
+      "grad_norm": 1.82527494430542,
+      "learning_rate": 4.213040734842444e-06,
+      "loss": 0.3823,
+      "step": 40580
+    },
+    {
+      "epoch": 4.06550808834577,
+      "grad_norm": 2.46215558052063,
+      "learning_rate": 4.204286667871859e-06,
+      "loss": 0.423,
+      "step": 40590
+    },
+    {
+      "epoch": 4.06650974107277,
+      "grad_norm": 2.0969185829162598,
+      "learning_rate": 4.1955408701509854e-06,
+      "loss": 0.4262,
+      "step": 40600
+    },
+    {
+      "epoch": 4.067511393799769,
+      "grad_norm": 1.8389031887054443,
+      "learning_rate": 4.186803345157517e-06,
+      "loss": 0.4392,
+      "step": 40610
+    },
+    {
+      "epoch": 4.068513046526769,
+      "grad_norm": 2.8035295009613037,
+      "learning_rate": 4.178074096365864e-06,
+      "loss": 0.4361,
+      "step": 40620
+    },
+    {
+      "epoch": 4.0695146992537685,
+      "grad_norm": 1.908402442932129,
+      "learning_rate": 4.169353127247139e-06,
+      "loss": 0.3687,
+      "step": 40630
+    },
+    {
+      "epoch": 4.0705163519807686,
+      "grad_norm": 1.9811326265335083,
+      "learning_rate": 4.160640441269168e-06,
+      "loss": 0.3685,
+      "step": 40640
+    },
+    {
+      "epoch": 4.071518004707768,
+      "grad_norm": 2.062203884124756,
+      "learning_rate": 4.151936041896482e-06,
+      "loss": 0.4121,
+      "step": 40650
+    },
+    {
+      "epoch": 4.072519657434768,
+      "grad_norm": 1.943377137184143,
+      "learning_rate": 4.143239932590312e-06,
+      "loss": 0.4364,
+      "step": 40660
+    },
+    {
+      "epoch": 4.073521310161767,
+      "grad_norm": 2.0794131755828857,
+      "learning_rate": 4.134552116808602e-06,
+      "loss": 0.3846,
+      "step": 40670
+    },
+    {
+      "epoch": 4.074522962888766,
+      "grad_norm": 2.561751365661621,
+      "learning_rate": 4.1258725980059865e-06,
+      "loss": 0.4058,
+      "step": 40680
+    },
+    {
+      "epoch": 4.075524615615766,
+      "grad_norm": 2.585106611251831,
+      "learning_rate": 4.117201379633809e-06,
+      "loss": 0.3959,
+      "step": 40690
+    },
+    {
+      "epoch": 4.076526268342765,
+      "grad_norm": 1.8804863691329956,
+      "learning_rate": 4.108538465140122e-06,
+      "loss": 0.391,
+      "step": 40700
+    },
+    {
+      "epoch": 4.077527921069765,
+      "grad_norm": 1.504548192024231,
+      "learning_rate": 4.099883857969647e-06,
+      "loss": 0.4662,
+      "step": 40710
+    },
+    {
+      "epoch": 4.078529573796764,
+      "grad_norm": 2.0218613147735596,
+      "learning_rate": 4.091237561563821e-06,
+      "loss": 0.4463,
+      "step": 40720
+    },
+    {
+      "epoch": 4.079531226523764,
+      "grad_norm": 1.8394734859466553,
+      "learning_rate": 4.082599579360794e-06,
+      "loss": 0.3977,
+      "step": 40730
+    },
+    {
+      "epoch": 4.080532879250764,
+      "grad_norm": 2.5202815532684326,
+      "learning_rate": 4.073969914795373e-06,
+      "loss": 0.4841,
+      "step": 40740
+    },
+    {
+      "epoch": 4.081534531977764,
+      "grad_norm": 2.3350720405578613,
+      "learning_rate": 4.065348571299071e-06,
+      "loss": 0.3872,
+      "step": 40750
+    },
+    {
+      "epoch": 4.082536184704763,
+      "grad_norm": 2.0294203758239746,
+      "learning_rate": 4.056735552300115e-06,
+      "loss": 0.349,
+      "step": 40760
+    },
+    {
+      "epoch": 4.083537837431763,
+      "grad_norm": 1.7993828058242798,
+      "learning_rate": 4.048130861223395e-06,
+      "loss": 0.3741,
+      "step": 40770
+    },
+    {
+      "epoch": 4.084539490158762,
+      "grad_norm": 1.6386586427688599,
+      "learning_rate": 4.0395345014904885e-06,
+      "loss": 0.3829,
+      "step": 40780
+    },
+    {
+      "epoch": 4.085541142885761,
+      "grad_norm": 2.2599360942840576,
+      "learning_rate": 4.030946476519684e-06,
+      "loss": 0.3841,
+      "step": 40790
+    },
+    {
+      "epoch": 4.086542795612761,
+      "grad_norm": 1.9661349058151245,
+      "learning_rate": 4.022366789725931e-06,
+      "loss": 0.3766,
+      "step": 40800
+    },
+    {
+      "epoch": 4.08754444833976,
+      "grad_norm": 2.0001981258392334,
+      "learning_rate": 4.0137954445208876e-06,
+      "loss": 0.4531,
+      "step": 40810
+    },
+    {
+      "epoch": 4.08854610106676,
+      "grad_norm": 1.8442391157150269,
+      "learning_rate": 4.0052324443128564e-06,
+      "loss": 0.3898,
+      "step": 40820
+    },
+    {
+      "epoch": 4.0895477537937595,
+      "grad_norm": 2.130378246307373,
+      "learning_rate": 3.99667779250687e-06,
+      "loss": 0.4561,
+      "step": 40830
+    },
+    {
+      "epoch": 4.0905494065207595,
+      "grad_norm": 2.1687450408935547,
+      "learning_rate": 3.98813149250461e-06,
+      "loss": 0.4071,
+      "step": 40840
+    },
+    {
+      "epoch": 4.091551059247759,
+      "grad_norm": 1.72993004322052,
+      "learning_rate": 3.979593547704433e-06,
+      "loss": 0.3537,
+      "step": 40850
+    },
+    {
+      "epoch": 4.092552711974759,
+      "grad_norm": 1.8413400650024414,
+      "learning_rate": 3.9710639615014e-06,
+      "loss": 0.4079,
+      "step": 40860
+    },
+    {
+      "epoch": 4.093554364701758,
+      "grad_norm": 2.123229742050171,
+      "learning_rate": 3.962542737287226e-06,
+      "loss": 0.3985,
+      "step": 40870
+    },
+    {
+      "epoch": 4.094556017428758,
+      "grad_norm": 1.9011422395706177,
+      "learning_rate": 3.954029878450311e-06,
+      "loss": 0.3487,
+      "step": 40880
+    },
+    {
+      "epoch": 4.095557670155757,
+      "grad_norm": 1.8646225929260254,
+      "learning_rate": 3.9455253883757234e-06,
+      "loss": 0.4002,
+      "step": 40890
+    },
+    {
+      "epoch": 4.096559322882756,
+      "grad_norm": 2.285891056060791,
+      "learning_rate": 3.937029270445206e-06,
+      "loss": 0.3964,
+      "step": 40900
+    },
+    {
+      "epoch": 4.097560975609756,
+      "grad_norm": 2.354170799255371,
+      "learning_rate": 3.928541528037172e-06,
+      "loss": 0.3985,
+      "step": 40910
+    },
+    {
+      "epoch": 4.098562628336755,
+      "grad_norm": 2.1918609142303467,
+      "learning_rate": 3.920062164526706e-06,
+      "loss": 0.4191,
+      "step": 40920
+    },
+    {
+      "epoch": 4.099564281063755,
+      "grad_norm": 2.174389362335205,
+      "learning_rate": 3.911591183285557e-06,
+      "loss": 0.3841,
+      "step": 40930
+    },
+    {
+      "epoch": 4.1005659337907545,
+      "grad_norm": 2.802354335784912,
+      "learning_rate": 3.903128587682147e-06,
+      "loss": 0.4132,
+      "step": 40940
+    },
+    {
+      "epoch": 4.1015675865177545,
+      "grad_norm": 1.9637892246246338,
+      "learning_rate": 3.894674381081556e-06,
+      "loss": 0.3879,
+      "step": 40950
+    },
+    {
+      "epoch": 4.102569239244754,
+      "grad_norm": 2.0116653442382812,
+      "learning_rate": 3.886228566845534e-06,
+      "loss": 0.4091,
+      "step": 40960
+    },
+    {
+      "epoch": 4.103570891971754,
+      "grad_norm": 2.062088966369629,
+      "learning_rate": 3.877791148332491e-06,
+      "loss": 0.4185,
+      "step": 40970
+    },
+    {
+      "epoch": 4.104572544698753,
+      "grad_norm": 2.834785223007202,
+      "learning_rate": 3.8693621288975e-06,
+      "loss": 0.3921,
+      "step": 40980
+    },
+    {
+      "epoch": 4.105574197425753,
+      "grad_norm": 1.9823628664016724,
+      "learning_rate": 3.860941511892291e-06,
+      "loss": 0.4165,
+      "step": 40990
+    },
+    {
+      "epoch": 4.106575850152752,
+      "grad_norm": 2.0273165702819824,
+      "learning_rate": 3.852529300665259e-06,
+      "loss": 0.4596,
+      "step": 41000
+    },
+    {
+      "epoch": 4.107577502879751,
+      "grad_norm": 1.7067807912826538,
+      "learning_rate": 3.844125498561449e-06,
+      "loss": 0.4131,
+      "step": 41010
+    },
+    {
+      "epoch": 4.108579155606751,
+      "grad_norm": 2.307382345199585,
+      "learning_rate": 3.8357301089225696e-06,
+      "loss": 0.4735,
+      "step": 41020
+    },
+    {
+      "epoch": 4.10958080833375,
+      "grad_norm": 2.346848964691162,
+      "learning_rate": 3.827343135086978e-06,
+      "loss": 0.4004,
+      "step": 41030
+    },
+    {
+      "epoch": 4.11058246106075,
+      "grad_norm": 2.429508686065674,
+      "learning_rate": 3.818964580389686e-06,
+      "loss": 0.4555,
+      "step": 41040
+    },
+    {
+      "epoch": 4.11158411378775,
+      "grad_norm": 2.165010929107666,
+      "learning_rate": 3.8105944481623578e-06,
+      "loss": 0.3992,
+      "step": 41050
+    },
+    {
+      "epoch": 4.11258576651475,
+      "grad_norm": 2.037492513656616,
+      "learning_rate": 3.802232741733325e-06,
+      "loss": 0.4485,
+      "step": 41060
+    },
+    {
+      "epoch": 4.113587419241749,
+      "grad_norm": 2.500692367553711,
+      "learning_rate": 3.7938794644275355e-06,
+      "loss": 0.4451,
+      "step": 41070
+    },
+    {
+      "epoch": 4.114589071968749,
+      "grad_norm": 1.6784147024154663,
+      "learning_rate": 3.7855346195666027e-06,
+      "loss": 0.3731,
+      "step": 41080
+    },
+    {
+      "epoch": 4.115590724695748,
+      "grad_norm": 2.7670488357543945,
+      "learning_rate": 3.777198210468802e-06,
+      "loss": 0.3972,
+      "step": 41090
+    },
+    {
+      "epoch": 4.116592377422748,
+      "grad_norm": 2.1402854919433594,
+      "learning_rate": 3.7688702404490406e-06,
+      "loss": 0.4219,
+      "step": 41100
+    },
+    {
+      "epoch": 4.117594030149747,
+      "grad_norm": 2.4333910942077637,
+      "learning_rate": 3.760550712818847e-06,
+      "loss": 0.3754,
+      "step": 41110
+    },
+    {
+      "epoch": 4.118595682876746,
+      "grad_norm": 2.354822874069214,
+      "learning_rate": 3.7522396308864367e-06,
+      "loss": 0.401,
+      "step": 41120
+    },
+    {
+      "epoch": 4.119597335603746,
+      "grad_norm": 1.6452901363372803,
+      "learning_rate": 3.743936997956646e-06,
+      "loss": 0.3948,
+      "step": 41130
+    },
+    {
+      "epoch": 4.1205989883307454,
+      "grad_norm": 2.6851041316986084,
+      "learning_rate": 3.7356428173309337e-06,
+      "loss": 0.4247,
+      "step": 41140
+    },
+    {
+      "epoch": 4.1216006410577455,
+      "grad_norm": 2.218315839767456,
+      "learning_rate": 3.7273570923074207e-06,
+      "loss": 0.3556,
+      "step": 41150
+    },
+    {
+      "epoch": 4.122602293784745,
+      "grad_norm": 2.302386522293091,
+      "learning_rate": 3.7190798261808657e-06,
+      "loss": 0.3815,
+      "step": 41160
+    },
+    {
+      "epoch": 4.123603946511745,
+      "grad_norm": 2.2736644744873047,
+      "learning_rate": 3.7108110222426636e-06,
+      "loss": 0.325,
+      "step": 41170
+    },
+    {
+      "epoch": 4.124605599238744,
+      "grad_norm": 1.8379311561584473,
+      "learning_rate": 3.7025506837808173e-06,
+      "loss": 0.3882,
+      "step": 41180
+    },
+    {
+      "epoch": 4.125607251965744,
+      "grad_norm": 1.8531453609466553,
+      "learning_rate": 3.6942988140800023e-06,
+      "loss": 0.382,
+      "step": 41190
+    },
+    {
+      "epoch": 4.126608904692743,
+      "grad_norm": 2.105767011642456,
+      "learning_rate": 3.686055416421508e-06,
+      "loss": 0.4276,
+      "step": 41200
+    },
+    {
+      "epoch": 4.127610557419743,
+      "grad_norm": 1.9618855714797974,
+      "learning_rate": 3.6778204940832394e-06,
+      "loss": 0.3956,
+      "step": 41210
+    },
+    {
+      "epoch": 4.128612210146742,
+      "grad_norm": 1.891269326210022,
+      "learning_rate": 3.6695940503397667e-06,
+      "loss": 0.3889,
+      "step": 41220
+    },
+    {
+      "epoch": 4.129613862873741,
+      "grad_norm": 1.8943488597869873,
+      "learning_rate": 3.661376088462262e-06,
+      "loss": 0.4319,
+      "step": 41230
+    },
+    {
+      "epoch": 4.130615515600741,
+      "grad_norm": 2.5317161083221436,
+      "learning_rate": 3.6531666117185335e-06,
+      "loss": 0.4331,
+      "step": 41240
+    },
+    {
+      "epoch": 4.1316171683277405,
+      "grad_norm": 1.882836937904358,
+      "learning_rate": 3.644965623373012e-06,
+      "loss": 0.4,
+      "step": 41250
+    },
+    {
+      "epoch": 4.1326188210547405,
+      "grad_norm": 1.9603805541992188,
+      "learning_rate": 3.636773126686757e-06,
+      "loss": 0.3485,
+      "step": 41260
+    },
+    {
+      "epoch": 4.13362047378174,
+      "grad_norm": 2.294837236404419,
+      "learning_rate": 3.6285891249174504e-06,
+      "loss": 0.4074,
+      "step": 41270
+    },
+    {
+      "epoch": 4.13462212650874,
+      "grad_norm": 2.164186477661133,
+      "learning_rate": 3.6204136213193935e-06,
+      "loss": 0.4174,
+      "step": 41280
+    },
+    {
+      "epoch": 4.135623779235739,
+      "grad_norm": 1.9150091409683228,
+      "learning_rate": 3.6122466191435112e-06,
+      "loss": 0.4222,
+      "step": 41290
+    },
+    {
+      "epoch": 4.136625431962739,
+      "grad_norm": 2.168137550354004,
+      "learning_rate": 3.6040881216373444e-06,
+      "loss": 0.4368,
+      "step": 41300
+    },
+    {
+      "epoch": 4.137627084689738,
+      "grad_norm": 2.411673069000244,
+      "learning_rate": 3.5959381320450596e-06,
+      "loss": 0.4143,
+      "step": 41310
+    },
+    {
+      "epoch": 4.138628737416737,
+      "grad_norm": 1.9745694398880005,
+      "learning_rate": 3.5877966536074283e-06,
+      "loss": 0.3476,
+      "step": 41320
+    },
+    {
+      "epoch": 4.139630390143737,
+      "grad_norm": 1.8774809837341309,
+      "learning_rate": 3.579663689561852e-06,
+      "loss": 0.3927,
+      "step": 41330
+    },
+    {
+      "epoch": 4.140632042870736,
+      "grad_norm": 1.9254579544067383,
+      "learning_rate": 3.5715392431423356e-06,
+      "loss": 0.3869,
+      "step": 41340
+    },
+    {
+      "epoch": 4.141633695597736,
+      "grad_norm": 2.5691428184509277,
+      "learning_rate": 3.563423317579498e-06,
+      "loss": 0.4281,
+      "step": 41350
+    },
+    {
+      "epoch": 4.142635348324736,
+      "grad_norm": 2.32572340965271,
+      "learning_rate": 3.5553159161005765e-06,
+      "loss": 0.4107,
+      "step": 41360
+    },
+    {
+      "epoch": 4.143637001051736,
+      "grad_norm": 2.5501484870910645,
+      "learning_rate": 3.547217041929413e-06,
+      "loss": 0.4378,
+      "step": 41370
+    },
+    {
+      "epoch": 4.144638653778735,
+      "grad_norm": 2.0283477306365967,
+      "learning_rate": 3.53912669828646e-06,
+      "loss": 0.3957,
+      "step": 41380
+    },
+    {
+      "epoch": 4.145640306505735,
+      "grad_norm": 2.043149709701538,
+      "learning_rate": 3.531044888388779e-06,
+      "loss": 0.4236,
+      "step": 41390
+    },
+    {
+      "epoch": 4.146641959232734,
+      "grad_norm": 1.4444963932037354,
+      "learning_rate": 3.5229716154500354e-06,
+      "loss": 0.3652,
+      "step": 41400
+    },
+    {
+      "epoch": 4.147643611959734,
+      "grad_norm": 2.416794776916504,
+      "learning_rate": 3.5149068826804993e-06,
+      "loss": 0.4204,
+      "step": 41410
+    },
+    {
+      "epoch": 4.148645264686733,
+      "grad_norm": 1.7935189008712769,
+      "learning_rate": 3.5068506932870616e-06,
+      "loss": 0.344,
+      "step": 41420
+    },
+    {
+      "epoch": 4.149646917413732,
+      "grad_norm": 2.3463051319122314,
+      "learning_rate": 3.498803050473187e-06,
+      "loss": 0.3847,
+      "step": 41430
+    },
+    {
+      "epoch": 4.150648570140732,
+      "grad_norm": 2.427424669265747,
+      "learning_rate": 3.490763957438953e-06,
+      "loss": 0.4119,
+      "step": 41440
+    },
+    {
+      "epoch": 4.151650222867731,
+      "grad_norm": 1.8780012130737305,
+      "learning_rate": 3.4827334173810565e-06,
+      "loss": 0.3799,
+      "step": 41450
+    },
+    {
+      "epoch": 4.1526518755947315,
+      "grad_norm": 2.543027639389038,
+      "learning_rate": 3.4747114334927777e-06,
+      "loss": 0.3964,
+      "step": 41460
+    },
+    {
+      "epoch": 4.153653528321731,
+      "grad_norm": 1.5053480863571167,
+      "learning_rate": 3.4666980089639783e-06,
+      "loss": 0.4012,
+      "step": 41470
+    },
+    {
+      "epoch": 4.154655181048731,
+      "grad_norm": 1.9852502346038818,
+      "learning_rate": 3.4586931469811373e-06,
+      "loss": 0.3834,
+      "step": 41480
+    },
+    {
+      "epoch": 4.15565683377573,
+      "grad_norm": 2.525178909301758,
+      "learning_rate": 3.450696850727339e-06,
+      "loss": 0.4066,
+      "step": 41490
+    },
+    {
+      "epoch": 4.15665848650273,
+      "grad_norm": 2.6303272247314453,
+      "learning_rate": 3.442709123382232e-06,
+      "loss": 0.3694,
+      "step": 41500
+    },
+    {
+      "epoch": 4.157660139229729,
+      "grad_norm": 1.9393866062164307,
+      "learning_rate": 3.4347299681220716e-06,
+      "loss": 0.4593,
+      "step": 41510
+    },
+    {
+      "epoch": 4.158661791956729,
+      "grad_norm": 1.95452082157135,
+      "learning_rate": 3.426759388119716e-06,
+      "loss": 0.4154,
+      "step": 41520
+    },
+    {
+      "epoch": 4.159663444683728,
+      "grad_norm": 2.5059969425201416,
+      "learning_rate": 3.4187973865446005e-06,
+      "loss": 0.4373,
+      "step": 41530
+    },
+    {
+      "epoch": 4.160665097410727,
+      "grad_norm": 2.185824394226074,
+      "learning_rate": 3.410843966562741e-06,
+      "loss": 0.3905,
+      "step": 41540
+    },
+    {
+      "epoch": 4.161666750137727,
+      "grad_norm": 2.5035550594329834,
+      "learning_rate": 3.402899131336762e-06,
+      "loss": 0.4118,
+      "step": 41550
+    },
+    {
+      "epoch": 4.1626684028647265,
+      "grad_norm": 1.755468726158142,
+      "learning_rate": 3.394962884025862e-06,
+      "loss": 0.4559,
+      "step": 41560
+    },
+    {
+      "epoch": 4.1636700555917265,
+      "grad_norm": 2.0787034034729004,
+      "learning_rate": 3.387035227785826e-06,
+      "loss": 0.4125,
+      "step": 41570
+    },
+    {
+      "epoch": 4.164671708318726,
+      "grad_norm": 2.5473291873931885,
+      "learning_rate": 3.3791161657690225e-06,
+      "loss": 0.4438,
+      "step": 41580
+    },
+    {
+      "epoch": 4.165673361045726,
+      "grad_norm": 2.753129243850708,
+      "learning_rate": 3.3712057011244023e-06,
+      "loss": 0.4606,
+      "step": 41590
+    },
+    {
+      "epoch": 4.166675013772725,
+      "grad_norm": 2.3650059700012207,
+      "learning_rate": 3.363303836997503e-06,
+      "loss": 0.3975,
+      "step": 41600
+    },
+    {
+      "epoch": 4.167676666499725,
+      "grad_norm": 2.2438313961029053,
+      "learning_rate": 3.3554105765304323e-06,
+      "loss": 0.4134,
+      "step": 41610
+    },
+    {
+      "epoch": 4.168678319226724,
+      "grad_norm": 2.5630788803100586,
+      "learning_rate": 3.3475259228618856e-06,
+      "loss": 0.3901,
+      "step": 41620
+    },
+    {
+      "epoch": 4.169679971953723,
+      "grad_norm": 1.7865986824035645,
+      "learning_rate": 3.3396498791271324e-06,
+      "loss": 0.431,
+      "step": 41630
+    },
+    {
+      "epoch": 4.170681624680723,
+      "grad_norm": 2.134347915649414,
+      "learning_rate": 3.3317824484580177e-06,
+      "loss": 0.4261,
+      "step": 41640
+    },
+    {
+      "epoch": 4.171683277407722,
+      "grad_norm": 1.7214415073394775,
+      "learning_rate": 3.3239236339829645e-06,
+      "loss": 0.4616,
+      "step": 41650
+    },
+    {
+      "epoch": 4.172684930134722,
+      "grad_norm": 1.7558671236038208,
+      "learning_rate": 3.3160734388269666e-06,
+      "loss": 0.4491,
+      "step": 41660
+    },
+    {
+      "epoch": 4.173686582861722,
+      "grad_norm": 2.4053850173950195,
+      "learning_rate": 3.308231866111591e-06,
+      "loss": 0.415,
+      "step": 41670
+    },
+    {
+      "epoch": 4.174688235588722,
+      "grad_norm": 2.2763030529022217,
+      "learning_rate": 3.300398918954978e-06,
+      "loss": 0.4713,
+      "step": 41680
+    },
+    {
+      "epoch": 4.175689888315721,
+      "grad_norm": 2.562437057495117,
+      "learning_rate": 3.2925746004718344e-06,
+      "loss": 0.4066,
+      "step": 41690
+    },
+    {
+      "epoch": 4.176691541042721,
+      "grad_norm": 1.9022046327590942,
+      "learning_rate": 3.284758913773442e-06,
+      "loss": 0.369,
+      "step": 41700
+    },
+    {
+      "epoch": 4.17769319376972,
+      "grad_norm": 1.930850625038147,
+      "learning_rate": 3.276951861967642e-06,
+      "loss": 0.4493,
+      "step": 41710
+    },
+    {
+      "epoch": 4.17869484649672,
+      "grad_norm": 2.1526312828063965,
+      "learning_rate": 3.269153448158846e-06,
+      "loss": 0.4159,
+      "step": 41720
+    },
+    {
+      "epoch": 4.179696499223719,
+      "grad_norm": 2.301459550857544,
+      "learning_rate": 3.261363675448037e-06,
+      "loss": 0.4742,
+      "step": 41730
+    },
+    {
+      "epoch": 4.180698151950718,
+      "grad_norm": 2.6906955242156982,
+      "learning_rate": 3.253582546932746e-06,
+      "loss": 0.4189,
+      "step": 41740
+    },
+    {
+      "epoch": 4.181699804677718,
+      "grad_norm": 1.7854713201522827,
+      "learning_rate": 3.2458100657070916e-06,
+      "loss": 0.4461,
+      "step": 41750
+    },
+    {
+      "epoch": 4.182701457404717,
+      "grad_norm": 2.6074023246765137,
+      "learning_rate": 3.2380462348617272e-06,
+      "loss": 0.4518,
+      "step": 41760
+    },
+    {
+      "epoch": 4.1837031101317175,
+      "grad_norm": 2.636751174926758,
+      "learning_rate": 3.230291057483875e-06,
+      "loss": 0.4189,
+      "step": 41770
+    },
+    {
+      "epoch": 4.184704762858717,
+      "grad_norm": 2.3315718173980713,
+      "learning_rate": 3.2225445366573376e-06,
+      "loss": 0.3592,
+      "step": 41780
+    },
+    {
+      "epoch": 4.185706415585717,
+      "grad_norm": 2.527083158493042,
+      "learning_rate": 3.214806675462442e-06,
+      "loss": 0.4188,
+      "step": 41790
+    },
+    {
+      "epoch": 4.186708068312716,
+      "grad_norm": 2.1696078777313232,
+      "learning_rate": 3.2070774769760892e-06,
+      "loss": 0.4505,
+      "step": 41800
+    },
+    {
+      "epoch": 4.187709721039716,
+      "grad_norm": 2.4959564208984375,
+      "learning_rate": 3.199356944271728e-06,
+      "loss": 0.4,
+      "step": 41810
+    },
+    {
+      "epoch": 4.188711373766715,
+      "grad_norm": 1.9425634145736694,
+      "learning_rate": 3.1916450804193865e-06,
+      "loss": 0.3721,
+      "step": 41820
+    },
+    {
+      "epoch": 4.189713026493715,
+      "grad_norm": 2.789562702178955,
+      "learning_rate": 3.1839418884856057e-06,
+      "loss": 0.3906,
+      "step": 41830
+    },
+    {
+      "epoch": 4.190714679220714,
+      "grad_norm": 2.2337393760681152,
+      "learning_rate": 3.176247371533503e-06,
+      "loss": 0.4394,
+      "step": 41840
+    },
+    {
+      "epoch": 4.191716331947713,
+      "grad_norm": 2.4459781646728516,
+      "learning_rate": 3.168561532622749e-06,
+      "loss": 0.4465,
+      "step": 41850
+    },
+    {
+      "epoch": 4.192717984674713,
+      "grad_norm": 2.3486621379852295,
+      "learning_rate": 3.1608843748095585e-06,
+      "loss": 0.3933,
+      "step": 41860
+    },
+    {
+      "epoch": 4.1937196374017125,
+      "grad_norm": 2.227614402770996,
+      "learning_rate": 3.1532159011466724e-06,
+      "loss": 0.4286,
+      "step": 41870
+    },
+    {
+      "epoch": 4.1947212901287125,
+      "grad_norm": 1.9478914737701416,
+      "learning_rate": 3.1455561146834178e-06,
+      "loss": 0.4396,
+      "step": 41880
+    },
+    {
+      "epoch": 4.195722942855712,
+      "grad_norm": 2.1462783813476562,
+      "learning_rate": 3.1379050184656477e-06,
+      "loss": 0.3786,
+      "step": 41890
+    },
+    {
+      "epoch": 4.196724595582712,
+      "grad_norm": 2.016328811645508,
+      "learning_rate": 3.1302626155357423e-06,
+      "loss": 0.462,
+      "step": 41900
+    },
+    {
+      "epoch": 4.197726248309711,
+      "grad_norm": 1.9448295831680298,
+      "learning_rate": 3.1226289089326593e-06,
+      "loss": 0.4233,
+      "step": 41910
+    },
+    {
+      "epoch": 4.198727901036711,
+      "grad_norm": 1.9269777536392212,
+      "learning_rate": 3.115003901691871e-06,
+      "loss": 0.3689,
+      "step": 41920
+    },
+    {
+      "epoch": 4.19972955376371,
+      "grad_norm": 2.250922679901123,
+      "learning_rate": 3.1073875968454096e-06,
+      "loss": 0.394,
+      "step": 41930
+    },
+    {
+      "epoch": 4.20073120649071,
+      "grad_norm": 2.7293055057525635,
+      "learning_rate": 3.099779997421831e-06,
+      "loss": 0.359,
+      "step": 41940
+    },
+    {
+      "epoch": 4.201732859217709,
+      "grad_norm": 2.0588979721069336,
+      "learning_rate": 3.0921811064462374e-06,
+      "loss": 0.352,
+      "step": 41950
+    },
+    {
+      "epoch": 4.202734511944708,
+      "grad_norm": 1.7878366708755493,
+      "learning_rate": 3.0845909269402756e-06,
+      "loss": 0.3826,
+      "step": 41960
+    },
+    {
+      "epoch": 4.203736164671708,
+      "grad_norm": 2.4124763011932373,
+      "learning_rate": 3.0770094619221036e-06,
+      "loss": 0.4297,
+      "step": 41970
+    },
+    {
+      "epoch": 4.2047378173987076,
+      "grad_norm": 1.2512212991714478,
+      "learning_rate": 3.0694367144064456e-06,
+      "loss": 0.4079,
+      "step": 41980
+    },
+    {
+      "epoch": 4.205739470125708,
+      "grad_norm": 1.5953153371810913,
+      "learning_rate": 3.061872687404538e-06,
+      "loss": 0.3754,
+      "step": 41990
+    },
+    {
+      "epoch": 4.206741122852707,
+      "grad_norm": 2.8178634643554688,
+      "learning_rate": 3.0543173839241546e-06,
+      "loss": 0.4456,
+      "step": 42000
+    },
+    {
+      "epoch": 4.207742775579707,
+      "grad_norm": 2.2212166786193848,
+      "learning_rate": 3.0467708069696037e-06,
+      "loss": 0.3952,
+      "step": 42010
+    },
+    {
+      "epoch": 4.208744428306706,
+      "grad_norm": 2.113723039627075,
+      "learning_rate": 3.039232959541724e-06,
+      "loss": 0.3944,
+      "step": 42020
+    },
+    {
+      "epoch": 4.209746081033706,
+      "grad_norm": 2.7201507091522217,
+      "learning_rate": 3.031703844637876e-06,
+      "loss": 0.3985,
+      "step": 42030
+    },
+    {
+      "epoch": 4.210747733760705,
+      "grad_norm": 2.516010284423828,
+      "learning_rate": 3.024183465251956e-06,
+      "loss": 0.4158,
+      "step": 42040
+    },
+    {
+      "epoch": 4.211749386487705,
+      "grad_norm": 2.367647647857666,
+      "learning_rate": 3.0166718243743785e-06,
+      "loss": 0.3951,
+      "step": 42050
+    },
+    {
+      "epoch": 4.212751039214704,
+      "grad_norm": 1.4118751287460327,
+      "learning_rate": 3.0091689249920923e-06,
+      "loss": 0.3785,
+      "step": 42060
+    },
+    {
+      "epoch": 4.213752691941703,
+      "grad_norm": 2.2263405323028564,
+      "learning_rate": 3.001674770088564e-06,
+      "loss": 0.3853,
+      "step": 42070
+    },
+    {
+      "epoch": 4.2147543446687035,
+      "grad_norm": 1.7431256771087646,
+      "learning_rate": 2.9941893626437817e-06,
+      "loss": 0.3816,
+      "step": 42080
+    },
+    {
+      "epoch": 4.215755997395703,
+      "grad_norm": 1.8744088411331177,
+      "learning_rate": 2.986712705634262e-06,
+      "loss": 0.4145,
+      "step": 42090
+    },
+    {
+      "epoch": 4.216757650122703,
+      "grad_norm": 1.928413987159729,
+      "learning_rate": 2.979244802033032e-06,
+      "loss": 0.3944,
+      "step": 42100
+    },
+    {
+      "epoch": 4.217759302849702,
+      "grad_norm": 2.396594762802124,
+      "learning_rate": 2.971785654809656e-06,
+      "loss": 0.4667,
+      "step": 42110
+    },
+    {
+      "epoch": 4.218760955576702,
+      "grad_norm": 2.273517608642578,
+      "learning_rate": 2.9643352669301933e-06,
+      "loss": 0.4142,
+      "step": 42120
+    },
+    {
+      "epoch": 4.219762608303701,
+      "grad_norm": 1.8810628652572632,
+      "learning_rate": 2.9568936413572363e-06,
+      "loss": 0.3943,
+      "step": 42130
+    },
+    {
+      "epoch": 4.220764261030701,
+      "grad_norm": 2.027733325958252,
+      "learning_rate": 2.9494607810498786e-06,
+      "loss": 0.4632,
+      "step": 42140
+    },
+    {
+      "epoch": 4.2217659137577,
+      "grad_norm": 2.457437753677368,
+      "learning_rate": 2.9420366889637592e-06,
+      "loss": 0.439,
+      "step": 42150
+    },
+    {
+      "epoch": 4.2227675664847,
+      "grad_norm": 2.446218252182007,
+      "learning_rate": 2.9346213680509882e-06,
+      "loss": 0.3905,
+      "step": 42160
+    },
+    {
+      "epoch": 4.223769219211699,
+      "grad_norm": 2.112877368927002,
+      "learning_rate": 2.927214821260213e-06,
+      "loss": 0.3998,
+      "step": 42170
+    },
+    {
+      "epoch": 4.2247708719386985,
+      "grad_norm": 2.958712577819824,
+      "learning_rate": 2.9198170515366023e-06,
+      "loss": 0.4534,
+      "step": 42180
+    },
+    {
+      "epoch": 4.2257725246656985,
+      "grad_norm": 2.4029061794281006,
+      "learning_rate": 2.9124280618218035e-06,
+      "loss": 0.4766,
+      "step": 42190
+    },
+    {
+      "epoch": 4.226774177392698,
+      "grad_norm": 2.6346077919006348,
+      "learning_rate": 2.905047855053991e-06,
+      "loss": 0.4245,
+      "step": 42200
+    },
+    {
+      "epoch": 4.227775830119698,
+      "grad_norm": 2.153205156326294,
+      "learning_rate": 2.8976764341678536e-06,
+      "loss": 0.4395,
+      "step": 42210
+    },
+    {
+      "epoch": 4.228777482846697,
+      "grad_norm": 1.5752555131912231,
+      "learning_rate": 2.8903138020945797e-06,
+      "loss": 0.3512,
+      "step": 42220
+    },
+    {
+      "epoch": 4.229779135573697,
+      "grad_norm": 2.1180450916290283,
+      "learning_rate": 2.882959961761847e-06,
+      "loss": 0.4136,
+      "step": 42230
+    },
+    {
+      "epoch": 4.230780788300696,
+      "grad_norm": 2.0195274353027344,
+      "learning_rate": 2.8756149160938635e-06,
+      "loss": 0.42,
+      "step": 42240
+    },
+    {
+      "epoch": 4.231782441027696,
+      "grad_norm": 2.2286198139190674,
+      "learning_rate": 2.8682786680113314e-06,
+      "loss": 0.42,
+      "step": 42250
+    },
+    {
+      "epoch": 4.232784093754695,
+      "grad_norm": 2.2767393589019775,
+      "learning_rate": 2.860951220431435e-06,
+      "loss": 0.4883,
+      "step": 42260
+    },
+    {
+      "epoch": 4.233785746481694,
+      "grad_norm": 2.395681858062744,
+      "learning_rate": 2.8536325762678895e-06,
+      "loss": 0.4004,
+      "step": 42270
+    },
+    {
+      "epoch": 4.234787399208694,
+      "grad_norm": 2.077942371368408,
+      "learning_rate": 2.8463227384308898e-06,
+      "loss": 0.4004,
+      "step": 42280
+    },
+    {
+      "epoch": 4.2357890519356935,
+      "grad_norm": 2.383152723312378,
+      "learning_rate": 2.8390217098271414e-06,
+      "loss": 0.422,
+      "step": 42290
+    },
+    {
+      "epoch": 4.236790704662694,
+      "grad_norm": 1.919995903968811,
+      "learning_rate": 2.8317294933598277e-06,
+      "loss": 0.391,
+      "step": 42300
+    },
+    {
+      "epoch": 4.237792357389693,
+      "grad_norm": 2.3750290870666504,
+      "learning_rate": 2.824446091928651e-06,
+      "loss": 0.4244,
+      "step": 42310
+    },
+    {
+      "epoch": 4.238794010116693,
+      "grad_norm": 2.0089025497436523,
+      "learning_rate": 2.8171715084297916e-06,
+      "loss": 0.381,
+      "step": 42320
+    },
+    {
+      "epoch": 4.239795662843692,
+      "grad_norm": 2.8085126876831055,
+      "learning_rate": 2.809905745755936e-06,
+      "loss": 0.3842,
+      "step": 42330
+    },
+    {
+      "epoch": 4.240797315570692,
+      "grad_norm": 2.5227251052856445,
+      "learning_rate": 2.8026488067962515e-06,
+      "loss": 0.4056,
+      "step": 42340
+    },
+    {
+      "epoch": 4.241798968297691,
+      "grad_norm": 2.0635979175567627,
+      "learning_rate": 2.7954006944364063e-06,
+      "loss": 0.4036,
+      "step": 42350
+    },
+    {
+      "epoch": 4.242800621024691,
+      "grad_norm": 2.093088150024414,
+      "learning_rate": 2.7881614115585485e-06,
+      "loss": 0.3837,
+      "step": 42360
+    },
+    {
+      "epoch": 4.24380227375169,
+      "grad_norm": 3.117953300476074,
+      "learning_rate": 2.780930961041328e-06,
+      "loss": 0.414,
+      "step": 42370
+    },
+    {
+      "epoch": 4.244803926478689,
+      "grad_norm": 2.0712902545928955,
+      "learning_rate": 2.77370934575987e-06,
+      "loss": 0.4608,
