diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,26486 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9999404631963325,
+  "eval_steps": 500,
+  "global_step": 37791,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007938240488995614,
+      "grad_norm": 0.44355273246765137,
+      "learning_rate": 4.998676933661454e-05,
+      "loss": 1.5273,
+      "step": 10
+    },
+    {
+      "epoch": 0.0015876480977991228,
+      "grad_norm": 0.27582135796546936,
+      "learning_rate": 4.9973538673229073e-05,
+      "loss": 1.3011,
+      "step": 20
+    },
+    {
+      "epoch": 0.0023814721466986844,
+      "grad_norm": 0.3760926425457001,
+      "learning_rate": 4.996030800984361e-05,
+      "loss": 1.2377,
+      "step": 30
+    },
+    {
+      "epoch": 0.0031752961955982457,
+      "grad_norm": 0.2921426594257355,
+      "learning_rate": 4.994707734645815e-05,
+      "loss": 1.2327,
+      "step": 40
+    },
+    {
+      "epoch": 0.003969120244497807,
+      "grad_norm": 0.5147169828414917,
+      "learning_rate": 4.99338466830727e-05,
+      "loss": 1.1691,
+      "step": 50
+    },
+    {
+      "epoch": 0.004762944293397369,
+      "grad_norm": 0.38984984159469604,
+      "learning_rate": 4.992061601968723e-05,
+      "loss": 1.1292,
+      "step": 60
+    },
+    {
+      "epoch": 0.00555676834229693,
+      "grad_norm": 0.4045616090297699,
+      "learning_rate": 4.990738535630177e-05,
+      "loss": 1.0903,
+      "step": 70
+    },
+    {
+      "epoch": 0.006350592391196491,
+      "grad_norm": 0.4870094656944275,
+      "learning_rate": 4.989415469291631e-05,
+      "loss": 1.1421,
+      "step": 80
+    },
+    {
+      "epoch": 0.007144416440096053,
+      "grad_norm": 0.6016878485679626,
+      "learning_rate": 4.9880924029530846e-05,
+      "loss": 1.1459,
+      "step": 90
+    },
+    {
+      "epoch": 0.007938240488995614,
+      "grad_norm": 0.5868497490882874,
+      "learning_rate": 4.986769336614538e-05,
+      "loss": 1.0909,
+      "step": 100
+    },
+    {
+      "epoch": 0.008732064537895176,
+      "grad_norm": 0.6827491521835327,
+      "learning_rate": 4.985446270275992e-05,
+      "loss": 1.065,
+      "step": 110
+    },
+    {
+      "epoch": 0.009525888586794738,
+      "grad_norm": 0.7056289315223694,
+      "learning_rate": 4.9841232039374456e-05,
+      "loss": 1.0764,
+      "step": 120
+    },
+    {
+      "epoch": 0.010319712635694298,
+      "grad_norm": 0.5899627208709717,
+      "learning_rate": 4.9828001375988995e-05,
+      "loss": 1.0369,
+      "step": 130
+    },
+    {
+      "epoch": 0.01111353668459386,
+      "grad_norm": 0.6351290345191956,
+      "learning_rate": 4.9814770712603534e-05,
+      "loss": 1.0332,
+      "step": 140
+    },
+    {
+      "epoch": 0.011907360733493421,
+      "grad_norm": 0.7403615117073059,
+      "learning_rate": 4.980154004921807e-05,
+      "loss": 1.0837,
+      "step": 150
+    },
+    {
+      "epoch": 0.012701184782392983,
+      "grad_norm": 0.7318956255912781,
+      "learning_rate": 4.978830938583261e-05,
+      "loss": 1.1196,
+      "step": 160
+    },
+    {
+      "epoch": 0.013495008831292545,
+      "grad_norm": 0.8113905787467957,
+      "learning_rate": 4.9775078722447144e-05,
+      "loss": 1.0629,
+      "step": 170
+    },
+    {
+      "epoch": 0.014288832880192106,
+      "grad_norm": 0.7525221109390259,
+      "learning_rate": 4.976184805906168e-05,
+      "loss": 1.0937,
+      "step": 180
+    },
+    {
+      "epoch": 0.015082656929091666,
+      "grad_norm": 0.7905648350715637,
+      "learning_rate": 4.974861739567622e-05,
+      "loss": 1.0263,
+      "step": 190
+    },
+    {
+      "epoch": 0.015876480977991228,
+      "grad_norm": 0.47737765312194824,
+      "learning_rate": 4.973538673229076e-05,
+      "loss": 1.1193,
+      "step": 200
+    },
+    {
+      "epoch": 0.016670305026890788,
+      "grad_norm": 0.6782676577568054,
+      "learning_rate": 4.972215606890529e-05,
+      "loss": 1.0208,
+      "step": 210
+    },
+    {
+      "epoch": 0.01746412907579035,
+      "grad_norm": 0.7456789016723633,
+      "learning_rate": 4.970892540551983e-05,
+      "loss": 1.039,
+      "step": 220
+    },
+    {
+      "epoch": 0.01825795312468991,
+      "grad_norm": 0.8023837804794312,
+      "learning_rate": 4.969569474213438e-05,
+      "loss": 1.0291,
+      "step": 230
+    },
+    {
+      "epoch": 0.019051777173589475,
+      "grad_norm": 0.6566641926765442,
+      "learning_rate": 4.968246407874891e-05,
+      "loss": 1.071,
+      "step": 240
+    },
+    {
+      "epoch": 0.019845601222489035,
+      "grad_norm": 0.8695980906486511,
+      "learning_rate": 4.966923341536345e-05,
+      "loss": 0.9391,
+      "step": 250
+    },
+    {
+      "epoch": 0.020639425271388595,
+      "grad_norm": 0.7018496990203857,
+      "learning_rate": 4.965600275197799e-05,
+      "loss": 1.0663,
+      "step": 260
+    },
+    {
+      "epoch": 0.02143324932028816,
+      "grad_norm": 0.7785579562187195,
+      "learning_rate": 4.9642772088592526e-05,
+      "loss": 1.0047,
+      "step": 270
+    },
+    {
+      "epoch": 0.02222707336918772,
+      "grad_norm": 0.6475448608398438,
+      "learning_rate": 4.962954142520706e-05,
+      "loss": 1.0443,
+      "step": 280
+    },
+    {
+      "epoch": 0.023020897418087282,
+      "grad_norm": 0.8628908395767212,
+      "learning_rate": 4.96163107618216e-05,
+      "loss": 0.9987,
+      "step": 290
+    },
+    {
+      "epoch": 0.023814721466986842,
+      "grad_norm": 0.8528658747673035,
+      "learning_rate": 4.9603080098436136e-05,
+      "loss": 1.0199,
+      "step": 300
+    },
+    {
+      "epoch": 0.024608545515886402,
+      "grad_norm": 0.8548170328140259,
+      "learning_rate": 4.9589849435050675e-05,
+      "loss": 1.0916,
+      "step": 310
+    },
+    {
+      "epoch": 0.025402369564785966,
+      "grad_norm": 0.8923755884170532,
+      "learning_rate": 4.9576618771665214e-05,
+      "loss": 0.996,
+      "step": 320
+    },
+    {
+      "epoch": 0.026196193613685526,
+      "grad_norm": 0.8691526651382446,
+      "learning_rate": 4.956338810827975e-05,
+      "loss": 1.0162,
+      "step": 330
+    },
+    {
+      "epoch": 0.02699001766258509,
+      "grad_norm": 0.8156205415725708,
+      "learning_rate": 4.955015744489429e-05,
+      "loss": 1.0359,
+      "step": 340
+    },
+    {
+      "epoch": 0.02778384171148465,
+      "grad_norm": 0.8026768565177917,
+      "learning_rate": 4.9536926781508824e-05,
+      "loss": 1.0605,
+      "step": 350
+    },
+    {
+      "epoch": 0.028577665760384213,
+      "grad_norm": 0.9509242177009583,
+      "learning_rate": 4.952369611812336e-05,
+      "loss": 1.004,
+      "step": 360
+    },
+    {
+      "epoch": 0.029371489809283773,
+      "grad_norm": 0.8537798523902893,
+      "learning_rate": 4.95104654547379e-05,
+      "loss": 0.9256,
+      "step": 370
+    },
+    {
+      "epoch": 0.030165313858183333,
+      "grad_norm": 0.7341266870498657,
+      "learning_rate": 4.949723479135244e-05,
+      "loss": 0.95,
+      "step": 380
+    },
+    {
+      "epoch": 0.030959137907082896,
+      "grad_norm": 0.9486088752746582,
+      "learning_rate": 4.948400412796697e-05,
+      "loss": 0.9779,
+      "step": 390
+    },
+    {
+      "epoch": 0.031752961955982456,
+      "grad_norm": 0.5967279672622681,
+      "learning_rate": 4.947077346458152e-05,
+      "loss": 1.0431,
+      "step": 400
+    },
+    {
+      "epoch": 0.03254678600488202,
+      "grad_norm": 0.6559339165687561,
+      "learning_rate": 4.945754280119606e-05,
+      "loss": 1.0192,
+      "step": 410
+    },
+    {
+      "epoch": 0.033340610053781576,
+      "grad_norm": 0.8226445317268372,
+      "learning_rate": 4.94443121378106e-05,
+      "loss": 1.0024,
+      "step": 420
+    },
+    {
+      "epoch": 0.03413443410268114,
+      "grad_norm": 1.0217857360839844,
+      "learning_rate": 4.943108147442513e-05,
+      "loss": 1.0036,
+      "step": 430
+    },
+    {
+      "epoch": 0.0349282581515807,
+      "grad_norm": 0.9626277089118958,
+      "learning_rate": 4.941785081103967e-05,
+      "loss": 0.9955,
+      "step": 440
+    },
+    {
+      "epoch": 0.03572208220048027,
+      "grad_norm": 0.6774649620056152,
+      "learning_rate": 4.940462014765421e-05,
+      "loss": 0.9775,
+      "step": 450
+    },
+    {
+      "epoch": 0.03651590624937982,
+      "grad_norm": 0.8078123927116394,
+      "learning_rate": 4.9391389484268746e-05,
+      "loss": 0.9618,
+      "step": 460
+    },
+    {
+      "epoch": 0.03730973029827939,
+      "grad_norm": 1.0022040605545044,
+      "learning_rate": 4.937815882088328e-05,
+      "loss": 0.9787,
+      "step": 470
+    },
+    {
+      "epoch": 0.03810355434717895,
+      "grad_norm": 0.7025789618492126,
+      "learning_rate": 4.936492815749782e-05,
+      "loss": 0.9842,
+      "step": 480
+    },
+    {
+      "epoch": 0.03889737839607851,
+      "grad_norm": 0.6538853049278259,
+      "learning_rate": 4.9351697494112356e-05,
+      "loss": 0.988,
+      "step": 490
+    },
+    {
+      "epoch": 0.03969120244497807,
+      "grad_norm": 0.6774317026138306,
+      "learning_rate": 4.9338466830726895e-05,
+      "loss": 0.939,
+      "step": 500
+    },
+    {
+      "epoch": 0.040485026493877634,
+      "grad_norm": 0.9341694116592407,
+      "learning_rate": 4.9325236167341433e-05,
+      "loss": 1.0455,
+      "step": 510
+    },
+    {
+      "epoch": 0.04127885054277719,
+      "grad_norm": 0.8977088332176208,
+      "learning_rate": 4.931200550395597e-05,
+      "loss": 1.0065,
+      "step": 520
+    },
+    {
+      "epoch": 0.042072674591676754,
+      "grad_norm": 0.6621760725975037,
+      "learning_rate": 4.929877484057051e-05,
+      "loss": 1.0243,
+      "step": 530
+    },
+    {
+      "epoch": 0.04286649864057632,
+      "grad_norm": 0.8938838243484497,
+      "learning_rate": 4.9285544177185043e-05,
+      "loss": 1.0233,
+      "step": 540
+    },
+    {
+      "epoch": 0.04366032268947588,
+      "grad_norm": 0.7396982908248901,
+      "learning_rate": 4.927231351379958e-05,
+      "loss": 0.9628,
+      "step": 550
+    },
+    {
+      "epoch": 0.04445414673837544,
+      "grad_norm": 0.8066884279251099,
+      "learning_rate": 4.925908285041412e-05,
+      "loss": 1.0092,
+      "step": 560
+    },
+    {
+      "epoch": 0.045247970787275,
+      "grad_norm": 0.8251396417617798,
+      "learning_rate": 4.924585218702866e-05,
+      "loss": 1.0094,
+      "step": 570
+    },
+    {
+      "epoch": 0.046041794836174564,
+      "grad_norm": 0.8999793529510498,
+      "learning_rate": 4.92326215236432e-05,
+      "loss": 0.9971,
+      "step": 580
+    },
+    {
+      "epoch": 0.04683561888507412,
+      "grad_norm": 0.7802185416221619,
+      "learning_rate": 4.921939086025774e-05,
+      "loss": 0.9441,
+      "step": 590
+    },
+    {
+      "epoch": 0.047629442933973684,
+      "grad_norm": 0.9204962849617004,
+      "learning_rate": 4.920616019687228e-05,
+      "loss": 1.0266,
+      "step": 600
+    },
+    {
+      "epoch": 0.04842326698287325,
+      "grad_norm": 0.8698873519897461,
+      "learning_rate": 4.919292953348681e-05,
+      "loss": 0.9666,
+      "step": 610
+    },
+    {
+      "epoch": 0.049217091031772804,
+      "grad_norm": 0.7144138813018799,
+      "learning_rate": 4.917969887010135e-05,
+      "loss": 1.0201,
+      "step": 620
+    },
+    {
+      "epoch": 0.05001091508067237,
+      "grad_norm": 0.9176733493804932,
+      "learning_rate": 4.916646820671589e-05,
+      "loss": 0.9937,
+      "step": 630
+    },
+    {
+      "epoch": 0.05080473912957193,
+      "grad_norm": 0.8604631423950195,
+      "learning_rate": 4.9153237543330426e-05,
+      "loss": 0.9882,
+      "step": 640
+    },
+    {
+      "epoch": 0.051598563178471495,
+      "grad_norm": 1.0573097467422485,
+      "learning_rate": 4.914000687994496e-05,
+      "loss": 0.9997,
+      "step": 650
+    },
+    {
+      "epoch": 0.05239238722737105,
+      "grad_norm": 0.9698209166526794,
+      "learning_rate": 4.91267762165595e-05,
+      "loss": 0.9492,
+      "step": 660
+    },
+    {
+      "epoch": 0.053186211276270615,
+      "grad_norm": 0.8090052604675293,
+      "learning_rate": 4.911354555317404e-05,
+      "loss": 0.9408,
+      "step": 670
+    },
+    {
+      "epoch": 0.05398003532517018,
+      "grad_norm": 0.9131582975387573,
+      "learning_rate": 4.910031488978858e-05,
+      "loss": 0.92,
+      "step": 680
+    },
+    {
+      "epoch": 0.054773859374069735,
+      "grad_norm": 0.8041892051696777,
+      "learning_rate": 4.9087084226403114e-05,
+      "loss": 1.021,
+      "step": 690
+    },
+    {
+      "epoch": 0.0555676834229693,
+      "grad_norm": 1.0868831872940063,
+      "learning_rate": 4.907385356301765e-05,
+      "loss": 1.0078,
+      "step": 700
+    },
+    {
+      "epoch": 0.05636150747186886,
+      "grad_norm": 0.7588531374931335,
+      "learning_rate": 4.906062289963219e-05,
+      "loss": 0.9926,
+      "step": 710
+    },
+    {
+      "epoch": 0.057155331520768425,
+      "grad_norm": 0.7499911189079285,
+      "learning_rate": 4.9047392236246724e-05,
+      "loss": 1.0117,
+      "step": 720
+    },
+    {
+      "epoch": 0.05794915556966798,
+      "grad_norm": 0.8259711265563965,
+      "learning_rate": 4.903416157286126e-05,
+      "loss": 0.9394,
+      "step": 730
+    },
+    {
+      "epoch": 0.058742979618567545,
+      "grad_norm": 0.7790316939353943,
+      "learning_rate": 4.90209309094758e-05,
+      "loss": 0.9598,
+      "step": 740
+    },
+    {
+      "epoch": 0.05953680366746711,
+      "grad_norm": 0.8743465542793274,
+      "learning_rate": 4.900770024609034e-05,
+      "loss": 0.9337,
+      "step": 750
+    },
+    {
+      "epoch": 0.060330627716366665,
+      "grad_norm": 0.6910851001739502,
+      "learning_rate": 4.899446958270488e-05,
+      "loss": 1.0239,
+      "step": 760
+    },
+    {
+      "epoch": 0.06112445176526623,
+      "grad_norm": 0.8479220271110535,
+      "learning_rate": 4.898123891931942e-05,
+      "loss": 0.94,
+      "step": 770
+    },
+    {
+      "epoch": 0.06191827581416579,
+      "grad_norm": 0.9144072532653809,
+      "learning_rate": 4.896800825593396e-05,
+      "loss": 0.9813,
+      "step": 780
+    },
+    {
+      "epoch": 0.06271209986306535,
+      "grad_norm": 0.7847743034362793,
+      "learning_rate": 4.8954777592548496e-05,
+      "loss": 1.0148,
+      "step": 790
+    },
+    {
+      "epoch": 0.06350592391196491,
+      "grad_norm": 0.6444619297981262,
+      "learning_rate": 4.894154692916303e-05,
+      "loss": 1.0174,
+      "step": 800
+    },
+    {
+      "epoch": 0.06429974796086448,
+      "grad_norm": 0.8129200339317322,
+      "learning_rate": 4.892831626577757e-05,
+      "loss": 0.9764,
+      "step": 810
+    },
+    {
+      "epoch": 0.06509357200976404,
+      "grad_norm": 0.8406952619552612,
+      "learning_rate": 4.8915085602392106e-05,
+      "loss": 0.9332,
+      "step": 820
+    },
+    {
+      "epoch": 0.0658873960586636,
+      "grad_norm": 0.778441846370697,
+      "learning_rate": 4.890185493900664e-05,
+      "loss": 0.9439,
+      "step": 830
+    },
+    {
+      "epoch": 0.06668122010756315,
+      "grad_norm": 0.9417208433151245,
+      "learning_rate": 4.8888624275621184e-05,
+      "loss": 0.9337,
+      "step": 840
+    },
+    {
+      "epoch": 0.06747504415646272,
+      "grad_norm": 0.8156752586364746,
+      "learning_rate": 4.887539361223572e-05,
+      "loss": 0.9909,
+      "step": 850
+    },
+    {
+      "epoch": 0.06826886820536228,
+      "grad_norm": 0.8329022526741028,
+      "learning_rate": 4.886216294885026e-05,
+      "loss": 0.9374,
+      "step": 860
+    },
+    {
+      "epoch": 0.06906269225426184,
+      "grad_norm": 0.816589891910553,
+      "learning_rate": 4.8848932285464794e-05,
+      "loss": 1.0349,
+      "step": 870
+    },
+    {
+      "epoch": 0.0698565163031614,
+      "grad_norm": 0.7630716562271118,
+      "learning_rate": 4.883570162207933e-05,
+      "loss": 0.9716,
+      "step": 880
+    },
+    {
+      "epoch": 0.07065034035206097,
+      "grad_norm": 0.978765606880188,
+      "learning_rate": 4.882247095869387e-05,
+      "loss": 1.0316,
+      "step": 890
+    },
+    {
+      "epoch": 0.07144416440096053,
+      "grad_norm": 0.7739888429641724,
+      "learning_rate": 4.880924029530841e-05,
+      "loss": 0.9596,
+      "step": 900
+    },
+    {
+      "epoch": 0.07223798844986008,
+      "grad_norm": 0.9134480357170105,
+      "learning_rate": 4.879600963192294e-05,
+      "loss": 0.9758,
+      "step": 910
+    },
+    {
+      "epoch": 0.07303181249875965,
+      "grad_norm": 0.8623645305633545,
+      "learning_rate": 4.878277896853748e-05,
+      "loss": 1.001,
+      "step": 920
+    },
+    {
+      "epoch": 0.07382563654765921,
+      "grad_norm": 0.8770903944969177,
+      "learning_rate": 4.876954830515202e-05,
+      "loss": 0.9698,
+      "step": 930
+    },
+    {
+      "epoch": 0.07461946059655877,
+      "grad_norm": 0.6688238382339478,
+      "learning_rate": 4.875631764176656e-05,
+      "loss": 1.0244,
+      "step": 940
+    },
+    {
+      "epoch": 0.07541328464545834,
+      "grad_norm": 0.7384278178215027,
+      "learning_rate": 4.87430869783811e-05,
+      "loss": 0.9876,
+      "step": 950
+    },
+    {
+      "epoch": 0.0762071086943579,
+      "grad_norm": 0.8874367475509644,
+      "learning_rate": 4.872985631499564e-05,
+      "loss": 0.9847,
+      "step": 960
+    },
+    {
+      "epoch": 0.07700093274325746,
+      "grad_norm": 0.6085987687110901,
+      "learning_rate": 4.871662565161018e-05,
+      "loss": 0.9255,
+      "step": 970
+    },
+    {
+      "epoch": 0.07779475679215701,
+      "grad_norm": 0.7393110394477844,
+      "learning_rate": 4.870339498822471e-05,
+      "loss": 0.9081,
+      "step": 980
+    },
+    {
+      "epoch": 0.07858858084105658,
+      "grad_norm": 0.9497765302658081,
+      "learning_rate": 4.869016432483925e-05,
+      "loss": 0.9948,
+      "step": 990
+    },
+    {
+      "epoch": 0.07938240488995614,
+      "grad_norm": 0.7365654110908508,
+      "learning_rate": 4.867693366145379e-05,
+      "loss": 1.0194,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0801762289388557,
+      "grad_norm": 0.9192026257514954,
+      "learning_rate": 4.8663702998068326e-05,
+      "loss": 0.9777,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08097005298775527,
+      "grad_norm": 0.9804342985153198,
+      "learning_rate": 4.8650472334682865e-05,
+      "loss": 0.9756,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08176387703665483,
+      "grad_norm": 0.976823091506958,
+      "learning_rate": 4.8637241671297403e-05,
+      "loss": 0.9507,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08255770108555438,
+      "grad_norm": 0.6893647909164429,
+      "learning_rate": 4.862401100791194e-05,
+      "loss": 1.003,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08335152513445394,
+      "grad_norm": 0.8077487945556641,
+      "learning_rate": 4.8610780344526475e-05,
+      "loss": 0.944,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08414534918335351,
+      "grad_norm": 0.6170672178268433,
+      "learning_rate": 4.8597549681141013e-05,
+      "loss": 0.9685,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08493917323225307,
+      "grad_norm": 0.7847235798835754,
+      "learning_rate": 4.858431901775555e-05,
+      "loss": 0.9448,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08573299728115263,
+      "grad_norm": 0.7903896570205688,
+      "learning_rate": 4.857108835437009e-05,
+      "loss": 1.0231,
+      "step": 1080
+    },
+    {
+      "epoch": 0.0865268213300522,
+      "grad_norm": 0.8650990724563599,
+      "learning_rate": 4.8557857690984623e-05,
+      "loss": 0.9676,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08732064537895176,
+      "grad_norm": 0.9188217520713806,
+      "learning_rate": 4.854462702759916e-05,
+      "loss": 0.9851,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08811446942785131,
+      "grad_norm": 0.9663834571838379,
+      "learning_rate": 4.853139636421371e-05,
+      "loss": 0.9442,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08890829347675087,
+      "grad_norm": 0.8045158386230469,
+      "learning_rate": 4.851816570082825e-05,
+      "loss": 0.9411,
+      "step": 1120
+    },
+    {
+      "epoch": 0.08970211752565044,
+      "grad_norm": 0.9113909006118774,
+      "learning_rate": 4.850493503744278e-05,
+      "loss": 0.8866,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09049594157455,
+      "grad_norm": 0.8947747349739075,
+      "learning_rate": 4.849170437405732e-05,
+      "loss": 0.9206,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09128976562344956,
+      "grad_norm": 0.7741749882698059,
+      "learning_rate": 4.847847371067186e-05,
+      "loss": 1.0066,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09208358967234913,
+      "grad_norm": 0.9636614918708801,
+      "learning_rate": 4.846524304728639e-05,
+      "loss": 0.9518,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09287741372124869,
+      "grad_norm": 0.7008812427520752,
+      "learning_rate": 4.845201238390093e-05,
+      "loss": 0.9543,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09367123777014824,
+      "grad_norm": 0.9602216482162476,
+      "learning_rate": 4.843878172051547e-05,
+      "loss": 0.915,
+      "step": 1180
+    },
+    {
+      "epoch": 0.0944650618190478,
+      "grad_norm": 0.9824376702308655,
+      "learning_rate": 4.8425551057130006e-05,
+      "loss": 0.9073,
+      "step": 1190
+    },
+    {
+      "epoch": 0.09525888586794737,
+      "grad_norm": 0.7355358600616455,
+      "learning_rate": 4.8412320393744545e-05,
+      "loss": 0.9487,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09605270991684693,
+      "grad_norm": 1.0072027444839478,
+      "learning_rate": 4.8399089730359084e-05,
+      "loss": 0.9659,
+      "step": 1210
+    },
+    {
+      "epoch": 0.0968465339657465,
+      "grad_norm": 0.78026282787323,
+      "learning_rate": 4.838585906697362e-05,
+      "loss": 0.9716,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09764035801464606,
+      "grad_norm": 0.8772688508033752,
+      "learning_rate": 4.837262840358816e-05,
+      "loss": 0.9621,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09843418206354561,
+      "grad_norm": 0.8083699345588684,
+      "learning_rate": 4.8359397740202694e-05,
+      "loss": 0.8598,
+      "step": 1240
+    },
+    {
+      "epoch": 0.09922800611244517,
+      "grad_norm": 0.8948017954826355,
+      "learning_rate": 4.834616707681723e-05,
+      "loss": 0.9425,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10002183016134474,
+      "grad_norm": 0.7335965633392334,
+      "learning_rate": 4.833293641343177e-05,
+      "loss": 0.9441,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1008156542102443,
+      "grad_norm": 0.7533779740333557,
+      "learning_rate": 4.831970575004631e-05,
+      "loss": 0.9918,
+      "step": 1270
+    },
+    {
+      "epoch": 0.10160947825914386,
+      "grad_norm": 0.8201155066490173,
+      "learning_rate": 4.830647508666085e-05,
+      "loss": 1.014,
+      "step": 1280
+    },
+    {
+      "epoch": 0.10240330230804343,
+      "grad_norm": 0.6296638250350952,
+      "learning_rate": 4.829324442327539e-05,
+      "loss": 0.962,
+      "step": 1290
+    },
+    {
+      "epoch": 0.10319712635694299,
+      "grad_norm": 0.8968091011047363,
+      "learning_rate": 4.828001375988993e-05,
+      "loss": 0.9825,
+      "step": 1300
+    },
+    {
+      "epoch": 0.10399095040584254,
+      "grad_norm": 0.9169098138809204,
+      "learning_rate": 4.826678309650446e-05,
+      "loss": 0.8705,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1047847744547421,
+      "grad_norm": 0.7152407169342041,
+      "learning_rate": 4.8253552433119e-05,
+      "loss": 0.9902,
+      "step": 1320
+    },
+    {
+      "epoch": 0.10557859850364167,
+      "grad_norm": 0.9684945940971375,
+      "learning_rate": 4.824032176973354e-05,
+      "loss": 1.0179,
+      "step": 1330
+    },
+    {
+      "epoch": 0.10637242255254123,
+      "grad_norm": 1.0346221923828125,
+      "learning_rate": 4.8227091106348076e-05,
+      "loss": 0.9245,
+      "step": 1340
+    },
+    {
+      "epoch": 0.10716624660144079,
+      "grad_norm": 0.9577662348747253,
+      "learning_rate": 4.821386044296261e-05,
+      "loss": 1.0263,
+      "step": 1350
+    },
+    {
+      "epoch": 0.10796007065034036,
+      "grad_norm": 0.756993293762207,
+      "learning_rate": 4.820062977957715e-05,
+      "loss": 0.9375,
+      "step": 1360
+    },
+    {
+      "epoch": 0.10875389469923992,
+      "grad_norm": 0.6969035267829895,
+      "learning_rate": 4.8187399116191686e-05,
+      "loss": 0.9871,
+      "step": 1370
+    },
+    {
+      "epoch": 0.10954771874813947,
+      "grad_norm": 1.1062333583831787,
+      "learning_rate": 4.817416845280623e-05,
+      "loss": 0.9659,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11034154279703903,
+      "grad_norm": 0.592980146408081,
+      "learning_rate": 4.8160937789420764e-05,
+      "loss": 1.0116,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1111353668459386,
+      "grad_norm": 0.8485913276672363,
+      "learning_rate": 4.81477071260353e-05,
+      "loss": 0.963,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11192919089483816,
+      "grad_norm": 0.8063706159591675,
+      "learning_rate": 4.813447646264984e-05,
+      "loss": 0.9862,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11272301494373772,
+      "grad_norm": 0.705464243888855,
+      "learning_rate": 4.8121245799264374e-05,
+      "loss": 0.9472,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11351683899263729,
+      "grad_norm": 0.8876354098320007,
+      "learning_rate": 4.810801513587891e-05,
+      "loss": 0.924,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11431066304153685,
+      "grad_norm": 1.0832111835479736,
+      "learning_rate": 4.809478447249345e-05,
+      "loss": 0.9387,
+      "step": 1440
+    },
+    {
+      "epoch": 0.1151044870904364,
+      "grad_norm": 0.8490377068519592,
+      "learning_rate": 4.808155380910799e-05,
+      "loss": 1.0323,
+      "step": 1450
+    },
+    {
+      "epoch": 0.11589831113933596,
+      "grad_norm": 0.8763223886489868,
+      "learning_rate": 4.806832314572253e-05,
+      "loss": 0.9555,
+      "step": 1460
+    },
+    {
+      "epoch": 0.11669213518823553,
+      "grad_norm": 0.9210900664329529,
+      "learning_rate": 4.805509248233707e-05,
+      "loss": 1.0096,
+      "step": 1470
+    },
+    {
+      "epoch": 0.11748595923713509,
+      "grad_norm": 0.7176244258880615,
+      "learning_rate": 4.804186181895161e-05,
+      "loss": 0.9415,
+      "step": 1480
+    },
+    {
+      "epoch": 0.11827978328603465,
+      "grad_norm": 0.9028609991073608,
+      "learning_rate": 4.802863115556615e-05,
+      "loss": 0.9449,
+      "step": 1490
+    },
+    {
+      "epoch": 0.11907360733493422,
+      "grad_norm": 0.779123067855835,
+      "learning_rate": 4.801540049218068e-05,
+      "loss": 0.899,
+      "step": 1500
+    },
+    {
+      "epoch": 0.11986743138383377,
+      "grad_norm": 0.7735058665275574,
+      "learning_rate": 4.800216982879522e-05,
+      "loss": 0.9138,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12066125543273333,
+      "grad_norm": 0.851739764213562,
+      "learning_rate": 4.798893916540976e-05,
+      "loss": 0.9476,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1214550794816329,
+      "grad_norm": 0.8166431188583374,
+      "learning_rate": 4.797570850202429e-05,
+      "loss": 0.9746,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12224890353053246,
+      "grad_norm": 0.8283756971359253,
+      "learning_rate": 4.796247783863883e-05,
+      "loss": 0.9072,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12304272757943202,
+      "grad_norm": 0.8357489705085754,
+      "learning_rate": 4.794924717525337e-05,
+      "loss": 0.9563,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12383655162833158,
+      "grad_norm": 0.9437370896339417,
+      "learning_rate": 4.793601651186791e-05,
+      "loss": 0.9709,
+      "step": 1560
+    },
+    {
+      "epoch": 0.12463037567723115,
+      "grad_norm": 0.8020033836364746,
+      "learning_rate": 4.7922785848482445e-05,
+      "loss": 0.9925,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1254241997261307,
+      "grad_norm": 0.7915470600128174,
+      "learning_rate": 4.7909555185096983e-05,
+      "loss": 0.9991,
+      "step": 1580
+    },
+    {
+      "epoch": 0.12621802377503027,
+      "grad_norm": 0.6978237628936768,
+      "learning_rate": 4.789632452171152e-05,
+      "loss": 0.9994,
+      "step": 1590
+    },
+    {
+      "epoch": 0.12701184782392982,
+      "grad_norm": 0.9151979088783264,
+      "learning_rate": 4.788309385832606e-05,
+      "loss": 0.9341,
+      "step": 1600
+    },
+    {
+      "epoch": 0.12780567187282937,
+      "grad_norm": 0.8755201697349548,
+      "learning_rate": 4.7869863194940593e-05,
+      "loss": 0.9303,
+      "step": 1610
+    },
+    {
+      "epoch": 0.12859949592172895,
+      "grad_norm": 0.9662395715713501,
+      "learning_rate": 4.785663253155513e-05,
+      "loss": 0.9088,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1293933199706285,
+      "grad_norm": 0.7612505555152893,
+      "learning_rate": 4.784340186816967e-05,
+      "loss": 1.0068,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13018714401952808,
+      "grad_norm": 0.8853713274002075,
+      "learning_rate": 4.783017120478421e-05,
+      "loss": 0.9877,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13098096806842763,
+      "grad_norm": 0.6972758173942566,
+      "learning_rate": 4.781694054139875e-05,
+      "loss": 0.9068,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1317747921173272,
+      "grad_norm": 0.8050327301025391,
+      "learning_rate": 4.780370987801329e-05,
+      "loss": 0.8972,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13256861616622675,
+      "grad_norm": 0.8144531846046448,
+      "learning_rate": 4.779047921462783e-05,
+      "loss": 0.889,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1333624402151263,
+      "grad_norm": 0.8536010980606079,
+      "learning_rate": 4.777724855124236e-05,
+      "loss": 0.9718,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13415626426402588,
+      "grad_norm": 0.7940757274627686,
+      "learning_rate": 4.77640178878569e-05,
+      "loss": 0.9628,
+      "step": 1690
+    },
+    {
+      "epoch": 0.13495008831292543,
+      "grad_norm": 0.6885228157043457,
+      "learning_rate": 4.775078722447144e-05,
+      "loss": 0.9482,
+      "step": 1700
+    },
+    {
+      "epoch": 0.135743912361825,
+      "grad_norm": 0.7801288366317749,
+      "learning_rate": 4.7737556561085976e-05,
+      "loss": 0.9084,
+      "step": 1710
+    },
+    {
+      "epoch": 0.13653773641072456,
+      "grad_norm": 0.7502837777137756,
+      "learning_rate": 4.772432589770051e-05,
+      "loss": 0.869,
+      "step": 1720
+    },
+    {
+      "epoch": 0.13733156045962414,
+      "grad_norm": 0.9921239614486694,
+      "learning_rate": 4.7711095234315054e-05,
+      "loss": 0.9065,
+      "step": 1730
+    },
+    {
+      "epoch": 0.13812538450852369,
+      "grad_norm": 0.8346391916275024,
+      "learning_rate": 4.769786457092959e-05,
+      "loss": 0.9831,
+      "step": 1740
+    },
+    {
+      "epoch": 0.13891920855742323,
+      "grad_norm": 0.8118425607681274,
+      "learning_rate": 4.7684633907544125e-05,
+      "loss": 0.9673,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1397130326063228,
+      "grad_norm": 0.8836124539375305,
+      "learning_rate": 4.7671403244158664e-05,
+      "loss": 0.9506,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14050685665522236,
+      "grad_norm": 0.8998447060585022,
+      "learning_rate": 4.76581725807732e-05,
+      "loss": 0.9443,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14130068070412194,
+      "grad_norm": 0.9058572053909302,
+      "learning_rate": 4.764494191738774e-05,
+      "loss": 0.8735,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1420945047530215,
+      "grad_norm": 0.9656883478164673,
+      "learning_rate": 4.7631711254002274e-05,
+      "loss": 0.9379,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14288832880192107,
+      "grad_norm": 0.8430660367012024,
+      "learning_rate": 4.761848059061681e-05,
+      "loss": 0.9405,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14368215285082062,
+      "grad_norm": 0.963220477104187,
+      "learning_rate": 4.760524992723135e-05,
+      "loss": 0.9306,
+      "step": 1810
+    },
+    {
+      "epoch": 0.14447597689972017,
+      "grad_norm": 0.806779146194458,
+      "learning_rate": 4.759201926384589e-05,
+      "loss": 0.964,
+      "step": 1820
+    },
+    {
+      "epoch": 0.14526980094861974,
+      "grad_norm": 0.6742879152297974,
+      "learning_rate": 4.757878860046043e-05,
+      "loss": 0.9398,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1460636249975193,
+      "grad_norm": 0.8305578827857971,
+      "learning_rate": 4.756555793707497e-05,
+      "loss": 0.9661,
+      "step": 1840
+    },
+    {
+      "epoch": 0.14685744904641887,
+      "grad_norm": 0.9429900646209717,
+      "learning_rate": 4.755232727368951e-05,
+      "loss": 0.9151,
+      "step": 1850
+    },
+    {
+      "epoch": 0.14765127309531842,
+      "grad_norm": 0.7211086750030518,
+      "learning_rate": 4.753909661030404e-05,
+      "loss": 0.9367,
+      "step": 1860
+    },
+    {
+      "epoch": 0.148445097144218,
+      "grad_norm": 0.932462751865387,
+      "learning_rate": 4.752586594691858e-05,
+      "loss": 0.9465,
+      "step": 1870
+    },
+    {
+      "epoch": 0.14923892119311755,
+      "grad_norm": 0.8956672549247742,
+      "learning_rate": 4.751263528353312e-05,
+      "loss": 0.9948,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1500327452420171,
+      "grad_norm": 0.78652423620224,
+      "learning_rate": 4.7499404620147656e-05,
+      "loss": 0.9323,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15082656929091667,
+      "grad_norm": 0.8287889361381531,
+      "learning_rate": 4.7486173956762195e-05,
+      "loss": 0.944,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15162039333981622,
+      "grad_norm": 1.0075963735580444,
+      "learning_rate": 4.7472943293376734e-05,
+      "loss": 0.9595,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1524142173887158,
+      "grad_norm": 1.075189471244812,
+      "learning_rate": 4.745971262999127e-05,
+      "loss": 0.9636,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15320804143761535,
+      "grad_norm": 0.9045613408088684,
+      "learning_rate": 4.744648196660581e-05,
+      "loss": 0.9544,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15400186548651493,
+      "grad_norm": 0.8848656415939331,
+      "learning_rate": 4.7433251303220344e-05,
+      "loss": 0.9707,
+      "step": 1940
+    },
+    {
+      "epoch": 0.15479568953541448,
+      "grad_norm": 1.1052658557891846,
+      "learning_rate": 4.742002063983488e-05,
+      "loss": 0.9881,
+      "step": 1950
+    },
+    {
+      "epoch": 0.15558951358431403,
+      "grad_norm": 0.8946327567100525,
+      "learning_rate": 4.740678997644942e-05,
+      "loss": 1.0025,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1563833376332136,
+      "grad_norm": 0.8792600035667419,
+      "learning_rate": 4.7393559313063954e-05,
+      "loss": 0.9993,
+      "step": 1970
+    },
+    {
+      "epoch": 0.15717716168211315,
+      "grad_norm": 0.9298778176307678,
+      "learning_rate": 4.738032864967849e-05,
+      "loss": 0.9169,
+      "step": 1980
+    },
+    {
+      "epoch": 0.15797098573101273,
+      "grad_norm": 0.7488551735877991,
+      "learning_rate": 4.736709798629303e-05,
+      "loss": 0.9047,
+      "step": 1990
+    },
+    {
+      "epoch": 0.15876480977991228,
+      "grad_norm": 0.7787390351295471,
+      "learning_rate": 4.735386732290758e-05,
+      "loss": 0.936,
+      "step": 2000
+    },
+    {
+      "epoch": 0.15955863382881183,
+      "grad_norm": 0.707550048828125,
+      "learning_rate": 4.734063665952211e-05,
+      "loss": 0.9948,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1603524578777114,
+      "grad_norm": 0.8335928916931152,
+      "learning_rate": 4.732740599613665e-05,
+      "loss": 0.9602,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16114628192661096,
+      "grad_norm": 1.0111210346221924,
+      "learning_rate": 4.731417533275119e-05,
+      "loss": 0.9417,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16194010597551053,
+      "grad_norm": 0.6672728657722473,
+      "learning_rate": 4.730094466936573e-05,
+      "loss": 0.9389,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16273393002441008,
+      "grad_norm": 0.8862355947494507,
+      "learning_rate": 4.728771400598026e-05,
+      "loss": 0.9071,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16352775407330966,
+      "grad_norm": 0.8846628069877625,
+      "learning_rate": 4.72744833425948e-05,
+      "loss": 0.9351,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1643215781222092,
+      "grad_norm": 0.833533763885498,
+      "learning_rate": 4.726125267920934e-05,
+      "loss": 0.9713,
+      "step": 2070
+    },
+    {
+      "epoch": 0.16511540217110876,
+      "grad_norm": 0.724327564239502,
+      "learning_rate": 4.7248022015823876e-05,
+      "loss": 0.9349,
+      "step": 2080
+    },
+    {
+      "epoch": 0.16590922622000834,
+      "grad_norm": 0.9286043643951416,
+      "learning_rate": 4.7234791352438415e-05,
+      "loss": 0.9144,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1667030502689079,
+      "grad_norm": 0.8603926301002502,
+      "learning_rate": 4.7221560689052953e-05,
+      "loss": 0.9156,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16749687431780746,
+      "grad_norm": 0.9300916194915771,
+      "learning_rate": 4.720833002566749e-05,
+      "loss": 0.8894,
+      "step": 2110
+    },
+    {
+      "epoch": 0.16829069836670701,
+      "grad_norm": 0.7799938917160034,
+      "learning_rate": 4.7195099362282025e-05,
+      "loss": 0.8955,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1690845224156066,
+      "grad_norm": 0.9887612462043762,
+      "learning_rate": 4.7181868698896563e-05,
+      "loss": 0.9712,
+      "step": 2130
+    },
+    {
+      "epoch": 0.16987834646450614,
+      "grad_norm": 0.9083105325698853,
+      "learning_rate": 4.71686380355111e-05,
+      "loss": 0.9434,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1706721705134057,
+      "grad_norm": 0.868770182132721,
+      "learning_rate": 4.715540737212564e-05,
+      "loss": 0.9522,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17146599456230527,
+      "grad_norm": 0.819227933883667,
+      "learning_rate": 4.7142176708740173e-05,
+      "loss": 0.9812,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17225981861120482,
+      "grad_norm": 0.9821423292160034,
+      "learning_rate": 4.712894604535472e-05,
+      "loss": 0.9075,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1730536426601044,
+      "grad_norm": 0.6770303845405579,
+      "learning_rate": 4.711571538196926e-05,
+      "loss": 0.9006,
+      "step": 2180
+    },
+    {
+      "epoch": 0.17384746670900394,
+      "grad_norm": 1.0133216381072998,
+      "learning_rate": 4.71024847185838e-05,
+      "loss": 0.9851,
+      "step": 2190
+    },
+    {
+      "epoch": 0.17464129075790352,
+      "grad_norm": 0.6719434261322021,
+      "learning_rate": 4.708925405519833e-05,
+      "loss": 0.9876,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17543511480680307,
+      "grad_norm": 0.9074175357818604,
+      "learning_rate": 4.707602339181287e-05,
+      "loss": 0.8679,
+      "step": 2210
+    },
+    {
+      "epoch": 0.17622893885570262,
+      "grad_norm": 0.7234490513801575,
+      "learning_rate": 4.706279272842741e-05,
+      "loss": 0.9469,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1770227629046022,
+      "grad_norm": 0.8408550024032593,
+      "learning_rate": 4.704956206504194e-05,
+      "loss": 0.8871,
+      "step": 2230
+    },
+    {
+      "epoch": 0.17781658695350175,
+      "grad_norm": 0.9565715789794922,
+      "learning_rate": 4.703633140165648e-05,
+      "loss": 0.9597,
+      "step": 2240
+    },
+    {
+      "epoch": 0.17861041100240133,
+      "grad_norm": 0.7657901644706726,
+      "learning_rate": 4.702310073827102e-05,
+      "loss": 0.912,
+      "step": 2250
+    },
+    {
+      "epoch": 0.17940423505130088,
+      "grad_norm": 0.7222371101379395,
+      "learning_rate": 4.7009870074885556e-05,
+      "loss": 0.9369,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18019805910020045,
+      "grad_norm": 0.8453822731971741,
+      "learning_rate": 4.6996639411500095e-05,
+      "loss": 0.94,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1809918831491,
+      "grad_norm": 0.6532242298126221,
+      "learning_rate": 4.6983408748114634e-05,
+      "loss": 0.8897,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18178570719799955,
+      "grad_norm": 0.8275949358940125,
+      "learning_rate": 4.697017808472917e-05,
+      "loss": 0.9235,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18257953124689913,
+      "grad_norm": 0.7926145792007446,
+      "learning_rate": 4.695694742134371e-05,
+      "loss": 0.9327,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18337335529579868,
+      "grad_norm": 0.8629859089851379,
+      "learning_rate": 4.6943716757958244e-05,
+      "loss": 0.9714,
+      "step": 2310
+    },
+    {
+      "epoch": 0.18416717934469826,
+      "grad_norm": 0.8121680021286011,
+      "learning_rate": 4.693048609457278e-05,
+      "loss": 0.9545,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1849610033935978,
+      "grad_norm": 0.9470918774604797,
+      "learning_rate": 4.691725543118732e-05,
+      "loss": 0.8992,
+      "step": 2330
+    },
+    {
+      "epoch": 0.18575482744249738,
+      "grad_norm": 0.8054450750350952,
+      "learning_rate": 4.690402476780186e-05,
+      "loss": 0.9984,
+      "step": 2340
+    },
+    {
+      "epoch": 0.18654865149139693,
+      "grad_norm": 0.6857488751411438,
+      "learning_rate": 4.68907941044164e-05,
+      "loss": 0.9411,
+      "step": 2350
+    },
+    {
+      "epoch": 0.18734247554029648,
+      "grad_norm": 0.8462006449699402,
+      "learning_rate": 4.687756344103094e-05,
+      "loss": 0.9527,
+      "step": 2360
+    },
+    {
+      "epoch": 0.18813629958919606,
+      "grad_norm": 0.8726380467414856,
+      "learning_rate": 4.686433277764548e-05,
+      "loss": 0.8983,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1889301236380956,
+      "grad_norm": 0.883798360824585,
+      "learning_rate": 4.685110211426001e-05,
+      "loss": 0.9382,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1897239476869952,
+      "grad_norm": 0.8085561394691467,
+      "learning_rate": 4.683787145087455e-05,
+      "loss": 0.9371,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19051777173589474,
+      "grad_norm": 0.816972017288208,
+      "learning_rate": 4.682464078748909e-05,
+      "loss": 0.9322,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1913115957847943,
+      "grad_norm": 0.8701701164245605,
+      "learning_rate": 4.6811410124103626e-05,
+      "loss": 0.9277,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19210541983369386,
+      "grad_norm": 0.9203164577484131,
+      "learning_rate": 4.679817946071816e-05,
+      "loss": 0.9365,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1928992438825934,
+      "grad_norm": 0.9870917797088623,
+      "learning_rate": 4.67849487973327e-05,
+      "loss": 0.973,
+      "step": 2430
+    },
+    {
+      "epoch": 0.193693067931493,
+      "grad_norm": 0.8381191492080688,
+      "learning_rate": 4.677171813394724e-05,
+      "loss": 0.9344,
+      "step": 2440
+    },
+    {
+      "epoch": 0.19448689198039254,
+      "grad_norm": 0.9467155933380127,
+      "learning_rate": 4.6758487470561775e-05,
+      "loss": 0.9162,
+      "step": 2450
+    },
+    {
+      "epoch": 0.19528071602929212,
+      "grad_norm": 0.8836456537246704,
+      "learning_rate": 4.6745256807176314e-05,
+      "loss": 0.9548,
+      "step": 2460
+    },
+    {
+      "epoch": 0.19607454007819167,
+      "grad_norm": 0.806165337562561,
+      "learning_rate": 4.673202614379085e-05,
+      "loss": 0.9615,
+      "step": 2470
+    },
+    {
+      "epoch": 0.19686836412709122,
+      "grad_norm": 0.8333777785301208,
+      "learning_rate": 4.671879548040539e-05,
+      "loss": 0.9331,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1976621881759908,
+      "grad_norm": 0.9667910933494568,
+      "learning_rate": 4.6705564817019924e-05,
+      "loss": 0.9031,
+      "step": 2490
+    },
+    {
+      "epoch": 0.19845601222489034,
+      "grad_norm": 0.8661583065986633,
+      "learning_rate": 4.669233415363446e-05,
+      "loss": 0.9208,
+      "step": 2500
+    },
+    {
+      "epoch": 0.19924983627378992,
+      "grad_norm": 1.0444538593292236,
+      "learning_rate": 4.6679103490249e-05,
+      "loss": 0.992,
+      "step": 2510
+    },
+    {
+      "epoch": 0.20004366032268947,
+      "grad_norm": 0.6848312616348267,
+      "learning_rate": 4.666587282686354e-05,
+      "loss": 0.9215,
+      "step": 2520
+    },
+    {
+      "epoch": 0.20083748437158905,
+      "grad_norm": 0.789969801902771,
+      "learning_rate": 4.665264216347808e-05,
+      "loss": 0.9371,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2016313084204886,
+      "grad_norm": 0.7210158109664917,
+      "learning_rate": 4.663941150009262e-05,
+      "loss": 0.9561,
+      "step": 2540
+    },
+    {
+      "epoch": 0.20242513246938815,
+      "grad_norm": 0.9324051737785339,
+      "learning_rate": 4.662618083670716e-05,
+      "loss": 0.9632,
+      "step": 2550
+    },
+    {
+      "epoch": 0.20321895651828772,
+      "grad_norm": 0.6545863151550293,
+      "learning_rate": 4.661295017332169e-05,
+      "loss": 0.9027,
+      "step": 2560
+    },
+    {
+      "epoch": 0.20401278056718727,
+      "grad_norm": 0.9494184255599976,
+      "learning_rate": 4.659971950993623e-05,
+      "loss": 0.8897,
+      "step": 2570
+    },
+    {
+      "epoch": 0.20480660461608685,
+      "grad_norm": 0.670927107334137,
+      "learning_rate": 4.658648884655077e-05,
+      "loss": 0.9166,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2056004286649864,
+      "grad_norm": 0.8145579695701599,
+      "learning_rate": 4.657325818316531e-05,
+      "loss": 0.9364,
+      "step": 2590
+    },
+    {
+      "epoch": 0.20639425271388598,
+      "grad_norm": 0.9413288831710815,
+      "learning_rate": 4.656002751977984e-05,
+      "loss": 0.8997,
+      "step": 2600
+    },
+    {
+      "epoch": 0.20718807676278553,
+      "grad_norm": 0.8547295331954956,
+      "learning_rate": 4.6546796856394385e-05,
+      "loss": 0.9594,
+      "step": 2610
+    },
+    {
+      "epoch": 0.20798190081168508,
+      "grad_norm": 0.711956799030304,
+      "learning_rate": 4.6533566193008923e-05,
+      "loss": 0.9728,
+      "step": 2620
+    },
+    {
+      "epoch": 0.20877572486058465,
+      "grad_norm": 0.7029785513877869,
+      "learning_rate": 4.652033552962346e-05,
+      "loss": 0.9561,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2095695489094842,
+      "grad_norm": 0.7861992120742798,
+      "learning_rate": 4.6507104866237995e-05,
+      "loss": 0.9288,
+      "step": 2640
+    },
+    {
+      "epoch": 0.21036337295838378,
+      "grad_norm": 0.9786797761917114,
+      "learning_rate": 4.6493874202852533e-05,
+      "loss": 0.9492,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21115719700728333,
+      "grad_norm": 0.8280521631240845,
+      "learning_rate": 4.648064353946707e-05,
+      "loss": 0.96,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2119510210561829,
+      "grad_norm": 0.5959718227386475,
+      "learning_rate": 4.6467412876081605e-05,
+      "loss": 0.989,
+      "step": 2670
+    },
+    {
+      "epoch": 0.21274484510508246,
+      "grad_norm": 0.7416049838066101,
+      "learning_rate": 4.6454182212696143e-05,
+      "loss": 1.0148,
+      "step": 2680
+    },
+    {
+      "epoch": 0.213538669153982,
+      "grad_norm": 0.915602445602417,
+      "learning_rate": 4.644095154931068e-05,
+      "loss": 0.9789,
+      "step": 2690
+    },
+    {
+      "epoch": 0.21433249320288159,
+      "grad_norm": 0.616190493106842,
+      "learning_rate": 4.642772088592522e-05,
+      "loss": 0.9163,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21512631725178114,
+      "grad_norm": 0.7318758368492126,
+      "learning_rate": 4.641449022253976e-05,
+      "loss": 0.9084,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2159201413006807,
+      "grad_norm": 0.7663441896438599,
+      "learning_rate": 4.64012595591543e-05,
+      "loss": 0.921,
+      "step": 2720
+    },
+    {
+      "epoch": 0.21671396534958026,
+      "grad_norm": 0.7115159034729004,
+      "learning_rate": 4.638802889576884e-05,
+      "loss": 0.9214,
+      "step": 2730
+    },
+    {
+      "epoch": 0.21750778939847984,
+      "grad_norm": 0.569379985332489,
+      "learning_rate": 4.637479823238338e-05,
+      "loss": 0.9658,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2183016134473794,
+      "grad_norm": 0.9785438776016235,
+      "learning_rate": 4.636156756899791e-05,
+      "loss": 0.8818,
+      "step": 2750
+    },
+    {
+      "epoch": 0.21909543749627894,
+      "grad_norm": 0.7893972396850586,
+      "learning_rate": 4.634833690561245e-05,
+      "loss": 0.9254,
+      "step": 2760
+    },
+    {
+      "epoch": 0.21988926154517852,
+      "grad_norm": 0.948264479637146,
+      "learning_rate": 4.633510624222699e-05,
+      "loss": 0.8619,
+      "step": 2770
+    },
+    {
+      "epoch": 0.22068308559407807,
+      "grad_norm": 0.8142668008804321,
+      "learning_rate": 4.6321875578841526e-05,
+      "loss": 0.9289,
+      "step": 2780
+    },
+    {
+      "epoch": 0.22147690964297764,
+      "grad_norm": 0.7517206072807312,
+      "learning_rate": 4.6308644915456065e-05,
+      "loss": 0.8759,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2222707336918772,
+      "grad_norm": 0.6064891219139099,
+      "learning_rate": 4.6295414252070604e-05,
+      "loss": 0.8558,
+      "step": 2800
+    },
+    {
+      "epoch": 0.22306455774077677,
+      "grad_norm": 0.7425729036331177,
+      "learning_rate": 4.628218358868514e-05,
+      "loss": 0.9281,
+      "step": 2810
+    },
+    {
+      "epoch": 0.22385838178967632,
+      "grad_norm": 0.7883281707763672,
+      "learning_rate": 4.6268952925299675e-05,
+      "loss": 0.8969,
+      "step": 2820
+    },
+    {
+      "epoch": 0.22465220583857587,
+      "grad_norm": 0.8569539189338684,
+      "learning_rate": 4.6255722261914214e-05,
+      "loss": 0.9111,
+      "step": 2830
+    },
+    {
+      "epoch": 0.22544602988747545,
+      "grad_norm": 0.8346198797225952,
+      "learning_rate": 4.624249159852875e-05,
+      "loss": 0.8691,
+      "step": 2840
+    },
+    {
+      "epoch": 0.226239853936375,
+      "grad_norm": 0.8474957346916199,
+      "learning_rate": 4.622926093514329e-05,
+      "loss": 0.8502,
+      "step": 2850
+    },
+    {
+      "epoch": 0.22703367798527457,
+      "grad_norm": 0.9480683207511902,
+      "learning_rate": 4.6216030271757824e-05,
+      "loss": 0.8832,
+      "step": 2860
+    },
+    {
+      "epoch": 0.22782750203417412,
+      "grad_norm": 0.9572345018386841,
+      "learning_rate": 4.620279960837236e-05,
+      "loss": 0.9159,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2286213260830737,
+      "grad_norm": 0.9414117932319641,
+      "learning_rate": 4.618956894498691e-05,
+      "loss": 0.9427,
+      "step": 2880
+    },
+    {
+      "epoch": 0.22941515013197325,
+      "grad_norm": 0.7920238375663757,
+      "learning_rate": 4.617633828160145e-05,
+      "loss": 0.8879,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2302089741808728,
+      "grad_norm": 0.736736536026001,
+      "learning_rate": 4.616310761821598e-05,
+      "loss": 0.9429,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23100279822977238,
+      "grad_norm": 0.7859109044075012,
+      "learning_rate": 4.614987695483052e-05,
+      "loss": 0.8835,
+      "step": 2910
+    },
+    {
+      "epoch": 0.23179662227867193,
+      "grad_norm": 0.9163410663604736,
+      "learning_rate": 4.613664629144506e-05,
+      "loss": 0.9017,
+      "step": 2920
+    },
+    {
+      "epoch": 0.2325904463275715,
+      "grad_norm": 0.9528617262840271,
+      "learning_rate": 4.612341562805959e-05,
+      "loss": 0.9572,
+      "step": 2930
+    },
+    {
+      "epoch": 0.23338427037647105,
+      "grad_norm": 0.9378358125686646,
+      "learning_rate": 4.611018496467413e-05,
+      "loss": 0.9002,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2341780944253706,
+      "grad_norm": 0.754435122013092,
+      "learning_rate": 4.609695430128867e-05,
+      "loss": 0.9026,
+      "step": 2950
+    },
+    {
+      "epoch": 0.23497191847427018,
+      "grad_norm": 0.906570315361023,
+      "learning_rate": 4.6083723637903206e-05,
+      "loss": 0.9109,
+      "step": 2960
+    },
+    {
+      "epoch": 0.23576574252316973,
+      "grad_norm": 0.9783695936203003,
+      "learning_rate": 4.6070492974517745e-05,
+      "loss": 0.9366,
+      "step": 2970
+    },
+    {
+      "epoch": 0.2365595665720693,
+      "grad_norm": 0.8152298331260681,
+      "learning_rate": 4.6057262311132284e-05,
+      "loss": 0.9555,
+      "step": 2980
+    },
+    {
+      "epoch": 0.23735339062096886,
+      "grad_norm": 0.8119134902954102,
+      "learning_rate": 4.604403164774682e-05,
+      "loss": 0.9418,
+      "step": 2990
+    },
+    {
+      "epoch": 0.23814721466986843,
+      "grad_norm": 0.5931934714317322,
+      "learning_rate": 4.603080098436136e-05,
+      "loss": 0.8831,
+      "step": 3000
+    },
+    {
+      "epoch": 0.23894103871876798,
+      "grad_norm": 1.0472118854522705,
+      "learning_rate": 4.6017570320975894e-05,
+      "loss": 0.9529,
+      "step": 3010
+    },
+    {
+      "epoch": 0.23973486276766753,
+      "grad_norm": 0.8576637506484985,
+      "learning_rate": 4.600433965759043e-05,
+      "loss": 0.9101,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2405286868165671,
+      "grad_norm": 0.7458338141441345,
+      "learning_rate": 4.599110899420497e-05,
+      "loss": 0.8409,
+      "step": 3030
+    },
+    {
+      "epoch": 0.24132251086546666,
+      "grad_norm": 0.8986132740974426,
+      "learning_rate": 4.5977878330819504e-05,
+      "loss": 0.9417,
+      "step": 3040
+    },
+    {
+      "epoch": 0.24211633491436624,
+      "grad_norm": 0.9694287180900574,
+      "learning_rate": 4.596464766743404e-05,
+      "loss": 0.9255,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2429101589632658,
+      "grad_norm": 0.9465477466583252,
+      "learning_rate": 4.595141700404859e-05,
+      "loss": 0.9307,
+      "step": 3060
+    },
+    {
+      "epoch": 0.24370398301216536,
+      "grad_norm": 0.6529215574264526,
+      "learning_rate": 4.593818634066313e-05,
+      "loss": 0.8807,
+      "step": 3070
+    },
+    {
+      "epoch": 0.24449780706106491,
+      "grad_norm": 0.9342437982559204,
+      "learning_rate": 4.592495567727766e-05,
+      "loss": 0.9448,
+      "step": 3080
+    },
+    {
+      "epoch": 0.24529163110996446,
+      "grad_norm": 0.7587674260139465,
+      "learning_rate": 4.59117250138922e-05,
+      "loss": 0.9655,
+      "step": 3090
+    },
+    {
+      "epoch": 0.24608545515886404,
+      "grad_norm": 0.7746047377586365,
+      "learning_rate": 4.589849435050674e-05,
+      "loss": 0.9213,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2468792792077636,
+      "grad_norm": 0.9817701578140259,
+      "learning_rate": 4.588526368712128e-05,
+      "loss": 0.9621,
+      "step": 3110
+    },
+    {
+      "epoch": 0.24767310325666317,
+      "grad_norm": 0.9328246116638184,
+      "learning_rate": 4.587203302373581e-05,
+      "loss": 0.8562,
+      "step": 3120
+    },
+    {
+      "epoch": 0.24846692730556272,
+      "grad_norm": 0.8459776043891907,
+      "learning_rate": 4.585880236035035e-05,
+      "loss": 0.9237,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2492607513544623,
+      "grad_norm": 0.8480767607688904,
+      "learning_rate": 4.584557169696489e-05,
+      "loss": 0.8931,
+      "step": 3140
+    },
+    {
+      "epoch": 0.25005457540336185,
+      "grad_norm": 0.8643662333488464,
+      "learning_rate": 4.5832341033579426e-05,
+      "loss": 0.9657,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2508483994522614,
+      "grad_norm": 0.7610037326812744,
+      "learning_rate": 4.5819110370193965e-05,
+      "loss": 0.9678,
+      "step": 3160
+    },
+    {
+      "epoch": 0.25164222350116094,
+      "grad_norm": 0.8325300812721252,
+      "learning_rate": 4.5805879706808503e-05,
+      "loss": 0.9242,
+      "step": 3170
+    },
+    {
+      "epoch": 0.25243604755006055,
+      "grad_norm": 0.8497787117958069,
+      "learning_rate": 4.579264904342304e-05,
+      "loss": 0.9659,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2532298715989601,
+      "grad_norm": 0.9831972122192383,
+      "learning_rate": 4.5779418380037575e-05,
+      "loss": 0.927,
+      "step": 3190
+    },
+    {
+      "epoch": 0.25402369564785965,
+      "grad_norm": 0.829680323600769,
+      "learning_rate": 4.5766187716652113e-05,
+      "loss": 0.8841,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2548175196967592,
+      "grad_norm": 0.8839125633239746,
+      "learning_rate": 4.575295705326665e-05,
+      "loss": 0.8945,
+      "step": 3210
+    },
+    {
+      "epoch": 0.25561134374565875,
+      "grad_norm": 1.0367532968521118,
+      "learning_rate": 4.573972638988119e-05,
+      "loss": 0.8785,
+      "step": 3220
+    },
+    {
+      "epoch": 0.25640516779455835,
+      "grad_norm": 0.9415130615234375,
+      "learning_rate": 4.572649572649573e-05,
+      "loss": 0.9348,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2571989918434579,
+      "grad_norm": 0.7434957027435303,
+      "learning_rate": 4.571326506311027e-05,
+      "loss": 0.8587,
+      "step": 3240
+    },
+    {
+      "epoch": 0.25799281589235745,
+      "grad_norm": 0.6429414749145508,
+      "learning_rate": 4.570003439972481e-05,
+      "loss": 0.9606,
+      "step": 3250
+    },
+    {
+      "epoch": 0.258786639941257,
+      "grad_norm": 1.4009513854980469,
+      "learning_rate": 4.568680373633934e-05,
+      "loss": 0.8492,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2595804639901566,
+      "grad_norm": 0.6804202198982239,
+      "learning_rate": 4.567357307295388e-05,
+      "loss": 0.9661,
+      "step": 3270
+    },
+    {
+      "epoch": 0.26037428803905616,
+      "grad_norm": 0.7768397331237793,
+      "learning_rate": 4.566034240956842e-05,
+      "loss": 0.8915,
+      "step": 3280
+    },
+    {
+      "epoch": 0.2611681120879557,
+      "grad_norm": 0.8182624578475952,
+      "learning_rate": 4.564711174618296e-05,
+      "loss": 0.914,
+      "step": 3290
+    },
+    {
+      "epoch": 0.26196193613685526,
+      "grad_norm": 0.8247461915016174,
+      "learning_rate": 4.563388108279749e-05,
+      "loss": 0.9191,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2627557601857548,
+      "grad_norm": 0.8486924171447754,
+      "learning_rate": 4.562065041941203e-05,
+      "loss": 0.9561,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2635495842346544,
+      "grad_norm": 0.7965856790542603,
+      "learning_rate": 4.560741975602657e-05,
+      "loss": 0.9409,
+      "step": 3320
+    },
+    {
+      "epoch": 0.26434340828355396,
+      "grad_norm": 1.0266188383102417,
+      "learning_rate": 4.559418909264111e-05,
+      "loss": 0.8715,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2651372323324535,
+      "grad_norm": 0.795364499092102,
+      "learning_rate": 4.5580958429255645e-05,
+      "loss": 0.8615,
+      "step": 3340
+    },
+    {
+      "epoch": 0.26593105638135306,
+      "grad_norm": 0.7666260600090027,
+      "learning_rate": 4.5567727765870184e-05,
+      "loss": 0.8847,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2667248804302526,
+      "grad_norm": 0.741375207901001,
+      "learning_rate": 4.555449710248472e-05,
+      "loss": 0.977,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2675187044791522,
+      "grad_norm": 0.8026778697967529,
+      "learning_rate": 4.5541266439099255e-05,
+      "loss": 1.012,
+      "step": 3370
+    },
+    {
+      "epoch": 0.26831252852805176,
+      "grad_norm": 0.828702986240387,
+      "learning_rate": 4.5528035775713794e-05,
+      "loss": 0.9397,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2691063525769513,
+      "grad_norm": 0.7557156085968018,
+      "learning_rate": 4.551480511232833e-05,
+      "loss": 0.9116,
+      "step": 3390
+    },
+    {
+      "epoch": 0.26990017662585086,
+      "grad_norm": 0.8590690493583679,
+      "learning_rate": 4.550157444894287e-05,
+      "loss": 0.899,
+      "step": 3400
+    },
+    {
+      "epoch": 0.27069400067475047,
+      "grad_norm": 1.0395570993423462,
+      "learning_rate": 4.548834378555741e-05,
+      "loss": 0.8633,
+      "step": 3410
+    },
+    {
+      "epoch": 0.27148782472365,
+      "grad_norm": 0.756775438785553,
+      "learning_rate": 4.547511312217195e-05,
+      "loss": 0.9149,
+      "step": 3420
+    },
+    {
+      "epoch": 0.27228164877254957,
+      "grad_norm": 1.0174977779388428,
+      "learning_rate": 4.546188245878649e-05,
+      "loss": 0.8964,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2730754728214491,
+      "grad_norm": 0.9375600218772888,
+      "learning_rate": 4.544865179540103e-05,
+      "loss": 1.0064,
+      "step": 3440
+    },
+    {
+      "epoch": 0.27386929687034867,
+      "grad_norm": 0.7545995116233826,
+      "learning_rate": 4.543542113201556e-05,
+      "loss": 0.8792,
+      "step": 3450
+    },
+    {
+      "epoch": 0.27466312091924827,
+      "grad_norm": 0.7980530261993408,
+      "learning_rate": 4.54221904686301e-05,
+      "loss": 0.9207,
+      "step": 3460
+    },
+    {
+      "epoch": 0.2754569449681478,
+      "grad_norm": 0.8304005265235901,
+      "learning_rate": 4.540895980524464e-05,
+      "loss": 0.867,
+      "step": 3470
+    },
+    {
+      "epoch": 0.27625076901704737,
+      "grad_norm": 0.9552587270736694,
+      "learning_rate": 4.539572914185917e-05,
+      "loss": 0.9498,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2770445930659469,
+      "grad_norm": 0.8936342000961304,
+      "learning_rate": 4.538249847847371e-05,
+      "loss": 0.9253,
+      "step": 3490
+    },
+    {
+      "epoch": 0.27783841711484647,
+      "grad_norm": 0.8894282579421997,
+      "learning_rate": 4.5369267815088254e-05,
+      "loss": 0.8884,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2786322411637461,
+      "grad_norm": 0.7489833831787109,
+      "learning_rate": 4.535603715170279e-05,
+      "loss": 0.9472,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2794260652126456,
+      "grad_norm": 0.7078631520271301,
+      "learning_rate": 4.5342806488317325e-05,
+      "loss": 0.9027,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2802198892615452,
+      "grad_norm": 0.7065866589546204,
+      "learning_rate": 4.5329575824931864e-05,
+      "loss": 0.9334,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2810137133104447,
+      "grad_norm": 0.5849603414535522,
+      "learning_rate": 4.53163451615464e-05,
+      "loss": 0.9532,
+      "step": 3540
+    },
+    {
+      "epoch": 0.28180753735934433,
+      "grad_norm": 0.7648496627807617,
+      "learning_rate": 4.530311449816094e-05,
+      "loss": 0.9062,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2826013614082439,
+      "grad_norm": 0.7615086436271667,
+      "learning_rate": 4.5289883834775474e-05,
+      "loss": 0.999,
+      "step": 3560
+    },
+    {
+      "epoch": 0.28339518545714343,
+      "grad_norm": 0.7692292928695679,
+      "learning_rate": 4.527665317139001e-05,
+      "loss": 0.9794,
+      "step": 3570
+    },
+    {
+      "epoch": 0.284189009506043,
+      "grad_norm": 0.6985743045806885,
+      "learning_rate": 4.526342250800455e-05,
+      "loss": 0.922,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2849828335549425,
+      "grad_norm": 0.8894872069358826,
+      "learning_rate": 4.525019184461909e-05,
+      "loss": 0.9348,
+      "step": 3590
+    },
+    {
+      "epoch": 0.28577665760384213,
+      "grad_norm": 1.1146018505096436,
+      "learning_rate": 4.523696118123363e-05,
+      "loss": 0.9343,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2865704816527417,
+      "grad_norm": 0.7811164855957031,
+      "learning_rate": 4.522373051784817e-05,
+      "loss": 0.9495,
+      "step": 3610
+    },
+    {
+      "epoch": 0.28736430570164123,
+      "grad_norm": 0.9659305810928345,
+      "learning_rate": 4.521049985446271e-05,
+      "loss": 0.9037,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2881581297505408,
+      "grad_norm": 0.8798300623893738,
+      "learning_rate": 4.519726919107724e-05,
+      "loss": 0.8077,
+      "step": 3630
+    },
+    {
+      "epoch": 0.28895195379944033,
+      "grad_norm": 0.9190603494644165,
+      "learning_rate": 4.518403852769178e-05,
+      "loss": 0.9834,
+      "step": 3640
+    },
+    {
+      "epoch": 0.28974577784833994,
+      "grad_norm": 1.0563730001449585,
+      "learning_rate": 4.517080786430632e-05,
+      "loss": 0.903,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2905396018972395,
+      "grad_norm": 0.9505723714828491,
+      "learning_rate": 4.515757720092086e-05,
+      "loss": 0.9106,
+      "step": 3660
+    },
+    {
+      "epoch": 0.29133342594613904,
+      "grad_norm": 0.7892711758613586,
+      "learning_rate": 4.5144346537535396e-05,
+      "loss": 0.9156,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2921272499950386,
+      "grad_norm": 0.8885009288787842,
+      "learning_rate": 4.5131115874149935e-05,
+      "loss": 0.9297,
+      "step": 3680
+    },
+    {
+      "epoch": 0.29292107404393813,
+      "grad_norm": 0.9264837503433228,
+      "learning_rate": 4.5117885210764473e-05,
+      "loss": 0.8791,
+      "step": 3690
+    },
+    {
+      "epoch": 0.29371489809283774,
+      "grad_norm": 0.8464304208755493,
+      "learning_rate": 4.5104654547379006e-05,
+      "loss": 0.9119,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2945087221417373,
+      "grad_norm": 0.8379583358764648,
+      "learning_rate": 4.5091423883993545e-05,
+      "loss": 0.9339,
+      "step": 3710
+    },
+    {
+      "epoch": 0.29530254619063684,
+      "grad_norm": 0.7604143619537354,
+      "learning_rate": 4.5078193220608083e-05,
+      "loss": 0.8827,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2960963702395364,
+      "grad_norm": 1.2368942499160767,
+      "learning_rate": 4.506496255722262e-05,
+      "loss": 0.8412,
+      "step": 3730
+    },
+    {
+      "epoch": 0.296890194288436,
+      "grad_norm": 0.8182504773139954,
+      "learning_rate": 4.5051731893837155e-05,
+      "loss": 0.9742,
+      "step": 3740
+    },
+    {
+      "epoch": 0.29768401833733554,
+      "grad_norm": 0.6445258259773254,
+      "learning_rate": 4.5038501230451693e-05,
+      "loss": 0.9485,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2984778423862351,
+      "grad_norm": 0.535454511642456,
+      "learning_rate": 4.502527056706623e-05,
+      "loss": 0.8326,
+      "step": 3760
+    },
+    {
+      "epoch": 0.29927166643513464,
+      "grad_norm": 0.9186378121376038,
+      "learning_rate": 4.501203990368078e-05,
+      "loss": 0.8988,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3000654904840342,
+      "grad_norm": 0.8731489181518555,
+      "learning_rate": 4.499880924029531e-05,
+      "loss": 0.8857,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3008593145329338,
+      "grad_norm": 0.9101518392562866,
+      "learning_rate": 4.498557857690985e-05,
+      "loss": 0.8697,
+      "step": 3790
+    },
+    {
+      "epoch": 0.30165313858183335,
+      "grad_norm": 1.0505417585372925,
+      "learning_rate": 4.497234791352439e-05,
+      "loss": 0.9239,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3024469626307329,
+      "grad_norm": 0.9743961095809937,
+      "learning_rate": 4.495911725013893e-05,
+      "loss": 0.9206,
+      "step": 3810
+    },
+    {
+      "epoch": 0.30324078667963245,
+      "grad_norm": 0.7029954195022583,
+      "learning_rate": 4.494588658675346e-05,
+      "loss": 0.9278,
+      "step": 3820
+    },
+    {
+      "epoch": 0.304034610728532,
+      "grad_norm": 1.1363028287887573,
+      "learning_rate": 4.4932655923368e-05,
+      "loss": 0.8596,
+      "step": 3830
+    },
+    {
+      "epoch": 0.3048284347774316,
+      "grad_norm": 0.6260054707527161,
+      "learning_rate": 4.491942525998254e-05,
+      "loss": 0.8998,
+      "step": 3840
+    },
+    {
+      "epoch": 0.30562225882633115,
+      "grad_norm": 0.9502087831497192,
+      "learning_rate": 4.4906194596597076e-05,
+      "loss": 0.9423,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3064160828752307,
+      "grad_norm": 0.6927551627159119,
+      "learning_rate": 4.4892963933211615e-05,
+      "loss": 0.9297,
+      "step": 3860
+    },
+    {
+      "epoch": 0.30720990692413025,
+      "grad_norm": 0.9120237231254578,
+      "learning_rate": 4.4879733269826154e-05,
+      "loss": 0.916,
+      "step": 3870
+    },
+    {
+      "epoch": 0.30800373097302985,
+      "grad_norm": 0.7541207075119019,
+      "learning_rate": 4.486650260644069e-05,
+      "loss": 0.8848,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3087975550219294,
+      "grad_norm": 0.8690943717956543,
+      "learning_rate": 4.4853271943055225e-05,
+      "loss": 0.8941,
+      "step": 3890
+    },
+    {
+      "epoch": 0.30959137907082895,
+      "grad_norm": 0.9503733515739441,
+      "learning_rate": 4.4840041279669764e-05,
+      "loss": 0.9158,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3103852031197285,
+      "grad_norm": 0.8072488903999329,
+      "learning_rate": 4.48268106162843e-05,
+      "loss": 0.9325,
+      "step": 3910
+    },
+    {
+      "epoch": 0.31117902716862805,
+      "grad_norm": 0.9710533618927002,
+      "learning_rate": 4.481357995289884e-05,
+      "loss": 0.9011,
+      "step": 3920
+    },
+    {
+      "epoch": 0.31197285121752766,
+      "grad_norm": 0.8638473749160767,
+      "learning_rate": 4.4800349289513374e-05,
+      "loss": 0.9435,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3127666752664272,
+      "grad_norm": 1.0046641826629639,
+      "learning_rate": 4.478711862612792e-05,
+      "loss": 0.9496,
+      "step": 3940
+    },
+    {
+      "epoch": 0.31356049931532676,
+      "grad_norm": 0.7358148097991943,
+      "learning_rate": 4.477388796274246e-05,
+      "loss": 0.9531,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3143543233642263,
+      "grad_norm": 0.9239563941955566,
+      "learning_rate": 4.476065729935699e-05,
+      "loss": 0.9164,
+      "step": 3960
+    },
+    {
+      "epoch": 0.31514814741312586,
+      "grad_norm": 0.919695258140564,
+      "learning_rate": 4.474742663597153e-05,
+      "loss": 0.891,
+      "step": 3970
+    },
+    {
+      "epoch": 0.31594197146202546,
+      "grad_norm": 0.6320668458938599,
+      "learning_rate": 4.473419597258607e-05,
+      "loss": 0.9372,
+      "step": 3980
+    },
+    {
+      "epoch": 0.316735795510925,
+      "grad_norm": 0.8896647691726685,
+      "learning_rate": 4.472096530920061e-05,
+      "loss": 0.9236,
+      "step": 3990
+    },
+    {
+      "epoch": 0.31752961955982456,
+      "grad_norm": 0.8284244537353516,
+      "learning_rate": 4.470773464581514e-05,
+      "loss": 0.8625,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3183234436087241,
+      "grad_norm": 0.760328471660614,
+      "learning_rate": 4.469450398242968e-05,
+      "loss": 0.9433,
+      "step": 4010
+    },
+    {
+      "epoch": 0.31911726765762366,
+      "grad_norm": 0.7320681810379028,
+      "learning_rate": 4.468127331904422e-05,
+      "loss": 0.8655,
+      "step": 4020
+    },
+    {
+      "epoch": 0.31991109170652327,
+      "grad_norm": 0.9130124449729919,
+      "learning_rate": 4.4668042655658756e-05,
+      "loss": 0.8828,
+      "step": 4030
+    },
+    {
+      "epoch": 0.3207049157554228,
+      "grad_norm": 0.8663350343704224,
+      "learning_rate": 4.4654811992273295e-05,
+      "loss": 0.9523,
+      "step": 4040
+    },
+    {
+      "epoch": 0.32149873980432236,
+      "grad_norm": 0.8587914705276489,
+      "learning_rate": 4.4641581328887834e-05,
+      "loss": 0.8242,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3222925638532219,
+      "grad_norm": 0.7651414275169373,
+      "learning_rate": 4.462835066550237e-05,
+      "loss": 0.8989,
+      "step": 4060
+    },
+    {
+      "epoch": 0.3230863879021215,
+      "grad_norm": 0.834144651889801,
+      "learning_rate": 4.4615120002116905e-05,
+      "loss": 0.9105,
+      "step": 4070
+    },
+    {
+      "epoch": 0.32388021195102107,
+      "grad_norm": 0.691362202167511,
+      "learning_rate": 4.4601889338731444e-05,
+      "loss": 0.8757,
+      "step": 4080
+    },
+    {
+      "epoch": 0.3246740359999206,
+      "grad_norm": 0.9557787179946899,
+      "learning_rate": 4.458865867534598e-05,
+      "loss": 0.9276,
+      "step": 4090
+    },
+    {
+      "epoch": 0.32546786004882017,
+      "grad_norm": 0.8461413383483887,
+      "learning_rate": 4.457542801196052e-05,
+      "loss": 0.9158,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3262616840977197,
+      "grad_norm": 0.8489325046539307,
+      "learning_rate": 4.456219734857506e-05,
+      "loss": 0.9529,
+      "step": 4110
+    },
+    {
+      "epoch": 0.3270555081466193,
+      "grad_norm": 0.8827877044677734,
+      "learning_rate": 4.45489666851896e-05,
+      "loss": 0.9152,
+      "step": 4120
+    },
+    {
+      "epoch": 0.32784933219551887,
+      "grad_norm": 0.69962078332901,
+      "learning_rate": 4.453573602180414e-05,
+      "loss": 0.8927,
+      "step": 4130
+    },
+    {
+      "epoch": 0.3286431562444184,
+      "grad_norm": 0.6991488337516785,
+      "learning_rate": 4.452250535841868e-05,
+      "loss": 0.9344,
+      "step": 4140
+    },
+    {
+      "epoch": 0.32943698029331797,
+      "grad_norm": 0.9323925375938416,
+      "learning_rate": 4.450927469503321e-05,
+      "loss": 0.945,
+      "step": 4150
+    },
+    {
+      "epoch": 0.3302308043422175,
+      "grad_norm": 0.8266938328742981,
+      "learning_rate": 4.449604403164775e-05,
+      "loss": 0.9045,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3310246283911171,
+      "grad_norm": 0.8383837342262268,
+      "learning_rate": 4.448281336826229e-05,
+      "loss": 0.8516,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3318184524400167,
+      "grad_norm": 0.72791987657547,
+      "learning_rate": 4.446958270487682e-05,
+      "loss": 0.8718,
+      "step": 4180
+    },
+    {
+      "epoch": 0.3326122764889162,
+      "grad_norm": 0.7107085585594177,
+      "learning_rate": 4.445635204149136e-05,
+      "loss": 0.9216,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3334061005378158,
+      "grad_norm": 0.7751030325889587,
+      "learning_rate": 4.44431213781059e-05,
+      "loss": 0.8512,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3341999245867154,
+      "grad_norm": 0.764103889465332,
+      "learning_rate": 4.4429890714720443e-05,
+      "loss": 0.9594,
+      "step": 4210
+    },
+    {
+      "epoch": 0.33499374863561493,
+      "grad_norm": 0.7244476079940796,
+      "learning_rate": 4.4416660051334976e-05,
+      "loss": 0.9194,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3357875726845145,
+      "grad_norm": 0.8353381156921387,
+      "learning_rate": 4.4403429387949515e-05,
+      "loss": 0.8394,
+      "step": 4230
+    },
+    {
+      "epoch": 0.33658139673341403,
+      "grad_norm": 0.6536258459091187,
+      "learning_rate": 4.4390198724564053e-05,
+      "loss": 0.9125,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3373752207823136,
+      "grad_norm": 0.8908824920654297,
+      "learning_rate": 4.437696806117859e-05,
+      "loss": 0.9419,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3381690448312132,
+      "grad_norm": 0.6466046571731567,
+      "learning_rate": 4.4363737397793125e-05,
+      "loss": 0.946,
+      "step": 4260
+    },
+    {
+      "epoch": 0.33896286888011273,
+      "grad_norm": 0.8315856456756592,
+      "learning_rate": 4.4350506734407663e-05,
+      "loss": 0.9282,
+      "step": 4270
+    },
+    {
+      "epoch": 0.3397566929290123,
+      "grad_norm": 0.7369117736816406,
+      "learning_rate": 4.43372760710222e-05,
+      "loss": 0.9398,
+      "step": 4280
+    },
+    {
+      "epoch": 0.34055051697791183,
+      "grad_norm": 0.841690719127655,
+      "learning_rate": 4.432404540763674e-05,
+      "loss": 0.8411,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3413443410268114,
+      "grad_norm": 0.8048957586288452,
+      "learning_rate": 4.431081474425128e-05,
+      "loss": 0.9204,
+      "step": 4300
+    },
+    {
+      "epoch": 0.342138165075711,
+      "grad_norm": 0.6272882223129272,
+      "learning_rate": 4.429758408086582e-05,
+      "loss": 0.8819,
+      "step": 4310
+    },
+    {
+      "epoch": 0.34293198912461054,
+      "grad_norm": 0.8420210480690002,
+      "learning_rate": 4.428435341748036e-05,
+      "loss": 0.9187,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3437258131735101,
+      "grad_norm": 0.873335063457489,
+      "learning_rate": 4.427112275409489e-05,
+      "loss": 0.8807,
+      "step": 4330
+    },
+    {
+      "epoch": 0.34451963722240964,
+      "grad_norm": 0.6721493005752563,
+      "learning_rate": 4.425789209070943e-05,
+      "loss": 0.9068,
+      "step": 4340
+    },
+    {
+      "epoch": 0.34531346127130924,
+      "grad_norm": 0.8227477669715881,
+      "learning_rate": 4.424466142732397e-05,
+      "loss": 0.9372,
+      "step": 4350
+    },
+    {
+      "epoch": 0.3461072853202088,
+      "grad_norm": 0.8997884392738342,
+      "learning_rate": 4.423143076393851e-05,
+      "loss": 0.8916,
+      "step": 4360
+    },
+    {
+      "epoch": 0.34690110936910834,
+      "grad_norm": 0.9526328444480896,
+      "learning_rate": 4.421820010055304e-05,
+      "loss": 0.857,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3476949334180079,
+      "grad_norm": 0.8911690711975098,
+      "learning_rate": 4.4204969437167585e-05,
+      "loss": 0.9323,
+      "step": 4380
+    },
+    {
+      "epoch": 0.34848875746690744,
+      "grad_norm": 0.9047393798828125,
+      "learning_rate": 4.4191738773782124e-05,
+      "loss": 0.8946,
+      "step": 4390
+    },
+    {
+      "epoch": 0.34928258151580704,
+      "grad_norm": 0.8549637794494629,
+      "learning_rate": 4.4178508110396656e-05,
+      "loss": 0.8755,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3500764055647066,
+      "grad_norm": 0.9609139561653137,
+      "learning_rate": 4.4165277447011195e-05,
+      "loss": 0.8262,
+      "step": 4410
+    },
+    {
+      "epoch": 0.35087022961360614,
+      "grad_norm": 0.5445454120635986,
+      "learning_rate": 4.4152046783625734e-05,
+      "loss": 0.8849,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3516640536625057,
+      "grad_norm": 0.9729016423225403,
+      "learning_rate": 4.413881612024027e-05,
+      "loss": 0.8396,
+      "step": 4430
+    },
+    {
+      "epoch": 0.35245787771140524,
+      "grad_norm": 1.071895718574524,
+      "learning_rate": 4.4125585456854805e-05,
+      "loss": 0.9218,
+      "step": 4440
+    },
+    {
+      "epoch": 0.35325170176030485,
+      "grad_norm": 0.6580398082733154,
+      "learning_rate": 4.4112354793469344e-05,
+      "loss": 1.0065,
+      "step": 4450
+    },
+    {
+      "epoch": 0.3540455258092044,
+      "grad_norm": 0.6029400825500488,
+      "learning_rate": 4.409912413008388e-05,
+      "loss": 0.8869,
+      "step": 4460
+    },
+    {
+      "epoch": 0.35483934985810395,
+      "grad_norm": 1.124426007270813,
+      "learning_rate": 4.408589346669842e-05,
+      "loss": 0.8604,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3556331739070035,
+      "grad_norm": 0.905108630657196,
+      "learning_rate": 4.407266280331296e-05,
+      "loss": 0.8818,
+      "step": 4480
+    },
+    {
+      "epoch": 0.35642699795590305,
+      "grad_norm": 0.7963013052940369,
+      "learning_rate": 4.40594321399275e-05,
+      "loss": 0.9271,
+      "step": 4490
+    },
+    {
+      "epoch": 0.35722082200480265,
+      "grad_norm": 0.8501315712928772,
+      "learning_rate": 4.404620147654204e-05,
+      "loss": 0.9084,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3580146460537022,
+      "grad_norm": 0.7018426656723022,
+      "learning_rate": 4.403297081315657e-05,
+      "loss": 0.9059,
+      "step": 4510
+    },
+    {
+      "epoch": 0.35880847010260175,
+      "grad_norm": 0.9202073216438293,
+      "learning_rate": 4.401974014977111e-05,
+      "loss": 0.846,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3596022941515013,
+      "grad_norm": 1.0749320983886719,
+      "learning_rate": 4.400650948638565e-05,
+      "loss": 0.9123,
+      "step": 4530
+    },
+    {
+      "epoch": 0.3603961182004009,
+      "grad_norm": 0.7524193525314331,
+      "learning_rate": 4.399327882300019e-05,
+      "loss": 0.9018,
+      "step": 4540
+    },
+    {
+      "epoch": 0.36118994224930046,
+      "grad_norm": 0.7798032164573669,
+      "learning_rate": 4.398004815961472e-05,
+      "loss": 0.9221,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3619837662982,
+      "grad_norm": 0.8113389015197754,
+      "learning_rate": 4.3966817496229265e-05,
+      "loss": 0.9005,
+      "step": 4560
+    },
+    {
+      "epoch": 0.36277759034709955,
+      "grad_norm": 0.7713738083839417,
+      "learning_rate": 4.3953586832843804e-05,
+      "loss": 0.9608,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3635714143959991,
+      "grad_norm": 0.7880844473838806,
+      "learning_rate": 4.394035616945834e-05,
+      "loss": 0.9077,
+      "step": 4580
+    },
+    {
+      "epoch": 0.3643652384448987,
+      "grad_norm": 0.8188297748565674,
+      "learning_rate": 4.3927125506072875e-05,
+      "loss": 0.9276,
+      "step": 4590
+    },
+    {
+      "epoch": 0.36515906249379826,
+      "grad_norm": 0.7484593987464905,
+      "learning_rate": 4.3913894842687414e-05,
+      "loss": 0.926,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3659528865426978,
+      "grad_norm": 0.7410342693328857,
+      "learning_rate": 4.390066417930195e-05,
+      "loss": 0.9059,
+      "step": 4610
+    },
+    {
+      "epoch": 0.36674671059159736,
+      "grad_norm": 0.897441565990448,
+      "learning_rate": 4.388743351591649e-05,
+      "loss": 0.8394,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3675405346404969,
+      "grad_norm": 0.7051084637641907,
+      "learning_rate": 4.3874202852531024e-05,
+      "loss": 0.8888,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3683343586893965,
+      "grad_norm": 0.8222568035125732,
+      "learning_rate": 4.386097218914556e-05,
+      "loss": 0.8424,
+      "step": 4640
+    },
+    {
+      "epoch": 0.36912818273829606,
+      "grad_norm": 0.9209256768226624,
+      "learning_rate": 4.38477415257601e-05,
+      "loss": 0.8892,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3699220067871956,
+      "grad_norm": 0.9021577835083008,
+      "learning_rate": 4.383451086237464e-05,
+      "loss": 0.831,
+      "step": 4660
+    },
+    {
+      "epoch": 0.37071583083609516,
+      "grad_norm": 0.7472784519195557,
+      "learning_rate": 4.382128019898918e-05,
+      "loss": 0.8574,
+      "step": 4670
+    },
+    {
+      "epoch": 0.37150965488499477,
+      "grad_norm": 0.7072018384933472,
+      "learning_rate": 4.380804953560372e-05,
+      "loss": 0.9618,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3723034789338943,
+      "grad_norm": 0.9218171238899231,
+      "learning_rate": 4.379481887221826e-05,
+      "loss": 0.8916,
+      "step": 4690
+    },
+    {
+      "epoch": 0.37309730298279387,
+      "grad_norm": 0.8794512152671814,
+      "learning_rate": 4.378158820883279e-05,
+      "loss": 0.9359,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3738911270316934,
+      "grad_norm": 0.8208445906639099,
+      "learning_rate": 4.376835754544733e-05,
+      "loss": 0.8878,
+      "step": 4710
+    },
+    {
+      "epoch": 0.37468495108059297,
+      "grad_norm": 0.5891085863113403,
+      "learning_rate": 4.375512688206187e-05,
+      "loss": 0.9369,
+      "step": 4720
+    },
+    {
+      "epoch": 0.37547877512949257,
+      "grad_norm": 0.8617219924926758,
