{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.483004093170166, "learning_rate": 1.8e-06, "loss": 0.2183, "step": 10 }, { "grad_norm": 1.2604761123657227, "learning_rate": 3.8e-06, "loss": 0.1928, "step": 20 }, { "grad_norm": 0.7669920325279236, "learning_rate": 5.8e-06, "loss": 0.1433, "step": 30 }, { "grad_norm": 0.7105371356010437, "learning_rate": 7.8e-06, "loss": 0.1175, "step": 40 }, { "grad_norm": 1.1575429439544678, "learning_rate": 9.800000000000001e-06, "loss": 0.1159, "step": 50 }, { "grad_norm": 0.8928622603416443, "learning_rate": 1.18e-05, "loss": 0.1356, "step": 60 }, { "grad_norm": 0.7495980858802795, "learning_rate": 1.3800000000000002e-05, "loss": 0.1011, "step": 70 }, { "grad_norm": 1.3270576000213623, "learning_rate": 1.58e-05, "loss": 0.1049, "step": 80 }, { "grad_norm": 1.1671878099441528, "learning_rate": 1.78e-05, "loss": 0.1042, "step": 90 }, { "grad_norm": 0.5815449357032776, "learning_rate": 1.9800000000000004e-05, "loss": 0.1056, "step": 100 }, { "grad_norm": 1.1978652477264404, "learning_rate": 2.18e-05, "loss": 0.0996, "step": 110 }, { "grad_norm": 0.9299995303153992, "learning_rate": 2.38e-05, "loss": 0.0955, "step": 120 }, { "grad_norm": 1.215566635131836, "learning_rate": 2.58e-05, "loss": 0.1061, "step": 130 }, { "grad_norm": 1.2217522859573364, "learning_rate": 2.7800000000000005e-05, "loss": 0.109, "step": 140 }, { "grad_norm": 0.8476248383522034, "learning_rate": 2.98e-05, "loss": 0.0966, "step": 150 }, { "grad_norm": 0.6920000314712524, "learning_rate": 3.18e-05, "loss": 0.108, "step": 160 }, { "grad_norm": 1.2156225442886353, "learning_rate": 3.38e-05, "loss": 0.1078, "step": 170 }, { "grad_norm": 0.994560718536377, "learning_rate": 3.58e-05, "loss": 0.1062, "step": 180 }, { "grad_norm": 0.8169199228286743, "learning_rate": 3.7800000000000004e-05, "loss": 0.1136, "step": 190 }, { "grad_norm": 1.4434038400650024, "learning_rate": 3.9800000000000005e-05, "loss": 0.1077, "step": 200 }, { "grad_norm": 1.0772769451141357, "learning_rate": 4.18e-05, "loss": 0.0968, "step": 210 }, { "grad_norm": 0.9031879901885986, "learning_rate": 4.38e-05, "loss": 0.0868, "step": 220 }, { "grad_norm": 0.800814151763916, "learning_rate": 4.58e-05, "loss": 0.0922, "step": 230 }, { "grad_norm": 1.0411266088485718, "learning_rate": 4.78e-05, "loss": 0.1046, "step": 240 }, { "grad_norm": 0.7798251509666443, "learning_rate": 4.9800000000000004e-05, "loss": 0.0817, "step": 250 }, { "grad_norm": 0.6314559578895569, "learning_rate": 5.1800000000000005e-05, "loss": 0.0996, "step": 260 }, { "grad_norm": 0.6789538264274597, "learning_rate": 5.380000000000001e-05, "loss": 0.097, "step": 270 }, { "grad_norm": 0.7939445376396179, "learning_rate": 5.580000000000001e-05, "loss": 0.1036, "step": 280 }, { "grad_norm": 0.8318786025047302, "learning_rate": 5.7799999999999995e-05, "loss": 0.1283, "step": 290 }, { "grad_norm": 0.531631588935852, "learning_rate": 5.9800000000000003e-05, "loss": 0.1082, "step": 300 }, { "grad_norm": 0.7565068602561951, "learning_rate": 6.18e-05, "loss": 0.0891, "step": 310 }, { "grad_norm": 1.4697751998901367, "learning_rate": 6.38e-05, "loss": 0.1018, "step": 320 }, { "grad_norm": 0.5695682764053345, "learning_rate": 6.58e-05, "loss": 0.1166, "step": 330 }, { "grad_norm": 1.1098891496658325, "learning_rate": 6.780000000000001e-05, "loss": 0.1149, "step": 340 }, { "grad_norm": 0.5062991976737976, "learning_rate": 6.98e-05, "loss": 0.0869, "step": 350 }, { "grad_norm": 0.6086779236793518, "learning_rate": 7.18e-05, "loss": 0.1128, "step": 360 }, { "grad_norm": 0.601087749004364, "learning_rate": 7.38e-05, "loss": 0.0961, "step": 370 }, { "grad_norm": 0.871684730052948, "learning_rate": 7.58e-05, "loss": 0.1175, "step": 380 }, { "grad_norm": 1.0449392795562744, "learning_rate": 7.780000000000001e-05, "loss": 0.0993, "step": 390 }, { "grad_norm": 0.5058386325836182, "learning_rate": 7.98e-05, "loss": 0.1148, "step": 400 }, { "grad_norm": 0.7068254947662354, "learning_rate": 8.18e-05, "loss": 0.1078, "step": 410 }, { "grad_norm": 0.5794460773468018, "learning_rate": 8.38e-05, "loss": 0.1004, "step": 420 }, { "grad_norm": 0.6006048917770386, "learning_rate": 8.58e-05, "loss": 0.1012, "step": 430 }, { "grad_norm": 0.7644090056419373, "learning_rate": 8.78e-05, "loss": 0.0988, "step": 440 }, { "grad_norm": 0.7550540566444397, "learning_rate": 8.98e-05, "loss": 0.1115, "step": 450 }, { "grad_norm": 0.7580723762512207, "learning_rate": 9.180000000000001e-05, "loss": 0.1073, "step": 460 }, { "grad_norm": 0.7475100159645081, "learning_rate": 9.38e-05, "loss": 0.0798, "step": 470 }, { "grad_norm": 0.6530067324638367, "learning_rate": 9.58e-05, "loss": 0.1042, "step": 480 }, { "grad_norm": 0.6424221992492676, "learning_rate": 9.78e-05, "loss": 0.1269, "step": 490 }, { "grad_norm": 0.5241517424583435, "learning_rate": 9.98e-05, "loss": 0.0954, "step": 500 }, { "grad_norm": 0.5798940658569336, "learning_rate": 9.9999778549206e-05, "loss": 0.0932, "step": 510 }, { "grad_norm": 1.3433797359466553, "learning_rate": 9.999901304280685e-05, "loss": 0.1037, "step": 520 }, { "grad_norm": 0.4437280595302582, "learning_rate": 9.999770075521164e-05, "loss": 0.0952, "step": 530 }, { "grad_norm": 0.6020622253417969, "learning_rate": 9.99958417007713e-05, "loss": 0.0997, "step": 540 }, { "grad_norm": 0.9712867736816406, "learning_rate": 9.999343589981615e-05, "loss": 0.1069, "step": 550 }, { "grad_norm": 1.1637797355651855, "learning_rate": 9.999048337865568e-05, "loss": 0.1198, "step": 560 }, { "grad_norm": 0.6334680914878845, "learning_rate": 9.998698416957815e-05, "loss": 0.0937, "step": 570 }, { "grad_norm": 0.5274465084075928, "learning_rate": 9.998293831085037e-05, "loss": 0.092, "step": 580 }, { "grad_norm": 0.4338470697402954, "learning_rate": 9.997834584671719e-05, "loss": 0.0968, "step": 590 }, { "grad_norm": 0.5243820548057556, "learning_rate": 9.997320682740107e-05, "loss": 0.0903, "step": 600 }, { "grad_norm": 0.5944190621376038, "learning_rate": 9.996752130910149e-05, "loss": 0.0739, "step": 610 }, { "grad_norm": 0.3870990574359894, "learning_rate": 9.99612893539944e-05, "loss": 0.0768, "step": 620 }, { "grad_norm": 0.6159060597419739, "learning_rate": 9.995451103023144e-05, "loss": 0.0819, "step": 630 }, { "grad_norm": 0.7722949385643005, "learning_rate": 9.994718641193928e-05, "loss": 0.0746, "step": 640 }, { "grad_norm": 0.5177273750305176, "learning_rate": 9.993931557921874e-05, "loss": 0.0863, "step": 650 }, { "grad_norm": 0.9104658961296082, "learning_rate": 9.993089861814402e-05, "loss": 0.105, "step": 660 }, { "grad_norm": 0.419617235660553, "learning_rate": 9.992193562076166e-05, "loss": 0.111, "step": 670 }, { "grad_norm": 0.5497177839279175, "learning_rate": 9.991242668508954e-05, "loss": 0.0919, "step": 680 }, { "grad_norm": 0.7154567837715149, "learning_rate": 9.990237191511587e-05, "loss": 0.1, "step": 690 }, { "grad_norm": 0.4858478307723999, "learning_rate": 9.989177142079802e-05, "loss": 0.0928, "step": 700 }, { "grad_norm": 0.4792208969593048, "learning_rate": 9.988062531806126e-05, "loss": 0.0946, "step": 710 }, { "grad_norm": 0.38089263439178467, "learning_rate": 9.986893372879762e-05, "loss": 0.0755, "step": 720 }, { "grad_norm": 0.8900238275527954, "learning_rate": 9.985669678086443e-05, "loss": 0.0997, "step": 730 }, { "grad_norm": 0.6000534892082214, "learning_rate": 9.984391460808298e-05, "loss": 0.0859, "step": 740 }, { "grad_norm": 0.4302968382835388, "learning_rate": 9.983058735023709e-05, "loss": 0.0874, "step": 750 }, { "grad_norm": 0.7062979340553284, "learning_rate": 9.98167151530715e-05, "loss": 0.0932, "step": 760 }, { "grad_norm": 0.6229379773139954, "learning_rate": 9.980229816829034e-05, "loss": 0.0829, "step": 770 }, { "grad_norm": 1.0596909523010254, "learning_rate": 9.978733655355544e-05, "loss": 0.0986, "step": 780 }, { "grad_norm": 0.960236132144928, "learning_rate": 9.977183047248464e-05, "loss": 0.1009, "step": 790 }, { "grad_norm": 0.5637387633323669, "learning_rate": 9.975578009464992e-05, "loss": 0.0701, "step": 800 }, { "grad_norm": 0.44942381978034973, "learning_rate": 9.97391855955757e-05, "loss": 0.1023, "step": 810 }, { "grad_norm": 0.5455422401428223, "learning_rate": 9.972204715673669e-05, "loss": 0.0983, "step": 820 }, { "grad_norm": 0.7341344356536865, "learning_rate": 9.970436496555617e-05, "loss": 0.0909, "step": 830 }, { "grad_norm": 0.3503268361091614, "learning_rate": 9.968613921540373e-05, "loss": 0.0938, "step": 840 }, { "grad_norm": 0.458376407623291, "learning_rate": 9.966737010559326e-05, "loss": 0.082, "step": 850 }, { "grad_norm": 0.596380352973938, "learning_rate": 9.964805784138072e-05, "loss": 0.0727, "step": 860 }, { "grad_norm": 0.8671439290046692, "learning_rate": 9.962820263396195e-05, "loss": 0.0959, "step": 870 }, { "grad_norm": 0.5051323771476746, "learning_rate": 9.960780470047033e-05, "loss": 0.0748, "step": 880 }, { "grad_norm": 0.7690367102622986, "learning_rate": 9.958686426397437e-05, "loss": 0.1017, "step": 890 }, { "grad_norm": 0.4848758280277252, "learning_rate": 9.956538155347534e-05, "loss": 0.0919, "step": 900 }, { "grad_norm": 0.6932089924812317, "learning_rate": 9.95433568039047e-05, "loss": 0.0967, "step": 910 }, { "grad_norm": 0.35151025652885437, "learning_rate": 9.952079025612162e-05, "loss": 0.071, "step": 920 }, { "grad_norm": 0.5889578461647034, "learning_rate": 9.949768215691022e-05, "loss": 0.0763, "step": 930 }, { "grad_norm": 0.3519958555698395, "learning_rate": 9.9474032758977e-05, "loss": 0.0863, "step": 940 }, { "grad_norm": 0.5173790454864502, "learning_rate": 9.944984232094794e-05, "loss": 0.0713, "step": 950 }, { "grad_norm": 0.49300429224967957, "learning_rate": 9.942511110736584e-05, "loss": 0.0717, "step": 960 }, { "grad_norm": 0.40944114327430725, "learning_rate": 9.939983938868726e-05, "loss": 0.0826, "step": 970 }, { "grad_norm": 0.46141117811203003, "learning_rate": 9.93740274412797e-05, "loss": 0.086, "step": 980 }, { "grad_norm": 0.961546003818512, "learning_rate": 9.934767554741846e-05, "loss": 0.0776, "step": 990 }, { "grad_norm": 0.6107563972473145, "learning_rate": 9.932078399528361e-05, "loss": 0.0849, "step": 1000 }, { "grad_norm": 0.3805157244205475, "learning_rate": 9.929335307895689e-05, "loss": 0.0767, "step": 1010 }, { "grad_norm": 0.4135192334651947, "learning_rate": 9.926538309841839e-05, "loss": 0.069, "step": 1020 }, { "grad_norm": 0.5375974774360657, "learning_rate": 9.923687435954334e-05, "loss": 0.0621, "step": 1030 }, { "grad_norm": 0.4287557601928711, "learning_rate": 9.920782717409873e-05, "loss": 0.0707, "step": 1040 }, { "grad_norm": 0.4022911787033081, "learning_rate": 9.917824185973994e-05, "loss": 0.0676, "step": 1050 }, { "grad_norm": 0.325527548789978, "learning_rate": 9.914811874000723e-05, "loss": 0.0675, "step": 1060 }, { "grad_norm": 0.4964507520198822, "learning_rate": 9.911745814432218e-05, "loss": 0.0787, "step": 1070 }, { "grad_norm": 0.452700674533844, "learning_rate": 9.90862604079842e-05, "loss": 0.0908, "step": 1080 }, { "grad_norm": 0.6627142429351807, "learning_rate": 9.90545258721667e-05, "loss": 0.073, "step": 1090 }, { "grad_norm": 0.455059289932251, "learning_rate": 9.90222548839135e-05, "loss": 0.0822, "step": 1100 }, { "grad_norm": 0.5313907861709595, "learning_rate": 9.898944779613495e-05, "loss": 0.0657, "step": 1110 }, { "grad_norm": 0.4485672116279602, "learning_rate": 9.89561049676041e-05, "loss": 0.0719, "step": 1120 }, { "grad_norm": 0.5575585961341858, "learning_rate": 9.89222267629528e-05, "loss": 0.0813, "step": 1130 }, { "grad_norm": 0.5916008949279785, "learning_rate": 9.888781355266763e-05, "loss": 0.0982, "step": 1140 }, { "grad_norm": 0.5012277364730835, "learning_rate": 9.885286571308598e-05, "loss": 0.0717, "step": 1150 }, { "grad_norm": 0.6890244483947754, "learning_rate": 9.881738362639182e-05, "loss": 0.0819, "step": 1160 }, { "grad_norm": 0.3464037775993347, "learning_rate": 9.878136768061154e-05, "loss": 0.0659, "step": 1170 }, { "grad_norm": 0.42304810881614685, "learning_rate": 9.874481826960979e-05, "loss": 0.0621, "step": 1180 }, { "grad_norm": 0.3439233601093292, "learning_rate": 9.870773579308503e-05, "loss": 0.0769, "step": 1190 }, { "grad_norm": 0.5315023064613342, "learning_rate": 9.867012065656533e-05, "loss": 0.0838, "step": 1200 }, { "grad_norm": 0.5417388677597046, "learning_rate": 9.863197327140376e-05, "loss": 0.0716, "step": 1210 }, { "grad_norm": 0.4340190887451172, "learning_rate": 9.859329405477403e-05, "loss": 0.0725, "step": 1220 }, { "grad_norm": 0.3918209969997406, "learning_rate": 9.855408342966585e-05, "loss": 0.0622, "step": 1230 }, { "grad_norm": 0.4345376491546631, "learning_rate": 9.851434182488033e-05, "loss": 0.076, "step": 1240 }, { "grad_norm": 0.35629889369010925, "learning_rate": 9.84740696750253e-05, "loss": 0.0593, "step": 1250 }, { "grad_norm": 0.456149160861969, "learning_rate": 9.843326742051055e-05, "loss": 0.0592, "step": 1260 }, { "grad_norm": 0.39540812373161316, "learning_rate": 9.839193550754297e-05, "loss": 0.0789, "step": 1270 }, { "grad_norm": 0.31466761231422424, "learning_rate": 9.835007438812177e-05, "loss": 0.0678, "step": 1280 }, { "grad_norm": 0.5016930103302002, "learning_rate": 9.830768452003341e-05, "loss": 0.0652, "step": 1290 }, { "grad_norm": 0.355256587266922, "learning_rate": 9.826476636684671e-05, "loss": 0.06, "step": 1300 }, { "grad_norm": 0.3938450217247009, "learning_rate": 