diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24640 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9995732696082618,
+  "eval_steps": 500,
+  "global_step": 17574,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008534607834769992,
+      "grad_norm": 99.42818782210857,
+      "learning_rate": 1.4220705346985212e-07,
+      "loss": 6.2791,
+      "step": 5
+    },
+    {
+      "epoch": 0.0017069215669539984,
+      "grad_norm": 104.39808292384225,
+      "learning_rate": 2.8441410693970424e-07,
+      "loss": 6.1853,
+      "step": 10
+    },
+    {
+      "epoch": 0.0025603823504309975,
+      "grad_norm": 103.48376567294774,
+      "learning_rate": 4.2662116040955633e-07,
+      "loss": 6.3846,
+      "step": 15
+    },
+    {
+      "epoch": 0.003413843133907997,
+      "grad_norm": 104.04994240213826,
+      "learning_rate": 5.688282138794085e-07,
+      "loss": 5.7934,
+      "step": 20
+    },
+    {
+      "epoch": 0.004267303917384996,
+      "grad_norm": 87.84439258281431,
+      "learning_rate": 7.110352673492606e-07,
+      "loss": 5.2152,
+      "step": 25
+    },
+    {
+      "epoch": 0.005120764700861995,
+      "grad_norm": 77.01441231565465,
+      "learning_rate": 8.532423208191127e-07,
+      "loss": 3.8525,
+      "step": 30
+    },
+    {
+      "epoch": 0.005974225484338995,
+      "grad_norm": 21.83631807774304,
+      "learning_rate": 9.954493742889649e-07,
+      "loss": 2.9729,
+      "step": 35
+    },
+    {
+      "epoch": 0.006827686267815994,
+      "grad_norm": 14.113464115950803,
+      "learning_rate": 1.137656427758817e-06,
+      "loss": 1.7155,
+      "step": 40
+    },
+    {
+      "epoch": 0.0076811470512929934,
+      "grad_norm": 4.7553020883795405,
+      "learning_rate": 1.279863481228669e-06,
+      "loss": 1.4743,
+      "step": 45
+    },
+    {
+      "epoch": 0.008534607834769992,
+      "grad_norm": 7.79414775358199,
+      "learning_rate": 1.4220705346985211e-06,
+      "loss": 1.3836,
+      "step": 50
+    },
+    {
+      "epoch": 0.009388068618246991,
+      "grad_norm": 2.4354082720834977,
+      "learning_rate": 1.5642775881683732e-06,
+      "loss": 1.2929,
+      "step": 55
+    },
+    {
+      "epoch": 0.01024152940172399,
+      "grad_norm": 2.2849690291702522,
+      "learning_rate": 1.7064846416382253e-06,
+      "loss": 1.1902,
+      "step": 60
+    },
+    {
+      "epoch": 0.01109499018520099,
+      "grad_norm": 2.895821184135636,
+      "learning_rate": 1.8486916951080774e-06,
+      "loss": 1.0828,
+      "step": 65
+    },
+    {
+      "epoch": 0.01194845096867799,
+      "grad_norm": 1.1877547807762439,
+      "learning_rate": 1.9908987485779297e-06,
+      "loss": 1.0027,
+      "step": 70
+    },
+    {
+      "epoch": 0.012801911752154988,
+      "grad_norm": 1.0268455953088027,
+      "learning_rate": 2.133105802047782e-06,
+      "loss": 0.9333,
+      "step": 75
+    },
+    {
+      "epoch": 0.013655372535631987,
+      "grad_norm": 5.481550209143814,
+      "learning_rate": 2.275312855517634e-06,
+      "loss": 0.8887,
+      "step": 80
+    },
+    {
+      "epoch": 0.014508833319108986,
+      "grad_norm": 0.8793984613448603,
+      "learning_rate": 2.417519908987486e-06,
+      "loss": 0.8555,
+      "step": 85
+    },
+    {
+      "epoch": 0.015362294102585987,
+      "grad_norm": 0.7225726544778598,
+      "learning_rate": 2.559726962457338e-06,
+      "loss": 0.8049,
+      "step": 90
+    },
+    {
+      "epoch": 0.016215754886062986,
+      "grad_norm": 0.8823937518440574,
+      "learning_rate": 2.7019340159271904e-06,
+      "loss": 0.8209,
+      "step": 95
+    },
+    {
+      "epoch": 0.017069215669539985,
+      "grad_norm": 0.829188508851399,
+      "learning_rate": 2.8441410693970423e-06,
+      "loss": 0.7801,
+      "step": 100
+    },
+    {
+      "epoch": 0.017922676453016984,
+      "grad_norm": 0.7270054979813305,
+      "learning_rate": 2.9863481228668946e-06,
+      "loss": 0.7423,
+      "step": 105
+    },
+    {
+      "epoch": 0.018776137236493982,
+      "grad_norm": 0.6801373549704796,
+      "learning_rate": 3.1285551763367464e-06,
+      "loss": 0.7323,
+      "step": 110
+    },
+    {
+      "epoch": 0.01962959801997098,
+      "grad_norm": 1.0385640589855576,
+      "learning_rate": 3.2707622298065988e-06,
+      "loss": 0.7392,
+      "step": 115
+    },
+    {
+      "epoch": 0.02048305880344798,
+      "grad_norm": 0.8671806836602034,
+      "learning_rate": 3.4129692832764506e-06,
+      "loss": 0.6864,
+      "step": 120
+    },
+    {
+      "epoch": 0.02133651958692498,
+      "grad_norm": 0.6716992405032944,
+      "learning_rate": 3.5551763367463025e-06,
+      "loss": 0.7209,
+      "step": 125
+    },
+    {
+      "epoch": 0.02218998037040198,
+      "grad_norm": 2.3299276907106763,
+      "learning_rate": 3.697383390216155e-06,
+      "loss": 0.6971,
+      "step": 130
+    },
+    {
+      "epoch": 0.02304344115387898,
+      "grad_norm": 0.6490988370494056,
+      "learning_rate": 3.8395904436860075e-06,
+      "loss": 0.7131,
+      "step": 135
+    },
+    {
+      "epoch": 0.02389690193735598,
+      "grad_norm": 0.6320268165032422,
+      "learning_rate": 3.981797497155859e-06,
+      "loss": 0.6888,
+      "step": 140
+    },
+    {
+      "epoch": 0.024750362720832978,
+      "grad_norm": 0.7161127122525036,
+      "learning_rate": 4.124004550625711e-06,
+      "loss": 0.6837,
+      "step": 145
+    },
+    {
+      "epoch": 0.025603823504309977,
+      "grad_norm": 0.6972994060141804,
+      "learning_rate": 4.266211604095564e-06,
+      "loss": 0.6523,
+      "step": 150
+    },
+    {
+      "epoch": 0.026457284287786976,
+      "grad_norm": 0.7062186818568618,
+      "learning_rate": 4.408418657565416e-06,
+      "loss": 0.6649,
+      "step": 155
+    },
+    {
+      "epoch": 0.027310745071263975,
+      "grad_norm": 0.6568377650981212,
+      "learning_rate": 4.550625711035268e-06,
+      "loss": 0.6876,
+      "step": 160
+    },
+    {
+      "epoch": 0.028164205854740974,
+      "grad_norm": 0.8430733551488065,
+      "learning_rate": 4.69283276450512e-06,
+      "loss": 0.6741,
+      "step": 165
+    },
+    {
+      "epoch": 0.029017666638217973,
+      "grad_norm": 0.6106681738824012,
+      "learning_rate": 4.835039817974972e-06,
+      "loss": 0.6777,
+      "step": 170
+    },
+    {
+      "epoch": 0.02987112742169497,
+      "grad_norm": 0.7901990923848753,
+      "learning_rate": 4.977246871444824e-06,
+      "loss": 0.6641,
+      "step": 175
+    },
+    {
+      "epoch": 0.030724588205171974,
+      "grad_norm": 0.7623032565109146,
+      "learning_rate": 5.119453924914676e-06,
+      "loss": 0.6035,
+      "step": 180
+    },
+    {
+      "epoch": 0.03157804898864897,
+      "grad_norm": 0.9433218529441518,
+      "learning_rate": 5.261660978384528e-06,
+      "loss": 0.6372,
+      "step": 185
+    },
+    {
+      "epoch": 0.03243150977212597,
+      "grad_norm": 0.6318034752486974,
+      "learning_rate": 5.403868031854381e-06,
+      "loss": 0.6517,
+      "step": 190
+    },
+    {
+      "epoch": 0.03328497055560297,
+      "grad_norm": 0.7984227582107997,
+      "learning_rate": 5.546075085324233e-06,
+      "loss": 0.6507,
+      "step": 195
+    },
+    {
+      "epoch": 0.03413843133907997,
+      "grad_norm": 0.6970637034010194,
+      "learning_rate": 5.6882821387940845e-06,
+      "loss": 0.6132,
+      "step": 200
+    },
+    {
+      "epoch": 0.03499189212255697,
+      "grad_norm": 0.693661339636327,
+      "learning_rate": 5.830489192263936e-06,
+      "loss": 0.6447,
+      "step": 205
+    },
+    {
+      "epoch": 0.03584535290603397,
+      "grad_norm": 0.5750112660425444,
+      "learning_rate": 5.972696245733789e-06,
+      "loss": 0.6354,
+      "step": 210
+    },
+    {
+      "epoch": 0.036698813689510966,
+      "grad_norm": 0.7707644501420617,
+      "learning_rate": 6.114903299203641e-06,
+      "loss": 0.613,
+      "step": 215
+    },
+    {
+      "epoch": 0.037552274472987965,
+      "grad_norm": 0.7610886999828581,
+      "learning_rate": 6.257110352673493e-06,
+      "loss": 0.5985,
+      "step": 220
+    },
+    {
+      "epoch": 0.038405735256464964,
+      "grad_norm": 0.7219024880249069,
+      "learning_rate": 6.399317406143345e-06,
+      "loss": 0.6147,
+      "step": 225
+    },
+    {
+      "epoch": 0.03925919603994196,
+      "grad_norm": 0.694773090873567,
+      "learning_rate": 6.5415244596131975e-06,
+      "loss": 0.6138,
+      "step": 230
+    },
+    {
+      "epoch": 0.04011265682341896,
+      "grad_norm": 0.6507600820281264,
+      "learning_rate": 6.683731513083049e-06,
+      "loss": 0.6358,
+      "step": 235
+    },
+    {
+      "epoch": 0.04096611760689596,
+      "grad_norm": 0.6821754118235962,
+      "learning_rate": 6.825938566552901e-06,
+      "loss": 0.6064,
+      "step": 240
+    },
+    {
+      "epoch": 0.04181957839037296,
+      "grad_norm": 0.6685037872512465,
+      "learning_rate": 6.968145620022753e-06,
+      "loss": 0.6141,
+      "step": 245
+    },
+    {
+      "epoch": 0.04267303917384996,
+      "grad_norm": 0.650855824979963,
+      "learning_rate": 7.110352673492605e-06,
+      "loss": 0.6099,
+      "step": 250
+    },
+    {
+      "epoch": 0.043526499957326964,
+      "grad_norm": 0.6283039075662383,
+      "learning_rate": 7.252559726962458e-06,
+      "loss": 0.6487,
+      "step": 255
+    },
+    {
+      "epoch": 0.04437996074080396,
+      "grad_norm": 0.7017487578081022,
+      "learning_rate": 7.39476678043231e-06,
+      "loss": 0.6117,
+      "step": 260
+    },
+    {
+      "epoch": 0.04523342152428096,
+      "grad_norm": 0.6021311846646746,
+      "learning_rate": 7.536973833902162e-06,
+      "loss": 0.6074,
+      "step": 265
+    },
+    {
+      "epoch": 0.04608688230775796,
+      "grad_norm": 0.705303126743963,
+      "learning_rate": 7.679180887372015e-06,
+      "loss": 0.6228,
+      "step": 270
+    },
+    {
+      "epoch": 0.04694034309123496,
+      "grad_norm": 0.6331398845315224,
+      "learning_rate": 7.821387940841867e-06,
+      "loss": 0.6186,
+      "step": 275
+    },
+    {
+      "epoch": 0.04779380387471196,
+      "grad_norm": 0.7364757969311059,
+      "learning_rate": 7.963594994311719e-06,
+      "loss": 0.6028,
+      "step": 280
+    },
+    {
+      "epoch": 0.04864726465818896,
+      "grad_norm": 0.5864444756610997,
+      "learning_rate": 8.10580204778157e-06,
+      "loss": 0.5572,
+      "step": 285
+    },
+    {
+      "epoch": 0.049500725441665956,
+      "grad_norm": 0.7188508735007383,
+      "learning_rate": 8.248009101251423e-06,
+      "loss": 0.5983,
+      "step": 290
+    },
+    {
+      "epoch": 0.050354186225142955,
+      "grad_norm": 0.5871960228317447,
+      "learning_rate": 8.390216154721274e-06,
+      "loss": 0.5697,
+      "step": 295
+    },
+    {
+      "epoch": 0.051207647008619954,
+      "grad_norm": 0.6415281272301155,
+      "learning_rate": 8.532423208191128e-06,
+      "loss": 0.5669,
+      "step": 300
+    },
+    {
+      "epoch": 0.05206110779209695,
+      "grad_norm": 0.7422606435155473,
+      "learning_rate": 8.67463026166098e-06,
+      "loss": 0.6099,
+      "step": 305
+    },
+    {
+      "epoch": 0.05291456857557395,
+      "grad_norm": 0.8031320770409729,
+      "learning_rate": 8.816837315130832e-06,
+      "loss": 0.5805,
+      "step": 310
+    },
+    {
+      "epoch": 0.05376802935905095,
+      "grad_norm": 0.6235161179887372,
+      "learning_rate": 8.959044368600684e-06,
+      "loss": 0.5829,
+      "step": 315
+    },
+    {
+      "epoch": 0.05462149014252795,
+      "grad_norm": 0.6368377190363229,
+      "learning_rate": 9.101251422070536e-06,
+      "loss": 0.586,
+      "step": 320
+    },
+    {
+      "epoch": 0.05547495092600495,
+      "grad_norm": 0.7355127518957415,
+      "learning_rate": 9.243458475540387e-06,
+      "loss": 0.6052,
+      "step": 325
+    },
+    {
+      "epoch": 0.05632841170948195,
+      "grad_norm": 0.6491383433315371,
+      "learning_rate": 9.38566552901024e-06,
+      "loss": 0.5641,
+      "step": 330
+    },
+    {
+      "epoch": 0.057181872492958946,
+      "grad_norm": 0.6074404231611151,
+      "learning_rate": 9.527872582480093e-06,
+      "loss": 0.5805,
+      "step": 335
+    },
+    {
+      "epoch": 0.058035333276435945,
+      "grad_norm": 0.621909618300053,
+      "learning_rate": 9.670079635949945e-06,
+      "loss": 0.5921,
+      "step": 340
+    },
+    {
+      "epoch": 0.058888794059912944,
+      "grad_norm": 0.7596871354310805,
+      "learning_rate": 9.812286689419797e-06,
+      "loss": 0.5905,
+      "step": 345
+    },
+    {
+      "epoch": 0.05974225484338994,
+      "grad_norm": 0.6852465582192806,
+      "learning_rate": 9.954493742889649e-06,
+      "loss": 0.5718,
+      "step": 350
+    },
+    {
+      "epoch": 0.06059571562686695,
+      "grad_norm": 0.6323485065460291,
+      "learning_rate": 1.00967007963595e-05,
+      "loss": 0.6071,
+      "step": 355
+    },
+    {
+      "epoch": 0.06144917641034395,
+      "grad_norm": 0.702431722756456,
+      "learning_rate": 1.0238907849829352e-05,
+      "loss": 0.5612,
+      "step": 360
+    },
+    {
+      "epoch": 0.062302637193820946,
+      "grad_norm": 0.7446033311458136,
+      "learning_rate": 1.0381114903299204e-05,
+      "loss": 0.595,
+      "step": 365
+    },
+    {
+      "epoch": 0.06315609797729795,
+      "grad_norm": 0.6942761058832195,
+      "learning_rate": 1.0523321956769056e-05,
+      "loss": 0.5742,
+      "step": 370
+    },
+    {
+      "epoch": 0.06400955876077494,
+      "grad_norm": 0.7488476031298855,
+      "learning_rate": 1.066552901023891e-05,
+      "loss": 0.5651,
+      "step": 375
+    },
+    {
+      "epoch": 0.06486301954425194,
+      "grad_norm": 0.7465993777161892,
+      "learning_rate": 1.0807736063708762e-05,
+      "loss": 0.5731,
+      "step": 380
+    },
+    {
+      "epoch": 0.06571648032772893,
+      "grad_norm": 0.6785228215376499,
+      "learning_rate": 1.0949943117178613e-05,
+      "loss": 0.58,
+      "step": 385
+    },
+    {
+      "epoch": 0.06656994111120594,
+      "grad_norm": 0.6009750474296774,
+      "learning_rate": 1.1092150170648465e-05,
+      "loss": 0.6299,
+      "step": 390
+    },
+    {
+      "epoch": 0.06742340189468293,
+      "grad_norm": 0.6112998501665518,
+      "learning_rate": 1.1234357224118317e-05,
+      "loss": 0.5448,
+      "step": 395
+    },
+    {
+      "epoch": 0.06827686267815994,
+      "grad_norm": 0.6141285561163018,
+      "learning_rate": 1.1376564277588169e-05,
+      "loss": 0.5536,
+      "step": 400
+    },
+    {
+      "epoch": 0.06913032346163694,
+      "grad_norm": 0.7039674370569571,
+      "learning_rate": 1.1518771331058021e-05,
+      "loss": 0.5846,
+      "step": 405
+    },
+    {
+      "epoch": 0.06998378424511394,
+      "grad_norm": 0.6628719873901499,
+      "learning_rate": 1.1660978384527873e-05,
+      "loss": 0.5996,
+      "step": 410
+    },
+    {
+      "epoch": 0.07083724502859094,
+      "grad_norm": 0.733225603137678,
+      "learning_rate": 1.1803185437997725e-05,
+      "loss": 0.5604,
+      "step": 415
+    },
+    {
+      "epoch": 0.07169070581206793,
+      "grad_norm": 0.7999324830307888,
+      "learning_rate": 1.1945392491467578e-05,
+      "loss": 0.5845,
+      "step": 420
+    },
+    {
+      "epoch": 0.07254416659554494,
+      "grad_norm": 0.6696390387855465,
+      "learning_rate": 1.208759954493743e-05,
+      "loss": 0.5663,
+      "step": 425
+    },
+    {
+      "epoch": 0.07339762737902193,
+      "grad_norm": 0.6495085711609115,
+      "learning_rate": 1.2229806598407282e-05,
+      "loss": 0.5823,
+      "step": 430
+    },
+    {
+      "epoch": 0.07425108816249894,
+      "grad_norm": 0.6508779382883285,
+      "learning_rate": 1.2372013651877134e-05,
+      "loss": 0.5692,
+      "step": 435
+    },
+    {
+      "epoch": 0.07510454894597593,
+      "grad_norm": 0.677992969506062,
+      "learning_rate": 1.2514220705346986e-05,
+      "loss": 0.5867,
+      "step": 440
+    },
+    {
+      "epoch": 0.07595800972945294,
+      "grad_norm": 0.6481542235806955,
+      "learning_rate": 1.2656427758816838e-05,
+      "loss": 0.5458,
+      "step": 445
+    },
+    {
+      "epoch": 0.07681147051292993,
+      "grad_norm": 0.6507351226676409,
+      "learning_rate": 1.279863481228669e-05,
+      "loss": 0.542,
+      "step": 450
+    },
+    {
+      "epoch": 0.07766493129640693,
+      "grad_norm": 0.6520426456185812,
+      "learning_rate": 1.2940841865756541e-05,
+      "loss": 0.561,
+      "step": 455
+    },
+    {
+      "epoch": 0.07851839207988393,
+      "grad_norm": 0.7153844095429435,
+      "learning_rate": 1.3083048919226395e-05,
+      "loss": 0.5792,
+      "step": 460
+    },
+    {
+      "epoch": 0.07937185286336093,
+      "grad_norm": 0.9169128648612137,
+      "learning_rate": 1.3225255972696247e-05,
+      "loss": 0.5663,
+      "step": 465
+    },
+    {
+      "epoch": 0.08022531364683792,
+      "grad_norm": 0.6346143572448639,
+      "learning_rate": 1.3367463026166099e-05,
+      "loss": 0.5454,
+      "step": 470
+    },
+    {
+      "epoch": 0.08107877443031493,
+      "grad_norm": 0.661156267249212,
+      "learning_rate": 1.350967007963595e-05,
+      "loss": 0.5284,
+      "step": 475
+    },
+    {
+      "epoch": 0.08193223521379192,
+      "grad_norm": 0.6570366689901866,
+      "learning_rate": 1.3651877133105803e-05,
+      "loss": 0.583,
+      "step": 480
+    },
+    {
+      "epoch": 0.08278569599726893,
+      "grad_norm": 0.6925487929822411,
+      "learning_rate": 1.3794084186575654e-05,
+      "loss": 0.5734,
+      "step": 485
+    },
+    {
+      "epoch": 0.08363915678074592,
+      "grad_norm": 0.6825610761421598,
+      "learning_rate": 1.3936291240045506e-05,
+      "loss": 0.5604,
+      "step": 490
+    },
+    {
+      "epoch": 0.08449261756422292,
+      "grad_norm": 0.6455121109752382,
+      "learning_rate": 1.407849829351536e-05,
+      "loss": 0.5577,
+      "step": 495
+    },
+    {
+      "epoch": 0.08534607834769992,
+      "grad_norm": 0.7057436426948772,
+      "learning_rate": 1.422070534698521e-05,
+      "loss": 0.5635,
+      "step": 500
+    },
+    {
+      "epoch": 0.08619953913117692,
+      "grad_norm": 0.6035482154147205,
+      "learning_rate": 1.4362912400455064e-05,
+      "loss": 0.5608,
+      "step": 505
+    },
+    {
+      "epoch": 0.08705299991465393,
+      "grad_norm": 0.6945499780138561,
+      "learning_rate": 1.4505119453924915e-05,
+      "loss": 0.5734,
+      "step": 510
+    },
+    {
+      "epoch": 0.08790646069813092,
+      "grad_norm": 0.6342408596975896,
+      "learning_rate": 1.4647326507394767e-05,
+      "loss": 0.5478,
+      "step": 515
+    },
+    {
+      "epoch": 0.08875992148160793,
+      "grad_norm": 0.669235415531277,
+      "learning_rate": 1.478953356086462e-05,
+      "loss": 0.563,
+      "step": 520
+    },
+    {
+      "epoch": 0.08961338226508492,
+      "grad_norm": 0.8846650505038849,
+      "learning_rate": 1.4931740614334471e-05,
+      "loss": 0.5926,
+      "step": 525
+    },
+    {
+      "epoch": 0.09046684304856192,
+      "grad_norm": 0.7901893564028634,
+      "learning_rate": 1.5073947667804325e-05,
+      "loss": 0.5368,
+      "step": 530
+    },
+    {
+      "epoch": 0.09132030383203892,
+      "grad_norm": 0.7257391495825279,
+      "learning_rate": 1.5216154721274175e-05,
+      "loss": 0.5116,
+      "step": 535
+    },
+    {
+      "epoch": 0.09217376461551592,
+      "grad_norm": 0.8055262915442394,
+      "learning_rate": 1.535836177474403e-05,
+      "loss": 0.5737,
+      "step": 540
+    },
+    {
+      "epoch": 0.09302722539899291,
+      "grad_norm": 0.7724777138200022,
+      "learning_rate": 1.550056882821388e-05,
+      "loss": 0.5531,
+      "step": 545
+    },
+    {
+      "epoch": 0.09388068618246992,
+      "grad_norm": 0.7277006149678841,
+      "learning_rate": 1.5642775881683734e-05,
+      "loss": 0.5399,
+      "step": 550
+    },
+    {
+      "epoch": 0.09473414696594691,
+      "grad_norm": 1.0778652880491968,
+      "learning_rate": 1.5784982935153582e-05,
+      "loss": 0.5721,
+      "step": 555
+    },
+    {
+      "epoch": 0.09558760774942392,
+      "grad_norm": 0.9104304283180978,
+      "learning_rate": 1.5927189988623438e-05,
+      "loss": 0.5572,
+      "step": 560
+    },
+    {
+      "epoch": 0.09644106853290091,
+      "grad_norm": 0.8149446361982572,
+      "learning_rate": 1.606939704209329e-05,
+      "loss": 0.534,
+      "step": 565
+    },
+    {
+      "epoch": 0.09729452931637791,
+      "grad_norm": 0.6508255643810232,
+      "learning_rate": 1.621160409556314e-05,
+      "loss": 0.5901,
+      "step": 570
+    },
+    {
+      "epoch": 0.0981479900998549,
+      "grad_norm": 0.7908571035291143,
+      "learning_rate": 1.6353811149032993e-05,
+      "loss": 0.5345,
+      "step": 575
+    },
+    {
+      "epoch": 0.09900145088333191,
+      "grad_norm": 0.8711338798337782,
+      "learning_rate": 1.6496018202502845e-05,
+      "loss": 0.5439,
+      "step": 580
+    },
+    {
+      "epoch": 0.0998549116668089,
+      "grad_norm": 0.8510864530724201,
+      "learning_rate": 1.6638225255972697e-05,
+      "loss": 0.5346,
+      "step": 585
+    },
+    {
+      "epoch": 0.10070837245028591,
+      "grad_norm": 0.6004623000077266,
+      "learning_rate": 1.678043230944255e-05,
+      "loss": 0.5671,
+      "step": 590
+    },
+    {
+      "epoch": 0.1015618332337629,
+      "grad_norm": 0.7836797955774423,
+      "learning_rate": 1.69226393629124e-05,
+      "loss": 0.5728,
+      "step": 595
+    },
+    {
+      "epoch": 0.10241529401723991,
+      "grad_norm": 0.7164324569628334,
+      "learning_rate": 1.7064846416382256e-05,
+      "loss": 0.5808,
+      "step": 600
+    },
+    {
+      "epoch": 0.10326875480071691,
+      "grad_norm": 0.658437507959372,
+      "learning_rate": 1.7207053469852105e-05,
+      "loss": 0.527,
+      "step": 605
+    },
+    {
+      "epoch": 0.1041222155841939,
+      "grad_norm": 0.648235707862094,
+      "learning_rate": 1.734926052332196e-05,
+      "loss": 0.5042,
+      "step": 610
+    },
+    {
+      "epoch": 0.10497567636767091,
+      "grad_norm": 0.7066278650620545,
+      "learning_rate": 1.749146757679181e-05,
+      "loss": 0.5484,
+      "step": 615
+    },
+    {
+      "epoch": 0.1058291371511479,
+      "grad_norm": 0.7306436553829413,
+      "learning_rate": 1.7633674630261664e-05,
+      "loss": 0.5233,
+      "step": 620
+    },
+    {
+      "epoch": 0.10668259793462491,
+      "grad_norm": 0.5940294914233756,
+      "learning_rate": 1.7775881683731512e-05,
+      "loss": 0.5497,
+      "step": 625
+    },
+    {
+      "epoch": 0.1075360587181019,
+      "grad_norm": 0.6766881884349177,
+      "learning_rate": 1.7918088737201367e-05,
+      "loss": 0.524,
+      "step": 630
+    },
+    {
+      "epoch": 0.10838951950157891,
+      "grad_norm": 0.6374323634721849,
+      "learning_rate": 1.806029579067122e-05,
+      "loss": 0.5576,
+      "step": 635
+    },
+    {
+      "epoch": 0.1092429802850559,
+      "grad_norm": 0.7284674831221161,
+      "learning_rate": 1.820250284414107e-05,
+      "loss": 0.5529,
+      "step": 640
+    },
+    {
+      "epoch": 0.1100964410685329,
+      "grad_norm": 0.6615725114492135,
+      "learning_rate": 1.8344709897610923e-05,
+      "loss": 0.5044,
+      "step": 645
+    },
+    {
+      "epoch": 0.1109499018520099,
+      "grad_norm": 0.7676539593033215,
+      "learning_rate": 1.8486916951080775e-05,
+      "loss": 0.5512,
+      "step": 650
+    },
+    {
+      "epoch": 0.1118033626354869,
+      "grad_norm": 0.5885868256001245,
+      "learning_rate": 1.8629124004550627e-05,
+      "loss": 0.5001,
+      "step": 655
+    },
+    {
+      "epoch": 0.1126568234189639,
+      "grad_norm": 0.5686249348694413,
+      "learning_rate": 1.877133105802048e-05,
+      "loss": 0.5332,
+      "step": 660
+    },
+    {
+      "epoch": 0.1135102842024409,
+      "grad_norm": 0.6097676324328544,
+      "learning_rate": 1.891353811149033e-05,
+      "loss": 0.5374,
+      "step": 665
+    },
+    {
+      "epoch": 0.11436374498591789,
+      "grad_norm": 0.7494192367995726,
+      "learning_rate": 1.9055745164960186e-05,
+      "loss": 0.5428,
+      "step": 670
+    },
+    {
+      "epoch": 0.1152172057693949,
+      "grad_norm": 0.6181360753587957,
+      "learning_rate": 1.9197952218430034e-05,
+      "loss": 0.4983,
+      "step": 675
+    },
+    {
+      "epoch": 0.11607066655287189,
+      "grad_norm": 0.7148707525002317,
+      "learning_rate": 1.934015927189989e-05,
+      "loss": 0.5474,
+      "step": 680
+    },
+    {
+      "epoch": 0.1169241273363489,
+      "grad_norm": 0.6890678940899136,
+      "learning_rate": 1.9482366325369738e-05,
+      "loss": 0.539,
+      "step": 685
+    },
+    {
+      "epoch": 0.11777758811982589,
+      "grad_norm": 1.5520380828368654,
+      "learning_rate": 1.9624573378839593e-05,
+      "loss": 0.5304,
+      "step": 690
+    },
+    {
+      "epoch": 0.1186310489033029,
+      "grad_norm": 0.6879630723852271,
+      "learning_rate": 1.9766780432309442e-05,
+      "loss": 0.5523,
+      "step": 695
+    },
+    {
+      "epoch": 0.11948450968677989,
+      "grad_norm": 0.5952074420310725,
+      "learning_rate": 1.9908987485779297e-05,
+      "loss": 0.546,
+      "step": 700
+    },
+    {
+      "epoch": 0.12033797047025689,
+      "grad_norm": 0.6568607864977468,
+      "learning_rate": 2.005119453924915e-05,
+      "loss": 0.535,
+      "step": 705
+    },
+    {
+      "epoch": 0.1211914312537339,
+      "grad_norm": 0.7573004872488551,
+      "learning_rate": 2.0193401592719e-05,
+      "loss": 0.5698,
+      "step": 710
+    },
+    {
+      "epoch": 0.12204489203721089,
+      "grad_norm": 0.6038232065159463,
+      "learning_rate": 2.0335608646188853e-05,
+      "loss": 0.5355,
+      "step": 715
+    },
+    {
+      "epoch": 0.1228983528206879,
+      "grad_norm": 0.8905543409020308,
+      "learning_rate": 2.0477815699658705e-05,
+      "loss": 0.5501,
+      "step": 720
+    },
+    {
+      "epoch": 0.12375181360416489,
+      "grad_norm": 0.6494164955676679,
+      "learning_rate": 2.0620022753128557e-05,
+      "loss": 0.597,
+      "step": 725
+    },
+    {
+      "epoch": 0.12460527438764189,
+      "grad_norm": 0.6068419377405997,
+      "learning_rate": 2.076222980659841e-05,
+      "loss": 0.5191,
+      "step": 730
+    },
+    {
+      "epoch": 0.12545873517111888,
+      "grad_norm": 0.6748891771108966,
+      "learning_rate": 2.090443686006826e-05,
+      "loss": 0.5624,
+      "step": 735
+    },
+    {
+      "epoch": 0.1263121959545959,
+      "grad_norm": 0.6003241284298655,
+      "learning_rate": 2.1046643913538112e-05,
+      "loss": 0.5199,
+      "step": 740
+    },
+    {
+      "epoch": 0.1271656567380729,
+      "grad_norm": 0.993073185386471,
+      "learning_rate": 2.1188850967007964e-05,
+      "loss": 0.5379,
+      "step": 745
+    },
+    {
+      "epoch": 0.12801911752154987,
+      "grad_norm": 0.6830807878540766,
+      "learning_rate": 2.133105802047782e-05,
+      "loss": 0.5718,
+      "step": 750
+    },
+    {
+      "epoch": 0.12887257830502688,
+      "grad_norm": 0.5637706434424037,
+      "learning_rate": 2.1473265073947668e-05,
+      "loss": 0.5329,
+      "step": 755
+    },
+    {
+      "epoch": 0.12972603908850389,
+      "grad_norm": 0.5901909162808682,
+      "learning_rate": 2.1615472127417523e-05,
+      "loss": 0.5005,
+      "step": 760
+    },
+    {
+      "epoch": 0.1305794998719809,
+      "grad_norm": 0.6384938080969448,
+      "learning_rate": 2.175767918088737e-05,
+      "loss": 0.5081,
+      "step": 765
+    },
+    {
+      "epoch": 0.13143296065545787,
+      "grad_norm": 0.6515001609818502,
+      "learning_rate": 2.1899886234357227e-05,
+      "loss": 0.5139,
+      "step": 770
+    },
+    {
+      "epoch": 0.13228642143893488,
+      "grad_norm": 0.6594065584484019,
+      "learning_rate": 2.204209328782708e-05,
+      "loss": 0.5584,
+      "step": 775
+    },
+    {
+      "epoch": 0.13313988222241188,
+      "grad_norm": 0.5788918184456333,
+      "learning_rate": 2.218430034129693e-05,
+      "loss": 0.4872,
+      "step": 780
+    },
+    {
+      "epoch": 0.1339933430058889,
+      "grad_norm": 0.7132592163471578,
+      "learning_rate": 2.2326507394766782e-05,
+      "loss": 0.526,
+      "step": 785
+    },
+    {
+      "epoch": 0.13484680378936587,
+      "grad_norm": 0.6701954948574094,
+      "learning_rate": 2.2468714448236634e-05,
+      "loss": 0.5369,
+      "step": 790
+    },
+    {
+      "epoch": 0.13570026457284287,
+      "grad_norm": 0.664429218741904,
+      "learning_rate": 2.2610921501706486e-05,
+      "loss": 0.5622,
+      "step": 795
+    },
+    {
+      "epoch": 0.13655372535631988,
+      "grad_norm": 0.7401195985099817,
+      "learning_rate": 2.2753128555176338e-05,
+      "loss": 0.5276,
+      "step": 800
+    },
+    {
+      "epoch": 0.13740718613979688,
+      "grad_norm": 0.5761092021531733,
+      "learning_rate": 2.289533560864619e-05,
+      "loss": 0.5003,
+      "step": 805
+    },
+    {
+      "epoch": 0.1382606469232739,
+      "grad_norm": 0.7211376155782887,
+      "learning_rate": 2.3037542662116042e-05,
+      "loss": 0.5346,
+      "step": 810
+    },
+    {
+      "epoch": 0.13911410770675087,
+      "grad_norm": 0.6870353313728967,
+      "learning_rate": 2.3179749715585894e-05,
+      "loss": 0.5412,
+      "step": 815
+    },
+    {
+      "epoch": 0.13996756849022787,
+      "grad_norm": 0.7484951113803281,
+      "learning_rate": 2.3321956769055746e-05,
+      "loss": 0.5352,
+      "step": 820
+    },
+    {
+      "epoch": 0.14082102927370488,
+      "grad_norm": 0.5633549500567068,
+      "learning_rate": 2.3464163822525598e-05,
+      "loss": 0.5373,
+      "step": 825
+    },
+    {
+      "epoch": 0.14167449005718188,
+      "grad_norm": 0.6175161903456321,
+      "learning_rate": 2.360637087599545e-05,
+      "loss": 0.5293,
+      "step": 830
+    },
+    {
+      "epoch": 0.14252795084065886,
+      "grad_norm": 0.5286458735194771,
+      "learning_rate": 2.37485779294653e-05,
+      "loss": 0.4924,
+      "step": 835
+    },
+    {
+      "epoch": 0.14338141162413587,
+      "grad_norm": 0.809888612811748,
+      "learning_rate": 2.3890784982935157e-05,
+      "loss": 0.5476,
+      "step": 840
+    },
+    {
+      "epoch": 0.14423487240761287,
+      "grad_norm": 0.6320119433876821,
+      "learning_rate": 2.403299203640501e-05,
+      "loss": 0.5175,
+      "step": 845
+    },
+    {
+      "epoch": 0.14508833319108988,
+      "grad_norm": 0.7042156271509452,
+      "learning_rate": 2.417519908987486e-05,
+      "loss": 0.5225,
+      "step": 850
+    },
+    {
+      "epoch": 0.14594179397456686,
+      "grad_norm": 0.6630329659099481,
+      "learning_rate": 2.4317406143344712e-05,
+      "loss": 0.5219,
+      "step": 855
+    },
+    {
+      "epoch": 0.14679525475804386,
+      "grad_norm": 0.8047566311916006,
+      "learning_rate": 2.4459613196814564e-05,
+      "loss": 0.5282,
+      "step": 860
+    },
+    {
+      "epoch": 0.14764871554152087,
+      "grad_norm": 0.6322398706113659,
+      "learning_rate": 2.4601820250284416e-05,
+      "loss": 0.5496,
+      "step": 865
+    },
+    {
+      "epoch": 0.14850217632499788,
+      "grad_norm": 0.5394073805212835,
+      "learning_rate": 2.4744027303754268e-05,
+      "loss": 0.5391,
+      "step": 870
+    },
+    {
+      "epoch": 0.14935563710847485,
+      "grad_norm": 0.8539124043452883,
+      "learning_rate": 2.488623435722412e-05,
+      "loss": 0.5615,
+      "step": 875
+    },
+    {
+      "epoch": 0.15020909789195186,
+      "grad_norm": 0.5973992591932995,
+      "learning_rate": 2.502844141069397e-05,
+      "loss": 0.5109,
+      "step": 880
+    },
+    {
+      "epoch": 0.15106255867542887,
+      "grad_norm": 0.7868095705273332,
+      "learning_rate": 2.5170648464163827e-05,
+      "loss": 0.4925,
+      "step": 885
+    },
+    {
+      "epoch": 0.15191601945890587,
+      "grad_norm": 0.5303131516550371,
+      "learning_rate": 2.5312855517633675e-05,
+      "loss": 0.5223,
+      "step": 890
+    },
+    {
+      "epoch": 0.15276948024238285,
+      "grad_norm": 0.6915392576850998,
+      "learning_rate": 2.5455062571103527e-05,
+      "loss": 0.5199,
+      "step": 895
+    },
+    {
+      "epoch": 0.15362294102585985,
+      "grad_norm": 0.6501693825506315,
+      "learning_rate": 2.559726962457338e-05,
+      "loss": 0.5196,
+      "step": 900
+    },
+    {
+      "epoch": 0.15447640180933686,
+      "grad_norm": 0.8308211131956251,
+      "learning_rate": 2.5739476678043234e-05,
+      "loss": 0.5797,
+      "step": 905
+    },
+    {
+      "epoch": 0.15532986259281387,
+      "grad_norm": 0.6205667096580658,
+      "learning_rate": 2.5881683731513083e-05,
+      "loss": 0.5626,
+      "step": 910
+    },
+    {
+      "epoch": 0.15618332337629087,
+      "grad_norm": 0.78307958464136,
+      "learning_rate": 2.6023890784982935e-05,
+      "loss": 0.5572,
+      "step": 915
+    },
+    {
+      "epoch": 0.15703678415976785,
+      "grad_norm": 0.6971854241217064,
+      "learning_rate": 2.616609783845279e-05,
+      "loss": 0.5283,
+      "step": 920
+    },
+    {
+      "epoch": 0.15789024494324486,
+      "grad_norm": 0.790435113721265,
+      "learning_rate": 2.6308304891922642e-05,
+      "loss": 0.5096,
+      "step": 925
+    },
+    {
+      "epoch": 0.15874370572672186,
+      "grad_norm": 0.6263290150135153,
+      "learning_rate": 2.6450511945392494e-05,
+      "loss": 0.511,
+      "step": 930
+    },
+    {
+      "epoch": 0.15959716651019887,
+      "grad_norm": 1.0182762203606528,
+      "learning_rate": 2.6592718998862342e-05,
+      "loss": 0.5176,
+      "step": 935
+    },
+    {
+      "epoch": 0.16045062729367585,
+      "grad_norm": 0.6764980698940198,
+      "learning_rate": 2.6734926052332198e-05,
+      "loss": 0.532,
+      "step": 940
+    },
+    {
+      "epoch": 0.16130408807715285,
+      "grad_norm": 0.617066330968239,
+      "learning_rate": 2.687713310580205e-05,
+      "loss": 0.5215,
+      "step": 945
+    },
+    {
+      "epoch": 0.16215754886062986,
+      "grad_norm": 0.6650618699860503,
+      "learning_rate": 2.70193401592719e-05,
+      "loss": 0.5173,
+      "step": 950
+    },
+    {
+      "epoch": 0.16301100964410686,
+      "grad_norm": 0.6880915197162699,
+      "learning_rate": 2.7161547212741757e-05,
+      "loss": 0.5183,
+      "step": 955
+    },
+    {
+      "epoch": 0.16386447042758384,
+      "grad_norm": 0.6334400197912421,
+      "learning_rate": 2.7303754266211605e-05,
+      "loss": 0.5068,
+      "step": 960
+    },
+    {
+      "epoch": 0.16471793121106085,
+      "grad_norm": 0.6614924655158634,
+      "learning_rate": 2.7445961319681457e-05,
+      "loss": 0.5034,
+      "step": 965
+    },
+    {
+      "epoch": 0.16557139199453785,
+      "grad_norm": 0.6410739649524626,
+      "learning_rate": 2.758816837315131e-05,
+      "loss": 0.543,
+      "step": 970
+    },
+    {
+      "epoch": 0.16642485277801486,
+      "grad_norm": 0.6153025614066233,
+      "learning_rate": 2.7730375426621164e-05,
+      "loss": 0.5204,
+      "step": 975
+    },
+    {
+      "epoch": 0.16727831356149184,
+      "grad_norm": 0.5942921081010195,
+      "learning_rate": 2.7872582480091013e-05,
+      "loss": 0.5047,
+      "step": 980
+    },
+    {
+      "epoch": 0.16813177434496884,
+      "grad_norm": 0.6022254575923327,
+      "learning_rate": 2.8014789533560864e-05,
+      "loss": 0.5077,
+      "step": 985
+    },
+    {
+      "epoch": 0.16898523512844585,
+      "grad_norm": 0.7132287202025721,
+      "learning_rate": 2.815699658703072e-05,
+      "loss": 0.5722,
+      "step": 990
+    },
+    {
+      "epoch": 0.16983869591192285,
+      "grad_norm": 1.1177874785166635,
+      "learning_rate": 2.829920364050057e-05,
+      "loss": 0.5189,
+      "step": 995
+    },
+    {
+      "epoch": 0.17069215669539983,
+      "grad_norm": 1.193082385576556,
+      "learning_rate": 2.844141069397042e-05,
+      "loss": 0.5347,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17154561747887684,
+      "grad_norm": 0.9273224155143907,
+      "learning_rate": 2.8583617747440272e-05,
+      "loss": 0.5199,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17239907826235384,
+      "grad_norm": 0.8009696728529175,
+      "learning_rate": 2.8725824800910127e-05,
+      "loss": 0.5413,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17325253904583085,
+      "grad_norm": 0.6389356197858123,
+      "learning_rate": 2.886803185437998e-05,
+      "loss": 0.5197,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17410599982930786,
+      "grad_norm": 0.5643611328367691,
+      "learning_rate": 2.901023890784983e-05,
+      "loss": 0.5416,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17495946061278483,
+      "grad_norm": 0.8203413707371002,
+      "learning_rate": 2.9152445961319686e-05,
+      "loss": 0.5517,
+      "step": 1025
+    },
+    {
+      "epoch": 0.17581292139626184,
+      "grad_norm": 0.5767388331401231,
+      "learning_rate": 2.9294653014789535e-05,
+      "loss": 0.5281,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17666638217973885,
+      "grad_norm": 0.7639414252803709,
+      "learning_rate": 2.9436860068259387e-05,
+      "loss": 0.5117,
+      "step": 1035
+    },
+    {
+      "epoch": 0.17751984296321585,
+      "grad_norm": 0.7435837191495894,
+      "learning_rate": 2.957906712172924e-05,
+      "loss": 0.5174,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17837330374669283,
+      "grad_norm": 0.7406989837738298,
+      "learning_rate": 2.9721274175199094e-05,
+      "loss": 0.5564,
+      "step": 1045
+    },
+    {
+      "epoch": 0.17922676453016984,
+      "grad_norm": 0.7609524854068137,
+      "learning_rate": 2.9863481228668942e-05,
+      "loss": 0.5393,
+      "step": 1050
+    },
+    {
+      "epoch": 0.18008022531364684,
+      "grad_norm": 0.7602059635412194,
+      "learning_rate": 3.0005688282138794e-05,
+      "loss": 0.5085,
+      "step": 1055
+    },
+    {
+      "epoch": 0.18093368609712385,
+      "grad_norm": 0.6041864111887089,
+      "learning_rate": 3.014789533560865e-05,
+      "loss": 0.5022,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18178714688060083,
+      "grad_norm": 0.6725193484426399,
+      "learning_rate": 3.02901023890785e-05,
+      "loss": 0.5255,
+      "step": 1065
+    },
+    {
+      "epoch": 0.18264060766407783,
+      "grad_norm": 0.5809953323844759,
+      "learning_rate": 3.043230944254835e-05,
+      "loss": 0.5222,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18349406844755484,
+      "grad_norm": 0.5561450815754222,
+      "learning_rate": 3.0574516496018205e-05,
+      "loss": 0.497,
+      "step": 1075
+    },
+    {
+      "epoch": 0.18434752923103184,
+      "grad_norm": 0.7384676137709544,
+      "learning_rate": 3.071672354948806e-05,
+      "loss": 0.5228,
+      "step": 1080
+    },
+    {
+      "epoch": 0.18520099001450882,
+      "grad_norm": 0.5611330862106643,
+      "learning_rate": 3.085893060295791e-05,
+      "loss": 0.5278,
+      "step": 1085
+    },
+    {
+      "epoch": 0.18605445079798583,
+      "grad_norm": 0.9520323231067833,
+      "learning_rate": 3.100113765642776e-05,
+      "loss": 0.5224,
+      "step": 1090
+    },
+    {
+      "epoch": 0.18690791158146283,
+      "grad_norm": 0.7633270460301356,
+      "learning_rate": 3.114334470989761e-05,
+      "loss": 0.5234,
+      "step": 1095
+    },
+    {
+      "epoch": 0.18776137236493984,
+      "grad_norm": 1.8224752677288898,
+      "learning_rate": 3.128555176336747e-05,
+      "loss": 0.5076,
+      "step": 1100
+    },
+    {
+      "epoch": 0.18861483314841684,
+      "grad_norm": 0.6362222349727383,
+      "learning_rate": 3.1427758816837316e-05,
+      "loss": 0.5458,
+      "step": 1105
+    },
+    {
+      "epoch": 0.18946829393189382,
+      "grad_norm": 0.8293350005314576,
+      "learning_rate": 3.1569965870307165e-05,
+      "loss": 0.5106,
+      "step": 1110
+    },
+    {
+      "epoch": 0.19032175471537083,
+      "grad_norm": 0.6021094733250301,
+      "learning_rate": 3.171217292377702e-05,
+      "loss": 0.5316,
+      "step": 1115
+    },
+    {
+      "epoch": 0.19117521549884783,
+      "grad_norm": 1.0643292393092736,
+      "learning_rate": 3.1854379977246875e-05,
+      "loss": 0.5596,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19202867628232484,
+      "grad_norm": 0.6299150040629254,
+      "learning_rate": 3.1996587030716724e-05,
+      "loss": 0.4812,
+      "step": 1125
+    },
+    {
+      "epoch": 0.19288213706580182,
+      "grad_norm": 0.5043450499219864,
+      "learning_rate": 3.213879408418658e-05,
+      "loss": 0.4895,
+      "step": 1130
+    },
+    {
+      "epoch": 0.19373559784927882,
+      "grad_norm": 0.5990024804144823,
+      "learning_rate": 3.228100113765643e-05,
+      "loss": 0.5318,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19458905863275583,
+      "grad_norm": 0.640259423124337,
+      "learning_rate": 3.242320819112628e-05,
+      "loss": 0.4925,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19544251941623283,
+      "grad_norm": 0.5721401072678143,
+      "learning_rate": 3.256541524459613e-05,
+      "loss": 0.5364,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1962959801997098,
+      "grad_norm": 0.6483175670557406,
+      "learning_rate": 3.270762229806599e-05,
+      "loss": 0.5097,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19714944098318682,
+      "grad_norm": 0.6304883041191003,
+      "learning_rate": 3.2849829351535835e-05,
+      "loss": 0.529,
+      "step": 1155
+    },
+    {
+      "epoch": 0.19800290176666382,
+      "grad_norm": 0.5626780535279388,
+      "learning_rate": 3.299203640500569e-05,
+      "loss": 0.5207,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19885636255014083,
+      "grad_norm": 0.6141866957416666,
+      "learning_rate": 3.3134243458475546e-05,
+      "loss": 0.4979,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1997098233336178,
+      "grad_norm": 0.5923434635716356,
+      "learning_rate": 3.3276450511945394e-05,
+      "loss": 0.5477,
+      "step": 1170
+    },
+    {
+      "epoch": 0.20056328411709481,
+      "grad_norm": 0.6022601757807821,
+      "learning_rate": 3.341865756541524e-05,
+      "loss": 0.5082,
+      "step": 1175
+    },
+    {
+      "epoch": 0.20141674490057182,
+      "grad_norm": 0.5816942821573442,
+      "learning_rate": 3.35608646188851e-05,
+      "loss": 0.5256,
+      "step": 1180
+    },
+    {
+      "epoch": 0.20227020568404883,
+      "grad_norm": 0.6388083158763254,
+      "learning_rate": 3.370307167235495e-05,
+      "loss": 0.5539,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2031236664675258,
+      "grad_norm": 0.5887921842154867,
+      "learning_rate": 3.38452787258248e-05,
+      "loss": 0.4929,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2039771272510028,
+      "grad_norm": 0.7673632049052999,
+      "learning_rate": 3.398748577929465e-05,
+      "loss": 0.5263,
+      "step": 1195
+    },
+    {
+      "epoch": 0.20483058803447982,
+      "grad_norm": 0.606375881456679,
+      "learning_rate": 3.412969283276451e-05,
+      "loss": 0.5263,
+      "step": 1200
+    },
+    {
+      "epoch": 0.20568404881795682,
+      "grad_norm": 0.5865526644035232,
+      "learning_rate": 3.427189988623436e-05,
+      "loss": 0.5384,
+      "step": 1205
+    },
+    {
+      "epoch": 0.20653750960143383,
+      "grad_norm": 0.6854151362919446,
+      "learning_rate": 3.441410693970421e-05,
+      "loss": 0.5061,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2073909703849108,
+      "grad_norm": 0.5159891247284191,
+      "learning_rate": 3.455631399317406e-05,
+      "loss": 0.484,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2082444311683878,
+      "grad_norm": 0.5624066702071717,
+      "learning_rate": 3.469852104664392e-05,
+      "loss": 0.5225,
+      "step": 1220
+    },
+    {
+      "epoch": 0.20909789195186482,
+      "grad_norm": 0.598719341617163,
+      "learning_rate": 3.484072810011377e-05,
+      "loss": 0.497,
+      "step": 1225
+    },
+    {
+      "epoch": 0.20995135273534182,
+      "grad_norm": 0.528097154070161,
+      "learning_rate": 3.498293515358362e-05,
+      "loss": 0.4962,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2108048135188188,
+      "grad_norm": 0.9598704728440921,
+      "learning_rate": 3.512514220705347e-05,
+      "loss": 0.5462,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2116582743022958,
+      "grad_norm": 0.6266977130887611,
+      "learning_rate": 3.526734926052333e-05,
+      "loss": 0.4967,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2125117350857728,
+      "grad_norm": 0.6043602329334674,
+      "learning_rate": 3.5409556313993176e-05,
+      "loss": 0.5505,
+      "step": 1245
+    },
+    {
+      "epoch": 0.21336519586924982,
+      "grad_norm": 0.611914515574889,
+      "learning_rate": 3.5551763367463024e-05,
+      "loss": 0.5037,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2142186566527268,
+      "grad_norm": 0.4802733556079159,
+      "learning_rate": 3.569397042093288e-05,
+      "loss": 0.4863,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2150721174362038,
+      "grad_norm": 0.7703363378874821,
+      "learning_rate": 3.5836177474402735e-05,
+      "loss": 0.559,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2159255782196808,
+      "grad_norm": 0.6511129669959277,
+      "learning_rate": 3.597838452787258e-05,
+      "loss": 0.5513,
+      "step": 1265
+    },
+    {
+      "epoch": 0.21677903900315781,
+      "grad_norm": 0.5230313645523024,
+      "learning_rate": 3.612059158134244e-05,
+      "loss": 0.5241,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2176324997866348,
+      "grad_norm": 0.7233367326250001,
+      "learning_rate": 3.626279863481229e-05,
+      "loss": 0.5109,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2184859605701118,
+      "grad_norm": 0.7801443277079605,
+      "learning_rate": 3.640500568828214e-05,
+      "loss": 0.5012,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2193394213535888,
+      "grad_norm": 0.5504196600204198,
+      "learning_rate": 3.654721274175199e-05,
+      "loss": 0.529,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2201928821370658,
+      "grad_norm": 0.6541346146815773,
+      "learning_rate": 3.6689419795221846e-05,
+      "loss": 0.5132,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2210463429205428,
+      "grad_norm": 0.6054161786868589,
+      "learning_rate": 3.6831626848691695e-05,
+      "loss": 0.5424,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2218998037040198,
+      "grad_norm": 0.5020466281770849,
+      "learning_rate": 3.697383390216155e-05,
+      "loss": 0.5178,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2227532644874968,
+      "grad_norm": 0.720100028158539,
+      "learning_rate": 3.7116040955631405e-05,
+      "loss": 0.5332,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2236067252709738,
+      "grad_norm": 0.6013562540953598,
+      "learning_rate": 3.7258248009101254e-05,
+      "loss": 0.492,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2244601860544508,
+      "grad_norm": 0.6014554688745792,
+      "learning_rate": 3.74004550625711e-05,
+      "loss": 0.5031,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2253136468379278,
+      "grad_norm": 0.5996694354616003,
+      "learning_rate": 3.754266211604096e-05,
+      "loss": 0.5083,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2261671076214048,
+      "grad_norm": 0.5402077348620818,
+      "learning_rate": 3.768486916951081e-05,
+      "loss": 0.5015,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2270205684048818,
+      "grad_norm": 0.5622664960418867,
+      "learning_rate": 3.782707622298066e-05,
+      "loss": 0.5029,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2278740291883588,
+      "grad_norm": 0.6200593249070808,
+      "learning_rate": 3.796928327645051e-05,
+      "loss": 0.5067,
+      "step": 1335
+    },
+    {
+      "epoch": 0.22872748997183578,
+      "grad_norm": 1.0366247721032074,
+      "learning_rate": 3.811149032992037e-05,
+      "loss": 0.5268,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2295809507553128,
+      "grad_norm": 0.6004382440555587,
+      "learning_rate": 3.825369738339022e-05,
+      "loss": 0.5191,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2304344115387898,
+      "grad_norm": 0.6329180454286693,
+      "learning_rate": 3.839590443686007e-05,
+      "loss": 0.5188,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2312878723222668,
+      "grad_norm": 0.5840877646471871,
+      "learning_rate": 3.853811149032992e-05,
+      "loss": 0.5373,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23214133310574378,
+      "grad_norm": 0.7714178352259251,
+      "learning_rate": 3.868031854379978e-05,
+      "loss": 0.5194,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23299479388922079,
+      "grad_norm": 0.7433746732202924,
+      "learning_rate": 3.882252559726963e-05,
+      "loss": 0.5263,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2338482546726978,
+      "grad_norm": 0.6800388915463906,
+      "learning_rate": 3.8964732650739476e-05,
+      "loss": 0.5085,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2347017154561748,
+      "grad_norm": 0.627799071795484,
+      "learning_rate": 3.910693970420933e-05,
+      "loss": 0.5252,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23555517623965178,
+      "grad_norm": 0.6550245229952462,
+      "learning_rate": 3.924914675767919e-05,
+      "loss": 0.5407,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23640863702312878,
+      "grad_norm": 0.4875389791747265,
+      "learning_rate": 3.9391353811149035e-05,
+      "loss": 0.5582,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2372620978066058,
+      "grad_norm": 0.7704861487727674,
+      "learning_rate": 3.9533560864618884e-05,
+      "loss": 0.5073,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2381155585900828,
+      "grad_norm": 0.5482257906332871,
+      "learning_rate": 3.967576791808874e-05,
+      "loss": 0.5321,
+      "step": 1395
+    },
+    {
+      "epoch": 0.23896901937355977,
+      "grad_norm": 0.6132459928728721,
+      "learning_rate": 3.9817974971558594e-05,
+      "loss": 0.5372,
+      "step": 1400
+    },
+    {
+      "epoch": 0.23982248015703678,
+      "grad_norm": 0.5444425945346957,
+      "learning_rate": 3.996018202502844e-05,
+      "loss": 0.4966,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24067594094051378,
+      "grad_norm": 0.5467973803559824,
+      "learning_rate": 4.01023890784983e-05,
+      "loss": 0.5108,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2415294017239908,
+      "grad_norm": 0.577440466198162,
+      "learning_rate": 4.0244596131968146e-05,
+      "loss": 0.4918,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2423828625074678,
+      "grad_norm": 0.5852142132880568,
+      "learning_rate": 4.0386803185438e-05,
+      "loss": 0.5174,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24323632329094477,
+      "grad_norm": 0.6956282360042076,
+      "learning_rate": 4.052901023890785e-05,
+      "loss": 0.4887,
+      "step": 1425
+    },
+    {
+      "epoch": 0.24408978407442178,
+      "grad_norm": 0.6291740557738249,
+      "learning_rate": 4.0671217292377706e-05,
+      "loss": 0.5317,
+      "step": 1430
+    },
+    {
+      "epoch": 0.24494324485789878,
+      "grad_norm": 0.6356720737025028,
+      "learning_rate": 4.0813424345847554e-05,
+      "loss": 0.5151,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2457967056413758,
+      "grad_norm": 0.5466555007670854,
+      "learning_rate": 4.095563139931741e-05,
+      "loss": 0.5315,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24665016642485277,
+      "grad_norm": 0.49685090209168625,
+      "learning_rate": 4.1097838452787265e-05,
+      "loss": 0.5087,
+      "step": 1445
+    },
+    {
+      "epoch": 0.24750362720832977,
+      "grad_norm": 0.6805393500926865,
+      "learning_rate": 4.124004550625711e-05,
+      "loss": 0.5096,
+      "step": 1450
+    },
+    {
+      "epoch": 0.24835708799180678,
+      "grad_norm": 0.5725498300350252,
+      "learning_rate": 4.138225255972696e-05,
+      "loss": 0.4944,
+      "step": 1455
+    },
+    {
+      "epoch": 0.24921054877528379,
+      "grad_norm": 0.6133397456710716,
+      "learning_rate": 4.152445961319682e-05,
+      "loss": 0.504,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2500640095587608,
+      "grad_norm": 0.718665494692223,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.5238,
+      "step": 1465
+    },
+    {
+      "epoch": 0.25091747034223777,
+      "grad_norm": 0.6247896227403243,
+      "learning_rate": 4.180887372013652e-05,
+      "loss": 0.5065,
+      "step": 1470
+    },
+    {
+      "epoch": 0.25177093112571475,
+      "grad_norm": 0.538557800481529,
+      "learning_rate": 4.195108077360637e-05,
+      "loss": 0.493,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2526243919091918,
+      "grad_norm": 0.6496369108151522,
+      "learning_rate": 4.2093287827076224e-05,
+      "loss": 0.5221,
+      "step": 1480
+    },
+    {
+      "epoch": 0.25347785269266876,
+      "grad_norm": 0.48828596472972324,
+      "learning_rate": 4.223549488054608e-05,
+      "loss": 0.5136,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2543313134761458,
+      "grad_norm": 0.5625364791015088,
+      "learning_rate": 4.237770193401593e-05,
+      "loss": 0.5145,
+      "step": 1490
+    },
+    {
+      "epoch": 0.25518477425962277,
+      "grad_norm": 0.5017063317201218,
+      "learning_rate": 4.2519908987485777e-05,
+      "loss": 0.5582,
+      "step": 1495
+    },
+    {
+      "epoch": 0.25603823504309975,
+      "grad_norm": 0.6288989261344257,
+      "learning_rate": 4.266211604095564e-05,
+      "loss": 0.53,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2568916958265768,
+      "grad_norm": 0.6328069624699362,
+      "learning_rate": 4.280432309442549e-05,
+      "loss": 0.5444,
+      "step": 1505
+    },
+    {
+      "epoch": 0.25774515661005376,
+      "grad_norm": 0.5393975697229143,
+      "learning_rate": 4.2946530147895336e-05,
+      "loss": 0.4809,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2585986173935308,
+      "grad_norm": 0.5408443482735174,
+      "learning_rate": 4.308873720136519e-05,
+      "loss": 0.509,
+      "step": 1515
+    },
+    {
+      "epoch": 0.25945207817700777,
+      "grad_norm": 0.7121021821053504,
+      "learning_rate": 4.3230944254835046e-05,
+      "loss": 0.5138,
+      "step": 1520
+    },
+    {
+      "epoch": 0.26030553896048475,
+      "grad_norm": 0.5929259032947689,
+      "learning_rate": 4.3373151308304895e-05,
+      "loss": 0.4904,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2611589997439618,
+      "grad_norm": 0.6827715964490704,
+      "learning_rate": 4.351535836177474e-05,
+      "loss": 0.5197,
+      "step": 1530
+    },
+    {
+      "epoch": 0.26201246052743876,
+      "grad_norm": 0.5893962009428101,
+      "learning_rate": 4.36575654152446e-05,
+      "loss": 0.5113,
+      "step": 1535
+    },
+    {
+      "epoch": 0.26286592131091574,
+      "grad_norm": 0.7665006113847211,
+      "learning_rate": 4.3799772468714454e-05,
+      "loss": 0.4894,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2637193820943928,
+      "grad_norm": 0.5784712623580104,
+      "learning_rate": 4.39419795221843e-05,
+      "loss": 0.5031,
+      "step": 1545
+    },
+    {
+      "epoch": 0.26457284287786975,
+      "grad_norm": 0.4779882713725444,
+      "learning_rate": 4.408418657565416e-05,
+      "loss": 0.4841,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2654263036613468,
+      "grad_norm": 0.5265214659857844,
+      "learning_rate": 4.4226393629124006e-05,
+      "loss": 0.5366,
+      "step": 1555
+    },
+    {
+      "epoch": 0.26627976444482376,
+      "grad_norm": 0.5062628084381423,
+      "learning_rate": 4.436860068259386e-05,
+      "loss": 0.516,
+      "step": 1560
+    },
+    {
+      "epoch": 0.26713322522830074,
+      "grad_norm": 0.5384657066787054,
+      "learning_rate": 4.451080773606371e-05,
+      "loss": 0.5437,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2679866860117778,
+      "grad_norm": 0.4975884755019286,
+      "learning_rate": 4.4653014789533565e-05,
+      "loss": 0.4922,
+      "step": 1570
+    },
+    {
+      "epoch": 0.26884014679525475,
+      "grad_norm": 0.5763335477313506,
+      "learning_rate": 4.4795221843003413e-05,
+      "loss": 0.5283,
+      "step": 1575
+    },
+    {
+      "epoch": 0.26969360757873173,
+      "grad_norm": 0.5978071784312756,
+      "learning_rate": 4.493742889647327e-05,
+      "loss": 0.5144,
+      "step": 1580
+    },
+    {
+      "epoch": 0.27054706836220876,
+      "grad_norm": 0.8359278465607787,
+      "learning_rate": 4.5079635949943124e-05,
+      "loss": 0.4983,
+      "step": 1585
+    },
+    {
+      "epoch": 0.27140052914568574,
+      "grad_norm": 0.6214765417774613,
+      "learning_rate": 4.522184300341297e-05,
+      "loss": 0.5367,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2722539899291628,
+      "grad_norm": 0.508626898726594,
+      "learning_rate": 4.536405005688282e-05,
+      "loss": 0.5132,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27310745071263975,
+      "grad_norm": 0.6022081309698558,
+      "learning_rate": 4.5506257110352676e-05,
+      "loss": 0.5004,
+      "step": 1600
+    },
+    {
+      "epoch": 0.27396091149611673,
+      "grad_norm": 0.5155913335233575,
+      "learning_rate": 4.564846416382253e-05,
+      "loss": 0.4656,
+      "step": 1605
+    },
+    {
+      "epoch": 0.27481437227959377,
+      "grad_norm": 0.5791109901377541,
+      "learning_rate": 4.579067121729238e-05,
+      "loss": 0.4984,
+      "step": 1610
+    },
+    {
+      "epoch": 0.27566783306307074,
+      "grad_norm": 0.5767520283303881,
+      "learning_rate": 4.593287827076223e-05,
+      "loss": 0.5158,
+      "step": 1615
+    },
+    {
+      "epoch": 0.2765212938465478,
+      "grad_norm": 0.48349328202135766,
+      "learning_rate": 4.6075085324232084e-05,
+      "loss": 0.516,
+      "step": 1620
+    },
+    {
+      "epoch": 0.27737475463002476,
+      "grad_norm": 0.5355748029090559,
+      "learning_rate": 4.621729237770194e-05,
+      "loss": 0.5548,
+      "step": 1625
+    },
+    {
+      "epoch": 0.27822821541350173,
+      "grad_norm": 0.5274726573128763,
+      "learning_rate": 4.635949943117179e-05,
+      "loss": 0.5171,
+      "step": 1630
+    },
+    {
+      "epoch": 0.27908167619697877,
+      "grad_norm": 0.6055861657371587,
+      "learning_rate": 4.6501706484641636e-05,
+      "loss": 0.5496,
+      "step": 1635
+    },
+    {
+      "epoch": 0.27993513698045575,
+      "grad_norm": 0.5525370956087899,
+      "learning_rate": 4.664391353811149e-05,
+      "loss": 0.4954,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2807885977639327,
+      "grad_norm": 0.7432424117951163,
+      "learning_rate": 4.6786120591581347e-05,
+      "loss": 0.5069,
+      "step": 1645
+    },
+    {
+      "epoch": 0.28164205854740976,
+      "grad_norm": 0.5428701235562144,
+      "learning_rate": 4.6928327645051195e-05,
+      "loss": 0.516,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28249551933088674,
+      "grad_norm": 0.5595977569856022,
+      "learning_rate": 4.707053469852105e-05,
+      "loss": 0.5287,
+      "step": 1655
+    },
+    {
+      "epoch": 0.28334898011436377,
+      "grad_norm": 0.5031490669266451,
+      "learning_rate": 4.72127417519909e-05,
+      "loss": 0.4772,
+      "step": 1660
+    },
+    {
+      "epoch": 0.28420244089784075,
+      "grad_norm": 0.5529373305843895,
+      "learning_rate": 4.7354948805460754e-05,
+      "loss": 0.5293,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2850559016813177,
+      "grad_norm": 0.5050081484901403,
+      "learning_rate": 4.74971558589306e-05,
+      "loss": 0.5124,
+      "step": 1670
+    },
+    {
+      "epoch": 0.28590936246479476,
+      "grad_norm": 0.5466603151465805,
+      "learning_rate": 4.763936291240046e-05,
+      "loss": 0.5161,
+      "step": 1675
+    },
+    {
+      "epoch": 0.28676282324827174,
+      "grad_norm": 0.5114919686427399,
+      "learning_rate": 4.778156996587031e-05,
+      "loss": 0.5122,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2876162840317487,
+      "grad_norm": 0.505864506408219,
+      "learning_rate": 4.792377701934016e-05,
+      "loss": 0.5015,
+      "step": 1685
+    },
+    {
+      "epoch": 0.28846974481522575,
+      "grad_norm": 0.5253048001147426,
+      "learning_rate": 4.806598407281002e-05,
+      "loss": 0.5189,
+      "step": 1690
+    },
+    {
+      "epoch": 0.2893232055987027,
+      "grad_norm": 0.577856690154839,
+      "learning_rate": 4.8208191126279865e-05,
+      "loss": 0.5189,
+      "step": 1695
+    },
+    {
+      "epoch": 0.29017666638217976,
+      "grad_norm": 0.6130819699326354,
+      "learning_rate": 4.835039817974972e-05,
+      "loss": 0.5302,
+      "step": 1700
+    },
+    {
+      "epoch": 0.29103012716565674,
+      "grad_norm": 0.5188890895398451,
+      "learning_rate": 4.849260523321957e-05,
+      "loss": 0.5273,
+      "step": 1705
+    },
+    {
+      "epoch": 0.2918835879491337,
+      "grad_norm": 0.5787667071281957,
+      "learning_rate": 4.8634812286689424e-05,
+      "loss": 0.5221,
+      "step": 1710
+    },
+    {
+      "epoch": 0.29273704873261075,
+      "grad_norm": 0.5207600512605698,
+      "learning_rate": 4.877701934015927e-05,
+      "loss": 0.5224,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2935905095160877,
+      "grad_norm": 0.47177284402958103,
+      "learning_rate": 4.891922639362913e-05,
+      "loss": 0.5052,
+      "step": 1720
+    },
+    {
+      "epoch": 0.29444397029956476,
+      "grad_norm": 0.5622381226349713,
+      "learning_rate": 4.906143344709898e-05,
+      "loss": 0.4944,
+      "step": 1725
+    },
+    {
+      "epoch": 0.29529743108304174,
+      "grad_norm": 0.5135889415167705,
+      "learning_rate": 4.920364050056883e-05,
+      "loss": 0.4853,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2961508918665187,
+      "grad_norm": 0.5568571026359952,
+      "learning_rate": 4.934584755403868e-05,
+      "loss": 0.5118,
+      "step": 1735
+    },
+    {
+      "epoch": 0.29700435264999575,
+      "grad_norm": 0.5358216069493535,
+      "learning_rate": 4.9488054607508536e-05,
+      "loss": 0.5026,
+      "step": 1740
+    },
+    {
+      "epoch": 0.29785781343347273,
+      "grad_norm": 0.530619616402973,
+      "learning_rate": 4.963026166097839e-05,
+      "loss": 0.5146,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2987112742169497,
+      "grad_norm": 0.499787376121546,
+      "learning_rate": 4.977246871444824e-05,
+      "loss": 0.4968,
+      "step": 1750
+    },
+    {
+      "epoch": 0.29956473500042674,
+      "grad_norm": 0.5101078686357036,
+      "learning_rate": 4.991467576791809e-05,
+      "loss": 0.4946,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3004181957839037,
+      "grad_norm": 0.6548359705660204,
+      "learning_rate": 4.999367728882145e-05,
+      "loss": 0.5244,
+      "step": 1760
+    },
+    {
+      "epoch": 0.30127165656738075,
+      "grad_norm": 0.48137998048070096,
+      "learning_rate": 4.997787051087506e-05,
+      "loss": 0.5162,
+      "step": 1765
+    },
+    {
+      "epoch": 0.30212511735085773,
+      "grad_norm": 0.48614869274456174,
+      "learning_rate": 4.9962063732928684e-05,
+      "loss": 0.5062,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3029785781343347,
+      "grad_norm": 0.497782874867819,
+      "learning_rate": 4.99462569549823e-05,
+      "loss": 0.5455,
+      "step": 1775
+    },
+    {
+      "epoch": 0.30383203891781174,
+      "grad_norm": 0.5230168871985573,
+      "learning_rate": 4.993045017703591e-05,
+      "loss": 0.5312,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3046854997012887,
+      "grad_norm": 0.582983660254052,
+      "learning_rate": 4.991464339908954e-05,
+      "loss": 0.5127,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3055389604847657,
+      "grad_norm": 0.5396571524519772,
+      "learning_rate": 4.989883662114315e-05,
+      "loss": 0.4799,
+      "step": 1790
+    },
+    {
+      "epoch": 0.30639242126824273,
+      "grad_norm": 0.4678056267208111,
+      "learning_rate": 4.9883029843196765e-05,
+      "loss": 0.4892,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3072458820517197,
+      "grad_norm": 0.5877677893826099,
+      "learning_rate": 4.986722306525038e-05,
+      "loss": 0.5267,
+      "step": 1800
+    },
+    {
+      "epoch": 0.30809934283519674,
+      "grad_norm": 0.47625277215848183,
+      "learning_rate": 4.9851416287304e-05,
+      "loss": 0.51,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3089528036186737,
+      "grad_norm": 0.4607978832628274,
+      "learning_rate": 4.983560950935762e-05,
+      "loss": 0.5005,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3098062644021507,
+      "grad_norm": 0.508720567496007,
+      "learning_rate": 4.981980273141123e-05,
+      "loss": 0.4917,
+      "step": 1815
+    },
+    {
+      "epoch": 0.31065972518562773,
+      "grad_norm": 0.5498811575656698,
+      "learning_rate": 4.980399595346485e-05,
+      "loss": 0.539,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3115131859691047,
+      "grad_norm": 0.5059386725583341,
+      "learning_rate": 4.978818917551846e-05,
+      "loss": 0.5072,
+      "step": 1825
+    },
+    {
+      "epoch": 0.31236664675258174,
+      "grad_norm": 0.4591900174762019,
+      "learning_rate": 4.977238239757208e-05,
+      "loss": 0.5296,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3132201075360587,
+      "grad_norm": 0.6389348492531627,
+      "learning_rate": 4.97565756196257e-05,
+      "loss": 0.4879,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3140735683195357,
+      "grad_norm": 0.5536804544450629,
+      "learning_rate": 4.9740768841679315e-05,
+      "loss": 0.4865,
+      "step": 1840
+    },
+    {
+      "epoch": 0.31492702910301273,
+      "grad_norm": 0.47776468085928686,
+      "learning_rate": 4.972496206373293e-05,
+      "loss": 0.5003,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3157804898864897,
+      "grad_norm": 0.5040049671237692,
+      "learning_rate": 4.970915528578654e-05,
+      "loss": 0.487,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3166339506699667,
+      "grad_norm": 0.48062488023690886,
+      "learning_rate": 4.969334850784017e-05,
+      "loss": 0.4891,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3174874114534437,
+      "grad_norm": 0.5093613218378991,
+      "learning_rate": 4.967754172989378e-05,
+      "loss": 0.4753,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3183408722369207,
+      "grad_norm": 0.5624601450965416,
+      "learning_rate": 4.9661734951947396e-05,
+      "loss": 0.5231,
+      "step": 1865
+    },
+    {
+      "epoch": 0.31919433302039774,
+      "grad_norm": 0.5871555056740465,
+      "learning_rate": 4.964592817400101e-05,
+      "loss": 0.502,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3200477938038747,
+      "grad_norm": 0.6212161463505985,
+      "learning_rate": 4.963012139605463e-05,
+      "loss": 0.5248,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3209012545873517,
+      "grad_norm": 0.5483362010525371,
+      "learning_rate": 4.961431461810825e-05,
+      "loss": 0.497,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3217547153708287,
+      "grad_norm": 0.5386127950130882,
+      "learning_rate": 4.959850784016186e-05,
+      "loss": 0.5255,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3226081761543057,
+      "grad_norm": 0.5878591256347312,
+      "learning_rate": 4.958270106221548e-05,
+      "loss": 0.4989,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3234616369377827,
+      "grad_norm": 0.5174651720515663,
+      "learning_rate": 4.9566894284269094e-05,
+      "loss": 0.5361,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3243150977212597,
+      "grad_norm": 0.5698729020989101,
+      "learning_rate": 4.955108750632271e-05,
+      "loss": 0.4916,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3251685585047367,
+      "grad_norm": 0.4601199986994053,
+      "learning_rate": 4.9535280728376335e-05,
+      "loss": 0.4879,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3260220192882137,
+      "grad_norm": 0.6866248215777093,
+      "learning_rate": 4.9519473950429946e-05,
+      "loss": 0.5312,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3268754800716907,
+      "grad_norm": 0.5412387753636492,
+      "learning_rate": 4.9503667172483563e-05,
+      "loss": 0.5012,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3277289408551677,
+      "grad_norm": 0.5873267793920224,
+      "learning_rate": 4.948786039453718e-05,
+      "loss": 0.509,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3285824016386447,
+      "grad_norm": 0.5189115320755102,
+      "learning_rate": 4.94720536165908e-05,
+      "loss": 0.484,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3294358624221217,
+      "grad_norm": 0.5022541830055647,
+      "learning_rate": 4.9456246838644416e-05,
+      "loss": 0.502,
+      "step": 1930
+    },
+    {
+      "epoch": 0.33028932320559873,
+      "grad_norm": 0.6636199777411177,
+      "learning_rate": 4.9440440060698027e-05,
+      "loss": 0.5468,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3311427839890757,
+      "grad_norm": 0.47976829113815256,
+      "learning_rate": 4.942463328275165e-05,
+      "loss": 0.5609,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3319962447725527,
+      "grad_norm": 0.5364386551157393,
+      "learning_rate": 4.940882650480526e-05,
+      "loss": 0.5277,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3328497055560297,
+      "grad_norm": 0.5284348268891784,
+      "learning_rate": 4.939301972685888e-05,
+      "loss": 0.5115,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3337031663395067,
+      "grad_norm": 0.5985767793816986,
+      "learning_rate": 4.9377212948912496e-05,
+      "loss": 0.4869,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3345566271229837,
+      "grad_norm": 0.49985285222802744,
+      "learning_rate": 4.9361406170966114e-05,
+      "loss": 0.5194,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3354100879064607,
+      "grad_norm": 0.47473725204758305,
+      "learning_rate": 4.934559939301973e-05,
+      "loss": 0.5156,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3362635486899377,
+      "grad_norm": 0.4813926004817435,
+      "learning_rate": 4.932979261507334e-05,
+      "loss": 0.511,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3371170094734147,
+      "grad_norm": 0.44287958712492753,
+      "learning_rate": 4.9313985837126966e-05,
+      "loss": 0.5052,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3379704702568917,
+      "grad_norm": 0.8696903571161176,
+      "learning_rate": 4.929817905918058e-05,
+      "loss": 0.4993,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3388239310403687,
+      "grad_norm": 0.4961065331077907,
+      "learning_rate": 4.9282372281234194e-05,
+      "loss": 0.4847,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3396773918238457,
+      "grad_norm": 0.4303833412454493,
+      "learning_rate": 4.926656550328781e-05,
+      "loss": 0.5273,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3405308526073227,
+      "grad_norm": 0.5307093252344055,
+      "learning_rate": 4.925075872534143e-05,
+      "loss": 0.4798,
+      "step": 1995
+    },
+    {
+      "epoch": 0.34138431339079967,
+      "grad_norm": 0.588253233323147,
+      "learning_rate": 4.923495194739505e-05,
+      "loss": 0.4944,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3422377741742767,
+      "grad_norm": 0.4260031992882977,
+      "learning_rate": 4.921914516944866e-05,
+      "loss": 0.498,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3430912349577537,
+      "grad_norm": 0.6840740508852956,
+      "learning_rate": 4.920333839150228e-05,
+      "loss": 0.5146,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3439446957412307,
+      "grad_norm": 0.515886091219417,
+      "learning_rate": 4.918753161355589e-05,
+      "loss": 0.5015,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3447981565247077,
+      "grad_norm": 0.5556290220178304,
+      "learning_rate": 4.917172483560951e-05,
+      "loss": 0.5045,
+      "step": 2020
+    },
+    {
+      "epoch": 0.34565161730818467,
+      "grad_norm": 0.49030475027530085,
+      "learning_rate": 4.9155918057663134e-05,
+      "loss": 0.5244,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3465050780916617,
+      "grad_norm": 0.5762252501319363,
+      "learning_rate": 4.9140111279716745e-05,
+      "loss": 0.5091,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3473585388751387,
+      "grad_norm": 0.5442309646550838,
+      "learning_rate": 4.912430450177036e-05,
+      "loss": 0.5201,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3482119996586157,
+      "grad_norm": 0.49245899017001116,
+      "learning_rate": 4.910849772382398e-05,
+      "loss": 0.4858,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3490654604420927,
+      "grad_norm": 0.4810738247547212,
+      "learning_rate": 4.90926909458776e-05,
+      "loss": 0.4851,
+      "step": 2045
+    },
+    {
+      "epoch": 0.34991892122556967,
+      "grad_norm": 0.5208290744758776,
+      "learning_rate": 4.907688416793121e-05,
+      "loss": 0.5017,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3507723820090467,
+      "grad_norm": 0.5332087434013345,
+      "learning_rate": 4.9061077389984825e-05,
+      "loss": 0.4798,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3516258427925237,
+      "grad_norm": 0.4661707910424319,
+      "learning_rate": 4.904527061203845e-05,
+      "loss": 0.4949,
+      "step": 2060
+    },
+    {
+      "epoch": 0.35247930357600066,
+      "grad_norm": 0.6661288913768391,
+      "learning_rate": 4.902946383409206e-05,
+      "loss": 0.523,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3533327643594777,
+      "grad_norm": 0.495108641158868,
+      "learning_rate": 4.901365705614568e-05,
+      "loss": 0.4852,
+      "step": 2070
+    },
+    {
+      "epoch": 0.35418622514295467,
+      "grad_norm": 0.4680124924286701,
+      "learning_rate": 4.8997850278199295e-05,
+      "loss": 0.4814,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3550396859264317,
+      "grad_norm": 0.39154448085109655,
+      "learning_rate": 4.898204350025291e-05,
+      "loss": 0.4762,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3558931467099087,
+      "grad_norm": 0.4315351278111706,
+      "learning_rate": 4.896623672230653e-05,
+      "loss": 0.5076,
+      "step": 2085
+    },
+    {
+      "epoch": 0.35674660749338566,
+      "grad_norm": 0.6209355293840191,
+      "learning_rate": 4.895042994436014e-05,
+      "loss": 0.4814,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3576000682768627,
+      "grad_norm": 0.58744008191501,
+      "learning_rate": 4.8934623166413765e-05,
+      "loss": 0.5386,
+      "step": 2095
+    },
+    {
+      "epoch": 0.35845352906033967,
+      "grad_norm": 0.523456459006555,
+      "learning_rate": 4.8918816388467376e-05,
+      "loss": 0.5407,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3593069898438167,
+      "grad_norm": 0.5858458920130489,
+      "learning_rate": 4.890300961052099e-05,
+      "loss": 0.5043,
+      "step": 2105
+    },
+    {
+      "epoch": 0.3601604506272937,
+      "grad_norm": 0.622924245935873,
+      "learning_rate": 4.888720283257461e-05,
+      "loss": 0.4852,
+      "step": 2110
+    },
+    {
+      "epoch": 0.36101391141077066,
+      "grad_norm": 0.5331367222638274,
+      "learning_rate": 4.887139605462823e-05,
+      "loss": 0.5179,
+      "step": 2115
+    },
+    {
+      "epoch": 0.3618673721942477,
+      "grad_norm": 0.6035015648064055,
+      "learning_rate": 4.8855589276681845e-05,
+      "loss": 0.4673,
+      "step": 2120
+    },
+    {
+      "epoch": 0.36272083297772467,
+      "grad_norm": 0.5614942447133979,
+      "learning_rate": 4.8839782498735456e-05,
+      "loss": 0.4953,
+      "step": 2125
+    },
+    {
+      "epoch": 0.36357429376120165,
+      "grad_norm": 0.5417698788101754,
+      "learning_rate": 4.882397572078908e-05,
+      "loss": 0.51,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3644277545446787,
+      "grad_norm": 0.43028318471622506,
+      "learning_rate": 4.880816894284269e-05,
+      "loss": 0.5143,
+      "step": 2135
+    },
+    {
+      "epoch": 0.36528121532815566,
+      "grad_norm": 0.46453203731114756,
+      "learning_rate": 4.879236216489631e-05,
+      "loss": 0.5029,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3661346761116327,
+      "grad_norm": 0.4414681810312383,
+      "learning_rate": 4.8776555386949926e-05,
+      "loss": 0.517,
+      "step": 2145
+    },
+    {
+      "epoch": 0.3669881368951097,
+      "grad_norm": 0.5440667658494504,
+      "learning_rate": 4.876074860900354e-05,
+      "loss": 0.4952,
+      "step": 2150
+    },
+    {
+      "epoch": 0.36784159767858665,
+      "grad_norm": 0.5719949542451221,
+      "learning_rate": 4.874494183105716e-05,
+      "loss": 0.5067,
+      "step": 2155
+    },
+    {
+      "epoch": 0.3686950584620637,
+      "grad_norm": 0.4987125482108236,
+      "learning_rate": 4.872913505311078e-05,
+      "loss": 0.5011,
+      "step": 2160
+    },
+    {
+      "epoch": 0.36954851924554066,
+      "grad_norm": 0.48371745245377856,
+      "learning_rate": 4.8713328275164396e-05,
+      "loss": 0.5104,
+      "step": 2165
+    },
+    {
+      "epoch": 0.37040198002901764,
+      "grad_norm": 0.5571059180363649,
+      "learning_rate": 4.8697521497218006e-05,
+      "loss": 0.4963,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3712554408124947,
+      "grad_norm": 0.43563340403560935,
+      "learning_rate": 4.8681714719271624e-05,
+      "loss": 0.5039,
+      "step": 2175
+    },
+    {
+      "epoch": 0.37210890159597165,
+      "grad_norm": 0.4691492163127887,
+      "learning_rate": 4.866590794132525e-05,
+      "loss": 0.4919,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3729623623794487,
+      "grad_norm": 0.6600911089239349,
+      "learning_rate": 4.865010116337886e-05,
+      "loss": 0.5053,
+      "step": 2185
+    },
+    {
+      "epoch": 0.37381582316292566,
+      "grad_norm": 0.5111477799911323,
+      "learning_rate": 4.8634294385432476e-05,
+      "loss": 0.4909,
+      "step": 2190
+    },
+    {
+      "epoch": 0.37466928394640264,
+      "grad_norm": 0.6057332546181553,
+      "learning_rate": 4.8618487607486094e-05,
+      "loss": 0.5082,
+      "step": 2195
+    },
+    {
+      "epoch": 0.3755227447298797,
+      "grad_norm": 0.44911636114761194,
+      "learning_rate": 4.860268082953971e-05,