+      "learning_rate": 4.374189621867641e-05,
+      "loss": 0.8791,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3762725991783921,
+      "grad_norm": 1.0510433912277222,
+      "learning_rate": 4.3728665555290946e-05,
+      "loss": 0.987,
+      "step": 4740
+    },
+    {
+      "epoch": 0.37706642322729167,
+      "grad_norm": 0.9196320176124573,
+      "learning_rate": 4.3715434891905485e-05,
+      "loss": 0.8332,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3778602472761912,
+      "grad_norm": 0.9651428461074829,
+      "learning_rate": 4.3702204228520023e-05,
+      "loss": 0.9301,
+      "step": 4760
+    },
+    {
+      "epoch": 0.37865407132509077,
+      "grad_norm": 0.7794479131698608,
+      "learning_rate": 4.3688973565134556e-05,
+      "loss": 0.8898,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3794478953739904,
+      "grad_norm": 0.708845853805542,
+      "learning_rate": 4.3675742901749095e-05,
+      "loss": 0.8781,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3802417194228899,
+      "grad_norm": 1.07254159450531,
+      "learning_rate": 4.3662512238363633e-05,
+      "loss": 0.8329,
+      "step": 4790
+    },
+    {
+      "epoch": 0.3810355434717895,
+      "grad_norm": 0.6615691184997559,
+      "learning_rate": 4.364928157497817e-05,
+      "loss": 0.9107,
+      "step": 4800
+    },
+    {
+      "epoch": 0.381829367520689,
+      "grad_norm": 0.7933416962623596,
+      "learning_rate": 4.3636050911592705e-05,
+      "loss": 0.8986,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3826231915695886,
+      "grad_norm": 0.7650178074836731,
+      "learning_rate": 4.3622820248207243e-05,
+      "loss": 0.9212,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3834170156184882,
+      "grad_norm": 0.8654028177261353,
+      "learning_rate": 4.360958958482179e-05,
+      "loss": 0.8686,
+      "step": 4830
+    },
+    {
+      "epoch": 0.3842108396673877,
+      "grad_norm": 0.8981553912162781,
+      "learning_rate": 4.359635892143633e-05,
+      "loss": 0.843,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3850046637162873,
+      "grad_norm": 0.8608548641204834,
+      "learning_rate": 4.358312825805086e-05,
+      "loss": 0.8958,
+      "step": 4850
+    },
+    {
+      "epoch": 0.3857984877651868,
+      "grad_norm": 0.8663391470909119,
+      "learning_rate": 4.35698975946654e-05,
+      "loss": 0.8992,
+      "step": 4860
+    },
+    {
+      "epoch": 0.38659231181408643,
+      "grad_norm": 0.7639971971511841,
+      "learning_rate": 4.355666693127994e-05,
+      "loss": 0.8592,
+      "step": 4870
+    },
+    {
+      "epoch": 0.387386135862986,
+      "grad_norm": 0.5844544768333435,
+      "learning_rate": 4.354343626789447e-05,
+      "loss": 0.8638,
+      "step": 4880
+    },
+    {
+      "epoch": 0.38817995991188553,
+      "grad_norm": 0.8091756701469421,
+      "learning_rate": 4.353020560450901e-05,
+      "loss": 0.934,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3889737839607851,
+      "grad_norm": 0.8555530905723572,
+      "learning_rate": 4.351697494112355e-05,
+      "loss": 0.8541,
+      "step": 4900
+    },
+    {
+      "epoch": 0.38976760800968463,
+      "grad_norm": 0.7854059934616089,
+      "learning_rate": 4.350374427773809e-05,
+      "loss": 0.9043,
+      "step": 4910
+    },
+    {
+      "epoch": 0.39056143205858423,
+      "grad_norm": 1.0440559387207031,
+      "learning_rate": 4.3490513614352626e-05,
+      "loss": 0.8941,
+      "step": 4920
+    },
+    {
+      "epoch": 0.3913552561074838,
+      "grad_norm": 0.7478209733963013,
+      "learning_rate": 4.3477282950967165e-05,
+      "loss": 0.8565,
+      "step": 4930
+    },
+    {
+      "epoch": 0.39214908015638333,
+      "grad_norm": 0.7952112555503845,
+      "learning_rate": 4.3464052287581704e-05,
+      "loss": 0.8953,
+      "step": 4940
+    },
+    {
+      "epoch": 0.3929429042052829,
+      "grad_norm": 0.8431933522224426,
+      "learning_rate": 4.345082162419624e-05,
+      "loss": 0.8433,
+      "step": 4950
+    },
+    {
+      "epoch": 0.39373672825418243,
+      "grad_norm": 0.7418366074562073,
+      "learning_rate": 4.3437590960810775e-05,
+      "loss": 0.8713,
+      "step": 4960
+    },
+    {
+      "epoch": 0.39453055230308204,
+      "grad_norm": 0.8180972933769226,
+      "learning_rate": 4.3424360297425314e-05,
+      "loss": 0.9263,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3953243763519816,
+      "grad_norm": 0.681885302066803,
+      "learning_rate": 4.341112963403985e-05,
+      "loss": 0.8812,
+      "step": 4980
+    },
+    {
+      "epoch": 0.39611820040088114,
+      "grad_norm": 0.6465054750442505,
+      "learning_rate": 4.3397898970654385e-05,
+      "loss": 0.8943,
+      "step": 4990
+    },
+    {
+      "epoch": 0.3969120244497807,
+      "grad_norm": 0.8901807069778442,
+      "learning_rate": 4.338466830726893e-05,
+      "loss": 0.9004,
+      "step": 5000
+    },
+    {
+      "epoch": 0.3977058484986803,
+      "grad_norm": 0.937714695930481,
+      "learning_rate": 4.337143764388347e-05,
+      "loss": 0.8341,
+      "step": 5010
+    },
+    {
+      "epoch": 0.39849967254757984,
+      "grad_norm": 0.6198152303695679,
+      "learning_rate": 4.335820698049801e-05,
+      "loss": 0.9793,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3992934965964794,
+      "grad_norm": 1.0210473537445068,
+      "learning_rate": 4.334497631711254e-05,
+      "loss": 0.8939,
+      "step": 5030
+    },
+    {
+      "epoch": 0.40008732064537894,
+      "grad_norm": 0.7850756645202637,
+      "learning_rate": 4.333174565372708e-05,
+      "loss": 0.8545,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4008811446942785,
+      "grad_norm": 0.9468925595283508,
+      "learning_rate": 4.331851499034162e-05,
+      "loss": 0.833,
+      "step": 5050
+    },
+    {
+      "epoch": 0.4016749687431781,
+      "grad_norm": 0.6961553692817688,
+      "learning_rate": 4.330528432695616e-05,
+      "loss": 0.887,
+      "step": 5060
+    },
+    {
+      "epoch": 0.40246879279207765,
+      "grad_norm": 0.7382469177246094,
+      "learning_rate": 4.329205366357069e-05,
+      "loss": 0.8148,
+      "step": 5070
+    },
+    {
+      "epoch": 0.4032626168409772,
+      "grad_norm": 0.8620008230209351,
+      "learning_rate": 4.327882300018523e-05,
+      "loss": 0.8493,
+      "step": 5080
+    },
+    {
+      "epoch": 0.40405644088987674,
+      "grad_norm": 0.6228137612342834,
+      "learning_rate": 4.326559233679977e-05,
+      "loss": 0.961,
+      "step": 5090
+    },
+    {
+      "epoch": 0.4048502649387763,
+      "grad_norm": 0.8809449672698975,
+      "learning_rate": 4.3252361673414306e-05,
+      "loss": 0.8982,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4056440889876759,
+      "grad_norm": 0.7898284196853638,
+      "learning_rate": 4.3239131010028845e-05,
+      "loss": 0.8696,
+      "step": 5110
+    },
+    {
+      "epoch": 0.40643791303657545,
+      "grad_norm": 0.5993707776069641,
+      "learning_rate": 4.3225900346643384e-05,
+      "loss": 0.9455,
+      "step": 5120
+    },
+    {
+      "epoch": 0.407231737085475,
+      "grad_norm": 0.5548946261405945,
+      "learning_rate": 4.321266968325792e-05,
+      "loss": 0.9056,
+      "step": 5130
+    },
+    {
+      "epoch": 0.40802556113437455,
+      "grad_norm": 0.5974055528640747,
+      "learning_rate": 4.3199439019872455e-05,
+      "loss": 0.933,
+      "step": 5140
+    },
+    {
+      "epoch": 0.40881938518327415,
+      "grad_norm": 0.77692711353302,
+      "learning_rate": 4.3186208356486994e-05,
+      "loss": 0.9088,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4096132092321737,
+      "grad_norm": 0.6619086265563965,
+      "learning_rate": 4.317297769310153e-05,
+      "loss": 0.9095,
+      "step": 5160
+    },
+    {
+      "epoch": 0.41040703328107325,
+      "grad_norm": 0.8819933533668518,
+      "learning_rate": 4.315974702971607e-05,
+      "loss": 0.8546,
+      "step": 5170
+    },
+    {
+      "epoch": 0.4112008573299728,
+      "grad_norm": 0.8302019834518433,
+      "learning_rate": 4.314651636633061e-05,
+      "loss": 0.883,
+      "step": 5180
+    },
+    {
+      "epoch": 0.41199468137887235,
+      "grad_norm": 0.9397719502449036,
+      "learning_rate": 4.313328570294515e-05,
+      "loss": 0.8768,
+      "step": 5190
+    },
+    {
+      "epoch": 0.41278850542777196,
+      "grad_norm": 1.071415901184082,
+      "learning_rate": 4.312005503955969e-05,
+      "loss": 0.9198,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4135823294766715,
+      "grad_norm": 0.7499683499336243,
+      "learning_rate": 4.310682437617422e-05,
+      "loss": 0.875,
+      "step": 5210
+    },
+    {
+      "epoch": 0.41437615352557106,
+      "grad_norm": 0.7693167328834534,
+      "learning_rate": 4.309359371278876e-05,
+      "loss": 0.898,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4151699775744706,
+      "grad_norm": 1.0537440776824951,
+      "learning_rate": 4.30803630494033e-05,
+      "loss": 0.9159,
+      "step": 5230
+    },
+    {
+      "epoch": 0.41596380162337016,
+      "grad_norm": 0.7521883249282837,
+      "learning_rate": 4.306713238601784e-05,
+      "loss": 0.9153,
+      "step": 5240
+    },
+    {
+      "epoch": 0.41675762567226976,
+      "grad_norm": 1.0054055452346802,
+      "learning_rate": 4.305390172263237e-05,
+      "loss": 0.8956,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4175514497211693,
+      "grad_norm": 0.7442034482955933,
+      "learning_rate": 4.304067105924691e-05,
+      "loss": 0.891,
+      "step": 5260
+    },
+    {
+      "epoch": 0.41834527377006886,
+      "grad_norm": 1.084252953529358,
+      "learning_rate": 4.3027440395861455e-05,
+      "loss": 0.9027,
+      "step": 5270
+    },
+    {
+      "epoch": 0.4191390978189684,
+      "grad_norm": 0.8177867531776428,
+      "learning_rate": 4.3014209732475993e-05,
+      "loss": 0.9018,
+      "step": 5280
+    },
+    {
+      "epoch": 0.419932921867868,
+      "grad_norm": 0.48281505703926086,
+      "learning_rate": 4.3000979069090526e-05,
+      "loss": 0.9254,
+      "step": 5290
+    },
+    {
+      "epoch": 0.42072674591676756,
+      "grad_norm": 0.7245819568634033,
+      "learning_rate": 4.2987748405705065e-05,
+      "loss": 0.9109,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4215205699656671,
+      "grad_norm": 0.7288056015968323,
+      "learning_rate": 4.2974517742319603e-05,
+      "loss": 0.8432,
+      "step": 5310
+    },
+    {
+      "epoch": 0.42231439401456666,
+      "grad_norm": 0.7283946871757507,
+      "learning_rate": 4.2961287078934136e-05,
+      "loss": 0.8788,
+      "step": 5320
+    },
+    {
+      "epoch": 0.4231082180634662,
+      "grad_norm": 0.7446802258491516,
+      "learning_rate": 4.2948056415548675e-05,
+      "loss": 0.9142,
+      "step": 5330
+    },
+    {
+      "epoch": 0.4239020421123658,
+      "grad_norm": 1.0653091669082642,
+      "learning_rate": 4.2934825752163213e-05,
+      "loss": 0.8688,
+      "step": 5340
+    },
+    {
+      "epoch": 0.42469586616126537,
+      "grad_norm": 0.702083170413971,
+      "learning_rate": 4.292159508877775e-05,
+      "loss": 0.9083,
+      "step": 5350
+    },
+    {
+      "epoch": 0.4254896902101649,
+      "grad_norm": 0.8971485495567322,
+      "learning_rate": 4.290836442539229e-05,
+      "loss": 0.8935,
+      "step": 5360
+    },
+    {
+      "epoch": 0.42628351425906447,
+      "grad_norm": 0.7836154103279114,
+      "learning_rate": 4.289513376200683e-05,
+      "loss": 0.814,
+      "step": 5370
+    },
+    {
+      "epoch": 0.427077338307964,
+      "grad_norm": 0.964614987373352,
+      "learning_rate": 4.288190309862137e-05,
+      "loss": 0.8964,
+      "step": 5380
+    },
+    {
+      "epoch": 0.4278711623568636,
+      "grad_norm": 0.8366977572441101,
+      "learning_rate": 4.286867243523591e-05,
+      "loss": 0.9089,
+      "step": 5390
+    },
+    {
+      "epoch": 0.42866498640576317,
+      "grad_norm": 0.720564603805542,
+      "learning_rate": 4.285544177185044e-05,
+      "loss": 0.905,
+      "step": 5400
+    },
+    {
+      "epoch": 0.4294588104546627,
+      "grad_norm": 0.7407063841819763,
+      "learning_rate": 4.284221110846498e-05,
+      "loss": 0.9191,
+      "step": 5410
+    },
+    {
+      "epoch": 0.43025263450356227,
+      "grad_norm": 0.7772669792175293,
+      "learning_rate": 4.282898044507952e-05,
+      "loss": 0.8633,
+      "step": 5420
+    },
+    {
+      "epoch": 0.4310464585524618,
+      "grad_norm": 0.7156360149383545,
+      "learning_rate": 4.281574978169406e-05,
+      "loss": 0.9515,
+      "step": 5430
+    },
+    {
+      "epoch": 0.4318402826013614,
+      "grad_norm": 0.7044875621795654,
+      "learning_rate": 4.2802519118308596e-05,
+      "loss": 0.9199,
+      "step": 5440
+    },
+    {
+      "epoch": 0.432634106650261,
+      "grad_norm": 0.7540838122367859,
+      "learning_rate": 4.2789288454923135e-05,
+      "loss": 0.9229,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4334279306991605,
+      "grad_norm": 0.6620701551437378,
+      "learning_rate": 4.2776057791537674e-05,
+      "loss": 0.8556,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4342217547480601,
+      "grad_norm": 0.8868663907051086,
+      "learning_rate": 4.2762827128152206e-05,
+      "loss": 0.8666,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4350155787969597,
+      "grad_norm": 0.7081343531608582,
+      "learning_rate": 4.2749596464766745e-05,
+      "loss": 0.8888,
+      "step": 5480
+    },
+    {
+      "epoch": 0.43580940284585923,
+      "grad_norm": 0.7983565926551819,
+      "learning_rate": 4.2736365801381284e-05,
+      "loss": 0.9041,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4366032268947588,
+      "grad_norm": 0.7562633156776428,
+      "learning_rate": 4.272313513799582e-05,
+      "loss": 0.908,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4373970509436583,
+      "grad_norm": 0.8235949873924255,
+      "learning_rate": 4.2709904474610355e-05,
+      "loss": 0.8907,
+      "step": 5510
+    },
+    {
+      "epoch": 0.4381908749925579,
+      "grad_norm": 0.6528277397155762,
+      "learning_rate": 4.2696673811224894e-05,
+      "loss": 0.8403,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4389846990414575,
+      "grad_norm": 0.7162795662879944,
+      "learning_rate": 4.268344314783943e-05,
+      "loss": 0.9001,
+      "step": 5530
+    },
+    {
+      "epoch": 0.43977852309035703,
+      "grad_norm": 0.8113960027694702,
+      "learning_rate": 4.267021248445398e-05,
+      "loss": 0.8785,
+      "step": 5540
+    },
+    {
+      "epoch": 0.4405723471392566,
+      "grad_norm": 0.6791692972183228,
+      "learning_rate": 4.265698182106851e-05,
+      "loss": 0.8744,
+      "step": 5550
+    },
+    {
+      "epoch": 0.44136617118815613,
+      "grad_norm": 0.7448859214782715,
+      "learning_rate": 4.264375115768305e-05,
+      "loss": 0.9191,
+      "step": 5560
+    },
+    {
+      "epoch": 0.4421599952370557,
+      "grad_norm": 0.8218660950660706,
+      "learning_rate": 4.263052049429759e-05,
+      "loss": 0.8693,
+      "step": 5570
+    },
+    {
+      "epoch": 0.4429538192859553,
+      "grad_norm": 0.8838844299316406,
+      "learning_rate": 4.261728983091212e-05,
+      "loss": 0.8701,
+      "step": 5580
+    },
+    {
+      "epoch": 0.44374764333485484,
+      "grad_norm": 0.9519463181495667,
+      "learning_rate": 4.260405916752666e-05,
+      "loss": 0.8621,
+      "step": 5590
+    },
+    {
+      "epoch": 0.4445414673837544,
+      "grad_norm": 0.9154829382896423,
+      "learning_rate": 4.25908285041412e-05,
+      "loss": 0.8917,
+      "step": 5600
+    },
+    {
+      "epoch": 0.44533529143265393,
+      "grad_norm": 1.03704833984375,
+      "learning_rate": 4.257759784075574e-05,
+      "loss": 0.9156,
+      "step": 5610
+    },
+    {
+      "epoch": 0.44612911548155354,
+      "grad_norm": 0.7216574549674988,
+      "learning_rate": 4.2564367177370276e-05,
+      "loss": 0.8858,
+      "step": 5620
+    },
+    {
+      "epoch": 0.4469229395304531,
+      "grad_norm": 0.6714047789573669,
+      "learning_rate": 4.2551136513984815e-05,
+      "loss": 0.8379,
+      "step": 5630
+    },
+    {
+      "epoch": 0.44771676357935264,
+      "grad_norm": 0.9155183434486389,
+      "learning_rate": 4.2537905850599354e-05,
+      "loss": 0.871,
+      "step": 5640
+    },
+    {
+      "epoch": 0.4485105876282522,
+      "grad_norm": 0.7604344487190247,
+      "learning_rate": 4.252467518721389e-05,
+      "loss": 0.9569,
+      "step": 5650
+    },
+    {
+      "epoch": 0.44930441167715174,
+      "grad_norm": 0.6832132935523987,
+      "learning_rate": 4.2511444523828425e-05,
+      "loss": 0.9031,
+      "step": 5660
+    },
+    {
+      "epoch": 0.45009823572605134,
+      "grad_norm": 0.5372611880302429,
+      "learning_rate": 4.2498213860442964e-05,
+      "loss": 0.9364,
+      "step": 5670
+    },
+    {
+      "epoch": 0.4508920597749509,
+      "grad_norm": 0.8363433480262756,
+      "learning_rate": 4.24849831970575e-05,
+      "loss": 0.8708,
+      "step": 5680
+    },
+    {
+      "epoch": 0.45168588382385044,
+      "grad_norm": 0.7453399896621704,
+      "learning_rate": 4.2471752533672035e-05,
+      "loss": 0.8366,
+      "step": 5690
+    },
+    {
+      "epoch": 0.45247970787275,
+      "grad_norm": 0.9587568044662476,
+      "learning_rate": 4.2458521870286574e-05,
+      "loss": 0.884,
+      "step": 5700
+    },
+    {
+      "epoch": 0.45327353192164954,
+      "grad_norm": 0.9268891215324402,
+      "learning_rate": 4.244529120690112e-05,
+      "loss": 0.8884,
+      "step": 5710
+    },
+    {
+      "epoch": 0.45406735597054915,
+      "grad_norm": 0.7946332693099976,
+      "learning_rate": 4.243206054351566e-05,
+      "loss": 0.8686,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4548611800194487,
+      "grad_norm": 0.7091253995895386,
+      "learning_rate": 4.241882988013019e-05,
+      "loss": 0.8849,
+      "step": 5730
+    },
+    {
+      "epoch": 0.45565500406834825,
+      "grad_norm": 0.7621028423309326,
+      "learning_rate": 4.240559921674473e-05,
+      "loss": 0.9082,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4564488281172478,
+      "grad_norm": 0.8027592301368713,
+      "learning_rate": 4.239236855335927e-05,
+      "loss": 0.8613,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4572426521661474,
+      "grad_norm": 0.7330886721611023,
+      "learning_rate": 4.237913788997381e-05,
+      "loss": 0.9076,
+      "step": 5760
+    },
+    {
+      "epoch": 0.45803647621504695,
+      "grad_norm": 0.9339755177497864,
+      "learning_rate": 4.236590722658834e-05,
+      "loss": 0.8422,
+      "step": 5770
+    },
+    {
+      "epoch": 0.4588303002639465,
+      "grad_norm": 0.810807466506958,
+      "learning_rate": 4.235267656320288e-05,
+      "loss": 1.0052,
+      "step": 5780
+    },
+    {
+      "epoch": 0.45962412431284605,
+      "grad_norm": 0.6460042595863342,
+      "learning_rate": 4.233944589981742e-05,
+      "loss": 0.9634,
+      "step": 5790
+    },
+    {
+      "epoch": 0.4604179483617456,
+      "grad_norm": 0.9872604012489319,
+      "learning_rate": 4.232621523643196e-05,
+      "loss": 0.8762,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4612117724106452,
+      "grad_norm": 0.6482599973678589,
+      "learning_rate": 4.2312984573046496e-05,
+      "loss": 0.8439,
+      "step": 5810
+    },
+    {
+      "epoch": 0.46200559645954475,
+      "grad_norm": 0.7754231691360474,
+      "learning_rate": 4.2299753909661035e-05,
+      "loss": 0.9037,
+      "step": 5820
+    },
+    {
+      "epoch": 0.4627994205084443,
+      "grad_norm": 0.7624156475067139,
+      "learning_rate": 4.2286523246275573e-05,
+      "loss": 0.8971,
+      "step": 5830
+    },
+    {
+      "epoch": 0.46359324455734385,
+      "grad_norm": 0.8388280272483826,
+      "learning_rate": 4.2273292582890106e-05,
+      "loss": 0.8161,
+      "step": 5840
+    },
+    {
+      "epoch": 0.4643870686062434,
+      "grad_norm": 0.9448870420455933,
+      "learning_rate": 4.2260061919504645e-05,
+      "loss": 0.8949,
+      "step": 5850
+    },
+    {
+      "epoch": 0.465180892655143,
+      "grad_norm": 0.9757259488105774,
+      "learning_rate": 4.2246831256119183e-05,
+      "loss": 0.8765,
+      "step": 5860
+    },
+    {
+      "epoch": 0.46597471670404256,
+      "grad_norm": 0.7201075553894043,
+      "learning_rate": 4.223360059273372e-05,
+      "loss": 0.9155,
+      "step": 5870
+    },
+    {
+      "epoch": 0.4667685407529421,
+      "grad_norm": 0.9070501923561096,
+      "learning_rate": 4.222036992934826e-05,
+      "loss": 0.9414,
+      "step": 5880
+    },
+    {
+      "epoch": 0.46756236480184166,
+      "grad_norm": 0.6421061754226685,
+      "learning_rate": 4.22071392659628e-05,
+      "loss": 0.8891,
+      "step": 5890
+    },
+    {
+      "epoch": 0.4683561888507412,
+      "grad_norm": 0.957310140132904,
+      "learning_rate": 4.219390860257734e-05,
+      "loss": 0.9296,
+      "step": 5900
+    },
+    {
+      "epoch": 0.4691500128996408,
+      "grad_norm": 0.9540489315986633,
+      "learning_rate": 4.218067793919187e-05,
+      "loss": 0.8695,
+      "step": 5910
+    },
+    {
+      "epoch": 0.46994383694854036,
+      "grad_norm": 0.7519145011901855,
+      "learning_rate": 4.216744727580641e-05,
+      "loss": 0.874,
+      "step": 5920
+    },
+    {
+      "epoch": 0.4707376609974399,
+      "grad_norm": 0.6650322675704956,
+      "learning_rate": 4.215421661242095e-05,
+      "loss": 0.9093,
+      "step": 5930
+    },
+    {
+      "epoch": 0.47153148504633946,
+      "grad_norm": 0.6322752833366394,
+      "learning_rate": 4.214098594903549e-05,
+      "loss": 0.8887,
+      "step": 5940
+    },
+    {
+      "epoch": 0.47232530909523907,
+      "grad_norm": 0.8493393063545227,
+      "learning_rate": 4.212775528565002e-05,
+      "loss": 0.858,
+      "step": 5950
+    },
+    {
+      "epoch": 0.4731191331441386,
+      "grad_norm": 0.8725616931915283,
+      "learning_rate": 4.211452462226456e-05,
+      "loss": 0.8437,
+      "step": 5960
+    },
+    {
+      "epoch": 0.47391295719303816,
+      "grad_norm": 1.01144540309906,
+      "learning_rate": 4.21012939588791e-05,
+      "loss": 0.8824,
+      "step": 5970
+    },
+    {
+      "epoch": 0.4747067812419377,
+      "grad_norm": 0.8524407744407654,
+      "learning_rate": 4.2088063295493644e-05,
+      "loss": 0.8566,
+      "step": 5980
+    },
+    {
+      "epoch": 0.47550060529083726,
+      "grad_norm": 1.0122267007827759,
+      "learning_rate": 4.2074832632108176e-05,
+      "loss": 0.8802,
+      "step": 5990
+    },
+    {
+      "epoch": 0.47629442933973687,
+      "grad_norm": 0.930794894695282,
+      "learning_rate": 4.2061601968722715e-05,
+      "loss": 0.8751,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4770882533886364,
+      "grad_norm": 1.0592617988586426,
+      "learning_rate": 4.2048371305337254e-05,
+      "loss": 0.8464,
+      "step": 6010
+    },
+    {
+      "epoch": 0.47788207743753597,
+      "grad_norm": 0.7701418995857239,
+      "learning_rate": 4.2035140641951786e-05,
+      "loss": 0.8491,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4786759014864355,
+      "grad_norm": 0.7711431384086609,
+      "learning_rate": 4.2021909978566325e-05,
+      "loss": 0.8827,
+      "step": 6030
+    },
+    {
+      "epoch": 0.47946972553533507,
+      "grad_norm": 0.5670927166938782,
+      "learning_rate": 4.2008679315180864e-05,
+      "loss": 0.911,
+      "step": 6040
+    },
+    {
+      "epoch": 0.4802635495842347,
+      "grad_norm": 0.8850795030593872,
+      "learning_rate": 4.19954486517954e-05,
+      "loss": 0.8703,
+      "step": 6050
+    },
+    {
+      "epoch": 0.4810573736331342,
+      "grad_norm": 0.7688451409339905,
+      "learning_rate": 4.198221798840994e-05,
+      "loss": 0.8591,
+      "step": 6060
+    },
+    {
+      "epoch": 0.48185119768203377,
+      "grad_norm": 0.7370510101318359,
+      "learning_rate": 4.196898732502448e-05,
+      "loss": 0.8814,
+      "step": 6070
+    },
+    {
+      "epoch": 0.4826450217309333,
+      "grad_norm": 0.8961367607116699,
+      "learning_rate": 4.195575666163902e-05,
+      "loss": 0.9584,
+      "step": 6080
+    },
+    {
+      "epoch": 0.4834388457798329,
+      "grad_norm": 0.547537624835968,
+      "learning_rate": 4.194252599825356e-05,
+      "loss": 0.8956,
+      "step": 6090
+    },
+    {
+      "epoch": 0.4842326698287325,
+      "grad_norm": 0.7516084909439087,
+      "learning_rate": 4.192929533486809e-05,
+      "loss": 0.8988,
+      "step": 6100
+    },
+    {
+      "epoch": 0.485026493877632,
+      "grad_norm": 0.856939971446991,
+      "learning_rate": 4.191606467148263e-05,
+      "loss": 0.9231,
+      "step": 6110
+    },
+    {
+      "epoch": 0.4858203179265316,
+      "grad_norm": 0.8620836734771729,
+      "learning_rate": 4.190283400809717e-05,
+      "loss": 0.8834,
+      "step": 6120
+    },
+    {
+      "epoch": 0.4866141419754311,
+      "grad_norm": 0.7581316828727722,
+      "learning_rate": 4.18896033447117e-05,
+      "loss": 0.8127,
+      "step": 6130
+    },
+    {
+      "epoch": 0.48740796602433073,
+      "grad_norm": 0.6257145404815674,
+      "learning_rate": 4.187637268132624e-05,
+      "loss": 0.8563,
+      "step": 6140
+    },
+    {
+      "epoch": 0.4882017900732303,
+      "grad_norm": 0.8719417452812195,
+      "learning_rate": 4.186314201794078e-05,
+      "loss": 0.9042,
+      "step": 6150
+    },
+    {
+      "epoch": 0.48899561412212983,
+      "grad_norm": 0.8335233926773071,
+      "learning_rate": 4.1849911354555324e-05,
+      "loss": 0.9324,
+      "step": 6160
+    },
+    {
+      "epoch": 0.4897894381710294,
+      "grad_norm": 0.8433477878570557,
+      "learning_rate": 4.1836680691169856e-05,
+      "loss": 0.8667,
+      "step": 6170
+    },
+    {
+      "epoch": 0.49058326221992893,
+      "grad_norm": 0.5534406900405884,
+      "learning_rate": 4.1823450027784395e-05,
+      "loss": 0.8776,
+      "step": 6180
+    },
+    {
+      "epoch": 0.49137708626882853,
+      "grad_norm": 0.6916033029556274,
+      "learning_rate": 4.1810219364398934e-05,
+      "loss": 0.8879,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4921709103177281,
+      "grad_norm": 1.0233070850372314,
+      "learning_rate": 4.179698870101347e-05,
+      "loss": 0.915,
+      "step": 6200
+    },
+    {
+      "epoch": 0.49296473436662763,
+      "grad_norm": 0.8879387974739075,
+      "learning_rate": 4.1783758037628005e-05,
+      "loss": 0.8883,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4937585584155272,
+      "grad_norm": 0.8336936235427856,
+      "learning_rate": 4.1770527374242544e-05,
+      "loss": 0.8417,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4945523824644268,
+      "grad_norm": 0.8911527395248413,
+      "learning_rate": 4.175729671085708e-05,
+      "loss": 0.7989,
+      "step": 6230
+    },
+    {
+      "epoch": 0.49534620651332634,
+      "grad_norm": 0.5336394906044006,
+      "learning_rate": 4.174406604747162e-05,
+      "loss": 0.9055,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4961400305622259,
+      "grad_norm": 0.8444200158119202,
+      "learning_rate": 4.173083538408616e-05,
+      "loss": 0.8582,
+      "step": 6250
+    },
+    {
+      "epoch": 0.49693385461112544,
+      "grad_norm": 0.9549778699874878,
+      "learning_rate": 4.17176047207007e-05,
+      "loss": 0.9263,
+      "step": 6260
+    },
+    {
+      "epoch": 0.497727678660025,
+      "grad_norm": 0.8041070699691772,
+      "learning_rate": 4.170437405731524e-05,
+      "loss": 0.8669,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4985215027089246,
+      "grad_norm": 0.7747896313667297,
+      "learning_rate": 4.169114339392977e-05,
+      "loss": 0.8525,
+      "step": 6280
+    },
+    {
+      "epoch": 0.49931532675782414,
+      "grad_norm": 0.6687043309211731,
+      "learning_rate": 4.167791273054431e-05,
+      "loss": 0.9159,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5001091508067237,
+      "grad_norm": 0.725878119468689,
+      "learning_rate": 4.166468206715885e-05,
+      "loss": 0.9252,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5009029748556233,
+      "grad_norm": 0.9067284464836121,
+      "learning_rate": 4.165145140377339e-05,
+      "loss": 0.8905,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5016967989045228,
+      "grad_norm": 0.8240019083023071,
+      "learning_rate": 4.163822074038792e-05,
+      "loss": 0.8641,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5024906229534224,
+      "grad_norm": 0.7333963513374329,
+      "learning_rate": 4.1624990077002466e-05,
+      "loss": 0.8823,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5032844470023219,
+      "grad_norm": 1.1056770086288452,
+      "learning_rate": 4.1611759413617005e-05,
+      "loss": 0.8817,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5040782710512215,
+      "grad_norm": 0.8191234469413757,
+      "learning_rate": 4.1598528750231543e-05,
+      "loss": 0.8986,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5048720951001211,
+      "grad_norm": 0.6650758981704712,
+      "learning_rate": 4.1585298086846076e-05,
+      "loss": 0.885,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5056659191490206,
+      "grad_norm": 0.8753699064254761,
+      "learning_rate": 4.1572067423460615e-05,
+      "loss": 0.8402,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5064597431979202,
+      "grad_norm": 0.8924400806427002,
+      "learning_rate": 4.1558836760075153e-05,
+      "loss": 0.9016,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5072535672468197,
+      "grad_norm": 0.7312005758285522,
+      "learning_rate": 4.1545606096689686e-05,
+      "loss": 0.914,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5080473912957193,
+      "grad_norm": 0.771732747554779,
+      "learning_rate": 4.1532375433304225e-05,
+      "loss": 0.8976,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5088412153446189,
+      "grad_norm": 0.7958862781524658,
+      "learning_rate": 4.1519144769918763e-05,
+      "loss": 0.8749,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5096350393935184,
+      "grad_norm": 0.751964271068573,
+      "learning_rate": 4.15059141065333e-05,
+      "loss": 0.852,
+      "step": 6420
+    },
+    {
+      "epoch": 0.510428863442418,
+      "grad_norm": 0.721443772315979,
+      "learning_rate": 4.149268344314784e-05,
+      "loss": 0.8887,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5112226874913175,
+      "grad_norm": 0.6288412809371948,
+      "learning_rate": 4.147945277976238e-05,
+      "loss": 0.8488,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5120165115402171,
+      "grad_norm": 0.7060637474060059,
+      "learning_rate": 4.146622211637692e-05,
+      "loss": 0.8506,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5128103355891167,
+      "grad_norm": 0.5948452949523926,
+      "learning_rate": 4.145299145299146e-05,
+      "loss": 0.8759,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5136041596380162,
+      "grad_norm": 0.7204840183258057,
+      "learning_rate": 4.143976078960599e-05,
+      "loss": 0.9107,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5143979836869158,
+      "grad_norm": 0.9579210877418518,
+      "learning_rate": 4.142653012622053e-05,
+      "loss": 0.7921,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5151918077358154,
+      "grad_norm": 0.8127973079681396,
+      "learning_rate": 4.141329946283507e-05,
+      "loss": 0.8468,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5159856317847149,
+      "grad_norm": 0.6830429434776306,
+      "learning_rate": 4.140006879944961e-05,
+      "loss": 0.9339,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5167794558336145,
+      "grad_norm": 0.8261887431144714,
+      "learning_rate": 4.1386838136064146e-05,
+      "loss": 0.9211,
+      "step": 6510
+    },
+    {
+      "epoch": 0.517573279882514,
+      "grad_norm": 0.803862988948822,
+      "learning_rate": 4.1373607472678685e-05,
+      "loss": 0.883,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5183671039314136,
+      "grad_norm": 0.971078097820282,
+      "learning_rate": 4.1360376809293224e-05,
+      "loss": 0.8913,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5191609279803132,
+      "grad_norm": 0.8098673224449158,
+      "learning_rate": 4.1347146145907756e-05,
+      "loss": 0.9281,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5199547520292127,
+      "grad_norm": 0.9599949717521667,
+      "learning_rate": 4.1333915482522295e-05,
+      "loss": 0.8459,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5207485760781123,
+      "grad_norm": 0.7749876379966736,
+      "learning_rate": 4.1320684819136834e-05,
+      "loss": 0.8717,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5215424001270118,
+      "grad_norm": 0.9038444757461548,
+      "learning_rate": 4.130745415575137e-05,
+      "loss": 0.8716,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5223362241759114,
+      "grad_norm": 0.9091252088546753,
+      "learning_rate": 4.1294223492365905e-05,
+      "loss": 0.8709,
+      "step": 6580
+    },
+    {
+      "epoch": 0.523130048224811,
+      "grad_norm": 0.8427479863166809,
+      "learning_rate": 4.1280992828980444e-05,
+      "loss": 0.905,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5239238722737105,
+      "grad_norm": 0.6537529230117798,
+      "learning_rate": 4.126776216559499e-05,
+      "loss": 0.8252,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5247176963226101,
+      "grad_norm": 0.7358630895614624,
+      "learning_rate": 4.125453150220952e-05,
+      "loss": 0.9075,
+      "step": 6610
+    },
+    {
+      "epoch": 0.5255115203715096,
+      "grad_norm": 0.9415682554244995,
+      "learning_rate": 4.124130083882406e-05,
+      "loss": 0.8233,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5263053444204092,
+      "grad_norm": 0.790571928024292,
+      "learning_rate": 4.12280701754386e-05,
+      "loss": 0.9097,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5270991684693088,
+      "grad_norm": 0.9315680861473083,
+      "learning_rate": 4.121483951205314e-05,
+      "loss": 0.8962,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5278929925182083,
+      "grad_norm": 0.8426125645637512,
+      "learning_rate": 4.120160884866767e-05,
+      "loss": 0.8931,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5286868165671079,
+      "grad_norm": 0.7593029141426086,
+      "learning_rate": 4.118837818528221e-05,
+      "loss": 0.8584,
+      "step": 6660
+    },
+    {
+      "epoch": 0.5294806406160074,
+      "grad_norm": 0.719986617565155,
+      "learning_rate": 4.117514752189675e-05,
+      "loss": 0.9337,
+      "step": 6670
+    },
+    {
+      "epoch": 0.530274464664907,
+      "grad_norm": 0.8458060026168823,
+      "learning_rate": 4.116191685851129e-05,
+      "loss": 0.8355,
+      "step": 6680
+    },
+    {
+      "epoch": 0.5310682887138066,
+      "grad_norm": 0.7757362127304077,
+      "learning_rate": 4.1148686195125826e-05,
+      "loss": 0.897,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5318621127627061,
+      "grad_norm": 0.7549418807029724,
+      "learning_rate": 4.1135455531740365e-05,
+      "loss": 0.8864,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5326559368116057,
+      "grad_norm": 0.8752275705337524,
+      "learning_rate": 4.1122224868354904e-05,
+      "loss": 0.902,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5334497608605052,
+      "grad_norm": 0.901665985584259,
+      "learning_rate": 4.1108994204969436e-05,
+      "loss": 0.8892,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5342435849094048,
+      "grad_norm": 0.8543225526809692,
+      "learning_rate": 4.1095763541583975e-05,
+      "loss": 0.8923,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5350374089583044,
+      "grad_norm": 0.8486155867576599,
+      "learning_rate": 4.1082532878198514e-05,
+      "loss": 0.9043,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5358312330072039,
+      "grad_norm": 0.869857668876648,
+      "learning_rate": 4.106930221481305e-05,
+      "loss": 0.8406,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5366250570561035,
+      "grad_norm": 0.9113563299179077,
+      "learning_rate": 4.1056071551427585e-05,
+      "loss": 0.9667,
+      "step": 6760
+    },
+    {
+      "epoch": 0.537418881105003,
+      "grad_norm": 0.9034312963485718,
+      "learning_rate": 4.104284088804213e-05,
+      "loss": 0.8792,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5382127051539026,
+      "grad_norm": 0.8775055408477783,
+      "learning_rate": 4.102961022465667e-05,
+      "loss": 0.8938,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5390065292028022,
+      "grad_norm": 0.8911844491958618,
+      "learning_rate": 4.101637956127121e-05,
+      "loss": 0.8922,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5398003532517017,
+      "grad_norm": 0.9125531315803528,
+      "learning_rate": 4.100314889788574e-05,
+      "loss": 0.9208,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5405941773006013,
+      "grad_norm": 0.8507636189460754,
+      "learning_rate": 4.098991823450028e-05,
+      "loss": 0.8493,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5413880013495009,
+      "grad_norm": 0.9445191621780396,
+      "learning_rate": 4.097668757111482e-05,
+      "loss": 0.8364,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5421818253984004,
+      "grad_norm": 0.6830033659934998,
+      "learning_rate": 4.096345690772935e-05,
+      "loss": 0.8302,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5429756494473,
+      "grad_norm": 0.9439083337783813,
+      "learning_rate": 4.095022624434389e-05,
+      "loss": 0.8545,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5437694734961995,
+      "grad_norm": 0.8083987236022949,
+      "learning_rate": 4.093699558095843e-05,
+      "loss": 0.9273,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5445632975450991,
+      "grad_norm": 0.7800227403640747,
+      "learning_rate": 4.092376491757297e-05,
+      "loss": 0.8564,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5453571215939987,
+      "grad_norm": 0.7036982178688049,
+      "learning_rate": 4.091053425418751e-05,
+      "loss": 0.8236,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5461509456428982,
+      "grad_norm": 0.6380200982093811,
+      "learning_rate": 4.0897303590802046e-05,
+      "loss": 0.8749,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5469447696917978,
+      "grad_norm": 0.9935095906257629,
+      "learning_rate": 4.0884072927416585e-05,
+      "loss": 0.8536,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5477385937406973,
+      "grad_norm": 0.9212964773178101,
+      "learning_rate": 4.0870842264031123e-05,
+      "loss": 0.88,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5485324177895969,
+      "grad_norm": 0.7856978178024292,
+      "learning_rate": 4.0857611600645656e-05,
+      "loss": 0.9521,
+      "step": 6910
+    },
+    {
+      "epoch": 0.5493262418384965,
+      "grad_norm": 0.8367793560028076,
+      "learning_rate": 4.0844380937260195e-05,
+      "loss": 0.8705,
+      "step": 6920
+    },
+    {
+      "epoch": 0.550120065887396,
+      "grad_norm": 0.735126256942749,
+      "learning_rate": 4.0831150273874733e-05,
+      "loss": 0.9276,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5509138899362956,
+      "grad_norm": 0.7849065065383911,
+      "learning_rate": 4.081791961048927e-05,
+      "loss": 0.8424,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5517077139851951,
+      "grad_norm": 0.9133653044700623,
+      "learning_rate": 4.080468894710381e-05,
+      "loss": 0.8708,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5525015380340947,
+      "grad_norm": 0.6829420924186707,
+      "learning_rate": 4.079145828371835e-05,
+      "loss": 0.8696,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5532953620829943,
+      "grad_norm": 0.8633001446723938,
+      "learning_rate": 4.077822762033289e-05,
+      "loss": 0.9033,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5540891861318938,
+      "grad_norm": 0.6927655935287476,
+      "learning_rate": 4.076499695694742e-05,
+      "loss": 0.8687,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5548830101807934,
+      "grad_norm": 0.9020519852638245,
+      "learning_rate": 4.075176629356196e-05,
+      "loss": 0.8289,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5556768342296929,
+      "grad_norm": 0.8436912894248962,
+      "learning_rate": 4.07385356301765e-05,
+      "loss": 0.8881,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5564706582785925,
+      "grad_norm": 0.7371892333030701,
+      "learning_rate": 4.072530496679104e-05,
+      "loss": 0.8995,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5572644823274921,
+      "grad_norm": 0.8149321675300598,
+      "learning_rate": 4.071207430340557e-05,
+      "loss": 0.9026,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5580583063763916,
+      "grad_norm": 0.6326161026954651,
+      "learning_rate": 4.069884364002011e-05,
+      "loss": 0.8795,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5588521304252912,
+      "grad_norm": 0.8339284062385559,
+      "learning_rate": 4.0685612976634655e-05,
+      "loss": 0.8199,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5596459544741907,
+      "grad_norm": 0.7637563347816467,
+      "learning_rate": 4.0672382313249194e-05,
+      "loss": 0.8635,
+      "step": 7050
+    },
+    {
+      "epoch": 0.5604397785230903,
+      "grad_norm": 0.8212976455688477,
+      "learning_rate": 4.0659151649863726e-05,
+      "loss": 0.8681,
+      "step": 7060
+    },
+    {
+      "epoch": 0.56123360257199,
+      "grad_norm": 0.7294253706932068,
+      "learning_rate": 4.0645920986478265e-05,
+      "loss": 0.8807,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5620274266208894,
+      "grad_norm": 0.7334606051445007,
+      "learning_rate": 4.0632690323092804e-05,
+      "loss": 0.8732,
+      "step": 7080
+    },
+    {
+      "epoch": 0.562821250669789,
+      "grad_norm": 0.8146048188209534,
+      "learning_rate": 4.0619459659707336e-05,
+      "loss": 0.8941,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5636150747186887,
+      "grad_norm": 0.8506765365600586,
+      "learning_rate": 4.0606228996321875e-05,
+      "loss": 0.8885,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5644088987675882,
+      "grad_norm": 1.0465277433395386,
+      "learning_rate": 4.0592998332936414e-05,
+      "loss": 0.877,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5652027228164878,
+      "grad_norm": 1.0059021711349487,
+      "learning_rate": 4.057976766955095e-05,
+      "loss": 0.9511,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5659965468653873,
+      "grad_norm": 0.7272640466690063,
+      "learning_rate": 4.056653700616549e-05,
+      "loss": 0.8562,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5667903709142869,
+      "grad_norm": 0.6533452868461609,
+      "learning_rate": 4.055330634278003e-05,
+      "loss": 0.8402,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5675841949631865,
+      "grad_norm": 0.8188508152961731,
+      "learning_rate": 4.054007567939457e-05,
+      "loss": 0.8589,
+      "step": 7150
+    },
+    {
+      "epoch": 0.568378019012086,
+      "grad_norm": 0.8505097031593323,
+      "learning_rate": 4.052684501600911e-05,
+      "loss": 0.8614,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5691718430609856,
+      "grad_norm": 0.7959850430488586,
+      "learning_rate": 4.051361435262364e-05,
+      "loss": 0.8628,
+      "step": 7170
+    },
+    {
+      "epoch": 0.569965667109885,
+      "grad_norm": 0.9886569976806641,
+      "learning_rate": 4.050038368923818e-05,
+      "loss": 0.8947,
+      "step": 7180
+    },
+    {
+      "epoch": 0.5707594911587847,
+      "grad_norm": 0.8581879138946533,
+      "learning_rate": 4.048715302585272e-05,
+      "loss": 0.8619,
+      "step": 7190
+    },
+    {
+      "epoch": 0.5715533152076843,
+      "grad_norm": 0.9623154401779175,
+      "learning_rate": 4.047392236246725e-05,
+      "loss": 0.8252,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5723471392565838,
+      "grad_norm": 0.8627947568893433,
+      "learning_rate": 4.0460691699081796e-05,
+      "loss": 0.8785,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5731409633054834,
+      "grad_norm": 0.7813371419906616,
+      "learning_rate": 4.0447461035696335e-05,
+      "loss": 0.9171,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5739347873543829,
+      "grad_norm": 1.0067614316940308,
+      "learning_rate": 4.0434230372310874e-05,
+      "loss": 0.836,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5747286114032825,
+      "grad_norm": 0.6856113076210022,
+      "learning_rate": 4.0420999708925406e-05,
+      "loss": 0.9001,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5755224354521821,
+      "grad_norm": 1.0112297534942627,
+      "learning_rate": 4.0407769045539945e-05,
+      "loss": 0.8726,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5763162595010816,
+      "grad_norm": 0.6633104085922241,
+      "learning_rate": 4.0394538382154484e-05,
+      "loss": 0.8487,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5771100835499812,
+      "grad_norm": 0.7768102884292603,
+      "learning_rate": 4.038130771876902e-05,
+      "loss": 0.8202,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5779039075988807,
+      "grad_norm": 0.7367390394210815,
+      "learning_rate": 4.0368077055383555e-05,
+      "loss": 0.9207,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5786977316477803,
+      "grad_norm": 0.9495264887809753,
+      "learning_rate": 4.0354846391998094e-05,
+      "loss": 0.9026,
+      "step": 7290
+    },
+    {
+      "epoch": 0.5794915556966799,
+      "grad_norm": 0.880790650844574,
+      "learning_rate": 4.034161572861263e-05,
+      "loss": 0.8974,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5802853797455794,
+      "grad_norm": 0.8407362103462219,
+      "learning_rate": 4.032838506522717e-05,
+      "loss": 0.8189,
+      "step": 7310
+    },
+    {
+      "epoch": 0.581079203794479,
+      "grad_norm": 0.6734247207641602,
+      "learning_rate": 4.031515440184171e-05,
+      "loss": 0.9121,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5818730278433785,
+      "grad_norm": 0.965093195438385,
+      "learning_rate": 4.030192373845625e-05,
+      "loss": 0.8464,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5826668518922781,
+      "grad_norm": 0.7038446068763733,
+      "learning_rate": 4.028869307507079e-05,
+      "loss": 0.9442,
+      "step": 7340
+    },
+    {
+      "epoch": 0.5834606759411777,
+      "grad_norm": 0.6789405345916748,
+      "learning_rate": 4.027546241168532e-05,
+      "loss": 0.9112,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5842544999900772,
+      "grad_norm": 0.8215457797050476,
+      "learning_rate": 4.026223174829986e-05,
+      "loss": 0.891,
+      "step": 7360
+    },
+    {
+      "epoch": 0.5850483240389768,
+      "grad_norm": 0.8473328948020935,
+      "learning_rate": 4.02490010849144e-05,
+      "loss": 0.8637,
+      "step": 7370
+    },
+    {
+      "epoch": 0.5858421480878763,
+      "grad_norm": 0.7100654244422913,
+      "learning_rate": 4.023577042152894e-05,
+      "loss": 0.9027,
+      "step": 7380
+    },
+    {
+      "epoch": 0.5866359721367759,
+      "grad_norm": 0.8264563083648682,
+      "learning_rate": 4.022253975814348e-05,
+      "loss": 0.8289,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5874297961856755,
+      "grad_norm": 0.595227062702179,
+      "learning_rate": 4.0209309094758016e-05,
+      "loss": 0.8588,
+      "step": 7400
+    },
+    {
+      "epoch": 0.588223620234575,
+      "grad_norm": 0.8084173202514648,
+      "learning_rate": 4.0196078431372555e-05,
+      "loss": 0.9038,
+      "step": 7410
+    },
+    {
+      "epoch": 0.5890174442834746,
+      "grad_norm": 0.8075309991836548,
+      "learning_rate": 4.018284776798709e-05,
+      "loss": 0.9102,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5898112683323742,
+      "grad_norm": 0.5747039318084717,
+      "learning_rate": 4.0169617104601626e-05,
+      "loss": 0.8329,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5906050923812737,
+      "grad_norm": 0.943231463432312,
+      "learning_rate": 4.0156386441216165e-05,
+      "loss": 0.8205,
+      "step": 7440
+    },
+    {
+      "epoch": 0.5913989164301733,
+      "grad_norm": 0.8682481050491333,
+      "learning_rate": 4.0143155777830703e-05,
+      "loss": 0.862,
+      "step": 7450
+    },
+    {
+      "epoch": 0.5921927404790728,
+      "grad_norm": 0.6591200232505798,
+      "learning_rate": 4.0129925114445236e-05,
+      "loss": 0.8736,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5929865645279724,
+      "grad_norm": 0.7325188517570496,
+      "learning_rate": 4.0116694451059775e-05,
+      "loss": 0.8911,
+      "step": 7470
+    },
+    {
+      "epoch": 0.593780388576872,
+      "grad_norm": 0.5145597457885742,
+      "learning_rate": 4.010346378767432e-05,
+      "loss": 0.8992,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5945742126257715,
+      "grad_norm": 0.6899420022964478,
+      "learning_rate": 4.009023312428886e-05,
+      "loss": 0.8651,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5953680366746711,
+      "grad_norm": 0.8163533806800842,
+      "learning_rate": 4.007700246090339e-05,
+      "loss": 0.8571,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5961618607235706,
+      "grad_norm": 0.8414540886878967,
+      "learning_rate": 4.006377179751793e-05,
+      "loss": 0.8866,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5969556847724702,
+      "grad_norm": 0.9524148106575012,
+      "learning_rate": 4.005054113413247e-05,
+      "loss": 0.871,
+      "step": 7520
+    },
+    {
+      "epoch": 0.5977495088213698,
+      "grad_norm": 0.8480884432792664,
+      "learning_rate": 4.0037310470747e-05,
+      "loss": 0.8807,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5985433328702693,
+      "grad_norm": 0.9701437950134277,
+      "learning_rate": 4.002407980736154e-05,
+      "loss": 0.8092,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5993371569191689,
+      "grad_norm": 0.9509230256080627,
+      "learning_rate": 4.001084914397608e-05,
+      "loss": 0.9041,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6001309809680684,
+      "grad_norm": 0.6511445045471191,
+      "learning_rate": 3.999761848059062e-05,
+      "loss": 0.8408,
+      "step": 7560
+    },
+    {
+      "epoch": 0.600924805016968,
+      "grad_norm": 0.7083035111427307,
+      "learning_rate": 3.998438781720516e-05,
+      "loss": 0.8712,
+      "step": 7570
+    },
+    {
+      "epoch": 0.6017186290658676,
+      "grad_norm": 0.8894531726837158,
+      "learning_rate": 3.9971157153819696e-05,
+      "loss": 0.8662,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6025124531147671,
+      "grad_norm": 0.7354892492294312,
+      "learning_rate": 3.9957926490434235e-05,
+      "loss": 0.7996,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6033062771636667,
+      "grad_norm": 0.8339380025863647,
+      "learning_rate": 3.9944695827048774e-05,
+      "loss": 0.8971,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6041001012125662,
+      "grad_norm": 0.7149907350540161,
+      "learning_rate": 3.9931465163663306e-05,
+      "loss": 0.8674,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6048939252614658,
+      "grad_norm": 0.6684656739234924,
+      "learning_rate": 3.9918234500277845e-05,
+      "loss": 0.8449,
+      "step": 7620
+    },
+    {
+      "epoch": 0.6056877493103654,
+      "grad_norm": 0.803089439868927,
+      "learning_rate": 3.9905003836892384e-05,
+      "loss": 0.8952,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6064815733592649,
+      "grad_norm": 0.7551273107528687,
+      "learning_rate": 3.9891773173506916e-05,
+      "loss": 0.8461,
+      "step": 7640
+    },
+    {
+      "epoch": 0.6072753974081645,
+      "grad_norm": 0.6433910131454468,
+      "learning_rate": 3.9878542510121455e-05,
+      "loss": 0.8795,
+      "step": 7650
+    },
+    {
+      "epoch": 0.608069221457064,
+      "grad_norm": 0.73484867811203,
+      "learning_rate": 3.9865311846736e-05,
+      "loss": 0.8916,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6088630455059636,
+      "grad_norm": 0.8747826218605042,
+      "learning_rate": 3.985208118335054e-05,
+      "loss": 0.8803,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6096568695548632,
+      "grad_norm": 0.9112239480018616,
+      "learning_rate": 3.983885051996507e-05,
+      "loss": 0.9397,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6104506936037627,
+      "grad_norm": 0.8457674384117126,
+      "learning_rate": 3.982561985657961e-05,
+      "loss": 0.8451,
+      "step": 7690
+    },
+    {
+      "epoch": 0.6112445176526623,
+      "grad_norm": 0.7593239545822144,
+      "learning_rate": 3.981238919319415e-05,
+      "loss": 0.9272,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6120383417015618,
+      "grad_norm": 0.7265938520431519,
+      "learning_rate": 3.979915852980869e-05,
+      "loss": 0.9268,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6128321657504614,
+      "grad_norm": 0.7465494871139526,
+      "learning_rate": 3.978592786642322e-05,
+      "loss": 0.872,
+      "step": 7720
+    },
+    {
+      "epoch": 0.613625989799361,
+      "grad_norm": 0.7977067828178406,
+      "learning_rate": 3.977269720303776e-05,
+      "loss": 0.8586,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6144198138482605,
+      "grad_norm": 0.76861572265625,
+      "learning_rate": 3.97594665396523e-05,
+      "loss": 0.9187,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6152136378971601,
+      "grad_norm": 0.9571815133094788,
+      "learning_rate": 3.974623587626684e-05,
+      "loss": 0.835,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6160074619460597,
+      "grad_norm": 0.7097697257995605,
+      "learning_rate": 3.9733005212881376e-05,
+      "loss": 0.8519,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6168012859949592,
+      "grad_norm": 0.6047300696372986,
+      "learning_rate": 3.9719774549495915e-05,
+      "loss": 0.8864,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6175951100438588,
+      "grad_norm": 0.7774935364723206,
+      "learning_rate": 3.9706543886110454e-05,
+      "loss": 0.8528,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6183889340927583,
+      "grad_norm": 0.7351526021957397,
+      "learning_rate": 3.9693313222724986e-05,
+      "loss": 0.8548,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6191827581416579,
+      "grad_norm": 0.8887933492660522,
+      "learning_rate": 3.9680082559339525e-05,
+      "loss": 0.9191,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6199765821905575,
+      "grad_norm": 0.6193240284919739,
+      "learning_rate": 3.9666851895954064e-05,
+      "loss": 0.8906,
+      "step": 7810
+    },
+    {
+      "epoch": 0.620770406239457,
+      "grad_norm": 1.0305780172348022,
+      "learning_rate": 3.96536212325686e-05,
+      "loss": 0.9161,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6215642302883566,
+      "grad_norm": 0.9303094744682312,
+      "learning_rate": 3.964039056918314e-05,
+      "loss": 0.923,
+      "step": 7830
+    },
+    {
+      "epoch": 0.6223580543372561,
+      "grad_norm": 0.7105975151062012,
+      "learning_rate": 3.962715990579768e-05,
+      "loss": 0.8918,
+      "step": 7840
+    },
+    {
+      "epoch": 0.6231518783861557,
+      "grad_norm": 0.7743216753005981,
+      "learning_rate": 3.961392924241222e-05,
+      "loss": 0.914,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6239457024350553,
+      "grad_norm": 1.069062352180481,
+      "learning_rate": 3.960069857902676e-05,
+      "loss": 0.9209,
+      "step": 7860
+    },
+    {
+      "epoch": 0.6247395264839548,
+      "grad_norm": 0.8283140659332275,
+      "learning_rate": 3.958746791564129e-05,
+      "loss": 0.8587,
+      "step": 7870
+    },
+    {
+      "epoch": 0.6255333505328544,
+      "grad_norm": 0.6875911355018616,
+      "learning_rate": 3.957423725225583e-05,
+      "loss": 0.8845,
+      "step": 7880
+    },
+    {
+      "epoch": 0.6263271745817539,
+      "grad_norm": 0.8538176417350769,
+      "learning_rate": 3.956100658887037e-05,
+      "loss": 0.8435,
+      "step": 7890
+    },
+    {
+      "epoch": 0.6271209986306535,
+      "grad_norm": 0.6891659498214722,
+      "learning_rate": 3.95477759254849e-05,
+      "loss": 0.8122,
+      "step": 7900
+    },
+    {
+      "epoch": 0.6279148226795531,
+      "grad_norm": 0.6774616837501526,
+      "learning_rate": 3.953454526209944e-05,
+      "loss": 0.9223,
+      "step": 7910
+    },
+    {
+      "epoch": 0.6287086467284526,
+      "grad_norm": 0.6957716345787048,
+      "learning_rate": 3.952131459871398e-05,
+      "loss": 0.9055,
+      "step": 7920
+    },
+    {
+      "epoch": 0.6295024707773522,
+      "grad_norm": 0.7864802479743958,
+      "learning_rate": 3.9508083935328525e-05,
+      "loss": 0.8758,
+      "step": 7930
+    },
+    {
+      "epoch": 0.6302962948262517,
+      "grad_norm": 0.6542963981628418,
+      "learning_rate": 3.949485327194306e-05,
+      "loss": 0.8862,
+      "step": 7940
+    },
+    {
+      "epoch": 0.6310901188751513,
+      "grad_norm": 0.8261880278587341,
+      "learning_rate": 3.9481622608557596e-05,
+      "loss": 0.8559,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6318839429240509,
+      "grad_norm": 0.7244775295257568,
+      "learning_rate": 3.9468391945172135e-05,
+      "loss": 0.896,
+      "step": 7960
+    },
+    {
+      "epoch": 0.6326777669729504,
+      "grad_norm": 0.8435840010643005,
+      "learning_rate": 3.9455161281786673e-05,
+      "loss": 0.8781,
+      "step": 7970
+    },
+    {
+      "epoch": 0.63347159102185,
+      "grad_norm": 0.8194109797477722,
+      "learning_rate": 3.9441930618401206e-05,
+      "loss": 0.9129,
+      "step": 7980
+    },
+    {
+      "epoch": 0.6342654150707495,
+      "grad_norm": 0.746361494064331,
+      "learning_rate": 3.9428699955015745e-05,
+      "loss": 0.9033,
+      "step": 7990
+    },
+    {
+      "epoch": 0.6350592391196491,
+      "grad_norm": 0.7431300282478333,
+      "learning_rate": 3.9415469291630283e-05,
+      "loss": 0.9065,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6358530631685487,
+      "grad_norm": 0.7559617757797241,
+      "learning_rate": 3.940223862824482e-05,
+      "loss": 0.8851,
+      "step": 8010
+    },
+    {
+      "epoch": 0.6366468872174482,
+      "grad_norm": 0.7204374074935913,
+      "learning_rate": 3.938900796485936e-05,
+      "loss": 0.9143,
+      "step": 8020
+    },
+    {
+      "epoch": 0.6374407112663478,
+      "grad_norm": 0.9044206142425537,
+      "learning_rate": 3.93757773014739e-05,
+      "loss": 0.9308,
+      "step": 8030
+    },
+    {
+      "epoch": 0.6382345353152473,
+      "grad_norm": 0.6706372499465942,
+      "learning_rate": 3.936254663808844e-05,
+      "loss": 0.8871,
+      "step": 8040
+    },
+    {
+      "epoch": 0.6390283593641469,
+      "grad_norm": 0.8371633291244507,
+      "learning_rate": 3.934931597470297e-05,
+      "loss": 0.8375,
+      "step": 8050
+    },
+    {
+      "epoch": 0.6398221834130465,
+      "grad_norm": 0.823395311832428,
+      "learning_rate": 3.933608531131751e-05,
+      "loss": 0.8351,
+      "step": 8060
+    },
+    {
+      "epoch": 0.640616007461946,
+      "grad_norm": 0.7647190690040588,
+      "learning_rate": 3.932417771427059e-05,
+      "loss": 0.8564,
+      "step": 8070
+    },
+    {
+      "epoch": 0.6414098315108456,
+      "grad_norm": 0.7140836119651794,
+      "learning_rate": 3.931094705088513e-05,
+      "loss": 0.8248,
+      "step": 8080
+    },
+    {
+      "epoch": 0.6422036555597452,
+      "grad_norm": 0.7830526232719421,
+      "learning_rate": 3.929771638749967e-05,
+      "loss": 0.8325,
+      "step": 8090
+    },
+    {
+      "epoch": 0.6429974796086447,
+      "grad_norm": 0.8376036286354065,
+      "learning_rate": 3.9284485724114215e-05,
+      "loss": 0.8787,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6437913036575443,
+      "grad_norm": 0.8224188685417175,
+      "learning_rate": 3.927125506072875e-05,
+      "loss": 0.8556,
+      "step": 8110
+    },
+    {
+      "epoch": 0.6445851277064438,
+      "grad_norm": 0.7094652056694031,
+      "learning_rate": 3.9258024397343286e-05,
+      "loss": 0.9008,
+      "step": 8120
+    },
+    {
+      "epoch": 0.6453789517553434,
+      "grad_norm": 0.9661831855773926,
+      "learning_rate": 3.9244793733957825e-05,
+      "loss": 0.9017,
+      "step": 8130
+    },
+    {
+      "epoch": 0.646172775804243,
+      "grad_norm": 0.5937822461128235,
+      "learning_rate": 3.923156307057236e-05,
+      "loss": 0.93,
+      "step": 8140
+    },
+    {
+      "epoch": 0.6469665998531425,
+      "grad_norm": 0.8832845091819763,
+      "learning_rate": 3.9218332407186896e-05,
+      "loss": 0.7535,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6477604239020421,
+      "grad_norm": 0.8030399084091187,
+      "learning_rate": 3.9205101743801435e-05,
+      "loss": 0.8628,
+      "step": 8160
+    },
+    {
+      "epoch": 0.6485542479509416,
+      "grad_norm": 0.6828364729881287,
+      "learning_rate": 3.9191871080415974e-05,
+      "loss": 0.8172,
+      "step": 8170
+    },
+    {
+      "epoch": 0.6493480719998412,
+      "grad_norm": 0.7362493872642517,
+      "learning_rate": 3.917864041703051e-05,
+      "loss": 0.8649,
+      "step": 8180
+    },
+    {
+      "epoch": 0.6501418960487408,
+      "grad_norm": 0.711621105670929,
+      "learning_rate": 3.916540975364505e-05,
+      "loss": 0.8908,
+      "step": 8190
+    },
+    {
+      "epoch": 0.6509357200976403,
+      "grad_norm": 0.7424710392951965,
+      "learning_rate": 3.915217909025959e-05,
+      "loss": 0.8684,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6517295441465399,
+      "grad_norm": 0.8908485770225525,
+      "learning_rate": 3.913894842687413e-05,
+      "loss": 0.867,
+      "step": 8210
+    },
+    {
+      "epoch": 0.6525233681954394,
+      "grad_norm": 0.8501769304275513,
+      "learning_rate": 3.912571776348866e-05,
+      "loss": 0.8688,
+      "step": 8220
+    },
+    {
+      "epoch": 0.653317192244339,
+      "grad_norm": 0.8504555225372314,
+      "learning_rate": 3.91124871001032e-05,
+      "loss": 0.8451,
+      "step": 8230
+    },
+    {
+      "epoch": 0.6541110162932386,
+      "grad_norm": 1.316743016242981,
+      "learning_rate": 3.909925643671774e-05,
+      "loss": 0.8375,
+      "step": 8240
+    },
+    {
+      "epoch": 0.6549048403421381,
+      "grad_norm": 0.7561972141265869,
+      "learning_rate": 3.908602577333227e-05,
+      "loss": 0.8597,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6556986643910377,
+      "grad_norm": 0.7232686877250671,
+      "learning_rate": 3.907279510994681e-05,
+      "loss": 0.8808,
+      "step": 8260
+    },
+    {
+      "epoch": 0.6564924884399372,
+      "grad_norm": 0.6502814888954163,
+      "learning_rate": 3.9059564446561356e-05,
+      "loss": 0.939,
+      "step": 8270
+    },
+    {
+      "epoch": 0.6572863124888368,
+      "grad_norm": 0.9864717721939087,
+      "learning_rate": 3.9046333783175895e-05,
+      "loss": 0.8091,
+      "step": 8280
+    },
+    {
+      "epoch": 0.6580801365377364,
+      "grad_norm": 0.8748832941055298,
+      "learning_rate": 3.903310311979043e-05,
+      "loss": 0.8676,
+      "step": 8290
+    },
+    {
+      "epoch": 0.6588739605866359,
+      "grad_norm": 0.8619644641876221,
+      "learning_rate": 3.9019872456404966e-05,
+      "loss": 0.8134,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6596677846355355,
+      "grad_norm": 0.7893159985542297,
+      "learning_rate": 3.9006641793019505e-05,
+      "loss": 0.8656,
+      "step": 8310
+    },
+    {
+      "epoch": 0.660461608684435,
+      "grad_norm": 0.8085225820541382,
+      "learning_rate": 3.8993411129634044e-05,
+      "loss": 0.9401,
+      "step": 8320
+    },
+    {
+      "epoch": 0.6612554327333346,
+      "grad_norm": 0.8929345011711121,
+      "learning_rate": 3.8980180466248576e-05,
+      "loss": 0.8665,
+      "step": 8330
+    },
+    {
+      "epoch": 0.6620492567822343,
+      "grad_norm": 0.9615473747253418,
+      "learning_rate": 3.8966949802863115e-05,
+      "loss": 0.8331,
+      "step": 8340
+    },
+    {
+      "epoch": 0.6628430808311337,
+      "grad_norm": 0.582528293132782,
+      "learning_rate": 3.8953719139477654e-05,
+      "loss": 0.8626,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6636369048800334,
+      "grad_norm": 0.7773953676223755,
+      "learning_rate": 3.894048847609219e-05,
+      "loss": 0.8119,
+      "step": 8360
+    },
+    {
+      "epoch": 0.664430728928933,
+      "grad_norm": 0.8851808309555054,
+      "learning_rate": 3.892725781270673e-05,
+      "loss": 0.8806,
+      "step": 8370
+    },
+    {
+      "epoch": 0.6652245529778325,
+      "grad_norm": 0.8507623076438904,
+      "learning_rate": 3.891402714932127e-05,
+      "loss": 0.8734,
+      "step": 8380
+    },
+    {
+      "epoch": 0.6660183770267321,
+      "grad_norm": 0.8153007626533508,
+      "learning_rate": 3.890079648593581e-05,
+      "loss": 0.8501,
+      "step": 8390
+    },
+    {
+      "epoch": 0.6668122010756315,
+      "grad_norm": 0.8047279715538025,
+      "learning_rate": 3.888756582255034e-05,
+      "loss": 0.8574,
+      "step": 8400
+    },
+    {
+      "epoch": 0.6676060251245312,
+      "grad_norm": 0.933725118637085,
+      "learning_rate": 3.887433515916488e-05,
+      "loss": 0.8638,
+      "step": 8410
+    },
+    {
+      "epoch": 0.6683998491734308,
+      "grad_norm": 0.7633134126663208,
+      "learning_rate": 3.886110449577942e-05,
+      "loss": 0.9276,
+      "step": 8420
+    },
+    {
+      "epoch": 0.6691936732223303,
+      "grad_norm": 0.8847464323043823,
+      "learning_rate": 3.884787383239396e-05,
+      "loss": 0.9414,
+      "step": 8430
+    },
+    {
+      "epoch": 0.6699874972712299,
+      "grad_norm": 0.9716514945030212,
+      "learning_rate": 3.88346431690085e-05,
+      "loss": 0.8823,
+      "step": 8440
+    },
+    {
+      "epoch": 0.6707813213201294,
+      "grad_norm": 0.8320568203926086,
+      "learning_rate": 3.8821412505623037e-05,
+      "loss": 0.812,
+      "step": 8450
+    },
+    {
+      "epoch": 0.671575145369029,
+      "grad_norm": 0.7961634397506714,
+      "learning_rate": 3.8808181842237575e-05,
+      "loss": 0.816,
+      "step": 8460
+    },
+    {
+      "epoch": 0.6723689694179286,
+      "grad_norm": 0.7325245141983032,
+      "learning_rate": 3.879495117885211e-05,
+      "loss": 0.8224,
+      "step": 8470
+    },
+    {
+      "epoch": 0.6731627934668281,
+      "grad_norm": 0.7438898086547852,
+      "learning_rate": 3.8781720515466647e-05,
+      "loss": 0.8102,
+      "step": 8480
+    },
+    {
+      "epoch": 0.6739566175157277,
+      "grad_norm": 0.7299096584320068,
+      "learning_rate": 3.8768489852081186e-05,
+      "loss": 0.9632,
+      "step": 8490
+    },
+    {
+      "epoch": 0.6747504415646272,
+      "grad_norm": 0.895844578742981,
+      "learning_rate": 3.8755259188695724e-05,
+      "loss": 0.9497,
+      "step": 8500
+    },
+    {
+      "epoch": 0.6755442656135268,
+      "grad_norm": 0.8177179098129272,
+      "learning_rate": 3.8742028525310257e-05,
+      "loss": 0.853,
+      "step": 8510
+    },
+    {
+      "epoch": 0.6763380896624264,
+      "grad_norm": 0.6874784827232361,
+      "learning_rate": 3.8728797861924796e-05,
+      "loss": 0.8464,
+      "step": 8520
+    },
+    {
+      "epoch": 0.6771319137113259,
+      "grad_norm": 0.8291011452674866,
+      "learning_rate": 3.8715567198539334e-05,
+      "loss": 0.9136,
+      "step": 8530
+    },
+    {
+      "epoch": 0.6779257377602255,
+      "grad_norm": 0.6445680260658264,
+      "learning_rate": 3.870233653515388e-05,
+      "loss": 0.8429,
+      "step": 8540
+    },
+    {
+      "epoch": 0.678719561809125,
+      "grad_norm": 0.7403009533882141,
+      "learning_rate": 3.868910587176841e-05,
+      "loss": 0.8473,
+      "step": 8550
+    },
+    {
+      "epoch": 0.6795133858580246,
+      "grad_norm": 0.5939794778823853,
+      "learning_rate": 3.867587520838295e-05,
+      "loss": 0.9001,
+      "step": 8560
+    },
+    {
+      "epoch": 0.6803072099069242,
+      "grad_norm": 0.6442236304283142,
+      "learning_rate": 3.866264454499749e-05,
+      "loss": 0.8819,
+      "step": 8570
+    },
+    {
+      "epoch": 0.6811010339558237,
+      "grad_norm": 0.7586227655410767,
+      "learning_rate": 3.864941388161202e-05,
+      "loss": 0.8705,
+      "step": 8580
+    },
+    {
+      "epoch": 0.6818948580047233,
+      "grad_norm": 0.7336118817329407,
+      "learning_rate": 3.863618321822656e-05,
+      "loss": 0.9112,
+      "step": 8590
+    },
+    {
+      "epoch": 0.6826886820536228,
+      "grad_norm": 0.7981590032577515,
+      "learning_rate": 3.86229525548411e-05,
+      "loss": 0.9474,
+      "step": 8600
+    },
+    {
+      "epoch": 0.6834825061025224,
+      "grad_norm": 0.5665314197540283,
+      "learning_rate": 3.860972189145564e-05,
+      "loss": 0.8787,
+      "step": 8610
+    },
+    {
+      "epoch": 0.684276330151422,
+      "grad_norm": 0.6617142558097839,
+      "learning_rate": 3.859649122807018e-05,
+      "loss": 0.9087,
+      "step": 8620
+    },
+    {
+      "epoch": 0.6850701542003215,
+      "grad_norm": 0.7118547558784485,
+      "learning_rate": 3.858326056468472e-05,
+      "loss": 0.8094,
+      "step": 8630
+    },
+    {
+      "epoch": 0.6858639782492211,
+      "grad_norm": 0.6482293605804443,
+      "learning_rate": 3.8570029901299256e-05,
+      "loss": 0.861,
+      "step": 8640
+    },
+    {
+      "epoch": 0.6866578022981206,
+      "grad_norm": 0.8651313781738281,
+      "learning_rate": 3.8556799237913795e-05,
+      "loss": 0.7867,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6874516263470202,
+      "grad_norm": 0.5534746646881104,
+      "learning_rate": 3.854356857452833e-05,
+      "loss": 0.9116,
+      "step": 8660
+    },
+    {
+      "epoch": 0.6882454503959198,
+      "grad_norm": 0.6434946060180664,
+      "learning_rate": 3.8530337911142866e-05,
+      "loss": 0.8464,
+      "step": 8670
+    },
+    {
+      "epoch": 0.6890392744448193,
+      "grad_norm": 0.7514382600784302,
+      "learning_rate": 3.8517107247757405e-05,
+      "loss": 0.8618,
+      "step": 8680
+    },
+    {
+      "epoch": 0.6898330984937189,
+      "grad_norm": 0.7134977579116821,
+      "learning_rate": 3.850387658437194e-05,
+      "loss": 0.8792,
+      "step": 8690
+    },
+    {
+      "epoch": 0.6906269225426185,
+      "grad_norm": 0.8967533707618713,
+      "learning_rate": 3.8490645920986476e-05,
+      "loss": 0.8489,
+      "step": 8700
+    },
+    {
+      "epoch": 0.691420746591518,
+      "grad_norm": 0.7297849655151367,
+      "learning_rate": 3.847741525760102e-05,
+      "loss": 0.8498,
+      "step": 8710
+    },
+    {
+      "epoch": 0.6922145706404176,
+      "grad_norm": 0.8720963001251221,
+      "learning_rate": 3.846418459421556e-05,
+      "loss": 0.8486,
+      "step": 8720
+    },
+    {
+      "epoch": 0.6930083946893171,
+      "grad_norm": 0.9449414610862732,
+      "learning_rate": 3.845095393083009e-05,
+      "loss": 0.8597,
+      "step": 8730
+    },
+    {
+      "epoch": 0.6938022187382167,
+      "grad_norm": 0.7103795409202576,
+      "learning_rate": 3.843772326744463e-05,
+      "loss": 0.8633,
+      "step": 8740
+    },
+    {
+      "epoch": 0.6945960427871163,
+      "grad_norm": 1.0056965351104736,
+      "learning_rate": 3.842449260405917e-05,
+      "loss": 0.8723,
+      "step": 8750
+    },
+    {
+      "epoch": 0.6953898668360158,
+      "grad_norm": 0.6581205725669861,
+      "learning_rate": 3.841126194067371e-05,
+      "loss": 0.8956,
+      "step": 8760
+    },
+    {
+      "epoch": 0.6961836908849154,
+      "grad_norm": 0.8664624691009521,
+      "learning_rate": 3.839803127728824e-05,
+      "loss": 0.8979,
+      "step": 8770
+    },
+    {
+      "epoch": 0.6969775149338149,
+      "grad_norm": 0.6672316789627075,
+      "learning_rate": 3.838480061390278e-05,
+      "loss": 0.896,
+      "step": 8780
+    },
+    {
+      "epoch": 0.6977713389827145,
+      "grad_norm": 0.7690501809120178,
+      "learning_rate": 3.837156995051732e-05,
+      "loss": 0.856,
+      "step": 8790
+    },
+    {
+      "epoch": 0.6985651630316141,
+      "grad_norm": 0.7820170521736145,
+      "learning_rate": 3.835833928713186e-05,
+      "loss": 0.8775,
+      "step": 8800
+    },
+    {
+      "epoch": 0.6993589870805136,
+      "grad_norm": 1.0023752450942993,
+      "learning_rate": 3.83451086237464e-05,
+      "loss": 0.8321,
+      "step": 8810
+    },
+    {
+      "epoch": 0.7001528111294132,
+      "grad_norm": 0.7955570816993713,
+      "learning_rate": 3.8331877960360936e-05,
+      "loss": 0.8198,
+      "step": 8820
+    },
+    {
+      "epoch": 0.7009466351783127,
+      "grad_norm": 0.7987425327301025,
+      "learning_rate": 3.8318647296975475e-05,
+      "loss": 0.8072,
+      "step": 8830
+    },
+    {
+      "epoch": 0.7017404592272123,
+      "grad_norm": 0.6776150465011597,
+      "learning_rate": 3.830541663359001e-05,
+      "loss": 0.8903,
+      "step": 8840
+    },
+    {
+      "epoch": 0.7025342832761119,
+      "grad_norm": 0.8060004711151123,
+      "learning_rate": 3.8292185970204546e-05,
+      "loss": 0.8328,
+      "step": 8850
+    },
+    {
+      "epoch": 0.7033281073250114,
+      "grad_norm": 0.6647984981536865,
+      "learning_rate": 3.8278955306819085e-05,
+      "loss": 0.9151,
+      "step": 8860
+    },
+    {
+      "epoch": 0.704121931373911,
+      "grad_norm": 0.7198352217674255,
+      "learning_rate": 3.8265724643433624e-05,
+      "loss": 0.8502,
+      "step": 8870
+    },
+    {
+      "epoch": 0.7049157554228105,
+      "grad_norm": 0.8274781107902527,
+      "learning_rate": 3.825249398004816e-05,
+      "loss": 0.8856,
+      "step": 8880
+    },
+    {
+      "epoch": 0.7057095794717101,
+      "grad_norm": 0.868508517742157,
+      "learning_rate": 3.82392633166627e-05,
+      "loss": 0.806,
+      "step": 8890
+    },
+    {
+      "epoch": 0.7065034035206097,
+      "grad_norm": 0.7774865627288818,
+      "learning_rate": 3.822603265327724e-05,
+      "loss": 0.844,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7072972275695092,
+      "grad_norm": 0.8134011030197144,
+      "learning_rate": 3.821280198989178e-05,
+      "loss": 0.946,
+      "step": 8910
+    },
+    {
+      "epoch": 0.7080910516184088,
+      "grad_norm": 0.7954188585281372,
+      "learning_rate": 3.819957132650631e-05,
+      "loss": 0.8422,
+      "step": 8920
+    },
+    {
+      "epoch": 0.7088848756673083,
+      "grad_norm": 0.7246106863021851,
+      "learning_rate": 3.818634066312085e-05,
+      "loss": 0.8401,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7096786997162079,
+      "grad_norm": 0.7810704112052917,
+      "learning_rate": 3.817310999973539e-05,
+      "loss": 0.8684,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7104725237651075,
+      "grad_norm": 0.74953293800354,
+      "learning_rate": 3.815987933634992e-05,
+      "loss": 0.8565,
+      "step": 8950
+    },
+    {
+      "epoch": 0.711266347814007,
+      "grad_norm": 0.9855328798294067,
+      "learning_rate": 3.814664867296446e-05,
+      "loss": 0.9263,
+      "step": 8960
+    },
+    {
+      "epoch": 0.7120601718629066,
+      "grad_norm": 0.5810590386390686,
+      "learning_rate": 3.8133418009579e-05,
+      "loss": 0.9298,
+      "step": 8970
+    },
+    {
+      "epoch": 0.7128539959118061,
+      "grad_norm": 0.817237377166748,
+      "learning_rate": 3.8120187346193545e-05,
+      "loss": 0.8469,
+      "step": 8980
+    },
+    {
+      "epoch": 0.7136478199607057,
+      "grad_norm": 0.7904085516929626,
+      "learning_rate": 3.810695668280808e-05,
+      "loss": 0.8513,
+      "step": 8990
+    },
+    {
+      "epoch": 0.7144416440096053,
+      "grad_norm": 0.6733309626579285,
+      "learning_rate": 3.8093726019422617e-05,
+      "loss": 0.8477,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7152354680585048,
+      "grad_norm": 0.7731225490570068,
+      "learning_rate": 3.8080495356037155e-05,
+      "loss": 0.9097,
+      "step": 9010
+    },
+    {
+      "epoch": 0.7160292921074044,
+      "grad_norm": 0.8163710236549377,
+      "learning_rate": 3.8067264692651694e-05,
+      "loss": 0.8878,
+      "step": 9020
+    },
+    {
+      "epoch": 0.716823116156304,
+      "grad_norm": 0.8576553463935852,
+      "learning_rate": 3.8054034029266227e-05,
+      "loss": 0.8706,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7176169402052035,
+      "grad_norm": 0.7297602295875549,
+      "learning_rate": 3.8040803365880766e-05,
+      "loss": 0.9051,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7184107642541031,
+      "grad_norm": 0.6882147192955017,
+      "learning_rate": 3.8027572702495304e-05,
+      "loss": 0.8919,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7192045883030026,
+      "grad_norm": 0.8439272046089172,
+      "learning_rate": 3.801434203910984e-05,
+      "loss": 0.898,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7199984123519022,
+      "grad_norm": 0.7839822173118591,
+      "learning_rate": 3.800111137572438e-05,
+      "loss": 0.9044,
+      "step": 9070
+    },
+    {
+      "epoch": 0.7207922364008018,
+      "grad_norm": 0.6823743581771851,
+      "learning_rate": 3.798788071233892e-05,
+      "loss": 0.8372,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7215860604497013,
+      "grad_norm": 0.7644033432006836,
+      "learning_rate": 3.797465004895346e-05,
+      "loss": 0.9316,
+      "step": 9090
+    },
+    {
+      "epoch": 0.7223798844986009,
+      "grad_norm": 0.826184093952179,
+      "learning_rate": 3.796141938556799e-05,
+      "loss": 0.8519,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7231737085475004,
+      "grad_norm": 0.8634784817695618,
+      "learning_rate": 3.794818872218253e-05,
+      "loss": 0.905,
+      "step": 9110
+    },
+    {
+      "epoch": 0.7239675325964,
+      "grad_norm": 0.6479964852333069,
+      "learning_rate": 3.793495805879707e-05,
+      "loss": 0.7854,
+      "step": 9120
+    },
+    {
+      "epoch": 0.7247613566452996,
+      "grad_norm": 0.7302672863006592,
+      "learning_rate": 3.792172739541161e-05,
+      "loss": 0.877,
+      "step": 9130
+    },
+    {
+      "epoch": 0.7255551806941991,
+      "grad_norm": 0.7126200199127197,
+      "learning_rate": 3.790849673202614e-05,
+      "loss": 0.9136,
+      "step": 9140
+    },
+    {
+      "epoch": 0.7263490047430987,
+      "grad_norm": 0.8808421492576599,
+      "learning_rate": 3.789526606864068e-05,
+      "loss": 0.8717,
+      "step": 9150
+    },
+    {
+      "epoch": 0.7271428287919982,
+      "grad_norm": 0.7039631605148315,
+      "learning_rate": 3.7882035405255226e-05,
+      "loss": 0.835,
+      "step": 9160
+    },
+    {
+      "epoch": 0.7279366528408978,
+      "grad_norm": 0.7323824167251587,
+      "learning_rate": 3.786880474186976e-05,
+      "loss": 0.8713,
+      "step": 9170
+    },
+    {
+      "epoch": 0.7287304768897974,
+      "grad_norm": 0.7466415762901306,
+      "learning_rate": 3.78555740784843e-05,
+      "loss": 0.874,
+      "step": 9180
+    },
+    {
+      "epoch": 0.7295243009386969,
+      "grad_norm": 0.6216104626655579,
+      "learning_rate": 3.7842343415098836e-05,
+      "loss": 0.8985,
+      "step": 9190
+    },
+    {
+      "epoch": 0.7303181249875965,
+      "grad_norm": 0.8570176362991333,
+      "learning_rate": 3.7829112751713375e-05,
+      "loss": 0.8755,
+      "step": 9200
+    },
+    {
+      "epoch": 0.731111949036496,
+      "grad_norm": 0.7818983197212219,
+      "learning_rate": 3.781588208832791e-05,
+      "loss": 0.8462,
+      "step": 9210
+    },
+    {
+      "epoch": 0.7319057730853956,
+      "grad_norm": 0.6952176094055176,
+      "learning_rate": 3.7802651424942446e-05,
+      "loss": 0.8286,
+      "step": 9220
+    },
+    {
+      "epoch": 0.7326995971342952,
+      "grad_norm": 0.6671289801597595,
+      "learning_rate": 3.7789420761556985e-05,
+      "loss": 0.8437,
+      "step": 9230
+    },
+    {
+      "epoch": 0.7334934211831947,
+      "grad_norm": 0.6920092701911926,
+      "learning_rate": 3.7776190098171524e-05,
+      "loss": 0.8691,
+      "step": 9240
+    },
+    {
+      "epoch": 0.7342872452320943,
+      "grad_norm": 0.5979323387145996,
+      "learning_rate": 3.776295943478606e-05,
+      "loss": 0.8728,
+      "step": 9250
+    },
+    {
+      "epoch": 0.7350810692809938,
+      "grad_norm": 0.8079019784927368,
+      "learning_rate": 3.77497287714006e-05,
+      "loss": 0.8504,
+      "step": 9260
+    },
+    {
+      "epoch": 0.7358748933298934,
+      "grad_norm": 0.6762669086456299,
+      "learning_rate": 3.773649810801514e-05,
+      "loss": 0.8608,
+      "step": 9270
+    },
+    {
+      "epoch": 0.736668717378793,
+      "grad_norm": 0.6366623044013977,
+      "learning_rate": 3.772326744462967e-05,
+      "loss": 0.873,
+      "step": 9280
+    },
+    {
+      "epoch": 0.7374625414276925,
+      "grad_norm": 0.8903300166130066,
+      "learning_rate": 3.771003678124421e-05,
+      "loss": 0.8963,
+      "step": 9290
+    },
+    {
+      "epoch": 0.7382563654765921,
+      "grad_norm": 0.8363783955574036,
+      "learning_rate": 3.769680611785875e-05,
+      "loss": 0.8462,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7390501895254917,
+      "grad_norm": 0.716432511806488,
+      "learning_rate": 3.768357545447329e-05,
+      "loss": 0.8599,
+      "step": 9310
+    },
+    {
+      "epoch": 0.7398440135743912,
+      "grad_norm": 0.8705762624740601,
+      "learning_rate": 3.767034479108782e-05,
+      "loss": 0.8659,
+      "step": 9320
+    },
+    {
+      "epoch": 0.7406378376232908,
+      "grad_norm": 0.7024847865104675,
+      "learning_rate": 3.765711412770237e-05,
+      "loss": 0.8697,
+      "step": 9330
+    },
+    {
+      "epoch": 0.7414316616721903,
+      "grad_norm": 0.7398332357406616,
+      "learning_rate": 3.7643883464316906e-05,
+      "loss": 0.8322,
+      "step": 9340
+    },
+    {
+      "epoch": 0.7422254857210899,
+      "grad_norm": 0.7438483834266663,
+      "learning_rate": 3.7630652800931445e-05,
+      "loss": 0.8447,
+      "step": 9350
+    },
+    {
+      "epoch": 0.7430193097699895,
+      "grad_norm": 0.7145562767982483,
+      "learning_rate": 3.761742213754598e-05,
+      "loss": 0.8426,
+      "step": 9360
+    },
+    {
+      "epoch": 0.743813133818889,
+      "grad_norm": 0.7920240759849548,
+      "learning_rate": 3.7604191474160516e-05,
+      "loss": 0.8861,
+      "step": 9370
+    },
+    {
+      "epoch": 0.7446069578677886,
+      "grad_norm": 0.7747379541397095,
+      "learning_rate": 3.7590960810775055e-05,
+      "loss": 0.864,
+      "step": 9380
+    },
+    {
+      "epoch": 0.7454007819166881,
+      "grad_norm": 0.8978133201599121,
+      "learning_rate": 3.757773014738959e-05,
+      "loss": 0.7788,
+      "step": 9390
+    },
+    {
+      "epoch": 0.7461946059655877,
+      "grad_norm": 0.6355377435684204,
+      "learning_rate": 3.7564499484004126e-05,
+      "loss": 0.8701,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7469884300144873,
+      "grad_norm": 0.6304174065589905,
+      "learning_rate": 3.7551268820618665e-05,
+      "loss": 0.8598,
+      "step": 9410
+    },
+    {
+      "epoch": 0.7477822540633868,
+      "grad_norm": 0.8747965693473816,
+      "learning_rate": 3.7538038157233204e-05,
+      "loss": 0.8457,
+      "step": 9420
+    },
+    {
+      "epoch": 0.7485760781122864,
+      "grad_norm": 0.7455962300300598,
+      "learning_rate": 3.752480749384774e-05,
+      "loss": 0.838,
+      "step": 9430
+    },
+    {
+      "epoch": 0.7493699021611859,
+      "grad_norm": 0.6928434371948242,
+      "learning_rate": 3.751289989680083e-05,
+      "loss": 0.8531,
+      "step": 9440
+    },
+    {
+      "epoch": 0.7501637262100855,
+      "grad_norm": 0.6415355801582336,
+      "learning_rate": 3.749966923341536e-05,
+      "loss": 0.836,
+      "step": 9450
+    },
+    {
+      "epoch": 0.7509575502589851,
+      "grad_norm": 0.7036964893341064,
+      "learning_rate": 3.74864385700299e-05,
+      "loss": 0.8683,
+      "step": 9460
+    },
+    {
+      "epoch": 0.7517513743078846,
+      "grad_norm": 0.546319305896759,
+      "learning_rate": 3.747320790664444e-05,
+      "loss": 0.864,
+      "step": 9470
+    },
+    {
+      "epoch": 0.7525451983567842,
+      "grad_norm": 0.7785167098045349,
