gpt2-sft-dutch / trainer_state.json
BrianGu's picture
Model save
6967d98 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 7454,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00013415615776764153,
"grad_norm": 8.1875,
"learning_rate": 2.6809651474530834e-08,
"loss": 5.4504,
"step": 1
},
{
"epoch": 0.0006707807888382077,
"grad_norm": 8.0625,
"learning_rate": 1.3404825737265417e-07,
"loss": 5.4299,
"step": 5
},
{
"epoch": 0.0013415615776764154,
"grad_norm": 8.8125,
"learning_rate": 2.6809651474530835e-07,
"loss": 5.4123,
"step": 10
},
{
"epoch": 0.002012342366514623,
"grad_norm": 8.9375,
"learning_rate": 4.021447721179625e-07,
"loss": 5.4355,
"step": 15
},
{
"epoch": 0.002683123155352831,
"grad_norm": 8.5,
"learning_rate": 5.361930294906167e-07,
"loss": 5.4206,
"step": 20
},
{
"epoch": 0.0033539039441910384,
"grad_norm": 9.125,
"learning_rate": 6.702412868632709e-07,
"loss": 5.4374,
"step": 25
},
{
"epoch": 0.004024684733029246,
"grad_norm": 8.5625,
"learning_rate": 8.04289544235925e-07,
"loss": 5.4034,
"step": 30
},
{
"epoch": 0.004695465521867454,
"grad_norm": 9.25,
"learning_rate": 9.383378016085791e-07,
"loss": 5.4501,
"step": 35
},
{
"epoch": 0.005366246310705662,
"grad_norm": 9.5,
"learning_rate": 1.0723860589812334e-06,
"loss": 5.4001,
"step": 40
},
{
"epoch": 0.006037027099543869,
"grad_norm": 8.9375,
"learning_rate": 1.2064343163538874e-06,
"loss": 5.4229,
"step": 45
},
{
"epoch": 0.006707807888382077,
"grad_norm": 9.0,
"learning_rate": 1.3404825737265418e-06,
"loss": 5.3579,
"step": 50
},
{
"epoch": 0.007378588677220284,
"grad_norm": 8.9375,
"learning_rate": 1.4745308310991958e-06,
"loss": 5.407,
"step": 55
},
{
"epoch": 0.008049369466058493,
"grad_norm": 9.0,
"learning_rate": 1.60857908847185e-06,
"loss": 5.3885,
"step": 60
},
{
"epoch": 0.0087201502548967,
"grad_norm": 9.0625,
"learning_rate": 1.7426273458445042e-06,
"loss": 5.4106,
"step": 65
},
{
"epoch": 0.009390931043734908,
"grad_norm": 9.125,
"learning_rate": 1.8766756032171582e-06,
"loss": 5.4105,
"step": 70
},
{
"epoch": 0.010061711832573116,
"grad_norm": 9.125,
"learning_rate": 2.0107238605898126e-06,
"loss": 5.4066,
"step": 75
},
{
"epoch": 0.010732492621411323,
"grad_norm": 9.0,
"learning_rate": 2.1447721179624668e-06,
"loss": 5.4043,
"step": 80
},
{
"epoch": 0.01140327341024953,
"grad_norm": 9.0,
"learning_rate": 2.278820375335121e-06,
"loss": 5.4473,
"step": 85
},
{
"epoch": 0.012074054199087738,
"grad_norm": 9.1875,
"learning_rate": 2.4128686327077747e-06,
"loss": 5.3858,
"step": 90
},
{
"epoch": 0.012744834987925946,
"grad_norm": 8.8125,
"learning_rate": 2.5469168900804294e-06,
"loss": 5.4037,
"step": 95
},
{
"epoch": 0.013415615776764154,
"grad_norm": 9.0,
"learning_rate": 2.6809651474530836e-06,
"loss": 5.393,
"step": 100
},
{
"epoch": 0.014086396565602361,
"grad_norm": 8.1875,
"learning_rate": 2.8150134048257378e-06,
"loss": 5.388,
"step": 105
},
{
"epoch": 0.014757177354440569,
"grad_norm": 8.1875,
"learning_rate": 2.9490616621983915e-06,
"loss": 5.3953,
"step": 110
},
{
"epoch": 0.015427958143278776,
"grad_norm": 8.1875,
"learning_rate": 3.0831099195710457e-06,
"loss": 5.3861,
"step": 115
},
{
"epoch": 0.016098738932116986,
"grad_norm": 8.4375,
"learning_rate": 3.2171581769437e-06,
"loss": 5.3817,
"step": 120
},
{
"epoch": 0.01676951972095519,
"grad_norm": 8.3125,
"learning_rate": 3.351206434316354e-06,
"loss": 5.3575,
"step": 125
},
{
"epoch": 0.0174403005097934,
"grad_norm": 8.9375,
"learning_rate": 3.4852546916890083e-06,
"loss": 5.4005,
"step": 130
},
{
"epoch": 0.018111081298631607,
"grad_norm": 8.5,
"learning_rate": 3.6193029490616625e-06,
"loss": 5.4164,
"step": 135
},
{
"epoch": 0.018781862087469816,
"grad_norm": 8.3125,
"learning_rate": 3.7533512064343163e-06,
"loss": 5.4165,
"step": 140
},
{
"epoch": 0.019452642876308022,
"grad_norm": 7.875,
"learning_rate": 3.8873994638069705e-06,
"loss": 5.3457,
"step": 145
},
{
"epoch": 0.02012342366514623,
"grad_norm": 8.5625,
"learning_rate": 4.021447721179625e-06,
"loss": 5.4649,
"step": 150
},
{
"epoch": 0.020794204453984437,
"grad_norm": 8.625,
"learning_rate": 4.155495978552279e-06,
"loss": 5.3878,
"step": 155
},
{
"epoch": 0.021464985242822646,
"grad_norm": 9.4375,
"learning_rate": 4.2895442359249335e-06,
"loss": 5.4015,
"step": 160
},
{
"epoch": 0.022135766031660852,
"grad_norm": 8.25,
"learning_rate": 4.423592493297587e-06,
"loss": 5.3703,
"step": 165
},
{
"epoch": 0.02280654682049906,
"grad_norm": 9.5625,
"learning_rate": 4.557640750670242e-06,
"loss": 5.3875,
"step": 170
},
{
"epoch": 0.023477327609337267,
"grad_norm": 9.125,
"learning_rate": 4.691689008042896e-06,
"loss": 5.3573,
"step": 175
},
{
"epoch": 0.024148108398175477,
"grad_norm": 8.5,
"learning_rate": 4.8257372654155495e-06,
"loss": 5.3552,
"step": 180
},
{
"epoch": 0.024818889187013683,
"grad_norm": 8.5625,
"learning_rate": 4.959785522788204e-06,
"loss": 5.3685,
"step": 185
},
{
"epoch": 0.025489669975851892,
"grad_norm": 7.96875,
"learning_rate": 5.093833780160859e-06,
"loss": 5.3749,
"step": 190
},
{
"epoch": 0.026160450764690098,
"grad_norm": 8.8125,
"learning_rate": 5.2278820375335125e-06,
"loss": 5.3618,
"step": 195
},
{
"epoch": 0.026831231553528307,
"grad_norm": 7.65625,
"learning_rate": 5.361930294906167e-06,
"loss": 5.3699,
"step": 200
},
{
"epoch": 0.027502012342366513,
"grad_norm": 8.5,
"learning_rate": 5.495978552278821e-06,
"loss": 5.3798,
"step": 205
},
{
"epoch": 0.028172793131204722,
"grad_norm": 8.125,
"learning_rate": 5.6300268096514755e-06,
"loss": 5.3516,
"step": 210
},
{
"epoch": 0.02884357392004293,
"grad_norm": 8.1875,
"learning_rate": 5.764075067024129e-06,
"loss": 5.3616,
"step": 215
},
{
"epoch": 0.029514354708881137,
"grad_norm": 8.1875,
"learning_rate": 5.898123324396783e-06,
"loss": 5.3436,
"step": 220
},
{
"epoch": 0.030185135497719347,
"grad_norm": 7.96875,
"learning_rate": 6.032171581769437e-06,
"loss": 5.3178,
"step": 225
},
{
"epoch": 0.030855916286557553,
"grad_norm": 8.25,
"learning_rate": 6.1662198391420915e-06,
"loss": 5.2885,
"step": 230
},
{
"epoch": 0.03152669707539576,
"grad_norm": 8.6875,
"learning_rate": 6.300268096514745e-06,
"loss": 5.3546,
"step": 235
},
{
"epoch": 0.03219747786423397,
"grad_norm": 8.5,
"learning_rate": 6.4343163538874e-06,
"loss": 5.3342,
"step": 240
},
{
"epoch": 0.03286825865307218,
"grad_norm": 7.75,
"learning_rate": 6.5683646112600545e-06,
"loss": 5.3256,
"step": 245
},
{
"epoch": 0.03353903944191038,
"grad_norm": 7.875,
"learning_rate": 6.702412868632708e-06,
"loss": 5.3461,
"step": 250
},
{
"epoch": 0.03420982023074859,
"grad_norm": 7.25,
"learning_rate": 6.836461126005363e-06,
"loss": 5.3205,
"step": 255
},
{
"epoch": 0.0348806010195868,
"grad_norm": 7.59375,
"learning_rate": 6.970509383378017e-06,
"loss": 5.2975,
"step": 260
},
{
"epoch": 0.03555138180842501,
"grad_norm": 7.90625,
"learning_rate": 7.104557640750671e-06,
"loss": 5.3069,
"step": 265
},
{
"epoch": 0.03622216259726321,
"grad_norm": 8.625,
"learning_rate": 7.238605898123325e-06,
"loss": 5.3156,
"step": 270
},
{
"epoch": 0.03689294338610142,
"grad_norm": 8.25,
"learning_rate": 7.37265415549598e-06,
"loss": 5.2831,
"step": 275
},
{
"epoch": 0.03756372417493963,
"grad_norm": 7.4375,
"learning_rate": 7.506702412868633e-06,
"loss": 5.3046,
"step": 280
},
{
"epoch": 0.03823450496377784,
"grad_norm": 8.0,
"learning_rate": 7.640750670241287e-06,
"loss": 5.2892,
"step": 285
},
{
"epoch": 0.038905285752616044,
"grad_norm": 8.5,
"learning_rate": 7.774798927613941e-06,
"loss": 5.3001,
"step": 290
},
{
"epoch": 0.03957606654145425,
"grad_norm": 7.375,
"learning_rate": 7.908847184986595e-06,
"loss": 5.273,
"step": 295
},
{
"epoch": 0.04024684733029246,
"grad_norm": 7.8125,
"learning_rate": 8.04289544235925e-06,
"loss": 5.2766,
"step": 300
},
{
"epoch": 0.04091762811913067,
"grad_norm": 7.9375,
"learning_rate": 8.176943699731904e-06,
"loss": 5.2722,
"step": 305
},
{
"epoch": 0.041588408907968874,
"grad_norm": 6.75,
"learning_rate": 8.310991957104558e-06,
"loss": 5.3008,
"step": 310
},
{
"epoch": 0.04225918969680709,
"grad_norm": 8.25,
"learning_rate": 8.445040214477213e-06,
"loss": 5.2501,
"step": 315
},
{
"epoch": 0.04292997048564529,
"grad_norm": 7.78125,
"learning_rate": 8.579088471849867e-06,
"loss": 5.2168,
"step": 320
},
{
"epoch": 0.0436007512744835,
"grad_norm": 7.625,
"learning_rate": 8.71313672922252e-06,
"loss": 5.2447,
"step": 325
},
{
"epoch": 0.044271532063321704,
"grad_norm": 7.875,
"learning_rate": 8.847184986595175e-06,
"loss": 5.2456,
"step": 330
},
{
"epoch": 0.04494231285215992,
"grad_norm": 8.3125,
"learning_rate": 8.98123324396783e-06,
"loss": 5.2333,
"step": 335
},
{
"epoch": 0.04561309364099812,
"grad_norm": 8.0625,
"learning_rate": 9.115281501340484e-06,
"loss": 5.2385,
"step": 340
},
{
"epoch": 0.04628387442983633,
"grad_norm": 8.8125,
"learning_rate": 9.249329758713138e-06,
"loss": 5.249,
"step": 345
},
{
"epoch": 0.046954655218674535,
"grad_norm": 7.78125,
"learning_rate": 9.383378016085791e-06,
"loss": 5.2488,
"step": 350
},
{
"epoch": 0.04762543600751275,
"grad_norm": 7.1875,
"learning_rate": 9.517426273458445e-06,
"loss": 5.1848,
"step": 355
},
{
"epoch": 0.04829621679635095,
"grad_norm": 7.375,
"learning_rate": 9.651474530831099e-06,
"loss": 5.209,
"step": 360
},
{
"epoch": 0.04896699758518916,
"grad_norm": 8.0,
"learning_rate": 9.785522788203754e-06,
"loss": 5.1832,
"step": 365
},
{
"epoch": 0.049637778374027365,
"grad_norm": 8.0625,
"learning_rate": 9.919571045576408e-06,
"loss": 5.1808,
"step": 370
},
{
"epoch": 0.05030855916286558,
"grad_norm": 7.15625,
"learning_rate": 1.0053619302949062e-05,
"loss": 5.1798,
"step": 375
},
{
"epoch": 0.050979339951703784,
"grad_norm": 7.125,
"learning_rate": 1.0187667560321717e-05,
"loss": 5.1462,
"step": 380
},
{
"epoch": 0.05165012074054199,
"grad_norm": 6.25,
"learning_rate": 1.0321715817694371e-05,
"loss": 5.1521,
"step": 385
},
{
"epoch": 0.052320901529380195,
"grad_norm": 6.96875,
"learning_rate": 1.0455764075067025e-05,
"loss": 5.1394,
"step": 390
},
{
"epoch": 0.05299168231821841,
"grad_norm": 6.6875,
"learning_rate": 1.0589812332439679e-05,
"loss": 5.1408,
"step": 395
},
{
"epoch": 0.053662463107056614,
"grad_norm": 6.8125,
"learning_rate": 1.0723860589812334e-05,
"loss": 5.0856,
"step": 400
},
{
"epoch": 0.05433324389589482,
"grad_norm": 7.25,
"learning_rate": 1.0857908847184988e-05,
"loss": 5.0621,
"step": 405
},
{
"epoch": 0.055004024684733026,
"grad_norm": 6.34375,
"learning_rate": 1.0991957104557642e-05,
"loss": 5.0859,
"step": 410
},
{
"epoch": 0.05567480547357124,
"grad_norm": 6.65625,
"learning_rate": 1.1126005361930297e-05,
"loss": 5.0641,
"step": 415
},
{
"epoch": 0.056345586262409444,
"grad_norm": 6.53125,
"learning_rate": 1.1260053619302951e-05,
"loss": 5.1037,
"step": 420
},
{
"epoch": 0.05701636705124765,
"grad_norm": 6.375,
"learning_rate": 1.1394101876675605e-05,
"loss": 5.0467,
"step": 425
},
{
"epoch": 0.05768714784008586,
"grad_norm": 6.40625,
"learning_rate": 1.1528150134048259e-05,
"loss": 5.072,
"step": 430
},
{
"epoch": 0.05835792862892407,
"grad_norm": 6.25,
"learning_rate": 1.1662198391420914e-05,
"loss": 5.0457,
"step": 435
},
{
"epoch": 0.059028709417762275,
"grad_norm": 6.28125,
"learning_rate": 1.1796246648793566e-05,
"loss": 5.0391,
"step": 440
},
{
"epoch": 0.05969949020660048,
"grad_norm": 6.09375,
"learning_rate": 1.193029490616622e-05,
"loss": 5.0458,
"step": 445
},
{
"epoch": 0.06037027099543869,
"grad_norm": 5.71875,
"learning_rate": 1.2064343163538874e-05,
"loss": 5.0164,
"step": 450
},
{
"epoch": 0.0610410517842769,
"grad_norm": 5.875,
"learning_rate": 1.219839142091153e-05,
"loss": 5.0291,
"step": 455
},
{
"epoch": 0.061711832573115105,
"grad_norm": 5.875,
"learning_rate": 1.2332439678284183e-05,
"loss": 4.9909,
"step": 460
},
{
"epoch": 0.06238261336195331,
"grad_norm": 6.09375,
"learning_rate": 1.2466487935656837e-05,
"loss": 5.0022,
"step": 465
},
{
"epoch": 0.06305339415079152,
"grad_norm": 6.0,
"learning_rate": 1.260053619302949e-05,
"loss": 5.0228,
"step": 470
},
{
"epoch": 0.06372417493962973,
"grad_norm": 5.6875,
"learning_rate": 1.2734584450402146e-05,
"loss": 4.9727,
"step": 475
},
{
"epoch": 0.06439495572846794,
"grad_norm": 5.375,
"learning_rate": 1.28686327077748e-05,
"loss": 4.9947,
"step": 480
},
{
"epoch": 0.06506573651730614,
"grad_norm": 5.5625,
"learning_rate": 1.3002680965147454e-05,
"loss": 4.9518,
"step": 485
},
{
"epoch": 0.06573651730614435,
"grad_norm": 5.75,
"learning_rate": 1.3136729222520109e-05,
"loss": 4.9473,
"step": 490
},
{
"epoch": 0.06640729809498255,
"grad_norm": 5.65625,
"learning_rate": 1.3270777479892763e-05,
"loss": 4.9456,
"step": 495
},
{
"epoch": 0.06707807888382077,
"grad_norm": 5.65625,
"learning_rate": 1.3404825737265417e-05,
"loss": 4.9316,
"step": 500
},
{
"epoch": 0.06774885967265898,
"grad_norm": 5.625,
"learning_rate": 1.353887399463807e-05,
"loss": 4.9074,
"step": 505
},
{
"epoch": 0.06841964046149718,
"grad_norm": 5.34375,
"learning_rate": 1.3672922252010726e-05,
"loss": 4.9208,
"step": 510
},
{
"epoch": 0.06909042125033539,
"grad_norm": 5.6875,
"learning_rate": 1.380697050938338e-05,
"loss": 4.9392,
"step": 515
},
{
"epoch": 0.0697612020391736,
"grad_norm": 5.53125,
"learning_rate": 1.3941018766756033e-05,
"loss": 4.9199,
"step": 520
},
{
"epoch": 0.0704319828280118,
"grad_norm": 5.53125,
"learning_rate": 1.4075067024128689e-05,
"loss": 4.9472,
"step": 525
},
{
"epoch": 0.07110276361685001,
"grad_norm": 5.40625,
"learning_rate": 1.4209115281501343e-05,
"loss": 4.9466,
"step": 530
},
{
"epoch": 0.07177354440568823,
"grad_norm": 5.9375,
"learning_rate": 1.4343163538873996e-05,
"loss": 4.8262,
"step": 535
},
{
"epoch": 0.07244432519452643,
"grad_norm": 5.65625,
"learning_rate": 1.447721179624665e-05,
"loss": 4.9246,
"step": 540
},
{
"epoch": 0.07311510598336464,
"grad_norm": 4.96875,
"learning_rate": 1.4611260053619306e-05,
"loss": 4.8883,
"step": 545
},
{
"epoch": 0.07378588677220284,
"grad_norm": 5.28125,
"learning_rate": 1.474530831099196e-05,
"loss": 4.842,
"step": 550
},
{
"epoch": 0.07445666756104105,
"grad_norm": 5.0625,
"learning_rate": 1.4879356568364613e-05,
"loss": 4.8574,
"step": 555
},
{
"epoch": 0.07512744834987926,
"grad_norm": 5.3125,
"learning_rate": 1.5013404825737265e-05,
"loss": 4.8697,
"step": 560
},
{
"epoch": 0.07579822913871746,
"grad_norm": 5.1875,
"learning_rate": 1.5147453083109919e-05,
"loss": 4.8338,
"step": 565
},
{
"epoch": 0.07646900992755568,
"grad_norm": 5.28125,
"learning_rate": 1.5281501340482574e-05,
"loss": 4.8414,
"step": 570
},
{
"epoch": 0.07713979071639389,
"grad_norm": 4.8125,
"learning_rate": 1.541554959785523e-05,
"loss": 4.8836,
"step": 575
},
{
"epoch": 0.07781057150523209,
"grad_norm": 5.375,
"learning_rate": 1.5549597855227882e-05,
"loss": 4.8591,
"step": 580
},
{
"epoch": 0.0784813522940703,
"grad_norm": 5.09375,
"learning_rate": 1.5683646112600538e-05,
"loss": 4.8049,
"step": 585
},
{
"epoch": 0.0791521330829085,
"grad_norm": 4.96875,
"learning_rate": 1.581769436997319e-05,
"loss": 4.8538,
"step": 590
},
{
"epoch": 0.07982291387174671,
"grad_norm": 5.4375,
"learning_rate": 1.5951742627345845e-05,
"loss": 4.8651,
"step": 595
},
{
"epoch": 0.08049369466058492,
"grad_norm": 5.3125,
"learning_rate": 1.60857908847185e-05,
"loss": 4.8247,
"step": 600
},
{
"epoch": 0.08116447544942312,
"grad_norm": 5.0,
"learning_rate": 1.6219839142091153e-05,
"loss": 4.8217,
"step": 605
},
{
"epoch": 0.08183525623826134,
"grad_norm": 5.1875,
"learning_rate": 1.6353887399463808e-05,
"loss": 4.8204,
"step": 610
},
{
"epoch": 0.08250603702709955,
"grad_norm": 5.0,
"learning_rate": 1.6487935656836464e-05,
"loss": 4.8164,
"step": 615
},
{
"epoch": 0.08317681781593775,
"grad_norm": 4.90625,
"learning_rate": 1.6621983914209116e-05,
"loss": 4.8127,
"step": 620
},
{
"epoch": 0.08384759860477596,
"grad_norm": 4.90625,
"learning_rate": 1.675603217158177e-05,
"loss": 4.7912,
"step": 625
},
{
"epoch": 0.08451837939361417,
"grad_norm": 5.21875,
"learning_rate": 1.6890080428954427e-05,
"loss": 4.8293,
"step": 630
},
{
"epoch": 0.08518916018245237,
"grad_norm": 5.125,
"learning_rate": 1.702412868632708e-05,
"loss": 4.8181,
"step": 635
},
{
"epoch": 0.08585994097129059,
"grad_norm": 4.8125,
"learning_rate": 1.7158176943699734e-05,
"loss": 4.7758,
"step": 640
},
{
"epoch": 0.08653072176012878,
"grad_norm": 4.53125,
"learning_rate": 1.7292225201072386e-05,
"loss": 4.8266,
"step": 645
},
{
"epoch": 0.087201502548967,
"grad_norm": 4.90625,
"learning_rate": 1.742627345844504e-05,
"loss": 4.7627,
"step": 650
},
{
"epoch": 0.08787228333780521,
"grad_norm": 4.71875,
"learning_rate": 1.7560321715817697e-05,
"loss": 4.7842,
"step": 655
},
{
"epoch": 0.08854306412664341,
"grad_norm": 4.6875,
"learning_rate": 1.769436997319035e-05,
"loss": 4.733,
"step": 660
},
{
"epoch": 0.08921384491548162,
"grad_norm": 4.8125,
"learning_rate": 1.7828418230563005e-05,
"loss": 4.7759,
"step": 665
},
{
"epoch": 0.08988462570431983,
"grad_norm": 4.5,
"learning_rate": 1.796246648793566e-05,
"loss": 4.7474,
"step": 670
},
{
"epoch": 0.09055540649315803,
"grad_norm": 5.03125,
"learning_rate": 1.8096514745308312e-05,
"loss": 4.7218,
"step": 675
},
{
"epoch": 0.09122618728199625,
"grad_norm": 4.875,
"learning_rate": 1.8230563002680968e-05,
"loss": 4.7831,
"step": 680
},
{
"epoch": 0.09189696807083444,
"grad_norm": 4.71875,
"learning_rate": 1.836461126005362e-05,
"loss": 4.7727,
"step": 685
},
{
"epoch": 0.09256774885967266,
"grad_norm": 4.65625,
"learning_rate": 1.8498659517426275e-05,
"loss": 4.7595,
"step": 690
},
{
"epoch": 0.09323852964851087,
"grad_norm": 4.78125,
"learning_rate": 1.8632707774798927e-05,
"loss": 4.7422,
"step": 695
},
{
"epoch": 0.09390931043734907,
"grad_norm": 4.71875,
"learning_rate": 1.8766756032171583e-05,
"loss": 4.6978,
"step": 700
},
{
"epoch": 0.09458009122618728,
"grad_norm": 4.5625,
"learning_rate": 1.8900804289544238e-05,
"loss": 4.7197,
"step": 705
},
{
"epoch": 0.0952508720150255,
"grad_norm": 4.65625,
"learning_rate": 1.903485254691689e-05,
"loss": 4.7553,
"step": 710
},
{
"epoch": 0.0959216528038637,
"grad_norm": 4.71875,
"learning_rate": 1.9168900804289546e-05,
"loss": 4.7196,
"step": 715
},
{
"epoch": 0.0965924335927019,
"grad_norm": 4.90625,
"learning_rate": 1.9302949061662198e-05,
"loss": 4.7149,
"step": 720
},
{
"epoch": 0.09726321438154012,
"grad_norm": 4.375,
"learning_rate": 1.9436997319034853e-05,
"loss": 4.7124,
"step": 725
},
{
"epoch": 0.09793399517037832,
"grad_norm": 4.5625,
"learning_rate": 1.957104557640751e-05,
"loss": 4.6549,
"step": 730
},
{
"epoch": 0.09860477595921653,
"grad_norm": 4.09375,
"learning_rate": 1.970509383378016e-05,
"loss": 4.7137,
"step": 735
},
{
"epoch": 0.09927555674805473,
"grad_norm": 4.65625,
"learning_rate": 1.9839142091152816e-05,
"loss": 4.6671,
"step": 740
},
{
"epoch": 0.09994633753689294,
"grad_norm": 4.5625,
"learning_rate": 1.9973190348525472e-05,
"loss": 4.7296,
"step": 745
},
{
"epoch": 0.10061711832573116,
"grad_norm": 4.34375,
"learning_rate": 1.9999982452974893e-05,
"loss": 4.7069,
"step": 750
},
{
"epoch": 0.10128789911456935,
"grad_norm": 3.953125,
"learning_rate": 1.9999911168290933e-05,
"loss": 4.689,
"step": 755
},
{
"epoch": 0.10195867990340757,
"grad_norm": 4.6875,
"learning_rate": 1.999978504964964e-05,
"loss": 4.6645,
"step": 760
},
{
"epoch": 0.10262946069224578,
"grad_norm": 4.59375,
"learning_rate": 1.999960409774257e-05,
"loss": 4.656,
"step": 765
},
{
"epoch": 0.10330024148108398,
"grad_norm": 4.5625,
"learning_rate": 1.9999368313561967e-05,
"loss": 4.6363,
"step": 770
},
{
"epoch": 0.10397102226992219,
"grad_norm": 4.3125,
"learning_rate": 1.9999077698400742e-05,
"loss": 4.6598,
"step": 775
},
{
"epoch": 0.10464180305876039,
"grad_norm": 4.09375,
"learning_rate": 1.9998732253852467e-05,
"loss": 4.6542,
"step": 780
},
{
"epoch": 0.1053125838475986,
"grad_norm": 4.5625,
"learning_rate": 1.999833198181137e-05,
"loss": 4.6505,
"step": 785
},
{
"epoch": 0.10598336463643682,
"grad_norm": 4.3125,
"learning_rate": 1.999787688447231e-05,
"loss": 4.6467,
"step": 790
},
{
"epoch": 0.10665414542527502,
"grad_norm": 4.28125,
"learning_rate": 1.99973669643308e-05,
"loss": 4.6537,
"step": 795
},
{
"epoch": 0.10732492621411323,
"grad_norm": 4.03125,
"learning_rate": 1.9996802224182953e-05,
"loss": 4.6373,
"step": 800
},
{
"epoch": 0.10799570700295144,
"grad_norm": 4.1875,
"learning_rate": 1.9996182667125495e-05,
"loss": 4.6019,
"step": 805
},
{
"epoch": 0.10866648779178964,
"grad_norm": 4.71875,
"learning_rate": 1.9995508296555725e-05,
"loss": 4.6458,
"step": 810
},
{
"epoch": 0.10933726858062785,
"grad_norm": 4.5625,
"learning_rate": 1.9994779116171524e-05,
"loss": 4.6334,
"step": 815
},
{
"epoch": 0.11000804936946605,
"grad_norm": 4.125,
"learning_rate": 1.9993995129971305e-05,
"loss": 4.6204,
"step": 820
},
{
"epoch": 0.11067883015830426,
"grad_norm": 3.96875,
"learning_rate": 1.999315634225402e-05,
"loss": 4.655,
"step": 825
},
{
"epoch": 0.11134961094714248,
"grad_norm": 4.09375,
"learning_rate": 1.999226275761911e-05,
"loss": 4.5946,
"step": 830
},
{
"epoch": 0.11202039173598068,
"grad_norm": 4.1875,
"learning_rate": 1.9991314380966496e-05,
"loss": 4.6585,
"step": 835
},
{
"epoch": 0.11269117252481889,
"grad_norm": 4.125,
"learning_rate": 1.9990311217496554e-05,
"loss": 4.6658,
"step": 840
},
{
"epoch": 0.1133619533136571,
"grad_norm": 3.890625,
"learning_rate": 1.998925327271007e-05,
"loss": 4.5976,
"step": 845
},
{
"epoch": 0.1140327341024953,
"grad_norm": 3.890625,
"learning_rate": 1.998814055240823e-05,
"loss": 4.5957,
"step": 850
},
{
"epoch": 0.11470351489133351,
"grad_norm": 4.4375,
"learning_rate": 1.998697306269257e-05,
"loss": 4.6033,
"step": 855
},
{
"epoch": 0.11537429568017173,
"grad_norm": 4.125,
"learning_rate": 1.998575080996496e-05,
"loss": 4.6202,
"step": 860
},
{
"epoch": 0.11604507646900993,
"grad_norm": 4.40625,
"learning_rate": 1.998447380092755e-05,
"loss": 4.5933,
"step": 865
},
{
"epoch": 0.11671585725784814,
"grad_norm": 4.125,
"learning_rate": 1.9983142042582745e-05,
"loss": 4.5676,
"step": 870
},
{
"epoch": 0.11738663804668634,
"grad_norm": 4.125,
"learning_rate": 1.9981755542233175e-05,
"loss": 4.5953,
"step": 875
},
{
"epoch": 0.11805741883552455,
"grad_norm": 4.15625,
"learning_rate": 1.9980314307481638e-05,
"loss": 4.588,
"step": 880
},
{
"epoch": 0.11872819962436276,
"grad_norm": 4.03125,
"learning_rate": 1.9978818346231058e-05,
"loss": 4.519,
"step": 885
},
{
"epoch": 0.11939898041320096,
"grad_norm": 4.09375,
"learning_rate": 1.997726766668446e-05,
"loss": 4.6255,
"step": 890
},
{
"epoch": 0.12006976120203917,
"grad_norm": 4.3125,
"learning_rate": 1.9975662277344908e-05,
"loss": 4.594,
"step": 895
},
{
"epoch": 0.12074054199087739,
"grad_norm": 4.125,
"learning_rate": 1.9974002187015467e-05,
"loss": 4.5651,
"step": 900
},
{
"epoch": 0.12141132277971559,
"grad_norm": 4.0625,
"learning_rate": 1.9972287404799146e-05,
"loss": 4.5711,
"step": 905
},
{
"epoch": 0.1220821035685538,
"grad_norm": 4.375,
"learning_rate": 1.9970517940098864e-05,
"loss": 4.5614,
"step": 910
},
{
"epoch": 0.122752884357392,
"grad_norm": 4.21875,
"learning_rate": 1.9968693802617377e-05,
"loss": 4.5412,
"step": 915
},
{
"epoch": 0.12342366514623021,
"grad_norm": 3.921875,
"learning_rate": 1.9966815002357243e-05,
"loss": 4.5596,
"step": 920
},
{
"epoch": 0.12409444593506842,
"grad_norm": 3.984375,
"learning_rate": 1.996488154962076e-05,
"loss": 4.5865,
"step": 925
},
{
"epoch": 0.12476522672390662,
"grad_norm": 4.125,
"learning_rate": 1.996289345500991e-05,
"loss": 4.5762,
"step": 930
},
{
"epoch": 0.12543600751274483,
