{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 6000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.32891911986707206, "learning_rate": 6e-06, "loss": 1.1219, "step": 10 }, { "grad_norm": 0.3098449768379839, "learning_rate": 1.2666666666666668e-05, "loss": 1.0973, "step": 20 }, { "grad_norm": 0.4181264541119447, "learning_rate": 1.9333333333333333e-05, "loss": 1.1477, "step": 30 }, { "grad_norm": 0.19579543583098616, "learning_rate": 2.6000000000000002e-05, "loss": 1.1164, "step": 40 }, { "grad_norm": 0.3284332677896478, "learning_rate": 3.266666666666667e-05, "loss": 1.11, "step": 50 }, { "grad_norm": 0.3727308019006094, "learning_rate": 3.933333333333333e-05, "loss": 1.102, "step": 60 }, { "grad_norm": 0.2919390957287582, "learning_rate": 4.600000000000001e-05, "loss": 1.1098, "step": 70 }, { "grad_norm": 0.2791731012903993, "learning_rate": 5.266666666666666e-05, "loss": 1.1014, "step": 80 }, { "grad_norm": 1.5006957040381796, "learning_rate": 5.9333333333333343e-05, "loss": 1.0455, "step": 90 }, { "grad_norm": 1.6697583640334281, "learning_rate": 6.6e-05, "loss": 1.0631, "step": 100 }, { "grad_norm": 1.0581249819843779, "learning_rate": 7.266666666666667e-05, "loss": 0.9525, "step": 110 }, { "grad_norm": 1.495679154954095, "learning_rate": 7.933333333333334e-05, "loss": 0.8818, "step": 120 }, { "grad_norm": 1.7002164446244465, "learning_rate": 8.6e-05, "loss": 0.6975, "step": 130 }, { "grad_norm": 1.4449488234731722, "learning_rate": 9.266666666666666e-05, "loss": 0.649, "step": 140 }, { "grad_norm": 0.8170425577719468, "learning_rate": 9.933333333333334e-05, "loss": 0.6441, "step": 150 }, { "grad_norm": 0.7357770437068394, "learning_rate": 9.999753945398704e-05, "loss": 0.6189, "step": 160 }, { "grad_norm": 1.7766889784950777, "learning_rate": 9.998903417374228e-05, "loss": 0.5883, "step": 170 }, { "grad_norm": 1.4756047439566644, "learning_rate": 9.997445481536973e-05, "loss": 0.5387, "step": 180 }, { "grad_norm": 0.8764409455410609, "learning_rate": 9.995380315038119e-05, "loss": 0.5031, "step": 190 }, { "grad_norm": 0.9924664451511096, "learning_rate": 9.99270816881235e-05, "loss": 0.5232, "step": 200 }, { "grad_norm": 1.4089022631518857, "learning_rate": 9.989429367547377e-05, "loss": 0.4816, "step": 210 }, { "grad_norm": 1.1808656693606936, "learning_rate": 9.985544309644475e-05, "loss": 0.4858, "step": 220 }, { "grad_norm": 0.9841995393277984, "learning_rate": 9.98105346717008e-05, "loss": 0.4658, "step": 230 }, { "grad_norm": 0.9248665540993384, "learning_rate": 9.97595738579843e-05, "loss": 0.4728, "step": 240 }, { "grad_norm": 0.6299008235049119, "learning_rate": 9.970256684745258e-05, "loss": 0.4468, "step": 250 }, { "grad_norm": 0.5953267637014983, "learning_rate": 9.963952056692549e-05, "loss": 0.4396, "step": 260 }, { "grad_norm": 0.7711853896316084, "learning_rate": 9.957044267704384e-05, "loss": 0.44, "step": 270 }, { "grad_norm": 0.8074427398365275, "learning_rate": 9.949534157133844e-05, "loss": 0.4768, "step": 280 }, { "grad_norm": 0.7120900824090433, "learning_rate": 9.941422637521035e-05, "loss": 0.4044, "step": 290 }, { "grad_norm": 1.0687579981315425, "learning_rate": 9.932710694482191e-05, "loss": 0.4854, "step": 300 }, { "grad_norm": 0.7665545463907436, "learning_rate": 9.923399386589933e-05, "loss": 0.4597, "step": 310 }, { "grad_norm": 0.9900244322431292, "learning_rate": 9.913489845244626e-05, "loss": 0.4478, "step": 320 }, { "grad_norm": 0.7068241634751296, "learning_rate": 9.902983274536912e-05, "loss": 0.4409, "step": 330 }, { "grad_norm": 0.7128985499020303, "learning_rate": 9.891880951101407e-05, "loss": 0.4004, "step": 340 }, { "grad_norm": 0.9602818335769777, "learning_rate": 9.880184223961573e-05, "loss": 0.4853, "step": 350 }, { "grad_norm": 0.788175422958031, "learning_rate": 9.867894514365802e-05, "loss": 0.4325, "step": 360 }, { "grad_norm": 0.8100614979471025, "learning_rate": 9.855013315614725e-05, "loss": 0.3659, "step": 370 }, { "grad_norm": 0.9012392872081991, "learning_rate": 9.841542192879762e-05, "loss": 0.4218, "step": 380 }, { "grad_norm": 0.8401109245455818, "learning_rate": 9.82748278301294e-05, "loss": 0.4396, "step": 390 }, { "grad_norm": 0.8786182406061516, "learning_rate": 9.812836794348004e-05, "loss": 0.3853, "step": 400 }, { "grad_norm": 0.858313028630255, "learning_rate": 9.797606006492841e-05, "loss": 0.4363, "step": 410 }, { "grad_norm": 0.670516197724105, "learning_rate": 9.781792270113241e-05, "loss": 0.3944, "step": 420 }, { "grad_norm": 0.46364295813930684, "learning_rate": 9.765397506708023e-05, "loss": 0.4254, "step": 430 }, { "grad_norm": 0.5443653764812083, "learning_rate": 9.748423708375563e-05, "loss": 0.4198, "step": 440 }, { "grad_norm": 0.8438781472769962, "learning_rate": 9.730872937571739e-05, "loss": 0.4578, "step": 450 }, { "grad_norm": 0.5063270562454627, "learning_rate": 9.712747326859315e-05, "loss": 0.4293, "step": 460 }, { "grad_norm": 0.6835439376770004, "learning_rate": 9.69404907864883e-05, "loss": 0.3997, "step": 470 }, { "grad_norm": 0.39639175041302077, "learning_rate": 9.674780464930979e-05, "loss": 0.3911, "step": 480 }, { "grad_norm": 0.7763841643676905, "learning_rate": 9.654943827000548e-05, "loss": 0.424, "step": 490 }, { "grad_norm": 0.5285780936709736, "learning_rate": 9.634541575171929e-05, "loss": 0.3793, "step": 500 }, { "grad_norm": 0.5361899627895629, "learning_rate": 9.613576188486253e-05, "loss": 0.4569, "step": 510 }, { "grad_norm": 0.5431711864470214, "learning_rate": 9.59205021441015e-05, "loss": 0.3986, "step": 520 }, { "grad_norm": 0.9352274828548598, "learning_rate": 9.569966268526232e-05, "loss": 0.4188, "step": 530 }, { "grad_norm": 0.5893501607547204, "learning_rate": 9.54732703421526e-05, "loss": 0.4694, "step": 540 }, { "grad_norm": 0.6557089384702188, "learning_rate": 9.524135262330098e-05, "loss": 0.4188, "step": 550 }, { "grad_norm": 0.4074521903845341, "learning_rate": 9.50039377086147e-05, "loss": 0.431, "step": 560 }, { "grad_norm": 0.807674419182726, "learning_rate": 9.476105444595534e-05, "loss": 0.4484, "step": 570 }, { "grad_norm": 0.5428174214895874, "learning_rate": 9.451273234763371e-05, "loss": 0.4318, "step": 580 }, { "grad_norm": 0.5651447014731802, "learning_rate": 9.425900158682385e-05, "loss": 0.3861, "step": 590 }, { "grad_norm": 0.5768624773564572, "learning_rate": 9.399989299389661e-05, "loss": 0.4383, "step": 600 }, { "grad_norm": 0.6335510721145432, "learning_rate": 9.373543805267368e-05, "loss": 0.4428, "step": 610 }, { "grad_norm": 0.4747736982292791, "learning_rate": 9.346566889660193e-05, "loss": 0.4204, "step": 620 }, { "grad_norm": 0.9224269884452101, "learning_rate": 9.319061830484898e-05, "loss": 0.4311, "step": 630 }, { "grad_norm": 0.5490877844459744, "learning_rate": 9.291031969832026e-05, "loss": 0.4143, "step": 640 }, { "grad_norm": 0.569696837170464, "learning_rate": 9.262480713559808e-05, "loss": 0.4477, "step": 650 }, { "grad_norm": 0.7174758360572557, "learning_rate": 9.233411530880326e-05, "loss": 0.4402, "step": 660 }, { "grad_norm": 0.9511876690055683, "learning_rate": 9.20382795393797e-05, "loss": 0.351, "step": 670 }, { "grad_norm": 0.5468780679526735, "learning_rate": 9.173733577380258e-05, "loss": 0.4402, "step": 680 }, { "grad_norm": 0.513725411976899, "learning_rate": 9.143132057921058e-05, "loss": 0.4079, "step": 690 }, { "grad_norm": 0.5544523963661264, "learning_rate": 9.112027113896262e-05, "loss": 0.4054, "step": 700 }, { "grad_norm": 0.6257646386900859, "learning_rate": 9.080422524811982e-05, "loss": 0.4201, "step": 710 }, { "grad_norm": 0.49134281020233195, "learning_rate": 9.048322130885305e-05, "loss": 0.3895, "step": 720 }, { "grad_norm": 0.5514139930549428, "learning_rate": 9.015729832577681e-05, "loss": 0.4292, "step": 730 }, { "grad_norm": 0.591406648502766, "learning_rate": 8.982649590120982e-05, "loss": 0.3767, "step": 740 }, { "grad_norm": 0.57901018226806, "learning_rate": 8.949085423036296e-05, "loss": 0.3966, "step": 750 }, { "grad_norm": 0.5942028030796065, "learning_rate": 8.91504140964553e-05, "loss": 0.4213, "step": 760 }, { "grad_norm": 0.6468413248638392, "learning_rate": 8.880521686575857e-05, "loss": 0.3838, "step": 770 }, { "grad_norm": 0.6104050075372115, "learning_rate": 8.845530448257085e-05, "loss": 0.3847, "step": 780 }, { "grad_norm": 0.49591808277621635, "learning_rate": 8.810071946411989e-05, "loss": 0.3652, "step": 790 }, { "grad_norm": 0.7105616648052399, "learning_rate": 8.774150489539707e-05, "loss": 0.4432, "step": 800 }, { "grad_norm": 0.5374681588991885, "learning_rate": 8.737770442392212e-05, "loss": 0.3882, "step": 810 }, { "grad_norm": 0.6971839316910557, "learning_rate": 8.700936225443959e-05, "loss": 0.4596, "step": 820 }, { "grad_norm": 0.6443790488497669, "learning_rate": 8.663652314354765e-05, "loss": 0.4235, "step": 830 }, { "grad_norm": 0.7752998515035786, "learning_rate": 8.625923239425978e-05, "loss": 0.3917, "step": 840 }, { "grad_norm": 0.5721013367090486, "learning_rate": 8.587753585050004e-05, "loss": 0.3741, "step": 850 }, { "grad_norm": 0.5175054601422239, "learning_rate": 8.549147989153276e-05, "loss": 0.446, "step": 860 }, { "grad_norm": 0.5249955057403916, "learning_rate": 8.510111142632698e-05, "loss": 0.4107, "step": 870 }, { "grad_norm": 0.4768163240560392, "learning_rate": 8.470647788785665e-05, "loss": 0.3631, "step": 880 }, { "grad_norm": 0.49316856034517426, "learning_rate": 8.430762722733714e-05, "loss": 0.3907, "step": 890 }, { "grad_norm": 0.4841369149123188, "learning_rate": 8.390460790839882e-05, "loss": 0.3946, "step": 900 }, { "grad_norm": 0.6258621497596762, "learning_rate": 8.349746890119826e-05, "loss": 0.3921, "step": 910 }, { "grad_norm": 0.4379088656974616, "learning_rate": 8.308625967646795e-05, "loss": 0.4244, "step": 920 }, { "grad_norm": 0.8040486929657406, "learning_rate": 8.267103019950529e-05, "loss": 0.4847, "step": 930 }, { "grad_norm": 0.6091336739708904, "learning_rate": 8.225183092410128e-05, "loss": 0.3898, "step": 940 }, { "grad_norm": 0.7052045014465564, "learning_rate": 8.182871278641009e-05, "loss": 0.4031, "step": 950 }, { "grad_norm": 0.4141336399169555, "learning_rate": 8.140172719875979e-05, "loss": 0.4101, "step": 960 }, { "grad_norm": 0.6343892664305733, "learning_rate": 8.097092604340542e-05, "loss": 0.4011, "step": 970 }, { "grad_norm": 0.4258768669513822, "learning_rate": 8.053636166622476e-05, "loss": 0.3894, "step": 980 }, { "grad_norm": 0.7442886571443741, "learning_rate": 8.009808687035798e-05, "loss": 0.3775, "step": 990 }, { "grad_norm": 0.5432565061457083, "learning_rate": 7.965615490979163e-05, "loss": 0.4056, "step": 1000 }, { "grad_norm": 0.5478317609562899, "learning_rate": 7.921061948288773e-05, "loss": 0.4378, "step": 1010 }, { "grad_norm": 0.6517636073390697, "learning_rate": 7.87615347258591e-05, "loss": 0.3564, "step": 1020 }, { "grad_norm": 0.42114687622382974, "learning_rate": 7.830895520619128e-05, "loss": 0.3889, "step": 1030 }, { "grad_norm": 0.49241335758021804, "learning_rate": 7.785293591601217e-05, "loss": 0.3747, "step": 1040 }, { "grad_norm": 0.5511725521343366, "learning_rate": 7.739353226541009e-05, "loss": 0.3806, "step": 1050 }, { "grad_norm": 0.48703133055193043, "learning_rate": 7.693080007570084e-05, "loss": 0.3998, "step": 1060 }, { "grad_norm": 0.48744693058604066, "learning_rate": 7.646479557264513e-05, "loss": 0.3493, "step": 1070 }, { "grad_norm": 0.6468318085032428, "learning_rate": 7.599557537961663e-05, "loss": 0.4098, "step": 1080 }, { "grad_norm": 0.5029988612910242, "learning_rate": 7.552319651072164e-05, "loss": 0.3981, "step": 1090 }, { "grad_norm": 0.8030149468673661, "learning_rate": 7.504771636387163e-05, "loss": 0.3685, "step": 1100 }, { "grad_norm": 0.5922838510741312, "learning_rate": 7.456919271380875e-05, "loss": 0.3727, "step": 1110 }, { "grad_norm": 0.4640284231465197, "learning_rate": 7.408768370508576e-05, "loss": 0.3632, "step": 1120 }, { "grad_norm": 0.464490141169679, "learning_rate": 7.36032478450011e-05, "loss": 0.3963, "step": 1130 }, { "grad_norm": 0.6565535257465489, "learning_rate": 7.311594399648957e-05, "loss": 0.3779, "step": 1140 }, { "grad_norm": 0.5454852182945461, "learning_rate": 7.262583137097018e-05, "loss": 0.3937, "step": 1150 }, { "grad_norm": 0.6603272546026309, "learning_rate": 7.213296952115144e-05, "loss": 0.3902, "step": 1160 }, { "grad_norm": 0.4207924682462282, "learning_rate": 7.16374183337951e-05, "loss": 0.4333, "step": 1170 }, { "grad_norm": 0.5331475944967922, "learning_rate": 7.113923802243957e-05, "loss": 0.426, "step": 1180 }, { "grad_norm": 0.5008474841640382, "learning_rate": 7.06384891200834e-05, "loss": 0.4039, "step": 1190 }, { "grad_norm": 0.6014599965760046, "learning_rate": 7.013523247183e-05, "loss": 0.4055, "step": 1200 }, { "grad_norm": 0.5701064209933435, "learning_rate": 6.962952922749457e-05, "loss": 0.444, "step": 1210 }, { "grad_norm": 0.5456782851870061, "learning_rate": 6.912144083417376e-05, "loss": 0.4453, "step": 1220 }, { "grad_norm": 0.5548604832656555, "learning_rate": 6.861102902877946e-05, "loss": 0.3869, "step": 1230 }, { "grad_norm": 0.4167451491019649, "learning_rate": 6.809835583053715e-05, "loss": 0.3879, "step": 1240 }, { "grad_norm": 0.5687360672126506, "learning_rate": 6.758348353345014e-05, "loss": 0.3794, "step": 1250 }, { "grad_norm": 0.3893735383088963, "learning_rate": 6.706647469873031e-05, "loss": 0.3591, "step": 1260 }, { "grad_norm": 0.4872065108814663, "learning_rate": 6.654739214719641e-05, "loss": 0.3502, "step": 1270 }, { "grad_norm": 0.6657768797277458, "learning_rate": 6.602629895164081e-05, "loss": 0.4112, "step": 1280 }, { "grad_norm": 0.78954710520931, "learning_rate": 6.550325842916559e-05, "loss": 0.4277, "step": 1290 }, { "grad_norm": 0.4413187868087259, "learning_rate": 6.497833413348909e-05, "loss": 0.4209, "step": 1300 }, { "grad_norm": 0.661878533452368, "learning_rate": 6.445158984722358e-05, "loss": 0.3809, "step": 1310 }, { "grad_norm": 0.5859572019350382, "learning_rate": 6.39230895741251e-05, "loss": 0.4197, "step": 1320 }, { "grad_norm": 0.5324974689501786, "learning_rate": 6.339289753131649e-05, "loss": 0.3896, "step": 1330 }, { "grad_norm": 0.40591577239013543, "learning_rate": 6.286107814148454e-05, "loss": 0.4304, "step": 1340 }, { "grad_norm": 0.3825141506111005, "learning_rate": 6.232769602505203e-05, "loss": 0.4013, "step": 1350 }, { "grad_norm": 0.5339582662898721, "learning_rate": 6.179281599232591e-05, "loss": 0.427, "step": 1360 }, { "grad_norm": 0.45983991897801046, "learning_rate": 6.125650303562221e-05, "loss": 0.3822, "step": 1370 }, { "grad_norm": 0.44477920726515213, "learning_rate": 6.071882232136901e-05, "loss": 0.3796, "step": 1380 }, { "grad_norm": 0.4305086680095017, "learning_rate": 6.017983918218812e-05, "loss": 0.3836, "step": 1390 }, { "grad_norm": 0.4011625803186596, "learning_rate": 5.963961910895676e-05, "loss": 0.4132, "step": 1400 }, { "grad_norm": 0.45016244743449424, "learning_rate": 5.909822774284971e-05, "loss": 0.4292, "step": 1410 }, { "grad_norm": 0.6506561624849786, "learning_rate": 5.85557308673635e-05, "loss": 0.3784, "step": 1420 }, { "grad_norm": 0.40114112830871923, "learning_rate": 5.8012194400323116e-05, "loss": 0.4375, "step": 1430 }, { "grad_norm": 0.5668284827174664, "learning_rate": 5.746768438587245e-05, "loss": 0.3907, "step": 1440 }, { "grad_norm": 0.475005072715766, "learning_rate": 5.692226698644938e-05, "loss": 0.3831, "step": 1450 }, { "grad_norm": 0.5790271336807717, "learning_rate": 5.637600847474656e-05, "loss": 0.3941, "step": 1460 }, { "grad_norm": 0.41214760111447596, "learning_rate": 5.5828975225658666e-05, "loss": 0.3475, "step": 1470 }, { "grad_norm": 0.42325735215830074, "learning_rate": 5.52812337082173e-05, "loss": 0.3556, "step": 1480 }, { "grad_norm": 0.42783716167434893, "learning_rate": 5.473285047751451e-05, "loss": 0.3799, "step": 1490 }, { "grad_norm": 0.4120856827992062, "learning_rate": 5.418389216661579e-05, "loss": 0.4021, "step": 1500 }, { "grad_norm": 0.5136582964093963, "learning_rate": 5.363442547846356e-05, "loss": 0.3268, "step": 1510 }, { "grad_norm": 0.4343113264221542, "learning_rate": 5.308451717777228e-05, "loss": 