+      "learning_rate": 3.745997724325898e-05,
+      "loss": 0.9178,
+      "step": 9480
+    },
+    {
+      "epoch": 0.7533390224056837,
+      "grad_norm": 0.6985113024711609,
+      "learning_rate": 3.744674657987351e-05,
+      "loss": 0.8726,
+      "step": 9490
+    },
+    {
+      "epoch": 0.7541328464545833,
+      "grad_norm": 0.7310826182365417,
+      "learning_rate": 3.743351591648806e-05,
+      "loss": 0.8742,
+      "step": 9500
+    },
+    {
+      "epoch": 0.7549266705034829,
+      "grad_norm": 0.8766258358955383,
+      "learning_rate": 3.7420285253102596e-05,
+      "loss": 0.8659,
+      "step": 9510
+    },
+    {
+      "epoch": 0.7557204945523824,
+      "grad_norm": 0.6016245484352112,
+      "learning_rate": 3.740705458971713e-05,
+      "loss": 0.8517,
+      "step": 9520
+    },
+    {
+      "epoch": 0.756514318601282,
+      "grad_norm": 0.6588014960289001,
+      "learning_rate": 3.739382392633167e-05,
+      "loss": 0.8621,
+      "step": 9530
+    },
+    {
+      "epoch": 0.7573081426501815,
+      "grad_norm": 0.8283513784408569,
+      "learning_rate": 3.7380593262946206e-05,
+      "loss": 0.8527,
+      "step": 9540
+    },
+    {
+      "epoch": 0.7581019666990811,
+      "grad_norm": 0.6860958337783813,
+      "learning_rate": 3.7367362599560745e-05,
+      "loss": 0.8549,
+      "step": 9550
+    },
+    {
+      "epoch": 0.7588957907479807,
+      "grad_norm": 0.8800874948501587,
+      "learning_rate": 3.735413193617528e-05,
+      "loss": 0.8435,
+      "step": 9560
+    },
+    {
+      "epoch": 0.7596896147968802,
+      "grad_norm": 0.6277894973754883,
+      "learning_rate": 3.7340901272789816e-05,
+      "loss": 0.845,
+      "step": 9570
+    },
+    {
+      "epoch": 0.7604834388457798,
+      "grad_norm": 0.8731912970542908,
+      "learning_rate": 3.7327670609404355e-05,
+      "loss": 0.8302,
+      "step": 9580
+    },
+    {
+      "epoch": 0.7612772628946793,
+      "grad_norm": 0.9646309018135071,
+      "learning_rate": 3.7314439946018894e-05,
+      "loss": 0.7982,
+      "step": 9590
+    },
+    {
+      "epoch": 0.762071086943579,
+      "grad_norm": 0.7538288831710815,
+      "learning_rate": 3.730120928263343e-05,
+      "loss": 0.8195,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7628649109924786,
+      "grad_norm": 0.8189886212348938,
+      "learning_rate": 3.728797861924797e-05,
+      "loss": 0.8968,
+      "step": 9610
+    },
+    {
+      "epoch": 0.763658735041378,
+      "grad_norm": 0.9274044036865234,
+      "learning_rate": 3.727474795586251e-05,
+      "loss": 0.8673,
+      "step": 9620
+    },
+    {
+      "epoch": 0.7644525590902777,
+      "grad_norm": 0.7127663493156433,
+      "learning_rate": 3.726151729247704e-05,
+      "loss": 0.8359,
+      "step": 9630
+    },
+    {
+      "epoch": 0.7652463831391773,
+      "grad_norm": 0.8801278471946716,
+      "learning_rate": 3.724828662909158e-05,
+      "loss": 0.8546,
+      "step": 9640
+    },
+    {
+      "epoch": 0.7660402071880767,
+      "grad_norm": 0.8994466066360474,
+      "learning_rate": 3.723505596570612e-05,
+      "loss": 0.8434,
+      "step": 9650
+    },
+    {
+      "epoch": 0.7668340312369764,
+      "grad_norm": 0.7613593339920044,
+      "learning_rate": 3.722182530232066e-05,
+      "loss": 0.8683,
+      "step": 9660
+    },
+    {
+      "epoch": 0.7676278552858758,
+      "grad_norm": 0.6378811597824097,
+      "learning_rate": 3.72085946389352e-05,
+      "loss": 0.8729,
+      "step": 9670
+    },
+    {
+      "epoch": 0.7684216793347755,
+      "grad_norm": 0.6833720207214355,
+      "learning_rate": 3.719536397554974e-05,
+      "loss": 0.8427,
+      "step": 9680
+    },
+    {
+      "epoch": 0.7692155033836751,
+      "grad_norm": 0.6543610692024231,
+      "learning_rate": 3.718213331216428e-05,
+      "loss": 0.8309,
+      "step": 9690
+    },
+    {
+      "epoch": 0.7700093274325746,
+      "grad_norm": 0.7500269412994385,
+      "learning_rate": 3.7168902648778816e-05,
+      "loss": 0.8621,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7708031514814742,
+      "grad_norm": 0.7766570448875427,
+      "learning_rate": 3.715567198539335e-05,
+      "loss": 0.8549,
+      "step": 9710
+    },
+    {
+      "epoch": 0.7715969755303737,
+      "grad_norm": 0.8536286950111389,
+      "learning_rate": 3.714244132200789e-05,
+      "loss": 0.8375,
+      "step": 9720
+    },
+    {
+      "epoch": 0.7723907995792733,
+      "grad_norm": 0.8572762608528137,
+      "learning_rate": 3.7129210658622426e-05,
+      "loss": 0.8188,
+      "step": 9730
+    },
+    {
+      "epoch": 0.7731846236281729,
+      "grad_norm": 1.0639240741729736,
+      "learning_rate": 3.711597999523696e-05,
+      "loss": 0.8433,
+      "step": 9740
+    },
+    {
+      "epoch": 0.7739784476770724,
+      "grad_norm": 0.7401818633079529,
+      "learning_rate": 3.71027493318515e-05,
+      "loss": 0.8178,
+      "step": 9750
+    },
+    {
+      "epoch": 0.774772271725972,
+      "grad_norm": 0.9968650341033936,
+      "learning_rate": 3.7089518668466036e-05,
+      "loss": 0.8486,
+      "step": 9760
+    },
+    {
+      "epoch": 0.7755660957748715,
+      "grad_norm": 0.9690655469894409,
+      "learning_rate": 3.707628800508058e-05,
+      "loss": 0.8331,
+      "step": 9770
+    },
+    {
+      "epoch": 0.7763599198237711,
+      "grad_norm": 1.0017821788787842,
+      "learning_rate": 3.7063057341695114e-05,
+      "loss": 0.8756,
+      "step": 9780
+    },
+    {
+      "epoch": 0.7771537438726707,
+      "grad_norm": 0.7528254389762878,
+      "learning_rate": 3.704982667830965e-05,
+      "loss": 0.9261,
+      "step": 9790
+    },
+    {
+      "epoch": 0.7779475679215702,
+      "grad_norm": 0.7826637625694275,
+      "learning_rate": 3.703659601492419e-05,
+      "loss": 0.8768,
+      "step": 9800
+    },
+    {
+      "epoch": 0.7787413919704698,
+      "grad_norm": 0.6399036049842834,
+      "learning_rate": 3.702336535153873e-05,
+      "loss": 0.8645,
+      "step": 9810
+    },
+    {
+      "epoch": 0.7795352160193693,
+      "grad_norm": 0.649523138999939,
+      "learning_rate": 3.701013468815326e-05,
+      "loss": 0.8536,
+      "step": 9820
+    },
+    {
+      "epoch": 0.7803290400682689,
+      "grad_norm": 0.955564558506012,
+      "learning_rate": 3.69969040247678e-05,
+      "loss": 0.8279,
+      "step": 9830
+    },
+    {
+      "epoch": 0.7811228641171685,
+      "grad_norm": 0.733024537563324,
+      "learning_rate": 3.698367336138234e-05,
+      "loss": 0.9158,
+      "step": 9840
+    },
+    {
+      "epoch": 0.781916688166068,
+      "grad_norm": 0.6793590784072876,
+      "learning_rate": 3.697044269799688e-05,
+      "loss": 0.8817,
+      "step": 9850
+    },
+    {
+      "epoch": 0.7827105122149676,
+      "grad_norm": 0.7847456932067871,
+      "learning_rate": 3.695721203461142e-05,
+      "loss": 0.8568,
+      "step": 9860
+    },
+    {
+      "epoch": 0.7835043362638671,
+      "grad_norm": 0.8460919260978699,
+      "learning_rate": 3.694398137122596e-05,
+      "loss": 0.8767,
+      "step": 9870
+    },
+    {
+      "epoch": 0.7842981603127667,
+      "grad_norm": 0.6977751851081848,
+      "learning_rate": 3.6930750707840496e-05,
+      "loss": 0.8671,
+      "step": 9880
+    },
+    {
+      "epoch": 0.7850919843616663,
+      "grad_norm": 0.7356807589530945,
+      "learning_rate": 3.691752004445503e-05,
+      "loss": 0.9341,
+      "step": 9890
+    },
+    {
+      "epoch": 0.7858858084105658,
+      "grad_norm": 0.7931004762649536,
+      "learning_rate": 3.690428938106957e-05,
+      "loss": 0.868,
+      "step": 9900
+    },
+    {
+      "epoch": 0.7866796324594654,
+      "grad_norm": 0.8790589570999146,
+      "learning_rate": 3.6891058717684106e-05,
+      "loss": 0.8516,
+      "step": 9910
+    },
+    {
+      "epoch": 0.7874734565083649,
+      "grad_norm": 0.7858555912971497,
+      "learning_rate": 3.6877828054298645e-05,
+      "loss": 0.8336,
+      "step": 9920
+    },
+    {
+      "epoch": 0.7882672805572645,
+      "grad_norm": 0.6692208647727966,
+      "learning_rate": 3.686459739091318e-05,
+      "loss": 0.8679,
+      "step": 9930
+    },
+    {
+      "epoch": 0.7890611046061641,
+      "grad_norm": 0.9050410389900208,
+      "learning_rate": 3.685136672752772e-05,
+      "loss": 0.7942,
+      "step": 9940
+    },
+    {
+      "epoch": 0.7898549286550636,
+      "grad_norm": 0.8038280010223389,
+      "learning_rate": 3.683813606414226e-05,
+      "loss": 0.8699,
+      "step": 9950
+    },
+    {
+      "epoch": 0.7906487527039632,
+      "grad_norm": 0.7280693054199219,
+      "learning_rate": 3.68249054007568e-05,
+      "loss": 0.8876,
+      "step": 9960
+    },
+    {
+      "epoch": 0.7914425767528628,
+      "grad_norm": 0.6296914219856262,
+      "learning_rate": 3.681167473737133e-05,
+      "loss": 0.8535,
+      "step": 9970
+    },
+    {
+      "epoch": 0.7922364008017623,
+      "grad_norm": 0.8135585784912109,
+      "learning_rate": 3.679844407398587e-05,
+      "loss": 0.8563,
+      "step": 9980
+    },
+    {
+      "epoch": 0.7930302248506619,
+      "grad_norm": 0.6743822693824768,
+      "learning_rate": 3.678521341060041e-05,
+      "loss": 0.9013,
+      "step": 9990
+    },
+    {
+      "epoch": 0.7938240488995614,
+      "grad_norm": 0.6657276153564453,
+      "learning_rate": 3.677198274721494e-05,
+      "loss": 0.8898,
+      "step": 10000
+    },
+    {
+      "epoch": 0.794617872948461,
+      "grad_norm": 0.7389101386070251,
+      "learning_rate": 3.675875208382948e-05,
+      "loss": 0.8925,
+      "step": 10010
+    },
+    {
+      "epoch": 0.7954116969973606,
+      "grad_norm": 0.802793562412262,
+      "learning_rate": 3.674552142044402e-05,
+      "loss": 0.8091,
+      "step": 10020
+    },
+    {
+      "epoch": 0.7962055210462601,
+      "grad_norm": 0.9091024398803711,
+      "learning_rate": 3.673229075705856e-05,
+      "loss": 0.8559,
+      "step": 10030
+    },
+    {
+      "epoch": 0.7969993450951597,
+      "grad_norm": 0.8592483997344971,
+      "learning_rate": 3.67190600936731e-05,
+      "loss": 0.8379,
+      "step": 10040
+    },
+    {
+      "epoch": 0.7977931691440592,
+      "grad_norm": 0.8377381563186646,
+      "learning_rate": 3.670582943028764e-05,
+      "loss": 0.9248,
+      "step": 10050
+    },
+    {
+      "epoch": 0.7985869931929588,
+      "grad_norm": 0.7962272763252258,
+      "learning_rate": 3.6692598766902176e-05,
+      "loss": 0.8846,
+      "step": 10060
+    },
+    {
+      "epoch": 0.7993808172418584,
+      "grad_norm": 0.5838531255722046,
+      "learning_rate": 3.6679368103516715e-05,
+      "loss": 0.8658,
+      "step": 10070
+    },
+    {
+      "epoch": 0.8001746412907579,
+      "grad_norm": 0.6025267839431763,
+      "learning_rate": 3.666613744013125e-05,
+      "loss": 0.8649,
+      "step": 10080
+    },
+    {
+      "epoch": 0.8009684653396575,
+      "grad_norm": 0.6443192362785339,
+      "learning_rate": 3.6652906776745786e-05,
+      "loss": 0.8555,
+      "step": 10090
+    },
+    {
+      "epoch": 0.801762289388557,
+      "grad_norm": 0.7511973977088928,
+      "learning_rate": 3.6639676113360325e-05,
+      "loss": 0.945,
+      "step": 10100
+    },
+    {
+      "epoch": 0.8025561134374566,
+      "grad_norm": 0.7938660979270935,
+      "learning_rate": 3.6626445449974864e-05,
+      "loss": 0.796,
+      "step": 10110
+    },
+    {
+      "epoch": 0.8033499374863562,
+      "grad_norm": 0.7330296039581299,
+      "learning_rate": 3.66132147865894e-05,
+      "loss": 0.7629,
+      "step": 10120
+    },
+    {
+      "epoch": 0.8041437615352557,
+      "grad_norm": 0.8548980355262756,
+      "learning_rate": 3.659998412320394e-05,
+      "loss": 0.8621,
+      "step": 10130
+    },
+    {
+      "epoch": 0.8049375855841553,
+      "grad_norm": 0.7704104781150818,
+      "learning_rate": 3.658675345981848e-05,
+      "loss": 0.8502,
+      "step": 10140
+    },
+    {
+      "epoch": 0.8057314096330548,
+      "grad_norm": 0.672539472579956,
+      "learning_rate": 3.657352279643301e-05,
+      "loss": 0.8981,
+      "step": 10150
+    },
+    {
+      "epoch": 0.8065252336819544,
+      "grad_norm": 0.8186033964157104,
+      "learning_rate": 3.656029213304755e-05,
+      "loss": 0.9071,
+      "step": 10160
+    },
+    {
+      "epoch": 0.807319057730854,
+      "grad_norm": 0.5315646529197693,
+      "learning_rate": 3.654706146966209e-05,
+      "loss": 0.8992,
+      "step": 10170
+    },
+    {
+      "epoch": 0.8081128817797535,
+      "grad_norm": 0.6407436728477478,
+      "learning_rate": 3.653383080627663e-05,
+      "loss": 0.8987,
+      "step": 10180
+    },
+    {
+      "epoch": 0.8089067058286531,
+      "grad_norm": 0.6754816770553589,
+      "learning_rate": 3.652060014289116e-05,
+      "loss": 0.8625,
+      "step": 10190
+    },
+    {
+      "epoch": 0.8097005298775526,
+      "grad_norm": 0.7109003663063049,
+      "learning_rate": 3.65073694795057e-05,
+      "loss": 0.8156,
+      "step": 10200
+    },
+    {
+      "epoch": 0.8104943539264522,
+      "grad_norm": 0.7097606658935547,
+      "learning_rate": 3.649413881612025e-05,
+      "loss": 0.8754,
+      "step": 10210
+    },
+    {
+      "epoch": 0.8112881779753518,
+      "grad_norm": 0.7361970543861389,
+      "learning_rate": 3.648090815273478e-05,
+      "loss": 0.8663,
+      "step": 10220
+    },
+    {
+      "epoch": 0.8120820020242513,
+      "grad_norm": 0.5668066740036011,
+      "learning_rate": 3.646767748934932e-05,
+      "loss": 0.8903,
+      "step": 10230
+    },
+    {
+      "epoch": 0.8128758260731509,
+      "grad_norm": 0.9411951899528503,
+      "learning_rate": 3.645444682596386e-05,
+      "loss": 0.8686,
+      "step": 10240
+    },
+    {
+      "epoch": 0.8136696501220505,
+      "grad_norm": 0.6737961769104004,
+      "learning_rate": 3.6441216162578396e-05,
+      "loss": 0.8497,
+      "step": 10250
+    },
+    {
+      "epoch": 0.81446347417095,
+      "grad_norm": 0.8757466673851013,
+      "learning_rate": 3.642798549919293e-05,
+      "loss": 0.8725,
+      "step": 10260
+    },
+    {
+      "epoch": 0.8152572982198496,
+      "grad_norm": 0.670192539691925,
+      "learning_rate": 3.641475483580747e-05,
+      "loss": 0.892,
+      "step": 10270
+    },
+    {
+      "epoch": 0.8160511222687491,
+      "grad_norm": 0.7439650893211365,
+      "learning_rate": 3.6401524172422006e-05,
+      "loss": 0.8589,
+      "step": 10280
+    },
+    {
+      "epoch": 0.8168449463176487,
+      "grad_norm": 0.795250654220581,
+      "learning_rate": 3.6388293509036545e-05,
+      "loss": 0.8543,
+      "step": 10290
+    },
+    {
+      "epoch": 0.8176387703665483,
+      "grad_norm": 0.7579794526100159,
+      "learning_rate": 3.6375062845651084e-05,
+      "loss": 0.8498,
+      "step": 10300
+    },
+    {
+      "epoch": 0.8184325944154478,
+      "grad_norm": 0.5933066606521606,
+      "learning_rate": 3.636183218226562e-05,
+      "loss": 0.8498,
+      "step": 10310
+    },
+    {
+      "epoch": 0.8192264184643474,
+      "grad_norm": 0.8285762667655945,
+      "learning_rate": 3.634860151888016e-05,
+      "loss": 0.9121,
+      "step": 10320
+    },
+    {
+      "epoch": 0.8200202425132469,
+      "grad_norm": 0.7152075171470642,
+      "learning_rate": 3.6335370855494694e-05,
+      "loss": 0.8678,
+      "step": 10330
+    },
+    {
+      "epoch": 0.8208140665621465,
+      "grad_norm": 0.933392345905304,
+      "learning_rate": 3.632214019210923e-05,
+      "loss": 0.8849,
+      "step": 10340
+    },
+    {
+      "epoch": 0.8216078906110461,
+      "grad_norm": 0.698727548122406,
+      "learning_rate": 3.630890952872377e-05,
+      "loss": 0.8407,
+      "step": 10350
+    },
+    {
+      "epoch": 0.8224017146599456,
+      "grad_norm": 0.6476231217384338,
+      "learning_rate": 3.629567886533831e-05,
+      "loss": 0.9283,
+      "step": 10360
+    },
+    {
+      "epoch": 0.8231955387088452,
+      "grad_norm": 0.6746705770492554,
+      "learning_rate": 3.628244820195284e-05,
+      "loss": 0.8968,
+      "step": 10370
+    },
+    {
+      "epoch": 0.8239893627577447,
+      "grad_norm": 0.654994547367096,
+      "learning_rate": 3.626921753856739e-05,
+      "loss": 0.871,
+      "step": 10380
+    },
+    {
+      "epoch": 0.8247831868066443,
+      "grad_norm": 0.9437777996063232,
+      "learning_rate": 3.625598687518193e-05,
+      "loss": 0.8428,
+      "step": 10390
+    },
+    {
+      "epoch": 0.8255770108555439,
+      "grad_norm": 0.6646280884742737,
+      "learning_rate": 3.6242756211796466e-05,
+      "loss": 0.9034,
+      "step": 10400
+    },
+    {
+      "epoch": 0.8263708349044434,
+      "grad_norm": 0.8529212474822998,
+      "learning_rate": 3.6229525548411e-05,
+      "loss": 0.8507,
+      "step": 10410
+    },
+    {
+      "epoch": 0.827164658953343,
+      "grad_norm": 0.7901684045791626,
+      "learning_rate": 3.621629488502554e-05,
+      "loss": 0.8191,
+      "step": 10420
+    },
+    {
+      "epoch": 0.8279584830022425,
+      "grad_norm": 0.6879044771194458,
+      "learning_rate": 3.6203064221640076e-05,
+      "loss": 0.8222,
+      "step": 10430
+    },
+    {
+      "epoch": 0.8287523070511421,
+      "grad_norm": 0.8270571827888489,
+      "learning_rate": 3.618983355825461e-05,
+      "loss": 0.8483,
+      "step": 10440
+    },
+    {
+      "epoch": 0.8295461311000417,
+      "grad_norm": 0.8302059769630432,
+      "learning_rate": 3.617660289486915e-05,
+      "loss": 0.8513,
+      "step": 10450
+    },
+    {
+      "epoch": 0.8303399551489412,
+      "grad_norm": 0.669792890548706,
+      "learning_rate": 3.6163372231483686e-05,
+      "loss": 0.8854,
+      "step": 10460
+    },
+    {
+      "epoch": 0.8311337791978408,
+      "grad_norm": 0.858910322189331,
+      "learning_rate": 3.6150141568098225e-05,
+      "loss": 0.8241,
+      "step": 10470
+    },
+    {
+      "epoch": 0.8319276032467403,
+      "grad_norm": 0.8056598901748657,
+      "learning_rate": 3.6136910904712764e-05,
+      "loss": 0.8923,
+      "step": 10480
+    },
+    {
+      "epoch": 0.8327214272956399,
+      "grad_norm": 0.733371913433075,
+      "learning_rate": 3.61236802413273e-05,
+      "loss": 0.8247,
+      "step": 10490
+    },
+    {
+      "epoch": 0.8335152513445395,
+      "grad_norm": 0.5836498141288757,
+      "learning_rate": 3.611044957794184e-05,
+      "loss": 0.8817,
+      "step": 10500
+    },
+    {
+      "epoch": 0.834309075393439,
+      "grad_norm": 0.8154313564300537,
+      "learning_rate": 3.609721891455638e-05,
+      "loss": 0.8183,
+      "step": 10510
+    },
+    {
+      "epoch": 0.8351028994423386,
+      "grad_norm": 0.7569496631622314,
+      "learning_rate": 3.608398825117091e-05,
+      "loss": 0.9073,
+      "step": 10520
+    },
+    {
+      "epoch": 0.8358967234912381,
+      "grad_norm": 0.6307011246681213,
+      "learning_rate": 3.607075758778545e-05,
+      "loss": 0.9599,
+      "step": 10530
+    },
+    {
+      "epoch": 0.8366905475401377,
+      "grad_norm": 0.7529721856117249,
+      "learning_rate": 3.605752692439999e-05,
+      "loss": 0.8456,
+      "step": 10540
+    },
+    {
+      "epoch": 0.8374843715890373,
+      "grad_norm": 0.8897116184234619,
+      "learning_rate": 3.604429626101452e-05,
+      "loss": 0.8309,
+      "step": 10550
+    },
+    {
+      "epoch": 0.8382781956379368,
+      "grad_norm": 0.8665919303894043,
+      "learning_rate": 3.603106559762907e-05,
+      "loss": 0.8252,
+      "step": 10560
+    },
+    {
+      "epoch": 0.8390720196868364,
+      "grad_norm": 0.7858225703239441,
+      "learning_rate": 3.601783493424361e-05,
+      "loss": 0.8592,
+      "step": 10570
+    },
+    {
+      "epoch": 0.839865843735736,
+      "grad_norm": 0.719430685043335,
+      "learning_rate": 3.6004604270858146e-05,
+      "loss": 0.8424,
+      "step": 10580
+    },
+    {
+      "epoch": 0.8406596677846355,
+      "grad_norm": 0.8680655360221863,
+      "learning_rate": 3.599137360747268e-05,
+      "loss": 0.8125,
+      "step": 10590
+    },
+    {
+      "epoch": 0.8414534918335351,
+      "grad_norm": 0.7084610462188721,
+      "learning_rate": 3.597814294408722e-05,
+      "loss": 0.878,
+      "step": 10600
+    },
+    {
+      "epoch": 0.8422473158824346,
+      "grad_norm": 0.9070943593978882,
+      "learning_rate": 3.5964912280701756e-05,
+      "loss": 0.8752,
+      "step": 10610
+    },
+    {
+      "epoch": 0.8430411399313342,
+      "grad_norm": 0.8004103302955627,
+      "learning_rate": 3.5951681617316295e-05,
+      "loss": 0.8462,
+      "step": 10620
+    },
+    {
+      "epoch": 0.8438349639802338,
+      "grad_norm": 0.7617958784103394,
+      "learning_rate": 3.593845095393083e-05,
+      "loss": 0.832,
+      "step": 10630
+    },
+    {
+      "epoch": 0.8446287880291333,
+      "grad_norm": 0.8618339896202087,
+      "learning_rate": 3.5925220290545366e-05,
+      "loss": 0.8751,
+      "step": 10640
+    },
+    {
+      "epoch": 0.8454226120780329,
+      "grad_norm": 0.6573039889335632,
+      "learning_rate": 3.5911989627159905e-05,
+      "loss": 0.8706,
+      "step": 10650
+    },
+    {
+      "epoch": 0.8462164361269324,
+      "grad_norm": 0.6365640163421631,
+      "learning_rate": 3.589875896377445e-05,
+      "loss": 0.8706,
+      "step": 10660
+    },
+    {
+      "epoch": 0.847010260175832,
+      "grad_norm": 0.712862491607666,
+      "learning_rate": 3.588552830038898e-05,
+      "loss": 0.7904,
+      "step": 10670
+    },
+    {
+      "epoch": 0.8478040842247316,
+      "grad_norm": 0.7353672981262207,
+      "learning_rate": 3.587229763700352e-05,
+      "loss": 0.8465,
+      "step": 10680
+    },
+    {
+      "epoch": 0.8485979082736311,
+      "grad_norm": 0.7351557612419128,
+      "learning_rate": 3.585906697361806e-05,
+      "loss": 0.8947,
+      "step": 10690
+    },
+    {
+      "epoch": 0.8493917323225307,
+      "grad_norm": 0.7744636535644531,
+      "learning_rate": 3.584583631023259e-05,
+      "loss": 0.8291,
+      "step": 10700
+    },
+    {
+      "epoch": 0.8501855563714302,
+      "grad_norm": 0.5523749589920044,
+      "learning_rate": 3.583260564684713e-05,
+      "loss": 0.9476,
+      "step": 10710
+    },
+    {
+      "epoch": 0.8509793804203298,
+      "grad_norm": 0.7999189496040344,
+      "learning_rate": 3.581937498346167e-05,
+      "loss": 0.7629,
+      "step": 10720
+    },
+    {
+      "epoch": 0.8517732044692294,
+      "grad_norm": 0.7351500988006592,
+      "learning_rate": 3.580614432007621e-05,
+      "loss": 0.8504,
+      "step": 10730
+    },
+    {
+      "epoch": 0.8525670285181289,
+      "grad_norm": 0.7090862393379211,
+      "learning_rate": 3.579291365669075e-05,
+      "loss": 0.9068,
+      "step": 10740
+    },
+    {
+      "epoch": 0.8533608525670285,
+      "grad_norm": 0.6834081411361694,
+      "learning_rate": 3.577968299330529e-05,
+      "loss": 0.8594,
+      "step": 10750
+    },
+    {
+      "epoch": 0.854154676615928,
+      "grad_norm": 0.7891108989715576,
+      "learning_rate": 3.576645232991983e-05,
+      "loss": 0.8918,
+      "step": 10760
+    },
+    {
+      "epoch": 0.8549485006648276,
+      "grad_norm": 0.7295839786529541,
+      "learning_rate": 3.5753221666534366e-05,
+      "loss": 0.8802,
+      "step": 10770
+    },
+    {
+      "epoch": 0.8557423247137272,
+      "grad_norm": 0.8786309361457825,
+      "learning_rate": 3.57399910031489e-05,
+      "loss": 0.8576,
+      "step": 10780
+    },
+    {
+      "epoch": 0.8565361487626267,
+      "grad_norm": 0.7351185083389282,
+      "learning_rate": 3.572676033976344e-05,
+      "loss": 0.8407,
+      "step": 10790
+    },
+    {
+      "epoch": 0.8573299728115263,
+      "grad_norm": 0.6069027185440063,
+      "learning_rate": 3.5713529676377976e-05,
+      "loss": 0.8566,
+      "step": 10800
+    },
+    {
+      "epoch": 0.8581237968604258,
+      "grad_norm": 0.8635075092315674,
+      "learning_rate": 3.570029901299251e-05,
+      "loss": 0.885,
+      "step": 10810
+    },
+    {
+      "epoch": 0.8589176209093254,
+      "grad_norm": 0.7878072261810303,
+      "learning_rate": 3.568706834960705e-05,
+      "loss": 0.8718,
+      "step": 10820
+    },
+    {
+      "epoch": 0.859711444958225,
+      "grad_norm": 0.7929555177688599,
+      "learning_rate": 3.567383768622159e-05,
+      "loss": 0.8686,
+      "step": 10830
+    },
+    {
+      "epoch": 0.8605052690071245,
+      "grad_norm": 0.8081250786781311,
+      "learning_rate": 3.566060702283613e-05,
+      "loss": 0.893,
+      "step": 10840
+    },
+    {
+      "epoch": 0.8612990930560241,
+      "grad_norm": 0.7977842688560486,
+      "learning_rate": 3.5647376359450664e-05,
+      "loss": 0.8753,
+      "step": 10850
+    },
+    {
+      "epoch": 0.8620929171049236,
+      "grad_norm": 0.7646862268447876,
+      "learning_rate": 3.56341456960652e-05,
+      "loss": 0.8658,
+      "step": 10860
+    },
+    {
+      "epoch": 0.8628867411538232,
+      "grad_norm": 0.7526537775993347,
+      "learning_rate": 3.562091503267974e-05,
+      "loss": 0.8557,
+      "step": 10870
+    },
+    {
+      "epoch": 0.8636805652027228,
+      "grad_norm": 0.7993239760398865,
+      "learning_rate": 3.560768436929428e-05,
+      "loss": 0.8521,
+      "step": 10880
+    },
+    {
+      "epoch": 0.8644743892516223,
+      "grad_norm": 0.9761634469032288,
+      "learning_rate": 3.559445370590881e-05,
+      "loss": 0.926,
+      "step": 10890
+    },
+    {
+      "epoch": 0.865268213300522,
+      "grad_norm": 0.7890929579734802,
+      "learning_rate": 3.558122304252335e-05,
+      "loss": 0.8633,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8660620373494216,
+      "grad_norm": 0.7080404758453369,
+      "learning_rate": 3.556799237913789e-05,
+      "loss": 0.9538,
+      "step": 10910
+    },
+    {
+      "epoch": 0.866855861398321,
+      "grad_norm": 0.8383463621139526,
+      "learning_rate": 3.555476171575243e-05,
+      "loss": 0.8548,
+      "step": 10920
+    },
+    {
+      "epoch": 0.8676496854472207,
+      "grad_norm": 0.8602036237716675,
+      "learning_rate": 3.554153105236697e-05,
+      "loss": 0.8676,
+      "step": 10930
+    },
+    {
+      "epoch": 0.8684435094961201,
+      "grad_norm": 0.7755318880081177,
+      "learning_rate": 3.552830038898151e-05,
+      "loss": 0.8335,
+      "step": 10940
+    },
+    {
+      "epoch": 0.8692373335450198,
+      "grad_norm": 0.7934598922729492,
+      "learning_rate": 3.5515069725596046e-05,
+      "loss": 0.7969,
+      "step": 10950
+    },
+    {
+      "epoch": 0.8700311575939194,
+      "grad_norm": 0.9328092932701111,
+      "learning_rate": 3.550183906221058e-05,
+      "loss": 0.7799,
+      "step": 10960
+    },
+    {
+      "epoch": 0.8708249816428189,
+      "grad_norm": 0.665879487991333,
+      "learning_rate": 3.548860839882512e-05,
+      "loss": 0.8617,
+      "step": 10970
+    },
+    {
+      "epoch": 0.8716188056917185,
+      "grad_norm": 0.775994598865509,
+      "learning_rate": 3.5475377735439656e-05,
+      "loss": 0.8282,
+      "step": 10980
+    },
+    {
+      "epoch": 0.872412629740618,
+      "grad_norm": 0.8753870129585266,
+      "learning_rate": 3.5462147072054195e-05,
+      "loss": 0.786,
+      "step": 10990
+    },
+    {
+      "epoch": 0.8732064537895176,
+      "grad_norm": 0.8270352482795715,
+      "learning_rate": 3.5448916408668734e-05,
+      "loss": 0.8371,
+      "step": 11000
+    },
+    {
+      "epoch": 0.8740002778384172,
+      "grad_norm": 0.7965131402015686,
+      "learning_rate": 3.543568574528327e-05,
+      "loss": 0.885,
+      "step": 11010
+    },
+    {
+      "epoch": 0.8747941018873167,
+      "grad_norm": 0.7161972522735596,
+      "learning_rate": 3.542245508189781e-05,
+      "loss": 0.7859,
+      "step": 11020
+    },
+    {
+      "epoch": 0.8755879259362163,
+      "grad_norm": 0.7473718523979187,
+      "learning_rate": 3.5409224418512344e-05,
+      "loss": 0.8075,
+      "step": 11030
+    },
+    {
+      "epoch": 0.8763817499851158,
+      "grad_norm": 0.8067750334739685,
+      "learning_rate": 3.539599375512688e-05,
+      "loss": 0.7951,
+      "step": 11040
+    },
+    {
+      "epoch": 0.8771755740340154,
+      "grad_norm": 0.8966103196144104,
+      "learning_rate": 3.538276309174142e-05,
+      "loss": 0.8467,
+      "step": 11050
+    },
+    {
+      "epoch": 0.877969398082915,
+      "grad_norm": 0.7432875037193298,
+      "learning_rate": 3.536953242835596e-05,
+      "loss": 0.8957,
+      "step": 11060
+    },
+    {
+      "epoch": 0.8787632221318145,
+      "grad_norm": 0.7818832993507385,
+      "learning_rate": 3.535630176497049e-05,
+      "loss": 0.9194,
+      "step": 11070
+    },
+    {
+      "epoch": 0.8795570461807141,
+      "grad_norm": 0.8966363668441772,
+      "learning_rate": 3.534307110158503e-05,
+      "loss": 0.8665,
+      "step": 11080
+    },
+    {
+      "epoch": 0.8803508702296136,
+      "grad_norm": 0.5509730577468872,
+      "learning_rate": 3.532984043819957e-05,
+      "loss": 0.9304,
+      "step": 11090
+    },
+    {
+      "epoch": 0.8811446942785132,
+      "grad_norm": 0.7345530986785889,
+      "learning_rate": 3.5316609774814116e-05,
+      "loss": 0.8524,
+      "step": 11100
+    },
+    {
+      "epoch": 0.8819385183274128,
+      "grad_norm": 0.9241631031036377,
+      "learning_rate": 3.530337911142865e-05,
+      "loss": 0.8748,
+      "step": 11110
+    },
+    {
+      "epoch": 0.8827323423763123,
+      "grad_norm": 0.7990071177482605,
+      "learning_rate": 3.529014844804319e-05,
+      "loss": 0.8633,
+      "step": 11120
+    },
+    {
+      "epoch": 0.8835261664252119,
+      "grad_norm": 0.8361101746559143,
+      "learning_rate": 3.5276917784657726e-05,
+      "loss": 0.8708,
+      "step": 11130
+    },
+    {
+      "epoch": 0.8843199904741114,
+      "grad_norm": 0.6737117767333984,
+      "learning_rate": 3.526368712127226e-05,
+      "loss": 0.8489,
+      "step": 11140
+    },
+    {
+      "epoch": 0.885113814523011,
+      "grad_norm": 0.8274135589599609,
+      "learning_rate": 3.52504564578868e-05,
+      "loss": 0.9062,
+      "step": 11150
+    },
+    {
+      "epoch": 0.8859076385719106,
+      "grad_norm": 0.5696777701377869,
+      "learning_rate": 3.5237225794501336e-05,
+      "loss": 0.8464,
+      "step": 11160
+    },
+    {
+      "epoch": 0.8867014626208101,
+      "grad_norm": 0.8153122067451477,
+      "learning_rate": 3.5223995131115875e-05,
+      "loss": 0.8514,
+      "step": 11170
+    },
+    {
+      "epoch": 0.8874952866697097,
+      "grad_norm": 0.7769964933395386,
+      "learning_rate": 3.5210764467730414e-05,
+      "loss": 0.8245,
+      "step": 11180
+    },
+    {
+      "epoch": 0.8882891107186092,
+      "grad_norm": 0.6569917798042297,
+      "learning_rate": 3.519753380434495e-05,
+      "loss": 0.799,
+      "step": 11190
+    },
+    {
+      "epoch": 0.8890829347675088,
+      "grad_norm": 0.5966030359268188,
+      "learning_rate": 3.518430314095949e-05,
+      "loss": 0.8563,
+      "step": 11200
+    },
+    {
+      "epoch": 0.8898767588164084,
+      "grad_norm": 0.6492002010345459,
+      "learning_rate": 3.517107247757403e-05,
+      "loss": 0.861,
+      "step": 11210
+    },
+    {
+      "epoch": 0.8906705828653079,
+      "grad_norm": 0.7151293158531189,
+      "learning_rate": 3.515784181418856e-05,
+      "loss": 0.8536,
+      "step": 11220
+    },
+    {
+      "epoch": 0.8914644069142075,
+      "grad_norm": 0.6742676496505737,
+      "learning_rate": 3.51446111508031e-05,
+      "loss": 0.8548,
+      "step": 11230
+    },
+    {
+      "epoch": 0.8922582309631071,
+      "grad_norm": 0.8235064148902893,
+      "learning_rate": 3.513138048741764e-05,
+      "loss": 0.8937,
+      "step": 11240
+    },
+    {
+      "epoch": 0.8930520550120066,
+      "grad_norm": 0.6672957539558411,
+      "learning_rate": 3.511814982403217e-05,
+      "loss": 0.8116,
+      "step": 11250
+    },
+    {
+      "epoch": 0.8938458790609062,
+      "grad_norm": 0.745836079120636,
+      "learning_rate": 3.510491916064671e-05,
+      "loss": 0.8183,
+      "step": 11260
+    },
+    {
+      "epoch": 0.8946397031098057,
+      "grad_norm": 0.93377286195755,
+      "learning_rate": 3.509168849726126e-05,
+      "loss": 0.87,
+      "step": 11270
+    },
+    {
+      "epoch": 0.8954335271587053,
+      "grad_norm": 0.7982069253921509,
+      "learning_rate": 3.50784578338758e-05,
+      "loss": 0.8557,
+      "step": 11280
+    },
+    {
+      "epoch": 0.8962273512076049,
+      "grad_norm": 0.7779275178909302,
+      "learning_rate": 3.506522717049033e-05,
+      "loss": 0.7948,
+      "step": 11290
+    },
+    {
+      "epoch": 0.8970211752565044,
+      "grad_norm": 0.7200841307640076,
+      "learning_rate": 3.505199650710487e-05,
+      "loss": 0.8031,
+      "step": 11300
+    },
+    {
+      "epoch": 0.897814999305404,
+      "grad_norm": 0.7799506783485413,
+      "learning_rate": 3.503876584371941e-05,
+      "loss": 0.8692,
+      "step": 11310
+    },
+    {
+      "epoch": 0.8986088233543035,
+      "grad_norm": 0.6927218437194824,
+      "learning_rate": 3.5025535180333946e-05,
+      "loss": 0.8854,
+      "step": 11320
+    },
+    {
+      "epoch": 0.8994026474032031,
+      "grad_norm": 0.8087165951728821,
+      "learning_rate": 3.501230451694848e-05,
+      "loss": 0.8593,
+      "step": 11330
+    },
+    {
+      "epoch": 0.9001964714521027,
+      "grad_norm": 0.817240834236145,
+      "learning_rate": 3.499907385356302e-05,
+      "loss": 0.8312,
+      "step": 11340
+    },
+    {
+      "epoch": 0.9009902955010022,
+      "grad_norm": 0.6710502505302429,
+      "learning_rate": 3.4985843190177556e-05,
+      "loss": 0.883,
+      "step": 11350
+    },
+    {
+      "epoch": 0.9017841195499018,
+      "grad_norm": 0.6229230761528015,
+      "learning_rate": 3.4972612526792095e-05,
+      "loss": 0.885,
+      "step": 11360
+    },
+    {
+      "epoch": 0.9025779435988013,
+      "grad_norm": 0.7668790221214294,
+      "learning_rate": 3.4959381863406634e-05,
+      "loss": 0.8206,
+      "step": 11370
+    },
+    {
+      "epoch": 0.9033717676477009,
+      "grad_norm": 0.668046236038208,
+      "learning_rate": 3.494615120002117e-05,
+      "loss": 0.8866,
+      "step": 11380
+    },
+    {
+      "epoch": 0.9041655916966005,
+      "grad_norm": 0.7366786003112793,
+      "learning_rate": 3.493292053663571e-05,
+      "loss": 0.8091,
+      "step": 11390
+    },
+    {
+      "epoch": 0.9049594157455,
+      "grad_norm": 0.9512502551078796,
+      "learning_rate": 3.4919689873250244e-05,
+      "loss": 0.7877,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9057532397943996,
+      "grad_norm": 0.8555247783660889,
+      "learning_rate": 3.490645920986478e-05,
+      "loss": 0.901,
+      "step": 11410
+    },
+    {
+      "epoch": 0.9065470638432991,
+      "grad_norm": 0.6285889148712158,
+      "learning_rate": 3.489322854647932e-05,
+      "loss": 0.8963,
+      "step": 11420
+    },
+    {
+      "epoch": 0.9073408878921987,
+      "grad_norm": 0.7358067631721497,
+      "learning_rate": 3.487999788309386e-05,
+      "loss": 0.8679,
+      "step": 11430
+    },
+    {
+      "epoch": 0.9081347119410983,
+      "grad_norm": 0.8041672110557556,
+      "learning_rate": 3.48667672197084e-05,
+      "loss": 0.8222,
+      "step": 11440
+    },
+    {
+      "epoch": 0.9089285359899978,
+      "grad_norm": 0.7826716899871826,
+      "learning_rate": 3.485353655632294e-05,
+      "loss": 0.8347,
+      "step": 11450
+    },
+    {
+      "epoch": 0.9097223600388974,
+      "grad_norm": 0.8296499252319336,
+      "learning_rate": 3.484030589293748e-05,
+      "loss": 0.8105,
+      "step": 11460
+    },
+    {
+      "epoch": 0.9105161840877969,
+      "grad_norm": 0.6100326776504517,
+      "learning_rate": 3.482707522955201e-05,
+      "loss": 0.8618,
+      "step": 11470
+    },
+    {
+      "epoch": 0.9113100081366965,
+      "grad_norm": 0.9201139807701111,
+      "learning_rate": 3.481384456616655e-05,
+      "loss": 0.905,
+      "step": 11480
+    },
+    {
+      "epoch": 0.9121038321855961,
+      "grad_norm": 0.8482776880264282,
+      "learning_rate": 3.480061390278109e-05,
+      "loss": 0.8644,
+      "step": 11490
+    },
+    {
+      "epoch": 0.9128976562344956,
+      "grad_norm": 1.077039122581482,
+      "learning_rate": 3.4787383239395626e-05,
+      "loss": 0.848,
+      "step": 11500
+    },
+    {
+      "epoch": 0.9136914802833952,
+      "grad_norm": 0.6531374454498291,
+      "learning_rate": 3.477415257601016e-05,
+      "loss": 0.8716,
+      "step": 11510
+    },
+    {
+      "epoch": 0.9144853043322948,
+      "grad_norm": 0.7523869276046753,
+      "learning_rate": 3.47609219126247e-05,
+      "loss": 0.9086,
+      "step": 11520
+    },
+    {
+      "epoch": 0.9152791283811943,
+      "grad_norm": 0.7146991491317749,
+      "learning_rate": 3.4747691249239236e-05,
+      "loss": 0.8156,
+      "step": 11530
+    },
+    {
+      "epoch": 0.9160729524300939,
+      "grad_norm": 0.916566789150238,
+      "learning_rate": 3.473446058585378e-05,
+      "loss": 0.9163,
+      "step": 11540
+    },
+    {
+      "epoch": 0.9168667764789934,
+      "grad_norm": 0.8944635987281799,
+      "learning_rate": 3.4721229922468314e-05,
+      "loss": 0.8811,
+      "step": 11550
+    },
+    {
+      "epoch": 0.917660600527893,
+      "grad_norm": 0.7929695248603821,
+      "learning_rate": 3.470799925908285e-05,
+      "loss": 0.8048,
+      "step": 11560
+    },
+    {
+      "epoch": 0.9184544245767926,
+      "grad_norm": 0.6163130402565002,
+      "learning_rate": 3.469476859569739e-05,
+      "loss": 0.904,
+      "step": 11570
+    },
+    {
+      "epoch": 0.9192482486256921,
+      "grad_norm": 0.7213409543037415,
+      "learning_rate": 3.468153793231193e-05,
+      "loss": 0.9098,
+      "step": 11580
+    },
+    {
+      "epoch": 0.9200420726745917,
+      "grad_norm": 0.8145052790641785,
+      "learning_rate": 3.466830726892646e-05,
+      "loss": 0.8596,
+      "step": 11590
+    },
+    {
+      "epoch": 0.9208358967234912,
+      "grad_norm": 0.7872694730758667,
+      "learning_rate": 3.4655076605541e-05,
+      "loss": 0.8214,
+      "step": 11600
+    },
+    {
+      "epoch": 0.9216297207723908,
+      "grad_norm": 0.745549201965332,
+      "learning_rate": 3.464184594215554e-05,
+      "loss": 0.885,
+      "step": 11610
+    },
+    {
+      "epoch": 0.9224235448212904,
+      "grad_norm": 0.8459563255310059,
+      "learning_rate": 3.462861527877008e-05,
+      "loss": 0.9197,
+      "step": 11620
+    },
+    {
+      "epoch": 0.9232173688701899,
+      "grad_norm": 0.9419437646865845,
+      "learning_rate": 3.461538461538462e-05,
+      "loss": 0.7556,
+      "step": 11630
+    },
+    {
+      "epoch": 0.9240111929190895,
+      "grad_norm": 0.7619801759719849,
+      "learning_rate": 3.460215395199916e-05,
+      "loss": 0.87,
+      "step": 11640
+    },
+    {
+      "epoch": 0.924805016967989,
+      "grad_norm": 0.5955981612205505,
+      "learning_rate": 3.4588923288613696e-05,
+      "loss": 0.922,
+      "step": 11650
+    },
+    {
+      "epoch": 0.9255988410168886,
+      "grad_norm": 0.7901360392570496,
+      "learning_rate": 3.457569262522823e-05,
+      "loss": 0.8939,
+      "step": 11660
+    },
+    {
+      "epoch": 0.9263926650657882,
+      "grad_norm": 0.6033757328987122,
+      "learning_rate": 3.456246196184277e-05,
+      "loss": 0.7984,
+      "step": 11670
+    },
+    {
+      "epoch": 0.9271864891146877,
+      "grad_norm": 0.7921754121780396,
+      "learning_rate": 3.4549231298457306e-05,
+      "loss": 0.8477,
+      "step": 11680
+    },
+    {
+      "epoch": 0.9279803131635873,
+      "grad_norm": 0.6191670298576355,
+      "learning_rate": 3.4536000635071845e-05,
+      "loss": 0.8468,
+      "step": 11690
+    },
+    {
+      "epoch": 0.9287741372124868,
+      "grad_norm": 0.6756204962730408,
+      "learning_rate": 3.452276997168638e-05,
+      "loss": 0.8113,
+      "step": 11700
+    },
+    {
+      "epoch": 0.9295679612613864,
+      "grad_norm": 0.8277655839920044,
+      "learning_rate": 3.450953930830092e-05,
+      "loss": 0.889,
+      "step": 11710
+    },
+    {
+      "epoch": 0.930361785310286,
+      "grad_norm": 0.6921846270561218,
+      "learning_rate": 3.449630864491546e-05,
+      "loss": 0.9316,
+      "step": 11720
+    },
+    {
+      "epoch": 0.9311556093591855,
+      "grad_norm": 0.6792823672294617,
+      "learning_rate": 3.4483077981529994e-05,
+      "loss": 0.8069,
+      "step": 11730
+    },
+    {
+      "epoch": 0.9319494334080851,
+      "grad_norm": 0.9177089929580688,
+      "learning_rate": 3.446984731814453e-05,
+      "loss": 0.8124,
+      "step": 11740
+    },
+    {
+      "epoch": 0.9327432574569846,
+      "grad_norm": 0.8078745603561401,
+      "learning_rate": 3.445661665475907e-05,
+      "loss": 0.8762,
+      "step": 11750
+    },
+    {
+      "epoch": 0.9335370815058842,
+      "grad_norm": 0.9108356833457947,
+      "learning_rate": 3.444338599137361e-05,
+      "loss": 0.7767,
+      "step": 11760
+    },
+    {
+      "epoch": 0.9343309055547838,
+      "grad_norm": 0.6663825511932373,
+      "learning_rate": 3.443015532798814e-05,
+      "loss": 0.8749,
+      "step": 11770
+    },
+    {
+      "epoch": 0.9351247296036833,
+      "grad_norm": 0.8398308753967285,
+      "learning_rate": 3.441692466460268e-05,
+      "loss": 0.8793,
+      "step": 11780
+    },
+    {
+      "epoch": 0.9359185536525829,
+      "grad_norm": 0.8784694075584412,
+      "learning_rate": 3.440369400121722e-05,
+      "loss": 0.8916,
+      "step": 11790
+    },
+    {
+      "epoch": 0.9367123777014824,
+      "grad_norm": 0.7957909107208252,
+      "learning_rate": 3.439046333783176e-05,
+      "loss": 0.8855,
+      "step": 11800
+    },
+    {
+      "epoch": 0.937506201750382,
+      "grad_norm": 0.6235615015029907,
+      "learning_rate": 3.43772326744463e-05,
+      "loss": 0.9407,
+      "step": 11810
+    },
+    {
+      "epoch": 0.9383000257992816,
+      "grad_norm": 0.912480354309082,
+      "learning_rate": 3.436400201106084e-05,
+      "loss": 0.8935,
+      "step": 11820
+    },
+    {
+      "epoch": 0.9390938498481811,
+      "grad_norm": 0.8489815592765808,
+      "learning_rate": 3.435077134767538e-05,
+      "loss": 0.8482,
+      "step": 11830
+    },
+    {
+      "epoch": 0.9398876738970807,
+      "grad_norm": 0.6857966780662537,
+      "learning_rate": 3.433754068428991e-05,
+      "loss": 0.8982,
+      "step": 11840
+    },
+    {
+      "epoch": 0.9406814979459803,
+      "grad_norm": 0.6412976384162903,
+      "learning_rate": 3.432431002090445e-05,
+      "loss": 0.8496,
+      "step": 11850
+    },
+    {
+      "epoch": 0.9414753219948798,
+      "grad_norm": 0.7496560215950012,
+      "learning_rate": 3.431107935751899e-05,
+      "loss": 0.8434,
+      "step": 11860
+    },
+    {
+      "epoch": 0.9422691460437794,
+      "grad_norm": 0.7269144058227539,
+      "learning_rate": 3.4297848694133526e-05,
+      "loss": 0.8561,
+      "step": 11870
+    },
+    {
+      "epoch": 0.9430629700926789,
+      "grad_norm": 0.8505829572677612,
+      "learning_rate": 3.4284618030748065e-05,
+      "loss": 0.8006,
+      "step": 11880
+    },
+    {
+      "epoch": 0.9438567941415785,
+      "grad_norm": 0.7966324090957642,
+      "learning_rate": 3.4271387367362604e-05,
+      "loss": 0.8968,
+      "step": 11890
+    },
+    {
+      "epoch": 0.9446506181904781,
+      "grad_norm": 0.7437507510185242,
+      "learning_rate": 3.425815670397714e-05,
+      "loss": 0.8193,
+      "step": 11900
+    },
+    {
+      "epoch": 0.9454444422393776,
+      "grad_norm": 0.8566707968711853,
+      "learning_rate": 3.424492604059168e-05,
+      "loss": 0.8084,
+      "step": 11910
+    },
+    {
+      "epoch": 0.9462382662882772,
+      "grad_norm": 0.6849654316902161,
+      "learning_rate": 3.4231695377206214e-05,
+      "loss": 0.8792,
+      "step": 11920
+    },
+    {
+      "epoch": 0.9470320903371767,
+      "grad_norm": 0.7506431341171265,
+      "learning_rate": 3.421846471382075e-05,
+      "loss": 0.8374,
+      "step": 11930
+    },
+    {
+      "epoch": 0.9478259143860763,
+      "grad_norm": 0.6591388583183289,
+      "learning_rate": 3.420523405043529e-05,
+      "loss": 0.9128,
+      "step": 11940
+    },
+    {
+      "epoch": 0.9486197384349759,
+      "grad_norm": 0.6423110365867615,
+      "learning_rate": 3.4192003387049824e-05,
+      "loss": 0.7918,
+      "step": 11950
+    },
+    {
+      "epoch": 0.9494135624838754,
+      "grad_norm": 0.6780298948287964,
+      "learning_rate": 3.417877272366436e-05,
+      "loss": 0.8715,
+      "step": 11960
+    },
+    {
+      "epoch": 0.950207386532775,
+      "grad_norm": 0.6950411200523376,
+      "learning_rate": 3.41655420602789e-05,
+      "loss": 0.8555,
+      "step": 11970
+    },
+    {
+      "epoch": 0.9510012105816745,
+      "grad_norm": 0.8524368405342102,
+      "learning_rate": 3.415231139689345e-05,
+      "loss": 0.8706,
+      "step": 11980
+    },
+    {
+      "epoch": 0.9517950346305741,
+      "grad_norm": 0.8601171374320984,
+      "learning_rate": 3.413908073350798e-05,
+      "loss": 0.8835,
+      "step": 11990
+    },
+    {
+      "epoch": 0.9525888586794737,
+      "grad_norm": 0.7460906505584717,
+      "learning_rate": 3.412585007012252e-05,
+      "loss": 0.8911,
+      "step": 12000
+    },
+    {
+      "epoch": 0.9533826827283732,
+      "grad_norm": 0.6877840757369995,
+      "learning_rate": 3.411261940673706e-05,
+      "loss": 0.8733,
+      "step": 12010
+    },
+    {
+      "epoch": 0.9541765067772728,
+      "grad_norm": 0.8841610550880432,
+      "learning_rate": 3.4099388743351596e-05,
+      "loss": 0.8988,
+      "step": 12020
+    },
+    {
+      "epoch": 0.9549703308261723,
+      "grad_norm": 0.7192667126655579,
+      "learning_rate": 3.408615807996613e-05,
+      "loss": 0.876,
+      "step": 12030
+    },
+    {
+      "epoch": 0.9557641548750719,
+      "grad_norm": 0.739669144153595,
+      "learning_rate": 3.407292741658067e-05,
+      "loss": 0.9147,
+      "step": 12040
+    },
+    {
+      "epoch": 0.9565579789239715,
+      "grad_norm": 0.9102823138237,
+      "learning_rate": 3.4059696753195206e-05,
+      "loss": 0.8335,
+      "step": 12050
+    },
+    {
+      "epoch": 0.957351802972871,
+      "grad_norm": 0.8623813986778259,
+      "learning_rate": 3.4046466089809745e-05,
+      "loss": 0.9067,
+      "step": 12060
+    },
+    {
+      "epoch": 0.9581456270217706,
+      "grad_norm": 0.6548823714256287,
+      "learning_rate": 3.4033235426424284e-05,
+      "loss": 0.8553,
+      "step": 12070
+    },
+    {
+      "epoch": 0.9589394510706701,
+      "grad_norm": 0.6950840950012207,
+      "learning_rate": 3.402000476303882e-05,
+      "loss": 0.8346,
+      "step": 12080
+    },
+    {
+      "epoch": 0.9597332751195697,
+      "grad_norm": 0.8395444750785828,
+      "learning_rate": 3.400677409965336e-05,
+      "loss": 0.8331,
+      "step": 12090
+    },
+    {
+      "epoch": 0.9605270991684693,
+      "grad_norm": 0.7684609889984131,
+      "learning_rate": 3.3993543436267894e-05,
+      "loss": 0.8309,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9613209232173688,
+      "grad_norm": 0.6737335324287415,
+      "learning_rate": 3.398031277288243e-05,
+      "loss": 0.8495,
+      "step": 12110
+    },
+    {
+      "epoch": 0.9621147472662684,
+      "grad_norm": 0.8039416670799255,
+      "learning_rate": 3.396708210949697e-05,
+      "loss": 0.8201,
+      "step": 12120
+    },
+    {
+      "epoch": 0.9629085713151679,
+      "grad_norm": 0.7795635461807251,
+      "learning_rate": 3.395385144611151e-05,
+      "loss": 0.8921,
+      "step": 12130
+    },
+    {
+      "epoch": 0.9637023953640675,
+      "grad_norm": 0.8788042068481445,
+      "learning_rate": 3.394062078272604e-05,
+      "loss": 0.849,
+      "step": 12140
+    },
+    {
+      "epoch": 0.9644962194129671,
+      "grad_norm": 0.8527970910072327,
+      "learning_rate": 3.392739011934058e-05,
+      "loss": 0.8119,
+      "step": 12150
+    },
+    {
+      "epoch": 0.9652900434618666,
+      "grad_norm": 0.7557294964790344,
+      "learning_rate": 3.391415945595513e-05,
+      "loss": 0.8121,
+      "step": 12160
+    },
+    {
+      "epoch": 0.9660838675107662,
+      "grad_norm": 0.871924877166748,
+      "learning_rate": 3.390092879256966e-05,
+      "loss": 0.9318,
+      "step": 12170
+    },
+    {
+      "epoch": 0.9668776915596659,
+      "grad_norm": 0.6994668245315552,
+      "learning_rate": 3.38876981291842e-05,
+      "loss": 0.8465,
+      "step": 12180
+    },
+    {
+      "epoch": 0.9676715156085653,
+      "grad_norm": 0.7652361989021301,
+      "learning_rate": 3.387446746579874e-05,
+      "loss": 0.8967,
+      "step": 12190
+    },
+    {
+      "epoch": 0.968465339657465,
+      "grad_norm": 0.7152222990989685,
+      "learning_rate": 3.3861236802413276e-05,
+      "loss": 0.9297,
+      "step": 12200
+    },
+    {
+      "epoch": 0.9692591637063644,
+      "grad_norm": 0.8487961292266846,
+      "learning_rate": 3.384800613902781e-05,
+      "loss": 0.8412,
+      "step": 12210
+    },
+    {
+      "epoch": 0.970052987755264,
+      "grad_norm": 0.7079365849494934,
+      "learning_rate": 3.383477547564235e-05,
+      "loss": 0.8746,
+      "step": 12220
+    },
+    {
+      "epoch": 0.9708468118041637,
+      "grad_norm": 0.830172598361969,
+      "learning_rate": 3.3821544812256886e-05,
+      "loss": 0.864,
+      "step": 12230
+    },
+    {
+      "epoch": 0.9716406358530632,
+      "grad_norm": 0.9805923104286194,
+      "learning_rate": 3.3808314148871425e-05,
+      "loss": 0.9199,
+      "step": 12240
+    },
+    {
+      "epoch": 0.9724344599019628,
+      "grad_norm": 0.7402176856994629,
+      "learning_rate": 3.3795083485485964e-05,
+      "loss": 0.8274,
+      "step": 12250
+    },
+    {
+      "epoch": 0.9732282839508622,
+      "grad_norm": 0.7366051077842712,
+      "learning_rate": 3.37818528221005e-05,
+      "loss": 0.8966,
+      "step": 12260
+    },
+    {
+      "epoch": 0.9740221079997619,
+      "grad_norm": 0.8370699286460876,
+      "learning_rate": 3.376862215871504e-05,
+      "loss": 0.7642,
+      "step": 12270
+    },
+    {
+      "epoch": 0.9748159320486615,
+      "grad_norm": 0.8621345162391663,
+      "learning_rate": 3.3755391495329574e-05,
+      "loss": 0.8868,
+      "step": 12280
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 0.6125873327255249,
+      "learning_rate": 3.374216083194411e-05,
+      "loss": 0.8572,
+      "step": 12290
+    },
+    {
+      "epoch": 0.9764035801464606,
+      "grad_norm": 0.708325207233429,
+      "learning_rate": 3.372893016855865e-05,
+      "loss": 0.811,
+      "step": 12300
+    },
+    {
+      "epoch": 0.97719740419536,
+      "grad_norm": 0.9190065264701843,
+      "learning_rate": 3.371569950517319e-05,
+      "loss": 0.7922,
+      "step": 12310
+    },
+    {
+      "epoch": 0.9779912282442597,
+      "grad_norm": 0.6722436547279358,
+      "learning_rate": 3.370246884178772e-05,
+      "loss": 0.8409,
+      "step": 12320
+    },
+    {
+      "epoch": 0.9787850522931593,
+      "grad_norm": 0.6370068788528442,
+      "learning_rate": 3.368923817840227e-05,
+      "loss": 0.8367,
+      "step": 12330
+    },
+    {
+      "epoch": 0.9795788763420588,
+      "grad_norm": 0.7187243103981018,
+      "learning_rate": 3.367600751501681e-05,
+      "loss": 0.8673,
+      "step": 12340
+    },
+    {
+      "epoch": 0.9803727003909584,
+      "grad_norm": 0.6579310297966003,
+      "learning_rate": 3.366277685163135e-05,
+      "loss": 0.8582,
+      "step": 12350
+    },
+    {
+      "epoch": 0.9811665244398579,
+      "grad_norm": 0.9233648777008057,
+      "learning_rate": 3.364954618824588e-05,
+      "loss": 0.8838,
+      "step": 12360
+    },
+    {
+      "epoch": 0.9819603484887575,
+      "grad_norm": 0.8820499181747437,
+      "learning_rate": 3.363631552486042e-05,
+      "loss": 0.8572,
+      "step": 12370
+    },
+    {
+      "epoch": 0.9827541725376571,
+      "grad_norm": 0.7715148329734802,
+      "learning_rate": 3.362308486147496e-05,
+      "loss": 0.8855,
+      "step": 12380
+    },
+    {
+      "epoch": 0.9835479965865566,
+      "grad_norm": 0.85701984167099,
+      "learning_rate": 3.3609854198089496e-05,
+      "loss": 0.8448,
+      "step": 12390
+    },
+    {
+      "epoch": 0.9843418206354562,
+      "grad_norm": 0.8934018015861511,
+      "learning_rate": 3.359662353470403e-05,
+      "loss": 0.7882,
+      "step": 12400
+    },
+    {
+      "epoch": 0.9851356446843557,
+      "grad_norm": 0.7508744597434998,
+      "learning_rate": 3.358339287131857e-05,
+      "loss": 0.8488,
+      "step": 12410
+    },
+    {
+      "epoch": 0.9859294687332553,
+      "grad_norm": 0.7345721125602722,
+      "learning_rate": 3.3570162207933106e-05,
+      "loss": 0.8662,
+      "step": 12420
+    },
+    {
+      "epoch": 0.9867232927821549,
+      "grad_norm": 0.7805215120315552,
+      "learning_rate": 3.3556931544547645e-05,
+      "loss": 0.7834,
+      "step": 12430
+    },
+    {
+      "epoch": 0.9875171168310544,
+      "grad_norm": 0.8609070181846619,
+      "learning_rate": 3.3543700881162184e-05,
+      "loss": 0.7946,
+      "step": 12440
+    },
+    {
+      "epoch": 0.988310940879954,
+      "grad_norm": 0.8504823446273804,
+      "learning_rate": 3.353047021777672e-05,
+      "loss": 0.8471,
+      "step": 12450
+    },
+    {
+      "epoch": 0.9891047649288536,
+      "grad_norm": 0.800946831703186,
+      "learning_rate": 3.351723955439126e-05,
+      "loss": 0.8686,
+      "step": 12460
+    },
+    {
+      "epoch": 0.9898985889777531,
+      "grad_norm": 0.770326554775238,
+      "learning_rate": 3.3504008891005794e-05,
+      "loss": 0.7954,
+      "step": 12470
+    },
+    {
+      "epoch": 0.9906924130266527,
+      "grad_norm": 0.7792863845825195,
+      "learning_rate": 3.349077822762033e-05,
+      "loss": 0.8636,
+      "step": 12480
+    },
+    {
+      "epoch": 0.9914862370755522,
+      "grad_norm": 0.7109197974205017,
+      "learning_rate": 3.347754756423487e-05,
+      "loss": 0.9198,
+      "step": 12490
+    },
+    {
+      "epoch": 0.9922800611244518,
+      "grad_norm": 0.6527384519577026,
+      "learning_rate": 3.346431690084941e-05,
+      "loss": 0.8457,
+      "step": 12500
+    },
+    {
+      "epoch": 0.9930738851733514,
+      "grad_norm": 0.9128293395042419,
+      "learning_rate": 3.345108623746395e-05,
+      "loss": 0.9064,
+      "step": 12510
+    },
+    {
+      "epoch": 0.9938677092222509,
+      "grad_norm": 0.7357638478279114,
+      "learning_rate": 3.343785557407849e-05,
+      "loss": 0.8497,
+      "step": 12520
+    },
+    {
+      "epoch": 0.9946615332711505,
+      "grad_norm": 0.8037011027336121,
+      "learning_rate": 3.342462491069303e-05,
+      "loss": 0.8659,
+      "step": 12530
+    },
+    {
+      "epoch": 0.99545535732005,
+      "grad_norm": 0.8320547938346863,
+      "learning_rate": 3.341139424730756e-05,
+      "loss": 0.8856,
+      "step": 12540
+    },
+    {
+      "epoch": 0.9962491813689496,
+      "grad_norm": 0.7249430418014526,
+      "learning_rate": 3.33981635839221e-05,
+      "loss": 0.8472,
+      "step": 12550
+    },
+    {
+      "epoch": 0.9970430054178492,
+      "grad_norm": 0.7746173143386841,
+      "learning_rate": 3.338493292053664e-05,
+      "loss": 0.8397,
+      "step": 12560
+    },
+    {
+      "epoch": 0.9978368294667487,
+      "grad_norm": 0.7750112414360046,
+      "learning_rate": 3.3371702257151176e-05,
+      "loss": 0.8164,
+      "step": 12570
+    },
+    {
+      "epoch": 0.9986306535156483,
+      "grad_norm": 0.9370802044868469,
+      "learning_rate": 3.335847159376571e-05,
+      "loss": 0.9081,
+      "step": 12580
+    },
+    {
+      "epoch": 0.9994244775645478,
+      "grad_norm": 0.7398301362991333,
+      "learning_rate": 3.334524093038025e-05,
+      "loss": 0.8786,
+      "step": 12590
+    },
+    {
+      "epoch": 1.0002183016134474,
+      "grad_norm": 0.811583399772644,
+      "learning_rate": 3.333201026699479e-05,
+      "loss": 0.7198,
+      "step": 12600
+    },
+    {
+      "epoch": 1.001012125662347,
+      "grad_norm": 0.8118659853935242,
+      "learning_rate": 3.331877960360933e-05,
+      "loss": 0.8307,
+      "step": 12610
+    },
+    {
+      "epoch": 1.0018059497112466,
+      "grad_norm": 0.6908761858940125,
+      "learning_rate": 3.3305548940223864e-05,
+      "loss": 0.8154,
+      "step": 12620
+    },
+    {
+      "epoch": 1.002599773760146,
+      "grad_norm": 0.5952050685882568,
+      "learning_rate": 3.32923182768384e-05,
+      "loss": 0.8398,
+      "step": 12630
+    },
+    {
+      "epoch": 1.0033935978090456,
+      "grad_norm": 0.7398274540901184,
+      "learning_rate": 3.327908761345294e-05,
+      "loss": 0.8172,
+      "step": 12640
+    },
+    {
+      "epoch": 1.0041874218579452,
+      "grad_norm": 0.6245648860931396,
+      "learning_rate": 3.3265856950067474e-05,
+      "loss": 0.8967,
+      "step": 12650
+    },
+    {
+      "epoch": 1.0049812459068448,
+      "grad_norm": 0.7542557716369629,
+      "learning_rate": 3.325262628668201e-05,
+      "loss": 0.763,
+      "step": 12660
+    },
+    {
+      "epoch": 1.0057750699557444,
+      "grad_norm": 0.6645131707191467,
+      "learning_rate": 3.323939562329655e-05,
+      "loss": 0.8822,
+      "step": 12670
+    },
+    {
+      "epoch": 1.0065688940046438,
+      "grad_norm": 0.8816360235214233,
+      "learning_rate": 3.322616495991109e-05,
+      "loss": 0.879,
+      "step": 12680
+    },
+    {
+      "epoch": 1.0073627180535434,
+      "grad_norm": 0.8549253344535828,
+      "learning_rate": 3.321293429652563e-05,
+      "loss": 0.836,
+      "step": 12690
+    },
+    {
+      "epoch": 1.008156542102443,
+      "grad_norm": 0.95768141746521,
+      "learning_rate": 3.319970363314017e-05,
+      "loss": 0.842,
+      "step": 12700
+    },
+    {
+      "epoch": 1.0089503661513426,
+      "grad_norm": 0.8039479851722717,
+      "learning_rate": 3.318647296975471e-05,
+      "loss": 0.7732,
+      "step": 12710
+    },
+    {
+      "epoch": 1.0097441902002422,
+      "grad_norm": 0.724469006061554,
+      "learning_rate": 3.3173242306369246e-05,
+      "loss": 0.797,
+      "step": 12720
+    },
+    {
+      "epoch": 1.0105380142491416,
+      "grad_norm": 0.7534793019294739,
+      "learning_rate": 3.316001164298378e-05,
+      "loss": 0.8058,
+      "step": 12730
+    },
+    {
+      "epoch": 1.0113318382980412,
+      "grad_norm": 0.9147580862045288,
+      "learning_rate": 3.314678097959832e-05,
+      "loss": 0.819,
+      "step": 12740
+    },
+    {
+      "epoch": 1.0121256623469408,
+      "grad_norm": 0.8486768007278442,
+      "learning_rate": 3.3133550316212856e-05,
+      "loss": 0.7997,
+      "step": 12750
+    },
+    {
+      "epoch": 1.0129194863958404,
+      "grad_norm": 0.7328081130981445,
+      "learning_rate": 3.312031965282739e-05,
+      "loss": 0.8433,
+      "step": 12760
+    },
+    {
+      "epoch": 1.01371331044474,
+      "grad_norm": 0.9267632961273193,
+      "learning_rate": 3.3107088989441934e-05,
+      "loss": 0.8013,
+      "step": 12770
+    },
+    {
+      "epoch": 1.0145071344936394,
+      "grad_norm": 0.7737709879875183,
+      "learning_rate": 3.309385832605647e-05,
+      "loss": 0.8186,
+      "step": 12780
+    },
+    {
+      "epoch": 1.015300958542539,
+      "grad_norm": 0.6647375822067261,
+      "learning_rate": 3.308062766267101e-05,
+      "loss": 0.8904,
+      "step": 12790
+    },
+    {
+      "epoch": 1.0160947825914386,
+      "grad_norm": 0.7600811123847961,
+      "learning_rate": 3.3067396999285544e-05,
+      "loss": 0.8189,
+      "step": 12800
+    },
+    {
+      "epoch": 1.0168886066403382,
+      "grad_norm": 0.7824926972389221,
+      "learning_rate": 3.305416633590008e-05,
+      "loss": 0.8189,
+      "step": 12810
+    },
+    {
+      "epoch": 1.0176824306892378,
+      "grad_norm": 0.5769202709197998,
+      "learning_rate": 3.304093567251462e-05,
+      "loss": 0.8361,
+      "step": 12820
+    },
+    {
+      "epoch": 1.0184762547381372,
+      "grad_norm": 0.8311686515808105,
+      "learning_rate": 3.302770500912916e-05,
+      "loss": 0.8081,
+      "step": 12830
+    },
+    {
+      "epoch": 1.0192700787870368,
+      "grad_norm": 0.7909318804740906,
+      "learning_rate": 3.301447434574369e-05,
+      "loss": 0.7855,
+      "step": 12840
+    },
+    {
+      "epoch": 1.0200639028359364,
+      "grad_norm": 0.8774951100349426,
+      "learning_rate": 3.300124368235823e-05,
+      "loss": 0.8729,
+      "step": 12850
+    },
+    {
+      "epoch": 1.020857726884836,
+      "grad_norm": 0.747606635093689,
+      "learning_rate": 3.298801301897277e-05,
+      "loss": 0.8321,
+      "step": 12860
+    },
+    {
+      "epoch": 1.0216515509337356,
+      "grad_norm": 0.8213605880737305,
+      "learning_rate": 3.297478235558731e-05,
+      "loss": 0.8039,
+      "step": 12870
+    },
+    {
+      "epoch": 1.022445374982635,
+      "grad_norm": 0.7250701189041138,
+      "learning_rate": 3.296155169220185e-05,
+      "loss": 0.7972,
+      "step": 12880
+    },
+    {
+      "epoch": 1.0232391990315346,
+      "grad_norm": 0.7011380195617676,
+      "learning_rate": 3.294832102881639e-05,
+      "loss": 0.8402,
+      "step": 12890
+    },
+    {
+      "epoch": 1.0240330230804342,
+      "grad_norm": 0.9204203486442566,
+      "learning_rate": 3.293509036543093e-05,
+      "loss": 0.8623,
+      "step": 12900
+    },
+    {
+      "epoch": 1.0248268471293338,
+      "grad_norm": 0.7364963293075562,
+      "learning_rate": 3.292185970204546e-05,
+      "loss": 0.8065,
+      "step": 12910
+    },
+    {
+      "epoch": 1.0256206711782334,
+      "grad_norm": 0.8210947513580322,
+      "learning_rate": 3.290862903866e-05,
+      "loss": 0.8116,
+      "step": 12920
+    },
+    {
+      "epoch": 1.026414495227133,
+      "grad_norm": 0.8079206347465515,
+      "learning_rate": 3.289539837527454e-05,
+      "loss": 0.8278,
+      "step": 12930
+    },
+    {
+      "epoch": 1.0272083192760324,
+      "grad_norm": 0.8758916258811951,
+      "learning_rate": 3.2882167711889076e-05,
+      "loss": 0.8202,
+      "step": 12940
+    },
+    {
+      "epoch": 1.028002143324932,
+      "grad_norm": 0.8359827399253845,
+      "learning_rate": 3.2868937048503615e-05,
+      "loss": 0.7738,
+      "step": 12950
+    },
+    {
+      "epoch": 1.0287959673738316,
+      "grad_norm": 0.7513866424560547,
+      "learning_rate": 3.2855706385118154e-05,
+      "loss": 0.8644,
+      "step": 12960
+    },
+    {
+      "epoch": 1.0295897914227312,
+      "grad_norm": 0.8743990063667297,
+      "learning_rate": 3.284247572173269e-05,
+      "loss": 0.8927,
+      "step": 12970
+    },
+    {
+      "epoch": 1.0303836154716308,
+      "grad_norm": 0.8003587126731873,
+      "learning_rate": 3.2829245058347225e-05,
+      "loss": 0.8206,
+      "step": 12980
+    },
+    {
+      "epoch": 1.0311774395205302,
+      "grad_norm": 0.7568554878234863,
+      "learning_rate": 3.2816014394961764e-05,
+      "loss": 0.8321,
+      "step": 12990
+    },
+    {
+      "epoch": 1.0319712635694298,
+      "grad_norm": 0.7184220552444458,
+      "learning_rate": 3.28027837315763e-05,
+      "loss": 0.8007,
+      "step": 13000
+    },
+    {
+      "epoch": 1.0327650876183294,
+      "grad_norm": 0.7958794236183167,
+      "learning_rate": 3.278955306819084e-05,
+      "loss": 0.886,
+      "step": 13010
+    },
+    {
+      "epoch": 1.033558911667229,
+      "grad_norm": 0.8230629563331604,
+      "learning_rate": 3.2776322404805374e-05,
+      "loss": 0.8296,
+      "step": 13020
+    },
+    {
+      "epoch": 1.0343527357161286,
+      "grad_norm": 0.8877654671669006,
+      "learning_rate": 3.276309174141991e-05,
+      "loss": 0.8461,
+      "step": 13030
+    },
+    {
+      "epoch": 1.035146559765028,
+      "grad_norm": 0.7817438244819641,
+      "learning_rate": 3.274986107803446e-05,
+      "loss": 0.8301,
+      "step": 13040
+    },
+    {
+      "epoch": 1.0359403838139276,
+      "grad_norm": 0.9634678959846497,
+      "learning_rate": 3.2736630414649e-05,
+      "loss": 0.8447,
+      "step": 13050
+    },
+    {
+      "epoch": 1.0367342078628272,
+      "grad_norm": 0.8039039373397827,
+      "learning_rate": 3.272339975126353e-05,
+      "loss": 0.8416,
+      "step": 13060
+    },
+    {
+      "epoch": 1.0375280319117268,
+      "grad_norm": 0.813490092754364,
+      "learning_rate": 3.271016908787807e-05,
+      "loss": 0.8281,
+      "step": 13070
+    },
+    {
+      "epoch": 1.0383218559606264,
+      "grad_norm": 0.8581125736236572,
+      "learning_rate": 3.269693842449261e-05,
+      "loss": 0.8527,
+      "step": 13080
+    },
+    {
+      "epoch": 1.0391156800095258,
+      "grad_norm": 0.7691791653633118,
+      "learning_rate": 3.268370776110714e-05,
+      "loss": 0.8385,
+      "step": 13090
+    },
+    {
+      "epoch": 1.0399095040584254,
+      "grad_norm": 0.788347601890564,
+      "learning_rate": 3.267047709772168e-05,
+      "loss": 0.8372,
+      "step": 13100
+    },
+    {
+      "epoch": 1.040703328107325,
+      "grad_norm": 0.8846760392189026,
+      "learning_rate": 3.265724643433622e-05,
+      "loss": 0.8794,
+      "step": 13110
+    },
+    {
+      "epoch": 1.0414971521562246,
+      "grad_norm": 0.8345226645469666,
+      "learning_rate": 3.2644015770950756e-05,
+      "loss": 0.848,
+      "step": 13120
+    },
+    {
+      "epoch": 1.0422909762051242,
+      "grad_norm": 0.7519625425338745,
+      "learning_rate": 3.2630785107565295e-05,
+      "loss": 0.8409,
+      "step": 13130
+    },
+    {
+      "epoch": 1.0430848002540236,
+      "grad_norm": 0.7269890904426575,
+      "learning_rate": 3.2617554444179834e-05,
+      "loss": 0.8638,
+      "step": 13140
+    },
+    {
+      "epoch": 1.0438786243029232,
+      "grad_norm": 0.8326950669288635,
+      "learning_rate": 3.260432378079437e-05,
+      "loss": 0.7581,
+      "step": 13150
+    },
+    {
+      "epoch": 1.0446724483518228,
+      "grad_norm": 0.8327310681343079,
+      "learning_rate": 3.259109311740891e-05,
+      "loss": 0.834,
+      "step": 13160
+    },
+    {
+      "epoch": 1.0454662724007224,
+      "grad_norm": 0.9904863238334656,
+      "learning_rate": 3.2577862454023444e-05,
+      "loss": 0.7524,
+      "step": 13170
+    },
+    {
+      "epoch": 1.046260096449622,
+      "grad_norm": 0.9039106965065002,
+      "learning_rate": 3.256463179063798e-05,
+      "loss": 0.7784,
+      "step": 13180
+    },
+    {
+      "epoch": 1.0470539204985214,
+      "grad_norm": 0.6900113821029663,
+      "learning_rate": 3.255140112725252e-05,
+      "loss": 0.8065,
+      "step": 13190
+    },
+    {
+      "epoch": 1.047847744547421,
+      "grad_norm": 0.8624986410140991,
+      "learning_rate": 3.253817046386706e-05,
+      "loss": 0.8827,
+      "step": 13200
+    },
+    {
+      "epoch": 1.0486415685963206,
+      "grad_norm": 0.7626116275787354,
+      "learning_rate": 3.25249398004816e-05,
+      "loss": 0.8464,
+      "step": 13210
+    },
+    {
+      "epoch": 1.0494353926452202,
+      "grad_norm": 0.7918316721916199,
+      "learning_rate": 3.251170913709614e-05,
+      "loss": 0.8258,
+      "step": 13220
+    },
+    {
+      "epoch": 1.0502292166941198,
+      "grad_norm": 0.7877132892608643,
+      "learning_rate": 3.249847847371068e-05,
+      "loss": 0.8688,
+      "step": 13230
+    },
+    {
+      "epoch": 1.0510230407430192,
+      "grad_norm": 0.7505654096603394,
+      "learning_rate": 3.248524781032521e-05,
+      "loss": 0.8173,
+      "step": 13240
+    },
+    {
+      "epoch": 1.0518168647919188,
+      "grad_norm": 0.8359394073486328,
+      "learning_rate": 3.247201714693975e-05,
+      "loss": 0.8227,
+      "step": 13250
+    },
+    {
+      "epoch": 1.0526106888408184,
+      "grad_norm": 0.9003983736038208,
+      "learning_rate": 3.245878648355429e-05,
+      "loss": 0.7917,
+      "step": 13260
+    },
+    {
+      "epoch": 1.053404512889718,
+      "grad_norm": 0.7913084030151367,
+      "learning_rate": 3.2445555820168826e-05,
+      "loss": 0.8602,
+      "step": 13270
+    },
+    {
+      "epoch": 1.0541983369386176,
+      "grad_norm": 0.7529349327087402,
+      "learning_rate": 3.243232515678336e-05,
+      "loss": 0.7734,
+      "step": 13280
+    },
+    {
+      "epoch": 1.054992160987517,
+      "grad_norm": 0.7655823230743408,
+      "learning_rate": 3.24190944933979e-05,
+      "loss": 0.8708,
+      "step": 13290
+    },
+    {
+      "epoch": 1.0557859850364166,
+      "grad_norm": 0.7675358057022095,
+      "learning_rate": 3.2405863830012436e-05,
+      "loss": 0.7815,
+      "step": 13300
+    },
+    {
+      "epoch": 1.0565798090853162,
+      "grad_norm": 0.8172122240066528,
+      "learning_rate": 3.239263316662698e-05,
+      "loss": 0.8038,
+      "step": 13310
+    },
+    {
+      "epoch": 1.0573736331342158,
+      "grad_norm": 0.8098911643028259,
+      "learning_rate": 3.2379402503241514e-05,
+      "loss": 0.8107,
+      "step": 13320
+    },
+    {
+      "epoch": 1.0581674571831154,
+      "grad_norm": 0.7456363439559937,
+      "learning_rate": 3.236617183985605e-05,
+      "loss": 0.8484,
+      "step": 13330
+    },
+    {
+      "epoch": 1.0589612812320148,
+      "grad_norm": 0.8800321221351624,
+      "learning_rate": 3.235294117647059e-05,
+      "loss": 0.7985,
+      "step": 13340
+    },
+    {
+      "epoch": 1.0597551052809144,
+      "grad_norm": 0.7727877497673035,
+      "learning_rate": 3.2339710513085124e-05,
+      "loss": 0.8827,
+      "step": 13350
+    },
+    {
+      "epoch": 1.060548929329814,
+      "grad_norm": 0.8835548162460327,
+      "learning_rate": 3.232647984969966e-05,
+      "loss": 0.7698,
+      "step": 13360
+    },
+    {
+      "epoch": 1.0613427533787136,
+      "grad_norm": 0.7106403708457947,
+      "learning_rate": 3.23132491863142e-05,
+      "loss": 0.863,
+      "step": 13370
+    },
+    {
+      "epoch": 1.0621365774276132,
+      "grad_norm": 0.6507337689399719,
+      "learning_rate": 3.230001852292874e-05,
+      "loss": 0.7971,
+      "step": 13380
+    },
+    {
+      "epoch": 1.0629304014765126,
+      "grad_norm": 0.7235475778579712,
+      "learning_rate": 3.228678785954328e-05,
+      "loss": 0.7664,
+      "step": 13390
+    },
+    {
+      "epoch": 1.0637242255254122,
+      "grad_norm": 0.7101955413818359,
+      "learning_rate": 3.227355719615782e-05,
+      "loss": 0.7829,
+      "step": 13400
+    },
+    {
+      "epoch": 1.0645180495743118,
+      "grad_norm": 0.7536014318466187,
+      "learning_rate": 3.226032653277236e-05,
+      "loss": 0.8732,
+      "step": 13410
+    },
+    {
+      "epoch": 1.0653118736232114,
+      "grad_norm": 0.7618986368179321,
+      "learning_rate": 3.22470958693869e-05,
+      "loss": 0.8343,
+      "step": 13420
+    },
+    {
+      "epoch": 1.066105697672111,
+      "grad_norm": 0.8681114315986633,
+      "learning_rate": 3.223386520600143e-05,
+      "loss": 0.8696,
+      "step": 13430
+    },
+    {
+      "epoch": 1.0668995217210104,
+      "grad_norm": 0.9769225120544434,
+      "learning_rate": 3.222063454261597e-05,
+      "loss": 0.8216,
+      "step": 13440
+    },
+    {
+      "epoch": 1.06769334576991,
+      "grad_norm": 0.775951087474823,
+      "learning_rate": 3.220872694556905e-05,
+      "loss": 0.8409,
+      "step": 13450
+    },
+    {
+      "epoch": 1.0684871698188096,
+      "grad_norm": 0.7240128517150879,
+      "learning_rate": 3.219549628218359e-05,
+      "loss": 0.881,
+      "step": 13460
+    },
+    {
+      "epoch": 1.0692809938677093,
+      "grad_norm": 0.7799546122550964,
+      "learning_rate": 3.218226561879813e-05,
+      "loss": 0.8269,
+      "step": 13470
+    },
+    {
+      "epoch": 1.0700748179166089,
+      "grad_norm": 0.7071607112884521,
+      "learning_rate": 3.2169034955412666e-05,
+      "loss": 0.8752,
+      "step": 13480
+    },
+    {
+      "epoch": 1.0708686419655082,
+      "grad_norm": 0.8684601783752441,
+      "learning_rate": 3.2155804292027205e-05,
+      "loss": 0.8156,
+      "step": 13490
+    },
+    {
+      "epoch": 1.0716624660144078,
+      "grad_norm": 0.7755899429321289,
+      "learning_rate": 3.2142573628641743e-05,
+      "loss": 0.8075,
+      "step": 13500
+    },
+    {
+      "epoch": 1.0724562900633074,
+      "grad_norm": 0.7392835021018982,
+      "learning_rate": 3.212934296525628e-05,
+      "loss": 0.8711,
+      "step": 13510
+    },
+    {
+      "epoch": 1.073250114112207,
+      "grad_norm": 0.8979812860488892,
+      "learning_rate": 3.2116112301870815e-05,
+      "loss": 0.8296,
+      "step": 13520
+    },
+    {
+      "epoch": 1.0740439381611067,
+      "grad_norm": 0.9251773357391357,
+      "learning_rate": 3.2102881638485354e-05,
+      "loss": 0.8392,
+      "step": 13530
+    },
+    {
+      "epoch": 1.0748377622100063,
+      "grad_norm": 0.6050375699996948,
+      "learning_rate": 3.208965097509989e-05,
+      "loss": 0.8432,
+      "step": 13540
+    },
+    {
+      "epoch": 1.0756315862589056,
+      "grad_norm": 0.8619400858879089,
+      "learning_rate": 3.207642031171443e-05,
+      "loss": 0.8651,
+      "step": 13550
+    },
+    {
+      "epoch": 1.0764254103078053,
+      "grad_norm": 0.9240981340408325,
+      "learning_rate": 3.206318964832897e-05,
+      "loss": 0.8467,
+      "step": 13560
+    },
+    {
+      "epoch": 1.0772192343567049,
+      "grad_norm": 0.9178832173347473,
+      "learning_rate": 3.204995898494351e-05,
+      "loss": 0.8079,
+      "step": 13570
+    },
+    {
+      "epoch": 1.0780130584056045,
+      "grad_norm": 0.7604396343231201,
+      "learning_rate": 3.203672832155805e-05,
+      "loss": 0.7611,
+      "step": 13580
+    },
+    {
+      "epoch": 1.0788068824545038,
+      "grad_norm": 0.5721964836120605,
+      "learning_rate": 3.202349765817258e-05,
+      "loss": 0.7863,
+      "step": 13590
+    },
+    {
+      "epoch": 1.0796007065034035,
+      "grad_norm": 0.5710390210151672,
+      "learning_rate": 3.201026699478712e-05,
+      "loss": 0.873,
+      "step": 13600
+    },
+    {
+      "epoch": 1.080394530552303,
+      "grad_norm": 0.72386234998703,
+      "learning_rate": 3.199703633140166e-05,
+      "loss": 0.866,
+      "step": 13610
+    },
+    {
+      "epoch": 1.0811883546012027,
+      "grad_norm": 0.6764307618141174,
+      "learning_rate": 3.19838056680162e-05,
+      "loss": 0.8335,
+      "step": 13620
+    },
+    {
+      "epoch": 1.0819821786501023,
+      "grad_norm": 0.796247124671936,
+      "learning_rate": 3.197057500463073e-05,
+      "loss": 0.8253,
+      "step": 13630
+    },
+    {
+      "epoch": 1.0827760026990019,
+      "grad_norm": 0.8703547120094299,
+      "learning_rate": 3.195734434124527e-05,
+      "loss": 0.8459,
+      "step": 13640
+    },
+    {
+      "epoch": 1.0835698267479013,
+      "grad_norm": 0.6727367639541626,
+      "learning_rate": 3.194411367785981e-05,
+      "loss": 0.8283,
+      "step": 13650
+    },
+    {
+      "epoch": 1.0843636507968009,
+      "grad_norm": 0.8048222064971924,
+      "learning_rate": 3.193088301447435e-05,
+      "loss": 0.8383,
+      "step": 13660
+    },
+    {
+      "epoch": 1.0851574748457005,
+      "grad_norm": 0.8705745339393616,
+      "learning_rate": 3.1917652351088885e-05,
+      "loss": 0.8146,
+      "step": 13670
+    },
+    {
+      "epoch": 1.0859512988946,
+      "grad_norm": 0.7636762261390686,
+      "learning_rate": 3.1904421687703424e-05,
+      "loss": 0.9398,
+      "step": 13680
+    },
+    {
+      "epoch": 1.0867451229434997,
+      "grad_norm": 0.806253969669342,
+      "learning_rate": 3.189119102431796e-05,
+      "loss": 0.8356,
+      "step": 13690
+    },
+    {
+      "epoch": 1.087538946992399,
+      "grad_norm": 0.8001126646995544,
+      "learning_rate": 3.1877960360932495e-05,
+      "loss": 0.8216,
+      "step": 13700
+    },
+    {
+      "epoch": 1.0883327710412987,
+      "grad_norm": 0.8212706446647644,
+      "learning_rate": 3.1864729697547034e-05,
+      "loss": 0.8063,
+      "step": 13710
+    },
+    {
+      "epoch": 1.0891265950901983,
+      "grad_norm": 0.8403292894363403,
+      "learning_rate": 3.185149903416157e-05,
+      "loss": 0.7797,
+      "step": 13720
+    },
+    {
+      "epoch": 1.0899204191390979,
+      "grad_norm": 0.876654326915741,
+      "learning_rate": 3.183826837077611e-05,
+      "loss": 0.8569,
+      "step": 13730
+    },
+    {
+      "epoch": 1.0907142431879975,
+      "grad_norm": 0.69620680809021,
+      "learning_rate": 3.182503770739065e-05,
+      "loss": 0.8604,
+      "step": 13740
+    },
+    {
+      "epoch": 1.0915080672368969,
+      "grad_norm": 0.6936290860176086,
+      "learning_rate": 3.181180704400519e-05,
+      "loss": 0.8722,
+      "step": 13750
+    },
+    {
+      "epoch": 1.0923018912857965,
+      "grad_norm": 0.848301112651825,
+      "learning_rate": 3.179857638061973e-05,
+      "loss": 0.8742,
+      "step": 13760
+    },
+    {
+      "epoch": 1.093095715334696,
+      "grad_norm": 0.9208377003669739,
+      "learning_rate": 3.178534571723427e-05,
+      "loss": 0.8333,
+      "step": 13770
+    },
+    {
+      "epoch": 1.0938895393835957,
+      "grad_norm": 0.8173259496688843,
+      "learning_rate": 3.17721150538488e-05,
+      "loss": 0.8459,
+      "step": 13780
+    },
+    {
+      "epoch": 1.0946833634324953,
+      "grad_norm": 0.8618404865264893,
+      "learning_rate": 3.175888439046334e-05,
+      "loss": 0.8343,
+      "step": 13790
+    },
+    {
+      "epoch": 1.0954771874813947,
+      "grad_norm": 0.7455660104751587,
+      "learning_rate": 3.174565372707788e-05,
+      "loss": 0.7877,
+      "step": 13800
+    },
+    {
+      "epoch": 1.0962710115302943,
+      "grad_norm": 0.7906826734542847,
+      "learning_rate": 3.173242306369241e-05,
+      "loss": 0.8711,
+      "step": 13810
+    },
+    {
+      "epoch": 1.0970648355791939,
+      "grad_norm": 0.8172614574432373,
+      "learning_rate": 3.171919240030695e-05,
+      "loss": 0.8373,
+      "step": 13820
+    },
+    {
+      "epoch": 1.0978586596280935,
+      "grad_norm": 0.7989473938941956,
+      "learning_rate": 3.1705961736921494e-05,
+      "loss": 0.8162,
+      "step": 13830
+    },
+    {
+      "epoch": 1.098652483676993,
+      "grad_norm": 0.8977952003479004,
+      "learning_rate": 3.169273107353603e-05,
+      "loss": 0.8057,
+      "step": 13840
+    },
+    {
+      "epoch": 1.0994463077258925,
+      "grad_norm": 0.6931014060974121,
+      "learning_rate": 3.1679500410150565e-05,
+      "loss": 0.8232,
+      "step": 13850
+    },
+    {
+      "epoch": 1.100240131774792,
+      "grad_norm": 0.8068207502365112,
+      "learning_rate": 3.1666269746765104e-05,
+      "loss": 0.8804,
+      "step": 13860
+    },
+    {
+      "epoch": 1.1010339558236917,
+      "grad_norm": 0.7539369463920593,
+      "learning_rate": 3.165303908337964e-05,
+      "loss": 0.8351,
+      "step": 13870
+    },
+    {
+      "epoch": 1.1018277798725913,
+      "grad_norm": 0.7581447958946228,
+      "learning_rate": 3.163980841999418e-05,
+      "loss": 0.846,
+      "step": 13880
+    },
+    {
+      "epoch": 1.102621603921491,
+      "grad_norm": 0.9124611616134644,
+      "learning_rate": 3.1626577756608714e-05,
+      "loss": 0.8507,
+      "step": 13890
+    },
+    {
+      "epoch": 1.1034154279703903,
+      "grad_norm": 0.7561845779418945,
+      "learning_rate": 3.161334709322325e-05,
+      "loss": 0.7908,
+      "step": 13900
+    },
+    {
+      "epoch": 1.1042092520192899,
+      "grad_norm": 0.718763530254364,
+      "learning_rate": 3.160011642983779e-05,
+      "loss": 0.832,
+      "step": 13910
+    },
+    {
+      "epoch": 1.1050030760681895,
+      "grad_norm": 0.873539924621582,
+      "learning_rate": 3.158688576645233e-05,
+      "loss": 0.7854,
+      "step": 13920
+    },
+    {
+      "epoch": 1.105796900117089,
+      "grad_norm": 0.8948822021484375,
+      "learning_rate": 3.157365510306687e-05,
+      "loss": 0.7936,
+      "step": 13930
+    },
+    {
+      "epoch": 1.1065907241659887,
+      "grad_norm": 0.7010993957519531,
+      "learning_rate": 3.156042443968141e-05,
+      "loss": 0.7964,
+      "step": 13940
+    },
+    {
+      "epoch": 1.107384548214888,
+      "grad_norm": 0.8644427061080933,
+      "learning_rate": 3.154719377629595e-05,
+      "loss": 0.8249,
+      "step": 13950
+    },
+    {
+      "epoch": 1.1081783722637877,
+      "grad_norm": 0.8382445573806763,
+      "learning_rate": 3.153396311291048e-05,
+      "loss": 0.8446,
+      "step": 13960
+    },
+    {
+      "epoch": 1.1089721963126873,
+      "grad_norm": 0.9029629230499268,
+      "learning_rate": 3.152073244952502e-05,
+      "loss": 0.8784,
+      "step": 13970
+    },
+    {
+      "epoch": 1.109766020361587,
+      "grad_norm": 0.8613706827163696,
+      "learning_rate": 3.150750178613956e-05,
+      "loss": 0.7617,
+      "step": 13980
+    },
+    {
+      "epoch": 1.1105598444104865,
+      "grad_norm": 1.0319517850875854,
+      "learning_rate": 3.14942711227541e-05,
+      "loss": 0.8107,
+      "step": 13990
+    },
+    {
+      "epoch": 1.1113536684593859,
+      "grad_norm": 0.8257167339324951,
+      "learning_rate": 3.1481040459368636e-05,
+      "loss": 0.7137,
+      "step": 14000
+    },
+    {
+      "epoch": 1.1121474925082855,
+      "grad_norm": 0.8850013017654419,
+      "learning_rate": 3.1467809795983175e-05,
+      "loss": 0.8005,
+      "step": 14010
+    },
+    {
+      "epoch": 1.112941316557185,
+      "grad_norm": 0.7616755366325378,
+      "learning_rate": 3.1454579132597713e-05,
+      "loss": 0.7917,
+      "step": 14020
+    },
+    {
+      "epoch": 1.1137351406060847,
+      "grad_norm": 0.891237735748291,
+      "learning_rate": 3.1441348469212246e-05,
+      "loss": 0.7803,
+      "step": 14030
+    },
+    {
+      "epoch": 1.1145289646549843,
+      "grad_norm": 0.7421796917915344,
+      "learning_rate": 3.1428117805826785e-05,
+      "loss": 0.8219,
+      "step": 14040
+    },
+    {
+      "epoch": 1.1153227887038837,
+      "grad_norm": 0.8168928027153015,
+      "learning_rate": 3.1414887142441324e-05,
+      "loss": 0.7825,
+      "step": 14050
+    },
+    {
+      "epoch": 1.1161166127527833,
+      "grad_norm": 0.7771959900856018,
+      "learning_rate": 3.140165647905586e-05,
+      "loss": 0.8165,
+      "step": 14060
+    },
+    {
+      "epoch": 1.116910436801683,
+      "grad_norm": 0.7545336484909058,
+      "learning_rate": 3.1388425815670395e-05,
+      "loss": 0.9121,
+      "step": 14070
+    },
+    {
+      "epoch": 1.1177042608505825,
+      "grad_norm": 0.626607358455658,
+      "learning_rate": 3.1375195152284934e-05,
+      "loss": 0.8899,
+      "step": 14080
+    },
+    {
+      "epoch": 1.118498084899482,
+      "grad_norm": 0.8673813343048096,
+      "learning_rate": 3.136196448889947e-05,
+      "loss": 0.8339,
+      "step": 14090
+    },
+    {
+      "epoch": 1.1192919089483815,
+      "grad_norm": 0.8372191786766052,
+      "learning_rate": 3.134873382551402e-05,
+      "loss": 0.7954,
+      "step": 14100
+    },
+    {
+      "epoch": 1.120085732997281,
+      "grad_norm": 0.9092103838920593,
+      "learning_rate": 3.133550316212855e-05,
+      "loss": 0.8326,
+      "step": 14110
+    },
+    {
+      "epoch": 1.1208795570461807,
+      "grad_norm": 0.7805484533309937,
+      "learning_rate": 3.132227249874309e-05,
+      "loss": 0.799,
+      "step": 14120
+    },
+    {
+      "epoch": 1.1216733810950803,
+      "grad_norm": 0.8048092126846313,
+      "learning_rate": 3.130904183535763e-05,
+      "loss": 0.8497,
+      "step": 14130
+    },
+    {
+      "epoch": 1.12246720514398,
+      "grad_norm": 0.922804594039917,
+      "learning_rate": 3.129581117197217e-05,
+      "loss": 0.7648,
+      "step": 14140
+    },
+    {
+      "epoch": 1.1232610291928795,
+      "grad_norm": 0.7149479985237122,
+      "learning_rate": 3.12825805085867e-05,
+      "loss": 0.8884,
+      "step": 14150
+    },
+    {
+      "epoch": 1.124054853241779,
+      "grad_norm": 0.6718799471855164,
+      "learning_rate": 3.126934984520124e-05,
+      "loss": 0.7914,
+      "step": 14160
+    },
+    {
+      "epoch": 1.1248486772906785,
+      "grad_norm": 0.6949440836906433,
+      "learning_rate": 3.125611918181578e-05,
+      "loss": 0.8319,
+      "step": 14170
+    },
+    {
+      "epoch": 1.125642501339578,
+      "grad_norm": 0.8410897254943848,
+      "learning_rate": 3.1242888518430316e-05,
+      "loss": 0.8594,
+      "step": 14180
+    },
+    {
+      "epoch": 1.1264363253884777,
+      "grad_norm": 0.6044743657112122,
+      "learning_rate": 3.1229657855044855e-05,
+      "loss": 0.8168,
+      "step": 14190
+    },
+    {
+      "epoch": 1.127230149437377,
+      "grad_norm": 1.078133225440979,
+      "learning_rate": 3.1216427191659394e-05,
+      "loss": 0.834,
+      "step": 14200
+    },