+      "step": 42380
+    },
+    {
+      "epoch": 4.2458055792056895,
+      "grad_norm": 2.027517080307007,
+      "learning_rate": 2.766496568585797e-06,
+      "loss": 0.4196,
+      "step": 42390
+    },
+    {
+      "epoch": 4.246807231932689,
+      "grad_norm": 2.0849087238311768,
+      "learning_rate": 2.7592926323872088e-06,
+      "loss": 0.4262,
+      "step": 42400
+    },
+    {
+      "epoch": 4.247808884659689,
+      "grad_norm": 1.7223390340805054,
+      "learning_rate": 2.7520975400286973e-06,
+      "loss": 0.3834,
+      "step": 42410
+    },
+    {
+      "epoch": 4.248810537386688,
+      "grad_norm": 2.2863972187042236,
+      "learning_rate": 2.7449112943713272e-06,
+      "loss": 0.4206,
+      "step": 42420
+    },
+    {
+      "epoch": 4.249812190113688,
+      "grad_norm": 1.9107893705368042,
+      "learning_rate": 2.73773389827266e-06,
+      "loss": 0.4077,
+      "step": 42430
+    },
+    {
+      "epoch": 4.250813842840687,
+      "grad_norm": 1.7265775203704834,
+      "learning_rate": 2.730565354586723e-06,
+      "loss": 0.4015,
+      "step": 42440
+    },
+    {
+      "epoch": 4.251815495567687,
+      "grad_norm": 2.400996208190918,
+      "learning_rate": 2.7234056661640335e-06,
+      "loss": 0.4373,
+      "step": 42450
+    },
+    {
+      "epoch": 4.252817148294686,
+      "grad_norm": 2.8759820461273193,
+      "learning_rate": 2.7162548358515865e-06,
+      "loss": 0.4431,
+      "step": 42460
+    },
+    {
+      "epoch": 4.253818801021686,
+      "grad_norm": 2.009953737258911,
+      "learning_rate": 2.709112866492847e-06,
+      "loss": 0.3766,
+      "step": 42470
+    },
+    {
+      "epoch": 4.254820453748685,
+      "grad_norm": 2.1563308238983154,
+      "learning_rate": 2.7019797609277696e-06,
+      "loss": 0.4626,
+      "step": 42480
+    },
+    {
+      "epoch": 4.2558221064756845,
+      "grad_norm": 2.7688074111938477,
+      "learning_rate": 2.69485552199277e-06,
+      "loss": 0.4418,
+      "step": 42490
+    },
+    {
+      "epoch": 4.2568237592026845,
+      "grad_norm": 2.4989662170410156,
+      "learning_rate": 2.6877401525207458e-06,
+      "loss": 0.39,
+      "step": 42500
+    },
+    {
+      "epoch": 4.257825411929684,
+      "grad_norm": 2.268092393875122,
+      "learning_rate": 2.6806336553410777e-06,
+      "loss": 0.4315,
+      "step": 42510
+    },
+    {
+      "epoch": 4.258827064656684,
+      "grad_norm": 2.241440773010254,
+      "learning_rate": 2.6735360332795917e-06,
+      "loss": 0.4064,
+      "step": 42520
+    },
+    {
+      "epoch": 4.259828717383683,
+      "grad_norm": 2.1825673580169678,
+      "learning_rate": 2.6664472891586077e-06,
+      "loss": 0.3797,
+      "step": 42530
+    },
+    {
+      "epoch": 4.260830370110683,
+      "grad_norm": 2.2951955795288086,
+      "learning_rate": 2.659367425796916e-06,
+      "loss": 0.4349,
+      "step": 42540
+    },
+    {
+      "epoch": 4.261832022837682,
+      "grad_norm": 1.7518614530563354,
+      "learning_rate": 2.6522964460097553e-06,
+      "loss": 0.3917,
+      "step": 42550
+    },
+    {
+      "epoch": 4.262833675564682,
+      "grad_norm": 2.920267343521118,
+      "learning_rate": 2.645234352608847e-06,
+      "loss": 0.412,
+      "step": 42560
+    },
+    {
+      "epoch": 4.263835328291681,
+      "grad_norm": 2.29345965385437,
+      "learning_rate": 2.638181148402383e-06,
+      "loss": 0.3755,
+      "step": 42570
+    },
+    {
+      "epoch": 4.26483698101868,
+      "grad_norm": 2.323241710662842,
+      "learning_rate": 2.631136836195014e-06,
+      "loss": 0.4501,
+      "step": 42580
+    },
+    {
+      "epoch": 4.26583863374568,
+      "grad_norm": 2.2796177864074707,
+      "learning_rate": 2.6241014187878417e-06,
+      "loss": 0.3953,
+      "step": 42590
+    },
+    {
+      "epoch": 4.2668402864726795,
+      "grad_norm": 2.3419172763824463,
+      "learning_rate": 2.6170748989784583e-06,
+      "loss": 0.3953,
+      "step": 42600
+    },
+    {
+      "epoch": 4.26784193919968,
+      "grad_norm": 2.1087424755096436,
+      "learning_rate": 2.610057279560896e-06,
+      "loss": 0.4198,
+      "step": 42610
+    },
+    {
+      "epoch": 4.268843591926679,
+      "grad_norm": 1.8118137121200562,
+      "learning_rate": 2.603048563325661e-06,
+      "loss": 0.4031,
+      "step": 42620
+    },
+    {
+      "epoch": 4.269845244653679,
+      "grad_norm": 2.0749223232269287,
+      "learning_rate": 2.5960487530597018e-06,
+      "loss": 0.4355,
+      "step": 42630
+    },
+    {
+      "epoch": 4.270846897380678,
+      "grad_norm": 2.66550612449646,
+      "learning_rate": 2.5890578515464476e-06,
+      "loss": 0.421,
+      "step": 42640
+    },
+    {
+      "epoch": 4.271848550107678,
+      "grad_norm": 1.960179090499878,
+      "learning_rate": 2.5820758615657776e-06,
+      "loss": 0.3849,
+      "step": 42650
+    },
+    {
+      "epoch": 4.272850202834677,
+      "grad_norm": 2.837324619293213,
+      "learning_rate": 2.575102785894007e-06,
+      "loss": 0.4429,
+      "step": 42660
+    },
+    {
+      "epoch": 4.273851855561677,
+      "grad_norm": 1.9536443948745728,
+      "learning_rate": 2.5681386273039426e-06,
+      "loss": 0.3737,
+      "step": 42670
+    },
+    {
+      "epoch": 4.274853508288676,
+      "grad_norm": 1.969543695449829,
+      "learning_rate": 2.5611833885648194e-06,
+      "loss": 0.4493,
+      "step": 42680
+    },
+    {
+      "epoch": 4.275855161015675,
+      "grad_norm": 2.2718586921691895,
+      "learning_rate": 2.554237072442331e-06,
+      "loss": 0.3987,
+      "step": 42690
+    },
+    {
+      "epoch": 4.2768568137426755,
+      "grad_norm": 2.5806398391723633,
+      "learning_rate": 2.5472996816986283e-06,
+      "loss": 0.4017,
+      "step": 42700
+    },
+    {
+      "epoch": 4.277858466469675,
+      "grad_norm": 1.761518120765686,
+      "learning_rate": 2.540371219092305e-06,
+      "loss": 0.3936,
+      "step": 42710
+    },
+    {
+      "epoch": 4.278860119196675,
+      "grad_norm": 2.4270286560058594,
+      "learning_rate": 2.533451687378413e-06,
+      "loss": 0.3815,
+      "step": 42720
+    },
+    {
+      "epoch": 4.279861771923674,
+      "grad_norm": 2.0313477516174316,
+      "learning_rate": 2.5265410893084485e-06,
+      "loss": 0.4022,
+      "step": 42730
+    },
+    {
+      "epoch": 4.280863424650674,
+      "grad_norm": 1.7014490365982056,
+      "learning_rate": 2.519639427630352e-06,
+      "loss": 0.4044,
+      "step": 42740
+    },
+    {
+      "epoch": 4.281865077377673,
+      "grad_norm": 2.5806448459625244,
+      "learning_rate": 2.51274670508852e-06,
+      "loss": 0.4435,
+      "step": 42750
+    },
+    {
+      "epoch": 4.282866730104673,
+      "grad_norm": 2.133436918258667,
+      "learning_rate": 2.505862924423785e-06,
+      "loss": 0.3617,
+      "step": 42760
+    },
+    {
+      "epoch": 4.283868382831672,
+      "grad_norm": 2.5142359733581543,
+      "learning_rate": 2.4989880883734273e-06,
+      "loss": 0.4125,
+      "step": 42770
+    },
+    {
+      "epoch": 4.284870035558672,
+      "grad_norm": 1.8463962078094482,
+      "learning_rate": 2.4921221996711707e-06,
+      "loss": 0.4152,
+      "step": 42780
+    },
+    {
+      "epoch": 4.285871688285671,
+      "grad_norm": 2.2918601036071777,
+      "learning_rate": 2.4852652610471842e-06,
+      "loss": 0.4054,
+      "step": 42790
+    },
+    {
+      "epoch": 4.2868733410126705,
+      "grad_norm": 2.54331374168396,
+      "learning_rate": 2.4784172752280733e-06,
+      "loss": 0.3861,
+      "step": 42800
+    },
+    {
+      "epoch": 4.2878749937396705,
+      "grad_norm": 1.8478418588638306,
+      "learning_rate": 2.471578244936881e-06,
+      "loss": 0.4161,
+      "step": 42810
+    },
+    {
+      "epoch": 4.28887664646667,
+      "grad_norm": 2.231584072113037,
+      "learning_rate": 2.4647481728931e-06,
+      "loss": 0.3572,
+      "step": 42820
+    },
+    {
+      "epoch": 4.28987829919367,
+      "grad_norm": 2.3038971424102783,
+      "learning_rate": 2.457927061812648e-06,
+      "loss": 0.3729,
+      "step": 42830
+    },
+    {
+      "epoch": 4.290879951920669,
+      "grad_norm": 1.7609002590179443,
+      "learning_rate": 2.45111491440789e-06,
+      "loss": 0.3543,
+      "step": 42840
+    },
+    {
+      "epoch": 4.291881604647669,
+      "grad_norm": 2.5485479831695557,
+      "learning_rate": 2.4443117333876194e-06,
+      "loss": 0.358,
+      "step": 42850
+    },
+    {
+      "epoch": 4.292883257374668,
+      "grad_norm": 2.667595148086548,
+      "learning_rate": 2.437517521457061e-06,
+      "loss": 0.3777,
+      "step": 42860
+    },
+    {
+      "epoch": 4.293884910101668,
+      "grad_norm": 2.162569522857666,
+      "learning_rate": 2.430732281317899e-06,
+      "loss": 0.4588,
+      "step": 42870
+    },
+    {
+      "epoch": 4.294886562828667,
+      "grad_norm": 2.4120516777038574,
+      "learning_rate": 2.4239560156682105e-06,
+      "loss": 0.4066,
+      "step": 42880
+    },
+    {
+      "epoch": 4.295888215555667,
+      "grad_norm": 2.75325345993042,
+      "learning_rate": 2.417188727202524e-06,
+      "loss": 0.4154,
+      "step": 42890
+    },
+    {
+      "epoch": 4.296889868282666,
+      "grad_norm": 1.9056499004364014,
+      "learning_rate": 2.4104304186118088e-06,
+      "loss": 0.4268,
+      "step": 42900
+    },
+    {
+      "epoch": 4.2978915210096655,
+      "grad_norm": 2.1397392749786377,
+      "learning_rate": 2.403681092583454e-06,
+      "loss": 0.3723,
+      "step": 42910
+    },
+    {
+      "epoch": 4.298893173736666,
+      "grad_norm": 2.5149714946746826,
+      "learning_rate": 2.3969407518012576e-06,
+      "loss": 0.5002,
+      "step": 42920
+    },
+    {
+      "epoch": 4.299894826463665,
+      "grad_norm": 1.9185551404953003,
+      "learning_rate": 2.390209398945478e-06,
+      "loss": 0.4161,
+      "step": 42930
+    },
+    {
+      "epoch": 4.300896479190665,
+      "grad_norm": 2.358262777328491,
+      "learning_rate": 2.383487036692786e-06,
+      "loss": 0.4237,
+      "step": 42940
+    },
+    {
+      "epoch": 4.301898131917664,
+      "grad_norm": 1.9108012914657593,
+      "learning_rate": 2.376773667716262e-06,
+      "loss": 0.4023,
+      "step": 42950
+    },
+    {
+      "epoch": 4.302899784644664,
+      "grad_norm": 2.148002862930298,
+      "learning_rate": 2.3700692946854286e-06,
+      "loss": 0.4459,
+      "step": 42960
+    },
+    {
+      "epoch": 4.303901437371663,
+      "grad_norm": 1.83115816116333,
+      "learning_rate": 2.363373920266229e-06,
+      "loss": 0.4049,
+      "step": 42970
+    },
+    {
+      "epoch": 4.304903090098663,
+      "grad_norm": 2.1353845596313477,
+      "learning_rate": 2.356687547121034e-06,
+      "loss": 0.4329,
+      "step": 42980
+    },
+    {
+      "epoch": 4.305904742825662,
+      "grad_norm": 2.241957902908325,
+      "learning_rate": 2.350010177908604e-06,
+      "loss": 0.3842,
+      "step": 42990
+    },
+    {
+      "epoch": 4.306906395552662,
+      "grad_norm": 2.3134796619415283,
+      "learning_rate": 2.343341815284164e-06,
+      "loss": 0.4726,
+      "step": 43000
+    },
+    {
+      "epoch": 4.3079080482796615,
+      "grad_norm": 1.6615601778030396,
+      "learning_rate": 2.336682461899328e-06,
+      "loss": 0.4071,
+      "step": 43010
+    },
+    {
+      "epoch": 4.308909701006661,
+      "grad_norm": 1.741769552230835,
+      "learning_rate": 2.3300321204021257e-06,
+      "loss": 0.331,
+      "step": 43020
+    },
+    {
+      "epoch": 4.309911353733661,
+      "grad_norm": 2.2427818775177,
+      "learning_rate": 2.323390793437022e-06,
+      "loss": 0.386,
+      "step": 43030
+    },
+    {
+      "epoch": 4.31091300646066,
+      "grad_norm": 3.0531322956085205,
+      "learning_rate": 2.3167584836448875e-06,
+      "loss": 0.3896,
+      "step": 43040
+    },
+    {
+      "epoch": 4.31191465918766,
+      "grad_norm": 1.993509292602539,
+      "learning_rate": 2.3101351936630047e-06,
+      "loss": 0.3983,
+      "step": 43050
+    },
+    {
+      "epoch": 4.312916311914659,
+      "grad_norm": 2.2616026401519775,
+      "learning_rate": 2.3035209261250716e-06,
+      "loss": 0.4106,
+      "step": 43060
+    },
+    {
+      "epoch": 4.313917964641659,
+      "grad_norm": 2.288247585296631,
+      "learning_rate": 2.296915683661202e-06,
+      "loss": 0.3551,
+      "step": 43070
+    },
+    {
+      "epoch": 4.314919617368658,
+      "grad_norm": 1.8383212089538574,
+      "learning_rate": 2.290319468897917e-06,
+      "loss": 0.4036,
+      "step": 43080
+    },
+    {
+      "epoch": 4.315921270095658,
+      "grad_norm": 2.345820665359497,
+      "learning_rate": 2.2837322844581454e-06,
+      "loss": 0.4122,
+      "step": 43090
+    },
+    {
+      "epoch": 4.316922922822657,
+      "grad_norm": 1.7840343713760376,
+      "learning_rate": 2.2771541329612317e-06,
+      "loss": 0.3799,
+      "step": 43100
+    },
+    {
+      "epoch": 4.317924575549657,
+      "grad_norm": 1.8226646184921265,
+      "learning_rate": 2.2705850170229246e-06,
+      "loss": 0.4293,
+      "step": 43110
+    },
+    {
+      "epoch": 4.3189262282766565,
+      "grad_norm": 1.8495908975601196,
+      "learning_rate": 2.2640249392553823e-06,
+      "loss": 0.3918,
+      "step": 43120
+    },
+    {
+      "epoch": 4.319927881003656,
+      "grad_norm": 2.4388835430145264,
+      "learning_rate": 2.257473902267165e-06,
+      "loss": 0.4003,
+      "step": 43130
+    },
+    {
+      "epoch": 4.320929533730656,
+      "grad_norm": 1.8674567937850952,
+      "learning_rate": 2.2509319086632425e-06,
+      "loss": 0.381,
+      "step": 43140
+    },
+    {
+      "epoch": 4.321931186457655,
+      "grad_norm": 1.9195691347122192,
+      "learning_rate": 2.2443989610449855e-06,
+      "loss": 0.383,
+      "step": 43150
+    },
+    {
+      "epoch": 4.322932839184655,
+      "grad_norm": 2.5186662673950195,
+      "learning_rate": 2.2378750620101667e-06,
+      "loss": 0.3618,
+      "step": 43160
+    },
+    {
+      "epoch": 4.323934491911654,
+      "grad_norm": 1.8908497095108032,
+      "learning_rate": 2.2313602141529668e-06,
+      "loss": 0.4159,
+      "step": 43170
+    },
+    {
+      "epoch": 4.324936144638654,
+      "grad_norm": 2.380554437637329,
+      "learning_rate": 2.22485442006396e-06,
+      "loss": 0.4289,
+      "step": 43180
+    },
+    {
+      "epoch": 4.325937797365653,
+      "grad_norm": 2.3493990898132324,
+      "learning_rate": 2.218357682330119e-06,
+      "loss": 0.4513,
+      "step": 43190
+    },
+    {
+      "epoch": 4.326939450092653,
+      "grad_norm": 2.776486396789551,
+      "learning_rate": 2.2118700035348328e-06,
+      "loss": 0.3659,
+      "step": 43200
+    },
+    {
+      "epoch": 4.327941102819652,
+      "grad_norm": 2.014625310897827,
+      "learning_rate": 2.2053913862578656e-06,
+      "loss": 0.3878,
+      "step": 43210
+    },
+    {
+      "epoch": 4.328942755546652,
+      "grad_norm": 2.328507423400879,
+      "learning_rate": 2.198921833075385e-06,
+      "loss": 0.4097,
+      "step": 43220
+    },
+    {
+      "epoch": 4.329944408273652,
+      "grad_norm": 2.105379104614258,
+      "learning_rate": 2.192461346559968e-06,
+      "loss": 0.3963,
+      "step": 43230
+    },
+    {
+      "epoch": 4.330946061000651,
+      "grad_norm": 2.268242120742798,
+      "learning_rate": 2.1860099292805664e-06,
+      "loss": 0.4335,
+      "step": 43240
+    },
+    {
+      "epoch": 4.331947713727651,
+      "grad_norm": 1.6855250597000122,
+      "learning_rate": 2.1795675838025333e-06,
+      "loss": 0.3711,
+      "step": 43250
+    },
+    {
+      "epoch": 4.33294936645465,
+      "grad_norm": 2.5439865589141846,
+      "learning_rate": 2.1731343126876276e-06,
+      "loss": 0.363,
+      "step": 43260
+    },
+    {
+      "epoch": 4.33395101918165,
+      "grad_norm": 1.8600542545318604,
+      "learning_rate": 2.1667101184939837e-06,
+      "loss": 0.3719,
+      "step": 43270
+    },
+    {
+      "epoch": 4.334952671908649,
+      "grad_norm": 2.798398494720459,
+      "learning_rate": 2.160295003776125e-06,
+      "loss": 0.3651,
+      "step": 43280
+    },
+    {
+      "epoch": 4.335954324635649,
+      "grad_norm": 1.7904213666915894,
+      "learning_rate": 2.153888971084969e-06,
+      "loss": 0.4151,
+      "step": 43290
+    },
+    {
+      "epoch": 4.336955977362648,
+      "grad_norm": 2.4006080627441406,
+      "learning_rate": 2.1474920229678396e-06,
+      "loss": 0.4356,
+      "step": 43300
+    },
+    {
+      "epoch": 4.337957630089648,
+      "grad_norm": 2.7240309715270996,
+      "learning_rate": 2.1411041619684186e-06,
+      "loss": 0.4338,
+      "step": 43310
+    },
+    {
+      "epoch": 4.3389592828166474,
+      "grad_norm": 1.9607129096984863,
+      "learning_rate": 2.134725390626785e-06,
+      "loss": 0.3561,
+      "step": 43320
+    },
+    {
+      "epoch": 4.3399609355436475,
+      "grad_norm": 2.6184070110321045,
+      "learning_rate": 2.1283557114794183e-06,
+      "loss": 0.4381,
+      "step": 43330
+    },
+    {
+      "epoch": 4.340962588270647,
+      "grad_norm": 1.8792089223861694,
+      "learning_rate": 2.1219951270591703e-06,
+      "loss": 0.3831,
+      "step": 43340
+    },
+    {
+      "epoch": 4.341964240997646,
+      "grad_norm": 1.6819710731506348,
+      "learning_rate": 2.1156436398952623e-06,
+      "loss": 0.3814,
+      "step": 43350
+    },
+    {
+      "epoch": 4.342965893724646,
+      "grad_norm": 2.5652620792388916,
+      "learning_rate": 2.1093012525133237e-06,
+      "loss": 0.4009,
+      "step": 43360
+    },
+    {
+      "epoch": 4.343967546451645,
+      "grad_norm": 1.6315181255340576,
+      "learning_rate": 2.102967967435354e-06,
+      "loss": 0.4335,
+      "step": 43370
+    },
+    {
+      "epoch": 4.344969199178645,
+      "grad_norm": 2.178917407989502,
+      "learning_rate": 2.0966437871797333e-06,
+      "loss": 0.4018,
+      "step": 43380
+    },
+    {
+      "epoch": 4.345970851905644,
+      "grad_norm": 1.893080711364746,
+      "learning_rate": 2.0903287142612193e-06,
+      "loss": 0.3939,
+      "step": 43390
+    },
+    {
+      "epoch": 4.346972504632644,
+      "grad_norm": 2.0967724323272705,
+      "learning_rate": 2.0840227511909504e-06,
+      "loss": 0.4255,
+      "step": 43400
+    },
+    {
+      "epoch": 4.347974157359643,
+      "grad_norm": 2.593022108078003,
+      "learning_rate": 2.077725900476446e-06,
+      "loss": 0.3903,
+      "step": 43410
+    },
+    {
+      "epoch": 4.348975810086643,
+      "grad_norm": 1.952826738357544,
+      "learning_rate": 2.071438164621595e-06,
+      "loss": 0.415,
+      "step": 43420
+    },
+    {
+      "epoch": 4.3499774628136425,
+      "grad_norm": 1.9698874950408936,
+      "learning_rate": 2.065159546126666e-06,
+      "loss": 0.3708,
+      "step": 43430
+    },
+    {
+      "epoch": 4.350979115540642,
+      "grad_norm": 2.352921962738037,
+      "learning_rate": 2.0588900474883017e-06,
+      "loss": 0.4387,
+      "step": 43440
+    },
+    {
+      "epoch": 4.351980768267642,
+      "grad_norm": 2.0334842205047607,
+      "learning_rate": 2.0526296711995184e-06,
+      "loss": 0.4526,
+      "step": 43450
+    },
+    {
+      "epoch": 4.352982420994641,
+      "grad_norm": 2.4401466846466064,
+      "learning_rate": 2.046378419749706e-06,
+      "loss": 0.3585,
+      "step": 43460
+    },
+    {
+      "epoch": 4.353984073721641,
+      "grad_norm": 2.077836751937866,
+      "learning_rate": 2.0401362956246256e-06,
+      "loss": 0.4503,
+      "step": 43470
+    },
+    {
+      "epoch": 4.35498572644864,
+      "grad_norm": 2.017932415008545,
+      "learning_rate": 2.033903301306403e-06,
+      "loss": 0.3723,
+      "step": 43480
+    },
+    {
+      "epoch": 4.35598737917564,
+      "grad_norm": 2.4183194637298584,
+      "learning_rate": 2.0276794392735444e-06,
+      "loss": 0.4098,
+      "step": 43490
+    },
+    {
+      "epoch": 4.356989031902639,
+      "grad_norm": 2.4014265537261963,
+      "learning_rate": 2.0214647120009173e-06,
+      "loss": 0.4118,
+      "step": 43500
+    },
+    {
+      "epoch": 4.357990684629639,
+      "grad_norm": 2.41623854637146,
+      "learning_rate": 2.015259121959759e-06,
+      "loss": 0.4008,
+      "step": 43510
+    },
+    {
+      "epoch": 4.358992337356638,
+      "grad_norm": 2.6848697662353516,
+      "learning_rate": 2.00906267161767e-06,
+      "loss": 0.4053,
+      "step": 43520
+    },
+    {
+      "epoch": 4.359993990083638,
+      "grad_norm": 2.032064437866211,
+      "learning_rate": 2.002875363438622e-06,
+      "loss": 0.3978,
+      "step": 43530
+    },
+    {
+      "epoch": 4.360995642810638,
+      "grad_norm": 1.95393967628479,
+      "learning_rate": 1.9966971998829463e-06,
+      "loss": 0.3873,
+      "step": 43540
+    },
+    {
+      "epoch": 4.361997295537637,
+      "grad_norm": 2.1058743000030518,
+      "learning_rate": 1.990528183407339e-06,
+      "loss": 0.4167,
+      "step": 43550
+    },
+    {
+      "epoch": 4.362998948264637,
+      "grad_norm": 2.3774189949035645,
+      "learning_rate": 1.984368316464874e-06,
+      "loss": 0.3911,
+      "step": 43560
+    },
+    {
+      "epoch": 4.364000600991636,
+      "grad_norm": 1.8962682485580444,
+      "learning_rate": 1.978217601504956e-06,
+      "loss": 0.3933,
+      "step": 43570
+    },
+    {
+      "epoch": 4.365002253718636,
+      "grad_norm": 2.469862937927246,
+      "learning_rate": 1.972076040973372e-06,
+      "loss": 0.4414,
+      "step": 43580
+    },
+    {
+      "epoch": 4.366003906445635,
+      "grad_norm": 2.292123794555664,
+      "learning_rate": 1.965943637312276e-06,
+      "loss": 0.3814,
+      "step": 43590
+    },
+    {
+      "epoch": 4.367005559172635,
+      "grad_norm": 2.4966185092926025,
+      "learning_rate": 1.9598203929601573e-06,
+      "loss": 0.4514,
+      "step": 43600
+    },
+    {
+      "epoch": 4.368007211899634,
+      "grad_norm": 2.058504104614258,
+      "learning_rate": 1.9537063103518772e-06,
+      "loss": 0.3639,
+      "step": 43610
+    },
+    {
+      "epoch": 4.369008864626634,
+      "grad_norm": 2.378732919692993,
+      "learning_rate": 1.947601391918649e-06,
+      "loss": 0.3669,
+      "step": 43620
+    },
+    {
+      "epoch": 4.370010517353633,
+      "grad_norm": 2.036379337310791,
+      "learning_rate": 1.9415056400880593e-06,
+      "loss": 0.3781,
+      "step": 43630
+    },
+    {
+      "epoch": 4.371012170080633,
+      "grad_norm": 2.2540555000305176,
+      "learning_rate": 1.9354190572840187e-06,
+      "loss": 0.4094,
+      "step": 43640
+    },
+    {
+      "epoch": 4.372013822807633,
+      "grad_norm": 1.785356879234314,
+      "learning_rate": 1.9293416459268134e-06,
+      "loss": 0.4068,
+      "step": 43650
+    },
+    {
+      "epoch": 4.373015475534632,
+      "grad_norm": 2.5779080390930176,
+      "learning_rate": 1.9232734084330824e-06,
+      "loss": 0.4351,
+      "step": 43660
+    },
+    {
+      "epoch": 4.374017128261632,
+      "grad_norm": 1.8296396732330322,
+      "learning_rate": 1.9172143472158122e-06,
+      "loss": 0.3672,
+      "step": 43670
+    },
+    {
+      "epoch": 4.375018780988631,
+      "grad_norm": 2.4608004093170166,
+      "learning_rate": 1.9111644646843284e-06,
+      "loss": 0.4167,
+      "step": 43680
+    },
+    {
+      "epoch": 4.376020433715631,
+      "grad_norm": 1.9104523658752441,
+      "learning_rate": 1.905123763244329e-06,
+      "loss": 0.375,
+      "step": 43690
+    },
+    {
+      "epoch": 4.37702208644263,
+      "grad_norm": 1.8407065868377686,
+      "learning_rate": 1.8990922452978565e-06,
+      "loss": 0.4115,
+      "step": 43700
+    },
+    {
+      "epoch": 4.37802373916963,
+      "grad_norm": 1.947320818901062,
+      "learning_rate": 1.8930699132432784e-06,
+      "loss": 0.4018,
+      "step": 43710
+    },
+    {
+      "epoch": 4.379025391896629,
+      "grad_norm": 2.074536085128784,
+      "learning_rate": 1.88705676947534e-06,
+      "loss": 0.4216,
+      "step": 43720
+    },
+    {
+      "epoch": 4.380027044623629,
+      "grad_norm": 2.215348958969116,
+      "learning_rate": 1.8810528163851177e-06,
+      "loss": 0.436,
+      "step": 43730
+    },
+    {
+      "epoch": 4.3810286973506285,
+      "grad_norm": 2.0011367797851562,
+      "learning_rate": 1.8750580563600351e-06,
+      "loss": 0.3784,
+      "step": 43740
+    },
+    {
+      "epoch": 4.382030350077628,
+      "grad_norm": 2.45017409324646,
+      "learning_rate": 1.8690724917838603e-06,
+      "loss": 0.3976,
+      "step": 43750
+    },
+    {
+      "epoch": 4.383032002804628,
+      "grad_norm": 2.329396963119507,
+      "learning_rate": 1.8630961250367062e-06,
+      "loss": 0.3553,
+      "step": 43760
+    },
+    {
+      "epoch": 4.384033655531627,
+      "grad_norm": 2.6597559452056885,
+      "learning_rate": 1.8571289584950302e-06,
+      "loss": 0.4332,
+      "step": 43770
+    },
+    {
+      "epoch": 4.385035308258627,
+      "grad_norm": 2.527765989303589,
+      "learning_rate": 1.8511709945316146e-06,
+      "loss": 0.4534,
+      "step": 43780
+    },
+    {
+      "epoch": 4.386036960985626,
+      "grad_norm": 2.524082899093628,
+      "learning_rate": 1.8452222355156146e-06,
+      "loss": 0.3753,
+      "step": 43790
+    },
+    {
+      "epoch": 4.387038613712626,
+      "grad_norm": 2.056520938873291,
+      "learning_rate": 1.8392826838124961e-06,
+      "loss": 0.4125,
+      "step": 43800
+    },
+    {
+      "epoch": 4.388040266439625,
+      "grad_norm": 1.816526174545288,
+      "learning_rate": 1.8333523417840782e-06,
+      "loss": 0.4369,
+      "step": 43810
+    },
+    {
+      "epoch": 4.389041919166625,
+      "grad_norm": 2.039801836013794,
+      "learning_rate": 1.8274312117885135e-06,
+      "loss": 0.3958,
+      "step": 43820
+    },
+    {
+      "epoch": 4.390043571893624,
+      "grad_norm": 2.485262155532837,
+      "learning_rate": 1.8215192961802907e-06,
+      "loss": 0.3797,
+      "step": 43830
+    },
+    {
+      "epoch": 4.391045224620624,
+      "grad_norm": 3.454650402069092,
+      "learning_rate": 1.8156165973102379e-06,
+      "loss": 0.3839,
+      "step": 43840
+    },
+    {
+      "epoch": 4.392046877347624,
+      "grad_norm": 2.490710735321045,
+      "learning_rate": 1.8097231175255163e-06,
+      "loss": 0.4095,
+      "step": 43850
+    },
+    {
+      "epoch": 4.393048530074623,
+      "grad_norm": 1.9275168180465698,
+      "learning_rate": 1.803838859169621e-06,
+      "loss": 0.3915,
+      "step": 43860
+    },
+    {
+      "epoch": 4.394050182801623,
+      "grad_norm": 1.8556205034255981,
+      "learning_rate": 1.7979638245823771e-06,
+      "loss": 0.457,
+      "step": 43870
+    },
+    {
+      "epoch": 4.395051835528622,
+      "grad_norm": 2.13421893119812,
+      "learning_rate": 1.7920980160999496e-06,
+      "loss": 0.4821,
+      "step": 43880
+    },
+    {
+      "epoch": 4.396053488255622,
+      "grad_norm": 1.8426682949066162,
+      "learning_rate": 1.786241436054828e-06,
+      "loss": 0.3621,
+      "step": 43890
+    },
+    {
+      "epoch": 4.397055140982621,
+      "grad_norm": 2.002091884613037,
+      "learning_rate": 1.7803940867758384e-06,
+      "loss": 0.4091,
+      "step": 43900
+    },
+    {
+      "epoch": 4.398056793709621,
+      "grad_norm": 2.000837564468384,
+      "learning_rate": 1.7745559705881236e-06,
+      "loss": 0.3843,
+      "step": 43910
+    },
+    {
+      "epoch": 4.39905844643662,
+      "grad_norm": 2.5199735164642334,
+      "learning_rate": 1.7687270898131796e-06,
+      "loss": 0.4583,
+      "step": 43920
+    },
+    {
+      "epoch": 4.40006009916362,
+      "grad_norm": 2.1583092212677,
+      "learning_rate": 1.7629074467687995e-06,
+      "loss": 0.3926,
+      "step": 43930
+    },
+    {
+      "epoch": 4.401061751890619,
+      "grad_norm": 1.873566746711731,
+      "learning_rate": 1.7570970437691238e-06,
+      "loss": 0.4649,
+      "step": 43940
+    },
+    {
+      "epoch": 4.4020634046176195,
+      "grad_norm": 1.9938627481460571,
+      "learning_rate": 1.7512958831246096e-06,
+      "loss": 0.4128,
+      "step": 43950
+    },
+    {
+      "epoch": 4.403065057344619,
+      "grad_norm": 2.516998529434204,
+      "learning_rate": 1.7455039671420537e-06,
+      "loss": 0.3994,
+      "step": 43960
+    },
+    {
+      "epoch": 4.404066710071618,
+      "grad_norm": 1.9853394031524658,
+      "learning_rate": 1.7397212981245526e-06,
+      "loss": 0.4124,
+      "step": 43970
+    },
+    {
+      "epoch": 4.405068362798618,
+      "grad_norm": 2.252943515777588,
+      "learning_rate": 1.733947878371539e-06,
+      "loss": 0.3939,
+      "step": 43980
+    },
+    {
+      "epoch": 4.406070015525617,
+      "grad_norm": 1.9540263414382935,
+      "learning_rate": 1.7281837101787797e-06,
+      "loss": 0.3688,
+      "step": 43990
+    },
+    {
+      "epoch": 4.407071668252617,
+      "grad_norm": 1.7471299171447754,
+      "learning_rate": 1.722428795838335e-06,
+      "loss": 0.3401,
+      "step": 44000
+    },
+    {
+      "epoch": 4.408073320979616,
+      "grad_norm": 1.9910095930099487,
+      "learning_rate": 1.7166831376386084e-06,
+      "loss": 0.4205,
+      "step": 44010
+    },
+    {
+      "epoch": 4.409074973706616,
+      "grad_norm": 2.40335750579834,
+      "learning_rate": 1.7109467378643135e-06,
+      "loss": 0.4245,
+      "step": 44020
+    },
+    {
+      "epoch": 4.410076626433615,
+      "grad_norm": 2.117269515991211,
+      "learning_rate": 1.7052195987964898e-06,
+      "loss": 0.3946,
+      "step": 44030
+    },
+    {
+      "epoch": 4.411078279160615,
+      "grad_norm": 1.9731988906860352,
+      "learning_rate": 1.6995017227124766e-06,
+      "loss": 0.3306,
+      "step": 44040
+    },
+    {
+      "epoch": 4.4120799318876145,
+      "grad_norm": 2.510343551635742,
+      "learning_rate": 1.6937931118859523e-06,
+      "loss": 0.4009,
+      "step": 44050
+    },
+    {
+      "epoch": 4.4130815846146145,
+      "grad_norm": 2.3110790252685547,
+      "learning_rate": 1.6880937685869013e-06,
+      "loss": 0.4243,
+      "step": 44060
+    },
+    {
+      "epoch": 4.414083237341614,
+      "grad_norm": 2.2889299392700195,
+      "learning_rate": 1.682403695081608e-06,
+      "loss": 0.3852,
+      "step": 44070
+    },
+    {
+      "epoch": 4.415084890068613,
+      "grad_norm": 2.237353563308716,
+      "learning_rate": 1.6767228936326984e-06,
+      "loss": 0.3906,
+      "step": 44080
+    },
+    {
+      "epoch": 4.416086542795613,
+      "grad_norm": 2.174852132797241,
+      "learning_rate": 1.6710513664990961e-06,
+      "loss": 0.3828,
+      "step": 44090
+    },
+    {
+      "epoch": 4.417088195522612,
+      "grad_norm": 2.5690200328826904,
+      "learning_rate": 1.6653891159360391e-06,
+      "loss": 0.4048,
+      "step": 44100
+    },
+    {
+      "epoch": 4.418089848249612,
+      "grad_norm": 2.094517230987549,
+      "learning_rate": 1.659736144195065e-06,
+      "loss": 0.4225,
+      "step": 44110
+    },
+    {
+      "epoch": 4.419091500976611,
+      "grad_norm": 2.416579246520996,
+      "learning_rate": 1.654092453524045e-06,
+      "loss": 0.4436,
+      "step": 44120
+    },
+    {
+      "epoch": 4.420093153703611,
+      "grad_norm": 2.3031387329101562,
+      "learning_rate": 1.648458046167145e-06,
+      "loss": 0.4566,
+      "step": 44130
+    },
+    {
+      "epoch": 4.42109480643061,
+      "grad_norm": 2.294541358947754,
+      "learning_rate": 1.64283292436484e-06,
+      "loss": 0.4522,
+      "step": 44140
+    },
+    {
+      "epoch": 4.42209645915761,
+      "grad_norm": 1.5399305820465088,
+      "learning_rate": 1.637217090353918e-06,
+      "loss": 0.409,
+      "step": 44150
+    },
+    {
+      "epoch": 4.4230981118846096,
+      "grad_norm": 2.1396875381469727,
+      "learning_rate": 1.631610546367468e-06,
+      "loss": 0.3919,
+      "step": 44160
+    },
+    {
+      "epoch": 4.42409976461161,
+      "grad_norm": 2.2197341918945312,
+      "learning_rate": 1.6260132946348899e-06,
+      "loss": 0.4115,
+      "step": 44170
+    },
+    {
+      "epoch": 4.425101417338609,
+      "grad_norm": 2.275446891784668,
+      "learning_rate": 1.620425337381884e-06,
+      "loss": 0.4239,
+      "step": 44180
+    },
+    {
+      "epoch": 4.426103070065608,
+      "grad_norm": 2.164278745651245,