0.3874, "step": 1520 }, { "grad_norm": 0.7559532951429224, "learning_rate": 5.2534234082915976e-05, "loss": 0.3926, "step": 1530 }, { "grad_norm": 0.4531249399534264, "learning_rate": 5.198364305780922e-05, "loss": 0.3652, "step": 1540 }, { "grad_norm": 0.3676309185559356, "learning_rate": 5.143281100378261e-05, "loss": 0.3999, "step": 1550 }, { "grad_norm": 0.502314567823685, "learning_rate": 5.088180485145378e-05, "loss": 0.3529, "step": 1560 }, { "grad_norm": 0.4008597844838749, "learning_rate": 5.033069155259471e-05, "loss": 0.3652, "step": 1570 }, { "grad_norm": 0.38221955681810893, "learning_rate": 4.97795380719966e-05, "loss": 0.3971, "step": 1580 }, { "grad_norm": 0.514195212347906, "learning_rate": 4.9228411379333014e-05, "loss": 0.3992, "step": 1590 }, { "grad_norm": 0.506444609566808, "learning_rate": 4.867737844102261e-05, "loss": 0.3782, "step": 1600 }, { "grad_norm": 0.5022376180599458, "learning_rate": 4.812650621209209e-05, "loss": 0.4225, "step": 1610 }, { "grad_norm": 0.372488244698932, "learning_rate": 4.7575861628040635e-05, "loss": 0.3889, "step": 1620 }, { "grad_norm": 0.5183410430700899, "learning_rate": 4.702551159670672e-05, "loss": 0.4337, "step": 1630 }, { "grad_norm": 0.8020065520501941, "learning_rate": 4.647552299013828e-05, "loss": 0.3991, "step": 1640 }, { "grad_norm": 0.4379403117006386, "learning_rate": 4.5925962636467126e-05, "loss": 0.3672, "step": 1650 }, { "grad_norm": 0.38911255759788993, "learning_rate": 4.537689731178883e-05, "loss": 0.4055, "step": 1660 }, { "grad_norm": 0.5811776934787003, "learning_rate": 4.482839373204891e-05, "loss": 0.3826, "step": 1670 }, { "grad_norm": 0.4674545545817927, "learning_rate": 4.428051854493623e-05, "loss": 0.3756, "step": 1680 }, { "grad_norm": 0.43782853364275837, "learning_rate": 4.373333832178478e-05, "loss": 0.4068, "step": 1690 }, { "grad_norm": 0.4921322439969929, "learning_rate": 4.3186919549484784e-05, "loss": 0.417, "step": 1700 }, { "grad_norm": 0.3715716566384778, "learning_rate": 4.264132862240387e-05, "loss": 0.3738, "step": 1710 }, { "grad_norm": 0.4455140693455599, "learning_rate": 4.209663183431969e-05, "loss": 0.4155, "step": 1720 }, { "grad_norm": 0.4405701573035707, "learning_rate": 4.155289537036466e-05, "loss": 0.3584, "step": 1730 }, { "grad_norm": 0.7669481505789998, "learning_rate": 4.1010185298983984e-05, "loss": 0.397, "step": 1740 }, { "grad_norm": 0.610261988778607, "learning_rate": 4.046856756390767e-05, "loss": 0.3807, "step": 1750 }, { "grad_norm": 0.48776426794284755, "learning_rate": 3.9928107976137906e-05, "loss": 0.3875, "step": 1760 }, { "grad_norm": 0.49321577369867364, "learning_rate": 3.9388872205952526e-05, "loss": 0.3672, "step": 1770 }, { "grad_norm": 0.5892170576667044, "learning_rate": 3.8850925774925425e-05, "loss": 0.3511, "step": 1780 }, { "grad_norm": 0.450083481802323, "learning_rate": 3.831433404796521e-05, "loss": 0.4209, "step": 1790 }, { "grad_norm": 0.4068548750362664, "learning_rate": 3.777916222537285e-05, "loss": 0.388, "step": 1800 }, { "grad_norm": 0.451378947949752, "learning_rate": 3.7245475334919246e-05, "loss": 0.4013, "step": 1810 }, { "grad_norm": 0.4062080818228897, "learning_rate": 3.6713338223943867e-05, "loss": 0.3336, "step": 1820 }, { "grad_norm": 0.35179979071321366, "learning_rate": 3.618281555147522e-05, "loss": 0.3849, "step": 1830 }, { "grad_norm": 0.36860252149181366, "learning_rate": 3.5653971780374295e-05, "loss": 0.4105, "step": 1840 }, { "grad_norm": 0.3516448724512693, "learning_rate": 3.512687116950182e-05, "loss": 0.3882, "step": 1850 }, { "grad_norm": 0.4543172360621974, "learning_rate": 3.460157776591018e-05, "loss": 0.3877, "step": 1860 }, { "grad_norm": 0.3588443958432708, "learning_rate": 3.407815539706124e-05, "loss": 0.3548, "step": 1870 }, { "grad_norm": 0.4288623298768492, "learning_rate": 3.355666766307084e-05, "loss": 0.3711, "step": 1880 }, { "grad_norm": 0.41722233803436054, "learning_rate": 3.3037177928980735e-05, "loss": 0.4394, "step": 1890 }, { "grad_norm": 0.3963644889744335, "learning_rate": 3.251974931705933e-05, "loss": 0.3909, "step": 1900 }, { "grad_norm": 0.42188404963808807, "learning_rate": 3.2004444699131727e-05, "loss": 0.3724, "step": 1910 }, { "grad_norm": 0.4916869956695671, "learning_rate": 3.1491326688940345e-05, "loss": 0.3317, "step": 1920 }, { "grad_norm": 0.3217948526892873, "learning_rate": 3.098045763453678e-05, "loss": 0.3789, "step": 1930 }, { "grad_norm": 0.4911198074158023, "learning_rate": 3.0471899610706038e-05, "loss": 0.3455, "step": 1940 }, { "grad_norm": 0.3812003393401496, "learning_rate": 2.9965714411423972e-05, "loss": 0.4118, "step": 1950 }, { "grad_norm": 0.40967352871648477, "learning_rate": 2.9461963542348737e-05, "loss": 0.3921, "step": 1960 }, { "grad_norm": 0.4409375861217166, "learning_rate": 2.8960708213347366e-05, "loss": 0.392, "step": 1970 }, { "grad_norm": 0.3386711362853152, "learning_rate": 2.846200933105829e-05, "loss": 0.3695, "step": 1980 }, { "grad_norm": 0.42904099524353656, "learning_rate": 2.7965927491490705e-05, "loss": 0.3927, "step": 1990 }, { "grad_norm": 0.4029328307252031, "learning_rate": 2.747252297266162e-05, "loss": 0.3613, "step": 2000 }, { "grad_norm": 0.6039300549943266, "learning_rate": 2.698185572727151e-05, "loss": 0.3583, "step": 2010 }, { "grad_norm": 0.3414696961993867, "learning_rate": 2.6493985375419778e-05, "loss": 0.3511, "step": 2020 }, { "grad_norm": 0.3318722775583879, "learning_rate": 2.6008971197360176e-05, "loss": 0.4089, "step": 2030 }, { "grad_norm": 0.43625669119802263, "learning_rate": 2.552687212629799e-05, "loss": 0.3659, "step": 2040 }, { "grad_norm": 0.4172201201713923, "learning_rate": 2.5047746741228978e-05, "loss": 0.3788, "step": 2050 }, { "grad_norm": 0.43361968656212774, "learning_rate": 2.4571653259821694e-05, "loss": 0.3722, "step": 2060 }, { "grad_norm": 0.39623078232292763, "learning_rate": 2.4098649531343497e-05, "loss": 0.4076, "step": 2070 }, { "grad_norm": 0.501851848826552, "learning_rate": 2.362879302963135e-05, "loss": 0.4006, "step": 2080 }, { "grad_norm": 0.5703110945642815, "learning_rate": 2.3162140846108366e-05, "loss": 0.3521, "step": 2090 }, { "grad_norm": 0.4358442068058637, "learning_rate": 2.2698749682846687e-05, "loss": 0.3632, "step": 2100 }, { "grad_norm": 0.30904300920371297, "learning_rate": 2.2238675845677663e-05, "loss": 0.3833, "step": 2110 }, { "grad_norm": 0.35328368209125, "learning_rate": 2.1781975237350366e-05, "loss": 0.4123, "step": 2120 }, { "grad_norm": 0.34113080543928337, "learning_rate": 2.1328703350738765e-05, "loss": 0.4196, "step": 2130 }, { "grad_norm": 0.33599240764625266, "learning_rate": 2.0878915262099098e-05, "loss": 0.3941, "step": 2140 }, { "grad_norm": 0.35695134718399285, "learning_rate": 2.0432665624377434e-05, "loss": 0.3648, "step": 2150 }, { "grad_norm": 0.3162560933004789, "learning_rate": 1.999000866056908e-05, "loss": 0.3515, "step": 2160 }, { "grad_norm": 0.34885679759789284, "learning_rate": 1.9550998157129946e-05, "loss": 0.3645, "step": 2170 }, { "grad_norm": 0.3231142852032699, "learning_rate": 1.9115687457441022e-05, "loss": 0.3254, "step": 2180 }, { "grad_norm": 0.3718936654720125, "learning_rate": 1.868412945532681e-05, "loss": 0.3846, "step": 2190 }, { "grad_norm": 0.44941123488383783, "learning_rate": 1.8256376588628238e-05, "loss": 0.3666, "step": 2200 }, { "grad_norm": 0.36955389782934217, "learning_rate": 1.7832480832830987e-05, "loss": 0.3811, "step": 2210 }, { "grad_norm": 0.361870477179932, "learning_rate": 1.7412493694750176e-05, "loss": 0.3633, "step": 2220 }, { "grad_norm": 0.3520593894407223, "learning_rate": 1.699646620627168e-05, "loss": 0.374, "step": 2230 }, { "grad_norm": 0.35560348717377765, "learning_rate": 1.658444891815152e-05, "loss": 0.3309, "step": 2240 }, { "grad_norm": 0.44538308670514687, "learning_rate": 1.617649189387337e-05, "loss": 0.4025, "step": 2250 }, { "grad_norm": 0.33803782431333723, "learning_rate": 1.5772644703565565e-05, "loss": 0.3874, "step": 2260 }, { "grad_norm": 0.3875104821889467, "learning_rate": 1.537295641797785e-05, "loss": 