9.822132039790773e-05, "loss": 0.0706, "step": 1310 }, { "grad_norm": 0.3223625421524048, "learning_rate": 9.817734708833461e-05, "loss": 0.0595, "step": 1320 }, { "grad_norm": 0.5285195708274841, "learning_rate": 9.813284691901243e-05, "loss": 0.0951, "step": 1330 }, { "grad_norm": 0.37530797719955444, "learning_rate": 9.808782037658792e-05, "loss": 0.0812, "step": 1340 }, { "grad_norm": 0.360025554895401, "learning_rate": 9.804226795346411e-05, "loss": 0.0867, "step": 1350 }, { "grad_norm": 0.4104430079460144, "learning_rate": 9.799619014779503e-05, "loss": 0.064, "step": 1360 }, { "grad_norm": 0.40579068660736084, "learning_rate": 9.794958746348013e-05, "loss": 0.0575, "step": 1370 }, { "grad_norm": 0.4490329623222351, "learning_rate": 9.790246041015896e-05, "loss": 0.0621, "step": 1380 }, { "grad_norm": 0.4080933928489685, "learning_rate": 9.785480950320538e-05, "loss": 0.068, "step": 1390 }, { "grad_norm": 0.39906758069992065, "learning_rate": 9.78066352637221e-05, "loss": 0.0615, "step": 1400 }, { "grad_norm": 0.3616662919521332, "learning_rate": 9.775793821853488e-05, "loss": 0.0655, "step": 1410 }, { "grad_norm": 0.46691909432411194, "learning_rate": 9.77087189001868e-05, "loss": 0.0664, "step": 1420 }, { "grad_norm": 0.4176214635372162, "learning_rate": 9.765897784693243e-05, "loss": 0.0692, "step": 1430 }, { "grad_norm": 0.3964342474937439, "learning_rate": 9.760871560273197e-05, "loss": 0.064, "step": 1440 }, { "grad_norm": 0.35268643498420715, "learning_rate": 9.755793271724526e-05, "loss": 0.0576, "step": 1450 }, { "grad_norm": 0.29759541153907776, "learning_rate": 9.750662974582584e-05, "loss": 0.0533, "step": 1460 }, { "grad_norm": 0.4741663634777069, "learning_rate": 9.745480724951473e-05, "loss": 0.0638, "step": 1470 }, { "grad_norm": 0.32099461555480957, "learning_rate": 9.740246579503447e-05, "loss": 0.0567, "step": 1480 }, { "grad_norm": 0.3192496597766876, "learning_rate": 9.734960595478284e-05, "loss": 0.0479, "step": 1490 }, { "grad_norm": 0.30460673570632935, "learning_rate": 9.729622830682657e-05, "loss": 0.0616, "step": 1500 }, { "grad_norm": 0.44033607840538025, "learning_rate": 9.724233343489504e-05, "loss": 0.0635, "step": 1510 }, { "grad_norm": 0.4035488963127136, "learning_rate": 9.718792192837396e-05, "loss": 0.0596, "step": 1520 }, { "grad_norm": 0.4594622552394867, "learning_rate": 9.713299438229886e-05, "loss": 0.0693, "step": 1530 }, { "grad_norm": 0.5418776869773865, "learning_rate": 9.707755139734855e-05, "loss": 0.0694, "step": 1540 }, { "grad_norm": 0.3242817223072052, "learning_rate": 9.702159357983866e-05, "loss": 0.0808, "step": 1550 }, { "grad_norm": 0.48201990127563477, "learning_rate": 9.696512154171492e-05, "loss": 0.0929, "step": 1560 }, { "grad_norm": 0.3746156394481659, "learning_rate": 9.690813590054645e-05, "loss": 0.0753, "step": 1570 }, { "grad_norm": 0.4728257358074188, "learning_rate": 9.685063727951914e-05, "loss": 0.0797, "step": 1580 }, { "grad_norm": 0.3726077675819397, "learning_rate": 9.679262630742865e-05, "loss": 0.0693, "step": 1590 }, { "grad_norm": 0.33868861198425293, "learning_rate": 9.673410361867373e-05, "loss": 0.055, "step": 1600 }, { "grad_norm": 0.4609202742576599, "learning_rate": 9.667506985324909e-05, "loss": 0.0608, "step": 1610 }, { "grad_norm": 0.653600811958313, "learning_rate": 9.661552565673855e-05, "loss": 0.0633, "step": 1620 }, { "grad_norm": 0.37413015961647034, "learning_rate": 9.655547168030789e-05, "loss": 0.0806, "step": 1630 }, { "grad_norm": 0.35416433215141296, "learning_rate": 9.649490858069777e-05, "loss": 0.0682, "step": 1640 }, { "grad_norm": 0.3177140951156616, "learning_rate": 9.643383702021658e-05, "loss": 0.0613, "step": 1650 }, { "grad_norm": 0.4641144871711731, "learning_rate": 9.637225766673307e-05, "loss": 0.0741, "step": 1660 }, { "grad_norm": 0.46640968322753906, "learning_rate": 9.631017119366922e-05, "loss": 0.0638, "step": 1670 }, { "grad_norm": 0.48948466777801514, "learning_rate": 9.624757827999273e-05, "loss": 0.0668, "step": 1680 }, { "grad_norm": 1.0846784114837646, "learning_rate": 9.618447961020971e-05, "loss": 0.0683, "step": 1690 }, { "grad_norm": 0.43397754430770874, "learning_rate": 9.612087587435707e-05, "loss": 0.0608, "step": 1700 }, { "grad_norm": 0.5121847987174988, "learning_rate": 9.605676776799508e-05, "loss": 0.0734, "step": 1710 }, { "grad_norm": 0.39016175270080566, "learning_rate": 9.599215599219973e-05, "loss": 0.0597, "step": 1720 }, { "grad_norm": 0.4705209732055664, "learning_rate": 9.592704125355505e-05, "loss": 0.0599, "step": 1730 }, { "grad_norm": 0.44376662373542786, "learning_rate": 9.586142426414538e-05, "loss": 0.0549, "step": 1740 }, { "grad_norm": 0.7253206968307495, "learning_rate": 9.57953057415476e-05, "loss": 0.085, "step": 1750 }, { "grad_norm": 0.3589021563529968, "learning_rate": 9.572868640882328e-05, "loss": 0.0688, "step": 1760 }, { "grad_norm": 0.4356394112110138, "learning_rate": 9.56615669945108e-05, "loss": 0.061, "step": 1770 }, { "grad_norm": 0.517203152179718, "learning_rate": 9.55939482326173e-05, "loss": 0.0693, "step": 1780 }, { "grad_norm": 0.49198028445243835, "learning_rate": 9.552583086261069e-05, "loss": 0.064, "step": 1790 }, { "grad_norm": 0.2453034371137619, "learning_rate": 9.545721562941168e-05, "loss": 0.0589, "step": 1800 }, { "grad_norm": 0.4335815906524658, "learning_rate": 9.538810328338543e-05, "loss": 0.049, "step": 1810 }, { "grad_norm": 0.9689037799835205, "learning_rate": 9.531849458033349e-05, "loss": 0.0655, "step": 1820 }, { "grad_norm": 0.4750818610191345, "learning_rate": 9.524839028148547e-05, "loss": 0.0679, "step": 1830 }, { "grad_norm": 0.4282602071762085, "learning_rate": 9.517779115349077e-05, "loss": 0.0645, "step": 1840 }, { "grad_norm": 0.28931593894958496, "learning_rate": 9.510669796841014e-05, "loss": 0.0754, "step": 1850 }, { "grad_norm": 0.48506951332092285, "learning_rate": 9.503511150370727e-05, "loss": 0.0807, "step": 1860 }, { "grad_norm": 0.32402390241622925, "learning_rate": 9.496303254224024e-05, "loss": 0.0578, "step": 1870 }, { "grad_norm": 0.39468804001808167, "learning_rate": 9.489046187225306e-05, "loss": 0.0687, "step": 1880 }, { "grad_norm": 0.34290972352027893, "learning_rate": 9.481740028736692e-05, "loss": 0.0695, "step": 1890 }, { "grad_norm": 0.47554826736450195, "learning_rate": 9.474384858657164e-05, "loss": 0.0526, "step": 1900 }, { "grad_norm": 0.40521764755249023, "learning_rate": 9.466980757421679e-05, "loss": 0.0806, "step": 1910 }, { "grad_norm": 0.4195276200771332, "learning_rate": 9.459527806000305e-05, "loss": 0.0625, "step": 1920 }, { "grad_norm": 0.28959736227989197, "learning_rate": 9.452026085897325e-05, "loss": 0.0481, "step": 1930 }, { "grad_norm": 0.26217034459114075, "learning_rate": 9.444475679150348e-05, "loss": 0.0503, "step": 1940 }, { "grad_norm": 0.3181614279747009, "learning_rate": 9.436876668329411e-05, "loss": 0.0461, "step": 1950 }, { "grad_norm": 0.625689685344696, "learning_rate": 9.429229136536079e-05, "loss": 0.0981, "step": 1960 }, { "grad_norm": 0.4893723428249359, "learning_rate": 9.421533167402534e-05, "loss": 0.0673, "step": 1970 }, { "grad_norm": 0.4205009341239929, "learning_rate": 9.413788845090666e-05, "loss": 0.0634, "step": 1980 }, { "grad_norm": 0.31198346614837646, "learning_rate": 9.405996254291136e-05, "loss": 0.0639, "step": 1990 }, { "grad_norm": 0.3892742693424225, "learning_rate": 9.398155480222474e-05, "loss": 0.0548, "step": 2000 }, { "grad_norm": 0.4362570345401764, "learning_rate": 9.390266608630128e-05, "loss": 0.0711, "step": 2010 }, { "grad_norm": 0.4395233690738678, "learning_rate": 9.38232972578553e-05, "loss": 0.0612, "step": 2020 }, { "grad_norm": 0.45024609565734863, "learning_rate": 9.374344918485164e-05, "loss": 0.0559, "step": 2030 }, { "grad_norm": 0.32473599910736084, "learning_rate": 9.366312274049602e-05, "loss": 0.0628, "step": 2040 }, { "grad_norm": 0.3600308299064636, "learning_rate": 9.358231880322554e-05, "loss": 0.0646, "step": 2050 }, { "grad_norm": 0.430912047624588, "learning_rate": 9.350103825669916e-05, "loss": 0.0602, "step": 2060 }, { "grad_norm": 0.44321903586387634, "learning_rate": 9.341928198978787e-05, "loss": 0.0685, "step": 2070 }, { "grad_norm": 0.5298194289207458, "learning_rate": 9.333705089656512e-05, "loss": 0.068, "step": 2080 }, { "grad_norm": 0.3317927420139313, "learning_rate": 9.325434587629698e-05, "loss": 0.0645, "step": 2090 }, { "grad_norm": 0.5142538547515869, "learning_rate": 9.31711678334323e-05, "loss": 0.0494, "step": 2100 }, { "grad_norm": 0.4302539825439453, "learning_rate": 9.308751767759282e-05, "loss": 0.0757, "step": 2110 }, { "grad_norm": 0.45203739404678345, "learning_rate": 9.300339632356325e-05, "loss": 0.0723, "step": 2120 }, { "grad_norm": 0.49534091353416443, "learning_rate": 9.291880469128124e-05, "loss": 0.061, "step": 2130 }, { "grad_norm": 0.3361993730068207, "learning_rate": 9.283374370582732e-05, "loss": 0.05, "step": 2140 }, { "grad_norm": 0.30240070819854736, "learning_rate": 9.274821429741482e-05, "loss": 0.0553, "step": 2150 }, { "grad_norm": 0.3601803481578827, "learning_rate": 9.266221740137961e-05, "loss": 0.0605, "step": 2160 }, { "grad_norm": 0.27230405807495117, "learning_rate": 9.257575395817001e-05, "loss": 0.0523, "step": 2170 }, { "grad_norm": 0.43246030807495117, "learning_rate": 9.248882491333637e-05, "loss": 0.0719, "step": 2180 }, { "grad_norm": 0.4542589485645294, "learning_rate": 9.240143121752076e-05, "loss": 0.0665, "step": 2190 }, { "grad_norm": 0.4290499985218048, "learning_rate": 9.23135738264467e-05, "loss": 0.0604, "step": 2200 }, { "grad_norm": 0.42102301120758057, "learning_rate": 9.222525370090849e-05, "loss": 0.0753, "step": 2210 }, { "grad_norm": 0.3329409956932068, "learning_rate": 9.213647180676088e-05, "loss": 0.0655, "step": 2220 }, { "grad_norm": 0.3311687409877777, "learning_rate": 9.204722911490846e-05, "loss": 0.0675, "step": 2230 }, { "grad_norm": 0.27192556858062744, "learning_rate": 9.1957526601295e-05, "loss": 0.0483, "step": 2240 }, { "grad_norm": 0.4126991629600525, "learning_rate": 9.186736524689281e-05, "loss": 0.0661, "step": 2250 }, { "grad_norm": 0.3393888771533966, "learning_rate": 9.177674603769204e-05, "loss": 0.0568, "step": 2260 }, { "grad_norm": 0.22622670233249664, "learning_rate": 9.168566996468983e-05, "loss": 0.0451, "step": 2270 }, { "grad_norm": 0.4101444482803345, "learning_rate": 9.159413802387951e-05, "loss": 0.0529, "step": 2280 }, { "grad_norm": 0.4291892647743225, "learning_rate": 9.150215121623974e-05, "loss": 0.0692, "step": 2290 }, { "grad_norm": 0.37264513969421387, "learning_rate": 9.140971054772349e-05, "loss": 0.0554, "step": 2300 }, { "grad_norm": 0.3080578148365021, "learning_rate": 9.131681702924713e-05, "loss": 0.0506, "step": 2310 }, { "grad_norm": 0.3260904550552368, "learning_rate": 9.122347167667926e-05, "loss": 0.0493, "step": 2320 }, { "grad_norm": 0.3017251789569855, "learning_rate": 9.112967551082973e-05, "loss": 0.0613, "step": 2330 }, { "grad_norm": 0.3142363131046295, "learning_rate": 9.103542955743835e-05, "loss": 0.0584, "step": 2340 }, { "grad_norm": 0.28809356689453125, "learning_rate": 9.094073484716381e-05, "loss": 0.0519, "step": 2350 }, { "grad_norm": 0.39731451869010925, "learning_rate": 9.084559241557226e-05, "loss": 0.0601, "step": 2360 }, { "grad_norm": 0.6038907766342163, "learning_rate": 9.075000330312608e-05, "loss": 0.0783, "step": 2370 }, { "grad_norm": 0.41357478499412537, "learning_rate": 9.065396855517253e-05, "loss": 0.0799, "step": 2380 }, { "grad_norm": 0.53999924659729, "learning_rate": 9.055748922193219e-05, "loss": 0.0664, "step": 2390 }, { "grad_norm": 0.46934521198272705, "learning_rate": 9.046056635848761e-05, "loss": 0.0528, "step": 2400 }, { "grad_norm": 0.3256746530532837, "learning_rate": 9.036320102477169e-05, "loss": 0.0575, "step": 2410 }, { "grad_norm": 0.4976422190666199, "learning_rate": 9.02653942855561e-05, "loss": 0.0544, "step": 2420 }, { "grad_norm": 0.34778955578804016, "learning_rate": 9.016714721043971e-05, "loss": 0.0446, "step": 2430 }, { "grad_norm": 0.28631681203842163, "learning_rate": 9.006846087383675e-05, "loss": 0.0499, "step": 2440 }, { "grad_norm": 0.27976754307746887, "learning_rate": 8.996933635496523e-05, "loss": 0.052, "step": 2450 }, { "grad_norm": 0.5386269688606262, "learning_rate": 8.986977473783498e-05, "loss": 0.0657, "step": 2460 }, { "grad_norm": 0.43381503224372864, "learning_rate": 8.97697771112359e-05, "loss": 0.0612, "step": 2470 }, { "grad_norm": 0.3160889744758606, "learning_rate": 8.966934456872602e-05, "loss": 0.0578, "step": 2480 }, { "grad_norm": 0.46731552481651306, "learning_rate": 8.95684782086195e-05, "loss": 0.0566, "step": 2490 }, { "grad_norm": 0.2606413662433624, "learning_rate": 8.946717913397476e-05, "loss": 0.0662, "step": 2500 }, { "grad_norm": 0.5511654019355774, "learning_rate": 8.93654484525822e-05, "loss": 0.0553, "step": 2510 }, { "grad_norm": 0.37461772561073303, "learning_rate": 8.926328727695226e-05, "loss": 0.0467, "step": 2520 }, { "grad_norm": 0.4292597770690918, "learning_rate": 8.916069672430319e-05, "loss": 0.0535, "step": 2530 }, { "grad_norm": 0.30002355575561523, "learning_rate": 8.905767791654884e-05, "loss": 0.0537, "step": 2540 }, { "grad_norm": 0.6624168753623962, "learning_rate": 8.895423198028638e-05, "loss": 0.0646, "step": 2550 }, { "grad_norm": 