gr00t-so101-eef-300eps / trainer_state.json
tingtu0721's picture
Add files using upload-large-folder tool
1ca4bc1 verified
Raw
History Blame Contribute Delete
140 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"grad_norm": 0.3207867443561554,
"learning_rate": 1.8e-06,
"loss": 1.1005,
"step": 10
},
{
"grad_norm": 0.2261800318956375,
"learning_rate": 3.8e-06,
"loss": 1.1111,
"step": 20
},
{
"grad_norm": 0.4193296730518341,
"learning_rate": 5.8e-06,
"loss": 1.0898,
"step": 30
},
{
"grad_norm": 0.21961848437786102,
"learning_rate": 7.8e-06,
"loss": 1.0998,
"step": 40
},
{
"grad_norm": 0.2328994870185852,
"learning_rate": 9.800000000000001e-06,
"loss": 1.095,
"step": 50
},
{
"grad_norm": 0.31275686621665955,
"learning_rate": 1.18e-05,
"loss": 1.0737,
"step": 60
},
{
"grad_norm": 0.49875926971435547,
"learning_rate": 1.3800000000000002e-05,
"loss": 1.0479,
"step": 70
},
{
"grad_norm": 0.3772468864917755,
"learning_rate": 1.58e-05,
"loss": 1.0288,
"step": 80
},
{
"grad_norm": 0.7008079290390015,
"learning_rate": 1.78e-05,
"loss": 1.0152,
"step": 90
},
{
"grad_norm": 1.7280652523040771,
"learning_rate": 1.9800000000000004e-05,
"loss": 0.985,
"step": 100
},
{
"grad_norm": 1.879516363143921,
"learning_rate": 2.18e-05,
"loss": 0.8571,
"step": 110
},
{
"grad_norm": 2.1656241416931152,
"learning_rate": 2.38e-05,
"loss": 0.7327,
"step": 120
},
{
"grad_norm": 2.6657841205596924,
"learning_rate": 2.58e-05,
"loss": 0.5899,
"step": 130
},
{
"grad_norm": 3.001236915588379,
"learning_rate": 2.7800000000000005e-05,
"loss": 0.4492,
"step": 140
},
{
"grad_norm": 2.5079903602600098,
"learning_rate": 2.98e-05,
"loss": 0.3171,
"step": 150
},
{
"grad_norm": 2.579601526260376,
"learning_rate": 3.18e-05,
"loss": 0.2487,
"step": 160
},
{
"grad_norm": 2.6017887592315674,
"learning_rate": 3.38e-05,
"loss": 0.1886,
"step": 170
},
{
"grad_norm": 2.400867462158203,
"learning_rate": 3.58e-05,
"loss": 0.1838,
"step": 180
},
{
"grad_norm": 1.6188538074493408,
"learning_rate": 3.7800000000000004e-05,
"loss": 0.1475,
"step": 190
},
{
"grad_norm": 2.1670868396759033,
"learning_rate": 3.9800000000000005e-05,
"loss": 0.1703,
"step": 200
},
{
"grad_norm": 1.6946394443511963,
"learning_rate": 4.18e-05,
"loss": 0.1597,
"step": 210
},
{
"grad_norm": 1.9171910285949707,
"learning_rate": 4.38e-05,
"loss": 0.1423,
"step": 220
},
{
"grad_norm": 1.6843936443328857,
"learning_rate": 4.58e-05,
"loss": 0.1357,
"step": 230
},
{
"grad_norm": 1.5568238496780396,
"learning_rate": 4.78e-05,
"loss": 0.1338,
"step": 240
},
{
"grad_norm": 1.5835602283477783,
"learning_rate": 4.9800000000000004e-05,
"loss": 0.1357,
"step": 250
},
{
"grad_norm": 1.192285180091858,
"learning_rate": 5.1800000000000005e-05,
"loss": 0.1244,
"step": 260
},
{
"grad_norm": 1.5154273509979248,
"learning_rate": 5.380000000000001e-05,
"loss": 0.1376,
"step": 270
},
{
"grad_norm": 0.9447847604751587,
"learning_rate": 5.580000000000001e-05,
"loss": 0.12,
"step": 280
},
{
"grad_norm": 1.3534984588623047,
"learning_rate": 5.7799999999999995e-05,
"loss": 0.1323,
"step": 290
},
{
"grad_norm": 1.8623212575912476,
"learning_rate": 5.9800000000000003e-05,
"loss": 0.1198,
"step": 300
},
{
"grad_norm": 1.0872575044631958,
"learning_rate": 6.18e-05,
"loss": 0.1263,
"step": 310
},
{
"grad_norm": 1.2258734703063965,
"learning_rate": 6.38e-05,
"loss": 0.1014,
"step": 320
},
{
"grad_norm": 1.2477049827575684,
"learning_rate": 6.58e-05,
"loss": 0.1009,
"step": 330
},
{
"grad_norm": 1.2992441654205322,
"learning_rate": 6.780000000000001e-05,
"loss": 0.1054,
"step": 340
},
{
"grad_norm": 0.8912233114242554,
"learning_rate": 6.98e-05,
"loss": 0.0954,
"step": 350
},
{
"grad_norm": 0.8545780777931213,
"learning_rate": 7.18e-05,
"loss": 0.1053,
"step": 360
},
{
"grad_norm": 1.6349998712539673,
"learning_rate": 7.38e-05,
"loss": 0.1122,
"step": 370
},
{
"grad_norm": 1.068824052810669,
"learning_rate": 7.58e-05,
"loss": 0.1066,
"step": 380
},
{
"grad_norm": 0.7284474968910217,
"learning_rate": 7.780000000000001e-05,
"loss": 0.1077,
"step": 390
},
{
"grad_norm": 0.840778112411499,
"learning_rate": 7.98e-05,
"loss": 0.0976,
"step": 400
},
{
"grad_norm": 0.8797901272773743,
"learning_rate": 8.18e-05,
"loss": 0.1034,
"step": 410
},
{
"grad_norm": 1.127791404724121,
"learning_rate": 8.38e-05,
"loss": 0.0958,
"step": 420
},
{
"grad_norm": 0.8316822052001953,
"learning_rate": 8.58e-05,
"loss": 0.0985,
"step": 430
},
{
"grad_norm": 0.8704973459243774,
"learning_rate": 8.78e-05,
"loss": 0.1086,
"step": 440
},
{
"grad_norm": 0.7539105415344238,
"learning_rate": 8.98e-05,
"loss": 0.1027,
"step": 450
},
{
"grad_norm": 1.0874912738800049,
"learning_rate": 9.180000000000001e-05,
"loss": 0.096,
"step": 460
},
{
"grad_norm": 0.8364002108573914,
"learning_rate": 9.38e-05,
"loss": 0.1116,
"step": 470
},
{
"grad_norm": 0.7446768283843994,
"learning_rate": 9.58e-05,
"loss": 0.0959,
"step": 480
},
{
"grad_norm": 0.954987645149231,
"learning_rate": 9.78e-05,
"loss": 0.0914,
"step": 490
},
{
"grad_norm": 1.2324256896972656,
"learning_rate": 9.98e-05,
"loss": 0.1003,
"step": 500
},
{
"grad_norm": 0.9119232892990112,
"learning_rate": 9.9999778549206e-05,
"loss": 0.1025,
"step": 510
},
{
"grad_norm": 0.7419020533561707,
"learning_rate": 9.999901304280685e-05,
"loss": 0.0934,
"step": 520
},
{
"grad_norm": 0.5527909994125366,
"learning_rate": 9.999770075521164e-05,
"loss": 0.0934,
"step": 530
},
{
"grad_norm": 0.5270976424217224,
"learning_rate": 9.99958417007713e-05,
"loss": 0.089,
"step": 540
},
{
"grad_norm": 0.5891151428222656,
"learning_rate": 9.999343589981615e-05,
"loss": 0.0854,
"step": 550
},
{
"grad_norm": 0.6041948795318604,
"learning_rate": 9.999048337865568e-05,
"loss": 0.0816,
"step": 560
},
{
"grad_norm": 0.5275608897209167,
"learning_rate": 9.998698416957815e-05,
"loss": 0.0747,
"step": 570
},
{
"grad_norm": 0.8493439555168152,
"learning_rate": 9.998293831085037e-05,
"loss": 0.078,
"step": 580
},
{
"grad_norm": 0.7588852643966675,
"learning_rate": 9.997834584671719e-05,
"loss": 0.0827,
"step": 590
},
{
"grad_norm": 0.7806046009063721,
"learning_rate": 9.997320682740107e-05,
"loss": 0.0831,
"step": 600
},
{
"grad_norm": 0.6732871532440186,
"learning_rate": 9.996752130910149e-05,
"loss": 0.0951,
"step": 610
},
{
"grad_norm": 0.6196691393852234,
"learning_rate": 9.99612893539944e-05,
"loss": 0.0792,
"step": 620
},
{
"grad_norm": 0.5885448455810547,
"learning_rate": 9.995451103023144e-05,
"loss": 0.0712,
"step": 630
},
{
"grad_norm": 0.6277177333831787,
"learning_rate": 9.994718641193928e-05,
"loss": 0.0696,
"step": 640
},
{
"grad_norm": 0.5323973894119263,
"learning_rate": 9.993931557921874e-05,
"loss": 0.0711,
"step": 650
},
{
"grad_norm": 0.6824409365653992,
"learning_rate": 9.993089861814402e-05,
"loss": 0.0803,
"step": 660
},
{
"grad_norm": 0.9927233457565308,
"learning_rate": 9.992193562076166e-05,
"loss": 0.0892,
"step": 670
},
{
"grad_norm": 0.609441876411438,
"learning_rate": 9.991242668508954e-05,
"loss": 0.0766,
"step": 680
},
{
"grad_norm": 0.6370614171028137,
"learning_rate": 9.990237191511587e-05,
"loss": 0.0696,
"step": 690
},
{
"grad_norm": 0.6155320405960083,
"learning_rate": 9.989177142079802e-05,
"loss": 0.0764,
"step": 700
},
{
"grad_norm": 0.5438863039016724,
"learning_rate": 9.988062531806126e-05,
"loss": 0.0806,
"step": 710
},
{
"grad_norm": 0.5776309370994568,
"learning_rate": 9.986893372879762e-05,
"loss": 0.076,
"step": 720
},
{
"grad_norm": 0.6240535974502563,
"learning_rate": 9.985669678086443e-05,
"loss": 0.0726,
"step": 730
},
{
"grad_norm": 0.5558215975761414,
"learning_rate": 9.984391460808298e-05,
"loss": 0.063,
"step": 740
},
{
"grad_norm": 0.47590723633766174,
"learning_rate": 9.983058735023709e-05,
"loss": 0.0698,
"step": 750
},
{
"grad_norm": 0.5383110642433167,
"learning_rate": 9.98167151530715e-05,
"loss": 0.061,
"step": 760
},
{
"grad_norm": 0.747265100479126,
"learning_rate": 9.980229816829034e-05,
"loss": 0.068,
"step": 770
},
{
"grad_norm": 0.7335792183876038,
"learning_rate": 9.978733655355544e-05,
"loss": 0.0687,
"step": 780
},
{
"grad_norm": 0.6182377338409424,
"learning_rate": 9.977183047248464e-05,
"loss": 0.0675,
"step": 790
},
{
"grad_norm": 0.6036187410354614,
"learning_rate": 9.975578009464992e-05,
"loss": 0.0771,
"step": 800
},
{
"grad_norm": 0.5219994783401489,
"learning_rate": 9.97391855955757e-05,
"loss": 0.0742,
"step": 810
},
{
"grad_norm": 0.5030124187469482,
"learning_rate": 9.972204715673669e-05,
"loss": 0.0749,
"step": 820
},
{
"grad_norm": 0.5388495326042175,
"learning_rate": 9.970436496555617e-05,
"loss": 0.0696,
"step": 830
},
{
"grad_norm": 0.6444914937019348,
"learning_rate": 9.968613921540373e-05,
"loss": 0.0638,
"step": 840
},
{
"grad_norm": 0.47062602639198303,
"learning_rate": 9.966737010559326e-05,
"loss": 0.0753,
"step": 850
},
{
"grad_norm": 0.6814469695091248,
"learning_rate": 9.964805784138072e-05,
"loss": 0.0694,
"step": 860
},
{
"grad_norm": 0.47789523005485535,
"learning_rate": 9.962820263396195e-05,
"loss": 0.0698,
"step": 870
},
{
"grad_norm": 0.5015204548835754,
"learning_rate": 9.960780470047033e-05,
"loss": 0.0685,
"step": 880
},
{
"grad_norm": 0.5280117392539978,
"learning_rate": 9.958686426397437e-05,
"loss": 0.0628,
"step": 890
},
{
"grad_norm": 0.35730186104774475,
"learning_rate": 9.956538155347534e-05,
"loss": 0.073,
"step": 900
},
{
"grad_norm": 0.4053404927253723,
"learning_rate": 9.95433568039047e-05,
"loss": 0.0611,
"step": 910
},
{
"grad_norm": 0.45370951294898987,
"learning_rate": 9.952079025612162e-05,
"loss": 0.0642,
"step": 920
},
{
"grad_norm": 0.4795547425746918,
"learning_rate": 9.949768215691022e-05,
"loss": 0.0622,
"step": 930
},
{
"grad_norm": 0.5141300559043884,
"learning_rate": 9.9474032758977e-05,
"loss": 0.0661,
"step": 940
},
{
"grad_norm": 0.5531869530677795,
"learning_rate": 9.944984232094794e-05,
"loss": 0.0729,
"step": 950
},
{
"grad_norm": 0.48658886551856995,
"learning_rate": 9.942511110736584e-05,
"loss": 0.0555,
"step": 960
},
{
"grad_norm": 0.40576183795928955,
"learning_rate": 9.939983938868726e-05,
"loss": 0.0594,
"step": 970
},
{
"grad_norm": 0.4921189546585083,
"learning_rate": 9.93740274412797e-05,
"loss": 0.0635,
"step": 980
},
{
"grad_norm": 0.3368489742279053,
"learning_rate": 9.934767554741846e-05,
"loss": 0.0668,
"step": 990
},
{
"grad_norm": 0.5900139212608337,
"learning_rate": 9.932078399528361e-05,
"loss": 0.0666,
"step": 1000
},
{
"grad_norm": 0.38995057344436646,
"learning_rate": 9.929335307895689e-05,
"loss": 0.0589,
"step": 1010
},
{
"grad_norm": 0.3530271649360657,
"learning_rate": 9.926538309841839e-05,
"loss": 0.0599,
"step": 1020
},
{
"grad_norm": 0.625306248664856,
"learning_rate": 9.923687435954334e-05,
"loss": 0.0612,
"step": 1030
},
{
"grad_norm": 0.3649205267429352,
"learning_rate": 9.920782717409873e-05,
"loss": 0.064,
"step": 1040
},
{
"grad_norm": 0.4765617847442627,
"learning_rate": 9.917824185973994e-05,
"loss": 0.0557,
"step": 1050
},
{
"grad_norm": 0.6244515180587769,
"learning_rate": 9.914811874000723e-05,
"loss": 0.068,
"step": 1060
},
{
"grad_norm": 0.5091502070426941,
"learning_rate": 9.911745814432218e-05,
"loss": 0.0622,
"step": 1070
},
{
"grad_norm": 0.3705998957157135,
"learning_rate": 9.90862604079842e-05,
"loss": 0.0657,
"step": 1080
},
{
"grad_norm": 0.432797908782959,
"learning_rate": 9.90545258721667e-05,
"loss": 0.0528,
"step": 1090
},
{
"grad_norm": 0.4613644480705261,
"learning_rate": 9.90222548839135e-05,
"loss": 0.0586,
"step": 1100
},
{
"grad_norm": 0.5109816193580627,
"learning_rate": 9.898944779613495e-05,
"loss": 0.0634,
"step": 1110
},
{
"grad_norm": 0.44418686628341675,
"learning_rate": 9.89561049676041e-05,
"loss": 0.0642,
"step": 1120
},
{
"grad_norm": 0.4306738078594208,
"learning_rate": 9.89222267629528e-05,
"loss": 0.0661,
"step": 1130
},
{
"grad_norm": 0.44220030307769775,
"learning_rate": 9.888781355266763e-05,
"loss": 0.066,
"step": 1140
},
{
"grad_norm": 0.47052425146102905,
"learning_rate": 9.885286571308598e-05,
"loss": 0.0596,
"step": 1150
},
{
"grad_norm": 0.4535216689109802,
"learning_rate": 9.881738362639182e-05,
"loss": 0.0583,
"step": 1160
},
{
"grad_norm": 0.4234335422515869,
"learning_rate": 9.878136768061154e-05,
"loss": 0.0568,
"step": 1170
},
{
"grad_norm": 0.45696863532066345,
"learning_rate": 9.874481826960979e-05,
"loss": 0.0627,
"step": 1180
},
{
"grad_norm": 0.3872640132904053,
"learning_rate": 9.870773579308503e-05,
"loss": 0.06,
"step": 1190
},
{
"grad_norm": 0.4766055643558502,
"learning_rate": 9.867012065656533e-05,
"loss": 0.0593,
"step": 1200
},
{
"grad_norm": 0.39494651556015015,
"learning_rate": 9.863197327140376e-05,
"loss": 0.058,
"step": 1210
},
{
"grad_norm": 0.4328981637954712,
"learning_rate": 9.859329405477403e-05,
"loss": 0.05,
"step": 1220
},
{
"grad_norm": 0.37711066007614136,
"learning_rate": 9.855408342966585e-05,
"loss": 0.0625,
"step": 1230
},
{
"grad_norm": 0.3636433780193329,
"learning_rate": 9.851434182488033e-05,
"loss": 0.0528,
"step": 1240
},
{
"grad_norm": 0.40212321281433105,
"learning_rate": 9.84740696750253e-05,
"loss": 0.0564,
"step": 1250
},
{
"grad_norm": 0.35969722270965576,
"learning_rate": 9.843326742051055e-05,
"loss": 0.0549,
"step": 1260
},
{
"grad_norm": 0.4876939058303833,
"learning_rate": 9.839193550754297e-05,
"loss": 0.0596,
"step": 1270
},
{
"grad_norm": 0.482296347618103,
"learning_rate": 9.835007438812177e-05,
"loss": 0.0561,
"step": 1280
},
{
"grad_norm": 0.3805181384086609,
"learning_rate": 9.830768452003341e-05,
"loss": 0.0613,