"grad_norm": 3.78125,
"learning_rate": 1.9960850729426296e-05,
"loss": 4.593,
"step": 935
},
{
"epoch": 0.12610678830158303,
"grad_norm": 4.09375,
"learning_rate": 1.995875338407109e-05,
"loss": 4.5774,
"step": 940
},
{
"epoch": 0.12677756909042126,
"grad_norm": 4.0625,
"learning_rate": 1.9956601430444973e-05,
"loss": 4.5131,
"step": 945
},
{
"epoch": 0.12744834987925946,
"grad_norm": 4.125,
"learning_rate": 1.995439488034806e-05,
"loss": 4.5451,
"step": 950
},
{
"epoch": 0.12811913066809766,
"grad_norm": 4.09375,
"learning_rate": 1.9952133745879844e-05,
"loss": 4.5449,
"step": 955
},
{
"epoch": 0.12878991145693588,
"grad_norm": 4.1875,
"learning_rate": 1.9949818039439137e-05,
"loss": 4.5772,
"step": 960
},
{
"epoch": 0.12946069224577408,
"grad_norm": 3.734375,
"learning_rate": 1.994744777372398e-05,
"loss": 4.5225,
"step": 965
},
{
"epoch": 0.13013147303461228,
"grad_norm": 3.921875,
"learning_rate": 1.9945022961731596e-05,
"loss": 4.5347,
"step": 970
},
{
"epoch": 0.1308022538234505,
"grad_norm": 4.1875,
"learning_rate": 1.9942543616758304e-05,
"loss": 4.5648,
"step": 975
},
{
"epoch": 0.1314730346122887,
"grad_norm": 4.03125,
"learning_rate": 1.9940009752399462e-05,
"loss": 4.5185,
"step": 980
},
{
"epoch": 0.1321438154011269,
"grad_norm": 3.78125,
"learning_rate": 1.9937421382549367e-05,
"loss": 4.5739,
"step": 985
},
{
"epoch": 0.1328145961899651,
"grad_norm": 3.71875,
"learning_rate": 1.99347785214012e-05,
"loss": 4.554,
"step": 990
},
{
"epoch": 0.13348537697880333,
"grad_norm": 3.90625,
"learning_rate": 1.9932081183446953e-05,
"loss": 4.5733,
"step": 995
},
{
"epoch": 0.13415615776764153,
"grad_norm": 3.921875,
"learning_rate": 1.992932938347732e-05,
"loss": 4.5142,
"step": 1000
},
{
"epoch": 0.13482693855647973,
"grad_norm": 3.828125,
"learning_rate": 1.9926523136581643e-05,
"loss": 4.5331,
"step": 1005
},
{
"epoch": 0.13549771934531796,
"grad_norm": 4.0,
"learning_rate": 1.992366245814783e-05,
"loss": 4.5364,
"step": 1010
},
{
"epoch": 0.13616850013415616,
"grad_norm": 4.09375,
"learning_rate": 1.992074736386224e-05,
"loss": 4.528,
"step": 1015
},
{
"epoch": 0.13683928092299436,
"grad_norm": 4.25,
"learning_rate": 1.991777786970964e-05,
"loss": 4.5612,
"step": 1020
},
{
"epoch": 0.13751006171183258,
"grad_norm": 3.765625,
"learning_rate": 1.9914753991973084e-05,
"loss": 4.4952,
"step": 1025
},
{
"epoch": 0.13818084250067078,
"grad_norm": 3.96875,
"learning_rate": 1.9911675747233834e-05,
"loss": 4.5122,
"step": 1030
},
{
"epoch": 0.13885162328950898,
"grad_norm": 3.71875,
"learning_rate": 1.9908543152371278e-05,
"loss": 4.5297,
"step": 1035
},
{
"epoch": 0.1395224040783472,
"grad_norm": 3.71875,
"learning_rate": 1.9905356224562824e-05,
"loss": 4.5079,
"step": 1040
},
{
"epoch": 0.1401931848671854,
"grad_norm": 4.03125,
"learning_rate": 1.9902114981283812e-05,
"loss": 4.5109,
"step": 1045
},
{
"epoch": 0.1408639656560236,
"grad_norm": 3.65625,
"learning_rate": 1.9898819440307427e-05,
"loss": 4.5694,
"step": 1050
},
{
"epoch": 0.14153474644486183,
"grad_norm": 4.03125,
"learning_rate": 1.9895469619704576e-05,
"loss": 4.5367,
"step": 1055
},
{
"epoch": 0.14220552723370003,
"grad_norm": 4.09375,
"learning_rate": 1.9892065537843817e-05,
"loss": 4.514,
"step": 1060
},
{
"epoch": 0.14287630802253823,
"grad_norm": 3.640625,
"learning_rate": 1.9888607213391246e-05,
"loss": 4.5282,
"step": 1065
},
{
"epoch": 0.14354708881137646,
"grad_norm": 3.859375,
"learning_rate": 1.988509466531039e-05,
"loss": 4.5126,
"step": 1070
},
{
"epoch": 0.14421786960021465,
"grad_norm": 3.6875,
"learning_rate": 1.988152791286211e-05,
"loss": 4.5547,
"step": 1075
},
{
"epoch": 0.14488865038905285,
"grad_norm": 3.5625,
"learning_rate": 1.9877906975604503e-05,
"loss": 4.5705,
"step": 1080
},
{
"epoch": 0.14555943117789105,
"grad_norm": 4.28125,
"learning_rate": 1.987423187339277e-05,
"loss": 4.5192,
"step": 1085
},
{
"epoch": 0.14623021196672928,
"grad_norm": 3.890625,
"learning_rate": 1.9870502626379127e-05,
"loss": 4.5349,
"step": 1090
},
{
"epoch": 0.14690099275556748,
"grad_norm": 4.0625,
"learning_rate": 1.9866719255012702e-05,
"loss": 4.472,
"step": 1095
},
{
"epoch": 0.14757177354440568,
"grad_norm": 3.78125,
"learning_rate": 1.9862881780039398e-05,
"loss": 4.5542,
"step": 1100
},
{
"epoch": 0.1482425543332439,
"grad_norm": 3.828125,
"learning_rate": 1.9858990222501796e-05,
"loss": 4.5082,
"step": 1105
},
{
"epoch": 0.1489133351220821,
"grad_norm": 3.84375,
"learning_rate": 1.985504460373903e-05,
"loss": 4.5119,
"step": 1110
},
{
"epoch": 0.1495841159109203,
"grad_norm": 3.90625,
"learning_rate": 1.9851044945386687e-05,
"loss": 4.5104,
"step": 1115
},
{
"epoch": 0.15025489669975853,
"grad_norm": 3.484375,
"learning_rate": 1.984699126937667e-05,
"loss": 4.476,
"step": 1120
},
{
"epoch": 0.15092567748859673,
"grad_norm": 3.984375,
"learning_rate": 1.9842883597937084e-05,
"loss": 4.5008,
"step": 1125
},
{
"epoch": 0.15159645827743493,
"grad_norm": 3.71875,
"learning_rate": 1.9838721953592122e-05,
"loss": 4.4806,
"step": 1130
},
{
"epoch": 0.15226723906627315,
"grad_norm": 4.0625,
"learning_rate": 1.983450635916193e-05,
"loss": 4.4887,
"step": 1135
},
{
"epoch": 0.15293801985511135,
"grad_norm": 3.90625,
"learning_rate": 1.983023683776248e-05,
"loss": 4.4838,
"step": 1140
},
{
"epoch": 0.15360880064394955,
"grad_norm": 3.90625,
"learning_rate": 1.9825913412805456e-05,
"loss": 4.519,
"step": 1145
},
{
"epoch": 0.15427958143278778,
"grad_norm": 3.890625,
"learning_rate": 1.9821536107998126e-05,
"loss": 4.5381,
"step": 1150
},
{
"epoch": 0.15495036222162598,
"grad_norm": 3.796875,
"learning_rate": 1.9817104947343188e-05,
"loss": 4.5032,
"step": 1155
},
{
"epoch": 0.15562114301046417,
"grad_norm": 3.875,
"learning_rate": 1.981261995513867e-05,
"loss": 4.4454,
"step": 1160
},
{
"epoch": 0.1562919237993024,
"grad_norm": 3.5625,
"learning_rate": 1.980808115597778e-05,
"loss": 4.4839,
"step": 1165
},
{
"epoch": 0.1569627045881406,
"grad_norm": 4.125,
"learning_rate": 1.9803488574748767e-05,
"loss": 4.4902,
"step": 1170
},
{
"epoch": 0.1576334853769788,
"grad_norm": 3.875,
"learning_rate": 1.9798842236634797e-05,
"loss": 4.486,
"step": 1175
},
{
"epoch": 0.158304266165817,
"grad_norm": 3.921875,
"learning_rate": 1.9794142167113802e-05,
"loss": 4.5046,
"step": 1180
},
{
"epoch": 0.15897504695465522,
"grad_norm": 3.921875,
"learning_rate": 1.9789388391958365e-05,
"loss": 4.506,
"step": 1185
},
{
"epoch": 0.15964582774349342,
"grad_norm": 3.921875,
"learning_rate": 1.9784580937235535e-05,
"loss": 4.4873,
"step": 1190
},
{
"epoch": 0.16031660853233162,
"grad_norm": 3.6875,
"learning_rate": 1.9779719829306725e-05,
"loss": 4.5101,
"step": 1195
},
{
"epoch": 0.16098738932116985,
"grad_norm": 3.53125,
"learning_rate": 1.9774805094827557e-05,
"loss": 4.4654,
"step": 1200
},
{
"epoch": 0.16165817011000805,
"grad_norm": 3.953125,
"learning_rate": 1.9769836760747693e-05,
"loss": 4.4927,
"step": 1205
},
{
"epoch": 0.16232895089884625,
"grad_norm": 3.75,
"learning_rate": 1.976481485431073e-05,
"loss": 4.5078,
"step": 1210
},
{
"epoch": 0.16299973168768447,
"grad_norm": 4.125,
"learning_rate": 1.9759739403054004e-05,
"loss": 4.4274,
"step": 1215
},
{
"epoch": 0.16367051247652267,
"grad_norm": 3.90625,
"learning_rate": 1.9754610434808476e-05,
"loss": 4.4607,
"step": 1220
},
{
"epoch": 0.16434129326536087,
"grad_norm": 3.71875,
"learning_rate": 1.9749427977698555e-05,
"loss": 4.4978,
"step": 1225
},
{
"epoch": 0.1650120740541991,
"grad_norm": 3.859375,
"learning_rate": 1.9744192060141956e-05,
"loss": 4.46,
"step": 1230
},
{
"epoch": 0.1656828548430373,
"grad_norm": 3.9375,
"learning_rate": 1.9738902710849545e-05,
"loss": 4.5099,
"step": 1235
},
{
"epoch": 0.1663536356318755,
"grad_norm": 3.921875,
"learning_rate": 1.9733559958825167e-05,
"loss": 4.505,
"step": 1240
},
{
"epoch": 0.16702441642071372,
"grad_norm": 3.65625,
"learning_rate": 1.972816383336552e-05,
"loss": 4.4777,
"step": 1245
},
{
"epoch": 0.16769519720955192,
"grad_norm": 3.984375,
"learning_rate": 1.9722714364059943e-05,
"loss": 4.4861,
"step": 1250
},
{
"epoch": 0.16836597799839012,
"grad_norm": 3.90625,
"learning_rate": 1.971721158079031e-05,
"loss": 4.4889,
"step": 1255
},
{
"epoch": 0.16903675878722835,
"grad_norm": 3.484375,
"learning_rate": 1.9711655513730822e-05,
"loss": 4.4638,
"step": 1260
},
{
"epoch": 0.16970753957606655,
"grad_norm": 3.625,
"learning_rate": 1.970604619334786e-05,
"loss": 4.4618,
"step": 1265
},
{
"epoch": 0.17037832036490475,
"grad_norm": 3.734375,
"learning_rate": 1.970038365039983e-05,
"loss": 4.4446,
"step": 1270
},
{
"epoch": 0.17104910115374294,
"grad_norm": 3.859375,
"learning_rate": 1.9694667915936966e-05,
"loss": 4.4818,
"step": 1275
},
{
"epoch": 0.17171988194258117,
"grad_norm": 3.75,
"learning_rate": 1.9688899021301182e-05,
"loss": 4.4632,
"step": 1280
},
{
"epoch": 0.17239066273141937,
"grad_norm": 3.8125,
"learning_rate": 1.968307699812589e-05,
"loss": 4.4962,
"step": 1285
},
{
"epoch": 0.17306144352025757,
"grad_norm": 3.703125,
"learning_rate": 1.967720187833583e-05,
"loss": 4.4323,
"step": 1290
},
{
"epoch": 0.1737322243090958,
"grad_norm": 3.734375,
"learning_rate": 1.9671273694146897e-05,
"loss": 4.5151,
"step": 1295
},
{
"epoch": 0.174403005097934,
"grad_norm": 3.796875,
"learning_rate": 1.9665292478065962e-05,
"loss": 4.5083,
"step": 1300
},
{
"epoch": 0.1750737858867722,
"grad_norm": 3.78125,
"learning_rate": 1.9659258262890683e-05,
"loss": 4.4574,
"step": 1305
},
{
"epoch": 0.17574456667561042,
"grad_norm": 3.734375,
"learning_rate": 1.9653171081709353e-05,
"loss": 4.4577,
"step": 1310
},
{
"epoch": 0.17641534746444862,
"grad_norm": 3.765625,
"learning_rate": 1.9647030967900685e-05,
"loss": 4.4341,
"step": 1315
},
{
"epoch": 0.17708612825328682,
"grad_norm": 3.625,
"learning_rate": 1.9640837955133655e-05,
"loss": 4.5131,
"step": 1320
},
{
"epoch": 0.17775690904212504,
"grad_norm": 3.828125,
"learning_rate": 1.9634592077367306e-05,
"loss": 4.4504,
"step": 1325
},
{
"epoch": 0.17842768983096324,
"grad_norm": 4.0,
"learning_rate": 1.9628293368850557e-05,
"loss": 4.4687,
"step": 1330
},
{
"epoch": 0.17909847061980144,
"grad_norm": 3.484375,
"learning_rate": 1.9621941864122028e-05,
"loss": 4.4425,
"step": 1335
},
{
"epoch": 0.17976925140863967,
"grad_norm": 3.703125,
"learning_rate": 1.9615537598009845e-05,
"loss": 4.4812,
"step": 1340
},
{
"epoch": 0.18044003219747787,
"grad_norm": 3.734375,
"learning_rate": 1.9609080605631444e-05,
"loss": 4.4735,
"step": 1345
},
{
"epoch": 0.18111081298631607,
"grad_norm": 3.796875,
"learning_rate": 1.9602570922393383e-05,
"loss": 4.421,
"step": 1350
},
{
"epoch": 0.1817815937751543,
"grad_norm": 3.703125,
"learning_rate": 1.959600858399115e-05,
"loss": 4.4084,
"step": 1355
},
{
"epoch": 0.1824523745639925,
"grad_norm": 3.578125,
"learning_rate": 1.958939362640896e-05,
"loss": 4.435,
"step": 1360
},
{
"epoch": 0.1831231553528307,
"grad_norm": 3.6875,
"learning_rate": 1.958272608591957e-05,
"loss": 4.4884,
"step": 1365
},
{
"epoch": 0.1837939361416689,
"grad_norm": 3.546875,
"learning_rate": 1.957600599908406e-05,
"loss": 4.4484,
"step": 1370
},
{
"epoch": 0.18446471693050712,
"grad_norm": 3.84375,
"learning_rate": 1.9569233402751663e-05,
"loss": 4.4781,
"step": 1375
},
{
"epoch": 0.18513549771934532,
"grad_norm": 4.09375,
"learning_rate": 1.9562408334059528e-05,
"loss": 4.5157,
"step": 1380
},
{
"epoch": 0.18580627850818351,
"grad_norm": 3.75,
"learning_rate": 1.9555530830432543e-05,
"loss": 4.4841,
"step": 1385
},
{
"epoch": 0.18647705929702174,
"grad_norm": 3.90625,
"learning_rate": 1.9548600929583116e-05,
"loss": 4.4756,
"step": 1390
},
{
"epoch": 0.18714784008585994,
"grad_norm": 3.84375,
"learning_rate": 1.954161866951097e-05,
"loss": 4.4691,
"step": 1395
},
{
"epoch": 0.18781862087469814,
"grad_norm": 3.703125,
"learning_rate": 1.9534584088502946e-05,
"loss": 4.4098,
"step": 1400
},
{
"epoch": 0.18848940166353637,
"grad_norm": 3.703125,
"learning_rate": 1.9527497225132776e-05,
"loss": 4.4561,
"step": 1405
},
{
"epoch": 0.18916018245237456,
"grad_norm": 3.65625,
"learning_rate": 1.9520358118260882e-05,
"loss": 4.429,
"step": 1410
},
{
"epoch": 0.18983096324121276,
"grad_norm": 3.765625,
"learning_rate": 1.9513166807034156e-05,
"loss": 4.4711,
"step": 1415
},
{
"epoch": 0.190501744030051,
"grad_norm": 3.6875,
"learning_rate": 1.950592333088576e-05,
"loss": 4.4349,
"step": 1420
},
{
"epoch": 0.1911725248188892,
"grad_norm": 3.9375,
"learning_rate": 1.9498627729534885e-05,
"loss": 4.4481,
"step": 1425
},
{
"epoch": 0.1918433056077274,
"grad_norm": 3.578125,
"learning_rate": 1.9491280042986566e-05,
"loss": 4.507,
"step": 1430
},
{
"epoch": 0.19251408639656561,
"grad_norm": 3.546875,
"learning_rate": 1.9483880311531424e-05,
"loss": 4.4605,
"step": 1435
},
{
"epoch": 0.1931848671854038,
"grad_norm": 4.28125,
"learning_rate": 1.947642857574548e-05,
"loss": 4.4604,
"step": 1440
},
{
"epoch": 0.193855647974242,
"grad_norm": 3.625,
"learning_rate": 1.9468924876489915e-05,
"loss": 4.4499,
"step": 1445
},
{
"epoch": 0.19452642876308024,
"grad_norm": 3.46875,
"learning_rate": 1.9461369254910844e-05,
"loss": 4.449,
"step": 1450
},
{
"epoch": 0.19519720955191844,
"grad_norm": 3.671875,
"learning_rate": 1.94537617524391e-05,
"loss": 4.4513,
"step": 1455
},
{
"epoch": 0.19586799034075664,
"grad_norm": 3.484375,
"learning_rate": 1.944610241079e-05,
"loss": 4.4322,
"step": 1460
},
{
"epoch": 0.19653877112959484,
"grad_norm": 3.828125,
"learning_rate": 1.943839127196311e-05,
"loss": 4.4345,
"step": 1465
},
{
"epoch": 0.19720955191843306,
"grad_norm": 3.8125,
"learning_rate": 1.9430628378242047e-05,
"loss": 4.4492,
"step": 1470
},
{
"epoch": 0.19788033270727126,
"grad_norm": 3.984375,
"learning_rate": 1.9422813772194197e-05,
"loss": 4.4094,
"step": 1475
},
{
"epoch": 0.19855111349610946,
"grad_norm": 3.796875,
"learning_rate": 1.9414947496670523e-05,
"loss": 4.4448,
"step": 1480
},
{
"epoch": 0.1992218942849477,
"grad_norm": 3.625,
"learning_rate": 1.94070295948053e-05,
"loss": 4.4932,
"step": 1485
},
{
"epoch": 0.19989267507378589,
"grad_norm": 3.609375,
"learning_rate": 1.9399060110015918e-05,
"loss": 4.4421,
"step": 1490
},
{
"epoch": 0.20056345586262408,
"grad_norm": 3.875,
"learning_rate": 1.9391039086002594e-05,
"loss": 4.4216,
"step": 1495
},
{
"epoch": 0.2012342366514623,
"grad_norm": 3.65625,
"learning_rate": 1.938296656674817e-05,
"loss": 4.479,
"step": 1500
},
{
"epoch": 0.2019050174403005,
"grad_norm": 3.640625,
"learning_rate": 1.9374842596517855e-05,
"loss": 4.436,
"step": 1505
},
{
"epoch": 0.2025757982291387,
"grad_norm": 3.859375,
"learning_rate": 1.936666721985899e-05,
"loss": 4.4568,
"step": 1510
},
{
"epoch": 0.20324657901797694,
"grad_norm": 3.6875,
"learning_rate": 1.9358440481600803e-05,
"loss": 4.4605,
"step": 1515
},
{
"epoch": 0.20391735980681513,
"grad_norm": 3.515625,
"learning_rate": 1.9350162426854152e-05,
"loss": 4.4594,
"step": 1520
},
{
"epoch": 0.20458814059565333,
"grad_norm": 3.90625,
"learning_rate": 1.9341833101011283e-05,
"loss": 4.4777,
"step": 1525
},
{
"epoch": 0.20525892138449156,
"grad_norm": 3.828125,
"learning_rate": 1.933345254974561e-05,
"loss": 4.4416,
"step": 1530
},
{
"epoch": 0.20592970217332976,
"grad_norm": 3.5625,
"learning_rate": 1.9325020819011403e-05,
"loss": 4.4037,
"step": 1535
},
{
"epoch": 0.20660048296216796,
"grad_norm": 3.703125,
"learning_rate": 1.9316537955043592e-05,
"loss": 4.4825,
"step": 1540
},
{
"epoch": 0.20727126375100619,
"grad_norm": 3.640625,
"learning_rate": 1.93080040043575e-05,
"loss": 4.4204,
"step": 1545
},
{
"epoch": 0.20794204453984438,
"grad_norm": 3.828125,
"learning_rate": 1.929941901374856e-05,
"loss": 4.4384,
"step": 1550
},
{
"epoch": 0.20861282532868258,
"grad_norm": 3.578125,
"learning_rate": 1.92907830302921e-05,
"loss": 4.4358,
"step": 1555
},
{
"epoch": 0.20928360611752078,
"grad_norm": 3.78125,
"learning_rate": 1.9282096101343052e-05,
"loss": 4.4517,
"step": 1560
},
{
"epoch": 0.209954386906359,
"grad_norm": 3.6875,
"learning_rate": 1.9273358274535703e-05,
"loss": 4.4393,
"step": 1565
},
{
"epoch": 0.2106251676951972,
"grad_norm": 3.625,
"learning_rate": 1.926456959778345e-05,
"loss": 4.4431,
"step": 1570
},
{
"epoch": 0.2112959484840354,
"grad_norm": 3.921875,
"learning_rate": 1.9255730119278507e-05,
"loss": 4.4922,
"step": 1575
},
{
"epoch": 0.21196672927287363,
"grad_norm": 3.71875,
"learning_rate": 1.924683988749167e-05,
"loss": 4.4552,
"step": 1580
},
{
"epoch": 0.21263751006171183,
"grad_norm": 3.921875,
"learning_rate": 1.9237898951172025e-05,
"loss": 4.452,
"step": 1585
},
{
"epoch": 0.21330829085055003,
"grad_norm": 3.625,
"learning_rate": 1.92289073593467e-05,
"loss": 4.4399,
"step": 1590
},
{
"epoch": 0.21397907163938826,
"grad_norm": 3.9375,
"learning_rate": 1.921986516132059e-05,
"loss": 4.4321,
"step": 1595
},
{
"epoch": 0.21464985242822646,
"grad_norm": 3.703125,
"learning_rate": 1.921077240667609e-05,
"loss": 4.4525,
"step": 1600
},
{
"epoch": 0.21532063321706466,
"grad_norm": 3.640625,
"learning_rate": 1.9201629145272807e-05,
"loss": 4.4522,
"step": 1605
},
{
"epoch": 0.21599141400590288,
"grad_norm": 3.515625,
"learning_rate": 1.9192435427247313e-05,
"loss": 4.459,
"step": 1610
},
{
"epoch": 0.21666219479474108,
"grad_norm": 3.890625,
"learning_rate": 1.918319130301285e-05,
"loss": 4.4189,
"step": 1615
},
{
"epoch": 0.21733297558357928,
"grad_norm": 3.640625,
"learning_rate": 1.9173896823259052e-05,
"loss": 4.4315,
"step": 1620
},
{
"epoch": 0.2180037563724175,
"grad_norm": 3.734375,
"learning_rate": 1.9164552038951692e-05,
"loss": 4.4628,
"step": 1625
},
{
"epoch": 0.2186745371612557,
"grad_norm": 3.5625,
"learning_rate": 1.9155157001332374e-05,
"loss": 4.4324,
"step": 1630
},
{
"epoch": 0.2193453179500939,
"grad_norm": 3.796875,
"learning_rate": 1.914571176191826e-05,
"loss": 4.4043,
"step": 1635
},
{
"epoch": 0.2200160987389321,
"grad_norm": 4.0625,
"learning_rate": 1.91362163725018e-05,
"loss": 4.441,
"step": 1640
},
{
"epoch": 0.22068687952777033,
"grad_norm": 3.53125,
"learning_rate": 1.912667088515043e-05,
"loss": 4.4331,
"step": 1645
},
{
"epoch": 0.22135766031660853,
"grad_norm": 3.734375,
"learning_rate": 1.9117075352206303e-05,
"loss": 4.452,
"step": 1650
},
{
"epoch": 0.22202844110544673,
"grad_norm": 3.546875,
"learning_rate": 1.9107429826285992e-05,
"loss": 4.4131,
"step": 1655
},
{
"epoch": 0.22269922189428495,
"grad_norm": 3.421875,
"learning_rate": 1.9097734360280197e-05,
"loss": 4.4637,
"step": 1660
},
{
"epoch": 0.22337000268312315,
"grad_norm": 3.734375,
"learning_rate": 1.908798900735347e-05,
"loss": 4.4253,
"step": 1665
},
{
"epoch": 0.22404078347196135,
"grad_norm": 3.65625,
"learning_rate": 1.9078193820943916e-05,
"loss": 4.4178,
"step": 1670
},
{
"epoch": 0.22471156426079958,
"grad_norm": 3.765625,
"learning_rate": 1.906834885476289e-05,
"loss": 4.4321,
"step": 1675
},
{
"epoch": 0.22538234504963778,
"grad_norm": 3.859375,
"learning_rate": 1.9058454162794723e-05,
"loss": 4.4382,
"step": 1680
},
{
"epoch": 0.22605312583847598,
"grad_norm": 3.703125,
"learning_rate": 1.9048509799296403e-05,
"loss": 4.4169,
"step": 1685
},
{
"epoch": 0.2267239066273142,
"grad_norm": 3.875,
"learning_rate": 1.90385158187973e-05,
"loss": 4.4473,
"step": 1690
},
{
"epoch": 0.2273946874161524,
"grad_norm": 3.859375,
"learning_rate": 1.9028472276098845e-05,
"loss": 4.3892,
"step": 1695
},
{
"epoch": 0.2280654682049906,
"grad_norm": 3.734375,
"learning_rate": 1.901837922627425e-05,
"loss": 4.4703,
"step": 1700
},
{
"epoch": 0.22873624899382883,
"grad_norm": 3.765625,
"learning_rate": 1.9008236724668193e-05,
"loss": 4.4481,
"step": 1705
},
{
"epoch": 0.22940702978266703,
"grad_norm": 3.75,
"learning_rate": 1.8998044826896515e-05,
"loss": 4.4369,
"step": 1710
},
{
"epoch": 0.23007781057150523,
"grad_norm": 3.828125,
"learning_rate": 1.8987803588845927e-05,
"loss": 4.4108,
"step": 1715
},
{
"epoch": 0.23074859136034345,
"grad_norm": 3.734375,
"learning_rate": 1.8977513066673685e-05,
"loss": 4.43,
"step": 1720
},
{
"epoch": 0.23141937214918165,
"grad_norm": 3.796875,
"learning_rate": 1.8967173316807293e-05,
"loss": 4.4198,
"step": 1725
},
{
"epoch": 0.23209015293801985,
"grad_norm": 3.875,
"learning_rate": 1.89567843959442e-05,
"loss": 4.4244,
"step": 1730
},
{
"epoch": 0.23276093372685805,
"grad_norm": 3.65625,
"learning_rate": 1.8946346361051474e-05,
"loss": 4.4152,
"step": 1735
},
{
"epoch": 0.23343171451569628,
"grad_norm": 3.59375,
"learning_rate": 1.89358592693655e-05,
"loss": 4.4371,
"step": 1740
},
{
"epoch": 0.23410249530453447,
"grad_norm": 3.75,
"learning_rate": 1.8925323178391664e-05,
"loss": 4.4153,
"step": 1745
},
{
"epoch": 0.23477327609337267,
"grad_norm": 3.65625,
"learning_rate": 1.8914738145904035e-05,
"loss": 4.4303,
"step": 1750
},
{
"epoch": 0.2354440568822109,
"grad_norm": 3.375,
"learning_rate": 1.8904104229945046e-05,
"loss": 4.4473,
"step": 1755
},
{
"epoch": 0.2361148376710491,
"grad_norm": 3.828125,
"learning_rate": 1.889342148882519e-05,
"loss": 4.4223,
"step": 1760
},
{
"epoch": 0.2367856184598873,
"grad_norm": 3.671875,
"learning_rate": 1.8882689981122682e-05,
"loss": 4.4492,
"step": 1765
},
{
"epoch": 0.23745639924872552,
"grad_norm": 3.671875,
"learning_rate": 1.887190976568315e-05,
"loss": 4.4353,
"step": 1770
},
{
"epoch": 0.23812718003756372,
"grad_norm": 3.3125,
"learning_rate": 1.8861080901619305e-05,
"loss": 4.4209,
"step": 1775
},
{
"epoch": 0.23879796082640192,
"grad_norm": 3.546875,
"learning_rate": 1.885020344831062e-05,
"loss": 4.4474,
"step": 1780
},
{
"epoch": 0.23946874161524015,
"grad_norm": 3.640625,
"learning_rate": 1.8839277465403007e-05,
"loss": 4.429,
"step": 1785
},
{
"epoch": 0.24013952240407835,
"grad_norm": 3.53125,
"learning_rate": 1.8828303012808492e-05,
"loss": 4.4374,
"step": 1790
},
{
"epoch": 0.24081030319291655,
"grad_norm": 3.484375,
"learning_rate": 1.8817280150704867e-05,
"loss": 4.3939,
"step": 1795
},
{
"epoch": 0.24148108398175477,
"grad_norm": 3.640625,
"learning_rate": 1.8806208939535387e-05,
"loss": 4.4141,
"step": 1800
},
{
"epoch": 0.24215186477059297,
"grad_norm": 3.890625,
"learning_rate": 1.8795089440008424e-05,
"loss": 4.443,
"step": 1805
},
{
"epoch": 0.24282264555943117,
"grad_norm": 3.71875,
"learning_rate": 1.878392171309714e-05,
"loss": 4.442,
"step": 1810
},
{
"epoch": 0.2434934263482694,
"grad_norm": 3.546875,
"learning_rate": 1.877270582003914e-05,
"loss": 4.4221,
"step": 1815
},
{
"epoch": 0.2441642071371076,
"grad_norm": 3.6875,
"learning_rate": 1.8761441822336152e-05,
"loss": 4.4152,
"step": 1820
},
{
"epoch": 0.2448349879259458,
"grad_norm": 3.71875,
"learning_rate": 1.8750129781753678e-05,
"loss": 4.4229,
"step": 1825
},
{
"epoch": 0.245505768714784,
"grad_norm": 3.625,
"learning_rate": 1.8738769760320673e-05,
"loss": 4.4364,
"step": 1830
},
{
"epoch": 0.24617654950362222,
"grad_norm": 3.53125,
"learning_rate": 1.8727361820329173e-05,
"loss": 4.4457,
"step": 1835
},
{
"epoch": 0.24684733029246042,
"grad_norm": 3.59375,
"learning_rate": 1.871590602433399e-05,
"loss": 4.4157,
"step": 1840
},
{
"epoch": 0.24751811108129862,
"grad_norm": 3.796875,
"learning_rate": 1.8704402435152344e-05,
"loss": 4.4635,
"step": 1845
},
{
"epoch": 0.24818889187013685,
"grad_norm": 3.796875,
"learning_rate": 1.8692851115863524e-05,
"loss": 4.4298,
"step": 1850
},
{
"epoch": 0.24885967265897505,
"grad_norm": 3.609375,
"learning_rate": 1.868125212980855e-05,
"loss": 4.4479,
"step": 1855
},
{
"epoch": 0.24953045344781324,
"grad_norm": 3.6875,
"learning_rate": 1.866960554058982e-05,
"loss": 4.4061,
"step": 1860
},
{
"epoch": 0.25020123423665147,