0.27232038974761963, "learning_rate": 8.885036004678402e-05, "loss": 0.0651, "step": 2560 }, { "grad_norm": 0.5404578447341919, "learning_rate": 8.874606325196857e-05, "loss": 0.062, "step": 2570 }, { "grad_norm": 0.3968512713909149, "learning_rate": 8.864134273641304e-05, "loss": 0.0543, "step": 2580 }, { "grad_norm": 0.30398818850517273, "learning_rate": 8.853619964532427e-05, "loss": 0.0599, "step": 2590 }, { "grad_norm": 0.43291252851486206, "learning_rate": 8.843063512853019e-05, "loss": 0.0618, "step": 2600 }, { "grad_norm": 0.31835436820983887, "learning_rate": 8.832465034046749e-05, "loss": 0.0591, "step": 2610 }, { "grad_norm": 0.33442452549934387, "learning_rate": 8.821824644016882e-05, "loss": 0.0585, "step": 2620 }, { "grad_norm": 0.3564557135105133, "learning_rate": 8.811142459125019e-05, "loss": 0.0613, "step": 2630 }, { "grad_norm": 0.3679230511188507, "learning_rate": 8.800418596189822e-05, "loss": 0.048, "step": 2640 }, { "grad_norm": 0.2966712415218353, "learning_rate": 8.789653172485737e-05, "loss": 0.0484, "step": 2650 }, { "grad_norm": 0.26649269461631775, "learning_rate": 8.778846305741715e-05, "loss": 0.0665, "step": 2660 }, { "grad_norm": 0.28444889187812805, "learning_rate": 8.767998114139918e-05, "loss": 0.0476, "step": 2670 }, { "grad_norm": 0.25331106781959534, "learning_rate": 8.757108716314429e-05, "loss": 0.0466, "step": 2680 }, { "grad_norm": 0.3416052758693695, "learning_rate": 8.746178231349962e-05, "loss": 0.0523, "step": 2690 }, { "grad_norm": 0.3273114860057831, "learning_rate": 8.735206778780549e-05, "loss": 0.0488, "step": 2700 }, { "grad_norm": 0.3353601396083832, "learning_rate": 8.724194478588234e-05, "loss": 0.0552, "step": 2710 }, { "grad_norm": 0.2853771150112152, "learning_rate": 8.713141451201772e-05, "loss": 0.0599, "step": 2720 }, { "grad_norm": 0.8858267664909363, "learning_rate": 8.702047817495295e-05, "loss": 0.0514, "step": 2730 }, { "grad_norm": 0.39907389879226685, "learning_rate": 8.69091369878701e-05, "loss": 0.0601, "step": 2740 }, { "grad_norm": 0.3272283971309662, "learning_rate": 8.679739216837849e-05, "loss": 0.0586, "step": 2750 }, { "grad_norm": 0.3093755543231964, "learning_rate": 8.66852449385016e-05, "loss": 0.0573, "step": 2760 }, { "grad_norm": 0.5089357495307922, "learning_rate": 8.657269652466356e-05, "loss": 0.0637, "step": 2770 }, { "grad_norm": 0.6865484118461609, "learning_rate": 8.645974815767577e-05, "loss": 0.0691, "step": 2780 }, { "grad_norm": 0.33544138073921204, "learning_rate": 8.634640107272351e-05, "loss": 0.0596, "step": 2790 }, { "grad_norm": 0.3245137929916382, "learning_rate": 8.623265650935234e-05, "loss": 0.0575, "step": 2800 }, { "grad_norm": 0.624232292175293, "learning_rate": 8.611851571145456e-05, "loss": 0.0729, "step": 2810 }, { "grad_norm": 0.3308275640010834, "learning_rate": 8.600397992725566e-05, "loss": 0.0565, "step": 2820 }, { "grad_norm": 0.3374408483505249, "learning_rate": 8.588905040930061e-05, "loss": 0.0575, "step": 2830 }, { "grad_norm": 0.37502506375312805, "learning_rate": 8.577372841444022e-05, "loss": 0.0647, "step": 2840 }, { "grad_norm": 0.46550050377845764, "learning_rate": 8.565801520381736e-05, "loss": 0.0515, "step": 2850 }, { "grad_norm": 0.49899986386299133, "learning_rate": 8.554191204285313e-05, "loss": 0.044, "step": 2860 }, { "grad_norm": 0.4710383415222168, "learning_rate": 8.542542020123315e-05, "loss": 0.0448, "step": 2870 }, { "grad_norm": 0.40419644117355347, "learning_rate": 8.530854095289347e-05, "loss": 0.0608, "step": 2880 }, { "grad_norm": 0.4625893235206604, "learning_rate": 8.519127557600688e-05, "loss": 0.0611, "step": 2890 }, { "grad_norm": 0.3422441780567169, "learning_rate": 8.507362535296871e-05, "loss": 0.0576, "step": 2900 }, { "grad_norm": 0.4221443235874176, "learning_rate": 8.495559157038299e-05, "loss": 0.0698, "step": 2910 }, { "grad_norm": 0.4001758396625519, "learning_rate": 8.483717551904823e-05, "loss": 0.0558, "step": 2920 }, { "grad_norm": 0.27740639448165894, "learning_rate": 8.47183784939434e-05, "loss": 0.0593, "step": 2930 }, { "grad_norm": 0.296661376953125, "learning_rate": 8.459920179421374e-05, "loss": 0.0467, "step": 2940 }, { "grad_norm": 0.27482205629348755, "learning_rate": 8.447964672315656e-05, "loss": 0.0399, "step": 2950 }, { "grad_norm": 0.31482210755348206, "learning_rate": 8.435971458820692e-05, "loss": 0.062, "step": 2960 }, { "grad_norm": 0.3010026812553406, "learning_rate": 8.423940670092345e-05, "loss": 0.046, "step": 2970 }, { "grad_norm": 0.3973766565322876, "learning_rate": 8.411872437697394e-05, "loss": 0.0424, "step": 2980 }, { "grad_norm": 0.3402259945869446, "learning_rate": 8.399766893612096e-05, "loss": 0.0483, "step": 2990 }, { "grad_norm": 0.4289913475513458, "learning_rate": 8.38762417022074e-05, "loss": 0.0423, "step": 3000 }, { "grad_norm": 0.35315629839897156, "learning_rate": 8.375444400314204e-05, "loss": 0.0466, "step": 3010 }, { "grad_norm": 0.3035411536693573, "learning_rate": 8.3632277170885e-05, "loss": 0.0428, "step": 3020 }, { "grad_norm": 0.32245349884033203, "learning_rate": 8.350974254143318e-05, "loss": 0.0449, "step": 3030 }, { "grad_norm": 0.34269288182258606, "learning_rate": 8.338684145480566e-05, "loss": 0.0448, "step": 3040 }, { "grad_norm": 0.30882328748703003, "learning_rate": 8.326357525502904e-05, "loss": 0.0459, "step": 3050 }, { "grad_norm": 0.45904475450515747, "learning_rate": 8.313994529012273e-05, "loss": 0.0403, "step": 3060 }, { "grad_norm": 0.38897281885147095, "learning_rate": 8.301595291208422e-05, "loss": 0.0566, "step": 3070 }, { "grad_norm": 0.4331461787223816, "learning_rate": 8.289159947687427e-05, "loss": 0.0458, "step": 3080 }, { "grad_norm": 0.3855319321155548, "learning_rate": 8.276688634440216e-05, "loss": 0.0732, "step": 3090 }, { "grad_norm": 0.2934146523475647, "learning_rate": 8.26418148785107e-05, "loss": 0.0553, "step": 3100 }, { "grad_norm": 0.2760293483734131, "learning_rate": 8.251638644696141e-05, "loss": 0.0618, "step": 3110 }, { "grad_norm": 0.3949689567089081, "learning_rate": 8.23906024214195e-05, "loss": 0.0548, "step": 3120 }, { "grad_norm": 0.32500118017196655, "learning_rate": 8.226446417743897e-05, "loss": 0.0464, "step": 3130 }, { "grad_norm": 0.29904112219810486, "learning_rate": 8.213797309444742e-05, "loss": 0.0521, "step": 3140 }, { "grad_norm": 0.3249894976615906, "learning_rate": 8.201113055573105e-05, "loss": 0.0463, "step": 3150 }, { "grad_norm": 0.30189311504364014, "learning_rate": 8.188393794841958e-05, "loss": 0.0451, "step": 3160 }, { "grad_norm": 0.2547624707221985, "learning_rate": 8.175639666347094e-05, "loss": 0.0413, "step": 3170 }, { "grad_norm": 0.45044976472854614, "learning_rate": 8.162850809565623e-05, "loss": 0.0623, "step": 3180 }, { "grad_norm": 0.36835864186286926, "learning_rate": 8.150027364354431e-05, "loss": 0.0481, "step": 3190 }, { "grad_norm": 0.2573576867580414, "learning_rate": 8.137169470948662e-05, "loss": 0.0517, "step": 3200 }, { "grad_norm": 0.2997172176837921, "learning_rate": 8.124277269960179e-05, "loss": 0.0413, "step": 3210 }, { "grad_norm": 0.39065656065940857, "learning_rate": 8.111350902376023e-05, "loss": 0.0482, "step": 3220 }, { "grad_norm": 0.36203548312187195, "learning_rate": 8.098390509556883e-05, "loss": 0.0497, "step": 3230 }, { "grad_norm": 0.36177006363868713, "learning_rate": 8.085396233235536e-05, "loss": 0.0505, "step": 3240 }, { "grad_norm": 0.20334990322589874, "learning_rate": 8.072368215515306e-05, "loss": 0.05, "step": 3250 }, { "grad_norm": 0.34968677163124084, "learning_rate": 8.059306598868506e-05, "loss": 0.0426, "step": 3260 }, { "grad_norm": 0.4203032851219177, "learning_rate": 8.046211526134888e-05, "loss": 0.0441, "step": 3270 }, { "grad_norm": 0.28516480326652527, "learning_rate": 8.033083140520065e-05, "loss": 0.0394, "step": 3280 }, { "grad_norm": 0.43890759348869324, "learning_rate": 8.019921585593962e-05, "loss": 0.047, "step": 3290 }, { "grad_norm": 0.2784634530544281, "learning_rate": 8.006727005289232e-05, "loss": 0.0527, "step": 3300 }, { "grad_norm": 0.3048313558101654, "learning_rate": 7.993499543899692e-05, "loss": 0.0604, "step": 3310 }, { "grad_norm": 0.31584277749061584, "learning_rate": 7.980239346078742e-05, "loss": 0.0507, "step": 3320 }, { "grad_norm": 0.4363664984703064, "learning_rate": 7.966946556837778e-05, "loss": 0.0545, "step": 3330 }, { "grad_norm": 0.366855263710022, "learning_rate": 7.953621321544616e-05, "loss": 0.0548, "step": 3340 }, { "grad_norm": 0.6630612015724182, "learning_rate": 7.940263785921896e-05, "loss": 0.0489, "step": 3350 }, { "grad_norm": 0.3984606862068176, "learning_rate": 7.926874096045482e-05, "loss": 0.0518, "step": 3360 }, { "grad_norm": 0.35021379590034485, "learning_rate": 7.913452398342881e-05, "loss": 0.0414, "step": 3370 }, { "grad_norm": 0.5306863188743591, "learning_rate": 7.89999883959163e-05, "loss": 0.0707, "step": 3380 }, { "grad_norm": 0.40779954195022583, "learning_rate": 7.886513566917687e-05, "loss": 0.0559, "step": 3390 }, { "grad_norm": 0.3225768208503723, "learning_rate": 7.872996727793838e-05, "loss": 0.0596, "step": 3400 }, { "grad_norm": 0.30012157559394836, "learning_rate": 7.859448470038069e-05, "loss": 0.0453, "step": 3410 }, { "grad_norm": 0.2541743516921997, "learning_rate": 7.845868941811956e-05, "loss": 0.0496, "step": 3420 }, { "grad_norm": 0.4318830668926239, "learning_rate": 7.832258291619043e-05, "loss": 0.0447, "step": 3430 }, { "grad_norm": 0.8927193880081177, "learning_rate": 7.81861666830322e-05, "loss": 0.0518, "step": 3440 }, { "grad_norm": 0.2713899314403534, "learning_rate": 7.804944221047097e-05, "loss": 0.0503, "step": 3450 }, { "grad_norm": 0.39461761713027954, "learning_rate": 7.791241099370364e-05, "loss": 0.0514, "step": 3460 }, { "grad_norm": 0.5582200884819031, "learning_rate": 7.777507453128163e-05, "loss": 0.0595, "step": 3470 }, { "grad_norm": 0.3216250538825989, "learning_rate": 7.763743432509451e-05, "loss": 0.0497, "step": 3480 }, { "grad_norm": 0.17822667956352234, "learning_rate": 7.749949188035353e-05, "loss": 0.0491, "step": 3490 }, { "grad_norm": 0.24286402761936188, "learning_rate": 7.736124870557516e-05, "loss": 0.0456, "step": 3500 }, { "grad_norm": 0.28441107273101807, "learning_rate": 7.722270631256459e-05, "loss": 0.0552, "step": 3510 }, { "grad_norm": 0.4222944378852844, "learning_rate": 7.708386621639925e-05, "loss": 0.0539, "step": 3520 }, { "grad_norm": 0.46774202585220337, "learning_rate": 7.694472993541219e-05, "loss": 0.0533, "step": 3530 }, { "grad_norm": 0.30795472860336304, "learning_rate": 7.680529899117547e-05, "loss": 0.0483, "step": 3540 }, { "grad_norm": 0.324521541595459, "learning_rate": 7.666557490848358e-05, "loss": 0.0451, "step": 3550 }, { "grad_norm": 0.44083595275878906, "learning_rate": 7.65255592153367e-05, "loss": 0.0434, "step": 3560 }, { "grad_norm": 0.365691214799881, "learning_rate": 7.638525344292402e-05, "loss": 0.0443, "step": 3570 }, { "grad_norm": 0.6039958000183105, "learning_rate": 7.624465912560697e-05, "loss": 0.0483, "step": 3580 }, { "grad_norm": 0.3660784065723419, "learning_rate": 7.610377780090249e-05, "loss": 0.0673, "step": 3590 }, { "grad_norm": 0.2858947813510895, "learning_rate": 7.596261100946618e-05, "loss": 0.0513, "step": 3600 }, { "grad_norm": 0.27577343583106995, "learning_rate": 7.582116029507542e-05, "loss": 0.0429, "step": 3610 }, { "grad_norm": 0.2801256477832794, "learning_rate": 7.56794272046126e-05, "loss": 0.0418, "step": 3620 }, { "grad_norm": 0.38995081186294556, "learning_rate": 7.55374132880481e-05, "loss": 0.0371, "step": 3630 }, { "grad_norm": 0.7078235149383545, "learning_rate": 7.539512009842333e-05, "loss": 0.0532, "step": 3640 }, { "grad_norm": 0.2438284307718277, "learning_rate": 7.525254919183382e-05, "loss": 0.0547, "step": 3650 }, { "grad_norm": 0.6589510440826416, "learning_rate": 7.510970212741215e-05, "loss": 0.0539, "step": 3660 }, { "grad_norm": 0.40066996216773987, "learning_rate": 7.496658046731096e-05, "loss": 0.0569, "step": 3670 }, { "grad_norm": 0.30550703406333923, "learning_rate": 7.482318577668578e-05, "loss": 0.0433, "step": 3680 }, { "grad_norm": 0.2855319082736969, "learning_rate": 7.467951962367796e-05, "loss": 0.0347, "step": 3690 }, { "grad_norm": 0.3086870610713959, "learning_rate": 7.453558357939755e-05, "loss": 0.042, "step": 3700 }, { "grad_norm": 0.3489810526371002, "learning_rate": 7.439137921790606e-05, "loss": 0.0428, "step": 3710 }, { "grad_norm": 0.22950643301010132, "learning_rate": 7.42469081161993e-05, "loss": 0.0431, "step": 3720 }, { "grad_norm": 0.39428192377090454, "learning_rate": 7.410217185419006e-05, "loss": 0.0576, "step": 3730 }, { "grad_norm": 0.41695839166641235, "learning_rate": 7.395717201469095e-05, "loss": 0.0486, "step": 3740 }, { "grad_norm": 0.3005598187446594, "learning_rate": 7.381191018339696e-05, "loss": 0.0478, "step": 3750 }, { "grad_norm": 0.3312225937843323, "learning_rate": 7.36663879488682e-05, "loss": 0.0364, "step": 3760 }, { "grad_norm": 0.2851124703884125, "learning_rate": 7.352060690251254e-05, "loss": 0.0429, "step": 3770 }, { "grad_norm": 0.25727149844169617, "learning_rate": 7.337456863856811e-05, "loss": 0.0406, "step": 3780 }, { "grad_norm": 0.45772242546081543, "learning_rate": 7.3228274754086e-05, "loss": 0.0442, "step": 3790 }, { "grad_norm": 0.6611658334732056, "learning_rate": 7.308172684891267e-05, "loss": 0.068, "step": 3800 }, { "grad_norm": 