"step": 1290
},
{
"grad_norm": 0.3932335078716278,
"learning_rate": 9.826476636684671e-05,
"loss": 0.0633,
"step": 1300
},
{
"grad_norm": 0.4378570318222046,
"learning_rate": 9.822132039790773e-05,
"loss": 0.053,
"step": 1310
},
{
"grad_norm": 0.3924146592617035,
"learning_rate": 9.817734708833461e-05,
"loss": 0.0566,
"step": 1320
},
{
"grad_norm": 0.35605448484420776,
"learning_rate": 9.813284691901243e-05,
"loss": 0.0532,
"step": 1330
},
{
"grad_norm": 0.40543317794799805,
"learning_rate": 9.808782037658792e-05,
"loss": 0.0557,
"step": 1340
},
{
"grad_norm": 0.31904837489128113,
"learning_rate": 9.804226795346411e-05,
"loss": 0.0493,
"step": 1350
},
{
"grad_norm": 0.46915408968925476,
"learning_rate": 9.799619014779503e-05,
"loss": 0.05,
"step": 1360
},
{
"grad_norm": 0.32613101601600647,
"learning_rate": 9.794958746348013e-05,
"loss": 0.0447,
"step": 1370
},
{
"grad_norm": 0.516217052936554,
"learning_rate": 9.790246041015896e-05,
"loss": 0.0492,
"step": 1380
},
{
"grad_norm": 0.3911663889884949,
"learning_rate": 9.785480950320538e-05,
"loss": 0.0493,
"step": 1390
},
{
"grad_norm": 0.3783031702041626,
"learning_rate": 9.78066352637221e-05,
"loss": 0.0561,
"step": 1400
},
{
"grad_norm": 0.4387311637401581,
"learning_rate": 9.775793821853488e-05,
"loss": 0.0505,
"step": 1410
},
{
"grad_norm": 0.41631126403808594,
"learning_rate": 9.77087189001868e-05,
"loss": 0.0613,
"step": 1420
},
{
"grad_norm": 0.32027342915534973,
"learning_rate": 9.765897784693243e-05,
"loss": 0.0519,
"step": 1430
},
{
"grad_norm": 0.39711493253707886,
"learning_rate": 9.760871560273197e-05,
"loss": 0.0526,
"step": 1440
},
{
"grad_norm": 0.39819058775901794,
"learning_rate": 9.755793271724526e-05,
"loss": 0.0564,
"step": 1450
},
{
"grad_norm": 0.5000616312026978,
"learning_rate": 9.750662974582584e-05,
"loss": 0.0567,
"step": 1460
},
{
"grad_norm": 0.34620797634124756,
"learning_rate": 9.745480724951473e-05,
"loss": 0.0521,
"step": 1470
},
{
"grad_norm": 0.47389352321624756,
"learning_rate": 9.740246579503447e-05,
"loss": 0.0564,
"step": 1480
},
{
"grad_norm": 0.40866824984550476,
"learning_rate": 9.734960595478284e-05,
"loss": 0.0455,
"step": 1490
},
{
"grad_norm": 0.46241334080696106,
"learning_rate": 9.729622830682657e-05,
"loss": 0.0614,
"step": 1500
},
{
"grad_norm": 0.40948155522346497,
"learning_rate": 9.724233343489504e-05,
"loss": 0.0505,
"step": 1510
},
{
"grad_norm": 0.4696265161037445,
"learning_rate": 9.718792192837396e-05,
"loss": 0.0573,
"step": 1520
},
{
"grad_norm": 0.3956542909145355,
"learning_rate": 9.713299438229886e-05,
"loss": 0.0549,
"step": 1530
},
{
"grad_norm": 0.46862635016441345,
"learning_rate": 9.707755139734855e-05,
"loss": 0.0575,
"step": 1540
},
{
"grad_norm": 0.2976231276988983,
"learning_rate": 9.702159357983866e-05,
"loss": 0.0541,
"step": 1550
},
{
"grad_norm": 0.3357617259025574,
"learning_rate": 9.696512154171492e-05,
"loss": 0.0549,
"step": 1560
},
{
"grad_norm": 0.32237687706947327,
"learning_rate": 9.690813590054645e-05,
"loss": 0.0543,
"step": 1570
},
{
"grad_norm": 0.45769360661506653,
"learning_rate": 9.685063727951914e-05,
"loss": 0.055,
"step": 1580
},
{
"grad_norm": 0.4333069920539856,
"learning_rate": 9.679262630742865e-05,
"loss": 0.057,
"step": 1590
},
{
"grad_norm": 0.4686541259288788,
"learning_rate": 9.673410361867373e-05,
"loss": 0.0561,
"step": 1600
},
{
"grad_norm": 0.45229238271713257,
"learning_rate": 9.667506985324909e-05,
"loss": 0.0581,
"step": 1610
},
{
"grad_norm": 0.40434762835502625,
"learning_rate": 9.661552565673855e-05,
"loss": 0.053,
"step": 1620
},
{
"grad_norm": 0.3540216386318207,
"learning_rate": 9.655547168030789e-05,
"loss": 0.0556,
"step": 1630
},
{
"grad_norm": 0.47013527154922485,
"learning_rate": 9.649490858069777e-05,
"loss": 0.0568,
"step": 1640
},
{
"grad_norm": 0.4146243929862976,
"learning_rate": 9.643383702021658e-05,
"loss": 0.0626,
"step": 1650
},
{
"grad_norm": 0.6327381730079651,
"learning_rate": 9.637225766673307e-05,
"loss": 0.0506,
"step": 1660
},
{
"grad_norm": 0.3726016879081726,
"learning_rate": 9.631017119366922e-05,
"loss": 0.0591,
"step": 1670
},
{
"grad_norm": 0.34331509470939636,
"learning_rate": 9.624757827999273e-05,
"loss": 0.0586,
"step": 1680
},
{
"grad_norm": 0.3801133334636688,
"learning_rate": 9.618447961020971e-05,
"loss": 0.0585,
"step": 1690
},
{
"grad_norm": 0.41130298376083374,
"learning_rate": 9.612087587435707e-05,
"loss": 0.0551,
"step": 1700
},
{
"grad_norm": 0.35909804701805115,
"learning_rate": 9.605676776799508e-05,
"loss": 0.0513,
"step": 1710
},
{
"grad_norm": 0.4176706373691559,
"learning_rate": 9.599215599219973e-05,
"loss": 0.0562,
"step": 1720
},
{
"grad_norm": 0.3385564386844635,
"learning_rate": 9.592704125355505e-05,
"loss": 0.0546,
"step": 1730
},
{
"grad_norm": 0.29991668462753296,
"learning_rate": 9.586142426414538e-05,
"loss": 0.051,
"step": 1740
},
{
"grad_norm": 0.4050476551055908,
"learning_rate": 9.57953057415476e-05,
"loss": 0.053,
"step": 1750
},
{
"grad_norm": 0.3702267110347748,
"learning_rate": 9.572868640882328e-05,
"loss": 0.0657,
"step": 1760
},
{
"grad_norm": 0.34253689646720886,
"learning_rate": 9.56615669945108e-05,
"loss": 0.0541,
"step": 1770
},
{
"grad_norm": 0.40993639826774597,
"learning_rate": 9.55939482326173e-05,
"loss": 0.0483,
"step": 1780
},
{
"grad_norm": 0.37866318225860596,
"learning_rate": 9.552583086261069e-05,
"loss": 0.0621,
"step": 1790
},
{
"grad_norm": 0.43934744596481323,
"learning_rate": 9.545721562941168e-05,
"loss": 0.0504,
"step": 1800
},
{
"grad_norm": 0.3429582417011261,
"learning_rate": 9.538810328338543e-05,
"loss": 0.0535,
"step": 1810
},
{
"grad_norm": 0.39602530002593994,
"learning_rate": 9.531849458033349e-05,
"loss": 0.0508,
"step": 1820
},
{
"grad_norm": 0.2815093994140625,
"learning_rate": 9.524839028148547e-05,
"loss": 0.0538,
"step": 1830
},
{
"grad_norm": 0.29502665996551514,
"learning_rate": 9.517779115349077e-05,
"loss": 0.0461,
"step": 1840
},
{
"grad_norm": 0.3675382733345032,
"learning_rate": 9.510669796841014e-05,
"loss": 0.0504,
"step": 1850
},
{
"grad_norm": 0.3640974462032318,
"learning_rate": 9.503511150370727e-05,
"loss": 0.0529,
"step": 1860
},
{
"grad_norm": 0.3881661891937256,
"learning_rate": 9.496303254224024e-05,
"loss": 0.0496,
"step": 1870
},
{
"grad_norm": 0.4098767638206482,
"learning_rate": 9.489046187225306e-05,
"loss": 0.0467,
"step": 1880
},
{
"grad_norm": 0.3381575345993042,
"learning_rate": 9.481740028736692e-05,
"loss": 0.0477,
"step": 1890
},
{
"grad_norm": 0.3819166421890259,
"learning_rate": 9.474384858657164e-05,
"loss": 0.0562,
"step": 1900
},
{
"grad_norm": 0.44135865569114685,
"learning_rate": 9.466980757421679e-05,
"loss": 0.0528,
"step": 1910
},
{
"grad_norm": 0.3618215024471283,
"learning_rate": 9.459527806000305e-05,
"loss": 0.0507,
"step": 1920
},
{
"grad_norm": 0.3703959584236145,
"learning_rate": 9.452026085897325e-05,
"loss": 0.0508,
"step": 1930
},
{
"grad_norm": 0.28048086166381836,
"learning_rate": 9.444475679150348e-05,
"loss": 0.0495,
"step": 1940
},
{
"grad_norm": 0.32608762383461,
"learning_rate": 9.436876668329411e-05,
"loss": 0.0511,
"step": 1950
},
{
"grad_norm": 0.31372350454330444,
"learning_rate": 9.429229136536079e-05,
"loss": 0.0419,
"step": 1960
},
{
"grad_norm": 0.49752604961395264,
"learning_rate": 9.421533167402534e-05,
"loss": 0.0519,
"step": 1970
},
{
"grad_norm": 0.6253518462181091,
"learning_rate": 9.413788845090666e-05,
"loss": 0.0615,
"step": 1980
},
{
"grad_norm": 0.5852185487747192,
"learning_rate": 9.405996254291136e-05,
"loss": 0.0552,
"step": 1990
},
{
"grad_norm": 0.38057762384414673,
"learning_rate": 9.398155480222474e-05,
"loss": 0.0494,
"step": 2000
},
{
"grad_norm": 0.2937912940979004,
"learning_rate": 9.390266608630128e-05,
"loss": 0.0436,
"step": 2010
},
{
"grad_norm": 0.4695729613304138,
"learning_rate": 9.38232972578553e-05,
"loss": 0.0502,
"step": 2020
},
{
"grad_norm": 0.27542877197265625,
"learning_rate": 9.374344918485164e-05,
"loss": 0.0528,
"step": 2030
},
{
"grad_norm": 0.3729042410850525,
"learning_rate": 9.366312274049602e-05,
"loss": 0.0558,
"step": 2040
},
{
"grad_norm": 0.4016962945461273,
"learning_rate": 9.358231880322554e-05,
"loss": 0.0497,
"step": 2050
},
{
"grad_norm": 0.3093472123146057,
"learning_rate": 9.350103825669916e-05,
"loss": 0.046,
"step": 2060
},
{
"grad_norm": 0.3629983365535736,
"learning_rate": 9.341928198978787e-05,
"loss": 0.0453,
"step": 2070
},
{
"grad_norm": 0.343095988035202,
"learning_rate": 9.333705089656512e-05,
"loss": 0.0502,
"step": 2080
},
{
"grad_norm": 0.24454210698604584,
"learning_rate": 9.325434587629698e-05,
"loss": 0.044,
"step": 2090
},
{
"grad_norm": 0.380744993686676,
"learning_rate": 9.31711678334323e-05,
"loss": 0.0484,
"step": 2100
},
{
"grad_norm": 0.35298046469688416,
"learning_rate": 9.308751767759282e-05,
"loss": 0.048,
"step": 2110
},
{
"grad_norm": 0.355131059885025,
"learning_rate": 9.300339632356325e-05,
"loss": 0.0474,
"step": 2120
},
{
"grad_norm": 0.379371702671051,
"learning_rate": 9.291880469128124e-05,
"loss": 0.0484,
"step": 2130
},
{
"grad_norm": 0.5716716051101685,
"learning_rate": 9.283374370582732e-05,
"loss": 0.0488,
"step": 2140
},
{
"grad_norm": 0.3022350072860718,
"learning_rate": 9.274821429741482e-05,
"loss": 0.0563,
"step": 2150
},
{
"grad_norm": 0.404402494430542,
"learning_rate": 9.266221740137961e-05,
"loss": 0.0467,
"step": 2160
},
{
"grad_norm": 0.21346545219421387,
"learning_rate": 9.257575395817001e-05,
"loss": 0.0451,
"step": 2170
},
{
"grad_norm": 0.3357575535774231,
"learning_rate": 9.248882491333637e-05,
"loss": 0.045,
"step": 2180
},
{
"grad_norm": 0.3501649498939514,
"learning_rate": 9.240143121752076e-05,
"loss": 0.0427,
"step": 2190
},
{
"grad_norm": 0.22892490029335022,
"learning_rate": 9.23135738264467e-05,
"loss": 0.0482,
"step": 2200
},
{
"grad_norm": 0.4455818831920624,
"learning_rate": 9.222525370090849e-05,
"loss": 0.0478,
"step": 2210
},
{
"grad_norm": 0.26360446214675903,
"learning_rate": 9.213647180676088e-05,
"loss": 0.0464,
"step": 2220
},
{
"grad_norm": 0.33578574657440186,
"learning_rate": 9.204722911490846e-05,
"loss": 0.0506,
"step": 2230
},
{
"grad_norm": 0.42269811034202576,
"learning_rate": 9.1957526601295e-05,
"loss": 0.0474,
"step": 2240
},
{
"grad_norm": 0.4483242332935333,
"learning_rate": 9.186736524689281e-05,
"loss": 0.0508,
"step": 2250
},
{
"grad_norm": 0.3417402505874634,
"learning_rate": 9.177674603769204e-05,
"loss": 0.0487,
"step": 2260
},
{
"grad_norm": 0.32713067531585693,
"learning_rate": 9.168566996468983e-05,
"loss": 0.0451,
"step": 2270
},
{
"grad_norm": 0.277864933013916,
"learning_rate": 9.159413802387951e-05,
"loss": 0.0472,
"step": 2280
},
{
"grad_norm": 0.38279271125793457,
"learning_rate": 9.150215121623974e-05,
"loss": 0.0461,
"step": 2290
},
{
"grad_norm": 0.40508410334587097,
"learning_rate": 9.140971054772349e-05,
"loss": 0.0419,
"step": 2300
},
{
"grad_norm": 0.335996150970459,
"learning_rate": 9.131681702924713e-05,
"loss": 0.0511,
"step": 2310
},
{
"grad_norm": 0.31667858362197876,
"learning_rate": 9.122347167667926e-05,
"loss": 0.0475,
"step": 2320
},
{
"grad_norm": 0.5489864945411682,
"learning_rate": 9.112967551082973e-05,
"loss": 0.0607,
"step": 2330
},
{
"grad_norm": 0.47898566722869873,
"learning_rate": 9.103542955743835e-05,
"loss": 0.0506,
"step": 2340
},
{
"grad_norm": 0.5498167276382446,
"learning_rate": 9.094073484716381e-05,
"loss": 0.0583,
"step": 2350
},
{
"grad_norm": 0.3407987058162689,
"learning_rate": 9.084559241557226e-05,
"loss": 0.0476,
"step": 2360
},
{
"grad_norm": 0.34748774766921997,
"learning_rate": 9.075000330312608e-05,
"loss": 0.0503,
"step": 2370
},
{
"grad_norm": 0.32787489891052246,
"learning_rate": 9.065396855517253e-05,
"loss": 0.0449,
"step": 2380
},
{
"grad_norm": 0.43296945095062256,
"learning_rate": 9.055748922193219e-05,
"loss": 0.0523,
"step": 2390
},
{
"grad_norm": 0.37504345178604126,
"learning_rate": 9.046056635848761e-05,
"loss": 0.0528,
"step": 2400
},
{
"grad_norm": 0.34991011023521423,
"learning_rate": 9.036320102477169e-05,
"loss": 0.0443,
"step": 2410
},
{
"grad_norm": 0.5025856494903564,
"learning_rate": 9.02653942855561e-05,
"loss": 0.0475,
"step": 2420
},
{
"grad_norm": 0.3575250804424286,
"learning_rate": 9.016714721043971e-05,
"loss": 0.0509,
"step": 2430
},
{
"grad_norm": 0.44044655561447144,
"learning_rate": 9.006846087383675e-05,
"loss": 0.0544,
"step": 2440
},
{
"grad_norm": 0.3654404282569885,
"learning_rate": 8.996933635496523e-05,
"loss": 0.0487,
"step": 2450
},
{
"grad_norm": 0.36741903424263,
"learning_rate": 8.986977473783498e-05,
"loss": 0.0576,
"step": 2460
},
{
"grad_norm": 0.3331684172153473,
"learning_rate": 8.97697771112359e-05,
"loss": 0.0488,
"step": 2470
},
{
"grad_norm": 0.322755366563797,
"learning_rate": 8.966934456872602e-05,
"loss": 0.0464,
"step": 2480
},
{
"grad_norm": 0.3070371448993683,
"learning_rate": 8.95684782086195e-05,
"loss": 0.0396,
"step": 2490
},
{
"grad_norm": 0.35789644718170166,
"learning_rate": 8.946717913397476e-05,
"loss": 0.05,
"step": 2500
},
{
"grad_norm": 0.3882046937942505,
"learning_rate": 8.93654484525822e-05,
"loss": 0.043,
"step": 2510
},
{
"grad_norm": 0.3256243169307709,
"learning_rate": 8.926328727695226e-05,
"loss": 0.0461,
"step": 2520
},
{
"grad_norm": 0.3392179608345032,
"learning_rate": 8.916069672430319e-05,
"loss": 0.0418,
"step": 2530
},
{
"grad_norm": 0.40785473585128784,
"learning_rate": 8.905767791654884e-05,
"loss": 0.0465,
"step": 2540