+      "loss": 0.4956,
+      "step": 2200
+    },
+    {
+      "epoch": 0.37637620551335665,
+      "grad_norm": 0.5266248333158463,
+      "learning_rate": 4.858687405159332e-05,
+      "loss": 0.4905,
+      "step": 2205
+    },
+    {
+      "epoch": 0.3772296662968337,
+      "grad_norm": 0.5209847386036796,
+      "learning_rate": 4.857106727364694e-05,
+      "loss": 0.5158,
+      "step": 2210
+    },
+    {
+      "epoch": 0.37808312708031067,
+      "grad_norm": 0.42162475531591376,
+      "learning_rate": 4.8555260495700563e-05,
+      "loss": 0.4957,
+      "step": 2215
+    },
+    {
+      "epoch": 0.37893658786378764,
+      "grad_norm": 0.5268786836115406,
+      "learning_rate": 4.8539453717754174e-05,
+      "loss": 0.5108,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3797900486472647,
+      "grad_norm": 0.5502003172812554,
+      "learning_rate": 4.852364693980779e-05,
+      "loss": 0.503,
+      "step": 2225
+    },
+    {
+      "epoch": 0.38064350943074166,
+      "grad_norm": 0.47184815550175846,
+      "learning_rate": 4.850784016186141e-05,
+      "loss": 0.4816,
+      "step": 2230
+    },
+    {
+      "epoch": 0.38149697021421863,
+      "grad_norm": 0.4914467821125124,
+      "learning_rate": 4.8492033383915027e-05,
+      "loss": 0.5124,
+      "step": 2235
+    },
+    {
+      "epoch": 0.38235043099769567,
+      "grad_norm": 0.5284192832763485,
+      "learning_rate": 4.8476226605968644e-05,
+      "loss": 0.527,
+      "step": 2240
+    },
+    {
+      "epoch": 0.38320389178117265,
+      "grad_norm": 0.43916650960746295,
+      "learning_rate": 4.8460419828022255e-05,
+      "loss": 0.4853,
+      "step": 2245
+    },
+    {
+      "epoch": 0.3840573525646497,
+      "grad_norm": 0.4822030333495498,
+      "learning_rate": 4.844461305007588e-05,
+      "loss": 0.5259,
+      "step": 2250
+    },
+    {
+      "epoch": 0.38491081334812666,
+      "grad_norm": 0.4976857587767006,
+      "learning_rate": 4.842880627212949e-05,
+      "loss": 0.5126,
+      "step": 2255
+    },
+    {
+      "epoch": 0.38576427413160363,
+      "grad_norm": 0.5028406437234613,
+      "learning_rate": 4.841299949418311e-05,
+      "loss": 0.5037,
+      "step": 2260
+    },
+    {
+      "epoch": 0.38661773491508067,
+      "grad_norm": 0.5703060961212163,
+      "learning_rate": 4.8397192716236725e-05,
+      "loss": 0.5136,
+      "step": 2265
+    },
+    {
+      "epoch": 0.38747119569855765,
+      "grad_norm": 0.5540371585065555,
+      "learning_rate": 4.838138593829034e-05,
+      "loss": 0.5219,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3883246564820346,
+      "grad_norm": 0.460303219646615,
+      "learning_rate": 4.836557916034396e-05,
+      "loss": 0.465,
+      "step": 2275
+    },
+    {
+      "epoch": 0.38917811726551166,
+      "grad_norm": 0.6376586309384963,
+      "learning_rate": 4.834977238239757e-05,
+      "loss": 0.5159,
+      "step": 2280
+    },
+    {
+      "epoch": 0.39003157804898864,
+      "grad_norm": 0.5385467971311747,
+      "learning_rate": 4.8333965604451194e-05,
+      "loss": 0.5133,
+      "step": 2285
+    },
+    {
+      "epoch": 0.39088503883246567,
+      "grad_norm": 0.45603340338427445,
+      "learning_rate": 4.8318158826504805e-05,
+      "loss": 0.4883,
+      "step": 2290
+    },
+    {
+      "epoch": 0.39173849961594265,
+      "grad_norm": 0.495022738683918,
+      "learning_rate": 4.830235204855842e-05,
+      "loss": 0.4919,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3925919603994196,
+      "grad_norm": 0.4198291402943894,
+      "learning_rate": 4.828654527061204e-05,
+      "loss": 0.4913,
+      "step": 2300
+    },
+    {
+      "epoch": 0.39344542118289666,
+      "grad_norm": 0.4897660003658656,
+      "learning_rate": 4.827073849266566e-05,
+      "loss": 0.4885,
+      "step": 2305
+    },
+    {
+      "epoch": 0.39429888196637364,
+      "grad_norm": 0.5880897395232438,
+      "learning_rate": 4.8254931714719275e-05,
+      "loss": 0.5182,
+      "step": 2310
+    },
+    {
+      "epoch": 0.39515234274985067,
+      "grad_norm": 0.5237830796008816,
+      "learning_rate": 4.823912493677289e-05,
+      "loss": 0.5066,
+      "step": 2315
+    },
+    {
+      "epoch": 0.39600580353332765,
+      "grad_norm": 0.46915821115914,
+      "learning_rate": 4.822331815882651e-05,
+      "loss": 0.5038,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3968592643168046,
+      "grad_norm": 0.489977862747463,
+      "learning_rate": 4.820751138088012e-05,
+      "loss": 0.5023,
+      "step": 2325
+    },
+    {
+      "epoch": 0.39771272510028166,
+      "grad_norm": 0.5519206595985753,
+      "learning_rate": 4.819170460293374e-05,
+      "loss": 0.5092,
+      "step": 2330
+    },
+    {
+      "epoch": 0.39856618588375864,
+      "grad_norm": 0.7853155996315747,
+      "learning_rate": 4.817589782498736e-05,
+      "loss": 0.4821,
+      "step": 2335
+    },
+    {
+      "epoch": 0.3994196466672356,
+      "grad_norm": 0.533071490199691,
+      "learning_rate": 4.816009104704097e-05,
+      "loss": 0.4623,
+      "step": 2340
+    },
+    {
+      "epoch": 0.40027310745071265,
+      "grad_norm": 0.4311014867792575,
+      "learning_rate": 4.814428426909459e-05,
+      "loss": 0.5367,
+      "step": 2345
+    },
+    {
+      "epoch": 0.40112656823418963,
+      "grad_norm": 0.5065115129965704,
+      "learning_rate": 4.812847749114821e-05,
+      "loss": 0.4755,
+      "step": 2350
+    },
+    {
+      "epoch": 0.40198002901766666,
+      "grad_norm": 0.6459279188772843,
+      "learning_rate": 4.8112670713201825e-05,
+      "loss": 0.4819,
+      "step": 2355
+    },
+    {
+      "epoch": 0.40283348980114364,
+      "grad_norm": 0.5234927153215261,
+      "learning_rate": 4.8096863935255436e-05,
+      "loss": 0.5055,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4036869505846206,
+      "grad_norm": 0.5002221843150646,
+      "learning_rate": 4.808105715730905e-05,
+      "loss": 0.4641,
+      "step": 2365
+    },
+    {
+      "epoch": 0.40454041136809765,
+      "grad_norm": 0.7926000042008581,
+      "learning_rate": 4.806525037936268e-05,
+      "loss": 0.4794,
+      "step": 2370
+    },
+    {
+      "epoch": 0.40539387215157463,
+      "grad_norm": 0.4819003808100341,
+      "learning_rate": 4.804944360141629e-05,
+      "loss": 0.5226,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4062473329350516,
+      "grad_norm": 0.7494131850862085,
+      "learning_rate": 4.8033636823469906e-05,
+      "loss": 0.4793,
+      "step": 2380
+    },
+    {
+      "epoch": 0.40710079371852864,
+      "grad_norm": 0.43027822412816397,
+      "learning_rate": 4.801783004552352e-05,
+      "loss": 0.4661,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4079542545020056,
+      "grad_norm": 0.5258254862312308,
+      "learning_rate": 4.800202326757714e-05,
+      "loss": 0.494,
+      "step": 2390
+    },
+    {
+      "epoch": 0.40880771528548265,
+      "grad_norm": 0.5089133616963131,
+      "learning_rate": 4.798621648963076e-05,
+      "loss": 0.4804,
+      "step": 2395
+    },
+    {
+      "epoch": 0.40966117606895963,
+      "grad_norm": 0.4832185604406258,
+      "learning_rate": 4.797040971168437e-05,
+      "loss": 0.4921,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4105146368524366,
+      "grad_norm": 0.4978598215249046,
+      "learning_rate": 4.795460293373799e-05,
+      "loss": 0.4759,
+      "step": 2405
+    },
+    {
+      "epoch": 0.41136809763591364,
+      "grad_norm": 0.48133388100148394,
+      "learning_rate": 4.7938796155791604e-05,
+      "loss": 0.5048,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4122215584193906,
+      "grad_norm": 0.4696469239540106,
+      "learning_rate": 4.792298937784522e-05,
+      "loss": 0.5042,
+      "step": 2415
+    },
+    {
+      "epoch": 0.41307501920286765,
+      "grad_norm": 0.5581633064738356,
+      "learning_rate": 4.790718259989884e-05,
+      "loss": 0.4927,
+      "step": 2420
+    },
+    {
+      "epoch": 0.41392847998634463,
+      "grad_norm": 0.6278902607749504,
+      "learning_rate": 4.7891375821952456e-05,
+      "loss": 0.503,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4147819407698216,
+      "grad_norm": 0.4517239717322925,
+      "learning_rate": 4.7875569044006074e-05,
+      "loss": 0.5011,
+      "step": 2430
+    },
+    {
+      "epoch": 0.41563540155329864,
+      "grad_norm": 0.6903762794458227,
+      "learning_rate": 4.785976226605969e-05,
+      "loss": 0.5233,
+      "step": 2435
+    },
+    {
+      "epoch": 0.4164888623367756,
+      "grad_norm": 3.2978680656603636,
+      "learning_rate": 4.784395548811331e-05,
+      "loss": 0.5109,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4173423231202526,
+      "grad_norm": 1.1420518689716137,
+      "learning_rate": 4.782814871016692e-05,
+      "loss": 0.4938,
+      "step": 2445
+    },
+    {
+      "epoch": 0.41819578390372963,
+      "grad_norm": 1.8300738320275258,
+      "learning_rate": 4.7812341932220537e-05,
+      "loss": 0.5105,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4190492446872066,
+      "grad_norm": 0.6481455070850186,
+      "learning_rate": 4.7796535154274154e-05,
+      "loss": 0.51,
+      "step": 2455
+    },
+    {
+      "epoch": 0.41990270547068365,
+      "grad_norm": 0.7913096084633933,
+      "learning_rate": 4.778072837632777e-05,
+      "loss": 0.4987,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4207561662541606,
+      "grad_norm": 1.2255098650997986,
+      "learning_rate": 4.776492159838139e-05,
+      "loss": 0.5009,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4216096270376376,
+      "grad_norm": 0.6795160628453747,
+      "learning_rate": 4.7749114820435006e-05,
+      "loss": 0.5186,
+      "step": 2470
+    },
+    {
+      "epoch": 0.42246308782111464,
+      "grad_norm": 0.4898195410241313,
+      "learning_rate": 4.7733308042488624e-05,
+      "loss": 0.495,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4233165486045916,
+      "grad_norm": 0.6029736724871774,
+      "learning_rate": 4.7717501264542235e-05,
+      "loss": 0.5097,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4241700093880686,
+      "grad_norm": 0.6029111635582989,
+      "learning_rate": 4.770169448659585e-05,
+      "loss": 0.4871,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4250234701715456,
+      "grad_norm": 0.49507690107864344,
+      "learning_rate": 4.7685887708649476e-05,
+      "loss": 0.4755,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4258769309550226,
+      "grad_norm": 0.6613373223792829,
+      "learning_rate": 4.767008093070309e-05,
+      "loss": 0.5154,
+      "step": 2495
+    },
+    {
+      "epoch": 0.42673039173849964,
+      "grad_norm": 0.49685624332129247,
+      "learning_rate": 4.7654274152756704e-05,
+      "loss": 0.4927,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4275838525219766,
+      "grad_norm": 0.4437325089367657,
+      "learning_rate": 4.763846737481032e-05,
+      "loss": 0.4792,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4284373133054536,
+      "grad_norm": 0.5576342725164943,
+      "learning_rate": 4.762266059686394e-05,
+      "loss": 0.5079,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4292907740889306,
+      "grad_norm": 0.46585229501787667,
+      "learning_rate": 4.760685381891755e-05,
+      "loss": 0.5006,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4301442348724076,
+      "grad_norm": 0.6845463777228801,
+      "learning_rate": 4.759104704097117e-05,
+      "loss": 0.5115,
+      "step": 2520
+    },
+    {
+      "epoch": 0.43099769565588464,
+      "grad_norm": 0.5057601413614787,
+      "learning_rate": 4.757524026302479e-05,
+      "loss": 0.4927,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4318511564393616,
+      "grad_norm": 0.5684767431715674,
+      "learning_rate": 4.75594334850784e-05,
+      "loss": 0.5094,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4327046172228386,
+      "grad_norm": 0.7288429534702419,
+      "learning_rate": 4.754362670713202e-05,
+      "loss": 0.5102,
+      "step": 2535
+    },
+    {
+      "epoch": 0.43355807800631563,
+      "grad_norm": 0.6797933337175003,
+      "learning_rate": 4.752781992918564e-05,
+      "loss": 0.5011,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4344115387897926,
+      "grad_norm": 0.7310663062589878,
+      "learning_rate": 4.7512013151239255e-05,
+      "loss": 0.5418,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4352649995732696,
+      "grad_norm": 1.048960668662016,
+      "learning_rate": 4.7496206373292865e-05,
+      "loss": 0.4918,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4361184603567466,
+      "grad_norm": 0.48307167584529137,
+      "learning_rate": 4.748039959534649e-05,
+      "loss": 0.4774,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4369719211402236,
+      "grad_norm": 0.5814531480371788,
+      "learning_rate": 4.746459281740011e-05,
+      "loss": 0.4581,
+      "step": 2560
+    },
+    {
+      "epoch": 0.43782538192370063,
+      "grad_norm": 0.42862234878740185,
+      "learning_rate": 4.744878603945372e-05,
+      "loss": 0.5208,
+      "step": 2565
+    },
+    {
+      "epoch": 0.4386788427071776,
+      "grad_norm": 0.4657196512442816,
+      "learning_rate": 4.7432979261507335e-05,
+      "loss": 0.5068,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4395323034906546,
+      "grad_norm": 0.5131288521906384,
+      "learning_rate": 4.741717248356095e-05,
+      "loss": 0.5498,
+      "step": 2575
+    },
+    {
+      "epoch": 0.4403857642741316,
+      "grad_norm": 0.5803920539575499,
+      "learning_rate": 4.740136570561457e-05,
+      "loss": 0.5143,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4412392250576086,
+      "grad_norm": 0.5181878370826999,
+      "learning_rate": 4.738555892766819e-05,
+      "loss": 0.4873,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4420926858410856,
+      "grad_norm": 0.6480775342678349,
+      "learning_rate": 4.7369752149721805e-05,
+      "loss": 0.4641,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4429461466245626,
+      "grad_norm": 0.4511955581917607,
+      "learning_rate": 4.735394537177542e-05,
+      "loss": 0.483,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4437996074080396,
+      "grad_norm": 0.44086238762815927,
+      "learning_rate": 4.733813859382903e-05,
+      "loss": 0.4931,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4446530681915166,
+      "grad_norm": 0.5191620095061635,
+      "learning_rate": 4.732233181588265e-05,
+      "loss": 0.4814,
+      "step": 2605
+    },
+    {
+      "epoch": 0.4455065289749936,
+      "grad_norm": 0.5151747201853522,
+      "learning_rate": 4.730652503793627e-05,
+      "loss": 0.506,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4463599897584706,
+      "grad_norm": 0.597319461144412,
+      "learning_rate": 4.7290718259989886e-05,
+      "loss": 0.5201,
+      "step": 2615
+    },
+    {
+      "epoch": 0.4472134505419476,
+      "grad_norm": 0.4871202179954294,
+      "learning_rate": 4.72749114820435e-05,
+      "loss": 0.4899,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4480669113254246,
+      "grad_norm": 0.6523872164581523,
+      "learning_rate": 4.725910470409712e-05,
+      "loss": 0.4962,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4489203721089016,
+      "grad_norm": 1.5505489626714442,
+      "learning_rate": 4.724329792615074e-05,
+      "loss": 0.5084,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4497738328923786,
+      "grad_norm": 0.5726929915232936,
+      "learning_rate": 4.722749114820435e-05,
+      "loss": 0.4919,
+      "step": 2635
+    },
+    {
+      "epoch": 0.4506272936758556,
+      "grad_norm": 0.4825017560375864,
+      "learning_rate": 4.7211684370257966e-05,
+      "loss": 0.4652,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4514807544593326,
+      "grad_norm": 0.530849848752361,
+      "learning_rate": 4.719587759231159e-05,
+      "loss": 0.5028,
+      "step": 2645
+    },
+    {
+      "epoch": 0.4523342152428096,
+      "grad_norm": 0.49903865492028343,
+      "learning_rate": 4.71800708143652e-05,
+      "loss": 0.545,
+      "step": 2650
+    },
+    {
+      "epoch": 0.45318767602628657,
+      "grad_norm": 0.43232891166576487,
+      "learning_rate": 4.716426403641882e-05,
+      "loss": 0.4847,
+      "step": 2655
+    },
+    {
+      "epoch": 0.4540411368097636,
+      "grad_norm": 0.4962269363045644,
+      "learning_rate": 4.7148457258472436e-05,
+      "loss": 0.495,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4548945975932406,
+      "grad_norm": 0.3904494391212347,
+      "learning_rate": 4.713265048052605e-05,
+      "loss": 0.5095,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4557480583767176,
+      "grad_norm": 0.4859232333362679,
+      "learning_rate": 4.7116843702579664e-05,
+      "loss": 0.499,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4566015191601946,
+      "grad_norm": 0.5253121206824946,
+      "learning_rate": 4.710103692463329e-05,
+      "loss": 0.512,
+      "step": 2675
+    },
+    {
+      "epoch": 0.45745497994367157,
+      "grad_norm": 0.44486107791609475,
+      "learning_rate": 4.7085230146686906e-05,
+      "loss": 0.4962,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4583084407271486,
+      "grad_norm": 2.0847784182788134,
+      "learning_rate": 4.7069423368740516e-05,
+      "loss": 0.4794,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4591619015106256,
+      "grad_norm": 0.49662006211171916,
+      "learning_rate": 4.7053616590794134e-05,
+      "loss": 0.4929,
+      "step": 2690
+    },
+    {
+      "epoch": 0.46001536229410256,
+      "grad_norm": 0.4588170270812257,
+      "learning_rate": 4.703780981284775e-05,
+      "loss": 0.5252,
+      "step": 2695
+    },
+    {
+      "epoch": 0.4608688230775796,
+      "grad_norm": 0.4238447364177512,
+      "learning_rate": 4.702200303490137e-05,
+      "loss": 0.4757,
+      "step": 2700
+    },
+    {
+      "epoch": 0.46172228386105657,
+      "grad_norm": 0.7349134241268634,
+      "learning_rate": 4.700619625695498e-05,
+      "loss": 0.4923,
+      "step": 2705
+    },
+    {
+      "epoch": 0.4625757446445336,
+      "grad_norm": 0.5620110128820872,
+      "learning_rate": 4.6990389479008604e-05,
+      "loss": 0.5046,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4634292054280106,
+      "grad_norm": 0.43117658804511005,
+      "learning_rate": 4.697458270106222e-05,
+      "loss": 0.4937,
+      "step": 2715
+    },
+    {
+      "epoch": 0.46428266621148756,
+      "grad_norm": 0.5258019562099506,
+      "learning_rate": 4.695877592311583e-05,
+      "loss": 0.5429,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4651361269949646,
+      "grad_norm": 0.7152247622932054,
+      "learning_rate": 4.694296914516945e-05,
+      "loss": 0.4617,
+      "step": 2725
+    },
+    {
+      "epoch": 0.46598958777844157,
+      "grad_norm": 0.5525572451632204,
+      "learning_rate": 4.692716236722307e-05,
+      "loss": 0.4953,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4668430485619186,
+      "grad_norm": 0.4592347168689562,
+      "learning_rate": 4.6911355589276684e-05,
+      "loss": 0.4777,
+      "step": 2735
+    },
+    {
+      "epoch": 0.4676965093453956,
+      "grad_norm": 0.4995452622352298,
+      "learning_rate": 4.68955488113303e-05,
+      "loss": 0.4842,
+      "step": 2740
+    },
+    {
+      "epoch": 0.46854997012887256,
+      "grad_norm": 0.4572243792841723,
+      "learning_rate": 4.687974203338392e-05,
+      "loss": 0.4788,
+      "step": 2745
+    },
+    {
+      "epoch": 0.4694034309123496,
+      "grad_norm": 0.4839684835554378,
+      "learning_rate": 4.6863935255437537e-05,
+      "loss": 0.4687,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4702568916958266,
+      "grad_norm": 0.47612394912228895,
+      "learning_rate": 4.684812847749115e-05,
+      "loss": 0.4854,
+      "step": 2755
+    },
+    {
+      "epoch": 0.47111035247930355,
+      "grad_norm": 0.7490520620929824,
+      "learning_rate": 4.6832321699544765e-05,
+      "loss": 0.5242,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4719638132627806,
+      "grad_norm": 0.47986695057589407,
+      "learning_rate": 4.681651492159838e-05,
+      "loss": 0.4879,
+      "step": 2765
+    },
+    {
+      "epoch": 0.47281727404625756,
+      "grad_norm": 0.4286048199133557,
+      "learning_rate": 4.6800708143652e-05,
+      "loss": 0.5156,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4736707348297346,
+      "grad_norm": 0.45170713322736494,
+      "learning_rate": 4.678490136570562e-05,
+      "loss": 0.4836,
+      "step": 2775
+    },
+    {
+      "epoch": 0.4745241956132116,
+      "grad_norm": 0.6234029954594431,
+      "learning_rate": 4.6769094587759235e-05,
+      "loss": 0.5,
+      "step": 2780
+    },
+    {
+      "epoch": 0.47537765639668855,
+      "grad_norm": 0.759643106803168,
+      "learning_rate": 4.675328780981285e-05,
+      "loss": 0.5208,
+      "step": 2785
+    },
+    {
+      "epoch": 0.4762311171801656,
+      "grad_norm": 1.8711982791961406,
+      "learning_rate": 4.673748103186646e-05,
+      "loss": 0.4743,
+      "step": 2790
+    },
+    {
+      "epoch": 0.47708457796364256,
+      "grad_norm": 0.5627847698519466,
+      "learning_rate": 4.672167425392009e-05,
+      "loss": 0.5029,
+      "step": 2795
+    },
+    {
+      "epoch": 0.47793803874711954,
+      "grad_norm": 0.6159815105546127,
+      "learning_rate": 4.6705867475973704e-05,
+      "loss": 0.512,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4787914995305966,
+      "grad_norm": 0.6492068132157409,
+      "learning_rate": 4.6690060698027315e-05,
+      "loss": 0.4964,
+      "step": 2805
+    },
+    {
+      "epoch": 0.47964496031407355,
+      "grad_norm": 0.4136980806781585,
+      "learning_rate": 4.667425392008093e-05,
+      "loss": 0.491,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4804984210975506,
+      "grad_norm": 0.6222877223645581,
+      "learning_rate": 4.665844714213455e-05,
+      "loss": 0.5039,
+      "step": 2815
+    },
+    {
+      "epoch": 0.48135188188102757,
+      "grad_norm": 0.42847295588776324,
+      "learning_rate": 4.664264036418817e-05,
+      "loss": 0.4892,
+      "step": 2820
+    },
+    {
+      "epoch": 0.48220534266450454,
+      "grad_norm": 0.47614961391215077,
+      "learning_rate": 4.662683358624178e-05,
+      "loss": 0.482,
+      "step": 2825
+    },
+    {
+      "epoch": 0.4830588034479816,
+      "grad_norm": 0.4725249173125883,
+      "learning_rate": 4.66110268082954e-05,
+      "loss": 0.5098,
+      "step": 2830
+    },
+    {
+      "epoch": 0.48391226423145856,
+      "grad_norm": 0.4697674688999553,
+      "learning_rate": 4.659522003034902e-05,
+      "loss": 0.5265,
+      "step": 2835
+    },
+    {
+      "epoch": 0.4847657250149356,
+      "grad_norm": 0.5982380189162461,
+      "learning_rate": 4.657941325240263e-05,
+      "loss": 0.4896,
+      "step": 2840
+    },
+    {
+      "epoch": 0.48561918579841257,
+      "grad_norm": 0.40133960414105363,
+      "learning_rate": 4.656360647445625e-05,
+      "loss": 0.5034,
+      "step": 2845
+    },
+    {
+      "epoch": 0.48647264658188955,
+      "grad_norm": 0.48525613870731715,
+      "learning_rate": 4.6547799696509865e-05,
+      "loss": 0.4745,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4873261073653666,
+      "grad_norm": 0.42933710960223315,
+      "learning_rate": 4.653199291856348e-05,
+      "loss": 0.5306,
+      "step": 2855
+    },
+    {
+      "epoch": 0.48817956814884356,
+      "grad_norm": 1.1063656861366253,
+      "learning_rate": 4.6516186140617094e-05,
+      "loss": 0.5048,
+      "step": 2860
+    },
+    {
+      "epoch": 0.48903302893232053,
+      "grad_norm": 0.5224278678720886,
+      "learning_rate": 4.650037936267072e-05,
+      "loss": 0.5039,
+      "step": 2865
+    },
+    {
+      "epoch": 0.48988648971579757,
+      "grad_norm": 0.5882672184596583,
+      "learning_rate": 4.6484572584724335e-05,
+      "loss": 0.4749,
+      "step": 2870
+    },
+    {
+      "epoch": 0.49073995049927455,
+      "grad_norm": 0.479519347249896,
+      "learning_rate": 4.6468765806777946e-05,
+      "loss": 0.5241,
+      "step": 2875
+    },
+    {
+      "epoch": 0.4915934112827516,
+      "grad_norm": 0.6728167983699064,
+      "learning_rate": 4.645295902883156e-05,
+      "loss": 0.5106,
+      "step": 2880
+    },
+    {
+      "epoch": 0.49244687206622856,
+      "grad_norm": 0.43211549845106334,
+      "learning_rate": 4.643715225088518e-05,
+      "loss": 0.4844,
+      "step": 2885
+    },
+    {
+      "epoch": 0.49330033284970554,
+      "grad_norm": 0.5784010762486059,
+      "learning_rate": 4.64213454729388e-05,
+      "loss": 0.4564,
+      "step": 2890
+    },
+    {
+      "epoch": 0.49415379363318257,
+      "grad_norm": 0.5342212683412771,
+      "learning_rate": 4.6405538694992416e-05,
+      "loss": 0.4614,
+      "step": 2895
+    },
+    {
+      "epoch": 0.49500725441665955,
+      "grad_norm": 0.6805185680905738,
+      "learning_rate": 4.638973191704603e-05,
+      "loss": 0.47,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4958607152001366,
+      "grad_norm": 0.5574973662984423,
+      "learning_rate": 4.637392513909965e-05,
+      "loss": 0.515,
+      "step": 2905
+    },
+    {
+      "epoch": 0.49671417598361356,
+      "grad_norm": 0.7435686355881591,
+      "learning_rate": 4.635811836115326e-05,
+      "loss": 0.4926,
+      "step": 2910
+    },
+    {
+      "epoch": 0.49756763676709054,
+      "grad_norm": 0.6874015511986851,
+      "learning_rate": 4.634231158320688e-05,
+      "loss": 0.4809,
+      "step": 2915
+    },
+    {
+      "epoch": 0.49842109755056757,
+      "grad_norm": 0.5383764757447236,
+      "learning_rate": 4.6326504805260496e-05,
+      "loss": 0.4861,
+      "step": 2920
+    },
+    {
+      "epoch": 0.49927455833404455,
+      "grad_norm": 0.4665152035108908,
+      "learning_rate": 4.6310698027314114e-05,
+      "loss": 0.4847,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5001280191175216,
+      "grad_norm": 0.5836806834915259,
+      "learning_rate": 4.629489124936773e-05,
+      "loss": 0.4998,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5009814799009985,
+      "grad_norm": 0.4361648214216843,
+      "learning_rate": 4.627908447142135e-05,
+      "loss": 0.4808,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5018349406844755,
+      "grad_norm": 0.49583680476975295,
+      "learning_rate": 4.6263277693474966e-05,
+      "loss": 0.4868,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5026884014679526,
+      "grad_norm": 0.4483956246468155,
+      "learning_rate": 4.624747091552858e-05,
+      "loss": 0.5048,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5035418622514295,
+      "grad_norm": 0.5294571252245484,
+      "learning_rate": 4.62316641375822e-05,
+      "loss": 0.5058,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5043953230349065,
+      "grad_norm": 0.5252090443214954,
+      "learning_rate": 4.621585735963581e-05,
+      "loss": 0.5169,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5052487838183836,
+      "grad_norm": 0.46760767827848043,
+      "learning_rate": 4.620005058168943e-05,
+      "loss": 0.4562,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5061022446018606,
+      "grad_norm": 0.8415597340930349,
+      "learning_rate": 4.6184243803743047e-05,
+      "loss": 0.4544,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5069557053853375,
+      "grad_norm": 0.6247587825872346,
+      "learning_rate": 4.6168437025796664e-05,
+      "loss": 0.4783,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5078091661688146,
+      "grad_norm": 0.43871676248593794,
+      "learning_rate": 4.615263024785028e-05,
+      "loss": 0.5082,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5086626269522916,
+      "grad_norm": 0.5148243512845597,
+      "learning_rate": 4.613682346990389e-05,
+      "loss": 0.5239,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5095160877357685,
+      "grad_norm": 0.5385515871310813,
+      "learning_rate": 4.6121016691957516e-05,
+      "loss": 0.516,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5103695485192455,
+      "grad_norm": 0.5138201657544148,
+      "learning_rate": 4.6105209914011134e-05,
+      "loss": 0.472,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5112230093027226,
+      "grad_norm": 0.4716007255261149,
+      "learning_rate": 4.6089403136064745e-05,
+      "loss": 0.5122,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5120764700861995,
+      "grad_norm": 0.5368887676130082,
+      "learning_rate": 4.607359635811836e-05,
+      "loss": 0.5264,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5129299308696765,
+      "grad_norm": 0.43583049564141374,
+      "learning_rate": 4.605778958017198e-05,
+      "loss": 0.4887,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5137833916531536,
+      "grad_norm": 0.5455177937923521,
+      "learning_rate": 4.60419828022256e-05,
+      "loss": 0.4742,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5146368524366305,
+      "grad_norm": 0.44161273126802225,
+      "learning_rate": 4.602617602427921e-05,
+      "loss": 0.4569,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5154903132201075,
+      "grad_norm": 0.5588901866324794,
+      "learning_rate": 4.601036924633283e-05,
+      "loss": 0.4948,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5163437740035846,
+      "grad_norm": 0.43294336422490237,
+      "learning_rate": 4.599456246838645e-05,
+      "loss": 0.4824,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5171972347870616,
+      "grad_norm": 0.4262908125239168,
+      "learning_rate": 4.597875569044006e-05,
+      "loss": 0.4674,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5180506955705385,
+      "grad_norm": 0.4134934926125551,
+      "learning_rate": 4.596294891249368e-05,
+      "loss": 0.478,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5189041563540155,
+      "grad_norm": 0.5146435082463794,
+      "learning_rate": 4.5947142134547295e-05,
+      "loss": 0.48,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5197576171374926,
+      "grad_norm": 0.5275388619042292,
+      "learning_rate": 4.593133535660091e-05,
+      "loss": 0.5283,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5206110779209695,
+      "grad_norm": 0.41285683367409964,
+      "learning_rate": 4.591552857865453e-05,
+      "loss": 0.4905,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5214645387044465,
+      "grad_norm": 0.4414766971805224,
+      "learning_rate": 4.589972180070815e-05,
+      "loss": 0.482,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5223179994879236,
+      "grad_norm": 0.5560472533327718,
+      "learning_rate": 4.5883915022761765e-05,
+      "loss": 0.4754,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5231714602714005,
+      "grad_norm": 0.47111606947459106,
+      "learning_rate": 4.5868108244815375e-05,
+      "loss": 0.4926,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5240249210548775,
+      "grad_norm": 0.433105229367798,
+      "learning_rate": 4.5852301466869e-05,
+      "loss": 0.4725,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5248783818383546,
+      "grad_norm": 0.4628901557642212,
+      "learning_rate": 4.583649468892261e-05,
+      "loss": 0.4785,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5257318426218315,
+      "grad_norm": 0.6049145619127523,
+      "learning_rate": 4.582068791097623e-05,
+      "loss": 0.4915,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5265853034053085,
+      "grad_norm": 0.6860256332793251,
+      "learning_rate": 4.5804881133029845e-05,
+      "loss": 0.5093,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5274387641887855,
+      "grad_norm": 0.4977844134723854,
+      "learning_rate": 4.578907435508346e-05,
+      "loss": 0.4954,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5282922249722625,
+      "grad_norm": 0.5739524858960392,
+      "learning_rate": 4.577326757713708e-05,
+      "loss": 0.4651,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5291456857557395,
+      "grad_norm": 0.5123685569383267,
+      "learning_rate": 4.575746079919069e-05,
+      "loss": 0.5025,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5299991465392165,
+      "grad_norm": 0.4796521979465388,
+      "learning_rate": 4.5741654021244315e-05,
+      "loss": 0.5068,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5308526073226936,
+      "grad_norm": 0.47308033426789425,
+      "learning_rate": 4.5725847243297926e-05,
+      "loss": 0.4909,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5317060681061705,
+      "grad_norm": 0.38443216822110887,
+      "learning_rate": 4.571004046535154e-05,
+      "loss": 0.4683,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5325595288896475,
+      "grad_norm": 0.5224091434335496,
+      "learning_rate": 4.569423368740516e-05,
+      "loss": 0.4884,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5334129896731246,
+      "grad_norm": 0.5429417668695532,
+      "learning_rate": 4.567842690945878e-05,
+      "loss": 0.4857,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5342664504566015,
+      "grad_norm": 0.41612083241826625,
+      "learning_rate": 4.5662620131512396e-05,
+      "loss": 0.4998,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5351199112400785,
+      "grad_norm": 0.5384787879906221,
+      "learning_rate": 4.5646813353566006e-05,
+      "loss": 0.5083,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5359733720235555,
+      "grad_norm": 0.5734922772320576,
+      "learning_rate": 4.563100657561963e-05,
+      "loss": 0.451,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5368268328070325,
+      "grad_norm": 0.4541825803148031,
+      "learning_rate": 4.561519979767325e-05,
+      "loss": 0.493,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5376802935905095,
+      "grad_norm": 0.4994486724205086,
+      "learning_rate": 4.559939301972686e-05,
+      "loss": 0.4685,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5385337543739865,
+      "grad_norm": 0.47708325681611224,
+      "learning_rate": 4.5583586241780476e-05,
+      "loss": 0.4609,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5393872151574635,
+      "grad_norm": 0.46817706499113726,
+      "learning_rate": 4.5567779463834094e-05,
+      "loss": 0.4758,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5402406759409405,
+      "grad_norm": 0.4902818781364108,
+      "learning_rate": 4.555197268588771e-05,
+      "loss": 0.5077,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5410941367244175,
+      "grad_norm": 0.45922084807176566,
+      "learning_rate": 4.553616590794132e-05,
+      "loss": 0.5119,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5419475975078946,
+      "grad_norm": 0.6712454332090487,
+      "learning_rate": 4.5520359129994946e-05,
+      "loss": 0.4795,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5428010582913715,
+      "grad_norm": 0.43605119685450705,
+      "learning_rate": 4.5504552352048563e-05,
+      "loss": 0.4677,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5436545190748485,
+      "grad_norm": 0.4925537431519938,
+      "learning_rate": 4.5488745574102174e-05,
+      "loss": 0.4914,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5445079798583256,
+      "grad_norm": 0.531654901487706,
+      "learning_rate": 4.54729387961558e-05,
+      "loss": 0.4965,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5453614406418025,
+      "grad_norm": 0.542627550461915,
+      "learning_rate": 4.545713201820941e-05,
+      "loss": 0.505,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5462149014252795,
+      "grad_norm": 0.5005989461431033,
+      "learning_rate": 4.5441325240263026e-05,
+      "loss": 0.4676,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5470683622087565,
+      "grad_norm": 0.8356311379255125,
+      "learning_rate": 4.5425518462316644e-05,
+      "loss": 0.4878,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5479218229922335,
+      "grad_norm": 0.6159083414874345,
+      "learning_rate": 4.540971168437026e-05,
+      "loss": 0.5381,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5487752837757105,
+      "grad_norm": 0.6463799250831206,
+      "learning_rate": 4.539390490642388e-05,
+      "loss": 0.5048,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5496287445591875,
+      "grad_norm": 0.514052351649874,
+      "learning_rate": 4.537809812847749e-05,
+      "loss": 0.4608,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5504822053426645,
+      "grad_norm": 0.43715311139472457,
+      "learning_rate": 4.5362291350531114e-05,
+      "loss": 0.5101,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5513356661261415,
+      "grad_norm": 0.6500744851584536,
+      "learning_rate": 4.5346484572584724e-05,
+      "loss": 0.488,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5521891269096185,
+      "grad_norm": 0.495178494171075,
+      "learning_rate": 4.533067779463834e-05,
+      "loss": 0.4887,
+      "step": 3235
+    },
+    {
+      "epoch": 0.5530425876930956,
+      "grad_norm": 0.798100029506453,
+      "learning_rate": 4.531487101669196e-05,
+      "loss": 0.491,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5538960484765725,
+      "grad_norm": 0.9902280186281872,
+      "learning_rate": 4.529906423874558e-05,
+      "loss": 0.4627,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5547495092600495,
+      "grad_norm": 1.7680027677139394,
+      "learning_rate": 4.5283257460799194e-05,
+      "loss": 0.4694,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5556029700435265,
+      "grad_norm": 0.6487867409422882,
+      "learning_rate": 4.5267450682852805e-05,
+      "loss": 0.5086,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5564564308270035,
+      "grad_norm": 0.4778060701450197,
+      "learning_rate": 4.525164390490643e-05,
+      "loss": 0.4672,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5573098916104805,
+      "grad_norm": 0.6933757779273402,
+      "learning_rate": 4.523583712696004e-05,
+      "loss": 0.4947,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5581633523939575,
+      "grad_norm": 0.8292966610796421,
+      "learning_rate": 4.522003034901366e-05,
+      "loss": 0.4817,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5590168131774345,
+      "grad_norm": 0.5753336708545032,
+      "learning_rate": 4.5204223571067275e-05,
+      "loss": 0.4603,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5598702739609115,
+      "grad_norm": 0.7087010804750883,
+      "learning_rate": 4.518841679312089e-05,
+      "loss": 0.4935,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5607237347443885,
+      "grad_norm": 0.4609635531108082,
+      "learning_rate": 4.517261001517451e-05,
+      "loss": 0.4767,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5615771955278654,
+      "grad_norm": 1.0131378296218625,
+      "learning_rate": 4.515680323722812e-05,
+      "loss": 0.4919,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5624306563113425,
+      "grad_norm": 0.5286009061837952,
+      "learning_rate": 4.5140996459281745e-05,
+      "loss": 0.4629,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5632841170948195,
+      "grad_norm": 0.5233857128596109,
+      "learning_rate": 4.512518968133536e-05,
+      "loss": 0.4787,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5641375778782965,
+      "grad_norm": 0.8557917738316622,
+      "learning_rate": 4.510938290338897e-05,
+      "loss": 0.4996,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5649910386617735,
+      "grad_norm": 0.4526387213257986,
+      "learning_rate": 4.50935761254426e-05,
+      "loss": 0.504,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5658444994452505,
+      "grad_norm": 2.9307163982033657,
+      "learning_rate": 4.507776934749621e-05,
+      "loss": 0.4833,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5666979602287275,
+      "grad_norm": 0.5201434642577162,
+      "learning_rate": 4.5061962569549825e-05,
+      "loss": 0.4934,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5675514210122045,
+      "grad_norm": 0.49175135020457955,
+      "learning_rate": 4.504615579160344e-05,
+      "loss": 0.5206,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5684048817956815,
+      "grad_norm": 0.3878627218856043,
+      "learning_rate": 4.503034901365706e-05,
+      "loss": 0.463,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5692583425791585,
+      "grad_norm": 0.38549872056761764,
+      "learning_rate": 4.501454223571068e-05,
+      "loss": 0.4836,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5701118033626354,
+      "grad_norm": 0.5321198076144228,
+      "learning_rate": 4.499873545776429e-05,
+      "loss": 0.4712,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5709652641461125,
+      "grad_norm": 0.4262011189522238,
+      "learning_rate": 4.498292867981791e-05,
+      "loss": 0.4926,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5718187249295895,
+      "grad_norm": 0.46600041458591407,
+      "learning_rate": 4.496712190187152e-05,
+      "loss": 0.4785,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5726721857130664,
+      "grad_norm": 0.6200806088331268,
+      "learning_rate": 4.495131512392514e-05,
+      "loss": 0.4992,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5735256464965435,
+      "grad_norm": 0.9996035433021478,
+      "learning_rate": 4.493550834597876e-05,
+      "loss": 0.4715,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5743791072800205,
+      "grad_norm": 0.49082524222147683,
+      "learning_rate": 4.4919701568032375e-05,
+      "loss": 0.4718,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5752325680634974,
+      "grad_norm": 0.8837435528802063,
+      "learning_rate": 4.490389479008599e-05,
+      "loss": 0.4734,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5760860288469745,
+      "grad_norm": 1.7445854952652542,
+      "learning_rate": 4.4888088012139604e-05,
+      "loss": 0.4883,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5769394896304515,
+      "grad_norm": 0.6301167609521606,
+      "learning_rate": 4.487228123419323e-05,
+      "loss": 0.4933,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5777929504139285,
+      "grad_norm": 0.3939525490223246,
+      "learning_rate": 4.485647445624684e-05,
+      "loss": 0.4684,
+      "step": 3385
+    },
+    {
+      "epoch": 0.5786464111974055,
+      "grad_norm": 0.46250171089281755,
+      "learning_rate": 4.4840667678300456e-05,
+      "loss": 0.4897,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5794998719808825,
+      "grad_norm": 0.45920802219487256,
+      "learning_rate": 4.4824860900354073e-05,
+      "loss": 0.5029,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5803533327643595,
+      "grad_norm": 0.4819306175315199,
+      "learning_rate": 4.480905412240769e-05,
+      "loss": 0.4412,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5812067935478364,
+      "grad_norm": 0.4256152645527321,
+      "learning_rate": 4.479324734446131e-05,
+      "loss": 0.444,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5820602543313135,
+      "grad_norm": 0.42620197572305174,
+      "learning_rate": 4.477744056651492e-05,
+      "loss": 0.4685,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5829137151147905,
+      "grad_norm": 0.5550364644460304,
+      "learning_rate": 4.476163378856854e-05,
+      "loss": 0.4907,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5837671758982674,
+      "grad_norm": 0.45984600307346984,
+      "learning_rate": 4.4745827010622154e-05,
+      "loss": 0.4982,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5846206366817445,
+      "grad_norm": 0.689863334490102,
+      "learning_rate": 4.473002023267577e-05,
+      "loss": 0.4893,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5854740974652215,
+      "grad_norm": 0.5275670181218156,
+      "learning_rate": 4.4714213454729396e-05,
+      "loss": 0.5077,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5863275582486984,
+      "grad_norm": 0.5796184441646595,
+      "learning_rate": 4.4698406676783006e-05,
+      "loss": 0.4639,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5871810190321755,
+      "grad_norm": 0.45709870220036597,
+      "learning_rate": 4.4682599898836624e-05,
+      "loss": 0.4673,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5880344798156525,
+      "grad_norm": 0.4538463488322591,
+      "learning_rate": 4.466679312089024e-05,
+      "loss": 0.4873,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5888879405991295,
+      "grad_norm": 0.44323971254558264,
+      "learning_rate": 4.465098634294386e-05,
+      "loss": 0.4845,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5897414013826064,
+      "grad_norm": 0.4801465832311005,
+      "learning_rate": 4.4635179564997476e-05,
+      "loss": 0.4614,
+      "step": 3455
+    },
+    {
+      "epoch": 0.5905948621660835,
+      "grad_norm": 0.40555326089246163,
+      "learning_rate": 4.461937278705109e-05,
+      "loss": 0.4798,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5914483229495605,
+      "grad_norm": 0.5648457333532083,
+      "learning_rate": 4.460356600910471e-05,
+      "loss": 0.4583,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5923017837330374,
+      "grad_norm": 0.4626371797805737,
+      "learning_rate": 4.458775923115832e-05,
+      "loss": 0.4556,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5931552445165145,
+      "grad_norm": 0.4645009958255317,
+      "learning_rate": 4.457195245321194e-05,
+      "loss": 0.4429,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5940087052999915,
+      "grad_norm": 0.5553987660550683,
+      "learning_rate": 4.455614567526556e-05,
+      "loss": 0.5151,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5948621660834684,
+      "grad_norm": 0.526142044453019,
+      "learning_rate": 4.4540338897319174e-05,
+      "loss": 0.4707,
+      "step": 3485
+    },
+    {
+      "epoch": 0.5957156268669455,
+      "grad_norm": 0.5645287259798881,
+      "learning_rate": 4.452453211937279e-05,
+      "loss": 0.5235,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5965690876504225,
+      "grad_norm": 0.46749936432388794,
+      "learning_rate": 4.45087253414264e-05,
+      "loss": 0.4797,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5974225484338994,
+      "grad_norm": 0.7045403748184238,
+      "learning_rate": 4.4492918563480026e-05,
+      "loss": 0.4866,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5982760092173764,
+      "grad_norm": 0.4342980770279321,
+      "learning_rate": 4.447711178553364e-05,
+      "loss": 0.4775,
+      "step": 3505
+    },
+    {
+      "epoch": 0.5991294700008535,
+      "grad_norm": 0.4806934305956506,
+      "learning_rate": 4.4461305007587255e-05,
+      "loss": 0.4686,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5999829307843305,
+      "grad_norm": 63.95029639895003,
+      "learning_rate": 4.444549822964087e-05,
+      "loss": 0.4532,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6008363915678074,
+      "grad_norm": 0.500224009602108,
+      "learning_rate": 4.442969145169449e-05,
+      "loss": 0.4426,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6016898523512845,
+      "grad_norm": 0.50512012301229,
+      "learning_rate": 4.441388467374811e-05,
+      "loss": 0.4478,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6025433131347615,
+      "grad_norm": 0.484096459835353,
+      "learning_rate": 4.439807789580172e-05,
+      "loss": 0.452,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6033967739182384,
+      "grad_norm": 0.4688040856385147,
+      "learning_rate": 4.438227111785534e-05,
+      "loss": 0.5168,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6042502347017155,
+      "grad_norm": 0.6679582478264732,
+      "learning_rate": 4.436646433990895e-05,
+      "loss": 0.4672,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6051036954851925,
+      "grad_norm": 0.47770775490769585,
+      "learning_rate": 4.435065756196257e-05,
+      "loss": 0.5019,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6059571562686694,
+      "grad_norm": 0.3880046321693871,
+      "learning_rate": 4.433485078401619e-05,
+      "loss": 0.4797,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6068106170521465,
+      "grad_norm": 0.5414538314833576,
+      "learning_rate": 4.4319044006069805e-05,
+      "loss": 0.4706,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6076640778356235,
+      "grad_norm": 0.4809579676224301,
+      "learning_rate": 4.430323722812342e-05,
+      "loss": 0.4838,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6085175386191004,
+      "grad_norm": 0.48144765077507257,
+      "learning_rate": 4.428743045017704e-05,
+      "loss": 0.5095,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6093709994025774,
+      "grad_norm": 0.5138696069532919,
+      "learning_rate": 4.427162367223066e-05,
+      "loss": 0.4944,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6102244601860545,
+      "grad_norm": 0.45587797333607455,
+      "learning_rate": 4.425581689428427e-05,
+      "loss": 0.4901,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6110779209695314,
+      "grad_norm": 0.48425921551782086,
+      "learning_rate": 4.4240010116337885e-05,
+      "loss": 0.4403,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6119313817530084,
+      "grad_norm": 0.43899108915854307,
+      "learning_rate": 4.422420333839151e-05,
+      "loss": 0.4636,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6127848425364855,
+      "grad_norm": 0.5141934864853146,
+      "learning_rate": 4.420839656044512e-05,
+      "loss": 0.4854,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6136383033199625,
+      "grad_norm": 0.5051172565590736,
+      "learning_rate": 4.419258978249874e-05,
+      "loss": 0.4811,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6144917641034394,
+      "grad_norm": 0.43738734095879495,
+      "learning_rate": 4.4176783004552355e-05,
+      "loss": 0.5052,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6153452248869165,
+      "grad_norm": 0.39331118925627184,
+      "learning_rate": 4.416097622660597e-05,
+      "loss": 0.4798,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6161986856703935,
+      "grad_norm": 0.4133684384057315,
+      "learning_rate": 4.4145169448659583e-05,
+      "loss": 0.4894,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6170521464538704,
+      "grad_norm": 0.45545934534082405,
+      "learning_rate": 4.41293626707132e-05,
+      "loss": 0.5195,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6179056072373474,
+      "grad_norm": 0.7681187073171971,
+      "learning_rate": 4.4113555892766825e-05,
+      "loss": 0.4689,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6187590680208245,
+      "grad_norm": 0.4325591179904035,
+      "learning_rate": 4.4097749114820436e-05,
+      "loss": 0.523,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6196125288043014,
+      "grad_norm": 0.5070197418845998,
+      "learning_rate": 4.408194233687405e-05,
+      "loss": 0.4583,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6204659895877784,
+      "grad_norm": 0.45243291995651214,
+      "learning_rate": 4.406613555892767e-05,
+      "loss": 0.4922,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6213194503712555,
+      "grad_norm": 0.449732217059902,
+      "learning_rate": 4.405032878098129e-05,
+      "loss": 0.4699,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6221729111547324,
+      "grad_norm": 0.7246861327311569,
+      "learning_rate": 4.4034522003034906e-05,
+      "loss": 0.4998,
+      "step": 3645
+    },
+    {
+      "epoch": 0.6230263719382094,
+      "grad_norm": 0.4280068084173374,
+      "learning_rate": 4.4018715225088516e-05,
+      "loss": 0.4978,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6238798327216865,
+      "grad_norm": 0.4014939390871873,
+      "learning_rate": 4.400290844714214e-05,
+      "loss": 0.474,
+      "step": 3655
+    },
+    {
+      "epoch": 0.6247332935051635,
+      "grad_norm": 0.4497317854323919,
+      "learning_rate": 4.398710166919575e-05,
+      "loss": 0.4929,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6255867542886404,
+      "grad_norm": 0.36552726366718447,
+      "learning_rate": 4.397129489124937e-05,
+      "loss": 0.4726,
+      "step": 3665
+    },
+    {
+      "epoch": 0.6264402150721174,
+      "grad_norm": 0.4328997047135238,
+      "learning_rate": 4.3955488113302986e-05,
+      "loss": 0.5043,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6272936758555945,
+      "grad_norm": 0.7066126750337065,
+      "learning_rate": 4.3939681335356604e-05,
+      "loss": 0.4773,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6281471366390714,
+      "grad_norm": 0.47512560430075906,
+      "learning_rate": 4.392387455741022e-05,
+      "loss": 0.4426,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6290005974225484,
+      "grad_norm": 0.42252564493650935,
+      "learning_rate": 4.390806777946383e-05,
+      "loss": 0.452,
+      "step": 3685
+    },
+    {
+      "epoch": 0.6298540582060255,
+      "grad_norm": 0.43013539904554504,
+      "learning_rate": 4.3892261001517456e-05,
+      "loss": 0.4888,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6307075189895024,
+      "grad_norm": 0.5071874143036144,
+      "learning_rate": 4.387645422357107e-05,
+      "loss": 0.479,
+      "step": 3695
+    },
+    {
+      "epoch": 0.6315609797729794,
+      "grad_norm": 0.5470278547942871,
+      "learning_rate": 4.3860647445624684e-05,
+      "loss": 0.478,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6324144405564565,
+      "grad_norm": 0.4167734022580235,
+      "learning_rate": 4.384484066767831e-05,
+      "loss": 0.4917,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6332679013399334,
+      "grad_norm": 0.4890306405293852,
+      "learning_rate": 4.382903388973192e-05,
+      "loss": 0.4493,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6341213621234104,
+      "grad_norm": 0.5470391911698399,
+      "learning_rate": 4.3813227111785537e-05,
+      "loss": 0.476,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6349748229068874,
+      "grad_norm": 0.47745469088807063,
+      "learning_rate": 4.3797420333839154e-05,
+      "loss": 0.4793,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6358282836903645,
+      "grad_norm": 0.5113844420452976,
+      "learning_rate": 4.378161355589277e-05,
+      "loss": 0.4656,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6366817444738414,
+      "grad_norm": 0.4458910221298414,
+      "learning_rate": 4.376580677794638e-05,
+      "loss": 0.4776,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6375352052573184,
+      "grad_norm": 0.4818584623158423,
+      "learning_rate": 4.375e-05,
+      "loss": 0.4918,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6383886660407955,
+      "grad_norm": 0.4492848878575235,
+      "learning_rate": 4.3734193222053624e-05,
+      "loss": 0.508,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6392421268242724,
+      "grad_norm": 0.45574504864146553,
+      "learning_rate": 4.3718386444107234e-05,
+      "loss": 0.4764,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6400955876077494,
+      "grad_norm": 0.42562056908018,
+      "learning_rate": 4.370257966616085e-05,
+      "loss": 0.5007,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6409490483912265,
+      "grad_norm": 0.708859438990614,
+      "learning_rate": 4.368677288821447e-05,
+      "loss": 0.4744,
+      "step": 3755
+    },
+    {
+      "epoch": 0.6418025091747034,
+      "grad_norm": 0.5326165948781,
+      "learning_rate": 4.367096611026809e-05,
+      "loss": 0.5026,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6426559699581804,
+      "grad_norm": 0.39341698760922,
+      "learning_rate": 4.36551593323217e-05,
+      "loss": 0.4491,
+      "step": 3765
+    },
+    {
+      "epoch": 0.6435094307416575,
+      "grad_norm": 0.40912745072852724,
+      "learning_rate": 4.3639352554375315e-05,
+      "loss": 0.4642,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6443628915251344,
+      "grad_norm": 0.4144693996623863,
+      "learning_rate": 4.362354577642894e-05,
+      "loss": 0.4859,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6452163523086114,
+      "grad_norm": 0.4881292009372189,
+      "learning_rate": 4.360773899848255e-05,
+      "loss": 0.4984,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6460698130920884,
+      "grad_norm": 0.45053933392519724,
+      "learning_rate": 4.359193222053617e-05,
+      "loss": 0.4768,
+      "step": 3785
+    },
+    {
+      "epoch": 0.6469232738755654,
+      "grad_norm": 0.47325450782615,
+      "learning_rate": 4.3576125442589785e-05,
+      "loss": 0.4493,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6477767346590424,
+      "grad_norm": 0.48811233595627407,
+      "learning_rate": 4.35603186646434e-05,
+      "loss": 0.4935,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6486301954425194,
+      "grad_norm": 0.421010723444544,
+      "learning_rate": 4.354451188669702e-05,
+      "loss": 0.4689,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6494836562259965,
+      "grad_norm": 0.5627086061065368,
+      "learning_rate": 4.352870510875063e-05,
+      "loss": 0.4753,
+      "step": 3805
+    },
+    {
+      "epoch": 0.6503371170094734,
+      "grad_norm": 0.4463014770154561,
+      "learning_rate": 4.3512898330804255e-05,
+      "loss": 0.4884,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6511905777929504,
+      "grad_norm": 0.4638012220494922,
+      "learning_rate": 4.3497091552857865e-05,
+      "loss": 0.4851,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6520440385764275,
+      "grad_norm": 0.7544906461658837,
+      "learning_rate": 4.348128477491148e-05,
+      "loss": 0.4444,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6528974993599044,
+      "grad_norm": 0.44167391706768905,
+      "learning_rate": 4.34654779969651e-05,
+      "loss": 0.4799,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6537509601433814,
+      "grad_norm": 0.43567757946878566,
+      "learning_rate": 4.344967121901872e-05,
+      "loss": 0.4941,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6546044209268584,
+      "grad_norm": 0.504526904318,
+      "learning_rate": 4.3433864441072335e-05,
+      "loss": 0.4414,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6554578817103354,
+      "grad_norm": 0.42110493335879345,
+      "learning_rate": 4.341805766312595e-05,
+      "loss": 0.4514,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6563113424938124,
+      "grad_norm": 0.3854733011231964,
+      "learning_rate": 4.340225088517957e-05,
+      "loss": 0.4715,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6571648032772894,
+      "grad_norm": 0.6822309966180405,
+      "learning_rate": 4.338644410723318e-05,
+      "loss": 0.4958,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6580182640607664,
+      "grad_norm": 0.5186683164282136,
+      "learning_rate": 4.33706373292868e-05,
+      "loss": 0.4722,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6588717248442434,
+      "grad_norm": 0.47780293082012615,
+      "learning_rate": 4.335483055134042e-05,
+      "loss": 0.5217,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6597251856277204,
+      "grad_norm": 0.4764829465540424,
+      "learning_rate": 4.333902377339403e-05,
+      "loss": 0.4605,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6605786464111975,
+      "grad_norm": 0.4541715815782311,
+      "learning_rate": 4.332321699544765e-05,
+      "loss": 0.4812,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6614321071946744,
+      "grad_norm": 0.5603266519555172,
+      "learning_rate": 4.330741021750127e-05,
+      "loss": 0.4651,
+      "step": 3875
+    },
+    {
+      "epoch": 0.6622855679781514,
+      "grad_norm": 0.48041672062511726,
+      "learning_rate": 4.3291603439554885e-05,
+      "loss": 0.4824,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6631390287616284,
+      "grad_norm": 0.47907954922802726,
+      "learning_rate": 4.3275796661608496e-05,
+      "loss": 0.4608,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6639924895451054,
+      "grad_norm": 0.44609812672729476,
+      "learning_rate": 4.3259989883662114e-05,
+      "loss": 0.462,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6648459503285824,
+      "grad_norm": 0.7993792487509449,
+      "learning_rate": 4.324418310571574e-05,
+      "loss": 0.4719,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6656994111120594,
+      "grad_norm": 0.4311820933387167,
+      "learning_rate": 4.322837632776935e-05,
+      "loss": 0.465,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6665528718955364,
+      "grad_norm": 0.4388514679700474,
+      "learning_rate": 4.3212569549822966e-05,
+      "loss": 0.4778,
+      "step": 3905
+    },
+    {
+      "epoch": 0.6674063326790134,
+      "grad_norm": 0.4575188747935784,
+      "learning_rate": 4.3196762771876583e-05,
+      "loss": 0.4956,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6682597934624904,
+      "grad_norm": 0.8660842719376415,
+      "learning_rate": 4.31809559939302e-05,
+      "loss": 0.4726,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6691132542459673,
+      "grad_norm": 0.5516487261948204,
+      "learning_rate": 4.316514921598381e-05,
+      "loss": 0.4762,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6699667150294444,
+      "grad_norm": 0.47455394040106036,
+      "learning_rate": 4.314934243803743e-05,
+      "loss": 0.4604,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6708201758129214,
+      "grad_norm": 0.5207320878622663,
+      "learning_rate": 4.313353566009105e-05,
+      "loss": 0.4804,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6716736365963984,
+      "grad_norm": 0.47742075584221105,
+      "learning_rate": 4.3117728882144664e-05,
+      "loss": 0.4556,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6725270973798754,
+      "grad_norm": 0.45314483542298994,
+      "learning_rate": 4.310192210419828e-05,
+      "loss": 0.4429,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6733805581633524,
+      "grad_norm": 0.4381070559951893,
+      "learning_rate": 4.30861153262519e-05,
+      "loss": 0.4627,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6742340189468294,
+      "grad_norm": 0.4071218465306306,
+      "learning_rate": 4.3070308548305516e-05,
+      "loss": 0.4602,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6750874797303064,
+      "grad_norm": 0.6499999822102005,
+      "learning_rate": 4.3054501770359134e-05,
+      "loss": 0.4684,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6759409405137834,
+      "grad_norm": 0.4846648689655542,
+      "learning_rate": 4.303869499241275e-05,
+      "loss": 0.4667,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6767944012972604,
+      "grad_norm": 0.4062643508656444,
+      "learning_rate": 4.302288821446637e-05,
+      "loss": 0.471,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6776478620807374,
+      "grad_norm": 0.4172505487094595,
+      "learning_rate": 4.300708143651998e-05,
+      "loss": 0.4447,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6785013228642144,
+      "grad_norm": 0.517751315224531,
+      "learning_rate": 4.29912746585736e-05,
+      "loss": 0.4495,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6793547836476914,
+      "grad_norm": 0.521540079831743,
+      "learning_rate": 4.2975467880627214e-05,
+      "loss": 0.4749,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6802082444311683,
+      "grad_norm": 0.5457293452763281,
+      "learning_rate": 4.295966110268083e-05,
+      "loss": 0.485,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6810617052146454,
+      "grad_norm": 0.4501647552450407,
+      "learning_rate": 4.294385432473445e-05,
+      "loss": 0.4869,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6819151659981224,
+      "grad_norm": 0.518406725291088,
+      "learning_rate": 4.292804754678807e-05,
+      "loss": 0.4719,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6827686267815993,
+      "grad_norm": 0.4457480223254278,
+      "learning_rate": 4.2912240768841684e-05,
+      "loss": 0.466,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6836220875650764,
+      "grad_norm": 0.6292094801052758,
+      "learning_rate": 4.2896433990895295e-05,
+      "loss": 0.5088,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6844755483485534,
+      "grad_norm": 0.5116550207377446,
+      "learning_rate": 4.288062721294891e-05,
+      "loss": 0.4723,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6853290091320304,
+      "grad_norm": 0.46845065186174567,
+      "learning_rate": 4.286482043500253e-05,
+      "loss": 0.4818,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6861824699155074,
+      "grad_norm": 0.7301404751357614,
+      "learning_rate": 4.284901365705615e-05,
+      "loss": 0.4917,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6870359306989844,
+      "grad_norm": 0.44690134946083243,
+      "learning_rate": 4.2833206879109765e-05,
+      "loss": 0.4556,
+      "step": 4025
+    },
+    {
+      "epoch": 0.6878893914824614,
+      "grad_norm": 0.413374957143859,
+      "learning_rate": 4.281740010116338e-05,
+      "loss": 0.4586,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6887428522659383,
+      "grad_norm": 0.4497745580402003,
+      "learning_rate": 4.2801593323217e-05,
+      "loss": 0.4523,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6895963130494154,
+      "grad_norm": 0.5756425206553233,
+      "learning_rate": 4.278578654527061e-05,
+      "loss": 0.4698,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6904497738328924,
+      "grad_norm": 0.39328406249104275,
+      "learning_rate": 4.276997976732423e-05,
+      "loss": 0.4576,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6913032346163693,
+      "grad_norm": 0.5660539542345869,
+      "learning_rate": 4.275417298937785e-05,
+      "loss": 0.4913,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6921566953998464,
+      "grad_norm": 0.7728511153118378,
+      "learning_rate": 4.273836621143146e-05,
+      "loss": 0.4834,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6930101561833234,
+      "grad_norm": 0.5208945931014687,
+      "learning_rate": 4.272255943348508e-05,
+      "loss": 0.4686,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6938636169668003,
+      "grad_norm": 0.4279123272455591,
+      "learning_rate": 4.27067526555387e-05,
+      "loss": 0.4632,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6947170777502774,
+      "grad_norm": 0.43309100601295614,
+      "learning_rate": 4.2690945877592315e-05,
+      "loss": 0.4789,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6955705385337544,
+      "grad_norm": 0.4800452561798458,
+      "learning_rate": 4.2675139099645926e-05,
+      "loss": 0.4556,
+      "step": 4075
+    },
+    {
+      "epoch": 0.6964239993172314,
+      "grad_norm": 0.43338273426642704,
+      "learning_rate": 4.265933232169955e-05,
+      "loss": 0.4703,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6972774601007083,
+      "grad_norm": 0.5380585209184109,
+      "learning_rate": 4.264352554375317e-05,
+      "loss": 0.4503,
+      "step": 4085
+    },
+    {
+      "epoch": 0.6981309208841854,
+      "grad_norm": 0.5593367082780836,
+      "learning_rate": 4.262771876580678e-05,
+      "loss": 0.4739,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6989843816676624,
+      "grad_norm": 0.4153747243000154,
+      "learning_rate": 4.2611911987860396e-05,
+      "loss": 0.5017,
+      "step": 4095
+    },
+    {
+      "epoch": 0.6998378424511393,
+      "grad_norm": 0.6526841312322189,
+      "learning_rate": 4.259610520991401e-05,
+      "loss": 0.4988,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7006913032346164,
+      "grad_norm": 0.5092368692530734,
+      "learning_rate": 4.258029843196763e-05,
+      "loss": 0.4733,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7015447640180934,
+      "grad_norm": 0.6272371729345999,
+      "learning_rate": 4.256449165402125e-05,
+      "loss": 0.4739,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7023982248015703,
+      "grad_norm": 0.7150201654065392,
+      "learning_rate": 4.2548684876074865e-05,
+      "loss": 0.4818,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7032516855850474,
+      "grad_norm": 0.6606867428972878,
+      "learning_rate": 4.253287809812848e-05,
+      "loss": 0.4625,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7041051463685244,
+      "grad_norm": 2.4043603785982803,
+      "learning_rate": 4.2517071320182093e-05,
+      "loss": 0.4724,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7049586071520013,
+      "grad_norm": 0.7085148116148299,
+      "learning_rate": 4.250126454223571e-05,
+      "loss": 0.498,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7058120679354783,
+      "grad_norm": 0.5309349094860171,
+      "learning_rate": 4.248545776428933e-05,
+      "loss": 0.4711,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7066655287189554,
+      "grad_norm": 1.4124045053824865,
+      "learning_rate": 4.2469650986342946e-05,
+      "loss": 0.4619,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7075189895024324,
+      "grad_norm": 0.5323652304800572,
+      "learning_rate": 4.245384420839656e-05,
+      "loss": 0.4813,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7083724502859093,
+      "grad_norm": 0.584382089693229,
+      "learning_rate": 4.243803743045018e-05,
+      "loss": 0.4609,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7092259110693864,
+      "grad_norm": 0.7538382751874487,
+      "learning_rate": 4.24222306525038e-05,
+      "loss": 0.4981,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7100793718528634,
+      "grad_norm": 0.6751026260244125,
+      "learning_rate": 4.240642387455741e-05,
+      "loss": 0.5001,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7109328326363403,
+      "grad_norm": 0.5439814041009411,
+      "learning_rate": 4.2390617096611026e-05,
+      "loss": 0.4721,
+      "step": 4165
+    },
+    {
+      "epoch": 0.7117862934198174,
+      "grad_norm": 0.5032188165856872,
+      "learning_rate": 4.2374810318664644e-05,
+      "loss": 0.4666,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7126397542032944,
+      "grad_norm": 0.6113467103393669,
+      "learning_rate": 4.235900354071826e-05,
+      "loss": 0.4541,
+      "step": 4175
+    },
+    {
+      "epoch": 0.7134932149867713,
+      "grad_norm": 0.5201854878427893,
+      "learning_rate": 4.234319676277188e-05,
+      "loss": 0.4756,
+      "step": 4180
+    },
+    {
+      "epoch": 0.7143466757702484,
+      "grad_norm": 0.4910717521923143,
+      "learning_rate": 4.2327389984825496e-05,
+      "loss": 0.4675,
+      "step": 4185
+    },
+    {
+      "epoch": 0.7152001365537254,
+      "grad_norm": 12.982698067516175,
+      "learning_rate": 4.2311583206879114e-05,
+      "loss": 0.4653,
+      "step": 4190
+    },
+    {
+      "epoch": 0.7160535973372023,
+      "grad_norm": 0.5000586078973243,
+      "learning_rate": 4.2295776428932724e-05,
+      "loss": 0.517,
+      "step": 4195
+    },
+    {
+      "epoch": 0.7169070581206793,
+      "grad_norm": 0.6077145038020854,
+      "learning_rate": 4.227996965098635e-05,
+      "loss": 0.4655,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7177605189041564,
+      "grad_norm": 0.4708526675368424,
+      "learning_rate": 4.2264162873039966e-05,
+      "loss": 0.4888,
+      "step": 4205
+    },
+    {
+      "epoch": 0.7186139796876334,
+      "grad_norm": 0.43314402639331606,
+      "learning_rate": 4.224835609509358e-05,
+      "loss": 0.4418,
+      "step": 4210
+    },
+    {
+      "epoch": 0.7194674404711103,
+      "grad_norm": 0.4044626975468846,
+      "learning_rate": 4.2232549317147194e-05,
+      "loss": 0.455,
+      "step": 4215
+    },
+    {
+      "epoch": 0.7203209012545874,
+      "grad_norm": 0.533450940376835,
+      "learning_rate": 4.221674253920081e-05,
+      "loss": 0.4699,
+      "step": 4220
+    },
+    {
+      "epoch": 0.7211743620380644,
+      "grad_norm": 1.367143406475424,
+      "learning_rate": 4.220093576125443e-05,
+      "loss": 0.4934,
+      "step": 4225
+    },
+    {
+      "epoch": 0.7220278228215413,
+      "grad_norm": 0.48597158199842877,
+      "learning_rate": 4.218512898330804e-05,
+      "loss": 0.4682,
+      "step": 4230
+    },
+    {
+      "epoch": 0.7228812836050184,
+      "grad_norm": 0.7095642481694201,
+      "learning_rate": 4.2169322205361664e-05,
+      "loss": 0.4755,
+      "step": 4235
+    },
+    {
+      "epoch": 0.7237347443884954,
+      "grad_norm": 0.6921834984338361,
+      "learning_rate": 4.215351542741528e-05,
+      "loss": 0.4764,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7245882051719723,
+      "grad_norm": 0.4669850856765737,
+      "learning_rate": 4.213770864946889e-05,
+      "loss": 0.4571,
+      "step": 4245
+    },
+    {
+      "epoch": 0.7254416659554493,
+      "grad_norm": 0.47156030803819626,
+      "learning_rate": 4.212190187152251e-05,
+      "loss": 0.4713,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7262951267389264,
+      "grad_norm": 0.4591076554168388,
+      "learning_rate": 4.210609509357613e-05,
+      "loss": 0.4559,
+      "step": 4255
+    },
+    {
+      "epoch": 0.7271485875224033,
+      "grad_norm": 0.619852825322946,
+      "learning_rate": 4.2090288315629745e-05,
+      "loss": 0.5021,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7280020483058803,
+      "grad_norm": 0.6062909723217509,
+      "learning_rate": 4.2074481537683355e-05,
+      "loss": 0.4624,
+      "step": 4265
+    },
+    {
+      "epoch": 0.7288555090893574,
+      "grad_norm": 0.4146634798044198,
+      "learning_rate": 4.205867475973698e-05,
+      "loss": 0.4807,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7297089698728343,
+      "grad_norm": 0.648514590985978,
+      "learning_rate": 4.20428679817906e-05,
+      "loss": 0.4488,
+      "step": 4275
+    },
+    {
+      "epoch": 0.7305624306563113,
+      "grad_norm": 0.5969831455917731,
+      "learning_rate": 4.202706120384421e-05,
+      "loss": 0.4757,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7314158914397884,
+      "grad_norm": 0.5313730082368189,
+      "learning_rate": 4.2011254425897825e-05,
+      "loss": 0.4571,
+      "step": 4285
+    },
+    {
+      "epoch": 0.7322693522232654,
+      "grad_norm": 0.4251657799991048,
+      "learning_rate": 4.199544764795144e-05,
+      "loss": 0.4786,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7331228130067423,
+      "grad_norm": 0.4738411097488647,
+      "learning_rate": 4.197964087000506e-05,
+      "loss": 0.4721,
+      "step": 4295
+    },
+    {
+      "epoch": 0.7339762737902193,
+      "grad_norm": 0.6266077393685796,
+      "learning_rate": 4.196383409205868e-05,
+      "loss": 0.5001,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7348297345736964,
+      "grad_norm": 0.6021787676695373,
+      "learning_rate": 4.1948027314112295e-05,
+      "loss": 0.4645,
+      "step": 4305
+    },
+    {
+      "epoch": 0.7356831953571733,
+      "grad_norm": 0.7975390508151426,
+      "learning_rate": 4.193222053616591e-05,
+      "loss": 0.4871,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7365366561406503,
+      "grad_norm": 0.5021933795796126,
+      "learning_rate": 4.191641375821952e-05,
+      "loss": 0.4755,
+      "step": 4315
+    },
+    {
+      "epoch": 0.7373901169241274,
+      "grad_norm": 0.47386820243446054,
+      "learning_rate": 4.190060698027314e-05,
+      "loss": 0.4782,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7382435777076043,
+      "grad_norm": 0.47399583150250074,
+      "learning_rate": 4.188480020232676e-05,
+      "loss": 0.5062,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7390970384910813,
+      "grad_norm": 0.9263847986919433,
+      "learning_rate": 4.1868993424380375e-05,
+      "loss": 0.4713,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7399504992745584,
+      "grad_norm": 0.38060764152791543,
+      "learning_rate": 4.185318664643399e-05,
+      "loss": 0.4931,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7408039600580353,
+      "grad_norm": 0.4408892386193912,
+      "learning_rate": 4.183737986848761e-05,
+      "loss": 0.4812,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7416574208415123,
+      "grad_norm": 0.5146022417429106,
+      "learning_rate": 4.182157309054123e-05,
+      "loss": 0.4908,
+      "step": 4345
+    },
+    {
+      "epoch": 0.7425108816249893,
+      "grad_norm": 0.5124824948297765,
+      "learning_rate": 4.180576631259484e-05,
+      "loss": 0.4333,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7433643424084664,
+      "grad_norm": 0.43519678035197396,
+      "learning_rate": 4.178995953464846e-05,
+      "loss": 0.5134,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7442178031919433,
+      "grad_norm": 0.40836843733058104,
+      "learning_rate": 4.177415275670208e-05,
+      "loss": 0.4759,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7450712639754203,
+      "grad_norm": 0.6539216082100642,
+      "learning_rate": 4.175834597875569e-05,
+      "loss": 0.491,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7459247247588974,
+      "grad_norm": 0.741333557246768,
+      "learning_rate": 4.174253920080931e-05,
+      "loss": 0.4521,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7467781855423743,
+      "grad_norm": 0.4584230945611621,
+      "learning_rate": 4.1726732422862926e-05,
+      "loss": 0.4787,
+      "step": 4375
+    },
+    {
+      "epoch": 0.7476316463258513,
+      "grad_norm": 0.5047238639694727,
+      "learning_rate": 4.171092564491654e-05,
+      "loss": 0.4596,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7484851071093284,
+      "grad_norm": 0.4388777622815176,
+      "learning_rate": 4.1695118866970154e-05,
+      "loss": 0.4342,
+      "step": 4385
+    },
+    {
+      "epoch": 0.7493385678928053,
+      "grad_norm": 0.6473129680064909,
+      "learning_rate": 4.167931208902378e-05,
+      "loss": 0.4496,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7501920286762823,
+      "grad_norm": 0.4667194781780377,
+      "learning_rate": 4.1663505311077396e-05,
+      "loss": 0.4729,
+      "step": 4395
+    },
+    {
+      "epoch": 0.7510454894597594,
+      "grad_norm": 0.4798661445158895,
+      "learning_rate": 4.1647698533131006e-05,
+      "loss": 0.4781,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7518989502432363,
+      "grad_norm": 0.5258313097710264,
+      "learning_rate": 4.1631891755184624e-05,
+      "loss": 0.4574,
+      "step": 4405
+    },
+    {
+      "epoch": 0.7527524110267133,
+      "grad_norm": 0.4367681199785881,