0.3872, "step": 2270 }, { "grad_norm": 0.41357843187063564, "learning_rate": 1.4977475602518876e-05, "loss": 0.3609, "step": 2280 }, { "grad_norm": 0.3733742182040787, "learning_rate": 1.4586250311355132e-05, "loss": 0.3598, "step": 2290 }, { "grad_norm": 0.4264278862512874, "learning_rate": 1.4199328081572e-05, "loss": 0.3681, "step": 2300 }, { "grad_norm": 0.4509137112261884, "learning_rate": 1.3816755927397502e-05, "loss": 0.3595, "step": 2310 }, { "grad_norm": 0.32369694644625163, "learning_rate": 1.343858033448982e-05, "loss": 0.3733, "step": 2320 }, { "grad_norm": 0.46587408173897793, "learning_rate": 1.3064847254288797e-05, "loss": 0.4173, "step": 2330 }, { "grad_norm": 0.42358906145456865, "learning_rate": 1.2695602098432502e-05, "loss": 0.3501, "step": 2340 }, { "grad_norm": 0.36667757830111064, "learning_rate": 1.233088973323937e-05, "loss": 0.3646, "step": 2350 }, { "grad_norm": 0.3896211745421265, "learning_rate": 1.1970754474256563e-05, "loss": 0.3729, "step": 2360 }, { "grad_norm": 0.5272010673506055, "learning_rate": 1.16152400808752e-05, "loss": 0.3607, "step": 2370 }, { "grad_norm": 0.4024134697307971, "learning_rate": 1.1264389751013326e-05, "loss": 0.352, "step": 2380 }, { "grad_norm": 0.3603106078255842, "learning_rate": 1.0918246115866964e-05, "loss": 0.3782, "step": 2390 }, { "grad_norm": 0.3691157619252364, "learning_rate": 1.0576851234730095e-05, "loss": 0.3599, "step": 2400 }, { "grad_norm": 0.5104764852607075, "learning_rate": 1.0240246589884044e-05, "loss": 0.3818, "step": 2410 }, { "grad_norm": 0.3696447725214109, "learning_rate": 9.908473081557151e-06, "loss": 0.364, "step": 2420 }, { "grad_norm": 0.32217128659199723, "learning_rate": 9.581571022954988e-06, "loss": 0.3583, "step": 2430 }, { "grad_norm": 0.3520731433103786, "learning_rate": 9.259580135361929e-06, "loss": 0.3632, "step": 2440 }, { "grad_norm": 0.4594438378072774, "learning_rate": 8.9425395433148e-06, "loss": 0.3755, "step": 2450 }, { "grad_norm": 0.406527783704118, "learning_rate": 8.630487769848877e-06, "loss": 0.3805, "step": 2460 }, { "grad_norm": 0.43058452666748237, "learning_rate": 8.323462731816961e-06, "loss": 0.3306, "step": 2470 }, { "grad_norm": 0.3025098226233087, "learning_rate": 8.021501735282266e-06, "loss": 0.3597, "step": 2480 }, { "grad_norm": 0.37851030745345665, "learning_rate": 7.724641470985378e-06, "loss": 0.3878, "step": 2490 }, { "grad_norm": 0.4518441526603634, "learning_rate": 7.432918009885997e-06, "loss": 0.3732, "step": 2500 }, { "grad_norm": 0.39962364927940536, "learning_rate": 7.146366798780096e-06, "loss": 0.4038, "step": 2510 }, { "grad_norm": 0.41893974273287166, "learning_rate": 6.865022655992798e-06, "loss": 0.3541, "step": 2520 }, { "grad_norm": 0.37562955141208276, "learning_rate": 6.588919767147639e-06, "loss": 0.3424, "step": 2530 }, { "grad_norm": 0.3228538408725492, "learning_rate": 6.318091681012772e-06, "loss": 0.4431, "step": 2540 }, { "grad_norm": 0.3215962168635294, "learning_rate": 6.052571305424531e-06, "loss": 0.3914, "step": 2550 }, { "grad_norm": 0.3568843213952275, "learning_rate": 5.79239090328883e-06, "loss": 0.3858, "step": 2560 }, { "grad_norm": 0.31573805342470757, "learning_rate": 5.537582088660937e-06, "loss": 0.3771, "step": 2570 }, { "grad_norm": 0.3352823181623894, "learning_rate": 5.28817582290414e-06, "loss": 0.402, "step": 2580 }, { "grad_norm": 0.32400014562935714, "learning_rate": 5.044202410927706e-06, "loss": 0.3392, "step": 2590 }, { "grad_norm": 0.33161920281675344, "learning_rate": 4.805691497504505e-06, "loss": 0.3725, "step": 2600 }, { "grad_norm": 0.35434556297292824, "learning_rate": 4.57267206366902e-06, "loss": 0.4131, "step": 2610 }, { "grad_norm": 0.4806235812087971, "learning_rate": 4.3451724231958644e-06, "loss": 0.3891, "step": 2620 }, { "grad_norm": 0.3686062566567837, "learning_rate": 4.123220219159418e-06, "loss": 0.3859, "step": 2630 }, { "grad_norm": 0.3819553305443249, "learning_rate": 3.90684242057498e-06, "loss": 0.377, "step": 2640 }, { "grad_norm": 0.38645564376866687, "learning_rate": 3.696065319121833e-06, "loss": 0.3433, "step": 2650 }, { "grad_norm": 0.3862754884347339, "learning_rate": 3.4909145259485744e-06, "loss": 0.3492, "step": 2660 }, { "grad_norm": 0.3438116939889349, "learning_rate": 3.2914149685611073e-06, "loss": 0.3814, "step": 2670 }, { "grad_norm": 0.4323218087421228, "learning_rate": 3.0975908877938277e-06, "loss": 0.3604, "step": 2680 }, { "grad_norm": 0.3144911455036637, "learning_rate": 2.9094658348640945e-06, "loss": 0.3514, "step": 2690 }, { "grad_norm": 0.31680973314520106, "learning_rate": 2.7270626685105828e-06, "loss": 0.3562, "step": 2700 }, { "grad_norm": 0.35789606050557055, "learning_rate": 2.5504035522157854e-06, "loss": 0.3381, "step": 2710 }, { "grad_norm": 0.3534238539283105, "learning_rate": 2.379509951512937e-06, "loss": 0.3808, "step": 2720 }, { "grad_norm": 0.33976927726458955, "learning_rate": 2.214402631377782e-06, "loss": 0.313, "step": 2730 }, { "grad_norm": 0.42660376334862016, "learning_rate": 2.0551016537054493e-06, "loss": 0.3569, "step": 2740 }, { "grad_norm": 0.4214962353320561, "learning_rate": 1.9016263748728114e-06, "loss": 0.3823, "step": 2750 }, { "grad_norm": 0.3998753698402858, "learning_rate": 1.7539954433864858e-06, "loss": 0.3817, "step": 2760 }, { "grad_norm": 0.30260742146182706, "learning_rate": 1.6122267976168781e-06, "loss": 0.3405, "step": 2770 }, { "grad_norm": 0.41280172706374546, "learning_rate": 1.4763376636185599e-06, "loss": 0.3663, "step": 2780 }, { "grad_norm": 0.33604241347602076, "learning_rate": 1.3463445530371488e-06, "loss": 0.3945, "step": 2790 }, { "grad_norm": 0.34043915962458693, "learning_rate": 1.222263261102985e-06, "loss": 0.3545, "step": 2800 }, { "grad_norm": 0.3260450284827044, "learning_rate": 1.1041088647119114e-06, "loss": 0.3659, "step": 2810 }, { "grad_norm": 0.37612371215908813, "learning_rate": 9.918957205933e-07, "loss": 0.368, "step": 2820 }, { "grad_norm": 0.4052660270258271, "learning_rate": 8.856374635655695e-07, "loss": 0.3665, "step": 2830 }, { "grad_norm": 0.35735268414366195, "learning_rate": 7.853470048794664e-07, "loss": 0.38, "step": 2840 }, { "grad_norm": 0.3059094600773396, "learning_rate": 6.910365306492416e-07, "loss": 0.3523, "step": 2850 }, { "grad_norm": 0.3608877177140587, "learning_rate": 6.027175003719354e-07, "loss": 0.389, "step": 2860 }, { "grad_norm": 0.2960936893278895, "learning_rate": 5.204006455349297e-07, "loss": 0.3976, "step": 2870 }, { "grad_norm": 0.4938512919991574, "learning_rate": 4.440959683120194e-07, "loss": 0.3366, "step": 2880 }, { "grad_norm": 0.36693392690964893, "learning_rate": 3.738127403480507e-07, "loss": 0.3627, "step": 2890 }, { "grad_norm": 0.3496669345978209, "learning_rate": 3.095595016323394e-07, "loss": 0.3688, "step": 2900 }, { "grad_norm": 0.304813232856759, "learning_rate": 2.51344059460995e-07, "loss": 0.3516, "step": 2910 }, { "grad_norm": 0.30683688680645477, "learning_rate": 1.9917348748826335e-07, "loss": 0.3639, "step": 2920 }, { "grad_norm": 0.3399231957790038, "learning_rate": 1.5305412486702474e-07, "loss": 0.3568, "step": 2930 }, { "grad_norm": 0.34643761387196265, "learning_rate": 1.1299157547854377e-07, "loss": 0.4083, "step": 2940 }, { "grad_norm": 0.36684871754149795, "learning_rate": 7.899070725153613e-08, "loss": 0.3916, "step": 2950 }, { "grad_norm": 0.41975265461708267, "learning_rate": 5.105565157068615e-08, "loss": 0.3758, "step": 2960 }, { "grad_norm": 0.3348368685535469, "learning_rate": 2.9189802774631792e-08, "loss": 0.3653, "step": 2970 }, { "grad_norm": 0.30325360159302284, "learning_rate": 1.3395817743561134e-08, "loss": 0.3524, "step": 2980 }, { "grad_norm": 0.3154182116946632, "learning_rate": 3.6756155763373323e-09, "loss": 0.4118, "step": 2990 }, { "grad_norm": 0.3366239577877631, "learning_rate": 3.037735734623404e-11, "loss": 0.3914, "step": 3000 }, { "grad_norm": 0.5143874303556566, "learning_rate": 5.3881744132384104e-05, "loss": 0.3362, "step": 3010 }, { "grad_norm": 0.4674007939126855, "learning_rate": 5.360693999650303e-05, "loss": 0.4177, "step": 