},
{
"grad_norm": 0.3217814862728119,
"learning_rate": 8.895423198028638e-05,
"loss": 0.0509,
"step": 2550
},
{
"grad_norm": 0.29712602496147156,
"learning_rate": 8.885036004678402e-05,
"loss": 0.0458,
"step": 2560
},
{
"grad_norm": 0.30837199091911316,
"learning_rate": 8.874606325196857e-05,
"loss": 0.0482,
"step": 2570
},
{
"grad_norm": 0.4253198206424713,
"learning_rate": 8.864134273641304e-05,
"loss": 0.0492,
"step": 2580
},
{
"grad_norm": 0.3407295346260071,
"learning_rate": 8.853619964532427e-05,
"loss": 0.0484,
"step": 2590
},
{
"grad_norm": 0.2488165944814682,
"learning_rate": 8.843063512853019e-05,
"loss": 0.0436,
"step": 2600
},
{
"grad_norm": 0.25318872928619385,
"learning_rate": 8.832465034046749e-05,
"loss": 0.0423,
"step": 2610
},
{
"grad_norm": 0.42971646785736084,
"learning_rate": 8.821824644016882e-05,
"loss": 0.049,
"step": 2620
},
{
"grad_norm": 0.2790759205818176,
"learning_rate": 8.811142459125019e-05,
"loss": 0.045,
"step": 2630
},
{
"grad_norm": 0.2801419198513031,
"learning_rate": 8.800418596189822e-05,
"loss": 0.0422,
"step": 2640
},
{
"grad_norm": 0.4057613015174866,
"learning_rate": 8.789653172485737e-05,
"loss": 0.0484,
"step": 2650
},
{
"grad_norm": 0.38167011737823486,
"learning_rate": 8.778846305741715e-05,
"loss": 0.0501,
"step": 2660
},
{
"grad_norm": 0.35908186435699463,
"learning_rate": 8.767998114139918e-05,
"loss": 0.048,
"step": 2670
},
{
"grad_norm": 0.3967783749103546,
"learning_rate": 8.757108716314429e-05,
"loss": 0.0544,
"step": 2680
},
{
"grad_norm": 0.28078755736351013,
"learning_rate": 8.746178231349962e-05,
"loss": 0.0437,
"step": 2690
},
{
"grad_norm": 0.3339881896972656,
"learning_rate": 8.735206778780549e-05,
"loss": 0.0505,
"step": 2700
},
{
"grad_norm": 0.20266804099082947,
"learning_rate": 8.724194478588234e-05,
"loss": 0.0452,
"step": 2710
},
{
"grad_norm": 0.2945866882801056,
"learning_rate": 8.713141451201772e-05,
"loss": 0.0441,
"step": 2720
},
{
"grad_norm": 0.29569557309150696,
"learning_rate": 8.702047817495295e-05,
"loss": 0.0431,
"step": 2730
},
{
"grad_norm": 0.38597384095191956,
"learning_rate": 8.69091369878701e-05,
"loss": 0.0455,
"step": 2740
},
{
"grad_norm": 0.2877079248428345,
"learning_rate": 8.679739216837849e-05,
"loss": 0.0395,
"step": 2750
},
{
"grad_norm": 0.4120080769062042,
"learning_rate": 8.66852449385016e-05,
"loss": 0.0445,
"step": 2760
},
{
"grad_norm": 0.4648508131504059,
"learning_rate": 8.657269652466356e-05,
"loss": 0.0436,
"step": 2770
},
{
"grad_norm": 0.31352418661117554,
"learning_rate": 8.645974815767577e-05,
"loss": 0.0441,
"step": 2780
},
{
"grad_norm": 0.3555717468261719,
"learning_rate": 8.634640107272351e-05,
"loss": 0.039,
"step": 2790
},
{
"grad_norm": 0.27151474356651306,
"learning_rate": 8.623265650935234e-05,
"loss": 0.0426,
"step": 2800
},
{
"grad_norm": 0.37681901454925537,
"learning_rate": 8.611851571145456e-05,
"loss": 0.0454,
"step": 2810
},
{
"grad_norm": 0.2861371636390686,
"learning_rate": 8.600397992725566e-05,
"loss": 0.0451,
"step": 2820
},
{
"grad_norm": 0.29237791895866394,
"learning_rate": 8.588905040930061e-05,
"loss": 0.0427,
"step": 2830
},
{
"grad_norm": 0.35147255659103394,
"learning_rate": 8.577372841444022e-05,
"loss": 0.0443,
"step": 2840
},
{
"grad_norm": 0.359683632850647,
"learning_rate": 8.565801520381736e-05,
"loss": 0.043,
"step": 2850
},
{
"grad_norm": 0.3914463222026825,
"learning_rate": 8.554191204285313e-05,
"loss": 0.0436,
"step": 2860
},
{
"grad_norm": 0.44010627269744873,
"learning_rate": 8.542542020123315e-05,
"loss": 0.0427,
"step": 2870
},
{
"grad_norm": 0.3706228733062744,
"learning_rate": 8.530854095289347e-05,
"loss": 0.0451,
"step": 2880
},
{
"grad_norm": 0.2067432552576065,
"learning_rate": 8.519127557600688e-05,
"loss": 0.0445,
"step": 2890
},
{
"grad_norm": 0.3595845699310303,
"learning_rate": 8.507362535296871e-05,
"loss": 0.0443,
"step": 2900
},
{
"grad_norm": 0.2836382985115051,
"learning_rate": 8.495559157038299e-05,
"loss": 0.0516,
"step": 2910
},
{
"grad_norm": 0.4540116786956787,
"learning_rate": 8.483717551904823e-05,
"loss": 0.0457,
"step": 2920
},
{
"grad_norm": 0.4011117219924927,
"learning_rate": 8.47183784939434e-05,
"loss": 0.0424,
"step": 2930
},
{
"grad_norm": 0.3454821705818176,
"learning_rate": 8.459920179421374e-05,
"loss": 0.0434,
"step": 2940
},
{
"grad_norm": 0.4366862177848816,
"learning_rate": 8.447964672315656e-05,
"loss": 0.045,
"step": 2950
},
{
"grad_norm": 0.3353109359741211,
"learning_rate": 8.435971458820692e-05,
"loss": 0.052,
"step": 2960
},
{
"grad_norm": 0.34972432255744934,
"learning_rate": 8.423940670092345e-05,
"loss": 0.0483,
"step": 2970
},
{
"grad_norm": 0.3265519142150879,
"learning_rate": 8.411872437697394e-05,
"loss": 0.0428,
"step": 2980
},
{
"grad_norm": 0.3052782416343689,
"learning_rate": 8.399766893612096e-05,
"loss": 0.04,
"step": 2990
},
{
"grad_norm": 0.32262122631073,
"learning_rate": 8.38762417022074e-05,
"loss": 0.0475,
"step": 3000
},
{
"grad_norm": 0.4137900471687317,
"learning_rate": 8.375444400314204e-05,
"loss": 0.0411,
"step": 3010
},
{
"grad_norm": 0.3818408250808716,
"learning_rate": 8.3632277170885e-05,
"loss": 0.048,
"step": 3020
},
{
"grad_norm": 0.4213651418685913,
"learning_rate": 8.350974254143318e-05,
"loss": 0.0414,
"step": 3030
},
{
"grad_norm": 0.2937682569026947,
"learning_rate": 8.338684145480566e-05,
"loss": 0.0398,
"step": 3040
},
{
"grad_norm": 0.3058902323246002,
"learning_rate": 8.326357525502904e-05,
"loss": 0.047,
"step": 3050
},
{
"grad_norm": 0.2876177728176117,
"learning_rate": 8.313994529012273e-05,
"loss": 0.043,
"step": 3060
},
{
"grad_norm": 0.33672255277633667,
"learning_rate": 8.301595291208422e-05,
"loss": 0.0441,
"step": 3070
},
{
"grad_norm": 0.27741292119026184,
"learning_rate": 8.289159947687427e-05,
"loss": 0.0432,
"step": 3080
},
{
"grad_norm": 0.29131191968917847,
"learning_rate": 8.276688634440216e-05,
"loss": 0.0429,
"step": 3090
},
{
"grad_norm": 0.29063108563423157,
"learning_rate": 8.26418148785107e-05,
"loss": 0.0425,
"step": 3100
},
{
"grad_norm": 0.25788575410842896,
"learning_rate": 8.251638644696141e-05,
"loss": 0.048,
"step": 3110
},
{
"grad_norm": 0.35483136773109436,
"learning_rate": 8.23906024214195e-05,
"loss": 0.0425,
"step": 3120
},
{
"grad_norm": 0.3214036822319031,
"learning_rate": 8.226446417743897e-05,
"loss": 0.0464,
"step": 3130
},
{
"grad_norm": 0.2682476341724396,
"learning_rate": 8.213797309444742e-05,
"loss": 0.0435,
"step": 3140
},
{
"grad_norm": 0.2817361354827881,
"learning_rate": 8.201113055573105e-05,
"loss": 0.0433,
"step": 3150
},
{
"grad_norm": 0.3765580952167511,
"learning_rate": 8.188393794841958e-05,
"loss": 0.0428,
"step": 3160
},
{
"grad_norm": 0.3851284384727478,
"learning_rate": 8.175639666347094e-05,
"loss": 0.0459,
"step": 3170
},
{
"grad_norm": 0.3516382575035095,
"learning_rate": 8.162850809565623e-05,
"loss": 0.0387,
"step": 3180
},
{
"grad_norm": 0.23620860278606415,
"learning_rate": 8.150027364354431e-05,
"loss": 0.0426,
"step": 3190
},
{
"grad_norm": 0.4271146059036255,
"learning_rate": 8.137169470948662e-05,
"loss": 0.0429,
"step": 3200
},
{
"grad_norm": 0.38874682784080505,
"learning_rate": 8.124277269960179e-05,
"loss": 0.0436,
"step": 3210
},
{
"grad_norm": 0.2055014967918396,
"learning_rate": 8.111350902376023e-05,
"loss": 0.0384,
"step": 3220
},
{
"grad_norm": 0.21492816507816315,
"learning_rate": 8.098390509556883e-05,
"loss": 0.0394,
"step": 3230
},
{
"grad_norm": 0.3927494287490845,
"learning_rate": 8.085396233235536e-05,
"loss": 0.0368,
"step": 3240
},
{
"grad_norm": 0.31042924523353577,
"learning_rate": 8.072368215515306e-05,
"loss": 0.0412,
"step": 3250
},
{
"grad_norm": 0.32237017154693604,
"learning_rate": 8.059306598868506e-05,
"loss": 0.0475,
"step": 3260
},
{
"grad_norm": 0.32005056738853455,
"learning_rate": 8.046211526134888e-05,
"loss": 0.043,
"step": 3270
},
{
"grad_norm": 0.2511746287345886,
"learning_rate": 8.033083140520065e-05,
"loss": 0.0392,
"step": 3280
},
{
"grad_norm": 0.37887656688690186,
"learning_rate": 8.019921585593962e-05,
"loss": 0.0416,
"step": 3290
},
{
"grad_norm": 0.2693057358264923,
"learning_rate": 8.006727005289232e-05,
"loss": 0.0399,
"step": 3300
},
{
"grad_norm": 0.2650562524795532,
"learning_rate": 7.993499543899692e-05,
"loss": 0.0406,
"step": 3310
},
{
"grad_norm": 0.20058368146419525,
"learning_rate": 7.980239346078742e-05,
"loss": 0.0382,
"step": 3320
},
{
"grad_norm": 0.2906408905982971,
"learning_rate": 7.966946556837778e-05,
"loss": 0.042,
"step": 3330
},
{
"grad_norm": 0.3585576117038727,
"learning_rate": 7.953621321544616e-05,
"loss": 0.0397,
"step": 3340
},
{
"grad_norm": 0.4038415849208832,
"learning_rate": 7.940263785921896e-05,
"loss": 0.0425,
"step": 3350
},
{
"grad_norm": 0.37583935260772705,
"learning_rate": 7.926874096045482e-05,
"loss": 0.0432,
"step": 3360
},
{
"grad_norm": 0.35287800431251526,
"learning_rate": 7.913452398342881e-05,
"loss": 0.0497,
"step": 3370
},
{
"grad_norm": 0.2622617185115814,
"learning_rate": 7.89999883959163e-05,
"loss": 0.0456,
"step": 3380
},
{
"grad_norm": 0.31066831946372986,
"learning_rate": 7.886513566917687e-05,
"loss": 0.0493,
"step": 3390
},
{
"grad_norm": 0.1938023716211319,
"learning_rate": 7.872996727793838e-05,
"loss": 0.0502,
"step": 3400
},
{
"grad_norm": 0.34381890296936035,
"learning_rate": 7.859448470038069e-05,
"loss": 0.0393,
"step": 3410
},
{
"grad_norm": 0.25354310870170593,
"learning_rate": 7.845868941811956e-05,
"loss": 0.0454,
"step": 3420
},
{
"grad_norm": 0.3401482403278351,
"learning_rate": 7.832258291619043e-05,
"loss": 0.0407,
"step": 3430
},
{
"grad_norm": 0.4771004915237427,
"learning_rate": 7.81861666830322e-05,
"loss": 0.0458,
"step": 3440
},
{
"grad_norm": 0.275635302066803,
"learning_rate": 7.804944221047097e-05,
"loss": 0.0394,
"step": 3450
},
{
"grad_norm": 0.25738394260406494,
"learning_rate": 7.791241099370364e-05,
"loss": 0.0476,
"step": 3460
},
{
"grad_norm": 0.2737746834754944,
"learning_rate": 7.777507453128163e-05,
"loss": 0.0376,
"step": 3470
},
{
"grad_norm": 0.24491152167320251,
"learning_rate": 7.763743432509451e-05,
"loss": 0.0413,
"step": 3480
},
{
"grad_norm": 0.293349027633667,
"learning_rate": 7.749949188035353e-05,
"loss": 0.0412,
"step": 3490
},
{
"grad_norm": 0.32892847061157227,
"learning_rate": 7.736124870557516e-05,
"loss": 0.0493,
"step": 3500
},
{
"grad_norm": 0.2998802661895752,
"learning_rate": 7.722270631256459e-05,
"loss": 0.0443,
"step": 3510
},
{
"grad_norm": 0.28482547402381897,
"learning_rate": 7.708386621639925e-05,
"loss": 0.0386,
"step": 3520
},
{
"grad_norm": 0.2953447997570038,
"learning_rate": 7.694472993541219e-05,
"loss": 0.0401,
"step": 3530
},
{
"grad_norm": 0.2643677592277527,
"learning_rate": 7.680529899117547e-05,
"loss": 0.0422,
"step": 3540
},
{
"grad_norm": 0.3331605792045593,
"learning_rate": 7.666557490848358e-05,
"loss": 0.0407,
"step": 3550
},
{
"grad_norm": 0.24552491307258606,
"learning_rate": 7.65255592153367e-05,
"loss": 0.0405,
"step": 3560
},
{
"grad_norm": 0.32412147521972656,
"learning_rate": 7.638525344292402e-05,
"loss": 0.0428,
"step": 3570
},
{
"grad_norm": 0.3231714069843292,
"learning_rate": 7.624465912560697e-05,
"loss": 0.048,
"step": 3580
},
{
"grad_norm": 0.28266048431396484,
"learning_rate": 7.610377780090249e-05,
"loss": 0.0431,
"step": 3590
},
{
"grad_norm": 0.34312570095062256,
"learning_rate": 7.596261100946618e-05,
"loss": 0.0463,
"step": 3600
},
{
"grad_norm": 0.28578826785087585,
"learning_rate": 7.582116029507542e-05,
"loss": 0.0374,
"step": 3610
},
{
"grad_norm": 0.21435998380184174,
"learning_rate": 7.56794272046126e-05,
"loss": 0.0424,
"step": 3620
},
{
"grad_norm": 0.2694740891456604,
"learning_rate": 7.55374132880481e-05,
"loss": 0.038,
"step": 3630
},
{
"grad_norm": 0.25812557339668274,
"learning_rate": 7.539512009842333e-05,
"loss": 0.0425,
"step": 3640
},
{
"grad_norm": 0.28829240798950195,
"learning_rate": 7.525254919183382e-05,
"loss": 0.0437,
"step": 3650
},
{
"grad_norm": 0.2935335040092468,
"learning_rate": 7.510970212741215e-05,
"loss": 0.0337,
"step": 3660
},
{
"grad_norm": 0.3696131706237793,
"learning_rate": 7.496658046731096e-05,
"loss": 0.0395,
"step": 3670
},
{
"grad_norm": 0.33474403619766235,
"learning_rate": 7.482318577668578e-05,
"loss": 0.044,
"step": 3680
},
{
"grad_norm": 0.39132723212242126,
"learning_rate": 7.467951962367796e-05,
"loss": 0.0463,
"step": 3690
},
{
"grad_norm": 0.3159390389919281,
"learning_rate": 7.453558357939755e-05,
"loss": 0.0425,
"step": 3700
},
{
"grad_norm": 0.25400304794311523,
"learning_rate": 7.439137921790606e-05,
"loss": 0.0498,
"step": 3710
},
{
"grad_norm": 0.2740709185600281,
"learning_rate": 7.42469081161993e-05,
"loss": 0.0452,
"step": 3720
},
{
"grad_norm": 0.2602589428424835,
"learning_rate": 7.410217185419006e-05,
"loss": 0.0372,
"step": 3730
},
{
"grad_norm": 0.2620036005973816,
"learning_rate": 7.395717201469095e-05,
"loss": 0.0457,
"step": 3740
},
{
"grad_norm": 0.24679666757583618,
"learning_rate": 7.381191018339696e-05,
"loss": 0.0411,
"step": 3750
},
{
"grad_norm": 0.3610756993293762,
"learning_rate": 7.36663879488682e-05,
"loss": 0.0416,
"step": 3760
},
{
"grad_norm": 0.3013175427913666,
"learning_rate": 7.352060690251254e-05,
"loss": 0.0461,
"step": 3770
},
{
"grad_norm": 0.1743604838848114,
"learning_rate": 7.337456863856811e-05,
"loss": 0.0399,
"step": 3780
},
{
"grad_norm": 0.3789909780025482,
"learning_rate": 7.3228274754086e-05,
"loss": 0.039,
"step": 3790
},
{