+    {
+      "epoch": 1.1280239734862767,
+      "grad_norm": 0.9174512624740601,
+      "learning_rate": 3.120319652827393e-05,
+      "loss": 0.7889,
+      "step": 14210
+    },
+    {
+      "epoch": 1.1288177975351763,
+      "grad_norm": 0.5952799916267395,
+      "learning_rate": 3.1189965864888465e-05,
+      "loss": 0.79,
+      "step": 14220
+    },
+    {
+      "epoch": 1.129611621584076,
+      "grad_norm": 0.8574535250663757,
+      "learning_rate": 3.1176735201503004e-05,
+      "loss": 0.8347,
+      "step": 14230
+    },
+    {
+      "epoch": 1.1304054456329755,
+      "grad_norm": 0.882088303565979,
+      "learning_rate": 3.116350453811754e-05,
+      "loss": 0.8774,
+      "step": 14240
+    },
+    {
+      "epoch": 1.1311992696818751,
+      "grad_norm": 0.8594134449958801,
+      "learning_rate": 3.115027387473208e-05,
+      "loss": 0.7897,
+      "step": 14250
+    },
+    {
+      "epoch": 1.1319930937307745,
+      "grad_norm": 0.739986002445221,
+      "learning_rate": 3.1137043211346614e-05,
+      "loss": 0.8021,
+      "step": 14260
+    },
+    {
+      "epoch": 1.132786917779674,
+      "grad_norm": 0.8186190724372864,
+      "learning_rate": 3.112381254796116e-05,
+      "loss": 0.8212,
+      "step": 14270
+    },
+    {
+      "epoch": 1.1335807418285737,
+      "grad_norm": 0.5985064506530762,
+      "learning_rate": 3.11105818845757e-05,
+      "loss": 0.8579,
+      "step": 14280
+    },
+    {
+      "epoch": 1.1343745658774733,
+      "grad_norm": 0.7839154601097107,
+      "learning_rate": 3.109735122119023e-05,
+      "loss": 0.8212,
+      "step": 14290
+    },
+    {
+      "epoch": 1.1351683899263727,
+      "grad_norm": 0.8646635413169861,
+      "learning_rate": 3.108412055780477e-05,
+      "loss": 0.8486,
+      "step": 14300
+    },
+    {
+      "epoch": 1.1359622139752723,
+      "grad_norm": 0.6865241527557373,
+      "learning_rate": 3.107088989441931e-05,
+      "loss": 0.8799,
+      "step": 14310
+    },
+    {
+      "epoch": 1.136756038024172,
+      "grad_norm": 0.7246029376983643,
+      "learning_rate": 3.105765923103385e-05,
+      "loss": 0.8432,
+      "step": 14320
+    },
+    {
+      "epoch": 1.1375498620730715,
+      "grad_norm": 0.8030770421028137,
+      "learning_rate": 3.104442856764838e-05,
+      "loss": 0.7673,
+      "step": 14330
+    },
+    {
+      "epoch": 1.1383436861219711,
+      "grad_norm": 0.7397379279136658,
+      "learning_rate": 3.103119790426292e-05,
+      "loss": 0.8608,
+      "step": 14340
+    },
+    {
+      "epoch": 1.1391375101708707,
+      "grad_norm": 0.7867853045463562,
+      "learning_rate": 3.101796724087746e-05,
+      "loss": 0.8263,
+      "step": 14350
+    },
+    {
+      "epoch": 1.13993133421977,
+      "grad_norm": 0.6703535914421082,
+      "learning_rate": 3.1004736577491996e-05,
+      "loss": 0.8149,
+      "step": 14360
+    },
+    {
+      "epoch": 1.1407251582686697,
+      "grad_norm": 1.018248200416565,
+      "learning_rate": 3.0991505914106535e-05,
+      "loss": 0.8686,
+      "step": 14370
+    },
+    {
+      "epoch": 1.1415189823175693,
+      "grad_norm": 0.8166862726211548,
+      "learning_rate": 3.0978275250721074e-05,
+      "loss": 0.8326,
+      "step": 14380
+    },
+    {
+      "epoch": 1.142312806366469,
+      "grad_norm": 0.8093626499176025,
+      "learning_rate": 3.096504458733561e-05,
+      "loss": 0.7861,
+      "step": 14390
+    },
+    {
+      "epoch": 1.1431066304153685,
+      "grad_norm": 1.1167017221450806,
+      "learning_rate": 3.0951813923950145e-05,
+      "loss": 0.8208,
+      "step": 14400
+    },
+    {
+      "epoch": 1.143900454464268,
+      "grad_norm": 0.6292011141777039,
+      "learning_rate": 3.0938583260564684e-05,
+      "loss": 0.7954,
+      "step": 14410
+    },
+    {
+      "epoch": 1.1446942785131675,
+      "grad_norm": 0.7709428071975708,
+      "learning_rate": 3.092535259717922e-05,
+      "loss": 0.8027,
+      "step": 14420
+    },
+    {
+      "epoch": 1.1454881025620671,
+      "grad_norm": 0.8300706744194031,
+      "learning_rate": 3.091212193379376e-05,
+      "loss": 0.8297,
+      "step": 14430
+    },
+    {
+      "epoch": 1.1462819266109667,
+      "grad_norm": 0.7700504064559937,
+      "learning_rate": 3.08988912704083e-05,
+      "loss": 0.8367,
+      "step": 14440
+    },
+    {
+      "epoch": 1.1470757506598663,
+      "grad_norm": 1.0189820528030396,
+      "learning_rate": 3.088566060702284e-05,
+      "loss": 0.7969,
+      "step": 14450
+    },
+    {
+      "epoch": 1.1478695747087657,
+      "grad_norm": 0.7077481150627136,
+      "learning_rate": 3.087242994363738e-05,
+      "loss": 0.7619,
+      "step": 14460
+    },
+    {
+      "epoch": 1.1486633987576653,
+      "grad_norm": 0.8861839175224304,
+      "learning_rate": 3.085919928025192e-05,
+      "loss": 0.7641,
+      "step": 14470
+    },
+    {
+      "epoch": 1.149457222806565,
+      "grad_norm": 0.8999897837638855,
+      "learning_rate": 3.084596861686645e-05,
+      "loss": 0.7736,
+      "step": 14480
+    },
+    {
+      "epoch": 1.1502510468554645,
+      "grad_norm": 0.9881305694580078,
+      "learning_rate": 3.083273795348099e-05,
+      "loss": 0.8508,
+      "step": 14490
+    },
+    {
+      "epoch": 1.1510448709043641,
+      "grad_norm": 0.6451756358146667,
+      "learning_rate": 3.081950729009553e-05,
+      "loss": 0.7832,
+      "step": 14500
+    },
+    {
+      "epoch": 1.1518386949532635,
+      "grad_norm": 0.9463132619857788,
+      "learning_rate": 3.080627662671006e-05,
+      "loss": 0.8006,
+      "step": 14510
+    },
+    {
+      "epoch": 1.1526325190021631,
+      "grad_norm": 0.8050596117973328,
+      "learning_rate": 3.07930459633246e-05,
+      "loss": 0.8835,
+      "step": 14520
+    },
+    {
+      "epoch": 1.1534263430510627,
+      "grad_norm": 0.6816156506538391,
+      "learning_rate": 3.077981529993914e-05,
+      "loss": 0.8281,
+      "step": 14530
+    },
+    {
+      "epoch": 1.1542201670999623,
+      "grad_norm": 0.7849035263061523,
+      "learning_rate": 3.0766584636553683e-05,
+      "loss": 0.7944,
+      "step": 14540
+    },
+    {
+      "epoch": 1.155013991148862,
+      "grad_norm": 0.7930827140808105,
+      "learning_rate": 3.0753353973168216e-05,
+      "loss": 0.8804,
+      "step": 14550
+    },
+    {
+      "epoch": 1.1558078151977613,
+      "grad_norm": 0.8753893971443176,
+      "learning_rate": 3.0740123309782755e-05,
+      "loss": 0.8789,
+      "step": 14560
+    },
+    {
+      "epoch": 1.156601639246661,
+      "grad_norm": 0.974399983882904,
+      "learning_rate": 3.0726892646397293e-05,
+      "loss": 0.8217,
+      "step": 14570
+    },
+    {
+      "epoch": 1.1573954632955605,
+      "grad_norm": 0.7714707255363464,
+      "learning_rate": 3.071366198301183e-05,
+      "loss": 0.767,
+      "step": 14580
+    },
+    {
+      "epoch": 1.1581892873444601,
+      "grad_norm": 0.7159197926521301,
+      "learning_rate": 3.0700431319626365e-05,
+      "loss": 0.7676,
+      "step": 14590
+    },
+    {
+      "epoch": 1.1589831113933597,
+      "grad_norm": 0.6954260468482971,
+      "learning_rate": 3.0687200656240904e-05,
+      "loss": 0.8319,
+      "step": 14600
+    },
+    {
+      "epoch": 1.1597769354422591,
+      "grad_norm": 0.6127654910087585,
+      "learning_rate": 3.067396999285544e-05,
+      "loss": 0.8533,
+      "step": 14610
+    },
+    {
+      "epoch": 1.1605707594911587,
+      "grad_norm": 0.6652526259422302,
+      "learning_rate": 3.066073932946998e-05,
+      "loss": 0.8525,
+      "step": 14620
+    },
+    {
+      "epoch": 1.1613645835400583,
+      "grad_norm": 0.9769663214683533,
+      "learning_rate": 3.064750866608452e-05,
+      "loss": 0.8263,
+      "step": 14630
+    },
+    {
+      "epoch": 1.162158407588958,
+      "grad_norm": 0.8271260261535645,
+      "learning_rate": 3.063427800269906e-05,
+      "loss": 0.8327,
+      "step": 14640
+    },
+    {
+      "epoch": 1.1629522316378575,
+      "grad_norm": 0.7566442489624023,
+      "learning_rate": 3.06210473393136e-05,
+      "loss": 0.7574,
+      "step": 14650
+    },
+    {
+      "epoch": 1.1637460556867572,
+      "grad_norm": 0.7964077591896057,
+      "learning_rate": 3.060781667592813e-05,
+      "loss": 0.7874,
+      "step": 14660
+    },
+    {
+      "epoch": 1.1645398797356565,
+      "grad_norm": 0.7571009993553162,
+      "learning_rate": 3.059458601254267e-05,
+      "loss": 0.7855,
+      "step": 14670
+    },
+    {
+      "epoch": 1.1653337037845561,
+      "grad_norm": 0.8210294842720032,
+      "learning_rate": 3.058135534915721e-05,
+      "loss": 0.8045,
+      "step": 14680
+    },
+    {
+      "epoch": 1.1661275278334557,
+      "grad_norm": 0.6790111064910889,
+      "learning_rate": 3.056812468577175e-05,
+      "loss": 0.8583,
+      "step": 14690
+    },
+    {
+      "epoch": 1.1669213518823554,
+      "grad_norm": 0.8836639523506165,
+      "learning_rate": 3.055489402238628e-05,
+      "loss": 0.81,
+      "step": 14700
+    },
+    {
+      "epoch": 1.1677151759312547,
+      "grad_norm": 0.92630535364151,
+      "learning_rate": 3.0541663359000825e-05,
+      "loss": 0.8531,
+      "step": 14710
+    },
+    {
+      "epoch": 1.1685089999801543,
+      "grad_norm": 0.681800365447998,
+      "learning_rate": 3.0528432695615364e-05,
+      "loss": 0.817,
+      "step": 14720
+    },
+    {
+      "epoch": 1.169302824029054,
+      "grad_norm": 0.7862175107002258,
+      "learning_rate": 3.0515202032229896e-05,
+      "loss": 0.8026,
+      "step": 14730
+    },
+    {
+      "epoch": 1.1700966480779535,
+      "grad_norm": 0.7809861898422241,
+      "learning_rate": 3.0501971368844435e-05,
+      "loss": 0.8225,
+      "step": 14740
+    },
+    {
+      "epoch": 1.1708904721268532,
+      "grad_norm": 0.9233412146568298,
+      "learning_rate": 3.0488740705458974e-05,
+      "loss": 0.8066,
+      "step": 14750
+    },
+    {
+      "epoch": 1.1716842961757528,
+      "grad_norm": 0.8254870772361755,
+      "learning_rate": 3.0475510042073513e-05,
+      "loss": 0.8538,
+      "step": 14760
+    },
+    {
+      "epoch": 1.1724781202246521,
+      "grad_norm": 0.8267969489097595,
+      "learning_rate": 3.046227937868805e-05,
+      "loss": 0.8346,
+      "step": 14770
+    },
+    {
+      "epoch": 1.1732719442735517,
+      "grad_norm": 0.7673428058624268,
+      "learning_rate": 3.0449048715302587e-05,
+      "loss": 0.8336,
+      "step": 14780
+    },
+    {
+      "epoch": 1.1740657683224514,
+      "grad_norm": 0.8175097107887268,
+      "learning_rate": 3.0435818051917126e-05,
+      "loss": 0.8939,
+      "step": 14790
+    },
+    {
+      "epoch": 1.174859592371351,
+      "grad_norm": 0.8476851582527161,
+      "learning_rate": 3.0422587388531665e-05,
+      "loss": 0.828,
+      "step": 14800
+    },
+    {
+      "epoch": 1.1756534164202503,
+      "grad_norm": 0.8046305179595947,
+      "learning_rate": 3.0409356725146197e-05,
+      "loss": 0.8243,
+      "step": 14810
+    },
+    {
+      "epoch": 1.17644724046915,
+      "grad_norm": 0.8776468634605408,
+      "learning_rate": 3.039612606176074e-05,
+      "loss": 0.8292,
+      "step": 14820
+    },
+    {
+      "epoch": 1.1772410645180496,
+      "grad_norm": 0.7931563258171082,
+      "learning_rate": 3.038289539837528e-05,
+      "loss": 0.7834,
+      "step": 14830
+    },
+    {
+      "epoch": 1.1780348885669492,
+      "grad_norm": 1.0599154233932495,
+      "learning_rate": 3.036966473498981e-05,
+      "loss": 0.7835,
+      "step": 14840
+    },
+    {
+      "epoch": 1.1788287126158488,
+      "grad_norm": 0.902905285358429,
+      "learning_rate": 3.035643407160435e-05,
+      "loss": 0.8089,
+      "step": 14850
+    },
+    {
+      "epoch": 1.1796225366647484,
+      "grad_norm": 0.670547604560852,
+      "learning_rate": 3.034320340821889e-05,
+      "loss": 0.7964,
+      "step": 14860
+    },
+    {
+      "epoch": 1.1804163607136477,
+      "grad_norm": 0.7922139763832092,
+      "learning_rate": 3.032997274483343e-05,
+      "loss": 0.8179,
+      "step": 14870
+    },
+    {
+      "epoch": 1.1812101847625474,
+      "grad_norm": 0.8534002304077148,
+      "learning_rate": 3.0316742081447963e-05,
+      "loss": 0.8396,
+      "step": 14880
+    },
+    {
+      "epoch": 1.182004008811447,
+      "grad_norm": 0.8879272937774658,
+      "learning_rate": 3.0303511418062502e-05,
+      "loss": 0.7891,
+      "step": 14890
+    },
+    {
+      "epoch": 1.1827978328603466,
+      "grad_norm": 0.711462676525116,
+      "learning_rate": 3.029028075467704e-05,
+      "loss": 0.7261,
+      "step": 14900
+    },
+    {
+      "epoch": 1.183591656909246,
+      "grad_norm": 0.790237545967102,
+      "learning_rate": 3.027705009129158e-05,
+      "loss": 0.8176,
+      "step": 14910
+    },
+    {
+      "epoch": 1.1843854809581456,
+      "grad_norm": 0.7441157102584839,
+      "learning_rate": 3.0263819427906115e-05,
+      "loss": 0.8004,
+      "step": 14920
+    },
+    {
+      "epoch": 1.1851793050070452,
+      "grad_norm": 0.6485238075256348,
+      "learning_rate": 3.0250588764520654e-05,
+      "loss": 0.8286,
+      "step": 14930
+    },
+    {
+      "epoch": 1.1859731290559448,
+      "grad_norm": 0.5834975242614746,
+      "learning_rate": 3.0237358101135193e-05,
+      "loss": 0.8362,
+      "step": 14940
+    },
+    {
+      "epoch": 1.1867669531048444,
+      "grad_norm": 0.7251124382019043,
+      "learning_rate": 3.0224127437749732e-05,
+      "loss": 0.8968,
+      "step": 14950
+    },
+    {
+      "epoch": 1.187560777153744,
+      "grad_norm": 0.9666823148727417,
+      "learning_rate": 3.0210896774364268e-05,
+      "loss": 0.8247,
+      "step": 14960
+    },
+    {
+      "epoch": 1.1883546012026434,
+      "grad_norm": 0.839264988899231,
+      "learning_rate": 3.0197666110978807e-05,
+      "loss": 0.8347,
+      "step": 14970
+    },
+    {
+      "epoch": 1.189148425251543,
+      "grad_norm": 0.6984066367149353,
+      "learning_rate": 3.0184435447593345e-05,
+      "loss": 0.9087,
+      "step": 14980
+    },
+    {
+      "epoch": 1.1899422493004426,
+      "grad_norm": 0.6566882729530334,
+      "learning_rate": 3.017120478420788e-05,
+      "loss": 0.8933,
+      "step": 14990
+    },
+    {
+      "epoch": 1.1907360733493422,
+      "grad_norm": 0.880733847618103,
+      "learning_rate": 3.015797412082242e-05,
+      "loss": 0.8226,
+      "step": 15000
+    },
+    {
+      "epoch": 1.1915298973982416,
+      "grad_norm": 0.6445071697235107,
+      "learning_rate": 3.014474345743696e-05,
+      "loss": 0.8331,
+      "step": 15010
+    },
+    {
+      "epoch": 1.1923237214471412,
+      "grad_norm": 1.1646579504013062,
+      "learning_rate": 3.0131512794051498e-05,
+      "loss": 0.8809,
+      "step": 15020
+    },
+    {
+      "epoch": 1.1931175454960408,
+      "grad_norm": 0.7741886973381042,
+      "learning_rate": 3.011828213066603e-05,
+      "loss": 0.8101,
+      "step": 15030
+    },
+    {
+      "epoch": 1.1939113695449404,
+      "grad_norm": 0.760037899017334,
+      "learning_rate": 3.0105051467280572e-05,
+      "loss": 0.7767,
+      "step": 15040
+    },
+    {
+      "epoch": 1.19470519359384,
+      "grad_norm": 1.031218409538269,
+      "learning_rate": 3.009182080389511e-05,
+      "loss": 0.8009,
+      "step": 15050
+    },
+    {
+      "epoch": 1.1954990176427396,
+      "grad_norm": 0.6918702125549316,
+      "learning_rate": 3.007859014050965e-05,
+      "loss": 0.7885,
+      "step": 15060
+    },
+    {
+      "epoch": 1.196292841691639,
+      "grad_norm": 1.0237363576889038,
+      "learning_rate": 3.0065359477124182e-05,
+      "loss": 0.8503,
+      "step": 15070
+    },
+    {
+      "epoch": 1.1970866657405386,
+      "grad_norm": 1.026094913482666,
+      "learning_rate": 3.005212881373872e-05,
+      "loss": 0.7394,
+      "step": 15080
+    },
+    {
+      "epoch": 1.1978804897894382,
+      "grad_norm": 0.8072192072868347,
+      "learning_rate": 3.0038898150353263e-05,
+      "loss": 0.8343,
+      "step": 15090
+    },
+    {
+      "epoch": 1.1986743138383378,
+      "grad_norm": 0.8054350018501282,
+      "learning_rate": 3.0025667486967796e-05,
+      "loss": 0.8339,
+      "step": 15100
+    },
+    {
+      "epoch": 1.1994681378872374,
+      "grad_norm": 0.7186553478240967,
+      "learning_rate": 3.0012436823582335e-05,
+      "loss": 0.7915,
+      "step": 15110
+    },
+    {
+      "epoch": 1.2002619619361368,
+      "grad_norm": 0.9386051297187805,
+      "learning_rate": 2.9999206160196874e-05,
+      "loss": 0.8306,
+      "step": 15120
+    },
+    {
+      "epoch": 1.2010557859850364,
+      "grad_norm": 0.7570390105247498,
+      "learning_rate": 2.9985975496811412e-05,
+      "loss": 0.832,
+      "step": 15130
+    },
+    {
+      "epoch": 1.201849610033936,
+      "grad_norm": 0.8414233326911926,
+      "learning_rate": 2.9972744833425948e-05,
+      "loss": 0.8668,
+      "step": 15140
+    },
+    {
+      "epoch": 1.2026434340828356,
+      "grad_norm": 0.790988564491272,
+      "learning_rate": 2.9959514170040487e-05,
+      "loss": 0.802,
+      "step": 15150
+    },
+    {
+      "epoch": 1.2034372581317352,
+      "grad_norm": 1.0100650787353516,
+      "learning_rate": 2.9946283506655026e-05,
+      "loss": 0.8767,
+      "step": 15160
+    },
+    {
+      "epoch": 1.2042310821806346,
+      "grad_norm": 0.8488495349884033,
+      "learning_rate": 2.9933052843269565e-05,
+      "loss": 0.8674,
+      "step": 15170
+    },
+    {
+      "epoch": 1.2050249062295342,
+      "grad_norm": 0.8193151354789734,
+      "learning_rate": 2.99198221798841e-05,
+      "loss": 0.8027,
+      "step": 15180
+    },
+    {
+      "epoch": 1.2058187302784338,
+      "grad_norm": 0.7268685698509216,
+      "learning_rate": 2.990659151649864e-05,
+      "loss": 0.8172,
+      "step": 15190
+    },
+    {
+      "epoch": 1.2066125543273334,
+      "grad_norm": 0.8344203233718872,
+      "learning_rate": 2.9893360853113178e-05,
+      "loss": 0.8125,
+      "step": 15200
+    },
+    {
+      "epoch": 1.207406378376233,
+      "grad_norm": 0.8553740978240967,
+      "learning_rate": 2.988013018972771e-05,
+      "loss": 0.7997,
+      "step": 15210
+    },
+    {
+      "epoch": 1.2082002024251324,
+      "grad_norm": 0.7145830988883972,
+      "learning_rate": 2.9866899526342253e-05,
+      "loss": 0.7697,
+      "step": 15220
+    },
+    {
+      "epoch": 1.208994026474032,
+      "grad_norm": 0.7507617473602295,
+      "learning_rate": 2.985366886295679e-05,
+      "loss": 0.8852,
+      "step": 15230
+    },
+    {
+      "epoch": 1.2097878505229316,
+      "grad_norm": 0.9048145413398743,
+      "learning_rate": 2.984043819957133e-05,
+      "loss": 0.8967,
+      "step": 15240
+    },
+    {
+      "epoch": 1.2105816745718312,
+      "grad_norm": 0.7048582434654236,
+      "learning_rate": 2.9827207536185863e-05,
+      "loss": 0.8613,
+      "step": 15250
+    },
+    {
+      "epoch": 1.2113754986207308,
+      "grad_norm": 0.6709416508674622,
+      "learning_rate": 2.98139768728004e-05,
+      "loss": 0.8725,
+      "step": 15260
+    },
+    {
+      "epoch": 1.2121693226696304,
+      "grad_norm": 0.7239158749580383,
+      "learning_rate": 2.9800746209414944e-05,
+      "loss": 0.8236,
+      "step": 15270
+    },
+    {
+      "epoch": 1.2129631467185298,
+      "grad_norm": 0.7127059102058411,
+      "learning_rate": 2.9787515546029483e-05,
+      "loss": 0.8302,
+      "step": 15280
+    },
+    {
+      "epoch": 1.2137569707674294,
+      "grad_norm": 0.8340943455696106,
+      "learning_rate": 2.9774284882644015e-05,
+      "loss": 0.7475,
+      "step": 15290
+    },
+    {
+      "epoch": 1.214550794816329,
+      "grad_norm": 0.9518420696258545,
+      "learning_rate": 2.9761054219258554e-05,
+      "loss": 0.8673,
+      "step": 15300
+    },
+    {
+      "epoch": 1.2153446188652286,
+      "grad_norm": 0.6714077591896057,
+      "learning_rate": 2.9747823555873093e-05,
+      "loss": 0.7967,
+      "step": 15310
+    },
+    {
+      "epoch": 1.216138442914128,
+      "grad_norm": 0.9402453303337097,
+      "learning_rate": 2.973459289248763e-05,
+      "loss": 0.8375,
+      "step": 15320
+    },
+    {
+      "epoch": 1.2169322669630276,
+      "grad_norm": 0.8551952242851257,
+      "learning_rate": 2.9721362229102167e-05,
+      "loss": 0.8432,
+      "step": 15330
+    },
+    {
+      "epoch": 1.2177260910119272,
+      "grad_norm": 0.9510105848312378,
+      "learning_rate": 2.9708131565716706e-05,
+      "loss": 0.8647,
+      "step": 15340
+    },
+    {
+      "epoch": 1.2185199150608268,
+      "grad_norm": 0.6727361679077148,
+      "learning_rate": 2.9694900902331245e-05,
+      "loss": 0.8573,
+      "step": 15350
+    },
+    {
+      "epoch": 1.2193137391097264,
+      "grad_norm": 0.8175778388977051,
+      "learning_rate": 2.968167023894578e-05,
+      "loss": 0.793,
+      "step": 15360
+    },
+    {
+      "epoch": 1.220107563158626,
+      "grad_norm": 0.771083652973175,
+      "learning_rate": 2.966843957556032e-05,
+      "loss": 0.8401,
+      "step": 15370
+    },
+    {
+      "epoch": 1.2209013872075254,
+      "grad_norm": 0.8669653534889221,
+      "learning_rate": 2.965520891217486e-05,
+      "loss": 0.7758,
+      "step": 15380
+    },
+    {
+      "epoch": 1.221695211256425,
+      "grad_norm": 0.7607120871543884,
+      "learning_rate": 2.9641978248789397e-05,
+      "loss": 0.8172,
+      "step": 15390
+    },
+    {
+      "epoch": 1.2224890353053246,
+      "grad_norm": 0.9306351542472839,
+      "learning_rate": 2.9628747585403933e-05,
+      "loss": 0.8082,
+      "step": 15400
+    },
+    {
+      "epoch": 1.2232828593542242,
+      "grad_norm": 0.8709003329277039,
+      "learning_rate": 2.9615516922018472e-05,
+      "loss": 0.7716,
+      "step": 15410
+    },
+    {
+      "epoch": 1.2240766834031236,
+      "grad_norm": 0.9127326011657715,
+      "learning_rate": 2.960228625863301e-05,
+      "loss": 0.8231,
+      "step": 15420
+    },
+    {
+      "epoch": 1.2248705074520232,
+      "grad_norm": 0.7594894766807556,
+      "learning_rate": 2.9589055595247543e-05,
+      "loss": 0.7788,
+      "step": 15430
+    },
+    {
+      "epoch": 1.2256643315009228,
+      "grad_norm": 0.7096917033195496,
+      "learning_rate": 2.9575824931862085e-05,
+      "loss": 0.8266,
+      "step": 15440
+    },
+    {
+      "epoch": 1.2264581555498224,
+      "grad_norm": 0.8250033855438232,
+      "learning_rate": 2.9562594268476624e-05,
+      "loss": 0.8823,
+      "step": 15450
+    },
+    {
+      "epoch": 1.227251979598722,
+      "grad_norm": 0.8161039352416992,
+      "learning_rate": 2.9550686671429705e-05,
+      "loss": 0.7683,
+      "step": 15460
+    },
+    {
+      "epoch": 1.2280458036476216,
+      "grad_norm": 0.9151235222816467,
+      "learning_rate": 2.9537456008044244e-05,
+      "loss": 0.7988,
+      "step": 15470
+    },
+    {
+      "epoch": 1.228839627696521,
+      "grad_norm": 0.696901798248291,
+      "learning_rate": 2.9524225344658783e-05,
+      "loss": 0.8231,
+      "step": 15480
+    },
+    {
+      "epoch": 1.2296334517454206,
+      "grad_norm": Infinity,
+      "learning_rate": 2.9512317747611867e-05,
+      "loss": 0.8496,
+      "step": 15490
+    },
+    {
+      "epoch": 1.2304272757943202,
+      "grad_norm": 0.7194181084632874,
+      "learning_rate": 2.9499087084226406e-05,
+      "loss": 0.8621,
+      "step": 15500
+    },
+    {
+      "epoch": 1.2312210998432198,
+      "grad_norm": 0.7951611280441284,
+      "learning_rate": 2.9485856420840942e-05,
+      "loss": 0.7799,
+      "step": 15510
+    },
+    {
+      "epoch": 1.2320149238921192,
+      "grad_norm": 0.8867210149765015,
+      "learning_rate": 2.947262575745548e-05,
+      "loss": 0.8452,
+      "step": 15520
+    },
+    {
+      "epoch": 1.2328087479410188,
+      "grad_norm": 0.842461109161377,
+      "learning_rate": 2.945939509407002e-05,
+      "loss": 0.8028,
+      "step": 15530
+    },
+    {
+      "epoch": 1.2336025719899184,
+      "grad_norm": 0.7903575897216797,
+      "learning_rate": 2.944616443068456e-05,
+      "loss": 0.8713,
+      "step": 15540
+    },
+    {
+      "epoch": 1.234396396038818,
+      "grad_norm": 0.7790343165397644,
+      "learning_rate": 2.943293376729909e-05,
+      "loss": 0.7852,
+      "step": 15550
+    },
+    {
+      "epoch": 1.2351902200877176,
+      "grad_norm": 0.8206233978271484,
+      "learning_rate": 2.9419703103913633e-05,
+      "loss": 0.7899,
+      "step": 15560
+    },
+    {
+      "epoch": 1.2359840441366172,
+      "grad_norm": 0.8968833684921265,
+      "learning_rate": 2.9406472440528172e-05,
+      "loss": 0.8477,
+      "step": 15570
+    },
+    {
+      "epoch": 1.2367778681855166,
+      "grad_norm": 0.9603748917579651,
+      "learning_rate": 2.9393241777142704e-05,
+      "loss": 0.79,
+      "step": 15580
+    },
+    {
+      "epoch": 1.2375716922344162,
+      "grad_norm": 0.7768341898918152,
+      "learning_rate": 2.9380011113757243e-05,
+      "loss": 0.7812,
+      "step": 15590
+    },
+    {
+      "epoch": 1.2383655162833158,
+      "grad_norm": 0.7468785047531128,
+      "learning_rate": 2.9366780450371782e-05,
+      "loss": 0.8008,
+      "step": 15600
+    },
+    {
+      "epoch": 1.2391593403322154,
+      "grad_norm": 1.0277005434036255,
+      "learning_rate": 2.9353549786986324e-05,
+      "loss": 0.8168,
+      "step": 15610
+    },
+    {
+      "epoch": 1.2399531643811148,
+      "grad_norm": 0.7243433594703674,
+      "learning_rate": 2.9340319123600857e-05,
+      "loss": 0.8091,
+      "step": 15620
+    },
+    {
+      "epoch": 1.2407469884300144,
+      "grad_norm": 0.8666514158248901,
+      "learning_rate": 2.9327088460215395e-05,
+      "loss": 0.8132,
+      "step": 15630
+    },
+    {
+      "epoch": 1.241540812478914,
+      "grad_norm": 0.80784010887146,
+      "learning_rate": 2.9313857796829934e-05,
+      "loss": 0.8182,
+      "step": 15640
+    },
+    {
+      "epoch": 1.2423346365278136,
+      "grad_norm": 0.743398129940033,
+      "learning_rate": 2.9300627133444473e-05,
+      "loss": 0.8264,
+      "step": 15650
+    },
+    {
+      "epoch": 1.2431284605767132,
+      "grad_norm": 0.6970568895339966,
+      "learning_rate": 2.928739647005901e-05,
+      "loss": 0.8156,
+      "step": 15660
+    },
+    {
+      "epoch": 1.2439222846256128,
+      "grad_norm": 0.640468955039978,
+      "learning_rate": 2.9274165806673548e-05,
+      "loss": 0.8191,
+      "step": 15670
+    },
+    {
+      "epoch": 1.2447161086745122,
+      "grad_norm": 0.9808381795883179,
+      "learning_rate": 2.9260935143288087e-05,
+      "loss": 0.8104,
+      "step": 15680
+    },
+    {
+      "epoch": 1.2455099327234118,
+      "grad_norm": 0.7460395693778992,
+      "learning_rate": 2.9247704479902622e-05,
+      "loss": 0.8131,
+      "step": 15690
+    },
+    {
+      "epoch": 1.2463037567723114,
+      "grad_norm": 0.8486602306365967,
+      "learning_rate": 2.923447381651716e-05,
+      "loss": 0.7889,
+      "step": 15700
+    },
+    {
+      "epoch": 1.247097580821211,
+      "grad_norm": 0.9149585962295532,
+      "learning_rate": 2.92212431531317e-05,
+      "loss": 0.782,
+      "step": 15710
+    },
+    {
+      "epoch": 1.2478914048701106,
+      "grad_norm": 0.7191526889801025,
+      "learning_rate": 2.920801248974624e-05,
+      "loss": 0.8407,
+      "step": 15720
+    },
+    {
+      "epoch": 1.24868522891901,
+      "grad_norm": 0.7577335834503174,
+      "learning_rate": 2.9194781826360775e-05,
+      "loss": 0.8428,
+      "step": 15730
+    },
+    {
+      "epoch": 1.2494790529679096,
+      "grad_norm": 0.8915782570838928,
+      "learning_rate": 2.9181551162975313e-05,
+      "loss": 0.7993,
+      "step": 15740
+    },
+    {
+      "epoch": 1.2502728770168092,
+      "grad_norm": 0.9318637847900391,
+      "learning_rate": 2.9168320499589852e-05,
+      "loss": 0.7586,
+      "step": 15750
+    },
+    {
+      "epoch": 1.2510667010657088,
+      "grad_norm": 0.7829806208610535,
+      "learning_rate": 2.915508983620439e-05,
+      "loss": 0.8027,
+      "step": 15760
+    },
+    {
+      "epoch": 1.2518605251146084,
+      "grad_norm": 0.7112168073654175,
+      "learning_rate": 2.9141859172818923e-05,
+      "loss": 0.8171,
+      "step": 15770
+    },
+    {
+      "epoch": 1.252654349163508,
+      "grad_norm": 0.935157060623169,
+      "learning_rate": 2.9128628509433466e-05,
+      "loss": 0.7728,
+      "step": 15780
+    },
+    {
+      "epoch": 1.2534481732124074,
+      "grad_norm": 0.8060632348060608,
+      "learning_rate": 2.9115397846048005e-05,
+      "loss": 0.835,
+      "step": 15790
+    },
+    {
+      "epoch": 1.254241997261307,
+      "grad_norm": 0.7491572499275208,
+      "learning_rate": 2.9102167182662537e-05,
+      "loss": 0.7851,
+      "step": 15800
+    },
+    {
+      "epoch": 1.2550358213102066,
+      "grad_norm": 0.844757080078125,
+      "learning_rate": 2.9088936519277076e-05,
+      "loss": 0.7442,
+      "step": 15810
+    },
+    {
+      "epoch": 1.255829645359106,
+      "grad_norm": 0.5904254913330078,
+      "learning_rate": 2.9075705855891615e-05,
+      "loss": 0.77,
+      "step": 15820
+    },
+    {
+      "epoch": 1.2566234694080056,
+      "grad_norm": 0.7826739549636841,
+      "learning_rate": 2.9062475192506157e-05,
+      "loss": 0.8496,
+      "step": 15830
+    },
+    {
+      "epoch": 1.2574172934569052,
+      "grad_norm": 0.7734588980674744,
+      "learning_rate": 2.904924452912069e-05,
+      "loss": 0.8367,
+      "step": 15840
+    },
+    {
+      "epoch": 1.2582111175058048,
+      "grad_norm": 0.712218701839447,
+      "learning_rate": 2.9036013865735228e-05,
+      "loss": 0.7709,
+      "step": 15850
+    },
+    {
+      "epoch": 1.2590049415547044,
+      "grad_norm": 0.7345622181892395,
+      "learning_rate": 2.9022783202349767e-05,
+      "loss": 0.8509,
+      "step": 15860
+    },
+    {
+      "epoch": 1.259798765603604,
+      "grad_norm": 0.9125804901123047,
+      "learning_rate": 2.9009552538964306e-05,
+      "loss": 0.8151,
+      "step": 15870
+    },
+    {
+      "epoch": 1.2605925896525036,
+      "grad_norm": 0.7302107214927673,
+      "learning_rate": 2.899632187557884e-05,
+      "loss": 0.8181,
+      "step": 15880
+    },
+    {
+      "epoch": 1.261386413701403,
+      "grad_norm": 0.921686589717865,
+      "learning_rate": 2.898309121219338e-05,
+      "loss": 0.8133,
+      "step": 15890
+    },
+    {
+      "epoch": 1.2621802377503026,
+      "grad_norm": 0.9053015112876892,
+      "learning_rate": 2.896986054880792e-05,
+      "loss": 0.822,
+      "step": 15900
+    },
+    {
+      "epoch": 1.2629740617992022,
+      "grad_norm": 0.8016523718833923,
+      "learning_rate": 2.8956629885422455e-05,
+      "loss": 0.8147,
+      "step": 15910
+    },
+    {
+      "epoch": 1.2637678858481018,
+      "grad_norm": 0.9724611043930054,
+      "learning_rate": 2.8943399222036994e-05,
+      "loss": 0.8608,
+      "step": 15920
+    },
+    {
+      "epoch": 1.2645617098970012,
+      "grad_norm": 0.9794032573699951,
+      "learning_rate": 2.8930168558651533e-05,
+      "loss": 0.7779,
+      "step": 15930
+    },
+    {
+      "epoch": 1.2653555339459008,
+      "grad_norm": 0.6316156387329102,
+      "learning_rate": 2.891693789526607e-05,
+      "loss": 0.8147,
+      "step": 15940
+    },
+    {
+      "epoch": 1.2661493579948004,
+      "grad_norm": 0.835666298866272,
+      "learning_rate": 2.8903707231880607e-05,
+      "loss": 0.8002,
+      "step": 15950
+    },
+    {
+      "epoch": 1.2669431820437,
+      "grad_norm": 0.6945637464523315,
+      "learning_rate": 2.8890476568495146e-05,
+      "loss": 0.9178,
+      "step": 15960
+    },
+    {
+      "epoch": 1.2677370060925996,
+      "grad_norm": 0.6459661722183228,
+      "learning_rate": 2.8877245905109685e-05,
+      "loss": 0.8289,
+      "step": 15970
+    },
+    {
+      "epoch": 1.2685308301414993,
+      "grad_norm": 0.8627393841743469,
+      "learning_rate": 2.8864015241724224e-05,
+      "loss": 0.8809,
+      "step": 15980
+    },
+    {
+      "epoch": 1.2693246541903986,
+      "grad_norm": 0.8296471834182739,
+      "learning_rate": 2.8850784578338756e-05,
+      "loss": 0.8025,
+      "step": 15990
+    },
+    {
+      "epoch": 1.2701184782392982,
+      "grad_norm": 0.8422446846961975,
+      "learning_rate": 2.88375539149533e-05,
+      "loss": 0.8378,
+      "step": 16000
+    },
+    {
+      "epoch": 1.2709123022881978,
+      "grad_norm": 0.8606308102607727,
+      "learning_rate": 2.8824323251567837e-05,
+      "loss": 0.8543,
+      "step": 16010
+    },
+    {
+      "epoch": 1.2717061263370975,
+      "grad_norm": 0.9312155842781067,
+      "learning_rate": 2.881109258818237e-05,
+      "loss": 0.8339,
+      "step": 16020
+    },
+    {
+      "epoch": 1.2724999503859968,
+      "grad_norm": 0.782143771648407,
+      "learning_rate": 2.879786192479691e-05,
+      "loss": 0.7537,
+      "step": 16030
+    },
+    {
+      "epoch": 1.2732937744348964,
+      "grad_norm": 0.7882851958274841,
+      "learning_rate": 2.8784631261411447e-05,
+      "loss": 0.8494,
+      "step": 16040
+    },
+    {
+      "epoch": 1.274087598483796,
+      "grad_norm": 0.8223291039466858,
+      "learning_rate": 2.877140059802599e-05,
+      "loss": 0.788,
+      "step": 16050
+    },
+    {
+      "epoch": 1.2748814225326957,
+      "grad_norm": 0.7009278535842896,
+      "learning_rate": 2.8758169934640522e-05,
+      "loss": 0.8456,
+      "step": 16060
+    },
+    {
+      "epoch": 1.2756752465815953,
+      "grad_norm": 0.7133784294128418,
+      "learning_rate": 2.874493927125506e-05,
+      "loss": 0.7725,
+      "step": 16070
+    },
+    {
+      "epoch": 1.2764690706304949,
+      "grad_norm": 0.7193998098373413,
+      "learning_rate": 2.87317086078696e-05,
+      "loss": 0.8465,
+      "step": 16080
+    },
+    {
+      "epoch": 1.2772628946793942,
+      "grad_norm": 0.6710327863693237,
+      "learning_rate": 2.871847794448414e-05,
+      "loss": 0.8016,
+      "step": 16090
+    },
+    {
+      "epoch": 1.2780567187282939,
+      "grad_norm": 0.8942824006080627,
+      "learning_rate": 2.8705247281098674e-05,
+      "loss": 0.8741,
+      "step": 16100
+    },
+    {
+      "epoch": 1.2788505427771935,
+      "grad_norm": 0.891715407371521,
+      "learning_rate": 2.8692016617713213e-05,
+      "loss": 0.792,
+      "step": 16110
+    },
+    {
+      "epoch": 1.279644366826093,
+      "grad_norm": 0.7947127819061279,
+      "learning_rate": 2.8678785954327752e-05,
+      "loss": 0.8215,
+      "step": 16120
+    },
+    {
+      "epoch": 1.2804381908749924,
+      "grad_norm": 0.7644555568695068,
+      "learning_rate": 2.866555529094229e-05,
+      "loss": 0.7671,
+      "step": 16130
+    },
+    {
+      "epoch": 1.281232014923892,
+      "grad_norm": 0.9834242463111877,
+      "learning_rate": 2.8652324627556827e-05,
+      "loss": 0.7838,
+      "step": 16140
+    },
+    {
+      "epoch": 1.2820258389727917,
+      "grad_norm": 0.851807177066803,
+      "learning_rate": 2.8639093964171365e-05,
+      "loss": 0.9058,
+      "step": 16150
+    },
+    {
+      "epoch": 1.2828196630216913,
+      "grad_norm": 0.7503984570503235,
+      "learning_rate": 2.8625863300785904e-05,
+      "loss": 0.783,
+      "step": 16160
+    },
+    {
+      "epoch": 1.2836134870705909,
+      "grad_norm": 0.7916014790534973,
+      "learning_rate": 2.861263263740044e-05,
+      "loss": 0.888,
+      "step": 16170
+    },
+    {
+      "epoch": 1.2844073111194905,
+      "grad_norm": 0.8114141225814819,
+      "learning_rate": 2.859940197401498e-05,
+      "loss": 0.8192,
+      "step": 16180
+    },
+    {
+      "epoch": 1.2852011351683899,
+      "grad_norm": 0.7763829827308655,
+      "learning_rate": 2.8586171310629518e-05,
+      "loss": 0.8159,
+      "step": 16190
+    },
+    {
+      "epoch": 1.2859949592172895,
+      "grad_norm": 0.7763867378234863,
+      "learning_rate": 2.8572940647244057e-05,
+      "loss": 0.83,
+      "step": 16200
+    },
+    {
+      "epoch": 1.286788783266189,
+      "grad_norm": 0.969960629940033,
+      "learning_rate": 2.855970998385859e-05,
+      "loss": 0.8671,
+      "step": 16210
+    },
+    {
+      "epoch": 1.2875826073150887,
+      "grad_norm": 0.7875334620475769,
+      "learning_rate": 2.854647932047313e-05,
+      "loss": 0.7545,
+      "step": 16220
+    },
+    {
+      "epoch": 1.288376431363988,
+      "grad_norm": 0.9192537665367126,
+      "learning_rate": 2.853324865708767e-05,
+      "loss": 0.8174,
+      "step": 16230
+    },
+    {
+      "epoch": 1.2891702554128877,
+      "grad_norm": 0.8982329368591309,
+      "learning_rate": 2.852001799370221e-05,
+      "loss": 0.8373,
+      "step": 16240
+    },
+    {
+      "epoch": 1.2899640794617873,
+      "grad_norm": 0.7992169857025146,
+      "learning_rate": 2.850678733031674e-05,
+      "loss": 0.8858,
+      "step": 16250
+    },
+    {
+      "epoch": 1.2907579035106869,
+      "grad_norm": 0.7234670519828796,
+      "learning_rate": 2.849355666693128e-05,
+      "loss": 0.8548,
+      "step": 16260
+    },
+    {
+      "epoch": 1.2915517275595865,
+      "grad_norm": 0.803986132144928,
+      "learning_rate": 2.8480326003545822e-05,
+      "loss": 0.843,
+      "step": 16270
+    },
+    {
+      "epoch": 1.292345551608486,
+      "grad_norm": 0.6676992177963257,
+      "learning_rate": 2.8467095340160355e-05,
+      "loss": 0.8432,
+      "step": 16280
+    },
+    {
+      "epoch": 1.2931393756573855,
+      "grad_norm": 0.7712359428405762,
+      "learning_rate": 2.8453864676774893e-05,
+      "loss": 0.7889,
+      "step": 16290
+    },
+    {
+      "epoch": 1.293933199706285,
+      "grad_norm": 0.9185086488723755,
+      "learning_rate": 2.8440634013389432e-05,
+      "loss": 0.8816,
+      "step": 16300
+    },
+    {
+      "epoch": 1.2947270237551847,
+      "grad_norm": 0.7256115078926086,
+      "learning_rate": 2.842740335000397e-05,
+      "loss": 0.8636,
+      "step": 16310
+    },
+    {
+      "epoch": 1.2955208478040843,
+      "grad_norm": 0.9436436295509338,
+      "learning_rate": 2.8414172686618507e-05,
+      "loss": 0.8138,
+      "step": 16320
+    },
+    {
+      "epoch": 1.2963146718529837,
+      "grad_norm": 0.8887737393379211,
+      "learning_rate": 2.8400942023233046e-05,
+      "loss": 0.7817,
+      "step": 16330
+    },
+    {
+      "epoch": 1.2971084959018833,
+      "grad_norm": 0.8305734395980835,
+      "learning_rate": 2.8387711359847585e-05,
+      "loss": 0.7737,
+      "step": 16340
+    },
+    {
+      "epoch": 1.2979023199507829,
+      "grad_norm": 0.8457329273223877,
+      "learning_rate": 2.8374480696462124e-05,
+      "loss": 0.7743,
+      "step": 16350
+    },
+    {
+      "epoch": 1.2986961439996825,
+      "grad_norm": 0.7918816208839417,
+      "learning_rate": 2.836125003307666e-05,
+      "loss": 0.8273,
+      "step": 16360
+    },
+    {
+      "epoch": 1.299489968048582,
+      "grad_norm": 0.6060484647750854,
+      "learning_rate": 2.8348019369691198e-05,
+      "loss": 0.7823,
+      "step": 16370
+    },
+    {
+      "epoch": 1.3002837920974817,
+      "grad_norm": 0.9133262038230896,
+      "learning_rate": 2.8334788706305737e-05,
+      "loss": 0.7733,
+      "step": 16380
+    },
+    {
+      "epoch": 1.3010776161463813,
+      "grad_norm": 0.8485836386680603,
+      "learning_rate": 2.8321558042920273e-05,
+      "loss": 0.7581,
+      "step": 16390
+    },
+    {
+      "epoch": 1.3018714401952807,
+      "grad_norm": 0.7613677978515625,
+      "learning_rate": 2.830832737953481e-05,
+      "loss": 0.8436,
+      "step": 16400
+    },
+    {
+      "epoch": 1.3026652642441803,
+      "grad_norm": 0.8914282917976379,
+      "learning_rate": 2.829509671614935e-05,
+      "loss": 0.8283,
+      "step": 16410
+    },
+    {
+      "epoch": 1.3034590882930799,
+      "grad_norm": 0.8663108944892883,
+      "learning_rate": 2.828186605276389e-05,
+      "loss": 0.7854,
+      "step": 16420
+    },
+    {
+      "epoch": 1.3042529123419793,
+      "grad_norm": 0.8851757049560547,
+      "learning_rate": 2.826863538937842e-05,
+      "loss": 0.824,
+      "step": 16430
+    },
+    {
+      "epoch": 1.3050467363908789,
+      "grad_norm": 0.8618358969688416,
+      "learning_rate": 2.8255404725992964e-05,
+      "loss": 0.8277,
+      "step": 16440
+    },
+    {
+      "epoch": 1.3058405604397785,
+      "grad_norm": 0.8172771334648132,
+      "learning_rate": 2.8242174062607503e-05,
+      "loss": 0.8929,
+      "step": 16450
+    },
+    {
+      "epoch": 1.306634384488678,
+      "grad_norm": 0.898499608039856,
+      "learning_rate": 2.822894339922204e-05,
+      "loss": 0.8683,
+      "step": 16460
+    },
+    {
+      "epoch": 1.3074282085375777,
+      "grad_norm": 0.7734599113464355,
+      "learning_rate": 2.8215712735836574e-05,
+      "loss": 0.8588,
+      "step": 16470
+    },
+    {
+      "epoch": 1.3082220325864773,
+      "grad_norm": 0.8357298374176025,
+      "learning_rate": 2.8202482072451113e-05,
+      "loss": 0.7965,
+      "step": 16480
+    },
+    {
+      "epoch": 1.309015856635377,
+      "grad_norm": 0.8872872591018677,
+      "learning_rate": 2.8189251409065655e-05,
+      "loss": 0.7839,
+      "step": 16490
+    },
+    {
+      "epoch": 1.3098096806842763,
+      "grad_norm": 0.9397867321968079,
+      "learning_rate": 2.8176020745680187e-05,
+      "loss": 0.8508,
+      "step": 16500
+    },
+    {
+      "epoch": 1.3106035047331759,
+      "grad_norm": 0.790989339351654,
+      "learning_rate": 2.8162790082294726e-05,
+      "loss": 0.8086,
+      "step": 16510
+    },
+    {
+      "epoch": 1.3113973287820755,
+      "grad_norm": 0.747873842716217,
+      "learning_rate": 2.8149559418909265e-05,
+      "loss": 0.8536,
+      "step": 16520
+    },
+    {
+      "epoch": 1.312191152830975,
+      "grad_norm": 0.8547879457473755,
+      "learning_rate": 2.8136328755523804e-05,
+      "loss": 0.7932,
+      "step": 16530
+    },
+    {
+      "epoch": 1.3129849768798745,
+      "grad_norm": 0.8623278141021729,
+      "learning_rate": 2.812309809213834e-05,
+      "loss": 0.765,
+      "step": 16540
+    },
+    {
+      "epoch": 1.313778800928774,
+      "grad_norm": 0.7240385413169861,
+      "learning_rate": 2.810986742875288e-05,
+      "loss": 0.8292,
+      "step": 16550
+    },
+    {
+      "epoch": 1.3145726249776737,
+      "grad_norm": 0.9323850274085999,
+      "learning_rate": 2.8096636765367417e-05,
+      "loss": 0.76,
+      "step": 16560
+    },
+    {
+      "epoch": 1.3153664490265733,
+      "grad_norm": 0.695256769657135,
+      "learning_rate": 2.8083406101981956e-05,
+      "loss": 0.8158,
+      "step": 16570
+    },
+    {
+      "epoch": 1.316160273075473,
+      "grad_norm": 0.6848450899124146,
+      "learning_rate": 2.8070175438596492e-05,
+      "loss": 0.7873,
+      "step": 16580
+    },
+    {
+      "epoch": 1.3169540971243725,
+      "grad_norm": 0.7952008247375488,
+      "learning_rate": 2.805694477521103e-05,
+      "loss": 0.8798,
+      "step": 16590
+    },
+    {
+      "epoch": 1.3177479211732719,
+      "grad_norm": 0.8499898314476013,
+      "learning_rate": 2.804371411182557e-05,
+      "loss": 0.7545,
+      "step": 16600
+    },
+    {
+      "epoch": 1.3185417452221715,
+      "grad_norm": 0.8064092993736267,
+      "learning_rate": 2.8030483448440102e-05,
+      "loss": 0.7962,
+      "step": 16610
+    },
+    {
+      "epoch": 1.319335569271071,
+      "grad_norm": 0.9621743559837341,
+      "learning_rate": 2.8017252785054644e-05,
+      "loss": 0.853,
+      "step": 16620
+    },
+    {
+      "epoch": 1.3201293933199707,
+      "grad_norm": 0.7509051561355591,
+      "learning_rate": 2.8004022121669183e-05,
+      "loss": 0.8417,
+      "step": 16630
+    },
+    {
+      "epoch": 1.32092321736887,
+      "grad_norm": 0.7417434453964233,
+      "learning_rate": 2.7990791458283722e-05,
+      "loss": 0.8704,
+      "step": 16640
+    },
+    {
+      "epoch": 1.3217170414177697,
+      "grad_norm": 0.8676928877830505,
+      "learning_rate": 2.7977560794898254e-05,
+      "loss": 0.789,
+      "step": 16650
+    },
+    {
+      "epoch": 1.3225108654666693,
+      "grad_norm": 0.907887876033783,
+      "learning_rate": 2.7964330131512793e-05,
+      "loss": 0.8448,
+      "step": 16660
+    },
+    {
+      "epoch": 1.323304689515569,
+      "grad_norm": 0.7819181084632874,
+      "learning_rate": 2.7951099468127335e-05,
+      "loss": 0.8467,
+      "step": 16670
+    },
+    {
+      "epoch": 1.3240985135644685,
+      "grad_norm": 0.7135961651802063,
+      "learning_rate": 2.7937868804741874e-05,
+      "loss": 0.8322,
+      "step": 16680
+    },
+    {
+      "epoch": 1.324892337613368,
+      "grad_norm": 0.7459518313407898,
+      "learning_rate": 2.7924638141356407e-05,
+      "loss": 0.8954,
+      "step": 16690
+    },
+    {
+      "epoch": 1.3256861616622675,
+      "grad_norm": 0.8480586409568787,
+      "learning_rate": 2.7911407477970945e-05,
+      "loss": 0.8421,
+      "step": 16700
+    },
+    {
+      "epoch": 1.326479985711167,
+      "grad_norm": 0.7946953773498535,
+      "learning_rate": 2.7898176814585484e-05,
+      "loss": 0.7894,
+      "step": 16710
+    },
+    {
+      "epoch": 1.3272738097600667,
+      "grad_norm": 0.9292244911193848,
+      "learning_rate": 2.788494615120002e-05,
+      "loss": 0.8569,
+      "step": 16720
+    },
+    {
+      "epoch": 1.3280676338089663,
+      "grad_norm": 0.8674454689025879,
+      "learning_rate": 2.787171548781456e-05,
+      "loss": 0.8093,
+      "step": 16730
+    },
+    {
+      "epoch": 1.3288614578578657,
+      "grad_norm": 0.9656598567962646,
+      "learning_rate": 2.7858484824429098e-05,
+      "loss": 0.7962,
+      "step": 16740
+    },
+    {
+      "epoch": 1.3296552819067653,
+      "grad_norm": 0.7918367385864258,
+      "learning_rate": 2.7845254161043637e-05,
+      "loss": 0.8784,
+      "step": 16750
+    },
+    {
+      "epoch": 1.330449105955665,
+      "grad_norm": 0.8442765474319458,
+      "learning_rate": 2.7832023497658172e-05,
+      "loss": 0.8143,
+      "step": 16760
+    },
+    {
+      "epoch": 1.3312429300045645,
+      "grad_norm": 0.8403950333595276,
+      "learning_rate": 2.781879283427271e-05,
+      "loss": 0.8452,
+      "step": 16770
+    },
+    {
+      "epoch": 1.3320367540534641,
+      "grad_norm": 0.8802147507667542,
+      "learning_rate": 2.780556217088725e-05,
+      "loss": 0.8828,
+      "step": 16780
+    },
+    {
+      "epoch": 1.3328305781023637,
+      "grad_norm": 0.8321970105171204,
+      "learning_rate": 2.779233150750179e-05,
+      "loss": 0.8389,
+      "step": 16790
+    },
+    {
+      "epoch": 1.333624402151263,
+      "grad_norm": 0.814605176448822,
+      "learning_rate": 2.7779100844116325e-05,
+      "loss": 0.8263,
+      "step": 16800
+    },
+    {
+      "epoch": 1.3344182262001627,
+      "grad_norm": 0.8342418670654297,
+      "learning_rate": 2.7765870180730863e-05,
+      "loss": 0.7823,
+      "step": 16810
+    },
+    {
+      "epoch": 1.3352120502490623,
+      "grad_norm": 0.8523891568183899,
+      "learning_rate": 2.7752639517345402e-05,
+      "loss": 0.8527,
+      "step": 16820
+    },
+    {
+      "epoch": 1.336005874297962,
+      "grad_norm": 0.9738366603851318,
+      "learning_rate": 2.7739408853959935e-05,
+      "loss": 0.8452,
+      "step": 16830
+    },
+    {
+      "epoch": 1.3367996983468613,
+      "grad_norm": 0.6618583798408508,
+      "learning_rate": 2.7726178190574477e-05,
+      "loss": 0.9022,
+      "step": 16840
+    },
+    {
+      "epoch": 1.337593522395761,
+      "grad_norm": 0.6916811466217041,
+      "learning_rate": 2.7712947527189016e-05,
+      "loss": 0.8484,
+      "step": 16850
+    },
+    {
+      "epoch": 1.3383873464446605,
+      "grad_norm": 0.8240322470664978,
+      "learning_rate": 2.7699716863803555e-05,
+      "loss": 0.7851,
+      "step": 16860
+    },
+    {
+      "epoch": 1.3391811704935601,
+      "grad_norm": 0.8770764470100403,
+      "learning_rate": 2.7686486200418087e-05,
+      "loss": 0.8042,
+      "step": 16870
+    },
+    {
+      "epoch": 1.3399749945424597,
+      "grad_norm": 0.8255194425582886,
+      "learning_rate": 2.7673255537032626e-05,
+      "loss": 0.8021,
+      "step": 16880
+    },
+    {
+      "epoch": 1.3407688185913593,
+      "grad_norm": 0.822505533695221,
+      "learning_rate": 2.7660024873647168e-05,
+      "loss": 0.7602,
+      "step": 16890
+    },
+    {
+      "epoch": 1.3415626426402587,
+      "grad_norm": 1.0661057233810425,
+      "learning_rate": 2.7646794210261707e-05,
+      "loss": 0.8531,
+      "step": 16900
+    },
+    {
+      "epoch": 1.3423564666891583,
+      "grad_norm": 0.8882428407669067,
+      "learning_rate": 2.763356354687624e-05,
+      "loss": 0.8681,
+      "step": 16910
+    },
+    {
+      "epoch": 1.343150290738058,
+      "grad_norm": 0.8333000540733337,
+      "learning_rate": 2.7620332883490778e-05,
+      "loss": 0.8016,
+      "step": 16920
+    },
+    {
+      "epoch": 1.3439441147869575,
+      "grad_norm": 0.7466040849685669,
+      "learning_rate": 2.7607102220105317e-05,
+      "loss": 0.8253,
+      "step": 16930
+    },
+    {
+      "epoch": 1.344737938835857,
+      "grad_norm": 0.9921952486038208,
+      "learning_rate": 2.759387155671986e-05,
+      "loss": 0.8058,
+      "step": 16940
+    },
+    {
+      "epoch": 1.3455317628847565,
+      "grad_norm": 0.7671175599098206,
+      "learning_rate": 2.758064089333439e-05,
+      "loss": 0.8876,
+      "step": 16950
+    },
+    {
+      "epoch": 1.3463255869336561,
+      "grad_norm": 0.5953800678253174,
+      "learning_rate": 2.756741022994893e-05,
+      "loss": 0.8096,
+      "step": 16960
+    },
+    {
+      "epoch": 1.3471194109825557,
+      "grad_norm": 0.8293882012367249,
+      "learning_rate": 2.755417956656347e-05,
+      "loss": 0.7875,
+      "step": 16970
+    },
+    {
+      "epoch": 1.3479132350314553,
+      "grad_norm": 0.7418809533119202,
+      "learning_rate": 2.7540948903178005e-05,
+      "loss": 0.7676,
+      "step": 16980
+    },
+    {
+      "epoch": 1.348707059080355,
+      "grad_norm": 0.8036984205245972,
+      "learning_rate": 2.7527718239792544e-05,
+      "loss": 0.8258,
+      "step": 16990
+    },
+    {
+      "epoch": 1.3495008831292545,
+      "grad_norm": 0.9726935625076294,
+      "learning_rate": 2.7514487576407083e-05,
+      "loss": 0.8252,
+      "step": 17000
+    },
+    {
+      "epoch": 1.350294707178154,
+      "grad_norm": 0.9548497796058655,
+      "learning_rate": 2.750125691302162e-05,
+      "loss": 0.7797,
+      "step": 17010
+    },
+    {
+      "epoch": 1.3510885312270535,
+      "grad_norm": 0.84930819272995,
+      "learning_rate": 2.7488026249636157e-05,
+      "loss": 0.8566,
+      "step": 17020
+    },
+    {
+      "epoch": 1.3518823552759531,
+      "grad_norm": 0.8506318926811218,
+      "learning_rate": 2.7474795586250696e-05,
+      "loss": 0.8499,
+      "step": 17030
+    },
+    {
+      "epoch": 1.3526761793248525,
+      "grad_norm": 0.7205038070678711,
+      "learning_rate": 2.7461564922865235e-05,
+      "loss": 0.8265,
+      "step": 17040
+    },
+    {
+      "epoch": 1.3534700033737521,
+      "grad_norm": 0.8020601868629456,
+      "learning_rate": 2.7448334259479774e-05,
+      "loss": 0.7768,
+      "step": 17050
+    },
+    {
+      "epoch": 1.3542638274226517,
+      "grad_norm": 1.0212651491165161,
+      "learning_rate": 2.743510359609431e-05,
+      "loss": 0.8562,
+      "step": 17060
+    },
+    {
+      "epoch": 1.3550576514715513,
+      "grad_norm": 0.9957228302955627,
+      "learning_rate": 2.742187293270885e-05,
+      "loss": 0.8412,
+      "step": 17070
+    },
+    {
+      "epoch": 1.355851475520451,
+      "grad_norm": 0.8726456165313721,
+      "learning_rate": 2.7408642269323387e-05,
+      "loss": 0.8624,
+      "step": 17080
+    },
+    {
+      "epoch": 1.3566452995693505,
+      "grad_norm": 0.9366957545280457,
+      "learning_rate": 2.739541160593792e-05,
+      "loss": 0.8055,
+      "step": 17090
+    },
+    {
+      "epoch": 1.3574391236182501,
+      "grad_norm": 0.7526766657829285,
+      "learning_rate": 2.738218094255246e-05,
+      "loss": 0.901,
+      "step": 17100
+    },
+    {
+      "epoch": 1.3582329476671495,
+      "grad_norm": 0.8756746053695679,
+      "learning_rate": 2.7368950279167e-05,
+      "loss": 0.8179,
+      "step": 17110
+    },
+    {
+      "epoch": 1.3590267717160491,
+      "grad_norm": 0.7645390629768372,
+      "learning_rate": 2.735571961578154e-05,
+      "loss": 0.8382,
+      "step": 17120
+    },
+    {
+      "epoch": 1.3598205957649487,
+      "grad_norm": 0.8134385943412781,
+      "learning_rate": 2.7342488952396072e-05,
+      "loss": 0.8029,
+      "step": 17130
+    },
+    {
+      "epoch": 1.3606144198138483,
+      "grad_norm": 0.8174660205841064,
+      "learning_rate": 2.732925828901061e-05,
+      "loss": 0.8108,
+      "step": 17140
+    },
+    {
+      "epoch": 1.3614082438627477,
+      "grad_norm": 0.8981174230575562,
+      "learning_rate": 2.731602762562515e-05,
+      "loss": 0.8636,
+      "step": 17150
+    },
+    {
+      "epoch": 1.3622020679116473,
+      "grad_norm": 0.7757834792137146,
+      "learning_rate": 2.7302796962239692e-05,
+      "loss": 0.8198,
+      "step": 17160
+    },
+    {
+      "epoch": 1.362995891960547,
+      "grad_norm": 0.850306510925293,
+      "learning_rate": 2.7289566298854224e-05,
+      "loss": 0.8118,
+      "step": 17170
+    },
+    {
+      "epoch": 1.3637897160094465,
+      "grad_norm": 0.7254217863082886,
+      "learning_rate": 2.7276335635468763e-05,
+      "loss": 0.827,
+      "step": 17180
+    },
+    {
+      "epoch": 1.3645835400583461,
+      "grad_norm": 1.0736218690872192,
+      "learning_rate": 2.7263104972083302e-05,
+      "loss": 0.8839,
+      "step": 17190
+    },
+    {
+      "epoch": 1.3653773641072458,
+      "grad_norm": 0.8943197131156921,
+      "learning_rate": 2.7249874308697838e-05,
+      "loss": 0.7996,
+      "step": 17200
+    },
+    {
+      "epoch": 1.3661711881561451,
+      "grad_norm": 0.9427874684333801,
+      "learning_rate": 2.7236643645312377e-05,
+      "loss": 0.8843,
+      "step": 17210
+    },
+    {
+      "epoch": 1.3669650122050447,
+      "grad_norm": 1.0540951490402222,
+      "learning_rate": 2.7223412981926915e-05,
+      "loss": 0.8258,
+      "step": 17220
+    },
+    {
+      "epoch": 1.3677588362539443,
+      "grad_norm": 0.827373206615448,
+      "learning_rate": 2.7210182318541454e-05,
+      "loss": 0.8492,
+      "step": 17230
+    },
+    {
+      "epoch": 1.368552660302844,
+      "grad_norm": 0.7452979683876038,
+      "learning_rate": 2.719695165515599e-05,
+      "loss": 0.8457,
+      "step": 17240
+    },
+    {
+      "epoch": 1.3693464843517433,
+      "grad_norm": 1.0606896877288818,
+      "learning_rate": 2.718372099177053e-05,
+      "loss": 0.8411,
+      "step": 17250
+    },
+    {
+      "epoch": 1.370140308400643,
+      "grad_norm": 0.6817187070846558,
+      "learning_rate": 2.7170490328385068e-05,
+      "loss": 0.7858,
+      "step": 17260
+    },
+    {
+      "epoch": 1.3709341324495425,
+      "grad_norm": 0.9781541228294373,
+      "learning_rate": 2.7157259664999607e-05,
+      "loss": 0.8377,
+      "step": 17270
+    },
+    {
+      "epoch": 1.3717279564984421,
+      "grad_norm": 0.787799060344696,
+      "learning_rate": 2.7144029001614142e-05,
+      "loss": 0.8435,
+      "step": 17280
+    },
+    {
+      "epoch": 1.3725217805473418,
+      "grad_norm": 0.7571372389793396,
+      "learning_rate": 2.713079833822868e-05,
+      "loss": 0.8434,
+      "step": 17290
+    },
+    {
+      "epoch": 1.3733156045962414,
+      "grad_norm": 0.6738792061805725,
+      "learning_rate": 2.711756767484322e-05,
+      "loss": 0.8412,
+      "step": 17300
+    },
+    {
+      "epoch": 1.3741094286451407,
+      "grad_norm": 1.037701964378357,
+      "learning_rate": 2.7104337011457752e-05,
+      "loss": 0.831,
+      "step": 17310
+    },
+    {
+      "epoch": 1.3749032526940403,
+      "grad_norm": 0.7781122326850891,
+      "learning_rate": 2.709110634807229e-05,
+      "loss": 0.8592,
+      "step": 17320
+    },
+    {
+      "epoch": 1.37569707674294,
+      "grad_norm": 0.9618025422096252,
+      "learning_rate": 2.7077875684686833e-05,
+      "loss": 0.808,
+      "step": 17330
+    },
+    {
+      "epoch": 1.3764909007918396,
+      "grad_norm": 0.768357515335083,
+      "learning_rate": 2.7064645021301372e-05,
+      "loss": 0.8701,
+      "step": 17340
+    },
+    {
+      "epoch": 1.377284724840739,
+      "grad_norm": 0.8587696552276611,
+      "learning_rate": 2.7051414357915905e-05,
+      "loss": 0.8703,
+      "step": 17350
+    },
+    {
+      "epoch": 1.3780785488896385,
+      "grad_norm": 0.9739010334014893,
+      "learning_rate": 2.7038183694530443e-05,
+      "loss": 0.7343,
+      "step": 17360
+    },
+    {
+      "epoch": 1.3788723729385381,
+      "grad_norm": 0.7748016119003296,
+      "learning_rate": 2.7024953031144982e-05,
+      "loss": 0.7732,
+      "step": 17370
+    },
+    {
+      "epoch": 1.3796661969874378,
+      "grad_norm": 0.7629042863845825,
+      "learning_rate": 2.7011722367759525e-05,
+      "loss": 0.7921,
+      "step": 17380
+    },
+    {
+      "epoch": 1.3804600210363374,
+      "grad_norm": 0.9152623414993286,
+      "learning_rate": 2.6998491704374057e-05,
+      "loss": 0.8426,
+      "step": 17390
+    },
+    {
+      "epoch": 1.381253845085237,
+      "grad_norm": 0.8051658868789673,
+      "learning_rate": 2.6985261040988596e-05,
+      "loss": 0.8433,
+      "step": 17400
+    },
+    {
+      "epoch": 1.3820476691341363,
+      "grad_norm": 0.7106397151947021,
+      "learning_rate": 2.6972030377603135e-05,
+      "loss": 0.8297,
+      "step": 17410
+    },
+    {
+      "epoch": 1.382841493183036,
+      "grad_norm": 0.7966054677963257,
+      "learning_rate": 2.695879971421767e-05,
+      "loss": 0.9135,
+      "step": 17420
+    },
+    {
+      "epoch": 1.3836353172319356,
+      "grad_norm": 0.8528908491134644,
+      "learning_rate": 2.694556905083221e-05,
+      "loss": 0.7658,
+      "step": 17430
+    },
+    {
+      "epoch": 1.3844291412808352,
+      "grad_norm": 0.7444737553596497,
+      "learning_rate": 2.6932338387446748e-05,
+      "loss": 0.7901,
+      "step": 17440
+    },
+    {
+      "epoch": 1.3852229653297345,
+      "grad_norm": 0.6893477439880371,
+      "learning_rate": 2.6919107724061287e-05,
+      "loss": 0.8823,
+      "step": 17450
+    },
+    {
+      "epoch": 1.3860167893786342,
+      "grad_norm": 0.8451089859008789,
+      "learning_rate": 2.6905877060675823e-05,
+      "loss": 0.8699,
+      "step": 17460
+    },
+    {
+      "epoch": 1.3868106134275338,
+      "grad_norm": 0.7939800024032593,
+      "learning_rate": 2.689264639729036e-05,
+      "loss": 0.8529,
+      "step": 17470
+    },
+    {
+      "epoch": 1.3876044374764334,
+      "grad_norm": 0.7365521788597107,
+      "learning_rate": 2.68794157339049e-05,
+      "loss": 0.8272,
+      "step": 17480
+    },
+    {
+      "epoch": 1.388398261525333,
+      "grad_norm": 0.8047791123390198,
+      "learning_rate": 2.686618507051944e-05,
+      "loss": 0.7692,
+      "step": 17490
+    },
+    {
+      "epoch": 1.3891920855742326,
+      "grad_norm": 0.7910019755363464,
+      "learning_rate": 2.6852954407133975e-05,
+      "loss": 0.8384,
+      "step": 17500
+    },
+    {
+      "epoch": 1.389985909623132,
+      "grad_norm": 0.8277596235275269,
+      "learning_rate": 2.6839723743748514e-05,
+      "loss": 0.7949,
+      "step": 17510
+    },
+    {
+      "epoch": 1.3907797336720316,
+      "grad_norm": 1.008076786994934,
+      "learning_rate": 2.6826493080363053e-05,
+      "loss": 0.8341,
+      "step": 17520
+    },
+    {
+      "epoch": 1.3915735577209312,
+      "grad_norm": 0.7078754305839539,
+      "learning_rate": 2.6813262416977585e-05,
+      "loss": 0.8233,
+      "step": 17530
+    },
+    {
+      "epoch": 1.3923673817698308,
+      "grad_norm": 0.8263536095619202,
+      "learning_rate": 2.6800031753592124e-05,
+      "loss": 0.8366,
+      "step": 17540
+    },
+    {
+      "epoch": 1.3931612058187302,
+      "grad_norm": 0.8170019388198853,
+      "learning_rate": 2.6786801090206666e-05,
+      "loss": 0.83,
+      "step": 17550
+    },
+    {
+      "epoch": 1.3939550298676298,
+      "grad_norm": 0.8896247744560242,
+      "learning_rate": 2.6773570426821205e-05,
+      "loss": 0.8493,
+      "step": 17560
+    },
+    {
+      "epoch": 1.3947488539165294,
+      "grad_norm": 0.6989545822143555,
+      "learning_rate": 2.6760339763435737e-05,
+      "loss": 0.8151,
+      "step": 17570
+    },
+    {
+      "epoch": 1.395542677965429,
+      "grad_norm": 0.956941545009613,
+      "learning_rate": 2.6747109100050276e-05,
+      "loss": 0.8058,
+      "step": 17580
+    },
+    {
+      "epoch": 1.3963365020143286,
+      "grad_norm": 0.8271597027778625,
+      "learning_rate": 2.6733878436664815e-05,
+      "loss": 0.7853,
+      "step": 17590
+    },
+    {
+      "epoch": 1.3971303260632282,
+      "grad_norm": 0.7426630258560181,
+      "learning_rate": 2.6720647773279357e-05,
+      "loss": 0.7689,
+      "step": 17600
+    },
+    {
+      "epoch": 1.3979241501121278,
+      "grad_norm": 0.8329174518585205,
+      "learning_rate": 2.670741710989389e-05,
+      "loss": 0.8266,
+      "step": 17610
+    },
+    {
+      "epoch": 1.3987179741610272,
+      "grad_norm": 0.918170154094696,
+      "learning_rate": 2.669418644650843e-05,
+      "loss": 0.8444,
+      "step": 17620
+    },
+    {
+      "epoch": 1.3995117982099268,
+      "grad_norm": 0.6863502264022827,
+      "learning_rate": 2.6680955783122967e-05,
+      "loss": 0.8654,
+      "step": 17630
+    },
+    {
+      "epoch": 1.4003056222588264,
+      "grad_norm": 0.8955294489860535,
+      "learning_rate": 2.6667725119737503e-05,
+      "loss": 0.8649,
+      "step": 17640
+    },
+    {
+      "epoch": 1.4010994463077258,
+      "grad_norm": 0.8615051507949829,
+      "learning_rate": 2.6654494456352042e-05,
+      "loss": 0.7899,
+      "step": 17650
+    },
+    {
+      "epoch": 1.4018932703566254,
+      "grad_norm": 0.9451674222946167,
+      "learning_rate": 2.664126379296658e-05,
+      "loss": 0.8824,
+      "step": 17660
+    },
+    {
+      "epoch": 1.402687094405525,
+      "grad_norm": 0.8407998085021973,
+      "learning_rate": 2.662803312958112e-05,
+      "loss": 0.7696,
+      "step": 17670
+    },
+    {
+      "epoch": 1.4034809184544246,
+      "grad_norm": 0.8204694390296936,
+      "learning_rate": 2.6614802466195655e-05,
+      "loss": 0.7892,
+      "step": 17680
+    },
+    {
+      "epoch": 1.4042747425033242,
+      "grad_norm": 0.7346405982971191,
+      "learning_rate": 2.6601571802810194e-05,
+      "loss": 0.794,
+      "step": 17690
+    },
+    {
+      "epoch": 1.4050685665522238,
+      "grad_norm": 0.7684462070465088,
+      "learning_rate": 2.6588341139424733e-05,
+      "loss": 0.8656,
+      "step": 17700
+    },
+    {
+      "epoch": 1.4058623906011234,
+      "grad_norm": 0.7464351058006287,
+      "learning_rate": 2.6575110476039272e-05,
+      "loss": 0.8215,
+      "step": 17710
+    },
+    {
+      "epoch": 1.4066562146500228,
+      "grad_norm": 0.8222455382347107,
+      "learning_rate": 2.6561879812653808e-05,
+      "loss": 0.835,
+      "step": 17720
+    },
+    {
+      "epoch": 1.4074500386989224,
+      "grad_norm": 0.9029430150985718,
+      "learning_rate": 2.6548649149268347e-05,
+      "loss": 0.8038,
+      "step": 17730
+    },
+    {
+      "epoch": 1.408243862747822,
+      "grad_norm": 0.6402363777160645,
+      "learning_rate": 2.6535418485882885e-05,
+      "loss": 0.8557,
+      "step": 17740
+    },
+    {
+      "epoch": 1.4090376867967216,
+      "grad_norm": 0.7686623930931091,
+      "learning_rate": 2.6522187822497424e-05,
+      "loss": 0.7969,
+      "step": 17750
+    },
+    {
+      "epoch": 1.409831510845621,
+      "grad_norm": 0.8864317536354065,
+      "learning_rate": 2.6508957159111957e-05,
+      "loss": 0.8525,
+      "step": 17760
+    },
+    {
+      "epoch": 1.4106253348945206,
+      "grad_norm": 0.8594456911087036,
+      "learning_rate": 2.64957264957265e-05,
+      "loss": 0.8271,
+      "step": 17770
+    },
+    {
+      "epoch": 1.4114191589434202,
+      "grad_norm": 0.8445634841918945,
+      "learning_rate": 2.6482495832341038e-05,
+      "loss": 0.761,
+      "step": 17780
+    },
+    {
+      "epoch": 1.4122129829923198,
+      "grad_norm": 0.8919235467910767,
+      "learning_rate": 2.646926516895557e-05,
+      "loss": 0.8448,
+      "step": 17790
+    },
+    {
+      "epoch": 1.4130068070412194,
+      "grad_norm": 0.9002474546432495,
+      "learning_rate": 2.645603450557011e-05,
+      "loss": 0.873,
+      "step": 17800
+    },
+    {
+      "epoch": 1.413800631090119,
+      "grad_norm": 0.8635540008544922,
+      "learning_rate": 2.6442803842184648e-05,
+      "loss": 0.8874,
+      "step": 17810
+    },
+    {
+      "epoch": 1.4145944551390184,
+      "grad_norm": 0.8504475355148315,
+      "learning_rate": 2.642957317879919e-05,
+      "loss": 0.8296,
+      "step": 17820
+    },
+    {
+      "epoch": 1.415388279187918,
+      "grad_norm": 0.8239719867706299,
+      "learning_rate": 2.6416342515413722e-05,
+      "loss": 0.7893,
+      "step": 17830
+    },
+    {
+      "epoch": 1.4161821032368176,
+      "grad_norm": 0.7520464062690735,
+      "learning_rate": 2.640311185202826e-05,
+      "loss": 0.8056,
+      "step": 17840
+    },
+    {
+      "epoch": 1.4169759272857172,
+      "grad_norm": 0.8690072894096375,
+      "learning_rate": 2.63898811886428e-05,
+      "loss": 0.8513,
+      "step": 17850
+    },
+    {
+      "epoch": 1.4177697513346166,
+      "grad_norm": 0.6335851550102234,
+      "learning_rate": 2.637665052525734e-05,
+      "loss": 0.8591,
+      "step": 17860
+    },
+    {
+      "epoch": 1.4185635753835162,
+      "grad_norm": 0.9497599005699158,
+      "learning_rate": 2.6363419861871875e-05,
+      "loss": 0.8655,
+      "step": 17870
+    },
+    {
+      "epoch": 1.4193573994324158,
+      "grad_norm": 0.7551174163818359,
+      "learning_rate": 2.6350189198486413e-05,
+      "loss": 0.7851,
+      "step": 17880
+    },
+    {
+      "epoch": 1.4201512234813154,
+      "grad_norm": 0.9593257904052734,
+      "learning_rate": 2.6336958535100952e-05,
+      "loss": 0.8049,
+      "step": 17890
+    },
+    {
+      "epoch": 1.420945047530215,
+      "grad_norm": 0.8998299837112427,
+      "learning_rate": 2.6323727871715488e-05,
+      "loss": 0.8266,
+      "step": 17900
+    },
+    {
+      "epoch": 1.4217388715791146,
+      "grad_norm": 0.8189812302589417,
+      "learning_rate": 2.6310497208330027e-05,
+      "loss": 0.8333,
+      "step": 17910
+    },
+    {
+      "epoch": 1.422532695628014,
+      "grad_norm": 0.7574518322944641,
+      "learning_rate": 2.6297266544944566e-05,
+      "loss": 0.8794,
+      "step": 17920
+    },
+    {
+      "epoch": 1.4233265196769136,
+      "grad_norm": 0.815168559551239,
+      "learning_rate": 2.6284035881559105e-05,
+      "loss": 0.8662,
+      "step": 17930
+    },
+    {
+      "epoch": 1.4241203437258132,
+      "grad_norm": 0.7286579012870789,
+      "learning_rate": 2.627080521817364e-05,
+      "loss": 0.8239,
+      "step": 17940
+    },
+    {
+      "epoch": 1.4249141677747128,
+      "grad_norm": 0.8906771540641785,
+      "learning_rate": 2.625757455478818e-05,
+      "loss": 0.8331,
+      "step": 17950
+    },
+    {
+      "epoch": 1.4257079918236122,
+      "grad_norm": 0.7556918859481812,
+      "learning_rate": 2.6244343891402718e-05,
+      "loss": 0.8623,
+      "step": 17960
+    },
+    {
+      "epoch": 1.4265018158725118,
+      "grad_norm": 0.8590714931488037,
+      "learning_rate": 2.6231113228017257e-05,
+      "loss": 0.7373,
+      "step": 17970
+    },
+    {
+      "epoch": 1.4272956399214114,
+      "grad_norm": 0.827892541885376,
+      "learning_rate": 2.621788256463179e-05,
+      "loss": 0.7714,
+      "step": 17980
+    },
+    {
+      "epoch": 1.428089463970311,
+      "grad_norm": 0.7814650535583496,
+      "learning_rate": 2.620465190124633e-05,
+      "loss": 0.8154,
+      "step": 17990
+    },
+    {
+      "epoch": 1.4288832880192106,
+      "grad_norm": 0.8106557130813599,
+      "learning_rate": 2.619142123786087e-05,
+      "loss": 0.7834,
+      "step": 18000
+    },
+    {
+      "epoch": 1.4296771120681102,
+      "grad_norm": 0.7437555193901062,
+      "learning_rate": 2.6178190574475403e-05,
+      "loss": 0.7773,
+      "step": 18010
+    },
+    {
+      "epoch": 1.4304709361170096,
+      "grad_norm": 0.9069659113883972,
+      "learning_rate": 2.616495991108994e-05,
+      "loss": 0.7767,
+      "step": 18020
+    },
+    {
+      "epoch": 1.4312647601659092,
+      "grad_norm": 0.8167043924331665,
+      "learning_rate": 2.615172924770448e-05,
+      "loss": 0.8079,
+      "step": 18030
+    },
+    {
+      "epoch": 1.4320585842148088,
+      "grad_norm": 0.8131256103515625,
+      "learning_rate": 2.6138498584319023e-05,
+      "loss": 0.806,
+      "step": 18040
+    },
+    {
+      "epoch": 1.4328524082637084,
+      "grad_norm": 0.8274345993995667,
+      "learning_rate": 2.6125267920933555e-05,
+      "loss": 0.8135,
+      "step": 18050
+    },
+    {
+      "epoch": 1.4336462323126078,
+      "grad_norm": 0.9551146626472473,
+      "learning_rate": 2.6112037257548094e-05,
+      "loss": 0.852,
+      "step": 18060
+    },
+    {
+      "epoch": 1.4344400563615074,
+      "grad_norm": 0.8088938593864441,
+      "learning_rate": 2.6098806594162633e-05,
+      "loss": 0.7876,
+      "step": 18070
+    },
+    {
+      "epoch": 1.435233880410407,
+      "grad_norm": 0.7884169816970825,
+      "learning_rate": 2.608557593077717e-05,
+      "loss": 0.7942,
+      "step": 18080
+    },
+    {
+      "epoch": 1.4360277044593066,
+      "grad_norm": 0.8303605914115906,
+      "learning_rate": 2.6072345267391707e-05,
+      "loss": 0.8792,
+      "step": 18090
+    },
+    {
+      "epoch": 1.4368215285082062,
+      "grad_norm": 0.75099778175354,
+      "learning_rate": 2.6059114604006246e-05,
+      "loss": 0.7786,
+      "step": 18100
+    },
+    {
+      "epoch": 1.4376153525571058,
+      "grad_norm": 0.7984867095947266,
+      "learning_rate": 2.6045883940620785e-05,
+      "loss": 0.786,
+      "step": 18110
+    },
+    {
+      "epoch": 1.4384091766060052,
+      "grad_norm": 0.7262865900993347,
+      "learning_rate": 2.603265327723532e-05,
+      "loss": 0.8207,
+      "step": 18120
+    },
+    {
+      "epoch": 1.4392030006549048,
+      "grad_norm": 0.8592838048934937,
+      "learning_rate": 2.601942261384986e-05,
+      "loss": 0.8127,
+      "step": 18130
+    },
+    {
+      "epoch": 1.4399968247038044,
+      "grad_norm": 0.6876896619796753,
+      "learning_rate": 2.60061919504644e-05,
+      "loss": 0.8638,
+      "step": 18140
+    },
+    {
+      "epoch": 1.440790648752704,
+      "grad_norm": 0.9342442154884338,
+      "learning_rate": 2.5992961287078937e-05,
+      "loss": 0.8032,
+      "step": 18150
+    },
+    {
+      "epoch": 1.4415844728016034,
+      "grad_norm": 0.9930727481842041,
+      "learning_rate": 2.597973062369347e-05,
+      "loss": 0.7582,
+      "step": 18160
+    },
+    {
+      "epoch": 1.442378296850503,
+      "grad_norm": 0.8912200331687927,
+      "learning_rate": 2.5966499960308012e-05,
+      "loss": 0.8152,
+      "step": 18170
+    },
+    {
+      "epoch": 1.4431721208994026,
+      "grad_norm": 0.7123863697052002,
+      "learning_rate": 2.595326929692255e-05,
+      "loss": 0.7756,
+      "step": 18180
+    },
+    {
+      "epoch": 1.4439659449483022,
+      "grad_norm": 0.8635473251342773,
+      "learning_rate": 2.594003863353709e-05,
+      "loss": 0.8072,
+      "step": 18190
+    },
+    {
+      "epoch": 1.4447597689972018,
+      "grad_norm": 0.7606240510940552,
+      "learning_rate": 2.5926807970151622e-05,
+      "loss": 0.8457,
+      "step": 18200
+    },
+    {
+      "epoch": 1.4455535930461014,
+      "grad_norm": 0.7793667912483215,
+      "learning_rate": 2.591357730676616e-05,
+      "loss": 0.8179,
+      "step": 18210
+    },
+    {
+      "epoch": 1.446347417095001,
+      "grad_norm": 0.6772461533546448,
+      "learning_rate": 2.5900346643380703e-05,
+      "loss": 0.8746,
+      "step": 18220
+    },
+    {
+      "epoch": 1.4471412411439004,
+      "grad_norm": 0.9796277284622192,
+      "learning_rate": 2.5887115979995235e-05,
+      "loss": 0.8008,
+      "step": 18230
+    },
+    {
+      "epoch": 1.4479350651928,
+      "grad_norm": 0.8049467206001282,
+      "learning_rate": 2.5873885316609774e-05,
+      "loss": 0.7848,
+      "step": 18240
+    },
+    {
+      "epoch": 1.4487288892416996,
+      "grad_norm": 0.8529013991355896,
+      "learning_rate": 2.5860654653224313e-05,
+      "loss": 0.7956,
+      "step": 18250
+    },
+    {
+      "epoch": 1.449522713290599,
+      "grad_norm": 1.007569432258606,
+      "learning_rate": 2.5847423989838852e-05,
+      "loss": 0.7736,
+      "step": 18260
+    },
+    {
+      "epoch": 1.4503165373394986,
+      "grad_norm": 0.8529340624809265,
+      "learning_rate": 2.5834193326453388e-05,
+      "loss": 0.8122,
+      "step": 18270
+    },
+    {
+      "epoch": 1.4511103613883982,
+      "grad_norm": 0.8169769644737244,
+      "learning_rate": 2.5820962663067927e-05,
+      "loss": 0.841,
+      "step": 18280
+    },
+    {
+      "epoch": 1.4519041854372978,
+      "grad_norm": 0.7556354999542236,
+      "learning_rate": 2.5807731999682465e-05,
+      "loss": 0.8136,
+      "step": 18290
+    },
+    {
+      "epoch": 1.4526980094861974,
+      "grad_norm": 0.867939293384552,
+      "learning_rate": 2.5794501336297004e-05,
+      "loss": 0.8153,
+      "step": 18300
+    },
+    {
+      "epoch": 1.453491833535097,
+      "grad_norm": 0.9463819861412048,
+      "learning_rate": 2.578127067291154e-05,
+      "loss": 0.8175,
+      "step": 18310
+    },
+    {
+      "epoch": 1.4542856575839966,
+      "grad_norm": 0.9238904714584351,
+      "learning_rate": 2.576804000952608e-05,
+      "loss": 0.7291,
+      "step": 18320
+    },
+    {
+      "epoch": 1.455079481632896,
+      "grad_norm": 0.8403037786483765,
+      "learning_rate": 2.5754809346140618e-05,
+      "loss": 0.864,
+      "step": 18330
+    },
+    {
+      "epoch": 1.4558733056817956,
+      "grad_norm": 0.8228583931922913,
+      "learning_rate": 2.5741578682755153e-05,
+      "loss": 0.8357,
+      "step": 18340
+    },
+    {
+      "epoch": 1.4566671297306952,
+      "grad_norm": 0.730689287185669,
+      "learning_rate": 2.5728348019369692e-05,
+      "loss": 0.8575,
+      "step": 18350
+    },
+    {
+      "epoch": 1.4574609537795946,
+      "grad_norm": 0.7443234920501709,
+      "learning_rate": 2.571511735598423e-05,
+      "loss": 0.8414,
+      "step": 18360
+    },
+    {
+      "epoch": 1.4582547778284942,
+      "grad_norm": 0.9750908017158508,
+      "learning_rate": 2.570188669259877e-05,
+      "loss": 0.8359,
+      "step": 18370
+    },
+    {
+      "epoch": 1.4590486018773938,
+      "grad_norm": 0.9453748464584351,
+      "learning_rate": 2.5688656029213302e-05,
+      "loss": 0.7975,
+      "step": 18380
+    },
+    {
+      "epoch": 1.4598424259262934,
+      "grad_norm": 0.826844334602356,
+      "learning_rate": 2.5675425365827845e-05,
+      "loss": 0.7952,
+      "step": 18390
+    },
+    {
+      "epoch": 1.460636249975193,
+      "grad_norm": 0.937788724899292,
+      "learning_rate": 2.5662194702442383e-05,
+      "loss": 0.8504,
+      "step": 18400
+    },
+    {
+      "epoch": 1.4614300740240926,
+      "grad_norm": 0.9471660852432251,
+      "learning_rate": 2.5648964039056922e-05,
+      "loss": 0.8499,
+      "step": 18410
+    },
+    {
+      "epoch": 1.4622238980729922,
+      "grad_norm": 0.8096206784248352,
+      "learning_rate": 2.5635733375671455e-05,
+      "loss": 0.8297,
+      "step": 18420
+    },
+    {
+      "epoch": 1.4630177221218916,
+      "grad_norm": 1.0029072761535645,
+      "learning_rate": 2.5622502712285993e-05,
+      "loss": 0.8496,
+      "step": 18430
+    },
+    {
+      "epoch": 1.4638115461707912,
+      "grad_norm": 0.9172188639640808,
+      "learning_rate": 2.5609272048900536e-05,
+      "loss": 0.823,
+      "step": 18440
+    },
+    {
+      "epoch": 1.4646053702196908,
+      "grad_norm": 0.8738644123077393,
+      "learning_rate": 2.5596041385515068e-05,
+      "loss": 0.8175,
+      "step": 18450
+    },
+    {
+      "epoch": 1.4653991942685904,
+      "grad_norm": 0.7507184743881226,
+      "learning_rate": 2.5582810722129607e-05,
+      "loss": 0.7805,
+      "step": 18460
+    },
+    {
+      "epoch": 1.4661930183174898,
+      "grad_norm": 0.8251667022705078,
+      "learning_rate": 2.5569580058744146e-05,
+      "loss": 0.8151,
+      "step": 18470
+    },
+    {
+      "epoch": 1.4669868423663894,
+      "grad_norm": 0.9038745164871216,
+      "learning_rate": 2.5556349395358685e-05,
+      "loss": 0.8222,
+      "step": 18480
+    },
+    {
+      "epoch": 1.467780666415289,
+      "grad_norm": 0.7549251914024353,
+      "learning_rate": 2.554311873197322e-05,
+      "loss": 0.7928,
+      "step": 18490
+    },
+    {
+      "epoch": 1.4685744904641886,
+      "grad_norm": 0.8208056688308716,
+      "learning_rate": 2.552988806858776e-05,
+      "loss": 0.8973,
+      "step": 18500
+    },
+    {
+      "epoch": 1.4693683145130882,
+      "grad_norm": 1.0672497749328613,
+      "learning_rate": 2.5516657405202298e-05,
+      "loss": 0.8407,
+      "step": 18510
+    },
+    {
+      "epoch": 1.4701621385619879,
+      "grad_norm": 0.8073993921279907,
+      "learning_rate": 2.5503426741816837e-05,
+      "loss": 0.8038,
+      "step": 18520
+    },
+    {
+      "epoch": 1.4709559626108872,
+      "grad_norm": 0.9057318568229675,
+      "learning_rate": 2.5490196078431373e-05,
+      "loss": 0.829,
+      "step": 18530
+    },
+    {
+      "epoch": 1.4717497866597868,
+      "grad_norm": 0.7318175435066223,
+      "learning_rate": 2.547696541504591e-05,
+      "loss": 0.8253,
+      "step": 18540
+    },
+    {
+      "epoch": 1.4725436107086864,
+      "grad_norm": 0.819756805896759,
+      "learning_rate": 2.546373475166045e-05,
+      "loss": 0.8265,
+      "step": 18550
+    },
+    {
+      "epoch": 1.473337434757586,
+      "grad_norm": 0.852188766002655,
+      "learning_rate": 2.5450504088274986e-05,
+      "loss": 0.7702,
+      "step": 18560
+    },
+    {
+      "epoch": 1.4741312588064854,
+      "grad_norm": 0.8986577391624451,
+      "learning_rate": 2.5437273424889525e-05,
+      "loss": 0.8488,
+      "step": 18570
+    },
+    {
+      "epoch": 1.474925082855385,
+      "grad_norm": 0.6667974591255188,
+      "learning_rate": 2.5424042761504064e-05,
+      "loss": 0.8424,
+      "step": 18580
+    },
+    {
+      "epoch": 1.4757189069042846,
+      "grad_norm": 0.7348021864891052,
+      "learning_rate": 2.5410812098118603e-05,
+      "loss": 0.7877,
+      "step": 18590
+    },
+    {
+      "epoch": 1.4765127309531842,
+      "grad_norm": 0.7171440720558167,
+      "learning_rate": 2.5397581434733135e-05,
+      "loss": 0.8142,
+      "step": 18600
+    },
+    {
+      "epoch": 1.4773065550020839,
+      "grad_norm": 0.7128421068191528,
+      "learning_rate": 2.5384350771347677e-05,
+      "loss": 0.8373,
+      "step": 18610
+    },
+    {
+      "epoch": 1.4781003790509835,
+      "grad_norm": 0.7781753540039062,
+      "learning_rate": 2.5371120107962216e-05,
+      "loss": 0.7994,
+      "step": 18620
+    },
+    {
+      "epoch": 1.4788942030998828,
+      "grad_norm": 0.8336313366889954,
+      "learning_rate": 2.5357889444576755e-05,
+      "loss": 0.8565,
+      "step": 18630
+    },
+    {
+      "epoch": 1.4796880271487824,
+      "grad_norm": 0.9292792677879333,
+      "learning_rate": 2.5344658781191287e-05,
+      "loss": 0.7621,
+      "step": 18640
+    },
+    {
+      "epoch": 1.480481851197682,
+      "grad_norm": 0.81519615650177,
+      "learning_rate": 2.5331428117805826e-05,
+      "loss": 0.8161,
+      "step": 18650
+    },
+    {
+      "epoch": 1.4812756752465817,
+      "grad_norm": 0.7871677279472351,
+      "learning_rate": 2.531819745442037e-05,
+      "loss": 0.8304,
+      "step": 18660
+    },
+    {
+      "epoch": 1.482069499295481,
+      "grad_norm": 0.8033188581466675,
+      "learning_rate": 2.5304966791034907e-05,
+      "loss": 0.859,
+      "step": 18670
+    },
+    {
+      "epoch": 1.4828633233443806,
+      "grad_norm": 0.8543998003005981,
+      "learning_rate": 2.529173612764944e-05,
+      "loss": 0.8521,
+      "step": 18680
+    },
+    {
+      "epoch": 1.4836571473932803,
+      "grad_norm": 0.7945504188537598,
+      "learning_rate": 2.527850546426398e-05,
+      "loss": 0.8477,
+      "step": 18690
+    },
+    {
+      "epoch": 1.4844509714421799,
+      "grad_norm": 0.9741247892379761,
+      "learning_rate": 2.5265274800878517e-05,
+      "loss": 0.8306,
+      "step": 18700
+    },
+    {
+      "epoch": 1.4852447954910795,
+      "grad_norm": 1.0190260410308838,
+      "learning_rate": 2.5252044137493053e-05,
+      "loss": 0.7352,
+      "step": 18710
+    },
+    {
+      "epoch": 1.486038619539979,
+      "grad_norm": 0.8218889832496643,
+      "learning_rate": 2.5238813474107592e-05,
+      "loss": 0.8306,
+      "step": 18720
+    },
+    {
+      "epoch": 1.4868324435888784,
+      "grad_norm": 0.9101859331130981,
+      "learning_rate": 2.522558281072213e-05,
+      "loss": 0.8644,
+      "step": 18730
+    },
+    {
+      "epoch": 1.487626267637778,
+      "grad_norm": 0.8741206526756287,
+      "learning_rate": 2.521235214733667e-05,
+      "loss": 0.811,
+      "step": 18740
+    },
+    {
+      "epoch": 1.4884200916866777,
+      "grad_norm": 0.7072809338569641,
+      "learning_rate": 2.5199121483951205e-05,
+      "loss": 0.8437,
+      "step": 18750
+    },
+    {
+      "epoch": 1.4892139157355773,
+      "grad_norm": 0.8772575855255127,
+      "learning_rate": 2.5185890820565744e-05,
+      "loss": 0.7746,
+      "step": 18760
+    },
+    {
+      "epoch": 1.4900077397844766,
+      "grad_norm": 0.8278160095214844,
+      "learning_rate": 2.5172660157180283e-05,
+      "loss": 0.8073,
+      "step": 18770
+    },
+    {
+      "epoch": 1.4908015638333763,
+      "grad_norm": 0.8926985859870911,
+      "learning_rate": 2.5159429493794822e-05,
+      "loss": 0.8103,
+      "step": 18780
+    },
+    {
+      "epoch": 1.4915953878822759,
+      "grad_norm": 1.0303518772125244,
+      "learning_rate": 2.5146198830409358e-05,
+      "loss": 0.8209,
+      "step": 18790
+    },
+    {
+      "epoch": 1.4923892119311755,
+      "grad_norm": 0.7462199926376343,
+      "learning_rate": 2.5132968167023897e-05,
+      "loss": 0.7996,
+      "step": 18800
+    },
+    {
+      "epoch": 1.493183035980075,
+      "grad_norm": 0.8517622351646423,
+      "learning_rate": 2.5119737503638435e-05,
+      "loss": 0.7538,
+      "step": 18810
+    },
+    {
+      "epoch": 1.4939768600289747,
+      "grad_norm": 0.8433437347412109,
+      "learning_rate": 2.5106506840252968e-05,
+      "loss": 0.8481,
+      "step": 18820
+    },
+    {
+      "epoch": 1.494770684077874,
+      "grad_norm": 0.8017115592956543,
+      "learning_rate": 2.509327617686751e-05,
+      "loss": 0.7835,
+      "step": 18830
+    },
+    {
+      "epoch": 1.4955645081267737,
+      "grad_norm": 0.8956550359725952,
+      "learning_rate": 2.508004551348205e-05,
+      "loss": 0.746,
+      "step": 18840
+    },
+    {
+      "epoch": 1.4963583321756733,
+      "grad_norm": 0.8071773648262024,
+      "learning_rate": 2.5066814850096588e-05,
+      "loss": 0.8029,
+      "step": 18850
+    },
+    {
+      "epoch": 1.4971521562245729,
+      "grad_norm": 0.8614389300346375,
+      "learning_rate": 2.505358418671112e-05,
+      "loss": 0.8419,
+      "step": 18860
+    },
+    {
+      "epoch": 1.4979459802734723,
+      "grad_norm": 0.8073408007621765,
+      "learning_rate": 2.504035352332566e-05,
+      "loss": 0.8295,
+      "step": 18870
+    },
+    {
+      "epoch": 1.4987398043223719,
+      "grad_norm": 0.8368566036224365,
+      "learning_rate": 2.50271228599402e-05,
+      "loss": 0.8155,
+      "step": 18880
+    },
+    {
+      "epoch": 1.4995336283712715,
+      "grad_norm": 0.7750986814498901,
+      "learning_rate": 2.501389219655474e-05,
+      "loss": 0.7991,
+      "step": 18890
+    },
+    {
+      "epoch": 1.500327452420171,
+      "grad_norm": 0.843612015247345,
+      "learning_rate": 2.5000661533169272e-05,
+      "loss": 0.8506,
+      "step": 18900
+    },
+    {
+      "epoch": 1.5011212764690707,
+      "grad_norm": 0.9953761100769043,
+      "learning_rate": 2.498743086978381e-05,
+      "loss": 0.8303,
+      "step": 18910
+    },
+    {
+      "epoch": 1.5019151005179703,