3020 }, { "grad_norm": 0.7171709420380833, "learning_rate": 5.3332026291583016e-05, "loss": 0.3883, "step": 3030 }, { "grad_norm": 0.45973444304672273, "learning_rate": 5.305701136875566e-05, "loss": 0.4045, "step": 3040 }, { "grad_norm": 0.5040774687272648, "learning_rate": 5.278190358222721e-05, "loss": 0.443, "step": 3050 }, { "grad_norm": 0.3764159664634708, "learning_rate": 5.25067112890249e-05, "loss": 0.3565, "step": 3060 }, { "grad_norm": 0.6533520402778248, "learning_rate": 5.2231442848743064e-05, "loss": 0.3955, "step": 3070 }, { "grad_norm": 0.42360670746504026, "learning_rate": 5.1956106623289145e-05, "loss": 0.4198, "step": 3080 }, { "grad_norm": 0.4739954968549972, "learning_rate": 5.168071097662972e-05, "loss": 0.379, "step": 3090 }, { "grad_norm": 0.5101573895516204, "learning_rate": 5.1405264274536445e-05, "loss": 0.3723, "step": 3100 }, { "grad_norm": 0.6555709240817196, "learning_rate": 5.112977488433188e-05, "loss": 0.4002, "step": 3110 }, { "grad_norm": 0.5943956482997271, "learning_rate": 5.085425117463533e-05, "loss": 0.383, "step": 3120 }, { "grad_norm": 0.39432403727383697, "learning_rate": 5.057870151510864e-05, "loss": 0.3906, "step": 3130 }, { "grad_norm": 0.6118709178672301, "learning_rate": 5.030313427620197e-05, "loss": 0.3991, "step": 3140 }, { "grad_norm": 0.5536795815154062, "learning_rate": 5.0027557828899426e-05, "loss": 0.3482, "step": 3150 }, { "grad_norm": 0.460466103209189, "learning_rate": 4.975198054446492e-05, "loss": 0.3759, "step": 3160 }, { "grad_norm": 0.3703685879233111, "learning_rate": 4.947641079418773e-05, "loss": 0.3848, "step": 3170 }, { "grad_norm": 0.45675902789577927, "learning_rate": 4.920085694912828e-05, "loss": 0.3848, "step": 3180 }, { "grad_norm": 0.5469602044639579, "learning_rate": 4.892532737986387e-05, "loss": 0.3694, "step": 3190 }, { "grad_norm": 0.4185600612692819, "learning_rate": 4.864983045623434e-05, "loss": 0.3834, "step": 3200 }, { "grad_norm": 0.3759680561327521, "learning_rate": 4.837437454708784e-05, "loss": 0.3396, "step": 3210 }, { "grad_norm": 0.49222346251892585, "learning_rate": 4.809896802002662e-05, "loss": 0.3771, "step": 3220 }, { "grad_norm": 0.4647380842531198, "learning_rate": 4.7823619241152854e-05, "loss": 0.3096, "step": 3230 }, { "grad_norm": 0.6713007994816915, "learning_rate": 4.754833657481445e-05, "loss": 0.3652, "step": 3240 }, { "grad_norm": 0.5264176439555625, "learning_rate": 4.7273128383351015e-05, "loss": 0.3885, "step": 3250 }, { "grad_norm": 0.6344433404067391, "learning_rate": 4.699800302683981e-05, "loss": 0.3754, "step": 3260 }, { "grad_norm": 0.5916087333719502, "learning_rate": 4.6722968862841806e-05, "loss": 0.3974, "step": 3270 }, { "grad_norm": 0.494239468982641, "learning_rate": 4.6448034246147754e-05, "loss": 0.3888, "step": 3280 }, { "grad_norm": 0.6469428674601687, "learning_rate": 4.6173207528524476e-05, "loss": 0.4113, "step": 3290 }, { "grad_norm": 0.6253938042283882, "learning_rate": 4.58984970584611e-05, "loss": 0.4079, "step": 3300 }, { "grad_norm": 0.525208794080512, "learning_rate": 4.562391118091544e-05, "loss": 0.3571, "step": 3310 }, { "grad_norm": 0.42238526751477135, "learning_rate": 4.534945823706056e-05, "loss": 0.3548, "step": 3320 }, { "grad_norm": 0.5013962403955494, "learning_rate": 4.507514656403137e-05, "loss": 0.3852, "step": 3330 }, { "grad_norm": 0.3698406365992168, "learning_rate": 4.480098449467132e-05, "loss": 0.3587, "step": 3340 }, { "grad_norm": 0.3808247937107974, "learning_rate": 4.452698035727929e-05, "loss": 0.3714, "step": 3350 }, { "grad_norm": 0.6485942639847168, "learning_rate": 4.425314247535668e-05, "loss": 0.3881, "step": 3360 }, { "grad_norm": 0.6782586443120437, "learning_rate": 4.3979479167354477e-05, "loss": 0.4058, "step": 3370 }, { "grad_norm": 0.3656533707359406, "learning_rate": 4.370599874642055e-05, "loss": 0.3371, "step": 3380 }, { "grad_norm": 0.41789178845928576, "learning_rate": 4.3432709520147205e-05, "loss": 0.3716, "step": 3390 }, { "grad_norm": 0.4465846686007457, "learning_rate": 4.315961979031875e-05, "loss": 0.3983, "step": 3400 }, { "grad_norm": 0.5491118901615767, "learning_rate": 4.2886737852659325e-05, "loss": 0.3622, "step": 3410 }, { "grad_norm": 0.45990660285198115, "learning_rate": 4.261407199658093e-05, "loss": 0.4363, "step": 3420 }, { "grad_norm": 0.4718203060163525, "learning_rate": 4.234163050493158e-05, "loss": 0.3624, "step": 3430 }, { "grad_norm": 0.44129551705581044, "learning_rate": 4.2069421653743706e-05, "loss": 0.3936, "step": 3440 }, { "grad_norm": 0.43283515174158754, "learning_rate": 4.179745371198276e-05, "loss": 0.3983, "step": 3450 }, { "grad_norm": 0.5561710103656611, "learning_rate": 4.1525734941296026e-05, "loss": 0.3796, "step": 3460 }, { "grad_norm": 0.5317939966652737, "learning_rate": 4.125427359576162e-05, "loss": 0.3884, "step": 3470 }, { "grad_norm": 0.7468908997659971, "learning_rate": 4.0983077921637815e-05, "loss": 0.3926, "step": 3480 }, { "grad_norm": 0.4877080982083682, "learning_rate": 4.07121561571125e-05, "loss": 0.4163, "step": 3490 }, { "grad_norm": 0.43661417443010475, "learning_rate": 4.044151653205292e-05, "loss": 0.3977, "step": 3500 }, { "grad_norm": 0.32641244300734795, "learning_rate": 4.0171167267755696e-05, "loss": 0.3686, "step": 3510 }, { "grad_norm": 0.4691998283323796, "learning_rate": 3.9901116576697083e-05, "loss": 0.4179, "step": 3520 }, { "grad_norm": 0.4189424670484654, "learning_rate": 3.963137266228349e-05, "loss": 0.3731, "step": 3530 }, { "grad_norm": 0.36083513406022927, "learning_rate": 3.93619437186023e-05, "loss": 0.412, "step": 3540 }, { "grad_norm": 0.44105336257977834, "learning_rate": 3.9092837930172884e-05, "loss": 0.3501, "step": 3550 }, { "grad_norm": 0.40186055201933096, "learning_rate": 3.8824063471698105e-05, "loss": 0.3893, "step": 3560 }, { "grad_norm": 0.5882387850262908, "learning_rate": 3.855562850781589e-05, "loss": 0.3629, "step": 3570 }, { "grad_norm": 0.4104795666247558, "learning_rate": 3.828754119285123e-05, "loss": 0.3957, "step": 3580 }, { "grad_norm": 0.5091122331514772, "learning_rate": 3.801980967056851e-05, "loss": 0.3921, "step": 3590 }, { "grad_norm": 0.3833415271618265, "learning_rate": 3.77524420739241e-05, "loss": 0.3967, "step": 3600 }, { "grad_norm": 0.46215022462093935, "learning_rate": 3.748544652481927e-05, "loss": 0.3806, "step": 3610 }, { "grad_norm": 0.5117243572055811, "learning_rate": 3.721883113385353e-05, "loss": 0.3901, "step": 3620 }, { "grad_norm": 0.48577684836785, "learning_rate": 3.695260400007819e-05, "loss": 0.3984, "step": 3630 }, { "grad_norm": 0.3193196258027079, "learning_rate": 3.6686773210750385e-05, "loss": 0.3396, "step": 3640 }, { "grad_norm": 0.4509513444569201, "learning_rate": 3.642134684108737e-05, "loss": 0.4061, "step": 3650 }, { "grad_norm": 0.34044665741164415, "learning_rate": 3.615633295402123e-05, "loss": 0.3773, "step": 3660 }, { "grad_norm": 0.43146068730612647, "learning_rate": 3.5891739599953945e-05, "loss": 0.4254, "step": 3670 }, { "grad_norm": 0.39992483472809565, "learning_rate": 3.5627574816512846e-05, "loss": 0.3465, "step": 3680 }, { "grad_norm": 0.43300126460400745, "learning_rate": 3.536384662830648e-05, "loss": 0.3776, "step": 3690 }, { "grad_norm": 0.44862227988218184, "learning_rate": 3.5100563046680764e-05, "loss": 0.3635, "step": 3700 }, { "grad_norm": 0.37149482198956685, "learning_rate": 3.483773206947572e-05, "loss": 0.3899, "step": 3710 }, { "grad_norm": 0.37198390267680626, "learning_rate": 3.457536168078247e-05, "loss": 0.4326, "step": 3720 }, { "grad_norm": 0.4023039486823056, "learning_rate": 3.431345985070067e-05, "loss": 0.3562, "step": 3730 }, { "grad_norm": 0.4817760613553965, "learning_rate": 3.40520345350965e-05, "loss": 0.378, "step": 3740 }, { "grad_norm": 0.45376242645698844, "learning_rate": 3.379109367536089e-05, "loss": 0.3548, "step": 3750 }, { "grad_norm": 0.49078676361589063, "learning_rate": 3.3530645198168295e-05, "loss": 0.37, "step": 3760 }, { "grad_norm": 0.5900174881846914, "learning_rate": 3.327069701523595e-05, "loss": 0.3345, "step": 3770 }, { "grad_norm": 0.3887737218890107, "learning_rate": 3.301125702308353e-05, "loss": 0.3251, "step": 3780 }, { "grad_norm": 0.4937441373669907, "learning_rate": 3.275233310279321e-05, "loss": 0.3684, "step": 3790 }, { "grad_norm": 0.42018896982144077, "learning_rate": 3.249393311977037e-05, "loss": 0.3771, "step": 3800 }, { "grad_norm": 0.41266379733367037, "learning_rate": 3.223606492350451e-05, "loss": 0.3639, "step": 3810 }, { "grad_norm": 0.4547409462987729, "learning_rate": 3.197873634733096e-05, "loss": 0.4112, "step": 3820 }, { "grad_norm": 0.5210825343742327, "learning_rate": 3.172195520819285e-05, "loss": 0.3803, "step": 3830 }, { "grad_norm": 0.4711377418002084, "learning_rate": 3.146572930640362e-05, "loss": 0.4154, "step": 3840 }, { "grad_norm": 0.5021643079461422, "learning_rate": 3.121006642541014e-05, "loss": 0.3955, "step": 3850 }, { "grad_norm": 0.43487782667880515, "learning_rate": 3.095497433155626e-05, "loss": 0.3743, "step": 3860 }, { "grad_norm": 0.43497499900459846, "learning_rate": 3.070046077384682e-05, "loss": 0.3957, "step": 3870 }, { "grad_norm": 0.4947138161549373, "learning_rate": 3.0446533483712304e-05, "loss": 0.4189, "step": 3880 }, { "grad_norm": 0.7040858451907958, "learning_rate": 3.0193200174774038e-05, "loss": 0.394, "step": 3890 }, { "grad_norm": 0.4518092804753387, "learning_rate": 2.994046854260974e-05, "loss": 0.3732, "step": 3900 }, { "grad_norm": 0.45475623976287877, "learning_rate": 2.9688346264519866e-05, "loss": 0.3569, "step": 3910 }, { "grad_norm": 0.45117406514426267, "learning_rate": 2.943684099929436e-05, "loss": 0.3495, "step": 3920 }, { "grad_norm": 0.4517394359453835, "learning_rate": 2.918596038697995e-05, "loss": 0.4045, "step": 3930 }, { "grad_norm": 0.41559185678840294, "learning_rate": 2.8935712048648112e-05, "loss": 0.4121, "step": 3940 }, { "grad_norm": 0.43365315169077545, "learning_rate": 2.8686103586163626e-05, "loss": 0.3463, "step": 3950 }, { "grad_norm": 0.5673533560393954, "learning_rate": 2.843714258195346e-05, "loss": 0.3932, "step": 3960 }, { "grad_norm": 0.39559631784640764, "learning_rate": 2.8188836598776662e-05, "loss": 0.4048, "step": 3970 }, { "grad_norm": 0.4093772161871664, "learning_rate": 2.7941193179494484e-05, "loss": 0.3665, "step": 3980 }, { "grad_norm": 0.385240584146908, "learning_rate": 2.7694219846841262e-05, "loss": 0.337, "step": 3990 }, { "grad_norm": 0.48260840694284063, "learning_rate": 2.7447924103195976e-05, "loss": 0.3686, "step": 4000 }, { "grad_norm": 0.36261894275279655, "learning_rate": 2.7202313430354253e-05, "loss": 0.3915, "step": 4010 }, { "grad_norm": 0.49100728717723635, "learning_rate": 2.695739528930111e-05, "loss": 0.3486, "step": 4020 }, { "grad_norm": 0.39558488469145986, "learning_rate": 2.67131771199844e-05, "loss": 0.3956, "step": 4030 }, { "grad_norm": 0.4188481240607268, "learning_rate": 2.6469666341088677e-05, "loss": 0.3986, "step": 4040 }, { "grad_norm": 0.3471246868708886, "learning_rate": 2.6226870349809885e-05, "loss": 0.3815, "step": 4050 }, { "grad_norm": 0.34856173431708565, "learning_rate": 2.5984796521630737e-05, "loss": 0.3796, "step": 4060 }, { "grad_norm": 0.39480581889190886, "learning_rate": 2.574345221009653e-05, "loss": 0.3968, "step": 4070 }, { "grad_norm": 0.4918501505084642, "learning_rate": 2.5502844746591804e-05, "loss": 0.4009, "step": 4080 }, { "grad_norm": 0.4049005933632862, "learning_rate": 2.526298144011775e-05, "loss": 0.382, "step": 4090 }, { "grad_norm": 0.46508334505296406, "learning_rate": 2.5023869577070013e-05, "loss": 0.3327, "step": 4100 }, { "grad_norm": 0.38545017523804675, "learning_rate": 2.478551642101743e-05, "loss": 0.3, "step": 4110 }, { "grad_norm": 0.44232287597375997, "learning_rate": 2.4547929212481435e-05, "loss": 0.3625, "step": 4120 }, { "grad_norm": 0.5458460952712424, "learning_rate": 2.4311115168716013e-05, "loss": 0.3631, "step": 4130 }, { "grad_norm": 0.41687517358772874, "learning_rate": 2.4075081483488494e-05, "loss": 0.3675, "step": 4140 }, { "grad_norm": 0.4365776907936673, "learning_rate": 2.3839835326861104e-05, "loss": 0.388, "step": 4150 }, { "grad_norm": 0.3780405675566688, "learning_rate": 2.3605383844972966e-05, "loss": 0.3557, "step": 4160 }, { "grad_norm": 0.435186616553367, "learning_rate": 2.3371734159823284e-05, "loss": 0.3275, "step": 4170 }, { "grad_norm": 0.4511249740892401, "learning_rate": 2.3138893369054766e-05, "loss": 0.3479, "step": 4180 }, { "grad_norm": 0.5026422632377036, "learning_rate": 2.2906868545738102e-05, "loss": 0.3224, "step": 4190 }, { "grad_norm": 0.4383924058368146, "learning_rate": 2.2675666738157186e-05, "loss": 0.3321, "step": 4200 }, { "grad_norm": 0.34546491381997313, "learning_rate": 2.2445294969594844e-05, "loss": 0.3368, "step": 4210 }, { "grad_norm": 0.3926687610581663, "learning_rate": 2.22157602381196e-05, "loss": 0.3699, "step": 4220 }, { "grad_norm": 0.5166171493843229, "learning_rate": 2.1987069516373098e-05, "loss": 0.3679, "step": 4230 }, { "grad_norm": 0.3895014160129566, "learning_rate": 2.1759229751358217e-05, "loss": 0.3285, "step": 4240 }, { "grad_norm": 0.34748283754160414, "learning_rate": 2.1532247864228084e-05, "loss": 0.3849, "step": 4250 }, { "grad_norm": 0.39578946739693177, "learning_rate": 2.1306130750075865e-05, "loss": 0.3386, "step": 4260 }, { "grad_norm": 0.36606814811872534, "learning_rate": 2.1080885277725236e-05, "loss": 0.4137, "step": 4270 }, { "grad_norm": 0.5258718619627164, "learning_rate": 2.085651828952175e-05, "loss": 0.3803, "step": 4280 }, { "grad_norm": 0.508771131716736, "learning_rate": 2.063303660112506e-05, "loss": 0.4293, "step": 4290 }, { "grad_norm": 0.36490183177613783, "learning_rate": 2.0410447001301753e-05, "loss": 0.3311, "step": 4300 }, { "grad_norm": 0.356092088838117, "learning_rate": 2.0188756251719203e-05, "loss": 0.3559, "step": 4310 }, { "grad_norm": 0.40860174012108336, "learning_rate": 1.9967971086740195e-05, "loss": 0.3908, "step": 4320 }, { "grad_norm": 0.33650175165004664, "learning_rate": 1.974809821321827e-05, "loss": 0.3735, "step": 4330 }, { "grad_norm": 0.3298312119775653, "learning_rate": 1.9529144310294023e-05, "loss": 0.3343, "step": 4340 }, { "grad_norm": 0.3362327600635854, "learning_rate": 1.9311116029192278e-05, "loss": 0.3484, "step": 4350 }, { "grad_norm": 0.3368025744869337, "learning_rate": 1.909401999301993e-05, "loss": 0.4065, "step": 4360 }, { "grad_norm": 0.40267175508882713, "learning_rate": 1.887786279656482e-05, "loss": 0.3396, "step": 4370 }, { "grad_norm": 0.3385181748472623, "learning_rate": 1.8662651006095387e-05, "loss": 0.3668, "step": 4380 }, { "grad_norm": 0.36307858928952574, "learning_rate": 1.8448391159161204e-05, "loss": 0.4009, "step": 4390 }, { "grad_norm": 0.39149584492899037, "learning_rate": 1.8235089764394408e-05, "loss": 0.3401, "step": 4400 }, { "grad_norm": 0.4428007090159073, "learning_rate": 1.8022753301311935e-05, "loss": 0.363, "step": 4410 }, { "grad_norm": 0.3264813182989556, "learning_rate": 1.7811388220118707e-05, "loss": 0.3599, "step": 4420 }, { "grad_norm": 0.359197847816617, "learning_rate": 1.7601000941511757e-05, "loss": 0.3373, "step": 4430 }, { "grad_norm": 0.3517267699949707, "learning_rate": 1.7391597856485083e-05, "loss": 0.3685, "step": 4440 }, { "grad_norm": 0.3708631446456292, "learning_rate": 1.7183185326135543e-05, "loss": 0.3628, "step": 4450 }, { "grad_norm": 0.49214581884103564, "learning_rate": 1.6975769681469705e-05, "loss": 0.3606, "step": 4460 }, { "grad_norm": 0.48989838692934284, "learning_rate": 1.676935722321139e-05, "loss": 0.3694, "step": 4470 }, { "grad_norm": 0.47777032218701065, "learning_rate": 1.6563954221610355e-05, "loss": 0.3283, "step": 4480 }, { "grad_norm": 0.3645949828113383, "learning_rate": 1.6359566916251845e-05, "loss": 0.3564, "step": 4490 }, { "grad_norm": 0.32656192227041686, "learning_rate": 1.615620151586697e-05, "loss": 0.3952, "step": 4500 }, { "grad_norm": 0.35322386889228297, "learning_rate": 1.5953864198144135e-05, "loss": 0.3372, "step": 4510 }, { "grad_norm": 0.3245450616391652, "learning_rate": 1.5752561109541447e-05, "loss": 0.3843, "step": 