0.23518788814544678, "learning_rate": 7.293492652567255e-05, "loss": 0.0452, "step": 3810 }, { "grad_norm": 0.3017369210720062, "learning_rate": 7.278787538975043e-05, "loss": 0.0544, "step": 3820 }, { "grad_norm": 0.31179139018058777, "learning_rate": 7.2640575049274e-05, "loss": 0.0446, "step": 3830 }, { "grad_norm": 0.33609893918037415, "learning_rate": 7.249302711509616e-05, "loss": 0.049, "step": 3840 }, { "grad_norm": 0.3160516917705536, "learning_rate": 7.23452332007775e-05, "loss": 0.0574, "step": 3850 }, { "grad_norm": 0.4546661972999573, "learning_rate": 7.219719492256858e-05, "loss": 0.0423, "step": 3860 }, { "grad_norm": 0.35542193055152893, "learning_rate": 7.20489138993923e-05, "loss": 0.0474, "step": 3870 }, { "grad_norm": 0.18799087405204773, "learning_rate": 7.190039175282614e-05, "loss": 0.0401, "step": 3880 }, { "grad_norm": 0.389781653881073, "learning_rate": 7.175163010708455e-05, "loss": 0.0445, "step": 3890 }, { "grad_norm": 0.23875433206558228, "learning_rate": 7.1602630589001e-05, "loss": 0.0476, "step": 3900 }, { "grad_norm": 0.4060862958431244, "learning_rate": 7.14533948280104e-05, "loss": 0.0414, "step": 3910 }, { "grad_norm": 0.6040109395980835, "learning_rate": 7.130392445613109e-05, "loss": 0.057, "step": 3920 }, { "grad_norm": 0.2816777527332306, "learning_rate": 7.115422110794711e-05, "loss": 0.0465, "step": 3930 }, { "grad_norm": 0.2720162868499756, "learning_rate": 7.100428642059033e-05, "loss": 0.0514, "step": 3940 }, { "grad_norm": 0.27545544505119324, "learning_rate": 7.08541220337224e-05, "loss": 0.0484, "step": 3950 }, { "grad_norm": 0.2656237781047821, "learning_rate": 7.070372958951706e-05, "loss": 0.0575, "step": 3960 }, { "grad_norm": 0.2884605824947357, "learning_rate": 7.055311073264194e-05, "loss": 0.0402, "step": 3970 }, { "grad_norm": 0.18910518288612366, "learning_rate": 7.040226711024077e-05, "loss": 0.0364, "step": 3980 }, { "grad_norm": 0.28063857555389404, "learning_rate": 7.02512003719152e-05, "loss": 0.0487, "step": 3990 }, { "grad_norm": 0.22612549364566803, "learning_rate": 7.00999121697069e-05, "loss": 0.0368, "step": 4000 }, { "grad_norm": 0.4328576326370239, "learning_rate": 6.99484041580794e-05, "loss": 0.052, "step": 4010 }, { "grad_norm": 0.36760014295578003, "learning_rate": 6.979667799390004e-05, "loss": 0.0504, "step": 4020 }, { "grad_norm": 0.7016072869300842, "learning_rate": 6.964473533642185e-05, "loss": 0.0428, "step": 4030 }, { "grad_norm": 0.3831039071083069, "learning_rate": 6.949257784726539e-05, "loss": 0.0419, "step": 4040 }, { "grad_norm": 0.3123524785041809, "learning_rate": 6.934020719040056e-05, "loss": 0.0469, "step": 4050 }, { "grad_norm": 0.3288634121417999, "learning_rate": 6.918762503212848e-05, "loss": 0.044, "step": 4060 }, { "grad_norm": 0.4789193868637085, "learning_rate": 6.903483304106319e-05, "loss": 0.0443, "step": 4070 }, { "grad_norm": 0.6809588670730591, "learning_rate": 6.888183288811341e-05, "loss": 0.048, "step": 4080 }, { "grad_norm": 0.29730698466300964, "learning_rate": 6.87286262464643e-05, "loss": 0.0397, "step": 4090 }, { "grad_norm": 0.6700552701950073, "learning_rate": 6.857521479155915e-05, "loss": 0.0357, "step": 4100 }, { "grad_norm": 0.37245431542396545, "learning_rate": 6.842160020108104e-05, "loss": 0.0404, "step": 4110 }, { "grad_norm": 0.38619565963745117, "learning_rate": 6.826778415493455e-05, "loss": 0.0511, "step": 4120 }, { "grad_norm": 0.3734796941280365, "learning_rate": 6.811376833522729e-05, "loss": 0.0428, "step": 4130 }, { "grad_norm": 0.4031915068626404, "learning_rate": 6.795955442625159e-05, "loss": 0.045, "step": 4140 }, { "grad_norm": 0.29657241702079773, "learning_rate": 6.780514411446608e-05, "loss": 0.0508, "step": 4150 }, { "grad_norm": 0.45698150992393494, "learning_rate": 6.765053908847716e-05, "loss": 0.0517, "step": 4160 }, { "grad_norm": 0.30829083919525146, "learning_rate": 6.749574103902064e-05, "loss": 0.0379, "step": 4170 }, { "grad_norm": 0.3815689980983734, "learning_rate": 6.734075165894317e-05, "loss": 0.041, "step": 4180 }, { "grad_norm": 0.3518606126308441, "learning_rate": 6.71855726431838e-05, "loss": 0.0468, "step": 4190 }, { "grad_norm": 0.2483089566230774, "learning_rate": 6.703020568875538e-05, "loss": 0.0404, "step": 4200 }, { "grad_norm": 0.3675616383552551, "learning_rate": 6.687465249472603e-05, "loss": 0.0326, "step": 4210 }, { "grad_norm": 0.31421416997909546, "learning_rate": 6.671891476220055e-05, "loss": 0.0679, "step": 4220 }, { "grad_norm": 0.2858869135379791, "learning_rate": 6.656299419430183e-05, "loss": 0.0512, "step": 4230 }, { "grad_norm": 0.2392619103193283, "learning_rate": 6.640689249615223e-05, "loss": 0.0378, "step": 4240 }, { "grad_norm": 0.32951438426971436, "learning_rate": 6.625061137485491e-05, "loss": 0.0478, "step": 4250 }, { "grad_norm": 0.18623381853103638, "learning_rate": 6.609415253947517e-05, "loss": 0.0411, "step": 4260 }, { "grad_norm": 0.2676140069961548, "learning_rate": 6.593751770102178e-05, "loss": 0.0368, "step": 4270 }, { "grad_norm": 0.2826608717441559, "learning_rate": 6.578070857242823e-05, "loss": 0.0317, "step": 4280 }, { "grad_norm": 0.27231642603874207, "learning_rate": 6.562372686853402e-05, "loss": 0.0325, "step": 4290 }, { "grad_norm": 0.2492620348930359, "learning_rate": 6.546657430606593e-05, "loss": 0.0356, "step": 4300 }, { "grad_norm": 0.3359788954257965, "learning_rate": 6.530925260361918e-05, "loss": 0.0324, "step": 4310 }, { "grad_norm": 0.25713908672332764, "learning_rate": 6.515176348163871e-05, "loss": 0.0551, "step": 4320 }, { "grad_norm": 0.22375008463859558, "learning_rate": 6.499410866240032e-05, "loss": 0.0311, "step": 4330 }, { "grad_norm": 0.5510855317115784, "learning_rate": 6.48362898699919e-05, "loss": 0.0315, "step": 4340 }, { "grad_norm": 0.30023783445358276, "learning_rate": 6.467830883029443e-05, "loss": 0.0457, "step": 4350 }, { "grad_norm": 0.3350319564342499, "learning_rate": 6.452016727096326e-05, "loss": 0.0383, "step": 4360 }, { "grad_norm": 0.249588742852211, "learning_rate": 6.436186692140916e-05, "loss": 0.0391, "step": 4370 }, { "grad_norm": 0.19402426481246948, "learning_rate": 6.420340951277938e-05, "loss": 0.038, "step": 4380 }, { "grad_norm": 0.29260846972465515, "learning_rate": 6.404479677793874e-05, "loss": 0.033, "step": 4390 }, { "grad_norm": 0.4553968608379364, "learning_rate": 6.388603045145075e-05, "loss": 0.0422, "step": 4400 }, { "grad_norm": 0.2660338580608368, "learning_rate": 6.372711226955843e-05, "loss": 0.0379, "step": 4410 }, { "grad_norm": 0.42290446162223816, "learning_rate": 6.356804397016564e-05, "loss": 0.0517, "step": 4420 }, { "grad_norm": 0.3735252320766449, "learning_rate": 6.340882729281779e-05, "loss": 0.0514, "step": 4430 }, { "grad_norm": 0.24494394659996033, "learning_rate": 6.324946397868294e-05, "loss": 0.048, "step": 4440 }, { "grad_norm": 0.3932195007801056, "learning_rate": 6.308995577053276e-05, "loss": 0.0409, "step": 4450 }, { "grad_norm": 0.39515602588653564, "learning_rate": 6.293030441272347e-05, "loss": 0.0403, "step": 4460 }, { "grad_norm": 0.5191615223884583, "learning_rate": 6.277051165117677e-05, "loss": 0.0411, "step": 4470 }, { "grad_norm": 0.21785695850849152, "learning_rate": 6.261057923336064e-05, "loss": 0.0371, "step": 4480 }, { "grad_norm": 0.23777277767658234, "learning_rate": 6.245050890827042e-05, "loss": 0.0385, "step": 4490 }, { "grad_norm": 0.23155346512794495, "learning_rate": 6.229030242640952e-05, "loss": 0.0311, "step": 4500 }, { "grad_norm": 0.23964309692382812, "learning_rate": 6.212996153977037e-05, "loss": 0.0475, "step": 4510 }, { "grad_norm": 0.36289918422698975, "learning_rate": 6.196948800181523e-05, "loss": 0.0352, "step": 4520 }, { "grad_norm": 0.19602380692958832, "learning_rate": 6.180888356745695e-05, "loss": 0.0383, "step": 4530 }, { "grad_norm": 0.28904667496681213, "learning_rate": 6.164814999303995e-05, "loss": 0.0347, "step": 4540 }, { "grad_norm": 0.4491634964942932, "learning_rate": 6.148728903632081e-05, "loss": 0.0418, "step": 4550 }, { "grad_norm": 0.28219833970069885, "learning_rate": 6.132630245644921e-05, "loss": 0.0302, "step": 4560 }, { "grad_norm": 0.2649332880973816, "learning_rate": 6.116519201394857e-05, "loss": 0.0354, "step": 4570 }, { "grad_norm": 0.2773434817790985, "learning_rate": 6.10039594706969e-05, "loss": 0.038, "step": 4580 }, { "grad_norm": 0.3036292791366577, "learning_rate": 6.084260658990744e-05, "loss": 0.0344, "step": 4590 }, { "grad_norm": 0.15181758999824524, "learning_rate": 6.068113513610943e-05, "loss": 0.0373, "step": 4600 }, { "grad_norm": 0.2643829584121704, "learning_rate": 6.0519546875128876e-05, "loss": 0.0406, "step": 4610 }, { "grad_norm": 0.30344170331954956, "learning_rate": 6.035784357406906e-05, "loss": 0.0365, "step": 4620 }, { "grad_norm": 0.38723570108413696, "learning_rate": 6.01960270012914e-05, "loss": 0.0441, "step": 4630 }, { "grad_norm": 0.3940103352069855, "learning_rate": 6.003409892639599e-05, "loss": 0.0321, "step": 4640 }, { "grad_norm": 0.38709867000579834, "learning_rate": 5.9872061120202336e-05, "loss": 0.0317, "step": 4650 }, { "grad_norm": 0.2447054237127304, "learning_rate": 5.9709915354729914e-05, "loss": 0.0276, "step": 4660 }, { "grad_norm": 0.29801687598228455, "learning_rate": 5.9547663403178824e-05, "loss": 0.038, "step": 4670 }, { "grad_norm": 0.2786604166030884, "learning_rate": 5.9385307039910445e-05, "loss": 0.0359, "step": 4680 }, { "grad_norm": 0.21282000839710236, "learning_rate": 5.922284804042792e-05, "loss": 0.0371, "step": 4690 }, { "grad_norm": 0.4068140685558319, "learning_rate": 5.906028818135687e-05, "loss": 0.0432, "step": 4700 }, { "grad_norm": 0.43392395973205566, "learning_rate": 5.889762924042585e-05, "loss": 0.0419, "step": 4710 }, { "grad_norm": 0.2545771896839142, "learning_rate": 5.873487299644699e-05, "loss": 0.0305, "step": 4720 }, { "grad_norm": 0.4044094979763031, "learning_rate": 5.857202122929649e-05, "loss": 0.0381, "step": 4730 }, { "grad_norm": 0.2631467878818512, "learning_rate": 5.840907571989518e-05, "loss": 0.0364, "step": 4740 }, { "grad_norm": 0.18844734132289886, "learning_rate": 5.824603825018904e-05, "loss": 0.0375, "step": 4750 }, { "grad_norm": 0.29656732082366943, "learning_rate": 5.808291060312975e-05, "loss": 0.042, "step": 4760 }, { "grad_norm": 0.23187580704689026, "learning_rate": 5.7919694562655083e-05, "loss": 0.0323, "step": 4770 }, { "grad_norm": 0.32454779744148254, "learning_rate": 5.775639191366954e-05, "loss": 0.0449, "step": 4780 }, { "grad_norm": 0.19900774955749512, "learning_rate": 5.75930044420247e-05, "loss": 0.0321, "step": 4790 }, { "grad_norm": 0.2118074595928192, "learning_rate": 5.74295339344998e-05, "loss": 0.0374, "step": 4800 }, { "grad_norm": 0.36435621976852417, "learning_rate": 5.726598217878211e-05, "loss": 0.0335, "step": 4810 }, { "grad_norm": 0.2857658267021179, "learning_rate": 5.71023509634474e-05, "loss": 0.0364, "step": 4820 }, { "grad_norm": 0.3319779932498932, "learning_rate": 5.693864207794049e-05, "loss": 0.0358, "step": 4830 }, { "grad_norm": 0.2846710681915283, "learning_rate": 5.677485731255545e-05, "loss": 0.04, "step": 4840 }, { "grad_norm": 0.2874966263771057, "learning_rate": 5.6610998458416296e-05, "loss": 0.033, "step": 4850 }, { "grad_norm": 0.19641466438770294, "learning_rate": 5.644706730745716e-05, "loss": 0.0381, "step": 4860 }, { "grad_norm": 0.2433190941810608, "learning_rate": 5.628306565240287e-05, "loss": 0.0382, "step": 4870 }, { "grad_norm": 0.22455795109272003, "learning_rate": 5.611899528674923e-05, "loss": 0.0309, "step": 4880 }, { "grad_norm": 0.2831731140613556, "learning_rate": 5.595485800474349e-05, "loss": 0.0405, "step": 4890 }, { "grad_norm": 0.23509128391742706, "learning_rate": 5.579065560136467e-05, "loss": 0.039, "step": 4900 }, { "grad_norm": 0.2643736004829407, "learning_rate": 5.562638987230392e-05, "loss": 0.0287, "step": 4910 }, { "grad_norm": 0.267294704914093, "learning_rate": 5.546206261394498e-05, "loss": 0.0328, "step": 4920 }, { "grad_norm": 0.2821327745914459, "learning_rate": 5.529767562334437e-05, "loss": 0.0306, "step": 4930 }, { "grad_norm": 0.2922841012477875, "learning_rate": 5.5133230698211926e-05, "loss": 0.037, "step": 4940 }, { "grad_norm": 0.25912240147590637, "learning_rate": 5.496872963689096e-05, "loss": 0.0387, "step": 4950 }, { "grad_norm": 0.3085579574108124, "learning_rate": 5.4804174238338756e-05, "loss": 0.0405, "step": 4960 }, { "grad_norm": 0.22725103795528412, "learning_rate": 5.463956630210678e-05, "loss": 0.0427, "step": 4970 }, { "grad_norm": 0.24448730051517487, "learning_rate": 5.4474907628321046e-05, "loss": 0.0352, "step": 4980 }, { "grad_norm": 0.21927978098392487, "learning_rate": 5.431020001766244e-05, "loss": 0.0313, "step": 4990 }, { "grad_norm": 0.26013532280921936, "learning_rate": 5.4145445271346986e-05, "loss": 0.0495, "step": 5000 }, { "grad_norm": 0.21806970238685608, "learning_rate": 5.398064519110622e-05, "loss": 0.0294, "step": 5010 }, { "grad_norm": 0.2216309905052185, "learning_rate": 5.3815801579167394e-05, "loss": 0.0329, "step": 5020 }, { "grad_norm": 0.46748965978622437, "learning_rate": 5.365091623823382e-05, "loss": 0.0308, "step": 5030 }, { "grad_norm": 0.2756463885307312, "learning_rate": 5.348599097146521e-05, "loss": 0.0407, "step": 5040 }, { "grad_norm": 0.27374759316444397, "learning_rate": 5.3321027582457836e-05, "loss": 0.0376, "step": 5050 }, { "grad_norm": 0.21862487494945526, "learning_rate": 5.315602787522491e-05, "loss": 0.0289, "step": 5060 }, { "grad_norm": 0.2831465005874634, "learning_rate": 5.299099365417678e-05, "loss": 0.0355, "step": 5070 }, { "grad_norm": 0.30860114097595215, "learning_rate": 5.2825926724101236e-05, "loss": 0.0471, "step": 5080 }, { "grad_norm": 0.16491234302520752, "learning_rate": 5.26608288901438e-05, "loss": 0.0387, "step": 5090 }, { "grad_norm": 0.2784714102745056, "learning_rate": 5.24957019577879e-05, "loss": 0.0366, "step": 5100 }, { "grad_norm": 0.25706684589385986, "learning_rate": 5.2330547732835266e-05, "loss": 0.0303, "step": 5110 }, { "grad_norm": 0.27837157249450684, "learning_rate": 5.2165368021385996e-05, "loss": 0.0447, "step": 5120 }, { "grad_norm": 0.3727225959300995, "learning_rate": 5.200016462981897e-05, "loss": 0.0387, "step": 5130 }, { "grad_norm": 0.21520785987377167, "learning_rate": 5.1834939364772015e-05, "loss": 0.0297, "step": 5140 }, { "grad_norm": 0.29367464780807495, "learning_rate": 5.166969403312214e-05, "loss": 0.0318, "step": 5150 }, { "grad_norm": 0.22095492482185364, "learning_rate": 5.1504430441965844e-05, "loss": 0.0386, "step": 5160 }, { "grad_norm": 0.24211366474628448, "learning_rate": 5.133915039859923e-05, "loss": 0.0285, "step": 5170 }, { "grad_norm": 0.20622171461582184, "learning_rate": 5.1173855710498444e-05, "loss": 0.0363, "step": 5180 }, { "grad_norm": 0.25345972180366516, "learning_rate": 5.100854818529967e-05, "loss": 0.0437, "step": 5190 }, { "grad_norm": 0.17755721509456635, "learning_rate": 5.084322963077951e-05, "loss": 0.0319, "step": 5200 }, { "grad_norm": 0.5081574320793152, "learning_rate": 5.067790185483522e-05, "loss": 0.038, "step": 5210 }, { "grad_norm": 0.17916293442249298, "learning_rate": 5.0512566665464844e-05, "loss": 0.03, "step": 5220 }, { "grad_norm": 0.4401004910469055, "learning_rate": 5.034722587074755e-05, "loss": 0.0393, "step": 5230 }, { "grad_norm": 0.3223535716533661, "learning_rate": 5.018188127882375e-05, "loss": 0.033, "step": 5240 }, { "grad_norm": 0.19792114198207855, "learning_rate": 5.0016534697875417e-05, "loss": 0.0286, "step": 5250 }, { "grad_norm": 0.2993556261062622, "learning_rate": 4.9851187936106294e-05, "loss": 0.0259, "step": 5260 }, { "grad_norm": 0.37946489453315735, "learning_rate": 4.968584280172206e-05, "loss": 0.0267, "step": 5270 }, { "grad_norm": 0.2651371359825134, "learning_rate": 4.95205011029106e-05, "loss": 0.0292, "step": 5280 }, { "grad_norm": 0.29063135385513306, "learning_rate": 4.935516464782227e-05, "loss": 0.037, "step": 5290 }, { "grad_norm": 0.2428380846977234, "learning_rate": 4.918983524455003e-05, "loss": 0.0305, "step": 5300 }, { "grad_norm": 0.22239631414413452, "learning_rate": 4.9024514701109766e-05, "loss": 0.0336, "step": 5310 }, { "grad_norm": 0.2198503464460373, "learning_rate": 4.885920482542043e-05, "loss": 0.0287, "step": 5320 }, { "grad_norm": 0.4152858555316925, "learning_rate": 4.869390742528438e-05, "loss": 0.0446, "step": 5330 }, { "grad_norm": 0.28536468744277954, "learning_rate": 4.852862430836744e-05, "loss": 0.0329, "step": 5340 }, { "grad_norm": 0.22123442590236664, "learning_rate": 4.836335728217933e-05, "loss": 0.0323, "step": 5350 }, { "grad_norm": 0.4595975875854492, "learning_rate": 4.819810815405379e-05, "loss": 0.0387, "step": 5360 }, { "grad_norm": 0.40020084381103516, "learning_rate": 4.803287873112877e-05, "loss": 0.0403, "step": 5370 }, { "grad_norm": 0.2148907482624054, "learning_rate": 4.786767082032681e-05, "loss": 0.0246, "step": 5380 }, { "grad_norm": 0.18793365359306335, "learning_rate": 4.77024862283351e-05, "loss": 0.0285, "step": 5390 }, { "grad_norm": 0.20205479860305786, "learning_rate": 4.753732676158593e-05, "loss": 0.0377, "step": 5400 }, { "grad_norm": 0.24851568043231964, "learning_rate": 4.737219422623672e-05, "loss": 0.0263, "step": 5410 }, { "grad_norm": 0.22022035717964172, "learning_rate": 4.720709042815044e-05, "loss": 0.0369, "step": 5420 }, { "grad_norm": 0.29483720660209656, "learning_rate": 4.704201717287578e-05, "loss": 0.0409, "step": 5430 }, { "grad_norm": 0.2774232029914856, "learning_rate": 4.6876976265627404e-05, "loss": 0.0415, "step": 5440 }, { "grad_norm": 0.3933103382587433, "learning_rate": 4.671196951126626e-05, "loss": 0.0389, "step": 5450 }, { "grad_norm": 0.2194719910621643, "learning_rate": 4.654699871427971e-05, "loss": 0.0342, "step": 5460 }, { "grad_norm": 0.39370372891426086, "learning_rate": 4.6382065678762034e-05, "loss": 0.0382, "step": 5470 }, { "grad_norm": 0.2720700204372406, "learning_rate": 4.6217172208394424e-05, "loss": 0.0275, "step": 5480 }, { "grad_norm": 0.25890782475471497, "learning_rate": 4.605232010642549e-05, "loss": 0.0422, "step": 5490 }, { "grad_norm": 0.21099437773227692, "learning_rate": 4.588751117565142e-05, "loss": 0.0261, "step": 5500 }, { "grad_norm": 0.3229185938835144, "learning_rate": 4.5722747218396214e-05, "loss": 0.027, "step": 5510 }, { "grad_norm": 0.30408990383148193, "learning_rate": 4.5558030036492194e-05, "loss": 0.0301, "step": 5520 }, { "grad_norm": 0.2151412069797516, "learning_rate": 4.539336143125999e-05, "loss": 0.0305, "step": 5530 }, { "grad_norm": 0.2549504339694977, "learning_rate": 4.522874320348916e-05, "loss": 0.0354, "step": 5540 }, { "grad_norm": 0.2315414398908615, "learning_rate": 4.506417715341821e-05, "loss": 0.0332, "step": 5550 }, { "grad_norm": 0.3185659348964691, "learning_rate": 4.489966508071511e-05, "loss": 0.0223, "step": 5560 }, { "grad_norm": 0.3004637658596039, "learning_rate": 4.4735208784457575e-05, "loss": 0.0287, "step": 5570 }, { "grad_norm": 0.27106961607933044, "learning_rate": 4.457081006311325e-05, "loss": 0.033, "step": 5580 }, { "grad_norm": 0.27395015954971313, "learning_rate": 4.440647071452027e-05, "loss": 0.033, "step": 5590 }, { "grad_norm": 0.2875633239746094, "learning_rate": 4.424219253586737e-05, "loss": 0.0345, "step": 5600 }, { "grad_norm": 0.3700510561466217, "learning_rate": 4.407797732367443e-05, "loss": 0.0321, "step": 5610 }, { "grad_norm": 0.18107275664806366, "learning_rate": 4.391382687377268e-05, "loss": 0.0389, "step": 5620 }, { "grad_norm": 0.2068130373954773, "learning_rate": 4.374974298128512e-05, "loss": 0.0221, "step": 5630 }, { "grad_norm": 0.1867276281118393, "learning_rate": 4.358572744060699e-05, "loss": 0.0288, "step": 5640 }, { "grad_norm": 0.3557949960231781, "learning_rate": 4.342178204538588e-05, "loss": 0.0415, "step": 5650 }, { "grad_norm": 0.27791720628738403, "learning_rate": 4.325790858850241e-05, "loss": 0.0267, "step": 5660 }, { "grad_norm": 0.3043263554573059, "learning_rate": 4.309410886205043e-05, "loss": 0.0279, "step": 5670 }, { "grad_norm": 0.1915428340435028, "learning_rate": 4.293038465731752e-05, "loss": 0.0255, "step": 5680 }, { "grad_norm": 0.33247771859169006, "learning_rate": 4.276673776476533e-05, "loss": 0.0336, "step": 5690 }, { "grad_norm": 0.2977507412433624, "learning_rate": 4.260316997401007e-05, "loss": 0.0293, "step": 5700 }, { "grad_norm": 0.2913121283054352, "learning_rate": 4.243968307380293e-05, "loss": 0.0288, "step": 5710 }, { "grad_norm": 0.2173949033021927, "learning_rate": 4.22762788520104e-05, "loss": 0.0245, "step": 5720 }, { "grad_norm": 0.24406494200229645, "learning_rate": 4.211295909559491e-05, "loss": 0.0279, "step": 5730 }, { "grad_norm": 0.1619112640619278, "learning_rate": 4.194972559059511e-05, "loss": 0.0272, "step": 5740 }, { "grad_norm": 0.26294293999671936, "learning_rate": 4.178658012210651e-05, "loss": 0.03, "step": 5750 }, { "grad_norm": 0.40339013934135437, "learning_rate": 4.162352447426177e-05, "loss": 0.0365, "step": 5760 }, { "grad_norm": 0.3051551282405853, "learning_rate": 4.146056043021135e-05, "loss": 0.0361, "step": 5770 }, { "grad_norm": 0.23379230499267578, "learning_rate": 4.1297689772103944e-05, "loss": 0.0372, "step": 5780 }, { "grad_norm": 0.2590567171573639, "learning_rate": 4.113491428106694e-05, "loss": 0.0312, "step": 5790 }, { "grad_norm": 0.27747851610183716, "learning_rate": 4.0972235737187055e-05, "loss": 0.0334, "step": 5800 }, { "grad_norm": 0.23636579513549805, "learning_rate": 4.080965591949076e-05, "loss": 0.0436, "step": 5810 }, { "grad_norm": 0.5845754742622375, "learning_rate": 4.0647176605924924e-05, "loss": 0.0325, "step": 5820 }, { "grad_norm": 0.19908304512500763, "learning_rate": 4.0484799573337255e-05, "loss": 0.0265, "step": 5830 }, { "grad_norm": 0.2301468551158905, "learning_rate": 4.032252659745699e-05, "loss": 0.0291, "step": 5840 }, { "grad_norm": 0.19236695766448975, "learning_rate": 4.016035945287539e-05, "loss": 0.0232, "step": 5850 }, { "grad_norm": 0.1407483071088791, "learning_rate": 3.999829991302635e-05, "loss": 0.0277, "step": 5860 }, { "grad_norm": 0.4871128797531128, "learning_rate": 3.983634975016707e-05, "loss": 0.0324, "step": 5870 }, { "grad_norm": 0.1708776205778122, "learning_rate": 3.967451073535854e-05, "loss": 0.0327, "step": 5880 }, { "grad_norm": 0.20979563891887665, "learning_rate": 3.951278463844633e-05, "loss": 0.0274, "step": 5890 }, { "grad_norm": 0.2779957950115204, "learning_rate": 3.935117322804111e-05, "loss": 0.0444, "step": 5900 }, { "grad_norm": 0.3578139543533325, "learning_rate": 3.918967827149938e-05, "loss": 0.0324, "step": 5910 }, { "grad_norm": 0.44276702404022217, "learning_rate": 3.9028301534904094e-05, "loss": 0.0426, "step": 5920 }, { "grad_norm": 0.18046914041042328, "learning_rate": 3.88670447830454e-05, "loss": 0.0264, "step": 5930 }, { "grad_norm": 0.2534346580505371, "learning_rate": 3.870590977940132e-05, "loss": 0.0395, "step": 5940 }, { "grad_norm": 0.18928661942481995, "learning_rate": 3.8544898286118404e-05, "loss": 0.0309, "step": 5950 }, { "grad_norm": 0.3789597153663635, "learning_rate": 3.838401206399257e-05, "loss": 0.0264, "step": 5960 }, { "grad_norm": 0.2720654010772705, "learning_rate": 3.822325287244975e-05, "loss": 0.031, "step": 5970 }, { "grad_norm": 0.18171793222427368, "learning_rate": 3.8062622469526725e-05, "loss": 0.0225, "step": 5980 }, { "grad_norm": 0.2897655963897705, "learning_rate": 3.790212261185183e-05, "loss": 0.0251, "step": 5990 }, { "grad_norm": 0.20128948986530304, "learning_rate": 3.7741755054625794e-05, "loss": 0.043, "step": 6000 }, { "grad_norm": 0.33488544821739197, "learning_rate": 3.758152155160255e-05, "loss": 0.0374, "step": 6010 }, { "grad_norm": 0.335475355386734, "learning_rate": 3.742142385506999e-05, "loss": 0.029, "step": 6020 }, { "grad_norm": 0.1850915402173996, "learning_rate": 3.72614637158309e-05, "loss": 0.0251, "step": 6030 }, { "grad_norm": 0.2532207667827606, "learning_rate": 3.710164288318371e-05, "loss": 0.028, "step": 6040 }, { "grad_norm": 0.2902134358882904, "learning_rate": 3.694196310490345e-05, "loss": 0.0377, "step": 6050 }, { "grad_norm": 0.2942584455013275, "learning_rate": 3.678242612722259e-05, "loss": 0.0318, "step": 6060 }, { "grad_norm": 0.19171607494354248, "learning_rate": 3.6623033694811953e-05, "loss": 0.0265, "step": 6070 }, { "grad_norm": 0.22808094322681427, "learning_rate": 3.6463787550761665e-05, "loss": 0.0292, "step": 6080 }, { "grad_norm": 0.2731122672557831, "learning_rate": 3.630468943656202e-05, "loss": 0.0278, "step": 6090 }, { "grad_norm": 0.2075517475605011, "learning_rate": 3.6145741092084523e-05, "loss": 0.0303, "step": 6100 }, { "grad_norm": 0.19924300909042358, "learning_rate": 3.598694425556278e-05, "loss": 0.0262, "step": 6110 }, { "grad_norm": 0.40640881657600403, "learning_rate": 3.58283006635736e-05, "loss": 0.0358, "step": 6120 }, { "grad_norm": 0.18152761459350586, "learning_rate": 3.566981205101781e-05, "loss": 0.0258, "step": 6130 }, { "grad_norm": 0.26074346899986267, "learning_rate": 3.5511480151101556e-05, "loss": 0.0263, "step": 6140 }, { "grad_norm": 0.3021116852760315, "learning_rate": 3.5353306695317104e-05, "loss": 0.0369, "step": 6150 }, { "grad_norm": 0.23547479510307312, "learning_rate": 3.519529341342402e-05, "loss": 0.0423, "step": 6160 }, { "grad_norm": 0.38080236315727234, "learning_rate": 3.503744203343026e-05, "loss": 0.0326, "step": 6170 }, { "grad_norm": 0.2871115803718567, "learning_rate": 3.487975428157318e-05, "loss": 0.0235, "step": 6180 }, { "grad_norm": 0.3646252453327179, "learning_rate": 3.472223188230083e-05, "loss": 0.0402, "step": 6190 }, { "grad_norm": 0.43427202105522156, "learning_rate": 3.4564876558252866e-05, "loss": 0.0302, "step": 6200 }, { "grad_norm": 0.18523265421390533, "learning_rate": 3.440769003024195e-05, "loss": 0.0313, "step": 6210 }, { "grad_norm": 0.23390622437000275, "learning_rate": 3.425067401723477e-05, "loss": 0.027, "step": 6220 }, { "grad_norm": 0.21183255314826965, "learning_rate": 3.409383023633325e-05, "loss": 0.0235, "step": 6230 }, { "grad_norm": 0.24631860852241516, "learning_rate": 3.3937160402755894e-05, "loss": 0.0259, "step": 6240 }, { "grad_norm": 0.22316347062587738, "learning_rate": 3.378066622981885e-05, "loss": 0.0292, "step": 6250 }, { "grad_norm": 0.28046584129333496, "learning_rate": 3.362434942891738e-05, "loss": 0.0247, "step": 6260 }, { "grad_norm": 0.21045738458633423, "learning_rate": 3.346821170950693e-05, "loss": 0.0235, "step": 6270 }, { "grad_norm": 0.293483167886734, "learning_rate": 3.3312254779084585e-05, "loss": 0.0256, "step": 6280 }, { "grad_norm": 0.26241621375083923, "learning_rate": 3.315648034317039e-05, "loss": 0.027, "step": 6290 }, { "grad_norm": 0.18486903607845306, "learning_rate": 