+      "grad_norm": 0.8304370641708374,
+      "learning_rate": 2.497420020639835e-05,
+      "loss": 0.7987,
+      "step": 18920
+    },
+    {
+      "epoch": 1.5027089245668699,
+      "grad_norm": 0.6083853840827942,
+      "learning_rate": 2.496096954301289e-05,
+      "loss": 0.7744,
+      "step": 18930
+    },
+    {
+      "epoch": 1.5035027486157693,
+      "grad_norm": 0.8940452933311462,
+      "learning_rate": 2.4947738879627425e-05,
+      "loss": 0.7647,
+      "step": 18940
+    },
+    {
+      "epoch": 1.5042965726646689,
+      "grad_norm": 0.723276674747467,
+      "learning_rate": 2.4934508216241963e-05,
+      "loss": 0.8258,
+      "step": 18950
+    },
+    {
+      "epoch": 1.5050903967135685,
+      "grad_norm": 0.8840285539627075,
+      "learning_rate": 2.49212775528565e-05,
+      "loss": 0.8022,
+      "step": 18960
+    },
+    {
+      "epoch": 1.5058842207624679,
+      "grad_norm": 0.775147020816803,
+      "learning_rate": 2.490804688947104e-05,
+      "loss": 0.8434,
+      "step": 18970
+    },
+    {
+      "epoch": 1.5066780448113675,
+      "grad_norm": 0.9234639406204224,
+      "learning_rate": 2.4894816226085577e-05,
+      "loss": 0.8079,
+      "step": 18980
+    },
+    {
+      "epoch": 1.507471868860267,
+      "grad_norm": 0.7386789321899414,
+      "learning_rate": 2.4881585562700116e-05,
+      "loss": 0.8194,
+      "step": 18990
+    },
+    {
+      "epoch": 1.5082656929091667,
+      "grad_norm": 0.9541153311729431,
+      "learning_rate": 2.486835489931465e-05,
+      "loss": 0.7625,
+      "step": 19000
+    },
+    {
+      "epoch": 1.5090595169580663,
+      "grad_norm": 0.9702879190444946,
+      "learning_rate": 2.485512423592919e-05,
+      "loss": 0.8113,
+      "step": 19010
+    },
+    {
+      "epoch": 1.5098533410069659,
+      "grad_norm": 0.7624096870422363,
+      "learning_rate": 2.484189357254373e-05,
+      "loss": 0.8299,
+      "step": 19020
+    },
+    {
+      "epoch": 1.5106471650558655,
+      "grad_norm": 0.734418511390686,
+      "learning_rate": 2.4828662909158265e-05,
+      "loss": 0.8394,
+      "step": 19030
+    },
+    {
+      "epoch": 1.5114409891047649,
+      "grad_norm": 0.8682621717453003,
+      "learning_rate": 2.4815432245772804e-05,
+      "loss": 0.8202,
+      "step": 19040
+    },
+    {
+      "epoch": 1.5122348131536645,
+      "grad_norm": 0.7648998498916626,
+      "learning_rate": 2.4802201582387343e-05,
+      "loss": 0.8419,
+      "step": 19050
+    },
+    {
+      "epoch": 1.513028637202564,
+      "grad_norm": 0.702462911605835,
+      "learning_rate": 2.478897091900188e-05,
+      "loss": 0.8125,
+      "step": 19060
+    },
+    {
+      "epoch": 1.5138224612514635,
+      "grad_norm": 0.7904557585716248,
+      "learning_rate": 2.4775740255616417e-05,
+      "loss": 0.8679,
+      "step": 19070
+    },
+    {
+      "epoch": 1.514616285300363,
+      "grad_norm": 1.0039995908737183,
+      "learning_rate": 2.4762509592230956e-05,
+      "loss": 0.8623,
+      "step": 19080
+    },
+    {
+      "epoch": 1.5154101093492627,
+      "grad_norm": 0.8769594430923462,
+      "learning_rate": 2.474927892884549e-05,
+      "loss": 0.8309,
+      "step": 19090
+    },
+    {
+      "epoch": 1.5162039333981623,
+      "grad_norm": 0.742843508720398,
+      "learning_rate": 2.473737133179858e-05,
+      "loss": 0.8068,
+      "step": 19100
+    },
+    {
+      "epoch": 1.516997757447062,
+      "grad_norm": 0.7758870720863342,
+      "learning_rate": 2.4724140668413115e-05,
+      "loss": 0.8617,
+      "step": 19110
+    },
+    {
+      "epoch": 1.5177915814959615,
+      "grad_norm": 0.8473050594329834,
+      "learning_rate": 2.4710910005027654e-05,
+      "loss": 0.8175,
+      "step": 19120
+    },
+    {
+      "epoch": 1.518585405544861,
+      "grad_norm": 0.6727568507194519,
+      "learning_rate": 2.469767934164219e-05,
+      "loss": 0.8389,
+      "step": 19130
+    },
+    {
+      "epoch": 1.5193792295937607,
+      "grad_norm": 0.7093884944915771,
+      "learning_rate": 2.468444867825673e-05,
+      "loss": 0.8226,
+      "step": 19140
+    },
+    {
+      "epoch": 1.52017305364266,
+      "grad_norm": 0.8359051942825317,
+      "learning_rate": 2.4671218014871267e-05,
+      "loss": 0.7772,
+      "step": 19150
+    },
+    {
+      "epoch": 1.5209668776915597,
+      "grad_norm": 0.6940776705741882,
+      "learning_rate": 2.4657987351485803e-05,
+      "loss": 0.7137,
+      "step": 19160
+    },
+    {
+      "epoch": 1.521760701740459,
+      "grad_norm": 0.8422114253044128,
+      "learning_rate": 2.464475668810034e-05,
+      "loss": 0.8425,
+      "step": 19170
+    },
+    {
+      "epoch": 1.5225545257893587,
+      "grad_norm": 1.108744740486145,
+      "learning_rate": 2.463152602471488e-05,
+      "loss": 0.8392,
+      "step": 19180
+    },
+    {
+      "epoch": 1.5233483498382583,
+      "grad_norm": 0.9370563626289368,
+      "learning_rate": 2.461829536132942e-05,
+      "loss": 0.8278,
+      "step": 19190
+    },
+    {
+      "epoch": 1.524142173887158,
+      "grad_norm": 0.8219224214553833,
+      "learning_rate": 2.4605064697943955e-05,
+      "loss": 0.7916,
+      "step": 19200
+    },
+    {
+      "epoch": 1.5249359979360575,
+      "grad_norm": 0.8902950882911682,
+      "learning_rate": 2.4591834034558494e-05,
+      "loss": 0.8494,
+      "step": 19210
+    },
+    {
+      "epoch": 1.525729821984957,
+      "grad_norm": 0.884885311126709,
+      "learning_rate": 2.4578603371173033e-05,
+      "loss": 0.8258,
+      "step": 19220
+    },
+    {
+      "epoch": 1.5265236460338567,
+      "grad_norm": 0.6714193224906921,
+      "learning_rate": 2.4565372707787572e-05,
+      "loss": 0.8377,
+      "step": 19230
+    },
+    {
+      "epoch": 1.5273174700827563,
+      "grad_norm": 0.9097526669502258,
+      "learning_rate": 2.4552142044402107e-05,
+      "loss": 0.8093,
+      "step": 19240
+    },
+    {
+      "epoch": 1.5281112941316557,
+      "grad_norm": 0.6362065672874451,
+      "learning_rate": 2.4538911381016646e-05,
+      "loss": 0.8245,
+      "step": 19250
+    },
+    {
+      "epoch": 1.5289051181805553,
+      "grad_norm": 0.7581408619880676,
+      "learning_rate": 2.4525680717631182e-05,
+      "loss": 0.8677,
+      "step": 19260
+    },
+    {
+      "epoch": 1.5296989422294547,
+      "grad_norm": 0.7132633328437805,
+      "learning_rate": 2.451245005424572e-05,
+      "loss": 0.8338,
+      "step": 19270
+    },
+    {
+      "epoch": 1.5304927662783543,
+      "grad_norm": 0.8469735383987427,
+      "learning_rate": 2.449921939086026e-05,
+      "loss": 0.9025,
+      "step": 19280
+    },
+    {
+      "epoch": 1.531286590327254,
+      "grad_norm": 0.8352085947990417,
+      "learning_rate": 2.4485988727474795e-05,
+      "loss": 0.8216,
+      "step": 19290
+    },
+    {
+      "epoch": 1.5320804143761535,
+      "grad_norm": 0.7701964378356934,
+      "learning_rate": 2.4472758064089334e-05,
+      "loss": 0.7969,
+      "step": 19300
+    },
+    {
+      "epoch": 1.532874238425053,
+      "grad_norm": 0.8008362054824829,
+      "learning_rate": 2.4459527400703873e-05,
+      "loss": 0.781,
+      "step": 19310
+    },
+    {
+      "epoch": 1.5336680624739527,
+      "grad_norm": 0.7611208558082581,
+      "learning_rate": 2.4446296737318412e-05,
+      "loss": 0.7799,
+      "step": 19320
+    },
+    {
+      "epoch": 1.5344618865228523,
+      "grad_norm": 0.7576231360435486,
+      "learning_rate": 2.4433066073932947e-05,
+      "loss": 0.8071,
+      "step": 19330
+    },
+    {
+      "epoch": 1.535255710571752,
+      "grad_norm": 0.9794348478317261,
+      "learning_rate": 2.4419835410547486e-05,
+      "loss": 0.8337,
+      "step": 19340
+    },
+    {
+      "epoch": 1.5360495346206513,
+      "grad_norm": 0.8349586725234985,
+      "learning_rate": 2.4406604747162022e-05,
+      "loss": 0.7267,
+      "step": 19350
+    },
+    {
+      "epoch": 1.536843358669551,
+      "grad_norm": 0.794275164604187,
+      "learning_rate": 2.4393374083776564e-05,
+      "loss": 0.7525,
+      "step": 19360
+    },
+    {
+      "epoch": 1.5376371827184503,
+      "grad_norm": 0.9255030155181885,
+      "learning_rate": 2.43801434203911e-05,
+      "loss": 0.7845,
+      "step": 19370
+    },
+    {
+      "epoch": 1.53843100676735,
+      "grad_norm": 0.7629058957099915,
+      "learning_rate": 2.436691275700564e-05,
+      "loss": 0.7849,
+      "step": 19380
+    },
+    {
+      "epoch": 1.5392248308162495,
+      "grad_norm": 0.8714119791984558,
+      "learning_rate": 2.4353682093620174e-05,
+      "loss": 0.8106,
+      "step": 19390
+    },
+    {
+      "epoch": 1.540018654865149,
+      "grad_norm": 0.9720879793167114,
+      "learning_rate": 2.4340451430234713e-05,
+      "loss": 0.868,
+      "step": 19400
+    },
+    {
+      "epoch": 1.5408124789140487,
+      "grad_norm": 0.848136842250824,
+      "learning_rate": 2.4327220766849252e-05,
+      "loss": 0.8204,
+      "step": 19410
+    },
+    {
+      "epoch": 1.5416063029629483,
+      "grad_norm": 0.8804022073745728,
+      "learning_rate": 2.4313990103463788e-05,
+      "loss": 0.8423,
+      "step": 19420
+    },
+    {
+      "epoch": 1.542400127011848,
+      "grad_norm": 0.7391209602355957,
+      "learning_rate": 2.4300759440078327e-05,
+      "loss": 0.8378,
+      "step": 19430
+    },
+    {
+      "epoch": 1.5431939510607475,
+      "grad_norm": 0.6705385446548462,
+      "learning_rate": 2.4287528776692866e-05,
+      "loss": 0.7564,
+      "step": 19440
+    },
+    {
+      "epoch": 1.543987775109647,
+      "grad_norm": 0.8526281118392944,
+      "learning_rate": 2.4274298113307404e-05,
+      "loss": 0.7472,
+      "step": 19450
+    },
+    {
+      "epoch": 1.5447815991585465,
+      "grad_norm": 0.7899938225746155,
+      "learning_rate": 2.426106744992194e-05,
+      "loss": 0.8292,
+      "step": 19460
+    },
+    {
+      "epoch": 1.5455754232074461,
+      "grad_norm": 0.9177964925765991,
+      "learning_rate": 2.424783678653648e-05,
+      "loss": 0.829,
+      "step": 19470
+    },
+    {
+      "epoch": 1.5463692472563455,
+      "grad_norm": 0.8367957472801208,
+      "learning_rate": 2.4234606123151014e-05,
+      "loss": 0.8395,
+      "step": 19480
+    },
+    {
+      "epoch": 1.547163071305245,
+      "grad_norm": 0.8440412878990173,
+      "learning_rate": 2.4221375459765557e-05,
+      "loss": 0.7762,
+      "step": 19490
+    },
+    {
+      "epoch": 1.5479568953541447,
+      "grad_norm": 0.8187562823295593,
+      "learning_rate": 2.4208144796380092e-05,
+      "loss": 0.8345,
+      "step": 19500
+    },
+    {
+      "epoch": 1.5487507194030443,
+      "grad_norm": 0.8997789025306702,
+      "learning_rate": 2.4194914132994628e-05,
+      "loss": 0.8156,
+      "step": 19510
+    },
+    {
+      "epoch": 1.549544543451944,
+      "grad_norm": 0.8826847672462463,
+      "learning_rate": 2.4181683469609167e-05,
+      "loss": 0.8558,
+      "step": 19520
+    },
+    {
+      "epoch": 1.5503383675008435,
+      "grad_norm": 0.8336629867553711,
+      "learning_rate": 2.4168452806223706e-05,
+      "loss": 0.9098,
+      "step": 19530
+    },
+    {
+      "epoch": 1.5511321915497431,
+      "grad_norm": 0.9518599510192871,
+      "learning_rate": 2.4155222142838245e-05,
+      "loss": 0.7868,
+      "step": 19540
+    },
+    {
+      "epoch": 1.5519260155986425,
+      "grad_norm": 0.6656529903411865,
+      "learning_rate": 2.414199147945278e-05,
+      "loss": 0.8164,
+      "step": 19550
+    },
+    {
+      "epoch": 1.5527198396475421,
+      "grad_norm": 0.8456233739852905,
+      "learning_rate": 2.412876081606732e-05,
+      "loss": 0.7531,
+      "step": 19560
+    },
+    {
+      "epoch": 1.5535136636964417,
+      "grad_norm": 0.8264173269271851,
+      "learning_rate": 2.4115530152681855e-05,
+      "loss": 0.8304,
+      "step": 19570
+    },
+    {
+      "epoch": 1.554307487745341,
+      "grad_norm": 0.9435857534408569,
+      "learning_rate": 2.4102299489296397e-05,
+      "loss": 0.8006,
+      "step": 19580
+    },
+    {
+      "epoch": 1.5551013117942407,
+      "grad_norm": 0.7315828204154968,
+      "learning_rate": 2.4089068825910932e-05,
+      "loss": 0.8037,
+      "step": 19590
+    },
+    {
+      "epoch": 1.5558951358431403,
+      "grad_norm": 0.8221999406814575,
+      "learning_rate": 2.407583816252547e-05,
+      "loss": 0.8318,
+      "step": 19600
+    },
+    {
+      "epoch": 1.55668895989204,
+      "grad_norm": 0.7483288645744324,
+      "learning_rate": 2.4062607499140007e-05,
+      "loss": 0.8202,
+      "step": 19610
+    },
+    {
+      "epoch": 1.5574827839409395,
+      "grad_norm": 0.7994224429130554,
+      "learning_rate": 2.4049376835754546e-05,
+      "loss": 0.7977,
+      "step": 19620
+    },
+    {
+      "epoch": 1.5582766079898391,
+      "grad_norm": 0.7807520031929016,
+      "learning_rate": 2.4036146172369085e-05,
+      "loss": 0.7979,
+      "step": 19630
+    },
+    {
+      "epoch": 1.5590704320387387,
+      "grad_norm": 0.771969199180603,
+      "learning_rate": 2.402291550898362e-05,
+      "loss": 0.8071,
+      "step": 19640
+    },
+    {
+      "epoch": 1.5598642560876381,
+      "grad_norm": 0.7460570931434631,
+      "learning_rate": 2.400968484559816e-05,
+      "loss": 0.7829,
+      "step": 19650
+    },
+    {
+      "epoch": 1.5606580801365377,
+      "grad_norm": 0.737625241279602,
+      "learning_rate": 2.3996454182212695e-05,
+      "loss": 0.8223,
+      "step": 19660
+    },
+    {
+      "epoch": 1.5614519041854373,
+      "grad_norm": 0.9518927335739136,
+      "learning_rate": 2.3983223518827237e-05,
+      "loss": 0.7475,
+      "step": 19670
+    },
+    {
+      "epoch": 1.5622457282343367,
+      "grad_norm": 0.719597578048706,
+      "learning_rate": 2.3969992855441773e-05,
+      "loss": 0.8264,
+      "step": 19680
+    },
+    {
+      "epoch": 1.5630395522832363,
+      "grad_norm": 0.8887612223625183,
+      "learning_rate": 2.395676219205631e-05,
+      "loss": 0.7934,
+      "step": 19690
+    },
+    {
+      "epoch": 1.563833376332136,
+      "grad_norm": 0.801966667175293,
+      "learning_rate": 2.3943531528670847e-05,
+      "loss": 0.7757,
+      "step": 19700
+    },
+    {
+      "epoch": 1.5646272003810355,
+      "grad_norm": 0.9171623587608337,
+      "learning_rate": 2.3930300865285386e-05,
+      "loss": 0.7704,
+      "step": 19710
+    },
+    {
+      "epoch": 1.5654210244299351,
+      "grad_norm": 0.8287226557731628,
+      "learning_rate": 2.3917070201899925e-05,
+      "loss": 0.8069,
+      "step": 19720
+    },
+    {
+      "epoch": 1.5662148484788347,
+      "grad_norm": 0.8021081686019897,
+      "learning_rate": 2.390383953851446e-05,
+      "loss": 0.8093,
+      "step": 19730
+    },
+    {
+      "epoch": 1.5670086725277343,
+      "grad_norm": 0.8900328874588013,
+      "learning_rate": 2.3890608875129e-05,
+      "loss": 0.7636,
+      "step": 19740
+    },
+    {
+      "epoch": 1.567802496576634,
+      "grad_norm": 0.9409016966819763,
+      "learning_rate": 2.387737821174354e-05,
+      "loss": 0.7915,
+      "step": 19750
+    },
+    {
+      "epoch": 1.5685963206255333,
+      "grad_norm": 0.8233746290206909,
+      "learning_rate": 2.3864147548358077e-05,
+      "loss": 0.8201,
+      "step": 19760
+    },
+    {
+      "epoch": 1.569390144674433,
+      "grad_norm": 0.8792815804481506,
+      "learning_rate": 2.3850916884972613e-05,
+      "loss": 0.7767,
+      "step": 19770
+    },
+    {
+      "epoch": 1.5701839687233323,
+      "grad_norm": 0.8257175087928772,
+      "learning_rate": 2.3837686221587152e-05,
+      "loss": 0.7879,
+      "step": 19780
+    },
+    {
+      "epoch": 1.570977792772232,
+      "grad_norm": 0.9405742883682251,
+      "learning_rate": 2.3824455558201687e-05,
+      "loss": 0.7247,
+      "step": 19790
+    },
+    {
+      "epoch": 1.5717716168211315,
+      "grad_norm": 0.8703904747962952,
+      "learning_rate": 2.381122489481623e-05,
+      "loss": 0.8193,
+      "step": 19800
+    },
+    {
+      "epoch": 1.5725654408700311,
+      "grad_norm": 0.9303200244903564,
+      "learning_rate": 2.3797994231430765e-05,
+      "loss": 0.7814,
+      "step": 19810
+    },
+    {
+      "epoch": 1.5733592649189307,
+      "grad_norm": 0.8841179609298706,
+      "learning_rate": 2.3784763568045304e-05,
+      "loss": 0.8572,
+      "step": 19820
+    },
+    {
+      "epoch": 1.5741530889678303,
+      "grad_norm": 0.7013158798217773,
+      "learning_rate": 2.377153290465984e-05,
+      "loss": 0.8371,
+      "step": 19830
+    },
+    {
+      "epoch": 1.57494691301673,
+      "grad_norm": 0.8166126012802124,
+      "learning_rate": 2.375830224127438e-05,
+      "loss": 0.7561,
+      "step": 19840
+    },
+    {
+      "epoch": 1.5757407370656296,
+      "grad_norm": 0.8469328284263611,
+      "learning_rate": 2.3745071577888917e-05,
+      "loss": 0.8203,
+      "step": 19850
+    },
+    {
+      "epoch": 1.576534561114529,
+      "grad_norm": 0.90618497133255,
+      "learning_rate": 2.3731840914503453e-05,
+      "loss": 0.808,
+      "step": 19860
+    },
+    {
+      "epoch": 1.5773283851634285,
+      "grad_norm": 0.9630061984062195,
+      "learning_rate": 2.3718610251117992e-05,
+      "loss": 0.7956,
+      "step": 19870
+    },
+    {
+      "epoch": 1.578122209212328,
+      "grad_norm": 0.8445176482200623,
+      "learning_rate": 2.3705379587732527e-05,
+      "loss": 0.8172,
+      "step": 19880
+    },
+    {
+      "epoch": 1.5789160332612275,
+      "grad_norm": 0.834750235080719,
+      "learning_rate": 2.369214892434707e-05,
+      "loss": 0.9094,
+      "step": 19890
+    },
+    {
+      "epoch": 1.5797098573101271,
+      "grad_norm": 0.8799012303352356,
+      "learning_rate": 2.3678918260961605e-05,
+      "loss": 0.8587,
+      "step": 19900
+    },
+    {
+      "epoch": 1.5805036813590267,
+      "grad_norm": 0.9250380992889404,
+      "learning_rate": 2.3665687597576144e-05,
+      "loss": 0.8404,
+      "step": 19910
+    },
+    {
+      "epoch": 1.5812975054079264,
+      "grad_norm": 0.9735758304595947,
+      "learning_rate": 2.365245693419068e-05,
+      "loss": 0.8144,
+      "step": 19920
+    },
+    {
+      "epoch": 1.582091329456826,
+      "grad_norm": 0.7764338850975037,
+      "learning_rate": 2.363922627080522e-05,
+      "loss": 0.7696,
+      "step": 19930
+    },
+    {
+      "epoch": 1.5828851535057256,
+      "grad_norm": 0.7918556332588196,
+      "learning_rate": 2.3625995607419758e-05,
+      "loss": 0.7751,
+      "step": 19940
+    },
+    {
+      "epoch": 1.5836789775546252,
+      "grad_norm": 0.980975866317749,
+      "learning_rate": 2.3612764944034297e-05,
+      "loss": 0.8112,
+      "step": 19950
+    },
+    {
+      "epoch": 1.5844728016035245,
+      "grad_norm": 0.6612667441368103,
+      "learning_rate": 2.3599534280648832e-05,
+      "loss": 0.7655,
+      "step": 19960
+    },
+    {
+      "epoch": 1.5852666256524242,
+      "grad_norm": 0.7851887941360474,
+      "learning_rate": 2.358630361726337e-05,
+      "loss": 0.7865,
+      "step": 19970
+    },
+    {
+      "epoch": 1.5860604497013235,
+      "grad_norm": 0.8260042071342468,
+      "learning_rate": 2.357307295387791e-05,
+      "loss": 0.7995,
+      "step": 19980
+    },
+    {
+      "epoch": 1.5868542737502231,
+      "grad_norm": 0.918483555316925,
+      "learning_rate": 2.3559842290492446e-05,
+      "loss": 0.7792,
+      "step": 19990
+    },
+    {
+      "epoch": 1.5876480977991227,
+      "grad_norm": 0.9035216569900513,
+      "learning_rate": 2.3546611627106984e-05,
+      "loss": 0.8465,
+      "step": 20000
+    },
+    {
+      "epoch": 1.5884419218480224,
+      "grad_norm": 0.7393106818199158,
+      "learning_rate": 2.353338096372152e-05,
+      "loss": 0.7915,
+      "step": 20010
+    },
+    {
+      "epoch": 1.589235745896922,
+      "grad_norm": 0.9715613722801208,
+      "learning_rate": 2.3520150300336062e-05,
+      "loss": 0.8255,
+      "step": 20020
+    },
+    {
+      "epoch": 1.5900295699458216,
+      "grad_norm": 0.79570472240448,
+      "learning_rate": 2.3506919636950598e-05,
+      "loss": 0.8283,
+      "step": 20030
+    },
+    {
+      "epoch": 1.5908233939947212,
+      "grad_norm": 0.738107442855835,
+      "learning_rate": 2.3493688973565137e-05,
+      "loss": 0.8292,
+      "step": 20040
+    },
+    {
+      "epoch": 1.5916172180436208,
+      "grad_norm": 1.0753180980682373,
+      "learning_rate": 2.3480458310179672e-05,
+      "loss": 0.8189,
+      "step": 20050
+    },
+    {
+      "epoch": 1.5924110420925202,
+      "grad_norm": 0.9600769877433777,
+      "learning_rate": 2.346722764679421e-05,
+      "loss": 0.855,
+      "step": 20060
+    },
+    {
+      "epoch": 1.5932048661414198,
+      "grad_norm": 0.7264878749847412,
+      "learning_rate": 2.345399698340875e-05,
+      "loss": 0.8333,
+      "step": 20070
+    },
+    {
+      "epoch": 1.5939986901903194,
+      "grad_norm": 1.0567606687545776,
+      "learning_rate": 2.3440766320023286e-05,
+      "loss": 0.842,
+      "step": 20080
+    },
+    {
+      "epoch": 1.5947925142392188,
+      "grad_norm": 0.8307549953460693,
+      "learning_rate": 2.3427535656637825e-05,
+      "loss": 0.8132,
+      "step": 20090
+    },
+    {
+      "epoch": 1.5955863382881184,
+      "grad_norm": 0.7764892578125,
+      "learning_rate": 2.341430499325236e-05,
+      "loss": 0.783,
+      "step": 20100
+    },
+    {
+      "epoch": 1.596380162337018,
+      "grad_norm": 0.8395814895629883,
+      "learning_rate": 2.3401074329866902e-05,
+      "loss": 0.8022,
+      "step": 20110
+    },
+    {
+      "epoch": 1.5971739863859176,
+      "grad_norm": 0.7872236371040344,
+      "learning_rate": 2.3387843666481438e-05,
+      "loss": 0.8471,
+      "step": 20120
+    },
+    {
+      "epoch": 1.5979678104348172,
+      "grad_norm": 0.9380735158920288,
+      "learning_rate": 2.3374613003095977e-05,
+      "loss": 0.7703,
+      "step": 20130
+    },
+    {
+      "epoch": 1.5987616344837168,
+      "grad_norm": 1.0980794429779053,
+      "learning_rate": 2.3361382339710512e-05,
+      "loss": 0.7735,
+      "step": 20140
+    },
+    {
+      "epoch": 1.5995554585326164,
+      "grad_norm": 0.8633868098258972,
+      "learning_rate": 2.334815167632505e-05,
+      "loss": 0.7795,
+      "step": 20150
+    },
+    {
+      "epoch": 1.6003492825815158,
+      "grad_norm": 0.8296395540237427,
+      "learning_rate": 2.333492101293959e-05,
+      "loss": 0.783,
+      "step": 20160
+    },
+    {
+      "epoch": 1.6011431066304154,
+      "grad_norm": 0.9974943995475769,
+      "learning_rate": 2.332169034955413e-05,
+      "loss": 0.8203,
+      "step": 20170
+    },
+    {
+      "epoch": 1.601936930679315,
+      "grad_norm": 0.932245671749115,
+      "learning_rate": 2.3308459686168665e-05,
+      "loss": 0.8744,
+      "step": 20180
+    },
+    {
+      "epoch": 1.6027307547282144,
+      "grad_norm": 0.8958556652069092,
+      "learning_rate": 2.3295229022783204e-05,
+      "loss": 0.7615,
+      "step": 20190
+    },
+    {
+      "epoch": 1.603524578777114,
+      "grad_norm": 0.8927448391914368,
+      "learning_rate": 2.3281998359397743e-05,
+      "loss": 0.8409,
+      "step": 20200
+    },
+    {
+      "epoch": 1.6043184028260136,
+      "grad_norm": 0.8587812185287476,
+      "learning_rate": 2.3268767696012278e-05,
+      "loss": 0.7895,
+      "step": 20210
+    },
+    {
+      "epoch": 1.6051122268749132,
+      "grad_norm": 0.9127293825149536,
+      "learning_rate": 2.3255537032626817e-05,
+      "loss": 0.7561,
+      "step": 20220
+    },
+    {
+      "epoch": 1.6059060509238128,
+      "grad_norm": 0.8766940832138062,
+      "learning_rate": 2.3242306369241353e-05,
+      "loss": 0.8178,
+      "step": 20230
+    },
+    {
+      "epoch": 1.6066998749727124,
+      "grad_norm": 0.9021517634391785,
+      "learning_rate": 2.3229075705855895e-05,
+      "loss": 0.7978,
+      "step": 20240
+    },
+    {
+      "epoch": 1.607493699021612,
+      "grad_norm": 0.7711321711540222,
+      "learning_rate": 2.321584504247043e-05,
+      "loss": 0.8566,
+      "step": 20250
+    },
+    {
+      "epoch": 1.6082875230705114,
+      "grad_norm": 0.7530736327171326,
+      "learning_rate": 2.320261437908497e-05,
+      "loss": 0.772,
+      "step": 20260
+    },
+    {
+      "epoch": 1.609081347119411,
+      "grad_norm": 0.7428927421569824,
+      "learning_rate": 2.3189383715699505e-05,
+      "loss": 0.8944,
+      "step": 20270
+    },
+    {
+      "epoch": 1.6098751711683106,
+      "grad_norm": 0.7444825768470764,
+      "learning_rate": 2.3176153052314044e-05,
+      "loss": 0.8346,
+      "step": 20280
+    },
+    {
+      "epoch": 1.61066899521721,
+      "grad_norm": 1.0045958757400513,
+      "learning_rate": 2.3162922388928583e-05,
+      "loss": 0.8444,
+      "step": 20290
+    },
+    {
+      "epoch": 1.6114628192661096,
+      "grad_norm": 1.0150105953216553,
+      "learning_rate": 2.3149691725543122e-05,
+      "loss": 0.7973,
+      "step": 20300
+    },
+    {
+      "epoch": 1.6122566433150092,
+      "grad_norm": 0.9908817410469055,
+      "learning_rate": 2.3136461062157657e-05,
+      "loss": 0.7855,
+      "step": 20310
+    },
+    {
+      "epoch": 1.6130504673639088,
+      "grad_norm": 0.8784881830215454,
+      "learning_rate": 2.3123230398772193e-05,
+      "loss": 0.7707,
+      "step": 20320
+    },
+    {
+      "epoch": 1.6138442914128084,
+      "grad_norm": 0.7329348921775818,
+      "learning_rate": 2.3109999735386735e-05,
+      "loss": 0.8465,
+      "step": 20330
+    },
+    {
+      "epoch": 1.614638115461708,
+      "grad_norm": 0.9049598574638367,
+      "learning_rate": 2.309676907200127e-05,
+      "loss": 0.8031,
+      "step": 20340
+    },
+    {
+      "epoch": 1.6154319395106076,
+      "grad_norm": 0.8993352651596069,
+      "learning_rate": 2.308353840861581e-05,
+      "loss": 0.8029,
+      "step": 20350
+    },
+    {
+      "epoch": 1.6162257635595072,
+      "grad_norm": 0.7339730858802795,
+      "learning_rate": 2.3070307745230345e-05,
+      "loss": 0.8168,
+      "step": 20360
+    },
+    {
+      "epoch": 1.6170195876084066,
+      "grad_norm": 0.8246588706970215,
+      "learning_rate": 2.3057077081844884e-05,
+      "loss": 0.8413,
+      "step": 20370
+    },
+    {
+      "epoch": 1.6178134116573062,
+      "grad_norm": 0.8873500227928162,
+      "learning_rate": 2.3043846418459423e-05,
+      "loss": 0.822,
+      "step": 20380
+    },
+    {
+      "epoch": 1.6186072357062056,
+      "grad_norm": 0.7149674892425537,
+      "learning_rate": 2.3030615755073962e-05,
+      "loss": 0.8343,
+      "step": 20390
+    },
+    {
+      "epoch": 1.6194010597551052,
+      "grad_norm": 0.8390538096427917,
+      "learning_rate": 2.3017385091688497e-05,
+      "loss": 0.7945,
+      "step": 20400
+    },
+    {
+      "epoch": 1.6201948838040048,
+      "grad_norm": 0.7026415467262268,
+      "learning_rate": 2.3004154428303036e-05,
+      "loss": 0.7255,
+      "step": 20410
+    },
+    {
+      "epoch": 1.6209887078529044,
+      "grad_norm": 0.8849766254425049,
+      "learning_rate": 2.2990923764917575e-05,
+      "loss": 0.7589,
+      "step": 20420
+    },
+    {
+      "epoch": 1.621782531901804,
+      "grad_norm": 0.8478531241416931,
+      "learning_rate": 2.297769310153211e-05,
+      "loss": 0.7977,
+      "step": 20430
+    },
+    {
+      "epoch": 1.6225763559507036,
+      "grad_norm": 0.8617717027664185,
+      "learning_rate": 2.296446243814665e-05,
+      "loss": 0.8657,
+      "step": 20440
+    },
+    {
+      "epoch": 1.6233701799996032,
+      "grad_norm": 0.913943350315094,
+      "learning_rate": 2.2951231774761185e-05,
+      "loss": 0.8567,
+      "step": 20450
+    },
+    {
+      "epoch": 1.6241640040485028,
+      "grad_norm": 0.6891298294067383,
+      "learning_rate": 2.2938001111375724e-05,
+      "loss": 0.806,
+      "step": 20460
+    },
+    {
+      "epoch": 1.6249578280974022,
+      "grad_norm": 0.914467990398407,
+      "learning_rate": 2.2924770447990263e-05,
+      "loss": 0.7876,
+      "step": 20470
+    },
+    {
+      "epoch": 1.6257516521463018,
+      "grad_norm": 0.8670864701271057,
+      "learning_rate": 2.2911539784604802e-05,
+      "loss": 0.7589,
+      "step": 20480
+    },
+    {
+      "epoch": 1.6265454761952012,
+      "grad_norm": 0.8266043663024902,
+      "learning_rate": 2.2898309121219338e-05,
+      "loss": 0.8038,
+      "step": 20490
+    },
+    {
+      "epoch": 1.6273393002441008,
+      "grad_norm": 0.7241661548614502,
+      "learning_rate": 2.2885078457833877e-05,
+      "loss": 0.8191,
+      "step": 20500
+    },
+    {
+      "epoch": 1.6281331242930004,
+      "grad_norm": 0.7914606332778931,
+      "learning_rate": 2.2871847794448416e-05,
+      "loss": 0.759,
+      "step": 20510
+    },
+    {
+      "epoch": 1.6289269483419,
+      "grad_norm": 0.8614340424537659,
+      "learning_rate": 2.2858617131062954e-05,
+      "loss": 0.7668,
+      "step": 20520
+    },
+    {
+      "epoch": 1.6297207723907996,
+      "grad_norm": 0.9252728223800659,
+      "learning_rate": 2.284538646767749e-05,
+      "loss": 0.7982,
+      "step": 20530
+    },
+    {
+      "epoch": 1.6305145964396992,
+      "grad_norm": 0.808665931224823,
+      "learning_rate": 2.2832155804292026e-05,
+      "loss": 0.8616,
+      "step": 20540
+    },
+    {
+      "epoch": 1.6313084204885988,
+      "grad_norm": 0.6369495987892151,
+      "learning_rate": 2.2818925140906568e-05,
+      "loss": 0.8265,
+      "step": 20550
+    },
+    {
+      "epoch": 1.6321022445374984,
+      "grad_norm": 0.8415317535400391,
+      "learning_rate": 2.2805694477521103e-05,
+      "loss": 0.8126,
+      "step": 20560
+    },
+    {
+      "epoch": 1.6328960685863978,
+      "grad_norm": 0.9225695133209229,
+      "learning_rate": 2.2792463814135642e-05,
+      "loss": 0.8399,
+      "step": 20570
+    },
+    {
+      "epoch": 1.6336898926352974,
+      "grad_norm": 0.6714004278182983,
+      "learning_rate": 2.2779233150750178e-05,
+      "loss": 0.8543,
+      "step": 20580
+    },
+    {
+      "epoch": 1.6344837166841968,
+      "grad_norm": 0.8270373940467834,
+      "learning_rate": 2.2766002487364717e-05,
+      "loss": 0.8255,
+      "step": 20590
+    },
+    {
+      "epoch": 1.6352775407330964,
+      "grad_norm": 0.7550914883613586,
+      "learning_rate": 2.2752771823979256e-05,
+      "loss": 0.7879,
+      "step": 20600
+    },
+    {
+      "epoch": 1.636071364781996,
+      "grad_norm": 0.865746796131134,
+      "learning_rate": 2.2739541160593795e-05,
+      "loss": 0.857,
+      "step": 20610
+    },
+    {
+      "epoch": 1.6368651888308956,
+      "grad_norm": 0.8743048906326294,
+      "learning_rate": 2.272631049720833e-05,
+      "loss": 0.8459,
+      "step": 20620
+    },
+    {
+      "epoch": 1.6376590128797952,
+      "grad_norm": 0.7982882261276245,
+      "learning_rate": 2.271307983382287e-05,
+      "loss": 0.8263,
+      "step": 20630
+    },
+    {
+      "epoch": 1.6384528369286948,
+      "grad_norm": 0.7997922897338867,
+      "learning_rate": 2.2699849170437408e-05,
+      "loss": 0.7892,
+      "step": 20640
+    },
+    {
+      "epoch": 1.6392466609775944,
+      "grad_norm": 1.0065598487854004,
+      "learning_rate": 2.2686618507051947e-05,
+      "loss": 0.7943,
+      "step": 20650
+    },
+    {
+      "epoch": 1.640040485026494,
+      "grad_norm": 0.828996479511261,
+      "learning_rate": 2.2673387843666482e-05,
+      "loss": 0.7368,
+      "step": 20660
+    },
+    {
+      "epoch": 1.6408343090753934,
+      "grad_norm": 0.8274086117744446,
+      "learning_rate": 2.2660157180281018e-05,
+      "loss": 0.8166,
+      "step": 20670
+    },
+    {
+      "epoch": 1.641628133124293,
+      "grad_norm": 0.8134365081787109,
+      "learning_rate": 2.2646926516895557e-05,
+      "loss": 0.8618,
+      "step": 20680
+    },
+    {
+      "epoch": 1.6424219571731926,
+      "grad_norm": 0.8807488679885864,
+      "learning_rate": 2.2633695853510096e-05,
+      "loss": 0.8119,
+      "step": 20690
+    },
+    {
+      "epoch": 1.643215781222092,
+      "grad_norm": 0.93861985206604,
+      "learning_rate": 2.2620465190124635e-05,
+      "loss": 0.8055,
+      "step": 20700
+    },
+    {
+      "epoch": 1.6440096052709916,
+      "grad_norm": 0.8421952724456787,
+      "learning_rate": 2.260723452673917e-05,
+      "loss": 0.8429,
+      "step": 20710
+    },
+    {
+      "epoch": 1.6448034293198912,
+      "grad_norm": 0.9165804982185364,
+      "learning_rate": 2.259400386335371e-05,
+      "loss": 0.7859,
+      "step": 20720
+    },
+    {
+      "epoch": 1.6455972533687908,
+      "grad_norm": 0.7905272841453552,
+      "learning_rate": 2.2580773199968248e-05,
+      "loss": 0.8527,
+      "step": 20730
+    },
+    {
+      "epoch": 1.6463910774176904,
+      "grad_norm": 0.9320306777954102,
+      "learning_rate": 2.2567542536582787e-05,
+      "loss": 0.7826,
+      "step": 20740
+    },
+    {
+      "epoch": 1.64718490146659,
+      "grad_norm": 0.8871251940727234,
+      "learning_rate": 2.2554311873197323e-05,
+      "loss": 0.7819,
+      "step": 20750
+    },
+    {
+      "epoch": 1.6479787255154896,
+      "grad_norm": 1.1218063831329346,
+      "learning_rate": 2.254108120981186e-05,
+      "loss": 0.8509,
+      "step": 20760
+    },
+    {
+      "epoch": 1.648772549564389,
+      "grad_norm": 0.890609622001648,
+      "learning_rate": 2.25278505464264e-05,
+      "loss": 0.8236,
+      "step": 20770
+    },
+    {
+      "epoch": 1.6495663736132886,
+      "grad_norm": 0.8825326561927795,
+      "learning_rate": 2.2514619883040936e-05,
+      "loss": 0.7444,
+      "step": 20780
+    },
+    {
+      "epoch": 1.6503601976621882,
+      "grad_norm": 0.8295640349388123,
+      "learning_rate": 2.2501389219655475e-05,
+      "loss": 0.7725,
+      "step": 20790
+    },
+    {
+      "epoch": 1.6511540217110876,
+      "grad_norm": 0.9995619654655457,
+      "learning_rate": 2.248815855627001e-05,
+      "loss": 0.7952,
+      "step": 20800
+    },
+    {
+      "epoch": 1.6519478457599872,
+      "grad_norm": 1.0375027656555176,
+      "learning_rate": 2.247492789288455e-05,
+      "loss": 0.7575,
+      "step": 20810
+    },
+    {
+      "epoch": 1.6527416698088868,
+      "grad_norm": 0.7434347867965698,
+      "learning_rate": 2.246169722949909e-05,
+      "loss": 0.7775,
+      "step": 20820
+    },
+    {
+      "epoch": 1.6535354938577864,
+      "grad_norm": 0.7898569107055664,
+      "learning_rate": 2.2448466566113627e-05,
+      "loss": 0.8458,
+      "step": 20830
+    },
+    {
+      "epoch": 1.654329317906686,
+      "grad_norm": 0.8854445815086365,
+      "learning_rate": 2.2435235902728163e-05,
+      "loss": 0.8005,
+      "step": 20840
+    },
+    {
+      "epoch": 1.6551231419555856,
+      "grad_norm": 0.8585209846496582,
+      "learning_rate": 2.2422005239342702e-05,
+      "loss": 0.7762,
+      "step": 20850
+    },
+    {
+      "epoch": 1.6559169660044852,
+      "grad_norm": 0.7355554103851318,
+      "learning_rate": 2.240877457595724e-05,
+      "loss": 0.7967,
+      "step": 20860
+    },
+    {
+      "epoch": 1.6567107900533846,
+      "grad_norm": 0.8480395078659058,
+      "learning_rate": 2.239554391257178e-05,
+      "loss": 0.8113,
+      "step": 20870
+    },
+    {
+      "epoch": 1.6575046141022842,
+      "grad_norm": 0.7819201946258545,
+      "learning_rate": 2.2382313249186315e-05,
+      "loss": 0.7618,
+      "step": 20880
+    },
+    {
+      "epoch": 1.6582984381511838,
+      "grad_norm": 0.7378450036048889,
+      "learning_rate": 2.236908258580085e-05,
+      "loss": 0.8375,
+      "step": 20890
+    },
+    {
+      "epoch": 1.6590922622000832,
+      "grad_norm": 0.8621236085891724,
+      "learning_rate": 2.235585192241539e-05,
+      "loss": 0.7767,
+      "step": 20900
+    },
+    {
+      "epoch": 1.6598860862489828,
+      "grad_norm": 0.720856785774231,
+      "learning_rate": 2.234262125902993e-05,
+      "loss": 0.7847,
+      "step": 20910
+    },
+    {
+      "epoch": 1.6606799102978824,
+      "grad_norm": 0.9445878863334656,
+      "learning_rate": 2.2329390595644467e-05,
+      "loss": 0.832,
+      "step": 20920
+    },
+    {
+      "epoch": 1.661473734346782,
+      "grad_norm": 0.7476319670677185,
+      "learning_rate": 2.2316159932259003e-05,
+      "loss": 0.8122,
+      "step": 20930
+    },
+    {
+      "epoch": 1.6622675583956816,
+      "grad_norm": 0.7706652283668518,
+      "learning_rate": 2.2302929268873542e-05,
+      "loss": 0.7505,
+      "step": 20940
+    },
+    {
+      "epoch": 1.6630613824445812,
+      "grad_norm": 1.0088225603103638,
+      "learning_rate": 2.228969860548808e-05,
+      "loss": 0.8042,
+      "step": 20950
+    },
+    {
+      "epoch": 1.6638552064934808,
+      "grad_norm": 0.906871497631073,
+      "learning_rate": 2.227646794210262e-05,
+      "loss": 0.8492,
+      "step": 20960
+    },
+    {
+      "epoch": 1.6646490305423804,
+      "grad_norm": 0.8510369062423706,
+      "learning_rate": 2.2263237278717155e-05,
+      "loss": 0.789,
+      "step": 20970
+    },
+    {
+      "epoch": 1.6654428545912798,
+      "grad_norm": 0.842056930065155,
+      "learning_rate": 2.2250006615331694e-05,
+      "loss": 0.7597,
+      "step": 20980
+    },
+    {
+      "epoch": 1.6662366786401794,
+      "grad_norm": 0.8941336274147034,
+      "learning_rate": 2.2236775951946233e-05,
+      "loss": 0.7614,
+      "step": 20990
+    },
+    {
+      "epoch": 1.6670305026890788,
+      "grad_norm": 0.6981929540634155,
+      "learning_rate": 2.222354528856077e-05,
+      "loss": 0.8553,
+      "step": 21000
+    },
+    {
+      "epoch": 1.6678243267379784,
+      "grad_norm": 0.7219326496124268,
+      "learning_rate": 2.2210314625175308e-05,
+      "loss": 0.8378,
+      "step": 21010
+    },
+    {
+      "epoch": 1.668618150786878,
+      "grad_norm": 0.6784833073616028,
+      "learning_rate": 2.2197083961789843e-05,
+      "loss": 0.87,
+      "step": 21020
+    },
+    {
+      "epoch": 1.6694119748357776,
+      "grad_norm": 0.7664601802825928,
+      "learning_rate": 2.2183853298404382e-05,
+      "loss": 0.8162,
+      "step": 21030
+    },
+    {
+      "epoch": 1.6702057988846772,
+      "grad_norm": 0.7743760347366333,
+      "learning_rate": 2.217062263501892e-05,
+      "loss": 0.7797,
+      "step": 21040
+    },
+    {
+      "epoch": 1.6709996229335768,
+      "grad_norm": 0.9794851541519165,
+      "learning_rate": 2.215739197163346e-05,
+      "loss": 0.8093,
+      "step": 21050
+    },
+    {
+      "epoch": 1.6717934469824764,
+      "grad_norm": 0.8463415503501892,
+      "learning_rate": 2.2144161308247996e-05,
+      "loss": 0.7824,
+      "step": 21060
+    },
+    {
+      "epoch": 1.672587271031376,
+      "grad_norm": 0.8329752087593079,
+      "learning_rate": 2.2130930644862534e-05,
+      "loss": 0.824,
+      "step": 21070
+    },
+    {
+      "epoch": 1.6733810950802754,
+      "grad_norm": 0.7661581039428711,
+      "learning_rate": 2.2117699981477073e-05,
+      "loss": 0.8068,
+      "step": 21080
+    },
+    {
+      "epoch": 1.674174919129175,
+      "grad_norm": 0.8593927621841431,
+      "learning_rate": 2.2104469318091612e-05,
+      "loss": 0.8332,
+      "step": 21090
+    },
+    {
+      "epoch": 1.6749687431780744,
+      "grad_norm": 0.7758104801177979,
+      "learning_rate": 2.2091238654706148e-05,
+      "loss": 0.8511,
+      "step": 21100
+    },
+    {
+      "epoch": 1.675762567226974,
+      "grad_norm": 0.8966826796531677,
+      "learning_rate": 2.2078007991320687e-05,
+      "loss": 0.8041,
+      "step": 21110
+    },
+    {
+      "epoch": 1.6765563912758736,
+      "grad_norm": 0.8776934146881104,
+      "learning_rate": 2.2064777327935222e-05,
+      "loss": 0.8093,
+      "step": 21120
+    },
+    {
+      "epoch": 1.6773502153247732,
+      "grad_norm": 0.6183434724807739,
+      "learning_rate": 2.205154666454976e-05,
+      "loss": 0.8896,
+      "step": 21130
+    },
+    {
+      "epoch": 1.6781440393736728,
+      "grad_norm": 0.7311633825302124,
+      "learning_rate": 2.20383160011643e-05,
+      "loss": 0.7782,
+      "step": 21140
+    },
+    {
+      "epoch": 1.6789378634225725,
+      "grad_norm": 0.8047720789909363,
+      "learning_rate": 2.2025085337778836e-05,
+      "loss": 0.8534,
+      "step": 21150
+    },
+    {
+      "epoch": 1.679731687471472,
+      "grad_norm": 0.7538871169090271,
+      "learning_rate": 2.2011854674393375e-05,
+      "loss": 0.791,
+      "step": 21160
+    },
+    {
+      "epoch": 1.6805255115203717,
+      "grad_norm": 0.7678347826004028,
+      "learning_rate": 2.1998624011007914e-05,
+      "loss": 0.8283,
+      "step": 21170
+    },
+    {
+      "epoch": 1.681319335569271,
+      "grad_norm": 0.7614843249320984,
+      "learning_rate": 2.1985393347622452e-05,
+      "loss": 0.7892,
+      "step": 21180
+    },
+    {
+      "epoch": 1.6821131596181707,
+      "grad_norm": 0.7208516597747803,
+      "learning_rate": 2.1972162684236988e-05,
+      "loss": 0.7965,
+      "step": 21190
+    },
+    {
+      "epoch": 1.68290698366707,
+      "grad_norm": 0.7032837867736816,
+      "learning_rate": 2.1958932020851527e-05,
+      "loss": 0.7784,
+      "step": 21200
+    },
+    {
+      "epoch": 1.6837008077159696,
+      "grad_norm": 0.705690860748291,
+      "learning_rate": 2.1945701357466062e-05,
+      "loss": 0.829,
+      "step": 21210
+    },
+    {
+      "epoch": 1.6844946317648692,
+      "grad_norm": 0.8606507778167725,
+      "learning_rate": 2.1932470694080605e-05,
+      "loss": 0.8002,
+      "step": 21220
+    },
+    {
+      "epoch": 1.6852884558137688,
+      "grad_norm": 0.7552010416984558,
+      "learning_rate": 2.191924003069514e-05,
+      "loss": 0.7753,
+      "step": 21230
+    },
+    {
+      "epoch": 1.6860822798626685,
+      "grad_norm": 0.9295687079429626,
+      "learning_rate": 2.1906009367309676e-05,
+      "loss": 0.8016,
+      "step": 21240
+    },
+    {
+      "epoch": 1.686876103911568,
+      "grad_norm": 0.8093982934951782,
+      "learning_rate": 2.1892778703924215e-05,
+      "loss": 0.8157,
+      "step": 21250
+    },
+    {
+      "epoch": 1.6876699279604677,
+      "grad_norm": 0.8069568872451782,
+      "learning_rate": 2.1879548040538754e-05,
+      "loss": 0.836,
+      "step": 21260
+    },
+    {
+      "epoch": 1.6884637520093673,
+      "grad_norm": 0.799312174320221,
+      "learning_rate": 2.1866317377153293e-05,
+      "loss": 0.8299,
+      "step": 21270
+    },
+    {
+      "epoch": 1.6892575760582667,
+      "grad_norm": 0.8665412068367004,
+      "learning_rate": 2.1853086713767828e-05,
+      "loss": 0.795,
+      "step": 21280
+    },
+    {
+      "epoch": 1.6900514001071663,
+      "grad_norm": 0.8677714467048645,
+      "learning_rate": 2.1839856050382367e-05,
+      "loss": 0.8833,
+      "step": 21290
+    },
+    {
+      "epoch": 1.6908452241560659,
+      "grad_norm": 0.7136849164962769,
+      "learning_rate": 2.1826625386996906e-05,
+      "loss": 0.83,
+      "step": 21300
+    },
+    {
+      "epoch": 1.6916390482049652,
+      "grad_norm": 0.8250473737716675,
+      "learning_rate": 2.1813394723611445e-05,
+      "loss": 0.8308,
+      "step": 21310
+    },
+    {
+      "epoch": 1.6924328722538649,
+      "grad_norm": 0.7840633988380432,
+      "learning_rate": 2.180016406022598e-05,
+      "loss": 0.8283,
+      "step": 21320
+    },
+    {
+      "epoch": 1.6932266963027645,
+      "grad_norm": 0.8209207653999329,
+      "learning_rate": 2.178693339684052e-05,
+      "loss": 0.7752,
+      "step": 21330
+    },
+    {
+      "epoch": 1.694020520351664,
+      "grad_norm": 0.7074013948440552,
+      "learning_rate": 2.1773702733455055e-05,
+      "loss": 0.7849,
+      "step": 21340
+    },
+    {
+      "epoch": 1.6948143444005637,
+      "grad_norm": 0.8547490835189819,
+      "learning_rate": 2.1760472070069594e-05,
+      "loss": 0.8125,
+      "step": 21350
+    },
+    {
+      "epoch": 1.6956081684494633,
+      "grad_norm": 0.7698444724082947,
+      "learning_rate": 2.1747241406684133e-05,
+      "loss": 0.766,
+      "step": 21360
+    },
+    {
+      "epoch": 1.6964019924983629,
+      "grad_norm": 0.9371459484100342,
+      "learning_rate": 2.173401074329867e-05,
+      "loss": 0.7663,
+      "step": 21370
+    },
+    {
+      "epoch": 1.6971958165472623,
+      "grad_norm": 0.8051425218582153,
+      "learning_rate": 2.1720780079913207e-05,
+      "loss": 0.8333,
+      "step": 21380
+    },
+    {
+      "epoch": 1.6979896405961619,
+      "grad_norm": 0.7680243253707886,
+      "learning_rate": 2.1707549416527746e-05,
+      "loss": 0.8129,
+      "step": 21390
+    },
+    {
+      "epoch": 1.6987834646450615,
+      "grad_norm": 0.6620742082595825,
+      "learning_rate": 2.1694318753142285e-05,
+      "loss": 0.8581,
+      "step": 21400
+    },
+    {
+      "epoch": 1.6995772886939609,
+      "grad_norm": 0.8950489163398743,
+      "learning_rate": 2.168108808975682e-05,
+      "loss": 0.8185,
+      "step": 21410
+    },
+    {
+      "epoch": 1.7003711127428605,
+      "grad_norm": 0.7099031805992126,
+      "learning_rate": 2.166785742637136e-05,
+      "loss": 0.8097,
+      "step": 21420
+    },
+    {
+      "epoch": 1.70116493679176,
+      "grad_norm": 0.9240164756774902,
+      "learning_rate": 2.1654626762985895e-05,
+      "loss": 0.8416,
+      "step": 21430
+    },
+    {
+      "epoch": 1.7019587608406597,
+      "grad_norm": 0.7131415009498596,
+      "learning_rate": 2.1641396099600437e-05,
+      "loss": 0.7934,
+      "step": 21440
+    },
+    {
+      "epoch": 1.7027525848895593,
+      "grad_norm": 0.8093000650405884,
+      "learning_rate": 2.1628165436214973e-05,
+      "loss": 0.8598,
+      "step": 21450
+    },
+    {
+      "epoch": 1.7035464089384589,
+      "grad_norm": 0.8821578025817871,
+      "learning_rate": 2.1614934772829512e-05,
+      "loss": 0.8821,
+      "step": 21460
+    },
+    {
+      "epoch": 1.7043402329873585,
+      "grad_norm": 0.835568904876709,
+      "learning_rate": 2.1601704109444047e-05,
+      "loss": 0.8338,
+      "step": 21470
+    },
+    {
+      "epoch": 1.7051340570362579,
+      "grad_norm": 0.9917017817497253,
+      "learning_rate": 2.1588473446058586e-05,
+      "loss": 0.8007,
+      "step": 21480
+    },
+    {
+      "epoch": 1.7059278810851575,
+      "grad_norm": 0.8382050395011902,
+      "learning_rate": 2.1575242782673125e-05,
+      "loss": 0.733,
+      "step": 21490
+    },
+    {
+      "epoch": 1.706721705134057,
+      "grad_norm": 0.9868109226226807,
+      "learning_rate": 2.156201211928766e-05,
+      "loss": 0.8729,
+      "step": 21500
+    },
+    {
+      "epoch": 1.7075155291829565,
+      "grad_norm": 0.8536074161529541,
+      "learning_rate": 2.15487814559022e-05,
+      "loss": 0.8462,
+      "step": 21510
+    },
+    {
+      "epoch": 1.708309353231856,
+      "grad_norm": 0.7851694822311401,
+      "learning_rate": 2.153555079251674e-05,
+      "loss": 0.8142,
+      "step": 21520
+    },
+    {
+      "epoch": 1.7091031772807557,
+      "grad_norm": 0.9547113180160522,
+      "learning_rate": 2.1522320129131278e-05,
+      "loss": 0.8597,
+      "step": 21530
+    },
+    {
+      "epoch": 1.7098970013296553,
+      "grad_norm": 0.9159030914306641,
+      "learning_rate": 2.1509089465745813e-05,
+      "loss": 0.8335,
+      "step": 21540
+    },
+    {
+      "epoch": 1.7106908253785549,
+      "grad_norm": 0.6976479887962341,
+      "learning_rate": 2.1495858802360352e-05,
+      "loss": 0.8117,
+      "step": 21550
+    },
+    {
+      "epoch": 1.7114846494274545,
+      "grad_norm": 0.7213908433914185,
+      "learning_rate": 2.1482628138974888e-05,
+      "loss": 0.8176,
+      "step": 21560
+    },
+    {
+      "epoch": 1.712278473476354,
+      "grad_norm": 0.8861874341964722,
+      "learning_rate": 2.146939747558943e-05,
+      "loss": 0.7886,
+      "step": 21570
+    },
+    {
+      "epoch": 1.7130722975252535,
+      "grad_norm": 0.87417072057724,
+      "learning_rate": 2.1456166812203966e-05,
+      "loss": 0.7848,
+      "step": 21580
+    },
+    {
+      "epoch": 1.713866121574153,
+      "grad_norm": 0.9183176755905151,
+      "learning_rate": 2.14429361488185e-05,
+      "loss": 0.8826,
+      "step": 21590
+    },
+    {
+      "epoch": 1.7146599456230527,
+      "grad_norm": 0.7000020146369934,
+      "learning_rate": 2.142970548543304e-05,
+      "loss": 0.8566,
+      "step": 21600
+    },
+    {
+      "epoch": 1.715453769671952,
+      "grad_norm": 0.7864513993263245,
+      "learning_rate": 2.141647482204758e-05,
+      "loss": 0.7844,
+      "step": 21610
+    },
+    {
+      "epoch": 1.7162475937208517,
+      "grad_norm": 0.8153032660484314,
+      "learning_rate": 2.1403244158662118e-05,
+      "loss": 0.8072,
+      "step": 21620
+    },
+    {
+      "epoch": 1.7170414177697513,
+      "grad_norm": 0.8709026575088501,
+      "learning_rate": 2.1390013495276653e-05,
+      "loss": 0.8276,
+      "step": 21630
+    },
+    {
+      "epoch": 1.7178352418186509,
+      "grad_norm": 0.8884004950523376,
+      "learning_rate": 2.1376782831891192e-05,
+      "loss": 0.8013,
+      "step": 21640
+    },
+    {
+      "epoch": 1.7186290658675505,
+      "grad_norm": 0.7052041292190552,
+      "learning_rate": 2.1363552168505728e-05,
+      "loss": 0.8273,
+      "step": 21650
+    },
+    {
+      "epoch": 1.71942288991645,
+      "grad_norm": 0.8561339378356934,
+      "learning_rate": 2.135032150512027e-05,
+      "loss": 0.8321,
+      "step": 21660
+    },
+    {
+      "epoch": 1.7202167139653497,
+      "grad_norm": 0.8964945077896118,
+      "learning_rate": 2.1337090841734806e-05,
+      "loss": 0.7822,
+      "step": 21670
+    },
+    {
+      "epoch": 1.7210105380142493,
+      "grad_norm": 0.702278733253479,
+      "learning_rate": 2.1323860178349345e-05,
+      "loss": 0.8043,
+      "step": 21680
+    },
+    {
+      "epoch": 1.7218043620631487,
+      "grad_norm": 0.8085154891014099,
+      "learning_rate": 2.131062951496388e-05,
+      "loss": 0.8019,
+      "step": 21690
+    },
+    {
+      "epoch": 1.7225981861120483,
+      "grad_norm": 0.8028178811073303,
+      "learning_rate": 2.129739885157842e-05,
+      "loss": 0.8023,
+      "step": 21700
+    },
+    {
+      "epoch": 1.7233920101609477,
+      "grad_norm": 0.8829013705253601,
+      "learning_rate": 2.1284168188192958e-05,
+      "loss": 0.7858,
+      "step": 21710
+    },
+    {
+      "epoch": 1.7241858342098473,
+      "grad_norm": 0.8301076889038086,
+      "learning_rate": 2.1270937524807494e-05,
+      "loss": 0.8117,
+      "step": 21720
+    },
+    {
+      "epoch": 1.7249796582587469,
+      "grad_norm": 0.8090790510177612,
+      "learning_rate": 2.1257706861422032e-05,
+      "loss": 0.7803,
+      "step": 21730
+    },
+    {
+      "epoch": 1.7257734823076465,
+      "grad_norm": 0.8940252065658569,
+      "learning_rate": 2.124447619803657e-05,
+      "loss": 0.7733,
+      "step": 21740
+    },
+    {
+      "epoch": 1.726567306356546,
+      "grad_norm": 0.8481318950653076,
+      "learning_rate": 2.123124553465111e-05,
+      "loss": 0.826,
+      "step": 21750
+    },
+    {
+      "epoch": 1.7273611304054457,
+      "grad_norm": 0.7668688297271729,
+      "learning_rate": 2.1218014871265646e-05,
+      "loss": 0.8235,
+      "step": 21760
+    },
+    {
+      "epoch": 1.7281549544543453,
+      "grad_norm": 0.8397562503814697,
+      "learning_rate": 2.1204784207880185e-05,
+      "loss": 0.8275,
+      "step": 21770
+    },
+    {
+      "epoch": 1.728948778503245,
+      "grad_norm": 0.786025881767273,
+      "learning_rate": 2.119155354449472e-05,
+      "loss": 0.862,
+      "step": 21780
+    },
+    {
+      "epoch": 1.7297426025521443,
+      "grad_norm": 0.7707594633102417,
+      "learning_rate": 2.1178322881109263e-05,
+      "loss": 0.8089,
+      "step": 21790
+    },
+    {
+      "epoch": 1.730536426601044,
+      "grad_norm": 0.7764220237731934,
+      "learning_rate": 2.1165092217723798e-05,
+      "loss": 0.7513,
+      "step": 21800
+    },
+    {
+      "epoch": 1.7313302506499433,
+      "grad_norm": 0.9090325832366943,
+      "learning_rate": 2.1151861554338334e-05,
+      "loss": 0.7636,
+      "step": 21810
+    },
+    {
+      "epoch": 1.7321240746988429,
+      "grad_norm": 0.7767049074172974,
+      "learning_rate": 2.1138630890952873e-05,
+      "loss": 0.8391,
+      "step": 21820
+    },
+    {
+      "epoch": 1.7329178987477425,
+      "grad_norm": 0.7917195558547974,
+      "learning_rate": 2.112540022756741e-05,
+      "loss": 0.8291,
+      "step": 21830
+    },
+    {
+      "epoch": 1.733711722796642,
+      "grad_norm": 0.9477527141571045,
+      "learning_rate": 2.111216956418195e-05,
+      "loss": 0.7697,
+      "step": 21840
+    },
+    {
+      "epoch": 1.7345055468455417,
+      "grad_norm": 0.8348498940467834,
+      "learning_rate": 2.1098938900796486e-05,
+      "loss": 0.7575,
+      "step": 21850
+    },
+    {
+      "epoch": 1.7352993708944413,
+      "grad_norm": 0.7253052592277527,
+      "learning_rate": 2.1085708237411025e-05,
+      "loss": 0.7476,
+      "step": 21860
+    },
+    {
+      "epoch": 1.736093194943341,
+      "grad_norm": 0.8680984973907471,
+      "learning_rate": 2.107247757402556e-05,
+      "loss": 0.7865,
+      "step": 21870
+    },
+    {
+      "epoch": 1.7368870189922405,
+      "grad_norm": 0.7522729635238647,
+      "learning_rate": 2.1059246910640103e-05,
+      "loss": 0.8271,
+      "step": 21880
+    },
+    {
+      "epoch": 1.73768084304114,
+      "grad_norm": 0.6878653764724731,
+      "learning_rate": 2.104601624725464e-05,
+      "loss": 0.8229,
+      "step": 21890
+    },
+    {
+      "epoch": 1.7384746670900395,
+      "grad_norm": 0.9120025038719177,
+      "learning_rate": 2.1032785583869177e-05,
+      "loss": 0.8646,
+      "step": 21900
+    },
+    {
+      "epoch": 1.7392684911389389,
+      "grad_norm": 0.7929124236106873,
+      "learning_rate": 2.1019554920483713e-05,
+      "loss": 0.8654,
+      "step": 21910
+    },
+    {
+      "epoch": 1.7400623151878385,
+      "grad_norm": 0.8900664448738098,
+      "learning_rate": 2.1006324257098252e-05,
+      "loss": 0.8068,
+      "step": 21920
+    },
+    {
+      "epoch": 1.740856139236738,
+      "grad_norm": 0.8378934860229492,
+      "learning_rate": 2.099309359371279e-05,
+      "loss": 0.7829,
+      "step": 21930
+    },
+    {
+      "epoch": 1.7416499632856377,
+      "grad_norm": 0.8364038467407227,
+      "learning_rate": 2.0979862930327326e-05,
+      "loss": 0.8372,
+      "step": 21940
+    },
+    {
+      "epoch": 1.7424437873345373,
+      "grad_norm": 0.7890764474868774,
+      "learning_rate": 2.0966632266941865e-05,
+      "loss": 0.8067,
+      "step": 21950
+    },
+    {
+      "epoch": 1.743237611383437,
+      "grad_norm": 0.8029731512069702,
+      "learning_rate": 2.09534016035564e-05,
+      "loss": 0.8142,
+      "step": 21960
+    },
+    {
+      "epoch": 1.7440314354323365,
+      "grad_norm": 0.9675547480583191,
+      "learning_rate": 2.0940170940170943e-05,
+      "loss": 0.8511,
+      "step": 21970
+    },
+    {
+      "epoch": 1.7448252594812361,
+      "grad_norm": 0.866611897945404,
+      "learning_rate": 2.092694027678548e-05,
+      "loss": 0.7468,
+      "step": 21980
+    },
+    {
+      "epoch": 1.7456190835301355,
+      "grad_norm": 0.7923464775085449,
+      "learning_rate": 2.0913709613400017e-05,
+      "loss": 0.8396,
+      "step": 21990
+    },
+    {
+      "epoch": 1.7464129075790351,
+      "grad_norm": 0.8513892292976379,
+      "learning_rate": 2.0900478950014553e-05,
+      "loss": 0.8128,
+      "step": 22000
+    },
+    {
+      "epoch": 1.7472067316279347,
+      "grad_norm": 0.7640386819839478,
+      "learning_rate": 2.0887248286629092e-05,
+      "loss": 0.7826,
+      "step": 22010
+    },
+    {
+      "epoch": 1.748000555676834,
+      "grad_norm": 0.8362756967544556,
+      "learning_rate": 2.087401762324363e-05,
+      "loss": 0.7284,
+      "step": 22020
+    },
+    {
+      "epoch": 1.7487943797257337,
+      "grad_norm": 0.78795325756073,
+      "learning_rate": 2.086078695985817e-05,
+      "loss": 0.8005,
+      "step": 22030
+    },
+    {
+      "epoch": 1.7495882037746333,
+      "grad_norm": 0.9655894637107849,
+      "learning_rate": 2.0847556296472705e-05,
+      "loss": 0.8236,
+      "step": 22040
+    },
+    {
+      "epoch": 1.750382027823533,
+      "grad_norm": 0.8761208653450012,
+      "learning_rate": 2.0834325633087244e-05,
+      "loss": 0.8187,
+      "step": 22050
+    },
+    {
+      "epoch": 1.7511758518724325,
+      "grad_norm": 0.6492605209350586,
+      "learning_rate": 2.0821094969701783e-05,
+      "loss": 0.922,
+      "step": 22060
+    },
+    {
+      "epoch": 1.7519696759213321,
+      "grad_norm": 0.6981455683708191,
+      "learning_rate": 2.080786430631632e-05,
+      "loss": 0.7863,
+      "step": 22070
+    },
+    {
+      "epoch": 1.7527634999702317,
+      "grad_norm": 0.978775680065155,
+      "learning_rate": 2.0794633642930858e-05,
+      "loss": 0.7996,
+      "step": 22080
+    },
+    {
+      "epoch": 1.7535573240191311,
+      "grad_norm": 0.7167590260505676,
+      "learning_rate": 2.0781402979545393e-05,
+      "loss": 0.8866,
+      "step": 22090
+    },
+    {
+      "epoch": 1.7543511480680307,
+      "grad_norm": 0.7744993567466736,
+      "learning_rate": 2.0768172316159935e-05,
+      "loss": 0.7559,
+      "step": 22100
+    },
+    {
+      "epoch": 1.7551449721169303,
+      "grad_norm": 0.8939769268035889,
+      "learning_rate": 2.075494165277447e-05,
+      "loss": 0.7956,
+      "step": 22110
+    },
+    {
+      "epoch": 1.7559387961658297,
+      "grad_norm": 0.74879390001297,
+      "learning_rate": 2.074171098938901e-05,
+      "loss": 0.8159,
+      "step": 22120
+    },
+    {
+      "epoch": 1.7567326202147293,
+      "grad_norm": 0.7568157911300659,
+      "learning_rate": 2.0728480326003546e-05,
+      "loss": 0.7673,
+      "step": 22130
+    },
+    {
+      "epoch": 1.757526444263629,
+      "grad_norm": 0.9827935695648193,
+      "learning_rate": 2.0715249662618084e-05,
+      "loss": 0.8079,
+      "step": 22140
+    },
+    {
+      "epoch": 1.7583202683125285,
+      "grad_norm": 0.8912889957427979,
+      "learning_rate": 2.0702018999232623e-05,
+      "loss": 0.821,
+      "step": 22150
+    },
+    {
+      "epoch": 1.7591140923614281,
+      "grad_norm": 0.8157911896705627,
+      "learning_rate": 2.068878833584716e-05,
+      "loss": 0.7612,
+      "step": 22160
+    },
+    {
+      "epoch": 1.7599079164103277,
+      "grad_norm": 0.745104968547821,
+      "learning_rate": 2.0675557672461698e-05,
+      "loss": 0.8183,
+      "step": 22170
+    },
+    {
+      "epoch": 1.7607017404592273,
+      "grad_norm": 0.846377432346344,
+      "learning_rate": 2.0662327009076233e-05,
+      "loss": 0.7756,
+      "step": 22180
+    },
+    {
+      "epoch": 1.7614955645081267,
+      "grad_norm": 0.8381850719451904,
+      "learning_rate": 2.0649096345690776e-05,
+      "loss": 0.8417,
+      "step": 22190
+    },
+    {
+      "epoch": 1.7622893885570263,
+      "grad_norm": 0.8737789392471313,
+      "learning_rate": 2.063586568230531e-05,
+      "loss": 0.7929,
+      "step": 22200
+    },
+    {
+      "epoch": 1.763083212605926,
+      "grad_norm": 0.8165319561958313,
+      "learning_rate": 2.062263501891985e-05,
+      "loss": 0.7364,
+      "step": 22210
+    },
+    {
+      "epoch": 1.7638770366548253,
+      "grad_norm": 0.9064796566963196,
+      "learning_rate": 2.0609404355534386e-05,
+      "loss": 0.8336,
+      "step": 22220
+    },
+    {
+      "epoch": 1.764670860703725,
+      "grad_norm": 0.8361899256706238,
+      "learning_rate": 2.0596173692148925e-05,
+      "loss": 0.7976,
+      "step": 22230
+    },
+    {
+      "epoch": 1.7654646847526245,
+      "grad_norm": 0.955283522605896,
+      "learning_rate": 2.0582943028763464e-05,
+      "loss": 0.781,
+      "step": 22240
+    },
+    {
+      "epoch": 1.7662585088015241,
+      "grad_norm": 0.87896329164505,
+      "learning_rate": 2.0569712365378002e-05,
+      "loss": 0.8037,
+      "step": 22250
+    },
+    {
+      "epoch": 1.7670523328504237,
+      "grad_norm": 0.91143399477005,
+      "learning_rate": 2.0556481701992538e-05,
+      "loss": 0.748,
+      "step": 22260
+    },
+    {
+      "epoch": 1.7678461568993233,
+      "grad_norm": 0.8325607180595398,
+      "learning_rate": 2.0543251038607077e-05,
+      "loss": 0.8072,
+      "step": 22270
+    },
+    {
+      "epoch": 1.768639980948223,
+      "grad_norm": 0.904513955116272,
+      "learning_rate": 2.0530020375221616e-05,
+      "loss": 0.7791,
+      "step": 22280
+    },
+    {
+      "epoch": 1.7694338049971226,
+      "grad_norm": 0.9098820686340332,
+      "learning_rate": 2.051678971183615e-05,
+      "loss": 0.8965,
+      "step": 22290
+    },
+    {
+      "epoch": 1.770227629046022,
+      "grad_norm": 0.900877833366394,
+      "learning_rate": 2.050355904845069e-05,
+      "loss": 0.8368,
+      "step": 22300
+    },
+    {
+      "epoch": 1.7710214530949215,
+      "grad_norm": 0.9795036911964417,
+      "learning_rate": 2.0490328385065226e-05,
+      "loss": 0.8214,
+      "step": 22310
+    },
+    {
+      "epoch": 1.771815277143821,
+      "grad_norm": 0.7888408303260803,
+      "learning_rate": 2.0477097721679768e-05,
+      "loss": 0.829,
+      "step": 22320
+    },
+    {
+      "epoch": 1.7726091011927205,
+      "grad_norm": 0.6869066953659058,
+      "learning_rate": 2.0463867058294304e-05,
+      "loss": 0.8656,
+      "step": 22330
+    },
+    {
+      "epoch": 1.7734029252416201,
+      "grad_norm": 0.891853928565979,
+      "learning_rate": 2.0450636394908843e-05,
+      "loss": 0.8304,
+      "step": 22340
+    },
+    {
+      "epoch": 1.7741967492905197,
+      "grad_norm": 0.6878489255905151,
+      "learning_rate": 2.0437405731523378e-05,
+      "loss": 0.8594,
+      "step": 22350
+    },
+    {
+      "epoch": 1.7749905733394193,
+      "grad_norm": 0.9317987561225891,
+      "learning_rate": 2.0424175068137917e-05,
+      "loss": 0.8681,
+      "step": 22360
+    },
+    {
+      "epoch": 1.775784397388319,
+      "grad_norm": 0.8260341882705688,
+      "learning_rate": 2.0410944404752456e-05,
+      "loss": 0.8144,
+      "step": 22370
+    },
+    {
+      "epoch": 1.7765782214372186,
+      "grad_norm": 0.7893334031105042,
+      "learning_rate": 2.0397713741366995e-05,
+      "loss": 0.8375,
+      "step": 22380
+    },
+    {
+      "epoch": 1.7773720454861182,
+      "grad_norm": 0.8867517113685608,
+      "learning_rate": 2.038448307798153e-05,
+      "loss": 0.817,
+      "step": 22390
+    },
+    {
+      "epoch": 1.7781658695350175,
+      "grad_norm": 0.843120813369751,
+      "learning_rate": 2.0371252414596066e-05,
+      "loss": 0.8542,
+      "step": 22400
+    },
+    {
+      "epoch": 1.7789596935839171,
+      "grad_norm": 0.8140029311180115,
+      "learning_rate": 2.035802175121061e-05,
+      "loss": 0.9021,
+      "step": 22410
+    },
+    {
+      "epoch": 1.7797535176328165,
+      "grad_norm": 0.9623441100120544,
+      "learning_rate": 2.0344791087825144e-05,
+      "loss": 0.8447,
+      "step": 22420
+    },
+    {
+      "epoch": 1.7805473416817161,
+      "grad_norm": 0.9128091335296631,
+      "learning_rate": 2.0331560424439683e-05,
+      "loss": 0.794,
+      "step": 22430
+    },
+    {
+      "epoch": 1.7813411657306157,
+      "grad_norm": 0.8841097354888916,
+      "learning_rate": 2.031832976105422e-05,
+      "loss": 0.8707,
+      "step": 22440
+    },
+    {
+      "epoch": 1.7821349897795153,
+      "grad_norm": 0.7501084208488464,
+      "learning_rate": 2.0305099097668757e-05,
+      "loss": 0.8113,
+      "step": 22450
+    },
+    {
+      "epoch": 1.782928813828415,
+      "grad_norm": 0.9023767113685608,
+      "learning_rate": 2.0291868434283296e-05,
+      "loss": 0.8584,
+      "step": 22460
+    },
+    {
+      "epoch": 1.7837226378773146,
+      "grad_norm": 0.9524368047714233,
+      "learning_rate": 2.0278637770897835e-05,
+      "loss": 0.7864,
+      "step": 22470
+    },
+    {
+      "epoch": 1.7845164619262142,
+      "grad_norm": 0.7797056436538696,
+      "learning_rate": 2.026540710751237e-05,
+      "loss": 0.7762,
+      "step": 22480
+    },
+    {
+      "epoch": 1.7853102859751138,
+      "grad_norm": 0.918021559715271,
+      "learning_rate": 2.025217644412691e-05,
+      "loss": 0.7703,
+      "step": 22490
+    },
+    {
+      "epoch": 1.7861041100240131,
+      "grad_norm": 0.9234730005264282,
+      "learning_rate": 2.023894578074145e-05,
+      "loss": 0.8419,
+      "step": 22500
+    },
+    {
+      "epoch": 1.7868979340729128,
+      "grad_norm": 0.6988440155982971,
+      "learning_rate": 2.0225715117355984e-05,
+      "loss": 0.8038,
+      "step": 22510
+    },
+    {
+      "epoch": 1.7876917581218121,
+      "grad_norm": 0.96622633934021,
+      "learning_rate": 2.0212484453970523e-05,
+      "loss": 0.792,
+      "step": 22520
+    },
+    {
+      "epoch": 1.7884855821707117,
+      "grad_norm": 0.7496207356452942,
+      "learning_rate": 2.019925379058506e-05,
+      "loss": 0.7702,
+      "step": 22530
+    },
+    {
+      "epoch": 1.7892794062196113,
+      "grad_norm": 0.9251072406768799,
+      "learning_rate": 2.01860231271996e-05,
+      "loss": 0.8013,
+      "step": 22540
+    },
+    {
+      "epoch": 1.790073230268511,
+      "grad_norm": 0.8455217480659485,
+      "learning_rate": 2.0172792463814136e-05,
+      "loss": 0.8374,
+      "step": 22550
+    },
+    {
+      "epoch": 1.7908670543174106,
+      "grad_norm": 0.8806858062744141,
+      "learning_rate": 2.0159561800428675e-05,
+      "loss": 0.8292,
+      "step": 22560
+    },
+    {
+      "epoch": 1.7916608783663102,
+      "grad_norm": 0.8762655258178711,
+      "learning_rate": 2.014633113704321e-05,
+      "loss": 0.774,
+      "step": 22570
+    },
+    {
+      "epoch": 1.7924547024152098,
+      "grad_norm": 0.8209755420684814,
+      "learning_rate": 2.013310047365775e-05,
+      "loss": 0.7664,
+      "step": 22580
+    },
+    {
+      "epoch": 1.7932485264641094,
+      "grad_norm": 0.8102571964263916,
+      "learning_rate": 2.011986981027229e-05,
+      "loss": 0.8032,
+      "step": 22590
+    },
+    {
+      "epoch": 1.7940423505130088,
+      "grad_norm": 0.7247342467308044,
+      "learning_rate": 2.0106639146886828e-05,
+      "loss": 0.832,
+      "step": 22600
+    },
+    {
+      "epoch": 1.7948361745619084,
+      "grad_norm": 0.9266355633735657,
+      "learning_rate": 2.0093408483501363e-05,
+      "loss": 0.7909,
+      "step": 22610
+    },
+    {
+      "epoch": 1.795629998610808,
+      "grad_norm": 0.7393051981925964,
+      "learning_rate": 2.00801778201159e-05,
+      "loss": 0.8408,
+      "step": 22620
+    },
+    {
+      "epoch": 1.7964238226597073,
+      "grad_norm": 0.8639968633651733,
+      "learning_rate": 2.006694715673044e-05,
+      "loss": 0.8002,
+      "step": 22630
+    },
+    {
+      "epoch": 1.797217646708607,
+      "grad_norm": 0.9779701232910156,
+      "learning_rate": 2.0053716493344977e-05,
+      "loss": 0.7608,
+      "step": 22640
+    },
+    {
+      "epoch": 1.7980114707575066,
+      "grad_norm": 0.8327713012695312,
+      "learning_rate": 2.0040485829959516e-05,
+      "loss": 0.8247,
+      "step": 22650
+    },
+    {
+      "epoch": 1.7988052948064062,
+      "grad_norm": 0.8809974193572998,
+      "learning_rate": 2.002725516657405e-05,
+      "loss": 0.8666,
+      "step": 22660
+    },
+    {
+      "epoch": 1.7995991188553058,
+      "grad_norm": 0.8804794549942017,
+      "learning_rate": 2.001402450318859e-05,
+      "loss": 0.8138,
+      "step": 22670
+    },
+    {
+      "epoch": 1.8003929429042054,
+      "grad_norm": 0.8360373377799988,
+      "learning_rate": 2.000079383980313e-05,
+      "loss": 0.8091,
+      "step": 22680
+    },
+    {
+      "epoch": 1.801186766953105,
+      "grad_norm": 0.8613433241844177,
+      "learning_rate": 1.9987563176417668e-05,
+      "loss": 0.7882,
+      "step": 22690
+    },
+    {
+      "epoch": 1.8019805910020044,
+      "grad_norm": 0.8613284826278687,
+      "learning_rate": 1.9974332513032203e-05,
+      "loss": 0.8012,
+      "step": 22700
+    },
+    {
+      "epoch": 1.802774415050904,
+      "grad_norm": 0.8369642496109009,
+      "learning_rate": 1.9961101849646742e-05,
+      "loss": 0.7765,
+      "step": 22710
+    },
+    {
+      "epoch": 1.8035682390998036,
+      "grad_norm": 0.8004109859466553,
+      "learning_rate": 1.994787118626128e-05,
+      "loss": 0.7961,
+      "step": 22720
+    },
+    {
+      "epoch": 1.804362063148703,
+      "grad_norm": 0.792927086353302,
+      "learning_rate": 1.993464052287582e-05,
+      "loss": 0.848,
+      "step": 22730
+    },
+    {
+      "epoch": 1.8051558871976026,
+      "grad_norm": 0.7693809866905212,
+      "learning_rate": 1.9921409859490356e-05,
+      "loss": 0.786,
+      "step": 22740
+    },
+    {
+      "epoch": 1.8059497112465022,
+      "grad_norm": 1.0815696716308594,
+      "learning_rate": 1.990817919610489e-05,
+      "loss": 0.8142,
+      "step": 22750
+    },
+    {
+      "epoch": 1.8067435352954018,
+      "grad_norm": 0.9764348268508911,
+      "learning_rate": 1.989494853271943e-05,
+      "loss": 0.8801,
+      "step": 22760
+    },
+    {
+      "epoch": 1.8075373593443014,
+      "grad_norm": 0.9634998440742493,
+      "learning_rate": 1.988171786933397e-05,
+      "loss": 0.7793,
+      "step": 22770
+    },
+    {
+      "epoch": 1.808331183393201,
+      "grad_norm": 0.8265504837036133,
+      "learning_rate": 1.9868487205948508e-05,
+      "loss": 0.7417,
+      "step": 22780
+    },
+    {
+      "epoch": 1.8091250074421006,
+      "grad_norm": 0.8806119561195374,
+      "learning_rate": 1.9855256542563044e-05,
+      "loss": 0.8474,
+      "step": 22790
+    },
+    {
+      "epoch": 1.809918831491,
+      "grad_norm": 1.0471200942993164,
+      "learning_rate": 1.9842025879177582e-05,
+      "loss": 0.8872,
+      "step": 22800
+    },
+    {
+      "epoch": 1.8107126555398996,
+      "grad_norm": 0.7680615782737732,
+      "learning_rate": 1.982879521579212e-05,
+      "loss": 0.8097,
+      "step": 22810
+    },
+    {
+      "epoch": 1.8115064795887992,
+      "grad_norm": 0.9899775385856628,
+      "learning_rate": 1.981556455240666e-05,
+      "loss": 0.8203,
+      "step": 22820
+    },
+    {
+      "epoch": 1.8123003036376986,
+      "grad_norm": 0.8771310448646545,
+      "learning_rate": 1.9802333889021196e-05,
+      "loss": 0.783,
+      "step": 22830
+    },
+    {
+      "epoch": 1.8130941276865982,
+      "grad_norm": 0.7819077372550964,
+      "learning_rate": 1.9789103225635735e-05,
+      "loss": 0.8192,
+      "step": 22840
+    },
+    {
+      "epoch": 1.8138879517354978,
+      "grad_norm": 0.9690881371498108,
+      "learning_rate": 1.9775872562250274e-05,
+      "loss": 0.7972,
+      "step": 22850
+    },
+    {
+      "epoch": 1.8146817757843974,
+      "grad_norm": 0.9323807954788208,
+      "learning_rate": 1.976264189886481e-05,
+      "loss": 0.8351,
+      "step": 22860
+    },
+    {
+      "epoch": 1.815475599833297,
+      "grad_norm": 0.7597420811653137,
+      "learning_rate": 1.9749411235479348e-05,
+      "loss": 0.7927,
+      "step": 22870
+    },
+    {
+      "epoch": 1.8162694238821966,
+      "grad_norm": 0.614967942237854,
+      "learning_rate": 1.9736180572093884e-05,
+      "loss": 0.8414,
+      "step": 22880
+    },
+    {
+      "epoch": 1.8170632479310962,
+      "grad_norm": 0.9562491178512573,
+      "learning_rate": 1.9722949908708423e-05,
+      "loss": 0.7689,
+      "step": 22890
+    },
+    {
+      "epoch": 1.8178570719799958,
+      "grad_norm": 0.8071816563606262,
+      "learning_rate": 1.970971924532296e-05,
+      "loss": 0.7618,
+      "step": 22900
+    },
+    {
+      "epoch": 1.8186508960288952,
+      "grad_norm": 0.9454165697097778,
+      "learning_rate": 1.96964885819375e-05,
+      "loss": 0.8387,
+      "step": 22910
+    },
+    {
+      "epoch": 1.8194447200777948,
+      "grad_norm": 0.8835422992706299,
+      "learning_rate": 1.9683257918552036e-05,
+      "loss": 0.8217,
+      "step": 22920
+    },
+    {
+      "epoch": 1.8202385441266942,
+      "grad_norm": 0.9334397912025452,
+      "learning_rate": 1.9670027255166575e-05,
+      "loss": 0.7968,
+      "step": 22930
+    },
+    {
+      "epoch": 1.8210323681755938,
+      "grad_norm": 0.9972975254058838,
+      "learning_rate": 1.9656796591781114e-05,
+      "loss": 0.7708,
+      "step": 22940
+    },
+    {
+      "epoch": 1.8218261922244934,
+      "grad_norm": 1.0508019924163818,
+      "learning_rate": 1.9643565928395653e-05,
+      "loss": 0.795,
+      "step": 22950
+    },
+    {
+      "epoch": 1.822620016273393,
+      "grad_norm": 0.8678830862045288,
+      "learning_rate": 1.963033526501019e-05,
+      "loss": 0.8326,
+      "step": 22960
+    },
+    {
+      "epoch": 1.8234138403222926,
+      "grad_norm": 0.7960614562034607,
+      "learning_rate": 1.9617104601624724e-05,
+      "loss": 0.8116,
+      "step": 22970
+    },
+    {
+      "epoch": 1.8242076643711922,
+      "grad_norm": 0.8850997090339661,
+      "learning_rate": 1.9603873938239263e-05,
+      "loss": 0.7878,
+      "step": 22980
+    },
+    {
+      "epoch": 1.8250014884200918,
+      "grad_norm": 0.7653430700302124,
+      "learning_rate": 1.9590643274853802e-05,
+      "loss": 0.7775,
+      "step": 22990
+    },
+    {
+      "epoch": 1.8257953124689914,
+      "grad_norm": 0.87236487865448,
+      "learning_rate": 1.957741261146834e-05,
+      "loss": 0.7627,
+      "step": 23000
+    },
+    {
+      "epoch": 1.8265891365178908,
+      "grad_norm": 1.0964977741241455,
+      "learning_rate": 1.9564181948082876e-05,
+      "loss": 0.7447,
+      "step": 23010
+    },
+    {
+      "epoch": 1.8273829605667904,
+      "grad_norm": 0.8569262623786926,
+      "learning_rate": 1.9550951284697415e-05,
+      "loss": 0.8153,
+      "step": 23020
+    },
+    {
+      "epoch": 1.8281767846156898,
+      "grad_norm": 0.9798825979232788,
+      "learning_rate": 1.9537720621311954e-05,
+      "loss": 0.7993,
+      "step": 23030
+    },
+    {
+      "epoch": 1.8289706086645894,
+      "grad_norm": 0.8593613505363464,
+      "learning_rate": 1.9524489957926493e-05,
+      "loss": 0.8205,
+      "step": 23040
+    },
+    {
+      "epoch": 1.829764432713489,
+      "grad_norm": 0.6907086372375488,
+      "learning_rate": 1.951125929454103e-05,
+      "loss": 0.7571,
+      "step": 23050
+    },
+    {
+      "epoch": 1.8305582567623886,
+      "grad_norm": 0.8665714263916016,
+      "learning_rate": 1.9498028631155567e-05,
+      "loss": 0.8783,
+      "step": 23060
+    },
+    {
+      "epoch": 1.8313520808112882,
+      "grad_norm": 0.8479406833648682,
+      "learning_rate": 1.9484797967770106e-05,
+      "loss": 0.8075,
+      "step": 23070
+    },
+    {
+      "epoch": 1.8321459048601878,
+      "grad_norm": 1.0376914739608765,
+      "learning_rate": 1.9471567304384645e-05,
+      "loss": 0.8745,
+      "step": 23080
+    },
+    {
+      "epoch": 1.8329397289090874,
+      "grad_norm": 0.7835286855697632,
+      "learning_rate": 1.945833664099918e-05,
+      "loss": 0.7769,
+      "step": 23090
+    },
+    {
+      "epoch": 1.833733552957987,
+      "grad_norm": 0.7931967973709106,
+      "learning_rate": 1.9445105977613716e-05,
+      "loss": 0.7833,
+      "step": 23100
+    },
+    {
+      "epoch": 1.8345273770068864,
+      "grad_norm": 0.9185943007469177,
+      "learning_rate": 1.9433198380566804e-05,
+      "loss": 0.8216,
+      "step": 23110
+    },
+    {
+      "epoch": 1.835321201055786,
+      "grad_norm": 0.9194105267524719,
+      "learning_rate": 1.941996771718134e-05,
+      "loss": 0.8119,
+      "step": 23120
+    },
+    {
+      "epoch": 1.8361150251046854,
+      "grad_norm": 0.9397788047790527,
+      "learning_rate": 1.940673705379588e-05,
+      "loss": 0.7952,
+      "step": 23130
+    },
+    {
+      "epoch": 1.836908849153585,
+      "grad_norm": 0.9412715435028076,
+      "learning_rate": 1.9393506390410414e-05,
+      "loss": 0.8362,
+      "step": 23140
+    },
+    {
+      "epoch": 1.8377026732024846,
+      "grad_norm": 0.8924200534820557,
+      "learning_rate": 1.9380275727024953e-05,
+      "loss": 0.8191,
+      "step": 23150
+    },
+    {
+      "epoch": 1.8384964972513842,
+      "grad_norm": 0.9370428323745728,
+      "learning_rate": 1.9367045063639492e-05,
+      "loss": 0.7919,
+      "step": 23160
+    },
+    {
+      "epoch": 1.8392903213002838,
+      "grad_norm": 0.9872609376907349,
+      "learning_rate": 1.935381440025403e-05,
+      "loss": 0.7711,
+      "step": 23170
+    },
+    {
+      "epoch": 1.8400841453491834,
+      "grad_norm": 0.7494041323661804,
+      "learning_rate": 1.9340583736868566e-05,
+      "loss": 0.8043,
+      "step": 23180
+    },
+    {
+      "epoch": 1.840877969398083,
+      "grad_norm": 0.768407940864563,
+      "learning_rate": 1.9327353073483105e-05,
+      "loss": 0.7851,
+      "step": 23190
+    },
+    {
+      "epoch": 1.8416717934469826,
+      "grad_norm": 0.7478283047676086,
+      "learning_rate": 1.9314122410097644e-05,
+      "loss": 0.8769,
+      "step": 23200
+    },
+    {
+      "epoch": 1.842465617495882,
+      "grad_norm": 1.001345157623291,
+      "learning_rate": 1.930089174671218e-05,
+      "loss": 0.8168,
+      "step": 23210
+    },
+    {
+      "epoch": 1.8432594415447816,
+      "grad_norm": 0.9208034873008728,
+      "learning_rate": 1.928766108332672e-05,
+      "loss": 0.7763,
+      "step": 23220
+    },
+    {
+      "epoch": 1.8440532655936812,
+      "grad_norm": 0.7579790949821472,
+      "learning_rate": 1.9274430419941254e-05,
+      "loss": 0.8256,
+      "step": 23230
+    },
+    {
+      "epoch": 1.8448470896425806,
+      "grad_norm": 0.8360084891319275,
+      "learning_rate": 1.9261199756555797e-05,
+      "loss": 0.8554,
+      "step": 23240
+    },
+    {
+      "epoch": 1.8456409136914802,
+      "grad_norm": 0.8228520154953003,
+      "learning_rate": 1.9247969093170332e-05,
+      "loss": 0.7981,
+      "step": 23250
+    },
+    {
+      "epoch": 1.8464347377403798,
+      "grad_norm": 0.7764120697975159,
+      "learning_rate": 1.923473842978487e-05,
+      "loss": 0.8218,
+      "step": 23260
+    },
+    {
+      "epoch": 1.8472285617892794,
+      "grad_norm": 0.730930507183075,
+      "learning_rate": 1.9221507766399407e-05,
+      "loss": 0.8442,
+      "step": 23270
+    },
+    {
+      "epoch": 1.848022385838179,
+      "grad_norm": 0.8573958277702332,
+      "learning_rate": 1.9208277103013946e-05,
+      "loss": 0.8032,
+      "step": 23280
+    },
+    {
+      "epoch": 1.8488162098870786,
+      "grad_norm": 1.118808388710022,
+      "learning_rate": 1.9195046439628485e-05,
+      "loss": 0.8619,
+      "step": 23290
+    },
+    {
+      "epoch": 1.8496100339359782,
+      "grad_norm": 0.7608010768890381,
+      "learning_rate": 1.9181815776243023e-05,
+      "loss": 0.8218,
+      "step": 23300
+    },
+    {
+      "epoch": 1.8504038579848776,
+      "grad_norm": 0.8794158697128296,
+      "learning_rate": 1.916858511285756e-05,
+      "loss": 0.8562,
+      "step": 23310
+    },
+    {
+      "epoch": 1.8511976820337772,
+      "grad_norm": 0.8464013934135437,
+      "learning_rate": 1.9155354449472098e-05,
+      "loss": 0.836,
+      "step": 23320
+    },
+    {
+      "epoch": 1.8519915060826768,
+      "grad_norm": 0.7255980372428894,
+      "learning_rate": 1.9142123786086637e-05,
+      "loss": 0.8553,
+      "step": 23330
+    },
+    {
+      "epoch": 1.8527853301315762,
+      "grad_norm": 0.9724112153053284,
+      "learning_rate": 1.9128893122701172e-05,
+      "loss": 0.8473,
+      "step": 23340
+    },
+    {
+      "epoch": 1.8535791541804758,
+      "grad_norm": 0.7873225212097168,
+      "learning_rate": 1.911566245931571e-05,
+      "loss": 0.8406,
+      "step": 23350
+    },
+    {
+      "epoch": 1.8543729782293754,
+      "grad_norm": 0.8537048697471619,
+      "learning_rate": 1.9102431795930247e-05,
+      "loss": 0.8082,
+      "step": 23360
+    },
+    {
+      "epoch": 1.855166802278275,
+      "grad_norm": 0.7261394262313843,
+      "learning_rate": 1.9089201132544786e-05,
+      "loss": 0.8488,
+      "step": 23370
+    },
+    {
+      "epoch": 1.8559606263271746,
+      "grad_norm": 0.8339587450027466,
+      "learning_rate": 1.9075970469159325e-05,
+      "loss": 0.8027,
+      "step": 23380
+    },
+    {
+      "epoch": 1.8567544503760742,
+      "grad_norm": 1.3965240716934204,
+      "learning_rate": 1.9062739805773864e-05,
+      "loss": 0.7967,
+      "step": 23390
+    },
+    {
+      "epoch": 1.8575482744249738,
+      "grad_norm": 0.789261519908905,
+      "learning_rate": 1.90495091423884e-05,
+      "loss": 0.7911,
+      "step": 23400
+    },
+    {
+      "epoch": 1.8583420984738732,
+      "grad_norm": 0.6335147619247437,
+      "learning_rate": 1.9036278479002938e-05,
+      "loss": 0.8299,
+      "step": 23410
+    },
+    {
+      "epoch": 1.8591359225227728,
+      "grad_norm": 0.7321164608001709,
+      "learning_rate": 1.9023047815617477e-05,
+      "loss": 0.8545,
+      "step": 23420
+    },
+    {
+      "epoch": 1.8599297465716724,
+      "grad_norm": 0.9617056250572205,
+      "learning_rate": 1.9009817152232016e-05,
+      "loss": 0.7811,
+      "step": 23430
+    },
+    {
+      "epoch": 1.8607235706205718,
+      "grad_norm": 0.8931480050086975,
+      "learning_rate": 1.899658648884655e-05,
+      "loss": 0.8112,
+      "step": 23440
+    },
+    {
+      "epoch": 1.8615173946694714,
+      "grad_norm": 1.094520926475525,
+      "learning_rate": 1.8983355825461087e-05,
+      "loss": 0.8078,
+      "step": 23450
+    },
+    {
+      "epoch": 1.862311218718371,
+      "grad_norm": 0.9582784175872803,
+      "learning_rate": 1.8970125162075626e-05,
+      "loss": 0.763,
+      "step": 23460
+    },
+    {
+      "epoch": 1.8631050427672706,
+      "grad_norm": 0.9741404056549072,
+      "learning_rate": 1.8956894498690165e-05,
+      "loss": 0.7808,
+      "step": 23470
+    },
+    {
+      "epoch": 1.8638988668161702,
+      "grad_norm": 0.758374035358429,
+      "learning_rate": 1.8943663835304704e-05,
+      "loss": 0.8281,
+      "step": 23480
+    },
+    {
+      "epoch": 1.8646926908650698,
+      "grad_norm": 0.815403938293457,
+      "learning_rate": 1.893043317191924e-05,
+      "loss": 0.7594,
+      "step": 23490
+    },
+    {
+      "epoch": 1.8654865149139694,
+      "grad_norm": 0.7608597278594971,
+      "learning_rate": 1.8917202508533778e-05,
+      "loss": 0.8072,
+      "step": 23500
+    },
+    {
+      "epoch": 1.866280338962869,
+      "grad_norm": 0.8437971472740173,
+      "learning_rate": 1.8903971845148317e-05,
+      "loss": 0.8153,
+      "step": 23510
+    },
+    {
+      "epoch": 1.8670741630117684,
+      "grad_norm": 0.8459624648094177,
+      "learning_rate": 1.8890741181762856e-05,
+      "loss": 0.7585,
+      "step": 23520
+    },
+    {
+      "epoch": 1.867867987060668,
+      "grad_norm": 0.9859129786491394,
+      "learning_rate": 1.887751051837739e-05,
+      "loss": 0.7822,
+      "step": 23530
+    },
+    {
+      "epoch": 1.8686618111095674,
+      "grad_norm": 0.7122694849967957,
+      "learning_rate": 1.886427985499193e-05,
+      "loss": 0.8052,
+      "step": 23540
+    },
+    {
+      "epoch": 1.869455635158467,
+      "grad_norm": 0.8045480251312256,
+      "learning_rate": 1.885104919160647e-05,
+      "loss": 0.7523,
+      "step": 23550
+    },
+    {
+      "epoch": 1.8702494592073666,
+      "grad_norm": 0.8591973781585693,
+      "learning_rate": 1.8837818528221005e-05,
+      "loss": 0.8087,
+      "step": 23560
+    },
+    {
+      "epoch": 1.8710432832562662,
+      "grad_norm": 0.8517307043075562,
+      "learning_rate": 1.8824587864835544e-05,
+      "loss": 0.7844,
+      "step": 23570
+    },
+    {
+      "epoch": 1.8718371073051658,
+      "grad_norm": 0.96706223487854,
+      "learning_rate": 1.881135720145008e-05,
+      "loss": 0.7595,
+      "step": 23580
+    },
+    {
+      "epoch": 1.8726309313540654,
+      "grad_norm": 0.966511070728302,
+      "learning_rate": 1.879812653806462e-05,
+      "loss": 0.8569,
+      "step": 23590
+    },
+    {
+      "epoch": 1.873424755402965,
+      "grad_norm": 0.8597899675369263,
+      "learning_rate": 1.8784895874679157e-05,
+      "loss": 0.7934,
+      "step": 23600
+    },
+    {
+      "epoch": 1.8742185794518647,
+      "grad_norm": 0.7398180961608887,
+      "learning_rate": 1.8771665211293696e-05,
+      "loss": 0.818,
+      "step": 23610
+    },
+    {
+      "epoch": 1.875012403500764,
+      "grad_norm": 0.8387320637702942,
+      "learning_rate": 1.8758434547908232e-05,
+      "loss": 0.8067,
+      "step": 23620
+    },
+    {
+      "epoch": 1.8758062275496636,
+      "grad_norm": 0.8745064735412598,