+      "learning_rate": 1.6148466768304587e-06,
+      "loss": 0.4411,
+      "step": 44190
+    },
+    {
+      "epoch": 4.427104722792608,
+      "grad_norm": 2.4946508407592773,
+      "learning_rate": 1.6092773151989204e-06,
+      "loss": 0.4211,
+      "step": 44200
+    },
+    {
+      "epoch": 4.428106375519607,
+      "grad_norm": 2.1352880001068115,
+      "learning_rate": 1.6037172547018863e-06,
+      "loss": 0.3815,
+      "step": 44210
+    },
+    {
+      "epoch": 4.429108028246607,
+      "grad_norm": 2.4429593086242676,
+      "learning_rate": 1.5981664975502685e-06,
+      "loss": 0.474,
+      "step": 44220
+    },
+    {
+      "epoch": 4.430109680973606,
+      "grad_norm": 1.741686463356018,
+      "learning_rate": 1.5926250459512793e-06,
+      "loss": 0.3641,
+      "step": 44230
+    },
+    {
+      "epoch": 4.431111333700606,
+      "grad_norm": 1.501874327659607,
+      "learning_rate": 1.5870929021084307e-06,
+      "loss": 0.3908,
+      "step": 44240
+    },
+    {
+      "epoch": 4.432112986427605,
+      "grad_norm": 2.2431552410125732,
+      "learning_rate": 1.5815700682215439e-06,
+      "loss": 0.4295,
+      "step": 44250
+    },
+    {
+      "epoch": 4.4331146391546055,
+      "grad_norm": 2.8315837383270264,
+      "learning_rate": 1.576056546486726e-06,
+      "loss": 0.4185,
+      "step": 44260
+    },
+    {
+      "epoch": 4.434116291881605,
+      "grad_norm": 2.2130463123321533,
+      "learning_rate": 1.570552339096376e-06,
+      "loss": 0.4278,
+      "step": 44270
+    },
+    {
+      "epoch": 4.435117944608605,
+      "grad_norm": 2.1421968936920166,
+      "learning_rate": 1.5650574482392183e-06,
+      "loss": 0.4138,
+      "step": 44280
+    },
+    {
+      "epoch": 4.436119597335604,
+      "grad_norm": 2.1757049560546875,
+      "learning_rate": 1.5595718761002325e-06,
+      "loss": 0.4759,
+      "step": 44290
+    },
+    {
+      "epoch": 4.437121250062603,
+      "grad_norm": 2.8125619888305664,
+      "learning_rate": 1.5540956248607213e-06,
+      "loss": 0.4304,
+      "step": 44300
+    },
+    {
+      "epoch": 4.438122902789603,
+      "grad_norm": 2.888702154159546,
+      "learning_rate": 1.5486286966982677e-06,
+      "loss": 0.4594,
+      "step": 44310
+    },
+    {
+      "epoch": 4.439124555516602,
+      "grad_norm": 2.215304136276245,
+      "learning_rate": 1.5431710937867633e-06,
+      "loss": 0.3982,
+      "step": 44320
+    },
+    {
+      "epoch": 4.440126208243602,
+      "grad_norm": 2.1499667167663574,
+      "learning_rate": 1.537722818296372e-06,
+      "loss": 0.421,
+      "step": 44330
+    },
+    {
+      "epoch": 4.441127860970601,
+      "grad_norm": 2.082578420639038,
+      "learning_rate": 1.53228387239355e-06,
+      "loss": 0.4226,
+      "step": 44340
+    },
+    {
+      "epoch": 4.442129513697601,
+      "grad_norm": 2.056445598602295,
+      "learning_rate": 1.52685425824107e-06,
+      "loss": 0.4324,
+      "step": 44350
+    },
+    {
+      "epoch": 4.4431311664246005,
+      "grad_norm": 2.5467376708984375,
+      "learning_rate": 1.5214339779979576e-06,
+      "loss": 0.4099,
+      "step": 44360
+    },
+    {
+      "epoch": 4.4441328191516005,
+      "grad_norm": 2.7463908195495605,
+      "learning_rate": 1.5160230338195497e-06,
+      "loss": 0.4159,
+      "step": 44370
+    },
+    {
+      "epoch": 4.4451344718786,
+      "grad_norm": 2.3661184310913086,
+      "learning_rate": 1.510621427857467e-06,
+      "loss": 0.4228,
+      "step": 44380
+    },
+    {
+      "epoch": 4.4461361246056,
+      "grad_norm": 2.045476198196411,
+      "learning_rate": 1.5052291622596166e-06,
+      "loss": 0.3782,
+      "step": 44390
+    },
+    {
+      "epoch": 4.447137777332599,
+      "grad_norm": 2.825392007827759,
+      "learning_rate": 1.49984623917018e-06,
+      "loss": 0.4446,
+      "step": 44400
+    },
+    {
+      "epoch": 4.448139430059598,
+      "grad_norm": 1.9934977293014526,
+      "learning_rate": 1.4944726607296456e-06,
+      "loss": 0.3826,
+      "step": 44410
+    },
+    {
+      "epoch": 4.449141082786598,
+      "grad_norm": 2.085172414779663,
+      "learning_rate": 1.4891084290747704e-06,
+      "loss": 0.4111,
+      "step": 44420
+    },
+    {
+      "epoch": 4.450142735513597,
+      "grad_norm": 2.056011199951172,
+      "learning_rate": 1.4837535463385982e-06,
+      "loss": 0.4359,
+      "step": 44430
+    },
+    {
+      "epoch": 4.451144388240597,
+      "grad_norm": 2.0101840496063232,
+      "learning_rate": 1.478408014650448e-06,
+      "loss": 0.3443,
+      "step": 44440
+    },
+    {
+      "epoch": 4.452146040967596,
+      "grad_norm": 2.3599886894226074,
+      "learning_rate": 1.473071836135939e-06,
+      "loss": 0.4214,
+      "step": 44450
+    },
+    {
+      "epoch": 4.453147693694596,
+      "grad_norm": 2.235658884048462,
+      "learning_rate": 1.4677450129169574e-06,
+      "loss": 0.4071,
+      "step": 44460
+    },
+    {
+      "epoch": 4.4541493464215955,
+      "grad_norm": 2.4682376384735107,
+      "learning_rate": 1.4624275471116638e-06,
+      "loss": 0.3985,
+      "step": 44470
+    },
+    {
+      "epoch": 4.455150999148596,
+      "grad_norm": 2.524515390396118,
+      "learning_rate": 1.4571194408345146e-06,
+      "loss": 0.3838,
+      "step": 44480
+    },
+    {
+      "epoch": 4.456152651875595,
+      "grad_norm": 2.4610915184020996,
+      "learning_rate": 1.451820696196235e-06,
+      "loss": 0.3972,
+      "step": 44490
+    },
+    {
+      "epoch": 4.457154304602594,
+      "grad_norm": 1.954127311706543,
+      "learning_rate": 1.4465313153038284e-06,
+      "loss": 0.4061,
+      "step": 44500
+    },
+    {
+      "epoch": 4.458155957329594,
+      "grad_norm": 1.9836078882217407,
+      "learning_rate": 1.441251300260571e-06,
+      "loss": 0.4063,
+      "step": 44510
+    },
+    {
+      "epoch": 4.459157610056593,
+      "grad_norm": 2.0435895919799805,
+      "learning_rate": 1.435980653166022e-06,
+      "loss": 0.4643,
+      "step": 44520
+    },
+    {
+      "epoch": 4.460159262783593,
+      "grad_norm": 2.140763998031616,
+      "learning_rate": 1.4307193761160131e-06,
+      "loss": 0.3666,
+      "step": 44530
+    },
+    {
+      "epoch": 4.461160915510592,
+      "grad_norm": 2.599942445755005,
+      "learning_rate": 1.4254674712026488e-06,
+      "loss": 0.4522,
+      "step": 44540
+    },
+    {
+      "epoch": 4.462162568237592,
+      "grad_norm": 1.9349007606506348,
+      "learning_rate": 1.4202249405143032e-06,
+      "loss": 0.3737,
+      "step": 44550
+    },
+    {
+      "epoch": 4.463164220964591,
+      "grad_norm": 2.34694766998291,
+      "learning_rate": 1.4149917861356331e-06,
+      "loss": 0.4058,
+      "step": 44560
+    },
+    {
+      "epoch": 4.4641658736915915,
+      "grad_norm": 2.036736249923706,
+      "learning_rate": 1.4097680101475553e-06,
+      "loss": 0.3698,
+      "step": 44570
+    },
+    {
+      "epoch": 4.465167526418591,
+      "grad_norm": 2.2654614448547363,
+      "learning_rate": 1.404553614627266e-06,
+      "loss": 0.4287,
+      "step": 44580
+    },
+    {
+      "epoch": 4.466169179145591,
+      "grad_norm": 1.855678915977478,
+      "learning_rate": 1.399348601648226e-06,
+      "loss": 0.3928,
+      "step": 44590
+    },
+    {
+      "epoch": 4.46717083187259,
+      "grad_norm": 2.315638303756714,
+      "learning_rate": 1.3941529732801662e-06,
+      "loss": 0.424,
+      "step": 44600
+    },
+    {
+      "epoch": 4.468172484599589,
+      "grad_norm": 2.044198989868164,
+      "learning_rate": 1.3889667315890948e-06,
+      "loss": 0.4122,
+      "step": 44610
+    },
+    {
+      "epoch": 4.469174137326589,
+      "grad_norm": 2.004559278488159,
+      "learning_rate": 1.3837898786372704e-06,
+      "loss": 0.3801,
+      "step": 44620
+    },
+    {
+      "epoch": 4.470175790053588,
+      "grad_norm": 1.7977226972579956,
+      "learning_rate": 1.3786224164832302e-06,
+      "loss": 0.3668,
+      "step": 44630
+    },
+    {
+      "epoch": 4.471177442780588,
+      "grad_norm": 1.8870209455490112,
+      "learning_rate": 1.3734643471817743e-06,
+      "loss": 0.4103,
+      "step": 44640
+    },
+    {
+      "epoch": 4.472179095507587,
+      "grad_norm": 1.5372698307037354,
+      "learning_rate": 1.368315672783968e-06,
+      "loss": 0.4171,
+      "step": 44650
+    },
+    {
+      "epoch": 4.473180748234587,
+      "grad_norm": 2.3017232418060303,
+      "learning_rate": 1.3631763953371402e-06,
+      "loss": 0.3999,
+      "step": 44660
+    },
+    {
+      "epoch": 4.4741824009615865,
+      "grad_norm": 2.015127182006836,
+      "learning_rate": 1.358046516884881e-06,
+      "loss": 0.4281,
+      "step": 44670
+    },
+    {
+      "epoch": 4.4751840536885865,
+      "grad_norm": 2.389350175857544,
+      "learning_rate": 1.3529260394670562e-06,
+      "loss": 0.4312,
+      "step": 44680
+    },
+    {
+      "epoch": 4.476185706415586,
+      "grad_norm": 2.0506579875946045,
+      "learning_rate": 1.347814965119773e-06,
+      "loss": 0.4292,
+      "step": 44690
+    },
+    {
+      "epoch": 4.477187359142585,
+      "grad_norm": 2.1258811950683594,
+      "learning_rate": 1.3427132958754057e-06,
+      "loss": 0.3413,
+      "step": 44700
+    },
+    {
+      "epoch": 4.478189011869585,
+      "grad_norm": 2.5169951915740967,
+      "learning_rate": 1.3376210337626037e-06,
+      "loss": 0.5072,
+      "step": 44710
+    },
+    {
+      "epoch": 4.479190664596584,
+      "grad_norm": 2.3843328952789307,
+      "learning_rate": 1.3325381808062641e-06,
+      "loss": 0.4088,
+      "step": 44720
+    },
+    {
+      "epoch": 4.480192317323584,
+      "grad_norm": 2.526643753051758,
+      "learning_rate": 1.3274647390275314e-06,
+      "loss": 0.3923,
+      "step": 44730
+    },
+    {
+      "epoch": 4.481193970050583,
+      "grad_norm": 2.592135429382324,
+      "learning_rate": 1.3224007104438273e-06,
+      "loss": 0.4091,
+      "step": 44740
+    },
+    {
+      "epoch": 4.482195622777583,
+      "grad_norm": 2.1877357959747314,
+      "learning_rate": 1.3173460970688251e-06,
+      "loss": 0.4152,
+      "step": 44750
+    },
+    {
+      "epoch": 4.483197275504582,
+      "grad_norm": 2.017637014389038,
+      "learning_rate": 1.3123009009124442e-06,
+      "loss": 0.471,
+      "step": 44760
+    },
+    {
+      "epoch": 4.484198928231582,
+      "grad_norm": 2.706550359725952,
+      "learning_rate": 1.3072651239808692e-06,
+      "loss": 0.4349,
+      "step": 44770
+    },
+    {
+      "epoch": 4.4852005809585815,
+      "grad_norm": 2.889103889465332,
+      "learning_rate": 1.3022387682765398e-06,
+      "loss": 0.3704,
+      "step": 44780
+    },
+    {
+      "epoch": 4.486202233685582,
+      "grad_norm": 2.3987677097320557,
+      "learning_rate": 1.2972218357981458e-06,
+      "loss": 0.3974,
+      "step": 44790
+    },
+    {
+      "epoch": 4.487203886412581,
+      "grad_norm": 2.7593445777893066,
+      "learning_rate": 1.2922143285406224e-06,
+      "loss": 0.4262,
+      "step": 44800
+    },
+    {
+      "epoch": 4.48820553913958,
+      "grad_norm": 2.0148863792419434,
+      "learning_rate": 1.2872162484951738e-06,
+      "loss": 0.4055,
+      "step": 44810
+    },
+    {
+      "epoch": 4.48920719186658,
+      "grad_norm": 2.1033871173858643,
+      "learning_rate": 1.2822275976492493e-06,
+      "loss": 0.4186,
+      "step": 44820
+    },
+    {
+      "epoch": 4.490208844593579,
+      "grad_norm": 2.587449312210083,
+      "learning_rate": 1.277248377986534e-06,
+      "loss": 0.3816,
+      "step": 44830
+    },
+    {
+      "epoch": 4.491210497320579,
+      "grad_norm": 2.1474623680114746,
+      "learning_rate": 1.2722785914869862e-06,
+      "loss": 0.3941,
+      "step": 44840
+    },
+    {
+      "epoch": 4.492212150047578,
+      "grad_norm": 2.459506034851074,
+      "learning_rate": 1.267318240126794e-06,
+      "loss": 0.361,
+      "step": 44850
+    },
+    {
+      "epoch": 4.493213802774578,
+      "grad_norm": 2.14664363861084,
+      "learning_rate": 1.262367325878408e-06,
+      "loss": 0.4469,
+      "step": 44860
+    },
+    {
+      "epoch": 4.494215455501577,
+      "grad_norm": 2.295531749725342,
+      "learning_rate": 1.2574258507105168e-06,
+      "loss": 0.5106,
+      "step": 44870
+    },
+    {
+      "epoch": 4.4952171082285775,
+      "grad_norm": 2.360304594039917,
+      "learning_rate": 1.2524938165880601e-06,
+      "loss": 0.4323,
+      "step": 44880
+    },
+    {
+      "epoch": 4.496218760955577,
+      "grad_norm": 2.631624937057495,
+      "learning_rate": 1.2475712254722188e-06,
+      "loss": 0.4318,
+      "step": 44890
+    },
+    {
+      "epoch": 4.497220413682577,
+      "grad_norm": 1.8856874704360962,
+      "learning_rate": 1.2426580793204246e-06,
+      "loss": 0.4143,
+      "step": 44900
+    },
+    {
+      "epoch": 4.498222066409576,
+      "grad_norm": 2.2703697681427,
+      "learning_rate": 1.2377543800863505e-06,
+      "loss": 0.4279,
+      "step": 44910
+    },
+    {
+      "epoch": 4.499223719136575,
+      "grad_norm": 1.7943007946014404,
+      "learning_rate": 1.2328601297199121e-06,
+      "loss": 0.3452,
+      "step": 44920
+    },
+    {
+      "epoch": 4.500225371863575,
+      "grad_norm": 1.8310256004333496,
+      "learning_rate": 1.2279753301672691e-06,
+      "loss": 0.4461,
+      "step": 44930
+    },
+    {
+      "epoch": 4.501227024590574,
+      "grad_norm": 2.538496494293213,
+      "learning_rate": 1.2230999833708262e-06,
+      "loss": 0.3722,
+      "step": 44940
+    },
+    {
+      "epoch": 4.502228677317574,
+      "grad_norm": 1.8670365810394287,
+      "learning_rate": 1.218234091269224e-06,
+      "loss": 0.3366,
+      "step": 44950
+    },
+    {
+      "epoch": 4.503230330044573,
+      "grad_norm": 2.470527410507202,
+      "learning_rate": 1.2133776557973458e-06,
+      "loss": 0.3661,
+      "step": 44960
+    },
+    {
+      "epoch": 4.504231982771573,
+      "grad_norm": 2.3969390392303467,
+      "learning_rate": 1.208530678886316e-06,
+      "loss": 0.3995,
+      "step": 44970
+    },
+    {
+      "epoch": 4.5052336354985725,
+      "grad_norm": 1.7786046266555786,
+      "learning_rate": 1.2036931624634928e-06,
+      "loss": 0.3834,
+      "step": 44980
+    },
+    {
+      "epoch": 4.5062352882255725,
+      "grad_norm": 1.8966312408447266,
+      "learning_rate": 1.1988651084524794e-06,
+      "loss": 0.4019,
+      "step": 44990
+    },
+    {
+      "epoch": 4.507236940952572,
+      "grad_norm": 2.385106325149536,
+      "learning_rate": 1.1940465187731093e-06,
+      "loss": 0.3839,
+      "step": 45000
+    },
+    {
+      "epoch": 4.508238593679572,
+      "grad_norm": 1.5654442310333252,
+      "learning_rate": 1.1892373953414638e-06,
+      "loss": 0.3852,
+      "step": 45010
+    },
+    {
+      "epoch": 4.509240246406571,
+      "grad_norm": 2.0553817749023438,
+      "learning_rate": 1.1844377400698437e-06,
+      "loss": 0.4238,
+      "step": 45020
+    },
+    {
+      "epoch": 4.51024189913357,
+      "grad_norm": 1.7214395999908447,
+      "learning_rate": 1.1796475548667945e-06,
+      "loss": 0.3799,
+      "step": 45030
+    },
+    {
+      "epoch": 4.51124355186057,
+      "grad_norm": 2.3973593711853027,
+      "learning_rate": 1.1748668416371067e-06,
+      "loss": 0.4252,
+      "step": 45040
+    },
+    {
+      "epoch": 4.512245204587569,
+      "grad_norm": 1.9291691780090332,
+      "learning_rate": 1.1700956022817788e-06,
+      "loss": 0.4175,
+      "step": 45050
+    },
+    {
+      "epoch": 4.513246857314569,
+      "grad_norm": 2.493912696838379,
+      "learning_rate": 1.16533383869806e-06,
+      "loss": 0.3822,
+      "step": 45060
+    },
+    {
+      "epoch": 4.514248510041568,
+      "grad_norm": 1.8617504835128784,
+      "learning_rate": 1.1605815527794329e-06,
+      "loss": 0.3491,
+      "step": 45070
+    },
+    {
+      "epoch": 4.515250162768568,
+      "grad_norm": 2.023247003555298,
+      "learning_rate": 1.1558387464156024e-06,
+      "loss": 0.4175,
+      "step": 45080
+    },
+    {
+      "epoch": 4.5162518154955675,
+      "grad_norm": 2.2523674964904785,
+      "learning_rate": 1.151105421492507e-06,
+      "loss": 0.4757,
+      "step": 45090
+    },
+    {
+      "epoch": 4.517253468222568,
+      "grad_norm": 1.8958468437194824,
+      "learning_rate": 1.1463815798923138e-06,
+      "loss": 0.3856,
+      "step": 45100
+    },
+    {
+      "epoch": 4.518255120949567,
+      "grad_norm": 1.961393117904663,
+      "learning_rate": 1.1416672234934283e-06,
+      "loss": 0.451,
+      "step": 45110
+    },
+    {
+      "epoch": 4.519256773676567,
+      "grad_norm": 1.5756324529647827,
+      "learning_rate": 1.1369623541704706e-06,
+      "loss": 0.4175,
+      "step": 45120
+    },
+    {
+      "epoch": 4.520258426403566,
+      "grad_norm": 1.8311176300048828,
+      "learning_rate": 1.1322669737942908e-06,
+      "loss": 0.4113,
+      "step": 45130
+    },
+    {
+      "epoch": 4.521260079130565,
+      "grad_norm": 2.2952687740325928,
+      "learning_rate": 1.1275810842319767e-06,
+      "loss": 0.3718,
+      "step": 45140
+    },
+    {
+      "epoch": 4.522261731857565,
+      "grad_norm": 1.6738089323043823,
+      "learning_rate": 1.1229046873468374e-06,
+      "loss": 0.372,
+      "step": 45150
+    },
+    {
+      "epoch": 4.523263384584564,
+      "grad_norm": 2.8291242122650146,
+      "learning_rate": 1.118237784998394e-06,
+      "loss": 0.4321,
+      "step": 45160
+    },
+    {
+      "epoch": 4.524265037311564,
+      "grad_norm": 2.192211866378784,
+      "learning_rate": 1.1135803790424115e-06,
+      "loss": 0.3983,
+      "step": 45170
+    },
+    {
+      "epoch": 4.525266690038563,
+      "grad_norm": 2.6538708209991455,
+      "learning_rate": 1.1089324713308674e-06,
+      "loss": 0.3733,
+      "step": 45180
+    },
+    {
+      "epoch": 4.5262683427655634,
+      "grad_norm": 2.111126661300659,
+      "learning_rate": 1.1042940637119665e-06,
+      "loss": 0.3748,
+      "step": 45190
+    },
+    {
+      "epoch": 4.527269995492563,
+      "grad_norm": 2.8170530796051025,
+      "learning_rate": 1.099665158030133e-06,
+      "loss": 0.4561,
+      "step": 45200
+    },
+    {
+      "epoch": 4.528271648219563,
+      "grad_norm": 1.8179787397384644,
+      "learning_rate": 1.0950457561260174e-06,
+      "loss": 0.3934,
+      "step": 45210
+    },
+    {
+      "epoch": 4.529273300946562,
+      "grad_norm": 3.770756721496582,
+      "learning_rate": 1.0904358598364833e-06,
+      "loss": 0.3988,
+      "step": 45220
+    },
+    {
+      "epoch": 4.530274953673562,
+      "grad_norm": 2.2152037620544434,
+      "learning_rate": 1.085835470994623e-06,
+      "loss": 0.386,
+      "step": 45230
+    },
+    {
+      "epoch": 4.531276606400561,
+      "grad_norm": 2.2403817176818848,
+      "learning_rate": 1.0812445914297447e-06,
+      "loss": 0.4372,
+      "step": 45240
+    },
+    {
+      "epoch": 4.53227825912756,
+      "grad_norm": 2.0329346656799316,
+      "learning_rate": 1.0766632229673724e-06,
+      "loss": 0.3938,
+      "step": 45250
+    },
+    {
+      "epoch": 4.53327991185456,
+      "grad_norm": 2.1116416454315186,
+      "learning_rate": 1.0720913674292509e-06,
+      "loss": 0.4002,
+      "step": 45260
+    },
+    {
+      "epoch": 4.534281564581559,
+      "grad_norm": 1.6225544214248657,
+      "learning_rate": 1.0675290266333433e-06,
+      "loss": 0.4409,
+      "step": 45270
+    },
+    {
+      "epoch": 4.535283217308559,
+      "grad_norm": 2.520085334777832,
+      "learning_rate": 1.0629762023938283e-06,
+      "loss": 0.3987,
+      "step": 45280
+    },
+    {
+      "epoch": 4.5362848700355585,
+      "grad_norm": 1.736752986907959,
+      "learning_rate": 1.0584328965211e-06,
+      "loss": 0.3976,
+      "step": 45290
+    },
+    {
+      "epoch": 4.5372865227625585,
+      "grad_norm": 2.065598726272583,
+      "learning_rate": 1.0538991108217682e-06,
+      "loss": 0.4387,
+      "step": 45300
+    },
+    {
+      "epoch": 4.538288175489558,
+      "grad_norm": 2.372802495956421,
+      "learning_rate": 1.0493748470986554e-06,
+      "loss": 0.4333,
+      "step": 45310
+    },
+    {
+      "epoch": 4.539289828216558,
+      "grad_norm": 2.1573617458343506,
+      "learning_rate": 1.0448601071507996e-06,
+      "loss": 0.3845,
+      "step": 45320
+    },
+    {
+      "epoch": 4.540291480943557,
+      "grad_norm": 2.4107489585876465,
+      "learning_rate": 1.040354892773454e-06,
+      "loss": 0.439,
+      "step": 45330
+    },
+    {
+      "epoch": 4.541293133670557,
+      "grad_norm": 2.2853431701660156,
+      "learning_rate": 1.0358592057580746e-06,
+      "loss": 0.393,
+      "step": 45340
+    },
+    {
+      "epoch": 4.542294786397556,
+      "grad_norm": 2.2682673931121826,
+      "learning_rate": 1.0313730478923422e-06,
+      "loss": 0.3983,
+      "step": 45350
+    },
+    {
+      "epoch": 4.543296439124555,
+      "grad_norm": 1.9406390190124512,
+      "learning_rate": 1.0268964209601328e-06,
+      "loss": 0.3568,
+      "step": 45360
+    },
+    {
+      "epoch": 4.544298091851555,
+      "grad_norm": 2.364281415939331,
+      "learning_rate": 1.0224293267415558e-06,
+      "loss": 0.4477,
+      "step": 45370
+    },
+    {
+      "epoch": 4.545299744578554,
+      "grad_norm": 2.5253348350524902,
+      "learning_rate": 1.0179717670129041e-06,
+      "loss": 0.425,
+      "step": 45380
+    },
+    {
+      "epoch": 4.546301397305554,
+      "grad_norm": 2.046555995941162,
+      "learning_rate": 1.0135237435466932e-06,
+      "loss": 0.4468,
+      "step": 45390
+    },
+    {
+      "epoch": 4.5473030500325535,
+      "grad_norm": 1.9803757667541504,
+      "learning_rate": 1.0090852581116473e-06,
+      "loss": 0.4173,
+      "step": 45400
+    },
+    {
+      "epoch": 4.548304702759554,
+      "grad_norm": 2.417950391769409,
+      "learning_rate": 1.004656312472693e-06,
+      "loss": 0.3773,
+      "step": 45410
+    },
+    {
+      "epoch": 4.549306355486553,
+      "grad_norm": 2.228677749633789,
+      "learning_rate": 1.0002369083909612e-06,
+      "loss": 0.4586,
+      "step": 45420
+    },
+    {
+      "epoch": 4.550308008213553,
+      "grad_norm": 2.234309196472168,
+      "learning_rate": 9.958270476237957e-07,
+      "loss": 0.4367,
+      "step": 45430
+    },
+    {
+      "epoch": 4.551309660940552,
+      "grad_norm": 1.536161184310913,
+      "learning_rate": 9.914267319247495e-07,
+      "loss": 0.4218,
+      "step": 45440
+    },
+    {
+      "epoch": 4.552311313667552,
+      "grad_norm": 2.13627552986145,
+      "learning_rate": 9.870359630435616e-07,
+      "loss": 0.397,
+      "step": 45450
+    },
+    {
+      "epoch": 4.553312966394551,
+      "grad_norm": 2.022158145904541,
+      "learning_rate": 9.826547427261913e-07,
+      "loss": 0.4395,
+      "step": 45460
+    },
+    {
+      "epoch": 4.55431461912155,
+      "grad_norm": 2.3157637119293213,
+      "learning_rate": 9.782830727147974e-07,
+      "loss": 0.4107,
+      "step": 45470
+    },
+    {
+      "epoch": 4.55531627184855,
+      "grad_norm": 2.565974235534668,
+      "learning_rate": 9.739209547477396e-07,
+      "loss": 0.3774,
+      "step": 45480
+    },
+    {
+      "epoch": 4.556317924575549,
+      "grad_norm": 2.3096580505371094,
+      "learning_rate": 9.695683905595748e-07,
+      "loss": 0.458,
+      "step": 45490
+    },
+    {
+      "epoch": 4.557319577302549,
+      "grad_norm": 2.6375136375427246,
+      "learning_rate": 9.652253818810686e-07,
+      "loss": 0.4138,
+      "step": 45500
+    },
+    {
+      "epoch": 4.558321230029549,
+      "grad_norm": 2.0477283000946045,
+      "learning_rate": 9.608919304391895e-07,
+      "loss": 0.4362,
+      "step": 45510
+    },
+    {
+      "epoch": 4.559322882756549,
+      "grad_norm": 2.114675521850586,
+      "learning_rate": 9.565680379570867e-07,
+      "loss": 0.4118,
+      "step": 45520
+    },
+    {
+      "epoch": 4.560324535483548,
+      "grad_norm": 1.7504016160964966,
+      "learning_rate": 9.522537061541353e-07,
+      "loss": 0.3752,
+      "step": 45530
+    },
+    {
+      "epoch": 4.561326188210547,
+      "grad_norm": 2.3009567260742188,
+      "learning_rate": 9.47948936745885e-07,
+      "loss": 0.3958,
+      "step": 45540
+    },
+    {
+      "epoch": 4.562327840937547,
+      "grad_norm": 2.306684732437134,
+      "learning_rate": 9.436537314440996e-07,
+      "loss": 0.4089,
+      "step": 45550
+    },
+    {
+      "epoch": 4.563329493664547,
+      "grad_norm": 3.0576114654541016,
+      "learning_rate": 9.393680919567299e-07,
+      "loss": 0.4074,
+      "step": 45560
+    },
+    {
+      "epoch": 4.564331146391546,
+      "grad_norm": 1.9671968221664429,
+      "learning_rate": 9.350920199879265e-07,
+      "loss": 0.3332,
+      "step": 45570
+    },
+    {
+      "epoch": 4.565332799118545,
+      "grad_norm": 1.918282389640808,
+      "learning_rate": 9.308255172380376e-07,
+      "loss": 0.4116,
+      "step": 45580
+    },
+    {
+      "epoch": 4.566334451845545,
+      "grad_norm": 1.908010721206665,
+      "learning_rate": 9.265685854035977e-07,
+      "loss": 0.3648,
+      "step": 45590
+    },
+    {
+      "epoch": 4.5673361045725445,
+      "grad_norm": 1.6069321632385254,
+      "learning_rate": 9.2232122617735e-07,
+      "loss": 0.4405,
+      "step": 45600
+    },
+    {
+      "epoch": 4.5683377572995445,
+      "grad_norm": 2.416090726852417,
+      "learning_rate": 9.180834412482187e-07,
+      "loss": 0.3838,
+      "step": 45610
+    },
+    {
+      "epoch": 4.569339410026544,
+      "grad_norm": 1.9915950298309326,
+      "learning_rate": 9.13855232301325e-07,
+      "loss": 0.4142,
+      "step": 45620
+    },
+    {
+      "epoch": 4.570341062753544,
+      "grad_norm": 2.693861722946167,
+      "learning_rate": 9.096366010179852e-07,
+      "loss": 0.4403,
+      "step": 45630
+    },
+    {
+      "epoch": 4.571342715480543,
+      "grad_norm": 1.9319853782653809,
+      "learning_rate": 9.054275490757019e-07,
+      "loss": 0.4234,
+      "step": 45640
+    },
+    {
+      "epoch": 4.572344368207542,
+      "grad_norm": 2.010544538497925,
+      "learning_rate": 9.012280781481725e-07,
+      "loss": 0.4433,
+      "step": 45650
+    },
+    {
+      "epoch": 4.573346020934542,
+      "grad_norm": 2.1950645446777344,
+      "learning_rate": 8.970381899052804e-07,
+      "loss": 0.4073,
+      "step": 45660
+    },
+    {
+      "epoch": 4.574347673661541,
+      "grad_norm": 1.9796150922775269,
+      "learning_rate": 8.928578860131043e-07,
+      "loss": 0.4217,
+      "step": 45670
+    },
+    {
+      "epoch": 4.575349326388541,
+      "grad_norm": 2.374180316925049,
+      "learning_rate": 8.886871681339087e-07,
+      "loss": 0.3626,
+      "step": 45680
+    },
+    {
+      "epoch": 4.57635097911554,
+      "grad_norm": 3.04028582572937,
+      "learning_rate": 8.845260379261449e-07,
+      "loss": 0.4371,
+      "step": 45690
+    },
+    {
+      "epoch": 4.57735263184254,
+      "grad_norm": 2.386300563812256,
+      "learning_rate": 8.803744970444533e-07,
+      "loss": 0.3989,
+      "step": 45700
+    },
+    {
+      "epoch": 4.5783542845695395,
+      "grad_norm": 2.734030246734619,
+      "learning_rate": 8.762325471396632e-07,
+      "loss": 0.4327,
+      "step": 45710
+    },
+    {
+      "epoch": 4.57935593729654,
+      "grad_norm": 2.3858745098114014,
+      "learning_rate": 8.721001898587822e-07,
+      "loss": 0.4408,
+      "step": 45720
+    },
+    {
+      "epoch": 4.580357590023539,
+      "grad_norm": 2.001619577407837,
+      "learning_rate": 8.67977426845018e-07,
+      "loss": 0.421,
+      "step": 45730
+    },
+    {
+      "epoch": 4.581359242750539,
+      "grad_norm": 2.0118606090545654,
+      "learning_rate": 8.638642597377483e-07,
+      "loss": 0.442,
+      "step": 45740
+    },
+    {
+      "epoch": 4.582360895477538,
+      "grad_norm": 2.007488489151001,
+      "learning_rate": 8.597606901725397e-07,
+      "loss": 0.4262,
+      "step": 45750
+    },
+    {
+      "epoch": 4.583362548204537,
+      "grad_norm": 2.105195999145508,
+      "learning_rate": 8.55666719781148e-07,
+      "loss": 0.3884,
+      "step": 45760
+    },
+    {
+      "epoch": 4.584364200931537,
+      "grad_norm": 1.8013733625411987,
+      "learning_rate": 8.515823501915126e-07,
+      "loss": 0.4489,
+      "step": 45770
+    },
+    {
+      "epoch": 4.585365853658536,
+      "grad_norm": 2.1725831031799316,
+      "learning_rate": 8.475075830277401e-07,
+      "loss": 0.4059,
+      "step": 45780
+    },
+    {
+      "epoch": 4.586367506385536,
+      "grad_norm": 1.905267357826233,
+      "learning_rate": 8.434424199101315e-07,
+      "loss": 0.3964,
+      "step": 45790
+    },
+    {
+      "epoch": 4.587369159112535,
+      "grad_norm": 2.351881265640259,
+      "learning_rate": 8.393868624551743e-07,
+      "loss": 0.3971,
+      "step": 45800
+    },
+    {
+      "epoch": 4.588370811839535,
+      "grad_norm": 1.9069644212722778,
+      "learning_rate": 8.353409122755202e-07,
+      "loss": 0.4131,
+      "step": 45810
+    },
+    {
+      "epoch": 4.589372464566535,
+      "grad_norm": 2.0071794986724854,
+      "learning_rate": 8.313045709800071e-07,
+      "loss": 0.3534,
+      "step": 45820
+    },
+    {
+      "epoch": 4.590374117293535,
+      "grad_norm": 1.9949153661727905,
+      "learning_rate": 8.272778401736652e-07,
+      "loss": 0.4288,
+      "step": 45830
+    },
+    {
+      "epoch": 4.591375770020534,
+      "grad_norm": 2.0804128646850586,
+      "learning_rate": 8.232607214576859e-07,
+      "loss": 0.3968,
+      "step": 45840
+    },
+    {
+      "epoch": 4.592377422747534,
+      "grad_norm": 2.215404987335205,
+      "learning_rate": 8.192532164294414e-07,
+      "loss": 0.4034,
+      "step": 45850
+    },
+    {
+      "epoch": 4.593379075474533,
+      "grad_norm": 2.0437800884246826,
+      "learning_rate": 8.152553266824875e-07,
+      "loss": 0.4487,
+      "step": 45860
+    },
+    {
+      "epoch": 4.594380728201532,
+      "grad_norm": 2.4165077209472656,
+      "learning_rate": 8.112670538065553e-07,
+      "loss": 0.3931,
+      "step": 45870
+    },
+    {
+      "epoch": 4.595382380928532,
+      "grad_norm": 2.1966311931610107,
+      "learning_rate": 8.072883993875429e-07,
+      "loss": 0.3606,
+      "step": 45880
+    },
+    {
+      "epoch": 4.596384033655531,
+      "grad_norm": 1.7312675714492798,
+      "learning_rate": 8.033193650075349e-07,
+      "loss": 0.4133,
+      "step": 45890
+    },
+    {
+      "epoch": 4.597385686382531,
+      "grad_norm": 1.9515161514282227,
+      "learning_rate": 7.993599522447881e-07,
+      "loss": 0.4031,
+      "step": 45900
+    },
+    {
+      "epoch": 4.5983873391095305,
+      "grad_norm": 2.2079174518585205,
+      "learning_rate": 7.954101626737321e-07,
+      "loss": 0.431,
+      "step": 45910
+    },
+    {
+      "epoch": 4.5993889918365305,
+      "grad_norm": 2.438955783843994,
+      "learning_rate": 7.914699978649604e-07,
+      "loss": 0.3663,
+      "step": 45920
+    },
+    {
+      "epoch": 4.60039064456353,
+      "grad_norm": 2.7964272499084473,
+      "learning_rate": 7.875394593852559e-07,
+      "loss": 0.4036,
+      "step": 45930
+    },
+    {
+      "epoch": 4.60139229729053,
+      "grad_norm": 2.09883713722229,
+      "learning_rate": 7.836185487975655e-07,
+      "loss": 0.4253,
+      "step": 45940
+    },
+    {
+      "epoch": 4.602393950017529,
+      "grad_norm": 2.234827995300293,
+      "learning_rate": 7.797072676610062e-07,
+      "loss": 0.3641,
+      "step": 45950
+    },
+    {
+      "epoch": 4.603395602744529,
+      "grad_norm": 2.1803348064422607,
+      "learning_rate": 7.75805617530867e-07,
+      "loss": 0.38,
+      "step": 45960
+    },
+    {
+      "epoch": 4.604397255471528,
+      "grad_norm": 2.11966872215271,
+      "learning_rate": 7.719135999586125e-07,
+      "loss": 0.402,
+      "step": 45970
+    },
+    {
+      "epoch": 4.605398908198527,
+      "grad_norm": 1.670632243156433,
+      "learning_rate": 7.680312164918657e-07,
+      "loss": 0.356,
+      "step": 45980
+    },
+    {
+      "epoch": 4.606400560925527,
+      "grad_norm": 2.887688636779785,
+      "learning_rate": 7.641584686744308e-07,
+      "loss": 0.4281,
+      "step": 45990
+    },
+    {