"grad_norm": 3.453125,
"learning_rate": 1.8657911412070754e-05,
"loss": 4.4029,
"step": 1865
},
{
"epoch": 0.25087201502548967,
"grad_norm": 3.765625,
"learning_rate": 1.8646169808375458e-05,
"loss": 4.4169,
"step": 1870
},
{
"epoch": 0.25154279581432787,
"grad_norm": 3.65625,
"learning_rate": 1.8634380793888366e-05,
"loss": 4.4411,
"step": 1875
},
{
"epoch": 0.25221357660316607,
"grad_norm": 3.890625,
"learning_rate": 1.8622544433253883e-05,
"loss": 4.4456,
"step": 1880
},
{
"epoch": 0.25288435739200427,
"grad_norm": 3.734375,
"learning_rate": 1.861066079137603e-05,
"loss": 4.4251,
"step": 1885
},
{
"epoch": 0.2535551381808425,
"grad_norm": 3.875,
"learning_rate": 1.8598729933418102e-05,
"loss": 4.4506,
"step": 1890
},
{
"epoch": 0.2542259189696807,
"grad_norm": 3.71875,
"learning_rate": 1.8586751924802297e-05,
"loss": 4.4298,
"step": 1895
},
{
"epoch": 0.2548966997585189,
"grad_norm": 3.9375,
"learning_rate": 1.8574726831209357e-05,
"loss": 4.4281,
"step": 1900
},
{
"epoch": 0.2555674805473571,
"grad_norm": 3.859375,
"learning_rate": 1.856265471857821e-05,
"loss": 4.4271,
"step": 1905
},
{
"epoch": 0.2562382613361953,
"grad_norm": 3.765625,
"learning_rate": 1.8550535653105625e-05,
"loss": 4.4123,
"step": 1910
},
{
"epoch": 0.2569090421250335,
"grad_norm": 4.0,
"learning_rate": 1.853836970124582e-05,
"loss": 4.4376,
"step": 1915
},
{
"epoch": 0.25757982291387177,
"grad_norm": 3.8125,
"learning_rate": 1.852615692971011e-05,
"loss": 4.416,
"step": 1920
},
{
"epoch": 0.25825060370270997,
"grad_norm": 3.796875,
"learning_rate": 1.8513897405466558e-05,
"loss": 4.4052,
"step": 1925
},
{
"epoch": 0.25892138449154817,
"grad_norm": 3.734375,
"learning_rate": 1.8501591195739585e-05,
"loss": 4.3873,
"step": 1930
},
{
"epoch": 0.25959216528038637,
"grad_norm": 3.609375,
"learning_rate": 1.8489238368009598e-05,
"loss": 4.4198,
"step": 1935
},
{
"epoch": 0.26026294606922457,
"grad_norm": 3.765625,
"learning_rate": 1.847683899001265e-05,
"loss": 4.4201,
"step": 1940
},
{
"epoch": 0.26093372685806276,
"grad_norm": 3.609375,
"learning_rate": 1.846439312974004e-05,
"loss": 4.4493,
"step": 1945
},
{
"epoch": 0.261604507646901,
"grad_norm": 3.734375,
"learning_rate": 1.845190085543795e-05,
"loss": 4.4111,
"step": 1950
},
{
"epoch": 0.2622752884357392,
"grad_norm": 3.90625,
"learning_rate": 1.8439362235607074e-05,
"loss": 4.3986,
"step": 1955
},
{
"epoch": 0.2629460692245774,
"grad_norm": 3.65625,
"learning_rate": 1.842677733900223e-05,
"loss": 4.446,
"step": 1960
},
{
"epoch": 0.2636168500134156,
"grad_norm": 3.6875,
"learning_rate": 1.8414146234632012e-05,
"loss": 4.3502,
"step": 1965
},
{
"epoch": 0.2642876308022538,
"grad_norm": 3.65625,
"learning_rate": 1.8401468991758368e-05,
"loss": 4.4034,
"step": 1970
},
{
"epoch": 0.264958411591092,
"grad_norm": 3.859375,
"learning_rate": 1.838874567989626e-05,
"loss": 4.4007,
"step": 1975
},
{
"epoch": 0.2656291923799302,
"grad_norm": 3.671875,
"learning_rate": 1.8375976368813262e-05,
"loss": 4.3984,
"step": 1980
},
{
"epoch": 0.26629997316876847,
"grad_norm": 3.8125,
"learning_rate": 1.836316112852918e-05,
"loss": 4.3978,
"step": 1985
},
{
"epoch": 0.26697075395760667,
"grad_norm": 3.71875,
"learning_rate": 1.8350300029315674e-05,
"loss": 4.4301,
"step": 1990
},
{
"epoch": 0.26764153474644486,
"grad_norm": 3.59375,
"learning_rate": 1.8337393141695865e-05,
"loss": 4.3932,
"step": 1995
},
{
"epoch": 0.26831231553528306,
"grad_norm": 3.6875,
"learning_rate": 1.8324440536443956e-05,
"loss": 4.4224,
"step": 2000
},
{
"epoch": 0.26898309632412126,
"grad_norm": 3.6875,
"learning_rate": 1.8311442284584836e-05,
"loss": 4.3716,
"step": 2005
},
{
"epoch": 0.26965387711295946,
"grad_norm": 3.671875,
"learning_rate": 1.82983984573937e-05,
"loss": 4.4076,
"step": 2010
},
{
"epoch": 0.2703246579017977,
"grad_norm": 3.578125,
"learning_rate": 1.828530912639565e-05,
"loss": 4.3844,
"step": 2015
},
{
"epoch": 0.2709954386906359,
"grad_norm": 3.609375,
"learning_rate": 1.82721743633653e-05,
"loss": 4.3935,
"step": 2020
},
{
"epoch": 0.2716662194794741,
"grad_norm": 3.578125,
"learning_rate": 1.82589942403264e-05,
"loss": 4.4344,
"step": 2025
},
{
"epoch": 0.2723370002683123,
"grad_norm": 3.890625,
"learning_rate": 1.8245768829551417e-05,
"loss": 4.4087,
"step": 2030
},
{
"epoch": 0.2730077810571505,
"grad_norm": 3.90625,
"learning_rate": 1.8232498203561163e-05,
"loss": 4.4298,
"step": 2035
},
{
"epoch": 0.2736785618459887,
"grad_norm": 3.953125,
"learning_rate": 1.8219182435124366e-05,
"loss": 4.418,
"step": 2040
},
{
"epoch": 0.27434934263482696,
"grad_norm": 3.71875,
"learning_rate": 1.820582159725732e-05,
"loss": 4.4244,
"step": 2045
},
{
"epoch": 0.27502012342366516,
"grad_norm": 3.796875,
"learning_rate": 1.8192415763223418e-05,
"loss": 4.3726,
"step": 2050
},
{
"epoch": 0.27569090421250336,
"grad_norm": 3.5,
"learning_rate": 1.8178965006532824e-05,
"loss": 4.3844,
"step": 2055
},
{
"epoch": 0.27636168500134156,
"grad_norm": 3.484375,
"learning_rate": 1.8165469400942e-05,
"loss": 4.421,
"step": 2060
},
{
"epoch": 0.27703246579017976,
"grad_norm": 3.765625,
"learning_rate": 1.815192902045336e-05,
"loss": 4.448,
"step": 2065
},
{
"epoch": 0.27770324657901796,
"grad_norm": 3.75,
"learning_rate": 1.8138343939314825e-05,
"loss": 4.4342,
"step": 2070
},
{
"epoch": 0.27837402736785616,
"grad_norm": 3.8125,
"learning_rate": 1.8124714232019432e-05,
"loss": 4.4152,
"step": 2075
},
{
"epoch": 0.2790448081566944,
"grad_norm": 3.734375,
"learning_rate": 1.8111039973304922e-05,
"loss": 4.4415,
"step": 2080
},
{
"epoch": 0.2797155889455326,
"grad_norm": 3.890625,
"learning_rate": 1.809732123815334e-05,
"loss": 4.4436,
"step": 2085
},
{
"epoch": 0.2803863697343708,
"grad_norm": 3.640625,
"learning_rate": 1.8083558101790598e-05,
"loss": 4.4152,
"step": 2090
},
{
"epoch": 0.281057150523209,
"grad_norm": 3.75,
"learning_rate": 1.8069750639686096e-05,
"loss": 4.4004,
"step": 2095
},
{
"epoch": 0.2817279313120472,
"grad_norm": 3.765625,
"learning_rate": 1.805589892755228e-05,
"loss": 4.421,
"step": 2100
},
{
"epoch": 0.2823987121008854,
"grad_norm": 3.875,
"learning_rate": 1.8042003041344257e-05,
"loss": 4.4347,
"step": 2105
},
{
"epoch": 0.28306949288972366,
"grad_norm": 4.1875,
"learning_rate": 1.8028063057259332e-05,
"loss": 4.4023,
"step": 2110
},
{
"epoch": 0.28374027367856186,
"grad_norm": 3.3125,
"learning_rate": 1.8014079051736637e-05,
"loss": 4.4294,
"step": 2115
},
{
"epoch": 0.28441105446740006,
"grad_norm": 4.0625,
"learning_rate": 1.800005110145669e-05,
"loss": 4.4377,
"step": 2120
},
{
"epoch": 0.28508183525623826,
"grad_norm": 3.96875,
"learning_rate": 1.7985979283340987e-05,
"loss": 4.4204,
"step": 2125
},
{
"epoch": 0.28575261604507646,
"grad_norm": 4.0,
"learning_rate": 1.797186367455155e-05,
"loss": 4.4229,
"step": 2130
},
{
"epoch": 0.28642339683391466,
"grad_norm": 3.703125,
"learning_rate": 1.7957704352490535e-05,
"loss": 4.3768,
"step": 2135
},
{
"epoch": 0.2870941776227529,
"grad_norm": 3.6875,
"learning_rate": 1.79435013947998e-05,
"loss": 4.3896,
"step": 2140
},
{
"epoch": 0.2877649584115911,
"grad_norm": 3.734375,
"learning_rate": 1.792925487936048e-05,
"loss": 4.4076,
"step": 2145
},
{
"epoch": 0.2884357392004293,
"grad_norm": 3.8125,
"learning_rate": 1.7914964884292543e-05,
"loss": 4.4216,
"step": 2150
},
{
"epoch": 0.2891065199892675,
"grad_norm": 3.890625,
"learning_rate": 1.7900631487954387e-05,
"loss": 4.4526,
"step": 2155
},
{
"epoch": 0.2897773007781057,
"grad_norm": 3.65625,
"learning_rate": 1.7886254768942394e-05,
"loss": 4.3844,
"step": 2160
},
{
"epoch": 0.2904480815669439,
"grad_norm": 3.875,
"learning_rate": 1.7871834806090502e-05,
"loss": 4.4265,
"step": 2165
},
{
"epoch": 0.2911188623557821,
"grad_norm": 3.671875,
"learning_rate": 1.7857371678469783e-05,
"loss": 4.4415,
"step": 2170
},
{
"epoch": 0.29178964314462036,
"grad_norm": 3.96875,
"learning_rate": 1.7842865465387985e-05,
"loss": 4.386,
"step": 2175
},
{
"epoch": 0.29246042393345856,
"grad_norm": 3.484375,
"learning_rate": 1.7828316246389125e-05,
"loss": 4.3972,
"step": 2180
},
{
"epoch": 0.29313120472229676,
"grad_norm": 3.484375,
"learning_rate": 1.781372410125304e-05,
"loss": 4.4124,
"step": 2185
},
{
"epoch": 0.29380198551113496,
"grad_norm": 3.8125,
"learning_rate": 1.779908910999494e-05,
"loss": 4.4112,
"step": 2190
},
{
"epoch": 0.29447276629997315,
"grad_norm": 3.765625,
"learning_rate": 1.7784411352864993e-05,
"loss": 4.4725,
"step": 2195
},
{
"epoch": 0.29514354708881135,
"grad_norm": 3.625,
"learning_rate": 1.7769690910347867e-05,
"loss": 4.3941,
"step": 2200
},
{
"epoch": 0.2958143278776496,
"grad_norm": 3.53125,
"learning_rate": 1.7754927863162283e-05,
"loss": 4.4291,
"step": 2205
},
{
"epoch": 0.2964851086664878,
"grad_norm": 3.8125,
"learning_rate": 1.7740122292260594e-05,
"loss": 4.4122,
"step": 2210
},
{
"epoch": 0.297155889455326,
"grad_norm": 3.71875,
"learning_rate": 1.7725274278828325e-05,
"loss": 4.4366,
"step": 2215
},
{
"epoch": 0.2978266702441642,
"grad_norm": 3.734375,
"learning_rate": 1.7710383904283737e-05,
"loss": 4.432,
"step": 2220
},
{
"epoch": 0.2984974510330024,
"grad_norm": 3.890625,
"learning_rate": 1.769545125027736e-05,
"loss": 4.4228,
"step": 2225
},
{
"epoch": 0.2991682318218406,
"grad_norm": 3.71875,
"learning_rate": 1.7680476398691587e-05,
"loss": 4.4053,
"step": 2230
},
{
"epoch": 0.29983901261067886,
"grad_norm": 3.703125,
"learning_rate": 1.7665459431640174e-05,
"loss": 4.3996,
"step": 2235
},
{
"epoch": 0.30050979339951706,
"grad_norm": 3.453125,
"learning_rate": 1.7650400431467832e-05,
"loss": 4.3972,
"step": 2240
},
{
"epoch": 0.30118057418835525,
"grad_norm": 3.5,
"learning_rate": 1.763529948074975e-05,
"loss": 4.3789,
"step": 2245
},
{
"epoch": 0.30185135497719345,
"grad_norm": 3.71875,
"learning_rate": 1.762015666229115e-05,
"loss": 4.4187,
"step": 2250
},
{
"epoch": 0.30252213576603165,
"grad_norm": 3.609375,
"learning_rate": 1.7604972059126835e-05,
"loss": 4.4495,
"step": 2255
},
{
"epoch": 0.30319291655486985,
"grad_norm": 3.703125,
"learning_rate": 1.7589745754520736e-05,
"loss": 4.4122,
"step": 2260
},
{
"epoch": 0.30386369734370805,
"grad_norm": 3.59375,
"learning_rate": 1.757447783196545e-05,
"loss": 4.4088,
"step": 2265
},
{
"epoch": 0.3045344781325463,
"grad_norm": 3.5,
"learning_rate": 1.7559168375181777e-05,
"loss": 4.3646,
"step": 2270
},
{
"epoch": 0.3052052589213845,
"grad_norm": 3.796875,
"learning_rate": 1.7543817468118277e-05,
"loss": 4.4268,
"step": 2275
},
{
"epoch": 0.3058760397102227,
"grad_norm": 3.609375,
"learning_rate": 1.7528425194950794e-05,
"loss": 4.3959,
"step": 2280
},
{
"epoch": 0.3065468204990609,
"grad_norm": 3.96875,
"learning_rate": 1.7512991640082004e-05,
"loss": 4.4487,
"step": 2285
},
{
"epoch": 0.3072176012878991,
"grad_norm": 3.609375,
"learning_rate": 1.7497516888140947e-05,
"loss": 4.4144,
"step": 2290
},
{
"epoch": 0.3078883820767373,
"grad_norm": 3.78125,
"learning_rate": 1.7482001023982565e-05,
"loss": 4.3803,
"step": 2295
},
{
"epoch": 0.30855916286557555,
"grad_norm": 3.515625,
"learning_rate": 1.746644413268724e-05,
"loss": 4.4051,
"step": 2300
},
{
"epoch": 0.30922994365441375,
"grad_norm": 3.796875,
"learning_rate": 1.7450846299560315e-05,
"loss": 4.3861,
"step": 2305
},
{
"epoch": 0.30990072444325195,
"grad_norm": 3.6875,
"learning_rate": 1.7435207610131647e-05,
"loss": 4.3963,
"step": 2310
},
{
"epoch": 0.31057150523209015,
"grad_norm": 3.515625,
"learning_rate": 1.7419528150155118e-05,
"loss": 4.3722,
"step": 2315
},
{
"epoch": 0.31124228602092835,
"grad_norm": 3.5625,
"learning_rate": 1.7403808005608178e-05,
"loss": 4.3887,
"step": 2320
},
{
"epoch": 0.31191306680976655,
"grad_norm": 4.09375,
"learning_rate": 1.7388047262691363e-05,
"loss": 4.3945,
"step": 2325
},
{
"epoch": 0.3125838475986048,
"grad_norm": 3.828125,
"learning_rate": 1.7372246007827837e-05,
"loss": 4.3913,
"step": 2330
},
{
"epoch": 0.313254628387443,
"grad_norm": 3.515625,
"learning_rate": 1.735640432766289e-05,
"loss": 4.3554,
"step": 2335
},
{
"epoch": 0.3139254091762812,
"grad_norm": 3.953125,
"learning_rate": 1.7340522309063513e-05,
"loss": 4.3625,
"step": 2340
},
{
"epoch": 0.3145961899651194,
"grad_norm": 3.59375,
"learning_rate": 1.7324600039117862e-05,
"loss": 4.4013,
"step": 2345
},
{
"epoch": 0.3152669707539576,
"grad_norm": 3.796875,
"learning_rate": 1.7308637605134822e-05,
"loss": 4.4282,
"step": 2350
},
{
"epoch": 0.3159377515427958,
"grad_norm": 3.890625,
"learning_rate": 1.729263509464351e-05,
"loss": 4.3801,
"step": 2355
},
{
"epoch": 0.316608532331634,
"grad_norm": 3.8125,
"learning_rate": 1.727659259539281e-05,
"loss": 4.3768,
"step": 2360
},
{
"epoch": 0.31727931312047225,
"grad_norm": 3.734375,
"learning_rate": 1.7260510195350866e-05,
"loss": 4.3706,
"step": 2365
},
{
"epoch": 0.31795009390931045,
"grad_norm": 3.546875,
"learning_rate": 1.7244387982704635e-05,
"loss": 4.4014,
"step": 2370
},
{
"epoch": 0.31862087469814865,
"grad_norm": 3.734375,
"learning_rate": 1.7228226045859363e-05,
"loss": 4.3957,
"step": 2375
},
{
"epoch": 0.31929165548698685,
"grad_norm": 3.9375,
"learning_rate": 1.7212024473438145e-05,
"loss": 4.3729,
"step": 2380
},
{
"epoch": 0.31996243627582505,
"grad_norm": 3.6875,
"learning_rate": 1.71957833542814e-05,
"loss": 4.4025,
"step": 2385
},
{
"epoch": 0.32063321706466325,
"grad_norm": 3.90625,
"learning_rate": 1.7179502777446395e-05,
"loss": 4.4161,
"step": 2390
},
{
"epoch": 0.3213039978535015,
"grad_norm": 3.921875,
"learning_rate": 1.7163182832206777e-05,
"loss": 4.402,
"step": 2395
},
{
"epoch": 0.3219747786423397,
"grad_norm": 3.78125,
"learning_rate": 1.7146823608052054e-05,
"loss": 4.3798,
"step": 2400
},
{
"epoch": 0.3226455594311779,
"grad_norm": 3.546875,
"learning_rate": 1.7130425194687124e-05,
"loss": 4.3807,
"step": 2405
},
{
"epoch": 0.3233163402200161,
"grad_norm": 3.65625,
"learning_rate": 1.711398768203178e-05,
"loss": 4.439,
"step": 2410
},
{
"epoch": 0.3239871210088543,
"grad_norm": 3.5,
"learning_rate": 1.7097511160220202e-05,
"loss": 4.4167,
"step": 2415
},
{
"epoch": 0.3246579017976925,
"grad_norm": 3.703125,
"learning_rate": 1.708099571960049e-05,
"loss": 4.3927,
"step": 2420
},
{
"epoch": 0.32532868258653075,
"grad_norm": 3.84375,
"learning_rate": 1.706444145073414e-05,
"loss": 4.4066,
"step": 2425
},
{
"epoch": 0.32599946337536895,
"grad_norm": 3.890625,
"learning_rate": 1.7047848444395566e-05,
"loss": 4.4076,
"step": 2430
},
{
"epoch": 0.32667024416420715,
"grad_norm": 3.796875,
"learning_rate": 1.7031216791571596e-05,
"loss": 4.4141,
"step": 2435
},
{
"epoch": 0.32734102495304535,
"grad_norm": 3.75,
"learning_rate": 1.7014546583460974e-05,
"loss": 4.4057,
"step": 2440
},
{
"epoch": 0.32801180574188354,
"grad_norm": 4.21875,
"learning_rate": 1.6997837911473865e-05,
"loss": 4.3548,
"step": 2445
},
{
"epoch": 0.32868258653072174,
"grad_norm": 3.71875,
"learning_rate": 1.6981090867231337e-05,
"loss": 4.4405,
"step": 2450
},
{
"epoch": 0.32935336731955994,
"grad_norm": 3.625,
"learning_rate": 1.6964305542564884e-05,
"loss": 4.4124,
"step": 2455
},
{
"epoch": 0.3300241481083982,
"grad_norm": 3.765625,
"learning_rate": 1.694748202951589e-05,
"loss": 4.3738,
"step": 2460
},
{
"epoch": 0.3306949288972364,
"grad_norm": 3.6875,
"learning_rate": 1.6930620420335165e-05,
"loss": 4.4134,
"step": 2465
},
{
"epoch": 0.3313657096860746,
"grad_norm": 3.625,
"learning_rate": 1.6913720807482402e-05,
"loss": 4.3845,
"step": 2470
},
{
"epoch": 0.3320364904749128,
"grad_norm": 3.8125,
"learning_rate": 1.689678328362569e-05,
"loss": 4.4129,
"step": 2475
},
{
"epoch": 0.332707271263751,
"grad_norm": 3.796875,
"learning_rate": 1.6879807941641e-05,
"loss": 4.4102,
"step": 2480
},
{
"epoch": 0.3333780520525892,
"grad_norm": 3.765625,
"learning_rate": 1.6862794874611673e-05,
"loss": 4.3801,
"step": 2485
},
{
"epoch": 0.33404883284142745,
"grad_norm": 3.703125,
"learning_rate": 1.6845744175827927e-05,
"loss": 4.4383,
"step": 2490
},
{
"epoch": 0.33471961363026564,
"grad_norm": 3.875,
"learning_rate": 1.6828655938786308e-05,
"loss": 4.3669,
"step": 2495
},
{
"epoch": 0.33539039441910384,
"grad_norm": 4.09375,
"learning_rate": 1.6811530257189224e-05,
"loss": 4.4448,
"step": 2500
},
{
"epoch": 0.33606117520794204,
"grad_norm": 3.625,
"learning_rate": 1.6794367224944395e-05,
"loss": 4.3914,
"step": 2505
},
{
"epoch": 0.33673195599678024,
"grad_norm": 3.609375,
"learning_rate": 1.6777166936164354e-05,
"loss": 4.3708,
"step": 2510
},
{
"epoch": 0.33740273678561844,
"grad_norm": 3.578125,
"learning_rate": 1.6759929485165935e-05,
"loss": 4.4079,
"step": 2515
},
{
"epoch": 0.3380735175744567,
"grad_norm": 3.65625,
"learning_rate": 1.6742654966469735e-05,
"loss": 4.3914,
"step": 2520
},
{
"epoch": 0.3387442983632949,
"grad_norm": 3.890625,
"learning_rate": 1.672534347479962e-05,
"loss": 4.3912,
"step": 2525
},
{
"epoch": 0.3394150791521331,
"grad_norm": 4.0625,
"learning_rate": 1.6707995105082196e-05,
"loss": 4.4489,
"step": 2530
},
{
"epoch": 0.3400858599409713,
"grad_norm": 3.6875,
"learning_rate": 1.6690609952446283e-05,
"loss": 4.3998,
"step": 2535
},
{
"epoch": 0.3407566407298095,
"grad_norm": 3.9375,
"learning_rate": 1.6673188112222394e-05,
"loss": 4.4053,
"step": 2540
},
{
"epoch": 0.3414274215186477,
"grad_norm": 3.25,
"learning_rate": 1.665572967994223e-05,
"loss": 4.3508,
"step": 2545
},
{
"epoch": 0.3420982023074859,
"grad_norm": 4.0,
"learning_rate": 1.6638234751338135e-05,
"loss": 4.4214,
"step": 2550
},
{
"epoch": 0.34276898309632414,
"grad_norm": 3.96875,
"learning_rate": 1.662070342234257e-05,
"loss": 4.3708,
"step": 2555
},
{
"epoch": 0.34343976388516234,
"grad_norm": 3.9375,
"learning_rate": 1.660313578908761e-05,
"loss": 4.3787,
"step": 2560
},
{
"epoch": 0.34411054467400054,
"grad_norm": 3.46875,
"learning_rate": 1.6585531947904394e-05,
"loss": 4.411,
"step": 2565
},
{
"epoch": 0.34478132546283874,
"grad_norm": 3.78125,
"learning_rate": 1.6567891995322605e-05,
"loss": 4.3579,
"step": 2570
},
{
"epoch": 0.34545210625167694,
"grad_norm": 3.625,
"learning_rate": 1.6550216028069947e-05,
"loss": 4.3835,
"step": 2575
},
{
"epoch": 0.34612288704051514,
"grad_norm": 3.53125,
"learning_rate": 1.6532504143071602e-05,
"loss": 4.3862,
"step": 2580
},
{
"epoch": 0.3467936678293534,
"grad_norm": 3.625,
"learning_rate": 1.6514756437449708e-05,
"loss": 4.4191,
"step": 2585
},
{
"epoch": 0.3474644486181916,
"grad_norm": 3.546875,
"learning_rate": 1.649697300852282e-05,
"loss": 4.384,
"step": 2590
},
{
"epoch": 0.3481352294070298,
"grad_norm": 3.53125,
"learning_rate": 1.647915395380539e-05,
"loss": 4.4173,
"step": 2595
},
{
"epoch": 0.348806010195868,
"grad_norm": 3.796875,
"learning_rate": 1.646129937100721e-05,
"loss": 4.4226,
"step": 2600
},
{
"epoch": 0.3494767909847062,
"grad_norm": 3.96875,
"learning_rate": 1.644340935803289e-05,
"loss": 4.4196,
"step": 2605
},
{
"epoch": 0.3501475717735444,
"grad_norm": 3.453125,
"learning_rate": 1.6425484012981322e-05,
"loss": 4.434,
"step": 2610
},
{
"epoch": 0.35081835256238264,
"grad_norm": 3.59375,
"learning_rate": 1.6407523434145146e-05,
"loss": 4.4156,
"step": 2615
},
{
"epoch": 0.35148913335122084,
"grad_norm": 3.765625,
"learning_rate": 1.6389527720010196e-05,
"loss": 4.3823,
"step": 2620
},
{
"epoch": 0.35215991414005904,
"grad_norm": 3.578125,
"learning_rate": 1.637149696925497e-05,
"loss": 4.3591,
"step": 2625
},
{
"epoch": 0.35283069492889724,
"grad_norm": 3.78125,
"learning_rate": 1.6353431280750083e-05,
"loss": 4.3899,
"step": 2630
},
{
"epoch": 0.35350147571773544,
"grad_norm": 3.921875,
"learning_rate": 1.633533075355774e-05,
"loss": 4.4156,
"step": 2635
},
{
"epoch": 0.35417225650657364,
"grad_norm": 3.578125,
"learning_rate": 1.631719548693118e-05,
"loss": 4.4327,
"step": 2640
},
{
"epoch": 0.35484303729541183,
"grad_norm": 3.78125,
"learning_rate": 1.6299025580314123e-05,
"loss": 4.3791,
"step": 2645
},
{
"epoch": 0.3555138180842501,
"grad_norm": 3.875,
"learning_rate": 1.6280821133340244e-05,
"loss": 4.4001,
"step": 2650
},
{
"epoch": 0.3561845988730883,
"grad_norm": 3.78125,
"learning_rate": 1.6262582245832617e-05,
"loss": 4.393,
"step": 2655
},
{
"epoch": 0.3568553796619265,
"grad_norm": 3.78125,
"learning_rate": 1.624430901780317e-05,
"loss": 4.3926,
"step": 2660
},
{
"epoch": 0.3575261604507647,
"grad_norm": 3.859375,
"learning_rate": 1.622600154945213e-05,
"loss": 4.3708,
"step": 2665
},
{
"epoch": 0.3581969412396029,
"grad_norm": 3.5625,
"learning_rate": 1.6207659941167485e-05,
"loss": 4.4015,
"step": 2670
},
{
"epoch": 0.3588677220284411,
"grad_norm": 3.71875,
"learning_rate": 1.618928429352443e-05,
"loss": 4.4183,
"step": 2675
},
{
"epoch": 0.35953850281727934,
"grad_norm": 4.09375,
"learning_rate": 1.6170874707284796e-05,
"loss": 4.4158,
"step": 2680
},
{
"epoch": 0.36020928360611754,
"grad_norm": 3.71875,
"learning_rate": 1.6152431283396536e-05,
"loss": 4.3971,
"step": 2685
},
{
"epoch": 0.36088006439495574,
"grad_norm": 3.703125,
"learning_rate": 1.613395412299314e-05,
"loss": 4.4236,
"step": 2690
},
{
"epoch": 0.36155084518379393,
"grad_norm": 3.875,
"learning_rate": 1.6115443327393087e-05,
"loss": 4.42,
"step": 2695
},
{
"epoch": 0.36222162597263213,
"grad_norm": 3.703125,
"learning_rate": 1.6096898998099298e-05,
"loss": 4.3931,
"step": 2700
},
{
"epoch": 0.36289240676147033,
"grad_norm": 3.96875,
"learning_rate": 1.607832123679857e-05,
"loss": 4.3664,
"step": 2705
},
{
"epoch": 0.3635631875503086,
"grad_norm": 3.6875,
"learning_rate": 1.605971014536103e-05,
"loss": 4.4262,
"step": 2710
},
{
"epoch": 0.3642339683391468,
"grad_norm": 3.71875,
"learning_rate": 1.604106582583956e-05,
"loss": 4.3482,
"step": 2715
},
{
"epoch": 0.364904749127985,
"grad_norm": 3.953125,
"learning_rate": 1.6022388380469248e-05,
"loss": 4.3982,
"step": 2720
},
{
"epoch": 0.3655755299168232,
"grad_norm": 3.625,
"learning_rate": 1.6003677911666834e-05,
"loss": 4.4264,
"step": 2725
},
{
"epoch": 0.3662463107056614,
"grad_norm": 3.71875,
"learning_rate": 1.598493452203013e-05,
"loss": 4.3806,
"step": 2730
},
{
"epoch": 0.3669170914944996,
"grad_norm": 3.84375,
"learning_rate": 1.5966158314337472e-05,
"loss": 4.3693,
"step": 2735
},
{
"epoch": 0.3675878722833378,
"grad_norm": 3.671875,
"learning_rate": 1.5947349391547153e-05,
"loss": 4.3819,
"step": 2740
},
{
"epoch": 0.36825865307217603,
"grad_norm": 3.75,
"learning_rate": 1.5928507856796856e-05,
"loss": 4.4167,
"step": 2745
},
{
"epoch": 0.36892943386101423,
"grad_norm": 3.65625,
"learning_rate": 1.5909633813403094e-05,
"loss": 4.3846,
"step": 2750
},
{
"epoch": 0.36960021464985243,
"grad_norm": 3.75,
"learning_rate": 1.5890727364860632e-05,
"loss": 4.3983,
"step": 2755
},
{
"epoch": 0.37027099543869063,
"grad_norm": 3.5,
"learning_rate": 1.5871788614841924e-05,
"loss": 4.3977,
"step": 2760
},
{
"epoch": 0.37094177622752883,
"grad_norm": 3.921875,
"learning_rate": 1.585281766719656e-05,
"loss": 4.3959,
"step": 2765
},
{
"epoch": 0.37161255701636703,
"grad_norm": 3.578125,
"learning_rate": 1.583381462595068e-05,
"loss": 4.3918,
"step": 2770
},
{
"epoch": 0.3722833378052053,
"grad_norm": 3.890625,
"learning_rate": 1.5814779595306403e-05,
"loss": 4.3682,
"step": 2775
},
{
"epoch": 0.3729541185940435,
"grad_norm": 3.75,
"learning_rate": 1.5795712679641257e-05,
"loss": 4.4032,
"step": 2780
},
{
"epoch": 0.3736248993828817,
"grad_norm": 3.75,
"learning_rate": 1.577661398350762e-05,
"loss": 4.4053,
"step": 2785
},
{
"epoch": 0.3742956801717199,
"grad_norm": 3.890625,
"learning_rate": 1.575748361163213e-05,
"loss": 4.4276,
"step": 2790
},
{
"epoch": 0.3749664609605581,
"grad_norm": 3.609375,