4520 }, { "grad_norm": 0.36226184996174854, "learning_rate": 1.5552298365099882e-05, "loss": 0.3653, "step": 4530 }, { "grad_norm": 0.38805613233424224, "learning_rate": 1.5353082048257596e-05, "loss": 0.3466, "step": 4540 }, { "grad_norm": 0.3773442249468262, "learning_rate": 1.5154918210665148e-05, "loss": 0.3558, "step": 4550 }, { "grad_norm": 0.41646002681021455, "learning_rate": 1.4957812872001614e-05, "loss": 0.399, "step": 4560 }, { "grad_norm": 0.38984164850034464, "learning_rate": 1.4761772019791748e-05, "loss": 0.3562, "step": 4570 }, { "grad_norm": 0.4588182870618412, "learning_rate": 1.4566801609224096e-05, "loss": 0.3722, "step": 4580 }, { "grad_norm": 0.3373481556774297, "learning_rate": 1.4372907562970079e-05, "loss": 0.3423, "step": 4590 }, { "grad_norm": 0.3791567648682786, "learning_rate": 1.4180095771004154e-05, "loss": 0.3397, "step": 4600 }, { "grad_norm": 0.435394416953223, "learning_rate": 1.3988372090424773e-05, "loss": 0.3663, "step": 4610 }, { "grad_norm": 0.34354563117141096, "learning_rate": 1.3797742345276521e-05, "loss": 0.366, "step": 4620 }, { "grad_norm": 0.32899128590382737, "learning_rate": 1.3608212326373249e-05, "loss": 0.3091, "step": 4630 }, { "grad_norm": 0.3684485910876113, "learning_rate": 1.3419787791122062e-05, "loss": 0.3652, "step": 4640 }, { "grad_norm": 0.35207117634780916, "learning_rate": 1.323247446334847e-05, "loss": 0.371, "step": 4650 }, { "grad_norm": 0.40069621378581843, "learning_rate": 1.3046278033122577e-05, "loss": 0.4111, "step": 4660 }, { "grad_norm": 0.48789773372246104, "learning_rate": 1.286120415658611e-05, "loss": 0.3826, "step": 4670 }, { "grad_norm": 0.3216996690414691, "learning_rate": 1.2677258455780683e-05, "loss": 0.3751, "step": 4680 }, { "grad_norm": 0.38166107867303917, "learning_rate": 1.2494446518477022e-05, "loss": 0.3729, "step": 4690 }, { "grad_norm": 0.4333441500714642, "learning_rate": 1.2312773898005175e-05, "loss": 0.3565, "step": 4700 }, { "grad_norm": 0.3191373082481251, "learning_rate": 1.2132246113085822e-05, "loss": 0.3549, "step": 4710 }, { "grad_norm": 0.31386559277077325, "learning_rate": 1.1952868647662696e-05, "loss": 0.3542, "step": 4720 }, { "grad_norm": 0.3759549426396945, "learning_rate": 1.1774646950735913e-05, "loss": 0.3433, "step": 4730 }, { "grad_norm": 0.39748591537658645, "learning_rate": 1.1597586436196473e-05, "loss": 0.3563, "step": 4740 }, { "grad_norm": 0.348483457082518, "learning_rate": 1.1421692482661856e-05, "loss": 0.3326, "step": 4750 }, { "grad_norm": 0.3655085400138021, "learning_rate": 1.124697043331256e-05, "loss": 0.3206, "step": 4760 }, { "grad_norm": 0.4372560554153719, "learning_rate": 1.107342559572977e-05, "loss": 0.3889, "step": 4770 }, { "grad_norm": 0.438670343993391, "learning_rate": 1.090106324173426e-05, "loss": 0.3876, "step": 4780 }, { "grad_norm": 0.36420528604847285, "learning_rate": 1.0729888607226113e-05, "loss": 0.3666, "step": 4790 }, { "grad_norm": 0.3384933721304882, "learning_rate": 1.0559906892025745e-05, "loss": 0.3491, "step": 4800 }, { "grad_norm": 0.365808240238384, "learning_rate": 1.0391123259715906e-05, "loss": 0.3696, "step": 4810 }, { "grad_norm": 0.3916150492726777, "learning_rate": 1.0223542837484839e-05, "loss": 0.3338, "step": 4820 }, { "grad_norm": 0.3494818261101757, "learning_rate": 1.0057170715970559e-05, "loss": 0.3585, "step": 4830 }, { "grad_norm": 0.37285163793134796, "learning_rate": 9.892011949106172e-06, "loss": 0.3573, "step": 4840 }, { "grad_norm": 0.3096720182474047, "learning_rate": 9.728071553966339e-06, "loss": 0.3845, "step": 4850 }, { "grad_norm": 0.43155974147200327, "learning_rate": 9.56535451061496e-06, "loss": 0.3605, "step": 4860 }, { "grad_norm": 0.4060103500775724, "learning_rate": 9.403865761953779e-06, "loss": 0.3972, "step": 4870 }, { "grad_norm": 0.4626982068221135, "learning_rate": 9.243610213572285e-06, "loss": 0.3397, "step": 4880 }, { "grad_norm": 0.3365532447807934, "learning_rate": 9.084592733598735e-06, "loss": 0.3943, "step": 4890 }, { "grad_norm": 0.33984059634176106, "learning_rate": 8.92681815255219e-06, "loss": 0.3454, "step": 4900 }, { "grad_norm": 0.4382192824138389, "learning_rate": 8.770291263195819e-06, "loss": 0.3592, "step": 4910 }, { "grad_norm": 0.34977392227340753, "learning_rate": 8.615016820391342e-06, "loss": 0.3849, "step": 4920 }, { "grad_norm": 0.3394468188167302, "learning_rate": 8.460999540954517e-06, "loss": 0.3436, "step": 4930 }, { "grad_norm": 0.3724872846512477, "learning_rate": 8.308244103511909e-06, "loss": 0.3492, "step": 4940 }, { "grad_norm": 0.43443174714279786, "learning_rate": 8.156755148358764e-06, "loss": 0.3363, "step": 4950 }, { "grad_norm": 0.3335829253112259, "learning_rate": 8.00653727731801e-06, "loss": 0.3505, "step": 4960 }, { "grad_norm": 0.3784039753389459, "learning_rate": 7.857595053600513e-06, "loss": 0.3625, "step": 4970 }, { "grad_norm": 0.3915498012185126, "learning_rate": 7.709933001666431e-06, "loss": 0.3443, "step": 4980 }, { "grad_norm": 0.37379972175603404, "learning_rate": 7.56355560708778e-06, "loss": 0.3641, "step": 4990 }, { "grad_norm": 0.3105501746171036, "learning_rate": 7.418467316412158e-06, "loss": 0.4035, "step": 5000 }, { "grad_norm": 0.3481860284683503, "learning_rate": 7.2746725370277435e-06, "loss": 0.3652, "step": 5010 }, { "grad_norm": 0.29115636720109683, "learning_rate": 7.132175637029293e-06, "loss": 0.3576, "step": 5020 }, { "grad_norm": 0.40402029323139244, "learning_rate": 6.9909809450855345e-06, "loss": 0.3637, "step": 5030 }, { "grad_norm": 0.3391913530958596, "learning_rate": 6.851092750307686e-06, "loss": 0.3487, "step": 5040 }, { "grad_norm": 0.46122727952875414, "learning_rate": 6.712515302119077e-06, "loss": 0.3694, "step": 5050 }, { "grad_norm": 0.4072326458760014, "learning_rate": 6.575252810126143e-06, "loss": 0.3052, "step": 5060 }, { "grad_norm": 0.3164805531864042, "learning_rate": 6.439309443990532e-06, "loss": 0.3682, "step": 5070 }, { "grad_norm": 0.37507028292664224, "learning_rate": 6.304689333302416e-06, "loss": 0.3673, "step": 5080 }, { "grad_norm": 0.3603932084689202, "learning_rate": 6.171396567455051e-06, "loss": 0.3662, "step": 5090 }, { "grad_norm": 0.36177731320679635, "learning_rate": 6.039435195520604e-06, "loss": 0.3955, "step": 5100 }, { "grad_norm": 0.4536102733607865, "learning_rate": 5.908809226127054e-06, "loss": 0.377, "step": 5110 }, { "grad_norm": 0.3301961510598887, "learning_rate": 5.779522627336537e-06, "loss": 0.3297, "step": 5120 }, { "grad_norm": 0.32402632288835104, "learning_rate": 5.651579326524709e-06, "loss": 0.3577, "step": 5130 }, { "grad_norm": 0.28479213439385515, "learning_rate": 5.524983210261481e-06, "loss": 0.3529, "step": 5140 }, { "grad_norm": 0.37610752014960935, "learning_rate": 5.399738124192988e-06, "loss": 0.3694, "step": 5150 }, { "grad_norm": 0.33962268259618794, "learning_rate": 5.2758478729247164e-06, "loss": 0.3535, "step": 5160 }, { "grad_norm": 0.3679533921287712, "learning_rate": 5.153316219905946e-06, "loss": 0.3688, "step": 5170 }, { "grad_norm": 0.35878387760551544, "learning_rate": 5.032146887315448e-06, "loss": 0.376, "step": 5180 }, { "grad_norm": 0.3218898581799965, "learning_rate": 4.91234355594839e-06, "loss": 0.3536, "step": 5190 }, { "grad_norm": 0.45312773772341053, "learning_rate": 4.7939098651045235e-06, "loss": 0.3784, "step": 5200 }, { "grad_norm": 0.3895016911064277, "learning_rate": 4.67684941247768e-06, "loss": 0.355, "step": 5210 }, { "grad_norm": 0.5091378966060223, "learning_rate": 4.5611657540464036e-06, "loss": 0.3214, "step": 5220 }, { "grad_norm": 0.34703203266708005, "learning_rate": 4.446862403965984e-06, "loss": 0.3408, "step": 5230 }, { "grad_norm": 0.3542007263022114, "learning_rate": 4.333942834461702e-06, "loss": 0.3505, "step": 5240 }, { "grad_norm": 0.3768976304847943, "learning_rate": 4.222410475723326e-06, "loss": 0.3186, "step": 5250 }, { "grad_norm": 0.3024482796617722, "learning_rate": 4.112268715800943e-06, "loss": 0.344, "step": 5260 }, { "grad_norm": 0.35743677532607987, "learning_rate": 4.003520900502028e-06, "loss": 0.3404, "step": 5270 }, { "grad_norm": 0.3723222892876854, "learning_rate": 3.8961703332898e-06, "loss": 0.3383, "step": 5280 }, { "grad_norm": 0.3697915508112912, "learning_rate": 3.790220275182854e-06, "loss": 0.3691, "step": 5290 }, { "grad_norm": 0.31738181887749806, "learning_rate": 3.685673944656176e-06, "loss": 0.4035, "step": 5300 }, { "grad_norm": 0.2943422722521425, "learning_rate": 3.582534517543268e-06, "loss": 0.3433, "step": 5310 }, { "grad_norm": 0.3509425914865178, "learning_rate": 3.4808051269397512e-06, "loss": 0.3659, "step": 5320 }, { "grad_norm": 0.3807327618582179, "learning_rate": 3.380488863108183e-06, "loss": 0.373, "step": 5330 }, { "grad_norm": 0.31062717068158835, "learning_rate": 3.2815887733841365e-06, "loss": 0.351, "step": 5340 }, { "grad_norm": 0.3053783634994152, "learning_rate": 3.1841078620836683e-06, "loss": 0.3479, "step": 5350 }, { "grad_norm": 0.2919442450620151, "learning_rate": 3.0880490904120874e-06, "loss": 0.342, "step": 5360 }, { "grad_norm": 0.2978548390645225, "learning_rate": 2.9934153763739205e-06, "loss": 0.3352, "step": 5370 }, { "grad_norm": 0.33859623369377123, "learning_rate": 2.9002095946843277e-06, "loss": 0.3909, "step": 5380 }, { "grad_norm": 0.31107731436828995, "learning_rate": 2.8084345766817676e-06, "loss": 0.3928, "step": 5390 }, { "grad_norm": 0.3432509302646241, "learning_rate": 2.718093110241976e-06, "loss": 0.3678, "step": 5400 }, { "grad_norm": 0.3543903754974993, "learning_rate": 2.6291879396933004e-06, "loss": 0.3825, "step": 5410 }, { "grad_norm": 0.32605477441982916, "learning_rate": 2.541721765733318e-06, "loss": 0.3215, "step": 5420 }, { "grad_norm": 0.3002966271692467, "learning_rate": 2.455697245346783e-06, "loss": 0.3701, "step": 5430 }, { "grad_norm": 0.40398934143830967, "learning_rate": 2.371116991724953e-06, "loss": 0.3734, "step": 5440 }, { "grad_norm": 0.32647553234997995, "learning_rate": 2.2879835741861586e-06, "loss": 0.4022, "step": 5450 }, { "grad_norm": 0.3903615669260566, "learning_rate": 2.206299518097804e-06, "loss": 0.3677, "step": 5460 }, { "grad_norm": 0.3977243186175328, "learning_rate": 2.1260673047996227e-06, "loss": 0.3896, "step": 5470 }, { "grad_norm": 0.41479611868422156, "learning_rate": 2.047289371528299e-06, "loss": 0.3679, "step": 5480 }, { "grad_norm": 0.37041706581226347, "learning_rate": 1.96996811134344e-06, "loss": 0.3548, "step": 5490 }, { "grad_norm": 0.2738330238327822, "learning_rate": 1.8941058730549132e-06, "loss": 0.3253, "step": 5500 }, { "grad_norm": 0.33052816964149007, "learning_rate": 1.8197049611514194e-06, "loss": 0.3491, "step": 5510 }, { "grad_norm": 0.37617440903981175, "learning_rate": 1.7467676357305561e-06, "loss": 0.3735, "step": 5520 }, { "grad_norm": 0.28372878402969465, "learning_rate": 1.6752961124301415e-06, "loss": 0.3601, "step": 5530 }, { "grad_norm": 0.3702028869048457, "learning_rate": 1.6052925623609049e-06, "loss": 0.3255, "step": 5540 }, { "grad_norm": 0.3212456561632761, "learning_rate": 1.5367591120405256e-06, "loss": 0.3454, "step": 5550 }, { "grad_norm": 0.2992497421050431, "learning_rate": 1.4696978433290653e-06, "loss": 0.3465, "step": 5560 }, { "grad_norm": 0.33701106030171296, "learning_rate": 1.4041107933656928e-06, "loss": 0.3707, "step": 5570 }, { "grad_norm": 0.34206484636823453, "learning_rate": 1.339999954506821e-06, "loss": 0.2979, "step": 5580 }, { "grad_norm": 0.3241506667680006, "learning_rate": 1.2773672742655784e-06, "loss": 0.3403, "step": 5590 }, { "grad_norm": 0.4037133229762182, "learning_rate": 1.2162146552526399e-06, "loss": 0.3817, "step": 5600 }, { "grad_norm": 0.41817268673865626, "learning_rate": 1.1565439551184664e-06, "loss": 0.3651, "step": 5610 }, { "grad_norm": 0.38251120644471465, "learning_rate": 1.0983569864968346e-06, "loss": 0.3435, "step": 5620 }, { "grad_norm": 0.31322071295704285, "learning_rate": 1.0416555169497688e-06, "loss": 0.2905, "step": 5630 }, { "grad_norm": 0.5351330756956421, "learning_rate": 9.864412689139123e-07, "loss": 0.3769, "step": 5640 }, { "grad_norm": 0.29768889708463475, "learning_rate": 9.327159196481138e-07, "loss": 0.3749, "step": 5650 }, { "grad_norm": 0.33935260113014143, "learning_rate": 8.804811011825398e-07, "loss": 0.3894, "step": 5660 }, { "grad_norm": 0.3064489467591253, "learning_rate": 8.297384002690866e-07, "loss": 0.2908, "step": 5670 }, { "grad_norm": 0.3197602967303336, "learning_rate": 7.804893583331696e-07, "loss": 0.3724, "step": 5680 }, { "grad_norm": 0.44650181997482613, "learning_rate": 7.32735471426893e-07, "loss": 0.3234, "step": 5690 }, { "grad_norm": 0.36393717650204266, "learning_rate": 6.864781901836259e-07, "loss": 0.3845, "step": 5700 }, { "grad_norm": 0.3167022066770035, "learning_rate": 6.417189197739093e-07, "loss": 0.3268, "step": 5710 }, { "grad_norm": 0.3193652147087475, "learning_rate": 5.984590198627849e-07, "loss": 0.3487, "step": 5720 }, { "grad_norm": 0.33636100173664407, "learning_rate": 5.566998045685112e-07, "loss": 0.3542, "step": 5730 }, { "grad_norm": 0.36880810723810353, "learning_rate": 5.164425424226016e-07, "loss": 0.3308, "step": 5740 }, { "grad_norm": 0.3366427666360023, "learning_rate": 4.776884563313266e-07, "loss": 0.3796, "step": 5750 }, { "grad_norm": 0.3066218991208089, "learning_rate": 4.404387235385443e-07, "loss": 0.3177, "step": 5760 }, { "grad_norm": 0.3651638708642201, "learning_rate": 4.0469447558995065e-07, "loss": 0.3688, "step": 5770 }, { "grad_norm": 0.30942630069477617, "learning_rate": 3.7045679829870175e-07, "loss": 0.3697, "step": 5780 }, { "grad_norm": 0.2926817937154156, "learning_rate": 3.377267317124233e-07, "loss": 0.3496, "step": 5790 }, { "grad_norm": 0.3654834776830413, "learning_rate": 3.0650527008162513e-07, "loss": 0.3685, "step": 5800 }, { "grad_norm": 0.34778705674297133, "learning_rate": 2.767933618295082e-07, "loss": 0.4183, "step": 5810 }, { "grad_norm": 0.27193714529120994, "learning_rate": 2.485919095231326e-07, "loss": 0.3779, "step": 5820 }, { "grad_norm": 0.38487354487689446, "learning_rate": 2.219017698460002e-07, "loss": 0.3735, "step": 5830 }, { "grad_norm": 0.3702509639912046, "learning_rate": 1.9672375357206452e-07, "loss": 0.3454, "step": 5840 }, { "grad_norm": 0.33237263382504373, "learning_rate": 1.73058625541056e-07, "loss": 0.3852, "step": 5850 }, { "grad_norm": 0.36032084554783694, "learning_rate": 1.5090710463527836e-07, "loss": 0.336, "step": 5860 }, { "grad_norm": 0.4438658397472768, "learning_rate": 1.3026986375776485e-07, "loss": 0.3345, "step": 5870 }, { "grad_norm": 0.31837174214818803, "learning_rate": 1.1114752981183917e-07, "loss": 0.3503, "step": 5880 }, { "grad_norm": 0.37272660552776066, "learning_rate": 9.354068368204739e-08, "loss": 0.3082, "step": 5890 }, { "grad_norm": 0.2843995705149283, "learning_rate": 7.744986021656076e-08, "loss": 0.3397, "step": 5900 }, { "grad_norm": 0.472538403896959, "learning_rate": 6.287554821087783e-08, "loss": 0.3828, "step": 5910 }, { "grad_norm": 0.40458913199122015, "learning_rate": 4.981819039300284e-08, "loss": 0.3554, "step": 5920 }, { "grad_norm": 0.31692593820544107, "learning_rate": 3.827818341000655e-08, "loss": 0.3569, "step": 5930 }, { "grad_norm": 0.39089391145247465, "learning_rate": 2.8255877815946963e-08, "loss": 0.346, "step": 5940 }, { "grad_norm": 0.3867325765403945, "learning_rate": 1.9751578061244504e-08, "loss": 0.3766, "step": 5950 }, { "grad_norm": 0.3910162069072075, "learning_rate": 1.2765542483417214e-08, "loss": 0.3372, "step": 5960 }, { "grad_norm": 0.4142721096411222, "learning_rate": 7.2979832992592365e-09, "loss": 0.3593, "step": 5970 }, { "grad_norm": 0.2857764204754828, "learning_rate": 3.349066598362649e-09, "loss": 0.3165, "step": 5980 }, { "grad_norm": 0.30031238033046825, "learning_rate": 9.189123380826114e-10, "loss": 0.3865, "step": 5990 }, { "grad_norm": 0.37948754176387356, "learning_rate": 7.594339912486703e-12, "loss": 0.3719, "step": 6000 } ], "logging_steps": 10, "max_steps": 6000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 500, "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": 16, "trial_name": null, "trial_params": null }