{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "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": 1.7000700235366821, "learning_rate": 3e-06, "loss": 1.1283, "step": 10 }, { "grad_norm": 0.8262935876846313, "learning_rate": 6.333333333333334e-06, "loss": 1.097, "step": 20 }, { "grad_norm": 0.7697754502296448, "learning_rate": 9.666666666666667e-06, "loss": 1.0692, "step": 30 }, { "grad_norm": 0.3612891733646393, "learning_rate": 1.3000000000000001e-05, "loss": 1.0448, "step": 40 }, { "grad_norm": 0.7433922290802002, "learning_rate": 1.6333333333333335e-05, "loss": 1.0697, "step": 50 }, { "grad_norm": 0.6279253959655762, "learning_rate": 1.9666666666666666e-05, "loss": 1.0638, "step": 60 }, { "grad_norm": 0.6754515767097473, "learning_rate": 2.3000000000000003e-05, "loss": 1.0625, "step": 70 }, { "grad_norm": 0.5152317881584167, "learning_rate": 2.633333333333333e-05, "loss": 1.068, "step": 80 }, { "grad_norm": 0.7765102386474609, "learning_rate": 2.9666666666666672e-05, "loss": 1.0458, "step": 90 }, { "grad_norm": 0.5475148558616638, "learning_rate": 3.3e-05, "loss": 1.0491, "step": 100 }, { "grad_norm": 0.9391897916793823, "learning_rate": 3.633333333333333e-05, "loss": 1.0308, "step": 110 }, { "grad_norm": 1.0357705354690552, "learning_rate": 3.966666666666667e-05, "loss": 0.9933, "step": 120 }, { "grad_norm": 0.7339250445365906, "learning_rate": 4.3e-05, "loss": 0.9393, "step": 130 }, { "grad_norm": 0.9334234595298767, "learning_rate": 4.633333333333333e-05, "loss": 0.9121, "step": 140 }, { "grad_norm": 4.375876426696777, "learning_rate": 4.966666666666667e-05, "loss": 0.85, "step": 150 }, { "grad_norm": 2.2090346813201904, "learning_rate": 5.300000000000001e-05, "loss": 0.7249, "step": 160 }, { "grad_norm": 1.9277912378311157, "learning_rate": 5.633333333333334e-05, "loss": 0.5533, "step": 170 }, { "grad_norm": 2.3054964542388916, "learning_rate": 5.966666666666667e-05, "loss": 0.4604, "step": 180 }, { "grad_norm": 1.3749431371688843, "learning_rate": 6.3e-05, "loss": 0.3421, "step": 190 }, { "grad_norm": 1.6281027793884277, "learning_rate": 6.633333333333334e-05, "loss": 0.2506, "step": 200 }, { "grad_norm": 1.410512089729309, "learning_rate": 6.966666666666668e-05, "loss": 0.189, "step": 210 }, { "grad_norm": 1.0307164192199707, "learning_rate": 7.3e-05, "loss": 0.1688, "step": 220 }, { "grad_norm": 1.0910820960998535, "learning_rate": 7.633333333333334e-05, "loss": 0.1543, "step": 230 }, { "grad_norm": 0.8725211024284363, "learning_rate": 7.966666666666666e-05, "loss": 0.1578, "step": 240 }, { "grad_norm": 1.1222732067108154, "learning_rate": 8.3e-05, "loss": 0.1432, "step": 250 }, { "grad_norm": 0.7612312436103821, "learning_rate": 8.633333333333334e-05, "loss": 0.13, "step": 260 }, { "grad_norm": 1.1903313398361206, "learning_rate": 8.966666666666666e-05, "loss": 0.148, "step": 270 }, { "grad_norm": 0.8089312314987183, "learning_rate": 9.300000000000001e-05, "loss": 0.1326, "step": 280 }, { "grad_norm": 0.9453487992286682, "learning_rate": 9.633333333333335e-05, "loss": 0.1467, "step": 290 }, { "grad_norm": 1.205849051475525, "learning_rate": 9.966666666666667e-05, "loss": 0.1273, "step": 300 }, { "grad_norm": 0.6407943964004517, "learning_rate": 9.999938485971279e-05, "loss": 0.1256, "step": 310 }, { "grad_norm": 0.696092963218689, "learning_rate": 9.999725846827562e-05, "loss": 0.113, "step": 320 }, { "grad_norm": 0.6457434892654419, "learning_rate": 9.999361329594254e-05, "loss": 0.1059, "step": 330 }, { "grad_norm": 0.769981324672699, "learning_rate": 9.998844945344405e-05, "loss": 0.102, "step": 340 }, { "grad_norm": 0.4546976387500763, "learning_rate": 9.99817670976436e-05, "loss": 0.0906, "step": 350 }, { "grad_norm": 0.7327650785446167, "learning_rate": 9.997356643153303e-05, "loss": 0.0935, "step": 360 }, { "grad_norm": 0.7533209919929504, "learning_rate": 9.996384770422629e-05, "loss": 0.1085, "step": 370 }, { "grad_norm": 0.6281214356422424, "learning_rate": 9.995261121095194e-05, "loss": 0.094, "step": 380 }, { "grad_norm": 0.45302775502204895, "learning_rate": 9.993985729304408e-05, "loss": 0.0973, "step": 390 }, { "grad_norm": 0.5847544074058533, "learning_rate": 9.992558633793212e-05, "loss": 0.0955, "step": 400 }, { "grad_norm": 0.5938292741775513, "learning_rate": 9.990979877912891e-05, "loss": 0.0927, "step": 410 }, { "grad_norm": 0.7282758355140686, "learning_rate": 9.989249509621759e-05, "loss": 0.0858, "step": 420 }, { "grad_norm": 0.6158447265625, "learning_rate": 9.987367581483705e-05, "loss": 0.0974, "step": 430 }, { "grad_norm": 0.4978296458721161, "learning_rate": 9.985334150666592e-05, "loss": 0.0946, "step": 440 }, { "grad_norm": 0.7747285962104797, "learning_rate": 9.983149278940526e-05, "loss": 0.0904, "step": 450 }, { "grad_norm": 0.5748892426490784, "learning_rate": 9.980813032675974e-05, "loss": 0.0938, "step": 460 }, { "grad_norm": 0.5600329637527466, "learning_rate": 9.978325482841753e-05, "loss": 0.0976, "step": 470 }, { "grad_norm": 0.4506227970123291, "learning_rate": 9.975686705002867e-05, "loss": 0.0875, "step": 480 }, { "grad_norm": 0.5217170715332031, "learning_rate": 9.972896779318219e-05, "loss": 0.085, "step": 490 }, { "grad_norm": 0.44347596168518066, "learning_rate": 9.969955790538175e-05, "loss": 0.0811, "step": 500 }, { "grad_norm": 0.7565658688545227, "learning_rate": 9.966863828001982e-05, "loss": 0.089, "step": 510 }, { "grad_norm": 0.4682891070842743, "learning_rate": 9.963620985635065e-05, "loss": 0.0801, "step": 520 }, { "grad_norm": 0.407950222492218, "learning_rate": 9.960227361946164e-05, "loss": 0.0855, "step": 530 }, { "grad_norm": 0.555381178855896, "learning_rate": 9.95668306002435e-05, "loss": 0.0769, "step": 540 }, { "grad_norm": 0.4560313820838928, "learning_rate": 9.952988187535886e-05, "loss": 0.0818, "step": 550 }, { "grad_norm": 0.46689373254776, "learning_rate": 9.949142856720961e-05, "loss": 0.0741, "step": 560 }, { "grad_norm": 0.5960952639579773, "learning_rate": 9.945147184390278e-05, "loss": 0.0744, "step": 570 }, { "grad_norm": 0.4881025552749634, "learning_rate": 9.941001291921512e-05, "loss": 0.0778, "step": 580 }, { "grad_norm": 0.44606727361679077, "learning_rate": 9.936705305255612e-05, "loss": 0.0805, "step": 590 }, { "grad_norm": 0.43473923206329346, "learning_rate": 9.932259354892984e-05, "loss": 0.0823, "step": 600 }, { "grad_norm": 0.39367181062698364, "learning_rate": 9.927663575889521e-05, "loss": 0.084, "step": 610 }, { "grad_norm": 0.47242826223373413, "learning_rate": 9.922918107852504e-05, "loss": 0.0729, "step": 620 }, { "grad_norm": 0.4946627616882324, "learning_rate": 9.918023094936363e-05, "loss": 0.0694, "step": 630 }, { "grad_norm": 0.4091627597808838, "learning_rate": 9.912978685838294e-05, "loss": 0.072, "step": 640 }, { "grad_norm": 0.2955147624015808, "learning_rate": 9.90778503379374e-05, "loss": 0.0704, "step": 650 }, { "grad_norm": 0.5645137429237366, "learning_rate": 9.902442296571743e-05, "loss": 0.0748, "step": 660 }, { "grad_norm": 0.5958395004272461, "learning_rate": 9.896950636470147e-05, "loss": 0.0883, "step": 670 }, { "grad_norm": 0.46314942836761475, "learning_rate": 9.891310220310666e-05, "loss": 0.077, "step": 680 }, { "grad_norm": 0.4550405442714691, "learning_rate": 9.885521219433823e-05, "loss": 0.0696, "step": 690 }, { "grad_norm": 0.36976656317710876, "learning_rate": 9.879583809693738e-05, "loss": 0.0719, "step": 700 }, { "grad_norm": 0.4276926815509796, "learning_rate": 9.873498171452789e-05, "loss": 0.0842, "step": 710 }, { "grad_norm": 0.6329816579818726, "learning_rate": 9.867264489576135e-05, "loss": 0.0767, "step": 720 }, { "grad_norm": 0.45453768968582153, "learning_rate": 9.860882953426099e-05, "loss": 0.0742, "step": 730 }, { "grad_norm": 0.4058501720428467, "learning_rate": 9.854353756856412e-05, "loss": 0.0643, "step": 740 }, { "grad_norm": 0.3971281349658966, "learning_rate": 9.847677098206332e-05, "loss": 0.0732, "step": 750 }, { "grad_norm": 0.44588446617126465, "learning_rate": 9.840853180294608e-05, "loss": 0.065, "step": 760 }, { "grad_norm": 