"learning_rate": 1.5738321668915117e-05,
"loss": 4.3954,
"step": 2795
},
{
"epoch": 0.3756372417493963,
"grad_norm": 3.75,
"learning_rate": 1.571912826043003e-05,
"loss": 4.4108,
"step": 2800
},
{
"epoch": 0.37630802253823453,
"grad_norm": 3.859375,
"learning_rate": 1.569990349142286e-05,
"loss": 4.3599,
"step": 2805
},
{
"epoch": 0.37697880332707273,
"grad_norm": 3.953125,
"learning_rate": 1.568064746731156e-05,
"loss": 4.3933,
"step": 2810
},
{
"epoch": 0.37764958411591093,
"grad_norm": 3.546875,
"learning_rate": 1.5661360293685465e-05,
"loss": 4.4253,
"step": 2815
},
{
"epoch": 0.37832036490474913,
"grad_norm": 3.484375,
"learning_rate": 1.564204207630472e-05,
"loss": 4.4272,
"step": 2820
},
{
"epoch": 0.37899114569358733,
"grad_norm": 3.484375,
"learning_rate": 1.56226929210997e-05,
"loss": 4.409,
"step": 2825
},
{
"epoch": 0.3796619264824255,
"grad_norm": 3.796875,
"learning_rate": 1.560331293417042e-05,
"loss": 4.3887,
"step": 2830
},
{
"epoch": 0.3803327072712637,
"grad_norm": 3.828125,
"learning_rate": 1.5583902221785966e-05,
"loss": 4.412,
"step": 2835
},
{
"epoch": 0.381003488060102,
"grad_norm": 3.78125,
"learning_rate": 1.5564460890383897e-05,
"loss": 4.3847,
"step": 2840
},
{
"epoch": 0.3816742688489402,
"grad_norm": 4.1875,
"learning_rate": 1.5544989046569673e-05,
"loss": 4.3818,
"step": 2845
},
{
"epoch": 0.3823450496377784,
"grad_norm": 3.75,
"learning_rate": 1.552548679711607e-05,
"loss": 4.4052,
"step": 2850
},
{
"epoch": 0.3830158304266166,
"grad_norm": 3.671875,
"learning_rate": 1.5505954248962588e-05,
"loss": 4.3718,
"step": 2855
},
{
"epoch": 0.3836866112154548,
"grad_norm": 3.59375,
"learning_rate": 1.548639150921487e-05,
"loss": 4.3968,
"step": 2860
},
{
"epoch": 0.384357392004293,
"grad_norm": 3.546875,
"learning_rate": 1.5466798685144114e-05,
"loss": 4.3778,
"step": 2865
},
{
"epoch": 0.38502817279313123,
"grad_norm": 3.53125,
"learning_rate": 1.544717588418648e-05,
"loss": 4.3772,
"step": 2870
},
{
"epoch": 0.38569895358196943,
"grad_norm": 3.703125,
"learning_rate": 1.542752321394251e-05,
"loss": 4.423,
"step": 2875
},
{
"epoch": 0.3863697343708076,
"grad_norm": 3.4375,
"learning_rate": 1.5407840782176523e-05,
"loss": 4.3827,
"step": 2880
},
{
"epoch": 0.3870405151596458,
"grad_norm": 4.03125,
"learning_rate": 1.5388128696816048e-05,
"loss": 4.376,
"step": 2885
},
{
"epoch": 0.387711295948484,
"grad_norm": 3.375,
"learning_rate": 1.5368387065951206e-05,
"loss": 4.4086,
"step": 2890
},
{
"epoch": 0.3883820767373222,
"grad_norm": 3.65625,
"learning_rate": 1.5348615997834135e-05,
"loss": 4.4191,
"step": 2895
},
{
"epoch": 0.3890528575261605,
"grad_norm": 3.875,
"learning_rate": 1.532881560087839e-05,
"loss": 4.3906,
"step": 2900
},
{
"epoch": 0.3897236383149987,
"grad_norm": 3.953125,
"learning_rate": 1.5308985983658336e-05,
"loss": 4.3939,
"step": 2905
},
{
"epoch": 0.3903944191038369,
"grad_norm": 3.171875,
"learning_rate": 1.5289127254908595e-05,
"loss": 4.3537,
"step": 2910
},
{
"epoch": 0.3910651998926751,
"grad_norm": 3.6875,
"learning_rate": 1.526923952352339e-05,
"loss": 4.4069,
"step": 2915
},
{
"epoch": 0.3917359806815133,
"grad_norm": 3.578125,
"learning_rate": 1.5249322898555998e-05,
"loss": 4.4163,
"step": 2920
},
{
"epoch": 0.3924067614703515,
"grad_norm": 3.984375,
"learning_rate": 1.522937748921812e-05,
"loss": 4.4153,
"step": 2925
},
{
"epoch": 0.39307754225918967,
"grad_norm": 3.6875,
"learning_rate": 1.5209403404879305e-05,
"loss": 4.3778,
"step": 2930
},
{
"epoch": 0.3937483230480279,
"grad_norm": 3.71875,
"learning_rate": 1.5189400755066332e-05,
"loss": 4.363,
"step": 2935
},
{
"epoch": 0.3944191038368661,
"grad_norm": 3.703125,
"learning_rate": 1.516936964946262e-05,
"loss": 4.4285,
"step": 2940
},
{
"epoch": 0.3950898846257043,
"grad_norm": 3.953125,
"learning_rate": 1.5149310197907623e-05,
"loss": 4.4109,
"step": 2945
},
{
"epoch": 0.3957606654145425,
"grad_norm": 3.5625,
"learning_rate": 1.5129222510396236e-05,
"loss": 4.4281,
"step": 2950
},
{
"epoch": 0.3964314462033807,
"grad_norm": 3.765625,
"learning_rate": 1.5109106697078167e-05,
"loss": 4.3507,
"step": 2955
},
{
"epoch": 0.3971022269922189,
"grad_norm": 3.8125,
"learning_rate": 1.5088962868257367e-05,
"loss": 4.3791,
"step": 2960
},
{
"epoch": 0.3977730077810572,
"grad_norm": 3.828125,
"learning_rate": 1.50687911343914e-05,
"loss": 4.41,
"step": 2965
},
{
"epoch": 0.3984437885698954,
"grad_norm": 3.5625,
"learning_rate": 1.504859160609085e-05,
"loss": 4.382,
"step": 2970
},
{
"epoch": 0.3991145693587336,
"grad_norm": 4.0,
"learning_rate": 1.502836439411871e-05,
"loss": 4.4214,
"step": 2975
},
{
"epoch": 0.39978535014757177,
"grad_norm": 3.875,
"learning_rate": 1.5008109609389769e-05,
"loss": 4.3487,
"step": 2980
},
{
"epoch": 0.40045613093640997,
"grad_norm": 3.734375,
"learning_rate": 1.4987827362970011e-05,
"loss": 4.3967,
"step": 2985
},
{
"epoch": 0.40112691172524817,
"grad_norm": 3.6875,
"learning_rate": 1.4967517766076017e-05,
"loss": 4.3919,
"step": 2990
},
{
"epoch": 0.4017976925140864,
"grad_norm": 3.8125,
"learning_rate": 1.4947180930074326e-05,
"loss": 4.4002,
"step": 2995
},
{
"epoch": 0.4024684733029246,
"grad_norm": 3.78125,
"learning_rate": 1.492681696648085e-05,
"loss": 4.3534,
"step": 3000
},
{
"epoch": 0.4031392540917628,
"grad_norm": 3.953125,
"learning_rate": 1.4906425986960251e-05,
"loss": 4.3921,
"step": 3005
},
{
"epoch": 0.403810034880601,
"grad_norm": 3.578125,
"learning_rate": 1.4886008103325336e-05,
"loss": 4.4077,
"step": 3010
},
{
"epoch": 0.4044808156694392,
"grad_norm": 3.953125,
"learning_rate": 1.4865563427536432e-05,
"loss": 4.4167,
"step": 3015
},
{
"epoch": 0.4051515964582774,
"grad_norm": 3.796875,
"learning_rate": 1.4845092071700786e-05,
"loss": 4.3825,
"step": 3020
},
{
"epoch": 0.4058223772471156,
"grad_norm": 3.375,
"learning_rate": 1.4824594148071936e-05,
"loss": 4.4115,
"step": 3025
},
{
"epoch": 0.40649315803595387,
"grad_norm": 3.515625,
"learning_rate": 1.480406976904911e-05,
"loss": 4.392,
"step": 3030
},
{
"epoch": 0.40716393882479207,
"grad_norm": 3.671875,
"learning_rate": 1.47835190471766e-05,
"loss": 4.4223,
"step": 3035
},
{
"epoch": 0.40783471961363027,
"grad_norm": 3.765625,
"learning_rate": 1.4762942095143147e-05,
"loss": 4.3812,
"step": 3040
},
{
"epoch": 0.40850550040246847,
"grad_norm": 3.625,
"learning_rate": 1.4742339025781328e-05,
"loss": 4.3885,
"step": 3045
},
{
"epoch": 0.40917628119130667,
"grad_norm": 3.90625,
"learning_rate": 1.4721709952066925e-05,
"loss": 4.3846,
"step": 3050
},
{
"epoch": 0.40984706198014487,
"grad_norm": 3.453125,
"learning_rate": 1.4701054987118319e-05,
"loss": 4.3952,
"step": 3055
},
{
"epoch": 0.4105178427689831,
"grad_norm": 3.53125,
"learning_rate": 1.4680374244195861e-05,
"loss": 4.4121,
"step": 3060
},
{
"epoch": 0.4111886235578213,
"grad_norm": 3.671875,
"learning_rate": 1.465966783670126e-05,
"loss": 4.4089,
"step": 3065
},
{
"epoch": 0.4118594043466595,
"grad_norm": 3.53125,
"learning_rate": 1.4638935878176948e-05,
"loss": 4.3886,
"step": 3070
},
{
"epoch": 0.4125301851354977,
"grad_norm": 3.75,
"learning_rate": 1.461817848230547e-05,
"loss": 4.4352,
"step": 3075
},
{
"epoch": 0.4132009659243359,
"grad_norm": 3.546875,
"learning_rate": 1.4597395762908848e-05,
"loss": 4.4082,
"step": 3080
},
{
"epoch": 0.4138717467131741,
"grad_norm": 3.765625,
"learning_rate": 1.4576587833947975e-05,
"loss": 4.3777,
"step": 3085
},
{
"epoch": 0.41454252750201237,
"grad_norm": 3.90625,
"learning_rate": 1.455575480952197e-05,
"loss": 4.4432,
"step": 3090
},
{
"epoch": 0.41521330829085057,
"grad_norm": 3.578125,
"learning_rate": 1.4534896803867568e-05,
"loss": 4.4069,
"step": 3095
},
{
"epoch": 0.41588408907968877,
"grad_norm": 3.671875,
"learning_rate": 1.4514013931358481e-05,
"loss": 4.432,
"step": 3100
},
{
"epoch": 0.41655486986852697,
"grad_norm": 3.6875,
"learning_rate": 1.4493106306504783e-05,
"loss": 4.3867,
"step": 3105
},
{
"epoch": 0.41722565065736517,
"grad_norm": 3.5,
"learning_rate": 1.4472174043952271e-05,
"loss": 4.4033,
"step": 3110
},
{
"epoch": 0.41789643144620336,
"grad_norm": 3.46875,
"learning_rate": 1.4451217258481846e-05,
"loss": 4.3926,
"step": 3115
},
{
"epoch": 0.41856721223504156,
"grad_norm": 3.8125,
"learning_rate": 1.4430236065008874e-05,
"loss": 4.4022,
"step": 3120
},
{
"epoch": 0.4192379930238798,
"grad_norm": 3.71875,
"learning_rate": 1.4409230578582566e-05,
"loss": 4.4122,
"step": 3125
},
{
"epoch": 0.419908773812718,
"grad_norm": 3.71875,
"learning_rate": 1.438820091438534e-05,
"loss": 4.3715,
"step": 3130
},
{
"epoch": 0.4205795546015562,
"grad_norm": 3.734375,
"learning_rate": 1.4367147187732192e-05,
"loss": 4.4037,
"step": 3135
},
{
"epoch": 0.4212503353903944,
"grad_norm": 3.75,
"learning_rate": 1.4346069514070055e-05,
"loss": 4.3814,
"step": 3140
},
{
"epoch": 0.4219211161792326,
"grad_norm": 3.984375,
"learning_rate": 1.4324968008977194e-05,
"loss": 4.3927,
"step": 3145
},
{
"epoch": 0.4225918969680708,
"grad_norm": 3.8125,
"learning_rate": 1.4303842788162531e-05,
"loss": 4.4446,
"step": 3150
},
{
"epoch": 0.42326267775690907,
"grad_norm": 3.5,
"learning_rate": 1.4282693967465044e-05,
"loss": 4.3785,
"step": 3155
},
{
"epoch": 0.42393345854574727,
"grad_norm": 3.65625,
"learning_rate": 1.4261521662853115e-05,
"loss": 4.3995,
"step": 3160
},
{
"epoch": 0.42460423933458546,
"grad_norm": 3.71875,
"learning_rate": 1.42403259904239e-05,
"loss": 4.4148,
"step": 3165
},
{
"epoch": 0.42527502012342366,
"grad_norm": 3.625,
"learning_rate": 1.4219107066402694e-05,
"loss": 4.3924,
"step": 3170
},
{
"epoch": 0.42594580091226186,
"grad_norm": 3.734375,
"learning_rate": 1.4197865007142284e-05,
"loss": 4.4051,
"step": 3175
},
{
"epoch": 0.42661658170110006,
"grad_norm": 3.46875,
"learning_rate": 1.4176599929122323e-05,
"loss": 4.4072,
"step": 3180
},
{
"epoch": 0.42728736248993826,
"grad_norm": 3.5,
"learning_rate": 1.4155311948948687e-05,
"loss": 4.4207,
"step": 3185
},
{
"epoch": 0.4279581432787765,
"grad_norm": 3.609375,
"learning_rate": 1.4134001183352833e-05,
"loss": 4.4253,
"step": 3190
},
{
"epoch": 0.4286289240676147,
"grad_norm": 3.84375,
"learning_rate": 1.4112667749191156e-05,
"loss": 4.4142,
"step": 3195
},
{
"epoch": 0.4292997048564529,
"grad_norm": 3.859375,
"learning_rate": 1.4091311763444358e-05,
"loss": 4.3874,
"step": 3200
},
{
"epoch": 0.4299704856452911,
"grad_norm": 3.953125,
"learning_rate": 1.4069933343216804e-05,
"loss": 4.3932,
"step": 3205
},
{
"epoch": 0.4306412664341293,
"grad_norm": 3.53125,
"learning_rate": 1.4048532605735872e-05,
"loss": 4.3913,
"step": 3210
},
{
"epoch": 0.4313120472229675,
"grad_norm": 3.59375,
"learning_rate": 1.4027109668351316e-05,
"loss": 4.3544,
"step": 3215
},
{
"epoch": 0.43198282801180576,
"grad_norm": 3.921875,
"learning_rate": 1.4005664648534622e-05,
"loss": 4.375,
"step": 3220
},
{
"epoch": 0.43265360880064396,
"grad_norm": 3.78125,
"learning_rate": 1.3984197663878367e-05,
"loss": 4.4052,
"step": 3225
},
{
"epoch": 0.43332438958948216,
"grad_norm": 3.75,
"learning_rate": 1.396270883209557e-05,
"loss": 4.3715,
"step": 3230
},
{
"epoch": 0.43399517037832036,
"grad_norm": 3.515625,
"learning_rate": 1.3941198271019043e-05,
"loss": 4.4174,
"step": 3235
},
{
"epoch": 0.43466595116715856,
"grad_norm": 3.671875,
"learning_rate": 1.3919666098600753e-05,
"loss": 4.4026,
"step": 3240
},
{
"epoch": 0.43533673195599676,
"grad_norm": 3.609375,
"learning_rate": 1.3898112432911173e-05,
"loss": 4.4225,
"step": 3245
},
{
"epoch": 0.436007512744835,
"grad_norm": 3.859375,
"learning_rate": 1.3876537392138629e-05,
"loss": 4.4141,
"step": 3250
},
{
"epoch": 0.4366782935336732,
"grad_norm": 3.640625,
"learning_rate": 1.3854941094588661e-05,
"loss": 4.4026,
"step": 3255
},
{
"epoch": 0.4373490743225114,
"grad_norm": 3.734375,
"learning_rate": 1.3833323658683364e-05,
"loss": 4.3962,
"step": 3260
},
{
"epoch": 0.4380198551113496,
"grad_norm": 3.6875,
"learning_rate": 1.3811685202960752e-05,
"loss": 4.3929,
"step": 3265
},
{
"epoch": 0.4386906359001878,
"grad_norm": 3.703125,
"learning_rate": 1.3790025846074091e-05,
"loss": 4.3999,
"step": 3270
},
{
"epoch": 0.439361416689026,
"grad_norm": 3.765625,
"learning_rate": 1.3768345706791257e-05,
"loss": 4.4429,
"step": 3275
},
{
"epoch": 0.4400321974778642,
"grad_norm": 3.921875,
"learning_rate": 1.3746644903994097e-05,
"loss": 4.3816,
"step": 3280
},
{
"epoch": 0.44070297826670246,
"grad_norm": 3.859375,
"learning_rate": 1.3724923556677757e-05,
"loss": 4.3834,
"step": 3285
},
{
"epoch": 0.44137375905554066,
"grad_norm": 3.5,
"learning_rate": 1.3703181783950032e-05,
"loss": 4.3509,
"step": 3290
},
{
"epoch": 0.44204453984437886,
"grad_norm": 3.6875,
"learning_rate": 1.3681419705030728e-05,
"loss": 4.4298,
"step": 3295
},
{
"epoch": 0.44271532063321706,
"grad_norm": 3.484375,
"learning_rate": 1.3659637439250998e-05,
"loss": 4.4165,
"step": 3300
},
{
"epoch": 0.44338610142205526,
"grad_norm": 3.65625,
"learning_rate": 1.3637835106052686e-05,
"loss": 4.4027,
"step": 3305
},
{
"epoch": 0.44405688221089346,
"grad_norm": 3.8125,
"learning_rate": 1.361601282498767e-05,
"loss": 4.4034,
"step": 3310
},
{
"epoch": 0.4447276629997317,
"grad_norm": 3.53125,
"learning_rate": 1.3594170715717222e-05,
"loss": 4.4133,
"step": 3315
},
{
"epoch": 0.4453984437885699,
"grad_norm": 3.765625,
"learning_rate": 1.3572308898011328e-05,
"loss": 4.3764,
"step": 3320
},
{
"epoch": 0.4460692245774081,
"grad_norm": 3.90625,
"learning_rate": 1.355042749174806e-05,
"loss": 4.4039,
"step": 3325
},
{
"epoch": 0.4467400053662463,
"grad_norm": 3.6875,
"learning_rate": 1.3528526616912885e-05,
"loss": 4.4143,
"step": 3330
},
{
"epoch": 0.4474107861550845,
"grad_norm": 3.65625,
"learning_rate": 1.3506606393598037e-05,
"loss": 4.4029,
"step": 3335
},
{
"epoch": 0.4480815669439227,
"grad_norm": 3.734375,
"learning_rate": 1.3484666942001843e-05,
"loss": 4.3849,
"step": 3340
},
{
"epoch": 0.44875234773276096,
"grad_norm": 3.84375,
"learning_rate": 1.3462708382428067e-05,
"loss": 4.3815,
"step": 3345
},
{
"epoch": 0.44942312852159916,
"grad_norm": 3.640625,
"learning_rate": 1.344073083528525e-05,
"loss": 4.3983,
"step": 3350
},
{
"epoch": 0.45009390931043736,
"grad_norm": 3.796875,
"learning_rate": 1.3418734421086049e-05,
"loss": 4.4092,
"step": 3355
},
{
"epoch": 0.45076469009927556,
"grad_norm": 3.578125,
"learning_rate": 1.3396719260446582e-05,
"loss": 4.426,
"step": 3360
},
{
"epoch": 0.45143547088811375,
"grad_norm": 3.53125,
"learning_rate": 1.337468547408576e-05,
"loss": 4.3939,
"step": 3365
},
{
"epoch": 0.45210625167695195,
"grad_norm": 3.78125,
"learning_rate": 1.3352633182824623e-05,
"loss": 4.408,
"step": 3370
},
{
"epoch": 0.45277703246579015,
"grad_norm": 3.84375,
"learning_rate": 1.3330562507585685e-05,
"loss": 4.412,
"step": 3375
},
{
"epoch": 0.4534478132546284,
"grad_norm": 3.765625,
"learning_rate": 1.3308473569392274e-05,
"loss": 4.3687,
"step": 3380
},
{
"epoch": 0.4541185940434666,
"grad_norm": 3.34375,
"learning_rate": 1.3286366489367848e-05,
"loss": 4.4033,
"step": 3385
},
{
"epoch": 0.4547893748323048,
"grad_norm": 3.5625,
"learning_rate": 1.3264241388735352e-05,
"loss": 4.3978,
"step": 3390
},
{
"epoch": 0.455460155621143,
"grad_norm": 3.8125,
"learning_rate": 1.324209838881655e-05,
"loss": 4.3656,
"step": 3395
},
{
"epoch": 0.4561309364099812,
"grad_norm": 3.6875,
"learning_rate": 1.3219937611031355e-05,
"loss": 4.3951,
"step": 3400
},
{
"epoch": 0.4568017171988194,
"grad_norm": 3.71875,
"learning_rate": 1.3197759176897153e-05,
"loss": 4.3775,
"step": 3405
},
{
"epoch": 0.45747249798765766,
"grad_norm": 3.734375,
"learning_rate": 1.3175563208028161e-05,
"loss": 4.395,
"step": 3410
},
{
"epoch": 0.45814327877649585,
"grad_norm": 3.640625,
"learning_rate": 1.3153349826134736e-05,
"loss": 4.3827,
"step": 3415
},
{
"epoch": 0.45881405956533405,
"grad_norm": 3.53125,
"learning_rate": 1.3131119153022728e-05,
"loss": 4.3814,
"step": 3420
},
{
"epoch": 0.45948484035417225,
"grad_norm": 3.765625,
"learning_rate": 1.3108871310592799e-05,
"loss": 4.3145,
"step": 3425
},
{
"epoch": 0.46015562114301045,
"grad_norm": 3.578125,
"learning_rate": 1.3086606420839754e-05,
"loss": 4.3922,
"step": 3430
},
{
"epoch": 0.46082640193184865,
"grad_norm": 3.84375,
"learning_rate": 1.3064324605851876e-05,
"loss": 4.4014,
"step": 3435
},
{
"epoch": 0.4614971827206869,
"grad_norm": 4.03125,
"learning_rate": 1.3042025987810267e-05,
"loss": 4.396,
"step": 3440
},
{
"epoch": 0.4621679635095251,
"grad_norm": 3.671875,
"learning_rate": 1.3019710688988157e-05,
"loss": 4.3681,
"step": 3445
},
{
"epoch": 0.4628387442983633,
"grad_norm": 3.765625,
"learning_rate": 1.2997378831750242e-05,
"loss": 4.3799,
"step": 3450
},
{
"epoch": 0.4635095250872015,
"grad_norm": 3.359375,
"learning_rate": 1.297503053855203e-05,
"loss": 4.4308,
"step": 3455
},
{
"epoch": 0.4641803058760397,
"grad_norm": 3.65625,
"learning_rate": 1.2952665931939142e-05,
"loss": 4.3964,
"step": 3460
},
{
"epoch": 0.4648510866648779,
"grad_norm": 3.71875,
"learning_rate": 1.2930285134546652e-05,
"loss": 4.4162,
"step": 3465
},
{
"epoch": 0.4655218674537161,
"grad_norm": 3.78125,
"learning_rate": 1.2907888269098418e-05,
"loss": 4.4147,
"step": 3470
},
{
"epoch": 0.46619264824255435,
"grad_norm": 3.53125,
"learning_rate": 1.2885475458406415e-05,
"loss": 4.3725,
"step": 3475
},
{
"epoch": 0.46686342903139255,
"grad_norm": 3.828125,
"learning_rate": 1.286304682537004e-05,
"loss": 4.3931,
"step": 3480
},
{
"epoch": 0.46753420982023075,
"grad_norm": 3.765625,
"learning_rate": 1.2840602492975458e-05,
"loss": 4.3986,
"step": 3485
},
{
"epoch": 0.46820499060906895,
"grad_norm": 3.625,
"learning_rate": 1.2818142584294914e-05,
"loss": 4.3968,
"step": 3490
},
{
"epoch": 0.46887577139790715,
"grad_norm": 3.53125,
"learning_rate": 1.2795667222486078e-05,
"loss": 4.3896,
"step": 3495
},
{
"epoch": 0.46954655218674535,
"grad_norm": 3.734375,
"learning_rate": 1.2773176530791342e-05,
"loss": 4.4133,
"step": 3500
},
{
"epoch": 0.4702173329755836,
"grad_norm": 3.546875,
"learning_rate": 1.275067063253717e-05,
"loss": 4.3691,
"step": 3505
},
{
"epoch": 0.4708881137644218,
"grad_norm": 3.6875,
"learning_rate": 1.2728149651133395e-05,
"loss": 4.3764,
"step": 3510
},
{
"epoch": 0.47155889455326,
"grad_norm": 3.625,
"learning_rate": 1.2705613710072575e-05,
"loss": 4.3814,
"step": 3515
},
{
"epoch": 0.4722296753420982,
"grad_norm": 3.65625,
"learning_rate": 1.2683062932929286e-05,
"loss": 4.3811,
"step": 3520
},
{
"epoch": 0.4729004561309364,
"grad_norm": 3.78125,
"learning_rate": 1.266049744335947e-05,
"loss": 4.3836,
"step": 3525
},
{
"epoch": 0.4735712369197746,
"grad_norm": 3.796875,
"learning_rate": 1.2637917365099726e-05,
"loss": 4.415,
"step": 3530
},
{
"epoch": 0.47424201770861285,
"grad_norm": 3.78125,
"learning_rate": 1.2615322821966665e-05,
"loss": 4.3702,
"step": 3535
},
{
"epoch": 0.47491279849745105,
"grad_norm": 3.53125,
"learning_rate": 1.2592713937856202e-05,
"loss": 4.403,
"step": 3540
},
{
"epoch": 0.47558357928628925,
"grad_norm": 3.828125,
"learning_rate": 1.2570090836742908e-05,
"loss": 4.4,
"step": 3545
},
{
"epoch": 0.47625436007512745,
"grad_norm": 3.8125,
"learning_rate": 1.2547453642679287e-05,
"loss": 4.423,
"step": 3550
},
{
"epoch": 0.47692514086396565,
"grad_norm": 3.578125,
"learning_rate": 1.2524802479795148e-05,
"loss": 4.4383,
"step": 3555
},
{
"epoch": 0.47759592165280385,
"grad_norm": 3.796875,
"learning_rate": 1.2502137472296879e-05,
"loss": 4.393,
"step": 3560
},
{
"epoch": 0.47826670244164204,
"grad_norm": 3.71875,
"learning_rate": 1.2479458744466788e-05,
"loss": 4.4247,
"step": 3565
},
{
"epoch": 0.4789374832304803,
"grad_norm": 3.625,
"learning_rate": 1.2456766420662416e-05,
"loss": 4.4019,
"step": 3570
},
{
"epoch": 0.4796082640193185,
"grad_norm": 3.46875,
"learning_rate": 1.2434060625315869e-05,
"loss": 4.3897,
"step": 3575
},
{
"epoch": 0.4802790448081567,
"grad_norm": 3.828125,
"learning_rate": 1.241134148293311e-05,
"loss": 4.4322,
"step": 3580
},
{
"epoch": 0.4809498255969949,
"grad_norm": 3.59375,
"learning_rate": 1.2388609118093293e-05,
"loss": 4.4191,
"step": 3585
},
{
"epoch": 0.4816206063858331,
"grad_norm": 3.625,
"learning_rate": 1.2365863655448076e-05,
"loss": 4.3682,
"step": 3590
},
{
"epoch": 0.4822913871746713,
"grad_norm": 3.65625,
"learning_rate": 1.2343105219720945e-05,
"loss": 4.3919,
"step": 3595
},
{
"epoch": 0.48296216796350955,
"grad_norm": 3.765625,
"learning_rate": 1.2320333935706514e-05,
"loss": 4.3884,
"step": 3600
},
{
"epoch": 0.48363294875234775,
"grad_norm": 3.546875,
"learning_rate": 1.2297549928269856e-05,
"loss": 4.4035,
"step": 3605
},
{
"epoch": 0.48430372954118595,
"grad_norm": 3.875,
"learning_rate": 1.2274753322345808e-05,
"loss": 4.372,
"step": 3610
},
{
"epoch": 0.48497451033002414,
"grad_norm": 3.546875,
"learning_rate": 1.2251944242938291e-05,
"loss": 4.423,
"step": 3615
},
{
"epoch": 0.48564529111886234,
"grad_norm": 3.84375,
"learning_rate": 1.2229122815119629e-05,
"loss": 4.4199,
"step": 3620
},
{
"epoch": 0.48631607190770054,
"grad_norm": 3.765625,
"learning_rate": 1.2206289164029852e-05,
"loss": 4.4277,
"step": 3625
},
{
"epoch": 0.4869868526965388,
"grad_norm": 3.84375,
"learning_rate": 1.2183443414876014e-05,
"loss": 4.4376,
"step": 3630
},
{
"epoch": 0.487657633485377,
"grad_norm": 3.875,
"learning_rate": 1.2160585692931517e-05,
"loss": 4.3829,
"step": 3635
},
{
"epoch": 0.4883284142742152,
"grad_norm": 3.828125,
"learning_rate": 1.2137716123535403e-05,
"loss": 4.3675,
"step": 3640
},
{
"epoch": 0.4889991950630534,
"grad_norm": 3.734375,
"learning_rate": 1.211483483209169e-05,
"loss": 4.3905,
"step": 3645
},
{
"epoch": 0.4896699758518916,
"grad_norm": 3.65625,
"learning_rate": 1.2091941944068666e-05,
"loss": 4.4176,
"step": 3650
},
{
"epoch": 0.4903407566407298,
"grad_norm": 3.484375,
"learning_rate": 1.2069037584998213e-05,
"loss": 4.4009,
"step": 3655
},
{
"epoch": 0.491011537429568,
"grad_norm": 3.828125,
"learning_rate": 1.204612188047511e-05,
"loss": 4.3636,
"step": 3660
},
{
"epoch": 0.49168231821840624,
"grad_norm": 3.796875,
"learning_rate": 1.2023194956156348e-05,
"loss": 4.393,
"step": 3665
},
{
"epoch": 0.49235309900724444,
"grad_norm": 3.890625,
"learning_rate": 1.2000256937760446e-05,
"loss": 4.4157,
"step": 3670
},
{
"epoch": 0.49302387979608264,
"grad_norm": 3.578125,
"learning_rate": 1.1977307951066752e-05,
"loss": 4.4211,
"step": 3675
},
{
"epoch": 0.49369466058492084,
"grad_norm": 3.640625,
"learning_rate": 1.1954348121914758e-05,
"loss": 4.3617,
"step": 3680
},
{
"epoch": 0.49436544137375904,
"grad_norm": 3.875,
"learning_rate": 1.1931377576203413e-05,
"loss": 4.4282,
"step": 3685
},
{
"epoch": 0.49503622216259724,
"grad_norm": 3.515625,
"learning_rate": 1.190839643989043e-05,
"loss": 4.3486,
"step": 3690
},
{
"epoch": 0.4957070029514355,
"grad_norm": 3.859375,
"learning_rate": 1.1885404838991592e-05,
"loss": 4.4314,
"step": 3695
},
{
"epoch": 0.4963777837402737,
"grad_norm": 3.84375,
"learning_rate": 1.1862402899580063e-05,
"loss": 4.3974,
"step": 3700
},
{
"epoch": 0.4970485645291119,
"grad_norm": 3.625,
"learning_rate": 1.1839390747785702e-05,
"loss": 4.4033,
"step": 3705
},
{
"epoch": 0.4977193453179501,
"grad_norm": 3.71875,
"learning_rate": 1.1816368509794365e-05,
"loss": 4.3641,
"step": 3710
},
{
"epoch": 0.4983901261067883,
"grad_norm": 3.734375,
"learning_rate": 1.1793336311847214e-05,
"loss": 4.3937,
"step": 3715
},
{
"epoch": 0.4990609068956265,
"grad_norm": 4.03125,
"learning_rate": 1.177029428024003e-05,
"loss": 4.4265,
"step": 3720
},
{
"epoch": 0.49973168768446474,
"grad_norm": 4.0,
"learning_rate": 1.1747242541322507e-05,