"grad_norm": 0.2807275354862213,
"learning_rate": 7.308172684891267e-05,
"loss": 0.0427,
"step": 3800
},
{
"grad_norm": 0.2820431590080261,
"learning_rate": 7.293492652567255e-05,
"loss": 0.0365,
"step": 3810
},
{
"grad_norm": 0.20893006026744843,
"learning_rate": 7.278787538975043e-05,
"loss": 0.0412,
"step": 3820
},
{
"grad_norm": 0.42322054505348206,
"learning_rate": 7.2640575049274e-05,
"loss": 0.0446,
"step": 3830
},
{
"grad_norm": 0.3860882818698883,
"learning_rate": 7.249302711509616e-05,
"loss": 0.0474,
"step": 3840
},
{
"grad_norm": 0.29109442234039307,
"learning_rate": 7.23452332007775e-05,
"loss": 0.0393,
"step": 3850
},
{
"grad_norm": 0.3240247964859009,
"learning_rate": 7.219719492256858e-05,
"loss": 0.044,
"step": 3860
},
{
"grad_norm": 0.2614988386631012,
"learning_rate": 7.20489138993923e-05,
"loss": 0.0411,
"step": 3870
},
{
"grad_norm": 0.22772513329982758,
"learning_rate": 7.190039175282614e-05,
"loss": 0.0414,
"step": 3880
},
{
"grad_norm": 0.1992558240890503,
"learning_rate": 7.175163010708455e-05,
"loss": 0.0409,
"step": 3890
},
{
"grad_norm": 0.19291351735591888,
"learning_rate": 7.1602630589001e-05,
"loss": 0.0385,
"step": 3900
},
{
"grad_norm": 0.18126238882541656,
"learning_rate": 7.14533948280104e-05,
"loss": 0.0401,
"step": 3910
},
{
"grad_norm": 0.32571858167648315,
"learning_rate": 7.130392445613109e-05,
"loss": 0.0388,
"step": 3920
},
{
"grad_norm": 0.33397889137268066,
"learning_rate": 7.115422110794711e-05,
"loss": 0.041,
"step": 3930
},
{
"grad_norm": 0.3632291555404663,
"learning_rate": 7.100428642059033e-05,
"loss": 0.0414,
"step": 3940
},
{
"grad_norm": 0.3965553641319275,
"learning_rate": 7.08541220337224e-05,
"loss": 0.0419,
"step": 3950
},
{
"grad_norm": 0.3390904366970062,
"learning_rate": 7.070372958951706e-05,
"loss": 0.0415,
"step": 3960
},
{
"grad_norm": 0.18588070571422577,
"learning_rate": 7.055311073264194e-05,
"loss": 0.0392,
"step": 3970
},
{
"grad_norm": 0.24716848134994507,
"learning_rate": 7.040226711024077e-05,
"loss": 0.0384,
"step": 3980
},
{
"grad_norm": 0.267456591129303,
"learning_rate": 7.02512003719152e-05,
"loss": 0.0359,
"step": 3990
},
{
"grad_norm": 0.22239471971988678,
"learning_rate": 7.00999121697069e-05,
"loss": 0.0408,
"step": 4000
},
{
"grad_norm": 0.2416866421699524,
"learning_rate": 6.99484041580794e-05,
"loss": 0.0357,
"step": 4010
},
{
"grad_norm": 0.25899538397789,
"learning_rate": 6.979667799390004e-05,
"loss": 0.0409,
"step": 4020
},
{
"grad_norm": 0.33594587445259094,
"learning_rate": 6.964473533642185e-05,
"loss": 0.0419,
"step": 4030
},
{
"grad_norm": 0.2432696521282196,
"learning_rate": 6.949257784726539e-05,
"loss": 0.0399,
"step": 4040
},
{
"grad_norm": 0.24899610877037048,
"learning_rate": 6.934020719040056e-05,
"loss": 0.0354,
"step": 4050
},
{
"grad_norm": 0.2921767234802246,
"learning_rate": 6.918762503212848e-05,
"loss": 0.0431,
"step": 4060
},
{
"grad_norm": 0.29085418581962585,
"learning_rate": 6.903483304106319e-05,
"loss": 0.0408,
"step": 4070
},
{
"grad_norm": 0.274190753698349,
"learning_rate": 6.888183288811341e-05,
"loss": 0.0348,
"step": 4080
},
{
"grad_norm": 0.2501881420612335,
"learning_rate": 6.87286262464643e-05,
"loss": 0.0414,
"step": 4090
},
{
"grad_norm": 0.3673791289329529,
"learning_rate": 6.857521479155915e-05,
"loss": 0.0434,
"step": 4100
},
{
"grad_norm": 0.3201288878917694,
"learning_rate": 6.842160020108104e-05,
"loss": 0.0437,
"step": 4110
},
{
"grad_norm": 0.37830421328544617,
"learning_rate": 6.826778415493455e-05,
"loss": 0.0391,
"step": 4120
},
{
"grad_norm": 0.38764116168022156,
"learning_rate": 6.811376833522729e-05,
"loss": 0.0442,
"step": 4130
},
{
"grad_norm": 0.31950128078460693,
"learning_rate": 6.795955442625159e-05,
"loss": 0.0425,
"step": 4140
},
{
"grad_norm": 0.2520832121372223,
"learning_rate": 6.780514411446608e-05,
"loss": 0.0428,
"step": 4150
},
{
"grad_norm": 0.33493414521217346,
"learning_rate": 6.765053908847716e-05,
"loss": 0.0406,
"step": 4160
},
{
"grad_norm": 0.3129543364048004,
"learning_rate": 6.749574103902064e-05,
"loss": 0.0436,
"step": 4170
},
{
"grad_norm": 0.4778180718421936,
"learning_rate": 6.734075165894317e-05,
"loss": 0.0427,
"step": 4180
},
{
"grad_norm": 0.2949196696281433,
"learning_rate": 6.71855726431838e-05,
"loss": 0.0458,
"step": 4190
},
{
"grad_norm": 0.23823823034763336,
"learning_rate": 6.703020568875538e-05,
"loss": 0.0382,
"step": 4200
},
{
"grad_norm": 0.20042912662029266,
"learning_rate": 6.687465249472603e-05,
"loss": 0.0346,
"step": 4210
},
{
"grad_norm": 0.3149445354938507,
"learning_rate": 6.671891476220055e-05,
"loss": 0.0383,
"step": 4220
},
{
"grad_norm": 0.3420006036758423,
"learning_rate": 6.656299419430183e-05,
"loss": 0.0364,
"step": 4230
},
{
"grad_norm": 0.28867608308792114,
"learning_rate": 6.640689249615223e-05,
"loss": 0.0403,
"step": 4240
},
{
"grad_norm": 0.24635563790798187,
"learning_rate": 6.625061137485491e-05,
"loss": 0.0421,
"step": 4250
},
{
"grad_norm": 0.27294376492500305,
"learning_rate": 6.609415253947517e-05,
"loss": 0.0375,
"step": 4260
},
{
"grad_norm": 0.34344911575317383,
"learning_rate": 6.593751770102178e-05,
"loss": 0.0392,
"step": 4270
},
{
"grad_norm": 0.20826786756515503,
"learning_rate": 6.578070857242823e-05,
"loss": 0.034,
"step": 4280
},
{
"grad_norm": 0.34785202145576477,
"learning_rate": 6.562372686853402e-05,
"loss": 0.0389,
"step": 4290
},
{
"grad_norm": 0.20657755434513092,
"learning_rate": 6.546657430606593e-05,
"loss": 0.0368,
"step": 4300
},
{
"grad_norm": 0.3082147538661957,
"learning_rate": 6.530925260361918e-05,
"loss": 0.0334,
"step": 4310
},
{
"grad_norm": 0.2364552617073059,
"learning_rate": 6.515176348163871e-05,
"loss": 0.0353,
"step": 4320
},
{
"grad_norm": 0.37811702489852905,
"learning_rate": 6.499410866240032e-05,
"loss": 0.0349,
"step": 4330
},
{
"grad_norm": 0.2733839750289917,
"learning_rate": 6.48362898699919e-05,
"loss": 0.0387,
"step": 4340
},
{
"grad_norm": 0.36886081099510193,
"learning_rate": 6.467830883029443e-05,
"loss": 0.0357,
"step": 4350
},
{
"grad_norm": 0.29669854044914246,
"learning_rate": 6.452016727096326e-05,
"loss": 0.0383,
"step": 4360
},
{
"grad_norm": 0.24818091094493866,
"learning_rate": 6.436186692140916e-05,
"loss": 0.04,
"step": 4370
},
{
"grad_norm": 0.36215534806251526,
"learning_rate": 6.420340951277938e-05,
"loss": 0.0356,
"step": 4380
},
{
"grad_norm": 0.3138059377670288,
"learning_rate": 6.404479677793874e-05,
"loss": 0.0413,
"step": 4390
},
{
"grad_norm": 0.312511146068573,
"learning_rate": 6.388603045145075e-05,
"loss": 0.04,
"step": 4400
},
{
"grad_norm": 0.3475371301174164,
"learning_rate": 6.372711226955843e-05,
"loss": 0.0367,
"step": 4410
},
{
"grad_norm": 0.29854604601860046,
"learning_rate": 6.356804397016564e-05,
"loss": 0.0407,
"step": 4420
},
{
"grad_norm": 0.29716578125953674,
"learning_rate": 6.340882729281779e-05,
"loss": 0.0358,
"step": 4430
},
{
"grad_norm": 0.2285289615392685,
"learning_rate": 6.324946397868294e-05,
"loss": 0.0411,
"step": 4440
},
{
"grad_norm": 0.2572365999221802,
"learning_rate": 6.308995577053276e-05,
"loss": 0.033,
"step": 4450
},
{
"grad_norm": 0.2779274582862854,
"learning_rate": 6.293030441272347e-05,
"loss": 0.0387,
"step": 4460
},
{
"grad_norm": 0.2334423065185547,
"learning_rate": 6.277051165117677e-05,
"loss": 0.0364,
"step": 4470
},
{
"grad_norm": 0.21409209072589874,
"learning_rate": 6.261057923336064e-05,
"loss": 0.0343,
"step": 4480
},
{
"grad_norm": 0.24509204924106598,
"learning_rate": 6.245050890827042e-05,
"loss": 0.041,
"step": 4490
},
{
"grad_norm": 0.20741361379623413,
"learning_rate": 6.229030242640952e-05,
"loss": 0.0396,
"step": 4500
},
{
"grad_norm": 0.25321847200393677,
"learning_rate": 6.212996153977037e-05,
"loss": 0.0404,
"step": 4510
},
{
"grad_norm": 0.21805378794670105,
"learning_rate": 6.196948800181523e-05,
"loss": 0.0337,
"step": 4520
},
{
"grad_norm": 0.24263311922550201,
"learning_rate": 6.180888356745695e-05,
"loss": 0.0429,
"step": 4530
},
{
"grad_norm": 0.3522562086582184,
"learning_rate": 6.164814999303995e-05,
"loss": 0.0339,
"step": 4540
},
{
"grad_norm": 0.20797550678253174,
"learning_rate": 6.148728903632081e-05,
"loss": 0.0328,
"step": 4550
},
{
"grad_norm": 0.3335641920566559,
"learning_rate": 6.132630245644921e-05,
"loss": 0.0357,
"step": 4560
},
{
"grad_norm": 0.3732931315898895,
"learning_rate": 6.116519201394857e-05,
"loss": 0.0417,
"step": 4570
},
{
"grad_norm": 0.2926104962825775,
"learning_rate": 6.10039594706969e-05,
"loss": 0.0366,
"step": 4580
},
{
"grad_norm": 0.26506686210632324,
"learning_rate": 6.084260658990744e-05,
"loss": 0.0345,
"step": 4590
},
{
"grad_norm": 0.3617144823074341,
"learning_rate": 6.068113513610943e-05,
"loss": 0.0362,
"step": 4600
},
{
"grad_norm": 0.18685516715049744,
"learning_rate": 6.0519546875128876e-05,
"loss": 0.0406,
"step": 4610
},
{
"grad_norm": 0.2997559905052185,
"learning_rate": 6.035784357406906e-05,
"loss": 0.0391,
"step": 4620
},
{
"grad_norm": 0.23543107509613037,
"learning_rate": 6.01960270012914e-05,
"loss": 0.0379,
"step": 4630
},
{
"grad_norm": 0.3187630772590637,
"learning_rate": 6.003409892639599e-05,
"loss": 0.0376,
"step": 4640
},
{
"grad_norm": 0.32667702436447144,
"learning_rate": 5.9872061120202336e-05,
"loss": 0.036,
"step": 4650
},
{
"grad_norm": 0.31323152780532837,
"learning_rate": 5.9709915354729914e-05,
"loss": 0.0471,
"step": 4660
},
{
"grad_norm": 0.224016010761261,
"learning_rate": 5.9547663403178824e-05,
"loss": 0.0394,
"step": 4670
},
{
"grad_norm": 0.2602682113647461,
"learning_rate": 5.9385307039910445e-05,
"loss": 0.0342,
"step": 4680
},
{
"grad_norm": 0.22987647354602814,
"learning_rate": 5.922284804042792e-05,
"loss": 0.0408,
"step": 4690
},
{
"grad_norm": 0.3253161907196045,
"learning_rate": 5.906028818135687e-05,
"loss": 0.0424,
"step": 4700
},
{
"grad_norm": 0.1963428258895874,
"learning_rate": 5.889762924042585e-05,
"loss": 0.0355,
"step": 4710
},
{
"grad_norm": 0.27450031042099,
"learning_rate": 5.873487299644699e-05,
"loss": 0.0362,
"step": 4720
},
{
"grad_norm": 0.18444731831550598,
"learning_rate": 5.857202122929649e-05,
"loss": 0.037,
"step": 4730
},
{
"grad_norm": 0.24543538689613342,
"learning_rate": 5.840907571989518e-05,
"loss": 0.035,
"step": 4740
},
{
"grad_norm": 0.30772364139556885,
"learning_rate": 5.824603825018904e-05,
"loss": 0.0377,
"step": 4750
},
{
"grad_norm": 0.22756722569465637,
"learning_rate": 5.808291060312975e-05,
"loss": 0.0413,
"step": 4760
},
{
"grad_norm": 0.183345764875412,
"learning_rate": 5.7919694562655083e-05,
"loss": 0.041,
"step": 4770
},
{
"grad_norm": 0.26453009247779846,
"learning_rate": 5.775639191366954e-05,
"loss": 0.0389,
"step": 4780
},
{
"grad_norm": 0.3596522808074951,
"learning_rate": 5.75930044420247e-05,
"loss": 0.0374,
"step": 4790
},
{
"grad_norm": 0.24653761088848114,
"learning_rate": 5.74295339344998e-05,
"loss": 0.0412,
"step": 4800
},
{
"grad_norm": 0.21216793358325958,
"learning_rate": 5.726598217878211e-05,
"loss": 0.0384,
"step": 4810
},
{
"grad_norm": 0.24145081639289856,
"learning_rate": 5.71023509634474e-05,
"loss": 0.0346,
"step": 4820
},
{
"grad_norm": 0.31848645210266113,
"learning_rate": 5.693864207794049e-05,
"loss": 0.0409,
"step": 4830
},
{
"grad_norm": 0.2501415014266968,
"learning_rate": 5.677485731255545e-05,
"loss": 0.0479,
"step": 4840
},
{
"grad_norm": 0.1878986954689026,
"learning_rate": 5.6610998458416296e-05,
"loss": 0.041,
"step": 4850
},
{
"grad_norm": 0.23605766892433167,
"learning_rate": 5.644706730745716e-05,
"loss": 0.0391,
"step": 4860
},
{
"grad_norm": 0.3446219861507416,
"learning_rate": 5.628306565240287e-05,
"loss": 0.0404,
"step": 4870
},
{
"grad_norm": 0.2836945354938507,
"learning_rate": 5.611899528674923e-05,
"loss": 0.0433,
"step": 4880
},
{
"grad_norm": 0.31638237833976746,
"learning_rate": 5.595485800474349e-05,
"loss": 0.0407,
"step": 4890
},
{
"grad_norm": 0.22898872196674347,
"learning_rate": 5.579065560136467e-05,
"loss": 0.038,
"step": 4900
},
{
"grad_norm": 0.26057061553001404,
"learning_rate": 5.562638987230392e-05,
"loss": 0.0353,
"step": 4910
},
{
"grad_norm": 0.22998417913913727,
"learning_rate": 5.546206261394498e-05,
"loss": 0.0323,
"step": 4920
},
{
"grad_norm": 0.2660549581050873,
"learning_rate": 5.529767562334437e-05,
"loss": 0.0379,
"step": 4930
},
{
"grad_norm": 0.25155380368232727,
"learning_rate": 5.5133230698211926e-05,
"loss": 0.0328,
"step": 4940
},
{
"grad_norm": 0.21588538587093353,
"learning_rate": 5.496872963689096e-05,
"loss": 0.0329,
"step": 4950
},
{
"grad_norm": 0.23521240055561066,
"learning_rate": 5.4804174238338756e-05,
"loss": 0.033,
"step": 4960
},
{
"grad_norm": 0.22621150314807892,
"learning_rate": 5.463956630210678e-05,
"loss": 0.0359,
"step": 4970
},
{
"grad_norm": 0.231257826089859,
"learning_rate": 5.4474907628321046e-05,
"loss": 0.0404,
"step": 4980
},
{
"grad_norm": 0.31288015842437744,
"learning_rate": 5.431020001766244e-05,
"loss": 0.0382,
"step": 4990
},
{
"grad_norm": 0.22130578756332397,
"learning_rate": 5.4145445271346986e-05,
"loss": 0.0414,
"step": 5000
},
{
"grad_norm": 0.18791089951992035,
"learning_rate": 5.398064519110622e-05,
"loss": 0.0396,
"step": 5010
},
{
"grad_norm": 0.3742504417896271,
"learning_rate": 5.3815801579167394e-05,
"loss": 0.038,
"step": 5020
},
{
"grad_norm": 0.2941673696041107,
"learning_rate": 5.365091623823382e-05,
"loss": 0.0435,
"step": 5030
},
{
"grad_norm": 0.37370261549949646,
"learning_rate": 5.348599097146521e-05,
"loss": 0.0375,
"step": 5040
},
{