+      "learning_rate": 4.161608497723824e-05,
+      "loss": 0.4629,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7536058718101903,
+      "grad_norm": 0.4058766887370437,
+      "learning_rate": 4.160027819929186e-05,
+      "loss": 0.4349,
+      "step": 4415
+    },
+    {
+      "epoch": 0.7544593325936674,
+      "grad_norm": 0.45389133719116836,
+      "learning_rate": 4.158447142134547e-05,
+      "loss": 0.4683,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7553127933771443,
+      "grad_norm": 0.9384019269281727,
+      "learning_rate": 4.1568664643399094e-05,
+      "loss": 0.4725,
+      "step": 4425
+    },
+    {
+      "epoch": 0.7561662541606213,
+      "grad_norm": 0.4171621248299094,
+      "learning_rate": 4.155285786545271e-05,
+      "loss": 0.4577,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7570197149440984,
+      "grad_norm": 0.4293985411718834,
+      "learning_rate": 4.153705108750632e-05,
+      "loss": 0.4583,
+      "step": 4435
+    },
+    {
+      "epoch": 0.7578731757275753,
+      "grad_norm": 0.5110467466987901,
+      "learning_rate": 4.152124430955994e-05,
+      "loss": 0.4726,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7587266365110523,
+      "grad_norm": 0.36866945596984196,
+      "learning_rate": 4.1505437531613557e-05,
+      "loss": 0.4706,
+      "step": 4445
+    },
+    {
+      "epoch": 0.7595800972945294,
+      "grad_norm": 0.3968297640314202,
+      "learning_rate": 4.1489630753667174e-05,
+      "loss": 0.4678,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7604335580780063,
+      "grad_norm": 0.38358786025829267,
+      "learning_rate": 4.147382397572079e-05,
+      "loss": 0.424,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7612870188614833,
+      "grad_norm": 0.44851307779408023,
+      "learning_rate": 4.145801719777441e-05,
+      "loss": 0.4786,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7621404796449603,
+      "grad_norm": 0.4372288670484839,
+      "learning_rate": 4.1442210419828026e-05,
+      "loss": 0.461,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7629939404284373,
+      "grad_norm": 0.3884950987593795,
+      "learning_rate": 4.142640364188164e-05,
+      "loss": 0.4601,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7638474012119143,
+      "grad_norm": 1.003434034563983,
+      "learning_rate": 4.141059686393526e-05,
+      "loss": 0.4751,
+      "step": 4475
+    },
+    {
+      "epoch": 0.7647008619953913,
+      "grad_norm": 0.41989132522212425,
+      "learning_rate": 4.139479008598887e-05,
+      "loss": 0.4405,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7655543227788683,
+      "grad_norm": 0.4421070338541073,
+      "learning_rate": 4.137898330804249e-05,
+      "loss": 0.4627,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7664077835623453,
+      "grad_norm": 0.4343925351326079,
+      "learning_rate": 4.136317653009611e-05,
+      "loss": 0.4718,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7672612443458223,
+      "grad_norm": 0.36233285655270203,
+      "learning_rate": 4.1347369752149724e-05,
+      "loss": 0.4532,
+      "step": 4495
+    },
+    {
+      "epoch": 0.7681147051292994,
+      "grad_norm": 0.42515097175129407,
+      "learning_rate": 4.133156297420334e-05,
+      "loss": 0.4685,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7689681659127763,
+      "grad_norm": 0.5464773560885858,
+      "learning_rate": 4.131575619625695e-05,
+      "loss": 0.4744,
+      "step": 4505
+    },
+    {
+      "epoch": 0.7698216266962533,
+      "grad_norm": 0.7900641435836122,
+      "learning_rate": 4.129994941831058e-05,
+      "loss": 0.4636,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7706750874797303,
+      "grad_norm": 0.38514288708300826,
+      "learning_rate": 4.1284142640364194e-05,
+      "loss": 0.4565,
+      "step": 4515
+    },
+    {
+      "epoch": 0.7715285482632073,
+      "grad_norm": 0.513740255362734,
+      "learning_rate": 4.1268335862417805e-05,
+      "loss": 0.4498,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7723820090466843,
+      "grad_norm": 0.4685507535202432,
+      "learning_rate": 4.125252908447142e-05,
+      "loss": 0.449,
+      "step": 4525
+    },
+    {
+      "epoch": 0.7732354698301613,
+      "grad_norm": 0.748890376912051,
+      "learning_rate": 4.123672230652504e-05,
+      "loss": 0.4908,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7740889306136383,
+      "grad_norm": 0.5297351959814006,
+      "learning_rate": 4.122091552857866e-05,
+      "loss": 0.4748,
+      "step": 4535
+    },
+    {
+      "epoch": 0.7749423913971153,
+      "grad_norm": 0.4004090062602965,
+      "learning_rate": 4.120510875063227e-05,
+      "loss": 0.456,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7757958521805923,
+      "grad_norm": 0.4603580473495154,
+      "learning_rate": 4.118930197268589e-05,
+      "loss": 0.4835,
+      "step": 4545
+    },
+    {
+      "epoch": 0.7766493129640692,
+      "grad_norm": 0.4091151289367216,
+      "learning_rate": 4.117349519473951e-05,
+      "loss": 0.4632,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7775027737475463,
+      "grad_norm": 0.4336955748413149,
+      "learning_rate": 4.115768841679312e-05,
+      "loss": 0.4466,
+      "step": 4555
+    },
+    {
+      "epoch": 0.7783562345310233,
+      "grad_norm": 0.4596120126536533,
+      "learning_rate": 4.114188163884674e-05,
+      "loss": 0.4658,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7792096953145004,
+      "grad_norm": 0.4537868261444298,
+      "learning_rate": 4.1126074860900355e-05,
+      "loss": 0.4484,
+      "step": 4565
+    },
+    {
+      "epoch": 0.7800631560979773,
+      "grad_norm": 0.49903144617561895,
+      "learning_rate": 4.111026808295397e-05,
+      "loss": 0.4902,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7809166168814543,
+      "grad_norm": 0.4300468454390534,
+      "learning_rate": 4.109446130500758e-05,
+      "loss": 0.4664,
+      "step": 4575
+    },
+    {
+      "epoch": 0.7817700776649313,
+      "grad_norm": 0.5268366623180529,
+      "learning_rate": 4.107865452706121e-05,
+      "loss": 0.4493,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7826235384484083,
+      "grad_norm": 0.46124267219055926,
+      "learning_rate": 4.1062847749114825e-05,
+      "loss": 0.4627,
+      "step": 4585
+    },
+    {
+      "epoch": 0.7834769992318853,
+      "grad_norm": 0.4226627432522082,
+      "learning_rate": 4.1047040971168436e-05,
+      "loss": 0.4368,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7843304600153623,
+      "grad_norm": 0.44261675551299395,
+      "learning_rate": 4.103123419322206e-05,
+      "loss": 0.463,
+      "step": 4595
+    },
+    {
+      "epoch": 0.7851839207988393,
+      "grad_norm": 0.8991940019280324,
+      "learning_rate": 4.101542741527567e-05,
+      "loss": 0.4683,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7860373815823163,
+      "grad_norm": 0.384751334196363,
+      "learning_rate": 4.099962063732929e-05,
+      "loss": 0.4768,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7868908423657933,
+      "grad_norm": 0.5951659423169229,
+      "learning_rate": 4.0983813859382906e-05,
+      "loss": 0.4961,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7877443031492702,
+      "grad_norm": 0.8089623040423884,
+      "learning_rate": 4.096800708143652e-05,
+      "loss": 0.4645,
+      "step": 4615
+    },
+    {
+      "epoch": 0.7885977639327473,
+      "grad_norm": 0.5242955484230468,
+      "learning_rate": 4.095220030349014e-05,
+      "loss": 0.4764,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7894512247162243,
+      "grad_norm": 0.45832328760285135,
+      "learning_rate": 4.093639352554375e-05,
+      "loss": 0.4693,
+      "step": 4625
+    },
+    {
+      "epoch": 0.7903046854997013,
+      "grad_norm": 0.4500365088370673,
+      "learning_rate": 4.0920586747597375e-05,
+      "loss": 0.4801,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7911581462831783,
+      "grad_norm": 0.5533871333970412,
+      "learning_rate": 4.0904779969650986e-05,
+      "loss": 0.4714,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7920116070666553,
+      "grad_norm": 0.6154994613950776,
+      "learning_rate": 4.0888973191704604e-05,
+      "loss": 0.4845,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7928650678501323,
+      "grad_norm": 0.46089491268483523,
+      "learning_rate": 4.087316641375822e-05,
+      "loss": 0.4745,
+      "step": 4645
+    },
+    {
+      "epoch": 0.7937185286336093,
+      "grad_norm": 0.5361431726193799,
+      "learning_rate": 4.085735963581184e-05,
+      "loss": 0.4538,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7945719894170863,
+      "grad_norm": 0.5220493537026821,
+      "learning_rate": 4.0841552857865456e-05,
+      "loss": 0.4396,
+      "step": 4655
+    },
+    {
+      "epoch": 0.7954254502005633,
+      "grad_norm": 0.6361482571581994,
+      "learning_rate": 4.0825746079919067e-05,
+      "loss": 0.4727,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7962789109840402,
+      "grad_norm": 0.40651694156272317,
+      "learning_rate": 4.080993930197269e-05,
+      "loss": 0.4816,
+      "step": 4665
+    },
+    {
+      "epoch": 0.7971323717675173,
+      "grad_norm": 0.6135578534897143,
+      "learning_rate": 4.07941325240263e-05,
+      "loss": 0.4748,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7979858325509943,
+      "grad_norm": 0.41497484588020045,
+      "learning_rate": 4.077832574607992e-05,
+      "loss": 0.4564,
+      "step": 4675
+    },
+    {
+      "epoch": 0.7988392933344712,
+      "grad_norm": 0.3853446441725627,
+      "learning_rate": 4.0762518968133536e-05,
+      "loss": 0.4515,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7996927541179483,
+      "grad_norm": 0.5824960605024038,
+      "learning_rate": 4.0746712190187154e-05,
+      "loss": 0.4524,
+      "step": 4685
+    },
+    {
+      "epoch": 0.8005462149014253,
+      "grad_norm": 0.49185525639089317,
+      "learning_rate": 4.073090541224077e-05,
+      "loss": 0.4626,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8013996756849022,
+      "grad_norm": 0.612221451639026,
+      "learning_rate": 4.071509863429438e-05,
+      "loss": 0.4769,
+      "step": 4695
+    },
+    {
+      "epoch": 0.8022531364683793,
+      "grad_norm": 0.3994074480084441,
+      "learning_rate": 4.0699291856348006e-05,
+      "loss": 0.4637,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8031065972518563,
+      "grad_norm": 0.7539597800025871,
+      "learning_rate": 4.0683485078401624e-05,
+      "loss": 0.4722,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8039600580353333,
+      "grad_norm": 0.5070926595384238,
+      "learning_rate": 4.0667678300455234e-05,
+      "loss": 0.4914,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8048135188188102,
+      "grad_norm": 0.5513153710741796,
+      "learning_rate": 4.065187152250886e-05,
+      "loss": 0.492,
+      "step": 4715
+    },
+    {
+      "epoch": 0.8056669796022873,
+      "grad_norm": 0.370937807426415,
+      "learning_rate": 4.063606474456247e-05,
+      "loss": 0.4894,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8065204403857643,
+      "grad_norm": 0.6682684562843318,
+      "learning_rate": 4.062025796661609e-05,
+      "loss": 0.4733,
+      "step": 4725
+    },
+    {
+      "epoch": 0.8073739011692412,
+      "grad_norm": 0.44302192953537,
+      "learning_rate": 4.0604451188669704e-05,
+      "loss": 0.4873,
+      "step": 4730
+    },
+    {
+      "epoch": 0.8082273619527183,
+      "grad_norm": 0.5333647000913362,
+      "learning_rate": 4.058864441072332e-05,
+      "loss": 0.4736,
+      "step": 4735
+    },
+    {
+      "epoch": 0.8090808227361953,
+      "grad_norm": 0.5505767154605103,
+      "learning_rate": 4.057283763277694e-05,
+      "loss": 0.4815,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8099342835196722,
+      "grad_norm": 0.4854973753670553,
+      "learning_rate": 4.055703085483055e-05,
+      "loss": 0.4616,
+      "step": 4745
+    },
+    {
+      "epoch": 0.8107877443031493,
+      "grad_norm": 0.4216977758902865,
+      "learning_rate": 4.0541224076884174e-05,
+      "loss": 0.4692,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8116412050866263,
+      "grad_norm": 0.6480224440676884,
+      "learning_rate": 4.0525417298937785e-05,
+      "loss": 0.4656,
+      "step": 4755
+    },
+    {
+      "epoch": 0.8124946658701032,
+      "grad_norm": 0.5196720864934075,
+      "learning_rate": 4.05096105209914e-05,
+      "loss": 0.4724,
+      "step": 4760
+    },
+    {
+      "epoch": 0.8133481266535803,
+      "grad_norm": 0.4142240721140529,
+      "learning_rate": 4.049380374304502e-05,
+      "loss": 0.4386,
+      "step": 4765
+    },
+    {
+      "epoch": 0.8142015874370573,
+      "grad_norm": 0.5044032502799161,
+      "learning_rate": 4.047799696509864e-05,
+      "loss": 0.4647,
+      "step": 4770
+    },
+    {
+      "epoch": 0.8150550482205343,
+      "grad_norm": 1.0326614960275178,
+      "learning_rate": 4.0462190187152255e-05,
+      "loss": 0.4406,
+      "step": 4775
+    },
+    {
+      "epoch": 0.8159085090040112,
+      "grad_norm": 0.502366249679762,
+      "learning_rate": 4.0446383409205865e-05,
+      "loss": 0.4639,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8167619697874883,
+      "grad_norm": 1.083322203978478,
+      "learning_rate": 4.043057663125949e-05,
+      "loss": 0.4567,
+      "step": 4785
+    },
+    {
+      "epoch": 0.8176154305709653,
+      "grad_norm": 0.4807750034947064,
+      "learning_rate": 4.04147698533131e-05,
+      "loss": 0.4534,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8184688913544422,
+      "grad_norm": 0.42630273833796006,
+      "learning_rate": 4.039896307536672e-05,
+      "loss": 0.4424,
+      "step": 4795
+    },
+    {
+      "epoch": 0.8193223521379193,
+      "grad_norm": 0.5471666962224171,
+      "learning_rate": 4.0383156297420335e-05,
+      "loss": 0.5013,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8201758129213963,
+      "grad_norm": 0.40088630057325625,
+      "learning_rate": 4.036734951947395e-05,
+      "loss": 0.4684,
+      "step": 4805
+    },
+    {
+      "epoch": 0.8210292737048732,
+      "grad_norm": 0.44204961810028137,
+      "learning_rate": 4.035154274152757e-05,
+      "loss": 0.4949,
+      "step": 4810
+    },
+    {
+      "epoch": 0.8218827344883503,
+      "grad_norm": 0.43055591123212844,
+      "learning_rate": 4.033573596358118e-05,
+      "loss": 0.4473,
+      "step": 4815
+    },
+    {
+      "epoch": 0.8227361952718273,
+      "grad_norm": 0.4270110673581626,
+      "learning_rate": 4.0319929185634805e-05,
+      "loss": 0.4974,
+      "step": 4820
+    },
+    {
+      "epoch": 0.8235896560553042,
+      "grad_norm": 0.508380546480638,
+      "learning_rate": 4.0304122407688416e-05,
+      "loss": 0.4386,
+      "step": 4825
+    },
+    {
+      "epoch": 0.8244431168387812,
+      "grad_norm": 0.8178606336757268,
+      "learning_rate": 4.028831562974203e-05,
+      "loss": 0.4791,
+      "step": 4830
+    },
+    {
+      "epoch": 0.8252965776222583,
+      "grad_norm": 10.544437701616161,
+      "learning_rate": 4.027250885179566e-05,
+      "loss": 0.4405,
+      "step": 4835
+    },
+    {
+      "epoch": 0.8261500384057353,
+      "grad_norm": 0.6872417396438704,
+      "learning_rate": 4.025670207384927e-05,
+      "loss": 0.4496,
+      "step": 4840
+    },
+    {
+      "epoch": 0.8270034991892122,
+      "grad_norm": 0.37016603686175553,
+      "learning_rate": 4.0240895295902885e-05,
+      "loss": 0.4566,
+      "step": 4845
+    },
+    {
+      "epoch": 0.8278569599726893,
+      "grad_norm": 0.43379088296174073,
+      "learning_rate": 4.02250885179565e-05,
+      "loss": 0.4756,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8287104207561663,
+      "grad_norm": 0.5484048511104972,
+      "learning_rate": 4.020928174001012e-05,
+      "loss": 0.4365,
+      "step": 4855
+    },
+    {
+      "epoch": 0.8295638815396432,
+      "grad_norm": 0.41841300895893785,
+      "learning_rate": 4.019347496206374e-05,
+      "loss": 0.4388,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8304173423231203,
+      "grad_norm": 0.4472193830259819,
+      "learning_rate": 4.017766818411735e-05,
+      "loss": 0.4765,
+      "step": 4865
+    },
+    {
+      "epoch": 0.8312708031065973,
+      "grad_norm": 0.35853153795034776,
+      "learning_rate": 4.016186140617097e-05,
+      "loss": 0.449,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8321242638900742,
+      "grad_norm": 0.4057918958397313,
+      "learning_rate": 4.014605462822458e-05,
+      "loss": 0.5011,
+      "step": 4875
+    },
+    {
+      "epoch": 0.8329777246735512,
+      "grad_norm": 0.3958130117885879,
+      "learning_rate": 4.01302478502782e-05,
+      "loss": 0.4667,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8338311854570283,
+      "grad_norm": 0.4738033473804357,
+      "learning_rate": 4.011444107233182e-05,
+      "loss": 0.4565,
+      "step": 4885
+    },
+    {
+      "epoch": 0.8346846462405052,
+      "grad_norm": 0.4025662142201199,
+      "learning_rate": 4.0098634294385436e-05,
+      "loss": 0.4978,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8355381070239822,
+      "grad_norm": 0.4434428406725396,
+      "learning_rate": 4.008282751643905e-05,
+      "loss": 0.4512,
+      "step": 4895
+    },
+    {
+      "epoch": 0.8363915678074593,
+      "grad_norm": 0.5150848531149659,
+      "learning_rate": 4.0067020738492664e-05,
+      "loss": 0.4427,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8372450285909362,
+      "grad_norm": 0.45758051602960453,
+      "learning_rate": 4.005121396054629e-05,
+      "loss": 0.4313,
+      "step": 4905
+    },
+    {
+      "epoch": 0.8380984893744132,
+      "grad_norm": 0.43431208239996316,
+      "learning_rate": 4.00354071825999e-05,
+      "loss": 0.4606,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8389519501578903,
+      "grad_norm": 0.36914787928203885,
+      "learning_rate": 4.0019600404653516e-05,
+      "loss": 0.4554,
+      "step": 4915
+    },
+    {
+      "epoch": 0.8398054109413673,
+      "grad_norm": 0.34338051155507815,
+      "learning_rate": 4.0003793626707134e-05,
+      "loss": 0.4671,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8406588717248442,
+      "grad_norm": 0.3423170869242225,
+      "learning_rate": 3.998798684876075e-05,
+      "loss": 0.4487,
+      "step": 4925
+    },
+    {
+      "epoch": 0.8415123325083212,
+      "grad_norm": 0.42559210239977796,
+      "learning_rate": 3.997218007081437e-05,
+      "loss": 0.4592,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8423657932917983,
+      "grad_norm": 0.4026293804737442,
+      "learning_rate": 3.995637329286798e-05,
+      "loss": 0.4699,
+      "step": 4935
+    },
+    {
+      "epoch": 0.8432192540752752,
+      "grad_norm": 0.5283079560242373,
+      "learning_rate": 3.9940566514921604e-05,
+      "loss": 0.4578,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8440727148587522,
+      "grad_norm": 0.576609671663474,
+      "learning_rate": 3.9924759736975214e-05,
+      "loss": 0.4704,
+      "step": 4945
+    },
+    {
+      "epoch": 0.8449261756422293,
+      "grad_norm": 0.3581273587093246,
+      "learning_rate": 3.990895295902883e-05,
+      "loss": 0.4376,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8457796364257062,
+      "grad_norm": 0.4994142001135852,
+      "learning_rate": 3.989314618108245e-05,
+      "loss": 0.4848,
+      "step": 4955
+    },
+    {
+      "epoch": 0.8466330972091832,
+      "grad_norm": 0.40318144070697504,
+      "learning_rate": 3.9877339403136067e-05,
+      "loss": 0.4286,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8474865579926603,
+      "grad_norm": 0.45045415873930383,
+      "learning_rate": 3.9861532625189684e-05,
+      "loss": 0.4813,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8483400187761372,
+      "grad_norm": 0.5210299348454007,
+      "learning_rate": 3.98457258472433e-05,
+      "loss": 0.4714,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8491934795596142,
+      "grad_norm": 0.39032612913864473,
+      "learning_rate": 3.982991906929692e-05,
+      "loss": 0.464,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8500469403430913,
+      "grad_norm": 0.4572815340178369,
+      "learning_rate": 3.981411229135053e-05,
+      "loss": 0.4681,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8509004011265683,
+      "grad_norm": 0.4248878833345479,
+      "learning_rate": 3.979830551340415e-05,
+      "loss": 0.4883,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8517538619100452,
+      "grad_norm": 0.4227418728003324,
+      "learning_rate": 3.978249873545777e-05,
+      "loss": 0.4746,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8526073226935222,
+      "grad_norm": 0.44564270844937715,
+      "learning_rate": 3.976669195751138e-05,
+      "loss": 0.4639,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8534607834769993,
+      "grad_norm": 0.43257274810461194,
+      "learning_rate": 3.9750885179565e-05,
+      "loss": 0.4761,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8543142442604762,
+      "grad_norm": 0.6510055910669625,
+      "learning_rate": 3.973507840161862e-05,
+      "loss": 0.483,
+      "step": 5005
+    },
+    {
+      "epoch": 0.8551677050439532,
+      "grad_norm": 0.4174446601702513,
+      "learning_rate": 3.9719271623672234e-05,
+      "loss": 0.4585,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8560211658274303,
+      "grad_norm": 0.46208025471964614,
+      "learning_rate": 3.970346484572585e-05,
+      "loss": 0.4531,
+      "step": 5015
+    },
+    {
+      "epoch": 0.8568746266109072,
+      "grad_norm": 0.38656337115836403,
+      "learning_rate": 3.968765806777946e-05,
+      "loss": 0.4773,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8577280873943842,
+      "grad_norm": 0.3995048395641776,
+      "learning_rate": 3.967185128983309e-05,
+      "loss": 0.492,
+      "step": 5025
+    },
+    {
+      "epoch": 0.8585815481778613,
+      "grad_norm": 0.646673653905275,
+      "learning_rate": 3.96560445118867e-05,
+      "loss": 0.4532,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8594350089613382,
+      "grad_norm": 0.4480550691124822,
+      "learning_rate": 3.9640237733940315e-05,
+      "loss": 0.4809,
+      "step": 5035
+    },
+    {
+      "epoch": 0.8602884697448152,
+      "grad_norm": 0.5130279783696443,
+      "learning_rate": 3.962443095599393e-05,
+      "loss": 0.4597,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8611419305282922,
+      "grad_norm": 0.5187207242807629,
+      "learning_rate": 3.960862417804755e-05,
+      "loss": 0.4516,
+      "step": 5045
+    },
+    {
+      "epoch": 0.8619953913117693,
+      "grad_norm": 0.3890802932011367,
+      "learning_rate": 3.959281740010117e-05,
+      "loss": 0.4479,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8628488520952462,
+      "grad_norm": 0.444180638493775,
+      "learning_rate": 3.957701062215478e-05,
+      "loss": 0.4724,
+      "step": 5055
+    },
+    {
+      "epoch": 0.8637023128787232,
+      "grad_norm": 0.44490943443342873,
+      "learning_rate": 3.95612038442084e-05,
+      "loss": 0.4796,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8645557736622003,
+      "grad_norm": 0.39454985098988465,
+      "learning_rate": 3.954539706626201e-05,
+      "loss": 0.4767,
+      "step": 5065
+    },
+    {
+      "epoch": 0.8654092344456772,
+      "grad_norm": 0.43075230815685706,
+      "learning_rate": 3.952959028831563e-05,
+      "loss": 0.4943,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8662626952291542,
+      "grad_norm": 0.579774474859966,
+      "learning_rate": 3.951378351036925e-05,
+      "loss": 0.4781,
+      "step": 5075
+    },
+    {
+      "epoch": 0.8671161560126313,
+      "grad_norm": 0.6050594246243188,
+      "learning_rate": 3.9497976732422865e-05,
+      "loss": 0.462,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8679696167961082,
+      "grad_norm": 0.4906701824581924,
+      "learning_rate": 3.948216995447648e-05,
+      "loss": 0.4717,
+      "step": 5085
+    },
+    {
+      "epoch": 0.8688230775795852,
+      "grad_norm": 0.3628131269499048,
+      "learning_rate": 3.9466363176530093e-05,
+      "loss": 0.4469,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8696765383630622,
+      "grad_norm": 0.3875750676153778,
+      "learning_rate": 3.945055639858372e-05,
+      "loss": 0.447,
+      "step": 5095
+    },
+    {
+      "epoch": 0.8705299991465392,
+      "grad_norm": 0.3785699572390181,
+      "learning_rate": 3.943474962063733e-05,
+      "loss": 0.4666,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8713834599300162,
+      "grad_norm": 0.41413344640524147,
+      "learning_rate": 3.9418942842690946e-05,
+      "loss": 0.4764,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8722369207134932,
+      "grad_norm": 0.37990935238738166,
+      "learning_rate": 3.940313606474457e-05,
+      "loss": 0.4341,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8730903814969703,
+      "grad_norm": 0.325922151892048,
+      "learning_rate": 3.938732928679818e-05,
+      "loss": 0.4464,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8739438422804472,
+      "grad_norm": 0.3625407227950122,
+      "learning_rate": 3.93715225088518e-05,
+      "loss": 0.4572,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8747973030639242,
+      "grad_norm": 0.4305070799506871,
+      "learning_rate": 3.9355715730905416e-05,
+      "loss": 0.4209,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8756507638474013,
+      "grad_norm": 0.34770288354175327,
+      "learning_rate": 3.933990895295903e-05,
+      "loss": 0.4957,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8765042246308782,
+      "grad_norm": 0.370649858584003,
+      "learning_rate": 3.9324102175012644e-05,
+      "loss": 0.4369,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8773576854143552,
+      "grad_norm": 0.6631930203969955,
+      "learning_rate": 3.930829539706626e-05,
+      "loss": 0.4637,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8782111461978322,
+      "grad_norm": 0.3821885520825206,
+      "learning_rate": 3.9292488619119885e-05,
+      "loss": 0.4468,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8790646069813092,
+      "grad_norm": 0.42015351568091,
+      "learning_rate": 3.9276681841173496e-05,
+      "loss": 0.4551,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8799180677647862,
+      "grad_norm": 0.39848695247719024,
+      "learning_rate": 3.9260875063227114e-05,
+      "loss": 0.4633,
+      "step": 5155
+    },
+    {
+      "epoch": 0.8807715285482632,
+      "grad_norm": 0.39805656884228074,
+      "learning_rate": 3.924506828528073e-05,
+      "loss": 0.453,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8816249893317402,
+      "grad_norm": 0.72730511596657,
+      "learning_rate": 3.922926150733435e-05,
+      "loss": 0.482,
+      "step": 5165
+    },
+    {
+      "epoch": 0.8824784501152172,
+      "grad_norm": 0.3835150212424347,
+      "learning_rate": 3.9213454729387966e-05,
+      "loss": 0.4268,
+      "step": 5170
+    },
+    {
+      "epoch": 0.8833319108986942,
+      "grad_norm": 0.5549806218770832,
+      "learning_rate": 3.919764795144158e-05,
+      "loss": 0.4976,
+      "step": 5175
+    },
+    {
+      "epoch": 0.8841853716821712,
+      "grad_norm": 0.45886131028562016,
+      "learning_rate": 3.91818411734952e-05,
+      "loss": 0.4771,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8850388324656482,
+      "grad_norm": 0.38461247343867677,
+      "learning_rate": 3.916603439554881e-05,
+      "loss": 0.4479,
+      "step": 5185
+    },
+    {
+      "epoch": 0.8858922932491252,
+      "grad_norm": 0.39432030577402233,
+      "learning_rate": 3.915022761760243e-05,
+      "loss": 0.4389,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8867457540326023,
+      "grad_norm": 0.4445775875626058,
+      "learning_rate": 3.9134420839656046e-05,
+      "loss": 0.4424,
+      "step": 5195
+    },
+    {
+      "epoch": 0.8875992148160792,
+      "grad_norm": 0.47542478261605625,
+      "learning_rate": 3.9118614061709664e-05,
+      "loss": 0.4603,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8884526755995562,
+      "grad_norm": 0.3887252750059788,
+      "learning_rate": 3.910280728376328e-05,
+      "loss": 0.446,
+      "step": 5205
+    },
+    {
+      "epoch": 0.8893061363830332,
+      "grad_norm": 0.5388718629032798,
+      "learning_rate": 3.908700050581689e-05,
+      "loss": 0.4911,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8901595971665102,
+      "grad_norm": 0.415472898172573,
+      "learning_rate": 3.9071193727870516e-05,
+      "loss": 0.4503,
+      "step": 5215
+    },
+    {
+      "epoch": 0.8910130579499872,
+      "grad_norm": 0.41063762995113084,
+      "learning_rate": 3.905538694992413e-05,
+      "loss": 0.4546,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8918665187334642,
+      "grad_norm": 0.44084348091256453,
+      "learning_rate": 3.9039580171977744e-05,
+      "loss": 0.4809,
+      "step": 5225
+    },
+    {
+      "epoch": 0.8927199795169412,
+      "grad_norm": 0.45463091201990335,
+      "learning_rate": 3.902377339403136e-05,
+      "loss": 0.4559,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8935734403004182,
+      "grad_norm": 0.38493395938624353,
+      "learning_rate": 3.900796661608498e-05,
+      "loss": 0.4572,
+      "step": 5235
+    },
+    {
+      "epoch": 0.8944269010838952,
+      "grad_norm": 0.3792702027768502,
+      "learning_rate": 3.89921598381386e-05,
+      "loss": 0.4438,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8952803618673721,
+      "grad_norm": 0.35552476732779537,
+      "learning_rate": 3.8976353060192214e-05,
+      "loss": 0.4857,
+      "step": 5245
+    },
+    {
+      "epoch": 0.8961338226508492,
+      "grad_norm": 0.4411242106417821,
+      "learning_rate": 3.896054628224583e-05,
+      "loss": 0.4431,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8969872834343262,
+      "grad_norm": 0.3501135290159158,
+      "learning_rate": 3.894473950429944e-05,
+      "loss": 0.475,
+      "step": 5255
+    },
+    {
+      "epoch": 0.8978407442178032,
+      "grad_norm": 0.46967490673458273,
+      "learning_rate": 3.892893272635306e-05,
+      "loss": 0.4732,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8986942050012802,
+      "grad_norm": 0.41543836571528997,
+      "learning_rate": 3.8913125948406684e-05,
+      "loss": 0.4693,
+      "step": 5265
+    },
+    {
+      "epoch": 0.8995476657847572,
+      "grad_norm": 0.3803928447120958,
+      "learning_rate": 3.8897319170460295e-05,
+      "loss": 0.4574,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9004011265682342,
+      "grad_norm": 0.4043183675600425,
+      "learning_rate": 3.888151239251391e-05,
+      "loss": 0.4446,
+      "step": 5275
+    },
+    {
+      "epoch": 0.9012545873517112,
+      "grad_norm": 0.35773284894988394,
+      "learning_rate": 3.886570561456753e-05,
+      "loss": 0.4506,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9021080481351882,
+      "grad_norm": 0.39756526693714445,
+      "learning_rate": 3.884989883662115e-05,
+      "loss": 0.4637,
+      "step": 5285
+    },
+    {
+      "epoch": 0.9029615089186652,
+      "grad_norm": 0.3968235423417824,
+      "learning_rate": 3.883409205867476e-05,
+      "loss": 0.4524,
+      "step": 5290
+    },
+    {
+      "epoch": 0.9038149697021421,
+      "grad_norm": 0.3976823096856294,
+      "learning_rate": 3.8818285280728375e-05,
+      "loss": 0.4617,
+      "step": 5295
+    },
+    {
+      "epoch": 0.9046684304856192,
+      "grad_norm": 0.37734719862477745,
+      "learning_rate": 3.8802478502782e-05,
+      "loss": 0.4396,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9055218912690962,
+      "grad_norm": 0.3713872249714145,
+      "learning_rate": 3.878667172483561e-05,
+      "loss": 0.4701,
+      "step": 5305
+    },
+    {
+      "epoch": 0.9063753520525731,
+      "grad_norm": 0.356091804946768,
+      "learning_rate": 3.877086494688923e-05,
+      "loss": 0.4368,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9072288128360502,
+      "grad_norm": 0.394160250063924,
+      "learning_rate": 3.8755058168942845e-05,
+      "loss": 0.4555,
+      "step": 5315
+    },
+    {
+      "epoch": 0.9080822736195272,
+      "grad_norm": 0.43170759667100317,
+      "learning_rate": 3.873925139099646e-05,
+      "loss": 0.4284,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9089357344030042,
+      "grad_norm": 0.4179349029183989,
+      "learning_rate": 3.872344461305007e-05,
+      "loss": 0.4523,
+      "step": 5325
+    },
+    {
+      "epoch": 0.9097891951864812,
+      "grad_norm": 0.3664464693848915,
+      "learning_rate": 3.870763783510369e-05,
+      "loss": 0.4502,
+      "step": 5330
+    },
+    {
+      "epoch": 0.9106426559699582,
+      "grad_norm": 0.38687957544969326,
+      "learning_rate": 3.8691831057157315e-05,
+      "loss": 0.4597,
+      "step": 5335
+    },
+    {
+      "epoch": 0.9114961167534352,
+      "grad_norm": 0.44979205797145444,
+      "learning_rate": 3.8676024279210926e-05,
+      "loss": 0.4613,
+      "step": 5340
+    },
+    {
+      "epoch": 0.9123495775369121,
+      "grad_norm": 0.38598553360485094,
+      "learning_rate": 3.866021750126454e-05,
+      "loss": 0.4427,
+      "step": 5345
+    },
+    {
+      "epoch": 0.9132030383203892,
+      "grad_norm": 1.6236580699720844,
+      "learning_rate": 3.864441072331816e-05,
+      "loss": 0.4739,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9140564991038662,
+      "grad_norm": 0.4246158964041592,
+      "learning_rate": 3.862860394537178e-05,
+      "loss": 0.4876,
+      "step": 5355
+    },
+    {
+      "epoch": 0.9149099598873431,
+      "grad_norm": 0.41328536026606627,
+      "learning_rate": 3.8612797167425395e-05,
+      "loss": 0.4887,
+      "step": 5360
+    },
+    {
+      "epoch": 0.9157634206708202,
+      "grad_norm": 0.7643525341386238,
+      "learning_rate": 3.859699038947901e-05,
+      "loss": 0.4387,
+      "step": 5365
+    },
+    {
+      "epoch": 0.9166168814542972,
+      "grad_norm": 0.39333252384942685,
+      "learning_rate": 3.858118361153263e-05,
+      "loss": 0.4603,
+      "step": 5370
+    },
+    {
+      "epoch": 0.9174703422377741,
+      "grad_norm": 0.36549208355741303,
+      "learning_rate": 3.856537683358624e-05,
+      "loss": 0.4313,
+      "step": 5375
+    },
+    {
+      "epoch": 0.9183238030212512,
+      "grad_norm": 0.38888946522034146,
+      "learning_rate": 3.854957005563986e-05,
+      "loss": 0.4526,
+      "step": 5380
+    },
+    {
+      "epoch": 0.9191772638047282,
+      "grad_norm": 0.39411286500370646,
+      "learning_rate": 3.8533763277693476e-05,
+      "loss": 0.4325,
+      "step": 5385
+    },
+    {
+      "epoch": 0.9200307245882051,
+      "grad_norm": 0.3927495233983897,
+      "learning_rate": 3.8517956499747093e-05,
+      "loss": 0.4575,
+      "step": 5390
+    },
+    {
+      "epoch": 0.9208841853716822,
+      "grad_norm": 0.42377440168134617,
+      "learning_rate": 3.850214972180071e-05,
+      "loss": 0.4673,
+      "step": 5395
+    },
+    {
+      "epoch": 0.9217376461551592,
+      "grad_norm": 0.4432199119619774,
+      "learning_rate": 3.848634294385433e-05,
+      "loss": 0.4875,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9225911069386362,
+      "grad_norm": 0.40585755422175995,
+      "learning_rate": 3.8470536165907946e-05,
+      "loss": 0.4389,
+      "step": 5405
+    },
+    {
+      "epoch": 0.9234445677221131,
+      "grad_norm": 0.4538103442430755,
+      "learning_rate": 3.8454729387961556e-05,
+      "loss": 0.4859,
+      "step": 5410
+    },
+    {
+      "epoch": 0.9242980285055902,
+      "grad_norm": 0.3621378142955139,
+      "learning_rate": 3.8438922610015174e-05,
+      "loss": 0.4815,
+      "step": 5415
+    },
+    {
+      "epoch": 0.9251514892890672,
+      "grad_norm": 0.3697916050267976,
+      "learning_rate": 3.84231158320688e-05,
+      "loss": 0.4686,
+      "step": 5420
+    },
+    {
+      "epoch": 0.9260049500725441,
+      "grad_norm": 0.3792912721600657,
+      "learning_rate": 3.840730905412241e-05,
+      "loss": 0.4481,
+      "step": 5425
+    },
+    {
+      "epoch": 0.9268584108560212,
+      "grad_norm": 0.34922627192313405,
+      "learning_rate": 3.8391502276176026e-05,
+      "loss": 0.446,
+      "step": 5430
+    },
+    {
+      "epoch": 0.9277118716394982,
+      "grad_norm": 0.3940130604158538,
+      "learning_rate": 3.8375695498229644e-05,
+      "loss": 0.4578,
+      "step": 5435
+    },
+    {
+      "epoch": 0.9285653324229751,
+      "grad_norm": 0.4821375933416552,
+      "learning_rate": 3.835988872028326e-05,
+      "loss": 0.4579,
+      "step": 5440
+    },
+    {
+      "epoch": 0.9294187932064522,
+      "grad_norm": 0.3960119821025436,
+      "learning_rate": 3.834408194233687e-05,
+      "loss": 0.4309,
+      "step": 5445
+    },
+    {
+      "epoch": 0.9302722539899292,
+      "grad_norm": 0.6492593695986181,
+      "learning_rate": 3.832827516439049e-05,
+      "loss": 0.443,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9311257147734061,
+      "grad_norm": 0.3866711989379687,
+      "learning_rate": 3.8312468386444114e-05,
+      "loss": 0.4596,
+      "step": 5455
+    },
+    {
+      "epoch": 0.9319791755568831,
+      "grad_norm": 0.4548216235756429,
+      "learning_rate": 3.8296661608497724e-05,
+      "loss": 0.4493,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9328326363403602,
+      "grad_norm": 0.3559021780563998,
+      "learning_rate": 3.828085483055134e-05,
+      "loss": 0.4338,
+      "step": 5465
+    },
+    {
+      "epoch": 0.9336860971238372,
+      "grad_norm": 0.34946564902279204,
+      "learning_rate": 3.826504805260496e-05,
+      "loss": 0.4472,
+      "step": 5470
+    },
+    {
+      "epoch": 0.9345395579073141,
+      "grad_norm": 0.37578990267881635,
+      "learning_rate": 3.824924127465858e-05,
+      "loss": 0.4496,
+      "step": 5475
+    },
+    {
+      "epoch": 0.9353930186907912,
+      "grad_norm": 0.4716502415591886,
+      "learning_rate": 3.823343449671219e-05,
+      "loss": 0.5159,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9362464794742682,
+      "grad_norm": 1.0266519725140473,
+      "learning_rate": 3.821762771876581e-05,
+      "loss": 0.4503,
+      "step": 5485
+    },
+    {
+      "epoch": 0.9370999402577451,
+      "grad_norm": 0.44412547391687285,
+      "learning_rate": 3.820182094081943e-05,
+      "loss": 0.4524,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9379534010412222,
+      "grad_norm": 0.503593542190173,
+      "learning_rate": 3.818601416287304e-05,
+      "loss": 0.495,
+      "step": 5495
+    },
+    {
+      "epoch": 0.9388068618246992,
+      "grad_norm": 0.390089010381301,
+      "learning_rate": 3.817020738492666e-05,
+      "loss": 0.4438,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9396603226081761,
+      "grad_norm": 0.5237476550840837,
+      "learning_rate": 3.8154400606980275e-05,
+      "loss": 0.4737,
+      "step": 5505
+    },
+    {
+      "epoch": 0.9405137833916531,
+      "grad_norm": 0.5745085161307547,
+      "learning_rate": 3.813859382903389e-05,
+      "loss": 0.4677,
+      "step": 5510
+    },
+    {
+      "epoch": 0.9413672441751302,
+      "grad_norm": 0.42441116822101493,
+      "learning_rate": 3.812278705108751e-05,
+      "loss": 0.433,
+      "step": 5515
+    },
+    {
+      "epoch": 0.9422207049586071,
+      "grad_norm": 0.4141184188803559,
+      "learning_rate": 3.810698027314113e-05,
+      "loss": 0.4731,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9430741657420841,
+      "grad_norm": 0.3863082733523815,
+      "learning_rate": 3.8091173495194744e-05,
+      "loss": 0.4562,
+      "step": 5525
+    },
+    {
+      "epoch": 0.9439276265255612,
+      "grad_norm": 0.4710681089385406,
+      "learning_rate": 3.8075366717248355e-05,
+      "loss": 0.4643,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9447810873090382,
+      "grad_norm": 0.4433783546619392,
+      "learning_rate": 3.805955993930197e-05,
+      "loss": 0.4793,
+      "step": 5535
+    },
+    {
+      "epoch": 0.9456345480925151,
+      "grad_norm": 0.40443344260421626,
+      "learning_rate": 3.804375316135559e-05,
+      "loss": 0.4623,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9464880088759922,
+      "grad_norm": 0.47758717691243485,
+      "learning_rate": 3.802794638340921e-05,
+      "loss": 0.4584,
+      "step": 5545
+    },
+    {
+      "epoch": 0.9473414696594692,
+      "grad_norm": 0.5070813150960025,
+      "learning_rate": 3.8012139605462825e-05,
+      "loss": 0.4451,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9481949304429461,
+      "grad_norm": 0.4787056345105152,
+      "learning_rate": 3.799633282751644e-05,
+      "loss": 0.4798,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9490483912264231,
+      "grad_norm": 0.41180507381024867,
+      "learning_rate": 3.798052604957006e-05,
+      "loss": 0.4603,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9499018520099002,
+      "grad_norm": 0.4275967348346644,
+      "learning_rate": 3.796471927162367e-05,
+      "loss": 0.4589,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9507553127933771,
+      "grad_norm": 0.409501042986901,
+      "learning_rate": 3.794891249367729e-05,
+      "loss": 0.4745,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9516087735768541,
+      "grad_norm": 0.4671380078413468,
+      "learning_rate": 3.793310571573091e-05,
+      "loss": 0.4755,
+      "step": 5575
+    },
+    {
+      "epoch": 0.9524622343603312,
+      "grad_norm": 0.4160105726538355,
+      "learning_rate": 3.791729893778452e-05,
+      "loss": 0.4587,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9533156951438081,
+      "grad_norm": 0.4816419888813679,
+      "learning_rate": 3.790149215983814e-05,
+      "loss": 0.4593,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9541691559272851,
+      "grad_norm": 0.41394130403273693,
+      "learning_rate": 3.788568538189176e-05,
+      "loss": 0.4666,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9550226167107622,
+      "grad_norm": 0.42424423245469944,
+      "learning_rate": 3.7869878603945375e-05,
+      "loss": 0.4545,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9558760774942391,
+      "grad_norm": 0.4639841340731286,
+      "learning_rate": 3.7854071825998986e-05,
+      "loss": 0.4488,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9567295382777161,
+      "grad_norm": 0.5060482156153165,
+      "learning_rate": 3.783826504805261e-05,
+      "loss": 0.4315,
+      "step": 5605
+    },
+    {
+      "epoch": 0.9575829990611932,
+      "grad_norm": 0.4145197030432268,
+      "learning_rate": 3.782245827010623e-05,
+      "loss": 0.4472,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9584364598446702,
+      "grad_norm": 0.42736718516659117,
+      "learning_rate": 3.780665149215984e-05,
+      "loss": 0.463,
+      "step": 5615
+    },
+    {
+      "epoch": 0.9592899206281471,
+      "grad_norm": 0.5099800714999574,
+      "learning_rate": 3.7790844714213456e-05,
+      "loss": 0.4494,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9601433814116241,
+      "grad_norm": 0.3932085090574082,
+      "learning_rate": 3.777503793626707e-05,
+      "loss": 0.4472,
+      "step": 5625
+    },
+    {
+      "epoch": 0.9609968421951012,
+      "grad_norm": 0.4041638575629075,
+      "learning_rate": 3.775923115832069e-05,
+      "loss": 0.4716,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9618503029785781,
+      "grad_norm": 0.35654399586608365,
+      "learning_rate": 3.77434243803743e-05,
+      "loss": 0.4256,
+      "step": 5635
+    },
+    {
+      "epoch": 0.9627037637620551,
+      "grad_norm": 0.40490143000313356,
+      "learning_rate": 3.7727617602427926e-05,
+      "loss": 0.4427,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9635572245455322,
+      "grad_norm": 0.4306969286472383,
+      "learning_rate": 3.771181082448154e-05,
+      "loss": 0.4599,
+      "step": 5645
+    },
+    {
+      "epoch": 0.9644106853290091,
+      "grad_norm": 0.44608416080202795,
+      "learning_rate": 3.7696004046535154e-05,
+      "loss": 0.4504,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9652641461124861,
+      "grad_norm": 0.4909756812121427,
+      "learning_rate": 3.768019726858877e-05,
+      "loss": 0.4454,
+      "step": 5655
+    },
+    {
+      "epoch": 0.9661176068959632,
+      "grad_norm": 0.42378492668995615,
+      "learning_rate": 3.766439049064239e-05,
+      "loss": 0.4792,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9669710676794401,
+      "grad_norm": 0.39735362969367277,
+      "learning_rate": 3.7648583712696006e-05,
+      "loss": 0.4881,
+      "step": 5665
+    },
+    {
+      "epoch": 0.9678245284629171,
+      "grad_norm": 0.3573636936700055,
+      "learning_rate": 3.7632776934749624e-05,
+      "loss": 0.431,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9686779892463941,
+      "grad_norm": 0.856707855907997,
+      "learning_rate": 3.761697015680324e-05,
+      "loss": 0.4403,
+      "step": 5675
+    },
+    {
+      "epoch": 0.9695314500298712,
+      "grad_norm": 0.3907319920812853,
+      "learning_rate": 3.760116337885686e-05,
+      "loss": 0.4591,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9703849108133481,
+      "grad_norm": 0.3617016366595509,
+      "learning_rate": 3.758535660091047e-05,
+      "loss": 0.4381,
+      "step": 5685
+    },
+    {
+      "epoch": 0.9712383715968251,
+      "grad_norm": 0.4184546987813738,
+      "learning_rate": 3.756954982296409e-05,
+      "loss": 0.5012,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9720918323803022,
+      "grad_norm": 4.862897933769316,
+      "learning_rate": 3.7553743045017704e-05,
+      "loss": 0.4867,
+      "step": 5695
+    },
+    {
+      "epoch": 0.9729452931637791,
+      "grad_norm": 0.5282636819263453,
+      "learning_rate": 3.753793626707132e-05,
+      "loss": 0.4659,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9737987539472561,
+      "grad_norm": 0.43193117699851213,
+      "learning_rate": 3.752212948912494e-05,
+      "loss": 0.4693,
+      "step": 5705
+    },
+    {
+      "epoch": 0.9746522147307332,
+      "grad_norm": 0.42975743283527323,
+      "learning_rate": 3.7506322711178557e-05,
+      "loss": 0.4738,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9755056755142101,
+      "grad_norm": 0.4343654508833946,
+      "learning_rate": 3.7490515933232174e-05,
+      "loss": 0.4388,
+      "step": 5715
+    },
+    {
+      "epoch": 0.9763591362976871,
+      "grad_norm": 0.365610873557326,
+      "learning_rate": 3.7474709155285785e-05,
+      "loss": 0.4799,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9772125970811641,
+      "grad_norm": 0.4683124404513186,
+      "learning_rate": 3.74589023773394e-05,
+      "loss": 0.4498,
+      "step": 5725
+    },
+    {
+      "epoch": 0.9780660578646411,
+      "grad_norm": 0.7878480520304444,
+      "learning_rate": 3.744309559939302e-05,
+      "loss": 0.4575,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9789195186481181,
+      "grad_norm": 0.4801039090352379,
+      "learning_rate": 3.742728882144664e-05,
+      "loss": 0.484,
+      "step": 5735
+    },
+    {
+      "epoch": 0.9797729794315951,
+      "grad_norm": 0.5152189238178845,
+      "learning_rate": 3.7411482043500254e-05,
+      "loss": 0.4705,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9806264402150722,
+      "grad_norm": 0.4867310274108711,
+      "learning_rate": 3.739567526555387e-05,
+      "loss": 0.4771,
+      "step": 5745
+    },
+    {
+      "epoch": 0.9814799009985491,
+      "grad_norm": 0.5645780923606522,
+      "learning_rate": 3.737986848760749e-05,
+      "loss": 0.4746,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9823333617820261,
+      "grad_norm": 0.38450727974879706,
+      "learning_rate": 3.73640617096611e-05,
+      "loss": 0.4484,
+      "step": 5755
+    },
+    {
+      "epoch": 0.9831868225655032,
+      "grad_norm": 0.41981751648706767,
+      "learning_rate": 3.7348254931714724e-05,
+      "loss": 0.4366,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9840402833489801,
+      "grad_norm": 0.5124579749342585,
+      "learning_rate": 3.733244815376834e-05,
+      "loss": 0.4655,
+      "step": 5765
+    },
+    {
+      "epoch": 0.9848937441324571,
+      "grad_norm": 0.43040573748634436,
+      "learning_rate": 3.731664137582195e-05,
+      "loss": 0.4458,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9857472049159341,
+      "grad_norm": 0.371097389052583,
+      "learning_rate": 3.730083459787557e-05,
+      "loss": 0.4646,
+      "step": 5775
+    },
+    {
+      "epoch": 0.9866006656994111,
+      "grad_norm": 0.8835102192758597,
+      "learning_rate": 3.728502781992919e-05,
+      "loss": 0.4987,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9874541264828881,
+      "grad_norm": 0.34985676739187793,
+      "learning_rate": 3.7269221041982805e-05,
+      "loss": 0.4307,
+      "step": 5785
+    },
+    {
+      "epoch": 0.9883075872663651,
+      "grad_norm": 0.4482204094674099,
+      "learning_rate": 3.7253414264036416e-05,
+      "loss": 0.462,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9891610480498421,
+      "grad_norm": 0.4641563365689893,
+      "learning_rate": 3.723760748609004e-05,
+      "loss": 0.4458,
+      "step": 5795
+    },
+    {
+      "epoch": 0.9900145088333191,
+      "grad_norm": 0.6375724302920852,
+      "learning_rate": 3.722180070814366e-05,
+      "loss": 0.4595,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9908679696167961,
+      "grad_norm": 0.5559232925637143,
+      "learning_rate": 3.720599393019727e-05,
+      "loss": 0.4439,
+      "step": 5805
+    },
+    {
+      "epoch": 0.9917214304002732,
+      "grad_norm": 0.37657819341882687,
+      "learning_rate": 3.7190187152250885e-05,
+      "loss": 0.4632,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9925748911837501,
+      "grad_norm": 0.3901766835132743,
+      "learning_rate": 3.71743803743045e-05,
+      "loss": 0.4596,
+      "step": 5815
+    },
+    {
+      "epoch": 0.9934283519672271,
+      "grad_norm": 0.4313043353543776,
+      "learning_rate": 3.715857359635812e-05,
+      "loss": 0.4735,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9942818127507042,
+      "grad_norm": 0.37318724262666314,
+      "learning_rate": 3.714276681841174e-05,
+      "loss": 0.4684,
+      "step": 5825
+    },
+    {
+      "epoch": 0.9951352735341811,
+      "grad_norm": 0.4813693313647458,
+      "learning_rate": 3.7126960040465355e-05,
+      "loss": 0.4623,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9959887343176581,
+      "grad_norm": 0.4486878697427319,
+      "learning_rate": 3.711115326251897e-05,
+      "loss": 0.4533,
+      "step": 5835
+    },
+    {
+      "epoch": 0.9968421951011351,
+      "grad_norm": 0.42413980193320633,
+      "learning_rate": 3.709534648457258e-05,
+      "loss": 0.4569,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9976956558846121,
+      "grad_norm": 0.4333397282299119,
+      "learning_rate": 3.70795397066262e-05,
+      "loss": 0.4722,
+      "step": 5845
+    },
+    {
+      "epoch": 0.9985491166680891,
+      "grad_norm": 0.3499741056409646,
+      "learning_rate": 3.706373292867982e-05,
+      "loss": 0.422,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9994025774515661,
+      "grad_norm": 0.4027209707705357,
+      "learning_rate": 3.7047926150733436e-05,
+      "loss": 0.4636,
+      "step": 5855
+    },
+    {
+      "epoch": 1.0001706921566953,
+      "grad_norm": 0.6288062417613335,
+      "learning_rate": 3.703211937278705e-05,
+      "loss": 0.4473,
+      "step": 5860
+    },
+    {
+      "epoch": 1.0010241529401724,
+      "grad_norm": 0.38754061072346385,
+      "learning_rate": 3.701631259484067e-05,
+      "loss": 0.3766,
+      "step": 5865
+    },
+    {
+      "epoch": 1.0018776137236494,
+      "grad_norm": 0.33960600570302646,
+      "learning_rate": 3.700050581689429e-05,
+      "loss": 0.3663,
+      "step": 5870
+    },
+    {
+      "epoch": 1.0027310745071265,
+      "grad_norm": 0.4136489943272252,
+      "learning_rate": 3.69846990389479e-05,
+      "loss": 0.4111,
+      "step": 5875
+    },
+    {
+      "epoch": 1.0035845352906034,
+      "grad_norm": 0.61166635664291,
+      "learning_rate": 3.696889226100152e-05,
+      "loss": 0.3794,
+      "step": 5880
+    },
+    {
+      "epoch": 1.0044379960740804,
+      "grad_norm": 0.42563074032686277,
+      "learning_rate": 3.6953085483055134e-05,
+      "loss": 0.3654,
+      "step": 5885
+    },
+    {
+      "epoch": 1.0052914568575575,
+      "grad_norm": 0.4164695520452888,
+      "learning_rate": 3.693727870510875e-05,
+      "loss": 0.4062,
+      "step": 5890
+    },
+    {
+      "epoch": 1.0061449176410344,
+      "grad_norm": 0.35877236247587296,
+      "learning_rate": 3.692147192716237e-05,
+      "loss": 0.3721,
+      "step": 5895
+    },
+    {
+      "epoch": 1.0069983784245113,
+      "grad_norm": 0.4070961912817456,
+      "learning_rate": 3.6905665149215986e-05,
+      "loss": 0.3549,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0078518392079885,
+      "grad_norm": 0.4514198162271714,
+      "learning_rate": 3.6889858371269603e-05,
+      "loss": 0.3823,
+      "step": 5905
+    },
+    {
+      "epoch": 1.0087052999914654,
+      "grad_norm": 0.35551701094292903,
+      "learning_rate": 3.6874051593323214e-05,
+      "loss": 0.3563,
+      "step": 5910
+    },
+    {
+      "epoch": 1.0095587607749423,
+      "grad_norm": 0.4206702553978487,
+      "learning_rate": 3.685824481537684e-05,
+      "loss": 0.39,
+      "step": 5915
+    },
+    {
+      "epoch": 1.0104122215584195,
+      "grad_norm": 0.4050493735514043,
+      "learning_rate": 3.6842438037430456e-05,
+      "loss": 0.3833,
+      "step": 5920
+    },
+    {
+      "epoch": 1.0112656823418964,
+      "grad_norm": 0.41089629306713504,
+      "learning_rate": 3.6826631259484067e-05,
+      "loss": 0.391,
+      "step": 5925
+    },
+    {
+      "epoch": 1.0121191431253733,
+      "grad_norm": 0.4847686846807888,
+      "learning_rate": 3.6810824481537684e-05,
+      "loss": 0.3576,
+      "step": 5930
+    },
+    {
+      "epoch": 1.0129726039088505,
+      "grad_norm": 0.4310668413326392,
+      "learning_rate": 3.67950177035913e-05,
+      "loss": 0.4042,
+      "step": 5935
+    },
+    {
+      "epoch": 1.0138260646923274,
+      "grad_norm": 0.3616617447543886,
+      "learning_rate": 3.677921092564492e-05,
+      "loss": 0.4214,
+      "step": 5940
+    },
+    {
+      "epoch": 1.0146795254758043,
+      "grad_norm": 0.3256661980486338,
+      "learning_rate": 3.676340414769853e-05,
+      "loss": 0.3656,
+      "step": 5945
+    },
+    {
+      "epoch": 1.0155329862592815,
+      "grad_norm": 0.37290726450601397,
+      "learning_rate": 3.6747597369752154e-05,
+      "loss": 0.3952,
+      "step": 5950
+    },
+    {
+      "epoch": 1.0163864470427584,
+      "grad_norm": 0.33989355267900506,
+      "learning_rate": 3.673179059180577e-05,
+      "loss": 0.3617,
+      "step": 5955
+    },
+    {
+      "epoch": 1.0172399078262353,
+      "grad_norm": 0.39699402875988227,
+      "learning_rate": 3.671598381385938e-05,
+      "loss": 0.3721,
+      "step": 5960
+    },
+    {
+      "epoch": 1.0180933686097124,
+      "grad_norm": 0.3264087946543946,
+      "learning_rate": 3.6700177035913e-05,
+      "loss": 0.3734,
+      "step": 5965
+    },
+    {
+      "epoch": 1.0189468293931894,
+      "grad_norm": 0.4111875246383104,
+      "learning_rate": 3.668437025796662e-05,
+      "loss": 0.3915,
+      "step": 5970
+    },
+    {
+      "epoch": 1.0198002901766663,
+      "grad_norm": 0.3995828309649097,
+      "learning_rate": 3.6668563480020234e-05,
+      "loss": 0.3882,
+      "step": 5975
+    },
+    {
+      "epoch": 1.0206537509601434,
+      "grad_norm": 0.3747886147565575,
+      "learning_rate": 3.6652756702073845e-05,
+      "loss": 0.3916,
+      "step": 5980
+    },
+    {
+      "epoch": 1.0215072117436204,
+      "grad_norm": 0.3372085408259476,
+      "learning_rate": 3.663694992412747e-05,
+      "loss": 0.3849,
+      "step": 5985
+    },
+    {
+      "epoch": 1.0223606725270973,
+      "grad_norm": 0.4160692716074461,
+      "learning_rate": 3.662114314618109e-05,
+      "loss": 0.3641,
+      "step": 5990
+    },
+    {
+      "epoch": 1.0232141333105744,
+      "grad_norm": 0.3729675376120794,
+      "learning_rate": 3.66053363682347e-05,
+      "loss": 0.4029,
+      "step": 5995
+    },
+    {
+      "epoch": 1.0240675940940513,
+      "grad_norm": 0.33227559391363237,
+      "learning_rate": 3.658952959028832e-05,
+      "loss": 0.3856,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0249210548775283,
+      "grad_norm": 0.35293680411023437,
+      "learning_rate": 3.657372281234193e-05,
+      "loss": 0.3484,
+      "step": 6005
+    },
+    {
+      "epoch": 1.0257745156610054,
+      "grad_norm": 0.4034296244720003,
+      "learning_rate": 3.655791603439555e-05,
+      "loss": 0.3546,
+      "step": 6010
+    },
+    {
+      "epoch": 1.0266279764444823,
+      "grad_norm": 0.38737538918982256,
+      "learning_rate": 3.654210925644917e-05,
+      "loss": 0.3696,
+      "step": 6015
+    },
+    {
+      "epoch": 1.0274814372279595,
+      "grad_norm": 0.35939055045698165,
+      "learning_rate": 3.6526302478502785e-05,
+      "loss": 0.386,
+      "step": 6020
+    },
+    {
+      "epoch": 1.0283348980114364,
+      "grad_norm": 0.3428720112959876,
+      "learning_rate": 3.65104957005564e-05,
+      "loss": 0.3796,
+      "step": 6025
+    },
+    {
+      "epoch": 1.0291883587949133,
+      "grad_norm": 0.33521610732253654,
+      "learning_rate": 3.649468892261001e-05,
+      "loss": 0.3781,
+      "step": 6030
+    },
+    {
+      "epoch": 1.0300418195783905,
+      "grad_norm": 0.37984871713555923,
+      "learning_rate": 3.647888214466364e-05,
+      "loss": 0.3796,
+      "step": 6035
+    },
+    {
+      "epoch": 1.0308952803618674,
+      "grad_norm": 0.5985744943668265,
+      "learning_rate": 3.646307536671725e-05,
+      "loss": 0.3668,
+      "step": 6040
+    },
+    {
+      "epoch": 1.0317487411453443,
+      "grad_norm": 0.3352657163150749,
+      "learning_rate": 3.6447268588770865e-05,
+      "loss": 0.409,
+      "step": 6045
+    },
+    {
+      "epoch": 1.0326022019288215,
+      "grad_norm": 0.3491380338159307,
+      "learning_rate": 3.643146181082448e-05,
+      "loss": 0.3683,
+      "step": 6050
+    },
+    {
+      "epoch": 1.0334556627122984,
+      "grad_norm": 0.46292870144229437,
+      "learning_rate": 3.64156550328781e-05,
+      "loss": 0.3837,
+      "step": 6055
+    },
+    {
+      "epoch": 1.0343091234957753,
+      "grad_norm": 0.30738126723971215,
+      "learning_rate": 3.639984825493172e-05,
+      "loss": 0.3796,
+      "step": 6060
+    },
+    {
+      "epoch": 1.0351625842792525,
+      "grad_norm": 0.34472849345856665,
+      "learning_rate": 3.638404147698533e-05,
+      "loss": 0.3911,
+      "step": 6065
+    },
+    {
+      "epoch": 1.0360160450627294,
+      "grad_norm": 0.3346180101688685,
+      "learning_rate": 3.636823469903895e-05,
+      "loss": 0.3668,
+      "step": 6070
+    },
+    {
+      "epoch": 1.0368695058462063,
+      "grad_norm": 0.35690751165072565,
+      "learning_rate": 3.635242792109257e-05,
+      "loss": 0.3778,
+      "step": 6075
+    },
+    {
+      "epoch": 1.0377229666296834,
+      "grad_norm": 0.3677037386510019,
+      "learning_rate": 3.633662114314618e-05,
+      "loss": 0.37,
+      "step": 6080
+    },
+    {
+      "epoch": 1.0385764274131604,
+      "grad_norm": 0.4255067817756184,
+      "learning_rate": 3.63208143651998e-05,
+      "loss": 0.3993,
+      "step": 6085
+    },
+    {
+      "epoch": 1.0394298881966373,
+      "grad_norm": 0.47237275766808884,
+      "learning_rate": 3.6305007587253416e-05,
+      "loss": 0.4001,
+      "step": 6090
+    },
+    {
+      "epoch": 1.0402833489801144,
+      "grad_norm": 0.3160522482641632,
+      "learning_rate": 3.628920080930703e-05,
+      "loss": 0.3811,
+      "step": 6095
+    },
+    {
+      "epoch": 1.0411368097635914,
+      "grad_norm": 0.420033345369002,
+      "learning_rate": 3.6273394031360644e-05,
+      "loss": 0.3561,
+      "step": 6100
+    },
+    {
+      "epoch": 1.0419902705470683,
+      "grad_norm": 0.45986808309255994,
+      "learning_rate": 3.625758725341427e-05,
+      "loss": 0.3515,
+      "step": 6105
+    },
+    {
+      "epoch": 1.0428437313305454,
+      "grad_norm": 0.4289388137838117,
+      "learning_rate": 3.6241780475467885e-05,
+      "loss": 0.3903,
+      "step": 6110
+    },
+    {
+      "epoch": 1.0436971921140223,
+      "grad_norm": 0.44362015679472544,
+      "learning_rate": 3.6225973697521496e-05,
+      "loss": 0.3781,
+      "step": 6115
+    },
+    {
+      "epoch": 1.0445506528974993,
+      "grad_norm": 0.415183964660911,
+      "learning_rate": 3.621016691957512e-05,
+      "loss": 0.39,
+      "step": 6120
+    },
+    {
+      "epoch": 1.0454041136809764,
+      "grad_norm": 0.31920896239376617,
+      "learning_rate": 3.619436014162873e-05,
+      "loss": 0.4002,
+      "step": 6125
+    },
+    {
+      "epoch": 1.0462575744644533,
+      "grad_norm": 0.34023466791164747,
+      "learning_rate": 3.617855336368235e-05,
+      "loss": 0.3885,
+      "step": 6130
+    },
+    {
+      "epoch": 1.0471110352479303,
+      "grad_norm": 0.398273235152284,
+      "learning_rate": 3.6162746585735966e-05,
+      "loss": 0.3781,
+      "step": 6135
+    },
+    {
+      "epoch": 1.0479644960314074,
+      "grad_norm": 0.3513130838940666,
+      "learning_rate": 3.614693980778958e-05,
+      "loss": 0.3908,
+      "step": 6140
+    },
+    {
+      "epoch": 1.0488179568148843,
+      "grad_norm": 0.28614314157111614,
+      "learning_rate": 3.61311330298432e-05,
+      "loss": 0.3478,
+      "step": 6145
+    },
+    {
+      "epoch": 1.0496714175983612,
+      "grad_norm": 0.34306163268100404,
+      "learning_rate": 3.611532625189681e-05,
+      "loss": 0.3856,
+      "step": 6150
+    },
+    {
+      "epoch": 1.0505248783818384,
+      "grad_norm": 0.36278534942222906,
+      "learning_rate": 3.6099519473950436e-05,
+      "loss": 0.4104,
+      "step": 6155
+    },
+    {
+      "epoch": 1.0513783391653153,
+      "grad_norm": 0.38037742648197687,
+      "learning_rate": 3.6083712696004046e-05,
+      "loss": 0.3489,
+      "step": 6160
+    },
+    {
+      "epoch": 1.0522317999487925,
+      "grad_norm": 0.37240475311150545,
+      "learning_rate": 3.6067905918057664e-05,
+      "loss": 0.3747,
+      "step": 6165
+    },
+    {
+      "epoch": 1.0530852607322694,
+      "grad_norm": 0.34139587197340454,
+      "learning_rate": 3.605209914011128e-05,
+      "loss": 0.3767,
+      "step": 6170
+    },
+    {
+      "epoch": 1.0539387215157463,
+      "grad_norm": 0.40863545347694546,
+      "learning_rate": 3.60362923621649e-05,
+      "loss": 0.3654,
+      "step": 6175
+    },
+    {
+      "epoch": 1.0547921822992234,
+      "grad_norm": 0.3753918782608841,
+      "learning_rate": 3.6020485584218516e-05,
+      "loss": 0.3605,
+      "step": 6180
+    },
+    {
+      "epoch": 1.0556456430827004,
+      "grad_norm": 0.32009725853280235,
+      "learning_rate": 3.600467880627213e-05,
+      "loss": 0.3655,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0564991038661773,
+      "grad_norm": 0.5092431911324296,
+      "learning_rate": 3.598887202832575e-05,
+      "loss": 0.3881,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0573525646496544,
+      "grad_norm": 0.43456819194948054,
+      "learning_rate": 3.597306525037936e-05,
+      "loss": 0.3669,
+      "step": 6195
+    },
+    {
+      "epoch": 1.0582060254331314,
+      "grad_norm": 0.3243068352215124,
+      "learning_rate": 3.595725847243298e-05,
+      "loss": 0.3541,
+      "step": 6200
+    },
+    {
+      "epoch": 1.0590594862166083,
+      "grad_norm": 0.5383990238343098,
+      "learning_rate": 3.59414516944866e-05,
+      "loss": 0.3623,
+      "step": 6205
+    },
+    {
+      "epoch": 1.0599129470000854,
+      "grad_norm": 0.3541289721102046,
+      "learning_rate": 3.5925644916540214e-05,
+      "loss": 0.4034,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0607664077835623,
+      "grad_norm": 0.38223980100958427,
+      "learning_rate": 3.590983813859383e-05,
+      "loss": 0.3959,
+      "step": 6215
+    },
+    {
+      "epoch": 1.0616198685670393,
+      "grad_norm": 0.3215840779602491,
+      "learning_rate": 3.589403136064744e-05,
+      "loss": 0.3737,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0624733293505164,
+      "grad_norm": 0.36266124166010544,
+      "learning_rate": 3.5878224582701067e-05,
+      "loss": 0.3575,
+      "step": 6225
+    },
+    {
+      "epoch": 1.0633267901339933,
+      "grad_norm": 0.3668600032359126,
+      "learning_rate": 3.5862417804754684e-05,
+      "loss": 0.3597,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0641802509174703,
+      "grad_norm": 0.2960195006335881,
+      "learning_rate": 3.5846611026808295e-05,
+      "loss": 0.3921,
+      "step": 6235
+    },
+    {
+      "epoch": 1.0650337117009474,
+      "grad_norm": 0.31845474075202623,
+      "learning_rate": 3.583080424886192e-05,
+      "loss": 0.3911,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0658871724844243,
+      "grad_norm": 0.3725948362521795,
+      "learning_rate": 3.581499747091553e-05,
+      "loss": 0.3888,
+      "step": 6245
+    },
+    {
+      "epoch": 1.0667406332679013,
+      "grad_norm": 0.34149111645486385,
+      "learning_rate": 3.579919069296915e-05,
+      "loss": 0.3674,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0675940940513784,
+      "grad_norm": 0.4858579069975639,
+      "learning_rate": 3.5783383915022765e-05,
+      "loss": 0.4212,
+      "step": 6255
+    },
+    {
+      "epoch": 1.0684475548348553,
+      "grad_norm": 0.37017559385021276,
+      "learning_rate": 3.576757713707638e-05,
+      "loss": 0.3764,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0693010156183322,
+      "grad_norm": 0.3501088365141956,
+      "learning_rate": 3.575177035913e-05,
+      "loss": 0.381,
+      "step": 6265
+    },
+    {
+      "epoch": 1.0701544764018094,
+      "grad_norm": 0.4536112262409545,
+      "learning_rate": 3.573596358118361e-05,
+      "loss": 0.3927,
+      "step": 6270
+    },
+    {
+      "epoch": 1.0710079371852863,
+      "grad_norm": 0.3581084690276881,
+      "learning_rate": 3.5720156803237234e-05,
+      "loss": 0.3531,
+      "step": 6275
+    },
+    {
+      "epoch": 1.0718613979687635,
+      "grad_norm": 0.35442213729757155,
+      "learning_rate": 3.5704350025290845e-05,
+      "loss": 0.3679,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0727148587522404,
+      "grad_norm": 0.3300824599413051,
+      "learning_rate": 3.568854324734446e-05,
+      "loss": 0.3536,
+      "step": 6285
+    },
+    {
+      "epoch": 1.0735683195357173,
+      "grad_norm": 0.37005100248415596,
+      "learning_rate": 3.567273646939808e-05,
+      "loss": 0.3753,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0744217803191942,
+      "grad_norm": 0.37767347033399445,
+      "learning_rate": 3.56569296914517e-05,
+      "loss": 0.3748,
+      "step": 6295
+    },
+    {
+      "epoch": 1.0752752411026714,
+      "grad_norm": 0.3549726169959677,
+      "learning_rate": 3.5641122913505315e-05,
+      "loss": 0.3704,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0761287018861483,
+      "grad_norm": 0.42642399641544837,
+      "learning_rate": 3.5625316135558926e-05,
+      "loss": 0.3947,
+      "step": 6305
+    },
+    {
+      "epoch": 1.0769821626696254,
+      "grad_norm": 0.4184430610863713,
+      "learning_rate": 3.560950935761255e-05,
+      "loss": 0.4113,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0778356234531024,
+      "grad_norm": 0.34839888808095437,
+      "learning_rate": 3.559370257966616e-05,
+      "loss": 0.3741,
+      "step": 6315
+    },
+    {
+      "epoch": 1.0786890842365793,
+      "grad_norm": 0.3541232583117419,
+      "learning_rate": 3.557789580171978e-05,
+      "loss": 0.3609,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0795425450200564,
+      "grad_norm": 0.4127048987151015,
+      "learning_rate": 3.5562089023773395e-05,
+      "loss": 0.3681,
+      "step": 6325
+    },
+    {
+      "epoch": 1.0803960058035333,
+      "grad_norm": 0.3967872211761154,
+      "learning_rate": 3.554628224582701e-05,
+      "loss": 0.3852,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0812494665870103,
+      "grad_norm": 0.40621764304381197,
+      "learning_rate": 3.553047546788063e-05,
+      "loss": 0.3631,
+      "step": 6335
+    },
+    {
+      "epoch": 1.0821029273704874,
+      "grad_norm": 0.3507900861501906,
+      "learning_rate": 3.551466868993424e-05,
+      "loss": 0.4227,
+      "step": 6340
+    },
+    {
+      "epoch": 1.0829563881539643,
+      "grad_norm": 0.34308174986128376,
+      "learning_rate": 3.5498861911987865e-05,
+      "loss": 0.3793,
+      "step": 6345
+    },
+    {
+      "epoch": 1.0838098489374413,
+      "grad_norm": 0.7806539078148241,
+      "learning_rate": 3.5483055134041476e-05,
+      "loss": 0.3627,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0846633097209184,
+      "grad_norm": 0.40689762094410925,
+      "learning_rate": 3.546724835609509e-05,
+      "loss": 0.3799,
+      "step": 6355
+    },
+    {
+      "epoch": 1.0855167705043953,
+      "grad_norm": 0.3372893848996216,
+      "learning_rate": 3.545144157814871e-05,
+      "loss": 0.3706,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0863702312878722,
+      "grad_norm": 0.5105115199964079,
+      "learning_rate": 3.543563480020233e-05,
+      "loss": 0.3517,
+      "step": 6365
+    },
+    {
+      "epoch": 1.0872236920713494,
+      "grad_norm": 0.49674907906461885,
+      "learning_rate": 3.5419828022255946e-05,
+      "loss": 0.3976,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0880771528548263,
+      "grad_norm": 0.3616314250841704,
+      "learning_rate": 3.540402124430956e-05,
+      "loss": 0.3624,
+      "step": 6375
+    },
+    {
+      "epoch": 1.0889306136383032,
+      "grad_norm": 0.3261095293861316,
+      "learning_rate": 3.538821446636318e-05,
+      "loss": 0.4025,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0897840744217804,
+      "grad_norm": 0.3995130514496934,
+      "learning_rate": 3.537240768841679e-05,
+      "loss": 0.3816,
+      "step": 6385
+    },
+    {
+      "epoch": 1.0906375352052573,
+      "grad_norm": 0.6368223818268357,
+      "learning_rate": 3.535660091047041e-05,
+      "loss": 0.3668,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0914909959887342,
+      "grad_norm": 0.40057929508373474,
+      "learning_rate": 3.534079413252403e-05,
+      "loss": 0.3464,
+      "step": 6395
+    },
+    {
+      "epoch": 1.0923444567722114,
+      "grad_norm": 0.4494989616048801,
+      "learning_rate": 3.5324987354577644e-05,
+      "loss": 0.3896,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0931979175556883,
+      "grad_norm": 0.3670399719103569,
+      "learning_rate": 3.530918057663126e-05,
+      "loss": 0.371,
+      "step": 6405
+    },
+    {
+      "epoch": 1.0940513783391652,
+      "grad_norm": 0.38767765917433505,
+      "learning_rate": 3.529337379868488e-05,
+      "loss": 0.355,
+      "step": 6410
+    },
+    {
+      "epoch": 1.0949048391226424,
+      "grad_norm": 0.3578300732267279,
+      "learning_rate": 3.5277567020738496e-05,
+      "loss": 0.3777,
+      "step": 6415
+    },
+    {
+      "epoch": 1.0957582999061193,
+      "grad_norm": 0.36797054114616534,
+      "learning_rate": 3.5261760242792114e-05,
+      "loss": 0.3748,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0966117606895964,
+      "grad_norm": 0.3752746066914836,
+      "learning_rate": 3.5245953464845724e-05,
+      "loss": 0.3951,
+      "step": 6425
+    },
+    {
+      "epoch": 1.0974652214730733,
+      "grad_norm": 0.4992245034330738,
+      "learning_rate": 3.523014668689935e-05,
+      "loss": 0.3996,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0983186822565503,
+      "grad_norm": 0.3802733826060877,
+      "learning_rate": 3.521433990895296e-05,
+      "loss": 0.3683,
+      "step": 6435
+    },
+    {
+      "epoch": 1.0991721430400272,
+      "grad_norm": 0.4479737303674627,
+      "learning_rate": 3.5198533131006577e-05,
+      "loss": 0.3849,
+      "step": 6440
+    },
+    {
+      "epoch": 1.1000256038235043,
+      "grad_norm": 0.42426668896227515,
+      "learning_rate": 3.5182726353060194e-05,
+      "loss": 0.3919,
+      "step": 6445
+    },
+    {
+      "epoch": 1.1008790646069813,
+      "grad_norm": 0.48513194649690067,
+      "learning_rate": 3.516691957511381e-05,
+      "loss": 0.3588,
+      "step": 6450
+    },
+    {
+      "epoch": 1.1017325253904584,
+      "grad_norm": 0.414365401738906,
+      "learning_rate": 3.515111279716743e-05,
+      "loss": 0.3514,
+      "step": 6455
+    },
+    {
+      "epoch": 1.1025859861739353,
+      "grad_norm": 0.3772732173172724,
+      "learning_rate": 3.513530601922104e-05,
+      "loss": 0.3982,
+      "step": 6460
+    },
+    {
+      "epoch": 1.1034394469574123,
+      "grad_norm": 0.5082277911704493,
+      "learning_rate": 3.5119499241274664e-05,
+      "loss": 0.3986,
+      "step": 6465
+    },
+    {
+      "epoch": 1.1042929077408894,
+      "grad_norm": 0.5876443202759482,
+      "learning_rate": 3.5103692463328275e-05,
+      "loss": 0.3659,
+      "step": 6470
+    },
+    {
+      "epoch": 1.1051463685243663,
+      "grad_norm": 0.3388469265049481,
+      "learning_rate": 3.508788568538189e-05,
+      "loss": 0.3699,
+      "step": 6475
+    },
+    {
+      "epoch": 1.1059998293078432,
+      "grad_norm": 0.4262580363531045,
+      "learning_rate": 3.507207890743551e-05,
+      "loss": 0.3553,
+      "step": 6480
+    },
+    {
+      "epoch": 1.1068532900913204,
+      "grad_norm": 0.4133261577814819,
+      "learning_rate": 3.505627212948913e-05,
+      "loss": 0.3917,
+      "step": 6485
+    },
+    {
+      "epoch": 1.1077067508747973,
+      "grad_norm": 0.29908511017730205,
+      "learning_rate": 3.5040465351542744e-05,
+      "loss": 0.3848,
+      "step": 6490
+    },
+    {
+      "epoch": 1.1085602116582742,
+      "grad_norm": 0.4519310135103528,
+      "learning_rate": 3.5024658573596355e-05,
+      "loss": 0.3669,
+      "step": 6495
+    },
+    {
+      "epoch": 1.1094136724417514,
+      "grad_norm": 0.36753690476682793,
+      "learning_rate": 3.500885179564998e-05,
+      "loss": 0.362,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1102671332252283,
+      "grad_norm": 0.43419707034688587,
+      "learning_rate": 3.499304501770359e-05,
+      "loss": 0.3717,
+      "step": 6505
+    },
+    {
+      "epoch": 1.1111205940087052,
+      "grad_norm": 0.3362000973234563,
+      "learning_rate": 3.497723823975721e-05,
+      "loss": 0.3707,
+      "step": 6510
+    },
+    {
+      "epoch": 1.1119740547921824,
+      "grad_norm": 0.41338929812634956,
+      "learning_rate": 3.496143146181083e-05,
+      "loss": 0.3636,
+      "step": 6515
+    },
+    {
+      "epoch": 1.1128275155756593,
+      "grad_norm": 0.38546860525336446,
+      "learning_rate": 3.494562468386444e-05,
+      "loss": 0.3432,
+      "step": 6520
+    },
+    {
+      "epoch": 1.1136809763591362,
+      "grad_norm": 0.4837974451704121,
+      "learning_rate": 3.492981790591806e-05,
+      "loss": 0.3856,
+      "step": 6525
+    },
+    {
+      "epoch": 1.1145344371426134,
+      "grad_norm": 0.4099945195244458,
+      "learning_rate": 3.491401112797168e-05,
+      "loss": 0.3991,
+      "step": 6530
+    },
+    {
+      "epoch": 1.1153878979260903,
+      "grad_norm": 0.4090729970162646,
+      "learning_rate": 3.4898204350025295e-05,
+      "loss": 0.399,
+      "step": 6535
+    },
+    {
+      "epoch": 1.1162413587095672,
+      "grad_norm": 0.3293074613655451,
+      "learning_rate": 3.4882397572078905e-05,
+      "loss": 0.3773,
+      "step": 6540
+    },
+    {
+      "epoch": 1.1170948194930443,
+      "grad_norm": 0.31970254478771015,
+      "learning_rate": 3.486659079413252e-05,
+      "loss": 0.4063,
+      "step": 6545
+    },
+    {
+      "epoch": 1.1179482802765213,
+      "grad_norm": 0.3404686139395737,
+      "learning_rate": 3.485078401618615e-05,
+      "loss": 0.3783,
+      "step": 6550
+    },
+    {
+      "epoch": 1.1188017410599982,
+      "grad_norm": 0.37921888621672023,
+      "learning_rate": 3.483497723823976e-05,
+      "loss": 0.3732,
+      "step": 6555
+    },
+    {
+      "epoch": 1.1196552018434753,
+      "grad_norm": 0.3937102414061394,
+      "learning_rate": 3.4819170460293375e-05,
+      "loss": 0.3667,
+      "step": 6560
+    },
+    {
+      "epoch": 1.1205086626269523,
+      "grad_norm": 0.34761818920877413,
+      "learning_rate": 3.480336368234699e-05,
+      "loss": 0.3588,
+      "step": 6565
+    },
+    {
+      "epoch": 1.1213621234104294,
+      "grad_norm": 0.34424410517904946,
+      "learning_rate": 3.478755690440061e-05,
+      "loss": 0.3841,
+      "step": 6570
+    },
+    {
+      "epoch": 1.1222155841939063,
+      "grad_norm": 0.4214224963242259,
+      "learning_rate": 3.477175012645423e-05,
+      "loss": 0.3739,
+      "step": 6575
+    },
+    {
+      "epoch": 1.1230690449773832,
+      "grad_norm": 0.3865334872789258,
+      "learning_rate": 3.475594334850784e-05,
+      "loss": 0.3947,
+      "step": 6580
+    },
+    {
+      "epoch": 1.1239225057608604,
+      "grad_norm": 0.40708734545542935,
+      "learning_rate": 3.474013657056146e-05,
+      "loss": 0.3706,
+      "step": 6585
+    },
+    {
+      "epoch": 1.1247759665443373,
+      "grad_norm": 0.9045066835294332,
+      "learning_rate": 3.472432979261507e-05,
+      "loss": 0.3699,
+      "step": 6590
+    },
+    {
+      "epoch": 1.1256294273278142,
+      "grad_norm": 0.345613816124089,
+      "learning_rate": 3.470852301466869e-05,
+      "loss": 0.3627,
+      "step": 6595
+    },
+    {
+      "epoch": 1.1264828881112914,
+      "grad_norm": 0.4417786367821901,
+      "learning_rate": 3.469271623672231e-05,
+      "loss": 0.3952,
+      "step": 6600
+    },
+    {
+      "epoch": 1.1273363488947683,
+      "grad_norm": 0.3488869958861024,
+      "learning_rate": 3.4676909458775926e-05,
+      "loss": 0.383,
+      "step": 6605
+    },
+    {
+      "epoch": 1.1281898096782452,
+      "grad_norm": 0.32968002715251665,
+      "learning_rate": 3.466110268082954e-05,
+      "loss": 0.3885,
+      "step": 6610
+    },
+    {
+      "epoch": 1.1290432704617224,
+      "grad_norm": 0.32465319986861174,