"loss": 4.4112,
"step": 3725
},
{
"epoch": 0.5004024684733029,
"grad_norm": 3.578125,
"learning_rate": 1.1724181221497581e-05,
"loss": 4.4055,
"step": 3730
},
{
"epoch": 0.5010732492621411,
"grad_norm": 3.984375,
"learning_rate": 1.1701110447220714e-05,
"loss": 4.4015,
"step": 3735
},
{
"epoch": 0.5017440300509793,
"grad_norm": 4.0,
"learning_rate": 1.167803034499922e-05,
"loss": 4.4224,
"step": 3740
},
{
"epoch": 0.5024148108398175,
"grad_norm": 3.96875,
"learning_rate": 1.165494104139155e-05,
"loss": 4.3543,
"step": 3745
},
{
"epoch": 0.5030855916286557,
"grad_norm": 3.65625,
"learning_rate": 1.1631842663006623e-05,
"loss": 4.4313,
"step": 3750
},
{
"epoch": 0.5037563724174939,
"grad_norm": 3.625,
"learning_rate": 1.160873533650311e-05,
"loss": 4.3988,
"step": 3755
},
{
"epoch": 0.5044271532063321,
"grad_norm": 3.625,
"learning_rate": 1.1585619188588757e-05,
"loss": 4.4288,
"step": 3760
},
{
"epoch": 0.5050979339951703,
"grad_norm": 3.734375,
"learning_rate": 1.1562494346019671e-05,
"loss": 4.3717,
"step": 3765
},
{
"epoch": 0.5057687147840085,
"grad_norm": 3.546875,
"learning_rate": 1.1539360935599646e-05,
"loss": 4.387,
"step": 3770
},
{
"epoch": 0.5064394955728468,
"grad_norm": 3.671875,
"learning_rate": 1.151621908417945e-05,
"loss": 4.3882,
"step": 3775
},
{
"epoch": 0.507110276361685,
"grad_norm": 4.0,
"learning_rate": 1.1493068918656145e-05,
"loss": 4.4193,
"step": 3780
},
{
"epoch": 0.5077810571505232,
"grad_norm": 3.796875,
"learning_rate": 1.146991056597237e-05,
"loss": 4.409,
"step": 3785
},
{
"epoch": 0.5084518379393614,
"grad_norm": 3.796875,
"learning_rate": 1.1446744153115668e-05,
"loss": 4.4023,
"step": 3790
},
{
"epoch": 0.5091226187281996,
"grad_norm": 3.65625,
"learning_rate": 1.1423569807117779e-05,
"loss": 4.3982,
"step": 3795
},
{
"epoch": 0.5097933995170378,
"grad_norm": 3.578125,
"learning_rate": 1.1400387655053942e-05,
"loss": 4.3624,
"step": 3800
},
{
"epoch": 0.510464180305876,
"grad_norm": 3.875,
"learning_rate": 1.1377197824042195e-05,
"loss": 4.379,
"step": 3805
},
{
"epoch": 0.5111349610947142,
"grad_norm": 3.796875,
"learning_rate": 1.1354000441242695e-05,
"loss": 4.3815,
"step": 3810
},
{
"epoch": 0.5118057418835524,
"grad_norm": 3.953125,
"learning_rate": 1.1330795633856992e-05,
"loss": 4.4081,
"step": 3815
},
{
"epoch": 0.5124765226723906,
"grad_norm": 3.546875,
"learning_rate": 1.130758352912737e-05,
"loss": 4.4034,
"step": 3820
},
{
"epoch": 0.5131473034612288,
"grad_norm": 3.625,
"learning_rate": 1.128436425433611e-05,
"loss": 4.4055,
"step": 3825
},
{
"epoch": 0.513818084250067,
"grad_norm": 3.703125,
"learning_rate": 1.1261137936804813e-05,
"loss": 4.4122,
"step": 3830
},
{
"epoch": 0.5144888650389053,
"grad_norm": 3.6875,
"learning_rate": 1.1237904703893706e-05,
"loss": 4.3663,
"step": 3835
},
{
"epoch": 0.5151596458277435,
"grad_norm": 3.328125,
"learning_rate": 1.1214664683000927e-05,
"loss": 4.4252,
"step": 3840
},
{
"epoch": 0.5158304266165817,
"grad_norm": 3.59375,
"learning_rate": 1.1191418001561837e-05,
"loss": 4.421,
"step": 3845
},
{
"epoch": 0.5165012074054199,
"grad_norm": 3.65625,
"learning_rate": 1.116816478704833e-05,
"loss": 4.3565,
"step": 3850
},
{
"epoch": 0.5171719881942581,
"grad_norm": 3.546875,
"learning_rate": 1.1144905166968111e-05,
"loss": 4.3452,
"step": 3855
},
{
"epoch": 0.5178427689830963,
"grad_norm": 3.578125,
"learning_rate": 1.1121639268864014e-05,
"loss": 4.4149,
"step": 3860
},
{
"epoch": 0.5185135497719345,
"grad_norm": 3.734375,
"learning_rate": 1.10983672203133e-05,
"loss": 4.4387,
"step": 3865
},
{
"epoch": 0.5191843305607727,
"grad_norm": 3.984375,
"learning_rate": 1.1075089148926956e-05,
"loss": 4.4295,
"step": 3870
},
{
"epoch": 0.5198551113496109,
"grad_norm": 3.625,
"learning_rate": 1.1051805182348992e-05,
"loss": 4.3991,
"step": 3875
},
{
"epoch": 0.5205258921384491,
"grad_norm": 3.453125,
"learning_rate": 1.1028515448255742e-05,
"loss": 4.3959,
"step": 3880
},
{
"epoch": 0.5211966729272873,
"grad_norm": 3.875,
"learning_rate": 1.1005220074355172e-05,
"loss": 4.3896,
"step": 3885
},
{
"epoch": 0.5218674537161255,
"grad_norm": 3.71875,
"learning_rate": 1.0981919188386171e-05,
"loss": 4.3578,
"step": 3890
},
{
"epoch": 0.5225382345049637,
"grad_norm": 3.484375,
"learning_rate": 1.0958612918117853e-05,
"loss": 4.3869,
"step": 3895
},
{
"epoch": 0.523209015293802,
"grad_norm": 3.6875,
"learning_rate": 1.0935301391348852e-05,
"loss": 4.3862,
"step": 3900
},
{
"epoch": 0.5238797960826402,
"grad_norm": 3.78125,
"learning_rate": 1.0911984735906635e-05,
"loss": 4.4063,
"step": 3905
},
{
"epoch": 0.5245505768714784,
"grad_norm": 3.671875,
"learning_rate": 1.0888663079646786e-05,
"loss": 4.3818,
"step": 3910
},
{
"epoch": 0.5252213576603166,
"grad_norm": 3.484375,
"learning_rate": 1.086533655045231e-05,
"loss": 4.4144,
"step": 3915
},
{
"epoch": 0.5258921384491548,
"grad_norm": 3.625,
"learning_rate": 1.0842005276232932e-05,
"loss": 4.3468,
"step": 3920
},
{
"epoch": 0.526562919237993,
"grad_norm": 3.609375,
"learning_rate": 1.0818669384924403e-05,
"loss": 4.3769,
"step": 3925
},
{
"epoch": 0.5272337000268312,
"grad_norm": 3.890625,
"learning_rate": 1.079532900448778e-05,
"loss": 4.4038,
"step": 3930
},
{
"epoch": 0.5279044808156694,
"grad_norm": 3.578125,
"learning_rate": 1.077198426290875e-05,
"loss": 4.407,
"step": 3935
},
{
"epoch": 0.5285752616045076,
"grad_norm": 3.671875,
"learning_rate": 1.07486352881969e-05,
"loss": 4.3713,
"step": 3940
},
{
"epoch": 0.5292460423933458,
"grad_norm": 3.5,
"learning_rate": 1.072528220838504e-05,
"loss": 4.3843,
"step": 3945
},
{
"epoch": 0.529916823182184,
"grad_norm": 3.765625,
"learning_rate": 1.0701925151528484e-05,
"loss": 4.4285,
"step": 3950
},
{
"epoch": 0.5305876039710222,
"grad_norm": 3.578125,
"learning_rate": 1.0678564245704357e-05,
"loss": 4.4273,
"step": 3955
},
{
"epoch": 0.5312583847598604,
"grad_norm": 3.78125,
"learning_rate": 1.0655199619010883e-05,
"loss": 4.3958,
"step": 3960
},
{
"epoch": 0.5319291655486987,
"grad_norm": 3.75,
"learning_rate": 1.06318313995667e-05,
"loss": 4.3411,
"step": 3965
},
{
"epoch": 0.5325999463375369,
"grad_norm": 3.71875,
"learning_rate": 1.060845971551014e-05,
"loss": 4.3776,
"step": 3970
},
{
"epoch": 0.5332707271263751,
"grad_norm": 3.859375,
"learning_rate": 1.0585084694998535e-05,
"loss": 4.4107,
"step": 3975
},
{
"epoch": 0.5339415079152133,
"grad_norm": 3.84375,
"learning_rate": 1.0561706466207506e-05,
"loss": 4.3821,
"step": 3980
},
{
"epoch": 0.5346122887040515,
"grad_norm": 3.90625,
"learning_rate": 1.053832515733028e-05,
"loss": 4.4181,
"step": 3985
},
{
"epoch": 0.5352830694928897,
"grad_norm": 4.0,
"learning_rate": 1.0514940896576957e-05,
"loss": 4.4138,
"step": 3990
},
{
"epoch": 0.5359538502817279,
"grad_norm": 3.90625,
"learning_rate": 1.0491553812173841e-05,
"loss": 4.4058,
"step": 3995
},
{
"epoch": 0.5366246310705661,
"grad_norm": 3.9375,
"learning_rate": 1.04681640323627e-05,
"loss": 4.4066,
"step": 4000
},
{
"epoch": 0.5372954118594043,
"grad_norm": 3.59375,
"learning_rate": 1.0444771685400099e-05,
"loss": 4.3804,
"step": 4005
},
{
"epoch": 0.5379661926482425,
"grad_norm": 3.625,
"learning_rate": 1.0421376899556668e-05,
"loss": 4.4235,
"step": 4010
},
{
"epoch": 0.5386369734370807,
"grad_norm": 3.453125,
"learning_rate": 1.0397979803116422e-05,
"loss": 4.3834,
"step": 4015
},
{
"epoch": 0.5393077542259189,
"grad_norm": 3.734375,
"learning_rate": 1.0374580524376035e-05,
"loss": 4.3586,
"step": 4020
},
{
"epoch": 0.5399785350147572,
"grad_norm": 3.640625,
"learning_rate": 1.0351179191644151e-05,
"loss": 4.3907,
"step": 4025
},
{
"epoch": 0.5406493158035954,
"grad_norm": 3.796875,
"learning_rate": 1.0327775933240679e-05,
"loss": 4.4153,
"step": 4030
},
{
"epoch": 0.5413200965924336,
"grad_norm": 3.484375,
"learning_rate": 1.0304370877496091e-05,
"loss": 4.396,
"step": 4035
},
{
"epoch": 0.5419908773812718,
"grad_norm": 3.5625,
"learning_rate": 1.0280964152750701e-05,
"loss": 4.3967,
"step": 4040
},
{
"epoch": 0.54266165817011,
"grad_norm": 3.84375,
"learning_rate": 1.025755588735399e-05,
"loss": 4.4083,
"step": 4045
},
{
"epoch": 0.5433324389589482,
"grad_norm": 3.671875,
"learning_rate": 1.0234146209663882e-05,
"loss": 4.3799,
"step": 4050
},
{
"epoch": 0.5440032197477864,
"grad_norm": 3.734375,
"learning_rate": 1.0210735248046042e-05,
"loss": 4.3845,
"step": 4055
},
{
"epoch": 0.5446740005366246,
"grad_norm": 3.515625,
"learning_rate": 1.0187323130873175e-05,
"loss": 4.4014,
"step": 4060
},
{
"epoch": 0.5453447813254628,
"grad_norm": 3.765625,
"learning_rate": 1.0163909986524326e-05,
"loss": 4.3884,
"step": 4065
},
{
"epoch": 0.546015562114301,
"grad_norm": 3.5,
"learning_rate": 1.014049594338417e-05,
"loss": 4.4042,
"step": 4070
},
{
"epoch": 0.5466863429031392,
"grad_norm": 3.75,
"learning_rate": 1.0117081129842319e-05,
"loss": 4.3762,
"step": 4075
},
{
"epoch": 0.5473571236919774,
"grad_norm": 3.671875,
"learning_rate": 1.0093665674292594e-05,
"loss": 4.4218,
"step": 4080
},
{
"epoch": 0.5480279044808156,
"grad_norm": 3.53125,
"learning_rate": 1.007024970513235e-05,
"loss": 4.3828,
"step": 4085
},
{
"epoch": 0.5486986852696539,
"grad_norm": 3.671875,
"learning_rate": 1.0046833350761746e-05,
"loss": 4.4037,
"step": 4090
},
{
"epoch": 0.5493694660584921,
"grad_norm": 3.515625,
"learning_rate": 1.0023416739583069e-05,
"loss": 4.3908,
"step": 4095
},
{
"epoch": 0.5500402468473303,
"grad_norm": 3.484375,
"learning_rate": 1e-05,
"loss": 4.4223,
"step": 4100
},
{
"epoch": 0.5507110276361685,
"grad_norm": 3.703125,
"learning_rate": 9.976583260416933e-06,
"loss": 4.3904,
"step": 4105
},
{
"epoch": 0.5513818084250067,
"grad_norm": 3.59375,
"learning_rate": 9.953166649238254e-06,
"loss": 4.3833,
"step": 4110
},
{
"epoch": 0.5520525892138449,
"grad_norm": 3.703125,
"learning_rate": 9.929750294867654e-06,
"loss": 4.3606,
"step": 4115
},
{
"epoch": 0.5527233700026831,
"grad_norm": 3.765625,
"learning_rate": 9.90633432570741e-06,
"loss": 4.4052,
"step": 4120
},
{
"epoch": 0.5533941507915213,
"grad_norm": 3.71875,
"learning_rate": 9.882918870157684e-06,
"loss": 4.4148,
"step": 4125
},
{
"epoch": 0.5540649315803595,
"grad_norm": 3.671875,
"learning_rate": 9.859504056615832e-06,
"loss": 4.4287,
"step": 4130
},
{
"epoch": 0.5547357123691977,
"grad_norm": 3.890625,
"learning_rate": 9.836090013475677e-06,
"loss": 4.3849,
"step": 4135
},
{
"epoch": 0.5554064931580359,
"grad_norm": 3.515625,
"learning_rate": 9.812676869126829e-06,
"loss": 4.3951,
"step": 4140
},
{
"epoch": 0.5560772739468741,
"grad_norm": 3.71875,
"learning_rate": 9.789264751953963e-06,
"loss": 4.3881,
"step": 4145
},
{
"epoch": 0.5567480547357123,
"grad_norm": 3.796875,
"learning_rate": 9.76585379033612e-06,
"loss": 4.432,
"step": 4150
},
{
"epoch": 0.5574188355245506,
"grad_norm": 3.859375,
"learning_rate": 9.742444112646011e-06,
"loss": 4.3632,
"step": 4155
},
{
"epoch": 0.5580896163133888,
"grad_norm": 3.6875,
"learning_rate": 9.719035847249304e-06,
"loss": 4.4386,
"step": 4160
},
{
"epoch": 0.558760397102227,
"grad_norm": 3.484375,
"learning_rate": 9.69562912250391e-06,
"loss": 4.3616,
"step": 4165
},
{
"epoch": 0.5594311778910652,
"grad_norm": 3.734375,
"learning_rate": 9.672224066759323e-06,
"loss": 4.392,
"step": 4170
},
{
"epoch": 0.5601019586799034,
"grad_norm": 3.703125,
"learning_rate": 9.648820808355852e-06,
"loss": 4.4165,
"step": 4175
},
{
"epoch": 0.5607727394687416,
"grad_norm": 3.953125,
"learning_rate": 9.62541947562397e-06,
"loss": 4.4154,
"step": 4180
},
{
"epoch": 0.5614435202575798,
"grad_norm": 3.6875,
"learning_rate": 9.602020196883583e-06,
"loss": 4.4188,
"step": 4185
},
{
"epoch": 0.562114301046418,
"grad_norm": 3.59375,
"learning_rate": 9.578623100443333e-06,
"loss": 4.4307,
"step": 4190
},
{
"epoch": 0.5627850818352562,
"grad_norm": 3.953125,
"learning_rate": 9.555228314599905e-06,
"loss": 4.4093,
"step": 4195
},
{
"epoch": 0.5634558626240944,
"grad_norm": 3.734375,
"learning_rate": 9.531835967637305e-06,
"loss": 4.4235,
"step": 4200
},
{
"epoch": 0.5641266434129326,
"grad_norm": 3.765625,
"learning_rate": 9.508446187826162e-06,
"loss": 4.3954,
"step": 4205
},
{
"epoch": 0.5647974242017708,
"grad_norm": 4.125,
"learning_rate": 9.485059103423046e-06,
"loss": 4.4077,
"step": 4210
},
{
"epoch": 0.565468204990609,
"grad_norm": 3.78125,
"learning_rate": 9.46167484266972e-06,
"loss": 4.4484,
"step": 4215
},
{
"epoch": 0.5661389857794473,
"grad_norm": 3.90625,
"learning_rate": 9.438293533792496e-06,
"loss": 4.4214,
"step": 4220
},
{
"epoch": 0.5668097665682855,
"grad_norm": 3.84375,
"learning_rate": 9.41491530500147e-06,
"loss": 4.4047,
"step": 4225
},
{
"epoch": 0.5674805473571237,
"grad_norm": 4.03125,
"learning_rate": 9.391540284489862e-06,
"loss": 4.377,
"step": 4230
},
{
"epoch": 0.5681513281459619,
"grad_norm": 3.671875,
"learning_rate": 9.368168600433301e-06,
"loss": 4.3916,
"step": 4235
},
{
"epoch": 0.5688221089348001,
"grad_norm": 4.125,
"learning_rate": 9.34480038098912e-06,
"loss": 4.3683,
"step": 4240
},
{
"epoch": 0.5694928897236383,
"grad_norm": 3.8125,
"learning_rate": 9.321435754295648e-06,
"loss": 4.4025,
"step": 4245
},
{
"epoch": 0.5701636705124765,
"grad_norm": 3.78125,
"learning_rate": 9.298074848471521e-06,
"loss": 4.3903,
"step": 4250
},
{
"epoch": 0.5708344513013147,
"grad_norm": 3.78125,
"learning_rate": 9.27471779161496e-06,
"loss": 4.4022,
"step": 4255
},
{
"epoch": 0.5715052320901529,
"grad_norm": 3.921875,
"learning_rate": 9.251364711803103e-06,
"loss": 4.4162,
"step": 4260
},
{
"epoch": 0.5721760128789911,
"grad_norm": 3.671875,
"learning_rate": 9.228015737091254e-06,
"loss": 4.3908,
"step": 4265
},
{
"epoch": 0.5728467936678293,
"grad_norm": 3.75,
"learning_rate": 9.20467099551222e-06,
"loss": 4.3929,
"step": 4270
},
{
"epoch": 0.5735175744566675,
"grad_norm": 3.515625,
"learning_rate": 9.1813306150756e-06,
"loss": 4.3807,
"step": 4275
},
{
"epoch": 0.5741883552455058,
"grad_norm": 3.625,
"learning_rate": 9.157994723767071e-06,
"loss": 4.3848,
"step": 4280
},
{
"epoch": 0.574859136034344,
"grad_norm": 3.59375,
"learning_rate": 9.134663449547693e-06,
"loss": 4.4414,
"step": 4285
},
{
"epoch": 0.5755299168231822,
"grad_norm": 3.6875,
"learning_rate": 9.111336920353216e-06,
"loss": 4.3939,
"step": 4290
},
{
"epoch": 0.5762006976120204,
"grad_norm": 3.65625,
"learning_rate": 9.088015264093365e-06,
"loss": 4.3556,
"step": 4295
},
{
"epoch": 0.5768714784008586,
"grad_norm": 3.640625,
"learning_rate": 9.06469860865115e-06,
"loss": 4.4139,
"step": 4300
},
{
"epoch": 0.5775422591896968,
"grad_norm": 3.65625,
"learning_rate": 9.041387081882152e-06,
"loss": 4.3931,
"step": 4305
},
{
"epoch": 0.578213039978535,
"grad_norm": 3.9375,
"learning_rate": 9.018080811613832e-06,
"loss": 4.3987,
"step": 4310
},
{
"epoch": 0.5788838207673732,
"grad_norm": 3.828125,
"learning_rate": 8.994779925644832e-06,
"loss": 4.3817,
"step": 4315
},
{
"epoch": 0.5795546015562114,
"grad_norm": 3.703125,
"learning_rate": 8.971484551744262e-06,
"loss": 4.3945,
"step": 4320
},
{
"epoch": 0.5802253823450496,
"grad_norm": 3.71875,
"learning_rate": 8.948194817651011e-06,
"loss": 4.3512,
"step": 4325
},
{
"epoch": 0.5808961631338878,
"grad_norm": 4.03125,
"learning_rate": 8.924910851073046e-06,
"loss": 4.3895,
"step": 4330
},
{
"epoch": 0.581566943922726,
"grad_norm": 3.625,
"learning_rate": 8.9016327796867e-06,
"loss": 4.4194,
"step": 4335
},
{
"epoch": 0.5822377247115642,
"grad_norm": 4.0625,
"learning_rate": 8.878360731135987e-06,
"loss": 4.3582,
"step": 4340
},
{
"epoch": 0.5829085055004025,
"grad_norm": 3.8125,
"learning_rate": 8.855094833031894e-06,
"loss": 4.4191,
"step": 4345
},
{
"epoch": 0.5835792862892407,
"grad_norm": 3.734375,
"learning_rate": 8.831835212951672e-06,
"loss": 4.3721,
"step": 4350
},
{
"epoch": 0.5842500670780789,
"grad_norm": 3.96875,
"learning_rate": 8.808581998438164e-06,
"loss": 4.3621,
"step": 4355
},
{
"epoch": 0.5849208478669171,
"grad_norm": 3.75,
"learning_rate": 8.785335316999078e-06,
"loss": 4.3346,
"step": 4360
},
{
"epoch": 0.5855916286557553,
"grad_norm": 3.796875,
"learning_rate": 8.762095296106298e-06,
"loss": 4.4219,
"step": 4365
},
{
"epoch": 0.5862624094445935,
"grad_norm": 3.6875,
"learning_rate": 8.73886206319519e-06,
"loss": 4.3795,
"step": 4370
},
{
"epoch": 0.5869331902334317,
"grad_norm": 3.65625,
"learning_rate": 8.715635745663896e-06,
"loss": 4.3993,
"step": 4375
},
{
"epoch": 0.5876039710222699,
"grad_norm": 3.796875,
"learning_rate": 8.692416470872631e-06,
"loss": 4.3923,
"step": 4380
},
{
"epoch": 0.5882747518111081,
"grad_norm": 3.609375,
"learning_rate": 8.66920436614301e-06,
"loss": 4.3734,
"step": 4385
},
{
"epoch": 0.5889455325999463,
"grad_norm": 3.421875,
"learning_rate": 8.645999558757307e-06,
"loss": 4.394,
"step": 4390
},
{
"epoch": 0.5896163133887845,
"grad_norm": 3.640625,
"learning_rate": 8.622802175957809e-06,
"loss": 4.3979,
"step": 4395
},
{
"epoch": 0.5902870941776227,
"grad_norm": 3.28125,
"learning_rate": 8.599612344946063e-06,
"loss": 4.3944,
"step": 4400
},
{
"epoch": 0.5909578749664609,
"grad_norm": 3.421875,
"learning_rate": 8.576430192882223e-06,
"loss": 4.4249,
"step": 4405
},
{
"epoch": 0.5916286557552992,
"grad_norm": 3.65625,
"learning_rate": 8.553255846884334e-06,
"loss": 4.4077,
"step": 4410
},
{
"epoch": 0.5922994365441374,
"grad_norm": 4.1875,
"learning_rate": 8.530089434027636e-06,
"loss": 4.3869,
"step": 4415
},
{
"epoch": 0.5929702173329756,
"grad_norm": 3.640625,
"learning_rate": 8.506931081343858e-06,
"loss": 4.4153,
"step": 4420
},
{
"epoch": 0.5936409981218138,
"grad_norm": 3.46875,
"learning_rate": 8.483780915820552e-06,
"loss": 4.4152,
"step": 4425
},
{
"epoch": 0.594311778910652,
"grad_norm": 3.59375,
"learning_rate": 8.460639064400354e-06,
"loss": 4.4143,
"step": 4430
},
{
"epoch": 0.5949825596994902,
"grad_norm": 3.90625,
"learning_rate": 8.43750565398033e-06,
"loss": 4.3965,
"step": 4435
},
{
"epoch": 0.5956533404883284,
"grad_norm": 3.625,
"learning_rate": 8.414380811411248e-06,
"loss": 4.398,
"step": 4440
},
{
"epoch": 0.5963241212771666,
"grad_norm": 3.640625,
"learning_rate": 8.391264663496891e-06,
"loss": 4.3566,
"step": 4445
},
{
"epoch": 0.5969949020660048,
"grad_norm": 3.734375,
"learning_rate": 8.36815733699338e-06,
"loss": 4.3862,
"step": 4450
},
{
"epoch": 0.597665682854843,
"grad_norm": 3.734375,
"learning_rate": 8.345058958608453e-06,
"loss": 4.4123,
"step": 4455
},
{
"epoch": 0.5983364636436812,
"grad_norm": 3.578125,
"learning_rate": 8.321969655000782e-06,
"loss": 4.4287,
"step": 4460
},
{
"epoch": 0.5990072444325194,
"grad_norm": 3.640625,
"learning_rate": 8.298889552779288e-06,
"loss": 4.4074,
"step": 4465
},
{
"epoch": 0.5996780252213577,
"grad_norm": 3.6875,
"learning_rate": 8.275818778502417e-06,
"loss": 4.4279,
"step": 4470
},
{
"epoch": 0.6003488060101959,
"grad_norm": 3.828125,
"learning_rate": 8.252757458677495e-06,
"loss": 4.439,
"step": 4475
},
{
"epoch": 0.6010195867990341,
"grad_norm": 3.828125,
"learning_rate": 8.229705719759974e-06,
"loss": 4.3711,
"step": 4480
},
{
"epoch": 0.6016903675878723,
"grad_norm": 3.625,
"learning_rate": 8.206663688152788e-06,
"loss": 4.3654,
"step": 4485
},
{
"epoch": 0.6023611483767105,
"grad_norm": 3.328125,
"learning_rate": 8.183631490205636e-06,
"loss": 4.391,
"step": 4490
},
{
"epoch": 0.6030319291655487,
"grad_norm": 3.75,
"learning_rate": 8.160609252214301e-06,
"loss": 4.3996,
"step": 4495
},
{
"epoch": 0.6037027099543869,
"grad_norm": 3.546875,
"learning_rate": 8.137597100419938e-06,
"loss": 4.4137,
"step": 4500
},
{
"epoch": 0.6043734907432251,
"grad_norm": 3.546875,
"learning_rate": 8.114595161008412e-06,
"loss": 4.4228,
"step": 4505
},
{
"epoch": 0.6050442715320633,
"grad_norm": 3.78125,
"learning_rate": 8.091603560109571e-06,
"loss": 4.3908,
"step": 4510
},
{
"epoch": 0.6057150523209015,
"grad_norm": 3.71875,
"learning_rate": 8.068622423796589e-06,
"loss": 4.4073,
"step": 4515
},
{
"epoch": 0.6063858331097397,
"grad_norm": 3.96875,
"learning_rate": 8.045651878085245e-06,
"loss": 4.3968,
"step": 4520
},
{
"epoch": 0.6070566138985779,
"grad_norm": 3.8125,
"learning_rate": 8.022692048933251e-06,
"loss": 4.3987,
"step": 4525
},
{
"epoch": 0.6077273946874161,
"grad_norm": 3.546875,
"learning_rate": 7.999743062239557e-06,
"loss": 4.4111,
"step": 4530
},
{
"epoch": 0.6083981754762544,
"grad_norm": 3.6875,
"learning_rate": 7.976805043843655e-06,
"loss": 4.3836,
"step": 4535
},
{
"epoch": 0.6090689562650926,
"grad_norm": 3.78125,
"learning_rate": 7.953878119524892e-06,
"loss": 4.4118,
"step": 4540
},
{
"epoch": 0.6097397370539308,
"grad_norm": 3.703125,
"learning_rate": 7.930962415001789e-06,
"loss": 4.3702,
"step": 4545
},
{
"epoch": 0.610410517842769,
"grad_norm": 3.9375,
"learning_rate": 7.908058055931334e-06,
"loss": 4.4395,
"step": 4550
},
{
"epoch": 0.6110812986316072,
"grad_norm": 3.59375,
"learning_rate": 7.885165167908311e-06,
"loss": 4.37,
"step": 4555
},
{
"epoch": 0.6117520794204454,
"grad_norm": 3.546875,
"learning_rate": 7.8622838764646e-06,
"loss": 4.3779,
"step": 4560
},
{
"epoch": 0.6124228602092836,
"grad_norm": 3.859375,
"learning_rate": 7.839414307068486e-06,
"loss": 4.4012,
"step": 4565
},
{
"epoch": 0.6130936409981218,
"grad_norm": 4.28125,
"learning_rate": 7.816556585123989e-06,
"loss": 4.3689,
"step": 4570
},
{
"epoch": 0.61376442178696,
"grad_norm": 3.734375,
"learning_rate": 7.793710835970152e-06,
"loss": 4.3972,
"step": 4575
},
{
"epoch": 0.6144352025757982,
"grad_norm": 3.53125,
"learning_rate": 7.770877184880373e-06,
"loss": 4.3873,
"step": 4580
},
{
"epoch": 0.6151059833646364,
"grad_norm": 3.671875,
"learning_rate": 7.74805575706171e-06,
"loss": 4.3851,
"step": 4585
},
{
"epoch": 0.6157767641534746,
"grad_norm": 3.578125,
"learning_rate": 7.725246677654194e-06,
"loss": 4.3724,
"step": 4590
},
{
"epoch": 0.6164475449423128,
"grad_norm": 3.625,
"learning_rate": 7.702450071730146e-06,
"loss": 4.4216,
"step": 4595
},
{
"epoch": 0.6171183257311511,
"grad_norm": 3.5625,
"learning_rate": 7.679666064293491e-06,
"loss": 4.4061,
"step": 4600
},
{
"epoch": 0.6177891065199893,
"grad_norm": 3.578125,
"learning_rate": 7.656894780279056e-06,
"loss": 4.3874,
"step": 4605
},
{
"epoch": 0.6184598873088275,
"grad_norm": 3.71875,
"learning_rate": 7.634136344551926e-06,
"loss": 4.4111,
"step": 4610
},
{
"epoch": 0.6191306680976657,
"grad_norm": 3.5625,
"learning_rate": 7.6113908819067114e-06,
"loss": 4.4211,
"step": 4615
},
{
"epoch": 0.6198014488865039,
"grad_norm": 3.75,
"learning_rate": 7.588658517066893e-06,
"loss": 4.3891,
"step": 4620
},
{
"epoch": 0.6204722296753421,
"grad_norm": 3.71875,
"learning_rate": 7.565939374684132e-06,
"loss": 4.4386,
"step": 4625
},
{
"epoch": 0.6211430104641803,
"grad_norm": 3.84375,
"learning_rate": 7.543233579337583e-06,
"loss": 4.4506,
"step": 4630
},
{
"epoch": 0.6218137912530185,
"grad_norm": 3.765625,
"learning_rate": 7.520541255533215e-06,
"loss": 4.426,
"step": 4635
},
{
"epoch": 0.6224845720418567,
"grad_norm": 3.890625,
"learning_rate": 7.497862527703127e-06,
"loss": 4.4202,
"step": 4640
},
{
"epoch": 0.6231553528306949,
"grad_norm": 3.640625,
"learning_rate": 7.475197520204852e-06,
"loss": 4.3753,
"step": 4645
},
{
"epoch": 0.6238261336195331,
"grad_norm": 3.640625,
"learning_rate": 7.452546357320714e-06,
"loss": 4.408,
"step": 4650
},
{
"epoch": 0.6244969144083713,
"grad_norm": 3.90625,
"learning_rate": 7.429909163257098e-06,
"loss": 4.3865,
"step": 4655
},
{
"epoch": 0.6251676951972096,
"grad_norm": 4.03125,
"learning_rate": 7.407286062143799e-06,
"loss": 4.3834,
"step": 4660
},
{
"epoch": 0.6258384759860478,