"grad_norm": 0.25983932614326477,
"learning_rate": 5.3321027582457836e-05,
"loss": 0.0406,
"step": 5050
},
{
"grad_norm": 0.24436895549297333,
"learning_rate": 5.315602787522491e-05,
"loss": 0.0396,
"step": 5060
},
{
"grad_norm": 0.3256576657295227,
"learning_rate": 5.299099365417678e-05,
"loss": 0.0367,
"step": 5070
},
{
"grad_norm": 0.25155988335609436,
"learning_rate": 5.2825926724101236e-05,
"loss": 0.0375,
"step": 5080
},
{
"grad_norm": 0.2639361321926117,
"learning_rate": 5.26608288901438e-05,
"loss": 0.0372,
"step": 5090
},
{
"grad_norm": 0.35629966855049133,
"learning_rate": 5.24957019577879e-05,
"loss": 0.0385,
"step": 5100
},
{
"grad_norm": 0.2046343833208084,
"learning_rate": 5.2330547732835266e-05,
"loss": 0.0353,
"step": 5110
},
{
"grad_norm": 0.304828941822052,
"learning_rate": 5.2165368021385996e-05,
"loss": 0.0396,
"step": 5120
},
{
"grad_norm": 0.25816217064857483,
"learning_rate": 5.200016462981897e-05,
"loss": 0.0324,
"step": 5130
},
{
"grad_norm": 0.233418270945549,
"learning_rate": 5.1834939364772015e-05,
"loss": 0.0368,
"step": 5140
},
{
"grad_norm": 0.22067779302597046,
"learning_rate": 5.166969403312214e-05,
"loss": 0.0349,
"step": 5150
},
{
"grad_norm": 0.20625650882720947,
"learning_rate": 5.1504430441965844e-05,
"loss": 0.0387,
"step": 5160
},
{
"grad_norm": 0.25634822249412537,
"learning_rate": 5.133915039859923e-05,
"loss": 0.0303,
"step": 5170
},
{
"grad_norm": 0.26570481061935425,
"learning_rate": 5.1173855710498444e-05,
"loss": 0.0369,
"step": 5180
},
{
"grad_norm": 0.3560854494571686,
"learning_rate": 5.100854818529967e-05,
"loss": 0.0413,
"step": 5190
},
{
"grad_norm": 0.34352895617485046,
"learning_rate": 5.084322963077951e-05,
"loss": 0.0428,
"step": 5200
},
{
"grad_norm": 0.1952911913394928,
"learning_rate": 5.067790185483522e-05,
"loss": 0.0411,
"step": 5210
},
{
"grad_norm": 0.27682581543922424,
"learning_rate": 5.0512566665464844e-05,
"loss": 0.0411,
"step": 5220
},
{
"grad_norm": 0.21189375221729279,
"learning_rate": 5.034722587074755e-05,
"loss": 0.0368,
"step": 5230
},
{
"grad_norm": 0.22940503060817719,
"learning_rate": 5.018188127882375e-05,
"loss": 0.0386,
"step": 5240
},
{
"grad_norm": 0.255561500787735,
"learning_rate": 5.0016534697875417e-05,
"loss": 0.033,
"step": 5250
},
{
"grad_norm": 0.20533734560012817,
"learning_rate": 4.9851187936106294e-05,
"loss": 0.0376,
"step": 5260
},
{
"grad_norm": 0.1934116631746292,
"learning_rate": 4.968584280172206e-05,
"loss": 0.0376,
"step": 5270
},
{
"grad_norm": 0.342820405960083,
"learning_rate": 4.95205011029106e-05,
"loss": 0.0355,
"step": 5280
},
{
"grad_norm": 0.2753123939037323,
"learning_rate": 4.935516464782227e-05,
"loss": 0.0387,
"step": 5290
},
{
"grad_norm": 0.3125678300857544,
"learning_rate": 4.918983524455003e-05,
"loss": 0.0356,
"step": 5300
},
{
"grad_norm": 0.23502302169799805,
"learning_rate": 4.9024514701109766e-05,
"loss": 0.0386,
"step": 5310
},
{
"grad_norm": 0.3033488094806671,
"learning_rate": 4.885920482542043e-05,
"loss": 0.0371,
"step": 5320
},
{
"grad_norm": 0.3101099133491516,
"learning_rate": 4.869390742528438e-05,
"loss": 0.0334,
"step": 5330
},
{
"grad_norm": 0.2722679078578949,
"learning_rate": 4.852862430836744e-05,
"loss": 0.0324,
"step": 5340
},
{
"grad_norm": 0.21368750929832458,
"learning_rate": 4.836335728217933e-05,
"loss": 0.0337,
"step": 5350
},
{
"grad_norm": 0.22733661532402039,
"learning_rate": 4.819810815405379e-05,
"loss": 0.0365,
"step": 5360
},
{
"grad_norm": 0.22253678739070892,
"learning_rate": 4.803287873112877e-05,
"loss": 0.0391,
"step": 5370
},
{
"grad_norm": 0.17248223721981049,
"learning_rate": 4.786767082032681e-05,
"loss": 0.033,
"step": 5380
},
{
"grad_norm": 0.1906975358724594,
"learning_rate": 4.77024862283351e-05,
"loss": 0.0343,
"step": 5390
},
{
"grad_norm": 0.21055501699447632,
"learning_rate": 4.753732676158593e-05,
"loss": 0.0339,
"step": 5400
},
{
"grad_norm": 0.2365766316652298,
"learning_rate": 4.737219422623672e-05,
"loss": 0.0334,
"step": 5410
},
{
"grad_norm": 0.29083865880966187,
"learning_rate": 4.720709042815044e-05,
"loss": 0.0324,
"step": 5420
},
{
"grad_norm": 0.3975061774253845,
"learning_rate": 4.704201717287578e-05,
"loss": 0.0373,
"step": 5430
},
{
"grad_norm": 0.24299997091293335,
"learning_rate": 4.6876976265627404e-05,
"loss": 0.0365,
"step": 5440
},
{
"grad_norm": 0.2981041371822357,
"learning_rate": 4.671196951126626e-05,
"loss": 0.0377,
"step": 5450
},
{
"grad_norm": 0.20610220730304718,
"learning_rate": 4.654699871427971e-05,
"loss": 0.0314,
"step": 5460
},
{
"grad_norm": 0.22911398112773895,
"learning_rate": 4.6382065678762034e-05,
"loss": 0.0302,
"step": 5470
},
{
"grad_norm": 0.2057241052389145,
"learning_rate": 4.6217172208394424e-05,
"loss": 0.0344,
"step": 5480
},
{
"grad_norm": 0.19260850548744202,
"learning_rate": 4.605232010642549e-05,
"loss": 0.0335,
"step": 5490
},
{
"grad_norm": 0.1900186985731125,
"learning_rate": 4.588751117565142e-05,
"loss": 0.0266,
"step": 5500
},
{
"grad_norm": 0.19529210031032562,
"learning_rate": 4.5722747218396214e-05,
"loss": 0.035,
"step": 5510
},
{
"grad_norm": 0.14910563826560974,
"learning_rate": 4.5558030036492194e-05,
"loss": 0.0302,
"step": 5520
},
{
"grad_norm": 0.23223315179347992,
"learning_rate": 4.539336143125999e-05,
"loss": 0.0404,
"step": 5530
},
{
"grad_norm": 0.276851087808609,
"learning_rate": 4.522874320348916e-05,
"loss": 0.0329,
"step": 5540
},
{
"grad_norm": 0.2226450890302658,
"learning_rate": 4.506417715341821e-05,
"loss": 0.0331,
"step": 5550
},
{
"grad_norm": 0.2843874394893646,
"learning_rate": 4.489966508071511e-05,
"loss": 0.0377,
"step": 5560
},
{
"grad_norm": 0.2196144312620163,
"learning_rate": 4.4735208784457575e-05,
"loss": 0.0324,
"step": 5570
},
{
"grad_norm": 0.23702110350131989,
"learning_rate": 4.457081006311325e-05,
"loss": 0.038,
"step": 5580
},
{
"grad_norm": 0.24752630293369293,
"learning_rate": 4.440647071452027e-05,
"loss": 0.0378,
"step": 5590
},
{
"grad_norm": 0.30340322852134705,
"learning_rate": 4.424219253586737e-05,
"loss": 0.0366,
"step": 5600
},
{
"grad_norm": 0.17964060604572296,
"learning_rate": 4.407797732367443e-05,
"loss": 0.0406,
"step": 5610
},
{
"grad_norm": 0.24079760909080505,
"learning_rate": 4.391382687377268e-05,
"loss": 0.0363,
"step": 5620
},
{
"grad_norm": 0.19189085066318512,
"learning_rate": 4.374974298128512e-05,
"loss": 0.0322,
"step": 5630
},
{
"grad_norm": 0.32452279329299927,
"learning_rate": 4.358572744060699e-05,
"loss": 0.0366,
"step": 5640
},
{
"grad_norm": 0.2274034023284912,
"learning_rate": 4.342178204538588e-05,
"loss": 0.0297,
"step": 5650
},
{
"grad_norm": 0.24880623817443848,
"learning_rate": 4.325790858850241e-05,
"loss": 0.0336,
"step": 5660
},
{
"grad_norm": 0.29345250129699707,
"learning_rate": 4.309410886205043e-05,
"loss": 0.0348,
"step": 5670
},
{
"grad_norm": 0.24763484299182892,
"learning_rate": 4.293038465731752e-05,
"loss": 0.0356,
"step": 5680
},
{
"grad_norm": 0.28137367963790894,
"learning_rate": 4.276673776476533e-05,
"loss": 0.0373,
"step": 5690
},
{
"grad_norm": 0.26791107654571533,
"learning_rate": 4.260316997401007e-05,
"loss": 0.0316,
"step": 5700
},
{
"grad_norm": 0.2903235852718353,
"learning_rate": 4.243968307380293e-05,
"loss": 0.0358,
"step": 5710
},
{
"grad_norm": 0.2530398964881897,
"learning_rate": 4.22762788520104e-05,
"loss": 0.0386,
"step": 5720
},
{
"grad_norm": 0.19159391522407532,
"learning_rate": 4.211295909559491e-05,
"loss": 0.0371,
"step": 5730
},
{
"grad_norm": 0.2652336359024048,
"learning_rate": 4.194972559059511e-05,
"loss": 0.0327,
"step": 5740
},
{
"grad_norm": 0.1957378387451172,
"learning_rate": 4.178658012210651e-05,
"loss": 0.0312,
"step": 5750
},
{
"grad_norm": 0.2904433012008667,
"learning_rate": 4.162352447426177e-05,
"loss": 0.0316,
"step": 5760
},
{
"grad_norm": 0.31744495034217834,
"learning_rate": 4.146056043021135e-05,
"loss": 0.0344,
"step": 5770
},
{
"grad_norm": 0.3251538872718811,
"learning_rate": 4.1297689772103944e-05,
"loss": 0.0372,
"step": 5780
},
{
"grad_norm": 0.30282530188560486,
"learning_rate": 4.113491428106694e-05,
"loss": 0.0324,
"step": 5790
},
{
"grad_norm": 0.28836318850517273,
"learning_rate": 4.0972235737187055e-05,
"loss": 0.0324,
"step": 5800
},
{
"grad_norm": 0.2510840892791748,
"learning_rate": 4.080965591949076e-05,
"loss": 0.0332,
"step": 5810
},
{
"grad_norm": 0.18844673037528992,
"learning_rate": 4.0647176605924924e-05,
"loss": 0.0289,
"step": 5820
},
{
"grad_norm": 0.24641014635562897,
"learning_rate": 4.0484799573337255e-05,
"loss": 0.033,
"step": 5830
},
{
"grad_norm": 0.22412146627902985,
"learning_rate": 4.032252659745699e-05,
"loss": 0.0327,
"step": 5840
},
{
"grad_norm": 0.27203699946403503,
"learning_rate": 4.016035945287539e-05,
"loss": 0.0404,
"step": 5850
},
{
"grad_norm": 0.19050034880638123,
"learning_rate": 3.999829991302635e-05,
"loss": 0.0311,
"step": 5860
},
{
"grad_norm": 0.2875731587409973,
"learning_rate": 3.983634975016707e-05,
"loss": 0.0316,
"step": 5870
},
{
"grad_norm": 0.19109110534191132,
"learning_rate": 3.967451073535854e-05,
"loss": 0.0427,
"step": 5880
},
{
"grad_norm": 0.25180962681770325,
"learning_rate": 3.951278463844633e-05,
"loss": 0.0347,
"step": 5890
},
{
"grad_norm": 0.22274623811244965,
"learning_rate": 3.935117322804111e-05,
"loss": 0.0337,
"step": 5900
},
{
"grad_norm": 0.3141174018383026,
"learning_rate": 3.918967827149938e-05,
"loss": 0.0382,
"step": 5910
},
{
"grad_norm": 0.1939334273338318,
"learning_rate": 3.9028301534904094e-05,
"loss": 0.0334,
"step": 5920
},
{
"grad_norm": 0.18623189628124237,
"learning_rate": 3.88670447830454e-05,
"loss": 0.0316,
"step": 5930
},
{
"grad_norm": 0.18647274374961853,
"learning_rate": 3.870590977940132e-05,
"loss": 0.0311,
"step": 5940
},
{
"grad_norm": 0.2672222852706909,
"learning_rate": 3.8544898286118404e-05,
"loss": 0.0303,
"step": 5950
},
{
"grad_norm": 0.2662316560745239,
"learning_rate": 3.838401206399257e-05,
"loss": 0.0304,
"step": 5960
},
{
"grad_norm": 0.2288493663072586,
"learning_rate": 3.822325287244975e-05,
"loss": 0.0305,
"step": 5970
},
{
"grad_norm": 0.26445239782333374,
"learning_rate": 3.8062622469526725e-05,
"loss": 0.0324,
"step": 5980
},
{
"grad_norm": 0.2491525113582611,
"learning_rate": 3.790212261185183e-05,
"loss": 0.032,
"step": 5990
},
{
"grad_norm": 0.231404110789299,
"learning_rate": 3.7741755054625794e-05,
"loss": 0.0327,
"step": 6000
},
{
"grad_norm": 0.22487586736679077,
"learning_rate": 3.758152155160255e-05,
"loss": 0.0294,
"step": 6010
},
{
"grad_norm": 0.20827744901180267,
"learning_rate": 3.742142385506999e-05,
"loss": 0.033,
"step": 6020
},
{
"grad_norm": 0.21554578840732574,
"learning_rate": 3.72614637158309e-05,
"loss": 0.0314,
"step": 6030
},
{
"grad_norm": 0.22519326210021973,
"learning_rate": 3.710164288318371e-05,
"loss": 0.0332,
"step": 6040
},
{
"grad_norm": 0.1589578539133072,
"learning_rate": 3.694196310490345e-05,
"loss": 0.0288,
"step": 6050
},
{
"grad_norm": 0.23213742673397064,
"learning_rate": 3.678242612722259e-05,
"loss": 0.0316,
"step": 6060
},
{
"grad_norm": 0.20701618492603302,
"learning_rate": 3.6623033694811953e-05,
"loss": 0.031,
"step": 6070
},
{
"grad_norm": 0.24319326877593994,
"learning_rate": 3.6463787550761665e-05,
"loss": 0.0354,
"step": 6080
},
{
"grad_norm": 0.22210799157619476,
"learning_rate": 3.630468943656202e-05,
"loss": 0.0334,
"step": 6090
},
{
"grad_norm": 0.2544934153556824,
"learning_rate": 3.6145741092084523e-05,
"loss": 0.0298,
"step": 6100
},
{
"grad_norm": 0.23841388523578644,
"learning_rate": 3.598694425556278e-05,
"loss": 0.0322,
"step": 6110
},
{
"grad_norm": 0.19115740060806274,
"learning_rate": 3.58283006635736e-05,
"loss": 0.0346,
"step": 6120
},
{
"grad_norm": 0.22685843706130981,
"learning_rate": 3.566981205101781e-05,
"loss": 0.0356,
"step": 6130
},
{
"grad_norm": 0.27599021792411804,
"learning_rate": 3.5511480151101556e-05,
"loss": 0.0326,
"step": 6140
},
{
"grad_norm": 0.15813547372817993,
"learning_rate": 3.5353306695317104e-05,
"loss": 0.0284,
"step": 6150
},
{
"grad_norm": 0.24084292352199554,
"learning_rate": 3.519529341342402e-05,
"loss": 0.0328,
"step": 6160
},
{
"grad_norm": 0.24921941757202148,
"learning_rate": 3.503744203343026e-05,
"loss": 0.0312,
"step": 6170
},
{
"grad_norm": 0.23112253844738007,
"learning_rate": 3.487975428157318e-05,
"loss": 0.0373,
"step": 6180
},
{
"grad_norm": 0.23314426839351654,
"learning_rate": 3.472223188230083e-05,
"loss": 0.0363,
"step": 6190
},
{
"grad_norm": 0.18143337965011597,
"learning_rate": 3.4564876558252866e-05,
"loss": 0.0311,
"step": 6200
},
{
"grad_norm": 0.25183209776878357,
"learning_rate": 3.440769003024195e-05,
"loss": 0.0315,
"step": 6210
},
{
"grad_norm": 0.22333580255508423,
"learning_rate": 3.425067401723477e-05,
"loss": 0.0287,
"step": 6220
},
{
"grad_norm": 0.26738110184669495,
"learning_rate": 3.409383023633325e-05,
"loss": 0.0332,
"step": 6230
},
{
"grad_norm": 0.24896834790706635,
"learning_rate": 3.3937160402755894e-05,
"loss": 0.032,
"step": 6240
},
{
"grad_norm": 0.2380266934633255,
"learning_rate": 3.378066622981885e-05,
"loss": 0.0315,
"step": 6250
},
{
"grad_norm": 0.29233479499816895,
"learning_rate": 3.362434942891738e-05,
"loss": 0.0368,
"step": 6260
},
{
"grad_norm": 0.166949063539505,
"learning_rate": 3.346821170950693e-05,
"loss": 0.0297,
"step": 6270
},
{
"grad_norm": 0.15828630328178406,
"learning_rate": 3.3312254779084585e-05,
"loss": 0.0316,
"step": 6280
},
{
"grad_norm": 0.21644681692123413,
"learning_rate": 3.315648034317039e-05,