0.4146692752838135, "learning_rate": 9.833882210413332e-05, "loss": 0.0704, "step": 770 }, { "grad_norm": 0.4290420413017273, "learning_rate": 9.826764400321633e-05, "loss": 0.0677, "step": 780 }, { "grad_norm": 0.4227924942970276, "learning_rate": 9.819499966239243e-05, "loss": 0.0703, "step": 790 }, { "grad_norm": 0.5033952593803406, "learning_rate": 9.812089128839938e-05, "loss": 0.0794, "step": 800 }, { "grad_norm": 0.3571436405181885, "learning_rate": 9.804532113244828e-05, "loss": 0.0717, "step": 810 }, { "grad_norm": 0.29655179381370544, "learning_rate": 9.796829149015517e-05, "loss": 0.0678, "step": 820 }, { "grad_norm": 0.43665337562561035, "learning_rate": 9.788980470147132e-05, "loss": 0.0668, "step": 830 }, { "grad_norm": 0.4904373586177826, "learning_rate": 9.780986315061218e-05, "loss": 0.0643, "step": 840 }, { "grad_norm": 0.3467118442058563, "learning_rate": 9.772846926598491e-05, "loss": 0.0752, "step": 850 }, { "grad_norm": 0.5342732667922974, "learning_rate": 9.76456255201146e-05, "loss": 0.0653, "step": 860 }, { "grad_norm": 0.5031263828277588, "learning_rate": 9.756133442956923e-05, "loss": 0.0638, "step": 870 }, { "grad_norm": 0.37514349818229675, "learning_rate": 9.747559855488313e-05, "loss": 0.063, "step": 880 }, { "grad_norm": 0.35820281505584717, "learning_rate": 9.73884205004793e-05, "loss": 0.0619, "step": 890 }, { "grad_norm": 0.5237784385681152, "learning_rate": 9.729980291459019e-05, "loss": 0.0694, "step": 900 }, { "grad_norm": 0.4646155536174774, "learning_rate": 9.720974848917735e-05, "loss": 0.0645, "step": 910 }, { "grad_norm": 0.4395473897457123, "learning_rate": 9.711825995984957e-05, "loss": 0.0687, "step": 920 }, { "grad_norm": 0.39834749698638916, "learning_rate": 9.702534010577991e-05, "loss": 0.0616, "step": 930 }, { "grad_norm": 0.43580004572868347, "learning_rate": 9.693099174962103e-05, "loss": 0.0602, "step": 940 }, { "grad_norm": 0.47095683217048645, "learning_rate": 9.683521775741977e-05, "loss": 0.0742, "step": 950 }, { "grad_norm": 0.4243863821029663, "learning_rate": 9.673802103852979e-05, "loss": 0.0569, "step": 960 }, { "grad_norm": 0.4031473398208618, "learning_rate": 9.663940454552342e-05, "loss": 0.0585, "step": 970 }, { "grad_norm": 0.47860366106033325, "learning_rate": 9.65393712741018e-05, "loss": 0.0634, "step": 980 }, { "grad_norm": 0.38193196058273315, "learning_rate": 9.6437924263004e-05, "loss": 0.0594, "step": 990 }, { "grad_norm": 0.4677029252052307, "learning_rate": 9.63350665939146e-05, "loss": 0.0617, "step": 1000 }, { "grad_norm": 0.3832956850528717, "learning_rate": 9.623080139137023e-05, "loss": 0.0616, "step": 1010 }, { "grad_norm": 0.36230915784835815, "learning_rate": 9.612513182266447e-05, "loss": 0.0615, "step": 1020 }, { "grad_norm": 0.4043103754520416, "learning_rate": 9.601806109775179e-05, "loss": 0.0664, "step": 1030 }, { "grad_norm": 0.43316778540611267, "learning_rate": 9.590959246914995e-05, "loss": 0.0663, "step": 1040 }, { "grad_norm": 0.32864686846733093, "learning_rate": 9.579972923184122e-05, "loss": 0.0603, "step": 1050 }, { "grad_norm": 0.3839098811149597, "learning_rate": 9.568847472317232e-05, "loss": 0.0642, "step": 1060 }, { "grad_norm": 0.28199896216392517, "learning_rate": 9.557583232275303e-05, "loss": 0.0604, "step": 1070 }, { "grad_norm": 0.3797306716442108, "learning_rate": 9.546180545235344e-05, "loss": 0.0601, "step": 1080 }, { "grad_norm": 0.2888736426830292, "learning_rate": 9.534639757580013e-05, "loss": 0.0495, "step": 1090 }, { "grad_norm": 0.2527846395969391, "learning_rate": 9.522961219887092e-05, "loss": 0.0542, "step": 1100 }, { "grad_norm": 0.26228803396224976, "learning_rate": 9.511145286918828e-05, "loss": 0.0571, "step": 1110 }, { "grad_norm": 0.5870105028152466, "learning_rate": 9.499192317611167e-05, "loss": 0.0649, "step": 1120 }, { "grad_norm": 0.39054936170578003, "learning_rate": 9.487102675062851e-05, "loss": 0.0637, "step": 1130 }, { "grad_norm": 0.3763654828071594, "learning_rate": 9.474876726524374e-05, "loss": 0.0612, "step": 1140 }, { "grad_norm": 0.2446664720773697, "learning_rate": 9.462514843386845e-05, "loss": 0.0572, "step": 1150 }, { "grad_norm": 0.29194676876068115, "learning_rate": 9.450017401170689e-05, "loss": 0.0516, "step": 1160 }, { "grad_norm": 0.51090008020401, "learning_rate": 9.437384779514256e-05, "loss": 0.0573, "step": 1170 }, { "grad_norm": 0.23249898850917816, "learning_rate": 9.424617362162271e-05, "loss": 0.0559, "step": 1180 }, { "grad_norm": 0.3077312409877777, "learning_rate": 9.411715536954196e-05, "loss": 0.054, "step": 1190 }, { "grad_norm": 0.42092403769493103, "learning_rate": 9.39867969581243e-05, "loss": 0.0574, "step": 1200 }, { "grad_norm": 0.45014843344688416, "learning_rate": 9.385510234730415e-05, "loss": 0.0518, "step": 1210 }, { "grad_norm": 0.45236918330192566, "learning_rate": 9.372207553760603e-05, "loss": 0.049, "step": 1220 }, { "grad_norm": 0.3155176341533661, "learning_rate": 9.358772057002312e-05, "loss": 0.0573, "step": 1230 }, { "grad_norm": 0.3163870573043823, "learning_rate": 9.345204152589428e-05, "loss": 0.0504, "step": 1240 }, { "grad_norm": 0.3211176097393036, "learning_rate": 9.331504252678037e-05, "loss": 0.0595, "step": 1250 }, { "grad_norm": 0.4440314769744873, "learning_rate": 9.317672773433876e-05, "loss": 0.0539, "step": 1260 }, { "grad_norm": 0.36051490902900696, "learning_rate": 9.30371013501972e-05, "loss": 0.06, "step": 1270 }, { "grad_norm": 0.43395301699638367, "learning_rate": 9.289616761582587e-05, "loss": 0.0592, "step": 1280 }, { "grad_norm": 0.40232110023498535, "learning_rate": 9.275393081240882e-05, "loss": 0.062, "step": 1290 }, { "grad_norm": 0.43771880865097046, "learning_rate": 9.261039526071374e-05, "loss": 0.0582, "step": 1300 }, { "grad_norm": 0.48542389273643494, "learning_rate": 9.246556532096078e-05, "loss": 0.0544, "step": 1310 }, { "grad_norm": 0.360439270734787, "learning_rate": 9.231944539269009e-05, "loss": 0.0562, "step": 1320 }, { "grad_norm": 0.3617100417613983, "learning_rate": 9.217203991462815e-05, "loss": 0.0521, "step": 1330 }, { "grad_norm": 0.3198077082633972, "learning_rate": 9.202335336455296e-05, "loss": 0.0524, "step": 1340 }, { "grad_norm": 0.36050671339035034, "learning_rate": 9.187339025915802e-05, "loss": 0.0496, "step": 1350 }, { "grad_norm": 0.309641033411026, "learning_rate": 9.17221551539151e-05, "loss": 0.0527, "step": 1360 }, { "grad_norm": 0.2965846359729767, "learning_rate": 9.156965264293586e-05, "loss": 0.0445, "step": 1370 }, { "grad_norm": 0.35571765899658203, "learning_rate": 9.141588735883232e-05, "loss": 0.0523, "step": 1380 }, { "grad_norm": 0.3687618374824524, "learning_rate": 9.126086397257612e-05, "loss": 0.0511, "step": 1390 }, { "grad_norm": 0.43302109837532043, "learning_rate": 9.110458719335659e-05, "loss": 0.0565, "step": 1400 }, { "grad_norm": 0.31573888659477234, "learning_rate": 9.094706176843777e-05, "loss": 0.0484, "step": 1410 }, { "grad_norm": 0.41654664278030396, "learning_rate": 9.078829248301417e-05, "loss": 0.0558, "step": 1420 }, { "grad_norm": 0.49469777941703796, "learning_rate": 9.062828416006539e-05, "loss": 0.0544, "step": 1430 }, { "grad_norm": 0.3401212990283966, "learning_rate": 9.046704166020961e-05, "loss": 0.0563, "step": 1440 }, { "grad_norm": 0.3731771409511566, "learning_rate": 9.030456988155596e-05, "loss": 0.0557, "step": 1450 }, { "grad_norm": 0.3299166262149811, "learning_rate": 9.014087375955573e-05, "loss": 0.0547, "step": 1460 }, { "grad_norm": 0.33593466877937317, "learning_rate": 8.997595826685243e-05, "loss": 0.0554, "step": 1470 }, { "grad_norm": 0.4328671097755432, "learning_rate": 8.980982841313074e-05, "loss": 0.0575, "step": 1480 }, { "grad_norm": 0.35234278440475464, "learning_rate": 8.964248924496435e-05, "loss": 0.0519, "step": 1490 }, { "grad_norm": 0.5080187320709229, "learning_rate": 8.947394584566258e-05, "loss": 0.0602, "step": 1500 }, { "grad_norm": 0.2649039626121521, "learning_rate": 8.930420333511606e-05, "loss": 0.053, "step": 1510 }, { "grad_norm": 0.42354223132133484, "learning_rate": 8.913326686964117e-05, "loss": 