+      "epoch": 4.607402213652526,
+      "grad_norm": 1.9670038223266602,
+      "learning_rate": 7.602953580462729e-07,
+      "loss": 0.4214,
+      "step": 46000
+    },
+    {
+      "epoch": 4.608403866379526,
+      "grad_norm": 2.14485239982605,
+      "learning_rate": 7.564418861435301e-07,
+      "loss": 0.3554,
+      "step": 46010
+    },
+    {
+      "epoch": 4.6094055191065255,
+      "grad_norm": 1.9791523218154907,
+      "learning_rate": 7.525980544984989e-07,
+      "loss": 0.3751,
+      "step": 46020
+    },
+    {
+      "epoch": 4.6104071718335256,
+      "grad_norm": 2.46907901763916,
+      "learning_rate": 7.487638646396539e-07,
+      "loss": 0.4055,
+      "step": 46030
+    },
+    {
+      "epoch": 4.611408824560525,
+      "grad_norm": 2.633350372314453,
+      "learning_rate": 7.449393180916281e-07,
+      "loss": 0.404,
+      "step": 46040
+    },
+    {
+      "epoch": 4.612410477287525,
+      "grad_norm": 2.210999011993408,
+      "learning_rate": 7.411244163752163e-07,
+      "loss": 0.4498,
+      "step": 46050
+    },
+    {
+      "epoch": 4.613412130014524,
+      "grad_norm": 2.5408952236175537,
+      "learning_rate": 7.373191610073965e-07,
+      "loss": 0.4118,
+      "step": 46060
+    },
+    {
+      "epoch": 4.614413782741524,
+      "grad_norm": 2.2600090503692627,
+      "learning_rate": 7.335235535012891e-07,
+      "loss": 0.411,
+      "step": 46070
+    },
+    {
+      "epoch": 4.615415435468523,
+      "grad_norm": 2.1320509910583496,
+      "learning_rate": 7.297375953661867e-07,
+      "loss": 0.4412,
+      "step": 46080
+    },
+    {
+      "epoch": 4.616417088195522,
+      "grad_norm": 2.4685146808624268,
+      "learning_rate": 7.259612881075517e-07,
+      "loss": 0.4401,
+      "step": 46090
+    },
+    {
+      "epoch": 4.617418740922522,
+      "grad_norm": 1.9967085123062134,
+      "learning_rate": 7.221946332269968e-07,
+      "loss": 0.4211,
+      "step": 46100
+    },
+    {
+      "epoch": 4.618420393649521,
+      "grad_norm": 1.9377949237823486,
+      "learning_rate": 7.184376322223019e-07,
+      "loss": 0.3698,
+      "step": 46110
+    },
+    {
+      "epoch": 4.619422046376521,
+      "grad_norm": 2.3374502658843994,
+      "learning_rate": 7.146902865874105e-07,
+      "loss": 0.3967,
+      "step": 46120
+    },
+    {
+      "epoch": 4.620423699103521,
+      "grad_norm": 3.4399402141571045,
+      "learning_rate": 7.10952597812431e-07,
+      "loss": 0.3948,
+      "step": 46130
+    },
+    {
+      "epoch": 4.621425351830521,
+      "grad_norm": 1.7531739473342896,
+      "learning_rate": 7.072245673836131e-07,
+      "loss": 0.4081,
+      "step": 46140
+    },
+    {
+      "epoch": 4.62242700455752,
+      "grad_norm": 2.1112916469573975,
+      "learning_rate": 7.03506196783385e-07,
+      "loss": 0.3658,
+      "step": 46150
+    },
+    {
+      "epoch": 4.62342865728452,
+      "grad_norm": 3.10821533203125,
+      "learning_rate": 6.997974874903334e-07,
+      "loss": 0.4272,
+      "step": 46160
+    },
+    {
+      "epoch": 4.624430310011519,
+      "grad_norm": 2.686772346496582,
+      "learning_rate": 6.960984409791871e-07,
+      "loss": 0.3991,
+      "step": 46170
+    },
+    {
+      "epoch": 4.625431962738519,
+      "grad_norm": 2.0568907260894775,
+      "learning_rate": 6.924090587208415e-07,
+      "loss": 0.4252,
+      "step": 46180
+    },
+    {
+      "epoch": 4.626433615465518,
+      "grad_norm": 2.369635581970215,
+      "learning_rate": 6.887293421823593e-07,
+      "loss": 0.4062,
+      "step": 46190
+    },
+    {
+      "epoch": 4.627435268192517,
+      "grad_norm": 2.34002947807312,
+      "learning_rate": 6.850592928269478e-07,
+      "loss": 0.3955,
+      "step": 46200
+    },
+    {
+      "epoch": 4.628436920919517,
+      "grad_norm": 2.856238603591919,
+      "learning_rate": 6.813989121139647e-07,
+      "loss": 0.4216,
+      "step": 46210
+    },
+    {
+      "epoch": 4.6294385736465165,
+      "grad_norm": 2.129307508468628,
+      "learning_rate": 6.7774820149894e-07,
+      "loss": 0.3689,
+      "step": 46220
+    },
+    {
+      "epoch": 4.6304402263735165,
+      "grad_norm": 2.2219908237457275,
+      "learning_rate": 6.741071624335459e-07,
+      "loss": 0.4228,
+      "step": 46230
+    },
+    {
+      "epoch": 4.631441879100516,
+      "grad_norm": 2.6033074855804443,
+      "learning_rate": 6.704757963656189e-07,
+      "loss": 0.4703,
+      "step": 46240
+    },
+    {
+      "epoch": 4.632443531827516,
+      "grad_norm": 2.086947441101074,
+      "learning_rate": 6.668541047391313e-07,
+      "loss": 0.3758,
+      "step": 46250
+    },
+    {
+      "epoch": 4.633445184554515,
+      "grad_norm": 1.4777419567108154,
+      "learning_rate": 6.632420889942287e-07,
+      "loss": 0.3655,
+      "step": 46260
+    },
+    {
+      "epoch": 4.634446837281515,
+      "grad_norm": 2.539104461669922,
+      "learning_rate": 6.596397505672009e-07,
+      "loss": 0.4336,
+      "step": 46270
+    },
+    {
+      "epoch": 4.635448490008514,
+      "grad_norm": 2.2465150356292725,
+      "learning_rate": 6.560470908904798e-07,
+      "loss": 0.4204,
+      "step": 46280
+    },
+    {
+      "epoch": 4.636450142735514,
+      "grad_norm": 2.106008529663086,
+      "learning_rate": 6.524641113926672e-07,
+      "loss": 0.4228,
+      "step": 46290
+    },
+    {
+      "epoch": 4.637451795462513,
+      "grad_norm": 3.1730639934539795,
+      "learning_rate": 6.488908134985011e-07,
+      "loss": 0.3906,
+      "step": 46300
+    },
+    {
+      "epoch": 4.638453448189512,
+      "grad_norm": 2.27172589302063,
+      "learning_rate": 6.453271986288812e-07,
+      "loss": 0.3889,
+      "step": 46310
+    },
+    {
+      "epoch": 4.639455100916512,
+      "grad_norm": 2.1306052207946777,
+      "learning_rate": 6.417732682008431e-07,
+      "loss": 0.4226,
+      "step": 46320
+    },
+    {
+      "epoch": 4.6404567536435115,
+      "grad_norm": 3.0258591175079346,
+      "learning_rate": 6.382290236275845e-07,
+      "loss": 0.4184,
+      "step": 46330
+    },
+    {
+      "epoch": 4.6414584063705115,
+      "grad_norm": 1.9228705167770386,
+      "learning_rate": 6.346944663184418e-07,
+      "loss": 0.3917,
+      "step": 46340
+    },
+    {
+      "epoch": 4.642460059097511,
+      "grad_norm": 2.0833024978637695,
+      "learning_rate": 6.311695976789073e-07,
+      "loss": 0.4407,
+      "step": 46350
+    },
+    {
+      "epoch": 4.643461711824511,
+      "grad_norm": 2.445340871810913,
+      "learning_rate": 6.276544191106154e-07,
+      "loss": 0.3477,
+      "step": 46360
+    },
+    {
+      "epoch": 4.64446336455151,
+      "grad_norm": 2.1367125511169434,
+      "learning_rate": 6.241489320113453e-07,
+      "loss": 0.3996,
+      "step": 46370
+    },
+    {
+      "epoch": 4.64546501727851,
+      "grad_norm": 2.3300392627716064,
+      "learning_rate": 6.206531377750319e-07,
+      "loss": 0.3829,
+      "step": 46380
+    },
+    {
+      "epoch": 4.646466670005509,
+      "grad_norm": 2.5135016441345215,
+      "learning_rate": 6.171670377917465e-07,
+      "loss": 0.3779,
+      "step": 46390
+    },
+    {
+      "epoch": 4.647468322732509,
+      "grad_norm": 2.5999858379364014,
+      "learning_rate": 6.136906334477111e-07,
+      "loss": 0.3559,
+      "step": 46400
+    },
+    {
+      "epoch": 4.648469975459508,
+      "grad_norm": 2.109624147415161,
+      "learning_rate": 6.102239261252862e-07,
+      "loss": 0.3592,
+      "step": 46410
+    },
+    {
+      "epoch": 4.649471628186507,
+      "grad_norm": 2.480128526687622,
+      "learning_rate": 6.067669172029888e-07,
+      "loss": 0.4487,
+      "step": 46420
+    },
+    {
+      "epoch": 4.650473280913507,
+      "grad_norm": 2.0874390602111816,
+      "learning_rate": 6.03319608055461e-07,
+      "loss": 0.4079,
+      "step": 46430
+    },
+    {
+      "epoch": 4.651474933640507,
+      "grad_norm": 2.6308510303497314,
+      "learning_rate": 5.998820000535005e-07,
+      "loss": 0.4327,
+      "step": 46440
+    },
+    {
+      "epoch": 4.652476586367507,
+      "grad_norm": 1.756169319152832,
+      "learning_rate": 5.964540945640501e-07,
+      "loss": 0.4047,
+      "step": 46450
+    },
+    {
+      "epoch": 4.653478239094506,
+      "grad_norm": 1.9640753269195557,
+      "learning_rate": 5.930358929501834e-07,
+      "loss": 0.4348,
+      "step": 46460
+    },
+    {
+      "epoch": 4.654479891821506,
+      "grad_norm": 1.7669875621795654,
+      "learning_rate": 5.896273965711213e-07,
+      "loss": 0.3855,
+      "step": 46470
+    },
+    {
+      "epoch": 4.655481544548505,
+      "grad_norm": 3.000284433364868,
+      "learning_rate": 5.86228606782227e-07,
+      "loss": 0.3811,
+      "step": 46480
+    },
+    {
+      "epoch": 4.656483197275505,
+      "grad_norm": 2.0170233249664307,
+      "learning_rate": 5.828395249350054e-07,
+      "loss": 0.3869,
+      "step": 46490
+    },
+    {
+      "epoch": 4.657484850002504,
+      "grad_norm": 2.0546176433563232,
+      "learning_rate": 5.794601523770926e-07,
+      "loss": 0.4393,
+      "step": 46500
+    },
+    {
+      "epoch": 4.658486502729504,
+      "grad_norm": 2.225003719329834,
+      "learning_rate": 5.76090490452269e-07,
+      "loss": 0.4438,
+      "step": 46510
+    },
+    {
+      "epoch": 4.659488155456503,
+      "grad_norm": 1.9720081090927124,
+      "learning_rate": 5.727305405004574e-07,
+      "loss": 0.3664,
+      "step": 46520
+    },
+    {
+      "epoch": 4.6604898081835024,
+      "grad_norm": 2.1545374393463135,
+      "learning_rate": 5.693803038577167e-07,
+      "loss": 0.3775,
+      "step": 46530
+    },
+    {
+      "epoch": 4.6614914609105025,
+      "grad_norm": 1.8879632949829102,
+      "learning_rate": 5.660397818562341e-07,
+      "loss": 0.4258,
+      "step": 46540
+    },
+    {
+      "epoch": 4.662493113637502,
+      "grad_norm": 2.4001457691192627,
+      "learning_rate": 5.627089758243498e-07,
+      "loss": 0.3909,
+      "step": 46550
+    },
+    {
+      "epoch": 4.663494766364502,
+      "grad_norm": 2.4746174812316895,
+      "learning_rate": 5.593878870865294e-07,
+      "loss": 0.4284,
+      "step": 46560
+    },
+    {
+      "epoch": 4.664496419091501,
+      "grad_norm": 2.5044631958007812,
+      "learning_rate": 5.56076516963372e-07,
+      "loss": 0.4111,
+      "step": 46570
+    },
+    {
+      "epoch": 4.665498071818501,
+      "grad_norm": 2.208523988723755,
+      "learning_rate": 5.527748667716243e-07,
+      "loss": 0.4011,
+      "step": 46580
+    },
+    {
+      "epoch": 4.6664997245455,
+      "grad_norm": 1.9320225715637207,
+      "learning_rate": 5.494829378241584e-07,
+      "loss": 0.4025,
+      "step": 46590
+    },
+    {
+      "epoch": 4.667501377272499,
+      "grad_norm": 2.7530622482299805,
+      "learning_rate": 5.462007314299883e-07,
+      "loss": 0.3827,
+      "step": 46600
+    },
+    {
+      "epoch": 4.668503029999499,
+      "grad_norm": 1.7465649843215942,
+      "learning_rate": 5.429282488942477e-07,
+      "loss": 0.3867,
+      "step": 46610
+    },
+    {
+      "epoch": 4.669504682726499,
+      "grad_norm": 1.7411669492721558,
+      "learning_rate": 5.396654915182209e-07,
+      "loss": 0.4054,
+      "step": 46620
+    },
+    {
+      "epoch": 4.670506335453498,
+      "grad_norm": 1.8928498029708862,
+      "learning_rate": 5.36412460599317e-07,
+      "loss": 0.4182,
+      "step": 46630
+    },
+    {
+      "epoch": 4.6715079881804975,
+      "grad_norm": 2.3412082195281982,
+      "learning_rate": 5.33169157431071e-07,
+      "loss": 0.4115,
+      "step": 46640
+    },
+    {
+      "epoch": 4.6725096409074975,
+      "grad_norm": 2.5557522773742676,
+      "learning_rate": 5.29935583303165e-07,
+      "loss": 0.3666,
+      "step": 46650
+    },
+    {
+      "epoch": 4.673511293634497,
+      "grad_norm": 2.372802972793579,
+      "learning_rate": 5.267117395014009e-07,
+      "loss": 0.4566,
+      "step": 46660
+    },
+    {
+      "epoch": 4.674512946361497,
+      "grad_norm": 2.2661385536193848,
+      "learning_rate": 5.234976273077147e-07,
+      "loss": 0.417,
+      "step": 46670
+    },
+    {
+      "epoch": 4.675514599088496,
+      "grad_norm": 2.0665102005004883,
+      "learning_rate": 5.202932480001699e-07,
+      "loss": 0.3484,
+      "step": 46680
+    },
+    {
+      "epoch": 4.676516251815496,
+      "grad_norm": 2.0566537380218506,
+      "learning_rate": 5.170986028529667e-07,
+      "loss": 0.3725,
+      "step": 46690
+    },
+    {
+      "epoch": 4.677517904542495,
+      "grad_norm": 1.9390207529067993,
+      "learning_rate": 5.139136931364252e-07,
+      "loss": 0.4094,
+      "step": 46700
+    },
+    {
+      "epoch": 4.678519557269494,
+      "grad_norm": 2.6145527362823486,
+      "learning_rate": 5.107385201170045e-07,
+      "loss": 0.4093,
+      "step": 46710
+    },
+    {
+      "epoch": 4.679521209996494,
+      "grad_norm": 1.9652702808380127,
+      "learning_rate": 5.075730850572835e-07,
+      "loss": 0.416,
+      "step": 46720
+    },
+    {
+      "epoch": 4.680522862723493,
+      "grad_norm": 2.30545973777771,
+      "learning_rate": 5.044173892159748e-07,
+      "loss": 0.3941,
+      "step": 46730
+    },
+    {
+      "epoch": 4.681524515450493,
+      "grad_norm": 1.984140157699585,
+      "learning_rate": 5.012714338479135e-07,
+      "loss": 0.4178,
+      "step": 46740
+    },
+    {
+      "epoch": 4.682526168177493,
+      "grad_norm": 2.0339691638946533,
+      "learning_rate": 4.981352202040628e-07,
+      "loss": 0.3753,
+      "step": 46750
+    },
+    {
+      "epoch": 4.683527820904493,
+      "grad_norm": 1.7999358177185059,
+      "learning_rate": 4.95008749531517e-07,
+      "loss": 0.3793,
+      "step": 46760
+    },
+    {
+      "epoch": 4.684529473631492,
+      "grad_norm": 1.8290388584136963,
+      "learning_rate": 4.91892023073487e-07,
+      "loss": 0.4243,
+      "step": 46770
+    },
+    {
+      "epoch": 4.685531126358492,
+      "grad_norm": 2.1667447090148926,
+      "learning_rate": 4.887850420693202e-07,
+      "loss": 0.4241,
+      "step": 46780
+    },
+    {
+      "epoch": 4.686532779085491,
+      "grad_norm": 2.310354471206665,
+      "learning_rate": 4.856878077544785e-07,
+      "loss": 0.4206,
+      "step": 46790
+    },
+    {
+      "epoch": 4.687534431812491,
+      "grad_norm": 2.7219927310943604,
+      "learning_rate": 4.826003213605545e-07,
+      "loss": 0.4295,
+      "step": 46800
+    },
+    {
+      "epoch": 4.68853608453949,
+      "grad_norm": 1.945172667503357,
+      "learning_rate": 4.795225841152579e-07,
+      "loss": 0.4499,
+      "step": 46810
+    },
+    {
+      "epoch": 4.689537737266489,
+      "grad_norm": 2.1988821029663086,
+      "learning_rate": 4.7645459724243444e-07,
+      "loss": 0.403,
+      "step": 46820
+    },
+    {
+      "epoch": 4.690539389993489,
+      "grad_norm": 2.19480562210083,
+      "learning_rate": 4.7339636196204184e-07,
+      "loss": 0.3844,
+      "step": 46830
+    },
+    {
+      "epoch": 4.691541042720488,
+      "grad_norm": 2.085426092147827,
+      "learning_rate": 4.703478794901572e-07,
+      "loss": 0.4018,
+      "step": 46840
+    },
+    {
+      "epoch": 4.6925426954474885,
+      "grad_norm": 2.0894999504089355,
+      "learning_rate": 4.673091510389943e-07,
+      "loss": 0.3954,
+      "step": 46850
+    },
+    {
+      "epoch": 4.693544348174488,
+      "grad_norm": 2.0696043968200684,
+      "learning_rate": 4.642801778168726e-07,
+      "loss": 0.4097,
+      "step": 46860
+    },
+    {
+      "epoch": 4.694546000901488,
+      "grad_norm": 2.10793137550354,
+      "learning_rate": 4.61260961028237e-07,
+      "loss": 0.3324,
+      "step": 46870
+    },
+    {
+      "epoch": 4.695547653628487,
+      "grad_norm": 2.361173629760742,
+      "learning_rate": 4.582515018736633e-07,
+      "loss": 0.4199,
+      "step": 46880
+    },
+    {
+      "epoch": 4.696549306355487,
+      "grad_norm": 2.2172489166259766,
+      "learning_rate": 4.552518015498386e-07,
+      "loss": 0.3685,
+      "step": 46890
+    },
+    {
+      "epoch": 4.697550959082486,
+      "grad_norm": 2.353889226913452,
+      "learning_rate": 4.522618612495588e-07,
+      "loss": 0.4178,
+      "step": 46900
+    },
+    {
+      "epoch": 4.698552611809486,
+      "grad_norm": 1.8032987117767334,
+      "learning_rate": 4.492816821617618e-07,
+      "loss": 0.4407,
+      "step": 46910
+    },
+    {
+      "epoch": 4.699554264536485,
+      "grad_norm": 2.0415968894958496,
+      "learning_rate": 4.463112654714885e-07,
+      "loss": 0.3827,
+      "step": 46920
+    },
+    {
+      "epoch": 4.700555917263484,
+      "grad_norm": 2.6588540077209473,
+      "learning_rate": 4.433506123598996e-07,
+      "loss": 0.3954,
+      "step": 46930
+    },
+    {
+      "epoch": 4.701557569990484,
+      "grad_norm": 2.428084373474121,
+      "learning_rate": 4.4039972400427286e-07,
+      "loss": 0.4364,
+      "step": 46940
+    },
+    {
+      "epoch": 4.7025592227174835,
+      "grad_norm": 2.2132630348205566,
+      "learning_rate": 4.374586015780113e-07,
+      "loss": 0.474,
+      "step": 46950
+    },
+    {
+      "epoch": 4.7035608754444835,
+      "grad_norm": 2.4782333374023438,
+      "learning_rate": 4.3452724625062946e-07,
+      "loss": 0.4194,
+      "step": 46960
+    },
+    {
+      "epoch": 4.704562528171483,
+      "grad_norm": 1.8694998025894165,
+      "learning_rate": 4.3160565918774767e-07,
+      "loss": 0.379,
+      "step": 46970
+    },
+    {
+      "epoch": 4.705564180898483,
+      "grad_norm": 1.9565788507461548,
+      "learning_rate": 4.286938415511227e-07,
+      "loss": 0.4134,
+      "step": 46980
+    },
+    {
+      "epoch": 4.706565833625482,
+      "grad_norm": 2.3455045223236084,
+      "learning_rate": 4.2579179449860896e-07,
+      "loss": 0.3826,
+      "step": 46990
+    },
+    {
+      "epoch": 4.707567486352482,
+      "grad_norm": 1.9453314542770386,
+      "learning_rate": 4.228995191841861e-07,
+      "loss": 0.3892,
+      "step": 47000
+    },
+    {
+      "epoch": 4.708569139079481,
+      "grad_norm": 1.8712366819381714,
+      "learning_rate": 4.200170167579426e-07,
+      "loss": 0.4426,
+      "step": 47010
+    },
+    {
+      "epoch": 4.709570791806481,
+      "grad_norm": 1.9517831802368164,
+      "learning_rate": 4.171442883660809e-07,
+      "loss": 0.371,
+      "step": 47020
+    },
+    {
+      "epoch": 4.71057244453348,
+      "grad_norm": 2.4578795433044434,
+      "learning_rate": 4.142813351509234e-07,
+      "loss": 0.4289,
+      "step": 47030
+    },
+    {
+      "epoch": 4.711574097260479,
+      "grad_norm": 1.4846292734146118,
+      "learning_rate": 4.114281582508955e-07,
+      "loss": 0.3575,
+      "step": 47040
+    },
+    {
+      "epoch": 4.712575749987479,
+      "grad_norm": 2.2234723567962646,
+      "learning_rate": 4.0858475880054537e-07,
+      "loss": 0.4322,
+      "step": 47050
+    },
+    {
+      "epoch": 4.713577402714479,
+      "grad_norm": 1.7275258302688599,
+      "learning_rate": 4.057511379305212e-07,
+      "loss": 0.3939,
+      "step": 47060
+    },
+    {
+      "epoch": 4.714579055441479,
+      "grad_norm": 2.4854557514190674,
+      "learning_rate": 4.029272967675968e-07,
+      "loss": 0.4135,
+      "step": 47070
+    },
+    {
+      "epoch": 4.715580708168478,
+      "grad_norm": 3.625199317932129,
+      "learning_rate": 4.0011323643464605e-07,
+      "loss": 0.4075,
+      "step": 47080
+    },
+    {
+      "epoch": 4.716582360895478,
+      "grad_norm": 1.9963372945785522,
+      "learning_rate": 3.9730895805066e-07,
+      "loss": 0.4061,
+      "step": 47090
+    },
+    {
+      "epoch": 4.717584013622477,
+      "grad_norm": 2.0822815895080566,
+      "learning_rate": 3.945144627307329e-07,
+      "loss": 0.42,
+      "step": 47100
+    },
+    {
+      "epoch": 4.718585666349477,
+      "grad_norm": 1.8344260454177856,
+      "learning_rate": 3.917297515860813e-07,
+      "loss": 0.4285,
+      "step": 47110
+    },
+    {
+      "epoch": 4.719587319076476,
+      "grad_norm": 2.9358978271484375,
+      "learning_rate": 3.8895482572401664e-07,
+      "loss": 0.4028,
+      "step": 47120
+    },
+    {
+      "epoch": 4.720588971803476,
+      "grad_norm": 1.8024274110794067,
+      "learning_rate": 3.861896862479675e-07,
+      "loss": 0.3655,
+      "step": 47130
+    },
+    {
+      "epoch": 4.721590624530475,
+      "grad_norm": 2.0380611419677734,
+      "learning_rate": 3.8343433425747365e-07,
+      "loss": 0.3879,
+      "step": 47140
+    },
+    {
+      "epoch": 4.722592277257474,
+      "grad_norm": 1.7357391119003296,
+      "learning_rate": 3.8068877084817267e-07,
+      "loss": 0.3821,
+      "step": 47150
+    },
+    {
+      "epoch": 4.7235939299844745,
+      "grad_norm": 2.1513285636901855,
+      "learning_rate": 3.7795299711182173e-07,
+      "loss": 0.4278,
+      "step": 47160
+    },
+    {
+      "epoch": 4.724595582711474,
+      "grad_norm": 2.3541009426116943,
+      "learning_rate": 3.752270141362729e-07,
+      "loss": 0.4123,
+      "step": 47170
+    },
+    {
+      "epoch": 4.725597235438474,
+      "grad_norm": 2.2936346530914307,
+      "learning_rate": 3.72510823005498e-07,
+      "loss": 0.4565,
+      "step": 47180
+    },
+    {
+      "epoch": 4.726598888165473,
+      "grad_norm": 2.1213440895080566,
+      "learning_rate": 3.6980442479956633e-07,
+      "loss": 0.3844,
+      "step": 47190
+    },
+    {
+      "epoch": 4.727600540892473,
+      "grad_norm": 2.0257651805877686,
+      "learning_rate": 3.671078205946532e-07,
+      "loss": 0.4299,
+      "step": 47200
+    },
+    {
+      "epoch": 4.728602193619472,
+      "grad_norm": 2.968496561050415,
+      "learning_rate": 3.644210114630481e-07,
+      "loss": 0.4463,
+      "step": 47210
+    },
+    {
+      "epoch": 4.729603846346472,
+      "grad_norm": 1.8193625211715698,
+      "learning_rate": 3.6174399847313525e-07,
+      "loss": 0.4118,
+      "step": 47220
+    },
+    {
+      "epoch": 4.730605499073471,
+      "grad_norm": 2.025061845779419,
+      "learning_rate": 3.5907678268940206e-07,
+      "loss": 0.3836,
+      "step": 47230
+    },
+    {
+      "epoch": 4.731607151800471,
+      "grad_norm": 1.9306206703186035,
+      "learning_rate": 3.564193651724557e-07,
+      "loss": 0.3626,
+      "step": 47240
+    },
+    {
+      "epoch": 4.73260880452747,
+      "grad_norm": 2.3328723907470703,
+      "learning_rate": 3.5377174697899253e-07,
+      "loss": 0.4453,
+      "step": 47250
+    },
+    {
+      "epoch": 4.7336104572544695,
+      "grad_norm": 2.4250965118408203,
+      "learning_rate": 3.5113392916181485e-07,
+      "loss": 0.4006,
+      "step": 47260
+    },
+    {
+      "epoch": 4.7346121099814695,
+      "grad_norm": 2.66080904006958,
+      "learning_rate": 3.485059127698309e-07,
+      "loss": 0.4236,
+      "step": 47270
+    },
+    {
+      "epoch": 4.735613762708469,
+      "grad_norm": 2.2034497261047363,
+      "learning_rate": 3.4588769884805473e-07,
+      "loss": 0.3963,
+      "step": 47280
+    },
+    {
+      "epoch": 4.736615415435469,
+      "grad_norm": 2.1475400924682617,
+      "learning_rate": 3.4327928843759517e-07,
+      "loss": 0.4313,
+      "step": 47290
+    },
+    {
+      "epoch": 4.737617068162468,
+      "grad_norm": 2.439648389816284,
+      "learning_rate": 3.406806825756614e-07,
+      "loss": 0.4619,
+      "step": 47300
+    },
+    {
+      "epoch": 4.738618720889468,
+      "grad_norm": 1.9085047245025635,
+      "learning_rate": 3.380918822955742e-07,
+      "loss": 0.3627,
+      "step": 47310
+    },
+    {
+      "epoch": 4.739620373616467,
+      "grad_norm": 2.736255168914795,
+      "learning_rate": 3.3551288862675167e-07,
+      "loss": 0.415,
+      "step": 47320
+    },
+    {
+      "epoch": 4.740622026343467,
+      "grad_norm": 2.4652960300445557,
+      "learning_rate": 3.329437025947013e-07,
+      "loss": 0.4147,
+      "step": 47330
+    },
+    {
+      "epoch": 4.741623679070466,
+      "grad_norm": 2.4739601612091064,
+      "learning_rate": 3.303843252210448e-07,
+      "loss": 0.4556,
+      "step": 47340
+    },
+    {
+      "epoch": 4.742625331797466,
+      "grad_norm": 2.4230849742889404,
+      "learning_rate": 3.278347575234986e-07,
+      "loss": 0.3888,
+      "step": 47350
+    },
+    {
+      "epoch": 4.743626984524465,
+      "grad_norm": 2.326728105545044,
+      "learning_rate": 3.252950005158767e-07,
+      "loss": 0.4818,
+      "step": 47360
+    },
+    {
+      "epoch": 4.7446286372514646,
+      "grad_norm": 2.3454596996307373,
+      "learning_rate": 3.2276505520809353e-07,
+      "loss": 0.3918,
+      "step": 47370
+    },
+    {
+      "epoch": 4.745630289978465,
+      "grad_norm": 2.66456937789917,
+      "learning_rate": 3.20244922606161e-07,
+      "loss": 0.4257,
+      "step": 47380
+    },
+    {
+      "epoch": 4.746631942705464,
+      "grad_norm": 2.293207883834839,
+      "learning_rate": 3.1773460371219144e-07,
+      "loss": 0.4507,
+      "step": 47390
+    },
+    {
+      "epoch": 4.747633595432464,
+      "grad_norm": 2.403235673904419,
+      "learning_rate": 3.1523409952439465e-07,
+      "loss": 0.4127,
+      "step": 47400
+    },
+    {
+      "epoch": 4.748635248159463,
+      "grad_norm": 2.0613138675689697,
+      "learning_rate": 3.1274341103706973e-07,
+      "loss": 0.351,
+      "step": 47410
+    },
+    {
+      "epoch": 4.749636900886463,
+      "grad_norm": 2.4735984802246094,
+      "learning_rate": 3.102625392406244e-07,
+      "loss": 0.44,
+      "step": 47420
+    },
+    {
+      "epoch": 4.750638553613462,
+      "grad_norm": 1.6152466535568237,
+      "learning_rate": 3.077914851215585e-07,
+      "loss": 0.4251,
+      "step": 47430
+    },
+    {
+      "epoch": 4.751640206340462,
+      "grad_norm": 1.6947752237319946,
+      "learning_rate": 3.05330249662461e-07,
+      "loss": 0.3719,
+      "step": 47440
+    },
+    {
+      "epoch": 4.752641859067461,
+      "grad_norm": 2.6465158462524414,
+      "learning_rate": 3.0287883384202965e-07,
+      "loss": 0.4549,
+      "step": 47450
+    },
+    {
+      "epoch": 4.753643511794461,
+      "grad_norm": 2.3784332275390625,
+      "learning_rate": 3.004372386350457e-07,
+      "loss": 0.4461,
+      "step": 47460
+    },
+    {
+      "epoch": 4.7546451645214605,
+      "grad_norm": 2.1705985069274902,
+      "learning_rate": 2.980054650123909e-07,
+      "loss": 0.3915,
+      "step": 47470
+    },
+    {
+      "epoch": 4.75564681724846,
+      "grad_norm": 2.2914929389953613,
+      "learning_rate": 2.955835139410418e-07,
+      "loss": 0.4398,
+      "step": 47480
+    },
+    {
+      "epoch": 4.75664846997546,
+      "grad_norm": 2.312525749206543,
+      "learning_rate": 2.9317138638406684e-07,
+      "loss": 0.4375,
+      "step": 47490
+    },
+    {
+      "epoch": 4.757650122702459,
+      "grad_norm": 1.8366079330444336,
+      "learning_rate": 2.9076908330062937e-07,
+      "loss": 0.3871,
+      "step": 47500
+    },
+    {
+      "epoch": 4.758651775429459,
+      "grad_norm": 2.4751739501953125,
+      "learning_rate": 2.8837660564598747e-07,
+      "loss": 0.4007,
+      "step": 47510
+    },
+    {
+      "epoch": 4.759653428156458,
+      "grad_norm": 1.7309223413467407,
+      "learning_rate": 2.859939543714912e-07,
+      "loss": 0.4545,
+      "step": 47520
+    },
+    {
+      "epoch": 4.760655080883458,
+      "grad_norm": 2.232672929763794,
+      "learning_rate": 2.836211304245773e-07,
+      "loss": 0.4088,
+      "step": 47530
+    },
+    {
+      "epoch": 4.761656733610457,
+      "grad_norm": 2.060687303543091,
+      "learning_rate": 2.812581347487908e-07,
+      "loss": 0.3797,
+      "step": 47540
+    },
+    {
+      "epoch": 4.762658386337456,
+      "grad_norm": 2.090240716934204,
+      "learning_rate": 2.789049682837469e-07,
+      "loss": 0.4377,
+      "step": 47550
+    },
+    {
+      "epoch": 4.763660039064456,
+      "grad_norm": 1.645607590675354,
+      "learning_rate": 2.765616319651693e-07,
+      "loss": 0.3978,
+      "step": 47560
+    },
+    {
+      "epoch": 4.764661691791456,
+      "grad_norm": 1.79109525680542,
+      "learning_rate": 2.742281267248681e-07,
+      "loss": 0.4024,
+      "step": 47570
+    },
+    {
+      "epoch": 4.7656633445184555,
+      "grad_norm": 2.5542709827423096,
+      "learning_rate": 2.7190445349074e-07,
+      "loss": 0.4798,
+      "step": 47580
+    },
+    {
+      "epoch": 4.766664997245455,
+      "grad_norm": 2.0201637744903564,
+      "learning_rate": 2.6959061318677645e-07,
+      "loss": 0.4226,
+      "step": 47590
+    },
+    {
+      "epoch": 4.767666649972455,
+      "grad_norm": 2.7899389266967773,
+      "learning_rate": 2.672866067330554e-07,
+      "loss": 0.4479,
+      "step": 47600
+    },
+    {
+      "epoch": 4.768668302699454,
+      "grad_norm": 2.2652668952941895,
+      "learning_rate": 2.649924350457522e-07,
+      "loss": 0.4183,
+      "step": 47610
+    },
+    {
+      "epoch": 4.769669955426454,
+      "grad_norm": 2.208167314529419,
+      "learning_rate": 2.6270809903712057e-07,
+      "loss": 0.379,
+      "step": 47620
+    },
+    {
+      "epoch": 4.770671608153453,
+      "grad_norm": 2.410855293273926,
+      "learning_rate": 2.604335996155088e-07,
+      "loss": 0.4361,
+      "step": 47630
+    },
+    {
+      "epoch": 4.771673260880453,
+      "grad_norm": 1.9786897897720337,
+      "learning_rate": 2.5816893768535744e-07,
+      "loss": 0.4211,
+      "step": 47640
+    },
+    {
+      "epoch": 4.772674913607452,
+      "grad_norm": 2.5731799602508545,
+      "learning_rate": 2.5591411414719046e-07,
+      "loss": 0.3395,
+      "step": 47650
+    },
+    {
+      "epoch": 4.773676566334451,
+      "grad_norm": 2.1998531818389893,
+      "learning_rate": 2.5366912989761573e-07,
+      "loss": 0.412,
+      "step": 47660
+    },
+    {
+      "epoch": 4.774678219061451,
+      "grad_norm": 2.2904112339019775,
+      "learning_rate": 2.5143398582933575e-07,
+      "loss": 0.3507,
+      "step": 47670
+    },
+    {
+      "epoch": 4.775679871788451,
+      "grad_norm": 2.1148123741149902,
+      "learning_rate": 2.4920868283114243e-07,
+      "loss": 0.4345,
+      "step": 47680
+    },
+    {
+      "epoch": 4.776681524515451,
+      "grad_norm": 2.2035341262817383,
+      "learning_rate": 2.4699322178790285e-07,
+      "loss": 0.3879,
+      "step": 47690
+    },
+    {
+      "epoch": 4.77768317724245,
+      "grad_norm": 2.235149621963501,
+      "learning_rate": 2.4478760358057904e-07,
+      "loss": 0.3872,
+      "step": 47700
+    },
+    {
+      "epoch": 4.77868482996945,
+      "grad_norm": 2.893347978591919,
+      "learning_rate": 2.4259182908622226e-07,
+      "loss": 0.4255,
+      "step": 47710
+    },
+    {
+      "epoch": 4.779686482696449,
+      "grad_norm": 2.0718188285827637,
+      "learning_rate": 2.4040589917796175e-07,
+      "loss": 0.4655,
+      "step": 47720
+    },
+    {
+      "epoch": 4.780688135423449,
+      "grad_norm": 2.6403911113739014,
+      "learning_rate": 2.382298147250106e-07,
+      "loss": 0.3789,
+      "step": 47730
+    },
+    {
+      "epoch": 4.781689788150448,
+      "grad_norm": 2.3864388465881348,
+      "learning_rate": 2.3606357659267942e-07,
+      "loss": 0.373,
+      "step": 47740
+    },
+    {
+      "epoch": 4.782691440877448,
+      "grad_norm": 2.413670539855957,
+      "learning_rate": 2.3390718564235137e-07,
+      "loss": 0.4291,
+      "step": 47750
+    },
+    {
+      "epoch": 4.783693093604447,
+      "grad_norm": 1.797200083732605,
+      "learning_rate": 2.3176064273149612e-07,
+      "loss": 0.3736,
+      "step": 47760
+    },
+    {
+      "epoch": 4.784694746331446,
+      "grad_norm": 2.2276856899261475,
+      "learning_rate": 2.2962394871367533e-07,
+      "loss": 0.4219,
+      "step": 47770
+    },
+    {
+      "epoch": 4.7856963990584465,
+      "grad_norm": 2.2116262912750244,
+      "learning_rate": 2.2749710443852047e-07,
+      "loss": 0.3695,
+      "step": 47780
+    },
+    {
+      "epoch": 4.786698051785446,
+      "grad_norm": 2.3747339248657227,
+      "learning_rate": 2.2538011075176059e-07,
+      "loss": 0.4255,
+      "step": 47790
+    },
+    {
+      "epoch": 4.787699704512446,
+      "grad_norm": 2.656726360321045,
+      "learning_rate": 2.2327296849520008e-07,
+      "loss": 0.3903,