+      "learning_rate": 3.4645295902883154e-05,
+      "loss": 0.3843,
+      "step": 6615
+    },
+    {
+      "epoch": 1.1298967312451993,
+      "grad_norm": 0.42478512459153034,
+      "learning_rate": 3.462948912493678e-05,
+      "loss": 0.3774,
+      "step": 6620
+    },
+    {
+      "epoch": 1.1307501920286762,
+      "grad_norm": 0.3691686696446473,
+      "learning_rate": 3.461368234699039e-05,
+      "loss": 0.3477,
+      "step": 6625
+    },
+    {
+      "epoch": 1.1316036528121534,
+      "grad_norm": 0.3294388018000737,
+      "learning_rate": 3.4597875569044006e-05,
+      "loss": 0.3521,
+      "step": 6630
+    },
+    {
+      "epoch": 1.1324571135956303,
+      "grad_norm": 0.5078573889585375,
+      "learning_rate": 3.4582068791097624e-05,
+      "loss": 0.331,
+      "step": 6635
+    },
+    {
+      "epoch": 1.1333105743791072,
+      "grad_norm": 0.3540787108069881,
+      "learning_rate": 3.456626201315124e-05,
+      "loss": 0.3548,
+      "step": 6640
+    },
+    {
+      "epoch": 1.1341640351625843,
+      "grad_norm": 0.425664971743762,
+      "learning_rate": 3.455045523520486e-05,
+      "loss": 0.3614,
+      "step": 6645
+    },
+    {
+      "epoch": 1.1350174959460613,
+      "grad_norm": 0.4519194755911954,
+      "learning_rate": 3.4534648457258476e-05,
+      "loss": 0.376,
+      "step": 6650
+    },
+    {
+      "epoch": 1.1358709567295382,
+      "grad_norm": 0.3792641841284432,
+      "learning_rate": 3.451884167931209e-05,
+      "loss": 0.3663,
+      "step": 6655
+    },
+    {
+      "epoch": 1.1367244175130153,
+      "grad_norm": 0.38336214452272527,
+      "learning_rate": 3.4503034901365704e-05,
+      "loss": 0.3729,
+      "step": 6660
+    },
+    {
+      "epoch": 1.1375778782964923,
+      "grad_norm": 0.44490882098941625,
+      "learning_rate": 3.448722812341932e-05,
+      "loss": 0.3898,
+      "step": 6665
+    },
+    {
+      "epoch": 1.1384313390799692,
+      "grad_norm": 0.4923210533563228,
+      "learning_rate": 3.4471421345472946e-05,
+      "loss": 0.3671,
+      "step": 6670
+    },
+    {
+      "epoch": 1.1392847998634463,
+      "grad_norm": 0.32576638199942004,
+      "learning_rate": 3.4455614567526556e-05,
+      "loss": 0.408,
+      "step": 6675
+    },
+    {
+      "epoch": 1.1401382606469233,
+      "grad_norm": 0.3698146388980988,
+      "learning_rate": 3.4439807789580174e-05,
+      "loss": 0.3656,
+      "step": 6680
+    },
+    {
+      "epoch": 1.1409917214304004,
+      "grad_norm": 0.4038930256287759,
+      "learning_rate": 3.442400101163379e-05,
+      "loss": 0.3809,
+      "step": 6685
+    },
+    {
+      "epoch": 1.1418451822138773,
+      "grad_norm": 0.3678224472834901,
+      "learning_rate": 3.440819423368741e-05,
+      "loss": 0.4072,
+      "step": 6690
+    },
+    {
+      "epoch": 1.1426986429973542,
+      "grad_norm": 0.37674876188896234,
+      "learning_rate": 3.439238745574102e-05,
+      "loss": 0.3597,
+      "step": 6695
+    },
+    {
+      "epoch": 1.1435521037808312,
+      "grad_norm": 0.36763677912434917,
+      "learning_rate": 3.437658067779464e-05,
+      "loss": 0.3937,
+      "step": 6700
+    },
+    {
+      "epoch": 1.1444055645643083,
+      "grad_norm": 0.34982750551205294,
+      "learning_rate": 3.436077389984826e-05,
+      "loss": 0.3937,
+      "step": 6705
+    },
+    {
+      "epoch": 1.1452590253477852,
+      "grad_norm": 0.769092786227432,
+      "learning_rate": 3.434496712190187e-05,
+      "loss": 0.3773,
+      "step": 6710
+    },
+    {
+      "epoch": 1.1461124861312624,
+      "grad_norm": 0.3615013250403249,
+      "learning_rate": 3.432916034395549e-05,
+      "loss": 0.3776,
+      "step": 6715
+    },
+    {
+      "epoch": 1.1469659469147393,
+      "grad_norm": 0.4250703191148808,
+      "learning_rate": 3.431335356600911e-05,
+      "loss": 0.3793,
+      "step": 6720
+    },
+    {
+      "epoch": 1.1478194076982162,
+      "grad_norm": 0.375871467769371,
+      "learning_rate": 3.4297546788062724e-05,
+      "loss": 0.3661,
+      "step": 6725
+    },
+    {
+      "epoch": 1.1486728684816931,
+      "grad_norm": 0.3965560488665287,
+      "learning_rate": 3.428174001011634e-05,
+      "loss": 0.3926,
+      "step": 6730
+    },
+    {
+      "epoch": 1.1495263292651703,
+      "grad_norm": 0.33386379563780616,
+      "learning_rate": 3.426593323216995e-05,
+      "loss": 0.3415,
+      "step": 6735
+    },
+    {
+      "epoch": 1.1503797900486472,
+      "grad_norm": 0.49744433445623903,
+      "learning_rate": 3.4250126454223577e-05,
+      "loss": 0.3984,
+      "step": 6740
+    },
+    {
+      "epoch": 1.1512332508321244,
+      "grad_norm": 0.35276686175575156,
+      "learning_rate": 3.423431967627719e-05,
+      "loss": 0.3796,
+      "step": 6745
+    },
+    {
+      "epoch": 1.1520867116156013,
+      "grad_norm": 0.3782105257502626,
+      "learning_rate": 3.4218512898330805e-05,
+      "loss": 0.3939,
+      "step": 6750
+    },
+    {
+      "epoch": 1.1529401723990782,
+      "grad_norm": 0.374200039079983,
+      "learning_rate": 3.420270612038442e-05,
+      "loss": 0.396,
+      "step": 6755
+    },
+    {
+      "epoch": 1.1537936331825553,
+      "grad_norm": 0.3836710306563444,
+      "learning_rate": 3.418689934243804e-05,
+      "loss": 0.3562,
+      "step": 6760
+    },
+    {
+      "epoch": 1.1546470939660323,
+      "grad_norm": 0.42364950051694567,
+      "learning_rate": 3.417109256449166e-05,
+      "loss": 0.3864,
+      "step": 6765
+    },
+    {
+      "epoch": 1.1555005547495092,
+      "grad_norm": 0.38861266994602445,
+      "learning_rate": 3.4155285786545275e-05,
+      "loss": 0.3844,
+      "step": 6770
+    },
+    {
+      "epoch": 1.1563540155329863,
+      "grad_norm": 0.6018078136039005,
+      "learning_rate": 3.413947900859889e-05,
+      "loss": 0.3489,
+      "step": 6775
+    },
+    {
+      "epoch": 1.1572074763164633,
+      "grad_norm": 0.5972874263639776,
+      "learning_rate": 3.41236722306525e-05,
+      "loss": 0.3945,
+      "step": 6780
+    },
+    {
+      "epoch": 1.1580609370999402,
+      "grad_norm": 0.4093857505193346,
+      "learning_rate": 3.410786545270612e-05,
+      "loss": 0.3791,
+      "step": 6785
+    },
+    {
+      "epoch": 1.1589143978834173,
+      "grad_norm": 0.42600590826537443,
+      "learning_rate": 3.409205867475974e-05,
+      "loss": 0.376,
+      "step": 6790
+    },
+    {
+      "epoch": 1.1597678586668942,
+      "grad_norm": 0.4065411805048917,
+      "learning_rate": 3.4076251896813355e-05,
+      "loss": 0.384,
+      "step": 6795
+    },
+    {
+      "epoch": 1.1606213194503712,
+      "grad_norm": 0.4244702012793157,
+      "learning_rate": 3.406044511886697e-05,
+      "loss": 0.3883,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1614747802338483,
+      "grad_norm": 0.374917508843201,
+      "learning_rate": 3.404463834092059e-05,
+      "loss": 0.3931,
+      "step": 6805
+    },
+    {
+      "epoch": 1.1623282410173252,
+      "grad_norm": 0.5211371296519884,
+      "learning_rate": 3.402883156297421e-05,
+      "loss": 0.3828,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1631817018008022,
+      "grad_norm": 0.7561755882811956,
+      "learning_rate": 3.401302478502782e-05,
+      "loss": 0.356,
+      "step": 6815
+    },
+    {
+      "epoch": 1.1640351625842793,
+      "grad_norm": 0.6436005006819279,
+      "learning_rate": 3.3997218007081436e-05,
+      "loss": 0.3848,
+      "step": 6820
+    },
+    {
+      "epoch": 1.1648886233677562,
+      "grad_norm": 0.42468254817266504,
+      "learning_rate": 3.398141122913506e-05,
+      "loss": 0.3802,
+      "step": 6825
+    },
+    {
+      "epoch": 1.1657420841512334,
+      "grad_norm": 0.3982367416624562,
+      "learning_rate": 3.396560445118867e-05,
+      "loss": 0.399,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1665955449347103,
+      "grad_norm": 0.5806335612905161,
+      "learning_rate": 3.394979767324229e-05,
+      "loss": 0.379,
+      "step": 6835
+    },
+    {
+      "epoch": 1.1674490057181872,
+      "grad_norm": 0.6539048969095816,
+      "learning_rate": 3.3933990895295905e-05,
+      "loss": 0.3814,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1683024665016641,
+      "grad_norm": 0.45131138747036637,
+      "learning_rate": 3.391818411734952e-05,
+      "loss": 0.3901,
+      "step": 6845
+    },
+    {
+      "epoch": 1.1691559272851413,
+      "grad_norm": 0.4857094349513415,
+      "learning_rate": 3.3902377339403134e-05,
+      "loss": 0.3933,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1700093880686182,
+      "grad_norm": 0.3439760439525917,
+      "learning_rate": 3.388657056145675e-05,
+      "loss": 0.3781,
+      "step": 6855
+    },
+    {
+      "epoch": 1.1708628488520954,
+      "grad_norm": 5.327471762596987,
+      "learning_rate": 3.3870763783510375e-05,
+      "loss": 0.3671,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1717163096355723,
+      "grad_norm": 0.37046626612176936,
+      "learning_rate": 3.3854957005563986e-05,
+      "loss": 0.3977,
+      "step": 6865
+    },
+    {
+      "epoch": 1.1725697704190492,
+      "grad_norm": 0.4327085457178763,
+      "learning_rate": 3.3839150227617603e-05,
+      "loss": 0.3748,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1734232312025263,
+      "grad_norm": 0.333034919237716,
+      "learning_rate": 3.382334344967122e-05,
+      "loss": 0.3933,
+      "step": 6875
+    },
+    {
+      "epoch": 1.1742766919860033,
+      "grad_norm": 0.35345434918838137,
+      "learning_rate": 3.380753667172484e-05,
+      "loss": 0.4103,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1751301527694802,
+      "grad_norm": 0.3421462474321589,
+      "learning_rate": 3.3791729893778456e-05,
+      "loss": 0.3832,
+      "step": 6885
+    },
+    {
+      "epoch": 1.1759836135529573,
+      "grad_norm": 0.3561427234439529,
+      "learning_rate": 3.377592311583207e-05,
+      "loss": 0.3839,
+      "step": 6890
+    },
+    {
+      "epoch": 1.1768370743364343,
+      "grad_norm": 0.3531319457485075,
+      "learning_rate": 3.376011633788569e-05,
+      "loss": 0.3836,
+      "step": 6895
+    },
+    {
+      "epoch": 1.1776905351199112,
+      "grad_norm": 0.8047701928164278,
+      "learning_rate": 3.37443095599393e-05,
+      "loss": 0.3878,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1785439959033883,
+      "grad_norm": 0.33725477031687834,
+      "learning_rate": 3.372850278199292e-05,
+      "loss": 0.3529,
+      "step": 6905
+    },
+    {
+      "epoch": 1.1793974566868652,
+      "grad_norm": 0.4488080219229851,
+      "learning_rate": 3.3712696004046536e-05,
+      "loss": 0.3735,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1802509174703422,
+      "grad_norm": 0.4004718951161043,
+      "learning_rate": 3.3696889226100154e-05,
+      "loss": 0.3768,
+      "step": 6915
+    },
+    {
+      "epoch": 1.1811043782538193,
+      "grad_norm": 0.3308681548917742,
+      "learning_rate": 3.368108244815377e-05,
+      "loss": 0.3966,
+      "step": 6920
+    },
+    {
+      "epoch": 1.1819578390372962,
+      "grad_norm": 0.41582127810301883,
+      "learning_rate": 3.366527567020739e-05,
+      "loss": 0.3705,
+      "step": 6925
+    },
+    {
+      "epoch": 1.1828112998207732,
+      "grad_norm": 0.4892347947895599,
+      "learning_rate": 3.3649468892261006e-05,
+      "loss": 0.3847,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1836647606042503,
+      "grad_norm": 0.49948370788401486,
+      "learning_rate": 3.363366211431462e-05,
+      "loss": 0.3879,
+      "step": 6935
+    },
+    {
+      "epoch": 1.1845182213877272,
+      "grad_norm": 0.5170019696339748,
+      "learning_rate": 3.3617855336368234e-05,
+      "loss": 0.378,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1853716821712041,
+      "grad_norm": 0.7060410855161342,
+      "learning_rate": 3.360204855842185e-05,
+      "loss": 0.3666,
+      "step": 6945
+    },
+    {
+      "epoch": 1.1862251429546813,
+      "grad_norm": 0.42389425542138154,
+      "learning_rate": 3.358624178047547e-05,
+      "loss": 0.3649,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1870786037381582,
+      "grad_norm": 0.36140348377650977,
+      "learning_rate": 3.357043500252909e-05,
+      "loss": 0.3897,
+      "step": 6955
+    },
+    {
+      "epoch": 1.1879320645216351,
+      "grad_norm": 0.565928565250766,
+      "learning_rate": 3.3554628224582704e-05,
+      "loss": 0.3707,
+      "step": 6960
+    },
+    {
+      "epoch": 1.1887855253051123,
+      "grad_norm": 0.458982871858116,
+      "learning_rate": 3.353882144663632e-05,
+      "loss": 0.4007,
+      "step": 6965
+    },
+    {
+      "epoch": 1.1896389860885892,
+      "grad_norm": 0.43602823762012966,
+      "learning_rate": 3.352301466868993e-05,
+      "loss": 0.3722,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1904924468720663,
+      "grad_norm": 0.5078272538835501,
+      "learning_rate": 3.350720789074355e-05,
+      "loss": 0.4005,
+      "step": 6975
+    },
+    {
+      "epoch": 1.1913459076555433,
+      "grad_norm": 0.47470167814658665,
+      "learning_rate": 3.3491401112797174e-05,
+      "loss": 0.3815,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1921993684390202,
+      "grad_norm": 0.496588669310637,
+      "learning_rate": 3.3475594334850785e-05,
+      "loss": 0.3485,
+      "step": 6985
+    },
+    {
+      "epoch": 1.1930528292224971,
+      "grad_norm": 0.4339679366968848,
+      "learning_rate": 3.34597875569044e-05,
+      "loss": 0.3881,
+      "step": 6990
+    },
+    {
+      "epoch": 1.1939062900059743,
+      "grad_norm": 0.8520920328281085,
+      "learning_rate": 3.344398077895802e-05,
+      "loss": 0.3791,
+      "step": 6995
+    },
+    {
+      "epoch": 1.1947597507894512,
+      "grad_norm": 0.39706879572408627,
+      "learning_rate": 3.342817400101164e-05,
+      "loss": 0.3711,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1956132115729283,
+      "grad_norm": 0.4629972024434518,
+      "learning_rate": 3.341236722306525e-05,
+      "loss": 0.3521,
+      "step": 7005
+    },
+    {
+      "epoch": 1.1964666723564052,
+      "grad_norm": 0.4278575426029068,
+      "learning_rate": 3.339656044511887e-05,
+      "loss": 0.4015,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1973201331398822,
+      "grad_norm": 1.5156192916688398,
+      "learning_rate": 3.338075366717249e-05,
+      "loss": 0.4231,
+      "step": 7015
+    },
+    {
+      "epoch": 1.1981735939233593,
+      "grad_norm": 0.5013914924256465,
+      "learning_rate": 3.33649468892261e-05,
+      "loss": 0.3706,
+      "step": 7020
+    },
+    {
+      "epoch": 1.1990270547068362,
+      "grad_norm": 0.7785649440633896,
+      "learning_rate": 3.334914011127972e-05,
+      "loss": 0.3604,
+      "step": 7025
+    },
+    {
+      "epoch": 1.1998805154903132,
+      "grad_norm": 0.408375287115701,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3798,
+      "step": 7030
+    },
+    {
+      "epoch": 1.2007339762737903,
+      "grad_norm": 0.6229062930638491,
+      "learning_rate": 3.331752655538695e-05,
+      "loss": 0.397,
+      "step": 7035
+    },
+    {
+      "epoch": 1.2015874370572672,
+      "grad_norm": 1.852308593695395,
+      "learning_rate": 3.330171977744056e-05,
+      "loss": 0.3776,
+      "step": 7040
+    },
+    {
+      "epoch": 1.2024408978407441,
+      "grad_norm": 0.41876048398721233,
+      "learning_rate": 3.328591299949419e-05,
+      "loss": 0.3771,
+      "step": 7045
+    },
+    {
+      "epoch": 1.2032943586242213,
+      "grad_norm": 0.43459573619031744,
+      "learning_rate": 3.3270106221547805e-05,
+      "loss": 0.3791,
+      "step": 7050
+    },
+    {
+      "epoch": 1.2041478194076982,
+      "grad_norm": 0.3696811553353602,
+      "learning_rate": 3.3254299443601415e-05,
+      "loss": 0.379,
+      "step": 7055
+    },
+    {
+      "epoch": 1.2050012801911751,
+      "grad_norm": 0.38109035981244466,
+      "learning_rate": 3.323849266565503e-05,
+      "loss": 0.3669,
+      "step": 7060
+    },
+    {
+      "epoch": 1.2058547409746523,
+      "grad_norm": 0.4723170910022811,
+      "learning_rate": 3.322268588770865e-05,
+      "loss": 0.3558,
+      "step": 7065
+    },
+    {
+      "epoch": 1.2067082017581292,
+      "grad_norm": 0.4897771860746794,
+      "learning_rate": 3.320687910976227e-05,
+      "loss": 0.3833,
+      "step": 7070
+    },
+    {
+      "epoch": 1.2075616625416061,
+      "grad_norm": 0.3298927449050978,
+      "learning_rate": 3.3191072331815885e-05,
+      "loss": 0.3643,
+      "step": 7075
+    },
+    {
+      "epoch": 1.2084151233250833,
+      "grad_norm": 0.4285871454607902,
+      "learning_rate": 3.31752655538695e-05,
+      "loss": 0.3762,
+      "step": 7080
+    },
+    {
+      "epoch": 1.2092685841085602,
+      "grad_norm": 0.6603932639317657,
+      "learning_rate": 3.315945877592312e-05,
+      "loss": 0.4047,
+      "step": 7085
+    },
+    {
+      "epoch": 1.2101220448920371,
+      "grad_norm": 0.4502007277418514,
+      "learning_rate": 3.314365199797673e-05,
+      "loss": 0.3846,
+      "step": 7090
+    },
+    {
+      "epoch": 1.2109755056755143,
+      "grad_norm": 0.4390536913234136,
+      "learning_rate": 3.312784522003035e-05,
+      "loss": 0.3868,
+      "step": 7095
+    },
+    {
+      "epoch": 1.2118289664589912,
+      "grad_norm": 1.4489440576914383,
+      "learning_rate": 3.3112038442083966e-05,
+      "loss": 0.3641,
+      "step": 7100
+    },
+    {
+      "epoch": 1.212682427242468,
+      "grad_norm": 0.809919753108125,
+      "learning_rate": 3.309623166413758e-05,
+      "loss": 0.4055,
+      "step": 7105
+    },
+    {
+      "epoch": 1.2135358880259453,
+      "grad_norm": 0.8926732603170776,
+      "learning_rate": 3.30804248861912e-05,
+      "loss": 0.386,
+      "step": 7110
+    },
+    {
+      "epoch": 1.2143893488094222,
+      "grad_norm": 0.665918402333069,
+      "learning_rate": 3.306461810824482e-05,
+      "loss": 0.3566,
+      "step": 7115
+    },
+    {
+      "epoch": 1.2152428095928993,
+      "grad_norm": 0.9914253719657289,
+      "learning_rate": 3.3048811330298436e-05,
+      "loss": 0.3832,
+      "step": 7120
+    },
+    {
+      "epoch": 1.2160962703763762,
+      "grad_norm": 0.44527666815736894,
+      "learning_rate": 3.3033004552352046e-05,
+      "loss": 0.3981,
+      "step": 7125
+    },
+    {
+      "epoch": 1.2169497311598532,
+      "grad_norm": 0.62674383923586,
+      "learning_rate": 3.3017197774405664e-05,
+      "loss": 0.3838,
+      "step": 7130
+    },
+    {
+      "epoch": 1.21780319194333,
+      "grad_norm": 0.5277348895958146,
+      "learning_rate": 3.300139099645929e-05,
+      "loss": 0.3705,
+      "step": 7135
+    },
+    {
+      "epoch": 1.2186566527268072,
+      "grad_norm": 0.35825319195406957,
+      "learning_rate": 3.29855842185129e-05,
+      "loss": 0.402,
+      "step": 7140
+    },
+    {
+      "epoch": 1.2195101135102842,
+      "grad_norm": 0.48097154642704415,
+      "learning_rate": 3.2969777440566516e-05,
+      "loss": 0.3589,
+      "step": 7145
+    },
+    {
+      "epoch": 1.2203635742937613,
+      "grad_norm": 0.4636097364299617,
+      "learning_rate": 3.2953970662620134e-05,
+      "loss": 0.3606,
+      "step": 7150
+    },
+    {
+      "epoch": 1.2212170350772382,
+      "grad_norm": 0.5011562959849015,
+      "learning_rate": 3.293816388467375e-05,
+      "loss": 0.3955,
+      "step": 7155
+    },
+    {
+      "epoch": 1.2220704958607151,
+      "grad_norm": 0.46403897845965236,
+      "learning_rate": 3.292235710672736e-05,
+      "loss": 0.3897,
+      "step": 7160
+    },
+    {
+      "epoch": 1.2229239566441923,
+      "grad_norm": 0.4948350216460918,
+      "learning_rate": 3.2906550328780986e-05,
+      "loss": 0.4061,
+      "step": 7165
+    },
+    {
+      "epoch": 1.2237774174276692,
+      "grad_norm": 0.3015142861226733,
+      "learning_rate": 3.2890743550834603e-05,
+      "loss": 0.3889,
+      "step": 7170
+    },
+    {
+      "epoch": 1.2246308782111461,
+      "grad_norm": 0.31718569415925846,
+      "learning_rate": 3.2874936772888214e-05,
+      "loss": 0.3646,
+      "step": 7175
+    },
+    {
+      "epoch": 1.2254843389946233,
+      "grad_norm": 0.505590576371758,
+      "learning_rate": 3.285912999494183e-05,
+      "loss": 0.3714,
+      "step": 7180
+    },
+    {
+      "epoch": 1.2263377997781002,
+      "grad_norm": 0.3377072659297116,
+      "learning_rate": 3.284332321699545e-05,
+      "loss": 0.366,
+      "step": 7185
+    },
+    {
+      "epoch": 1.2271912605615771,
+      "grad_norm": 0.34519374819683335,
+      "learning_rate": 3.2827516439049066e-05,
+      "loss": 0.3797,
+      "step": 7190
+    },
+    {
+      "epoch": 1.2280447213450543,
+      "grad_norm": 0.5307861769223624,
+      "learning_rate": 3.281170966110268e-05,
+      "loss": 0.3635,
+      "step": 7195
+    },
+    {
+      "epoch": 1.2288981821285312,
+      "grad_norm": 0.3803821448197012,
+      "learning_rate": 3.27959028831563e-05,
+      "loss": 0.3877,
+      "step": 7200
+    },
+    {
+      "epoch": 1.2297516429120081,
+      "grad_norm": 0.38378582601697386,
+      "learning_rate": 3.278009610520992e-05,
+      "loss": 0.3716,
+      "step": 7205
+    },
+    {
+      "epoch": 1.2306051036954853,
+      "grad_norm": 0.3722126409807913,
+      "learning_rate": 3.276428932726353e-05,
+      "loss": 0.3494,
+      "step": 7210
+    },
+    {
+      "epoch": 1.2314585644789622,
+      "grad_norm": 0.4701650733043367,
+      "learning_rate": 3.274848254931715e-05,
+      "loss": 0.3735,
+      "step": 7215
+    },
+    {
+      "epoch": 1.232312025262439,
+      "grad_norm": 0.360950752183332,
+      "learning_rate": 3.2732675771370764e-05,
+      "loss": 0.3798,
+      "step": 7220
+    },
+    {
+      "epoch": 1.2331654860459162,
+      "grad_norm": 0.38238358550781987,
+      "learning_rate": 3.271686899342438e-05,
+      "loss": 0.4007,
+      "step": 7225
+    },
+    {
+      "epoch": 1.2340189468293932,
+      "grad_norm": 0.3832560631518119,
+      "learning_rate": 3.2701062215478e-05,
+      "loss": 0.3862,
+      "step": 7230
+    },
+    {
+      "epoch": 1.23487240761287,
+      "grad_norm": 0.5471877311072296,
+      "learning_rate": 3.268525543753162e-05,
+      "loss": 0.3706,
+      "step": 7235
+    },
+    {
+      "epoch": 1.2357258683963472,
+      "grad_norm": 0.3265459517779948,
+      "learning_rate": 3.2669448659585234e-05,
+      "loss": 0.3964,
+      "step": 7240
+    },
+    {
+      "epoch": 1.2365793291798242,
+      "grad_norm": 0.43050611655984045,
+      "learning_rate": 3.2653641881638845e-05,
+      "loss": 0.3713,
+      "step": 7245
+    },
+    {
+      "epoch": 1.237432789963301,
+      "grad_norm": 0.3151548630949105,
+      "learning_rate": 3.263783510369246e-05,
+      "loss": 0.3584,
+      "step": 7250
+    },
+    {
+      "epoch": 1.2382862507467782,
+      "grad_norm": 0.4371362314804908,
+      "learning_rate": 3.262202832574608e-05,
+      "loss": 0.3903,
+      "step": 7255
+    },
+    {
+      "epoch": 1.2391397115302552,
+      "grad_norm": 0.358335905265931,
+      "learning_rate": 3.26062215477997e-05,
+      "loss": 0.3847,
+      "step": 7260
+    },
+    {
+      "epoch": 1.2399931723137323,
+      "grad_norm": 0.3764231398756448,
+      "learning_rate": 3.2590414769853315e-05,
+      "loss": 0.3907,
+      "step": 7265
+    },
+    {
+      "epoch": 1.2408466330972092,
+      "grad_norm": 0.3794729091152979,
+      "learning_rate": 3.257460799190693e-05,
+      "loss": 0.3689,
+      "step": 7270
+    },
+    {
+      "epoch": 1.2417000938806861,
+      "grad_norm": 0.3329388410145465,
+      "learning_rate": 3.255880121396055e-05,
+      "loss": 0.3619,
+      "step": 7275
+    },
+    {
+      "epoch": 1.242553554664163,
+      "grad_norm": 0.391398848493733,
+      "learning_rate": 3.254299443601416e-05,
+      "loss": 0.3711,
+      "step": 7280
+    },
+    {
+      "epoch": 1.2434070154476402,
+      "grad_norm": 0.4583419085661663,
+      "learning_rate": 3.2527187658067785e-05,
+      "loss": 0.3812,
+      "step": 7285
+    },
+    {
+      "epoch": 1.2442604762311171,
+      "grad_norm": 0.6730001261007689,
+      "learning_rate": 3.25113808801214e-05,
+      "loss": 0.384,
+      "step": 7290
+    },
+    {
+      "epoch": 1.2451139370145943,
+      "grad_norm": 0.40309795667742904,
+      "learning_rate": 3.249557410217501e-05,
+      "loss": 0.3574,
+      "step": 7295
+    },
+    {
+      "epoch": 1.2459673977980712,
+      "grad_norm": 0.3682695719098797,
+      "learning_rate": 3.247976732422863e-05,
+      "loss": 0.3632,
+      "step": 7300
+    },
+    {
+      "epoch": 1.2468208585815481,
+      "grad_norm": 0.42014217847259433,
+      "learning_rate": 3.246396054628225e-05,
+      "loss": 0.3517,
+      "step": 7305
+    },
+    {
+      "epoch": 1.2476743193650253,
+      "grad_norm": 0.40438319288989455,
+      "learning_rate": 3.2448153768335865e-05,
+      "loss": 0.3863,
+      "step": 7310
+    },
+    {
+      "epoch": 1.2485277801485022,
+      "grad_norm": 0.4019821499230842,
+      "learning_rate": 3.2432346990389476e-05,
+      "loss": 0.3457,
+      "step": 7315
+    },
+    {
+      "epoch": 1.249381240931979,
+      "grad_norm": 0.3613372896117948,
+      "learning_rate": 3.24165402124431e-05,
+      "loss": 0.3433,
+      "step": 7320
+    },
+    {
+      "epoch": 1.2502347017154563,
+      "grad_norm": 0.3026609174834825,
+      "learning_rate": 3.240073343449672e-05,
+      "loss": 0.3884,
+      "step": 7325
+    },
+    {
+      "epoch": 1.2510881624989332,
+      "grad_norm": 0.3138234605619824,
+      "learning_rate": 3.238492665655033e-05,
+      "loss": 0.3431,
+      "step": 7330
+    },
+    {
+      "epoch": 1.25194162328241,
+      "grad_norm": 0.30825661611686755,
+      "learning_rate": 3.2369119878603946e-05,
+      "loss": 0.3995,
+      "step": 7335
+    },
+    {
+      "epoch": 1.2527950840658872,
+      "grad_norm": 0.35993177196568976,
+      "learning_rate": 3.235331310065756e-05,
+      "loss": 0.3853,
+      "step": 7340
+    },
+    {
+      "epoch": 1.2536485448493642,
+      "grad_norm": 0.36522745740979284,
+      "learning_rate": 3.233750632271118e-05,
+      "loss": 0.3562,
+      "step": 7345
+    },
+    {
+      "epoch": 1.254502005632841,
+      "grad_norm": 0.45001752961029345,
+      "learning_rate": 3.232169954476479e-05,
+      "loss": 0.3431,
+      "step": 7350
+    },
+    {
+      "epoch": 1.2553554664163182,
+      "grad_norm": 0.3788022310918792,
+      "learning_rate": 3.2305892766818415e-05,
+      "loss": 0.354,
+      "step": 7355
+    },
+    {
+      "epoch": 1.2562089271997952,
+      "grad_norm": 0.36572687727000264,
+      "learning_rate": 3.229008598887203e-05,
+      "loss": 0.3784,
+      "step": 7360
+    },
+    {
+      "epoch": 1.257062387983272,
+      "grad_norm": 0.3401462065540811,
+      "learning_rate": 3.2274279210925644e-05,
+      "loss": 0.3618,
+      "step": 7365
+    },
+    {
+      "epoch": 1.2579158487667492,
+      "grad_norm": 0.3118211840094879,
+      "learning_rate": 3.225847243297926e-05,
+      "loss": 0.3703,
+      "step": 7370
+    },
+    {
+      "epoch": 1.2587693095502261,
+      "grad_norm": 0.44163497557223913,
+      "learning_rate": 3.224266565503288e-05,
+      "loss": 0.3798,
+      "step": 7375
+    },
+    {
+      "epoch": 1.2596227703337033,
+      "grad_norm": 0.48378884941491146,
+      "learning_rate": 3.2226858877086496e-05,
+      "loss": 0.3822,
+      "step": 7380
+    },
+    {
+      "epoch": 1.2604762311171802,
+      "grad_norm": 0.33225096808340443,
+      "learning_rate": 3.2211052099140113e-05,
+      "loss": 0.3956,
+      "step": 7385
+    },
+    {
+      "epoch": 1.2613296919006571,
+      "grad_norm": 0.39311029190076513,
+      "learning_rate": 3.219524532119373e-05,
+      "loss": 0.3842,
+      "step": 7390
+    },
+    {
+      "epoch": 1.262183152684134,
+      "grad_norm": 0.36608142885222406,
+      "learning_rate": 3.217943854324735e-05,
+      "loss": 0.4025,
+      "step": 7395
+    },
+    {
+      "epoch": 1.2630366134676112,
+      "grad_norm": 0.45751117064550084,
+      "learning_rate": 3.216363176530096e-05,
+      "loss": 0.3525,
+      "step": 7400
+    },
+    {
+      "epoch": 1.2638900742510881,
+      "grad_norm": 2.3313256094843053,
+      "learning_rate": 3.214782498735458e-05,
+      "loss": 0.3887,
+      "step": 7405
+    },
+    {
+      "epoch": 1.2647435350345653,
+      "grad_norm": 0.47541991430934355,
+      "learning_rate": 3.2132018209408194e-05,
+      "loss": 0.3965,
+      "step": 7410
+    },
+    {
+      "epoch": 1.2655969958180422,
+      "grad_norm": 0.5849483567906478,
+      "learning_rate": 3.211621143146181e-05,
+      "loss": 0.3697,
+      "step": 7415
+    },
+    {
+      "epoch": 1.2664504566015191,
+      "grad_norm": 0.5057974177473267,
+      "learning_rate": 3.210040465351543e-05,
+      "loss": 0.3925,
+      "step": 7420
+    },
+    {
+      "epoch": 1.267303917384996,
+      "grad_norm": 0.4481431566018363,
+      "learning_rate": 3.2084597875569046e-05,
+      "loss": 0.3666,
+      "step": 7425
+    },
+    {
+      "epoch": 1.2681573781684732,
+      "grad_norm": 0.376348907461285,
+      "learning_rate": 3.2068791097622664e-05,
+      "loss": 0.4056,
+      "step": 7430
+    },
+    {
+      "epoch": 1.26901083895195,
+      "grad_norm": 0.6364161503465311,
+      "learning_rate": 3.2052984319676274e-05,
+      "loss": 0.372,
+      "step": 7435
+    },
+    {
+      "epoch": 1.2698642997354272,
+      "grad_norm": 0.3804417117420967,
+      "learning_rate": 3.20371775417299e-05,
+      "loss": 0.3755,
+      "step": 7440
+    },
+    {
+      "epoch": 1.2707177605189042,
+      "grad_norm": 0.36191706587531747,
+      "learning_rate": 3.202137076378351e-05,
+      "loss": 0.3541,
+      "step": 7445
+    },
+    {
+      "epoch": 1.271571221302381,
+      "grad_norm": 0.5558036802892538,
+      "learning_rate": 3.200556398583713e-05,
+      "loss": 0.3661,
+      "step": 7450
+    },
+    {
+      "epoch": 1.272424682085858,
+      "grad_norm": 0.382324393894525,
+      "learning_rate": 3.1989757207890744e-05,
+      "loss": 0.3626,
+      "step": 7455
+    },
+    {
+      "epoch": 1.2732781428693352,
+      "grad_norm": 0.3563296482156405,
+      "learning_rate": 3.197395042994436e-05,
+      "loss": 0.3669,
+      "step": 7460
+    },
+    {
+      "epoch": 1.274131603652812,
+      "grad_norm": 0.4647270589907194,
+      "learning_rate": 3.195814365199798e-05,
+      "loss": 0.3901,
+      "step": 7465
+    },
+    {
+      "epoch": 1.2749850644362892,
+      "grad_norm": 0.44178128241828774,
+      "learning_rate": 3.194233687405159e-05,
+      "loss": 0.374,
+      "step": 7470
+    },
+    {
+      "epoch": 1.2758385252197662,
+      "grad_norm": 0.4185597275471627,
+      "learning_rate": 3.1926530096105214e-05,
+      "loss": 0.3818,
+      "step": 7475
+    },
+    {
+      "epoch": 1.276691986003243,
+      "grad_norm": 0.3937336884055784,
+      "learning_rate": 3.191072331815883e-05,
+      "loss": 0.383,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2775454467867202,
+      "grad_norm": 0.39788730239138864,
+      "learning_rate": 3.189491654021244e-05,
+      "loss": 0.4014,
+      "step": 7485
+    },
+    {
+      "epoch": 1.2783989075701971,
+      "grad_norm": 0.34102604629935046,
+      "learning_rate": 3.187910976226606e-05,
+      "loss": 0.3661,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2792523683536743,
+      "grad_norm": 0.32253182333469643,
+      "learning_rate": 3.186330298431968e-05,
+      "loss": 0.3918,
+      "step": 7495
+    },
+    {
+      "epoch": 1.2801058291371512,
+      "grad_norm": 0.5846100005497011,
+      "learning_rate": 3.1847496206373295e-05,
+      "loss": 0.3791,
+      "step": 7500
+    },
+    {
+      "epoch": 1.2809592899206281,
+      "grad_norm": 0.434655248897244,
+      "learning_rate": 3.1831689428426905e-05,
+      "loss": 0.367,
+      "step": 7505
+    },
+    {
+      "epoch": 1.281812750704105,
+      "grad_norm": 0.3518144063630111,
+      "learning_rate": 3.181588265048053e-05,
+      "loss": 0.3685,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2826662114875822,
+      "grad_norm": 0.44386299533464313,
+      "learning_rate": 3.180007587253415e-05,
+      "loss": 0.3712,
+      "step": 7515
+    },
+    {
+      "epoch": 1.2835196722710591,
+      "grad_norm": 0.5085192195133006,
+      "learning_rate": 3.178426909458776e-05,
+      "loss": 0.3672,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2843731330545363,
+      "grad_norm": 0.4561094288931108,
+      "learning_rate": 3.176846231664138e-05,
+      "loss": 0.3886,
+      "step": 7525
+    },
+    {
+      "epoch": 1.2852265938380132,
+      "grad_norm": 0.3477676921430173,
+      "learning_rate": 3.175265553869499e-05,
+      "loss": 0.3848,
+      "step": 7530
+    },
+    {
+      "epoch": 1.28608005462149,
+      "grad_norm": 0.5728733842271062,
+      "learning_rate": 3.173684876074861e-05,
+      "loss": 0.3666,
+      "step": 7535
+    },
+    {
+      "epoch": 1.286933515404967,
+      "grad_norm": 0.40073643354033833,
+      "learning_rate": 3.172104198280223e-05,
+      "loss": 0.3537,
+      "step": 7540
+    },
+    {
+      "epoch": 1.2877869761884442,
+      "grad_norm": 0.6490800328728087,
+      "learning_rate": 3.1705235204855845e-05,
+      "loss": 0.3622,
+      "step": 7545
+    },
+    {
+      "epoch": 1.288640436971921,
+      "grad_norm": 0.3555309507557241,
+      "learning_rate": 3.168942842690946e-05,
+      "loss": 0.4001,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2894938977553982,
+      "grad_norm": 0.41740412935467214,
+      "learning_rate": 3.167362164896307e-05,
+      "loss": 0.3628,
+      "step": 7555
+    },
+    {
+      "epoch": 1.2903473585388752,
+      "grad_norm": 0.39299843024898157,
+      "learning_rate": 3.16578148710167e-05,
+      "loss": 0.3696,
+      "step": 7560
+    },
+    {
+      "epoch": 1.291200819322352,
+      "grad_norm": 0.34495576011293283,
+      "learning_rate": 3.164200809307031e-05,
+      "loss": 0.3658,
+      "step": 7565
+    },
+    {
+      "epoch": 1.292054280105829,
+      "grad_norm": 0.3684441928525252,
+      "learning_rate": 3.1626201315123925e-05,
+      "loss": 0.3639,
+      "step": 7570
+    },
+    {
+      "epoch": 1.2929077408893062,
+      "grad_norm": 0.45395816645691744,
+      "learning_rate": 3.161039453717754e-05,
+      "loss": 0.3933,
+      "step": 7575
+    },
+    {
+      "epoch": 1.293761201672783,
+      "grad_norm": 0.38866887520272303,
+      "learning_rate": 3.159458775923116e-05,
+      "loss": 0.4058,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2946146624562602,
+      "grad_norm": 0.43712059420894067,
+      "learning_rate": 3.157878098128478e-05,
+      "loss": 0.3581,
+      "step": 7585
+    },
+    {
+      "epoch": 1.2954681232397371,
+      "grad_norm": 0.3760028727190519,
+      "learning_rate": 3.156297420333839e-05,
+      "loss": 0.3825,
+      "step": 7590
+    },
+    {
+      "epoch": 1.296321584023214,
+      "grad_norm": 0.3670961920825581,
+      "learning_rate": 3.154716742539201e-05,
+      "loss": 0.3859,
+      "step": 7595
+    },
+    {
+      "epoch": 1.2971750448066912,
+      "grad_norm": 0.2919355002339907,
+      "learning_rate": 3.1531360647445623e-05,
+      "loss": 0.3808,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2980285055901681,
+      "grad_norm": 0.4068715415938785,
+      "learning_rate": 3.151555386949924e-05,
+      "loss": 0.3913,
+      "step": 7605
+    },
+    {
+      "epoch": 1.298881966373645,
+      "grad_norm": 0.35915673766672923,
+      "learning_rate": 3.149974709155286e-05,
+      "loss": 0.3963,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2997354271571222,
+      "grad_norm": 0.3702895029464927,
+      "learning_rate": 3.1483940313606476e-05,
+      "loss": 0.3971,
+      "step": 7615
+    },
+    {
+      "epoch": 1.3005888879405991,
+      "grad_norm": 0.3265715421342069,
+      "learning_rate": 3.146813353566009e-05,
+      "loss": 0.3879,
+      "step": 7620
+    },
+    {
+      "epoch": 1.301442348724076,
+      "grad_norm": 0.43103727650242346,
+      "learning_rate": 3.1452326757713704e-05,
+      "loss": 0.3736,
+      "step": 7625
+    },
+    {
+      "epoch": 1.3022958095075532,
+      "grad_norm": 0.48845220319095556,
+      "learning_rate": 3.143651997976733e-05,
+      "loss": 0.4164,
+      "step": 7630
+    },
+    {
+      "epoch": 1.3031492702910301,
+      "grad_norm": 0.3409680039457544,
+      "learning_rate": 3.1420713201820946e-05,
+      "loss": 0.3614,
+      "step": 7635
+    },
+    {
+      "epoch": 1.3040027310745073,
+      "grad_norm": 0.5580659730454305,
+      "learning_rate": 3.1404906423874556e-05,
+      "loss": 0.3589,
+      "step": 7640
+    },
+    {
+      "epoch": 1.3048561918579842,
+      "grad_norm": 0.47995922826383103,
+      "learning_rate": 3.138909964592818e-05,
+      "loss": 0.3756,
+      "step": 7645
+    },
+    {
+      "epoch": 1.305709652641461,
+      "grad_norm": 0.41730841016773473,
+      "learning_rate": 3.137329286798179e-05,
+      "loss": 0.3621,
+      "step": 7650
+    },
+    {
+      "epoch": 1.306563113424938,
+      "grad_norm": 0.3934331300158891,
+      "learning_rate": 3.135748609003541e-05,
+      "loss": 0.3921,
+      "step": 7655
+    },
+    {
+      "epoch": 1.3074165742084152,
+      "grad_norm": 0.5423469201599359,
+      "learning_rate": 3.1341679312089026e-05,
+      "loss": 0.36,
+      "step": 7660
+    },
+    {
+      "epoch": 1.308270034991892,
+      "grad_norm": 0.3373234551343298,
+      "learning_rate": 3.1325872534142644e-05,
+      "loss": 0.3905,
+      "step": 7665
+    },
+    {
+      "epoch": 1.3091234957753692,
+      "grad_norm": 1.3857392438443161,
+      "learning_rate": 3.131006575619626e-05,
+      "loss": 0.3626,
+      "step": 7670
+    },
+    {
+      "epoch": 1.3099769565588462,
+      "grad_norm": 0.33705392183486976,
+      "learning_rate": 3.129425897824987e-05,
+      "loss": 0.3598,
+      "step": 7675
+    },
+    {
+      "epoch": 1.310830417342323,
+      "grad_norm": 0.48904999688941714,
+      "learning_rate": 3.1278452200303496e-05,
+      "loss": 0.3938,
+      "step": 7680
+    },
+    {
+      "epoch": 1.3116838781258,
+      "grad_norm": 0.3708751012077216,
+      "learning_rate": 3.126264542235711e-05,
+      "loss": 0.3829,
+      "step": 7685
+    },
+    {
+      "epoch": 1.3125373389092772,
+      "grad_norm": 0.3811349233254703,
+      "learning_rate": 3.1246838644410724e-05,
+      "loss": 0.4035,
+      "step": 7690
+    },
+    {
+      "epoch": 1.313390799692754,
+      "grad_norm": 0.45778200280902426,
+      "learning_rate": 3.123103186646434e-05,
+      "loss": 0.3635,
+      "step": 7695
+    },
+    {
+      "epoch": 1.3142442604762312,
+      "grad_norm": 0.44592689042058387,
+      "learning_rate": 3.121522508851796e-05,
+      "loss": 0.4066,
+      "step": 7700
+    },
+    {
+      "epoch": 1.3150977212597081,
+      "grad_norm": 0.3944483419896346,
+      "learning_rate": 3.1199418310571577e-05,
+      "loss": 0.3518,
+      "step": 7705
+    },
+    {
+      "epoch": 1.315951182043185,
+      "grad_norm": 0.38297744340106776,
+      "learning_rate": 3.118361153262519e-05,
+      "loss": 0.376,
+      "step": 7710
+    },
+    {
+      "epoch": 1.316804642826662,
+      "grad_norm": 0.6446635723826704,
+      "learning_rate": 3.116780475467881e-05,
+      "loss": 0.3775,
+      "step": 7715
+    },
+    {
+      "epoch": 1.3176581036101391,
+      "grad_norm": 0.428645446586942,
+      "learning_rate": 3.115199797673242e-05,
+      "loss": 0.3939,
+      "step": 7720
+    },
+    {
+      "epoch": 1.318511564393616,
+      "grad_norm": 0.32052118851991424,
+      "learning_rate": 3.113619119878604e-05,
+      "loss": 0.3718,
+      "step": 7725
+    },
+    {
+      "epoch": 1.3193650251770932,
+      "grad_norm": 0.4274457410022372,
+      "learning_rate": 3.112038442083966e-05,
+      "loss": 0.3769,
+      "step": 7730
+    },
+    {
+      "epoch": 1.3202184859605701,
+      "grad_norm": 0.4300172034421353,
+      "learning_rate": 3.1104577642893274e-05,
+      "loss": 0.4138,
+      "step": 7735
+    },
+    {
+      "epoch": 1.321071946744047,
+      "grad_norm": 0.39964267758451666,
+      "learning_rate": 3.108877086494689e-05,
+      "loss": 0.3845,
+      "step": 7740
+    },
+    {
+      "epoch": 1.3219254075275242,
+      "grad_norm": 0.38500495269783735,
+      "learning_rate": 3.10729640870005e-05,
+      "loss": 0.3873,
+      "step": 7745
+    },
+    {
+      "epoch": 1.322778868311001,
+      "grad_norm": 0.482078988228766,
+      "learning_rate": 3.105715730905413e-05,
+      "loss": 0.3831,
+      "step": 7750
+    },
+    {
+      "epoch": 1.323632329094478,
+      "grad_norm": 0.5148872754346172,
+      "learning_rate": 3.104135053110774e-05,
+      "loss": 0.3491,
+      "step": 7755
+    },
+    {
+      "epoch": 1.3244857898779552,
+      "grad_norm": 0.6946469363221514,
+      "learning_rate": 3.1025543753161355e-05,
+      "loss": 0.4165,
+      "step": 7760
+    },
+    {
+      "epoch": 1.325339250661432,
+      "grad_norm": 0.3418292401578764,
+      "learning_rate": 3.100973697521497e-05,
+      "loss": 0.3764,
+      "step": 7765
+    },
+    {
+      "epoch": 1.326192711444909,
+      "grad_norm": 0.32461219033431804,
+      "learning_rate": 3.099393019726859e-05,
+      "loss": 0.3401,
+      "step": 7770
+    },
+    {
+      "epoch": 1.3270461722283862,
+      "grad_norm": 0.3522364921006484,
+      "learning_rate": 3.097812341932221e-05,
+      "loss": 0.3625,
+      "step": 7775
+    },
+    {
+      "epoch": 1.327899633011863,
+      "grad_norm": 0.359786914606667,
+      "learning_rate": 3.0962316641375825e-05,
+      "loss": 0.3669,
+      "step": 7780
+    },
+    {
+      "epoch": 1.3287530937953402,
+      "grad_norm": 0.4588973269434624,
+      "learning_rate": 3.094650986342944e-05,
+      "loss": 0.3745,
+      "step": 7785
+    },
+    {
+      "epoch": 1.3296065545788172,
+      "grad_norm": 0.3743763419328188,
+      "learning_rate": 3.093070308548306e-05,
+      "loss": 0.382,
+      "step": 7790
+    },
+    {
+      "epoch": 1.330460015362294,
+      "grad_norm": 0.39118720849370636,
+      "learning_rate": 3.091489630753667e-05,
+      "loss": 0.3999,
+      "step": 7795
+    },
+    {
+      "epoch": 1.331313476145771,
+      "grad_norm": 0.35444759440072005,
+      "learning_rate": 3.0899089529590295e-05,
+      "loss": 0.3687,
+      "step": 7800
+    },
+    {
+      "epoch": 1.3321669369292481,
+      "grad_norm": 0.9234039532835139,
+      "learning_rate": 3.0883282751643905e-05,
+      "loss": 0.3862,
+      "step": 7805
+    },
+    {
+      "epoch": 1.333020397712725,
+      "grad_norm": 0.667172260558289,
+      "learning_rate": 3.086747597369752e-05,
+      "loss": 0.3987,
+      "step": 7810
+    },
+    {
+      "epoch": 1.3338738584962022,
+      "grad_norm": 0.3452488818103813,
+      "learning_rate": 3.085166919575114e-05,
+      "loss": 0.3747,
+      "step": 7815
+    },
+    {
+      "epoch": 1.3347273192796791,
+      "grad_norm": 0.3223161181571908,
+      "learning_rate": 3.083586241780476e-05,
+      "loss": 0.356,
+      "step": 7820
+    },
+    {
+      "epoch": 1.335580780063156,
+      "grad_norm": 0.3458890990408114,
+      "learning_rate": 3.0820055639858375e-05,
+      "loss": 0.3811,
+      "step": 7825
+    },
+    {
+      "epoch": 1.336434240846633,
+      "grad_norm": 0.3597519874961351,
+      "learning_rate": 3.0804248861911986e-05,
+      "loss": 0.3749,
+      "step": 7830
+    },
+    {
+      "epoch": 1.3372877016301101,
+      "grad_norm": 0.3524589306255129,
+      "learning_rate": 3.078844208396561e-05,
+      "loss": 0.3783,
+      "step": 7835
+    },
+    {
+      "epoch": 1.338141162413587,
+      "grad_norm": 0.32905689659120674,
+      "learning_rate": 3.077263530601922e-05,
+      "loss": 0.3988,
+      "step": 7840
+    },
+    {
+      "epoch": 1.3389946231970642,
+      "grad_norm": 0.4535575415728494,
+      "learning_rate": 3.075682852807284e-05,
+      "loss": 0.3842,
+      "step": 7845
+    },
+    {
+      "epoch": 1.3398480839805411,
+      "grad_norm": 0.387173113373346,
+      "learning_rate": 3.0741021750126456e-05,
+      "loss": 0.3858,
+      "step": 7850
+    },
+    {
+      "epoch": 1.340701544764018,
+      "grad_norm": 0.5401931885388208,
+      "learning_rate": 3.072521497218007e-05,
+      "loss": 0.3626,
+      "step": 7855
+    },
+    {
+      "epoch": 1.341555005547495,
+      "grad_norm": 0.3974982171224348,
+      "learning_rate": 3.070940819423369e-05,
+      "loss": 0.3898,
+      "step": 7860
+    },
+    {
+      "epoch": 1.342408466330972,
+      "grad_norm": 0.4010112756869794,
+      "learning_rate": 3.06936014162873e-05,
+      "loss": 0.3498,
+      "step": 7865
+    },
+    {
+      "epoch": 1.343261927114449,
+      "grad_norm": 0.398779730142885,
+      "learning_rate": 3.0677794638340926e-05,
+      "loss": 0.3668,
+      "step": 7870
+    },
+    {
+      "epoch": 1.3441153878979262,
+      "grad_norm": 0.3388769073868379,
+      "learning_rate": 3.0661987860394536e-05,
+      "loss": 0.3694,
+      "step": 7875
+    },
+    {
+      "epoch": 1.344968848681403,
+      "grad_norm": 0.44410138691153744,
+      "learning_rate": 3.0646181082448154e-05,
+      "loss": 0.3487,
+      "step": 7880
+    },
+    {
+      "epoch": 1.34582230946488,
+      "grad_norm": 0.44520525990871945,
+      "learning_rate": 3.063037430450177e-05,
+      "loss": 0.4044,
+      "step": 7885
+    },
+    {
+      "epoch": 1.3466757702483572,
+      "grad_norm": 0.40870853561648024,
+      "learning_rate": 3.061456752655539e-05,
+      "loss": 0.3996,
+      "step": 7890
+    },
+    {
+      "epoch": 1.347529231031834,
+      "grad_norm": 0.45739908444749894,
+      "learning_rate": 3.0598760748609006e-05,
+      "loss": 0.4049,
+      "step": 7895
+    },
+    {
+      "epoch": 1.348382691815311,
+      "grad_norm": 0.4344239547490103,
+      "learning_rate": 3.058295397066262e-05,
+      "loss": 0.3908,
+      "step": 7900
+    },
+    {
+      "epoch": 1.3492361525987882,
+      "grad_norm": 0.5573287877268787,
+      "learning_rate": 3.056714719271624e-05,
+      "loss": 0.3423,
+      "step": 7905
+    },
+    {
+      "epoch": 1.350089613382265,
+      "grad_norm": 0.3558345062805742,
+      "learning_rate": 3.055134041476985e-05,
+      "loss": 0.3697,
+      "step": 7910
+    },
+    {
+      "epoch": 1.350943074165742,
+      "grad_norm": 0.388868854432985,
+      "learning_rate": 3.053553363682347e-05,
+      "loss": 0.381,
+      "step": 7915
+    },
+    {
+      "epoch": 1.3517965349492191,
+      "grad_norm": 0.3249176781431105,
+      "learning_rate": 3.051972685887709e-05,
+      "loss": 0.3692,
+      "step": 7920
+    },
+    {
+      "epoch": 1.352649995732696,
+      "grad_norm": 0.4996246353360835,
+      "learning_rate": 3.0503920080930704e-05,
+      "loss": 0.3598,
+      "step": 7925
+    },
+    {
+      "epoch": 1.3535034565161732,
+      "grad_norm": 0.39933935553381533,
+      "learning_rate": 3.048811330298432e-05,
+      "loss": 0.3527,
+      "step": 7930
+    },
+    {
+      "epoch": 1.3543569172996501,
+      "grad_norm": 0.3681641698853502,
+      "learning_rate": 3.0472306525037936e-05,
+      "loss": 0.3881,
+      "step": 7935
+    },
+    {
+      "epoch": 1.355210378083127,
+      "grad_norm": 0.3653004963141978,
+      "learning_rate": 3.0456499747091556e-05,
+      "loss": 0.3753,
+      "step": 7940
+    },
+    {
+      "epoch": 1.356063838866604,
+      "grad_norm": 0.38532162730812364,
+      "learning_rate": 3.0440692969145174e-05,
+      "loss": 0.3747,
+      "step": 7945
+    },
+    {
+      "epoch": 1.3569172996500811,
+      "grad_norm": 0.41075651148330905,
+      "learning_rate": 3.0424886191198788e-05,
+      "loss": 0.3954,
+      "step": 7950
+    },
+    {
+      "epoch": 1.357770760433558,
+      "grad_norm": 0.36456529148274497,
+      "learning_rate": 3.0409079413252405e-05,
+      "loss": 0.3533,
+      "step": 7955
+    },
+    {
+      "epoch": 1.3586242212170352,
+      "grad_norm": 0.33276477226522555,
+      "learning_rate": 3.039327263530602e-05,
+      "loss": 0.3599,
+      "step": 7960
+    },
+    {
+      "epoch": 1.359477682000512,
+      "grad_norm": 0.32834117277130925,
+      "learning_rate": 3.0377465857359637e-05,
+      "loss": 0.3572,
+      "step": 7965
+    },
+    {
+      "epoch": 1.360331142783989,
+      "grad_norm": 0.40041837294621246,
+      "learning_rate": 3.036165907941325e-05,
+      "loss": 0.3721,
+      "step": 7970
+    },
+    {
+      "epoch": 1.361184603567466,
+      "grad_norm": 0.7082390310178429,
+      "learning_rate": 3.0345852301466872e-05,
+      "loss": 0.3756,
+      "step": 7975
+    },
+    {
+      "epoch": 1.362038064350943,
+      "grad_norm": 0.3581008506419958,
+      "learning_rate": 3.033004552352049e-05,
+      "loss": 0.3867,
+      "step": 7980
+    },
+    {
+      "epoch": 1.36289152513442,
+      "grad_norm": 0.4005690622739677,
+      "learning_rate": 3.0314238745574103e-05,
+      "loss": 0.3464,
+      "step": 7985
+    },
+    {
+      "epoch": 1.3637449859178972,
+      "grad_norm": 0.42662455749251676,
+      "learning_rate": 3.029843196762772e-05,
+      "loss": 0.3822,
+      "step": 7990
+    },
+    {
+      "epoch": 1.364598446701374,
+      "grad_norm": 0.36182851158669216,
+      "learning_rate": 3.0282625189681335e-05,
+      "loss": 0.3903,
+      "step": 7995
+    },
+    {
+      "epoch": 1.365451907484851,
+      "grad_norm": 0.4182698715308899,
+      "learning_rate": 3.0266818411734956e-05,
+      "loss": 0.3781,
+      "step": 8000
+    },
+    {
+      "epoch": 1.366305368268328,
+      "grad_norm": 0.38527955794425545,
+      "learning_rate": 3.0251011633788566e-05,
+      "loss": 0.3793,
+      "step": 8005
+    },
+    {
+      "epoch": 1.367158829051805,
+      "grad_norm": 0.41184995139370656,
+      "learning_rate": 3.0235204855842187e-05,
+      "loss": 0.3643,
+      "step": 8010
+    },
+    {
+      "epoch": 1.368012289835282,
+      "grad_norm": 0.37017177931689527,
+      "learning_rate": 3.0219398077895805e-05,
+      "loss": 0.3746,
+      "step": 8015
+    },
+    {
+      "epoch": 1.3688657506187591,
+      "grad_norm": 0.3517497261409417,
+      "learning_rate": 3.020359129994942e-05,
+      "loss": 0.4003,
+      "step": 8020
+    },
+    {
+      "epoch": 1.369719211402236,
+      "grad_norm": 0.36285052188375777,
+      "learning_rate": 3.0187784522003036e-05,
+      "loss": 0.3829,
+      "step": 8025
+    },
+    {
+      "epoch": 1.370572672185713,
+      "grad_norm": 0.407723005803761,
+      "learning_rate": 3.017197774405665e-05,
+      "loss": 0.3904,
+      "step": 8030
+    },
+    {
+      "epoch": 1.3714261329691901,
+      "grad_norm": 0.4660175917055238,
+      "learning_rate": 3.015617096611027e-05,
+      "loss": 0.3852,
+      "step": 8035
+    },
+    {
+      "epoch": 1.372279593752667,
+      "grad_norm": 0.35277520394249384,
+      "learning_rate": 3.014036418816389e-05,
+      "loss": 0.3805,
+      "step": 8040
+    },
+    {
+      "epoch": 1.373133054536144,
+      "grad_norm": 0.4244683993930311,
+      "learning_rate": 3.0124557410217503e-05,
+      "loss": 0.3724,
+      "step": 8045
+    },
+    {
+      "epoch": 1.3739865153196211,
+      "grad_norm": 0.534391863994486,
+      "learning_rate": 3.010875063227112e-05,
+      "loss": 0.3877,
+      "step": 8050
+    },
+    {
+      "epoch": 1.374839976103098,
+      "grad_norm": 0.3877045020895413,
+      "learning_rate": 3.0092943854324734e-05,
+      "loss": 0.3642,
+      "step": 8055
+    },
+    {
+      "epoch": 1.375693436886575,
+      "grad_norm": 0.3923341331422191,
+      "learning_rate": 3.007713707637835e-05,
+      "loss": 0.3816,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3765468976700521,
+      "grad_norm": 0.40198899955815753,
+      "learning_rate": 3.0061330298431966e-05,
+      "loss": 0.3757,
+      "step": 8065
+    },
+    {
+      "epoch": 1.377400358453529,
+      "grad_norm": 0.3077601297800156,
+      "learning_rate": 3.0045523520485587e-05,
+      "loss": 0.3779,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3782538192370062,
+      "grad_norm": 0.5506486589825703,
+      "learning_rate": 3.0029716742539204e-05,
+      "loss": 0.3771,
+      "step": 8075
+    },
+    {
+      "epoch": 1.379107280020483,
+      "grad_norm": 0.416658660112756,
+      "learning_rate": 3.0013909964592818e-05,
+      "loss": 0.3881,
+      "step": 8080
+    },
+    {
+      "epoch": 1.37996074080396,
+      "grad_norm": 0.3910076290601425,
+      "learning_rate": 2.9998103186646436e-05,
+      "loss": 0.4073,
+      "step": 8085
+    },
+    {
+      "epoch": 1.380814201587437,
+      "grad_norm": 0.3926307963488214,
+      "learning_rate": 2.998229640870005e-05,
+      "loss": 0.3822,
+      "step": 8090
+    },
+    {
+      "epoch": 1.381667662370914,
+      "grad_norm": 0.5148249677887993,
+      "learning_rate": 2.996648963075367e-05,
+      "loss": 0.3846,
+      "step": 8095
+    },
+    {
+      "epoch": 1.382521123154391,
+      "grad_norm": 0.4227190091549857,
+      "learning_rate": 2.995068285280728e-05,
+      "loss": 0.3882,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3833745839378682,
+      "grad_norm": 0.39891875888144984,
+      "learning_rate": 2.9934876074860902e-05,
+      "loss": 0.3739,
+      "step": 8105
+    },
+    {
+      "epoch": 1.384228044721345,
+      "grad_norm": 0.3650419019224378,
+      "learning_rate": 2.991906929691452e-05,
+      "loss": 0.3711,
+      "step": 8110
+    },
+    {
+      "epoch": 1.385081505504822,
+      "grad_norm": 0.42465523808031613,
+      "learning_rate": 2.9903262518968134e-05,
+      "loss": 0.3791,
+      "step": 8115
+    },
+    {
+      "epoch": 1.385934966288299,
+      "grad_norm": 0.2805002865152103,
+      "learning_rate": 2.988745574102175e-05,
+      "loss": 0.3663,
+      "step": 8120
+    },
+    {
+      "epoch": 1.386788427071776,
+      "grad_norm": 0.44417964430675694,
+      "learning_rate": 2.9871648963075365e-05,
+      "loss": 0.367,
+      "step": 8125
+    },
+    {
+      "epoch": 1.387641887855253,
+      "grad_norm": 0.38110904409638974,
+      "learning_rate": 2.9855842185128986e-05,
+      "loss": 0.3814,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3884953486387301,
+      "grad_norm": 0.4473363162850736,
+      "learning_rate": 2.9840035407182603e-05,
+      "loss": 0.3703,
+      "step": 8135
+    },
+    {
+      "epoch": 1.389348809422207,
+      "grad_norm": 0.370077716856123,
+      "learning_rate": 2.9824228629236217e-05,
+      "loss": 0.3822,
+      "step": 8140
+    },
+    {
+      "epoch": 1.390202270205684,
+      "grad_norm": 0.4409116111956241,
+      "learning_rate": 2.9808421851289835e-05,
+      "loss": 0.3613,
+      "step": 8145
+    },
+    {
+      "epoch": 1.391055730989161,
+      "grad_norm": 0.45142254858558284,
+      "learning_rate": 2.979261507334345e-05,
+      "loss": 0.3893,
+      "step": 8150
+    },
+    {
+      "epoch": 1.391909191772638,
+      "grad_norm": 0.3649482987051066,
+      "learning_rate": 2.977680829539707e-05,
+      "loss": 0.3629,
+      "step": 8155
+    },
+    {
+      "epoch": 1.392762652556115,
+      "grad_norm": 0.38984275503151034,
+      "learning_rate": 2.976100151745068e-05,
+      "loss": 0.3921,
+      "step": 8160
+    },
+    {
+      "epoch": 1.3936161133395921,
+      "grad_norm": 0.35334520856945323,
+      "learning_rate": 2.97451947395043e-05,
+      "loss": 0.3642,
+      "step": 8165
+    },
+    {
+      "epoch": 1.394469574123069,
+      "grad_norm": 0.39985450271284095,
+      "learning_rate": 2.972938796155792e-05,
+      "loss": 0.3732,
+      "step": 8170
+    },
+    {
+      "epoch": 1.395323034906546,
+      "grad_norm": 0.35479709027549755,
+      "learning_rate": 2.9713581183611533e-05,
+      "loss": 0.3711,
+      "step": 8175
+    },
+    {
+      "epoch": 1.3961764956900231,
+      "grad_norm": 0.29189648002820756,
+      "learning_rate": 2.969777440566515e-05,
+      "loss": 0.3884,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3970299564735,
+      "grad_norm": 0.46385100833457055,
+      "learning_rate": 2.9681967627718764e-05,
+      "loss": 0.376,
+      "step": 8185
+    },
+    {
+      "epoch": 1.3978834172569772,
+      "grad_norm": 0.4131075562028953,
+      "learning_rate": 2.9666160849772385e-05,
+      "loss": 0.3801,
+      "step": 8190
+    },
+    {
+      "epoch": 1.398736878040454,
+      "grad_norm": 0.33214764010056536,
+      "learning_rate": 2.9650354071826003e-05,
+      "loss": 0.3681,
+      "step": 8195
+    },
+    {
+      "epoch": 1.399590338823931,
+      "grad_norm": 0.53525220637501,
+      "learning_rate": 2.9634547293879617e-05,
+      "loss": 0.3984,
+      "step": 8200
+    },
+    {
+      "epoch": 1.400443799607408,
+      "grad_norm": 0.3525049517732988,
+      "learning_rate": 2.9618740515933234e-05,
+      "loss": 0.3993,
+      "step": 8205
+    },
+    {
+      "epoch": 1.401297260390885,
+      "grad_norm": 0.2862840539021341,
+      "learning_rate": 2.9602933737986848e-05,
+      "loss": 0.3694,
+      "step": 8210
+    },
+    {
+      "epoch": 1.402150721174362,
+      "grad_norm": 0.4106370473387779,
+      "learning_rate": 2.958712696004047e-05,
+      "loss": 0.3542,
+      "step": 8215
+    },
+    {
+      "epoch": 1.4030041819578392,
+      "grad_norm": 0.3709496464873719,
+      "learning_rate": 2.957132018209408e-05,
+      "loss": 0.3848,
+      "step": 8220
+    },
+    {
+      "epoch": 1.403857642741316,
+      "grad_norm": 0.4866566318354896,
+      "learning_rate": 2.95555134041477e-05,
+      "loss": 0.386,
+      "step": 8225
+    },
+    {
+      "epoch": 1.404711103524793,
+      "grad_norm": 0.3574489462236936,
+      "learning_rate": 2.9539706626201318e-05,
+      "loss": 0.3569,
+      "step": 8230
+    },
+    {
+      "epoch": 1.40556456430827,
+      "grad_norm": 0.5124624087288366,
+      "learning_rate": 2.9523899848254932e-05,
+      "loss": 0.387,
+      "step": 8235
+    },
+    {
+      "epoch": 1.406418025091747,
+      "grad_norm": 0.45569556196234207,
+      "learning_rate": 2.950809307030855e-05,
+      "loss": 0.4009,
+      "step": 8240
+    },
+    {
+      "epoch": 1.407271485875224,
+      "grad_norm": 0.4171523942953767,
+      "learning_rate": 2.9492286292362164e-05,
+      "loss": 0.3642,
+      "step": 8245
+    },
+    {
+      "epoch": 1.4081249466587011,
+      "grad_norm": 0.29889924634374965,
+      "learning_rate": 2.9476479514415785e-05,
+      "loss": 0.3834,
+      "step": 8250
+    },
+    {
+      "epoch": 1.408978407442178,
+      "grad_norm": 0.3360512543164952,
+      "learning_rate": 2.9460672736469395e-05,
+      "loss": 0.3748,
+      "step": 8255
+    },
+    {
+      "epoch": 1.409831868225655,
+      "grad_norm": 0.38582001935622595,
+      "learning_rate": 2.9444865958523016e-05,
+      "loss": 0.3472,
+      "step": 8260
+    },
+    {
+      "epoch": 1.410685329009132,
+      "grad_norm": 0.3130260111344319,
+      "learning_rate": 2.9429059180576634e-05,
+      "loss": 0.3665,
+      "step": 8265
+    },
+    {
+      "epoch": 1.411538789792609,
+      "grad_norm": 0.3624333359753043,
+      "learning_rate": 2.9413252402630248e-05,
+      "loss": 0.3813,
+      "step": 8270
+    },
+    {
+      "epoch": 1.412392250576086,
+      "grad_norm": 0.3637937324332378,
+      "learning_rate": 2.939744562468387e-05,
+      "loss": 0.3851,
+      "step": 8275
+    },
+    {
+      "epoch": 1.4132457113595631,
+      "grad_norm": 0.3912603378405003,
+      "learning_rate": 2.938163884673748e-05,
+      "loss": 0.3752,
+      "step": 8280
+    },
+    {
+      "epoch": 1.41409917214304,
+      "grad_norm": 0.4295669156261878,
+      "learning_rate": 2.93658320687911e-05,
+      "loss": 0.3699,
+      "step": 8285
+    },
+    {
+      "epoch": 1.414952632926517,
+      "grad_norm": 0.3834328296660925,
+      "learning_rate": 2.9350025290844717e-05,
+      "loss": 0.3644,
+      "step": 8290
+    },
+    {
+      "epoch": 1.415806093709994,
+      "grad_norm": 0.3746515593442747,
+      "learning_rate": 2.933421851289833e-05,
+      "loss": 0.365,
+      "step": 8295
+    },
+    {
+      "epoch": 1.416659554493471,
+      "grad_norm": 0.7066203168458007,
+      "learning_rate": 2.931841173495195e-05,
+      "loss": 0.3813,
+      "step": 8300
+    },
+    {
+      "epoch": 1.417513015276948,
+      "grad_norm": 0.34315167367633687,
+      "learning_rate": 2.9302604957005563e-05,
+      "loss": 0.3794,
+      "step": 8305
+    },
+    {
+      "epoch": 1.418366476060425,
+      "grad_norm": 0.38712370941731256,
+      "learning_rate": 2.9286798179059184e-05,
+      "loss": 0.3894,
+      "step": 8310
+    },
+    {
+      "epoch": 1.419219936843902,
+      "grad_norm": 0.3673660256432574,
+      "learning_rate": 2.9270991401112795e-05,
+      "loss": 0.3686,
+      "step": 8315
+    },
+    {
+      "epoch": 1.420073397627379,
+      "grad_norm": 0.3677150794575681,
+      "learning_rate": 2.9255184623166415e-05,
+      "loss": 0.413,
+      "step": 8320
+    },
+    {
+      "epoch": 1.420926858410856,
+      "grad_norm": 0.42763016123464537,
+      "learning_rate": 2.9239377845220033e-05,
+      "loss": 0.35,
+      "step": 8325
+    },
+    {
+      "epoch": 1.421780319194333,
+      "grad_norm": 0.3446902671135802,
+      "learning_rate": 2.9223571067273647e-05,
+      "loss": 0.37,
+      "step": 8330
+    },
+    {
+      "epoch": 1.4226337799778102,
+      "grad_norm": 0.34060123260542946,
+      "learning_rate": 2.9207764289327268e-05,
+      "loss": 0.386,
+      "step": 8335
+    },
+    {
+      "epoch": 1.423487240761287,
+      "grad_norm": 0.45806486515155437,
+      "learning_rate": 2.919195751138088e-05,
+      "loss": 0.384,
+      "step": 8340
+    },
+    {
+      "epoch": 1.424340701544764,
+      "grad_norm": 0.3739214262633945,
+      "learning_rate": 2.91761507334345e-05,
+      "loss": 0.3522,
+      "step": 8345
+    },
+    {
+      "epoch": 1.425194162328241,
+      "grad_norm": 0.40813115432879005,
+      "learning_rate": 2.9160343955488113e-05,
+      "loss": 0.3904,
+      "step": 8350
+    },
+    {
+      "epoch": 1.426047623111718,
+      "grad_norm": 0.3647379706539541,
+      "learning_rate": 2.914453717754173e-05,
+      "loss": 0.359,
+      "step": 8355
+    },
+    {
+      "epoch": 1.426901083895195,
+      "grad_norm": 0.3112262907787899,
+      "learning_rate": 2.9128730399595348e-05,
+      "loss": 0.3527,
+      "step": 8360
+    },
+    {
+      "epoch": 1.4277545446786721,
+      "grad_norm": 0.380474743802569,
+      "learning_rate": 2.9112923621648962e-05,
+      "loss": 0.375,
+      "step": 8365
+    },
+    {
+      "epoch": 1.428608005462149,
+      "grad_norm": 0.40453440065650126,
+      "learning_rate": 2.9097116843702583e-05,
+      "loss": 0.3797,
+      "step": 8370
+    },
+    {
+      "epoch": 1.429461466245626,
+      "grad_norm": 0.29419053059544575,
+      "learning_rate": 2.9081310065756194e-05,
+      "loss": 0.3748,
+      "step": 8375
+    },
+    {
+      "epoch": 1.430314927029103,
+      "grad_norm": 0.29630117830668,
+      "learning_rate": 2.9065503287809815e-05,
+      "loss": 0.3729,
+      "step": 8380
+    },
+    {
+      "epoch": 1.43116838781258,
+      "grad_norm": 0.47399249555344675,
+      "learning_rate": 2.9049696509863432e-05,
+      "loss": 0.3898,
+      "step": 8385
+    },
+    {
+      "epoch": 1.432021848596057,
+      "grad_norm": 0.4194551512756209,
+      "learning_rate": 2.9033889731917046e-05,
+      "loss": 0.346,
+      "step": 8390
+    },
+    {
+      "epoch": 1.4328753093795341,
+      "grad_norm": 0.3505631403613706,
+      "learning_rate": 2.9018082953970667e-05,
+      "loss": 0.3649,
+      "step": 8395
+    },
+    {
+      "epoch": 1.433728770163011,
+      "grad_norm": 0.3338652798914307,
+      "learning_rate": 2.9002276176024278e-05,
+      "loss": 0.378,
+      "step": 8400
+    },
+    {
+      "epoch": 1.434582230946488,
+      "grad_norm": 0.4133051714483541,
+      "learning_rate": 2.89864693980779e-05,
+      "loss": 0.3764,
+      "step": 8405
+    },
+    {
+      "epoch": 1.4354356917299649,
+      "grad_norm": 0.3221303004684508,
+      "learning_rate": 2.8970662620131513e-05,
+      "loss": 0.3594,
+      "step": 8410
+    },
+    {
+      "epoch": 1.436289152513442,
+      "grad_norm": 0.5510652482425935,
+      "learning_rate": 2.895485584218513e-05,
+      "loss": 0.3817,
+      "step": 8415
+    },
+    {
+      "epoch": 1.437142613296919,
+      "grad_norm": 0.3615619462568365,
+      "learning_rate": 2.8939049064238748e-05,
+      "loss": 0.384,
+      "step": 8420
+    },
+    {
+      "epoch": 1.437996074080396,
+      "grad_norm": 0.4800111548347184,
+      "learning_rate": 2.892324228629236e-05,
+      "loss": 0.3847,
+      "step": 8425
+    },
+    {
+      "epoch": 1.438849534863873,
+      "grad_norm": 0.34094030765100325,
+      "learning_rate": 2.8907435508345983e-05,
+      "loss": 0.3643,
+      "step": 8430
+    },
+    {
+      "epoch": 1.43970299564735,
+      "grad_norm": 0.33032663807125756,
+      "learning_rate": 2.8891628730399593e-05,
+      "loss": 0.3856,
+      "step": 8435
+    },
+    {
+      "epoch": 1.440556456430827,
+      "grad_norm": 0.3093031003852718,
+      "learning_rate": 2.8875821952453214e-05,
+      "loss": 0.3618,
+      "step": 8440
+    },
+    {
+      "epoch": 1.441409917214304,
+      "grad_norm": 0.39872434410355373,
+      "learning_rate": 2.886001517450683e-05,
+      "loss": 0.378,
+      "step": 8445
+    },
+    {
+      "epoch": 1.442263377997781,
+      "grad_norm": 0.6054554551753893,
+      "learning_rate": 2.8844208396560446e-05,
+      "loss": 0.3892,
+      "step": 8450
+    },
+    {
+      "epoch": 1.443116838781258,
+      "grad_norm": 0.39556706642501605,
+      "learning_rate": 2.8828401618614066e-05,
+      "loss": 0.3792,
+      "step": 8455
+    },
+    {
+      "epoch": 1.443970299564735,
+      "grad_norm": 0.3272752075394602,
+      "learning_rate": 2.8812594840667677e-05,
+      "loss": 0.3737,
+      "step": 8460
+    },
+    {
+      "epoch": 1.444823760348212,
+      "grad_norm": 0.37048934763363117,
+      "learning_rate": 2.8796788062721298e-05,
+      "loss": 0.3649,
+      "step": 8465
+    },
+    {
+      "epoch": 1.445677221131689,
+      "grad_norm": 0.4596849364173726,
+      "learning_rate": 2.8780981284774912e-05,
+      "loss": 0.3632,
+      "step": 8470
+    },
+    {
+      "epoch": 1.446530681915166,
+      "grad_norm": 2.293224386936095,
+      "learning_rate": 2.876517450682853e-05,
+      "loss": 0.3635,
+      "step": 8475
+    },
+    {
+      "epoch": 1.4473841426986431,
+      "grad_norm": 0.33651129082014747,
+      "learning_rate": 2.8749367728882147e-05,
+      "loss": 0.381,
+      "step": 8480
+    },
+    {
+      "epoch": 1.44823760348212,
+      "grad_norm": 0.3456310747207851,
+      "learning_rate": 2.873356095093576e-05,
+      "loss": 0.3634,
+      "step": 8485
+    },
+    {
+      "epoch": 1.449091064265597,
+      "grad_norm": 0.3995683842923382,
+      "learning_rate": 2.8717754172989382e-05,
+      "loss": 0.3782,
+      "step": 8490
+    },
+    {
+      "epoch": 1.449944525049074,
+      "grad_norm": 0.3458136741834777,
+      "learning_rate": 2.8701947395042993e-05,
+      "loss": 0.3617,
+      "step": 8495
+    },
+    {
+      "epoch": 1.450797985832551,
+      "grad_norm": 0.43859808049241406,
+      "learning_rate": 2.8686140617096613e-05,
+      "loss": 0.3676,
+      "step": 8500
+    },
+    {
+      "epoch": 1.451651446616028,
+      "grad_norm": 0.34478162797919504,
+      "learning_rate": 2.8670333839150227e-05,
+      "loss": 0.3695,
+      "step": 8505
+    },
+    {
+      "epoch": 1.452504907399505,
+      "grad_norm": 0.3697671344377753,
+      "learning_rate": 2.8654527061203845e-05,
+      "loss": 0.3793,
+      "step": 8510
+    },
+    {
+      "epoch": 1.453358368182982,
+      "grad_norm": 0.3823214798575854,
+      "learning_rate": 2.8638720283257466e-05,
+      "loss": 0.3784,
+      "step": 8515
+    },
+    {
+      "epoch": 1.454211828966459,
+      "grad_norm": 0.38784507192907286,
+      "learning_rate": 2.8622913505311076e-05,
+      "loss": 0.3837,
+      "step": 8520
+    },
+    {
+      "epoch": 1.4550652897499359,
+      "grad_norm": 0.41557813313362896,
+      "learning_rate": 2.8607106727364697e-05,
+      "loss": 0.3378,
+      "step": 8525
+    },
+    {
+      "epoch": 1.455918750533413,
+      "grad_norm": 0.3275493993112729,
+      "learning_rate": 2.859129994941831e-05,
+      "loss": 0.3715,
+      "step": 8530
+    },
+    {
+      "epoch": 1.45677221131689,
+      "grad_norm": 0.6038453812193175,
+      "learning_rate": 2.857549317147193e-05,
+      "loss": 0.3962,
+      "step": 8535
+    },
+    {
+      "epoch": 1.457625672100367,
+      "grad_norm": 0.531888329219478,
+      "learning_rate": 2.8559686393525546e-05,
+      "loss": 0.3635,
+      "step": 8540
+    },
+    {
+      "epoch": 1.458479132883844,
+      "grad_norm": 0.4508160296461119,
+      "learning_rate": 2.854387961557916e-05,
+      "loss": 0.3694,
+      "step": 8545
+    },
+    {
+      "epoch": 1.459332593667321,
+      "grad_norm": 0.3785246712895831,
+      "learning_rate": 2.852807283763278e-05,
+      "loss": 0.3818,
+      "step": 8550
+    },
+    {
+      "epoch": 1.4601860544507979,
+      "grad_norm": 0.36557857481825257,
+      "learning_rate": 2.8512266059686392e-05,
+      "loss": 0.3697,
+      "step": 8555
+    },
+    {
+      "epoch": 1.461039515234275,
+      "grad_norm": 0.3965680377021494,
+      "learning_rate": 2.8496459281740013e-05,
+      "loss": 0.4041,
+      "step": 8560
+    },
+    {
+      "epoch": 1.461892976017752,
+      "grad_norm": 0.33612947628386897,
+      "learning_rate": 2.8480652503793627e-05,
+      "loss": 0.3515,
+      "step": 8565
+    },
+    {
+      "epoch": 1.462746436801229,
+      "grad_norm": 0.3365847961971949,
+      "learning_rate": 2.8464845725847244e-05,
+      "loss": 0.346,
+      "step": 8570
+    },
+    {
+      "epoch": 1.463599897584706,
+      "grad_norm": 0.3670891194765111,
+      "learning_rate": 2.8449038947900865e-05,
+      "loss": 0.3412,
+      "step": 8575
+    },
+    {
+      "epoch": 1.464453358368183,
+      "grad_norm": 0.357725309927502,
+      "learning_rate": 2.8433232169954476e-05,
+      "loss": 0.3561,
+      "step": 8580
+    },
+    {
+      "epoch": 1.46530681915166,
+      "grad_norm": 0.6195894783397705,
+      "learning_rate": 2.8417425392008097e-05,
+      "loss": 0.3661,
+      "step": 8585
+    },
+    {
+      "epoch": 1.466160279935137,
+      "grad_norm": 0.3755371747074425,
+      "learning_rate": 2.840161861406171e-05,
+      "loss": 0.4006,
+      "step": 8590
+    },
+    {
+      "epoch": 1.467013740718614,
+      "grad_norm": 0.44702736073352417,
+      "learning_rate": 2.8385811836115328e-05,
+      "loss": 0.4065,
+      "step": 8595
+    },
+    {
+      "epoch": 1.467867201502091,
+      "grad_norm": 0.34275483033089893,
+      "learning_rate": 2.8370005058168946e-05,
+      "loss": 0.3735,
+      "step": 8600
+    },
+    {
+      "epoch": 1.468720662285568,
+      "grad_norm": 0.4273574089186848,
+      "learning_rate": 2.835419828022256e-05,
+      "loss": 0.3963,
+      "step": 8605
+    },
+    {
+      "epoch": 1.469574123069045,
+      "grad_norm": 0.40888070045662017,
+      "learning_rate": 2.833839150227618e-05,
+      "loss": 0.3936,
+      "step": 8610
+    },
+    {
+      "epoch": 1.470427583852522,
+      "grad_norm": 0.34082565712865803,
+      "learning_rate": 2.832258472432979e-05,
+      "loss": 0.3645,
+      "step": 8615
+    },
+    {
+      "epoch": 1.471281044635999,
+      "grad_norm": 0.48554931370305,
+      "learning_rate": 2.8306777946383412e-05,
+      "loss": 0.3707,
+      "step": 8620
+    },
+    {
+      "epoch": 1.472134505419476,
+      "grad_norm": 0.3271541302725915,
+      "learning_rate": 2.8290971168437026e-05,
+      "loss": 0.3519,
+      "step": 8625
+    },
+    {
+      "epoch": 1.472987966202953,
+      "grad_norm": 0.3903889081674045,
+      "learning_rate": 2.8275164390490644e-05,
+      "loss": 0.3632,
+      "step": 8630
+    },
+    {
+      "epoch": 1.47384142698643,
+      "grad_norm": 0.4284030291366842,
+      "learning_rate": 2.8259357612544264e-05,
+      "loss": 0.3633,
+      "step": 8635
+    },
+    {
+      "epoch": 1.4746948877699069,
+      "grad_norm": 0.3216128934656074,
+      "learning_rate": 2.8243550834597875e-05,
+      "loss": 0.3686,
+      "step": 8640
+    },
+    {
+      "epoch": 1.475548348553384,
+      "grad_norm": 0.3895542158384611,
+      "learning_rate": 2.8227744056651496e-05,
+      "loss": 0.4004,
+      "step": 8645
+    },
+    {
+      "epoch": 1.476401809336861,
+      "grad_norm": 0.38894558511697247,
+      "learning_rate": 2.821193727870511e-05,
+      "loss": 0.3757,
+      "step": 8650
+    },
+    {
+      "epoch": 1.477255270120338,
+      "grad_norm": 0.454030090547461,
+      "learning_rate": 2.8196130500758727e-05,
+      "loss": 0.3672,
+      "step": 8655
+    },
+    {
+      "epoch": 1.478108730903815,
+      "grad_norm": 0.3678890776042101,
+      "learning_rate": 2.818032372281234e-05,
+      "loss": 0.3656,
+      "step": 8660
+    },
+    {
+      "epoch": 1.478962191687292,
+      "grad_norm": 0.3679379479229768,
+      "learning_rate": 2.816451694486596e-05,
+      "loss": 0.3611,
+      "step": 8665
+    },
+    {
+      "epoch": 1.4798156524707688,
+      "grad_norm": 0.36577399619786805,
+      "learning_rate": 2.814871016691958e-05,
+      "loss": 0.372,
+      "step": 8670
+    },
+    {
+      "epoch": 1.480669113254246,
+      "grad_norm": 0.41365093673266295,
+      "learning_rate": 2.813290338897319e-05,
+      "loss": 0.3916,
+      "step": 8675
+    },
+    {
+      "epoch": 1.481522574037723,
+      "grad_norm": 0.5075863875682871,
+      "learning_rate": 2.811709661102681e-05,
+      "loss": 0.3795,
+      "step": 8680
+    },
+    {
+      "epoch": 1.4823760348212,
+      "grad_norm": 0.4153661164502053,
+      "learning_rate": 2.8101289833080425e-05,
+      "loss": 0.4021,
+      "step": 8685
+    },
+    {
+      "epoch": 1.483229495604677,
+      "grad_norm": 0.3878147450114199,
+      "learning_rate": 2.8085483055134043e-05,
+      "loss": 0.3908,
+      "step": 8690
+    },
+    {
+      "epoch": 1.484082956388154,
+      "grad_norm": 0.41708422810946044,
+      "learning_rate": 2.806967627718766e-05,
+      "loss": 0.3482,
+      "step": 8695
+    },
+    {
+      "epoch": 1.4849364171716308,
+      "grad_norm": 0.39439490995397697,
+      "learning_rate": 2.8053869499241274e-05,
+      "loss": 0.3647,
+      "step": 8700
+    },
+    {
+      "epoch": 1.485789877955108,
+      "grad_norm": 0.5110644577060773,
+      "learning_rate": 2.8038062721294895e-05,
+      "loss": 0.3442,
+      "step": 8705
+    },
+    {
+      "epoch": 1.486643338738585,
+      "grad_norm": 0.307577556175372,
+      "learning_rate": 2.802225594334851e-05,
+      "loss": 0.3564,
+      "step": 8710
+    },
+    {
+      "epoch": 1.487496799522062,
+      "grad_norm": 0.4158390196218934,
+      "learning_rate": 2.8006449165402127e-05,
+      "loss": 0.3971,
+      "step": 8715
+    },
+    {
+      "epoch": 1.488350260305539,
+      "grad_norm": 0.39969357057334653,
+      "learning_rate": 2.799064238745574e-05,
+      "loss": 0.3935,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4892037210890159,
+      "grad_norm": 0.3722754066149683,
+      "learning_rate": 2.7974835609509358e-05,
+      "loss": 0.3716,
+      "step": 8725
+    },
+    {
+      "epoch": 1.490057181872493,
+      "grad_norm": 0.33067377058218317,
+      "learning_rate": 2.795902883156298e-05,
+      "loss": 0.3746,
+      "step": 8730
+    },
+    {
+      "epoch": 1.49091064265597,
+      "grad_norm": 0.405252723193417,
+      "learning_rate": 2.794322205361659e-05,
+      "loss": 0.356,
+      "step": 8735
+    },
+    {
+      "epoch": 1.4917641034394469,
+      "grad_norm": 0.3694165764691589,
+      "learning_rate": 2.792741527567021e-05,
+      "loss": 0.366,
+      "step": 8740
+    },
+    {
+      "epoch": 1.492617564222924,
+      "grad_norm": 0.426916135457793,
+      "learning_rate": 2.7911608497723825e-05,
+      "loss": 0.3824,
+      "step": 8745
+    },
+    {
+      "epoch": 1.493471025006401,
+      "grad_norm": 0.754366531660231,
+      "learning_rate": 2.7895801719777442e-05,
+      "loss": 0.4195,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4943244857898779,
+      "grad_norm": 0.5434925483779819,
+      "learning_rate": 2.7879994941831056e-05,
+      "loss": 0.3823,
+      "step": 8755
+    },
+    {
+      "epoch": 1.495177946573355,
+      "grad_norm": 0.29984507565571433,
+      "learning_rate": 2.7864188163884674e-05,
+      "loss": 0.3804,
+      "step": 8760
+    },
+    {
+      "epoch": 1.496031407356832,
+      "grad_norm": 0.32650876732692063,
+      "learning_rate": 2.7848381385938295e-05,
+      "loss": 0.3849,
+      "step": 8765
+    },
+    {
+      "epoch": 1.496884868140309,
+      "grad_norm": 0.3331507148161785,
+      "learning_rate": 2.783257460799191e-05,
+      "loss": 0.3668,
+      "step": 8770
+    },
+    {
+      "epoch": 1.497738328923786,
+      "grad_norm": 0.37201454830785047,
+      "learning_rate": 2.7816767830045526e-05,
+      "loss": 0.3724,
+      "step": 8775
+    },
+    {
+      "epoch": 1.498591789707263,
+      "grad_norm": 0.3883608809915117,
+      "learning_rate": 2.780096105209914e-05,
+      "loss": 0.3742,
+      "step": 8780
+    },
+    {
+      "epoch": 1.4994452504907398,
+      "grad_norm": 0.4341517061568229,
+      "learning_rate": 2.7785154274152758e-05,
+      "loss": 0.3796,
+      "step": 8785
+    },
+    {
+      "epoch": 1.500298711274217,
+      "grad_norm": 0.43802481543794175,
+      "learning_rate": 2.776934749620638e-05,
+      "loss": 0.3863,
+      "step": 8790
+    },
+    {
+      "epoch": 1.501152172057694,
+      "grad_norm": 0.422125554109543,
+      "learning_rate": 2.775354071825999e-05,
+      "loss": 0.3795,
+      "step": 8795
+    },
+    {
+      "epoch": 1.502005632841171,
+      "grad_norm": 0.5053590727369254,
+      "learning_rate": 2.773773394031361e-05,
+      "loss": 0.3596,
+      "step": 8800
+    },
+    {
+      "epoch": 1.502859093624648,
+      "grad_norm": 2.003664078861723,
+      "learning_rate": 2.7721927162367224e-05,
+      "loss": 0.3823,
+      "step": 8805
+    },
+    {
+      "epoch": 1.503712554408125,
+      "grad_norm": 0.503061980341984,
+      "learning_rate": 2.770612038442084e-05,
+      "loss": 0.3435,
+      "step": 8810
+    },
+    {
+      "epoch": 1.5045660151916018,
+      "grad_norm": 0.3839031362370193,
+      "learning_rate": 2.7690313606474456e-05,
+      "loss": 0.3787,
+      "step": 8815
+    },
+    {
+      "epoch": 1.505419475975079,
+      "grad_norm": 0.6262488963362381,
+      "learning_rate": 2.7674506828528073e-05,
+      "loss": 0.3775,