+      "learning_rate": 1.874520388452277e-05,
+      "loss": 0.7378,
+      "step": 23630
+    },
+    {
+      "epoch": 1.876600051598563,
+      "grad_norm": 0.7294278740882874,
+      "learning_rate": 1.873197322113731e-05,
+      "loss": 0.8538,
+      "step": 23640
+    },
+    {
+      "epoch": 1.8773938756474626,
+      "grad_norm": 0.8370188474655151,
+      "learning_rate": 1.871874255775185e-05,
+      "loss": 0.837,
+      "step": 23650
+    },
+    {
+      "epoch": 1.8781876996963622,
+      "grad_norm": 0.8423630595207214,
+      "learning_rate": 1.8705511894366384e-05,
+      "loss": 0.7942,
+      "step": 23660
+    },
+    {
+      "epoch": 1.8789815237452618,
+      "grad_norm": 0.8420581817626953,
+      "learning_rate": 1.8692281230980923e-05,
+      "loss": 0.8273,
+      "step": 23670
+    },
+    {
+      "epoch": 1.8797753477941614,
+      "grad_norm": 0.9311301708221436,
+      "learning_rate": 1.867905056759546e-05,
+      "loss": 0.832,
+      "step": 23680
+    },
+    {
+      "epoch": 1.880569171843061,
+      "grad_norm": 0.7972429990768433,
+      "learning_rate": 1.8665819904209998e-05,
+      "loss": 0.8131,
+      "step": 23690
+    },
+    {
+      "epoch": 1.8813629958919607,
+      "grad_norm": 0.7214037775993347,
+      "learning_rate": 1.8652589240824536e-05,
+      "loss": 0.8298,
+      "step": 23700
+    },
+    {
+      "epoch": 1.8821568199408603,
+      "grad_norm": 0.8672298789024353,
+      "learning_rate": 1.8639358577439072e-05,
+      "loss": 0.8283,
+      "step": 23710
+    },
+    {
+      "epoch": 1.8829506439897596,
+      "grad_norm": 0.8380793333053589,
+      "learning_rate": 1.862612791405361e-05,
+      "loss": 0.7733,
+      "step": 23720
+    },
+    {
+      "epoch": 1.8837444680386592,
+      "grad_norm": 0.8176583051681519,
+      "learning_rate": 1.861289725066815e-05,
+      "loss": 0.8774,
+      "step": 23730
+    },
+    {
+      "epoch": 1.8845382920875586,
+      "grad_norm": 0.8315280079841614,
+      "learning_rate": 1.859966658728269e-05,
+      "loss": 0.8133,
+      "step": 23740
+    },
+    {
+      "epoch": 1.8853321161364582,
+      "grad_norm": 0.7591937184333801,
+      "learning_rate": 1.8586435923897224e-05,
+      "loss": 0.8648,
+      "step": 23750
+    },
+    {
+      "epoch": 1.8861259401853578,
+      "grad_norm": 0.9065371155738831,
+      "learning_rate": 1.8573205260511763e-05,
+      "loss": 0.7526,
+      "step": 23760
+    },
+    {
+      "epoch": 1.8869197642342574,
+      "grad_norm": 0.8125611543655396,
+      "learning_rate": 1.8559974597126302e-05,
+      "loss": 0.7813,
+      "step": 23770
+    },
+    {
+      "epoch": 1.887713588283157,
+      "grad_norm": 0.7974992990493774,
+      "learning_rate": 1.854674393374084e-05,
+      "loss": 0.8515,
+      "step": 23780
+    },
+    {
+      "epoch": 1.8885074123320567,
+      "grad_norm": 0.6835601925849915,
+      "learning_rate": 1.8533513270355377e-05,
+      "loss": 0.8259,
+      "step": 23790
+    },
+    {
+      "epoch": 1.8893012363809563,
+      "grad_norm": 0.82981938123703,
+      "learning_rate": 1.8520282606969912e-05,
+      "loss": 0.7867,
+      "step": 23800
+    },
+    {
+      "epoch": 1.8900950604298559,
+      "grad_norm": 0.7962732911109924,
+      "learning_rate": 1.850705194358445e-05,
+      "loss": 0.8032,
+      "step": 23810
+    },
+    {
+      "epoch": 1.8908888844787552,
+      "grad_norm": 0.7996245622634888,
+      "learning_rate": 1.849382128019899e-05,
+      "loss": 0.7464,
+      "step": 23820
+    },
+    {
+      "epoch": 1.8916827085276549,
+      "grad_norm": 0.8889716267585754,
+      "learning_rate": 1.848059061681353e-05,
+      "loss": 0.7978,
+      "step": 23830
+    },
+    {
+      "epoch": 1.8924765325765545,
+      "grad_norm": 0.8882633447647095,
+      "learning_rate": 1.8467359953428065e-05,
+      "loss": 0.7114,
+      "step": 23840
+    },
+    {
+      "epoch": 1.8932703566254538,
+      "grad_norm": 0.8519427180290222,
+      "learning_rate": 1.8454129290042603e-05,
+      "loss": 0.8326,
+      "step": 23850
+    },
+    {
+      "epoch": 1.8940641806743534,
+      "grad_norm": 1.117575764656067,
+      "learning_rate": 1.8440898626657142e-05,
+      "loss": 0.8373,
+      "step": 23860
+    },
+    {
+      "epoch": 1.894858004723253,
+      "grad_norm": 0.871019184589386,
+      "learning_rate": 1.842766796327168e-05,
+      "loss": 0.8082,
+      "step": 23870
+    },
+    {
+      "epoch": 1.8956518287721527,
+      "grad_norm": 0.8169146776199341,
+      "learning_rate": 1.8414437299886217e-05,
+      "loss": 0.7634,
+      "step": 23880
+    },
+    {
+      "epoch": 1.8964456528210523,
+      "grad_norm": 0.8163456320762634,
+      "learning_rate": 1.8401206636500756e-05,
+      "loss": 0.8704,
+      "step": 23890
+    },
+    {
+      "epoch": 1.8972394768699519,
+      "grad_norm": 0.912982165813446,
+      "learning_rate": 1.838797597311529e-05,
+      "loss": 0.7831,
+      "step": 23900
+    },
+    {
+      "epoch": 1.8980333009188515,
+      "grad_norm": 0.8707506656646729,
+      "learning_rate": 1.837474530972983e-05,
+      "loss": 0.7522,
+      "step": 23910
+    },
+    {
+      "epoch": 1.8988271249677509,
+      "grad_norm": 0.7773687839508057,
+      "learning_rate": 1.836151464634437e-05,
+      "loss": 0.8301,
+      "step": 23920
+    },
+    {
+      "epoch": 1.8996209490166505,
+      "grad_norm": 0.846040666103363,
+      "learning_rate": 1.8348283982958905e-05,
+      "loss": 0.8274,
+      "step": 23930
+    },
+    {
+      "epoch": 1.90041477306555,
+      "grad_norm": 0.890678882598877,
+      "learning_rate": 1.8335053319573444e-05,
+      "loss": 0.8097,
+      "step": 23940
+    },
+    {
+      "epoch": 1.9012085971144495,
+      "grad_norm": 0.8786852955818176,
+      "learning_rate": 1.8321822656187983e-05,
+      "loss": 0.8448,
+      "step": 23950
+    },
+    {
+      "epoch": 1.902002421163349,
+      "grad_norm": 0.8807387948036194,
+      "learning_rate": 1.830859199280252e-05,
+      "loss": 0.8331,
+      "step": 23960
+    },
+    {
+      "epoch": 1.9027962452122487,
+      "grad_norm": 0.6707026958465576,
+      "learning_rate": 1.8295361329417057e-05,
+      "loss": 0.8109,
+      "step": 23970
+    },
+    {
+      "epoch": 1.9035900692611483,
+      "grad_norm": 0.9399641156196594,
+      "learning_rate": 1.8282130666031596e-05,
+      "loss": 0.7463,
+      "step": 23980
+    },
+    {
+      "epoch": 1.9043838933100479,
+      "grad_norm": 0.6597705483436584,
+      "learning_rate": 1.8268900002646135e-05,
+      "loss": 0.8391,
+      "step": 23990
+    },
+    {
+      "epoch": 1.9051777173589475,
+      "grad_norm": 0.869948148727417,
+      "learning_rate": 1.8255669339260674e-05,
+      "loss": 0.8085,
+      "step": 24000
+    },
+    {
+      "epoch": 1.905971541407847,
+      "grad_norm": 0.9739574193954468,
+      "learning_rate": 1.824243867587521e-05,
+      "loss": 0.8292,
+      "step": 24010
+    },
+    {
+      "epoch": 1.9067653654567465,
+      "grad_norm": 0.8347158432006836,
+      "learning_rate": 1.8229208012489745e-05,
+      "loss": 0.8148,
+      "step": 24020
+    },
+    {
+      "epoch": 1.907559189505646,
+      "grad_norm": 0.8817285299301147,
+      "learning_rate": 1.8215977349104284e-05,
+      "loss": 0.8662,
+      "step": 24030
+    },
+    {
+      "epoch": 1.9083530135545457,
+      "grad_norm": 0.8079895377159119,
+      "learning_rate": 1.8202746685718823e-05,
+      "loss": 0.8082,
+      "step": 24040
+    },
+    {
+      "epoch": 1.909146837603445,
+      "grad_norm": 0.9502777457237244,
+      "learning_rate": 1.818951602233336e-05,
+      "loss": 0.8375,
+      "step": 24050
+    },
+    {
+      "epoch": 1.9099406616523447,
+      "grad_norm": 0.7885468602180481,
+      "learning_rate": 1.8176285358947897e-05,
+      "loss": 0.7887,
+      "step": 24060
+    },
+    {
+      "epoch": 1.9107344857012443,
+      "grad_norm": 0.861359179019928,
+      "learning_rate": 1.8163054695562436e-05,
+      "loss": 0.8925,
+      "step": 24070
+    },
+    {
+      "epoch": 1.9115283097501439,
+      "grad_norm": 0.8542492389678955,
+      "learning_rate": 1.8149824032176975e-05,
+      "loss": 0.7872,
+      "step": 24080
+    },
+    {
+      "epoch": 1.9123221337990435,
+      "grad_norm": 0.97325199842453,
+      "learning_rate": 1.8136593368791514e-05,
+      "loss": 0.8129,
+      "step": 24090
+    },
+    {
+      "epoch": 1.913115957847943,
+      "grad_norm": 0.8216763734817505,
+      "learning_rate": 1.812336270540605e-05,
+      "loss": 0.8226,
+      "step": 24100
+    },
+    {
+      "epoch": 1.9139097818968427,
+      "grad_norm": 0.8220085501670837,
+      "learning_rate": 1.811013204202059e-05,
+      "loss": 0.8222,
+      "step": 24110
+    },
+    {
+      "epoch": 1.914703605945742,
+      "grad_norm": 0.8182690739631653,
+      "learning_rate": 1.8096901378635124e-05,
+      "loss": 0.843,
+      "step": 24120
+    },
+    {
+      "epoch": 1.9154974299946417,
+      "grad_norm": 1.0019487142562866,
+      "learning_rate": 1.8083670715249666e-05,
+      "loss": 0.826,
+      "step": 24130
+    },
+    {
+      "epoch": 1.9162912540435413,
+      "grad_norm": 0.9827296137809753,
+      "learning_rate": 1.8070440051864202e-05,
+      "loss": 0.876,
+      "step": 24140
+    },
+    {
+      "epoch": 1.9170850780924407,
+      "grad_norm": 0.848641037940979,
+      "learning_rate": 1.8057209388478737e-05,
+      "loss": 0.7753,
+      "step": 24150
+    },
+    {
+      "epoch": 1.9178789021413403,
+      "grad_norm": 0.9888964891433716,
+      "learning_rate": 1.8043978725093276e-05,
+      "loss": 0.7491,
+      "step": 24160
+    },
+    {
+      "epoch": 1.9186727261902399,
+      "grad_norm": 0.8973375558853149,
+      "learning_rate": 1.8030748061707815e-05,
+      "loss": 0.7887,
+      "step": 24170
+    },
+    {
+      "epoch": 1.9194665502391395,
+      "grad_norm": 0.8632115721702576,
+      "learning_rate": 1.8017517398322354e-05,
+      "loss": 0.8052,
+      "step": 24180
+    },
+    {
+      "epoch": 1.920260374288039,
+      "grad_norm": 0.9406533241271973,
+      "learning_rate": 1.800428673493689e-05,
+      "loss": 0.7834,
+      "step": 24190
+    },
+    {
+      "epoch": 1.9210541983369387,
+      "grad_norm": 0.7621222734451294,
+      "learning_rate": 1.799105607155143e-05,
+      "loss": 0.8211,
+      "step": 24200
+    },
+    {
+      "epoch": 1.9218480223858383,
+      "grad_norm": 0.7019477486610413,
+      "learning_rate": 1.7977825408165964e-05,
+      "loss": 0.7818,
+      "step": 24210
+    },
+    {
+      "epoch": 1.922641846434738,
+      "grad_norm": 0.7617840766906738,
+      "learning_rate": 1.7964594744780506e-05,
+      "loss": 0.7691,
+      "step": 24220
+    },
+    {
+      "epoch": 1.9234356704836373,
+      "grad_norm": 0.8671677112579346,
+      "learning_rate": 1.7951364081395042e-05,
+      "loss": 0.8076,
+      "step": 24230
+    },
+    {
+      "epoch": 1.9242294945325369,
+      "grad_norm": 0.7963345646858215,
+      "learning_rate": 1.793813341800958e-05,
+      "loss": 0.8549,
+      "step": 24240
+    },
+    {
+      "epoch": 1.9250233185814363,
+      "grad_norm": 0.7418803572654724,
+      "learning_rate": 1.7924902754624116e-05,
+      "loss": 0.8776,
+      "step": 24250
+    },
+    {
+      "epoch": 1.9258171426303359,
+      "grad_norm": 0.991839587688446,
+      "learning_rate": 1.7911672091238655e-05,
+      "loss": 0.8098,
+      "step": 24260
+    },
+    {
+      "epoch": 1.9266109666792355,
+      "grad_norm": 0.8886794447898865,
+      "learning_rate": 1.7898441427853194e-05,
+      "loss": 0.7906,
+      "step": 24270
+    },
+    {
+      "epoch": 1.927404790728135,
+      "grad_norm": 0.8058855533599854,
+      "learning_rate": 1.788521076446773e-05,
+      "loss": 0.7909,
+      "step": 24280
+    },
+    {
+      "epoch": 1.9281986147770347,
+      "grad_norm": 0.9611883163452148,
+      "learning_rate": 1.787198010108227e-05,
+      "loss": 0.7466,
+      "step": 24290
+    },
+    {
+      "epoch": 1.9289924388259343,
+      "grad_norm": 0.9158162474632263,
+      "learning_rate": 1.7858749437696808e-05,
+      "loss": 0.772,
+      "step": 24300
+    },
+    {
+      "epoch": 1.929786262874834,
+      "grad_norm": 0.7327207922935486,
+      "learning_rate": 1.7845518774311347e-05,
+      "loss": 0.832,
+      "step": 24310
+    },
+    {
+      "epoch": 1.9305800869237335,
+      "grad_norm": 0.8581303954124451,
+      "learning_rate": 1.7832288110925882e-05,
+      "loss": 0.8193,
+      "step": 24320
+    },
+    {
+      "epoch": 1.931373910972633,
+      "grad_norm": 0.8568522930145264,
+      "learning_rate": 1.781905744754042e-05,
+      "loss": 0.8105,
+      "step": 24330
+    },
+    {
+      "epoch": 1.9321677350215325,
+      "grad_norm": 0.9713951945304871,
+      "learning_rate": 1.7805826784154957e-05,
+      "loss": 0.7699,
+      "step": 24340
+    },
+    {
+      "epoch": 1.9329615590704319,
+      "grad_norm": 0.8259567618370056,
+      "learning_rate": 1.77925961207695e-05,
+      "loss": 0.8278,
+      "step": 24350
+    },
+    {
+      "epoch": 1.9337553831193315,
+      "grad_norm": 0.9013030529022217,
+      "learning_rate": 1.7779365457384035e-05,
+      "loss": 0.8677,
+      "step": 24360
+    },
+    {
+      "epoch": 1.934549207168231,
+      "grad_norm": 0.8050611019134521,
+      "learning_rate": 1.776613479399857e-05,
+      "loss": 0.7818,
+      "step": 24370
+    },
+    {
+      "epoch": 1.9353430312171307,
+      "grad_norm": 0.8828538060188293,
+      "learning_rate": 1.775290413061311e-05,
+      "loss": 0.8062,
+      "step": 24380
+    },
+    {
+      "epoch": 1.9361368552660303,
+      "grad_norm": 0.9584046602249146,
+      "learning_rate": 1.7739673467227648e-05,
+      "loss": 0.8004,
+      "step": 24390
+    },
+    {
+      "epoch": 1.93693067931493,
+      "grad_norm": 1.0166678428649902,
+      "learning_rate": 1.7726442803842187e-05,
+      "loss": 0.8063,
+      "step": 24400
+    },
+    {
+      "epoch": 1.9377245033638295,
+      "grad_norm": 0.8781585693359375,
+      "learning_rate": 1.7713212140456722e-05,
+      "loss": 0.8428,
+      "step": 24410
+    },
+    {
+      "epoch": 1.9385183274127291,
+      "grad_norm": 0.9329415559768677,
+      "learning_rate": 1.769998147707126e-05,
+      "loss": 0.8114,
+      "step": 24420
+    },
+    {
+      "epoch": 1.9393121514616285,
+      "grad_norm": 0.9052295684814453,
+      "learning_rate": 1.7686750813685797e-05,
+      "loss": 0.8206,
+      "step": 24430
+    },
+    {
+      "epoch": 1.940105975510528,
+      "grad_norm": 0.8311532139778137,
+      "learning_rate": 1.767352015030034e-05,
+      "loss": 0.83,
+      "step": 24440
+    },
+    {
+      "epoch": 1.9408997995594277,
+      "grad_norm": 0.8196767568588257,
+      "learning_rate": 1.7660289486914875e-05,
+      "loss": 0.8425,
+      "step": 24450
+    },
+    {
+      "epoch": 1.941693623608327,
+      "grad_norm": 0.7230108976364136,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 0.8507,
+      "step": 24460
+    },
+    {
+      "epoch": 1.9424874476572267,
+      "grad_norm": 1.056864857673645,
+      "learning_rate": 1.763382816014395e-05,
+      "loss": 0.7798,
+      "step": 24470
+    },
+    {
+      "epoch": 1.9432812717061263,
+      "grad_norm": 1.0241467952728271,
+      "learning_rate": 1.7620597496758488e-05,
+      "loss": 0.8083,
+      "step": 24480
+    },
+    {
+      "epoch": 1.944075095755026,
+      "grad_norm": 1.016441822052002,
+      "learning_rate": 1.7607366833373027e-05,
+      "loss": 0.8201,
+      "step": 24490
+    },
+    {
+      "epoch": 1.9448689198039255,
+      "grad_norm": 0.9095987677574158,
+      "learning_rate": 1.7594136169987563e-05,
+      "loss": 0.8211,
+      "step": 24500
+    },
+    {
+      "epoch": 1.9456627438528251,
+      "grad_norm": 0.7234585881233215,
+      "learning_rate": 1.75809055066021e-05,
+      "loss": 0.8225,
+      "step": 24510
+    },
+    {
+      "epoch": 1.9464565679017247,
+      "grad_norm": 0.8463301062583923,
+      "learning_rate": 1.756767484321664e-05,
+      "loss": 0.8663,
+      "step": 24520
+    },
+    {
+      "epoch": 1.947250391950624,
+      "grad_norm": 0.8381713628768921,
+      "learning_rate": 1.755444417983118e-05,
+      "loss": 0.8309,
+      "step": 24530
+    },
+    {
+      "epoch": 1.9480442159995237,
+      "grad_norm": 1.2629505395889282,
+      "learning_rate": 1.7541213516445715e-05,
+      "loss": 0.8193,
+      "step": 24540
+    },
+    {
+      "epoch": 1.9488380400484233,
+      "grad_norm": 0.889024555683136,
+      "learning_rate": 1.7527982853060254e-05,
+      "loss": 0.7851,
+      "step": 24550
+    },
+    {
+      "epoch": 1.9496318640973227,
+      "grad_norm": 1.1809180974960327,
+      "learning_rate": 1.751475218967479e-05,
+      "loss": 0.8625,
+      "step": 24560
+    },
+    {
+      "epoch": 1.9504256881462223,
+      "grad_norm": 0.973340630531311,
+      "learning_rate": 1.750152152628933e-05,
+      "loss": 0.7819,
+      "step": 24570
+    },
+    {
+      "epoch": 1.951219512195122,
+      "grad_norm": 0.9036343693733215,
+      "learning_rate": 1.7488290862903867e-05,
+      "loss": 0.7523,
+      "step": 24580
+    },
+    {
+      "epoch": 1.9520133362440215,
+      "grad_norm": 0.9107938408851624,
+      "learning_rate": 1.7475060199518406e-05,
+      "loss": 0.8277,
+      "step": 24590
+    },
+    {
+      "epoch": 1.9528071602929211,
+      "grad_norm": 0.7978836297988892,
+      "learning_rate": 1.746182953613294e-05,
+      "loss": 0.8058,
+      "step": 24600
+    },
+    {
+      "epoch": 1.9536009843418207,
+      "grad_norm": 0.9246600270271301,
+      "learning_rate": 1.744859887274748e-05,
+      "loss": 0.862,
+      "step": 24610
+    },
+    {
+      "epoch": 1.9543948083907203,
+      "grad_norm": 0.96808922290802,
+      "learning_rate": 1.743536820936202e-05,
+      "loss": 0.9027,
+      "step": 24620
+    },
+    {
+      "epoch": 1.9551886324396197,
+      "grad_norm": 0.7712015509605408,
+      "learning_rate": 1.7422137545976555e-05,
+      "loss": 0.836,
+      "step": 24630
+    },
+    {
+      "epoch": 1.9559824564885193,
+      "grad_norm": 0.9321126937866211,
+      "learning_rate": 1.7408906882591094e-05,
+      "loss": 0.836,
+      "step": 24640
+    },
+    {
+      "epoch": 1.956776280537419,
+      "grad_norm": 0.8697831630706787,
+      "learning_rate": 1.739567621920563e-05,
+      "loss": 0.8271,
+      "step": 24650
+    },
+    {
+      "epoch": 1.9575701045863183,
+      "grad_norm": 0.8572379350662231,
+      "learning_rate": 1.7382445555820172e-05,
+      "loss": 0.7872,
+      "step": 24660
+    },
+    {
+      "epoch": 1.958363928635218,
+      "grad_norm": 0.7709526419639587,
+      "learning_rate": 1.7369214892434707e-05,
+      "loss": 0.8315,
+      "step": 24670
+    },
+    {
+      "epoch": 1.9591577526841175,
+      "grad_norm": 0.867958128452301,
+      "learning_rate": 1.7355984229049246e-05,
+      "loss": 0.7875,
+      "step": 24680
+    },
+    {
+      "epoch": 1.9599515767330171,
+      "grad_norm": 0.883253276348114,
+      "learning_rate": 1.7342753565663782e-05,
+      "loss": 0.8871,
+      "step": 24690
+    },
+    {
+      "epoch": 1.9607454007819167,
+      "grad_norm": 0.8314810395240784,
+      "learning_rate": 1.732952290227832e-05,
+      "loss": 0.7559,
+      "step": 24700
+    },
+    {
+      "epoch": 1.9615392248308163,
+      "grad_norm": 0.7700093984603882,
+      "learning_rate": 1.731629223889286e-05,
+      "loss": 0.8198,
+      "step": 24710
+    },
+    {
+      "epoch": 1.962333048879716,
+      "grad_norm": 0.7861685156822205,
+      "learning_rate": 1.7303061575507395e-05,
+      "loss": 0.8823,
+      "step": 24720
+    },
+    {
+      "epoch": 1.9631268729286153,
+      "grad_norm": 0.9376586675643921,
+      "learning_rate": 1.7289830912121934e-05,
+      "loss": 0.8059,
+      "step": 24730
+    },
+    {
+      "epoch": 1.963920696977515,
+      "grad_norm": 0.9346411228179932,
+      "learning_rate": 1.7276600248736473e-05,
+      "loss": 0.8167,
+      "step": 24740
+    },
+    {
+      "epoch": 1.9647145210264145,
+      "grad_norm": 0.7460438013076782,
+      "learning_rate": 1.7263369585351012e-05,
+      "loss": 0.7703,
+      "step": 24750
+    },
+    {
+      "epoch": 1.965508345075314,
+      "grad_norm": 0.7430590391159058,
+      "learning_rate": 1.7250138921965548e-05,
+      "loss": 0.7716,
+      "step": 24760
+    },
+    {
+      "epoch": 1.9663021691242135,
+      "grad_norm": 0.9722541570663452,
+      "learning_rate": 1.7236908258580086e-05,
+      "loss": 0.7665,
+      "step": 24770
+    },
+    {
+      "epoch": 1.9670959931731131,
+      "grad_norm": 0.8619949817657471,
+      "learning_rate": 1.7223677595194622e-05,
+      "loss": 0.7947,
+      "step": 24780
+    },
+    {
+      "epoch": 1.9678898172220127,
+      "grad_norm": 0.864704966545105,
+      "learning_rate": 1.7210446931809164e-05,
+      "loss": 0.8233,
+      "step": 24790
+    },
+    {
+      "epoch": 1.9686836412709123,
+      "grad_norm": 0.8963067531585693,
+      "learning_rate": 1.71972162684237e-05,
+      "loss": 0.8249,
+      "step": 24800
+    },
+    {
+      "epoch": 1.969477465319812,
+      "grad_norm": 0.7496664524078369,
+      "learning_rate": 1.718398560503824e-05,
+      "loss": 0.7904,
+      "step": 24810
+    },
+    {
+      "epoch": 1.9702712893687115,
+      "grad_norm": 0.9141893982887268,
+      "learning_rate": 1.7170754941652774e-05,
+      "loss": 0.8094,
+      "step": 24820
+    },
+    {
+      "epoch": 1.9710651134176111,
+      "grad_norm": 0.8252711296081543,
+      "learning_rate": 1.7157524278267313e-05,
+      "loss": 0.744,
+      "step": 24830
+    },
+    {
+      "epoch": 1.9718589374665105,
+      "grad_norm": 0.7899823188781738,
+      "learning_rate": 1.7144293614881852e-05,
+      "loss": 0.7799,
+      "step": 24840
+    },
+    {
+      "epoch": 1.9726527615154101,
+      "grad_norm": 0.7791532874107361,
+      "learning_rate": 1.7131062951496388e-05,
+      "loss": 0.8205,
+      "step": 24850
+    },
+    {
+      "epoch": 1.9734465855643095,
+      "grad_norm": 0.8147314786911011,
+      "learning_rate": 1.7117832288110927e-05,
+      "loss": 0.7866,
+      "step": 24860
+    },
+    {
+      "epoch": 1.9742404096132091,
+      "grad_norm": 0.9811384677886963,
+      "learning_rate": 1.7104601624725462e-05,
+      "loss": 0.8046,
+      "step": 24870
+    },
+    {
+      "epoch": 1.9750342336621087,
+      "grad_norm": 0.6570988893508911,
+      "learning_rate": 1.7091370961340004e-05,
+      "loss": 0.8115,
+      "step": 24880
+    },
+    {
+      "epoch": 1.9758280577110083,
+      "grad_norm": 0.8411725759506226,
+      "learning_rate": 1.707814029795454e-05,
+      "loss": 0.8163,
+      "step": 24890
+    },
+    {
+      "epoch": 1.976621881759908,
+      "grad_norm": 0.9080935120582581,
+      "learning_rate": 1.706490963456908e-05,
+      "loss": 0.7894,
+      "step": 24900
+    },
+    {
+      "epoch": 1.9774157058088075,
+      "grad_norm": 0.8399326801300049,
+      "learning_rate": 1.7051678971183615e-05,
+      "loss": 0.8409,
+      "step": 24910
+    },
+    {
+      "epoch": 1.9782095298577071,
+      "grad_norm": 0.7995646595954895,
+      "learning_rate": 1.7038448307798153e-05,
+      "loss": 0.8039,
+      "step": 24920
+    },
+    {
+      "epoch": 1.9790033539066068,
+      "grad_norm": 0.7588172554969788,
+      "learning_rate": 1.7025217644412692e-05,
+      "loss": 0.8694,
+      "step": 24930
+    },
+    {
+      "epoch": 1.9797971779555061,
+      "grad_norm": 0.8422114253044128,
+      "learning_rate": 1.701198698102723e-05,
+      "loss": 0.8497,
+      "step": 24940
+    },
+    {
+      "epoch": 1.9805910020044057,
+      "grad_norm": 0.9750681519508362,
+      "learning_rate": 1.6998756317641767e-05,
+      "loss": 0.809,
+      "step": 24950
+    },
+    {
+      "epoch": 1.9813848260533051,
+      "grad_norm": 0.8295944929122925,
+      "learning_rate": 1.6985525654256302e-05,
+      "loss": 0.828,
+      "step": 24960
+    },
+    {
+      "epoch": 1.9821786501022047,
+      "grad_norm": 0.7455227375030518,
+      "learning_rate": 1.6972294990870845e-05,
+      "loss": 0.774,
+      "step": 24970
+    },
+    {
+      "epoch": 1.9829724741511043,
+      "grad_norm": 0.7973991632461548,
+      "learning_rate": 1.695906432748538e-05,
+      "loss": 0.833,
+      "step": 24980
+    },
+    {
+      "epoch": 1.983766298200004,
+      "grad_norm": 0.9970287084579468,
+      "learning_rate": 1.694583366409992e-05,
+      "loss": 0.7929,
+      "step": 24990
+    },
+    {
+      "epoch": 1.9845601222489035,
+      "grad_norm": 1.006111741065979,
+      "learning_rate": 1.6932603000714455e-05,
+      "loss": 0.7835,
+      "step": 25000
+    },
+    {
+      "epoch": 1.9853539462978032,
+      "grad_norm": 0.7539806962013245,
+      "learning_rate": 1.6919372337328994e-05,
+      "loss": 0.791,
+      "step": 25010
+    },
+    {
+      "epoch": 1.9861477703467028,
+      "grad_norm": 0.6668902635574341,
+      "learning_rate": 1.6906141673943533e-05,
+      "loss": 0.8413,
+      "step": 25020
+    },
+    {
+      "epoch": 1.9869415943956024,
+      "grad_norm": 0.7992756366729736,
+      "learning_rate": 1.689291101055807e-05,
+      "loss": 0.8181,
+      "step": 25030
+    },
+    {
+      "epoch": 1.9877354184445017,
+      "grad_norm": 0.836247980594635,
+      "learning_rate": 1.6879680347172607e-05,
+      "loss": 0.8255,
+      "step": 25040
+    },
+    {
+      "epoch": 1.9885292424934013,
+      "grad_norm": 0.701686680316925,
+      "learning_rate": 1.6866449683787146e-05,
+      "loss": 0.8523,
+      "step": 25050
+    },
+    {
+      "epoch": 1.9893230665423007,
+      "grad_norm": 0.8680205941200256,
+      "learning_rate": 1.6853219020401685e-05,
+      "loss": 0.8737,
+      "step": 25060
+    },
+    {
+      "epoch": 1.9901168905912003,
+      "grad_norm": 1.0040700435638428,
+      "learning_rate": 1.683998835701622e-05,
+      "loss": 0.8068,
+      "step": 25070
+    },
+    {
+      "epoch": 1.9909107146401,
+      "grad_norm": 0.7312654852867126,
+      "learning_rate": 1.682675769363076e-05,
+      "loss": 0.7744,
+      "step": 25080
+    },
+    {
+      "epoch": 1.9917045386889995,
+      "grad_norm": 0.8141999244689941,
+      "learning_rate": 1.6813527030245295e-05,
+      "loss": 0.8031,
+      "step": 25090
+    },
+    {
+      "epoch": 1.9924983627378992,
+      "grad_norm": 0.9395736455917358,
+      "learning_rate": 1.6800296366859837e-05,
+      "loss": 0.8556,
+      "step": 25100
+    },
+    {
+      "epoch": 1.9932921867867988,
+      "grad_norm": 1.0903936624526978,
+      "learning_rate": 1.6788388769812918e-05,
+      "loss": 0.7715,
+      "step": 25110
+    },
+    {
+      "epoch": 1.9940860108356984,
+      "grad_norm": 0.9055798053741455,
+      "learning_rate": 1.6775158106427457e-05,
+      "loss": 0.8161,
+      "step": 25120
+    },
+    {
+      "epoch": 1.994879834884598,
+      "grad_norm": 1.0567915439605713,
+      "learning_rate": 1.6761927443041993e-05,
+      "loss": 0.7872,
+      "step": 25130
+    },
+    {
+      "epoch": 1.9956736589334974,
+      "grad_norm": 0.8715507388114929,
+      "learning_rate": 1.6748696779656535e-05,
+      "loss": 0.847,
+      "step": 25140
+    },
+    {
+      "epoch": 1.996467482982397,
+      "grad_norm": 0.9350098371505737,
+      "learning_rate": 1.673546611627107e-05,
+      "loss": 0.7932,
+      "step": 25150
+    },
+    {
+      "epoch": 1.9972613070312966,
+      "grad_norm": 0.8718578219413757,
+      "learning_rate": 1.672223545288561e-05,
+      "loss": 0.8038,
+      "step": 25160
+    },
+    {
+      "epoch": 1.998055131080196,
+      "grad_norm": 0.987808883190155,
+      "learning_rate": 1.6709004789500145e-05,
+      "loss": 0.8141,
+      "step": 25170
+    },
+    {
+      "epoch": 1.9988489551290956,
+      "grad_norm": 0.8378719091415405,
+      "learning_rate": 1.6695774126114684e-05,
+      "loss": 0.7715,
+      "step": 25180
+    },
+    {
+      "epoch": 1.9996427791779952,
+      "grad_norm": 0.8598037958145142,
+      "learning_rate": 1.6682543462729223e-05,
+      "loss": 0.7998,
+      "step": 25190
+    },
+    {
+      "epoch": 2.0004366032268948,
+      "grad_norm": 0.8673446774482727,
+      "learning_rate": 1.666931279934376e-05,
+      "loss": 0.8147,
+      "step": 25200
+    },
+    {
+      "epoch": 2.0012304272757944,
+      "grad_norm": 0.8430289030075073,
+      "learning_rate": 1.6656082135958297e-05,
+      "loss": 0.801,
+      "step": 25210
+    },
+    {
+      "epoch": 2.002024251324694,
+      "grad_norm": 0.8227368593215942,
+      "learning_rate": 1.6642851472572836e-05,
+      "loss": 0.7521,
+      "step": 25220
+    },
+    {
+      "epoch": 2.0028180753735936,
+      "grad_norm": 0.8883048295974731,
+      "learning_rate": 1.6629620809187375e-05,
+      "loss": 0.8145,
+      "step": 25230
+    },
+    {
+      "epoch": 2.003611899422493,
+      "grad_norm": 0.6865272521972656,
+      "learning_rate": 1.661639014580191e-05,
+      "loss": 0.776,
+      "step": 25240
+    },
+    {
+      "epoch": 2.004405723471393,
+      "grad_norm": 0.8548721075057983,
+      "learning_rate": 1.660315948241645e-05,
+      "loss": 0.7732,
+      "step": 25250
+    },
+    {
+      "epoch": 2.005199547520292,
+      "grad_norm": 0.8403987884521484,
+      "learning_rate": 1.6589928819030985e-05,
+      "loss": 0.7927,
+      "step": 25260
+    },
+    {
+      "epoch": 2.0059933715691916,
+      "grad_norm": 0.7534387111663818,
+      "learning_rate": 1.6576698155645527e-05,
+      "loss": 0.8369,
+      "step": 25270
+    },
+    {
+      "epoch": 2.006787195618091,
+      "grad_norm": 0.8503125309944153,
+      "learning_rate": 1.6563467492260063e-05,
+      "loss": 0.7955,
+      "step": 25280
+    },
+    {
+      "epoch": 2.0075810196669908,
+      "grad_norm": 0.6855154633522034,
+      "learning_rate": 1.6550236828874602e-05,
+      "loss": 0.7958,
+      "step": 25290
+    },
+    {
+      "epoch": 2.0083748437158904,
+      "grad_norm": 0.7701917290687561,
+      "learning_rate": 1.6537006165489137e-05,
+      "loss": 0.7644,
+      "step": 25300
+    },
+    {
+      "epoch": 2.00916866776479,
+      "grad_norm": 0.9168187379837036,
+      "learning_rate": 1.6523775502103676e-05,
+      "loss": 0.8316,
+      "step": 25310
+    },
+    {
+      "epoch": 2.0099624918136896,
+      "grad_norm": 0.833348274230957,
+      "learning_rate": 1.6510544838718215e-05,
+      "loss": 0.7625,
+      "step": 25320
+    },
+    {
+      "epoch": 2.010756315862589,
+      "grad_norm": 0.8914290070533752,
+      "learning_rate": 1.649731417533275e-05,
+      "loss": 0.7627,
+      "step": 25330
+    },
+    {
+      "epoch": 2.011550139911489,
+      "grad_norm": 1.0038100481033325,
+      "learning_rate": 1.648408351194729e-05,
+      "loss": 0.784,
+      "step": 25340
+    },
+    {
+      "epoch": 2.0123439639603884,
+      "grad_norm": 0.8875173330307007,
+      "learning_rate": 1.6470852848561825e-05,
+      "loss": 0.8098,
+      "step": 25350
+    },
+    {
+      "epoch": 2.0131377880092876,
+      "grad_norm": 0.9379634857177734,
+      "learning_rate": 1.6457622185176368e-05,
+      "loss": 0.8373,
+      "step": 25360
+    },
+    {
+      "epoch": 2.013931612058187,
+      "grad_norm": 0.7726089954376221,
+      "learning_rate": 1.6444391521790903e-05,
+      "loss": 0.7727,
+      "step": 25370
+    },
+    {
+      "epoch": 2.0147254361070868,
+      "grad_norm": 0.7188129425048828,
+      "learning_rate": 1.6431160858405442e-05,
+      "loss": 0.7634,
+      "step": 25380
+    },
+    {
+      "epoch": 2.0155192601559864,
+      "grad_norm": 0.7986645698547363,
+      "learning_rate": 1.6417930195019978e-05,
+      "loss": 0.8123,
+      "step": 25390
+    },
+    {
+      "epoch": 2.016313084204886,
+      "grad_norm": 0.720370888710022,
+      "learning_rate": 1.6404699531634517e-05,
+      "loss": 0.7428,
+      "step": 25400
+    },
+    {
+      "epoch": 2.0171069082537856,
+      "grad_norm": 0.9025735259056091,
+      "learning_rate": 1.6391468868249055e-05,
+      "loss": 0.817,
+      "step": 25410
+    },
+    {
+      "epoch": 2.017900732302685,
+      "grad_norm": 0.8990045785903931,
+      "learning_rate": 1.637823820486359e-05,
+      "loss": 0.7254,
+      "step": 25420
+    },
+    {
+      "epoch": 2.018694556351585,
+      "grad_norm": 0.9403473138809204,
+      "learning_rate": 1.636500754147813e-05,
+      "loss": 0.7958,
+      "step": 25430
+    },
+    {
+      "epoch": 2.0194883804004844,
+      "grad_norm": 1.0497170686721802,
+      "learning_rate": 1.635177687809267e-05,
+      "loss": 0.7647,
+      "step": 25440
+    },
+    {
+      "epoch": 2.020282204449384,
+      "grad_norm": 1.1026662588119507,
+      "learning_rate": 1.6338546214707208e-05,
+      "loss": 0.7541,
+      "step": 25450
+    },
+    {
+      "epoch": 2.021076028498283,
+      "grad_norm": 0.9010255336761475,
+      "learning_rate": 1.6325315551321743e-05,
+      "loss": 0.7501,
+      "step": 25460
+    },
+    {
+      "epoch": 2.0218698525471828,
+      "grad_norm": 0.6994714736938477,
+      "learning_rate": 1.6312084887936282e-05,
+      "loss": 0.7985,
+      "step": 25470
+    },
+    {
+      "epoch": 2.0226636765960824,
+      "grad_norm": 0.7155155539512634,
+      "learning_rate": 1.6298854224550818e-05,
+      "loss": 0.7136,
+      "step": 25480
+    },
+    {
+      "epoch": 2.023457500644982,
+      "grad_norm": 0.9782127141952515,
+      "learning_rate": 1.628562356116536e-05,
+      "loss": 0.782,
+      "step": 25490
+    },
+    {
+      "epoch": 2.0242513246938816,
+      "grad_norm": 0.8731641173362732,
+      "learning_rate": 1.6272392897779896e-05,
+      "loss": 0.7801,
+      "step": 25500
+    },
+    {
+      "epoch": 2.025045148742781,
+      "grad_norm": 0.8580078482627869,
+      "learning_rate": 1.6259162234394435e-05,
+      "loss": 0.833,
+      "step": 25510
+    },
+    {
+      "epoch": 2.025838972791681,
+      "grad_norm": 0.9071916937828064,
+      "learning_rate": 1.624593157100897e-05,
+      "loss": 0.8026,
+      "step": 25520
+    },
+    {
+      "epoch": 2.0266327968405804,
+      "grad_norm": 0.9354455471038818,
+      "learning_rate": 1.623270090762351e-05,
+      "loss": 0.7792,
+      "step": 25530
+    },
+    {
+      "epoch": 2.02742662088948,
+      "grad_norm": 0.8776410222053528,
+      "learning_rate": 1.6219470244238048e-05,
+      "loss": 0.7958,
+      "step": 25540
+    },
+    {
+      "epoch": 2.0282204449383796,
+      "grad_norm": 0.7995261549949646,
+      "learning_rate": 1.6206239580852584e-05,
+      "loss": 0.7811,
+      "step": 25550
+    },
+    {
+      "epoch": 2.0290142689872788,
+      "grad_norm": 0.8553417325019836,
+      "learning_rate": 1.6193008917467122e-05,
+      "loss": 0.7713,
+      "step": 25560
+    },
+    {
+      "epoch": 2.0298080930361784,
+      "grad_norm": 0.9338173270225525,
+      "learning_rate": 1.6179778254081658e-05,
+      "loss": 0.7451,
+      "step": 25570
+    },
+    {
+      "epoch": 2.030601917085078,
+      "grad_norm": 0.9235241413116455,
+      "learning_rate": 1.61665475906962e-05,
+      "loss": 0.7126,
+      "step": 25580
+    },
+    {
+      "epoch": 2.0313957411339776,
+      "grad_norm": 0.864325761795044,
+      "learning_rate": 1.6153316927310736e-05,
+      "loss": 0.7988,
+      "step": 25590
+    },
+    {
+      "epoch": 2.032189565182877,
+      "grad_norm": 0.7986742854118347,
+      "learning_rate": 1.6140086263925275e-05,
+      "loss": 0.7959,
+      "step": 25600
+    },
+    {
+      "epoch": 2.032983389231777,
+      "grad_norm": 0.9655138254165649,
+      "learning_rate": 1.612685560053981e-05,
+      "loss": 0.8415,
+      "step": 25610
+    },
+    {
+      "epoch": 2.0337772132806764,
+      "grad_norm": 0.6859591603279114,
+      "learning_rate": 1.611362493715435e-05,
+      "loss": 0.7661,
+      "step": 25620
+    },
+    {
+      "epoch": 2.034571037329576,
+      "grad_norm": 0.7954578399658203,
+      "learning_rate": 1.6100394273768888e-05,
+      "loss": 0.811,
+      "step": 25630
+    },
+    {
+      "epoch": 2.0353648613784756,
+      "grad_norm": 0.9830471873283386,
+      "learning_rate": 1.6087163610383427e-05,
+      "loss": 0.7648,
+      "step": 25640
+    },
+    {
+      "epoch": 2.036158685427375,
+      "grad_norm": 0.9153994917869568,
+      "learning_rate": 1.6073932946997963e-05,
+      "loss": 0.7447,
+      "step": 25650
+    },
+    {
+      "epoch": 2.0369525094762744,
+      "grad_norm": 0.9104596972465515,
+      "learning_rate": 1.60607022836125e-05,
+      "loss": 0.8326,
+      "step": 25660
+    },
+    {
+      "epoch": 2.037746333525174,
+      "grad_norm": 0.8871384859085083,
+      "learning_rate": 1.604747162022704e-05,
+      "loss": 0.8012,
+      "step": 25670
+    },
+    {
+      "epoch": 2.0385401575740736,
+      "grad_norm": 0.800845205783844,
+      "learning_rate": 1.6034240956841576e-05,
+      "loss": 0.7772,
+      "step": 25680
+    },
+    {
+      "epoch": 2.039333981622973,
+      "grad_norm": 1.036254644393921,
+      "learning_rate": 1.6021010293456115e-05,
+      "loss": 0.7825,
+      "step": 25690
+    },
+    {
+      "epoch": 2.040127805671873,
+      "grad_norm": 0.8596107363700867,
+      "learning_rate": 1.600777963007065e-05,
+      "loss": 0.7614,
+      "step": 25700
+    },
+    {
+      "epoch": 2.0409216297207724,
+      "grad_norm": 0.7805830240249634,
+      "learning_rate": 1.5994548966685193e-05,
+      "loss": 0.8176,
+      "step": 25710
+    },
+    {
+      "epoch": 2.041715453769672,
+      "grad_norm": 1.0922433137893677,
+      "learning_rate": 1.598131830329973e-05,
+      "loss": 0.7965,
+      "step": 25720
+    },
+    {
+      "epoch": 2.0425092778185716,
+      "grad_norm": 0.7439164519309998,
+      "learning_rate": 1.5968087639914267e-05,
+      "loss": 0.8144,
+      "step": 25730
+    },
+    {
+      "epoch": 2.043303101867471,
+      "grad_norm": 0.7576305270195007,
+      "learning_rate": 1.5954856976528803e-05,
+      "loss": 0.8107,
+      "step": 25740
+    },
+    {
+      "epoch": 2.044096925916371,
+      "grad_norm": 0.7325326800346375,
+      "learning_rate": 1.5941626313143342e-05,
+      "loss": 0.7942,
+      "step": 25750
+    },
+    {
+      "epoch": 2.04489074996527,
+      "grad_norm": 0.7904483079910278,
+      "learning_rate": 1.592839564975788e-05,
+      "loss": 0.8301,
+      "step": 25760
+    },
+    {
+      "epoch": 2.0456845740141696,
+      "grad_norm": 0.8427342772483826,
+      "learning_rate": 1.5915164986372416e-05,
+      "loss": 0.7408,
+      "step": 25770
+    },
+    {
+      "epoch": 2.046478398063069,
+      "grad_norm": 0.9607779383659363,
+      "learning_rate": 1.5901934322986955e-05,
+      "loss": 0.7796,
+      "step": 25780
+    },
+    {
+      "epoch": 2.047272222111969,
+      "grad_norm": 0.9316497445106506,
+      "learning_rate": 1.588870365960149e-05,
+      "loss": 0.8237,
+      "step": 25790
+    },
+    {
+      "epoch": 2.0480660461608684,
+      "grad_norm": 0.8862087726593018,
+      "learning_rate": 1.5875472996216033e-05,
+      "loss": 0.7505,
+      "step": 25800
+    },
+    {
+      "epoch": 2.048859870209768,
+      "grad_norm": 0.9712191820144653,
+      "learning_rate": 1.586224233283057e-05,
+      "loss": 0.7987,
+      "step": 25810
+    },
+    {
+      "epoch": 2.0496536942586676,
+      "grad_norm": 0.8523909449577332,
+      "learning_rate": 1.5849011669445107e-05,
+      "loss": 0.7585,
+      "step": 25820
+    },
+    {
+      "epoch": 2.050447518307567,
+      "grad_norm": 0.9257214069366455,
+      "learning_rate": 1.5835781006059643e-05,
+      "loss": 0.7431,
+      "step": 25830
+    },
+    {
+      "epoch": 2.051241342356467,
+      "grad_norm": 0.9414131045341492,
+      "learning_rate": 1.5822550342674182e-05,
+      "loss": 0.7289,
+      "step": 25840
+    },
+    {
+      "epoch": 2.0520351664053664,
+      "grad_norm": 0.7903403639793396,
+      "learning_rate": 1.580931967928872e-05,
+      "loss": 0.7282,
+      "step": 25850
+    },
+    {
+      "epoch": 2.052828990454266,
+      "grad_norm": 0.7305004000663757,
+      "learning_rate": 1.579608901590326e-05,
+      "loss": 0.7352,
+      "step": 25860
+    },
+    {
+      "epoch": 2.053622814503165,
+      "grad_norm": 0.8797780275344849,
+      "learning_rate": 1.5782858352517795e-05,
+      "loss": 0.7377,
+      "step": 25870
+    },
+    {
+      "epoch": 2.054416638552065,
+      "grad_norm": 1.0097732543945312,
+      "learning_rate": 1.5769627689132334e-05,
+      "loss": 0.7732,
+      "step": 25880
+    },
+    {
+      "epoch": 2.0552104626009644,
+      "grad_norm": 0.8882811069488525,
+      "learning_rate": 1.5756397025746873e-05,
+      "loss": 0.8136,
+      "step": 25890
+    },
+    {
+      "epoch": 2.056004286649864,
+      "grad_norm": 0.7051288485527039,
+      "learning_rate": 1.574316636236141e-05,
+      "loss": 0.7464,
+      "step": 25900
+    },
+    {
+      "epoch": 2.0567981106987636,
+      "grad_norm": 0.9617118239402771,
+      "learning_rate": 1.5729935698975948e-05,
+      "loss": 0.7851,
+      "step": 25910
+    },
+    {
+      "epoch": 2.057591934747663,
+      "grad_norm": 0.8823207020759583,
+      "learning_rate": 1.5716705035590483e-05,
+      "loss": 0.8155,
+      "step": 25920
+    },
+    {
+      "epoch": 2.058385758796563,
+      "grad_norm": 1.0698038339614868,
+      "learning_rate": 1.5703474372205022e-05,
+      "loss": 0.798,
+      "step": 25930
+    },
+    {
+      "epoch": 2.0591795828454624,
+      "grad_norm": 1.0213465690612793,
+      "learning_rate": 1.569024370881956e-05,
+      "loss": 0.7397,
+      "step": 25940
+    },
+    {
+      "epoch": 2.059973406894362,
+      "grad_norm": 0.8252881765365601,
+      "learning_rate": 1.56770130454341e-05,
+      "loss": 0.7786,
+      "step": 25950
+    },
+    {
+      "epoch": 2.0607672309432616,
+      "grad_norm": 0.79705810546875,
+      "learning_rate": 1.5663782382048635e-05,
+      "loss": 0.7649,
+      "step": 25960
+    },
+    {
+      "epoch": 2.061561054992161,
+      "grad_norm": 0.7776365280151367,
+      "learning_rate": 1.5650551718663174e-05,
+      "loss": 0.7646,
+      "step": 25970
+    },
+    {
+      "epoch": 2.0623548790410604,
+      "grad_norm": 0.7812894582748413,
+      "learning_rate": 1.5637321055277713e-05,
+      "loss": 0.7598,
+      "step": 25980
+    },
+    {
+      "epoch": 2.06314870308996,
+      "grad_norm": 0.9928243160247803,
+      "learning_rate": 1.5624090391892252e-05,
+      "loss": 0.7991,
+      "step": 25990
+    },
+    {
+      "epoch": 2.0639425271388596,
+      "grad_norm": 0.9473987221717834,
+      "learning_rate": 1.5610859728506788e-05,
+      "loss": 0.7417,
+      "step": 26000
+    },
+    {
+      "epoch": 2.0647363511877592,
+      "grad_norm": 0.9683297276496887,
+      "learning_rate": 1.5597629065121323e-05,
+      "loss": 0.7986,
+      "step": 26010
+    },
+    {
+      "epoch": 2.065530175236659,
+      "grad_norm": 0.9280180335044861,
+      "learning_rate": 1.5584398401735866e-05,
+      "loss": 0.7455,
+      "step": 26020
+    },
+    {
+      "epoch": 2.0663239992855584,
+      "grad_norm": 0.8765127658843994,
+      "learning_rate": 1.55711677383504e-05,
+      "loss": 0.8731,
+      "step": 26030
+    },
+    {
+      "epoch": 2.067117823334458,
+      "grad_norm": 0.7415598034858704,
+      "learning_rate": 1.555793707496494e-05,
+      "loss": 0.8217,
+      "step": 26040
+    },
+    {
+      "epoch": 2.0679116473833576,
+      "grad_norm": 0.8780352473258972,
+      "learning_rate": 1.554602947791802e-05,
+      "loss": 0.8236,
+      "step": 26050
+    },
+    {
+      "epoch": 2.0687054714322572,
+      "grad_norm": 0.9112974405288696,
+      "learning_rate": 1.5532798814532563e-05,
+      "loss": 0.7893,
+      "step": 26060
+    },
+    {
+      "epoch": 2.0694992954811564,
+      "grad_norm": 0.726959228515625,
+      "learning_rate": 1.55195681511471e-05,
+      "loss": 0.7708,
+      "step": 26070
+    },
+    {
+      "epoch": 2.070293119530056,
+      "grad_norm": 0.9605491161346436,
+      "learning_rate": 1.5506337487761638e-05,
+      "loss": 0.7306,
+      "step": 26080
+    },
+    {
+      "epoch": 2.0710869435789556,
+      "grad_norm": 0.6991214752197266,
+      "learning_rate": 1.5493106824376173e-05,
+      "loss": 0.8029,
+      "step": 26090
+    },
+    {
+      "epoch": 2.0718807676278552,
+      "grad_norm": 0.947589099407196,
+      "learning_rate": 1.5479876160990712e-05,
+      "loss": 0.7885,
+      "step": 26100
+    },
+    {
+      "epoch": 2.072674591676755,
+      "grad_norm": 0.9293213486671448,
+      "learning_rate": 1.546664549760525e-05,
+      "loss": 0.7285,
+      "step": 26110
+    },
+    {
+      "epoch": 2.0734684157256544,
+      "grad_norm": 1.076295018196106,
+      "learning_rate": 1.545341483421979e-05,
+      "loss": 0.7704,
+      "step": 26120
+    },
+    {
+      "epoch": 2.074262239774554,
+      "grad_norm": 0.894410252571106,
+      "learning_rate": 1.5440184170834326e-05,
+      "loss": 0.7298,
+      "step": 26130
+    },
+    {
+      "epoch": 2.0750560638234536,
+      "grad_norm": 0.9316887855529785,
+      "learning_rate": 1.5426953507448865e-05,
+      "loss": 0.7989,
+      "step": 26140
+    },
+    {
+      "epoch": 2.0758498878723532,
+      "grad_norm": 0.8590110540390015,
+      "learning_rate": 1.5413722844063404e-05,
+      "loss": 0.7369,
+      "step": 26150
+    },
+    {
+      "epoch": 2.076643711921253,
+      "grad_norm": 0.8195733428001404,
+      "learning_rate": 1.540049218067794e-05,
+      "loss": 0.7837,
+      "step": 26160
+    },
+    {
+      "epoch": 2.077437535970152,
+      "grad_norm": 0.8563514351844788,
+      "learning_rate": 1.5387261517292478e-05,
+      "loss": 0.7514,
+      "step": 26170
+    },
+    {
+      "epoch": 2.0782313600190516,
+      "grad_norm": 0.8873001933097839,
+      "learning_rate": 1.5374030853907014e-05,
+      "loss": 0.7356,
+      "step": 26180
+    },
+    {
+      "epoch": 2.0790251840679512,
+      "grad_norm": 0.9553200006484985,
+      "learning_rate": 1.5360800190521556e-05,
+      "loss": 0.7611,
+      "step": 26190
+    },
+    {
+      "epoch": 2.079819008116851,
+      "grad_norm": 0.782842218875885,
+      "learning_rate": 1.534756952713609e-05,
+      "loss": 0.7383,
+      "step": 26200
+    },
+    {
+      "epoch": 2.0806128321657504,
+      "grad_norm": 0.9627733826637268,
+      "learning_rate": 1.533433886375063e-05,
+      "loss": 0.7796,
+      "step": 26210
+    },
+    {
+      "epoch": 2.08140665621465,
+      "grad_norm": 0.7810627818107605,
+      "learning_rate": 1.5321108200365166e-05,
+      "loss": 0.7968,
+      "step": 26220
+    },
+    {
+      "epoch": 2.0822004802635496,
+      "grad_norm": 0.8176783919334412,
+      "learning_rate": 1.5307877536979705e-05,
+      "loss": 0.7525,
+      "step": 26230
+    },
+    {
+      "epoch": 2.0829943043124493,
+      "grad_norm": 0.9359534382820129,
+      "learning_rate": 1.5294646873594244e-05,
+      "loss": 0.8248,
+      "step": 26240
+    },
+    {
+      "epoch": 2.083788128361349,
+      "grad_norm": 0.7079921960830688,
+      "learning_rate": 1.528141621020878e-05,
+      "loss": 0.8115,
+      "step": 26250
+    },
+    {
+      "epoch": 2.0845819524102485,
+      "grad_norm": 0.8684226870536804,
+      "learning_rate": 1.5268185546823318e-05,
+      "loss": 0.7866,
+      "step": 26260
+    },
+    {
+      "epoch": 2.0853757764591476,
+      "grad_norm": 0.8505408763885498,
+      "learning_rate": 1.5254954883437855e-05,
+      "loss": 0.7144,
+      "step": 26270
+    },
+    {
+      "epoch": 2.0861696005080472,
+      "grad_norm": 0.953670084476471,
+      "learning_rate": 1.5241724220052394e-05,
+      "loss": 0.7999,
+      "step": 26280
+    },
+    {
+      "epoch": 2.086963424556947,
+      "grad_norm": 1.1968601942062378,
+      "learning_rate": 1.5228493556666932e-05,
+      "loss": 0.7863,
+      "step": 26290
+    },
+    {
+      "epoch": 2.0877572486058464,
+      "grad_norm": 0.7209346294403076,
+      "learning_rate": 1.521526289328147e-05,
+      "loss": 0.843,
+      "step": 26300
+    },
+    {
+      "epoch": 2.088551072654746,
+      "grad_norm": 0.8510392904281616,
+      "learning_rate": 1.5202032229896008e-05,
+      "loss": 0.831,
+      "step": 26310
+    },
+    {
+      "epoch": 2.0893448967036456,
+      "grad_norm": 0.899815559387207,
+      "learning_rate": 1.5188801566510547e-05,
+      "loss": 0.7577,
+      "step": 26320
+    },
+    {
+      "epoch": 2.0901387207525453,
+      "grad_norm": 0.8000263571739197,
+      "learning_rate": 1.5175570903125084e-05,
+      "loss": 0.7918,
+      "step": 26330
+    },
+    {
+      "epoch": 2.090932544801445,
+      "grad_norm": 0.7844998240470886,
+      "learning_rate": 1.5162340239739623e-05,
+      "loss": 0.7903,
+      "step": 26340
+    },
+    {
+      "epoch": 2.0917263688503445,
+      "grad_norm": 1.0130077600479126,
+      "learning_rate": 1.5149109576354158e-05,
+      "loss": 0.7545,
+      "step": 26350
+    },
+    {
+      "epoch": 2.092520192899244,
+      "grad_norm": 0.8703743815422058,
+      "learning_rate": 1.5135878912968696e-05,
+      "loss": 0.8129,
+      "step": 26360
+    },
+    {
+      "epoch": 2.0933140169481432,
+      "grad_norm": 0.6894111633300781,
+      "learning_rate": 1.5122648249583235e-05,
+      "loss": 0.7017,
+      "step": 26370
+    },
+    {
+      "epoch": 2.094107840997043,
+      "grad_norm": 0.9596577882766724,
+      "learning_rate": 1.5109417586197772e-05,
+      "loss": 0.6868,
+      "step": 26380
+    },
+    {
+      "epoch": 2.0949016650459424,
+      "grad_norm": 0.878044843673706,
+      "learning_rate": 1.509618692281231e-05,
+      "loss": 0.8629,
+      "step": 26390
+    },
+    {
+      "epoch": 2.095695489094842,
+      "grad_norm": 0.9057536125183105,
+      "learning_rate": 1.5082956259426848e-05,
+      "loss": 0.7812,
+      "step": 26400
+    },
+    {
+      "epoch": 2.0964893131437417,
+      "grad_norm": 0.8363611698150635,
+      "learning_rate": 1.5069725596041387e-05,
+      "loss": 0.7738,
+      "step": 26410
+    },
+    {
+      "epoch": 2.0972831371926413,
+      "grad_norm": 0.7889995574951172,
+      "learning_rate": 1.5056494932655924e-05,
+      "loss": 0.8144,
+      "step": 26420
+    },
+    {
+      "epoch": 2.098076961241541,
+      "grad_norm": 0.9280676245689392,
+      "learning_rate": 1.5043264269270463e-05,
+      "loss": 0.7792,
+      "step": 26430
+    },
+    {
+      "epoch": 2.0988707852904405,
+      "grad_norm": 1.015638828277588,
+      "learning_rate": 1.5030033605885e-05,
+      "loss": 0.7745,
+      "step": 26440
+    },
+    {
+      "epoch": 2.09966460933934,
+      "grad_norm": 0.9778096079826355,
+      "learning_rate": 1.501680294249954e-05,
+      "loss": 0.7636,
+      "step": 26450
+    },
+    {
+      "epoch": 2.1004584333882397,
+      "grad_norm": 0.9082860946655273,
+      "learning_rate": 1.5003572279114075e-05,
+      "loss": 0.7399,
+      "step": 26460
+    },
+    {
+      "epoch": 2.101252257437139,
+      "grad_norm": 0.8969444632530212,
+      "learning_rate": 1.4990341615728615e-05,
+      "loss": 0.7326,
+      "step": 26470
+    },
+    {
+      "epoch": 2.1020460814860384,
+      "grad_norm": 0.9246445298194885,
+      "learning_rate": 1.4977110952343151e-05,
+      "loss": 0.789,
+      "step": 26480
+    },
+    {
+      "epoch": 2.102839905534938,
+      "grad_norm": 0.9216938614845276,
+      "learning_rate": 1.4963880288957688e-05,
+      "loss": 0.821,
+      "step": 26490
+    },
+    {
+      "epoch": 2.1036337295838377,
+      "grad_norm": 0.7816628813743591,
+      "learning_rate": 1.4950649625572227e-05,
+      "loss": 0.7705,
+      "step": 26500
+    },
+    {
+      "epoch": 2.1044275536327373,
+      "grad_norm": 1.04630446434021,
+      "learning_rate": 1.4937418962186764e-05,
+      "loss": 0.7733,
+      "step": 26510
+    },
+    {
+      "epoch": 2.105221377681637,
+      "grad_norm": 0.9890500903129578,
+      "learning_rate": 1.4924188298801303e-05,
+      "loss": 0.8013,
+      "step": 26520
+    },
+    {
+      "epoch": 2.1060152017305365,
+      "grad_norm": 0.8225802779197693,
+      "learning_rate": 1.491095763541584e-05,
+      "loss": 0.7539,
+      "step": 26530
+    },
+    {
+      "epoch": 2.106809025779436,
+      "grad_norm": 0.7966319918632507,
+      "learning_rate": 1.489772697203038e-05,
+      "loss": 0.7814,
+      "step": 26540
+    },
+    {
+      "epoch": 2.1076028498283357,
+      "grad_norm": 0.9166170954704285,
+      "learning_rate": 1.4884496308644915e-05,
+      "loss": 0.6954,
+      "step": 26550
+    },
+    {
+      "epoch": 2.1083966738772353,
+      "grad_norm": 1.0586912631988525,
+      "learning_rate": 1.4871265645259456e-05,
+      "loss": 0.7287,
+      "step": 26560
+    },
+    {
+      "epoch": 2.109190497926135,
+      "grad_norm": 0.9350046515464783,
+      "learning_rate": 1.4858034981873991e-05,
+      "loss": 0.7704,
+      "step": 26570
+    },
+    {
+      "epoch": 2.109984321975034,
+      "grad_norm": 0.9269388914108276,
+      "learning_rate": 1.4844804318488532e-05,
+      "loss": 0.8284,
+      "step": 26580
+    },
+    {
+      "epoch": 2.1107781460239337,
+      "grad_norm": 0.7918737530708313,
+      "learning_rate": 1.4831573655103067e-05,
+      "loss": 0.7947,
+      "step": 26590
+    },
+    {
+      "epoch": 2.1115719700728333,
+      "grad_norm": 0.8778436183929443,
+      "learning_rate": 1.4818342991717604e-05,
+      "loss": 0.7774,
+      "step": 26600
+    },
+    {
+      "epoch": 2.112365794121733,
+      "grad_norm": 1.0038212537765503,
+      "learning_rate": 1.4805112328332143e-05,
+      "loss": 0.7356,
+      "step": 26610
+    },
+    {
+      "epoch": 2.1131596181706325,
+      "grad_norm": 0.811038076877594,
+      "learning_rate": 1.479188166494668e-05,
+      "loss": 0.811,
+      "step": 26620
+    },
+    {
+      "epoch": 2.113953442219532,
+      "grad_norm": 1.1587707996368408,
+      "learning_rate": 1.477865100156122e-05,
+      "loss": 0.8549,
+      "step": 26630
+    },
+    {
+      "epoch": 2.1147472662684317,
+      "grad_norm": 0.74503493309021,
+      "learning_rate": 1.4765420338175757e-05,
+      "loss": 0.7897,
+      "step": 26640
+    },
+    {
+      "epoch": 2.1155410903173313,
+      "grad_norm": 0.9020684361457825,
+      "learning_rate": 1.4752189674790296e-05,
+      "loss": 0.7894,
+      "step": 26650
+    },
+    {
+      "epoch": 2.116334914366231,
+      "grad_norm": 0.7925030589103699,
+      "learning_rate": 1.4738959011404831e-05,
+      "loss": 0.8061,
+      "step": 26660
+    },
+    {
+      "epoch": 2.1171287384151305,
+      "grad_norm": 0.8654730916023254,
+      "learning_rate": 1.4725728348019372e-05,
+      "loss": 0.7569,
+      "step": 26670
+    },
+    {
+      "epoch": 2.1179225624640297,
+      "grad_norm": 0.9839343428611755,
+      "learning_rate": 1.4712497684633907e-05,
+      "loss": 0.8012,
+      "step": 26680
+    },
+    {
+      "epoch": 2.1187163865129293,
+      "grad_norm": 0.8837997317314148,
+      "learning_rate": 1.4699267021248448e-05,
+      "loss": 0.8263,
+      "step": 26690
+    },
+    {
+      "epoch": 2.119510210561829,
+      "grad_norm": 1.0431504249572754,
+      "learning_rate": 1.4686036357862984e-05,
+      "loss": 0.8411,
+      "step": 26700
+    },
+    {
+      "epoch": 2.1203040346107285,
+      "grad_norm": 0.9431108236312866,
+      "learning_rate": 1.467280569447752e-05,
+      "loss": 0.786,
+      "step": 26710
+    },
+    {
+      "epoch": 2.121097858659628,
+      "grad_norm": 0.8475701212882996,
+      "learning_rate": 1.465957503109206e-05,
+      "loss": 0.8422,
+      "step": 26720
+    },
+    {
+      "epoch": 2.1218916827085277,
+      "grad_norm": 0.9545729160308838,
+      "learning_rate": 1.4646344367706597e-05,
+      "loss": 0.7607,
+      "step": 26730
+    },
+    {
+      "epoch": 2.1226855067574273,
+      "grad_norm": 0.9317882061004639,
+      "learning_rate": 1.4633113704321136e-05,
+      "loss": 0.7591,
+      "step": 26740
+    },
+    {
+      "epoch": 2.123479330806327,
+      "grad_norm": 0.8636899590492249,
+      "learning_rate": 1.4619883040935673e-05,
+      "loss": 0.7443,
+      "step": 26750
+    },
+    {
+      "epoch": 2.1242731548552265,
+      "grad_norm": 0.9780907034873962,
+      "learning_rate": 1.4606652377550212e-05,
+      "loss": 0.7513,
+      "step": 26760
+    },
+    {
+      "epoch": 2.125066978904126,
+      "grad_norm": 0.9252169132232666,
+      "learning_rate": 1.4593421714164748e-05,
+      "loss": 0.8244,
+      "step": 26770
+    },
+    {
+      "epoch": 2.1258608029530253,
+      "grad_norm": 0.8407522439956665,
+      "learning_rate": 1.4580191050779288e-05,
+      "loss": 0.7896,
+      "step": 26780
+    },
+    {
+      "epoch": 2.126654627001925,
+      "grad_norm": 0.8842277526855469,
+      "learning_rate": 1.4566960387393824e-05,
+      "loss": 0.8509,
+      "step": 26790
+    },
+    {
+      "epoch": 2.1274484510508245,
+      "grad_norm": 0.8318625092506409,
+      "learning_rate": 1.4553729724008364e-05,
+      "loss": 0.7577,
+      "step": 26800
+    },
+    {
+      "epoch": 2.128242275099724,
+      "grad_norm": 1.0206186771392822,
+      "learning_rate": 1.45404990606229e-05,
+      "loss": 0.7706,
+      "step": 26810
+    },
+    {
+      "epoch": 2.1290360991486237,
+      "grad_norm": 0.7975717782974243,
+      "learning_rate": 1.4527268397237437e-05,
+      "loss": 0.7187,
+      "step": 26820
+    },
+    {
+      "epoch": 2.1298299231975233,
+      "grad_norm": 0.9346230030059814,
+      "learning_rate": 1.4514037733851976e-05,
+      "loss": 0.7833,
+      "step": 26830
+    },
+    {
+      "epoch": 2.130623747246423,
+      "grad_norm": 0.8083703517913818,
+      "learning_rate": 1.4500807070466513e-05,
+      "loss": 0.8184,
+      "step": 26840
+    },
+    {
+      "epoch": 2.1314175712953225,
+      "grad_norm": 1.0353641510009766,
+      "learning_rate": 1.4487576407081052e-05,
+      "loss": 0.8394,
+      "step": 26850
+    },
+    {
+      "epoch": 2.132211395344222,
+      "grad_norm": 0.8950655460357666,
+      "learning_rate": 1.447434574369559e-05,
+      "loss": 0.8077,
+      "step": 26860
+    },
+    {
+      "epoch": 2.1330052193931217,
+      "grad_norm": 1.034495234489441,
+      "learning_rate": 1.4461115080310128e-05,
+      "loss": 0.7665,
+      "step": 26870
+    },
+    {
+      "epoch": 2.133799043442021,
+      "grad_norm": 0.8213537335395813,
+      "learning_rate": 1.4447884416924664e-05,
+      "loss": 0.8224,
+      "step": 26880
+    },
+    {
+      "epoch": 2.1345928674909205,
+      "grad_norm": 1.1027780771255493,
+      "learning_rate": 1.4434653753539205e-05,
+      "loss": 0.7717,
+      "step": 26890
+    },
+    {
+      "epoch": 2.13538669153982,
+      "grad_norm": 1.02985417842865,
+      "learning_rate": 1.442142309015374e-05,
+      "loss": 0.7465,
+      "step": 26900
+    },
+    {
+      "epoch": 2.1361805155887197,
+      "grad_norm": 0.7433731555938721,
+      "learning_rate": 1.440819242676828e-05,
+      "loss": 0.7882,
+      "step": 26910
+    },
+    {
+      "epoch": 2.1369743396376193,
+      "grad_norm": 1.0033385753631592,
+      "learning_rate": 1.4394961763382816e-05,
+      "loss": 0.8659,
+      "step": 26920
+    },
+    {
+      "epoch": 2.137768163686519,
+      "grad_norm": 0.7365638613700867,
+      "learning_rate": 1.4381731099997355e-05,
+      "loss": 0.7062,
+      "step": 26930
+    },
+    {
+      "epoch": 2.1385619877354185,
+      "grad_norm": 1.0388671159744263,
+      "learning_rate": 1.4368500436611892e-05,
+      "loss": 0.8148,
+      "step": 26940
+    },
+    {
+      "epoch": 2.139355811784318,
+      "grad_norm": 0.9410164952278137,
+      "learning_rate": 1.435526977322643e-05,
+      "loss": 0.6894,
+      "step": 26950
+    },
+    {
+      "epoch": 2.1401496358332177,
+      "grad_norm": 0.903447151184082,
+      "learning_rate": 1.4342039109840969e-05,
+      "loss": 0.7527,
+      "step": 26960
+    },
+    {
+      "epoch": 2.1409434598821173,
+      "grad_norm": 1.121335744857788,
+      "learning_rate": 1.4328808446455506e-05,
+      "loss": 0.8063,
+      "step": 26970
+    },
+    {
+      "epoch": 2.1417372839310165,
+      "grad_norm": 0.8166680335998535,
+      "learning_rate": 1.4315577783070045e-05,
+      "loss": 0.8217,
+      "step": 26980
+    },
+    {
+      "epoch": 2.142531107979916,
+      "grad_norm": 0.8533350825309753,
+      "learning_rate": 1.430234711968458e-05,
+      "loss": 0.8218,
+      "step": 26990
+    },
+    {
+      "epoch": 2.1433249320288157,
+      "grad_norm": 0.9793769121170044,
+      "learning_rate": 1.4289116456299121e-05,
+      "loss": 0.7345,
+      "step": 27000
+    },
+    {
+      "epoch": 2.1441187560777153,
+      "grad_norm": 0.9998401403427124,
+      "learning_rate": 1.4275885792913656e-05,
+      "loss": 0.7736,
+      "step": 27010
+    },
+    {
+      "epoch": 2.144912580126615,
+      "grad_norm": 1.0122779607772827,
+      "learning_rate": 1.4262655129528197e-05,
+      "loss": 0.7703,
+      "step": 27020
+    },
+    {
+      "epoch": 2.1457064041755145,
+      "grad_norm": 1.040964961051941,
+      "learning_rate": 1.4249424466142733e-05,
+      "loss": 0.7899,
+      "step": 27030
+    },
+    {
+      "epoch": 2.146500228224414,
+      "grad_norm": 0.8178542852401733,
+      "learning_rate": 1.4236193802757272e-05,
+      "loss": 0.7742,
+      "step": 27040
+    },
+    {
+      "epoch": 2.1472940522733137,
+      "grad_norm": 0.8384308815002441,
+      "learning_rate": 1.4222963139371809e-05,
+      "loss": 0.8275,
+      "step": 27050
+    },
+    {
+      "epoch": 2.1480878763222133,
+      "grad_norm": 0.9705979228019714,
+      "learning_rate": 1.4209732475986346e-05,
+      "loss": 0.7532,
+      "step": 27060
+    },
+    {
+      "epoch": 2.148881700371113,
+      "grad_norm": 0.9038788080215454,
+      "learning_rate": 1.4196501812600885e-05,
+      "loss": 0.8162,
+      "step": 27070
+    },
+    {
+      "epoch": 2.1496755244200125,
+      "grad_norm": 0.9565449953079224,
+      "learning_rate": 1.4183271149215422e-05,
+      "loss": 0.7742,
+      "step": 27080
+    },
+    {
+      "epoch": 2.1504693484689117,
+      "grad_norm": 0.9967783093452454,
+      "learning_rate": 1.4170040485829961e-05,
+      "loss": 0.8269,
+      "step": 27090
+    },
+    {
+      "epoch": 2.1512631725178113,
+      "grad_norm": 0.8932458162307739,
+      "learning_rate": 1.4156809822444497e-05,
+      "loss": 0.746,
+      "step": 27100
+    },
+    {
+      "epoch": 2.152056996566711,
+      "grad_norm": 0.9235972762107849,
+      "learning_rate": 1.4143579159059037e-05,
+      "loss": 0.8389,
+      "step": 27110
+    },
+    {
+      "epoch": 2.1528508206156105,
+      "grad_norm": 0.9918680787086487,
+      "learning_rate": 1.4130348495673573e-05,
+      "loss": 0.7368,
+      "step": 27120
+    },
+    {
+      "epoch": 2.15364464466451,
+      "grad_norm": 1.0589817762374878,
+      "learning_rate": 1.4117117832288113e-05,
+      "loss": 0.708,
+      "step": 27130
+    },
+    {
+      "epoch": 2.1544384687134097,
+      "grad_norm": 0.8974472284317017,
+      "learning_rate": 1.4103887168902649e-05,
+      "loss": 0.8292,
+      "step": 27140
+    },
+    {
+      "epoch": 2.1552322927623093,
+      "grad_norm": 0.8955181837081909,
+      "learning_rate": 1.4090656505517188e-05,
+      "loss": 0.7798,
+      "step": 27150
+    },
+    {
+      "epoch": 2.156026116811209,
+      "grad_norm": 1.0225322246551514,
+      "learning_rate": 1.4077425842131725e-05,
+      "loss": 0.7567,
+      "step": 27160
+    },
+    {
+      "epoch": 2.1568199408601085,
+      "grad_norm": 0.7848957180976868,
+      "learning_rate": 1.4064195178746262e-05,
+      "loss": 0.8015,
+      "step": 27170
+    },
+    {
+      "epoch": 2.1576137649090077,
+      "grad_norm": 0.916002631187439,
+      "learning_rate": 1.4050964515360801e-05,
+      "loss": 0.8266,
+      "step": 27180
+    },
+    {
+      "epoch": 2.1584075889579073,
+      "grad_norm": 0.8571447134017944,
+      "learning_rate": 1.4037733851975337e-05,
+      "loss": 0.7878,
+      "step": 27190
+    },
+    {
+      "epoch": 2.159201413006807,
+      "grad_norm": 0.8290430903434753,
+      "learning_rate": 1.4024503188589877e-05,
+      "loss": 0.7503,
+      "step": 27200
+    },
+    {
+      "epoch": 2.1599952370557065,
+      "grad_norm": 0.9717968702316284,
+      "learning_rate": 1.4011272525204413e-05,
+      "loss": 0.7839,
+      "step": 27210
+    },
+    {
+      "epoch": 2.160789061104606,
+      "grad_norm": 0.8097538948059082,
+      "learning_rate": 1.3998041861818954e-05,
+      "loss": 0.784,
+      "step": 27220
+    },
+    {
+      "epoch": 2.1615828851535057,
+      "grad_norm": 1.0527442693710327,
+      "learning_rate": 1.3984811198433489e-05,
+      "loss": 0.7795,
+      "step": 27230
+    },
+    {
+      "epoch": 2.1623767092024053,
+      "grad_norm": 0.7761133909225464,
+      "learning_rate": 1.397158053504803e-05,
+      "loss": 0.796,
+      "step": 27240
+    },
+    {
+      "epoch": 2.163170533251305,
+      "grad_norm": 1.009090781211853,
+      "learning_rate": 1.3958349871662565e-05,
+      "loss": 0.7895,
+      "step": 27250
+    },
+    {
+      "epoch": 2.1639643573002045,
+      "grad_norm": 0.9476297497749329,
+      "learning_rate": 1.3945119208277104e-05,
+      "loss": 0.805,
+      "step": 27260
+    },
+    {
+      "epoch": 2.164758181349104,
+      "grad_norm": 0.8424776792526245,
+      "learning_rate": 1.3931888544891641e-05,
+      "loss": 0.7752,
+      "step": 27270
+    },
+    {
+      "epoch": 2.1655520053980037,
+      "grad_norm": 1.0687159299850464,
+      "learning_rate": 1.391865788150618e-05,
+      "loss": 0.7333,
+      "step": 27280
+    },
+    {
+      "epoch": 2.166345829446903,
+      "grad_norm": 0.7562177777290344,
+      "learning_rate": 1.3905427218120718e-05,
+      "loss": 0.766,
+      "step": 27290
+    },
+    {
+      "epoch": 2.1671396534958025,
+      "grad_norm": 0.9093291163444519,
+      "learning_rate": 1.3892196554735253e-05,
+      "loss": 0.7164,
+      "step": 27300
+    },
+    {
+      "epoch": 2.167933477544702,
+      "grad_norm": 0.8972833156585693,
+      "learning_rate": 1.3878965891349794e-05,
+      "loss": 0.7667,
+      "step": 27310
+    },
+    {
+      "epoch": 2.1687273015936017,
+      "grad_norm": 0.9251981377601624,
+      "learning_rate": 1.386573522796433e-05,
+      "loss": 0.7996,
+      "step": 27320
+    },
+    {
+      "epoch": 2.1695211256425013,
+      "grad_norm": 0.8145211338996887,
+      "learning_rate": 1.385250456457887e-05,
+      "loss": 0.7725,
+      "step": 27330
+    },
+    {
+      "epoch": 2.170314949691401,
+      "grad_norm": 0.8409168124198914,
+      "learning_rate": 1.3839273901193405e-05,
+      "loss": 0.7804,
+      "step": 27340
+    },
+    {
+      "epoch": 2.1711087737403005,
+      "grad_norm": 0.9111427068710327,
+      "learning_rate": 1.3826043237807944e-05,
+      "loss": 0.7775,
+      "step": 27350
+    },
+    {
+      "epoch": 2.1719025977892,
+      "grad_norm": 0.91303551197052,
+      "learning_rate": 1.3812812574422482e-05,
+      "loss": 0.8227,
+      "step": 27360
+    },
+    {
+      "epoch": 2.1726964218380997,
+      "grad_norm": 0.8870943188667297,
+      "learning_rate": 1.379958191103702e-05,
+      "loss": 0.7965,
+      "step": 27370
+    },
+    {
+      "epoch": 2.1734902458869994,
+      "grad_norm": 0.8687139749526978,
+      "learning_rate": 1.3786351247651558e-05,
+      "loss": 0.799,
+      "step": 27380
+    },
+    {
+      "epoch": 2.1742840699358985,
+      "grad_norm": 0.9708724617958069,
+      "learning_rate": 1.3773120584266097e-05,
+      "loss": 0.6977,
+      "step": 27390
+    },
+    {
+      "epoch": 2.175077893984798,
+      "grad_norm": 0.8507083654403687,
+      "learning_rate": 1.3759889920880634e-05,
+      "loss": 0.8515,
+      "step": 27400
+    },
+    {
+      "epoch": 2.1758717180336977,
+      "grad_norm": 0.9613845348358154,
+      "learning_rate": 1.374665925749517e-05,
+      "loss": 0.8635,
+      "step": 27410
+    },
+    {
+      "epoch": 2.1766655420825973,
+      "grad_norm": 0.8473123908042908,
+      "learning_rate": 1.373342859410971e-05,
+      "loss": 0.7451,
+      "step": 27420
+    },
+    {
+      "epoch": 2.177459366131497,
+      "grad_norm": 0.9432036280632019,
+      "learning_rate": 1.3720197930724246e-05,
+      "loss": 0.8018,
+      "step": 27430
+    },
+    {
+      "epoch": 2.1782531901803965,
+      "grad_norm": 0.8323158025741577,
+      "learning_rate": 1.3706967267338786e-05,
+      "loss": 0.7799,
+      "step": 27440
+    },
+    {
+      "epoch": 2.179047014229296,
+      "grad_norm": 1.166176438331604,
+      "learning_rate": 1.3693736603953322e-05,
+      "loss": 0.7891,
+      "step": 27450
+    },
+    {
+      "epoch": 2.1798408382781957,
+      "grad_norm": 0.781601071357727,
+      "learning_rate": 1.368050594056786e-05,
+      "loss": 0.7851,
+      "step": 27460
+    },
+    {
+      "epoch": 2.1806346623270954,
+      "grad_norm": 1.0222710371017456,
+      "learning_rate": 1.3667275277182398e-05,
+      "loss": 0.7159,
+      "step": 27470
+    },
+    {
+      "epoch": 2.181428486375995,
+      "grad_norm": 0.758436381816864,
+      "learning_rate": 1.3654044613796937e-05,
+      "loss": 0.7342,
+      "step": 27480
+    },
+    {
+      "epoch": 2.182222310424894,
+      "grad_norm": 0.8809502124786377,
+      "learning_rate": 1.3640813950411474e-05,
+      "loss": 0.8097,
+      "step": 27490
+    },
+    {
+      "epoch": 2.1830161344737937,
+      "grad_norm": 0.8321342468261719,
+      "learning_rate": 1.3627583287026013e-05,
+      "loss": 0.756,
+      "step": 27500
+    },
+    {
+      "epoch": 2.1838099585226933,
+      "grad_norm": 0.8868312835693359,
+      "learning_rate": 1.361435262364055e-05,
+      "loss": 0.8246,
+      "step": 27510
+    },
+    {
+      "epoch": 2.184603782571593,
+      "grad_norm": 0.8987812995910645,
+      "learning_rate": 1.3601121960255086e-05,
+      "loss": 0.8034,
+      "step": 27520
+    },
+    {
+      "epoch": 2.1853976066204925,
+      "grad_norm": 0.8325653076171875,
+      "learning_rate": 1.3587891296869626e-05,
+      "loss": 0.8197,
+      "step": 27530
+    },
+    {
+      "epoch": 2.186191430669392,
+      "grad_norm": 0.9166516065597534,
+      "learning_rate": 1.3574660633484162e-05,
+      "loss": 0.7413,
+      "step": 27540
+    },
+    {
+      "epoch": 2.1869852547182917,
+      "grad_norm": 0.9565515518188477,
+      "learning_rate": 1.3561429970098703e-05,
+      "loss": 0.7971,
+      "step": 27550
+    },
+    {
+      "epoch": 2.1877790787671914,
+      "grad_norm": 0.8458524346351624,
+      "learning_rate": 1.3548199306713238e-05,
+      "loss": 0.756,
+      "step": 27560
+    },
+    {
+      "epoch": 2.188572902816091,
+      "grad_norm": 0.7264953255653381,
+      "learning_rate": 1.3534968643327777e-05,
+      "loss": 0.798,
+      "step": 27570
+    },
+    {
+      "epoch": 2.1893667268649906,
+      "grad_norm": 0.9435407519340515,
+      "learning_rate": 1.3521737979942314e-05,
+      "loss": 0.7956,
+      "step": 27580
+    },
+    {
+      "epoch": 2.19016055091389,
+      "grad_norm": 0.9455521106719971,
+      "learning_rate": 1.3508507316556853e-05,
+      "loss": 0.7712,
+      "step": 27590
+    },
+    {
+      "epoch": 2.1909543749627893,
+      "grad_norm": 1.1080188751220703,
+      "learning_rate": 1.349527665317139e-05,
+      "loss": 0.7544,
+      "step": 27600
+    },
+    {
+      "epoch": 2.191748199011689,
+      "grad_norm": 0.98456871509552,
+      "learning_rate": 1.348204598978593e-05,
+      "loss": 0.8196,
+      "step": 27610
+    },
+    {
+      "epoch": 2.1925420230605885,
+      "grad_norm": 0.8413136005401611,
+      "learning_rate": 1.3468815326400467e-05,
+      "loss": 0.7864,
+      "step": 27620
+    },
+    {
+      "epoch": 2.193335847109488,
+      "grad_norm": 0.7672378420829773,
+      "learning_rate": 1.3455584663015002e-05,
+      "loss": 0.7624,
+      "step": 27630
+    },
+    {
+      "epoch": 2.1941296711583878,
+      "grad_norm": 0.9120532274246216,
+      "learning_rate": 1.3442353999629543e-05,
+      "loss": 0.7425,
+      "step": 27640
+    },
+    {
+      "epoch": 2.1949234952072874,
+      "grad_norm": 0.9745380878448486,
+      "learning_rate": 1.3429123336244078e-05,
+      "loss": 0.8149,
+      "step": 27650
+    },
+    {
+      "epoch": 2.195717319256187,
+      "grad_norm": 0.8304915428161621,
+      "learning_rate": 1.3415892672858619e-05,
+      "loss": 0.7609,
+      "step": 27660
+    },
+    {
+      "epoch": 2.1965111433050866,
+      "grad_norm": 0.9230349659919739,
+      "learning_rate": 1.3402662009473154e-05,
+      "loss": 0.7648,
+      "step": 27670
+    },
+    {
+      "epoch": 2.197304967353986,
+      "grad_norm": 1.0619224309921265,
+      "learning_rate": 1.3389431346087693e-05,
+      "loss": 0.7887,
+      "step": 27680
+    },
+    {
+      "epoch": 2.1980987914028853,
+      "grad_norm": 1.0033013820648193,
+      "learning_rate": 1.337620068270223e-05,
+      "loss": 0.805,
+      "step": 27690
+    },
+    {
+      "epoch": 2.198892615451785,
+      "grad_norm": 0.9245965480804443,
+      "learning_rate": 1.336297001931677e-05,
+      "loss": 0.8012,
+      "step": 27700
+    },
+    {
+      "epoch": 2.1996864395006845,
+      "grad_norm": 1.0526070594787598,
+      "learning_rate": 1.3349739355931307e-05,
+      "loss": 0.7792,
+      "step": 27710
+    },
+    {
+      "epoch": 2.200480263549584,
+      "grad_norm": 0.7733340859413147,
+      "learning_rate": 1.3336508692545846e-05,
+      "loss": 0.7708,
+      "step": 27720
+    },
+    {
+      "epoch": 2.2012740875984838,
+      "grad_norm": 1.1390515565872192,
+      "learning_rate": 1.3323278029160383e-05,
+      "loss": 0.7797,
+      "step": 27730
+    },
+    {
+      "epoch": 2.2020679116473834,
+      "grad_norm": 0.8812433481216431,
+      "learning_rate": 1.3310047365774922e-05,
+      "loss": 0.719,
+      "step": 27740
+    },
+    {
+      "epoch": 2.202861735696283,
+      "grad_norm": 0.8924884796142578,
+      "learning_rate": 1.3296816702389459e-05,
+      "loss": 0.836,
+      "step": 27750
+    },
+    {
+      "epoch": 2.2036555597451826,
+      "grad_norm": 0.9707908630371094,
+      "learning_rate": 1.3283586039003995e-05,
+      "loss": 0.7732,
+      "step": 27760
+    },
+    {
+      "epoch": 2.204449383794082,
+      "grad_norm": 0.904464602470398,
+      "learning_rate": 1.3270355375618535e-05,
+      "loss": 0.8084,
+      "step": 27770
+    },
+    {
+      "epoch": 2.205243207842982,
+      "grad_norm": 0.8692517280578613,
+      "learning_rate": 1.325712471223307e-05,
+      "loss": 0.8074,
+      "step": 27780
+    },
+    {
+      "epoch": 2.2060370318918814,
+      "grad_norm": 0.9948639869689941,
+      "learning_rate": 1.324389404884761e-05,
+      "loss": 0.7936,
+      "step": 27790
+    },
+    {
+      "epoch": 2.2068308559407805,
+      "grad_norm": 0.915266215801239,
+      "learning_rate": 1.3230663385462147e-05,
+      "loss": 0.8084,
+      "step": 27800
+    },
+    {
+      "epoch": 2.20762467998968,
+      "grad_norm": 0.8853785395622253,
+      "learning_rate": 1.3217432722076686e-05,
+      "loss": 0.774,
+      "step": 27810
+    },
+    {
+      "epoch": 2.2084185040385798,
+      "grad_norm": 1.0445677042007446,
+      "learning_rate": 1.3204202058691223e-05,
+      "loss": 0.8096,
+      "step": 27820
+    },
+    {
+      "epoch": 2.2092123280874794,
+      "grad_norm": 0.9219695925712585,
+      "learning_rate": 1.3190971395305762e-05,
+      "loss": 0.8174,
+      "step": 27830
+    },
+    {
+      "epoch": 2.210006152136379,
+      "grad_norm": 0.727891206741333,
+      "learning_rate": 1.31777407319203e-05,
+      "loss": 0.7495,
+      "step": 27840
+    },
+    {
+      "epoch": 2.2107999761852786,
+      "grad_norm": 0.8732167482376099,
+      "learning_rate": 1.3164510068534838e-05,
+      "loss": 0.7907,
+      "step": 27850
+    },
+    {
+      "epoch": 2.211593800234178,
+      "grad_norm": 0.9940975904464722,
+      "learning_rate": 1.3151279405149375e-05,
+      "loss": 0.8013,
+      "step": 27860
+    },
+    {
+      "epoch": 2.212387624283078,
+      "grad_norm": 0.9644079804420471,
+      "learning_rate": 1.3138048741763911e-05,
+      "loss": 0.7781,
+      "step": 27870
+    },
+    {
+      "epoch": 2.2131814483319774,
+      "grad_norm": 0.957815945148468,
+      "learning_rate": 1.3124818078378452e-05,
+      "loss": 0.7853,
+      "step": 27880
+    },
+    {
+      "epoch": 2.2139752723808765,
+      "grad_norm": 0.9299752712249756,
+      "learning_rate": 1.3111587414992987e-05,
+      "loss": 0.7654,
+      "step": 27890
+    },
+    {
+      "epoch": 2.214769096429776,
+      "grad_norm": 0.763888418674469,
+      "learning_rate": 1.3098356751607526e-05,
+      "loss": 0.8254,
+      "step": 27900
+    },
+    {
+      "epoch": 2.2155629204786758,
+      "grad_norm": 0.9485118389129639,
+      "learning_rate": 1.3085126088222063e-05,
+      "loss": 0.7565,
+      "step": 27910
+    },
+    {
+      "epoch": 2.2163567445275754,
+      "grad_norm": 0.7794759273529053,
+      "learning_rate": 1.3071895424836602e-05,
+      "loss": 0.7794,
+      "step": 27920
+    },
+    {
+      "epoch": 2.217150568576475,
+      "grad_norm": 0.8612314462661743,
+      "learning_rate": 1.305866476145114e-05,
+      "loss": 0.8535,
+      "step": 27930
+    },
+    {
+      "epoch": 2.2179443926253746,
+      "grad_norm": 0.926816463470459,
+      "learning_rate": 1.3045434098065678e-05,
+      "loss": 0.7421,
+      "step": 27940
+    },
+    {
+      "epoch": 2.218738216674274,
+      "grad_norm": 0.8914726376533508,
+      "learning_rate": 1.3032203434680216e-05,
+      "loss": 0.7918,
+      "step": 27950
+    },
+    {
+      "epoch": 2.219532040723174,
+      "grad_norm": 0.7953301072120667,
+      "learning_rate": 1.3018972771294755e-05,
+      "loss": 0.7813,
+      "step": 27960
+    },
+    {
+      "epoch": 2.2203258647720734,
+      "grad_norm": 1.100185513496399,
+      "learning_rate": 1.3005742107909292e-05,
+      "loss": 0.8074,
+      "step": 27970
+    },
+    {
+      "epoch": 2.221119688820973,
+      "grad_norm": 0.9314887523651123,
+      "learning_rate": 1.2992511444523827e-05,
+      "loss": 0.7986,
+      "step": 27980
+    },
+    {
+      "epoch": 2.2219135128698726,
+      "grad_norm": 0.9188593626022339,
+      "learning_rate": 1.2979280781138366e-05,
+      "loss": 0.7485,
+      "step": 27990
+    },
+    {
+      "epoch": 2.2227073369187718,
+      "grad_norm": 1.0341116189956665,
+      "learning_rate": 1.2966050117752903e-05,
+      "loss": 0.7407,
+      "step": 28000
+    },
+    {
+      "epoch": 2.2235011609676714,
+      "grad_norm": 0.8388242125511169,
+      "learning_rate": 1.2952819454367442e-05,
+      "loss": 0.8283,
+      "step": 28010
+    },
+    {
+      "epoch": 2.224294985016571,
+      "grad_norm": 0.9826040267944336,
+      "learning_rate": 1.293958879098198e-05,
+      "loss": 0.7214,
+      "step": 28020
+    },
+    {
+      "epoch": 2.2250888090654706,
+      "grad_norm": 0.7271533608436584,
+      "learning_rate": 1.2926358127596519e-05,
+      "loss": 0.836,
+      "step": 28030
+    },
+    {
+      "epoch": 2.22588263311437,
+      "grad_norm": 0.795552134513855,
+      "learning_rate": 1.2913127464211056e-05,
+      "loss": 0.8255,
+      "step": 28040
+    },
+    {
+      "epoch": 2.22667645716327,
+      "grad_norm": 0.8182919025421143,
+      "learning_rate": 1.2899896800825595e-05,
+      "loss": 0.7078,
+      "step": 28050
+    },
+    {
+      "epoch": 2.2274702812121694,
+      "grad_norm": 0.9969826340675354,
+      "learning_rate": 1.2886666137440132e-05,
+      "loss": 0.7659,
+      "step": 28060
+    },
+    {
+      "epoch": 2.228264105261069,
+      "grad_norm": 0.8955338001251221,
+      "learning_rate": 1.2873435474054671e-05,
+      "loss": 0.8228,
+      "step": 28070
+    },
+    {
+      "epoch": 2.2290579293099686,
+      "grad_norm": 0.9254652261734009,
+      "learning_rate": 1.2860204810669208e-05,
+      "loss": 0.8226,
+      "step": 28080
+    },
+    {
+      "epoch": 2.229851753358868,
+      "grad_norm": 0.8870140910148621,
+      "learning_rate": 1.2846974147283744e-05,
+      "loss": 0.8239,
+      "step": 28090
+    },
+    {
+      "epoch": 2.2306455774077674,
+      "grad_norm": 1.0438228845596313,
+      "learning_rate": 1.2833743483898283e-05,
+      "loss": 0.7559,
+      "step": 28100
+    },
+    {
+      "epoch": 2.231439401456667,
+      "grad_norm": 1.0301085710525513,
+      "learning_rate": 1.282051282051282e-05,
+      "loss": 0.7513,
+      "step": 28110
+    },
+    {
+      "epoch": 2.2322332255055666,
+      "grad_norm": 0.9675314426422119,
+      "learning_rate": 1.2807282157127359e-05,
+      "loss": 0.8013,
+      "step": 28120
+    },
+    {
+      "epoch": 2.233027049554466,
+      "grad_norm": 1.017149806022644,
+      "learning_rate": 1.2794051493741896e-05,
+      "loss": 0.7762,
+      "step": 28130
+    },
+    {
+      "epoch": 2.233820873603366,
+      "grad_norm": 0.8916369080543518,
+      "learning_rate": 1.2780820830356435e-05,
+      "loss": 0.8421,
+      "step": 28140
+    },
+    {
+      "epoch": 2.2346146976522654,
+      "grad_norm": 0.8855636119842529,
+      "learning_rate": 1.2767590166970972e-05,
+      "loss": 0.792,
+      "step": 28150
+    },
+    {
+      "epoch": 2.235408521701165,
+      "grad_norm": 0.758443295955658,
+      "learning_rate": 1.2754359503585511e-05,
+      "loss": 0.7603,
+      "step": 28160
+    },
+    {
+      "epoch": 2.2362023457500646,
+      "grad_norm": 1.0338447093963623,
+      "learning_rate": 1.2741128840200048e-05,
+      "loss": 0.787,
+      "step": 28170
+    },
+    {
+      "epoch": 2.236996169798964,
+      "grad_norm": 1.0957199335098267,
+      "learning_rate": 1.2727898176814587e-05,
+      "loss": 0.7863,
+      "step": 28180
+    },
+    {
+      "epoch": 2.237789993847864,
+      "grad_norm": 0.8449088335037231,
+      "learning_rate": 1.2714667513429124e-05,
+      "loss": 0.782,
+      "step": 28190
+    },
+    {
+      "epoch": 2.238583817896763,
+      "grad_norm": 0.9781031012535095,
+      "learning_rate": 1.2701436850043663e-05,
+      "loss": 0.777,
+      "step": 28200
+    },
+    {
+      "epoch": 2.2393776419456626,
+      "grad_norm": 0.9978182315826416,
+      "learning_rate": 1.2688206186658199e-05,
+      "loss": 0.7937,
+      "step": 28210
+    },
+    {
+      "epoch": 2.240171465994562,
+      "grad_norm": 1.0001929998397827,
+      "learning_rate": 1.2674975523272736e-05,
+      "loss": 0.8321,
+      "step": 28220
+    },
+    {
+      "epoch": 2.240965290043462,
+      "grad_norm": 0.9343948364257812,
+      "learning_rate": 1.2661744859887275e-05,
+      "loss": 0.7676,
+      "step": 28230
+    },
+    {
+      "epoch": 2.2417591140923614,
+      "grad_norm": 1.0100466012954712,
+      "learning_rate": 1.2648514196501812e-05,
+      "loss": 0.7627,
+      "step": 28240
+    },
+    {
+      "epoch": 2.242552938141261,
+      "grad_norm": 0.7554891109466553,
+      "learning_rate": 1.2635283533116351e-05,
+      "loss": 0.7973,
+      "step": 28250
+    },
+    {
+      "epoch": 2.2433467621901606,
+      "grad_norm": 0.9241369366645813,
+      "learning_rate": 1.2622052869730888e-05,
+      "loss": 0.7823,
+      "step": 28260
+    },
+    {
+      "epoch": 2.24414058623906,
+      "grad_norm": 0.9405050277709961,
+      "learning_rate": 1.2608822206345427e-05,
+      "loss": 0.7655,
+      "step": 28270
+    },
+    {
+      "epoch": 2.24493441028796,
+      "grad_norm": 0.8088013529777527,
+      "learning_rate": 1.2595591542959965e-05,
+      "loss": 0.7726,
+      "step": 28280
+    },
+    {
+      "epoch": 2.2457282343368594,
+      "grad_norm": 0.9160634279251099,
+      "learning_rate": 1.2582360879574504e-05,
+      "loss": 0.7966,
+      "step": 28290
+    },
+    {
+      "epoch": 2.246522058385759,
+      "grad_norm": 0.9134334921836853,
+      "learning_rate": 1.256913021618904e-05,
+      "loss": 0.7318,
+      "step": 28300
+    },
+    {
+      "epoch": 2.247315882434658,
+      "grad_norm": 0.9910647869110107,
+      "learning_rate": 1.255589955280358e-05,
+      "loss": 0.8372,
+      "step": 28310
+    },
+    {
+      "epoch": 2.248109706483558,
+      "grad_norm": 0.9692395329475403,
+      "learning_rate": 1.2542668889418115e-05,
+      "loss": 0.7783,
+      "step": 28320
+    },
+    {
+      "epoch": 2.2489035305324574,
+      "grad_norm": 0.9558752179145813,
+      "learning_rate": 1.2529438226032653e-05,
+      "loss": 0.7304,
+      "step": 28330
+    },
+    {
+      "epoch": 2.249697354581357,
+      "grad_norm": 1.0415302515029907,
+      "learning_rate": 1.2516207562647191e-05,
+      "loss": 0.8635,
+      "step": 28340
+    },
+    {
+      "epoch": 2.2504911786302566,