"grad_norm": 3.53125,
"learning_rate": 7.384677178033339e-06,
"loss": 4.4292,
"step": 4665
},
{
"epoch": 0.626509256774886,
"grad_norm": 3.859375,
"learning_rate": 7.362082634900276e-06,
"loss": 4.3479,
"step": 4670
},
{
"epoch": 0.6271800375637242,
"grad_norm": 4.1875,
"learning_rate": 7.339502556640532e-06,
"loss": 4.4081,
"step": 4675
},
{
"epoch": 0.6278508183525624,
"grad_norm": 3.703125,
"learning_rate": 7.3169370670707155e-06,
"loss": 4.39,
"step": 4680
},
{
"epoch": 0.6285215991414006,
"grad_norm": 3.609375,
"learning_rate": 7.294386289927425e-06,
"loss": 4.3982,
"step": 4685
},
{
"epoch": 0.6291923799302388,
"grad_norm": 3.640625,
"learning_rate": 7.271850348866608e-06,
"loss": 4.361,
"step": 4690
},
{
"epoch": 0.629863160719077,
"grad_norm": 3.703125,
"learning_rate": 7.249329367462836e-06,
"loss": 4.4032,
"step": 4695
},
{
"epoch": 0.6305339415079152,
"grad_norm": 3.796875,
"learning_rate": 7.226823469208659e-06,
"loss": 4.3901,
"step": 4700
},
{
"epoch": 0.6312047222967534,
"grad_norm": 4.15625,
"learning_rate": 7.204332777513922e-06,
"loss": 4.394,
"step": 4705
},
{
"epoch": 0.6318755030855916,
"grad_norm": 3.671875,
"learning_rate": 7.181857415705087e-06,
"loss": 4.4286,
"step": 4710
},
{
"epoch": 0.6325462838744298,
"grad_norm": 3.6875,
"learning_rate": 7.1593975070245444e-06,
"loss": 4.4245,
"step": 4715
},
{
"epoch": 0.633217064663268,
"grad_norm": 3.875,
"learning_rate": 7.136953174629964e-06,
"loss": 4.4219,
"step": 4720
},
{
"epoch": 0.6338878454521063,
"grad_norm": 3.828125,
"learning_rate": 7.114524541593586e-06,
"loss": 4.4459,
"step": 4725
},
{
"epoch": 0.6345586262409445,
"grad_norm": 3.734375,
"learning_rate": 7.0921117309015855e-06,
"loss": 4.4021,
"step": 4730
},
{
"epoch": 0.6352294070297827,
"grad_norm": 3.8125,
"learning_rate": 7.069714865453354e-06,
"loss": 4.4067,
"step": 4735
},
{
"epoch": 0.6359001878186209,
"grad_norm": 3.765625,
"learning_rate": 7.0473340680608625e-06,
"loss": 4.4501,
"step": 4740
},
{
"epoch": 0.6365709686074591,
"grad_norm": 3.734375,
"learning_rate": 7.024969461447973e-06,
"loss": 4.3329,
"step": 4745
},
{
"epoch": 0.6372417493962973,
"grad_norm": 3.6875,
"learning_rate": 7.002621168249759e-06,
"loss": 4.4149,
"step": 4750
},
{
"epoch": 0.6379125301851355,
"grad_norm": 3.71875,
"learning_rate": 6.9802893110118455e-06,
"loss": 4.4014,
"step": 4755
},
{
"epoch": 0.6385833109739737,
"grad_norm": 3.75,
"learning_rate": 6.957974012189734e-06,
"loss": 4.397,
"step": 4760
},
{
"epoch": 0.6392540917628119,
"grad_norm": 3.84375,
"learning_rate": 6.935675394148122e-06,
"loss": 4.3629,
"step": 4765
},
{
"epoch": 0.6399248725516501,
"grad_norm": 3.90625,
"learning_rate": 6.913393579160248e-06,
"loss": 4.362,
"step": 4770
},
{
"epoch": 0.6405956533404883,
"grad_norm": 3.703125,
"learning_rate": 6.891128689407205e-06,
"loss": 4.4079,
"step": 4775
},
{
"epoch": 0.6412664341293265,
"grad_norm": 3.5625,
"learning_rate": 6.8688808469772725e-06,
"loss": 4.4518,
"step": 4780
},
{
"epoch": 0.6419372149181647,
"grad_norm": 3.578125,
"learning_rate": 6.846650173865266e-06,
"loss": 4.371,
"step": 4785
},
{
"epoch": 0.642607995707003,
"grad_norm": 3.71875,
"learning_rate": 6.8244367919718445e-06,
"loss": 4.4096,
"step": 4790
},
{
"epoch": 0.6432787764958412,
"grad_norm": 3.484375,
"learning_rate": 6.8022408231028495e-06,
"loss": 4.3522,
"step": 4795
},
{
"epoch": 0.6439495572846794,
"grad_norm": 3.703125,
"learning_rate": 6.780062388968649e-06,
"loss": 4.4125,
"step": 4800
},
{
"epoch": 0.6446203380735176,
"grad_norm": 3.65625,
"learning_rate": 6.757901611183448e-06,
"loss": 4.4193,
"step": 4805
},
{
"epoch": 0.6452911188623558,
"grad_norm": 3.703125,
"learning_rate": 6.7357586112646485e-06,
"loss": 4.4057,
"step": 4810
},
{
"epoch": 0.645961899651194,
"grad_norm": 3.828125,
"learning_rate": 6.713633510632158e-06,
"loss": 4.4303,
"step": 4815
},
{
"epoch": 0.6466326804400322,
"grad_norm": 3.75,
"learning_rate": 6.691526430607729e-06,
"loss": 4.4038,
"step": 4820
},
{
"epoch": 0.6473034612288704,
"grad_norm": 3.671875,
"learning_rate": 6.669437492414317e-06,
"loss": 4.3863,
"step": 4825
},
{
"epoch": 0.6479742420177086,
"grad_norm": 3.6875,
"learning_rate": 6.647366817175383e-06,
"loss": 4.4104,
"step": 4830
},
{
"epoch": 0.6486450228065468,
"grad_norm": 3.46875,
"learning_rate": 6.625314525914244e-06,
"loss": 4.3808,
"step": 4835
},
{
"epoch": 0.649315803595385,
"grad_norm": 3.703125,
"learning_rate": 6.6032807395534205e-06,
"loss": 4.4392,
"step": 4840
},
{
"epoch": 0.6499865843842232,
"grad_norm": 3.671875,
"learning_rate": 6.581265578913952e-06,
"loss": 4.4142,
"step": 4845
},
{
"epoch": 0.6506573651730615,
"grad_norm": 3.59375,
"learning_rate": 6.559269164714753e-06,
"loss": 4.363,
"step": 4850
},
{
"epoch": 0.6513281459618997,
"grad_norm": 3.546875,
"learning_rate": 6.537291617571938e-06,
"loss": 4.431,
"step": 4855
},
{
"epoch": 0.6519989267507379,
"grad_norm": 3.53125,
"learning_rate": 6.515333057998159e-06,
"loss": 4.4049,
"step": 4860
},
{
"epoch": 0.6526697075395761,
"grad_norm": 3.703125,
"learning_rate": 6.4933936064019665e-06,
"loss": 4.424,
"step": 4865
},
{
"epoch": 0.6533404883284143,
"grad_norm": 3.796875,
"learning_rate": 6.471473383087119e-06,
"loss": 4.4004,
"step": 4870
},
{
"epoch": 0.6540112691172525,
"grad_norm": 3.640625,
"learning_rate": 6.449572508251945e-06,
"loss": 4.4021,
"step": 4875
},
{
"epoch": 0.6546820499060907,
"grad_norm": 3.8125,
"learning_rate": 6.427691101988673e-06,
"loss": 4.4447,
"step": 4880
},
{
"epoch": 0.6553528306949289,
"grad_norm": 3.375,
"learning_rate": 6.405829284282781e-06,
"loss": 4.3861,
"step": 4885
},
{
"epoch": 0.6560236114837671,
"grad_norm": 3.796875,
"learning_rate": 6.383987175012332e-06,
"loss": 4.4009,
"step": 4890
},
{
"epoch": 0.6566943922726053,
"grad_norm": 4.0625,
"learning_rate": 6.36216489394732e-06,
"loss": 4.4331,
"step": 4895
},
{
"epoch": 0.6573651730614435,
"grad_norm": 3.5625,
"learning_rate": 6.3403625607490025e-06,
"loss": 4.3998,
"step": 4900
},
{
"epoch": 0.6580359538502817,
"grad_norm": 3.921875,
"learning_rate": 6.318580294969274e-06,
"loss": 4.4288,
"step": 4905
},
{
"epoch": 0.6587067346391199,
"grad_norm": 3.90625,
"learning_rate": 6.296818216049973e-06,
"loss": 4.4509,
"step": 4910
},
{
"epoch": 0.6593775154279582,
"grad_norm": 3.953125,
"learning_rate": 6.2750764433222465e-06,
"loss": 4.3966,
"step": 4915
},
{
"epoch": 0.6600482962167964,
"grad_norm": 3.75,
"learning_rate": 6.2533550960059045e-06,
"loss": 4.3664,
"step": 4920
},
{
"epoch": 0.6607190770056346,
"grad_norm": 3.59375,
"learning_rate": 6.231654293208745e-06,
"loss": 4.3573,
"step": 4925
},
{
"epoch": 0.6613898577944728,
"grad_norm": 3.78125,
"learning_rate": 6.2099741539259125e-06,
"loss": 4.4433,
"step": 4930
},
{
"epoch": 0.662060638583311,
"grad_norm": 3.578125,
"learning_rate": 6.1883147970392506e-06,
"loss": 4.4083,
"step": 4935
},
{
"epoch": 0.6627314193721492,
"grad_norm": 3.71875,
"learning_rate": 6.1666763413166355e-06,
"loss": 4.4263,
"step": 4940
},
{
"epoch": 0.6634022001609874,
"grad_norm": 4.0625,
"learning_rate": 6.145058905411343e-06,
"loss": 4.3954,
"step": 4945
},
{
"epoch": 0.6640729809498256,
"grad_norm": 3.6875,
"learning_rate": 6.123462607861374e-06,
"loss": 4.4228,
"step": 4950
},
{
"epoch": 0.6647437617386638,
"grad_norm": 3.921875,
"learning_rate": 6.101887567088831e-06,
"loss": 4.4016,
"step": 4955
},
{
"epoch": 0.665414542527502,
"grad_norm": 3.640625,
"learning_rate": 6.080333901399252e-06,
"loss": 4.3724,
"step": 4960
},
{
"epoch": 0.6660853233163402,
"grad_norm": 3.796875,
"learning_rate": 6.058801728980962e-06,
"loss": 4.3588,
"step": 4965
},
{
"epoch": 0.6667561041051784,
"grad_norm": 3.734375,
"learning_rate": 6.0372911679044334e-06,
"loss": 4.3668,
"step": 4970
},
{
"epoch": 0.6674268848940166,
"grad_norm": 3.9375,
"learning_rate": 6.015802336121635e-06,
"loss": 4.4079,
"step": 4975
},
{
"epoch": 0.6680976656828549,
"grad_norm": 3.734375,
"learning_rate": 5.994335351465379e-06,
"loss": 4.4178,
"step": 4980
},
{
"epoch": 0.6687684464716931,
"grad_norm": 3.71875,
"learning_rate": 5.972890331648686e-06,
"loss": 4.4,
"step": 4985
},
{
"epoch": 0.6694392272605313,
"grad_norm": 3.984375,
"learning_rate": 5.951467394264132e-06,
"loss": 4.3958,
"step": 4990
},
{
"epoch": 0.6701100080493695,
"grad_norm": 3.84375,
"learning_rate": 5.9300666567831975e-06,
"loss": 4.3187,
"step": 4995
},
{
"epoch": 0.6707807888382077,
"grad_norm": 3.71875,
"learning_rate": 5.908688236555644e-06,
"loss": 4.3914,
"step": 5000
},
{
"epoch": 0.6714515696270459,
"grad_norm": 3.796875,
"learning_rate": 5.887332250808848e-06,
"loss": 4.3597,
"step": 5005
},
{
"epoch": 0.6721223504158841,
"grad_norm": 3.9375,
"learning_rate": 5.8659988166471715e-06,
"loss": 4.3844,
"step": 5010
},
{
"epoch": 0.6727931312047223,
"grad_norm": 3.71875,
"learning_rate": 5.844688051051315e-06,
"loss": 4.4128,
"step": 5015
},
{
"epoch": 0.6734639119935605,
"grad_norm": 3.71875,
"learning_rate": 5.823400070877678e-06,
"loss": 4.3962,
"step": 5020
},
{
"epoch": 0.6741346927823987,
"grad_norm": 3.515625,
"learning_rate": 5.80213499285772e-06,
"loss": 4.3736,
"step": 5025
},
{
"epoch": 0.6748054735712369,
"grad_norm": 3.453125,
"learning_rate": 5.780892933597309e-06,
"loss": 4.3908,
"step": 5030
},
{
"epoch": 0.6754762543600751,
"grad_norm": 3.625,
"learning_rate": 5.7596740095760994e-06,
"loss": 4.4189,
"step": 5035
},
{
"epoch": 0.6761470351489134,
"grad_norm": 3.671875,
"learning_rate": 5.738478337146889e-06,
"loss": 4.4422,
"step": 5040
},
{
"epoch": 0.6768178159377516,
"grad_norm": 3.546875,
"learning_rate": 5.717306032534962e-06,
"loss": 4.4209,
"step": 5045
},
{
"epoch": 0.6774885967265898,
"grad_norm": 3.921875,
"learning_rate": 5.69615721183747e-06,
"loss": 4.376,
"step": 5050
},
{
"epoch": 0.678159377515428,
"grad_norm": 3.671875,
"learning_rate": 5.675031991022809e-06,
"loss": 4.3824,
"step": 5055
},
{
"epoch": 0.6788301583042662,
"grad_norm": 3.640625,
"learning_rate": 5.653930485929944e-06,
"loss": 4.4306,
"step": 5060
},
{
"epoch": 0.6795009390931044,
"grad_norm": 3.84375,
"learning_rate": 5.632852812267814e-06,
"loss": 4.3702,
"step": 5065
},
{
"epoch": 0.6801717198819426,
"grad_norm": 3.515625,
"learning_rate": 5.611799085614663e-06,
"loss": 4.3973,
"step": 5070
},
{
"epoch": 0.6808425006707808,
"grad_norm": 3.578125,
"learning_rate": 5.590769421417435e-06,
"loss": 4.4049,
"step": 5075
},
{
"epoch": 0.681513281459619,
"grad_norm": 3.796875,
"learning_rate": 5.5697639349911305e-06,
"loss": 4.4051,
"step": 5080
},
{
"epoch": 0.6821840622484572,
"grad_norm": 3.546875,
"learning_rate": 5.5487827415181615e-06,
"loss": 4.3982,
"step": 5085
},
{
"epoch": 0.6828548430372954,
"grad_norm": 3.8125,
"learning_rate": 5.52782595604773e-06,
"loss": 4.3795,
"step": 5090
},
{
"epoch": 0.6835256238261336,
"grad_norm": 3.65625,
"learning_rate": 5.506893693495221e-06,
"loss": 4.3762,
"step": 5095
},
{
"epoch": 0.6841964046149718,
"grad_norm": 3.9375,
"learning_rate": 5.48598606864152e-06,
"loss": 4.4103,
"step": 5100
},
{
"epoch": 0.6848671854038101,
"grad_norm": 3.8125,
"learning_rate": 5.465103196132437e-06,
"loss": 4.3863,
"step": 5105
},
{
"epoch": 0.6855379661926483,
"grad_norm": 3.515625,
"learning_rate": 5.4442451904780325e-06,
"loss": 4.4238,
"step": 5110
},
{
"epoch": 0.6862087469814865,
"grad_norm": 3.78125,
"learning_rate": 5.423412166052026e-06,
"loss": 4.3979,
"step": 5115
},
{
"epoch": 0.6868795277703247,
"grad_norm": 3.734375,
"learning_rate": 5.402604237091156e-06,
"loss": 4.4053,
"step": 5120
},
{
"epoch": 0.6875503085591629,
"grad_norm": 3.71875,
"learning_rate": 5.381821517694533e-06,
"loss": 4.4324,
"step": 5125
},
{
"epoch": 0.6882210893480011,
"grad_norm": 3.875,
"learning_rate": 5.361064121823052e-06,
"loss": 4.4192,
"step": 5130
},
{
"epoch": 0.6888918701368393,
"grad_norm": 3.828125,
"learning_rate": 5.340332163298743e-06,
"loss": 4.4003,
"step": 5135
},
{
"epoch": 0.6895626509256775,
"grad_norm": 3.671875,
"learning_rate": 5.3196257558041386e-06,
"loss": 4.3659,
"step": 5140
},
{
"epoch": 0.6902334317145157,
"grad_norm": 3.59375,
"learning_rate": 5.298945012881685e-06,
"loss": 4.3819,
"step": 5145
},
{
"epoch": 0.6909042125033539,
"grad_norm": 3.6875,
"learning_rate": 5.278290047933077e-06,
"loss": 4.4069,
"step": 5150
},
{
"epoch": 0.6915749932921921,
"grad_norm": 3.765625,
"learning_rate": 5.257660974218672e-06,
"loss": 4.39,
"step": 5155
},
{
"epoch": 0.6922457740810303,
"grad_norm": 3.59375,
"learning_rate": 5.237057904856855e-06,
"loss": 4.3851,
"step": 5160
},
{
"epoch": 0.6929165548698685,
"grad_norm": 3.703125,
"learning_rate": 5.216480952823402e-06,
"loss": 4.4171,
"step": 5165
},
{
"epoch": 0.6935873356587068,
"grad_norm": 3.8125,
"learning_rate": 5.1959302309508905e-06,
"loss": 4.3813,
"step": 5170
},
{
"epoch": 0.694258116447545,
"grad_norm": 3.59375,
"learning_rate": 5.175405851928068e-06,
"loss": 4.4198,
"step": 5175
},
{
"epoch": 0.6949288972363832,
"grad_norm": 3.578125,
"learning_rate": 5.154907928299217e-06,
"loss": 4.449,
"step": 5180
},
{
"epoch": 0.6955996780252214,
"grad_norm": 3.875,
"learning_rate": 5.134436572463572e-06,
"loss": 4.3747,
"step": 5185
},
{
"epoch": 0.6962704588140596,
"grad_norm": 3.6875,
"learning_rate": 5.113991896674667e-06,
"loss": 4.4072,
"step": 5190
},
{
"epoch": 0.6969412396028978,
"grad_norm": 3.65625,
"learning_rate": 5.093574013039749e-06,
"loss": 4.4413,
"step": 5195
},
{
"epoch": 0.697612020391736,
"grad_norm": 3.8125,
"learning_rate": 5.073183033519155e-06,
"loss": 4.4025,
"step": 5200
},
{
"epoch": 0.6982828011805742,
"grad_norm": 3.96875,
"learning_rate": 5.052819069925676e-06,
"loss": 4.4095,
"step": 5205
},
{
"epoch": 0.6989535819694124,
"grad_norm": 3.5,
"learning_rate": 5.032482233923984e-06,
"loss": 4.4403,
"step": 5210
},
{
"epoch": 0.6996243627582506,
"grad_norm": 3.65625,
"learning_rate": 5.012172637029991e-06,
"loss": 4.3786,
"step": 5215
},
{
"epoch": 0.7002951435470888,
"grad_norm": 3.59375,
"learning_rate": 4.991890390610238e-06,
"loss": 4.3771,
"step": 5220
},
{
"epoch": 0.700965924335927,
"grad_norm": 3.578125,
"learning_rate": 4.971635605881291e-06,
"loss": 4.4071,
"step": 5225
},
{
"epoch": 0.7016367051247653,
"grad_norm": 3.984375,
"learning_rate": 4.951408393909152e-06,
"loss": 4.3989,
"step": 5230
},
{
"epoch": 0.7023074859136035,
"grad_norm": 3.796875,
"learning_rate": 4.9312088656086e-06,
"loss": 4.4137,
"step": 5235
},
{
"epoch": 0.7029782667024417,
"grad_norm": 3.609375,
"learning_rate": 4.911037131742637e-06,
"loss": 4.3949,
"step": 5240
},
{
"epoch": 0.7036490474912799,
"grad_norm": 3.65625,
"learning_rate": 4.890893302921837e-06,
"loss": 4.4129,
"step": 5245
},
{
"epoch": 0.7043198282801181,
"grad_norm": 3.78125,
"learning_rate": 4.870777489603768e-06,
"loss": 4.3866,
"step": 5250
},
{
"epoch": 0.7049906090689563,
"grad_norm": 3.375,
"learning_rate": 4.850689802092378e-06,
"loss": 4.4095,
"step": 5255
},
{
"epoch": 0.7056613898577945,
"grad_norm": 3.859375,
"learning_rate": 4.830630350537387e-06,
"loss": 4.4244,
"step": 5260
},
{
"epoch": 0.7063321706466327,
"grad_norm": 3.484375,
"learning_rate": 4.8105992449336695e-06,
"loss": 4.3818,
"step": 5265
},
{
"epoch": 0.7070029514354709,
"grad_norm": 3.640625,
"learning_rate": 4.790596595120699e-06,
"loss": 4.3756,
"step": 5270
},
{
"epoch": 0.7076737322243091,
"grad_norm": 3.75,
"learning_rate": 4.770622510781882e-06,
"loss": 4.3763,
"step": 5275
},
{
"epoch": 0.7083445130131473,
"grad_norm": 3.828125,
"learning_rate": 4.750677101444007e-06,
"loss": 4.3842,
"step": 5280
},
{
"epoch": 0.7090152938019855,
"grad_norm": 3.78125,
"learning_rate": 4.730760476476611e-06,
"loss": 4.3864,
"step": 5285
},
{
"epoch": 0.7096860745908237,
"grad_norm": 3.703125,
"learning_rate": 4.710872745091406e-06,
"loss": 4.3919,
"step": 5290
},
{
"epoch": 0.710356855379662,
"grad_norm": 3.515625,
"learning_rate": 4.691014016341664e-06,
"loss": 4.4083,
"step": 5295
},
{
"epoch": 0.7110276361685002,
"grad_norm": 3.65625,
"learning_rate": 4.671184399121619e-06,
"loss": 4.4256,
"step": 5300
},
{
"epoch": 0.7116984169573384,
"grad_norm": 4.0,
"learning_rate": 4.651384002165865e-06,
"loss": 4.3952,
"step": 5305
},
{
"epoch": 0.7123691977461766,
"grad_norm": 3.640625,
"learning_rate": 4.6316129340487946e-06,
"loss": 4.395,
"step": 5310
},
{
"epoch": 0.7130399785350148,
"grad_norm": 3.765625,
"learning_rate": 4.611871303183952e-06,
"loss": 4.3943,
"step": 5315
},
{
"epoch": 0.713710759323853,
"grad_norm": 3.78125,
"learning_rate": 4.59215921782348e-06,
"loss": 4.3839,
"step": 5320
},
{
"epoch": 0.7143815401126912,
"grad_norm": 3.609375,
"learning_rate": 4.572476786057495e-06,
"loss": 4.3957,
"step": 5325
},
{
"epoch": 0.7150523209015294,
"grad_norm": 3.421875,
"learning_rate": 4.552824115813522e-06,
"loss": 4.4095,
"step": 5330
},
{
"epoch": 0.7157231016903676,
"grad_norm": 3.671875,
"learning_rate": 4.533201314855891e-06,
"loss": 4.4004,
"step": 5335
},
{
"epoch": 0.7163938824792058,
"grad_norm": 3.8125,
"learning_rate": 4.513608490785133e-06,
"loss": 4.4022,
"step": 5340
},
{
"epoch": 0.717064663268044,
"grad_norm": 3.59375,
"learning_rate": 4.494045751037414e-06,
"loss": 4.3979,
"step": 5345
},
{
"epoch": 0.7177354440568822,
"grad_norm": 3.75,
"learning_rate": 4.474513202883934e-06,
"loss": 4.3709,
"step": 5350
},
{
"epoch": 0.7184062248457204,
"grad_norm": 3.890625,
"learning_rate": 4.45501095343033e-06,
"loss": 4.4321,
"step": 5355
},
{
"epoch": 0.7190770056345587,
"grad_norm": 3.6875,
"learning_rate": 4.4355391096161085e-06,
"loss": 4.396,
"step": 5360
},
{
"epoch": 0.7197477864233969,
"grad_norm": 3.8125,
"learning_rate": 4.416097778214037e-06,
"loss": 4.4502,
"step": 5365
},
{
"epoch": 0.7204185672122351,
"grad_norm": 3.671875,
"learning_rate": 4.396687065829579e-06,
"loss": 4.3786,
"step": 5370
},
{
"epoch": 0.7210893480010733,
"grad_norm": 3.703125,
"learning_rate": 4.377307078900303e-06,
"loss": 4.3977,
"step": 5375
},
{
"epoch": 0.7217601287899115,
"grad_norm": 3.5625,
"learning_rate": 4.3579579236952815e-06,
"loss": 4.389,
"step": 5380
},
{
"epoch": 0.7224309095787497,
"grad_norm": 3.828125,
"learning_rate": 4.338639706314537e-06,
"loss": 4.3925,
"step": 5385
},
{
"epoch": 0.7231016903675879,
"grad_norm": 3.9375,
"learning_rate": 4.319352532688444e-06,
"loss": 4.3831,
"step": 5390
},
{
"epoch": 0.7237724711564261,
"grad_norm": 3.671875,
"learning_rate": 4.3000965085771405e-06,
"loss": 4.4045,
"step": 5395
},
{
"epoch": 0.7244432519452643,
"grad_norm": 3.65625,
"learning_rate": 4.280871739569972e-06,
"loss": 4.3704,
"step": 5400
},
{
"epoch": 0.7251140327341025,
"grad_norm": 3.6875,
"learning_rate": 4.261678331084884e-06,
"loss": 4.4079,
"step": 5405
},
{
"epoch": 0.7257848135229407,
"grad_norm": 3.6875,
"learning_rate": 4.242516388367871e-06,
"loss": 4.3972,
"step": 5410
},
{
"epoch": 0.7264555943117789,
"grad_norm": 3.765625,
"learning_rate": 4.223386016492382e-06,
"loss": 4.4009,
"step": 5415
},
{
"epoch": 0.7271263751006172,
"grad_norm": 3.78125,
"learning_rate": 4.204287320358744e-06,
"loss": 4.3831,
"step": 5420
},
{
"epoch": 0.7277971558894554,
"grad_norm": 3.828125,
"learning_rate": 4.185220404693598e-06,
"loss": 4.3798,
"step": 5425
},
{
"epoch": 0.7284679366782936,
"grad_norm": 3.59375,
"learning_rate": 4.166185374049322e-06,
"loss": 4.3876,
"step": 5430
},
{
"epoch": 0.7291387174671318,
"grad_norm": 3.609375,
"learning_rate": 4.1471823328034384e-06,
"loss": 4.4057,
"step": 5435
},
{
"epoch": 0.72980949825597,
"grad_norm": 3.921875,
"learning_rate": 4.1282113851580765e-06,
"loss": 4.411,
"step": 5440
},
{
"epoch": 0.7304802790448082,
"grad_norm": 3.671875,
"learning_rate": 4.109272635139373e-06,
"loss": 4.3814,
"step": 5445
},
{
"epoch": 0.7311510598336464,
"grad_norm": 3.609375,
"learning_rate": 4.090366186596907e-06,
"loss": 4.3982,
"step": 5450
},
{
"epoch": 0.7318218406224846,
"grad_norm": 3.71875,
"learning_rate": 4.0714921432031455e-06,
"loss": 4.4104,
"step": 5455
},
{
"epoch": 0.7324926214113228,
"grad_norm": 3.828125,
"learning_rate": 4.052650608452848e-06,
"loss": 4.3788,
"step": 5460
},
{
"epoch": 0.733163402200161,
"grad_norm": 3.703125,
"learning_rate": 4.033841685662529e-06,
"loss": 4.4357,
"step": 5465
},
{
"epoch": 0.7338341829889992,
"grad_norm": 3.65625,
"learning_rate": 4.015065477969875e-06,
"loss": 4.4251,
"step": 5470
},
{
"epoch": 0.7345049637778374,
"grad_norm": 3.828125,
"learning_rate": 3.996322088333174e-06,
"loss": 4.4003,
"step": 5475
},
{
"epoch": 0.7351757445666756,
"grad_norm": 3.78125,
"learning_rate": 3.977611619530755e-06,
"loss": 4.3826,
"step": 5480
},
{
"epoch": 0.7358465253555139,
"grad_norm": 3.984375,
"learning_rate": 3.958934174160446e-06,
"loss": 4.4137,
"step": 5485
},
{
"epoch": 0.7365173061443521,
"grad_norm": 3.78125,
"learning_rate": 3.940289854638974e-06,
"loss": 4.3749,
"step": 5490
},
{
"epoch": 0.7371880869331903,
"grad_norm": 3.71875,
"learning_rate": 3.921678763201434e-06,
"loss": 4.4361,
"step": 5495
},
{
"epoch": 0.7378588677220285,
"grad_norm": 3.515625,
"learning_rate": 3.903101001900706e-06,
"loss": 4.4128,
"step": 5500
},
{
"epoch": 0.7385296485108667,
"grad_norm": 3.4375,
"learning_rate": 3.884556672606914e-06,
"loss": 4.3806,
"step": 5505
},
{
"epoch": 0.7392004292997049,
"grad_norm": 3.796875,
"learning_rate": 3.866045877006863e-06,
"loss": 4.3959,
"step": 5510
},
{
"epoch": 0.7398712100885431,
"grad_norm": 3.796875,
"learning_rate": 3.8475687166034684e-06,
"loss": 4.3771,
"step": 5515
},
{
"epoch": 0.7405419908773813,
"grad_norm": 3.796875,
"learning_rate": 3.8291252927152034e-06,
"loss": 4.3975,
"step": 5520
},
{
"epoch": 0.7412127716662195,
"grad_norm": 3.921875,
"learning_rate": 3.810715706475575e-06,
"loss": 4.3581,
"step": 5525
},
{
"epoch": 0.7418835524550577,
"grad_norm": 3.734375,
"learning_rate": 3.7923400588325156e-06,
"loss": 4.3414,
"step": 5530
},
{
"epoch": 0.7425543332438959,
"grad_norm": 3.671875,
"learning_rate": 3.7739984505478743e-06,
"loss": 4.3786,
"step": 5535
},
{
"epoch": 0.7432251140327341,
"grad_norm": 3.859375,
"learning_rate": 3.7556909821968335e-06,
"loss": 4.4311,
"step": 5540
},
{
"epoch": 0.7438958948215723,
"grad_norm": 3.90625,
"learning_rate": 3.737417754167385e-06,
"loss": 4.3919,
"step": 5545
},
{
"epoch": 0.7445666756104106,
"grad_norm": 3.640625,
"learning_rate": 3.7191788666597604e-06,
"loss": 4.4056,
"step": 5550
},
{
"epoch": 0.7452374563992488,
"grad_norm": 3.84375,
"learning_rate": 3.7009744196858797e-06,
"loss": 4.4135,
"step": 5555
},
{
"epoch": 0.745908237188087,
"grad_norm": 3.8125,
"learning_rate": 3.6828045130688205e-06,
"loss": 4.4051,
"step": 5560
},
{
"epoch": 0.7465790179769252,
"grad_norm": 3.484375,
"learning_rate": 3.6646692464422608e-06,
"loss": 4.428,
"step": 5565
},
{
"epoch": 0.7472497987657634,
"grad_norm": 3.671875,
"learning_rate": 3.6465687192499176e-06,
"loss": 4.4121,
"step": 5570
},
{
"epoch": 0.7479205795546016,
"grad_norm": 3.734375,
"learning_rate": 3.6285030307450365e-06,
"loss": 4.3923,
"step": 5575
},
{
"epoch": 0.7485913603434398,
"grad_norm": 3.390625,
"learning_rate": 3.6104722799898075e-06,
"loss": 4.4248,
"step": 5580
},
{
"epoch": 0.749262141132278,
"grad_norm": 3.859375,
"learning_rate": 3.5924765658548544e-06,
"loss": 4.4009,
"step": 5585
},
{
"epoch": 0.7499329219211162,
"grad_norm": 3.8125,
"learning_rate": 3.574515987018681e-06,
"loss": 4.3861,
"step": 5590
},
{
"epoch": 0.7506037027099544,
"grad_norm": 3.6875,
"learning_rate": 3.556590641967115e-06,
"loss": 4.408,
"step": 5595
},
{
"epoch": 0.7512744834987926,
"grad_norm": 3.546875,
"learning_rate": 3.538700628992794e-06,