+      "step": 8820
+    },
+    {
+      "epoch": 1.506272936758556,
+      "grad_norm": 0.3837957990180745,
+      "learning_rate": 2.7658700050581694e-05,
+      "loss": 0.3558,
+      "step": 8825
+    },
+    {
+      "epoch": 1.507126397542033,
+      "grad_norm": 0.5708127621802429,
+      "learning_rate": 2.7642893272635305e-05,
+      "loss": 0.3633,
+      "step": 8830
+    },
+    {
+      "epoch": 1.50797985832551,
+      "grad_norm": 0.37810240026882774,
+      "learning_rate": 2.7627086494688925e-05,
+      "loss": 0.3696,
+      "step": 8835
+    },
+    {
+      "epoch": 1.5088333191089869,
+      "grad_norm": 0.4379948264672438,
+      "learning_rate": 2.761127971674254e-05,
+      "loss": 0.3966,
+      "step": 8840
+    },
+    {
+      "epoch": 1.5096867798924638,
+      "grad_norm": 0.4864435508523675,
+      "learning_rate": 2.7595472938796157e-05,
+      "loss": 0.3626,
+      "step": 8845
+    },
+    {
+      "epoch": 1.510540240675941,
+      "grad_norm": 0.37753863697846646,
+      "learning_rate": 2.7579666160849778e-05,
+      "loss": 0.3773,
+      "step": 8850
+    },
+    {
+      "epoch": 1.511393701459418,
+      "grad_norm": 0.42430613680275475,
+      "learning_rate": 2.756385938290339e-05,
+      "loss": 0.3775,
+      "step": 8855
+    },
+    {
+      "epoch": 1.512247162242895,
+      "grad_norm": 0.544003009322885,
+      "learning_rate": 2.754805260495701e-05,
+      "loss": 0.3859,
+      "step": 8860
+    },
+    {
+      "epoch": 1.513100623026372,
+      "grad_norm": 0.4074885598009095,
+      "learning_rate": 2.7532245827010623e-05,
+      "loss": 0.3484,
+      "step": 8865
+    },
+    {
+      "epoch": 1.5139540838098489,
+      "grad_norm": 0.3988185137624454,
+      "learning_rate": 2.751643904906424e-05,
+      "loss": 0.377,
+      "step": 8870
+    },
+    {
+      "epoch": 1.5148075445933258,
+      "grad_norm": 0.5744208283221645,
+      "learning_rate": 2.7500632271117855e-05,
+      "loss": 0.3988,
+      "step": 8875
+    },
+    {
+      "epoch": 1.515661005376803,
+      "grad_norm": 0.4850703376232935,
+      "learning_rate": 2.7484825493171472e-05,
+      "loss": 0.3533,
+      "step": 8880
+    },
+    {
+      "epoch": 1.51651446616028,
+      "grad_norm": 0.48494000229877243,
+      "learning_rate": 2.7469018715225093e-05,
+      "loss": 0.3627,
+      "step": 8885
+    },
+    {
+      "epoch": 1.517367926943757,
+      "grad_norm": 0.3332980410205891,
+      "learning_rate": 2.7453211937278704e-05,
+      "loss": 0.3879,
+      "step": 8890
+    },
+    {
+      "epoch": 1.518221387727234,
+      "grad_norm": 0.3294632163479773,
+      "learning_rate": 2.7437405159332325e-05,
+      "loss": 0.3924,
+      "step": 8895
+    },
+    {
+      "epoch": 1.5190748485107108,
+      "grad_norm": 0.338661944064282,
+      "learning_rate": 2.742159838138594e-05,
+      "loss": 0.3641,
+      "step": 8900
+    },
+    {
+      "epoch": 1.519928309294188,
+      "grad_norm": 0.3636548212436217,
+      "learning_rate": 2.7405791603439556e-05,
+      "loss": 0.3822,
+      "step": 8905
+    },
+    {
+      "epoch": 1.520781770077665,
+      "grad_norm": 0.463133124848011,
+      "learning_rate": 2.738998482549317e-05,
+      "loss": 0.3697,
+      "step": 8910
+    },
+    {
+      "epoch": 1.521635230861142,
+      "grad_norm": 0.42984954316114005,
+      "learning_rate": 2.7374178047546788e-05,
+      "loss": 0.3764,
+      "step": 8915
+    },
+    {
+      "epoch": 1.522488691644619,
+      "grad_norm": 0.3745966593564767,
+      "learning_rate": 2.735837126960041e-05,
+      "loss": 0.3734,
+      "step": 8920
+    },
+    {
+      "epoch": 1.523342152428096,
+      "grad_norm": 0.38169440992259,
+      "learning_rate": 2.7342564491654023e-05,
+      "loss": 0.3773,
+      "step": 8925
+    },
+    {
+      "epoch": 1.5241956132115728,
+      "grad_norm": 0.6318648268520636,
+      "learning_rate": 2.732675771370764e-05,
+      "loss": 0.408,
+      "step": 8930
+    },
+    {
+      "epoch": 1.52504907399505,
+      "grad_norm": 0.4723539663658168,
+      "learning_rate": 2.7310950935761254e-05,
+      "loss": 0.3749,
+      "step": 8935
+    },
+    {
+      "epoch": 1.5259025347785269,
+      "grad_norm": 0.34942791277198937,
+      "learning_rate": 2.7295144157814872e-05,
+      "loss": 0.3689,
+      "step": 8940
+    },
+    {
+      "epoch": 1.526755995562004,
+      "grad_norm": 0.41284349215649063,
+      "learning_rate": 2.7279337379868493e-05,
+      "loss": 0.3786,
+      "step": 8945
+    },
+    {
+      "epoch": 1.527609456345481,
+      "grad_norm": 0.3161233364871583,
+      "learning_rate": 2.7263530601922103e-05,
+      "loss": 0.3678,
+      "step": 8950
+    },
+    {
+      "epoch": 1.5284629171289579,
+      "grad_norm": 0.39487784144550014,
+      "learning_rate": 2.7247723823975724e-05,
+      "loss": 0.3562,
+      "step": 8955
+    },
+    {
+      "epoch": 1.5293163779124348,
+      "grad_norm": 0.3068547919022294,
+      "learning_rate": 2.7231917046029338e-05,
+      "loss": 0.3689,
+      "step": 8960
+    },
+    {
+      "epoch": 1.530169838695912,
+      "grad_norm": 10.742316504226467,
+      "learning_rate": 2.7216110268082956e-05,
+      "loss": 0.3945,
+      "step": 8965
+    },
+    {
+      "epoch": 1.531023299479389,
+      "grad_norm": 0.655424979342714,
+      "learning_rate": 2.720030349013657e-05,
+      "loss": 0.3879,
+      "step": 8970
+    },
+    {
+      "epoch": 1.531876760262866,
+      "grad_norm": 0.40648843586294353,
+      "learning_rate": 2.7184496712190187e-05,
+      "loss": 0.3979,
+      "step": 8975
+    },
+    {
+      "epoch": 1.532730221046343,
+      "grad_norm": 0.4803385811688802,
+      "learning_rate": 2.7168689934243808e-05,
+      "loss": 0.3637,
+      "step": 8980
+    },
+    {
+      "epoch": 1.5335836818298199,
+      "grad_norm": 0.4245263455311067,
+      "learning_rate": 2.7152883156297422e-05,
+      "loss": 0.3618,
+      "step": 8985
+    },
+    {
+      "epoch": 1.5344371426132968,
+      "grad_norm": 0.39932903050533153,
+      "learning_rate": 2.713707637835104e-05,
+      "loss": 0.3761,
+      "step": 8990
+    },
+    {
+      "epoch": 1.535290603396774,
+      "grad_norm": 0.44406625007850303,
+      "learning_rate": 2.7121269600404654e-05,
+      "loss": 0.3915,
+      "step": 8995
+    },
+    {
+      "epoch": 1.536144064180251,
+      "grad_norm": 0.32849238846478673,
+      "learning_rate": 2.710546282245827e-05,
+      "loss": 0.3668,
+      "step": 9000
+    },
+    {
+      "epoch": 1.536997524963728,
+      "grad_norm": 0.3925200210031384,
+      "learning_rate": 2.7089656044511892e-05,
+      "loss": 0.3905,
+      "step": 9005
+    },
+    {
+      "epoch": 1.537850985747205,
+      "grad_norm": 0.7415926818661323,
+      "learning_rate": 2.7073849266565503e-05,
+      "loss": 0.3888,
+      "step": 9010
+    },
+    {
+      "epoch": 1.5387044465306818,
+      "grad_norm": 0.3707764983758511,
+      "learning_rate": 2.7058042488619123e-05,
+      "loss": 0.3605,
+      "step": 9015
+    },
+    {
+      "epoch": 1.5395579073141588,
+      "grad_norm": 0.3933176309338654,
+      "learning_rate": 2.7042235710672737e-05,
+      "loss": 0.3583,
+      "step": 9020
+    },
+    {
+      "epoch": 1.540411368097636,
+      "grad_norm": 0.3297557693413401,
+      "learning_rate": 2.7026428932726355e-05,
+      "loss": 0.3479,
+      "step": 9025
+    },
+    {
+      "epoch": 1.541264828881113,
+      "grad_norm": 0.4398309704400533,
+      "learning_rate": 2.701062215477997e-05,
+      "loss": 0.4116,
+      "step": 9030
+    },
+    {
+      "epoch": 1.54211828966459,
+      "grad_norm": 0.42724442868004253,
+      "learning_rate": 2.6994815376833586e-05,
+      "loss": 0.3671,
+      "step": 9035
+    },
+    {
+      "epoch": 1.542971750448067,
+      "grad_norm": 0.3989452060612025,
+      "learning_rate": 2.6979008598887207e-05,
+      "loss": 0.3718,
+      "step": 9040
+    },
+    {
+      "epoch": 1.5438252112315438,
+      "grad_norm": 0.37428979509096105,
+      "learning_rate": 2.696320182094082e-05,
+      "loss": 0.3786,
+      "step": 9045
+    },
+    {
+      "epoch": 1.544678672015021,
+      "grad_norm": 0.5076240642768487,
+      "learning_rate": 2.694739504299444e-05,
+      "loss": 0.377,
+      "step": 9050
+    },
+    {
+      "epoch": 1.5455321327984979,
+      "grad_norm": 0.7285376829530452,
+      "learning_rate": 2.6931588265048053e-05,
+      "loss": 0.3886,
+      "step": 9055
+    },
+    {
+      "epoch": 1.546385593581975,
+      "grad_norm": 0.5204765982447884,
+      "learning_rate": 2.691578148710167e-05,
+      "loss": 0.4023,
+      "step": 9060
+    },
+    {
+      "epoch": 1.547239054365452,
+      "grad_norm": 0.3569364007758203,
+      "learning_rate": 2.6899974709155284e-05,
+      "loss": 0.3562,
+      "step": 9065
+    },
+    {
+      "epoch": 1.5480925151489289,
+      "grad_norm": 0.3811351720837915,
+      "learning_rate": 2.6884167931208902e-05,
+      "loss": 0.3824,
+      "step": 9070
+    },
+    {
+      "epoch": 1.5489459759324058,
+      "grad_norm": 0.32962360850539596,
+      "learning_rate": 2.6868361153262523e-05,
+      "loss": 0.3924,
+      "step": 9075
+    },
+    {
+      "epoch": 1.549799436715883,
+      "grad_norm": 0.432679451843628,
+      "learning_rate": 2.6852554375316137e-05,
+      "loss": 0.3895,
+      "step": 9080
+    },
+    {
+      "epoch": 1.5506528974993599,
+      "grad_norm": 0.4945031259578414,
+      "learning_rate": 2.6836747597369754e-05,
+      "loss": 0.3868,
+      "step": 9085
+    },
+    {
+      "epoch": 1.551506358282837,
+      "grad_norm": 0.3701671834144567,
+      "learning_rate": 2.682094081942337e-05,
+      "loss": 0.3808,
+      "step": 9090
+    },
+    {
+      "epoch": 1.552359819066314,
+      "grad_norm": 0.3636808660556998,
+      "learning_rate": 2.6805134041476986e-05,
+      "loss": 0.3966,
+      "step": 9095
+    },
+    {
+      "epoch": 1.5532132798497909,
+      "grad_norm": 0.36849988756897306,
+      "learning_rate": 2.6789327263530607e-05,
+      "loss": 0.3855,
+      "step": 9100
+    },
+    {
+      "epoch": 1.5540667406332678,
+      "grad_norm": 0.5038956078553172,
+      "learning_rate": 2.677352048558422e-05,
+      "loss": 0.3869,
+      "step": 9105
+    },
+    {
+      "epoch": 1.554920201416745,
+      "grad_norm": 0.34468219300425484,
+      "learning_rate": 2.6757713707637838e-05,
+      "loss": 0.3686,
+      "step": 9110
+    },
+    {
+      "epoch": 1.555773662200222,
+      "grad_norm": 0.401718519770233,
+      "learning_rate": 2.6741906929691452e-05,
+      "loss": 0.3738,
+      "step": 9115
+    },
+    {
+      "epoch": 1.556627122983699,
+      "grad_norm": 0.4171816020041063,
+      "learning_rate": 2.672610015174507e-05,
+      "loss": 0.366,
+      "step": 9120
+    },
+    {
+      "epoch": 1.557480583767176,
+      "grad_norm": 0.39786329975089246,
+      "learning_rate": 2.6710293373798684e-05,
+      "loss": 0.3761,
+      "step": 9125
+    },
+    {
+      "epoch": 1.5583340445506528,
+      "grad_norm": 0.7040786337578375,
+      "learning_rate": 2.66944865958523e-05,
+      "loss": 0.3979,
+      "step": 9130
+    },
+    {
+      "epoch": 1.5591875053341298,
+      "grad_norm": 0.557664904636061,
+      "learning_rate": 2.6678679817905922e-05,
+      "loss": 0.3655,
+      "step": 9135
+    },
+    {
+      "epoch": 1.560040966117607,
+      "grad_norm": 0.3497624468193592,
+      "learning_rate": 2.6662873039959536e-05,
+      "loss": 0.3738,
+      "step": 9140
+    },
+    {
+      "epoch": 1.560894426901084,
+      "grad_norm": 0.40404210994387985,
+      "learning_rate": 2.6647066262013154e-05,
+      "loss": 0.3607,
+      "step": 9145
+    },
+    {
+      "epoch": 1.561747887684561,
+      "grad_norm": 0.36077575739029377,
+      "learning_rate": 2.6631259484066768e-05,
+      "loss": 0.371,
+      "step": 9150
+    },
+    {
+      "epoch": 1.5626013484680379,
+      "grad_norm": 0.4729302920624471,
+      "learning_rate": 2.6615452706120385e-05,
+      "loss": 0.3864,
+      "step": 9155
+    },
+    {
+      "epoch": 1.5634548092515148,
+      "grad_norm": 0.2874203120762668,
+      "learning_rate": 2.6599645928174e-05,
+      "loss": 0.3856,
+      "step": 9160
+    },
+    {
+      "epoch": 1.5643082700349917,
+      "grad_norm": 0.36858966565331214,
+      "learning_rate": 2.658383915022762e-05,
+      "loss": 0.3872,
+      "step": 9165
+    },
+    {
+      "epoch": 1.5651617308184689,
+      "grad_norm": 0.38098337930660564,
+      "learning_rate": 2.6568032372281237e-05,
+      "loss": 0.3847,
+      "step": 9170
+    },
+    {
+      "epoch": 1.566015191601946,
+      "grad_norm": 0.800233524254945,
+      "learning_rate": 2.655222559433485e-05,
+      "loss": 0.3685,
+      "step": 9175
+    },
+    {
+      "epoch": 1.566868652385423,
+      "grad_norm": 0.48202523987743545,
+      "learning_rate": 2.653641881638847e-05,
+      "loss": 0.3699,
+      "step": 9180
+    },
+    {
+      "epoch": 1.5677221131688999,
+      "grad_norm": 0.5688858346090507,
+      "learning_rate": 2.6520612038442083e-05,
+      "loss": 0.379,
+      "step": 9185
+    },
+    {
+      "epoch": 1.5685755739523768,
+      "grad_norm": 0.3767776776816685,
+      "learning_rate": 2.65048052604957e-05,
+      "loss": 0.3842,
+      "step": 9190
+    },
+    {
+      "epoch": 1.569429034735854,
+      "grad_norm": 0.3629471220319574,
+      "learning_rate": 2.648899848254932e-05,
+      "loss": 0.3647,
+      "step": 9195
+    },
+    {
+      "epoch": 1.5702824955193309,
+      "grad_norm": 0.3949093492225141,
+      "learning_rate": 2.6473191704602935e-05,
+      "loss": 0.3515,
+      "step": 9200
+    },
+    {
+      "epoch": 1.571135956302808,
+      "grad_norm": 0.43793661800851835,
+      "learning_rate": 2.6457384926656553e-05,
+      "loss": 0.3881,
+      "step": 9205
+    },
+    {
+      "epoch": 1.571989417086285,
+      "grad_norm": 0.3727516948465313,
+      "learning_rate": 2.6441578148710167e-05,
+      "loss": 0.3831,
+      "step": 9210
+    },
+    {
+      "epoch": 1.5728428778697618,
+      "grad_norm": 0.37882602636586166,
+      "learning_rate": 2.6425771370763784e-05,
+      "loss": 0.3851,
+      "step": 9215
+    },
+    {
+      "epoch": 1.5736963386532388,
+      "grad_norm": 0.33033538606628826,
+      "learning_rate": 2.64099645928174e-05,
+      "loss": 0.3765,
+      "step": 9220
+    },
+    {
+      "epoch": 1.574549799436716,
+      "grad_norm": 0.3784344567080915,
+      "learning_rate": 2.639415781487102e-05,
+      "loss": 0.39,
+      "step": 9225
+    },
+    {
+      "epoch": 1.5754032602201928,
+      "grad_norm": 0.35185440416602554,
+      "learning_rate": 2.6378351036924637e-05,
+      "loss": 0.3646,
+      "step": 9230
+    },
+    {
+      "epoch": 1.57625672100367,
+      "grad_norm": 0.3219427366807726,
+      "learning_rate": 2.636254425897825e-05,
+      "loss": 0.376,
+      "step": 9235
+    },
+    {
+      "epoch": 1.577110181787147,
+      "grad_norm": 0.35824132884480425,
+      "learning_rate": 2.634673748103187e-05,
+      "loss": 0.3897,
+      "step": 9240
+    },
+    {
+      "epoch": 1.5779636425706238,
+      "grad_norm": 0.34487482225992905,
+      "learning_rate": 2.6330930703085482e-05,
+      "loss": 0.3953,
+      "step": 9245
+    },
+    {
+      "epoch": 1.5788171033541007,
+      "grad_norm": 0.362644522036318,
+      "learning_rate": 2.63151239251391e-05,
+      "loss": 0.3762,
+      "step": 9250
+    },
+    {
+      "epoch": 1.579670564137578,
+      "grad_norm": 0.3780291497619649,
+      "learning_rate": 2.629931714719272e-05,
+      "loss": 0.376,
+      "step": 9255
+    },
+    {
+      "epoch": 1.580524024921055,
+      "grad_norm": 0.40088734596731784,
+      "learning_rate": 2.6283510369246335e-05,
+      "loss": 0.3823,
+      "step": 9260
+    },
+    {
+      "epoch": 1.581377485704532,
+      "grad_norm": 0.36782407616549584,
+      "learning_rate": 2.6267703591299952e-05,
+      "loss": 0.3907,
+      "step": 9265
+    },
+    {
+      "epoch": 1.5822309464880089,
+      "grad_norm": 0.5086138052713359,
+      "learning_rate": 2.6251896813353566e-05,
+      "loss": 0.3942,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5830844072714858,
+      "grad_norm": 0.36939975099414485,
+      "learning_rate": 2.6236090035407184e-05,
+      "loss": 0.3888,
+      "step": 9275
+    },
+    {
+      "epoch": 1.5839378680549627,
+      "grad_norm": 0.3539713587920976,
+      "learning_rate": 2.6220283257460798e-05,
+      "loss": 0.3494,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5847913288384399,
+      "grad_norm": 0.35250448460289285,
+      "learning_rate": 2.620447647951442e-05,
+      "loss": 0.3757,
+      "step": 9285
+    },
+    {
+      "epoch": 1.585644789621917,
+      "grad_norm": 0.5006557278450434,
+      "learning_rate": 2.6188669701568036e-05,
+      "loss": 0.37,
+      "step": 9290
+    },
+    {
+      "epoch": 1.586498250405394,
+      "grad_norm": 0.29962215222651417,
+      "learning_rate": 2.617286292362165e-05,
+      "loss": 0.3692,
+      "step": 9295
+    },
+    {
+      "epoch": 1.5873517111888709,
+      "grad_norm": 0.31878668347807404,
+      "learning_rate": 2.6157056145675268e-05,
+      "loss": 0.3486,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5882051719723478,
+      "grad_norm": 0.290472656585368,
+      "learning_rate": 2.6141249367728882e-05,
+      "loss": 0.3823,
+      "step": 9305
+    },
+    {
+      "epoch": 1.5890586327558247,
+      "grad_norm": 0.34575943425850814,
+      "learning_rate": 2.61254425897825e-05,
+      "loss": 0.3585,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5899120935393019,
+      "grad_norm": 0.6272976523179152,
+      "learning_rate": 2.6109635811836113e-05,
+      "loss": 0.3926,
+      "step": 9315
+    },
+    {
+      "epoch": 1.590765554322779,
+      "grad_norm": 0.43107217052336677,
+      "learning_rate": 2.6093829033889734e-05,
+      "loss": 0.3863,
+      "step": 9320
+    },
+    {
+      "epoch": 1.591619015106256,
+      "grad_norm": 0.43085347816235414,
+      "learning_rate": 2.607802225594335e-05,
+      "loss": 0.3793,
+      "step": 9325
+    },
+    {
+      "epoch": 1.5924724758897328,
+      "grad_norm": 0.34121218921916524,
+      "learning_rate": 2.6062215477996966e-05,
+      "loss": 0.3767,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5933259366732098,
+      "grad_norm": 0.434036624709166,
+      "learning_rate": 2.6046408700050583e-05,
+      "loss": 0.3608,
+      "step": 9335
+    },
+    {
+      "epoch": 1.594179397456687,
+      "grad_norm": 0.4432817785731385,
+      "learning_rate": 2.6030601922104197e-05,
+      "loss": 0.3775,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5950328582401638,
+      "grad_norm": 0.45732496141433143,
+      "learning_rate": 2.6014795144157818e-05,
+      "loss": 0.3737,
+      "step": 9345
+    },
+    {
+      "epoch": 1.595886319023641,
+      "grad_norm": 0.3275847167701077,
+      "learning_rate": 2.5998988366211435e-05,
+      "loss": 0.3585,
+      "step": 9350
+    },
+    {
+      "epoch": 1.596739779807118,
+      "grad_norm": 0.4462161939109244,
+      "learning_rate": 2.598318158826505e-05,
+      "loss": 0.3812,
+      "step": 9355
+    },
+    {
+      "epoch": 1.5975932405905948,
+      "grad_norm": 0.36722959104251734,
+      "learning_rate": 2.5967374810318667e-05,
+      "loss": 0.361,
+      "step": 9360
+    },
+    {
+      "epoch": 1.5984467013740717,
+      "grad_norm": 0.3708369371991551,
+      "learning_rate": 2.595156803237228e-05,
+      "loss": 0.3524,
+      "step": 9365
+    },
+    {
+      "epoch": 1.5993001621575489,
+      "grad_norm": 0.38301506596702944,
+      "learning_rate": 2.59357612544259e-05,
+      "loss": 0.3771,
+      "step": 9370
+    },
+    {
+      "epoch": 1.6001536229410258,
+      "grad_norm": 0.3112861622370914,
+      "learning_rate": 2.5919954476479513e-05,
+      "loss": 0.3828,
+      "step": 9375
+    },
+    {
+      "epoch": 1.601007083724503,
+      "grad_norm": 0.5593603068779465,
+      "learning_rate": 2.5904147698533133e-05,
+      "loss": 0.3909,
+      "step": 9380
+    },
+    {
+      "epoch": 1.6018605445079799,
+      "grad_norm": 0.324086537459784,
+      "learning_rate": 2.588834092058675e-05,
+      "loss": 0.3806,
+      "step": 9385
+    },
+    {
+      "epoch": 1.6027140052914568,
+      "grad_norm": 0.3909912111430699,
+      "learning_rate": 2.5872534142640365e-05,
+      "loss": 0.3779,
+      "step": 9390
+    },
+    {
+      "epoch": 1.6035674660749337,
+      "grad_norm": 0.3699949482868292,
+      "learning_rate": 2.5856727364693982e-05,
+      "loss": 0.3646,
+      "step": 9395
+    },
+    {
+      "epoch": 1.6044209268584109,
+      "grad_norm": 0.28406496348136745,
+      "learning_rate": 2.5840920586747596e-05,
+      "loss": 0.3579,
+      "step": 9400
+    },
+    {
+      "epoch": 1.605274387641888,
+      "grad_norm": 0.3616251180170039,
+      "learning_rate": 2.5825113808801217e-05,
+      "loss": 0.3526,
+      "step": 9405
+    },
+    {
+      "epoch": 1.606127848425365,
+      "grad_norm": 0.809971232093663,
+      "learning_rate": 2.5809307030854828e-05,
+      "loss": 0.3855,
+      "step": 9410
+    },
+    {
+      "epoch": 1.6069813092088419,
+      "grad_norm": 0.3519168231586514,
+      "learning_rate": 2.579350025290845e-05,
+      "loss": 0.3811,
+      "step": 9415
+    },
+    {
+      "epoch": 1.6078347699923188,
+      "grad_norm": 0.3721702280093411,
+      "learning_rate": 2.5777693474962066e-05,
+      "loss": 0.3781,
+      "step": 9420
+    },
+    {
+      "epoch": 1.6086882307757957,
+      "grad_norm": 0.3459379061704033,
+      "learning_rate": 2.576188669701568e-05,
+      "loss": 0.3824,
+      "step": 9425
+    },
+    {
+      "epoch": 1.6095416915592728,
+      "grad_norm": 0.38834628814085237,
+      "learning_rate": 2.5746079919069298e-05,
+      "loss": 0.3668,
+      "step": 9430
+    },
+    {
+      "epoch": 1.61039515234275,
+      "grad_norm": 0.3766126582753333,
+      "learning_rate": 2.5730273141122912e-05,
+      "loss": 0.3779,
+      "step": 9435
+    },
+    {
+      "epoch": 1.611248613126227,
+      "grad_norm": 0.5430538563590361,
+      "learning_rate": 2.5714466363176533e-05,
+      "loss": 0.3662,
+      "step": 9440
+    },
+    {
+      "epoch": 1.6121020739097038,
+      "grad_norm": 0.5057585205733404,
+      "learning_rate": 2.569865958523015e-05,
+      "loss": 0.3974,
+      "step": 9445
+    },
+    {
+      "epoch": 1.6129555346931808,
+      "grad_norm": 0.3993719986455356,
+      "learning_rate": 2.5682852807283764e-05,
+      "loss": 0.3776,
+      "step": 9450
+    },
+    {
+      "epoch": 1.6138089954766577,
+      "grad_norm": 0.43665620531008,
+      "learning_rate": 2.5667046029337382e-05,
+      "loss": 0.3669,
+      "step": 9455
+    },
+    {
+      "epoch": 1.6146624562601348,
+      "grad_norm": 0.37226881957182756,
+      "learning_rate": 2.5651239251390996e-05,
+      "loss": 0.3794,
+      "step": 9460
+    },
+    {
+      "epoch": 1.615515917043612,
+      "grad_norm": 0.630966720469816,
+      "learning_rate": 2.5635432473444613e-05,
+      "loss": 0.3592,
+      "step": 9465
+    },
+    {
+      "epoch": 1.616369377827089,
+      "grad_norm": 0.3939621369972374,
+      "learning_rate": 2.5619625695498227e-05,
+      "loss": 0.3896,
+      "step": 9470
+    },
+    {
+      "epoch": 1.6172228386105658,
+      "grad_norm": 0.43487915001420685,
+      "learning_rate": 2.5603818917551848e-05,
+      "loss": 0.3566,
+      "step": 9475
+    },
+    {
+      "epoch": 1.6180762993940427,
+      "grad_norm": 0.35916149088239563,
+      "learning_rate": 2.5588012139605466e-05,
+      "loss": 0.3883,
+      "step": 9480
+    },
+    {
+      "epoch": 1.6189297601775199,
+      "grad_norm": 0.36317699165579764,
+      "learning_rate": 2.557220536165908e-05,
+      "loss": 0.3516,
+      "step": 9485
+    },
+    {
+      "epoch": 1.6197832209609968,
+      "grad_norm": 0.351646812340972,
+      "learning_rate": 2.5556398583712697e-05,
+      "loss": 0.3682,
+      "step": 9490
+    },
+    {
+      "epoch": 1.620636681744474,
+      "grad_norm": 0.5051537961986949,
+      "learning_rate": 2.554059180576631e-05,
+      "loss": 0.3839,
+      "step": 9495
+    },
+    {
+      "epoch": 1.6214901425279509,
+      "grad_norm": 0.37903178035144436,
+      "learning_rate": 2.5524785027819932e-05,
+      "loss": 0.3817,
+      "step": 9500
+    },
+    {
+      "epoch": 1.6223436033114278,
+      "grad_norm": 0.41297845376305514,
+      "learning_rate": 2.550897824987355e-05,
+      "loss": 0.3823,
+      "step": 9505
+    },
+    {
+      "epoch": 1.6231970640949047,
+      "grad_norm": 0.36913854812775393,
+      "learning_rate": 2.5493171471927164e-05,
+      "loss": 0.3683,
+      "step": 9510
+    },
+    {
+      "epoch": 1.6240505248783819,
+      "grad_norm": 0.3610013396592195,
+      "learning_rate": 2.547736469398078e-05,
+      "loss": 0.3749,
+      "step": 9515
+    },
+    {
+      "epoch": 1.6249039856618588,
+      "grad_norm": 0.3870563224746467,
+      "learning_rate": 2.5461557916034395e-05,
+      "loss": 0.3783,
+      "step": 9520
+    },
+    {
+      "epoch": 1.625757446445336,
+      "grad_norm": 0.41329546305715414,
+      "learning_rate": 2.5445751138088013e-05,
+      "loss": 0.3624,
+      "step": 9525
+    },
+    {
+      "epoch": 1.6266109072288129,
+      "grad_norm": 0.3316826147883295,
+      "learning_rate": 2.5429944360141627e-05,
+      "loss": 0.3789,
+      "step": 9530
+    },
+    {
+      "epoch": 1.6274643680122898,
+      "grad_norm": 0.34889825755925274,
+      "learning_rate": 2.5414137582195248e-05,
+      "loss": 0.3531,
+      "step": 9535
+    },
+    {
+      "epoch": 1.6283178287957667,
+      "grad_norm": 0.35270280103669066,
+      "learning_rate": 2.5398330804248865e-05,
+      "loss": 0.3634,
+      "step": 9540
+    },
+    {
+      "epoch": 1.6291712895792438,
+      "grad_norm": 0.3818055527627773,
+      "learning_rate": 2.538252402630248e-05,
+      "loss": 0.367,
+      "step": 9545
+    },
+    {
+      "epoch": 1.630024750362721,
+      "grad_norm": 0.37334001874481865,
+      "learning_rate": 2.5366717248356097e-05,
+      "loss": 0.3504,
+      "step": 9550
+    },
+    {
+      "epoch": 1.630878211146198,
+      "grad_norm": 0.36463766396357167,
+      "learning_rate": 2.535091047040971e-05,
+      "loss": 0.3832,
+      "step": 9555
+    },
+    {
+      "epoch": 1.6317316719296748,
+      "grad_norm": 0.5269109992538331,
+      "learning_rate": 2.533510369246333e-05,
+      "loss": 0.3837,
+      "step": 9560
+    },
+    {
+      "epoch": 1.6325851327131518,
+      "grad_norm": 0.3894795153131237,
+      "learning_rate": 2.5319296914516942e-05,
+      "loss": 0.3826,
+      "step": 9565
+    },
+    {
+      "epoch": 1.6334385934966287,
+      "grad_norm": 0.4911555396539967,
+      "learning_rate": 2.5303490136570563e-05,
+      "loss": 0.3696,
+      "step": 9570
+    },
+    {
+      "epoch": 1.6342920542801058,
+      "grad_norm": 0.35874576935761504,
+      "learning_rate": 2.528768335862418e-05,
+      "loss": 0.3484,
+      "step": 9575
+    },
+    {
+      "epoch": 1.635145515063583,
+      "grad_norm": 0.3238490838476584,
+      "learning_rate": 2.5271876580677794e-05,
+      "loss": 0.3627,
+      "step": 9580
+    },
+    {
+      "epoch": 1.6359989758470599,
+      "grad_norm": 0.42527871728802813,
+      "learning_rate": 2.5256069802731412e-05,
+      "loss": 0.3923,
+      "step": 9585
+    },
+    {
+      "epoch": 1.6368524366305368,
+      "grad_norm": 0.32991180591923897,
+      "learning_rate": 2.5240263024785026e-05,
+      "loss": 0.3655,
+      "step": 9590
+    },
+    {
+      "epoch": 1.6377058974140137,
+      "grad_norm": 0.37893644741260296,
+      "learning_rate": 2.5224456246838647e-05,
+      "loss": 0.3674,
+      "step": 9595
+    },
+    {
+      "epoch": 1.6385593581974909,
+      "grad_norm": 0.36987743038688486,
+      "learning_rate": 2.5208649468892264e-05,
+      "loss": 0.3351,
+      "step": 9600
+    },
+    {
+      "epoch": 1.6394128189809678,
+      "grad_norm": 0.3811319455423694,
+      "learning_rate": 2.519284269094588e-05,
+      "loss": 0.3729,
+      "step": 9605
+    },
+    {
+      "epoch": 1.640266279764445,
+      "grad_norm": 0.35332370354771675,
+      "learning_rate": 2.5177035912999496e-05,
+      "loss": 0.3748,
+      "step": 9610
+    },
+    {
+      "epoch": 1.6411197405479219,
+      "grad_norm": 0.37714158965201894,
+      "learning_rate": 2.516122913505311e-05,
+      "loss": 0.3599,
+      "step": 9615
+    },
+    {
+      "epoch": 1.6419732013313988,
+      "grad_norm": 0.3532253534495296,
+      "learning_rate": 2.514542235710673e-05,
+      "loss": 0.4038,
+      "step": 9620
+    },
+    {
+      "epoch": 1.6428266621148757,
+      "grad_norm": 0.43917181291586693,
+      "learning_rate": 2.512961557916034e-05,
+      "loss": 0.3575,
+      "step": 9625
+    },
+    {
+      "epoch": 1.6436801228983529,
+      "grad_norm": 0.38639765123345415,
+      "learning_rate": 2.5113808801213962e-05,
+      "loss": 0.3534,
+      "step": 9630
+    },
+    {
+      "epoch": 1.6445335836818298,
+      "grad_norm": 0.3277553454765587,
+      "learning_rate": 2.509800202326758e-05,
+      "loss": 0.3479,
+      "step": 9635
+    },
+    {
+      "epoch": 1.645387044465307,
+      "grad_norm": 0.40405233020628756,
+      "learning_rate": 2.5082195245321194e-05,
+      "loss": 0.3726,
+      "step": 9640
+    },
+    {
+      "epoch": 1.6462405052487838,
+      "grad_norm": 0.40336960689920276,
+      "learning_rate": 2.506638846737481e-05,
+      "loss": 0.3843,
+      "step": 9645
+    },
+    {
+      "epoch": 1.6470939660322608,
+      "grad_norm": 0.3713185152846159,
+      "learning_rate": 2.5050581689428425e-05,
+      "loss": 0.3696,
+      "step": 9650
+    },
+    {
+      "epoch": 1.6479474268157377,
+      "grad_norm": 0.30365171938062613,
+      "learning_rate": 2.5034774911482046e-05,
+      "loss": 0.3871,
+      "step": 9655
+    },
+    {
+      "epoch": 1.6488008875992148,
+      "grad_norm": 0.39868229296031327,
+      "learning_rate": 2.5018968133535664e-05,
+      "loss": 0.3832,
+      "step": 9660
+    },
+    {
+      "epoch": 1.649654348382692,
+      "grad_norm": 0.31783690689314664,
+      "learning_rate": 2.5003161355589278e-05,
+      "loss": 0.3587,
+      "step": 9665
+    },
+    {
+      "epoch": 1.650507809166169,
+      "grad_norm": 0.44308631034536705,
+      "learning_rate": 2.4987354577642895e-05,
+      "loss": 0.3931,
+      "step": 9670
+    },
+    {
+      "epoch": 1.6513612699496458,
+      "grad_norm": 0.358684922290691,
+      "learning_rate": 2.497154779969651e-05,
+      "loss": 0.3561,
+      "step": 9675
+    },
+    {
+      "epoch": 1.6522147307331227,
+      "grad_norm": 0.3480662742675701,
+      "learning_rate": 2.495574102175013e-05,
+      "loss": 0.3917,
+      "step": 9680
+    },
+    {
+      "epoch": 1.6530681915165997,
+      "grad_norm": 0.344709757767127,
+      "learning_rate": 2.4939934243803744e-05,
+      "loss": 0.3533,
+      "step": 9685
+    },
+    {
+      "epoch": 1.6539216523000768,
+      "grad_norm": 0.5403638550208598,
+      "learning_rate": 2.492412746585736e-05,
+      "loss": 0.3756,
+      "step": 9690
+    },
+    {
+      "epoch": 1.654775113083554,
+      "grad_norm": 0.49560256448326895,
+      "learning_rate": 2.4908320687910976e-05,
+      "loss": 0.3619,
+      "step": 9695
+    },
+    {
+      "epoch": 1.6556285738670309,
+      "grad_norm": 0.4479970829130771,
+      "learning_rate": 2.4892513909964593e-05,
+      "loss": 0.3785,
+      "step": 9700
+    },
+    {
+      "epoch": 1.6564820346505078,
+      "grad_norm": 0.3643797818915992,
+      "learning_rate": 2.487670713201821e-05,
+      "loss": 0.3768,
+      "step": 9705
+    },
+    {
+      "epoch": 1.6573354954339847,
+      "grad_norm": 0.30986452031482137,
+      "learning_rate": 2.4860900354071828e-05,
+      "loss": 0.3684,
+      "step": 9710
+    },
+    {
+      "epoch": 1.6581889562174617,
+      "grad_norm": 0.34834214775622496,
+      "learning_rate": 2.4845093576125446e-05,
+      "loss": 0.3846,
+      "step": 9715
+    },
+    {
+      "epoch": 1.6590424170009388,
+      "grad_norm": 0.3240949698223232,
+      "learning_rate": 2.482928679817906e-05,
+      "loss": 0.3486,
+      "step": 9720
+    },
+    {
+      "epoch": 1.659895877784416,
+      "grad_norm": 0.4986974751176869,
+      "learning_rate": 2.4813480020232677e-05,
+      "loss": 0.3601,
+      "step": 9725
+    },
+    {
+      "epoch": 1.6607493385678929,
+      "grad_norm": 0.30454103767734464,
+      "learning_rate": 2.4797673242286294e-05,
+      "loss": 0.3635,
+      "step": 9730
+    },
+    {
+      "epoch": 1.6616027993513698,
+      "grad_norm": 0.4645322349398634,
+      "learning_rate": 2.478186646433991e-05,
+      "loss": 0.3974,
+      "step": 9735
+    },
+    {
+      "epoch": 1.6624562601348467,
+      "grad_norm": 0.33063087630227694,
+      "learning_rate": 2.476605968639353e-05,
+      "loss": 0.3595,
+      "step": 9740
+    },
+    {
+      "epoch": 1.6633097209183239,
+      "grad_norm": 0.3202227448808457,
+      "learning_rate": 2.4750252908447143e-05,
+      "loss": 0.3595,
+      "step": 9745
+    },
+    {
+      "epoch": 1.6641631817018008,
+      "grad_norm": 1.0984498791246415,
+      "learning_rate": 2.473444613050076e-05,
+      "loss": 0.3525,
+      "step": 9750
+    },
+    {
+      "epoch": 1.665016642485278,
+      "grad_norm": 0.3007709975602925,
+      "learning_rate": 2.4718639352554375e-05,
+      "loss": 0.3554,
+      "step": 9755
+    },
+    {
+      "epoch": 1.6658701032687548,
+      "grad_norm": 0.472316702108902,
+      "learning_rate": 2.4702832574607992e-05,
+      "loss": 0.3762,
+      "step": 9760
+    },
+    {
+      "epoch": 1.6667235640522318,
+      "grad_norm": 0.31508308960368503,
+      "learning_rate": 2.468702579666161e-05,
+      "loss": 0.3643,
+      "step": 9765
+    },
+    {
+      "epoch": 1.6675770248357087,
+      "grad_norm": 0.37722753928340996,
+      "learning_rate": 2.4671219018715227e-05,
+      "loss": 0.3902,
+      "step": 9770
+    },
+    {
+      "epoch": 1.6684304856191858,
+      "grad_norm": 0.3586764574812393,
+      "learning_rate": 2.4655412240768845e-05,
+      "loss": 0.3601,
+      "step": 9775
+    },
+    {
+      "epoch": 1.6692839464026628,
+      "grad_norm": 0.6380925765420232,
+      "learning_rate": 2.463960546282246e-05,
+      "loss": 0.3617,
+      "step": 9780
+    },
+    {
+      "epoch": 1.67013740718614,
+      "grad_norm": 0.3383616224692915,
+      "learning_rate": 2.4623798684876076e-05,
+      "loss": 0.3576,
+      "step": 9785
+    },
+    {
+      "epoch": 1.6709908679696168,
+      "grad_norm": 0.31425957298125445,
+      "learning_rate": 2.4607991906929694e-05,
+      "loss": 0.3393,
+      "step": 9790
+    },
+    {
+      "epoch": 1.6718443287530937,
+      "grad_norm": 0.4147864909222179,
+      "learning_rate": 2.4592185128983308e-05,
+      "loss": 0.3428,
+      "step": 9795
+    },
+    {
+      "epoch": 1.6726977895365707,
+      "grad_norm": 0.45264259568950715,
+      "learning_rate": 2.4576378351036925e-05,
+      "loss": 0.3886,
+      "step": 9800
+    },
+    {
+      "epoch": 1.6735512503200478,
+      "grad_norm": 0.36370608206159005,
+      "learning_rate": 2.4560571573090543e-05,
+      "loss": 0.3701,
+      "step": 9805
+    },
+    {
+      "epoch": 1.674404711103525,
+      "grad_norm": 0.3361008102105946,
+      "learning_rate": 2.454476479514416e-05,
+      "loss": 0.3765,
+      "step": 9810
+    },
+    {
+      "epoch": 1.6752581718870019,
+      "grad_norm": 0.31285208882339277,
+      "learning_rate": 2.4528958017197774e-05,
+      "loss": 0.3756,
+      "step": 9815
+    },
+    {
+      "epoch": 1.6761116326704788,
+      "grad_norm": 0.329371451846838,
+      "learning_rate": 2.4513151239251392e-05,
+      "loss": 0.3731,
+      "step": 9820
+    },
+    {
+      "epoch": 1.6769650934539557,
+      "grad_norm": 0.34101086796868185,
+      "learning_rate": 2.449734446130501e-05,
+      "loss": 0.3692,
+      "step": 9825
+    },
+    {
+      "epoch": 1.6778185542374326,
+      "grad_norm": 0.327415797309251,
+      "learning_rate": 2.4481537683358623e-05,
+      "loss": 0.3579,
+      "step": 9830
+    },
+    {
+      "epoch": 1.6786720150209098,
+      "grad_norm": 0.36212891124970775,
+      "learning_rate": 2.4465730905412244e-05,
+      "loss": 0.3544,
+      "step": 9835
+    },
+    {
+      "epoch": 1.679525475804387,
+      "grad_norm": 0.4808949754230029,
+      "learning_rate": 2.4449924127465858e-05,
+      "loss": 0.3717,
+      "step": 9840
+    },
+    {
+      "epoch": 1.6803789365878639,
+      "grad_norm": 0.35727492749761064,
+      "learning_rate": 2.4434117349519476e-05,
+      "loss": 0.37,
+      "step": 9845
+    },
+    {
+      "epoch": 1.6812323973713408,
+      "grad_norm": 0.5483344985996038,
+      "learning_rate": 2.441831057157309e-05,
+      "loss": 0.3586,
+      "step": 9850
+    },
+    {
+      "epoch": 1.6820858581548177,
+      "grad_norm": 0.38437005134873453,
+      "learning_rate": 2.4402503793626707e-05,
+      "loss": 0.3656,
+      "step": 9855
+    },
+    {
+      "epoch": 1.6829393189382946,
+      "grad_norm": 0.38040213577450993,
+      "learning_rate": 2.4386697015680325e-05,
+      "loss": 0.3733,
+      "step": 9860
+    },
+    {
+      "epoch": 1.6837927797217718,
+      "grad_norm": 0.35000993626904336,
+      "learning_rate": 2.4370890237733942e-05,
+      "loss": 0.355,
+      "step": 9865
+    },
+    {
+      "epoch": 1.684646240505249,
+      "grad_norm": 0.4048599449467024,
+      "learning_rate": 2.435508345978756e-05,
+      "loss": 0.3836,
+      "step": 9870
+    },
+    {
+      "epoch": 1.6854997012887258,
+      "grad_norm": 0.3751379685656599,
+      "learning_rate": 2.4339276681841174e-05,
+      "loss": 0.3967,
+      "step": 9875
+    },
+    {
+      "epoch": 1.6863531620722028,
+      "grad_norm": 0.3165261830400712,
+      "learning_rate": 2.432346990389479e-05,
+      "loss": 0.3419,
+      "step": 9880
+    },
+    {
+      "epoch": 1.6872066228556797,
+      "grad_norm": 0.38120809943595424,
+      "learning_rate": 2.430766312594841e-05,
+      "loss": 0.3962,
+      "step": 9885
+    },
+    {
+      "epoch": 1.6880600836391568,
+      "grad_norm": 0.3165567647867523,
+      "learning_rate": 2.4291856348002023e-05,
+      "loss": 0.3799,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6889135444226338,
+      "grad_norm": 0.35431128900669323,
+      "learning_rate": 2.4276049570055643e-05,
+      "loss": 0.3442,
+      "step": 9895
+    },
+    {
+      "epoch": 1.689767005206111,
+      "grad_norm": 0.3404306623210263,
+      "learning_rate": 2.4260242792109258e-05,
+      "loss": 0.3576,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6906204659895878,
+      "grad_norm": 0.38923418157520556,
+      "learning_rate": 2.4244436014162875e-05,
+      "loss": 0.3702,
+      "step": 9905
+    },
+    {
+      "epoch": 1.6914739267730647,
+      "grad_norm": 0.3701957123731767,
+      "learning_rate": 2.422862923621649e-05,
+      "loss": 0.3564,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6923273875565417,
+      "grad_norm": 0.3662805891531023,
+      "learning_rate": 2.4212822458270107e-05,
+      "loss": 0.394,
+      "step": 9915
+    },
+    {
+      "epoch": 1.6931808483400188,
+      "grad_norm": 0.3420394235920587,
+      "learning_rate": 2.4197015680323724e-05,
+      "loss": 0.3418,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6940343091234957,
+      "grad_norm": 0.30379527638119164,
+      "learning_rate": 2.418120890237734e-05,
+      "loss": 0.347,
+      "step": 9925
+    },
+    {
+      "epoch": 1.6948877699069729,
+      "grad_norm": 0.3383450236304016,
+      "learning_rate": 2.416540212443096e-05,
+      "loss": 0.3895,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6957412306904498,
+      "grad_norm": 0.36141397006934706,
+      "learning_rate": 2.4149595346484573e-05,
+      "loss": 0.3588,
+      "step": 9935
+    },
+    {
+      "epoch": 1.6965946914739267,
+      "grad_norm": 0.34458445452666825,
+      "learning_rate": 2.413378856853819e-05,
+      "loss": 0.3729,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6974481522574036,
+      "grad_norm": 0.3007875269036116,
+      "learning_rate": 2.4117981790591808e-05,
+      "loss": 0.3794,
+      "step": 9945
+    },
+    {
+      "epoch": 1.6983016130408808,
+      "grad_norm": 0.3532242665277833,
+      "learning_rate": 2.4102175012645422e-05,
+      "loss": 0.3651,
+      "step": 9950
+    },
+    {
+      "epoch": 1.699155073824358,
+      "grad_norm": 0.3374537867998225,
+      "learning_rate": 2.408636823469904e-05,
+      "loss": 0.3726,
+      "step": 9955
+    },
+    {
+      "epoch": 1.7000085346078349,
+      "grad_norm": 0.3328228201959643,
+      "learning_rate": 2.4070561456752657e-05,
+      "loss": 0.36,
+      "step": 9960
+    },
+    {
+      "epoch": 1.7008619953913118,
+      "grad_norm": 0.29198609814295523,
+      "learning_rate": 2.4054754678806274e-05,
+      "loss": 0.3924,
+      "step": 9965
+    },
+    {
+      "epoch": 1.7017154561747887,
+      "grad_norm": 0.3613528956300614,
+      "learning_rate": 2.403894790085989e-05,
+      "loss": 0.3762,
+      "step": 9970
+    },
+    {
+      "epoch": 1.7025689169582656,
+      "grad_norm": 0.42149928313668533,
+      "learning_rate": 2.4023141122913506e-05,
+      "loss": 0.387,
+      "step": 9975
+    },
+    {
+      "epoch": 1.7034223777417428,
+      "grad_norm": 0.4093811404104952,
+      "learning_rate": 2.4007334344967123e-05,
+      "loss": 0.3757,
+      "step": 9980
+    },
+    {
+      "epoch": 1.70427583852522,
+      "grad_norm": 0.31232798392425404,
+      "learning_rate": 2.3991527567020737e-05,
+      "loss": 0.3588,
+      "step": 9985
+    },
+    {
+      "epoch": 1.7051292993086968,
+      "grad_norm": 0.39676720303151275,
+      "learning_rate": 2.3975720789074358e-05,
+      "loss": 0.3714,
+      "step": 9990
+    },
+    {
+      "epoch": 1.7059827600921738,
+      "grad_norm": 1.2822911601946423,
+      "learning_rate": 2.3959914011127972e-05,
+      "loss": 0.3758,
+      "step": 9995
+    },
+    {
+      "epoch": 1.7068362208756507,
+      "grad_norm": 0.38618418843218183,
+      "learning_rate": 2.394410723318159e-05,
+      "loss": 0.356,
+      "step": 10000
+    },
+    {
+      "epoch": 1.7076896816591276,
+      "grad_norm": 0.28221632964104154,
+      "learning_rate": 2.3928300455235207e-05,
+      "loss": 0.3503,
+      "step": 10005
+    },
+    {
+      "epoch": 1.7085431424426047,
+      "grad_norm": 1.8906529696817969,
+      "learning_rate": 2.391249367728882e-05,
+      "loss": 0.3981,
+      "step": 10010
+    },
+    {
+      "epoch": 1.709396603226082,
+      "grad_norm": 0.37767485910480847,
+      "learning_rate": 2.389668689934244e-05,
+      "loss": 0.3476,
+      "step": 10015
+    },
+    {
+      "epoch": 1.7102500640095588,
+      "grad_norm": 0.5819221743981302,
+      "learning_rate": 2.3880880121396056e-05,
+      "loss": 0.3376,
+      "step": 10020
+    },
+    {
+      "epoch": 1.7111035247930357,
+      "grad_norm": 0.36288759811979004,
+      "learning_rate": 2.3865073343449674e-05,
+      "loss": 0.3754,
+      "step": 10025
+    },
+    {
+      "epoch": 1.7119569855765127,
+      "grad_norm": 0.3691645217803475,
+      "learning_rate": 2.3849266565503288e-05,
+      "loss": 0.3642,
+      "step": 10030
+    },
+    {
+      "epoch": 1.7128104463599898,
+      "grad_norm": 0.3892005134188453,
+      "learning_rate": 2.3833459787556905e-05,
+      "loss": 0.3643,
+      "step": 10035
+    },
+    {
+      "epoch": 1.7136639071434667,
+      "grad_norm": 0.3789287552931439,
+      "learning_rate": 2.3817653009610523e-05,
+      "loss": 0.3512,
+      "step": 10040
+    },
+    {
+      "epoch": 1.7145173679269439,
+      "grad_norm": 0.497272194873337,
+      "learning_rate": 2.3801846231664137e-05,
+      "loss": 0.3353,
+      "step": 10045
+    },
+    {
+      "epoch": 1.7153708287104208,
+      "grad_norm": 0.7177210423529197,
+      "learning_rate": 2.3786039453717758e-05,
+      "loss": 0.3646,
+      "step": 10050
+    },
+    {
+      "epoch": 1.7162242894938977,
+      "grad_norm": 0.35545288550050264,
+      "learning_rate": 2.377023267577137e-05,
+      "loss": 0.3839,
+      "step": 10055
+    },
+    {
+      "epoch": 1.7170777502773746,
+      "grad_norm": 0.3475567969581151,
+      "learning_rate": 2.375442589782499e-05,
+      "loss": 0.3621,
+      "step": 10060
+    },
+    {
+      "epoch": 1.7179312110608518,
+      "grad_norm": 0.32022819698741734,
+      "learning_rate": 2.3738619119878607e-05,
+      "loss": 0.367,
+      "step": 10065
+    },
+    {
+      "epoch": 1.7187846718443287,
+      "grad_norm": 0.38584311787546577,
+      "learning_rate": 2.372281234193222e-05,
+      "loss": 0.3487,
+      "step": 10070
+    },
+    {
+      "epoch": 1.7196381326278058,
+      "grad_norm": 0.385576863467266,
+      "learning_rate": 2.3707005563985838e-05,
+      "loss": 0.3812,
+      "step": 10075
+    },
+    {
+      "epoch": 1.7204915934112828,
+      "grad_norm": 0.29632475338105196,
+      "learning_rate": 2.3691198786039452e-05,
+      "loss": 0.369,
+      "step": 10080
+    },
+    {
+      "epoch": 1.7213450541947597,
+      "grad_norm": 0.3455973736066388,
+      "learning_rate": 2.3675392008093073e-05,
+      "loss": 0.358,
+      "step": 10085
+    },
+    {
+      "epoch": 1.7221985149782366,
+      "grad_norm": 0.3218490085562857,
+      "learning_rate": 2.3659585230146687e-05,
+      "loss": 0.3736,
+      "step": 10090
+    },
+    {
+      "epoch": 1.7230519757617138,
+      "grad_norm": 0.40162109561154596,
+      "learning_rate": 2.3643778452200305e-05,
+      "loss": 0.3573,
+      "step": 10095
+    },
+    {
+      "epoch": 1.723905436545191,
+      "grad_norm": 0.4538109362924726,
+      "learning_rate": 2.3627971674253922e-05,
+      "loss": 0.3653,
+      "step": 10100
+    },
+    {
+      "epoch": 1.7247588973286678,
+      "grad_norm": 0.33572753039570163,
+      "learning_rate": 2.3612164896307536e-05,
+      "loss": 0.354,
+      "step": 10105
+    },
+    {
+      "epoch": 1.7256123581121448,
+      "grad_norm": 0.32031756723298443,
+      "learning_rate": 2.3596358118361154e-05,
+      "loss": 0.3789,
+      "step": 10110
+    },
+    {
+      "epoch": 1.7264658188956217,
+      "grad_norm": 0.29450177423269897,
+      "learning_rate": 2.358055134041477e-05,
+      "loss": 0.3616,
+      "step": 10115
+    },
+    {
+      "epoch": 1.7273192796790986,
+      "grad_norm": 0.3565780967071999,
+      "learning_rate": 2.356474456246839e-05,
+      "loss": 0.3696,
+      "step": 10120
+    },
+    {
+      "epoch": 1.7281727404625757,
+      "grad_norm": 0.4310022631641016,
+      "learning_rate": 2.3548937784522006e-05,
+      "loss": 0.3505,
+      "step": 10125
+    },
+    {
+      "epoch": 1.7290262012460529,
+      "grad_norm": 0.3691592372099381,
+      "learning_rate": 2.353313100657562e-05,
+      "loss": 0.3567,
+      "step": 10130
+    },
+    {
+      "epoch": 1.7298796620295298,
+      "grad_norm": 0.316140033477628,
+      "learning_rate": 2.3517324228629237e-05,
+      "loss": 0.3406,
+      "step": 10135
+    },
+    {
+      "epoch": 1.7307331228130067,
+      "grad_norm": 0.3632478706450496,
+      "learning_rate": 2.350151745068285e-05,
+      "loss": 0.3545,
+      "step": 10140
+    },
+    {
+      "epoch": 1.7315865835964837,
+      "grad_norm": 0.4417561062140959,
+      "learning_rate": 2.3485710672736472e-05,
+      "loss": 0.3608,
+      "step": 10145
+    },
+    {
+      "epoch": 1.7324400443799606,
+      "grad_norm": 0.3284369454386695,
+      "learning_rate": 2.3469903894790086e-05,
+      "loss": 0.3714,
+      "step": 10150
+    },
+    {
+      "epoch": 1.7332935051634377,
+      "grad_norm": 0.3769972935790573,
+      "learning_rate": 2.3454097116843704e-05,
+      "loss": 0.3675,
+      "step": 10155
+    },
+    {
+      "epoch": 1.7341469659469149,
+      "grad_norm": 0.39666950231256687,
+      "learning_rate": 2.343829033889732e-05,
+      "loss": 0.3866,
+      "step": 10160
+    },
+    {
+      "epoch": 1.7350004267303918,
+      "grad_norm": 0.4108714616727913,
+      "learning_rate": 2.3422483560950935e-05,
+      "loss": 0.3868,
+      "step": 10165
+    },
+    {
+      "epoch": 1.7358538875138687,
+      "grad_norm": 0.36737787273895234,
+      "learning_rate": 2.3406676783004553e-05,
+      "loss": 0.3769,
+      "step": 10170
+    },
+    {
+      "epoch": 1.7367073482973456,
+      "grad_norm": 0.4113399543149594,
+      "learning_rate": 2.339087000505817e-05,
+      "loss": 0.388,
+      "step": 10175
+    },
+    {
+      "epoch": 1.7375608090808228,
+      "grad_norm": 0.39169442625744216,
+      "learning_rate": 2.3375063227111788e-05,
+      "loss": 0.3444,
+      "step": 10180
+    },
+    {
+      "epoch": 1.7384142698642997,
+      "grad_norm": 0.29177123848011854,
+      "learning_rate": 2.3359256449165405e-05,
+      "loss": 0.3753,
+      "step": 10185
+    },
+    {
+      "epoch": 1.7392677306477768,
+      "grad_norm": 0.3383770494978673,
+      "learning_rate": 2.334344967121902e-05,
+      "loss": 0.3561,
+      "step": 10190
+    },
+    {
+      "epoch": 1.7401211914312538,
+      "grad_norm": 0.3724319975484265,
+      "learning_rate": 2.3327642893272637e-05,
+      "loss": 0.3926,
+      "step": 10195
+    },
+    {
+      "epoch": 1.7409746522147307,
+      "grad_norm": 0.4986854755701425,
+      "learning_rate": 2.331183611532625e-05,
+      "loss": 0.3941,
+      "step": 10200
+    },
+    {
+      "epoch": 1.7418281129982076,
+      "grad_norm": 0.4390558287126891,
+      "learning_rate": 2.3296029337379868e-05,
+      "loss": 0.3591,
+      "step": 10205
+    },
+    {
+      "epoch": 1.7426815737816848,
+      "grad_norm": 0.33959687704720026,
+      "learning_rate": 2.3280222559433486e-05,
+      "loss": 0.3671,
+      "step": 10210
+    },
+    {
+      "epoch": 1.7435350345651617,
+      "grad_norm": 0.29763719970702696,
+      "learning_rate": 2.3264415781487103e-05,
+      "loss": 0.3561,
+      "step": 10215
+    },
+    {
+      "epoch": 1.7443884953486388,
+      "grad_norm": 0.475731508454673,
+      "learning_rate": 2.324860900354072e-05,
+      "loss": 0.3844,
+      "step": 10220
+    },
+    {
+      "epoch": 1.7452419561321157,
+      "grad_norm": 0.3356547866961049,
+      "learning_rate": 2.3232802225594335e-05,
+      "loss": 0.387,
+      "step": 10225
+    },
+    {
+      "epoch": 1.7460954169155927,
+      "grad_norm": 0.3505708880260287,
+      "learning_rate": 2.3216995447647952e-05,
+      "loss": 0.3617,
+      "step": 10230
+    },
+    {
+      "epoch": 1.7469488776990696,
+      "grad_norm": 0.39851400765382256,
+      "learning_rate": 2.3201188669701566e-05,
+      "loss": 0.3761,
+      "step": 10235
+    },
+    {
+      "epoch": 1.7478023384825467,
+      "grad_norm": 0.3762568602632757,
+      "learning_rate": 2.3185381891755187e-05,
+      "loss": 0.3758,
+      "step": 10240
+    },
+    {
+      "epoch": 1.7486557992660239,
+      "grad_norm": 0.35144964132478174,
+      "learning_rate": 2.3169575113808805e-05,
+      "loss": 0.3809,
+      "step": 10245
+    },
+    {
+      "epoch": 1.7495092600495008,
+      "grad_norm": 0.41360431860790936,
+      "learning_rate": 2.315376833586242e-05,
+      "loss": 0.3545,
+      "step": 10250
+    },
+    {
+      "epoch": 1.7503627208329777,
+      "grad_norm": 0.3154093255362032,
+      "learning_rate": 2.3137961557916036e-05,
+      "loss": 0.3491,
+      "step": 10255
+    },
+    {
+      "epoch": 1.7512161816164546,
+      "grad_norm": 1.4137719194929725,
+      "learning_rate": 2.312215477996965e-05,
+      "loss": 0.3721,
+      "step": 10260
+    },
+    {
+      "epoch": 1.7520696423999316,
+      "grad_norm": 0.34016453789967843,
+      "learning_rate": 2.3106348002023268e-05,
+      "loss": 0.3647,
+      "step": 10265
+    },
+    {
+      "epoch": 1.7529231031834087,
+      "grad_norm": 0.38216600334837486,
+      "learning_rate": 2.3090541224076885e-05,
+      "loss": 0.3609,
+      "step": 10270
+    },
+    {
+      "epoch": 1.7537765639668859,
+      "grad_norm": 0.33344039798406455,
+      "learning_rate": 2.3074734446130503e-05,
+      "loss": 0.3817,
+      "step": 10275
+    },
+    {
+      "epoch": 1.7546300247503628,
+      "grad_norm": 0.3505756778067142,
+      "learning_rate": 2.305892766818412e-05,
+      "loss": 0.3999,
+      "step": 10280
+    },
+    {
+      "epoch": 1.7554834855338397,
+      "grad_norm": 0.2908220891462724,
+      "learning_rate": 2.3043120890237734e-05,
+      "loss": 0.376,
+      "step": 10285
+    },
+    {
+      "epoch": 1.7563369463173166,
+      "grad_norm": 0.4531864302133631,
+      "learning_rate": 2.302731411229135e-05,
+      "loss": 0.353,
+      "step": 10290
+    },
+    {
+      "epoch": 1.7571904071007938,
+      "grad_norm": 0.3960793007598124,
+      "learning_rate": 2.3011507334344966e-05,
+      "loss": 0.3773,
+      "step": 10295
+    },
+    {
+      "epoch": 1.7580438678842707,
+      "grad_norm": 0.2987489593726992,
+      "learning_rate": 2.2995700556398586e-05,
+      "loss": 0.3791,
+      "step": 10300
+    },
+    {
+      "epoch": 1.7588973286677478,
+      "grad_norm": 0.39906326924904945,
+      "learning_rate": 2.2979893778452204e-05,
+      "loss": 0.3499,
+      "step": 10305
+    },
+    {
+      "epoch": 1.7597507894512248,
+      "grad_norm": 0.4200226635689522,
+      "learning_rate": 2.2964087000505818e-05,
+      "loss": 0.3922,
+      "step": 10310
+    },
+    {
+      "epoch": 1.7606042502347017,
+      "grad_norm": 0.35742914237215445,
+      "learning_rate": 2.2948280222559435e-05,
+      "loss": 0.3775,
+      "step": 10315
+    },
+    {
+      "epoch": 1.7614577110181786,
+      "grad_norm": 0.3327129777054162,
+      "learning_rate": 2.293247344461305e-05,
+      "loss": 0.3762,
+      "step": 10320
+    },
+    {
+      "epoch": 1.7623111718016558,
+      "grad_norm": 0.36120457351243423,
+      "learning_rate": 2.2916666666666667e-05,
+      "loss": 0.3462,
+      "step": 10325
+    },
+    {
+      "epoch": 1.7631646325851327,
+      "grad_norm": 0.3472423813510895,
+      "learning_rate": 2.2900859888720284e-05,
+      "loss": 0.3787,
+      "step": 10330
+    },
+    {
+      "epoch": 1.7640180933686098,
+      "grad_norm": 0.36961950942276844,
+      "learning_rate": 2.2885053110773902e-05,
+      "loss": 0.3682,
+      "step": 10335
+    },
+    {
+      "epoch": 1.7648715541520867,
+      "grad_norm": 0.32673319213139956,
+      "learning_rate": 2.286924633282752e-05,
+      "loss": 0.3489,
+      "step": 10340
+    },
+    {
+      "epoch": 1.7657250149355637,
+      "grad_norm": 0.3268261260976071,
+      "learning_rate": 2.2853439554881133e-05,
+      "loss": 0.3739,
+      "step": 10345
+    },
+    {
+      "epoch": 1.7665784757190406,
+      "grad_norm": 0.3530261635729621,
+      "learning_rate": 2.283763277693475e-05,
+      "loss": 0.3879,
+      "step": 10350
+    },
+    {
+      "epoch": 1.7674319365025177,
+      "grad_norm": 0.3268436411952027,
+      "learning_rate": 2.2821825998988365e-05,
+      "loss": 0.3636,
+      "step": 10355
+    },
+    {
+      "epoch": 1.7682853972859947,
+      "grad_norm": 0.3229942323059562,
+      "learning_rate": 2.2806019221041982e-05,
+      "loss": 0.3821,
+      "step": 10360
+    },
+    {
+      "epoch": 1.7691388580694718,
+      "grad_norm": 0.5568367398126395,
+      "learning_rate": 2.2790212443095603e-05,
+      "loss": 0.3725,
+      "step": 10365
+    },
+    {
+      "epoch": 1.7699923188529487,
+      "grad_norm": 0.3977205017910189,
+      "learning_rate": 2.2774405665149217e-05,
+      "loss": 0.3722,
+      "step": 10370
+    },
+    {
+      "epoch": 1.7708457796364256,
+      "grad_norm": 0.3460952652909466,
+      "learning_rate": 2.2758598887202835e-05,
+      "loss": 0.3771,
+      "step": 10375
+    },
+    {
+      "epoch": 1.7716992404199026,
+      "grad_norm": 0.43227517832685763,
+      "learning_rate": 2.274279210925645e-05,
+      "loss": 0.361,
+      "step": 10380
+    },
+    {
+      "epoch": 1.7725527012033797,
+      "grad_norm": 0.3524632847384698,
+      "learning_rate": 2.2726985331310066e-05,
+      "loss": 0.3521,
+      "step": 10385
+    },
+    {
+      "epoch": 1.7734061619868569,
+      "grad_norm": 0.7478051106054819,
+      "learning_rate": 2.2711178553363684e-05,
+      "loss": 0.3458,
+      "step": 10390
+    },
+    {
+      "epoch": 1.7742596227703338,
+      "grad_norm": 0.4109190829078148,
+      "learning_rate": 2.26953717754173e-05,
+      "loss": 0.351,
+      "step": 10395
+    },
+    {
+      "epoch": 1.7751130835538107,
+      "grad_norm": 1.5678713067526842,
+      "learning_rate": 2.267956499747092e-05,
+      "loss": 0.4199,
+      "step": 10400
+    },
+    {
+      "epoch": 1.7759665443372876,
+      "grad_norm": 0.42333237544214836,
+      "learning_rate": 2.2663758219524533e-05,
+      "loss": 0.3572,
+      "step": 10405
+    },
+    {
+      "epoch": 1.7768200051207645,
+      "grad_norm": 0.39219567327864413,
+      "learning_rate": 2.264795144157815e-05,
+      "loss": 0.3654,
+      "step": 10410
+    },
+    {
+      "epoch": 1.7776734659042417,
+      "grad_norm": 0.49609649621588464,
+      "learning_rate": 2.2632144663631764e-05,
+      "loss": 0.3465,
+      "step": 10415
+    },
+    {
+      "epoch": 1.7785269266877188,
+      "grad_norm": 0.37267128866203747,
+      "learning_rate": 2.261633788568538e-05,
+      "loss": 0.3754,
+      "step": 10420
+    },
+    {
+      "epoch": 1.7793803874711958,
+      "grad_norm": 0.3748502000740906,
+      "learning_rate": 2.2600531107739003e-05,
+      "loss": 0.3533,
+      "step": 10425
+    },
+    {
+      "epoch": 1.7802338482546727,
+      "grad_norm": 0.4463158896622891,
+      "learning_rate": 2.2584724329792617e-05,
+      "loss": 0.3623,
+      "step": 10430
+    },
+    {
+      "epoch": 1.7810873090381496,
+      "grad_norm": 0.3555591871261391,
+      "learning_rate": 2.2568917551846234e-05,
+      "loss": 0.3414,
+      "step": 10435
+    },
+    {
+      "epoch": 1.7819407698216267,
+      "grad_norm": 0.4765275492824118,
+      "learning_rate": 2.2553110773899848e-05,
+      "loss": 0.3873,
+      "step": 10440
+    },
+    {
+      "epoch": 1.7827942306051037,
+      "grad_norm": 0.46616758415086174,
+      "learning_rate": 2.2537303995953466e-05,
+      "loss": 0.3629,
+      "step": 10445
+    },
+    {
+      "epoch": 1.7836476913885808,
+      "grad_norm": 0.34573027591378713,
+      "learning_rate": 2.2521497218007083e-05,
+      "loss": 0.3782,
+      "step": 10450
+    },
+    {
+      "epoch": 1.7845011521720577,
+      "grad_norm": 0.5251500050749887,
+      "learning_rate": 2.25056904400607e-05,
+      "loss": 0.3665,
+      "step": 10455
+    },
+    {
+      "epoch": 1.7853546129555347,
+      "grad_norm": 0.3764774923808963,
+      "learning_rate": 2.2489883662114318e-05,
+      "loss": 0.3656,
+      "step": 10460
+    },
+    {
+      "epoch": 1.7862080737390116,
+      "grad_norm": 0.3382205983972903,
+      "learning_rate": 2.2474076884167932e-05,
+      "loss": 0.3604,
+      "step": 10465
+    },
+    {
+      "epoch": 1.7870615345224887,
+      "grad_norm": 0.5317388106722926,
+      "learning_rate": 2.245827010622155e-05,
+      "loss": 0.356,
+      "step": 10470
+    },
+    {
+      "epoch": 1.7879149953059656,
+      "grad_norm": 1.4903783114176634,
+      "learning_rate": 2.2442463328275164e-05,
+      "loss": 0.3596,
+      "step": 10475
+    },
+    {
+      "epoch": 1.7887684560894428,
+      "grad_norm": 0.36496507612857976,
+      "learning_rate": 2.242665655032878e-05,
+      "loss": 0.3502,
+      "step": 10480
+    },
+    {
+      "epoch": 1.7896219168729197,
+      "grad_norm": 0.3937612529189051,
+      "learning_rate": 2.24108497723824e-05,
+      "loss": 0.3995,
+      "step": 10485
+    },
+    {
+      "epoch": 1.7904753776563966,
+      "grad_norm": 0.31508253953639564,
+      "learning_rate": 2.2395042994436016e-05,
+      "loss": 0.36,
+      "step": 10490
+    },
+    {
+      "epoch": 1.7913288384398736,
+      "grad_norm": 0.4404646582186003,
+      "learning_rate": 2.2379236216489633e-05,
+      "loss": 0.3781,
+      "step": 10495
+    },
+    {
+      "epoch": 1.7921822992233507,
+      "grad_norm": 0.32583844450645083,
+      "learning_rate": 2.2363429438543247e-05,
+      "loss": 0.3931,
+      "step": 10500
+    },
+    {
+      "epoch": 1.7930357600068279,
+      "grad_norm": 0.902758638180003,
+      "learning_rate": 2.2347622660596865e-05,
+      "loss": 0.3594,
+      "step": 10505
+    },
+    {
+      "epoch": 1.7938892207903048,
+      "grad_norm": 0.3365733522219216,
+      "learning_rate": 2.2331815882650482e-05,
+      "loss": 0.3591,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7947426815737817,
+      "grad_norm": 0.4934569504171031,
+      "learning_rate": 2.2316009104704096e-05,
+      "loss": 0.3699,
+      "step": 10515
+    },
+    {
+      "epoch": 1.7955961423572586,
+      "grad_norm": 0.34520444249585464,
+      "learning_rate": 2.2300202326757717e-05,
+      "loss": 0.358,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7964496031407355,
+      "grad_norm": 0.40578450288272916,
+      "learning_rate": 2.228439554881133e-05,
+      "loss": 0.3727,
+      "step": 10525
+    },
+    {
+      "epoch": 1.7973030639242127,
+      "grad_norm": 0.4141773602465772,
+      "learning_rate": 2.226858877086495e-05,
+      "loss": 0.3776,
+      "step": 10530
+    },
+    {
+      "epoch": 1.7981565247076898,
+      "grad_norm": 0.39931050960669756,
+      "learning_rate": 2.2252781992918563e-05,
+      "loss": 0.3614,
+      "step": 10535
+    },
+    {
+      "epoch": 1.7990099854911668,
+      "grad_norm": 0.5136945097329725,
+      "learning_rate": 2.223697521497218e-05,
+      "loss": 0.364,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7998634462746437,
+      "grad_norm": 0.4337614573293977,
+      "learning_rate": 2.2221168437025798e-05,
+      "loss": 0.3793,
+      "step": 10545
+    },
+    {
+      "epoch": 1.8007169070581206,
+      "grad_norm": 0.31625048284765106,
+      "learning_rate": 2.2205361659079415e-05,
+      "loss": 0.3454,
+      "step": 10550
+    },
+    {
+      "epoch": 1.8015703678415975,
+      "grad_norm": 0.7410211808156124,
+      "learning_rate": 2.2189554881133033e-05,
+      "loss": 0.3663,
+      "step": 10555
+    },
+    {
+      "epoch": 1.8024238286250747,
+      "grad_norm": 0.4919923140230198,
+      "learning_rate": 2.2173748103186647e-05,
+      "loss": 0.3439,
+      "step": 10560
+    },
+    {
+      "epoch": 1.8032772894085518,
+      "grad_norm": 0.37579765211598076,
+      "learning_rate": 2.2157941325240264e-05,
+      "loss": 0.3683,
+      "step": 10565
+    },
+    {
+      "epoch": 1.8041307501920287,
+      "grad_norm": 0.36895523353830173,
+      "learning_rate": 2.214213454729388e-05,
+      "loss": 0.3569,
+      "step": 10570
+    },
+    {
+      "epoch": 1.8049842109755057,
+      "grad_norm": 0.3033833823701229,
+      "learning_rate": 2.2126327769347496e-05,
+      "loss": 0.3732,
+      "step": 10575
+    },
+    {
+      "epoch": 1.8058376717589826,
+      "grad_norm": 0.39266161317165993,
+      "learning_rate": 2.2110520991401117e-05,
+      "loss": 0.3621,
+      "step": 10580
+    },
+    {
+      "epoch": 1.8066911325424597,
+      "grad_norm": 0.42813518689174307,
+      "learning_rate": 2.209471421345473e-05,
+      "loss": 0.3716,
+      "step": 10585
+    },
+    {
+      "epoch": 1.8075445933259366,
+      "grad_norm": 0.3638748905898181,
+      "learning_rate": 2.2078907435508348e-05,
+      "loss": 0.3718,
+      "step": 10590
+    },
+    {
+      "epoch": 1.8083980541094138,
+      "grad_norm": 0.5950632824365684,
+      "learning_rate": 2.2063100657561962e-05,
+      "loss": 0.3659,
+      "step": 10595
+    },
+    {
+      "epoch": 1.8092515148928907,
+      "grad_norm": 0.3486629271103143,
+      "learning_rate": 2.204729387961558e-05,
+      "loss": 0.353,
+      "step": 10600
+    },
+    {
+      "epoch": 1.8101049756763676,
+      "grad_norm": 0.3942577678706724,
+      "learning_rate": 2.2031487101669197e-05,
+      "loss": 0.3758,
+      "step": 10605
+    },
+    {
+      "epoch": 1.8109584364598446,
+      "grad_norm": 0.3631065961712478,
+      "learning_rate": 2.201568032372281e-05,
+      "loss": 0.3628,
+      "step": 10610
+    },
+    {
+      "epoch": 1.8118118972433217,
+      "grad_norm": 0.28850470186236804,
+      "learning_rate": 2.1999873545776432e-05,
+      "loss": 0.3583,
+      "step": 10615
+    },
+    {
+      "epoch": 1.8126653580267986,
+      "grad_norm": 0.3084338342525628,
+      "learning_rate": 2.1984066767830046e-05,
+      "loss": 0.3747,
+      "step": 10620
+    },
+    {
+      "epoch": 1.8135188188102758,
+      "grad_norm": 0.48109290660100135,
+      "learning_rate": 2.1968259989883664e-05,
+      "loss": 0.357,
+      "step": 10625
+    },
+    {
+      "epoch": 1.8143722795937527,
+      "grad_norm": 0.593473543871764,
+      "learning_rate": 2.195245321193728e-05,
+      "loss": 0.3727,
+      "step": 10630
+    },
+    {
+      "epoch": 1.8152257403772296,
+      "grad_norm": 0.4100937199773082,
+      "learning_rate": 2.1936646433990895e-05,
+      "loss": 0.3697,
+      "step": 10635
+    },
+    {
+      "epoch": 1.8160792011607065,
+      "grad_norm": 0.32877977611372144,
+      "learning_rate": 2.1920839656044513e-05,
+      "loss": 0.3617,
+      "step": 10640
+    },
+    {
+      "epoch": 1.8169326619441837,
+      "grad_norm": 0.35612619937395745,
+      "learning_rate": 2.190503287809813e-05,
+      "loss": 0.3466,
+      "step": 10645
+    },
+    {
+      "epoch": 1.8177861227276608,
+      "grad_norm": 0.35396257019449956,
+      "learning_rate": 2.1889226100151747e-05,
+      "loss": 0.3571,
+      "step": 10650
+    },
+    {
+      "epoch": 1.8186395835111377,
+      "grad_norm": 0.39690438673164563,
+      "learning_rate": 2.187341932220536e-05,
+      "loss": 0.3674,
+      "step": 10655
+    },
+    {
+      "epoch": 1.8194930442946147,
+      "grad_norm": 0.5242482990321449,
+      "learning_rate": 2.185761254425898e-05,
+      "loss": 0.3953,
+      "step": 10660
+    },
+    {
+      "epoch": 1.8203465050780916,
+      "grad_norm": 0.4214529237838405,
+      "learning_rate": 2.1841805766312596e-05,
+      "loss": 0.3727,
+      "step": 10665
+    },
+    {
+      "epoch": 1.8211999658615685,
+      "grad_norm": 1.1136726595991726,
+      "learning_rate": 2.182599898836621e-05,
+      "loss": 0.3728,
+      "step": 10670
+    },
+    {
+      "epoch": 1.8220534266450457,
+      "grad_norm": 0.327682234110669,
+      "learning_rate": 2.181019221041983e-05,
+      "loss": 0.3747,
+      "step": 10675
+    },
+    {
+      "epoch": 1.8229068874285228,
+      "grad_norm": 0.28611775290909447,
+      "learning_rate": 2.1794385432473445e-05,
+      "loss": 0.3587,
+      "step": 10680
+    },
+    {
+      "epoch": 1.8237603482119997,
+      "grad_norm": 0.4407991387242986,
+      "learning_rate": 2.1778578654527063e-05,
+      "loss": 0.364,
+      "step": 10685
+    },
+    {
+      "epoch": 1.8246138089954766,
+      "grad_norm": 0.47526465046818295,
+      "learning_rate": 2.176277187658068e-05,
+      "loss": 0.3497,
+      "step": 10690
+    },
+    {
+      "epoch": 1.8254672697789536,
+      "grad_norm": 0.3928753576304463,
+      "learning_rate": 2.1746965098634294e-05,
+      "loss": 0.359,
+      "step": 10695
+    },
+    {
+      "epoch": 1.8263207305624305,
+      "grad_norm": 0.36232606753636554,
+      "learning_rate": 2.1731158320687912e-05,
+      "loss": 0.3623,
+      "step": 10700
+    },
+    {
+      "epoch": 1.8271741913459076,
+      "grad_norm": 0.46836363032120787,
+      "learning_rate": 2.171535154274153e-05,
+      "loss": 0.3518,
+      "step": 10705
+    },
+    {
+      "epoch": 1.8280276521293848,
+      "grad_norm": 0.3895529808865236,
+      "learning_rate": 2.1699544764795147e-05,
+      "loss": 0.3751,
+      "step": 10710
+    },
+    {
+      "epoch": 1.8288811129128617,
+      "grad_norm": 0.5460199251051238,
+      "learning_rate": 2.168373798684876e-05,
+      "loss": 0.3708,
+      "step": 10715
+    },
+    {
+      "epoch": 1.8297345736963386,
+      "grad_norm": 0.40974357623987206,
+      "learning_rate": 2.166793120890238e-05,
+      "loss": 0.3667,
+      "step": 10720
+    },
+    {
+      "epoch": 1.8305880344798156,
+      "grad_norm": 0.38992645911807666,
+      "learning_rate": 2.1652124430955996e-05,
+      "loss": 0.366,
+      "step": 10725
+    },
+    {
+      "epoch": 1.8314414952632927,
+      "grad_norm": 0.41047123669309143,
+      "learning_rate": 2.163631765300961e-05,
+      "loss": 0.3931,
+      "step": 10730
+    },
+    {
+      "epoch": 1.8322949560467696,
+      "grad_norm": 0.40673065790434665,
+      "learning_rate": 2.1620510875063227e-05,
+      "loss": 0.3805,
+      "step": 10735
+    },
+    {
+      "epoch": 1.8331484168302468,
+      "grad_norm": 0.5322917028214879,
+      "learning_rate": 2.1604704097116845e-05,
+      "loss": 0.3715,
+      "step": 10740
+    },
+    {
+      "epoch": 1.8340018776137237,
+      "grad_norm": 0.3181019613142807,
+      "learning_rate": 2.1588897319170462e-05,
+      "loss": 0.3868,
+      "step": 10745
+    },
+    {
+      "epoch": 1.8348553383972006,
+      "grad_norm": 0.31096990946905845,
+      "learning_rate": 2.157309054122408e-05,
+      "loss": 0.3615,
+      "step": 10750
+    },
+    {
+      "epoch": 1.8357087991806775,
+      "grad_norm": 0.36823862849584077,
+      "learning_rate": 2.1557283763277694e-05,
+      "loss": 0.368,
+      "step": 10755
+    },
+    {
+      "epoch": 1.8365622599641547,
+      "grad_norm": 0.3250417469863049,
+      "learning_rate": 2.154147698533131e-05,
+      "loss": 0.3659,
+      "step": 10760
+    },
+    {
+      "epoch": 1.8374157207476316,
+      "grad_norm": 0.45584925598743514,
+      "learning_rate": 2.1525670207384925e-05,
+      "loss": 0.3677,
+      "step": 10765
+    },
+    {
+      "epoch": 1.8382691815311087,
+      "grad_norm": 0.36190482925923284,
+      "learning_rate": 2.1509863429438546e-05,
+      "loss": 0.3798,
+      "step": 10770
+    },
+    {
+      "epoch": 1.8391226423145857,
+      "grad_norm": 0.4258078428713466,
+      "learning_rate": 2.149405665149216e-05,
+      "loss": 0.3453,
+      "step": 10775
+    },
+    {
+      "epoch": 1.8399761030980626,
+      "grad_norm": 0.5146661868072934,
+      "learning_rate": 2.1478249873545778e-05,
+      "loss": 0.3835,
+      "step": 10780
+    },
+    {
+      "epoch": 1.8408295638815395,
+      "grad_norm": 0.35699074579188195,
+      "learning_rate": 2.1462443095599395e-05,
+      "loss": 0.3917,
+      "step": 10785
+    },
+    {
+      "epoch": 1.8416830246650167,
+      "grad_norm": 0.7331207287649647,
+      "learning_rate": 2.144663631765301e-05,
+      "loss": 0.3935,
+      "step": 10790
+    },
+    {
+      "epoch": 1.8425364854484938,
+      "grad_norm": 0.3419839330270435,
+      "learning_rate": 2.1430829539706627e-05,
+      "loss": 0.3599,
+      "step": 10795
+    },
+    {
+      "epoch": 1.8433899462319707,
+      "grad_norm": 0.3822457077968837,
+      "learning_rate": 2.1415022761760244e-05,
+      "loss": 0.3644,
+      "step": 10800
+    },
+    {
+      "epoch": 1.8442434070154476,
+      "grad_norm": 0.4084306894441126,
+      "learning_rate": 2.139921598381386e-05,
+      "loss": 0.3534,
+      "step": 10805
+    },
+    {
+      "epoch": 1.8450968677989246,
+      "grad_norm": 0.4874881885032455,
+      "learning_rate": 2.138340920586748e-05,
+      "loss": 0.3841,
+      "step": 10810
+    },
+    {
+      "epoch": 1.8459503285824015,
+      "grad_norm": 0.4127729887949137,
+      "learning_rate": 2.1367602427921093e-05,
+      "loss": 0.3733,
+      "step": 10815
+    },
+    {
+      "epoch": 1.8468037893658786,
+      "grad_norm": 0.43915191784894403,
+      "learning_rate": 2.135179564997471e-05,
+      "loss": 0.3719,
+      "step": 10820
+    },
+    {
+      "epoch": 1.8476572501493558,
+      "grad_norm": 0.4940688648002275,
+      "learning_rate": 2.1335988872028325e-05,
+      "loss": 0.361,
+      "step": 10825
+    },
+    {
+      "epoch": 1.8485107109328327,
+      "grad_norm": 0.45004907244050096,
+      "learning_rate": 2.1320182094081945e-05,
+      "loss": 0.3734,
+      "step": 10830
+    },
+    {
+      "epoch": 1.8493641717163096,
+      "grad_norm": 0.5525318440453658,
+      "learning_rate": 2.130437531613556e-05,
+      "loss": 0.3438,
+      "step": 10835
+    },
+    {
+      "epoch": 1.8502176324997865,
+      "grad_norm": 0.326606963379231,
+      "learning_rate": 2.1288568538189177e-05,
+      "loss": 0.3491,
+      "step": 10840
+    },
+    {
+      "epoch": 1.8510710932832635,
+      "grad_norm": 0.37527385318494266,
+      "learning_rate": 2.1272761760242794e-05,
+      "loss": 0.4038,
+      "step": 10845
+    },
+    {
+      "epoch": 1.8519245540667406,
+      "grad_norm": 0.3939938913670757,
+      "learning_rate": 2.125695498229641e-05,
+      "loss": 0.3715,
+      "step": 10850
+    },
+    {
+      "epoch": 1.8527780148502178,
+      "grad_norm": 0.326301072046663,
+      "learning_rate": 2.1241148204350026e-05,
+      "loss": 0.3764,
+      "step": 10855
+    },
+    {
+      "epoch": 1.8536314756336947,
+      "grad_norm": 0.4400152642190683,
+      "learning_rate": 2.122534142640364e-05,
+      "loss": 0.3747,
+      "step": 10860
+    },
+    {
+      "epoch": 1.8544849364171716,
+      "grad_norm": 0.43012907394312316,
+      "learning_rate": 2.120953464845726e-05,
+      "loss": 0.3546,
+      "step": 10865
+    },
+    {
+      "epoch": 1.8553383972006485,
+      "grad_norm": 0.3541062863654811,
+      "learning_rate": 2.1193727870510875e-05,
+      "loss": 0.385,
+      "step": 10870
+    },
+    {
+      "epoch": 1.8561918579841257,
+      "grad_norm": 0.42864551170374077,
+      "learning_rate": 2.1177921092564492e-05,
+      "loss": 0.3632,
+      "step": 10875
+    },
+    {
+      "epoch": 1.8570453187676026,
+      "grad_norm": 0.6092328779930328,
+      "learning_rate": 2.116211431461811e-05,
+      "loss": 0.3545,
+      "step": 10880
+    },
+    {
+      "epoch": 1.8578987795510797,
+      "grad_norm": 0.36373812884671775,
+      "learning_rate": 2.1146307536671724e-05,
+      "loss": 0.3811,
+      "step": 10885
+    },
+    {
+      "epoch": 1.8587522403345567,
+      "grad_norm": 0.48966929587361846,
+      "learning_rate": 2.113050075872534e-05,
+      "loss": 0.3629,
+      "step": 10890
+    },
+    {
+      "epoch": 1.8596057011180336,
+      "grad_norm": 0.3294541247900809,
+      "learning_rate": 2.111469398077896e-05,
+      "loss": 0.3589,
+      "step": 10895
+    },
+    {
+      "epoch": 1.8604591619015105,
+      "grad_norm": 0.3762098114886841,
+      "learning_rate": 2.1098887202832576e-05,
+      "loss": 0.3624,
+      "step": 10900
+    },
+    {
+      "epoch": 1.8613126226849876,
+      "grad_norm": 0.4498972865269915,
+      "learning_rate": 2.1083080424886194e-05,
+      "loss": 0.3742,
+      "step": 10905
+    },
+    {
+      "epoch": 1.8621660834684646,
+      "grad_norm": 0.3806443462128479,
+      "learning_rate": 2.1067273646939808e-05,
+      "loss": 0.3544,
+      "step": 10910
+    },
+    {
+      "epoch": 1.8630195442519417,
+      "grad_norm": 0.4187900337274261,
+      "learning_rate": 2.1051466868993425e-05,
+      "loss": 0.3744,
+      "step": 10915
+    },
+    {
+      "epoch": 1.8638730050354186,
+      "grad_norm": 0.43709248143647517,
+      "learning_rate": 2.103566009104704e-05,
+      "loss": 0.3653,
+      "step": 10920
+    },
+    {
+      "epoch": 1.8647264658188956,
+      "grad_norm": 0.5847621244254114,
+      "learning_rate": 2.101985331310066e-05,
+      "loss": 0.3741,
+      "step": 10925
+    },
+    {
+      "epoch": 1.8655799266023725,
+      "grad_norm": 0.3726576210378853,
+      "learning_rate": 2.1004046535154274e-05,
+      "loss": 0.3444,
+      "step": 10930
+    },
+    {
+      "epoch": 1.8664333873858496,
+      "grad_norm": 0.7331755128492993,
+      "learning_rate": 2.0988239757207892e-05,
+      "loss": 0.3854,
+      "step": 10935
+    },
+    {
+      "epoch": 1.8672868481693268,
+      "grad_norm": 0.3396553895603976,
+      "learning_rate": 2.097243297926151e-05,
+      "loss": 0.3815,
+      "step": 10940
+    },
+    {
+      "epoch": 1.8681403089528037,
+      "grad_norm": 0.361490335839466,
+      "learning_rate": 2.0956626201315123e-05,
+      "loss": 0.3724,
+      "step": 10945
+    },
+    {
+      "epoch": 1.8689937697362806,
+      "grad_norm": 0.3673689734763417,
+      "learning_rate": 2.094081942336874e-05,
+      "loss": 0.349,
+      "step": 10950
+    },
+    {
+      "epoch": 1.8698472305197575,
+      "grad_norm": 0.39744288396915534,
+      "learning_rate": 2.0925012645422358e-05,
+      "loss": 0.3663,
+      "step": 10955
+    },
+    {
+      "epoch": 1.8707006913032345,
+      "grad_norm": 0.42295654199751714,
+      "learning_rate": 2.0909205867475976e-05,
+      "loss": 0.3801,
+      "step": 10960
+    },
+    {
+      "epoch": 1.8715541520867116,
+      "grad_norm": 0.360461220134643,
+      "learning_rate": 2.0893399089529593e-05,
+      "loss": 0.3642,
+      "step": 10965
+    },
+    {
+      "epoch": 1.8724076128701888,
+      "grad_norm": 0.2789941075918898,
+      "learning_rate": 2.0877592311583207e-05,
+      "loss": 0.3565,
+      "step": 10970
+    },
+    {
+      "epoch": 1.8732610736536657,
+      "grad_norm": 0.46125387391646,
+      "learning_rate": 2.0861785533636825e-05,
+      "loss": 0.3521,
+      "step": 10975
+    },
+    {
+      "epoch": 1.8741145344371426,
+      "grad_norm": 0.41105185493252544,
+      "learning_rate": 2.084597875569044e-05,
+      "loss": 0.3831,
+      "step": 10980
+    },
+    {
+      "epoch": 1.8749679952206195,
+      "grad_norm": 0.354426825365594,
+      "learning_rate": 2.0830171977744056e-05,
+      "loss": 0.373,
+      "step": 10985
+    },
+    {
+      "epoch": 1.8758214560040967,
+      "grad_norm": 0.34640586063871975,
+      "learning_rate": 2.0814365199797674e-05,
+      "loss": 0.3784,
+      "step": 10990
+    },
+    {
+      "epoch": 1.8766749167875736,
+      "grad_norm": 0.409642869323863,
+      "learning_rate": 2.079855842185129e-05,
+      "loss": 0.3844,
+      "step": 10995
+    },
+    {
+      "epoch": 1.8775283775710507,
+      "grad_norm": 0.3194906214235848,
+      "learning_rate": 2.078275164390491e-05,
+      "loss": 0.3612,
+      "step": 11000
+    },
+    {
+      "epoch": 1.8783818383545277,
+      "grad_norm": 0.36984853554037195,
+      "learning_rate": 2.0766944865958523e-05,
+      "loss": 0.3589,
+      "step": 11005
+    },
+    {
+      "epoch": 1.8792352991380046,
+      "grad_norm": 0.37674468309420256,
+      "learning_rate": 2.075113808801214e-05,
+      "loss": 0.3684,
+      "step": 11010
+    },
+    {
+      "epoch": 1.8800887599214815,
+      "grad_norm": 0.3577464377599332,