3.3000890105288564e-05, "loss": 0.0344, "step": 6300 }, { "grad_norm": 0.23903749883174896, "learning_rate": 3.284548576694908e-05, "loss": 0.0293, "step": 6310 }, { "grad_norm": 0.24166221916675568, "learning_rate": 3.2690269027628815e-05, "loss": 0.0235, "step": 6320 }, { "grad_norm": 0.22714166343212128, "learning_rate": 3.253524158475324e-05, "loss": 0.0265, "step": 6330 }, { "grad_norm": 0.3007940649986267, "learning_rate": 3.238040513367757e-05, "loss": 0.0505, "step": 6340 }, { "grad_norm": 0.22384606301784515, "learning_rate": 3.222576136766843e-05, "loss": 0.0319, "step": 6350 }, { "grad_norm": 0.18857574462890625, "learning_rate": 3.2071311977885324e-05, "loss": 0.0302, "step": 6360 }, { "grad_norm": 0.23991218209266663, "learning_rate": 3.191705865336197e-05, "loss": 0.0296, "step": 6370 }, { "grad_norm": 0.24769237637519836, "learning_rate": 3.1763003080988075e-05, "loss": 0.0302, "step": 6380 }, { "grad_norm": 0.2303563356399536, "learning_rate": 3.160914694549063e-05, "loss": 0.0303, "step": 6390 }, { "grad_norm": 0.17761586606502533, "learning_rate": 3.145549192941573e-05, "loss": 0.0282, "step": 6400 }, { "grad_norm": 0.19297486543655396, "learning_rate": 3.130203971310999e-05, "loss": 0.0256, "step": 6410 }, { "grad_norm": 0.3301508128643036, "learning_rate": 3.114879197470225e-05, "loss": 0.042, "step": 6420 }, { "grad_norm": 0.2834303677082062, "learning_rate": 3.0995750390085285e-05, "loss": 0.0242, "step": 6430 }, { "grad_norm": 0.22953759133815765, "learning_rate": 3.084291663289728e-05, "loss": 0.0239, "step": 6440 }, { "grad_norm": 0.25911012291908264, "learning_rate": 3.069029237450375e-05, "loss": 0.0296, "step": 6450 }, { "grad_norm": 0.30016112327575684, "learning_rate": 3.053787928397911e-05, "loss": 0.0253, "step": 6460 }, { "grad_norm": 0.4305129051208496, "learning_rate": 3.0385679028088526e-05, "loss": 0.039, "step": 6470 }, { "grad_norm": 0.17147257924079895, "learning_rate": 3.023369327126959e-05, "loss": 0.0232, "step": 6480 }, { "grad_norm": 0.23515070974826813, "learning_rate": 3.0081923675614198e-05, "loss": 0.025, "step": 6490 }, { "grad_norm": 0.30130869150161743, "learning_rate": 2.993037190085034e-05, "loss": 0.0297, "step": 6500 }, { "grad_norm": 0.20307253301143646, "learning_rate": 2.977903960432392e-05, "loss": 0.0251, "step": 6510 }, { "grad_norm": 0.1888335645198822, "learning_rate": 2.9627928440980722e-05, "loss": 0.0266, "step": 6520 }, { "grad_norm": 0.27895158529281616, "learning_rate": 2.9477040063348183e-05, "loss": 0.0264, "step": 6530 }, { "grad_norm": 0.46164700388908386, "learning_rate": 2.9326376121517456e-05, "loss": 0.031, "step": 6540 }, { "grad_norm": 0.2360396534204483, "learning_rate": 2.9175938263125236e-05, "loss": 0.0293, "step": 6550 }, { "grad_norm": 0.21015746891498566, "learning_rate": 2.9025728133335873e-05, "loss": 0.0199, "step": 6560 }, { "grad_norm": 0.38158395886421204, "learning_rate": 2.8875747374823288e-05, "loss": 0.0277, "step": 6570 }, { "grad_norm": 0.23185719549655914, "learning_rate": 2.872599762775298e-05, "loss": 0.0212, "step": 6580 }, { "grad_norm": 0.28961655497550964, "learning_rate": 2.857648052976425e-05, "loss": 0.0255, "step": 6590 }, { "grad_norm": 0.22700928151607513, "learning_rate": 2.8427197715952047e-05, "loss": 0.0305, "step": 6600 }, { "grad_norm": 0.1816316395998001, "learning_rate": 2.8278150818849393e-05, "loss": 0.0257, "step": 6610 }, { "grad_norm": 0.3382924795150757, "learning_rate": 2.812934146840922e-05, "loss": 0.0268, "step": 6620 }, { "grad_norm": 0.22625279426574707, "learning_rate": 2.7980771291986764e-05, "loss": 0.0268, "step": 6630 }, { "grad_norm": 0.2742982804775238, "learning_rate": 2.783244191432167e-05, "loss": 0.0279, "step": 6640 }, { "grad_norm": 0.3318300247192383, "learning_rate": 2.768435495752022e-05, "loss": 0.0316, "step": 6650 }, { "grad_norm": 0.22798657417297363, "learning_rate": 2.753651204103771e-05, "loss": 0.0208, "step": 6660 }, { "grad_norm": 0.2050807625055313, "learning_rate": 2.7388914781660523e-05, "loss": 0.0276, "step": 6670 }, { "grad_norm": 0.3849182724952698, "learning_rate": 2.7241564793488693e-05, "loss": 0.0344, "step": 6680 }, { "grad_norm": 0.16767466068267822, "learning_rate": 2.7094463687918037e-05, "loss": 0.0309, "step": 6690 }, { "grad_norm": 0.27049970626831055, "learning_rate": 2.694761307362268e-05, "loss": 0.0268, "step": 6700 }, { "grad_norm": 0.3123959004878998, "learning_rate": 2.6801014556537467e-05, "loss": 0.0271, "step": 6710 }, { "grad_norm": 0.16415123641490936, "learning_rate": 2.6654669739840243e-05, "loss": 0.0256, "step": 6720 }, { "grad_norm": 0.17787528038024902, "learning_rate": 2.650858022393451e-05, "loss": 0.0376, "step": 6730 }, { "grad_norm": 0.35154667496681213, "learning_rate": 2.6362747606431747e-05, "loss": 0.0396, "step": 6740 }, { "grad_norm": 0.19925552606582642, "learning_rate": 2.6217173482134172e-05, "loss": 0.0286, "step": 6750 }, { "grad_norm": 0.3724563419818878, "learning_rate": 2.6071859443017044e-05, "loss": 0.0227, "step": 6760 }, { "grad_norm": 0.2245984971523285, "learning_rate": 2.5926807078211414e-05, "loss": 0.0204, "step": 6770 }, { "grad_norm": 0.23046311736106873, "learning_rate": 2.5782017973986728e-05, "loss": 0.0239, "step": 6780 }, { "grad_norm": 0.3315809667110443, "learning_rate": 2.5637493713733374e-05, "loss": 0.0251, "step": 6790 }, { "grad_norm": 0.27738234400749207, "learning_rate": 2.549323587794559e-05, "loss": 0.0245, "step": 6800 }, { "grad_norm": 0.30052852630615234, "learning_rate": 2.5349246044203895e-05, "loss": 0.0238, "step": 6810 }, { "grad_norm": 0.3303302824497223, "learning_rate": 2.520552578715808e-05, "loss": 0.0271, "step": 6820 }, { "grad_norm": 0.20158299803733826, "learning_rate": 2.506207667850981e-05, "loss": 0.0187, "step": 6830 }, { "grad_norm": 0.2564351558685303, "learning_rate": 2.4918900286995555e-05, "loss": 0.0241, "step": 6840 }, { "grad_norm": 0.27842944860458374, "learning_rate": 2.4775998178369458e-05, "loss": 0.0281, "step": 6850 }, { "grad_norm": 0.27970221638679504, "learning_rate": 2.4633371915386017e-05, "loss": 0.0256, "step": 6860 }, { "grad_norm": 0.24285854399204254, "learning_rate": 2.4491023057783235e-05, "loss": 0.0368, "step": 6870 }, { "grad_norm": 0.2039748579263687, "learning_rate": 2.4348953162265375e-05, "loss": 0.0314, "step": 6880 }, { "grad_norm": 0.13576377928256989, "learning_rate": 2.420716378248607e-05, "loss": 0.0193, "step": 6890 }, { "grad_norm": 0.23244483768939972, "learning_rate": 2.4065656469031266e-05, "loss": 0.0234, "step": 6900 }, { "grad_norm": 0.3031342625617981, "learning_rate": 2.3924432769402268e-05, "loss": 0.0235, "step": 6910 }, { "grad_norm": 0.26926127076148987, "learning_rate": 2.3783494227998844e-05, "loss": 0.0227, "step": 6920 }, { "grad_norm": 0.28122660517692566, "learning_rate": 2.3642842386102264e-05, "loss": 0.0213, "step": 6930 }, { "grad_norm": 0.21838457882404327, "learning_rate": 2.3502478781858567e-05, "loss": 0.0204, "step": 6940 }, { "grad_norm": 0.2226761281490326, "learning_rate": 2.3362404950261628e-05, "loss": 0.0197, "step": 6950 }, { "grad_norm": 0.25371214747428894, "learning_rate": 2.3222622423136458e-05, "loss": 0.0186, "step": 6960 }, { "grad_norm": 0.1674366295337677, "learning_rate": 2.3083132729122332e-05, "loss": 0.0192, "step": 6970 }, { "grad_norm": 0.16318254172801971, "learning_rate": 2.294393739365621e-05, "loss": 0.0267, "step": 6980 }, { "grad_norm": 0.19804884493350983, "learning_rate": 2.2805037938956e-05, "loss": 0.0205, "step": 6990 }, { "grad_norm": 0.20122772455215454, "learning_rate": 2.266643588400386e-05, "loss": 0.0284, "step": 7000 }, { "grad_norm": 0.22899574041366577, "learning_rate": 2.252813274452969e-05, "loss": 0.0361, "step": 7010 }, { "grad_norm": 0.37030965089797974, "learning_rate": 2.2390130032994427e-05, "loss": 0.0229, "step": 7020 }, { "grad_norm": 0.276943176984787, "learning_rate": 2.2252429258573633e-05, "loss": 0.0293, "step": 7030 }, { "grad_norm": 0.3368491530418396, "learning_rate": 2.2115031927140904e-05, "loss": 0.0214, "step": 7040 }, { "grad_norm": 0.33794718980789185, "learning_rate": 2.1977939541251463e-05, "loss": 0.028, "step": 7050 }, { "grad_norm": 0.2878977060317993, "learning_rate": 2.1841153600125684e-05, "loss": 0.0222, "step": 7060 }, { "grad_norm": 0.17448770999908447, "learning_rate": 2.170467559963267e-05, "loss": 0.0218, "step": 7070 }, { "grad_norm": 0.15542273223400116, "learning_rate": 2.1568507032273982e-05, "loss": 0.0355, "step": 7080 }, { "grad_norm": 0.2354625016450882, "learning_rate": 2.1432649387167264e-05, "loss": 0.0196, "step": 7090 }, { "grad_norm": 0.16816364228725433, "learning_rate": 2.1297104150029973e-05, "loss": 0.0204, "step": 7100 }, { "grad_norm": 0.2640521228313446, "learning_rate": 2.116187280316307e-05, "loss": 0.0228, "step": 7110 }, { "grad_norm": 0.1861780285835266, "learning_rate": 2.1026956825434908e-05, "loss": 0.0175, "step": 7120 }, { "grad_norm": 0.2305373251438141, "learning_rate": 2.0892357692265017e-05, "loss": 0.022, "step": 7130 }, { "grad_norm": 0.31543591618537903, "learning_rate": 2.0758076875607947e-05, "loss": 0.0348, "step": 7140 }, { "grad_norm": 0.17948749661445618, "learning_rate": 2.0624115843937207e-05, "loss": 0.0259, "step": 7150 }, { "grad_norm": 0.28883659839630127, "learning_rate": 2.0490476062229157e-05, "loss": 0.0258, "step": 7160 }, { "grad_norm": 0.1696421205997467, "learning_rate": 2.035715899194704e-05, "loss": 0.0259, "step": 7170 }, { "grad_norm": 0.2641899287700653, "learning_rate": 2.022416609102499e-05, "loss": 0.0196, "step": 7180 }, { "grad_norm": 0.2483811229467392, "learning_rate": 2.009149881385205e-05, "loss": 0.0247, "step": 7190 }, { "grad_norm": 0.30771681666374207, "learning_rate": 1.995915861125634e-05, "loss": 0.0243, "step": 7200 }, { "grad_norm": 0.3011758625507355, "learning_rate": 1.9827146930489065e-05, "loss": 0.0259, "step": 7210 }, { "grad_norm": 0.20340734720230103, "learning_rate": 1.9695465215208848e-05, "loss": 0.0218, "step": 7220 }, { "grad_norm": 0.25753843784332275, "learning_rate": 1.9564114905465813e-05, "loss": 0.027, "step": 7230 }, { "grad_norm": 0.17963698506355286, "learning_rate": 1.9433097437685936e-05, "loss": 0.0186, "step": 7240 }, { "grad_norm": 0.6462975144386292, "learning_rate": 1.930241424465521e-05, "loss": 0.0304, "step": 7250 }, { "grad_norm": 0.17870105803012848, "learning_rate": 1.9172066755504115e-05, "loss": 0.0277, "step": 7260 }, { "grad_norm": 0.6116870045661926, "learning_rate": 1.9042056395691914e-05, "loss": 0.0235, "step": 7270 }, { "grad_norm": 0.27199944853782654, "learning_rate": 1.8912384586991066e-05, "loss": 0.0204, "step": 7280 }, { "grad_norm": 0.21804150938987732, "learning_rate": 1.8783052747471717e-05, "loss": 0.0246, "step": 7290 }, { "grad_norm": 0.4563230276107788, "learning_rate": 1.865406229148611e-05, "loss": 0.0238, "step": 7300 }, { "grad_norm": 0.4072667062282562, "learning_rate": 1.8525414629653233e-05, "loss": 0.0254, "step": 7310 }, { "grad_norm": 0.15672551095485687, "learning_rate": 1.8397111168843255e-05, "loss": 0.0171, "step": 7320 }, { "grad_norm": 0.1302461326122284, "learning_rate": 1.8269153312162323e-05, "loss": 0.0199, "step": 7330 }, { "grad_norm": 0.21455837786197662, "learning_rate": 1.8141542458937054e-05, "loss": 0.0211, "step": 7340 }, { "grad_norm": 0.21119919419288635, "learning_rate": 1.8014280004699268e-05, "loss": 0.0212, "step": 7350 }, { "grad_norm": 0.2465796172618866, "learning_rate": 1.788736734117078e-05, "loss": 0.0182, "step": 7360 }, { "grad_norm": 0.25872689485549927, "learning_rate": 1.7760805856248152e-05, "loss": 0.024, "step": 7370 }, { "grad_norm": 0.22911085188388824, "learning_rate": 1.7634596933987518e-05, "loss": 0.0169, "step": 7380 }, { "grad_norm": 0.4515232741832733, "learning_rate": 1.7508741954589404e-05, "loss": 0.0239, "step": 7390 }, { "grad_norm": 0.19319535791873932, "learning_rate": 1.7383242294383717e-05, "loss": 0.0192, "step": 7400 }, { "grad_norm": 0.17884545028209686, "learning_rate": 1.7258099325814632e-05, "loss": 0.0172, "step": 7410 }, { "grad_norm": 0.2722393274307251, "learning_rate": 1.7133314417425594e-05, "loss": 0.0315, "step": 7420 }, { "grad_norm": 0.25529342889785767, "learning_rate": 1.7008888933844408e-05, "loss": 0.0195, "step": 7430 }, { "grad_norm": 0.1635976880788803, "learning_rate": 1.6884824235768172e-05, "loss": 0.0191, "step": 7440 }, { "grad_norm": 0.1438387781381607, "learning_rate": 1.6761121679948592e-05, "loss": 0.022, "step": 7450 }, { "grad_norm": 0.1406402885913849, "learning_rate": 1.663778261917695e-05, "loss": 0.0189, "step": 7460 }, { "grad_norm": 0.15140388906002045, "learning_rate": 1.651480840226952e-05, "loss": 0.0215, "step": 7470 }, { "grad_norm": 0.35472726821899414, "learning_rate": 1.639220037405258e-05, "loss": 0.0244, "step": 7480 }, { "grad_norm": 0.21781112253665924, "learning_rate": 1.6269959875347906e-05, "loss": 0.0194, "step": 7490 }, { "grad_norm": 0.21820303797721863, "learning_rate": 1.614808824295802e-05, "loss": 0.0194, "step": 7500 }, { "grad_norm": 0.3313082754611969, "learning_rate": 1.602658680965152e-05, "loss": 0.017, "step": 7510 }, { "grad_norm": 0.16093221306800842, "learning_rate": 1.5905456904148686e-05, "loss": 0.0165, "step": 7520 }, { "grad_norm": 0.23664490878582, "learning_rate": 1.57846998511067e-05, "loss": 0.0199, "step": 7530 }, { "grad_norm": 