"loss": 0.0328,
"step": 6290
},
{
"grad_norm": 0.19851650297641754,
"learning_rate": 3.3000890105288564e-05,
"loss": 0.0298,
"step": 6300
},
{
"grad_norm": 0.20566806197166443,
"learning_rate": 3.284548576694908e-05,
"loss": 0.0314,
"step": 6310
},
{
"grad_norm": 0.20094391703605652,
"learning_rate": 3.2690269027628815e-05,
"loss": 0.0303,
"step": 6320
},
{
"grad_norm": 0.20310550928115845,
"learning_rate": 3.253524158475324e-05,
"loss": 0.0302,
"step": 6330
},
{
"grad_norm": 0.13095231354236603,
"learning_rate": 3.238040513367757e-05,
"loss": 0.0305,
"step": 6340
},
{
"grad_norm": 0.24317970871925354,
"learning_rate": 3.222576136766843e-05,
"loss": 0.0286,
"step": 6350
},
{
"grad_norm": 0.17637597024440765,
"learning_rate": 3.2071311977885324e-05,
"loss": 0.0323,
"step": 6360
},
{
"grad_norm": 0.1991087943315506,
"learning_rate": 3.191705865336197e-05,
"loss": 0.0333,
"step": 6370
},
{
"grad_norm": 0.17794857919216156,
"learning_rate": 3.1763003080988075e-05,
"loss": 0.0307,
"step": 6380
},
{
"grad_norm": 0.21128028631210327,
"learning_rate": 3.160914694549063e-05,
"loss": 0.0296,
"step": 6390
},
{
"grad_norm": 0.18709604442119598,
"learning_rate": 3.145549192941573e-05,
"loss": 0.0315,
"step": 6400
},
{
"grad_norm": 0.2321898490190506,
"learning_rate": 3.130203971310999e-05,
"loss": 0.0338,
"step": 6410
},
{
"grad_norm": 0.18489141762256622,
"learning_rate": 3.114879197470225e-05,
"loss": 0.0321,
"step": 6420
},
{
"grad_norm": 0.18594172596931458,
"learning_rate": 3.0995750390085285e-05,
"loss": 0.0269,
"step": 6430
},
{
"grad_norm": 0.2447895109653473,
"learning_rate": 3.084291663289728e-05,
"loss": 0.0315,
"step": 6440
},
{
"grad_norm": 0.2385556548833847,
"learning_rate": 3.069029237450375e-05,
"loss": 0.0276,
"step": 6450
},
{
"grad_norm": 0.17444922029972076,
"learning_rate": 3.053787928397911e-05,
"loss": 0.0281,
"step": 6460
},
{
"grad_norm": 0.16294556856155396,
"learning_rate": 3.0385679028088526e-05,
"loss": 0.0328,
"step": 6470
},
{
"grad_norm": 0.20392334461212158,
"learning_rate": 3.023369327126959e-05,
"loss": 0.0288,
"step": 6480
},
{
"grad_norm": 0.18127469718456268,
"learning_rate": 3.0081923675614198e-05,
"loss": 0.03,
"step": 6490
},
{
"grad_norm": 0.20642398297786713,
"learning_rate": 2.993037190085034e-05,
"loss": 0.0287,
"step": 6500
},
{
"grad_norm": 0.19302089512348175,
"learning_rate": 2.977903960432392e-05,
"loss": 0.0345,
"step": 6510
},
{
"grad_norm": 0.2642991840839386,
"learning_rate": 2.9627928440980722e-05,
"loss": 0.031,
"step": 6520
},
{
"grad_norm": 0.21077126264572144,
"learning_rate": 2.9477040063348183e-05,
"loss": 0.0313,
"step": 6530
},
{
"grad_norm": 0.147378608584404,
"learning_rate": 2.9326376121517456e-05,
"loss": 0.0274,
"step": 6540
},
{
"grad_norm": 0.22836551070213318,
"learning_rate": 2.9175938263125236e-05,
"loss": 0.0302,
"step": 6550
},
{
"grad_norm": 0.2703534662723541,
"learning_rate": 2.9025728133335873e-05,
"loss": 0.0349,
"step": 6560
},
{
"grad_norm": 0.2877998352050781,
"learning_rate": 2.8875747374823288e-05,
"loss": 0.0333,
"step": 6570
},
{
"grad_norm": 0.2315242886543274,
"learning_rate": 2.872599762775298e-05,
"loss": 0.0338,
"step": 6580
},
{
"grad_norm": 0.18308284878730774,
"learning_rate": 2.857648052976425e-05,
"loss": 0.0298,
"step": 6590
},
{
"grad_norm": 0.2217554748058319,
"learning_rate": 2.8427197715952047e-05,
"loss": 0.0295,
"step": 6600
},
{
"grad_norm": 0.21651163697242737,
"learning_rate": 2.8278150818849393e-05,
"loss": 0.0283,
"step": 6610
},
{
"grad_norm": 0.17601701617240906,
"learning_rate": 2.812934146840922e-05,
"loss": 0.0284,
"step": 6620
},
{
"grad_norm": 0.2138030081987381,
"learning_rate": 2.7980771291986764e-05,
"loss": 0.0323,
"step": 6630
},
{
"grad_norm": 0.20729990303516388,
"learning_rate": 2.783244191432167e-05,
"loss": 0.03,
"step": 6640
},
{
"grad_norm": 0.3634009063243866,
"learning_rate": 2.768435495752022e-05,
"loss": 0.0292,
"step": 6650
},
{
"grad_norm": 0.2113521248102188,
"learning_rate": 2.753651204103771e-05,
"loss": 0.0285,
"step": 6660
},
{
"grad_norm": 0.19870157539844513,
"learning_rate": 2.7388914781660523e-05,
"loss": 0.0299,
"step": 6670
},
{
"grad_norm": 0.17803676426410675,
"learning_rate": 2.7241564793488693e-05,
"loss": 0.0324,
"step": 6680
},
{
"grad_norm": 0.27665045857429504,
"learning_rate": 2.7094463687918037e-05,
"loss": 0.0323,
"step": 6690
},
{
"grad_norm": 0.3654191792011261,
"learning_rate": 2.694761307362268e-05,
"loss": 0.0314,
"step": 6700
},
{
"grad_norm": 0.2820141911506653,
"learning_rate": 2.6801014556537467e-05,
"loss": 0.0305,
"step": 6710
},
{
"grad_norm": 0.21494096517562866,
"learning_rate": 2.6654669739840243e-05,
"loss": 0.0286,
"step": 6720
},
{
"grad_norm": 0.2526452839374542,
"learning_rate": 2.650858022393451e-05,
"loss": 0.0308,
"step": 6730
},
{
"grad_norm": 0.19835573434829712,
"learning_rate": 2.6362747606431747e-05,
"loss": 0.0359,
"step": 6740
},
{
"grad_norm": 0.3186538815498352,
"learning_rate": 2.6217173482134172e-05,
"loss": 0.0308,
"step": 6750
},
{
"grad_norm": 0.27317216992378235,
"learning_rate": 2.6071859443017044e-05,
"loss": 0.0287,
"step": 6760
},
{
"grad_norm": 0.20776128768920898,
"learning_rate": 2.5926807078211414e-05,
"loss": 0.0262,
"step": 6770
},
{
"grad_norm": 0.22316332161426544,
"learning_rate": 2.5782017973986728e-05,
"loss": 0.0262,
"step": 6780
},
{
"grad_norm": 0.20760706067085266,
"learning_rate": 2.5637493713733374e-05,
"loss": 0.0285,
"step": 6790
},
{
"grad_norm": 0.2864840626716614,
"learning_rate": 2.549323587794559e-05,
"loss": 0.0326,
"step": 6800
},
{
"grad_norm": 0.15268519520759583,
"learning_rate": 2.5349246044203895e-05,
"loss": 0.027,
"step": 6810
},
{
"grad_norm": 0.233912855386734,
"learning_rate": 2.520552578715808e-05,
"loss": 0.0283,
"step": 6820
},
{
"grad_norm": 0.21493370831012726,
"learning_rate": 2.506207667850981e-05,
"loss": 0.027,
"step": 6830
},
{
"grad_norm": 0.18121784925460815,
"learning_rate": 2.4918900286995555e-05,
"loss": 0.0272,
"step": 6840
},
{
"grad_norm": 0.21178673207759857,
"learning_rate": 2.4775998178369458e-05,
"loss": 0.0275,
"step": 6850
},
{
"grad_norm": 0.23063145577907562,
"learning_rate": 2.4633371915386017e-05,
"loss": 0.0333,
"step": 6860
},
{
"grad_norm": 0.2095630019903183,
"learning_rate": 2.4491023057783235e-05,
"loss": 0.0326,
"step": 6870
},
{
"grad_norm": 0.21951556205749512,
"learning_rate": 2.4348953162265375e-05,
"loss": 0.0353,
"step": 6880
},
{
"grad_norm": 0.16206666827201843,
"learning_rate": 2.420716378248607e-05,
"loss": 0.0288,
"step": 6890
},
{
"grad_norm": 0.17747406661510468,
"learning_rate": 2.4065656469031266e-05,
"loss": 0.0286,
"step": 6900
},
{
"grad_norm": 0.2104213535785675,
"learning_rate": 2.3924432769402268e-05,
"loss": 0.0281,
"step": 6910
},
{
"grad_norm": 0.24779659509658813,
"learning_rate": 2.3783494227998844e-05,
"loss": 0.034,
"step": 6920
},
{
"grad_norm": 0.1603226214647293,
"learning_rate": 2.3642842386102264e-05,
"loss": 0.0245,
"step": 6930
},
{
"grad_norm": 0.25845393538475037,
"learning_rate": 2.3502478781858567e-05,
"loss": 0.028,
"step": 6940
},
{
"grad_norm": 0.23870128393173218,
"learning_rate": 2.3362404950261628e-05,
"loss": 0.0325,
"step": 6950
},
{
"grad_norm": 0.21474549174308777,
"learning_rate": 2.3222622423136458e-05,
"loss": 0.0303,
"step": 6960
},
{
"grad_norm": 0.23277056217193604,
"learning_rate": 2.3083132729122332e-05,
"loss": 0.0257,
"step": 6970
},
{
"grad_norm": 0.24012663960456848,
"learning_rate": 2.294393739365621e-05,
"loss": 0.029,
"step": 6980
},
{
"grad_norm": 0.17798693478107452,
"learning_rate": 2.2805037938956e-05,
"loss": 0.0312,
"step": 6990
},
{
"grad_norm": 0.19275571405887604,
"learning_rate": 2.266643588400386e-05,
"loss": 0.0289,
"step": 7000
},
{
"grad_norm": 0.23350587487220764,
"learning_rate": 2.252813274452969e-05,
"loss": 0.0327,
"step": 7010
},
{
"grad_norm": 0.1445411741733551,
"learning_rate": 2.2390130032994427e-05,
"loss": 0.0253,
"step": 7020
},
{
"grad_norm": 0.1882321685552597,
"learning_rate": 2.2252429258573633e-05,
"loss": 0.0271,
"step": 7030
},
{
"grad_norm": 0.1621200144290924,
"learning_rate": 2.2115031927140904e-05,
"loss": 0.0298,
"step": 7040
},
{
"grad_norm": 0.3082856833934784,
"learning_rate": 2.1977939541251463e-05,
"loss": 0.0371,
"step": 7050
},
{
"grad_norm": 0.168369859457016,
"learning_rate": 2.1841153600125684e-05,
"loss": 0.0272,
"step": 7060
},
{
"grad_norm": 0.17478293180465698,
"learning_rate": 2.170467559963267e-05,
"loss": 0.0302,
"step": 7070
},
{
"grad_norm": 0.19040048122406006,
"learning_rate": 2.1568507032273982e-05,
"loss": 0.0299,
"step": 7080
},
{
"grad_norm": 0.15172909200191498,
"learning_rate": 2.1432649387167264e-05,
"loss": 0.0278,
"step": 7090
},
{
"grad_norm": 0.15691806375980377,
"learning_rate": 2.1297104150029973e-05,
"loss": 0.0272,
"step": 7100
},
{
"grad_norm": 0.1737809032201767,
"learning_rate": 2.116187280316307e-05,
"loss": 0.0307,
"step": 7110
},
{
"grad_norm": 0.2123764157295227,
"learning_rate": 2.1026956825434908e-05,
"loss": 0.0286,
"step": 7120
},
{
"grad_norm": 0.25922951102256775,
"learning_rate": 2.0892357692265017e-05,
"loss": 0.0276,
"step": 7130
},
{
"grad_norm": 0.1568475365638733,
"learning_rate": 2.0758076875607947e-05,
"loss": 0.03,
"step": 7140
},
{
"grad_norm": 0.22746144235134125,
"learning_rate": 2.0624115843937207e-05,
"loss": 0.0309,
"step": 7150
},
{
"grad_norm": 0.20488545298576355,
"learning_rate": 2.0490476062229157e-05,
"loss": 0.0314,
"step": 7160
},
{
"grad_norm": 0.23041017353534698,
"learning_rate": 2.035715899194704e-05,
"loss": 0.0326,
"step": 7170
},
{
"grad_norm": 0.255893349647522,
"learning_rate": 2.022416609102499e-05,
"loss": 0.0276,
"step": 7180
},
{
"grad_norm": 0.21499620378017426,
"learning_rate": 2.009149881385205e-05,
"loss": 0.0331,
"step": 7190
},
{
"grad_norm": 0.1767837405204773,
"learning_rate": 1.995915861125634e-05,
"loss": 0.0285,
"step": 7200
},
{
"grad_norm": 0.1782473623752594,
"learning_rate": 1.9827146930489065e-05,
"loss": 0.0256,
"step": 7210
},
{
"grad_norm": 0.1710137575864792,
"learning_rate": 1.9695465215208848e-05,
"loss": 0.0272,
"step": 7220
},
{
"grad_norm": 0.16107825934886932,
"learning_rate": 1.9564114905465813e-05,
"loss": 0.0266,
"step": 7230
},
{
"grad_norm": 0.19516757130622864,
"learning_rate": 1.9433097437685936e-05,
"loss": 0.0278,
"step": 7240
},
{
"grad_norm": 0.1714981347322464,
"learning_rate": 1.930241424465521e-05,
"loss": 0.0232,
"step": 7250
},
{
"grad_norm": 0.20657700300216675,
"learning_rate": 1.9172066755504115e-05,
"loss": 0.0292,
"step": 7260
},
{
"grad_norm": 0.23087969422340393,
"learning_rate": 1.9042056395691914e-05,
"loss": 0.0303,
"step": 7270
},
{
"grad_norm": 0.18433725833892822,
"learning_rate": 1.8912384586991066e-05,
"loss": 0.0266,
"step": 7280
},
{
"grad_norm": 0.20770221948623657,
"learning_rate": 1.8783052747471717e-05,
"loss": 0.027,
"step": 7290
},
{
"grad_norm": 0.18887269496917725,
"learning_rate": 1.865406229148611e-05,
"loss": 0.0325,
"step": 7300
},
{
"grad_norm": 0.18146908283233643,
"learning_rate": 1.8525414629653233e-05,
"loss": 0.033,
"step": 7310
},
{
"grad_norm": 0.14821194112300873,
"learning_rate": 1.8397111168843255e-05,
"loss": 0.03,
"step": 7320
},
{
"grad_norm": 0.29156258702278137,
"learning_rate": 1.8269153312162323e-05,
"loss": 0.0297,
"step": 7330
},
{
"grad_norm": 0.21394626796245575,
"learning_rate": 1.8141542458937054e-05,
"loss": 0.0283,
"step": 7340
},
{
"grad_norm": 0.184283047914505,
"learning_rate": 1.8014280004699268e-05,
"loss": 0.026,
"step": 7350
},
{
"grad_norm": 0.23677630722522736,
"learning_rate": 1.788736734117078e-05,
"loss": 0.0265,
"step": 7360
},
{
"grad_norm": 0.2302027940750122,
"learning_rate": 1.7760805856248152e-05,
"loss": 0.0278,
"step": 7370
},
{
"grad_norm": 0.1684088408946991,
"learning_rate": 1.7634596933987518e-05,
"loss": 0.0299,
"step": 7380
},
{
"grad_norm": 0.2044224590063095,
"learning_rate": 1.7508741954589404e-05,
"loss": 0.0291,
"step": 7390
},
{
"grad_norm": 0.22018086910247803,
"learning_rate": 1.7383242294383717e-05,
"loss": 0.0281,
"step": 7400
},
{
"grad_norm": 0.2093113362789154,
"learning_rate": 1.7258099325814632e-05,
"loss": 0.029,
"step": 7410
},
{
"grad_norm": 0.20827700197696686,
"learning_rate": 1.7133314417425594e-05,
"loss": 0.0321,
"step": 7420
},
{
"grad_norm": 0.22063322365283966,
"learning_rate": 1.7008888933844408e-05,
"loss": 0.0283,
"step": 7430
},
{
"grad_norm": 0.2052403688430786,
"learning_rate": 1.6884824235768172e-05,
"loss": 0.027,
"step": 7440
},
{
"grad_norm": 0.24670685827732086,
"learning_rate": 1.6761121679948592e-05,
"loss": 0.0267,
"step": 7450
},
{
"grad_norm": 0.1991979479789734,
"learning_rate": 1.663778261917695e-05,
"loss": 0.0294,
"step": 7460
},
{
"grad_norm": 0.3023832142353058,
"learning_rate": 1.651480840226952e-05,
"loss": 0.0277,
"step": 7470
},
{
"grad_norm": 0.23743179440498352,
"learning_rate": 1.639220037405258e-05,
"loss": 0.0305,
"step": 7480
},
{
"grad_norm": 0.19982585310935974,
"learning_rate": 1.6269959875347906e-05,
"loss": 0.0298,
"step": 7490
},
{
"grad_norm": 0.17956045269966125,
"learning_rate": 1.614808824295802e-05,
"loss": 0.0271,
"step": 7500
},
{
"grad_norm": 0.2056272178888321,
"learning_rate": 1.602658680965152e-05,
"loss": 0.0252,
"step": 7510
},
{
"grad_norm": 0.2476661205291748,
"learning_rate": 1.5905456904148686e-05,
"loss": 0.0275,
"step": 7520
},
{
"grad_norm": 0.23972822725772858,
"learning_rate": 1.57846998511067e-05,