"loss": 4.4252,
"step": 5600
},
{
"epoch": 0.7519452642876308,
"grad_norm": 3.796875,
"learning_rate": 3.5208460461946136e-06,
"loss": 4.3986,
"step": 5605
},
{
"epoch": 0.7526160450764691,
"grad_norm": 3.8125,
"learning_rate": 3.503026991477181e-06,
"loss": 4.3918,
"step": 5610
},
{
"epoch": 0.7532868258653073,
"grad_norm": 3.890625,
"learning_rate": 3.485243562550297e-06,
"loss": 4.4013,
"step": 5615
},
{
"epoch": 0.7539576066541455,
"grad_norm": 3.6875,
"learning_rate": 3.4674958569284024e-06,
"loss": 4.4146,
"step": 5620
},
{
"epoch": 0.7546283874429837,
"grad_norm": 3.71875,
"learning_rate": 3.4497839719300553e-06,
"loss": 4.3925,
"step": 5625
},
{
"epoch": 0.7552991682318219,
"grad_norm": 3.671875,
"learning_rate": 3.4321080046773993e-06,
"loss": 4.3892,
"step": 5630
},
{
"epoch": 0.7559699490206601,
"grad_norm": 3.609375,
"learning_rate": 3.4144680520956095e-06,
"loss": 4.4124,
"step": 5635
},
{
"epoch": 0.7566407298094983,
"grad_norm": 3.84375,
"learning_rate": 3.3968642109123917e-06,
"loss": 4.3954,
"step": 5640
},
{
"epoch": 0.7573115105983365,
"grad_norm": 3.71875,
"learning_rate": 3.3792965776574337e-06,
"loss": 4.4232,
"step": 5645
},
{
"epoch": 0.7579822913871747,
"grad_norm": 3.65625,
"learning_rate": 3.361765248661868e-06,
"loss": 4.4099,
"step": 5650
},
{
"epoch": 0.7586530721760129,
"grad_norm": 3.75,
"learning_rate": 3.3442703200577687e-06,
"loss": 4.374,
"step": 5655
},
{
"epoch": 0.759323852964851,
"grad_norm": 3.625,
"learning_rate": 3.326811887777607e-06,
"loss": 4.3952,
"step": 5660
},
{
"epoch": 0.7599946337536893,
"grad_norm": 3.59375,
"learning_rate": 3.30939004755372e-06,
"loss": 4.4108,
"step": 5665
},
{
"epoch": 0.7606654145425275,
"grad_norm": 3.65625,
"learning_rate": 3.2920048949178074e-06,
"loss": 4.3965,
"step": 5670
},
{
"epoch": 0.7613361953313658,
"grad_norm": 3.796875,
"learning_rate": 3.2746565252003816e-06,
"loss": 4.4189,
"step": 5675
},
{
"epoch": 0.762006976120204,
"grad_norm": 3.71875,
"learning_rate": 3.257345033530267e-06,
"loss": 4.4313,
"step": 5680
},
{
"epoch": 0.7626777569090422,
"grad_norm": 3.546875,
"learning_rate": 3.2400705148340696e-06,
"loss": 4.4077,
"step": 5685
},
{
"epoch": 0.7633485376978804,
"grad_norm": 3.671875,
"learning_rate": 3.2228330638356463e-06,
"loss": 4.4143,
"step": 5690
},
{
"epoch": 0.7640193184867186,
"grad_norm": 3.796875,
"learning_rate": 3.205632775055606e-06,
"loss": 4.3684,
"step": 5695
},
{
"epoch": 0.7646900992755568,
"grad_norm": 3.703125,
"learning_rate": 3.1884697428107793e-06,
"loss": 4.393,
"step": 5700
},
{
"epoch": 0.765360880064395,
"grad_norm": 3.484375,
"learning_rate": 3.1713440612136925e-06,
"loss": 4.3604,
"step": 5705
},
{
"epoch": 0.7660316608532332,
"grad_norm": 3.609375,
"learning_rate": 3.1542558241720776e-06,
"loss": 4.3907,
"step": 5710
},
{
"epoch": 0.7667024416420714,
"grad_norm": 3.953125,
"learning_rate": 3.137205125388326e-06,
"loss": 4.3713,
"step": 5715
},
{
"epoch": 0.7673732224309096,
"grad_norm": 3.71875,
"learning_rate": 3.1201920583590005e-06,
"loss": 4.3858,
"step": 5720
},
{
"epoch": 0.7680440032197478,
"grad_norm": 3.6875,
"learning_rate": 3.103216716374312e-06,
"loss": 4.411,
"step": 5725
},
{
"epoch": 0.768714784008586,
"grad_norm": 4.0,
"learning_rate": 3.0862791925175984e-06,
"loss": 4.4053,
"step": 5730
},
{
"epoch": 0.7693855647974241,
"grad_norm": 3.78125,
"learning_rate": 3.0693795796648352e-06,
"loss": 4.3741,
"step": 5735
},
{
"epoch": 0.7700563455862625,
"grad_norm": 3.84375,
"learning_rate": 3.0525179704841113e-06,
"loss": 4.4025,
"step": 5740
},
{
"epoch": 0.7707271263751007,
"grad_norm": 3.875,
"learning_rate": 3.03569445743512e-06,
"loss": 4.3617,
"step": 5745
},
{
"epoch": 0.7713979071639389,
"grad_norm": 3.703125,
"learning_rate": 3.018909132768666e-06,
"loss": 4.4312,
"step": 5750
},
{
"epoch": 0.772068687952777,
"grad_norm": 3.765625,
"learning_rate": 3.0021620885261383e-06,
"loss": 4.4055,
"step": 5755
},
{
"epoch": 0.7727394687416153,
"grad_norm": 4.0625,
"learning_rate": 2.985453416539026e-06,
"loss": 4.4152,
"step": 5760
},
{
"epoch": 0.7734102495304535,
"grad_norm": 3.484375,
"learning_rate": 2.9687832084284084e-06,
"loss": 4.4156,
"step": 5765
},
{
"epoch": 0.7740810303192917,
"grad_norm": 3.578125,
"learning_rate": 2.9521515556044376e-06,
"loss": 4.3955,
"step": 5770
},
{
"epoch": 0.7747518111081299,
"grad_norm": 4.0,
"learning_rate": 2.9355585492658622e-06,
"loss": 4.3971,
"step": 5775
},
{
"epoch": 0.775422591896968,
"grad_norm": 3.859375,
"learning_rate": 2.9190042803995135e-06,
"loss": 4.3765,
"step": 5780
},
{
"epoch": 0.7760933726858062,
"grad_norm": 3.6875,
"learning_rate": 2.902488839779798e-06,
"loss": 4.4271,
"step": 5785
},
{
"epoch": 0.7767641534746444,
"grad_norm": 3.46875,
"learning_rate": 2.8860123179682244e-06,
"loss": 4.372,
"step": 5790
},
{
"epoch": 0.7774349342634826,
"grad_norm": 3.8125,
"learning_rate": 2.869574805312877e-06,
"loss": 4.4268,
"step": 5795
},
{
"epoch": 0.778105715052321,
"grad_norm": 3.6875,
"learning_rate": 2.8531763919479473e-06,
"loss": 4.3664,
"step": 5800
},
{
"epoch": 0.7787764958411592,
"grad_norm": 3.71875,
"learning_rate": 2.8368171677932267e-06,
"loss": 4.3766,
"step": 5805
},
{
"epoch": 0.7794472766299974,
"grad_norm": 3.59375,
"learning_rate": 2.8204972225536066e-06,
"loss": 4.4061,
"step": 5810
},
{
"epoch": 0.7801180574188356,
"grad_norm": 3.4375,
"learning_rate": 2.8042166457186027e-06,
"loss": 4.4081,
"step": 5815
},
{
"epoch": 0.7807888382076738,
"grad_norm": 3.765625,
"learning_rate": 2.7879755265618558e-06,
"loss": 4.3727,
"step": 5820
},
{
"epoch": 0.781459618996512,
"grad_norm": 3.703125,
"learning_rate": 2.7717739541406363e-06,
"loss": 4.3778,
"step": 5825
},
{
"epoch": 0.7821303997853501,
"grad_norm": 3.65625,
"learning_rate": 2.755612017295367e-06,
"loss": 4.3904,
"step": 5830
},
{
"epoch": 0.7828011805741883,
"grad_norm": 3.609375,
"learning_rate": 2.739489804649136e-06,
"loss": 4.3754,
"step": 5835
},
{
"epoch": 0.7834719613630265,
"grad_norm": 3.625,
"learning_rate": 2.7234074046071933e-06,
"loss": 4.4107,
"step": 5840
},
{
"epoch": 0.7841427421518647,
"grad_norm": 3.78125,
"learning_rate": 2.707364905356492e-06,
"loss": 4.4124,
"step": 5845
},
{
"epoch": 0.784813522940703,
"grad_norm": 3.640625,
"learning_rate": 2.69136239486518e-06,
"loss": 4.3771,
"step": 5850
},
{
"epoch": 0.7854843037295411,
"grad_norm": 3.875,
"learning_rate": 2.675399960882138e-06,
"loss": 4.3464,
"step": 5855
},
{
"epoch": 0.7861550845183793,
"grad_norm": 3.296875,
"learning_rate": 2.6594776909364884e-06,
"loss": 4.4236,
"step": 5860
},
{
"epoch": 0.7868258653072177,
"grad_norm": 3.359375,
"learning_rate": 2.643595672337108e-06,
"loss": 4.3676,
"step": 5865
},
{
"epoch": 0.7874966460960559,
"grad_norm": 3.734375,
"learning_rate": 2.6277539921721663e-06,
"loss": 4.3905,
"step": 5870
},
{
"epoch": 0.788167426884894,
"grad_norm": 3.921875,
"learning_rate": 2.6119527373086386e-06,
"loss": 4.3828,
"step": 5875
},
{
"epoch": 0.7888382076737323,
"grad_norm": 3.9375,
"learning_rate": 2.5961919943918237e-06,
"loss": 4.395,
"step": 5880
},
{
"epoch": 0.7895089884625704,
"grad_norm": 3.953125,
"learning_rate": 2.5804718498448844e-06,
"loss": 4.3578,
"step": 5885
},
{
"epoch": 0.7901797692514086,
"grad_norm": 3.96875,
"learning_rate": 2.564792389868356e-06,
"loss": 4.426,
"step": 5890
},
{
"epoch": 0.7908505500402468,
"grad_norm": 3.8125,
"learning_rate": 2.549153700439686e-06,
"loss": 4.4104,
"step": 5895
},
{
"epoch": 0.791521330829085,
"grad_norm": 3.828125,
"learning_rate": 2.5335558673127647e-06,
"loss": 4.4126,
"step": 5900
},
{
"epoch": 0.7921921116179232,
"grad_norm": 3.71875,
"learning_rate": 2.517998976017436e-06,
"loss": 4.4048,
"step": 5905
},
{
"epoch": 0.7928628924067614,
"grad_norm": 3.515625,
"learning_rate": 2.502483111859053e-06,
"loss": 4.3782,
"step": 5910
},
{
"epoch": 0.7935336731955996,
"grad_norm": 3.734375,
"learning_rate": 2.4870083599179974e-06,
"loss": 4.3431,
"step": 5915
},
{
"epoch": 0.7942044539844378,
"grad_norm": 3.78125,
"learning_rate": 2.471574805049206e-06,
"loss": 4.3973,
"step": 5920
},
{
"epoch": 0.794875234773276,
"grad_norm": 3.984375,
"learning_rate": 2.4561825318817254e-06,
"loss": 4.365,
"step": 5925
},
{
"epoch": 0.7955460155621144,
"grad_norm": 3.8125,
"learning_rate": 2.440831624818224e-06,
"loss": 4.3674,
"step": 5930
},
{
"epoch": 0.7962167963509525,
"grad_norm": 3.484375,
"learning_rate": 2.425522168034552e-06,
"loss": 4.3753,
"step": 5935
},
{
"epoch": 0.7968875771397907,
"grad_norm": 3.96875,
"learning_rate": 2.4102542454792665e-06,
"loss": 4.3955,
"step": 5940
},
{
"epoch": 0.797558357928629,
"grad_norm": 3.625,
"learning_rate": 2.3950279408731682e-06,
"loss": 4.4208,
"step": 5945
},
{
"epoch": 0.7982291387174671,
"grad_norm": 3.578125,
"learning_rate": 2.3798433377088535e-06,
"loss": 4.3843,
"step": 5950
},
{
"epoch": 0.7988999195063053,
"grad_norm": 3.5625,
"learning_rate": 2.364700519250255e-06,
"loss": 4.4331,
"step": 5955
},
{
"epoch": 0.7995707002951435,
"grad_norm": 3.953125,
"learning_rate": 2.3495995685321705e-06,
"loss": 4.4346,
"step": 5960
},
{
"epoch": 0.8002414810839817,
"grad_norm": 3.65625,
"learning_rate": 2.334540568359829e-06,
"loss": 4.4483,
"step": 5965
},
{
"epoch": 0.8009122618728199,
"grad_norm": 3.765625,
"learning_rate": 2.3195236013084156e-06,
"loss": 4.4429,
"step": 5970
},
{
"epoch": 0.8015830426616581,
"grad_norm": 3.578125,
"learning_rate": 2.304548749722639e-06,
"loss": 4.399,
"step": 5975
},
{
"epoch": 0.8022538234504963,
"grad_norm": 4.09375,
"learning_rate": 2.289616095716268e-06,
"loss": 4.3782,
"step": 5980
},
{
"epoch": 0.8029246042393345,
"grad_norm": 3.6875,
"learning_rate": 2.2747257211716754e-06,
"loss": 4.4022,
"step": 5985
},
{
"epoch": 0.8035953850281728,
"grad_norm": 3.484375,
"learning_rate": 2.2598777077394063e-06,
"loss": 4.4023,
"step": 5990
},
{
"epoch": 0.804266165817011,
"grad_norm": 3.46875,
"learning_rate": 2.2450721368377194e-06,
"loss": 4.4303,
"step": 5995
},
{
"epoch": 0.8049369466058492,
"grad_norm": 3.875,
"learning_rate": 2.2303090896521352e-06,
"loss": 4.4109,
"step": 6000
},
{
"epoch": 0.8056077273946874,
"grad_norm": 3.84375,
"learning_rate": 2.2155886471350073e-06,
"loss": 4.3667,
"step": 6005
},
{
"epoch": 0.8062785081835256,
"grad_norm": 3.953125,
"learning_rate": 2.2009108900050613e-06,
"loss": 4.3359,
"step": 6010
},
{
"epoch": 0.8069492889723638,
"grad_norm": 3.65625,
"learning_rate": 2.1862758987469634e-06,
"loss": 4.3656,
"step": 6015
},
{
"epoch": 0.807620069761202,
"grad_norm": 3.875,
"learning_rate": 2.171683753610878e-06,
"loss": 4.4017,
"step": 6020
},
{
"epoch": 0.8082908505500402,
"grad_norm": 3.640625,
"learning_rate": 2.1571345346120188e-06,
"loss": 4.4229,
"step": 6025
},
{
"epoch": 0.8089616313388784,
"grad_norm": 3.90625,
"learning_rate": 2.1426283215302204e-06,
"loss": 4.3755,
"step": 6030
},
{
"epoch": 0.8096324121277166,
"grad_norm": 3.921875,
"learning_rate": 2.1281651939094996e-06,
"loss": 4.3852,
"step": 6035
},
{
"epoch": 0.8103031929165548,
"grad_norm": 3.640625,
"learning_rate": 2.113745231057608e-06,
"loss": 4.384,
"step": 6040
},
{
"epoch": 0.810973973705393,
"grad_norm": 3.71875,
"learning_rate": 2.0993685120456143e-06,
"loss": 4.3756,
"step": 6045
},
{
"epoch": 0.8116447544942312,
"grad_norm": 3.6875,
"learning_rate": 2.08503511570746e-06,
"loss": 4.3656,
"step": 6050
},
{
"epoch": 0.8123155352830695,
"grad_norm": 4.09375,
"learning_rate": 2.0707451206395223e-06,
"loss": 4.4327,
"step": 6055
},
{
"epoch": 0.8129863160719077,
"grad_norm": 3.703125,
"learning_rate": 2.0564986052002022e-06,
"loss": 4.3937,
"step": 6060
},
{
"epoch": 0.8136570968607459,
"grad_norm": 4.0625,
"learning_rate": 2.0422956475094678e-06,
"loss": 4.4104,
"step": 6065
},
{
"epoch": 0.8143278776495841,
"grad_norm": 3.609375,
"learning_rate": 2.028136325448453e-06,
"loss": 4.4209,
"step": 6070
},
{
"epoch": 0.8149986584384223,
"grad_norm": 3.71875,
"learning_rate": 2.0140207166590165e-06,
"loss": 4.4011,
"step": 6075
},
{
"epoch": 0.8156694392272605,
"grad_norm": 3.53125,
"learning_rate": 1.999948898543308e-06,
"loss": 4.4007,
"step": 6080
},
{
"epoch": 0.8163402200160987,
"grad_norm": 3.8125,
"learning_rate": 1.9859209482633634e-06,
"loss": 4.3775,
"step": 6085
},
{
"epoch": 0.8170110008049369,
"grad_norm": 3.703125,
"learning_rate": 1.9719369427406732e-06,
"loss": 4.4116,
"step": 6090
},
{
"epoch": 0.8176817815937751,
"grad_norm": 3.703125,
"learning_rate": 1.957996958655748e-06,
"loss": 4.433,
"step": 6095
},
{
"epoch": 0.8183525623826133,
"grad_norm": 3.65625,
"learning_rate": 1.9441010724477207e-06,
"loss": 4.4168,
"step": 6100
},
{
"epoch": 0.8190233431714515,
"grad_norm": 3.71875,
"learning_rate": 1.930249360313907e-06,
"loss": 4.4494,
"step": 6105
},
{
"epoch": 0.8196941239602897,
"grad_norm": 3.515625,
"learning_rate": 1.9164418982094036e-06,
"loss": 4.4221,
"step": 6110
},
{
"epoch": 0.8203649047491279,
"grad_norm": 3.609375,
"learning_rate": 1.9026787618466646e-06,
"loss": 4.4614,
"step": 6115
},
{
"epoch": 0.8210356855379662,
"grad_norm": 3.625,
"learning_rate": 1.888960026695078e-06,
"loss": 4.3725,
"step": 6120
},
{
"epoch": 0.8217064663268044,
"grad_norm": 3.5,
"learning_rate": 1.8752857679805681e-06,
"loss": 4.4424,
"step": 6125
},
{
"epoch": 0.8223772471156426,
"grad_norm": 3.828125,
"learning_rate": 1.8616560606851763e-06,
"loss": 4.4192,
"step": 6130
},
{
"epoch": 0.8230480279044808,
"grad_norm": 3.578125,
"learning_rate": 1.8480709795466401e-06,
"loss": 4.387,
"step": 6135
},
{
"epoch": 0.823718808693319,
"grad_norm": 3.75,
"learning_rate": 1.834530599058001e-06,
"loss": 4.3954,
"step": 6140
},
{
"epoch": 0.8243895894821572,
"grad_norm": 3.484375,
"learning_rate": 1.82103499346718e-06,
"loss": 4.3832,
"step": 6145
},
{
"epoch": 0.8250603702709954,
"grad_norm": 3.65625,
"learning_rate": 1.8075842367765817e-06,
"loss": 4.397,
"step": 6150
},
{
"epoch": 0.8257311510598336,
"grad_norm": 3.84375,
"learning_rate": 1.7941784027426856e-06,
"loss": 4.3856,
"step": 6155
},
{
"epoch": 0.8264019318486718,
"grad_norm": 3.625,
"learning_rate": 1.7808175648756342e-06,
"loss": 4.3776,
"step": 6160
},
{
"epoch": 0.82707271263751,
"grad_norm": 3.75,
"learning_rate": 1.7675017964388408e-06,
"loss": 4.4442,
"step": 6165
},
{
"epoch": 0.8277434934263482,
"grad_norm": 3.75,
"learning_rate": 1.754231170448586e-06,
"loss": 4.4072,
"step": 6170
},
{
"epoch": 0.8284142742151864,
"grad_norm": 3.4375,
"learning_rate": 1.741005759673603e-06,
"loss": 4.4095,
"step": 6175
},
{
"epoch": 0.8290850550040247,
"grad_norm": 3.625,
"learning_rate": 1.7278256366347034e-06,
"loss": 4.3951,
"step": 6180
},
{
"epoch": 0.8297558357928629,
"grad_norm": 3.703125,
"learning_rate": 1.7146908736043543e-06,
"loss": 4.4436,
"step": 6185
},
{
"epoch": 0.8304266165817011,
"grad_norm": 3.609375,
"learning_rate": 1.7016015426063016e-06,
"loss": 4.4158,
"step": 6190
},
{
"epoch": 0.8310973973705393,
"grad_norm": 3.359375,
"learning_rate": 1.6885577154151667e-06,
"loss": 4.3742,
"step": 6195
},
{
"epoch": 0.8317681781593775,
"grad_norm": 3.8125,
"learning_rate": 1.6755594635560468e-06,
"loss": 4.4058,
"step": 6200
},
{
"epoch": 0.8324389589482157,
"grad_norm": 3.640625,
"learning_rate": 1.6626068583041366e-06,
"loss": 4.368,
"step": 6205
},
{
"epoch": 0.8331097397370539,
"grad_norm": 3.8125,
"learning_rate": 1.6496999706843297e-06,
"loss": 4.3562,
"step": 6210
},
{
"epoch": 0.8337805205258921,
"grad_norm": 3.5625,
"learning_rate": 1.6368388714708229e-06,
"loss": 4.373,
"step": 6215
},
{
"epoch": 0.8344513013147303,
"grad_norm": 4.0625,
"learning_rate": 1.6240236311867419e-06,
"loss": 4.3973,
"step": 6220
},
{
"epoch": 0.8351220821035685,
"grad_norm": 3.6875,
"learning_rate": 1.6112543201037434e-06,
"loss": 4.3796,
"step": 6225
},
{
"epoch": 0.8357928628924067,
"grad_norm": 3.671875,
"learning_rate": 1.598531008241635e-06,
"loss": 4.3776,
"step": 6230
},
{
"epoch": 0.8364636436812449,
"grad_norm": 3.796875,
"learning_rate": 1.5858537653679929e-06,
"loss": 4.3906,
"step": 6235
},
{
"epoch": 0.8371344244700831,
"grad_norm": 3.71875,
"learning_rate": 1.5732226609977718e-06,
"loss": 4.3952,
"step": 6240
},
{
"epoch": 0.8378052052589214,
"grad_norm": 3.609375,
"learning_rate": 1.5606377643929305e-06,
"loss": 4.3985,
"step": 6245
},
{
"epoch": 0.8384759860477596,
"grad_norm": 3.53125,
"learning_rate": 1.5480991445620541e-06,
"loss": 4.3837,
"step": 6250
},
{
"epoch": 0.8391467668365978,
"grad_norm": 3.65625,
"learning_rate": 1.5356068702599626e-06,
"loss": 4.3504,
"step": 6255
},
{
"epoch": 0.839817547625436,
"grad_norm": 3.59375,
"learning_rate": 1.5231610099873505e-06,
"loss": 4.4061,
"step": 6260
},
{
"epoch": 0.8404883284142742,
"grad_norm": 3.828125,
"learning_rate": 1.510761631990404e-06,
"loss": 4.3666,
"step": 6265
},
{
"epoch": 0.8411591092031124,
"grad_norm": 3.90625,
"learning_rate": 1.4984088042604183e-06,
"loss": 4.3922,
"step": 6270
},
{
"epoch": 0.8418298899919506,
"grad_norm": 3.71875,
"learning_rate": 1.4861025945334429e-06,
"loss": 4.3993,
"step": 6275
},
{
"epoch": 0.8425006707807888,
"grad_norm": 3.6875,
"learning_rate": 1.4738430702898898e-06,
"loss": 4.3985,
"step": 6280
},
{
"epoch": 0.843171451569627,
"grad_norm": 3.828125,
"learning_rate": 1.461630298754183e-06,
"loss": 4.3809,
"step": 6285
},
{
"epoch": 0.8438422323584652,
"grad_norm": 3.4375,
"learning_rate": 1.4494643468943782e-06,
"loss": 4.4072,
"step": 6290
},
{
"epoch": 0.8445130131473034,
"grad_norm": 3.671875,
"learning_rate": 1.4373452814217902e-06,
"loss": 4.4207,
"step": 6295
},
{
"epoch": 0.8451837939361416,
"grad_norm": 3.78125,
"learning_rate": 1.4252731687906462e-06,
"loss": 4.3916,
"step": 6300
},
{
"epoch": 0.8458545747249798,
"grad_norm": 3.75,
"learning_rate": 1.4132480751977073e-06,
"loss": 4.378,
"step": 6305
},
{
"epoch": 0.8465253555138181,
"grad_norm": 3.453125,
"learning_rate": 1.401270066581899e-06,
"loss": 4.3816,
"step": 6310
},
{
"epoch": 0.8471961363026563,
"grad_norm": 3.578125,
"learning_rate": 1.3893392086239733e-06,
"loss": 4.3827,
"step": 6315
},
{
"epoch": 0.8478669170914945,
"grad_norm": 3.671875,
"learning_rate": 1.377455566746121e-06,
"loss": 4.3736,
"step": 6320
},
{
"epoch": 0.8485376978803327,
"grad_norm": 3.71875,
"learning_rate": 1.3656192061116359e-06,
"loss": 4.4332,
"step": 6325
},
{
"epoch": 0.8492084786691709,
"grad_norm": 3.65625,
"learning_rate": 1.3538301916245432e-06,
"loss": 4.3595,
"step": 6330
},
{
"epoch": 0.8498792594580091,
"grad_norm": 4.03125,
"learning_rate": 1.3420885879292477e-06,
"loss": 4.4125,
"step": 6335
},
{
"epoch": 0.8505500402468473,
"grad_norm": 3.921875,
"learning_rate": 1.3303944594101813e-06,
"loss": 4.4203,
"step": 6340
},
{
"epoch": 0.8512208210356855,
"grad_norm": 3.890625,
"learning_rate": 1.3187478701914514e-06,
"loss": 4.3966,
"step": 6345
},
{
"epoch": 0.8518916018245237,
"grad_norm": 3.546875,
"learning_rate": 1.3071488841364776e-06,
"loss": 4.3657,
"step": 6350
},
{
"epoch": 0.8525623826133619,
"grad_norm": 3.75,
"learning_rate": 1.2955975648476593e-06,
"loss": 4.3691,
"step": 6355
},
{
"epoch": 0.8532331634022001,
"grad_norm": 3.84375,
"learning_rate": 1.284093975666012e-06,
"loss": 4.3664,
"step": 6360
},
{
"epoch": 0.8539039441910383,
"grad_norm": 3.75,
"learning_rate": 1.2726381796708287e-06,
"loss": 4.4085,
"step": 6365
},
{
"epoch": 0.8545747249798765,
"grad_norm": 3.84375,
"learning_rate": 1.2612302396793319e-06,
"loss": 4.4047,
"step": 6370
},
{
"epoch": 0.8552455057687148,
"grad_norm": 3.609375,
"learning_rate": 1.249870218246323e-06,
"loss": 4.4046,
"step": 6375
},
{
"epoch": 0.855916286557553,
"grad_norm": 3.734375,
"learning_rate": 1.2385581776638512e-06,
"loss": 4.4097,
"step": 6380
},
{
"epoch": 0.8565870673463912,
"grad_norm": 3.984375,
"learning_rate": 1.2272941799608628e-06,
"loss": 4.3767,
"step": 6385
},
{
"epoch": 0.8572578481352294,
"grad_norm": 3.953125,
"learning_rate": 1.216078286902862e-06,
"loss": 4.3908,
"step": 6390
},
{
"epoch": 0.8579286289240676,
"grad_norm": 3.640625,
"learning_rate": 1.2049105599915768e-06,
"loss": 4.3875,
"step": 6395
},
{
"epoch": 0.8585994097129058,
"grad_norm": 3.671875,
"learning_rate": 1.1937910604646152e-06,
"loss": 4.4117,
"step": 6400
},
{
"epoch": 0.859270190501744,
"grad_norm": 3.984375,
"learning_rate": 1.1827198492951353e-06,
"loss": 4.3847,
"step": 6405
},
{
"epoch": 0.8599409712905822,
"grad_norm": 3.671875,
"learning_rate": 1.171696987191513e-06,
"loss": 4.4047,
"step": 6410
},
{
"epoch": 0.8606117520794204,
"grad_norm": 3.75,
"learning_rate": 1.160722534596993e-06,
"loss": 4.3749,
"step": 6415
},
{
"epoch": 0.8612825328682586,
"grad_norm": 3.796875,
"learning_rate": 1.1497965516893816e-06,
"loss": 4.3845,
"step": 6420
},
{
"epoch": 0.8619533136570968,
"grad_norm": 3.828125,
"learning_rate": 1.1389190983806986e-06,
"loss": 4.4241,
"step": 6425
},
{
"epoch": 0.862624094445935,
"grad_norm": 3.71875,
"learning_rate": 1.1280902343168521e-06,
"loss": 4.3735,
"step": 6430
},
{
"epoch": 0.8632948752347733,
"grad_norm": 3.53125,
"learning_rate": 1.11731001887732e-06,
"loss": 4.395,
"step": 6435
},
{
"epoch": 0.8639656560236115,
"grad_norm": 3.9375,
"learning_rate": 1.1065785111748117e-06,
"loss": 4.4219,
"step": 6440
},
{
"epoch": 0.8646364368124497,
"grad_norm": 3.859375,
"learning_rate": 1.0958957700549543e-06,
"loss": 4.4098,
"step": 6445
},
{
"epoch": 0.8653072176012879,
"grad_norm": 3.546875,
"learning_rate": 1.0852618540959692e-06,
"loss": 4.3588,
"step": 6450
},
{
"epoch": 0.8659779983901261,
"grad_norm": 3.9375,
"learning_rate": 1.0746768216083369e-06,
"loss": 4.3495,
"step": 6455
},
{
"epoch": 0.8666487791789643,
"grad_norm": 3.640625,
"learning_rate": 1.0641407306344997e-06,
"loss": 4.3889,
"step": 6460
},
{
"epoch": 0.8673195599678025,
"grad_norm": 3.84375,
"learning_rate": 1.0536536389485275e-06,
"loss": 4.4105,
"step": 6465
},
{
"epoch": 0.8679903407566407,
"grad_norm": 3.8125,
"learning_rate": 1.0432156040558018e-06,
"loss": 4.4021,
"step": 6470
},
{
"epoch": 0.8686611215454789,
"grad_norm": 3.703125,
"learning_rate": 1.032826683192708e-06,
"loss": 4.3493,
"step": 6475
},
{
"epoch": 0.8693319023343171,
"grad_norm": 3.859375,
"learning_rate": 1.0224869333263187e-06,
"loss": 4.3967,
"step": 6480
},
{
"epoch": 0.8700026831231553,
"grad_norm": 3.828125,
"learning_rate": 1.012196411154075e-06,
"loss": 4.3875,
"step": 6485
},
{
"epoch": 0.8706734639119935,
"grad_norm": 3.828125,
"learning_rate": 1.0019551731034871e-06,
"loss": 4.4063,
"step": 6490
},
{
"epoch": 0.8713442447008317,
"grad_norm": 3.609375,
"learning_rate": 9.917632753318095e-07,
"loss": 4.4293,
"step": 6495
},
{
"epoch": 0.87201502548967,
"grad_norm": 3.875,
"learning_rate": 9.81620773725751e-07,
"loss": 4.3773,
"step": 6500
},
{
"epoch": 0.8726858062785082,
"grad_norm": 3.671875,
"learning_rate": 9.715277239011577e-07,
"loss": 4.3746,
"step": 6505
},
{
"epoch": 0.8733565870673464,
"grad_norm": 3.75,
"learning_rate": 9.614841812027032e-07,
"loss": 4.4339,
"step": 6510
},
{
"epoch": 0.8740273678561846,
"grad_norm": 3.453125,
"learning_rate": 9.514902007035964e-07,
"loss": 4.4179,
"step": 6515
},
{
"epoch": 0.8746981486450228,
"grad_norm": 3.359375,
"learning_rate": 9.415458372052788e-07,
"loss": 4.3804,
"step": 6520
},
{
"epoch": 0.875368929433861,
"grad_norm": 3.84375,
"learning_rate": 9.316511452371091e-07,
"loss": 4.417,
"step": 6525
},
{
"epoch": 0.8760397102226992,
"grad_norm": 3.640625,
"learning_rate": 9.218061790560862e-07,
"loss": 4.3884,
"step": 6530
},
{
"epoch": 0.8767104910115374,
"grad_norm": 3.59375,
"learning_rate": 9.120109926465304e-07,