+      "learning_rate": 2.0735331310065757e-05,
+      "loss": 0.3433,
+      "step": 11015
+    },
+    {
+      "epoch": 1.8809422207049586,
+      "grad_norm": 0.3657277409746499,
+      "learning_rate": 2.0719524532119375e-05,
+      "loss": 0.3678,
+      "step": 11020
+    },
+    {
+      "epoch": 1.8817956814884356,
+      "grad_norm": 0.3239080611529195,
+      "learning_rate": 2.0703717754172992e-05,
+      "loss": 0.3668,
+      "step": 11025
+    },
+    {
+      "epoch": 1.8826491422719127,
+      "grad_norm": 0.4133691952735056,
+      "learning_rate": 2.0687910976226606e-05,
+      "loss": 0.3417,
+      "step": 11030
+    },
+    {
+      "epoch": 1.8835026030553896,
+      "grad_norm": 0.31025729042046996,
+      "learning_rate": 2.0672104198280224e-05,
+      "loss": 0.3651,
+      "step": 11035
+    },
+    {
+      "epoch": 1.8843560638388666,
+      "grad_norm": 0.2653544568477746,
+      "learning_rate": 2.0656297420333838e-05,
+      "loss": 0.396,
+      "step": 11040
+    },
+    {
+      "epoch": 1.8852095246223435,
+      "grad_norm": 0.34035557157096347,
+      "learning_rate": 2.0640490642387455e-05,
+      "loss": 0.3692,
+      "step": 11045
+    },
+    {
+      "epoch": 1.8860629854058206,
+      "grad_norm": 0.25176173499633664,
+      "learning_rate": 2.0624683864441073e-05,
+      "loss": 0.3688,
+      "step": 11050
+    },
+    {
+      "epoch": 1.8869164461892975,
+      "grad_norm": 0.4855932237933519,
+      "learning_rate": 2.060887708649469e-05,
+      "loss": 0.3518,
+      "step": 11055
+    },
+    {
+      "epoch": 1.8877699069727747,
+      "grad_norm": 0.7704487813945038,
+      "learning_rate": 2.0593070308548308e-05,
+      "loss": 0.3626,
+      "step": 11060
+    },
+    {
+      "epoch": 1.8886233677562516,
+      "grad_norm": 0.397441661431441,
+      "learning_rate": 2.0577263530601922e-05,
+      "loss": 0.3453,
+      "step": 11065
+    },
+    {
+      "epoch": 1.8894768285397285,
+      "grad_norm": 0.3122045807760479,
+      "learning_rate": 2.056145675265554e-05,
+      "loss": 0.3464,
+      "step": 11070
+    },
+    {
+      "epoch": 1.8903302893232055,
+      "grad_norm": 0.3885736967225049,
+      "learning_rate": 2.0545649974709157e-05,
+      "loss": 0.3596,
+      "step": 11075
+    },
+    {
+      "epoch": 1.8911837501066826,
+      "grad_norm": 0.33420689549505933,
+      "learning_rate": 2.0529843196762774e-05,
+      "loss": 0.3389,
+      "step": 11080
+    },
+    {
+      "epoch": 1.8920372108901597,
+      "grad_norm": 0.4701672640614866,
+      "learning_rate": 2.0514036418816392e-05,
+      "loss": 0.3685,
+      "step": 11085
+    },
+    {
+      "epoch": 1.8928906716736367,
+      "grad_norm": 0.3229552993180945,
+      "learning_rate": 2.0498229640870006e-05,
+      "loss": 0.3536,
+      "step": 11090
+    },
+    {
+      "epoch": 1.8937441324571136,
+      "grad_norm": 0.9219103914160934,
+      "learning_rate": 2.0482422862923623e-05,
+      "loss": 0.3542,
+      "step": 11095
+    },
+    {
+      "epoch": 1.8945975932405905,
+      "grad_norm": 0.32298789544258444,
+      "learning_rate": 2.0466616084977237e-05,
+      "loss": 0.3662,
+      "step": 11100
+    },
+    {
+      "epoch": 1.8954510540240674,
+      "grad_norm": 0.4619059980890969,
+      "learning_rate": 2.0450809307030855e-05,
+      "loss": 0.3539,
+      "step": 11105
+    },
+    {
+      "epoch": 1.8963045148075446,
+      "grad_norm": 0.3683441404553503,
+      "learning_rate": 2.0435002529084472e-05,
+      "loss": 0.3582,
+      "step": 11110
+    },
+    {
+      "epoch": 1.8971579755910217,
+      "grad_norm": 0.6612297982391425,
+      "learning_rate": 2.041919575113809e-05,
+      "loss": 0.3232,
+      "step": 11115
+    },
+    {
+      "epoch": 1.8980114363744987,
+      "grad_norm": 0.5625119257799257,
+      "learning_rate": 2.0403388973191707e-05,
+      "loss": 0.3498,
+      "step": 11120
+    },
+    {
+      "epoch": 1.8988648971579756,
+      "grad_norm": 0.3541435603942317,
+      "learning_rate": 2.038758219524532e-05,
+      "loss": 0.3626,
+      "step": 11125
+    },
+    {
+      "epoch": 1.8997183579414525,
+      "grad_norm": 0.3155807293604779,
+      "learning_rate": 2.037177541729894e-05,
+      "loss": 0.371,
+      "step": 11130
+    },
+    {
+      "epoch": 1.9005718187249296,
+      "grad_norm": 0.3826111273138795,
+      "learning_rate": 2.0355968639352556e-05,
+      "loss": 0.3849,
+      "step": 11135
+    },
+    {
+      "epoch": 1.9014252795084066,
+      "grad_norm": 0.37475780016940374,
+      "learning_rate": 2.034016186140617e-05,
+      "loss": 0.388,
+      "step": 11140
+    },
+    {
+      "epoch": 1.9022787402918837,
+      "grad_norm": 0.32392024173798867,
+      "learning_rate": 2.032435508345979e-05,
+      "loss": 0.3687,
+      "step": 11145
+    },
+    {
+      "epoch": 1.9031322010753606,
+      "grad_norm": 0.3030706062232896,
+      "learning_rate": 2.0308548305513405e-05,
+      "loss": 0.3674,
+      "step": 11150
+    },
+    {
+      "epoch": 1.9039856618588376,
+      "grad_norm": 0.47629616214997345,
+      "learning_rate": 2.0292741527567023e-05,
+      "loss": 0.371,
+      "step": 11155
+    },
+    {
+      "epoch": 1.9048391226423145,
+      "grad_norm": 0.352930507119559,
+      "learning_rate": 2.0276934749620637e-05,
+      "loss": 0.3804,
+      "step": 11160
+    },
+    {
+      "epoch": 1.9056925834257916,
+      "grad_norm": 0.457987910274872,
+      "learning_rate": 2.0261127971674254e-05,
+      "loss": 0.3407,
+      "step": 11165
+    },
+    {
+      "epoch": 1.9065460442092685,
+      "grad_norm": 0.31367635452942055,
+      "learning_rate": 2.024532119372787e-05,
+      "loss": 0.3606,
+      "step": 11170
+    },
+    {
+      "epoch": 1.9073995049927457,
+      "grad_norm": 0.3238065727672519,
+      "learning_rate": 2.022951441578149e-05,
+      "loss": 0.3878,
+      "step": 11175
+    },
+    {
+      "epoch": 1.9082529657762226,
+      "grad_norm": 0.32758691010339364,
+      "learning_rate": 2.0213707637835106e-05,
+      "loss": 0.3306,
+      "step": 11180
+    },
+    {
+      "epoch": 1.9091064265596995,
+      "grad_norm": 0.3119731116231921,
+      "learning_rate": 2.019790085988872e-05,
+      "loss": 0.3471,
+      "step": 11185
+    },
+    {
+      "epoch": 1.9099598873431765,
+      "grad_norm": 0.36084222880900263,
+      "learning_rate": 2.0182094081942338e-05,
+      "loss": 0.3712,
+      "step": 11190
+    },
+    {
+      "epoch": 1.9108133481266536,
+      "grad_norm": 0.4090976778708613,
+      "learning_rate": 2.0166287303995955e-05,
+      "loss": 0.3518,
+      "step": 11195
+    },
+    {
+      "epoch": 1.9116668089101307,
+      "grad_norm": 0.37520141214522906,
+      "learning_rate": 2.015048052604957e-05,
+      "loss": 0.3704,
+      "step": 11200
+    },
+    {
+      "epoch": 1.9125202696936077,
+      "grad_norm": 0.37574938620862963,
+      "learning_rate": 2.013467374810319e-05,
+      "loss": 0.3622,
+      "step": 11205
+    },
+    {
+      "epoch": 1.9133737304770846,
+      "grad_norm": 0.44480299534892875,
+      "learning_rate": 2.0118866970156804e-05,
+      "loss": 0.3363,
+      "step": 11210
+    },
+    {
+      "epoch": 1.9142271912605615,
+      "grad_norm": 0.35464180850066146,
+      "learning_rate": 2.0103060192210422e-05,
+      "loss": 0.3519,
+      "step": 11215
+    },
+    {
+      "epoch": 1.9150806520440384,
+      "grad_norm": 0.3324121298201554,
+      "learning_rate": 2.0087253414264036e-05,
+      "loss": 0.3676,
+      "step": 11220
+    },
+    {
+      "epoch": 1.9159341128275156,
+      "grad_norm": 0.3533966232549052,
+      "learning_rate": 2.0071446636317653e-05,
+      "loss": 0.3652,
+      "step": 11225
+    },
+    {
+      "epoch": 1.9167875736109927,
+      "grad_norm": 0.36029559671898753,
+      "learning_rate": 2.005563985837127e-05,
+      "loss": 0.365,
+      "step": 11230
+    },
+    {
+      "epoch": 1.9176410343944696,
+      "grad_norm": 0.4350178376187806,
+      "learning_rate": 2.003983308042489e-05,
+      "loss": 0.3608,
+      "step": 11235
+    },
+    {
+      "epoch": 1.9184944951779466,
+      "grad_norm": 0.35699404918806876,
+      "learning_rate": 2.0024026302478506e-05,
+      "loss": 0.3931,
+      "step": 11240
+    },
+    {
+      "epoch": 1.9193479559614235,
+      "grad_norm": 0.3446151856091827,
+      "learning_rate": 2.000821952453212e-05,
+      "loss": 0.3812,
+      "step": 11245
+    },
+    {
+      "epoch": 1.9202014167449004,
+      "grad_norm": 0.3526287131807039,
+      "learning_rate": 1.9992412746585737e-05,
+      "loss": 0.3908,
+      "step": 11250
+    },
+    {
+      "epoch": 1.9210548775283776,
+      "grad_norm": 0.38504624802916554,
+      "learning_rate": 1.997660596863935e-05,
+      "loss": 0.3554,
+      "step": 11255
+    },
+    {
+      "epoch": 1.9219083383118547,
+      "grad_norm": 0.37777764566097194,
+      "learning_rate": 1.996079919069297e-05,
+      "loss": 0.3651,
+      "step": 11260
+    },
+    {
+      "epoch": 1.9227617990953316,
+      "grad_norm": 0.3125281065863975,
+      "learning_rate": 1.9944992412746586e-05,
+      "loss": 0.3639,
+      "step": 11265
+    },
+    {
+      "epoch": 1.9236152598788085,
+      "grad_norm": 0.6245378333272329,
+      "learning_rate": 1.9929185634800204e-05,
+      "loss": 0.3501,
+      "step": 11270
+    },
+    {
+      "epoch": 1.9244687206622855,
+      "grad_norm": 0.3280900134912351,
+      "learning_rate": 1.991337885685382e-05,
+      "loss": 0.368,
+      "step": 11275
+    },
+    {
+      "epoch": 1.9253221814457626,
+      "grad_norm": 0.398874931108072,
+      "learning_rate": 1.9897572078907435e-05,
+      "loss": 0.3731,
+      "step": 11280
+    },
+    {
+      "epoch": 1.9261756422292395,
+      "grad_norm": 0.35962176717808686,
+      "learning_rate": 1.9881765300961053e-05,
+      "loss": 0.355,
+      "step": 11285
+    },
+    {
+      "epoch": 1.9270291030127167,
+      "grad_norm": 0.3338696595537415,
+      "learning_rate": 1.986595852301467e-05,
+      "loss": 0.3613,
+      "step": 11290
+    },
+    {
+      "epoch": 1.9278825637961936,
+      "grad_norm": 0.3065677342530692,
+      "learning_rate": 1.9850151745068284e-05,
+      "loss": 0.3326,
+      "step": 11295
+    },
+    {
+      "epoch": 1.9287360245796705,
+      "grad_norm": 0.32622888385701815,
+      "learning_rate": 1.9834344967121905e-05,
+      "loss": 0.3662,
+      "step": 11300
+    },
+    {
+      "epoch": 1.9295894853631474,
+      "grad_norm": 0.3292226957746066,
+      "learning_rate": 1.981853818917552e-05,
+      "loss": 0.3492,
+      "step": 11305
+    },
+    {
+      "epoch": 1.9304429461466246,
+      "grad_norm": 0.3667746710987046,
+      "learning_rate": 1.9802731411229137e-05,
+      "loss": 0.359,
+      "step": 11310
+    },
+    {
+      "epoch": 1.9312964069301015,
+      "grad_norm": 0.38086219300338536,
+      "learning_rate": 1.978692463328275e-05,
+      "loss": 0.3716,
+      "step": 11315
+    },
+    {
+      "epoch": 1.9321498677135787,
+      "grad_norm": 0.4722974686907586,
+      "learning_rate": 1.9771117855336368e-05,
+      "loss": 0.3744,
+      "step": 11320
+    },
+    {
+      "epoch": 1.9330033284970556,
+      "grad_norm": 0.3304308918934173,
+      "learning_rate": 1.9755311077389986e-05,
+      "loss": 0.3722,
+      "step": 11325
+    },
+    {
+      "epoch": 1.9338567892805325,
+      "grad_norm": 0.39895671630370916,
+      "learning_rate": 1.9739504299443603e-05,
+      "loss": 0.3566,
+      "step": 11330
+    },
+    {
+      "epoch": 1.9347102500640094,
+      "grad_norm": 0.33445870300135433,
+      "learning_rate": 1.972369752149722e-05,
+      "loss": 0.3688,
+      "step": 11335
+    },
+    {
+      "epoch": 1.9355637108474866,
+      "grad_norm": 0.3624243432384039,
+      "learning_rate": 1.9707890743550835e-05,
+      "loss": 0.3741,
+      "step": 11340
+    },
+    {
+      "epoch": 1.9364171716309637,
+      "grad_norm": 0.38588661759205267,
+      "learning_rate": 1.9692083965604452e-05,
+      "loss": 0.3764,
+      "step": 11345
+    },
+    {
+      "epoch": 1.9372706324144406,
+      "grad_norm": 0.3861164155920311,
+      "learning_rate": 1.967627718765807e-05,
+      "loss": 0.3538,
+      "step": 11350
+    },
+    {
+      "epoch": 1.9381240931979176,
+      "grad_norm": 0.3571311832146459,
+      "learning_rate": 1.9660470409711684e-05,
+      "loss": 0.3737,
+      "step": 11355
+    },
+    {
+      "epoch": 1.9389775539813945,
+      "grad_norm": 0.3503066025720848,
+      "learning_rate": 1.9644663631765304e-05,
+      "loss": 0.37,
+      "step": 11360
+    },
+    {
+      "epoch": 1.9398310147648714,
+      "grad_norm": 0.31192881974980474,
+      "learning_rate": 1.962885685381892e-05,
+      "loss": 0.3645,
+      "step": 11365
+    },
+    {
+      "epoch": 1.9406844755483486,
+      "grad_norm": 0.4104030121912594,
+      "learning_rate": 1.9613050075872536e-05,
+      "loss": 0.3631,
+      "step": 11370
+    },
+    {
+      "epoch": 1.9415379363318257,
+      "grad_norm": 0.3066578523567166,
+      "learning_rate": 1.959724329792615e-05,
+      "loss": 0.3606,
+      "step": 11375
+    },
+    {
+      "epoch": 1.9423913971153026,
+      "grad_norm": 0.261312140891305,
+      "learning_rate": 1.9581436519979768e-05,
+      "loss": 0.3952,
+      "step": 11380
+    },
+    {
+      "epoch": 1.9432448578987795,
+      "grad_norm": 0.3413334933644581,
+      "learning_rate": 1.9565629742033385e-05,
+      "loss": 0.3779,
+      "step": 11385
+    },
+    {
+      "epoch": 1.9440983186822565,
+      "grad_norm": 0.3786158496861722,
+      "learning_rate": 1.9549822964087e-05,
+      "loss": 0.3623,
+      "step": 11390
+    },
+    {
+      "epoch": 1.9449517794657334,
+      "grad_norm": 0.3694445662662908,
+      "learning_rate": 1.953401618614062e-05,
+      "loss": 0.3544,
+      "step": 11395
+    },
+    {
+      "epoch": 1.9458052402492105,
+      "grad_norm": 0.3624189966994706,
+      "learning_rate": 1.9518209408194234e-05,
+      "loss": 0.3631,
+      "step": 11400
+    },
+    {
+      "epoch": 1.9466587010326877,
+      "grad_norm": 0.28941849668007225,
+      "learning_rate": 1.950240263024785e-05,
+      "loss": 0.339,
+      "step": 11405
+    },
+    {
+      "epoch": 1.9475121618161646,
+      "grad_norm": 0.7991996892591037,
+      "learning_rate": 1.948659585230147e-05,
+      "loss": 0.3741,
+      "step": 11410
+    },
+    {
+      "epoch": 1.9483656225996415,
+      "grad_norm": 0.3605776512297038,
+      "learning_rate": 1.9470789074355083e-05,
+      "loss": 0.3637,
+      "step": 11415
+    },
+    {
+      "epoch": 1.9492190833831184,
+      "grad_norm": 0.3319018707245592,
+      "learning_rate": 1.94549822964087e-05,
+      "loss": 0.3642,
+      "step": 11420
+    },
+    {
+      "epoch": 1.9500725441665956,
+      "grad_norm": 0.3925895291989305,
+      "learning_rate": 1.9439175518462318e-05,
+      "loss": 0.3571,
+      "step": 11425
+    },
+    {
+      "epoch": 1.9509260049500725,
+      "grad_norm": 0.4445945431545767,
+      "learning_rate": 1.9423368740515935e-05,
+      "loss": 0.3602,
+      "step": 11430
+    },
+    {
+      "epoch": 1.9517794657335497,
+      "grad_norm": 0.3263330072466279,
+      "learning_rate": 1.940756196256955e-05,
+      "loss": 0.3799,
+      "step": 11435
+    },
+    {
+      "epoch": 1.9526329265170266,
+      "grad_norm": 0.9395242224578074,
+      "learning_rate": 1.9391755184623167e-05,
+      "loss": 0.3598,
+      "step": 11440
+    },
+    {
+      "epoch": 1.9534863873005035,
+      "grad_norm": 0.33484251821398536,
+      "learning_rate": 1.9375948406676784e-05,
+      "loss": 0.3762,
+      "step": 11445
+    },
+    {
+      "epoch": 1.9543398480839804,
+      "grad_norm": 0.49024773217209083,
+      "learning_rate": 1.93601416287304e-05,
+      "loss": 0.395,
+      "step": 11450
+    },
+    {
+      "epoch": 1.9551933088674576,
+      "grad_norm": 0.3717301732218937,
+      "learning_rate": 1.934433485078402e-05,
+      "loss": 0.3876,
+      "step": 11455
+    },
+    {
+      "epoch": 1.9560467696509345,
+      "grad_norm": 0.33147738637956553,
+      "learning_rate": 1.9328528072837633e-05,
+      "loss": 0.3644,
+      "step": 11460
+    },
+    {
+      "epoch": 1.9569002304344116,
+      "grad_norm": 0.4014253422212822,
+      "learning_rate": 1.931272129489125e-05,
+      "loss": 0.3651,
+      "step": 11465
+    },
+    {
+      "epoch": 1.9577536912178886,
+      "grad_norm": 0.3400205004163349,
+      "learning_rate": 1.9296914516944868e-05,
+      "loss": 0.3463,
+      "step": 11470
+    },
+    {
+      "epoch": 1.9586071520013655,
+      "grad_norm": 0.3264483566767699,
+      "learning_rate": 1.9281107738998482e-05,
+      "loss": 0.3701,
+      "step": 11475
+    },
+    {
+      "epoch": 1.9594606127848424,
+      "grad_norm": 0.38572487164686403,
+      "learning_rate": 1.92653009610521e-05,
+      "loss": 0.3384,
+      "step": 11480
+    },
+    {
+      "epoch": 1.9603140735683195,
+      "grad_norm": 0.4467785592513792,
+      "learning_rate": 1.9249494183105717e-05,
+      "loss": 0.3617,
+      "step": 11485
+    },
+    {
+      "epoch": 1.9611675343517967,
+      "grad_norm": 0.3129488961883966,
+      "learning_rate": 1.9233687405159335e-05,
+      "loss": 0.3857,
+      "step": 11490
+    },
+    {
+      "epoch": 1.9620209951352736,
+      "grad_norm": 0.3432421500027621,
+      "learning_rate": 1.921788062721295e-05,
+      "loss": 0.3481,
+      "step": 11495
+    },
+    {
+      "epoch": 1.9628744559187505,
+      "grad_norm": 0.8997788869734178,
+      "learning_rate": 1.9202073849266566e-05,
+      "loss": 0.3686,
+      "step": 11500
+    },
+    {
+      "epoch": 1.9637279167022275,
+      "grad_norm": 0.36006822237742797,
+      "learning_rate": 1.9186267071320184e-05,
+      "loss": 0.3591,
+      "step": 11505
+    },
+    {
+      "epoch": 1.9645813774857044,
+      "grad_norm": 0.3324090120282234,
+      "learning_rate": 1.9170460293373798e-05,
+      "loss": 0.357,
+      "step": 11510
+    },
+    {
+      "epoch": 1.9654348382691815,
+      "grad_norm": 0.3485251661481423,
+      "learning_rate": 1.9154653515427415e-05,
+      "loss": 0.3762,
+      "step": 11515
+    },
+    {
+      "epoch": 1.9662882990526587,
+      "grad_norm": 0.383412188275394,
+      "learning_rate": 1.9138846737481033e-05,
+      "loss": 0.3594,
+      "step": 11520
+    },
+    {
+      "epoch": 1.9671417598361356,
+      "grad_norm": 0.3666698926714686,
+      "learning_rate": 1.912303995953465e-05,
+      "loss": 0.3641,
+      "step": 11525
+    },
+    {
+      "epoch": 1.9679952206196125,
+      "grad_norm": 0.6887105354347631,
+      "learning_rate": 1.9107233181588268e-05,
+      "loss": 0.3577,
+      "step": 11530
+    },
+    {
+      "epoch": 1.9688486814030894,
+      "grad_norm": 0.3299880013438571,
+      "learning_rate": 1.909142640364188e-05,
+      "loss": 0.3584,
+      "step": 11535
+    },
+    {
+      "epoch": 1.9697021421865664,
+      "grad_norm": 0.3361477361924093,
+      "learning_rate": 1.90756196256955e-05,
+      "loss": 0.3472,
+      "step": 11540
+    },
+    {
+      "epoch": 1.9705556029700435,
+      "grad_norm": 0.619409623572859,
+      "learning_rate": 1.9059812847749113e-05,
+      "loss": 0.3783,
+      "step": 11545
+    },
+    {
+      "epoch": 1.9714090637535207,
+      "grad_norm": 0.3462415307913492,
+      "learning_rate": 1.9044006069802734e-05,
+      "loss": 0.3455,
+      "step": 11550
+    },
+    {
+      "epoch": 1.9722625245369976,
+      "grad_norm": 0.2879089288950731,
+      "learning_rate": 1.9028199291856348e-05,
+      "loss": 0.3663,
+      "step": 11555
+    },
+    {
+      "epoch": 1.9731159853204745,
+      "grad_norm": 0.3862808024430762,
+      "learning_rate": 1.9012392513909966e-05,
+      "loss": 0.3743,
+      "step": 11560
+    },
+    {
+      "epoch": 1.9739694461039514,
+      "grad_norm": 0.3952032121696526,
+      "learning_rate": 1.8996585735963583e-05,
+      "loss": 0.3604,
+      "step": 11565
+    },
+    {
+      "epoch": 1.9748229068874286,
+      "grad_norm": 0.2765527056826614,
+      "learning_rate": 1.8980778958017197e-05,
+      "loss": 0.3346,
+      "step": 11570
+    },
+    {
+      "epoch": 1.9756763676709055,
+      "grad_norm": 0.3262199224171904,
+      "learning_rate": 1.8964972180070814e-05,
+      "loss": 0.3548,
+      "step": 11575
+    },
+    {
+      "epoch": 1.9765298284543826,
+      "grad_norm": 0.3555939118146572,
+      "learning_rate": 1.8949165402124432e-05,
+      "loss": 0.3752,
+      "step": 11580
+    },
+    {
+      "epoch": 1.9773832892378596,
+      "grad_norm": 0.38534503417504223,
+      "learning_rate": 1.893335862417805e-05,
+      "loss": 0.3717,
+      "step": 11585
+    },
+    {
+      "epoch": 1.9782367500213365,
+      "grad_norm": 0.3525087826480515,
+      "learning_rate": 1.8917551846231667e-05,
+      "loss": 0.3435,
+      "step": 11590
+    },
+    {
+      "epoch": 1.9790902108048134,
+      "grad_norm": 0.41355689384936367,
+      "learning_rate": 1.890174506828528e-05,
+      "loss": 0.3452,
+      "step": 11595
+    },
+    {
+      "epoch": 1.9799436715882905,
+      "grad_norm": 0.32554251750894886,
+      "learning_rate": 1.88859382903389e-05,
+      "loss": 0.3674,
+      "step": 11600
+    },
+    {
+      "epoch": 1.9807971323717675,
+      "grad_norm": 0.38617312349391203,
+      "learning_rate": 1.8870131512392512e-05,
+      "loss": 0.3961,
+      "step": 11605
+    },
+    {
+      "epoch": 1.9816505931552446,
+      "grad_norm": 0.35810908236896655,
+      "learning_rate": 1.8854324734446133e-05,
+      "loss": 0.36,
+      "step": 11610
+    },
+    {
+      "epoch": 1.9825040539387215,
+      "grad_norm": 0.3959067679105283,
+      "learning_rate": 1.8838517956499747e-05,
+      "loss": 0.3812,
+      "step": 11615
+    },
+    {
+      "epoch": 1.9833575147221985,
+      "grad_norm": 0.44481170611894405,
+      "learning_rate": 1.8822711178553365e-05,
+      "loss": 0.3563,
+      "step": 11620
+    },
+    {
+      "epoch": 1.9842109755056754,
+      "grad_norm": 0.29757742501818696,
+      "learning_rate": 1.8806904400606982e-05,
+      "loss": 0.3575,
+      "step": 11625
+    },
+    {
+      "epoch": 1.9850644362891525,
+      "grad_norm": 0.34151394562783016,
+      "learning_rate": 1.8791097622660596e-05,
+      "loss": 0.3673,
+      "step": 11630
+    },
+    {
+      "epoch": 1.9859178970726297,
+      "grad_norm": 0.4115597667769673,
+      "learning_rate": 1.8775290844714214e-05,
+      "loss": 0.373,
+      "step": 11635
+    },
+    {
+      "epoch": 1.9867713578561066,
+      "grad_norm": 0.5589463365602205,
+      "learning_rate": 1.875948406676783e-05,
+      "loss": 0.3663,
+      "step": 11640
+    },
+    {
+      "epoch": 1.9876248186395835,
+      "grad_norm": 0.32174338662763924,
+      "learning_rate": 1.874367728882145e-05,
+      "loss": 0.3693,
+      "step": 11645
+    },
+    {
+      "epoch": 1.9884782794230604,
+      "grad_norm": 0.3992664638486375,
+      "learning_rate": 1.8727870510875066e-05,
+      "loss": 0.3649,
+      "step": 11650
+    },
+    {
+      "epoch": 1.9893317402065374,
+      "grad_norm": 0.3004563209664813,
+      "learning_rate": 1.871206373292868e-05,
+      "loss": 0.3669,
+      "step": 11655
+    },
+    {
+      "epoch": 1.9901852009900145,
+      "grad_norm": 0.2891256520749365,
+      "learning_rate": 1.8696256954982298e-05,
+      "loss": 0.3501,
+      "step": 11660
+    },
+    {
+      "epoch": 1.9910386617734916,
+      "grad_norm": 0.3523769019050455,
+      "learning_rate": 1.8680450177035912e-05,
+      "loss": 0.3661,
+      "step": 11665
+    },
+    {
+      "epoch": 1.9918921225569686,
+      "grad_norm": 0.3736500940750435,
+      "learning_rate": 1.866464339908953e-05,
+      "loss": 0.3867,
+      "step": 11670
+    },
+    {
+      "epoch": 1.9927455833404455,
+      "grad_norm": 0.35462316288841506,
+      "learning_rate": 1.8648836621143147e-05,
+      "loss": 0.3488,
+      "step": 11675
+    },
+    {
+      "epoch": 1.9935990441239224,
+      "grad_norm": 0.3031867758368206,
+      "learning_rate": 1.8633029843196764e-05,
+      "loss": 0.3447,
+      "step": 11680
+    },
+    {
+      "epoch": 1.9944525049073996,
+      "grad_norm": 0.3263771067325506,
+      "learning_rate": 1.861722306525038e-05,
+      "loss": 0.365,
+      "step": 11685
+    },
+    {
+      "epoch": 1.9953059656908765,
+      "grad_norm": 0.3174180681613223,
+      "learning_rate": 1.8601416287303996e-05,
+      "loss": 0.356,
+      "step": 11690
+    },
+    {
+      "epoch": 1.9961594264743536,
+      "grad_norm": 0.3758309298391433,
+      "learning_rate": 1.8585609509357613e-05,
+      "loss": 0.3618,
+      "step": 11695
+    },
+    {
+      "epoch": 1.9970128872578305,
+      "grad_norm": 0.37441528860296364,
+      "learning_rate": 1.8569802731411227e-05,
+      "loss": 0.3449,
+      "step": 11700
+    },
+    {
+      "epoch": 1.9978663480413075,
+      "grad_norm": 0.4118868612211099,
+      "learning_rate": 1.8553995953464848e-05,
+      "loss": 0.3535,
+      "step": 11705
+    },
+    {
+      "epoch": 1.9987198088247844,
+      "grad_norm": 1.8941609581505108,
+      "learning_rate": 1.8538189175518466e-05,
+      "loss": 0.3636,
+      "step": 11710
+    },
+    {
+      "epoch": 1.9995732696082615,
+      "grad_norm": 0.33030098032605626,
+      "learning_rate": 1.852238239757208e-05,
+      "loss": 0.3583,
+      "step": 11715
+    },
+    {
+      "epoch": 2.0003413843133906,
+      "grad_norm": 0.36601217529631147,
+      "learning_rate": 1.8506575619625697e-05,
+      "loss": 0.3355,
+      "step": 11720
+    },
+    {
+      "epoch": 2.001194845096868,
+      "grad_norm": 0.4087276302441495,
+      "learning_rate": 1.849076884167931e-05,
+      "loss": 0.274,
+      "step": 11725
+    },
+    {
+      "epoch": 2.002048305880345,
+      "grad_norm": 0.4409873629189802,
+      "learning_rate": 1.847496206373293e-05,
+      "loss": 0.2851,
+      "step": 11730
+    },
+    {
+      "epoch": 2.002901766663822,
+      "grad_norm": 0.3332216524255506,
+      "learning_rate": 1.8459155285786546e-05,
+      "loss": 0.2731,
+      "step": 11735
+    },
+    {
+      "epoch": 2.0037552274472987,
+      "grad_norm": 0.3616256277252148,
+      "learning_rate": 1.8443348507840163e-05,
+      "loss": 0.2801,
+      "step": 11740
+    },
+    {
+      "epoch": 2.0046086882307756,
+      "grad_norm": 0.3424298707047687,
+      "learning_rate": 1.842754172989378e-05,
+      "loss": 0.2736,
+      "step": 11745
+    },
+    {
+      "epoch": 2.005462149014253,
+      "grad_norm": 0.3445224253336677,
+      "learning_rate": 1.8411734951947395e-05,
+      "loss": 0.2557,
+      "step": 11750
+    },
+    {
+      "epoch": 2.00631560979773,
+      "grad_norm": 0.3351716189441447,
+      "learning_rate": 1.8395928174001012e-05,
+      "loss": 0.2615,
+      "step": 11755
+    },
+    {
+      "epoch": 2.007169070581207,
+      "grad_norm": 0.3332607228646462,
+      "learning_rate": 1.8380121396054627e-05,
+      "loss": 0.269,
+      "step": 11760
+    },
+    {
+      "epoch": 2.008022531364684,
+      "grad_norm": 0.36985606309509017,
+      "learning_rate": 1.8364314618108247e-05,
+      "loss": 0.269,
+      "step": 11765
+    },
+    {
+      "epoch": 2.0088759921481607,
+      "grad_norm": 0.34963039815133334,
+      "learning_rate": 1.8348507840161865e-05,
+      "loss": 0.2551,
+      "step": 11770
+    },
+    {
+      "epoch": 2.0097294529316376,
+      "grad_norm": 0.36421348478429594,
+      "learning_rate": 1.833270106221548e-05,
+      "loss": 0.2785,
+      "step": 11775
+    },
+    {
+      "epoch": 2.010582913715115,
+      "grad_norm": 0.3200502728315867,
+      "learning_rate": 1.8316894284269096e-05,
+      "loss": 0.2778,
+      "step": 11780
+    },
+    {
+      "epoch": 2.011436374498592,
+      "grad_norm": 0.3567813226557995,
+      "learning_rate": 1.830108750632271e-05,
+      "loss": 0.262,
+      "step": 11785
+    },
+    {
+      "epoch": 2.012289835282069,
+      "grad_norm": 0.38018272176428664,
+      "learning_rate": 1.8285280728376328e-05,
+      "loss": 0.2643,
+      "step": 11790
+    },
+    {
+      "epoch": 2.0131432960655458,
+      "grad_norm": 0.3196365171312658,
+      "learning_rate": 1.8269473950429945e-05,
+      "loss": 0.2939,
+      "step": 11795
+    },
+    {
+      "epoch": 2.0139967568490227,
+      "grad_norm": 0.3749386596391697,
+      "learning_rate": 1.8253667172483563e-05,
+      "loss": 0.2609,
+      "step": 11800
+    },
+    {
+      "epoch": 2.0148502176324996,
+      "grad_norm": 0.3320723067712335,
+      "learning_rate": 1.823786039453718e-05,
+      "loss": 0.2845,
+      "step": 11805
+    },
+    {
+      "epoch": 2.015703678415977,
+      "grad_norm": 0.3567384530855997,
+      "learning_rate": 1.8222053616590794e-05,
+      "loss": 0.2837,
+      "step": 11810
+    },
+    {
+      "epoch": 2.016557139199454,
+      "grad_norm": 0.5478809009656719,
+      "learning_rate": 1.8206246838644412e-05,
+      "loss": 0.2701,
+      "step": 11815
+    },
+    {
+      "epoch": 2.017410599982931,
+      "grad_norm": 0.4015324014536703,
+      "learning_rate": 1.8190440060698026e-05,
+      "loss": 0.2561,
+      "step": 11820
+    },
+    {
+      "epoch": 2.0182640607664077,
+      "grad_norm": 0.3082295737208152,
+      "learning_rate": 1.8174633282751643e-05,
+      "loss": 0.2556,
+      "step": 11825
+    },
+    {
+      "epoch": 2.0191175215498847,
+      "grad_norm": 0.35450120044324074,
+      "learning_rate": 1.8158826504805264e-05,
+      "loss": 0.2607,
+      "step": 11830
+    },
+    {
+      "epoch": 2.0199709823333616,
+      "grad_norm": 0.37204359376393903,
+      "learning_rate": 1.8143019726858878e-05,
+      "loss": 0.2509,
+      "step": 11835
+    },
+    {
+      "epoch": 2.020824443116839,
+      "grad_norm": 0.327493742049166,
+      "learning_rate": 1.8127212948912496e-05,
+      "loss": 0.2727,
+      "step": 11840
+    },
+    {
+      "epoch": 2.021677903900316,
+      "grad_norm": 0.34014078032855066,
+      "learning_rate": 1.811140617096611e-05,
+      "loss": 0.2891,
+      "step": 11845
+    },
+    {
+      "epoch": 2.022531364683793,
+      "grad_norm": 0.3714956426888765,
+      "learning_rate": 1.8095599393019727e-05,
+      "loss": 0.2714,
+      "step": 11850
+    },
+    {
+      "epoch": 2.0233848254672697,
+      "grad_norm": 0.34384771814885784,
+      "learning_rate": 1.8079792615073345e-05,
+      "loss": 0.2765,
+      "step": 11855
+    },
+    {
+      "epoch": 2.0242382862507466,
+      "grad_norm": 0.37971783255947067,
+      "learning_rate": 1.8063985837126962e-05,
+      "loss": 0.2569,
+      "step": 11860
+    },
+    {
+      "epoch": 2.0250917470342236,
+      "grad_norm": 0.35988065541812736,
+      "learning_rate": 1.804817905918058e-05,
+      "loss": 0.2596,
+      "step": 11865
+    },
+    {
+      "epoch": 2.025945207817701,
+      "grad_norm": 0.3249526782969596,
+      "learning_rate": 1.8032372281234194e-05,
+      "loss": 0.2762,
+      "step": 11870
+    },
+    {
+      "epoch": 2.026798668601178,
+      "grad_norm": 0.3514693210668101,
+      "learning_rate": 1.801656550328781e-05,
+      "loss": 0.2708,
+      "step": 11875
+    },
+    {
+      "epoch": 2.0276521293846548,
+      "grad_norm": 0.48699825557546833,
+      "learning_rate": 1.8000758725341425e-05,
+      "loss": 0.2978,
+      "step": 11880
+    },
+    {
+      "epoch": 2.0285055901681317,
+      "grad_norm": 0.3302433246221382,
+      "learning_rate": 1.7984951947395043e-05,
+      "loss": 0.2723,
+      "step": 11885
+    },
+    {
+      "epoch": 2.0293590509516086,
+      "grad_norm": 0.3494563980866781,
+      "learning_rate": 1.796914516944866e-05,
+      "loss": 0.2567,
+      "step": 11890
+    },
+    {
+      "epoch": 2.030212511735086,
+      "grad_norm": 0.5078171681533615,
+      "learning_rate": 1.7953338391502278e-05,
+      "loss": 0.3068,
+      "step": 11895
+    },
+    {
+      "epoch": 2.031065972518563,
+      "grad_norm": 0.33115962232712237,
+      "learning_rate": 1.7937531613555895e-05,
+      "loss": 0.2774,
+      "step": 11900
+    },
+    {
+      "epoch": 2.03191943330204,
+      "grad_norm": 0.3601496012552449,
+      "learning_rate": 1.792172483560951e-05,
+      "loss": 0.2873,
+      "step": 11905
+    },
+    {
+      "epoch": 2.0327728940855168,
+      "grad_norm": 0.2972739692797199,
+      "learning_rate": 1.7905918057663127e-05,
+      "loss": 0.2746,
+      "step": 11910
+    },
+    {
+      "epoch": 2.0336263548689937,
+      "grad_norm": 0.34598117202222756,
+      "learning_rate": 1.7890111279716744e-05,
+      "loss": 0.2788,
+      "step": 11915
+    },
+    {
+      "epoch": 2.0344798156524706,
+      "grad_norm": 0.2932483984070083,
+      "learning_rate": 1.7874304501770358e-05,
+      "loss": 0.2541,
+      "step": 11920
+    },
+    {
+      "epoch": 2.035333276435948,
+      "grad_norm": 0.2863647603035123,
+      "learning_rate": 1.785849772382398e-05,
+      "loss": 0.2627,
+      "step": 11925
+    },
+    {
+      "epoch": 2.036186737219425,
+      "grad_norm": 0.37976530277574055,
+      "learning_rate": 1.7842690945877593e-05,
+      "loss": 0.293,
+      "step": 11930
+    },
+    {
+      "epoch": 2.037040198002902,
+      "grad_norm": 0.31537747743198213,
+      "learning_rate": 1.782688416793121e-05,
+      "loss": 0.2723,
+      "step": 11935
+    },
+    {
+      "epoch": 2.0378936587863787,
+      "grad_norm": 0.3300382729548969,
+      "learning_rate": 1.7811077389984825e-05,
+      "loss": 0.2643,
+      "step": 11940
+    },
+    {
+      "epoch": 2.0387471195698557,
+      "grad_norm": 0.447352553052991,
+      "learning_rate": 1.7795270612038442e-05,
+      "loss": 0.2677,
+      "step": 11945
+    },
+    {
+      "epoch": 2.0396005803533326,
+      "grad_norm": 0.36568175470105174,
+      "learning_rate": 1.777946383409206e-05,
+      "loss": 0.2759,
+      "step": 11950
+    },
+    {
+      "epoch": 2.04045404113681,
+      "grad_norm": 0.34674038246091743,
+      "learning_rate": 1.7763657056145677e-05,
+      "loss": 0.2775,
+      "step": 11955
+    },
+    {
+      "epoch": 2.041307501920287,
+      "grad_norm": 0.3099595102646737,
+      "learning_rate": 1.7747850278199294e-05,
+      "loss": 0.2406,
+      "step": 11960
+    },
+    {
+      "epoch": 2.042160962703764,
+      "grad_norm": 0.34062152209947577,
+      "learning_rate": 1.773204350025291e-05,
+      "loss": 0.2546,
+      "step": 11965
+    },
+    {
+      "epoch": 2.0430144234872407,
+      "grad_norm": 0.3212507029051799,
+      "learning_rate": 1.7716236722306526e-05,
+      "loss": 0.2781,
+      "step": 11970
+    },
+    {
+      "epoch": 2.0438678842707176,
+      "grad_norm": 0.3669402570578361,
+      "learning_rate": 1.7700429944360143e-05,
+      "loss": 0.2508,
+      "step": 11975
+    },
+    {
+      "epoch": 2.0447213450541946,
+      "grad_norm": 0.3256430454535454,
+      "learning_rate": 1.7684623166413757e-05,
+      "loss": 0.2707,
+      "step": 11980
+    },
+    {
+      "epoch": 2.045574805837672,
+      "grad_norm": 0.3644091779560888,
+      "learning_rate": 1.7668816388467378e-05,
+      "loss": 0.2611,
+      "step": 11985
+    },
+    {
+      "epoch": 2.046428266621149,
+      "grad_norm": 0.29339094551071926,
+      "learning_rate": 1.7653009610520992e-05,
+      "loss": 0.277,
+      "step": 11990
+    },
+    {
+      "epoch": 2.0472817274046258,
+      "grad_norm": 0.32224421179142887,
+      "learning_rate": 1.763720283257461e-05,
+      "loss": 0.2854,
+      "step": 11995
+    },
+    {
+      "epoch": 2.0481351881881027,
+      "grad_norm": 0.6105808733353931,
+      "learning_rate": 1.7621396054628224e-05,
+      "loss": 0.29,
+      "step": 12000
+    },
+    {
+      "epoch": 2.0489886489715796,
+      "grad_norm": 0.6269079828473192,
+      "learning_rate": 1.760558927668184e-05,
+      "loss": 0.2836,
+      "step": 12005
+    },
+    {
+      "epoch": 2.0498421097550565,
+      "grad_norm": 0.33382413769115177,
+      "learning_rate": 1.758978249873546e-05,
+      "loss": 0.2804,
+      "step": 12010
+    },
+    {
+      "epoch": 2.050695570538534,
+      "grad_norm": 0.3833260788413062,
+      "learning_rate": 1.7573975720789076e-05,
+      "loss": 0.2761,
+      "step": 12015
+    },
+    {
+      "epoch": 2.051549031322011,
+      "grad_norm": 0.7526300116993044,
+      "learning_rate": 1.7558168942842694e-05,
+      "loss": 0.2675,
+      "step": 12020
+    },
+    {
+      "epoch": 2.0524024921054878,
+      "grad_norm": 0.4095339100076474,
+      "learning_rate": 1.7542362164896308e-05,
+      "loss": 0.2531,
+      "step": 12025
+    },
+    {
+      "epoch": 2.0532559528889647,
+      "grad_norm": 0.3324065066305095,
+      "learning_rate": 1.7526555386949925e-05,
+      "loss": 0.2664,
+      "step": 12030
+    },
+    {
+      "epoch": 2.0541094136724416,
+      "grad_norm": 0.31380277400476436,
+      "learning_rate": 1.7510748609003543e-05,
+      "loss": 0.2596,
+      "step": 12035
+    },
+    {
+      "epoch": 2.054962874455919,
+      "grad_norm": 0.4274697022704499,
+      "learning_rate": 1.7494941831057157e-05,
+      "loss": 0.2463,
+      "step": 12040
+    },
+    {
+      "epoch": 2.055816335239396,
+      "grad_norm": 0.35626841520038793,
+      "learning_rate": 1.7479135053110774e-05,
+      "loss": 0.2719,
+      "step": 12045
+    },
+    {
+      "epoch": 2.056669796022873,
+      "grad_norm": 0.3974235988514715,
+      "learning_rate": 1.746332827516439e-05,
+      "loss": 0.2709,
+      "step": 12050
+    },
+    {
+      "epoch": 2.0575232568063497,
+      "grad_norm": 0.30990091059678676,
+      "learning_rate": 1.744752149721801e-05,
+      "loss": 0.2596,
+      "step": 12055
+    },
+    {
+      "epoch": 2.0583767175898267,
+      "grad_norm": 0.33322796671815513,
+      "learning_rate": 1.7431714719271623e-05,
+      "loss": 0.2913,
+      "step": 12060
+    },
+    {
+      "epoch": 2.0592301783733036,
+      "grad_norm": 0.3171079041098653,
+      "learning_rate": 1.741590794132524e-05,
+      "loss": 0.2807,
+      "step": 12065
+    },
+    {
+      "epoch": 2.060083639156781,
+      "grad_norm": 0.3588034806290993,
+      "learning_rate": 1.7400101163378858e-05,
+      "loss": 0.2531,
+      "step": 12070
+    },
+    {
+      "epoch": 2.060937099940258,
+      "grad_norm": 0.33898417082895527,
+      "learning_rate": 1.7384294385432472e-05,
+      "loss": 0.2732,
+      "step": 12075
+    },
+    {
+      "epoch": 2.061790560723735,
+      "grad_norm": 0.3092639727199814,
+      "learning_rate": 1.7368487607486093e-05,
+      "loss": 0.2898,
+      "step": 12080
+    },
+    {
+      "epoch": 2.0626440215072117,
+      "grad_norm": 0.31338268131353825,
+      "learning_rate": 1.7352680829539707e-05,
+      "loss": 0.3038,
+      "step": 12085
+    },
+    {
+      "epoch": 2.0634974822906886,
+      "grad_norm": 0.29827852349198464,
+      "learning_rate": 1.7336874051593325e-05,
+      "loss": 0.2584,
+      "step": 12090
+    },
+    {
+      "epoch": 2.0643509430741656,
+      "grad_norm": 0.3657097608792625,
+      "learning_rate": 1.7321067273646942e-05,
+      "loss": 0.2759,
+      "step": 12095
+    },
+    {
+      "epoch": 2.065204403857643,
+      "grad_norm": 0.3765895570517396,
+      "learning_rate": 1.7305260495700556e-05,
+      "loss": 0.2534,
+      "step": 12100
+    },
+    {
+      "epoch": 2.06605786464112,
+      "grad_norm": 0.4322246859086725,
+      "learning_rate": 1.7289453717754174e-05,
+      "loss": 0.2672,
+      "step": 12105
+    },
+    {
+      "epoch": 2.0669113254245968,
+      "grad_norm": 0.3118296114497835,
+      "learning_rate": 1.727364693980779e-05,
+      "loss": 0.2722,
+      "step": 12110
+    },
+    {
+      "epoch": 2.0677647862080737,
+      "grad_norm": 0.31531418590925914,
+      "learning_rate": 1.725784016186141e-05,
+      "loss": 0.2729,
+      "step": 12115
+    },
+    {
+      "epoch": 2.0686182469915506,
+      "grad_norm": 0.29201923605182517,
+      "learning_rate": 1.7242033383915023e-05,
+      "loss": 0.2588,
+      "step": 12120
+    },
+    {
+      "epoch": 2.0694717077750275,
+      "grad_norm": 0.3403082569544838,
+      "learning_rate": 1.722622660596864e-05,
+      "loss": 0.2729,
+      "step": 12125
+    },
+    {
+      "epoch": 2.070325168558505,
+      "grad_norm": 0.35162950149295324,
+      "learning_rate": 1.7210419828022257e-05,
+      "loss": 0.2631,
+      "step": 12130
+    },
+    {
+      "epoch": 2.071178629341982,
+      "grad_norm": 0.4200351820771278,
+      "learning_rate": 1.719461305007587e-05,
+      "loss": 0.2719,
+      "step": 12135
+    },
+    {
+      "epoch": 2.0720320901254587,
+      "grad_norm": 0.37018489015593226,
+      "learning_rate": 1.7178806272129492e-05,
+      "loss": 0.2532,
+      "step": 12140
+    },
+    {
+      "epoch": 2.0728855509089357,
+      "grad_norm": 0.2948475725320786,
+      "learning_rate": 1.7162999494183106e-05,
+      "loss": 0.2305,
+      "step": 12145
+    },
+    {
+      "epoch": 2.0737390116924126,
+      "grad_norm": 0.31880448697394564,
+      "learning_rate": 1.7147192716236724e-05,
+      "loss": 0.2916,
+      "step": 12150
+    },
+    {
+      "epoch": 2.0745924724758895,
+      "grad_norm": 0.30086387149635874,
+      "learning_rate": 1.713138593829034e-05,
+      "loss": 0.2728,
+      "step": 12155
+    },
+    {
+      "epoch": 2.075445933259367,
+      "grad_norm": 0.35225846509084374,
+      "learning_rate": 1.7115579160343955e-05,
+      "loss": 0.2755,
+      "step": 12160
+    },
+    {
+      "epoch": 2.076299394042844,
+      "grad_norm": 0.34709944866279024,
+      "learning_rate": 1.7099772382397573e-05,
+      "loss": 0.2615,
+      "step": 12165
+    },
+    {
+      "epoch": 2.0771528548263207,
+      "grad_norm": 0.333291664285541,
+      "learning_rate": 1.708396560445119e-05,
+      "loss": 0.2624,
+      "step": 12170
+    },
+    {
+      "epoch": 2.0780063156097976,
+      "grad_norm": 0.3360796911333201,
+      "learning_rate": 1.7068158826504808e-05,
+      "loss": 0.2841,
+      "step": 12175
+    },
+    {
+      "epoch": 2.0788597763932746,
+      "grad_norm": 0.40102339611112403,
+      "learning_rate": 1.7052352048558422e-05,
+      "loss": 0.2472,
+      "step": 12180
+    },
+    {
+      "epoch": 2.079713237176752,
+      "grad_norm": 0.3524313454573767,
+      "learning_rate": 1.703654527061204e-05,
+      "loss": 0.2738,
+      "step": 12185
+    },
+    {
+      "epoch": 2.080566697960229,
+      "grad_norm": 0.38390466639979504,
+      "learning_rate": 1.7020738492665657e-05,
+      "loss": 0.2764,
+      "step": 12190
+    },
+    {
+      "epoch": 2.081420158743706,
+      "grad_norm": 0.3616939115523455,
+      "learning_rate": 1.700493171471927e-05,
+      "loss": 0.2676,
+      "step": 12195
+    },
+    {
+      "epoch": 2.0822736195271827,
+      "grad_norm": 0.34575626616021204,
+      "learning_rate": 1.6989124936772888e-05,
+      "loss": 0.2717,
+      "step": 12200
+    },
+    {
+      "epoch": 2.0831270803106596,
+      "grad_norm": 0.5088527280069074,
+      "learning_rate": 1.6973318158826506e-05,
+      "loss": 0.2601,
+      "step": 12205
+    },
+    {
+      "epoch": 2.0839805410941366,
+      "grad_norm": 0.3481856832523284,
+      "learning_rate": 1.6957511380880123e-05,
+      "loss": 0.2659,
+      "step": 12210
+    },
+    {
+      "epoch": 2.084834001877614,
+      "grad_norm": 0.3584208310807397,
+      "learning_rate": 1.694170460293374e-05,
+      "loss": 0.2682,
+      "step": 12215
+    },
+    {
+      "epoch": 2.085687462661091,
+      "grad_norm": 0.33623978474962535,
+      "learning_rate": 1.6925897824987355e-05,
+      "loss": 0.274,
+      "step": 12220
+    },
+    {
+      "epoch": 2.0865409234445678,
+      "grad_norm": 0.3430855451783835,
+      "learning_rate": 1.6910091047040972e-05,
+      "loss": 0.2723,
+      "step": 12225
+    },
+    {
+      "epoch": 2.0873943842280447,
+      "grad_norm": 0.3239022405078139,
+      "learning_rate": 1.6894284269094586e-05,
+      "loss": 0.2657,
+      "step": 12230
+    },
+    {
+      "epoch": 2.0882478450115216,
+      "grad_norm": 0.43544523014277337,
+      "learning_rate": 1.6878477491148207e-05,
+      "loss": 0.2632,
+      "step": 12235
+    },
+    {
+      "epoch": 2.0891013057949985,
+      "grad_norm": 0.4528751490778456,
+      "learning_rate": 1.686267071320182e-05,
+      "loss": 0.2576,
+      "step": 12240
+    },
+    {
+      "epoch": 2.089954766578476,
+      "grad_norm": 0.3585745420304474,
+      "learning_rate": 1.684686393525544e-05,
+      "loss": 0.2739,
+      "step": 12245
+    },
+    {
+      "epoch": 2.090808227361953,
+      "grad_norm": 0.311146872238834,
+      "learning_rate": 1.6831057157309056e-05,
+      "loss": 0.2575,
+      "step": 12250
+    },
+    {
+      "epoch": 2.0916616881454297,
+      "grad_norm": 0.3552959349700159,
+      "learning_rate": 1.681525037936267e-05,
+      "loss": 0.2805,
+      "step": 12255
+    },
+    {
+      "epoch": 2.0925151489289067,
+      "grad_norm": 0.32835550560208493,
+      "learning_rate": 1.6799443601416288e-05,
+      "loss": 0.2516,
+      "step": 12260
+    },
+    {
+      "epoch": 2.0933686097123836,
+      "grad_norm": 0.3347494591407445,
+      "learning_rate": 1.6783636823469905e-05,
+      "loss": 0.2804,
+      "step": 12265
+    },
+    {
+      "epoch": 2.0942220704958605,
+      "grad_norm": 0.6363722617507217,
+      "learning_rate": 1.6767830045523523e-05,
+      "loss": 0.275,
+      "step": 12270
+    },
+    {
+      "epoch": 2.095075531279338,
+      "grad_norm": 0.34387693365627625,
+      "learning_rate": 1.6752023267577137e-05,
+      "loss": 0.2639,
+      "step": 12275
+    },
+    {
+      "epoch": 2.095928992062815,
+      "grad_norm": 0.35194711797385764,
+      "learning_rate": 1.6736216489630754e-05,
+      "loss": 0.2815,
+      "step": 12280
+    },
+    {
+      "epoch": 2.0967824528462917,
+      "grad_norm": 0.3278093489773815,
+      "learning_rate": 1.672040971168437e-05,
+      "loss": 0.3025,
+      "step": 12285
+    },
+    {
+      "epoch": 2.0976359136297686,
+      "grad_norm": 0.32201858704326325,
+      "learning_rate": 1.6704602933737986e-05,
+      "loss": 0.26,
+      "step": 12290
+    },
+    {
+      "epoch": 2.0984893744132456,
+      "grad_norm": 0.43088177963709523,
+      "learning_rate": 1.6688796155791606e-05,
+      "loss": 0.2664,
+      "step": 12295
+    },
+    {
+      "epoch": 2.0993428351967225,
+      "grad_norm": 0.35898411380628825,
+      "learning_rate": 1.667298937784522e-05,
+      "loss": 0.2829,
+      "step": 12300
+    },
+    {
+      "epoch": 2.1001962959802,
+      "grad_norm": 0.36341604943601946,
+      "learning_rate": 1.6657182599898838e-05,
+      "loss": 0.2787,
+      "step": 12305
+    },
+    {
+      "epoch": 2.101049756763677,
+      "grad_norm": 0.3801967985876427,
+      "learning_rate": 1.6641375821952455e-05,
+      "loss": 0.2525,
+      "step": 12310
+    },
+    {
+      "epoch": 2.1019032175471537,
+      "grad_norm": 0.3484349896151566,
+      "learning_rate": 1.662556904400607e-05,
+      "loss": 0.2703,
+      "step": 12315
+    },
+    {
+      "epoch": 2.1027566783306306,
+      "grad_norm": 0.6689341642203409,
+      "learning_rate": 1.6609762266059687e-05,
+      "loss": 0.2767,
+      "step": 12320
+    },
+    {
+      "epoch": 2.1036101391141075,
+      "grad_norm": 0.42341943337546845,
+      "learning_rate": 1.65939554881133e-05,
+      "loss": 0.2605,
+      "step": 12325
+    },
+    {
+      "epoch": 2.104463599897585,
+      "grad_norm": 0.32348121955528175,
+      "learning_rate": 1.6578148710166922e-05,
+      "loss": 0.2735,
+      "step": 12330
+    },
+    {
+      "epoch": 2.105317060681062,
+      "grad_norm": 0.3205700673140748,
+      "learning_rate": 1.6562341932220536e-05,
+      "loss": 0.2685,
+      "step": 12335
+    },
+    {
+      "epoch": 2.1061705214645388,
+      "grad_norm": 0.34986014628067635,
+      "learning_rate": 1.6546535154274153e-05,
+      "loss": 0.2567,
+      "step": 12340
+    },
+    {
+      "epoch": 2.1070239822480157,
+      "grad_norm": 0.43101180839913,
+      "learning_rate": 1.653072837632777e-05,
+      "loss": 0.2679,
+      "step": 12345
+    },
+    {
+      "epoch": 2.1078774430314926,
+      "grad_norm": 0.3627838318748922,
+      "learning_rate": 1.6514921598381385e-05,
+      "loss": 0.2455,
+      "step": 12350
+    },
+    {
+      "epoch": 2.1087309038149695,
+      "grad_norm": 0.323420809694066,
+      "learning_rate": 1.6499114820435002e-05,
+      "loss": 0.2886,
+      "step": 12355
+    },
+    {
+      "epoch": 2.109584364598447,
+      "grad_norm": 0.3552575219189849,
+      "learning_rate": 1.648330804248862e-05,
+      "loss": 0.2668,
+      "step": 12360
+    },
+    {
+      "epoch": 2.110437825381924,
+      "grad_norm": 0.35622871312938437,
+      "learning_rate": 1.6467501264542237e-05,
+      "loss": 0.2722,
+      "step": 12365
+    },
+    {
+      "epoch": 2.1112912861654007,
+      "grad_norm": 0.3758135642451043,
+      "learning_rate": 1.6451694486595855e-05,
+      "loss": 0.29,
+      "step": 12370
+    },
+    {
+      "epoch": 2.1121447469488777,
+      "grad_norm": 0.5326988667885914,
+      "learning_rate": 1.643588770864947e-05,
+      "loss": 0.2707,
+      "step": 12375
+    },
+    {
+      "epoch": 2.1129982077323546,
+      "grad_norm": 1.3780005011298029,
+      "learning_rate": 1.6420080930703086e-05,
+      "loss": 0.2974,
+      "step": 12380
+    },
+    {
+      "epoch": 2.1138516685158315,
+      "grad_norm": 0.3316592274921884,
+      "learning_rate": 1.64042741527567e-05,
+      "loss": 0.2607,
+      "step": 12385
+    },
+    {
+      "epoch": 2.114705129299309,
+      "grad_norm": 0.32763820515024833,
+      "learning_rate": 1.638846737481032e-05,
+      "loss": 0.2716,
+      "step": 12390
+    },
+    {
+      "epoch": 2.115558590082786,
+      "grad_norm": 0.37938700687076043,
+      "learning_rate": 1.6372660596863935e-05,
+      "loss": 0.2753,
+      "step": 12395
+    },
+    {
+      "epoch": 2.1164120508662627,
+      "grad_norm": 0.311938748878213,
+      "learning_rate": 1.6356853818917553e-05,
+      "loss": 0.2682,
+      "step": 12400
+    },
+    {
+      "epoch": 2.1172655116497396,
+      "grad_norm": 0.3436823316900887,
+      "learning_rate": 1.634104704097117e-05,
+      "loss": 0.2773,
+      "step": 12405
+    },
+    {
+      "epoch": 2.1181189724332166,
+      "grad_norm": 0.3693852853641911,
+      "learning_rate": 1.6325240263024784e-05,
+      "loss": 0.2697,
+      "step": 12410
+    },
+    {
+      "epoch": 2.1189724332166935,
+      "grad_norm": 0.30534991718254095,
+      "learning_rate": 1.63094334850784e-05,
+      "loss": 0.2944,
+      "step": 12415
+    },
+    {
+      "epoch": 2.119825894000171,
+      "grad_norm": 0.46673971087004035,
+      "learning_rate": 1.629362670713202e-05,
+      "loss": 0.2775,
+      "step": 12420
+    },
+    {
+      "epoch": 2.1206793547836478,
+      "grad_norm": 0.40866693342916965,
+      "learning_rate": 1.6277819929185637e-05,
+      "loss": 0.2769,
+      "step": 12425
+    },
+    {
+      "epoch": 2.1215328155671247,
+      "grad_norm": 0.3348396839675023,
+      "learning_rate": 1.6262013151239254e-05,
+      "loss": 0.2585,
+      "step": 12430
+    },
+    {
+      "epoch": 2.1223862763506016,
+      "grad_norm": 0.3458705634911477,
+      "learning_rate": 1.6246206373292868e-05,
+      "loss": 0.2822,
+      "step": 12435
+    },
+    {
+      "epoch": 2.1232397371340785,
+      "grad_norm": 0.34401778783058806,
+      "learning_rate": 1.6230399595346486e-05,
+      "loss": 0.277,
+      "step": 12440
+    },
+    {
+      "epoch": 2.1240931979175555,
+      "grad_norm": 0.3155670129755863,
+      "learning_rate": 1.62145928174001e-05,
+      "loss": 0.2474,
+      "step": 12445
+    },
+    {
+      "epoch": 2.124946658701033,
+      "grad_norm": 0.4164779372341016,
+      "learning_rate": 1.6198786039453717e-05,
+      "loss": 0.2674,
+      "step": 12450
+    },
+    {
+      "epoch": 2.1258001194845098,
+      "grad_norm": 0.30873321446598606,
+      "learning_rate": 1.6182979261507335e-05,
+      "loss": 0.2356,
+      "step": 12455
+    },
+    {
+      "epoch": 2.1266535802679867,
+      "grad_norm": 0.33359895490839675,
+      "learning_rate": 1.6167172483560952e-05,
+      "loss": 0.2396,
+      "step": 12460
+    },
+    {
+      "epoch": 2.1275070410514636,
+      "grad_norm": 0.34527339376243676,
+      "learning_rate": 1.615136570561457e-05,
+      "loss": 0.2721,
+      "step": 12465
+    },
+    {
+      "epoch": 2.1283605018349405,
+      "grad_norm": 0.3467447095430177,
+      "learning_rate": 1.6135558927668184e-05,
+      "loss": 0.261,
+      "step": 12470
+    },
+    {
+      "epoch": 2.129213962618418,
+      "grad_norm": 0.3336284624075867,
+      "learning_rate": 1.61197521497218e-05,
+      "loss": 0.2873,
+      "step": 12475
+    },
+    {
+      "epoch": 2.130067423401895,
+      "grad_norm": 0.3427654798703003,
+      "learning_rate": 1.610394537177542e-05,
+      "loss": 0.2747,
+      "step": 12480
+    },
+    {
+      "epoch": 2.1309208841853717,
+      "grad_norm": 0.2954348761039776,
+      "learning_rate": 1.6088138593829036e-05,
+      "loss": 0.2646,
+      "step": 12485
+    },
+    {
+      "epoch": 2.1317743449688487,
+      "grad_norm": 0.3508140722897979,
+      "learning_rate": 1.6072331815882653e-05,
+      "loss": 0.2632,
+      "step": 12490
+    },
+    {
+      "epoch": 2.1326278057523256,
+      "grad_norm": 0.3202599343544827,
+      "learning_rate": 1.6056525037936267e-05,
+      "loss": 0.2658,
+      "step": 12495
+    },
+    {
+      "epoch": 2.1334812665358025,
+      "grad_norm": 0.3465604693790608,
+      "learning_rate": 1.6040718259989885e-05,
+      "loss": 0.244,
+      "step": 12500
+    },
+    {
+      "epoch": 2.13433472731928,
+      "grad_norm": 0.36359975874338335,
+      "learning_rate": 1.60249114820435e-05,
+      "loss": 0.2608,
+      "step": 12505
+    },
+    {
+      "epoch": 2.135188188102757,
+      "grad_norm": 0.308184805668463,
+      "learning_rate": 1.6009104704097116e-05,
+      "loss": 0.2698,
+      "step": 12510
+    },
+    {
+      "epoch": 2.1360416488862337,
+      "grad_norm": 0.3247399308382604,
+      "learning_rate": 1.5993297926150734e-05,
+      "loss": 0.2738,
+      "step": 12515
+    },
+    {
+      "epoch": 2.1368951096697106,
+      "grad_norm": 0.3907883620181,
+      "learning_rate": 1.597749114820435e-05,
+      "loss": 0.2655,
+      "step": 12520
+    },
+    {
+      "epoch": 2.1377485704531876,
+      "grad_norm": 0.33844432676774516,
+      "learning_rate": 1.596168437025797e-05,
+      "loss": 0.2627,
+      "step": 12525
+    },
+    {
+      "epoch": 2.1386020312366645,
+      "grad_norm": 0.3170816006943378,
+      "learning_rate": 1.5945877592311583e-05,
+      "loss": 0.2641,
+      "step": 12530
+    },
+    {
+      "epoch": 2.139455492020142,
+      "grad_norm": 0.3397399340019784,
+      "learning_rate": 1.59300708143652e-05,
+      "loss": 0.2834,
+      "step": 12535
+    },
+    {
+      "epoch": 2.1403089528036188,
+      "grad_norm": 0.31757903497442197,
+      "learning_rate": 1.5914264036418818e-05,
+      "loss": 0.2898,
+      "step": 12540
+    },
+    {
+      "epoch": 2.1411624135870957,
+      "grad_norm": 0.4508104912347626,
+      "learning_rate": 1.5898457258472435e-05,
+      "loss": 0.2583,
+      "step": 12545
+    },
+    {
+      "epoch": 2.1420158743705726,
+      "grad_norm": 0.3312541425875485,
+      "learning_rate": 1.5882650480526053e-05,
+      "loss": 0.2638,
+      "step": 12550
+    },
+    {
+      "epoch": 2.1428693351540495,
+      "grad_norm": 0.3454086840727479,
+      "learning_rate": 1.5866843702579667e-05,
+      "loss": 0.2949,
+      "step": 12555
+    },
+    {
+      "epoch": 2.143722795937527,
+      "grad_norm": 0.35233563685593,
+      "learning_rate": 1.5851036924633284e-05,
+      "loss": 0.2789,
+      "step": 12560
+    },
+    {
+      "epoch": 2.144576256721004,
+      "grad_norm": 0.35464891316951724,
+      "learning_rate": 1.5835230146686898e-05,
+      "loss": 0.2554,
+      "step": 12565
+    },
+    {
+      "epoch": 2.1454297175044807,
+      "grad_norm": 0.3423594887215211,
+      "learning_rate": 1.5819423368740516e-05,
+      "loss": 0.2672,
+      "step": 12570
+    },
+    {
+      "epoch": 2.1462831782879577,
+      "grad_norm": 0.364168129489653,
+      "learning_rate": 1.5803616590794133e-05,
+      "loss": 0.2628,
+      "step": 12575
+    },
+    {
+      "epoch": 2.1471366390714346,
+      "grad_norm": 0.3512621654880636,
+      "learning_rate": 1.578780981284775e-05,
+      "loss": 0.2714,
+      "step": 12580
+    },
+    {
+      "epoch": 2.1479900998549115,
+      "grad_norm": 0.32118410562314237,
+      "learning_rate": 1.5772003034901368e-05,
+      "loss": 0.2814,
+      "step": 12585
+    },
+    {
+      "epoch": 2.1488435606383884,
+      "grad_norm": 0.32706659710720487,
+      "learning_rate": 1.5756196256954982e-05,
+      "loss": 0.2612,
+      "step": 12590
+    },
+    {
+      "epoch": 2.149697021421866,
+      "grad_norm": 0.3491312081639566,
+      "learning_rate": 1.57403894790086e-05,
+      "loss": 0.2349,
+      "step": 12595
+    },
+    {
+      "epoch": 2.1505504822053427,
+      "grad_norm": 0.4315399093375859,
+      "learning_rate": 1.5724582701062217e-05,
+      "loss": 0.2883,
+      "step": 12600
+    },
+    {
+      "epoch": 2.1514039429888197,
+      "grad_norm": 0.31659750259238256,
+      "learning_rate": 1.570877592311583e-05,
+      "loss": 0.2831,
+      "step": 12605
+    },
+    {
+      "epoch": 2.1522574037722966,
+      "grad_norm": 0.35447492593381486,
+      "learning_rate": 1.5692969145169452e-05,
+      "loss": 0.2772,
+      "step": 12610
+    },
+    {
+      "epoch": 2.1531108645557735,
+      "grad_norm": 0.3252605965205536,
+      "learning_rate": 1.5677162367223066e-05,
+      "loss": 0.2711,
+      "step": 12615
+    },
+    {
+      "epoch": 2.153964325339251,
+      "grad_norm": 0.33529255229945437,
+      "learning_rate": 1.5661355589276684e-05,
+      "loss": 0.2564,
+      "step": 12620
+    },
+    {
+      "epoch": 2.154817786122728,
+      "grad_norm": 0.3092992567695559,
+      "learning_rate": 1.5645548811330298e-05,
+      "loss": 0.2872,
+      "step": 12625
+    },
+    {
+      "epoch": 2.1556712469062047,
+      "grad_norm": 0.36336461468734477,
+      "learning_rate": 1.5629742033383915e-05,
+      "loss": 0.2922,
+      "step": 12630
+    },
+    {
+      "epoch": 2.1565247076896816,
+      "grad_norm": 0.3432044815496815,
+      "learning_rate": 1.5613935255437533e-05,
+      "loss": 0.2567,
+      "step": 12635
+    },
+    {
+      "epoch": 2.1573781684731586,
+      "grad_norm": 0.4069346298755619,
+      "learning_rate": 1.559812847749115e-05,
+      "loss": 0.2445,
+      "step": 12640
+    },
+    {
+      "epoch": 2.1582316292566355,
+      "grad_norm": 0.40033171137575607,
+      "learning_rate": 1.5582321699544767e-05,
+      "loss": 0.2596,
+      "step": 12645
+    },
+    {
+      "epoch": 2.159085090040113,
+      "grad_norm": 0.2854063154369835,
+      "learning_rate": 1.556651492159838e-05,
+      "loss": 0.2698,
+      "step": 12650
+    },
+    {
+      "epoch": 2.1599385508235898,
+      "grad_norm": 0.4636509583084091,
+      "learning_rate": 1.5550708143652e-05,
+      "loss": 0.2866,
+      "step": 12655
+    },
+    {
+      "epoch": 2.1607920116070667,
+      "grad_norm": 0.34682881592035003,
+      "learning_rate": 1.5534901365705613e-05,
+      "loss": 0.2823,
+      "step": 12660
+    },
+    {
+      "epoch": 2.1616454723905436,
+      "grad_norm": 0.3290308156842946,
+      "learning_rate": 1.551909458775923e-05,
+      "loss": 0.2759,
+      "step": 12665
+    },
+    {
+      "epoch": 2.1624989331740205,
+      "grad_norm": 0.36783948757997753,
+      "learning_rate": 1.550328780981285e-05,
+      "loss": 0.2405,
+      "step": 12670
+    },
+    {
+      "epoch": 2.1633523939574975,
+      "grad_norm": 0.35236119660393694,
+      "learning_rate": 1.5487481031866465e-05,
+      "loss": 0.2803,
+      "step": 12675
+    },
+    {
+      "epoch": 2.164205854740975,
+      "grad_norm": 0.3754353607264287,
+      "learning_rate": 1.5471674253920083e-05,
+      "loss": 0.2575,
+      "step": 12680
+    },
+    {
+      "epoch": 2.1650593155244517,
+      "grad_norm": 0.3861606015120469,
+      "learning_rate": 1.5455867475973697e-05,
+      "loss": 0.2592,
+      "step": 12685
+    },
+    {
+      "epoch": 2.1659127763079287,
+      "grad_norm": 0.3675140543631446,
+      "learning_rate": 1.5440060698027314e-05,
+      "loss": 0.2601,
+      "step": 12690
+    },
+    {
+      "epoch": 2.1667662370914056,
+      "grad_norm": 0.32619482564351854,
+      "learning_rate": 1.5424253920080932e-05,
+      "loss": 0.2616,
+      "step": 12695
+    },
+    {
+      "epoch": 2.1676196978748825,
+      "grad_norm": 0.3884753490924257,
+      "learning_rate": 1.5408447142134546e-05,
+      "loss": 0.2773,
+      "step": 12700
+    },
+    {
+      "epoch": 2.16847315865836,
+      "grad_norm": 0.34120808204855113,
+      "learning_rate": 1.5392640364188167e-05,
+      "loss": 0.2652,
+      "step": 12705
+    },
+    {
+      "epoch": 2.169326619441837,
+      "grad_norm": 0.37573300008074934,
+      "learning_rate": 1.537683358624178e-05,
+      "loss": 0.2796,
+      "step": 12710
+    },
+    {
+      "epoch": 2.1701800802253137,
+      "grad_norm": 0.32540937938049824,
+      "learning_rate": 1.53610268082954e-05,
+      "loss": 0.3074,
+      "step": 12715
+    },
+    {
+      "epoch": 2.1710335410087906,
+      "grad_norm": 0.3082161890237831,
+      "learning_rate": 1.5345220030349012e-05,
+      "loss": 0.2566,
+      "step": 12720
+    },
+    {
+      "epoch": 2.1718870017922676,
+      "grad_norm": 0.32691786850356097,
+      "learning_rate": 1.532941325240263e-05,
+      "loss": 0.2825,
+      "step": 12725
+    },
+    {
+      "epoch": 2.1727404625757445,
+      "grad_norm": 0.3575672326173452,
+      "learning_rate": 1.5313606474456247e-05,
+      "loss": 0.2758,
+      "step": 12730
+    },
+    {
+      "epoch": 2.1735939233592214,
+      "grad_norm": 0.3228001556590004,
+      "learning_rate": 1.5297799696509865e-05,
+      "loss": 0.2923,
+      "step": 12735
+    },
+    {
+      "epoch": 2.174447384142699,
+      "grad_norm": 0.3680400486658653,
+      "learning_rate": 1.5281992918563482e-05,
+      "loss": 0.2645,
+      "step": 12740
+    },
+    {
+      "epoch": 2.1753008449261757,
+      "grad_norm": 0.2988167970677839,
+      "learning_rate": 1.5266186140617096e-05,
+      "loss": 0.2574,
+      "step": 12745
+    },
+    {
+      "epoch": 2.1761543057096526,
+      "grad_norm": 0.3481716359669406,
+      "learning_rate": 1.5250379362670714e-05,
+      "loss": 0.2704,
+      "step": 12750
+    },
+    {
+      "epoch": 2.1770077664931295,
+      "grad_norm": 0.6012227455690394,
+      "learning_rate": 1.523457258472433e-05,
+      "loss": 0.2756,
+      "step": 12755
+    },
+    {
+      "epoch": 2.1778612272766065,
+      "grad_norm": 0.3418187429879156,
+      "learning_rate": 1.5218765806777945e-05,
+      "loss": 0.2823,
+      "step": 12760
+    },
+    {
+      "epoch": 2.178714688060084,
+      "grad_norm": 0.3116854425472003,
+      "learning_rate": 1.5202959028831564e-05,
+      "loss": 0.2636,
+      "step": 12765
+    },
+    {
+      "epoch": 2.1795681488435608,
+      "grad_norm": 0.3524114961325397,
+      "learning_rate": 1.518715225088518e-05,
+      "loss": 0.2911,
+      "step": 12770
+    },
+    {
+      "epoch": 2.1804216096270377,
+      "grad_norm": 0.352148168631824,
+      "learning_rate": 1.5171345472938798e-05,
+      "loss": 0.2962,
+      "step": 12775
+    },
+    {
+      "epoch": 2.1812750704105146,
+      "grad_norm": 0.338679396703598,
+      "learning_rate": 1.5155538694992413e-05,
+      "loss": 0.2564,
+      "step": 12780
+    },
+    {
+      "epoch": 2.1821285311939915,
+      "grad_norm": 0.32320553715208916,
+      "learning_rate": 1.513973191704603e-05,
+      "loss": 0.2732,
+      "step": 12785
+    },
+    {
+      "epoch": 2.1829819919774685,
+      "grad_norm": 0.35736734339987114,
+      "learning_rate": 1.5123925139099645e-05,
+      "loss": 0.2631,
+      "step": 12790
+    },
+    {
+      "epoch": 2.183835452760946,
+      "grad_norm": 0.34756966658806226,
+      "learning_rate": 1.5108118361153264e-05,
+      "loss": 0.2557,
+      "step": 12795
+    },
+    {
+      "epoch": 2.1846889135444227,
+      "grad_norm": 0.37993879830853133,
+      "learning_rate": 1.509231158320688e-05,
+      "loss": 0.2503,
+      "step": 12800
+    },
+    {
+      "epoch": 2.1855423743278997,
+      "grad_norm": 0.3152500178102411,
+      "learning_rate": 1.5076504805260497e-05,
+      "loss": 0.2726,
+      "step": 12805
+    },
+    {
+      "epoch": 2.1863958351113766,
+      "grad_norm": 0.3530014163806682,
+      "learning_rate": 1.5060698027314113e-05,
+      "loss": 0.2509,
+      "step": 12810
+    },
+    {
+      "epoch": 2.1872492958948535,
+      "grad_norm": 0.3375905873609432,
+      "learning_rate": 1.5044891249367729e-05,
+      "loss": 0.2601,
+      "step": 12815
+    },
+    {
+      "epoch": 2.1881027566783304,
+      "grad_norm": 0.3208022206971321,
+      "learning_rate": 1.5029084471421345e-05,
+      "loss": 0.2641,
+      "step": 12820
+    },
+    {
+      "epoch": 2.188956217461808,
+      "grad_norm": 0.32239717872885754,
+      "learning_rate": 1.5013277693474964e-05,
+      "loss": 0.2709,
+      "step": 12825
+    },
+    {
+      "epoch": 2.1898096782452847,
+      "grad_norm": 0.3667722488933535,
+      "learning_rate": 1.499747091552858e-05,
+      "loss": 0.2588,
+      "step": 12830
+    },
+    {
+      "epoch": 2.1906631390287616,
+      "grad_norm": 0.3700698531378623,
+      "learning_rate": 1.4981664137582197e-05,
+      "loss": 0.2773,
+      "step": 12835
+    },
+    {
+      "epoch": 2.1915165998122386,
+      "grad_norm": 0.36627608522584637,
+      "learning_rate": 1.4965857359635813e-05,
+      "loss": 0.2595,
+      "step": 12840
+    },
+    {
+      "epoch": 2.1923700605957155,
+      "grad_norm": 0.43087433500491623,
+      "learning_rate": 1.4950050581689428e-05,
+      "loss": 0.2764,
+      "step": 12845
+    },
+    {
+      "epoch": 2.193223521379193,
+      "grad_norm": 0.3093574838883957,
+      "learning_rate": 1.4934243803743044e-05,
+      "loss": 0.266,
+      "step": 12850
+    },
+    {
+      "epoch": 2.1940769821626698,
+      "grad_norm": 0.5303469339533032,
+      "learning_rate": 1.4918437025796662e-05,
+      "loss": 0.2541,
+      "step": 12855
+    },
+    {
+      "epoch": 2.1949304429461467,
+      "grad_norm": 0.326459257267585,
+      "learning_rate": 1.490263024785028e-05,
+      "loss": 0.2816,
+      "step": 12860
+    },
+    {
+      "epoch": 2.1957839037296236,
+      "grad_norm": 0.31829004688424295,
+      "learning_rate": 1.4886823469903897e-05,
+      "loss": 0.2688,
+      "step": 12865
+    },
+    {
+      "epoch": 2.1966373645131005,
+      "grad_norm": 0.34921997998373205,
+      "learning_rate": 1.4871016691957512e-05,
+      "loss": 0.262,
+      "step": 12870
+    },
+    {
+      "epoch": 2.1974908252965775,
+      "grad_norm": 0.30883359187418896,
+      "learning_rate": 1.4855209914011128e-05,
+      "loss": 0.2882,
+      "step": 12875
+    },
+    {
+      "epoch": 2.1983442860800544,
+      "grad_norm": 0.33608701417398557,
+      "learning_rate": 1.4839403136064744e-05,
+      "loss": 0.2862,
+      "step": 12880
+    },
+    {
+      "epoch": 2.1991977468635318,
+      "grad_norm": 0.3203124453274556,
+      "learning_rate": 1.4823596358118361e-05,
+      "loss": 0.2484,
+      "step": 12885
+    },
+    {
+      "epoch": 2.2000512076470087,
+      "grad_norm": 0.32154313066050594,
+      "learning_rate": 1.4807789580171979e-05,
+      "loss": 0.2464,
+      "step": 12890
+    },
+    {
+      "epoch": 2.2009046684304856,
+      "grad_norm": 0.40498751721611115,
+      "learning_rate": 1.4791982802225596e-05,
+      "loss": 0.2595,
+      "step": 12895
+    },
+    {
+      "epoch": 2.2017581292139625,
+      "grad_norm": 0.328804572442942,
+      "learning_rate": 1.4776176024279212e-05,
+      "loss": 0.2595,
+      "step": 12900
+    },
+    {
+      "epoch": 2.2026115899974394,
+      "grad_norm": 0.3953120231902593,
+      "learning_rate": 1.4760369246332828e-05,
+      "loss": 0.2844,
+      "step": 12905
+    },
+    {
+      "epoch": 2.203465050780917,
+      "grad_norm": 0.31917535029732214,
+      "learning_rate": 1.4744562468386444e-05,
+      "loss": 0.293,
+      "step": 12910
+    },
+    {
+      "epoch": 2.2043185115643937,
+      "grad_norm": 0.3390405116903046,
+      "learning_rate": 1.4728755690440061e-05,
+      "loss": 0.2824,
+      "step": 12915
+    },
+    {
+      "epoch": 2.2051719723478707,
+      "grad_norm": 0.3022370342046419,
+      "learning_rate": 1.4712948912493679e-05,
+      "loss": 0.2439,
+      "step": 12920
+    },
+    {
+      "epoch": 2.2060254331313476,
+      "grad_norm": 0.34719223033193247,
+      "learning_rate": 1.4697142134547296e-05,
+      "loss": 0.2752,
+      "step": 12925
+    },
+    {
+      "epoch": 2.2068788939148245,
+      "grad_norm": 0.3338113211641244,
+      "learning_rate": 1.4681335356600912e-05,
+      "loss": 0.2725,
+      "step": 12930
+    },
+    {
+      "epoch": 2.2077323546983014,
+      "grad_norm": 0.3085027669451087,
+      "learning_rate": 1.4665528578654527e-05,
+      "loss": 0.2726,
+      "step": 12935
+    },
+    {
+      "epoch": 2.208585815481779,
+      "grad_norm": 0.3491580780264821,
+      "learning_rate": 1.4649721800708143e-05,
+      "loss": 0.2746,
+      "step": 12940
+    },
+    {
+      "epoch": 2.2094392762652557,
+      "grad_norm": 0.3381555139448322,
+      "learning_rate": 1.463391502276176e-05,
+      "loss": 0.2679,
+      "step": 12945
+    },
+    {
+      "epoch": 2.2102927370487326,
+      "grad_norm": 0.3620677646590881,
+      "learning_rate": 1.4618108244815378e-05,
+      "loss": 0.2748,
+      "step": 12950
+    },
+    {
+      "epoch": 2.2111461978322096,
+      "grad_norm": 0.33091174734322143,
+      "learning_rate": 1.4602301466868996e-05,
+      "loss": 0.2558,
+      "step": 12955
+    },
+    {
+      "epoch": 2.2119996586156865,
+      "grad_norm": 0.32578075958840164,
+      "learning_rate": 1.4586494688922611e-05,
+      "loss": 0.2563,
+      "step": 12960
+    },
+    {
+      "epoch": 2.2128531193991634,
+      "grad_norm": 0.3148099453779239,
+      "learning_rate": 1.4570687910976227e-05,
+      "loss": 0.2746,
+      "step": 12965
+    },
+    {
+      "epoch": 2.2137065801826408,
+      "grad_norm": 0.3020408957231368,
+      "learning_rate": 1.4554881133029843e-05,
+      "loss": 0.3052,
+      "step": 12970
+    },
+    {
+      "epoch": 2.2145600409661177,
+      "grad_norm": 0.33901774599802326,
+      "learning_rate": 1.453907435508346e-05,
+      "loss": 0.2809,
+      "step": 12975
+    },
+    {
+      "epoch": 2.2154135017495946,
+      "grad_norm": 0.3672776534404785,
+      "learning_rate": 1.4523267577137076e-05,
+      "loss": 0.2614,
+      "step": 12980
+    },
+    {
+      "epoch": 2.2162669625330715,
+      "grad_norm": 0.35687879044065507,
+      "learning_rate": 1.4507460799190695e-05,
+      "loss": 0.2523,
+      "step": 12985
+    },
+    {
+      "epoch": 2.2171204233165485,
+      "grad_norm": 0.33417825892532527,
+      "learning_rate": 1.4491654021244311e-05,
+      "loss": 0.2649,
+      "step": 12990
+    },
+    {
+      "epoch": 2.217973884100026,
+      "grad_norm": 0.3248124831537634,
+      "learning_rate": 1.4475847243297927e-05,
+      "loss": 0.2606,
+      "step": 12995
+    },
+    {
+      "epoch": 2.2188273448835028,
+      "grad_norm": 0.328358316916069,
+      "learning_rate": 1.4460040465351543e-05,
+      "loss": 0.2684,
+      "step": 13000
+    },
+    {
+      "epoch": 2.2196808056669797,
+      "grad_norm": 0.32013619138294613,
+      "learning_rate": 1.444423368740516e-05,
+      "loss": 0.2745,
+      "step": 13005
+    },
+    {
+      "epoch": 2.2205342664504566,
+      "grad_norm": 0.4578169853135187,
+      "learning_rate": 1.4428426909458776e-05,
+      "loss": 0.2596,
+      "step": 13010
+    },
+    {
+      "epoch": 2.2213877272339335,
+      "grad_norm": 0.32431737220922735,
+      "learning_rate": 1.4412620131512395e-05,
+      "loss": 0.2789,
+      "step": 13015
+    },
+    {
+      "epoch": 2.2222411880174104,
+      "grad_norm": 0.3505141108910819,
+      "learning_rate": 1.439681335356601e-05,
+      "loss": 0.2821,
+      "step": 13020
+    },
+    {
+      "epoch": 2.223094648800888,
+      "grad_norm": 0.33086072608819334,
+      "learning_rate": 1.4381006575619626e-05,
+      "loss": 0.2905,
+      "step": 13025
+    },
+    {
+      "epoch": 2.2239481095843647,
+      "grad_norm": 0.33884909980139977,
+      "learning_rate": 1.4365199797673242e-05,
+      "loss": 0.2553,
+      "step": 13030
+    },
+    {
+      "epoch": 2.2248015703678417,
+      "grad_norm": 0.37831012435066824,
+      "learning_rate": 1.434939301972686e-05,
+      "loss": 0.2592,
+      "step": 13035
+    },
+    {
+      "epoch": 2.2256550311513186,
+      "grad_norm": 0.31582713332257867,
+      "learning_rate": 1.4333586241780475e-05,
+      "loss": 0.2833,
+      "step": 13040
+    },
+    {
+      "epoch": 2.2265084919347955,
+      "grad_norm": 0.3455450182611844,
+      "learning_rate": 1.4317779463834095e-05,
+      "loss": 0.2675,
+      "step": 13045
+    },
+    {
+      "epoch": 2.2273619527182724,
+      "grad_norm": 0.3374489300728345,
+      "learning_rate": 1.430197268588771e-05,
+      "loss": 0.2979,
+      "step": 13050
+    },
+    {
+      "epoch": 2.22821541350175,
+      "grad_norm": 0.45405171174075065,
+      "learning_rate": 1.4286165907941326e-05,
+      "loss": 0.2867,
+      "step": 13055
+    },
+    {
+      "epoch": 2.2290688742852267,
+      "grad_norm": 0.3394515701351571,
+      "learning_rate": 1.4270359129994942e-05,
+      "loss": 0.2699,
+      "step": 13060
+    },
+    {
+      "epoch": 2.2299223350687036,
+      "grad_norm": 0.36895329729472887,
+      "learning_rate": 1.425455235204856e-05,
+      "loss": 0.2709,
+      "step": 13065
+    },
+    {
+      "epoch": 2.2307757958521806,
+      "grad_norm": 0.35443931340420204,
+      "learning_rate": 1.4238745574102175e-05,
+      "loss": 0.2612,
+      "step": 13070
+    },
+    {
+      "epoch": 2.2316292566356575,
+      "grad_norm": 0.37439768952743185,
+      "learning_rate": 1.4222938796155794e-05,
+      "loss": 0.2566,
+      "step": 13075
+    },
+    {
+      "epoch": 2.2324827174191344,
+      "grad_norm": 0.3989028279151154,
+      "learning_rate": 1.420713201820941e-05,
+      "loss": 0.267,
+      "step": 13080
+    },
+    {
+      "epoch": 2.2333361782026118,
+      "grad_norm": 0.3145608456005138,
+      "learning_rate": 1.4191325240263026e-05,
+      "loss": 0.2595,
+      "step": 13085
+    },
+    {
+      "epoch": 2.2341896389860887,
+      "grad_norm": 0.404315313993326,
+      "learning_rate": 1.4175518462316642e-05,
+      "loss": 0.2476,
+      "step": 13090
+    },
+    {
+      "epoch": 2.2350430997695656,
+      "grad_norm": 0.33175989786238314,
+      "learning_rate": 1.4159711684370259e-05,
+      "loss": 0.2445,
+      "step": 13095
+    },
+    {
+      "epoch": 2.2358965605530425,
+      "grad_norm": 0.3673847446231519,
+      "learning_rate": 1.4143904906423875e-05,
+      "loss": 0.2795,
+      "step": 13100
+    },
+    {
+      "epoch": 2.2367500213365195,
+      "grad_norm": 0.32920478196154535,
+      "learning_rate": 1.412809812847749e-05,
+      "loss": 0.2612,
+      "step": 13105
+    },
+    {
+      "epoch": 2.2376034821199964,
+      "grad_norm": 0.3942709927381735,
+      "learning_rate": 1.411229135053111e-05,
+      "loss": 0.2664,
+      "step": 13110
+    },
+    {
+      "epoch": 2.2384569429034737,
+      "grad_norm": 0.34105379614174075,
+      "learning_rate": 1.4096484572584725e-05,
+      "loss": 0.2713,
+      "step": 13115
+    },
+    {
+      "epoch": 2.2393104036869507,
+      "grad_norm": 0.3319067913692869,
+      "learning_rate": 1.4080677794638341e-05,
+      "loss": 0.2797,
+      "step": 13120
+    },
+    {
+      "epoch": 2.2401638644704276,
+      "grad_norm": 0.33108165381903953,
+      "learning_rate": 1.4064871016691957e-05,
+      "loss": 0.274,
+      "step": 13125
+    },
+    {
+      "epoch": 2.2410173252539045,
+      "grad_norm": 0.3389736881258419,
+      "learning_rate": 1.4049064238745574e-05,
+      "loss": 0.28,
+      "step": 13130
+    },
+    {
+      "epoch": 2.2418707860373814,
+      "grad_norm": 0.3936727543709566,
+      "learning_rate": 1.403325746079919e-05,
+      "loss": 0.2851,
+      "step": 13135
+    },
+    {
+      "epoch": 2.242724246820859,
+      "grad_norm": 0.35429129447576024,
+      "learning_rate": 1.401745068285281e-05,
+      "loss": 0.2819,
+      "step": 13140
+    },
+    {
+      "epoch": 2.2435777076043357,
+      "grad_norm": 0.3327919813822197,
+      "learning_rate": 1.4001643904906425e-05,
+      "loss": 0.2712,
+      "step": 13145
+    },
+    {
+      "epoch": 2.2444311683878126,
+      "grad_norm": 0.36825000296835053,
+      "learning_rate": 1.3985837126960041e-05,
+      "loss": 0.2704,
+      "step": 13150
+    },
+    {
+      "epoch": 2.2452846291712896,
+      "grad_norm": 0.3037327517059426,
+      "learning_rate": 1.3970030349013657e-05,
+      "loss": 0.2845,
+      "step": 13155
+    },
+    {
+      "epoch": 2.2461380899547665,
+      "grad_norm": 0.3463895684505155,
+      "learning_rate": 1.3954223571067274e-05,
+      "loss": 0.2614,
+      "step": 13160
+    },
+    {
+      "epoch": 2.2469915507382434,
+      "grad_norm": 0.3716484778076624,
+      "learning_rate": 1.393841679312089e-05,
+      "loss": 0.265,
+      "step": 13165
+    },
+    {
+      "epoch": 2.247845011521721,
+      "grad_norm": 0.34069370568224044,
+      "learning_rate": 1.3922610015174509e-05,
+      "loss": 0.272,
+      "step": 13170
+    },
+    {
+      "epoch": 2.2486984723051977,
+      "grad_norm": 0.3560029147034675,
+      "learning_rate": 1.3906803237228125e-05,
+      "loss": 0.2796,
+      "step": 13175
+    },
+    {
+      "epoch": 2.2495519330886746,
+      "grad_norm": 0.38746486241761824,
+      "learning_rate": 1.389099645928174e-05,
+      "loss": 0.2818,
+      "step": 13180
+    },
+    {
+      "epoch": 2.2504053938721515,
+      "grad_norm": 0.3482362639054943,
+      "learning_rate": 1.3875189681335356e-05,
+      "loss": 0.2777,
+      "step": 13185
+    },
+    {
+      "epoch": 2.2512588546556285,
+      "grad_norm": 0.3255965079568052,
+      "learning_rate": 1.3859382903388974e-05,
+      "loss": 0.2851,
+      "step": 13190
+    },
+    {
+      "epoch": 2.2521123154391054,
+      "grad_norm": 0.31144665726446935,
+      "learning_rate": 1.384357612544259e-05,
+      "loss": 0.2717,
+      "step": 13195
+    },
+    {
+      "epoch": 2.2529657762225828,
+      "grad_norm": 0.3313669197754686,