0.0501, "step": 1520 }, { "grad_norm": 0.4319063425064087, "learning_rate": 8.89611416418234e-05, "loss": 0.0538, "step": 1530 }, { "grad_norm": 0.4024607539176941, "learning_rate": 8.878783288035957e-05, "loss": 0.0616, "step": 1540 }, { "grad_norm": 0.34141480922698975, "learning_rate": 8.86133458498991e-05, "loss": 0.0514, "step": 1550 }, { "grad_norm": 0.3257942795753479, "learning_rate": 8.843768585088393e-05, "loss": 0.0569, "step": 1560 }, { "grad_norm": 0.22518017888069153, "learning_rate": 8.82608582193877e-05, "loss": 0.048, "step": 1570 }, { "grad_norm": 0.358802855014801, "learning_rate": 8.80828683269535e-05, "loss": 0.0516, "step": 1580 }, { "grad_norm": 0.2343483865261078, "learning_rate": 8.790372158043074e-05, "loss": 0.0494, "step": 1590 }, { "grad_norm": 0.37448617815971375, "learning_rate": 8.772342342181095e-05, "loss": 0.0477, "step": 1600 }, { "grad_norm": 0.36448949575424194, "learning_rate": 8.75419793280624e-05, "loss": 0.0499, "step": 1610 }, { "grad_norm": 0.2531074583530426, "learning_rate": 8.735939481096378e-05, "loss": 0.0425, "step": 1620 }, { "grad_norm": 0.32438674569129944, "learning_rate": 8.717567541693673e-05, "loss": 0.0433, "step": 1630 }, { "grad_norm": 0.3196386992931366, "learning_rate": 8.699082672687734e-05, "loss": 0.051, "step": 1640 }, { "grad_norm": 0.4035435616970062, "learning_rate": 8.680485435598673e-05, "loss": 0.0541, "step": 1650 }, { "grad_norm": 0.2992634177207947, "learning_rate": 8.661776395360029e-05, "loss": 0.0448, "step": 1660 }, { "grad_norm": 0.3280009329319, "learning_rate": 8.642956120301626e-05, "loss": 0.047, "step": 1670 }, { "grad_norm": 0.25291767716407776, "learning_rate": 8.624025182132292e-05, "loss": 0.0473, "step": 1680 }, { "grad_norm": 0.28412434458732605, "learning_rate": 8.604984155922506e-05, "loss": 0.0518, "step": 1690 }, { "grad_norm": 0.3839319944381714, "learning_rate": 8.585833620086918e-05, "loss": 0.0467, "step": 1700 }, { "grad_norm": 0.2821938991546631, "learning_rate": 8.566574156366784e-05, "loss": 0.0486, "step": 1710 }, { "grad_norm": 0.33787041902542114, "learning_rate": 8.547206349812298e-05, "loss": 0.0483, "step": 1720 }, { "grad_norm": 0.39908427000045776, "learning_rate": 8.527730788764805e-05, "loss": 0.0487, "step": 1730 }, { "grad_norm": 0.35582125186920166, "learning_rate": 8.508148064838948e-05, "loss": 0.045, "step": 1740 }, { "grad_norm": 0.42031681537628174, "learning_rate": 8.488458772904684e-05, "loss": 0.05, "step": 1750 }, { "grad_norm": 0.3604871928691864, "learning_rate": 8.468663511069217e-05, "loss": 0.0508, "step": 1760 }, { "grad_norm": 0.3763245940208435, "learning_rate": 8.448762880658825e-05, "loss": 0.0523, "step": 1770 }, { "grad_norm": 0.2665746212005615, "learning_rate": 8.428757486200603e-05, "loss": 0.0466, "step": 1780 }, { "grad_norm": 0.30607786774635315, "learning_rate": 8.40864793540409e-05, "loss": 0.0488, "step": 1790 }, { "grad_norm": 0.42583027482032776, "learning_rate": 8.388434839142813e-05, "loss": 0.0446, "step": 1800 }, { "grad_norm": 0.28605785965919495, "learning_rate": 8.368118811435726e-05, "loss": 0.0463, "step": 1810 }, { "grad_norm": 0.32007771730422974, "learning_rate": 8.347700469428564e-05, "loss": 0.0464, "step": 1820 }, { "grad_norm": 0.4149059057235718, "learning_rate": 8.327180433375091e-05, "loss": 0.0469, "step": 1830 }, { "grad_norm": 0.3714878559112549, "learning_rate": 8.306559326618259e-05, "loss": 0.0468, "step": 1840 }, { "grad_norm": 0.33420297503471375, "learning_rate": 8.285837775571276e-05, "loss": 0.0435, "step": 1850 }, { "grad_norm": 0.22721770405769348, "learning_rate": 8.265016409698573e-05, "loss": 0.047, "step": 1860 }, { "grad_norm": 0.3329220116138458, "learning_rate": 8.244095861496686e-05, "loss": 0.0504, "step": 1870 }, { "grad_norm": 0.479600727558136, "learning_rate": 8.223076766475035e-05, "loss": 0.042, "step": 1880 }, { "grad_norm": 0.4081374704837799, "learning_rate": 8.201959763136633e-05, "loss": 0.0433, "step": 1890 }, { "grad_norm": 0.2546713054180145, "learning_rate": 8.180745492958674e-05, "loss": 0.0533, "step": 1900 }, { "grad_norm": 0.25530946254730225, "learning_rate": 8.159434600373061e-05, "loss": 0.0486, "step": 1910 }, { "grad_norm": 0.370567262172699, "learning_rate": 8.138027732746818e-05, "loss": 0.0452, "step": 1920 }, { "grad_norm": 0.34710201621055603, "learning_rate": 8.116525540362434e-05, "loss": 0.0463, "step": 1930 }, { "grad_norm": 0.30122607946395874, "learning_rate": 8.094928676398101e-05, "loss": 0.0455, "step": 1940 }, { "grad_norm": 0.24021995067596436, "learning_rate": 8.073237796907882e-05, "loss": 0.0472, "step": 1950 }, { "grad_norm": 0.3200050890445709, "learning_rate": 8.051453560801772e-05, "loss": 0.0413, "step": 1960 }, { "grad_norm": 0.2531821131706238, "learning_rate": 8.029576629825687e-05, "loss": 0.0489, "step": 1970 }, { "grad_norm": 0.3085343539714813, "learning_rate": 8.007607668541362e-05, "loss": 0.0532, "step": 1980 }, { "grad_norm": 0.3176669776439667, "learning_rate": 7.985547344306161e-05, "loss": 0.0438, "step": 1990 }, { "grad_norm": 0.3424891531467438, "learning_rate": 7.963396327252812e-05, "loss": 0.0417, "step": 2000 }, { "grad_norm": 0.2856842279434204, "learning_rate": 7.941155290269038e-05, "loss": 0.0398, "step": 2010 }, { "grad_norm": 0.2625652551651001, "learning_rate": 7.918824908977123e-05, "loss": 0.0404, "step": 2020 }, { "grad_norm": 0.33125776052474976, "learning_rate": 7.896405861713394e-05, "loss": 0.0455, "step": 2030 }, { "grad_norm": 0.36468076705932617, "learning_rate": 7.873898829507606e-05, "loss": 0.0499, "step": 2040 }, { "grad_norm": 0.3898219168186188, "learning_rate": 7.851304496062254e-05, "loss": 0.0433, "step": 2050 }, { "grad_norm": 0.3178541958332062, "learning_rate": 7.828623547731818e-05, "loss": 0.0375, "step": 2060 }, { "grad_norm": 0.2662155032157898, "learning_rate": 7.80585667350189e-05, "loss": 0.0431, "step": 2070 }, { "grad_norm": 0.19700299203395844, "learning_rate": 7.783004564968263e-05, "loss": 0.0359, "step": 2080 }, { "grad_norm": 0.2499379962682724, "learning_rate": 7.760067916315921e-05, "loss": 0.035, "step": 2090 }, { "grad_norm": 0.18432468175888062, "learning_rate": 7.737047424297941e-05, "loss": 0.0352, "step": 2100 }, { "grad_norm": 0.3939158022403717, "learning_rate": 7.713943788214337e-05, "loss": 0.0394, "step": 2110 }, { "grad_norm": 0.430083692073822, "learning_rate": 7.690757709890812e-05, "loss": 0.0378, "step": 2120 }, { "grad_norm": 0.23813720047473907, "learning_rate": 7.66748989365744e-05, "loss": 0.041, "step": 2130 }, { "grad_norm": 0.30467575788497925, "learning_rate": 7.644141046327271e-05, "loss": 0.0469, "step": 2140 }, { "grad_norm": 0.3030410408973694, "learning_rate": 7.620711877174866e-05, "loss": 0.0424, "step": 2150 }, { "grad_norm": 0.23633800446987152, "learning_rate": 7.597203097914732e-05, "loss": 0.0393, "step": 2160 }, { "grad_norm": 0.31460145115852356, "learning_rate": 7.573615422679726e-05, "loss": 0.0362, "step": 2170 }, { "grad_norm": 0.4296618700027466, "learning_rate": 7.549949567999345e-05, "loss": 0.0437, "step": 2180 }, { "grad_norm": 0.2874586582183838, "learning_rate": 7.526206252777968e-05, "loss": 0.0341, "step": 2190 }, { "grad_norm": 0.36963751912117004, "learning_rate": 7.50238619827301e-05, "loss": 0.0387, "step": 2200 }, { "grad_norm": 0.38531920313835144, "learning_rate": 7.478490128073022e-05, "loss": 0.0367, "step": 2210 }, { "grad_norm": 0.22951769828796387, "learning_rate": 7.454518768075704e-05, "loss": 0.0385, "step": 2220 }, { "grad_norm": 0.32662105560302734, "learning_rate": 7.430472846465856e-05, "loss": 0.0453, "step": 2230 }, { "grad_norm": 0.22467322647571564, "learning_rate": 7.406353093693253e-05, "loss": 0.0361, "step": 2240 }, { "grad_norm": 0.20656687021255493, "learning_rate": 7.382160242450469e-05, "loss": 0.0408, "step": 2250 }, { "grad_norm": 0.340875506401062, "learning_rate": 7.357895027650598e-05, "loss": 0.0388, "step": 2260 }, { "grad_norm": 0.18724791705608368, "learning_rate": 7.333558186404958e-05, "loss": 0.0375, "step": 2270 }, { "grad_norm": 