"loss": 4.428,
"step": 6535
},
{
"epoch": 0.8773812718003756,
"grad_norm": 3.703125,
"learning_rate": 9.022656397198049e-07,
"loss": 4.4363,
"step": 6540
},
{
"epoch": 0.8780520525892138,
"grad_norm": 3.953125,
"learning_rate": 8.925701737140136e-07,
"loss": 4.3825,
"step": 6545
},
{
"epoch": 0.878722833378052,
"grad_norm": 3.734375,
"learning_rate": 8.829246477936992e-07,
"loss": 4.4015,
"step": 6550
},
{
"epoch": 0.8793936141668902,
"grad_norm": 3.328125,
"learning_rate": 8.733291148495726e-07,
"loss": 4.3787,
"step": 6555
},
{
"epoch": 0.8800643949557284,
"grad_norm": 3.765625,
"learning_rate": 8.637836274982048e-07,
"loss": 4.3698,
"step": 6560
},
{
"epoch": 0.8807351757445667,
"grad_norm": 4.0625,
"learning_rate": 8.542882380817419e-07,
"loss": 4.3971,
"step": 6565
},
{
"epoch": 0.8814059565334049,
"grad_norm": 3.546875,
"learning_rate": 8.448429986676298e-07,
"loss": 4.3729,
"step": 6570
},
{
"epoch": 0.8820767373222431,
"grad_norm": 3.484375,
"learning_rate": 8.35447961048309e-07,
"loss": 4.3558,
"step": 6575
},
{
"epoch": 0.8827475181110813,
"grad_norm": 3.578125,
"learning_rate": 8.261031767409478e-07,
"loss": 4.4293,
"step": 6580
},
{
"epoch": 0.8834182988999195,
"grad_norm": 3.578125,
"learning_rate": 8.168086969871547e-07,
"loss": 4.3793,
"step": 6585
},
{
"epoch": 0.8840890796887577,
"grad_norm": 3.671875,
"learning_rate": 8.075645727526893e-07,
"loss": 4.3796,
"step": 6590
},
{
"epoch": 0.8847598604775959,
"grad_norm": 3.6875,
"learning_rate": 7.983708547271951e-07,
"loss": 4.4287,
"step": 6595
},
{
"epoch": 0.8854306412664341,
"grad_norm": 3.6875,
"learning_rate": 7.892275933239146e-07,
"loss": 4.4191,
"step": 6600
},
{
"epoch": 0.8861014220552723,
"grad_norm": 3.6875,
"learning_rate": 7.801348386794116e-07,
"loss": 4.4102,
"step": 6605
},
{
"epoch": 0.8867722028441105,
"grad_norm": 3.390625,
"learning_rate": 7.710926406533037e-07,
"loss": 4.3783,
"step": 6610
},
{
"epoch": 0.8874429836329487,
"grad_norm": 3.65625,
"learning_rate": 7.62101048827979e-07,
"loss": 4.3875,
"step": 6615
},
{
"epoch": 0.8881137644217869,
"grad_norm": 3.6875,
"learning_rate": 7.531601125083321e-07,
"loss": 4.3873,
"step": 6620
},
{
"epoch": 0.8887845452106252,
"grad_norm": 3.53125,
"learning_rate": 7.442698807214932e-07,
"loss": 4.4178,
"step": 6625
},
{
"epoch": 0.8894553259994634,
"grad_norm": 3.734375,
"learning_rate": 7.354304022165515e-07,
"loss": 4.3897,
"step": 6630
},
{
"epoch": 0.8901261067883016,
"grad_norm": 3.671875,
"learning_rate": 7.266417254642966e-07,
"loss": 4.3889,
"step": 6635
},
{
"epoch": 0.8907968875771398,
"grad_norm": 3.65625,
"learning_rate": 7.17903898656952e-07,
"loss": 4.3893,
"step": 6640
},
{
"epoch": 0.891467668365978,
"grad_norm": 3.78125,
"learning_rate": 7.092169697079021e-07,
"loss": 4.3921,
"step": 6645
},
{
"epoch": 0.8921384491548162,
"grad_norm": 3.625,
"learning_rate": 7.005809862514412e-07,
"loss": 4.3964,
"step": 6650
},
{
"epoch": 0.8928092299436544,
"grad_norm": 3.8125,
"learning_rate": 6.919959956425038e-07,
"loss": 4.4055,
"step": 6655
},
{
"epoch": 0.8934800107324926,
"grad_norm": 3.5625,
"learning_rate": 6.834620449564077e-07,
"loss": 4.3996,
"step": 6660
},
{
"epoch": 0.8941507915213308,
"grad_norm": 3.59375,
"learning_rate": 6.74979180988602e-07,
"loss": 4.3954,
"step": 6665
},
{
"epoch": 0.894821572310169,
"grad_norm": 3.796875,
"learning_rate": 6.665474502543944e-07,
"loss": 4.3729,
"step": 6670
},
{
"epoch": 0.8954923530990072,
"grad_norm": 3.515625,
"learning_rate": 6.581668989887158e-07,
"loss": 4.3782,
"step": 6675
},
{
"epoch": 0.8961631338878454,
"grad_norm": 3.59375,
"learning_rate": 6.498375731458529e-07,
"loss": 4.3743,
"step": 6680
},
{
"epoch": 0.8968339146766836,
"grad_norm": 3.703125,
"learning_rate": 6.415595183991985e-07,
"loss": 4.4123,
"step": 6685
},
{
"epoch": 0.8975046954655219,
"grad_norm": 4.0625,
"learning_rate": 6.333327801410094e-07,
"loss": 4.4441,
"step": 6690
},
{
"epoch": 0.8981754762543601,
"grad_norm": 4.1875,
"learning_rate": 6.251574034821461e-07,
"loss": 4.3585,
"step": 6695
},
{
"epoch": 0.8988462570431983,
"grad_norm": 3.75,
"learning_rate": 6.170334332518325e-07,
"loss": 4.4259,
"step": 6700
},
{
"epoch": 0.8995170378320365,
"grad_norm": 3.6875,
"learning_rate": 6.089609139974084e-07,
"loss": 4.3913,
"step": 6705
},
{
"epoch": 0.9001878186208747,
"grad_norm": 3.9375,
"learning_rate": 6.009398899840835e-07,
"loss": 4.3986,
"step": 6710
},
{
"epoch": 0.9008585994097129,
"grad_norm": 4.03125,
"learning_rate": 5.929704051946982e-07,
"loss": 4.3924,
"step": 6715
},
{
"epoch": 0.9015293801985511,
"grad_norm": 3.6875,
"learning_rate": 5.850525033294818e-07,
"loss": 4.3844,
"step": 6720
},
{
"epoch": 0.9022001609873893,
"grad_norm": 3.75,
"learning_rate": 5.771862278058038e-07,
"loss": 4.3865,
"step": 6725
},
{
"epoch": 0.9028709417762275,
"grad_norm": 3.609375,
"learning_rate": 5.693716217579536e-07,
"loss": 4.4349,
"step": 6730
},
{
"epoch": 0.9035417225650657,
"grad_norm": 3.53125,
"learning_rate": 5.616087280368898e-07,
"loss": 4.4081,
"step": 6735
},
{
"epoch": 0.9042125033539039,
"grad_norm": 3.890625,
"learning_rate": 5.538975892100041e-07,
"loss": 4.3878,
"step": 6740
},
{
"epoch": 0.9048832841427421,
"grad_norm": 3.828125,
"learning_rate": 5.462382475609029e-07,
"loss": 4.4002,
"step": 6745
},
{
"epoch": 0.9055540649315803,
"grad_norm": 3.90625,
"learning_rate": 5.386307450891581e-07,
"loss": 4.4117,
"step": 6750
},
{
"epoch": 0.9062248457204186,
"grad_norm": 3.890625,
"learning_rate": 5.310751235100864e-07,
"loss": 4.4077,
"step": 6755
},
{
"epoch": 0.9068956265092568,
"grad_norm": 3.921875,
"learning_rate": 5.235714242545209e-07,
"loss": 4.3939,
"step": 6760
},
{
"epoch": 0.907566407298095,
"grad_norm": 3.53125,
"learning_rate": 5.161196884685782e-07,
"loss": 4.3202,
"step": 6765
},
{
"epoch": 0.9082371880869332,
"grad_norm": 3.5,
"learning_rate": 5.087199570134371e-07,
"loss": 4.4227,
"step": 6770
},
{
"epoch": 0.9089079688757714,
"grad_norm": 3.4375,
"learning_rate": 5.013722704651158e-07,
"loss": 4.4416,
"step": 6775
},
{
"epoch": 0.9095787496646096,
"grad_norm": 3.75,
"learning_rate": 4.940766691142429e-07,
"loss": 4.4173,
"step": 6780
},
{
"epoch": 0.9102495304534478,
"grad_norm": 3.71875,
"learning_rate": 4.868331929658454e-07,
"loss": 4.3757,
"step": 6785
},
{
"epoch": 0.910920311242286,
"grad_norm": 3.625,
"learning_rate": 4.796418817391191e-07,
"loss": 4.4111,
"step": 6790
},
{
"epoch": 0.9115910920311242,
"grad_norm": 3.734375,
"learning_rate": 4.725027748672251e-07,
"loss": 4.4394,
"step": 6795
},
{
"epoch": 0.9122618728199624,
"grad_norm": 3.875,
"learning_rate": 4.654159114970547e-07,
"loss": 4.3961,
"step": 6800
},
{
"epoch": 0.9129326536088006,
"grad_norm": 4.0,
"learning_rate": 4.5838133048903057e-07,
"loss": 4.359,
"step": 6805
},
{
"epoch": 0.9136034343976388,
"grad_norm": 3.703125,
"learning_rate": 4.513990704168869e-07,
"loss": 4.3779,
"step": 6810
},
{
"epoch": 0.9142742151864771,
"grad_norm": 3.421875,
"learning_rate": 4.444691695674597e-07,
"loss": 4.3937,
"step": 6815
},
{
"epoch": 0.9149449959753153,
"grad_norm": 3.703125,
"learning_rate": 4.3759166594047333e-07,
"loss": 4.4453,
"step": 6820
},
{
"epoch": 0.9156157767641535,
"grad_norm": 3.796875,
"learning_rate": 4.307665972483399e-07,
"loss": 4.4053,
"step": 6825
},
{
"epoch": 0.9162865575529917,
"grad_norm": 3.671875,
"learning_rate": 4.2399400091594154e-07,
"loss": 4.3894,
"step": 6830
},
{
"epoch": 0.9169573383418299,
"grad_norm": 3.96875,
"learning_rate": 4.172739140804338e-07,
"loss": 4.41,
"step": 6835
},
{
"epoch": 0.9176281191306681,
"grad_norm": 3.5625,
"learning_rate": 4.1060637359104373e-07,
"loss": 4.4375,
"step": 6840
},
{
"epoch": 0.9182988999195063,
"grad_norm": 3.703125,
"learning_rate": 4.0399141600885314e-07,
"loss": 4.412,
"step": 6845
},
{
"epoch": 0.9189696807083445,
"grad_norm": 3.8125,
"learning_rate": 3.974290776066192e-07,
"loss": 4.3971,
"step": 6850
},
{
"epoch": 0.9196404614971827,
"grad_norm": 4.0,
"learning_rate": 3.909193943685585e-07,
"loss": 4.3867,
"step": 6855
},
{
"epoch": 0.9203112422860209,
"grad_norm": 3.59375,
"learning_rate": 3.844624019901566e-07,
"loss": 4.4049,
"step": 6860
},
{
"epoch": 0.9209820230748591,
"grad_norm": 3.359375,
"learning_rate": 3.780581358779745e-07,
"loss": 4.4568,
"step": 6865
},
{
"epoch": 0.9216528038636973,
"grad_norm": 3.34375,
"learning_rate": 3.717066311494466e-07,
"loss": 4.4242,
"step": 6870
},
{
"epoch": 0.9223235846525355,
"grad_norm": 3.8125,
"learning_rate": 3.6540792263269784e-07,
"loss": 4.396,
"step": 6875
},
{
"epoch": 0.9229943654413738,
"grad_norm": 3.890625,
"learning_rate": 3.591620448663469e-07,
"loss": 4.3961,
"step": 6880
},
{
"epoch": 0.923665146230212,
"grad_norm": 3.484375,
"learning_rate": 3.529690320993162e-07,
"loss": 4.3741,
"step": 6885
},
{
"epoch": 0.9243359270190502,
"grad_norm": 3.640625,
"learning_rate": 3.4682891829064923e-07,
"loss": 4.3834,
"step": 6890
},
{
"epoch": 0.9250067078078884,
"grad_norm": 3.578125,
"learning_rate": 3.4074173710931804e-07,
"loss": 4.4064,
"step": 6895
},
{
"epoch": 0.9256774885967266,
"grad_norm": 3.6875,
"learning_rate": 3.347075219340412e-07,
"loss": 4.4015,
"step": 6900
},
{
"epoch": 0.9263482693855648,
"grad_norm": 3.53125,
"learning_rate": 3.2872630585310315e-07,
"loss": 4.405,
"step": 6905
},
{
"epoch": 0.927019050174403,
"grad_norm": 3.828125,
"learning_rate": 3.227981216641707e-07,
"loss": 4.4143,
"step": 6910
},
{
"epoch": 0.9276898309632412,
"grad_norm": 3.8125,
"learning_rate": 3.1692300187411205e-07,
"loss": 4.4154,
"step": 6915
},
{
"epoch": 0.9283606117520794,
"grad_norm": 3.640625,
"learning_rate": 3.1110097869882063e-07,
"loss": 4.375,
"step": 6920
},
{
"epoch": 0.9290313925409176,
"grad_norm": 3.453125,
"learning_rate": 3.053320840630358e-07,
"loss": 4.4135,
"step": 6925
},
{
"epoch": 0.9297021733297558,
"grad_norm": 3.78125,
"learning_rate": 2.996163496001714e-07,
"loss": 4.3985,
"step": 6930
},
{
"epoch": 0.930372954118594,
"grad_norm": 3.59375,
"learning_rate": 2.939538066521408e-07,
"loss": 4.3818,
"step": 6935
},
{
"epoch": 0.9310437349074322,
"grad_norm": 3.875,
"learning_rate": 2.88344486269182e-07,
"loss": 4.3913,
"step": 6940
},
{
"epoch": 0.9317145156962705,
"grad_norm": 3.734375,
"learning_rate": 2.82788419209693e-07,
"loss": 4.4219,
"step": 6945
},
{
"epoch": 0.9323852964851087,
"grad_norm": 3.75,
"learning_rate": 2.772856359400589e-07,
"loss": 4.4144,
"step": 6950
},
{
"epoch": 0.9330560772739469,
"grad_norm": 3.609375,
"learning_rate": 2.7183616663448373e-07,
"loss": 4.3653,
"step": 6955
},
{
"epoch": 0.9337268580627851,
"grad_norm": 3.578125,
"learning_rate": 2.6644004117483357e-07,
"loss": 4.3984,
"step": 6960
},
{
"epoch": 0.9343976388516233,
"grad_norm": 3.8125,
"learning_rate": 2.610972891504593e-07,
"loss": 4.3968,
"step": 6965
},
{
"epoch": 0.9350684196404615,
"grad_norm": 3.609375,
"learning_rate": 2.5580793985804596e-07,
"loss": 4.4347,
"step": 6970
},
{
"epoch": 0.9357392004292997,
"grad_norm": 3.578125,
"learning_rate": 2.505720223014474e-07,
"loss": 4.4479,
"step": 6975
},
{
"epoch": 0.9364099812181379,
"grad_norm": 3.84375,
"learning_rate": 2.45389565191525e-07,
"loss": 4.4033,
"step": 6980
},
{
"epoch": 0.9370807620069761,
"grad_norm": 3.578125,
"learning_rate": 2.4026059694599477e-07,
"loss": 4.427,
"step": 6985
},
{
"epoch": 0.9377515427958143,
"grad_norm": 3.734375,
"learning_rate": 2.3518514568927065e-07,
"loss": 4.3964,
"step": 6990
},
{
"epoch": 0.9384223235846525,
"grad_norm": 3.75,
"learning_rate": 2.3016323925230678e-07,
"loss": 4.438,
"step": 6995
},
{
"epoch": 0.9390931043734907,
"grad_norm": 3.515625,
"learning_rate": 2.251949051724478e-07,
"loss": 4.3737,
"step": 7000
},
{
"epoch": 0.939763885162329,
"grad_norm": 3.703125,
"learning_rate": 2.202801706932767e-07,
"loss": 4.3943,
"step": 7005
},
{
"epoch": 0.9404346659511672,
"grad_norm": 3.734375,
"learning_rate": 2.1541906276446922e-07,
"loss": 4.3773,
"step": 7010
},
{
"epoch": 0.9411054467400054,
"grad_norm": 3.859375,
"learning_rate": 2.1061160804163984e-07,
"loss": 4.3354,
"step": 7015
},
{
"epoch": 0.9417762275288436,
"grad_norm": 3.734375,
"learning_rate": 2.058578328861971e-07,
"loss": 4.3882,
"step": 7020
},
{
"epoch": 0.9424470083176818,
"grad_norm": 3.96875,
"learning_rate": 2.0115776336520622e-07,
"loss": 4.4414,
"step": 7025
},
{
"epoch": 0.94311778910652,
"grad_norm": 3.5625,
"learning_rate": 1.9651142525123567e-07,
"loss": 4.4141,
"step": 7030
},
{
"epoch": 0.9437885698953582,
"grad_norm": 3.5,
"learning_rate": 1.9191884402222305e-07,
"loss": 4.3865,
"step": 7035
},
{
"epoch": 0.9444593506841964,
"grad_norm": 3.609375,
"learning_rate": 1.8738004486133166e-07,
"loss": 4.41,
"step": 7040
},
{
"epoch": 0.9451301314730346,
"grad_norm": 3.5625,
"learning_rate": 1.8289505265681407e-07,
"loss": 4.3861,
"step": 7045
},
{
"epoch": 0.9458009122618728,
"grad_norm": 3.796875,
"learning_rate": 1.7846389200187775e-07,
"loss": 4.4448,
"step": 7050
},
{
"epoch": 0.946471693050711,
"grad_norm": 3.484375,
"learning_rate": 1.7408658719454518e-07,
"loss": 4.4203,
"step": 7055
},
{
"epoch": 0.9471424738395492,
"grad_norm": 3.546875,
"learning_rate": 1.6976316223752288e-07,
"loss": 4.4004,
"step": 7060
},
{
"epoch": 0.9478132546283874,
"grad_norm": 3.609375,
"learning_rate": 1.654936408380725e-07,
"loss": 4.4079,
"step": 7065
},
{
"epoch": 0.9484840354172257,
"grad_norm": 3.703125,
"learning_rate": 1.6127804640787781e-07,
"loss": 4.4534,
"step": 7070
},
{
"epoch": 0.9491548162060639,
"grad_norm": 3.890625,
"learning_rate": 1.5711640206291567e-07,
"loss": 4.3794,
"step": 7075
},
{
"epoch": 0.9498255969949021,
"grad_norm": 3.5625,
"learning_rate": 1.53008730623333e-07,
"loss": 4.4101,
"step": 7080
},
{
"epoch": 0.9504963777837403,
"grad_norm": 3.609375,
"learning_rate": 1.489550546133156e-07,
"loss": 4.3742,
"step": 7085
},
{
"epoch": 0.9511671585725785,
"grad_norm": 3.515625,
"learning_rate": 1.4495539626097289e-07,
"loss": 4.3824,
"step": 7090
},
{
"epoch": 0.9518379393614167,
"grad_norm": 3.703125,
"learning_rate": 1.4100977749820887e-07,
"loss": 4.4421,
"step": 7095
},
{
"epoch": 0.9525087201502549,
"grad_norm": 3.84375,
"learning_rate": 1.371182199606036e-07,
"loss": 4.4194,
"step": 7100
},
{
"epoch": 0.9531795009390931,
"grad_norm": 3.71875,
"learning_rate": 1.3328074498729749e-07,
"loss": 4.4078,
"step": 7105
},
{
"epoch": 0.9538502817279313,
"grad_norm": 3.859375,
"learning_rate": 1.2949737362087156e-07,
"loss": 4.3763,
"step": 7110
},
{
"epoch": 0.9545210625167695,
"grad_norm": 3.515625,
"learning_rate": 1.257681266072319e-07,
"loss": 4.3965,
"step": 7115
},
{
"epoch": 0.9551918433056077,
"grad_norm": 3.671875,
"learning_rate": 1.2209302439549764e-07,
"loss": 4.4488,
"step": 7120
},
{
"epoch": 0.9558626240944459,
"grad_norm": 3.71875,
"learning_rate": 1.184720871378886e-07,
"loss": 4.3942,
"step": 7125
},
{
"epoch": 0.9565334048832841,
"grad_norm": 3.703125,
"learning_rate": 1.1490533468961118e-07,
"loss": 4.3801,
"step": 7130
},
{
"epoch": 0.9572041856721224,
"grad_norm": 3.671875,
"learning_rate": 1.1139278660875608e-07,
"loss": 4.3799,
"step": 7135
},
{
"epoch": 0.9578749664609606,
"grad_norm": 3.703125,
"learning_rate": 1.0793446215618397e-07,
"loss": 4.4207,
"step": 7140
},
{
"epoch": 0.9585457472497988,
"grad_norm": 3.609375,
"learning_rate": 1.0453038029542561e-07,
"loss": 4.3428,
"step": 7145
},
{
"epoch": 0.959216528038637,
"grad_norm": 3.546875,
"learning_rate": 1.0118055969257523e-07,
"loss": 4.4319,
"step": 7150
},
{
"epoch": 0.9598873088274752,
"grad_norm": 3.71875,
"learning_rate": 9.788501871618727e-08,
"loss": 4.3968,
"step": 7155
},
{
"epoch": 0.9605580896163134,
"grad_norm": 3.921875,
"learning_rate": 9.464377543717651e-08,
"loss": 4.3763,
"step": 7160
},
{
"epoch": 0.9612288704051516,
"grad_norm": 3.609375,
"learning_rate": 9.145684762872365e-08,
"loss": 4.4095,
"step": 7165
},
{
"epoch": 0.9618996511939898,
"grad_norm": 3.875,
"learning_rate": 8.832425276616763e-08,
"loss": 4.3961,
"step": 7170
},
{
"epoch": 0.962570431982828,
"grad_norm": 3.65625,
"learning_rate": 8.524600802691907e-08,
"loss": 4.3966,
"step": 7175
},
{
"epoch": 0.9632412127716662,
"grad_norm": 3.765625,
"learning_rate": 8.222213029036141e-08,
"loss": 4.4085,
"step": 7180
},
{
"epoch": 0.9639119935605044,
"grad_norm": 3.734375,
"learning_rate": 7.925263613776102e-08,
"loss": 4.3662,
"step": 7185
},
{
"epoch": 0.9645827743493426,
"grad_norm": 3.75,
"learning_rate": 7.63375418521739e-08,
"loss": 4.3742,
"step": 7190
},
{
"epoch": 0.9652535551381809,
"grad_norm": 3.625,
"learning_rate": 7.347686341835803e-08,
"loss": 4.3853,
"step": 7195
},
{
"epoch": 0.9659243359270191,
"grad_norm": 4.0625,
"learning_rate": 7.067061652268337e-08,
"loss": 4.3483,
"step": 7200
},
{
"epoch": 0.9665951167158573,
"grad_norm": 3.71875,
"learning_rate": 6.791881655305087e-08,
"loss": 4.3992,
"step": 7205
},
{
"epoch": 0.9672658975046955,
"grad_norm": 3.84375,
"learning_rate": 6.522147859880035e-08,
"loss": 4.416,
"step": 7210
},
{
"epoch": 0.9679366782935337,
"grad_norm": 3.828125,
"learning_rate": 6.257861745063709e-08,
"loss": 4.4027,
"step": 7215
},
{
"epoch": 0.9686074590823719,
"grad_norm": 3.375,
"learning_rate": 5.999024760054095e-08,
"loss": 4.4127,
"step": 7220
},
{
"epoch": 0.9692782398712101,
"grad_norm": 3.765625,
"learning_rate": 5.745638324169633e-08,
"loss": 4.4147,
"step": 7225
},
{
"epoch": 0.9699490206600483,
"grad_norm": 3.390625,
"learning_rate": 5.4977038268406724e-08,
"loss": 4.3556,
"step": 7230
},
{
"epoch": 0.9706198014488865,
"grad_norm": 3.46875,
"learning_rate": 5.2552226276021414e-08,
"loss": 4.4057,
"step": 7235
},
{
"epoch": 0.9712905822377247,
"grad_norm": 3.8125,
"learning_rate": 5.018196056086444e-08,
"loss": 4.3939,
"step": 7240
},
{
"epoch": 0.9719613630265629,
"grad_norm": 3.8125,
"learning_rate": 4.7866254120155774e-08,
"loss": 4.4177,
"step": 7245
},
{
"epoch": 0.9726321438154011,
"grad_norm": 3.734375,
"learning_rate": 4.5605119651941366e-08,
"loss": 4.3928,
"step": 7250
},
{
"epoch": 0.9733029246042393,
"grad_norm": 4.03125,
"learning_rate": 4.339856955502875e-08,
"loss": 4.3753,
"step": 7255
},
{
"epoch": 0.9739737053930776,
"grad_norm": 3.984375,
"learning_rate": 4.124661592891044e-08,
"loss": 4.4159,
"step": 7260
},
{
"epoch": 0.9746444861819158,
"grad_norm": 3.75,
"learning_rate": 3.914927057370621e-08,
"loss": 4.4484,
"step": 7265
},
{
"epoch": 0.975315266970754,
"grad_norm": 3.625,
"learning_rate": 3.710654499009203e-08,
"loss": 4.4105,
"step": 7270
},
{
"epoch": 0.9759860477595922,
"grad_norm": 3.875,
"learning_rate": 3.511845037924011e-08,
"loss": 4.3446,
"step": 7275
},
{
"epoch": 0.9766568285484304,
"grad_norm": 3.859375,
"learning_rate": 3.318499764275784e-08,
"loss": 4.4147,
"step": 7280
},
{
"epoch": 0.9773276093372686,
"grad_norm": 4.0625,
"learning_rate": 3.1306197382624525e-08,
"loss": 4.4198,
"step": 7285
},
{
"epoch": 0.9779983901261068,
"grad_norm": 3.5625,
"learning_rate": 2.948205990113806e-08,
"loss": 4.4451,
"step": 7290
},
{
"epoch": 0.978669170914945,
"grad_norm": 3.625,
"learning_rate": 2.771259520085501e-08,
"loss": 4.3992,
"step": 7295
},
{
"epoch": 0.9793399517037832,
"grad_norm": 3.546875,
"learning_rate": 2.5997812984535076e-08,
"loss": 4.3659,
"step": 7300
},
{
"epoch": 0.9800107324926214,
"grad_norm": 3.734375,
"learning_rate": 2.4337722655093376e-08,
"loss": 4.3878,
"step": 7305
},
{
"epoch": 0.9806815132814596,
"grad_norm": 3.84375,
"learning_rate": 2.273233331554159e-08,
"loss": 4.411,
"step": 7310
},
{
"epoch": 0.9813522940702978,
"grad_norm": 3.5625,
"learning_rate": 2.1181653768943543e-08,
"loss": 4.4082,
"step": 7315
},
{
"epoch": 0.982023074859136,
"grad_norm": 3.59375,
"learning_rate": 1.9685692518365274e-08,
"loss": 4.4145,
"step": 7320
},
{
"epoch": 0.9826938556479743,
"grad_norm": 3.671875,
"learning_rate": 1.824445776682504e-08,
"loss": 4.3883,
"step": 7325
},
{
"epoch": 0.9833646364368125,
"grad_norm": 3.46875,
"learning_rate": 1.68579574172556e-08,
"loss": 4.3861,
"step": 7330
},
{
"epoch": 0.9840354172256507,
"grad_norm": 3.515625,
"learning_rate": 1.5526199072453117e-08,
"loss": 4.3479,
"step": 7335
},
{
"epoch": 0.9847061980144889,
"grad_norm": 3.84375,
"learning_rate": 1.424919003504166e-08,
"loss": 4.4215,
"step": 7340
},
{
"epoch": 0.9853769788033271,
"grad_norm": 3.59375,
"learning_rate": 1.3026937307428766e-08,
"loss": 4.406,
"step": 7345
},
{
"epoch": 0.9860477595921653,
"grad_norm": 3.703125,
"learning_rate": 1.1859447591769934e-08,
"loss": 4.3822,
"step": 7350
},
{
"epoch": 0.9867185403810035,
"grad_norm": 3.53125,
"learning_rate": 1.0746727289928649e-08,
"loss": 4.4112,
"step": 7355
},
{
"epoch": 0.9873893211698417,
"grad_norm": 3.96875,
"learning_rate": 9.688782503446403e-09,
"loss": 4.3845,
"step": 7360
},
{
"epoch": 0.9880601019586799,
"grad_norm": 3.984375,
"learning_rate": 8.685619033503844e-09,
"loss": 4.4134,
"step": 7365
},
{
"epoch": 0.9887308827475181,
"grad_norm": 3.3125,
"learning_rate": 7.737242380891907e-09,
"loss": 4.3812,
"step": 7370
},
{
"epoch": 0.9894016635363563,
"grad_norm": 4.03125,
"learning_rate": 6.8436577459829454e-09,
"loss": 4.3623,
"step": 7375
},
{
"epoch": 0.9900724443251945,
"grad_norm": 3.53125,
"learning_rate": 6.004870028696319e-09,
"loss": 4.4073,
"step": 7380
},
{
"epoch": 0.9907432251140328,
"grad_norm": 3.703125,
"learning_rate": 5.220883828479517e-09,
"loss": 4.3742,
"step": 7385
},
{
"epoch": 0.991414005902871,
"grad_norm": 3.609375,
"learning_rate": 4.4917034442770736e-09,
"loss": 4.4221,
"step": 7390
},
{
"epoch": 0.9920847866917092,
"grad_norm": 3.578125,
"learning_rate": 3.817332874508361e-09,
"loss": 4.4118,
"step": 7395
},
{
"epoch": 0.9927555674805474,
"grad_norm": 3.671875,
"learning_rate": 3.1977758170476105e-09,
"loss": 4.4171,
"step": 7400
},
{
"epoch": 0.9934263482693856,
"grad_norm": 3.625,
"learning_rate": 2.6330356692017e-09,
"loss": 4.389,
"step": 7405
},
{
"epoch": 0.9940971290582238,
"grad_norm": 3.59375,
"learning_rate": 2.1231155276912886e-09,
"loss": 4.3805,
"step": 7410
},
{
"epoch": 0.994767909847062,
"grad_norm": 3.96875,
"learning_rate": 1.6680181886352676e-09,
"loss": 4.3677,
"step": 7415
},
{
"epoch": 0.9954386906359002,
"grad_norm": 3.46875,
"learning_rate": 1.2677461475352205e-09,
"loss": 4.4102,
"step": 7420
},
{
"epoch": 0.9961094714247384,
"grad_norm": 3.671875,
"learning_rate": 9.223015992598783e-10,
"loss": 4.4212,
"step": 7425
},
{
"epoch": 0.9967802522135766,
"grad_norm": 3.703125,
"learning_rate": 6.316864380351284e-10,
"loss": 4.3931,
"step": 7430
},
{
"epoch": 0.9974510330024148,
"grad_norm": 3.53125,
"learning_rate": 3.959022574329119e-10,
"loss": 4.4243,
"step": 7435
},
{
"epoch": 0.998121813791253,
"grad_norm": 3.6875,
"learning_rate": 2.1495035036456225e-10,
"loss": 4.3631,
"step": 7440
},
{
"epoch": 0.9987925945800912,
"grad_norm": 3.71875,
"learning_rate": 8.883170906748284e-11,
"loss": 4.4098,
"step": 7445
},
{
"epoch": 0.9994633753689295,
"grad_norm": 3.546875,
"learning_rate": 1.7547025107367276e-11,
"loss": 4.4341,
"step": 7450
},
{
"epoch": 1.0,
"eval_loss": 4.2098798751831055,
"eval_runtime": 48.3409,
"eval_samples_per_second": 548.004,
"eval_steps_per_second": 17.128,
"step": 7454
},
{
"epoch": 1.0,
"step": 7454,
"total_flos": 1.24650931617792e+17,
"train_loss": 4.485118892066819,
"train_runtime": 1567.5346,
"train_samples_per_second": 152.149,
"train_steps_per_second": 4.755
}
],
"logging_steps": 5,
"max_steps": 7454,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.24650931617792e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}