"loss": 0.0295,
"step": 7530
},
{
"grad_norm": 0.1646782010793686,
"learning_rate": 1.566431697110538e-05,
"loss": 0.0275,
"step": 7540
},
{
"grad_norm": 0.15701362490653992,
"learning_rate": 1.554430958063259e-05,
"loss": 0.0262,
"step": 7550
},
{
"grad_norm": 0.2679762840270996,
"learning_rate": 1.5424678992069912e-05,
"loss": 0.0252,
"step": 7560
},
{
"grad_norm": 0.2671467661857605,
"learning_rate": 1.5305426513678362e-05,
"loss": 0.0298,
"step": 7570
},
{
"grad_norm": 0.24908775091171265,
"learning_rate": 1.518655344958388e-05,
"loss": 0.0303,
"step": 7580
},
{
"grad_norm": 0.26669302582740784,
"learning_rate": 1.5068061099763275e-05,
"loss": 0.0297,
"step": 7590
},
{
"grad_norm": 0.1559380143880844,
"learning_rate": 1.494995076002988e-05,
"loss": 0.0288,
"step": 7600
},
{
"grad_norm": 0.20336051285266876,
"learning_rate": 1.4832223722019456e-05,
"loss": 0.025,
"step": 7610
},
{
"grad_norm": 0.17910243570804596,
"learning_rate": 1.4714881273176035e-05,
"loss": 0.0274,
"step": 7620
},
{
"grad_norm": 0.2607031464576721,
"learning_rate": 1.4597924696737835e-05,
"loss": 0.0261,
"step": 7630
},
{
"grad_norm": 0.15345190465450287,
"learning_rate": 1.4481355271723252e-05,
"loss": 0.0267,
"step": 7640
},
{
"grad_norm": 0.20041489601135254,
"learning_rate": 1.4365174272916809e-05,
"loss": 0.0292,
"step": 7650
},
{
"grad_norm": 0.17829930782318115,
"learning_rate": 1.4249382970855319e-05,
"loss": 0.028,
"step": 7660
},
{
"grad_norm": 0.21642664074897766,
"learning_rate": 1.4133982631813903e-05,
"loss": 0.0283,
"step": 7670
},
{
"grad_norm": 0.14973731338977814,
"learning_rate": 1.4018974517792194e-05,
"loss": 0.0244,
"step": 7680
},
{
"grad_norm": 0.21429570019245148,
"learning_rate": 1.390435988650048e-05,
"loss": 0.0263,
"step": 7690
},
{
"grad_norm": 0.18472836911678314,
"learning_rate": 1.3790139991346006e-05,
"loss": 0.0258,
"step": 7700
},
{
"grad_norm": 0.23076964914798737,
"learning_rate": 1.367631608141926e-05,
"loss": 0.0264,
"step": 7710
},
{
"grad_norm": 0.2141284942626953,
"learning_rate": 1.3562889401480278e-05,
"loss": 0.0268,
"step": 7720
},
{
"grad_norm": 0.21197544038295746,
"learning_rate": 1.3449861191945074e-05,
"loss": 0.0306,
"step": 7730
},
{
"grad_norm": 0.22252681851387024,
"learning_rate": 1.3337232688872009e-05,
"loss": 0.0269,
"step": 7740
},
{
"grad_norm": 0.2053757607936859,
"learning_rate": 1.3225005123948364e-05,
"loss": 0.0289,
"step": 7750
},
{
"grad_norm": 0.19560173153877258,
"learning_rate": 1.311317972447681e-05,
"loss": 0.0258,
"step": 7760
},
{
"grad_norm": 0.15975621342658997,
"learning_rate": 1.3001757713361996e-05,
"loss": 0.0225,
"step": 7770
},
{
"grad_norm": 0.21871204674243927,
"learning_rate": 1.2890740309097204e-05,
"loss": 0.0304,
"step": 7780
},
{
"grad_norm": 0.1690676212310791,
"learning_rate": 1.2780128725750944e-05,
"loss": 0.0237,
"step": 7790
},
{
"grad_norm": 0.1615976095199585,
"learning_rate": 1.266992417295379e-05,
"loss": 0.0257,
"step": 7800
},
{
"grad_norm": 0.18102748692035675,
"learning_rate": 1.2560127855885073e-05,
"loss": 0.0286,
"step": 7810
},
{
"grad_norm": 0.17890459299087524,
"learning_rate": 1.2450740975259745e-05,
"loss": 0.0283,
"step": 7820
},
{
"grad_norm": 0.14003805816173553,
"learning_rate": 1.234176472731517e-05,
"loss": 0.0318,
"step": 7830
},
{
"grad_norm": 0.21664176881313324,
"learning_rate": 1.2233200303798158e-05,
"loss": 0.0268,
"step": 7840
},
{
"grad_norm": 0.1764824241399765,
"learning_rate": 1.2125048891951846e-05,
"loss": 0.0242,
"step": 7850
},
{
"grad_norm": 0.1710733324289322,
"learning_rate": 1.2017311674502745e-05,
"loss": 0.0236,
"step": 7860
},
{
"grad_norm": 0.1569499522447586,
"learning_rate": 1.1909989829647822e-05,
"loss": 0.0239,
"step": 7870
},
{
"grad_norm": 0.2005121260881424,
"learning_rate": 1.1803084531041553e-05,
"loss": 0.0259,
"step": 7880
},
{
"grad_norm": 0.19949199259281158,
"learning_rate": 1.1696596947783162e-05,
"loss": 0.0266,
"step": 7890
},
{
"grad_norm": 0.13878698647022247,
"learning_rate": 1.1590528244403803e-05,
"loss": 0.0235,
"step": 7900
},
{
"grad_norm": 0.19747662544250488,
"learning_rate": 1.148487958085382e-05,
"loss": 0.0326,
"step": 7910
},
{
"grad_norm": 0.20122241973876953,
"learning_rate": 1.1379652112490086e-05,
"loss": 0.0261,
"step": 7920
},
{
"grad_norm": 0.16324187815189362,
"learning_rate": 1.1274846990063315e-05,
"loss": 0.0271,
"step": 7930
},
{
"grad_norm": 0.17504747211933136,
"learning_rate": 1.117046535970554e-05,
"loss": 0.0293,
"step": 7940
},
{
"grad_norm": 0.20539623498916626,
"learning_rate": 1.106650836291755e-05,
"loss": 0.0355,
"step": 7950
},
{
"grad_norm": 0.19083498418331146,
"learning_rate": 1.0962977136556418e-05,
"loss": 0.0322,
"step": 7960
},
{
"grad_norm": 0.1924605667591095,
"learning_rate": 1.0859872812823024e-05,
"loss": 0.0268,
"step": 7970
},
{
"grad_norm": 0.2575099468231201,
"learning_rate": 1.0757196519249747e-05,
"loss": 0.0331,
"step": 7980
},
{
"grad_norm": 0.1368202567100525,
"learning_rate": 1.0654949378688077e-05,
"loss": 0.0294,
"step": 7990
},
{
"grad_norm": 0.19401443004608154,
"learning_rate": 1.0553132509296376e-05,
"loss": 0.0271,
"step": 8000
},
{
"grad_norm": 0.1929926872253418,
"learning_rate": 1.0451747024527613e-05,
"loss": 0.0297,
"step": 8010
},
{
"grad_norm": 0.16329045593738556,
"learning_rate": 1.0350794033117189e-05,
"loss": 0.027,
"step": 8020
},
{
"grad_norm": 0.21634945273399353,
"learning_rate": 1.0250274639070856e-05,
"loss": 0.0284,
"step": 8030
},
{
"grad_norm": 0.1562228947877884,
"learning_rate": 1.0150189941652599e-05,
"loss": 0.027,
"step": 8040
},
{
"grad_norm": 0.18230652809143066,
"learning_rate": 1.0050541035372635e-05,
"loss": 0.0269,
"step": 8050
},
{
"grad_norm": 0.19346484541893005,
"learning_rate": 9.951329009975458e-06,
"loss": 0.0295,
"step": 8060
},
{
"grad_norm": 0.20279110968112946,
"learning_rate": 9.852554950427845e-06,
"loss": 0.0237,
"step": 8070
},
{
"grad_norm": 0.14290191233158112,
"learning_rate": 9.754219936907105e-06,
"loss": 0.025,
"step": 8080
},
{
"grad_norm": 0.17854970693588257,
"learning_rate": 9.656325044789194e-06,
"loss": 0.0247,
"step": 8090
},
{
"grad_norm": 0.23673054575920105,
"learning_rate": 9.55887134463697e-06,
"loss": 0.0237,
"step": 8100
},
{
"grad_norm": 0.15271058678627014,
"learning_rate": 9.461859902188475e-06,
"loss": 0.025,
"step": 8110
},
{
"grad_norm": 0.12986095249652863,
"learning_rate": 9.365291778345303e-06,
"loss": 0.027,
"step": 8120
},
{
"grad_norm": 0.13045966625213623,
"learning_rate": 9.269168029160991e-06,
"loss": 0.0234,
"step": 8130
},
{
"grad_norm": 0.222456693649292,
"learning_rate": 9.173489705829447e-06,
"loss": 0.0272,
"step": 8140
},
{
"grad_norm": 0.17325180768966675,
"learning_rate": 9.078257854673516e-06,
"loss": 0.027,
"step": 8150
},
{
"grad_norm": 0.15894733369350433,
"learning_rate": 8.983473517133429e-06,
"loss": 0.0263,
"step": 8160
},
{
"grad_norm": 0.2092178463935852,
"learning_rate": 8.889137729755537e-06,
"loss": 0.0267,
"step": 8170
},
{
"grad_norm": 0.18789401650428772,
"learning_rate": 8.79525152418087e-06,
"loss": 0.0273,
"step": 8180
},
{
"grad_norm": 0.23926988244056702,
"learning_rate": 8.701815927133961e-06,
"loss": 0.0251,
"step": 8190
},
{
"grad_norm": 0.1841491460800171,
"learning_rate": 8.608831960411534e-06,
"loss": 0.0262,
"step": 8200
},
{
"grad_norm": 0.21668025851249695,
"learning_rate": 8.516300640871321e-06,
"loss": 0.0265,
"step": 8210
},
{
"grad_norm": 0.16964370012283325,
"learning_rate": 8.424222980421038e-06,
"loss": 0.026,
"step": 8220
},
{
"grad_norm": 0.19786661863327026,
"learning_rate": 8.332599986007184e-06,
"loss": 0.0258,
"step": 8230
},
{
"grad_norm": 0.10745269805192947,
"learning_rate": 8.241432659604203e-06,
"loss": 0.0253,
"step": 8240
},
{
"grad_norm": 0.20551584661006927,
"learning_rate": 8.150721998203331e-06,
"loss": 0.0255,
"step": 8250
},
{
"grad_norm": 0.1453953981399536,
"learning_rate": 8.06046899380184e-06,
"loss": 0.0232,
"step": 8260
},
{
"grad_norm": 0.31428804993629456,
"learning_rate": 7.970674633392133e-06,
"loss": 0.0281,
"step": 8270
},
{
"grad_norm": 0.18517866730690002,
"learning_rate": 7.881339898950924e-06,
"loss": 0.0284,
"step": 8280
},
{
"grad_norm": 0.19145368039608002,
"learning_rate": 7.792465767428597e-06,
"loss": 0.0248,
"step": 8290
},
{
"grad_norm": 0.1350008249282837,
"learning_rate": 7.704053210738376e-06,
"loss": 0.0203,
"step": 8300
},
{
"grad_norm": 0.2560305893421173,
"learning_rate": 7.6161031957458494e-06,
"loss": 0.0294,
"step": 8310
},
{
"grad_norm": 0.24567309021949768,
"learning_rate": 7.5286166842582605e-06,
"loss": 0.0288,
"step": 8320
},
{
"grad_norm": 0.21309012174606323,
"learning_rate": 7.4415946330140814e-06,
"loss": 0.0292,
"step": 8330
},
{
"grad_norm": 0.20877450704574585,
"learning_rate": 7.3550379936725644e-06,
"loss": 0.0256,
"step": 8340
},
{
"grad_norm": 0.20123223960399628,
"learning_rate": 7.2689477128032035e-06,
"loss": 0.0299,
"step": 8350
},
{
"grad_norm": 0.19062873721122742,
"learning_rate": 7.183324731875551e-06,
"loss": 0.0276,
"step": 8360
},
{
"grad_norm": 0.17903867363929749,
"learning_rate": 7.098169987248782e-06,
"loss": 0.0239,
"step": 8370
},
{
"grad_norm": 0.1556374728679657,
"learning_rate": 7.013484410161553e-06,
"loss": 0.0261,
"step": 8380
},
{
"grad_norm": 0.18347126245498657,
"learning_rate": 6.92926892672176e-06,
"loss": 0.0252,
"step": 8390
},
{
"grad_norm": 0.14952725172042847,
"learning_rate": 6.845524457896446e-06,
"loss": 0.0228,
"step": 8400
},
{
"grad_norm": 0.15896227955818176,
"learning_rate": 6.7622519195017165e-06,
"loss": 0.0249,
"step": 8410
},
{
"grad_norm": 0.13131366670131683,
"learning_rate": 6.679452222192684e-06,
"loss": 0.0221,
"step": 8420
},
{
"grad_norm": 0.15154506266117096,
"learning_rate": 6.597126271453579e-06,
"loss": 0.0238,
"step": 8430
},
{
"grad_norm": 0.17291532456874847,
"learning_rate": 6.51527496758782e-06,
"loss": 0.024,
"step": 8440
},
{
"grad_norm": 0.1727951020002365,
"learning_rate": 6.433899205708155e-06,
"loss": 0.0245,
"step": 8450
},
{
"grad_norm": 0.21756398677825928,
"learning_rate": 6.352999875726856e-06,
"loss": 0.0253,
"step": 8460
},
{
"grad_norm": 0.17791414260864258,
"learning_rate": 6.272577862346052e-06,
"loss": 0.0225,
"step": 8470
},
{
"grad_norm": 0.19491951167583466,
"learning_rate": 6.192634045047996e-06,
"loss": 0.0307,
"step": 8480
},
{
"grad_norm": 0.1367735117673874,
"learning_rate": 6.113169298085458e-06,
"loss": 0.0245,
"step": 8490
},
{
"grad_norm": 0.20764757692813873,
"learning_rate": 6.034184490472195e-06,
"loss": 0.0246,
"step": 8500
},
{
"grad_norm": 0.17366479337215424,
"learning_rate": 5.955680485973386e-06,
"loss": 0.0254,
"step": 8510
},
{
"grad_norm": 0.20417249202728271,
"learning_rate": 5.877658143096265e-06,
"loss": 0.0232,
"step": 8520
},
{
"grad_norm": 0.2845112681388855,
"learning_rate": 5.800118315080661e-06,
"loss": 0.0223,
"step": 8530
},
{
"grad_norm": 0.19180867075920105,
"learning_rate": 5.723061849889716e-06,
"loss": 0.024,
"step": 8540
},
{
"grad_norm": 0.17623895406723022,
"learning_rate": 5.646489590200604e-06,
"loss": 0.0279,
"step": 8550
},
{
"grad_norm": 0.17861409485340118,
"learning_rate": 5.570402373395256e-06,
"loss": 0.0229,
"step": 8560
},
{
"grad_norm": 0.20771996676921844,
"learning_rate": 5.494801031551305e-06,
"loss": 0.0292,
"step": 8570
},
{
"grad_norm": 0.15694555640220642,
"learning_rate": 5.41968639143291e-06,
"loss": 0.0223,
"step": 8580
},
{
"grad_norm": 0.17930053174495697,
"learning_rate": 5.345059274481751e-06,
"loss": 0.0248,
"step": 8590
},
{
"grad_norm": 0.22845885157585144,
"learning_rate": 5.270920496808002e-06,
"loss": 0.0298,
"step": 8600
},
{
"grad_norm": 0.19219282269477844,
"learning_rate": 5.1972708691814695e-06,
"loss": 0.0276,
"step": 8610
},
{
"grad_norm": 0.12660208344459534,
"learning_rate": 5.124111197022674e-06,
"loss": 0.0269,
"step": 8620
},
{
"grad_norm": 0.19566069543361664,
"learning_rate": 5.051442280394081e-06,
"loss": 0.0251,
"step": 8630
},
{
"grad_norm": 0.21735556423664093,
"learning_rate": 4.979264913991322e-06,
"loss": 0.024,
"step": 8640
},
{
"grad_norm": 0.2024160623550415,
"learning_rate": 4.907579887134489e-06,
"loss": 0.0263,
"step": 8650
},
{
"grad_norm": 0.18530689179897308,
"learning_rate": 4.836387983759572e-06,
"loss": 0.0274,
"step": 8660
},
{
"grad_norm": 0.13919097185134888,
"learning_rate": 4.765689982409816e-06,
"loss": 0.0254,
"step": 8670
},
{
"grad_norm": 0.16162839531898499,
"learning_rate": 4.695486656227233e-06,
"loss": 0.0272,
"step": 8680
},
{
"grad_norm": 0.20709174871444702,
"learning_rate": 4.625778772944156e-06,
"loss": 0.0258,
"step": 8690
},
{
"grad_norm": 0.15888750553131104,
"learning_rate": 4.556567094874825e-06,
"loss": 0.0221,
"step": 8700
},
{
"grad_norm": 0.1863916963338852,
"learning_rate": 4.487852378907059e-06,
"loss": 0.0246,
"step": 8710
},
{
"grad_norm": 0.1884496510028839,
"learning_rate": 4.419635376493986e-06,
"loss": 0.0234,
"step": 8720
},
{
"grad_norm": 0.19046708941459656,
"learning_rate": 4.351916833645825e-06,
"loss": 0.0264,
"step": 8730
},
{
"grad_norm": 0.2117229402065277,
"learning_rate": 4.284697490921691e-06,
"loss": 0.0254,
"step": 8740
},
{
"grad_norm": 0.15471944212913513,
"learning_rate": 4.2179780834215585e-06,
"loss": 0.0245,
"step": 8750
},
{
"grad_norm": 0.17657896876335144,
"learning_rate": 4.151759340778178e-06,
"loss": 0.0273,
"step": 8760
},
{
"grad_norm": 0.1651591807603836,
"learning_rate": 4.086041987149109e-06,