+      "step": 47800
+    },
+    {
+      "epoch": 4.788701357239445,
+      "grad_norm": 2.5891129970550537,
+      "learning_rate": 2.2117567850672705e-07,
+      "loss": 0.4349,
+      "step": 47810
+    },
+    {
+      "epoch": 4.789703009966445,
+      "grad_norm": 2.4755101203918457,
+      "learning_rate": 2.190882416203105e-07,
+      "loss": 0.3904,
+      "step": 47820
+    },
+    {
+      "epoch": 4.790704662693444,
+      "grad_norm": 2.4286439418792725,
+      "learning_rate": 2.1701065866600312e-07,
+      "loss": 0.3761,
+      "step": 47830
+    },
+    {
+      "epoch": 4.791706315420444,
+      "grad_norm": 2.4156734943389893,
+      "learning_rate": 2.149429304699413e-07,
+      "loss": 0.4208,
+      "step": 47840
+    },
+    {
+      "epoch": 4.792707968147443,
+      "grad_norm": 2.1973776817321777,
+      "learning_rate": 2.1288505785433954e-07,
+      "loss": 0.3716,
+      "step": 47850
+    },
+    {
+      "epoch": 4.793709620874443,
+      "grad_norm": 2.4909114837646484,
+      "learning_rate": 2.108370416374933e-07,
+      "loss": 0.3857,
+      "step": 47860
+    },
+    {
+      "epoch": 4.794711273601442,
+      "grad_norm": 2.1599183082580566,
+      "learning_rate": 2.0879888263378167e-07,
+      "loss": 0.4253,
+      "step": 47870
+    },
+    {
+      "epoch": 4.7957129263284415,
+      "grad_norm": 1.7978217601776123,
+      "learning_rate": 2.067705816536647e-07,
+      "loss": 0.367,
+      "step": 47880
+    },
+    {
+      "epoch": 4.7967145790554415,
+      "grad_norm": 2.390063524246216,
+      "learning_rate": 2.0475213950367221e-07,
+      "loss": 0.4664,
+      "step": 47890
+    },
+    {
+      "epoch": 4.797716231782441,
+      "grad_norm": 2.1187353134155273,
+      "learning_rate": 2.0274355698643166e-07,
+      "loss": 0.4457,
+      "step": 47900
+    },
+    {
+      "epoch": 4.798717884509441,
+      "grad_norm": 2.258744239807129,
+      "learning_rate": 2.007448349006319e-07,
+      "loss": 0.4192,
+      "step": 47910
+    },
+    {
+      "epoch": 4.79971953723644,
+      "grad_norm": 2.4565742015838623,
+      "learning_rate": 1.9875597404105383e-07,
+      "loss": 0.4753,
+      "step": 47920
+    },
+    {
+      "epoch": 4.80072118996344,
+      "grad_norm": 2.1519105434417725,
+      "learning_rate": 1.967769751985482e-07,
+      "loss": 0.4224,
+      "step": 47930
+    },
+    {
+      "epoch": 4.801722842690439,
+      "grad_norm": 1.9558358192443848,
+      "learning_rate": 1.9480783916005217e-07,
+      "loss": 0.3739,
+      "step": 47940
+    },
+    {
+      "epoch": 4.802724495417439,
+      "grad_norm": 2.899057388305664,
+      "learning_rate": 1.9284856670857276e-07,
+      "loss": 0.4697,
+      "step": 47950
+    },
+    {
+      "epoch": 4.803726148144438,
+      "grad_norm": 2.3547844886779785,
+      "learning_rate": 1.908991586232006e-07,
+      "loss": 0.3781,
+      "step": 47960
+    },
+    {
+      "epoch": 4.804727800871438,
+      "grad_norm": 2.6609480381011963,
+      "learning_rate": 1.8895961567910737e-07,
+      "loss": 0.388,
+      "step": 47970
+    },
+    {
+      "epoch": 4.805729453598437,
+      "grad_norm": 2.470571994781494,
+      "learning_rate": 1.8702993864752883e-07,
+      "loss": 0.4113,
+      "step": 47980
+    },
+    {
+      "epoch": 4.8067311063254365,
+      "grad_norm": 2.2260477542877197,
+      "learning_rate": 1.8511012829578733e-07,
+      "loss": 0.3831,
+      "step": 47990
+    },
+    {
+      "epoch": 4.807732759052437,
+      "grad_norm": 1.9835573434829712,
+      "learning_rate": 1.832001853872861e-07,
+      "loss": 0.4476,
+      "step": 48000
+    },
+    {
+      "epoch": 4.808734411779436,
+      "grad_norm": 2.08286190032959,
+      "learning_rate": 1.8130011068149266e-07,
+      "loss": 0.3969,
+      "step": 48010
+    },
+    {
+      "epoch": 4.809736064506436,
+      "grad_norm": 2.29129958152771,
+      "learning_rate": 1.7940990493395815e-07,
+      "loss": 0.4017,
+      "step": 48020
+    },
+    {
+      "epoch": 4.810737717233435,
+      "grad_norm": 2.6982805728912354,
+      "learning_rate": 1.775295688963091e-07,
+      "loss": 0.4785,
+      "step": 48030
+    },
+    {
+      "epoch": 4.811739369960435,
+      "grad_norm": 2.0841140747070312,
+      "learning_rate": 1.7565910331624468e-07,
+      "loss": 0.4117,
+      "step": 48040
+    },
+    {
+      "epoch": 4.812741022687434,
+      "grad_norm": 2.39800763130188,
+      "learning_rate": 1.7379850893754212e-07,
+      "loss": 0.4054,
+      "step": 48050
+    },
+    {
+      "epoch": 4.813742675414434,
+      "grad_norm": 1.6849093437194824,
+      "learning_rate": 1.719477865000513e-07,
+      "loss": 0.4053,
+      "step": 48060
+    },
+    {
+      "epoch": 4.814744328141433,
+      "grad_norm": 2.5003740787506104,
+      "learning_rate": 1.7010693673969736e-07,
+      "loss": 0.3939,
+      "step": 48070
+    },
+    {
+      "epoch": 4.815745980868433,
+      "grad_norm": 2.4821770191192627,
+      "learning_rate": 1.6827596038848092e-07,
+      "loss": 0.4269,
+      "step": 48080
+    },
+    {
+      "epoch": 4.8167476335954325,
+      "grad_norm": 1.7561688423156738,
+      "learning_rate": 1.664548581744696e-07,
+      "loss": 0.3697,
+      "step": 48090
+    },
+    {
+      "epoch": 4.817749286322432,
+      "grad_norm": 2.127955436706543,
+      "learning_rate": 1.6464363082181467e-07,
+      "loss": 0.3987,
+      "step": 48100
+    },
+    {
+      "epoch": 4.818750939049432,
+      "grad_norm": 1.8868879079818726,
+      "learning_rate": 1.6284227905073722e-07,
+      "loss": 0.4484,
+      "step": 48110
+    },
+    {
+      "epoch": 4.819752591776431,
+      "grad_norm": 1.6073931455612183,
+      "learning_rate": 1.6105080357752822e-07,
+      "loss": 0.3723,
+      "step": 48120
+    },
+    {
+      "epoch": 4.820754244503431,
+      "grad_norm": 2.3502357006073,
+      "learning_rate": 1.592692051145539e-07,
+      "loss": 0.3868,
+      "step": 48130
+    },
+    {
+      "epoch": 4.82175589723043,
+      "grad_norm": 2.4274137020111084,
+      "learning_rate": 1.5749748437025314e-07,
+      "loss": 0.423,
+      "step": 48140
+    },
+    {
+      "epoch": 4.82275754995743,
+      "grad_norm": 2.140169143676758,
+      "learning_rate": 1.557356420491346e-07,
+      "loss": 0.4156,
+      "step": 48150
+    },
+    {
+      "epoch": 4.823759202684429,
+      "grad_norm": 2.5988030433654785,
+      "learning_rate": 1.539836788517851e-07,
+      "loss": 0.4077,
+      "step": 48160
+    },
+    {
+      "epoch": 4.824760855411429,
+      "grad_norm": 2.1025750637054443,
+      "learning_rate": 1.5224159547485573e-07,
+      "loss": 0.4326,
+      "step": 48170
+    },
+    {
+      "epoch": 4.825762508138428,
+      "grad_norm": 1.7013978958129883,
+      "learning_rate": 1.505093926110729e-07,
+      "loss": 0.396,
+      "step": 48180
+    },
+    {
+      "epoch": 4.826764160865428,
+      "grad_norm": 2.384352207183838,
+      "learning_rate": 1.4878707094923283e-07,
+      "loss": 0.3865,
+      "step": 48190
+    },
+    {
+      "epoch": 4.8277658135924275,
+      "grad_norm": 2.697436571121216,
+      "learning_rate": 1.470746311742044e-07,
+      "loss": 0.3775,
+      "step": 48200
+    },
+    {
+      "epoch": 4.828767466319427,
+      "grad_norm": 2.4065449237823486,
+      "learning_rate": 1.4537207396692343e-07,
+      "loss": 0.4029,
+      "step": 48210
+    },
+    {
+      "epoch": 4.829769119046427,
+      "grad_norm": 1.876576542854309,
+      "learning_rate": 1.436794000043984e-07,
+      "loss": 0.4067,
+      "step": 48220
+    },
+    {
+      "epoch": 4.830770771773426,
+      "grad_norm": 2.1361072063446045,
+      "learning_rate": 1.419966099597103e-07,
+      "loss": 0.3742,
+      "step": 48230
+    },
+    {
+      "epoch": 4.831772424500426,
+      "grad_norm": 2.7625808715820312,
+      "learning_rate": 1.4032370450200728e-07,
+      "loss": 0.4557,
+      "step": 48240
+    },
+    {
+      "epoch": 4.832774077227425,
+      "grad_norm": 2.028794527053833,
+      "learning_rate": 1.386606842965016e-07,
+      "loss": 0.4347,
+      "step": 48250
+    },
+    {
+      "epoch": 4.833775729954425,
+      "grad_norm": 2.240504026412964,
+      "learning_rate": 1.3700755000448373e-07,
+      "loss": 0.4257,
+      "step": 48260
+    },
+    {
+      "epoch": 4.834777382681424,
+      "grad_norm": 2.295073986053467,
+      "learning_rate": 1.3536430228331122e-07,
+      "loss": 0.3854,
+      "step": 48270
+    },
+    {
+      "epoch": 4.835779035408424,
+      "grad_norm": 2.3853859901428223,
+      "learning_rate": 1.3373094178640576e-07,
+      "loss": 0.3957,
+      "step": 48280
+    },
+    {
+      "epoch": 4.836780688135423,
+      "grad_norm": 2.4576685428619385,
+      "learning_rate": 1.32107469163259e-07,
+      "loss": 0.3913,
+      "step": 48290
+    },
+    {
+      "epoch": 4.837782340862423,
+      "grad_norm": 1.7085213661193848,
+      "learning_rate": 1.3049388505943504e-07,
+      "loss": 0.3868,
+      "step": 48300
+    },
+    {
+      "epoch": 4.838783993589423,
+      "grad_norm": 2.529207706451416,
+      "learning_rate": 1.2889019011655955e-07,
+      "loss": 0.4383,
+      "step": 48310
+    },
+    {
+      "epoch": 4.839785646316422,
+      "grad_norm": 2.305237054824829,
+      "learning_rate": 1.2729638497233077e-07,
+      "loss": 0.4403,
+      "step": 48320
+    },
+    {
+      "epoch": 4.840787299043422,
+      "grad_norm": 1.8845841884613037,
+      "learning_rate": 1.2571247026051392e-07,
+      "loss": 0.3387,
+      "step": 48330
+    },
+    {
+      "epoch": 4.841788951770421,
+      "grad_norm": 2.760739803314209,
+      "learning_rate": 1.241384466109413e-07,
+      "loss": 0.4136,
+      "step": 48340
+    },
+    {
+      "epoch": 4.842790604497421,
+      "grad_norm": 2.0073342323303223,
+      "learning_rate": 1.2257431464950396e-07,
+      "loss": 0.4204,
+      "step": 48350
+    },
+    {
+      "epoch": 4.84379225722442,
+      "grad_norm": 2.3097691535949707,
+      "learning_rate": 1.2102007499817103e-07,
+      "loss": 0.3863,
+      "step": 48360
+    },
+    {
+      "epoch": 4.84479390995142,
+      "grad_norm": 2.4525158405303955,
+      "learning_rate": 1.1947572827497588e-07,
+      "loss": 0.425,
+      "step": 48370
+    },
+    {
+      "epoch": 4.845795562678419,
+      "grad_norm": 1.7297872304916382,
+      "learning_rate": 1.1794127509401065e-07,
+      "loss": 0.3882,
+      "step": 48380
+    },
+    {
+      "epoch": 4.846797215405419,
+      "grad_norm": 2.2076919078826904,
+      "learning_rate": 1.164167160654428e-07,
+      "loss": 0.4148,
+      "step": 48390
+    },
+    {
+      "epoch": 4.8477988681324184,
+      "grad_norm": 2.070524215698242,
+      "learning_rate": 1.1490205179549851e-07,
+      "loss": 0.3745,
+      "step": 48400
+    },
+    {
+      "epoch": 4.8488005208594185,
+      "grad_norm": 2.7982497215270996,
+      "learning_rate": 1.1339728288647378e-07,
+      "loss": 0.4531,
+      "step": 48410
+    },
+    {
+      "epoch": 4.849802173586418,
+      "grad_norm": 1.602648377418518,
+      "learning_rate": 1.1190240993672607e-07,
+      "loss": 0.3816,
+      "step": 48420
+    },
+    {
+      "epoch": 4.850803826313417,
+      "grad_norm": 1.842578411102295,
+      "learning_rate": 1.1041743354067991e-07,
+      "loss": 0.3725,
+      "step": 48430
+    },
+    {
+      "epoch": 4.851805479040417,
+      "grad_norm": 2.1316118240356445,
+      "learning_rate": 1.0894235428882682e-07,
+      "loss": 0.4026,
+      "step": 48440
+    },
+    {
+      "epoch": 4.852807131767416,
+      "grad_norm": 2.16373348236084,
+      "learning_rate": 1.0747717276771707e-07,
+      "loss": 0.3965,
+      "step": 48450
+    },
+    {
+      "epoch": 4.853808784494416,
+      "grad_norm": 2.0309717655181885,
+      "learning_rate": 1.0602188955996795e-07,
+      "loss": 0.4534,
+      "step": 48460
+    },
+    {
+      "epoch": 4.854810437221415,
+      "grad_norm": 1.7845313549041748,
+      "learning_rate": 1.0457650524426654e-07,
+      "loss": 0.4438,
+      "step": 48470
+    },
+    {
+      "epoch": 4.855812089948415,
+      "grad_norm": 1.8658897876739502,
+      "learning_rate": 1.0314102039535312e-07,
+      "loss": 0.3743,
+      "step": 48480
+    },
+    {
+      "epoch": 4.856813742675414,
+      "grad_norm": 2.0582423210144043,
+      "learning_rate": 1.0171543558403774e-07,
+      "loss": 0.4112,
+      "step": 48490
+    },
+    {
+      "epoch": 4.857815395402414,
+      "grad_norm": 2.1272170543670654,
+      "learning_rate": 1.0029975137719472e-07,
+      "loss": 0.3999,
+      "step": 48500
+    },
+    {
+      "epoch": 4.8588170481294135,
+      "grad_norm": 2.131007432937622,
+      "learning_rate": 9.88939683377571e-08,
+      "loss": 0.384,
+      "step": 48510
+    },
+    {
+      "epoch": 4.8598187008564135,
+      "grad_norm": 2.36995005607605,
+      "learning_rate": 9.749808702472774e-08,
+      "loss": 0.3686,
+      "step": 48520
+    },
+    {
+      "epoch": 4.860820353583413,
+      "grad_norm": 2.426204204559326,
+      "learning_rate": 9.611210799316262e-08,
+      "loss": 0.4627,
+      "step": 48530
+    },
+    {
+      "epoch": 4.861822006310412,
+      "grad_norm": 1.8704348802566528,
+      "learning_rate": 9.473603179418756e-08,
+      "loss": 0.4085,
+      "step": 48540
+    },
+    {
+      "epoch": 4.862823659037412,
+      "grad_norm": 2.140582799911499,
+      "learning_rate": 9.336985897498706e-08,
+      "loss": 0.3729,
+      "step": 48550
+    },
+    {
+      "epoch": 4.863825311764411,
+      "grad_norm": 1.9879549741744995,
+      "learning_rate": 9.201359007881271e-08,
+      "loss": 0.4406,
+      "step": 48560
+    },
+    {
+      "epoch": 4.864826964491411,
+      "grad_norm": 2.478294610977173,
+      "learning_rate": 9.066722564496921e-08,
+      "loss": 0.3528,
+      "step": 48570
+    },
+    {
+      "epoch": 4.86582861721841,
+      "grad_norm": 2.1455304622650146,
+      "learning_rate": 8.93307662088283e-08,
+      "loss": 0.3928,
+      "step": 48580
+    },
+    {
+      "epoch": 4.86683026994541,
+      "grad_norm": 2.190230369567871,
+      "learning_rate": 8.800421230182599e-08,
+      "loss": 0.3745,
+      "step": 48590
+    },
+    {
+      "epoch": 4.867831922672409,
+      "grad_norm": 2.3964619636535645,
+      "learning_rate": 8.668756445145421e-08,
+      "loss": 0.4348,
+      "step": 48600
+    },
+    {
+      "epoch": 4.8688335753994085,
+      "grad_norm": 2.1755192279815674,
+      "learning_rate": 8.53808231812664e-08,
+      "loss": 0.3612,
+      "step": 48610
+    },
+    {
+      "epoch": 4.869835228126409,
+      "grad_norm": 2.1091182231903076,
+      "learning_rate": 8.408398901087466e-08,
+      "loss": 0.4825,
+      "step": 48620
+    },
+    {
+      "epoch": 4.870836880853409,
+      "grad_norm": 2.6433656215667725,
+      "learning_rate": 8.279706245596375e-08,
+      "loss": 0.4309,
+      "step": 48630
+    },
+    {
+      "epoch": 4.871838533580408,
+      "grad_norm": 2.1513772010803223,
+      "learning_rate": 8.152004402826319e-08,
+      "loss": 0.3659,
+      "step": 48640
+    },
+    {
+      "epoch": 4.872840186307407,
+      "grad_norm": 2.3869216442108154,
+      "learning_rate": 8.025293423556956e-08,
+      "loss": 0.4321,
+      "step": 48650
+    },
+    {
+      "epoch": 4.873841839034407,
+      "grad_norm": 2.6791770458221436,
+      "learning_rate": 7.899573358174094e-08,
+      "loss": 0.4111,
+      "step": 48660
+    },
+    {
+      "epoch": 4.874843491761406,
+      "grad_norm": 2.2372169494628906,
+      "learning_rate": 7.774844256669134e-08,
+      "loss": 0.4593,
+      "step": 48670
+    },
+    {
+      "epoch": 4.875845144488406,
+      "grad_norm": 2.1360697746276855,
+      "learning_rate": 7.651106168639344e-08,
+      "loss": 0.443,
+      "step": 48680
+    },
+    {
+      "epoch": 4.876846797215405,
+      "grad_norm": 2.072798490524292,
+      "learning_rate": 7.528359143288977e-08,
+      "loss": 0.3797,
+      "step": 48690
+    },
+    {
+      "epoch": 4.877848449942405,
+      "grad_norm": 2.050501823425293,
+      "learning_rate": 7.406603229427045e-08,
+      "loss": 0.3611,
+      "step": 48700
+    },
+    {
+      "epoch": 4.878850102669404,
+      "grad_norm": 2.0490472316741943,
+      "learning_rate": 7.285838475468431e-08,
+      "loss": 0.4052,
+      "step": 48710
+    },
+    {
+      "epoch": 4.879851755396404,
+      "grad_norm": 2.0858280658721924,
+      "learning_rate": 7.166064929434446e-08,
+      "loss": 0.4111,
+      "step": 48720
+    },
+    {
+      "epoch": 4.880853408123404,
+      "grad_norm": 1.834701657295227,
+      "learning_rate": 7.047282638952545e-08,
+      "loss": 0.4064,
+      "step": 48730
+    },
+    {
+      "epoch": 4.881855060850404,
+      "grad_norm": 1.9594968557357788,
+      "learning_rate": 6.929491651255226e-08,
+      "loss": 0.4009,
+      "step": 48740
+    },
+    {
+      "epoch": 4.882856713577403,
+      "grad_norm": 1.9518842697143555,
+      "learning_rate": 6.812692013181132e-08,
+      "loss": 0.3914,
+      "step": 48750
+    },
+    {
+      "epoch": 4.883858366304402,
+      "grad_norm": 3.005319118499756,
+      "learning_rate": 6.6968837711745e-08,
+      "loss": 0.4469,
+      "step": 48760
+    },
+    {
+      "epoch": 4.884860019031402,
+      "grad_norm": 2.5086545944213867,
+      "learning_rate": 6.582066971285995e-08,
+      "loss": 0.4403,
+      "step": 48770
+    },
+    {
+      "epoch": 4.885861671758401,
+      "grad_norm": 2.4275851249694824,
+      "learning_rate": 6.468241659171315e-08,
+      "loss": 0.4217,
+      "step": 48780
+    },
+    {
+      "epoch": 4.886863324485401,
+      "grad_norm": 2.5610122680664062,
+      "learning_rate": 6.355407880092313e-08,
+      "loss": 0.4648,
+      "step": 48790
+    },
+    {
+      "epoch": 4.8878649772124,
+      "grad_norm": 1.8257973194122314,
+      "learning_rate": 6.24356567891643e-08,
+      "loss": 0.4074,
+      "step": 48800
+    },
+    {
+      "epoch": 4.8888666299394,
+      "grad_norm": 1.4888192415237427,
+      "learning_rate": 6.132715100116704e-08,
+      "loss": 0.4077,
+      "step": 48810
+    },
+    {
+      "epoch": 4.8898682826663995,
+      "grad_norm": 2.0692968368530273,
+      "learning_rate": 6.022856187772041e-08,
+      "loss": 0.4498,
+      "step": 48820
+    },
+    {
+      "epoch": 4.890869935393399,
+      "grad_norm": 1.9948792457580566,
+      "learning_rate": 5.913988985566943e-08,
+      "loss": 0.3604,
+      "step": 48830
+    },
+    {
+      "epoch": 4.891871588120399,
+      "grad_norm": 2.2797932624816895,
+      "learning_rate": 5.806113536791779e-08,
+      "loss": 0.4214,
+      "step": 48840
+    },
+    {
+      "epoch": 4.892873240847398,
+      "grad_norm": 1.757158637046814,
+      "learning_rate": 5.699229884341961e-08,
+      "loss": 0.4587,
+      "step": 48850
+    },
+    {
+      "epoch": 4.893874893574398,
+      "grad_norm": 2.277294874191284,
+      "learning_rate": 5.593338070719323e-08,
+      "loss": 0.3885,
+      "step": 48860
+    },
+    {
+      "epoch": 4.894876546301397,
+      "grad_norm": 2.3876211643218994,
+      "learning_rate": 5.488438138030738e-08,
+      "loss": 0.365,
+      "step": 48870
+    },
+    {
+      "epoch": 4.895878199028397,
+      "grad_norm": 2.418525457382202,
+      "learning_rate": 5.3845301279889514e-08,
+      "loss": 0.3895,
+      "step": 48880
+    },
+    {
+      "epoch": 4.896879851755396,
+      "grad_norm": 2.0426852703094482,
+      "learning_rate": 5.2816140819120233e-08,
+      "loss": 0.4293,
+      "step": 48890
+    },
+    {
+      "epoch": 4.897881504482396,
+      "grad_norm": 2.254586696624756,
+      "learning_rate": 5.179690040723606e-08,
+      "loss": 0.3793,
+      "step": 48900
+    },
+    {
+      "epoch": 4.898883157209395,
+      "grad_norm": 2.491412878036499,
+      "learning_rate": 5.078758044952947e-08,
+      "loss": 0.3904,
+      "step": 48910
+    },
+    {
+      "epoch": 4.899884809936395,
+      "grad_norm": 2.3205726146698,
+      "learning_rate": 4.978818134735164e-08,
+      "loss": 0.3826,
+      "step": 48920
+    },
+    {
+      "epoch": 4.900886462663395,
+      "grad_norm": 2.237753391265869,
+      "learning_rate": 4.879870349810689e-08,
+      "loss": 0.3971,
+      "step": 48930
+    },
+    {
+      "epoch": 4.901888115390394,
+      "grad_norm": 1.7564687728881836,
+      "learning_rate": 4.781914729524717e-08,
+      "loss": 0.4064,
+      "step": 48940
+    },
+    {
+      "epoch": 4.902889768117394,
+      "grad_norm": 2.1869213581085205,
+      "learning_rate": 4.684951312828867e-08,
+      "loss": 0.4179,
+      "step": 48950
+    },
+    {
+      "epoch": 4.903891420844393,
+      "grad_norm": 1.9542244672775269,
+      "learning_rate": 4.588980138279797e-08,
+      "loss": 0.4201,
+      "step": 48960
+    },
+    {
+      "epoch": 4.904893073571393,
+      "grad_norm": 2.3261213302612305,
+      "learning_rate": 4.4940012440397583e-08,
+      "loss": 0.3918,
+      "step": 48970
+    },
+    {
+      "epoch": 4.905894726298392,
+      "grad_norm": 2.3012595176696777,
+      "learning_rate": 4.400014667876318e-08,
+      "loss": 0.4794,
+      "step": 48980
+    },
+    {
+      "epoch": 4.906896379025392,
+      "grad_norm": 2.335749387741089,
+      "learning_rate": 4.307020447162635e-08,
+      "loss": 0.3636,
+      "step": 48990
+    },
+    {
+      "epoch": 4.907898031752391,
+      "grad_norm": 2.5763540267944336,
+      "learning_rate": 4.215018618876632e-08,
+      "loss": 0.388,
+      "step": 49000
+    },
+    {
+      "epoch": 4.908899684479391,
+      "grad_norm": 2.0762269496917725,
+      "learning_rate": 4.124009219602654e-08,
+      "loss": 0.4073,
+      "step": 49010
+    },
+    {
+      "epoch": 4.90990133720639,
+      "grad_norm": 2.1076955795288086,
+      "learning_rate": 4.033992285529809e-08,
+      "loss": 0.4247,
+      "step": 49020
+    },
+    {
+      "epoch": 4.9109029899333905,
+      "grad_norm": 2.1044840812683105,
+      "learning_rate": 3.9449678524522415e-08,
+      "loss": 0.449,
+      "step": 49030
+    },
+    {
+      "epoch": 4.91190464266039,
+      "grad_norm": 2.3766860961914062,
+      "learning_rate": 3.856935955769969e-08,
+      "loss": 0.4182,
+      "step": 49040
+    },
+    {
+      "epoch": 4.912906295387389,
+      "grad_norm": 2.158463716506958,
+      "learning_rate": 3.769896630488323e-08,
+      "loss": 0.4147,
+      "step": 49050
+    },
+    {
+      "epoch": 4.913907948114389,
+      "grad_norm": 2.246264696121216,
+      "learning_rate": 3.683849911217674e-08,
+      "loss": 0.4409,
+      "step": 49060
+    },
+    {
+      "epoch": 4.914909600841388,
+      "grad_norm": 1.723870873451233,
+      "learning_rate": 3.5987958321739844e-08,
+      "loss": 0.4226,
+      "step": 49070
+    },
+    {
+      "epoch": 4.915911253568388,
+      "grad_norm": 2.071007490158081,
+      "learning_rate": 3.514734427177979e-08,
+      "loss": 0.5269,
+      "step": 49080
+    },
+    {
+      "epoch": 4.916912906295387,
+      "grad_norm": 2.3110177516937256,
+      "learning_rate": 3.431665729656253e-08,
+      "loss": 0.3906,
+      "step": 49090
+    },
+    {
+      "epoch": 4.917914559022387,
+      "grad_norm": 1.8781073093414307,
+      "learning_rate": 3.349589772640715e-08,
+      "loss": 0.3827,
+      "step": 49100
+    },
+    {
+      "epoch": 4.918916211749386,
+      "grad_norm": 2.104565143585205,
+      "learning_rate": 3.2685065887674834e-08,
+      "loss": 0.4316,
+      "step": 49110
+    },
+    {
+      "epoch": 4.919917864476386,
+      "grad_norm": 2.219674587249756,
+      "learning_rate": 3.188416210279099e-08,
+      "loss": 0.363,
+      "step": 49120
+    },
+    {
+      "epoch": 4.9209195172033855,
+      "grad_norm": 2.114689588546753,
+      "learning_rate": 3.1093186690228645e-08,
+      "loss": 0.4162,
+      "step": 49130
+    },
+    {
+      "epoch": 4.9219211699303855,
+      "grad_norm": 2.0361156463623047,
+      "learning_rate": 3.031213996451121e-08,
+      "loss": 0.4122,
+      "step": 49140
+    },
+    {
+      "epoch": 4.922922822657385,
+      "grad_norm": 1.689947247505188,
+      "learning_rate": 2.954102223621802e-08,
+      "loss": 0.4286,
+      "step": 49150
+    },
+    {
+      "epoch": 4.923924475384384,
+      "grad_norm": 1.9771041870117188,
+      "learning_rate": 2.877983381197602e-08,
+      "loss": 0.3823,
+      "step": 49160
+    },
+    {
+      "epoch": 4.924926128111384,
+      "grad_norm": 2.4345974922180176,
+      "learning_rate": 2.8028574994465317e-08,
+      "loss": 0.4479,
+      "step": 49170
+    },
+    {
+      "epoch": 4.925927780838383,
+      "grad_norm": 2.2519659996032715,
+      "learning_rate": 2.728724608241917e-08,
+      "loss": 0.3989,
+      "step": 49180
+    },
+    {
+      "epoch": 4.926929433565383,
+      "grad_norm": 2.584491729736328,
+      "learning_rate": 2.6555847370621222e-08,
+      "loss": 0.441,
+      "step": 49190
+    },
+    {
+      "epoch": 4.927931086292382,
+      "grad_norm": 2.2947561740875244,
+      "learning_rate": 2.5834379149905495e-08,
+      "loss": 0.4375,
+      "step": 49200
+    },
+    {
+      "epoch": 4.928932739019382,
+      "grad_norm": 2.1762804985046387,
+      "learning_rate": 2.5122841707159172e-08,
+      "loss": 0.3976,
+      "step": 49210
+    },
+    {
+      "epoch": 4.929934391746381,
+      "grad_norm": 2.127589225769043,
+      "learning_rate": 2.4421235325319815e-08,
+      "loss": 0.3811,
+      "step": 49220
+    },
+    {
+      "epoch": 4.930936044473381,
+      "grad_norm": 2.0069942474365234,
+      "learning_rate": 2.3729560283372586e-08,
+      "loss": 0.4151,
+      "step": 49230
+    },
+    {
+      "epoch": 4.9319376972003806,
+      "grad_norm": 1.6562203168869019,
+      "learning_rate": 2.3047816856358595e-08,
+      "loss": 0.3697,
+      "step": 49240
+    },
+    {
+      "epoch": 4.932939349927381,
+      "grad_norm": 2.022047996520996,
+      "learning_rate": 2.2376005315369318e-08,
+      "loss": 0.4432,
+      "step": 49250
+    },
+    {
+      "epoch": 4.93394100265438,
+      "grad_norm": 2.255925416946411,
+      "learning_rate": 2.1714125927543848e-08,
+      "loss": 0.3806,
+      "step": 49260
+    },
+    {
+      "epoch": 4.934942655381379,
+      "grad_norm": 1.9878329038619995,
+      "learning_rate": 2.1062178956071655e-08,
+      "loss": 0.3847,
+      "step": 49270
+    },
+    {
+      "epoch": 4.935944308108379,
+      "grad_norm": 1.935551404953003,
+      "learning_rate": 2.0420164660195362e-08,
+      "loss": 0.4273,
+      "step": 49280
+    },
+    {
+      "epoch": 4.936945960835378,
+      "grad_norm": 2.561936378479004,
+      "learning_rate": 1.978808329520798e-08,
+      "loss": 0.4051,
+      "step": 49290
+    },
+    {
+      "epoch": 4.937947613562378,
+      "grad_norm": 2.0825181007385254,
+      "learning_rate": 1.916593511245013e-08,
+      "loss": 0.4425,
+      "step": 49300
+    },
+    {
+      "epoch": 4.938949266289377,
+      "grad_norm": 1.867803692817688,
+      "learning_rate": 1.8553720359315574e-08,
+      "loss": 0.442,
+      "step": 49310
+    },
+    {
+      "epoch": 4.939950919016377,
+      "grad_norm": 2.4296388626098633,
+      "learning_rate": 1.7951439279245697e-08,
+      "loss": 0.4022,
+      "step": 49320
+    },
+    {
+      "epoch": 4.940952571743376,
+      "grad_norm": 1.7822569608688354,
+      "learning_rate": 1.7359092111732255e-08,
+      "loss": 0.42,
+      "step": 49330
+    },
+    {
+      "epoch": 4.9419542244703765,
+      "grad_norm": 2.034141778945923,
+      "learning_rate": 1.6776679092320168e-08,
+      "loss": 0.4095,
+      "step": 49340
+    },
+    {
+      "epoch": 4.942955877197376,
+      "grad_norm": 2.480480194091797,
+      "learning_rate": 1.6204200452596407e-08,
+      "loss": 0.4123,
+      "step": 49350
+    },
+    {
+      "epoch": 4.943957529924376,
+      "grad_norm": 1.9543697834014893,
+      "learning_rate": 1.564165642020665e-08,
+      "loss": 0.4178,
+      "step": 49360
+    },
+    {
+      "epoch": 4.944959182651375,
+      "grad_norm": 2.275780439376831,
+      "learning_rate": 1.5089047218838637e-08,
+      "loss": 0.4003,
+      "step": 49370
+    },
+    {
+      "epoch": 4.945960835378374,
+      "grad_norm": 2.82968807220459,
+      "learning_rate": 1.454637306823603e-08,
+      "loss": 0.4349,
+      "step": 49380
+    },
+    {
+      "epoch": 4.946962488105374,
+      "grad_norm": 1.861443042755127,
+      "learning_rate": 1.4013634184190105e-08,
+      "loss": 0.3798,
+      "step": 49390
+    },
+    {
+      "epoch": 4.947964140832373,
+      "grad_norm": 2.3817360401153564,
+      "learning_rate": 1.3490830778534192e-08,
+      "loss": 0.3854,
+      "step": 49400
+    },
+    {
+      "epoch": 4.948965793559373,
+      "grad_norm": 2.194943904876709,
+      "learning_rate": 1.2977963059163101e-08,
+      "loss": 0.422,
+      "step": 49410
+    },
+    {
+      "epoch": 4.949967446286372,
+      "grad_norm": 2.164625644683838,
+      "learning_rate": 1.24750312300137e-08,
+      "loss": 0.399,
+      "step": 49420
+    },
+    {
+      "epoch": 4.950969099013372,
+      "grad_norm": 2.4973015785217285,
+      "learning_rate": 1.198203549106769e-08,
+      "loss": 0.4427,
+      "step": 49430
+    },
+    {
+      "epoch": 4.9519707517403715,
+      "grad_norm": 1.9972591400146484,
+      "learning_rate": 1.1498976038368248e-08,
+      "loss": 0.3763,
+      "step": 49440
+    },
+    {
+      "epoch": 4.9529724044673715,
+      "grad_norm": 2.178032875061035,
+      "learning_rate": 1.1025853063992287e-08,
+      "loss": 0.4024,
+      "step": 49450
+    },
+    {
+      "epoch": 4.953974057194371,
+      "grad_norm": 1.7938523292541504,
+      "learning_rate": 1.0562666756080974e-08,
+      "loss": 0.3719,
+      "step": 49460
+    },
+    {
+      "epoch": 4.954975709921371,
+      "grad_norm": 1.7838163375854492,
+      "learning_rate": 1.0109417298811985e-08,
+      "loss": 0.3619,
+      "step": 49470
+    },
+    {
+      "epoch": 4.95597736264837,
+      "grad_norm": 1.8792906999588013,
+      "learning_rate": 9.666104872416148e-09,
+      "loss": 0.4221,
+      "step": 49480
+    },
+    {
+      "epoch": 4.956979015375369,
+      "grad_norm": 1.9721959829330444,
+      "learning_rate": 9.232729653177452e-09,
+      "loss": 0.3949,
+      "step": 49490
+    },
+    {
+      "epoch": 4.957980668102369,
+      "grad_norm": 2.2299644947052,
+      "learning_rate": 8.809291813419163e-09,
+      "loss": 0.4221,
+      "step": 49500
+    },
+    {
+      "epoch": 4.958982320829368,
+      "grad_norm": 2.5148696899414062,
+      "learning_rate": 8.39579152152048e-09,
+      "loss": 0.4239,
+      "step": 49510
+    },
+    {
+      "epoch": 4.959983973556368,
+      "grad_norm": 2.1798460483551025,
+      "learning_rate": 7.992228941905433e-09,
+      "loss": 0.427,
+      "step": 49520
+    },
+    {
+      "epoch": 4.960985626283367,
+      "grad_norm": 2.080162525177002,
+      "learning_rate": 7.598604235048434e-09,
+      "loss": 0.4076,
+      "step": 49530
+    },
+    {
+      "epoch": 4.961987279010367,
+      "grad_norm": 1.9533112049102783,
+      "learning_rate": 7.214917557471501e-09,
+      "loss": 0.3875,
+      "step": 49540
+    },
+    {
+      "epoch": 4.9629889317373665,
+      "grad_norm": 2.1065080165863037,
+      "learning_rate": 6.841169061744257e-09,
+      "loss": 0.4761,
+      "step": 49550
+    },
+    {
+      "epoch": 4.963990584464367,
+      "grad_norm": 2.6746320724487305,
+      "learning_rate": 6.477358896483932e-09,
+      "loss": 0.3773,
+      "step": 49560
+    },
+    {
+      "epoch": 4.964992237191366,
+      "grad_norm": 1.8172719478607178,
+      "learning_rate": 6.1234872063553604e-09,
+      "loss": 0.3882,
+      "step": 49570
+    },
+    {
+      "epoch": 4.965993889918366,
+      "grad_norm": 2.7890677452087402,
+      "learning_rate": 5.779554132076537e-09,
+      "loss": 0.3805,
+      "step": 49580
+    },
+    {
+      "epoch": 4.966995542645365,
+      "grad_norm": 2.2248497009277344,
+      "learning_rate": 5.445559810407508e-09,
+      "loss": 0.4279,
+      "step": 49590
+    },
+    {
+      "epoch": 4.967997195372364,
+      "grad_norm": 2.016270399093628,
+      "learning_rate": 5.1215043741587034e-09,
+      "loss": 0.3731,
+      "step": 49600
+    },
+    {
+      "epoch": 4.968998848099364,
+      "grad_norm": 2.032270669937134,