0.21643663942813873, "learning_rate": 7.309150458000668e-05, "loss": 0.0427, "step": 2280 }, { "grad_norm": 0.22064824402332306, "learning_rate": 7.284672583878219e-05, "loss": 0.0375, "step": 2290 }, { "grad_norm": 0.25080978870391846, "learning_rate": 7.260125307608929e-05, "loss": 0.0337, "step": 2300 }, { "grad_norm": 0.3313097357749939, "learning_rate": 7.235509374872373e-05, "loss": 0.0431, "step": 2310 }, { "grad_norm": 0.28288960456848145, "learning_rate": 7.210825533433719e-05, "loss": 0.0375, "step": 2320 }, { "grad_norm": 0.3031257390975952, "learning_rate": 7.186074533121013e-05, "loss": 0.048, "step": 2330 }, { "grad_norm": 0.2575356960296631, "learning_rate": 7.161257125802413e-05, "loss": 0.0375, "step": 2340 }, { "grad_norm": 0.264274001121521, "learning_rate": 7.136374065363334e-05, "loss": 0.0399, "step": 2350 }, { "grad_norm": 0.3352709412574768, "learning_rate": 7.11142610768356e-05, "loss": 0.0345, "step": 2360 }, { "grad_norm": 0.2115754336118698, "learning_rate": 7.086414010614276e-05, "loss": 0.0374, "step": 2370 }, { "grad_norm": 0.2517990171909332, "learning_rate": 7.061338533955043e-05, "loss": 0.0324, "step": 2380 }, { "grad_norm": 0.24857991933822632, "learning_rate": 7.036200439430725e-05, "loss": 0.0391, "step": 2390 }, { "grad_norm": 0.23163624107837677, "learning_rate": 7.01100049066835e-05, "loss": 0.0389, "step": 2400 }, { "grad_norm": 0.33151012659072876, "learning_rate": 6.985739453173903e-05, "loss": 0.033, "step": 2410 }, { "grad_norm": 0.26381349563598633, "learning_rate": 6.960418094309085e-05, "loss": 0.0341, "step": 2420 }, { "grad_norm": 0.30055221915245056, "learning_rate": 6.93503718326799e-05, "loss": 0.0336, "step": 2430 }, { "grad_norm": 0.25649207830429077, "learning_rate": 6.909597491053751e-05, "loss": 0.0389, "step": 2440 }, { "grad_norm": 0.2667904794216156, "learning_rate": 6.884099790455113e-05, "loss": 0.0321, "step": 2450 }, { "grad_norm": 0.38824063539505005, "learning_rate": 6.858544856022952e-05, "loss": 0.0393, "step": 2460 }, { "grad_norm": 0.2738566994667053, "learning_rate": 6.83293346404676e-05, "loss": 0.0377, "step": 2470 }, { "grad_norm": 0.2428845465183258, "learning_rate": 6.80726639253105e-05, "loss": 0.0332, "step": 2480 }, { "grad_norm": 0.27254635095596313, "learning_rate": 6.781544421171732e-05, "loss": 0.0311, "step": 2490 }, { "grad_norm": 0.3009340465068817, "learning_rate": 6.755768331332424e-05, "loss": 0.0361, "step": 2500 }, { "grad_norm": 0.27296382188796997, "learning_rate": 6.729938906020713e-05, "loss": 0.0388, "step": 2510 }, { "grad_norm": 0.19399593770503998, "learning_rate": 6.704056929864376e-05, "loss": 0.0348, "step": 2520 }, { "grad_norm": 0.3452855050563812, "learning_rate": 6.67812318908754e-05, "loss": 0.0313, "step": 2530 }, { "grad_norm": 0.2279820591211319, "learning_rate": 6.6521384714868e-05, "loss": 0.0372, "step": 2540 }, { "grad_norm": 0.23924601078033447, "learning_rate": 6.626103566407295e-05, "loss": 0.0421, "step": 2550 }, { "grad_norm": 0.26451361179351807, "learning_rate": 6.600019264718713e-05, "loss": 0.0364, "step": 2560 }, { "grad_norm": 0.24230974912643433, "learning_rate": 6.573886358791285e-05, "loss": 0.0356, "step": 2570 }, { "grad_norm": 0.21219466626644135, "learning_rate": 6.547705642471703e-05, "loss": 0.0357, "step": 2580 }, { "grad_norm": 0.21990950405597687, "learning_rate": 6.521477911059008e-05, "loss": 0.0369, "step": 2590 }, { "grad_norm": 0.25543391704559326, "learning_rate": 6.495203961280434e-05, "loss": 0.0332, "step": 2600 }, { "grad_norm": 0.23028869926929474, "learning_rate": 6.468884591267204e-05, "loss": 0.0317, "step": 2610 }, { "grad_norm": 0.25860467553138733, "learning_rate": 6.44252060053028e-05, "loss": 0.0341, "step": 2620 }, { "grad_norm": 0.23476362228393555, "learning_rate": 6.416112789936086e-05, "loss": 0.0304, "step": 2630 }, { "grad_norm": 0.24847416579723358, "learning_rate": 6.389661961682173e-05, "loss": 0.0315, "step": 2640 }, { "grad_norm": 0.31609058380126953, "learning_rate": 6.363168919272846e-05, "loss": 0.039, "step": 2650 }, { "grad_norm": 0.2490130513906479, "learning_rate": 6.336634467494768e-05, "loss": 0.0414, "step": 2660 }, { "grad_norm": 0.26495233178138733, "learning_rate": 6.310059412392505e-05, "loss": 0.0372, "step": 2670 }, { "grad_norm": 0.30787259340286255, "learning_rate": 6.283444561244042e-05, "loss": 0.0396, "step": 2680 }, { "grad_norm": 0.29595232009887695, "learning_rate": 6.256790722536251e-05, "loss": 0.0364, "step": 2690 }, { "grad_norm": 0.30548182129859924, "learning_rate": 6.230098705940354e-05, "loss": 0.0384, "step": 2700 }, { "grad_norm": 0.1984514445066452, "learning_rate": 6.203369322287306e-05, "loss": 0.0342, "step": 2710 }, { "grad_norm": 0.2916681468486786, "learning_rate": 6.17660338354317e-05, "loss": 0.0315, "step": 2720 }, { "grad_norm": 0.26730629801750183, "learning_rate": 6.149801702784456e-05, "loss": 0.0327, "step": 2730 }, { "grad_norm": 0.20792292058467865, "learning_rate": 6.122965094173424e-05, "loss": 0.0351, "step": 2740 }, { "grad_norm": 0.2516806721687317, "learning_rate": 6.0960943729333374e-05, "loss": 0.0307, "step": 2750 }, { "grad_norm": 0.32123464345932007, "learning_rate": 6.069190355323717e-05, "loss": 0.0324, "step": 2760 }, { "grad_norm": 0.282929003238678, "learning_rate": 6.042253858615532e-05, "loss": 0.0321, "step": 2770 }, { "grad_norm": 0.288335919380188, "learning_rate": 6.015285701066382e-05, "loss": 0.0318, "step": 2780 }, { "grad_norm": 0.2876138389110565, "learning_rate": 5.988286701895631e-05, "loss": 0.0299, "step": 2790 }, { "grad_norm": 0.2926459014415741, "learning_rate": 5.961257681259535e-05, "loss": 0.0332, "step": 2800 }, { "grad_norm": 0.242830291390419, "learning_rate": 5.934199460226317e-05, "loss": 0.0324, "step": 2810 }, { "grad_norm": 0.34635767340660095, "learning_rate": 5.9071128607512285e-05, "loss": 0.0319, "step": 2820 }, { "grad_norm": 0.2989971339702606, "learning_rate": 5.8799987056515804e-05, "loss": 0.0341, "step": 2830 }, { "grad_norm": 0.2593464255332947, "learning_rate": 5.8528578185817514e-05, "loss": 0.0378, "step": 2840 }, { "grad_norm": 0.3017624020576477, "learning_rate": 5.825691024008162e-05, "loss": 0.0337, "step": 2850 }, { "grad_norm": 0.23172958195209503, "learning_rate": 5.798499147184233e-05, "loss": 0.0334, "step": 2860 }, { "grad_norm": 0.3011959195137024, "learning_rate": 5.771283014125317e-05, "loss": 0.0343, "step": 2870 }, { "grad_norm": 0.19752350449562073, "learning_rate": 5.7440434515836064e-05, "loss": 0.0329, "step": 2880 }, { "grad_norm": 0.20669515430927277, "learning_rate": 5.7167812870230094e-05, "loss": 0.0325, "step": 2890 }, { "grad_norm": 0.329367995262146, "learning_rate": 5.689497348594035e-05, "loss": 0.0353, "step": 2900 }, { "grad_norm": 0.2281043380498886, "learning_rate": 5.662192465108613e-05, "loss": 0.0458, "step": 2910 }, { "grad_norm": 0.2360934019088745, "learning_rate": 5.634867466014932e-05, "loss": 0.0302, "step": 2920 }, { "grad_norm": 0.2888210117816925, "learning_rate": 5.607523181372234e-05, "loss": 0.0333, "step": 2930 }, { "grad_norm": 0.19709353148937225, "learning_rate": 5.5801604418256117e-05, "loss": 0.0326, "step": 2940 }, { "grad_norm": 0.3951866328716278, "learning_rate": 5.552780078580756e-05, "loss": 0.0331, "step": 2950 }, { "grad_norm": 0.3455570936203003, "learning_rate": 5.525382923378728e-05, "loss": 0.0358, "step": 2960 }, { "grad_norm": 0.40430018305778503, "learning_rate": 5.49796980847068e-05, "loss": 0.0355, "step": 2970 }, { "grad_norm": 0.25933536887168884, "learning_rate": 5.470541566592573e-05, "loss": 0.0303, "step": 2980 }, { "grad_norm": 0.22030013799667358, "learning_rate": 5.443099030939887e-05, "loss": 0.0296, "step": 2990 }, { "grad_norm": 0.27321019768714905, "learning_rate": 5.415643035142309e-05, "loss": 0.037, "step": 3000 }, { "grad_norm": 0.4012751281261444, "learning_rate": 5.3881744132384104e-05, "loss": 0.0305, "step": 3010 }, { "grad_norm": 0.2611137628555298, "learning_rate": 5.360693999650303e-05, "loss": 0.0352, "step": 3020 }, { "grad_norm": 0.24199552834033966, "learning_rate": 