0.1621006727218628, "learning_rate": 1.566431697110538e-05, "loss": 0.0225, "step": 7540 }, { "grad_norm": 0.23100721836090088, "learning_rate": 1.554430958063259e-05, "loss": 0.0176, "step": 7550 }, { "grad_norm": 0.1643836498260498, "learning_rate": 1.5424678992069912e-05, "loss": 0.0228, "step": 7560 }, { "grad_norm": 0.2134196013212204, "learning_rate": 1.5305426513678362e-05, "loss": 0.0191, "step": 7570 }, { "grad_norm": 0.2343544214963913, "learning_rate": 1.518655344958388e-05, "loss": 0.0204, "step": 7580 }, { "grad_norm": 0.26475465297698975, "learning_rate": 1.5068061099763275e-05, "loss": 0.0227, "step": 7590 }, { "grad_norm": 0.23978599905967712, "learning_rate": 1.494995076002988e-05, "loss": 0.0245, "step": 7600 }, { "grad_norm": 0.20001424849033356, "learning_rate": 1.4832223722019456e-05, "loss": 0.0188, "step": 7610 }, { "grad_norm": 0.1758461743593216, "learning_rate": 1.4714881273176035e-05, "loss": 0.0281, "step": 7620 }, { "grad_norm": 0.15663458406925201, "learning_rate": 1.4597924696737835e-05, "loss": 0.0201, "step": 7630 }, { "grad_norm": 0.13410815596580505, "learning_rate": 1.4481355271723252e-05, "loss": 0.0244, "step": 7640 }, { "grad_norm": 0.23470665514469147, "learning_rate": 1.4365174272916809e-05, "loss": 0.0183, "step": 7650 }, { "grad_norm": 0.15982842445373535, "learning_rate": 1.4249382970855319e-05, "loss": 0.0307, "step": 7660 }, { "grad_norm": 0.4767357110977173, "learning_rate": 1.4133982631813903e-05, "loss": 0.0228, "step": 7670 }, { "grad_norm": 0.21625514328479767, "learning_rate": 1.4018974517792194e-05, "loss": 0.0204, "step": 7680 }, { "grad_norm": 0.1709865778684616, "learning_rate": 1.390435988650048e-05, "loss": 0.0203, "step": 7690 }, { "grad_norm": 0.22934754192829132, "learning_rate": 1.3790139991346006e-05, "loss": 0.0182, "step": 7700 }, { "grad_norm": 0.21213029325008392, "learning_rate": 1.367631608141926e-05, "loss": 0.0205, "step": 7710 }, { "grad_norm": 0.2824610769748688, "learning_rate": 1.3562889401480278e-05, "loss": 0.0282, "step": 7720 }, { "grad_norm": 0.15210497379302979, "learning_rate": 1.3449861191945074e-05, "loss": 0.0221, "step": 7730 }, { "grad_norm": 0.31539323925971985, "learning_rate": 1.3337232688872009e-05, "loss": 0.0253, "step": 7740 }, { "grad_norm": 0.14817117154598236, "learning_rate": 1.3225005123948364e-05, "loss": 0.0203, "step": 7750 }, { "grad_norm": 0.2413870245218277, "learning_rate": 1.311317972447681e-05, "loss": 0.0217, "step": 7760 }, { "grad_norm": 0.15943080186843872, "learning_rate": 1.3001757713361996e-05, "loss": 0.0196, "step": 7770 }, { "grad_norm": 0.22361941635608673, "learning_rate": 1.2890740309097204e-05, "loss": 0.0196, "step": 7780 }, { "grad_norm": 0.23159140348434448, "learning_rate": 1.2780128725750944e-05, "loss": 0.0187, "step": 7790 }, { "grad_norm": 0.1692468523979187, "learning_rate": 1.266992417295379e-05, "loss": 0.0208, "step": 7800 }, { "grad_norm": 0.2517269253730774, "learning_rate": 1.2560127855885073e-05, "loss": 0.0212, "step": 7810 }, { "grad_norm": 0.1715618371963501, "learning_rate": 1.2450740975259745e-05, "loss": 0.0192, "step": 7820 }, { "grad_norm": 0.19144675135612488, "learning_rate": 1.234176472731517e-05, "loss": 0.02, "step": 7830 }, { "grad_norm": 0.1983812004327774, "learning_rate": 1.2233200303798158e-05, "loss": 0.019, "step": 7840 }, { "grad_norm": 0.19309787452220917, "learning_rate": 1.2125048891951846e-05, "loss": 0.0187, "step": 7850 }, { "grad_norm": 0.15086416900157928, "learning_rate": 1.2017311674502745e-05, "loss": 0.0194, "step": 7860 }, { "grad_norm": 0.22929087281227112, "learning_rate": 1.1909989829647822e-05, "loss": 0.0169, "step": 7870 }, { "grad_norm": 0.15502966940402985, "learning_rate": 1.1803084531041553e-05, "loss": 0.0214, "step": 7880 }, { "grad_norm": 0.24078314006328583, "learning_rate": 1.1696596947783162e-05, "loss": 0.0212, "step": 7890 }, { "grad_norm": 0.16591036319732666, "learning_rate": 1.1590528244403803e-05, "loss": 0.0283, "step": 7900 }, { "grad_norm": 0.2427297979593277, "learning_rate": 1.148487958085382e-05, "loss": 0.0187, "step": 7910 }, { "grad_norm": 0.25897398591041565, "learning_rate": 1.1379652112490086e-05, "loss": 0.0238, "step": 7920 }, { "grad_norm": 0.4003790020942688, "learning_rate": 1.1274846990063315e-05, "loss": 0.0184, "step": 7930 }, { "grad_norm": 0.23119792342185974, "learning_rate": 1.117046535970554e-05, "loss": 0.0196, "step": 7940 }, { "grad_norm": 0.2545106112957001, "learning_rate": 1.106650836291755e-05, "loss": 0.0267, "step": 7950 }, { "grad_norm": 0.159910649061203, "learning_rate": 1.0962977136556418e-05, "loss": 0.0131, "step": 7960 }, { "grad_norm": 0.4565497636795044, "learning_rate": 1.0859872812823024e-05, "loss": 0.0203, "step": 7970 }, { "grad_norm": 0.4160275161266327, "learning_rate": 1.0757196519249747e-05, "loss": 0.0212, "step": 7980 }, { "grad_norm": 0.14112485945224762, "learning_rate": 1.0654949378688077e-05, "loss": 0.0189, "step": 7990 }, { "grad_norm": 0.2439354509115219, "learning_rate": 1.0553132509296376e-05, "loss": 0.0156, "step": 8000 }, { "grad_norm": 0.1694277822971344, "learning_rate": 1.0451747024527613e-05, "loss": 0.0194, "step": 8010 }, { "grad_norm": 0.11691729724407196, "learning_rate": 1.0350794033117189e-05, "loss": 0.0185, "step": 8020 }, { "grad_norm": 0.16513925790786743, "learning_rate": 1.0250274639070856e-05, "loss": 0.0187, "step": 8030 }, { "grad_norm": 0.20386062562465668, "learning_rate": 1.0150189941652599e-05, "loss": 0.0182, "step": 8040 }, { "grad_norm": 0.17677254974842072, "learning_rate": 1.0050541035372635e-05, "loss": 0.0194, "step": 8050 }, { "grad_norm": 0.18914099037647247, "learning_rate": 9.951329009975458e-06, "loss": 0.0129, "step": 8060 }, { "grad_norm": 0.1724333018064499, "learning_rate": 9.852554950427845e-06, "loss": 0.0161, "step": 8070 }, { "grad_norm": 0.29509150981903076, "learning_rate": 9.754219936907105e-06, "loss": 0.03, "step": 8080 }, { "grad_norm": 0.16798822581768036, "learning_rate": 9.656325044789194e-06, "loss": 0.0198, "step": 8090 }, { "grad_norm": 0.19473522901535034, "learning_rate": 9.55887134463697e-06, "loss": 0.02, "step": 8100 }, { "grad_norm": 0.18005290627479553, "learning_rate": 9.461859902188475e-06, "loss": 0.0251, "step": 8110 }, { "grad_norm": 0.13496847450733185, "learning_rate": 9.365291778345303e-06, "loss": 0.0252, "step": 8120 }, { "grad_norm": 0.528339684009552, "learning_rate": 9.269168029160991e-06, "loss": 0.0413, "step": 8130 }, { "grad_norm": 0.11373227834701538, "learning_rate": 9.173489705829447e-06, "loss": 0.0223, "step": 8140 }, { "grad_norm": 0.16203755140304565, "learning_rate": 9.078257854673516e-06, "loss": 0.0175, "step": 8150 }, { "grad_norm": 0.19423604011535645, "learning_rate": 8.983473517133429e-06, "loss": 0.0185, "step": 8160 }, { "grad_norm": 0.2075670212507248, "learning_rate": 8.889137729755537e-06, "loss": 0.0335, "step": 8170 }, { "grad_norm": 0.2564495801925659, "learning_rate": 8.79525152418087e-06, "loss": 0.0153, "step": 8180 }, { "grad_norm": 0.18873554468154907, "learning_rate": 8.701815927133961e-06, "loss": 0.0215, "step": 8190 }, { "grad_norm": 0.16108211874961853, "learning_rate": 8.608831960411534e-06, "loss": 0.0225, "step": 8200 }, { "grad_norm": 0.14270196855068207, "learning_rate": 8.516300640871321e-06, "loss": 0.0204, "step": 8210 }, { "grad_norm": 0.11156225204467773, "learning_rate": 8.424222980421038e-06, "loss": 0.0183, "step": 8220 }, { "grad_norm": 0.2599072754383087, "learning_rate": 8.332599986007184e-06, "loss": 0.0229, "step": 8230 }, { "grad_norm": 0.23409691452980042, "learning_rate": 8.241432659604203e-06, "loss": 0.0189, "step": 8240 }, { "grad_norm": 0.22522665560245514, "learning_rate": 8.150721998203331e-06, "loss": 0.023, "step": 8250 }, { "grad_norm": 0.1250986009836197, "learning_rate": 8.06046899380184e-06, "loss": 0.0213, "step": 8260 }, { "grad_norm": 0.20645646750926971, "learning_rate": 7.970674633392133e-06, "loss": 0.0195, "step": 8270 }, { "grad_norm": 0.3955833911895752, "learning_rate": 7.881339898950924e-06, "loss": 0.0255, "step": 8280 }, { "grad_norm": 0.2075974941253662, "learning_rate": 7.792465767428597e-06, "loss": 0.018, "step": 8290 }, { "grad_norm": 0.16943125426769257, "learning_rate": 7.704053210738376e-06, "loss": 0.0201, "step": 8300 }, { "grad_norm": 0.20055049657821655, "learning_rate": 7.6161031957458494e-06, "loss": 0.0221, "step": 8310 }, { "grad_norm": 0.1241411417722702, "learning_rate": 7.5286166842582605e-06, "loss": 0.02, "step": 8320 }, { "grad_norm": 0.20695409178733826, "learning_rate": 7.4415946330140814e-06, "loss": 0.0203, "step": 8330 }, { "grad_norm": 0.20760363340377808, "learning_rate": 7.3550379936725644e-06, "loss": 0.0234, "step": 8340 }, { "grad_norm": 0.11204303801059723, "learning_rate": 7.2689477128032035e-06, "loss": 0.0196, "step": 8350 }, { "grad_norm": 0.24513210356235504, "learning_rate": 7.183324731875551e-06, "loss": 0.0163, "step": 8360 }, { "grad_norm": 0.1317310333251953, "learning_rate": 7.098169987248782e-06, "loss": 0.0208, "step": 8370 }, { "grad_norm": 0.124283067882061, "learning_rate": 7.013484410161553e-06, "loss": 0.0235, "step": 8380 }, { "grad_norm": 0.33234789967536926, "learning_rate": 6.92926892672176e-06, "loss": 0.0169, "step": 8390 }, { "grad_norm": 0.14468254148960114, "learning_rate": 6.845524457896446e-06, "loss": 0.0168, "step": 8400 }, { "grad_norm": 0.16299068927764893, "learning_rate": 6.7622519195017165e-06, "loss": 0.0171, "step": 8410 }, { "grad_norm": 0.30965808033943176, "learning_rate": 6.679452222192684e-06, "loss": 0.0184, "step": 8420 }, { "grad_norm": 0.1459573358297348, "learning_rate": 6.597126271453579e-06, "loss": 0.0226, "step": 8430 }, { "grad_norm": 0.1939426064491272, "learning_rate": 6.51527496758782e-06, "loss": 0.0164, "step": 8440 }, { "grad_norm": 0.21373961865901947, "learning_rate": 6.433899205708155e-06, "loss": 0.0164, "step": 8450 }, { "grad_norm": 0.1405775099992752, "learning_rate": 6.352999875726856e-06, "loss": 0.0216, "step": 8460 }, { "grad_norm": 0.1503438502550125, "learning_rate": 6.272577862346052e-06, "loss": 0.0221, "step": 8470 }, { "grad_norm": 0.2544173300266266, "learning_rate": 6.192634045047996e-06, "loss": 0.0251, "step": 8480 }, { "grad_norm": 0.2627699375152588, "learning_rate": 6.113169298085458e-06, "loss": 0.0174, "step": 8490 }, { "grad_norm": 0.2215712070465088, "learning_rate": 6.034184490472195e-06, "loss": 0.0191, "step": 8500 }, { "grad_norm": 0.1559782773256302, "learning_rate": 5.955680485973386e-06, "loss": 0.0164, "step": 8510 }, { "grad_norm": 0.24071262776851654, "learning_rate": 5.877658143096265e-06, "loss": 0.0162, "step": 8520 }, { "grad_norm": 0.11391042172908783, "learning_rate": 5.800118315080661e-06, "loss": 0.0121, "step": 8530 }, { "grad_norm": 0.27424702048301697, "learning_rate": 5.723061849889716e-06, "loss": 0.0174, "step": 8540 }, { "grad_norm": 0.17970024049282074, "learning_rate": 5.646489590200604e-06, "loss": 0.0253, "step": 8550 }, { "grad_norm": 0.13378426432609558, "learning_rate": 5.570402373395256e-06, "loss": 0.015, "step": 8560 }, { "grad_norm": 0.20034243166446686, "learning_rate": 5.494801031551305e-06, "loss": 0.0245, "step": 8570 }, { "grad_norm": 0.16769438982009888, "learning_rate": 5.41968639143291e-06, "loss": 0.0161, "step": 8580 }, { "grad_norm": 0.17584598064422607, "learning_rate": 5.345059274481751e-06, "loss": 0.0222, "step": 8590 }, { "grad_norm": 0.20084048807621002, "learning_rate": 5.270920496808002e-06, "loss": 0.0206, "step": 8600 }, { "grad_norm": 0.2491016536951065, "learning_rate": 5.1972708691814695e-06, "loss": 0.0142, "step": 8610 }, { "grad_norm": 0.1246127262711525, "learning_rate": 5.124111197022674e-06, "loss": 0.0173, "step": 8620 }, { "grad_norm": 0.14954860508441925, "learning_rate": 5.051442280394081e-06, "loss": 0.0139, "step": 8630 }, { "grad_norm": 0.13885726034641266, "learning_rate": 4.979264913991322e-06, "loss": 0.0187, "step": 8640 }, { "grad_norm": 0.1686953902244568, "learning_rate": 4.907579887134489e-06, "loss": 0.0188, "step": 8650 }, { "grad_norm": 0.1913326531648636, "learning_rate": 4.836387983759572e-06, "loss": 0.0162, "step": 8660 }, { "grad_norm": 0.1838378757238388, "learning_rate": 4.765689982409816e-06, "loss": 0.0213, "step": 8670 }, { "grad_norm": 0.21530814468860626, "learning_rate": 4.695486656227233e-06, "loss": 0.015, "step": 8680 }, { "grad_norm": 0.14688540995121002, "learning_rate": 4.625778772944156e-06, "loss": 0.0151, "step": 8690 }, { "grad_norm": 0.08302649110555649, "learning_rate": 4.556567094874825e-06, "loss": 0.0141, "step": 8700 }, { "grad_norm": 0.5314264297485352, "learning_rate": 4.487852378907059e-06, "loss": 0.0282, "step": 8710 }, { "grad_norm": 0.24112309515476227, "learning_rate": 4.419635376493986e-06, "loss": 0.023, "step": 8720 }, { "grad_norm": 0.11274424940347672, "learning_rate": 4.351916833645825e-06, "loss": 0.0184, "step": 8730 }, { "grad_norm": 0.1458204686641693, "learning_rate": 4.284697490921691e-06, "loss": 0.0139, "step": 8740 }, { "grad_norm": 0.1555270403623581, "learning_rate": 4.2179780834215585e-06, "loss": 0.0178, "step": 8750 }, { "grad_norm": 0.13533522188663483, "learning_rate": 4.151759340778178e-06, "loss": 0.0174, "step": 8760 }, { "grad_norm": 0.13748352229595184, "learning_rate": 4.086041987149109e-06, "loss": 0.016, "step": 8770 }, { "grad_norm": 0.13409939408302307, "learning_rate": 