+      "learning_rate": 1.3827769347496209e-05,
+      "loss": 0.2776,
+      "step": 13200
+    },
+    {
+      "epoch": 2.2538192370060597,
+      "grad_norm": 0.2952076521595467,
+      "learning_rate": 1.3811962569549824e-05,
+      "loss": 0.2752,
+      "step": 13205
+    },
+    {
+      "epoch": 2.2546726977895366,
+      "grad_norm": 0.3530678406931364,
+      "learning_rate": 1.379615579160344e-05,
+      "loss": 0.2576,
+      "step": 13210
+    },
+    {
+      "epoch": 2.2555261585730135,
+      "grad_norm": 0.36836585151120216,
+      "learning_rate": 1.3780349013657056e-05,
+      "loss": 0.2628,
+      "step": 13215
+    },
+    {
+      "epoch": 2.2563796193564905,
+      "grad_norm": 0.33551894077187827,
+      "learning_rate": 1.3764542235710673e-05,
+      "loss": 0.2684,
+      "step": 13220
+    },
+    {
+      "epoch": 2.257233080139968,
+      "grad_norm": 0.4224354356132112,
+      "learning_rate": 1.374873545776429e-05,
+      "loss": 0.2868,
+      "step": 13225
+    },
+    {
+      "epoch": 2.2580865409234447,
+      "grad_norm": 0.4123832020951836,
+      "learning_rate": 1.3732928679817905e-05,
+      "loss": 0.2767,
+      "step": 13230
+    },
+    {
+      "epoch": 2.2589400017069217,
+      "grad_norm": 0.3906636906336848,
+      "learning_rate": 1.3717121901871524e-05,
+      "loss": 0.2656,
+      "step": 13235
+    },
+    {
+      "epoch": 2.2597934624903986,
+      "grad_norm": 0.34283897290348253,
+      "learning_rate": 1.370131512392514e-05,
+      "loss": 0.2786,
+      "step": 13240
+    },
+    {
+      "epoch": 2.2606469232738755,
+      "grad_norm": 0.3163011115260496,
+      "learning_rate": 1.3685508345978756e-05,
+      "loss": 0.2606,
+      "step": 13245
+    },
+    {
+      "epoch": 2.2615003840573524,
+      "grad_norm": 0.39600564834207624,
+      "learning_rate": 1.3669701568032373e-05,
+      "loss": 0.2605,
+      "step": 13250
+    },
+    {
+      "epoch": 2.2623538448408294,
+      "grad_norm": 0.3083749239635374,
+      "learning_rate": 1.3653894790085989e-05,
+      "loss": 0.265,
+      "step": 13255
+    },
+    {
+      "epoch": 2.2632073056243067,
+      "grad_norm": 0.3760609634293888,
+      "learning_rate": 1.3638088012139605e-05,
+      "loss": 0.2462,
+      "step": 13260
+    },
+    {
+      "epoch": 2.2640607664077836,
+      "grad_norm": 0.3757475388409267,
+      "learning_rate": 1.3622281234193224e-05,
+      "loss": 0.2803,
+      "step": 13265
+    },
+    {
+      "epoch": 2.2649142271912606,
+      "grad_norm": 0.48846304920007216,
+      "learning_rate": 1.360647445624684e-05,
+      "loss": 0.2603,
+      "step": 13270
+    },
+    {
+      "epoch": 2.2657676879747375,
+      "grad_norm": 0.3514453908540707,
+      "learning_rate": 1.3590667678300455e-05,
+      "loss": 0.287,
+      "step": 13275
+    },
+    {
+      "epoch": 2.2666211487582144,
+      "grad_norm": 0.3470686782406026,
+      "learning_rate": 1.3574860900354073e-05,
+      "loss": 0.3474,
+      "step": 13280
+    },
+    {
+      "epoch": 2.2674746095416918,
+      "grad_norm": 0.3508841337950813,
+      "learning_rate": 1.3559054122407689e-05,
+      "loss": 0.2558,
+      "step": 13285
+    },
+    {
+      "epoch": 2.2683280703251687,
+      "grad_norm": 0.3139648663004901,
+      "learning_rate": 1.3543247344461304e-05,
+      "loss": 0.2645,
+      "step": 13290
+    },
+    {
+      "epoch": 2.2691815311086456,
+      "grad_norm": 0.32112362909304765,
+      "learning_rate": 1.3527440566514923e-05,
+      "loss": 0.2534,
+      "step": 13295
+    },
+    {
+      "epoch": 2.2700349918921225,
+      "grad_norm": 0.40308720353751076,
+      "learning_rate": 1.351163378856854e-05,
+      "loss": 0.249,
+      "step": 13300
+    },
+    {
+      "epoch": 2.2708884526755995,
+      "grad_norm": 0.3711046566966942,
+      "learning_rate": 1.3495827010622155e-05,
+      "loss": 0.2605,
+      "step": 13305
+    },
+    {
+      "epoch": 2.2717419134590764,
+      "grad_norm": 0.3396794923247863,
+      "learning_rate": 1.3480020232675772e-05,
+      "loss": 0.2707,
+      "step": 13310
+    },
+    {
+      "epoch": 2.2725953742425533,
+      "grad_norm": 0.35625958725160256,
+      "learning_rate": 1.3464213454729388e-05,
+      "loss": 0.2635,
+      "step": 13315
+    },
+    {
+      "epoch": 2.2734488350260307,
+      "grad_norm": 0.31309163461644807,
+      "learning_rate": 1.3448406676783004e-05,
+      "loss": 0.2773,
+      "step": 13320
+    },
+    {
+      "epoch": 2.2743022958095076,
+      "grad_norm": 0.3445215453677931,
+      "learning_rate": 1.3432599898836623e-05,
+      "loss": 0.2555,
+      "step": 13325
+    },
+    {
+      "epoch": 2.2751557565929845,
+      "grad_norm": 0.36659337926037683,
+      "learning_rate": 1.3416793120890239e-05,
+      "loss": 0.2702,
+      "step": 13330
+    },
+    {
+      "epoch": 2.2760092173764614,
+      "grad_norm": 0.35224502414420666,
+      "learning_rate": 1.3400986342943855e-05,
+      "loss": 0.2683,
+      "step": 13335
+    },
+    {
+      "epoch": 2.2768626781599384,
+      "grad_norm": 0.3050596078699797,
+      "learning_rate": 1.3385179564997472e-05,
+      "loss": 0.2742,
+      "step": 13340
+    },
+    {
+      "epoch": 2.2777161389434157,
+      "grad_norm": 0.3445279792044359,
+      "learning_rate": 1.3369372787051088e-05,
+      "loss": 0.284,
+      "step": 13345
+    },
+    {
+      "epoch": 2.2785695997268927,
+      "grad_norm": 0.4521506535764485,
+      "learning_rate": 1.3353566009104704e-05,
+      "loss": 0.2514,
+      "step": 13350
+    },
+    {
+      "epoch": 2.2794230605103696,
+      "grad_norm": 0.303913246496587,
+      "learning_rate": 1.3337759231158323e-05,
+      "loss": 0.2633,
+      "step": 13355
+    },
+    {
+      "epoch": 2.2802765212938465,
+      "grad_norm": 0.3502568952160238,
+      "learning_rate": 1.3321952453211939e-05,
+      "loss": 0.2594,
+      "step": 13360
+    },
+    {
+      "epoch": 2.2811299820773234,
+      "grad_norm": 0.3717036717618989,
+      "learning_rate": 1.3306145675265554e-05,
+      "loss": 0.253,
+      "step": 13365
+    },
+    {
+      "epoch": 2.281983442860801,
+      "grad_norm": 0.31335236449414805,
+      "learning_rate": 1.3290338897319172e-05,
+      "loss": 0.289,
+      "step": 13370
+    },
+    {
+      "epoch": 2.2828369036442777,
+      "grad_norm": 0.47642486009392315,
+      "learning_rate": 1.3274532119372788e-05,
+      "loss": 0.312,
+      "step": 13375
+    },
+    {
+      "epoch": 2.2836903644277546,
+      "grad_norm": 0.3706481152626613,
+      "learning_rate": 1.3258725341426403e-05,
+      "loss": 0.2734,
+      "step": 13380
+    },
+    {
+      "epoch": 2.2845438252112316,
+      "grad_norm": 0.335699018913271,
+      "learning_rate": 1.3242918563480019e-05,
+      "loss": 0.2588,
+      "step": 13385
+    },
+    {
+      "epoch": 2.2853972859947085,
+      "grad_norm": 0.39600171459187933,
+      "learning_rate": 1.3227111785533638e-05,
+      "loss": 0.2668,
+      "step": 13390
+    },
+    {
+      "epoch": 2.2862507467781854,
+      "grad_norm": 0.31398579074501,
+      "learning_rate": 1.3211305007587254e-05,
+      "loss": 0.2751,
+      "step": 13395
+    },
+    {
+      "epoch": 2.2871042075616623,
+      "grad_norm": 0.33944474559657767,
+      "learning_rate": 1.3195498229640871e-05,
+      "loss": 0.2719,
+      "step": 13400
+    },
+    {
+      "epoch": 2.2879576683451397,
+      "grad_norm": 0.3552249703706572,
+      "learning_rate": 1.3179691451694487e-05,
+      "loss": 0.2578,
+      "step": 13405
+    },
+    {
+      "epoch": 2.2888111291286166,
+      "grad_norm": 0.34558029914006605,
+      "learning_rate": 1.3163884673748103e-05,
+      "loss": 0.253,
+      "step": 13410
+    },
+    {
+      "epoch": 2.2896645899120935,
+      "grad_norm": 0.34898837234878644,
+      "learning_rate": 1.3148077895801719e-05,
+      "loss": 0.2707,
+      "step": 13415
+    },
+    {
+      "epoch": 2.2905180506955705,
+      "grad_norm": 0.3293566621281475,
+      "learning_rate": 1.3132271117855338e-05,
+      "loss": 0.2694,
+      "step": 13420
+    },
+    {
+      "epoch": 2.2913715114790474,
+      "grad_norm": 0.33463051881200645,
+      "learning_rate": 1.3116464339908954e-05,
+      "loss": 0.2584,
+      "step": 13425
+    },
+    {
+      "epoch": 2.2922249722625248,
+      "grad_norm": 0.32214858707844063,
+      "learning_rate": 1.3100657561962571e-05,
+      "loss": 0.265,
+      "step": 13430
+    },
+    {
+      "epoch": 2.2930784330460017,
+      "grad_norm": 0.2894622609386176,
+      "learning_rate": 1.3084850784016187e-05,
+      "loss": 0.2721,
+      "step": 13435
+    },
+    {
+      "epoch": 2.2939318938294786,
+      "grad_norm": 0.3351765964390977,
+      "learning_rate": 1.3069044006069803e-05,
+      "loss": 0.2602,
+      "step": 13440
+    },
+    {
+      "epoch": 2.2947853546129555,
+      "grad_norm": 0.4465919787461111,
+      "learning_rate": 1.3053237228123418e-05,
+      "loss": 0.2686,
+      "step": 13445
+    },
+    {
+      "epoch": 2.2956388153964324,
+      "grad_norm": 0.3568598133398877,
+      "learning_rate": 1.3037430450177038e-05,
+      "loss": 0.2734,
+      "step": 13450
+    },
+    {
+      "epoch": 2.2964922761799094,
+      "grad_norm": 0.6384608561412038,
+      "learning_rate": 1.3021623672230653e-05,
+      "loss": 0.2655,
+      "step": 13455
+    },
+    {
+      "epoch": 2.2973457369633863,
+      "grad_norm": 0.33839005812425016,
+      "learning_rate": 1.300581689428427e-05,
+      "loss": 0.2868,
+      "step": 13460
+    },
+    {
+      "epoch": 2.2981991977468637,
+      "grad_norm": 0.3483743337197857,
+      "learning_rate": 1.2990010116337887e-05,
+      "loss": 0.2533,
+      "step": 13465
+    },
+    {
+      "epoch": 2.2990526585303406,
+      "grad_norm": 0.3389495049205933,
+      "learning_rate": 1.2974203338391502e-05,
+      "loss": 0.2666,
+      "step": 13470
+    },
+    {
+      "epoch": 2.2999061193138175,
+      "grad_norm": 0.33493171178968356,
+      "learning_rate": 1.2958396560445118e-05,
+      "loss": 0.246,
+      "step": 13475
+    },
+    {
+      "epoch": 2.3007595800972944,
+      "grad_norm": 0.3279772510257887,
+      "learning_rate": 1.2942589782498737e-05,
+      "loss": 0.2824,
+      "step": 13480
+    },
+    {
+      "epoch": 2.3016130408807713,
+      "grad_norm": 0.36115760396041924,
+      "learning_rate": 1.2926783004552353e-05,
+      "loss": 0.2621,
+      "step": 13485
+    },
+    {
+      "epoch": 2.3024665016642487,
+      "grad_norm": 0.3373729497737809,
+      "learning_rate": 1.291097622660597e-05,
+      "loss": 0.2767,
+      "step": 13490
+    },
+    {
+      "epoch": 2.3033199624477256,
+      "grad_norm": 0.35983074089182177,
+      "learning_rate": 1.2895169448659586e-05,
+      "loss": 0.2633,
+      "step": 13495
+    },
+    {
+      "epoch": 2.3041734232312026,
+      "grad_norm": 0.36750791033347413,
+      "learning_rate": 1.2879362670713202e-05,
+      "loss": 0.297,
+      "step": 13500
+    },
+    {
+      "epoch": 2.3050268840146795,
+      "grad_norm": 0.37034125219454295,
+      "learning_rate": 1.2863555892766818e-05,
+      "loss": 0.2723,
+      "step": 13505
+    },
+    {
+      "epoch": 2.3058803447981564,
+      "grad_norm": 0.31177905436906944,
+      "learning_rate": 1.2847749114820433e-05,
+      "loss": 0.2711,
+      "step": 13510
+    },
+    {
+      "epoch": 2.3067338055816338,
+      "grad_norm": 0.33355398096509026,
+      "learning_rate": 1.2831942336874053e-05,
+      "loss": 0.2648,
+      "step": 13515
+    },
+    {
+      "epoch": 2.3075872663651107,
+      "grad_norm": 0.3337734453400322,
+      "learning_rate": 1.281613555892767e-05,
+      "loss": 0.2897,
+      "step": 13520
+    },
+    {
+      "epoch": 2.3084407271485876,
+      "grad_norm": 0.35520779216386267,
+      "learning_rate": 1.2800328780981286e-05,
+      "loss": 0.2554,
+      "step": 13525
+    },
+    {
+      "epoch": 2.3092941879320645,
+      "grad_norm": 0.32716894576133215,
+      "learning_rate": 1.2784522003034902e-05,
+      "loss": 0.2681,
+      "step": 13530
+    },
+    {
+      "epoch": 2.3101476487155415,
+      "grad_norm": 0.38456397670233244,
+      "learning_rate": 1.2768715225088517e-05,
+      "loss": 0.2503,
+      "step": 13535
+    },
+    {
+      "epoch": 2.3110011094990184,
+      "grad_norm": 0.354510794826201,
+      "learning_rate": 1.2752908447142133e-05,
+      "loss": 0.2566,
+      "step": 13540
+    },
+    {
+      "epoch": 2.3118545702824953,
+      "grad_norm": 0.3240400186178979,
+      "learning_rate": 1.2737101669195752e-05,
+      "loss": 0.2767,
+      "step": 13545
+    },
+    {
+      "epoch": 2.3127080310659727,
+      "grad_norm": 0.3151313415868495,
+      "learning_rate": 1.272129489124937e-05,
+      "loss": 0.2611,
+      "step": 13550
+    },
+    {
+      "epoch": 2.3135614918494496,
+      "grad_norm": 0.3645467755833886,
+      "learning_rate": 1.2705488113302986e-05,
+      "loss": 0.2738,
+      "step": 13555
+    },
+    {
+      "epoch": 2.3144149526329265,
+      "grad_norm": 0.33916988264617853,
+      "learning_rate": 1.2689681335356601e-05,
+      "loss": 0.2685,
+      "step": 13560
+    },
+    {
+      "epoch": 2.3152684134164034,
+      "grad_norm": 0.35312157453609383,
+      "learning_rate": 1.2673874557410217e-05,
+      "loss": 0.2651,
+      "step": 13565
+    },
+    {
+      "epoch": 2.3161218741998804,
+      "grad_norm": 0.3486353831660235,
+      "learning_rate": 1.2658067779463833e-05,
+      "loss": 0.2562,
+      "step": 13570
+    },
+    {
+      "epoch": 2.3169753349833577,
+      "grad_norm": 0.35167361210599335,
+      "learning_rate": 1.2642261001517452e-05,
+      "loss": 0.2624,
+      "step": 13575
+    },
+    {
+      "epoch": 2.3178287957668346,
+      "grad_norm": 0.3675526623335268,
+      "learning_rate": 1.262645422357107e-05,
+      "loss": 0.2447,
+      "step": 13580
+    },
+    {
+      "epoch": 2.3186822565503116,
+      "grad_norm": 0.3801187038171107,
+      "learning_rate": 1.2610647445624685e-05,
+      "loss": 0.2543,
+      "step": 13585
+    },
+    {
+      "epoch": 2.3195357173337885,
+      "grad_norm": 0.3260726023256078,
+      "learning_rate": 1.2594840667678301e-05,
+      "loss": 0.2593,
+      "step": 13590
+    },
+    {
+      "epoch": 2.3203891781172654,
+      "grad_norm": 0.337574532861758,
+      "learning_rate": 1.2579033889731917e-05,
+      "loss": 0.2794,
+      "step": 13595
+    },
+    {
+      "epoch": 2.3212426389007423,
+      "grad_norm": 1.7267629397768802,
+      "learning_rate": 1.2563227111785532e-05,
+      "loss": 0.2702,
+      "step": 13600
+    },
+    {
+      "epoch": 2.3220960996842193,
+      "grad_norm": 0.3955834078147852,
+      "learning_rate": 1.2547420333839152e-05,
+      "loss": 0.2465,
+      "step": 13605
+    },
+    {
+      "epoch": 2.3229495604676966,
+      "grad_norm": 0.3262499618665922,
+      "learning_rate": 1.2531613555892769e-05,
+      "loss": 0.277,
+      "step": 13610
+    },
+    {
+      "epoch": 2.3238030212511736,
+      "grad_norm": 0.35351031166159325,
+      "learning_rate": 1.2515806777946385e-05,
+      "loss": 0.2865,
+      "step": 13615
+    },
+    {
+      "epoch": 2.3246564820346505,
+      "grad_norm": 0.42411178148525197,
+      "learning_rate": 1.25e-05,
+      "loss": 0.2516,
+      "step": 13620
+    },
+    {
+      "epoch": 2.3255099428181274,
+      "grad_norm": 0.3531445964873752,
+      "learning_rate": 1.2484193222053616e-05,
+      "loss": 0.268,
+      "step": 13625
+    },
+    {
+      "epoch": 2.3263634036016043,
+      "grad_norm": 0.41456787080265933,
+      "learning_rate": 1.2468386444107234e-05,
+      "loss": 0.2745,
+      "step": 13630
+    },
+    {
+      "epoch": 2.3272168643850817,
+      "grad_norm": 0.28403448926776254,
+      "learning_rate": 1.245257966616085e-05,
+      "loss": 0.2858,
+      "step": 13635
+    },
+    {
+      "epoch": 2.3280703251685586,
+      "grad_norm": 0.32908123683743307,
+      "learning_rate": 1.2436772888214467e-05,
+      "loss": 0.2875,
+      "step": 13640
+    },
+    {
+      "epoch": 2.3289237859520355,
+      "grad_norm": 0.4612657030725096,
+      "learning_rate": 1.2420966110268084e-05,
+      "loss": 0.2496,
+      "step": 13645
+    },
+    {
+      "epoch": 2.3297772467355125,
+      "grad_norm": 0.38676606881990483,
+      "learning_rate": 1.24051593323217e-05,
+      "loss": 0.2556,
+      "step": 13650
+    },
+    {
+      "epoch": 2.3306307075189894,
+      "grad_norm": 0.3337040823652374,
+      "learning_rate": 1.2389352554375316e-05,
+      "loss": 0.2713,
+      "step": 13655
+    },
+    {
+      "epoch": 2.3314841683024667,
+      "grad_norm": 0.3898313745285282,
+      "learning_rate": 1.2373545776428933e-05,
+      "loss": 0.2587,
+      "step": 13660
+    },
+    {
+      "epoch": 2.3323376290859437,
+      "grad_norm": 0.40488411353562653,
+      "learning_rate": 1.235773899848255e-05,
+      "loss": 0.281,
+      "step": 13665
+    },
+    {
+      "epoch": 2.3331910898694206,
+      "grad_norm": 0.3341003634320264,
+      "learning_rate": 1.2341932220536167e-05,
+      "loss": 0.2711,
+      "step": 13670
+    },
+    {
+      "epoch": 2.3340445506528975,
+      "grad_norm": 0.3280264582320497,
+      "learning_rate": 1.2326125442589784e-05,
+      "loss": 0.2566,
+      "step": 13675
+    },
+    {
+      "epoch": 2.3348980114363744,
+      "grad_norm": 0.3941865741053653,
+      "learning_rate": 1.23103186646434e-05,
+      "loss": 0.2895,
+      "step": 13680
+    },
+    {
+      "epoch": 2.3357514722198514,
+      "grad_norm": 0.3545181728225511,
+      "learning_rate": 1.2294511886697016e-05,
+      "loss": 0.2442,
+      "step": 13685
+    },
+    {
+      "epoch": 2.3366049330033283,
+      "grad_norm": 0.3993830195808206,
+      "learning_rate": 1.2278705108750631e-05,
+      "loss": 0.2502,
+      "step": 13690
+    },
+    {
+      "epoch": 2.3374583937868056,
+      "grad_norm": 0.3491825617229024,
+      "learning_rate": 1.2262898330804249e-05,
+      "loss": 0.2605,
+      "step": 13695
+    },
+    {
+      "epoch": 2.3383118545702826,
+      "grad_norm": 0.3220282909490161,
+      "learning_rate": 1.2247091552857866e-05,
+      "loss": 0.2784,
+      "step": 13700
+    },
+    {
+      "epoch": 2.3391653153537595,
+      "grad_norm": 0.3062997099550015,
+      "learning_rate": 1.2231284774911482e-05,
+      "loss": 0.2562,
+      "step": 13705
+    },
+    {
+      "epoch": 2.3400187761372364,
+      "grad_norm": 0.45019869708623317,
+      "learning_rate": 1.22154779969651e-05,
+      "loss": 0.2731,
+      "step": 13710
+    },
+    {
+      "epoch": 2.3408722369207133,
+      "grad_norm": 0.3647848167556784,
+      "learning_rate": 1.2199671219018715e-05,
+      "loss": 0.2635,
+      "step": 13715
+    },
+    {
+      "epoch": 2.3417256977041907,
+      "grad_norm": 0.31568148779307076,
+      "learning_rate": 1.2183864441072331e-05,
+      "loss": 0.2543,
+      "step": 13720
+    },
+    {
+      "epoch": 2.3425791584876676,
+      "grad_norm": 0.39945924823601864,
+      "learning_rate": 1.2168057663125949e-05,
+      "loss": 0.2768,
+      "step": 13725
+    },
+    {
+      "epoch": 2.3434326192711445,
+      "grad_norm": 0.3292869322699895,
+      "learning_rate": 1.2152250885179566e-05,
+      "loss": 0.2529,
+      "step": 13730
+    },
+    {
+      "epoch": 2.3442860800546215,
+      "grad_norm": 0.35104556195991654,
+      "learning_rate": 1.2136444107233182e-05,
+      "loss": 0.2749,
+      "step": 13735
+    },
+    {
+      "epoch": 2.3451395408380984,
+      "grad_norm": 0.3458157948154376,
+      "learning_rate": 1.21206373292868e-05,
+      "loss": 0.2653,
+      "step": 13740
+    },
+    {
+      "epoch": 2.3459930016215753,
+      "grad_norm": 0.3701665159887945,
+      "learning_rate": 1.2104830551340415e-05,
+      "loss": 0.2372,
+      "step": 13745
+    },
+    {
+      "epoch": 2.3468464624050527,
+      "grad_norm": 0.3500403307761384,
+      "learning_rate": 1.208902377339403e-05,
+      "loss": 0.2522,
+      "step": 13750
+    },
+    {
+      "epoch": 2.3476999231885296,
+      "grad_norm": 0.3699869244805946,
+      "learning_rate": 1.2073216995447648e-05,
+      "loss": 0.2631,
+      "step": 13755
+    },
+    {
+      "epoch": 2.3485533839720065,
+      "grad_norm": 0.3297373602356559,
+      "learning_rate": 1.2057410217501266e-05,
+      "loss": 0.2721,
+      "step": 13760
+    },
+    {
+      "epoch": 2.3494068447554834,
+      "grad_norm": 0.36704741894542436,
+      "learning_rate": 1.2041603439554881e-05,
+      "loss": 0.2497,
+      "step": 13765
+    },
+    {
+      "epoch": 2.3502603055389604,
+      "grad_norm": 0.3752754677463977,
+      "learning_rate": 1.2025796661608499e-05,
+      "loss": 0.2519,
+      "step": 13770
+    },
+    {
+      "epoch": 2.3511137663224373,
+      "grad_norm": 0.3466693760766624,
+      "learning_rate": 1.2009989883662115e-05,
+      "loss": 0.2529,
+      "step": 13775
+    },
+    {
+      "epoch": 2.3519672271059147,
+      "grad_norm": 0.4805276702632299,
+      "learning_rate": 1.199418310571573e-05,
+      "loss": 0.2829,
+      "step": 13780
+    },
+    {
+      "epoch": 2.3528206878893916,
+      "grad_norm": 0.3959562211082513,
+      "learning_rate": 1.1978376327769348e-05,
+      "loss": 0.2693,
+      "step": 13785
+    },
+    {
+      "epoch": 2.3536741486728685,
+      "grad_norm": 0.35924528379883774,
+      "learning_rate": 1.1962569549822965e-05,
+      "loss": 0.265,
+      "step": 13790
+    },
+    {
+      "epoch": 2.3545276094563454,
+      "grad_norm": 0.2774038864591207,
+      "learning_rate": 1.1946762771876581e-05,
+      "loss": 0.2575,
+      "step": 13795
+    },
+    {
+      "epoch": 2.3553810702398223,
+      "grad_norm": 0.33781070283823095,
+      "learning_rate": 1.1930955993930199e-05,
+      "loss": 0.2841,
+      "step": 13800
+    },
+    {
+      "epoch": 2.3562345310232997,
+      "grad_norm": 0.35215720977487114,
+      "learning_rate": 1.1915149215983814e-05,
+      "loss": 0.2337,
+      "step": 13805
+    },
+    {
+      "epoch": 2.3570879918067766,
+      "grad_norm": 0.3380343209709989,
+      "learning_rate": 1.189934243803743e-05,
+      "loss": 0.2493,
+      "step": 13810
+    },
+    {
+      "epoch": 2.3579414525902536,
+      "grad_norm": 0.5667691024279196,
+      "learning_rate": 1.1883535660091048e-05,
+      "loss": 0.2467,
+      "step": 13815
+    },
+    {
+      "epoch": 2.3587949133737305,
+      "grad_norm": 0.35442986611563343,
+      "learning_rate": 1.1867728882144665e-05,
+      "loss": 0.2622,
+      "step": 13820
+    },
+    {
+      "epoch": 2.3596483741572074,
+      "grad_norm": 0.4573056954650897,
+      "learning_rate": 1.185192210419828e-05,
+      "loss": 0.2958,
+      "step": 13825
+    },
+    {
+      "epoch": 2.3605018349406843,
+      "grad_norm": 0.3464846363619561,
+      "learning_rate": 1.1836115326251897e-05,
+      "loss": 0.2557,
+      "step": 13830
+    },
+    {
+      "epoch": 2.3613552957241613,
+      "grad_norm": 0.31107449399274584,
+      "learning_rate": 1.1820308548305514e-05,
+      "loss": 0.2569,
+      "step": 13835
+    },
+    {
+      "epoch": 2.3622087565076386,
+      "grad_norm": 0.31223474529197637,
+      "learning_rate": 1.180450177035913e-05,
+      "loss": 0.2636,
+      "step": 13840
+    },
+    {
+      "epoch": 2.3630622172911155,
+      "grad_norm": 0.3423766402813196,
+      "learning_rate": 1.1788694992412747e-05,
+      "loss": 0.254,
+      "step": 13845
+    },
+    {
+      "epoch": 2.3639156780745925,
+      "grad_norm": 0.4642496449605801,
+      "learning_rate": 1.1772888214466365e-05,
+      "loss": 0.2397,
+      "step": 13850
+    },
+    {
+      "epoch": 2.3647691388580694,
+      "grad_norm": 0.34953469626709727,
+      "learning_rate": 1.175708143651998e-05,
+      "loss": 0.2986,
+      "step": 13855
+    },
+    {
+      "epoch": 2.3656225996415463,
+      "grad_norm": 0.35708543589868114,
+      "learning_rate": 1.1741274658573596e-05,
+      "loss": 0.2582,
+      "step": 13860
+    },
+    {
+      "epoch": 2.3664760604250237,
+      "grad_norm": 0.3698635896251769,
+      "learning_rate": 1.1725467880627214e-05,
+      "loss": 0.2721,
+      "step": 13865
+    },
+    {
+      "epoch": 2.3673295212085006,
+      "grad_norm": 0.3302748261797436,
+      "learning_rate": 1.170966110268083e-05,
+      "loss": 0.2894,
+      "step": 13870
+    },
+    {
+      "epoch": 2.3681829819919775,
+      "grad_norm": 0.31393263378581593,
+      "learning_rate": 1.1693854324734447e-05,
+      "loss": 0.2562,
+      "step": 13875
+    },
+    {
+      "epoch": 2.3690364427754544,
+      "grad_norm": 0.5538836388276289,
+      "learning_rate": 1.1678047546788064e-05,
+      "loss": 0.2488,
+      "step": 13880
+    },
+    {
+      "epoch": 2.3698899035589314,
+      "grad_norm": 0.34646042875522115,
+      "learning_rate": 1.166224076884168e-05,
+      "loss": 0.258,
+      "step": 13885
+    },
+    {
+      "epoch": 2.3707433643424083,
+      "grad_norm": 0.3567861355385066,
+      "learning_rate": 1.1646433990895296e-05,
+      "loss": 0.2684,
+      "step": 13890
+    },
+    {
+      "epoch": 2.3715968251258857,
+      "grad_norm": 0.3622011831997506,
+      "learning_rate": 1.1630627212948913e-05,
+      "loss": 0.279,
+      "step": 13895
+    },
+    {
+      "epoch": 2.3724502859093626,
+      "grad_norm": 0.3329058907180884,
+      "learning_rate": 1.1614820435002529e-05,
+      "loss": 0.2448,
+      "step": 13900
+    },
+    {
+      "epoch": 2.3733037466928395,
+      "grad_norm": 0.47032289909689734,
+      "learning_rate": 1.1599013657056147e-05,
+      "loss": 0.2876,
+      "step": 13905
+    },
+    {
+      "epoch": 2.3741572074763164,
+      "grad_norm": 0.3102544235708446,
+      "learning_rate": 1.1583206879109764e-05,
+      "loss": 0.2461,
+      "step": 13910
+    },
+    {
+      "epoch": 2.3750106682597933,
+      "grad_norm": 0.31929100036653174,
+      "learning_rate": 1.156740010116338e-05,
+      "loss": 0.2419,
+      "step": 13915
+    },
+    {
+      "epoch": 2.3758641290432703,
+      "grad_norm": 0.35223857287658794,
+      "learning_rate": 1.1551593323216996e-05,
+      "loss": 0.2897,
+      "step": 13920
+    },
+    {
+      "epoch": 2.3767175898267476,
+      "grad_norm": 0.31929075368848847,
+      "learning_rate": 1.1535786545270613e-05,
+      "loss": 0.2653,
+      "step": 13925
+    },
+    {
+      "epoch": 2.3775710506102246,
+      "grad_norm": 0.5705191250133481,
+      "learning_rate": 1.1519979767324229e-05,
+      "loss": 0.2505,
+      "step": 13930
+    },
+    {
+      "epoch": 2.3784245113937015,
+      "grad_norm": 0.32864593989844176,
+      "learning_rate": 1.1504172989377846e-05,
+      "loss": 0.2477,
+      "step": 13935
+    },
+    {
+      "epoch": 2.3792779721771784,
+      "grad_norm": 0.3760333483534563,
+      "learning_rate": 1.1488366211431464e-05,
+      "loss": 0.242,
+      "step": 13940
+    },
+    {
+      "epoch": 2.3801314329606553,
+      "grad_norm": 0.30221859215315,
+      "learning_rate": 1.147255943348508e-05,
+      "loss": 0.2675,
+      "step": 13945
+    },
+    {
+      "epoch": 2.3809848937441327,
+      "grad_norm": 0.3094129613854578,
+      "learning_rate": 1.1456752655538695e-05,
+      "loss": 0.2855,
+      "step": 13950
+    },
+    {
+      "epoch": 2.3818383545276096,
+      "grad_norm": 0.3118239798941746,
+      "learning_rate": 1.1440945877592311e-05,
+      "loss": 0.2509,
+      "step": 13955
+    },
+    {
+      "epoch": 2.3826918153110865,
+      "grad_norm": 0.33807031265894355,
+      "learning_rate": 1.1425139099645928e-05,
+      "loss": 0.2519,
+      "step": 13960
+    },
+    {
+      "epoch": 2.3835452760945635,
+      "grad_norm": 0.3589548365905337,
+      "learning_rate": 1.1409332321699546e-05,
+      "loss": 0.2626,
+      "step": 13965
+    },
+    {
+      "epoch": 2.3843987368780404,
+      "grad_norm": 0.3601196386452014,
+      "learning_rate": 1.1393525543753162e-05,
+      "loss": 0.2368,
+      "step": 13970
+    },
+    {
+      "epoch": 2.3852521976615173,
+      "grad_norm": 0.37435014821697704,
+      "learning_rate": 1.1377718765806779e-05,
+      "loss": 0.2737,
+      "step": 13975
+    },
+    {
+      "epoch": 2.3861056584449942,
+      "grad_norm": 0.39186184407312485,
+      "learning_rate": 1.1361911987860395e-05,
+      "loss": 0.2549,
+      "step": 13980
+    },
+    {
+      "epoch": 2.3869591192284716,
+      "grad_norm": 0.3480509116803668,
+      "learning_rate": 1.134610520991401e-05,
+      "loss": 0.2747,
+      "step": 13985
+    },
+    {
+      "epoch": 2.3878125800119485,
+      "grad_norm": 0.3108613164406919,
+      "learning_rate": 1.1330298431967628e-05,
+      "loss": 0.2721,
+      "step": 13990
+    },
+    {
+      "epoch": 2.3886660407954254,
+      "grad_norm": 0.33249119855304893,
+      "learning_rate": 1.1314491654021246e-05,
+      "loss": 0.2617,
+      "step": 13995
+    },
+    {
+      "epoch": 2.3895195015789024,
+      "grad_norm": 2.6530979762285525,
+      "learning_rate": 1.1298684876074861e-05,
+      "loss": 0.301,
+      "step": 14000
+    },
+    {
+      "epoch": 2.3903729623623793,
+      "grad_norm": 0.658749679558768,
+      "learning_rate": 1.1282878098128479e-05,
+      "loss": 0.2676,
+      "step": 14005
+    },
+    {
+      "epoch": 2.3912264231458566,
+      "grad_norm": 0.36243360222353593,
+      "learning_rate": 1.1267071320182095e-05,
+      "loss": 0.2522,
+      "step": 14010
+    },
+    {
+      "epoch": 2.3920798839293336,
+      "grad_norm": 0.3763675627463445,
+      "learning_rate": 1.125126454223571e-05,
+      "loss": 0.2512,
+      "step": 14015
+    },
+    {
+      "epoch": 2.3929333447128105,
+      "grad_norm": 0.4142890979332789,
+      "learning_rate": 1.1235457764289328e-05,
+      "loss": 0.2677,
+      "step": 14020
+    },
+    {
+      "epoch": 2.3937868054962874,
+      "grad_norm": 0.3439817539904734,
+      "learning_rate": 1.1219650986342945e-05,
+      "loss": 0.2736,
+      "step": 14025
+    },
+    {
+      "epoch": 2.3946402662797643,
+      "grad_norm": 1.8155640993639846,
+      "learning_rate": 1.1203844208396561e-05,
+      "loss": 0.2408,
+      "step": 14030
+    },
+    {
+      "epoch": 2.3954937270632413,
+      "grad_norm": 0.32729231033143624,
+      "learning_rate": 1.1188037430450178e-05,
+      "loss": 0.2655,
+      "step": 14035
+    },
+    {
+      "epoch": 2.3963471878467186,
+      "grad_norm": 0.3273285197702505,
+      "learning_rate": 1.1172230652503794e-05,
+      "loss": 0.26,
+      "step": 14040
+    },
+    {
+      "epoch": 2.3972006486301956,
+      "grad_norm": 0.30793617073489615,
+      "learning_rate": 1.115642387455741e-05,
+      "loss": 0.2903,
+      "step": 14045
+    },
+    {
+      "epoch": 2.3980541094136725,
+      "grad_norm": 0.36727247145711506,
+      "learning_rate": 1.1140617096611027e-05,
+      "loss": 0.2849,
+      "step": 14050
+    },
+    {
+      "epoch": 2.3989075701971494,
+      "grad_norm": 0.4386468565369713,
+      "learning_rate": 1.1124810318664645e-05,
+      "loss": 0.2312,
+      "step": 14055
+    },
+    {
+      "epoch": 2.3997610309806263,
+      "grad_norm": 0.3906762379869074,
+      "learning_rate": 1.110900354071826e-05,
+      "loss": 0.2691,
+      "step": 14060
+    },
+    {
+      "epoch": 2.4006144917641032,
+      "grad_norm": 0.38600382005800393,
+      "learning_rate": 1.1093196762771878e-05,
+      "loss": 0.2972,
+      "step": 14065
+    },
+    {
+      "epoch": 2.4014679525475806,
+      "grad_norm": 0.3425366869475592,
+      "learning_rate": 1.1077389984825494e-05,
+      "loss": 0.2817,
+      "step": 14070
+    },
+    {
+      "epoch": 2.4023214133310575,
+      "grad_norm": 0.370235777382759,
+      "learning_rate": 1.106158320687911e-05,
+      "loss": 0.2596,
+      "step": 14075
+    },
+    {
+      "epoch": 2.4031748741145345,
+      "grad_norm": 0.30431484417216215,
+      "learning_rate": 1.1045776428932727e-05,
+      "loss": 0.2554,
+      "step": 14080
+    },
+    {
+      "epoch": 2.4040283348980114,
+      "grad_norm": 0.35170746169540634,
+      "learning_rate": 1.1029969650986345e-05,
+      "loss": 0.273,
+      "step": 14085
+    },
+    {
+      "epoch": 2.4048817956814883,
+      "grad_norm": 0.3469725693929598,
+      "learning_rate": 1.101416287303996e-05,
+      "loss": 0.2758,
+      "step": 14090
+    },
+    {
+      "epoch": 2.4057352564649657,
+      "grad_norm": 0.37712041067128005,
+      "learning_rate": 1.0998356095093576e-05,
+      "loss": 0.2824,
+      "step": 14095
+    },
+    {
+      "epoch": 2.4065887172484426,
+      "grad_norm": 0.3538335750117501,
+      "learning_rate": 1.0982549317147194e-05,
+      "loss": 0.2575,
+      "step": 14100
+    },
+    {
+      "epoch": 2.4074421780319195,
+      "grad_norm": 0.3320536986234714,
+      "learning_rate": 1.096674253920081e-05,
+      "loss": 0.2782,
+      "step": 14105
+    },
+    {
+      "epoch": 2.4082956388153964,
+      "grad_norm": 0.47800365739823614,
+      "learning_rate": 1.0950935761254425e-05,
+      "loss": 0.2607,
+      "step": 14110
+    },
+    {
+      "epoch": 2.4091490995988734,
+      "grad_norm": 0.3550919226992691,
+      "learning_rate": 1.0935128983308044e-05,
+      "loss": 0.2737,
+      "step": 14115
+    },
+    {
+      "epoch": 2.4100025603823503,
+      "grad_norm": 0.319973289046971,
+      "learning_rate": 1.091932220536166e-05,
+      "loss": 0.2918,
+      "step": 14120
+    },
+    {
+      "epoch": 2.410856021165827,
+      "grad_norm": 0.6546241723107031,
+      "learning_rate": 1.0903515427415276e-05,
+      "loss": 0.2589,
+      "step": 14125
+    },
+    {
+      "epoch": 2.4117094819493046,
+      "grad_norm": 0.34912701414952313,
+      "learning_rate": 1.0887708649468893e-05,
+      "loss": 0.2749,
+      "step": 14130
+    },
+    {
+      "epoch": 2.4125629427327815,
+      "grad_norm": 0.31161437564403516,
+      "learning_rate": 1.0871901871522509e-05,
+      "loss": 0.2402,
+      "step": 14135
+    },
+    {
+      "epoch": 2.4134164035162584,
+      "grad_norm": 0.3596705941470503,
+      "learning_rate": 1.0856095093576125e-05,
+      "loss": 0.2608,
+      "step": 14140
+    },
+    {
+      "epoch": 2.4142698642997353,
+      "grad_norm": 0.38807872345735817,
+      "learning_rate": 1.0840288315629742e-05,
+      "loss": 0.2654,
+      "step": 14145
+    },
+    {
+      "epoch": 2.4151233250832123,
+      "grad_norm": 0.3543470511620338,
+      "learning_rate": 1.082448153768336e-05,
+      "loss": 0.2535,
+      "step": 14150
+    },
+    {
+      "epoch": 2.4159767858666896,
+      "grad_norm": 0.30110137348341887,
+      "learning_rate": 1.0808674759736975e-05,
+      "loss": 0.2792,
+      "step": 14155
+    },
+    {
+      "epoch": 2.4168302466501665,
+      "grad_norm": 0.34315261887349163,
+      "learning_rate": 1.0792867981790593e-05,
+      "loss": 0.2511,
+      "step": 14160
+    },
+    {
+      "epoch": 2.4176837074336435,
+      "grad_norm": 0.35974940683996065,
+      "learning_rate": 1.0777061203844209e-05,
+      "loss": 0.2568,
+      "step": 14165
+    },
+    {
+      "epoch": 2.4185371682171204,
+      "grad_norm": 0.3416035040366801,
+      "learning_rate": 1.0761254425897824e-05,
+      "loss": 0.2677,
+      "step": 14170
+    },
+    {
+      "epoch": 2.4193906290005973,
+      "grad_norm": 0.40959594994649545,
+      "learning_rate": 1.0745447647951442e-05,
+      "loss": 0.2619,
+      "step": 14175
+    },
+    {
+      "epoch": 2.4202440897840742,
+      "grad_norm": 0.34899014164854364,
+      "learning_rate": 1.072964087000506e-05,
+      "loss": 0.2788,
+      "step": 14180
+    },
+    {
+      "epoch": 2.4210975505675516,
+      "grad_norm": 0.34914763029264057,
+      "learning_rate": 1.0713834092058675e-05,
+      "loss": 0.263,
+      "step": 14185
+    },
+    {
+      "epoch": 2.4219510113510285,
+      "grad_norm": 0.38033095181588245,
+      "learning_rate": 1.0698027314112293e-05,
+      "loss": 0.2658,
+      "step": 14190
+    },
+    {
+      "epoch": 2.4228044721345054,
+      "grad_norm": 0.3394889343943498,
+      "learning_rate": 1.0682220536165908e-05,
+      "loss": 0.2848,
+      "step": 14195
+    },
+    {
+      "epoch": 2.4236579329179824,
+      "grad_norm": 0.3760457990875017,
+      "learning_rate": 1.0666413758219524e-05,
+      "loss": 0.2537,
+      "step": 14200
+    },
+    {
+      "epoch": 2.4245113937014593,
+      "grad_norm": 0.3380987792591969,
+      "learning_rate": 1.0650606980273141e-05,
+      "loss": 0.2644,
+      "step": 14205
+    },
+    {
+      "epoch": 2.425364854484936,
+      "grad_norm": 0.318549850666751,
+      "learning_rate": 1.0634800202326759e-05,
+      "loss": 0.2574,
+      "step": 14210
+    },
+    {
+      "epoch": 2.4262183152684136,
+      "grad_norm": 0.31342348190936237,
+      "learning_rate": 1.0618993424380375e-05,
+      "loss": 0.2642,
+      "step": 14215
+    },
+    {
+      "epoch": 2.4270717760518905,
+      "grad_norm": 0.3462677540696492,
+      "learning_rate": 1.060318664643399e-05,
+      "loss": 0.2537,
+      "step": 14220
+    },
+    {
+      "epoch": 2.4279252368353674,
+      "grad_norm": 0.3826866159391629,
+      "learning_rate": 1.0587379868487608e-05,
+      "loss": 0.2742,
+      "step": 14225
+    },
+    {
+      "epoch": 2.4287786976188444,
+      "grad_norm": 0.3263531749919483,
+      "learning_rate": 1.0571573090541224e-05,
+      "loss": 0.2882,
+      "step": 14230
+    },
+    {
+      "epoch": 2.4296321584023213,
+      "grad_norm": 0.28544342560666724,
+      "learning_rate": 1.0555766312594841e-05,
+      "loss": 0.2674,
+      "step": 14235
+    },
+    {
+      "epoch": 2.4304856191857986,
+      "grad_norm": 0.3396665329122308,
+      "learning_rate": 1.0539959534648459e-05,
+      "loss": 0.2787,
+      "step": 14240
+    },
+    {
+      "epoch": 2.4313390799692756,
+      "grad_norm": 0.3399252000182434,
+      "learning_rate": 1.0524152756702074e-05,
+      "loss": 0.2667,
+      "step": 14245
+    },
+    {
+      "epoch": 2.4321925407527525,
+      "grad_norm": 0.3247161010293173,
+      "learning_rate": 1.050834597875569e-05,
+      "loss": 0.2591,
+      "step": 14250
+    },
+    {
+      "epoch": 2.4330460015362294,
+      "grad_norm": 0.36080600895528747,
+      "learning_rate": 1.0492539200809308e-05,
+      "loss": 0.258,
+      "step": 14255
+    },
+    {
+      "epoch": 2.4338994623197063,
+      "grad_norm": 0.3464012982629838,
+      "learning_rate": 1.0476732422862923e-05,
+      "loss": 0.2562,
+      "step": 14260
+    },
+    {
+      "epoch": 2.4347529231031833,
+      "grad_norm": 0.328033884048764,
+      "learning_rate": 1.046092564491654e-05,
+      "loss": 0.2595,
+      "step": 14265
+    },
+    {
+      "epoch": 2.43560638388666,
+      "grad_norm": 0.34044030311146867,
+      "learning_rate": 1.0445118866970158e-05,
+      "loss": 0.2701,
+      "step": 14270
+    },
+    {
+      "epoch": 2.4364598446701375,
+      "grad_norm": 0.3787577646980057,
+      "learning_rate": 1.0429312089023774e-05,
+      "loss": 0.2757,
+      "step": 14275
+    },
+    {
+      "epoch": 2.4373133054536145,
+      "grad_norm": 0.4621132454628748,
+      "learning_rate": 1.041350531107739e-05,
+      "loss": 0.2706,
+      "step": 14280
+    },
+    {
+      "epoch": 2.4381667662370914,
+      "grad_norm": 0.3699920523962295,
+      "learning_rate": 1.0397698533131007e-05,
+      "loss": 0.2467,
+      "step": 14285
+    },
+    {
+      "epoch": 2.4390202270205683,
+      "grad_norm": 0.37567265417832557,
+      "learning_rate": 1.0381891755184623e-05,
+      "loss": 0.2841,
+      "step": 14290
+    },
+    {
+      "epoch": 2.4398736878040452,
+      "grad_norm": 0.31674098891327795,
+      "learning_rate": 1.036608497723824e-05,
+      "loss": 0.2765,
+      "step": 14295
+    },
+    {
+      "epoch": 2.4407271485875226,
+      "grad_norm": 0.3683870689428794,
+      "learning_rate": 1.0350278199291858e-05,
+      "loss": 0.2566,
+      "step": 14300
+    },
+    {
+      "epoch": 2.4415806093709995,
+      "grad_norm": 0.3312211455701073,
+      "learning_rate": 1.0334471421345474e-05,
+      "loss": 0.2854,
+      "step": 14305
+    },
+    {
+      "epoch": 2.4424340701544764,
+      "grad_norm": 0.3047702631079437,
+      "learning_rate": 1.031866464339909e-05,
+      "loss": 0.2713,
+      "step": 14310
+    },
+    {
+      "epoch": 2.4432875309379534,
+      "grad_norm": 0.29809655107369376,
+      "learning_rate": 1.0302857865452707e-05,
+      "loss": 0.2606,
+      "step": 14315
+    },
+    {
+      "epoch": 2.4441409917214303,
+      "grad_norm": 0.3742642086400778,
+      "learning_rate": 1.0287051087506323e-05,
+      "loss": 0.2688,
+      "step": 14320
+    },
+    {
+      "epoch": 2.444994452504907,
+      "grad_norm": 0.3722878341115334,
+      "learning_rate": 1.027124430955994e-05,
+      "loss": 0.2607,
+      "step": 14325
+    },
+    {
+      "epoch": 2.4458479132883846,
+      "grad_norm": 0.31636039883162703,
+      "learning_rate": 1.0255437531613558e-05,
+      "loss": 0.2713,
+      "step": 14330
+    },
+    {
+      "epoch": 2.4467013740718615,
+      "grad_norm": 0.3528678978404139,
+      "learning_rate": 1.0239630753667173e-05,
+      "loss": 0.269,
+      "step": 14335
+    },
+    {
+      "epoch": 2.4475548348553384,
+      "grad_norm": 0.33220518954122014,
+      "learning_rate": 1.0223823975720789e-05,
+      "loss": 0.2773,
+      "step": 14340
+    },
+    {
+      "epoch": 2.4484082956388153,
+      "grad_norm": 0.3433486949077163,
+      "learning_rate": 1.0208017197774405e-05,
+      "loss": 0.224,
+      "step": 14345
+    },
+    {
+      "epoch": 2.4492617564222923,
+      "grad_norm": 0.31614126849766067,
+      "learning_rate": 1.0192210419828022e-05,
+      "loss": 0.269,
+      "step": 14350
+    },
+    {
+      "epoch": 2.450115217205769,
+      "grad_norm": 0.3171829857680147,
+      "learning_rate": 1.017640364188164e-05,
+      "loss": 0.284,
+      "step": 14355
+    },
+    {
+      "epoch": 2.4509686779892466,
+      "grad_norm": 0.36265633846142,
+      "learning_rate": 1.0160596863935256e-05,
+      "loss": 0.2498,
+      "step": 14360
+    },
+    {
+      "epoch": 2.4518221387727235,
+      "grad_norm": 0.3282258036126791,
+      "learning_rate": 1.0144790085988873e-05,
+      "loss": 0.2704,
+      "step": 14365
+    },
+    {
+      "epoch": 2.4526755995562004,
+      "grad_norm": 0.35214370660567523,
+      "learning_rate": 1.0128983308042489e-05,
+      "loss": 0.2869,
+      "step": 14370
+    },
+    {
+      "epoch": 2.4535290603396773,
+      "grad_norm": 0.3363280362651704,
+      "learning_rate": 1.0113176530096105e-05,
+      "loss": 0.2637,
+      "step": 14375
+    },
+    {
+      "epoch": 2.4543825211231542,
+      "grad_norm": 0.39213689524643197,
+      "learning_rate": 1.0097369752149722e-05,
+      "loss": 0.2701,
+      "step": 14380
+    },
+    {
+      "epoch": 2.4552359819066316,
+      "grad_norm": 0.343905126836761,
+      "learning_rate": 1.008156297420334e-05,
+      "loss": 0.253,
+      "step": 14385
+    },
+    {
+      "epoch": 2.4560894426901085,
+      "grad_norm": 0.34344106048561895,
+      "learning_rate": 1.0065756196256955e-05,
+      "loss": 0.258,
+      "step": 14390
+    },
+    {
+      "epoch": 2.4569429034735855,
+      "grad_norm": 0.29586278047989606,
+      "learning_rate": 1.0049949418310573e-05,
+      "loss": 0.273,
+      "step": 14395
+    },
+    {
+      "epoch": 2.4577963642570624,
+      "grad_norm": 0.4027576698874579,
+      "learning_rate": 1.0034142640364188e-05,
+      "loss": 0.2648,
+      "step": 14400
+    },
+    {
+      "epoch": 2.4586498250405393,
+      "grad_norm": 0.36593665379768425,
+      "learning_rate": 1.0018335862417804e-05,
+      "loss": 0.2557,
+      "step": 14405
+    },
+    {
+      "epoch": 2.4595032858240162,
+      "grad_norm": 0.3598055453128078,
+      "learning_rate": 1.0002529084471422e-05,
+      "loss": 0.2439,
+      "step": 14410
+    },
+    {
+      "epoch": 2.460356746607493,
+      "grad_norm": 0.38253852146152434,
+      "learning_rate": 9.986722306525039e-06,
+      "loss": 0.2587,
+      "step": 14415
+    },
+    {
+      "epoch": 2.4612102073909705,
+      "grad_norm": 0.30682301559292546,
+      "learning_rate": 9.970915528578655e-06,
+      "loss": 0.255,
+      "step": 14420
+    },
+    {
+      "epoch": 2.4620636681744474,
+      "grad_norm": 0.346241220452941,
+      "learning_rate": 9.955108750632272e-06,
+      "loss": 0.2609,
+      "step": 14425
+    },
+    {
+      "epoch": 2.4629171289579244,
+      "grad_norm": 0.39098463135896994,
+      "learning_rate": 9.939301972685888e-06,
+      "loss": 0.2614,
+      "step": 14430
+    },
+    {
+      "epoch": 2.4637705897414013,
+      "grad_norm": 0.36311594138575676,
+      "learning_rate": 9.923495194739504e-06,
+      "loss": 0.2663,
+      "step": 14435
+    },
+    {
+      "epoch": 2.464624050524878,
+      "grad_norm": 0.3694187382426125,
+      "learning_rate": 9.907688416793121e-06,
+      "loss": 0.2725,
+      "step": 14440
+    },
+    {
+      "epoch": 2.4654775113083556,
+      "grad_norm": 0.3181368980124899,
+      "learning_rate": 9.891881638846739e-06,
+      "loss": 0.273,
+      "step": 14445
+    },
+    {
+      "epoch": 2.4663309720918325,
+      "grad_norm": 0.3001019744638354,
+      "learning_rate": 9.876074860900355e-06,
+      "loss": 0.2496,
+      "step": 14450
+    },
+    {
+      "epoch": 2.4671844328753094,
+      "grad_norm": 0.31428730167056335,
+      "learning_rate": 9.860268082953972e-06,
+      "loss": 0.2554,
+      "step": 14455
+    },
+    {
+      "epoch": 2.4680378936587863,
+      "grad_norm": 0.3712724482982424,
+      "learning_rate": 9.844461305007588e-06,
+      "loss": 0.2833,
+      "step": 14460
+    },
+    {
+      "epoch": 2.4688913544422633,
+      "grad_norm": 0.44433449011826265,
+      "learning_rate": 9.828654527061204e-06,
+      "loss": 0.2721,
+      "step": 14465
+    },
+    {
+      "epoch": 2.46974481522574,
+      "grad_norm": 0.32849313014573395,
+      "learning_rate": 9.812847749114821e-06,
+      "loss": 0.2854,
+      "step": 14470
+    },
+    {
+      "epoch": 2.4705982760092176,
+      "grad_norm": 0.3440790071717394,
+      "learning_rate": 9.797040971168438e-06,
+      "loss": 0.2573,
+      "step": 14475
+    },
+    {
+      "epoch": 2.4714517367926945,
+      "grad_norm": 0.3587060980032648,
+      "learning_rate": 9.781234193222054e-06,
+      "loss": 0.2358,
+      "step": 14480
+    },
+    {
+      "epoch": 2.4723051975761714,
+      "grad_norm": 0.3348288455825785,
+      "learning_rate": 9.76542741527567e-06,
+      "loss": 0.2786,
+      "step": 14485
+    },
+    {
+      "epoch": 2.4731586583596483,
+      "grad_norm": 0.34934828195955436,
+      "learning_rate": 9.749620637329287e-06,
+      "loss": 0.2677,
+      "step": 14490
+    },
+    {
+      "epoch": 2.4740121191431252,
+      "grad_norm": 0.8305033574248635,
+      "learning_rate": 9.733813859382903e-06,
+      "loss": 0.2619,
+      "step": 14495
+    },
+    {
+      "epoch": 2.474865579926602,
+      "grad_norm": 0.3580936459587481,
+      "learning_rate": 9.71800708143652e-06,
+      "loss": 0.2549,
+      "step": 14500
+    },
+    {
+      "epoch": 2.4757190407100795,
+      "grad_norm": 0.31674144817736993,
+      "learning_rate": 9.702200303490138e-06,
+      "loss": 0.2696,
+      "step": 14505
+    },
+    {
+      "epoch": 2.4765725014935565,
+      "grad_norm": 0.33296233907632666,
+      "learning_rate": 9.686393525543754e-06,
+      "loss": 0.277,
+      "step": 14510
+    },
+    {
+      "epoch": 2.4774259622770334,
+      "grad_norm": 0.3625925072432881,
+      "learning_rate": 9.67058674759737e-06,
+      "loss": 0.2552,
+      "step": 14515
+    },
+    {
+      "epoch": 2.4782794230605103,
+      "grad_norm": 0.3052866481204275,
+      "learning_rate": 9.654779969650987e-06,
+      "loss": 0.2745,
+      "step": 14520
+    },
+    {
+      "epoch": 2.4791328838439872,
+      "grad_norm": 0.3436194097126335,
+      "learning_rate": 9.638973191704603e-06,
+      "loss": 0.2821,
+      "step": 14525
+    },
+    {
+      "epoch": 2.4799863446274646,
+      "grad_norm": 0.3954215609284519,
+      "learning_rate": 9.623166413758219e-06,
+      "loss": 0.2558,
+      "step": 14530
+    },
+    {
+      "epoch": 2.4808398054109415,
+      "grad_norm": 0.4587324591471796,
+      "learning_rate": 9.607359635811838e-06,
+      "loss": 0.2469,
+      "step": 14535
+    },
+    {
+      "epoch": 2.4816932661944184,
+      "grad_norm": 0.3269705637376026,
+      "learning_rate": 9.591552857865454e-06,
+      "loss": 0.2672,
+      "step": 14540
+    },
+    {
+      "epoch": 2.4825467269778954,
+      "grad_norm": 0.3725948745994985,
+      "learning_rate": 9.57574607991907e-06,
+      "loss": 0.2642,
+      "step": 14545
+    },
+    {
+      "epoch": 2.4834001877613723,
+      "grad_norm": 0.4251797756192436,
+      "learning_rate": 9.559939301972687e-06,
+      "loss": 0.2442,
+      "step": 14550
+    },
+    {
+      "epoch": 2.484253648544849,
+      "grad_norm": 0.35811671563058445,
+      "learning_rate": 9.544132524026303e-06,
+      "loss": 0.3074,
+      "step": 14555
+    },
+    {
+      "epoch": 2.485107109328326,
+      "grad_norm": 0.3747974354254869,
+      "learning_rate": 9.528325746079918e-06,
+      "loss": 0.2693,
+      "step": 14560
+    },
+    {
+      "epoch": 2.4859605701118035,
+      "grad_norm": 0.33567451968078693,
+      "learning_rate": 9.512518968133537e-06,
+      "loss": 0.2384,
+      "step": 14565
+    },
+    {
+      "epoch": 2.4868140308952804,
+      "grad_norm": 0.3189977713488994,
+      "learning_rate": 9.496712190187153e-06,
+      "loss": 0.2626,
+      "step": 14570
+    },
+    {
+      "epoch": 2.4876674916787573,
+      "grad_norm": 0.3489836368302477,
+      "learning_rate": 9.480905412240769e-06,
+      "loss": 0.2523,
+      "step": 14575
+    },
+    {
+      "epoch": 2.4885209524622343,
+      "grad_norm": 0.3613479235819589,
+      "learning_rate": 9.465098634294386e-06,
+      "loss": 0.2729,
+      "step": 14580
+    },
+    {
+      "epoch": 2.489374413245711,
+      "grad_norm": 0.40565449622708155,
+      "learning_rate": 9.449291856348002e-06,
+      "loss": 0.2444,
+      "step": 14585
+    },
+    {
+      "epoch": 2.4902278740291885,
+      "grad_norm": 0.3680615304096154,
+      "learning_rate": 9.433485078401618e-06,
+      "loss": 0.2566,
+      "step": 14590
+    },
+    {
+      "epoch": 2.4910813348126655,
+      "grad_norm": 0.3128924900072794,
+      "learning_rate": 9.417678300455237e-06,
+      "loss": 0.2318,
+      "step": 14595
+    },
+    {
+      "epoch": 2.4919347955961424,
+      "grad_norm": 0.3186196091002627,
+      "learning_rate": 9.401871522508853e-06,
+      "loss": 0.2822,
+      "step": 14600
+    },
+    {
+      "epoch": 2.4927882563796193,
+      "grad_norm": 0.3287136069133009,
+      "learning_rate": 9.386064744562469e-06,
+      "loss": 0.2533,
+      "step": 14605
+    },
+    {
+      "epoch": 2.4936417171630962,
+      "grad_norm": 0.4396657971919846,
+      "learning_rate": 9.370257966616084e-06,
+      "loss": 0.2827,
+      "step": 14610
+    },
+    {
+      "epoch": 2.494495177946573,
+      "grad_norm": 0.3124085405558122,
+      "learning_rate": 9.354451188669702e-06,
+      "loss": 0.2409,
+      "step": 14615
+    },
+    {
+      "epoch": 2.4953486387300505,
+      "grad_norm": 0.36168954024164707,
+      "learning_rate": 9.338644410723318e-06,
+      "loss": 0.2791,
+      "step": 14620
+    },
+    {
+      "epoch": 2.4962020995135275,
+      "grad_norm": 0.3471248448213307,
+      "learning_rate": 9.322837632776935e-06,
+      "loss": 0.278,
+      "step": 14625
+    },
+    {
+      "epoch": 2.4970555602970044,
+      "grad_norm": 0.3520183299863951,
+      "learning_rate": 9.307030854830553e-06,
+      "loss": 0.2674,
+      "step": 14630
+    },
+    {
+      "epoch": 2.4979090210804813,
+      "grad_norm": 0.326191389335285,
+      "learning_rate": 9.291224076884168e-06,
+      "loss": 0.2542,
+      "step": 14635
+    },
+    {
+      "epoch": 2.498762481863958,
+      "grad_norm": 0.3220048844992483,
+      "learning_rate": 9.275417298937784e-06,
+      "loss": 0.2532,
+      "step": 14640
+    },
+    {
+      "epoch": 2.499615942647435,
+      "grad_norm": 0.33818063700540074,
+      "learning_rate": 9.259610520991402e-06,
+      "loss": 0.279,
+      "step": 14645
+    },
+    {
+      "epoch": 2.5004694034309125,
+      "grad_norm": 0.33944084356888454,
+      "learning_rate": 9.243803743045017e-06,
+      "loss": 0.2603,
+      "step": 14650
+    },
+    {
+      "epoch": 2.5013228642143894,
+      "grad_norm": 0.7220901229701975,
+      "learning_rate": 9.227996965098635e-06,
+      "loss": 0.2567,
+      "step": 14655
+    },
+    {
+      "epoch": 2.5021763249978664,
+      "grad_norm": 0.3115755105531366,
+      "learning_rate": 9.212190187152252e-06,
+      "loss": 0.2495,
+      "step": 14660
+    },
+    {
+      "epoch": 2.5030297857813433,
+      "grad_norm": 0.3498764937461434,
+      "learning_rate": 9.196383409205868e-06,
+      "loss": 0.2939,
+      "step": 14665
+    },
+    {
+      "epoch": 2.50388324656482,
+      "grad_norm": 0.4011131591215743,
+      "learning_rate": 9.180576631259484e-06,
+      "loss": 0.2477,
+      "step": 14670
+    },
+    {
+      "epoch": 2.5047367073482976,
+      "grad_norm": 0.3447402466363007,
+      "learning_rate": 9.164769853313101e-06,
+      "loss": 0.2782,
+      "step": 14675
+    },
+    {
+      "epoch": 2.5055901681317745,
+      "grad_norm": 0.3408437723228539,
+      "learning_rate": 9.148963075366717e-06,
+      "loss": 0.2883,
+      "step": 14680
+    },
+    {
+      "epoch": 2.5064436289152514,
+      "grad_norm": 0.3598563871864662,
+      "learning_rate": 9.133156297420334e-06,
+      "loss": 0.2661,
+      "step": 14685
+    },
+    {
+      "epoch": 2.5072970896987283,
+      "grad_norm": 0.3524261962338911,
+      "learning_rate": 9.117349519473952e-06,
+      "loss": 0.2299,
+      "step": 14690
+    },
+    {
+      "epoch": 2.5081505504822053,
+      "grad_norm": 0.35548469805642413,
+      "learning_rate": 9.101542741527568e-06,
+      "loss": 0.2725,
+      "step": 14695
+    },
+    {
+      "epoch": 2.509004011265682,
+      "grad_norm": 0.312906336303122,
+      "learning_rate": 9.085735963581183e-06,
+      "loss": 0.26,
+      "step": 14700
+    },
+    {
+      "epoch": 2.509857472049159,
+      "grad_norm": 0.3389670840404128,
+      "learning_rate": 9.069929185634801e-06,
+      "loss": 0.2735,
+      "step": 14705
+    },
+    {
+      "epoch": 2.5107109328326365,
+      "grad_norm": 0.3364038657527453,
+      "learning_rate": 9.054122407688417e-06,
+      "loss": 0.27,
+      "step": 14710
+    },
+    {
+      "epoch": 2.5115643936161134,
+      "grad_norm": 0.32104461280395497,
+      "learning_rate": 9.038315629742034e-06,
+      "loss": 0.2838,
+      "step": 14715
+    },
+    {
+      "epoch": 2.5124178543995903,
+      "grad_norm": 0.4144744341346533,
+      "learning_rate": 9.022508851795652e-06,
+      "loss": 0.2545,
+      "step": 14720
+    },
+    {
+      "epoch": 2.5132713151830672,
+      "grad_norm": 0.3319490971303601,
+      "learning_rate": 9.006702073849267e-06,
+      "loss": 0.2625,
+      "step": 14725
+    },
+    {
+      "epoch": 2.514124775966544,
+      "grad_norm": 0.37661552718672137,
+      "learning_rate": 8.990895295902883e-06,
+      "loss": 0.2497,
+      "step": 14730
+    },
+    {
+      "epoch": 2.5149782367500215,
+      "grad_norm": 0.32876186506762023,
+      "learning_rate": 8.9750885179565e-06,
+      "loss": 0.2433,
+      "step": 14735
+    },
+    {
+      "epoch": 2.5158316975334984,
+      "grad_norm": 0.43033153828122256,
+      "learning_rate": 8.959281740010116e-06,
+      "loss": 0.2522,
+      "step": 14740
+    },
+    {
+      "epoch": 2.5166851583169754,
+      "grad_norm": 0.3640377089384684,
+      "learning_rate": 8.943474962063734e-06,
+      "loss": 0.2555,
+      "step": 14745
+    },
+    {
+      "epoch": 2.5175386191004523,
+      "grad_norm": 0.31362572989653836,
+      "learning_rate": 8.92766818411735e-06,
+      "loss": 0.2837,
+      "step": 14750
+    },
+    {
+      "epoch": 2.518392079883929,
+      "grad_norm": 0.40458398017180186,
+      "learning_rate": 8.911861406170967e-06,
+      "loss": 0.2629,
+      "step": 14755
+    },
+    {
+      "epoch": 2.5192455406674066,
+      "grad_norm": 0.3248341879661295,
+      "learning_rate": 8.896054628224583e-06,
+      "loss": 0.2743,
+      "step": 14760
+    },
+    {
+      "epoch": 2.520099001450883,
+      "grad_norm": 0.3578432284103998,
+      "learning_rate": 8.880247850278198e-06,
+      "loss": 0.2507,
+      "step": 14765
+    },
+    {
+      "epoch": 2.5209524622343604,
+      "grad_norm": 0.35467076421052923,
+      "learning_rate": 8.864441072331816e-06,
+      "loss": 0.2837,
+      "step": 14770
+    },
+    {
+      "epoch": 2.5218059230178373,
+      "grad_norm": 0.4006464151945454,
+      "learning_rate": 8.848634294385433e-06,
+      "loss": 0.2741,
+      "step": 14775
+    },
+    {
+      "epoch": 2.5226593838013143,
+      "grad_norm": 0.29135652611912566,
+      "learning_rate": 8.83282751643905e-06,
+      "loss": 0.2738,
+      "step": 14780
+    },
+    {
+      "epoch": 2.523512844584791,
+      "grad_norm": 0.3323270796757813,
+      "learning_rate": 8.817020738492667e-06,
+      "loss": 0.2684,
+      "step": 14785
+    },
+    {
+      "epoch": 2.524366305368268,
+      "grad_norm": 0.3291607431817088,
+      "learning_rate": 8.801213960546282e-06,
+      "loss": 0.2821,
+      "step": 14790
+    },
+    {
+      "epoch": 2.5252197661517455,
+      "grad_norm": 0.32039698349844725,
+      "learning_rate": 8.785407182599898e-06,
+      "loss": 0.2633,
+      "step": 14795
+    },
+    {
+      "epoch": 2.5260732269352224,
+      "grad_norm": 0.3604688035888555,
+      "learning_rate": 8.769600404653516e-06,
+      "loss": 0.2634,
+      "step": 14800
+    },
+    {
+      "epoch": 2.5269266877186993,
+      "grad_norm": 0.34790212358787853,
+      "learning_rate": 8.753793626707133e-06,
+      "loss": 0.2556,
+      "step": 14805
+    },
+    {
+      "epoch": 2.5277801485021762,
+      "grad_norm": 0.33700470901798213,
+      "learning_rate": 8.737986848760749e-06,
+      "loss": 0.2818,
+      "step": 14810
+    },
+    {
+      "epoch": 2.528633609285653,
+      "grad_norm": 0.3044950344194103,
+      "learning_rate": 8.722180070814366e-06,
+      "loss": 0.2748,
+      "step": 14815
+    },
+    {
+      "epoch": 2.5294870700691305,
+      "grad_norm": 0.7149306608806377,
+      "learning_rate": 8.706373292867982e-06,
+      "loss": 0.2246,
+      "step": 14820
+    },
+    {
+      "epoch": 2.5303405308526075,
+      "grad_norm": 0.30250017016183345,
+      "learning_rate": 8.690566514921598e-06,
+      "loss": 0.2555,
+      "step": 14825
+    },
+    {
+      "epoch": 2.5311939916360844,
+      "grad_norm": 0.3278541640816871,
+      "learning_rate": 8.674759736975215e-06,
+      "loss": 0.2981,
+      "step": 14830
+    },
+    {
+      "epoch": 2.5320474524195613,
+      "grad_norm": 0.33474662014247003,
+      "learning_rate": 8.658952959028833e-06,
+      "loss": 0.2437,
+      "step": 14835
+    },
+    {
+      "epoch": 2.5329009132030382,
+      "grad_norm": 0.36515057289873015,
+      "learning_rate": 8.643146181082449e-06,
+      "loss": 0.2771,
+      "step": 14840
+    },
+    {
+      "epoch": 2.5337543739865156,
+      "grad_norm": 0.31446094410200376,
+      "learning_rate": 8.627339403136066e-06,
+      "loss": 0.2545,
+      "step": 14845
+    },
+    {
+      "epoch": 2.534607834769992,
+      "grad_norm": 0.32208686775598677,
+      "learning_rate": 8.611532625189682e-06,
+      "loss": 0.2594,
+      "step": 14850
+    },
+    {
+      "epoch": 2.5354612955534694,
+      "grad_norm": 0.3204019032092192,
+      "learning_rate": 8.595725847243297e-06,
+      "loss": 0.2582,
+      "step": 14855
+    },
+    {
+      "epoch": 2.5363147563369464,
+      "grad_norm": 0.31283853850881443,
+      "learning_rate": 8.579919069296915e-06,
+      "loss": 0.2997,
+      "step": 14860
+    },
+    {
+      "epoch": 2.5371682171204233,
+      "grad_norm": 0.2888961506768333,
+      "learning_rate": 8.564112291350532e-06,
+      "loss": 0.2618,
+      "step": 14865
+    },
+    {
+      "epoch": 2.5380216779039,
+      "grad_norm": 0.33297411246830283,
+      "learning_rate": 8.548305513404148e-06,
+      "loss": 0.2738,
+      "step": 14870
+    },
+    {
+      "epoch": 2.538875138687377,
+      "grad_norm": 0.4111717314073259,
+      "learning_rate": 8.532498735457764e-06,
+      "loss": 0.2666,
+      "step": 14875
+    },
+    {
+      "epoch": 2.5397285994708545,
+      "grad_norm": 0.3093128801039588,
+      "learning_rate": 8.516691957511381e-06,
+      "loss": 0.2788,
+      "step": 14880
+    },
+    {
+      "epoch": 2.5405820602543314,
+      "grad_norm": 0.3471700671659906,
+      "learning_rate": 8.500885179564997e-06,
+      "loss": 0.2499,
+      "step": 14885
+    },
+    {
+      "epoch": 2.5414355210378083,
+      "grad_norm": 0.34457282907570624,
+      "learning_rate": 8.485078401618615e-06,
+      "loss": 0.279,
+      "step": 14890
+    },
+    {
+      "epoch": 2.5422889818212853,
+      "grad_norm": 0.3703156588027841,
+      "learning_rate": 8.469271623672232e-06,
+      "loss": 0.2805,
+      "step": 14895
+    },
+    {
+      "epoch": 2.543142442604762,
+      "grad_norm": 0.3854068611469067,
+      "learning_rate": 8.453464845725848e-06,
+      "loss": 0.2684,
+      "step": 14900
+    },
+    {
+      "epoch": 2.5439959033882396,
+      "grad_norm": 0.3258732154948925,
+      "learning_rate": 8.437658067779464e-06,
+      "loss": 0.2783,
+      "step": 14905
+    },
+    {
+      "epoch": 2.544849364171716,
+      "grad_norm": 0.3257588125906597,
+      "learning_rate": 8.421851289833081e-06,
+      "loss": 0.2478,
+      "step": 14910
+    },
+    {
+      "epoch": 2.5457028249551934,
+      "grad_norm": 0.3327368767692538,
+      "learning_rate": 8.406044511886697e-06,
+      "loss": 0.2658,
+      "step": 14915
+    },
+    {
+      "epoch": 2.5465562857386703,
+      "grad_norm": 0.308838692022257,
+      "learning_rate": 8.390237733940314e-06,
+      "loss": 0.2681,
+      "step": 14920
+    },
+    {
+      "epoch": 2.5474097465221472,
+      "grad_norm": 0.31348836145437325,
+      "learning_rate": 8.374430955993932e-06,
+      "loss": 0.2677,
+      "step": 14925
+    },
+    {
+      "epoch": 2.548263207305624,
+      "grad_norm": 0.34955246448296834,
+      "learning_rate": 8.358624178047547e-06,
+      "loss": 0.2652,
+      "step": 14930
+    },
+    {
+      "epoch": 2.549116668089101,
+      "grad_norm": 0.3622847088386915,
+      "learning_rate": 8.342817400101163e-06,
+      "loss": 0.2554,
+      "step": 14935
+    },
+    {
+      "epoch": 2.5499701288725785,
+      "grad_norm": 0.3255722255573887,
+      "learning_rate": 8.32701062215478e-06,
+      "loss": 0.2649,
+      "step": 14940
+    },
+    {
+      "epoch": 2.5508235896560554,
+      "grad_norm": 0.36982499193159035,
+      "learning_rate": 8.311203844208396e-06,
+      "loss": 0.2948,
+      "step": 14945
+    },
+    {
+      "epoch": 2.5516770504395323,
+      "grad_norm": 0.36951547737636453,
+      "learning_rate": 8.295397066262014e-06,
+      "loss": 0.2585,
+      "step": 14950
+    },
+    {
+      "epoch": 2.5525305112230092,
+      "grad_norm": 0.3590117026965425,
+      "learning_rate": 8.279590288315631e-06,
+      "loss": 0.2559,
+      "step": 14955
+    },
+    {
+      "epoch": 2.553383972006486,
+      "grad_norm": 0.3636835050984558,
+      "learning_rate": 8.263783510369247e-06,
+      "loss": 0.2513,
+      "step": 14960
+    },
+    {
+      "epoch": 2.5542374327899635,
+      "grad_norm": 0.38929351543522134,
+      "learning_rate": 8.247976732422863e-06,
+      "loss": 0.2822,
+      "step": 14965
+    },
+    {
+      "epoch": 2.5550908935734404,
+      "grad_norm": 0.2865223071614123,
+      "learning_rate": 8.23216995447648e-06,
+      "loss": 0.2585,
+      "step": 14970
+    },
+    {
+      "epoch": 2.5559443543569174,
+      "grad_norm": 0.33793589845307387,
+      "learning_rate": 8.216363176530096e-06,
+      "loss": 0.2494,
+      "step": 14975
+    },
+    {
+      "epoch": 2.5567978151403943,
+      "grad_norm": 0.42855016920914785,
+      "learning_rate": 8.200556398583714e-06,
+      "loss": 0.2586,
+      "step": 14980
+    },
+    {
+      "epoch": 2.557651275923871,
+      "grad_norm": 0.3710576936912798,
+      "learning_rate": 8.184749620637331e-06,
+      "loss": 0.2653,
+      "step": 14985
+    },
+    {
+      "epoch": 2.5585047367073486,
+      "grad_norm": 0.3192702670716028,
+      "learning_rate": 8.168942842690947e-06,
+      "loss": 0.2742,
+      "step": 14990
+    },
+    {
+      "epoch": 2.559358197490825,
+      "grad_norm": 0.326286449535892,
+      "learning_rate": 8.153136064744563e-06,
+      "loss": 0.269,
+      "step": 14995
+    },
+    {
+      "epoch": 2.5602116582743024,
+      "grad_norm": 0.36306778061934497,
+      "learning_rate": 8.13732928679818e-06,
+      "loss": 0.267,
+      "step": 15000
+    },
+    {
+      "epoch": 2.5610651190577793,
+      "grad_norm": 0.3674230986760816,
+      "learning_rate": 8.121522508851796e-06,
+      "loss": 0.2556,
+      "step": 15005
+    },
+    {
+      "epoch": 2.5619185798412563,
+      "grad_norm": 0.3886694315470275,
+      "learning_rate": 8.105715730905413e-06,
+      "loss": 0.2586,
+      "step": 15010
+    },
+    {
+      "epoch": 2.562772040624733,
+      "grad_norm": 0.31734254900201264,
+      "learning_rate": 8.089908952959029e-06,
+      "loss": 0.2752,
+      "step": 15015
+    },
+    {
+      "epoch": 2.56362550140821,
+      "grad_norm": 0.3462925041349604,
+      "learning_rate": 8.074102175012646e-06,
+      "loss": 0.2601,
+      "step": 15020
+    },
+    {
+      "epoch": 2.5644789621916875,
+      "grad_norm": 0.36436655644205235,
+      "learning_rate": 8.058295397066262e-06,
+      "loss": 0.2734,
+      "step": 15025
+    },
+    {
+      "epoch": 2.5653324229751644,
+      "grad_norm": 0.3349267498877912,
+      "learning_rate": 8.042488619119878e-06,
+      "loss": 0.2777,
+      "step": 15030
+    },
+    {
+      "epoch": 2.5661858837586413,
+      "grad_norm": 0.353138665294069,
+      "learning_rate": 8.026681841173495e-06,
+      "loss": 0.287,
+      "step": 15035
+    },
+    {
+      "epoch": 2.5670393445421182,
+      "grad_norm": 0.3345752313165088,
+      "learning_rate": 8.010875063227111e-06,
+      "loss": 0.2488,
+      "step": 15040
+    },
+    {
+      "epoch": 2.567892805325595,
+      "grad_norm": 0.3606955744346641,
+      "learning_rate": 7.995068285280729e-06,
+      "loss": 0.2786,
+      "step": 15045
+    },
+    {
+      "epoch": 2.5687462661090725,
+      "grad_norm": 0.3412036584846417,
+      "learning_rate": 7.979261507334346e-06,
+      "loss": 0.258,
+      "step": 15050
+    },
+    {
+      "epoch": 2.5695997268925495,
+      "grad_norm": 0.35693653636226713,
+      "learning_rate": 7.963454729387962e-06,
+      "loss": 0.2642,
+      "step": 15055
+    },
+    {
+      "epoch": 2.5704531876760264,
+      "grad_norm": 0.3280891820451852,
+      "learning_rate": 7.947647951441578e-06,
+      "loss": 0.2845,
+      "step": 15060
+    },
+    {
+      "epoch": 2.5713066484595033,
+      "grad_norm": 0.3579927841500715,
+      "learning_rate": 7.931841173495195e-06,
+      "loss": 0.2667,
+      "step": 15065
+    },
+    {
+      "epoch": 2.57216010924298,
+      "grad_norm": 0.3384300716371121,
+      "learning_rate": 7.916034395548811e-06,
+      "loss": 0.2602,
+      "step": 15070
+    },
+    {
+      "epoch": 2.573013570026457,
+      "grad_norm": 0.35418106860695336,
+      "learning_rate": 7.900227617602428e-06,
+      "loss": 0.2673,
+      "step": 15075
+    },
+    {
+      "epoch": 2.573867030809934,
+      "grad_norm": 0.3698438742352702,
+      "learning_rate": 7.884420839656046e-06,
+      "loss": 0.2743,
+      "step": 15080
+    },
+    {
+      "epoch": 2.5747204915934114,
+      "grad_norm": 0.363613103217298,
+      "learning_rate": 7.868614061709662e-06,
+      "loss": 0.2549,
+      "step": 15085
+    },
+    {
+      "epoch": 2.5755739523768884,
+      "grad_norm": 0.3394372180741034,
+      "learning_rate": 7.852807283763277e-06,
+      "loss": 0.2718,
+      "step": 15090
+    },
+    {
+      "epoch": 2.5764274131603653,
+      "grad_norm": 0.36423762991683717,
+      "learning_rate": 7.837000505816895e-06,
+      "loss": 0.2949,
+      "step": 15095
+    },
+    {
+      "epoch": 2.577280873943842,
+      "grad_norm": 0.31743731897346655,
+      "learning_rate": 7.82119372787051e-06,
+      "loss": 0.2222,
+      "step": 15100
+    },
+    {
+      "epoch": 2.578134334727319,
+      "grad_norm": 0.32554087737361476,
+      "learning_rate": 7.805386949924128e-06,
+      "loss": 0.2709,
+      "step": 15105
+    },
+    {
+      "epoch": 2.5789877955107965,
+      "grad_norm": 0.31559740264213476,
+      "learning_rate": 7.789580171977745e-06,
+      "loss": 0.2755,
+      "step": 15110
+    },
+    {
+      "epoch": 2.5798412562942734,
+      "grad_norm": 0.3063018528733892,
+      "learning_rate": 7.773773394031361e-06,
+      "loss": 0.2489,
+      "step": 15115
+    },
+    {
+      "epoch": 2.5806947170777503,
+      "grad_norm": 0.31891238791816007,
+      "learning_rate": 7.757966616084977e-06,
+      "loss": 0.2815,
+      "step": 15120
+    },
+    {
+      "epoch": 2.5815481778612273,
+      "grad_norm": 0.3482705279801674,
+      "learning_rate": 7.742159838138594e-06,
+      "loss": 0.2372,
+      "step": 15125
+    },
+    {
+      "epoch": 2.582401638644704,
+      "grad_norm": 0.3468335021357962,
+      "learning_rate": 7.72635306019221e-06,
+      "loss": 0.2815,
+      "step": 15130
+    },
+    {
+      "epoch": 2.5832550994281815,
+      "grad_norm": 0.3465393777242725,
+      "learning_rate": 7.710546282245828e-06,
+      "loss": 0.2731,
+      "step": 15135
+    },
+    {
+      "epoch": 2.584108560211658,
+      "grad_norm": 0.3913655087141834,
+      "learning_rate": 7.694739504299443e-06,
+      "loss": 0.2575,
+      "step": 15140
+    },
+    {
+      "epoch": 2.5849620209951354,
+      "grad_norm": 0.3190687438763158,
+      "learning_rate": 7.678932726353061e-06,
+      "loss": 0.2669,
+      "step": 15145
+    },
+    {
+      "epoch": 2.5858154817786123,
+      "grad_norm": 0.358078662616606,
+      "learning_rate": 7.663125948406677e-06,
+      "loss": 0.2901,
+      "step": 15150
+    },
+    {
+      "epoch": 2.5866689425620892,
+      "grad_norm": 0.34333694637136325,
+      "learning_rate": 7.647319170460292e-06,
+      "loss": 0.2697,
+      "step": 15155
+    },
+    {
+      "epoch": 2.587522403345566,
+      "grad_norm": 0.35238615533564754,
+      "learning_rate": 7.63151239251391e-06,
+      "loss": 0.2598,
+      "step": 15160
+    },
+    {
+      "epoch": 2.588375864129043,
+      "grad_norm": 0.34817701038063625,
+      "learning_rate": 7.6157056145675265e-06,
+      "loss": 0.2519,
+      "step": 15165
+    },
+    {
+      "epoch": 2.5892293249125204,
+      "grad_norm": 0.31586534444356135,
+      "learning_rate": 7.599898836621143e-06,
+      "loss": 0.2504,
+      "step": 15170
+    },
+    {
+      "epoch": 2.5900827856959974,
+      "grad_norm": 0.37410390706038105,
+      "learning_rate": 7.5840920586747606e-06,
+      "loss": 0.2859,
+      "step": 15175
+    },
+    {
+      "epoch": 2.5909362464794743,
+      "grad_norm": 0.4076676227039792,
+      "learning_rate": 7.568285280728376e-06,
+      "loss": 0.2664,
+      "step": 15180
+    },
+    {
+      "epoch": 2.591789707262951,
+      "grad_norm": 0.328956954930043,
+      "learning_rate": 7.552478502781993e-06,
+      "loss": 0.2795,
+      "step": 15185
+    },
+    {
+      "epoch": 2.592643168046428,
+      "grad_norm": 0.3242313957454638,
+      "learning_rate": 7.53667172483561e-06,
+      "loss": 0.2605,
+      "step": 15190
+    },
+    {
+      "epoch": 2.5934966288299055,
+      "grad_norm": 0.35632982846401845,
+      "learning_rate": 7.520864946889226e-06,
+      "loss": 0.2711,
+      "step": 15195
+    },
+    {
+      "epoch": 2.5943500896133824,
+      "grad_norm": 0.3673822981048876,
+      "learning_rate": 7.505058168942843e-06,
+      "loss": 0.2431,
+      "step": 15200
+    },
+    {
+      "epoch": 2.5952035503968593,
+      "grad_norm": 0.38628552518056947,
+      "learning_rate": 7.48925139099646e-06,
+      "loss": 0.2666,
+      "step": 15205
+    },
+    {
+      "epoch": 2.5960570111803363,
+      "grad_norm": 0.3477240224200369,
+      "learning_rate": 7.473444613050076e-06,
+      "loss": 0.2615,
+      "step": 15210
+    },
+    {
+      "epoch": 2.596910471963813,
+      "grad_norm": 0.3444502479801836,
+      "learning_rate": 7.457637835103693e-06,
+      "loss": 0.2654,
+      "step": 15215
+    },
+    {
+      "epoch": 2.59776393274729,
+      "grad_norm": 0.3889329248670148,
+      "learning_rate": 7.44183105715731e-06,
+      "loss": 0.2655,
+      "step": 15220
+    },
+    {
+      "epoch": 2.598617393530767,
+      "grad_norm": 0.7088173666283505,
+      "learning_rate": 7.426024279210926e-06,
+      "loss": 0.2666,
+      "step": 15225
+    },
+    {
+      "epoch": 2.5994708543142444,
+      "grad_norm": 0.37264826890797625,
+      "learning_rate": 7.4102175012645424e-06,
+      "loss": 0.2666,
+      "step": 15230
+    },
+    {
+      "epoch": 2.6003243150977213,
+      "grad_norm": 0.3639291886301692,
+      "learning_rate": 7.39441072331816e-06,
+      "loss": 0.28,
+      "step": 15235
+    },
+    {
+      "epoch": 2.6011777758811983,
+      "grad_norm": 0.3427815985351985,
+      "learning_rate": 7.378603945371776e-06,
+      "loss": 0.2731,
+      "step": 15240
+    },
+    {
+      "epoch": 2.602031236664675,
+      "grad_norm": 0.32112484418580056,
+      "learning_rate": 7.362797167425392e-06,
+      "loss": 0.2465,
+      "step": 15245
+    },
+    {
+      "epoch": 2.602884697448152,
+      "grad_norm": 0.3613054889577417,
+      "learning_rate": 7.34699038947901e-06,
+      "loss": 0.2826,
+      "step": 15250
+    },
+    {
+      "epoch": 2.6037381582316295,
+      "grad_norm": 0.30921220095070084,
+      "learning_rate": 7.3311836115326255e-06,
+      "loss": 0.2635,
+      "step": 15255
+    },
+    {
+      "epoch": 2.6045916190151064,
+      "grad_norm": 0.35104175864231874,
+      "learning_rate": 7.315376833586242e-06,
+      "loss": 0.2503,
+      "step": 15260
+    },
+    {
+      "epoch": 2.6054450797985833,
+      "grad_norm": 0.3290929790912007,
+      "learning_rate": 7.2995700556398596e-06,
+      "loss": 0.2818,
+      "step": 15265
+    },
+    {
+      "epoch": 2.6062985405820602,
+      "grad_norm": 0.38330817541779394,
+      "learning_rate": 7.283763277693475e-06,
+      "loss": 0.2515,
+      "step": 15270
+    },
+    {
+      "epoch": 2.607152001365537,
+      "grad_norm": 0.3732479596804675,
+      "learning_rate": 7.267956499747092e-06,
+      "loss": 0.2756,
+      "step": 15275
+    },
+    {
+      "epoch": 2.6080054621490145,
+      "grad_norm": 0.34507633371042173,
+      "learning_rate": 7.252149721800708e-06,
+      "loss": 0.2564,
+      "step": 15280
+    },
+    {
+      "epoch": 2.608858922932491,
+      "grad_norm": 0.35935976480326104,
+      "learning_rate": 7.236342943854325e-06,
+      "loss": 0.2627,
+      "step": 15285
+    },
+    {
+      "epoch": 2.6097123837159684,
+      "grad_norm": 0.3126641707142665,
+      "learning_rate": 7.220536165907942e-06,
+      "loss": 0.2649,
+      "step": 15290
+    },
+    {
+      "epoch": 2.6105658444994453,
+      "grad_norm": 0.3924070249485932,
+      "learning_rate": 7.2047293879615575e-06,
+      "loss": 0.2532,
+      "step": 15295
+    },
+    {
+      "epoch": 2.611419305282922,
+      "grad_norm": 0.39147945187954714,
+      "learning_rate": 7.188922610015175e-06,
+      "loss": 0.2594,
+      "step": 15300
+    },
+    {
+      "epoch": 2.612272766066399,
+      "grad_norm": 0.33173543761510493,
+      "learning_rate": 7.173115832068792e-06,
+      "loss": 0.2505,
+      "step": 15305
+    },
+    {
+      "epoch": 2.613126226849876,
+      "grad_norm": 0.3320536524500648,
+      "learning_rate": 7.157309054122407e-06,
+      "loss": 0.2704,
+      "step": 15310
+    },
+    {
+      "epoch": 2.6139796876333534,
+      "grad_norm": 0.31324033158088654,
+      "learning_rate": 7.141502276176025e-06,
+      "loss": 0.2588,
+      "step": 15315
+    },
+    {
+      "epoch": 2.6148331484168303,
+      "grad_norm": 0.3673039365008783,
+      "learning_rate": 7.1256954982296414e-06,
+      "loss": 0.2577,
+      "step": 15320
+    },
+    {
+      "epoch": 2.6156866092003073,
+      "grad_norm": 0.35635933014280124,
+      "learning_rate": 7.109888720283257e-06,
+      "loss": 0.2847,
+      "step": 15325
+    },
+    {
+      "epoch": 2.616540069983784,
+      "grad_norm": 0.32938069554069876,
+      "learning_rate": 7.094081942336875e-06,
+      "loss": 0.2482,
+      "step": 15330
+    },
+    {
+      "epoch": 2.617393530767261,
+      "grad_norm": 0.32267895465365887,
+      "learning_rate": 7.078275164390491e-06,
+      "loss": 0.2595,
+      "step": 15335
+    },
+    {
+      "epoch": 2.6182469915507385,
+      "grad_norm": 0.338747778656447,
+      "learning_rate": 7.062468386444107e-06,
+      "loss": 0.2308,
+      "step": 15340
+    },
+    {
+      "epoch": 2.6191004523342154,
+      "grad_norm": 2.6701997243958746,
+      "learning_rate": 7.0466616084977245e-06,
+      "loss": 0.2452,
+      "step": 15345
+    },
+    {
+      "epoch": 2.6199539131176923,
+      "grad_norm": 0.3638867152831773,
+      "learning_rate": 7.030854830551341e-06,
+      "loss": 0.2659,
+      "step": 15350
+    },
+    {
+      "epoch": 2.6208073739011692,
+      "grad_norm": 0.3411134390071248,
+      "learning_rate": 7.015048052604957e-06,
+      "loss": 0.2765,
+      "step": 15355
+    },
+    {
+      "epoch": 2.621660834684646,
+      "grad_norm": 0.33325268514592343,
+      "learning_rate": 6.999241274658574e-06,
+      "loss": 0.2389,
+      "step": 15360
+    },
+    {
+      "epoch": 2.622514295468123,
+      "grad_norm": 0.3303227362918107,
+      "learning_rate": 6.983434496712191e-06,
+      "loss": 0.2505,
+      "step": 15365
+    },
+    {
+      "epoch": 2.6233677562516,
+      "grad_norm": 0.3383050930967925,
+      "learning_rate": 6.967627718765807e-06,
+      "loss": 0.2783,
+      "step": 15370
+    },
+    {
+      "epoch": 2.6242212170350774,
+      "grad_norm": 0.3280058631343651,
+      "learning_rate": 6.951820940819424e-06,
+      "loss": 0.2689,
+      "step": 15375
+    },
+    {
+      "epoch": 2.6250746778185543,
+      "grad_norm": 0.35144419432689084,
+      "learning_rate": 6.936014162873041e-06,
+      "loss": 0.2519,
+      "step": 15380
+    },
+    {
+      "epoch": 2.6259281386020312,
+      "grad_norm": 0.37545184334564646,
+      "learning_rate": 6.9202073849266565e-06,
+      "loss": 0.2723,
+      "step": 15385
+    },
+    {