5.3332026291583016e-05, "loss": 0.0278, "step": 3030 }, { "grad_norm": 0.16219401359558105, "learning_rate": 5.305701136875566e-05, "loss": 0.026, "step": 3040 }, { "grad_norm": 0.25862357020378113, "learning_rate": 5.278190358222721e-05, "loss": 0.033, "step": 3050 }, { "grad_norm": 0.1735977679491043, "learning_rate": 5.25067112890249e-05, "loss": 0.0328, "step": 3060 }, { "grad_norm": 0.19945716857910156, "learning_rate": 5.2231442848743064e-05, "loss": 0.0341, "step": 3070 }, { "grad_norm": 0.19215127825737, "learning_rate": 5.1956106623289145e-05, "loss": 0.0289, "step": 3080 }, { "grad_norm": 0.24303321540355682, "learning_rate": 5.168071097662972e-05, "loss": 0.0344, "step": 3090 }, { "grad_norm": 0.20462214946746826, "learning_rate": 5.1405264274536445e-05, "loss": 0.0305, "step": 3100 }, { "grad_norm": 0.19431976974010468, "learning_rate": 5.112977488433188e-05, "loss": 0.0389, "step": 3110 }, { "grad_norm": 0.1659228801727295, "learning_rate": 5.085425117463533e-05, "loss": 0.032, "step": 3120 }, { "grad_norm": 0.3553987145423889, "learning_rate": 5.057870151510864e-05, "loss": 0.0354, "step": 3130 }, { "grad_norm": 0.17798061668872833, "learning_rate": 5.030313427620197e-05, "loss": 0.0283, "step": 3140 }, { "grad_norm": 0.2874560058116913, "learning_rate": 5.0027557828899426e-05, "loss": 0.0302, "step": 3150 }, { "grad_norm": 0.22314925491809845, "learning_rate": 4.975198054446492e-05, "loss": 0.0292, "step": 3160 }, { "grad_norm": 0.21092987060546875, "learning_rate": 4.947641079418773e-05, "loss": 0.0308, "step": 3170 }, { "grad_norm": 0.2341562956571579, "learning_rate": 4.920085694912828e-05, "loss": 0.0265, "step": 3180 }, { "grad_norm": 0.2401692271232605, "learning_rate": 4.892532737986387e-05, "loss": 0.0297, "step": 3190 }, { "grad_norm": 0.3158665597438812, "learning_rate": 4.864983045623434e-05, "loss": 0.0333, "step": 3200 }, { "grad_norm": 0.3427079916000366, "learning_rate": 4.837437454708784e-05, "loss": 0.0289, "step": 3210 }, { "grad_norm": 0.2586159408092499, "learning_rate": 4.809896802002662e-05, "loss": 0.0266, "step": 3220 }, { "grad_norm": 0.23725765943527222, "learning_rate": 4.7823619241152854e-05, "loss": 0.0291, "step": 3230 }, { "grad_norm": 0.3046949803829193, "learning_rate": 4.754833657481445e-05, "loss": 0.027, "step": 3240 }, { "grad_norm": 0.24730272591114044, "learning_rate": 4.7273128383351015e-05, "loss": 0.027, "step": 3250 }, { "grad_norm": 0.2536727786064148, "learning_rate": 4.699800302683981e-05, "loss": 0.0347, "step": 3260 }, { "grad_norm": 0.25115495920181274, "learning_rate": 4.6722968862841806e-05, "loss": 0.0294, "step": 3270 }, { "grad_norm": 0.234017014503479, "learning_rate": 4.6448034246147754e-05, "loss": 0.0278, "step": 3280 }, { "grad_norm": 0.19399166107177734, "learning_rate": 4.6173207528524476e-05, "loss": 0.029, "step": 3290 }, { "grad_norm": 0.28640416264533997, "learning_rate": 4.58984970584611e-05, "loss": 0.0259, "step": 3300 }, { "grad_norm": 0.26729804277420044, "learning_rate": 4.562391118091544e-05, "loss": 0.0275, "step": 3310 }, { "grad_norm": 0.20602551102638245, "learning_rate": 4.534945823706056e-05, "loss": 0.029, "step": 3320 }, { "grad_norm": 0.3043207824230194, "learning_rate": 4.507514656403137e-05, "loss": 0.0277, "step": 3330 }, { "grad_norm": 0.18039357662200928, "learning_rate": 4.480098449467132e-05, "loss": 0.026, "step": 3340 }, { "grad_norm": 0.21921129524707794, "learning_rate": 4.452698035727929e-05, "loss": 0.0306, "step": 3350 }, { "grad_norm": 0.24986359477043152, "learning_rate": 4.425314247535668e-05, "loss": 0.0267, "step": 3360 }, { "grad_norm": 0.33457106351852417, "learning_rate": 4.3979479167354477e-05, "loss": 0.0373, "step": 3370 }, { "grad_norm": 0.30760475993156433, "learning_rate": 4.370599874642055e-05, "loss": 0.0323, "step": 3380 }, { "grad_norm": 0.26259493827819824, "learning_rate": 4.3432709520147205e-05, "loss": 0.0302, "step": 3390 }, { "grad_norm": 0.2644908130168915, "learning_rate": 4.315961979031875e-05, "loss": 0.0323, "step": 3400 }, { "grad_norm": 0.24664807319641113, "learning_rate": 4.2886737852659325e-05, "loss": 0.0282, "step": 3410 }, { "grad_norm": 0.2367314249277115, "learning_rate": 4.261407199658093e-05, "loss": 0.0271, "step": 3420 }, { "grad_norm": 0.32822489738464355, "learning_rate": 4.234163050493158e-05, "loss": 0.0253, "step": 3430 }, { "grad_norm": 0.2989010214805603, "learning_rate": 4.2069421653743706e-05, "loss": 0.0285, "step": 3440 }, { "grad_norm": 0.2728559076786041, "learning_rate": 4.179745371198276e-05, "loss": 0.0241, "step": 3450 }, { "grad_norm": 0.20925480127334595, "learning_rate": 4.1525734941296026e-05, "loss": 0.0291, "step": 3460 }, { "grad_norm": 0.2235332429409027, "learning_rate": 4.125427359576162e-05, "loss": 0.0244, "step": 3470 }, { "grad_norm": 0.23195625841617584, "learning_rate": 4.0983077921637815e-05, "loss": 0.0269, "step": 3480 }, { "grad_norm": 0.282319575548172, "learning_rate": 4.07121561571125e-05, "loss": 0.0263, "step": 3490 }, { "grad_norm": 0.29623618721961975, "learning_rate": 4.044151653205292e-05, "loss": 0.0312, "step": 3500 }, { "grad_norm": 0.22592341899871826, "learning_rate": 4.0171167267755696e-05, "loss": 0.0294, "step": 3510 }, { "grad_norm": 0.21023550629615784, "learning_rate": 3.9901116576697083e-05, "loss": 0.0246, "step": 3520 }, { "grad_norm": 0.21181227266788483, "learning_rate": 3.963137266228349e-05, "loss": 0.0249, "step": 3530 }, { "grad_norm": 0.30305394530296326, "learning_rate": 3.93619437186023e-05, "loss": 0.0313, "step": 3540 }, { "grad_norm": 0.21164947748184204, "learning_rate": 3.9092837930172884e-05, "loss": 0.0262, "step": 3550 }, { "grad_norm": 0.29877039790153503, "learning_rate": 3.8824063471698105e-05, "loss": 0.028, "step": 3560 }, { "grad_norm": 0.31443488597869873, "learning_rate": 3.855562850781589e-05, "loss": 0.0292, "step": 3570 }, { "grad_norm": 0.26347023248672485, "learning_rate": 3.828754119285123e-05, "loss": 0.0273, "step": 3580 }, { "grad_norm": 0.23366351425647736, "learning_rate": 3.801980967056851e-05, "loss": 0.0253, "step": 3590 }, { "grad_norm": 0.2442861944437027, "learning_rate": 3.77524420739241e-05, "loss": 0.0274, "step": 3600 }, { "grad_norm": 0.23381778597831726, "learning_rate": 3.748544652481927e-05, "loss": 0.022, "step": 3610 }, { "grad_norm": 0.17044523358345032, "learning_rate": 3.721883113385353e-05, "loss": 0.0275, "step": 3620 }, { "grad_norm": 0.1821439266204834, "learning_rate": 3.695260400007819e-05, "loss": 0.0274, "step": 3630 }, { "grad_norm": 0.272661954164505, "learning_rate": 3.6686773210750385e-05, "loss": 0.0279, "step": 3640 }, { "grad_norm": 0.25035589933395386, "learning_rate": 3.642134684108737e-05, "loss": 0.0306, "step": 3650 }, { "grad_norm": 0.27742305397987366, "learning_rate": 3.615633295402123e-05, "loss": 0.0217, "step": 3660 }, { "grad_norm": 0.18357422947883606, "learning_rate": 3.5891739599953945e-05, "loss": 0.0264, "step": 3670 }, { "grad_norm": 0.2029154896736145, "learning_rate": 3.5627574816512846e-05, "loss": 0.0273, "step": 3680 }, { "grad_norm": 0.25645092129707336, "learning_rate": 3.536384662830648e-05, "loss": 0.0293, "step": 3690 }, { "grad_norm": 0.2826383709907532, "learning_rate": 3.5100563046680764e-05, "loss": 0.0249, "step": 3700 }, { "grad_norm": 0.22233398258686066, "learning_rate": 3.483773206947572e-05, "loss": 0.028, "step": 3710 }, { "grad_norm": 0.21906347572803497, "learning_rate": 3.457536168078247e-05, "loss": 0.0268, "step": 3720 }, { "grad_norm": 0.22737747430801392, "learning_rate": 3.431345985070067e-05, "loss": 0.0231, "step": 3730 }, { "grad_norm": 0.24586769938468933, "learning_rate": 3.40520345350965e-05, "loss": 0.0287, "step": 3740 }, { "grad_norm": 0.1879250854253769, "learning_rate": 3.379109367536089e-05, "loss": 0.0271, "step": 3750 }, { "grad_norm": 0.15110410749912262, "learning_rate": 3.3530645198168295e-05, "loss": 0.0252, "step": 3760 }, { "grad_norm": 0.24863441288471222, "learning_rate": 3.327069701523595e-05, "loss": 0.0296, "step": 3770 }, { "grad_norm": 