+      "grad_norm": 0.7867638468742371,
+      "learning_rate": 1.2502976899261729e-05,
+      "loss": 0.8164,
+      "step": 28350
+    },
+    {
+      "epoch": 2.251285002679156,
+      "grad_norm": 0.7861484885215759,
+      "learning_rate": 1.2489746235876268e-05,
+      "loss": 0.8794,
+      "step": 28360
+    },
+    {
+      "epoch": 2.252078826728056,
+      "grad_norm": 0.8788695335388184,
+      "learning_rate": 1.2476515572490807e-05,
+      "loss": 0.8134,
+      "step": 28370
+    },
+    {
+      "epoch": 2.2528726507769554,
+      "grad_norm": 0.9511646032333374,
+      "learning_rate": 1.2463284909105342e-05,
+      "loss": 0.7574,
+      "step": 28380
+    },
+    {
+      "epoch": 2.253666474825855,
+      "grad_norm": 1.1273367404937744,
+      "learning_rate": 1.2450054245719881e-05,
+      "loss": 0.7506,
+      "step": 28390
+    },
+    {
+      "epoch": 2.254460298874754,
+      "grad_norm": 0.773710310459137,
+      "learning_rate": 1.2436823582334418e-05,
+      "loss": 0.8182,
+      "step": 28400
+    },
+    {
+      "epoch": 2.255254122923654,
+      "grad_norm": 0.8454147577285767,
+      "learning_rate": 1.2423592918948957e-05,
+      "loss": 0.7827,
+      "step": 28410
+    },
+    {
+      "epoch": 2.2560479469725534,
+      "grad_norm": 0.7909261584281921,
+      "learning_rate": 1.2410362255563494e-05,
+      "loss": 0.854,
+      "step": 28420
+    },
+    {
+      "epoch": 2.256841771021453,
+      "grad_norm": 1.1431065797805786,
+      "learning_rate": 1.2397131592178032e-05,
+      "loss": 0.7735,
+      "step": 28430
+    },
+    {
+      "epoch": 2.2576355950703526,
+      "grad_norm": 0.8554138541221619,
+      "learning_rate": 1.238390092879257e-05,
+      "loss": 0.7886,
+      "step": 28440
+    },
+    {
+      "epoch": 2.258429419119252,
+      "grad_norm": 0.801655113697052,
+      "learning_rate": 1.2370670265407108e-05,
+      "loss": 0.8245,
+      "step": 28450
+    },
+    {
+      "epoch": 2.259223243168152,
+      "grad_norm": 0.8901875019073486,
+      "learning_rate": 1.2357439602021647e-05,
+      "loss": 0.7802,
+      "step": 28460
+    },
+    {
+      "epoch": 2.2600170672170514,
+      "grad_norm": 1.0075801610946655,
+      "learning_rate": 1.2344208938636184e-05,
+      "loss": 0.763,
+      "step": 28470
+    },
+    {
+      "epoch": 2.260810891265951,
+      "grad_norm": 1.1321099996566772,
+      "learning_rate": 1.2330978275250723e-05,
+      "loss": 0.8004,
+      "step": 28480
+    },
+    {
+      "epoch": 2.2616047153148506,
+      "grad_norm": 0.7835102081298828,
+      "learning_rate": 1.231774761186526e-05,
+      "loss": 0.7653,
+      "step": 28490
+    },
+    {
+      "epoch": 2.2623985393637502,
+      "grad_norm": 0.9481699466705322,
+      "learning_rate": 1.2304516948479797e-05,
+      "loss": 0.7393,
+      "step": 28500
+    },
+    {
+      "epoch": 2.2631923634126494,
+      "grad_norm": 0.876270055770874,
+      "learning_rate": 1.2291286285094335e-05,
+      "loss": 0.7817,
+      "step": 28510
+    },
+    {
+      "epoch": 2.263986187461549,
+      "grad_norm": 0.9594854712486267,
+      "learning_rate": 1.2278055621708873e-05,
+      "loss": 0.8727,
+      "step": 28520
+    },
+    {
+      "epoch": 2.2647800115104486,
+      "grad_norm": 0.919558584690094,
+      "learning_rate": 1.226482495832341e-05,
+      "loss": 0.8242,
+      "step": 28530
+    },
+    {
+      "epoch": 2.265573835559348,
+      "grad_norm": 0.7271903157234192,
+      "learning_rate": 1.2251594294937948e-05,
+      "loss": 0.7551,
+      "step": 28540
+    },
+    {
+      "epoch": 2.266367659608248,
+      "grad_norm": 1.0002189874649048,
+      "learning_rate": 1.2238363631552487e-05,
+      "loss": 0.7631,
+      "step": 28550
+    },
+    {
+      "epoch": 2.2671614836571474,
+      "grad_norm": 0.8755606412887573,
+      "learning_rate": 1.2225132968167024e-05,
+      "loss": 0.7551,
+      "step": 28560
+    },
+    {
+      "epoch": 2.267955307706047,
+      "grad_norm": 0.8737064599990845,
+      "learning_rate": 1.2211902304781563e-05,
+      "loss": 0.7563,
+      "step": 28570
+    },
+    {
+      "epoch": 2.2687491317549466,
+      "grad_norm": 1.0743333101272583,
+      "learning_rate": 1.21986716413961e-05,
+      "loss": 0.7898,
+      "step": 28580
+    },
+    {
+      "epoch": 2.2695429558038462,
+      "grad_norm": 0.7818424105644226,
+      "learning_rate": 1.218544097801064e-05,
+      "loss": 0.8337,
+      "step": 28590
+    },
+    {
+      "epoch": 2.2703367798527454,
+      "grad_norm": 0.8167740702629089,
+      "learning_rate": 1.2172210314625176e-05,
+      "loss": 0.7188,
+      "step": 28600
+    },
+    {
+      "epoch": 2.271130603901645,
+      "grad_norm": 1.0544767379760742,
+      "learning_rate": 1.2158979651239714e-05,
+      "loss": 0.7083,
+      "step": 28610
+    },
+    {
+      "epoch": 2.2719244279505446,
+      "grad_norm": 0.9623509049415588,
+      "learning_rate": 1.2145748987854251e-05,
+      "loss": 0.8609,
+      "step": 28620
+    },
+    {
+      "epoch": 2.272718251999444,
+      "grad_norm": 0.9412724375724792,
+      "learning_rate": 1.213251832446879e-05,
+      "loss": 0.7785,
+      "step": 28630
+    },
+    {
+      "epoch": 2.273512076048344,
+      "grad_norm": 0.9624713063240051,
+      "learning_rate": 1.2119287661083327e-05,
+      "loss": 0.7225,
+      "step": 28640
+    },
+    {
+      "epoch": 2.2743059000972434,
+      "grad_norm": 0.958265483379364,
+      "learning_rate": 1.2106056997697864e-05,
+      "loss": 0.7591,
+      "step": 28650
+    },
+    {
+      "epoch": 2.275099724146143,
+      "grad_norm": 1.0588797330856323,
+      "learning_rate": 1.2092826334312403e-05,
+      "loss": 0.7488,
+      "step": 28660
+    },
+    {
+      "epoch": 2.2758935481950426,
+      "grad_norm": 0.8400679230690002,
+      "learning_rate": 1.207959567092694e-05,
+      "loss": 0.7603,
+      "step": 28670
+    },
+    {
+      "epoch": 2.2766873722439422,
+      "grad_norm": 0.8208017349243164,
+      "learning_rate": 1.206636500754148e-05,
+      "loss": 0.7729,
+      "step": 28680
+    },
+    {
+      "epoch": 2.277481196292842,
+      "grad_norm": 0.8255361914634705,
+      "learning_rate": 1.2053134344156017e-05,
+      "loss": 0.7548,
+      "step": 28690
+    },
+    {
+      "epoch": 2.2782750203417415,
+      "grad_norm": 1.0541584491729736,
+      "learning_rate": 1.2039903680770556e-05,
+      "loss": 0.7856,
+      "step": 28700
+    },
+    {
+      "epoch": 2.2790688443906406,
+      "grad_norm": 1.1183170080184937,
+      "learning_rate": 1.2026673017385093e-05,
+      "loss": 0.7884,
+      "step": 28710
+    },
+    {
+      "epoch": 2.27986266843954,
+      "grad_norm": 0.8430930376052856,
+      "learning_rate": 1.201344235399963e-05,
+      "loss": 0.8003,
+      "step": 28720
+    },
+    {
+      "epoch": 2.28065649248844,
+      "grad_norm": 0.872404932975769,
+      "learning_rate": 1.2000211690614167e-05,
+      "loss": 0.7212,
+      "step": 28730
+    },
+    {
+      "epoch": 2.2814503165373394,
+      "grad_norm": 0.8843950033187866,
+      "learning_rate": 1.1986981027228704e-05,
+      "loss": 0.7738,
+      "step": 28740
+    },
+    {
+      "epoch": 2.282244140586239,
+      "grad_norm": 0.8415934443473816,
+      "learning_rate": 1.1973750363843243e-05,
+      "loss": 0.7609,
+      "step": 28750
+    },
+    {
+      "epoch": 2.2830379646351386,
+      "grad_norm": 0.7827284932136536,
+      "learning_rate": 1.196051970045778e-05,
+      "loss": 0.7876,
+      "step": 28760
+    },
+    {
+      "epoch": 2.2838317886840382,
+      "grad_norm": 1.0107696056365967,
+      "learning_rate": 1.194728903707232e-05,
+      "loss": 0.7452,
+      "step": 28770
+    },
+    {
+      "epoch": 2.284625612732938,
+      "grad_norm": 0.802697479724884,
+      "learning_rate": 1.1934058373686857e-05,
+      "loss": 0.8176,
+      "step": 28780
+    },
+    {
+      "epoch": 2.2854194367818375,
+      "grad_norm": 0.8944302797317505,
+      "learning_rate": 1.1920827710301396e-05,
+      "loss": 0.7346,
+      "step": 28790
+    },
+    {
+      "epoch": 2.286213260830737,
+      "grad_norm": 0.8373863101005554,
+      "learning_rate": 1.1907597046915933e-05,
+      "loss": 0.7585,
+      "step": 28800
+    },
+    {
+      "epoch": 2.2870070848796367,
+      "grad_norm": 0.964019238948822,
+      "learning_rate": 1.1894366383530472e-05,
+      "loss": 0.7473,
+      "step": 28810
+    },
+    {
+      "epoch": 2.287800908928536,
+      "grad_norm": 1.0196411609649658,
+      "learning_rate": 1.1881135720145009e-05,
+      "loss": 0.7552,
+      "step": 28820
+    },
+    {
+      "epoch": 2.2885947329774354,
+      "grad_norm": 0.758881688117981,
+      "learning_rate": 1.1867905056759546e-05,
+      "loss": 0.8031,
+      "step": 28830
+    },
+    {
+      "epoch": 2.289388557026335,
+      "grad_norm": 0.768775999546051,
+      "learning_rate": 1.1854674393374085e-05,
+      "loss": 0.7993,
+      "step": 28840
+    },
+    {
+      "epoch": 2.2901823810752346,
+      "grad_norm": 0.9451555013656616,
+      "learning_rate": 1.184144372998862e-05,
+      "loss": 0.8086,
+      "step": 28850
+    },
+    {
+      "epoch": 2.2909762051241342,
+      "grad_norm": 0.9371961951255798,
+      "learning_rate": 1.182821306660316e-05,
+      "loss": 0.7808,
+      "step": 28860
+    },
+    {
+      "epoch": 2.291770029173034,
+      "grad_norm": 0.7734199166297913,
+      "learning_rate": 1.1814982403217697e-05,
+      "loss": 0.7633,
+      "step": 28870
+    },
+    {
+      "epoch": 2.2925638532219335,
+      "grad_norm": 0.787481427192688,
+      "learning_rate": 1.1801751739832236e-05,
+      "loss": 0.837,
+      "step": 28880
+    },
+    {
+      "epoch": 2.293357677270833,
+      "grad_norm": 0.9613273739814758,
+      "learning_rate": 1.1788521076446773e-05,
+      "loss": 0.726,
+      "step": 28890
+    },
+    {
+      "epoch": 2.2941515013197327,
+      "grad_norm": 0.9527220129966736,
+      "learning_rate": 1.1775290413061312e-05,
+      "loss": 0.7615,
+      "step": 28900
+    },
+    {
+      "epoch": 2.294945325368632,
+      "grad_norm": 0.9209556579589844,
+      "learning_rate": 1.176205974967585e-05,
+      "loss": 0.8216,
+      "step": 28910
+    },
+    {
+      "epoch": 2.2957391494175314,
+      "grad_norm": 0.9544232487678528,
+      "learning_rate": 1.1748829086290388e-05,
+      "loss": 0.794,
+      "step": 28920
+    },
+    {
+      "epoch": 2.296532973466431,
+      "grad_norm": 0.9990895390510559,
+      "learning_rate": 1.1735598422904925e-05,
+      "loss": 0.7551,
+      "step": 28930
+    },
+    {
+      "epoch": 2.2973267975153306,
+      "grad_norm": 1.1743451356887817,
+      "learning_rate": 1.1722367759519463e-05,
+      "loss": 0.7476,
+      "step": 28940
+    },
+    {
+      "epoch": 2.2981206215642302,
+      "grad_norm": 0.9132225513458252,
+      "learning_rate": 1.1709137096134002e-05,
+      "loss": 0.7687,
+      "step": 28950
+    },
+    {
+      "epoch": 2.29891444561313,
+      "grad_norm": 0.7507365345954895,
+      "learning_rate": 1.1695906432748537e-05,
+      "loss": 0.7469,
+      "step": 28960
+    },
+    {
+      "epoch": 2.2997082696620295,
+      "grad_norm": 0.8947992920875549,
+      "learning_rate": 1.1682675769363076e-05,
+      "loss": 0.8074,
+      "step": 28970
+    },
+    {
+      "epoch": 2.300502093710929,
+      "grad_norm": 1.213215708732605,
+      "learning_rate": 1.1669445105977613e-05,
+      "loss": 0.7489,
+      "step": 28980
+    },
+    {
+      "epoch": 2.3012959177598287,
+      "grad_norm": 0.8981990218162537,
+      "learning_rate": 1.1656214442592152e-05,
+      "loss": 0.7278,
+      "step": 28990
+    },
+    {
+      "epoch": 2.3020897418087283,
+      "grad_norm": 0.820511519908905,
+      "learning_rate": 1.164298377920669e-05,
+      "loss": 0.7792,
+      "step": 29000
+    },
+    {
+      "epoch": 2.302883565857628,
+      "grad_norm": 1.1118419170379639,
+      "learning_rate": 1.1629753115821228e-05,
+      "loss": 0.7267,
+      "step": 29010
+    },
+    {
+      "epoch": 2.303677389906527,
+      "grad_norm": 1.0339609384536743,
+      "learning_rate": 1.1616522452435766e-05,
+      "loss": 0.7933,
+      "step": 29020
+    },
+    {
+      "epoch": 2.3044712139554266,
+      "grad_norm": 0.8476637601852417,
+      "learning_rate": 1.1603291789050305e-05,
+      "loss": 0.7775,
+      "step": 29030
+    },
+    {
+      "epoch": 2.3052650380043263,
+      "grad_norm": 0.7746336460113525,
+      "learning_rate": 1.1590061125664842e-05,
+      "loss": 0.7619,
+      "step": 29040
+    },
+    {
+      "epoch": 2.306058862053226,
+      "grad_norm": 1.1253118515014648,
+      "learning_rate": 1.1576830462279379e-05,
+      "loss": 0.7724,
+      "step": 29050
+    },
+    {
+      "epoch": 2.3068526861021255,
+      "grad_norm": 1.1791478395462036,
+      "learning_rate": 1.1563599798893918e-05,
+      "loss": 0.794,
+      "step": 29060
+    },
+    {
+      "epoch": 2.307646510151025,
+      "grad_norm": 0.935023307800293,
+      "learning_rate": 1.1550369135508455e-05,
+      "loss": 0.7548,
+      "step": 29070
+    },
+    {
+      "epoch": 2.3084403341999247,
+      "grad_norm": 0.8298285603523254,
+      "learning_rate": 1.1537138472122992e-05,
+      "loss": 0.7939,
+      "step": 29080
+    },
+    {
+      "epoch": 2.3092341582488243,
+      "grad_norm": 0.8414050340652466,
+      "learning_rate": 1.152390780873753e-05,
+      "loss": 0.8132,
+      "step": 29090
+    },
+    {
+      "epoch": 2.310027982297724,
+      "grad_norm": 0.9045025110244751,
+      "learning_rate": 1.1510677145352069e-05,
+      "loss": 0.7828,
+      "step": 29100
+    },
+    {
+      "epoch": 2.310821806346623,
+      "grad_norm": 0.8417425751686096,
+      "learning_rate": 1.1497446481966606e-05,
+      "loss": 0.8046,
+      "step": 29110
+    },
+    {
+      "epoch": 2.3116156303955226,
+      "grad_norm": 0.9332298040390015,
+      "learning_rate": 1.1484215818581145e-05,
+      "loss": 0.8187,
+      "step": 29120
+    },
+    {
+      "epoch": 2.3124094544444223,
+      "grad_norm": 1.017712116241455,
+      "learning_rate": 1.1470985155195682e-05,
+      "loss": 0.8168,
+      "step": 29130
+    },
+    {
+      "epoch": 2.313203278493322,
+      "grad_norm": 0.9712084531784058,
+      "learning_rate": 1.145775449181022e-05,
+      "loss": 0.7914,
+      "step": 29140
+    },
+    {
+      "epoch": 2.3139971025422215,
+      "grad_norm": 0.8326614499092102,
+      "learning_rate": 1.1444523828424758e-05,
+      "loss": 0.757,
+      "step": 29150
+    },
+    {
+      "epoch": 2.314790926591121,
+      "grad_norm": 0.7456022500991821,
+      "learning_rate": 1.1431293165039295e-05,
+      "loss": 0.8309,
+      "step": 29160
+    },
+    {
+      "epoch": 2.3155847506400207,
+      "grad_norm": 0.895487368106842,
+      "learning_rate": 1.1418062501653834e-05,
+      "loss": 0.7881,
+      "step": 29170
+    },
+    {
+      "epoch": 2.3163785746889203,
+      "grad_norm": 0.9077318906784058,
+      "learning_rate": 1.1404831838268372e-05,
+      "loss": 0.8497,
+      "step": 29180
+    },
+    {
+      "epoch": 2.31717239873782,
+      "grad_norm": 0.7892618775367737,
+      "learning_rate": 1.1391601174882909e-05,
+      "loss": 0.7859,
+      "step": 29190
+    },
+    {
+      "epoch": 2.3179662227867195,
+      "grad_norm": 0.8005306720733643,
+      "learning_rate": 1.1378370511497446e-05,
+      "loss": 0.7707,
+      "step": 29200
+    },
+    {
+      "epoch": 2.318760046835619,
+      "grad_norm": 1.0797233581542969,
+      "learning_rate": 1.1365139848111985e-05,
+      "loss": 0.7446,
+      "step": 29210
+    },
+    {
+      "epoch": 2.3195538708845183,
+      "grad_norm": 0.9487477540969849,
+      "learning_rate": 1.135323225106507e-05,
+      "loss": 0.8089,
+      "step": 29220
+    },
+    {
+      "epoch": 2.320347694933418,
+      "grad_norm": 0.9373265504837036,
+      "learning_rate": 1.1340001587679607e-05,
+      "loss": 0.7926,
+      "step": 29230
+    },
+    {
+      "epoch": 2.3211415189823175,
+      "grad_norm": 1.0408942699432373,
+      "learning_rate": 1.1326770924294144e-05,
+      "loss": 0.8057,
+      "step": 29240
+    },
+    {
+      "epoch": 2.321935343031217,
+      "grad_norm": 0.9446254968643188,
+      "learning_rate": 1.1313540260908683e-05,
+      "loss": 0.8202,
+      "step": 29250
+    },
+    {
+      "epoch": 2.3227291670801167,
+      "grad_norm": 1.1571518182754517,
+      "learning_rate": 1.130030959752322e-05,
+      "loss": 0.7437,
+      "step": 29260
+    },
+    {
+      "epoch": 2.3235229911290163,
+      "grad_norm": 0.8599859476089478,
+      "learning_rate": 1.1287078934137759e-05,
+      "loss": 0.7217,
+      "step": 29270
+    },
+    {
+      "epoch": 2.324316815177916,
+      "grad_norm": 0.9582074880599976,
+      "learning_rate": 1.1273848270752296e-05,
+      "loss": 0.8472,
+      "step": 29280
+    },
+    {
+      "epoch": 2.3251106392268155,
+      "grad_norm": 0.9550096392631531,
+      "learning_rate": 1.1260617607366835e-05,
+      "loss": 0.7503,
+      "step": 29290
+    },
+    {
+      "epoch": 2.325904463275715,
+      "grad_norm": 0.972760796546936,
+      "learning_rate": 1.1247386943981372e-05,
+      "loss": 0.7296,
+      "step": 29300
+    },
+    {
+      "epoch": 2.3266982873246143,
+      "grad_norm": 0.9969944953918457,
+      "learning_rate": 1.123415628059591e-05,
+      "loss": 0.7982,
+      "step": 29310
+    },
+    {
+      "epoch": 2.3274921113735143,
+      "grad_norm": 0.9653575420379639,
+      "learning_rate": 1.1220925617210447e-05,
+      "loss": 0.8543,
+      "step": 29320
+    },
+    {
+      "epoch": 2.3282859354224135,
+      "grad_norm": 0.8924707174301147,
+      "learning_rate": 1.1207694953824986e-05,
+      "loss": 0.7601,
+      "step": 29330
+    },
+    {
+      "epoch": 2.329079759471313,
+      "grad_norm": 0.9261003136634827,
+      "learning_rate": 1.1194464290439523e-05,
+      "loss": 0.7984,
+      "step": 29340
+    },
+    {
+      "epoch": 2.3298735835202127,
+      "grad_norm": 0.678521454334259,
+      "learning_rate": 1.118123362705406e-05,
+      "loss": 0.8383,
+      "step": 29350
+    },
+    {
+      "epoch": 2.3306674075691123,
+      "grad_norm": 1.122550368309021,
+      "learning_rate": 1.1168002963668599e-05,
+      "loss": 0.7964,
+      "step": 29360
+    },
+    {
+      "epoch": 2.331461231618012,
+      "grad_norm": 0.8651812672615051,
+      "learning_rate": 1.1154772300283136e-05,
+      "loss": 0.7974,
+      "step": 29370
+    },
+    {
+      "epoch": 2.3322550556669115,
+      "grad_norm": 1.018391489982605,
+      "learning_rate": 1.1141541636897675e-05,
+      "loss": 0.7908,
+      "step": 29380
+    },
+    {
+      "epoch": 2.333048879715811,
+      "grad_norm": 0.8569789528846741,
+      "learning_rate": 1.1128310973512212e-05,
+      "loss": 0.7332,
+      "step": 29390
+    },
+    {
+      "epoch": 2.3338427037647107,
+      "grad_norm": 0.7940601706504822,
+      "learning_rate": 1.1115080310126751e-05,
+      "loss": 0.8628,
+      "step": 29400
+    },
+    {
+      "epoch": 2.3346365278136103,
+      "grad_norm": 0.9954172372817993,
+      "learning_rate": 1.1101849646741289e-05,
+      "loss": 0.7837,
+      "step": 29410
+    },
+    {
+      "epoch": 2.3354303518625095,
+      "grad_norm": 0.9305082559585571,
+      "learning_rate": 1.1088618983355826e-05,
+      "loss": 0.7587,
+      "step": 29420
+    },
+    {
+      "epoch": 2.336224175911409,
+      "grad_norm": 1.003757357597351,
+      "learning_rate": 1.1075388319970365e-05,
+      "loss": 0.8239,
+      "step": 29430
+    },
+    {
+      "epoch": 2.3370179999603087,
+      "grad_norm": 0.8799276947975159,
+      "learning_rate": 1.1062157656584902e-05,
+      "loss": 0.728,
+      "step": 29440
+    },
+    {
+      "epoch": 2.3378118240092083,
+      "grad_norm": 0.758741021156311,
+      "learning_rate": 1.104892699319944e-05,
+      "loss": 0.806,
+      "step": 29450
+    },
+    {
+      "epoch": 2.338605648058108,
+      "grad_norm": 0.9049004912376404,
+      "learning_rate": 1.1035696329813976e-05,
+      "loss": 0.7504,
+      "step": 29460
+    },
+    {
+      "epoch": 2.3393994721070075,
+      "grad_norm": 1.0414828062057495,
+      "learning_rate": 1.1022465666428515e-05,
+      "loss": 0.77,
+      "step": 29470
+    },
+    {
+      "epoch": 2.340193296155907,
+      "grad_norm": 0.8456864356994629,
+      "learning_rate": 1.1009235003043053e-05,
+      "loss": 0.7447,
+      "step": 29480
+    },
+    {
+      "epoch": 2.3409871202048067,
+      "grad_norm": 0.7757733464241028,
+      "learning_rate": 1.0996004339657591e-05,
+      "loss": 0.7494,
+      "step": 29490
+    },
+    {
+      "epoch": 2.3417809442537063,
+      "grad_norm": 1.131042718887329,
+      "learning_rate": 1.0982773676272129e-05,
+      "loss": 0.8584,
+      "step": 29500
+    },
+    {
+      "epoch": 2.342574768302606,
+      "grad_norm": 0.8263581395149231,
+      "learning_rate": 1.0969543012886668e-05,
+      "loss": 0.741,
+      "step": 29510
+    },
+    {
+      "epoch": 2.3433685923515055,
+      "grad_norm": 1.135141134262085,
+      "learning_rate": 1.0956312349501205e-05,
+      "loss": 0.7473,
+      "step": 29520
+    },
+    {
+      "epoch": 2.3441624164004047,
+      "grad_norm": 1.2775719165802002,
+      "learning_rate": 1.0943081686115742e-05,
+      "loss": 0.7666,
+      "step": 29530
+    },
+    {
+      "epoch": 2.3449562404493043,
+      "grad_norm": 0.8533919453620911,
+      "learning_rate": 1.0929851022730281e-05,
+      "loss": 0.7309,
+      "step": 29540
+    },
+    {
+      "epoch": 2.345750064498204,
+      "grad_norm": 0.7968776822090149,
+      "learning_rate": 1.0916620359344817e-05,
+      "loss": 0.8046,
+      "step": 29550
+    },
+    {
+      "epoch": 2.3465438885471035,
+      "grad_norm": 0.9204419851303101,
+      "learning_rate": 1.0903389695959356e-05,
+      "loss": 0.7839,
+      "step": 29560
+    },
+    {
+      "epoch": 2.347337712596003,
+      "grad_norm": 1.0441004037857056,
+      "learning_rate": 1.0890159032573893e-05,
+      "loss": 0.7742,
+      "step": 29570
+    },
+    {
+      "epoch": 2.3481315366449027,
+      "grad_norm": 0.9476767182350159,
+      "learning_rate": 1.0876928369188432e-05,
+      "loss": 0.7271,
+      "step": 29580
+    },
+    {
+      "epoch": 2.3489253606938023,
+      "grad_norm": 0.8613219857215881,
+      "learning_rate": 1.0863697705802969e-05,
+      "loss": 0.7584,
+      "step": 29590
+    },
+    {
+      "epoch": 2.349719184742702,
+      "grad_norm": 0.8703140616416931,
+      "learning_rate": 1.0850467042417508e-05,
+      "loss": 0.7311,
+      "step": 29600
+    },
+    {
+      "epoch": 2.3505130087916015,
+      "grad_norm": 0.8106393218040466,
+      "learning_rate": 1.0837236379032045e-05,
+      "loss": 0.8058,
+      "step": 29610
+    },
+    {
+      "epoch": 2.3513068328405007,
+      "grad_norm": 1.057108759880066,
+      "learning_rate": 1.0824005715646584e-05,
+      "loss": 0.79,
+      "step": 29620
+    },
+    {
+      "epoch": 2.3521006568894003,
+      "grad_norm": 1.0854650735855103,
+      "learning_rate": 1.0810775052261121e-05,
+      "loss": 0.8291,
+      "step": 29630
+    },
+    {
+      "epoch": 2.3528944809383,
+      "grad_norm": 0.9281707406044006,
+      "learning_rate": 1.0797544388875658e-05,
+      "loss": 0.7736,
+      "step": 29640
+    },
+    {
+      "epoch": 2.3536883049871995,
+      "grad_norm": 0.7807760238647461,
+      "learning_rate": 1.0784313725490197e-05,
+      "loss": 0.7909,
+      "step": 29650
+    },
+    {
+      "epoch": 2.354482129036099,
+      "grad_norm": 0.9227698445320129,
+      "learning_rate": 1.0771083062104735e-05,
+      "loss": 0.8431,
+      "step": 29660
+    },
+    {
+      "epoch": 2.3552759530849987,
+      "grad_norm": 0.8814806938171387,
+      "learning_rate": 1.0757852398719272e-05,
+      "loss": 0.8081,
+      "step": 29670
+    },
+    {
+      "epoch": 2.3560697771338983,
+      "grad_norm": 0.7971763014793396,
+      "learning_rate": 1.0744621735333809e-05,
+      "loss": 0.7334,
+      "step": 29680
+    },
+    {
+      "epoch": 2.356863601182798,
+      "grad_norm": 0.8405244946479797,
+      "learning_rate": 1.0731391071948348e-05,
+      "loss": 0.7273,
+      "step": 29690
+    },
+    {
+      "epoch": 2.3576574252316975,
+      "grad_norm": 1.0113767385482788,
+      "learning_rate": 1.0718160408562885e-05,
+      "loss": 0.832,
+      "step": 29700
+    },
+    {
+      "epoch": 2.358451249280597,
+      "grad_norm": 0.9246639013290405,
+      "learning_rate": 1.0704929745177424e-05,
+      "loss": 0.8362,
+      "step": 29710
+    },
+    {
+      "epoch": 2.3592450733294967,
+      "grad_norm": 1.0259578227996826,
+      "learning_rate": 1.0691699081791961e-05,
+      "loss": 0.7198,
+      "step": 29720
+    },
+    {
+      "epoch": 2.360038897378396,
+      "grad_norm": 0.9511812925338745,
+      "learning_rate": 1.06784684184065e-05,
+      "loss": 0.7828,
+      "step": 29730
+    },
+    {
+      "epoch": 2.3608327214272955,
+      "grad_norm": 0.87998366355896,
+      "learning_rate": 1.0665237755021038e-05,
+      "loss": 0.7572,
+      "step": 29740
+    },
+    {
+      "epoch": 2.361626545476195,
+      "grad_norm": 0.6550692915916443,
+      "learning_rate": 1.0652007091635575e-05,
+      "loss": 0.7808,
+      "step": 29750
+    },
+    {
+      "epoch": 2.3624203695250947,
+      "grad_norm": 0.8000962734222412,
+      "learning_rate": 1.0638776428250114e-05,
+      "loss": 0.784,
+      "step": 29760
+    },
+    {
+      "epoch": 2.3632141935739943,
+      "grad_norm": 0.8203931450843811,
+      "learning_rate": 1.0625545764864651e-05,
+      "loss": 0.8397,
+      "step": 29770
+    },
+    {
+      "epoch": 2.364008017622894,
+      "grad_norm": 0.9608734250068665,
+      "learning_rate": 1.0612315101479188e-05,
+      "loss": 0.7318,
+      "step": 29780
+    },
+    {
+      "epoch": 2.3648018416717935,
+      "grad_norm": 0.9658749103546143,
+      "learning_rate": 1.0599084438093725e-05,
+      "loss": 0.8222,
+      "step": 29790
+    },
+    {
+      "epoch": 2.365595665720693,
+      "grad_norm": 0.9885733127593994,
+      "learning_rate": 1.0585853774708264e-05,
+      "loss": 0.7645,
+      "step": 29800
+    },
+    {
+      "epoch": 2.3663894897695927,
+      "grad_norm": 0.9032514691352844,
+      "learning_rate": 1.0572623111322802e-05,
+      "loss": 0.8039,
+      "step": 29810
+    },
+    {
+      "epoch": 2.367183313818492,
+      "grad_norm": 0.9113055467605591,
+      "learning_rate": 1.055939244793734e-05,
+      "loss": 0.8142,
+      "step": 29820
+    },
+    {
+      "epoch": 2.3679771378673915,
+      "grad_norm": 0.9187504649162292,
+      "learning_rate": 1.0546161784551878e-05,
+      "loss": 0.7926,
+      "step": 29830
+    },
+    {
+      "epoch": 2.368770961916291,
+      "grad_norm": 0.8882543444633484,
+      "learning_rate": 1.0532931121166417e-05,
+      "loss": 0.7267,
+      "step": 29840
+    },
+    {
+      "epoch": 2.3695647859651907,
+      "grad_norm": 0.9198811054229736,
+      "learning_rate": 1.0519700457780954e-05,
+      "loss": 0.7839,
+      "step": 29850
+    },
+    {
+      "epoch": 2.3703586100140903,
+      "grad_norm": 0.8237046003341675,
+      "learning_rate": 1.0506469794395491e-05,
+      "loss": 0.7327,
+      "step": 29860
+    },
+    {
+      "epoch": 2.37115243406299,
+      "grad_norm": 0.9879904985427856,
+      "learning_rate": 1.049323913101003e-05,
+      "loss": 0.759,
+      "step": 29870
+    },
+    {
+      "epoch": 2.3719462581118895,
+      "grad_norm": 0.7641343474388123,
+      "learning_rate": 1.0480008467624567e-05,
+      "loss": 0.7876,
+      "step": 29880
+    },
+    {
+      "epoch": 2.372740082160789,
+      "grad_norm": 0.8497799634933472,
+      "learning_rate": 1.0466777804239106e-05,
+      "loss": 0.788,
+      "step": 29890
+    },
+    {
+      "epoch": 2.3735339062096887,
+      "grad_norm": 0.7273193001747131,
+      "learning_rate": 1.0453547140853642e-05,
+      "loss": 0.8189,
+      "step": 29900
+    },
+    {
+      "epoch": 2.3743277302585883,
+      "grad_norm": 0.9262228012084961,
+      "learning_rate": 1.044031647746818e-05,
+      "loss": 0.8058,
+      "step": 29910
+    },
+    {
+      "epoch": 2.375121554307488,
+      "grad_norm": 0.8716014623641968,
+      "learning_rate": 1.0427085814082718e-05,
+      "loss": 0.7813,
+      "step": 29920
+    },
+    {
+      "epoch": 2.375915378356387,
+      "grad_norm": 1.047127366065979,
+      "learning_rate": 1.0413855150697257e-05,
+      "loss": 0.8362,
+      "step": 29930
+    },
+    {
+      "epoch": 2.3767092024052867,
+      "grad_norm": 1.0029412508010864,
+      "learning_rate": 1.0400624487311794e-05,
+      "loss": 0.7892,
+      "step": 29940
+    },
+    {
+      "epoch": 2.3775030264541863,
+      "grad_norm": 0.8537968397140503,
+      "learning_rate": 1.0387393823926331e-05,
+      "loss": 0.7837,
+      "step": 29950
+    },
+    {
+      "epoch": 2.378296850503086,
+      "grad_norm": 0.8679432272911072,
+      "learning_rate": 1.037416316054087e-05,
+      "loss": 0.7785,
+      "step": 29960
+    },
+    {
+      "epoch": 2.3790906745519855,
+      "grad_norm": 1.063770055770874,
+      "learning_rate": 1.0360932497155407e-05,
+      "loss": 0.7029,
+      "step": 29970
+    },
+    {
+      "epoch": 2.379884498600885,
+      "grad_norm": 0.8927010297775269,
+      "learning_rate": 1.0347701833769946e-05,
+      "loss": 0.7595,
+      "step": 29980
+    },
+    {
+      "epoch": 2.3806783226497847,
+      "grad_norm": 0.8152598738670349,
+      "learning_rate": 1.0334471170384484e-05,
+      "loss": 0.8162,
+      "step": 29990
+    },
+    {
+      "epoch": 2.3814721466986843,
+      "grad_norm": 0.9757214188575745,
+      "learning_rate": 1.0321240506999023e-05,
+      "loss": 0.7675,
+      "step": 30000
+    },
+    {
+      "epoch": 2.382265970747584,
+      "grad_norm": 0.8075728416442871,
+      "learning_rate": 1.0308009843613558e-05,
+      "loss": 0.8226,
+      "step": 30010
+    },
+    {
+      "epoch": 2.383059794796483,
+      "grad_norm": 0.7815035581588745,
+      "learning_rate": 1.0294779180228097e-05,
+      "loss": 0.8277,
+      "step": 30020
+    },
+    {
+      "epoch": 2.383853618845383,
+      "grad_norm": 0.8908788561820984,
+      "learning_rate": 1.0281548516842634e-05,
+      "loss": 0.773,
+      "step": 30030
+    },
+    {
+      "epoch": 2.3846474428942823,
+      "grad_norm": 0.8850157260894775,
+      "learning_rate": 1.0268317853457173e-05,
+      "loss": 0.7643,
+      "step": 30040
+    },
+    {
+      "epoch": 2.385441266943182,
+      "grad_norm": 0.802002489566803,
+      "learning_rate": 1.025508719007171e-05,
+      "loss": 0.8312,
+      "step": 30050
+    },
+    {
+      "epoch": 2.3862350909920815,
+      "grad_norm": 0.9820096492767334,
+      "learning_rate": 1.0241856526686248e-05,
+      "loss": 0.8148,
+      "step": 30060
+    },
+    {
+      "epoch": 2.387028915040981,
+      "grad_norm": 0.934891402721405,
+      "learning_rate": 1.0228625863300787e-05,
+      "loss": 0.7546,
+      "step": 30070
+    },
+    {
+      "epoch": 2.3878227390898807,
+      "grad_norm": 0.9393154978752136,
+      "learning_rate": 1.0215395199915324e-05,
+      "loss": 0.7793,
+      "step": 30080
+    },
+    {
+      "epoch": 2.3886165631387803,
+      "grad_norm": 0.9475399255752563,
+      "learning_rate": 1.0202164536529863e-05,
+      "loss": 0.7756,
+      "step": 30090
+    },
+    {
+      "epoch": 2.38941038718768,
+      "grad_norm": 0.8171314001083374,
+      "learning_rate": 1.01889338731444e-05,
+      "loss": 0.78,
+      "step": 30100
+    },
+    {
+      "epoch": 2.3902042112365796,
+      "grad_norm": 0.85038822889328,
+      "learning_rate": 1.0175703209758939e-05,
+      "loss": 0.7777,
+      "step": 30110
+    },
+    {
+      "epoch": 2.390998035285479,
+      "grad_norm": 0.9316538572311401,
+      "learning_rate": 1.0162472546373476e-05,
+      "loss": 0.8058,
+      "step": 30120
+    },
+    {
+      "epoch": 2.3917918593343783,
+      "grad_norm": 0.7901173233985901,
+      "learning_rate": 1.0149241882988013e-05,
+      "loss": 0.7778,
+      "step": 30130
+    },
+    {
+      "epoch": 2.392585683383278,
+      "grad_norm": 0.907632052898407,
+      "learning_rate": 1.013601121960255e-05,
+      "loss": 0.7577,
+      "step": 30140
+    },
+    {
+      "epoch": 2.3933795074321775,
+      "grad_norm": 0.7853071689605713,
+      "learning_rate": 1.012278055621709e-05,
+      "loss": 0.7905,
+      "step": 30150
+    },
+    {
+      "epoch": 2.394173331481077,
+      "grad_norm": 0.9845629930496216,
+      "learning_rate": 1.0109549892831627e-05,
+      "loss": 0.8268,
+      "step": 30160
+    },
+    {
+      "epoch": 2.3949671555299767,
+      "grad_norm": 0.8937403559684753,
+      "learning_rate": 1.0096319229446164e-05,
+      "loss": 0.842,
+      "step": 30170
+    },
+    {
+      "epoch": 2.3957609795788763,
+      "grad_norm": 0.9141705632209778,
+      "learning_rate": 1.0083088566060703e-05,
+      "loss": 0.7376,
+      "step": 30180
+    },
+    {
+      "epoch": 2.396554803627776,
+      "grad_norm": 0.8283079266548157,
+      "learning_rate": 1.006985790267524e-05,
+      "loss": 0.7864,
+      "step": 30190
+    },
+    {
+      "epoch": 2.3973486276766756,
+      "grad_norm": 1.2408968210220337,
+      "learning_rate": 1.0056627239289779e-05,
+      "loss": 0.8002,
+      "step": 30200
+    },
+    {
+      "epoch": 2.398142451725575,
+      "grad_norm": 0.8985006809234619,
+      "learning_rate": 1.0043396575904316e-05,
+      "loss": 0.7865,
+      "step": 30210
+    },
+    {
+      "epoch": 2.3989362757744748,
+      "grad_norm": 0.8631697297096252,
+      "learning_rate": 1.0030165912518855e-05,
+      "loss": 0.8899,
+      "step": 30220
+    },
+    {
+      "epoch": 2.3997300998233744,
+      "grad_norm": 0.8567438721656799,
+      "learning_rate": 1.0016935249133392e-05,
+      "loss": 0.7726,
+      "step": 30230
+    },
+    {
+      "epoch": 2.4005239238722735,
+      "grad_norm": 0.780764102935791,
+      "learning_rate": 1.0003704585747931e-05,
+      "loss": 0.8075,
+      "step": 30240
+    },
+    {
+      "epoch": 2.401317747921173,
+      "grad_norm": 0.9728577136993408,
+      "learning_rate": 9.990473922362467e-06,
+      "loss": 0.7686,
+      "step": 30250
+    },
+    {
+      "epoch": 2.4021115719700727,
+      "grad_norm": 0.9801949858665466,
+      "learning_rate": 9.977243258977006e-06,
+      "loss": 0.7584,
+      "step": 30260
+    },
+    {
+      "epoch": 2.4029053960189724,
+      "grad_norm": 1.1175134181976318,
+      "learning_rate": 9.964012595591543e-06,
+      "loss": 0.7502,
+      "step": 30270
+    },
+    {
+      "epoch": 2.403699220067872,
+      "grad_norm": 1.078339695930481,
+      "learning_rate": 9.95078193220608e-06,
+      "loss": 0.8411,
+      "step": 30280
+    },
+    {
+      "epoch": 2.4044930441167716,
+      "grad_norm": 1.1007136106491089,
+      "learning_rate": 9.93755126882062e-06,
+      "loss": 0.8423,
+      "step": 30290
+    },
+    {
+      "epoch": 2.405286868165671,
+      "grad_norm": 0.8614094853401184,
+      "learning_rate": 9.924320605435157e-06,
+      "loss": 0.8043,
+      "step": 30300
+    },
+    {
+      "epoch": 2.4060806922145708,
+      "grad_norm": 0.723034679889679,
+      "learning_rate": 9.911089942049695e-06,
+      "loss": 0.822,
+      "step": 30310
+    },
+    {
+      "epoch": 2.4068745162634704,
+      "grad_norm": 0.899814784526825,
+      "learning_rate": 9.897859278664233e-06,
+      "loss": 0.8256,
+      "step": 30320
+    },
+    {
+      "epoch": 2.4076683403123695,
+      "grad_norm": 0.8916921615600586,
+      "learning_rate": 9.884628615278772e-06,
+      "loss": 0.8253,
+      "step": 30330
+    },
+    {
+      "epoch": 2.408462164361269,
+      "grad_norm": 0.8144512176513672,
+      "learning_rate": 9.871397951893309e-06,
+      "loss": 0.8021,
+      "step": 30340
+    },
+    {
+      "epoch": 2.4092559884101687,
+      "grad_norm": 0.9181259274482727,
+      "learning_rate": 9.858167288507846e-06,
+      "loss": 0.8385,
+      "step": 30350
+    },
+    {
+      "epoch": 2.4100498124590684,
+      "grad_norm": 0.9962033629417419,
+      "learning_rate": 9.844936625122383e-06,
+      "loss": 0.8356,
+      "step": 30360
+    },
+    {
+      "epoch": 2.410843636507968,
+      "grad_norm": 1.0186583995819092,
+      "learning_rate": 9.831705961736922e-06,
+      "loss": 0.7458,
+      "step": 30370
+    },
+    {
+      "epoch": 2.4116374605568676,
+      "grad_norm": 0.9851090312004089,
+      "learning_rate": 9.81847529835146e-06,
+      "loss": 0.8241,
+      "step": 30380
+    },
+    {
+      "epoch": 2.412431284605767,
+      "grad_norm": 0.8016131520271301,
+      "learning_rate": 9.805244634965997e-06,
+      "loss": 0.7743,
+      "step": 30390
+    },
+    {
+      "epoch": 2.4132251086546668,
+      "grad_norm": 0.937201976776123,
+      "learning_rate": 9.792013971580536e-06,
+      "loss": 0.7902,
+      "step": 30400
+    },
+    {
+      "epoch": 2.4140189327035664,
+      "grad_norm": 0.6954747438430786,
+      "learning_rate": 9.778783308195073e-06,
+      "loss": 0.8559,
+      "step": 30410
+    },
+    {
+      "epoch": 2.414812756752466,
+      "grad_norm": 0.8315379023551941,
+      "learning_rate": 9.765552644809612e-06,
+      "loss": 0.7645,
+      "step": 30420
+    },
+    {
+      "epoch": 2.4156065808013656,
+      "grad_norm": 0.86857008934021,
+      "learning_rate": 9.752321981424149e-06,
+      "loss": 0.7826,
+      "step": 30430
+    },
+    {
+      "epoch": 2.4164004048502647,
+      "grad_norm": 0.9260841012001038,
+      "learning_rate": 9.739091318038688e-06,
+      "loss": 0.8496,
+      "step": 30440
+    },
+    {
+      "epoch": 2.4171942288991644,
+      "grad_norm": 1.021768569946289,
+      "learning_rate": 9.725860654653225e-06,
+      "loss": 0.7887,
+      "step": 30450
+    },
+    {
+      "epoch": 2.417988052948064,
+      "grad_norm": 0.9660463929176331,
+      "learning_rate": 9.712629991267762e-06,
+      "loss": 0.7776,
+      "step": 30460
+    },
+    {
+      "epoch": 2.4187818769969636,
+      "grad_norm": 0.7337642908096313,
+      "learning_rate": 9.699399327882301e-06,
+      "loss": 0.786,
+      "step": 30470
+    },
+    {
+      "epoch": 2.419575701045863,
+      "grad_norm": 1.0514928102493286,
+      "learning_rate": 9.686168664496839e-06,
+      "loss": 0.7785,
+      "step": 30480
+    },
+    {
+      "epoch": 2.4203695250947628,
+      "grad_norm": 1.1055537462234497,
+      "learning_rate": 9.672938001111376e-06,
+      "loss": 0.7879,
+      "step": 30490
+    },
+    {
+      "epoch": 2.4211633491436624,
+      "grad_norm": 0.9621986746788025,
+      "learning_rate": 9.659707337725913e-06,
+      "loss": 0.7828,
+      "step": 30500
+    },
+    {
+      "epoch": 2.421957173192562,
+      "grad_norm": 0.9570415019989014,
+      "learning_rate": 9.646476674340452e-06,
+      "loss": 0.7945,
+      "step": 30510
+    },
+    {
+      "epoch": 2.4227509972414616,
+      "grad_norm": 0.831770122051239,
+      "learning_rate": 9.63324601095499e-06,
+      "loss": 0.772,
+      "step": 30520
+    },
+    {
+      "epoch": 2.4235448212903608,
+      "grad_norm": 0.7580466270446777,
+      "learning_rate": 9.620015347569528e-06,
+      "loss": 0.8456,
+      "step": 30530
+    },
+    {
+      "epoch": 2.424338645339261,
+      "grad_norm": 1.027631163597107,
+      "learning_rate": 9.606784684184065e-06,
+      "loss": 0.7945,
+      "step": 30540
+    },
+    {
+      "epoch": 2.42513246938816,
+      "grad_norm": 1.0360970497131348,
+      "learning_rate": 9.593554020798604e-06,
+      "loss": 0.8,
+      "step": 30550
+    },
+    {
+      "epoch": 2.4259262934370596,
+      "grad_norm": 0.9129590392112732,
+      "learning_rate": 9.580323357413142e-06,
+      "loss": 0.8057,
+      "step": 30560
+    },
+    {
+      "epoch": 2.426720117485959,
+      "grad_norm": 0.7934035658836365,
+      "learning_rate": 9.567092694027679e-06,
+      "loss": 0.8296,
+      "step": 30570
+    },
+    {
+      "epoch": 2.4275139415348588,
+      "grad_norm": 0.9812830090522766,
+      "learning_rate": 9.553862030642218e-06,
+      "loss": 0.7877,
+      "step": 30580
+    },
+    {
+      "epoch": 2.4283077655837584,
+      "grad_norm": 0.9974318146705627,
+      "learning_rate": 9.540631367256755e-06,
+      "loss": 0.7479,
+      "step": 30590
+    },
+    {
+      "epoch": 2.429101589632658,
+      "grad_norm": 0.9436841011047363,
+      "learning_rate": 9.527400703871292e-06,
+      "loss": 0.8279,
+      "step": 30600
+    },
+    {
+      "epoch": 2.4298954136815576,
+      "grad_norm": 0.9886923432350159,
+      "learning_rate": 9.51417004048583e-06,
+      "loss": 0.7854,
+      "step": 30610
+    },
+    {
+      "epoch": 2.430689237730457,
+      "grad_norm": 0.9457038044929504,
+      "learning_rate": 9.500939377100368e-06,
+      "loss": 0.827,
+      "step": 30620
+    },
+    {
+      "epoch": 2.431483061779357,
+      "grad_norm": 0.93290776014328,
+      "learning_rate": 9.487708713714906e-06,
+      "loss": 0.7439,
+      "step": 30630
+    },
+    {
+      "epoch": 2.432276885828256,
+      "grad_norm": 0.9726458787918091,
+      "learning_rate": 9.474478050329444e-06,
+      "loss": 0.7855,
+      "step": 30640
+    },
+    {
+      "epoch": 2.4330707098771556,
+      "grad_norm": 1.1130033731460571,
+      "learning_rate": 9.461247386943982e-06,
+      "loss": 0.7976,
+      "step": 30650
+    },
+    {
+      "epoch": 2.433864533926055,
+      "grad_norm": 0.7888174653053284,
+      "learning_rate": 9.44801672355852e-06,
+      "loss": 0.7857,
+      "step": 30660
+    },
+    {
+      "epoch": 2.434658357974955,
+      "grad_norm": 0.9603310227394104,
+      "learning_rate": 9.434786060173058e-06,
+      "loss": 0.7558,
+      "step": 30670
+    },
+    {
+      "epoch": 2.4354521820238544,
+      "grad_norm": 0.9045382738113403,
+      "learning_rate": 9.421555396787595e-06,
+      "loss": 0.7363,
+      "step": 30680
+    },
+    {
+      "epoch": 2.436246006072754,
+      "grad_norm": 1.1511313915252686,
+      "learning_rate": 9.408324733402134e-06,
+      "loss": 0.8566,
+      "step": 30690
+    },
+    {
+      "epoch": 2.4370398301216536,
+      "grad_norm": 0.9795113205909729,
+      "learning_rate": 9.395094070016671e-06,
+      "loss": 0.7888,
+      "step": 30700
+    },
+    {
+      "epoch": 2.437833654170553,
+      "grad_norm": 1.0283374786376953,
+      "learning_rate": 9.381863406631208e-06,
+      "loss": 0.8044,
+      "step": 30710
+    },
+    {
+      "epoch": 2.438627478219453,
+      "grad_norm": 1.023207664489746,
+      "learning_rate": 9.368632743245746e-06,
+      "loss": 0.818,
+      "step": 30720
+    },
+    {
+      "epoch": 2.4394213022683524,
+      "grad_norm": 0.7632606029510498,
+      "learning_rate": 9.355402079860285e-06,
+      "loss": 0.8481,
+      "step": 30730
+    },
+    {
+      "epoch": 2.440215126317252,
+      "grad_norm": 0.8076527118682861,
+      "learning_rate": 9.342171416474822e-06,
+      "loss": 0.772,
+      "step": 30740
+    },
+    {
+      "epoch": 2.441008950366151,
+      "grad_norm": 0.9960797429084778,
+      "learning_rate": 9.32894075308936e-06,
+      "loss": 0.7411,
+      "step": 30750
+    },
+    {
+      "epoch": 2.441802774415051,
+      "grad_norm": 0.8365696668624878,
+      "learning_rate": 9.315710089703898e-06,
+      "loss": 0.804,
+      "step": 30760
+    },
+    {
+      "epoch": 2.4425965984639504,
+      "grad_norm": 0.9081131815910339,
+      "learning_rate": 9.302479426318437e-06,
+      "loss": 0.797,
+      "step": 30770
+    },
+    {
+      "epoch": 2.44339042251285,
+      "grad_norm": 0.9612648487091064,
+      "learning_rate": 9.289248762932974e-06,
+      "loss": 0.7829,
+      "step": 30780
+    },
+    {
+      "epoch": 2.4441842465617496,
+      "grad_norm": 0.8103266954421997,
+      "learning_rate": 9.276018099547511e-06,
+      "loss": 0.8183,
+      "step": 30790
+    },
+    {
+      "epoch": 2.444978070610649,
+      "grad_norm": 0.8949236273765564,
+      "learning_rate": 9.26278743616205e-06,
+      "loss": 0.7561,
+      "step": 30800
+    },
+    {
+      "epoch": 2.445771894659549,
+      "grad_norm": 0.8719674348831177,
+      "learning_rate": 9.249556772776588e-06,
+      "loss": 0.7777,
+      "step": 30810
+    },
+    {
+      "epoch": 2.4465657187084484,
+      "grad_norm": 0.959321916103363,
+      "learning_rate": 9.236326109391125e-06,
+      "loss": 0.7687,
+      "step": 30820
+    },
+    {
+      "epoch": 2.447359542757348,
+      "grad_norm": 0.8504953980445862,
+      "learning_rate": 9.223095446005662e-06,
+      "loss": 0.7785,
+      "step": 30830
+    },
+    {
+      "epoch": 2.448153366806247,
+      "grad_norm": 1.073606014251709,
+      "learning_rate": 9.209864782620201e-06,
+      "loss": 0.8652,
+      "step": 30840
+    },
+    {
+      "epoch": 2.448947190855147,
+      "grad_norm": 1.1981532573699951,
+      "learning_rate": 9.196634119234738e-06,
+      "loss": 0.7746,
+      "step": 30850
+    },
+    {
+      "epoch": 2.4497410149040464,
+      "grad_norm": 0.819255530834198,
+      "learning_rate": 9.183403455849277e-06,
+      "loss": 0.8058,
+      "step": 30860
+    },
+    {
+      "epoch": 2.450534838952946,
+      "grad_norm": 1.0335006713867188,
+      "learning_rate": 9.170172792463814e-06,
+      "loss": 0.8036,
+      "step": 30870
+    },
+    {
+      "epoch": 2.4513286630018456,
+      "grad_norm": 0.7458271980285645,
+      "learning_rate": 9.156942129078353e-06,
+      "loss": 0.7471,
+      "step": 30880
+    },
+    {
+      "epoch": 2.452122487050745,
+      "grad_norm": 0.7569029331207275,
+      "learning_rate": 9.14371146569289e-06,
+      "loss": 0.768,
+      "step": 30890
+    },
+    {
+      "epoch": 2.452916311099645,
+      "grad_norm": 1.0082522630691528,
+      "learning_rate": 9.130480802307428e-06,
+      "loss": 0.7802,
+      "step": 30900
+    },
+    {
+      "epoch": 2.4537101351485444,
+      "grad_norm": 1.024379014968872,
+      "learning_rate": 9.117250138921967e-06,
+      "loss": 0.7639,
+      "step": 30910
+    },
+    {
+      "epoch": 2.454503959197444,
+      "grad_norm": 1.1425983905792236,
+      "learning_rate": 9.104019475536504e-06,
+      "loss": 0.7593,
+      "step": 30920
+    },
+    {
+      "epoch": 2.4552977832463436,
+      "grad_norm": 0.9849503040313721,
+      "learning_rate": 9.090788812151043e-06,
+      "loss": 0.8317,
+      "step": 30930
+    },
+    {
+      "epoch": 2.4560916072952432,
+      "grad_norm": 1.0086652040481567,
+      "learning_rate": 9.077558148765578e-06,
+      "loss": 0.7548,
+      "step": 30940
+    },
+    {
+      "epoch": 2.4568854313441424,
+      "grad_norm": 0.8764373660087585,
+      "learning_rate": 9.064327485380117e-06,
+      "loss": 0.7411,
+      "step": 30950
+    },
+    {
+      "epoch": 2.457679255393042,
+      "grad_norm": 1.1179012060165405,
+      "learning_rate": 9.051096821994655e-06,
+      "loss": 0.734,
+      "step": 30960
+    },
+    {
+      "epoch": 2.4584730794419416,
+      "grad_norm": 1.0787594318389893,
+      "learning_rate": 9.037866158609193e-06,
+      "loss": 0.8439,
+      "step": 30970
+    },
+    {
+      "epoch": 2.459266903490841,
+      "grad_norm": 0.9047144651412964,
+      "learning_rate": 9.02463549522373e-06,
+      "loss": 0.7876,
+      "step": 30980
+    },
+    {
+      "epoch": 2.460060727539741,
+      "grad_norm": 0.771873414516449,
+      "learning_rate": 9.01140483183827e-06,
+      "loss": 0.8258,
+      "step": 30990
+    },
+    {
+      "epoch": 2.4608545515886404,
+      "grad_norm": 0.8397606611251831,
+      "learning_rate": 8.998174168452807e-06,
+      "loss": 0.8014,
+      "step": 31000
+    },
+    {
+      "epoch": 2.46164837563754,
+      "grad_norm": 1.1035096645355225,
+      "learning_rate": 8.984943505067344e-06,
+      "loss": 0.7098,
+      "step": 31010
+    },
+    {
+      "epoch": 2.4624421996864396,
+      "grad_norm": 0.8682432770729065,
+      "learning_rate": 8.971712841681883e-06,
+      "loss": 0.7,
+      "step": 31020
+    },
+    {
+      "epoch": 2.4632360237353392,
+      "grad_norm": 0.9221312403678894,
+      "learning_rate": 8.95848217829642e-06,
+      "loss": 0.8001,
+      "step": 31030
+    },
+    {
+      "epoch": 2.4640298477842384,
+      "grad_norm": 1.064479947090149,
+      "learning_rate": 8.94525151491096e-06,
+      "loss": 0.727,
+      "step": 31040
+    },
+    {
+      "epoch": 2.464823671833138,
+      "grad_norm": 1.1401010751724243,
+      "learning_rate": 8.932020851525495e-06,
+      "loss": 0.7556,
+      "step": 31050
+    },
+    {
+      "epoch": 2.4656174958820376,
+      "grad_norm": 0.9812748432159424,
+      "learning_rate": 8.918790188140034e-06,
+      "loss": 0.7788,
+      "step": 31060
+    },
+    {
+      "epoch": 2.466411319930937,
+      "grad_norm": 0.8414878845214844,
+      "learning_rate": 8.905559524754571e-06,
+      "loss": 0.751,
+      "step": 31070
+    },
+    {
+      "epoch": 2.467205143979837,
+      "grad_norm": 0.9719010591506958,
+      "learning_rate": 8.89232886136911e-06,
+      "loss": 0.7413,
+      "step": 31080
+    },
+    {
+      "epoch": 2.4679989680287364,
+      "grad_norm": 0.8958083987236023,
+      "learning_rate": 8.879098197983647e-06,
+      "loss": 0.7575,
+      "step": 31090
+    },
+    {
+      "epoch": 2.468792792077636,
+      "grad_norm": 0.94130939245224,
+      "learning_rate": 8.865867534598184e-06,
+      "loss": 0.7973,
+      "step": 31100
+    },
+    {
+      "epoch": 2.4695866161265356,
+      "grad_norm": 0.8929863572120667,
+      "learning_rate": 8.852636871212723e-06,
+      "loss": 0.8096,
+      "step": 31110
+    },
+    {
+      "epoch": 2.4703804401754352,
+      "grad_norm": 0.9234357476234436,
+      "learning_rate": 8.83940620782726e-06,
+      "loss": 0.8093,
+      "step": 31120
+    },
+    {
+      "epoch": 2.471174264224335,
+      "grad_norm": 0.8647427558898926,
+      "learning_rate": 8.8261755444418e-06,
+      "loss": 0.7498,
+      "step": 31130
+    },
+    {
+      "epoch": 2.4719680882732344,
+      "grad_norm": 0.9515289068222046,
+      "learning_rate": 8.812944881056337e-06,
+      "loss": 0.814,
+      "step": 31140
+    },
+    {
+      "epoch": 2.4727619123221336,
+      "grad_norm": 1.1594178676605225,
+      "learning_rate": 8.799714217670876e-06,
+      "loss": 0.8102,
+      "step": 31150
+    },
+    {
+      "epoch": 2.473555736371033,
+      "grad_norm": 0.7505339980125427,
+      "learning_rate": 8.786483554285413e-06,
+      "loss": 0.8073,
+      "step": 31160
+    },
+    {
+      "epoch": 2.474349560419933,
+      "grad_norm": 1.0167568922042847,
+      "learning_rate": 8.77325289089995e-06,
+      "loss": 0.7955,
+      "step": 31170
+    },
+    {
+      "epoch": 2.4751433844688324,
+      "grad_norm": 0.8681299090385437,
+      "learning_rate": 8.760022227514487e-06,
+      "loss": 0.8455,
+      "step": 31180
+    },
+    {
+      "epoch": 2.475937208517732,
+      "grad_norm": 0.9660631418228149,
+      "learning_rate": 8.746791564129026e-06,
+      "loss": 0.7721,
+      "step": 31190
+    },
+    {
+      "epoch": 2.4767310325666316,
+      "grad_norm": 1.0215500593185425,
+      "learning_rate": 8.733560900743563e-06,
+      "loss": 0.7798,
+      "step": 31200
+    },
+    {
+      "epoch": 2.4775248566155312,
+      "grad_norm": 1.1119939088821411,
+      "learning_rate": 8.7203302373581e-06,
+      "loss": 0.7486,
+      "step": 31210
+    },
+    {
+      "epoch": 2.478318680664431,
+      "grad_norm": 0.8434147238731384,
+      "learning_rate": 8.70709957397264e-06,
+      "loss": 0.7727,
+      "step": 31220
+    },
+    {
+      "epoch": 2.4791125047133304,
+      "grad_norm": 0.9059457778930664,
+      "learning_rate": 8.693868910587177e-06,
+      "loss": 0.7878,
+      "step": 31230
+    },
+    {
+      "epoch": 2.4799063287622296,
+      "grad_norm": 0.9416773319244385,
+      "learning_rate": 8.680638247201716e-06,
+      "loss": 0.7827,
+      "step": 31240
+    },
+    {
+      "epoch": 2.4807001528111297,
+      "grad_norm": 1.0320826768875122,
+      "learning_rate": 8.667407583816253e-06,
+      "loss": 0.803,
+      "step": 31250
+    },
+    {
+      "epoch": 2.481493976860029,
+      "grad_norm": 1.0947669744491577,
+      "learning_rate": 8.654176920430792e-06,
+      "loss": 0.7813,
+      "step": 31260
+    },
+    {
+      "epoch": 2.4822878009089284,
+      "grad_norm": 0.9328486919403076,
+      "learning_rate": 8.640946257045329e-06,
+      "loss": 0.7644,
+      "step": 31270
+    },
+    {
+      "epoch": 2.483081624957828,
+      "grad_norm": 1.0494132041931152,
+      "learning_rate": 8.627715593659868e-06,
+      "loss": 0.8421,
+      "step": 31280
+    },
+    {
+      "epoch": 2.4838754490067276,
+      "grad_norm": 0.7501383423805237,
+      "learning_rate": 8.614484930274404e-06,
+      "loss": 0.7802,
+      "step": 31290
+    },
+    {
+      "epoch": 2.4846692730556272,
+      "grad_norm": 0.9962626099586487,
+      "learning_rate": 8.601254266888942e-06,
+      "loss": 0.7712,
+      "step": 31300
+    },
+    {
+      "epoch": 2.485463097104527,
+      "grad_norm": 0.8576439023017883,
+      "learning_rate": 8.58802360350348e-06,
+      "loss": 0.8233,
+      "step": 31310
+    },
+    {
+      "epoch": 2.4862569211534264,
+      "grad_norm": 1.1776164770126343,
+      "learning_rate": 8.574792940118017e-06,
+      "loss": 0.7753,
+      "step": 31320
+    },
+    {
+      "epoch": 2.487050745202326,
+      "grad_norm": 0.8087422847747803,
+      "learning_rate": 8.561562276732556e-06,
+      "loss": 0.7315,
+      "step": 31330
+    },
+    {
+      "epoch": 2.4878445692512257,
+      "grad_norm": 0.8869718909263611,
+      "learning_rate": 8.548331613347093e-06,
+      "loss": 0.7489,
+      "step": 31340
+    },
+    {
+      "epoch": 2.488638393300125,
+      "grad_norm": 0.9648873805999756,
+      "learning_rate": 8.535100949961632e-06,
+      "loss": 0.7823,
+      "step": 31350
+    },
+    {
+      "epoch": 2.4894322173490244,
+      "grad_norm": 0.9656558632850647,
+      "learning_rate": 8.52187028657617e-06,
+      "loss": 0.7554,
+      "step": 31360
+    },
+    {
+      "epoch": 2.490226041397924,
+      "grad_norm": 1.0111730098724365,
+      "learning_rate": 8.508639623190708e-06,
+      "loss": 0.7656,
+      "step": 31370
+    },
+    {
+      "epoch": 2.4910198654468236,
+      "grad_norm": 0.9099489450454712,
+      "learning_rate": 8.495408959805245e-06,
+      "loss": 0.8378,
+      "step": 31380
+    },
+    {
+      "epoch": 2.4918136894957232,
+      "grad_norm": 0.9399808645248413,
+      "learning_rate": 8.482178296419784e-06,
+      "loss": 0.8008,
+      "step": 31390
+    },
+    {
+      "epoch": 2.492607513544623,
+      "grad_norm": 0.6809582710266113,
+      "learning_rate": 8.46894763303432e-06,
+      "loss": 0.8124,
+      "step": 31400
+    },
+    {
+      "epoch": 2.4934013375935224,
+      "grad_norm": 0.8552061915397644,
+      "learning_rate": 8.455716969648859e-06,
+      "loss": 0.7942,
+      "step": 31410
+    },
+    {
+      "epoch": 2.494195161642422,
+      "grad_norm": 0.8102262616157532,
+      "learning_rate": 8.442486306263396e-06,
+      "loss": 0.7476,
+      "step": 31420
+    },
+    {
+      "epoch": 2.4949889856913217,
+      "grad_norm": 0.8236813545227051,
+      "learning_rate": 8.429255642877933e-06,
+      "loss": 0.8093,
+      "step": 31430
+    },
+    {
+      "epoch": 2.4957828097402213,
+      "grad_norm": 0.8333600163459778,
+      "learning_rate": 8.416024979492472e-06,
+      "loss": 0.7869,
+      "step": 31440
+    },
+    {
+      "epoch": 2.496576633789121,
+      "grad_norm": 1.0917775630950928,
+      "learning_rate": 8.40279431610701e-06,
+      "loss": 0.7584,
+      "step": 31450
+    },
+    {
+      "epoch": 2.49737045783802,
+      "grad_norm": 1.0433661937713623,
+      "learning_rate": 8.389563652721548e-06,
+      "loss": 0.8002,
+      "step": 31460
+    },
+    {
+      "epoch": 2.4981642818869196,
+      "grad_norm": 0.8462589979171753,
+      "learning_rate": 8.376332989336086e-06,
+      "loss": 0.7642,
+      "step": 31470
+    },
+    {
+      "epoch": 2.4989581059358192,
+      "grad_norm": 0.8918420076370239,
+      "learning_rate": 8.363102325950625e-06,
+      "loss": 0.7486,
+      "step": 31480
+    },
+    {
+      "epoch": 2.499751929984719,
+      "grad_norm": 0.9466501474380493,
+      "learning_rate": 8.349871662565162e-06,
+      "loss": 0.8042,
+      "step": 31490
+    },
+    {
+      "epoch": 2.5005457540336185,
+      "grad_norm": 0.9120933413505554,
+      "learning_rate": 8.336640999179699e-06,
+      "loss": 0.761,
+      "step": 31500
+    },
+    {
+      "epoch": 2.501339578082518,
+      "grad_norm": 0.8320806622505188,
+      "learning_rate": 8.323410335794238e-06,
+      "loss": 0.7678,
+      "step": 31510
+    },
+    {
+      "epoch": 2.5021334021314177,
+      "grad_norm": 0.9464184641838074,
+      "learning_rate": 8.310179672408775e-06,
+      "loss": 0.728,
+      "step": 31520
+    },
+    {
+      "epoch": 2.5029272261803173,
+      "grad_norm": 0.831498384475708,
+      "learning_rate": 8.296949009023312e-06,
+      "loss": 0.7438,
+      "step": 31530
+    },
+    {
+      "epoch": 2.503721050229217,
+      "grad_norm": 0.7958276271820068,
+      "learning_rate": 8.28371834563785e-06,
+      "loss": 0.7579,
+      "step": 31540
+    },
+    {
+      "epoch": 2.504514874278116,
+      "grad_norm": 0.859502375125885,
+      "learning_rate": 8.270487682252389e-06,
+      "loss": 0.8299,
+      "step": 31550
+    },
+    {
+      "epoch": 2.505308698327016,
+      "grad_norm": 0.9718039035797119,
+      "learning_rate": 8.257257018866926e-06,
+      "loss": 0.7772,
+      "step": 31560
+    },
+    {
+      "epoch": 2.5061025223759152,
+      "grad_norm": 0.8938490152359009,
+      "learning_rate": 8.244026355481465e-06,
+      "loss": 0.823,
+      "step": 31570
+    },
+    {
+      "epoch": 2.506896346424815,
+      "grad_norm": 0.9374427199363708,
+      "learning_rate": 8.230795692096002e-06,
+      "loss": 0.8246,
+      "step": 31580
+    },
+    {
+      "epoch": 2.5076901704737145,
+      "grad_norm": 0.8214637637138367,
+      "learning_rate": 8.217565028710541e-06,
+      "loss": 0.7747,
+      "step": 31590
+    },
+    {
+      "epoch": 2.508483994522614,
+      "grad_norm": 1.0425516366958618,
+      "learning_rate": 8.204334365325078e-06,
+      "loss": 0.7211,
+      "step": 31600
+    },
+    {
+      "epoch": 2.5092778185715137,
+      "grad_norm": 0.8850832581520081,
+      "learning_rate": 8.191103701939615e-06,
+      "loss": 0.7707,
+      "step": 31610
+    },
+    {
+      "epoch": 2.5100716426204133,
+      "grad_norm": 0.7210223078727722,
+      "learning_rate": 8.177873038554154e-06,
+      "loss": 0.7833,
+      "step": 31620
+    },
+    {
+      "epoch": 2.510865466669313,
+      "grad_norm": 0.9423604011535645,
+      "learning_rate": 8.164642375168692e-06,
+      "loss": 0.7697,
+      "step": 31630
+    },
+    {
+      "epoch": 2.511659290718212,
+      "grad_norm": 0.937473475933075,
+      "learning_rate": 8.151411711783229e-06,
+      "loss": 0.7924,
+      "step": 31640
+    },
+    {
+      "epoch": 2.512453114767112,
+      "grad_norm": 1.0225754976272583,
+      "learning_rate": 8.138181048397766e-06,
+      "loss": 0.8051,
+      "step": 31650
+    },
+    {
+      "epoch": 2.5132469388160112,
+      "grad_norm": 0.9610739350318909,
+      "learning_rate": 8.124950385012305e-06,
+      "loss": 0.7844,
+      "step": 31660
+    },
+    {
+      "epoch": 2.514040762864911,
+      "grad_norm": 0.6740348935127258,
+      "learning_rate": 8.111719721626842e-06,
+      "loss": 0.7502,
+      "step": 31670
+    },
+    {
+      "epoch": 2.5148345869138105,
+      "grad_norm": 0.9693344235420227,
+      "learning_rate": 8.098489058241381e-06,
+      "loss": 0.7821,
+      "step": 31680
+    },
+    {
+      "epoch": 2.51562841096271,
+      "grad_norm": 1.042784333229065,
+      "learning_rate": 8.085258394855918e-06,
+      "loss": 0.8042,
+      "step": 31690
+    },
+    {
+      "epoch": 2.5164222350116097,
+      "grad_norm": 0.726180911064148,
+      "learning_rate": 8.072027731470457e-06,
+      "loss": 0.7687,
+      "step": 31700
+    },
+    {
+      "epoch": 2.5172160590605093,
+      "grad_norm": 0.8935758471488953,
+      "learning_rate": 8.058797068084994e-06,
+      "loss": 0.7846,
+      "step": 31710
+    },
+    {
+      "epoch": 2.518009883109409,
+      "grad_norm": 0.884282648563385,
+      "learning_rate": 8.045566404699532e-06,
+      "loss": 0.7613,
+      "step": 31720
+    },
+    {
+      "epoch": 2.5188037071583085,
+      "grad_norm": 0.8129022717475891,
+      "learning_rate": 8.03233574131407e-06,
+      "loss": 0.8585,
+      "step": 31730
+    },
+    {
+      "epoch": 2.519597531207208,
+      "grad_norm": 0.9726604223251343,
+      "learning_rate": 8.019105077928608e-06,
+      "loss": 0.7648,
+      "step": 31740
+    },
+    {
+      "epoch": 2.5203913552561072,
+      "grad_norm": 0.7345698475837708,
+      "learning_rate": 8.005874414543145e-06,
+      "loss": 0.8031,
+      "step": 31750
+    },
+    {
+      "epoch": 2.5211851793050073,
+      "grad_norm": 0.9011138081550598,
+      "learning_rate": 7.992643751157682e-06,
+      "loss": 0.7917,
+      "step": 31760
+    },
+    {
+      "epoch": 2.5219790033539065,
+      "grad_norm": 0.9308391213417053,
+      "learning_rate": 7.979413087772221e-06,
+      "loss": 0.8107,
+      "step": 31770
+    },
+    {
+      "epoch": 2.522772827402806,
+      "grad_norm": 0.9620044827461243,
+      "learning_rate": 7.966182424386758e-06,
+      "loss": 0.7515,
+      "step": 31780
+    },
+    {
+      "epoch": 2.5235666514517057,
+      "grad_norm": 0.7784889936447144,
+      "learning_rate": 7.952951761001297e-06,
+      "loss": 0.7873,
+      "step": 31790
+    },
+    {
+      "epoch": 2.5243604755006053,
+      "grad_norm": 0.9141636490821838,
+      "learning_rate": 7.939721097615835e-06,
+      "loss": 0.8912,
+      "step": 31800
+    },
+    {
+      "epoch": 2.525154299549505,
+      "grad_norm": 0.7752735614776611,
+      "learning_rate": 7.926490434230374e-06,
+      "loss": 0.7585,
+      "step": 31810
+    },
+    {
+      "epoch": 2.5259481235984045,
+      "grad_norm": 0.7791392803192139,
+      "learning_rate": 7.91325977084491e-06,
+      "loss": 0.8186,
+      "step": 31820
+    },
+    {
+      "epoch": 2.526741947647304,
+      "grad_norm": 1.1256967782974243,
+      "learning_rate": 7.900029107459448e-06,
+      "loss": 0.7574,
+      "step": 31830
+    },
+    {
+      "epoch": 2.5275357716962037,
+      "grad_norm": 1.1258056163787842,
+      "learning_rate": 7.886798444073987e-06,
+      "loss": 0.7826,
+      "step": 31840
+    },
+    {
+      "epoch": 2.5283295957451033,
+      "grad_norm": 1.118212103843689,
+      "learning_rate": 7.873567780688524e-06,
+      "loss": 0.8267,
+      "step": 31850
+    },
+    {
+      "epoch": 2.5291234197940025,
+      "grad_norm": 1.0277642011642456,
+      "learning_rate": 7.860337117303061e-06,
+      "loss": 0.7603,
+      "step": 31860
+    },
+    {
+      "epoch": 2.529917243842902,
+      "grad_norm": 0.9534600377082825,
+      "learning_rate": 7.847106453917599e-06,
+      "loss": 0.7872,
+      "step": 31870
+    },
+    {
+      "epoch": 2.5307110678918017,
+      "grad_norm": 0.9541226625442505,
+      "learning_rate": 7.833875790532138e-06,
+      "loss": 0.7756,
+      "step": 31880
+    },
+    {
+      "epoch": 2.5315048919407013,
+      "grad_norm": 0.8130003213882446,
+      "learning_rate": 7.820645127146675e-06,
+      "loss": 0.8042,
+      "step": 31890
+    },
+    {
+      "epoch": 2.532298715989601,
+      "grad_norm": 1.0782946348190308,
+      "learning_rate": 7.807414463761214e-06,
+      "loss": 0.7568,
+      "step": 31900
+    },
+    {
+      "epoch": 2.5330925400385005,
+      "grad_norm": 1.0216178894042969,
+      "learning_rate": 7.794183800375751e-06,
+      "loss": 0.7805,
+      "step": 31910
+    },
+    {
+      "epoch": 2.5338863640874,
+      "grad_norm": 1.0545295476913452,
+      "learning_rate": 7.78095313699029e-06,
+      "loss": 0.7286,
+      "step": 31920
+    },
+    {
+      "epoch": 2.5346801881362997,
+      "grad_norm": 0.8440236449241638,
+      "learning_rate": 7.767722473604827e-06,
+      "loss": 0.731,
+      "step": 31930
+    },
+    {
+      "epoch": 2.5354740121851993,
+      "grad_norm": 1.0648548603057861,
+      "learning_rate": 7.754491810219364e-06,
+      "loss": 0.8121,
+      "step": 31940
+    },
+    {
+      "epoch": 2.5362678362340985,
+      "grad_norm": 1.0364627838134766,
+      "learning_rate": 7.741261146833903e-06,
+      "loss": 0.7799,
+      "step": 31950
+    },
+    {
+      "epoch": 2.5370616602829985,
+      "grad_norm": 1.0399836301803589,
+      "learning_rate": 7.72803048344844e-06,
+      "loss": 0.8083,
+      "step": 31960
+    },
+    {
+      "epoch": 2.5378554843318977,
+      "grad_norm": 0.8876070380210876,
+      "learning_rate": 7.71479982006298e-06,
+      "loss": 0.8043,
+      "step": 31970
+    },
+    {
+      "epoch": 2.5386493083807973,
+      "grad_norm": 0.9416910409927368,
+      "learning_rate": 7.701569156677515e-06,
+      "loss": 0.7913,
+      "step": 31980
+    },
+    {
+      "epoch": 2.539443132429697,
+      "grad_norm": 0.7986873984336853,
+      "learning_rate": 7.688338493292054e-06,
+      "loss": 0.7693,
+      "step": 31990
+    },
+    {
+      "epoch": 2.5402369564785965,
+      "grad_norm": 1.0350062847137451,
+      "learning_rate": 7.675107829906591e-06,
+      "loss": 0.7858,
+      "step": 32000
+    },
+    {
+      "epoch": 2.541030780527496,
+      "grad_norm": 0.820899486541748,
+      "learning_rate": 7.66187716652113e-06,
+      "loss": 0.781,
+      "step": 32010
+    },
+    {
+      "epoch": 2.5418246045763957,
+      "grad_norm": 0.9244166612625122,
+      "learning_rate": 7.648646503135667e-06,
+      "loss": 0.7373,
+      "step": 32020
+    },
+    {
+      "epoch": 2.5426184286252953,
+      "grad_norm": 0.9774866104125977,
+      "learning_rate": 7.635415839750206e-06,
+      "loss": 0.7804,
+      "step": 32030
+    },
+    {
+      "epoch": 2.543412252674195,
+      "grad_norm": 0.9632622599601746,
+      "learning_rate": 7.6221851763647435e-06,
+      "loss": 0.8138,
+      "step": 32040
+    },
+    {
+      "epoch": 2.5442060767230945,
+      "grad_norm": 0.9499511122703552,
+      "learning_rate": 7.6089545129792815e-06,
+      "loss": 0.8123,
+      "step": 32050
+    },
+    {
+      "epoch": 2.5449999007719937,
+      "grad_norm": 0.9002213478088379,
+      "learning_rate": 7.59572384959382e-06,
+      "loss": 0.7807,
+      "step": 32060
+    },
+    {
+      "epoch": 2.5457937248208937,
+      "grad_norm": 0.965908944606781,
+      "learning_rate": 7.582493186208358e-06,
+      "loss": 0.8199,
+      "step": 32070
+    },
+    {
+      "epoch": 2.546587548869793,
+      "grad_norm": 0.9211716055870056,
+      "learning_rate": 7.569262522822895e-06,
+      "loss": 0.6892,
+      "step": 32080
+    },
+    {
+      "epoch": 2.5473813729186925,
+      "grad_norm": 1.01425039768219,
+      "learning_rate": 7.556031859437433e-06,
+      "loss": 0.7395,
+      "step": 32090
+    },
+    {
+      "epoch": 2.548175196967592,
+      "grad_norm": 0.9339504241943359,
+      "learning_rate": 7.54280119605197e-06,
+      "loss": 0.7235,
+      "step": 32100
+    },
+    {
+      "epoch": 2.5489690210164917,
+      "grad_norm": 1.2052608728408813,
+      "learning_rate": 7.5295705326665075e-06,
+      "loss": 0.7632,
+      "step": 32110
+    },
+    {
+      "epoch": 2.5497628450653913,
+      "grad_norm": 0.9481042623519897,
+      "learning_rate": 7.5163398692810456e-06,
+      "loss": 0.7757,
+      "step": 32120
+    },
+    {
+      "epoch": 2.550556669114291,
+      "grad_norm": 1.060491681098938,
+      "learning_rate": 7.503109205895584e-06,
+      "loss": 0.8118,
+      "step": 32130
+    },
+    {
+      "epoch": 2.5513504931631905,
+      "grad_norm": 0.9081485867500305,
+      "learning_rate": 7.489878542510122e-06,
+      "loss": 0.7584,
+      "step": 32140
+    },
+    {
+      "epoch": 2.5521443172120897,
+      "grad_norm": 0.8219790458679199,
+      "learning_rate": 7.47664787912466e-06,
+      "loss": 0.8306,
+      "step": 32150
+    },
+    {
+      "epoch": 2.5529381412609897,
+      "grad_norm": 0.9022849202156067,
+      "learning_rate": 7.463417215739198e-06,
+      "loss": 0.7768,
+      "step": 32160
+    },
+    {
+      "epoch": 2.553731965309889,
+      "grad_norm": 1.0158053636550903,
+      "learning_rate": 7.450186552353736e-06,
+      "loss": 0.7782,
+      "step": 32170
+    },
+    {
+      "epoch": 2.5545257893587885,
+      "grad_norm": 0.8737016916275024,
+      "learning_rate": 7.436955888968273e-06,
+      "loss": 0.744,
+      "step": 32180
+    },
+    {
+      "epoch": 2.555319613407688,
+      "grad_norm": 0.9380881190299988,
+      "learning_rate": 7.423725225582811e-06,
+      "loss": 0.7312,
+      "step": 32190
+    },
+    {
+      "epoch": 2.5561134374565877,
+      "grad_norm": 0.9312672019004822,
+      "learning_rate": 7.410494562197349e-06,
+      "loss": 0.7955,
+      "step": 32200
+    },
+    {
+      "epoch": 2.5569072615054873,
+      "grad_norm": 1.055057168006897,
+      "learning_rate": 7.397263898811886e-06,
+      "loss": 0.7535,
+      "step": 32210
+    },
+    {
+      "epoch": 2.557701085554387,
+      "grad_norm": 0.9864413142204285,
+      "learning_rate": 7.384033235426424e-06,
+      "loss": 0.8113,
+      "step": 32220
+    },
+    {
+      "epoch": 2.5584949096032865,
+      "grad_norm": 0.8412654995918274,
+      "learning_rate": 7.370802572040962e-06,
+      "loss": 0.77,
+      "step": 32230
+    },
+    {
+      "epoch": 2.559288733652186,
+      "grad_norm": 0.8634358048439026,
+      "learning_rate": 7.3575719086555e-06,
+      "loss": 0.7868,
+      "step": 32240
+    },
+    {
+      "epoch": 2.5600825577010857,
+      "grad_norm": 1.081010103225708,
+      "learning_rate": 7.344341245270038e-06,
+      "loss": 0.8323,
+      "step": 32250
+    },
+    {
+      "epoch": 2.560876381749985,
+      "grad_norm": 0.9374126195907593,
+      "learning_rate": 7.331110581884576e-06,
+      "loss": 0.78,
+      "step": 32260
+    },
+    {
+      "epoch": 2.561670205798885,
+      "grad_norm": 1.0620479583740234,
+      "learning_rate": 7.317879918499114e-06,
+      "loss": 0.7825,
+      "step": 32270
+    },
+    {
+      "epoch": 2.562464029847784,
+      "grad_norm": 0.8730692863464355,
+      "learning_rate": 7.304649255113652e-06,
+      "loss": 0.7922,
+      "step": 32280
+    },
+    {
+      "epoch": 2.5632578538966837,
+      "grad_norm": 1.0309388637542725,
+      "learning_rate": 7.2914185917281895e-06,
+      "loss": 0.79,
+      "step": 32290
+    },
+    {
+      "epoch": 2.5640516779455833,
+      "grad_norm": 0.8236114382743835,
+      "learning_rate": 7.278187928342728e-06,
+      "loss": 0.7463,
+      "step": 32300
+    },
+    {
+      "epoch": 2.564845501994483,
+      "grad_norm": 0.9980899095535278,
+      "learning_rate": 7.264957264957266e-06,
+      "loss": 0.7699,
+      "step": 32310
+    },
+    {
+      "epoch": 2.5656393260433825,
+      "grad_norm": 0.8803249597549438,
+      "learning_rate": 7.251726601571804e-06,
+      "loss": 0.8637,
+      "step": 32320
+    },
+    {
+      "epoch": 2.566433150092282,
+      "grad_norm": 0.9461043477058411,
+      "learning_rate": 7.23849593818634e-06,
+      "loss": 0.8058,
+      "step": 32330
+    },
+    {
+      "epoch": 2.5672269741411817,
+      "grad_norm": 1.0341662168502808,
+      "learning_rate": 7.225265274800878e-06,
+      "loss": 0.8059,
+      "step": 32340
+    },
+    {
+      "epoch": 2.5680207981900813,
+      "grad_norm": 0.7638232111930847,
+      "learning_rate": 7.212034611415416e-06,
+      "loss": 0.8157,
+      "step": 32350
+    },
+    {
+      "epoch": 2.568814622238981,
+      "grad_norm": 0.8748646974563599,
+      "learning_rate": 7.198803948029954e-06,
+      "loss": 0.7982,
+      "step": 32360
+    },
+    {
+      "epoch": 2.56960844628788,
+      "grad_norm": 0.9040183424949646,
+      "learning_rate": 7.1855732846444925e-06,
+      "loss": 0.8113,
+      "step": 32370
+    },
+    {
+      "epoch": 2.5704022703367797,
+      "grad_norm": 0.7992666959762573,
+      "learning_rate": 7.1723426212590306e-06,
+      "loss": 0.791,
+      "step": 32380
+    },
+    {
+      "epoch": 2.5711960943856793,
+      "grad_norm": 0.9834904074668884,
+      "learning_rate": 7.159111957873569e-06,
+      "loss": 0.8094,
+      "step": 32390
+    },
+    {
+      "epoch": 2.571989918434579,
+      "grad_norm": 1.0673480033874512,
+      "learning_rate": 7.145881294488106e-06,
+      "loss": 0.7499,
+      "step": 32400
+    },
+    {
+      "epoch": 2.5727837424834785,
+      "grad_norm": 1.0334373712539673,
+      "learning_rate": 7.132650631102644e-06,
+      "loss": 0.7715,
+      "step": 32410
+    },
+    {
+      "epoch": 2.573577566532378,
+      "grad_norm": 0.9567040205001831,
+      "learning_rate": 7.119419967717182e-06,
+      "loss": 0.7818,
+      "step": 32420
+    },
+    {
+      "epoch": 2.5743713905812777,
+      "grad_norm": 0.890590488910675,
+      "learning_rate": 7.10618930433172e-06,
+      "loss": 0.7509,
+      "step": 32430
+    },
+    {
+      "epoch": 2.5751652146301773,
+      "grad_norm": 0.8644120693206787,
+      "learning_rate": 7.0929586409462565e-06,
+      "loss": 0.7556,
+      "step": 32440
+    },
+    {
+      "epoch": 2.575959038679077,
+      "grad_norm": 0.8161708116531372,
+      "learning_rate": 7.079727977560795e-06,
+      "loss": 0.8253,
+      "step": 32450
+    },
+    {
+      "epoch": 2.576752862727976,
+      "grad_norm": 0.8945040106773376,
+      "learning_rate": 7.066497314175333e-06,
+      "loss": 0.754,
+      "step": 32460
+    },
+    {
+      "epoch": 2.577546686776876,
+      "grad_norm": 0.9900794625282288,
+      "learning_rate": 7.053266650789871e-06,
+      "loss": 0.7775,
+      "step": 32470
+    },
+    {
+      "epoch": 2.5783405108257753,
+      "grad_norm": 0.7387442588806152,
+      "learning_rate": 7.040035987404409e-06,
+      "loss": 0.8277,
+      "step": 32480
+    },
+    {
+      "epoch": 2.579134334874675,
+      "grad_norm": 0.8887678980827332,
+      "learning_rate": 7.026805324018947e-06,
+      "loss": 0.8033,
+      "step": 32490
+    },
+    {
+      "epoch": 2.5799281589235745,
+      "grad_norm": 1.0268609523773193,
+      "learning_rate": 7.013574660633485e-06,
+      "loss": 0.7224,
+      "step": 32500
+    },
+    {
+      "epoch": 2.580721982972474,
+      "grad_norm": 0.7121632099151611,
+      "learning_rate": 7.000343997248022e-06,
+      "loss": 0.8311,
+      "step": 32510
+    },
+    {
+      "epoch": 2.5815158070213737,
+      "grad_norm": 0.9221680164337158,
+      "learning_rate": 6.98711333386256e-06,
+      "loss": 0.8105,
+      "step": 32520
+    },
+    {
+      "epoch": 2.5823096310702733,
+      "grad_norm": 0.8981099128723145,
+      "learning_rate": 6.973882670477098e-06,
+      "loss": 0.7886,
+      "step": 32530
+    },
+    {
+      "epoch": 2.583103455119173,
+      "grad_norm": 0.8529419302940369,
+      "learning_rate": 6.9606520070916365e-06,
+      "loss": 0.7879,
+      "step": 32540
+    },
+    {
+      "epoch": 2.5838972791680725,
+      "grad_norm": 0.9864525198936462,
+      "learning_rate": 6.9474213437061745e-06,
+      "loss": 0.7829,
+      "step": 32550
+    },
+    {
+      "epoch": 2.584691103216972,
+      "grad_norm": 1.1552358865737915,
+      "learning_rate": 6.934190680320711e-06,
+      "loss": 0.737,
+      "step": 32560
+    },
+    {
+      "epoch": 2.5854849272658713,
+      "grad_norm": 0.9277712106704712,
+      "learning_rate": 6.920960016935249e-06,
+      "loss": 0.773,
+      "step": 32570
+    },
+    {
+      "epoch": 2.586278751314771,
+      "grad_norm": 1.1222898960113525,
+      "learning_rate": 6.907729353549787e-06,
+      "loss": 0.7899,
+      "step": 32580
+    },
+    {
+      "epoch": 2.5870725753636705,
+      "grad_norm": 0.9069304466247559,
+      "learning_rate": 6.894498690164325e-06,
+      "loss": 0.7647,
+      "step": 32590
+    },
+    {
+      "epoch": 2.58786639941257,
+      "grad_norm": 1.0726667642593384,
+      "learning_rate": 6.881268026778863e-06,
+      "loss": 0.752,
+      "step": 32600
+    },
+    {
+      "epoch": 2.5886602234614697,
+      "grad_norm": 1.020435094833374,
+      "learning_rate": 6.8680373633934005e-06,
+      "loss": 0.7844,
+      "step": 32610
+    },
+    {
+      "epoch": 2.5894540475103693,
+      "grad_norm": 0.928993821144104,
+      "learning_rate": 6.8548067000079386e-06,
+      "loss": 0.8021,
+      "step": 32620
+    },
+    {
+      "epoch": 2.590247871559269,
+      "grad_norm": 0.8852306604385376,
+      "learning_rate": 6.841576036622477e-06,
+      "loss": 0.7672,
+      "step": 32630
+    },
+    {
+      "epoch": 2.5910416956081685,
+      "grad_norm": 0.8123989701271057,
+      "learning_rate": 6.828345373237015e-06,
+      "loss": 0.8049,
+      "step": 32640
+    },
+    {
+      "epoch": 2.591835519657068,
+      "grad_norm": 0.8990057110786438,
+      "learning_rate": 6.815114709851553e-06,
+      "loss": 0.72,
+      "step": 32650
+    },
+    {
+      "epoch": 2.5926293437059673,
+      "grad_norm": 1.0066652297973633,
+      "learning_rate": 6.801884046466091e-06,
+      "loss": 0.7786,
+      "step": 32660
+    },
+    {
+      "epoch": 2.5934231677548674,
+      "grad_norm": 0.984686017036438,
+      "learning_rate": 6.788653383080627e-06,
+      "loss": 0.7876,
+      "step": 32670
+    },
+    {
+      "epoch": 2.5942169918037665,
+      "grad_norm": 0.9421548247337341,
+      "learning_rate": 6.775422719695165e-06,
+      "loss": 0.7983,
+      "step": 32680
+    },
+    {
+      "epoch": 2.595010815852666,
+      "grad_norm": 0.9953796863555908,
+      "learning_rate": 6.762192056309703e-06,
+      "loss": 0.7933,
+      "step": 32690
+    },
+    {
+      "epoch": 2.5958046399015657,
+      "grad_norm": 1.0435398817062378,
+      "learning_rate": 6.7489613929242415e-06,
+      "loss": 0.8087,
+      "step": 32700
+    },
+    {
+      "epoch": 2.5965984639504653,
+      "grad_norm": 0.8738863468170166,
+      "learning_rate": 6.73573072953878e-06,
+      "loss": 0.7782,
+      "step": 32710
+    },
+    {
+      "epoch": 2.597392287999365,
+      "grad_norm": 1.0245190858840942,
+      "learning_rate": 6.722500066153317e-06,
+      "loss": 0.8013,
+      "step": 32720
+    },
+    {
+      "epoch": 2.5981861120482646,
+      "grad_norm": 0.965116560459137,
+      "learning_rate": 6.709269402767855e-06,
+      "loss": 0.8208,
+      "step": 32730
+    },
+    {
+      "epoch": 2.598979936097164,
+      "grad_norm": 1.0972223281860352,
+      "learning_rate": 6.696038739382393e-06,
+      "loss": 0.7588,
+      "step": 32740
+    },
+    {
+      "epoch": 2.5997737601460638,
+      "grad_norm": 1.1245334148406982,
+      "learning_rate": 6.682808075996931e-06,
+      "loss": 0.8014,
+      "step": 32750
+    },
+    {
+      "epoch": 2.6005675841949634,
+      "grad_norm": 0.9737082719802856,
+      "learning_rate": 6.669577412611469e-06,
+      "loss": 0.7959,
+      "step": 32760
+    },
+    {
+      "epoch": 2.6013614082438625,
+      "grad_norm": 0.9256748557090759,
+      "learning_rate": 6.656346749226007e-06,
+      "loss": 0.7476,
+      "step": 32770
+    },
+    {
+      "epoch": 2.6021552322927626,
+      "grad_norm": 0.9990182518959045,
+      "learning_rate": 6.643116085840545e-06,
+      "loss": 0.8093,
+      "step": 32780
+    },
+    {
+      "epoch": 2.6029490563416617,
+      "grad_norm": 1.174641489982605,
+      "learning_rate": 6.629885422455082e-06,
+      "loss": 0.7579,
+      "step": 32790
+    },
+    {
+      "epoch": 2.6037428803905613,
+      "grad_norm": 0.8486537933349609,
+      "learning_rate": 6.61665475906962e-06,
+      "loss": 0.755,
+      "step": 32800
+    },
+    {
+      "epoch": 2.604536704439461,
+      "grad_norm": 1.0289491415023804,
+      "learning_rate": 6.603424095684158e-06,
+      "loss": 0.7426,
+      "step": 32810
+    },
+    {
+      "epoch": 2.6053305284883606,
+      "grad_norm": 0.9940427541732788,
+      "learning_rate": 6.590193432298696e-06,
+      "loss": 0.802,
+      "step": 32820
+    },
+    {
+      "epoch": 2.60612435253726,
+      "grad_norm": 0.9397895336151123,
+      "learning_rate": 6.576962768913233e-06,
+      "loss": 0.7455,
+      "step": 32830
+    },
+    {
+      "epoch": 2.6069181765861598,
+      "grad_norm": 0.9215680360794067,
+      "learning_rate": 6.563732105527771e-06,
+      "loss": 0.8156,
+      "step": 32840
+    },
+    {
+      "epoch": 2.6077120006350594,
+      "grad_norm": 0.9983633160591125,
+      "learning_rate": 6.550501442142309e-06,
+      "loss": 0.749,
+      "step": 32850
+    },
+    {
+      "epoch": 2.6085058246839585,
+      "grad_norm": 0.9694678783416748,
+      "learning_rate": 6.537270778756847e-06,
+      "loss": 0.7625,
+      "step": 32860
+    },
+    {
+      "epoch": 2.6092996487328586,
+      "grad_norm": 0.9712207913398743,
+      "learning_rate": 6.5240401153713855e-06,
+      "loss": 0.7712,
+      "step": 32870
+    },
+    {
+      "epoch": 2.6100934727817577,
+      "grad_norm": 0.8602631688117981,
+      "learning_rate": 6.5108094519859236e-06,
+      "loss": 0.7941,
+      "step": 32880
+    },
+    {
+      "epoch": 2.6108872968306573,
+      "grad_norm": 1.1966452598571777,
+      "learning_rate": 6.497578788600462e-06,
+      "loss": 0.7736,
+      "step": 32890
+    },
+    {
+      "epoch": 2.611681120879557,
+      "grad_norm": 1.0157437324523926,
+      "learning_rate": 6.484348125214998e-06,
+      "loss": 0.7439,
+      "step": 32900
+    },
+    {
+      "epoch": 2.6124749449284566,
+      "grad_norm": 0.9984763264656067,
+      "learning_rate": 6.471117461829536e-06,
+      "loss": 0.7874,
+      "step": 32910
+    },
+    {
+      "epoch": 2.613268768977356,
+      "grad_norm": 0.9150610566139221,
+      "learning_rate": 6.457886798444074e-06,
+      "loss": 0.6659,
+      "step": 32920
+    },
+    {
+      "epoch": 2.6140625930262558,
+      "grad_norm": 1.074073076248169,
+      "learning_rate": 6.444656135058611e-06,
+      "loss": 0.8417,
+      "step": 32930
+    },
+    {
+      "epoch": 2.6148564170751554,
+      "grad_norm": 0.8862860798835754,
+      "learning_rate": 6.4314254716731495e-06,
+      "loss": 0.7755,
+      "step": 32940
+    },
+    {
+      "epoch": 2.615650241124055,
+      "grad_norm": 0.9611344933509827,
+      "learning_rate": 6.4181948082876876e-06,
+      "loss": 0.7053,
+      "step": 32950
+    },
+    {
+      "epoch": 2.6164440651729546,
+      "grad_norm": 0.8691285252571106,
+      "learning_rate": 6.404964144902226e-06,
+      "loss": 0.7922,
+      "step": 32960
+    },
+    {
+      "epoch": 2.6172378892218537,
+      "grad_norm": 0.7891274690628052,
+      "learning_rate": 6.391733481516764e-06,
+      "loss": 0.7849,
+      "step": 32970
+    },
+    {
+      "epoch": 2.618031713270754,
+      "grad_norm": 0.9660747647285461,
+      "learning_rate": 6.378502818131302e-06,
+      "loss": 0.8229,
+      "step": 32980
+    },
+    {
+      "epoch": 2.618825537319653,
+      "grad_norm": 0.9171955585479736,
+      "learning_rate": 6.36527215474584e-06,
+      "loss": 0.8054,
+      "step": 32990
+    },
+    {
+      "epoch": 2.6196193613685526,
+      "grad_norm": 1.118066668510437,
+      "learning_rate": 6.352041491360378e-06,
+      "loss": 0.7319,
+      "step": 33000
+    },
+    {
+      "epoch": 2.620413185417452,
+      "grad_norm": 1.0057705640792847,
+      "learning_rate": 6.338810827974915e-06,
+      "loss": 0.7856,
+      "step": 33010
+    },
+    {
+      "epoch": 2.6212070094663518,
+      "grad_norm": 0.8103092908859253,
+      "learning_rate": 6.3255801645894524e-06,
+      "loss": 0.804,
+      "step": 33020
+    },
+    {
+      "epoch": 2.6220008335152514,
+      "grad_norm": 0.926285982131958,
+      "learning_rate": 6.3123495012039905e-06,
+      "loss": 0.8147,
+      "step": 33030
+    },
+    {
+      "epoch": 2.622794657564151,
+      "grad_norm": 0.8489839434623718,
+      "learning_rate": 6.299118837818528e-06,
+      "loss": 0.7151,
+      "step": 33040
+    },
+    {
+      "epoch": 2.6235884816130506,
+      "grad_norm": 0.8944807648658752,
+      "learning_rate": 6.285888174433066e-06,
+      "loss": 0.7859,
+      "step": 33050
+    },
+    {
+      "epoch": 2.62438230566195,
+      "grad_norm": 0.8567325472831726,
+      "learning_rate": 6.272657511047604e-06,
+      "loss": 0.8029,
+      "step": 33060
+    },
+    {
+      "epoch": 2.62517612971085,
+      "grad_norm": 0.9069671034812927,
+      "learning_rate": 6.259426847662142e-06,
+      "loss": 0.7756,
+      "step": 33070
+    },
+    {