+      "learning_rate": 4.8073879521909335e-09,
+      "loss": 0.4135,
+      "step": 49610
+    },
+    {
+      "epoch": 4.970000500826363,
+      "grad_norm": 2.85500431060791,
+      "learning_rate": 4.50321066940429e-09,
+      "loss": 0.4544,
+      "step": 49620
+    },
+    {
+      "epoch": 4.971002153553363,
+      "grad_norm": 1.6594706773757935,
+      "learning_rate": 4.2089726467547945e-09,
+      "loss": 0.4161,
+      "step": 49630
+    },
+    {
+      "epoch": 4.972003806280362,
+      "grad_norm": 2.2570676803588867,
+      "learning_rate": 3.9246740012488515e-09,
+      "loss": 0.4116,
+      "step": 49640
+    },
+    {
+      "epoch": 4.9730054590073625,
+      "grad_norm": 2.1982297897338867,
+      "learning_rate": 3.6503148459265944e-09,
+      "loss": 0.394,
+      "step": 49650
+    },
+    {
+      "epoch": 4.974007111734362,
+      "grad_norm": 2.184145450592041,
+      "learning_rate": 3.3858952898924156e-09,
+      "loss": 0.4166,
+      "step": 49660
+    },
+    {
+      "epoch": 4.975008764461361,
+      "grad_norm": 1.760334849357605,
+      "learning_rate": 3.1314154382872108e-09,
+      "loss": 0.4014,
+      "step": 49670
+    },
+    {
+      "epoch": 4.976010417188361,
+      "grad_norm": 2.06929612159729,
+      "learning_rate": 2.8868753923022573e-09,
+      "loss": 0.4085,
+      "step": 49680
+    },
+    {
+      "epoch": 4.977012069915361,
+      "grad_norm": 2.60774827003479,
+      "learning_rate": 2.6522752491792147e-09,
+      "loss": 0.3767,
+      "step": 49690
+    },
+    {
+      "epoch": 4.97801372264236,
+      "grad_norm": 2.44492769241333,
+      "learning_rate": 2.4276151022045724e-09,
+      "loss": 0.4283,
+      "step": 49700
+    },
+    {
+      "epoch": 4.979015375369359,
+      "grad_norm": 2.499879837036133,
+      "learning_rate": 2.212895040712426e-09,
+      "loss": 0.4121,
+      "step": 49710
+    },
+    {
+      "epoch": 4.980017028096359,
+      "grad_norm": 1.796566367149353,
+      "learning_rate": 2.008115150081702e-09,
+      "loss": 0.3966,
+      "step": 49720
+    },
+    {
+      "epoch": 4.981018680823358,
+      "grad_norm": 2.4814465045928955,
+      "learning_rate": 1.8132755117444832e-09,
+      "loss": 0.3729,
+      "step": 49730
+    },
+    {
+      "epoch": 4.982020333550358,
+      "grad_norm": 2.5805459022521973,
+      "learning_rate": 1.628376203177684e-09,
+      "loss": 0.3886,
+      "step": 49740
+    },
+    {
+      "epoch": 4.9830219862773575,
+      "grad_norm": 2.3878934383392334,
+      "learning_rate": 1.453417297903048e-09,
+      "loss": 0.42,
+      "step": 49750
+    },
+    {
+      "epoch": 4.9840236390043575,
+      "grad_norm": 1.974474549293518,
+      "learning_rate": 1.2883988654927015e-09,
+      "loss": 0.4083,
+      "step": 49760
+    },
+    {
+      "epoch": 4.985025291731357,
+      "grad_norm": 2.5575461387634277,
+      "learning_rate": 1.1333209715636006e-09,
+      "loss": 0.4061,
+      "step": 49770
+    },
+    {
+      "epoch": 4.986026944458356,
+      "grad_norm": 1.7831733226776123,
+      "learning_rate": 9.881836777830832e-10,
+      "loss": 0.4559,
+      "step": 49780
+    },
+    {
+      "epoch": 4.987028597185356,
+      "grad_norm": 2.3388864994049072,
+      "learning_rate": 8.529870418633179e-10,
+      "loss": 0.4285,
+      "step": 49790
+    },
+    {
+      "epoch": 4.988030249912356,
+      "grad_norm": 2.1213908195495605,
+      "learning_rate": 7.277311175640789e-10,
+      "loss": 0.4158,
+      "step": 49800
+    },
+    {
+      "epoch": 4.989031902639355,
+      "grad_norm": 2.5922374725341797,
+      "learning_rate": 6.124159546899711e-10,
+      "loss": 0.4285,
+      "step": 49810
+    },
+    {
+      "epoch": 4.990033555366354,
+      "grad_norm": 1.7242733240127563,
+      "learning_rate": 5.070415990987565e-10,
+      "loss": 0.421,
+      "step": 49820
+    },
+    {
+      "epoch": 4.991035208093354,
+      "grad_norm": 2.315945863723755,
+      "learning_rate": 4.1160809269025213e-10,
+      "loss": 0.4358,
+      "step": 49830
+    },
+    {
+      "epoch": 4.992036860820353,
+      "grad_norm": 2.1397175788879395,
+      "learning_rate": 3.261154734146565e-10,
+      "loss": 0.4419,
+      "step": 49840
+    },
+    {
+      "epoch": 4.993038513547353,
+      "grad_norm": 1.9984654188156128,
+      "learning_rate": 2.505637752642231e-10,
+      "loss": 0.4377,
+      "step": 49850
+    },
+    {
+      "epoch": 4.9940401662743525,
+      "grad_norm": 1.9078891277313232,
+      "learning_rate": 1.84953028281587e-10,
+      "loss": 0.4617,
+      "step": 49860
+    },
+    {
+      "epoch": 4.995041819001353,
+      "grad_norm": 2.45648455619812,
+      "learning_rate": 1.2928325855976475e-10,
+      "loss": 0.4741,
+      "step": 49870
+    },
+    {
+      "epoch": 4.996043471728352,
+      "grad_norm": 1.3880661725997925,
+      "learning_rate": 8.355448823105238e-11,
+      "loss": 0.3786,
+      "step": 49880
+    },
+    {
+      "epoch": 4.997045124455351,
+      "grad_norm": 1.9489227533340454,
+      "learning_rate": 4.77667354836786e-11,
+      "loss": 0.3738,
+      "step": 49890
+    },
+    {
+      "epoch": 4.998046777182351,
+      "grad_norm": 2.0824508666992188,
+      "learning_rate": 2.192001454515147e-11,
+      "loss": 0.4642,
+      "step": 49900
+    },
+    {
+      "epoch": 4.99904842990935,
+      "grad_norm": 2.111239433288574,
+      "learning_rate": 6.014335693360629e-12,
+      "loss": 0.3691,
+      "step": 49910
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.7227387428283691,
+      "learning_rate": 4.970525657732594e-14,
+      "loss": 0.4032,
+      "step": 49920
+    },
+    {
+      "epoch": 5.001001652726999,
+      "grad_norm": 1.9057866334915161,
+      "learning_rate": 9.429981463315144e-06,
+      "loss": 0.3777,
+      "step": 49930
+    },
+    {
+      "epoch": 5.002003305453999,
+      "grad_norm": 2.094399929046631,
+      "learning_rate": 9.421178086170654e-06,
+      "loss": 0.3618,
+      "step": 49940
+    },
+    {
+      "epoch": 5.003004958180998,
+      "grad_norm": 2.0568583011627197,
+      "learning_rate": 9.412377866013538e-06,
+      "loss": 0.3926,
+      "step": 49950
+    },
+    {
+      "epoch": 5.004006610907998,
+      "grad_norm": 2.5123448371887207,
+      "learning_rate": 9.403580804627127e-06,
+      "loss": 0.4442,
+      "step": 49960
+    },
+    {
+      "epoch": 5.0050082636349975,
+      "grad_norm": 2.6346914768218994,
+      "learning_rate": 9.394786903794133e-06,
+      "loss": 0.4134,
+      "step": 49970
+    },
+    {
+      "epoch": 5.0060099163619975,
+      "grad_norm": 2.1011178493499756,
+      "learning_rate": 9.385996165296584e-06,
+      "loss": 0.3734,
+      "step": 49980
+    },
+    {
+      "epoch": 5.007011569088997,
+      "grad_norm": 2.0437891483306885,
+      "learning_rate": 9.377208590915892e-06,
+      "loss": 0.3885,
+      "step": 49990
+    },
+    {
+      "epoch": 5.008013221815997,
+      "grad_norm": 2.111875295639038,
+      "learning_rate": 9.368424182432825e-06,
+      "loss": 0.4173,
+      "step": 50000
+    },
+    {
+      "epoch": 5.008013221815997,
+      "eval_bleu": 0.4009660835826073,
+      "eval_loss": 0.508514940738678,
+      "eval_rouge1": 0.715396568133781,
+      "eval_rouge2": 0.550282125226329,
+      "eval_rougeL": 0.6740186461553936,
+      "eval_runtime": 77959.3622,
+      "eval_samples_per_second": 0.228,
+      "eval_steps_per_second": 0.028,
+      "eval_wer": 0.6738413384611066,
+      "step": 50000
+    },
+    {
+      "epoch": 5.009014874542996,
+      "grad_norm": 2.5804548263549805,
+      "learning_rate": 9.359642941627524e-06,
+      "loss": 0.4246,
+      "step": 50010
+    },
+    {
+      "epoch": 5.010016527269996,
+      "grad_norm": 2.5093562602996826,
+      "learning_rate": 9.350864870279457e-06,
+      "loss": 0.3901,
+      "step": 50020
+    },
+    {
+      "epoch": 5.011018179996995,
+      "grad_norm": 2.047625780105591,
+      "learning_rate": 9.342089970167458e-06,
+      "loss": 0.3293,
+      "step": 50030
+    },
+    {
+      "epoch": 5.012019832723994,
+      "grad_norm": 2.4648852348327637,
+      "learning_rate": 9.33331824306975e-06,
+      "loss": 0.47,
+      "step": 50040
+    },
+    {
+      "epoch": 5.013021485450994,
+      "grad_norm": 2.168367862701416,
+      "learning_rate": 9.324549690763887e-06,
+      "loss": 0.4254,
+      "step": 50050
+    },
+    {
+      "epoch": 5.014023138177993,
+      "grad_norm": 2.1796538829803467,
+      "learning_rate": 9.31578431502676e-06,
+      "loss": 0.4144,
+      "step": 50060
+    },
+    {
+      "epoch": 5.015024790904993,
+      "grad_norm": 2.1317245960235596,
+      "learning_rate": 9.307022117634646e-06,
+      "loss": 0.4444,
+      "step": 50070
+    },
+    {
+      "epoch": 5.016026443631993,
+      "grad_norm": 2.2961223125457764,
+      "learning_rate": 9.298263100363188e-06,
+      "loss": 0.4129,
+      "step": 50080
+    },
+    {
+      "epoch": 5.017028096358993,
+      "grad_norm": 2.3871800899505615,
+      "learning_rate": 9.289507264987348e-06,
+      "loss": 0.3687,
+      "step": 50090
+    },
+    {
+      "epoch": 5.018029749085992,
+      "grad_norm": 2.189589500427246,
+      "learning_rate": 9.280754613281456e-06,
+      "loss": 0.3685,
+      "step": 50100
+    },
+    {
+      "epoch": 5.019031401812992,
+      "grad_norm": 2.174671173095703,
+      "learning_rate": 9.272005147019225e-06,
+      "loss": 0.388,
+      "step": 50110
+    },
+    {
+      "epoch": 5.020033054539991,
+      "grad_norm": 1.7764248847961426,
+      "learning_rate": 9.263258867973696e-06,
+      "loss": 0.4141,
+      "step": 50120
+    },
+    {
+      "epoch": 5.021034707266991,
+      "grad_norm": 2.3782589435577393,
+      "learning_rate": 9.254515777917253e-06,
+      "loss": 0.4332,
+      "step": 50130
+    },
+    {
+      "epoch": 5.02203635999399,
+      "grad_norm": 1.9943517446517944,
+      "learning_rate": 9.245775878621649e-06,
+      "loss": 0.4191,
+      "step": 50140
+    },
+    {
+      "epoch": 5.023038012720989,
+      "grad_norm": 2.292137384414673,
+      "learning_rate": 9.237039171858006e-06,
+      "loss": 0.3959,
+      "step": 50150
+    },
+    {
+      "epoch": 5.024039665447989,
+      "grad_norm": 2.6247379779815674,
+      "learning_rate": 9.228305659396785e-06,
+      "loss": 0.3956,
+      "step": 50160
+    },
+    {
+      "epoch": 5.0250413181749884,
+      "grad_norm": 1.960654854774475,
+      "learning_rate": 9.219575343007771e-06,
+      "loss": 0.3702,
+      "step": 50170
+    },
+    {
+      "epoch": 5.0260429709019885,
+      "grad_norm": 1.9130282402038574,
+      "learning_rate": 9.210848224460158e-06,
+      "loss": 0.3923,
+      "step": 50180
+    },
+    {
+      "epoch": 5.027044623628988,
+      "grad_norm": 2.1862452030181885,
+      "learning_rate": 9.202124305522462e-06,
+      "loss": 0.3833,
+      "step": 50190
+    },
+    {
+      "epoch": 5.028046276355988,
+      "grad_norm": 1.7818164825439453,
+      "learning_rate": 9.193403587962527e-06,
+      "loss": 0.3926,
+      "step": 50200
+    },
+    {
+      "epoch": 5.029047929082987,
+      "grad_norm": 2.3955230712890625,
+      "learning_rate": 9.184686073547576e-06,
+      "loss": 0.4152,
+      "step": 50210
+    },
+    {
+      "epoch": 5.030049581809987,
+      "grad_norm": 2.290144681930542,
+      "learning_rate": 9.175971764044202e-06,
+      "loss": 0.4072,
+      "step": 50220
+    },
+    {
+      "epoch": 5.031051234536986,
+      "grad_norm": 1.9459701776504517,
+      "learning_rate": 9.167260661218322e-06,
+      "loss": 0.413,
+      "step": 50230
+    },
+    {
+      "epoch": 5.032052887263985,
+      "grad_norm": 1.623977541923523,
+      "learning_rate": 9.158552766835176e-06,
+      "loss": 0.3658,
+      "step": 50240
+    },
+    {
+      "epoch": 5.033054539990985,
+      "grad_norm": 2.4459660053253174,
+      "learning_rate": 9.149848082659417e-06,
+      "loss": 0.4204,
+      "step": 50250
+    },
+    {
+      "epoch": 5.034056192717984,
+      "grad_norm": 1.9712905883789062,
+      "learning_rate": 9.141146610455006e-06,
+      "loss": 0.4554,
+      "step": 50260
+    },
+    {
+      "epoch": 5.035057845444984,
+      "grad_norm": 2.139141798019409,
+      "learning_rate": 9.13244835198526e-06,
+      "loss": 0.3885,
+      "step": 50270
+    },
+    {
+      "epoch": 5.0360594981719835,
+      "grad_norm": 2.5080270767211914,
+      "learning_rate": 9.123753309012848e-06,
+      "loss": 0.4311,
+      "step": 50280
+    },
+    {
+      "epoch": 5.0370611508989835,
+      "grad_norm": 1.8112452030181885,
+      "learning_rate": 9.115061483299786e-06,
+      "loss": 0.3111,
+      "step": 50290
+    },
+    {
+      "epoch": 5.038062803625983,
+      "grad_norm": 2.099536180496216,
+      "learning_rate": 9.10637287660745e-06,
+      "loss": 0.3742,
+      "step": 50300
+    },
+    {
+      "epoch": 5.039064456352983,
+      "grad_norm": 2.1379730701446533,
+      "learning_rate": 9.097687490696522e-06,
+      "loss": 0.3914,
+      "step": 50310
+    },
+    {
+      "epoch": 5.040066109079982,
+      "grad_norm": 2.5523993968963623,
+      "learning_rate": 9.089005327327088e-06,
+      "loss": 0.4206,
+      "step": 50320
+    },
+    {
+      "epoch": 5.041067761806982,
+      "grad_norm": 1.9698617458343506,
+      "learning_rate": 9.08032638825855e-06,
+      "loss": 0.3978,
+      "step": 50330
+    },
+    {
+      "epoch": 5.042069414533981,
+      "grad_norm": 2.262073516845703,
+      "learning_rate": 9.071650675249658e-06,
+      "loss": 0.4593,
+      "step": 50340
+    },
+    {
+      "epoch": 5.04307106726098,
+      "grad_norm": 2.4841954708099365,
+      "learning_rate": 9.06297819005851e-06,
+      "loss": 0.3381,
+      "step": 50350
+    },
+    {
+      "epoch": 5.04407271998798,
+      "grad_norm": 2.5123629570007324,
+      "learning_rate": 9.054308934442554e-06,
+      "loss": 0.4633,
+      "step": 50360
+    },
+    {
+      "epoch": 5.045074372714979,
+      "grad_norm": 2.472378969192505,
+      "learning_rate": 9.045642910158581e-06,
+      "loss": 0.3572,
+      "step": 50370
+    },
+    {
+      "epoch": 5.046076025441979,
+      "grad_norm": 2.2146644592285156,
+      "learning_rate": 9.036980118962723e-06,
+      "loss": 0.4085,
+      "step": 50380
+    },
+    {
+      "epoch": 5.047077678168979,
+      "grad_norm": 2.2886135578155518,
+      "learning_rate": 9.028320562610465e-06,
+      "loss": 0.4098,
+      "step": 50390
+    },
+    {
+      "epoch": 5.048079330895979,
+      "grad_norm": 2.3172385692596436,
+      "learning_rate": 9.019664242856632e-06,
+      "loss": 0.401,
+      "step": 50400
+    },
+    {
+      "epoch": 5.049080983622978,
+      "grad_norm": 2.2763006687164307,
+      "learning_rate": 9.01101116145539e-06,
+      "loss": 0.4277,
+      "step": 50410
+    },
+    {
+      "epoch": 5.050082636349978,
+      "grad_norm": 2.075080633163452,
+      "learning_rate": 9.002361320160255e-06,
+      "loss": 0.4076,
+      "step": 50420
+    },
+    {
+      "epoch": 5.051084289076977,
+      "grad_norm": 2.275829315185547,
+      "learning_rate": 8.993714720724084e-06,
+      "loss": 0.4366,
+      "step": 50430
+    },
+    {
+      "epoch": 5.052085941803977,
+      "grad_norm": 2.0190999507904053,
+      "learning_rate": 8.985071364899072e-06,
+      "loss": 0.4001,
+      "step": 50440
+    },
+    {
+      "epoch": 5.053087594530976,
+      "grad_norm": 1.6728298664093018,
+      "learning_rate": 8.976431254436769e-06,
+      "loss": 0.3575,
+      "step": 50450
+    },
+    {
+      "epoch": 5.054089247257975,
+      "grad_norm": 2.0061964988708496,
+      "learning_rate": 8.967794391088052e-06,
+      "loss": 0.3538,
+      "step": 50460
+    },
+    {
+      "epoch": 5.055090899984975,
+      "grad_norm": 2.0933334827423096,
+      "learning_rate": 8.959160776603152e-06,
+      "loss": 0.3712,
+      "step": 50470
+    },
+    {
+      "epoch": 5.056092552711974,
+      "grad_norm": 2.1422982215881348,
+      "learning_rate": 8.950530412731634e-06,
+      "loss": 0.3951,
+      "step": 50480
+    },
+    {
+      "epoch": 5.0570942054389745,
+      "grad_norm": 2.5245361328125,
+      "learning_rate": 8.941903301222412e-06,
+      "loss": 0.3951,
+      "step": 50490
+    },
+    {
+      "epoch": 5.058095858165974,
+      "grad_norm": 1.8500747680664062,
+      "learning_rate": 8.933279443823733e-06,
+      "loss": 0.3443,
+      "step": 50500
+    },
+    {
+      "epoch": 5.059097510892974,
+      "grad_norm": 1.7166526317596436,
+      "learning_rate": 8.92465884228319e-06,
+      "loss": 0.4024,
+      "step": 50510
+    },
+    {
+      "epoch": 5.060099163619973,
+      "grad_norm": 1.9748985767364502,
+      "learning_rate": 8.916041498347712e-06,
+      "loss": 0.3806,
+      "step": 50520
+    },
+    {
+      "epoch": 5.061100816346973,
+      "grad_norm": 2.0966179370880127,
+      "learning_rate": 8.907427413763573e-06,
+      "loss": 0.3958,
+      "step": 50530
+    },
+    {
+      "epoch": 5.062102469073972,
+      "grad_norm": 3.496039867401123,
+      "learning_rate": 8.898816590276379e-06,
+      "loss": 0.3929,
+      "step": 50540
+    },
+    {
+      "epoch": 5.063104121800972,
+      "grad_norm": 1.860559105873108,
+      "learning_rate": 8.890209029631086e-06,
+      "loss": 0.3346,
+      "step": 50550
+    },
+    {
+      "epoch": 5.064105774527971,
+      "grad_norm": 2.681044340133667,
+      "learning_rate": 8.881604733571977e-06,
+      "loss": 0.3904,
+      "step": 50560
+    },
+    {
+      "epoch": 5.06510742725497,
+      "grad_norm": 1.6064103841781616,
+      "learning_rate": 8.873003703842681e-06,
+      "loss": 0.4072,
+      "step": 50570
+    },
+    {
+      "epoch": 5.06610907998197,
+      "grad_norm": 2.7219972610473633,
+      "learning_rate": 8.864405942186163e-06,
+      "loss": 0.4058,
+      "step": 50580
+    },
+    {
+      "epoch": 5.0671107327089695,
+      "grad_norm": 2.393876314163208,
+      "learning_rate": 8.855811450344729e-06,
+      "loss": 0.455,
+      "step": 50590
+    },
+    {
+      "epoch": 5.0681123854359695,
+      "grad_norm": 2.1746304035186768,
+      "learning_rate": 8.847220230060014e-06,
+      "loss": 0.4109,
+      "step": 50600
+    },
+    {
+      "epoch": 5.069114038162969,
+      "grad_norm": 1.8715713024139404,
+      "learning_rate": 8.838632283072998e-06,
+      "loss": 0.346,
+      "step": 50610
+    },
+    {
+      "epoch": 5.070115690889969,
+      "grad_norm": 2.0678677558898926,
+      "learning_rate": 8.830047611123992e-06,
+      "loss": 0.3984,
+      "step": 50620
+    },
+    {
+      "epoch": 5.071117343616968,
+      "grad_norm": 1.8897209167480469,
+      "learning_rate": 8.821466215952651e-06,
+      "loss": 0.3794,
+      "step": 50630
+    },
+    {
+      "epoch": 5.072118996343968,
+      "grad_norm": 3.4548091888427734,
+      "learning_rate": 8.81288809929796e-06,
+      "loss": 0.3808,
+      "step": 50640
+    },
+    {
+      "epoch": 5.073120649070967,
+      "grad_norm": 1.9121617078781128,
+      "learning_rate": 8.804313262898234e-06,
+      "loss": 0.4143,
+      "step": 50650
+    },
+    {
+      "epoch": 5.074122301797967,
+      "grad_norm": 2.0208704471588135,
+      "learning_rate": 8.795741708491139e-06,
+      "loss": 0.3962,
+      "step": 50660
+    },
+    {
+      "epoch": 5.075123954524966,
+      "grad_norm": 2.1051366329193115,
+      "learning_rate": 8.787173437813664e-06,
+      "loss": 0.3973,
+      "step": 50670
+    },
+    {
+      "epoch": 5.076125607251965,
+      "grad_norm": 2.504681348800659,
+      "learning_rate": 8.778608452602136e-06,
+      "loss": 0.406,
+      "step": 50680
+    },
+    {
+      "epoch": 5.077127259978965,
+      "grad_norm": 2.3707287311553955,
+      "learning_rate": 8.770046754592211e-06,
+      "loss": 0.4546,
+      "step": 50690
+    },
+    {
+      "epoch": 5.078128912705965,
+      "grad_norm": 2.108637809753418,
+      "learning_rate": 8.761488345518893e-06,
+      "loss": 0.3693,
+      "step": 50700
+    },
+    {
+      "epoch": 5.079130565432965,
+      "grad_norm": 2.4608542919158936,
+      "learning_rate": 8.752933227116503e-06,
+      "loss": 0.4373,
+      "step": 50710
+    },
+    {
+      "epoch": 5.080132218159964,
+      "grad_norm": 2.6883091926574707,
+      "learning_rate": 8.7443814011187e-06,
+      "loss": 0.4489,
+      "step": 50720
+    },
+    {
+      "epoch": 5.081133870886964,
+      "grad_norm": 1.7556604146957397,
+      "learning_rate": 8.735832869258486e-06,
+      "loss": 0.3975,
+      "step": 50730
+    },
+    {
+      "epoch": 5.082135523613963,
+      "grad_norm": 2.1072566509246826,
+      "learning_rate": 8.727287633268182e-06,
+      "loss": 0.4055,
+      "step": 50740
+    },
+    {
+      "epoch": 5.083137176340963,
+      "grad_norm": 2.7997405529022217,
+      "learning_rate": 8.718745694879451e-06,
+      "loss": 0.4077,
+      "step": 50750
+    },
+    {
+      "epoch": 5.084138829067962,
+      "grad_norm": 2.5607924461364746,
+      "learning_rate": 8.710207055823272e-06,
+      "loss": 0.4069,
+      "step": 50760
+    },
+    {
+      "epoch": 5.085140481794961,
+      "grad_norm": 2.0641298294067383,
+      "learning_rate": 8.701671717829993e-06,
+      "loss": 0.371,
+      "step": 50770
+    },
+    {
+      "epoch": 5.086142134521961,
+      "grad_norm": 2.205183982849121,
+      "learning_rate": 8.69313968262924e-06,
+      "loss": 0.4074,
+      "step": 50780
+    },
+    {
+      "epoch": 5.08714378724896,
+      "grad_norm": 2.3521621227264404,
+      "learning_rate": 8.684610951950006e-06,
+      "loss": 0.3741,
+      "step": 50790
+    },
+    {
+      "epoch": 5.0881454399759605,
+      "grad_norm": 2.01767635345459,
+      "learning_rate": 8.676085527520605e-06,
+      "loss": 0.4264,
+      "step": 50800
+    },
+    {
+      "epoch": 5.08914709270296,
+      "grad_norm": 2.336679458618164,
+      "learning_rate": 8.66756341106868e-06,
+      "loss": 0.4016,
+      "step": 50810
+    },
+    {
+      "epoch": 5.09014874542996,
+      "grad_norm": 2.2592053413391113,
+      "learning_rate": 8.659044604321206e-06,
+      "loss": 0.4801,
+      "step": 50820
+    },
+    {
+      "epoch": 5.091150398156959,
+      "grad_norm": 2.573158025741577,
+      "learning_rate": 8.65052910900448e-06,
+      "loss": 0.4175,
+      "step": 50830
+    },
+    {
+      "epoch": 5.092152050883959,
+      "grad_norm": 2.452357292175293,
+      "learning_rate": 8.642016926844154e-06,
+      "loss": 0.4282,
+      "step": 50840
+    },
+    {
+      "epoch": 5.093153703610958,
+      "grad_norm": 1.6255464553833008,
+      "learning_rate": 8.633508059565166e-06,
+      "loss": 0.4136,
+      "step": 50850
+    },
+    {
+      "epoch": 5.094155356337958,
+      "grad_norm": 1.7420393228530884,
+      "learning_rate": 8.625002508891813e-06,
+      "loss": 0.3775,
+      "step": 50860
+    },
+    {
+      "epoch": 5.095157009064957,
+      "grad_norm": 2.3295178413391113,
+      "learning_rate": 8.6165002765477e-06,
+      "loss": 0.4032,
+      "step": 50870
+    },
+    {
+      "epoch": 5.096158661791956,
+      "grad_norm": 1.9889898300170898,
+      "learning_rate": 8.608001364255803e-06,
+      "loss": 0.4529,
+      "step": 50880
+    },
+    {
+      "epoch": 5.097160314518956,
+      "grad_norm": 2.4790945053100586,
+      "learning_rate": 8.599505773738365e-06,
+      "loss": 0.3701,
+      "step": 50890
+    },
+    {
+      "epoch": 5.0981619672459555,
+      "grad_norm": 2.510486602783203,
+      "learning_rate": 8.59101350671698e-06,
+      "loss": 0.4356,
+      "step": 50900
+    },
+    {
+      "epoch": 5.0991636199729555,
+      "grad_norm": 2.439235210418701,
+      "learning_rate": 8.582524564912604e-06,
+      "loss": 0.4114,
+      "step": 50910
+    },
+    {
+      "epoch": 5.100165272699955,
+      "grad_norm": 2.2638766765594482,
+      "learning_rate": 8.574038950045457e-06,
+      "loss": 0.3951,
+      "step": 50920
+    },
+    {
+      "epoch": 5.101166925426955,
+      "grad_norm": 2.317436933517456,
+      "learning_rate": 8.565556663835131e-06,
+      "loss": 0.4255,
+      "step": 50930
+    },
+    {
+      "epoch": 5.102168578153954,
+      "grad_norm": 2.248082160949707,
+      "learning_rate": 8.557077708000514e-06,
+      "loss": 0.445,
+      "step": 50940
+    },
+    {
+      "epoch": 5.103170230880954,
+      "grad_norm": 2.4589478969573975,
+      "learning_rate": 8.54860208425986e-06,
+      "loss": 0.3887,
+      "step": 50950
+    },
+    {
+      "epoch": 5.104171883607953,
+      "grad_norm": 2.3282318115234375,
+      "learning_rate": 8.540129794330699e-06,
+      "loss": 0.3674,
+      "step": 50960
+    },
+    {
+      "epoch": 5.105173536334953,
+      "grad_norm": 1.6831501722335815,
+      "learning_rate": 8.5316608399299e-06,
+      "loss": 0.4535,
+      "step": 50970
+    },
+    {
+      "epoch": 5.106175189061952,
+      "grad_norm": 1.7259416580200195,
+      "learning_rate": 8.523195222773688e-06,
+      "loss": 0.3896,
+      "step": 50980
+    },
+    {
+      "epoch": 5.107176841788951,
+      "grad_norm": 2.5421462059020996,
+      "learning_rate": 8.514732944577583e-06,
+      "loss": 0.3938,
+      "step": 50990
+    },
+    {
+      "epoch": 5.108178494515951,
+      "grad_norm": 2.577516794204712,
+      "learning_rate": 8.506274007056412e-06,
+      "loss": 0.4579,
+      "step": 51000
+    },
+    {
+      "epoch": 5.1091801472429506,
+      "grad_norm": 2.2617828845977783,
+      "learning_rate": 8.497818411924363e-06,
+      "loss": 0.4584,
+      "step": 51010
+    },
+    {
+      "epoch": 5.110181799969951,
+      "grad_norm": 2.734596014022827,
+      "learning_rate": 8.489366160894937e-06,
+      "loss": 0.4102,
+      "step": 51020
+    },
+    {
+      "epoch": 5.11118345269695,
+      "grad_norm": 2.1711714267730713,
+      "learning_rate": 8.480917255680929e-06,
+      "loss": 0.4107,
+      "step": 51030
+    },
+    {
+      "epoch": 5.11218510542395,
+      "grad_norm": 2.1046245098114014,
+      "learning_rate": 8.472471697994478e-06,
+      "loss": 0.4126,
+      "step": 51040
+    },
+    {
+      "epoch": 5.113186758150949,
+      "grad_norm": 2.018583059310913,
+      "learning_rate": 8.464029489547057e-06,
+      "loss": 0.3639,
+      "step": 51050
+    },
+    {
+      "epoch": 5.114188410877949,
+      "grad_norm": 2.4939472675323486,
+      "learning_rate": 8.455590632049451e-06,
+      "loss": 0.391,
+      "step": 51060
+    },
+    {
+      "epoch": 5.115190063604948,
+      "grad_norm": 2.2978515625,
+      "learning_rate": 8.447155127211734e-06,
+      "loss": 0.3951,
+      "step": 51070
+    },
+    {
+      "epoch": 5.116191716331948,
+      "grad_norm": 2.4101288318634033,
+      "learning_rate": 8.438722976743352e-06,
+      "loss": 0.4014,
+      "step": 51080
+    },
+    {
+      "epoch": 5.117193369058947,
+      "grad_norm": 1.3514302968978882,
+      "learning_rate": 8.430294182353049e-06,
+      "loss": 0.3879,
+      "step": 51090
+    },
+    {
+      "epoch": 5.118195021785946,
+      "grad_norm": 3.136054277420044,
+      "learning_rate": 8.421868745748873e-06,
+      "loss": 0.4316,
+      "step": 51100
+    },
+    {
+      "epoch": 5.1191966745129465,
+      "grad_norm": 2.85766339302063,
+      "learning_rate": 8.4134466686382e-06,
+      "loss": 0.3821,
+      "step": 51110
+    },
+    {
+      "epoch": 5.120198327239946,
+      "grad_norm": 2.4241206645965576,
+      "learning_rate": 8.405027952727754e-06,
+      "loss": 0.5232,
+      "step": 51120
+    },
+    {
+      "epoch": 5.121199979966946,
+      "grad_norm": 1.8085269927978516,
+      "learning_rate": 8.39661259972355e-06,
+      "loss": 0.4257,
+      "step": 51130
+    },
+    {
+      "epoch": 5.122201632693945,
+      "grad_norm": 1.9889925718307495,
+      "learning_rate": 8.388200611330902e-06,
+      "loss": 0.4042,
+      "step": 51140
+    },
+    {
+      "epoch": 5.123203285420945,
+      "grad_norm": 1.835249662399292,
+      "learning_rate": 8.379791989254493e-06,
+      "loss": 0.3501,
+      "step": 51150
+    },
+    {
+      "epoch": 5.124204938147944,
+      "grad_norm": 2.209379196166992,
+      "learning_rate": 8.371386735198292e-06,
+      "loss": 0.3846,
+      "step": 51160
+    },
+    {
+      "epoch": 5.125206590874944,
+      "grad_norm": 1.8162431716918945,
+      "learning_rate": 8.36298485086559e-06,
+      "loss": 0.4411,
+      "step": 51170
+    },
+    {
+      "epoch": 5.126208243601943,
+      "grad_norm": 2.113241195678711,
+      "learning_rate": 8.354586337958983e-06,
+      "loss": 0.4041,
+      "step": 51180
+    },
+    {
+      "epoch": 5.127209896328942,
+      "grad_norm": 1.9983274936676025,
+      "learning_rate": 8.346191198180414e-06,
+      "loss": 0.379,
+      "step": 51190
+    },
+    {
+      "epoch": 5.128211549055942,
+      "grad_norm": 2.5187878608703613,
+      "learning_rate": 8.337799433231126e-06,
+      "loss": 0.4103,
+      "step": 51200
+    },
+    {
+      "epoch": 5.1292132017829415,
+      "grad_norm": 2.1544573307037354,
+      "learning_rate": 8.329411044811653e-06,
+      "loss": 0.4325,
+      "step": 51210
+    },
+    {
+      "epoch": 5.1302148545099415,
+      "grad_norm": 2.4598071575164795,
+      "learning_rate": 8.321026034621896e-06,
+      "loss": 0.3789,
+      "step": 51220
+    },
+    {
+      "epoch": 5.131216507236941,
+      "grad_norm": 2.138986349105835,
+      "learning_rate": 8.312644404361033e-06,
+      "loss": 0.4073,
+      "step": 51230
+    },
+    {
+      "epoch": 5.132218159963941,
+      "grad_norm": 2.2500813007354736,
+      "learning_rate": 8.30426615572758e-06,
+      "loss": 0.4308,
+      "step": 51240
+    },
+    {
+      "epoch": 5.13321981269094,
+      "grad_norm": 2.7701327800750732,
+      "learning_rate": 8.295891290419334e-06,
+      "loss": 0.3836,
+      "step": 51250
+    },
+    {
+      "epoch": 5.13422146541794,
+      "grad_norm": 1.6952491998672485,
+      "learning_rate": 8.287519810133443e-06,
+      "loss": 0.4298,
+      "step": 51260
+    },
+    {
+      "epoch": 5.135223118144939,
+      "grad_norm": 1.9191524982452393,
+      "learning_rate": 8.279151716566358e-06,
+      "loss": 0.3448,
+      "step": 51270
+    },
+    {
+      "epoch": 5.136224770871939,
+      "grad_norm": 1.756516695022583,
+      "learning_rate": 8.270787011413833e-06,
+      "loss": 0.4178,
+      "step": 51280
+    },
+    {
+      "epoch": 5.137226423598938,
+      "grad_norm": 1.841568112373352,
+      "learning_rate": 8.262425696370949e-06,
+      "loss": 0.3687,
+      "step": 51290
+    },
+    {
+      "epoch": 5.138228076325937,
+      "grad_norm": 2.1284594535827637,
+      "learning_rate": 8.254067773132085e-06,
+      "loss": 0.3736,
+      "step": 51300
+    },
+    {
+      "epoch": 5.139229729052937,
+      "grad_norm": 1.7997220754623413,
+      "learning_rate": 8.24571324339096e-06,
+      "loss": 0.3807,
+      "step": 51310
+    },
+    {
+      "epoch": 5.1402313817799365,
+      "grad_norm": 2.3964970111846924,
+      "learning_rate": 8.237362108840555e-06,
+      "loss": 0.4019,
+      "step": 51320
+    },
+    {
+      "epoch": 5.141233034506937,
+      "grad_norm": 2.325812816619873,
+      "learning_rate": 8.22901437117322e-06,
+      "loss": 0.3861,
+      "step": 51330
+    },
+    {
+      "epoch": 5.142234687233936,
+      "grad_norm": 2.397308826446533,
+      "learning_rate": 8.220670032080587e-06,
+      "loss": 0.4339,
+      "step": 51340
+    },
+    {
+      "epoch": 5.143236339960936,
+      "grad_norm": 2.3338468074798584,
+      "learning_rate": 8.212329093253605e-06,
+      "loss": 0.4321,
+      "step": 51350
+    },
+    {
+      "epoch": 5.144237992687935,
+      "grad_norm": 1.8569867610931396,
+      "learning_rate": 8.203991556382523e-06,
+      "loss": 0.3943,
+      "step": 51360
+    },
+    {
+      "epoch": 5.145239645414935,
+      "grad_norm": 1.970132827758789,
+      "learning_rate": 8.195657423156921e-06,
+      "loss": 0.3841,
+      "step": 51370
+    },
+    {
+      "epoch": 5.146241298141934,
+      "grad_norm": 2.166137933731079,
+      "learning_rate": 8.187326695265671e-06,
+      "loss": 0.4063,
+      "step": 51380
+    },
+    {
+      "epoch": 5.147242950868934,
+      "grad_norm": 2.590151786804199,
+      "learning_rate": 8.178999374396967e-06,
+      "loss": 0.4072,
+      "step": 51390
+    },
+    {
+      "epoch": 5.148244603595933,