0.25430187582969666, "learning_rate": 3.301125702308353e-05, "loss": 0.0255, "step": 3780 }, { "grad_norm": 0.21225595474243164, "learning_rate": 3.275233310279321e-05, "loss": 0.0238, "step": 3790 }, { "grad_norm": 0.1806585192680359, "learning_rate": 3.249393311977037e-05, "loss": 0.0258, "step": 3800 }, { "grad_norm": 0.25602349638938904, "learning_rate": 3.223606492350451e-05, "loss": 0.0232, "step": 3810 }, { "grad_norm": 0.19176100194454193, "learning_rate": 3.197873634733096e-05, "loss": 0.026, "step": 3820 }, { "grad_norm": 0.22703717648983002, "learning_rate": 3.172195520819285e-05, "loss": 0.0282, "step": 3830 }, { "grad_norm": 0.15395799279212952, "learning_rate": 3.146572930640362e-05, "loss": 0.0277, "step": 3840 }, { "grad_norm": 0.1866094321012497, "learning_rate": 3.121006642541014e-05, "loss": 0.023, "step": 3850 }, { "grad_norm": 0.20458878576755524, "learning_rate": 3.095497433155626e-05, "loss": 0.0279, "step": 3860 }, { "grad_norm": 0.32518133521080017, "learning_rate": 3.070046077384682e-05, "loss": 0.0269, "step": 3870 }, { "grad_norm": 0.18735451996326447, "learning_rate": 3.0446533483712304e-05, "loss": 0.0236, "step": 3880 }, { "grad_norm": 0.1771942526102066, "learning_rate": 3.0193200174774038e-05, "loss": 0.0222, "step": 3890 }, { "grad_norm": 0.1862131953239441, "learning_rate": 2.994046854260974e-05, "loss": 0.0293, "step": 3900 }, { "grad_norm": 0.15679983794689178, "learning_rate": 2.9688346264519866e-05, "loss": 0.0235, "step": 3910 }, { "grad_norm": 0.16667503118515015, "learning_rate": 2.943684099929436e-05, "loss": 0.0233, "step": 3920 }, { "grad_norm": 0.22355644404888153, "learning_rate": 2.918596038697995e-05, "loss": 0.0249, "step": 3930 }, { "grad_norm": 0.22866907715797424, "learning_rate": 2.8935712048648112e-05, "loss": 0.0285, "step": 3940 }, { "grad_norm": 0.18918220698833466, "learning_rate": 2.8686103586163626e-05, "loss": 0.0261, "step": 3950 }, { "grad_norm": 0.21910104155540466, "learning_rate": 2.843714258195346e-05, "loss": 0.0227, "step": 3960 }, { "grad_norm": 0.2601633071899414, "learning_rate": 2.8188836598776662e-05, "loss": 0.0277, "step": 3970 }, { "grad_norm": 0.1436263620853424, "learning_rate": 2.7941193179494484e-05, "loss": 0.0226, "step": 3980 }, { "grad_norm": 0.15300126373767853, "learning_rate": 2.7694219846841262e-05, "loss": 0.0242, "step": 3990 }, { "grad_norm": 0.1818075329065323, "learning_rate": 2.7447924103195976e-05, "loss": 0.0271, "step": 4000 }, { "grad_norm": 0.2131088674068451, "learning_rate": 2.7202313430354253e-05, "loss": 0.0217, "step": 4010 }, { "grad_norm": 0.2225767970085144, "learning_rate": 2.695739528930111e-05, "loss": 0.0231, "step": 4020 }, { "grad_norm": 0.2219320386648178, "learning_rate": 2.67131771199844e-05, "loss": 0.0255, "step": 4030 }, { "grad_norm": 0.19305405020713806, "learning_rate": 2.6469666341088677e-05, "loss": 0.0241, "step": 4040 }, { "grad_norm": 0.1855151206254959, "learning_rate": 2.6226870349809885e-05, "loss": 0.0213, "step": 4050 }, { "grad_norm": 0.21167217195034027, "learning_rate": 2.5984796521630737e-05, "loss": 0.0257, "step": 4060 }, { "grad_norm": 0.18498139083385468, "learning_rate": 2.574345221009653e-05, "loss": 0.0244, "step": 4070 }, { "grad_norm": 0.2016761749982834, "learning_rate": 2.5502844746591804e-05, "loss": 0.0211, "step": 4080 }, { "grad_norm": 0.225340336561203, "learning_rate": 2.526298144011775e-05, "loss": 0.0234, "step": 4090 }, { "grad_norm": 0.23794017732143402, "learning_rate": 2.5023869577070013e-05, "loss": 0.0261, "step": 4100 }, { "grad_norm": 0.1910785734653473, "learning_rate": 2.478551642101743e-05, "loss": 0.0233, "step": 4110 }, { "grad_norm": 0.211565762758255, "learning_rate": 2.4547929212481435e-05, "loss": 0.0193, "step": 4120 }, { "grad_norm": 0.18004284799098969, "learning_rate": 2.4311115168716013e-05, "loss": 0.0217, "step": 4130 }, { "grad_norm": 0.2306419312953949, "learning_rate": 2.4075081483488494e-05, "loss": 0.0269, "step": 4140 }, { "grad_norm": 0.20246738195419312, "learning_rate": 2.3839835326861104e-05, "loss": 0.0215, "step": 4150 }, { "grad_norm": 0.18215186893939972, "learning_rate": 2.3605383844972966e-05, "loss": 0.0199, "step": 4160 }, { "grad_norm": 0.2206883728504181, "learning_rate": 2.3371734159823284e-05, "loss": 0.0259, "step": 4170 }, { "grad_norm": 0.24595214426517487, "learning_rate": 2.3138893369054766e-05, "loss": 0.0268, "step": 4180 }, { "grad_norm": 0.2135511040687561, "learning_rate": 2.2906868545738102e-05, "loss": 0.0253, "step": 4190 }, { "grad_norm": 0.18752355873584747, "learning_rate": 2.2675666738157186e-05, "loss": 0.0186, "step": 4200 }, { "grad_norm": 0.17715811729431152, "learning_rate": 2.2445294969594844e-05, "loss": 0.0171, "step": 4210 }, { "grad_norm": 0.2948547303676605, "learning_rate": 2.22157602381196e-05, "loss": 0.02, "step": 4220 }, { "grad_norm": 0.12416701763868332, "learning_rate": 2.1987069516373098e-05, "loss": 0.0202, "step": 4230 }, { "grad_norm": 0.1834097057580948, "learning_rate": 2.1759229751358217e-05, "loss": 0.0252, "step": 4240 }, { "grad_norm": 0.16251210868358612, "learning_rate": 2.1532247864228084e-05, "loss": 0.0226, "step": 4250 }, { "grad_norm": 0.11320555210113525, "learning_rate": 2.1306130750075865e-05, "loss": 0.0264, "step": 4260 }, { "grad_norm": 0.16259628534317017, "learning_rate": 2.1080885277725236e-05, "loss": 0.0203, "step": 4270 }, { "grad_norm": 0.16590796411037445, "learning_rate": 2.085651828952175e-05, "loss": 0.0218, "step": 4280 }, { "grad_norm": 0.19988055527210236, "learning_rate": 2.063303660112506e-05, "loss": 0.0237, "step": 4290 }, { "grad_norm": 0.2420492172241211, "learning_rate": 2.0410447001301753e-05, "loss": 0.0227, "step": 4300 }, { "grad_norm": 0.2309999167919159, "learning_rate": 2.0188756251719203e-05, "loss": 0.0194, "step": 4310 }, { "grad_norm": 0.18164952099323273, "learning_rate": 1.9967971086740195e-05, "loss": 0.0209, "step": 4320 }, { "grad_norm": 0.2178877294063568, "learning_rate": 1.974809821321827e-05, "loss": 0.0198, "step": 4330 }, { "grad_norm": 0.21368789672851562, "learning_rate": 1.9529144310294023e-05, "loss": 0.0198, "step": 4340 }, { "grad_norm": 0.22808904945850372, "learning_rate": 1.9311116029192278e-05, "loss": 0.0219, "step": 4350 }, { "grad_norm": 0.2036302238702774, "learning_rate": 1.909401999301993e-05, "loss": 0.0178, "step": 4360 }, { "grad_norm": 0.2856689989566803, "learning_rate": 1.887786279656482e-05, "loss": 0.0213, "step": 4370 }, { "grad_norm": 0.18134640157222748, "learning_rate": 1.8662651006095387e-05, "loss": 0.0201, "step": 4380 }, { "grad_norm": 0.278046190738678, "learning_rate": 1.8448391159161204e-05, "loss": 0.0244, "step": 4390 }, { "grad_norm": 0.26307907700538635, "learning_rate": 1.8235089764394408e-05, "loss": 0.0274, "step": 4400 }, { "grad_norm": 0.22492247819900513, "learning_rate": 1.8022753301311935e-05, "loss": 0.0212, "step": 4410 }, { "grad_norm": 0.19130860269069672, "learning_rate": 1.7811388220118707e-05, "loss": 0.0237, "step": 4420 }, { "grad_norm": 0.18481294810771942, "learning_rate": 1.7601000941511757e-05, "loss": 0.0232, "step": 4430 }, { "grad_norm": 0.1777241826057434, "learning_rate": 1.7391597856485083e-05, "loss": 0.0245, "step": 4440 }, { "grad_norm": 0.2160668522119522, "learning_rate": 1.7183185326135543e-05, "loss": 0.0171, "step": 4450 }, { "grad_norm": 0.20380406081676483, "learning_rate": 1.6975769681469705e-05, "loss": 0.0188, "step": 4460 }, { "grad_norm": 0.20712777972221375, "learning_rate": 1.676935722321139e-05, "loss": 0.0207, "step": 4470 }, { "grad_norm": 0.20267875492572784, "learning_rate": 1.6563954221610355e-05, "loss": 0.0191, "step": 4480 }, { "grad_norm": 0.21463575959205627, "learning_rate": 1.6359566916251845e-05, "loss": 0.0208, "step": 4490 }, { "grad_norm": 0.1477004885673523, "learning_rate": 1.615620151586697e-05, "loss": 0.0215, "step": 4500 }, { "grad_norm": 0.21711723506450653, "learning_rate": 1.5953864198144135e-05, "loss": 0.0241, "step": 4510 }, { "grad_norm": 0.1844751238822937, "learning_rate": 1.5752561109541447e-05, "loss": 0.0208, "step": 