+      "epoch": 2.626781599385508,
+      "grad_norm": 0.37231014008438706,
+      "learning_rate": 6.904400606980274e-06,
+      "loss": 0.2524,
+      "step": 15390
+    },
+    {
+      "epoch": 2.627635060168985,
+      "grad_norm": 0.33419225335162134,
+      "learning_rate": 6.888593829033891e-06,
+      "loss": 0.258,
+      "step": 15395
+    },
+    {
+      "epoch": 2.6284885209524624,
+      "grad_norm": 0.29567237571501975,
+      "learning_rate": 6.872787051087506e-06,
+      "loss": 0.2567,
+      "step": 15400
+    },
+    {
+      "epoch": 2.6293419817359394,
+      "grad_norm": 0.30445908419779605,
+      "learning_rate": 6.856980273141123e-06,
+      "loss": 0.2596,
+      "step": 15405
+    },
+    {
+      "epoch": 2.6301954425194163,
+      "grad_norm": 0.30655934744530733,
+      "learning_rate": 6.84117349519474e-06,
+      "loss": 0.2509,
+      "step": 15410
+    },
+    {
+      "epoch": 2.631048903302893,
+      "grad_norm": 0.5928998708875786,
+      "learning_rate": 6.825366717248356e-06,
+      "loss": 0.2709,
+      "step": 15415
+    },
+    {
+      "epoch": 2.63190236408637,
+      "grad_norm": 0.3644074401787618,
+      "learning_rate": 6.809559939301973e-06,
+      "loss": 0.2633,
+      "step": 15420
+    },
+    {
+      "epoch": 2.6327558248698475,
+      "grad_norm": 0.417807151329196,
+      "learning_rate": 6.79375316135559e-06,
+      "loss": 0.2622,
+      "step": 15425
+    },
+    {
+      "epoch": 2.633609285653324,
+      "grad_norm": 0.3564349216934015,
+      "learning_rate": 6.777946383409206e-06,
+      "loss": 0.2412,
+      "step": 15430
+    },
+    {
+      "epoch": 2.6344627464368013,
+      "grad_norm": 0.3645799274924827,
+      "learning_rate": 6.762139605462823e-06,
+      "loss": 0.2539,
+      "step": 15435
+    },
+    {
+      "epoch": 2.6353162072202783,
+      "grad_norm": 0.46329426878383684,
+      "learning_rate": 6.74633282751644e-06,
+      "loss": 0.2878,
+      "step": 15440
+    },
+    {
+      "epoch": 2.636169668003755,
+      "grad_norm": 0.41205339021908743,
+      "learning_rate": 6.730526049570056e-06,
+      "loss": 0.2789,
+      "step": 15445
+    },
+    {
+      "epoch": 2.637023128787232,
+      "grad_norm": 0.36220454124846574,
+      "learning_rate": 6.7147192716236725e-06,
+      "loss": 0.2563,
+      "step": 15450
+    },
+    {
+      "epoch": 2.637876589570709,
+      "grad_norm": 0.36394619076485407,
+      "learning_rate": 6.69891249367729e-06,
+      "loss": 0.2702,
+      "step": 15455
+    },
+    {
+      "epoch": 2.6387300503541864,
+      "grad_norm": 0.31510412525522824,
+      "learning_rate": 6.683105715730906e-06,
+      "loss": 0.2705,
+      "step": 15460
+    },
+    {
+      "epoch": 2.6395835111376633,
+      "grad_norm": 0.35959438946296624,
+      "learning_rate": 6.6672989377845214e-06,
+      "loss": 0.2638,
+      "step": 15465
+    },
+    {
+      "epoch": 2.6404369719211402,
+      "grad_norm": 0.3719073881386995,
+      "learning_rate": 6.65149215983814e-06,
+      "loss": 0.2614,
+      "step": 15470
+    },
+    {
+      "epoch": 2.641290432704617,
+      "grad_norm": 0.34851867562967587,
+      "learning_rate": 6.6356853818917555e-06,
+      "loss": 0.2704,
+      "step": 15475
+    },
+    {
+      "epoch": 2.642143893488094,
+      "grad_norm": 0.3412509124045147,
+      "learning_rate": 6.619878603945371e-06,
+      "loss": 0.2965,
+      "step": 15480
+    },
+    {
+      "epoch": 2.6429973542715715,
+      "grad_norm": 0.2969856118052839,
+      "learning_rate": 6.60407182599899e-06,
+      "loss": 0.2606,
+      "step": 15485
+    },
+    {
+      "epoch": 2.6438508150550484,
+      "grad_norm": 0.33960837625779033,
+      "learning_rate": 6.588265048052605e-06,
+      "loss": 0.2827,
+      "step": 15490
+    },
+    {
+      "epoch": 2.6447042758385253,
+      "grad_norm": 0.3435040294889023,
+      "learning_rate": 6.572458270106221e-06,
+      "loss": 0.2543,
+      "step": 15495
+    },
+    {
+      "epoch": 2.645557736622002,
+      "grad_norm": 0.4138041998758704,
+      "learning_rate": 6.556651492159839e-06,
+      "loss": 0.2535,
+      "step": 15500
+    },
+    {
+      "epoch": 2.646411197405479,
+      "grad_norm": 0.30733760212114253,
+      "learning_rate": 6.540844714213455e-06,
+      "loss": 0.2723,
+      "step": 15505
+    },
+    {
+      "epoch": 2.647264658188956,
+      "grad_norm": 0.34787237731846093,
+      "learning_rate": 6.525037936267071e-06,
+      "loss": 0.2574,
+      "step": 15510
+    },
+    {
+      "epoch": 2.648118118972433,
+      "grad_norm": 0.3549540085305931,
+      "learning_rate": 6.509231158320689e-06,
+      "loss": 0.2653,
+      "step": 15515
+    },
+    {
+      "epoch": 2.6489715797559104,
+      "grad_norm": 0.37233586082518344,
+      "learning_rate": 6.493424380374305e-06,
+      "loss": 0.2551,
+      "step": 15520
+    },
+    {
+      "epoch": 2.6498250405393873,
+      "grad_norm": 0.323166190585263,
+      "learning_rate": 6.477617602427921e-06,
+      "loss": 0.2646,
+      "step": 15525
+    },
+    {
+      "epoch": 2.650678501322864,
+      "grad_norm": 0.34153440033426474,
+      "learning_rate": 6.461810824481539e-06,
+      "loss": 0.2538,
+      "step": 15530
+    },
+    {
+      "epoch": 2.651531962106341,
+      "grad_norm": 0.34495479489164316,
+      "learning_rate": 6.446004046535155e-06,
+      "loss": 0.2454,
+      "step": 15535
+    },
+    {
+      "epoch": 2.652385422889818,
+      "grad_norm": 0.2990419238606832,
+      "learning_rate": 6.430197268588771e-06,
+      "loss": 0.2983,
+      "step": 15540
+    },
+    {
+      "epoch": 2.6532388836732954,
+      "grad_norm": 0.3373808033775284,
+      "learning_rate": 6.414390490642387e-06,
+      "loss": 0.2648,
+      "step": 15545
+    },
+    {
+      "epoch": 2.6540923444567723,
+      "grad_norm": 0.30005592281826204,
+      "learning_rate": 6.398583712696005e-06,
+      "loss": 0.2769,
+      "step": 15550
+    },
+    {
+      "epoch": 2.6549458052402493,
+      "grad_norm": 0.2950390468483142,
+      "learning_rate": 6.3827769347496204e-06,
+      "loss": 0.2502,
+      "step": 15555
+    },
+    {
+      "epoch": 2.655799266023726,
+      "grad_norm": 0.7361620948064368,
+      "learning_rate": 6.366970156803237e-06,
+      "loss": 0.226,
+      "step": 15560
+    },
+    {
+      "epoch": 2.656652726807203,
+      "grad_norm": 0.35702882770988104,
+      "learning_rate": 6.3511633788568545e-06,
+      "loss": 0.2887,
+      "step": 15565
+    },
+    {
+      "epoch": 2.6575061875906805,
+      "grad_norm": 0.40098849474500975,
+      "learning_rate": 6.33535660091047e-06,
+      "loss": 0.2595,
+      "step": 15570
+    },
+    {
+      "epoch": 2.658359648374157,
+      "grad_norm": 0.33299844553405433,
+      "learning_rate": 6.319549822964087e-06,
+      "loss": 0.2624,
+      "step": 15575
+    },
+    {
+      "epoch": 2.6592131091576343,
+      "grad_norm": 0.9552251998611956,
+      "learning_rate": 6.303743045017704e-06,
+      "loss": 0.2657,
+      "step": 15580
+    },
+    {
+      "epoch": 2.6600665699411112,
+      "grad_norm": 0.33097635247691237,
+      "learning_rate": 6.28793626707132e-06,
+      "loss": 0.2329,
+      "step": 15585
+    },
+    {
+      "epoch": 2.660920030724588,
+      "grad_norm": 0.33191527265529913,
+      "learning_rate": 6.272129489124937e-06,
+      "loss": 0.2682,
+      "step": 15590
+    },
+    {
+      "epoch": 2.661773491508065,
+      "grad_norm": 0.3159377513980589,
+      "learning_rate": 6.256322711178554e-06,
+      "loss": 0.2587,
+      "step": 15595
+    },
+    {
+      "epoch": 2.662626952291542,
+      "grad_norm": 0.3141137315073459,
+      "learning_rate": 6.24051593323217e-06,
+      "loss": 0.288,
+      "step": 15600
+    },
+    {
+      "epoch": 2.6634804130750194,
+      "grad_norm": 0.3116398274186534,
+      "learning_rate": 6.224709155285787e-06,
+      "loss": 0.2686,
+      "step": 15605
+    },
+    {
+      "epoch": 2.6643338738584963,
+      "grad_norm": 0.3848915059332201,
+      "learning_rate": 6.208902377339403e-06,
+      "loss": 0.2774,
+      "step": 15610
+    },
+    {
+      "epoch": 2.665187334641973,
+      "grad_norm": 0.3917115023600133,
+      "learning_rate": 6.19309559939302e-06,
+      "loss": 0.2635,
+      "step": 15615
+    },
+    {
+      "epoch": 2.66604079542545,
+      "grad_norm": 0.37598166690722307,
+      "learning_rate": 6.177288821446637e-06,
+      "loss": 0.2359,
+      "step": 15620
+    },
+    {
+      "epoch": 2.666894256208927,
+      "grad_norm": 0.3427634213516318,
+      "learning_rate": 6.161482043500253e-06,
+      "loss": 0.2621,
+      "step": 15625
+    },
+    {
+      "epoch": 2.6677477169924044,
+      "grad_norm": 0.32992098983283125,
+      "learning_rate": 6.14567526555387e-06,
+      "loss": 0.2628,
+      "step": 15630
+    },
+    {
+      "epoch": 2.6686011777758814,
+      "grad_norm": 0.38083250474499536,
+      "learning_rate": 6.129868487607486e-06,
+      "loss": 0.2593,
+      "step": 15635
+    },
+    {
+      "epoch": 2.6694546385593583,
+      "grad_norm": 0.3737949530874446,
+      "learning_rate": 6.114061709661103e-06,
+      "loss": 0.2485,
+      "step": 15640
+    },
+    {
+      "epoch": 2.670308099342835,
+      "grad_norm": 0.36254880895682357,
+      "learning_rate": 6.098254931714719e-06,
+      "loss": 0.2741,
+      "step": 15645
+    },
+    {
+      "epoch": 2.671161560126312,
+      "grad_norm": 0.3679555211260355,
+      "learning_rate": 6.082448153768336e-06,
+      "loss": 0.2622,
+      "step": 15650
+    },
+    {
+      "epoch": 2.672015020909789,
+      "grad_norm": 0.32060345874403146,
+      "learning_rate": 6.066641375821953e-06,
+      "loss": 0.2662,
+      "step": 15655
+    },
+    {
+      "epoch": 2.672868481693266,
+      "grad_norm": 0.3079424030638837,
+      "learning_rate": 6.050834597875569e-06,
+      "loss": 0.2518,
+      "step": 15660
+    },
+    {
+      "epoch": 2.6737219424767433,
+      "grad_norm": 0.33856647147712504,
+      "learning_rate": 6.035027819929186e-06,
+      "loss": 0.2501,
+      "step": 15665
+    },
+    {
+      "epoch": 2.6745754032602203,
+      "grad_norm": 0.3653565329375958,
+      "learning_rate": 6.0192210419828025e-06,
+      "loss": 0.2542,
+      "step": 15670
+    },
+    {
+      "epoch": 2.675428864043697,
+      "grad_norm": 0.3595713968478812,
+      "learning_rate": 6.003414264036419e-06,
+      "loss": 0.2538,
+      "step": 15675
+    },
+    {
+      "epoch": 2.676282324827174,
+      "grad_norm": 0.33627575601187915,
+      "learning_rate": 5.987607486090036e-06,
+      "loss": 0.267,
+      "step": 15680
+    },
+    {
+      "epoch": 2.677135785610651,
+      "grad_norm": 0.4062703192670018,
+      "learning_rate": 5.971800708143652e-06,
+      "loss": 0.249,
+      "step": 15685
+    },
+    {
+      "epoch": 2.6779892463941284,
+      "grad_norm": 0.3594122030085205,
+      "learning_rate": 5.955993930197269e-06,
+      "loss": 0.2798,
+      "step": 15690
+    },
+    {
+      "epoch": 2.6788427071776053,
+      "grad_norm": 0.3895681673156118,
+      "learning_rate": 5.9401871522508855e-06,
+      "loss": 0.2818,
+      "step": 15695
+    },
+    {
+      "epoch": 2.6796961679610822,
+      "grad_norm": 0.43038452358267465,
+      "learning_rate": 5.924380374304502e-06,
+      "loss": 0.2598,
+      "step": 15700
+    },
+    {
+      "epoch": 2.680549628744559,
+      "grad_norm": 0.3995685056530944,
+      "learning_rate": 5.908573596358119e-06,
+      "loss": 0.2933,
+      "step": 15705
+    },
+    {
+      "epoch": 2.681403089528036,
+      "grad_norm": 0.3734290041378778,
+      "learning_rate": 5.892766818411735e-06,
+      "loss": 0.2743,
+      "step": 15710
+    },
+    {
+      "epoch": 2.6822565503115134,
+      "grad_norm": 0.31717697701338593,
+      "learning_rate": 5.876960040465352e-06,
+      "loss": 0.2742,
+      "step": 15715
+    },
+    {
+      "epoch": 2.68311001109499,
+      "grad_norm": 0.37878051677289487,
+      "learning_rate": 5.861153262518968e-06,
+      "loss": 0.2395,
+      "step": 15720
+    },
+    {
+      "epoch": 2.6839634718784673,
+      "grad_norm": 0.3502105273684265,
+      "learning_rate": 5.845346484572585e-06,
+      "loss": 0.2554,
+      "step": 15725
+    },
+    {
+      "epoch": 2.684816932661944,
+      "grad_norm": 0.36504635411920455,
+      "learning_rate": 5.829539706626202e-06,
+      "loss": 0.2799,
+      "step": 15730
+    },
+    {
+      "epoch": 2.685670393445421,
+      "grad_norm": 0.32571179571423386,
+      "learning_rate": 5.8137329286798176e-06,
+      "loss": 0.2738,
+      "step": 15735
+    },
+    {
+      "epoch": 2.686523854228898,
+      "grad_norm": 0.3450408141278497,
+      "learning_rate": 5.797926150733435e-06,
+      "loss": 0.2568,
+      "step": 15740
+    },
+    {
+      "epoch": 2.687377315012375,
+      "grad_norm": 0.35448784854975235,
+      "learning_rate": 5.782119372787052e-06,
+      "loss": 0.27,
+      "step": 15745
+    },
+    {
+      "epoch": 2.6882307757958523,
+      "grad_norm": 0.33889855398398216,
+      "learning_rate": 5.766312594840667e-06,
+      "loss": 0.2481,
+      "step": 15750
+    },
+    {
+      "epoch": 2.6890842365793293,
+      "grad_norm": 0.35149130323176103,
+      "learning_rate": 5.750505816894285e-06,
+      "loss": 0.2301,
+      "step": 15755
+    },
+    {
+      "epoch": 2.689937697362806,
+      "grad_norm": 0.3660742482701346,
+      "learning_rate": 5.7346990389479015e-06,
+      "loss": 0.2823,
+      "step": 15760
+    },
+    {
+      "epoch": 2.690791158146283,
+      "grad_norm": 0.3411786259643136,
+      "learning_rate": 5.718892261001517e-06,
+      "loss": 0.2463,
+      "step": 15765
+    },
+    {
+      "epoch": 2.69164461892976,
+      "grad_norm": 0.3730178861789147,
+      "learning_rate": 5.703085483055135e-06,
+      "loss": 0.2242,
+      "step": 15770
+    },
+    {
+      "epoch": 2.6924980797132374,
+      "grad_norm": 0.3267702608488623,
+      "learning_rate": 5.6872787051087505e-06,
+      "loss": 0.2437,
+      "step": 15775
+    },
+    {
+      "epoch": 2.6933515404967143,
+      "grad_norm": 0.3303506142224966,
+      "learning_rate": 5.671471927162367e-06,
+      "loss": 0.2716,
+      "step": 15780
+    },
+    {
+      "epoch": 2.6942050012801912,
+      "grad_norm": 0.3483159930338373,
+      "learning_rate": 5.6556651492159845e-06,
+      "loss": 0.2964,
+      "step": 15785
+    },
+    {
+      "epoch": 2.695058462063668,
+      "grad_norm": 0.34365776948938526,
+      "learning_rate": 5.6398583712696e-06,
+      "loss": 0.2501,
+      "step": 15790
+    },
+    {
+      "epoch": 2.695911922847145,
+      "grad_norm": 0.5091088648901413,
+      "learning_rate": 5.624051593323217e-06,
+      "loss": 0.2498,
+      "step": 15795
+    },
+    {
+      "epoch": 2.696765383630622,
+      "grad_norm": 0.3219123113570026,
+      "learning_rate": 5.608244815376834e-06,
+      "loss": 0.2548,
+      "step": 15800
+    },
+    {
+      "epoch": 2.697618844414099,
+      "grad_norm": 0.36789902114116035,
+      "learning_rate": 5.59243803743045e-06,
+      "loss": 0.2388,
+      "step": 15805
+    },
+    {
+      "epoch": 2.6984723051975763,
+      "grad_norm": 0.40298010696935843,
+      "learning_rate": 5.576631259484067e-06,
+      "loss": 0.2751,
+      "step": 15810
+    },
+    {
+      "epoch": 2.6993257659810532,
+      "grad_norm": 0.4110009799120327,
+      "learning_rate": 5.560824481537684e-06,
+      "loss": 0.2667,
+      "step": 15815
+    },
+    {
+      "epoch": 2.70017922676453,
+      "grad_norm": 0.33758326531005173,
+      "learning_rate": 5.5450177035913e-06,
+      "loss": 0.2563,
+      "step": 15820
+    },
+    {
+      "epoch": 2.701032687548007,
+      "grad_norm": 0.33169441729925814,
+      "learning_rate": 5.5292109256449166e-06,
+      "loss": 0.2477,
+      "step": 15825
+    },
+    {
+      "epoch": 2.701886148331484,
+      "grad_norm": 0.3708573039512998,
+      "learning_rate": 5.513404147698533e-06,
+      "loss": 0.2853,
+      "step": 15830
+    },
+    {
+      "epoch": 2.7027396091149614,
+      "grad_norm": 0.32450941495953645,
+      "learning_rate": 5.49759736975215e-06,
+      "loss": 0.2648,
+      "step": 15835
+    },
+    {
+      "epoch": 2.7035930698984383,
+      "grad_norm": 0.5847723828809334,
+      "learning_rate": 5.481790591805766e-06,
+      "loss": 0.272,
+      "step": 15840
+    },
+    {
+      "epoch": 2.704446530681915,
+      "grad_norm": 0.3362346277513503,
+      "learning_rate": 5.465983813859383e-06,
+      "loss": 0.2465,
+      "step": 15845
+    },
+    {
+      "epoch": 2.705299991465392,
+      "grad_norm": 0.35963561018554313,
+      "learning_rate": 5.450177035913e-06,
+      "loss": 0.2742,
+      "step": 15850
+    },
+    {
+      "epoch": 2.706153452248869,
+      "grad_norm": 0.3016792294913857,
+      "learning_rate": 5.434370257966616e-06,
+      "loss": 0.28,
+      "step": 15855
+    },
+    {
+      "epoch": 2.7070069130323464,
+      "grad_norm": 0.3303518225889094,
+      "learning_rate": 5.418563480020233e-06,
+      "loss": 0.2682,
+      "step": 15860
+    },
+    {
+      "epoch": 2.707860373815823,
+      "grad_norm": 0.31989733013292765,
+      "learning_rate": 5.4027567020738494e-06,
+      "loss": 0.2792,
+      "step": 15865
+    },
+    {
+      "epoch": 2.7087138345993003,
+      "grad_norm": 0.34750150943486774,
+      "learning_rate": 5.386949924127466e-06,
+      "loss": 0.2613,
+      "step": 15870
+    },
+    {
+      "epoch": 2.709567295382777,
+      "grad_norm": 0.35145405872984736,
+      "learning_rate": 5.371143146181083e-06,
+      "loss": 0.2453,
+      "step": 15875
+    },
+    {
+      "epoch": 2.710420756166254,
+      "grad_norm": 0.3140012579713542,
+      "learning_rate": 5.355336368234699e-06,
+      "loss": 0.2592,
+      "step": 15880
+    },
+    {
+      "epoch": 2.711274216949731,
+      "grad_norm": 0.33591125148705864,
+      "learning_rate": 5.339529590288316e-06,
+      "loss": 0.2669,
+      "step": 15885
+    },
+    {
+      "epoch": 2.712127677733208,
+      "grad_norm": 0.3087501936993232,
+      "learning_rate": 5.3237228123419325e-06,
+      "loss": 0.2506,
+      "step": 15890
+    },
+    {
+      "epoch": 2.7129811385166853,
+      "grad_norm": 0.36517175964213555,
+      "learning_rate": 5.307916034395549e-06,
+      "loss": 0.2564,
+      "step": 15895
+    },
+    {
+      "epoch": 2.7138345993001622,
+      "grad_norm": 0.3475024929941956,
+      "learning_rate": 5.292109256449166e-06,
+      "loss": 0.2758,
+      "step": 15900
+    },
+    {
+      "epoch": 2.714688060083639,
+      "grad_norm": 0.334050236491831,
+      "learning_rate": 5.276302478502782e-06,
+      "loss": 0.2882,
+      "step": 15905
+    },
+    {
+      "epoch": 2.715541520867116,
+      "grad_norm": 0.33305463616687425,
+      "learning_rate": 5.260495700556399e-06,
+      "loss": 0.2417,
+      "step": 15910
+    },
+    {
+      "epoch": 2.716394981650593,
+      "grad_norm": 0.3399104079077125,
+      "learning_rate": 5.2446889226100155e-06,
+      "loss": 0.2563,
+      "step": 15915
+    },
+    {
+      "epoch": 2.7172484424340704,
+      "grad_norm": 0.35992749919916034,
+      "learning_rate": 5.228882144663632e-06,
+      "loss": 0.2698,
+      "step": 15920
+    },
+    {
+      "epoch": 2.7181019032175473,
+      "grad_norm": 0.37463460779301216,
+      "learning_rate": 5.213075366717249e-06,
+      "loss": 0.2644,
+      "step": 15925
+    },
+    {
+      "epoch": 2.718955364001024,
+      "grad_norm": 0.34481025499311296,
+      "learning_rate": 5.197268588770865e-06,
+      "loss": 0.2767,
+      "step": 15930
+    },
+    {
+      "epoch": 2.719808824784501,
+      "grad_norm": 0.31824100918089626,
+      "learning_rate": 5.181461810824482e-06,
+      "loss": 0.2768,
+      "step": 15935
+    },
+    {
+      "epoch": 2.720662285567978,
+      "grad_norm": 0.39036827158201115,
+      "learning_rate": 5.165655032878099e-06,
+      "loss": 0.2468,
+      "step": 15940
+    },
+    {
+      "epoch": 2.721515746351455,
+      "grad_norm": 0.35354453353195064,
+      "learning_rate": 5.149848254931715e-06,
+      "loss": 0.2574,
+      "step": 15945
+    },
+    {
+      "epoch": 2.722369207134932,
+      "grad_norm": 0.3651947064412998,
+      "learning_rate": 5.134041476985332e-06,
+      "loss": 0.2598,
+      "step": 15950
+    },
+    {
+      "epoch": 2.7232226679184093,
+      "grad_norm": 0.3200578358249782,
+      "learning_rate": 5.1182346990389484e-06,
+      "loss": 0.2524,
+      "step": 15955
+    },
+    {
+      "epoch": 2.724076128701886,
+      "grad_norm": 0.32106339624302854,
+      "learning_rate": 5.102427921092565e-06,
+      "loss": 0.2834,
+      "step": 15960
+    },
+    {
+      "epoch": 2.724929589485363,
+      "grad_norm": 0.35695751957917327,
+      "learning_rate": 5.086621143146182e-06,
+      "loss": 0.2335,
+      "step": 15965
+    },
+    {
+      "epoch": 2.72578305026884,
+      "grad_norm": 0.3458854817400443,
+      "learning_rate": 5.070814365199797e-06,
+      "loss": 0.266,
+      "step": 15970
+    },
+    {
+      "epoch": 2.726636511052317,
+      "grad_norm": 0.3768715108688561,
+      "learning_rate": 5.055007587253414e-06,
+      "loss": 0.2441,
+      "step": 15975
+    },
+    {
+      "epoch": 2.7274899718357943,
+      "grad_norm": 0.401465053041461,
+      "learning_rate": 5.0392008093070315e-06,
+      "loss": 0.2586,
+      "step": 15980
+    },
+    {
+      "epoch": 2.7283434326192713,
+      "grad_norm": 0.35909876371708843,
+      "learning_rate": 5.023394031360647e-06,
+      "loss": 0.2668,
+      "step": 15985
+    },
+    {
+      "epoch": 2.729196893402748,
+      "grad_norm": 0.3404648389384397,
+      "learning_rate": 5.007587253414264e-06,
+      "loss": 0.2799,
+      "step": 15990
+    },
+    {
+      "epoch": 2.730050354186225,
+      "grad_norm": 0.3559294370769041,
+      "learning_rate": 4.991780475467881e-06,
+      "loss": 0.2634,
+      "step": 15995
+    },
+    {
+      "epoch": 2.730903814969702,
+      "grad_norm": 0.3254507220933428,
+      "learning_rate": 4.975973697521497e-06,
+      "loss": 0.2711,
+      "step": 16000
+    },
+    {
+      "epoch": 2.7317572757531794,
+      "grad_norm": 0.3289249884006375,
+      "learning_rate": 4.960166919575114e-06,
+      "loss": 0.2674,
+      "step": 16005
+    },
+    {
+      "epoch": 2.732610736536656,
+      "grad_norm": 0.35530721566450874,
+      "learning_rate": 4.944360141628731e-06,
+      "loss": 0.2764,
+      "step": 16010
+    },
+    {
+      "epoch": 2.7334641973201332,
+      "grad_norm": 0.3322818786935116,
+      "learning_rate": 4.928553363682347e-06,
+      "loss": 0.2568,
+      "step": 16015
+    },
+    {
+      "epoch": 2.73431765810361,
+      "grad_norm": 0.4173584135881613,
+      "learning_rate": 4.9127465857359635e-06,
+      "loss": 0.3052,
+      "step": 16020
+    },
+    {
+      "epoch": 2.735171118887087,
+      "grad_norm": 0.34952729726552456,
+      "learning_rate": 4.896939807789581e-06,
+      "loss": 0.2538,
+      "step": 16025
+    },
+    {
+      "epoch": 2.736024579670564,
+      "grad_norm": 0.36015010665089303,
+      "learning_rate": 4.881133029843197e-06,
+      "loss": 0.2727,
+      "step": 16030
+    },
+    {
+      "epoch": 2.736878040454041,
+      "grad_norm": 0.36300589588357357,
+      "learning_rate": 4.865326251896813e-06,
+      "loss": 0.2749,
+      "step": 16035
+    },
+    {
+      "epoch": 2.7377315012375183,
+      "grad_norm": 0.3523399666869812,
+      "learning_rate": 4.84951947395043e-06,
+      "loss": 0.2735,
+      "step": 16040
+    },
+    {
+      "epoch": 2.738584962020995,
+      "grad_norm": 0.32157226769536656,
+      "learning_rate": 4.833712696004047e-06,
+      "loss": 0.2646,
+      "step": 16045
+    },
+    {
+      "epoch": 2.739438422804472,
+      "grad_norm": 0.31282308394905234,
+      "learning_rate": 4.817905918057663e-06,
+      "loss": 0.2716,
+      "step": 16050
+    },
+    {
+      "epoch": 2.740291883587949,
+      "grad_norm": 0.35253803863997824,
+      "learning_rate": 4.80209914011128e-06,
+      "loss": 0.2824,
+      "step": 16055
+    },
+    {
+      "epoch": 2.741145344371426,
+      "grad_norm": 0.3998407459813653,
+      "learning_rate": 4.786292362164896e-06,
+      "loss": 0.2555,
+      "step": 16060
+    },
+    {
+      "epoch": 2.7419988051549034,
+      "grad_norm": 0.3330651152990024,
+      "learning_rate": 4.770485584218513e-06,
+      "loss": 0.234,
+      "step": 16065
+    },
+    {
+      "epoch": 2.7428522659383803,
+      "grad_norm": 0.34791339272080873,
+      "learning_rate": 4.75467880627213e-06,
+      "loss": 0.2464,
+      "step": 16070
+    },
+    {
+      "epoch": 2.743705726721857,
+      "grad_norm": 0.39063260922066667,
+      "learning_rate": 4.738872028325746e-06,
+      "loss": 0.2463,
+      "step": 16075
+    },
+    {
+      "epoch": 2.744559187505334,
+      "grad_norm": 0.32492035824152354,
+      "learning_rate": 4.723065250379363e-06,
+      "loss": 0.2537,
+      "step": 16080
+    },
+    {
+      "epoch": 2.745412648288811,
+      "grad_norm": 0.3731137115036364,
+      "learning_rate": 4.7072584724329795e-06,
+      "loss": 0.2679,
+      "step": 16085
+    },
+    {
+      "epoch": 2.746266109072288,
+      "grad_norm": 0.3456348621232528,
+      "learning_rate": 4.691451694486596e-06,
+      "loss": 0.2378,
+      "step": 16090
+    },
+    {
+      "epoch": 2.747119569855765,
+      "grad_norm": 0.6624254071052443,
+      "learning_rate": 4.675644916540213e-06,
+      "loss": 0.2481,
+      "step": 16095
+    },
+    {
+      "epoch": 2.7479730306392423,
+      "grad_norm": 0.355392307772178,
+      "learning_rate": 4.659838138593829e-06,
+      "loss": 0.2532,
+      "step": 16100
+    },
+    {
+      "epoch": 2.748826491422719,
+      "grad_norm": 0.3152096224980243,
+      "learning_rate": 4.644031360647446e-06,
+      "loss": 0.2586,
+      "step": 16105
+    },
+    {
+      "epoch": 2.749679952206196,
+      "grad_norm": 0.4539853050997346,
+      "learning_rate": 4.6282245827010625e-06,
+      "loss": 0.255,
+      "step": 16110
+    },
+    {
+      "epoch": 2.750533412989673,
+      "grad_norm": 0.36829194095607254,
+      "learning_rate": 4.612417804754679e-06,
+      "loss": 0.2704,
+      "step": 16115
+    },
+    {
+      "epoch": 2.75138687377315,
+      "grad_norm": 0.33596245320254203,
+      "learning_rate": 4.596611026808296e-06,
+      "loss": 0.2722,
+      "step": 16120
+    },
+    {
+      "epoch": 2.7522403345566273,
+      "grad_norm": 0.3484077352550164,
+      "learning_rate": 4.580804248861912e-06,
+      "loss": 0.2832,
+      "step": 16125
+    },
+    {
+      "epoch": 2.7530937953401042,
+      "grad_norm": 0.30581596719805965,
+      "learning_rate": 4.564997470915529e-06,
+      "loss": 0.2872,
+      "step": 16130
+    },
+    {
+      "epoch": 2.753947256123581,
+      "grad_norm": 0.35869044112332943,
+      "learning_rate": 4.5491906929691456e-06,
+      "loss": 0.2911,
+      "step": 16135
+    },
+    {
+      "epoch": 2.754800716907058,
+      "grad_norm": 0.3351541729240264,
+      "learning_rate": 4.533383915022762e-06,
+      "loss": 0.2431,
+      "step": 16140
+    },
+    {
+      "epoch": 2.755654177690535,
+      "grad_norm": 0.3364680381630783,
+      "learning_rate": 4.517577137076379e-06,
+      "loss": 0.2764,
+      "step": 16145
+    },
+    {
+      "epoch": 2.7565076384740124,
+      "grad_norm": 0.38641523054155913,
+      "learning_rate": 4.501770359129995e-06,
+      "loss": 0.2741,
+      "step": 16150
+    },
+    {
+      "epoch": 2.757361099257489,
+      "grad_norm": 0.46147946741974905,
+      "learning_rate": 4.485963581183612e-06,
+      "loss": 0.2459,
+      "step": 16155
+    },
+    {
+      "epoch": 2.758214560040966,
+      "grad_norm": 0.346631136401872,
+      "learning_rate": 4.470156803237229e-06,
+      "loss": 0.2617,
+      "step": 16160
+    },
+    {
+      "epoch": 2.759068020824443,
+      "grad_norm": 0.3563142377923633,
+      "learning_rate": 4.454350025290844e-06,
+      "loss": 0.2447,
+      "step": 16165
+    },
+    {
+      "epoch": 2.75992148160792,
+      "grad_norm": 0.3205055929160683,
+      "learning_rate": 4.438543247344462e-06,
+      "loss": 0.2811,
+      "step": 16170
+    },
+    {
+      "epoch": 2.760774942391397,
+      "grad_norm": 0.3543129423537271,
+      "learning_rate": 4.4227364693980785e-06,
+      "loss": 0.2739,
+      "step": 16175
+    },
+    {
+      "epoch": 2.761628403174874,
+      "grad_norm": 0.7640148553266405,
+      "learning_rate": 4.406929691451694e-06,
+      "loss": 0.2907,
+      "step": 16180
+    },
+    {
+      "epoch": 2.7624818639583513,
+      "grad_norm": 0.30945303636667076,
+      "learning_rate": 4.391122913505312e-06,
+      "loss": 0.2674,
+      "step": 16185
+    },
+    {
+      "epoch": 2.763335324741828,
+      "grad_norm": 0.3032270814697051,
+      "learning_rate": 4.375316135558928e-06,
+      "loss": 0.2625,
+      "step": 16190
+    },
+    {
+      "epoch": 2.764188785525305,
+      "grad_norm": 0.3517085066160985,
+      "learning_rate": 4.359509357612544e-06,
+      "loss": 0.2671,
+      "step": 16195
+    },
+    {
+      "epoch": 2.765042246308782,
+      "grad_norm": 0.3879520393018638,
+      "learning_rate": 4.3437025796661615e-06,
+      "loss": 0.2581,
+      "step": 16200
+    },
+    {
+      "epoch": 2.765895707092259,
+      "grad_norm": 0.4244100788756175,
+      "learning_rate": 4.327895801719778e-06,
+      "loss": 0.2507,
+      "step": 16205
+    },
+    {
+      "epoch": 2.7667491678757363,
+      "grad_norm": 0.3199529571535488,
+      "learning_rate": 4.312089023773394e-06,
+      "loss": 0.2713,
+      "step": 16210
+    },
+    {
+      "epoch": 2.7676026286592132,
+      "grad_norm": 0.5247639264974975,
+      "learning_rate": 4.296282245827011e-06,
+      "loss": 0.2678,
+      "step": 16215
+    },
+    {
+      "epoch": 2.76845608944269,
+      "grad_norm": 0.3445060865346845,
+      "learning_rate": 4.280475467880628e-06,
+      "loss": 0.2706,
+      "step": 16220
+    },
+    {
+      "epoch": 2.769309550226167,
+      "grad_norm": 0.3944631570601775,
+      "learning_rate": 4.264668689934244e-06,
+      "loss": 0.2693,
+      "step": 16225
+    },
+    {
+      "epoch": 2.770163011009644,
+      "grad_norm": 0.36783086003405474,
+      "learning_rate": 4.24886191198786e-06,
+      "loss": 0.2809,
+      "step": 16230
+    },
+    {
+      "epoch": 2.7710164717931214,
+      "grad_norm": 0.3638614241928813,
+      "learning_rate": 4.233055134041477e-06,
+      "loss": 0.255,
+      "step": 16235
+    },
+    {
+      "epoch": 2.771869932576598,
+      "grad_norm": 0.5399309369816859,
+      "learning_rate": 4.2172483560950935e-06,
+      "loss": 0.2785,
+      "step": 16240
+    },
+    {
+      "epoch": 2.7727233933600752,
+      "grad_norm": 0.3501567327898987,
+      "learning_rate": 4.20144157814871e-06,
+      "loss": 0.2741,
+      "step": 16245
+    },
+    {
+      "epoch": 2.773576854143552,
+      "grad_norm": 0.31510586430391474,
+      "learning_rate": 4.185634800202327e-06,
+      "loss": 0.2707,
+      "step": 16250
+    },
+    {
+      "epoch": 2.774430314927029,
+      "grad_norm": 0.3432568121675401,
+      "learning_rate": 4.169828022255943e-06,
+      "loss": 0.2728,
+      "step": 16255
+    },
+    {
+      "epoch": 2.775283775710506,
+      "grad_norm": 0.3380908042725392,
+      "learning_rate": 4.15402124430956e-06,
+      "loss": 0.286,
+      "step": 16260
+    },
+    {
+      "epoch": 2.776137236493983,
+      "grad_norm": 0.3451036213791636,
+      "learning_rate": 4.138214466363177e-06,
+      "loss": 0.2497,
+      "step": 16265
+    },
+    {
+      "epoch": 2.7769906972774603,
+      "grad_norm": 0.3671258981765301,
+      "learning_rate": 4.122407688416793e-06,
+      "loss": 0.2403,
+      "step": 16270
+    },
+    {
+      "epoch": 2.777844158060937,
+      "grad_norm": 0.6107452247947716,
+      "learning_rate": 4.10660091047041e-06,
+      "loss": 0.257,
+      "step": 16275
+    },
+    {
+      "epoch": 2.778697618844414,
+      "grad_norm": 0.3152322065227391,
+      "learning_rate": 4.0907941325240264e-06,
+      "loss": 0.2649,
+      "step": 16280
+    },
+    {
+      "epoch": 2.779551079627891,
+      "grad_norm": 0.35977789232362667,
+      "learning_rate": 4.074987354577643e-06,
+      "loss": 0.2451,
+      "step": 16285
+    },
+    {
+      "epoch": 2.780404540411368,
+      "grad_norm": 0.3628960931576398,
+      "learning_rate": 4.05918057663126e-06,
+      "loss": 0.2593,
+      "step": 16290
+    },
+    {
+      "epoch": 2.7812580011948453,
+      "grad_norm": 0.3123003825387298,
+      "learning_rate": 4.043373798684876e-06,
+      "loss": 0.261,
+      "step": 16295
+    },
+    {
+      "epoch": 2.782111461978322,
+      "grad_norm": 0.334998792544953,
+      "learning_rate": 4.027567020738493e-06,
+      "loss": 0.2563,
+      "step": 16300
+    },
+    {
+      "epoch": 2.782964922761799,
+      "grad_norm": 0.311287963703417,
+      "learning_rate": 4.0117602427921095e-06,
+      "loss": 0.2662,
+      "step": 16305
+    },
+    {
+      "epoch": 2.783818383545276,
+      "grad_norm": 0.400043297606707,
+      "learning_rate": 3.995953464845726e-06,
+      "loss": 0.2537,
+      "step": 16310
+    },
+    {
+      "epoch": 2.784671844328753,
+      "grad_norm": 0.33348227720765716,
+      "learning_rate": 3.980146686899343e-06,
+      "loss": 0.2489,
+      "step": 16315
+    },
+    {
+      "epoch": 2.78552530511223,
+      "grad_norm": 0.32342434498550293,
+      "learning_rate": 3.964339908952959e-06,
+      "loss": 0.2484,
+      "step": 16320
+    },
+    {
+      "epoch": 2.786378765895707,
+      "grad_norm": 0.3383306821346228,
+      "learning_rate": 3.948533131006576e-06,
+      "loss": 0.2678,
+      "step": 16325
+    },
+    {
+      "epoch": 2.7872322266791842,
+      "grad_norm": 0.333702445040295,
+      "learning_rate": 3.9327263530601925e-06,
+      "loss": 0.2498,
+      "step": 16330
+    },
+    {
+      "epoch": 2.788085687462661,
+      "grad_norm": 0.3529633567573724,
+      "learning_rate": 3.916919575113809e-06,
+      "loss": 0.2629,
+      "step": 16335
+    },
+    {
+      "epoch": 2.788939148246138,
+      "grad_norm": 0.3188183771237849,
+      "learning_rate": 3.901112797167426e-06,
+      "loss": 0.2715,
+      "step": 16340
+    },
+    {
+      "epoch": 2.789792609029615,
+      "grad_norm": 0.3186689929202362,
+      "learning_rate": 3.885306019221042e-06,
+      "loss": 0.2703,
+      "step": 16345
+    },
+    {
+      "epoch": 2.790646069813092,
+      "grad_norm": 0.359627613155441,
+      "learning_rate": 3.869499241274659e-06,
+      "loss": 0.2653,
+      "step": 16350
+    },
+    {
+      "epoch": 2.7914995305965693,
+      "grad_norm": 0.3679817694329449,
+      "learning_rate": 3.853692463328276e-06,
+      "loss": 0.2694,
+      "step": 16355
+    },
+    {
+      "epoch": 2.7923529913800462,
+      "grad_norm": 0.3084239161690286,
+      "learning_rate": 3.837885685381891e-06,
+      "loss": 0.2804,
+      "step": 16360
+    },
+    {
+      "epoch": 2.793206452163523,
+      "grad_norm": 0.3551961472696471,
+      "learning_rate": 3.822078907435509e-06,
+      "loss": 0.2701,
+      "step": 16365
+    },
+    {
+      "epoch": 2.794059912947,
+      "grad_norm": 0.3466176637809046,
+      "learning_rate": 3.8062721294891254e-06,
+      "loss": 0.2719,
+      "step": 16370
+    },
+    {
+      "epoch": 2.794913373730477,
+      "grad_norm": 0.3539991568159895,
+      "learning_rate": 3.7904653515427416e-06,
+      "loss": 0.2529,
+      "step": 16375
+    },
+    {
+      "epoch": 2.7957668345139544,
+      "grad_norm": 0.33891118964849615,
+      "learning_rate": 3.7746585735963582e-06,
+      "loss": 0.2636,
+      "step": 16380
+    },
+    {
+      "epoch": 2.796620295297431,
+      "grad_norm": 0.38201978122494606,
+      "learning_rate": 3.7588517956499753e-06,
+      "loss": 0.2641,
+      "step": 16385
+    },
+    {
+      "epoch": 2.797473756080908,
+      "grad_norm": 0.38292117731613895,
+      "learning_rate": 3.7430450177035914e-06,
+      "loss": 0.2561,
+      "step": 16390
+    },
+    {
+      "epoch": 2.798327216864385,
+      "grad_norm": 0.375796386857593,
+      "learning_rate": 3.727238239757208e-06,
+      "loss": 0.2706,
+      "step": 16395
+    },
+    {
+      "epoch": 2.799180677647862,
+      "grad_norm": 0.3432949267689113,
+      "learning_rate": 3.711431461810825e-06,
+      "loss": 0.2422,
+      "step": 16400
+    },
+    {
+      "epoch": 2.800034138431339,
+      "grad_norm": 0.34469894777136945,
+      "learning_rate": 3.6956246838644413e-06,
+      "loss": 0.256,
+      "step": 16405
+    },
+    {
+      "epoch": 2.800887599214816,
+      "grad_norm": 0.3848539159422196,
+      "learning_rate": 3.679817905918058e-06,
+      "loss": 0.2747,
+      "step": 16410
+    },
+    {
+      "epoch": 2.8017410599982933,
+      "grad_norm": 0.3604161648728695,
+      "learning_rate": 3.664011127971675e-06,
+      "loss": 0.2712,
+      "step": 16415
+    },
+    {
+      "epoch": 2.80259452078177,
+      "grad_norm": 0.33661256175350984,
+      "learning_rate": 3.648204350025291e-06,
+      "loss": 0.2836,
+      "step": 16420
+    },
+    {
+      "epoch": 2.803447981565247,
+      "grad_norm": 0.3746634874315744,
+      "learning_rate": 3.6323975720789077e-06,
+      "loss": 0.2765,
+      "step": 16425
+    },
+    {
+      "epoch": 2.804301442348724,
+      "grad_norm": 0.3412400860143434,
+      "learning_rate": 3.616590794132524e-06,
+      "loss": 0.2618,
+      "step": 16430
+    },
+    {
+      "epoch": 2.805154903132201,
+      "grad_norm": 0.4037360350688074,
+      "learning_rate": 3.600784016186141e-06,
+      "loss": 0.2663,
+      "step": 16435
+    },
+    {
+      "epoch": 2.8060083639156783,
+      "grad_norm": 0.36083859316854466,
+      "learning_rate": 3.5849772382397575e-06,
+      "loss": 0.2419,
+      "step": 16440
+    },
+    {
+      "epoch": 2.8068618246991552,
+      "grad_norm": 0.31857929374958394,
+      "learning_rate": 3.5691704602933737e-06,
+      "loss": 0.2917,
+      "step": 16445
+    },
+    {
+      "epoch": 2.807715285482632,
+      "grad_norm": 0.345950627257534,
+      "learning_rate": 3.5533636823469903e-06,
+      "loss": 0.2633,
+      "step": 16450
+    },
+    {
+      "epoch": 2.808568746266109,
+      "grad_norm": 0.33337471948836517,
+      "learning_rate": 3.5375569044006074e-06,
+      "loss": 0.2529,
+      "step": 16455
+    },
+    {
+      "epoch": 2.809422207049586,
+      "grad_norm": 0.33150891064007126,
+      "learning_rate": 3.5217501264542236e-06,
+      "loss": 0.261,
+      "step": 16460
+    },
+    {
+      "epoch": 2.810275667833063,
+      "grad_norm": 0.34937316105036914,
+      "learning_rate": 3.50594334850784e-06,
+      "loss": 0.2729,
+      "step": 16465
+    },
+    {
+      "epoch": 2.81112912861654,
+      "grad_norm": 0.36289531075289244,
+      "learning_rate": 3.490136570561457e-06,
+      "loss": 0.2439,
+      "step": 16470
+    },
+    {
+      "epoch": 2.811982589400017,
+      "grad_norm": 0.2887422562013696,
+      "learning_rate": 3.4743297926150734e-06,
+      "loss": 0.2507,
+      "step": 16475
+    },
+    {
+      "epoch": 2.812836050183494,
+      "grad_norm": 0.34289352944168966,
+      "learning_rate": 3.45852301466869e-06,
+      "loss": 0.253,
+      "step": 16480
+    },
+    {
+      "epoch": 2.813689510966971,
+      "grad_norm": 0.30363644256498634,
+      "learning_rate": 3.442716236722307e-06,
+      "loss": 0.2637,
+      "step": 16485
+    },
+    {
+      "epoch": 2.814542971750448,
+      "grad_norm": 0.34353385014258925,
+      "learning_rate": 3.4269094587759232e-06,
+      "loss": 0.277,
+      "step": 16490
+    },
+    {
+      "epoch": 2.815396432533925,
+      "grad_norm": 0.38427960033906927,
+      "learning_rate": 3.41110268082954e-06,
+      "loss": 0.2897,
+      "step": 16495
+    },
+    {
+      "epoch": 2.8162498933174023,
+      "grad_norm": 0.3616312249215327,
+      "learning_rate": 3.395295902883156e-06,
+      "loss": 0.2602,
+      "step": 16500
+    },
+    {
+      "epoch": 2.817103354100879,
+      "grad_norm": 0.37667058584403595,
+      "learning_rate": 3.379489124936773e-06,
+      "loss": 0.2548,
+      "step": 16505
+    },
+    {
+      "epoch": 2.817956814884356,
+      "grad_norm": 0.32923855347540354,
+      "learning_rate": 3.3636823469903897e-06,
+      "loss": 0.2705,
+      "step": 16510
+    },
+    {
+      "epoch": 2.818810275667833,
+      "grad_norm": 0.3135200057575021,
+      "learning_rate": 3.347875569044006e-06,
+      "loss": 0.253,
+      "step": 16515
+    },
+    {
+      "epoch": 2.81966373645131,
+      "grad_norm": 0.3142372830706972,
+      "learning_rate": 3.332068791097623e-06,
+      "loss": 0.2647,
+      "step": 16520
+    },
+    {
+      "epoch": 2.8205171972347873,
+      "grad_norm": 0.5160670686122224,
+      "learning_rate": 3.3162620131512395e-06,
+      "loss": 0.2705,
+      "step": 16525
+    },
+    {
+      "epoch": 2.821370658018264,
+      "grad_norm": 0.3462859455541581,
+      "learning_rate": 3.3004552352048557e-06,
+      "loss": 0.2583,
+      "step": 16530
+    },
+    {
+      "epoch": 2.822224118801741,
+      "grad_norm": 0.3189931770031935,
+      "learning_rate": 3.2846484572584727e-06,
+      "loss": 0.2486,
+      "step": 16535
+    },
+    {
+      "epoch": 2.823077579585218,
+      "grad_norm": 0.3672142254284251,
+      "learning_rate": 3.2688416793120893e-06,
+      "loss": 0.2425,
+      "step": 16540
+    },
+    {
+      "epoch": 2.823931040368695,
+      "grad_norm": 0.32596923087577295,
+      "learning_rate": 3.2530349013657055e-06,
+      "loss": 0.2783,
+      "step": 16545
+    },
+    {
+      "epoch": 2.824784501152172,
+      "grad_norm": 0.3467893622117871,
+      "learning_rate": 3.2372281234193226e-06,
+      "loss": 0.2707,
+      "step": 16550
+    },
+    {
+      "epoch": 2.825637961935649,
+      "grad_norm": 0.3396071877068457,
+      "learning_rate": 3.2214213454729387e-06,
+      "loss": 0.2367,
+      "step": 16555
+    },
+    {
+      "epoch": 2.8264914227191262,
+      "grad_norm": 0.3294933065611471,
+      "learning_rate": 3.2056145675265554e-06,
+      "loss": 0.2577,
+      "step": 16560
+    },
+    {
+      "epoch": 2.827344883502603,
+      "grad_norm": 0.32399162837887974,
+      "learning_rate": 3.1898077895801724e-06,
+      "loss": 0.2442,
+      "step": 16565
+    },
+    {
+      "epoch": 2.82819834428608,
+      "grad_norm": 0.3586333716102596,
+      "learning_rate": 3.1740010116337886e-06,
+      "loss": 0.2592,
+      "step": 16570
+    },
+    {
+      "epoch": 2.829051805069557,
+      "grad_norm": 0.3454188547978276,
+      "learning_rate": 3.158194233687405e-06,
+      "loss": 0.2542,
+      "step": 16575
+    },
+    {
+      "epoch": 2.829905265853034,
+      "grad_norm": 0.3560102356813727,
+      "learning_rate": 3.1423874557410222e-06,
+      "loss": 0.2559,
+      "step": 16580
+    },
+    {
+      "epoch": 2.8307587266365113,
+      "grad_norm": 0.3033730297464928,
+      "learning_rate": 3.1265806777946384e-06,
+      "loss": 0.2606,
+      "step": 16585
+    },
+    {
+      "epoch": 2.831612187419988,
+      "grad_norm": 0.3531667605024932,
+      "learning_rate": 3.110773899848255e-06,
+      "loss": 0.258,
+      "step": 16590
+    },
+    {
+      "epoch": 2.832465648203465,
+      "grad_norm": 0.3052730678155858,
+      "learning_rate": 3.0949671219018716e-06,
+      "loss": 0.2556,
+      "step": 16595
+    },
+    {
+      "epoch": 2.833319108986942,
+      "grad_norm": 0.43194185162288795,
+      "learning_rate": 3.0791603439554882e-06,
+      "loss": 0.251,
+      "step": 16600
+    },
+    {
+      "epoch": 2.834172569770419,
+      "grad_norm": 0.39939057437506964,
+      "learning_rate": 3.063353566009105e-06,
+      "loss": 0.2577,
+      "step": 16605
+    },
+    {
+      "epoch": 2.835026030553896,
+      "grad_norm": 0.36369109290470947,
+      "learning_rate": 3.0475467880627215e-06,
+      "loss": 0.2647,
+      "step": 16610
+    },
+    {
+      "epoch": 2.835879491337373,
+      "grad_norm": 0.38320813822635386,
+      "learning_rate": 3.031740010116338e-06,
+      "loss": 0.2514,
+      "step": 16615
+    },
+    {
+      "epoch": 2.83673295212085,
+      "grad_norm": 0.323973442141843,
+      "learning_rate": 3.0159332321699547e-06,
+      "loss": 0.2558,
+      "step": 16620
+    },
+    {
+      "epoch": 2.837586412904327,
+      "grad_norm": 0.4320919774217402,
+      "learning_rate": 3.0001264542235713e-06,
+      "loss": 0.2424,
+      "step": 16625
+    },
+    {
+      "epoch": 2.838439873687804,
+      "grad_norm": 0.32919915583200104,
+      "learning_rate": 2.984319676277188e-06,
+      "loss": 0.2478,
+      "step": 16630
+    },
+    {
+      "epoch": 2.839293334471281,
+      "grad_norm": 0.3450598144167833,
+      "learning_rate": 2.968512898330804e-06,
+      "loss": 0.2364,
+      "step": 16635
+    },
+    {
+      "epoch": 2.840146795254758,
+      "grad_norm": 0.35180820216469266,
+      "learning_rate": 2.952706120384421e-06,
+      "loss": 0.2591,
+      "step": 16640
+    },
+    {
+      "epoch": 2.8410002560382352,
+      "grad_norm": 0.37189345161365905,
+      "learning_rate": 2.9368993424380377e-06,
+      "loss": 0.2577,
+      "step": 16645
+    },
+    {
+      "epoch": 2.841853716821712,
+      "grad_norm": 0.40593001371536785,
+      "learning_rate": 2.921092564491654e-06,
+      "loss": 0.2836,
+      "step": 16650
+    },
+    {
+      "epoch": 2.842707177605189,
+      "grad_norm": 0.4050608018511317,
+      "learning_rate": 2.905285786545271e-06,
+      "loss": 0.2838,
+      "step": 16655
+    },
+    {
+      "epoch": 2.843560638388666,
+      "grad_norm": 0.30154619102257724,
+      "learning_rate": 2.8894790085988876e-06,
+      "loss": 0.2705,
+      "step": 16660
+    },
+    {
+      "epoch": 2.844414099172143,
+      "grad_norm": 0.4456602608003933,
+      "learning_rate": 2.8736722306525038e-06,
+      "loss": 0.2516,
+      "step": 16665
+    },
+    {
+      "epoch": 2.8452675599556203,
+      "grad_norm": 0.3489930788552161,
+      "learning_rate": 2.8578654527061204e-06,
+      "loss": 0.2623,
+      "step": 16670
+    },
+    {
+      "epoch": 2.846121020739097,
+      "grad_norm": 0.4369218078659227,
+      "learning_rate": 2.8420586747597374e-06,
+      "loss": 0.2539,
+      "step": 16675
+    },
+    {
+      "epoch": 2.846974481522574,
+      "grad_norm": 0.3779573979887877,
+      "learning_rate": 2.8262518968133536e-06,
+      "loss": 0.271,
+      "step": 16680
+    },
+    {
+      "epoch": 2.847827942306051,
+      "grad_norm": 0.33284706908380346,
+      "learning_rate": 2.81044511886697e-06,
+      "loss": 0.246,
+      "step": 16685
+    },
+    {
+      "epoch": 2.848681403089528,
+      "grad_norm": 0.3510187464918839,
+      "learning_rate": 2.7946383409205872e-06,
+      "loss": 0.2645,
+      "step": 16690
+    },
+    {
+      "epoch": 2.849534863873005,
+      "grad_norm": 0.35695375118711864,
+      "learning_rate": 2.7788315629742034e-06,
+      "loss": 0.2436,
+      "step": 16695
+    },
+    {
+      "epoch": 2.850388324656482,
+      "grad_norm": 0.31463851753808714,
+      "learning_rate": 2.76302478502782e-06,
+      "loss": 0.263,
+      "step": 16700
+    },
+    {
+      "epoch": 2.851241785439959,
+      "grad_norm": 0.3171378429621654,
+      "learning_rate": 2.7472180070814366e-06,
+      "loss": 0.2733,
+      "step": 16705
+    },
+    {
+      "epoch": 2.852095246223436,
+      "grad_norm": 0.36093161149175196,
+      "learning_rate": 2.7314112291350532e-06,
+      "loss": 0.2594,
+      "step": 16710
+    },
+    {
+      "epoch": 2.852948707006913,
+      "grad_norm": 0.326078521195147,
+      "learning_rate": 2.71560445118867e-06,
+      "loss": 0.2743,
+      "step": 16715
+    },
+    {
+      "epoch": 2.85380216779039,
+      "grad_norm": 0.3291895030132031,
+      "learning_rate": 2.6997976732422865e-06,
+      "loss": 0.2506,
+      "step": 16720
+    },
+    {
+      "epoch": 2.854655628573867,
+      "grad_norm": 0.31928657109893294,
+      "learning_rate": 2.683990895295903e-06,
+      "loss": 0.2527,
+      "step": 16725
+    },
+    {
+      "epoch": 2.8555090893573443,
+      "grad_norm": 0.3426980151395178,
+      "learning_rate": 2.6681841173495197e-06,
+      "loss": 0.2786,
+      "step": 16730
+    },
+    {
+      "epoch": 2.856362550140821,
+      "grad_norm": 0.3754166980773152,
+      "learning_rate": 2.6523773394031363e-06,
+      "loss": 0.2655,
+      "step": 16735
+    },
+    {
+      "epoch": 2.857216010924298,
+      "grad_norm": 0.3203125267306217,
+      "learning_rate": 2.6365705614567525e-06,
+      "loss": 0.2729,
+      "step": 16740
+    },
+    {
+      "epoch": 2.858069471707775,
+      "grad_norm": 0.3178226957464267,
+      "learning_rate": 2.6207637835103695e-06,
+      "loss": 0.281,
+      "step": 16745
+    },
+    {
+      "epoch": 2.858922932491252,
+      "grad_norm": 0.659502083111794,
+      "learning_rate": 2.604957005563986e-06,
+      "loss": 0.2517,
+      "step": 16750
+    },
+    {
+      "epoch": 2.859776393274729,
+      "grad_norm": 0.331750346845881,
+      "learning_rate": 2.5891502276176023e-06,
+      "loss": 0.2807,
+      "step": 16755
+    },
+    {
+      "epoch": 2.860629854058206,
+      "grad_norm": 0.37287672831585017,
+      "learning_rate": 2.5733434496712194e-06,
+      "loss": 0.2557,
+      "step": 16760
+    },
+    {
+      "epoch": 2.861483314841683,
+      "grad_norm": 0.3048070529596822,
+      "learning_rate": 2.557536671724836e-06,
+      "loss": 0.3214,
+      "step": 16765
+    },
+    {
+      "epoch": 2.86233677562516,
+      "grad_norm": 0.36514272157012023,
+      "learning_rate": 2.541729893778452e-06,
+      "loss": 0.2559,
+      "step": 16770
+    },
+    {
+      "epoch": 2.863190236408637,
+      "grad_norm": 0.31132155666325784,
+      "learning_rate": 2.5259231158320688e-06,
+      "loss": 0.2613,
+      "step": 16775
+    },
+    {
+      "epoch": 2.864043697192114,
+      "grad_norm": 0.34363123991986055,
+      "learning_rate": 2.510116337885686e-06,
+      "loss": 0.2509,
+      "step": 16780
+    },
+    {
+      "epoch": 2.864897157975591,
+      "grad_norm": 0.3512546391917189,
+      "learning_rate": 2.494309559939302e-06,
+      "loss": 0.2804,
+      "step": 16785
+    },
+    {
+      "epoch": 2.8657506187590682,
+      "grad_norm": 0.42972027282074027,
+      "learning_rate": 2.4785027819929186e-06,
+      "loss": 0.2461,
+      "step": 16790
+    },
+    {
+      "epoch": 2.866604079542545,
+      "grad_norm": 0.4125362084616382,
+      "learning_rate": 2.4626960040465356e-06,
+      "loss": 0.2688,
+      "step": 16795
+    },
+    {
+      "epoch": 2.867457540326022,
+      "grad_norm": 0.36012044835888873,
+      "learning_rate": 2.446889226100152e-06,
+      "loss": 0.2803,
+      "step": 16800
+    },
+    {
+      "epoch": 2.868311001109499,
+      "grad_norm": 0.316236201957246,
+      "learning_rate": 2.4310824481537684e-06,
+      "loss": 0.2704,
+      "step": 16805
+    },
+    {
+      "epoch": 2.869164461892976,
+      "grad_norm": 0.33919948307445347,
+      "learning_rate": 2.415275670207385e-06,
+      "loss": 0.2616,
+      "step": 16810
+    },
+    {
+      "epoch": 2.8700179226764533,
+      "grad_norm": 0.3961019397081436,
+      "learning_rate": 2.3994688922610016e-06,
+      "loss": 0.2572,
+      "step": 16815
+    },
+    {
+      "epoch": 2.8708713834599298,
+      "grad_norm": 0.3383384992169964,
+      "learning_rate": 2.3836621143146183e-06,
+      "loss": 0.2871,
+      "step": 16820
+    },
+    {
+      "epoch": 2.871724844243407,
+      "grad_norm": 0.3522245154276703,
+      "learning_rate": 2.367855336368235e-06,
+      "loss": 0.2704,
+      "step": 16825
+    },
+    {
+      "epoch": 2.872578305026884,
+      "grad_norm": 0.36351140878303867,
+      "learning_rate": 2.3520485584218515e-06,
+      "loss": 0.2625,
+      "step": 16830
+    },
+    {
+      "epoch": 2.873431765810361,
+      "grad_norm": 0.35581151645056946,
+      "learning_rate": 2.336241780475468e-06,
+      "loss": 0.2558,
+      "step": 16835
+    },
+    {
+      "epoch": 2.874285226593838,
+      "grad_norm": 0.39000427984634917,
+      "learning_rate": 2.3204350025290847e-06,
+      "loss": 0.2587,
+      "step": 16840
+    },
+    {
+      "epoch": 2.875138687377315,
+      "grad_norm": 0.3493985030961467,
+      "learning_rate": 2.304628224582701e-06,
+      "loss": 0.2688,
+      "step": 16845
+    },
+    {
+      "epoch": 2.875992148160792,
+      "grad_norm": 0.36280522890102784,
+      "learning_rate": 2.288821446636318e-06,
+      "loss": 0.2424,
+      "step": 16850
+    },
+    {
+      "epoch": 2.876845608944269,
+      "grad_norm": 0.3612496948579779,
+      "learning_rate": 2.2730146686899345e-06,
+      "loss": 0.2569,
+      "step": 16855
+    },
+    {
+      "epoch": 2.877699069727746,
+      "grad_norm": 0.33287369626609126,
+      "learning_rate": 2.2572078907435507e-06,
+      "loss": 0.27,
+      "step": 16860
+    },
+    {
+      "epoch": 2.878552530511223,
+      "grad_norm": 0.3196880230200889,
+      "learning_rate": 2.2414011127971678e-06,
+      "loss": 0.2624,
+      "step": 16865
+    },
+    {
+      "epoch": 2.8794059912947,
+      "grad_norm": 0.34624656208866267,
+      "learning_rate": 2.2255943348507844e-06,
+      "loss": 0.2607,
+      "step": 16870
+    },
+    {
+      "epoch": 2.8802594520781772,
+      "grad_norm": 0.34565250818861276,
+      "learning_rate": 2.2097875569044005e-06,
+      "loss": 0.2632,
+      "step": 16875
+    },
+    {
+      "epoch": 2.881112912861654,
+      "grad_norm": 0.3560040622984322,
+      "learning_rate": 2.193980778958017e-06,
+      "loss": 0.2438,
+      "step": 16880
+    },
+    {
+      "epoch": 2.881966373645131,
+      "grad_norm": 0.3001588959553635,
+      "learning_rate": 2.178174001011634e-06,
+      "loss": 0.2578,
+      "step": 16885
+    },
+    {
+      "epoch": 2.882819834428608,
+      "grad_norm": 0.3456646370534301,
+      "learning_rate": 2.1623672230652504e-06,
+      "loss": 0.2495,
+      "step": 16890
+    },
+    {
+      "epoch": 2.883673295212085,
+      "grad_norm": 0.3647852415506097,
+      "learning_rate": 2.146560445118867e-06,
+      "loss": 0.2531,
+      "step": 16895
+    },
+    {
+      "epoch": 2.884526755995562,
+      "grad_norm": 0.36332599424420253,
+      "learning_rate": 2.1307536671724836e-06,
+      "loss": 0.2629,
+      "step": 16900
+    },
+    {
+      "epoch": 2.8853802167790388,
+      "grad_norm": 0.3608278828622979,
+      "learning_rate": 2.1149468892261002e-06,
+      "loss": 0.24,
+      "step": 16905
+    },
+    {
+      "epoch": 2.886233677562516,
+      "grad_norm": 0.35888182245405303,
+      "learning_rate": 2.099140111279717e-06,
+      "loss": 0.271,
+      "step": 16910
+    },
+    {
+      "epoch": 2.887087138345993,
+      "grad_norm": 0.3297762540169189,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.2596,
+      "step": 16915
+    },
+    {
+      "epoch": 2.88794059912947,
+      "grad_norm": 0.2938186937645521,
+      "learning_rate": 2.06752655538695e-06,
+      "loss": 0.2693,
+      "step": 16920
+    },
+    {
+      "epoch": 2.888794059912947,
+      "grad_norm": 0.34128051267106463,
+      "learning_rate": 2.0517197774405667e-06,
+      "loss": 0.259,
+      "step": 16925
+    },
+    {
+      "epoch": 2.889647520696424,
+      "grad_norm": 0.34065879725858617,
+      "learning_rate": 2.0359129994941833e-06,
+      "loss": 0.2574,
+      "step": 16930
+    },
+    {
+      "epoch": 2.890500981479901,
+      "grad_norm": 0.2959496947932216,
+      "learning_rate": 2.0201062215477995e-06,
+      "loss": 0.2538,
+      "step": 16935
+    },
+    {
+      "epoch": 2.891354442263378,
+      "grad_norm": 0.3368931588753245,
+      "learning_rate": 2.0042994436014165e-06,
+      "loss": 0.2465,
+      "step": 16940
+    },
+    {
+      "epoch": 2.892207903046855,
+      "grad_norm": 0.3385532028442737,
+      "learning_rate": 1.988492665655033e-06,
+      "loss": 0.2625,
+      "step": 16945
+    },
+    {
+      "epoch": 2.893061363830332,
+      "grad_norm": 0.3209120610716774,
+      "learning_rate": 1.9726858877086493e-06,
+      "loss": 0.2509,
+      "step": 16950
+    },
+    {
+      "epoch": 2.893914824613809,
+      "grad_norm": 0.36205688214370463,
+      "learning_rate": 1.9568791097622663e-06,
+      "loss": 0.2636,
+      "step": 16955
+    },
+    {
+      "epoch": 2.8947682853972863,
+      "grad_norm": 0.350272241148083,
+      "learning_rate": 1.941072331815883e-06,
+      "loss": 0.2445,
+      "step": 16960
+    },
+    {
+      "epoch": 2.8956217461807627,
+      "grad_norm": 0.3831033921951975,
+      "learning_rate": 1.925265553869499e-06,
+      "loss": 0.2839,
+      "step": 16965
+    },
+    {
+      "epoch": 2.89647520696424,
+      "grad_norm": 0.366810305348369,
+      "learning_rate": 1.9094587759231157e-06,
+      "loss": 0.2661,
+      "step": 16970
+    },
+    {
+      "epoch": 2.897328667747717,
+      "grad_norm": 0.4126959808151922,
+      "learning_rate": 1.8936519979767325e-06,
+      "loss": 0.253,
+      "step": 16975
+    },
+    {
+      "epoch": 2.898182128531194,
+      "grad_norm": 0.3442295663869524,
+      "learning_rate": 1.8778452200303492e-06,
+      "loss": 0.2575,
+      "step": 16980
+    },
+    {
+      "epoch": 2.899035589314671,
+      "grad_norm": 0.3357700060647457,
+      "learning_rate": 1.8620384420839656e-06,
+      "loss": 0.2742,
+      "step": 16985
+    },
+    {
+      "epoch": 2.899889050098148,
+      "grad_norm": 0.3563811992415023,
+      "learning_rate": 1.8462316641375824e-06,
+      "loss": 0.2613,
+      "step": 16990
+    },
+    {
+      "epoch": 2.900742510881625,
+      "grad_norm": 0.33392508061801196,
+      "learning_rate": 1.830424886191199e-06,
+      "loss": 0.2704,
+      "step": 16995
+    },
+    {
+      "epoch": 2.901595971665102,
+      "grad_norm": 0.3498820219330395,
+      "learning_rate": 1.8146181082448154e-06,
+      "loss": 0.2575,
+      "step": 17000
+    },
+    {
+      "epoch": 2.902449432448579,
+      "grad_norm": 0.39194232479705393,
+      "learning_rate": 1.798811330298432e-06,
+      "loss": 0.2553,
+      "step": 17005
+    },
+    {
+      "epoch": 2.903302893232056,
+      "grad_norm": 0.322665717495448,
+      "learning_rate": 1.7830045523520488e-06,
+      "loss": 0.2599,
+      "step": 17010
+    },
+    {
+      "epoch": 2.904156354015533,
+      "grad_norm": 0.3199064344376355,
+      "learning_rate": 1.7671977744056652e-06,
+      "loss": 0.2604,
+      "step": 17015
+    },
+    {
+      "epoch": 2.90500981479901,
+      "grad_norm": 0.33304783509978897,
+      "learning_rate": 1.7513909964592818e-06,
+      "loss": 0.288,
+      "step": 17020
+    },
+    {
+      "epoch": 2.905863275582487,
+      "grad_norm": 0.2988482159195373,
+      "learning_rate": 1.7355842185128987e-06,
+      "loss": 0.2516,
+      "step": 17025
+    },
+    {
+      "epoch": 2.906716736365964,
+      "grad_norm": 0.36098320106595283,
+      "learning_rate": 1.719777440566515e-06,
+      "loss": 0.2678,
+      "step": 17030
+    },
+    {
+      "epoch": 2.907570197149441,
+      "grad_norm": 0.3147338768925288,
+      "learning_rate": 1.7039706626201315e-06,
+      "loss": 0.2499,
+      "step": 17035
+    },
+    {
+      "epoch": 2.908423657932918,
+      "grad_norm": 0.3286458368796154,
+      "learning_rate": 1.688163884673748e-06,
+      "loss": 0.2542,
+      "step": 17040
+    },
+    {
+      "epoch": 2.909277118716395,
+      "grad_norm": 0.3300835746620113,
+      "learning_rate": 1.6723571067273649e-06,
+      "loss": 0.2674,
+      "step": 17045
+    },
+    {
+      "epoch": 2.9101305794998718,
+      "grad_norm": 0.4587706915837577,
+      "learning_rate": 1.6565503287809813e-06,
+      "loss": 0.2483,
+      "step": 17050
+    },
+    {
+      "epoch": 2.910984040283349,
+      "grad_norm": 0.3182933111004528,
+      "learning_rate": 1.640743550834598e-06,
+      "loss": 0.2463,
+      "step": 17055
+    },
+    {
+      "epoch": 2.911837501066826,
+      "grad_norm": 0.32706183790213167,
+      "learning_rate": 1.6249367728882147e-06,
+      "loss": 0.2671,
+      "step": 17060
+    },
+    {
+      "epoch": 2.912690961850303,
+      "grad_norm": 0.36289049183139876,
+      "learning_rate": 1.6091299949418311e-06,
+      "loss": 0.2708,
+      "step": 17065
+    },
+    {
+      "epoch": 2.91354442263378,
+      "grad_norm": 0.5199517659672486,
+      "learning_rate": 1.5933232169954477e-06,
+      "loss": 0.2502,
+      "step": 17070
+    },
+    {
+      "epoch": 2.914397883417257,
+      "grad_norm": 0.3400098933724761,
+      "learning_rate": 1.5775164390490641e-06,
+      "loss": 0.2827,
+      "step": 17075
+    },
+    {
+      "epoch": 2.915251344200734,
+      "grad_norm": 0.42551756690623826,
+      "learning_rate": 1.561709661102681e-06,
+      "loss": 0.2649,
+      "step": 17080
+    },
+    {
+      "epoch": 2.916104804984211,
+      "grad_norm": 0.3257057211231263,
+      "learning_rate": 1.5459028831562976e-06,
+      "loss": 0.2587,
+      "step": 17085
+    },
+    {
+      "epoch": 2.916958265767688,
+      "grad_norm": 2.89360211234992,
+      "learning_rate": 1.530096105209914e-06,
+      "loss": 0.2646,
+      "step": 17090
+    },
+    {
+      "epoch": 2.917811726551165,
+      "grad_norm": 0.3178695407416205,
+      "learning_rate": 1.5142893272635306e-06,
+      "loss": 0.266,
+      "step": 17095
+    },
+    {
+      "epoch": 2.918665187334642,
+      "grad_norm": 0.35538425322722705,
+      "learning_rate": 1.4984825493171474e-06,
+      "loss": 0.2762,
+      "step": 17100
+    },
+    {
+      "epoch": 2.9195186481181192,
+      "grad_norm": 0.32448846647232016,
+      "learning_rate": 1.4826757713707638e-06,
+      "loss": 0.2418,
+      "step": 17105
+    },
+    {
+      "epoch": 2.9203721089015957,
+      "grad_norm": 0.32022695751267866,
+      "learning_rate": 1.4668689934243804e-06,
+      "loss": 0.2519,
+      "step": 17110
+    },
+    {
+      "epoch": 2.921225569685073,
+      "grad_norm": 0.3735232904592393,
+      "learning_rate": 1.451062215477997e-06,
+      "loss": 0.2594,
+      "step": 17115
+    },
+    {
+      "epoch": 2.92207903046855,
+      "grad_norm": 0.2901142010923986,
+      "learning_rate": 1.4352554375316136e-06,
+      "loss": 0.2619,
+      "step": 17120
+    },
+    {
+      "epoch": 2.922932491252027,
+      "grad_norm": 0.33541596937195634,
+      "learning_rate": 1.4194486595852302e-06,
+      "loss": 0.2653,
+      "step": 17125
+    },
+    {
+      "epoch": 2.923785952035504,
+      "grad_norm": 0.3377068527301117,
+      "learning_rate": 1.4036418816388468e-06,
+      "loss": 0.2497,
+      "step": 17130
+    },
+    {
+      "epoch": 2.9246394128189808,
+      "grad_norm": 0.4479816694601586,
+      "learning_rate": 1.3878351036924635e-06,
+      "loss": 0.2608,
+      "step": 17135
+    },
+    {
+      "epoch": 2.925492873602458,
+      "grad_norm": 0.5782450043876725,
+      "learning_rate": 1.37202832574608e-06,
+      "loss": 0.25,
+      "step": 17140
+    },
+    {
+      "epoch": 2.926346334385935,
+      "grad_norm": 0.37424540658123895,
+      "learning_rate": 1.3562215477996967e-06,
+      "loss": 0.2582,
+      "step": 17145
+    },
+    {
+      "epoch": 2.927199795169412,
+      "grad_norm": 0.30976988672860206,
+      "learning_rate": 1.340414769853313e-06,
+      "loss": 0.2431,
+      "step": 17150
+    },
+    {
+      "epoch": 2.928053255952889,
+      "grad_norm": 0.4082064861242491,
+      "learning_rate": 1.3246079919069297e-06,
+      "loss": 0.2761,
+      "step": 17155
+    },
+    {
+      "epoch": 2.928906716736366,
+      "grad_norm": 0.3785968340688524,
+      "learning_rate": 1.3088012139605463e-06,
+      "loss": 0.2608,
+      "step": 17160
+    },
+    {
+      "epoch": 2.929760177519843,
+      "grad_norm": 0.35607863049253763,
+      "learning_rate": 1.292994436014163e-06,
+      "loss": 0.2339,
+      "step": 17165
+    },
+    {
+      "epoch": 2.93061363830332,
+      "grad_norm": 0.31075015355334745,
+      "learning_rate": 1.2771876580677795e-06,
+      "loss": 0.2698,
+      "step": 17170
+    },
+    {
+      "epoch": 2.931467099086797,
+      "grad_norm": 0.5132625909881907,
+      "learning_rate": 1.2613808801213961e-06,
+      "loss": 0.2538,
+      "step": 17175
+    },
+    {
+      "epoch": 2.932320559870274,
+      "grad_norm": 0.35694491064633377,
+      "learning_rate": 1.2455741021750127e-06,
+      "loss": 0.2597,
+      "step": 17180
+    },
+    {
+      "epoch": 2.933174020653751,
+      "grad_norm": 0.3568641308814207,
+      "learning_rate": 1.2297673242286293e-06,
+      "loss": 0.2628,
+      "step": 17185
+    },
+    {
+      "epoch": 2.934027481437228,
+      "grad_norm": 0.33791318392593883,
+      "learning_rate": 1.213960546282246e-06,
+      "loss": 0.2628,
+      "step": 17190
+    },
+    {
+      "epoch": 2.9348809422207047,
+      "grad_norm": 0.3405716955626368,
+      "learning_rate": 1.1981537683358624e-06,
+      "loss": 0.269,
+      "step": 17195
+    },
+    {
+      "epoch": 2.935734403004182,
+      "grad_norm": 0.346110735300808,
+      "learning_rate": 1.1823469903894792e-06,
+      "loss": 0.2488,
+      "step": 17200
+    },
+    {
+      "epoch": 2.936587863787659,
+      "grad_norm": 0.33703040230967274,
+      "learning_rate": 1.1665402124430956e-06,
+      "loss": 0.2538,
+      "step": 17205
+    },
+    {
+      "epoch": 2.937441324571136,
+      "grad_norm": 0.3777437358981184,
+      "learning_rate": 1.1507334344967122e-06,
+      "loss": 0.2559,
+      "step": 17210
+    },
+    {
+      "epoch": 2.938294785354613,
+      "grad_norm": 0.3288510930223717,
+      "learning_rate": 1.134926656550329e-06,
+      "loss": 0.2232,
+      "step": 17215
+    },
+    {
+      "epoch": 2.93914824613809,
+      "grad_norm": 0.39222128486985425,
+      "learning_rate": 1.1191198786039454e-06,
+      "loss": 0.2493,
+      "step": 17220
+    },
+    {
+      "epoch": 2.940001706921567,
+      "grad_norm": 0.3671423223654542,
+      "learning_rate": 1.103313100657562e-06,
+      "loss": 0.2552,
+      "step": 17225
+    },
+    {
+      "epoch": 2.940855167705044,
+      "grad_norm": 0.29071214402041584,
+      "learning_rate": 1.0875063227111786e-06,
+      "loss": 0.2609,
+      "step": 17230
+    },
+    {
+      "epoch": 2.941708628488521,
+      "grad_norm": 0.37206901536039744,
+      "learning_rate": 1.0716995447647952e-06,
+      "loss": 0.261,
+      "step": 17235
+    },
+    {
+      "epoch": 2.942562089271998,
+      "grad_norm": 0.33367800371886974,
+      "learning_rate": 1.0558927668184116e-06,
+      "loss": 0.2517,
+      "step": 17240
+    },
+    {
+      "epoch": 2.943415550055475,
+      "grad_norm": 0.4043754763601572,
+      "learning_rate": 1.0400859888720285e-06,
+      "loss": 0.236,
+      "step": 17245
+    },
+    {
+      "epoch": 2.944269010838952,
+      "grad_norm": 0.3134703397311146,
+      "learning_rate": 1.024279210925645e-06,
+      "loss": 0.2445,
+      "step": 17250
+    },
+    {
+      "epoch": 2.9451224716224287,
+      "grad_norm": 0.3158404085722153,
+      "learning_rate": 1.0084724329792615e-06,
+      "loss": 0.2694,
+      "step": 17255
+    },
+    {
+      "epoch": 2.945975932405906,
+      "grad_norm": 0.35844186338513584,
+      "learning_rate": 9.926656550328783e-07,
+      "loss": 0.2828,
+      "step": 17260
+    },
+    {
+      "epoch": 2.946829393189383,
+      "grad_norm": 0.36568115466039464,
+      "learning_rate": 9.768588770864947e-07,
+      "loss": 0.2555,
+      "step": 17265
+    },
+    {
+      "epoch": 2.94768285397286,
+      "grad_norm": 0.344447356357639,
+      "learning_rate": 9.610520991401113e-07,
+      "loss": 0.2666,
+      "step": 17270
+    },
+    {
+      "epoch": 2.948536314756337,
+      "grad_norm": 0.6918163596519813,
+      "learning_rate": 9.452453211937278e-07,
+      "loss": 0.2632,
+      "step": 17275
+    },
+    {
+      "epoch": 2.9493897755398137,
+      "grad_norm": 0.342203389149714,
+      "learning_rate": 9.294385432473445e-07,
+      "loss": 0.2548,
+      "step": 17280
+    },
+    {
+      "epoch": 2.950243236323291,
+      "grad_norm": 0.31059695350905575,
+      "learning_rate": 9.136317653009611e-07,
+      "loss": 0.262,
+      "step": 17285
+    },
+    {
+      "epoch": 2.951096697106768,
+      "grad_norm": 0.3588886087799026,
+      "learning_rate": 8.978249873545776e-07,
+      "loss": 0.2443,
+      "step": 17290
+    },
+    {
+      "epoch": 2.951950157890245,
+      "grad_norm": 0.3725143712962683,
+      "learning_rate": 8.820182094081944e-07,
+      "loss": 0.2575,
+      "step": 17295
+    },
+    {
+      "epoch": 2.952803618673722,
+      "grad_norm": 0.33495674773065304,
+      "learning_rate": 8.662114314618109e-07,
+      "loss": 0.2647,
+      "step": 17300
+    },
+    {
+      "epoch": 2.953657079457199,
+      "grad_norm": 0.32461708186233,
+      "learning_rate": 8.504046535154275e-07,
+      "loss": 0.2672,
+      "step": 17305
+    },
+    {
+      "epoch": 2.954510540240676,
+      "grad_norm": 0.35713058747720877,
+      "learning_rate": 8.34597875569044e-07,
+      "loss": 0.2515,
+      "step": 17310
+    },
+    {
+      "epoch": 2.955364001024153,
+      "grad_norm": 0.3030379252571698,
+      "learning_rate": 8.187910976226607e-07,
+      "loss": 0.2534,
+      "step": 17315
+    },
+    {
+      "epoch": 2.95621746180763,
+      "grad_norm": 0.36601393888480166,
+      "learning_rate": 8.029843196762771e-07,
+      "loss": 0.2649,
+      "step": 17320
+    },
+    {
+      "epoch": 2.957070922591107,
+      "grad_norm": 0.390598927269396,
+      "learning_rate": 7.871775417298938e-07,
+      "loss": 0.2526,
+      "step": 17325
+    },
+    {
+      "epoch": 2.957924383374584,
+      "grad_norm": 0.3233437749380773,
+      "learning_rate": 7.713707637835104e-07,
+      "loss": 0.2621,
+      "step": 17330
+    },
+    {
+      "epoch": 2.9587778441580608,
+      "grad_norm": 0.3719088518172723,
+      "learning_rate": 7.555639858371269e-07,
+      "loss": 0.2597,
+      "step": 17335
+    },
+    {
+      "epoch": 2.9596313049415377,
+      "grad_norm": 0.35190840373585636,
+      "learning_rate": 7.397572078907435e-07,
+      "loss": 0.2503,
+      "step": 17340
+    },
+    {
+      "epoch": 2.960484765725015,
+      "grad_norm": 0.3542481971089052,
+      "learning_rate": 7.239504299443603e-07,
+      "loss": 0.2276,
+      "step": 17345
+    },
+    {
+      "epoch": 2.961338226508492,
+      "grad_norm": 0.3634321686781663,
+      "learning_rate": 7.081436519979768e-07,
+      "loss": 0.2655,
+      "step": 17350
+    },
+    {
+      "epoch": 2.962191687291969,
+      "grad_norm": 0.35586040496530635,
+      "learning_rate": 6.923368740515934e-07,
+      "loss": 0.2702,
+      "step": 17355
+    },
+    {
+      "epoch": 2.963045148075446,
+      "grad_norm": 0.4031032163040328,
+      "learning_rate": 6.7653009610521e-07,
+      "loss": 0.2475,
+      "step": 17360
+    },
+    {
+      "epoch": 2.9638986088589228,
+      "grad_norm": 0.2984585847340631,
+      "learning_rate": 6.607233181588265e-07,
+      "loss": 0.2579,
+      "step": 17365
+    },
+    {
+      "epoch": 2.9647520696424,
+      "grad_norm": 0.3714213545869481,
+      "learning_rate": 6.449165402124431e-07,
+      "loss": 0.2617,
+      "step": 17370
+    },
+    {
+      "epoch": 2.965605530425877,
+      "grad_norm": 0.38048976333390644,
+      "learning_rate": 6.291097622660597e-07,
+      "loss": 0.2612,
+      "step": 17375
+    },
+    {
+      "epoch": 2.966458991209354,
+      "grad_norm": 0.3764636736515448,
+      "learning_rate": 6.133029843196763e-07,
+      "loss": 0.2476,
+      "step": 17380
+    },
+    {
+      "epoch": 2.967312451992831,
+      "grad_norm": 0.3209064073897257,
+      "learning_rate": 5.974962063732929e-07,
+      "loss": 0.2663,
+      "step": 17385
+    },
+    {
+      "epoch": 2.968165912776308,
+      "grad_norm": 0.3659162352363195,
+      "learning_rate": 5.816894284269095e-07,
+      "loss": 0.2647,
+      "step": 17390
+    },
+    {
+      "epoch": 2.969019373559785,
+      "grad_norm": 0.33640959187301384,
+      "learning_rate": 5.65882650480526e-07,
+      "loss": 0.2714,
+      "step": 17395
+    },
+    {
+      "epoch": 2.9698728343432617,
+      "grad_norm": 0.32997317618874944,
+      "learning_rate": 5.500758725341426e-07,
+      "loss": 0.2637,
+      "step": 17400
+    },
+    {
+      "epoch": 2.970726295126739,
+      "grad_norm": 0.367167190417402,
+      "learning_rate": 5.342690945877593e-07,
+      "loss": 0.2638,
+      "step": 17405
+    },
+    {
+      "epoch": 2.971579755910216,
+      "grad_norm": 0.31105922748958315,
+      "learning_rate": 5.184623166413759e-07,
+      "loss": 0.2589,
+      "step": 17410
+    },
+    {
+      "epoch": 2.972433216693693,
+      "grad_norm": 0.32146559627153726,
+      "learning_rate": 5.026555386949924e-07,
+      "loss": 0.2556,
+      "step": 17415
+    },
+    {
+      "epoch": 2.97328667747717,
+      "grad_norm": 0.35793847937600215,
+      "learning_rate": 4.868487607486091e-07,
+      "loss": 0.263,
+      "step": 17420
+    },
+    {
+      "epoch": 2.9741401382606467,
+      "grad_norm": 0.35914293332697467,
+      "learning_rate": 4.7104198280222565e-07,
+      "loss": 0.2454,
+      "step": 17425
+    },
+    {
+      "epoch": 2.974993599044124,
+      "grad_norm": 0.2887747176193576,
+      "learning_rate": 4.552352048558422e-07,
+      "loss": 0.2372,
+      "step": 17430
+    },
+    {
+      "epoch": 2.975847059827601,
+      "grad_norm": 0.35855886226897943,
+      "learning_rate": 4.394284269094588e-07,
+      "loss": 0.2507,
+      "step": 17435
+    },
+    {
+      "epoch": 2.976700520611078,
+      "grad_norm": 0.34818857261223235,
+      "learning_rate": 4.236216489630754e-07,
+      "loss": 0.2767,
+      "step": 17440
+    },
+    {
+      "epoch": 2.977553981394555,
+      "grad_norm": 0.2878549501916922,
+      "learning_rate": 4.0781487101669193e-07,
+      "loss": 0.2726,
+      "step": 17445
+    },
+    {
+      "epoch": 2.9784074421780318,
+      "grad_norm": 0.3111344748617036,
+      "learning_rate": 3.9200809307030854e-07,
+      "loss": 0.2559,
+      "step": 17450
+    },
+    {
+      "epoch": 2.979260902961509,
+      "grad_norm": 0.3719374434219703,
+      "learning_rate": 3.7620131512392515e-07,
+      "loss": 0.2605,
+      "step": 17455
+    },
+    {
+      "epoch": 2.980114363744986,
+      "grad_norm": 0.3466211071811781,
+      "learning_rate": 3.603945371775417e-07,
+      "loss": 0.2701,
+      "step": 17460
+    },
+    {
+      "epoch": 2.980967824528463,
+      "grad_norm": 0.3734051181873891,
+      "learning_rate": 3.445877592311583e-07,
+      "loss": 0.2702,
+      "step": 17465
+    },
+    {
+      "epoch": 2.98182128531194,
+      "grad_norm": 0.35577982396358515,
+      "learning_rate": 3.2878098128477493e-07,
+      "loss": 0.2412,
+      "step": 17470
+    },
+    {
+      "epoch": 2.982674746095417,
+      "grad_norm": 0.3250053297942569,
+      "learning_rate": 3.129742033383915e-07,
+      "loss": 0.2798,
+      "step": 17475
+    },
+    {
+      "epoch": 2.9835282068788938,
+      "grad_norm": 0.2980448644295891,
+      "learning_rate": 2.971674253920081e-07,
+      "loss": 0.2579,
+      "step": 17480
+    },
+    {
+      "epoch": 2.9843816676623707,
+      "grad_norm": 0.308352770806981,
+      "learning_rate": 2.813606474456247e-07,
+      "loss": 0.2782,
+      "step": 17485
+    },
+    {
+      "epoch": 2.985235128445848,
+      "grad_norm": 0.3503886222854535,
+      "learning_rate": 2.6555386949924127e-07,
+      "loss": 0.259,
+      "step": 17490
+    },
+    {
+      "epoch": 2.986088589229325,
+      "grad_norm": 0.38014445587318535,
+      "learning_rate": 2.497470915528579e-07,
+      "loss": 0.2489,
+      "step": 17495
+    },
+    {
+      "epoch": 2.986942050012802,
+      "grad_norm": 0.34281879430471657,
+      "learning_rate": 2.3394031360647446e-07,
+      "loss": 0.2329,
+      "step": 17500
+    },
+    {
+      "epoch": 2.987795510796279,
+      "grad_norm": 0.3149722235884717,
+      "learning_rate": 2.1813353566009108e-07,
+      "loss": 0.2476,
+      "step": 17505
+    },
+    {
+      "epoch": 2.9886489715797557,
+      "grad_norm": 0.4067252040037474,
+      "learning_rate": 2.0232675771370766e-07,
+      "loss": 0.2771,
+      "step": 17510
+    },
+    {
+      "epoch": 2.989502432363233,
+      "grad_norm": 0.3822581744090084,
+      "learning_rate": 1.8651997976732424e-07,
+      "loss": 0.2473,
+      "step": 17515
+    },
+    {
+      "epoch": 2.99035589314671,
+      "grad_norm": 0.34724226273280534,
+      "learning_rate": 1.7071320182094083e-07,
+      "loss": 0.2642,
+      "step": 17520
+    },
+    {
+      "epoch": 2.991209353930187,
+      "grad_norm": 0.3194994143020438,
+      "learning_rate": 1.549064238745574e-07,
+      "loss": 0.2702,
+      "step": 17525
+    },
+    {
+      "epoch": 2.992062814713664,
+      "grad_norm": 0.31852349595999996,
+      "learning_rate": 1.3909964592817402e-07,
+      "loss": 0.2559,
+      "step": 17530
+    },
+    {
+      "epoch": 2.992916275497141,
+      "grad_norm": 0.32612990576667494,
+      "learning_rate": 1.2329286798179058e-07,
+      "loss": 0.2558,
+      "step": 17535
+    },
+    {
+      "epoch": 2.993769736280618,
+      "grad_norm": 0.3423943025137262,
+      "learning_rate": 1.0748609003540719e-07,
+      "loss": 0.2718,
+      "step": 17540
+    },
+    {
+      "epoch": 2.9946231970640946,
+      "grad_norm": 0.3751142492781964,
+      "learning_rate": 9.167931208902377e-08,
+      "loss": 0.244,
+      "step": 17545
+    },
+    {
+      "epoch": 2.995476657847572,
+      "grad_norm": 0.33765844355783725,
+      "learning_rate": 7.587253414264037e-08,
+      "loss": 0.2514,
+      "step": 17550
+    },
+    {
+      "epoch": 2.996330118631049,
+      "grad_norm": 0.34932096769176885,
+      "learning_rate": 6.006575619625696e-08,
+      "loss": 0.2481,
+      "step": 17555
+    },
+    {
+      "epoch": 2.997183579414526,
+      "grad_norm": 0.41118628511885635,
+      "learning_rate": 4.425897824987355e-08,
+      "loss": 0.2469,
+      "step": 17560
+    },
+    {
+      "epoch": 2.9980370401980028,
+      "grad_norm": 0.36482301233773756,
+      "learning_rate": 2.8452200303490138e-08,
+      "loss": 0.2756,
+      "step": 17565
+    },
+    {
+      "epoch": 2.9988905009814797,
+      "grad_norm": 0.323345446655115,
+      "learning_rate": 1.2645422357106729e-08,
+      "loss": 0.2363,
+      "step": 17570
+    },
+    {
+      "epoch": 2.9995732696082618,
+      "step": 17574,
+      "total_flos": 2.7506794053018583e+18,
+      "train_loss": 0.39039740704073767,
+      "train_runtime": 61759.1154,
+      "train_samples_per_second": 4.553,
+      "train_steps_per_second": 0.285
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 17574,
+  "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": 2.7506794053018583e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}