+      "grad_norm": 2.1858596801757812,
+      "learning_rate": 8.170675462238306e-06,
+      "loss": 0.3971,
+      "step": 51400
+    },
+    {
+      "epoch": 5.149246256322932,
+      "grad_norm": 2.6428606510162354,
+      "learning_rate": 8.162354960476498e-06,
+      "loss": 0.4001,
+      "step": 51410
+    },
+    {
+      "epoch": 5.1502479090499325,
+      "grad_norm": 2.1872785091400146,
+      "learning_rate": 8.154037870797657e-06,
+      "loss": 0.3665,
+      "step": 51420
+    },
+    {
+      "epoch": 5.151249561776932,
+      "grad_norm": 2.6396920680999756,
+      "learning_rate": 8.14572419488721e-06,
+      "loss": 0.3787,
+      "step": 51430
+    },
+    {
+      "epoch": 5.152251214503932,
+      "grad_norm": 2.438502073287964,
+      "learning_rate": 8.137413934429893e-06,
+      "loss": 0.3889,
+      "step": 51440
+    },
+    {
+      "epoch": 5.153252867230931,
+      "grad_norm": 2.2407338619232178,
+      "learning_rate": 8.12910709110975e-06,
+      "loss": 0.4352,
+      "step": 51450
+    },
+    {
+      "epoch": 5.154254519957931,
+      "grad_norm": 1.8680403232574463,
+      "learning_rate": 8.120803666610122e-06,
+      "loss": 0.367,
+      "step": 51460
+    },
+    {
+      "epoch": 5.15525617268493,
+      "grad_norm": 2.1846930980682373,
+      "learning_rate": 8.112503662613672e-06,
+      "loss": 0.3574,
+      "step": 51470
+    },
+    {
+      "epoch": 5.15625782541193,
+      "grad_norm": 1.9968862533569336,
+      "learning_rate": 8.104207080802361e-06,
+      "loss": 0.4072,
+      "step": 51480
+    },
+    {
+      "epoch": 5.157259478138929,
+      "grad_norm": 2.1309328079223633,
+      "learning_rate": 8.09591392285746e-06,
+      "loss": 0.4554,
+      "step": 51490
+    },
+    {
+      "epoch": 5.158261130865929,
+      "grad_norm": 2.3871166706085205,
+      "learning_rate": 8.087624190459545e-06,
+      "loss": 0.4323,
+      "step": 51500
+    },
+    {
+      "epoch": 5.159262783592928,
+      "grad_norm": 2.2008261680603027,
+      "learning_rate": 8.079337885288496e-06,
+      "loss": 0.4549,
+      "step": 51510
+    },
+    {
+      "epoch": 5.1602644363199275,
+      "grad_norm": 2.131897211074829,
+      "learning_rate": 8.071055009023505e-06,
+      "loss": 0.4221,
+      "step": 51520
+    },
+    {
+      "epoch": 5.1612660890469275,
+      "grad_norm": 2.741530418395996,
+      "learning_rate": 8.062775563343056e-06,
+      "loss": 0.4023,
+      "step": 51530
+    },
+    {
+      "epoch": 5.162267741773927,
+      "grad_norm": 1.911015272140503,
+      "learning_rate": 8.054499549924955e-06,
+      "loss": 0.4323,
+      "step": 51540
+    },
+    {
+      "epoch": 5.163269394500927,
+      "grad_norm": 2.1528337001800537,
+      "learning_rate": 8.046226970446299e-06,
+      "loss": 0.4609,
+      "step": 51550
+    },
+    {
+      "epoch": 5.164271047227926,
+      "grad_norm": 2.0719454288482666,
+      "learning_rate": 8.037957826583497e-06,
+      "loss": 0.3626,
+      "step": 51560
+    },
+    {
+      "epoch": 5.165272699954926,
+      "grad_norm": 2.4185614585876465,
+      "learning_rate": 8.029692120012255e-06,
+      "loss": 0.4371,
+      "step": 51570
+    },
+    {
+      "epoch": 5.166274352681925,
+      "grad_norm": 2.706132411956787,
+      "learning_rate": 8.021429852407592e-06,
+      "loss": 0.4581,
+      "step": 51580
+    },
+    {
+      "epoch": 5.167276005408925,
+      "grad_norm": 2.2206037044525146,
+      "learning_rate": 8.013171025443816e-06,
+      "loss": 0.3972,
+      "step": 51590
+    },
+    {
+      "epoch": 5.168277658135924,
+      "grad_norm": 1.9738082885742188,
+      "learning_rate": 8.004915640794553e-06,
+      "loss": 0.3954,
+      "step": 51600
+    },
+    {
+      "epoch": 5.169279310862924,
+      "grad_norm": 2.3320820331573486,
+      "learning_rate": 7.996663700132723e-06,
+      "loss": 0.4122,
+      "step": 51610
+    },
+    {
+      "epoch": 5.170280963589923,
+      "grad_norm": 1.627105474472046,
+      "learning_rate": 7.988415205130545e-06,
+      "loss": 0.3838,
+      "step": 51620
+    },
+    {
+      "epoch": 5.1712826163169225,
+      "grad_norm": 1.9464679956436157,
+      "learning_rate": 7.980170157459549e-06,
+      "loss": 0.3813,
+      "step": 51630
+    },
+    {
+      "epoch": 5.172284269043923,
+      "grad_norm": 2.1501245498657227,
+      "learning_rate": 7.971928558790562e-06,
+      "loss": 0.4326,
+      "step": 51640
+    },
+    {
+      "epoch": 5.173285921770922,
+      "grad_norm": 1.8441542387008667,
+      "learning_rate": 7.963690410793709e-06,
+      "loss": 0.3666,
+      "step": 51650
+    },
+    {
+      "epoch": 5.174287574497922,
+      "grad_norm": 1.986935019493103,
+      "learning_rate": 7.955455715138419e-06,
+      "loss": 0.4267,
+      "step": 51660
+    },
+    {
+      "epoch": 5.175289227224921,
+      "grad_norm": 2.072371006011963,
+      "learning_rate": 7.94722447349342e-06,
+      "loss": 0.4365,
+      "step": 51670
+    },
+    {
+      "epoch": 5.176290879951921,
+      "grad_norm": 1.5192880630493164,
+      "learning_rate": 7.938996687526745e-06,
+      "loss": 0.3503,
+      "step": 51680
+    },
+    {
+      "epoch": 5.17729253267892,
+      "grad_norm": 2.655179500579834,
+      "learning_rate": 7.930772358905719e-06,
+      "loss": 0.4375,
+      "step": 51690
+    },
+    {
+      "epoch": 5.17829418540592,
+      "grad_norm": 2.249799966812134,
+      "learning_rate": 7.92255148929697e-06,
+      "loss": 0.3607,
+      "step": 51700
+    },
+    {
+      "epoch": 5.179295838132919,
+      "grad_norm": 2.191415548324585,
+      "learning_rate": 7.914334080366428e-06,
+      "loss": 0.3986,
+      "step": 51710
+    },
+    {
+      "epoch": 5.180297490859919,
+      "grad_norm": 2.2049078941345215,
+      "learning_rate": 7.906120133779318e-06,
+      "loss": 0.3843,
+      "step": 51720
+    },
+    {
+      "epoch": 5.1812991435869185,
+      "grad_norm": 3.0092198848724365,
+      "learning_rate": 7.897909651200152e-06,
+      "loss": 0.4309,
+      "step": 51730
+    },
+    {
+      "epoch": 5.182300796313918,
+      "grad_norm": 2.194514751434326,
+      "learning_rate": 7.889702634292785e-06,
+      "loss": 0.4127,
+      "step": 51740
+    },
+    {
+      "epoch": 5.183302449040918,
+      "grad_norm": 2.2221009731292725,
+      "learning_rate": 7.881499084720301e-06,
+      "loss": 0.3811,
+      "step": 51750
+    },
+    {
+      "epoch": 5.184304101767917,
+      "grad_norm": 1.7878801822662354,
+      "learning_rate": 7.873299004145136e-06,
+      "loss": 0.3726,
+      "step": 51760
+    },
+    {
+      "epoch": 5.185305754494917,
+      "grad_norm": 2.5555198192596436,
+      "learning_rate": 7.86510239422899e-06,
+      "loss": 0.4467,
+      "step": 51770
+    },
+    {
+      "epoch": 5.186307407221916,
+      "grad_norm": 2.4053938388824463,
+      "learning_rate": 7.8569092566329e-06,
+      "loss": 0.3929,
+      "step": 51780
+    },
+    {
+      "epoch": 5.187309059948916,
+      "grad_norm": 2.1048076152801514,
+      "learning_rate": 7.84871959301715e-06,
+      "loss": 0.3874,
+      "step": 51790
+    },
+    {
+      "epoch": 5.188310712675915,
+      "grad_norm": 2.1604602336883545,
+      "learning_rate": 7.840533405041343e-06,
+      "loss": 0.381,
+      "step": 51800
+    },
+    {
+      "epoch": 5.189312365402915,
+      "grad_norm": 2.2942380905151367,
+      "learning_rate": 7.8323506943644e-06,
+      "loss": 0.4038,
+      "step": 51810
+    },
+    {
+      "epoch": 5.190314018129914,
+      "grad_norm": 2.568305015563965,
+      "learning_rate": 7.824171462644493e-06,
+      "loss": 0.4428,
+      "step": 51820
+    },
+    {
+      "epoch": 5.1913156708569135,
+      "grad_norm": 2.6570141315460205,
+      "learning_rate": 7.81599571153912e-06,
+      "loss": 0.4239,
+      "step": 51830
+    },
+    {
+      "epoch": 5.1923173235839135,
+      "grad_norm": 2.121030807495117,
+      "learning_rate": 7.807823442705056e-06,
+      "loss": 0.4091,
+      "step": 51840
+    },
+    {
+      "epoch": 5.193318976310913,
+      "grad_norm": 2.315417528152466,
+      "learning_rate": 7.799654657798402e-06,
+      "loss": 0.39,
+      "step": 51850
+    },
+    {
+      "epoch": 5.194320629037913,
+      "grad_norm": 2.200780153274536,
+      "learning_rate": 7.79148935847451e-06,
+      "loss": 0.3821,
+      "step": 51860
+    },
+    {
+      "epoch": 5.195322281764912,
+      "grad_norm": 2.2050414085388184,
+      "learning_rate": 7.783327546388045e-06,
+      "loss": 0.414,
+      "step": 51870
+    },
+    {
+      "epoch": 5.196323934491912,
+      "grad_norm": 2.513406991958618,
+      "learning_rate": 7.77516922319298e-06,
+      "loss": 0.4066,
+      "step": 51880
+    },
+    {
+      "epoch": 5.197325587218911,
+      "grad_norm": 2.4052371978759766,
+      "learning_rate": 7.767014390542565e-06,
+      "loss": 0.3507,
+      "step": 51890
+    },
+    {
+      "epoch": 5.198327239945911,
+      "grad_norm": 2.228341579437256,
+      "learning_rate": 7.758863050089337e-06,
+      "loss": 0.4601,
+      "step": 51900
+    },
+    {
+      "epoch": 5.19932889267291,
+      "grad_norm": 2.076115608215332,
+      "learning_rate": 7.750715203485127e-06,
+      "loss": 0.3405,
+      "step": 51910
+    },
+    {
+      "epoch": 5.20033054539991,
+      "grad_norm": 2.2756433486938477,
+      "learning_rate": 7.742570852381092e-06,
+      "loss": 0.4388,
+      "step": 51920
+    },
+    {
+      "epoch": 5.201332198126909,
+      "grad_norm": 1.7769001722335815,
+      "learning_rate": 7.734429998427626e-06,
+      "loss": 0.3725,
+      "step": 51930
+    },
+    {
+      "epoch": 5.2023338508539085,
+      "grad_norm": 2.7252211570739746,
+      "learning_rate": 7.726292643274441e-06,
+      "loss": 0.4404,
+      "step": 51940
+    },
+    {
+      "epoch": 5.203335503580909,
+      "grad_norm": 2.693631887435913,
+      "learning_rate": 7.718158788570557e-06,
+      "loss": 0.3966,
+      "step": 51950
+    },
+    {
+      "epoch": 5.204337156307908,
+      "grad_norm": 1.9786999225616455,
+      "learning_rate": 7.71002843596427e-06,
+      "loss": 0.3878,
+      "step": 51960
+    },
+    {
+      "epoch": 5.205338809034908,
+      "grad_norm": 2.3402976989746094,
+      "learning_rate": 7.701901587103146e-06,
+      "loss": 0.391,
+      "step": 51970
+    },
+    {
+      "epoch": 5.206340461761907,
+      "grad_norm": 1.6999900341033936,
+      "learning_rate": 7.69377824363406e-06,
+      "loss": 0.4084,
+      "step": 51980
+    },
+    {
+      "epoch": 5.207342114488907,
+      "grad_norm": 2.1598548889160156,
+      "learning_rate": 7.685658407203192e-06,
+      "loss": 0.3995,
+      "step": 51990
+    },
+    {
+      "epoch": 5.208343767215906,
+      "grad_norm": 2.034207582473755,
+      "learning_rate": 7.677542079455994e-06,
+      "loss": 0.4061,
+      "step": 52000
+    },
+    {
+      "epoch": 5.209345419942906,
+      "grad_norm": 2.343937635421753,
+      "learning_rate": 7.669429262037183e-06,
+      "loss": 0.355,
+      "step": 52010
+    },
+    {
+      "epoch": 5.210347072669905,
+      "grad_norm": 2.3966968059539795,
+      "learning_rate": 7.661319956590817e-06,
+      "loss": 0.3928,
+      "step": 52020
+    },
+    {
+      "epoch": 5.211348725396905,
+      "grad_norm": 3.0318691730499268,
+      "learning_rate": 7.653214164760217e-06,
+      "loss": 0.4378,
+      "step": 52030
+    },
+    {
+      "epoch": 5.2123503781239044,
+      "grad_norm": 2.464893341064453,
+      "learning_rate": 7.645111888187967e-06,
+      "loss": 0.3915,
+      "step": 52040
+    },
+    {
+      "epoch": 5.213352030850904,
+      "grad_norm": 1.6739379167556763,
+      "learning_rate": 7.637013128515966e-06,
+      "loss": 0.3769,
+      "step": 52050
+    },
+    {
+      "epoch": 5.214353683577904,
+      "grad_norm": 1.666397213935852,
+      "learning_rate": 7.6289178873854086e-06,
+      "loss": 0.427,
+      "step": 52060
+    },
+    {
+      "epoch": 5.215355336304903,
+      "grad_norm": 2.2799434661865234,
+      "learning_rate": 7.62082616643677e-06,
+      "loss": 0.3725,
+      "step": 52070
+    },
+    {
+      "epoch": 5.216356989031903,
+      "grad_norm": 2.4019501209259033,
+      "learning_rate": 7.612737967309777e-06,
+      "loss": 0.3981,
+      "step": 52080
+    },
+    {
+      "epoch": 5.217358641758902,
+      "grad_norm": 2.1799919605255127,
+      "learning_rate": 7.604653291643496e-06,
+      "loss": 0.4186,
+      "step": 52090
+    },
+    {
+      "epoch": 5.218360294485902,
+      "grad_norm": 2.0130274295806885,
+      "learning_rate": 7.59657214107625e-06,
+      "loss": 0.4132,
+      "step": 52100
+    },
+    {
+      "epoch": 5.219361947212901,
+      "grad_norm": 1.9630812406539917,
+      "learning_rate": 7.588494517245656e-06,
+      "loss": 0.351,
+      "step": 52110
+    },
+    {
+      "epoch": 5.220363599939901,
+      "grad_norm": 2.2186408042907715,
+      "learning_rate": 7.5804204217885925e-06,
+      "loss": 0.4274,
+      "step": 52120
+    },
+    {
+      "epoch": 5.2213652526669,
+      "grad_norm": 1.6884628534317017,
+      "learning_rate": 7.572349856341265e-06,
+      "loss": 0.3911,
+      "step": 52130
+    },
+    {
+      "epoch": 5.2223669053939,
+      "grad_norm": 2.2494137287139893,
+      "learning_rate": 7.564282822539143e-06,
+      "loss": 0.4608,
+      "step": 52140
+    },
+    {
+      "epoch": 5.2233685581208995,
+      "grad_norm": 2.3293933868408203,
+      "learning_rate": 7.556219322016958e-06,
+      "loss": 0.4025,
+      "step": 52150
+    },
+    {
+      "epoch": 5.224370210847899,
+      "grad_norm": 2.4375548362731934,
+      "learning_rate": 7.548159356408766e-06,
+      "loss": 0.4536,
+      "step": 52160
+    },
+    {
+      "epoch": 5.225371863574899,
+      "grad_norm": 1.501153826713562,
+      "learning_rate": 7.540102927347883e-06,
+      "loss": 0.393,
+      "step": 52170
+    },
+    {
+      "epoch": 5.226373516301898,
+      "grad_norm": 2.5104174613952637,
+      "learning_rate": 7.53205003646692e-06,
+      "loss": 0.4326,
+      "step": 52180
+    },
+    {
+      "epoch": 5.227375169028898,
+      "grad_norm": 2.1039373874664307,
+      "learning_rate": 7.524000685397739e-06,
+      "loss": 0.4187,
+      "step": 52190
+    },
+    {
+      "epoch": 5.228376821755897,
+      "grad_norm": 2.2166614532470703,
+      "learning_rate": 7.515954875771533e-06,
+      "loss": 0.4832,
+      "step": 52200
+    },
+    {
+      "epoch": 5.229378474482897,
+      "grad_norm": 2.8615593910217285,
+      "learning_rate": 7.507912609218759e-06,
+      "loss": 0.3768,
+      "step": 52210
+    },
+    {
+      "epoch": 5.230380127209896,
+      "grad_norm": 2.372994899749756,
+      "learning_rate": 7.499873887369119e-06,
+      "loss": 0.4151,
+      "step": 52220
+    },
+    {
+      "epoch": 5.231381779936896,
+      "grad_norm": 2.192148447036743,
+      "learning_rate": 7.491838711851659e-06,
+      "loss": 0.4568,
+      "step": 52230
+    },
+    {
+      "epoch": 5.232383432663895,
+      "grad_norm": 2.341484785079956,
+      "learning_rate": 7.483807084294664e-06,
+      "loss": 0.3927,
+      "step": 52240
+    },
+    {
+      "epoch": 5.2333850853908945,
+      "grad_norm": 1.8965469598770142,
+      "learning_rate": 7.475779006325723e-06,
+      "loss": 0.4777,
+      "step": 52250
+    },
+    {
+      "epoch": 5.234386738117895,
+      "grad_norm": 2.039476156234741,
+      "learning_rate": 7.467754479571667e-06,
+      "loss": 0.4218,
+      "step": 52260
+    },
+    {
+      "epoch": 5.235388390844894,
+      "grad_norm": 1.3191801309585571,
+      "learning_rate": 7.459733505658661e-06,
+      "loss": 0.3756,
+      "step": 52270
+    },
+    {
+      "epoch": 5.236390043571894,
+      "grad_norm": 2.1452765464782715,
+      "learning_rate": 7.4517160862121185e-06,
+      "loss": 0.4151,
+      "step": 52280
+    },
+    {
+      "epoch": 5.237391696298893,
+      "grad_norm": 1.9325929880142212,
+      "learning_rate": 7.443702222856735e-06,
+      "loss": 0.3947,
+      "step": 52290
+    },
+    {
+      "epoch": 5.238393349025893,
+      "grad_norm": 2.707630157470703,
+      "learning_rate": 7.435691917216489e-06,
+      "loss": 0.4178,
+      "step": 52300
+    },
+    {
+      "epoch": 5.239395001752892,
+      "grad_norm": 2.664762020111084,
+      "learning_rate": 7.427685170914636e-06,
+      "loss": 0.3769,
+      "step": 52310
+    },
+    {
+      "epoch": 5.240396654479892,
+      "grad_norm": 1.8347071409225464,
+      "learning_rate": 7.4196819855737255e-06,
+      "loss": 0.4197,
+      "step": 52320
+    },
+    {
+      "epoch": 5.241398307206891,
+      "grad_norm": 1.889101266860962,
+      "learning_rate": 7.411682362815542e-06,
+      "loss": 0.411,
+      "step": 52330
+    },
+    {
+      "epoch": 5.242399959933891,
+      "grad_norm": 3.006286144256592,
+      "learning_rate": 7.403686304261204e-06,
+      "loss": 0.4188,
+      "step": 52340
+    },
+    {
+      "epoch": 5.24340161266089,
+      "grad_norm": 2.348801851272583,
+      "learning_rate": 7.3956938115310734e-06,
+      "loss": 0.4278,
+      "step": 52350
+    },
+    {
+      "epoch": 5.24440326538789,
+      "grad_norm": 2.0983736515045166,
+      "learning_rate": 7.387704886244798e-06,
+      "loss": 0.4544,
+      "step": 52360
+    },
+    {
+      "epoch": 5.24540491811489,
+      "grad_norm": 3.7822070121765137,
+      "learning_rate": 7.3797195300213005e-06,
+      "loss": 0.4355,
+      "step": 52370
+    },
+    {
+      "epoch": 5.246406570841889,
+      "grad_norm": 2.2507216930389404,
+      "learning_rate": 7.371737744478785e-06,
+      "loss": 0.4098,
+      "step": 52380
+    },
+    {
+      "epoch": 5.247408223568889,
+      "grad_norm": 2.440915107727051,
+      "learning_rate": 7.363759531234729e-06,
+      "loss": 0.3955,
+      "step": 52390
+    },
+    {
+      "epoch": 5.248409876295888,
+      "grad_norm": 1.8399525880813599,
+      "learning_rate": 7.355784891905882e-06,
+      "loss": 0.3885,
+      "step": 52400
+    },
+    {
+      "epoch": 5.249411529022888,
+      "grad_norm": 2.2914631366729736,
+      "learning_rate": 7.347813828108277e-06,
+      "loss": 0.411,
+      "step": 52410
+    },
+    {
+      "epoch": 5.250413181749887,
+      "grad_norm": 2.0582337379455566,
+      "learning_rate": 7.339846341457221e-06,
+      "loss": 0.3832,
+      "step": 52420
+    },
+    {
+      "epoch": 5.251414834476887,
+      "grad_norm": 1.9887348413467407,
+      "learning_rate": 7.331882433567289e-06,
+      "loss": 0.4259,
+      "step": 52430
+    },
+    {
+      "epoch": 5.252416487203886,
+      "grad_norm": 2.355299472808838,
+      "learning_rate": 7.323922106052339e-06,
+      "loss": 0.3626,
+      "step": 52440
+    },
+    {
+      "epoch": 5.253418139930886,
+      "grad_norm": 2.0583131313323975,
+      "learning_rate": 7.315965360525498e-06,
+      "loss": 0.4364,
+      "step": 52450
+    },
+    {
+      "epoch": 5.2544197926578855,
+      "grad_norm": 2.732943296432495,
+      "learning_rate": 7.308012198599173e-06,
+      "loss": 0.4869,
+      "step": 52460
+    },
+    {
+      "epoch": 5.255421445384885,
+      "grad_norm": 2.507206439971924,
+      "learning_rate": 7.300062621885037e-06,
+      "loss": 0.4206,
+      "step": 52470
+    },
+    {
+      "epoch": 5.256423098111885,
+      "grad_norm": 2.7191555500030518,
+      "learning_rate": 7.292116631994045e-06,
+      "loss": 0.411,
+      "step": 52480
+    },
+    {
+      "epoch": 5.257424750838884,
+      "grad_norm": 1.7782527208328247,
+      "learning_rate": 7.284174230536417e-06,
+      "loss": 0.3589,
+      "step": 52490
+    },
+    {
+      "epoch": 5.258426403565884,
+      "grad_norm": 2.6875455379486084,
+      "learning_rate": 7.276235419121649e-06,
+      "loss": 0.3732,
+      "step": 52500
+    },
+    {
+      "epoch": 5.259428056292883,
+      "grad_norm": 2.146796464920044,
+      "learning_rate": 7.268300199358516e-06,
+      "loss": 0.3572,
+      "step": 52510
+    },
+    {
+      "epoch": 5.260429709019883,
+      "grad_norm": 2.5086498260498047,
+      "learning_rate": 7.260368572855053e-06,
+      "loss": 0.4919,
+      "step": 52520
+    },
+    {
+      "epoch": 5.261431361746882,
+      "grad_norm": 2.6334359645843506,
+      "learning_rate": 7.2524405412185775e-06,
+      "loss": 0.383,
+      "step": 52530
+    },
+    {
+      "epoch": 5.262433014473882,
+      "grad_norm": 2.3560729026794434,
+      "learning_rate": 7.244516106055671e-06,
+      "loss": 0.4374,
+      "step": 52540
+    },
+    {
+      "epoch": 5.263434667200881,
+      "grad_norm": 1.7651573419570923,
+      "learning_rate": 7.23659526897219e-06,
+      "loss": 0.3754,
+      "step": 52550
+    },
+    {
+      "epoch": 5.264436319927881,
+      "grad_norm": 2.338216781616211,
+      "learning_rate": 7.228678031573263e-06,
+      "loss": 0.3835,
+      "step": 52560
+    },
+    {
+      "epoch": 5.265437972654881,
+      "grad_norm": 2.0452308654785156,
+      "learning_rate": 7.2207643954632865e-06,
+      "loss": 0.4685,
+      "step": 52570
+    },
+    {
+      "epoch": 5.26643962538188,
+      "grad_norm": 1.7141194343566895,
+      "learning_rate": 7.212854362245924e-06,
+      "loss": 0.3844,
+      "step": 52580
+    },
+    {
+      "epoch": 5.26744127810888,
+      "grad_norm": 2.3181238174438477,
+      "learning_rate": 7.20494793352412e-06,
+      "loss": 0.4083,
+      "step": 52590
+    },
+    {
+      "epoch": 5.268442930835879,
+      "grad_norm": 2.631800889968872,
+      "learning_rate": 7.197045110900069e-06,
+      "loss": 0.359,
+      "step": 52600
+    },
+    {
+      "epoch": 5.269444583562879,
+      "grad_norm": 1.7754448652267456,
+      "learning_rate": 7.189145895975272e-06,
+      "loss": 0.3636,
+      "step": 52610
+    },
+    {
+      "epoch": 5.270446236289878,
+      "grad_norm": 2.115633964538574,
+      "learning_rate": 7.181250290350447e-06,
+      "loss": 0.3871,
+      "step": 52620
+    },
+    {
+      "epoch": 5.271447889016878,
+      "grad_norm": 2.290174961090088,
+      "learning_rate": 7.173358295625621e-06,
+      "loss": 0.4456,
+      "step": 52630
+    },
+    {
+      "epoch": 5.272449541743877,
+      "grad_norm": 2.639821767807007,
+      "learning_rate": 7.165469913400072e-06,
+      "loss": 0.4184,
+      "step": 52640
+    },
+    {
+      "epoch": 5.273451194470877,
+      "grad_norm": 2.0272226333618164,
+      "learning_rate": 7.1575851452723496e-06,
+      "loss": 0.4,
+      "step": 52650
+    },
+    {
+      "epoch": 5.274452847197876,
+      "grad_norm": 2.293170213699341,
+      "learning_rate": 7.149703992840276e-06,
+      "loss": 0.3766,
+      "step": 52660
+    },
+    {
+      "epoch": 5.2754544999248765,
+      "grad_norm": 3.4424662590026855,
+      "learning_rate": 7.141826457700923e-06,
+      "loss": 0.351,
+      "step": 52670
+    },
+    {
+      "epoch": 5.276456152651876,
+      "grad_norm": 1.8073848485946655,
+      "learning_rate": 7.133952541450669e-06,
+      "loss": 0.4577,
+      "step": 52680
+    },
+    {
+      "epoch": 5.277457805378875,
+      "grad_norm": 2.518890619277954,
+      "learning_rate": 7.126082245685106e-06,
+      "loss": 0.3885,
+      "step": 52690
+    },
+    {
+      "epoch": 5.278459458105875,
+      "grad_norm": 2.2708356380462646,
+      "learning_rate": 7.118215571999129e-06,
+      "loss": 0.3841,
+      "step": 52700
+    },
+    {
+      "epoch": 5.279461110832874,
+      "grad_norm": 1.920397162437439,
+      "learning_rate": 7.1103525219868795e-06,
+      "loss": 0.4228,
+      "step": 52710
+    },
+    {
+      "epoch": 5.280462763559874,
+      "grad_norm": 2.008230686187744,
+      "learning_rate": 7.102493097241797e-06,
+      "loss": 0.4011,
+      "step": 52720
+    },
+    {
+      "epoch": 5.281464416286873,
+      "grad_norm": 1.935886025428772,
+      "learning_rate": 7.094637299356544e-06,
+      "loss": 0.3789,
+      "step": 52730
+    },
+    {
+      "epoch": 5.282466069013873,
+      "grad_norm": 2.69047212600708,
+      "learning_rate": 7.0867851299230595e-06,
+      "loss": 0.4479,
+      "step": 52740
+    },
+    {
+      "epoch": 5.283467721740872,
+      "grad_norm": 1.8354910612106323,
+      "learning_rate": 7.078936590532584e-06,
+      "loss": 0.3983,
+      "step": 52750
+    },
+    {
+      "epoch": 5.284469374467872,
+      "grad_norm": 2.452239990234375,
+      "learning_rate": 7.071091682775569e-06,
+      "loss": 0.4333,
+      "step": 52760
+    },
+    {
+      "epoch": 5.2854710271948715,
+      "grad_norm": 2.979823112487793,
+      "learning_rate": 7.063250408241761e-06,
+      "loss": 0.427,
+      "step": 52770
+    },
+    {
+      "epoch": 5.2864726799218715,
+      "grad_norm": 2.0987963676452637,
+      "learning_rate": 7.055412768520156e-06,
+      "loss": 0.3653,
+      "step": 52780
+    },
+    {
+      "epoch": 5.287474332648871,
+      "grad_norm": 2.366989850997925,
+      "learning_rate": 7.047578765199048e-06,
+      "loss": 0.4078,
+      "step": 52790
+    },
+    {
+      "epoch": 5.28847598537587,
+      "grad_norm": 2.4074864387512207,
+      "learning_rate": 7.039748399865937e-06,
+      "loss": 0.4018,
+      "step": 52800
+    },
+    {
+      "epoch": 5.28947763810287,
+      "grad_norm": 2.3981125354766846,
+      "learning_rate": 7.031921674107622e-06,
+      "loss": 0.4609,
+      "step": 52810
+    },
+    {
+      "epoch": 5.290479290829869,
+      "grad_norm": 2.6520919799804688,
+      "learning_rate": 7.024098589510181e-06,
+      "loss": 0.4054,
+      "step": 52820
+    },
+    {
+      "epoch": 5.291480943556869,
+      "grad_norm": 2.526301145553589,
+      "learning_rate": 7.016279147658903e-06,
+      "loss": 0.4026,
+      "step": 52830
+    },
+    {
+      "epoch": 5.292482596283868,
+      "grad_norm": 1.9764511585235596,
+      "learning_rate": 7.008463350138381e-06,
+      "loss": 0.406,
+      "step": 52840
+    },
+    {
+      "epoch": 5.293484249010868,
+      "grad_norm": 2.1861753463745117,
+      "learning_rate": 7.000651198532446e-06,
+      "loss": 0.4219,
+      "step": 52850
+    },
+    {
+      "epoch": 5.294485901737867,
+      "grad_norm": 1.8499925136566162,
+      "learning_rate": 6.992842694424221e-06,
+      "loss": 0.3846,
+      "step": 52860
+    },
+    {
+      "epoch": 5.295487554464867,
+      "grad_norm": 2.1721975803375244,
+      "learning_rate": 6.9850378393960495e-06,
+      "loss": 0.4378,
+      "step": 52870
+    },
+    {
+      "epoch": 5.2964892071918666,
+      "grad_norm": 2.3228940963745117,
+      "learning_rate": 6.977236635029552e-06,
+      "loss": 0.4515,
+      "step": 52880
+    },
+    {
+      "epoch": 5.297490859918867,
+      "grad_norm": 2.2590081691741943,
+      "learning_rate": 6.96943908290563e-06,
+      "loss": 0.4124,
+      "step": 52890
+    },
+    {
+      "epoch": 5.298492512645866,
+      "grad_norm": 2.0431101322174072,
+      "learning_rate": 6.961645184604423e-06,
+      "loss": 0.3689,
+      "step": 52900
+    },
+    {
+      "epoch": 5.299494165372865,
+      "grad_norm": 1.8751521110534668,
+      "learning_rate": 6.953854941705323e-06,
+      "loss": 0.4161,
+      "step": 52910
+    },
+    {
+      "epoch": 5.300495818099865,
+      "grad_norm": 2.33186936378479,
+      "learning_rate": 6.946068355786992e-06,
+      "loss": 0.4703,
+      "step": 52920
+    },
+    {
+      "epoch": 5.301497470826864,
+      "grad_norm": 2.136018991470337,
+      "learning_rate": 6.9382854284273715e-06,
+      "loss": 0.3905,
+      "step": 52930
+    },
+    {
+      "epoch": 5.302499123553864,
+      "grad_norm": 1.790964126586914,
+      "learning_rate": 6.930506161203618e-06,
+      "loss": 0.4172,
+      "step": 52940
+    },
+    {
+      "epoch": 5.303500776280863,
+      "grad_norm": 1.893381118774414,
+      "learning_rate": 6.922730555692172e-06,
+      "loss": 0.4716,
+      "step": 52950
+    },
+    {
+      "epoch": 5.304502429007863,
+      "grad_norm": 2.6119513511657715,
+      "learning_rate": 6.914958613468744e-06,
+      "loss": 0.4113,
+      "step": 52960
+    },
+    {
+      "epoch": 5.305504081734862,
+      "grad_norm": 1.9697766304016113,
+      "learning_rate": 6.907190336108288e-06,
+      "loss": 0.419,
+      "step": 52970
+    },
+    {
+      "epoch": 5.3065057344618625,
+      "grad_norm": 2.429619312286377,
+      "learning_rate": 6.899425725184999e-06,
+      "loss": 0.3799,
+      "step": 52980
+    },
+    {
+      "epoch": 5.307507387188862,
+      "grad_norm": 2.6430892944335938,
+      "learning_rate": 6.891664782272347e-06,
+      "loss": 0.3845,
+      "step": 52990
+    },
+    {
+      "epoch": 5.308509039915861,
+      "grad_norm": 2.1307921409606934,
+      "learning_rate": 6.883907508943069e-06,
+      "loss": 0.3771,
+      "step": 53000
+    },
+    {
+      "epoch": 5.309510692642861,
+      "grad_norm": 2.107192277908325,
+      "learning_rate": 6.876153906769148e-06,
+      "loss": 0.4124,
+      "step": 53010
+    },
+    {
+      "epoch": 5.31051234536986,
+      "grad_norm": 2.150707721710205,
+      "learning_rate": 6.868403977321799e-06,
+      "loss": 0.3855,
+      "step": 53020
+    },
+    {
+      "epoch": 5.31151399809686,
+      "grad_norm": 2.0903549194335938,
+      "learning_rate": 6.860657722171534e-06,
+      "loss": 0.4535,
+      "step": 53030
+    },
+    {
+      "epoch": 5.312515650823859,
+      "grad_norm": 2.294050455093384,
+      "learning_rate": 6.852915142888108e-06,
+      "loss": 0.4359,
+      "step": 53040
+    },
+    {
+      "epoch": 5.313517303550859,
+      "grad_norm": 2.1415610313415527,
+      "learning_rate": 6.845176241040505e-06,
+      "loss": 0.4042,
+      "step": 53050
+    },
+    {
+      "epoch": 5.314518956277858,
+      "grad_norm": 2.3479254245758057,
+      "learning_rate": 6.83744101819698e-06,
+      "loss": 0.4422,
+      "step": 53060
+    },
+    {
+      "epoch": 5.315520609004858,
+      "grad_norm": 2.104673385620117,
+      "learning_rate": 6.8297094759250665e-06,
+      "loss": 0.3555,
+      "step": 53070
+    },
+    {
+      "epoch": 5.3165222617318575,
+      "grad_norm": 3.1974856853485107,
+      "learning_rate": 6.8219816157915286e-06,
+      "loss": 0.3773,
+      "step": 53080
+    },
+    {
+      "epoch": 5.3175239144588575,
+      "grad_norm": 2.344932794570923,
+      "learning_rate": 6.814257439362368e-06,
+      "loss": 0.3502,
+      "step": 53090
+    },
+    {
+      "epoch": 5.318525567185857,
+      "grad_norm": 2.2202727794647217,
+      "learning_rate": 6.806536948202874e-06,
+      "loss": 0.4425,
+      "step": 53100
+    },
+    {
+      "epoch": 5.319527219912856,
+      "grad_norm": 3.03460955619812,
+      "learning_rate": 6.798820143877574e-06,
+      "loss": 0.5061,
+      "step": 53110
+    },
+    {
+      "epoch": 5.320528872639856,
+      "grad_norm": 1.9732153415679932,
+      "learning_rate": 6.791107027950244e-06,
+      "loss": 0.3563,
+      "step": 53120
+    },
+    {
+      "epoch": 5.321530525366855,
+      "grad_norm": 2.2756731510162354,
+      "learning_rate": 6.783397601983918e-06,
+      "loss": 0.3928,
+      "step": 53130
+    },
+    {
+      "epoch": 5.322532178093855,
+      "grad_norm": 1.9389190673828125,
+      "learning_rate": 6.775691867540882e-06,
+      "loss": 0.4561,
+      "step": 53140
+    },
+    {
+      "epoch": 5.323533830820854,
+      "grad_norm": 2.05364727973938,
+      "learning_rate": 6.7679898261826775e-06,
+      "loss": 0.4577,
+      "step": 53150
+    },
+    {
+      "epoch": 5.324535483547854,
+      "grad_norm": 2.402559995651245,
+      "learning_rate": 6.760291479470074e-06,
+      "loss": 0.3989,
+      "step": 53160
+    },
+    {
+      "epoch": 5.325537136274853,
+      "grad_norm": 2.256415843963623,
+      "learning_rate": 6.752596828963132e-06,
+      "loss": 0.4353,
+      "step": 53170
+    },
+    {
+      "epoch": 5.326538789001853,
+      "grad_norm": 2.008063793182373,
+      "learning_rate": 6.744905876221133e-06,
+      "loss": 0.4673,
+      "step": 53180
+    },
+    {
+      "epoch": 5.3275404417288525,
+      "grad_norm": 1.8382278680801392,
+      "learning_rate": 6.737218622802621e-06,
+      "loss": 0.3791,
+      "step": 53190
+    },
+    {
+      "epoch": 5.328542094455852,
+      "grad_norm": 3.4138710498809814,
+      "learning_rate": 6.729535070265389e-06,
+      "loss": 0.4359,
+      "step": 53200
+    },
+    {
+      "epoch": 5.329543747182852,