4520 }, { "grad_norm": 0.1820620745420456, "learning_rate": 1.5552298365099882e-05, "loss": 0.0218, "step": 4530 }, { "grad_norm": 0.23407231271266937, "learning_rate": 1.5353082048257596e-05, "loss": 0.0191, "step": 4540 }, { "grad_norm": 0.21293601393699646, "learning_rate": 1.5154918210665148e-05, "loss": 0.0195, "step": 4550 }, { "grad_norm": 0.20897363126277924, "learning_rate": 1.4957812872001614e-05, "loss": 0.0186, "step": 4560 }, { "grad_norm": 0.13340842723846436, "learning_rate": 1.4761772019791748e-05, "loss": 0.0192, "step": 4570 }, { "grad_norm": 0.24015949666500092, "learning_rate": 1.4566801609224096e-05, "loss": 0.0187, "step": 4580 }, { "grad_norm": 0.186085045337677, "learning_rate": 1.4372907562970079e-05, "loss": 0.0193, "step": 4590 }, { "grad_norm": 0.18757310509681702, "learning_rate": 1.4180095771004154e-05, "loss": 0.0184, "step": 4600 }, { "grad_norm": 0.2534388303756714, "learning_rate": 1.3988372090424773e-05, "loss": 0.0202, "step": 4610 }, { "grad_norm": 0.16483885049819946, "learning_rate": 1.3797742345276521e-05, "loss": 0.0243, "step": 4620 }, { "grad_norm": 0.1554243564605713, "learning_rate": 1.3608212326373249e-05, "loss": 0.0183, "step": 4630 }, { "grad_norm": 0.22151866555213928, "learning_rate": 1.3419787791122062e-05, "loss": 0.0231, "step": 4640 }, { "grad_norm": 0.21161040663719177, "learning_rate": 1.323247446334847e-05, "loss": 0.0201, "step": 4650 }, { "grad_norm": 0.186412513256073, "learning_rate": 1.3046278033122577e-05, "loss": 0.025, "step": 4660 }, { "grad_norm": 0.18747615814208984, "learning_rate": 1.286120415658611e-05, "loss": 0.0203, "step": 4670 }, { "grad_norm": 0.1292790323495865, "learning_rate": 1.2677258455780683e-05, "loss": 0.0145, "step": 4680 }, { "grad_norm": 0.17994563281536102, "learning_rate": 1.2494446518477022e-05, "loss": 0.0231, "step": 4690 }, { "grad_norm": 0.1702570915222168, "learning_rate": 1.2312773898005175e-05, "loss": 0.0223, "step": 4700 }, { "grad_norm": 0.13994930684566498, "learning_rate": 1.2132246113085822e-05, "loss": 0.0187, "step": 4710 }, { "grad_norm": 0.14905129373073578, "learning_rate": 1.1952868647662696e-05, "loss": 0.0219, "step": 4720 }, { "grad_norm": 0.16883762180805206, "learning_rate": 1.1774646950735913e-05, "loss": 0.0189, "step": 4730 }, { "grad_norm": 0.23552639782428741, "learning_rate": 1.1597586436196473e-05, "loss": 0.0176, "step": 4740 }, { "grad_norm": 0.24354813992977142, "learning_rate": 1.1421692482661856e-05, "loss": 0.0183, "step": 4750 }, { "grad_norm": 0.19600911438465118, "learning_rate": 1.124697043331256e-05, "loss": 0.0209, "step": 4760 }, { "grad_norm": 0.16714124381542206, "learning_rate": 1.107342559572977e-05, "loss": 0.0202, "step": 4770 }, { "grad_norm": 0.12511172890663147, "learning_rate": 1.090106324173426e-05, "loss": 0.0204, "step": 4780 }, { "grad_norm": 0.2145424783229828, "learning_rate": 1.0729888607226113e-05, "loss": 0.0203, "step": 4790 }, { "grad_norm": 0.21264775097370148, "learning_rate": 1.0559906892025745e-05, "loss": 0.0209, "step": 4800 }, { "grad_norm": 0.18948693573474884, "learning_rate": 1.0391123259715906e-05, "loss": 0.017, "step": 4810 }, { "grad_norm": 0.16713643074035645, "learning_rate": 1.0223542837484839e-05, "loss": 0.017, "step": 4820 }, { "grad_norm": 0.21101868152618408, "learning_rate": 1.0057170715970559e-05, "loss": 0.0241, "step": 4830 }, { "grad_norm": 0.14855772256851196, "learning_rate": 9.892011949106172e-06, "loss": 0.0248, "step": 4840 }, { "grad_norm": 0.1694764643907547, "learning_rate": 9.728071553966339e-06, "loss": 0.0199, "step": 4850 }, { "grad_norm": 0.17370817065238953, "learning_rate": 9.56535451061496e-06, "loss": 0.0193, "step": 4860 }, { "grad_norm": 0.14265552163124084, "learning_rate": 9.403865761953779e-06, "loss": 0.0176, "step": 4870 }, { "grad_norm": 0.1497550755739212, "learning_rate": 9.243610213572285e-06, "loss": 0.0228, "step": 4880 }, { "grad_norm": 0.21949000656604767, "learning_rate": 9.084592733598735e-06, "loss": 0.019, "step": 4890 }, { "grad_norm": 0.13839106261730194, "learning_rate": 8.92681815255219e-06, "loss": 0.0178, "step": 4900 }, { "grad_norm": 0.14562994241714478, "learning_rate": 8.770291263195819e-06, "loss": 0.0185, "step": 4910 }, { "grad_norm": 0.19483783841133118, "learning_rate": 8.615016820391342e-06, "loss": 0.0158, "step": 4920 }, { "grad_norm": 0.16348987817764282, "learning_rate": 8.460999540954517e-06, "loss": 0.0182, "step": 4930 }, { "grad_norm": 0.18384204804897308, "learning_rate": 8.308244103511909e-06, "loss": 0.0152, "step": 4940 }, { "grad_norm": 0.16230851411819458, "learning_rate": 8.156755148358764e-06, "loss": 0.0196, "step": 4950 }, { "grad_norm": 0.19277586042881012, "learning_rate": 8.00653727731801e-06, "loss": 0.0156, "step": 4960 }, { "grad_norm": 0.21639972925186157, "learning_rate": 7.857595053600513e-06, "loss": 0.024, "step": 4970 }, { "grad_norm": 0.21124367415905, "learning_rate": 7.709933001666431e-06, "loss": 0.0224, "step": 4980 }, { "grad_norm": 0.17293235659599304, "learning_rate": 7.56355560708778e-06, "loss": 0.016, "step": 4990 }, { "grad_norm": 0.1931348741054535, "learning_rate": 7.418467316412158e-06, "loss": 0.0177, "step": 5000 }, { "grad_norm": 0.14251281321048737, "learning_rate": 7.2746725370277435e-06, "loss": 0.0186, "step": 5010 }, { "grad_norm": 0.15622611343860626, "learning_rate": 7.132175637029293e-06, "loss": 0.02, "step": 5020 }, { "grad_norm": 0.18546737730503082, "learning_rate": 6.9909809450855345e-06, "loss": 0.0193, "step": 5030 }, { "grad_norm": 0.18201150000095367, "learning_rate": 6.851092750307686e-06, "loss": 0.0159, "step": 5040 }, { "grad_norm": 0.20087608695030212, "learning_rate": 6.712515302119077e-06, "loss": 0.0185, "step": 5050 }, { "grad_norm": 0.21938590705394745, "learning_rate": 6.575252810126143e-06, "loss": 0.0181, "step": 5060 }, { "grad_norm": 0.1656438261270523, "learning_rate": 6.439309443990532e-06, "loss": 0.0146, "step": 5070 }, { "grad_norm": 0.1890212744474411, "learning_rate": 6.304689333302416e-06, "loss": 0.0194, "step": 5080 }, { "grad_norm": 0.10458321869373322, "learning_rate": 6.171396567455051e-06, "loss": 0.0186, "step": 5090 }, { "grad_norm": 0.1596899926662445, "learning_rate": 6.039435195520604e-06, "loss": 0.0197, "step": 5100 }, { "grad_norm": 0.14461831748485565, "learning_rate": 5.908809226127054e-06, "loss": 0.0182, "step": 5110 }, { "grad_norm": 0.15653514862060547, "learning_rate": 5.779522627336537e-06, "loss": 0.0263, "step": 5120 }, { "grad_norm": 0.12199040502309799, "learning_rate": 5.651579326524709e-06, "loss": 0.0158, "step": 5130 }, { "grad_norm": 0.2367057204246521, "learning_rate": 5.524983210261481e-06, "loss": 0.0184, "step": 5140 }, { "grad_norm": 0.1951921284198761, "learning_rate": 5.399738124192988e-06, "loss": 0.0171, "step": 5150 }, { "grad_norm": 0.1722964495420456, "learning_rate": 5.2758478729247164e-06, "loss": 0.0203, "step": 5160 }, { "grad_norm": 0.15574219822883606, "learning_rate": 5.153316219905946e-06, "loss": 0.0128, "step": 5170 }, { "grad_norm": 0.16298823058605194, "learning_rate": 5.032146887315448e-06, "loss": 0.0167, "step": 5180 }, { "grad_norm": 0.19788341224193573, "learning_rate": 4.91234355594839e-06, "loss": 0.0198, "step": 5190 }, { "grad_norm": 0.1816922277212143, "learning_rate": 4.7939098651045235e-06, "loss": 0.0212, "step": 5200 }, { "grad_norm": 0.2591005265712738, "learning_rate": 4.67684941247768e-06, "loss": 0.0191, "step": 5210 }, { "grad_norm": 0.1402284801006317, "learning_rate": 4.5611657540464036e-06, "loss": 0.0181, "step": 5220 }, { "grad_norm": 0.17382720112800598, "learning_rate": 4.446862403965984e-06, "loss": 0.0184, "step": 5230 }, { "grad_norm": 0.1312701553106308, "learning_rate": 4.333942834461702e-06, "loss": 0.0153, "step": 5240 }, { "grad_norm": 0.20298051834106445, "learning_rate": 4.222410475723326e-06, "loss": 0.0158, "step": 5250 }, { "grad_norm": 0.10563281178474426, "learning_rate": 4.112268715800943e-06, "loss": 0.0196, "step": 5260 }, { "grad_norm": 0.15568211674690247, "learning_rate": 4.003520900502028e-06, "loss": 