+      "epoch": 2.625969953759749,
+      "grad_norm": 0.924453854560852,
+      "learning_rate": 6.24619618427668e-06,
+      "loss": 0.7929,
+      "step": 33080
+    },
+    {
+      "epoch": 2.6267637778086486,
+      "grad_norm": 0.9638664126396179,
+      "learning_rate": 6.232965520891218e-06,
+      "loss": 0.8279,
+      "step": 33090
+    },
+    {
+      "epoch": 2.627557601857548,
+      "grad_norm": 0.9644804000854492,
+      "learning_rate": 6.219734857505756e-06,
+      "loss": 0.7289,
+      "step": 33100
+    },
+    {
+      "epoch": 2.6283514259064478,
+      "grad_norm": 0.8256565928459167,
+      "learning_rate": 6.2065041941202935e-06,
+      "loss": 0.7667,
+      "step": 33110
+    },
+    {
+      "epoch": 2.6291452499553474,
+      "grad_norm": 0.9192927479743958,
+      "learning_rate": 6.1932735307348315e-06,
+      "loss": 0.7884,
+      "step": 33120
+    },
+    {
+      "epoch": 2.629939074004247,
+      "grad_norm": 0.8394224047660828,
+      "learning_rate": 6.180042867349369e-06,
+      "loss": 0.7822,
+      "step": 33130
+    },
+    {
+      "epoch": 2.6307328980531466,
+      "grad_norm": 0.9371026754379272,
+      "learning_rate": 6.166812203963907e-06,
+      "loss": 0.7951,
+      "step": 33140
+    },
+    {
+      "epoch": 2.631526722102046,
+      "grad_norm": 0.9811797738075256,
+      "learning_rate": 6.153581540578445e-06,
+      "loss": 0.7974,
+      "step": 33150
+    },
+    {
+      "epoch": 2.632320546150946,
+      "grad_norm": 1.0210554599761963,
+      "learning_rate": 6.140350877192982e-06,
+      "loss": 0.7797,
+      "step": 33160
+    },
+    {
+      "epoch": 2.633114370199845,
+      "grad_norm": 1.1501883268356323,
+      "learning_rate": 6.12712021380752e-06,
+      "loss": 0.8254,
+      "step": 33170
+    },
+    {
+      "epoch": 2.633908194248745,
+      "grad_norm": 1.004550576210022,
+      "learning_rate": 6.113889550422058e-06,
+      "loss": 0.7635,
+      "step": 33180
+    },
+    {
+      "epoch": 2.634702018297644,
+      "grad_norm": 0.8310564756393433,
+      "learning_rate": 6.100658887036596e-06,
+      "loss": 0.835,
+      "step": 33190
+    },
+    {
+      "epoch": 2.6354958423465438,
+      "grad_norm": 0.8766066431999207,
+      "learning_rate": 6.0874282236511345e-06,
+      "loss": 0.8109,
+      "step": 33200
+    },
+    {
+      "epoch": 2.6362896663954434,
+      "grad_norm": 1.1657209396362305,
+      "learning_rate": 6.0741975602656726e-06,
+      "loss": 0.7254,
+      "step": 33210
+    },
+    {
+      "epoch": 2.637083490444343,
+      "grad_norm": 0.9214816689491272,
+      "learning_rate": 6.062289963218756e-06,
+      "loss": 0.7934,
+      "step": 33220
+    },
+    {
+      "epoch": 2.6378773144932426,
+      "grad_norm": 0.9850200414657593,
+      "learning_rate": 6.049059299833293e-06,
+      "loss": 0.814,
+      "step": 33230
+    },
+    {
+      "epoch": 2.638671138542142,
+      "grad_norm": 0.7920827865600586,
+      "learning_rate": 6.035828636447831e-06,
+      "loss": 0.7924,
+      "step": 33240
+    },
+    {
+      "epoch": 2.639464962591042,
+      "grad_norm": 0.9591189026832581,
+      "learning_rate": 6.0225979730623695e-06,
+      "loss": 0.793,
+      "step": 33250
+    },
+    {
+      "epoch": 2.6402587866399414,
+      "grad_norm": 0.8499795198440552,
+      "learning_rate": 6.0093673096769076e-06,
+      "loss": 0.8147,
+      "step": 33260
+    },
+    {
+      "epoch": 2.641052610688841,
+      "grad_norm": 0.9025987982749939,
+      "learning_rate": 5.996136646291446e-06,
+      "loss": 0.7892,
+      "step": 33270
+    },
+    {
+      "epoch": 2.64184643473774,
+      "grad_norm": 0.8324596881866455,
+      "learning_rate": 5.982905982905984e-06,
+      "loss": 0.7845,
+      "step": 33280
+    },
+    {
+      "epoch": 2.64264025878664,
+      "grad_norm": 0.948970377445221,
+      "learning_rate": 5.969675319520521e-06,
+      "loss": 0.7481,
+      "step": 33290
+    },
+    {
+      "epoch": 2.6434340828355394,
+      "grad_norm": 0.7791500687599182,
+      "learning_rate": 5.956444656135059e-06,
+      "loss": 0.7557,
+      "step": 33300
+    },
+    {
+      "epoch": 2.644227906884439,
+      "grad_norm": 0.7366840243339539,
+      "learning_rate": 5.943213992749596e-06,
+      "loss": 0.8133,
+      "step": 33310
+    },
+    {
+      "epoch": 2.6450217309333386,
+      "grad_norm": 0.9585718512535095,
+      "learning_rate": 5.929983329364134e-06,
+      "loss": 0.7364,
+      "step": 33320
+    },
+    {
+      "epoch": 2.645815554982238,
+      "grad_norm": 0.8997151851654053,
+      "learning_rate": 5.916752665978672e-06,
+      "loss": 0.8047,
+      "step": 33330
+    },
+    {
+      "epoch": 2.646609379031138,
+      "grad_norm": 1.0085241794586182,
+      "learning_rate": 5.9035220025932105e-06,
+      "loss": 0.7328,
+      "step": 33340
+    },
+    {
+      "epoch": 2.6474032030800374,
+      "grad_norm": 1.000828742980957,
+      "learning_rate": 5.890291339207748e-06,
+      "loss": 0.756,
+      "step": 33350
+    },
+    {
+      "epoch": 2.648197027128937,
+      "grad_norm": 0.913764476776123,
+      "learning_rate": 5.877060675822286e-06,
+      "loss": 0.8004,
+      "step": 33360
+    },
+    {
+      "epoch": 2.648990851177836,
+      "grad_norm": 0.861119270324707,
+      "learning_rate": 5.863830012436824e-06,
+      "loss": 0.8222,
+      "step": 33370
+    },
+    {
+      "epoch": 2.649784675226736,
+      "grad_norm": 0.7925031185150146,
+      "learning_rate": 5.850599349051362e-06,
+      "loss": 0.7273,
+      "step": 33380
+    },
+    {
+      "epoch": 2.6505784992756354,
+      "grad_norm": 1.0246551036834717,
+      "learning_rate": 5.8373686856659e-06,
+      "loss": 0.8253,
+      "step": 33390
+    },
+    {
+      "epoch": 2.651372323324535,
+      "grad_norm": 0.970195472240448,
+      "learning_rate": 5.824138022280438e-06,
+      "loss": 0.7488,
+      "step": 33400
+    },
+    {
+      "epoch": 2.6521661473734346,
+      "grad_norm": 0.955636203289032,
+      "learning_rate": 5.810907358894975e-06,
+      "loss": 0.7855,
+      "step": 33410
+    },
+    {
+      "epoch": 2.652959971422334,
+      "grad_norm": 0.837679386138916,
+      "learning_rate": 5.797676695509513e-06,
+      "loss": 0.7337,
+      "step": 33420
+    },
+    {
+      "epoch": 2.653753795471234,
+      "grad_norm": 0.9340305328369141,
+      "learning_rate": 5.784446032124051e-06,
+      "loss": 0.7628,
+      "step": 33430
+    },
+    {
+      "epoch": 2.6545476195201334,
+      "grad_norm": 0.8416723608970642,
+      "learning_rate": 5.771215368738589e-06,
+      "loss": 0.818,
+      "step": 33440
+    },
+    {
+      "epoch": 2.655341443569033,
+      "grad_norm": 0.9004825353622437,
+      "learning_rate": 5.757984705353127e-06,
+      "loss": 0.7861,
+      "step": 33450
+    },
+    {
+      "epoch": 2.6561352676179326,
+      "grad_norm": 0.7813234925270081,
+      "learning_rate": 5.744754041967665e-06,
+      "loss": 0.804,
+      "step": 33460
+    },
+    {
+      "epoch": 2.656929091666832,
+      "grad_norm": 1.1214011907577515,
+      "learning_rate": 5.731523378582202e-06,
+      "loss": 0.7651,
+      "step": 33470
+    },
+    {
+      "epoch": 2.6577229157157314,
+      "grad_norm": 0.9582983255386353,
+      "learning_rate": 5.71829271519674e-06,
+      "loss": 0.7695,
+      "step": 33480
+    },
+    {
+      "epoch": 2.6585167397646314,
+      "grad_norm": 0.7920512557029724,
+      "learning_rate": 5.705062051811278e-06,
+      "loss": 0.8515,
+      "step": 33490
+    },
+    {
+      "epoch": 2.6593105638135306,
+      "grad_norm": 0.904657781124115,
+      "learning_rate": 5.691831388425816e-06,
+      "loss": 0.8063,
+      "step": 33500
+    },
+    {
+      "epoch": 2.66010438786243,
+      "grad_norm": 0.9910812377929688,
+      "learning_rate": 5.678600725040354e-06,
+      "loss": 0.7586,
+      "step": 33510
+    },
+    {
+      "epoch": 2.66089821191133,
+      "grad_norm": 1.2386099100112915,
+      "learning_rate": 5.665370061654892e-06,
+      "loss": 0.764,
+      "step": 33520
+    },
+    {
+      "epoch": 2.6616920359602294,
+      "grad_norm": 0.9596011638641357,
+      "learning_rate": 5.652139398269429e-06,
+      "loss": 0.7142,
+      "step": 33530
+    },
+    {
+      "epoch": 2.662485860009129,
+      "grad_norm": 1.1034601926803589,
+      "learning_rate": 5.638908734883967e-06,
+      "loss": 0.8422,
+      "step": 33540
+    },
+    {
+      "epoch": 2.6632796840580286,
+      "grad_norm": 0.8817473649978638,
+      "learning_rate": 5.625678071498505e-06,
+      "loss": 0.7953,
+      "step": 33550
+    },
+    {
+      "epoch": 2.6640735081069282,
+      "grad_norm": 0.9314897656440735,
+      "learning_rate": 5.612447408113043e-06,
+      "loss": 0.8269,
+      "step": 33560
+    },
+    {
+      "epoch": 2.6648673321558274,
+      "grad_norm": 0.8691668510437012,
+      "learning_rate": 5.599216744727581e-06,
+      "loss": 0.7674,
+      "step": 33570
+    },
+    {
+      "epoch": 2.6656611562047274,
+      "grad_norm": 0.869050145149231,
+      "learning_rate": 5.5859860813421185e-06,
+      "loss": 0.7608,
+      "step": 33580
+    },
+    {
+      "epoch": 2.6664549802536266,
+      "grad_norm": 0.7281513214111328,
+      "learning_rate": 5.5727554179566566e-06,
+      "loss": 0.7388,
+      "step": 33590
+    },
+    {
+      "epoch": 2.667248804302526,
+      "grad_norm": 0.9663460850715637,
+      "learning_rate": 5.559524754571195e-06,
+      "loss": 0.7931,
+      "step": 33600
+    },
+    {
+      "epoch": 2.668042628351426,
+      "grad_norm": 0.8808895945549011,
+      "learning_rate": 5.546294091185733e-06,
+      "loss": 0.7915,
+      "step": 33610
+    },
+    {
+      "epoch": 2.6688364524003254,
+      "grad_norm": 1.0292035341262817,
+      "learning_rate": 5.53306342780027e-06,
+      "loss": 0.7188,
+      "step": 33620
+    },
+    {
+      "epoch": 2.669630276449225,
+      "grad_norm": 0.8974072337150574,
+      "learning_rate": 5.519832764414808e-06,
+      "loss": 0.7602,
+      "step": 33630
+    },
+    {
+      "epoch": 2.6704241004981246,
+      "grad_norm": 0.8630101084709167,
+      "learning_rate": 5.506602101029345e-06,
+      "loss": 0.806,
+      "step": 33640
+    },
+    {
+      "epoch": 2.6712179245470242,
+      "grad_norm": 0.958050549030304,
+      "learning_rate": 5.493371437643883e-06,
+      "loss": 0.7824,
+      "step": 33650
+    },
+    {
+      "epoch": 2.672011748595924,
+      "grad_norm": 0.9174726605415344,
+      "learning_rate": 5.4801407742584214e-06,
+      "loss": 0.848,
+      "step": 33660
+    },
+    {
+      "epoch": 2.6728055726448234,
+      "grad_norm": 0.9651468992233276,
+      "learning_rate": 5.4669101108729595e-06,
+      "loss": 0.7625,
+      "step": 33670
+    },
+    {
+      "epoch": 2.6735993966937226,
+      "grad_norm": 0.8684685230255127,
+      "learning_rate": 5.453679447487498e-06,
+      "loss": 0.7831,
+      "step": 33680
+    },
+    {
+      "epoch": 2.6743932207426226,
+      "grad_norm": 0.9022814631462097,
+      "learning_rate": 5.440448784102036e-06,
+      "loss": 0.7718,
+      "step": 33690
+    },
+    {
+      "epoch": 2.675187044791522,
+      "grad_norm": 0.9168913960456848,
+      "learning_rate": 5.427218120716573e-06,
+      "loss": 0.87,
+      "step": 33700
+    },
+    {
+      "epoch": 2.6759808688404214,
+      "grad_norm": 1.065454125404358,
+      "learning_rate": 5.413987457331111e-06,
+      "loss": 0.7865,
+      "step": 33710
+    },
+    {
+      "epoch": 2.676774692889321,
+      "grad_norm": 1.0886614322662354,
+      "learning_rate": 5.400756793945649e-06,
+      "loss": 0.7529,
+      "step": 33720
+    },
+    {
+      "epoch": 2.6775685169382206,
+      "grad_norm": 1.0097647905349731,
+      "learning_rate": 5.387526130560186e-06,
+      "loss": 0.7951,
+      "step": 33730
+    },
+    {
+      "epoch": 2.6783623409871202,
+      "grad_norm": 0.7832012176513672,
+      "learning_rate": 5.374295467174724e-06,
+      "loss": 0.7762,
+      "step": 33740
+    },
+    {
+      "epoch": 2.67915616503602,
+      "grad_norm": 0.7780231833457947,
+      "learning_rate": 5.3610648037892625e-06,
+      "loss": 0.8242,
+      "step": 33750
+    },
+    {
+      "epoch": 2.6799499890849194,
+      "grad_norm": 0.9866840243339539,
+      "learning_rate": 5.3478341404038e-06,
+      "loss": 0.7879,
+      "step": 33760
+    },
+    {
+      "epoch": 2.680743813133819,
+      "grad_norm": 1.0353221893310547,
+      "learning_rate": 5.334603477018338e-06,
+      "loss": 0.7558,
+      "step": 33770
+    },
+    {
+      "epoch": 2.6815376371827186,
+      "grad_norm": 1.043462872505188,
+      "learning_rate": 5.321372813632876e-06,
+      "loss": 0.7572,
+      "step": 33780
+    },
+    {
+      "epoch": 2.682331461231618,
+      "grad_norm": 0.7620270848274231,
+      "learning_rate": 5.308142150247414e-06,
+      "loss": 0.8275,
+      "step": 33790
+    },
+    {
+      "epoch": 2.6831252852805174,
+      "grad_norm": 0.835926353931427,
+      "learning_rate": 5.294911486861952e-06,
+      "loss": 0.8125,
+      "step": 33800
+    },
+    {
+      "epoch": 2.683919109329417,
+      "grad_norm": 1.1094342470169067,
+      "learning_rate": 5.281680823476489e-06,
+      "loss": 0.8078,
+      "step": 33810
+    },
+    {
+      "epoch": 2.6847129333783166,
+      "grad_norm": 0.8570294380187988,
+      "learning_rate": 5.268450160091027e-06,
+      "loss": 0.7878,
+      "step": 33820
+    },
+    {
+      "epoch": 2.6855067574272162,
+      "grad_norm": 0.7822785377502441,
+      "learning_rate": 5.255219496705565e-06,
+      "loss": 0.7786,
+      "step": 33830
+    },
+    {
+      "epoch": 2.686300581476116,
+      "grad_norm": 1.0392347574234009,
+      "learning_rate": 5.241988833320103e-06,
+      "loss": 0.7568,
+      "step": 33840
+    },
+    {
+      "epoch": 2.6870944055250154,
+      "grad_norm": 0.9113541841506958,
+      "learning_rate": 5.228758169934641e-06,
+      "loss": 0.82,
+      "step": 33850
+    },
+    {
+      "epoch": 2.687888229573915,
+      "grad_norm": 0.928196370601654,
+      "learning_rate": 5.215527506549179e-06,
+      "loss": 0.7763,
+      "step": 33860
+    },
+    {
+      "epoch": 2.6886820536228146,
+      "grad_norm": 1.0967717170715332,
+      "learning_rate": 5.202296843163716e-06,
+      "loss": 0.7269,
+      "step": 33870
+    },
+    {
+      "epoch": 2.689475877671714,
+      "grad_norm": 1.093342900276184,
+      "learning_rate": 5.189066179778254e-06,
+      "loss": 0.8291,
+      "step": 33880
+    },
+    {
+      "epoch": 2.690269701720614,
+      "grad_norm": 1.088734745979309,
+      "learning_rate": 5.175835516392792e-06,
+      "loss": 0.7851,
+      "step": 33890
+    },
+    {
+      "epoch": 2.691063525769513,
+      "grad_norm": 0.8813207149505615,
+      "learning_rate": 5.16260485300733e-06,
+      "loss": 0.7603,
+      "step": 33900
+    },
+    {
+      "epoch": 2.6918573498184126,
+      "grad_norm": 0.9411490559577942,
+      "learning_rate": 5.149374189621868e-06,
+      "loss": 0.8174,
+      "step": 33910
+    },
+    {
+      "epoch": 2.6926511738673122,
+      "grad_norm": 0.6878808736801147,
+      "learning_rate": 5.1361435262364064e-06,
+      "loss": 0.7783,
+      "step": 33920
+    },
+    {
+      "epoch": 2.693444997916212,
+      "grad_norm": 1.0253195762634277,
+      "learning_rate": 5.122912862850944e-06,
+      "loss": 0.8235,
+      "step": 33930
+    },
+    {
+      "epoch": 2.6942388219651114,
+      "grad_norm": 1.0071438550949097,
+      "learning_rate": 5.109682199465481e-06,
+      "loss": 0.7548,
+      "step": 33940
+    },
+    {
+      "epoch": 2.695032646014011,
+      "grad_norm": 0.876988410949707,
+      "learning_rate": 5.096451536080019e-06,
+      "loss": 0.7692,
+      "step": 33950
+    },
+    {
+      "epoch": 2.6958264700629107,
+      "grad_norm": 0.8041113615036011,
+      "learning_rate": 5.083220872694557e-06,
+      "loss": 0.7844,
+      "step": 33960
+    },
+    {
+      "epoch": 2.6966202941118103,
+      "grad_norm": 0.8075045943260193,
+      "learning_rate": 5.069990209309095e-06,
+      "loss": 0.7855,
+      "step": 33970
+    },
+    {
+      "epoch": 2.69741411816071,
+      "grad_norm": 0.9511701464653015,
+      "learning_rate": 5.056759545923633e-06,
+      "loss": 0.7444,
+      "step": 33980
+    },
+    {
+      "epoch": 2.698207942209609,
+      "grad_norm": 0.9707065224647522,
+      "learning_rate": 5.0435288825381705e-06,
+      "loss": 0.757,
+      "step": 33990
+    },
+    {
+      "epoch": 2.699001766258509,
+      "grad_norm": 1.0104913711547852,
+      "learning_rate": 5.0302982191527085e-06,
+      "loss": 0.7089,
+      "step": 34000
+    },
+    {
+      "epoch": 2.6997955903074082,
+      "grad_norm": 0.9136759042739868,
+      "learning_rate": 5.017067555767247e-06,
+      "loss": 0.7715,
+      "step": 34010
+    },
+    {
+      "epoch": 2.700589414356308,
+      "grad_norm": 1.179523229598999,
+      "learning_rate": 5.003836892381785e-06,
+      "loss": 0.7927,
+      "step": 34020
+    },
+    {
+      "epoch": 2.7013832384052074,
+      "grad_norm": 0.7838528156280518,
+      "learning_rate": 4.990606228996322e-06,
+      "loss": 0.7484,
+      "step": 34030
+    },
+    {
+      "epoch": 2.702177062454107,
+      "grad_norm": 1.034165859222412,
+      "learning_rate": 4.97737556561086e-06,
+      "loss": 0.8256,
+      "step": 34040
+    },
+    {
+      "epoch": 2.7029708865030067,
+      "grad_norm": 1.1382567882537842,
+      "learning_rate": 4.964144902225397e-06,
+      "loss": 0.7398,
+      "step": 34050
+    },
+    {
+      "epoch": 2.7037647105519063,
+      "grad_norm": 0.9856430292129517,
+      "learning_rate": 4.950914238839935e-06,
+      "loss": 0.7579,
+      "step": 34060
+    },
+    {
+      "epoch": 2.704558534600806,
+      "grad_norm": 0.8700586557388306,
+      "learning_rate": 4.937683575454473e-06,
+      "loss": 0.7864,
+      "step": 34070
+    },
+    {
+      "epoch": 2.705352358649705,
+      "grad_norm": 0.9212909936904907,
+      "learning_rate": 4.9244529120690115e-06,
+      "loss": 0.7534,
+      "step": 34080
+    },
+    {
+      "epoch": 2.706146182698605,
+      "grad_norm": 0.9201529026031494,
+      "learning_rate": 4.912545315022096e-06,
+      "loss": 0.7437,
+      "step": 34090
+    },
+    {
+      "epoch": 2.7069400067475042,
+      "grad_norm": 0.9818313121795654,
+      "learning_rate": 4.899314651636634e-06,
+      "loss": 0.7595,
+      "step": 34100
+    },
+    {
+      "epoch": 2.707733830796404,
+      "grad_norm": 0.9335930347442627,
+      "learning_rate": 4.886083988251171e-06,
+      "loss": 0.7705,
+      "step": 34110
+    },
+    {
+      "epoch": 2.7085276548453034,
+      "grad_norm": 1.0011483430862427,
+      "learning_rate": 4.872853324865708e-06,
+      "loss": 0.7944,
+      "step": 34120
+    },
+    {
+      "epoch": 2.709321478894203,
+      "grad_norm": 0.9489617943763733,
+      "learning_rate": 4.8596226614802465e-06,
+      "loss": 0.7808,
+      "step": 34130
+    },
+    {
+      "epoch": 2.7101153029431027,
+      "grad_norm": 1.0148978233337402,
+      "learning_rate": 4.8463919980947845e-06,
+      "loss": 0.7909,
+      "step": 34140
+    },
+    {
+      "epoch": 2.7109091269920023,
+      "grad_norm": 0.89469313621521,
+      "learning_rate": 4.833161334709323e-06,
+      "loss": 0.8341,
+      "step": 34150
+    },
+    {
+      "epoch": 2.711702951040902,
+      "grad_norm": 1.135300874710083,
+      "learning_rate": 4.819930671323861e-06,
+      "loss": 0.785,
+      "step": 34160
+    },
+    {
+      "epoch": 2.7124967750898015,
+      "grad_norm": 0.9394294619560242,
+      "learning_rate": 4.806700007938398e-06,
+      "loss": 0.7499,
+      "step": 34170
+    },
+    {
+      "epoch": 2.713290599138701,
+      "grad_norm": 0.9445950984954834,
+      "learning_rate": 4.793469344552936e-06,
+      "loss": 0.7784,
+      "step": 34180
+    },
+    {
+      "epoch": 2.7140844231876002,
+      "grad_norm": 0.7616167664527893,
+      "learning_rate": 4.780238681167474e-06,
+      "loss": 0.8297,
+      "step": 34190
+    },
+    {
+      "epoch": 2.7148782472365003,
+      "grad_norm": 0.9969817399978638,
+      "learning_rate": 4.767008017782012e-06,
+      "loss": 0.771,
+      "step": 34200
+    },
+    {
+      "epoch": 2.7156720712853994,
+      "grad_norm": 0.9528368711471558,
+      "learning_rate": 4.75377735439655e-06,
+      "loss": 0.8011,
+      "step": 34210
+    },
+    {
+      "epoch": 2.716465895334299,
+      "grad_norm": 0.9359155893325806,
+      "learning_rate": 4.7405466910110875e-06,
+      "loss": 0.7494,
+      "step": 34220
+    },
+    {
+      "epoch": 2.7172597193831987,
+      "grad_norm": 0.9772827625274658,
+      "learning_rate": 4.727316027625625e-06,
+      "loss": 0.7775,
+      "step": 34230
+    },
+    {
+      "epoch": 2.7180535434320983,
+      "grad_norm": 0.8860666751861572,
+      "learning_rate": 4.714085364240163e-06,
+      "loss": 0.7459,
+      "step": 34240
+    },
+    {
+      "epoch": 2.718847367480998,
+      "grad_norm": 0.9986729621887207,
+      "learning_rate": 4.700854700854701e-06,
+      "loss": 0.7979,
+      "step": 34250
+    },
+    {
+      "epoch": 2.7196411915298975,
+      "grad_norm": 1.0801081657409668,
+      "learning_rate": 4.687624037469239e-06,
+      "loss": 0.765,
+      "step": 34260
+    },
+    {
+      "epoch": 2.720435015578797,
+      "grad_norm": 1.139674186706543,
+      "learning_rate": 4.674393374083777e-06,
+      "loss": 0.7262,
+      "step": 34270
+    },
+    {
+      "epoch": 2.7212288396276967,
+      "grad_norm": 1.0824925899505615,
+      "learning_rate": 4.661162710698315e-06,
+      "loss": 0.7435,
+      "step": 34280
+    },
+    {
+      "epoch": 2.7220226636765963,
+      "grad_norm": 1.0776777267456055,
+      "learning_rate": 4.647932047312852e-06,
+      "loss": 0.741,
+      "step": 34290
+    },
+    {
+      "epoch": 2.7228164877254954,
+      "grad_norm": 0.7067087292671204,
+      "learning_rate": 4.6347013839273904e-06,
+      "loss": 0.7718,
+      "step": 34300
+    },
+    {
+      "epoch": 2.723610311774395,
+      "grad_norm": 0.9206405282020569,
+      "learning_rate": 4.6214707205419285e-06,
+      "loss": 0.8099,
+      "step": 34310
+    },
+    {
+      "epoch": 2.7244041358232947,
+      "grad_norm": 0.9427090287208557,
+      "learning_rate": 4.608240057156466e-06,
+      "loss": 0.8307,
+      "step": 34320
+    },
+    {
+      "epoch": 2.7251979598721943,
+      "grad_norm": 0.9077566862106323,
+      "learning_rate": 4.595009393771004e-06,
+      "loss": 0.7284,
+      "step": 34330
+    },
+    {
+      "epoch": 2.725991783921094,
+      "grad_norm": 1.0393189191818237,
+      "learning_rate": 4.581778730385542e-06,
+      "loss": 0.76,
+      "step": 34340
+    },
+    {
+      "epoch": 2.7267856079699935,
+      "grad_norm": 0.7990016937255859,
+      "learning_rate": 4.568548067000079e-06,
+      "loss": 0.8172,
+      "step": 34350
+    },
+    {
+      "epoch": 2.727579432018893,
+      "grad_norm": 0.9199971556663513,
+      "learning_rate": 4.555317403614617e-06,
+      "loss": 0.8014,
+      "step": 34360
+    },
+    {
+      "epoch": 2.7283732560677927,
+      "grad_norm": 1.028395175933838,
+      "learning_rate": 4.542086740229155e-06,
+      "loss": 0.7645,
+      "step": 34370
+    },
+    {
+      "epoch": 2.7291670801166923,
+      "grad_norm": 0.9020273089408875,
+      "learning_rate": 4.528856076843693e-06,
+      "loss": 0.7711,
+      "step": 34380
+    },
+    {
+      "epoch": 2.7299609041655915,
+      "grad_norm": 1.0992895364761353,
+      "learning_rate": 4.5156254134582315e-06,
+      "loss": 0.7861,
+      "step": 34390
+    },
+    {
+      "epoch": 2.7307547282144915,
+      "grad_norm": 0.8821068406105042,
+      "learning_rate": 4.502394750072769e-06,
+      "loss": 0.7572,
+      "step": 34400
+    },
+    {
+      "epoch": 2.7315485522633907,
+      "grad_norm": 0.8414868712425232,
+      "learning_rate": 4.489164086687307e-06,
+      "loss": 0.8069,
+      "step": 34410
+    },
+    {
+      "epoch": 2.7323423763122903,
+      "grad_norm": 1.0444607734680176,
+      "learning_rate": 4.475933423301845e-06,
+      "loss": 0.8252,
+      "step": 34420
+    },
+    {
+      "epoch": 2.73313620036119,
+      "grad_norm": 0.7695857882499695,
+      "learning_rate": 4.462702759916382e-06,
+      "loss": 0.8368,
+      "step": 34430
+    },
+    {
+      "epoch": 2.7339300244100895,
+      "grad_norm": 0.9509304761886597,
+      "learning_rate": 4.44947209653092e-06,
+      "loss": 0.796,
+      "step": 34440
+    },
+    {
+      "epoch": 2.734723848458989,
+      "grad_norm": 0.798905074596405,
+      "learning_rate": 4.436241433145458e-06,
+      "loss": 0.7507,
+      "step": 34450
+    },
+    {
+      "epoch": 2.7355176725078887,
+      "grad_norm": 0.9972504377365112,
+      "learning_rate": 4.4230107697599955e-06,
+      "loss": 0.8479,
+      "step": 34460
+    },
+    {
+      "epoch": 2.7363114965567883,
+      "grad_norm": 0.9366284012794495,
+      "learning_rate": 4.4097801063745336e-06,
+      "loss": 0.7644,
+      "step": 34470
+    },
+    {
+      "epoch": 2.737105320605688,
+      "grad_norm": 1.0526267290115356,
+      "learning_rate": 4.396549442989072e-06,
+      "loss": 0.7578,
+      "step": 34480
+    },
+    {
+      "epoch": 2.7378991446545875,
+      "grad_norm": 0.8934128880500793,
+      "learning_rate": 4.38331877960361e-06,
+      "loss": 0.8356,
+      "step": 34490
+    },
+    {
+      "epoch": 2.7386929687034867,
+      "grad_norm": 1.0690593719482422,
+      "learning_rate": 4.370088116218148e-06,
+      "loss": 0.78,
+      "step": 34500
+    },
+    {
+      "epoch": 2.7394867927523863,
+      "grad_norm": 0.9687326550483704,
+      "learning_rate": 4.356857452832686e-06,
+      "loss": 0.7375,
+      "step": 34510
+    },
+    {
+      "epoch": 2.740280616801286,
+      "grad_norm": 0.9082489609718323,
+      "learning_rate": 4.343626789447223e-06,
+      "loss": 0.7743,
+      "step": 34520
+    },
+    {
+      "epoch": 2.7410744408501855,
+      "grad_norm": 0.951102614402771,
+      "learning_rate": 4.330396126061761e-06,
+      "loss": 0.797,
+      "step": 34530
+    },
+    {
+      "epoch": 2.741868264899085,
+      "grad_norm": 1.032860517501831,
+      "learning_rate": 4.3171654626762984e-06,
+      "loss": 0.8604,
+      "step": 34540
+    },
+    {
+      "epoch": 2.7426620889479847,
+      "grad_norm": 0.9105594158172607,
+      "learning_rate": 4.3039347992908365e-06,
+      "loss": 0.7811,
+      "step": 34550
+    },
+    {
+      "epoch": 2.7434559129968843,
+      "grad_norm": 0.9474354982376099,
+      "learning_rate": 4.290704135905375e-06,
+      "loss": 0.7872,
+      "step": 34560
+    },
+    {
+      "epoch": 2.744249737045784,
+      "grad_norm": 0.7172473669052124,
+      "learning_rate": 4.277473472519913e-06,
+      "loss": 0.7295,
+      "step": 34570
+    },
+    {
+      "epoch": 2.7450435610946835,
+      "grad_norm": 0.8522198796272278,
+      "learning_rate": 4.26424280913445e-06,
+      "loss": 0.7326,
+      "step": 34580
+    },
+    {
+      "epoch": 2.7458373851435827,
+      "grad_norm": 1.0782872438430786,
+      "learning_rate": 4.251012145748988e-06,
+      "loss": 0.7579,
+      "step": 34590
+    },
+    {
+      "epoch": 2.7466312091924827,
+      "grad_norm": 1.02603280544281,
+      "learning_rate": 4.237781482363526e-06,
+      "loss": 0.7964,
+      "step": 34600
+    },
+    {
+      "epoch": 2.747425033241382,
+      "grad_norm": 0.9840404391288757,
+      "learning_rate": 4.224550818978064e-06,
+      "loss": 0.8247,
+      "step": 34610
+    },
+    {
+      "epoch": 2.7482188572902815,
+      "grad_norm": 0.8115825653076172,
+      "learning_rate": 4.211320155592602e-06,
+      "loss": 0.7881,
+      "step": 34620
+    },
+    {
+      "epoch": 2.749012681339181,
+      "grad_norm": 1.053634524345398,
+      "learning_rate": 4.1980894922071395e-06,
+      "loss": 0.7765,
+      "step": 34630
+    },
+    {
+      "epoch": 2.7498065053880807,
+      "grad_norm": 0.8972168564796448,
+      "learning_rate": 4.184858828821677e-06,
+      "loss": 0.7956,
+      "step": 34640
+    },
+    {
+      "epoch": 2.7506003294369803,
+      "grad_norm": 0.9098021388053894,
+      "learning_rate": 4.171628165436215e-06,
+      "loss": 0.7835,
+      "step": 34650
+    },
+    {
+      "epoch": 2.75139415348588,
+      "grad_norm": 0.8237542510032654,
+      "learning_rate": 4.158397502050753e-06,
+      "loss": 0.7629,
+      "step": 34660
+    },
+    {
+      "epoch": 2.7521879775347795,
+      "grad_norm": 1.1170654296875,
+      "learning_rate": 4.145166838665291e-06,
+      "loss": 0.7887,
+      "step": 34670
+    },
+    {
+      "epoch": 2.752981801583679,
+      "grad_norm": 1.165892243385315,
+      "learning_rate": 4.131936175279829e-06,
+      "loss": 0.7624,
+      "step": 34680
+    },
+    {
+      "epoch": 2.7537756256325787,
+      "grad_norm": 0.8696096539497375,
+      "learning_rate": 4.118705511894366e-06,
+      "loss": 0.7976,
+      "step": 34690
+    },
+    {
+      "epoch": 2.754569449681478,
+      "grad_norm": 1.0606629848480225,
+      "learning_rate": 4.105474848508904e-06,
+      "loss": 0.7863,
+      "step": 34700
+    },
+    {
+      "epoch": 2.755363273730378,
+      "grad_norm": 0.9789169430732727,
+      "learning_rate": 4.092244185123442e-06,
+      "loss": 0.7347,
+      "step": 34710
+    },
+    {
+      "epoch": 2.756157097779277,
+      "grad_norm": 0.9391206502914429,
+      "learning_rate": 4.0790135217379805e-06,
+      "loss": 0.8182,
+      "step": 34720
+    },
+    {
+      "epoch": 2.7569509218281767,
+      "grad_norm": 0.9895641803741455,
+      "learning_rate": 4.0657828583525186e-06,
+      "loss": 0.7514,
+      "step": 34730
+    },
+    {
+      "epoch": 2.7577447458770763,
+      "grad_norm": 0.8437541723251343,
+      "learning_rate": 4.052552194967056e-06,
+      "loss": 0.7866,
+      "step": 34740
+    },
+    {
+      "epoch": 2.758538569925976,
+      "grad_norm": 1.0394833087921143,
+      "learning_rate": 4.039321531581593e-06,
+      "loss": 0.8191,
+      "step": 34750
+    },
+    {
+      "epoch": 2.7593323939748755,
+      "grad_norm": 0.9118010997772217,
+      "learning_rate": 4.026090868196131e-06,
+      "loss": 0.7347,
+      "step": 34760
+    },
+    {
+      "epoch": 2.760126218023775,
+      "grad_norm": 0.8801964521408081,
+      "learning_rate": 4.012860204810669e-06,
+      "loss": 0.7648,
+      "step": 34770
+    },
+    {
+      "epoch": 2.7609200420726747,
+      "grad_norm": 0.9062778353691101,
+      "learning_rate": 3.999629541425207e-06,
+      "loss": 0.7519,
+      "step": 34780
+    },
+    {
+      "epoch": 2.761713866121574,
+      "grad_norm": 0.9668712615966797,
+      "learning_rate": 3.986398878039745e-06,
+      "loss": 0.7754,
+      "step": 34790
+    },
+    {
+      "epoch": 2.762507690170474,
+      "grad_norm": 0.7536428570747375,
+      "learning_rate": 3.9731682146542834e-06,
+      "loss": 0.8258,
+      "step": 34800
+    },
+    {
+      "epoch": 2.763301514219373,
+      "grad_norm": 0.9052528142929077,
+      "learning_rate": 3.959937551268821e-06,
+      "loss": 0.829,
+      "step": 34810
+    },
+    {
+      "epoch": 2.7640953382682727,
+      "grad_norm": 0.9865114688873291,
+      "learning_rate": 3.946706887883359e-06,
+      "loss": 0.8256,
+      "step": 34820
+    },
+    {
+      "epoch": 2.7648891623171723,
+      "grad_norm": 0.8684812188148499,
+      "learning_rate": 3.933476224497897e-06,
+      "loss": 0.7742,
+      "step": 34830
+    },
+    {
+      "epoch": 2.765682986366072,
+      "grad_norm": 1.075271725654602,
+      "learning_rate": 3.920245561112434e-06,
+      "loss": 0.8008,
+      "step": 34840
+    },
+    {
+      "epoch": 2.7664768104149715,
+      "grad_norm": 0.9537501335144043,
+      "learning_rate": 3.907014897726972e-06,
+      "loss": 0.8286,
+      "step": 34850
+    },
+    {
+      "epoch": 2.767270634463871,
+      "grad_norm": 0.8851303458213806,
+      "learning_rate": 3.89378423434151e-06,
+      "loss": 0.815,
+      "step": 34860
+    },
+    {
+      "epoch": 2.7680644585127707,
+      "grad_norm": 0.8173339366912842,
+      "learning_rate": 3.8805535709560474e-06,
+      "loss": 0.8254,
+      "step": 34870
+    },
+    {
+      "epoch": 2.7688582825616703,
+      "grad_norm": 1.0855127573013306,
+      "learning_rate": 3.8673229075705855e-06,
+      "loss": 0.7537,
+      "step": 34880
+    },
+    {
+      "epoch": 2.76965210661057,
+      "grad_norm": 1.0475999116897583,
+      "learning_rate": 3.854092244185124e-06,
+      "loss": 0.7576,
+      "step": 34890
+    },
+    {
+      "epoch": 2.770445930659469,
+      "grad_norm": 0.9665488004684448,
+      "learning_rate": 3.840861580799662e-06,
+      "loss": 0.728,
+      "step": 34900
+    },
+    {
+      "epoch": 2.771239754708369,
+      "grad_norm": 0.8024313449859619,
+      "learning_rate": 3.8276309174142e-06,
+      "loss": 0.7808,
+      "step": 34910
+    },
+    {
+      "epoch": 2.7720335787572683,
+      "grad_norm": 0.955835223197937,
+      "learning_rate": 3.814400254028737e-06,
+      "loss": 0.7961,
+      "step": 34920
+    },
+    {
+      "epoch": 2.772827402806168,
+      "grad_norm": 0.996760368347168,
+      "learning_rate": 3.8011695906432747e-06,
+      "loss": 0.7857,
+      "step": 34930
+    },
+    {
+      "epoch": 2.7736212268550675,
+      "grad_norm": 0.9988394975662231,
+      "learning_rate": 3.7879389272578127e-06,
+      "loss": 0.7821,
+      "step": 34940
+    },
+    {
+      "epoch": 2.774415050903967,
+      "grad_norm": 0.9682359099388123,
+      "learning_rate": 3.774708263872351e-06,
+      "loss": 0.8208,
+      "step": 34950
+    },
+    {
+      "epoch": 2.7752088749528667,
+      "grad_norm": 0.8783732652664185,
+      "learning_rate": 3.761477600486889e-06,
+      "loss": 0.8191,
+      "step": 34960
+    },
+    {
+      "epoch": 2.7760026990017663,
+      "grad_norm": 1.049475073814392,
+      "learning_rate": 3.7482469371014266e-06,
+      "loss": 0.733,
+      "step": 34970
+    },
+    {
+      "epoch": 2.776796523050666,
+      "grad_norm": 0.9063088893890381,
+      "learning_rate": 3.7350162737159638e-06,
+      "loss": 0.8224,
+      "step": 34980
+    },
+    {
+      "epoch": 2.7775903470995655,
+      "grad_norm": 0.9383750557899475,
+      "learning_rate": 3.721785610330502e-06,
+      "loss": 0.7338,
+      "step": 34990
+    },
+    {
+      "epoch": 2.778384171148465,
+      "grad_norm": 0.849460244178772,
+      "learning_rate": 3.70855494694504e-06,
+      "loss": 0.7359,
+      "step": 35000
+    },
+    {
+      "epoch": 2.7791779951973643,
+      "grad_norm": 1.0074882507324219,
+      "learning_rate": 3.695324283559578e-06,
+      "loss": 0.7969,
+      "step": 35010
+    },
+    {
+      "epoch": 2.779971819246264,
+      "grad_norm": 0.9304841160774231,
+      "learning_rate": 3.6820936201741157e-06,
+      "loss": 0.7987,
+      "step": 35020
+    },
+    {
+      "epoch": 2.7807656432951635,
+      "grad_norm": 1.1180306673049927,
+      "learning_rate": 3.6688629567886538e-06,
+      "loss": 0.7637,
+      "step": 35030
+    },
+    {
+      "epoch": 2.781559467344063,
+      "grad_norm": 0.9960020184516907,
+      "learning_rate": 3.655632293403191e-06,
+      "loss": 0.7664,
+      "step": 35040
+    },
+    {
+      "epoch": 2.7823532913929627,
+      "grad_norm": 1.1792720556259155,
+      "learning_rate": 3.642401630017729e-06,
+      "loss": 0.7887,
+      "step": 35050
+    },
+    {
+      "epoch": 2.7831471154418623,
+      "grad_norm": 0.7898070216178894,
+      "learning_rate": 3.629170966632267e-06,
+      "loss": 0.8356,
+      "step": 35060
+    },
+    {
+      "epoch": 2.783940939490762,
+      "grad_norm": 0.7194439172744751,
+      "learning_rate": 3.6159403032468052e-06,
+      "loss": 0.7888,
+      "step": 35070
+    },
+    {
+      "epoch": 2.7847347635396615,
+      "grad_norm": 1.077726125717163,
+      "learning_rate": 3.602709639861343e-06,
+      "loss": 0.8264,
+      "step": 35080
+    },
+    {
+      "epoch": 2.785528587588561,
+      "grad_norm": 1.00077486038208,
+      "learning_rate": 3.589478976475881e-06,
+      "loss": 0.7843,
+      "step": 35090
+    },
+    {
+      "epoch": 2.7863224116374603,
+      "grad_norm": 1.037901520729065,
+      "learning_rate": 3.576248313090418e-06,
+      "loss": 0.7137,
+      "step": 35100
+    },
+    {
+      "epoch": 2.7871162356863604,
+      "grad_norm": 0.8961439728736877,
+      "learning_rate": 3.5630176497049563e-06,
+      "loss": 0.7605,
+      "step": 35110
+    },
+    {
+      "epoch": 2.7879100597352595,
+      "grad_norm": 0.9663600325584412,
+      "learning_rate": 3.5497869863194944e-06,
+      "loss": 0.732,
+      "step": 35120
+    },
+    {
+      "epoch": 2.788703883784159,
+      "grad_norm": 0.8328473567962646,
+      "learning_rate": 3.536556322934032e-06,
+      "loss": 0.7589,
+      "step": 35130
+    },
+    {
+      "epoch": 2.7894977078330587,
+      "grad_norm": 1.0931073427200317,
+      "learning_rate": 3.52332565954857e-06,
+      "loss": 0.8035,
+      "step": 35140
+    },
+    {
+      "epoch": 2.7902915318819583,
+      "grad_norm": 0.8528993725776672,
+      "learning_rate": 3.5100949961631073e-06,
+      "loss": 0.818,
+      "step": 35150
+    },
+    {
+      "epoch": 2.791085355930858,
+      "grad_norm": 0.9126871228218079,
+      "learning_rate": 3.4968643327776454e-06,
+      "loss": 0.7092,
+      "step": 35160
+    },
+    {
+      "epoch": 2.7918791799797575,
+      "grad_norm": 0.8456578850746155,
+      "learning_rate": 3.4836336693921835e-06,
+      "loss": 0.7571,
+      "step": 35170
+    },
+    {
+      "epoch": 2.792673004028657,
+      "grad_norm": 1.0316028594970703,
+      "learning_rate": 3.470403006006721e-06,
+      "loss": 0.7413,
+      "step": 35180
+    },
+    {
+      "epoch": 2.7934668280775568,
+      "grad_norm": 0.8580575585365295,
+      "learning_rate": 3.4571723426212592e-06,
+      "loss": 0.7685,
+      "step": 35190
+    },
+    {
+      "epoch": 2.7942606521264564,
+      "grad_norm": 0.9380501508712769,
+      "learning_rate": 3.4439416792357973e-06,
+      "loss": 0.7756,
+      "step": 35200
+    },
+    {
+      "epoch": 2.7950544761753555,
+      "grad_norm": 0.8025434613227844,
+      "learning_rate": 3.4307110158503345e-06,
+      "loss": 0.7646,
+      "step": 35210
+    },
+    {
+      "epoch": 2.7958483002242556,
+      "grad_norm": 0.828920841217041,
+      "learning_rate": 3.4174803524648726e-06,
+      "loss": 0.8187,
+      "step": 35220
+    },
+    {
+      "epoch": 2.7966421242731547,
+      "grad_norm": 0.8058600425720215,
+      "learning_rate": 3.4042496890794107e-06,
+      "loss": 0.7068,
+      "step": 35230
+    },
+    {
+      "epoch": 2.7974359483220543,
+      "grad_norm": 0.9694280028343201,
+      "learning_rate": 3.3910190256939484e-06,
+      "loss": 0.7607,
+      "step": 35240
+    },
+    {
+      "epoch": 2.798229772370954,
+      "grad_norm": 0.9409134984016418,
+      "learning_rate": 3.3777883623084864e-06,
+      "loss": 0.7601,
+      "step": 35250
+    },
+    {
+      "epoch": 2.7990235964198535,
+      "grad_norm": 1.1917532682418823,
+      "learning_rate": 3.3645576989230245e-06,
+      "loss": 0.7952,
+      "step": 35260
+    },
+    {
+      "epoch": 2.799817420468753,
+      "grad_norm": 0.8570424318313599,
+      "learning_rate": 3.3513270355375618e-06,
+      "loss": 0.8205,
+      "step": 35270
+    },
+    {
+      "epoch": 2.8006112445176528,
+      "grad_norm": 1.0113235712051392,
+      "learning_rate": 3.3380963721521e-06,
+      "loss": 0.8328,
+      "step": 35280
+    },
+    {
+      "epoch": 2.8014050685665524,
+      "grad_norm": 0.9469982385635376,
+      "learning_rate": 3.3248657087666375e-06,
+      "loss": 0.7525,
+      "step": 35290
+    },
+    {
+      "epoch": 2.8021988926154515,
+      "grad_norm": 1.03362238407135,
+      "learning_rate": 3.3116350453811756e-06,
+      "loss": 0.7338,
+      "step": 35300
+    },
+    {
+      "epoch": 2.8029927166643516,
+      "grad_norm": 1.1091042757034302,
+      "learning_rate": 3.2984043819957137e-06,
+      "loss": 0.7698,
+      "step": 35310
+    },
+    {
+      "epoch": 2.8037865407132507,
+      "grad_norm": 1.1740446090698242,
+      "learning_rate": 3.2851737186102517e-06,
+      "loss": 0.7686,
+      "step": 35320
+    },
+    {
+      "epoch": 2.8045803647621503,
+      "grad_norm": 1.0045807361602783,
+      "learning_rate": 3.271943055224789e-06,
+      "loss": 0.7481,
+      "step": 35330
+    },
+    {
+      "epoch": 2.80537418881105,
+      "grad_norm": 0.9175392985343933,
+      "learning_rate": 3.258712391839327e-06,
+      "loss": 0.7549,
+      "step": 35340
+    },
+    {
+      "epoch": 2.8061680128599495,
+      "grad_norm": 1.221364974975586,
+      "learning_rate": 3.2454817284538647e-06,
+      "loss": 0.7778,
+      "step": 35350
+    },
+    {
+      "epoch": 2.806961836908849,
+      "grad_norm": 0.9719924330711365,
+      "learning_rate": 3.2322510650684028e-06,
+      "loss": 0.7635,
+      "step": 35360
+    },
+    {
+      "epoch": 2.8077556609577488,
+      "grad_norm": 0.9859107732772827,
+      "learning_rate": 3.219020401682941e-06,
+      "loss": 0.8012,
+      "step": 35370
+    },
+    {
+      "epoch": 2.8085494850066484,
+      "grad_norm": 1.052498698234558,
+      "learning_rate": 3.2057897382974785e-06,
+      "loss": 0.8148,
+      "step": 35380
+    },
+    {
+      "epoch": 2.809343309055548,
+      "grad_norm": 0.7815044522285461,
+      "learning_rate": 3.192559074912016e-06,
+      "loss": 0.7499,
+      "step": 35390
+    },
+    {
+      "epoch": 2.8101371331044476,
+      "grad_norm": 1.2309107780456543,
+      "learning_rate": 3.179328411526554e-06,
+      "loss": 0.8063,
+      "step": 35400
+    },
+    {
+      "epoch": 2.8109309571533467,
+      "grad_norm": 1.1573877334594727,
+      "learning_rate": 3.166097748141092e-06,
+      "loss": 0.7626,
+      "step": 35410
+    },
+    {
+      "epoch": 2.811724781202247,
+      "grad_norm": 1.0697295665740967,
+      "learning_rate": 3.15286708475563e-06,
+      "loss": 0.7613,
+      "step": 35420
+    },
+    {
+      "epoch": 2.812518605251146,
+      "grad_norm": 0.9957389235496521,
+      "learning_rate": 3.139636421370168e-06,
+      "loss": 0.7382,
+      "step": 35430
+    },
+    {
+      "epoch": 2.8133124293000455,
+      "grad_norm": 0.9990561604499817,
+      "learning_rate": 3.1264057579847053e-06,
+      "loss": 0.7597,
+      "step": 35440
+    },
+    {
+      "epoch": 2.814106253348945,
+      "grad_norm": 1.1225290298461914,
+      "learning_rate": 3.1131750945992434e-06,
+      "loss": 0.817,
+      "step": 35450
+    },
+    {
+      "epoch": 2.8149000773978448,
+      "grad_norm": 1.0921093225479126,
+      "learning_rate": 3.099944431213781e-06,
+      "loss": 0.7977,
+      "step": 35460
+    },
+    {
+      "epoch": 2.8156939014467444,
+      "grad_norm": 0.9584634304046631,
+      "learning_rate": 3.086713767828319e-06,
+      "loss": 0.7256,
+      "step": 35470
+    },
+    {
+      "epoch": 2.816487725495644,
+      "grad_norm": 0.8979785442352295,
+      "learning_rate": 3.073483104442857e-06,
+      "loss": 0.8171,
+      "step": 35480
+    },
+    {
+      "epoch": 2.8172815495445436,
+      "grad_norm": 0.9863408803939819,
+      "learning_rate": 3.060252441057395e-06,
+      "loss": 0.8187,
+      "step": 35490
+    },
+    {
+      "epoch": 2.818075373593443,
+      "grad_norm": 0.9419934153556824,
+      "learning_rate": 3.0470217776719325e-06,
+      "loss": 0.7236,
+      "step": 35500
+    },
+    {
+      "epoch": 2.818869197642343,
+      "grad_norm": 0.7588933706283569,
+      "learning_rate": 3.0337911142864706e-06,
+      "loss": 0.7996,
+      "step": 35510
+    },
+    {
+      "epoch": 2.819663021691242,
+      "grad_norm": 0.9996737837791443,
+      "learning_rate": 3.0205604509010083e-06,
+      "loss": 0.7806,
+      "step": 35520
+    },
+    {
+      "epoch": 2.8204568457401415,
+      "grad_norm": 1.1076947450637817,
+      "learning_rate": 3.0073297875155463e-06,
+      "loss": 0.7945,
+      "step": 35530
+    },
+    {
+      "epoch": 2.821250669789041,
+      "grad_norm": 0.942015528678894,
+      "learning_rate": 2.9940991241300844e-06,
+      "loss": 0.7777,
+      "step": 35540
+    },
+    {
+      "epoch": 2.8220444938379408,
+      "grad_norm": 0.9087643027305603,
+      "learning_rate": 2.9808684607446216e-06,
+      "loss": 0.8318,
+      "step": 35550
+    },
+    {
+      "epoch": 2.8228383178868404,
+      "grad_norm": 1.026443600654602,
+      "learning_rate": 2.9676377973591597e-06,
+      "loss": 0.7698,
+      "step": 35560
+    },
+    {
+      "epoch": 2.82363214193574,
+      "grad_norm": 0.9382554888725281,
+      "learning_rate": 2.9544071339736974e-06,
+      "loss": 0.7869,
+      "step": 35570
+    },
+    {
+      "epoch": 2.8244259659846396,
+      "grad_norm": 0.7953215837478638,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.7498,
+      "step": 35580
+    },
+    {
+      "epoch": 2.825219790033539,
+      "grad_norm": 0.9464923143386841,
+      "learning_rate": 2.9279458072027735e-06,
+      "loss": 0.7229,
+      "step": 35590
+    },
+    {
+      "epoch": 2.826013614082439,
+      "grad_norm": 0.9360369443893433,
+      "learning_rate": 2.914715143817311e-06,
+      "loss": 0.7409,
+      "step": 35600
+    },
+    {
+      "epoch": 2.826807438131338,
+      "grad_norm": 0.9968975186347961,
+      "learning_rate": 2.901484480431849e-06,
+      "loss": 0.8111,
+      "step": 35610
+    },
+    {
+      "epoch": 2.827601262180238,
+      "grad_norm": 0.7860965132713318,
+      "learning_rate": 2.888253817046387e-06,
+      "loss": 0.8074,
+      "step": 35620
+    },
+    {
+      "epoch": 2.828395086229137,
+      "grad_norm": 0.8920134902000427,
+      "learning_rate": 2.8750231536609246e-06,
+      "loss": 0.7975,
+      "step": 35630
+    },
+    {
+      "epoch": 2.8291889102780368,
+      "grad_norm": 0.802579939365387,
+      "learning_rate": 2.8617924902754627e-06,
+      "loss": 0.7383,
+      "step": 35640
+    },
+    {
+      "epoch": 2.8299827343269364,
+      "grad_norm": 0.790934681892395,
+      "learning_rate": 2.8485618268900003e-06,
+      "loss": 0.7811,
+      "step": 35650
+    },
+    {
+      "epoch": 2.830776558375836,
+      "grad_norm": 0.9820284247398376,
+      "learning_rate": 2.835331163504538e-06,
+      "loss": 0.7466,
+      "step": 35660
+    },
+    {
+      "epoch": 2.8315703824247356,
+      "grad_norm": 1.035416841506958,
+      "learning_rate": 2.822100500119076e-06,
+      "loss": 0.7927,
+      "step": 35670
+    },
+    {
+      "epoch": 2.832364206473635,
+      "grad_norm": 1.0442098379135132,
+      "learning_rate": 2.808869836733614e-06,
+      "loss": 0.7825,
+      "step": 35680
+    },
+    {
+      "epoch": 2.833158030522535,
+      "grad_norm": 1.0599391460418701,
+      "learning_rate": 2.795639173348152e-06,
+      "loss": 0.8229,
+      "step": 35690
+    },
+    {
+      "epoch": 2.8339518545714344,
+      "grad_norm": 1.2767279148101807,
+      "learning_rate": 2.78240850996269e-06,
+      "loss": 0.7614,
+      "step": 35700
+    },
+    {
+      "epoch": 2.834745678620334,
+      "grad_norm": 1.054215431213379,
+      "learning_rate": 2.7691778465772275e-06,
+      "loss": 0.799,
+      "step": 35710
+    },
+    {
+      "epoch": 2.835539502669233,
+      "grad_norm": 0.8956333994865417,
+      "learning_rate": 2.755947183191765e-06,
+      "loss": 0.7311,
+      "step": 35720
+    },
+    {
+      "epoch": 2.8363333267181328,
+      "grad_norm": 1.0596696138381958,
+      "learning_rate": 2.7427165198063033e-06,
+      "loss": 0.7634,
+      "step": 35730
+    },
+    {
+      "epoch": 2.8371271507670324,
+      "grad_norm": 1.0594090223312378,
+      "learning_rate": 2.7294858564208414e-06,
+      "loss": 0.8017,
+      "step": 35740
+    },
+    {
+      "epoch": 2.837920974815932,
+      "grad_norm": 1.278930902481079,
+      "learning_rate": 2.716255193035379e-06,
+      "loss": 0.7555,
+      "step": 35750
+    },
+    {
+      "epoch": 2.8387147988648316,
+      "grad_norm": 1.1408640146255493,
+      "learning_rate": 2.7030245296499167e-06,
+      "loss": 0.7985,
+      "step": 35760
+    },
+    {
+      "epoch": 2.839508622913731,
+      "grad_norm": 0.8677951693534851,
+      "learning_rate": 2.6897938662644547e-06,
+      "loss": 0.7973,
+      "step": 35770
+    },
+    {
+      "epoch": 2.840302446962631,
+      "grad_norm": 0.9274892807006836,
+      "learning_rate": 2.6765632028789924e-06,
+      "loss": 0.81,
+      "step": 35780
+    },
+    {
+      "epoch": 2.8410962710115304,
+      "grad_norm": 0.9693270921707153,
+      "learning_rate": 2.6633325394935305e-06,
+      "loss": 0.7442,
+      "step": 35790
+    },
+    {
+      "epoch": 2.84189009506043,
+      "grad_norm": 0.9193242192268372,
+      "learning_rate": 2.6501018761080686e-06,
+      "loss": 0.7411,
+      "step": 35800
+    },
+    {
+      "epoch": 2.842683919109329,
+      "grad_norm": 0.9392445087432861,
+      "learning_rate": 2.636871212722606e-06,
+      "loss": 0.7334,
+      "step": 35810
+    },
+    {
+      "epoch": 2.843477743158229,
+      "grad_norm": 0.9353786706924438,
+      "learning_rate": 2.623640549337144e-06,
+      "loss": 0.8122,
+      "step": 35820
+    },
+    {
+      "epoch": 2.8442715672071284,
+      "grad_norm": 0.7990022301673889,
+      "learning_rate": 2.6104098859516815e-06,
+      "loss": 0.7969,
+      "step": 35830
+    },
+    {
+      "epoch": 2.845065391256028,
+      "grad_norm": 0.7755838632583618,
+      "learning_rate": 2.5971792225662196e-06,
+      "loss": 0.8085,
+      "step": 35840
+    },
+    {
+      "epoch": 2.8458592153049276,
+      "grad_norm": 0.9300201535224915,
+      "learning_rate": 2.5839485591807577e-06,
+      "loss": 0.8288,
+      "step": 35850
+    },
+    {
+      "epoch": 2.846653039353827,
+      "grad_norm": 1.0524238348007202,
+      "learning_rate": 2.5707178957952953e-06,
+      "loss": 0.7261,
+      "step": 35860
+    },
+    {
+      "epoch": 2.847446863402727,
+      "grad_norm": 1.0584592819213867,
+      "learning_rate": 2.557487232409833e-06,
+      "loss": 0.779,
+      "step": 35870
+    },
+    {
+      "epoch": 2.8482406874516264,
+      "grad_norm": 1.093838095664978,
+      "learning_rate": 2.544256569024371e-06,
+      "loss": 0.8021,
+      "step": 35880
+    },
+    {
+      "epoch": 2.849034511500526,
+      "grad_norm": 0.6603565216064453,
+      "learning_rate": 2.5310259056389087e-06,
+      "loss": 0.8322,
+      "step": 35890
+    },
+    {
+      "epoch": 2.8498283355494256,
+      "grad_norm": 0.9231084585189819,
+      "learning_rate": 2.517795242253447e-06,
+      "loss": 0.7669,
+      "step": 35900
+    },
+    {
+      "epoch": 2.850622159598325,
+      "grad_norm": 0.9503127336502075,
+      "learning_rate": 2.5045645788679845e-06,
+      "loss": 0.7969,
+      "step": 35910
+    },
+    {
+      "epoch": 2.8514159836472244,
+      "grad_norm": 0.9831972122192383,
+      "learning_rate": 2.491333915482522e-06,
+      "loss": 0.7686,
+      "step": 35920
+    },
+    {
+      "epoch": 2.8522098076961244,
+      "grad_norm": 0.9379409551620483,
+      "learning_rate": 2.4781032520970602e-06,
+      "loss": 0.7717,
+      "step": 35930
+    },
+    {
+      "epoch": 2.8530036317450236,
+      "grad_norm": 0.9157143235206604,
+      "learning_rate": 2.4648725887115983e-06,
+      "loss": 0.7746,
+      "step": 35940
+    },
+    {
+      "epoch": 2.853797455793923,
+      "grad_norm": 0.896962583065033,
+      "learning_rate": 2.451641925326136e-06,
+      "loss": 0.7007,
+      "step": 35950
+    },
+    {
+      "epoch": 2.854591279842823,
+      "grad_norm": 1.0108494758605957,
+      "learning_rate": 2.438411261940674e-06,
+      "loss": 0.7847,
+      "step": 35960
+    },
+    {
+      "epoch": 2.8553851038917224,
+      "grad_norm": 1.0298715829849243,
+      "learning_rate": 2.4251805985552117e-06,
+      "loss": 0.7752,
+      "step": 35970
+    },
+    {
+      "epoch": 2.856178927940622,
+      "grad_norm": 1.0451425313949585,
+      "learning_rate": 2.4119499351697493e-06,
+      "loss": 0.7909,
+      "step": 35980
+    },
+    {
+      "epoch": 2.8569727519895216,
+      "grad_norm": 0.9338813424110413,
+      "learning_rate": 2.3987192717842874e-06,
+      "loss": 0.7663,
+      "step": 35990
+    },
+    {
+      "epoch": 2.857766576038421,
+      "grad_norm": 0.905019998550415,
+      "learning_rate": 2.3854886083988255e-06,
+      "loss": 0.746,
+      "step": 36000
+    },
+    {
+      "epoch": 2.8585604000873204,
+      "grad_norm": 0.8356999754905701,
+      "learning_rate": 2.372257945013363e-06,
+      "loss": 0.7736,
+      "step": 36010
+    },
+    {
+      "epoch": 2.8593542241362204,
+      "grad_norm": 1.016014575958252,
+      "learning_rate": 2.359027281627901e-06,
+      "loss": 0.8181,
+      "step": 36020
+    },
+    {
+      "epoch": 2.8601480481851196,
+      "grad_norm": 0.9665111303329468,
+      "learning_rate": 2.345796618242439e-06,
+      "loss": 0.7668,
+      "step": 36030
+    },
+    {
+      "epoch": 2.860941872234019,
+      "grad_norm": 0.8725518584251404,
+      "learning_rate": 2.3325659548569766e-06,
+      "loss": 0.8637,
+      "step": 36040
+    },
+    {
+      "epoch": 2.861735696282919,
+      "grad_norm": 1.1010116338729858,
+      "learning_rate": 2.3193352914715146e-06,
+      "loss": 0.7869,
+      "step": 36050
+    },
+    {
+      "epoch": 2.8625295203318184,
+      "grad_norm": 1.0355409383773804,
+      "learning_rate": 2.3061046280860527e-06,
+      "loss": 0.7937,
+      "step": 36060
+    },
+    {
+      "epoch": 2.863323344380718,
+      "grad_norm": 1.042080283164978,
+      "learning_rate": 2.29287396470059e-06,
+      "loss": 0.7468,
+      "step": 36070
+    },
+    {
+      "epoch": 2.8641171684296176,
+      "grad_norm": 0.9209195375442505,
+      "learning_rate": 2.279643301315128e-06,
+      "loss": 0.765,
+      "step": 36080
+    },
+    {
+      "epoch": 2.864910992478517,
+      "grad_norm": 0.9951355457305908,
+      "learning_rate": 2.2664126379296657e-06,
+      "loss": 0.7323,
+      "step": 36090
+    },
+    {
+      "epoch": 2.865704816527417,
+      "grad_norm": 0.969089925289154,
+      "learning_rate": 2.2531819745442038e-06,
+      "loss": 0.7319,
+      "step": 36100
+    },
+    {
+      "epoch": 2.8664986405763164,
+      "grad_norm": 1.1039822101593018,
+      "learning_rate": 2.239951311158742e-06,
+      "loss": 0.7683,
+      "step": 36110
+    },
+    {
+      "epoch": 2.8672924646252156,
+      "grad_norm": 1.0060070753097534,
+      "learning_rate": 2.2267206477732795e-06,
+      "loss": 0.7516,
+      "step": 36120
+    },
+    {
+      "epoch": 2.8680862886741156,
+      "grad_norm": 0.8177821636199951,
+      "learning_rate": 2.213489984387817e-06,
+      "loss": 0.8189,
+      "step": 36130
+    },
+    {
+      "epoch": 2.868880112723015,
+      "grad_norm": 0.8943607807159424,
+      "learning_rate": 2.2002593210023552e-06,
+      "loss": 0.7652,
+      "step": 36140
+    },
+    {
+      "epoch": 2.8696739367719144,
+      "grad_norm": 0.8499715328216553,
+      "learning_rate": 2.187028657616893e-06,
+      "loss": 0.8638,
+      "step": 36150
+    },
+    {
+      "epoch": 2.870467760820814,
+      "grad_norm": 0.8518121838569641,
+      "learning_rate": 2.173797994231431e-06,
+      "loss": 0.8079,
+      "step": 36160
+    },
+    {
+      "epoch": 2.8712615848697136,
+      "grad_norm": 1.0183836221694946,
+      "learning_rate": 2.1605673308459686e-06,
+      "loss": 0.7418,
+      "step": 36170
+    },
+    {
+      "epoch": 2.872055408918613,
+      "grad_norm": 1.092712640762329,
+      "learning_rate": 2.1473366674605063e-06,
+      "loss": 0.7641,
+      "step": 36180
+    },
+    {
+      "epoch": 2.872849232967513,
+      "grad_norm": 1.0446547269821167,
+      "learning_rate": 2.1341060040750444e-06,
+      "loss": 0.79,
+      "step": 36190
+    },
+    {
+      "epoch": 2.8736430570164124,
+      "grad_norm": 0.9164413213729858,
+      "learning_rate": 2.1208753406895824e-06,
+      "loss": 0.8226,
+      "step": 36200
+    },
+    {
+      "epoch": 2.874436881065312,
+      "grad_norm": 0.9824967980384827,
+      "learning_rate": 2.10764467730412e-06,
+      "loss": 0.7573,
+      "step": 36210
+    },
+    {
+      "epoch": 2.8752307051142116,
+      "grad_norm": 1.1960909366607666,
+      "learning_rate": 2.094414013918658e-06,
+      "loss": 0.7664,
+      "step": 36220
+    },
+    {
+      "epoch": 2.876024529163111,
+      "grad_norm": 1.140021562576294,
+      "learning_rate": 2.081183350533196e-06,
+      "loss": 0.7775,
+      "step": 36230
+    },
+    {
+      "epoch": 2.8768183532120104,
+      "grad_norm": 1.0924111604690552,
+      "learning_rate": 2.0679526871477335e-06,
+      "loss": 0.7845,
+      "step": 36240
+    },
+    {
+      "epoch": 2.87761217726091,
+      "grad_norm": 0.8583391308784485,
+      "learning_rate": 2.0547220237622716e-06,
+      "loss": 0.8183,
+      "step": 36250
+    },
+    {
+      "epoch": 2.8784060013098096,
+      "grad_norm": 1.004576563835144,
+      "learning_rate": 2.0414913603768097e-06,
+      "loss": 0.7346,
+      "step": 36260
+    },
+    {
+      "epoch": 2.879199825358709,
+      "grad_norm": 0.8908334374427795,
+      "learning_rate": 2.0282606969913473e-06,
+      "loss": 0.86,
+      "step": 36270
+    },
+    {
+      "epoch": 2.879993649407609,
+      "grad_norm": 1.0126510858535767,
+      "learning_rate": 2.015030033605885e-06,
+      "loss": 0.8082,
+      "step": 36280
+    },
+    {
+      "epoch": 2.8807874734565084,
+      "grad_norm": 1.0101633071899414,
+      "learning_rate": 2.001799370220423e-06,
+      "loss": 0.796,
+      "step": 36290
+    },
+    {
+      "epoch": 2.881581297505408,
+      "grad_norm": 0.8697376847267151,
+      "learning_rate": 1.9885687068349607e-06,
+      "loss": 0.7772,
+      "step": 36300
+    },
+    {
+      "epoch": 2.8823751215543076,
+      "grad_norm": 1.0588469505310059,
+      "learning_rate": 1.9753380434494988e-06,
+      "loss": 0.7941,
+      "step": 36310
+    },
+    {
+      "epoch": 2.883168945603207,
+      "grad_norm": 0.8329793810844421,
+      "learning_rate": 1.962107380064037e-06,
+      "loss": 0.7721,
+      "step": 36320
+    },
+    {
+      "epoch": 2.883962769652107,
+      "grad_norm": 1.057317852973938,
+      "learning_rate": 1.948876716678574e-06,
+      "loss": 0.7572,
+      "step": 36330
+    },
+    {
+      "epoch": 2.884756593701006,
+      "grad_norm": 0.87096107006073,
+      "learning_rate": 1.935646053293112e-06,
+      "loss": 0.7499,
+      "step": 36340
+    },
+    {
+      "epoch": 2.8855504177499056,
+      "grad_norm": 0.9541813135147095,
+      "learning_rate": 1.92241538990765e-06,
+      "loss": 0.7405,
+      "step": 36350
+    },
+    {
+      "epoch": 2.886344241798805,
+      "grad_norm": 0.7810908555984497,
+      "learning_rate": 1.909184726522188e-06,
+      "loss": 0.7629,
+      "step": 36360
+    },
+    {
+      "epoch": 2.887138065847705,
+      "grad_norm": 0.8539981842041016,
+      "learning_rate": 1.8959540631367258e-06,
+      "loss": 0.7746,
+      "step": 36370
+    },
+    {
+      "epoch": 2.8879318898966044,
+      "grad_norm": 1.008750081062317,
+      "learning_rate": 1.8827233997512634e-06,
+      "loss": 0.8307,
+      "step": 36380
+    },
+    {
+      "epoch": 2.888725713945504,
+      "grad_norm": 0.9009593725204468,
+      "learning_rate": 1.8694927363658015e-06,
+      "loss": 0.7134,
+      "step": 36390
+    },
+    {
+      "epoch": 2.8895195379944036,
+      "grad_norm": 1.03337562084198,
+      "learning_rate": 1.8562620729803394e-06,
+      "loss": 0.7966,
+      "step": 36400
+    },
+    {
+      "epoch": 2.8903133620433032,
+      "grad_norm": 0.9321558475494385,
+      "learning_rate": 1.843031409594877e-06,
+      "loss": 0.8321,
+      "step": 36410
+    },
+    {
+      "epoch": 2.891107186092203,
+      "grad_norm": 0.8705295324325562,
+      "learning_rate": 1.829800746209415e-06,
+      "loss": 0.821,
+      "step": 36420
+    },
+    {
+      "epoch": 2.891901010141102,
+      "grad_norm": 1.093181848526001,
+      "learning_rate": 1.816570082823953e-06,
+      "loss": 0.7454,
+      "step": 36430
+    },
+    {
+      "epoch": 2.892694834190002,
+      "grad_norm": 1.1398403644561768,
+      "learning_rate": 1.8033394194384906e-06,
+      "loss": 0.8026,
+      "step": 36440
+    },
+    {
+      "epoch": 2.8934886582389012,
+      "grad_norm": 0.8322063088417053,
+      "learning_rate": 1.7901087560530285e-06,
+      "loss": 0.8001,
+      "step": 36450
+    },
+    {
+      "epoch": 2.894282482287801,
+      "grad_norm": 0.9073172807693481,
+      "learning_rate": 1.7768780926675666e-06,
+      "loss": 0.8225,
+      "step": 36460
+    },
+    {
+      "epoch": 2.8950763063367004,
+      "grad_norm": 0.8599694967269897,
+      "learning_rate": 1.7636474292821043e-06,
+      "loss": 0.7355,
+      "step": 36470
+    },
+    {
+      "epoch": 2.8958701303856,
+      "grad_norm": 0.9327001571655273,
+      "learning_rate": 1.7504167658966421e-06,
+      "loss": 0.7628,
+      "step": 36480
+    },
+    {
+      "epoch": 2.8966639544344996,
+      "grad_norm": 1.1032301187515259,
+      "learning_rate": 1.7371861025111802e-06,
+      "loss": 0.7719,
+      "step": 36490
+    },
+    {
+      "epoch": 2.8974577784833992,
+      "grad_norm": 1.0107958316802979,
+      "learning_rate": 1.7239554391257176e-06,
+      "loss": 0.7785,
+      "step": 36500
+    },
+    {
+      "epoch": 2.898251602532299,
+      "grad_norm": 0.9559915065765381,
+      "learning_rate": 1.7107247757402557e-06,
+      "loss": 0.7749,
+      "step": 36510
+    },
+    {
+      "epoch": 2.899045426581198,
+      "grad_norm": 0.9474267363548279,
+      "learning_rate": 1.6974941123547936e-06,
+      "loss": 0.8143,
+      "step": 36520
+    },
+    {
+      "epoch": 2.899839250630098,
+      "grad_norm": 0.7721654772758484,
+      "learning_rate": 1.6842634489693313e-06,
+      "loss": 0.7488,
+      "step": 36530
+    },
+    {
+      "epoch": 2.9006330746789972,
+      "grad_norm": 0.8206939697265625,
+      "learning_rate": 1.6710327855838693e-06,
+      "loss": 0.7174,
+      "step": 36540
+    },
+    {
+      "epoch": 2.901426898727897,
+      "grad_norm": 0.8247297406196594,
+      "learning_rate": 1.6578021221984072e-06,
+      "loss": 0.7827,
+      "step": 36550
+    },
+    {
+      "epoch": 2.9022207227767964,
+      "grad_norm": 1.261728286743164,
+      "learning_rate": 1.6445714588129449e-06,
+      "loss": 0.7458,
+      "step": 36560
+    },
+    {
+      "epoch": 2.903014546825696,
+      "grad_norm": 0.9776942729949951,
+      "learning_rate": 1.631340795427483e-06,
+      "loss": 0.7707,
+      "step": 36570
+    },
+    {
+      "epoch": 2.9038083708745956,
+      "grad_norm": 1.0205057859420776,
+      "learning_rate": 1.6181101320420208e-06,
+      "loss": 0.7757,
+      "step": 36580
+    },
+    {
+      "epoch": 2.9046021949234953,
+      "grad_norm": 1.0153295993804932,
+      "learning_rate": 1.6048794686565585e-06,
+      "loss": 0.7932,
+      "step": 36590
+    },
+    {
+      "epoch": 2.905396018972395,
+      "grad_norm": 1.073330044746399,
+      "learning_rate": 1.5916488052710963e-06,
+      "loss": 0.742,
+      "step": 36600
+    },
+    {
+      "epoch": 2.9061898430212945,
+      "grad_norm": 1.0096873044967651,
+      "learning_rate": 1.578418141885634e-06,
+      "loss": 0.8228,
+      "step": 36610
+    },
+    {
+      "epoch": 2.906983667070194,
+      "grad_norm": 1.1333979368209839,
+      "learning_rate": 1.565187478500172e-06,
+      "loss": 0.7219,
+      "step": 36620
+    },
+    {
+      "epoch": 2.9077774911190932,
+      "grad_norm": 1.1649788618087769,
+      "learning_rate": 1.55195681511471e-06,
+      "loss": 0.8223,
+      "step": 36630
+    },
+    {
+      "epoch": 2.9085713151679933,
+      "grad_norm": 0.8990685939788818,
+      "learning_rate": 1.5387261517292478e-06,
+      "loss": 0.7569,
+      "step": 36640
+    },
+    {
+      "epoch": 2.9093651392168924,
+      "grad_norm": 1.097396731376648,
+      "learning_rate": 1.5254954883437857e-06,
+      "loss": 0.7456,
+      "step": 36650
+    },
+    {
+      "epoch": 2.910158963265792,
+      "grad_norm": 0.9843643307685852,
+      "learning_rate": 1.5122648249583233e-06,
+      "loss": 0.7238,
+      "step": 36660
+    },
+    {
+      "epoch": 2.9109527873146916,
+      "grad_norm": 1.0941895246505737,
+      "learning_rate": 1.4990341615728614e-06,
+      "loss": 0.8031,
+      "step": 36670
+    },
+    {
+      "epoch": 2.9117466113635913,
+      "grad_norm": 0.9502320289611816,
+      "learning_rate": 1.485803498187399e-06,
+      "loss": 0.8127,
+      "step": 36680
+    },
+    {
+      "epoch": 2.912540435412491,
+      "grad_norm": 0.8268551826477051,
+      "learning_rate": 1.472572834801937e-06,
+      "loss": 0.7619,
+      "step": 36690
+    },
+    {
+      "epoch": 2.9133342594613905,
+      "grad_norm": 1.1277645826339722,
+      "learning_rate": 1.459342171416475e-06,
+      "loss": 0.8046,
+      "step": 36700
+    },
+    {
+      "epoch": 2.91412808351029,
+      "grad_norm": 0.9089832901954651,
+      "learning_rate": 1.4461115080310127e-06,
+      "loss": 0.7497,
+      "step": 36710
+    },
+    {
+      "epoch": 2.9149219075591892,
+      "grad_norm": 0.9186846017837524,
+      "learning_rate": 1.4328808446455505e-06,
+      "loss": 0.771,
+      "step": 36720
+    },
+    {
+      "epoch": 2.9157157316080893,
+      "grad_norm": 1.0367939472198486,
+      "learning_rate": 1.4196501812600884e-06,
+      "loss": 0.7857,
+      "step": 36730
+    },
+    {
+      "epoch": 2.9165095556569884,
+      "grad_norm": 0.9849185347557068,
+      "learning_rate": 1.4064195178746263e-06,
+      "loss": 0.8304,
+      "step": 36740
+    },
+    {
+      "epoch": 2.917303379705888,
+      "grad_norm": 1.022588849067688,
+      "learning_rate": 1.3931888544891641e-06,
+      "loss": 0.7862,
+      "step": 36750
+    },
+    {
+      "epoch": 2.9180972037547876,
+      "grad_norm": 0.8609658479690552,
+      "learning_rate": 1.379958191103702e-06,
+      "loss": 0.7718,
+      "step": 36760
+    },
+    {
+      "epoch": 2.9188910278036873,
+      "grad_norm": 0.9891222715377808,
+      "learning_rate": 1.3667275277182399e-06,
+      "loss": 0.7544,
+      "step": 36770
+    },
+    {
+      "epoch": 2.919684851852587,
+      "grad_norm": 1.002481460571289,
+      "learning_rate": 1.3534968643327777e-06,
+      "loss": 0.8537,
+      "step": 36780
+    },
+    {
+      "epoch": 2.9204786759014865,
+      "grad_norm": 1.0346823930740356,
+      "learning_rate": 1.3402662009473156e-06,
+      "loss": 0.7605,
+      "step": 36790
+    },
+    {
+      "epoch": 2.921272499950386,
+      "grad_norm": 1.154091715812683,
+      "learning_rate": 1.3270355375618535e-06,
+      "loss": 0.7222,
+      "step": 36800
+    },
+    {
+      "epoch": 2.9220663239992857,
+      "grad_norm": 1.07093346118927,
+      "learning_rate": 1.3138048741763914e-06,
+      "loss": 0.8214,
+      "step": 36810
+    },
+    {
+      "epoch": 2.9228601480481853,
+      "grad_norm": 0.9682141542434692,
+      "learning_rate": 1.300574210790929e-06,
+      "loss": 0.7921,
+      "step": 36820
+    },
+    {
+      "epoch": 2.9236539720970844,
+      "grad_norm": 0.9235316514968872,
+      "learning_rate": 1.287343547405467e-06,
+      "loss": 0.8082,
+      "step": 36830
+    },
+    {
+      "epoch": 2.9244477961459845,
+      "grad_norm": 1.0996876955032349,
+      "learning_rate": 1.2741128840200047e-06,
+      "loss": 0.7969,
+      "step": 36840
+    },
+    {
+      "epoch": 2.9252416201948837,
+      "grad_norm": 0.8696537017822266,
+      "learning_rate": 1.2608822206345426e-06,
+      "loss": 0.7344,
+      "step": 36850
+    },
+    {
+      "epoch": 2.9260354442437833,
+      "grad_norm": 1.0605179071426392,
+      "learning_rate": 1.2476515572490805e-06,
+      "loss": 0.7609,
+      "step": 36860
+    },
+    {
+      "epoch": 2.926829268292683,
+      "grad_norm": 1.034128189086914,
+      "learning_rate": 1.2344208938636184e-06,
+      "loss": 0.8119,
+      "step": 36870
+    },
+    {
+      "epoch": 2.9276230923415825,
+      "grad_norm": 1.083617091178894,
+      "learning_rate": 1.2211902304781562e-06,
+      "loss": 0.7628,
+      "step": 36880
+    },
+    {
+      "epoch": 2.928416916390482,
+      "grad_norm": 1.001076102256775,
+      "learning_rate": 1.207959567092694e-06,
+      "loss": 0.7721,
+      "step": 36890
+    },
+    {
+      "epoch": 2.9292107404393817,
+      "grad_norm": 0.8138130903244019,
+      "learning_rate": 1.194728903707232e-06,
+      "loss": 0.7045,
+      "step": 36900
+    },
+    {
+      "epoch": 2.9300045644882813,
+      "grad_norm": 0.8781431317329407,
+      "learning_rate": 1.1814982403217698e-06,
+      "loss": 0.8074,
+      "step": 36910
+    },
+    {
+      "epoch": 2.930798388537181,
+      "grad_norm": 0.8041318655014038,
+      "learning_rate": 1.1682675769363077e-06,
+      "loss": 0.7763,
+      "step": 36920
+    },
+    {
+      "epoch": 2.9315922125860805,
+      "grad_norm": 0.9655971527099609,
+      "learning_rate": 1.1550369135508456e-06,
+      "loss": 0.7511,
+      "step": 36930
+    },
+    {
+      "epoch": 2.9323860366349797,
+      "grad_norm": 1.0426075458526611,
+      "learning_rate": 1.1418062501653834e-06,
+      "loss": 0.7663,
+      "step": 36940
+    },
+    {
+      "epoch": 2.9331798606838793,
+      "grad_norm": 1.0765414237976074,
+      "learning_rate": 1.128575586779921e-06,
+      "loss": 0.8248,
+      "step": 36950
+    },
+    {
+      "epoch": 2.933973684732779,
+      "grad_norm": 0.8941267132759094,
+      "learning_rate": 1.1153449233944592e-06,
+      "loss": 0.8233,
+      "step": 36960
+    },
+    {
+      "epoch": 2.9347675087816785,
+      "grad_norm": 1.1485576629638672,
+      "learning_rate": 1.1021142600089968e-06,
+      "loss": 0.7031,
+      "step": 36970
+    },
+    {
+      "epoch": 2.935561332830578,
+      "grad_norm": 1.0902643203735352,
+      "learning_rate": 1.0888835966235347e-06,
+      "loss": 0.7641,
+      "step": 36980
+    },
+    {
+      "epoch": 2.9363551568794777,
+      "grad_norm": 0.8667743802070618,
+      "learning_rate": 1.0756529332380728e-06,
+      "loss": 0.7518,
+      "step": 36990
+    },
+    {
+      "epoch": 2.9371489809283773,
+      "grad_norm": 1.2408496141433716,
+      "learning_rate": 1.0624222698526104e-06,
+      "loss": 0.8086,
+      "step": 37000
+    },
+    {
+      "epoch": 2.937942804977277,
+      "grad_norm": 0.9038211703300476,
+      "learning_rate": 1.0491916064671483e-06,
+      "loss": 0.7384,
+      "step": 37010
+    },
+    {
+      "epoch": 2.9387366290261765,
+      "grad_norm": 1.0854371786117554,
+      "learning_rate": 1.0359609430816862e-06,
+      "loss": 0.73,
+      "step": 37020
+    },
+    {
+      "epoch": 2.9395304530750757,
+      "grad_norm": 0.986851155757904,
+      "learning_rate": 1.022730279696224e-06,
+      "loss": 0.7916,
+      "step": 37030
+    },
+    {
+      "epoch": 2.9403242771239757,
+      "grad_norm": 0.9306720495223999,
+      "learning_rate": 1.009499616310762e-06,
+      "loss": 0.7434,
+      "step": 37040
+    },
+    {
+      "epoch": 2.941118101172875,
+      "grad_norm": 1.075705647468567,
+      "learning_rate": 9.962689529252998e-07,
+      "loss": 0.7953,
+      "step": 37050
+    },
+    {
+      "epoch": 2.9419119252217745,
+      "grad_norm": 0.9013558626174927,
+      "learning_rate": 9.830382895398376e-07,
+      "loss": 0.7916,
+      "step": 37060
+    },
+    {
+      "epoch": 2.942705749270674,
+      "grad_norm": 1.0537384748458862,
+      "learning_rate": 9.698076261543755e-07,
+      "loss": 0.768,
+      "step": 37070
+    },
+    {
+      "epoch": 2.9434995733195737,
+      "grad_norm": 0.8394224643707275,
+      "learning_rate": 9.565769627689132e-07,
+      "loss": 0.8091,
+      "step": 37080
+    },
+    {
+      "epoch": 2.9442933973684733,
+      "grad_norm": 0.9808106422424316,
+      "learning_rate": 9.433462993834511e-07,
+      "loss": 0.7533,
+      "step": 37090
+    },
+    {
+      "epoch": 2.945087221417373,
+      "grad_norm": 0.8724024295806885,
+      "learning_rate": 9.30115635997989e-07,
+      "loss": 0.736,
+      "step": 37100
+    },
+    {
+      "epoch": 2.9458810454662725,
+      "grad_norm": 1.0221234560012817,
+      "learning_rate": 9.168849726125268e-07,
+      "loss": 0.8032,
+      "step": 37110
+    },
+    {
+      "epoch": 2.946674869515172,
+      "grad_norm": 0.8556252121925354,
+      "learning_rate": 9.036543092270647e-07,
+      "loss": 0.7819,
+      "step": 37120
+    },
+    {
+      "epoch": 2.9474686935640717,
+      "grad_norm": 0.9892777800559998,
+      "learning_rate": 8.904236458416025e-07,
+      "loss": 0.7875,
+      "step": 37130
+    },
+    {
+      "epoch": 2.948262517612971,
+      "grad_norm": 0.8527591824531555,
+      "learning_rate": 8.771929824561404e-07,
+      "loss": 0.7543,
+      "step": 37140
+    },
+    {
+      "epoch": 2.949056341661871,
+      "grad_norm": 1.1402502059936523,
+      "learning_rate": 8.639623190706783e-07,
+      "loss": 0.8077,
+      "step": 37150
+    },
+    {
+      "epoch": 2.94985016571077,
+      "grad_norm": 0.9675154089927673,
+      "learning_rate": 8.507316556852161e-07,
+      "loss": 0.7796,
+      "step": 37160
+    },
+    {
+      "epoch": 2.9506439897596697,
+      "grad_norm": 0.8908880352973938,
+      "learning_rate": 8.375009922997539e-07,
+      "loss": 0.7795,
+      "step": 37170
+    },
+    {
+      "epoch": 2.9514378138085693,
+      "grad_norm": 0.9992043972015381,
+      "learning_rate": 8.242703289142918e-07,
+      "loss": 0.801,
+      "step": 37180
+    },
+    {
+      "epoch": 2.952231637857469,
+      "grad_norm": 0.9588622450828552,
+      "learning_rate": 8.110396655288297e-07,
+      "loss": 0.83,
+      "step": 37190
+    },
+    {
+      "epoch": 2.9530254619063685,
+      "grad_norm": 0.9657321572303772,
+      "learning_rate": 7.978090021433675e-07,
+      "loss": 0.8267,
+      "step": 37200
+    },
+    {
+      "epoch": 2.953819285955268,
+      "grad_norm": 0.8427868485450745,
+      "learning_rate": 7.845783387579052e-07,
+      "loss": 0.8107,
+      "step": 37210
+    },
+    {
+      "epoch": 2.9546131100041677,
+      "grad_norm": 1.1387544870376587,
+      "learning_rate": 7.713476753724432e-07,
+      "loss": 0.7857,
+      "step": 37220
+    },
+    {
+      "epoch": 2.955406934053067,
+      "grad_norm": 0.8458208441734314,
+      "learning_rate": 7.581170119869811e-07,
+      "loss": 0.815,
+      "step": 37230
+    },
+    {
+      "epoch": 2.956200758101967,
+      "grad_norm": 1.2061816453933716,
+      "learning_rate": 7.44886348601519e-07,
+      "loss": 0.7483,
+      "step": 37240
+    },
+    {
+      "epoch": 2.956994582150866,
+      "grad_norm": 0.93003910779953,
+      "learning_rate": 7.316556852160567e-07,
+      "loss": 0.7802,
+      "step": 37250
+    },
+    {
+      "epoch": 2.9577884061997657,
+      "grad_norm": 0.8982664346694946,
+      "learning_rate": 7.184250218305946e-07,
+      "loss": 0.8041,
+      "step": 37260
+    },
+    {
+      "epoch": 2.9585822302486653,
+      "grad_norm": 1.058793306350708,
+      "learning_rate": 7.051943584451326e-07,
+      "loss": 0.7814,
+      "step": 37270
+    },
+    {
+      "epoch": 2.959376054297565,
+      "grad_norm": 0.9171466827392578,
+      "learning_rate": 6.919636950596703e-07,
+      "loss": 0.7798,
+      "step": 37280
+    },
+    {
+      "epoch": 2.9601698783464645,
+      "grad_norm": 0.8719449043273926,
+      "learning_rate": 6.787330316742082e-07,
+      "loss": 0.8213,
+      "step": 37290
+    },
+    {
+      "epoch": 2.960963702395364,
+      "grad_norm": 0.8667718172073364,
+      "learning_rate": 6.655023682887459e-07,
+      "loss": 0.7089,
+      "step": 37300
+    },
+    {
+      "epoch": 2.9617575264442637,
+      "grad_norm": 0.8706735968589783,
+      "learning_rate": 6.522717049032839e-07,
+      "loss": 0.7605,
+      "step": 37310
+    },
+    {
+      "epoch": 2.9625513504931633,
+      "grad_norm": 0.9747303128242493,
+      "learning_rate": 6.390410415178218e-07,
+      "loss": 0.7329,
+      "step": 37320
+    },
+    {
+      "epoch": 2.963345174542063,
+      "grad_norm": 1.0735442638397217,
+      "learning_rate": 6.258103781323596e-07,
+      "loss": 0.784,
+      "step": 37330
+    },
+    {
+      "epoch": 2.964138998590962,
+      "grad_norm": 0.8519766330718994,
+      "learning_rate": 6.125797147468974e-07,
+      "loss": 0.7985,
+      "step": 37340
+    },
+    {
+      "epoch": 2.964932822639862,
+      "grad_norm": 1.0868384838104248,
+      "learning_rate": 5.993490513614353e-07,
+      "loss": 0.8058,
+      "step": 37350
+    },
+    {
+      "epoch": 2.9657266466887613,
+      "grad_norm": 1.1832605600357056,
+      "learning_rate": 5.861183879759732e-07,
+      "loss": 0.7284,
+      "step": 37360
+    },
+    {
+      "epoch": 2.966520470737661,
+      "grad_norm": 1.0526885986328125,
+      "learning_rate": 5.72887724590511e-07,
+      "loss": 0.8052,
+      "step": 37370
+    },
+    {
+      "epoch": 2.9673142947865605,
+      "grad_norm": 1.137607455253601,
+      "learning_rate": 5.596570612050488e-07,
+      "loss": 0.8004,
+      "step": 37380
+    },
+    {
+      "epoch": 2.96810811883546,
+      "grad_norm": 0.8867448568344116,
+      "learning_rate": 5.464263978195867e-07,
+      "loss": 0.7579,
+      "step": 37390
+    },
+    {
+      "epoch": 2.9689019428843597,
+      "grad_norm": 0.992607057094574,
+      "learning_rate": 5.331957344341246e-07,
+      "loss": 0.7234,
+      "step": 37400
+    },
+    {
+      "epoch": 2.9696957669332593,
+      "grad_norm": 0.9389944076538086,
+      "learning_rate": 5.199650710486624e-07,
+      "loss": 0.7204,
+      "step": 37410
+    },
+    {
+      "epoch": 2.970489590982159,
+      "grad_norm": 1.1689140796661377,
+      "learning_rate": 5.067344076632003e-07,
+      "loss": 0.7956,
+      "step": 37420
+    },
+    {
+      "epoch": 2.9712834150310585,
+      "grad_norm": 1.0063467025756836,
+      "learning_rate": 4.93503744277738e-07,
+      "loss": 0.8114,
+      "step": 37430
+    },
+    {
+      "epoch": 2.972077239079958,
+      "grad_norm": 0.9847369194030762,
+      "learning_rate": 4.815961472308222e-07,
+      "loss": 0.7854,
+      "step": 37440
+    },
+    {
+      "epoch": 2.9728710631288573,
+      "grad_norm": 0.8891132473945618,
+      "learning_rate": 4.6836548384536e-07,
+      "loss": 0.8131,
+      "step": 37450
+    },
+    {
+      "epoch": 2.973664887177757,
+      "grad_norm": 1.1381394863128662,
+      "learning_rate": 4.551348204598979e-07,
+      "loss": 0.7877,
+      "step": 37460
+    },
+    {
+      "epoch": 2.9744587112266565,
+      "grad_norm": 1.0094081163406372,
+      "learning_rate": 4.4190415707443577e-07,
+      "loss": 0.7812,
+      "step": 37470
+    },
+    {
+      "epoch": 2.975252535275556,
+      "grad_norm": 1.1680759191513062,
+      "learning_rate": 4.286734936889736e-07,
+      "loss": 0.7642,
+      "step": 37480
+    },
+    {
+      "epoch": 2.9760463593244557,
+      "grad_norm": 1.0051449537277222,
+      "learning_rate": 4.1544283030351145e-07,
+      "loss": 0.7755,
+      "step": 37490
+    },
+    {
+      "epoch": 2.9768401833733553,
+      "grad_norm": 1.1129390001296997,
+      "learning_rate": 4.0221216691804927e-07,
+      "loss": 0.7441,
+      "step": 37500
+    },
+    {
+      "epoch": 2.977634007422255,
+      "grad_norm": 0.8162450194358826,
+      "learning_rate": 3.8898150353258714e-07,
+      "loss": 0.7353,
+      "step": 37510
+    },
+    {
+      "epoch": 2.9784278314711545,
+      "grad_norm": 0.8343786001205444,
+      "learning_rate": 3.7575084014712495e-07,
+      "loss": 0.8653,
+      "step": 37520
+    },
+    {
+      "epoch": 2.979221655520054,
+      "grad_norm": 0.9422026872634888,
+      "learning_rate": 3.625201767616629e-07,
+      "loss": 0.7788,
+      "step": 37530
+    },
+    {
+      "epoch": 2.9800154795689533,
+      "grad_norm": 0.7513265013694763,
+      "learning_rate": 3.492895133762007e-07,
+      "loss": 0.7978,
+      "step": 37540
+    },
+    {
+      "epoch": 2.9808093036178533,
+      "grad_norm": 1.1932988166809082,
+      "learning_rate": 3.3605884999073856e-07,
+      "loss": 0.7766,
+      "step": 37550
+    },
+    {
+      "epoch": 2.9816031276667525,
+      "grad_norm": 0.8924500942230225,
+      "learning_rate": 3.228281866052764e-07,
+      "loss": 0.816,
+      "step": 37560
+    },
+    {
+      "epoch": 2.982396951715652,
+      "grad_norm": 1.2411748170852661,
+      "learning_rate": 3.095975232198143e-07,
+      "loss": 0.7699,
+      "step": 37570
+    },
+    {
+      "epoch": 2.9831907757645517,
+      "grad_norm": 0.9249628186225891,
+      "learning_rate": 2.963668598343521e-07,
+      "loss": 0.8282,
+      "step": 37580
+    },
+    {
+      "epoch": 2.9839845998134513,
+      "grad_norm": 0.9491875171661377,
+      "learning_rate": 2.8313619644889e-07,
+      "loss": 0.848,
+      "step": 37590
+    },
+    {
+      "epoch": 2.984778423862351,
+      "grad_norm": 1.0644991397857666,
+      "learning_rate": 2.699055330634278e-07,
+      "loss": 0.7872,
+      "step": 37600
+    },
+    {
+      "epoch": 2.9855722479112505,
+      "grad_norm": 1.1010690927505493,
+      "learning_rate": 2.5667486967796566e-07,
+      "loss": 0.7806,
+      "step": 37610
+    },
+    {
+      "epoch": 2.98636607196015,
+      "grad_norm": 0.8743943572044373,
+      "learning_rate": 2.4344420629250353e-07,
+      "loss": 0.7946,
+      "step": 37620
+    },
+    {
+      "epoch": 2.9871598960090497,
+      "grad_norm": 0.909243106842041,
+      "learning_rate": 2.3021354290704137e-07,
+      "loss": 0.8322,
+      "step": 37630
+    },
+    {
+      "epoch": 2.9879537200579493,
+      "grad_norm": 0.8433053493499756,
+      "learning_rate": 2.1698287952157921e-07,
+      "loss": 0.8142,
+      "step": 37640
+    },
+    {
+      "epoch": 2.9887475441068485,
+      "grad_norm": 0.8749028444290161,
+      "learning_rate": 2.0375221613611706e-07,
+      "loss": 0.7644,
+      "step": 37650
+    },
+    {
+      "epoch": 2.989541368155748,
+      "grad_norm": 1.0767362117767334,
+      "learning_rate": 1.9052155275065492e-07,
+      "loss": 0.7983,
+      "step": 37660
+    },
+    {
+      "epoch": 2.9903351922046477,
+      "grad_norm": 1.0285117626190186,
+      "learning_rate": 1.772908893651928e-07,
+      "loss": 0.7426,
+      "step": 37670
+    },
+    {
+      "epoch": 2.9911290162535473,
+      "grad_norm": 0.986137330532074,
+      "learning_rate": 1.6406022597973063e-07,
+      "loss": 0.7596,
+      "step": 37680
+    },
+    {
+      "epoch": 2.991922840302447,
+      "grad_norm": 0.9621859192848206,
+      "learning_rate": 1.508295625942685e-07,
+      "loss": 0.7735,
+      "step": 37690
+    },
+    {
+      "epoch": 2.9927166643513465,
+      "grad_norm": 0.9703050851821899,
+      "learning_rate": 1.3759889920880634e-07,
+      "loss": 0.7648,
+      "step": 37700
+    },
+    {
+      "epoch": 2.993510488400246,
+      "grad_norm": 0.7976683378219604,
+      "learning_rate": 1.2436823582334419e-07,
+      "loss": 0.7838,
+      "step": 37710
+    },
+    {
+      "epoch": 2.9943043124491457,
+      "grad_norm": 0.8955792188644409,
+      "learning_rate": 1.1113757243788204e-07,
+      "loss": 0.7638,
+      "step": 37720
+    },
+    {
+      "epoch": 2.9950981364980453,
+      "grad_norm": 0.8139486312866211,
+      "learning_rate": 9.79069090524199e-08,
+      "loss": 0.7634,
+      "step": 37730
+    },
+    {
+      "epoch": 2.9958919605469445,
+      "grad_norm": 1.1136054992675781,
+      "learning_rate": 8.467624566695774e-08,
+      "loss": 0.7807,
+      "step": 37740
+    },
+    {
+      "epoch": 2.9966857845958446,
+      "grad_norm": 0.8526073098182678,
+      "learning_rate": 7.144558228149559e-08,
+      "loss": 0.742,
+      "step": 37750
+    },
+    {
+      "epoch": 2.9974796086447437,
+      "grad_norm": 0.9740519523620605,
+      "learning_rate": 5.821491889603345e-08,
+      "loss": 0.7485,
+      "step": 37760
+    },
+    {
+      "epoch": 2.9982734326936433,
+      "grad_norm": 0.8623798489570618,
+      "learning_rate": 4.4984255510571303e-08,
+      "loss": 0.8157,
+      "step": 37770
+    },
+    {
+      "epoch": 2.999067256742543,
+      "grad_norm": 1.0553157329559326,
+      "learning_rate": 3.175359212510916e-08,
+      "loss": 0.7943,
+      "step": 37780
+    },
+    {
+      "epoch": 2.9998610807914425,
+      "grad_norm": 0.9887831807136536,
+      "learning_rate": 1.8522928739647007e-08,
+      "loss": 0.7406,
+      "step": 37790
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 37791,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.60495148153779e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}