+      "grad_norm": 2.198563575744629,
+      "learning_rate": 6.721855220166479e-06,
+      "loss": 0.4072,
+      "step": 53210
+    },
+    {
+      "epoch": 5.330545399909851,
+      "grad_norm": 3.2445294857025146,
+      "learning_rate": 6.71417907406218e-06,
+      "loss": 0.4526,
+      "step": 53220
+    },
+    {
+      "epoch": 5.331547052636851,
+      "grad_norm": 2.4834206104278564,
+      "learning_rate": 6.706506633508033e-06,
+      "loss": 0.4159,
+      "step": 53230
+    },
+    {
+      "epoch": 5.33254870536385,
+      "grad_norm": 2.6208958625793457,
+      "learning_rate": 6.69883790005883e-06,
+      "loss": 0.4058,
+      "step": 53240
+    },
+    {
+      "epoch": 5.33355035809085,
+      "grad_norm": 2.0602033138275146,
+      "learning_rate": 6.691172875268609e-06,
+      "loss": 0.3935,
+      "step": 53250
+    },
+    {
+      "epoch": 5.334552010817849,
+      "grad_norm": 2.335693359375,
+      "learning_rate": 6.683511560690658e-06,
+      "loss": 0.4759,
+      "step": 53260
+    },
+    {
+      "epoch": 5.335553663544849,
+      "grad_norm": 3.5803706645965576,
+      "learning_rate": 6.675853957877512e-06,
+      "loss": 0.4427,
+      "step": 53270
+    },
+    {
+      "epoch": 5.336555316271848,
+      "grad_norm": 1.9978817701339722,
+      "learning_rate": 6.668200068380953e-06,
+      "loss": 0.4553,
+      "step": 53280
+    },
+    {
+      "epoch": 5.3375569689988485,
+      "grad_norm": 2.189866065979004,
+      "learning_rate": 6.660549893752013e-06,
+      "loss": 0.4083,
+      "step": 53290
+    },
+    {
+      "epoch": 5.338558621725848,
+      "grad_norm": 1.6272187232971191,
+      "learning_rate": 6.652903435540972e-06,
+      "loss": 0.373,
+      "step": 53300
+    },
+    {
+      "epoch": 5.339560274452847,
+      "grad_norm": 2.4417366981506348,
+      "learning_rate": 6.6452606952973534e-06,
+      "loss": 0.4228,
+      "step": 53310
+    },
+    {
+      "epoch": 5.340561927179847,
+      "grad_norm": 2.1256210803985596,
+      "learning_rate": 6.637621674569925e-06,
+      "loss": 0.4603,
+      "step": 53320
+    },
+    {
+      "epoch": 5.341563579906846,
+      "grad_norm": 2.3436925411224365,
+      "learning_rate": 6.629986374906707e-06,
+      "loss": 0.4088,
+      "step": 53330
+    },
+    {
+      "epoch": 5.342565232633846,
+      "grad_norm": 1.9916789531707764,
+      "learning_rate": 6.622354797854965e-06,
+      "loss": 0.4398,
+      "step": 53340
+    },
+    {
+      "epoch": 5.343566885360845,
+      "grad_norm": 3.09892201423645,
+      "learning_rate": 6.614726944961208e-06,
+      "loss": 0.4802,
+      "step": 53350
+    },
+    {
+      "epoch": 5.344568538087845,
+      "grad_norm": 2.0518476963043213,
+      "learning_rate": 6.607102817771191e-06,
+      "loss": 0.4275,
+      "step": 53360
+    },
+    {
+      "epoch": 5.345570190814844,
+      "grad_norm": 2.4747159481048584,
+      "learning_rate": 6.599482417829908e-06,
+      "loss": 0.457,
+      "step": 53370
+    },
+    {
+      "epoch": 5.346571843541844,
+      "grad_norm": 2.272169351577759,
+      "learning_rate": 6.591865746681608e-06,
+      "loss": 0.3775,
+      "step": 53380
+    },
+    {
+      "epoch": 5.3475734962688435,
+      "grad_norm": 2.412099599838257,
+      "learning_rate": 6.584252805869781e-06,
+      "loss": 0.3946,
+      "step": 53390
+    },
+    {
+      "epoch": 5.3485751489958435,
+      "grad_norm": 2.5232436656951904,
+      "learning_rate": 6.576643596937157e-06,
+      "loss": 0.4018,
+      "step": 53400
+    },
+    {
+      "epoch": 5.349576801722843,
+      "grad_norm": 2.147260904312134,
+      "learning_rate": 6.569038121425711e-06,
+      "loss": 0.4361,
+      "step": 53410
+    },
+    {
+      "epoch": 5.350578454449842,
+      "grad_norm": 3.0588362216949463,
+      "learning_rate": 6.561436380876668e-06,
+      "loss": 0.4696,
+      "step": 53420
+    },
+    {
+      "epoch": 5.351580107176842,
+      "grad_norm": 2.22107195854187,
+      "learning_rate": 6.553838376830485e-06,
+      "loss": 0.3865,
+      "step": 53430
+    },
+    {
+      "epoch": 5.352581759903841,
+      "grad_norm": 1.650066614151001,
+      "learning_rate": 6.546244110826874e-06,
+      "loss": 0.3675,
+      "step": 53440
+    },
+    {
+      "epoch": 5.353583412630841,
+      "grad_norm": 2.833719253540039,
+      "learning_rate": 6.5386535844047775e-06,
+      "loss": 0.3676,
+      "step": 53450
+    },
+    {
+      "epoch": 5.35458506535784,
+      "grad_norm": 1.93025803565979,
+      "learning_rate": 6.53106679910239e-06,
+      "loss": 0.4342,
+      "step": 53460
+    },
+    {
+      "epoch": 5.35558671808484,
+      "grad_norm": 2.8691656589508057,
+      "learning_rate": 6.523483756457144e-06,
+      "loss": 0.4971,
+      "step": 53470
+    },
+    {
+      "epoch": 5.356588370811839,
+      "grad_norm": 2.115009069442749,
+      "learning_rate": 6.515904458005709e-06,
+      "loss": 0.417,
+      "step": 53480
+    },
+    {
+      "epoch": 5.357590023538839,
+      "grad_norm": 1.9475973844528198,
+      "learning_rate": 6.508328905284006e-06,
+      "loss": 0.3852,
+      "step": 53490
+    },
+    {
+      "epoch": 5.3585916762658385,
+      "grad_norm": 2.8739306926727295,
+      "learning_rate": 6.500757099827187e-06,
+      "loss": 0.3797,
+      "step": 53500
+    },
+    {
+      "epoch": 5.359593328992839,
+      "grad_norm": 2.277933359146118,
+      "learning_rate": 6.493189043169651e-06,
+      "loss": 0.3827,
+      "step": 53510
+    },
+    {
+      "epoch": 5.360594981719838,
+      "grad_norm": 2.5828804969787598,
+      "learning_rate": 6.485624736845031e-06,
+      "loss": 0.3898,
+      "step": 53520
+    },
+    {
+      "epoch": 5.361596634446837,
+      "grad_norm": 1.9495731592178345,
+      "learning_rate": 6.478064182386212e-06,
+      "loss": 0.406,
+      "step": 53530
+    },
+    {
+      "epoch": 5.362598287173837,
+      "grad_norm": 1.9667346477508545,
+      "learning_rate": 6.470507381325303e-06,
+      "loss": 0.4068,
+      "step": 53540
+    },
+    {
+      "epoch": 5.363599939900836,
+      "grad_norm": 1.973042368888855,
+      "learning_rate": 6.462954335193666e-06,
+      "loss": 0.4642,
+      "step": 53550
+    },
+    {
+      "epoch": 5.364601592627836,
+      "grad_norm": 2.1213204860687256,
+      "learning_rate": 6.455405045521892e-06,
+      "loss": 0.3665,
+      "step": 53560
+    },
+    {
+      "epoch": 5.365603245354835,
+      "grad_norm": 2.8752615451812744,
+      "learning_rate": 6.4478595138398185e-06,
+      "loss": 0.4114,
+      "step": 53570
+    },
+    {
+      "epoch": 5.366604898081835,
+      "grad_norm": 1.7899949550628662,
+      "learning_rate": 6.440317741676513e-06,
+      "loss": 0.3377,
+      "step": 53580
+    },
+    {
+      "epoch": 5.367606550808834,
+      "grad_norm": 2.2464749813079834,
+      "learning_rate": 6.432779730560292e-06,
+      "loss": 0.3849,
+      "step": 53590
+    },
+    {
+      "epoch": 5.3686082035358345,
+      "grad_norm": 2.21551513671875,
+      "learning_rate": 6.425245482018702e-06,
+      "loss": 0.3827,
+      "step": 53600
+    },
+    {
+      "epoch": 5.369609856262834,
+      "grad_norm": 2.1476612091064453,
+      "learning_rate": 6.41771499757852e-06,
+      "loss": 0.3988,
+      "step": 53610
+    },
+    {
+      "epoch": 5.370611508989834,
+      "grad_norm": 2.1985154151916504,
+      "learning_rate": 6.4101882787657916e-06,
+      "loss": 0.4127,
+      "step": 53620
+    },
+    {
+      "epoch": 5.371613161716833,
+      "grad_norm": 2.32558012008667,
+      "learning_rate": 6.402665327105756e-06,
+      "loss": 0.455,
+      "step": 53630
+    },
+    {
+      "epoch": 5.372614814443832,
+      "grad_norm": 1.937166690826416,
+      "learning_rate": 6.3951461441229185e-06,
+      "loss": 0.3793,
+      "step": 53640
+    },
+    {
+      "epoch": 5.373616467170832,
+      "grad_norm": 3.1271660327911377,
+      "learning_rate": 6.38763073134101e-06,
+      "loss": 0.4362,
+      "step": 53650
+    },
+    {
+      "epoch": 5.374618119897831,
+      "grad_norm": 2.246852159500122,
+      "learning_rate": 6.3801190902829985e-06,
+      "loss": 0.3768,
+      "step": 53660
+    },
+    {
+      "epoch": 5.375619772624831,
+      "grad_norm": 2.1570510864257812,
+      "learning_rate": 6.372611222471092e-06,
+      "loss": 0.3875,
+      "step": 53670
+    },
+    {
+      "epoch": 5.37662142535183,
+      "grad_norm": 2.5383236408233643,
+      "learning_rate": 6.365107129426723e-06,
+      "loss": 0.3872,
+      "step": 53680
+    },
+    {
+      "epoch": 5.37762307807883,
+      "grad_norm": 2.381948709487915,
+      "learning_rate": 6.3576068126705855e-06,
+      "loss": 0.3625,
+      "step": 53690
+    },
+    {
+      "epoch": 5.3786247308058295,
+      "grad_norm": 2.5966947078704834,
+      "learning_rate": 6.350110273722571e-06,
+      "loss": 0.4216,
+      "step": 53700
+    },
+    {
+      "epoch": 5.3796263835328295,
+      "grad_norm": 2.2546310424804688,
+      "learning_rate": 6.342617514101826e-06,
+      "loss": 0.4408,
+      "step": 53710
+    },
+    {
+      "epoch": 5.380628036259829,
+      "grad_norm": 2.3283772468566895,
+      "learning_rate": 6.335128535326726e-06,
+      "loss": 0.4264,
+      "step": 53720
+    },
+    {
+      "epoch": 5.381629688986829,
+      "grad_norm": 2.2052276134490967,
+      "learning_rate": 6.3276433389149045e-06,
+      "loss": 0.436,
+      "step": 53730
+    },
+    {
+      "epoch": 5.382631341713828,
+      "grad_norm": 2.5837795734405518,
+      "learning_rate": 6.320161926383186e-06,
+      "loss": 0.423,
+      "step": 53740
+    },
+    {
+      "epoch": 5.383632994440827,
+      "grad_norm": 1.8887791633605957,
+      "learning_rate": 6.312684299247648e-06,
+      "loss": 0.3555,
+      "step": 53750
+    },
+    {
+      "epoch": 5.384634647167827,
+      "grad_norm": 2.3950350284576416,
+      "learning_rate": 6.305210459023625e-06,
+      "loss": 0.4179,
+      "step": 53760
+    },
+    {
+      "epoch": 5.385636299894826,
+      "grad_norm": 3.177776336669922,
+      "learning_rate": 6.297740407225638e-06,
+      "loss": 0.4403,
+      "step": 53770
+    },
+    {
+      "epoch": 5.386637952621826,
+      "grad_norm": 2.376591920852661,
+      "learning_rate": 6.290274145367475e-06,
+      "loss": 0.5005,
+      "step": 53780
+    },
+    {
+      "epoch": 5.387639605348825,
+      "grad_norm": 2.027867078781128,
+      "learning_rate": 6.282811674962135e-06,
+      "loss": 0.3684,
+      "step": 53790
+    },
+    {
+      "epoch": 5.388641258075825,
+      "grad_norm": 2.313084602355957,
+      "learning_rate": 6.27535299752188e-06,
+      "loss": 0.3768,
+      "step": 53800
+    },
+    {
+      "epoch": 5.3896429108028245,
+      "grad_norm": 2.342231273651123,
+      "learning_rate": 6.267898114558157e-06,
+      "loss": 0.4171,
+      "step": 53810
+    },
+    {
+      "epoch": 5.390644563529825,
+      "grad_norm": 2.253715991973877,
+      "learning_rate": 6.260447027581676e-06,
+      "loss": 0.4569,
+      "step": 53820
+    },
+    {
+      "epoch": 5.391646216256824,
+      "grad_norm": 2.658900022506714,
+      "learning_rate": 6.252999738102381e-06,
+      "loss": 0.436,
+      "step": 53830
+    },
+    {
+      "epoch": 5.392647868983824,
+      "grad_norm": 2.1993579864501953,
+      "learning_rate": 6.245556247629436e-06,
+      "loss": 0.3874,
+      "step": 53840
+    },
+    {
+      "epoch": 5.393649521710823,
+      "grad_norm": 2.270911931991577,
+      "learning_rate": 6.238116557671217e-06,
+      "loss": 0.3653,
+      "step": 53850
+    },
+    {
+      "epoch": 5.394651174437822,
+      "grad_norm": 2.0035555362701416,
+      "learning_rate": 6.230680669735361e-06,
+      "loss": 0.3763,
+      "step": 53860
+    },
+    {
+      "epoch": 5.395652827164822,
+      "grad_norm": 3.234086513519287,
+      "learning_rate": 6.223248585328734e-06,
+      "loss": 0.4419,
+      "step": 53870
+    },
+    {
+      "epoch": 5.396654479891821,
+      "grad_norm": 2.189328193664551,
+      "learning_rate": 6.215820305957393e-06,
+      "loss": 0.389,
+      "step": 53880
+    },
+    {
+      "epoch": 5.397656132618821,
+      "grad_norm": 2.4266152381896973,
+      "learning_rate": 6.208395833126657e-06,
+      "loss": 0.3803,
+      "step": 53890
+    },
+    {
+      "epoch": 5.39865778534582,
+      "grad_norm": 2.4305834770202637,
+      "learning_rate": 6.200975168341081e-06,
+      "loss": 0.3779,
+      "step": 53900
+    },
+    {
+      "epoch": 5.3996594380728204,
+      "grad_norm": 2.4199655055999756,
+      "learning_rate": 6.193558313104425e-06,
+      "loss": 0.417,
+      "step": 53910
+    },
+    {
+      "epoch": 5.40066109079982,
+      "grad_norm": 2.372927188873291,
+      "learning_rate": 6.186145268919677e-06,
+      "loss": 0.4141,
+      "step": 53920
+    },
+    {
+      "epoch": 5.40166274352682,
+      "grad_norm": 2.0953314304351807,
+      "learning_rate": 6.178736037289074e-06,
+      "loss": 0.394,
+      "step": 53930
+    },
+    {
+      "epoch": 5.402664396253819,
+      "grad_norm": 2.165076971054077,
+      "learning_rate": 6.1713306197140605e-06,
+      "loss": 0.3865,
+      "step": 53940
+    },
+    {
+      "epoch": 5.403666048980819,
+      "grad_norm": 2.225846290588379,
+      "learning_rate": 6.163929017695328e-06,
+      "loss": 0.3702,
+      "step": 53950
+    },
+    {
+      "epoch": 5.404667701707818,
+      "grad_norm": 2.3943235874176025,
+      "learning_rate": 6.156531232732756e-06,
+      "loss": 0.444,
+      "step": 53960
+    },
+    {
+      "epoch": 5.405669354434817,
+      "grad_norm": 2.743006706237793,
+      "learning_rate": 6.1491372663255025e-06,
+      "loss": 0.4253,
+      "step": 53970
+    },
+    {
+      "epoch": 5.406671007161817,
+      "grad_norm": 2.2035491466522217,
+      "learning_rate": 6.141747119971925e-06,
+      "loss": 0.4603,
+      "step": 53980
+    },
+    {
+      "epoch": 5.407672659888816,
+      "grad_norm": 2.0817651748657227,
+      "learning_rate": 6.1343607951695805e-06,
+      "loss": 0.4126,
+      "step": 53990
+    },
+    {
+      "epoch": 5.408674312615816,
+      "grad_norm": 2.3498198986053467,
+      "learning_rate": 6.126978293415306e-06,
+      "loss": 0.3927,
+      "step": 54000
+    },
+    {
+      "epoch": 5.4096759653428155,
+      "grad_norm": 2.3940587043762207,
+      "learning_rate": 6.1195996162051295e-06,
+      "loss": 0.4416,
+      "step": 54010
+    },
+    {
+      "epoch": 5.4106776180698155,
+      "grad_norm": 1.8981224298477173,
+      "learning_rate": 6.112224765034316e-06,
+      "loss": 0.3833,
+      "step": 54020
+    },
+    {
+      "epoch": 5.411679270796815,
+      "grad_norm": 2.426468849182129,
+      "learning_rate": 6.104853741397332e-06,
+      "loss": 0.4342,
+      "step": 54030
+    },
+    {
+      "epoch": 5.412680923523815,
+      "grad_norm": 3.046752452850342,
+      "learning_rate": 6.097486546787903e-06,
+      "loss": 0.4214,
+      "step": 54040
+    },
+    {
+      "epoch": 5.413682576250814,
+      "grad_norm": 1.7482635974884033,
+      "learning_rate": 6.090123182698965e-06,
+      "loss": 0.4063,
+      "step": 54050
+    },
+    {
+      "epoch": 5.414684228977813,
+      "grad_norm": 2.8719656467437744,
+      "learning_rate": 6.082763650622655e-06,
+      "loss": 0.4556,
+      "step": 54060
+    },
+    {
+      "epoch": 5.415685881704813,
+      "grad_norm": 2.3082518577575684,
+      "learning_rate": 6.075407952050374e-06,
+      "loss": 0.4332,
+      "step": 54070
+    },
+    {
+      "epoch": 5.416687534431812,
+      "grad_norm": 1.7820711135864258,
+      "learning_rate": 6.068056088472715e-06,
+      "loss": 0.3846,
+      "step": 54080
+    },
+    {
+      "epoch": 5.417689187158812,
+      "grad_norm": 2.5791473388671875,
+      "learning_rate": 6.060708061379516e-06,
+      "loss": 0.4377,
+      "step": 54090
+    },
+    {
+      "epoch": 5.418690839885811,
+      "grad_norm": 2.1430938243865967,
+      "learning_rate": 6.053363872259802e-06,
+      "loss": 0.3801,
+      "step": 54100
+    },
+    {
+      "epoch": 5.419692492612811,
+      "grad_norm": 2.625718355178833,
+      "learning_rate": 6.046023522601868e-06,
+      "loss": 0.3715,
+      "step": 54110
+    },
+    {
+      "epoch": 5.4206941453398105,
+      "grad_norm": 2.7916579246520996,
+      "learning_rate": 6.038687013893199e-06,
+      "loss": 0.3915,
+      "step": 54120
+    },
+    {
+      "epoch": 5.421695798066811,
+      "grad_norm": 1.9555777311325073,
+      "learning_rate": 6.031354347620508e-06,
+      "loss": 0.3917,
+      "step": 54130
+    },
+    {
+      "epoch": 5.42269745079381,
+      "grad_norm": 2.3919904232025146,
+      "learning_rate": 6.024025525269733e-06,
+      "loss": 0.4028,
+      "step": 54140
+    },
+    {
+      "epoch": 5.42369910352081,
+      "grad_norm": 2.119253635406494,
+      "learning_rate": 6.016700548326029e-06,
+      "loss": 0.3997,
+      "step": 54150
+    },
+    {
+      "epoch": 5.424700756247809,
+      "grad_norm": 1.9236420392990112,
+      "learning_rate": 6.0093794182737866e-06,
+      "loss": 0.4038,
+      "step": 54160
+    },
+    {
+      "epoch": 5.425702408974808,
+      "grad_norm": 1.5495247840881348,
+      "learning_rate": 6.002062136596578e-06,
+      "loss": 0.3885,
+      "step": 54170
+    },
+    {
+      "epoch": 5.426704061701808,
+      "grad_norm": 2.620678424835205,
+      "learning_rate": 5.9947487047772425e-06,
+      "loss": 0.4028,
+      "step": 54180
+    },
+    {
+      "epoch": 5.427705714428807,
+      "grad_norm": 2.5425150394439697,
+      "learning_rate": 5.987439124297814e-06,
+      "loss": 0.4059,
+      "step": 54190
+    },
+    {
+      "epoch": 5.428707367155807,
+      "grad_norm": 2.143843173980713,
+      "learning_rate": 5.980133396639551e-06,
+      "loss": 0.4257,
+      "step": 54200
+    },
+    {
+      "epoch": 5.429709019882806,
+      "grad_norm": 2.344255208969116,
+      "learning_rate": 5.972831523282927e-06,
+      "loss": 0.4211,
+      "step": 54210
+    },
+    {
+      "epoch": 5.4307106726098064,
+      "grad_norm": 1.8500442504882812,
+      "learning_rate": 5.965533505707641e-06,
+      "loss": 0.3479,
+      "step": 54220
+    },
+    {
+      "epoch": 5.431712325336806,
+      "grad_norm": 1.6079330444335938,
+      "learning_rate": 5.958239345392605e-06,
+      "loss": 0.3915,
+      "step": 54230
+    },
+    {
+      "epoch": 5.432713978063806,
+      "grad_norm": 2.237523317337036,
+      "learning_rate": 5.950949043815956e-06,
+      "loss": 0.4472,
+      "step": 54240
+    },
+    {
+      "epoch": 5.433715630790805,
+      "grad_norm": 1.8473663330078125,
+      "learning_rate": 5.943662602455044e-06,
+      "loss": 0.3923,
+      "step": 54250
+    },
+    {
+      "epoch": 5.434717283517804,
+      "grad_norm": 2.521827220916748,
+      "learning_rate": 5.936380022786436e-06,
+      "loss": 0.4243,
+      "step": 54260
+    },
+    {
+      "epoch": 5.435718936244804,
+      "grad_norm": 1.908718466758728,
+      "learning_rate": 5.929101306285919e-06,
+      "loss": 0.3434,
+      "step": 54270
+    },
+    {
+      "epoch": 5.436720588971803,
+      "grad_norm": 2.581207752227783,
+      "learning_rate": 5.9218264544285e-06,
+      "loss": 0.4207,
+      "step": 54280
+    },
+    {
+      "epoch": 5.437722241698803,
+      "grad_norm": 2.1982924938201904,
+      "learning_rate": 5.914555468688393e-06,
+      "loss": 0.3784,
+      "step": 54290
+    },
+    {
+      "epoch": 5.438723894425802,
+      "grad_norm": 2.5535848140716553,
+      "learning_rate": 5.9072883505390395e-06,
+      "loss": 0.4427,
+      "step": 54300
+    },
+    {
+      "epoch": 5.439725547152802,
+      "grad_norm": 2.093137741088867,
+      "learning_rate": 5.900025101453089e-06,
+      "loss": 0.4415,
+      "step": 54310
+    },
+    {
+      "epoch": 5.4407271998798015,
+      "grad_norm": 2.2132139205932617,
+      "learning_rate": 5.892765722902413e-06,
+      "loss": 0.4316,
+      "step": 54320
+    },
+    {
+      "epoch": 5.4417288526068015,
+      "grad_norm": 1.9173887968063354,
+      "learning_rate": 5.885510216358098e-06,
+      "loss": 0.4156,
+      "step": 54330
+    },
+    {
+      "epoch": 5.442730505333801,
+      "grad_norm": 1.8375977277755737,
+      "learning_rate": 5.878258583290441e-06,
+      "loss": 0.3712,
+      "step": 54340
+    },
+    {
+      "epoch": 5.443732158060801,
+      "grad_norm": 2.173393726348877,
+      "learning_rate": 5.871010825168957e-06,
+      "loss": 0.3781,
+      "step": 54350
+    },
+    {
+      "epoch": 5.4447338107878,
+      "grad_norm": 2.204589605331421,
+      "learning_rate": 5.863766943462379e-06,
+      "loss": 0.3748,
+      "step": 54360
+    },
+    {
+      "epoch": 5.445735463514799,
+      "grad_norm": 2.3114230632781982,
+      "learning_rate": 5.856526939638646e-06,
+      "loss": 0.3875,
+      "step": 54370
+    },
+    {
+      "epoch": 5.446737116241799,
+      "grad_norm": 2.2886545658111572,
+      "learning_rate": 5.849290815164921e-06,
+      "loss": 0.4018,
+      "step": 54380
+    },
+    {
+      "epoch": 5.447738768968798,
+      "grad_norm": 2.2720205783843994,
+      "learning_rate": 5.8420585715075744e-06,
+      "loss": 0.4316,
+      "step": 54390
+    },
+    {
+      "epoch": 5.448740421695798,
+      "grad_norm": 2.6107401847839355,
+      "learning_rate": 5.8348302101321944e-06,
+      "loss": 0.435,
+      "step": 54400
+    },
+    {
+      "epoch": 5.449742074422797,
+      "grad_norm": 2.0635311603546143,
+      "learning_rate": 5.827605732503577e-06,
+      "loss": 0.4727,
+      "step": 54410
+    },
+    {
+      "epoch": 5.450743727149797,
+      "grad_norm": 2.109706401824951,
+      "learning_rate": 5.820385140085735e-06,
+      "loss": 0.3891,
+      "step": 54420
+    },
+    {
+      "epoch": 5.4517453798767965,
+      "grad_norm": 2.1004483699798584,
+      "learning_rate": 5.813168434341898e-06,
+      "loss": 0.4171,
+      "step": 54430
+    },
+    {
+      "epoch": 5.452747032603797,
+      "grad_norm": 2.433253765106201,
+      "learning_rate": 5.805955616734496e-06,
+      "loss": 0.4314,
+      "step": 54440
+    },
+    {
+      "epoch": 5.453748685330796,
+      "grad_norm": 1.5983706712722778,
+      "learning_rate": 5.798746688725182e-06,
+      "loss": 0.4282,
+      "step": 54450
+    },
+    {
+      "epoch": 5.454750338057796,
+      "grad_norm": 1.6324114799499512,
+      "learning_rate": 5.791541651774817e-06,
+      "loss": 0.4018,
+      "step": 54460
+    },
+    {
+      "epoch": 5.455751990784795,
+      "grad_norm": 2.1272900104522705,
+      "learning_rate": 5.784340507343472e-06,
+      "loss": 0.3705,
+      "step": 54470
+    },
+    {
+      "epoch": 5.456753643511794,
+      "grad_norm": 1.5174928903579712,
+      "learning_rate": 5.777143256890435e-06,
+      "loss": 0.3686,
+      "step": 54480
+    },
+    {
+      "epoch": 5.457755296238794,
+      "grad_norm": 2.0583887100219727,
+      "learning_rate": 5.769949901874194e-06,
+      "loss": 0.4044,
+      "step": 54490
+    },
+    {
+      "epoch": 5.458756948965793,
+      "grad_norm": 2.4094724655151367,
+      "learning_rate": 5.76276044375246e-06,
+      "loss": 0.3918,
+      "step": 54500
+    },
+    {
+      "epoch": 5.459758601692793,
+      "grad_norm": 3.136207103729248,
+      "learning_rate": 5.75557488398214e-06,
+      "loss": 0.374,
+      "step": 54510
+    },
+    {
+      "epoch": 5.460760254419792,
+      "grad_norm": 1.848882794380188,
+      "learning_rate": 5.74839322401938e-06,
+      "loss": 0.3436,
+      "step": 54520
+    },
+    {
+      "epoch": 5.461761907146792,
+      "grad_norm": 2.511730432510376,
+      "learning_rate": 5.741215465319494e-06,
+      "loss": 0.4469,
+      "step": 54530
+    },
+    {
+      "epoch": 5.462763559873792,
+      "grad_norm": 1.784735918045044,
+      "learning_rate": 5.7340416093370345e-06,
+      "loss": 0.3822,
+      "step": 54540
+    },
+    {
+      "epoch": 5.463765212600792,
+      "grad_norm": 2.1342051029205322,
+      "learning_rate": 5.726871657525751e-06,
+      "loss": 0.4132,
+      "step": 54550
+    },
+    {
+      "epoch": 5.464766865327791,
+      "grad_norm": 3.0335421562194824,
+      "learning_rate": 5.7197056113386215e-06,
+      "loss": 0.4171,
+      "step": 54560
+    },
+    {
+      "epoch": 5.465768518054791,
+      "grad_norm": 2.3769993782043457,
+      "learning_rate": 5.712543472227797e-06,
+      "loss": 0.3798,
+      "step": 54570
+    },
+    {
+      "epoch": 5.46677017078179,
+      "grad_norm": 2.6821632385253906,
+      "learning_rate": 5.705385241644662e-06,
+      "loss": 0.3973,
+      "step": 54580
+    },
+    {
+      "epoch": 5.467771823508789,
+      "grad_norm": 2.379179000854492,
+      "learning_rate": 5.698230921039821e-06,
+      "loss": 0.3875,
+      "step": 54590
+    },
+    {
+      "epoch": 5.468773476235789,
+      "grad_norm": 2.252995729446411,
+      "learning_rate": 5.691080511863051e-06,
+      "loss": 0.3976,
+      "step": 54600
+    },
+    {
+      "epoch": 5.469775128962788,
+      "grad_norm": 2.438122510910034,
+      "learning_rate": 5.683934015563358e-06,
+      "loss": 0.4273,
+      "step": 54610
+    },
+    {
+      "epoch": 5.470776781689788,
+      "grad_norm": 2.041632652282715,
+      "learning_rate": 5.676791433588946e-06,
+      "loss": 0.4138,
+      "step": 54620
+    },
+    {
+      "epoch": 5.4717784344167875,
+      "grad_norm": 2.2904317378997803,
+      "learning_rate": 5.669652767387254e-06,
+      "loss": 0.4218,
+      "step": 54630
+    },
+    {
+      "epoch": 5.4727800871437875,
+      "grad_norm": 2.181467294692993,
+      "learning_rate": 5.662518018404883e-06,
+      "loss": 0.3501,
+      "step": 54640
+    },
+    {
+      "epoch": 5.473781739870787,
+      "grad_norm": 1.8415119647979736,
+      "learning_rate": 5.6553871880876575e-06,
+      "loss": 0.3538,
+      "step": 54650
+    },
+    {
+      "epoch": 5.474783392597787,
+      "grad_norm": 2.321047782897949,
+      "learning_rate": 5.64826027788064e-06,
+      "loss": 0.4236,
+      "step": 54660
+    },
+    {
+      "epoch": 5.475785045324786,
+      "grad_norm": 2.017359972000122,
+      "learning_rate": 5.641137289228049e-06,
+      "loss": 0.4154,
+      "step": 54670
+    },
+    {
+      "epoch": 5.476786698051786,
+      "grad_norm": 2.092787504196167,
+      "learning_rate": 5.6340182235733315e-06,
+      "loss": 0.3638,
+      "step": 54680
+    },
+    {
+      "epoch": 5.477788350778785,
+      "grad_norm": 2.5516960620880127,
+      "learning_rate": 5.626903082359139e-06,
+      "loss": 0.4228,
+      "step": 54690
+    },
+    {
+      "epoch": 5.478790003505784,
+      "grad_norm": 2.765859603881836,
+      "learning_rate": 5.619791867027343e-06,
+      "loss": 0.3729,
+      "step": 54700
+    },
+    {
+      "epoch": 5.479791656232784,
+      "grad_norm": 1.8943431377410889,
+      "learning_rate": 5.612684579018984e-06,
+      "loss": 0.3901,
+      "step": 54710
+    },
+    {
+      "epoch": 5.480793308959783,
+      "grad_norm": 2.0830743312835693,
+      "learning_rate": 5.605581219774325e-06,
+      "loss": 0.449,
+      "step": 54720
+    },
+    {
+      "epoch": 5.481794961686783,
+      "grad_norm": 2.621382474899292,
+      "learning_rate": 5.598481790732851e-06,
+      "loss": 0.3602,
+      "step": 54730
+    },
+    {
+      "epoch": 5.4827966144137825,
+      "grad_norm": 2.1246063709259033,
+      "learning_rate": 5.591386293333231e-06,
+      "loss": 0.4236,
+      "step": 54740
+    },
+    {
+      "epoch": 5.483798267140783,
+      "grad_norm": 1.4775830507278442,
+      "learning_rate": 5.584294729013325e-06,
+      "loss": 0.3476,
+      "step": 54750
+    },
+    {
+      "epoch": 5.484799919867782,
+      "grad_norm": 2.423421859741211,
+      "learning_rate": 5.577207099210216e-06,
+      "loss": 0.3928,
+      "step": 54760
+    },
+    {
+      "epoch": 5.485801572594782,
+      "grad_norm": 2.1227428913116455,
+      "learning_rate": 5.570123405360198e-06,
+      "loss": 0.4252,
+      "step": 54770
+    },
+    {
+      "epoch": 5.486803225321781,
+      "grad_norm": 2.768158435821533,
+      "learning_rate": 5.563043648898738e-06,
+      "loss": 0.4303,
+      "step": 54780
+    },
+    {
+      "epoch": 5.487804878048781,
+      "grad_norm": 3.048062324523926,
+      "learning_rate": 5.55596783126052e-06,
+      "loss": 0.4353,
+      "step": 54790
+    },
+    {
+      "epoch": 5.48880653077578,
+      "grad_norm": 2.053222179412842,
+      "learning_rate": 5.548895953879443e-06,
+      "loss": 0.4145,
+      "step": 54800
+    },
+    {
+      "epoch": 5.489808183502779,
+      "grad_norm": 3.174201011657715,
+      "learning_rate": 5.541828018188599e-06,
+      "loss": 0.4039,
+      "step": 54810
+    },
+    {
+      "epoch": 5.490809836229779,
+      "grad_norm": 1.8603386878967285,
+      "learning_rate": 5.5347640256202595e-06,
+      "loss": 0.4172,
+      "step": 54820
+    },
+    {
+      "epoch": 5.491811488956778,
+      "grad_norm": 2.1023452281951904,
+      "learning_rate": 5.527703977605919e-06,
+      "loss": 0.4082,
+      "step": 54830
+    },
+    {
+      "epoch": 5.492813141683778,
+      "grad_norm": 2.3197402954101562,
+      "learning_rate": 5.520647875576279e-06,
+      "loss": 0.3994,
+      "step": 54840
+    },
+    {
+      "epoch": 5.493814794410778,
+      "grad_norm": 1.8299857378005981,
+      "learning_rate": 5.513595720961231e-06,
+      "loss": 0.3643,
+      "step": 54850
+    },
+    {
+      "epoch": 5.494816447137778,
+      "grad_norm": 1.9272540807724,
+      "learning_rate": 5.50654751518985e-06,
+      "loss": 0.4442,
+      "step": 54860
+    },
+    {
+      "epoch": 5.495818099864777,
+      "grad_norm": 2.175856113433838,
+      "learning_rate": 5.499503259690442e-06,
+      "loss": 0.3711,
+      "step": 54870
+    },
+    {
+      "epoch": 5.496819752591777,
+      "grad_norm": 2.690469741821289,
+      "learning_rate": 5.492462955890504e-06,
+      "loss": 0.4243,
+      "step": 54880
+    },
+    {
+      "epoch": 5.497821405318776,
+      "grad_norm": 2.87835693359375,
+      "learning_rate": 5.4854266052167065e-06,
+      "loss": 0.4216,
+      "step": 54890
+    },
+    {
+      "epoch": 5.498823058045776,
+      "grad_norm": 2.505415439605713,
+      "learning_rate": 5.478394209094942e-06,
+      "loss": 0.4195,
+      "step": 54900
+    },
+    {
+      "epoch": 5.499824710772775,
+      "grad_norm": 2.257429361343384,
+      "learning_rate": 5.471365768950313e-06,
+      "loss": 0.3791,
+      "step": 54910
+    },
+    {
+      "epoch": 5.500826363499774,
+      "grad_norm": 1.8191579580307007,
+      "learning_rate": 5.464341286207103e-06,
+      "loss": 0.4309,
+      "step": 54920
+    },
+    {
+      "epoch": 5.501828016226774,
+      "grad_norm": 2.082589626312256,
+      "learning_rate": 5.4573207622887755e-06,
+      "loss": 0.4079,
+      "step": 54930
+    },
+    {
+      "epoch": 5.5028296689537735,
+      "grad_norm": 2.1472620964050293,
+      "learning_rate": 5.450304198618034e-06,
+      "loss": 0.3812,
+      "step": 54940
+    },
+    {
+      "epoch": 5.5038313216807735,
+      "grad_norm": 2.3610379695892334,
+      "learning_rate": 5.443291596616748e-06,
+      "loss": 0.4051,
+      "step": 54950
+    },
+    {
+      "epoch": 5.504832974407773,
+      "grad_norm": 1.5136091709136963,
+      "learning_rate": 5.436282957706004e-06,
+      "loss": 0.3947,
+      "step": 54960
+    },
+    {
+      "epoch": 5.505834627134773,
+      "grad_norm": 2.1798224449157715,
+      "learning_rate": 5.429278283306055e-06,
+      "loss": 0.3891,
+      "step": 54970
+    },
+    {
+      "epoch": 5.506836279861772,
+      "grad_norm": 2.1945409774780273,
+      "learning_rate": 5.42227757483639e-06,
+      "loss": 0.3617,
+      "step": 54980
+    },
+    {
+      "epoch": 5.507837932588772,
+      "grad_norm": 1.7211298942565918,
+      "learning_rate": 5.415280833715675e-06,
+      "loss": 0.4423,
+      "step": 54990
+    },
+    {
+      "epoch": 5.508839585315771,
+      "grad_norm": 2.1400156021118164,
+      "learning_rate": 5.408288061361749e-06,
+      "loss": 0.4237,
+      "step": 55000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 69888,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 7,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.02157876235862e+21,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}