4.020826741208811e-06, "loss": 0.0268, "step": 8780 }, { "grad_norm": 0.20482176542282104, "learning_rate": 3.956114316140746e-06, "loss": 0.0215, "step": 8790 }, { "grad_norm": 0.19346670806407928, "learning_rate": 3.891905419629643e-06, "loss": 0.0141, "step": 8800 }, { "grad_norm": 0.1302545964717865, "learning_rate": 3.8282007538536946e-06, "loss": 0.0224, "step": 8810 }, { "grad_norm": 0.13775375485420227, "learning_rate": 3.7650010154769265e-06, "loss": 0.0141, "step": 8820 }, { "grad_norm": 0.21999360620975494, "learning_rate": 3.7023068956415608e-06, "loss": 0.0185, "step": 8830 }, { "grad_norm": 0.1638166308403015, "learning_rate": 3.6401190799604303e-06, "loss": 0.0147, "step": 8840 }, { "grad_norm": 0.23473572731018066, "learning_rate": 3.578438248509536e-06, "loss": 0.017, "step": 8850 }, { "grad_norm": 0.3557720482349396, "learning_rate": 3.5172650758205583e-06, "loss": 0.0163, "step": 8860 }, { "grad_norm": 0.23691177368164062, "learning_rate": 3.45660023087353e-06, "loss": 0.0154, "step": 8870 }, { "grad_norm": 0.19637039303779602, "learning_rate": 3.3964443770894528e-06, "loss": 0.0151, "step": 8880 }, { "grad_norm": 0.10761084407567978, "learning_rate": 3.3367981723231245e-06, "loss": 0.0229, "step": 8890 }, { "grad_norm": 0.13189582526683807, "learning_rate": 3.2776622688558746e-06, "loss": 0.016, "step": 8900 }, { "grad_norm": 0.15529175102710724, "learning_rate": 3.2190373133884677e-06, "loss": 0.0129, "step": 8910 }, { "grad_norm": 0.20802469551563263, "learning_rate": 3.1609239470340446e-06, "loss": 0.0181, "step": 8920 }, { "grad_norm": 0.18191738426685333, "learning_rate": 3.1033228053110373e-06, "loss": 0.0197, "step": 8930 }, { "grad_norm": 0.24461568892002106, "learning_rate": 3.0462345181363314e-06, "loss": 0.021, "step": 8940 }, { "grad_norm": 0.23142533004283905, "learning_rate": 2.9896597098182654e-06, "loss": 0.0174, "step": 8950 }, { "grad_norm": 0.1537734568119049, "learning_rate": 2.933598999049891e-06, "loss": 0.0237, "step": 8960 }, { "grad_norm": 0.15508203208446503, "learning_rate": 2.8780529989021697e-06, "loss": 0.0169, "step": 8970 }, { "grad_norm": 0.1938699632883072, "learning_rate": 2.823022316817242e-06, "loss": 0.021, "step": 8980 }, { "grad_norm": 0.17183320224285126, "learning_rate": 2.7685075546018456e-06, "loss": 0.0261, "step": 8990 }, { "grad_norm": 0.13777856528759003, "learning_rate": 2.7145093084206598e-06, "loss": 0.0153, "step": 9000 }, { "grad_norm": 0.18122753500938416, "learning_rate": 2.661028168789892e-06, "loss": 0.0153, "step": 9010 }, { "grad_norm": 0.2142951786518097, "learning_rate": 2.6080647205706855e-06, "loss": 0.0171, "step": 9020 }, { "grad_norm": 0.2175462543964386, "learning_rate": 2.555619542962834e-06, "loss": 0.0209, "step": 9030 }, { "grad_norm": 0.1220434382557869, "learning_rate": 2.503693209498409e-06, "loss": 0.0184, "step": 9040 }, { "grad_norm": 0.1253872811794281, "learning_rate": 2.452286288035449e-06, "loss": 0.0147, "step": 9050 }, { "grad_norm": 0.1226811408996582, "learning_rate": 2.4013993407518363e-06, "loss": 0.0211, "step": 9060 }, { "grad_norm": 0.14448240399360657, "learning_rate": 2.351032924139063e-06, "loss": 0.0154, "step": 9070 }, { "grad_norm": 0.14516210556030273, "learning_rate": 2.30118758899619e-06, "loss": 0.0155, "step": 9080 }, { "grad_norm": 0.1784232258796692, "learning_rate": 2.2518638804238157e-06, "loss": 0.0191, "step": 9090 }, { "grad_norm": 0.34005358815193176, "learning_rate": 2.203062337818118e-06, "loss": 0.0244, "step": 9100 }, { "grad_norm": 0.1420242041349411, "learning_rate": 2.1547834948649483e-06, "loss": 0.0177, "step": 9110 }, { "grad_norm": 0.18501514196395874, "learning_rate": 2.1070278795340017e-06, "loss": 0.017, "step": 9120 }, { "grad_norm": 0.20835857093334198, "learning_rate": 2.059796014073029e-06, "loss": 0.014, "step": 9130 }, { "grad_norm": 0.1955883800983429, "learning_rate": 2.01308841500214e-06, "loss": 0.0167, "step": 9140 }, { "grad_norm": 0.22925156354904175, "learning_rate": 1.9669055931081704e-06, "loss": 0.0204, "step": 9150 }, { "grad_norm": 0.09967335313558578, "learning_rate": 1.9212480534390507e-06, "loss": 0.0129, "step": 9160 }, { "grad_norm": 0.24186433851718903, "learning_rate": 1.8761162952983246e-06, "loss": 0.0186, "step": 9170 }, { "grad_norm": 0.17105622589588165, "learning_rate": 1.8315108122396618e-06, "loss": 0.0168, "step": 9180 }, { "grad_norm": 0.23244814574718475, "learning_rate": 1.787432092061475e-06, "loss": 0.0161, "step": 9190 }, { "grad_norm": 0.2814110219478607, "learning_rate": 1.743880616801602e-06, "loss": 0.0188, "step": 9200 }, { "grad_norm": 0.14840053021907806, "learning_rate": 1.7008568627319865e-06, "loss": 0.0251, "step": 9210 }, { "grad_norm": 0.4783870577812195, "learning_rate": 1.6583613003535226e-06, "loss": 0.0217, "step": 9220 }, { "grad_norm": 0.1967959851026535, "learning_rate": 1.6163943943908522e-06, "loss": 0.0181, "step": 9230 }, { "grad_norm": 0.1938934475183487, "learning_rate": 1.5749566037873476e-06, "loss": 0.0227, "step": 9240 }, { "grad_norm": 0.13417820632457733, "learning_rate": 1.5340483817000428e-06, "loss": 0.0199, "step": 9250 }, { "grad_norm": 0.10733755677938461, "learning_rate": 1.4936701754947101e-06, "loss": 0.0164, "step": 9260 }, { "grad_norm": 0.11440839618444443, "learning_rate": 1.4538224267409361e-06, "loss": 0.0156, "step": 9270 }, { "grad_norm": 0.11559891700744629, "learning_rate": 1.414505571207314e-06, "loss": 0.0175, "step": 9280 }, { "grad_norm": 0.23322202265262604, "learning_rate": 1.3757200388566816e-06, "loss": 0.0208, "step": 9290 }, { "grad_norm": 0.14128169417381287, "learning_rate": 1.3374662538414074e-06, "loss": 0.0231, "step": 9300 }, { "grad_norm": 0.16274330019950867, "learning_rate": 1.2997446344987617e-06, "loss": 0.014, "step": 9310 }, { "grad_norm": 0.21071688830852509, "learning_rate": 1.262555593346315e-06, "loss": 0.0248, "step": 9320 }, { "grad_norm": 0.17492417991161346, "learning_rate": 1.2258995370774685e-06, "loss": 0.0161, "step": 9330 }, { "grad_norm": 0.3967443108558655, "learning_rate": 1.1897768665569798e-06, "loss": 0.0232, "step": 9340 }, { "grad_norm": 0.20004737377166748, "learning_rate": 1.1541879768165954e-06, "loss": 0.0213, "step": 9350 }, { "grad_norm": 0.14310091733932495, "learning_rate": 1.1191332570507085e-06, "loss": 0.0164, "step": 9360 }, { "grad_norm": 0.263023316860199, "learning_rate": 1.0846130906121132e-06, "loss": 0.0224, "step": 9370 }, { "grad_norm": 0.11879894882440567, "learning_rate": 1.0506278550078131e-06, "loss": 0.0154, "step": 9380 }, { "grad_norm": 0.18597765266895294, "learning_rate": 1.0171779218949185e-06, "loss": 0.0128, "step": 9390 }, { "grad_norm": 0.1382218599319458, "learning_rate": 9.842636570765174e-07, "loss": 0.0189, "step": 9400 }, { "grad_norm": 0.13492447137832642, "learning_rate": 9.518854204977612e-07, "loss": 0.0139, "step": 9410 }, { "grad_norm": 0.20060350000858307, "learning_rate": 9.200435662418349e-07, "loss": 0.0177, "step": 9420 }, { "grad_norm": 0.1585240215063095, "learning_rate": 8.887384425261658e-07, "loss": 0.0129, "step": 9430 }, { "grad_norm": 0.1854269951581955, "learning_rate": 8.579703916985648e-07, "loss": 0.013, "step": 9440 }, { "grad_norm": 0.14933155477046967, "learning_rate": 8.277397502335194e-07, "loss": 0.0194, "step": 9450 }, { "grad_norm": 0.14595569670200348, "learning_rate": 7.980468487284675e-07, "loss": 0.0154, "step": 9460 }, { "grad_norm": 0.16428782045841217, "learning_rate": 7.688920119002297e-07, "loss": 0.0181, "step": 9470 }, { "grad_norm": 0.07394246757030487, "learning_rate": 7.402755585814269e-07, "loss": 0.0136, "step": 9480 }, { "grad_norm": 0.24877406656742096, "learning_rate": 7.121978017170073e-07, "loss": 0.0159, "step": 9490 }, { "grad_norm": 0.21356633305549622, "learning_rate": 6.846590483608306e-07, "loss": 0.0194, "step": 9500 }, { "grad_norm": 0.1482456475496292, "learning_rate": 6.576595996722834e-07, "loss": 0.0163, "step": 9510 }, { "grad_norm": 0.16996891796588898, "learning_rate": 6.311997509130141e-07, "loss": 0.0157, "step": 9520 }, { "grad_norm": 0.1558944582939148, "learning_rate": 6.052797914436803e-07, "loss": 0.0173, "step": 9530 }, { "grad_norm": 0.27410322427749634, "learning_rate": 5.799000047208181e-07, "loss": 0.02, "step": 9540 }, { "grad_norm": 0.10518340021371841, "learning_rate": 5.550606682937054e-07, "loss": 0.012, "step": 9550 }, { "grad_norm": 0.15488716959953308, "learning_rate": 5.307620538013481e-07, "loss": 0.0193, "step": 9560 }, { "grad_norm": 0.1941915899515152, "learning_rate": 5.070044269694874e-07, "loss": 0.0211, "step": 9570 }, { "grad_norm": 0.11499693989753723, "learning_rate": 4.837880476077417e-07, "loss": 0.0188, "step": 9580 }, { "grad_norm": 0.1590038388967514, "learning_rate": 4.6111316960670835e-07, "loss": 0.0133, "step": 9590 }, { "grad_norm": 0.1285114735364914, "learning_rate": 4.389800409352218e-07, "loss": 0.0203, "step": 9600 }, { "grad_norm": 0.12174399942159653, "learning_rate": 4.173889036376277e-07, "loss": 0.0124, "step": 9610 }, { "grad_norm": 0.07428070902824402, "learning_rate": 3.963399938311463e-07, "loss": 0.0186, "step": 9620 }, { "grad_norm": 0.143952876329422, "learning_rate": 3.7583354170328545e-07, "loss": 0.0133, "step": 9630 }, { "grad_norm": 0.15110772848129272, "learning_rate": 3.558697715093207e-07, "loss": 0.0173, "step": 9640 }, { "grad_norm": 0.4599926173686981, "learning_rate": 3.3644890156983576e-07, "loss": 0.0196, "step": 9650 }, { "grad_norm": 0.09601344168186188, "learning_rate": 3.175711442683638e-07, "loss": 0.0144, "step": 9660 }, { "grad_norm": 0.19387409090995789, "learning_rate": 2.9923670604902197e-07, "loss": 0.018, "step": 9670 }, { "grad_norm": 0.13468682765960693, "learning_rate": 2.814457874143028e-07, "loss": 0.0135, "step": 9680 }, { "grad_norm": 0.19535931944847107, "learning_rate": 2.641985829228366e-07, "loss": 0.0147, "step": 9690 }, { "grad_norm": 0.11346573382616043, "learning_rate": 2.474952811872877e-07, "loss": 0.0149, "step": 9700 }, { "grad_norm": 0.3119041323661804, "learning_rate": 2.3133606487228397e-07, "loss": 0.0148, "step": 9710 }, { "grad_norm": 0.17175710201263428, "learning_rate": 2.157211106924295e-07, "loss": 0.0161, "step": 9720 }, { "grad_norm": 0.18806205689907074, "learning_rate": 2.006505894103672e-07, "loss": 0.0156, "step": 9730 }, { "grad_norm": 0.2095193862915039, "learning_rate": 1.8612466583489696e-07, "loss": 0.0178, "step": 9740 }, { "grad_norm": 0.17248676717281342, "learning_rate": 1.7214349881918834e-07, "loss": 0.0173, "step": 9750 }, { "grad_norm": 0.33959922194480896, "learning_rate": 1.5870724125904845e-07, "loss": 0.0156, "step": 9760 }, { "grad_norm": 0.1554356813430786, "learning_rate": 1.4581604009124006e-07, "loss": 0.0131, "step": 9770 }, { "grad_norm": 0.13514424860477448, "learning_rate": 1.334700362918717e-07, "loss": 0.0161, "step": 9780 }, { "grad_norm": 0.16772912442684174, "learning_rate": 1.2166936487486015e-07, "loss": 0.0143, "step": 9790 }, { "grad_norm": 0.08267197757959366, "learning_rate": 1.1041415489045914e-07, "loss": 0.0137, "step": 9800 }, { "grad_norm": 0.19693174958229065, "learning_rate": 9.970452942384412e-08, "loss": 0.0205, "step": 9810 }, { "grad_norm": 0.175587996840477, "learning_rate": 8.954060559375754e-08, "loss": 0.0197, "step": 9820 }, { "grad_norm": 0.2839118242263794, "learning_rate": 7.99224945512489e-08, "loss": 0.0197, "step": 9830 }, { "grad_norm": 0.14380759000778198, "learning_rate": 7.085030147843675e-08, "loss": 0.0157, "step": 9840 }, { "grad_norm": 0.2006370723247528, "learning_rate": 6.232412558736523e-08, "loss": 0.0211, "step": 9850 }, { "grad_norm": 0.13094712793827057, "learning_rate": 5.434406011893822e-08, "loss": 0.012, "step": 9860 }, { "grad_norm": 0.320928156375885, "learning_rate": 4.6910192341864664e-08, "loss": 0.0164, "step": 9870 }, { "grad_norm": 0.12064913660287857, "learning_rate": 4.0022603551737035e-08, "loss": 0.0161, "step": 9880 }, { "grad_norm": 0.50347900390625, "learning_rate": 3.3681369070120985e-08, "loss": 0.0193, "step": 9890 }, { "grad_norm": 0.11336753517389297, "learning_rate": 2.7886558243744866e-08, "loss": 0.0169, "step": 9900 }, { "grad_norm": 0.1335654854774475, "learning_rate": 2.2638234443722596e-08, "loss": 0.0187, "step": 9910 }, { "grad_norm": 0.34617719054222107, "learning_rate": 1.7936455064887504e-08, "loss": 0.0208, "step": 9920 }, { "grad_norm": 0.11118677258491516, "learning_rate": 1.378127152514841e-08, "loss": 0.0155, "step": 9930 }, { "grad_norm": 0.16639424860477448, "learning_rate": 1.0172729264917857e-08, "loss": 0.0167, "step": 9940 }, { "grad_norm": 0.21694594621658325, "learning_rate": 7.1108677466458215e-09, "loss": 0.0147, "step": 9950 }, { "grad_norm": 0.12292720377445221, "learning_rate": 4.595720454353414e-09, "loss": 0.0131, "step": 9960 }, { "grad_norm": 0.14061234891414642, "learning_rate": 2.627314893294264e-09, "loss": 0.0226, "step": 9970 }, { "grad_norm": 0.10936050862073898, "learning_rate": 1.2056725896270048e-09, "loss": 0.0171, "step": 9980 }, { "grad_norm": 0.15498457849025726, "learning_rate": 3.308090902098826e-10, "loss": 0.0123, "step": 9990 }, { "grad_norm": 0.42714211344718933, "learning_rate": 2.7339624120159555e-12, "loss": 0.0168, "step": 10000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "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": 8, "trial_name": null, "trial_params": null }