0.0189, "step": 5270 }, { "grad_norm": 0.15784397721290588, "learning_rate": 3.8961703332898e-06, "loss": 0.0184, "step": 5280 }, { "grad_norm": 0.1736660897731781, "learning_rate": 3.790220275182854e-06, "loss": 0.0214, "step": 5290 }, { "grad_norm": 0.19348296523094177, "learning_rate": 3.685673944656176e-06, "loss": 0.0175, "step": 5300 }, { "grad_norm": 0.17948998510837555, "learning_rate": 3.582534517543268e-06, "loss": 0.0166, "step": 5310 }, { "grad_norm": 0.24264298379421234, "learning_rate": 3.4808051269397512e-06, "loss": 0.0206, "step": 5320 }, { "grad_norm": 0.15983110666275024, "learning_rate": 3.380488863108183e-06, "loss": 0.0164, "step": 5330 }, { "grad_norm": 0.20391100645065308, "learning_rate": 3.2815887733841365e-06, "loss": 0.0224, "step": 5340 }, { "grad_norm": 0.18342769145965576, "learning_rate": 3.1841078620836683e-06, "loss": 0.0168, "step": 5350 }, { "grad_norm": 0.1805984526872635, "learning_rate": 3.0880490904120874e-06, "loss": 0.0205, "step": 5360 }, { "grad_norm": 0.18293620645999908, "learning_rate": 2.9934153763739205e-06, "loss": 0.0226, "step": 5370 }, { "grad_norm": 0.1255396008491516, "learning_rate": 2.9002095946843277e-06, "loss": 0.0164, "step": 5380 }, { "grad_norm": 0.13082446157932281, "learning_rate": 2.8084345766817676e-06, "loss": 0.0146, "step": 5390 }, { "grad_norm": 0.18769370019435883, "learning_rate": 2.718093110241976e-06, "loss": 0.017, "step": 5400 }, { "grad_norm": 0.13325811922550201, "learning_rate": 2.6291879396933004e-06, "loss": 0.017, "step": 5410 }, { "grad_norm": 0.1486356258392334, "learning_rate": 2.541721765733318e-06, "loss": 0.0136, "step": 5420 }, { "grad_norm": 0.22683610022068024, "learning_rate": 2.455697245346783e-06, "loss": 0.0185, "step": 5430 }, { "grad_norm": 0.14576876163482666, "learning_rate": 2.371116991724953e-06, "loss": 0.017, "step": 5440 }, { "grad_norm": 0.28364524245262146, "learning_rate": 2.2879835741861586e-06, "loss": 0.0224, "step": 5450 }, { "grad_norm": 0.1426982879638672, "learning_rate": 2.206299518097804e-06, "loss": 0.0148, "step": 5460 }, { "grad_norm": 0.14432327449321747, "learning_rate": 2.1260673047996227e-06, "loss": 0.0153, "step": 5470 }, { "grad_norm": 0.15404032170772552, "learning_rate": 2.047289371528299e-06, "loss": 0.018, "step": 5480 }, { "grad_norm": 0.16218610107898712, "learning_rate": 1.96996811134344e-06, "loss": 0.0126, "step": 5490 }, { "grad_norm": 0.14973685145378113, "learning_rate": 1.8941058730549132e-06, "loss": 0.014, "step": 5500 }, { "grad_norm": 0.16633783280849457, "learning_rate": 1.8197049611514194e-06, "loss": 0.0176, "step": 5510 }, { "grad_norm": 0.11517109721899033, "learning_rate": 1.7467676357305561e-06, "loss": 0.0141, "step": 5520 }, { "grad_norm": 0.18079155683517456, "learning_rate": 1.6752961124301415e-06, "loss": 0.0235, "step": 5530 }, { "grad_norm": 0.12392091751098633, "learning_rate": 1.6052925623609049e-06, "loss": 0.0171, "step": 5540 }, { "grad_norm": 0.14315803349018097, "learning_rate": 1.5367591120405256e-06, "loss": 0.0128, "step": 5550 }, { "grad_norm": 0.21578791737556458, "learning_rate": 1.4696978433290653e-06, "loss": 0.0159, "step": 5560 }, { "grad_norm": 0.20014861226081848, "learning_rate": 1.4041107933656928e-06, "loss": 0.0148, "step": 5570 }, { "grad_norm": 0.17368756234645844, "learning_rate": 1.339999954506821e-06, "loss": 0.0174, "step": 5580 }, { "grad_norm": 0.14039455354213715, "learning_rate": 1.2773672742655784e-06, "loss": 0.0165, "step": 5590 }, { "grad_norm": 0.22623248398303986, "learning_rate": 1.2162146552526399e-06, "loss": 0.0205, "step": 5600 }, { "grad_norm": 0.17202813923358917, "learning_rate": 1.1565439551184664e-06, "loss": 0.0179, "step": 5610 }, { "grad_norm": 0.19978828728199005, "learning_rate": 1.0983569864968346e-06, "loss": 0.0189, "step": 5620 }, { "grad_norm": 0.11077994853258133, "learning_rate": 1.0416555169497688e-06, "loss": 0.017, "step": 5630 }, { "grad_norm": 0.17590732872486115, "learning_rate": 9.864412689139123e-07, "loss": 0.0229, "step": 5640 }, { "grad_norm": 0.1582922339439392, "learning_rate": 9.327159196481138e-07, "loss": 0.011, "step": 5650 }, { "grad_norm": 0.16256555914878845, "learning_rate": 8.804811011825398e-07, "loss": 0.0188, "step": 5660 }, { "grad_norm": 0.1413237303495407, "learning_rate": 8.297384002690866e-07, "loss": 0.0163, "step": 5670 }, { "grad_norm": 0.18030285835266113, "learning_rate": 7.804893583331696e-07, "loss": 0.0156, "step": 5680 }, { "grad_norm": 0.1868339627981186, "learning_rate": 7.32735471426893e-07, "loss": 0.0199, "step": 5690 }, { "grad_norm": 0.1561463177204132, "learning_rate": 6.864781901836259e-07, "loss": 0.0145, "step": 5700 }, { "grad_norm": 0.1667407900094986, "learning_rate": 6.417189197739093e-07, "loss": 0.0192, "step": 5710 }, { "grad_norm": 0.19925780594348907, "learning_rate": 5.984590198627849e-07, "loss": 0.0199, "step": 5720 }, { "grad_norm": 0.19619081914424896, "learning_rate": 5.566998045685112e-07, "loss": 0.022, "step": 5730 }, { "grad_norm": 0.15238729119300842, "learning_rate": 5.164425424226016e-07, "loss": 0.0161, "step": 5740 }, { "grad_norm": 0.16857373714447021, "learning_rate": 4.776884563313266e-07, "loss": 0.0163, "step": 5750 }, { "grad_norm": 0.11748020350933075, "learning_rate": 4.404387235385443e-07, "loss": 0.0163, "step": 5760 }, { "grad_norm": 0.13694846630096436, "learning_rate": 4.0469447558995065e-07, "loss": 0.0181, "step": 5770 }, { "grad_norm": 0.1884409785270691, "learning_rate": 3.7045679829870175e-07, "loss": 0.0188, "step": 5780 }, { "grad_norm": 0.1570352464914322, "learning_rate": 3.377267317124233e-07, "loss": 0.0172, "step": 5790 }, { "grad_norm": 0.17494681477546692, "learning_rate": 3.0650527008162513e-07, "loss": 0.0165, "step": 5800 }, { "grad_norm": 0.1305803656578064, "learning_rate": 2.767933618295082e-07, "loss": 0.0144, "step": 5810 }, { "grad_norm": 0.19473201036453247, "learning_rate": 2.485919095231326e-07, "loss": 0.0156, "step": 5820 }, { "grad_norm": 0.1486295610666275, "learning_rate": 2.219017698460002e-07, "loss": 0.0201, "step": 5830 }, { "grad_norm": 0.19880466163158417, "learning_rate": 1.9672375357206452e-07, "loss": 0.0176, "step": 5840 }, { "grad_norm": 0.19250106811523438, "learning_rate": 1.73058625541056e-07, "loss": 0.0212, "step": 5850 }, { "grad_norm": 0.17275409400463104, "learning_rate": 1.5090710463527836e-07, "loss": 0.0164, "step": 5860 }, { "grad_norm": 0.11281181871891022, "learning_rate": 1.3026986375776485e-07, "loss": 0.0174, "step": 5870 }, { "grad_norm": 0.15450893342494965, "learning_rate": 1.1114752981183917e-07, "loss": 0.0247, "step": 5880 }, { "grad_norm": 0.15461589395999908, "learning_rate": 9.354068368204739e-08, "loss": 0.0153, "step": 5890 }, { "grad_norm": 0.15060506761074066, "learning_rate": 7.744986021656076e-08, "loss": 0.0182, "step": 5900 }, { "grad_norm": 0.12657040357589722, "learning_rate": 6.287554821087783e-08, "loss": 0.0173, "step": 5910 }, { "grad_norm": 0.1401786506175995, "learning_rate": 4.981819039300284e-08, "loss": 0.0192, "step": 5920 }, { "grad_norm": 0.15238840878009796, "learning_rate": 3.827818341000655e-08, "loss": 0.0157, "step": 5930 }, { "grad_norm": 0.10546451807022095, "learning_rate": 2.8255877815946963e-08, "loss": 0.0147, "step": 5940 }, { "grad_norm": 0.15890178084373474, "learning_rate": 1.9751578061244504e-08, "loss": 0.0149, "step": 5950 }, { "grad_norm": 0.17914965748786926, "learning_rate": 1.2765542483417214e-08, "loss": 0.0151, "step": 5960 }, { "grad_norm": 0.20467838644981384, "learning_rate": 7.2979832992592365e-09, "loss": 0.0157, "step": 5970 }, { "grad_norm": 0.21326901018619537, "learning_rate": 3.349066598362649e-09, "loss": 0.0195, "step": 5980 }, { "grad_norm": 0.20198704302310944, "learning_rate": 9.189123380826114e-10, "loss": 0.0161, "step": 5990 }, { "grad_norm": 0.1654694527387619, "learning_rate": 7.594339912486703e-12, "loss": 0.0169, "step": 6000 } ], "logging_steps": 10, "max_steps": 6000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 2000, "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 }