"loss": 0.0257,
"step": 8770
},
{
"grad_norm": 0.22000476717948914,
"learning_rate": 4.020826741208811e-06,
"loss": 0.0277,
"step": 8780
},
{
"grad_norm": 0.17889368534088135,
"learning_rate": 3.956114316140746e-06,
"loss": 0.0269,
"step": 8790
},
{
"grad_norm": 0.19767482578754425,
"learning_rate": 3.891905419629643e-06,
"loss": 0.0261,
"step": 8800
},
{
"grad_norm": 0.1601017266511917,
"learning_rate": 3.8282007538536946e-06,
"loss": 0.025,
"step": 8810
},
{
"grad_norm": 0.19517087936401367,
"learning_rate": 3.7650010154769265e-06,
"loss": 0.0311,
"step": 8820
},
{
"grad_norm": 0.19349902868270874,
"learning_rate": 3.7023068956415608e-06,
"loss": 0.0252,
"step": 8830
},
{
"grad_norm": 0.19826580584049225,
"learning_rate": 3.6401190799604303e-06,
"loss": 0.022,
"step": 8840
},
{
"grad_norm": 0.2037934958934784,
"learning_rate": 3.578438248509536e-06,
"loss": 0.0256,
"step": 8850
},
{
"grad_norm": 0.1612161546945572,
"learning_rate": 3.5172650758205583e-06,
"loss": 0.0256,
"step": 8860
},
{
"grad_norm": 0.19523052871227264,
"learning_rate": 3.45660023087353e-06,
"loss": 0.0222,
"step": 8870
},
{
"grad_norm": 0.15548686683177948,
"learning_rate": 3.3964443770894528e-06,
"loss": 0.0254,
"step": 8880
},
{
"grad_norm": 0.1960304081439972,
"learning_rate": 3.3367981723231245e-06,
"loss": 0.0221,
"step": 8890
},
{
"grad_norm": 0.28092971444129944,
"learning_rate": 3.2776622688558746e-06,
"loss": 0.0266,
"step": 8900
},
{
"grad_norm": 0.1964091807603836,
"learning_rate": 3.2190373133884677e-06,
"loss": 0.0238,
"step": 8910
},
{
"grad_norm": 0.2174736112356186,
"learning_rate": 3.1609239470340446e-06,
"loss": 0.0251,
"step": 8920
},
{
"grad_norm": 0.1292739063501358,
"learning_rate": 3.1033228053110373e-06,
"loss": 0.0238,
"step": 8930
},
{
"grad_norm": 0.22161857783794403,
"learning_rate": 3.0462345181363314e-06,
"loss": 0.0207,
"step": 8940
},
{
"grad_norm": 0.17582115530967712,
"learning_rate": 2.9896597098182654e-06,
"loss": 0.0273,
"step": 8950
},
{
"grad_norm": 0.18183830380439758,
"learning_rate": 2.933598999049891e-06,
"loss": 0.0271,
"step": 8960
},
{
"grad_norm": 0.17828331887722015,
"learning_rate": 2.8780529989021697e-06,
"loss": 0.0237,
"step": 8970
},
{
"grad_norm": 0.13881993293762207,
"learning_rate": 2.823022316817242e-06,
"loss": 0.0283,
"step": 8980
},
{
"grad_norm": 0.17303673923015594,
"learning_rate": 2.7685075546018456e-06,
"loss": 0.0245,
"step": 8990
},
{
"grad_norm": 0.22630803287029266,
"learning_rate": 2.7145093084206598e-06,
"loss": 0.027,
"step": 9000
},
{
"grad_norm": 0.1338975429534912,
"learning_rate": 2.661028168789892e-06,
"loss": 0.0207,
"step": 9010
},
{
"grad_norm": 0.14717504382133484,
"learning_rate": 2.6080647205706855e-06,
"loss": 0.0223,
"step": 9020
},
{
"grad_norm": 0.1378103345632553,
"learning_rate": 2.555619542962834e-06,
"loss": 0.024,
"step": 9030
},
{
"grad_norm": 0.1646696925163269,
"learning_rate": 2.503693209498409e-06,
"loss": 0.0234,
"step": 9040
},
{
"grad_norm": 0.13002635538578033,
"learning_rate": 2.452286288035449e-06,
"loss": 0.0227,
"step": 9050
},
{
"grad_norm": 0.14199601113796234,
"learning_rate": 2.4013993407518363e-06,
"loss": 0.024,
"step": 9060
},
{
"grad_norm": 0.18208181858062744,
"learning_rate": 2.351032924139063e-06,
"loss": 0.0248,
"step": 9070
},
{
"grad_norm": 0.17981268465518951,
"learning_rate": 2.30118758899619e-06,
"loss": 0.026,
"step": 9080
},
{
"grad_norm": 0.18774117529392242,
"learning_rate": 2.2518638804238157e-06,
"loss": 0.0255,
"step": 9090
},
{
"grad_norm": 0.11918246746063232,
"learning_rate": 2.203062337818118e-06,
"loss": 0.0263,
"step": 9100
},
{
"grad_norm": 0.16200493276119232,
"learning_rate": 2.1547834948649483e-06,
"loss": 0.0266,
"step": 9110
},
{
"grad_norm": 0.21640022099018097,
"learning_rate": 2.1070278795340017e-06,
"loss": 0.0231,
"step": 9120
},
{
"grad_norm": 0.21495278179645538,
"learning_rate": 2.059796014073029e-06,
"loss": 0.0221,
"step": 9130
},
{
"grad_norm": 0.19244349002838135,
"learning_rate": 2.01308841500214e-06,
"loss": 0.026,
"step": 9140
},
{
"grad_norm": 0.1763620376586914,
"learning_rate": 1.9669055931081704e-06,
"loss": 0.0261,
"step": 9150
},
{
"grad_norm": 0.15732762217521667,
"learning_rate": 1.9212480534390507e-06,
"loss": 0.0225,
"step": 9160
},
{
"grad_norm": 0.20162920653820038,
"learning_rate": 1.8761162952983246e-06,
"loss": 0.0274,
"step": 9170
},
{
"grad_norm": 0.17763090133666992,
"learning_rate": 1.8315108122396618e-06,
"loss": 0.0247,
"step": 9180
},
{
"grad_norm": 0.179799884557724,
"learning_rate": 1.787432092061475e-06,
"loss": 0.0265,
"step": 9190
},
{
"grad_norm": 0.19381265342235565,
"learning_rate": 1.743880616801602e-06,
"loss": 0.0314,
"step": 9200
},
{
"grad_norm": 0.18946729600429535,
"learning_rate": 1.7008568627319865e-06,
"loss": 0.0235,
"step": 9210
},
{
"grad_norm": 0.2259599268436432,
"learning_rate": 1.6583613003535226e-06,
"loss": 0.0278,
"step": 9220
},
{
"grad_norm": 0.16870799660682678,
"learning_rate": 1.6163943943908522e-06,
"loss": 0.0269,
"step": 9230
},
{
"grad_norm": 0.1372016966342926,
"learning_rate": 1.5749566037873476e-06,
"loss": 0.025,
"step": 9240
},
{
"grad_norm": 0.1781427413225174,
"learning_rate": 1.5340483817000428e-06,
"loss": 0.0252,
"step": 9250
},
{
"grad_norm": 0.14164814352989197,
"learning_rate": 1.4936701754947101e-06,
"loss": 0.0243,
"step": 9260
},
{
"grad_norm": 0.2262454330921173,
"learning_rate": 1.4538224267409361e-06,
"loss": 0.0255,
"step": 9270
},
{
"grad_norm": 0.14385128021240234,
"learning_rate": 1.414505571207314e-06,
"loss": 0.0235,
"step": 9280
},
{
"grad_norm": 0.14189815521240234,
"learning_rate": 1.3757200388566816e-06,
"loss": 0.0241,
"step": 9290
},
{
"grad_norm": 0.1388234943151474,
"learning_rate": 1.3374662538414074e-06,
"loss": 0.027,
"step": 9300
},
{
"grad_norm": 0.2098885476589203,
"learning_rate": 1.2997446344987617e-06,
"loss": 0.0225,
"step": 9310
},
{
"grad_norm": 0.21225465834140778,
"learning_rate": 1.262555593346315e-06,
"loss": 0.0246,
"step": 9320
},
{
"grad_norm": 0.2089921236038208,
"learning_rate": 1.2258995370774685e-06,
"loss": 0.0264,
"step": 9330
},
{
"grad_norm": 0.2232697308063507,
"learning_rate": 1.1897768665569798e-06,
"loss": 0.0253,
"step": 9340
},
{
"grad_norm": 0.16245770454406738,
"learning_rate": 1.1541879768165954e-06,
"loss": 0.0253,
"step": 9350
},
{
"grad_norm": 0.20593994855880737,
"learning_rate": 1.1191332570507085e-06,
"loss": 0.0262,
"step": 9360
},
{
"grad_norm": 0.15458564460277557,
"learning_rate": 1.0846130906121132e-06,
"loss": 0.0281,
"step": 9370
},
{
"grad_norm": 0.18079373240470886,
"learning_rate": 1.0506278550078131e-06,
"loss": 0.0279,
"step": 9380
},
{
"grad_norm": 0.2194763422012329,
"learning_rate": 1.0171779218949185e-06,
"loss": 0.0247,
"step": 9390
},
{
"grad_norm": 0.1311703622341156,
"learning_rate": 9.842636570765174e-07,
"loss": 0.0225,
"step": 9400
},
{
"grad_norm": 0.1503073275089264,
"learning_rate": 9.518854204977612e-07,
"loss": 0.0249,
"step": 9410
},
{
"grad_norm": 0.13787901401519775,
"learning_rate": 9.200435662418349e-07,
"loss": 0.0234,
"step": 9420
},
{
"grad_norm": 0.17687959969043732,
"learning_rate": 8.887384425261658e-07,
"loss": 0.0261,
"step": 9430
},
{
"grad_norm": 0.2061752825975418,
"learning_rate": 8.579703916985648e-07,
"loss": 0.025,
"step": 9440
},
{
"grad_norm": 0.11024852097034454,
"learning_rate": 8.277397502335194e-07,
"loss": 0.0236,
"step": 9450
},
{
"grad_norm": 0.22748751938343048,
"learning_rate": 7.980468487284675e-07,
"loss": 0.0252,
"step": 9460
},
{
"grad_norm": 0.16331540048122406,
"learning_rate": 7.688920119002297e-07,
"loss": 0.0255,
"step": 9470
},
{
"grad_norm": 0.13891828060150146,
"learning_rate": 7.402755585814269e-07,
"loss": 0.025,
"step": 9480
},
{
"grad_norm": 0.17650681734085083,
"learning_rate": 7.121978017170073e-07,
"loss": 0.0274,
"step": 9490
},
{
"grad_norm": 0.1940966695547104,
"learning_rate": 6.846590483608306e-07,
"loss": 0.0259,
"step": 9500
},
{
"grad_norm": 0.15129396319389343,
"learning_rate": 6.576595996722834e-07,
"loss": 0.0287,
"step": 9510
},
{
"grad_norm": 0.1747410148382187,
"learning_rate": 6.311997509130141e-07,
"loss": 0.0267,
"step": 9520
},
{
"grad_norm": 0.13517805933952332,
"learning_rate": 6.052797914436803e-07,
"loss": 0.0247,
"step": 9530
},
{
"grad_norm": 0.16980281472206116,
"learning_rate": 5.799000047208181e-07,
"loss": 0.029,
"step": 9540
},
{
"grad_norm": 0.2221260517835617,
"learning_rate": 5.550606682937054e-07,
"loss": 0.0254,
"step": 9550
},
{
"grad_norm": 0.136415496468544,
"learning_rate": 5.307620538013481e-07,
"loss": 0.0261,
"step": 9560
},
{
"grad_norm": 0.1942778080701828,
"learning_rate": 5.070044269694874e-07,
"loss": 0.025,
"step": 9570
},
{
"grad_norm": 0.1478036493062973,
"learning_rate": 4.837880476077417e-07,
"loss": 0.0255,
"step": 9580
},
{
"grad_norm": 0.1439913660287857,
"learning_rate": 4.6111316960670835e-07,
"loss": 0.0223,
"step": 9590
},
{
"grad_norm": 0.226765438914299,
"learning_rate": 4.389800409352218e-07,
"loss": 0.0244,
"step": 9600
},
{
"grad_norm": 0.1964910328388214,
"learning_rate": 4.173889036376277e-07,
"loss": 0.0262,
"step": 9610
},
{
"grad_norm": 0.15166446566581726,
"learning_rate": 3.963399938311463e-07,
"loss": 0.0279,
"step": 9620
},
{
"grad_norm": 0.16576242446899414,
"learning_rate": 3.7583354170328545e-07,
"loss": 0.0264,
"step": 9630
},
{
"grad_norm": 0.11313388496637344,
"learning_rate": 3.558697715093207e-07,
"loss": 0.0206,
"step": 9640
},
{
"grad_norm": 0.2559225261211395,
"learning_rate": 3.3644890156983576e-07,
"loss": 0.0281,
"step": 9650
},
{
"grad_norm": 0.18023236095905304,
"learning_rate": 3.175711442683638e-07,
"loss": 0.0265,
"step": 9660
},
{
"grad_norm": 0.1540888249874115,
"learning_rate": 2.9923670604902197e-07,
"loss": 0.0256,
"step": 9670
},
{
"grad_norm": 0.1585042029619217,
"learning_rate": 2.814457874143028e-07,
"loss": 0.0245,
"step": 9680
},
{
"grad_norm": 0.17425575852394104,
"learning_rate": 2.641985829228366e-07,
"loss": 0.0225,
"step": 9690
},
{
"grad_norm": 0.20115868747234344,
"learning_rate": 2.474952811872877e-07,
"loss": 0.0217,
"step": 9700
},
{
"grad_norm": 0.1685631275177002,
"learning_rate": 2.3133606487228397e-07,
"loss": 0.0285,
"step": 9710
},
{
"grad_norm": 0.19537140429019928,
"learning_rate": 2.157211106924295e-07,
"loss": 0.0257,
"step": 9720
},
{
"grad_norm": 0.20747897028923035,
"learning_rate": 2.006505894103672e-07,
"loss": 0.0262,
"step": 9730
},
{
"grad_norm": 0.15045535564422607,
"learning_rate": 1.8612466583489696e-07,
"loss": 0.0256,
"step": 9740
},
{
"grad_norm": 0.1590084433555603,
"learning_rate": 1.7214349881918834e-07,
"loss": 0.0238,
"step": 9750
},
{
"grad_norm": 0.15026725828647614,
"learning_rate": 1.5870724125904845e-07,
"loss": 0.0244,
"step": 9760
},
{
"grad_norm": 0.18402710556983948,
"learning_rate": 1.4581604009124006e-07,
"loss": 0.0239,
"step": 9770
},
{
"grad_norm": 0.1279493123292923,
"learning_rate": 1.334700362918717e-07,
"loss": 0.0256,
"step": 9780
},
{
"grad_norm": 0.14363504946231842,
"learning_rate": 1.2166936487486015e-07,
"loss": 0.0252,
"step": 9790
},
{
"grad_norm": 0.14517320692539215,
"learning_rate": 1.1041415489045914e-07,
"loss": 0.0243,
"step": 9800
},
{
"grad_norm": 0.13801448047161102,
"learning_rate": 9.970452942384412e-08,
"loss": 0.0277,
"step": 9810
},
{
"grad_norm": 0.15440776944160461,
"learning_rate": 8.954060559375754e-08,
"loss": 0.0223,
"step": 9820
},
{
"grad_norm": 0.14275850355625153,
"learning_rate": 7.99224945512489e-08,
"loss": 0.0289,
"step": 9830
},
{
"grad_norm": 0.22071236371994019,
"learning_rate": 7.085030147843675e-08,
"loss": 0.025,
"step": 9840
},
{
"grad_norm": 0.20374628901481628,
"learning_rate": 6.232412558736523e-08,
"loss": 0.0284,
"step": 9850
},
{
"grad_norm": 0.19557805359363556,
"learning_rate": 5.434406011893822e-08,
"loss": 0.0254,
"step": 9860
},
{
"grad_norm": 0.18048694729804993,
"learning_rate": 4.6910192341864664e-08,
"loss": 0.0227,
"step": 9870
},
{
"grad_norm": 0.13578152656555176,
"learning_rate": 4.0022603551737035e-08,
"loss": 0.0258,
"step": 9880
},
{
"grad_norm": 0.1913282424211502,
"learning_rate": 3.3681369070120985e-08,
"loss": 0.0277,
"step": 9890
},
{
"grad_norm": 0.16268610954284668,
"learning_rate": 2.7886558243744866e-08,
"loss": 0.0235,
"step": 9900
},
{
"grad_norm": 0.1985759139060974,
"learning_rate": 2.2638234443722596e-08,
"loss": 0.0255,
"step": 9910
},
{
"grad_norm": 0.231684610247612,
"learning_rate": 1.7936455064887504e-08,
"loss": 0.0249,
"step": 9920
},
{
"grad_norm": 0.17760765552520752,
"learning_rate": 1.378127152514841e-08,
"loss": 0.0205,
"step": 9930
},
{
"grad_norm": 0.15855513513088226,
"learning_rate": 1.0172729264917857e-08,
"loss": 0.0275,
"step": 9940
},
{
"grad_norm": 0.22830337285995483,
"learning_rate": 7.1108677466458215e-09,
"loss": 0.0265,
"step": 9950
},
{
"grad_norm": 0.17130911350250244,
"learning_rate": 4.595720454353414e-09,
"loss": 0.0251,
"step": 9960
},
{
"grad_norm": 0.23396386206150055,
"learning_rate": 2.627314893294264e-09,
"loss": 0.0276,
"step": 9970
},
{
"grad_norm": 0.1631205528974533,
"learning_rate": 1.2056725896270048e-09,
"loss": 0.025,
"step": 9980
},
{
"grad_norm": 0.20658209919929504,
"learning_rate": 3.308090902098826e-10,
"loss": 0.0278,
"step": 9990
},
{
"grad_norm": 0.14629384875297546,
"learning_rate": 2.7339624120159555e-12,
"loss": 0.0289,
"step": 10000
}
],
"logging_steps": 10,
"max_steps": 10000,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}